Biochemical factors determining tumor response to 5,10-dideazatetrahydrofolate, a folate antimetabolite inhibitory to de novo purine biosynthesis

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, som e thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality o f th e copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bieedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send U M I a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. ProQuest Information and Learning 300 North Zeeb Road, Ann Arbor, M l 48106-1346 USA 800-521-0600 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. NOTE TO USERS Page(s) not included in the original manuscript are unavailable from the author or university. The manuscript was microfilmed as received. ix This reproduction is the best copy available UMI Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BIOCHEMICAL FACTORS DETERMINING TUMOR RESPONSE TO 5,10- DEDEAZATETRAHYDROFOLATE, A FOLATE ANTIMETABOLITE INHIBITORY TO D E N O V O PURINE BIOSYNTHESIS by ARCHIE TSE,MD A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (BIOCHEMISTRY AND MOLECULAR BIOLOGY) D ecem ber 2001 Copyright 2001 Archie Tse, MD Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3073860 Copyright 2001 by Tse, Archie Ngai-Chio All rights reserved. ___ ® UMI UMI Microform 3073860 Copyright 2003 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA The G raduate School U niversity Park LOS ANGELES, CALIFORNIA 90089-1695 This dissertation, w ritten b y 7 5 £ , MCfr/g UM'tfftQ Under the direction o f D issertation Committee, and approved b y a ll its members, has been presented to and accepted b y The Graduate School, in p a rtia l fulfillm ent o f requirements for the degree o f DOCTOR OF PHILOSOPHY Dean o f Graduate Studies D a te ___________ DISSER TA TION COMMITTEE ______ ^ ■ S c "-!”” Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lina Choueiri ABSTRACT ISSUES IN THE SYNTAX OF RESUMPTION: RESTRICTIVE RELATIVES IN LEBANESE ARABIC In this dissertation, 1 examine the mechanisms that underlie the formation of long A-dependencies in resumptive restrictive relatives (RRRs). I argue that such dependencies between an antecedent and the related resumptive site may be derived in one of two ways: either via Move or via base-generation. The former display the properties in (1), whereas the latter have the properties in (2). (1) a. Island sensitivity' b. Reconstruction effects c. Same range of interpretations for resumptive as for A-traces d. Resumptive element must be a (complex) agreement affix (2) a. No island sensitivity b. No reconstruction effects c. Fixed interpretation for resumptive: definite pronoun d. Resumptive element can be weak or strong Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The antecedent and the resumptive site form a chain, in RRRs derived by movement, but not in base-generated RRRs. The properties in (1) and (2) fall out naturally under this proposal. The behavior of Indefinite RRRs, which display only the properties in (2), further supports this dual analysis. Those constructions are shown to pose a problem for a raising-only analysis of relativization. The proposal makes correct predictions with respect to the interpretive properties of RRRs and cases of feature mismatch between the antecedent and the resumptive element. 1 provide a detailed proposal linking resumption, movement, and agreement morphology in Lebanese Arabic. Resumptives in the context of movement are the result of an agreement relation between a noun phrase and the strong head that selects it. Not only does the analysis account for the distribution of agreement morphology in Lebanese Arabic, it accounts for the difference between strong pronouns and clitics with respect to resumption. Generally, the dissertation provides answers to questions relating to the nature of resumptive elements and their role in the syntax of A-constructions by arguing that those elements do not form a natural class in triggering movement effects and that their morphological properties plav an important role in determining the availability of those effects. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGMENTS The author of this thesis is indebted to: Dr. R ich ard M oran, m y Ph.D. advisor, for his i n s p i r a t i o n , patience, an d intellectual guid an ce. His love for science is e n l i g h t e n i n g . Members of my Ph.D. com m ittee, Drs Robert Stellwagen, Peter Jones Robert Farley, and Pradip Roy-Burman, for their support a n d s u p e rv is io n . Drs. Shirley Taylor, Paresh Sanghani, Sonal Sanghani, Julie Kan, and Sarah Freemantle, for their advice and helpful discussions. Dr. Chuan Shih for his advice on the synthesis of t r i t iu m - labeled DDATHF and for his generous supply of DDATHF an alo g s, which were critical for the success of this work; Drs. I. D a v id Goldman and Kevin Brigle for their suggestions on folate transport. Ms. Valerie Evans for her superb assistance in the d e v e l o p m e n t of and in the initial characterization of the DDATHF-resistant cell lines; Ms. Rachel Cain, Jennifer Fergurson, Allison Null and Mr. V an Nguyen, for their technical asssitance. My parents for their encouragement. My wife, Rosemary, for her love that makes e v e ry th in g w o rth w h ile . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CONTENTS ACKNOW LEDGEM ENTS...................................................................................ii LIST OF FIGURES..............................................................................................viii LIST OF TABLES.................................................................................................. xi C H A P T E R 1................................................................................................................................... 1 B A C K G R O U N D IN F O R M A T IO N AND L IT E R A T U R E R E V IE W ANTIFOLATES AS ONCOLYTIC AGENTS AND BIOCHEMICAL PROBES - METHOTREXATE AS A PROTOTYPE............................................................................................ 1 B iochem istry and pharm acology o f M T X .............................................................................7 T herapeutic limitations of M T X ..............................................................................................17 Theoretical lim itations of M T X .............................................................................................. 24 ANTIFOLATES TARGETED AGAINST FOLATE-UTILIZING ENZYMES OTHER THAN D H F R ....................................................................................................................................................27 5 ,10-DIDEAZATETRAHYDROFOLATE: A POTENT INHIBITOR O F D E NO VO PURINE SYNTHESIS......................................................................................................................................... 30 Early studies of biochemical processes involved in the action of D D A TH F 3 1 C ellular T arget....................................................................................................................31 Stereochem istry of D D A TH F........................................................................................ 32 M em brane transport..........................................................................................................35 Polyglutam ation o f D D A TH F....................................................................................... 36 Effect o f DDATHF on nucleotide m etabolism ........................................................36 Cytotoxicity of D D ATHF................................................................................................37 Pre-clinical and clinical pharm acology of D D A T H F ......................................................38 First generation phase I clinical trials..........................................................................39 Second generation phase I clinical trails....................................................................40 MULTITARGETED ANTIFOLATE: ANOTHER NEW CLASS OF FOLATE ANTIMETABOLITES..........................................................................................................................4 1 SU M M A R Y O F D IS S E R T A T IO N ..................................................................................43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. i v C H A P T E R 2 ................................................................................................................................4 5 S T R U C T U R A L FE A T U R E S OF D D A TH F T H A T D E T E R M IN E IN H IB ITIO N O F M A M M A L IA N G L Y C IN A M ID E R IB O N U C L E O T ID E F O R M Y L T R A N S F E R A S I N T R O D U C T I O N ................................................................................................................... 45 M olecular genetics o f G A RFT................................................................................................ 46 Biochem ical characterization...................................................................................................48 Structural studies o f G A RFT....................................................................................................55 M echanism o f catalysis..............................................................................................................60 M A T E R IA L S AND M E T H O D S ..................................................................................... 6 3 C hem icals.......................................................................................................................................63 A nalog concentrations...............................................................................................................64 Cell C ulture....................................................................................................................................66 Protein assays................................................................................................................................ 67 Protein separation by SD S-PA G E.......................................................................................... 67 Enzym e p u rific a tio n ..................................................................................................................67 Enzym e assay s..............................................................................................................................70 Kinetic analysis............................................................................................................................ 71 R E S U L T S .................................................................................................................................... 74 Enzym e p u rificatio n ..................................................................................................................74 Steady state k in etics................................................................................................................... 76 Inhibition of G A R FT by DDATHF analogs....................................................................... 77 M odifications in the tetrahydropyridopyridine rin g .......................................................83 A lterations in the bridge and phenyl ring regions............................................................84 R equirem ent o f the glutam ic acid side chain for G A RFT inhibition......................... 86 D I S C U S S I O N ............................................................................................................................91 Structural determ inants of DDATHF for GARFT in h ib itio n ....................................... 91 Regions of D D A TH F that can tolerate substantial m odifications without com prom ising activity against G A R FT................................................................................ 96 Structural changes resulted in enhanced potency against G A RFT - the role of D D A TH F polyglutam ates in mediating cytotoxicity.....................................................102 Kinetics of the GARFT reaction revisited.......................................................................... 107 CHAPTER 3........................................................................................................... 109 BIOCHEMICAL CHARACTERIZATION OF MURINE LEUKEMIC L1210 CELLS SELECTED FOR RESISTANCE TO DDATHF INTRODUCTION.................................................................................................109 MATERIALS AND METHODS.......................................................................114 Chemicals...............................................................................................................................114 Synthesis of [3H](6R)-DDATHF..................................................................................... 115 Enzymatic synthesis of radiolabelled DDATHF tetraglutamate...............................117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cell Culture.............................................................................................................................118 Selection of DDATHF-resistant L1210 cells..................................................................119 Growth inhibition studies....................................................................................................120 Purification of GAFRT for kinetic analysis....................................................................120 Enzyme Assays......................................................................................................................121 GAFRT assay.................................................................................................................... 121 FPGS assay........................................................................................................................122 Conjugase assay................................................................................................................123 Purification of FPGS............................................................................................................123 Uptake of [14C]DDATHF.................................................................................................. 125 Transport Studies..................................................................................................................126 Analysis of Intracellular Accumulation of (6R)-DDATHF Polyglutamates by hplc...........................................................................................................................................127 Determination of Folate Cofactor Pools..........................................................................129 Northern analysis..................................................................................................................131 R E SU L T S............................................................................................................... 133 Development of L1210 sublines resistant to DDATHF..............................................133 Cross-resistance and collateral sensitivity of L1210/D3 cells to related antimetabolites....................................................................................................................... 138 Characteristics of glycinamide ribonucleotide formyltransferase (GARFT) expressed by L1210/D3 cells..............................................................................................139 FPGS and conjugase activities........................................................................................... 147 Uptake of [14C](6R)-DDATHF in wild-type and DDATHF-resistant L1210 cells...........................................................................................................................................154 Synthesis of high specific activity [^H](6R)-DDATHF..............................................156 Membrane transport of (6R)-DDATHF..........................................................................160 Polyglutamylation of (6R)-DDATHF in whole cells...................................................163 Folate cofactor pools in L1210 and L1210/D3 cells...................................................166 Effect of the expanded folate pools on the accumulation of intracellular (6R)- DDATHF polyglutamates in DDATHF-resistant cells................................................ 172 Polyglutamation of (6R)-DDATHF in cells grown in folinic acid medium 174 Folic acid transport in LI210 and L1210/D3 cells.......................................................174 D ISC U SSIO N ........................................................................................................ 178 C H A P T E R 4 .............................................................................................................................. 189 M O L E C U L A R G E N E T IC S O F D D A T H F -R E SIST A N T L1210 C E L L S I N T R O D U C T I O N ..................................................................................................................189 RFC-MEDIATED TRANSPORT.................................................................................................... 190 B iochem istry................................................................................................................................ 190 M olecular genetics.....................................................................................................................194 Biodistribution of R F C ............................................................................................................. 198 MEMBRANE FOLATE BINDING PROTEIN-MEDIATED TRANSPORT.............................. 199 B iochem istry................................................................................................................................ 199 M olecular genetics.................................................................................................................... 201 B iodistribution............................................................................................................................202 PHYSIOLOGICAL FUNCTIONS OF RFC AND M FBP............................................................ 203 EFFLUX SYSTEMS OF FOLATES...............................................................................................206 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. v i M A T E R IA L S AND M E T H O D S ................................................................................... 208 Chemicals...............................................................................................................................208 Cell culture.............................................................................................................................211 Transport studies...................................................................................................................211 Isolation of total RNA from cultured cells.....................................................................212 Northern blotting................................................................................................................. 214 Preparation of DNA probes................................................................................................215 DNA ligation reaction.........................................................................................................215 Purification of DNA fragments from agarose gels......................................................216 Amplification of RFC cDNA from wild-type and resistant L1210 cells by RT-PCR...................................................................................................................................217 Sequencing analysis.............................................................................................................219 Construction of expression vectors..................................................................................220 Transfection of L 1210 and MTXrA cells......................................................................221 Dilution cloning of L1210 transfectants........................................................................ 223 Genomic PCR........................................................................................................................223 R E S U L T S ..................................................................................................................................225 Substrate specificity of folate transport in wild-type and DDATHF-resistant L1210 cells....................................................................................................................... 225 Inhibition of folic acid transport in L1210/D3 cells by MTX and monensin nd northern blot analysis of expression of mFBPl and mFBP2 in L1210 and......... L1210/D3 cells.......................................................................................................................227 Amplification of RFC-1 cDNA from parental and DDATHF-resistant L1210 cells by RT-PCR......................................................... 232 Identification of mutations in RFC-1 cDNA derived from DDATHF-resistant L1210 sublines.....................................................................................................................234 Transfection of wild-type and mutant RFC-1 cDNA into parental and transport- defective L 1210 cells.......................................................................................................... 242 Expression of exogenous RFC in L1210 transfectants...............................................245 Sensitivity of L 1210 transfectants to (6R)-DDATHF.................................................252 Transport of folate compounds in L1210 transfectants............................................. 255 Expression of mutant RFC-1 in wild-type L1210 cells...............................................259 Correlation between folate pools, transport specificity and growth inhibition by (6R)-DDATHF...................................................................................................................... 262 Genotype of the L1210 and L1210/D3 cells.................................................................265 Acquisition of multi-step resistance to DDATHF........................................................ 270 D I S C U S S I O N ..........................................................................................................................2 7 3 CHAPTER 5 ............................................................................................................288 GENERAL DISCUSSION PRE-TARGET BIOCHEMICAL EVENTS.....................................................................................290 (1) M em brane Transport.........................................................................................................290 (2) Retention o f intracellular (6R)-D D A TH F polyglutam ates...................................297 (3) Effects o f total cellular folate p o o ls.............................................................................302 (4) Inhibition o f GARFT by (6R )-D D A TH F polyglutam ates.................................... 308 POST-TARGET BIOCHEMICAL EVENTS.................................................................................. 309 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. REFERENCES......................................................................................................318 APPENDIX A: ABBREVIATION USED.................................................................................351 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES Fig. 1.1 Folate metabolic pathways......................................................................................4 Fig. 1.2 De novo purine biosynthesis............................................................................... 34 Fig. 2.1 Kinetics of the GARFT reaction............................................................................. 54 Fig. 2.2 X-ray crystallographic structure of GARFT........................................................59 Fig. 2.3 SDS-PAGE of L1210 GARFT during purification..............................................75 Fig. 2.4 Substrate inhibition of GARFT by GAR............................................................. 78 Fig. 2.5 Stereo view showing the binding of folate inhibitor, 5-deazatetrahydrofolae and substrate GAR to the active site of GARFT......................................................95 Fig. 2.6 Folate compounds bind with different conformations to folate-dependent enzym es............................................................................................................................97 Fig. 3.1 Inhibition of the growth of L1210 cell sublines by (6R,S)-DDATHF 134 Fig. 3.2 Northern analysis of expression of GARFT and FPGS m essages in wild- type and DDATHF-resistant L1210 cells..........................................................................143 Fig. 3.3 Purification of GARFT from DDATHF-resistant L1210 cells.......................... 146 Fig. 3.4 Kinetics of FPGS partially purified from L1210 and L1210/D3 cells using aminopterin, (6S)-DDATHF, and (6R)-DDATHF as substrates.........................149 Fig. 3.5 Disparate substrate specificities of FPGS isolated from mouse leukemic L1210 cells and mouse liver........................................................................................152 Fig 3.6 Time course of uptake of [” '4C](6R)-DDATHF in wild-type and DDATHF- resistant L1210 cells..................................................................................................... 155 3 Fig. 3.7 Schematic of synthesis of high specific activity [ H](6R)-DDATHF............157 Fig. 3.8 Stability of [3H](6R)-DDATHF............................................................................ 159 Fig. 3.9 Transport of (6R)-DDATHF in L1210 and L1210/D3 c e lls...........................161 Fig. 3.10 Accumulation of (6R)-DDATHF polyglutamates in L1210 and L1210/D3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. NOTE TO USERS Page(s) not included in the original manuscript are unavailable from the author or university. The manuscript was microfilmed as received. ix This reproduction is the best copy available U M I" Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. X Fig. 4.12 Sensitivity of growth inhibition by (6R)-DDATHF in L1210 sublines is correlated with transport preference of folate substrates but not with total folate p o o ls.................................................................................................................... 267 Fig. 4.13 Sequencing of the RFC-1 genomic locus in L1210and L1210/D3 c e lls..................................................................................................................................269 Fig. 4.14 Transport of folic acid was increased in L1210/D0.5 cells but that of (6R)-DDATHF was unaffected...................................................................................272 Fig. 5.1 Energetics and kinetics of folate transport..........................................................294 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES Table 1.1 Structures of folates and analogs.......................................................................5 Table 2.1 Activity of DDATHF Analogs as Inhibitors of GARFT...............................80 Table 2.2 Inhibition of GARFT by (6R,S)-DDATHF Polyglutamates........................88 Table 2.3 Structures of DDATHF analogues as GARFT inhibitors............................. 98 Table 3.1 Sensitivity of wild-type and DDATHF-resistant L1210 cells to growth inhibition by inhibitors of folate-dependent enzym es......................................... 136 Table 3.2 Activity and Kinetics of GARFT, FPGS and conjugase in L1210 and L1210/D3 cells..............................................................................................................141 Table 3.3 Folate cofactor pools in wild-type and DDATHF-resistant L1210 cells.. 171 Table 4.1 Sequences of oligonucleotide primers used for PCR and sequencing analysis......................................................................................................................... 210 Table 4.2 Transport of folate derivatives into L1210 and L1210/D3 cells................226 Table 4.4 Summary of selection of L1210 transfectants...............................................244 Table 4.5 Expression levels of RFC-1 transcript in L1210 transfectants.................250 Table 4.6 Growth Inhibition of L1210 Cell Transfectants by (6R)-DDATHF.......... 254 Table 4.7 Folate Transport Characteristics in L1210 Transfectants.......................... 257 Table 4.8 Transfer of the Resistant Phenotype into Wild Type L1210 cells by Transfection........................................................................................................... 261 Table 4.9 Correlation between transport substrate specificities, total folate pools,.... and sensitivities to (6R)-DDATHF growth inhibition in L1210 transfectants... 263 Table 4.10 Phenotypes of RFC-mutatants.................................................................. 286 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 CHAPTER 1 BACKGROUND INFORMATION AND LITERATURE REVIEW A ntifolates as oncolytic agents and biochem ical probes - m ethotrexate as a prototype Folates are a class of cofactors that function as o n e -c a r b o n carriers in the biosynthesis of purines, thym idylate and am ino acids (Fig. 1.1). These products, in turn, are precursors of nucleic acids a n d proteins which are essential for the proliferation of both normal s te m cells and neoplastic cells. Thus, inhibitors of f o la t e - d e p e n d e n t enzymes have long been exploited in anti-cancer therapy. In fact, folate antagonists represent the first class of antim etabolites t h a t were used in the treatment of childhood leukemia with so m e measure of success (Farber et al., 1948). Since their clinical debut in 1948, num erous antifolates have been synthesized and studied as anti-tum or agents. Methotrexate (MTX), the 4-amino, N 10-m ethyl derivative of folic acid (Table 1.1), has been the most ex ten siv ely studied antifol, in terms of its clinical efficacy as well as of the basic biochemical steps involved in its cytotoxic action. Todate, in addition to its clinical use as an oncolytic agent, MTX is utilized as a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. first-line agent in the treatm ent of au to im m u n e diseases such rheumatoid arthritis and psoriasis (Jolivet et al., 1983). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 Fig. 1.1 Folate metabolic pathways. Uptake of 5-m ethyltetrahydrofolate, the predom inate form of fo late in the plasma, is mediated by the reduced folate carrier (RFC) a n d / o r the folate receptor (RF) (reaction 1). Once inside the cell, e n z y m a t ic interconversions of various folate cofactors are mediated by: (2) Methionine syntase or 5-m ethyltetrahydrofolate: h o m o c y s te in e methyltransferase, a folate- and m e th y l c o b a l a m i n - d e p e n d e n t enzyme that catalyzes the resynthesis of methionine fro m hom ocysteine. (3) The conversion of te tra h y d r o fo ly l m on o g lu tam ates to polyglutam ates is catalyzed by folylpoly-y- glutamate synthetase (FPGS). (4) The regeneration of tetrahydrofolates by dihydrofolate reductase (DHFR), the c e llu la r target of m ethotrexate MTX. (5) Thym idylate syntase (TS) c a ta ly z e s the reaction that results in the loss of reduced folates as dihydrofolates during thym idylate synthesis. This enzyme is inhibited by the nucleotide analog 5-fluorodeoxyuridylate (FdUMP), as well as by folate antagonists CB3717 and ZD-1694 (Table 1.1). (6) Serine hydroxymethyltransferase. (7) 5,10-methylenetetrahydrofolate dehydrogenase. (8) 5,10-methenyltetrahydrofolate cyclohydrolase. (9) 5,10- methylenetetrahydrofolate reductase. (10) Glycinam ide rib o n u c le o tid e formyltransferase (GARFT), the target enzyme of DDATHF. (11) Aminoimidazole carboxamide formyltransferase (AICARFT). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 3 } " »i u o £ ■ a c l < 3 00 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5-CH v H4 PtcGlu 5 Table 1.1 Structures of folates and analogs O COOH N H 10 V H Folic Acid HN N CH3 0 COOH N ' Methotrexate HN h2 n^ nA n O COOH 5,6,7,8-Tetrahydrofolic acid HN H N H 5,10-Dideazatetrahydrofolic acid (DDATHF) COOH COOH COOH COOH Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 Table 1.1 Structures of folates and analogs (cont'd) O COOH ‘ COOH NH; HN Edatrexate HN CH H2N CH 0 COOH “COOH 10-PropargyI-5,8-dideazafolic acid (CB3717) COOH COOH HN H D1694, Tomudex OCH. OCH N I HN Trimetrexate H sN ^ n^ ^ nh m t A o COOH COOH Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 B iochem istry and pharmacology of MTX. The effectiveness of MTX as an antineoplastic agent is r e la te d to a num ber of biochemical factors including mem brane tra n s p o rt, intracellular metabolism, and interaction with cellular targets. Because of the divalent anionic nature of MTX at physiological pH , membrane translocation of the molecule by passive diffusion alone is inefficient. Cellular influx of MTX is m ediated by specific m e m b r a n e transporters. The reduced folate carrier (RFC), a c a rr ie r - m e d ia te d system which also transports cellular folates, is the primary t r a n s p o r t system in m am m alian cells (Goldman et al., 1968). In addition to the RFC, some m am m alian cells have a second transport route for natural folates and MTX, the folate receptor (FR) (see " I n tr o d u c tio n " of Chapter 4 for a more detailed description of various folate transport systems). Like the naturally-occurring folate cofactors, MTX is extensively metabolized intracellularly into long c h a in polyglutam ates (Whitehead, 1977; Kisliuk, 1981; McGuire a n d Bertino, 1981; and McGuire and Coward, 1984). T h is polyglutamation process involves the successive addition of m u l t i p l e glutamate residues (up to 9) to the y-carboxyl of the folate m o le c u le and is catalyzed by the enzyme folylpoly-y-glutam ate s y n th e ta s e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 (FPGS) (E.C. 6.3.2.17) (McGuire et al., 1980; Moran 1983; a n d Cichowicz and Shane, 1987): Q r COOH nh2 , oer& r1 HN u m L I J CH3 J. 11 J c h 3 h2n ^ n h2n n n + + u r u v ju n O . COOH . MgATP MgADP + + L-glutamic acid p. M am m alian FPGS is present in both the cytoplasm and th e m itochondria (Garrow et al. 1992). Both species of enzyme are expressed from the same gene, and the two translated proteins differ only by the presence of a m itochondria leader sequence required for proper subcellular targeting (Freemantle et al., 1995; Taylor et al., 1995; Freemantle and Moran, 1997). The polyglutam ate d e riv a tiv e s of cellular folates or antifolates such as that of MTX are m e ta b o lic a lly trapped inside the cell presumably because of the a d d i t i o n a l negative charges imparted on the molecules; thus, rendering t h e m poor substrates for the efflux system (Moran et al., 1976; Jolivet et al., 1982; and Chabner et al., 1985). It has been dem onstrated t h a t poly g lu tam atio n of cellular folates result in cofactors that b i n d folate-dependent enzymes at a higher affinity than is seen with th e m o n o g lu tam ate forms (for reviews see McGuire and Bertino, 1981; Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 and McGuire and Coward, 1984). Retention of folate cofactors in t h e m itochondria is required to support the function of t h e m itochondria-specific serine-hydroxym ethyltransferase, a folate- dependent enzym e involves in glycine m etabolism (Stover et al, 1997). Until recently, little was known about the in tr a c e llu la r catabolism of the polyglutamates of the natural folates and that of MTX. Sirotnak and co-workers have proposed that, in S180 s a c r o m a cells, MTX polyglutam ates were actively transported into lysosom es where they were hydrolyzed by folylpolyglutam yl hydrolase (y- glutamyl hydrolase (EC 3.4.19.9)), and exited from the lysosome as m o n o g lu tam ates (Barrueco and Sirotnak, 1991). In H35 h e p a t o m a cells, both an intracellular and an excreted form of hydrolase h a v e been isolated, and both species were able to utilize M T X -d ig lu tam ate as a substrate. Since hydrolysis take place at the most proximal y- glutamyl bond, these enzymes were best classified as e n d o p e p ti d a s e s (O'Connor, 1991; and reviewed in Galivan et al., 1999). The biochem ical mode of action of M TX has been extensively studied but is still under dispute. Classically, the cytotoxic action of MTX and its polyglutam ates is thought to be primarily effected by inhibition of dihydrofolate reductase (DHFR) (E.C. 1.5.1.3) (Werkheiser, 1961, for review see Jackson and Grindey, 1984), t h e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 0 enzyme responsible for regeneration of tetrahydrofolates fro m dihydrofolates and, hence, m aintenance of intracellular r e d u c e d folate cofactor pools (Fig. 1.1). This was based on the e a rly observations that high level of dihydrofolate accu m u lated in cells treated with M TX (Nixon et al., 1976; White and Goldman, 1976), a n d the fact that MTX was a potent inhibitor of DHFR in cell-free ex tra cts (Zakrzewski and Nichol, 1958). The consequence of DHFR i n h ib itio n is a depletion of reduced folates and, presumably, an im p airm en t of de novo purine and thym idylate (dTMP) synthesis, which re q u ire reduced folate cofactors. This scheme, although reasonable, is n o t the only possible m echanism of action of MTX. Allegra et al. (1 9 8 7 ) have reported that the majority of cellular tetrahydrofolate co facto rs were preserved after exposing breast cancer MCF-7 cells to concentrations of MTX that totally suppressed reduced folate- dependent processes. It has also been shown that MTX polyglutamates, and dihydrofolate polyglutamates that a c c u m u l a t e d as a result of DHFR blockade could inhibit thym idylate (TS) a n d am in o im id azo le carboxamide ribonucleotide transform ylase (AICAR transformylase), the second folate-dependent enzyme in de n o v o purine synthesis, suggesting that MTX might have additional sites of action other than DHFR inhibition and reduced folate pool d e p l e t i o n (Allegra et al., 1985a; Allegra et al., 1985b; and Allegra et al., 1986). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11 Goldman and co-workers (1991) have suggested that the presence of a preserved tetrahydrofolate pool despite com plete inhibition of purine and pyrim idine biosynthesis after antifolate treatment was due to partitioning of cellular folates between a cytosolic and a m itochondrial com partm ent. Following treatm ent of L1210 cells with trimetrexate, a lipophilic inhibitor of DHFR, there was no c h a n g e in folate pools in the mitochondrial fraction, but a rapid co n v ersio n of tetrahydrofolate cofactors to dihydrofolate in the cytosolic fraction (45%), a level much greater than that observed in whole cell extract (25-30%). The effects of MTX on nucleotide m etabolism has been s tu d ie d at length. An almost invariable finding has been depletion of thymidine triphosphate (dTTP) (Tattersall et al., 1975; Skoog et al., 1976; Lowe and Grindey, 1976; and for review also see Jackson a n d Grindey, 1984). The next most frequent observation is a decrease in dGTP level in cells treated with MTX (Skoog et al., 1976; Taylor et al., 1981). Effects of MTX on dATP and dCTP pools have been variable and complex (Jackson and Grindey, 1984). Interpretation of drug- induced changes in deoxyribonucleotide and ribonucleotide pools has been com plicated by the fact that m etabolism of purine a n d pyrimidine is different in various phases of the cell cycle w hereas studies perform ed in unsynchronized cells reflects concentrations of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12 these DNA and RNA precursors in the entire cell population; selective changes in nucleotide pool size in S-phase cells, for example, may be greater than the measured average value (Hordern and H e n d erso n , 1981). The issue is further confounded by the fact that, within a subpopulation of cells in the same cell cycle phase, purine a n d pyrimidine nucleotide pools are probably not homogenous b u t compartmentalized "directly" in cellular organelles or "indirectly" as substrates channeled via m ultienzym e complexes (Moyer a n d Henderson, 1985). The precise biochemical events leading to cell death as a re s u lt of perturbation of deoxyribonucleotide and ribonucleotide pools are also poorly understood. An increase in deoxyuridylate (dUMP) a n d deoxyuridylate triphosphate (dUTP) has been reported in MTX- treated cells (Tattersall et al., 1973). It has been suggested that a n increase in dUTP/dTTP ratio could lead to m isincoporation of u ra cil into DNA. The removal of uracil by DNA-uracil glycosylase results i n reinsertion of dUMP in a futile cycle, causing DNA strand breaks a n d cytotoxicity (Goulian et al., 1980). However, this a ttr a c tiv e hypothesis has now been disputed by the following studies; (1) Lorico et al (1988) reported that MTX caused not only DNA single-strand, but double-strand breaks in NIH/3T3R murine fibroblasts. This is i n line with the observation that TS-negative murine FM3A cells Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 3 accu m u lated DNA double-strand breaks during thymineless d e a t h (Ayusawa et al., 1983). One would not expect double-strand break to accum ulate if DNA-uracil glycosylase were indeed responsible for th e fragm entation of DNA. (2) Although the folate-based TS in h i b it o r CB3717 (see below) caused build up of dUTP, its cytotoxicity was n o t potentiated by added deoxyuridine, nor antagonized by N- p h o sphonacetyl-L -aspartate, an inhibitor of dUTP synthesis (J a c k s o n et al., 1983). There has been increasing evidence that DNA fragm entation in cells treated with MTX represent a c h a ra c te ris tic feature of a general process induced by a wide variety of ch e m o th e rap e u tic agents known as apoptosis or program m ed cell death (reviewed in Eastman, 1993; Reed, 1995; Korsmeyer, 1999; Reed, 1999a; Reed, 1999b; Bossy-Wetzel and Green, 1999; Eastman a n d Rigas, 1999). (see also Chapter 5 "General Discussion") In addition to their pivotal role in com bating cancer, MTX a n d related antim etabolites has continued to be used as p o w e rfu l biochemical tools to unravel im portant principles in e n z y m o lo g y , nucleotide metabolism, cell death, and the genetic basis of d r u g re sistan c e. One im portant concept in enzymology that we have l e a r n e d from studies on the interaction between MTX and DHFR is t h e "stoichiom etric inhibition" phenom enon. Thus, initial k in etic Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 4 analysis on inhibition of DHFR by MTX and am inopterin p r o d u c e d apparent non-com petitive pattern of inhibition with respect to th e substrate folic acid as a substrate (Osborn et al., 1958). This was unexpected based on the close structural resemblance of th ese com p o u n d to folic acid (Table 1.1). This discrepancy was la te r resolved by Werkheiser (1961), who demonstrated that the kinetics of inhibition of DHFR by MTX was indeed competitive with folic a c id but gave an apparent non-com petitive pattern because of th e extremely tight binding of the drug to the enzyme. W erkheiser extended the treatm ent of tight binding inhibition by Strauss a n d Goldstein (1943) and coined the term "stoichiometric in h ib itio n " : when inhibition is caused by a co m p o u n d that binds e x tre m e ly tightly to the enzyme, at concentrations of inhibitor inadequate to cause com plete inhibition of enzyme activity, practically all inhibitors are enzyme-bound. Classical steady state analysis of inhibition kinetics assumes that enzym e-bound inhibitor is negligible com pared to total inhibitor concentration. Analysis based on th is assum ption, such as that by Lineweaver-Burke, is th e re fo re inappropriate for stoichiometric inhibitors. Another im portant principle illustrated by MTX is t h e biochem istry of tight-binding reversible inhibitors. Free i n tr a c e llu la r MTX does not readily efflux from the cell because after it disso ciates Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 5 from DHFR, M TX reassociates with the target enzyme much m o r e rapidly than translocating across the cell m em brane. However, in the presence of the activity of thym idylate syntase, DHFR i n h ib itio n results in a dram atic build-up of dihydrofolate in the form of polyglutam ates (Moran et al., 1975), which can c o m p e titiv e ly displace MTX from DHFR (White et al., 1975; White and G o l d m a n , 1976). M athem atical modeling studies have shown that it re q u ir e d at least 95% inhibition of DHFR activity in order to cause d e p le tio n of reduced folate pools and inhibition of cell growth (Jackson a n d Harrap, 1979). Moreover, a more recent study has suggested that, in cells treated with MTX, the level of DHFR might increase due to a post-transcriptional mechanism (Chu et al., 1993). The synthesis of DHFR was ordinarily repressed by the binding of the enzyme to DHFR mRNA; the interaction of MTX with DHFR prevented this re p re ssio n and, hence, directly stimulates DHFR synthesis (Chu et al., 1993). Therefore, an excess of unbound, or "free" intracellular MTX a b o v e that for stoichiometric titration of DHFR, is required to k e e p inhibition of the enzyme rate limiting for the cell cycle (White et al., 1975). The accum ulation of excess free drug is achieved by tw o processes. (1) A carrier-m ediated transport m echanism carried o u t by the RFC that allows accumulation of free intracellular MTX a b o v e the level that would otherwise expect from the existence of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16 electrochemical equilibrium with extracellular drug (G oldm an et al., 1968; and Goldman, 1969.) (2) Metabolism by FPGS into long c h a in polyglutam ates that allows cellular retention even in the absence of extracellular drug (reviewed in McGuire and Coward, 1984). Investigation of the biochemical mechanisms of resistance to MTX by tumors cells has revealed several fundam entally i m p o r t a n t processes in biology; one of which is the intriguing p h en o m en o n of gene amplification in eukaryotic cells (Schimke, 1992). In th e ir classic paper , Alt et al. (1978) dem onstrated that cultured cells selected for resistance to MTX after continuous exposure to step-w ise increasing concentrations of drug selectively amplified the DHFR gene and overexpressed DHFR enzyme activity. Since then, DHFR am plification after MTX treatm ent has been docum ented in m a n y other cultured cell system, and also in tumor samples from p a t i e n t s treated with MTX (Carman et al., 1984; Curt et al., 1983; Matherly et al., 1995). The configuration of the amplified DNA varies, with th e am plified DHFR genes located either within expanded c h r o m o s o m a l segments called homogeneously staining regions (HSRs) (Nunberg e t al., 1978 and Biedler et al., 1979) or in small e x t r a c h r o m o s o m a l fragments of DNA known as double minutes (Brown et al., 1981). Inasm uch as the im portance of oncogene amplification found in hum an cancers, research in gene amplification associated w ith Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 7 antifolate resistance has had profound implications in th e understanding of the process of carcinogenesis. Thus, an tifo la te s remain a class of interesting molecules from both the th e r a p e u t i c and fundamental research points of view. Therapeutic limitations of MTX Clinical response to MTX is highly tumor-dependent. The d r u g assumes a major cytoreductive and often curative role in th e treatment of acute lymphoblastic leukemia, n o n -H o d g k in 's lym phom a, osteosacrom a, and ch o riocarcinom a (Bertino et al., 1997). A second group of tumors which includes head and n ec k cancer, and breast carcinom a are responsive to initial treatment b u t soon become refractory to further therapy. Finally, MTX seems to have no efficacy in a third tumor group such as colorectal and n o n small cell lung carcinoma, which are also resistant to most o t h e r cytotoxic drugs (Bertino et al., 1997). The clinical usefulness of MTX is limited in part by the exhibition of either intrinsic or a c q u ir e d resistance by tumor cells (reviewed in Gorlick et al., 1996). Acquired resistance refers to the emergence of resistant clone(s) from a tum or population that was initially sensitive to d ru g treatment. Presumably, this represents the selection of a s u b Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 8 population of pre-existing resistant m utants in the presence of d ru g . Several m echanism s of acquired resistance to MTX have been established in cell culture models, which include: (1) Defective membrane transport of MTX is a fre q u e n t m echanism of drug resistance. Alterations in the RFC as a result of decreased expression of carrier protein, impaired translocation of ligand across the cell membrane, and changes in the kinetics of m em brane transport, i.e. an increased K m , and com bined Km a n d Vmax defects have all been described in the literature (Hill et al., 1979; Sirotnak et al., 1981; McCormick et al., 1981; N ietham m er a n d Jackson, 1975; Sirotnak, 1987; Schuetz et al., 1988). (2) Overexpression of DHFR by gene am plification in the fo rm of HSR's or double minutes results in an elevated level of DHFR transcripts and protein (Nunberg et al., 1978, Brown et al., 1981; a n d Biedler et al., 1979). (3) Altered inhibition kinetics of DHFR by MTX. This could be caused by a decrease in affinity for binding of MTX to the e n z y m e (Jackson et al., 1976) or by changes in the kinetic properties of th e enzyme that showed higher affinity for its substrates (Dedhar a n d Goldie, 1985). (4) Im paired accum ulation of MTX polyglutam ates. In m o s t clinical protocols, MTX is adm inistered to patients as r e p e a te d , Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 9 interm ittent doses rather than as continuous incubation as employed in traditional laboratory methods for selection of re s is ta n t variants. When resistant mutants were derived by subjecting wild- type CCRF-CEM leukemic cells to repeated pulsed (24 h) i n c u b a t i o n with MTX, sublines were found to be resistant to short-term (24 h), but not to continuous (120 h) exposure with MTX (Pizzorno et al., 1988). Further, defective accum ulation of MTX polyglutam ates as a result of decreased FPGS activity appeared to be the only m e c h a n i s m of antifolate resistance identified in these cells (Pizzorno et al., 1991). A decreased intracellular accum ulation of MTX polyglutam ates h a s also been reported to be associated with an increase in the activity of folylpolyglutamyl hydrolase (Rhee et al., 1993). M echanism s of acquired resistance to MTX described in cell culture systems seem to be operative also in vivo. C ry o p re s e rv e d leukemic blast cells obtained as bone marrow aspirates from p a t i e n t s with acute hem atological malignancies serve as a convenient so u rc e of tum or samples for studying acquired clinical resistance. T h e fluorescent MTX analogue, N<4-amino4-deoxy-N10-methypteroyl)-N5- (4'fluorescein thiocarbamyl)-Lrlysine (PT-430), binds quantitatively to cellular DHFR and, hence, allows m easu rem en t of DHFR content i n intact blast cells by flow cytometry. Further, comparing the a m o u n t of competitive displacement of PT-430 from fluorochrome p r e - lo a d e d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20 blasts by added MTX with that obtained by added trimetrexate ( a lipophilic DHFR inhibitor that enters cell by diffusion) provides a n indirect estimate of MTX uptake. Using this m ethodology, T r i p p e t t et al. (1992) have reported that defective transport of drug was found in blastic cells from 14 of the 27 patients with relapses of acute lym phoblastic leukemia (ALL). To date, amplification of th e DHFR gene appears to be the second most com m on m echanism of acquired clinical resistance to MTX. Using a DNA dot blot assay, DHFR am plification was detected in 9 of 29 patients (31%) w h o relapsed from ALL, whereas only 4 of 38 patients (11%) with new ly diagnosed disease showed detectable gene amplification (Goker et al., 1995). Efforts have been made to circum vent these m echanism s of drug resistance. One approach is by using lipophilic analogs of MTX lacking the glutam ate side chain which do not depend on RFC or FPGS for activity, e.g. trimetrexate, (Table 1.1) (Lin and Bertino, 1991). Another approach is to develop new antifols that ta rg e t folate-dependent enzymes other than DHFR (see below). Com pared with the chemosensitive cancers, such as c h i l d h o o d acute lym phoblastic leukemia (ALL), some tumor types simply d o not respond to MTX even during the first exposure to the drug a n d are considered intrinsically resistant (Pizzorno et al., 1989; Lin et al., 1991; Li et al., 1993). The extent of resistance is usually less d ra stic Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 1 than that found in cell culture models of acquired resistance w here the degree of resistance is accentuated by the purposeful a p p lic a tio n of a strong selective pressure (exposure to escalating levels of drug u p to very high concentrations). However, even a modest level of resistance (3 to 4-fo!d) is of clinical relevance given the n a rro w therapeutic window of most anti-cancer agents. The basis of in trin sic resistance has been studied in neoplastic diseases that are c o n s id e re d to be refractory or only partially responsive to MTX, e.g. a c u te nonlym phoblastic leukemia (ANLL) (Lin et al., 1991), squamous cell carcinoma of the head and neck (Pizzorno et al., 1989) and soft tissue sacroma (Li et al., 1993). When com pared with tumor cells isolated from patients with childhood ALL, a chemosensitive disease, leukemic blasts derived from most patients with ANLL were found to accumulate a lower level of MTX long chain polyglutam ates (Lin et al., 1991). This lack of p o ly glutam ation may be due to a d e c re a s e d level of FPGS activity in tumor cells refractory to antifolates. B arredo et al. (1994) have reported that levels of FPGS activity in ALL blasts (689 pm ol/h/m g protein) were significantly higher than those f o u n d in AM L blasts (181 pmol/h/mg protein). Conversely, an increased in the catabolism of MTX polyglutamates can be responsible for the low level of long chain metabolites in ANLL. This possibility is suggested by W hitehead and co-workers (1988), who found longer chain MTX Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22 polyglutam ates in ANLL blasts following incubation of cells with a y- glutamyl hydrolase inhibitor. In line with this observation, stu d ies on human sacrom a cell lines naturally resistant to MTX h a v e revealed an elevated activity of y-glutamyl hydrolase and a n intrinsically lower expression of FPGS (Li et al., 1993). In a n o t h e r study, Pizzorno et al. (1989) had reported that, among th ree squamous carcinom a cell lines of the head and neck, two were inherently resistance to short-term exposure to M TX (4 and 24 h). In contrast, all three cell lines were equally sensitive to trim e tre x a te (Table 1.1), a lipophilic antifolate that does not utilize the RFC for cell entry and is not polyglutam ated. The only alteration id en tified in the resistant lines was a decreased accum ulation of MTX polyglutam ates. Taken together, it appears that w hereas overexpression of DHFR and defective transport are freq u en t m echanism s of acquired resistance to MTX, i m p a i r e d polyglutam ation may be an important m echanism of intrinsic resistance to this drug. The ability of blast cells from patients with ALL to a c c u m u la te MTX polyglutam ates is of prognostic im portance. Thus, W h ite h e a d et al. (1992) have reported that a longer event-free survival is f o u n d in pediatric patients with ALL whose blasts at diagnosis re ta in e d higher levels of MTX polyglutamates. In addition, a higher FPGS Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23 activity and a higher percentage of long chain MTX metabolites are found in B-lineage ALL than in T-lineage ALL (Barredo et al., 1994). This is in good correlation with the better prognosis associated w ith the former condition treated with MTX (Goker et al., 1993). Therefore, there has been compelling evidence that f o r m a ti o n of MTX poly g lu tam ates is a critical determ inant of the clinical usefulness of this agent. Given the fact that MTX is a poor s u b s tra te for FPGS relative to the physiological folates (Moran, 1983), developm ent of new folate antagonists that are more efficiently polyglutamated may represent a valuable area of research. The n ew DHFR inhibitor, edatrexate (Table 1.1) was selected for clinical development, in part, on this basis and because of the fact that it is a better substrate for the RFC than MTX (Schmid et al., 1985; S iro tn a k et al., 1993; Schornagel et al., 1995). Bertino and co-workers have reported a rather in trig u in g mechanism of intrinsic resistance to MTX in tumor cells which la c k e d the retinoblastom a (RB) tumor suppressor. Cells defective in pRB function was found to contain elevated levels of DHFR and were relatively insensitive to MTX. These investigators proposed that cells lacking functional pRB would contain an increased level of E2F proteins, a transcription factor that is otherwise inactivated by b i n d to pRB, and have enhanced transcription of E2F target genes such as DHFR (Li et al., 1995). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24 Theoretical limitations of MTX Despite extensive investigation of the biochemical steps involved in the cytotoxic action of MTX, the precise molecular e v e n ts leading to cell death after DHFR inhibition remain hazy. Classically, it has been proposed that MTX depletes cellular reduced folate p o o ls and subsequently blocks the production of both thym idylate a n d purines. However, the relative contribution of M T X -in d u c e d an tithym idylate and antipurine effect to cytotoxicity r e m a i n s unclear. Studies on reversal of MTX effects by addition of e n d - products had yielded conflicting results. The antiproliferative effect of MTX could be partially eliminated by thymidine in some cell lines (Borsa and W hitmore, 1969; Rueckert and Mueller, 1960; ) but by purines in others (Hryniuk, 1972). Yet a universal finding was t h a t both thym idine and purines were necessary for complete p r o t e c t io n from the lethal effect of the drug (Hakala and Taylor, 1959; Borsa and Whitmore, 1969). Moran et al. (1979) has proposed that th is apparent cell line dependence of protection of MTX cytotoxicity by end-products appears to be related to the activity of t h y m i d y l a t e syntase (TS) (E.C. 2.1.1.45), the enzyme that catalyzes the re d u c tiv e methylation of deoxyuridylate to thym idylate using 5,10- methylenetetrahydrofolic acid as a source of methyl group (Fig. 1.1): Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25 The sole function of DHFR is to regenerate bioactive reduced folate from dihydrofolate produced by TS from 5,10- m eth y len etetrah y d ro fo la te during thym idylate synthesis (Fig. 1.1). In the absence of TS activity, even complete inhibition of DHFR is probably irrelevant to overall folate metabolism inside the cell because all cofactors will remain in the tetrahydrofolate forms. T h e rate of thym idylate synthesis appears to be controlled primarily by the availability of the substrate deoxyuridylate (dUMP) w hose production, in turn, is carried out by two enzymes, rib o n u c le o tid e reductase and deoxycytidylate deaminase. In eukaryotic cells, t h e activity of these two enzymes is regulated by feed-back inhibition by cellular thym idylate triphosphate (TTP). Hence, an expanded TTP pool furnished by an exogenous supply of thym idine will reduce th e level of dUMP, which will, in turn, decrease TS activity, a n d subsequently, the consu m p tio n of reduced folates. These c o m p le x interactions of metabolic pathways serve to explain the p a r ti a l protection of MTX cytotoxicity by thym idine which, in addition to alleviating the an tith y m id y late effect of the MTX, protects a g a in s t the antipurine effect of the drug by turning off TS. Moreover, it h a s been reported that, in the presence of a source of p re -fo r m e d thymine, the cytotoxicity of MTX can virtually be eliminated by deleting TS either genetically (Ayusawa et al., 1981) o r Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26 pharm acologically using 5-fluorodeoxyuridine (FUdR) (Moran et al., 1979). It has been postulated that the cytotoxicity of MTX is self- limiting as a result of simultaneous depletion of both t h y m i d y l a t e and purine. Several investigators have d em o n strated that the e x te n t of cell kill induced by MTX can be enhanced by an exogenous s u p p l y of purine, suggesting that the anti-purine effect of MTX might re d u c e the overall cytotoxic effect of the drug (Borsa and Whitmore, 1969; Taylor et al., 1982). One possible explanation is that p u r i n e depletion results in inhibition of RNA synthesis, causing cells to a rre st in G1 and, hence, protecting them from the phase-specific cy to to x ic action of MTX in the S-phase. Addition of hypoxanthine ab ro g ates the inhibitory effect of MTX on RNA synthesis and allows cell cycle progression (Fairchild et al., 1988). Another explanation is t h a t purine depletion result in an attenuated form of apoptosis (see Chapter 5 “General Discussion). Although the cytotoxic effects of MTX can be accounted for b y its interaction with DHFR, it has been suggested the drug might h a v e additional sites of action. Thus, it has been shown that t h e polyglutam ates of MTX were also inhibitors of thym idylate sy n ta se (Allegra et al., 1985a) and am in o im id azo le c a rb o x a m id e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 27 ribonucleotide transformylase, the second folate-dependent e n z y m e in de novo purine synthesis (Allegra et al., 1985b). Further, it h as been dem onstrated that dihydrofolate polyglutam ates, w h ich accum ulated as a result of the DHFR blockade, could also in h ib it these two enzymes (Allegra et al., 1985b). Thus, the multiple d irect and indirect effects of MTX on biosynthesis of DNA precursors m a k e s it difficult, if not impossible, to delineate the relative contribution of inhibition of each pathway to cell death induced by the drug. Antifolates targeted against folate-utilizing enzymes other than DHFR In view of the theoretical and therapeutic limitations of MTX and related DHFR inhibitors, attempts have been made to d e v e lo p novel classes of antifolates targeted against other f o la te - d e p e n d e n t enzymes, with emphasis on the developm ent of better substrates for m am m alian FPGS and ideally pure inhibitors of either t h y m i d y l a t e or purine synthesis. In early 1980's, Jones and his colleagues evaluated a series of quinazoline (5,8 -dideazafolic acid) analogs of folic acid as inhibitors of TS. Thym idylate syntase is not a n ew target for antim etabolic action as it is a major locus of the a n t i tumor action of the fluorinated pyrim idines (5-fluorouracil and 5- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28 fluoro-2'-deoxyuridine (FUdR)), where the metabolite FUdR monophosphate (FdUM P) forms a covalent ternary complex with t h e enzyme and the folate cofactor (Langenbach et al., 1972; Santi a n d McHenry, 1972; and Danenberg, 1977). However, folate-based inhibitors of TS offer a theoretical advantage over t h e fluoropyrimidines: the accumulation of deoxyuridylate as a result of TS blockade can overcome the inhibition of enzyme by FdUMP, b u t will not com pete with a folate antagonist. This is because substantial build up of 5 , 10-m ethylenetetrahydrofolic acid does n o t occur due to enzymatic conversion to other cofactor forms and to the intrinsic limitation of the size of the total folate pools (Fig. 1.1). In addition, unlike the fluoropyrimidines, folate-based TS in h ib ito rs do not incorporate into nucleic acids and thus they a p p r o x im a te pure inhibitors of thymidylate synthesis. Of the series of N'° substituted 5,8-dideaza analogs of folic acid studied, the N 10- propargyl com pound, or CB3717 (Table 1.1), was chosen for therapeutic developm ent. Although CB3717 showed so m e encouraging anti-tum or activity in vitro and in vivo , the clinical developm ent of this com pound was d iscontinued because of th e unacceptable nephrotoxicity reported in early clinical trials (C alvert et al., 1986; Sessa et al., 1988; and, for review see Clarke, 1993). Nevertheless, the introduction of CB3717 has established th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29 precedent that antifolates inhibitory to TS were viable alternatives to DHFR inhibitors for cancer chem otherapy. Continued o rg a n ic synthesis and structural analysis soon led to the discovery of a folate-based TS inhibitor, ZD-1694 (tomudex, Ralitrexed) (Table 1.1) that is not nephrotoxic in patients. Moreover, this drug also s h o w e d improved anti-tum or activity com pared to CB 3717 in pre-clin ical studies, presum ably owing to its better utilization of the RFC for cellular entry and its higher avidity for FPGS (Jackman et al., 1991a; Jackman et al., 1991b). Results of phase III clinical trials in E u ro p e has shown that Tomudex has activity against colorectal c a r c i n o m a comparable to that found with 5-FU, the standard c h e m o t h e r a p e u t i c agent used for this class of human cancers for the past thirty y ears (Jackman et al., 1995). The drug was recently licensed for clin ical use in Europe as an antineoplastic agent against colorectal cancer. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 5,10-dideazatetrahydrofolate: a potent inhibitor of de n o v o purine synthesis Another major advance in the search for novel classes of antifols was the synthesis of the tetrahydrofolate analog 5,10- dideaztetrahydrofolate (Table 1.1) by Taylor and his colleagues (1984). The replacement of nitrogen by carbon atoms at positions 5 and 10 renders this com pound incapable of participating in a n y normal biochemical reactions that utilize folate as o n e-carb o n carrier. The C/N substitution at position 5 imparts ex c ep tio n al chemical stability to this com pound. The absence of 4-amino g ro u p in the pyrimidine ring suggested that DDATHF would not be a n inhibitor of DHFR, which was proven to be the case (Taylor et al., 1984). This co m p o u n d was originally synthesized as an inhibitor of TS. but was found to be virtually inactive against this e n z y m e (Taylor et al., 1984). However, DDATHF dem o n strated potent a n t i proliferative effects on tumor cells in culture (Taylor et al., 1984), suggesting another folate-dependent process was the target of th is agent. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 1 Early studies of biochemical processes involved in the action of DDATHF Cellular Target. The enzymatic target of DDATHF is now established to be glycinamide ribonucleotide formyltransferase (GARFT), the first folate dependent enzyme in de novo purine synthesis (Fig. 1.2), based on the following biochemical evidence (Beardsley et al., 1989; Moran et al., 1989): 1) The growth inhibitory of DDATHF on mouse leukemic cells can be prevented by a source of performed purine, e.g. h y p o x a n t h i n e , but not by thym idine. This result suggested that either one or b o th of the folate-dependent enzymes in de novo purine synthesis, GARFT and a m in o im id a zo leca rb o x am id e ribonucleotide fo rm y ltra n sfe ra se (A1CAFT) would be the target(s) of this agent. 2) A m inoim idazolecarboxam ide, the purine precursor t h a t enters the purine synthesis pathway before AICAFT, protected cells from DDATHF, indicating that the enzymatic inhibition takes p la c e in the biosynthesis pathway prior to AICAFT, presumably in the ste p catalyzed by GARFT. 3) A more definitive proof of this point was the use of a technique previously employed by Divekar and Hakala (1975) for Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32 studying the flow of substrates through the early steps in de novo purine synthesis. l^c.Q ly cin e is utilized by the enzyme i m m e d i a t e l y upstream to GARFT to label the early interm ediates in the p a t h w a y . In the presence of the glutamine antagonist azaserine (Fig. 1.2), th e enzyme im mediately after GARFT is blocked and the i n c o r p o r a t e d l^C-glycine will build up as form ylglycinam ide ribonucleotide. In cells treated with DDATHF, this accum ulation was specifically elim inated, demonstrating that GARFT was the target for DDATHF i n whole cells. 4) Using purified GARFT from mouse L1210 cells, DDATHF was shown to be a competitive inhibitor of the enzyme (initial e s tim a te of K| = 100 nM), constituting direct evidence for the in t e r a c t io n between D DA THF and GARFT (Moran et al., 1989). Stereochemistry of DDATHF. DDATHF was synthesized as a mixture of two d ia s te re o m e rs , differing in chiralty at the 6-carbon. The 6R- isomer has an a b s o lu te configuration the same as that of the n a tu r a lly - o c c u r r in g tetrahydrofolate derivatives. Surprisingly, both d ia s te r e o m e rs showed remarkably similar activities as cytotoxic agents, inhibitors of GARFT, substrate for FPGS (Moran et al., 1989) and RFC (Pizzorno et Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 33 al., 1993). The lack of stereospecificity of interaction of the tw o isomers with GARFT and with FPGS can be rationalized based u p o n molecular modeling studies on the lowest energy conformations of the isomers and structural knowledge about the two enzymes (see Discussion of Chapter 2 for more details). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 34 Lometrexol g lu ta m in e g lu ta m a te H jO jP O C H , 10 -F o n tiy I - H i P te G I u„ PRPP H ,C H jP te G lu ,, " X T ATP ADP i j / U f I GARFT f j: fiboxe-5'-phu.»phale g ly c in e o ^ ^ q h g ly c in a m id c r ib o n u c le o tid e a s p a r ta te rih « » e-5 '-p h o sp h a ic .V -fo rm y f-g ly c in a m id e rib o n u c le o tid e A T P . h:n n h n b o > e o '-p h o > irih ate lu m a ra te H jN '•n h ADP + Pi " g lu ta m in e — — A zaserine g lu ta m a te “ T V - ribu»c-V -phosphntc A D P + Pi ATP ^N H I riho»e-5‘-phosphalc - f sc H N ^N H nbiK c-5'-phti> phuic lO -P o m iv l-II.P ic G lu ,, H 4P ieG lu„ A m m o n n id a z o le c a rb o x a m id e r i b o n u c le o tid e ■ H’ O | x!x> In o sin ic " cm a c id ribose-.S pbiispli.ilc H jN ' ^ O H C , AICARFT - | f ilH » c - 5 '- p h o s p h m r rih o sc -5 -p tu t-.p h aic N -F o rm y l-am i n o im id a z o le c a rb o x a m id e rib o n u c le o tid e Fig. 1.2 De novo purine biosynthesis. The two folate-dependent reactions, catalyzed by GARFT a n d AICARFT, are denoted by open arrows. The principle site of inhibition of this pathway by azaserine is shown, as are the sites of entry of l4C-glycine and am inoim idazole carboxam ide (AICA) in to the pathway. Lometrexol, (6R)-DDATHF; and PRPP, p h o s p h o rib o sy l p y r o p h o s p h a t e . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35 M em brane transport. Several lines of evidence suggested that DDATHF gained e n t r y into m am m alian cells primarily by the same reduced folate c a rrie r system utilized by MTX and physiological folates. First, the s a m e number of specific surface binding sites was found for b o th diastereomers of DDATHF and MTX in CCRF-CEM cells (Pizzorno et al., 1993). Second, the influx of both [3H](6R)-DDATHF and PH ]M TX were m arkedly inhibited by covalent inhibitors of the RFC, N- hydroxysuccinim ide esters of (6R)-DDATHF and MTX (Pizzorno et al., 1993). Third, CCRF-CEM cells selected for resistance to MTX by v irtu e of impaired membrane folate transport showed dim inished t r a n s p o r t of PH]DD ATHF and cross-resistance to this agent (Pizzorno et al., 1993). Fourth, transfection of the RFC-1 cDNA into t r a n s p o r t- defective L1210 cells restored the transport of DDATHF (Ch. 4 and Tse et al., 1998). In addition, DDATHF appeared to be utilized also by a n alternate class of folate transporters, the m em brane folate b in d in g proteins or folate receptors (Pizzorno et al., 1993), which d i s p l a y e d different substrate specificity, tissue distribution and m echanism of internalization from the RFC (reviewed in Chapter 4 of this thesis). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36 Polyglutamation of DDATHF. Much of the enthusiasm aroused during early developm ent of DDATHF stem m ed from the fact that the co m p o u n d was a s u p e rb substrate for m am m alian FPGS. In fact, DDATHF was the first folate analog found to be as efficient a substrate for the enzyme as the m o s t favorable physiological folate, with a Vm ax/K m value 30 times t h a t of MTX. Both diastereomers are utilized similarly by mouse liver FPGS and human CCRF-CEM cells except for some subtle difference i n kinetics; the (6S)-isomer displayed substrate inhibition at h ig h concentration of drug. (6R)-DDATHF (lometrexol), the isomer t h a t did not show this behavior was chosen for clinical trial solely b a s e d on this observation (Moran et al., 1989). Effect of D D A T H F on nucleotide metabolism. The effects of DDATHF on ribonucleotide a n d deoxyribonucleotide metabolism have been studied in CCRF-CEM human T -lym phoblastic leukemia cells (Pizzorno et al., 1991). After a 4-h incubation with Im M (6R)-DDATHF, marked reductions in ATP and GTP pools were observed, with little effects on CTP, UTP, dATP and dGTP pools. Changes in nucleotide pools were associated with a dramatic decrease in incoporation of [3H ]thym idine a n d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37 [3H]deoxyuridine into DNA and RNA, respectively. Selective b lo c k a d e of GTP synthesis in the presence of (6R)-DDATHF, hypoxanthine, a n d m ycophenolate, an inhibitor of inosinate dehydrogenase, s u p p re s s e d incorporation of precursors into nucleic acid to the same extent, suggesting that diminished m acrom olecule synthesis was re la te d primarily to GTP rather than ATP depletion. Cytotoxicity of DDATHF. In their classic studies, Borsa and Whitmore (1969) h a v e reported that cell kill induced by MTX in L-cells could be p o t e n t i a t e d by an exogenous supply of hypoxanthine, suggesting that the overall cytotoxicity of MTX was limited by the antipurine effect of the d ru g . Thus, it is of theoretical and therapeutic interest to d e t e r m i n e whether a pure de novo purine synthesis inhibitor such as DDATHF is cytocidal, or is merely cytostatic. Our laboratory has studied th e cytotoxcity of (6R)-DDATHF and of ZD-1694 in WiDr hum an colonic carcinom a cells (Smith et al., 1993). Using carefully d e s ig n e d clonogenic assays, Smith et al. (1993) have dem o n strated t h a t DDATHF is indeed cytotoxic. However, a number of in te re stin g findings were noted regarding cell kill induced by this a n tifo la te . First, a maximal cell kill of 5-6 logs was achieved with Z D -1694, consistent with expected frequency of preexisting drug re s is ta n t Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38 mutant; whereas a maximal of only 2-3 logs was observed with (6R)- DDATHF. Second, morphological changes in cells treated by the tw o drugs were remarkably different. Cells treated with Z D -1694 underwent megalocytic changes and became detached from t h e tissue culture dishes after 1-2 days. On the other hand, cells e x p o se d to (6R)-DDATHF remained adherent to the dishes for >10 days a fte r treatment. These observations suggested that the cytotoxicity of inhibition of purine and thym idylate synthesis was mediated by distinct biochem ical events. Pre-clinical and clinical pharmacology of DDATHF DDATHF was found to be extremely active against tu m o r cell lines in culture including a M TX-resistant subline of CCRF- CEM cells that overexpressed DHFR. (Beardsley et al., 1986 a n d Beardsley et al., 1989). When screened against a variety of murine transplantable tumors, the drug was found to b e remarkably active even in tumor types where MTX showed little or no activity (Beardsley et al., 1986). In addition, (6R)-DDATHF also showed impressive anti-tum or activities against a b ro a d spectrum of human tumor xenografts in nude mice, and again Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39 was effective in tumor models that were refractory to MTX (Shih et al, 1988). First generation phase 1 clinical trials. Based on its broad spectrum of in vitro anti-tum or a c tiv ity and on its unique m echanism of action, (6R)-DDATHF (lo m etre x o l) was advanced to clinical trials in the U.S and in Europe. The results of early phase I trials of lometrexol were rather surprising (Nelson e t al., 1990; Young et al., 1990; and Ray et al., 1993). Whereas objective tumor responses were reported in several trials, the pattern of d ru g toxicity was found to be difficult to m anage and was not re a d ily explained by previous preclinical animal studies. Man is by far t h e most sensitive species to this drug with major toxicity seen at doses as low as 6 m g / m 2 on a weekly x 3 schedule or 4 m g / m 2 for 3 consecutive days every 4 week schedule (Nelson et al., 1990; Young e t al., 1990; Ray et al., 1993; and Sessa et al., 1996). M y elo su p p re ssio n was the dose-lim iting toxicity with th ro m b o c y to p e n ia and a n e m i a predom inating. In contrast, although th ro m b o cy to p en ia has been reported in patients treated with MTX, it was always preceded by leukopenia and was never the dose-lim iting toxicity by itself (Ensminger, 1984). Moreover, unlike treatm ent with MTX w h ere toxicity occurs early and shows no cumulative effects, treatment w ith Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40 lometrexol showed delayed toxicity which increased in severity w ith repeated dosing (Nelson et al., 1990; and Ray et al., 1993). Second generation phase I clinical trails. The unusual toxicological profile of lometrexol has made it difficult to devise a feasible dosing schedule for phase II clinical trial of this drug. Several years ago, the late Dr. Gerald Grindey and his then co-workers at the Lilly Research Laboratories (Indianapolis, IN) had noted that the co-adm inistration of folic acid with DDATHF to animals protected against the toxicity of this drug w i t h o u t com prom ising its anti-tum or activity (Alati et al., 1992; Grindey e t al., 1991). The complete description of this work has only re c e n tly appeared due largely to Dr Grindey's death (Alati et al., 1996). These results had prom pted the initiation of second g e n e r a tio n phase I trials which now appear to dem onstrate that oral folic a c id supplement would improve clinical tolerance of lometrexol (Young e t al., 1992; Wedge et al., 1995; and Roberts et al., 2000). Another P hase I trial using folinic acid rescue given on day 7 and 9 after th e adm inistration of a single dose of lometrexol has allowed e s c a la tio n of the dose up to 6 0 m g / m 2 with m anageable toxicities (Sessa et al., 1996). The advantages of using folinic acid rescue over pre- a n d post-treatment with folic acid, however, have not been defined. T h e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 1 interplay between folate supplem entation, lometrexol toxicity a n d anti-tum or activity constitutes the focus of discussion in this thesis (see "Discussion" of Chapter 3 and “General Discussion” Chapter 5). Multitargeted antifolate: another new class of folate a n t i m e t a b o 1 i t e s M ultitargeted antifolate (MTA) is a p y rro lo p y rim id in e a n a lo g of folic acid that was originally designed to be an inhibitor of GARFT (Table 1.1). Initial pre-clinical studies suggested that it was essentially a TS inhibitor. However, further testing showed that MTA was an excellent substrate for FPGS and the p o l y g l u ta m a t e s derivatives of this co m p o u n d also inhibited other f o l a te - d e p e n d e n t enzymes: the Kj values of the pentaglutam ate of MTA for TS, DHFR, and GARFT are 1.3, 7.1, and 65 nM, respectively. Reversal of cytotoxicity using end-products revealed that at low doses of MTA (5- 10 pM), inhibition of cell growth could be protected by the a d d i t i o n of thymidine alone, suggesting that TS was the major target; w hereas at higher doses (50 -100 pM), a purine source was required for reversal, indicating involvement of inhibition of DHFR and/or GARFT (Chen et al., 1999). The existence of multiple enzyme targets m a y offer advantages com pared with those antifolates inhibiting only o n e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 42 because the developm ent of drug resistance by virtue of ta rg e t enzyme overexpression would be less likely. It is not clear w h e th e r the prom ising therapeutic value of MTA constitutes a clinical advance or represent a reversed step from mechanistic based d r u g design back to empiricism (Rinaldi et al, 1999; O'Dwyer et al., 1999; and Calvert, 1999). Nevertheless, knowledge obtained f ro m pharmacological and biochemical studies using selective t h y m i d y l a t e and de n o v o purine biosynthesis inhibitors such as ZD-1694 a n d DDATHF, respectively, has undoubtedly facilitated the r a p i d advancement in the clinical use of MTA. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 43 SUMMARY OF DISSERTATION In sum m ary, 5,10-d id eazatetrah y d ro fo late represents the first clinically investigated antifolate whose anti-tum or effect was r e la t e d to inhibition of de n o v o purine synthesis. A more s o p h is tic a te d understanding of the molecular d eterm inants of tumor response t o DDATHF would undoubtedly go far toward successful application of this class of agents to cancer treatment. The overall objective of th is dissertation was to define the biochemical factors that d e t e r m i n e whether a tumor cell would be sensitive or resistant to DDATHF. T w o approaches were undertaken: (1) Using a series of DDATHF a n a lo g s as inhibitors of GARFT, the topology of the active site of this e n z y m e was probed and the structural features of DDATHF that were necessary for GARFT inhibition were identified. (2) A n o th e r approach is to study the molecular m echanism s by which t u m o r cells become resistant to DDATHF. Variants of mouse leukemic cells selected for resistance to this drug were used as in vitro m o d e ls . Biochemical dissection of these m utant cells has revealed a n o v el m echanism of resistance to antifolates, which has implications o n the conceptual understanding of folate hom eostasis as well as t h e clinical use of DDA THF for cancer chemotherapy. Further, analysis of the molecular genetics of these drug resistance m utants has p r o v i d e d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 44 valuable insights into the structure and function relationship of t h e membrane folate transporter encoded by the R F C -1 gene. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45 CHAPTER 2 STR U C TU R A L FEATURES OF D D A T H F TH AT DETERMINE INHIBITION OF MAMMALIAN GLYCINAM IDE RIBONUCLEOTIDE FO RM Y LTRANSFERA SE INTRODUCTION The dem onstration that the folate antagonist 5,10- d id eaz atetra h y d ro fo late exerts its potent anti-tum or activities by inhibiting glycinam ide ribonucleotide formyltransferase (GARFT) (EC 2.1.2.2) of de n o v o purine synthesis has prom pted i n - d e p t h investigations of the biochemical, genetic and structural properties of this biological target (Beardsley et al., 1989). In the present studies, a series of DDATHF analogs were evaluated as inhibitors of p u rifie d m am m alian GARFT. The nature of interation of DDATHF to th e active site of GARFT is deduced by relating the structures of th ese analogs to their relative potency as inhibitors of the enzyme. Our studies represent an important arm of a m u lti- d is c ip lin a ry approach to gaining insights into the topography of the active site of GARFT. Other technique include active site m apping with affinity label, site-directed mutagenesis (Inglese et al., 1990) and after o u r Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 46 studies was published, X-ray crystallography (Almassy et al., 1992; Chen et al., 1992; Klein et al., 1995). The implications of these o th e r studies on the understanding of the interactions between GARFT a n d DDATHF will also be discussed. The following are b a c k g ro u n d information regarding the genetics and biochemistry of this pharm aco lo g ically important enzyme. Molecular genetics of GARFT. De n o v o purine synthesis consists of ten sequential e n z y m a tic reactions leading to inosinic acid (Fig. 1.2). G ly c in a m id e ribonucleotide formyltransferase (GARFT) (EC 2.1.2.2) catalyzes th e third step and is the first of the two folate-dependent enzymes of th e pathway. In the vertebrate species studied to date, GARFT activity is encoded by a trifunctional protein that possesses two other e n z y m e activities, glycinamide ribonucleotide synthetase (GARS) (EC 6 .3.4.13) and am in o im id azo le ribonucleotide synthetase (AIRS) (EC 6.3.3.1), catalyzing the second and fifth reactions of the pathway, respectively (Daubner et al., 1985, Daubner et al., 1986). In their studies of GARFT purified from chicken liver, mouse L1210 and human HeLa cells, Benkovic and co-workers reported that AIRS and GARS enzymatic activities inevitably copurified with GARFT even w h en Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 47 affinity ch ro m a to g rap h y based on ligands that bind specifically to GARFT was em ployed (Daubner et al., 1985; Daubner et al., 1986). Moreover, GARS and GARFT functions were separable only after limited digest by chymotrypsin. Thus, these investigators c o n c l u d e d that all three enzyme activities probably resided on the s a m e polypeptide. In E. coli, these enzyme activities are found as s e p a ra te gene products (Smith and Daum, 1986; Shen et, al., 1990); whereas in yeast, based on cDNA studies, a m onofunctional GARFT and a bifunctional GARS-AIRS are expressed (White et al, 1985; Henikoff, 1986). Definitive evidence for a m ultifunctional protein in vertebrates was provided by the isolation of cDNAs from ch ick e n , mouse, and hum an libraries which contained a single large o p e n reading frame that could be divided into three dom ains h o m o lo g o u s to bacterial GARS, AIRS, and GARFT genes (Aimi et al., 1990; Kan e t al., 1993). Apparently, polyfunctionality of protein participating in multi-step biosynthetic pathways evolves as a result of gene fusion. Surprisingly, in addition to the 3.4 kb mRNA for the t r i f u n c t i o n a l protein, another class of messages (1.7 - 1.9 kb) corresponding to a m o n o functional GARS is also expressed in mouse tissues, and th ese shorter transcripts appears to be translated into a catalytically activ e GARS (Kan et al., 1993; Kan and Moran, 1995). Analysis of g e n o m ic and cDNA clones encoding these messages has d em o n strate d t h a t Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48 both the m onofunctional and trifunctional transcripts are g e n e ra te d from a single genomic locus in mouse (Kan et al., 1993; Kan a n d Moran, 1995). The expression of this m o n o fu n ctio n al p ro te in appears to be conserved during evolution since a separate m essage for GARS is also found in birds and human (Aimi et al., 1990). However, the reason for the retention of this apparently r e d u n d a n t monofunctional enzym e is presently unclear. Biochemical characterization. GARFT, the cellular target of DDATHF, resides at the carboxyl terminus of this trifunctional protein. The enzyme has been p u rified to electrophoretic hom ogeneity from chicken liver (Young et al., 1984), as well as from a num ber of mouse and hum an tumor cell lines (Daubner and Benkovic, 1985; Daubner et al., 1986; Caperelli, 1989). Cell fractionations performed on chicken liver and on L I 210 cells indicated cytosolic localization of the enzym e (Daubner a n d Benkovic, 1985). GARFT was initially identified as a dimeric p ro te in consisted of two 55 Kd subunits. This assignment was made on th e basis of the results of sucrose density ultracentrifugation and SDS- PAGE, which showed a functional enzyme of M r 110 Kd and a band of Mr 55 Kd, respectively (Caperelli et al., 1980). However, using Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 9 polyclonal antibodies against avian GARFT as probes, Western b lo t analysis of trichloroacetic acid hom ogenized chicken liver ex tracts revealed that GARFT was actually a m onom eric protein of 117 Kd (Young et al., 1984). The M r 55000 protein band id e n tifie d previously on SDS-PAGE was found to be a degradation product of t h e native protein. Interestingly, this proteolytic fragment appeared to retain GARFT activity (Young et al., 1984; Baldwin and M o ran , unpublished observations). G A RFT catalyzes the formylation of glycinamide rib o n u c le o tid e (GAR) to formyl-GAR using (6R)-10-form yltetrahydrofolate as the o n e carbon donor (Smith et al., 1981): o COOH N H COOH ribose-5'-phosphate Glycinamide ribonucleotide (6R)-10-formyl-tetrahydrofolic acid O COOH COOH H ribose-5'-phosphate N-formyl glycinamide ribonucleotide (6S)-tetrahydrofolic acid Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50 In their early characterization of GARFT partially purified fro m chicken liver, Buchanan and co-workers reported that the folate cofactor was 5 ,1 0 -m eth en y ltetrah y d ro fo late. However, it is n o w known that this cofactor assignment was an artifact resulted fro m contam ination by other folate-dependent enzymes in th e preparation used for these studies (Smith et al., 1981). It was n o t e d by Benkovic and his colleagues that in their purification of GARFT from the same source, four other folate-requiring enzymes 5,10- m eth en y ltetrah y d ro fo late cyclohydrolase (EC 3.5.4.9), a m in o im id a zo leca rb o x am id e ribonucleotide formyltransferase (EC 2.2.2.3), serine h ydroxym ethyltransferase (EC 2.1.2.1) and fo rm y l- m e th e n y l-m e th y le n e te tra h y d ro fo la te synthetase (combined) (EC 6.3.4.3, EC 1.5.1.5, and EC 3.5.4.9) tended to copurify with GARFT (Caperelli et al., 1980). (This is not to be confused with th e copurification of GARS and AIRS activities with GARFT activity, w h ic h is now known to be due to the fact that all three catalytic sites are part of the same protein.) It appeared that the contam inant 5,10- m eth en y ltetrah y d ro fo late cyclohydrolase in early preparation of GARFT hydrolyzed 5 ,1 0 -m eth en y ltetrah y d ro fo late to (6R)-10- form yltetrahydrofolate and, thereby, furnished the true cofactor to GARFT. (6R)-10-formyltetrahydrofolate by GARFT was initially f o u n d to be poorly, if at all, utilized as a substrate by GARFT because t h e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 1 unnatural diastereom er present in the assay, (6S)-10- form yltetrahydrofolate, is an potent competitive inhibitor of th e enzyme. When stereochem ically pure (6S)-10-formyltetrahydrofolate was prepared, it was found to be an excellent substrate (Smith et al., 1981). Smith et al. reported that the synthetic substrate 10-form yl- 5.8-dideazafolate was, surprisingly, more efficiently utilized by GARFT than was the natural folate cofactor (apparent K m values for 10- formyl-5,8-dideazafolate and 10-form yl-tetrahydrofolate in th e presence of 0.23 pM a,(3-GAR were 29 pM and 68 pM, respectively). In view of its superior chemical stability, this alternate substrate is n o w routinely used as the formyl donor in GARFT assays (Smith et al., 1981a). For purified m am m alian enzyme, using a c o n v e n ti o n a l spectrophotom etric a ssa y ', the Km values of 10-form yl- 5,8,dideazafolate and a,(3-GAR (only the (3- isomer is active) r a n g e d from 1.3 to 7 pM and 2.5 to 250 pM, respectively (Daubner a n d Benkovic, 1985; Caperelli, 1989; Baldwin et al., 1991). 1 GARFT activity was measured using a spectrophotometric assay described by Smith et al (1981a), which followed the rate of conversion of 10-CHO-5,8-dideazafolate to 5.8-dideazafolate at 295 nm in a 1ml volume, 1cm path-length cuvette. More recent studies in this laboratory have suggested that, using this assay, technical difficulties associated with measuring enzyme activity under true initial velocity conditions have confounded these estimates (see below) (Sanghani and Moran, 1997). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52 Based on the patterns of steady-state kinetic studies, Caperelli (1989) has proposed that the GARFT reaction proceeds via a sequential m echanism of substrate binding. Thus, using 10-formyl- 5,8-dideazafolate and a,(3-GAR as substrates, Linew eaver-Burke analysis of initial velocity as a function of one substrate at d ifferent fixed concentrations of the other substrate yielded intersecting lines. Using 10-acetyl-5,8-dideazafolate and a,(3 -hydroxyacetam ide ribonucleotide (GAR-OH) as dead-end inhibitors, both c o m p o u n d s were found to be competitive against their respective substrates, as expected. When 10-acetyl-5,8,dideazafolate was evaluated as a n inhibitor with GAR as the changing substrate, a n o n - c o m p e titiv e pattern was obtained. However, GAR-OH was found to be a n uncompetitive inhibitor against folate as the varied substrate. T a k e n together, these early results seemed to suggest an o rd e re d -s e q u e n tia l mechanism, with the folate substrate binding first (Fig. 2.1A). In contrast, recent studies in our laboratory using equilibrium dialysis technique to directly examine the binding of GAR and folate to GARFT have demonstrated that either substrate binds to the e n z y m e equally well in the presence or absence of saturating levels of th e other substrate, implying a random, rather than an ordered, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53 Fig. 2.1 Kinetics of the GARFT reaction. (A ) Based on steady-state kinetic experiments using d e a d - e n d inhibitors, it has been proposed that the GARFT reaction follows a n ordered sequential mechanism of substrate binding, with the b in d in g of folate precedes that of GAR (Caperelli, 1989). However, e q u ilib riu m dialysis experim ents perform ed in our laboratory dem onstrated t h a t the binding of DDATHF or GAR to recom binant GARFT was independent of the absence or presence of saturating c o n c e n t r a t i o n of the other ligand, implying random substrate binding (B). Using more sensitive kinetic assays, the K m value of the com m only u s e d folate substrate, 10-formyl-5,8-dideazafolate, was found to be m u c h lower than that previously reported. Thus, in e x p e rim e n ts perform ed at high concentrations of folate substrate (relative to its K m), the GARFT reaction will proceed via a preferred pathway (th ic k arrow) dictated by the essentially irreversible binding of folate lig a n d to the enzyme. Under such conditions, an enzyme operating by a random pathw ay (B) will behave like an ordered sequential sy stem (A ). E, free enzyme; fTHF, 10-form yltetrahydrofolate; GAR, glycinamide ribonucleotide; fGAR, formyl GAR; THF, tetrahydrofolate. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54 fTHF GAR fGAR THF E EfTHF EfTHFGAR ETHF B ETHFfGAR fTHF GAR E EfTHF E EGAR E fTHFGAR ^ ----— ETHF fGAR I fGAR THF ETHF E EfGAR E GAR fTHF ▼ t THF fGAR Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 55 sequential m echanism of ligand binding to enzyme (see below ) (Sanghani and Moran, 1997) (Fig. 2 .IB). Structural studies of GARFT, Much of the structural inform ation of GARFT has b e e n obtained from studies with enzyme derived from E. coli. Unlike its trifunctional avian or m am m alian counterparts, bacterial GARFT is encoded by a small gene of 636 bases and possesses t r a n s f o r m y la t io n as the sole catalytic activity (Smith and Daum, 1987). The s m a ll size (M r= 23 Kd) and lack of structural com plexity of this b a c te ria l enzyme makes it more amenable to genetic m anipulation a n d overexpression in a prokaryotic background. Substantial se q u e n c e identity (35%) exists between the bacterial enzym e and the GARFT domain of the trifunctional avian protein. Based on this high d eg ree of sequence homology and other com m on features such as k in e tic m echanism , substrate specificity, and sensitivity to in h ib ito rs between the two enzymes (Inglese, et al., 1990), their 3 - d im e n s io n a l structures are expected to be similar. Soon after the results presented in this chapter were p u b l i s h e d (Baldwin et al., 1990), the crystal structure of recom binant E. coli GARFT was solved sim ultaneously by two groups (Almassy et al., 1992; Chen et al., 1992). In the paper by Almassy et al., t h e structure of both the apoenzym e and that of a ternary c o m p le x Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56 containing GAR and the folate antagonist, 5 -d e a z a te tra h y d ro f o la te , were reported. More recently, the second group also reported th e structure of the bacterial enzyme com plexed with a m u l ti s u b s t r a t e adduct (BW1476U89) (Table 2.3) solved at 1.96A (Klein et al., 1995). These crystallographic studies greatly reinforce our u n d e r s t a n d i n g about the structure of the active site of GARFT as well as basis of th e binding between DDATHF and the enzyme. The general structure of the E. coli GARFT is outlined below and the details on e n z y m e - lig a n d interaction will be discussed in conjunction with the data p re s e n te d in this chapter under "Discussion". C rystallographic data indicated that bacterial GARFT fo rm s closely packed dimers as a unit cell (Chen et al., 1992; Almassy et al, 1992) although biochemical evidence suggested m o n o m e r i c organization in solution (Inglese et al., 1990a). The enzyme has b een divided into two domains. The N-terminal domain consists of d ouble-w ound p - a - p - a - p - a - p - c t sheet made of four parallel P -strands flanked on both sides by two pairs of a-helixes. The carboxyl- terminal dom ain contains three more P-strands (P5-P7) with P6 running antiparallel to all the other P-strands. Following P7 is a long a-helix which shields the mixed P-sheet on one side (Fig. 2.2). T h e ligand binding site was defined as a long narrow cleft of approxim ately 20 A at the interface between the two domains. T h e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57 folate cofactor and GAR bind adjacent to the C-terminal end of (34 and (35. The binding sites for the two substrates are contiguous to each other and appear non-overlapping. The phosphate of GAR binds close to a loop connecting the end of (31 and the N-terminus of a l . This phosphate binding motif is com m only found in n u c le o tid e - binding site of proteins where negative charges of phosphate are stabilized by the positive dipole of the N-terminus of a a-helix (Schulz and Schirmer, 1979). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58 187 101 <X6 161 133 B Fig. 2.2 X-ray crystallographic structure of GARFT Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59 Fig. 2.3 (cont’d) (A) Topology of the secondary structure of GARFT. Arrows and cylinders represent P-sheets and a-helices, respectively. The approximate beginning and ending residue numbers of each element are given. (B) Ribbon diagram of GARFT. The seven P-sheets (red arrows), the six a-helices (blue), and the coil (yellow) are shown. Binding of the substrate GAR and the inhibitor 5-deazatetrahydrofolate (white bonds) are also seen. (C) Stereo diagram of GARFT. (From Almassy et a l 1992 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 Mechanism of catalysis. Available evidence suggests that the transform ylation re a c tio n catalyzed by GARFT proceeds via a direct nucleophilic attack by th e a-am ine of GAR on the formyl carbon of the folate co su b stra te , forming a h em ioxyorthoam ide tetrahedral transition state (Smith et al., 1982; Caperelli and McKellar, 1991). In a report by Inglese et al. (1990), NlO-(bromoacetyl)-dideazafolate was used as an affinity label in order to identify amino acid residues in or around the active site of E. coli GARFT. Proteolytic digest of labeled enzyme followed by sequencing of purified peptide containing the label had identified a n evolutionary conserved (from bacteria to human) amino acid re sid u e Asp 144. Subsequent site-directed mutagenesis of Asp 144 to Asn resulted in a 10^ times less active enzyme that was still capable of binding substrates and inhibitors (Inglese et al., 1990). It was initially proposed that Asp 144 functioned as a general base by e ith e r abstracting a proton from the am inium group of GAR or dcprotonating the proposed tetrahedral transition state. However, when 10-form yl-tetrahydrofolate was modeled into the active site of the enzyme based on crystallography data, Asp 144 was found to be sterically blocked from the amine of GAR (Almassy et al., 1992). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 1 Another strictly conserved amino acid residue that may p a r tic ip a te in catalysis is His-108. This residue is found in close proximity to both the 10-position of bound folate and the amine of GAR. It is n o w believed that His-108 either accepts a proton from the amino g ro u p of GAR or forms a hydrogen bond with the carbonyl oxygen of th e formyl group and thereby activate the carbonyl for direct attack by GAR (Almassy et al., 1992; Klein et al., 1995). The important role of His 108 (the equivalent of His-915 in mouse and human has been confirmed by experiments in which catalytic activity was e l i m in a t e d when this amino residue was mutated in the single-domain h u m a n GARFT (Kan et al., 1992). In this chapter, the kinetics of inhibition of GARFT p u rifie d from mouse leukemic L I 210 cells by DDATHF and a series of closely related compounds are presented. Specifically, our purposes were: 1) to determine the structural feature(s) of DDATHF that are critical for binding to and, hence, inhibition of GARFT; 2) to identify region(s) in the DDATHF molecule that can tolerate substantial a lte r a tio n without com prom ising inhibitory activity against GARFT. T h ese regions therefore represent potential sites where modifications can be explored to improve other p h a rm a c o d y n a m ic parameters such as poly g lu tam atio n and membrane transport; 3) to recognize p o te n tia l substitutions that may lead to more potent inhibitor of GARFT; a n d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 62 4) to map the topology of the substrate binding pocket within t h e active site of this enzyme Note: The work presented in this C hapter was performed i n collaboration with Dr. Samuel W. Baldwin, a former p o s t-d o c to ra l fellow in the laboratory. Dr. Baldwin was involved in t h e purification of enzyme and a portion of the studies of GRAFT inhibition by D D A T H F analogs. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 CHAPTER 2 MATERIALS AND METHODS Chemicals RPMI 1640 tissue culture and fetal calf serum were p u r c h a s e d from GIBCO (Grand Island, NY). Cynogen bromide a c ti v a te d Sepharose 4B and all other general chemicals were from S igm a Chemical Co. (St. Louis, MO). (6R,S)-Tetrahydrofolate was prepared by reduction of folic a c id with d im e th y la m in e -b o ra n e complex (Martinelli and C h ay k o v sk y , 1980), whereas (6S)-tetrahydrofolate was prepared by stereospecific reduction of folic acid by partially purified DHFR isolated from MTX- resistant L. casei (Moran et al., 1976). Both compounds were p u rifie d by DEAE-cellulose chrom atography and stored in lyophilized form a t -20 °C as previously described (Moran et al., 1976). (6S)-5- F o rm yltetrahydrofolate was synthesized by formylation of tetrahydrofolate in the presence of a water-soluble carbodiimide a n d was purified by chromatography as described by Moran and C o le m a n (1985). (6R )-10-Form yltetrahydrofolate was obtained by acidifying a solution of (6S )-5-form yltetrahydrofolate to pH 1 for 20 min a n d then adjusting the pH to 8 for 30 min in the presence of 1% 2- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 64 m ercaptoethanol (P-ME) (Rabinowitz, 1970). (6R,S)-DDATHF a n d (6R,S)-5-deazatetrahydrofolate were synthesized and provided by Dr. Edward Taylor from Princeton University (Princeton, NJ). (6R,S)- 5,8,10-trideazatetrahydrofolate was provided by Dr. Andre Rosowsky of the Dana-Farber Cancer Institute (Boston, MA). The synthesis of all other DDATHF analogs and DDATHF polyglutam ates were by Dr. Chuan Shih of Lilly Research Laboratories (Indianapolis, IN). Except where noted, all analogs containing a tetrahydropyridine ring were present as a mixture of (6R)- and (6S)- diastereomers. a,P- G lycinamide ribonucleotide was a generous gift of Dr. Homer Pearce also of the Lilly Research Laboratories. 10-Formyl-5,8- d id eaz atetra h y d ro fo late was initially supplied by Dr. Terrence J o n e s of Agouron Pharm aceuticals Inc. (La Jolla, CA) and subsequently by Dr Shih. Analog concentrations DDATHF analogs were dissolved in assay buffer and t h e concentrations determined spectro p h o to m etrically using t h e extinction coefficients (measured in 0.1 N NaOH) listed below: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 65 Compound e (cm 1 M-1) (X (6R,S)-DDATHF 11,700 (272) (6R)-DDATHF 11,700 (272) (6S)-DDATHF 11,700 (272) 2 -d esam in o -D D A T H F 9,000 (269) (6R,S)-tetrahydrofolic acid 9,200 (292) (6S)-tetrahydrofolic acid 9,200 (292) 5-d eazatetrahydrofolic acid 23,800 (278) 5 ,8 ,10-trideazatetrahydrofolic acid 21,000 8,300 (237) (269) 5,10-dideazafolic acid 6,400 (335) acyclo-DDA THF 12,300 (273) nor-DDATHF 10,970 (272) ho m o -D D A T H F 12.000 (272) abenzyl-DD A TH F (n=2) 10400 (274) abenzyl-DDA THF (n=3) 10400 (274) abenzyl-D DA TH F (n=4) 10400 (274) cyclohexyl-D D ATH F 10400 (274) 5 , 1 0 -dideazatetrahydropteroic acid 11,700 (272) 5 ,1 0 -d id eaz atetrah y d ro asp artic acid 11700 (273) ax (nm)) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 66 Cell Culture Mouse leukemic L 1210 cells were originally obtained from Dr. George Moore, formerly of Roswell Park Memorial Institute, Buffalo, NY. Cells were grown in suspension culture at 37°C in a h u m i d i f i e d 5% C O 2 atm osphere. The growth m edium em ployed was RPMI 1640 supplemented with 10% and 5% fetal calf serum (from GIBCO, G r a n d Island, NY), for routine passage of stock cultures and large scale enzyme purification, respectively. Stock cultures were m aintained in logarithmic growth phase by splitting to a density of =10^ c e lls /m l every other day. Cell density was d eterm ined electronically using a Coulter counter, Model Z f (Coulter Electronics, Inc., Hialeah, FL). T h e doubling time was 10-12 h under these growth conditions. In a typical large scale purification, two 1-L cultures at a density of 8-10 x 10^ cells/ml were used to inoculate 20-L of media contained in tw o 10-L spinner flasks. Cells were allowed to grow to a final density of 8- 10 x 10^ cells/ml. Stock cultures were consistently found free of M y c o p l a s m a upon screening with the Gen-Probe detection s y s te m (San Diego, CA). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 67 Protein assays Protein concentrations were determ ined by the method of Hartee using bovine serum albumin as a standard. Protein separation by SDS-PAGE Twelve percent separating and 4% stacking gels were run by the method of Laemmli (1970) in an 8.0 cm x 7.3 cm m ini-gel apparatus (Bio-Rad Laboratories, Hercules, CA). The following p ro te in standards were used: phosphorylase B (974 kd), bovine s e r u m albumin (660 kd), ovalbumin (430 kd), bovine carbonic a n h y d r a s e (300 kd) and soybean trypsin inhibitor (201 kd). Enzyme purification Purification of GARFT was performed by a two-step p ro c e d u re similar to that described by Young et al. (1984), a scheme o riginally designed for purifying thymidylate synthase (Rode et al., 1979). T his procedure involved an initial fractionation of cytosolic extracts by am m onium sulfate precipitation, followed by affinity ch rom atography with 10-form yl-5,8-dideazafolate as the lin k e d ligand. The affinity column was prepared in the following m a n n e r : Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 68 Three grams of cyanogen bromide (CNBr) Sepharose 4B was m i x e d with 500 ml of 1 mM HC1 for 15 min. After washing with 100 ml of coupling buffer (0.1 M NaHC0 3 , 0.5 M NaCl, pH 8.3), the a c ti v a te d CNBr-Sepharose was transferred to a 50 ml centrifuge tube, to which 5 ml of coupling buffer and 2.5 ml of ethylenediam ine were then added. The tube was cooled in an ice bath, capped and mixed b y tumbling at room temperature (RT) for 8 h. The e t h y l e n e d ia m in e - linked Sepharose was filtered and washed with 4 cycles of 250 ml of 1M NaCl, followed by another 4 washings of 250 ml of w ater. Thirteen mg of 10-formyl-5,8-dideazfolate in 3 ml of 50% (v/v) dimethylacetamide-water, pH 8.0, was then added to the resin, which was suspended in a small amount of water (=5 ml). Two h u n d r e d mg of the coupling agent, N -e th y l-N -[3 -(d im e th y la m in o )- propyl]carboiimide hydrochloride was added to the mixture and t h e PH was adjusted to 5.6. The reaction was tum bled in a capped tube for 10 h at RT. After another addition of 50 mg of carbodiimide, t h e pH was again adjusted to 5.6, and the reaction was allowed to continue for an additional 10 h. The resin was then filtered a n d washed 4 times in the following sequence: 25 ml of 0.1 M NaHCC>3 containing 0.5 M NaCl, pH 8.5, then 25 ml of water, and then 25 m 1 of 0.1 M sodium acetate containing 0.5 M NaCl, pH 4.6. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 69 In a typical large scale purification, 20-L of cells grown to late log-phase were harvested by centrifugation at 500 x g ( 1 1 0 0 rpm) for 10 min at 4 °C using a Beckman J6 centrifuge (Fullerton, CA), washed twice with ice cold PBS, resuspended in low salt buffer (1 0 mM sodium phosphate, pH 7.5, containing 0.1% Triton X-100 and 10 mM (3-ME) and disrupted by sonication on ice. All subsequent steps were carried out at 4 °C. Cell extracts were centrifuged at 165,000 x g for 1 h in a Ti 45 rotor (Beckman, Fullerton, CA) at 4° C. A m m o n i u m sulfate was then slowly added to the su p ern atan t with c o n s t a n t stirring until 30% s a tu ra tio n (176m g/m L) was reached. T h e suspension was centrifuged at 20,000 x g for 20 min, and the p ellet was discarded. The supernatant was then brought to 70% s a t u r a ti o n by (N H 4 )2 S 0 4 (270mg/mL) and the suspension was again c e n trifu g e d at 20,000 x g for 20 min. The pellet was dissolved in a m i n i m u m volume of low salt buffer and the solution was dialyzed o v e rn ig h t against 4-L of the same buffer. The dialyzed 30-70% (NH4 )2SC>4 fraction was then applied slowly (0.3 m l/m in) to a 1 x 5 cm 10- form yl-5,8-dideazafo!ate-linked affinity colum n equilibrated in low salt buffer. The enzym e-bound column was washed with 30 ml of low salt buffer followed by a overnight wash of 500 ml of high s a lt buffer (0.2 M sodium phosphate, pH 7.5, 0.5 M KC1, 0.1% Triton X- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 70 100 and 10 mM P-ME). To elute, 3 ml of a 3 mM solution of (6R)-10- form yl-tetrahydrofolate was added to the column, flow through th e column was stopped for 30 min to allow equilibration b etw e en enzyme bound to the column and that free in the mobile p h a s e . The enzyme was then eluted by another 3 ml of the same s o lu tio n . Half-ml fractions were collected and assayed for GARFT activity. Enzyme assays GARFT activity was measured as described by Smith et al. (1981b). Except where noted, the synthetic substrate, 10-form yl-5,8- dideazafolate was used as the formyl donor. The course of th e reaction was followed by monitoring absorbance increases at 295 n m due to the production of 5,8-dideazafolate using a Gilford re c o rd in g spectrophotom eter. The initial linear increase in absorbance was used for initial velocity calculation. Product formation was determ ined by Beer's Law using a As value of 18,900 M ' l c m ' l and a light path of 1 cm . For experiments using (6R)-10- formyltetrahydrofolate as a substrate, reaction was m onitored at 312 nm, and Ae = 12,000 M ‘ ^cm '1. Reactions were perform ed at 25 °C i n 75 mM Hepes buffer, pH 7.5, containing 20% glycerol and 45 mM a- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 1 thioglycerol in a final volume of 1 ml. A typical assay mixture for enzyme purification purposes contained 11 pM 10-form yl-5,8- dideazafolate and 10 pM a,P-glycinam ide ribonucleotide. R eactions were initiated by adding either GAR or en z y m e . One unit of GARFT activity was defined as 1 pmol of product formed per min. The (6R)- 10-form yltetrahydrofolate used for eluting enzyme from the affin ity column was removed prior to assays for kinetic studies by passage of enzyme through a 10-ml Sephadex G-50 gel filtration column. Kinetic analysis For K m and V max determ inations, substrate c o n c e n t r a t i o n s were varied from 0.5 to 10 times their respective K m values. K inetic param eters were obtained by fitting data to a rectangular h y p e r b o la using the program of Cleland (1967). For Ki determ inations, initial estimates were first obtained by measuring enzyme activity as a function of inhibitor c o n c e n t r a t i o n s at fixed concentration of folate substrate and GAR. Five in h i b it o r concentrations were used and each m easurem ent perform ed in duplicate. Assuming competitive inhibition with folate: Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 72 1/v - (K m /V m ax-S.K j,initial) [I] + 1/Vmax ^ + Km/S) where v and V niax are initial velocity and maximal velocity; K m a n d Kj,initial* are M ichaelis-Menten and initial estimate of in h ib itio n constant, respectively; S is the concentration of 10-formyI-5,8- dideazafolate used. Ki.initial. was determ ined graphically by plotting 1/v versus [I] (Dixon plot). This value was s u b s e q u e n tly used to determ ine the range of substrate concentration employed i n the final estim ation of K, by Lineweaver-Burke analysis, in w h ich activity was m easured with varied concentrations of folate s u b s tra te in three or four fixed concentrations of inhibitors. In the absence of inhibitor, concentration of the folate substrate was varied from 0.5 to 10 times its Km values. In the presence of inhibitor, the range was 0.5 to 9 times its apparent K m (K mapp), where K m,app = K m (1 + [I]/ Kj,initial) f°r com petitive inhibition. After fitting data to a rectangular hyperbola, final estimates of Kj were determined from t h e intercept of a replot of slope versus inhibitor concentration. All Kj values determ ined by Lineweaver-Burke analysis were performed a t least two separate experiments with duplicate points at five d ifferen t concentrations of the folate substrates. For analysis of tight b in d in g inhibitors, i.e. DDATHF polyglutamates, Kj values were obtained by Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 73 fitting data to the Morrison equation (see Results) with the ENZFTTTER program (Elsevier Science Publishers, Amsterdam). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 74 CHAPTER 2 RESULTS Enzyme purification It has been shown that mouse leukemic L1210 cells contained a 6-fold elevated GARFT activity com pared to mouse lym phocytes a n d thus represented a rich and convenient m am m alian source of th is enzyme (Daubner and Benkovic, 1985). Using affin ity ch ro m a to g rap h y as the principle step, GARFT was purified 4 3 5 -fo ld from a 160,000 x g supernatant to eletrophoretic hom ogeneity w ith 60-80% yield (Fig. 2.3). The specific activity of the purified e n z y m e was 1.2 units/m g of protein when assayed with 11 pM 10-form yl-5,8- dideazafolate and 10 pM a,(1-GAR. It has been reported that during purification of GARFT fro m chicken liver, the enzyme was susceptible to proteolytic d e g r a d a t i o n to lower m olecular weight fragments (primarily 55 kd species) t h a t appeared to retain catalytic activity (Daubner and Benkovic, 1985). During our initial studies with murine GARFT, it was found that e v e n highly (but, apparently, incompletely) purified enzyme was re a d ily proteolyzed. At least one of the degradation products was found to be a more efficient catalyst with a higher turn over number than the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 5 1 97.4 6 6 43 30 20.1 Fig. 2.3 SDS-PAGE of L1210 GARFT during purification. Aliquots of protein were subjected to SDS-PAGE at each step of enzyme purification. (Lane 1) Protein standards of the i n d i c a t e d molecular weight in kd. (Lane2) 30 pg of cytosolic L1210 cell extract. (Lane 3) 30 pg of 30-70% (NH4)2S04 pellet. (Lane 4) 5 pg of GARFT eluted from the affinity column. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 76 native 110 kd enzyme (Baldwin and Moran, u n p u b l i s h e d observation). Hence, it was concluded that GARFT preparations for kinetic experiments must be very highly purified and stable d u rin g incubations. Enzyme prepared according to the procedure d escrib ed above was stable to storage at -25 °C in assay buffer containing 50% glycerol for at least 2 months. No proteolytic fragments were detectable on SDS gels after incubation at 25 °C for up to 4 h u n d e r assay conditions (data not shown). Steady state kinetics Assuming an ordered-sequential kinetic m echanism , the stead y state initial velocity is given by the following expression: Vmax [F] [G] v _ K. K _ + K r [F] + K [G] + [F] [G] il- mG mG L J mh L J l j i j where [F] and [G] are the concentrations of folate substrate and a,P- GAR, respectively; Kjp is the dissociation constant between free enzyme and folate; K np and K represent the true M ichaelis-M eten constants for folate and a,P-G A R , respectively. Using 10 pM a,P-GAR as the cosubstrate, the apparent M ichaelis constants for 10-formyl-5,8-dideazafolate and 10- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 77 form yltetrahydrofolate were 1.3 ± 0.2 pM (n = 16) and 6.1 ± 1.1 pM (n = 3), respectively. In the presence of 1.1 pM 10-formyl-5,8- dideazafolate, the apparent K m of GAR was 4.3 ± 0.4 pM (n = 8 ) . Substrate inhibition was observed with GAR when either of the folate cosubstrate was used (Fig. 2.4). It should be noted that the velocity did not approach zero even at very high concentration of GAR, indicating only partial substrate inhibition. This is inconsistent w ith an ordered sequential mechanism where com plete s u b s tra te inhibition is typically produced by binding of the second substrate to free enzyme, resulting in the formation of a dead-end c o m p le x . Recent work in this laboratory has d em o n strated that, when th e interaction between GARFT and its ligands was studied directly u sin g equilibrium dialysis, the binding of [3h ]-(6R)-DDATHF to recom binant mouse GARFT was inhibited by only 50% even a t concentration of GAR of 750 pM (Sanghani and Moran, 1997). T his observation suggested that the previously proposed o rd e red - sequential model for this enzyme (Caperelli, 1989) may req u ire modification (see “Discussion” of this Chapter). Inhibition of GARFT by DDATHF analogs A series of DDATHF analogs modified at various regions of th e parent compound were studied as inhibitors of GARFT. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 78 0.08- 0.00 O.tO 0.20 0.30 0.40 i ;g a r . 1/uM 0.06- e E "5 O 0.04- E > 0.0 2 - 0.00 100 200 300 650 GAR, |iM Fig. 2.4 Substrate inhibition of GARFT by GAR. Enzyme activity was m easured in the presence of the i n d i c a t e d concentrations of GAR and with either 11 pM 10-formyl-5,8- dideazafolate (open circle) or 60 pM ( 6 S )- 1 0 -fo rm y lte trah y d ro fo la te (filled circle). The inset is a replot of 1/v (m in /p m o l) versus 1/GAR (1/pM ) (from Baldwin et al., 1991). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 79 C O O H C O O H Using 10-formyl-5,8-dideazafoIate as the varied substrate, as h a d been reported with DDATHF itself, all of the analogs described i n Table 2.1 were competitive inhibitors of the enzyme based o n Lineweaver-Burke analysis (Moran et al., 1989). Results for h o m o - DDATHF was shown in Fig. 2.5 as a representative example of th is pattern. With GAR as the variable substrate, Lineweaver-Burk p lo ts intersected on the abscissa, indicating that inhibition was n o n c o m p e t i t i v e . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 80 Table 2.1 Activity of DDATHF Analogs as Inhibitors of GARFT® Structure Compound K, (pM ) O COOH HN5 0 HN' H (6R,S)-DDATHF “COOH (6R)-DDATHF (6S)-DDATHF 0.12 ± 0.02 0.10 ± 0.02 0.029 ± 0 .0 1 2 2-desamino-DDATHF 27 (23,31) COOH O COOH H,N N N O COOH COOH (6R,S)-tetrahydrofolate 6.0 (5.8,6.1) (6S)-tetrahydrofolate 6.3 (5.0,7.6) 5-deazatetrahydro- 0.065 (0.064, COOH folic acid Y ' V V ' ^ A ^ H H2N^^N n 0.066) HN H O O COOH COOH 5,8,10-trideaza- tetrahydrofolic acid HN' H H O COOH "COOH 5,10-dideaza- folic 12(9.0,15) 13(12,14) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 1 Table 2.1 (c o n t’d) Structure Compound K, (pM) O COOH COOH acyclo-D DATHF 0.38 (0.47,0.29) nor-DDATHF .COOH COOH hom o-DDATHF COOH COOH COOH COOH abenzyl-D D A T H F 0.049(0.061, (n=2) 0.036) H2N n COOH COOH abenzyl-DDATHF (n=3) 0.028 (0.035, 0 .021) COOH H2N n COOH abenzyl-D DATHF (n=4) 0.018 ± 0 .0 0 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 82 Table 2.1 (c o n t’d) Structure Compound Kj (pM) COOH COOH HN ,COOH HN H cyclohexyl- 0.025 (0.022, DD ATHF 0.027) 5,10-dideazaletra- 0.15 (0.16,0.14) hydropteroic acid HN 5,10-dideazatetra- 0.047 (0.032, hydropteroyl 0.062) aspartic acid ^Mixtures of 6R and 6S disstereomers is indicated by an asterisk in the structure. Kinetic constants were obtained from relots of t h e slope of Lineweaver-Burk plots versus inhibitor concentrations with a t least three concentrations of inhibitors. Kinetic constants were determ ined with the program of Cleland (1967). The values liste d are the means of three or more determinations (±SD) or the means of two experiments with individual values in parentheses. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 83 Modifications in the tetrahydropyridopyridine ring The availability of analogs with substitutions in t h e heterobicyclic ring allowed probing of the nature of i n te r a c tio n between this region of the analog and the active site of GARFT. For example, the 2-desamino analog was >200-fold less active t h a n DDATHF itself as inhibitor of GARFT (Table 2.1), suggesting t h e involvem ent of hydrogen bond(s) between the 2-amino group of DDATHF and certain amino acid residue(s) in the active site. DDATHF is best viewed as a derivative of tetrahydrofolate, t h e natural product of the GARFT reaction. Tetrahydrofolate itself was a weak inhibitor of the enzyme, whose potency did not change w ith the absolute configuration at C-6 of the heterocyclic ring (Kj = 6 .0 and 6.3 pM for the (6R,S)-mixture and the (6S)-isomer, re sp ectiv ely ) (Table 2.1). In contrast, both of the diastereomers of DDATHF were rather potent inhibitors of GARFT, with the (6S)-isomer s o m e w h a t more active (Kj values for (6R)- and (6S)-DDATHF were 0.10 ± 0 .0 2 pM and 0.029 ± 0.012 pM, respectively) (Moran et al., 1989) (T ab le 2.1). Therefore, it appeared that the near-isosteric replacem ent of both nitrogen atoms at position 5 and 10 by carbons in tetrahydrofolate gave rise to an inhibitor that binds 60-fold tighter to the enzyme. Further, it appeared that only one of th e se Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 84 substitutions was sufficient for GARFT inhibition, as the 5 -d e a z a analog was found to be even more potent than DDATHF as a n inhibitor of the enzyme (Table 2.1). The replacem ent of N-8 as well by carbon resulted in the compound, 5 ,8 ,1 0 -trid e a z a te tra h y d ro fo la te , which, remarkably, showed virtually no activity against GARFT, w ith a Kj value 100-fold higher than that of DDATHF (Table 2.1). A reduced pyridine ring seemed to be required for tight in te ra c tio n with the folate binding pocket of the enzyme, as suggested by t h e poor inhibitory effect of 5,10-dideazafolic acid (Table 2.1). D esp ite the stringent requirem ent for specific m odifications in t h e tetrah y d ro p y rid in e ring of DDATHF, the acyclic analog with the 7- carbon rem oved was comparable in potency to DDATHF as a n inhibitor of GARFT (Table 2.1). Alterations in the bridge and phenyl ring regions The methylene bridge and the phenyl ring portions of DDATHF are relatively non-polar, and can participate in h y d r o p h o b ic interaction with the enzyme and/or function as spacers to p o s itio n the bicyclic ring and the glutamic acid moiety in fav o rab le orientations in the active site. While lengthening the bridge region by one m ethylene unit (5 ,1 0 -d id eaz atetra h y d ro h o m o fo late) r e s u lte d in a 6-fold increase in potency against GARFT, removing o n e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 85 m ethylene group in this region ( 5 ,1 0 - d id a z a te tr a h y d r o n o r f o la te ) caused a 5-fold decrease in activity (Table 2.1). Replacement of th e phenyl ring of DDATHF by either two, three, or four methylene g ro u p s not only allowed retention of GARFT inhibition but resulted in analogs that were significantly more active than the p a r e n t com pound (Table 2.1). Likewise, substituting the phenyl group w ith a fully saturated cyclohexyl ring gave rise to a somewhat m o r e potent analog, cyclohexyl-DDATHF. This latter observation a r g u e d against the presence of any specific Tt-rc-interaction between th e phenyl ring and any aromatic side chains in the folate binding site of GARFT. Taken together, it was concluded that the precise geometry in the bridge/phenyl ring region was not critical for b in d in g to GARFT. However, it appeared that there was a trend to w a r d enhancing association of analogs to GARFT as the sp acin g /flex ib ility between the heterocyclic ring and the glutamic acid moiety was extended by methylene groups (homo-DDATHF, abenzyl-DDATHF, n = 3; n = 4 ) or by a less rigid cyclohexyl ring. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 86 Requirement of the glutamic acid side chain for GARFT inhibition DDATHF was recognized as an excellent substrate for m am m alian FPGS soon after its synthesis (Taylor et al., 1985; Beardsley et al., 1989) and has subsequently been shown to be metabolized extensively into long chain polyglutam ates in i n t a c t cells (Pizzorno et al., 1990). These findings raise the question of whether DDATHF itself function as a "prodrug" in v iv o , with t h e polyglutam ate metabolites as the actual cytotoxic mediators. T w o classes of DDATHF derivatives were available for assessing t h e contribution of these metabolites to GARFT inhibition, and hence to overall cytotoxicity induced by the drug. The first class consisted of analogs that showed no s u b s tr a te activity for FPGS as a result of having the glutamic acid side chain of DDATHF rem oved (5,10 -d id eazatetrah y d ro p tero ic acid) or r e p la c e d by an aspartic acid ( 5 , 10-dideazatetrahydropteroy! aspartic ac id ). These non-polyglutam atable com pounds were also poor inhibitors of tumor cell growth (Moran et al., 1989; Dr Chuan Shih, Lilly R esearch Laboratories, unpublished observation). It was of interest to determine w hether their lack of anti-tum or activity was due to p o o r inhibition of GARFT. Interestingly, 5 ,1 0 -d id e a z a te tr a h y d r o p te r o ic acid was found to be a competitive inhibitor of GARFT with p o te n c y Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 87 comparable to that of DDATHF, a clear indication that the g l u t a m a t e side chain of DDATHF was not critical in binding to the e n z y m e (Table 2.1). In accord with this finding, 5,10- dideazatetrahydropteroyl aspartic acid was equivalent to DDATHF as inhibitor of GARFT (Table 2.1). The second class of analogs comprised of c h e m ic a lly synthesized polyglutam ate derivatives of DDATHF, with up to 4 glutamate residues successively added to the y-carboxy group of th e parent c o m p o u n d (DDATHF-di, tri, tetra and pentaglutam ate). T h e activity of these analogs as inhibitors of GARFT was examined in order to determine the potential involvement of these metabolites i n the cytotoxicity of DDATHF directly. When the binding affinity of DDATHF polyglutam ates to GARFT were initially estimated by D ixon plots (data not shown), it was apparent that the potency a g a in s t GARFT increased as a function of chain length, with calculated K j i n the s u b n an o m o lar range for the pen tag lu tam ate derivative (T ab le 2.2). Given the fact that the concentration of enzyme used in th e se experiments was 2-4 nM, tight enzym e-inhibitor interaction as s u c h would result in most inhibitor being enzym e-bound. T h u s, estimation of Kj values based on steady state kinetic analysis, w h ic h assumes enzyme-bound inhibitor is a negligible fraction of the total Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 88 Table 2.2 Inhibition o f GARFT by (6R,S)-DDATHF Poly glutamates No. of Glutamates Kj (nM) Dixon Plot” Henderson plotb Morrison equation" l d 39 ± 4 (n = 6) 33 37 ± 5 (n = 6) 2 5.8 (5.6,6.0) 6.1 (5.9,6.2) 3 2.7 ± 1.6 (n = 3) 3.4 ±1.5 (n = 3) 4 1.4 (1.0,1.7) 1.2 (1.1,1.3) 5 0.39 ±0.15 (n = 5) 0.40 0.47 ±0.01 (n = 5) a K, values were determ ined from the slopes of the plots o f 1/v versus inhibitor concentrations with five concentrations of inhibitors. Assum ing that inhibition is competitive w ith the folate substrate, the slope o f D ixon plot equals K n/V „mS K , where K m and S are the M achaelis-M eten constant and the fixed concentration o f the folate substrate, repectively. Values listed are the mean (± S.D.) of n replicate experim ents or the mean of two experim ents with individual values in parenthesis. Data were fitted to eq 1 with EN ZFITTER to determine the Km ,app. and the K was calculated from eq 2. The goodness of fit o f the data to eq 1 was indicated by the average SE of deviation from the function after the residuals were minimized; this value was 5.1, 0.7, 0.46, 0.22, and 0.04 when applied to the K, values from the mono- to pentaglutam ate derivatives, respectively. d Linew eaver-Burke analysis indicated a K value of 0.12 pM for DDATHF m onoglutam ate (Table 2.1). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 89 inhibitor, will not be valid here. This includes any graphical an a ly s is that was based on the classical M ichaelis-M enten equation such as Lineweaver-Burke or Dixon plots. In view of this, kinetic data were iteratively fit to the non-linear equation described by M o rriso n (1969) using the ENZFITTER microcomputer software, vi = ( V 2 E t){ Et - K j app - It + [ ( E t - K j a pp - It )2 + 4 K j app E t ] 1/2} ( 1) where v 0 and vj are uninhibited and inhibited reaction velocities, respectively; E t and It are the total concentrations of enzyme a n d inhibitor, respectively; and K j a pp is an apparent d is s o c ia tio n constant generated whose meaning depends on the mode of inhibition. Assuming inhibition of enzyme activity is c o m p e t i t i v e with the folate substrate, K jap p = Kj (1 + S/ Km ) ( 2 ) where Kj is the kinetically determ ined dissociation constant b e tw e e n enzyme and inhibitor; S and K m are the concentration of the fo late Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 90 substrate, 10-formyl-5,8-dideazfoIate (11 pM) and its M ichaelis constant (1.3 p M ). Data were also fit to the relationship derived by H e n d e rs o n (1972) : It / l-(vi/v0 ) = Kjapp (v0 /vj) + (3) Using a fixed E t of 2.7 nM, the Kj estimates for DDATHF m o n o g lu tam ate to pentaglutam ate obtained with the M orrison equation were shown in Table 2.3. The Kj of the p e n t a g l u t a m a t e derivative was almost 100-fold lower than that of DDATHF. Thus, we concluded that polyglutam ation of DDATHF greatly enhanced th e inhibitory potency Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 1 CHAPTER 2 DISCUSSION In the present study, we em ployed a s tru ctu re -ac ti vity relationship approach to probe the active site of GARFT. A systematic com parison of the potency of a series of DDATHF an a lo g s as inhibitors of GARFT allowed deductions to be made about th e basis of ligand recognition by this enzyme. Classification of th e effects of various structural modifications of DDATHF on GARFT inhibition also forms the framework for rational design of clinically useful chemotherapeutic agents directed against this enzyme. Structural determinants of DDATHF for GARFT inhibition What critical structural features of DDATHF are required for GARFT inhibition? This question is best approached by c o m p a r i n g the structures and the relative inhibitory activities of three closely related com pounds: DDATHF, 5-deazatetrahydrofolate, a n d tetrahydrofolate. The order of their relative potency against GARFT was DDATHF = 5-deazatetrahydrofolate » tetrahydrofolate. Thus, it appeared that the replacement of nitrogen at position 5 by c a rb o n was necessary and sufficient for inhibition of GARFT. This single Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 92 modification resulted in an inhibitor with a 100-fold decrease in Kj relative to tetrahydrofolate, a change equivalent to a difference of 2.7 kcal/mol in binding energy. Since C/N substitution creates o n ly minor variation in molecular geometry, the drastic change in binding affinity probably resulted from alteration in elec tro static property of the folate molecule, most likely related to hydrogen b o n d formation. The therm odynam ics of ligand binding depends on th e overall energetic changes result from breaking interaction with th e solvent and establishing new contact with the m a c r o m o le c u le . Therefore, a binding event can be unfavorable if the ligand has to desolvate before associating with the enzyme and there is no new interaction in the enzyme-ligand complex to com pensate (Jones el al., 1989). This appears to be the case with tetrahydrofolate a n d other reduced folates bearing a nitrogen at position 5. It is energetically costly for a NH group to desolvate; crystallographic d a t a provides direct evidence that there is no interaction forms betw een N-5 and any hydrogen bond accepting residues on GARFT (Almassy et al., 1992). On the other hand, DDATHF and other 5-deaza an a lo g s desolvate more readily before binding to G ARFT because of the lack of hydrogen bonds between C-5 and solvent molecules. Our studies indicate that at least two other functional g ro u p s in the bicyclic ring of DDATHF are critical for binding to GARFT. T h e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 93 2-amino group of the pyrim idine ring was particularly important as removal of this functional group virtually elim inated in h ib ito ry activity (Table 2.1). We postulated that this functional group was involved in one or more hydrogen bonds to the active site residues o r to bridging water molecules. In addition, replacem ent of N-8 by carbon ( 5 ,8 ,10-trideazatetrahydrofolate) resulted in a 100-fold decrease in binding affinity, suggesting the participation of this NH group in hydrogen bonding with the enzyme. We fu r th e r hypothesized that it was the hydrogen of this secondary amine t h a t served as the hydrogen bond donor because of the fact that 5,10- dideazafolic acid, in which no polarized hydrogen would be av a ila b le at position-5, bound poorly to GARFT (Table 2.1). Of course, o t h e r factors can also be responsible for the inactivity of 5 , 10-dideazafolic acid such as the rigidity of the oxidized pyridine ring as com pared to the flexible tetrahydropyridine ring of D D A TH F (see below). Our analysis thus far suggested that the inhibition of GARFT by DDATHF analogs was very sensitive to modification in th e tetrahydropyridopyrim idine ring, especially the composition of p o la r groups. This was later confirmed by X-ray crystallographic stu d ies. Thus, in the 5-deazatetrahydrofolate-GA R-GARFT ternary co m p lex , the bicyclic portion of the inhibitor is well ordered (Almassy et al., 1992) (Fig. 2.6). This is also the case with the binary c o m p le x Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 94 between GARFT and the multisubstrate inhibitor BW1476U89, w here the bicyclic ring is mimicked by the pyrim idone ring (Klein et al., 1995). The heterocyclic ring binds to a hydrophobic pocket fo r m e d by Leu85, Phe88, Leu92, Val97, Leul04 and V all39 (Fig. 2.6). T h e polar atoms of the heterocyclic ring forms hydrogen bonds (< 3.1 A) with residues Arg90 (to N-8), Leu 92 (to N-l and N-2a ), T h rl4 0 (to N- 3), Asp 141 (to N-2a ), and Aspl44 (to 0-4). Interestingly, contacts via hydrogen bond are made exclusively with backbone of the p ro te in without any involvem ent of side chain atoms (Almassy et al., 1992; Klein et al., 1995) Remarkably, this binding pattern is also found in the folate-binding site of DHFR and TS where backbone atoms fo rm hydrogen bonds with polar groups of the bicyclic ring (Matthews et al., 1985; Matthews et al., 1989). Since there is no obvious sequence similarity among the folate-binding sites of these three enzymes, th e functional resemblance of these binding motifs for p terid in e-lik e rings probably signifies the consequence of convergent evolution. Exam ination of the interaction of TS, DHFR, and GARFT w ith their respective folate ligands reveals that the folate moiety b in d s with a different conformation to GARFT as compared to the other Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 95 Fig. 2.5 Stereo view showing the binding of folate inhibitor, 5- deazatetrahydrofolae and substrate GAR (both in green) to the active site of GARFT. Selected backbone atoms are shown in orange. The heterocyclic ring of the inhibitor is well ordered. This portion of the folate m o lecu le binds to a hydrophobic pocket formed by Leu85, Phe88, Leu92, Val97, L eul04 and V all39 (yellow). Six hydrogen bonds (purple) are fo rm e d between the heterocyclic ring and backbone atoms of the e n z y m e ( N ' — 92N, N :a,*»920, N :“»»*1410, 0 4“***144N, N 8" « 9 0 O a n d N -■’•••140O ). Hydrophilic residues are shown in blue (from Alm assy et al., 1992). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 96 two folate-requiring enzymes. In TS and DHFR, folate binds with a preferred ga u ch e * conform ation (50-69°) defined by the N-5/C-6/C- 9/N-10 atoms while in GARFT the preferred conform ation is trans (166-174°) (Fig. 2.7). This observation is of particular relevance to the unexpected results obtained with the p y rro lo p y rim id in e an a lo g of D D A THF (Table 2.3). Although this compound possesses a n u m b e r of structural features of DDATHF important for GARFT inhibition: 2- amino, 5-deaza, and 8-NH; it was totally inactive against m o u s e enzyme (K( > 8 uM) but, surprisingly, was found to be an excellent inhibitor of TS (Taylor, 1993). The lack of GARFT inhibitory activ ity with this analog is presum ably because of its inability to adopt th e trans conformation required for binding. Regions of DDATHF that can tolerate substantial modifications without compromising activity against GARFT In contrast to the heterocyclic ring, the bridge and p h e n y l regions of DDATHF seem to be able to tolerate considerable s tr u c tu r a l changes without resulting in substantial decrease in GARFT inhibition. In accord with our results, X-ray diffraction studies h a v e demonstrated that the phenyl group associates with amino acid side chains of G ARFT mainly by hydrophobic interactions (but not by n - n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 97 Trans Pteridine N 10 C9 N1 Trans> I Pteridine N10 C9 N1 Fig. 2.6 Folate compounds bind with different conformations to folate- dependent enzymes. T etrahydrofolate (black) binds to DHFR with a g a u c h e + r o t a m e r conform ation (50° to 69°) for the dihedral angle defined by the N- 5/C-6/C-9/N-10 atoms, whereas 5-deazatetrah y d ro fo late (red) b in d s to GARFT with a trans (166° to 174°) conform ation for the s a m e dihedral angle (from Klein et al., 1995). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 98 Table 2.3 Structures o f DDATHF analogues as GARFT inhibitors HN H2 N' COOH COOH 5,10-Dideazatetrahydrofolic acid (DDATHF) ,OH OH ‘ 0 P 0 32 nh2 Multisubstrate adduct BW 14761)89 HN H2I COOH COOH N -[4-[2-(2-am ino-3,4-dihydro-4-oxo-7H -pyrrolo[2,3-d]pyrim idin-5-yl)ethyl]benzoyl]- L-glutamic acid (pyrrolopyrimidine analogue of DDATHF) COOH COOH HN' H2 Thiophene analogue (LY254155) COOH COOH HN' H2N' Furan analogue (LY222306) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 99 interactions), in which precise geometry is not critical. Previous studies from this laboratory have shown that both (6R)- and (6S)- diastereom ers of DDATHF are equiactive inhibitors of GARFT. T his lack of stereospecificity in ligand binding was initially surprising. However, previous studies on avian GARFT have dem onstrated t h a t the enzym e utilized (6R )-10-form yltetrahydrofolate as a substrate, but was strongly inhibited by the unnatural diastereom er, (6S)-10- form yltetrahydrofolate, a clear indication that both isomers c o u ld bind to the enzyme even though precise geometry at C-6 was required for catalysis to occur. Given the considerable tolerance of GARFT for alterations in the bridge and phenyl regions and that th e first glutamate side chain of DDATHF does not seem to be crucial for binding, it is no longer surprising for both diastereom ers of DDATHF to be active inhibitors. A series of second generation GARFT inhibitors have been synthesized and evaluated as anti-neoplastic agents. Of p a r tic u la r interest are the thiophene (LY254155) and furan (LY222306) an alo g s of DDATHF (Table 2.3). These two com pounds retain the critical structural features of DDATHF required for GARFT inhibition i.e., 2- amino, 5-deaza, and 8-NH. They differ from DDATHF in th e replacem ent of the phenyl ring (a region shown to be relatively u n im p o rtan t by our structure-activity relationship studies) by a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 thiophene and a furan moiety respectively. Surprisingly, th ese analogs were found to be 25- (thiophene) and 75-fold (furan) m o r e active than (6R)-DDATHF against genetically engineered h u m a n m o n o functional GARFT (Habeck et al., 1994). M olecular m o d e l i n g analysis suggested that the geometry of the 2'-5'-thiophene ring specifically positioned the a-carboxyl of the glutam ate side chain of the antifolate to interact favorably with an arginine just outside t h e binding pocket (Dr Chuan Shih, Lilly Research Laboratories, p e r s o n a l co m m u n icatio n ). This binding model is supported by the fact t h a t the l'-3'-phenyl analog (meta-DDATHF) also bound 40-fold better to single domain human GARFT than did DDATHF (Habeck et al., 1995). Interestingly, this interaction is not apparent from our studies, as deletion of the glutamate side chain did not result in substantial loss in affinity (Table 2.1). One possible explanation is that the p h e n y l carboxyl group in 5 ,1 0 -dideazatetrahydropteroic acid had a s s u m e d the role of the a-carboxyl of the glutamate side chain in fo rm in g interaction with the arginine outside the binding pocket. M oreover, it is tempted to speculate that the increase in potency obtained w ith analogs where there was an increase in spacing/flexibility between t h e heterocycle and glutam ate group (homo-DDATHF, ab e n c y l- DDATHF,n=4, and cyclohexyl-DDATHF), also prom oted th is in te r a c tio n . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 101 The surprising results obtained with the thiophene and fu r a n analogs also signify a ca v ea t in interpreting s tru c tu re -a c ti vity relationship studies. Modification made in one region of the lig a n d often produces an effect that concerns a different portion of t h e molecule. In this case, the increased binding affinity observed w ith the thiophene analog did not appear to be due to the replacement of the l'-4' phenyl by thiophene per se, but rather to a b e tte r positioning of the a-carboxy group of the glutamic acid after su c h modification. It is therefore inappropriate to attribute the difference in affinity to the relative contribution of phenyl versus thiophene to binding unless other parameters such as conform ations a n d m olecular geometry are carefully controlled. The fact that 5 ,1 0 -d id eazatetrah y d ro p tero ic acid r e ta i n e d substantial activity against GARFT (Table 2.1) clearly indicated t h a t the glutamate side chain of DDATHF m o n o g lu ta m a te did not play a critical role in determ ining the overall enzym e-in h ib ito r in te ra c tio n . This argum ent is supported by X-ray diffraction studies on th e ternary complex formed among E. coli GARFT, GAR and 5- d eazatetrah y d ro fo late. The glutamate end of the antifolate was found to be at the enzyme surface and was particularly d is o rd e re d (Almassy et al., 1992). On the contrary, Klein et al. (1995) r e p o r te d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 102 that the glutam ate side chain of the folate portion of t h e m ultisubstrate inhibitor was remarkably well-defined and its s- carboxyl group was tightly associated with the enzyme by h y d r o g e n bond with the amide nitrogen of Ile-91, and by e le c tro sta tic interaction with Arg-64. It is not clear what is the basis of t h e discrepancy between the two reports. Although one can s p e c u la te that the relative orderliness of the glutam ate side chain observed by the latter group is related to the low temperature (-150 °C) e m p l o y e d during data collection. Structural changes resulted in enhanced potency against GARFT - the role of DDATHF polyglutamates in mediating c y to to x ic ity Although the glutamate moiety of DDATHF does not seem to contribute to inhibition of GARFT activity, it is required for t h e conversion of any potential drug to folylpolyglutam ates by FPGS. There has been increasing evidence to suggest that p o l y g l u t a m a t i o n of DDATHF are critical for cytotoxicity induced by this drug. For example, it has been shown that the parent com pound was r a p id l y m etabolized into long chain polyglutam ates in intact tumor cells and tum or cells resistant to MTX by virtue of dim inished FPGS activity were cross-resistant to DDATHF (Pizzorno et al., 1991). Additional support for this concept is furnished by the p r e s e n t Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 103 study. Thus, the p en tag lu tam ate of DDATHF bound to GARFT a b o u t 100-fold more tightly than DDATHF itself (Table 2.2). This is reminiscent of the enhanced binding of the polyglutam ates of the TS inhibitors CB317 and D1694 to TS their target enzyme (Cheng et al., 1985; Siroka et al., 1988; Ward et al., 1992). In contrast, MTX polyglutam ates are only 2-5 fold more potent than MTX itself as inhibitor of DHFR (Kumar et al., 1986; Appleman et al., 1988). T h e 3-dim ensional structure of the complex between GARFT and th e p o ly g lu ta m a ted form of the folate inhibitor would u n d o ubtedly be an invaluable addition to our understanding of the nature of th is interaction. However, the E. coli enzyme might not be an ideal model system for such purpose. This is because, unlike in e u k a ry o tic cells where folylpolyglutam ates exist up the level of o c t a g l u t a m a t e all linked via the y-carboxyl group of the preceding glutam ate, in E. coli only the first two glutam ate residues (di- and tri-glutam ate) are y-linked, whereas residues 4-8 were each linked to the p o l y g l u t a m a t e chain at the a-carboxyl group (Ferone et al., 1986). Thus, one w o u ld not expect to find binding sites for higher chain polyglutam ates in mammalian GARFT to be conserved in the E. coli enzyme. The fact that the n o n -polyglutam atable analogs 5,10- did eaz atetra h y d ro p tero ic acid and 5 ,1 0 -d id e a z a te tr a h y d r o p te r o y l Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 104 aspartic acid were as inhibitory as DDATHF against GARFT and y e t were poor inhibitors of tumor cell growth, suggested that the lack of polyglutam ate formation was responsible for the d is c r e p a n c y between anti-tum or activities and GARFT inhibition exhibited by these analogs. However, the possibility that these c o m p o u n d s translocate across the plasma m em brane poorly could not be r u l e d o u t. It should be noted that although DDATHF polyglutam ates are probably the biologically relevant mediators of drug toxicity in s id e the cells, it is not clear whether more efficiently p o l y g l u t a m a t e d analogs will necessarily be better and more selective a n t i - c a n c e r agents than DDATHF. Nevertheless, should clinical data i n d i c a t e that this is the case, rational design of such agents can be i n i t i a t e d based on the information generated by the present s tu d y . Specifically, the regions of DDATHF that can tolerate s u b s ta n tia l alterations without com prom ising GARFT inhibition (i.e. bridge a n d phenyl ring region) can be explored for modifications to i m p r o v e polyglutamation. To this end, a report on the substrate specificity of mouse and hog liver FPGS for a series of DDATHF analogs h a s recognized a num ber of com pounds that are superior to DDATHF as substrates for FPGS, two of which contain modifications at the 10- position, 10-hydroxymethyl-DDATHF and 10-m ethyl-D D A TH F Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 105 (Habeck et al., 1995). These analogs are expected to be equally potent inhibitors of GARFT as DDATHF, and, in addition, to accumulate in tumor cells more readily than DD A TH F. Several preliminary reports regarding the kinetics of in h ib itio n of GARFT by DDATHF and related antifolates appeared to be perplexing. Pizzorno et al. (1990) have reported that, when th e pentaglutam ate of 10-form yl-tetrahydrofolate was used as th e substrate for GARFT, the kinetically d eterm ined Kj for DDATHF m o noglutam ate was 100-fold higher than that obtained when 10- form yltetrahydrofolate m onoglutam ate was used. Ferone et al. (1990) have reported the same observation with a DDATHF re la ted GRAFT inhibitor. A similar effect was found with inhibition of AICARFT and TS by MTX polyglutamates (Allegra et al., 1985a; Allegra et al., 1985b). These observations are at odds with what th eo ry would predict: if a kinetically determined Kj indeed represent a dissociation constant describing a binding equilibrium betw een inhibitor and enzyme, its value should be in d ependent of th e substrate used. Recent studies in this laboratory by Dr. Sonal Sanghani have refuted these reports (Sanghani and Moran, 1997). When the binding between (6R)-DDATHF and recombinant GARFT was Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 106 studied directly by equilibrium dialysis, the dissociation c o n s t a n t (K d) was found to be 1.8 nM, a value that is =50-fold lower than t h a t estimated by kinetic experiments (Kj = lOOnM) as described in th is report (Table 2.1). In equilibrium dialysis experiments where t h e folate substrate 10-CHO-5,8-dideazafolate was used as an inhibitor of binding of radiolabeled (6R)-DDATHF to GARFT, the Kj was e s t i m a t e d to be 30 nM, a value that is, again, much lower than the k in etically determined K m for this substrate (1.3 - 7 pM) (Sanghani and M o ran , 1997). The discrepancy between the Kd m easured by e q u ilib riu m dialysis and the Kj or Km estimated by kinetic analysis was shown to be due to a limitation of the spectrophotom etric assay used in t h e kinetic experiments which tends to overestimate the K m of the fo late substrate. It was found that using a conventional 1cm light p a t h cuvette system, significant production inhibition by 5 ,8 -d id eaz afo late was occuring. This happened even during the first 10-15 s of t h e reaction where it was previously thought to represent initial velocity condition. When the dependency of the rate of the GARFT re a c tio n on substrate 10-CHO-5,8-dideazafoiate concentration was s t u d i e d spectrophotom etrically using a 10 cm cuvette system to amplify t h e signal generated from product formation and to ensure true in itia l velocity m easurem ent, the K m was found to be 75 nM, c o n s is te n t Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 107 with that measured by equilibrium binding. Using this e n h a n c e d spectro p h o to m etric assay, the Kj of (6R)-DDATHF was determ ined to be 5.6 nM, which is also com patible with the K d m easured by equilibrium dialysis (1.8 nM). Using a highly sensitive HPLC assay where initial velocity condition was m aintained, kinetic an a ly s e s using either mono- or hexaglutamate substrate did not yield d iffe ren t values for the Kj for DDATHF hexaglutam ate (Kj = 0.3 nM) (S a n g h a n i and Moran, 1997), in contradiction with previous reports (Pizzorno et al., 1990; and Ferone et al, 1990). When inhibition of GARFT by various DDATHF analogues was reassessed using the enhanced spectrophotom etric assay, the s a m e conclusion is reached regarding the structural requirem ent for inhibition of GARFT. However, the Kj values are uniformly lower than those reported in the present study, consistent with th e difference in K m of the folate substrate (Sanghani and Moran, 1997). Kinetics of the GARFT reaction revisited Previous kinetic experiments on m am m alian GARFT, based o n the use of dead-end inhibitors, have suggested an ordered s e q u e n tia l pathw ay for the binding of substrates with folate binding p re c e d in g that of GAR (Caperelli, 1985 and Caperalli, 1989). However, th o s e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 108 experim ents were performed at high concentrations of folate substrate relative to the Km reported by Sanghani and Moran (1 9 9 7 ) so that the enzym e binding would be essentially irreversible. U n d e r such conditions, a random order binding m echanism could be masked and would appear to be ordered sequential. Using equilibrium dialysis technique, the binding of (6R)-DDATHF and GAR to G ARFT was found to be independent to each other, an o b s e rv a tio n that is at variance with an ordered sequential m echanism of substrate binding. It is therefore suggested that the kinetics of t h e GARFT reaction are more compatible with a random m echanism (Fig. 2.1) (Sanghani and Moran, 1997). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 109 CHAPTER 3 BIOCHEMICAL CHARACTERIZATION OF MURINE LEUKEMIC L1210 CELLS SELECTED FOR RESISTANCE TO DDATHF INTRODUCTION Drug resistance of tumor cells to cytotoxic agents remains a major obstacle in the treatment of human malignancies (M orrow and Cowan, 1997; Gorlick et al., 1996). An understanding of th e mechanisms of acquired and intrinsic resistance is necessary for optimal use of ch em o th erap y . In theory, drug resistance can be circumvented, on the basis of the underlying m echanism , by ra tio n a l design of ch e m o th e rap e u tic regimen using non-cross-resistant ag en ts or drugs to which the tumor is collaterally sensitive. Drug resistance is a relative term. In cell culture models, it can be m easured quantitatively by com paring the cytotoxic o r growth inhibitory effect of a drug on the resistant cells and that on a reference cell population, generally the wild-type cell line from w h ich the resistant mutants are derived. However, in most clinical settings, resistance can only be estimated qualitatively by examining t u m o r response and host toxicities induced by a chemotherapeutic reg im en . The tumor is deem ed resistant to treatment if an objective response Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 110 cannot be achieved at dosages that would otherwise impart high grade toxicities to the host. The majority of in vitro studies of d ru g resistance have used selection methods based on c o n t i n u o u s exposure of mass culture of tumor cells to increasing c o n c e n t r a t i o n s of the drug or single exposure to very high concentrations of d ru g with or without mutagen pre-treatm ent. There has been so m e skepticism about the relevance of these in vitro models of d ru g resistance to treatment of cancers (Pizzorno et al., 1988). Nevertheless, biochemical analysis of the m echanisms of d ru g resistance has, historically, been a powerful source of i n f o r m a t i o n about which steps in the action of a drug are necessary for cytotoxicity. Moreover, such analyses often reveal unforeseen b u t fu ndam entally important processes in biology. A prime example is the phenomenon of gene amplification in eukaryotic cells d is c o v e re d in MTX-resistant mutant cells containing elevated levels of DHFR (Alt et al., 1978). Another example is the multi-drug resistant (MDR) phenotype mediated by the P-glycoprotein, an e n e r g y - d e p e n d e n t efflux pum p which prevents intracellular accum ulation of an a r ra y of structurally diverse cytotoxic agents (Biedler and Riehm, 1970; Juliano and Ling, 1976; Riordan and Ling, 1985). The diverse biochemical and molecular m echanism s of b o th acquired and intrinsic resistance to MTX have been reviewed i n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 111 Chapter 1 of this thesis. To recapitulate, resistance can occur by (1) impaired m em b ran e transport of MTX (i.e. increased K m a n d / o r decreased V m a x) (Hill et al., 1979; Sirotnak et al., 1981; M cC o rm ick et al., 1981; N ieth am m er and Jackson, 1975; Sirotnak, 1987; S ch u etz et al., 1988); (2) decreased in accum ulation of MTX p o ly g lu ta m a te s as a result of diminished FPGS activity (Pizzorno et al., 1988; Pizzorno et al., 1991); or increased folylpolyglutamyl hydrolase activity (Rhee et al., 1993; Li et al., 1993); (3) elevated target enzyme level as a result of one of the following m echanisms: i) amplification of t h e DHFR gene (most com m on) (Alt et al., 1978; Nunberg et al., 1978, Brown et al., 1981; Biedler et al., 1979)., ii) increased translation of DHFR, a process that is ordinarily inhibited by binding of free DHFR to its mRNA, iii) enhanced transcription of DHFR gene by E2F-like proteins in cells that lack functional retinoblastom a p ro tein ; mutations in D H FR leading either to (4) a decreased affinity for MTX (Jackson et al., 1976); or (5) changes in the kinetics of the e n z y m e showing higher affinity for its substrate (D edhar and Goldie, 1985); and (6) decreased TS activity (Moran et al., 1979; Ayusawa et al., 1981). Since M TX and DDATHF utilize the same mem brane carrier systems for cellular entry, and both are substrates of FPGS a n d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 112 folylpolyglutam yl hydrolase, one would expect a similar pattern of resistance m echanism s to be seen with both agents. Conversely, o n the basis of distinct cellular targets, and the differential role of polyglutamation in mediating their cytotoxicity, i.e., p o ly g l u ta m a t e s derivatives of DDATHF are substantially more potent (> 10-fold) inhibitors against GRAFT than the parent drug (Baldwin et al., 1991; Sanghani and M o ran ,1997; and Chapter 2), whereas that of MTX are only slightly better inhibitors of DHFR (Cheng et al., 1985; Sikora et al., 1988), one would anticipate that the resistant mutants selected using the two drugs to be phenotypically different. To date, several reports of acquired resistance to (6R)-DDATHF have been d escrib ed . Prior to our studies, the causes of resistance have been attributed to decreased drug transport (Matherly et al., 1993), decreased FPGS activity (Pizzorno et al., 1995), and increased fo ly lp o ly g lu ta m y l hydrolase activity (Rhee et al., 1993); all of which result in th e reduction in the steady state level of cellular (6R)-DDATHF polyglutam ates. A distinct pattern is becoming clear: given th e central role of polyglutamation in the action of these newer classes of antifolates, it appears that any biochemical alteration that p re c lu d e s the accum ulation of polyglutam ates of DDATHF would allow t u m o r cell escape from cytotoxicity with these agents. In this report, we furnish additional evidence for this concept. We now describe a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 113 unique m echanism by which murine leukemic L 1210 cells d e v e lo p resistance to (6R)-DDATHF: an expansion of the intracellular fo late pool with consequent blockade of the synthesis of (6R)-DDATHF polyglutam ates. This unexpected m utant phenotype c o n s titu te s evidence for the existence of a feedback mechanism for the control of cellular (anti)folate polyglutamates, presum ably by direct effects o n FPGS. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 14 CHAPTER 3 MATERIALS AND METHODS C h em icals (6R)- , 6S)-, (6R,S)-DDATHF, (6R,S)-DDATHF polyglutamates, ZD- 1694, l-chloro-3,5-dim ethoxytriazene and [^C ](6 R )-D D A T H F were kindly provided by Dr. Chuan Shih of Lilly Research Laboratories (Indianapolis, IN). a,(3-Glycinamide ribonucleotide was provided by Dr. Homer Pearce, also of Lilly Research Laboratories. 10-formyl-5,8 dideazafolate was initially obtained from Dr. Pearce and was subsequently purchased from Dr John Hynes of the M edical University of South Carolina (Columbia, SC). [3',5',7,9-3h ]-(6S)-5- form yltetrahydrofolic acid, [3',5',7,9-3H]-folic acid and [6-^H]- fluorodeoxyuridylate (FdUMP) were purchased from M o rav ek Biochemicals (Brea, CA). [3,4-3H]-L-glutamic acid were obtained fro m either NEN DuPont (Boston, MA) or Am ersham (Arlington Heights, IL). 5-fluorodeoxyuridine (FUdR), MTX, and folinic acid were from S igm a Chemical C om pany (St. Louis, MO). [^H]folic acid and [3h ](6S)-5- form yltetrahydrofolic acid were purified by reverse phase hplc using a linear gradient of methanol in 0.03 M sodium acetate (4 to 12 % Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 1 5 methanol over 12 min. for [^HJfolic acid; 3 to 15 % methanol over 15 min. for [3H ](6S)-5-form yltetrahydrofolic acid) and were stored at - 20 °C for no more than 2 weeks prior to experim ents. [^H l-FdU M P was purified by DEAE column c h ro m a to g rap h y as p re v io u s ly described (Moran et al., 1979). (6S)-tetrahydrofolate and 10- form y ltetrah y d ro fo late were chemically synthesized (Moran et al., 1976). Fetal bovine sera were purchased from Gemini Bio-Products, Inc. (Calabasas, CA). Synthesis of [3H](6R)-DD ATHF [3H](6R)-DDATHF was synthesized based on a protocol su g g ested by Dr. Chuan Shih of Lilly Research Laboratories, which involved t h e coupling of (6 R )-5,10-dideazatetrahydropteroic acid with L- [^H]glutamic acid diethyl ester in the presence of the coupling a g e n t l-chloro-3,5-dim ethoxytriazene. (6 R )- 5 ,1 0 -d id e a z a te tr a h y d ro p te r o ic acid was prepared by hydrolyzing (6R)-DDATHF in 6N HC1 at 100 °C in a sealed vessel for 4h, followed by purification using DEAE cellulose column ch ro m ato g rap h y . The diethyl ester of L -[3,4-3H ]]G lutam ic acid was synthesized by reacting 18 pmol of lyophilized Lr [^H jjglutam ic acid (54Ci/mm ol) with 0.1M ethyl p - to lu e n e s u lf o n a te in anhydrous ethanol under reflux for 24h; the product was t h e n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 16 purified on a silica gel column eluted using 5% M e O H /c h lo ro fo rm , and the pooled fractions were dried under a stream of N2- In a typical coupling reaction, 3.4 pmol of (6R)-5,10- dideazatetrahydropteroic acid was added to an equimolar amount of l-ch loro-3,5-dim ethoxytriazene in 30 pi of anhydrous DMSO m ix e d with 2 pi of 4-methyl morpholine and allowed to react for 30 min. a t room tem perature (RT). This reaction mixture was then added to the dried L-[3,4-3H]glutamic acid diethyl ester and reaction was allowed to proceed for 6h at RT. The diethyl ester of [3h ](6R)- D D A T H F formed was hydrolyzed in IN NaOH for 6h at RT to give th e final product, [^H](6R)-DDATHF, which was first purified by ch ro m a to g rap h y first on a DEAE-cellulose column, then on a 10 x 0.46 cm Luna 3 pm C-18 (Phenomenex, Torrance, CA) eluted with a m ultiphase gradient of methanol in aqueous t e t r a b u t y l a m m o n i u m hydrogen sulfate (Pic A reagent, Waters Associates) run at 0.6 ml/min. To generate this gradient, mobile phase m e t h a n o l concentration was initially 27 %, then was increased to 35 % in a linear gradient over 10 min; subsequently, a less steep linear g r a d ie n t was initiated which reached a methanol concentration of 42 % afte r an additional 15 min; the methanol concentration was then held a t 42 % for the next 10 min. Pooled hplc fractions containing 3h_(6R)_ Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 17 DDATHF were dried under nitrogen. The te tra b u ty lam m o n iu m ion present in the mobile phase was subsequently removed by chrom atography on a 1 ml Dowex 50 column in sodium form; product was eluted with PBS. The purity and stability of the final product was determ ined by co-chrom atography with authentic (6R)- DDATHF standard and scintillation counting. A typical yield obtained using this protocol was 5-10%, relative to the r a d io a c tiv e glutamic acid precursor. The [3h ](6R)-DDATHF synthesized was found to be stable (radioactive impurities <2% detected by hplc) o v er two weeks when stored at -20 °C in 33% EtOH/66% PBS. Enzymatic synthesis of radiolabelled DDATHF tetr a g lu ta m a te (6R )-D D ATH F-[3,4-3 h ]G1u3 was synthesized by incubating 0 .145 mM 3 H-glutamic acid (200 f i d ) at 37 °C with 200 fiM (GR)-DDATHF, 20 mM 2-m ercaptoethanol, 10 mM ATP, 20 mM MgCl2 and 30 m M KC1, and 2.8 fig of recom binant cytosolic human fo ly lp o ly g lu ta m a te synthetase (Sanghani et al., 1999) in a total volume of 30 fi\ of 0.2 M Tris, pH 9.0 containing 0.2 mg/ml of bovine serum albumin. T h e reaction was stopped after 2 hr by heating at 100 °C for 3 min. T h e major product was the tetraglutam ate derivative, which was identified by cochrom atography with authentic DDATHF Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 118 polyglutamate standards on a reverse-phase, paired-ion hplc c o l u m n . Product was purified, first on a Sep-Pak C-18 column (W aters Associates, Milford, MA), then by ion-pair reverse phase hplc as described above for the purification of [3H](6R)-DDATHF. Pooled h p lc fractions containing [^H](6R)-DDATHF tetrag lu tam ate were d rie d under vacuum, dissolved in 100 pi water, and passed through a 1 m 1 column of AG 50W-x8 (Bio-Rad Laboratories, Hercules, CA) to re m o v e the ion-pairing reagent; the product was concentrated and stored at - 20 °C in 20 % ethanol. The amount of tetrag lu tam ate product in a n hplc run was quantitated against the area of a known amount of m o n o g lu ta m a te standard. Cell Culture Wild type L I 210 cells were originally obtained from Dr. George Moore, formerly of Roswell Park Memorial Institute (Buffalo, NY). Except where noted, both cell lines were routinely cultured in RPMI 1640 m edium containing 2.3 pM folic acid su p p lem en ted with 10% dialyzed fetal bovine serum at 37 °C in an atmosphere containing 5% C 0 2 - For studies on folate-depleted cultures, cells were grown in folate-free m edium containing 5.6 pM thym idine and 32 pM hypoxanthine for at least a week prior to use. In experiments w here Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 119 folinic acid was used as the folate source, cells were adapted to growth in 60 nM (6R ,S)-5-form yltetrahydrofolate (folinic acid). Cell density was quantitated electronically using a Coulter counter, M odel Z p (Coulter Electronics, Inc., Hialeah, FL). Stock cultures were consistently found free of M y c o p la s m a upon screening with the Gen- Probe detection system (San Diego, CA). Selection of DDATHF-resistant L1210 cells Drug resistant variants of L1210 cells were selected by c o n t i n u o u s exposure of parental cells to (6R,S)-DDATHF at an initial concentration of 0.05 pM (=1.7x I C 50)- When the d r u g - tr e a te d culture had achieved a growth rate com parable to that of u n t r e a t e d wild-type, the concentration of (6R,S)-DDATHF was escalated to 0.1 pM, then to 0.5 pM, then to 3 pM. C ontinuous passage of cells resistant to 3 pM (6R,S)-DDATHF for an additional five months in 10 pM drug did not allow the emergence of a phenotype which c o u ld grow rapidly at that concentration of selective agent. Three c lo n al sublines, designated L1210/D0.5, L1210/D3 and L1210/D10, w h ich grew in 0.5, 3, and 10 pM (6R,S)-DDATHF, respectively, were established from the mass culture at various stages of the selection process by soft agar cloning. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 120 Growth inhibition studies Exponentially growing cells were transferred into 24-well c u ltu re plates at an initial density of ~2 x 104 cells/ml. Each well c o n t a i n e d 1.35 ml of cell suspension and 150 ml of either PBS or k n o w n concentration of the indicated drug dissolved in PBS. Cell d e n s ity was quantitated electronically after 72 h. For short term exposure to MTX, cells were incubated with drug for 6 h, pelleted by centrifugation, washed with pre-warmed PBS and resuspended in drug-free medium. Cell density was then determ ined after a total of 72 h. Cell proliferation during the entire incubation period was expressed as the ratio between final and initial densities (N /N 0 ). I C 50 values were extrapolated as the drug concentrations required to inhibit growth by 50% com pared to untreated control cells u n d e r identical conditions. Purification of GARFT for kinetic analysis Purification of GARFT was purified to e le tro p h o re tic hom ogeneity by affinity chrom atography as described in Chapter 2 in this thesis. In short, exponentially growing cells were harvested by centrifugation, washed twice with ice cold PBS, resuspended in low salt buffer (10 mM sodium phosphate, pH 7.5, containing 0.1% Triton X-100 and 10 mM (3-ME) and disrupted by sonication on ice. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 121 Extracts were then centrifuged at 165,000 x g for 30 min at 4° C a n d the supernatant was brought to 30% and then to 70% saturation by (NH 4 ) 2 S 0 4 on ice. The 30-70% (NH 4 )2 S 0 4 precipitates was d i a ly z e d overnight against low salt buffer and then applied to a 3 ml affin ity column of Sepharose 4B to which 10-formyl-5,8-dideazafolic acid was covalently attached via ethylene linkages. The column was w a s h e d with 50 ml of low salt buffer followed by a overnight wash with 3 0 0 ml of high salt buffer (0.2 M sodium phosphate, pH 7.5, 0.5 M KC1, 0.1% Triton X-100 and 10 mM _-ME). GARFT was then eluted with a 3 mM solution of 10-formyl-5,8-dideazafolic acid. Enzyme Assays GARFT assay GARFT activity and kinetic analysis were performed as d e s c rib e d previously (Chapter 2 and Smith et al., 1981b). For comparison of GARFT activity between cell lines, enzyme assays were performed o n high speed supernatant fraction prepared by centrifugation at full speed (=200,000 x g) in a Beckman Airfuge (Fullerton, CA) for 30 m i n or in a Beckman Ti50 rotor at 160,000 x g for 1 h. Kinetic p a r a m e t e r s were obtained from Lineweaver-Burke analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 122 FPGS assay Cytosolic FPGS activity was m easured using a m ic ro a ss a y developed previously in our laboratory (Antonsson et al., 1990). Briefly, exponentially growing cells were harvested, pelleted cells were suspended in 200 mM Tris-HCl buffer at pH 8.5, and disrupted by sonication on ice. Five J of the high speed supernatant o b t a i n e d after full speed centrifugation in a Beckm an Airfuge for 30 min was incubated with 10 _M (6S)-tetrahydrofolate, 5 mM ATP, 10 m M MgC12. 30 m M KC1, 1 mM L-[3H]glutamic acid and 200 mM Tris-HCl, pH 8.5 in a total volume of 10 J at 37 °C for 30 min. T h e [3 H ]tetrah y d ro p te ro y ld ig lu tam a te formed was then q u a n t i t a t i v e l y trapped in a ternary covalent complex in the form of [3H ]5,10- m e th y le n e te tr a h y d ro p te ro y ld ig lu ta m a te by adding 100 J of solution containing excess L. casei thym idylate synthase TS (1 _ M ). FdUMP (2 _M), and formaldehyde (15 mM), in 30 mM N a2H P04, pH 7.2, 8 m M m ercaptoethanol and 0.1 mg/ml bovine s e r u m albumin. The incubation was continued for an additional 30 min a t 37 °C. The reaction mixture was then loaded onto a Sephadex G 50 gel-filtration column and the ternary complex was separated fro m the unin co rp o rated L-[3H]glutamic acid upon elution by centrifugation. The eluate was then processed for scin tilla tio n c o u n tin g . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 123 Conjugase assay y -G lutam ylcarboxypeptidase (conjugase, fo ly lp o ly g lu ta m y l hydrolase) activity was measured by the release of glutamic a c id from (6R)-DDATHF tetraglutam ate, which was tritium -labeled in th e second through fourth side chain glutam ate moieties. Briefly, 8 x 107 cells were harvested by centrifugation, suspended in 1 ml of 5 0 mM Tris-acetate buffer, pH 6.0, containing 50 mM _-ME, and the cells were broken by 3 x 20 strokes of a hand-held Dounce h o m o g e n iz e r. The lysate was centrifuged for 20 min at 14,000 rpm at 4 (C in a microfuge, and the supernatant was used for assays. Protein (0-30 fig) was incubated with 100 piM. (6R)-DDATHF-[3,4-3H]Glu3 (0.08 jttCi/assay) in lysate buffer for up to 30 min. and the reactions were stopped by the addition of 500 jtl of a suspension of a c ti v a te d charcoal in 10 mM glutamate, 10 mM _-ME and 150 mM KH2P04, p H 5.0. The charcoal had been pretreated with BSA and Dextran T-70 (Moran and Coleman. 1984). The reaction mixes were centrifuged in a microfuge for 5 min and the supernatant added to s c in tilla tio n fluid for determination of radioactivity. Purification of FPGS FPGS was partially purified from mouse liver and L I 210 cells as previously reported (Moran and Colman, 1983). DBA/2 female m ice Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 124 (Simonsen Laboratories) were sacrificed by cervical dislocation a n d the livers were perfused in situ through the hepatic vein with an ice- cold solution of 50 mM Hepes pH 7.4 containing 0.25 M sucrose. Livers were removed, weighed and minced in two volumes of th e above buffer containing 50 mM _-ME and were hom ogenized with 10 strokes of a motor driven pestle. The hom ogenates were th e n centrifuged at 160,000 x g for 1 h at 4 °C. For L1210 and L1210/D 3 cells, 5-10 g of log-phase cells were harvested by c e n trifu g atio n , resuspended in 2 volumes of the same h om ogenization buffer, a n d disrupted by three 10-s bursts of sonication on ice. High speed cytosolic fractions were prepared similarly by centrifugation. T he high speed supernatants were brought to 30% saturation by the slow addition of 172 g/liter of (NH4)2S04. The precipitate was pelleted by centrifugation at = 1000 x g for 20 min in a Sorvall centrifuge. T h e pellet was dissolved in a minimal volume of 50 mM Tris, pH 7.8 containing 50 mM _-ME and was desalted on a 0.62 cm2 x 38 c m column of Sephadex G-25 prior to kinetic experiments. Kinetic analysis on FPGS partially purified from mouse liver and L1210 cells was performed using the standard FPGS ch arco al assay (Moran and Colman, 1984). Briefly, desalted enzyme was incubated with various concentrations of (6R)-DDATHF or aminopterin (AMT), 1 mM L-[3H]gIutamic acid (4m C i/m m ol), 5 m M Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 125 ATP, 10 mM MgC12, 30 mM KC1, 20 mM _-thiog!ycerol, and 200 m M Tris-HCl, pH 8.6, in a total volume of 1 ml for 1 h at 37 °C. T h e radiolabelled diglutamyl product was isolated by adsorption o n to activated charcoal. Following one 5-ml and three 10-ml washes of the charcoal with 10 mM glutamate containing 10 m M _-ME, th e product was eluted with 1.5 ml of ethanolic am m onia (3 M N H 40H in 60% ethanol). Radioactivity was determ ined after the addition of 1.5 ml of H 2 0 and 15 ml of scintillaton cocktail. Under these conditions, reaction rates approximated initial velocity and s u b stra te consum ption was less than 20% even at the lowest s u b stra te concentration. Kinetic parameters were obtained from Lineweaver- Burke analysis. Uptake of [14C ]D D A T H F A limited am ount of [14C](6R)-DDATHF of m oderate specific activity (=4.4 Ci/mol) was provided by Dr Shih of Lilly Research Laboratories for some preliminary studies on the time course of d ru g uptake in L1210 and L1210/D3 cells. [14C](6R)-DDATHF was first purified by reverse phase hplc as described above for the p u rific a tio n of [3H](6R)-DDATHF, it was then passed through a AG 50W-x8 c a tio n exchange column to remove the ion-pair reagent. Log-phase cells a t Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 126 = 4 x 1 0 5 cells/ml were incubated with 2 pM [14c](6R)-DDATHF a n d 100 pM h y p oxanthine for various time periods. Duplicate s a m p le s were used for each time point. At the end of each incubation p e rio d , 1 ml of cell suspension was taken out for cell count, the r e m a i n in g cells were harvested by centrifugation, and pellets were washed twice with ice-cold PBS containing 5% FCS. Cell pellets were then d is s o lv e d in IN NaOH, neutralized with HC1 and processed for s c in tilla tio n c o u n tin g . Transport Studies Exponentially growing cells were harvested by c e n tr ifu g a tio n , washed once with pre-warmed PBS containing 5% FCS and once w ith transport buffer (107 mM NaCl, 20 mM Tris-HCl, 26.2 mM N aH C 03, 5.3 mM KC1, 1.9 mM C a C b , 1 mM M g C b and 7 mM glucose, pH 7.4 at 37 °C). All studies were performed at 37 °C unless o th erw ise stated. For studies on influx rate and total uptake of folate, cells were resuspended in 0.25 ml transport buffer at a density of =10^ cells/ml in 15-ml centrifuge tubes. Transport was initiated by forceful addition of 0.25 ml radiolabelled folates in p r e - w a r m e d transport buffer and quenched at indicated time points by adding 10 ml ice-cold PBS containing 5% FCS. Cells were then washed tw ice Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 127 with serum-containing ice-cold PBS. Prior to the final ce n trifu g a tio n , a 1-ml aliquot of cell suspension was taken from each tube for cell number determ in atio n using a Coulter counter. The final pellets were dissolved in 0.5 ml 1 N NaOH, neutralized with c o n c e n t r a t e d HC1, and counted for radioactivity by scintillation method. For m easurem ents of folic acid transport, cells were pretreated with 1 0 pM trimetrexate for 10 min at 37 °C in order to block c e llu la r metabolism of folic acid by DHFR. For studies on (6R)-DDATHF efflux, cells were pre-loaded with radiolabelled drug for 2 0 min, w a s h e d twice with ice-cold PBS, resuspended in drug-free buffer at 37 °C, a n d aliquots of cell suspension were withdrawn at various time p o in ts and processed for scintillation counting. Analysis of Intracellular Accumulation of (6R)-DDATHF Polyglutamates by hplc Log-phase cells grown in either folate-containing or folate-free media were treated with various concentration of [3h](6R)-DDATHF for 16h in the presence of 5.6 pM thym idine and 32 pM hypoxanthine. For studies on folate-depleted cells, the culture was passaged in RPMI 1640 medium formulated without folic acid a n d containing 10% dFCS, 5.6 pM thymidine and 32 pM hypoxanthine for 6 days prior to experiments. After incubation with ra d io la b e lle d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 128 drug, cells were harvested by centrifugation, washed twice with ice- cold PBS containing 5% FCS. A known number of cells were resuspended in 100 pi of 5 mM te tra b u ty la m m o n iu m h y d ro g e n sulfate (Pic A reagent, Waters Instruments, Millipore, MA). Cells were disrupted by 8-10 one-min pulses of sonication using a Heat System s- Ultrasonics sonic device with a cup-horn attachm ent, the lysed cells were then boiled for 3 min. Cell extracts were filtered through a microcon 10 filter (Amicon Corp, Beverly, MA) and an aliquot was analyzed by hplc. The percentage recovery after the filtration ste p was noted for each sample and any loss was corrected in s u b s e q u e n t calculation of intracellular drug amount. [3H](6R)-DDATHF polyglutamates were analyzed by an ion-pair reversed-phase m e t h o d using a 100 x 3.2 cm C i 8 column of 3-micron particle size (A p p lie d Biosystems, CA). After injection, the column was eluted a t 0.6m l/m in with a linear gradient from 27% to 35% methanol in 5mM te tra b u ty la m m o n iu m hydrogen sulfate over 10 min, follow ed by another linear gradient to 42% m ethanol over 15 min. F ractio n s (2 0 0 pi) were collected directly into scintillation vials a n d radioactivity was determined by scintillation counting. The i d e n tity of labeled peaks was determined from the retention times of c o injected authentic standards of D D A T H F polyglutamates. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 129 Determination of Folate Cofactor Pools For measurement of total cellular folates, cells were cultured i n RPMI 1640 m edium formulated without folate but contained e ith e r 2.0 pM pH Jfolic acid or 60 nM folinic acid ((6R,S)-5- form y ltetrah y d ro fo late) spiked with [3h ](6S)-5- formyltetrahydrofolate. Following one week of exponential growth in radiolabeled folate, the specific activity of intracellular cofactors becomes equivalent to that in the medium, and folate pool size c a n be estimated by the level of intracellular radioactivity (Moran et al., 1976). Cells were pelleted by centrifugation and washed twice w ith ice cold PBS. After the second wash, an aliquot of the cell su sp en sio n was counted electronically. The remaining suspension was p e lle te d and lysed in 0.5 ml of IN NaOH. The lysate was then n e u tra liz e d with HC1 and the radioactivity was determ ined by scin tilla tio n counting. A known amount of tritium -labeled com pound was counted under the same conditions to allow conversion from cpm to pmol of folates. The level of 10-CHO-H4PteGlun was quantitated using a variation of the ternary complex formation assay which based on th e entrapment of 5,1 0 -C H 2 -H 4 P teG lu n by excess TS and [^HJFdUMP in to a covalent ternary complex (Kesavan et al., 1986; Keyomarsi a n d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 130 Moran, 1988; Schimitz et al., 1994). The content of H 4 P te G lu n a n d 5 ,1 0 -C H 2 -H 4 P teG lu n was determ ined by this method. The c e llu la r levels of 10-CH O -H 4PteG lun was then determ ined indirectly by q uantitating the a d d itio n a l H 4 P te G lu n formed in the presence of excess GARFT and glycinamide ribonucleotide. Briefly, pellets containing 4 x 10^ cells were rapidly resuspended in 200 pi of b o iling 10 m M sodium phosphate buffer, pH 7.5, containing 0.1% T rito n - X100, 1% P-ME and 1% freshly-prepared sodium ascorbate. After boiling for 3 min, the disrupted cell suspensions were quickly b r o u g h t to 0 °C, and the supernatants were collected following c e n tr ifu g a tio n at 14,000 x g for 2 min at 4 °C. The pellets were extracted w ith another 2 0 0 pi of the above solution and the two supernatants were combined. Five or twenty-five pi of the extracts were added either to a 125 pi solution containing 10 mlU of purified L. ca sei t h y m i d y l a t e syntase, 0.16 pM [^HjFdUMP , and 20 pM formaldehyde in 10 m M sodium phosphate, pH 7.4 containing 1% P-ME and 0.2 m g / m l bovine serum albumin in a total volume of 2 0 0 pi, or to a 2 0 0 pi aliquot of this same solution containing, in addition, lm lU of recom binant mouse GARFT and 1 m M glycinamide ribonucleotide . The reaction mixtures were incubated at 30 °C for 2 h. and were t h e n boiled for 10 min to denature the ternary complex formed. Protein Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 131 was precipitated by the addition of 4 ml of 8 % TCA so lu tio n containing 200 pi of 1% bovine serum albumin carrier at 0 °C. Following centrifugation of the precipitates and aspiration of th e supernatant, the pellets were dissolved in 100 pi of 1 N NaOH, re precipitated with 4 ml of 8% TCA and centrifuged. The final pellets were then dissolved in 100 pi 1 N NaOH, neutralized with 750 pi of 0.1 N HC1, 0.2 M KC1 and quantitated for radioactivity by s c in tilla tio n counting. Standard curves were generated in each experim ent using known amount of ( 6 S)-H 4 PteGlu and 10-CHO-H4PteGlu to allow conversion of cpm to pmol of intracellular folates. Northern analysis Total RNA from L I 210 and L1210/D3 cells were isolated using the Triazol reagent (Gibco-BRL, Gaitherburg, MD), denatured w ith glyoxal, and fractionated by electrophoresis on a 1.2 % agarose gel in 10 mM sodium phosphate buffer, pH 7.0. The RNA was t h e n transferred onto nylon m em brane (Biotrans, 1CN, Irvine, CA) by capillary action in 20 x SSC and immobilized by UV crosslinking in a Stratalinker apparatus (Stratagene, CA). The blot was th e n hybridized with a cDNA probe of either a 1.7 kb fragm ent of m o u s e FPGS cDNA (Spinella et al., 1996), or a 800 bp fragment c o rre s p o n d in g to the glycinamide ribonucleotide synthetase dom ain of th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 132 trifunctional GARFT cDNA (probe pQ 0.9; Kan et al., 1993) at 65 °C overnight. Blots were washed to a final stringency of 0.2 xS SC , 0.5% SDS at 65 °C and then exposed to film. A hum an GADPH probe was included for normalization of RNA loading. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 133 CHAPTER 3 RESULTS Development of L1210 sublines resistant to DDATHF Clonal sublines of leukemic L1210 cells were selected by continuous exposure of wild-type cells to stepwise increments of (6R,S)-DDATHF concentration. Whereas the parental L I 210 cells were half-maximally inhibited by 2.8 x 1 0 '8 M (6R,S)-DDATHF, L 1 2 1 0 /D 0 .5 and L1210/D3 cells proliferate in 0.5 and 3 pM drug, respectively, with no detectable change in growth rate (double time of 10-12 hours) (Fig. 3.1). Despite continuous passage of cells resistant to 3 pM (6R,S)-DDATHF for several additional months in 10 pM drug, m o re highly resistant cells did not emerge, although a cell line with a slower growth rate (doubling time 18 hours), the L1210/D10, was isolated (Fig. 3.1). The highly resistant line, L1210/D3 was selected for further characterization. L1210/D3 cells were found to be equally resistant to b o t h diastereom ers of DDATHF differing in chirality at the C6 p o s itio n (Table 3.1). This lack of stereospecificity was not surprising since Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 134 4 0 - 2 0 - 1 0 - 0.001 0.0001 0.01 0.1 10 100 D D A TH F, fiM Fig. 3.1 Inhibition of the growth of L1210 cell sublines by (6R,S)-DDATHF. After continuous exposure of wild-type L1210 cells to stepw ise increments in concentration of (6R,S)-DDATHF, three re s is ta n t sublines were derived and cloned which grew in the presence of 0.5 pM (open triangles), 3 pM (filled triangles), and 10 pM (filled circles), respectively (see text). One and a half-ml of log phase wild ty p e (open circles) and resistant cells were exposed to the i n d i c a t e d concentrations of (6R,S)-DDATHF for a 72-h period. Cells were counted electronically at the end and cell densities were expressed as multiples (N /N o ) of the initial cell density when drug exposure was initiated (N /N q ). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3.1 Sensitivity o f wild-type a n d D D A T H F -resistan t L1210 cells to growth inhibition by in h ib ito rs o f f o la te -d e p e n d e n t e n zy m e s IC50 v alu es (n M ) = i= Drug L1210 (n) L1210/D3 (n) Ratio (mutant/wt) (6R,S)-DDATHF 28 + 10 (14) 5 , 600 ± 960 (3) 200 (6R)-DDATHF 23 + 3 (6 ) 9, 000 + 1 , 000 (3) 400 (6S)-DDATHF 29 ± 4 (7) 7 , 000 (1 ) 240 (6R)-DDATHF (folic ~> § folinic ) 220 + 80 (3) 71 + 25 (3) 0 .32 (6R)-DDATHF (folinic — > folicS 21 + 4 (2 ) 11,000 + 1, 30 0 (2 ) 510 D 1 6 9 4 11 + 12 (3) 160 + 140 (3) 15 M TX 72 h exposure 8 . 5 ± 0 . 5 (3) 17 ± 4 (2 ) 1 . 9 M TX 6h pulsed exposure^ 490 + 31 0 (2 ) 4 , 100 + 2 , 900 (2 ) 8 . 4 FUdR 0 .5 1 + 0 . 02 (3) 0.20 + 0 . 00 (3) 0.3 9 u> 136 Table 3.1 (cont’d) * Growth inhibition was measured after continuous exposure of cells to the listed drugs for a 72 h growth p erio d , except electronically. Values are represented as means ± SD or 0.5 x range (for n =2) obtained from n independent ex p e rim e n ts where noted, and c u ltu re densities were determ ined electronically. Values are represented as means ± SD or 0.5 x range (for n =2) obtained from n i n d e p e n d e n t ex p e rim e n ts . ^ These experiments were performed in RPMI 1640 m e d i u m -8 formulated without folic acid but containing 6.0 x 10" M folinic acid. Cells were grown in this medium for a week prior to grow th inhibition studies. • 5 - After two weeks of growth in folinic acid medium, cells were transferred back to standard medium containing 2.0 _M folic acid and used for growth inhibition studies in this medium. II In these experiments, cells were exposed to MTX for 6 h, w ashed, and then resuspended in drug-free medium and culture densities were measured after a total incubation of 72 h. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 137 previous studies have indicated that the two diastereomers d is p la y e d remarkably similar activities as inhibitors of GARFT and as su b stra te s for FPGS and RFC (Moran et al., 1989; Pizzorno et al., 1993). T h e phenotype of these variant cells was very unusual in that resistan ce to (6R)-DDATHF was observed only when cells were grown in m e d i u m containing folic acid, the medium in which they were derived. W h e n cells were grown in folate-free medium su p p lem en ted with 60 n M folinic a c id 2, the L1210/D3 subline was found to be, p a ra d o x ic a lly , more sensitive to drug than the wild-type L1210. However, even after a prolonged culture period (> 3 weeks) in folinic acid, resistance to (6R)-DDATHF resumed im mediately when cells were transferred b ack to folic acid containing medium (Table 3.1). Although th e manifestation of drug resistance was dependent on the specific folate source, L1210/D3 cells remained resistant to (6R)-DDATHF for at least six months after continued growth in the absence of drug. Thus, we concluded from the stability of this phenotype in the absence of selection, that one or more mutations had been fixed in gene(s) critical to the action of (6R)-DDATHF and that the re s is ta n t 2 T his co n cen tratio n o f folinic acid i.e. (6 R ,S )-5 -fo rm y ltetrah y d ro fo la te was equivalent to the standard level o f folic acid (2.3 pM ) ' n R PM I 1640 m edium , both in term s o f support of grow th o f L 1210 cells and total cellular pools o f folates form ed by these cells (Keyom arsi and M oran, 1986). The proliferation o f L 1 2 1 0 /D 3 cells in folinic acid m edium was also identical to that o f parent L 1 210 cells. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 138 phenotype is conditional to the folate source on which the cells were grown. Because this peculiarity suggested a previously u n d e f i n e d mode of resistance to antifolates, the m echanism underlying th is resistance was determined. Cross-resistance and collateral sensitivity of L1210/D3 cells to related antimetabolites. The sensitivities of L1210 and L1210/D3 cells to o t h e r antim etabolites inhibitory to folate-dependent enzymes were com pared (Table 3.1). With continuous exposure to drug during a 72 h period of growth, L1210/D3 cells showed a 15-fold resistance to the thym idylate synthase inhibitor ZD-1694, but only a 2-fold resistance to the dihydrofolate reductase inhibitor MTX. R esistance to MTX was substantially higher (8-fold) when drug exposure was limited to the first six hours of a 72 h culture period. This s ch e d u le - d ependent pattern of resistance to MTX was reminiscent of th e characteristics of tumor cells with an impaired capacity for po ly g lu tam atio n of this drug (Pizzorno et al., 1988). It was interesting to note that, although deficiency in polyglutam ation h a s been implicated as a mechanism of resistance to FUdR (Aschele et al., 1992), L1210/D3 cells were, in fact, collaterally more sensitive to FUdR com pared with parent L1210 cells. Since L1210/D3 cells Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 139 showed cross-resistance to antim etabolites targeted at d iffe ren t folate-dependent enzymes, we concluded that one or m o r e biochemical processes, which are both critical and common to t h e action of all three classes of antifolates, have been altered in th e s e cells. Characteristics of glycinamide ribonucleotide formyltransferase (GARFT) expressed by L1210/D3 cells. Overexpression of the target enzyme GARFT or alteration in affinity for DDATHF are two potential m echanism s of resistance to this drug. Hence, GARFT activity in cell-free extracts and kinetics of inhibition by (6R)-DDATHF on isolated enzyme were c o m p a r e d between wild-type and resistant cells. The specific activities of th is enzyme in cytosolic preparations from L1210 and L1210/D3 cells were not appreciably different (Table 3.2). Northern blots a n a ly s is also d em o n strated that the size and abundance of message for b o t h the trifunctional GARFT and m onofunctional g ly c in a m id e ribonucleotide synthetase were identical between resistant and w ild- type L I 210 cells (Fig 3.2). The kinetic characteristics of GARFT purified to near homogeneity from L1210/D3 cells were the same as those found in enzyme isolated form L1210 cells. Analysis of t h e kinetics of inhibition of GARFT derived from L1210/D3 indicated t h a t R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 140 the Kj values of both (6R)-DDATHF and (6R,S)-DDATHF- p en tag lu tam ate were similar to those obtained previously w ith enzyme purified from wild-type cells when a K m value of 75 nM for 10-formyl-5,8-dideazafrolate was used R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 141 Table 3.2 Activity and Kinetics o f GARFT, FPGS and conjugase in L I210 and L1210/D3 cells L1210 (n) L 1210/D 3 (n) GARFT Activity (nmol/min-mg protein) 2.8 ± 1.0 (3) 2.2 + 0.2 (3) Kj values (6R)-DDATHF (nM) 5 . 6 * 5 . 2 + 0.1 (2 ) (6R)-DDATHF(G1u)5 (nM) 0.11 * (6R,S)-DDATHF(GIu)5 (nM) 0.025 ± 0.01 (6 ) 0.042 + 0.01 (2 ) FPGS Activity (pmol/hr-mg protein) 350 ± 160 (4) 540 + 260 (4) Aminopterin: Km (fi M) 22 ± 3 (5) 24 + 5 (3) (6S)-DDATHF Km (MM) 1.1 ± 0.01 (2 ) 1 .3 + 0.1 (2 ) relative Vmax 0.12 ± 0.02 (2 ) 0.10 + 0.002 (2 ) relative k’ 2.0 ± 0.1 (2 ) 1 . 9 + 0 .5 (2 ) (6R)-DDATHF Km in M) 1 .3 ± 0.6 (5) 1 . 4 + 0 .5 (3) relative Vmax 0.14 ± 0.05 (5) 0.13 + 0.02 (3) relative k' 2.6 ± 1.0 (5) 2.2 + 0.6 (3) C onjugase Activity (nmol/hr-mg protein) § <0 . 13 § <0 .13 § R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 142 Table 3.2 (cont’d) The folate substrates used for the GARFT and FPGS a c tiv ity m easurem ents were 10-form yl-5,8-dideazafolate and (6S)- tetrahydrofolate, respectively. Kinetic analysis of GARFT i n h ib itio n was perform ed on GARFT purified by affinity ch ro m ato g rap h y f ro m both cell lines. Kj values were obtained from Dixon plots using K m values of 75 nM (Sanghani and Moran, 1997). Values are expressed as means ± SD or 0.5 x range (for n = 2) from n experiments. T h e kinetic param eters for FPGS expressed by L1210 and L1210/D3 cells were d eterm ined on enzyme concentrated and purified = 20-fold by (N H 4)2S 04 precipitation (Moran and Coleman, 1982); the Vmax a n d first order rate constants (k') for DDATHF isomers are ex p ressed relative to the values for am inopterin measured in each e x p e r i m e n t . DDATHF p en tag lu tam ate was available as a mixture of (6R)- a n d (6 S )-d ia ste reo m ers. * From (Sanghani and Moran, 1997). § The limits of detection in these assays. In the same e x p e rim e n ts , protein from CEM cells cleaved glutamic acid from DDATHF tetraglutamate at 4.9 ± 0.2 nmol/hr-mg protein. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 143 Fig. 3.2 Northern analysis of expression of GARFT (A) and FPGS (B) messages in wild-type and DDATHF-resistant L1210 cells. Poly(A+)-selected RNA (2_g/lane) was separated on a 1.2% agarose gel under denaturing condition, transferred to a nylon m e m b r a n e , and the m em brane was hybridized successively with cDNA p ro b e corresponding to the upstream domain of the mouse tr ifu n c tio n a l GARFT (A), cDNA probe corresponding to a 1.4 kb d o w n s t r e a m fragment of mouse FPGS (B), and a probe for the h u m a n g ly cerald eh y d e-3 -p h o sp h ate dehydrogenase (DAPDH) gene (C). T h e 1.7- and 3.0-kb transcripts hybridized to the GARFT probe re p re s e n t the m o n o fu n ctio n al glycinamide ribonucleotide synthetase a n d trifunctional G AR FT transcripts, respectively (Kan and Moran, 1995). Difference in RNA loading among lanes was corrected by n o rm a liz in g against the corresponding GAPDH signal. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 144 o C O O L O 0 1 C \ 1 C M C M — I — I _1 k b 4 . 4 2 . 3 7 1 . 3 5 4 . 4 2 . 3 7 1 . 3 5 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 145 (Sanghani and Moran, 1997) (Table 3.2). Taken together, the GARFT protein did not seem to be the locus responsible for drug resistan ce in the L1210/D3 line. Interestingly, whereas GARFT can be p r e p a r e d from parental L1210 cells in high purity (>90%) by direct a p p l i c a ti o n of crude cytosolic extract onto a 1 0 -fo rm y l-5 ,8 -d id e aza fo late-lin k ed column, enzym e from L1210/D3 cells could not be purified in o n e single step using the same technique. When high speed s u p e r n a t a n t of the cytosolic preparations from L1210/D3 cells were applied to th e affinity column, more than 90% of GARFT activity was found in th e “flow through” fractions, suggesting a lack of binding to the affinity ligand (Fig. 3.3A). Addition of more affinity resin did not result in any increase in binding (Fig. 3.3A). At an initial glance, th ese observations suggested the presence of a m utated enzyme in th e resistant cells. However, it was later found that near c o m p le te binding to the affinity resin could be obtained by subjecting th e cytosolic extracts from L1210/D3 cells to an am m onium sulfate precipitation step before loading onto the column (data not s h o w n ). Thus, the apparent lack of interaction of G A RFT to the affinity lig a n d was not due a structurally altered enzyme per se R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 146 B 1.251 ’> o < tt. 0.751 a < ° 0.51 1) > 0 '*5- d u -— n ED L1210 ED L1210/D 3 1.25-1 1 - 0.75' D.5- 0.25- 0 II III IV VI VII Fig. 3.3 Purification of GARFT from DDATHF-resistant L1210 cells. (A) Apparent lack of binding of L1210/D3-derived GARFT to affin ity chrom atography column. GARFT activity was m easured in the h ig h speed supernatant before applying to the affinity column (I) and in the “flow-through” eluate after binding to the column for 30 min (II) and 4 hour (III), respectively. GARFT activity was again measured i n the eluate after the addition of more resin and incubation for another 30 min. (IV) (B) Heat-stable factor(s) in high s p e e d supernatant from L1210/D3 cells causing dissociation of GARFT fr o m affinity column. Affinity column containing bound r e c o m b i n a n t L1210 was washed with either low salt phosphate buffer (V) or b o iled high speed supernatant derived from L1210/D3 cells (VI), GRAFT activity was measured in the eluate. No GARFT activity was found i n the boiled high speed supernatant derived from L1210/D3 cells (VII) (negative control). GARFT activity was expressed relative to t h a t found in the high speed supernatant derived from wild-type L1210 cells. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 147 but to inhibition of binding by some soluble factor(s) present in cytosolic extracts from the resistant cells. Further in v e s tig a tio n revealed that heat-treated cytosolic preparations from L1210/D3 cells could interfere with the binding of purified recom binant L1210 GARFT to the affinity column; strongly suggesting that some h e a t- stable substance was present in the cytosolic extract from L 1210/D 3 cells which inhibited GARFT binding to the affinity column (Fig. 3.3B). It has now become apparent that this inhibition of GARFT binding is secondary to an elevated endogenous folate polyglutamates pool in the resistant cells (see below). FPGS and conjugase activities The characteristics of the two key enzymes involved i n m ain ten an ce of steady state levels of folylpolyglutam ates, FPGS a n d folylpolyglutam yl hydrolase (conjugase), were com pared b etw e en wild-type and DDATHF-resistant L1210 cells. Using (6S)- tetrahydrofolate as a standard substrate, the activities of FPGS f o u n d in high-speed supernatant fractions of L1210 and L1210/D3 cells were not significantly different (Table 3.2). Likewise, northern b lo t analysis also revealed comparable levels of FPGS messages in the tw o cell lines (Fig. 3.2). Kinetic studies performed on FPGS p a r tia lly purified from both cell lines using aminopterin as a substrate s h o w e d R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 148 indistinguishable characteristics (Km values were 22 ± 3 and 24 ± 5 for enzyme derived from L1210 and L1210/D3 cells, respectively) (Fig. 3.4 & Table 3.2). The kinetics of the utilization of either (6R)- DDATHF or (6S)-DDATHF were also identical for FPGS isolated fro m both lines, albeit a more complex kinetic pattern was observed w ith this substrate (Fig. 3.4) (see below). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 149 Fig. 3.4 Kinetics of FPGS partially purified from L1210 (A) and L1210/D3 (B) cells using aminopterin, (6S)-DDATHF, and (6R)-DDATHF as substrates. FPGS was partially purified from high speed supernatant fraction of cell extracts by am m onium sulfate precipitation. C h a rc o a l adsorption assay was performed on desalted enzyme using t h e indicated concentrations of am inopterin (squares), (6R)-DDATHF (diam onds) and (6S)-DDATHF (circles). Double reciprocal plots are shown in insets. For details, see “Materials and Methods”. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 150 0 3 - 2 .0 - O e G r> > ) 0.5 1 1/[Substrate], l/(a,M o .o 80 100 40 60 20 0 [Substrate], (iM 3.0-1 0.3 - 2.0 - I 0.5 1 l/[Substrate], 1/j l iM o .o 60 20 40 80 100 0 [Substrate], [iM Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15 1 Two salient observations were made pertaining to the kin etics of L1210 (and L1210/D3) FPGS with DDATHF. First, the velocity versus substrate concentration plots did not display the typical h y p e rb o lic curve of M ichaelis-Menten kinetics (Fig 3.4). Double reciprocal p lo ts of the data suggested either more complex kinetic behavior of a single enzyme or the presence of more than one isoforms of FPGS i n both L1210 and L1210/D3 cells (Fig 3.4, insets). Second, there was a striking difference in substrate specificities between FPGS is o la te d from L1210 cells and that from mouse liver. Whereas a m i n o p t e r i n was utilized similarly by FPGS isolated from both sources, (6R)- DDATHF exhibited lower Km and Vmax values with enzyme d e r iv e d from L1210 than with that obtained from mouse liver (compare Fig. 3.4 with Fig. 3.5, also see Table 3.2). These s e r e n d ip ito u s observations have led to the discovery of tissue-specific expression of FPGS isoforms (Turner et al., 1999 and "Discussion" Ch. 3). Conjugase activity was not detectable in either L1210 or L1210/D3 cells, using experimental conditions under which e n z y m e was found at 4.9 ± 0.2 nm ol/m in-m g protein (n=2) in CCRF-CEM cells, a value similar to that previously reported from this cell line (Table 3.2) (O'Conner et al., 1991). R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 152 Fig. 3.5 Disparate substrate specificities of FPGS isolated from mouse leukemic L1210 cells (A) and mouse liver (B). Livers from DBA/2 mice were perfused in situ with PBS, excised, a n d homogenized. After disruption of cells by sonication, high sp ee d cytosolic fractions were prepared by centrifugation, followed by ammonium sulfate precipitation. Enzyme activity was measured by charcoal adsorption as in Fig. 3.4. Substrates used were a m i n o p t e r i n (squares), (6R)-DDATHF (diamonds) and (6S)-DDATHF (circles). For details, see “Materials and Methods”. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. , nmol/hr nm ol/hr A 153 3 2 1 0 80 100 [Substrate], juM B 4 2 0 0 0.5 l/[Substrate], 1/pM 0 .5-1 0 .4 - 0 .2 - > 200 100 1 0 - > 0 0.1 0.2 0.3 l/[Substrate], l/(iM [Substrate], |iiM Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 154 Uptake of [14C](6R)-DDATHF in wild-type and DDATHF- resistant L1210 cells A limited amount of moderate specific activity (=4.4 C i /m o l) [14C](6R)-DDATHF was available to allow comparison of the t i m e course of drug accum ulation between the two cell lines. At 2 _ M extracellular [14C](6R)-DDATHF, a concentration which c o m p le te ly inhibited the growth of L 1210 but had no effect on that of L 1210/D 3 cells, L I 210 cells showed a more rapid initial uptake and a h ig h e r plateau level of radiolabelled drug. At 12 hour, L I 210 and L 1 2 1 0 /D 3 contained 12.3 pmol and 2.81 pmol of drug per 106 cells, respectively (Fig 3.6). Interestingly, both levels were higher than t h e cellular content of GARFT (=1.5 pm ol/106 cells), implying that a n excess amount of drug over the level of the target enzyme is r e q u ir e d for inhibition of growth. This result provided the first piece of biochem ical evidence that the two cell lines differed in their ab ility to accumulate (6R)-DDATHF. However this study did not d is tin g u is h whether the difference in uptake was due to alteration in m e m b r a n e transport, efflux, or polyglutam ation of (6R)-DDATHF. M oreover, since the long chain polyglutamate derivatives of DDATHF were m o r e potent inhibitors of GARFT (Chapter 3, Baldwin et al., 1991), th e effect on cell proliferation would be profoundly different if L 1210/D3 cells contained only the parent drug. To address these q u estio n s, each biochemical process has to be studied in iso la tio n . R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 155 15 1 8 10 - o O o p Q. 0 5 10 15 20 25 incubation time, h Fig 3.6 Time course of uptake of [^4C](6R)-DDATHF in wild-type (squares) and DDATHF-resistant (diamonds) L1210 cells. After incubating with 2 pM for the indicated time periods, cells were harvested and washed with ice-cold PBS. Cell pellets were d isso lv e d in IN N aO H and processed for scintillation counting. Cellular level of GARFT in L1210 cells is indicated by the dash line. For details, see “Materials and M ethods”. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 156 Synthesis of high specific activity [3H](6R)-DDATH F In order to study the m em brane transport of (6R)-DDATHF and th e accumulation of its polyglutam ate metabolites in the resistant cells, radiolabelled drug of high specific activity was synthesized. A procedure for the synthesis of [3H](6R)-DDATHF was developed w ith the advice of Dr. Chuan Shih of Eli Lilly Research Laboratories, w h ich involved the coupling of [3,4-3 H]glutamic acid diethyl ester to (6R)- 5,10-dideazapteroic acid (Fig. 3.7). In this scheme, (6R)5,10- dideazapteroic was prepared by hydrolyzing (6R)-DDATHF in HC1, followed by purification on a DEAE cellulose column. It was necessary to avoid overloading of the DEAE colum n because a sm all chemical am ount of unhydrolyzed (6R)-DDATHF that was carried to subsequent step would substantially com prom ise the high specificity of the final product. The diethyl ester of [3 ,4 -3 H]glutamic acid was synthesized by reacting [3,4-3 H]glutamic acid with excess ethyltosylate in dry ethanol under reflux. The crude diethyl ester was purified on a silica gel colum n eluted using a m eth an o l/ch lo ro fo rm mobile phase and pooled fractions were d rie d under nitrogen. In the coupling reaction, the carboxyl group in 5,10- dideazapteroic acid was first activated by the addition of a n equimolar am ount of l-chloro-3,5-dim ethoxytriazene in anhydrous Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 157 COOH COOH 6R-DDATHF H2 COOH 6 N HCL CH3Cr N OCHj 1 -c h lo ro -3 ,5 -d im e th o \y tria z in e 6R-DDTHPtc H?N N DEAE chrom atography COOEt , / s i l i r chi COOH [3,4- H]-L-glutamic acid 0.1M ethyl p-toluene- sulfonate COOEt I3,4-3H]-L-glutamate 'silic a gel colum n diethyl ester ch ro m ato g rap h y COOEt COOEt 3 H-6R-DDATHF diethyl ester NaOH ion-pair HPLC COOH. h2n^ n^ n^ 3h COOH 6R-DDATHF Fig. 3.7 Schematic of synthesis of high specific activity [ H](6R)-DDATHF. For details, see “Materials and Methods” . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 158 DMSO containing the organic base, 4 -m eth y lm o rp h o lin e. This mixture was added to the dried [^H ]glutam ate diethyl ester and th e reaction was allowed to proceed for 6h at RT. [^H](6R)-DDATHF was obtained by hydrolysis of the [^HJdiethyl ester of (6R)-DDATHF. T h e product was purified on a DEAE column, followed by reverse p h a s e hplc ion pair system using a mobile phase of t e t r a b u t y l a m m o n i u m hydrogen sulfate containing methanol. After removal of the p a ire d ion in the mobile phase via a cation exchange AG 50W-XS c o lu m n , the purified product was stored at -20 °C in 33% ethanol. The p u rity of the synthetic product was determined by co -ch ro m ato g rap h y w ith authentic (6R)-DDATHF standard followed by scintillation co u n tin g , and was estim ated to be > 98%. The stability of the [3h ](6R)- DDATHF was studied by hplc under various conditions; purity was >98% for at least 2 weeks when stored at -20 °C, was 95% after incubation at 37 °C in transport buffer after 24 h, but dropped to about 70% after 72 h. (Fig 3.8). The specific activity of the final product ranged from to . The chemical am ount of ra d io la b e lle d (6R)-DDATHF was considered insignificant and was generally ignored in the calculation of total drug concentrations in most ex p e rim e n ts unless when it constituted >15% of total drug. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 159 100 8 0 - 6 0 - > 3 a 4 0 - 0 1 2 3 4 5 6 -80 *C 10% EtOH 37 'C Transport buffer Tim e, days Fig. 3.8 Stability of [3H](6R)-DDATHF. Radioactive purity of the synthesized [3H](6R)-DDATHF was follow ed under the indicated conditions. An aliquot of the label was c o ch ro m a to g rap h ed with (6R)-DDATHF standard using hplc c o n d itio n s described under "Materials and Methods". The purity of the label was estim ated from the percentage of radioactivity associated w ith the eluate fractions that corresponded to the standard compound. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 160 Membrane transport of (6R)-DDATHF In order to determine whether resistance to (6R)-DDATHF in L1210/D3 cells was due to defective m em brane tr a n s p o r t, translocation of drug across the plasma membrane was studied using high specific activity [3h ](6R)-DDATHF. At 2pM e x tra c e llu la r [3H](6R)-DDATHF, a concentration which completely suppressed th e growth of L 1210 but had no effect on the growth of L1210/D3 cells, the uptake of radiolabelled drug was linear in both cell lines over th e first 6 min. The initial velocity of drug uptake was 0.63 ± 0 . 1 0 p m o l / m i n - l O ^ cells for L 1210 and 0.38 ± 0.01 pmol/min-106 cells for L1210/D3 respectively (n=3) (Fig 3.9A). Influx of [^H](6R)-DDATHF was m easured over the initial 4 min after the addition of v a rio u s concentrations of drug. The influx K m for (6R)-DDATHF in L 1210/D 3 cells was 3-fold higher than that in L 1210 cells (5.7 ± 0.2 pM and 1.7 ± 0.6 pM, respectively). The values for influx V m ax were n o t significantly different between the two cell lines (0.93 ± 0 .1 2 and 1.2 ± 0.1 p m o l/m in - lO ^ in L I 210 and L1210/D3 cells, respectively (n = 3 )) (Fig. 3.9B). L 1210 and L1210/D3 cells were preloaded with 2 and 6 pM [3h ](6R)-DDATHF, respectively, for 20 min to allow accum ulation of comparable intracellular levels of radiolabelled drug. The rate of (6R)-DDATHF efflux was then measured in drug-free buffer over a 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 161 Fig. 3.9 Transport of (6R)-DDATHF in L1210 (open squares) and L1210/D3 cells (closed squares). (A ) Time course of influx of (6R)-DDATHF at 2pM e x tra c e llu la r drug. (B ) Concentration dependence of rate of (6R)-DDATHF influx into cells m easured over 4 min. The inset is a replot of th e reciprocal of initial velocity versus that of drug concentration. ( C ) Efflux of radiolabelled drug from cells were m easured after a 2 0 - m in pre-loading of L I 210 and L1210/D3 cells with 2 and 6 pM [3h](6R )- DDATHF. The dashed arrow represents the level of the target e n z y m e GARFT in these cells; hence, virtually all of the intracellular drug are expected to be in the form of free drug. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 162 o o u 'fe O £ a. C O 13 o 'fe I c £ o £ a. co 13 o o £ a. 5.0-1 3.0- 2 .0- 1.0 - o.a 10 0 20 time (min) 1.0 - , 10 - 0 .6 - 0.4- 0 .2 - 0.0 10 0 5 15 10 (6R)-DDATHF {iM) 3 0 20 10 40 time (min) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 163 min period. The initial rate of drug efflux was similar in both cell lines (Fig 3.9C). However, it should be noted that, in wild type L1210 cells, about one-third of the intracellular drug appeared to be nonexchangeable and probably represented p o l y g l u t a m a t e metabolites of (6R)-DDATHF, and this fraction was noticeably a b s e n t in the resistant cells (Fig. 3.9C). Taken together, there was a 3-fold increase in influx K m for (6R)-DDATHF in L1210/D3 which would contribute to the o v erall level of resistance in these cells. However, it was unlikely that such a moderate change alone could account for the extent of re sistan c e observed in L1210/D3 cells (Table 3.1). Polyglutamylation of (6R)-DDATHF in whole cells There has been increasing evidence that the formation of intracellular polyglutamates of (6R)-DDATHF is an i m p o r t a n t determ inant of its cytotoxicity, we therefore examined whether d r u g resistance in L1210/D3 cells was caused by i m p a i r e d polyglutam ation. When cells were treated with 0.5 pM [3h ](6R)- DDATHF for 16 h, the wild-type line accum ulated 13 pmol of d ru g per 106 cells, 84% of which were in the form of long c h a i n metabolites ( G lu ^ G lu g ) , with the hexaglutamate m e ta b o l i t e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 164 predo m in atin g (Fig. 3.10A). On the other hand, the resistant line accum ulated only 0.82 pmol of drug per 10^ cells, with long c h a in polyglutamates accounted for less than 0.34 pmol per 10^ cells (42% of total intracellular drug) (Fig. 3.10B). Hence, long chain m e ta b o lite s of DDATHF were found in L1210/D3 cells at less than 3% of the level in parental L1210 cells. Assuming 0.67 pi of intracellular water p e r 10^ L1210 cells, total long chain polyglutamates of (6R)-DDATHF were present at 0.51 and 10.9 pM concentrations in L1210/D3 and L1210 cells, respectively; the former is less than the cellular content of GARFT in wild-type cells (estimated to be 1 pM (Baldwin et a l ., 1991)), the latter substantially higher than target enzyme levels. In these experim ents, the level of u n m etab o lized (6R)-DDATHF in L1210/D3 cells was also 3-fold lower than that in L1210 cells, presumably due to the higher K m for drug transport. To allow legitimate com parison of the rate of p o ly g lu ta m a tio n between t h e two cell lines, L1210/D3 cells were treated with higher c o n c e n t r a t i o n s of drug (2 and 5 pM) in order to compensate for the transport d efect in these cells and to achieve a steady state level of intracellular (6R)- DDATHF m o n o g lu ta m a te that was com parable to the level found i n L I 210 cells treated with 0.5 pM (6R)-DDATHF. Even at (6R)-DDATHF Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 165 A B 2 - G. 1 2 3 4 5 6 7 8 4 2 1 2 3 4 5 6 7 8 0 Moles of Glutamate per mole of (6R)-DDATHF Fig. 3.10 Accumulation of (6R)-DDATHF polyglutamates in L1210 (A) and L1210/D3 (B) cells grown in standard culture medium. After exposure of cells to the indicated concentrations of (6R)- DDATHF, cell-free extracts were prepared and the p o l y g l u ta m a t e metabolites of the parent drug were separated by reverse p h a s e paired-ion hplc. Peaks were identified by co-injection of a u t h e n t i c (6R )-D D A TH F polyglutam ate standards. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 166 monoglutamate levels in L1210/D3 cells equal to or somewhat g reater than those in L1210 cells, a marked deficit in the accum ulation of (6R)-DDATHF polyglutamates remained evident in the L1210/D3 cells (Fig. 3.10B). Hence, it appeared that the impaired formation of (6R)- D D A TH F polyglutamates in the resistant cells was not secondary to a decrease in drug transport, but rather it reflected a s u b s ta n tia l dysfunction in the polyglutam ation process, per se. This posed a basic dilemma: cellular polyglutam ation of (6R)-DDATHF was low enough in L1210/D3 cells to explain drug resistance, but FPGS isolated from these cells appeared normal. Folate cofactor pools in L1210 and L1210/D3 cells Formation of folylpolyglutam ates is essential for th e intracellular retention of both natural folates and antifolates, th e fact that L1210/D3 cells failed to accum ulate (6R)-DDATHF polyglutam ates prom pted us to suspect that in tra c e llu la r endogenous folates would also be low in these cells. When to ta l cellular folates were assessed in cells cultured in [3H]folic acid, surprisingly, L1210/D3 cells showed a 4-5-fold elevation over that in parentla L1210 cells (Table 3.3). The majority of the cofactors in these cells appeared to be in polyglutamate forms since about 80% of the total radiolabelled material were retained after incubation w as continued in medium containing unlabeled folic acid for 12 h ( d a t a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 167 not shown). Conversely, when cells were grown in folate-free m e d i u m supplem ented with [3H](6S)-5-formytetrahydrofolate, the total folate level in the wild-type cells was 4-fold higher than that in th e resistant cells (Table 3.3). Since L1210/D3 cells were resistant to (6R)- DDATHF only when they were grown in m edium containing folic b u t not folinic acid, this correlation between drug resistance a n d endogenous folate levels led us to speculate that the resistan ce phenotype of L1210/D3 cells was caused by the aberrant increase in cofactor pools in these cells. One possible consequence was that th e cellular content of 10-formyltetrahydrofolate, the natural s u b s tra te for the GARFT reaction, would increase and antagonize the in h ib itio n by (6R)-DDATHF. The level of 10-form yltetrahydrofolate, w hich, presumably, was present mainly in the form of polyglutam ates, was measured using a modified version of the thym idylate s y n th a s e ternary-complex formation assay (Fig. 3.11) (Kesavan et al., 1986, Schmitz et al., 1994). The cellular content of the sum of T H F (G lu )n and 5,10-C H 2-T H F (G lu )n was first determined by this method. T h e levels of 10-CHO-THF(Glu)n was then determ ined indirectly by quantitating the additional T H F (G lu )n formed in the presence of excess GARFT and glycinamide ribonucleotide. The concentration of 10-C H O -T H F(G lu)n increased somewhat disproportionally to th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 168 increase in the total folate pool, and was 7-fold higher in L 1210/D 3 cells than in wild-type L1210 (Table3.3) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 169 Fig. 3.11 Schematic of determination of folate cofactor pools in L1210 and L1210/D3 cells. Cellular levels of the sum of H 4P te G lu n and 5,1 0 -C H 2-H 4PteG lun were determ ined using the ternary complex formation assay. H 4P te G lu n was converted, in the presence of formadehyde, into 5,10-C H 2- H 4P te G lu n, which, in turn, was covalently trapped by excess L. casei TS and [3H]FdUMP in a ternary complex. (Kesavan et al., 1986; Schimitz et al., 1994; Keyomarsi and Moran, 1998). The levels of 10- C H O -H 4PteGIun were determined indirectly by measuring t h e additional H 4P te G lu n formed in the presence of excess of GARFT a n d glycinam ide ribonucleotide (GAR). CHO-GAR, formyl g l y c in a m id e ribonucleotide. For details, see “Materials and Methods”. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 170 CHO-GAR GAR 10-C H O -H 4PteGlun GRAFT H 4PteG lun V ” 5,10-C H 2-H 4PteGlu„ 5,10-CH2-H 4PteGlun • [3H]FdUMP • TS ternary covalent complex Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 171 Table 3.3 Folate cofactor pools in wild-type and DDATH F-resistant L1210 cells L 1210 (n) L 1 2 1 0 /D 3 (n) ( p m o l / 1 0 ^ c e l l s ) Cells grown in folic acid Total folate 3 . 6 ± 1 .4 (3) 16 ± 2 .2 (3) 10-CHO-THF(Glu)n 1 . 1 ± 0 .3 (3) 7 . 9 ± 0 . 8 (3) 5,10-CH2-THF(Glu)n + THF(Glu)n 2 . 0 ± 0 .6 (3) 8 . 9 ± 1 . 6 (3) Cells grown in folinic acid Total folate 6 . 9 ± 0 .2 (2) 1 . 6 ± 0 . 1 (2) Cells were passaged in folate-free RPMI standard medium supplemented with either 2.3 pM [% ]folic acid or 6.0 x 10"^ M folinic acid containing [^H](6S)-5- formyltetrahydrofolate for one week prior to quantitation of total cellular folates. Cells were then harvested by centrifugation, washed with PBS, dissolved in 1 N NaOH, and radioactivity was determined by scintillation counting. For measurement of cofactor levels, cells were grown in the indicated medium without radiolabelled material, folate cofactors were extracted from cells in boiling 10 mM phosphate buffer, containing 1% ascorbate and 1% P-ME, and cofactor levels were determined using the ternary complex formation assay (Kesavan et al., 1986; Keyomarsi and Moran, 1988; Schimitz et al., 1994) as described under "Materials and Methods". Values represents mean ± SD or 0.5 x range (for n=2) from n experiments. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 172 Effect of the expanded folate pools on the accumulation of intracellular (6R)-DDATHF polyglutamates in DDATHF- resistant cells To investigate whether the increased folate pools in L 121 0 /D 3 cells was responsible for the impaired p o ly g lu ta m a tio n of c e llu la r (6R)-DDATHF in these cells, drug p o ly g lu tam atio n in whole cells was studied in folate-depleted conditions. Both wild-type and r e s is ta n t cells were grown in folate-free m edium su p p lem en ted with 5.6 pM thym idine and 32 pM hypoxanthine for at least 10 generations to allow complete depletion of intracellular folates. After i n c u b a t i o n with 0.5 pM [3H](6R)-DDATHF for 16 h in folate-depleted m e d i u m , wild-type L1210 cells contained 14 pmol of drug per 10^ cells a n d 74% of which consisted of long chain polyglutam ates, an extent of drug accum ulation similar to that observed in cells c o n t a i n i n g normal levels of folate pools (com pared Fig. 3.10A and 3.12A). Following incubation with either 0.5 or 2 pM [3h ](6R)-DDATHF, folate-depleted L1210/D3 cells accu m u lated 10.5 and 11.4 pmol p e r 10^ cells, respectively. The profile of long chain derivatives of (6R)- DDATHF accum ulated in these cells was essentially in d is tin g u is h a b le from that found in L1210 cells (Fig. 3.12A and 3.12B). Thus, we concluded that the apparent lack of (6R)-DDATHF p o l y g l u t a m a t i o n in L1210/D3 cells was caused by the elevated folate pools in th ese cells (Table 3.3). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 173 o o o B2 a 0 1 2 3 4 5 6 7 8 4 2 0 1 2 3 4 5 6 7 8 ^ 2 - 1 2 3 4 5 6 7 8 M oles of G lutam ate per mole of (6R)-DDATHF Fig. 3.12 Accumulation of (6R)-DDATHF polyglutamates in L1210 (C) and L1210/D3 (D) cells grown in folate-depleted culture medium. After exposure of cells to the indicated concentrations of (6R)- DDATHF in the presence of 5.6 pM thym idine and 32 pM hypoxanthine, intracellular polyglutam ate metabolites were analyzed by hplc as described in Fig. 3.10. Results of studies on (6R)- DDATHF poly g lu tam atio n in wild-type (A) and drug-resistant (B) L I 210 cells grown in folate-containing m edium are included for c o m p a r i s o n . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 174 Polyglutamation of (6R)-DDATHF in cells grown in folinic acid medium When cells were passaged in m edium supplem ented w ith folinic acid rather than folic acid, L1210/D3 cells were p a ra d o x ic a lly more sensitive to (6R)-DDATHF than were parental cells (Table 3.1). If the expanded folate pools in the resistant cells had prevented th e accum ulation of intracellular (6R)-DDATHF polyglutamates, we predicted that a 4-fold decrease in cellular folates in these cells, w h e n grown in folinic acid (Table 3.3) would prom ote polyglutam ation of the parent drug. This was indeed the case: following incubation w ith 0.5 pM [3h ](6R)-DDATHF, the polyglutam ate derivatives of d ru g formed in L1210/D3 cells were of longer chain length than those in wild-type cells when both cell lines were grown in m e d i u m containing folinic acid as the sole folate source (Fig. 3.13). Folic acid transport in L1210 and L1210/D3 cells. To determine the cause of the expanded folate pools in L 1210/D 3 cells, the m embrane transport of [3H]folic acid in both cell lines was compared. In these studies, cells were pre-treated with 10 pM trimetrexate, a lipophilic analog of MTX, in order to block th e metabolism of [^H]folic acid by DHFR (Assaraf and Goldman, 1997). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 175 In the presence of 5 pM [^Hjfolic acid, the initial uptake of radiolabelled material at 37 °C was barely m easurable in the L1210 line, a finding which was consistent with the previous report t h a t folic acid is a poor substrate for m em brane transport in these cells (Kt > 200 pM ) (Sirotnak et al., 1987; Goldman et al., 1968). However, under the same conditions, L1210/D3 showed a 50-fold increase in the initial uptake and a 15-fold increase in the steady state level of the radiolabelled com pound, measured at 30 min (Fig. 3.14). T hese results provided a basis for the elevated level of intracellular folates in L1210/D3 cells: an altered m em brane transport system t h a t allowed efficient uptake of folic acid, a folate com pound that is, otherwise, poorly utilized for transport in wild-type L I 210 cells. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 176 3 - I 1 2 3 4 5 6 7 8 Moles of Glutamate per mole of (6R)-DDATHF Fig. 3.13 Metabolism of (6R)-DDATHF to polyglutamates by L1210 (A) and L1210/D3 (B) cells grown in culture medium containing folinic acid as the source of folate. Cells grown in the presence of 60 nM folinic acid were exposed to 0.5 pM (6R)-DDATHF for 16 h and cellular polyglutam ates of drug was analyzed as in Fig. 3.10. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 177 o < u a • 4 — » C l , D j|3 "3 0.5- (U U a — b — o 10 20 30 40 time (min) Fig. 3.14 Time course of uptake of folic acid in L1210 (open symbols) and L1210/D3 cells (closed symbols) at 0 °C (triangles) and 37 °C (squares). Cells were pretreated with 10 pM trimetrexate for 10 min at 37 °C, harvested, and exposed to 5 p M folic acid for the indicated times. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 178 CHAPTER 3 DISCUSSION In this study, we characterized the phenotype of an L1210 subline selected for resistance to (6R,S)-DDATHF. We h a v e determined that the primary event causative of drug resistance is a n alteration in the substrate specificity of folate transport across th e plasma m em brane. Resistance is not a direct result of a d e c re a s e d transport of (6R)-DDATHF; the influx K m of (6R)-DDATHF in c re a s e d by only 3-fold. However, influx of folic acid, a poor substrate for transport in wild-type L1210 cells, has become remarkably efficient, leading to a substantially expanded cellular folate pool and a subsequent blockade of (6R)-DDATHF polyglutam ate synthesis in t h e resistant cells. In experiments in which the intracellular folates were removed by growth in the absence of extracellular folic acid, polyglutam ation of (6R)-DDATHF was normal in the resistant cells. This established that the expanded folate pool was responsible for the poly g lu tam atio n defect and, hence, the drug resistance of th is cell line. To the best of our knowledge, this represents the first r e p o r t of an unique m echanism of antifolate resistance in which d r u g R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 179 sensitivity is modulated by endogenous folate pools (Tse et al., 1994; Tse and Moran, 1998). A related m echanism was s u b s e q u e n tly reported when it was shown that CHO cells selected for resistance to pyrimethamine, a lipophilic DHFR inhibitor, showed augm ented folic acid accum ulation as a result of a reduced folic acid efflux a c tiv ity (Assaraf and Goldman, 1997). There are aspects of the phenotype displayed by the L 1210/D 3 cells which are unique but which are nicely explained by the c h a n g e s in the folate cofactor pools in these cells. Thus, the drug re sistan c e of L1210/D3 cells was displayed only when cells were grown in folic acid-containing medium; when the folate requirement for growth w as sufficed by folinic acid, L1210/D3 cells were, paradoxically, m o re sensitive to (6R)-DDATHF. Resistance appears to be conditional to the folate source because the intracellular pool was only expanded i n folic acid-containing medium; in folinic acid medium, c e llu la r folates were diminished. In accord with this, when in tra c e llu la r p o lyglutam ation of (6R)-DDATHF was studied in cells passaged i n folinic acid, L1210/D3 cells in fact accu m u lated p o l y g l u t a m a t e metabolites of slightly longer chain length. Previous literature h a s shown that m am m alian tumor cells are more sensitive to FUdR in the presence of folinic acid because of an expansion of the c e llu la r 5,10-m eth y len etetrah y d ro fo late pool and a kinetically s ta b iliz a tio n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 180 of the thym idylate s y n th a s e - 5 ,1 0 - m e th y le n e te tr a h y d ro fo la te - fluorodeoxy-uridylate ternary complex (Keyomarsi and Moran, 1986; Keyomarsi and Moran, 1988; Evans et al., 1981). Hence, it would be predicted that the elevated cellular folate pool in L1210/D3 cells would render them collaterally more sensitive to FUdR; this p r o v e d to be the case (Table 3.1). Analysis of resistant mutants has provided im portant in sig h ts into the biochemical processes critical to the cytotoxicity induced by a drug. There has been increasing evidence for the central role DDATHF polyglutam ates play in m ediating the anti-tum or effect of this drug. A series of cell lines selected for resistance to m e t h o t r e x a te were reported which had decreased levels of FPGS ; these cells were cross-resistant to (6R)-DDATHF, and the degree of resistance w as directly related to the decrease in enzyme levels (McGuire et al., 1993). Two studies have previously reported the characterization of tumor cells selected directly for resistance to (6R)-DDATHF. In t h e one study, an increased activity of folyl-y-glutamyl h y d r o la s e (conjugase), a carboxypeptidase responsible for the catabolism of folyipolyglutam ates, was implicated as a cause of resistance to (6R)- DDATHF in H35 hepatom a cells. In a recent report, two h u m a n leukemic CCRF-CEM cell lines with substantial acquired resistance to (6R)-DDATHF (24- and 217-fold) were shown to have severe Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 181 im pairm ent of the polyglutam ation of the parent drug in spite of the fact that very little change was noted in the activity of FPGS in vitro. The m echanism of decreased polyg lu tam atio n in those cells remains unexplained. It is interesting to note that these re sistan t CCRF-CEM sublines also showed a som ew hat elevated e n d o g e n o u s folate pool, suggesting that the p h en o m en o n we now report m i g h t have been involved. A third subline from that study expressed very low levels of FPGS activity, in spite of a steady state content of mRNA for this protein equivalent to that in wild type cells, suggesting a mutation in the coding region of this enzym e that rendered cells resistant to DDATHF. Hence, it is clear that any of several m echanism s which lead to a decreased cellular content of DDATHF polyglutam ates will compromise the cytotoxicity of this a g e n t. Interestingly, although an increased level of the target e n z y m e , GARFT represents a potential mechanism of resistance to DDATHF, i t has not been reported in the literature. In m am m alian cells, different gene loci may have different propensity to a m p lific a tio n ; thus, it is possible that the trifunctional GARFT-AIRS-GARS is lo c a te d in a region in the genome where gene amplification is infrequent. The m echanism responsible for the limited accum ulation of (6R)-DDATHF polyglutam ates in L1210/D3 cells is f u n d a m e n t a l l y different from those reported previously. The defect i n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 182 p o ly glutam ation in these cells is not a direct consequence of a n y biochemical alteration in FPGS per se nor an increase in folylpoly-y- glutamate hydrolase activity but rather appears due to a d irec t effect of intracellular folates on metabolism of (6R)-DDATHF to polyglutam ates. The exact nature of this control m echanism is n o t yet clear, although, it is reasonable to assume that regulation takes place at the level of FPGS. Thus, previous studies have shown t h a t the Km values for FPGS of most of the m onoglutam ate forms of th e physiological folates were com parable or even lower than that of (6R)-DDATHF (Taylor et al., 1985; Moran and Colman 1984) and t h a t Km values for FPGS within a series of homologous c o m p o u n d s generally decrease with increasing polyglutam ate chain len g th (Cichowicz and Shane, 1987; Chen et al., 1996). Hence, a sim p le mechanism that would explain this effect is that high level of folate cofactors (or one particular form of cofactor) function as c o m p e titiv e inhibitors for the FPGS reaction blocking the conversion of DDATHF to its polyglutam ate derivatives. On the other hand, m a m m a l i a n FPGS has been commonly found to display distinct s u b s tra te inhibition at higher concentrations of folate com pounds (Moran et al., 1984), suggesting the possibility of an allosteric binding site of folates in this enzyme with regulatory significance. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 183 Although kinetic studies of FPGS isolated from wild-type a n d DDATHF-resistance LI 210 cells showed an identical pattern of substrate utilization, a striking disparity in substrate specificity was noted between FPGS derived from mouse liver and that from L1210 cells (Fig. 3.5 and table 3.2). These findings were similar to th o se reported by Rumberger et al. (1990) where different s u b s tra te preference for various MTX analogs was found with FPGS is o la te d from mouse small intestine, L1210 cells, and S-180 sarcoma cells. Recently, explanations for these results have been sought at th e molecular level in our laboratory. When sequences of FPGS transcripts from mouse and L I 210 cells were compared, m a r k e d differences were found in the cDNA sequences upstream from th e border of exon 1 and 2 (Fig. 3.15 and Turner et al., 1999). T h e am ino-term inus of mouse liver FPGS contained 18 amino acids n o t found in the leukemic cell enzyme. To address whether th e disparities in substrate specificity of FPGS derived from mouse liver and L1210 cells are due to differences in amino acid sequences in th e amino term inus of the protein, cDNAs specific for FPGS from m o u s e and from L1210 cells were over-expressed in insect cells via a baculoviral expression system. The tissue-specific enzyme species were then purified to near homogeneity. Kinetic studies p e r f o r m e d using these recombinant proteins precisely recapitulated the p a t t e r n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 184 observed with partially purified FPGS derived form mouse liver a n d L I 210 cells (Turner et al., 1999). This tissue-specific expression of FPGS species with different substrate specificities has raised t h e possibility of design of antifolates that are selectively metabolized by isoforms expressed in some tumors but not normal stem cells (see "General Discussion" Ch. 5). We initially anticipated that our results were more of theoretical significance than clinical relevance, because the exact mechanism of resistance would only occur when folic acid was a nutritional source of intracellular folate cofactors, and folic acid d oes not occur naturally. However, recent preclinical and clinical w ork on (6R)-DDATHF have suggested direct therapeutic relevance of o u r results. Thus, the toxicity of (6R)-DDATHF during early phase I trials had dem o n strated that the drug was substantially more toxic to patients than was predictable from preceding preclinical experience in animals. The total dose of (6R)-DDATHF which could be safely given was found to be only 10-12 mg/m2 per course (Nelson et al., 1990; Pagani et al., 1992) This was in m arked contrast to a maxim um tolerated dose of 600 mg/kg/day given daily for 10 d a y s to mice in early toxicological studies. However, when mice were fed a folate-deficient diet for a short period, the toxicity of (6R)-DDATHF increased 1000-fold, mirroring the clinical pattern in animals for th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 185 first time. This preclinical studies has prom pted the initiation of a second generation of phase I trials of (6R)-DDATHF given in conjunction with folate supplem ent. Remarkably, Calvert and co workers have dem onstrated that (6R)-DDATHF was well tolerated u p to 170 mg/m2 every 3 weeks when given with oral folic acid at 5 mg/day for 7 days prior to and 7 days following each infusion of drug (Laohavinij et al., 1996). Several hypotheses have b een proposed to explain this striking effect of dietary folate supplementation on (6 R)-DDATHF toxicity. One possible m e c h a n i s m is that the transport of (6R)-DDATHF into normal stem cells is u p - regulated in folate-deficient animals. However, this has not b e e n substantiated by any in vitro or in vivo studies. A s e c o n d explanation is based on the observation that different p l a s m a clearance of (6R)-DDATHF was found in animals with d iffe ren t dietary folate intake. Thus, it has been shown that using [14C ](6 R)- DDATHF and autoradiography, the drug accum ulated in liver of folate-deficient animals as polyglutam ates and these hepatic d ru g metabolites were then slowly released to the circulation, mimicking a very toxic continuous infusion; this does not seem to occur in animals with higher folate intake (Pohland et al., 1994). A t h i r d possible m echanism would be that the p o lyglutam ation a n d accum ulation of (6R)-DDATHF in the intestinal and bone m a r r o w Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 186 stem cells is prevented by higher tissue folate cofactor pools r e s u lt a n t from higher dietary intake of folic acid. Regardless whether th is dietary modulation involves changes in plasma pharm acokinetics of drug clearance or in p h arm aco d y n am ics of drug sensitivity in normal tissues, the toxicity of (6R)-DDATHF appears to be affected by a m echanism which is formally identical to that uncovered by th e L1210/D3 mutant, namely, a blockade of the polyglutam ation of (6R)-DDATHF by intracellular folates com pounds, (also see “ G eneral Discussion, Chapter 5). Three distinct routes have been im plicated in the m e m b r a n e transport of folic acid in L I 210 cells, including the reduced folate carrier (G oldm an et al., 1968), a carrier- m ediated system o p e r a tiv e at low pH (Henderson and Strauss, 1990), and, after selection for th e ability for growth at low concentrations of folates, the m e m b r a n e - bound folate receptor (Jansen et al., 1989a; Jansen et al., 1989b; a n d Kamen et al., 1988). In the next chapter, we present genetic a n d biochemical evidence leading to the conclusion that alterations in the reduced folate carrier system are involved in the resistan ce phenotype of L1210/D3 cells. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 187 Fig. 3.15 Generation of tissue-specific FPGS isoforms by differential exons usage. Murine liver (top) and leukemic (bottom ) FPGS are generated by alternative splicing involving the initial exons. T r a n s la t io n initiation at the up stream ATG allows synthesis of FPGS species t h a t contain the leader sequence for m itochondrial targeting; w h e reas enzyme species produced using the down stream ATG results i n cytosolic FPGS. (From Turner et al., 1999) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 189 CHAPTER 4 MOLECULAR GENETICS OF DDATHF-RESISTANT L1210 CELLS INTRODUCTION In the preceding chapter, the biochemical characterization of the phenotype of an L1210 subline selected for resistance to DDATHF was described. The underlying m echanism of resistance was determ ined to be due to changes in the substrate specificity of membrane transport of folate com pounds. Resistance to DDATHF by virtue of an impaired m em brane transport of the drug has been described in CCRF-CEM cells (Pizzorno et al., 1993). However, in th e case of L1210/D3 cells, (6R)-DDATHF influx was only m o d e r a t e l y reduced, whereas that of folic acid was markedly e n h a n c e d . Consequently, there was an expansion of the intracellular folate pools, which in turn prevented the metabolism of (6R)-DDATHF to its p o l y g l u ta m a t e s . M em brane passage of physiologic folates and antifolates in m am m alian cells is mediated by at least two biochem ically d is tin c t systems - the reduced folate carrier (RFC) and the m em brane folate binding protein (mFBP). (Henderson, 1990; Antony, 1992; A nto n y , 1996; Spinella et al., 1995; also see below). It has been shown t h a t Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 190 DDATHF can gain cellular entry via both of these routes (Pizzorno et al., 1993). A third transport route that functions optimally at lower pH (6 - 6.5) has also been implicated (Henderson and Strauss, 1990; Sierra et al., 1997; Assaraf et al., 1998). The focus of the p re s e n t study is to identify the transport system that has been altered in th e resistant cells and to determine the specific genetic m u t a t io n ( s ) involved that is/are responsible for the functional changes observed. Our analysis suggested that the acquisition of the highly re sistan t phenotype of L1210/D3 cells involves genetic or epigenetic changes in the RFC-1 , a gene that encodes the reduced folate carrier in th e murine system. The biochemistry, molecular genetics and the role of the tw o transport systems in tumor sensitivity/resistance to antifolates are outlined below. RFC-mediated transport B ioch em istry The RFC has been studied in several tum or cell lines a n d various normal tissues using primarily MTX as a model s u b stra te (Goldm an et al., 1968; Sirotnak, 1985; Ratnam and Freisheim, 1990; Henderson, 1990). MTX offers a distinct advantage for studying th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 191 transport process because of its tight association with its c e llu la r target, DHFR. Prior to saturation of all DHFR binding sites, free MTX does not exist intracellularly and consequently drug efflux a n d metabolism due to polyglutam ation are negligible. This allows accurate m easurem ent of unidirectional influx rates (Goldman et al., 1968). The RFC displays several biochemical properties of a m em brane transport carrier. Thus, it exhibits M ich ae lis-M e n ten kinetics of influx of folate com pounds that can be c o m p e titiv e ly inhibited by structurally related analogs (Goldm an et al., 1968). T h e influx velocity is highly tem perature dependent with Q 10 (27-37 °C) of 3.8 for MTX (Spinella et al., 1995). Further, trans-stim ulation of MTX influx can be dem onstrated by pre-loading intact cells with 5- formyltetrahydrofolate. The energetics of transport appears to be a n uphill process in which MTX can be concentrated in tra c e llu la rly against an electrochemical potential (G oldm an et al., 1968; S iro tn a k et al., 1968). However, RFC-mediated transport does not seem to couple directly to ATP hydrolysis or downhill m ovem ent of s o d i u m ions since metabolic poisons (e.g. azide and dinitrophenol) a n d Na/K-ATPase inhibitor, ouabain, have little effect on MTX influx (Goldman, 1969; Sirotnak, 1985). A characteristic feature of RFC- mediated transport is that it can be inhibited by a variety of n o n - folyl anions such as probenecid, b rom osulfophthalein, p h o s p h a t e , Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 192 bicarbonate, NADP, AMP and ADP (Henderson and Zevely, 1981; Henderson and Zevely, 1983) . When these anions were tested for a n effect on MTX efflux, a process that appears to be m ediated in p a r t by the RFC, trans-stimulation, was observed with some anions. T hese observations have led to the hypothesis that transport via the RFC is m ediated by coupling the uphill influx of folate com p o u n d s to t h e extrusion of intracellular anions down a concentration g r a d ie n t (Henderson and Zevely, 1981; Henderson and Zevely, 1983). However, the intracellular anion responsible for this coupling has yet to be identified. RFC utilizes MTX and 5-substituted reduced folates as t h e preferred substrates, with influx Kt values of 1-26 pM (Freisheim et al., 1989), but shows poor affinity for folic acid (Kt > 100 pM) (Henderson et al., 1986; Sirotnak et al., 1987). The V m a x for MTX influx ranges from 1 to 12 pmol/min-mg dry weight (Sirotnak, 1985). Biochemical study of purified RFC has been ham pered by th e relatively low level of expression of the carrier on the cell surface (=0.04 p m o l / l O ^ mouse L1210 cells) (Henderson and Zevely, 1984), and also by the lack of an effective assay for following the p ro te in during purification. Affinity labeling of the carrier by folate an a lo g s was initially developed as a method for tracing the RFC after Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 193 solubilization from the membrane. Using N -h y d ro x y -su c c in im id e esters of [^H]MTX as a specific probe, Henderson and Zevely (1984) have identified a single radioactive band of 38 kd after fr a c tio n a tio n of labeled m em brane proteins from mouse L1210 cells by SDS-PAGE. However, a m em brane protein of 45-48 kd m olecular weight (MW) was identified by another laboratory using N -h y d ro x y -su c c in im id e esters of [^Hlaminopterin as affinity label in the same cell line (Yang et al., 1988). It was not clear whether the lower M W reported in th e former study was the result of degradative proteolysis as suggested by Yang et al. (1988). The MW of the RFC had been determ ined in o th e r cultured tumor cells using various affinity labeling agents such as biotin-conjugates of MTX (Fan et al., 1991) and p h o t o c h e m i c a ll y activated folate analogs (Price et al., 1988; Freisheim et al., 1988). The estimated MW of RFC was 46-48 kd in rodent cells (Yang et al., 1988; Fan et al., 1991; Price et al., 1988) and 76-85 kd in h u m a n erythroleukemia K562 cells (Matherly et al., 1991). The s u b s ta n tia lly higher MW reported in human RFC is presum ably due to glycosylation of the protein since treatm ent of the affin ity -lab eled RFC derived from K562 cells with glycosidase resulted in a shift to a lower MW band on SDS-PAGE that ap p roxim ated that of mouse RFC (Matherly et al., 1991). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 194 Molecular genetics Two laboratories have independently reported the isolation of cDNAs from mouse and hamster that, when transfected into folate transport-defective cells, were able to restore RFC function (Dixon et al., 1994; William et al., 1994). Both cDNAs encode a protein with a predicted MW of 58 kd (518 amino acids) and were designated RFC-1 in the murine system. More recently, several reports have d e s c rib e d the isolation of hum an cDNAs, from human placenta, K562 a n d HSC93 cell libraries, that were able to restore RFC activity u p o n transfection into MTX transport defective cells (Prasad et al., 1995; Wong et al., 1995; Williams and Flintoff, 1995). The p r e d ic t e d molecular mass of the human homologue is 65 kD, and the d e d u c e d amino acid sequence was about 70% identical or similar to that of mouse and hamster R F C -1 (Wong et al., 1995). The human RFC has a significantly longer carboxy-terminal tail than the r o d e n t counterparts. In addition, the human carrier is glycosylated at a n Asn58 residue in the first extracellular loop (Matherly et al., 1991; Matherly and Angeles, 1994; Wong et al., 1998); the m urine RFC lacks this Asn in the corresponding location and is not glycosylated. In L1210 cells, the principal mRNA transcript was 2.3 kb although m i n o r cDNA species of smaller sizes have been described and m i g h t represent alternate splice forms of the gene (Brigle et al., 1995). It Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 195 should be noted that previous affinity labeling studies h a v e suggested that the MW of mouse RFC was 46 - 48 kd, s u b s ta n tia lly smaller than that predicted from the amino acid sequences (Yang et al., 1988). This discrepancy in the molecular sizes of the m e m b r a n e protein identified by affinity labeling techniques and the o n e encoded by the cloned cDNA would suggest that either there was substantial proteolytic degradation of the carrier protein during th e affinity labeling studies or the translated protein had u n d e r g o n e proteolytic modification prior to function. The latter does not seem to occur with the human RFC because, in experim ents where a n epitope tag was inserted into various region of the carrier, t r a n s p o r t function was found to be normal in those insertion m utants t h a t had the amino and carboxyl terminus tagged (Ferguson and Flintoff, 1999) On the basis of the hydropathy profile of the protein, RFC-1 appears to belong to a superfamily of m em b ran e proteins, the 12- transm em brane helix transporters (Dixon et al., 1994; for review see Henderson, 1993). Other members of the superfam ily include th e human glucose transporter (Glutl), the bacterial lactose-H + symporter, and the bacterial amino acid transporters (H en d erso n , 1993). The topology of the hum an RFC has recently been studied by single insertion of hem agglutinin epitope into various sites of th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 196 protein, and by probing with anti-epitope monoclonal antibodies in n o n-perm eabilized and permeabilized cells that expressed t h e insertion m utants. The results confirm the 1 2 - tr a n s m e m e b r a n e topology predicted based on a hydropathy profile. Thus, the a m i n o and carboxyl terminus, the loops between transm em brane segment 4 and 5, 6 and 7, and 10 and 11 are intracellular, whereas the lo o p between segment 11 and 12 is extracellular ((Fig. 4.1.), Ferguson a n d Flintoff (1999)). Analysis of the primary amino acid sequence of t h e carrier does not reveal any definable nucleotide binding site, consistent with the observation that carrier function is not d ire c tly linked to A TP hydrolysis (Goldman 1969). The genomic organization of the RFC gene has been r e p o r t e d for ham ster (Murray et al., 1996), murine (Brigle et al., 1997) a n d human (Tolner et al., 1997). The hamster gene contains 7 exons a n d spans a total of 15.3 kb. Two alternatively spiced mRNAs of e q u a l abundance were identified in CHO cells. The longer 2.6 kb m essag e contains all 7 exons, whereas the shorter 2.5 kb message lacks e x o n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 197 glycosylation site 1 0 * V k ■ (OUTSIDE) ! ’ ’ / \ [ ! A I I , « « , { " « S E p T *M E 0 « V 0 ht p 8 » « b 0 0 V V L V R „ L S - H Y < C ^ ‘l 'o- v s *U L LL \ K Ay L. A, v V ■ V..l \ X v,f V ; . 0 T L a , a Y A 0 A 0 A A 1 1 0 'V * A "L "A -V F A I 0 P If T A ' v s -■?'■ < v1 ?■?** \ \ W - V \ W C \ ■ " \ N Iv ” ■ " “ ■ ■ « , R a . V ^ M l° ; ; A « ' L K L E ? ^ ‘ P s ^ P P 0 ! j / V j * 1 1 1 ; | t • R i S R O n* A R \ / £ a R 0 I V f f c F X r k * w I ? V f ? 3 S “ 1 8- “ 5 i i r p 8 t 3 "5 3 I ? 9 » E 8 HOC D A J J j! V e o 0 a p 2 0 { ( * » 9 f ki ! ! 1 R v ! N A a N ^ " * k ! i a \ 0 4 t (C 0 0 I,) U L ; Q ft 1 G p G ® * I ^ { a8 rl3 ' Tpg p Fig. 4.1 Proposed topology of human RFC. The predicted topology was based upon hydro p ath y analysis of RFC and results of epitope-tagging experiments using h e m a g g lu tin in insertions into different sites of the transporter (arrows). Sim ilar hydropathy profiles were obtained in m urine and hamster RFCs (from Ferguson and Flintoff, 1999). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 198 2 in the 5’-UTR but encodes the same protein. The prom oter region of the gene contains no TATA box-like element but has a c o n se n su s Spl binding site. (Murray et al., 1996). The murine and h u m a n genes are both com posed of six exons and span 21.4 and 13.1 kb, respectively. Five splice variants have been identified in human RFC- 1 transcripts, each a different 5 ’ UTR but containing the same c o d in g region (Tolner et al., 1997). Four species of murine RFC-1 tra n s c rip ts generated as a result of alternate splicing have been reported; one of which contain an “in fram e” deletion in exon 3 and the other tw o incorporate alternates to exon 1 and 5, respectively (Brigle et al., 1997). The physiological significance of these splice variants has y et to be elucidated. Biodistribution of RFC On the basis of transport kinetics, RFC expression has been dem onstrated in a wide variety of rodent and hum an d e r iv e d cultured cell lines as well as in a number of norm al tissues in c lu d in g small intestine, rat hepatocytes and rabbit kidney (reviewed in Sirotnak, 1985). Using antibodies against the protein, Matherly et al. (1992) has reported that the transporter was ubiquitously expressed in both normal and neoplastic murine tissues. However, there h a s been some skepticism about the specificity of the antibodies used in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 199 those im m u n o h isto ch em ical studies. Nonetheless, with t h e m olecular cloning of the RFC, more specific biochemical probes c a n be generated and definitive data on the biodistribution of the c a rrier should be available shortly. Membrane folate binding protein-mediated transport B ioch em istry The membrane folate binding protein (mFBP) or folate re c e p to r (FR) represents an alternate route for folate uptake (reviewed i n Antony, 1992; and, Antony, 1996). Folate binding proteins in itially referred to a class of proteins that exhibited no intrinsic e n z y m a t ic activity but bind folic acid with high affinity. These folate b in d e rs exist in both a m em brane-associated particulate form (mFBP) and a soluble form (sFBP). The m em b ran e-asso ciated form is anchored to the outer layer of plasma m em brane by a glycosyl- p h o sp hatidylinositol (GPI) linkage; whereas, the soluble form is probably derived from cleavage of the GPI-linked precursor by t h e action of a M g2+-dependent protease (Antony et al., 1989). The role of FR in mediating transport of the physiological folate, 5- m eth y ltetrah y d ro fo late, was first d em o n strated in KB cells (K a m e n and Capdevila, 1986) and later confirmed in monkey kidney M A 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 200 cells (Kamen et al., 1988). It has been proposed that th e internalization of folate ligands takes place by a process ca lle d "potocytosis" via the following sequence of steps: ( 1) high -affin ity binding of folate to receptor; (2) a coated p it-in d ep en d en t, receptor- mediated endocytotic m echanism in which ligand-bound recep to rs move into m em brane invaginations known as caveolae; (3) dissociation of the folate from the receptor in response to acidification followed by m ovement of released folate into th e cytoplasm via anion channels; (4) m etabolism of folate in to polyglutam ates that become retained within the cell (Anderson et al., 1993; Kamen et al., 1991; and Rothberg et al., 1990). However, this potocytosis model of folate uptake has recently been ch allen g ed , interestingly, by the initial proponents of the model. Thus, Mayor et al. (1994). have suggested that the sequestration of FBP in cav eo lae might be artifactually induced by cross-linking antibody used for detection and not by folate binding per se. The calculated cycling rate of the receptor has been found to be 100-fold slower than that of the RFC (Spinella et al., 1995). However, when expressed to a sufficient level, the receptor can allow efficient accum ulation of cofactors present at low e x tra cellu lar concentration (10-50 nM) by virtue of a strong re c e p to r-lig a n d interaction. The folate receptor exhibits tight binding with folic acid, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 201 DDATHF, and 5-methyltetrahydrofolate (K b = 0.1-1 nM, 2-38 nM, a n d 3-30 nM, respectively) but a relatively lower affinity for 5- fo rm yltetrahydrofolate and MTX (Kb = 13-1200 nM and 100-1700 nM, respectively) (Kamen et al., 1988; Brigle et al., 1994; Jansen et al., 1989). Molecular genetics Three hum an mFBP isoforms, F R -a (Coney et al., 1991), FR- (3 (Elwood, 1989), and FR-y (Shen et al., 1994) have been id e n tifie d and their cDNAs cloned. FR-a was initially reported as a n overexpressed surface marker in ovarian cancer (Campbell et al., 1991) whereas FR-P was isolated from nasopharyngeal carcinoma KB cells and hum an placenta; both FBP's are GPI-anchored m e m b r a n e - associated proteins. On the other hand, FR-y was found to be a secretory form predominantly expressed in hem atopoietic cells (S h en et al., 1995). In addition, two cDNA's, FBP1 and FBP2, which a re mouse hom ologue of FR-a and FR-P , respectively, have been c l o n e d (Brigle et al., 1991). There is 68% homology in amino acid se q u e n c e between FR-a and FR-p. The two isoforms exhibit sig n ifican tly different binding affinities and stereospecificities for the v a r io u s folate co m p o u n d s (Wang et al., 1992; Brigle et al., 1993). For Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 202 example, FR-a bound (6 S )-m ethyltetrahydrofolate and (6S)-5- fo rm y tetrah y d ro fo late more avidly than FR-P did by 50- and 100- fold, respectively. However, the affinity for the (6 R )-d ia ste re o m e rs was only 2- and 4-fold higher. Conversely, FR-P bound the (6 R )-form s tighter than the (6 S)-forms. DDATHF showed a 10-fold h i g h e r affinity for FR -a (or murine FBP1) without significant stereo sp ecificity (Brigle et al., 1993). The human FR-a, FR-P and two related genes o r pseudogenes are located within a 140-kb region in c h r o m o s o m e 1 lq 13 (Ragoussis et al., 1992). B io d istrib u tio n Using monoclonal antibody against FR-a, the expression of th is isoform was detected in a wide variety of normal and m a l i g n a n t human tissues. Normal tissues that showed intense re a c tiv ity include the epithelium of choroid plexus, fallopian tube, uterus, a n d epididymis; alveolar lining of type I and II pneum ocytes; acinar cells of m am m ary , subm andibular and bronchial glands; and p r o x i m a l tubules of the kidneys (Weitman et al., 1992). Within the c h o r o i d plexus, bilam inar pattern in both luminal and basal surfaces w as noted in some foci. Limited but focal reactivity was found in the v as deferens, ovary, thyroid, and pancreas (Weitman et al., 1992). M alignant tissues where consistently high expression of FR-a were Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 203 detected by im m u n o h isto ch em ical means include n a s o p h a ry n g e a l, ovarian, and cervical carcinoma, as well as c h o r io c a r c in o m a (Antony, 1996). Physiological functions of RFC and mFBP. Comparison of amino acid sequences of the RFC and mFBP d o e s not reveal significant similarity between them. As h ig h lig h te d earlier, the m echanism s of transport of folate com pounds are also fu n d am en tally different between the two systems. A number of im portant questions arise: ( 1) what physiologic function does e a c h folate transporter carry? (2 ) is there any functional i n t e r a c t io n between the two systems? (3) is there any specific coupling b e tw e e n RFC- or FR-m ediated transport to other intracellular processes of folate m etabolism such as poly g lu tam atio n ? (4) what is the role of RFC or FR in antifolate resistance? The high affinity, concentrative transport properties of mFBP has made it an attractive candidate for transcellular delivery of folate against a concentration gradient from one com partm ent to another that are separated by an epithelium . Thus, it had b e e n proposed that FR-mediated m echanism is responsible for transplacental transport of folates to the fetus (Henderson et al., Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 204 1995) and also folate reabsorption across proximal renal tubular cells (Selhub, 1994). The abundant expression of FR in epithelial tissues such as placenta, proximal kidney tubules and choroid plexus is also consistent with this proposed function of transcellular transport of folates. On the other hand, the ubiquitous expression of the RFC would favor its role as the primary transporter for uptake of fo late for intracellular utilization. Functional coupling between the RFC and FR was in itially proposed by Kamen and co-workers (Kamen et al., 1991). T h e observation that probenecid, a known inhibitor of RFC function, also interfered with FR-mediated internalization of folate in MA104 cells has p rom pted these investigators to hypothesize that the RFC w as involved in trans-caveolar translocation of ligand to the cytosol a fte r it had been released from the FR. However, recent studies by tw o laboratories have argued against such a model. Thus, Spinella et al. (1995) has reported that probenecid also inhibited F R -m e d ia te d folate uptake in a dose-dependent manner. Furthermore, Dixon et al. (1992) have shown that, after transfection of FR-a cDNA, w ild -ty p e and RFC transport-defective human breast cancer cells were able to accum ulate similar amounts of folic acid, 5 - m e t h y l t e t r a h y d r o f o l a t e and MTX, suggesting that FR could work independently of RFC. Taken together, FR does not appear to work in tandem with the RFC. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 205 In an attempt to determine whether there is any c o u p lin g between FR- or RFC-mediated transport and p o l y g l u ta m a t io n , Goldman and co-workers compared the rate of accum ulation of MTX polyglutam ates in RFC transport-defective L1210 cells tra n s fe c te d with either RFC-1 or FR-a cDNA (Spinella et al., 1996). It a p p e a r e d that the formation of MTX polyglutam ates depend only on the level of intracellular MTX m o n o g lu tam ate achieved, irrespective of th e route of entry. Therefore, no apparent preferential linkage betw een route of transport and polyglutamation could be demonstrated. Defects in membrane transport have been im plicated as a com m on mechanism of acquired resistance to antifolates both in vitro and in vivo (reviewed in Sirotnak, 1987; see also Chapter 1). Alterations in the RFC, in particular, have been well d ocum ented in tumor cells resistant to MTX; changes such as increased K m (S iro tn ak et al., 1968; Jackson et al., 1976), decreased V m ax> (Sirotnak et al., 1981; Schuetz et al., 1988; ) and combined kinetic defects (S iro tn ak et al., 1981; Niethammer and Jackson, 1975) have all been re p o rte d . The molecular cloning of putative RFC genes now allows th e identification of specific m utations that are responsible for th o se phenotypic alterations. Such analysis would undoubtedly p ro v id e invaluable insights into the structure-function relationship about th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 206 protein(s) involved in folate transport as well as the genetic c o n tro l of the RFC. The role of FR in antifolate resistance is less clear. The o n ly report was the one by Elwood and co-workers where reduction in FR expression was noted to correlate well with the extent of MTX resistance in KB cells that primarily depend on FR for fo late transport (Saikawa et al., 1995). Efflux systems of folates The steady state level of a folate permeate is determ ined by three processes: influx, efflux, and intracellular m e ta b o l i s m . Transport of folate compounds out of the cell is in part mediated by the bidirectional exchange function of the RFC. However, the m a j o r efflux route is mediated by an energ y -d ep en d en t pump system t h a t utilizes the energy derived from ATP hydrolysis to transport fo late out of the cell (Goldman et al., 1968; Goldm an, 1969). It has b e e n reported that the loss of this folate efflux activity was associated w ith an au g m en ted folic acid accum ulation in CHO that was re sp o n sib le for resistance to the lipophilic DHFR inhibitor, p y r i m e t h a m i n e (Assaraf and Goldman, 1997). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 207 In this chapter, we report the genetic changes acquired in t h e RFC in L1210 cell lines that had been selected for progressive resistance to DDATHF. Such changes included point m utations i n one allele of the gene causing functional change in the recognition of substrates for transport, as well as the loss of expression of t h e normal allele. To date, this phenotype is unique among all of t h e variant cells selected for antifolate resistance showing alterations i n the RFC system. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 208 CHAPTER 4 MATERIALS AND METHODS C hem icals [3',5',9-3H]M ethotrexate and [3',5',9-3H ](6S)-5- form yltetrahydrofolate and [3',5',9-3H]folic acid were p u r c h a s e d from Morevek Biochemicals (Brea, CA). [^H](6R)-DDATHF was synthesized in this laboratory as described in Chapter 3. All radiolabelled folates were purified by hplc on a 10 x 0.32 cm C l 8 reverse phase colum n of 3 pm pelicular size (Applied B iosystem s) and were stored at -20 °C for no more than 2 weeks prior to use. [^Hjfolic acid and [3H ](6S)-5-form yltetrahydrofolate were p u r ifie d using chro m ato g rap h ic conditions described in Chapter 3. [^H ]M ethotrexate was eluted using a linear gradient of 12 to 15% MeOH in 0.03M sodium acetate buffer over 15 min. Monensin was from CalBioChem (San Diego, CA). Total RNA isolated from tw o mouse erythroleukem ia (MEL) sublines that overexpress folate binding protein isotype a and P (FBP-a and FBP-P), respectively, were kindly provided by Dr. Goldman of Albert Einstein Cancer Center, as R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 209 were the bacterial clones harboring the corresponding cDNAs for FBP-a and FBP-p. Restriction enzymes were form Promega (Madison, WI). Oligonucleotide primers used for sequencing or PCR were o b t a i n e d either from the microbiology core facility at Medical College of Virginia or from Gibco-BRL (Gaitherburg, MD). Primer sequences are shown in Table 4.1. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 210 Table 4.1 Sequences o f oligonucleotide primers used f o r PCR and sequ en cin g an alysis P r i m e r n a m e P r i m e r s e q u e n c e s e n s e (S ) o r a n t i s e n s e (A ) A T I TGATGGATCCTGAGACCTGGGCAACATG (S ) A T2 GTCACTCGAGTCAGAAGGATAAGCTAACAG (A) AT3 CATGAAGTAGGCTGGGTTTAG (A) A T4 TTGGCTTCGACTCTTAAGTC (A) A T5 AGGCAAAGCAGGTGATAGAC (A) A T 6 GTCCCGGATCTGGAACATACAG (A) A T7 TTCCACCAGTTCACTTAGC (A) AT 8 ATATGTCCTACGGTGACCAG (A) AT 9 ACCAAGACTGGCTTGTATCGC (A) A T I 0 TGAGTTGCATGTGCACGACAG ( g e n o m ic PCR) (A) A T I 1 TGTGCAAGCGATAAGCCTAC (S ) AT 12 GCCTGGTGTTCTATCTCTCTGCTTCTTCGG ( g e n o m i c (S ) AT 5 4 CCTGCGATACAAGCCAGTCT PCR) (S ) AT 5 5 CATATTCTCCCTGGTGCACC (S ) AT 5 6 CTGTGAGCTGGATCAGATGC (S ) AT 5 7 ACACTGCTGAGCGCCATCAC (S ) AT 5 8 GGATCAACACTTTCCTAGCT (S ) M13 L 1 A - - 1 R EVE R SE M l 3 u -t J FORWARD CAGGAAACAGCTATGAC GTAAAACGACGGCCAG Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21 1 Cell culture M y c o p l a s m a -free mouse L1210 cells were passaged in RPMI 1640 m edium supplem ented with 10% dialyzed FCS unless o therw ise stated. The derivation of DDATHF-resistant L1210 sublines was described in detail in Chapter 3. The M T X rA, a tra n sp o rt-d e fe c tiv e cloned subline that was isolated by selection of parental L1210s w ith methotrexate, was kindly provided by Dr. Goldm an of Albert Einstein Medical College (Bronx, NY). Growth inhibition studies were performed as described in Chapter 3. Transport studies Transport experim ents were perform ed as described in "Materials and Methods" in Chapter 3. T ransport rates of (6R)- DDATHF, MTX, (6S)-5-CHO-THF, and folic acid were measured over 4,2,1, and 2 min, respectively, to approxim ate initial velocities. M easurem ent of folic acid influx was made in cells pre-treated w ith 10 pM trim etrexate for 10 min prior to transport studies to p re v e n t metabolism by dihydrofolate reductase. In experim ents with (6S)-5- fo rm yltetrahydrofolate ((6S)-5-CHO-THF), [^H]-(6S)-5-CHO-THF w as added to a mixture of non-radiolabelled (6R,S)-5-CHO-THF a n d kinetic constants were determined based on concentrations of th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 212 natural (6 S)-diastereom er in the mixture. For experim ents on th e effect of m onensin on folic acid transport, cells were pretreated w ith trimetrexate for 10 min, washed, and incubated with 25 pM monensin for 30 min at 37 °C and washed again with buffer before transport of folic acid was initiated. Isolation of total RNA from cultured cells Total RNA was isolated from cells by either a m o d ifie d guanidium isothiocynate method (Puissant and Houdebine, 1990) or using the Trizol Reagent (Biorad, Hercules, CA)) according to th e m anufacturer's protocol. All solutions were m ade using b a k e d glassware and HPLC grade water. The former m ethod was used in all cases except for RNA isolation from the L1210 transfectants. Briefly, =5 x 10^ log-phase cells were pelleted by centrifugation at 500 x g and were mixed with 15 ml of 4M guanidium thiocyanate buffer containing 1M sodium citrate, pH 7.0, 20% sarkosyl, and 0.1M (3-ME. Cells were broken open by shearing with a 19 gauge needle. One a n d a half-milliliter of 2M sodium acetate, pH 4.0, were added and th e tube was mixed. To this solution, 15 ml of water saturated p h e n o l was then added followed by 3 ml of C H C I3 . The mixture was transferred to a 30 ml Cortex tube and centrifuged at 6000 x g for 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 213 min at 4 °C. The upper aqueous phase was collected and an e q u a l volume of isopropanol was added. The solution was kept at -20 °C for at least 30 min. The precipitated RNA was pelleted by centrifugation at 6000 x g for 10 min. After removal of th e supernatant, the pellet was washed with 70% EtOH and was p a r tia lly air-dried. The RNA pellet was resuspended in 500 pi of 10 mM Tris buffer, pH 7.5, containing 1 mM EDTA, and 0.5% SDS. The so lu tio n was then transferred to a 1.7 ml Eppendorf tube and extracted w ith 500 pi of acidic p h e n o l:C H C l3 (1:1) followed by another 500 pi of CHC13. Extraction with CH Clj was repeated until the interface was clear. To the final aqueous phase collected, 1/10 volume of 2M sodium acetate, pH 5.0, and an equal volume of isopropanol was added. The solution was kept at -20 °C for at least 30 min. T h e precipitated RNA was centrifuged at 14,000 x g for 2 min. The p ellet was washed with 70% EtOH, partially dried, and resuspended in 200- 400 ml of 10 mM Tris, pH 7.5, 1 mM EDTA buffer. The c o n c e n t r a t i o n of RNA was determined spectrophotom etrically at 260 nm, using th e relationship 1 O.D. unit equals 40 mg/ml. The purified RNA so lu tio n was stored at -70 °C after addition of three volumes of 100% EtOH and was found to be stable for up to 18 months. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 214 Northern blotting RNA was denatured with glyoxal and DMSO (Sambrook et al., 1989) before fractionated on a 1% agarose gel in 10 mM p h o s p h a t e buffer, pH 7.0. Gels were run at =35 mA to a total of =650 volt-hr. The RNA was transferred to nylon membrane (Biotran™ from ICN) by capillary action in the presence of 20 x SSC (3M sodium ch loride, 0.3M sodium citrate, pH 7.0). The blots were treated with a 0 .1 2 Joules of UV radiation at 254 nm (Stratalinker™ a p p a r a t u s , Stratagene (La Jolla, CA)) to crosslink the RNA to the m e m b r a n e . RNA standards were visualized by pretreating an excised portion of the m em brane in 5% acetic acid followed by staining with 0.04% methylene blue in 0.5M sodium acetate (pH 5.2). Blots were prehybridized in 0.5M sodium phosphate, pH 7.0, 7% SDS, 1% BSA, and 1 mM EDTA (Church and Gilbert, 1984) for at least 3 h to overnight at 63 °C. Radiolabelled probes were then added a n d hybridization was carried out overnight at 63 °C in the same buffer. Blots were then washed with 2 x SSC, 0.2% SDS for 20 min at 50 °C and these washes were repeated until the background r a d io a c tiv ity was below 200 c p m / c m : ; for most experiments, three washes were done. Blots were then exposed on films (Kodak). Q uantitation of transcript abundance was performed using a M olecular D y n a m ic s 445 SI Phospholmager (Sunnyvale, CA). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 215 Preparation of DNA probes Radiolabelled DNA probes used for northern blots were prepared using a random primer DNA labeling system (Gibco-BRL, Gaithersburg, MD). Briefly, 1-2 ng DNA insert were boiled for 5 m i n and then im mediately chilled on ice. The DNA solution were t h e n incubated with dGTP, dTTP, [32p]dATP and [32p]CTP (for p re p a rin g the probe for human GADPH, only dATP was radiolabelled) in t h e presence of random hexamer primers and Klenow fragment at r o o m temperature for 30 min. The radiolabelled DNA was separated fro m the uninco rp o rated dNTP's using a 1 ml Sephadex G50 gel filtratio n column equilibrated in TE, pH 8.0. Sample was eluted by centrifuging of the column for 1 min at 500 rpm (Beckman J6, Fullerton, CA). The eluted probe was then boiled for 5 min follow ed by rapid chilling on ice and was ready to be used for hybridization. DNA ligation reaction In a typical ligation reaction, a known am ount of DNA was precipitated by ethanol, pelleted by centrifugation 14,000 rpm in a m icrocentrifuge (Eppendorf) for 10 min, washed once with ice c o ld 70% ethanol, and reconstituted in buffer containing 10 mM MgCH, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 216 10 mM DTT andl mM ATP in 30 mM Tris-HCl (pH 7.8) and 1-3 u n its of T4 ligase (Promega, Madison, WI) in a total volume of 10 pi. T h e reaction was incubated at 12 °C overnight. Purification of DNA fragments from agarose gels Recovery of DNA fragments after fractionation by gel electrophoresis was performed using the Geneclean II system (Bio 101 (Qbiogene). Carlsbad, CA). The DNA fragments were visualized by ethidium bromide staining and the DNA band of interest was excised from the agarose gel. The gel slice was then treated with 3 v o lu m e s of 4M sodium iodide to dissolve the agrose matrix. Ten microliter of Glassmilk™ was added and the mixture was kept on ice for 10 m i n to allow binding of DNA to the glass particles in the G lassm ilk™ . After centrifugation at 140,000 x g for 5 s, the sodium io d id e supernatant was aspirated. The pellet was washed and spun th re e times using 500 pi of sodium acetate-ethanol based buffer. The DNA was eluted twice from the Glassmilk™ by incubation in 10 pi of w ater or TE at 50 °C for 5 min. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 217 Amplification of RFC cDNA from wild-type and resistant L1210 cells by RT-PCR P o ly (A )+ -selected RNA was isolated from 100 pg of purified to ta l RNA using the oligo-dT column provided in the FastTrack™ s y ste m (Promega, Madison, WI). First strand synthesis (reverse t r a n s c r i p t io n ) was carried out utilizing reagents in the 5' RACE kit (GIBCO Biologicals, Inc, Grand Island, NY) on one-half of this p o!y(A )+ - selected RNA and 130 ng of a RFC-specific primer (AT3) co m p lem en tary to sequences in the 3' untranslated region of th e message based on the published cDNA (Dickson et al, 1994). T h e reversed transcribed cDNA was purified on a GlassMAX c a rtrid g e according to the m anufacturer’s protocol (Stratagene). The c o d in g region of the RFC-1 cDNA was amplified by PCR using primers ATI and AT2, containing engineered restriction sites for B am H l and X hol, respectively. PCR conditions: 4.5 pi of cDNA products in 10 mM KC1, 20 mM Tris-HCl, pH 8.8, 10 mM (N H 4)2S 0 4 1, 1.5, or 2 mM M g S 0 4; 0.1% Triton X-100, 200 mM dNTP’s, 180 ng of each primers and 3 units of Taq polymerase (Promega) in a 60 pi reaction. Samples were overlaid with mineral oil to prevent evaporation. After an in itial melting at 93 °C for 3 min, am plification was carried out in a thermal cycler (MJ Research, Waltham, MA) for 3 cycles of 60s at 94 °C, 60s at 52 °C, and 240s at 72 °C; for another 35 cycles of 60s at 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 218 °C, 60s at 58 °C, and 240s at 72 °C: and a final extension step at 75 °C for 5 min. For each cell line, products derived from th re e independent PCR experiments were pooled. The pooled PCR p r o d u c ts were purified by gel electrophoresis. Attempts for u n id ir e c tio n a l cloning of PCR products directly into expression vector using th e B a m H 1 and X ho 1 restriction sites were unsuccessful. Thus, PCR products were ligated instead into the pCRII TA-cloning vector as described by the m anufacturer (Invitrogen, Carlsbad, CA). Vectors harboring the correct 1800-bp insert were prepared in large scale after confirmation of authenticity using the colony PCR technique as follows: following transform ation of com petent E. coli strain INVaF (Invitrogen), colonies were picked from a re-streaked plate using a sterile pipette tip, and were then suspended in 50 pi of sterile w ater. Polymerase chain reactions were performed using the s a m e conditions as m entioned above. Each reaction contained 10 pi of bacterial cell suspension, 100 ng of each ATI and AT2 primers, a n d 1.5 mM M gCl; in a 50 pi reaction mixture. The authenticity of th e PCR products was verified by diagnostic restriction with S m a I. In a separate set of RT-PCR experiments, only the sequences that s p a n n e d the two m utated codons were amplified using primers ATI a n d AT10. In these experiments, 5 pg of total RNA from L1210, L1210/D0.5 and L1210/D3 cells were subjected to reverse Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 219 transcription followed by PCR amplification using the s a m e condition em ployed for isolation of fragments that spanned th e entire coding region. A single 350-bp PCR fragment was obtained in each reaction, the fragments were purified and cloned into pCRII vectors as described above. Three PCR clones from L1210/D0.5 cells and one clone from L1210 and L1210/D3 cells were sequenced. Sequencing analysis The nucleotide sequence of the cloned PCR products was determ ined by a dideoxy chain term ination m ethod (Sanger et al., 1977) using Sequenase 2.0 (United States Biochemical C o rp o ra tio n , Cleveland, OH) according to the m anufacturer's protocol. - Five microgram of double stranded DNA plasmid was used in e a c h reaction. The samples were electrophoresed on a 6% d e n a tu r i n g polyacrylam ide gel in 1 x TBE buffer (0.09 M Tris borate, 1 m M EDTA). The complete sequence of three clones from L1210 cells a n d that of seven clones from L1210/D3 cells were d e te r m in e d . Deviations from the published sequence were confirmed by sequencing runs from the opposite direction. For R F C -1 cDNA d e riv e d from L1210/D0.5 cells, sequences spanning the two mutations at n t 183 and 354 (see Results) were determ ined from three of the RT-PCR clones. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 220 Construction of expression vectors To construct the various RFC-1 expression vectors, two pCRII plasmid clones, 0-3 and 3-7, containing RFC-1 cDNA derived fro m wild-type and L1210/D3 cells, respectively, were digested by B a m H l to release the cDNA inserts. The expression vector, pcDNA3, was c u t with the same restriction enzyme and was treated with calf in te s tin e phosphatase (5 pg DNA/0.05 units of enzyme) in order to reduce th e frequency of self-ligation of vector. Approximately 30 ng of gel- purified cDNA insert was mixed with 50 ng of linearized vector in a 2:1 molar ratio in the ligation reaction. T ransfom ation of com petent E. coli strain INVaF' (Invitrogen) was carried out as described by the manufacturer. These cDNA clones were s e q u e n c e d in their entirety and were free from PCR-derived fidelity errors. T o construct RFC-1 expression vector harboring only one of the tw o mutations found in L1210/D3 cells, 5 mg of pCRII plasmid clone 0-3 and 3-7 were digested with B a m H l and Pst 1, which allowed th e isolation of the 1800-bp coding region of the RFC-1 cDNA as a 3 0 0 -b p fragment containing the mutation at codon 48 and the 5’- portion of the cDNA; and a 1500-bp fragment containing the mutation at c o d o n 105 and the remainder of the cDNA. These two fragments were gel- purified and one fragment harboring each of the two m utations was Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22 1 mixed with the opposite fragment from wild-type RFC-1 and th e mixtures were used in triple ligation reaction into pcDNAIII. However, it was later decided that transfection of L1210 cells s h o u l d best be carried out using the pTK-PGK vector (Brigle et al., 1995) for which optimal experimental conditions had already b een established. Thus, the cloned RFC constructs were released from th e pcDNAIII vector with B a m H l and X h o I, the fragm ents were gel purified, treated with Klenow fragment, and ligated into the pTK-PGK vector that had been linearized with £c/136II. In this vector, expression of RFC-1 constructs was driven by the m u r i n e phosphoglycerate kinase promoter; a neom ycin resistance cassette was also present for selection of stable transfectants. Escherichia coli strain MC/1061/P3 was transfected with these expression p la s m id s , and the orientation of the insert was verified by restriction d ig estio n and a subsequent sequencing reaction. Transfection of L1210 and MTXrA cells Expression vectors harboring different RFC-1 constructs were transfected into wild-type L I 210 and M TXrA cells. In a ty p ic a l transfection experiment, =1.1 x 10? exponentially growing re c ip ie n t cells were harvested and resuspended in 0.8 ml of DEAE-dextran Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 222 buffer containing 40 pg of pTK-PGK plasmid or the same p l a s m i d containing an insert representing wild-type, double mutant, or single mutant RFC-1 cDNA. The suspensions were subjected to electroporation at 250 V and 350 microfarads, following which each transfection mixture was brought to 10 ml with RPMI 1640 m e d i u m containing 10% FCS, and were incubated at 37 °C for 36 h to allow for recovery and expression of the neomycin resistance gene. Each cultures was adjusted to 1 x 105 cells/ml in complete m e d i u m containing 20 p M (3-ME, and G418 was added to a final c o n c e n t r a t i o n of 800 pg/ml; 200 ml of cell suspension were distributed to five 98- well plates at =20,000 cells per well. After 8-10 days, the num ber of wells containing G418-resistant cells, as judged by an o b v io u s increase in healthy-looking cells, was scored. To facilitate th e identification of stable transfectants that expressed exogenous RFC-1, all MTXrA transfectants that were G418-resistant were subjected to a quick screen with 50 nM MTX. At this drug concentration, MTXrA cells could proliferate at ease, whereas transfectants that ex pressed functional exogenous RFC would become growth inhibited. C lonal transfectants were then obtained from the MTX-sensitive cultures by the limiting dilution method and were expanded for further s tu d y . This screening method could not be applied to the wild-type L1210 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 223 transfectants; therefore, in those cases, cells from several wells w h ich showed G418-resistance were random ly selected for dilution c lo n in g and the expression of exogenous RFC-1 was later verified by n o r t h e r n blots. The sequence of steps in developing the various stable transfectants was depicted in Fig. 4.6. Dilution cloning of L1210 transfectants L1210 Transfectants were cloned by limiting dilution; Each cell population of interest was serially diluted to a final density of 1 cell/ml in RPMI 1640 medium containing 20% FCS, 800 mg/ml G418, and 20 mM p-ME (the presence of P-ME was found to be crucial for this procedure). An average of 0.2 cell (200 pi of diluted culture) was distributed into each well of a 96-well plates and one cloned cell line was expanded for further study from each cloning plate. Genomic PCR Genomic DNA was isolated from L1210 and L1210/D3 cells by PCR using primers which straddled the position of both m u t a t i o n s and an intron (Tolner et al., 1997)., to ensure that the products were not representatives of cDNA co ntam inants. The primers used were; AT 12, a sense primer in exon 2 of m urine RFC-1 and AT 10, an a n ti- sense primer in exon 3. A mixture of T a g polymerase (2 units) a n d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 224 Pfu polym erase (1 unit) was used in a 50-pl reaction containing 2 mM M gC l; 0.2 mM of each deoxynucleotide triphosphate, and 60 n g of each primer. The reaction was initially held at 94 °C for 30 s, annealing at 55 °C for 30 s, and elongation for at 68 °C for 3 m in , and then held at 72 °C for 5 min after a total of 35 cycles. The single 3.4-kb PCR products were isolated on a low melting temperature Tris- acetate-agarose gel, and the excised band was purified on a W iz a rd PCR purification system (Promega) according to directions s u p p l i e d by the m anufacturer. The primers used to amplify these PCR products were end-labeled with [y-33P]ATP using T4-kinase and t h e end-labeled primers were used to prime cycle sequencing re a c tio n s on the purified PCR product using Fmol cycle sequencing re ag en ts (P ro m eg a). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 225 CHAPTER .4 RESULTS Substrate specificity of folate transport in wild-type and DDATHF-resistant L1210 cells L1210/D3 is a highly resistant clonal subline selected by continuous exposure of parental L1210 cells to in cre asin g concentrations of (6R,S)-DDATHF. The m echanism of resistance h a s been d eterm in ed to involve alterations in the m e m b r a n e translocation of folates in these cells, primarily due to an increase in the transport of folic acid (Chapter 3; Tse and Moran, 1999). T o further characterize the underlying biochemical changes, the influx kinetics of various folate com pounds were com pared between wild- type and resistant cells. (Table 4.2). Overall, significant differences in transport kinetics were observed with all the folates s tu d ie d , although the extent of changes varied with each substrate. T h ese changes seem ed to involve primarily the influx M ic h a e lis -M e n te n constants (K m ) rather than the influx V m a x . The influx K m for (6R)- DDATHF was moderately increased (3.3-fold) (Table 4.2; Tse a n d Moran, 1999), whereas that for MTX was slightly (1.6-fold) d e c re a s e d in the resistant cells. Remarkably, a 29-fold reduction in K m was observed with folic acid in L1210/D3 cells (Table 4.2), which Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4.2 Transport o f fo la te derivatives into L1210 a n d L I 2 10/D3 cells L 1210 L 1210/D 3 K H L - V m : i y k' t K UL. Vm; mux T ransport Substrate D D A TH F (3) M TX (2) Folic Acid (2) (/iM ) (p m o l/lO ^ cells x min) 1.7 ± 0.6 0.93 ± 0 .1 4 3.3 ± 0.1 0.53 ± 0.02 (340)* not m easurable* k ' t (/iM ) (pmol/lO^ cells x min) 0 .5 5 5.7 ± 0 .1 6 1.23 ± 0 .1 2 0.22 0 .1 6 2.0 ± 0 .1 0.60 ± 0 .0 4 0.30 5 x 10’4 * 19.7 ± 2.9 1.20 ± 0.27 0.061 4 k' represents the First order rate constant for transport, defined as the V m a x /K m and is in units o f (p m o l/1 0 ^ cells x m in) /iM 'l. W ell below the K m for transport, the rate o f entry into the cells w ould be represented by k' x substrate concentration, and hence, k' reflects the efficiency o f transport at low concentrations o f each substrate. T he transport o f folic acid into L1210 cells did not saturate at achievable concentrations o f folic acid in L1210 cells, disallow ing m easurem ent o f eith er K,„ or Vm a, in these experim ents. T he value listed in parentheses for K,„ w as actually a K, derived by a D ixon analysis o f the inhibition o f D D A T H F transport by folic acid; for com parison, the K, value for folic acid in L 1210/D 3 cells was 18 /iM . The k' value listed w as estim ated from the initial slope o f v vs S plots such as Fig. 4.2 226 227 accounted reasonably well for the 40-fold increase in the in itia l transport of this folate substrate in these cells (Chapter 3; Tse a n d Moran, 1999). Interestingly, the transport of (6S)-5-CHO-THF w as noted to be significantly reduced in the resistant line, involving b o t h a 2-fold increase in K m and a mild decrease in V max (Fig. 4.2). T h is decrease in (6S)-5-CHO-THF transport therefore provided a n explanation for the parallel diminution in total folate pools in th e se cells compared to parental L1210 when both cell lines were grown in media containing (6S)-5-CHO-THF as the folate source (Chapter 3; Tse and Moran, 1999). Taken together, kinetics analysis r e v e a le d alterations in substrate specificity of folate transport in L 1210/D 3 cells that were consistent with the observed phenotype of these cells. Studies were then undertaken to identify the transport system a n d the molecular change(s) causative of these alterations. Inhibition of folic acid transport in L1210/D3 cells by MTX and monensin and northern blot analysis of expression of mFBPl and mFBP2 in L1210 and L1210/D3 cells A num ber of transport routes for folates have been i m p l i c a t e d in L1210 cells, including the RFC (Goldman et al., 1968; Spinel la et al., 1995) FR (Kamen et al., 1988; Jansen et al., 1990; Antony 1996) a n d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 228 0 .6 -i i G , ( a o 0.2- a a 0 2.5 5 7.5 10 [5-CHO-THF], |uM Fig. 4.2 Diminished membrane transport of (6S)-5-CHO-THF in DDATHF- resistant L1210 cells. C o n cen tra tio n -d ep en d e n ce of (6S)-5-CHO-THF influx was m e a s u r e d over 1 min in L1210 (squares) and L1210/D3 cells (diamonds). For details, see “Materials and Methods” . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 229 a less well-characterized transporter that become predominant at low pH (Henderson and Strauss, 1990). It has been shown that m e m b e r s of the FR all display characteristically high affinity for folic a c id (Kamen et al., 1988; Brigle et al., 1994); thus, the specific increase in transport of this substrate in the variant cells raised the p o ssib ility that an upregulation of one of the FR isoforms was resp o n sib le. However, several lines of evidence suggested that this was not th e case. The mRNA expression patterns of the two murine FRs (m FBPl (homologue of FR -a) and mFBP2 (homologue of FR-P)) were c o m p a r e d between parental and DDATHF-resistant cells. In these e x p e r im e n ts RNAs isolated from two mouse erythroleukem ic lines, MEL-Fa a n d MEL-La, which have been shown to overexpress m FBl and mFBP2, respectively, were also included as positive controls (Brigle et al., 1994). Northern analysis using cDNA probes specific for either mFBPl or mFBP2 revealed similar expression level between wild-type a n d resistant lines (Fig. 4.3). C om pared with that in the m o u s e erythroleukemic cell lines, the steady state levels of both messages i n L I 210 and L1210/D3 cells were very low, requiring a 6-day e x p o su re for the m FB Pl message to be visible on the radiographs (Fig. 4.3). It has been shown that FR-mediated uptake of folates i n mammalian cells can be inhibited by ionophores, such as monensin. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 230 Fig 4.3 Northern analysis of transcript levels for folate binding protein isotypes 1 and 2. Fifteen ng of total RNA from wild-type L1210 or L1210/D3 cells or 3 pg of RNA from two lines of mouse erythroleukem ia cells overexpressing FBP-1 and FBP-2 (MEL-Fa and MEL-La, respectively) were loaded onto a denaturing 1% agarose containing glyoxal. Blots from these gels were hybridized with either a 500-bp SacI fragment of the FBP-1 cDNA or a 700-bp B am H l fragment of the FBP-2 cDNA. Equal RNA loading was confirmed by reprobing of the stripped blots with the g ly c e ra ld e h y d e - 3-phosphate dehydrogenase cDNA. The upper band in the FBP-1 blo t was at the position of the 28S ribosomal RNA. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 231 rRNAi 1700 nt fO Q o o tH tH \ • ! < V - ro © § iH © ^ 1 -H h J i J U J S W i m $V iil $ S :-9 FB P -1 # 4 ' F B P -2 2450 n t ^ GADPH RFC-1 1800 nt 1500 nt Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 232 It has been proposed that such inhibition was due to dissipation of the proton gradient across endosomal c o m p artm e n t which is essential for the release of ligand from internalized re c e p to r-lig a n d complex (Kamen et al., 1988). However, pre-incubation of cells w ith 25 pM m onensin had no effect on the increased transport of folic acid in L1210/D3 cells; thus, arguing against the involvem ent of a n endocytotic process characteristic of that of the FR (Fig. 4.4). In contrast, the influx of folic acid in L1210/D3 cells was n e a rly abolished by 20 pM of MTX, indicating a higher affinity for th is substrate over folic acid, a feature reminiscent of the "classical" RFC rather than the FR. Thus, results of these biochemical and genetic experiments strongly suggested that the aberrant increase in folic acid transport in the resistant cells was not m ediated by the FR ro u te. Amplification of RFC-1 cDNA from parental and DDATHF- resistant L1210 cells by RT-PCR The RFC is the predom inant folate transport system found i n L1210 cells. To determine whether the changes in s u b s tra te specificity of transport in the resistant cells were due to mutation i n the RFC, the cDNAs encoding this protein were isolated from b o th wild-type and resistant cells, and their sequences were compared. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 233 C/3 13 o VO o 0 a 01 0.81 16- 3.4- 3.2- B L1210 0 0 L1210/D3 control MTX Monesin Fig. 4.4 Effects of monensin and MTX on folic acid transport in L1210 and L1210/D3 cells. Cells were pretreated with 10 pM trim etrexate for 10 min prior to transport studies. Transport of [^HJfolic acid at 5 pM was fo llo w ed over a 4-min interval at 37 °C either in the presence of 20 pM MTX o r after incubation with 25 p M monensin for 30 min at 37 °C. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 234 The coding region of the RFC cDNA was amplified from both cell lines by RT-PCR using primers based on the published mouse RFC-1 sequence (Dixon et al. 1994), (Fig. 4.5). The resulting 1.8 kb PCR product spanned nt 27 to 1809 (num bering is based on t h e published mouse R F C -1 sequence)3, consisted of a 15-nt 5' UTR w h ich contained the consensus sequence for ribosome recognition, t h e entire coding region and portion of the 3' UTR (Fig. 4.6A). The yield of the 1.8 kb PCR fragment was found to be exquisitely sensitive to the concentration of M g2+ ion; optimal result obtained at 1.5 m M (Fig. 4.6A). Identification of mutations in RFC-1 cDNA derived from DDATHF-resistant L1210 sublines For each cell line, products derived from three i n d e p e n d e n t PCR experiments were pooled. The pooled PCR products were p u r ifie d by gel electrophoresis and were ligated into pCRII vector for subsequent sequencing analysis. The authenticity of the PCR products was confirmed by diagnostic restriction analysis using S m a I which resulted in the expected 700 and 1100 bp fragments (Fig. 4.6B and 4.7). Since it was conceivable that more than one allele of RFC-1 3 Please note that in Tse et al. (1998), nucleotide 1 is the A of the translation start codon of the published mouse RFC-1 (Dixon et al., 1994). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 235 poly(A+) mRNA ' AAAAAAAA reverse transcription purification of cDNA on GlassMax cartridge AT3 ATI PCR amplification "AT2 PCR fragment ligated into pCRII vector Sequencing analysis expression vector construction Fig. 4.5 Isolation of the RFC-1 cDNA by RT-PCR. First strand synthesis was carried out on p o ly (A )+ -selected R N A fr o m L I 210 and L1210/D3 cells using a gene specific primer (AT3) co m p lem en tary to the sequences in the 5'-UTR of the message. T h e coding region of RFC-1 was amplified by PCR using primers ATI a n d AT2. The 1800-bp PCR fragment was gel-purified and ligated into a pCRII TA-cloning vector for subsequent sequencing analysis a n d construction of expression vector. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 236 L1210/D3 L1210 L1210 L1210/D3 Mr 1 2 3 4 5 6 Mr 1 2 3 4 5 Mr 6 7 8 9 10 Fig. 4.6 (A) PCR amplification of RFC-1 and (B) Restriction analysis of PCR fragments. cDNA for RFC-1 was isolated by RT-PCR from p o ly (A +) RNA p r e p a r e d from L1210 and L1210/D3 cells as described in Fig. 4.5 and u n d e r "Materials and Methods". The products were fractionated by electrophoresis on a 1% agarose gel. The yield of the desired 1800 b p PCR product varied with different M g^+ concentrations: ImM, lanes 1 and 4; 1.5 mM, lanes 2 and 5; 2 mM, lanes 3 and 6; Mr, m o l e c u l a r weight markers. After ligation of the 1.8 kb PCR product into pCRII vector, authenticity of the insert was confirmed using the colony PCR technique as described under “Materials and M ethods” . Restriction digest of the PCR product by Sma I resulted in the expected 700- a n d 1100-bp fragments. Lanes 1, 2, 3, 4, and 5, clones 0-1, 0-3, 0-4, 0-5 and 0-7, respectively; lanes 6, 7, 8, 9, and 10, clones 3-1, 3-2, 3-3, 3-4, and 3-5, respectively; Mr, molecular markers. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 237 2 kb Fig. 4.7 Restriction map and sequencing strategy of RFC-1 cDNA. B = BamH I, S = Sma I, P = Pst I, X = Xho I. BamH I and Xho I were restriction sites engineered in the PCR primers used for a m p lif ic a tio n of the cDNAs. Positions of the point m utations found in L 1210/D 3- derived R F C -1 cDNA were indicated (*). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 238 were expressed in L1210/D3 cells and that only one of the ex pressed alleles was mutated, sequencing of multiple PCR-derived clones w as necessary to ensure detection of m utation(s). Com plete d id e o x y sequencing was performed on three of the L1210-derived and seven of the L12l0/D3-derived clones (Fig. 4.7). Results of sequencing an a ly sis were su m m arized in Table 4.3. Several m utations were f o u n d throughout the cDNA. However, a majority of these m utations were present in only one of the clones derived from one particular cell line, e.g. m utation at nt 749 was found in clone 0-5 and was a b s e n t in the remaining two clones derived from L1210. Likewise, m u t a t i o n at nt 68 was present in only clone 3-1. These m utations p r e s u m a b ly represented artifacts generated during PCR as a result of the error- prone nature of Taq polymerase. The nucleotide sequence of th e three L1210-derived PCR clones analyzed was identical to that of th e published RFC-1 sequence (Dixon et al., 1994), except for th re e presum ed PCR-related mutations at nt 747 and 1475 in clone 0-5, and nt 1111 in clone 0-4 (Table 4.3). Twelve point m utations were also identified in the L1210/D3-derived clones. Ten of which were found in only one of the seven clones and were therefore assumed to represent PCR artifacts. However, two sim ultaneously present p o i n t mutations, at nt 183 (A —> T) and nt 354 (T —> G), were present in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 239 Table 4.3 Summary o f Sequencing Data N u c le o tid e P o sitio n ! Clone # Base C h a n g e Amino Acid C h a n g e C o m m e n t s 68 3-1 A — > G Glu no change 144 3-9 T — > C Phe — > Leu 183 3-1,3-3,3-4,3-7 3-9,3-10,3-13 A — > T He — > Phe confirmed i n b o t h d ir e c tio n 209 3-4 A — > G Lys no change 286 3-4 T — > C Phe — > Ser 35 4 3-1,3-3,3-4,3-7, 3-9, 3-10,3-13 T — > G Trp — > Gly confirmed in b o t h d ir e c tio n 382 3-4 T — > C Val — > Ala 4 0 4 3-13 A — > T Glu — > Asp 7 4 9 0-5 C — > A Thr no change 1015 3-3 C - > T Arg — > Cys 1027 3-4 T — > C Leu — > Pro 1111 0-4 T — > C Val — > Ala 1475 0-5 A — > G Leu no change 1530 3-10 G — > C Asp — > His 1681 3-4 A — > G 3' UTR The entire coding region of ten RFC-1 cDNA clones were sequenced by the dideoxy chain termination method. Clones 0-3, 0-4, and 0-5 were derived from wild-type L I 210 cells; Clones 3-1, 3-3, 3-4, 3-7, 3-9, 3-10 and 3-13 were derived from the L1210/D3 cells. UTR, untranslated region. 1 Nucleotide number system is based on the published mouse RFC-1 sequence (Dixon et al., 1994). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 240 Fig. 4.8 Identification of double point mutations in RFC-1 cDNAs in L1210/D3 cells. Two point m utations were identified in all (7/7) RFC-1 cDNA clones derived from L1210/D3 cells. Lanes were loaded in the order GATC. The sequence in panel B represents a sequencing run using a n antisense primer (AT10) viewed with the film flipped backward, t h e sequence should be read from the top to bottom to directly read i n the sense direction. (A) Mutation at 183 resulted in amino a c id change from an isoleucine (ATC) to a phenylalanine (TTC) at c o d o n 48. (B) Nucleotide change at 354 led to substitution of a glycine (GGG) for a tryptophan (TGG) at codon 105. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A G A T C 241 I -nt 354 m m -nt 358 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 242 all seven clones derived from the L1210/D3 cells (Fig. 4.8). T h ese mutations resulted in substitution of Phe for He, and Gly for Trp, a t amino acid 48 and 105, respectively. Both residues are found to be conserved in mouse, hamster and hum an and are located in t h e putative first and third transm em brane dom ain of the protein (Fig. 4.2). Sequencing of the opposite strands confirmed the presence of these two mutations. Transfection of wild-type and mutant RFC-1 cDNA into parental and transport-defective L1210 cells To determ ine whether the point m utations in RFC-1 were causative of drug resistance in L1210/D3 cells, the mutant cDNA was expressed in a transport-defective L1210 cell line, the MTXrA, in a n effort to reproduce the resistant phenotype. The MTXrA cell line was a MTX-resistant L I 210 variant that showed im paired translocation of MTX and reduced folates across the cell m em b ran e without c h a n g e s in substrate binding to the carrier (Schuetz et al., 1989). T h e transport defect has been determined to be due to a single p o i n t m utation in the RFC gene that resulted in a substitution of p ro lin e for alanine at amino acid residue 130 (Schuetz et al., 1989; Brigle et al., 1995). The m utant RFC-1 was subcloned into the pTK-PGK eukaryotic expression vector (Brigle et al., 1995), which harbored a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 243 neomycin resistance cassette and allowed good expression of exogenous genes in the L1210, and was transfected into M TXrA cells by electroporation. To assess the effect of each individual p o i n t mutation, expression vectors that contained only one of the tw o mutations were engineered and transfected. M TXrA cells tr a n s f e c te d with vector harboring the wild-type R F C -1 were used as controls. T h e double m utant cDNA was also expressed in wild-type L I 210 cells in order to determine the effect of mutant RFC-1 in cells that p r o d u c e d functional endogenous carrier; L1210 cells transfected with w ild -ty p e R F C -1 were used as controls for those experiments. After transfection of L1210 and M TXrA cells with th e corresponding cDNAs, cells expressing the neom ycin resistance gene were selected. The number of wells containing G418-resistant cells, as judged by an obvious growth of healthy-looking cells, ranged from 9 to 116 per five 96-well plates (Table 4.4). Such a variation in transfection efficiency was not uncom m on with electroporation, a technique that is very sensitive to slight changes in t e m p e r a t u r e , volume, etc. To facilitate the identification of stable tr a n s f e c ta n ts that expressed exogenous RFC-1, all MTXrA transfectants that were G418-resistant were subjected to a quick screen with 50 nM MTX. At this drug concentration, M T X rA cells could proliferate at ease, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 244 Table 4.4 Sum m ary o f selection o f LI 210 transfectants cDNA tran sfected~ recipien t line~ cell # o f wells con tain in g G 4 1 8 - resistan t cells'll # M TX -sensitive^ # G 418-resistan t # o f wells con tain in g transfectants selected f o r dilu tion clon in g I 4 8 F , W 1 0 5 G - MTXr A 23 1 3 / 2 3 13 r f c I 4 8 F - r f c MTXr A 2 3 9 / 2 3 7 W 1 0 5 G - r f c MTXr A 16 7 / 1 6 6 w t - r f c e x p t #1 MTXr A 9 4 / 9 20 e x p t #2 MTXr A 1 1 5 6 3 / 1 1 5 I 4 8 F , W105G- L 1 210 34 - 10 r f c w t - r f c L 1 210 48 - 10 'Recipient cells that were transport-competent (L1210) or showed no RFC function (MTXrA) were transfected with RFC-1 cDNA derived from either wild-type L 1210 cells (wt-rfc) or L1210D3 cells (I48F,W105G-r/c). or transfected with recombinant RFC-1 cDNA harboring only mutation at codon 48 (I48F-r/c) or at codon 105 (W105G-r/c). ^After electroporation, cells were allow to recover for 36 h, G418 was then added at 800 pg/ml. and cell suspension was distributed into wells of five 96-well plates. The number of wells containing G418-resistant cells was scored after 8-10 days. ^MTXrA transfectants that were G418-resistant were subjected to a screen with 50 nM MTX. Cell populations that were G418-resistant but had regained sensitivity to MTX were considered to be expressing functional exogenous RFC and were selected for dilution cloning and further studies. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 245 whereas transfectants that expressed functional exogenous RFC w o u ld become growth inhibited. As shown in Table 4.4, about half of t h e wells containing cells that were G418-resistant were MTX-sensitive, indicating that functional RFC were expressed. Clonal isolates were then derived from cells taken from these wells by a limiting d i l u ti o n m ethod and were expanded for further study. In the other half of the G418-resistant population that remained MTX-resistant, the RFC- 1 cDNA inserts were probably disrupted during integration of t h e expression vectors. This result also suggested that both w ild -ty p e and mutant RFC-1, when expressed, could restore MTX transport in the M T X rA cells. This screening method could not be applied to t h e wild-type L1210 transfectants; therefore, in those cases, G 4 1 8 -resistan t cells from several wells were randomly selected for dilution clo n in g and the expression of transfected R F C -1 was later verified by n o r t h e r n blots (see below). Fig. 4.9 depicts the sequence of steps in d e v e lo p in g the various stable transfectant cell lines. Expression of exogenous RFC in L1210 transfectants The expression of the exogenous RFC-1 in stable c lo n al transfectants was studied by northern analysis. The m essage generated from the transfected cDNA was diagnostically shorter (1 8 0 0 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 246 nt) than that produced by the endogenous gene (2400 nt) (Fig. 4.10). The levels of RFC-1 transcripts (either exogenous or endogenous) i n each transfectants cell line were quantitated by p h o s p h o i m a g e analysis of the radioactive blots and were corrected for RNA lo ad in g by normalizing to message levels of GADPH. Northern blots of to tal RNA from two sets of transfectant lines (M TXrA and L 1210 cells transfected with double mutant and wild-type RFC-1, respectively) were shown as examples (Fig. 4.10). The am ount of tra n s c rip t derived from the exogenous RFC-1 was 40 to 300 % of that from th e endogenous gene. Level of the endogenous RFC-1 message were 50 % lower in the MTXrA transfectants compared with the L1210 transfectants (Fig. 4.10 and Table 4.5), consistent with a d ec rea sed expression of endogenous RFC-1 in MTXrA cells com pared w ith parental L 1210 (Brigle et al., 1995) The levels of RFC-1 transcripts in the various cell lines are shown in Table 4.5; values are n o r m a liz e d to that of parental L 1210 cells which is set as 1. Clones w hich expressed no more than 2-fold higher levels of RFC-1 mRNA t h a n found in L1210 cells were selected for study. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 247 Fig. 4.9 Schematic of the development of transfectant cell lines. Expression vector pTK-PGK harboring either wild-type (wt-r/c), d o u b le mutant (I48F,W 105G-r/c) or single mutant (I48F-r/c\ W 105G -r/c) RFC-1 was transfected into transport-defective M TX rA (A) or wild-type L1210 (B) by electroporation. For transfectants using MTXrA as recipients, G418-resistant cultures that were sensitive to 50 nM MTX were deemed expressing the transfected gene and were selected for d ilu tio n cloning and further analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. J O z f-t-D S O lA X 'J L S P l 248 0£ fi CO c J •- >> fi « M © 53 s 'Z f i fi \ 00 lH Tf O * C I o in o P N 00 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 249 2 4 5 0 n t GADPH Fig. 4.10 Northern analysis of the expression of endogenous and transfected RFC-1. Eight |ig of total RNA from the indicated cell lines were loaded onto a denaturing 1.4 % agarose gel containing glyoxal. Northern blot was hybridized with a 700-bp Smal/EcoRl cDNA fragment of the m o u s e RFC-1 gene and a glyceraldehyde-3-phosphate d e h y d r o g e n a s e (GAPDH) cDNA probe. The transcripts generated from the tr a n s fe c te d R F C -1 (1800nt) were characteristically shorter than that produced by the endogenous gene (2450nt). For the nom enclature of th e transfectant cell lines, see Table 4.5, legend. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 250 Expression Table levels o f RFC-1 t r a n s f e e ta 4.5 transcript in n ts L1210 Transfected cDNA Recipient Cell Cell Line RFC-1 mRNA level ^ - - L 1 2 1 0 1 - - L 1 2 1 0 / D 0 . 5 1 - - L 1 2 1 0 / D 3 0 . 6 - - MTXr A 0 . 5 w t - r f c MTXr A w t - A 01 2 . 1 w t - r f c MTXr A w t - A 07 1 . 5 w t - r f c MTXr A w t - A 14 0 . 5 w t - r f c MTXr A w t - A 16 0 . 3 I 4 8 F , W 1 0 5 G - r f c MTXr A dm-A 01 1 . 3 I 4 8 F , W 1 0 5 G - r f c MTXr A dm-A 06 0 . 2 I 4 8 F , W 1 0 5 G - r f c MTXr A dm-A 17 0 . 9 l 4 8 F - r f c MTXr A I 4 8 F - A 02 1 . 4 I 4 8 F - r f c MTXr A I 4 8 F - A 12 0 . 1 W 1 0 5 G - r f c MTXr A W105G-A 12 2 W 1 0 5 G - r f c MTXr A W105G-A 15 1 . 5 w t - r f c L 1 2 1 0 w t - w t 01 0 . 6 w t - r f c L 1 2 1 0 w t - w t 03 0 . 8 I 4 8 F , W 1 0 5 G - r f c L 1 2 1 0 d m -w t 03 0 . 7 I 4 8 F , W 1 0 5 G - r f c L 1 2 1 0 d m -w t 06 0 . 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25 1 Table 4.5 ( c o n t ’d ) t Transcript levels were quantitated by northern analysis on a phosphoimager. For L1210, L1210/D0.5, L1210/D3, and MTXrA cells, data represent levels of the endogenous RFC-1 message. For all L1210 transfectants, levels of the exogenous RFC- 1 were measured. All values were normalized to that of parental L1210 cells which is set as 1. The transfectant cell lines were named for the cDNA transfected (also see Table 4.4, legend) (wild-type (wt), double mutant (dm), single mutant (I48F or W105G)), the cell line transfected (wild-type L1210 (wt) or MTXrA (A)), and a numerical code for the different clones isolated. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 252 Sensitivity of L1210 transfectants to (6R)-DDATHF To address the role of m utant RFC-1 in drug resistance, t h e sensitivities of M T X rA transfectants to (6R)-DDATHF were studied. As one would predict, the parental transport-defective MTXrA cells exhibited cross-resistant to (6R)-DDATHF since M TX and (6R)-DDATHF utilized the same carrier system for m em b ran e transport (Fig. 4.11 and Table 4.6). Transfection of wild-type RFC-1 cDNA into MTXrA cells restored sensitivity to (6R)-DDATHF in all four clonal isolates, suggesting the presence of functional carriers in the transfected cells. Interestingly, the different expression levels of RFC-1 in th e s e transfectant lines did not significantly affect their drug sensitivities (Table 4.5 and Table 4.6). Transfection of double mutant RFC-7 n o t only did not restore the (6R)-DDATHF sensitivity of wild-type L1210 cells, but resulted in cell lines that were 4-fold more resistant t h a n the M T X rA line (Fig. 4.11 and Table 4.6). The extent of resistance, in fact, approximated that observed in the L1210/D3 cells (Fig. 4.11 a n d Table 4.6). M TX rA cells transfected with RFC-1 containing p o i n t mutation at nt 183 (I48F-r/c) or nt 354 (W 105G -r/c) were m o r e resistant than cells transfected with wt RFC-1 by 50- and 10-fold, respectively (Table 4.6), suggesting that each m utation c o n t r i b u t e d to the overall resistance. Taken to g eth er, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 253 * o U 0 JU 2 1 0 /D 3 o 0 .6 - U o | 0.4- ~o C 4 w t-rfc — >MTXrA 0.001 0.1 0.01 1 10 100 (6R)-DDATHF, Fig. 4.11 Sensitivity of L1210 transfectant cell lines to growth inhibition by (6R)-DDATHF. MTXrA cells, which lack reduced folate carrier function, were transfected with RFC-1 cDNA isolated either from wild-type L1210 (wt-rfc: filled squares) or from L1210/D3 cells that contained p o i n t mutations at codons 48 and 105 ( d m -rfc\ filled triangles). T h e sensitivity of representative cloned transfectant lines to grow th inhibition by (6R)-DDATHF was com pared with that of w ild-type L I 210 (open squares), M TX rA (open circles) and L1210/D3 cells (o p e n triangles) during a 72-h growth period in the presence of th e indicated drug concentrations. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 254 Table 4.6 Growth Inhibition o f L I2I0 Cell Transfectants by (6R)- D D A TH F R ecip ien t T ra n s fe c te d Cell Line IC 5 0 f°r Cell cDNA (6R)-DDATHF nM (n) L1210 23 ± 3 (6) - L 1 2 1 0 /D 0 .5 3100 ± 300 (2) - L 12 1 0 /D 3 9000 ± 1000 (4) _ L1210/A 1700 ± 520 (4) MTXrA w t-rfc wt-A 01 60 ± 2 (3) MTXrA w t-rfc wt-A 07 47 ± 4 (3) MTXrA wt -rfc wt-A 14 25 ± 4 (3) MTXrA w t-rfc wt-A 16 38 ± 4 (3) MTXrA I48F,W105G- rfc dm-A 01 6200 ± 500 (3) MTXrA I48F,W105G- rfc dm-A 06 7800 ± 2300 (3) MTXrA I48F,W105G- rfc dm-A 17 6330 ± 1890 (3) MTXrA I48F-r/c I48F-A 02 1950 ± 500 (3) MTXrA I48F-//C I48F-A-12 2000 (1) MTXrA W 105G -r/c W105G-A 12 470 ± 120 (3) MTXrA W 105G -r/c W105G-A 15 450 (1) Multiple clonal transfectant cell lines derived from wide-type L 1210 and transport-defective MTXrA cells were tested for sensitivity to growth inhibition by (6R)-DDATHF, as described in the text. T h e numbers in parentheses indicate the num ber of experiments (n ) averaged for each estimate. For the nom enclature of the tr a n s f e c te d cell lines see Table 4.4 and 4.5, legends. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 255 transfection of double mutant, but not wild-type, RFC-1 into MTXrA cells conferred resistance to (6R)-DDATHF in a manner w h ic h recapitulated the phenotype of L1210/D3 cells, proving that t h e identified m utations were responsible for the resistance p h e n o t y p e observed in L1210/D3 cells. Transport of folate compounds in L1210 transfectants To verify that the point m utations in RFC-1 was causative of the altered substrate specificity of folate transport in L1210/D3 cells, influx of (6R)-DDATHF and folic acid were studied in the MTXrA transfectants. The influx rates of (6R)-DDATHF and folic acid were m easured at lp M and 5pM, respectively. These concentrations were below the Km values for the two substrates and were selected so t h a t any difference in transport that was due to a K m effect could be identified. The ratio of the influx rates provided an estimate of t h e substrate preference of transport in that particular cell line. A nalysis based on ratios of transport rates also allowed correction for t h e different levels of expression of the exogenous cDNAs in the v a r io u s transfectant cell lines. (6R)-DDATHF was a much preferred s u b s tr a te in wild-type L1210 cells (ratio = 44). In L1210/D3 cells, the influx of the antifolate was moderately decreased whereas that of folic a c id Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 256 was markedly increased, reducing the ratio of transport rates to 1.7. In parental M TX rA cells, the transport of (6R)-DDATHF was greatly impaired as a result of a mutated endogenous RFC-1 gene (Brigle e t al., 1995). Expression of the wild-type RFC in M TXrA cells c o m p le te ly restored (6R)-DDATHF transport. Notably, in wt-A 01 and wt-A 0 7 cells, the initial uptake rate of (6R)-DDATHF were =3-fold h ig h e r com pared with wild-type L1210, presum ably because of the h ig h expression of the exogenous cDNAs in these transfectants (Table 4 .7 ). Similarly, transport of folic acid was also significantly higher in th e s e two lines, suggesting that, despite being a poor substrate, folic a c id could gain cellular entry via the RFC. However, expression of t h e w ild -ty p e RFC-1 in all four clonal isolates resulted in ratios of influx rates that were not appreciably different from that observed i n L1210, indicating the substrate preference of transport c h a ra c te ris tic of that of wild-type L1210 was preserved. Transfection of the d o u b le m utant RFC-1 into MTXrA resulted in transfectants that e x h ib ite d substrate specificity of transport almost identical to that of L1210/D3 cells (ratio = 0.82 and 1.2 for dm-A 01 and dm-A 06, respectively) (Table 4.7). Thus, com parison between tr a n s f e c ta n ts that showed similar expression of the exogenous cDNA Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 257 Table 4.7 Folate Transport Characteristics in L1210 Transfectants Cell Line T ra n s p o rt rate (pm ol/1 0 6 cells - m i n ) T ra n s p o rt Preference (DDATHF/Folic DDATHF Folic acid Acid) LI 210 0.35 0.008 43 LI 210/ 3D 0.19 0.11 1.8 M T tfA 0.011 0.005 2.2 wt-A 01 0.82 0.026 31 wt-A 07 0.87 0.033 26 wt-A 14 0.36 0.007 51 wt-A 16 0.20 0.006 33 m ean = 36 dm -A 01 0.23 0.28 0.82 d m -A 06 0.146 0.128 1.1 m ean = 0.97 I48F-A 02 0.34 0.12 2.8 W105G-A 12 0.81 0.12 6.6 The rate of transport of (6R)-DDATHF (at 1 pM ) or folic acid (at 5 pM) in the indicated cell lines was determ ined as described in th e text. The ratio of transport rate of DDATHF to that of folic acid is listed as the Transport Preference of each cell line. Each value represents the average of two determ inations per experiment and a t least two experiments. For the nom enclature of the cell lines, see Table 4.4 and 4.5, legends. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 258 (e.g. wt-A 07 and dm-A 01; wt-A 03 and dm-A 06) revealed th a t cells transfected with mutant RFC-1 showed a 2 to 5-fold decrease in (6R)-DDATHF influx compared to cells tra n sfe c te d with wild-type cDNA (Table 4.7). Conversely, the transport of folic acid was 7 to 20-fold higher in transfectants expressing the mutant RFC-1 (Table 4.7). We therefore conclude that transfection of the double m utant RFC-1 into M TXr A cells also transferred the altered folate transport characteristic found in the DDATHF-resistant L1210/D3 cells, proving that the RFC-1 mutations were causative of the resistance phenotype. The effect of each mutation on transport substrate specificity was studied in transfectant lines I48F-A 02 and W105G-A, w h ich e x p r e s s e d RFC-1 containing only a single mutation at either codon 48 or 105, respectively. It appeared that both m utations resulted in a n increase in folic acid transport. Thus, the influx of folic acid in M T X rA cells transfected with RFC-1 containing either m utation was = 3-fold higher than that in cells that expressed similar level of th e wild-type counterpart (com pared I48F-A 02 with wt-A 07, a n d W105G-A 012 with wt-A 01) (Table 4.7). Notably, the Ile^S -> Phe^S m utation also caused a substantial reduction (= 3-fold) in (6R)- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 59 DDATHF influx in addition to its effect in folic acid t r a n s p o r t (com pared I48F-A 02 with wt-A 07); however, the T r p l 0 5 -> G l y l 0 5 m utation did not alter (6R)-DDATHF transport sig n ifican tly (com pared W105G-A 012 with wt-A 01). (Table 4.7). Therefore, t h e Phe48 for Ile48 substitution appeared to have a greater impact o n the alteration in transport preference than the G lylO ^ for T r p l O ^ substitution (ratios of influx rates were 2.8 and 6.8, respectively); paralleling the higher level of drug resistance conferred by the f o r m e r mutation (Table 4.7). Expression of mutant RFC-1 in wild-type L1210 cells O ur analysis thus far have attested that the point m u t a t i o n s in RFC-1 resulted in two changes in folate transport: a m a r k e d increase in folic acid influx and a m oderate decrease in DDATHF influx. It appeared that it was the former effect that was re sp o n sib le for the lack of accum ulation of DDATHF polyglutam ates in t h e L1210/D3 cells (Fig 3.10). Thus, we predict that expression of m utant RFC-7 would confer resistance to DDATHF even in cells t h a t express functional endogenous carrier, i.e., the effect of t h e m utations would be dominant. This was indeed the case: transfection of double mutant RFC-1 cDNA into parental L1210 cells Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 260 resulted in substantial level of drug resistance (IC50 = 1030 ± 580 n M and 660 ± 190 nM for dm-wt 03 and dm-wt 06, respectively) (T ab le 4.8). Conversely, transfection of wild-type R F C -1 into L1210 cells h a d no effect on DDA THF sensitivity (IC50 = 49 ± 4 nM and 25 ± 2 nM for wt-wt 01 and wt-wt 03, respectively) (Table 4.8). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 261 Table 4.8 Transfer o f the Resistant Phenotype into Wild Type L1210 Cells by Transfection R e c i p i e n t C e l l T r a n s f e c t e d cDNA C e l l L in e IC50 f o r nM DDATHF (n) _ _ L1210 23 ± 3 (6) - L 1210/D 0.5 3100 ± 300 (2) - L1210/D3 9000 ± 1000 (4) L1210 w t - r f c w t-L l2 1 0 -0 1 49 ± 4 (3) L1210 w t - r f c w t-L l2 1 0 -0 3 25 ± 2 (3) L1210 I48F,W 105G -rfc dm-Ll210-03 1030 ± 580 (3) L1210 I48F,W 105G -rfc dm -Ll210-06 660 ± 190 (3) L1210 cells were transfected with either the RFC-1 clo n e d from L1210 cells (w t-rfc) or that from L1210/D3 cells (I48F,W105G-r/c), m u l ti p l e independent clonal cell lines were isolated from these tran sfectio n s, and the sensitivity of each cell line to growth inhibition by (6R)- DDATHF was determ ined. The numbers in parentheses indicate t h e number (n) of replicate experiments pooled for each mean v alu e. The cell lines resultant from transfection and cloning were named for the cDNA transfected (wt (wild-type), dm (double mutant)), t h e recipient cell line (L1210), and a numerical code for the d ifferen t clones isolated. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 262 Correlation between folate pools, transport specificity and growth inhibition by (6R)-DDATHF Total folate pools were determ ined in transport-defective a n d transfectant L 1210 cell lines cultured in folic acid (Table 4.9). Level of total cofactor pool in the transport-defective M TXrA line was slightly higher than that in wild-type L1210 (6.0 ± 0.9 and 4.0 ± 1.5 p m o l / l O ^ cells, respectively) (Table 4.9). This was consistent w ith the assumption that folic acid was transported primarily by a mechanism distinct from the RFC (Yang et al., 1983). Surprisingly, M TXrA ceils transfected with wild-type RFC-1 cDNA sh o w ed substantial elevation in total folate pools, suggesting that expression of exogenous R F C -1 could transport folic acid. In accord with results of the experiments on folic acid transport, cofactor level were a b o u t 3-fold higher in M TXrA cells expressing RFC-1 containing the d o u b le mutations (Table 4.9). Transfection of MTXrA cells with RFC-1 cDNA bearing the I48F and W105G resulted in a m oderate (= 2-fold) elevation of folate pools (Table 4.9). As predicted, expression of th e mutated RFC-1 had a dom inant effect on total folate pool. T hus, transfection of R F C -1 cDNA harboring mutations in codon 48 and 105 into wild-type L 1210 cells resulted in a 2-fold increase in co facto r level when compared with the same cells transfected with wild-type Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 263 Table 4.9 Correlation between transport substrate specificities, total folate pools, and sensitivities to (6R)-DDATHF growth inhibition in L1210 transfectants Cell line Transport Preference ( D D A T H F /F o li c A c id ) Total Folate Pool (pmol/lO^ cells) (n) IC 50 for (6R)- DDATHF, (nM) (n) L 1210 44 4 . 0 ± 1 . 5 23 ± 3 (8) (6) L1210/D0.5 2 . 5 16 ± 0 . 9 3100 ± 300 (2) (2) L1210/D3 1 . 8 16 ± 2 . 1 9000 ± 1000 (8) (4) MTXrA 2 . 2 6 . 0 ± 0 . 9 1700 ± 520 (3) (4) wt-A 01 32 8 . 9 ± 0 . 1 60 ± 2 (2) (3) wt-A 07 26 11 ± 0 47 ± 4 (2) (3) wt-A 14 51 10 ± 0 25 ± 4 (2) (3) wt-A 16 33 9 . 1 ± 0 .1 38 ± 3 (2) (3) dm-A 01 0 .8 2 19 ± 3 . 8 6200 ± 500 (3) (3) dm-A 06 1 .2 18 ± 3 . 0 7 8 0 0 ± 2300 (2) (3) I48F-A 02 2 . 8 11 ± 0 . 4 2000 ± 500 (2) (3) I48F-A 12 6 . 6 14 ± 3 .7 470 ± 120 (3) (3) wt-L1210 01 16 7 . 6 ± 0 . 1 49 ± 4 (2) (3) wt-L1210 03 21 7 . 5 ± 0 . 6 25 ± 2 (2) (3) dm-L1210 03 1 .3 15 ± 0 . 0 1030 ± 580 (2) (3) dm-L1210 06 1 . 7 15 ± 0 . 0 660 ± 190 (2) (3) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 264 Table 4.9 (cont’d) The rate of transport of (6R)-DDATHF (at 1 pM) or folic acid (at 5 pM) in the indicated cell lines was measured. The ratio of t r a n s p o r t rate of DDATHF to that of folic acid is listed as the T r a n s p o r t Preference of each cell line. Cells were passaged in folate-free RPM1 1640 m edium supplem ented with 2.3 pM [3H]foIic acid for 1 week prior to quantitation of total cellular folates. Sensitivity to g ro w th inhibition by (6R)-DDATHF was determ ined after a 72 hour g ro w th period as described in the text. For details, see “Materials a n d M e t h o d s ” . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 265 RFC-1 cDNA (Table 4.9). The correlation between total folate p o o ls and sensitivity to growth inhibition by (6R)-DDATHF was shown in Fig. 4.11. It appeared that the 1C50 values were independent of cofactor levels. The most reasonable explanation for the lack of correlation was that in cells that contained elevated levels of t o t a l folates as a result of over-expression of wild-type RFC-1, there was a concom itant increase in (6R)-DDATHF transport which offset t h e inhibitory effect of the expanded folate pools on (6R)-DDATHF polyglutamation. On the other hand, when I C 50 values were p l o t t e d against the ratios of influx of (6R)-DDATHF to that of folic acid, a n inverse correlation was noted. Thus, we conclude that sensitivity to (6R)-DDATHF was primarily determ ined by the substrate p re fere n ce of the transport system for the drug relative to folic acid. Genotype of the L1210 and L1210/D3 cells4 It was of particular interest that every RFC-1 cDNA clone ( 7 / 7 ) isolated from the L1210/D3 cell line contained both mutations a t codon 48 and 105. This would imply that, in the genome of th e s e cells, the wild-type R F C -1 allele was either deleted or, if present, it w as not expressed, due to gene silencing m echanisms. To clarify this, 4 Experiments on the analysis of the RFC-1 genomic loci were performed by Rachel Cain and Jennifer Fergurson in Dr Shirley Taylor’s laboratory. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 266 genomic DNA of RFC-1 was isolated from both cell lines by PCR using primers that spanned both m utations and an intron (Tolner et a l ., 1997). A single 3.4-kb PCR fragment was obtained, somewhat larger than the 2.9 kb product expected from sequences straddled by th e two primers (about 100 nt from the 3'-end of exon 2, an in te rv e n in g intron 2 of 2.6 kb (Tolner et al., 1997), and about 180 nt into exon 3). However, the identity of the amplified product was c o r r o b o ra te d by strong hybridization of the fragment with a RFC-1 specific cDNA probe on southern blots which was washed to very high s trin g en cy (0.2 x SSC at 70 °C). Direct sequencing of the PCR products d e riv e d from L1210 cells revealed that only wild-type sequences were p r e s e n t in the R F C -1 locus at codons 48 (ATC) and 105 (TGG) (Fig 4.13). For genomic sequences obtained from L1210/D3 cells, however, both wild- type and mutant R F C -1 configurations were found at codons 48 (ATC and TTC) (Fig. 4.12) and 105 (TGG and GGG), indicating that th e RFC-1 genomic locus is heterozygous in these cells. We c o n c l u d e d that the expression of the wild-type allele of the carrier had been silenced in the DDATHF-resistant cells. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Transport preference 267 100 n 10 - _ o 2 s r° a 100 1000 10000 IC50 Fig. 4.12 Sensitivity of growth inhibition by (6R)-DDATHF in L1210 sublines is correlated with transport preference of folate substrates (squares) but not with total folate pools (circles). For details, see Table 4.9, legends. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 268 Fig. 4.13 Sequencing of the RFC-1 genomic locus in L1210 and L1210/D3 cells. The DNA sequences from L1210 (left sequence) and L1210/D3 (rig h t sequence) cells were determined in the region of genomic DNA corresponding to the cDNA sequences that spanned the p o i n t m utations at nt 183 (panel A) and nt 354 (panel B), respectively (arrows). Lanes were loaded in the order GATC. The sequence i n panel B represents a sequencing run using an antisense p r i m e r viewed with the film flipped backward, the sequence should be r e a d from the top to bottom to directly read in the sense direction. Note that the genomic DNA sequences in L1210/D3 cells at codon 48 were both ATC and TTC; likewise, at codon 105 were both TGG and GGG (boxes), indicating the presence of both the wild-type and m u t a n t R F C -1 alleles in the genome of these cells. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 269 A t T \ c c c A c A ~C~ T A T T C B A G— T ,_GL T G Lg j c T G T r Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 270 Multi-step acquisition of resistance to DDATHF The L1210/D3 cell line was derived by exposing parental L1210 cells to step-wise increasing concentration of DDATHF. In view of th e nature of this selection method and the complexity of b io c h e m ic a l events involved in DDATHF cytotoxicity, the emergence of the h ig h ly resistant phenotype of the L12I0/D3 cells was likely due to m u l ti p l e genetic changes. L1210/D0.5 was a clonal L1210 subline developed a t an interm ediate stage during the selection process and was c a p a b le of uninhibited growth in the presence of 0.5 pM (6R,S)-DDATHF (Tse and Moran, 1999, and Ch. 3). The inclusion of this interm ediate line in our analysis has provided further insights into the m o l e c u l a r changes during the developm ent of resistance to DDATHF. O u r initial presum ption was that this line would have harbored e ith e r the I48F or the W105G mutation; this did not turn out to be the case. Thus, sequencing of PCR-amplified genomic DNA from L I 2 1 0 / D 0 .5 clearly d em onstrated that mutation was present in both codons 4 8 and 105, as were the wild-type sequences at these positions — find in g s that were identical to that with genomic D N A of L1210/D3 cells ( d a t a not shown). Sequencing of a limited set of reverse transcribed PCR clones derived from L1210/D0.5 cells using primers that s p a n n e d both m utated codons dem onstrated that both m utations were present in two clones and neither was found in another. Hence, th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 271 RFC-1 genomic locus in L1210/D0.5 cells was heterozygous w ith expression of both alleles and both m utations were present in t h e same allele. Consistent with these results was that northern analysis of RFC- 1 expression in L1210, L1210/D0.5 and L1210/D3 cells re v e a le d similar abundance of a 2450-nt transcript between L1210 a n d L1210/D0.5 cells, but a 50% decrease in message level in L 1 2 1 0 /D 3 cells (Fig. 4.10 and Table 4.5), indicating a selective loss of ex p ressio n of the wild-type allele in the L1210/D3 cells. The question arises of how these genetic changes in RFC-1 correlate with alterations in folate transport in the L1210/D0.5 cells. Some prelim inary transport experim ents had provided s o m e interesting results: L1210/D0.5 cells also exhibited a marked in cre ase in folic acid influx indistinguishable from that observed with t h e more resistant L1210/D3 cells (Fig. 4.14). However, the transport of (6R)-DDATHF was not affected in the L1210/D0.5 cells (Fig. 4 .1 4 ). Hence, it appeared that the critical transport-related changes in t h e L1210/D0.5 cell line pertained to that of folic acid. In the p re s e n c e of further selective pressure, there was a decline in (6R)-DDATHF transport, resulting in the more resistant phenotype of the L 1 2 1 0 /D 3 line. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 272 A 0.5-1 0 .4 - 0 .3 - o 0 .2 - 0 20 40 60 80 100 [Folic acid], p.M B 0 .3 13 0 . 2 - GL [(6 R )-D D A T H F ], n M Fig. 4.14 Transport of folic acid (panel A) was increased in L1210/D0.5 cells but that of (6R)-DDATHF (panel B) was unaffected. The rate of transport of folic acid and (6R)-DDATHF were m e a s u r e d as a function of the indicated concentrations in L1210 (squares), L1210/D0.5 (diamonds), and L1210/D3 cells (circles). The influx of folic acid was measured over a 2-min interval and cells were pretreated with 10 pM trimetrexate prior to transport studies. T h e rate of (6R)-DDATHF transport was followed over a 4-min period to assure unidirectional influx measurement. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 273 CHAPTER 4 DISCUSSION In the preceding chapter, we describe the biochemical basis of an unusual mechanism by which leukemic L1210 cells b e c o m e resistance to the folate antimetabolite, 5 , 1 0 -d id e a z a te tra h y d ro fo la te . In the present study, evidence for the specific molecular changes t h a t are responsible for the resistance phenotype was provided. A lth o u g h altered membrane transport of drug has been a well recognized c a u s e of resistance to antifolates such as MTX, the tr a n s p o r t- r e l a t e d changes identified in the L1210/D3 cell line were very unique. U n lik e the co m m o n ly reported m echanism s of tra n s p o r t- a s s o c ia te d resistance where drug uptake was severely impaired, the transport of (6R)-DDATHF in L1210/D3 cells was only moderately r e d u c e d . Rather, there was a marked increase in folic acid influx, causing a n expanded total folate pool and, subsequently, a blockade in th e formation of (6R)-DDATHF polyglutam ates. M embrane transport of other folate substrates was also found to be altered (Table 4.2). Of interest was the finding that the influx of folinic acid was reduced i n L1210/D3 cells (Table 4.2). This served to explain why these cells contained a dim inished level of cellular folates and, hence, were paradoxically more sensitive to DDATHF than wild-type L1210 w h e n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 274 folinic acid was used as the folate source in the culture medium (Tse and Moran, 1999 and Ch. 3 Table 3.1). The peculiar transport-related changes in the resistant cells was unlikely to be mediated by the FR system because the increase in folic acid transport was not affected by the ionophore, monensin, a known inhibitor of FR-dependent uptake; and the levels of expression of both mouse FBP1 and FBP2 mRNAs were similar between w ild-type and DDATHF-resistant cells. We have d em onstrated that the resistance phenotype of L1210/D3 cells was due to expression of an aberrant RFC based on th e following: 1) cloning and sequence analysis identified tw o sim ultaneous m utations in all R F C -l-c ncoding cDNAs derived fro m the L1210/D3 cell line; 2) transfection of m utant RFC-1 in to transport-defective M TX rA cells conferred resistance to (6R)-DDATHF, whereas transfection of wild-type RFC-1 cDNA did not; and, 3) expression of m utant RFC-1 in MTXrA cells reproduced the a lte re d folate metabolism observed in L1210/D3 cells, i.e. increased folic a c id transport, decreased (6R)-DDATHF transport, and elevated folate pools. Transfection of mutant RFC-1 into wild-type L 1210 cells t h a t exhibited efficient (6R)-DDATHF transport via functional e n d o g e n o u s Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 275 RFC also caused these cells to become resistant (Table 4.8). T h is finding is in accord with our impression that it was the in c re a s e d folic acid transport but not the m inim ally decreased (6R)-DDATHF transport that was responsible for the substantial im pairm ent of DDATHF polyglutam ation in L1210/D3 cells. In genetics terms, th is would indicate a dom inant effect of the m utations, i.e. the d r u g resistance phenotype was determ ined by expression of the m u t a n t but not the wild-type R F C -1 allele. This was in contrast to the loss-of- function mutations in the RFC reported by others for which the effect would be recessive (Brigle et al., 1995; Wong et al., 1995). However, it should be noted that the dom inance was incomplete in our case because the L I 210 transfectants were significantly less resistant to (6R)-DDATHF than L1210/D3 cells, suggesting that the p h e n o t y p e imparted by the double mutant R F C - 1 was tempered by co-expression of wild-type RFC. H ydropathy plot of the RFC has predicted that it comprises of 12 m em brane spanning a-helices and belongs to the 12- tran sm em b ran e helix transporter superfamily (Dixon et al., 1994). Based on their hydropathic profiles, there are now over 100 transporters that are conceived to a m em ber of this s u p e rfa m ily , although biochemical evidence supportive of the 12-helix structure is available for only a handful of them (for review see H e n d e rs o n , Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 276 1993). Protein chemistry and structural studies of the RFC have been ham pered by the low expression of the carrier on the cell surface. However, biochemical analysis of drug resistant m utants over th e past two decades had provide a wealth of variant cell lines t h a t exhibited various transport-related changes involving the RFC (for review see Sirotnak, 1987). The alterations in folate t r a n s p o r t include an increased influx K m (Sirotnak et al., 1968; Jackson et al., 1976), a decreased influx V m ax> (Sirotnak et al., 1981; Schuetz et al., 1988) and a com bined kinetic defect (Sirotnak et al., 1981; N ietham m er and Jackson, 1975). A reduction in V m a x can be explained by either a diminished number of carrier protein or a n impaired translocation process (Schuetz et al., 1988; Brigle et al., 1995). The former mechanism can be resulted from a number of genetic/epigenetic changes such as genomic deletion, d e c re a se d transcription, splicing defect, unstable mRNA, frame shift m u t a t i o n , poorly translated message, and unstable protein. Recent genetic a n d cDNA cloning of the RFC from various species would allow identification of the specific molecular alterations responsible for th e generation of these mutants and, hence, a better u n d e r s t a n d i n g about the stru ctu re-an d -fu n ctio n -relatio n sh ip as well as the genetic control of this transporter. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 277 Such combined pharmacological and genetic approaches h a v e already yielded some useful structural information about the RFC: The M TXrA line was a transport-defective L1210 variant that w as used as the recipient cell line in our transfection studies. T h e underlying genetic defect in the RFC-1 gene has been determ ined to involve the substitution of a proline for an alanine at amino a c id residue 130 in the fourth putative tran sm em b ran e dom ain of t h e protein (Brigle et al., 1995 and Talbe 4.10). The total loss of RFC function as a result of this m utation was not surprising b e c a u s e random insertion of a proline would be expected to have serio u s structural perturbation by the introduction of a kink in the p r o t e i n backbone. Interestingly, despite a 100-fold decrease in the influ x V m ax for MTX, there was no change in the influx K m , suggesting t h a t substrate binding to the carrier was preserved (Schuetz et al., 1988). The two RFC-1 mutations identified in the L1210/D3 cell line resulted in amino acid substitution at positions 48 and 105, tw o highly conserved residues located in the first and third p r e d i c t e d tran sm em b ran e domains, respectively. Of particular interest were the functional consequences of these mutations: they did not re s u lt in complete loss of RFC function but led to subtle changes i n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 278 substrate specificity. Kinetic analysis of influx rates using the v a rio u s folate com p o u n d s revealed changes primarily associated with K m rather than with V m ax values, suggesting that the mutations affected mainly the affinities of substrate binding instead of the tra n s lo c a tio n process (Table 4.2). Although it might not be possible to predict t h e biophysical consequence of each mutation in the absence of a th ree - dim ensional organization or other structural information about t h e carrier, it is reasonable to hypothesize that the m utated amino ac id s represent residues comprising the substrate recognition site of th e transporter. However, the possibility of a remote effect leading to repositioning of the substrate binding site can not be c o m p le te ly excluded. Since both mutations resulted in substitution of n e u t r a l side chain that did not involve any gain or loss in electrical charges, if the m utated residues are indeed part of the substrate b in d in g pocket, the effect produced by the replacem ent will probably be steric or hydrophobic in nature. Transport studies of tra n s f e c ta n ts that expressed RFC-1 harboring single m utation have p r o v i d e d additional inform ation about the functional significance of e a c h mutation. It appeared that, whereas both m utations contribute to accelerated folic acid influx, the Ile48 -> Phe48 change also re s u lte d in substantial reduction in (6R)-DDATHF influx. Detail k in etic Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 279 analysis w ould undoubtedly help to delineate the molecular basis of these transport changes and, hence, the structure and function of t h e RFC. The mutated amino acids identified in the present report, a n d their adjacent residues, nevertheless, serve as good starting points for future m utagenesis studies. In a recent report, the pattern of murine RFC mutations w as studied in a panel of MTX transport-defective L1210 mutants d e r iv e d from pre-treatment with chemical mutagens followed by a single-step selection with continuous exposure to MTX in the presence of folic acid as the sole folate source (Zhao et al., 1999). Of the 34 t r a n s p o r t- defective sublines analyzed, 12 distinct point mutations were identified that resulted in amino acid substitution. In the r e m a i n in g transport-defective cell lines where m utation data were av ailab le, there were nucleotide changes causing expression of either no c a rrie r or truncated proteins. Of the 12 point m utations identified, all b u t one were found in or adjacent to predicted transm em brane d o m a i n s , with the highest frequencies in the first, fifth, and eighth. There were no m utations in the sixth, ninth, and twelfth t r a n s m e m b r a n e domains (Zhao et al., 1999). Together with a Ser46 -> Asn46 RFC-1 m utation identified earlier by the same group (Zhao et al., 1 9 8 8 a) and the Ile48 -> Phe48 reported here, there were a total of fo u r separate studies reporting 3 m utations clustered in the first Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 280 transm em brane domain, suggesting an im portant role of this region in the interaction with folate substrates (Fig. 4.1 and Table 4.10). It is interesting to note that amino acid substitutions that result in changes in substrate specificity are found within a t r a n s m e m b r a n e domain rather that at an extracellularly located position. However, a potential caveat is that, the mutagen N -m eth y l-N -n itro so u rea employed in these studies, which selectively caused G to A (70%) a n d C to T (30%) transitions (Lee et al, 1992), might bias toward th e production of a certain amino acid substitutions. Of note, a Glu45 - > Lys45 mutation resulted in an increased affinity for 5-CHO-THF a n d folic acid (3.6-fold decrease in influx Kt and 7-fold decrease in Kj based on inhibition of 5-CHO-THF, respectively); but a d e c re a s e d affinity for MTX (7-fold increase in influx Kj). These changes, alo n g with a m arked decrease in carrier mobility, resulted in MTX resistance (Fig. 4.1 and Table 4.10) (Zhao 1988b). Roy et al. (1998) have recently reported a single amino acid differrence between tw o murine cell lines at codon 297, which encodes either Ser or Asn in L1210 or S180 cells, respectively. This substitution located betw een the 7 lh and 8 'h transm em brance helices appears to account for th e higher influx K m for MTX and, thus, intrinsic resistance to th is antifolate in S180 cells (Fig. 4.1 and Table 4.10). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28 1 The clinical relevance of transport-related resistance to antifolates has been a subject of controversy. It has been su ggested that this mode of resistance might be an artifactual p h e n o m e n o n demonstrable only in cell culture systems. This is because tum or cell lines were generally grown in m edium containing folic acid as t h e sole source of folate owing to its superior chemical stability d e s p ite the fact that the circulating form of folate in the plasma is (6S)-5- methyltetrahydrofolate. Folic acid is a poor substrate for the R F C .it gain cellular entry via RFC-independent routes, either by p assive diffusion or a poorly characterized transport system (Yang et al., 1983; Sirotnak et al., 1987). Thus, in the presence of folic acid, resistant cells can sustain a large defect in RFC, w i t h o u t com prom ising their requirement for exogenous folates. However, transport-related resistance to antifolate has been dem onstrated in leukemia-bearing mice dosed with MTX (Sirtonak et al., 1967), a n d more recently, in patients with hematological malignancies t r e a t e d with this agent (Trippett et al., 1993). The molecular basis of th e transport alterations in these resistant cells established in v iv o h a s yet to be determined. Based on our results, it is possible that subtle changes in substrate specificities can decrease the transport of MTX, without compromising that of the natural folates, or that there is a n over-production of a different folate transporter such as the FR. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 282 Alternatively, tumor cells may be able to maintain s u b s ta n tia l surplus in capacity for accum ulating cellular folates to support cell division. To develop a more clinically relevant in vitro model, Z h a o et al. (1998a) studied a L I 210 subline selected for resistance to MTX and that had been m aintained in m edium containing 5-CHO-THF as the sole folate source. (Although 5-CH 3 -THF is the most ph y sio lo g ical folate, its poor chemical stability renders it an impractical source in cell culture. Moreover, in the absence of transcobalam in II, t r a n s p o r t of B ): into cultured cells is also reduced, thereby limiting entry of 5- C H 3-THF into cellular folate pools). This m utant line showed a 10- fold resistance to MTX by virtue of a 44-fold decrease in MTX influx V m a x without appreciable changes in influx Kt- However, for 5-CHO- THF and 5 -C H 3-THF, influx V m ax decreased by only 8- and 7-fold, respectively; and influx K m were also unchanged. Thus, th e re seemed to be a substrate-specific decline in carrier translocation t h a t disfavored M TX in particular. The molecular basis for the changes i n carrier mobility was a single point m u tation in codon 46 t h a t resulted in a serine to asparagine substitution within the first predicted tran sm em b ran e segment, again, highlighting t h e importance of this region in carrier function (Fig. 4.1). T h ese findings also dem onstrated that point m utation in the RFC c o u ld , Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 283 indeed, impart very specific changes in substrate preference of th is transport system . Another im portant question that arises is whether tumor cells expressing high level of RFC are more sensitive to antifolates. O u r analysis suggested that this might not be the case, or at least, it d o e s not seem to apply to agents that depend heavily on p o l y g l u t a m a t i o n for cytotoxicity such as DDATHF. Thus, transfectants that ex p ressed different levels of RFC-1 were not different from wild-type L1210 cells with respect to (6R)-DDATHF sensitivity (Table 4.6 and Table 4.8). A probable explanation was that, whereas cells expressed high level of RFC showed an increased transport of (6R)-DDATHF, they also to o k up more folic acid and, hence, m ain tain ed a higher e n d o g e n o u s folate pool which intervened with the accum ulation of (6R)-DDATHF p o l y g l u ta m a t e s . It was of considerable interest that only the mutant RFC-1 allele was expressed in the L120/D3 cells. Analysis of genomic DNA from these cells indicated that the wild-type allele was presen t; hence, the wild-type allele was selectively silenced. Consistent w ith this was that northern analysis revealed a 50% decrease i n abundance of the RFC-1 message. Differential expression b etw een wild-type and m utant RFC-1 alleles has been described in a c q u ir e d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 284 MTX resistance due to a mutated carrier. Brigle et al. (1995) h av e reported that in the M TXrA line, whereas both wild-type and m u t a n t RFC alleles were present in the genome, expression was restricted to the one containing the G 4 2 9 c 4 29 m utation. The basis of th is selective silencing of the wild-type allele has yet to be e lu c id a te d . Possible mechanisms include mutations in the prom oter of the RFC-1 gene and epigenetic mechanisms such as DNA h y p e r m e t h y l a t i o n (Nyce 1989; Nyce et al., 1993; Hsueh and Dolnick, 1994). It would be of interest to determine whether exposing the L1210/D3 cells to h y p o m eth y latio n agent such as 5'- azacytidine could reactivate th e expression of the parental RFC-1 allele in these resistant cells. If epigenetic mechanisms are indeed responsible for these gene silencing events, the frequency of acquired drug resistance will therefore be determ ined by the spontaneous mutation rate at a single allele, followed by the as yet unmeasured rate of gene silencing of the wild- type allele. The inclusion of the intermediate resistant line L1210/D0.5 in our analysis has proven to be invaluable to understanding the m u lti- step acquisition of resistance to DDATHF. Contrary to our initial presum ption, this line did not contain only one of the two RFC-1 mutations; instead, both mutations were found and resided in th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 285 same allele. Further, the wild-type allele of the gene was also expressed. Transport studies revealed that the L1210/D0.5 cells also exhibited a dramatic increase in folic acid transport as in th e L1210/D3 cells. On the other hand, (6R)-DDATHF transport was unaffected in the L1210/D0.5 cells. Taken together, we h y p o th e s iz e that, during the multi-step selection process, the RFC-1 gene h a d accum ulated, presumably in a sequentially manner, two p o i n t m utations in one of the alleles, giving rise to the 135-fold re s is ta n t L1210/D0.5 cell line. It appeared that the critical t r a n s p o r t- r e l a t e d change in the L1210/D0.5 cell line pertained to that of folic acid. In the presence of further selective pressure, expression of the w ild -ty p e allele was lost, resulting in a decline in (6R)-DDATHF transport a n d the more resistant phenotype of the L1210/D3 line. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 286 Table 4.10 Phenotypes o f R F C -m u tatants Amino a c id s u b s titu tio n Cell line T r a n s p o r t - r e la te d P h e n o ty p e Reference 45Glu 45Lys CEM/MTX-LF Wild-type K m value for MTX; 31- and 9-fold decrease in influx Km for folic acid a n d leu co v o rin , re sp ectiv ely Jansen e t al, 1998 45Glu 45Lys MTXrA-E45K¥ 3.6-fold increase in influx K m for 5-CHO- THF; 7- fold decrease i n influx Kj for folic a c id based on inhibition of for 5-CHO-THF influx; 7-fold increase in influx Kj for MTX Zhao e t al, 1998b 46Ser 46Asn L 1 2 1 0 -G la + D i s p r o p o r t i o n a t e decrease in influx Vm ax for MTX relative to reduced folates; n o changes in influx K m Zhao e t al, 1998a 48Ile 48Phe 105Trp-> 10 5 G ly L 1210/D 3 £ 17-fold decrease in influx K m for folic acid; 3.4- and 1.8-fold increase in K m v alu es for (6R)-DDATHF a n d 5-CHO-THF, respectively; 1.7-fold decrease in for MTX; n o appreciable changes in influx Vmax Tse et al., 1998 130Ala->130Pro MTXrA Loss of ca rrie r tr a n s l o c a t io n Brigle e t al., 1995 A sn 2 9 7 s SI 805 4-fold higher influx Km for MTX in S I 80 cells Roy e t al., 1998 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 287 Table 4.10 (cont’d) 1CEM/MTX-LF (low folate) was derived by subjecting a prev io u sly derived MTX-resistant cell line, CEM/MTX, to gradual deprivation of folic acid from 2.3 /tM to 2 nM in the culture medium , resulted in a >20-fold overexpression of a structurally altered RFC. +L1210 grown in the presence of 5-CHO-THF as the folate source, selected for M TX resistance by chemical mutagenesis. Y L1210-derived, resistant to MTX. £See also Table 4.1 *Codon 297 encodes either Ser or Asn in L 1210 or S180, respectively. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 288 CHAPTER 5 GENERAL DISCUSSION In 1948, the folate antagonist, aminopterin, was successfully introduced into clinical use by Dr. Sidney Farber for the treatment of childhood leukemia (Farber et al., 1948). This represents the first example of antim etabolite-based chem otherapy utilized a g a in s t human cancers. Numerous folate analogs have been synthesized a n d studied since then. Of all the antifolates, MTX is the agent that h a s been subjected to the most extensively clinical and b io c h e m ic a l investigation. Using MTX as a molecular probe, a great deal h a v e been learned about the fundamentals of various im portant biological processes such as gene amplification, nucleotide metabolism, cell cycle progression, and cell death, just to name a few. To d a te , antifolates, as a class, remain the paradigm for design a n d development of new anti-cancer agents. Over the past two decades, considerable interest has b een focused on the development of antifolates targeted at enzymes o t h e r DHFR. CB3717 and ZD-1694 (raltirexed, Tomudex) are specific thym idylate syntase inhibitors that have been introduced i n to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 289 clinical trails (Calvert et al., 1986; Cunningham et al., 1995; J a c k m a n et al., 1995; Clarke et al, 1993). Folate-dependent enzymes of the d e n o v o purine synthesis pathway are attractive alternatives to DHFR as targets for antifolate chem otherapy. DDATHF is the lead c o m p o u n d of a class of selective inhibitors of GARFT, the first of the two folate- dependent enzym es of the de n o v o purine biosynthesis (Beardsley et al., 1989). Pre-clinical studies using tumor cell lines and t u m o r xenograft models have established that (6R)-DDATHF is an ac tiv e folate-based chem otherapeutic agent that merits vigorous clinical testing (Beardsley et al., 1986; Shih et al., 1988). The focus of th is dissertation has been to understand the biochemical factors t h a t determ ine tum or sensitivity or resistance to this prototypic GARFT in h ib ito r . To help defining the m olecular determ inants of t u m o r response to a cytotoxic anti-neoplastic agent, it is necessary t o discrim inate between two stages of biochemical events: (1) a p r e target phase which encompasses the m olecular steps leading to interaction of the drug with its cellular target. This in c lu d e s translocation of drug across the plasma m em brane, i n t r a c e l lu l a r m etabolism , and binding/reacting with the target m a c r o m o le c u le s . (2) a post-target phase which involves biochemical processes t h a t Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 290 causes cell cycle perturbation, leading either to cytostasis or cell death. For most anti-cancer agents, the former processes are relatively well characterized; whereas, until recently, little was k n o w n about the latter set of molecular events (see below). Pre-target biochemical events Several pre-target processes have been identified that are critical for the cytotoxic action of DDATHF. These processes re p re s e n t potential targets that can be exploited to im prove selective t u m o r cell kill by this drug. (1) Membrane Transport M em brane permeation of (6R)-DDATHF is m ediated by transport systems utilized by other classical antifolates, i.e. th e reduced folate carrier and the folate receptor (Pizzorno et al., 1993; Tse et al., 1998). In most tumor cell types, RFC is the p r e d o m i n a n t route of transport (Sirotnak and Toner, 1999). (6R)-DDATHF is a n excellent substrate for the RFC, with an influx K m of 1.7 pM and 1.1 pM in mouse L 1210 and human CCRF-CEM cells, resp ectiv ely (Pizzorno et al., 1993; Tse et al., 1998). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 291 With the recent cloning of cDNAs encoding for the RFC f r o m several species (Dixon et al., 1994; Williams and Flintoff, 1995; W ong et al., 1995), molecular probes are now available to allow exam ination of the expression of this protein in various normal a n d tumor tissues. Several important issues arise: what is the impact of RFC expression on (i) the net transmembrane transport of folates a n d (ii) on the accum ulation of intracellular folylpolyglutamates; a n d , (iii) can intratum oral expression of RFC serve as a genetic determinant of sensitivity or resistance to folate antimetabolites? The effect of carrier expression on MTX transport has b een examined using two experimental approaches. The first a p p r o a c h , undertaken by investigators prior to cDNA cloning of the RFC, involves the selection for variant cell lines that overexpress t h e carrier following gradual adaptation to low e x tra c e llu la r concentrations of 5-form yltetrahydrofoIate (Sirotnak et al., 1984; Jansen et al., 1990; Matherly et al., 1991). Transport studies in th ese sublines have dem onstrated that there was a marked e n h a n c e m e n t of MTX influx as a result of an increase in influx Vmax, with little or n o change in the efflux rate constant, thereby causing a substantial n e t increase in intracellular free drug level (Sirotnak et al., 1984; J a n s e n et al., 1990; Matherly et al., 1991). Surprisingly, a different o u t c o m e is obtained when carrier expression is exaggerated by transfection of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 292 transport-defective L1210 cells with wild-type RFC cDNA (Zhao et al., 1997). Under these conditions, there was a 9-fold increase in influx of MTX in the transfectants; however, the efflux rate constant also increased 5-fold relative to parental L1210 cells, resulting in only a 2- fold elevation in free intracellular MTX (Zhao et al., 1997). It h a s been proposed that the asymmetrical increase in bidirectional fluxes of M TX obtained in variant cell lines selected under low folate g ro w th condition may involve additional modifications in the RFC such t h a t carrier-m ediated efflux is suppressed, or alternatively, it may re s u lt from alterations in the energy-dependent exit pum p distinctive f r o m the RFC (Zhao et al., 1997; Assaraf and Goldman, 1997). Conversely, this discrepancy can be explained solely on t h e basis of the energetics and kinetics of the carrier-m ediated t r a n s p o r t process (Fig. 5.1). Studies from several laboratories have suggested that the uphill inward transport of MTX and other folate c o m p o u n d s is coupled to the outward counterflow of organic anions (o rg an ic phosphates?) down a concentration gradient (Goldman 1971; Henderson and Zevely, 1981; Henderson and Zevely, 1983; Yang et al., 1984). The change in free energy per mole of folate t r a n s p o r t e d inward, AG, is a function of the composite electrochemical g r a d ie n t of the folate as well as the putative organic anion. The kinetics of transport in each direction is inversely related to the energy b a rrie r Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 293 that needs to be overcome, i.e., for carrier-m ediated influx, it is th e activation energy, EA; whereas, for carrier-m ediated efflux, it is th e sum of change in free energy and activation energy, AG + EA. Increasing the num ber of functional carriers effectively lowers th e activation energy per mole of folate transport (EA -> EA’), th e re b y increases the rate of bidirectional fluxes across the cell m e m b r a n e . However, the relative enhancem ent of unidirectional influx versus efflux depend on the overall energetic of the transport process. T hus, in systems where energetics of inward transport is highly fa v o rab le (AG > EA), carrier-m edicated influx is m arkedly augm ented (EA/ EA’ > 1); whereas, efflux is not appreciably affected (AG + EA / AG + EA’ = 1) (Fig. 5.1A). This may be the case in variant cell lines obtained w ith low folate-selective pressure. In contrast, in systems where th e energetics of inward transport is less favorable (L1210 cells transfected with RFC?) both influx and efflux will be enhanced in a more sym m etrical fashion as a result of RFC overexpression (Fig. 5.IB). Thus, symmetrical versus asym m etrical alteration in bidirectional fluxes as a result of changes in carrier numbers seems Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 294 Fig. 5.1 Energetics and kinetics of folate transport. The change in free energy per mole of folate transported inward (AG) is a function of the composite electrochemical gradient of the folate and the counter-transported putative organic anion. The kinetics of unidirectional influx and efflux is inversely related to the energy barrier encountered in each direction, which are EA and AG + EA , respectively. Overexpression of the RFC results in little change in efflux rate in system where influx is highly favorable (AG + EA / AG + E a ’ = 1) (A) ; but causes a more symmetrical increase in b id ire c tio n a l transport in system where influx is less favorable (AG + EA / AG + EA’ > 1) (B). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 295 A out AG B AG AG AG Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 296 to depend on the composite electrochemical gradient of folate com pound and inorganic ions of the system. The impact of RFC expression on the net transport of folate substrates other than MTX has not been systematically studied. Zhao et al (1997) have dem onstrated that, in a L1210 transfectant cell line that overexpresses RFC, changes i n accum ulation of cellular MTX polyglutamates mirror the changes in intracellular free parent drug levels but not that in MTX influx. Thus, an augmented 8-fold increase in MTX influx resulted in only a 2-fold increase in M T X polyglutamates (Zhao et al., 1997). This is in accord with the concept that, at such extracellular concentration of drug (0.1 pM), influx is still much faster than the rate of in tr a c e llu la r metabolism to its im perm eable polyglutamate derivatives. W h e th e r this notion also applies to antifolates that are more superior to MTX as a substrate for FPGS remains to be elucidated. The next query is on whether intratum oral RFC expression is predictive of sensitivity to folate antim etabolites. Our analysis suggested that high level of RFC expression does not necessarily confer increased tumor sensitivity to (6R)-DDATHF. Thus, in both p a r e n t and transport-defective L I 210 cells that overexpress the carrier by virtue of transfection of wild-type RFC cDNA, sensitivities to (6R)- DDATHF were identical to that of parent L1210 (Tables 4.4, 4.6 a n d Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 297 4.7). Notably, there was a con co m itan t elevation of cellular folate pools in these transfectants, suggesting that the effect of e n h a n c e d drug transport is counterbalanced by inhibition of d ru g polyglutamation by the expanded cofactor pools (Table 4.8). (2) Retention of intracellular (6R)-DDATHF polyglutamates There has been increasing evidence to support the concept t h a t (6R)-DDATHF is a ‘pro-drug’, with its polyglutam ates derivatives as the active cytotoxic metabolites: (i) (6R)-DDATHF is a s u p e rb substrate for m am m alian FPGS. It was the first folate analog f o u n d to be as efficient a substrate for the enzyme as the most fa v o rab le physiological folate, with a V max/K m value 30 times that of MTX observation (Moran et al., 1993). (ii) The drug is r a p id ly m etabolized into long chain polyglutam ates (G lu 5.7) in whole cells (Pizzorno et al., 1991; Tse and Moran, 1998). This form of m e ta b o lic trapping is particularly im portant for cell cycle-dependent, p h a s e - specific agents like DDATHF because it permits drug exposure to be shorter than the doubling time of tumor cells and thereby avoids t h e more labor intensive continuous infusion of chem otherapy (Smith e t al., 1993). (iii) Polyglutamation of (6R)-DDATHF increases its ac tiv ity as inhibitor of GARFT. In the original report form our la b o ra to r y Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 298 using purified mouse GARFT, the p en tag iu tam ate of DDATHF was about 100-fold more potent than the parent drug (Table 2.2, B aldw in et al., 1990). Our more recent re-investigation on this subject using a more sensitive spectrophotom etric assay has indicated that th e strength of this interaction increases with the h e x a g lu t a m a te com pared with (6R)-DDATHF itself, but only by a factor of 10-fold (Sanghani and Moran, 1997). In contrast, polyglutamates of MTX are only slightly better inhibitors of DHFR (Cheng et al., 1985; Sikora et al., 1988). (iv) In several studies on the biochemical basis of a c q u ir e d resistance to (6R)-DDATHF, it has been dem onstrated that i m p a i r e d accum ulation of intracellular drug polyglutam ates is a re c u rr e n t phenomenon. Thus, in a hepatoma cell line selected for resistance to the drug, the level of y-glutamyl hydrolase, the enzyme re sp o n sib le for degradation of folylpolyglutamates, was found to be elev ated ; that was acco m p an ied by a decrease in the amount of long c h a in (6R)-DDATHF polyglutam ates (Rhee et al., 1993). Pizzorno et al. (1995) have reported that in two CCRF-CEM mutant cell lines w ith different levels of resistance to (6R)-DDATHF, there was a p arallel decrease in FPGS activity. A third subline from that study ex pressed very low levels of FPGS activity, in spite of a steady state content of mRNA for this protein equivalent to that in wild type cells, suggesting a mutation in the coding region of this enzyme that rendered cells Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 299 resistant to D DA TH F (Pizzorno et al., 1995). Indeed, mutated form s of FPGS causing dim inished polyglutam ation of (6R)-DDATHF h a v e recently been reported as the underlying m echanism responsible for antifolate resistance in four L1210 sublines (Zhao et al., 2000). In these m utant cell lines, FPGS mRNA expression was u n c h a n g e d ; whereas, FPGS activity was decreased by >98%. It was d e t e r m i n e d that both alleles of the FPGS gene harbored a point mutation i n highly conserved dom ains of the coding region (Zhao et al., 2000). Finally, a unique m echanism uncovered by the L1210/D3 m u t a n t reported herein further supports this com m on scheme: tumor esc ap e by virtue of an impaired accum ulation of (6R)-DDATHF polyglutamates; in this case, caused by a p o lyglutam ation b lo c k a d e induced by an expanded cofactor pool (Chapter 3, and Tse a n d Moran, 1998). Having realized the critical role of (6R)-DDATHF polyglutam ates, efforts have been made to identify b io c h e m ic a l factors that m odulate the intracellular accum ulation of th ese cytotoxic metabolites. The steady state level of DDATHF polyglutam ates is determ ined primarily by two processes: (i) rate of formation of polyglutam ates formation catalyzed by FPGS; and, (ii) rate of their break down mediated by y-glutamyl h y d ro la s e (conjugase), a lyzosomal enzyme that is also constitutively secreted Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 300 (Galivan et al., 2000). Synthesis of polyglutam ates is driven by concentration of free m on o g lu tam ates available for FPGS, which, in turn, is determ ined by net uptake of drug from the e x tra c e llu la r space. M am m alian FPGS exists as both a cytosolic and a m itochondrial form. Transcription initiation using the m o re upstream start sites located in exon 1 of the gene results in th e synthesis of mRNAs that encode an additional leader p e p t i d e responsible for targeting the enzyme to the m i t o c h o n d r i o n (Freemantle et al, 1995; and Freemantle and Moran, 1997). Of d ire c t relevance to antifolate metabolism is the level of FPGS present in t h e cytosolic com partm ent. Based on estimates obtained fro m ribonucleotide protection assays, only one third of the FPGS transcripts in CEM cells corresponds to mRNA encoding cytosolic enzyme (Freemantle and Moran, 1997). Further, our a n a ly sis suggests that functional FPGS activity can be m odulated by the level of cellular folate pools, thereby adding another level of complexity to the system. AUXB1 is a mutant Chinese ham ster ovary cell line th a t expresses no detectable FPGS activity and is therefore auxotrophic for glycine, adenosine, and thym idine (McBurney and Whitmore, 1974; and Taylor and Hanna, 1977). The effect of FPGS levels on folate accum ulation as well as on sensitivity to folate antim etabolites h a s Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 301 been examined in a series of transfectants expressing various levels of human FPGS obtained upon transfection of AUXB1 with h u m a n genomic DNA (Foo and Shane, 1982; and Lowe et al., 1993). O n e interesting observation is that clonal cell growth can be sustained by an extremely low level (= 2% of wild-type cells) of FPGS activity (Foo and Shane, 1982). Of note, a similarly low level of enzyme activity is found to be sufficient to support proliferation in variant cell lines selected for resistant to (6R)-DDATHF (Zhao et al. 2000). Lowe et al. (1993) have reported that, at low m edium folate c o n c e n tr a tio n s (<200 nM folic acid), cellular folate accum ulation was found to be independent of FPGS activity except in cells expressing extremely low levels of FPGS (< 2% of wild-type). With higher m edium fo late concentration (2 - 20 pM folic acid or 20 nM folinic acid), co fa c to r accum ulation became proportional to FPGS activity and the c h a i n length of intracellular folates also decreased. It appears that u n d e r low extracellular folate conditions, FPGS activity is in excess, a n d accum ulation of total folylpolyglutam ates is limited by m e m b r a n e transport. This concept serves as the basis that might explain t h e selective m odulation of (6R)-DDATHF toxicity by folate s u p p l e m e n t a t i o n . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 302 (3) Effects of total cellular folate pools Phase I trials of (6R)-DDATHF (lometrexol) began almost te n years ago. However, further clinical ad vancem ent of this agent h a s been halted by the unexpected degree of prolonged and c u m u l a t i v e m yelosuppression, usually occurring after the second or third co u rse of treatm ent (Nelson et al., 1990; Young et al., 1990; and Ray et al., 1993). It was subsequently shown by the late Dr Grindey and his then colleagues that when mice were fed on a folate-restricted diet, the lethality of lometrexol increased by a 1000-fold, mimicking th e unexpected toxicity seen in hum ans. Interestingly, the c o administration of an intermediate level of folic acid with lo m etrex o l to animals ameliorated drug toxicity without com prom ising its a n t i tumor activity. At high level of folic acid intake, the drug is b o th non-toxic and non-therapeutic (Alati, 1996). These results h a v e prom pted the initiation of a series of second generation phase I studies which involve the co-adm inistration supplem ental folate with lometrexol (Young et al., 1992, Laohavinij et al., 1996; Synold et al., 1998; and Roberts et al., 2000). Three highly tolerable sc h e d u le s and dose com binations of lo m etrex o l/su p p lem en tal folate have been re co m m en d ed for phase II evaluation (Laohavinij et al., 1996; Sessa et al., 1996; and Roberts et al., 2000). The most dose in te n s e combination consists of lometrexol given at 170 mg/m2 every 3 weeks Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 303 with folic acid 5 mg administered orally for 1 week preceding a n d following each dose of drug (Laohavinij et al., 1996). Roberts et al. (2000) proposed a more frequent adm in istratio n of drug at 10.4 m g / m 2 per week with folic acid 3 m g / m 2 orally per day. A t h i r d recom m ended phase II schedule and dose com bination in v o lv es lometrexol 60 m g / m 2 every 4 weeks with delayed leucovorin rescu e given 10 mg orally four times daily on days 5-7 (Sessa et al., 1996). It is conceivable that, among these lom etrexol/folate c o m b in a tio n s , there exists one that has the highest therapeutic value; this will o n ly be discernible from phase II/III studies. It has been hypothesized that the cum ulative toxicity was d u e to accum ulation of drug and its polyglutam ate derivatives in ta rg e t tissue or in non-target tissue {e.g. liver) with subsequently r e distribution to target tissue. In accord with the later phenomenon, it has been shown that using [^C ](6R )-D D A T H F and a u t o r a d i o g r a p h y , the drug accu m u lated in liver of folate-deficient animals as polyglutamates and these hepatic drug m etabolites were then slowly released to the circulation, mimicking a very toxic c o n t i n u o u s infusion; this does not occur in animals with higher folate i n t a k e (Pohland et al., 1994). In hum ans, lometrexol p l a s m a pharm acokineitcs are best described by a three-co m p artm en t m o d e l (Wedge et al, 1995; and Synold et al., 1998). However, unlike w h a t Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 304 was observed in animals, the pharm acokinetics of plasma lom etrexol in patients, including the terminal elim ination phase, were found to be independent of drug dose and folic acid su p p lem entation (W edge et al., 1995; Synold et al., 1998; and Roberts et al., 2000). This w o u ld be at odd with the hypothesis that folic acid s u p p l e m e n t a t i o n modulates toxicities of lometrexol by changing the p h a r m a c o k in e tic s of the antimetabolite. Another hypothesis is that cum ulative toxicity of lometrexol is due to the accumulation of drug polyglutamates in stem cells. Since mature erythrocytes lack FPGS activity and therefore do not activ ely accum ulate folylpolyglutam ates (Barredo and Moran, 1992), it h a s been proposed that RBC lometrexol polyglutam ates can be used as a surrogate for drug content in hem atopoeitic progenitor cells. In tw o phase I studies, RBC lometrexol was found to rise over several weeks following a bolus infusion of drug, long after plasm a lometrexol level became undetectable (Synold et al., 1998; and Roberts et al., 2000). This observation most likely represents m aturation and release of erythrocytes that are loaded with lometrexol polyglutamates into th e circulation. Synold et al. proposed that since rising RBC lom etrexol levels correlated with a fall in several hem atological parameters, RBC drug level m ight serve as a predictor of cum ulative toxicity. T h is type of analysis is conceptually incorrect. Although RBC folates m a y Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 305 be a better reflection of long term physiological folate status of th e host, one cannot extrapolate this to using RBC antifolates as a surrogate for drug metabolites accum ulated in stem cell p o p u la tio n s during treatment. This is because, once the intracellular d ru g content in progenitor cells has reached a cytotoxic level, m a t u r a t i o n stops and cell death ensures in the bone marrow; only th e subpopulation of stem cells that have accumulated a sub-lethal level of drug would appear in the circulation. Thus, RBC lometrexol m a y be a poor indicator of the total amount of drug being actively ta k e n up by hematopoietic stem cells during drug exposure. Whereas the exact nature of the interaction between n a t u r a l folates and lometrexol has yet to be defined, m odulation of (6R)- DDATHF toxicity by supplem entary folate can be explained by a mechanism which is formally identical to that uncovered by th e L1210/D3 mutant, namely, a blockade of the polyglutam ation of (6R)-DDATHF by intracellular folates co m p o u n d s. One possible scenario is that target tissues susceptible to drug toxicity contain a high level of FPGS. Under physiological condition where p l a s m a concentration of 5 -m ethyltetrahydrofolate concentration is at 2-10 nM, cellular folate accum ulation is limited by transport and is independent of FPGS level. However, under most p h a r m a c o lo g ic a l conditions (e.g. during a bolus infusion of lometrexol), transport of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 306 antifolate is not rate-limiting, and the accum ulation of its polyglutamate metabolites would become proportional to FPGS level. Thus, cells containing higher level of FPGS are predicted to be m o re sensitive to antifolate. This concept is supported by a recent re p o rt by Aghi et al. (1999), who have dem o n strated that the sensitivity of glioma cells to two classical folate-based DHFR inhibitors, MTX a n d edatrexate, can be enhanced upon stable transfection of human FPGS cDNA into these cells. These results have also provided the proof-of- principle for this particular approach to gene therapy. Of note, these experiments were carried in standard culture m edium c o n ta in in g folic acid, a condition where it appears that accum ulation of ce llu lar folylpolyglutam ates is limited by transport but not FPGS level. Although changes in cofactor levels after FPGS overexpression h a v e not been directly measured, the enhanced cytotoxicity of folate antagonists observed in these glioma cells would argue against a n y counteraction of antifolate polyglutam ation by an expanded folate pool. What happen following pre-treatm ent with p h a r m a c o lo g ic a l dose of supplem ental folate? Under these conditions th e accum ulation of cofactors is no longer transport-lim ited; cells containing high level of FPGS will therefore enjoy an expansion of total folate pools, which can subsequently limits the accum ulation of (6R)-DDATHF. Based on this model, “folate priming” w o u ld Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 307 selectively protect cells that express high level of FPGS. In a c c o r d with this, it has been shown that purified normoblasts express very high levels of FPGS activity (Barredo and Moran, 1992; Barredo et al., 1994). The exact biochemical nature of the inhibition of (6R)-DDATHF polyglutam ate formation by cellular folates remains unclear. It is reasonable to assume that this control mechanism takes place at t h e level of FPGS. Thus, previous studies have shown that the K m v a lu e s for FPGS of most of the m onoglutam ate forms of the p h y sio lo g ic al folates were com parable or even lower than that of (6R)-DDATHF (Taylor et al., 1985; Moran and Colman 1984a); and, as the c h a i n length increases, the Km values for FPGS within a series of homologous com pounds generally decrease (Cichowicz and S hane, 1987; Chen et al., 1996). Hence, a simple mechanism that w o u ld explain this effect is that folate cofactor function as c o m p e titiv e inhibitors for the FPGS reaction, blocking the conversion of DDATHF to retainable polyglutam ate derivatives. Alternatively, m a m m a l i a n FPGS has been com m only found to display distinct s u b s tra te inhibition at higher concentrations of folate com pounds (Moran e t al., 1984), suggesting the possibility of an allosteric binding site of folates in this enzyme with regulatory significance. Regardless of t h e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 308 exact nature of inhibition of FPGS by cellular folates, we s p e c u la te that this type of feed-back control is likely to reflect a m e c h a n i s m that is in place to maintain homeostasis of folate cofactors r e q u ir e d for one-carbon metabolism. (4) Inhibition of GARFT by (6R)-DDATHF polyglutamates C om pared with the near stoichiometric binding between MTX and DHFR (Kj in the order of 1 0 '1 1 M) (Werkheiser, 1961), (6R)- DDATHF is a relatively weak inhibitor against its target e n z y m e GARFT, with K D and Kj values of 2-9 1 0 ’9 M (obtained u sin g equilibrium dialysis and sensitive kinetic assays that allowed in itia l velocity m easu rem en t) (Sanghani and Moran, 1997). W hereas p olyglutam ation of MTX has little effect its binding with DHFR (Cheng et al., 1985; Sikora et al., 1988), the interaction with GARFT is enhanced by about 10-fold upon metabolic conversion of (6R)- DDATHF into long chain p o ly glutam ates (Kj of DDATHF pentaglutamate = 0.5 nM) (Baldwin et al., 1990). Another i m p o r t a n t distinction is that, with MTX, DHFR inhibition results in a d r a m a t i c build-up of dihydrofolate in the form of polyglutam ates (Moran et al., 1975), which can competitively displace MTX from DHFR (W hite et al., 1975). However, (6R)-DDATHF, due to interconversion of 10- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 309 form ylletrah y d ro fo late into other forms of cofactors, it is p r e d ic t e d that inhibition of GARFT would not lead to an accum ulation of th e competing substrate which can displace the inhibitor and reestab lish an uninhibited flux for the de n o v o purine synthesis pathway (Fig. 1.1). This allowed perturbation of the pathway to a point such t h a t no new steady state is possible, and the inhibited cell u l t i m a t e l y would stop dividing and eventually dies. Post-target biochemical events Unlike MTX, which causes inhibition of both purine a n d th y m idylate synthesis following DHFR inhibition as a result of depletion of cellular reduced folate pools, (6R)-DDATHF selectively blocks purine synthesis by inhibiting GARFT, with little direct effect on thym idylate production (Beardsley et al., 1989). A num ber of classical purine analogs such as 6 -m ercaptopurine also inhibits t h e de n o v o pathway, but their nucleotide metabolites are also incorporated into nucleic acids, thus com pounding their m e c h a n i s m of action (Hitchings and Elion, 1954). Hence, for the first time, (6R)- DDATHF offers the opportunity to study the pharm acological effects of a selective inhibitor of de novo purine biosynthesis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 310 Perhaps the central questions regarding folate a n ta g o n is t inhibitory to purine synthesis are the issues of (1) whether su ch agent is cytotoxic, or is merely cytostatic; (2) if it is indeed cytotoxic, how purine synthesis inhibition leads to cell death; and (3) how to employ knowledge of post-target m olecular events to d esig n in g tumor-selective and synergistic therapeutic regimens. Effects of GARFT inhibition on nucleotide metabolism a p p e a r s to be cell type-specific and/or dose-dependent. Thus, in MOLT-4 human T-cell leukemia cells, a 6 h exposure to 5 pM of 5- deazaacyclotetrahydrofolate, a closely related (6R)-DDATHF a n a lo g u e (Table 2.1), resulted in 60% and 90% depletion of cellular GTP a n d ATP, respectively. After a transitory increase, UTP and CTP were depleted to 30% of control. Deoxynucleotides were also depleted by the drug; dCTP was redueced to the greatest extent, followed by dATP, dTTP, and dGTP, respectively. On the other hand, in CCRF-CEM human T-lym phoblastic leukemia, after a 4-h incubation with I m M (6R)-DDATHF, marked reductions in ATP and GTP pools were observed, with little effects on CTP, UTP, dATP and dGTP pools cells (Pizzorno et al., 1991). Borsa and Whitmore (1969) have reported that cell kill induced by MTX in L-cells could be potentiated by an exo g en o u s supply of hypoxanthine, suggesting that the anti-purine effect of th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 1 1 drug might in fact be antagonistic to its overall cytotoxicity. This has raised the question whether a pure de n o v o purine synthesis inhibitor such as (6R)-DDATHF is cytocidal. We have studied th e cytotoxicity of (6R)-DDATHF and of the thymidylate syntase in h ib ito r ZD-1694 to W iDr human colonic carcinoma cells (Smith et al., 1993). Using carefully designed clonogenic assays, our laboratory h av e d em onstrated that DDATHF is indeed cytotoxic. However, th e cytotoxic properties of (6R)-DDATHF were markedly different from that associated with ZD-1694. First, a maximal cell kill of 5-6 logs was achieved with ZD-1694, consistent with the elim ination of all viable cells but preexisting drug resistant mutants; whereas a maximal of only 2-3 logs was found with (6R)-DDATHF. Second, morphological changes in cells treated by the two drugs were remarkably different. Cells treated with ZD-1694 u n d e r w e n t megalocytic changes and became detached from the tissue cu ltu re dishes after 1-2 days. On the other hand, cells exposed to (6R)- DDATHF rem ained adherent to the dishes for >10 days after treatment, even though they failed to form viable colonies. Thus, it appears that selective folate inhibitor of purine and th y m id y l a t e synthesis each produce distinct cytotoxic end-points, presumably via fu ndam entally different biochemical m echanism s. What t h e n account for their disparate cytocidal aftermath? Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 1 2 Cellular demise in biological systems is more active a n d structured than previously thought. Apoptosis and necrosis are considered the two major modes of cell death. It is now w id ely accepted that apoptosis accounts for most physiological cell d e a t h , and that, most, if not all, chem o th erap eu tic agents exert t h e i r cytotoxicity by apoptosis. Programmed cell death, a term often u s e d interchangeably with apoptosis pertains to the complex b io c h e m ic a l processes that underlie apoptotic cell death. The original d e f in itio n of apoptosis referred to the characteristic m orphological c h a n g e s associated with the dying cell including DNA f r a g m e n ta tio n , chromatin condensation, mem brane blebbing, cell shrinkage, a n d disassembly of cells into m em brane-bound particles or a p o p t o t i c bodies (Wyllie et al., 1980). In contrast, necrotic cell death is characterized by intact chromatin, organelle, loss of p l a s m a m embrane integrity, and inflammatory response. A comprehensive review of the topic of apoptosis is beyond t h e scope of this thesis. In brief, the core players of this complex cell death pathway can be reduced to a few critical families of homologous proteins that are conserved throughout the evolution of metazoans (the caspases family, the bcl-2 family, the inhibitor of apoptosis protein (IAP) family, and the tum or necrosis factor d e a t h Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 313 receptor family) (reviewed in Korsmeyer, 1999; Reed, 1999a; Reed, 1999b; Bossy-Wetzel and Green, 1999; Eastman and Rigas, 1999). In response to cellular damage imparted by a variety of toxic stimuli, apoptosis activation usually begins at the level of th e m itochondrion, causing release of cytochrome c into the c y to p la s m . Upon binding to an adaptor molecule called apoptotic p ro tea se activating factor-1 (Apaf-1), c y to c h ro m e c activates its d o w n s t r e a m targets which encom pass a family of intracellular cysteine pro teases known as caspases. These proteases cleave substrates after a s p a ra tic acid and are produced initially as inactive zymogens. O nce activated, they can function as initiators (caspase-8, caspase-9, a n d caspase-10) by activating downstream caspases or as effectors (caspase-3, caspase-6, caspase-7) by cleaving substrates resulted in cellular disassembly. Known “death substrates” of effector caspases include inhibitor of deoxyribonuclease responsible for DNA fragmentation (I0AD), nuclear structural proteins (lamin A and la m in B), cytoskeleton regulatory proteins (gelsolin, focal adhesion kinase, and p21-activated kinase 2), as well as proteins involved in DNA repair and replication (DNA-PKcS, poly(ADP-ribose) polymerase, a n d replication factor C) (for review see Thornberry and Lazebnik, 1998). The apoptotic pathways are under stringent regulation by products of oncogenes and tumor suppressor genes. One of th e Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 314 major regulatory gene families is represented by bcl-2 and its homologues. The bcl-2 gene was first discovered by virtue of its involvement in the t( 14; 18) chrom osom al translocations c o m m o n l y found in indolent lym phom as (Tsujimoto et al., 1985). It fu n ctio n s as an inhibitor of apoptosis, thus contributes to tumorogenesis b y presumably disrupting physiological turnover of genetically in ju re d cells (McDonnell et al., 1989). In humans, at least sixteen bcl-2 homologues have been identified, some of which, like bcl-2 itself, operate as inhibitors of apoptosis (Bcl-XL, Mcl-1, and Al), and others as promoters of cell death (Bax, Bak, Bcl-Xs, and BAD) (Korsmeyer, 1999; Reed, 1999a; and Adams and Cory, 1998). Bcl-2 and its p r o survival relatives inhibit cell death by several m echanisms: (1) bcl-XL directly binds Apaf-1, preventing it from activating caspase-9; (2) bcl- 2 can inhibit cytochrom e c release from the m itochondrion, th ereb y preventing the activation of Apaf-1 and upstream caspases; and (3) by interfering with the function of its proapoptotic h o m o lo g u e s (Korsmeyer, 1999; Reed, 1999a; and Adams and Cory, 1998). Since most, if not all, ch em otherapeutic agents cause cell death by inducing apoptosis, bcl-2 family proteins confer a novel form of multidrug resistance phenotype by increasing cellular threshold to undergoing apoptosis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 15 A salient question is therefore whether folate inhibitor of purine synthesis causes apoptotic cell death? The answer is yes, a t least in a certain cell types. Thus, it has been shown that in MOLT-4 cells treated with 5-deazaacyclotetrahydrofolate, biochemical a n d morphological changes reminiscent of apoptosis indeed occurred, i.e. internucleosomal DNA fragmentation, chromatin condensation, a n d proteolysis of poly(ADP-ribose) polymerase and lamin B (Smith et al., 1993). If selective GARFT inhibitor can indeed trigger apoptosis, w h a t then account for the disparate morphologies and extent of cytotoxicities found in WiDr colonic carcinom a treated by ZD-1694 and (6R)-DDATHF, respectively? One possible explanation is that cell death induced by inhibition of thymidylate synthesis in WiDr cells is mediated by a non-apoptotic m echanism; this is unlikely to be th e case, since it has been well docum ented in other systems that folate- based TS inhibitor-induced cell death by apoptosis (Fisher et al., 1993; and Peters et al., 2000). Another explanation is that th e relative ineffective cell kill by (6R)-DDATHF in WiDr cells rep resen ts an altered form of apoptotic death. Traditionally, apoptosis and necrosis are c o n sid e re d conceptually and morphologically distinct forms of cell d e a th . However, it is evident from recent studies that there are s u b s ta n tia l overlaps between the two pathways. For instance, it has been sh o w n Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 16 that tissue hypoxia can induce both necrosis and apoptosis, and t h a t the proportion of these two modes is highly cell t y p e - d e p e n d e n t (Shimizu et al., 1996a). In addition, known anti-apoptotic m o le c u le s such as bcl-2, bcl-XL, and caspase inhibitors also inhibit necrotic cell death induced by hypoxia, suggesting com m on m ediators are s h a r e d by these two pathways (Shimizu et al., 1996a; Shimizu et al., 1996b; and Tsujimoto et al., 1997). Of particular interest, it has been dem onstrated that cell death induced by the same stimulus c a n proceed by either apoptosis or necrosis, depending on i n tr a c e llu la r ATP levels. Thus, in human lym phoid Jurkat cells, w h e n intracellular ATP was depleted with an inhibitor of m i t o c h o n d r i a l F0FiATPases, cell death caused by two apoptotic in d u cers, staurosporin and anti-CD95 antibodies (Fas/Apo-1 s t i m u l a t i o n ) , switched from an apoptotic to a necrotic m orphology (Leist et al., 1997). Using a similar experimental system, Tsujim oto and c o workers. have shown that ATP depletion in Jurkat and in HeLa cells completely inhibited both upstream caspase activation a n d apoptosis induced by Fas/Apo-1 stimulation, providing b io c h e m ic a l evidence for existence of ATP-dependent steps in the a p o p t o t i c pathway (Eguchi et al., 1997; Eguchi et al., 1999). These results might be of direct relevance to the mode of cell death induced by (6R)-DDATHF: since marked reduction in intracellular ATP levels is Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 317 often observed in cells treated with this class of agents, it is t e m p t i n g to speculate that the distinct morphology and relative ineffective cell kill associated with (6R)-DDATHF represent an attenuated form of apoptosis more akin to that of necrosis. However, it is important to emphasize that the intrinsic potency of a drug is of s e c o n d a ry interest com pared with its tumor-selective cytotoxicity. S uch selectivity will only be discernible from a better understanding of t h e biochemical factors affecting both pre-target and post-target e v e n ts in tumors and in normal stem cells. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 318 REFERENCES Adams, J. M., and Cory, S. (1998). The bcl-2 protein family: arbiters of cell survival. S cien ce, 281, 1322-26. Aghi, M., Kram m, C. M., and Breakefield, X. O. (1999). Folylpolyglutam yl synthetase gene transfer and glioma a n tifo la te sensitivity in culture and in vivo. J. Natl. Cancer Inst., 91, 1233-41. Aimi, J., Qiu, H., Williams, J., Zalkin, H. and Dixon, J. E. (1990). De n o v o purine nucleotide biosynthesis: cloning of human and a v ia n cDNAs encoding the trifunctional glycim am ide rib o n u c le o tid e sy n th etase-am in o im id az o le ribonucleotide s y n th e ta s e - g ly c in a m id e ribonucleotide transform ylase by functional co m p lem en tatio n in E. coli. Nucl. Acids Res. 18, 6665-72. Alati, T., Worzalla, J. F., Shih, C., Bewley, J. R., Sidney, L„ Moran, R. G., and Grindey, G. B. (1996). A ugm entation of the t h e r a p e u t i c activity of lometrexol [(6-R )5,10-dideazatetrahydrofolate] by oral folic acid. Cancer Res., 56, 2331-35. Allegra, C. J., Chabner, B. A., Drake, J. C., Lutz, R., Rodbard, D., a n d Jolivet, J. (1985a). Enhanced inhibition of thym idylate synthase by methotrexate polyglutamates. J. Biol. Chem., 2 6 0 , 720-6. Allegra, C. J., Drake, J. C., Joli vet, J., and Chabner, B. A. (1985b). Inhibition of phosphoribosylaminoimidazolecarboxamide transformylase by methotrexate and dihydrofolate acid polyglutam ates. Proc. Natl. Acad. Sci. (USA), 82, 4881-85. Allegra, C. J., Fine, J. R. L., Drake, C., and Chabner, B. A. (1987). T h e effect of methotrexate on intracellular folate pools in human MCF-7 breast cancer cells. Evidence for direct inhibition of p u r i n e synthesis. J. Biol. Chem, 261, 6478-85. Allegra, C. J., Fine, J. R. L., Drake, C., and Chabner, B. A. (1986). T h e effect of methotrexate on intracellular folate pools in human MCF-7 breast cancer cells. Evidence for direct inhibition of p u r i n e synthesis. J. Biol. Chem., 261, 6478-85. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 319 Almassy, R. J., Janson, C. A., Kan, C. C. and Hostomska, Z. (1992). Structures of apo and complexed Escherichia coli g ly c in a m id e ribonucleotide transformylase. Proc. Natl. Acad. Sci., 89, 6114-18. Alt, F. W., Kellems, R. E., Bertino, J. R., and Schimke, R. T. (1978). Selective m ultiplication of dihydrofolate reductase genes in m ethotrexate-resistant variants of cultured murine cells. J. Biol. C hem ., 253, 1357-70. Anderson, R. G., Kamen, B. A., Rothberg, K. G., and Lacey, S. W. (1993). Potocytosis: sequestration and transport of small m olecules by caveolae. S cien c e, 255, 410-11. Antonsson, B., Barredo, J., and Moran, R. G. (1990). A m icro assay for m am m alian folylpolyglutamate synthetase. Anal. B iochem ., 186, 8-13. Antony, A. C. (1992). The biological chemistry of folate receptors. B lood, 79, 2807-20. Antony, A. C. (1996). Folate receptors. Annu. Rev. Nutr., 16, 501- 2 1 . Antony, A. C., Verma, R. S., Unune, A. R., and LaRossa, J. A. (1989). Identification of a M g2+-dependent protease in human p la c e n ta which cleaves hydrophobic folate binding proteins to h y d ro p h ilic forms. J. Biol. Chem., 264, 1911-14. Appleman J. R., Prendergast, N., Delcamp, T. J., Freisheim, J. H., Blakley, R. L. (1988). Kinetics of the formation and isomerization of methotrexate complexes of recombinant hum an d ih y d ro fo la te reductase. J. Biol Chem. 2 6 3 ,1 0 3 0 4 -1 3 . Assaraf Y. G., Babani, S., and Goldman, I. D. (1998). Increased activity of a novel low pH folate transport associated with lipophilic antifolate resistance in Chinese hamster ovary cells. J. Biol. Chem ., 273, 8106-11. Assaraf, Y. G., and Goldman, I. D. (1997). Loss of folic acid ex p o rter function with markedly augmented folate accum ulation in lipophilic antifolate-resistant m am m alian cells. J. Biol. Chem ., 272, 17460-6. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 320 Baggott, J. E. (1983). Inhibition of purified avian liver am in oim idazole carboxamide ribonucleotide transform ylase by polyglutamates of methotrexate and oxidized folates. Fed. Proc., 4 2 , 662. Baldwin, S. W., Tse, A., Taylor, E. C., Rosowsky, A., Shih, C., a n d Moran, R. G. (1991). Structural features of 5,10-dideaza-5,6,7,8- tetrahydrofolate that determ ine inhibition of m a m m a l i a n glycinam ide ribonucleotide formyltransferase. B io ch e m istry , 3 0 , 1997-2006. Baldwin, S. W., Tse, A., Taylor, E. C., Rosowsky, A., Shih, C., a n d Moran, R. G. (1990). Inhibition of glycinamide rib o n u c le o tid e form yltransferase by 5,10-D ideazatetrahydrofolate Analogs. Proc. Am. Assoc. Cancer Res., 31, 341. Barredo, J. and Moran, R. G (1992). D eterm inants of a n tif o la te cytotoxicity: folylpolyglutam ate synthetase activity during c e llu la r proliferation and development. Mol. Pharmacol. 4 2 ,6 8 7 - 9 4 . Barredo, J. C., Synold, T. W., Laver, J., Relling, M. V., Pui, C-H., Priest, D. G., and Evans, W. E. (1994). Differences in constitutive and p o s t m ethotrexate folylpolyglutam ate synthetase activity in B-lineage and T-lineage leukemia. B lo o d , 84, 564-9. Barrueco, J. R., and Sirotnak, F. M. (1991). Evidence for th e facilitated transport of m ethotrexate polyglutam ates into ly so so m es derived from S180 cells. J. Biol. Chem., 266, 11732-7. Beardsley, G. P., Moroson, B. A., Taylor, E. C. and Moran, R. G. (1989). A new folate antim etabolite, 5,10-dideaza-5,6,7,8- tetrahydrofolate, is a potent inhibitor of de novo purine synthesis. J. Biol. Chem., 264, 328-33. Beardsley, G. P., Taylor, E. C., Grindey, G. B., and Moran, R. G. (1986). Deaza derivatives of tetrahydrofolic acid. A new class of folate antim etabolites. In: chemistry and Biology of the P te rid in es (Eds.: B. A. Cooper and V. M. W hitehead). De Gruyter and Co., New York, pp 953-7. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 321 Bertino, J. R., Kamen, B., and Romanini, A. (1997). F olate antagonists. In: Cancer medicine 4th edition (Eds: J. F. Holland, R. C. Bast, Jr., D. Morton, E. Frei III, D. W. Kufe, and R. R. Weichselbaum). Williams & Wilkins, pp 907-22. Biedler, J. L., and Riehm, H. (1970). Cellular resistance to actinom ycin D in Chinese hamster ovary cells in vitro: cross resistance, radioautographic, and cytogenetic studies. Cancer Res., 30, 1174. Biedler, J. L„ Melera, P. W., and Spengler, B. A. (1980). Specifically altered m etaphase chromosomes in antifolate-resistant C h in ese hamster cells that overproduce dihydrofolate re d u ctase. C a n cer Gene. & Cyto., 2, 47-60. Borsa, J., and Whitmore, G. F. (1969). Cell killing studies on t h e mode of action of methotrexate on L-cells in vitro. Cancer Res., 29, 737-44. Bossy-Wetzel, E., and Green, D. R. (1999). Apoptosis: checkpoint a t the mitochondrial frontier. Mutation Res., 4 3 4 , 243-51. Brigle, K. E., Westin, E. H., Houghton, M. T., and Goldman, I. D. (1991). C haracterization of two cDNAs encoding fo la te -b in d in g proteins from L1210 murine leukemia cells. J Biol. C h e m . 2 6 6 , 17243-9. Brigle, K. E., Spinella, M. J., Sierra, E. E., and Goldman, I. D. (1 9 9 7 ). Organization of the murine reduced folate carrier gene a n d identification of variants splice forms. Biochim. Biophys. Acta, 1 3 5 3 , 191-8. Brigle, K. E., Spinella, M. J., Esteban, E. S., and Goldman, I. D. (1 9 9 5 ). C haracterization of a mutation in the reduced folate carrier in a transport defective L1210 murine leukemic cell line. J. Biol. C h e m ., 270, 22974-9. Brigle, K. E., Spinella, M., Westin, E. H., and G oldm an, I. D. (1 9 9 4 ). Increased expression and characterization of two distinct folate binding proteins in murine ery th ro leu k em ia cells. B io c h e m . P h a r m a c o l. 47, 337-345. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 322 Brown, P. C., Beverley, S. M., and Schimke, R. T. (1981). R ela tio n s h ip of am plified dihydrofolate reductase genes to double m i n u t e chromosomes in unstably resistant mouse fibroblast cell lines. Mol. Cell. Biol., 1, 1077-83. Calvert, A. H., Alison, D. L., Harland, S. J., Robinson, B. A., J a c k m a n , A. L. , Jones, T. R., Newell, D. R., Siddik, Z. H., Wiltshaw, E., McElwain, T. J., Smith, 1. E., and Harrap, K. R. (1986). A phase I evaluation of the quinazoline antifolate thym idylate synthase inhibitor, N10- propargyl-5,8-didezafolic acid, CB3717. J. Clin. Oncol, 4, 1245-52. Calvert, H. (1999). MTA, a novel m ultitargeted antifolate, fro m preclinical to phase I and beyond: sum m ary and con clu sio n s. Semin O ncol., 26(2 Suppl. 6), 105-8. Campbell I. G., Jones, T. A., Foulkers, W. D., and Tro, J. (1991). Folate-binding protein is a marker for ovarian cancer. Cancer Res., 51, 5329-38. Caperelli, C. A. (1989). M am m alian glycinam ide rib o n u c le o tid e transformylase. J. Biol. Chem., 264, 5053-7. Carman, M. D., Schornagel, J. H., Rivest, R. S., Srim atkandada, S. Portlock, C. S., Duffy, T., and Bertino, J. R. (1984). Resistance to methotrexate due to gene am plification in a patient with a c u te leukemia. J. Clin. Oncol., 2, 16-20. Chabner, B. A., Allegra, C. J., Curt, G. A., Clendeninn, N. J., B ara m , J., Koizumi, S., Drake, J. C., and Jolivet, J. (1985). P o ly g lu ta m a tio n of m ethotrexate. Is methotrexate a prodrug? J. Clin. Invest., 7 6 , 907-12. Chen, L., Qi, H., Korenberg, J., Garrow, T. A., Choi, Y., and Shane, B. (1996). Purification and properties of hum an cytosolic folylpoly- gam m a-g lu tam a te synthetase and organization, localization, a n d differential splicing of the gene. J. Biol. Chem., 2 7 1 , 13077-87. Chen, P., Schulze-Gahmen, E. U., Stura, A., Inglese, J., Johnson, D. L., Marolewski, A., Benkovic, S. J., and Wilson, I. A. (1992). Crystal structure of glycinamide ribonucleotide transformylase from Escherichia coli at 3.0 A resolution. A target enzyme for chemotherapy. J Mol Biol, 2 2 7 ,2 8 3 -9 2 . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 323 Chen, V. J., Bewley, J. R., Andis, S. L., Schultz, R. M., Iversen, P. W., Shih, C., Mendelsohn, L. G., Seitz, D. E., and Tonkinson, ,J. L. (1999). Cellular p harm acology of MTA: a correlation of M T A -in d u c ed cellular toxicity and in vitro enzyme inhibition with its effect o n intracellular folate and nucleoside triphosphate pools in CCRF-CEM cells. Sem in Oncol., 26(2 Suppl 6), 48-54. Cheng, Y. C., Dutschm an, G. E., Starnes, M. C., Fisher, M. H., Nanavathi, N. T., and Nair, M. G. (1985). Activity of the n ew antifolates N 10-propargyl-5,8-dideazafolate and its p o l y g l u ta m a t e s against hum an dihydrofolate reductase, human t h y m i d y l a t e synthetase, and KB cells containing different levels of d ih y d ro fo la te reductase. Cancer Res., 45, 598-600. Chu, E., T akim oto, C. H., Voeller, D., Germ, J. L., and Allegra, C. J. (1993). Specific binding of human dihydrofolate reductase p r o t e i n to dihydrofolate reductase RNA in vitro. B io c h e m is tr y , 32, 4756-60. Church, G. M, and Gilbert, W. (1984). Genomic sequencing. Proc. Natl. Acad. Sci. (U S A), 7, 1991-5. Cichowicz, D. J., and Shane, B. (1987). M am m alian folylpoly- g am m a-g lu tam a te synthetase. 2. Substrate specificity and k in etic properties. B io c h e m is tr y , 26, 513-21. Clarke, S. J., Jackman, A. L., and Judson, I. R. (1993). The toxicity of ICI D1694 in man and mouse. Adv Expt Med. Biol., 338, 601-4. Clarke, S. J., Jackman, A. L., and Judson, I. R. (1993). The history of the dev elo p m en t and clinical use of CB3717 and ICI D1694. In: Novel ap p ro ach ed to selective treatments of human solid tu m o r s : laboratory and clinical correlation (Ed: Y. M. Rustum). P le n u m Press, New York, pp. 277-90. Cleland, W. W. (1967). The statistical analysis of enzyme kinetic data. Adv. E nzym ol. Relat. A rea s Mol. Biol., 2 9 , 1-3. Coney, L. R., Tomassetti, A., C arayannopoulos, L., Frasca, V., K a m e n , B. A., Colnaghi, M., and Zurawski, V. R. (1991). Cloning of a t u m o r - associated antigen: M O v l8 and M O v l9 antibodies recognize a folate binding protein. Cancer Res., 52, 3396-401. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 324 C unningham , D., Zalcberg, J. R., Rath, U., Oliver, I., Van-Cutsem, E., Stevenson, C., Seitz, J. F., Harper, P., Kerr, D., Perez-Manga, G. et al. (1995). T om udex' (ZD1694): results of a random ized trial in advanced colorectal cancer dem onstrate efficacy and r e d u c e d mucositis and leukopenia. The 'Tomudex' Colorectal Cancer S tu d y Group. Eu. J. Cancer, 31A , 1945-54. Curt, G. A., Carney, D. N., Cowan, K. H., Jolivet, J., Bailey, B. D., Drake, J. C., Kao-Shan, C. S., Minna, J. D., and Chabner, B. A. (1 9 8 3 ). Unstable m ethotrexate resistance in hum an small-cell c a r c i n o m a associated with double-m inute chrom osom e. N. Engl. J. Mecl., 3 0 8 , 199-202. Danenberg, P. V. (1977). Thym idylate synthetase - a target e n z y m e in cancer chemotherapy. Biochim. Biophys. Acta. 4 7 3 , 73-92. Daubner, S. C. and Benkovic, S. J. (1985). Characterization of m am m alian p h o sp h o rib o sy lg ly cin am id e formyltransferase fr o m transformed cells. Cancer Res., 45, 4990-7. Daubner, S. C., Young, M., Sammons, R. D., Courtney, L. F., a n d Benkovic, S. J. (1986). Structural and m echanistic studies on t h e HeLa and chicken liver proteins that catalyze g ly c in a m id e ribonucleotide synthesis and form ylation and a m i n o i m i d a z o l e ribonucleotide synthesis. B io c h e m istr y , 25, 2951-7. Dedhar, S., and Goldie, J. H. (1985). M ethotrexate-resistant h u m a n prom yelocytic leukemia (HL-60) cells express a d ih y d ro fo la te reductase with altered properties associated with increased e n z y m e activity. Biochem. Biophys. Res. Commun., 1 29, 536-45. Divekar, A. Y. and Hakala, M. T. (1975). Inhibition of t h e biosynthesis of 5'-phosphoribosyI-N -form ylglycinam ide in s a c r o m a 180 cells by homofolate. Mol. Pharmacol. 11, 319-25. Dixon, K. H., Lanpher, B. C., Chiu, J., Kelly, K., and Cowan, K. H. (1994). A novel cDNA restores reduced folate carrier activity a n d m ethotrexate sensitivity to transport deficient cells. J Biol. C h e m . 2 69, 17-20. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 325 Drori S., Jasen, G., Mauritz, R., Peters, G. J., and Assaraf, Y. G. (2000). Clustering of m utations in the first tran sm em b ran e domain of th e human reduced folate carrier in GW 1843U89-resistant leukemic cells with im paired antifolate transport and au g m en ted folate u p ta k e . J. Biol. Chem., 2 7 5 , 30855-63. Eastman, A. Apoptosis: A product of program m ed a n d unprogrammed cell death. (1993). Toxicol. Appl. P harm acol., 1 2 1 , 160-4. Eastman, A., and Rigas, R. R. (1999). M odulation of a p o p to s is signaling pathw ays and cell cycle regulation. Semin. Oncol., 2 6 (su p p l. 16), 7-16. Eguchi, Y., Srinivasan, A., Tomaselli, K. J., Shimizu, S., a n d Tsujimoto, Y. (1999). ATP-dependent steps in apoptotic signal transduction. Cancer Res., 59, 2174-81. Eguchi, Y., Shimizu, S., and Tsujimoto, Y. (1997). Intracellular ATP levels determ ine cell death fate by apoptosis or necrosis. C ancer Res., 57, 1835-40. Elwood, P. C. (1989). Molecular cloning and characterization of th e human folate-binding protein cDNA from placenta and m a l i g n a n t tissue culture (KB) cells. J. Biol. Chem., 2 6 4 , 14893-901. Ensminger, W. D (1984). Clinical pharm acology of folate an alo g s. In: Folate antagonists as therapeutic agents (Eds.: F. M. Sirotnak, J. J. Burchall, W. B. Ensminger, J. A. Montgomery)., pp 133-63. Evans, R. M., Laskin, J. D., and Hakala, M. T. (1981). Effect of excess folates and deoxyinosine on the activity and site of action of 5- fluorouracil. Cancer Res., 41, 3288-95. Fan, J. F., Vitols, K. S., and Huennekens, F. M. (1991). Biotin derivatives of methotrexate and folate. J. Biol. Chem., 266, 14862-5. Farber, S, Diamond, L. K., Mercer, R. D., Sylvester Jr, R. F., and Wolff, J. A. (1948). T em porary remissions in acute leukemia in c h ild re n produced by folic acid antagonist, 4-am in o p tero y l-g lu tam ic a c id (aminopterin). N. Engl. J. Med., 238, 787-93. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 326 Ferguson, P. L., and Flintoff, W. F (1999). Topological and f u n c tio n a l analysis of the human reduced folate carrier by h e m a g g lu tin in epitope insertion. J. Biol. Chem. 2 7 4 , 16269-78. Ferone, R, Hanlon, M. H., Singer, S. C., and Hunt, D. F. (1986). alpha-C arboxyl-linked glutamates in the folylpolyglutamates of Escherichia coli. J Biol. Chem., 2 61, 16356-62. Ferone, R., Hanlon, M., Kelly, J. K., McLean, E. W,, Cohn, N. K., D u c h , D., Edelstein, M., and Smith, G. (1990). Chemistry and Biology of Pteridine and Folate Derivatives, de Gruyter, Berlin. Fisher, T. C., Milner, A. E., Gregory, C. D., Jackman, A. L., Aherne, G. W., Hartley, J. A., Dive, C., and Hickman, J. A. (1993). Bcl-2 m odulation of apoptosis induced by anticancer : resistance to thym idylate stress is independent of classical resistance p a th w a y s . Cancer Res., 53, 3321-6. Foo, S. K. and Shane, B. (1982). Regulation of fo ly lp o ly -g a m m a - glutam ate synthesis in m am m alian cells. In v iv o and in v itro synthesis of p tero y lp o ly -g a m m a-g lu ta m a tes by Chinese h a m s t e r ovary cells. J Biol. Chem., 2 57, 13587-92. Freisheim, J. H., Price, E. M., and Ratnam, M. (1989). F olate coenzyme and antifolate transport proteins in normal a n d neoplastic cells. Adv Enz. Regul. 29, 13-16. Freisheim, J. H., Ratnam, M., Smith, P. L., Delcamp, T. J., and Price, E. M. (1988). Photoaffinity analogues of m ethotrexate as folate antagonist binding probes. Adv. Enz. Regul., 27, 25-29. Galivan J, Ryan, T. J., Chave, K., Rhee, M., Yao, R., and Yin, D. (2000). Glutamyl hydrolase, pharm acological role and e n z y m a t ic characterization. Pharmacol. Ther., 85, 207-15. Galivan, J. Ryan, T., Rhee, M., Yao, R., and Chave, K. (1999). Glutamyl hydrolase: properties and pharm acologic impact. S e m in . O n c o l., 26(2 S uppl 6), 33-7. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 327 Garrow, T. A., Admon, A., and Shane, B. (1992). Expression clo n in g of a hum an cDNA encoding fo ly lp o ly -g am m a-g lu tam ate s y n th e ta s e and determ ination of primary structure. Proc. Natl. Acad. Sci. (USA), 89, 9151-5. Goker, E., Lin, J. T., Trippett, T., Elisseyeff, Y., Tong, W. P., Niedzwiecki, D., Tan Charlotte, Steinherz, P., Schweitzer, B. I., a n d Bertino ,J. R. (1993). Decreased polyglutamation of m ethotrexate in acute lymphoblastic leukemia blasts in adults compared to c h ild r e n with this disease. L e u k e m ia , 7, 1000-4. Goldm an, I. D. (1969). Transport energetics of the folic a c id analogue, methotrexate, in L1210 cells. Enhanced accum ulation by metabolic inhibitors. J. Biol. Chem., 2 4 4 , 3779-85. G oldm an, I. D. (1971). The characteristics of the m e m b r a n e transport of am ethopterin and the naturally occurring folates. A n n N Y Acad Sci., 186, 400-22. G oldm an, I. D., Lichtenstein, N. S., and Oliverio, V. T. (1968). Carrier-m ediated transport of the folic acid analogue, m e th o tr e x a te , in the L1210 leukemia cell. J. Biol. Chem., 243, 5007-17. Gorlick, R., Goker, E., Trippett, T., Waltham, M., Banerjee, D., a n d Bertino, J. R. (1996). Intrinsic and acquired resistance to methotrexate in acute leukemia. N. Engl. J. Med., 3 3 5 , 1041-8. Goulian, M., Bleile, B., and Tseng, B. Y. (1980). M e th o tre x a te - induced m isincoporation of uracil into DNA. Proc. Natl. Acad. Sci. (USA)., 77, 1956-60. Habeck, L. L., Mendelsohn, L. G., Shih, C., Taylor, E. C., Colman, P. D., Gossett, L. S., Leitner, T. A., Schultz, R. M., Andis, S. L., a n d Moran, R. G. (1995). Substrate specificity of m a m m a l i a n folypolyglutam ate synthetase for 5 , 1 0 -d id e a z a te tr a h y d ro f o la te analogs. Mol. Pharmacol., 48, 326-33. Habeck, L. L., Leitner, T. A., Shackelford, K. A., Gossett, L. S., Schultz, R. M., Andis, S. L., Shih, C., Grindey, G. B., and M endelsohn, L. G. (1994). A novel class of m o n o g lu ta m a ted antifolates exhibits tight- binding inhibition of hum an glycinamide r i b o n u c le o tid e form yltransferase and potent activity against solid tumors. C ancer Res., 54, 1021-6. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 328 Henderson G. B. and Strauss, B. P. (1990). characteristics of a n o v el transport system for folate com p o u n d s in wild-type a n d methotrexate-resistant L1210 cells. Cancer Res., 50, 1709-14. Henderson, G. B. (1990). Folate-binding proteins. Annu. Rev. N utr., 10, 319-35. Henderson, G. B. and Zevely, E. M. (1981). Anion e x c h a n g e m echanism for transport of m ethotrexate in L1210 cells. B io c h e m . Biophy. Res. Commun. 99, 163-9. Henderson, G. B., and Zevely, E. M. (1983). Structural r e q u i r e m e n t for anion substrates of the m em brane transport system in L1210 cells. Arch. Biochem. Biophys., 2 2 1 , 438-46. Henderson, G. B., and Zevely, E. M. (1984). Affinity labeling of t h e 5 -m e th y lte tra h y d ro fo la te /m e th o tre x a te transport protein of L1210 cells by treatm ent with an N -hydroxysuccim ide ester of [^H Jm ethotrexate. J. Biol. Chem., 2 5 9 , 4558-62. Henderson, G. B., Suresh, M. R., Vitols, K. S., and Huennekens, F. M. (1986). T ransport of folate com pounds in L1210 cells: k in e tic evidence that folate influx proceeds via the high-affinity t r a n s p o r t system for 5-m eth y ltetrah y d ro fo late and m ethotrexate. Cancer Res. 46, 1639-43. Henderson, G. I., Perez, T., Schenker, S., Mackins, J., and Antony, A. C. (1995). M aternal-to-fetal transfer of 5 -m eth y ltetrah y d ro fo late b y the perfused human placental cotyledon: evidence for a concentrative role by placental folate receptors in fetal fo late delivery. J. Lab. Clin. Med., 126, 184-203. Hill B. T., Bailey, B. D., White, J. C., and Goldm an, I. D. (1 9 7 9 ). Characteristics of transport of 4-amino antifolates and fo late com pounds by two lines of L51178Y lymphoblasts, one w ith impaired transport of methotrexate. Cancer Res., 39, 2440-6. Hitchings, G. H., and Elion, G. B. (1954). The chemistry a n d biochemistry of purine analogs. Ann. N.Y. Acad. Sci., 60, 195. Hryniuk, W. M. (1972). Purineless death as a link between g ro w th rate and cytotoxicity by methotrexate. Cancer Res., 32, 1506-11. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 329 Jackman, A. L., Farrugia, D. C., Gibson, W., Kinbell, R., Harrap, K. R., Stephens, T. C., Azab, M., and Boyle, F. T. (1995). Z D 1 6 9 4 (Tomudex): a new thym idylate synthase inhibitor with activity in colorectal cancer. Eu. J. Cancer, 31A (7-8), 1277-82. Jackman, A. L., Jodrell, D. I., Gibson, W., and Stephens, T. C. (1991a). ICI D1694, an inhibitor of thym idylate synthase for clinical study. Adv. in Expt Med. Biol. 3 0 9 A, 19-23. Jackman, A. L., Taylor, G. A., Gibson, W., Kimbell, R., Brown, M., Calvert, A. H., Judson, I. R., and Hughes, L. R. (1991b). ICI D1694, a quinazoline antifolate thym idylate synthase inhibitor that is a potent inhibitor of L1210 tumor cell growth in vitro and in vivo: a new agent for clinical study. Cancer Res., 51, 5579-86. Jackson, R. C., Jackman, A. L., and Calvert, A. H. (1 9 8 3 ). Biochemical effects of the quinazoline inhibitor of t h y m i d y l a t e synthetase, CB3717, on human lym phoblastoid cells. B io c h e m . P h a r m a c o l ., 33, 3783-90. Jackson, R. C., and Grindey, G. B. (1984). The biochemical basis for m ethotrexate cytotoxicity. In: Folate anatagonists as t h e r a p e u t i c agents. (Ed: F.M. Sirotnak). Academic Press, Orlando, FL, pp. 289- 315. Jackson, R. C., and Harrap, K. R. (1979). Com puter models of anticancer drug interaction. Pharm acol Ther., 4, 245-80. Jackson, R. C., and Harrap, K. R. (1980). Studies with a mathematical model of folate metabolism. Arch. Biomed. B io p h ys., 158, 827-41. Jackson, R. C., Hart, L. I., and Harrap, K. R. (1976). Intrinsic resistance to methotrexate of cultured mammalian cells in relation to the inhibition kinetics of their dihydrololate reductases. Cancer Res., 36,1991-7. Jackson, R. C., Jackman, A. L., and Calvert, A. H. (1 9 8 3 ). Biochemical effects of a quinazoline inhibitor of t h y m i d y l a t e synthetase, N -(4 -(N -((2-am ino-4-hydroxy-6- q u in azo lin y l)m e th y l)p ro p -2 -y n y la m in o )b e n z o y l)-L -g lu ta m ic a c id (CB3717), on hum an lym phoblastoid cells. Biochem P h a rm acol., 32, 3783-90. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 330 Jansen, G., Westerhof, G. W., K athm ann, I., Rademaker, B. C., Rijksen, G., and Schornagel, J. H. (1989a). Identification of a m em b ran e-asso ciated folate-binding protein in human leu k em ic CCRF-CEM cells with transport-related m ethotrexate resistance. Cancer Res., 49, 2455-9. Jansen, G., Kathm ann, I., Rademaker, B. C., Braakhuis, B. J., Westerhof, G. R., Rijksen, G., and Schornagel, J. H. (1989b). Expression of a folate binding protein in L1210 cells grown in low folate medium. Cancer Res., 49, 1959-63. Jansen, G., Mauritz, R., Drori, S., Sprecher, H., K athm ann, I., B unni, M., Priest, D. G., Noordhuis, P., Schornagel, J., H., Pinedo, H. M., Peters, G. J., and Assaraf, Y. G. A structurally altered h u m a n reduced folate carrier with increased folic acid transport mediates a novel m ech an ism of antifolate resistance. J. Biol. Chem., 2 7 3 , 3 0 189-98. Joli vet, J., Cowan, K. H., Curt, G. A., Lendeninn, N. J., and C h ab n er, B. A. (1983). The pharmacology and clinical use of m e th o tre x a te . N. Engl. J. Med., 309, 1094-1104. Jolivet, J., Schilsky, R. L., Bailey, B. D., Drake, J. C., and Chabner, B. A. (1982). Synthesis, retention, and biological activity of methotrexate polyglutamates in cultured hum an breast cancer cells. J. Clin. Invest., 7 0 , 351-60. Jones, T. R„ Calvert, A. H., Jackman, A. L., Brown, S. J., Jones, M., Harrap, K. R. (1981). A potent antitum or quinazoline inhibitor of thym idylate syntase: synthesis, biological properties, a n d therapeutic results in mice. Eur. J. Cancer, 17, 11-19. Juliano, R. L., and Ling, V. (1976). A surface g ly co p ro tein m odulating drug permeability in Chinese hamster ovary cell m u tan ts. Biochim. Biophys. Acta, 4 5 5 , 152-62. Kamen, B. A., Smith, A. K., and Anderson, R. G. W. (1991). T h e folate receptor works in tandem with a probenecid-sensitive carrier in MA104 cells in vitro. J. Clin. Invest., 87, 1442-9. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 331 Kamen, B. A., and Capdevila, A. (1986). R eceptor-m ediated folate accumulation is regulated by the cellular folate content. Proc. Natl. Acad. Sci. (USA), 83, 5983-7. Kamen, B. A., Wang, M., Strekfuss, A. J., Peryea, X., and Anderson, R. G. W. (1988). Delivery of folates to the cytoplasm of MA104 cells is m ediated by a surface m em brane receptor that recycles. J, Biol. C h em ., 2 6 3 , 13602-9. Kan, C.-C., Gehring, M. R., Nodes, B. R., Janson, C. A., Almassey, R. J., and Hostomska, Z. (1992). Heterologous expression and p u r if ic a tio n of active hum an ph o sp h o rib o sy lg ly cin am id e form yltransferase as a single domain. J. Protein Chem., 11, 467-73. Kan, J. L.-C., and Moran, R. G. (1995). Analysis of a mouse gene encoding three steps of purine synthesis reveals use of an in tro n ic polyad en y latio n signal without alternative exon usage. J. Biol. C h em ., 2 7 0 , 1823-32. Kan, J. L.-C., Jannatipour, M., Taylor, S. M., and Moran, R. G. (1993). Mouse cDNAs encoding a trifunctional protein of de n o v o purine synthesis and a related single-domain g l y c in a m id e ribonucleotide synthetase. G e n e , 137, 195-202. Kesavan, V., Doig, M. T., and Priest, D. G. (1986). A r a d i o e n z y m a t i c m ethod for the determ ination of tissue 1 0 -f o r m y lte tr a h y d ro f o la te . J. Biochem Biophys. Methods, 12, 311-7. Keyomarsi K., and Moran, R. G. (1988). Mechanism of cy to to x ic synergism of fluoropyrim idines and folinic acid in mouse le u k e m ic cells. J. Biol. Chem., 263, 14402-14409. Kisliuk, R. L. (1981). Pteroylpolyglutam ates. Mol. Cell. B io ch e m ., 39, 331-45. Korsmeyer, S. J. (1999). BCL-2 gene family and the regulation of programmed cell death. Cancer Res., 59 (suppl), 1693S-1700S. Kumar P., Kisliuk, R. L., Gaum ont, Y., Freisheim, J. H., Nair, M. G. (1989). Inhibition of human dihydrofolate reductase by an tifo ly l polyglutamates. B iochem Pharm acol., 38, 541-43. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 332 Laemmli, U. K (1970). Cleavage of structural proteins during t h e assembly of the head of bacteriophage T4. N a tu re , 227, 680-5. Langenbach, R. J., Danenberg, P. V., and Heidelberger, C. (1 9 7 2 ) Thym idylate synthetase: M echanism of inhibition by 5-fluoro-2'- deoxyuridylate. Biochem. Biophys. Res. Commun. 48, 1565-71. Laohavinij, S., Wedge, S. R., Lind, M. J., Nigel, B., Humphreys, A., Proctor, M., C hapm an, F., Simmons, D., Oakley, A., Robson, L., Gumbrell, L., Taylor, G. A., Thomas, H. D., Boddy, A. V., Newell, D. R., and Calvert, A. H. (1996). A Phase I clinical study of th e antipurine antifolate lometrexol (DDATHF) given with oral folic acid. Invest. New Drugs, 14, 325-35. Lee, G. S., Blonsky, K. S., Van On, D. L., Savage, E. A., Morgan, A. R., and von Borstel, R. C. (1992). Base alterations in yeast induced by alkylating agents with different Swain-Scott substrate c o n s ta n ts . J. M o l B io l, 223, 617-26. Leist, M., Single, B., Castoldi, A. F., Kuhnle, S., and Nicotera, P. (1997). Intracellular adenosine triphosphate (ATP) concentration: a switch in the decision between apoptosis and necrosis. J Exp M e d ., 185, 1481-6. Li, W. W., William, M., Tong, W., Schweitzer, B. I., and Bertino, J. R. (1993). Increased activity of y-glutamyl hydrolase in h u m a n sacroma cell lines: a novel m echanism of intrinsic resistance to methotrexate (MTX). Adv. Exp. Med. Biol., 338, 635-8. Li, W., Fan, J., Hochhauser, D., Banerjee, D., Zielinski, Z., Almasan, A., Yin, Y., Kelly, R., Wahl, G. M., and Bertino, J. R. (1995). Lack of functional retinoblastom a protein mediates increased resistance to antimetabolites in human sarcoma cell lines. Proc. Natl. Acad. Sci. (USA), 92, 10436-40. Lin, B. F., Huang, R. F., and Shane, B. (1993). Regulation of folate and one-carbon metabolism in m am m alian cells. III. Role of m itochondrial fo ly l-gam m a-glutam ate synthetase. J. Biol. C h e m ., 268, 21674-9. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 333 Lin, J. T., Tong, P., Trippett, T. M., Niedzwiecki, D., Tao, Y., Tan, C., Steinherz, P., Schweitzer, B, I., and Bertino, J. R. (1991). Basis for natural resistance to methotrexate in human acute n o n - lymphocytic leukemia. Leukemia Res., 15, 1191-6. Lin. J. T., and Bertino, J. R. (1991). Update on trimetrexate, a folate antagonist with antineoplastic and antiprotozoal p ro p e rtie s . Cancer Invest., 9, 159-72. Lowe, J. K., and Grindey, G. B. (1976). Inhibition of growth rate a n d deoxynucleoside triphosphate concentrations in cultured l e u k e m ia L1210 cells. Mol. Pharmol., 12, 177-84. Matherly, L H., Czajkowski, C. A., and Angeles, S. M. (1991). Identification of a highly glycosylated m ethotrexate m e m b r a n e carrier in K562 human erythroleukem ia cells up-regulated for tetrahydrofolate cofactor and methotrexate transport. Cancer Res., 51, 3420-6. Matherly, L. H., Angeles, S. M., and Czajkowski, C. A. (1992). Characterization of transport-m ediated m ethotrexate resistance in human tumor cells with antibodies to the m em brane carrier for m ethotrexate and tetrahydrofolate cofactors. J. Biol. C h e m ., 2 6 7 , 23253-60. Matherly, L. H., Angeles, S. M., McGuire, J. J. (1993). D e te r m i n a n t s of the disparate antitum or activities of (6R )-5,10-dideaza-5,6,7,8- tetrahydrofolate and methotrexate toward human ly m p h o b la s tic leukemia cells, characterized by severely impaired a n tif o la te membrane transport. Biochem. Pharmacol., 4 6 , 2185-95. Matherly, L. H., Taub, J. W.,Y. Ravindranath, Proefke., S. A, Wong, S. C., Phyllis, G., Buck, S., Wright, J. E., and Rosowsky, A. (1995). Elevated dihydrofolate reductase in m ethotrexate resistance in children acute lymphoblastic leukemia. B lo o d , 85, 500-9. Matthews, D. A, Bolin, J. T., Burridge, J. M., Filman, D. J., Volz, K. W., and Kraut, J. (1985). Dihydrofolate reductase, th e stereochemistry of inhibitor selectivity. J. Biol. Chem., 2 6 0 , 392-9. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 334 Matthews, D. A., Appelt, K., and Oatley, S. J. (1989). C rystal structure of Escherichia coli thym idylate synthetase with FdUMP and 10 propargyl-5,8-dideazafolate. Adv. Enzyme Regul., 29, 47-60. Mayor, S. K., Rothberg, G., and Maxfield, F. R. (1994). S e q u e s tra tio n of GPI-anchored proteins in caveolae triggered by cross-linking. Science, 264, 1948-51. McBurney, M W., and Whitmore, G. F. (1974). Isolation a n d biochemical characterization of folate deficient m utants of C hinese hamster cells. Cell , 2, 173-82. McCloskey, D. E., McGuire, J. J., Russel, C. A., Rowan, B. G., Bertino, J. R., Pizzorno, G., and Mini, E. (1991). D ecreased folylpolyglutam ate synthetase activity as a m echanism of m ethotrexate resistance in CCRF-CEM human leukemia sublines. J. Biol. Chem., 2 6 6 , 6181-7. McCormick, J. I., Susten, S. S., and Freisheim, J. H. (1981). C haracterization of the m ethotrexate transport defect in a re s is ta n t L1210 lymphoma cell line. Arch. Biochem. Biophys., 2 1 2 , 311-8. McDonnell, T. J., Deane, N., Platt, F. M., Nunez, G., Jaeger, U., McKearn, J. P. and Korsmeyer, S. J. (1989). B c /-2 -im m u n o g lo b u lin transgenic mice dem onstrate extended B cell survival and fo llicular lymphoproliferation. Cell, 57, 79-88. McGuire, J. J. and Bertino, J. R. (1981). Enzymatic synthesis a n d function of folylpolyglutamates. Mol. Cell. Biochem., 38, 19-48. McGuire, J. J., Heitzman, K. J., Haile, W. H., Cynthia, A. R„ McCloskey, D. E„ and Piper, J. R. (1993). Cross-resistance studies of folylpolyglutam ate synthetase-deficient, m e th o tr e x a te - r e s is ta n t CCRF-CEM human leukemia sublines. L eu kem ia , 7, 1996-2003. McGuire, J. J., Hsieh, P., Coward, J. K. and Bertino, J. R. (1980). Enzymatic synthesis of folylpolyglutam ates. C haracterization of th e reaction and its products. J Biol. Chem., 2 5 5 , 5776-88. Miyashita, T., and Reed, J. (1993). Bcl-2 oncoprotein blocks ch em o th e rap y -in d u c ed apoptosis in a human leukemia cell line. Blood, 81, 151-7. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 335 Moran R. G., Colman, P. D., Rosowsky, A., Forsch, R. A., and Chan, K. K. (1984). A m echanism for the addition of multiple moles of glutam ate by folylpolyglutam ate synthetase. Mol. P h a rm a c o l, 27, 156-66. Moran, R. G. (1983). Characterization of the function of m am m alian folylpolyglutam ate synthetase (FPGS). In: Folyl a n d antifolyl polyglutam ates (Eds.: I.D. Goldman, B. A. Chabner, and J. R. Bertino). pp. 327-339. Moran, R. G., and Colman, P. D. (1982). A simple procedure for th e synthesis of high specific activity tritiated (6S)-5- formyltetrahydrofolate. Anal. Biochem., 1 2 2 , 70-8. Moran, R. G., and Colman, P. D. (1984). M easurem ent of folylpolyglutam ate synthetase in m am m alian tissues. A n a l. B io c h e m . 1 4 0 , 326-42. Moran, R. G., Mulkins, M., and Heidelberger, C. (1979). Role of thym idylate synthetase activity in developm ent of m e t h o t r e x a te cytotoxicity. Proc. Natl. Acad. Sci. (USA), 76, 5924-8. Moran, R. G., Colman, P. D., and Jones, T. R. (1989). Relative substrate activities of structurally related pterine, quinazoline, a n d pyrim idine analogs for mouse liver folylpolyglutam ate s y n th e ta s e . Mol. Pharmacol., 36, 736-43. Moran, R. G., Baldwin, S. W., Taylor, E. C., and Shih, C. (1989). T h e 6 S- and 6 R-diastereomers of 5,10-dideaza-5,6,7,8-tetrahydrofolate are equiactive inhibitors of de n o v o purine synthesis. J. Biol. C hem ., 2 6 4 , 21047-51. Moran, R. G., Werkheiser, W. C., and Zakrzewski, S. F. (1976). Folate m etabolism in m am m alian cells in culture. I Partial characterization of the folate derivatives present in L1210 m o u s e leukemia cells. J. Biol. Chem., 251, 3569-75. Morrison, J. F. (1969). Kinetics of the reversible inhibition of en zym e-catalysed reactions by tight-binding inhibitors. B io c h im . Biophys. Acta, 185, 269-86. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 336 Morrow, C. S., and Cowan, K. H. (1997). Drug resistance and its clinical circum vention. In: Cancer medicine 4th edition (Eds: J. F. Holland, R. C. Bast, Jr., D. Morton, E. Frei III, D. W. Kufe, and R. R. Weichselbaum). Baltimore, Williams & Wilkins, pp 799-815. Moscow, J. A., Gong, M., He, R., Sgagias, M K., Dixon, K. H., Anzick, S. L., Meotzer, P. S., and Cowan, K. H. (1995). Isolation of a gene encoding a human reduced folate carrier (RFC1) and analysis of its expression in transport-deficient, m ethotrexate-resistant h u m a n breast cancer cells. Cancer Res., 55, 3798-804. Moyer, J. D., and Henderson, J. F. (1985). C o m p artm en tatio n of intracellular nucleotides in m am m alian cells. CRC Cri. Rev. B io ch em ., 19, 45-61. Murray, R. C., Williams, F. M. R„ and Flintoff, W. F. (1996). Structural organization of the reduced folate carrier gene in C hinese hamster ovary cells. J. Biol. Chem., 2 7 1 , 19174-9. Nair MG, Murthy, B. R., Patil, S. D., Kisliuk, R. L., Thorndike, J., Gaumont, Y„ Ferone, R„ Duch, D. S., and Edelstein, M. P. (1989). Folate analogues. 31. Synthesis of the reduced derivatives of 11- deazahom ofolic acid, 10-m ethyl-l 1-deazahom ofolic acid, and th e ir evaluation as inhibitors of glycinam ide rib o n u c le o tid e formyltransferase. J Med. Chem., 32, 1277-83. Nelson, R., Butler, F., Dugan, Jr., W., David-Land, C., Stone, M., a n d Dyke, R. (1990). Phase I clinical trial of LY264618 (dideazatetrahydrofolic acid, DDATHF). Proc. Am. Soc. Clin. Oncol., 9, 76. Niethammer, D., and Jackson, R. C. (1975). Changes in m o le c u la r properties associated with the developm ent of resistance ag a in st methotrexate in hum an lym phoblastoid cells. Eur. J. Cancer, 11, 45-854. Nunberg, J. H., Kaugman, R. J., Schimke, R. T., Urlaub, G., a n d Chasin, L. A. (1978). Amplified dihydrofolate reductase genes are localized to a hom ogeneously staining region of a single chromosome in a m ethotrexate-resistant Chinese hamster ovary cell line. Proc. Natl. Acad. Sci.. (USA), 75, 5553-6. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 337 O'Connor, B. M., Rotundo, R. F., Nimec, Z., McGuire, J. J., a n d Galivan, J. (1991). Secretion of y-glutamyl hydrolase in vitro. Cancer Res., 51, 3874-81. O'Dwyer, P. J., Nelson, K., and Thornton, D. E. (1999). Overview of phase II trials of MTA in solid tumors. Semin. Oncol., 26 ( S u p p l 6 ), 99-104. Pagani, O., Sessa, C., Dejong, J., Kern, H., Hatty, H., Schmit, H., a n d Cavalli, F. (1992). Phase I study of lometrexol (DDATHF) given in com bination with leucovorin. Proc. Amer. Soc. Clin. O ncol., 11, 109. Peters, G. J., van Triest, B., Backus, H. H., Kuiper, C. M., van der Wilt, C. L., and Pinedo, H. M. (2000). Molecular dow nstream events a n d induction of thym idylate synthase in mutant and wild-type p5 3 colon cancer cell lines after treatment with 5-fluorouracil and th e thym idylate synthase inhibitor raltitrexed. Ear J Cancer, 3 6 , 9 lb- 24. Pizzorno, G.. Cashmore, A. R., Moroson, B. A., Cross, A. D., Smith, A. D.. Marling-Cason, M., Kamen, B. A. and Beardsley, G. P. (1993). 5,10-dideazatetrahydrofolic acid (DDATHF) transport in CCRF-CEM and M A I04 cell lines. J. Biol. Chem., 268, 1017-23. Pizzorno. G., Moroson, B. A., Cashmore, A. R., and Beardsley, G. P. (1991). Effects of 5,10-dideaza5,6,7,8-tetrahydrofolate on n u c le o tid e metabolism in CCRF-CEM cells. Cancer Res., 51, 2291-5. Pizzorno, G., Moroson, B. A., Cashmore, A. R., Russello, O. Mayer, J. R., Galivan, J., Bunni, M. A., Priest, D. G., and Beardsley, G. P. (1995). Multifactorial resistance to 5,10-dideazatetrahydrofolic a c id in cell lines derived from human lymphoblastic leukemia CCRF- CEM . Cancer Res., 55, 566-73. Pizzorno, G., Mini, E., Coronneilo, M., McGuire, J. J., Moroson, B. A., Cashmore, A. R., Dreyer, R. N., Lin, J. T., Mazzei, T., Periti, P., a n d Bertino, J. R. (1988). Impaired polyglutam ylation of m e th o tr e x a te as a cause of resistance in CCRF-CEM cells after short-term, h ig h -d o se treatment with this drug. Cancer Res., 48, 2149-55. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 338 Pizzorno, G., Sokoloski, J. A., Cashmore, A, R., Moroson, B. A., Cross, A. D., and Beardsley, G. P. (1991). Polyglutam ylation of 5,10- dideazatetrahydrofolate. Mol. Pharmacol., 39, 85-89. Pizzorno, G., Russello, 0., Cashmore, A. R., Morrison, B. A., Cross, A. D., Coronnello, M., and Beardsley, G. P. (1990). Polyglutam ation: a n essential step in the activation of 5,10-dideazatetrahydrofolic ac id (D D A T H F ).Proc. Am. Assoc. Cancer Res., 31, 339. Pizzorno, G., Chang, Y.-M., McGuire, J. J., and Bertino, J. R. (1989). Inherent resistance of hum an squam ous carcinoma cell lines to methotrexate as a result of decreased polyglutam ylation of this drug. Cancer Res., 49, 5274-80. Pohland, R. C., Alati, T., Lantz, R. J., and Grindey, G. B. (1994). Whole-body autoradiographic and plasma p h arm acokinetics of 5,10-dideazatetrahydrofolic acid in mice fed folic acid-deficient o r regular diets. J. Pharm. Sci., 83, 1396-9. Prasad, P. D., R am am oorthy, S., Leibach, F. H., and G anapathy, V. (1995). Molecular cloning of the human placental folate transporter. Biochem. Biophys. Res. Commun., 2 0 6 , 681-7. Price, E. M., Ratnam, M., Rodeman, K. M., and Freisheim, J. H. (1986) Characterization of the m ethotrexate transport pathway in m u r i n e L1210 leukemia cells: involvem ent of a m em brane receptor a n d cystolic protein. B io c h e m is tr y , 27, 7853-8. Ragoussis, J., Senger, G., Trowsdale, J., and Campbell, I. G. (1992). Genomic organization of the hum an folate receptor genes o n chromosome 1 lq 13. G e n o m ic s , 14, 423-30. Ratnam, M., and Freisheim, J. M. (1990). Proteins involved in th e transport of folates and antifolates by normal and neoplastic cells. In: Folic Acid Metabolism in Health and Disease (Eds: Picciano, M. F., Stokstad, E. L. R., and Gregory, J. F.). Wiley-Liss, Inc. pp91-120. Ray, M. S., Muggia, F. M., Leichman, C. G., Grunberg, S. G., Nelson, R. L., Dyke, R. W., and Moran, R. G. (1993). Phase I study of (6R)-5,10- dideazatetrahydrofolate: a folate antim etabolite inhibitory to d e n ovo purine synthesis. J. Natl. Cancer Inst., 85, 1154-9. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 339 Reed, J. C. Mechanisms of apoptosis avoidance in cancer (1 9 9 9 ). Curr. Opin. Oncol., 11, 68-75 Reed, J. C. Regulation of apoptosis by Bcl-2 family proteins and its role in cancer and chemoresistance (1995). Curr. Opin, Oncol., 7, 541-6. Reed. J. C. (1999a). Mechanisms of apoptosis avoidance in ca n c e r. Curr. Opin. Oncol., 11, 68-75. Reed, J. C. (1999b). Dysregulation of apoptosis in cancer. J. Clin. O ncol., 17, 2941-53. Rhee, M. S., Wang, Y., Nair, M. G., and Galivan, J. (1993). Acquisition of resistance to antifolates caused by enhanced g a m m a - glutamyl hydrolase activity. Cancer Res., 53, 2227-30. Rinaldi. D. A., Kuhn, J. G., Burris, H. A., Dorr, F. A.. Rodriguez, G., Eckhardt, S. G., Jones, S., Woodworth, J. R., Baker, S., Langley, C., Mascorro, D., Abrahams, T., and Von Hoff, D. D. (1999). A phase I evaluation of m ultitargeted antifolate (MTA, LY231514), adm inistered every 21 days, utilizing the modified c o n t i n u a l reassessm ent m ethod for dose escalation. Cancer C h e m o th e r . P h a rm a c o l., 44. 372-80. Riordan, J. R., and Ling, V. (1985). Genetic and b io c h e m ic a l characterization of multidrug resistance. Pharmacol. Ther., 2 8 , 51- 75. Riordan, J. R.. Deuchars, K., Kartner, N., Alon, N.. Trent, J., and Ling, V. (1985). Amplification of P-glycoprotein genes in m u l ti d r u g - resistant mammalian cell lines. N a tu r e , 3 1 6 , 817-9. Roberts, J. D., Poplin, E. A., Tombes, M. B., Kyle, B., Spicer, D. V., Grant, S., Synoid, T., and Moran, R. (2000). Weekly lometrexol w ith daily oral folic acid is appropriate for phase II e v a lu a t io n . C ancer Chemother. Pharmacol, 45, 103-10. Rode, W. Scanlon, K. J., Hynes, J., and Bertino, J. R. (1979). Purification of m am m alian tumor (L1210) thym idylate s y n th e ta s e by affinity chromatography on stable biospecific adsorbent. J. Biol. C h em ., 2 5 4 , 11538-43. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 340 Rothberg, K. G., Yang, Y., Kolhouse, J. F., Kamen, B. A., a n d Anderson, R. G. W. (1990). The glycophospholipid-linked folate receptor internalizes folate without entering the clathrin-coated p i t endocytotic pathway. J. Cell Biol., 110, 637-49. Roy, K., Tolner, B., Chiao, J., and Sirotnak, F. M. (1998). A single amino acid difference within the folate transporter encoded by th e murine RFC-1 gene selectively alters its interaction with folate analogues. Implications for intrinsic antifolate resistance a n d directional orientation of the transporter within the p l a s m a membrane of tumor cells. J. Biol. Chem., 2 7 3 , 2526-31. Rueckert, R. R., and Mueller, G. (1960). Studies on u n b a l a n c e d growth in tissue culture. I. Induction and consequences of thymidine deficiency. Cancer Res., 20, 1584-91. Rumberger, R. G., Barrueco, J. R., and Sirotnak, F. M. (1990). Differing specificities for 4-aminofolate analogues of folylpolyglutamyl synthetase from tumors and p ro liferativ e intestinal epithelium of the mouse with significance for selective antitumor action. Cancer Res., 50, 4639-43. Saikawa, Y., Knight, C. B., Saikawa, T., Page, S. T., Chabner, B. A., Elwood, P. C. (1995). Decreased expression of the human folate receptor mediates transport-defective m ethotrexate resistance in KB cells. J. Biol. Chem., 268, 5293-301. Sambrook, J., Fritsch E. F., and Maniatis, T. (1989). M o le c u la r Cloning: A Laboratory Manual, 2 nd Ed., Cold Spring H a rb o r Laboratory Press, Cold Spring Harbor, NY. Sanger, F., Nicklen, S., and Coulson, A. R. (1977). DNA s eq u e n cin g with ch ain -term in atin g inhibitors. Proc. Natl. Acad. Sci. (USA), 74, 5463-7. Sanghani, S. and Moran, R. G. (1997). Tight binding of folate substrates and inhibitors to recom binant mouse g ly c in a m id e ribonucleotide formyltransferase. B io c h e m is tr y , 36, 10506-16. Sanghani, S. P., and Sanghani, P. C., and Moran, R. G. (1999). Identification of three key active site residues in the C -te rm in a l domain of hum an recom binant folypoly-y-glutam ate synthetase by site-directed mutagenesis. J. Biol. Chem., 2 7 4 , 27018-27. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 341 Santi, D. V., and McHenry, C. S. (1972). 5-F lu o ro -2 '-d eo x y u rid y late: covalent complex with thymidylate synthetase. Proc. Natl. A ca d . Sci. (USA). 69, 1855-7. Schimke, R. T. (1992). Gene amplification; What are we learning? Mutation Res., 2 7 6 , 145-9. Schmid, F. A., Sirotnak, F. M., Otter, G. M., and DeGraw, J. I. (1985). New folate antagonists of 10-deaza-am inopterin series: m a rk e d ly increased anti-tum or activity of the 10-ethyl analogue compared to the parent c o m p o u n d and m ethotrexate against some h u m a n tumor xenografts in nude mice. Cancer Treat. Rep., 69, 551-3. Schmitz, J. C., Grindey, G. B., Schultz, and. Priest, D. G (1994). Impact of dietary folic acid on reduced folates in mouse p l a s m a and tissues. Biochem . Pharmacol., 48, 319-25. Schober, C., Gibbs, J. F., Yin, M. B., Slocum, H. K., and Rustum, Y., M. (1994). Cellular heterogeneity in DNA damage and grow th inhibition induced by ICI D1694, thym idylate synthase inh ib ito r, using single cell assays. Biochem. Pharmacol., 48, 997-1002. Schornagel, J. H., Verweji, J., de Mulder, P. H. M., Cognetti, F., Vermorken, J. B., Cappelacre, F., Armand, J. P., Wilders, J., de Graeff, A., Clavel, M., Sahmoud, T., Kirkpatrick, A., and Lefebvre, J. L. (1995). R andom ized phase III trial of edatrexate versus methotrexate in patients with m etastatic and/or re c u rre n t squamous cell carcinom a of the head and neck: a E u ro p e a n organization for research and treatment of head and neck c a n c e r cooperative group study. J. Clin. Oncol., 13, 1649-55. Schuetz, J. D., Matherly, L. H., Westin, E. H., and Goldman, I. D. (1988). Evidence for a functional defect in the translocation of th e methotrexate transport carrier in a m ethotrexate-resistant m u r i n e L 1210 leukemia cell line. J. Biol. Chem., 2 6 3 , 9840-7. Schulz, G., and Schirmer, R. H. (1979). Principles of p ro te in structure. Springer, New York. Selhub, J (1994). Folate binding proteins: m echanism s for p la c e n ta l and intestinal uptake. In: Nutrient regulation during p re g n an cy , lactation and infant growth (Eds.: L. Allen, J. King, and B. Lonerdahl). New York, Plenum, ppl41-9. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 342 Sessa, C., de Jong, J., D'Incalci, M., Hatty, S., Pangani, O., a n d Cavalli, F. (1996). Phase I study of the antipurine a n tifo la te Lometrexol (DDATHF) with folinic acid rescue. Clin Cancer Res., 2, 1123-7. Sessa, C., Zucchetti, M., Ginier, M., Willems, Y., D'Incalci, M., a n d Cavalli, F. (1988). Phase I study of the antifolate N 10-propargyl-5,8- dideazafolic acid, CB 3717. Eu. J. Cancer & Clin. One., 24, 769-75. Shen, F., Ross, J. F., Wang, X., and Ratnam, M. (1994). I d e n tific a tio n of a novel folate receptor, a truncated receptor, and receptor t y p e beta in hem atopoietic cells: cDNA cloning, expression, im m unoreactivity, and tissue specificity. B io ch em istry, 3 3 , 1209- 15. Shen, F., Wu, M., Ross, J. F., Miller, D., and Ratnam, M. (1995). Folate receptor y is primarily a secretory protein due to lack of a n efficient signal for glycosylphosphatisylinositol modification: P ro tein characterization and cell type specificity. B io c h e m is tr y , 5 4 , 3905- 11. Shen, Y., Rudolph, J., Stern, M., and Stubbe, J. (1990). G ly c in a m id e ribonucleotide synthetase from Escherichia coli: cloning, overproduction, sequencing isolation, and c h a r a c te riz a tio n . B io c h e m is tr y , 29, 218-27. Shih, C., Grindey, G. B., Houghton, P. J., and Houghton, J. (1988). In vivo antitum or activity of 5,10-dideazatetrahydrofolic a c id (DDATHF) and its diastereomeric isomers. Proc. Am. Assoc. C ancer Res., 29. 283. Shimizu. S., Eguchi, Y„ Kamiike, W., Waguri, S., Uchiyama, Y„ Matsuda, H., and Tsujimoto, Y. (1996b). Retardation of c h e m ic a l hypoxia-induced necrotic cell death by Bcl-2 and ICE in h ib ito rs: possible involvem ent of common m ediators in apoptotic a n d necrotic signal transductions. Oncogene, 12, 2045-50. Shimizu, S., Eguchi, Y., Kamiike, W., Itoh, Y., Hasegawa, J., Y am ab e, K., Otsuki, Y., Matsuda, H., and Tsujimoto, Y. (1996a). Induction of apoptosis as well as necrosis by hypoxia and prevention of a p o p to s is by Bcl-2 and Bcl-XL. Cancer Res., 56, 2161-6. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 343 Sierra, E. E., Brigle, K. E., Spinella, M. J., and Goldm an, I. D. (1997). pH dependence of m ethotrexate transport by the reduced folate carrier and the folate receptor in L1210 leukemia cells - F u rth e r evidence for a third route mediated at low pH. B io ch e m . P h a r m a c o l .. 53, 223-31. Sikora E, Jackm an, A. L„ Newell, D. R., Calvert, A. H. (1988). Formation and retention and biological activity of N lO -propargyl- 5,8-dideazafolic acid (CB3717) polyglutam ates in L1210 cells in vitro. B iochem Pharmacol., 37, 4047-54. Sirotnak. F. M (1987). D eterm inants of resistance to a n tifo lates: Biochemical phenotypes, their frequency of occurrence a n d circumvention. Natl. Cancer Inst. Monogr. 5, 27-35. Sirotnak, F. M.. and Toner, B. (1999). C arrier-m ediated m e m b r a n e transporter of folates in mammalian cells. Anna. Rev. Nutr., 19, 91- 122 . Sirotnak, F. M., Moccio, D. M., Kelleher, L. E„ and Goutas, L. J. (1981). Relative frequency and kinetic properties of tr a n s p o r t- defective phenotypes among m ethotrexate-resistant L 1210 c lo n al cell lines derived in vivo. Cancer Res., 41, 4447-52. Sirotnak. F. M., Otter, G. M.. and Schmid, F. A. (1993). M a rk e d ly improved efficacy of edatrexate com pared to methotrexate in a high-dose regimen with leucovorin rescue against metastatic m u r i n e solid tumors. Cancer Res. 53, 587-91. Sirotnak, F. M., Kurita, S., and Hutchison, D. J (1968). On the n a t u r e of a transport alteration determining resistance to am ethopterin in the L1210 leukemia. Cancer Res., 28, 75-80. Sirotnak, F. M., Goutas, L. J., Jacobsen, D. M., Mines, L. S., Barrueco, J. R., G aum ont, Y., and Kisliuk, R. L. (1987). C a rr ie r -m e d ia te d transport of folate compounds in L 1210 cells. Biochem. P harm acol., 36, 1659-67. Sirotnak, F. M., Kurita, S., Sargent, M. G., Robinson, D. L., H u tch iso n , D. J. (1967). Sequential biochemical alteration to a n tifo la te resistance in L1210 leukemia. Nature, 2 1 6 , 1236-7. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 344 Skoog, L., Nordenskjeold, B., Humla, S., and Heagerstreom, T. (1976). Effects of m ethotrexate on deoxyribonucleotide pools and n u cleic acid synthesis in human osteosacroma cells. Eu. J. Cancer, 12, 839- 45. Smith, G. K., Duch, D. S., Dev, I. K., and Kaufm ann, S. H. (1992). Metabolic effects and kill of human T-cell leukemia by 5- deazaacy clo tetrah y d ro fo late, a specific inhibitor of g ly c in a m id e ribonucleotide transformylase. Cancer Res., 52, 4895-903. Smith, G. K., Benkovic, P. A., and Benkovic, S. J. (1981b). L(-)-10- form yltetrah y d ro fo late is the cofactor for g ly c in a m id e ribonucleotide transformylase from chicken liver. B io ch em istry, 2 0 , 4 034-6. Smith, G. K., Mueller, T., Benkovic, P. A. and Benkovic, S. J. (1 9 8 1 a ). On the cofactor specificity of glycinamide riboucleotide and 5- a m in o im id azo le-4 -carb o x im id e ribonucleotide transformylase fr o m chicken liver. B io ch e m istry , 20, 1241-5. Smith, J. M., and Daum III, H. A. (1986). Nucleotide sequence of t h e purM gene encoding 5'-phosphoribosyl-5-aminoimidazole s y n th e ta s e of Escherichia coli K-12. J. Biol. Chem., 2 6 2 , 10632-6. Smith, S. G., Lehman, N. L. and Moran, R. G. (1993). Cytotoxicity of antifolate inhibitors of thymidylate and purine synthesis to W iDr clonic carcinoma cells. Cancer Res., 53, 5697-706. Spinella, M. J., Brigle, K. E„ Sierra, E. E., and Goldman, I. D. (1995). Distinguishing between folate re c e p to r-a -m e d ia te d transport a n d reduced folate carrier-m ediated transport in L1210 leukemia cells. J. Biol. Chem., 2 7 0 , 7842-9. Spinella, M. J., Brigle, K. E., Freemantle, S. J., Sierra, E. E., a n d Goldman, E. D. (1996). Com parison of m e th o tr e x a te p o ly g lu tam y latio n in L 1210 leukemia cells when influx is m e d i a t e d by the reduced folate carrier or the folate receptor. B io c h e m . P h a rm a c o l., 52, 703-12. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 345 Synold, T. W., Newman, E. M., Carroll, M., Muggia, F. M., Groshen, S. Johnson, K., and Doroshow, J. H. (1998). Cellular but not p l a s m a pharm acokinetics of lometrexol correlate with the occurrence of cumulative hematological toxicity. Clin. Cancer Res., 4, 2349-2355. Szeto, D. W., Cheng, Y.-C., Rosowsky, A., Yu, C.-Y., Modest, E. J., Piper, J. R., Temple, Jr., C., Elliott, R. D., Rose, J. D., and Montgomery, J. A. (1977). H um an thym idylate synthetase. III. effect of m e th o tr e x a te and methotrexate analogs. Biochem. Pharm acol., 28, 2633-7. Taylor I. W., Slowiaczek, P., Francis, P. and Tattersall, M. H. N. (1982). Purine modulation of m ethotrexate cytotoxicity in mammalian cell lines. Cancer Res., 42, 5159-64. Taylor, E. C (1993). Design and synthesis of inhibitors of folate- dependent enzymes as antitumor agents. In: chemistry and biology of the pteridines and folates (Eds.: Ayling, J. E., Nair, M. G., a n d Baugh, C. M.). Plenum Press, New York, Adv. Exp. Med., 3 3 8 , 387- 408. Taylor, E. C., Harrington, P. J., Pletcher, S. R., Beardsley, G. P. (1985). Synthesis of the antileukemic agents, 5,1 0 -d id eazaam in o p terin a n d 5,6 ,7,8 dideazatetrahydroam inopterin. J. Med. Chem., 28, 914-21. Taylor, I. W., Slowiaczek, P., Francis, P. R., and Tattersall, M. H. N. (1982). Biochemical and cell cycle perturbations in m e th o tre x a te - treated cells. Mol. Pharmacol., 21, 204-10. Taylor, R. T.. and Hanna, M. L. (1977). Folate-dependent e n z y m e s in cultured Chinese hamster cells: folypolyglutam ate s y n th e ta s e and its absence in mutant auxotrophic for glycine + adenosine + thymidine. Arch. Biochem. Biophys, 181, 331-4. Thornberry, N. A., and Lazebnik, Y. (1998). Caspases: Enem ies within. Science, 281, 1312-6. Trent, D. F., Seither, R. L., and G oldm an, I. D. (1991). C o m p artm en tatio n of intracellular folates. Biochem. P h arm ac o l., 42, 1015-9. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 346 Trippett, T., Schlemmer, S., Elisseyeff, Y., Goker, E., Wachter, M., Steinherz, P., Tan, C., Berman, E., Wright, J. E., and Rosowsky, A. (1992). Defective transport as a m echanism of acquired re sistan c e to m ethotrexate in patients with acute lym phocytic le u k e m ia . Blood, 80, 1158-62. Tse, A., and Moran, R. G (1994). Control of polyglutamation of 5,10- d id eaz atetra h y d ro fo late by endogenous folate pools: a no v el m echanism of resistance to antifolates. Proc. Am. Assoc. C a n c e r Res., 35, 304. Tse, A., and Moran, R. G (1998). Cellular folates p r e v e n t p olyg lu tam atio n of 5 , 10-dideazatetrahydrofolate. A n o v el m echanism of resistance to folate antim etabolites. J Biol C h e m ., 273, 25944-52. Tse, A., Brigle, K. E., and Moran, R. G (1997). Mutations in t h e reduced folate carrier which confer d o m inant resistance to 5 ,1 0- dideazatetrahydrofolate by allowing efficient transport of folic a c id . Proc. Am. Assoc. Cancer Res., 38, 162. Tse, A., Brigle, K. E., Taylor, S. M., and Moran, R. G (1 9 9 8 ). Mutations in the reduced folate carrier which confer d o m i n a n t resistance to 5 , 10-dideazatetrahydrofolate. J Biol Chem., 273, 25953-60. Tsujimoto, Y., Cossman, J., Jaffe, E., and Croce, C. M. (1 9 8 5 ) Involvem ent of the bcl-2 gene in hum an follicular l y m p h o m a . Science, 228, 1440-3. Tsujimoto, Y., Shimizu, S., Eguchi, Y., Kamiike, W., Matsuda, H. (1997). Bcl-2 and Bcl-XL block apoptosis as well as necrosis: involvement of common mediators in apoptotic and necrotic signal transduction pathways. Leukemia., ll(S u p p l. 3), 380-2. Turner, F. B., Andreassi II, J. L., Ferguson, J., Titus, S., Tse, A., T aylor, S. M., and Moran, R. G. (1999). Tissue-specific expression of functional isoforms of mouse folylpoly-_-glutam ate synthetase: a basis for targeting folate antimetabolites. Cancer Res., 59, 6074-9. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 347 van der Laan, B. F„ Jansen, G., Kathm ann, G. A., Westerhof, G. R., Schornagel, J. H., and Hordijk, G. J. (1992). In vitro activity of novel antifolates against human squam ous carcinom a cell lines of the head and neck with inherent resistance to methotrexate. Int J Cancer, 51, 909-14. Wang, X., Shen, F., Freisheim, J. H., Gentry, L. E., and Ratnam, M. (1992). Differential stereospecificities and affinities of folate receptor isoforms for folate com pounds and antifolates. B io ch em . Pharmacol., 44, 1898-901. Ward, W. H., Kimbell, R., and Jackm an, A. L. (1992). Kinetic characteristics of ICI D1694: a quinazoline antifolate which in h ib its thymidylate synthase. Biochem. Pharmacol. 43, 2029-31. Wedge, S. R., Laohavinji, S., G. Taylor, A., Boddy, A., Calvert, A. H. and Newell D. R. (1995). Clinical pharm acokinetics of t h e antipurine antifolate (6R )-5,10-dideazatetrahydrofolate (lo m etre x o l) administered with an oral folic acid supplement. Clin. Cancer. Res., 1, 1479-86. Weitman S. D., Lark, R. H., Coney, L. R., Fort, D. W„ Frasca, V., Zuravvski, Jr., V. R. and, Kamen, B. A. (1992). Distribution of t h e folate receptor GP38 in normal and malignant cell lines and tissues. Cancer Res., 52, 3396-401. Werkheiser, W. C (1961). Specific binding to 4-aminoflic a c id analogues by folic acid reductase. J. Biol. Chem., 236, 888-93. White, J. C., Loftfield, S., and Goldm an, I. D. (1975). T h e m echanism of action of methotrexate. III. Requirement of free intracellular methotrexate for maximal suppression of [1 4 C ]fo rm ate incorporation into nucleic acids and protein. Mol. Pharmacol, 11, 287-97. Whitehead, V. M. (1977). Synthesis of m ethotrexate p o l y g l u ta m a t e s in L I 210 murine leukemia cells. Cancer Res., 37, 408-12. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 348 Whitehead, V. M., Rosenblatt, D. S., Vuchich, M. J., Shuster, J., Witte, A., and Beaulieu, D. (1992). A ccum ulation of m ethotrexate a n d m ethotrexate polyglutamates in lym phoblasts diagnosis of c h ild r e n acute lym phoblastic leukemia pilot prognostic factor an a ly sis. Blood, 76, 44-9. Whitehead, V. M., Kalman, T. I., Rosenblatt, D. S., Vuchich, M J., a n d Payment, C. (1988). Regulation of methotrexate p o l y g l u t a m a t e (MTXPG) formation in human leukemic cells. Proc. Am. Assoc. Cancer Res., 29, 287. Williams, F. M. R., and Flintoff, W. F. (1995). Isolation of a h u m a n cDNA that com plem ents a m utant hamster cell defective i n methotrexate uptake. J. Biol. Chem., 2 7 0 , 2987-92. Williams, F. M., Murray, R. C., Underhill, T. M., and Flintoff, W. F. (1994). Isolation of a hamster cDNA clone coding for a f u n c tio n involved in methotrexate uptake. J. Biol. Chem., 269, 5810-6. Wong, S. C., Zhang, L., Proefke, S. A., and Matherly, L. H. (1998). Effects of the loss of capacity for N-glycosylation on the transport activity and cellular localization of the human reduced folate carrier. Biochim Biophys Acta., 1 3 7 5 , 6-12. Wong. S. C., Proefke, S. A., Bhushan, A., and Matherly, L. H. (1995). Isolation of human cDNAs that restore methotrexate sensitivity a n d reduced folate carrier activity in m ethotrexate tra n s p o rt-d e fe c tiv e Chinese hamster ovary cells. J. Biol. Chem., 270, 17468-75. Wong, S. C., Proefke, S. A., Bhushen, A., and Matherly, L. H. (1997). Mutations of the reduced folate carrier in transport impaired CCRF- CEM cells. Proc. Am. Assoc. Cancer. Res., 38, 162. Wyllie, A. H., Kerr, J. K., and Currie, A. R. (1980). Cell death: th e significance of apoptosis. Int. Rev. Cytol., 6 8 , 251-306. Yang CH, Sirotnak, F. M., Dembo, M. (1984). Interaction b etw e en anions and the reduced folate/m ethotrexate transport system in L1210 cell plasma m embrane vesicles: directional sym m etry a n d anion specificity for differential mobility of loaded and u n l o a d e d carrier. J Membr Biol., 79, 285-92. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 349 Yang, C.-H., Sirotnak, F. M., and Mines, L. S. (1988). Further studies on a novel class of genetic variants of the L1210 cell with in cre ased folate analogue transport inward. J. Biol. Chem., 2 6 3 , 9703-9. Yang, C.-H., Dembo, M., Sirotnak, F. M. (1983). Relationship betw een carrier-m ediated transport of folate com pound by L1210 le u k e m ia cells. Evidence for multiplicity entry routes with different kinetic properties expressed plasma membrane vesicles. J. Membr. Biol., 25, 11-28. Young, C. W., Currie, V. E., Muindi, J. F., Saltz, L. B., Pisters, K. M. W., Esposito, A. J., and Dyke, R. W. (1992). Phase I study of lom etrexol (LY264618 or LTX)-improved clinical tolerance with folic acid. Proceedings of the 7th NCI-EORTC Symposium on New Drugs in Cancer Therapy, p 136. Young, C., Currie, V., Baltzer, L., Trochanowski, B., Eton, 0., Dyke, R., and Bowsher, R. (1990). Phase I and clinical pharmacologic study of LY264618, 5 , 1 0 -d id e a z a te tra -h y d ro fo la te . Proc. Am. Assoc. Cancer Res., 31, 117. Young, M., Sammons, R. D., Mueller, W. T., and Benkovic, S. J. (1984). An antibody probe to determine the native species of glycinamide ribonucleotide transformylase in chicken liver. B io c h e m is tr y , 23, 3979-86. Zhao, R., Sharina, I., and Goldman, I. D. (1999). Pattern of mutations that results in loss of reduced folate carrier fu n c tio n under antifolate selective pressure augm ented by ch e m ic a l mutagenesis. Mol. Pharmacol., 56, 68-76. Zhao, R., Seither, R., Brigle, K. E., Sharina, I. G., Wang, P. J., a n d Goldman, I. D. (1997). Impact of overexpression of the re d u c e d folate carrier (RFCl), an anion exchanger, on concentrative t r a n s p o r t in murine L 1210 leukemic cells. J. Biol. Chem., 2 7 2 , 21207-12. Zhao, R., Titus, S., Gao, F., Moran R. G., and G oldm an, I. D. (2000). Molecular analysis of murine leukemia cell lines resistant to 5, 10- dideazatetrahydrofolate identifies several amino acids critical to th e function of folylpolyglutamate. Clin. Cancer Res. 6, 3687-95. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 350 Zhao, R., Assaraf, Y., and Goldm an, I. D. (1998a). A reduced folate carrier mutation produces substrate-dependent alterations in c a rrier mobility in murine leukemia cells and methotrexate resistance with conservation of growth in 5-form yltetrahydrofolate. J. Biol. C h em ., 2 7 3 , 7873-9. Zhao, R., Assaraf, Y., and Goldm an, I. D. (1998b). A muta ted murine reduced folate carrier (RFC1) with increased affinity for folic acid, decreased affinity for m ethotrexate, and an obligatory anion requirement for transport function. J. Biol. Chem., 2 7 3 , 19065-71. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 351 APPENDIX A ABBREVIATION USED (G lu )n , folyl polyglutamate derivatives 10-CHO-THF, 10-formyl tetrahydrofolate 5- CH 3 -TH F, 5-methyltetrahydrofolate 5 , 10-C H 2 -THF^ 5 , 1 0 -m ethylenetetrahydrofolate 5-CH O-TH F, 5-formyltetrahydrofolate ADP, adenosine diphosphate AICARFT, am inoimidazole carboxamide formyltransferase AIRS, am inoim idazole ribonucleotide synthetase ALL, acute lymphoblastic leukemia AML, acute myelogenous leukemia AMP, adenosine monophosphate ATP, adenosine triphosphate bp, base pair BSA, bovine serum albumin cDNA, complementary DNA CHO, Chinese hamster ovary CMP, cytosine m onophosphate dADP, 2 ’-deoxyadenosine 5 ’-diphosphate dATP, 2 ’-deoxyadenosine 5 ’-triphosphate dCTP, 2 ’-deoxycytidine 5 ’-triphosphate D D A TH F, 5,10-didezatetrahydrofolate DEAE, diethylam inoethyl dFCS, dialyzed fetal calf serum dGTP, 2 ’-deoxyguanidine 5 ’-triphosphate DHFR, dihydrofolate reductase DNA, deoxyribose nucleic acid dTTP, 2 ’-deoxythym idine 5 ’-m onophosphate dUMP, 2 ’-deoxyuridine 5 ’-monophosphate EDTA, ethylenediaminetetraacetic acid FCS, fetal calf serum FdUM P, 5 -flu o ro -2 ’-deoxyuridine 5 ’-m onophosphate FPGS, folylpoly-y-glutamate synthetase FR, folate receptor FUdR, 5-fluorodeoxyuridine GAR, a,b-glycinam ide ribonucleotide GARFT, glycinam ide ribonucleotide formyltransferase GARS, glycinamise ribonucleotide synthetase GDP, guanidine diphosphate Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 352 G M P, guanidine m onophosphate GTP, guanidine triphosphate h, hour hplc, high performance liquid chrom atography Kd, kilo Dalton K m, Michaelis-Menten constant mFBP, membrane binding protein min, minute M r, molecular weight M TA, multi-targeted antifolate M TX, methotrexate, nt, nuleotide PBS, phosphate buffer saline PCR, polymerase chain reaction PRPP, phosphoribosyl pyrophosphate. RFC, reduced folate carrier RNA, ribose nucleic acid rpm, revolution per minute RT, room temperature s, second SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis TAE, Tris acetate-EDTA TBE, Tris borate-EDTA TE, Tris-EDTA THF, tetrahydrofolate TM Q , trimetrexate TS, thymidylate syntase UTR, untranslated region V max, maximal velocity x g, times gravity P-ME, (3-mercaptoethnaol Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Asset Metadata

Creator Tse, Archie Ngai-Chio (author)

Core Title Biochemical factors determining tumor response to 5,10-dideazatetrahydrofolate, a folate antimetabolite inhibitory to de novo purine biosynthesis

Contributor Digitized by ProQuest (provenance)

School Graduate School

Degree Doctor of Philosophy

Degree Program Biochemistry and Molecular Biology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag biology, molecular,chemistry, biochemistry,health sciences, oncology,OAI-PMH Harvest

Language English

Advisor Stellwagen, Robert H. (committee chair), Farley, Robert (committee member), Jones, Peter A. (committee member), Roy-Burman, Pradip (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-234784

Unique identifier UC11339024

Identifier 3073860.pdf (filename),usctheses-c16-234784 (legacy record id)

Legacy Identifier 3073860.pdf

Dmrecord 234784

Document Type Dissertation

Rights Tse, Archie Ngai-Chio

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA