Development of HA -pseudotyped retroviral vectors for cell -specific gene delivery

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, some 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 of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI 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. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. ProQuest Information and Learning 300 North Zeeb Road, Ann Arbor, Ml 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. DEVELOPMENT OF HA-PSEUDOTYPED RETROVIRAL VECTORS FOR CELL-SPECIFIC GENE DELIVERY by Amy Hsiu-Ti Lin 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) August 2000 © 2000 Amy Hsiu-Ti Lin Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3018100 _ _ _ ® UMI UMI Microform 3018100 Copyright 2001 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. Bell & Howell 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 GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES. CALIFORNIA 90007 This dissertation, written by Ifeiu Ti Lin under the direction of tef.. Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School in partial fulfillment of re quirements for the degree of DOCTOR OF PHILOSOPHY Dean of Graduate Studies Date ....tfa y ..2 2 ji...2 .Q G Q . DISSERTATION COMMITTEE Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Development of HA-pseudotyped MuLV for gene delivery MuLV-based retroviral vectors are currently the most utilized delivery system for gene therapy via ex vivo procedures. In the past decade, numerous researchers have attempted to fully exploit the use of these vectors by restricting tropism and improving transduction efficiency, so that the vectors could be delivered into humans using in vivo protocols. However, the problems are considerable. Although specific-receptor targeting has been reported with chimeric MuLV envelope proteins, in general, target cells were transduced at an efficiency that was 3-5 logs lower than that obtained with unmodified viral vectors. The problem with such chimeric proteins was found to be at a post binding step, specifically, the inability of chimeric MuLV envelope proteins to trigger membrane fusion. Here, I devised two novel approaches to overcome the lack of fusion in previous targeting attempts. My strategies utilized a non-specific fusion protein, influenza HA, in MuLV-based vectors, in order to enable membrane fusion. One approach was to directly engineer a binding-defective HA protein (HAtmt) itself to achieve targeting. Another approach is to co-express HAtmt protein with a fusion-defective chimeric MuLV envelope protein. In both cases, the targeting receptors chosen are known to be endocytosed after ligand binding. My studies have demonstrated that a chimeric HAtmt protein (apoE- HAtmt) containing the apoE binding moiety at the N-terminus of the HAtmt was able to bind to its targeted receptor, LDL receptor, and to retain its membrane Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fusion function. However, MuLV vectors containing apoE-HAtmt were unable to efficiently transduce target cells. In contrast, the combination of HAtmt as a fusogenic protein, together with a chimeric MuLV envelope protein directed to the Flt-3 receptor, was able to trigger membrane fusion and enhance the transduction efficiency of Flt-3 expressing cells by 10 fold. The positive outcome of this approach suggests that this targeting strategy could be expanded by using different binding proteins in addition to chimeric MuLV envelope proteins. These data constitutes a proof of principle that to develop a targeted gene therapy vector is feasible. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Dedication I dedicate this thesis to my parents, who have always be supporting and loving; and to God, who gives strength to help me accomplish my goal. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Acknowledgement There are many peoples that I am very grateful to because of their supports during my graduate studies. It was them that made this work less painful. I would like to thank both Drs. W. French Anderson, Paula M. Cannon and Noriyuki Kasahara for their guidance throughout my graduate studies. I would like to especially thank Dr. Paula Cannon for training me to gain the skills a scientist needs. I would like to thank Drs, Dr. Mark Ginsberg Hans-Dieter Klenk, and Jonathan Yewdell who kindly provided the reagents needed for this study. I would also like to thank the past members (Drs. Cyril Empig, Jin-Young Han, Jung-Joo Hwang, Sunyoung Lee, Ling Li and Richard Tun) and the present members (all the graduate students, Stella Diaz, Kathy Burke and Pia Mabre) of Gene Therapy Laboratory that I had the opportunity to interact with them scientifically as well as do fun things with them. Thank you. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents Dedication ii Acknowledgements iii List of Tables vi List of Figures vii Chapter 1 Introduction 1 1.1 Retroviral entry, tropism and envelope function 2 1.2 Targeted retroviral vectors: previous attempts and problems 8 1.3 Retroviral vectors 14 1.4 Experimental design and preliminary data 16 Chapter 2 Characterization of the receptor binding pocket of influenza hemagglutinin 22 2.1 Abstract 23 2.2 Introduction 24 2.3 Material and Methods 27 2.4 Results 31 2.5 Discussion 45 Chapter 3 Conversion of H1 subtype hemagglutinin to endogenously cleaved protein does not result in functional HA: separation of cleavability and pathogenicity 50 3.1 Abstract 51 3.2 Introduction 53 3.3 Material and Methods 61 3.4 Results 67 3.5 Discussion 80 Chapter 4 Targeted retroviral vectors pseudotyped by hemagglutinin glycoprotein 84 4A Co-expression of binding-defective HA with fusion-defective chimeric MuLV envelope proteins to separate binding and fusion functions on two different molecules 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. V 4A.1. Abstract 86 4A.2. Introduction 88 4A.3. Material and Methods 95 4A.4. Results 101 4A.5. Discussion 115 4B. Modification of the HA protein for targeting 119 4B.1. Abstract 120 4B.2. Introduction 122 4B.3. Material and Methods 125 Part 1: Insertion of small ligand at the internal sites in the HA protein 140 4B.4. Results 140 4B.5. Discussion 148 Part 2: Insertion of a large ligand at the N-terminus of Hi subunit of the HA protein 151 4B.6. Results 151 4N.7. Discussion 159 Chapter 5 What we have learned from designing a targeted retroviral vector using the HA protein 161 5.1. Overall conclusions 162 References 167 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. vi Table 1.1. Table 1.2. Table 2.1. Table 2.2. Table 2.3. Table 3.1. Table 3.2. Table 3.3. Table 3.4. Table 3.5. Table 4A.1. Table 4B.1. Table 4B.2. Table 4B.3. List o f Tables Titer of MuLV particles pseudotyped by different viral fusion proteins. HA pseudotyped MuLV vectors enter cells through an endocytic pathway. Summary of receptor binding pocket mutants of FPVHA Fusion function of mutant FPV HAs at pH 5. H183F/L194A and Y98F/H183F/L194A are temperature-sensitive mutants. Cleavage sites of influenza virus HA. Cleavage site of influenza HA variants. Cell surface expression of HA variants. Infectivity of HA-pseudotyped MuLV vectors in 239Tand QT6 cells. Summary of binding and fusion properties of HA variants. HAtmt retains its fusion function when co expressed with chimeric MuLV Env. Characterization of RGD-containing chimeric HA proteins. Fusion property of chimeric apoE-HAtmt protein. Transduction of apoE-HAtmt-pseudotyped MuLV particles. 21 21 33 39 44 55 71 72 76 79 107 144 158 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Figures Figure 1.1. Figure 1.2. Figure 1.3. Figure 1.4. Figure 1.5. Figure 2.1. Figure 2.2. Figure 2.3. Figure 2.4. Figure 2.5. Figure 3.1. Figure 3.2. Figure 3.3. Envelope protein of Moloney murine leukemia virus virus (MoMuLV). Fusion of viral and cellular membranes. Possible barriers to prevent efficient transduction by modified MoMuLV retroviral vector displaying a new binding moiety directed to a specific receptor. Three-plasmid replication-defective retroviral vector system. HIV- and MuLV-based vectors can be pseudotyped by HA protein. Crystal structure of the receptor binding pocket site of a monomeric HA molecule of H3 subtype. Effects of HA protein mutations on virion incorporation. Polykaryon formation by HA-expressing NIH 3T3 cell (hyper-fusogenic HA mutants) Polykaryon formation by HA-expressing NIH 3T3 cells (hypo-fusogenic HA mutants). Incorporation of hypo-fusogenic HA mutants into MuLV particles. Crystal structure of the monomeric uncleaved HA molecule (H3 subtype). Immunoblot of cell lysates of HA variants. Vector incorporation and susceptibility to exogenous trypsin treatment of HA variants. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.4. Figure 4A.1. Figure 4A.2. Figure 4A.3. Figure 4A.4. Figure 4A.5. Figure 4A.6. Figure 4A.7. Figure 4A.8. Figure 4A.9. Possible reasons for defects in H1(7b) variant. Co-express a binding-defective HA with a fusion-defective chimeric MuLV envelope protein. Schematic diagram of chimeric MuLV envelope proteins. Co-expression of chimeric MuLV envelope protein and HAtmt protein. Receptor binding activity of MuLV viral particles containing a chimeric MuLV envelope protein and the HAtmt protein (Co-expressions of PC7Env + HAtmt and D84KE(31 + HAtmt). Receptor binding activity of MuLV viral particles containing a chimeric MuLV envelope protein and the HAtmt protein (Co-expressions of FIEnvBB + HAtmt and FIEnvSN + HAtmt). Transduction of target cells by MuLV viral particles containing a chimeric MuLV envelope protein and the HAtmt protein (Co-expressions of PC7Env + HAtmt and D84KEP1 + HAtmt). Transduction of 293A/Flt3 cells by MuLV viral particles containing chimeric MuLV envelope proteins and the HAtmt protein targeted to Flt-3 receptor-expressing cells. Ligand competition by soluble Flt-3 ligand (FL) to prevent enhancement of transduction efficiency on 293A/Flt3 cells. Co-expression FIEnvSN chimeric MuLV envelope protein and a fusion-defective HA protein (HAfd) prevents enhancement of transduction efficiency on 293A/Flt3 cells. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4B.1. Figure 4B.2a. Figure 4B.2b. Figure 4B.2c. Figure 4B.2d. Figure 4B.3. Figure 4B.4. Figure 4B.5. Figure 4B. 6. Figure 4B.7. Figure 4B.8. Figure 4B.9. Figure 5.1. Amino acid sequence alignment of the HAi subunit of HA proteins between the H3 and H7 subtypes. Cloing scheme for RGD peptide insertion. Snthesis of RGD peptide ligand by PCR. Cloning of HAapoE-expression plasmids. Potential insertion sites on the surface of the HA molecule. Insertion of a large ligand at the N-terminus of the HA! subunit of the HA protein. Insertion of apoE peptide ligand. Insertion of RGD peptide into the HA molecule and virion incorporation. MuLV viral particles pseudotyped with RGD- containing chimeric protein do not bind to a»bP 3 integrin receptors on CHO/a»b p3 cells. Immunoblots of chimeric HA protein incorporated into MuLV viral particles. Cell surface expression of apoE-HAmt chimeric protein. apoE-HAtmt binds specifically to the LDLR receptor. Possible barriers to efficient viral entry of MuLV particles that possess both receptor-binding and fusion functions. 124 127 129 130 131 135 141 143 147 152 154 156 166 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 Chapter 1 Introduction Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 1.1. Retroviral entry, tropism and envelope protein function Retroviral entry requires two processes: binding to a specific receptor on the cell surface and fusion between the viral and cellular membrane. Both processes are performed by the envelope protein, present on the surface of the retrovirus. Following fusion between the viral and cellular membranes, the viral core uncoats in the cytoplasm and the genome of the retrovirus is transported into the nucleus, where it is integrated into the host chromosome. Receptor binding is the initial step in viral entry. In the murine leukemia viruses (MuLVs) there are at least five groups, classified according to the receptors that the viruses recognize among species and according to the pattern of viral interference (Weiss 1993, review). The receptor (MCAT-1) for the ecotropic MuLVs is only present on murine and a few rat cells. In contrast, the xenotropic MuLVs recognize receptors present on cell types of all species except mouse cells. Amphotropic and polytropic MuLVs recognize receptors present in many cell types of all species including mice, but they do not interfere with each other, indicating that the receptors for amphotopic (Pit- 2) and polytropic MuLVs are two distinct molecules. 10A1 MuLVs display dual tropism characteristics; they interfere with amphotropic MuLVs indicating that one of the receptors which 10A1 MuLVs utilize is the same receptor molecule that amphotropic MuLVs use (Han et al., 1997, Wilson et al., 1994, Wilson et al„ 1995). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In spite of extensive research efforts, the mechanism by which the retroviral envelope protein induces fusion remains unclear. Current models are based on the influenza virus fusion protein hemagglutinin (HA). X-ray crystallography and biophysical studies have demonstrated that the fusion process mediated by the HA protein requires a gross conformational change in the protein (Carr and Kim 1993). After binding to its receptor on the cell surface, influenza virus is internalized into an endosome. The low pH environment of the mature endosome causes the conformational changes in the HA protein that transformed from a fusion inactive into a fusion active state (Ellens et al., 1990; Lamb 1993; Stegmann et al., 1990; and White 1994). These changes result in the extrusion of the fusion peptide located at the N-terminus of the HA2 subunit, and it is believed that the HA1 subunit is pushed away to the side to allow the viral and endosomal membranes to come into close proximity. A decrease in pH is therefore critical for the HA protein to carry out its fusion function, and influenza entry is described as a pH-dependent process. Viruses that gain entry into host cells by pH-dependent routes are sensitive to lysosomotropic agents that prevent the acidification of endocytic compartments (March and Pelchen-Mathews, 1994; and White et al., 1983). Rhabdoviruses, e.g. vesicular stomatitis virus (VSV), togaviruses, e.g. semliki forest virus, as well as ecotropic Moloney murine leukemia virus (MoMuLV) are all believed to enter cells via a pH-dependent pathway (Marsh and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 Pelchen-Matthews, 1994, Ragheb and Anderson, 1994b). However, it is important to note that the evidence supporting the conclusion that ecotropic MoMuLV enters host cells via a pH-dependent pathway remains equivocal (Katen 1998 thesis; McClure et al., 1990; Nussbaum et al., 1993; and Ragheb etal., 1994b). In contrast, viruses that are not sensitive to lysosomotropic agents are thought to enter host cells in a pH-independent pathway in which the fusion between the viral and cellular membranes occurs at the cell surface immediately following receptor binding. Retroviruses such as Rous sarcoma virus, human immunodeficiency virus (HIV) and amphotropic murine leukemia virus (A-MuLV) are believed to fuse directly at the surface membrane (Gilbert et al., 1990 and 1995; McClure et al., 1988; McClure et al., 1990; Ragheb and Anderson, 1995; Stein et al., 1987). In contrast to viruses that enter host cells via a pH-dependent pathway in which a low pH environment is critical to trigger the conformational change necessary for membrane fusion, pH- independent viruses probably rely on the interaction with the receptor alone to trigger the conformational change necessary for fusion. The envelope protein (Env) of MoMuLV consists of two subunits, gp70 (SU) which is a surface glycoprotein, and p15E (TM) which is a transmembrane protein (Figure 1.1.). Both proteins are derived from an 85- kDa precursor glycoprotein (Pr85) which is cleaved by a cellular protease Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 (Pinter and Honnen, 1983) during its transport to the surface of the cell membrane, where it is incorporated into budding viral cores. The mature SU and TM proteins remain associated with each other in the virions and, together, they mediate viral attachment and entry into target cells. Mutational analyses have mapped functional domains in both gp70 and p15E. The N- terminus of gp70 is involved in receptor binding (Battini et al., 1995 and 1996; and McKrell et al., 1996) and several regions of p15E have been shown to be involved in the fusion process (Ragheb et al., 1994a; Januszeski et al., 1997; Zhao et al., 1998; and Zhu et al., 1998). The current model of how MuLV envelope protein binding to its receptor results in fusion (Figure 1.2.) is based on the paradigm established for the HA protein of influenza virus described above. The SU and TM subunits of MoMuLV envelope protein are analogous to the HAi and HA2 subunits of the HA protein. There have been a number of studies of MoMuLV, HIV and other retroviruses that have suggested that a conformational change occurs in the envelope proteins following interaction with a receptor. For HIV, this conformational change results in the exposure of the “ buried" fusion peptide in the N-terminal region of TM and promotes fusion of viral and host membranes (Binley et al., 1997; Chan et al., 1997; Furuta, et al., 1998; Sattentau et al., 1991, 1993, and 1995; Stein et al., 1987). Recently, a similar observation was made for MoMuLV in a study by Ikeda et al., (2000). Here, a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 conformational change was detected in envelope protein following binding to its receptor, as detected by differential antibody recognition. Fusion peptide Figure 1.1. Envelope protein of Moloney murine leukemia virus (MoMuLV). MoMuLV envelope protein consists of two subunits, SU and TM. The N-terminus domain of SU is involved in receptor binding and possibly transmitting this signal to the TM subunit. The C-terminus domain of SU is highly conserved among MuLVs, but the functionalsignificance of this domain is unknown. The TM subunit contains anectodomain and a cytoplasmic domain. A stretch of hydrophobic residues (fusion peptide) located at the N-terminus of TM subunit is important for the fusion process. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7 Receptor binding Fusion opr Figure 1.2. Fusion of viral and cellular membranes. Based on the influenza HA model, binding of the MuLV envelope protein to its receptor is thought to cause an overall conformational change in the envelope protein which subsequently leads to reorientation of the fusion peptide and its insertion into the target cell membrane. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 1.2. Targeted MuLV retroviral vectors: previous attempts and problems The ability of retroviruss to integrate their genome into a host chromosome has attracted researchers to exploit these viruses as a vehicle for gene delivery. To date, the most commonly used retroviral vectors for gene therapy protocols are MuLV retroviral vectors that contain the following characteristics: (1) they integrate stably into the host genome; (2) they can be produced safely, to minimize recombination and the regeneration of replication competent virus; (3) they produce relatively high titer, but minimal immune response; and (4) they are stable to concentration. However, they also possess the following characteristics that preclude them from becoming the ideal vector for gene therapy: (1) the vector genome has a limited capacity for insertion of a trans gene; (2) the expression of the trans gene is not always appropriately regulated in targeted cells; (3) they cannot transduce non-dividing cells efficiently; (4) most vectors in use display a wide tropism which renders them unable to target a specific cell type; and (5) integration occurs at random sites and hence may result in insertional mutagenesis. Attempts to regulate the expression of the trans gene incorporated into a vector genome have been made. Tissue-specific promoters have been utilized to allow specific expression of the trans gene in target cells (Hwang et al., 1997; Waltherand Stein, 1996). Another approach to achieve cell-type specific gene delivery, which has received considerable attention in the past several years, is to regulate Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 viral entry at the receptor binding step. Many groups, including my laboratory, have attempted to redirect the tropism of MuLV retroviral vectors by engineering either the envelope protein of MoMuLV to broaden its tropism to human cells or the envelope protein of amphotropic MuLV to restrict its tropism to specific human cells (reviewed in Anderson 1998; and Cosset and Russell 1996). The ability to redirect the MoMuLV envelope protein to specific human cell types would allow the in vivo use of retroviral vectors and avoid the tedious ex vivo transduction procedures which many of current gene therapy protocols require. One of the earliest attempts at targeting retroviral vectors was to direct MoMuLV envelope protein to EGF receptor by antibody bridging (Goud et al., in 1988). Although internalization of the virus-receptor complex was observed, the vectors were unable to transduce the target cells. Later, Roux et al., (1989) and Etienne-Julan et al., (1992) also used a similar strategy to redirect the MoMuLV envelope protein to the histocompatibility complex (MHC) class I and class II antigens, EGF receptor, transferrin receptor, LDL receptor, and galactose receptor. In these studies, gene transfer was observed only for targeting of the EGF receptor and the MHC class I and class II antigens, and the transduction efficiency was very low (< 103 cfu/ml). These early experiments provided insights into the feasibility of targeting specific cell surface molecules, and also suggested that the choice of receptor could play a critical role in the success of a targeting strategy. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 0 Subsequently, other strategies that were attempted used modification of the MuLV envelope protein itself. In one common approach, a new binding moiety was incorporated into the envelope protein, keeping the native ’ envelope protein intact. MuLV envelope proteins that have been modified by this approach are referred to as "insertion" envelope proteins. The problems encountered using this strategy include steric hindrance of the inserted binding moiety by the surrounding host protein residues, or inefficient processing or virion incorporation of the modified envelope protein (Battini et al., unpublished; Wu et al., 2000). An improvement of this strategy was to insert the binding moiety at the N-terminus of the MoMuLV SU, which allowed the independent folding from the remaining envelope protein. Binding moieties that have been incorporated at the N-terminus of MoMuLV envelope protein include the receptor binding domain of the amphotropic MuLV envelope protein (Cosset and Russell, 1995), and a variety of a single chain antibodies (scFv) directed to specific receptors (Benedict et al., 1999; Marin et al., 1996; Martin et al., 1998; Russell et al., 1993; Scheierle et al., 1996; Somia et al., 1995; Yajima et al, 1998; and Zhao et al., 1999). All these chimeric MoMuLV envelope proteins demonstrated specific binding to their cognate receptors. However, only targeting of the EGF receptor, (Cosset and Russell 1995), MHC class I antigen (Marin et al., 1996), and LDL receptor (Somia et al., 1995) were found able to confer titer for target cells, and frequently at only low efficiency. Furthermore, these initial reports have often Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 1 been difficult to confirm in subsequent studies. In particular, the high titer (10s cfu/ml) reported by Somia et al., for the LDL receptor targeted MoMuLV envelope protein is contested (Empig et al., unpublished; Cosset et al., personal communication). In another approach, replacement of the entire receptor-binding domain of the MoMuLV SU with a new binding moiety has also been reported (Han et al., 1995; Kasahara et al., 1994; and Schnierle et al, 1996). MuLV envelope proteins that have been modified by this approach are referred to as "replacement" envelope proteins. One limitation of this strategy is that the expression of those modified envelope proteins is usually low and requires co-expression of wild type envelope protein in order to rescue expression and virion incorporation. Although several reports have demonstrated successful cell targeting, the interpretation of some of these experiments also remains controversial due to the difficulty many laboratories have experienced in trying to repeat these data (Kabat 1995). In general, the studies that have addressed the feasibility of specific receptor targeting have demonstrated that the attachment to a targeted receptor by chimeric envelope proteins can be achieved. However, these modified envelope proteins are not able to efficiently transduce the targeted cells. Barriers to efficient transduction could be either due to the chimeric envelope protein, the targeted receptor, or both, which will all lead to a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12 blockage at a post-binding step. One possible explanation is that the binding of a chimeric envelope protein to its targeted receptor does not trigger the necessary conformational change in the chimeric envelope protein that leads to fusion between the viral and cellular membranes (Benedict et al., 1999; and Zhao et al., 1999). In addition, the bound virus may be internalized with the targeted receptor and sequestered into a pathway that is not conducive to viral infection (Cosset et al., 1995). In both cases, the results would be inefficient gene delivery (Figure 1.3.). An alternate strategy employed in improving the generation of targeted retroviral MuLV envelope proteins is by a two-step "inverse targeting" approach, in which a new binding moiety is added on to an amphotropic MuLV envelope protein, and the two binding domains are separated by a protease activatable linker. In this approach, the new binding moiety competes with the amphotropic MuLV envelope protein for binding to their receptors, and only after cleavage of the activatable linker by a specific protease releases the new binding moiety is the remaining intact amphotropic MuLV envelope protein able to infect cells through the natural envelope- receptor interaction. Although the feasibility of this approach was demonstrated (Fielding et a!., 1998, Valsesia-Wittmann et al., 1997, Nilson et al., 1996), the availability of the specific protease present at the surface of the target cell limits the application of this strategy. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Nucleus , 1 Fusion at cell g"rfa^° p Binding ^ Gene delivery ^ 2 Internalization ^ Fusion in endosome Figure 1.3. Possible barriers to prevent efficient transduction by modified MoMuLV retroviral vectors displaying a new binding moiety directed to a specific receptor. 1. Fusion between the viral and cellular membranes does not occur at the cell surface following receptor binding. 2. Fusion between the viral and endosomal membranes does not take place following internalization of the bound virus. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14 1.3. Retroviral Vectors Using a vector system to study aspects of retroviral entry provides several advantages over the use of a natural replication-competent retrovirus (RCR). First, as a safety concern, the vector system avoids working with a replication-competent retroviral system. Second, it produces a high titer (107 cfu/ml). Third, it also provides a convenient way to analyze envelope protein function through the use of pseudotyped retroviral vectors. A natural retrovirus carries functional gag-pol and env genes in its viral genome. Most importantly, it contains a packaging signal (designated psi or \\r in the vector genomes) which allows the packaging of its viral genome into viral particles. Following infection, the integrated proviral DNA expresses Gag-Pol and Env proteins, which produce infectious particles. Therefore, RCR has the ability to continue carrying out infection. On the other hand, in a replication-defective vector system, the essential viral components (gag-pol and env) are expressed from two separate plasmids, and the viral genome has been replaced with a trans gene or a selection marker gene (P-gal or neor ), (Figure 1.4.). Following the initial infection, the integrated viral genome, which lacks gag-pol and env genes, can no longer produce viral particles in the infected cell. Therefore, this replication-defective retroviral system is a “ one-round" vector system. Because the infection is one round, it is usually Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15 referred to as transduction rather than infection. All the experiments described in this thesis were based on a replication-defective retroviral vector system. MuLV CMV SV40 ^g a^-“ ^ N e o MuLV vector U. i R Fusion protein HA or other viral fusion proteins (MoMuLV Env, VSV-G) Figure 1.4. Three-plasmid replication-defective retroviral vector system. 293T cells transiently transfected with three plasmids expressing the viral components produce retroviral vectors carrying a trans gene in its vector genome. The retroviral particles produced by the vector system are limited to one-round of infection (Soneoka et al., 1995). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16 1.4. Experimental design and preliminary data Among all the modifications of MuLV envelope proteins described above, insertion of a new binding moiety at the N-terminus of MuLV SU is currently the most effective approach to targeting. These modified envelope proteins are efficiently expressed on viral particles and are able to redirect virus binding to target cells. However, their poor efficiency in transducing target cells calls for a solution, which will allow efficient fusion triggering after receptor binding. I have attempted to overcome the problem of lack of fusion obtained in previous targeting approaches. I have devised two novel strategies to attempt to improve transduction efficiency. One strategy is to uncouple the binding and fusion functions of the fusion protein by co-expressing a targeted MuLV envelope protein with an influenza HA protein. In HA, binding occurs on the cell surface, whereas fusion takes place in response to a non-specific low pH trigger in the endosome. In contrast to the MuLV envelope, where binding and fusion are two related events, my new approach is to separate the binding and fusion processes temporally and spatially so that the two events are independent of each other. In this approach, the modified MuLV envelope protein could mediate receptor-specific binding, whereas the HA Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17 protein should function as a complementary fusion protein. However, as the trigger to a fusogenic function of HA is a low pH environment, the nature of the binding reaction may not matter. My second strategy is to target viral entry by pseudotyping MuLV particles with a modified HA protein, which displays a new binding moiety targeted to a specific receptor. In this case, both the binding and fusion processes will be carried out by the modified HA protein. There are two glycoproteins present on the surface of an influenza virus. They are HA and neuraminidase (NA). These two proteins are highly mutatable allowing an effective response to host immune pressure. Due to this high mutation rate, HA and NA are classified serologically into subtypes. To date, 15 HA and 9 NA subtypes have been identified in nature. The HA of an H2 subtype has been demonstrated able to pseudotype RSV particles (Dong et al., 1992). Recently, HA of an H7 subtype has also been shown to pseudotype MuLV particles (Haziioannou et al., 1998). So, I have first attempted to repeat the observation made by the latter group, using an H7 subtype HA protein. The data shown in Figure 1.5. and Table 1.1. demonstrate that MuLV particles can be pseudotyped by the H7 HA protein and gave titer of 104 cfu/ml, which was comparable to the previous report. In addition, MuLV particles pseudotyped by the amphotropic envelope protein and protein G of vesicular stomatitis virus, respectively, were included in the experiment for comparison. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18 Second, as influenza virus has been shown to enter host cells via a pH- dependent pathway, I therefore asked whether MuLV vectors pseudotyped by the HA protein also entered cells via a pH-dependent pathway. In order to do this, I pre-treated target cells with NH4CI prior to and during transduction. As a negative control in this experiment, MuLV vectors pseudotyped by amphotropic envelope protein were included, as they are not sensitive to NH4CI treatment. In addition, since vesicular stomatitis virus is known to enter cell via a pH-dependent pathway, I also included MuLV particles pseudotyped by the VSV-G protein as a positive control (Data from my lab has shown that MuLV particles pseudotyped with VSV-G are sensitive to NH4CI treatment). As expected, MuLV particles pseudotyped by the HA protein were also sensitive to NH4CI treatment. As shown in Table 1.2., the titers of MuLV particles pseudotyped by the HA protein were reduced by approximately 2 orders of magnitude in the presence of NH4CI, whereas the titer of MuLV particles pseudotyped by the amphotropic MuLV envelope protein was only slightly affected. The results obtained from this experiment suggest that HA-pseudotyped MuLV particles go through an endocytic pathway. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19 a-VSV-G a-HA a-ampho Fusion protein 12 3 4 5 6 a- HIV CA a-MuLV CA CA 1 3 5 2 4 6 Figure 1.5. HIV- and MuLV-based vectors can be pseudotyped by HA protein. HA protein can pseudotype both HIV- and MuLV- based vectors as efficiently as VSV-G and amphotropic fusion proteins. Top panel. Lanes 1 and 2, VSV-G protein; lanes 3 and 4, HA protein; lanes 5 and 6, amphotropic protein. Bottom panel. Lanes 1, 3 and 5, HIV capsid protein (CA); lanes 2, 4 and 6, MuLV capsid protein. The figures shows an immunblot of virion particles pelleted through 20% sucrose cushion to remove free proteins (for details,, see Chapter 2.3. Materials and Methods). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20 In order to utilize the HA protein solely as a fusion protein, it will be important to prevent its natural receptor binding activity. HA normally binds to sialic acid residues on cell surface glycoproteins. I have therefore attempted to eliminate the receptor binding site in the HA protein (Chapter 2). In addition, I have investigated the fusion mechanism of HA in order to gain insight into how structural modifications of the HA protein could affect its fusion function (Chapter 3). With the information I gathered from these studies, I then constructed several modified HA proteins targeted to different receptors and tested them in the two approaches I described above (Chapter 4). Knowledge gained from this work provides insight to further improve the design of a cell-specific MuLV vector (Chapter 5). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21 Table 1.1. Titer of MuLV particles pseudotyped by different viral fusion proteins. lane Vector fusion protein titer on 293T ce!ls(cfu/m!)a 1 HIV VSV-G 2 x 10® 2 MuLV VSV-G 3 x 10® 3 HIV HA 8 x 104 4 MuLV HA 7 x 104 5 HIV ampho 5x10® 6 MuLV ampho 3 x 10® " ■ Transduction assays were performed as described in Chapter 2.3. Materials and Methods. Table 1.2. HA pseudotyped MuLV vectors enter cells through an endocytic pathway. Titer on target cells (cfu/ml)a Fusion protein NIH 3T3 (%) NIH 3T3 (50 mM NH4CI) (%) ampho 3x10° (100) 2 x 10* (67) VSV-G 1 x 107 (100) 4 x 105 (4.0) HA 4 x 103 (100) 2 x 101 (0.5) *' Transduction assays were performed as described in Chapter 2.3. Materials and Methods. For NH4Ci treatment, cells were incubated with 50 mM NH4CI @ 37°C for 2 hrs prior and an additional 2 hrs during transduction. At the end of incubation, cells then were replaced with regular growth medium until p-gal staining. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22 Chapter 2 Characterization of the receptor binding pocket of influenza hemagglutinin Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23 2.1. Abstract The receptor binding pocket of influenza hemagglutinin (HA) binds to sialic acid residues on cell surface glycoproteins. The residues involved in contacting the sialic acid in the H3 subtype HA protein have previously been identified in the crystal structure of HA complexed with sialic acid. In order to investigate the role of the residues in the receptor binding pocket of the H7 subtype, A/Fowl Plaque Virus/Rostock/34 (FPV), I made mutations of six residues (Y98, T136, H183, E190, L194, S228) suggested by comparison to the H3 subtype structure. The resulting HA molecules were assessed for their ability to be expressed on the cell surface, to bind to sialic acid and for infectivity, using a pseudotyped retroviral vector system. The analysis revealed that all of the mutants were expressed on the cell surface and incorporated into viral particles, except for the substitutions Y98K and Y98F/T136K. Using retroviral vectors, I was able to demonstrate that the receptor binding activity of these mutants was affected to various degrees. In particular, the substitutions Y98F/E190Q, Y98F/S228K, and Y98F/E190Q/S228K greatly impaired receptor binding activity. I also identified two interesting subgroups of mutants whose fusogenicity was affected. Mutations T136A, L194A, Y98Fm36A, and Y98F/L194A had a hyper-fusogenic phenotype, whereas the substitutions H183F/L194A and Y98F/H183F/L194A resulted in hypo-fusogenic proteins. The defect in fusion of the two hypo-fusogenic mutants could be rescued by expressing the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24 proteins at 32°C, and HA-pseudotyped MuLV particles produced at 32°C also showed 5-10 fold increased infectivity. Taken together, these data suggest that mutations in the receptor binding pocket of HA not only can reduce its receptor binding activity, but can also influence the fusogenicity of the protein. 2.2. Introduction The entry of influenza virus into susceptible cells is mediated by the viral glycoprotein hemagglutinin (HA). This protein is comprised of two disulfide-linked subunits, HAi and HA2 , produced by post-translational proteolytic cleavage of the HA precursor (HAo). The function of the HAi subunit is to attach to cell surface molecules containing terminal sialic acid residues during the initial step in viral entry, whereas the HA2 subunit plays an essential role in promoting fusion between the viral and endosomal membranes (Wiley et al., 1987). X-ray crystallography studies have revealed the three-dimensional structure of the ectodomain of HA from the H3N2 X-31 influenza A strain (Wilson et al., 1981). In addition, the structure of H3N2 HA complexed to a receptor analogue has been determined by X-ray crystallography and nuclear magnetic resonance and has provided insights into the receptor binding site and its interaction with sialic acid residues (Eisen et a!., 1997; Niles et al., 1993; Sauter et al., 1989 & 1992; Takemoto et al., 1996; Watowich et al., 1994; and Weis et al., 1988). Information from these studies suggested that residues Y98, S136, W153, H183, E190, L194A, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25 L226, and S228 form the receptor-binding pocket (Figure 2.1.)- Of these residues, Y98, S136, W153, H183, E190, and L194, are conserved among various strains of influenza virus HA, and all except residue L194 are involved in forming hydrogen-bonds with sialic acid (Kelm et al., 1992; and Eisen et al.,1997). Martin et al., (1998) have previously reported that conservative substitutions of residues in the receptor binding pocket had various effects on the receptor binding activity of an H3 subtype protein. In particular, they have shown that substitutions Y98F, H183F and L194A severely impaired erythrocyte binding activity. However, infectivity tests were only performed with the Y98F mutation which, surprisingly, was found to be just as infectious as the wild type virus in MDCK cells, despite its decreased affinity for erythrocytes. In order to more directly examine the contribution of residues in the receptor binding pocket to HA function, I have made mutations of the five conserved (Y98, T136, H183, E190, and L194) and one non-conserved (S228) residue in the receptor binding pocket site of an H7 subtype protein. Also, I have taken advantage of the fact that HA can pseudoytpe retroviral vectors (Hatziioannou et al., 1998) in order to examine the ability of the mutant proteins to bind to and promote transduction of susceptible cells. Combined with cell-cell fusion assays, I have been able to examine the overall properties of these mutants. My study has revealed that mutations in 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. (H3) IFCLALGQDLPGNDN-STATLCLGHHAVPN (H7) MNTQILVFALVAVIPTNADKICLGHHAVSN GTLVKTITDDQIEVTNATELVQSSSTGKIC GTKVNTLTERGVEVVNATETVERTNIPKIC N-NPHRILDGIDCTLIDALLGDPHCD-VFQ S-KGKRTTDLGQCGLLGTITGPPQCD-QFL NETWDLFVERSKAF-SNC EFSADLIIERREGN-DVC jPYDVPDYASLR IPGKFVNEEALR slvassgtlefitegftwtgvtqnggSnac QILRGSGGIDKETMGFTYSGIRTNGT|SAC KRGPGSGFFSRLNML— TKSGSTYPVLNVT RRSGSS-FYAEHEiLLSNTDNASFPQMTKS hpnndnfdklyiwgiHhpstnqIqtsHyvq YKNTRRES ALI VWGI||hSGSTt|qTK§YGS ASGRVTVSTRRSQQTIIPNIGSRPWVRGAS GNKLITVGSSKYHQSFVPSPGTRPQINGffS SRISIYWTIVKPGDVLVINSNGNLIAPRGY GRIDFHWLILDPNDTVTFSFNGAFIAPNRA FKM-RTGKSSIMRSDAPIDTCISECITPNG SFL-RGKSMGIQSDVQVDANCEE-CYHSGG SIPNDKPFQNVNKITYGACPKYVKQNTLKLATGMRNVPEKQTR TITSRLPFQNINSRAVGKCPRYVKQESLLLATGMKNVPEPSKKRKKR 'L19A1. mm*, Figure 2.1. Crystal structure of the receptor binding pocket site of a monomeric HA molecule of H3 subtype. Shown on the left is the amino acid sequence alignment of the HAt subunit of HA proteins between the H3 and H7 subtypes. Residues highlighted are conserved among all HA subtypes. Italicized are residues involved in determining linkage specificity. Underlined are conserved cysteine residues in both the H3 and H7 subtypes. N > o > 27 the receptor-binding pocket affected receptor binding activity for most the mutants and that some substitutions also affected the fusion function of the protein. 2.3. Materials and Methods cDNA and mutagenesis. The cDNA of A/fowl plaque virus/Rostock/34 (FPV) HA was obtained from Dr. Klenk (Institute of Virology, Philipps- University, Marburg, Germany) and used to produce the CMV expression plasmid, pCMVHA. Mutations were generated by splice overlap PCR (Ho et al., 1989), and all plasmids were fully sequenced. (Sequencing Core Facility, USC, CA). Cell lines. NIH 3T3 cells are murine fibroblast cells (ATCC). 293T are human kidney epithelial cells transformed with the large T antigen of SV40 (ATCC). Both cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS). Transient transfection, transduction and determination of viral titer. Retroviral vectors were produced by transient transfection of pCgp, pCnBG and pCMVHA, essentially as described (Soneoka et al., 1995). The pCgp plasmid expresses MuLV Gag-Pol and the pCnBg plasmid is a retroviral vector, which expresses nuclear p-galactosidase downstream of the MuLV LTR and a neomycin resistance gene from an internal SV40 promoter. Plasmids were co-transfected into 60-70% confluent 293T cells in 60 mm Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28 plates by a calcium phosphate precipitation method. Five micrograms of each the pCgp and pCnBg plasmids were used, together with 1 pg of appropriate pCMVHA. At 15 to 17 hours post-transfection, the media was replaced with 3 ml of fresh media. The viral supernatants were harvested 24 hours later, filtered through a 0.45 pm filter (Millipore, CA) to remove cell debris and aliquots were stored at -80°C. Viral titer was analyzed by transduction of NIH 3T3 cells which were plated onto six-well culture dishes the day before at a density of 3 x 104 cells per well. Serial dilutions of the original viral supernatant were prepared, and 1 ml of the diluted viral supernatant with 8 pg/ml polybrene (Sigma, MO) was added to each well, followed by 10-12 hours incubation. Fresh media was then applied and viral titer was determined by X-gal staining at 48 hours post transduction at which time transduced ceils were fixed with 0.5 % glutaraldehyde for 10 min, followed by two subsequent PBS washes of 10 min for each. Cells were then incubated with staining solution (4 mM potassium ferricyanide, 4 mM potassium ferrocyanide, 2 mM MgCfe, 0.4 mg/ml X-gal, Sigma, MO) at 37°C over night). Titer was determined by counting the number of blue colonies under a microscope and multiplying by the appropriate dilution factor. Celt surface expression. Sixty to seventy percent confluent 293T cells were transfected with 1 pg of appropriate pCMVHA and 14 pg of pBlueScript Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29 plasmid (Strategene, CA) as a filler DNA, using the calcium phosphate precipitation method. HA-expressing 293T cells were harvested 36-48 hr post-transfection to analyze for cell surface expression leveis. Five to ten x 105 cells were pelleted (10 seconds at 12,000 r.p.m., eppendorf centrifuge 5415C) and washed with 1 ml of cold wash buffer (10% goat serum in PBS). Pelleted cells were re-suspended in 100 pi of cold wash buffer containing a rabbit polyclonal primary antibody against FPV HA (provided by Dr. Klenk, Institute of Virology, Philipps-University, Marburg, Germany) at 1:500 dilution for 1 hr at 4°C followed by 1 ml wash of cold wash buffer. After washing, the cells were then incubated with a goat anti-rabbit secondary antibody conjugated with FITC (Pierce, iL) at 1:100 dilution in the same conditions as described for the primary antibody incubation. The resulting pellet was then re-suspended in 4% paraformaldehyde and subjected to fluorescent activated cell sorting (FACS) to measure cell surface expression. A mean channel number (MN) was obtained for each sample and converted to fluorescence intensity (f) according to the formula log(/)=aMN+b, in which a and b are constants derived from a linear regression of a standard curve generated by using fluorescein-labeled RCP-70-5 microbeads (Spherotech Inc., IL) with known fluorescence intensities. Virion incorporation and binding. To determine virion incorporation, 1 ml of filtered viral supernatant was further purified through a 20% sucrose cushion by centrifugation at 14,000 r.p.m. for 30 min at 4°C (eppendorf Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30 centrifuge 5415C). Due to the similar molecular weights of the HA2 subunit (~p27 kDa) of HA and the capsid protein (~30 kDa) of the MuLV viral particles, the resulting pellet was then re-suspended in 2 x loading buffer (Novex, CA) split into two halves and subjected to SDS-PAGE. For detection of HA proteins, one blot was probed with anti-FPV HA as the primary antibody (1:5000) and HRP-conjugated goat anti-rabbit IgG as secondary antibody (1:10,000), (Pierce, IL). For detection of MuLV capsid protein (CA), the second blot was incubated with anti-CA polyclonal antibody (Quality Biotech, NJ) at 1:10,000 dilution as the primary antibody and HRP-conjugated rabbit anti-goat IgG as secondary antibody (1:10,000), (Pierce, IL). Protein signals were detected by an ECL kit (Amersham, CA). For detection of sialic acid binding activity, 1 ml of filtered viral supernatant was mixed with 5 x 105 NIH 3T3 cells and incubated for 2 hrs at 4°C to prevent receptor internalization. At the end of the incubation, each sample was washed with 1 ml of wash buffer once. Following the wash, the bound viral particles were detected with anti-FPV HA antibody and FITC- conjugated goat anti-rabbit IgG (Pierce, IL) as the secondary antibody, then subjected to FACS analysis as described for cell surface expression analysis. Polykaryon formation of NIH 3T3 cells. Eight x 104 NIH 3T3 cells were seeded in each well of a 6-well plate the night before transfection and co transfected with 0.1 pg of appropriate pCMVHA and 0.9 pg of pBlueScript Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31 plasmid the following day, using Lipofectamine Plus Reagent Kit (Gibco BRL, CA). Transfection procedures were carried out as described in the standard transfection protocol (Gibco BRL, CA). Thirty-six hrs post-transfection, the cells were washed once with room temperature PBS followed by a 3 min incubation in 1 ml pH 5 PBS buffer (PBS buffer with 10 mM HEPES (Gibco BRL, CA) and 10mM 2-[N-morpholino]ethanesulfonic acid, Sigma, MO) at 37°C. Cells were then washed twice with 2 ml of room temperature DMEM, replaced with 2 ml of DMEM containing 10% fetal bovine serum and incubated at 37°C for 5 hrs to allow cell-cell fusion to occur. At the end of incubation, cells were fixed with 0.5% glutaraldehyde (Sigma, MO) for 10 min at room temperature and stained with 0.1% methylene blue (Sigma, MO) at room temperature for 10 min. Poiykaryons (syncytia) were counted using a light microscope at 40x magnification. A polykaryon was defined as a cell mass containing 4 or more nuclei. 2.4. Results Cell surface expression and virion incorporation o f mutant HAs. To determine whether mutations made in the receptor binding pocket of HA affected the level of expression of the proteins on the cell surface, 293T cells transiently transfected with appropriate HA-expression plasmids were assayed by FACS analysis. Results obtained from these cell surface expression assays showed that most of the mutants, except for mutants Y98K Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32 and Y98F/T136K, were expressed at levels comparable to the wild type HA (Table 2.1 .)• Immunoblots of the cell lysates of the Y98K and Y98F/T136K mutants also revealed poor expression (data not shown), suggesting that their low cell surface expression were due to gross defects in protein transport and processing. Among the mutants that were detected on the cell surface, some mutants (H183F, E190K, L194A, Y98F/H183F, Y98F/E190Q, Y98F/L194A and Y98F/S228A) showed approximately 2-fold higher expression levels than the wild type HA. In addition, other mutants showed a reduction in cell surface levels of between 57 and 88% the wild type level (Table 2.1.). I also assessed the relative efficiency of the incorporation of the HA proteins into retroviral particles. Appropriate HA-expression plasmids were co-transfected with plasmids expressing MuLV Gag-Pol and a vector genome encoding p-galactosidase to produce HA-pseudotyped MuLV particles. Viral supernatant for each mutant HA was then partially purified through 20% sucrose and subjected to Western analysis. The immunoblot results revealed that most of the HA mutants that were expressed on the cell surface were also incorporated into viral particles (Figure 2.2.). However the efficiency of virion incorporation varied and did not always correlate exactly with cell surface expression. For instance, mutants Y98F/E190Q and Y98F/S228A, which showed approximately 2-fold higher cell surface expression levels than the wild type HA, and mutant Y98F/T136A, which showed 2-fold decrease in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 2.1 Summary of receptor binding pocket mutants of FPV HA. Hemagglutinin Incorporation'' Binding- Infectivity^ wild type 100 + + + TO O (1.20 ± 0.30) x 10H S228A 80 ± 7 + + + 107 ± 29 (2,00 ± 2.77) x 104 E190K 1 8 2 ± 15 + + 47 ± 3 (2.90 ± 4.41) x 10" E190Q 121 ± 9 + + + 51 ± 1 5 (1.40 ± 1.44) x 10? S228K 100 ± 17 + + + 55 ± 1 9 (6.67 ± 8.08) x 103 Y98F/H183F 189 ± 3 3 + 55 ± 1 3 (3,67 ± 1.52) x 102 Y98F/S228A 228 ± 25 + + + 53 ±11 (2.23 ± 1.32) x 102 Y98F/H183F/L194A 82 ± 8 ± 55 ± 1 (2.33 ± 1.52) x 101 Y98F 103 ± 1 7 + + + 32 ± 6 (5.00 ± 4.35) x 102 T136A 108 ± 2 6 + + + 39 ± 12 (1.06 ± 1.68) x 102 T136K 134 ± 2 2 + + + 40 ± 1 2 (7.70 ± 10.7) X102 H183F 169 ± 3 7 + 36 ± 1 6 (2.67 ± 1.15) x 102 L194A 220 ± 20 + + 36 ± 9 (1.77 ± 2.80) x 102 Y98F/T136A 66 ± 3 + + + 21 ± 7 (4.33 ± 1.15) x 101 Y98F/E190K 57 ± 8 + + 35 ± 1 7 (2.67 ± 1.53) x 101 Y98F/L194A 220 ± 37 + + 32 ± 1 0 (2.67 ± 1.53) x 102 H183F/L194A 107 ± 2 2 ± 26 ± 9 (1.33 ± 5.77) x 101 Y98F/E190Q8 214 ± 9 + + + 25 ± 3 (6.33 ± 2.31) x 10' Y98F/S228K8 119 ± 14 + + + 17 ± 6 (5.33 ± 2.00) x 101 Y98F/E190Q/S228K® 88 ± 8 + + + 14 ± 4 (2.33 ± 1,52) x 101 Y98K 9 ± 7 - 16 ± 9 (0.33 ± 0.58) x 101 Y98F/T136K 14 ± 1 - 1 3 ± 6 0 *• Percent of cell surface expression (CSE) level of HA mutants In transfected 293T cells ± SD. b‘ Comparison of virion incorporation level of HA proteins relative to the wild type HA. +++, wild type; ++, intermediate; +, low; ±, trace amount; undetectable. c' Receptor binding activity of HA-pseudotyped MuLV particles on NIH 3T3 cells. Percent of receptor binding activity of HA proteins relative to the wild type HA ±SD, d. P-gal titer reported in cfu/ml on NIH 3T3 cells ± SD. 8 HA proteins in bold are binding-defective mutants with wild type cell surface expression, wild type virion incorporation and low infectivity. 34 x8zzs/oo6i.3/d86A Vfr6n/dC8LH/d86A vwn/desm X8ZZS/d86A V8ZZS/d86A Vfr 6 H/d86A D06 L3/d86A X06 l>3/d86A d£8LH/d86A X9£H/d86A V9£kL/d86A X 8 Z Z S V8ZZS vwn 00613 >1061.3 dcsm X £9H V 9 e n X86A d86A adA* phm t t t r* C M < < X X < o c o jl) ^ o to E Q < C O X Q-oo ^ < It s x ■ £ c x: .2 i - < D 03 > * _Q . T J Q 0 ) S r irr v . O o CL 'F u C O £ o o — -£ C O .E "o aj 0 ) o *> c o (0 -m © to > d ) < D '© 03 -g 2 Q . - I < O 5 X ~ ir n < E c E - C O = 3 .E E -Q m ^ 3 E C O C O C O > Q > v 2 Q . _ c u O (0 S .& 0 4 w O < O 3* X © 3 _ C D ~ 0 LU oL t§ oi oi o k . 3 O) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. antibody; th e capsid protein o f th e M uLV viral c o re (CA) w as detected by anti-CA polyclonal antibody. 35 cell surface expression, exhibited similar levels of virion incorporation as the wild type HA. Furthermore, mutants H183F, L194A, Y98F/H183F, H183F/L194A and Y98F/H183F/L194A, which showed similar or higher cell surface expression levels than the wild type HA, were incorporated into virions less efficiently than the wild type HA. These data indicate that the overall cellsurface expression levels of the HA proteins are not the only factors determining the efficiency of incorporation of HA into retroviral particles. It is important to note that the reduction in virion incorporation observed for mutant H183F was also observed for other mutants that carried this particular mutation (mutants Y98F/H183F, H183F/L194A, and Y98F/H183F/L194A). In addition, an adverse effect on virion incorporation was also observed for mutants that contained the L194A or E190K substitutions (mutants Y98F/L194A, H183F/L194A, Y98F/H183F/L194A, E190K and Y98F/E190K). In all cases, these mutations caused a greatly reduced efficiency of virion incorporation. Effects of mutations in the sialic acid binding pocket on receptor binding activity and infectivity. I assessed the receptor binding activity of those HA mutants that could be efficiently incorporated into MuLV particles, using virus-cell incubation followed by FACS analysis. In addition, effects on HA protein function were assessed using transduction of NIH 3T3 cells by pseudotyped retroviral vectors. The results indicated that mutations in the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36 receptor-binding pocket had various effects on binding and infectivity. I have categorized ai! of the HA proteins into five different groups according to their receptor binding activity, regardless of their levels of virion incorporation (Table 2.1.). However, only the ones that were incorporated into virions as efficiently as the wild type HA proteins will be discussed. The first group consists of mutant S228A. This substitution retained wild type receptor binding activity and infectivity, suggesting that the removal of the hydroxyl group of S228 did not affect the overall affinity of HA for sialic acid, despite the fact that this group forms hydrogen bond with the hydroxyl group of C9 of sialic acid residue. The second group contains mutants E190Q, S228K, and Y98F/S228A. These three mutants showed good correlation between binding activity and infectivity, with an intermediate level of receptor binding activity and a corresponding 1 ~2 logs decrease in infectivity. The third group contains mutants Y98F, T136A, T136K, and Y98F/T136A, which displayed significant defects in receptor binding activity, but not a drastic reduction in infectivity. The low receptor binding affinity observed for mutants Y98F and T136A is in good agreement with the data reported by Martin et al., (1998). Furthermore, the double substitution Y98F/T136A that I constructed further reduced the level of receptor binding activity compared to the single substitutions and led to an additional one-log decrease in infectivity. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37 Mutants in the fourth group were primarily severely binding defective mutants. They are mutants Y98F/E190Q, Y98F/S228K, and Y98F/E190Q/S228K. All three binding-defective mutants displayed wild type cell surface expression levels and were incorporated into virions at wild type levels. The triple mutation (Y98F/E190Q/S228K) in particular reduced receptor binding activity by an additional 2-fold over the two double mutants. Also, the levels of infectivity shown by MuLV particles pseudotyped with this HA protein were 3-fold and 6-fold lower than those observed in both mutants Y98F/S228K and Y98F/E190Q, respectively. The last group includes the mutants Y98K and Y98F/T136K that had abnormally low levels of cell surface expression, and as expected, immunoblot results showed that minimal amounts of these two mutants were incorporated into virions. However, MuLV viral particles pseudotyped with these mutants showed a small amount of receptor binding activity detected by FACS, and mutant Y98K showed a very low level of infectivity. Mutations in the receptor binding pocket can affect fusion. It has been shown that the fusion function of the HA protein does not require prior receptor binding as HA proteins incorporated into liposomes which lack sialic acid residues are able to induce membrane fusion following a low pH pulse (Wharton et al.,1986). To find out whether mutations in the receptor binding pocket of HA would cause any effect on the fusion function of the protein, I examined the fusion ability of all the HA mutants in a cell-cell fusion assay. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38 The results showed that most HA mutants displayed normal fusion function, including several that were severely binding defective (e.g. Y98F/E190Q, Y98F/S228K, and Y98F/E190Q/S228K). But others displayed interesting phenotypes (Table 2.2.). There were two groups of HA mutants that affected fusion. One group of mutants exhibited a hyper-fusogenic phenotype and consisted of mutants T136A, L194A, Y98F/T136A, and Y98F/L194A. As shown in Figure 2.3., the number of nuclei per syncytium formed was higher than for the wild type HA. Although it is possible that the hyper-fusogenicity of these mutants is a direct consequence of higher cell surface expression, as for mutants L194A and Y98F/L194A, it does not explain the hyper-fusogenicity observed for mutants T 136A and Y98F/T136A, whose cell surface expression levels were similar to and lower than the wild type HA protein, respectively. In contrast to the above mutants, the mutations of H183F/L194A and Y98F/H183F/L194A resulted in hypo-fusogenic proteins (Figure 2.4.). Several reports have described pH-sensitive HA mutants that are unable to carry out fusion at normal pH (which the wild type HA proteins fuse most efficiently), but are able to carry out fusion at an elevated pH (Daniels et al., 1985; and Dorns et al.,1986). To address the possibility that these hypo- fusogenic mutants could be pH-sensitive mutants, I also performed the cell cell fusion assays at pH 5.2 and 5.5. The results showed that these mutants 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 2.2. Fusion function of mutant FPV HAs at pH 5. Polykaryon formation on NIH 3T3 cells'* Hemagglutinin @ 32°C @ 37°C Wild type + + + + u + + + S228A n.d. + + + E190Q n.d. + + + L194A n.d. + + + + Y98F/H183F n.d. + + + Y98F/L194A n.d. + + + + Y98F/S228A n.d. + + + Y98F/H183F/L194A + + + ± Y98F n.d. + + + T136A n.d. + + + + T136K n.d. + + + H183F n.d. + + + E190K n.d. + + + S228K n.d. + + + Y98F/T136A n.d. + + + + Y98F/E190K n.d. + + + H183F/L194A + + + ± Y98F/Ef90Q n.d. + + + Y98F/S228K n.d. + + + Y98F/E190Q/S228K + + + + + + + T98R '........... n.d. - Y98F/T136K n.d. - a' HA proteins were expressed at either 32°C or 3 7 °C . Polykaryon formation was performed at 37°C. b' Levels of fusion are relative to wild type HA at 37°C. + + + +, 200% ; + + + , 100%; ± , < 10%; no fusion. U > V O Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Y98F/T136A Figure 2.3. Polykaryon formation by HA-expressing NIH 3T3 cells. Hyper-fusogenic HA mutants expressed at 37°C. At 36 hr post transfection, HA-expressing NIH 3T3 cells were washed twice with PBS, and treated with prewarmed pH5 PBS for 3 min. Then cells were washed twice with DMEM lacking serum and incubated further in DMEM for 5 hr. The cells were fixed and stained with methyleneblue. O Wild type H183F/L194A Y98F/H183F/L194A Figure 2.4. Polykaryon formation by HA-expressing NIH 3T3 cells. Hypo-fusogenic HA mutants expressed at either 32°C or 37°C. At 36 hr post transfection, HA-expressing NIH 3T3 cells were washed twice with PBS, and treated with prewarmed pH5 PBS for 3 min. Then cells were washed twice with DMEM lacking serum and incubated further in DMEM for 5 hr. The cells were fixed and stained with methyleneblue. 42 remained hypo-fusogenic at the higher pHs (data not shown). However, I was able to show that the defect in fusion function of these mutants could be fully rescued by expressing the proteins at 32°C (Table 2.3. and Figure 2.4.). Since the expression of these hypo-fusogenic mutants at 32°C could rescue their fusion function, I next asked whether the recovery of their fusion activity was due to an increase in cell surface expression, and whether the production of MuLV particles pseudotyped with these HA proteins at 32°C could lead to enhanced infectivity. As shown in Table 2.3., the cell surface levels of the wild type HA, H183F/L194A and Y98F/H183F/L194A proteins expressed at 32°C were slightly lower than those observed at 37°C. However the relative levels of expression with respective to the wild type protein remained similar. Although immunoblot results showed that levels of virion incorporation of all three proteins were slightly higher at 32°C than those observed at 37°C (Figure 2.5.), only MuLV particles pseudotyped with H183F/L194A and Y98F/H183F/L194A had enhanced titers at this temperature with a 5 and 10 fold increased infectivity, respectively. These data suggest that the increased levels of infectivity seen for mutants H183F/L194A and Y98F/H183F/L194A expressed at 32°C were not due to higher levels of incorporation, but due to an enhancement of their fusion ability. However, the fact that the recovery of fusion function for these two mutants did not lead to a significant increase of infectivity suggests that the cell- cell fusion reaction may not fully reflect the virus-cell fusion reaction. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. 37°C 32°C HAi h a 2 - ► CA -► Figure 2.5. Incorporation of hypo-fusogenic HA mutants into MuLV particles. Immunoblot of HA-pseudotyped viral particles produced at either 32°C or 37°C. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 2.3. H183F/L194A and Y98F/H183F/L194A are temperature-sensitive mutants. Hemagglutinin Polykaryon formation on NIH 3T3 cells' CSE Titer (cfu/ml) of vector on 293T cellsv @ 32“C @ 37U C @ 32“C @37U C @ 32“C @37“C wild type + + + + + + + 85 ± 7 100 (2.50 ± 0.71) x 10“ * (4.50 ±0,50) x 10'1 H183F/L194A + + + ± 86± 5 108 ± 13 (3.50 ± 0.50) x 102 (4.00 ± 2.83) x 101 Y98F/H183F/L194A d . r». LJ A - + + + ± ______ j . 4 a I i U a . o n O o ___r*-rO r> n - i . . i. ____ 56 ± 7 76 ± 8 (2.95 ± 2.90) x 102 (2.00 ± 1.41) x 101 * Cell surface expression of HA mutants expressed on 293T cells either at 32°C or 37°C. The percentage is normalized to wild type HA expression at 37°C, '■ P-gal titer reported in cfu/ml are the average of three independent experiments ± SD. HA-pseudotyped vectors were produced at either 32°C or 37°C . Transduction assays were performed at 37°C, 45 2.5. Discussion I have made several mutations in the receptor-binding pocket of an H7 subtype HA from FPV. Most of these mutations were well tolerated and all of the mutants were expressed at the cell surface, and cleaved endogenously into their two subunits, HAi and HA2, except for mutants Y98K and Y98F/T136K. The poor protein expression seen in cell lysates for the two mutants suggest that these proteins were probably mis-folded and degraded. The interaction between the receptor binding pocket in HA and its receptor, sialic acid, involves the formation of a hydrogen-bonded network (Eisen et al., 1997; Sauter et al., 1992; Watowich et al., 1994; and Weis et al., 1998). The results of my analysis have revealed that residues Y98F, T136, and to some extent, H183 appeared to be important for sialic acid binding. In particular, the single substitution of Y98F greatly reduced receptor binding activity by 70%. The loss of hydrogen bond formation between Y98 and the hydroxyl group of C8 of sialic acid residue by the Y98F substitution probably accounts for the defect in receptor binding activity. Likewise, the loss of the hydrogen bond between the hydroxyl group of T136 and the carbonyl group of C1 of sialic acid residue could accounts for the defect in the T136A substitution. Furthermore, the double substitutions of Y98F/T136A showed a further reduced binding and infectivity, thus confirming that these two residues are important in contributing to receptor binding affinity. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 46 Although the substitution of H183F would abolish hydrogen bond between the hydroxyl group of H183 and the hydroxyl group of C9, it is difficult to interpret this mutant because the reduced receptor binding activity observed for mutant H183F could also be due to reduction in virion incorporation into MuLV particles. However, surprisingly, the combination of Y98F/H183F resulted in a compensatory mutant, which partially alleviated the defect in receptor binding activity seen in both the Y98F and H183F mutants, and the cell surface expression and virion incorporation levels of this double mutant resembled that of mutant H183F. Taken together, these data suggest that residue H183 might not contribute significantly to receptor binding activity. Rather, I propose the substitution of H183F causes a structural alteration in the HA molecule in such a way that it may compensate for the defect observed in the Y98F mutant. In my study, mutant L194A showed a 60% decrease in receptor binding activity, whereas the same mutation generated in an H3 subtype (Martin et al., 1998) showed an almost complete loss in binding to human erythrocytes, although it retained the ability to agglutinate turkey (avian) erythrocytes. Several reports have provided evidence that different influenza strains have different affinities toward a-2,3 and a-2,6 sialic acid linkages, and that this is determined by the residues present at positions 226 and 228 (Rogers et al., 1983; and Vines et al., 1998). The HA protein studied by Martin et al., (1998) was a human H3 subtype which is known to bind Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 47 preferentially to a-2,6 sialic acid. On the other hand, the HA protein which I describe here is an avian H7 subtype which is known to bind preferentially to a-2,3 sialic acid. The discrepancy between the two observations could therefore be due to the differences in the preference for the a-2,3 sialic residue linkage. Therefore, it is possible that residue L194 functions as an accessory residue in modulating linkage specificity in addition to residues 226 and 228. The interaction between HA and its receptor is thought to be multi- valent and cooperative. The data present in this study demonstrate that mutations in the receptor binding pocket of HA reduced receptor binding activity, and all of the mutants with reduced binding affinity also demonstrated decreased infectivity. However, the FACS-based virus-cell binding assay carried out in this study was not a good indicator of performance in infectivity assays as assessed by transduction efficiency. For example, mutants E190Q, S228K, and Y98F/S228A had higher levels of receptor binding activity than mutants Y98F, T136A, and T136K, but the infectivity levels of all six mutants were similar. The lack of a direct correlation between receptor binding activity and infectivity observed for these mutants is probably as result of the complexity of the process of transduction, which requires both binding and post-binding functions. It is possible that different mutants have different effects on post-binding functions. Alternatively, different mutants could also Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48 have different sensitivities to the density of receptors on the ceil surface. It will be interesting to find out whether there is any cell-type specific differences in the ability of these mutants to transduce cells. Data from this study unexpectedly revealed that mutations in the receptor binding pocket could affect the fusion function of the HA protein. A hyper-fusogenic phenotype was observed for mutants containing the T136A or L194A substitutions. In contrast, mutants that contained the substitution .H183F and L194A (mutants H183F/L194A and Y98F/H183F/L194A) exhibited a hypo-fusogenic phenotype. The ability to restore fusion function to these two mutants when expressed at 32°C, but not at higher pHs, indicates that these mutants are 'temperature-sensitive' mutants. Taken together these data suggest that the extreme fusion phenotypes observed for these mutants may reflect instability in the protein. The mechanism of how these mutations affected the fusion function remains an open question. It has been controversial whether receptor binding of HA enhances fusion and indeed whether binding and fusion occur on the same trimer, or whether the two processes are independent of each other. Studies by Millar et al., (1999) and Schoen et al., (1996) reported an increase of fusion efficiency with virus and liposomes that contained sialylated receptors, suggesting bound HA is involved in facilitating fusion. In contrast, Ellen et al., (1990) and Alford et al.(1994) reported that bound HA precluded the fusion function for that HA, suggesting that receptor bound HA is not involved in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 49 membrane fusion. My triple mutant (Y98F/E190Q/S228K) directly addressed this controversy. This triple mutant revealed a wild type level of fusion that had almost no receptor binding activity. My data therefore indicate that there is no requirement for binding to allow wild type fusion, and is therefore consistent with the reports by Alford et al., (1994) and Ellen et al., (1990) that receptor binding by HA does not enhance its membrane fusion function. However, since there is a residual receptor binding activity retained for the triple mutant (Y98F/E190Q/S228K), it could be that, as a formal possibility, minimal (-10%) binding is sufficient for fusion activity. In conclusion, this study has dissected the interactions that occur between the residues that form the receptor binding pocket site in the HA molecule and sialic acid. Most importantly, the results obtained from this study have identified a purely binding-defective mutant, Y98F/E190Q/S228K, that has wild type cell surface expression, is incorporated efficiently into vectors, and has the ability to promote cell-cell fusion, but has severely reduced ability to bind to its receptor. This molecule, which is designated as HAtmt, will be used in the targeting strategies described in Chapter 4. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 50 Chapter 3 Conversion of H1 subtype hemagglutinin to endogenously cleaved protein does not result in functional HA: separation of cleavability and pathogenicity. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51 3.1. Abstract Cleavage of the influenza virus hemagglutinin (HA) polyprotein into two disulfide-linked subunits, HA, and HA2 , is necessary for the function of the protein and viral infectivity. A direct correlation has been established between the nature of the cleavage site and viral pathogenicity, as virulent strains of influenza contain polybasic cleavage sites that are recognized by endogenous furin-like proteases, whereas avirulent strains contain only a single basic residue and are cleaved extracellularly in a restricted tissue pattern. Human strains of influenza are avirulent, but concern exists that the acquisition of a polybasic cleavage site by a human virus could lead to the emergence of a highly virulent strain of influenza. Here, I have investigated this possibility by engineering the cleavage site of the HA of the human H1 subtype virus, A/PuertoRico/8/34 (PR8) and by examining the properties of the resulting HA proteins using pseudotyped retroviral vector particles. In construct H1(7a), the single basic residue at the cleavage site was replaced by the KKREKR polybasic motif present in the virulent H7 subtype virus, A/Fowl plaque virus/Rostock/34 (FPV). In addition, and upstream amino acids were additionally deleted in order to conserve the spacing at the H1 cleavage site. Although this arrangement resulted in some endogenous cleavage of HA, it was very inefficient when compared to the level of cleavage obtained with the H7 protein. Furthermore, unlike the H1 protein, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52 H1(7a) was partially resistant to cleavage by exogenous trypsin treatment, and even trypsin treated HA-pseudotyped retroviral particles gave very low titers on target cells. In construct H1(7b), a longer stretch of the H7 sequence, EPSKKREKR, was inserted in place of the arginine and five upstream amino acids at the H1 cleavage site. This arrangement retained the spacing present in the H7 protein. H1(7b) protein was found to be efficiently cleaved in 293T cells. However, despite its ability to be cleaved endogenously, the H1(7b) protein was not able to form functional pseudotyped vectors. Further analysis revealed that H1(7b) protein had defects in both binding and fusion functions. These data demonstrate for the first time that alterations in the HA protein of avirulent human strains that allow endogenous cleavage will not necessarily allow the formation of a functional protein. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53 3.2. Introduction The entry of influenza viruses into susceptible cells requires the fusion of the lipid membrane surrounding the virus with host endosomal membranes. Both the initial binding event and subsequent fusion within the endosome are mediated by the action of the viral glycoprotein, hemagglutinin (HA) (reviewed in White 1992, 1994 and 1995). HA is initially synthesized as a precursor polypeptide (HAJ that has receptor binding activity, but lacks fusion activity. The protein is post-translationally cleaved into two disulfide- linked subunits, HA, and HA^ The HA, subunit binds to sialic acid-containing receptors on the cell surface, while HA2 mediates fusion following exposure to the low pH environment of the endosome. The cleavage of HA,,, which converts HA into a meta-stable fusogenic form, is essential to produce a functional protein (Chen et al., 1998), (Figure 3.1.). The nature of the cleavage motif present in HA plays an important role in determining the tissue tropism and overall pathogenicity of different subtypes of the influenza viruses (Table 3.1). For example, the avian avirulent and mammalian strains contain a single arginine residue at the cleavage site and are cleaved only when HA is present at the cell surface. The proteases involved are thought to be secreted trypsin-like proteases, such as plasmin and tryptase Clara, in addition to secreted bacterial proteases (Kido et al., 1992, Tashiro et al., 1992). The restricted availability Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Cleavage site Figure 3.1. Crystal structure of the monomeric uncleaved HA molecule (H3 subtype). Arrow indicates the position of the cleavage site on an exposed loop. Cleavage of HA0 converts HA into a meta-stable fusogenic form. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3.1. Cleavage sites of influenza virus H A . Virus Cleavage motif Site of proteolytic cleavage Identified proteases Human intluenza viruses (H I, H2 and H3 subtypes) Avirulent avian influenza viruses (all subtypes) Single arginine Cell surface Tryptase Clara, plasmin, bacterial proteases and trypsin Virulent avain influenza viruses (H5 and H7 subtypes) Multi-basic residues with RXR/KR motif Trans-golgi network (TG N) Furin, and furin-related proteases U ) u 56 of such proteases leads to the characteristic tissue tropism of these viruses, which typically cause limited respiratory tract infections rather than more systemic disease (reviewed in Tashiro and Rott 1996). In contrast, the HA cleavage sites of the avian virulent strains contain multiple basic residues (RXR/KR) and are recognized by furin (Barr, 1991 and Stieneke-Grober et al., 1992) or furin-related endoproteases (Horimoto et al., 1994a; and Molly et al., 1994). For these proteins, cleavage occurs in the trans golgi network. Additional studies have shown that recognition by furin is sensitive to both the actual amino acid sequence and also its position within HA (Horimoto et al., 1994b; Kawaoka et al., 1988; Morsy et al., 1994; Ohuchi, et al., 1991; Tashiro & Rott 1996, review; and Vey et al., 1992). This is in agreement with data from the crystal structure of the H3 subtype of HA, showing that the cleavage site is located on an exposed loop (Chen et al., 1998). Finally, it has been shown that exogenous trypsin treatment can be used to cleave HA from avirulent strains, and also to cleave variants of virulent strains whose HA can no longer be cleaved endogenously (Klenk et al., 1975, Ohuchi etal., 1991, Vey etal., 1992. Horimoto etal., 1994b). The virulent avian influenza viruses are only found in the H5 and H7 subtypes and the presence of the polybasic cleavage motif has been shown to correlate strongly with pathogenicity for these avian viruses (Hiromoto et al., 1994b; Khatchikian et al., 1989; Li et al., 1990; Ohuchi et al., 1989; and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57 Orlich et al., 1990). HA proteins containing such cleavage sites can be cleaved endogenously by both avian and mammalian cells, as well as by invertebrate cells (Kuroda et al., 1986). In contrast, the influenza viruses that infect human cells are either subtypes H1, H2 or H3. These strains have HA proteins that contain only a single basic residue at the cleavage site and are not cleaved endogenously. Considerable interest has been shown in studies addressing the question of whether mutations in the cleavage motif of human influenza HA could produce an endogenously cleaved form which could potentially expand the tissue tropism of the virus and result in a more virulent disease in humans. It is widely accepted that all subtypes of influenza virus originated from wild aquatic birds (reviewed in Steinhauer 1999). It is also known that the avian viruses do not replicate efficiently in human tissues. This is probably due to the different linkages present in the terminal sialic acid residues in human and avian glycoproteins, and the specificity of avian HA for a-2,3 galactoside linkages. Several mechanisms have been proposed to account for the evolution of new human viruses, including genetic re-assortment between avian and human viruses in a common host such as pigs (Castrucci, et al., 1993, Claas et al., 1994). It has recently been shown that the lung cells of pigs contain both a-2,3 and a-2,6 linked galactosides, explaining why pigs are susceptible to infection by both human and avian influenza viruses. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58 However, only influenza subtypes H1, H3, and H9 have been found to cause infections in pigs and, to date, there have been no cases reported of infection in pigs by the virulent avian strains of the H5 and H7 subtypes. Therefore, genetic re-assortment in pigs between human strains and the virulent avian viruses of the H5 and H7 subtypes producing a virulent human strain is unlikely to occur. Another possible mechanism of influenza evolution is that avian viruses may adapt to directly infect mammalian/human cells by switching the HA linkage specificity to a-2,6 galactoside, the receptor present on human lung cells. The presence of a polybasic cleavage site in HA could then allow the establishment of a systemic infection. The Hong Kong outbreak of influenza in 1997 was the first reported case of an H5 virulent human virus (Claas et al., 1998), suggesting that transmission of an H5 virulent avian virus from infected chickens to humans had occurred without another intermediate mammalian host acting as a 'mixing vessel'. Fortunately, transmission between people was poor. It is also possible that a human avirulent virus could be mutated de novo to a virulent form (by genetic recombination or RNA slippage events). Precedents exist for the acquisition of a polybasic cleavage site by the avirulent strains of avian viruses. For example, it has been demonstrated that non-homologous recombination between the HA and NP genes of the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 59 mammalian influenza virus A/seal/Mass/1/80 (H7N7) resulted in an endogenously cleaved protein and a corresponding increase in pathogenicity for chickens (Orlich et al., 1994). A similar observation was made following the insertion of a 28S ribosomal RNA sequence into the HA gene of the avian influenza virus A/turkey/Oregon/71 (Khatchikian et al.,1989). These results suggest possible origins for the development of pathogenic avian viruses, although why the virulent viruses are only found in the H5 and H7 subtypes remains an open question. Previous studies using in vitro selection and site-directed mutagenesis have attempted to correlate cleavage patterns with viral pathogenicity. These studies have been performed mostly with avian influenza viruses, and only a few have utilized mammalian influenza viruses. With avian viruses, in vitro studies have indeed shown a correlation between HA cleavability and the resulting pathogenicity in chickens (Khatchikian et al., 1989; Li et al., 1990; Ohuchi et al., 1989; and Orlich et al., 1990). In contrast, studies with mammalian viruses have been less straightforward. For example, although in vitro selection of the mammalian influenza virus A/seal/Mass/1/80 (H7) in MDCK and chicken embryonic cells resulted in the selection of variants that could be cleaved endogenously, further analysis revealed that only the chicken cells-adapted virus had acquired a polybasic cleavage site that could be cleaved endogenously. Intriguingly, the HA variant selected in MDCK Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 cells had acquired a mutation distant from the actual cleavage site. Furthermore, although the HA variant selected in chicken cells could be cleaved endogenously in cell types from different species and was pathogenic in chickens (Li et al., 1990), the MDCK adapted variant was only cleavable in MDCK cells and was not pathogenic in chickens. Similar results reported by Li et al., had also been shown for the X-31 strain (a recombinant influenza virus bearing the HA from the A/Aichi/68 strain) of an H3 subtype selected for growth in MDCK cells (Rott et al., 1984). The inability of an avirulent mammalian virus grown in mammalian cells (MDCK) to acquire a polybasic cleavage site, and the inability of an adapted virus to transmit across species suggests that the possibility of an avirulent influenza virus evolving into a virulent form in human or in other mammalian hosts is small. More direct studies have addressed the possibility of creating an endogenously cleaved virulent human influenza virus by directly creating a polybasic cleavage site by site-directed mutagenesis. For example, Ohuchi et. al., (1991) demonstrated that the insertion of four additional arginines at the cleavage site of the human influenza virus A/Port Chalmers/1/73 (H3 subtype) rendered it susceptible to cleavage by endoproteases in CV-1 cells. This HA protein was also shown to be fusogenic, although the infectivity of this human variant was not tested. I wondered whether the same mechanism could occur in the human H1 subtype viruses to give rise to virulent variants. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Accordingly I took the human avirulent virus, A/PuertoRico/8/34 (H1 subtype) and altered its cleavage site to that of the avian H7 subtypes. In this way I sought to create an HA protein that could be cleaved endogeously by human cells and remain functional. Through the use of pseudodtyped retroviral vectors, I was able to assess the function of the protein. However, the data from this study revealed that in contrast to the studies with the H3 protein (Ohuchi et al., 1991), the acquisition of a sequence that allowed efficient cleavage was not correlated with the maintenance of protein function. The lack of function of this protein provides an explanation as to why no such variants have arisen in nature. This surprising result also suggests that the intermediate meta-stable stage of HA processed after cleavage plays an important role in determining the membrane fusion function of HA. 3.3. Material and Methods cDNA and mutagenesis. The cDNA of A/PuertoRico/8/34 influenza virus HA was obtained from Dr. Yewdell (National Institute of Health, MD). The cDNAs of A/Fowl plaque virus/Rostock/34 HA wild type and mutant 10 (in my study, it is designated as H7cmt) were obtained from Dr. Klenk (Institute of Virology, Philipps-University, Marburg, Germany). These cDNAs were used to produce the CMV expression plasmids, pCMVHA-H1, pCMVHA-H7 and pCMVHA-H7cmt. Mutations were generated by splice overlap PCR (Ho et Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 62 al., 1989), and all plasmids were fully sequenced (Sequencing Core Facility, USC, CA). Cell lines. NIH 3T3 cells are murine fibroblast cells (ATCC). 293T are human kidney epithelial cells transformed with the large T antigen of SV40 (ATCC). Both cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS). QT6 cells are quail fibrosarcoma cells (ATCC) cultured in HAM's F10 (BRL Gibco, CA) with 10% FBS. Transient transfection, transduction and determination of viral titer. Retroviral vectors were produced by transient transfection of pCgp, pCnBG, together with the panel of different HA-expression plasmids, essentially as described (Soneoka et al., 1995). The pCgp plasmid expresses MuLV Gag- Pol and the pCnBg plasmid is a retroviral vector, which expresses nuclear (3 - gal downstream of the MuLV LTR and a neomycin resistant gene from an internal SV40 promoter. Plasmids were co-transfected into 60-70% confluent 293T cells in 60 mm plates by a calcium phosphate precipitation method. Five micrograms each of the pCgp and pCnBg plasmids were used, together with 1 jag of the panel of different HA-expression plasmids. At 15 to 17 hours post-transfection, the media was replaced with 3 ml of fresh media. The viral supernatants were harvested 24 hours later, filtered through a 0.45 jam filter (Millipore, CA) to remove cell debris and aliquots were stored at -80°C. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 Viral titer was analyzed by transduction of NIH 3T3 cells which were plated onto six-well culture dishes the day before at a density of 3 x 104 cells per well. Serial dilutions of the original viral supernatant were prepared, and 1 ml of the diluted viral supernatant with 8 pg/ml polybrene (Sigma, MO) was added to each well, followed by 10-12 hours incubation. Fresh media was then applied and viral titer was determined by X-gal staining at 48 hours post transduction at which time transduced cells were fixed with 0.5 % glutaraldehyde for 10 min, followed by two subsequent PBS washes of 10 min for each. Cells were then incubated with staining solution (4 mM potassium ferricyanide, 4 mM potassium ferrocyanide, 2 mM MgCI2, 0.4 mg/ml X-gal, Sigma, MO) at 37°C over night). Titer was determined by counting the number of blue colonies under a microscope and multiplying by the appropriate dilution factor. Cell surface expression. Sixty to seventy percent confluent 293T cells were transfected with 1 jj.g of the panel of different HA-expression plasmid and 14 ixg of pBlueScript plasmid (Strategene, CA) as a filler DNA, using the calcium phosphate precipitation method. HA-expression 293T cells were harvested 36-48 hr post-transfection to analyze for cell surface expression levels. 5-10 x 105 cells were pelleted (10 seconds at 12,000 r.p.m., eppendorf, centrifuge 5415C) and washed with 1 ml of cold wash buffer (10% goat serum in PBS). Pelleted cells were resuspended and incubated with monocloanl primary Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. antibody (for H1 HA, 200 p .l of H23-E23, provided by Dr. Yewdell, National Institute of Health, MD; for H7 HA, 200 pi of cold wash buffer with anti-FPV HA at 1:500 dilution, provided by Dr. Klenk, Institute of Virology, Philipps- University, Marburg, Germany) for 1 hr at 4°C followed by 1 ml wash of cold wash buffer. After washing, the cells were then incubated with secondary antibody conjugated with FITC (Pierce, IL) at 1:100 dilutions in the same condition as described for primary antibody incubation (goat anti-mouse for H1 HA, and goat anti-rabbit for H7 HA). The resulting pellet was then re suspended in 4% paraformaldehyde and subjected to fluorescent activated cell sorting (FACS) to measure cell surface expression. A mean channel number (MN) was obtained for each sample and converted to fluorescence intensity (/) according to the formula \og(f)=aMN+b, in which a and b are constants derived from a linear regression of a standard curve generated by using fluorescein-labeled RCP-70-5 microbeads (Spherotech Inc., IL) with known fluorescence intensities. Virion incorporation and binding. For virion incorporation, 1 ml of filtered viral supernatant was further purified through a 20% sucrose cushion by centrifugation at 14,000 r.p.m. for 30 min at4°C (eppendorf centrifuge 5415C). Due to the similar molecular weights of the HA2 subunit (~p27 kDa) of HA and the capsid protein (-30 kDa) of the MuLV viral particles, the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 65 resulting pellet was then resuspended in 2 x loading buffer (Novex, CA) split into two halves and subjected to SDS-PAGE. The H1 HA expression (HAo, HA-,, and HA^ was detected with monoclonal antibodies H28-E23 for HA, and RA5-22 for HA2 subunit (provided by Dr. Yewdell, National Institute of Health). For detection of H7 HA protein, the blot was incubated with anti-FPV HA as the primary antibody (1:5000) and HRP-conjugated goat anti-rabbit IgG as secondary antibody (1:10,000), (Pierce, IL). For detection of MuLV capsid protein (CA), the second blot was incubated with anti-CA polyclonal antibody (Quality Biotech, NJ) at 1:10,000 dilution as the primary antibody and HRP-conjugated rabbit anti-goat IgG as secondary antibody (1:10,000), (Pierce, IL). Protein signals were detected by an ECL kit (Amersham, CA). For detection of sialic acid binding activity, 1 ml of filtered viral supernatant was mixed with 5 x 10s NIH 3T3 cells and incubated for 2 hrs at 4°C to prevent receptor internalization. At the end of incubation, each sample was washed with 1 ml of wash buffer once. Following the wash, the bound viral particles were resuspended and incubated with primary antibody (for H1 HA, 200 pi of H28-E23 monocloanl antibody; for H7 HA, 200 pi of cold wash buffer with anti-FPV HA at 1:500 dilution) for 1 hr at 4°C followed by 1 ml wash with cold wash buffer. After washing, the cells were incubated with FITC-conjugated secondary antibody, then subjected to FACS analysis as described for cell surface expression analysis. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 6 Exogenous trypsin treatment Filtered viral supernatant was pelleted as described in the virion incorporation assay. Viral pellets of duplicate samples were resuspend in 200 jllI of DMEM with or without 5 pg/ml trypsin and incubated at 37°C for 10 min. At the end of trypsin incubation, 800 pi of DMEM with 10% FBS was added into the samples and subsequently subjected to micro-centrifugation in the absence of sucrose cushion at 14,000 r.p.m. for 20 min at 4°C (eppendorf centrifuge 5415C). The resulting pellet was then subjected to SDS-PAGE or titer assay. Polykaryon formation o f NIH 3T3 cells. 8 x 104 NIH 3T3 cells were plated in each well of a 6-well plate the night before transfection and co transfected with 0.1 pg of different HA-expression plasmids, together with 0.9 pg of pBlueScript plasmid the following day, using Lipofectamine Plus Reagent Kit (Gibco BRL, CA). Transfection procedures were carried out as described in the standard transfection protocol (Gibco BRL, CA). Thirty-six hrs post-transfection, cells were washed once with room temperature PBS followed by 3 min incubation of 1 ml pH 5 PBS buffer (PBS buffer with 10 mM HEPES and 10mM 2-[N-morpholino]ethanesulfonic acid, Sigma, MO) at 37°C. Cells were then washed twice with 2 ml of room temperature DMEM, replaced with 2 ml of DMEM containing 10% fetal bovine serum and incubated at 37°C for 5 hrs to allow cell-cell fusion to occur. At the end of incubation, cells were fixed with 0.5% glutaraldehyde (Sigma, MO) for 10 min Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. at room temperature and stained with 0.1% methylene blue (Sigma, MO) at room temperature for 10 min. Polykaryons (syncytia) were counted using a light microscope at 40* magnification. A polykaryon was defined as a cell mass containing 4 or more nuclei. 3.4. Results Generation of H1 subtype HA proteins with H7 poiybasic cleavage site. The H1 subtype virus A/PR/8/34 contains a single arginine residue at its cleavage site that is not cleaved by an endogenous protease when expressed in human 293T cells (Table 3.2. and Figure 3.2.). In contrast, the HA from the H7 subtype virus A/Fowl plaque virus/Rostock/34 contains a polybasic cleavage site (KKREKR), and is cleaved endogenously in these cells. In order to create an H1 protein that could be endogenously cleaved, I generated two chimeric HA proteins based on the H1 variant but containing the H7 cleavage site. In H1(7a), I replaced the 4 amino acids preceding the cleavage site with the polybasic sequence, while in H1(7b), I substituted a longer stretch of 10 amino acids in place of the 5 amino acids preceding the cleavage site. In this configuration, I reasoned that an insertion of 10 amino acids would probably produce a more authentic H7-like loop sequence. I examined the abilities of these variants to be cleaved endogenously in 293T cells. The results showed that, as expected, H1 remained mostly Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6 8 uncleaved in 293T cells. Variant H1(7a) which contains the KKREKR motif was not rendered cleavable to endoprotease, whereas variant H1(7b) which contains the PEPSKKREKR motif, was highly cleavable in 293T cells (Table 3.2. and Figure 3.2.). As a comparison, I also included the wild type H7 HA and its variant H7cmt, which contains an insertion of an amino acid glycine at the carboxyl- terminus of the HA! subunit (Vey et al., 1992). This insertion caused a shift of the KKREKR cleavage motif one residue toward the amino terminus, and was not rendered cleavable to endoprotease. My data, which are in accordance with Vey et al., (1992) showed that H7 is highly cleavable in 293T cells. However, when such a cleavage motif is not correctly present in the HA molecules as a result of one amino acid insertion at the carboxyl-terminus of the HAt subunit, susceptibility to endoprotease in 293T cells was mostly abolished (Figure 3.2.). All HA variants are expressed on the cell surface and incorporated into MuLV particles. To assess the ability of these HA variants to express on the cell surface and to incorporate into MuLV particles, cell surface expression level of the H1(7a) and H1(7b) proteins was measured by FACS analysis. It has been recently shown by Hatziioannou et al. (1998) that MuLV viral cores can be pseudotyped by FPV HA to transduce a variety of cell lines with titers ranging from 102to 104 . Accordingly, I generated MuLV pseudotyped Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 69 retroviral particles with the HA of H1 subtype (A/Puerto Rico/8/34) by employing a three-plasmid transient vector expression system (Soneoka et al., 1995). The results showed that the replacement of a single arginine with polybasic residues at the cleavage site reduced cell surface expression levels for both variants by 2-fold compared to the wild type HA (Table 3.3.). In addition, these two variants also seemed to affect viral production; the amount of viral particles were observed to be lower than those pseudotyped with the wild type HA. In contrast, the insertion of a glycine residue at the carboxyl terminus of the cleavage site in mutant H7cmt did not affect its cell surface expression nor its ability to be incorporated into MuLV particles, which were comparable to the parental H7 protein (Table 3.3. and Figure 3.3.). Variant H1(7a) is resistant to exogenous trypsin treatment in the context of HA-pseudotyped MuLV particles. Many studies have shown that HA of avirulent viruses, as well as variants of virulent viruses, which contain a single arginine at the cleavage site, can be cleaved by exogenous trypsin treatment (Horimoto et al., 1994b; Kawaokaetal., 1988; Morsyetal., 1994; Ohuchi, et al., 1991; and Vey et al., 1992). In addition, Dong et al., (1992) have previously shown that HA of human influenza strain A/Japan/305/57 (H2 subtype) can pseudotype the avian retrovirus, Rous sarcoma virus (RSV). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. if S a S S fS ffi. HA0 H V HA, Figure 3.2. Immunobiot of cell lysates of HA variants. Cell lysates of 293T cells transfected with HA-expressing plasmids were collected 36 hrs post transfection and subjected to SDS-PAGE. H1 derived proteins were detected with monoclonal antibodies H28-E23 and RA5-22 to HA! and HA2 subunits, respectively. H7 expression was detected with anti-FPV HA. -j o Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3.2. Cleavage sites of influenza HA variants. Hemagglutinin HA cleavage site® Cleavability HAi ha2 in 293T cells Trypsin0 A/PR/8/34(HI) HI NIPSIQS-----R GL - + HI (7a) NI-----KKREKR GL ± ± HI (7b) NIPEP-SKKREKR GL + n.a. A/FPV/Rostock/34(H7) H7 NVPEP-SKKREKR GL + n.a. H7cmtd NVPEP-SKKREKRG GL - + “ ■ Cleavage motifs (underlined) of HA of H I and H7 subtypes were aligned with respect to the asparagine in the carboxyl terminus of the HAj subunit. b ' Cleavage of HA variants in 293T cells. +, majority cleaved; ±, reduced cleavage; not cleaved. G Cleavage of HA incorporated into MuLV retroviral particles and treated with 5 mg/ml trypsin at 37°C for 5 min. +, majority cleaved; ±, reduced cleavage; not cleaved, n. a. = not applicable. * H7 cleavage defective mutant previously described by Vey et al. (mutant 10). Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3.3. Cell surface expression of HA variants Hemagglutinin Cell surface expression on 293T cells9 H1 100 H1(7a) 46 ± 9 H1(7b) 44 ± 5 H7 100 H7cmt 96 ± 4 a Levels of cell surface expression are made relative to the respective wild type HA. Results obtained were the average of three independent experiments ± SD. •^4 N ) Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. h- h- + . < -• +* +* E E E t »- 0 O O + ' N N N N N 1 I I I I I I- H + . S'S'S* K N K r r r X X X H h- + ■ ( B (0 (0 f c . 1- 1 - 1- N N N + 1 r r r T " r t~ X X X X X X HA0 HA1 HA2 Figure 3.3. Vector incorporation into virions and susceptibility to exogenous trypsin treatment of HA variants. H1 derived proteins were detected with monoclonal antibodies H28-E23 and RA5-22 to HA1 and HA2 subunits, respectively. H7 expression was detected with anti-FPV HA. The capsid protein (CA) of MuLV particles was detected with anti-CA antibody. Tr = trypsin treatment. 74 However, HA-pseudotyped RSV particles failed to transduce target cells efficiently (7-9 x101 cfu/ml) even when exogenous trypsin was used to cleave HAq I was then interested to address the question whether the cleavage of HA proteins, H1, H1(7a), and H7cmt, which were not cleaved endogenously, could be rescued by exogenous trypsin in the context of HA-pseudotyped retroviral particles. The results showed that H1, which contains a single arginine at the cleavage site, was cleaved by exogenous trypsin treatment. In contrast, variant H1(7a) which contains the KKREKR motif was resistant to exogenous trypsin treatment, with only a slight increase in the intensity of both the HA., and HA2 subunits. Similarly to H1 protein, the cleavage mutant H7cmt, which could not be cleaved endogenously, was highly cleavable by exogenous trypsin treatment, in good agreement with Vey et al., (1992). HA-pseudotyped MuLV particles are stable to exogenous trypsin treatment Since the exogenous trypsin treatment procedure requires the absence of serum in order for trypsin to be active, serum present in viral supernatant was removed by micro-centrifugation through 20% sucrose cushion, followed by trypsin treatment and an additional centrifugation procedure. I was interested to find out whether HA pseudotyped viral particles that proceed through these procedures would remain viable to be able to transduce cells at comparable efficiency as those that did not go Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. through the procedures. The H7 wild type protein, which can be cleaved endogenously, is the protein that can be used to answer this question. The titer results revealed that the difference in titer of MuLV particles pseudotyped with the H7 HA protein between those that had been treated with trypsin and those that had not was less than 2 fold (Table 3.4.). This suggests that the exogenous trypsin treatment did not severely damage the retroviral vectors since 60-70% of the viral particles remained viable. H1(7b) shows no correlation between cleavability and infectivity of HA pseudotyped MuLV particles. In order to study the correlation between cleavability and infectivity, MuLV particles pseudotyped with the panel of different HA proteins were generated, treated with trypsin then assessed in transduction assays. These data showed that MuLV viral particles pseudotyped with the H1 HA protein, due to the lack of HA cleavage, transduced target cells poorly in the absence of exogenous trypsin treatment. However, when treated with exogenous trypsin prior to transduction, they were able to transduce target cells with high efficiency, and the titer was only 3-5 fold less than those given by MuLV particles pseudotyped with the H7 HA protein. As expected, MuLV particles pseudotyped with the H1(7a) HA protein, which showed resistance to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3.4. Infectivity of HA-pseudotyped MuLV vectors in 293T and QT6 cells. Titer (cfu/ml) of vector on cell linea Hemagglutinin 293T QT6 H1 0 0 H1 -Tr* 0 0 H1 +Tr* (1.67 ± 0.15) x 104 (9.72 ± 5.20) x 103 H1(7a) 0 0 H1(7a)-Tr 0 0 H1(7a) +Tr (5.00 ± 1.15) x 101 (1.50 ± 2.34) x 102 H1(7b) 0 0 H1(7b) -Tr 0 0 H1(7b) +Tr 0 0 H7 (8.42 ± 2.47) x 104 (7.22 ± 1.98) x 104 H7 -Tr (5.03 ± 2.69) x 104 (4.57 ± 2.42) x 104 H7 +Tr (4.68 ± 2.42) x 104 (5.08 ± 2.54) x 104 H7cmt (7.06 ± 2.54) x 10' 0 H7cmt-Tr (3.38 ± 2.88) x 101 0 H7cmt +Tr (3.00 ± 0.70) x 102 (2.76± 1.76) x 101 a ' Titer results w ere averaged from at least three independent experim ents and are presented as the mean number of P-expressing colonies ± the standard deviation of the mean. b ' T r = Trypsin treatm ent. + Tr, trypsin treatm ent at 37°C for 5 min; -Tr, m ock trypsin treatment. *-4 0 \ 77 exogenous trypsin treatment, transduced target cells at an efficiency approximately 2-log less than that of H1 HA protein (Figure 3.3. and Table 3.4.). Surprisingly, although the H1(7b) HA protein was shown to be highly cleavable by endogenous protease, MuLV particles pseudotyped with H1(7b), unlike that of the H7 HA protein, failed to transduce target cells (Table 3.4.). Similarly, the variant H7cmt, which was shown to be highly cleavable by exogenous trypsin, gave two-log lower titer than its parental H7. However, the low transduction ability of this mutant could be explained by a defect in fusion, previously reported by Klenk's group in 1992 (Vey et al., 1992). H1(7a) and H1(7b) have reduced receptor binding activity. To explain why the variants H1(7a) and H1(7b) which could be partially cleaved by exogenous trypsin and by endoproteases, respectively, were unable to transduce cells as efficiently as their parent H1 protein, I posed the question whether an insertion of polybasic residues at the cleavage site would affect the receptor binding ability of these HA variants. To examine the binding ability of the HA-pseudotyped MuLV viral particles to sialic-acid residues on the surface of 293T cells, FACS analysis was performed. Since exogenous trypsin treatment caused a reduction in the amount of viable viral particles to some extent, levels of receptor binding were normalized according to the uncleaved form of the respective wild type HA Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. proteins. The data indicated that neither H1 (7a) nor the endogenously cleaved H1(7b) variant bound efficiently to sialic acid compared to the wild type HA (Table 3.5.). In contrast, variant H7cmt bound to sialic acid just as efficiently as its parent H7 protein. Importantly, this data shows that, unlike the fusion process which requires proper cleavage of the HA precursor, HA binding to its receptor is independent of cleavage. H1(7a) and H1(7b) HA variants lack membrane fusion function. To assess the fusion ability of the variants H1(7a), and H1(7b), cell-cell fusion assays were performed. NIH 3T3 cells expressing HA were pulsed with pH 5 media and examined for polykaryon formation. Unexpected, the results showed that the membrane fusion function of not only the H1 (7a), but the H1(7b) proteins as well, was severely compromised (Table 3.5.). It has been shown that an H3 HA binding mutant L226P is capable of going through membrane fusion process at a slightly higher pH (5.7-5.9) than the wild type (pH 5.0-5.3), (Daniels et al., 1985; and Martin et al., 1998). I therefore speculated that the membrane fusion function of the H1(7a) and H1(7b) variants might require a different pH to trigger fusion. However, the results obtained from this experiment showed that at pH 5.5 and 5.8, both the H1(7a) and the H1(7b) variant remains non-fusogenic (data not shown). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3.5. Summary of binding and fusion properties of HA variants. Hemagglutinin Binding to 293T cells* Polykaryon formation of NIH 3T3 cells* H1 100 0 H1 +Tr* n.d. 100 H1(7a) 40 ±10 0 H1(7a) + Tr n.d. 1.5 ±0.7 H1(7b) 68 ±5 4 ±1.4 H7 100 100 H7cmt 90 ±8 0 H7cmt + Tr n.d. 0.7± 0.4 a Tr = trypsin treatment. “ ■ Percentage was calculated relative to the respective wild type HA. Results obtained were the average of three independent experiments, n.d.= not determined. * - 4 VO 80 3.5. Discussion !n this study, I analyzed the cleavability of an H1 subtype HA and correlated it with infectivity using a retroviral vector system. I demonstrated that cleavage does not always correlate with infectivity. The difference observed in cleavability between the H1(7a) and H1(7b) variants supports a model which suggests that the carboxyl terminus of the HAn subunit has a spacer function important in determining the cleavability of HA precursor molecules (Horimoto et al., 1994b; and Kawaoka et al., 1988). Therefore, it is likely that the smaller cleavage motif of H1 (7a) is present in an incorrect position such that in 293T cells it is not accessible by cellular proteases. Furthermore, there is substantial evidence supporting the 'carbohydrate' hypothesis in which the number of basic residues influences the structure and therefore the susceptibility of the connecting peptide by the endoprotease (Molly et al., 1994). The glycosylation state of residue Asn 22 in H3 subtype HA and residue Asn 11 in H5 subtype HA, respectively, has also been shown to play an important role in determining protease accessibility when a polybasic insertion is introduced to the cleavage site between HA1 and HA2 subunits (Kawaoka et al., 1989; Ohuchi et al., 1989 and 1991). In addition, structural studies have revealed that the position of residue Asn 22 in H3 subtype of HA is in close proximity to the cleavage site (Chen et al., 1998). Therefore, it is reasonable to speculate that the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. resistance of the H1(7a) to exogenous trypsin treatment couid be that the secondary structure of its cleavage site is altered in a such way that even trypsin fails to perform cleavage. Since mutations made in H1(7a) resulted in a deletion of a conserved proline residue (Table 3.2.), it is possible that the lack of this conserved proline residue in particular is responsible for the secondary structure of the cleavage site in H1(7a). Proper cleavage of HA precursor is required for fusion, thus the inability of fusion process of H1(7a) could be explained by the lack of cleavage. It is well known that cleavage is required to transform HA into a fusion competent form. However, cleaved HA is in a meta-stable form that is labile (Bullough et al., 1994, Chen et al., 1998). It is surprising that the H1(7b) variant, which displayed a high endogenous cleavability, was incapable of going through membrane fusion process at pH 5 .1 propose that it is unlikely that the endogenous cleavage occurring for the H1(7b) variant is different from that of the wild type H7 HA protein. Rather, I speculate that there could be an intermediate fusion-competent configuration that the cleaved HA is required to adapt in order to trigger its fusion function under low pH conditions, and this particular configuration could be cellular compartment- specific. HA of H1 subtype is known to be cleaved extracellularly. However, by forcing the cleavage of H1(7a) variant to take place intracellularly may prevent normal processing and adoption of a fusion-competent configuration Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 82 in the protein. Alternatively, it could be due to the different amino acids around the C-terminus of HA1 subunit affecting normal folding and subsequent adoption of a fusion-competent configuration in the protein. In both cases, they result in a 'pre-matured configuration' (Figure 3.4.). The data that revealed reduced levels of cell surface expression for both variants, H1(7a) and H1(7b), using the monoclonal antibody E28-E23 could reflect a structural alteration in the variant HA molecules. This may possibly explain the low receptor binding activity observed in my studies. One can answer the question by using a panel of monoclonal antibodies, which bind to structurally distinct antigenic regions of wild type HA (Bachi, et al., 1985; and Caton et al., 1982). In conclusion, this study provides insight into how the meta-stable form of cleaved HA also plays a crucial role in enabling HA fusion function. The current model, which suggests a direct correlation between cleavage and infectivity, may be too simplistic and require modification. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Phenotypes H1 i\\u\ar (fusion competent) low pH ▼ .fusion ■ ■ ► peptide “meta-stable” HA tryP ,tase ttVPs'n Fusion HAo low pH fusion incompetent ♦ No fusion Figure 3.4. Possible reasons for defects in H1(7b) variant. 84 Chapter 4 Targeted retroviral vectors pseudotyped by hemagglutinin protein Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 85 Chapter 4A Receptor-specific targeting mediated by chimeric MuLV envelope protein and hemagglutinin protein: separation of receptor binding and fusion functions on two different molecules Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 86 4A.1. Abstract Gene delivery by MuLV-based retroviral vectors requires efficient entry which consists of two processes: binding to receptor and fusion between viral and cellular membranes. Previous attempts to generate chimeric MuLV envelope protein demonstrated that these modified proteins lack fusion function. Here, I have attempted to utilize a binding defective hemagglutinin (HAtmt) protein of influenza virus as a membrane fusion protein to complement the defect in fusion function in four chimeric MuLV envelope protiens: PC7Env, directed to the LDL receptor; D84KEP1, directed to the Her-4 receptor; and FIEnvBB and FIEnvSN, directed to the Flt-3 receptor. In this approach, the specific receptor binding activity would be carried by chimeric MuLV envelope proteins, whereas the HAtmt protein would mediate membrane fusion function. In order to generate MuLV vectors that contain both binding and fusion functions, appropriate chimeric MuLV proteins were co-expressed with the HAtmt protein to produce such vectors. The results showed that both chimeric MuLV envelope proteins and HA protein were incorporated into MuLV particles and that each of the two proteins was able to retain its complementary function. However, of these four combinations of co-expression (PC7Env + HAtmt; D84KEp1 + HAtmt; FIEnvBB + HAtmt; and FIEnvSN + HAtmt), only MuLV vectors containing the FIEnvSN and HA tmt proteins were able to enhance transduction efficiency on target cells that express the appropriate receptor. The data demonstrated that MuLV vectors Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 87 containing the FIEnvSN and HAtmt protein transudced 293A expressing mouse Flt-3 receptor (293A/Flt3) at approximately 10 times the efficiency of the parental 293A cells. These data consitute a proof of principle that the strategy of utilizing two different molecules to carrying binding and fusion processes is a feasible approach. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 88 4A.2. Introduction The ability of a retrovirus to integrate its genome into a host cellular genome attracts researchers to use it as a vehicle for gene delivery. To date, the most commonly used retroviral vectors for gene therapy protocols are MuLV retroviral vectors. However, they lack receptor-specificity, which therefore requires that the gene delivery procedure be done ex vivo. One approach to achieve cell-type specific gene delivery, which has received considerable attention in the past several years, is to regulate viral entry at the receptor binding step. Several strategies have been attempted to achieve this goal, mostly by engineering the Moloney MuLV envelope protein (reviewed in Anderson 1998; Cosset and Russell 1996; and Schnierle and Groner 1996). One of the earliest attempts to redirect binding of the MoMuLV envelope proteins to various receptors was by antibody bridging (Goud et al., 1988; and Etienne-Julan et al., 1992). Alternatively, redirecting the tropism of MuLV envelope protein by inserting a new binding moiety onto the envelope protein (internal insertion, N-terminus insertion and domain replacement), have also been attempted (Benedict et al., 1997; Cosset et al., 1995; Han et al., 1995; Kasahara et al., 1994; Marin et al., 1996; Martin et al., 1998; Nguyen et al., 1998; Valsesia-Wittmann et al., 1996; Yajima et al.,1998; and Zhao et al., 1999). Subsequently, another strategy employed to generate targeted retroviral MuLV envelope protein as well as to improve transduction Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 89 efficiency was a two-step inverse targeting approach in which a new binding moiety is added on to an amphotropic MuLV envelope protein with a protease activatable linker between the new binding moiety and the remaining envelope protein (Fielding et al., 1998; Morling et al., 1997; Nilson et al., 1996). In this approach, the attachment of two binding moieties to their targeted receptors compete with each other on the cell surface and the cleavage of the activatable linker by a specific protease to release one of the ligands allow the remaining intact natural envelope protein to transduce cells as normally. Although the feasibility of this approach has been demonstrated, the availability of a specific protease present at the surface of the target cell limits the general applicability of such a strategy. In general, many studies that had assessed the feasibility of specific receptor targeting have demonstrated that attachment to a targeted receptor by a modified envelope protein can be achieved. However, transduction of the target cells is very inefficient. Possible barriers to efficient transduction could be due either to defects in the modified envelope protein, problems caused by the choice of targeted receptor or both, which would all lead to a blockage at the post-binding step. Nevertheless, these experiments provided insight into the feasibility of targeting to a specific cell surface molecule. They also suggested that choice of receptors could play a critical role in conferring virus entry. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 90 Studies from our laboratory have identified a post-binding defect in chimeric MuLV envelope proteins, presumably caused by an inability to trigger the conformational change in the envelope protein presumed to be required to trigger fusion (Benedict et al., 1999; Empig et al., unpublished; Zhao et al., 1999). Furthermore, our laboratory also obtained data suggesting that the N-terminus of SU may be an important interactive domain between the SU and the TM subunits of the envelope protein, responsible for transmitting the signal to the TM subunit to trigger fusion (Empig and Burke, unpublished). This finding may explain why some of the chimeric envelope proteins displaying a new binding moiety at the N-terminus of SU (insertion between a.a. 6 and 7) were incapable of triggering the fusion process following binding to either ecotropic or targeted receptors (Empig et al., unpublished; Zhao et al., 1999), although it does not account for all the failures reported in the literature (Benedict et al., 1997; Cosset et al., 1995; Marin et al., 1996; Martin et al., 1998; Yajima et al.,1998; and Zhao et al., 1999). Among the modifications of the MuLV envelope proteins described above, insertion of a new binding moiety at the N-terminus of MuLV SU is currently the most effective approach. These modified envelope proteins are efficiently expressed on viral particles and are able to redirect virion binding to target cells. However, their poor efficiency in transducing target cells calls for Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 91 a solution to the problem of how to allow efficient fusion triggering after receptor binding. To circumvent the fusion problem identified in previous targeting approaches, I have devised a novel strategy to uncouple the binding and fusion functions. In contrast to MuLV envelope protein, in which binding and fusion are two related events, my approach is to separate the binding and fusion processes by co-expressing a fusion-defective chimeric MuLV envelope protein whose targeting receptor is known to be internalized upon ligand binding (Cristiano and Curiel 1996), with a binding-defective HA protein to function as a complementary fusion protein. Thus, the binding would be directed by the ligand on the chimeric MuLV envelope protein, and the fusion process would be carried out by HA following exposure to low pH in the endosome (Figure 4A.1.). For these studies, I chose four chimeric MuLV envelope proteins (PC7Env, D84KE01, FIEnvBB, and FIEnvSN), (Figure 4A.2.), directed to three different receptors: LDL, heregulin and Flt-3. PC7 Env, which displays a single chain antibody (scFv) between amino acids 15-16 of the MuLV envelope protein, is directed to the LDL receptor, which is known to go through an endocytic pathway (Brown et al., 1983; and Marsh and Pelchen- Matthews, 1994). Although PC7 Env has been reported to transduce target cells that express the LDL receptor (Somia et al., 1995), to date, the data remain controversial, and my laboratory as well as others have been unable Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Binding-defective HA Chimeric MuLV envelope protein internalization fusion by HA in tow pH endosome Target cell Figure 4A.1. Co-expression of a binding-defective HA with a fusion-defective chimeric MuLV envelope protein. The binding and fusion functions are performed by two separate molecules. The specific-receptor binding is carried out by a chimeric MuLV envelope protein, and the fusion is mediated by the binding-defective HA protein. o to Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Domain insertion: Sfi I Notl PC7Env D84KEp1 FIEnvSN CMV 617 632 I W ivSU'S'U: insertion between a.a. 6-7 ScFv a LDLR Ep1 a Her4 Fit a Flt3 receptor Domain replacement: BstE II Bam HI FIEnvBB CMV V / . • ' V ' Q 617 632 n Deletion between a.a. 17-227/replacement Figure 4A.2. Schematic diagram of chimeric MuLV envelope proteins. PC7Env contains a scFv directed to the LDL receptor. Eb1Env67 contains a EGF-like domain from the p isoform of heregulin directed to the heregulin receptor. FIEnvBB and FIEnvSN contain a Flt-3 ligand directed to the Flt-3 receptor. s O U > to repeat these observations (Cosset et al., personal communication; and Empig, thesis 1999). D84KEp1 is a chimeric MuLV envelope protein that contains the EGF-like domain from the p isoform of heregulin, also inserted between amino acids 6-7 (Empig, thesis 1999). Several reports have demonstrated viral particles displaying the EGF-like portion of the ligand interacts with Her-4 and is sufficient to trigger heregulin-mediated signaling (Cohen et al., 1996; Holmes et al., 1992; and Wen et al., 1994). Other reports have also demonstrated that a virus displaying this portion of the ligand or a single chain antibody directed to the Her-2 or Her-4 receptor was able to gain entry into target cells at low efficiency (Han et al., 1995; Poul et al., 1999; and Schnierle eta!., 1996; Sepp-Lorenzino et al., 1996). The FIEnvBB and FIEnvSN, kindly provided by Dr. Nori Kasahara, both display a mouse recombinant Flt-3 ligand on the MuLV envelope protien. FIEnvBB is a domain-replacement construct in which the binding domain in SU has been replaced with a Flt-3 ligand (a.a. 37-239), whereas the Flt-3 ligand is inserted between amino acids 6-7 in the FIEnvSN construct. The Flt-3 receptor has also been shown to be internalized rapidly after binding of soluble Flt-3 ligand (Lyman 1995, review; Turner et al., 1996). HA has been demonstrated to pseudotype RSV particles (Dong et al., 1992) as well as MuLV particles (Hatziioannou et al., 1998; Lin et al., unpublished). However, in order to utilize the HA protein solely as a complementary membrane fusion protein and to avoid the dual-receptor Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 95 binding activity contributed by chimeric MuLV envelope and HA protein, I have used a binding defective HA protein whose native receptor binding activity towards sialic acid has been impaired (Lin et al., unpublished). Knowledge gained from this work provides insight into understanding the mechanism causing the post-binding blockage in targeting receptors. 4A.3. Materials and Methods Cell lines. 293 are human kidney epithelial cells. 293/LDLR are 293 cells that stably express the LDL receptor at an intermediate level of expression. HT1080 are human sarcoma cells that express a high level of LDL receptor. NIH 3T3 is a murine fibroblast cell line. 3T3/Her4 cells (provided by K. Zhang, Amgen, CA) are derived from NIH 3T3 cells and stably express Her-4 receptor. The 293A cell line is derived from a sub-population of 293 cells, which are adapted to be more adhesive. 293A/Flt3, which stably expresses mouse Flt-3 receptor, is derived from the parental cell line 293A (provided by Dr. Nori Kasahara). All cell lines were cultured with DMEM containing 10% fetal bovine serum. In addition, 293/LDLR cells were cultured in the presence of 2.5 jig/ml of puromycin (Sigma, MO), and 293/Flt-3 cells were cultured in the presence of G418 (BRL Gibco, CA) at a concentration of 0.5 mg/ml. Plasmids and cloning. The pCgp plasmid expresses MuLV Gag-Pol. The pCnBG plasmid is a retroviral vector which expresses nuclear p-gal downstream of a CMV-MuLV LTR hybrid promoter and a neomycin resistant Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 96 gene from an internal SV40 promoter. D84KEnv carries a mutation in the MuLV envelope gene and lacks ecotropic receptor binding activity (McKrell et al., 1994). CEE67 plasmid contains the wild type MuLV envelope with unique Sfi I and Not I sites between residues 6 and 7 (Benedict et al., 1999). D84KEnv67 carries the Sfi I and Not I site between residues 6 and 7, and the D84K mutation (Zhao et al., 1999). The plasmid pCEEAEnv contains the backbone of CEE+ plasmid (which contains the wild type MuLV envelope gene) and a CMV promoter, but not the envelope gene. The plasmid pHIT123 contains MuLV envelope gene driven by a CMV promoter (Soneoka et al., 1995). The pHITAEnv plasmid contains the backbone of pHIT123 and a CMV promoter, but not the envelope gene. The pC7Env plasmid was obtained form Dr. Verma (Salk Institute, CA). It contains a scFv directed to the LDL receptor, inserted between amino acids 6 and 7 in the MuLV envelope protein. D84KEJ31 was constructed by insertion of a 63-amino acid EGF-like domain (residues 177-239) from the ( 3 isoform of heregulin ligand and inserted into D84KEnv67 between amino acids 6 and 7 (Sfi I and Not I sites) in D84KEnv67 backbone. The FIEnvSN plasmid (provided by Dr. Kasahara) contains a mouse recombinant Flt-3 domain (residues 28-163) directed to the Flt-3 receptor, and the ligand was inserted between residues 6 and 7 in the CEE67 backbone. The FIEnvBB plasmid contains the same mouse recombinant Flt-3 domain, and the ligand Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 97 was inserted between the BstE II and BamH I sites in the pHIT123 backbone to replace the receptor-binding domain of SU. The HAtmt plasmid is a triple mutations HA molecule constructed by Amy Lin, as described in Chapter 2. HAfd is a fusion-defective HA as described in Chapter 3. Transient transfection, transduction and determination of viral titer. Pseudotyped MuLV viral particles were produced by transient transfection using a calcium phosphate precipitation method, essentially as described (Soneoka et al., 1995). Sixty to seventy percent confluent 293T cells in 60 mm plates were co-transfected with 5 pg of each the pCgp and pCnBG plasmids, together with 5 pg of plasmid expressing MuLV Env (pC7Env, D84KEp1, FIEnvBB, or FIEnvSN) and/or 1 pg of plasmid expressing HA (pCMVHAtmt or pCMVHAfd). To make sure each sample contains the same amount of DNA to avoid deviations in transfection efficiency and promoter competition, pCEEAEnv plasmid and pHITAEnv were used as the filler DNA, so that a total of 16 pg of DNA was included in each transfection sample. At 15 to 17-hrs post transfection, the medium was replaced with 3 ml of fresh medium. The viral supernatants were harvested 24 hrs later, filtered through a 0.45 pm filter to remove cell debris and aliquots were stored at -80°C. Viral titer was analyzed by transduction of target cells which were plated onto six-well culture dishes the day before transduction at a density of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 98 1 x 105 cell per well. Serial dilutions of the original viral supernatant were prepared, and 1 ml of the diluted viral supernatant with 8 (j.g/ml polybrene (Sigma, MO) was added to each well, followed by 10-12 hrs incubation. Fresh media was then applied and viral titer was determined by X-gal staining at 48 hours post transduction, at which time transduced cells were fixed with 0.5 % of glutaraldehyde PBS buffer for 10 min, followed by two subsequent PBS washes of 10 min for each. Cells were then incubated with staining solution (4 mM potassium ferricyanide, 4 mM potassium ferrocyanide, 2 mM MgCfe, 0.4 mg/ml X-gal, Sigma, MO) at 37°C over night. Titer was determined by counting the number of blue colonies under a microscope and multiplying by the appropriate dilution factor. Ligand competition for binding of FIEnvSN to Flt-3 receptor by soluble Flt-3 ligand. Prior to transduction of FIEnvSN-expressing pseudotyped MuLV viral particles, 293A and 293A/Flt3 cells were incubated with soluble recombinant mouse Flt-3 ligand (R & D Systems, MN) at 10 ng/ml for 10 min at 37°C followed by the usual transduction procedure. Target cells were incubated with 10 ng/ml of the soluble mFlt-3 ligand during the transduction process until the viral supernatant was removed. Virion incorporation. One ml of filtered viral supernatant was purified through a 20% sucrose cushion by centrifugation at 14,000 r.p.m. (eppendorf centrifuge 5145C) for 30 min at 4°C. Due to the similar molecular weights of the HA2 subunit (~p27 kDa) of HA and the capsid protein (~30 kDa) of the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 99 MuLV viral particles, the resulting pellet was then re-suspended in 2 x loading buffer (Novex, CA) split into two halves and subjected to SDS-PAGE. For detection of HA proteins, one blot was probed with anti-FPV HA as the primary antibody (1:5000) and HRP-conjugated goat anti-rabbit IgG as secondary antibody (1:10,000), (Pierce, IL). For detection of MuLV capsid protein (CA) and MuLV Env protein, the second blot was incubated with primary polyclonal antibodies 78S-221 (Quality Biotech, NJ.) at 1:10,000 dilution and anti-SU (anti-Rauscher MuLV), (Quality Biotech, NJ.) at 1:3,000 dilution. HRP-conjugated rabbit anti-goat IgG was used as secondary antibody (Pierce, IL) at 1:10,000 dilution. Protein signals were detected by an ECL kit (Amersham, CA). Virus-receptor binding. One ml of filtered viral supernatant was mixed with 1 x 105 target cells and incubated for 2 hrs at 4°C to prevent receptor internalization. At the end of incubation, each sample was washed with 1 ml of wash buffer once. Following the wash, the bound viral particles were incubated with 200 pi of rat monoclonal anti-SU antibody, 83A25 (a stock of hybridoma supernatant from Dr. Evans, Rocky Mountain Laboratory, MT) for 1 hr at 4°C followed by 1 ml wash with wash buffer. FITC-conjugated goat anti-rat IgG was used as secondary antibody at 1:100 dilution and incubated with the cells for another 1 hr at 4°C as described for incubation with primary antibody. After washing, the cells were fixed with 4% paraformaldehyde and subjected to FACS analysis to measure the receptor binding activity. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 The HT1080 cell line, which expresses the LDL receptor, was used as the target cells for binding of PC7Env-pseudotyped MuLV viral particles. An NIH 3T3 stable cell line expressing Her-4 receptor (3T3/Her4) was used as the target cells for binding of Ep1Env67-pseudotyped MuLV viral particles, whereas the 293A/Flt3 was used as the target cells for binding of FIEnvBB- and FIEnvSN-containing pseudotyped MuLV viral particles, respectively. Polykaryon formation ofNIH3T3 cells. 8 x 104 NIH 3T3 cells were plated on each well of a 6-well plate the night before transfection. The next day cells were transfected alone with 0.1 pg of pCMVHAtmt or 0.5 pg of pC7Env, D84KEP1, FIEnvBB or FIEnvSN. For co-expression of chimeric MuLV envelope protein and the HAtmt protein, cells were co-transfected with 0.1 pg of pCMVHA and 0.5 pg of PC7Env, D84KEP1, FIEnvBB or pFIEnvSN. The total amount of DNA in each sample was 1 pg, using pBlueScript plasmid (Strategene, CA) as the filler DNA. Transfection of NIH 3T3 cells was performed using Lipofectamine Plus Reagent Kit, and transfection procedures were earned out as described in the standard transfection protocol (Gibco BRL, CA). Thirty-six hrs post transfection, cells were washed once with room temperature PBS followed by 3 min incubation of 1 ml pH 5 PBS buffer (10 mM HEPES, 10 mM 2-[N- morpholino]ethanesulfonic acid, Sigma, MO) at 37°C. Cells were then washed twice with 2 ml of room temperature DMEM, replaced with 2 ml of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 101 DMEM containing 10% fetal bovine serum and incubated at 37°C for 5 hrs to allow cell-cell fusion to occur. At the end of the incubation, the cells were fixed with 0.5% glutaraldehyde for 10 min at room temperature and stained with 0.1% methylene blue at room temperature for 10 min. Polykaryons (syncytia) were counted using a light microscope at 40x magnification. A polykaryon was defined as a cell mass containing 4 or more nuclei. 4.A.4. Results MuLV Env and HAtmt proteins compete with each other for virion incorporation into MuLV particles. It has been shown that MuLV viral particles can be pseudotyped efficiently by HA protein (Haziioannu et al., 1998, Lin et al., unpublished). In order for the chimeric MuLV envelope protein and the HAtmt protein to each carry out its function to gain entry into target cells, I addressed the question whether MuLV envelope protein and the HA protein could both be incorporated into MuLV particles. Accordingly, co expression of various combinations of MuLV envelope protein and the HA protein was carried out. As shown in Figure 4A.3., all MuLV envelope proteins (D84KEnv, PC7Env, D84KEP1, FIEnvBB, or FIEnvSN) and the HAtmt fusion proteins were incorporated into MuLV particles. However, there seems to be a competition between the two proteins for virion incorporation, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 0 2 L L I > »UJJVH + NSAU3U N S A U 3 U P W H + 9 Q A U 3 id S8AU3id ifa y iP B Q IUJ1VH + ZOd ZOd IUQVH+ AU3**8C] A U 3 ^ 8 Q < O > c Ul > ■ 8 •» < a i 8 liu iV H + N S A U 3 U NSAU3ld P W H + QQAU3IJ g s A U 3 u ;u i)V H + ifa W B Q I-03**80 liu;vH + ZOd ZOd P W H + AU3>«fr8a ; u i» v h < x i 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 4A. 3 . Co-expression o f chimeric M u L V envelope protein a n d HAtm t protein. HT1080 NIH 3T3/Her4 PC7Env HAtmt PC7Env + HAtmt D84KEp1 D84KEp1 + HAtmt TTr F L l FL 1 Figure 4A.4. Receptor binding activity of MuLV viral particles containing a chimeric MuLV envelope protein and the HAtmt protein. A. Co-expression of PC7Env + HAtmt. B. Co-expression of EpiEnv + HAtmt. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. 293A/Flt3 293A/Flt3 HAtmt HAtmt FIEnvSN . FIEnvBB FIEnvBB + HAtmt FIEnvSN + HAtmt 1 • TTTf 10 10 1 0 * i 10 10 FL 1 Figure 4A.5. Receptor binding activity of MuLV viral particles containing a chimeric MuLV envelope protein and the HAtmt protein. A. Co-expression of FIEnvBB + HAtmt. B. Co-expression of FIEnvSN + HAtmt. 104 105 particularly, for co-expression of FIEnvBB and HAtmt proteins in which a marked decrease of both proteins was observed. The co-expression of the two proteins does not seem to inhibit viral production, as there was no difference in levels of capsid signal. The presence of the HAtmt protein does not affect the receptor binding activity of chimeric MuLV envelope protein. To examine whether co expression of MuLV chimeric envelope protein and HAtmt fusion protein could interfere with the binding of MuLV chimeric envelope protein to its targeted receptor, the receptor binding activity of co-pseudotyped MuLV viral particles was measured by FACS analysis. Using antibody specifically for the MuLV envelope protein, the histograms shown in Figures 4A.4. demonstrated that both PC7Env and D84KE|31 bind specifically to their cognate receptors, and that co-expression of the HAtmt protein did not interfere with the receptor binding activity of these chimeric MuLV envelope proteins. Likewise, FIEnvBB and FIEnvSN , which bind specificity to the Flt-3 receptor, were also able to retain their receptor binding activity in the presence of co-expressed HAtmt protein (Figure 4A.5.). Presence of chimeric MuLV envelope protein does not interfere with the membrane fusion function o f the HAtmt protein. Since HAtmt is used as a complementary membrane fusion protein in this targeting strategy, I asked the question whether co-expression of MuLV chimeric envelope protein and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. HAtmt fusion protein would interfere with the membrane fusion function of HAtmt protein. In order to do this, I performed cell-cell fusion assays. The data revealed that the HAtmt protein was able to retain its membrane fusion function in the presence of any of the three MuLV chimeric envelope proteins tested (Table 4A.1.). The inclusion of chimeric MuLV Env alone serves as a negative control in the experiment, and the HA wild type and HAtmt alone were included as positive controls. The results showed that none of the chimeric MuLV envelope proteins was able to cause cell-cell membrane fusion at pH 5 in the absence of the HAtmt protein, whereas inclusion of the HAtmt protein allowed approximately the same efficiency of cell-cell membrane fusion as the HA wild type or HAtmt protein alone. Only MuLV particles expressing the FIEnvSN and HAtmt were able to enhance transduction efficiency on target cells. Results obtained from the immunoblot, receptor binding and cell-cell fusion assays, indicated that all four chimeric MuLV envelope proteins could co-pseudotype with the HAtmt protein to produce MuLV particles. Furthermore, each of the two proteins retained its complementary function, i. e. specific receptor binding activity by a chimeric MuLV envelope protein and membrane fusion function by the HAtmt protein. I then assayed the transduction efficiency of these co pseudotyped MuLV viral particles on cell lines that expressed receptors to which each chimeric MuLV envelope protein binds. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4A. 1. HAtmt retains its fusion function when co-expressed with chimeric MuLV Env Protein Polykaryon formation of NIH 3T3 cells3 mock H A wild type + + + HAtmt + + + PC7Env — PC7Env + HAtmt + + + D 8 4 K E f3 1 — D 84K EP1 + HAtmt + + + FLEnvBB — FLEnvBB + HAtmt + + + FIEnvSN — FIEnvSN + HAtmt + + + a - + + +, wild type HA syncytia; no syncytium O 108 As a negative control, MuLV viral particles pseudotyped with a chimeric envelope protein alone were included in each experiment. Since the HAtmt protein retains minimal receptor binding activity toward sialic acid residues (Chapter 2), MuLV particles pseudotyped with the HAtmt protein alone were also included in each experiment for background titer measurement. Another control included in the transduction assay is the co expression of a non-targeted MuLV envelope protein, D84KEnv, with the HAtmt protein. The reason to include this control in the experiment is that the immunoblot results showed that MuLV envelope protein and HAtmt protein competed with each other for virion incorporation (Figure 4A.3.). Therefore, inclusion of this control ensures that co-expression of a MuLV-based protein and the HAtmt protein did not affect titer of HAtmt. For co-expression of chimeric PC7Env and HAtmt proteins, transduction efficiency of MuLV viral particles pseudotyped with these two proteins was examined on several cell lines that express different levels of the LDL receptor. FACS analysis showed that cell surface expression levels were lowest on 293 cells, intermediate on 293/LDLR and highest on HT1080 cells (data not shown). As shown in Figure 4A.6., there was no significant difference in titer given by MuLV viral particles pseudotyped with HAtmt and those by HAtmt and D84KEnv. When comparing the titer given by MuLV viral Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 109 particles pseudotyped with PC7Env versus that given by PC7Env and HAtmt on target cells, no enhancement of titer was observed on any of the three cell lines tested. To assess whether choice of receptor could be an important factor in determining the enhancement of transduction efficiency on target cells, another combination of MuLV-pseudotyped viral particles I assayed was the co-expression of the D84KE01 and the HAtmt. However, the results obtained from this experiment were similar to those obtained from co-expression of the PC7Env HAtmt, as co-expression of D84KE{31 and HAtmt failed to enhance titer on the 3T3/Her4 cell line which stably expresses the Her-4 receptor (Figure 4A.6.) I next examined MuLV viral particles containing both the HAtmt protein and a chimeric MuLV envelope protein displaying the Flt3 ligand (FIEnvBB or FIEnvSN). Co-expression of FIEnvBB and HAtmt pseudotyped MuLV viral particles greatly reduced the transduction efficiency on both 293A and 293A/Flt3 cell lines. This inhibitory effect was caused by the FIEnvBB protein because the reduction in titer was observed only in the presence of FIEnvBB and not HAtmt alone. In contrast, the co-expression of FIEnvSN and HAtmt pseudotyped MuLV viral particles conferred an increase of a log in the titer on 293A/Flt3 cells but not on 293A cells (Figure 4A.7.). These data Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. E 2 t m a > *-» I M 1.00E+02 8.00E+01 6.00E+01 4.00E+01 2.00E+01 0.00E+00 D84KEnv HAtmt D84KEnv + PC7Env PC7Env + HAtmt HAtmt ■ 293T □ 293/LDLR ■ HT1080 B 4.00E+01 n 3.00E+01 - E 2.00E+01 - 2 o 1.00E+01 k . a > + ■ » 0.00E+00 - D84KEnv HAtmt D84KEnv EblEnv Eb1Env + + HAtmt HAtmt ■ NIH 3T3 □ NIH 3T3/Her4 Figure 4A.6. Transduction of target cells by MuLV viral particles containing a chimeric MuLV envelope protein and the HAtmt protein. A. Co-expression of PC7Env and HAtmt. B. Co-expression of D84KEp1 and HAtmt. 110 Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. E £ k. £ ■ MB I - 1.00E+04 1.00E+03 1.00E+02 1.00E+01 1.00E+00 293A 293A/Flt3 •k Significant, P < 0.01 Figure 4A.7. Transduction of 293A/FU3 cells by MuLV viral particles containing chimeric MuLV envelope proteins and the HAtmt protein targeted to Flt-3 receptor- expressing cells. 1 1 2 demonstrated that a co-expression of two heterologous proteins, each carrying a different function, could complement each other in trans to allow receptor-specific gene delivery. Soluble Flt-3 ligand is able to compete with FIEnvSN for receptor binding. To confirm the enhancement of transduction efficiency observed for FIEnvSN and HAtmt pseudotyped MuLV viral particles, soluble Flt-3 ligand was included in the transduction assay to compete with FIEnvSN for Flt-3 receptor binding. As shown in Figure 4A.8., MuLV viral particles co expressing FIEnvSN and HAtmt protein failed to enhance titer on 239A/Flt3 cells pre-incubated with soluble Flt-3 ligand. On the other hand, the background titer on 293A cells was unaffected by the presence of the soluble Flt-3 ligand. As expected, titer given by MuLV viral particles co-expressing the D84KEnv and HAtmt on 293A and 293A/Flt3 cells, respectively, were also unaffected by the presence of the Flt-3 ligand. Co-expression of FIEnvSN and HAfd fail to enhance transduction efficiency on 293A/Flt3 cells. An additional control experiment examined the requirement for HA fusion function in this targeting strategy was performed. MuLV particles were co-pseudotyped with FIEnvSN and a fusion- defective HA protein (Vey et al., 1992). The results showed that MuLV particles co-pseudotyped with FIEnvSN envelope protein and HAfd protein gave no enhancement of titer on either 293A or 293A/Flt3 cells (Figure 4A.9.). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. 1.00E+04 1.00E+03 1.00E+02 1.00E+01 1.00E+00 ■ 293A □ 293A+FL m 293A/F!t3 □ 293A/Flt3 + FL D84KEnv + HAtmt FlenvSN + HAtmt * Significant, P < 0.05 Figure 4A.8. Ligand competition by soluble Flt-3 ligand (FL) to prevent enhancement of transduction efficiency on 293A/Flt3 cells. 113 Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. £ 3 4- O, f t . < D 4* ■ M B H 1.00E+04 1.00E+03 1.00E+02 1.00E+01 1.00E+00 H293A H 293A/F It3 D84KEnv+ FlenvSN+ D84KEnv + FlenvSN + HAtmt HAtmt HAfd HAfd * Significant, P < 0.01 Figure 4A.9. Co-expression FIEnvSN chimeric MuLV envelope protein and a fusion-defective HA protein (HAfd) prevents enhancement of transduction efficiency on 293A/Flt3 cells. 114 115 4A.5. Discussion in my study, I utilized the membrane fusion function of the HA protein to complement chimeric MuLV envelope proteins that displayed a ligand directed to a specific receptor. In this approach, the binding and fusion processes of entry were separated and performed by two different molecules. A similar approach had been attempted and discussed by Patterson et al., in 1999, but the approach they took and the results they obtained were different from mine. In their study, an single chain antiboy directed to hapten nitrophenyl acetic acid was inserted between residues 139 and 145 in the HAt subunit of an H3 subtype HA protein. Although this chimeric HA protein was shown to bind specifically to its targeted receptor, it was fusion defective and required a co-expression of the wild type HA protein to exert the fusion function. This approach raised at least two concerns. First, the use a HA protein whose native receptor binding activity is retained, and the requirement of the wild type HA protein to co-express with the chimeric HA protien in order to mediate fusion introduce a dual-receptor binding activity (via the scFv and the sialic acid binding pocket). As a consequence, it compromises the receptor-specificity ability earned out by the chimeric HA protein. Second, the HA protein used in the study was an H3 subtype which is not cleaved endogenously, thus requiring an exogenous trypsin treatment to cleave the HA precursor into its two subunits, HAi and HA2 . This requirement for exogenous trypsin treatment not only adds an additional procedure in the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 116 experiment, but may possibly inactivate the receptor binding ability of new ligand in the HA molecule. On the other hand, in my approach, I took advantage of the fact that MuLV viral particles can be co-pseudotyped with both the MuLV envelope and HAtmt proteins and used these two different proteins to complement each other to gain entry into target cells. MuLV viral particles displaying the FIEnvSN and HAtmt proteins bound to 293A/Flt3 cells and enhanced transduction efficiency on 293A/Flt3 cells by 10-fold. This 10-fold increase of titer observed on target cells was further shown to be Flt-3 receptor-specific, as inclusion of soluble Flt-3 ligand blocked the enhancement of transduction efficiency on Flt-3-expressing cells. Futhermore, the use of a fusion-defective HA protein also prevented the enhancement of tranduction efficiency. Taken together, these data demonstrate that an HA co-expression approach to targeting is feasible. The mechanism by which co-expression of FIEnvSN and HAtmt proteins enhances transduction efficiency is unclear. An intriguing question that arises from the observations made in this study is why there is only an one-log increase of titer on target cells. Where could the rate-limiting step be that prevents MuLV particles, which carry both the receptor binding and membrane fusion functions, from giving a higher level of transduction efficiency. Although the cell-cell fusion data indicated that the HAtmt protein remains fusion-competent in the presence of chimeric MuLV envelope proteins, it may not truly reflect the interactions between Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 117 virus-cell fusion; the density of the two proteins on the cell surface could be different from the viral surface. It has been shown that the density of the HA protein on membrane surface plays a role in fusion (Gunther-Ausbom et a!., 2000). The data obtained from the cell-cell fusion assay might not reflect the ratio of the density of the two proteins in an endosomal compartment. Therefore, it is possible that the density of the HA proteins in the endosome in the presence of MuLV chimeric envelope protein does not allow optimal fusion. Alternatively, the rate-limiting step could still be at the internalization step, which was suggested by that fact that not all receptors can allow efficient viral entry. Toward that end, it is reasonable to postulate that the lack of significant enhancement on titer given by MuLV viral particles displaying FIEnvSN and HAtmt proteins could be due to an inefficient internalization rate of the Flt-3 receptor. Furthermore, since I do not have data to indicate that these MuLV particles were actually internalized after binding to their cognate receptor, it is possible that the receptor binding affinity given by the new ligands was too weak compared to the receptor's natural ligand, and therefore could not cause an efficient intemailzation of the receptor. This explanation would also be applicable to the chimeric MuLV envelope protein PC7Env, in which a scFv directed to the LDLR receptor was use as the ligand, and D84KEp1, in which a minimal receptor binding doamin directed to the Her4 receptor was used as the ligand. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 118 In conclusion, to improve the transduction efficiency of the chimeric MuLV viral particles that possess both the receptor binding and membrane fusion functions, a better understanding of the trafficking pathway of targeted receptors of interest as well as the interaction of ligand-receptor is needed to advance the utilization of this targeting strategy. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 119 C hapter 4 B Modification of the HA protein for targeting Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 2 0 4B.1. Abstract Using the genetic engineering approach previously employed to generate chimeric MuLV envelope proteins for receptor-specific targeting, I generated a targeted HA protein that possessed specific receptor binding activity while retaining its membrane fusion function. Two approaches for modifying the receptor-binding specificity of the HA protein have been explored. In one approach, a dimeric apoE peptide and a small RGD- containing peptide were incorporated into five potential insertion sites on the HA molecule. The data revealed that none of the five sites was able to tolerate a dimeric apoE peptide insertion, but the RGD-containing peptide was tolerated between residues 135-136 of the HAi subunit. The resulting HA protein remained fusogenic. Subsquently, the RGD-containing peptide was inserted into three different binding-defective mutants of HA at the same position in order to avoid dual-sialic acid binding activity. All three chimeric HA proteins ( H A - R G D y 98f / e i 9o q , H A - R G D y 98f / s 228k . and HA- RGDy98f/ei9 o q /s 2 2 8 k ) were found to be processed correctly but incorporated into MuLV viral particles less efficiently than their parentals. In contrast to the HA-RGD backbone, the RGD peptide insertion into the backbone of these three binding defective mutants resulted in a hypo-fusogenic phenotype. The defect in fusion was rescued when they were expressed at 32°C. However, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 121 none of the chimeric HA proteins displaying the RGD peptide was able to bind to CHO cells expressing the integrin receptor anbfb to confer enhancement of titer. In another approach, a large ligand was inserted at the N-terminus of the HA1 subunit of a binding-defective mutant HA (HAtmt), separated by a spacer to allow independent folding of the new binding moiety and the remaining intact protein. The ligands used included apoE, for targeting to the LDL receptor and heregulin, for targeting to the Her-4 receptor. The heregulin N-terminus insertion impaired processing of the HAtmt protein, but the apoE N-terminus insertion was correctly processed and was shown to possess LDL receptor binding acitivity and remained fusion competent. Taken together, these data strongly suggest that the generation of a chimeric HA protein that possesses a new receptor binding function without compromising its fusion function is feasible. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 122 4B.2. Introduction Attempts to modify the envelope protein of Moloney murine leukemia virus (MoMuLV) to target a specific receptor have had very limited success (Anderson 1998; and Cosset and Russell, 1996). Although a receptor- specific interaction between chimeric envelope proteins and their cognate receptors have been achieved, the transduction efficiency of such chimeric MuLV-based retroviral vectors was much lower than the wild type. As mentioned in the previous section (Chapter 4A), the barrier to targeted retroviral vectors appears to be at a post-binding step, and my laboratory has shown evidence that the defect is at the fusion step (Benedict et a., 1999; Empig & Burke, unpublished; Zhao etal., 1999). In Chapter 4A, I suggested a possible solution to overcome the defect in fusion was to use the HA protein as a complementary membrane fusion protein together with a receptor- targeted chimeric MuLV envelope protein as a binding protein. Toward this end, I have attempted another strategy to overcome the fusion problem- that is to directly modify the HA protein itself by inserting a new binding moiety into the HA protein. This strategy is similar to the modifications previously made in the MoMuLV envelope proteins (Benedict et al., 1999; Cosset et al.,1995; Martin et al., 1998; Marin et al., 1996; Valsesia-Wittmann, et al., 1996; and Zhao et al., 1999). There are two parts to this strategy: first, I explored the possibility of finding an internal site in the HA molecule that could tolerate a small peptide Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 123 insertion; second, I explored the possibility of whether the large ligands could be inserted at the N-terminus of the HAi subunit of the HA molecule. In contrast to the lack of adequate structural information available regarding the 3-D structure of the MoMuLV envelope protein, the X-ray crystal structure of the HA protein of an H3 subtype was solved in early the 1980s (Wilson et al., 1981). The reason I did not chose this particular HA protein in my study is that this HA protein can not be cleaved endogenously, whereas the HA protein (H7 subtype) that I used can. Based on the structural information of the HA protein of the H3 subtype (Wilson et al., 1981), I performed a sequence alignment between the HAs of the H3 and H7 subtypes (Figure 4B.1.). The sequence alignment showed that there is an approximately 70% homology in the HAi subunit between the H3 and H7 subtypes. In particular, the cysteine residues and the residues which form the receptor binding site are conserved in both subtypes. Also, the HA2 subunit amino acid sequence between the two subtypes is highly conserved (not shown). Comparison of the sequences from the two subtypes suggested three surface loops to be present in the HA-i subunit of the H7 subtype. Among these three loops, five potential insertion sites were chosen for ligand insertion. These five sites are between residues 129-130,135-136,142-143, 155-156, and 160-161, respectively. In this study, I chose two small ligands to insert into those five sites. One is a dimeric form of an apoE ligand which Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. (H3) IFCLALGQDLPGNDN-STATLCLGHHAVPN (H7) MNTQILVFALVAVIPTNADKICLGHHAVSN GTLVKTITDDQIEVTNATELVQSSSTGKIC GTKVNTLTERGVEVVNATETVERTNIPKIC N-NPHRILDGIDCTLIDALLGDPHCD-VFQ S-KGKRTTDLGQCGLLGTITGPPQCD-QFL NETWDLFVERSKAF-SNCYPYDVPDYASLR EFSADLIIERREGN-DVCYPGKFVNEEALR SLVASSGTLEFI QILRGSGGIDKE KgarngMSHBNWL ftfljjfiSMHKjLNVT MPNNDNFDKLYIWGIHHPSTNQEQTSLYVQ YKNTRRESALIVWGIHHSGSTTEQTKLYGS ASGRVTVSTRRSQQTIIPNIGSRPWVRGLS GNKLITVGSSKYHQSFVPSPGTRPQINGQS SRISIYWTIVKPGDVLVINSNGNLIAPRGY GRIDFHWLILDPNDTVTFSFNGAFXAPNRA FKM-RTGKSSIMRSDAPIDTCISECITPNG SFL-RGKSMGIQSDVQVDANCEE-CYHSGG SIPNDKPFQNVNKITYGACPKYVKQNTLKLATGMRNVPEKQTR TITSRLPFQNINSRAVGKCPRYVKQESLLLATGMKNVPEPSKKRKKR Figure 4B.1. Amino acid sequence alignment of the HA1 subunit of HA proteins between the H3 and H7 subtypes. Highlighted are residues that form exposed loops on the surface of the H3 HA molecule. Underlined are conserved cysteine residues in both the H3 and H7 subtypes. Diagram shown on the right is the crystal structure of the monomeric HAt subunit of the H3 subtype HA. 124 125 containing peptide directed to the anb p3 integrin receptor (O'Neil et al., 1992). Both peptides have been shown to have receptor binding affinity with dissociation constants in the nanomolar range. From the sequence alignment of the different subtypes of HAs, it appears that the N-terminus of the HAi subunit of all HA subtypes does not seem to have a defined secondary structure, and therefore, I have attempted to insert a large ligand at the N-terminus of the HAt subunit. I chose two ligands insert at the N-terminus of the HA1 subunit. One is the apoE ligand (residues 34-329) directed to the LDL receptor. The other is the heregulin ligand (residue 33-326) directed to the heregulin receptor. I also used a spacer to improve the chances of the ligand and the HA protein able to fold independently and maintain the correct structure of each protein. Since there are two parts to this study, I will first discuss the insertion of the small ligand/peptide in the five potential insertion sites in the HAi subunit of the H7 HA molecule (Part 1). Then the second part will include the large ligand inserted at the N-terminus of the subunit of the HA molecule (Part 2). 4B.3. Materials and Methods Cell lines. 293T, 293/LDLR, and HT1080 cells were cultured in Dulbecco's modified Eagle's medium (DMEM), (Cell Culture Core Facility, University of Southern California, CA) with 10% fetal bovine serum (FBS), (Hyclone, Logan Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 126 UT) added. 293/LDLR cells are cultured in the presence of 2.5 pg/ml of puromycin (Sigma, MO). 293T cells express a low level of the LDL receptor. 293/LDLR are 293 cells that stably express the LDL receptor at an intermediate level. HT1080 cells express a high level of the LDL receptor. The CHO cell line (provided by Dr. Ginsberg at Scripps Institute, CA) was cultured in DMEM with non-essential amino acids (Sigma, CA) and 10% FBS added. The CHO/anbP3 cell line that stably expresses the integrin anbP 3 receptor (OToole et al., 1989) was cultured in the same medium as the CHO cell line with the addition of 0.7 mg/ml of G418 (BRL Gibco, CA). Mutagenesis to introduce unique restriction enzyme sites into the HA backbone for peptide insertion. The HA(Y98F/E190Q), HA(Y98F/S228K), and HA(Y98F/E190Q/S228K) plasmids were generated by splice overlap PCR (soPCR), (Ho et al., 1989) with mutant primers as described in Chapter 2. The triple mutation of Y98F/E190Q/S228K is designated as HAtmt. In order to insert a peptide into the five potential insertion sites identified on the surface of the HA molecule, a fragment of HA (Xba l-Xba I) containing the natural Spe I site from the pCMVHA (H7 subtype FPV HA, see Chapter 2) plasmid was subcloned into the pGEM7-7Zf plasmid (Promega) at the Xba I site to create pGEMAHAi3 6-i3 7- Mutagenesis by soPCR was performed to introduce an unique restriction enzyme site at each potential site except for a natural Spe I site between residue 136-137 (Figure 4B.2a.). An Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. . Further reproduction prohibited without permission. 1st-round PCR products splice overlapping PCR Sphi ^ Stul 2nd-round PCR products I I Figure 4B.2a. Cloning scheme for RGD peptide insertion. Unique restriction enzyme sites were introduced into pGEMAHA by splice overlap PCR. Arrows indicate primers used for each PCR reaction.^ indicates mutations incorporated in the primers and the resulting PCR products. to 128 Afl II site was introduced between residues 130-131 to produce PGEMAHA130- 1 3 1; a Hpa I site was introduced between residues 155-156 to produce pGEMAHAi55-i5 6> ' a Nhe I site was introduced between residues 142-143 and between residues 160-161 to produce PGEMAHA1 4 2 -1 4 3 and pGEMAHAi6 o-i6i, respectively (Figures 4B.2b, 4B.2c, and 4B.2d.). For insertion of the dimeric apoE peptide (LRKLRKRLLRDADDL)2 , a pair of oligomers encoding the dimeric apoE peptide sequence was generated with a Spe I site flanking each end (Core Facility, University of California, CA). The sense sequence of the peptide is: 5'-P04 CTAGTCTCAGAAAGCTTAGAAAGAGACTCCTCAGGGATGCAGACGA TCTATTACGAAAATTGCGAAAACGCTTACTCCGCGACGCTGATGACTT GA-3', and the anti-sense sequence of the peptide is: 5'-P04 CTAGTCAA GTCATCAGCGTCGCGGAGTAAGCGTTTTCGCAATTTTCGTAATAGA TCGTCTGCATCCCTGAGGAGTCTCTTTCTAAGCTTTCTGAGA-3'. The sequence of Spe I site is underlined. The two complementary oligomers were hybridized, purified and inserted into the pGEMAHAi36-137 plasmid. Subsequently, this plasmid was used as a template to amplify the dimeric apoE peptide DNA sequence by the PCR method, using primers with an Alf II, Nhe I or Hpa I restriction enzyme site flanking each end of the primer site to generate a dimeric apoE peptide DNA sequence with a unique restriction enzyme site. Each resulting PCR product was then cloned separately into the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. pGEM AHA(135-136)apoE Sphl Xbal Peptide DNA sequence Synthesis of apoE peptide sequence using pGEMAHA(135-136)apoE as a template Afl II -Afl II Nhel ■Nhel Hpal -Hpal Figure 4B.2b. Synthesis of RGD peptide ligand by PCR. DNA sequence of apoE peptide with a flanking Spe I site at each end was initially cloned into pGEMDHA at the Spe I site. The resulting clone was used as a template to amplify apoE peptide DNA sequence flanked with Alf II, Nhe I or Hpa I restriction enzyme sites, respectively. Each PCR fragment was cloned into pGEMAHA at the corresponding restriction enzyme site. 129 pGEMAHAapoE142-143 pGEM AH AapoEl 5S-156 pGEMAHAapoE160-161 Xbal Xbal Clone into a CMV-driven expression vector at Xbal site Figure 4B.2c. Cloning of HAapoE-expression plasmids. Insertion of Xbal-Xbal HA fragment into a CMV-driven expression plasmid to produce pHAapoE130-131, pHAapoE135-136, pHAapoE142-143, pHAapoE155-156 and pHAapoE160-161, respectively. Solid box indicates the dimeric apoE peptide. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. 142-143 (Nhel) 135-136 (Spel) 160-161 (Nhel) 155-156 (Hpal) 129-130 (Afl II) Figure 4B.2d. Potential insertion sites on the surface of the HA molecule. Crystal structure of a monomeric HA1 subunit of the H3 subtype HA. Arrows indicate sites at which restriction enzyme sites were introduced for peptide insertion. 132 pGEMAHA plasmid at each corresponding restriction enzyme site. All fragments generated by PCR products were sequenced to ensure that there were no additional mutations being introduced. Following each subcloning, the Xba I - Xba I HA fragment now containing the dimeric apoE peptide sequence was then inserted into the pCMVAHA plasmid at the Xba I site. The HA molecule containing the apoE peptide at the Afl II site was designated as pHAapoE130-131; the HA molecule containing the apoE peptide at Nhe I site between residues 142-143 and between residues 160-161 were designated as pHAapoE1 4 2 -i4 3 and pHAapoE-i6 o-i6-i, respectively; and the HA molecule containing the apoE peptide at Hpa I site was designated as p H A a p o E -1 5 5 - 1 5 6 . For the RGD-containing peptide (GGCRGDMFGC) insertion, a similar approach was used with a slight modification. Instead of using a pGEMAHA containing the RGD peptide at the Spe I site as the template, pairs of RGD oligmers with Afl II, Nhel, or Hpal sequence flanking each end of the paired oligomers, respectively, were generated and directly cloned into pGEMAHA at each corresponding restriction enzyme site. The sense sequence of Aif II flanking RGD peptide is 5'-P04 TTAAGGGGTGGCTGCCGTGGCGATATG TTCGGTTGCGGC-3' and the anti-sense sequence is 5'-P04 TTAAGCCG CAACCGAAC ATATCG CCACGGCAGC CACCCC 3'. The sense sequence of Nhe I flanking RGD peptide is 5'-P04 CTAGCGGTGGCTGCCGTGGCG Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 133 ATATGTTCGGTTGCG-3' and the anti-sense sequence is 5'-PCXiCTAGC GCAACCGAACATATCGCCACGGCAGCCACCG-3'. The sense sequence of Spe I flanking RGD peptide is 5'-PQaCTAGTGGTGGCTGCCGTGGCGAT ATGTTCGGTTGCA-3' and the anti-sense sequence is 5'-P04 CTAGTGCAA CCGAACATATCGCCACGGCAGGCACCA-3'. The sense sequence of Hpa I flanking RGD peptide is 5'-P04 AACGGTGGCTGCC GTGGCGATATG TTCGGTTGCGGT-3' and anti-sense sequence is 5 -PO4AACGCAACC GAACATATCGCCACGGCAGCCACCGTT-3'. Underlined are the sequences of the Afl II, Nhe I, Spe I and Hpa I sites, respectively. The wild type HA protein containing the RGD peptide insertion at the Spe I site is designated as HArgd, whereas the binding-defective HA mutant proteins containing the RGD peptide insertion at Spe I site, respectively, are designated as HA(Y98F/E190Q)rgd, HA(Y98F/S228K)rgd, and HA(Y98F/E190Q/S228K)rgd. Construction of N-terminus insertion chimeric HA proteins. For construction of the chimeric HA protein that contains the apoE ligand at the N- terminus of the HAtmt (Y98F/E190Q/ S228K mutations) backbone, a glycine- rich sequence (G 3S)4 was introduced between the C-terminus of the apoE ligand and the N-terminus of the HAt subunit. The original signal peptide of the HAi subunit was deleted and replaced with a signal peptide (MKVLWAALLVTPLAGCQA) of the apoE ligand (Figure 4B.3.). For construction of a chimeric HA protein that contains the heregulin ligand at the N-terminus of the HAtmt backbone, a similar cloning approach Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 134 was employed except that the signal peptide (MARSTLSKPLKNKVNPPGP LIPLILLMLRGVST) and the first six amino acids following the signal peptide incorporated at the amino end of the chimeric HA protein is from the MuLV envelope protein, and a different spacer (AAAIEGRQDPLPGNDNS) was introduced between the C-terminus of the heregulin ligand and the N-terminus of the HAi subunit (Figure 4B.3.). Transient transfection, transduction and determination o f viral titer. Pseudotyped MuLV viral particles were produced by transient transfection of 293T cells with pCgp, pCnBG and appropriate HA-expression plasmid, essentially as described (Soneoka, 1995). The pCgp plasmid expresses MuLV gag-pol. The pCnBg plasmid expresses nuclear p-gal downstream of a CMV-MuLV LTR hybrid promoter and a neomycin resistant gene from an internal SV40 promoter. Plasmids were co-transfected into 60-70% confluent 293T cells in 60 mm plates by a calcium phosphate precipitation method. Five micrograms of each pCgp and pCnBg plasmids together were co transfected with 1 pg of the chimeric HA plasmid (pApoE-HAtmt or pHer- HAtmt). At 15-17 hrs post- transfection, the medium was replaced with 3 ml of fresh medium. The viral supernatants were harvested 24 hours later, filtered through a 0.45 pm filter (Millipore, CA) to remove cell debris and the viral titer analyzed by transduction of target cells. Target cells were plated onto six-well culture dishes the day before transduction at a density of 1 x 105 cells per well. One ml of the diluted viral Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. (G3S)4 spacer n / d k ic ... apoE-HAtmt ApoE * * Signal peptide of apoE Y98F/E190Q/S228K(HAtmt) Spacer: AAAIEGRQDPLPGNDNS Her-HAtmt Her < t t Signal peptide + ASPGS sequence of MoMuLV envelope protein Y98F/E190Q/S228K(HAtmt) Figure 4B.3. Insertion of a large ligand at the N-terminus of the HA1 subunit of the HA protein. * indicates mutations in the HAtmt backbone. u > L / l 136 supernatant with 8 pg/ml polybrene (Sigma, MO) was added to each well, followed by 10-12 hours incubation. Fresh media was then applied and the viral titer was determined by X-gal staining at 48 hours post-transduction at which time transduced cells were fixed with 0.5 % glutaraldehyde PBS buffer for 10 min, followed by two subsequent PBS washes of 10 min for each. Cells were then incubated with staining solution (4 mM potassium ferricyanide, 4 mM potasium ferrocyanide, 2 mM MgCl2 , 0.4 mg/ml X-gal, Sigma, MO), at 37°C over night. Titer was determined by counting the number of blue colonies under a microscope. Transduction of HT1080 cells by MuLV viral particles pseudotyped with apoE-HAtmt protein was carried out in a serum-free condition (DMEM). Cell lines expressing the LDLR receptor were subjected to serum starvation for 24 hr prior to transduction. Cell surface expression. Sixty to seventy percent confluent 293T cells were transfected with 1 pg of HA-expression plasmid and 14 pg of pBlueScript plasmid (Strategne, CA) as a filler DNA (to make up a total amount of 15 pg of DNA), using the calcium phosphate precipitation method. HA-expressing 293T cells were harvested 36-48 hr post-transfection to analyze for cell surface expression level. Five x 105 cells were pelleted (10 seconds at 12,000 r.p.m., eppendorf centrifuge 5415C) and washed with 1 ml of cold wash buffer (10% goat serum in PBS). Pelleted cells were re-suspended in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 137 100 pi of cold wash buffer containing primary antibody for 1 hr at 4°C followed by 1 ml wash with cold wash buffer. The primary antibody used to detect the presence of all chimeric HA proteins at the cell surface is a rabbit antiserum against FPV HA (provided by Dr. Klenk at Institute of Virology, Philipps- University, Marburg, Germany) at 1:500 dilution. The 1° antibody used for detection of the apoE ligand present in the apoE-HAtmt chimeric protein is a mouse anti-apoE antibody at 1:100 dilution (Perlmmune, MD). After washing, the cells were then incubated with a secondary antibody conjugated with FITC at 1:100 dilution under the same conditions as described for primary antibody incubation. The secondary antibody used to detect all chimeric HA proteins is a FITC-conjugated goat anti-rabbit antibody (Pierce, CA). The secondary antibody used to detect apoE ligand is an FITC-conjugated goat anti-mouse antibody (Pierce, IL). The resulting pellet was then re-suspended in 4% paraformaldehyde and subjected to fluorescent activated cell sorting (FACS) to measure cell surface expression. Virion incorporation. To measure for virion incorporation, 1 ml of filtered viral supernatant was purified through a 20% sucrose cushion by centrifugation at 14,000 r.p.m. for 30 min at 4°C. Due to the similar molecular weights of the HA2 subunit (~p27 kDa) of HA and the capsid (CA) protein (~30 kDa) of the MuLV viral particles, the resulting pellet was then re suspended in 2 x loading buffer (Novex, CA), split into two halves and subjected to SDS-PAGE. For detection of HA proteins, one blot was probed Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 138 with anti-FPV HA as the primary antibody (1:5000) and HRP-conjugated goat anti-rabbit IgG as secondary antibody (1:10,000), (Pierce, IL). For detection of MuLV capsid protein and MuLV env protein, the second blot was incubated with poly-clonal antibodies 78S-221 at 1:10,000 dilution and anti-SU (anti- Rauscher MuLV) at 1:3,000 dilution, respectively. Following the primary Ab incubation and wash, the blot was then incubated with an HRP-conjugated rabbit anti-goat IgG secondary antibody at 1:10,000 dilution (Pierce, IL) to detect both the capsid and the MuLV envelope signals. Protein signals were detected by an ECL kit (Amersham, CA). Virus-receptor binding. For detection of receptor binding activity of MuLV viral particles pseuodtyped with chimeric HA-RGD protein, 1 ml of filtered viral supernatant was mixed with 5 x 105 CHO/anbP3 cells and incubated at4°C for 2 hrs to prevent receptor internalization. At the end of incubation, each sample was washed once with 1 ml of washing buffer (PBS containing 10% goat serum), (BRL Gibco, CA). Following the wash, the bound viral particles were resuspended in 100 pi of wash buffer containing anti-FPV HA at 1:500 dilution for 1 hr at 4°C followed by 1 ml wash with cold washing buffer. FITC- conjugated goat anti-rabbit IgG was used as a secondary antibody at 1:100 dilution and incubated with the cells for another 1 hr at 4°C as described for incubation with primary antibody. After washing, the cells were fixed with 4% paraformaldehyde and subjected to FACS analysis to measure the receptor binding activity. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 139 For detection of the receptor binding activity of MuLV viral particles pseudotyped with apoE-HAtmt protein, similar procedures were followed, except that the target cells used were HT1080 cells and the initial virus-cell incubation was done in serum-free medium (DMEM). To remove lipids in the serum in the viral supernatant collected from viral-producing 293T cells, 1 ml of viral supernatant was subjected to purification by micro-centrifugation through a 20% sucrose cushion as described above. The resulting viral pellet was then resuspended in 1 ml of DMEM. Polykaryon formation ofNIH 3T3 cells. 8 x 104 NIH 3T3 cells were seeded on 6-well plate the night before transfection. The next day, cells were co-transfected with 0.1 pg of plasmids expressing wild type HA, HAtmt or chimeric HA, respectively, and 0.9 pg of pBlueScript plasmid (Strategene, CA) as the filler DNA, using the Lipofectamine Plus Reagent Kit (Gibco BRL, CA. #109864-013). Transfection procedures were carried out as described in the standard transfection protocol (Gibco BRL, CA). Thirty-six hrs post transfection, cells were washed once with room temperature PBS followed by 3 min incubation of 1 ml pH 5 PBS buffer (10 mM HEPES, 10 mM 2-[N- morpholinojethanesulfonic acid, MES), (Sigma, MO) at 37°C. Cells were then washed twice with 2 ml of room temperature DMEM, replaced with 2 ml of DMEM containing 10% fetal bovine serum and incubated at 37°C for 5 hrs to allow cell-cell fusion to occur. At the end of incubation, cells were fixed with 0.5% glutaraldehyde for 10 min at room temperature and stained with 0.1% Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 140 methylene blue at room temperature for 10 min. Polykaryons (syncytia) were counted using a light microscope at40x magnification. A polykaryon was defined as a cell containing 4 or more nuclei. Part 1: Insertion of small liaand at the internal sites in the HA protein. 4B.4. Results Insertion of dimeric apoE peptide and RGD peptide, expression and virion incorporation. To examine whether the five potential insertion sites identified on the surface of the HA molecule could tolerate a small peptide insertion, a dimeric apoE peptide directed to the LDL receptor and a RGD- containing peptide directed to the anb 0 3 integrin receptor, respectively, were cloned into each site. These insertions were carried out initially in the wild type HA backbone with the expectation of repeating the analysis in binding-defective HA molecules once they had been identified (Chapter 2). The immunoblot results of the cell lysates showed that none of the five potential sites could tolerate the dimeric apoE peptide insertion to produce properly cleaved HA subunits. Rather, they were expressed only in the precursor form and at a much lower level than for the wild type HA (Figure 4B.4.). Furthermore, their incorporation into MuLV viral particles was not Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. a > a * ■ ■ i n o IO <o W (O CO CO P CO CO CO Tf 10 CD T " t - r r r 1 r 1 i ■ i i 0 1 0 10 M CO CO CO ^ r r r r sr or iir nr 9 9 0 0 a a a a W W (0 (0 < << < 1 I X X HA0 HA! HA, Figure 4B.4. Insertion of apoE peptide ligand. Cell lysates of HAapoE chimeric proteins. None of the potential insertion sites in HA was able to tolerate an apoE peptide insertion to produce properly cleaved HA subunits. 142 detected (data not shown). A cleavage-defective mutant (HAcmt) was included in the immunoblot (Figure 4B.4.) to show that the uncleaved form of the HAapoE chimeric protein migrated at the same position as the HAcmt protein (due to their similar molecular weights). In contrast, three out of the five sites could tolerate the RGD peptide. This includes insertions between residues 130-131 (Alf II site), at the natural Spe I site between residues 135-136 and between residues 160-161 (Nhe I site). However, insertions of the RGD-containing peptide at the Afl II and Nhe I sites, respectively, seem to have an adverse effect on processing of the two chimeric HA proteins. Although the HA2 subunit was shown to be incorporated into viral particles, the signal for the HAt subunit was barely detectable. On the other hand, the HA(135-136) RGD chimeric HA protein was processed correctly and incorporated efficiently into MuLV viral particles (Figure 4B.5A.). In order to avoid the dual receptor binding activity that would be present in HA(i3 5 .i36)RGD contributed by the native HA towards the sialic acid residue and the RGD peptide towards anb p3 integrin receptor, three HA binding defective mutants were chosen at a later time to serve as the HA backbone for RGD peptide insertions at the Spe I site. The three binding mutants used were HAY98F/E190Q, HAY98F/S228K., and HAY98F/190Q/S228K. These HA binding-defective HA mutants containing the RGD peptide insertion at Spe I site were designated as Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Q Q Q Q Q CD CD CD CD CD E E Q £ E E < ! > P C D C O C D P Q- C O C O ^ to C D fy T “ T “ - T “ - T ” I I I I I T3 o m cm tn o = co co id co 5 T” T - T“ r - T - < < < < < < X I X I X I B HA0 HA, HA, CA HA0 HA? HA2 CA T3 E ? O' ^ * * 00 00 ^ ■ & a a O) E> w w O O o o c o o o O O CM CM O) 0) O) O) CM CM t- T - T — C O C O LU L U ■D 1 ± J y j E C i f u: s: ~ LL. LL 00 00 00 CO > 0 ) 0 ) 0 ) 0 1 0 ) 0 ) * b b b b b b < < < < < < < X X X X X X X 0 a & Figure 4B S. Insertion of RGD peptide into the HA molecule and virion incorporation. Immunoblots of MuLV particles pseudotyped with RGD-containing RGD peptide. Viral particles were produced at 32°C and the viral supernatant was pelleted through 20% sucrose and subjected to SDS-PAGE. A. Immunoblot of insertion of the RGD peptide into backbone of the wild type HA. B. Immunoblot of insertion of the RGD peptide into backbone of three different binding-defective HA mutants. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4B.1. Characterization of RGD-containing chimeric HA proteins Polykaryon formation of NIH 3T3 cells6 Titer of HA-oseudotvoed MuLV vectors on cell line0 fusion protein CSEa 32°C 37°C 293T CHO CHO/a|)b p3 HA 100 + + + + + + + (6.00±1.00)x104 (3.67±0.58)x102 (3.3310.58) x 102 HAY 9 8 F /E 1 9 0 Q 153 ±11 + + + + + + (2.67±1.15)x102 (3.0011.00)x101 (2.3310.58) x 101 H A (Y98F/E190Q)rgd 91 ±16 + + + + (1.13±0.58)x102 (1.33±0.57)x101 0 HAY 9 8 F /S 2 2 8 K 79 ±15 + + + + + + (2.67±0.58)x102 (2.33±0.58)x101 (2.33±0.58)x 101 HA(Y98F/S228K)rgd 96 ±12 + + + + (1.67±4.33)x102 (1.3310.58) x101 (1.6810.58) x 101 HAY 9 8 F /E 1 9 0 Q tS 2 2 8 K 94 ±18 + + + + + + (4.33 ±1.53)x101 0 0 H A(Y98F/E190Q/S228K)rgd 93 ±10 + + + + (2.33 ±1.53)x101 0 0 *• Percent of cell surface expression (CSE) level of HAs in transfected 293T cells relative to the wild type HA ± SD, b - Level of fusion was evaluated relative to the wild type HA at 37°C. + + + +, 200%; + + +, 75-100%, + < 25%. c - Transudctions were performed at 37°C as described in 4B.3. Materials and Methods, p-gal titer reported in cfu/ml ± SD. HA(Y98F/E190Q)rgd, HA(Y98F/S228K)rgd, and HA(Y98F/E190Q/S228K)rgd. As shown in Figure 4B.5B., all three RGD-containing chimeric HA proteins were also processed correctly and incorporated into MuLV viral particles, but the incorporation of all three RGD-containing chimeric HA proteins into MuLV particles was lower than their parental HAs lacking the RGD peptide. In addition, cell surface expression assays were performed to find out whether this reduction in virion incorporation observed for the chimeric HAs was due to decrease in cell surface expression. The results showed that surface expression of these three chimeric HAs was comparable to their parental HAs, indicating that the reduction in virion incorporation was due to other factors caused by the insertion of the RGD peptide into the HA backbone (Table 4B.1.). Fusion function o f RGD-containing chimeric HAs. To answer the question whether an insertion of the RGD peptide into the HA backbone would impair the fusion function of HA protein, cell-cell fusion assays were performed. The results showed that all three RGD-containing chimeric HA proteins were hypo-fusogenic. From the studies described in chapter 2, it was known that hypo-fusogenic mutants of HA could be compensated by expression at 32°C rather than 37°C prior to performing the fusion assay. Consequently, the RGD-containing chimeric HAs were expressed at 32°C instead of 37°C post-transfection of 293T cells. The results showed that Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 146 expression of these RGD-containing chimeric HAs at 32°C was able to restore their membrane fusion function (Table 4B.1 .). Transduction ability and integrin receptor binding activity of RGD- containing chimeric HAs. The transduction assay was performed to ask the question whether these RGD-containing chimeric HA proteins in the context of MuLV viral particles could transduce efficiently into cells that expressed the targeted integrin anbP3 receptor. The two cell lines that were used in the transduction assay to measure the transduction efficiency of RGD-containing vectors were CHO and CHO/aubP3 cells. It is worth noting here that the overall transduction efficiency of MuLV viral particles pseudotyped with any of the three chimeric HA or the wild type HA proteins on CHO-based cell lines is 1-2 logs lower than on the 293T cell line (Table 4B.1.). Since the three RGD- containing chimeric HA proteins still retain minimal receptor binding activity towards sialic acid residues, MuLV viral particles pseudotyped with each of their parental HA protein mutant were included in the experiment to assess whether the RGD-containing chimeric HA could enhance transduction efficiency on cells that express the integrin receptor. In the transduction experiment, the CHO cell line that lacks expression of aubfb integrin receptor was also included as a negative control. The titer results indicated that none of the three RGD-containing chimeric HAs were able to enhance transduction efficiency on CHO/a»bP3 (Table 4B.1.). Subsequently, I asked the question 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. HA wt \ ; 4 -H A w t HA mock mock Y98F/E190Q Y98F/S228K Y98F/E290Qrgd Y98F/S228Krgd i i rmi i i m u I I I IIIV m mock HA wt F U » Figure 4B. 6. MuLV viral particles pseudotyped with RGD-containing chimeric protein do not bind to a„b p3 integrin receptors on CHO/a„b p3 cells. A. Receptor binding activity of HA(Y98F/E190Q)rgd. B. Receptor binding activity of HA(Y98F/S228K)rgd. C. Receptor binding activity of HA(Y98F/E190Q/S228K)rgd. A 148 whether the inability of these RGD-containing chimeric HAs to enhance transduction efficiency on target cells could due the lack of receptor binding activity. I therefore performed virus-cells binding assays to assess the receptor binding activity of these chimeric HAs. The results revealed that the failure of MuLV viral particles pseudotyped with these RGD-containing chimeric HA proteins to enhance transduction efficiency on target cells indeed was due to their lack of binding to the integrin receptor (Figure 4B.6.). 4B.5. Discussion I have attempted to identify sites in the HA molecule that could tolerate a small ligand insertion and confer receptor binding specificity. The insertion of dimeric apoE and RGD-containing peptide ligands showed different results. None of the five potential sites identified on the surface of the HA molecule could tolerate the insertion of the dimeric apoE peptide, whereas the Spe I site located between residues 135-136 could tolerate the insertion of the RGD- containing peptide. In addition, insertion of the RGD peptide between residues 130-131 and 160-161 failed to allow proper processing of the HA precursor. The failure of the HA protein to tolerate the dimeric apoE peptide is most likely due to the size of the peptide (30 amino acids in length) and the secondary structure that is known to be present in this dimeric apoE peptide. It has been revealed by circular dichroism analysis that this dimeric apoE Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 149 peptide displays a-helix content (Dyer et al., 1991 and 1995). Therefore, it is reasonable to infer that the insertion of this peptide into the HA molecule might have impaired the proper folding of the protein. In addition, the fact that the overall expression levels of the precursor form of the HAapoE chimeric proteins were lower than the wild type HA in the cell lysates, and no HAapoE chimeric proteins were shown to be incorporated into the MuLV viral particles, suggest that these chimeric HA proteins are probably mis-folded in a way that prevents trimerization. Therefore, they were transported inefficiently to the cell surface, and the low expression level of the precursor form observed in the cell lysates could be the monomeric form of the chimeric HAapoE proteins. From the available structural information, I proposed that the lack of integrin receptor binding activity observed for these chimeric HA proteins unlikely due to steric hindrance caused by neighboring residues in the protein or by a carbohydrate group since the site of peptide insertions were shown to be distant from all known glycosylation sites. Although numerous linear or constrained peptides with high receptor binding affinity have been reported in the literature, a potential problem with the use of these peptides incorporated into a large protein is their configuration in the context of the entire protein. Hart et al., 1994 have demonstrated that a filamentous phage displaying the same RGD-containing peptide binds to its targeted receptor and was internalized efficiently into target cells. In my study, the same RGD- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 150 containing peptide incorporated into the HA molecule lacks receptor-binding activity. Presumably, the folding of the RGD peptide in the context of the HA protein did not result in the right configuration to contribute a high integrin receptor binding affinity. Furthermore, since an approximately 10-20% sialic acid binding activity (Chapter 2) is still retained in these chimeric HA proteins, the use of the anti- FPV HA in the binding assay to detect the integrin receptor binding activity might not be a good choice due to high background of HA binding to sialic acid. Therefore, it would be interesting to obtain an antibody against this particular peptide and perform a FACS analysis to re examine for their integrin receptor binding activity and to find out whether the peptide is exposed on the surface of the chimeric HA molecule, as been previously done for RGD peptide insertions in MuLV envelope protein (Wu et al., 2000). An inability of HA to tolerate an internal single chain antibody insertion has been reported (Patterson et al., 1999). In the report, it was shown that an insertion of a single chain antibody directed to the hapten NIP between residues 139-145 in the HAt subunit of an H3 subtype HA protein resulted in a fusion-defective chimeric protein. On the other hand, the partial defect in fusion caused by the inclusion of a RGD peptide into the backbone of the three binding-defective mutant HAs observed in my study was not as detrimental as the former one. However, the necessity of the RGD-containing chimeric HA protein to be expressed at 32°C in order to restore its fusion Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 151 function may suggest an structural instability of the protein, thus preventing efficient virion incorporation. In conclusion, although insertion of dimeric apoE and RGD-containing peptides into the HA backbone has failed to generate a targeted chimeric HA protein, nevertheless, it provides additional structural information about the HA protein of the H7 subtype. One site has been identified that is able to tolerate a small peptide insertion; other peptide candidates whose targeted receptors are of interest may be used in the future at this site. Part 2: Insertion of a large ligand at the N-terminus of the HA^ subunit of HA. 4B.6. Results Expression and virion incorporation of chimeric proteins apoE-HAtmt and her-HAtmt To examine whether insertion of an apoE ligand or an heregulin ligand at the N-terminus of the HAi subunit of the HAtmt protein would impair their protein processing or virion incorporation, MuLV viral particles pseudotyped by apoE-HAtmt and her-HAtmt, respectively, were subjected to SDS-PAGE. Immunoblot results showed that the apoE-HAtmt chimeric protein was processed correctly and incorporated efficiently into MuLV viral particles. The shift in the molecular weights of both the precursor (HAo) form and the HAi subunit is caused by the apoE ligand (~70 kDa) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. apoE-HA0 HA0 /apoE-HA1 HA1 HA, CA B Her-HA0 HA0 HA1 HA, CA Figure 4B.7. Immunoblots of chimeric HA protein incorporated into MuLV viral particles. A. apoE-HAtmt. B. her-HAtmt. Tr = trypsin treatment of her-HAtmt pseudotyped MuLV particles at 37°C for 5 min. 153 added at the N-terminus of the subunit (Figure 4B.7A). In contrast, insertion of an heregulin ligand at the N-terminus of the HAi subunit resulted in an uncleaved form of the chimeric HA protein. But, like the uncleaved form of the wild type HA (see mutant HAcmt in Chapter 3), the her-HAtmt chimeric protein also incorporated into MuLV viral particles (Figure 4B.7B). As noted previously in Chapter 2, the uncleaved form of HA can sometimes be rescued by exogenous trypsin treatment (Klenk et al.,1994). Also, since the uncleaved form of HA protein lacks the ability to carry out membrane fusion function, I tested here whether exogenous trypsin treatment of MuLV viral particles could cleave the precursor form of the chimeric her- HAtmt protein properly into its two subunits, the her-HA-i and HA2. Results from this experiment showed that exogenous trypsin treatment not only failed to cleave the chimeric her-HAtmt protein, it also degraded the protein, suggesting that the protein might be mis-folded in a way which still allowed cell surface expression and virion incorporation, but made the protein susceptible to degradation by trypsin (Figure 4B.7B.). Since the her-HAtmt chimeric protein failed to express and process correctly, the rest of my study focused on the chimeric apoE-HAtmt protein. Cell surface expression and accessibilty of apoE ligand by an anti-apoE antibody. The next question I asked was whether the apoE ligand present at the N-teminus of the chimeric apoE-HAtmt protein was exposed on the surface of the molecule in such a way as to allow accessiblity by an anti-apoE Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 154 r« - i u . I LU 0 Q. (0 1 a r a < X I a o o E E < E < X I LU o Q. C O Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F ig u re 4B.8. Cell surface expression o f apoE-HAmt chimeric protein. 155 antibody. To answer this question, a cell surface expression assay was performed using two different antibodies. Detection of cell surface expression using anti- FPV HA antibody was included as a control to ensure that both the HAtmt and apoE-HAtmt proteins were expressed at the cell surface. Results showed that the apoE ligand present in the chimeric apoE-HAtmt protein was accessible by the anti-apoE antibody (Figure 4B.8.), suggestig that the apoE ligand is exposed on the surface of the apoE-HAtmt molecule. Receptor binding activity. A virus-binding assay was performed to assess the receptor binding activity of the chimeric apoE-HAtmt protein for the LDLR receptor. HT1080 cells, which express a high level of LDL receptor, were used as target cells in the binding assay. Since lipids in the growth medium down-regulates the cell surface expression of the LDL receptor (Bailey 1973; Bates and Rothblat, 1974; Rothblat 1972; and Williams and Avigan 1972), HT1080 cells were incubated in serum-free growth medium 24 hr before harvesting the cells. In addition, to prevent ligand competition between the apoE ligand on the chimeric apoE-HAtmt protein and the lipids present in the serum, initial incubation of HT1080 cells with viral supernatant was also carried out in serum- free medium. The results showed that MuLV viral particles pseudotyped with the apoE-HAtmt protein were able to bind specifically to the LDL receptor, and the receptor binding signal was detected using either anti- FPV HA or anti-apoE antibody (Figure 4B.9.). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 156 L U 0 Q . C O 1 a Q. < D O 2 0C - I Q -J < D J O 4 -4 o >» 75 o o Q ) Q. tf) (/> ■o c I L U O Q. C O o> cri £ 3 O) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 157 ApoE-HAtmt is able to retain its membrane fusion function. Despite the aibility to confer high receptor binding specificy that numerous chimeric MuLV envelope proteins have been reported to achieve, a defect in fusion is commonly observed that is caused by the engineering of the MuLV envelope protein. To ensure that the chimeric apoE-HAtmt protein remained fusion competent, cell-cell fusion assays were carried out. The results indicated that the attachment of the apoE ligand at the N-terminus of the HAi subunit did not affect the membrane fusion function of the remaining intact HAtmt protein (Table 4B.2.). Transduction of cell lines expressing the LDL receptor. To find out whether MuLV viral particles pseudotyped with the chimeric apoE-HAtmt protein could use the LDL receptor to gain entry into the cell, the transduction assays were performed. Since the apoE-HAtmt chimeric protein still retains minimal receptor binding activity for sialic acid, MuLV viral particles pseudotyped with the HAtmt protein were included in the experiment as controls. In addition, to examine whether a difference in the density of receptors on the cell surface could influence transduction efficiency, three different cell lines (293, 293/LDLR and HT1080) that express different levels of the LDL receptor as determined by FACS analysis (data not shown) were tested. 293 cells express a low level of the LDL receptor; 293/LDLR cells express an intermediate level of the LDL receptor; and HT1080 cells express a high level of LDL receptor on their cell surface. The titer results indicated Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission. Table 4B.2. Fusion property of chimeric apoE-HAtmt protein. Construct Polykaryon formation on NIH 3T3 cells9 32°C 37°C HA wt + + + + + + + HAtmt + + + + + + apoE-HAtmt + + + + + + a • Levels of fusion were evaluated relative to the wild type HA at 37°C . + + + +, 200% ; + + +, 75-100% , + < 25%. Table 4B.3. Transduction of apoE-HAtmt-pseudotyped MuLV particles. Cell surface expression level of LDLR MuLV vectors Titers on cell linea pseudotyped with 293 293/LDLR HT1080 HAtmt 43.3 ±13.3 26.7 ± 5.8 33.3 ± 5.8 apoE-H A tm t.. ..... JLQ0.1.2.QQ. ... 3J7_ ±J.J5 5.67 ±1.15 a- p-gal titer reported in cfu/ml ± SD. U ) 0 0 159 that MuLV viral particles pseudotyped with the chimeric apoE-HAtmt protein not only were unable to enhance transduction efficiency on all three cell lines, but transduced these cell lines at approximately one order of the magnitute less than those observed by vectors pseudotyped with the parental HAtmt protein (Table 4B.3.). 4B.7. Discussion In this part of the study, I have described attempts to engineer a targeted chimeric HA protein through the approach of using a large ligand insertion at the N-terminus of the HAi subunit of a binding-defective HA protein, HAtmt. A similar approach taken by F.-L. Cosset's group recently demonstrated that the N-terminus of the HAi subunit could tolerate large ligand insertions (Hatziioannou etal., 1999). However, the HAbackone into which the new ligand was incorporated into was the wild type HA. The problem with this particular approach is that it creates dual tropsim; the chimeric HA proteins bind to both sialic acid residues and the receptor which the new ligand targets to. Although it has been shown that these chimeric HA proteins were able to bind specifically to their targeted receptor, only a minimal enhancement of titer was observed on target cells. In my approach, the sialic acid binding activity of HA has been impaired so that the initial Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 6 0 attachment of the viral particles to target cells solely depends on the binding of the new ligand to its targeted receptor. The results in this study which showed that the N-terminus of the HAi subunit of the HA protein is able to tolerate a large ligand insertion is in good agreement with observations reported by Hatzioannou et al., 1999. However, the lack of cleavability of the chimeric her-HAtmt protein, in contrast to the high cleavability of the apoE-HAtmt protein constructed in my study, suggests that not ail ligands inserted at the N-terminus of the HAi subunit of the HA protein can be processed correctly. In conclusion, I have made a chimeric HA protein that possess both binding and fusion functions, but failed to transduce target cells efficiently. However, unless a series of different binding moieties added at the N- terminus of the HAi subunit of the HA molecules and is tested to be unable to produce an enhancement of titer on target cells, it is too early to draw any negative conclusion about the feasibility of this particular approach. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 5 What we have learned from designing a targeted retroviral vector using the HA protein Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 162 5.1. Overall conclusions MuLV-based retroviral vectors are currently the most utilized delivery system for gene therapy via ex vivo procedures. In the past decade, numerous researchers have attempted to fully exploit the use of these vectors by restricting tropism and improving transduction efficiency, so that the vectors could be delivered into humans using in vivo protocols. However, the problems are considerable. Although specific-receptor targeting has been reported with chimeric MuLV envelope proteins, in general, target cells were transduced at an efficiency that was 3-5 logs lower than that obtained with unmodified viral particles. The problem with such chimeric proteins was found to be at a post-binding step, specifically, the inability of chimeric MuLV envelope proteins to trigger membrane fusion. I have attempted to overcome the problem of triggering of fusion. My strategy utilized a non-specific fusion protein, influenza HA, in MuLV-based vectors, in order to enable membrane fusion. My studies have demonstrated that the combination of HAtmt as a fusogenic protein, together with a chimeric MuLV envelope protein directed to Flt-3 receptor is able to trigger membrane fusion and enhance the transduction efficiency of Flt-3 expressing cells. This constitutes a proof of principle that such a strategy is feasible. The positive outcome of this approach suggests that this targeting strategy could be expanded by using different binding proteins in addition to chimeric MuLV envelope proteins. Several groups have reported that Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 163 numerous host proteins can efficiently be incorporated into retroviral particles (Balliet et al., 1998; Bastiani, etal., 1997; Bubbers and Lilly, 1977; and Calafat et al., 1983). Therefore, it would be interesting to test whether combinations of different binding proteins with the HAtmt fusogenic protein could also enhance transduction efficiency on cells that express the receptors to which such proteins could bind. The possible barriers to efficient entry by targeted HA pseudotyped MuLV vectors. Although I was able to demonstrate the applicability of this targeting strategy, a concern arising from this study is the low efficiency of viral entry. There are at least two possible blocks that could prevent efficient entry of MuLV vectors that possess both a specific-receptor binding and fusion function, either on two separate molecules (in the case of co expression of a chimeric MuLV envelope protein and the HAtmt protein) or on the same molecule (in the case of a chimeric HAtmt). 1. The chimeric MuLV protein fails to trigger efficient internalization of the targeted receptor. One possibility is that the receptor-ligand interaction fails to trigger internalization of the targeted receptor. The HA protein is known to promote membrane fusion in a non-specific, pH-triggered fashion. However in order to gain entry into a host cell, this fusion event must take place in a low pH endosome. Therefore, the MuLV particles must bind to their targeted receptor and then be internalized to allow fusion to take place in the endosome. In Chapters 4 and 5 ,1 presented data that showed that both a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 164 chimeric MuLV envelope protein (PC7 or D84KEP1), together with a binding- defective HA protein, HAtmt) and a chimeric HA protein, apoE-HAtmt (where binding and fusion modules are on the same molecule), were able to bind to their cognate receptors on target cells at greater than background levels. However, in both cases, the target cells were transduced inefficiently. Thus, the failure could be due the difference in the intrinsic properties between these ligands and the natural ligands (Figure 5.1.). If the barrier to efficient transduction were due to inefficient internalization of the targeted receptor, it would be interesting to test other fusion proteins which are known to fuse on the cell surface after viral attachment by their partner protein. For example, the F protein from Sendai, simian and measles virus (paramyxovirus family) have recently been shown to pseudotype MuLV particles. However, not only were these F proteins not able to promote fusion, but also caused a drastic reduction in infectivity when co-expressed with chimeric MuLV envelope proteins (Hatziioannou and Cosset, unpublished data; Hatziioannou et al., 2000; and Spiegel et al, 1998). 2. The HAtmt protein fails to trigger virus-cell fusion. Based on the cell cell fusion assay, my data suggested that both the co-expression of a chimeric MuLV envelope protein with the HAtmt and the addition of a binding moiety to HAtmt itself do not compromise the fusion function of the HA protein. However, whether the cell-cell fusion assay reflects the natural virus cell fusion event that occurs in the endosome remains unclear. Therefore, it Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 165 is possible that although vectors containing both the chimeric MuLV envelope protein and HAtmt protein or a chimeric HAtmt protein are efficiently internalized into an endosome following receptor binding, the HAtmt moiety is unable to trigger fusion. Possibly this could be caused by a steric hindrance between the co-expressed MuLV envelope protein or the additional binding domain that prevents mixing of the two lipid membranes. For example, the interaction between the binding domain and the receptor may somehow 'anchor1 the surface of the virus and prevent free movement that may be necessary to allow lipid mixing and the formation of a fusion pore. (Figure 5.1.). In this study, I have attempted to overcome the lack of fusion seen in previous approaches to targeting retroviral vectors by using the HA protein. However, my studies have revealed additional potential barriers and problems that may be present even when a pH-dependent protein is engineered. Thus, engineering fusion proteins to achieve receptor-specific entry remains a challenge to all the researchers in this field. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 166 Binding-defective HA Receptor-specific binding via chimeric MuLV env Targeted chimeric HA protein Binding defective HA fusion protein internalization fusion by HA in low pH endosome Target cell Figure 5.1. Possible barriers to efficient viral entry of MuLV particles that possess both receptor-specific binding and fusion functions. I. Co-expression of a fusion defective chimeric MuLV envelope protein with a binding- defective HA protein. II. A targetable HA protein that contains both receptor-specific binding and fusion functions. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 167 References Alford, D., H. Ellens, and J. Bentz. 1994. Fusion of influenza virus with sialic acid-bearing target membranes. Biochem. 33:1977-1987. Anderson W. F. 1998. Human gene therapy. Nature 392[supp]:25-30. Bachi, T., W. Gerhard, and J. W. Yewdell. 1985 . Monoclonal antibodies detect different forms of influenza virus hemagglutinin during viral penetration and biosynthesis. J. Virol. 55:307-313. Bailey, J. M. 1973. Regulation of cell cholesterol content. In "Atherogenesis: initiating factors" Ciba Found. Symp 12 (New Ser.) pp. 63-92, Amsterdam. Balliet, J. W., and P. Bates. 1998. Efficient infection mediated by viral receptors incorporated into retroviral particles. J. Virol. 72:671-676. Barr, P. J. 1991. Mammalian subtilisins: the long-sought dibasic processing endoproteases. Cell 66:1-3. Bastiani, L., S. Laal, M. Kim, and S. Zolla-Pazner. 1997. Host cell- dependent alterations in envelope components of human immunodeficiency virus type I virions. J. Virol. 71:3444-3450. Bates, S. R., G. H. Rothblat 1974. Regulation of cellular sterol flux and synthesis of human serum lipoprotiens. Biochem. Biophys. Acta 360:38-55. Battini, J. L. P. Rodrigues, R. Muller, O. Danos, and J. M. Heard. 1996. Receptor-binding properties of a purified fragment of the 4070A amphotropic murine leukemia virus envelope glycoprotein. J. Virol. 70:4387-4393. Battini, J. L., O. Danos, and J. M. Heard. 1995. Receptor-binding domain of mureine leukemia virus envelope glycoproteins. J. Virol. 69:713-719. Benedict, C. A., R. Y. M. Tun, D, Rubinstein, T. Guillanume, P. M. Cannon, and W. F. Anderson. 1999. Targeting of retroviral vectors to CD34 expressing cells. Hum. Gene Ther. 10:545-557. Binley, J., and J. P. Moore. 1997. HIV-cell fusion: the viral mousetrap. Nature 387:346-348. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 168 Brown, M. S., R. G. W. Anderson, and J. L. Goldstein. 1983. Recycling receptors: the round-trip itinerary of migrant membrane proteins. Cell 32:663- 667. Bubbers, J. E., and F. Lilly. 1977. Selective incorporation of H-2 antigenic determinants into Friend virus particles. Nature 266:458-459. Bullough, P. A., F. M., Hughson, J. J. Skehel, and D. C. Wiley. 1994. Structure of influenza haemagglutinin at the pH of membrane fusion. Nature 371:37-43. Calafat, J., H. Janssen, P. Demant, J. Hilgers, and J. Zavada. 1983. Specific selection of host cell glycoproteins during assembly of murine leukemia virus and vesicular stomatitis virus: presence of Thy-1 glycoprotein and absence of H-2, Pgp-1 and T-200 glycoproteins on the envelopes of these virus particles. J. Gen. Virol. 64:1241-1253. Carr, C. M. and P. S. Kim. 1993. A spring-loeaded mechanism for the conformational change of influenza hemagglutinin. 1993. Cell 73:823-832. Castrucci, M. R., I. Donatelli, L. Sidoli, Y. Kawaoka, Y., and R. G. Webster. 1993. Genetic reassortment between avian and human influenza A viruses in Italian pigs. Virology 193:503-506. Caton, A. J., and G. G. Browniee. 1982 . The antigenic structure of the influenza virus A/PR/8/34 hemagglutinin (H1 subtype). Cell 31:417-427. Chan, D. C., D. Fass, J. M. Berger, and P. S. Kim. 1997. Core structure of gp41 from HIV envelope glycoprotein. Cell 89:263-273. Chen, J., K. H. Lee, D. A. Steinhauer, D. J. Stevens, J. J. Skehel, and D. C. Wiley. 1998. Structure of the hemagglutinin precursor cleavage stie, a determinant of influenza pathogenicity and the origin of the labile conformation. Cell 95:409-17. Class, E. C. J., Osterhaus, A. D. M. E., Beek, R. Jong, J. C. J., Rimmelzwaan, G. F. R., Senne, D. A., Krauss, S., Shortridge, K. F., and Webster, R. G. 1998. Human influenza A H5N1 virus related to a highly pathogenic avian influenza virus. LACENT 351:472-477. Class, E. C. J., Kawaoka, Y., Jong, J. C., Masurel, N., and Webster, R. G. 1994. Infection of children with avain-human reassortant influenza virus from pigs in Europe. Virology 204:453-457. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 169 Cohen , B. D., J. M. Green, L. Foy, and H. P. Fell. 1996. Her4-mediated biological and biochemical properties in NIH 3T3 cells: evidence for Her1- Her4 heterodimers. J. Biol. Chem. 271:4813-4818. Cosset, F-L, S. J. Russell. 1996. Targeting retrovirus entry. Gene Ther. 3:946-956. Cosset, F-L, F. J. Morling, Y. Takeuchi, R. A. Weiss, M. K. L. Collins, and S. Russell. 1995. Retroviral retargeting by envelopes expressing an N- terminal binding domain. J. Virol. 69:6314-6322. Cristiano, R.J., D. T. Curiel. 1996. Strategies to accomplish gene delivery via the receptor-mediated endocytosis pathway. Cancer Gene Ther. 1996; 3:49-57. Daniels, R. S., J. C. Downie, A. J. Hay, M. Knossow, J. J. Skehel, M. L. Wang, and D. C. Wiley. 1985. Fusion mutants of the influenza virus hemagllutinin glycoprotein. Cell 40:431-439. Domas, R. W., M. J. Gething, J. Henneberry, J. White, A. Helenius. 1986. Variant influenza virus hemagglutinin that induces fusion at elevated pH. J. Virol. 57:603-613. Dong, J., M. G. Roth, and E. Hunter. 1992. A chimeric avian retrovirus containing the influenza virus hemagglutinin gene has an expanded host range. J. Virol. 66:7374-7382. Dyer, C. A., D. P. Cistola, G. C. Parry, L. K. and Curtiss. 1995. Structural features of synthetic peptides of apolipoprotein E that bind the LDL receptor. J. Lipid Res. 36:80-88. Dyer, C. A., and L. K.Curtiss. 1991. A synthetic peptide mimic of plasma apolipoprotein E that binds the LDL receptor. J. Biol. Chem. 266:22803- 22806. Eisen, M. B., S. Sabesan, J. J. Skehel, and D. C. Wiley. 1997. Binding of the influenza A virus to cell-surface receptors: structures of five hemagglutinin-sialyloligosaccharide complexes determined by X-ray cyrstallography. Virology 232:19-31. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 170 Ellens, H., J. Bents, D. Mason, F. Zhang, and J. M. White. 1990. Fusion of influenza hemagglutinin-expressing fibroblasts with glycophorin-bearing liposomes: role of hemaggluinin surface density. Biochemistry 29:9697-9707. Empig, C. L. 1999. Moloney murine leukemia virus (MoMuLV) envelope protein: interaction with receptor and engineering for targeted gene transfer. Thesis. Etienne-Julan, M., P. Roux, S. Carillo, P. Heanteur, and M. Piechaczyk. 1992. The efficiency of cell targeting by recobinant retroviruses depends on the nature of the receptor and the composition of the artificial cell-virus-liner. J. Gen. Virol. 73:3251-3255. Fielding, A. K., M. Maurice, F. J. Morling, F.-L. Cosset, and S. J. Russell. 1998. Inverse targeting of retroviral vectors: selective gene transfer in a mixed population of hematopoietic and nonhematopoietic cells. Blood 91:1802-1809. Furuta, R. A., C. T. Wild, Y. Weng, and C. D. Weiss. 1998. Capture of an early fusion-active conformation of HIV-1 gp41. Nature Struc. Biol. 5:276- 279. Gilbert, J. M., L. D. Hernandez, J. W. Balliet, P. Bates, and J. M. White. 1995. Receptor-induced conformational changes in the subgroup A avian leukosis and sarcoma virus envelope glycoproetein. J. Virol. 69:7410-7415. Gilbert, J. M. D. Mason, and J. M. White. 1990. Fusion of Rous sarcoma virus with host cells does not require exposure to low pH. J. Virol. 64:5106- 5113. Goud, B., P. Legrain, and G. Buttin. 1988. Chemical modification of a murine ecotropic Moloney retroviral vector to human cells allows internalization but not theestablishment of the proviral state. Virology 163:251-254. Gunther-Ausborn, S., P. Schoen, I. Bartoldus, J. Wilschut, and T. Stegmann. 2000. Role of hemagglutinin surface density in the initial stages of influenza virus fusion: lack of evidence for cooperativity. J. Virol. 74:2714- 2720. Han J. Y., Cannon P.M., Lai, K.M., Zhao Y, Eiden, M.V., W. F. Anderson. 1997. Identification of envelope protein residues required for the expanded host range of 10A1 murine leukemia virus. J. Virol. 71:8103-8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 171 Han, X., N. Kasahara, and Y. W. Kan. 1995. Ligand-directed retroviral targeting of human breast cancer cells. Proc, Natl. Acad. Sci. USA 92:9747- 9751. Hart, S. L., A. M. Knight, R. P. Harbottle, A. Mistry, H.-D. Hunger, D. F. Cutler, R. Williamson, and C. Coutelie. 1994. Cell binding and internalization by filamentous phage displaying a cyclic RGD-containing peptide. J. Biol. Chem. 269:12468-12474. Hatziioannou T., S. J.Russell, and F.-L., Cosset 2000. Incorporation of simian virus 5 fusion protein into murine leukemia virus particles and its effect on the co-incorporation of retroviral envelope glycoproteins. Virology 267:49- 57. Hatziioannou, T., E. Delahaye, F. Martin, S. J. Russell, and F.-L.Cosset. 1999. Retroviral display of functional binding domains fused to the amino terminus of influenza hemagglutinin. Hum. Gene. Ther. 10:1533-1544. Hatziioannou T., S. Valsesia-Wittmann, S. J. Russell, and F-L Cosset 1998. Incorporation of fowl plaque virus hemagglutinin into murine leukemia virus particles and analysis of the infectivity of the pseudotyped retroviruses. J. Virol. 72:5313-5317. Ho, S. N., H. D. Hunt, R. M. Norton, J. L. Pullen, and L. R. Please. 1989. Site-directed mutagenesis by overlap extension using polymerase chain reaction. Gene 77:51-59. Holmes, W. E., M. X. Sliwkowski, R. W. Akita, W. J. Henzel, J. Lee, and J. W. Park, 0. Yansura, N. Abadi, H. Raab, G. D. Lewis, H. M. Shepard, W.-J. Kuang, W. I. Wood, D. V. Goeddel, and R. L. Vandlen. 1992. Identification of heregulin, a specific activator of p185e r b B 2 . Science 256:1205-1210. Horimoto, T., K. Nakayama, S. P. Smeekens, and Y. Kawaoka. 1994a. Proprotein-processing endoproteases PC6 and furin both activate hemagglutinin of virulent avain influenza viruses. J. Virol. 68:6074-6078. Horimoto, T., and Y. Kawaoka. 1994b. Reverse genetics provides direct evidence for a correlation of hemagglutinin cleavability and virulence of an avain influenza A virus. J. Virol. 68:3120-3128. Hwang, J.-J., L. Li, and W. F. Anderson. 1997. A conditional self inactivated retrovirus vector that uses a tetracycline-responsive expression system. J. Virol. 71:7128-7131. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 172 Ikeda, H., K. Kato, T. Suzuki, H. Kitani, Y. Matsubara, S. Takase-Yoden, R. Watanabe, M. Kitagawa, and S. Aizawa. 2000. Properties of the naturally occurring soluble suface glycoprotein of ecotropic murine leukemia virus: binding specificity and possible conformational change after binding to receptor. J. Virol. 74:1815-1826. Januszeski, M. M., P. M. Cannon, D. Chen, Y. Rozenberg, and W. F. Anderson. 1997. Functional analysis of the cytoplasmic tail of Moloney murine leukemia virus envelope protein. J. Virol. 71:3613-3619. Kabat, D. Targeting retroviral vectors to specific cells. 1995. Science 269:417. Kasahara, N., A. M. Dozy, and Y. W. Kan. 1994. Tissue-specific targeting of retroviral vectors through ligand-receptor interaction. Science 266:1373- 1376. Katen, L. J. 1998. Characterization of target cell entry by murine leukemia viruses. Thesis. Kawaoka Y., and R. G. Webster. 1988. Sequence requirements for cleavage activation of influenza virus hemagglutinin expressed in mammalian cells. Proc. Natl. Acad. Sci. USA. 85:324-328. Kelm, S., J. C. Paulson, U. Rose, R. Brossmer, W. Schmid, B. P. Bandgar, E. Schreiner, M. Hartmann, and E. Zbiral. 1992. Use of sialic acid analogues to define functional groups involved in binding to the influenza virus hemagglutinin. Eur. J. Biochem. 205:147-153. Khatchikian, D., M. Orlich, and R. RotL 1989. Increased viral pathogenicity after insertion of a 28S ribosomal RNA sequence into the haemagglutinin gene of an influenza virus. Nature 340:156-157. Kido, H., Y. Yokogoshi, K. Sakai, M. Tashiro, Y. Kishino, A. Fukutomi, and N. Katunuma. 1992. Isolation and characterization of a novel trypsin like protease found in rat bronchiolar epithelial Clara cells: a possible activator of the viral fusion glycoprotein. J. Biol. Chem. 267:13573-13579. Klenk, H. D., and W. Garten. 1994. Host cell proteases controlling virus pathogenicity. Trends Microbiol. 2:39-43. Klenk, H.-D., R. Rott, M. Orlich, and J. Blodorn. 1975. Activation of influenza A viruses by trypsin treatment. Virology 68:426-439. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 173 Kuroda K, C. Hauser, R. Rott, H.-D. Klenk, and W. Doerfler. 1986. Expression of the influenza virus haemagglutinin in insect cells by a baculovirus vector. EMBO J. 5:1359-65. Lamb, R. A. 1993. Paramyxovirus fusion: a hypothesis for change. Virology 197:1-11. Li, S. M. Orlich, and R. Rott 1990. Generation of seal influenza variants pathogeic for chickens, because of hemagglutinin cleavage site changes. J. Viol. 64:3297-3303. Lyman, S. D. 1995. Biology of flt3 ligand and receptor. Int J. Hematol. 62:63-73. Mahley, R. W. 1998. Apolipoprotein E: cholesterol transport protein with expanding role in cell biology. Science 340:622-630. Marin, M., D. Noel, and M. Piechaczyk. 1997. Towards efficient cell targeting by recombinant retroviruses. Mol. Med. Today 3:396-403. Marin, M., D., Noel, S. Valselsia-Wittman, F. Brockly, M. Etienne-Julan, S. Russell, F.-L. Cosset, and M. Piechaczyk. 1996. Targeted infection of human cells via major histocompatabiiity complex class I molecules by Moloney murine leukemia virus-derived viruses displaying single-chain antibody fragment-envelope fusion proteins. J. Virol. 70:2957-2962. Marsh, M., and A. Pelchen-Matthews. 1994. The endocytic pathway and virus entry. In Cellular receptors for animal viruses (Wimmer, E. Ed), pp. 281- 301. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Martin, F. M„ Y, Kupsch, K. Y. Takeuchi, S. Russell, F.-L. Cosset, and M. Collins. 1998. Retroviral vector targeting to melanoma cells by single-chain antibody incorporation in envelope. Hum. Gene Ther. 9:737-746. Martin, J., S. A. Wharton, Y. P. Lin, D. K. Takemoto, J. J. Skehel, J. J., D. C. Wiley, and D. A. Steinhauer. 1998. Studies of the binding properties of Influenza Hemagglutinin receptor-site mutants. Virology. 241:101-111. Martin, F., J. Kupsch, Y. Takeuchi, S. Russell, F.-L. Cosset, and M. Collins. 1998. Retroviral vector targeting to melanoma cells by single-chain antibody incorporation in envelope. Hum. Gene Ther. 9:737-746. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 174 McClure, M. O., M. Marsh, and R. A. Weiss. 1988. Human immunodeficiency virus infection of CD-4-bearing cells occurs by a pH- indedepent mechanism. EMBO J. 7:513-518. McClure, M. O., M. A. Sommerfelt, M. Marsh, and R. A. Weiss. 1990. The pH independence of mammalian retrovirus infection. J. Gen. Virol. 71:767- 773. McKrell, A. J, N. W. Soong, C. M. Curtis, and W. F. Anderson. 1996. Identification of a subdomain in the Moloney murine leukemia envelope protein involved in receptor binding. J. Virol. 70:1768-1774. Millar, B. M. G., L. J. Caider, J. J. Skehel, and D. C. Wiley. 1999. Membrane fusion by surrogate receptor-bound influenza haemagglutinin. Virology 257:415-23. Molloy, S. S., L. Thomas, J. K. Van Shyke, P. E. Stenberg, and G. Thomas. 1994. Intracellular trafficking and activation of the furin proprotein convertase: locolization to the TGN and recycling from the cell surface. EMBO J. 13:18-33. Morling, F. J, K.-W. Peng, F.-L. Cosset, and S. J. Russell. 1997. Masking of retroviral envelope functions by oligomerizing polypeptide adaptors. Virol. 234:51-61. Morsy, J., W. Garten, and R. Rott. 1994. Activation of an influenza virus A/Turkey/Oregon/71 HA insertion variant by the subtilisin-like endoprotease furin. Virology 202:988-991. Nguyen, T. H., J.-C. Pages, D. Farge, P. Briand, and A. Webber. 1998. Amphotropic retroviral vectors displaying hepatocyte growth factor-envelope fusion protein improve transduction efficiency of primary hepatocytes. Hum Gene Ther. 9:2469-2479. Niles, W. D., F. S. Cohen. 1993. Single event recording shows that docking onto receptor alters the kinetics of membrane fusion mediated by influenza hemagglutin. Biophy. J. 65:171-6. Nilson, B. H. K., F. J. Morling, F.-L. Cosset, and S. J. Russell. 1996. Targeting of retroviral vectors through protease-substrate interactions. Gene Ther. 3:280-286. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 175 Nussbaum, O., A. Roop, and W. F. Anderson. 1993. Sequences determining the pH dependence of viral entry are distinct from the host range- determining region of the murine ecotropic and amphotropic retrovirus envelope proteins. J. Virol. 67:7402-7405. Ohuchi, R., M. Ohuchi, W. Garten, and H-D Klenk. 1991. Human influenza virus hemagglutinin with high sensitivity to proteolytic activation. J. Viol. 65:3530- 3537. Ohuchi, M., Orlich, M., Ohuchi, R., B. E. J. Simpson, W. Garten, H-D. Klenk, and R. Rott 1989. Mutations at the cleavage site of the hemagglutinin alter the pathogenicity of influenza virus A/Chick/Penn/83 (H5N2). Virology 168:274-280. O'Neil, K. T., R. H. Hoess, S. A. Jackson, N. S. Ramachandran, S. A. Mousa, S. A., and W. F. DeGrado. 1992. Identification of novel peptide antagonists for GPI lb/ll la from a conformationally constrained phage peptide library. Protein 14:509-515. Orlich M., H. Gottwald, and R. Rott 1994. Nonhomologous recombination between the hemagglutinin gene and the nucleoportein gene of an influenza virus. Virology 204:462-265. Orlich M., D. Khatchikian, A. Teigler, and R. Rott 1990. Structural variation occurring in the hemagglutinin of influenza virus A/Turkey/Oregon/71 during adaptation to different cell types. Virology 176:531-538. O'Toole T. E., J. C. Loftus, E. F. Plow, A. A. Glass, J. R. Harper, J. R., and M. H. Ginsberg. 1989. Efficient surface expression of platelet GPIIb-ISIa requires both subunits. Blood. 74:14-18. Patterson, S. M., R. Swainsbury, and E. G. Routledge. 1999. Antigen- specific membrane fusion mediated by the haemagglutin protein of influenza A virus: separation of attachment and fusion functions on different molecules. Gene Ther. 6:694-702. Pinter A, and W. J. Honnen. 1983. Comparison of structural domains of gp70s of ecotropic AKV and dual tropic MCF-247 MuLVs. Virology 129:40-50. Poul, M.-A., J. D. Marks. 1999. Targeted gene delivery to mammalian cells by filamentous bacteriophage. J. Mol. Biol. 288:203-211. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 176 Ragheb, J. A. and W. F. Anderson. 1994a. Uncoupled expression of Moloney murine leukemia virus envelope polypeptides SU and TM: a functional analysis of the role of TM domains in viral entry. J. Virol. 68:3207- 3219. Ragheb, J. A. and W. F. Anderson. 1994b. pH-independent murine leukemia virus ecotropic envelope-mediated cell fusion: implications for the role of the R-peptide and p12E TM in viral entry. J. Virol. 69:3220-3231. Ragheb, J. A. H. Yu, T. Hofmann, and W. F. Anderson. 1995. The amphotropic and ecotropic murine leukemia virus envelope TM subunits are equivalent mediators of direct membrane fusion: implications for the role of the ecotropic envelope and receptor in syncitium formation and viral entry. J. Virol. 69:7205-7215. Rogers, G. N., J. C. Paulson, R. S. Daniels, J. J. Skehel, I . A. Wilson, and D. C. Wiley. 1983. Single amino acid substitutions in influenza haemagglutinin change receptor binding specificity. Nature 304:76-78. Rothblat, G. H. 1972. Cellular sterol metabolism. In "Growth Nutrition and Metabolism of Cells in Culture," G. H. Rothblat and V. J. Cristofalo, eds., Vol 1, pp.297-325. Academic Press, New York. Rott, R., M. Orlich, H.-D. Klenk, M. L. Wang, J. J. Skehel, and D. C. Wiley. 1982. Studies on adaptation of influenza viruses to MDCK cells. EMBO J. 3:3329-3332. Roux, P., P. Jeanteur, and M. Piechaczyk. 1989. A versatile and potentially general approach to the targeting of specific cell types by retroviruses: application to the infection of human cells by means of major histocompatibility complex class I and class II antigens by mouse ecotropic murine leukemia virus-derived virus. Proc. Natl. Acad. Sci. USA 86:9079- 9083. Russell. S. J., R, E. Hawkins, and F. Winter. 1993. Retroviral vectors displaying functional antibody fragments. Nucleic Acids Res. 21:1081-1085. Sattentau, Q. J., S. Zolla-Pazner, and P. Poignard. 1995. Epitope exposure on functional oligomeric HIV-1 gp41 molecules. Viorology 206:713- 717. Sattentau, Q. J., J. P. Moore, F. Vignanx, F. Traincard, and P. Poignard. 1993. Conformational changes induced in the envelope glycoproteins of the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 177 human and simian immunodificiency vims by soluble receptor binding. J. Virol. 67:7383-7393. Sattentau, Q. J., J. P. Moore. 1991. Conformational changes indeuced in the human immunodeficiency vims envelope glycoprotein by soluble CD4 binding. J. Exp. Med. 174:407-415. Sauter, N., J. E. Hanson, G. D. Glick, J. H. Brown, R. L. Crowther, S-J Park, J. J. Skehel, and D. C. Wiley. 1992. Binding of influenza vims hemagglutinin to analogs of its cell-surface receptor, sialic acid: analysis by proton nuclear magnetic resonance spectroscopy and X-ray crystallography. Biochem. 31:9609-21. Sauter, N. K., M. D. Bednarski, B. A. Wurzburg, J. E. Hanson, G. M. Whitesides, J. J. Skehel, and D. C. Wiley. 1989. Hemagglutinins from two influenza vims variants bind to sialic acid derivatives with millimolar dissociation constants: a 500-MHz proton nuclear magnetic resonance study. Biochem. 28:8388-96. Schnierle, B. S., and Groner, B. 1996. Retroviral targeted delivery. Gene Ther. 3:1069-1073. Schnierle, B. S., D. Moritz, M. Jeschke, and B. Groner. 1996. Expression of chimeric envelope proteins in helper cell lines and integration into Moloney murine leukemia virus particles. Gene Ther. 3: 334-342. Somia, N. V., M. Zoppe, and I. M. Verma. 1995. Generation of targeted retroviral vectors by using single-chain variable fragment: An approach to in vivo gene therapy. Proc. Natl. Acad. Sci. USA 92:7570-7574. Schoen, P., L. Leserman, and J. Wilschut 1996. Fusion of reconstituted influenza virus envelopes with liposomes mediated by streptavidin/biotin interactions. FEBS Lett. 390:315-18. Sepp-Lorenzino, L., I. Eberhard, Z. Ma, C. Cho, J. Serve, F. Liu, N. Rosen, and R. Lupu. 1996. Signal transduction pathways induced by heregulin in MDA-MD-453 breast cancer cells. Oncogene 12:1679-87. Somia, N. V., M. Zoppe, and I. M. Verma. 1995. Generation of targeted retroviral vectors by using single-chain variable fragment: An approach to in vivo gene therapy. Proc. Natl. Acad. Sci. USA 92:7570-7574. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 178 Soneoka, Y., P. M. Cannon, E. E. Ramsdale, J. C. Griffiths, G. Romano, S. M. Kingsman and A. J. Kinsman. 1995. A transient three-plasmid expression system for the production of high titer retroviral vectors. Nucleic Acid Res. 23:628-633. Spiegel, M., M. Bitzer, A. Schenk, H. Rossmann, W. J. Neubert, U. Seidler, M. Gregor, and U. Lauer. 1998. Pseudotype formation of Moloney murine leukemia virus with Sendai virus glycoprotein F. J. Virol. 72:5269- 5302. Stegmann, T., J. M. White, and A. Helenius. 1990. Intermediates in influenza induced membrane fusion. EMBO J. 9:4231-4241. Stein, B. S., S. D. Gowda, J. D. Lifson, R. C. Penhallow, K. G. Bensch, and E. G. Engleman. 1987. pH-independent HIV entry into CD4-positive T cells via envelope fusion to the plasma membrane. Cell 49:659-668. Steinhauer, D. A. 1999. Role of hemagglutinin cleavage for the pathogenicity of influenza virus. Virology 258:1-20. Stieneke-Grober A., M. Vey, H. Angliker, E. Shaw, G. thomas, C. Roberts, H-D. Klenk, and W. Garten. 1992. Influenza virus hemagglutinin with multibasic cleavage site is activated by furin, a subtilisin-like endoprotease. EMBO J. 11:2407-2414. Takemoto, D. K., J. J., Skehel, and D. C. Wiley. 1996. A surface plasmon resonance assay for binding of influenza virus hemagglutinin to its sialic acid receptor. Virology 217:452-58. Tashiro, M., and R. Rott. 1996. The role of proteolytic cleavage of viral glycoproteins in the pathogenesis of influenza virus infections. Seminars in Virology 7:237-243. Tashiro, M., Y. Yokogoshi, K. Tobita, J. T. Seto, R. Rott, and H. Kido. 1992. Tryptase Clara, an activating protease for Sendai virus in rat lungs, is involved in pneumopathogenicity. J. Virol. 66:721-7216. Tuner, A. M., N. L. Lin, S. Issarachai, S. D. Lyman, and V. C. Broudy. 1996. Flt3 receptor expression on the surface of normal and malignant human hematopoietic cells. Blood 88:3383-90. Valselsia-Wittmann, S., F. J. Morling, B. H. K. Nilson, Y. Takeuchi, S. J. Russell, and F.-L. Cosset 1996. Improvement of retroviral retargeting by Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 179 using amino acid spacers between an additional binding domain and the N- terminus of Moloney murine leudkemia virus SU. J. Virol. 70:2059-2064. Vey, M., M. Orlich, S. Adler, H-D, Klenk, R. Rott, and W. Garten. 1992. Hemagglutinin activation of pathogenic avian influenza viruses of serotype H7 requires the protease recognition motif R-X-K/R-R. Virology 188:408-413. Vines, A., K. Wells, M. Matrosovich, M. R. Castrucci, T. Ito, and Y. Kawaoka. 1998. The role of influenza A virus hemagglutinin residues 226 and 228 in receptor specificity and host range restriction. J. Virol. 62:7626- 7631. Valsesia-Wittmann, S., F. J., Morling, B. H. K., Nilson, Y, Takeuchi, S. J. Russell, and F.-L. Cosset 1996. Improvement of retroviral retargeting by using amino acid spacers between an additional binding domain and the N terminus of Moloney murine leukemia virus SU. J. Virol. 1996; 68:4609- 4619. Valsesia-Wittmann, S., A. Drynda, G. Deleage, M. Aumailley, J.-M. Heard, O. Danos, G. Verdier, and F.-L.Cosset. 1994. Modifications in the binding domain of avian retrovirus envelope protein to redirect the host range of retroviral vectors. J. Virol. 68:4609-4619. Walther, W., and U. Stein. 1996. Targeted vectors for gene therapy of cancer and retroviral infections. Mol. Biotech. 6:267-286. Watowich, S. J., J. J. and Skehel. 1994. Crystal structures of influenza virus hemagglutinin in complex with high-affinity receptor analogs. Structure 2:719-731. Weiss. R. A. 1993. Cellular receptors and viral glycoproteins involved in retrovirus entry. In The retroviridae, 2nd edition (ed. J. A. Levy), pp. 1-108. Plenum Press, New York. Weis, W., J. H. Brown, S. Cusack, J. C. Paulson, J. J. Skehel, and D. C. Wiley. 1988. Structure of the influenza virus haemagglutinin complexed with its receptor, sialic acid. Nature 333:426-431. Wen, D., S. V. Suggs, D. Karunagaran, N. Liu, R. L. Cupples, Y. Luo, A. M. Janssen, N. Ben-Baruch, D. B. Trollinger, V. L. Jacobsen, S.-Y. Meng, H. S. Lu, S. Hu, D. Chang, W. Yang, D. Yanigahara, R. A. Koski, and Y. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 180 Harden. 1994. Structural and functional aspects of the multiplicity of Neu differentiation factors. Mol. Cell Biol. 14:1909-1919. Wharton, S. A., J. J. Skehel, and D. C. Wiley. 1986. Studies of influenza haemagglutinin-mediated membrane fusion. Virology 149:27-35. White, J. M. 1995. Membrane fusion: the influenza paradigm. CSHSQB 60:581-588. White, J. M. 1994. Fusion of influenza virus in endosomes: role of the hemagglutinin. In Cellular Receptors for Animal Viruses (Wimmer, E., Ed.), pp. 281-301. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. White, J. M. 1992. Membrane fusion. Science 258:917-924. White, J., M. Kielian, and Helenius. 1983. Membrane fusion proteins of enveloped animal viruses. Quarterly Review of Biophysics 16:151-195. Wiley. D. C., and J. J. Skehel. 1987. The structure and function of the hemagglutinin membrane glycoprotein of influenza virus. Annu. Rev. Biochem. 56:365-94. Williams, C. D., and J. Avigan. 1972. In vitro effects of serum proteins and lipids on lipid synthesis in human skin fibroblasts and leukocytes grown in culture. Biochm. Biophys. Acta 260:413-423. Wilson, C., M. R. Warded, K. H. Weisgraber, R. W. Mahley, and D. A. Agard. 1991. Three-dimensional structure of the LDL receptor-binding domain of human apolipoprotein E. Science 252:1817-1822. Wilson C. A., M. V. Eiden, W. B. Anderson, C. Lehel, and Z. Olah. 1995. The dual-function hamster receptor for amphotropic murine leukemia virus (MuLV), 10A1 MuLV, and gibbon ape leukemia virus is a phosphate symporter. J. Virol. 69:534-537. Wilson C. A. K. B. Farrell, and M. V. Eiden, 1994. Properties of a unique form of the murine amphotropic leukemia virus receptor expressed on hamster cell. J. Virol. 68:7697-703 Wilson, I. A., J. J. Skehel, and D. C. Wiley. 1981. Structure of the haemagglutinin membrane glycoprotein of influenza virus at 3 A resolution. Nature 289:366-72. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 181 Wu, B. W., J. Lu, T. K. Gallaher, W. F. Anderson, W. F., and P. M. Cannon. 2000. identification of regions in the Moloney murine leukemia virus SU protein that tolerate the insertion of an integrin-binding peptide. Virology 269:7-17. Yajima, T., T. Kanda, K. Yoshike, and Y. Kitamura. 1998. Retroviral vector tarting of human cells via c-Kit-Stem cell factor interaction. Hum. Gene Ther. 9:779-787. Zhao, Y., L. Zhu, S. Lee, L. Li, E. Changa, N.-W. Soong, D. Douer, and W. F. Anderson. 1999. Indentification of the block in targeted retroviral- mediated gene transfer. Proc. Natl. Acad. Sci. USA 96:4005-4010. Zhao, Y., L. Zhu, Benedict C. A., Chen, D., Anderson, W. F., and P. M. Cannon. 1998. Functional Domains in the retroviral transmembrane protein. J. Virol. 72:5392-5398. Zhu, N.-L., P. M. Cannon, D. Chen, and W. F. Anderson. 1998. Mutational analysis of the fusion peptide of Moloney murine leukemia virus transmembrane protein p15E. J. Virol. 72:1632-1639. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Asset Metadata

Creator Lin, Amy Hsiu-Ti (author)

Core Title Development of HA -pseudotyped retroviral vectors for cell -specific gene delivery

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 chemistry, biochemistry,OAI-PMH Harvest

Language English

Advisor Anderson, W. French (committee chair), Kasahara, Noriyuki (committee member), Stallcup, Michael R. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-81168

Unique identifier UC11337753

Identifier 3018100.pdf (filename),usctheses-c16-81168 (legacy record id)

Legacy Identifier 3018100.pdf

Dmrecord 81168

Document Type Dissertation

Rights Lin, Amy Hsiu-Ti

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