Studies of iron binding by the human serum protein, transferrin

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content STUDIES OP IRON BINDING BY THE HUMAN SERUM PROTEIN, TRANSFERRIN by- Ward Benjamin Davis A Thesis Presented to the FACULTY OP THE GRADUATE SCHOOL UNIVERSITY OP SOUTHERN CALIFORNIA In^Partial Fulfillment of the Requirements for the Degree MASTER OF SCIENCE (Biochemistry) January 1965 UMI Number: EP41355 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. IJM I Dissertation Pk&tbNng UMI EP41355 Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code ProQuest* ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Mi 4 8 1 0 6 -1 3 4 6 UNIVERSITY OF SOUTHERN CALIFORNIA T H E G R A D U A T E S C H O O L U N IV E R S IT Y P A R K L O S A N G E L E S 7 , C A L IF O R N IA ;0 o This thesis, written by WARD BENJAMIN DAVIS under the direction of h.%S...Thesis Committee, and approved by all its members, has been pre sented to and accepted by the Dean of The Graduate School, in partial fulfillment of the requirements for the degree of MASTER OP SCIENCE Dean D ate.... THESISCOMMITTEE Chairman 2-63— 2M— G 4 / ?99 ° TABLE OP CONSENTS Page INTRODUCTION . . ......... 1 STATEMENT OP THE PROBLEM ..... .................. b EXPERIMENTAL PROCEDURES. ........... 7 THEORETICAL 16 , DISCUSSION 18 ; SUMMARY............................................. 32 APPENDIX ........................................35 LIST OP BEPERENCES................................... bl ii LIST OF TABLES Table E jQ 1, Equilibrium Dialysis of Iron'^-Transferrin against Increasing Concentrations of EDTA at pH 7.If............................. 2. Calculated Values for Computing and K2 . Page . 15 . 39 iii IOTROKJCTIOH Recent experiments on the absorption of iron by mammals have shown that transferrin participates directly in the regulation and control of such absorption (1, 2). It has also been shown that the transport of iron from plasma to liver cells across the cell membrane is enhanced by low molecular weight chelates (3). In order to inter pret the role of transferrin and such exogenous chelates, it is of primaiy importance to know accurately the dis sociation constant (or constants) of the transferrin-iron complex. Physical and Chemical Proper ties of Transferrin Since its discovery (*+, 5)» an intensive investi gation of the chemical and physical properties of trans ferrin has been made, fransferrin (also called siderophi- lin) is the metal-binding, ^-pseudoglobulin component of plasma. Iransferrin has a molecular weight of 68,000 (6) and is an ellipsoidal molecule whose length to width ratio is *+.9 (170 A by 35 A). Although electrophoretieally a /^-globulin, transferrin is not precipitated by 50 per cent saturated ammonium sulfate and is soluble in distilled 1 ; water. The saturated transferrin-iron complex is more soluble than the apo-protein. Its isoelectric point is at ;a pH of 5.6 (7, 8). Pure transferrin can be crystallized and binds two atoms of ferric iron per molecule, which gives the color less protein a salmon-pink color. The spectrum of a 1 per cent water solution of the transferrin-iron complex at pH 7 shows an absorption maximum at a wavelength of **70 m/< and an extinction coefficient of 0,**8 to 0.55 when satu rated with iron (9). Transferrin can also bind copper and zinc in the same ratio of two atoms per molecule. The binding of copper and zinc, however, is much weaker than that of iron. Iron will displace both copper and zinc quite readily while copper will displace zinc. It has been shown that transferrin does not play a significant role in carrying either copper or zinc while it does play a key role in transport and utilization of iron (10, 11). The binding of iron to transferrin has been shown to be ionic with the iron bound in the ferric state. Perrous iron either is not bound or else yields a color less complex with transferrin (12). Carbon dioxide, in the form of bicarbonate ion, is required for the formation of the ferric-transferrin complex under aerobic conditions. One molecule of bicarbonate is required for each atom of iron bound to the protein, releasing two to three titratable hydrogen ions during the formation of the com plex (13)- Comprehensive reviews by Laurell (lH-, 15) of both the early studies and the more recent work are avail able. STATEMEFT OF THE PROBLEM It was desired that the strength of the binding of iron by transferrin be determined accurately for human transferrin. Early measurements of the binding capacity gave estimates of the dissociation constant for the com plex of 10-7 (i6), in a preliminary study of whole plasma, using the technique of equilibrium dialysis, Rubin, et al. (17) showed that iron is bound to transferrin with an 27 affinity constant greater than 10 • Since there was no accurate value for the binding of iron by transferrin in the literature, it was decided that such a determination should be made. The first step was to obtain pure human trans ferrin, without alterations which might impair its capacity for binding iron or introduce misleading artifacts. The second step was to evaluate the dissociation constant of the complex formed at each of the two iron-binding sites on transferrin. There are a variety of methods available for the isolation and purification of transferrin. One of the first sources of transferrin investigated was commercially available plasma fraction IY-S,^, isolated by the methods i f of Cohn, et al> (18). Samples of transferrin from this source were unsatisfactory since they required further purification and did not meet the first test for integrity,, that of full color development when saturated with iron. The method used by Laurell (8, 9) introduced ethanol as a precipitating agent for the transferrin-iron complex, making protein denaturation a possibility, 1 recently discovered method using a specific pre- ' cipitant for serum proteins was considered. This method ; (19)> originally designed for the preparation of pure gamma globulin by precipitation of all other serum proteins with 2-ethoxy-6,9“diaminoacridine lactate (Rivanol), was modi fied by Boettcher, Kistler and Hitschmann (20) to isolate transferrin from human plasma. At a pH of 9«*+, Rivanol quantitatively precipitates albumin while gamma globulin and transferrin remain soluble in the reagent (two parts of plasma to seven parts of a 0.H- per cent solution of Rivanol in water at room temperature), Charcoal adsorp tion removes the Rivanol and 25 per cent ethanol precipi tates the gamma globulins at a pH of 6.8 in the cold (-6° C). Pure transferrin precipitates in *K) per cent ethanol at pH 5.8 in the cold. The major objections to this method were the unknown effects that Rivanol might have on the properties of transferrin in binding iron, the high pH in the first step, and the introduction of ethanol as the precipitating agent. fiie isolation procedure finally chosen was a modi fication of the method of Sober (21) as follows. The gamma globulins were precipitated with half-saturated ammonium sulfate solution which was then removed by dialysis against running water. The remaining protein was concentrated by lyophilization and the transferrin separated from the other proteins by chromatography on diethylaminoethyl cellulose (DEAE cellulose) columns using gradient elution, fhe distinctive salmon-pink color of the transferrin-iron complex made it extremely simple to follow in the separa tion procedure, further purification was obtained by repeating the chromatography. fhe dissociation constants of the transferrin-iron complex were determined by equilibrium dialysis of the metal-protein complex against a solution of a chelate whose stability constant with iron had been determined previously. Ethylenediaminetetraaeetie acid (DBTA) was chosen for the small molecule chelate and the distribution of iron was followed by using radioactive iron (Fe^) as a tracer. EXPERIMENTAL PROCEDURES In a typical purification procedure, approximately 1 500 ml of pooled human serum was made 50 per cent saturated in ammonium sulfate by adding a saturated ammonium sulfate solution until the volume was twice the initial volume. The mixture was allowed to stand for about thirty minutes at room temperature and then it was centrifuged to remove ■ i the precipitated protein. The supernatant liquid was placed in Yisking dialysis tubing bags containing approximately 150 ml each and having a diameter of about one inch. The solutions were dialysed against running tap water for thirty-six hours, at which time a test sample of the supernatant gave 1 no positive test for sulfate. The solution was again centrifuged to remove traces of precipitate formed as a result of dialysis, placed in lyophilization flasks, frozen by swirling in a mixture of acetone and solid carbon dioxide, and lyophilized. DEAE cellulose chromatographic columns were pre pared by washing 100 gm of DEAB cellulose with four liters of distilled water, two liters of 1 M sodium chloride, two liters of 0.5 M sodium hydroxide, distilled water until 7 free of base, and then two liters of 100 per eent ethanol made 0.1 M in hydrochloric acid. Washing with two liters of 0,5 M sodium hydroxide and distilled water until neutral; was repeated, fhe DEAE cellulose was then washed with the starting buffer for the column until the pH of the super natant buffer remained at 7.23. fhe changes of volume which took place in the DEAE cellulose as the wash solutions were changed made it impossible to wash the DEAE cellulose on the column, fhe fluffy character and nearly equivalent density of the DEAE cellulose with the wash solutions made it impossible to filter it in the normal manner or to let it settle and decant the supernatant wash solutions, fhe washing pro cedure was made a great deal easier when it was discovered that a coarse cloth towel could be used as a filter in a large filtering funnel. fhe washed DEAE cellulose was suspended in the starting buffer and poured into a column made of 2 cm (I. D.) Pyrex tubing to a height of 60 cm. fhe buffers used in eluting the protein were all 0,02 M tris (2-amino- 2-hydroxymethyl-l,3“P**opanediol) and differed only in their sodium chloride concentrations which were 0.06 M, 0.08 M, 0.13 M, and 0.18 M, respectively. All the buffers were adjusted to a pH of 7.23 just prior to use. fhirteen grams of the lyophilized protein remaining after the ammonium sulfate precipitation were dissolved in , the minimum amount of starting buffer (about 75 ml of the 0.06 M sodium chloride buffer) and added to the DEAE cellu- I lose column. Air-tight connections from the top of the column were made to a mixing chamber containing 1900 ml of starting buffer and stirred by a magnetic stirring bar. Attached to the mixing chamber, and draining into it as the mixing chamber emptied, was a reservoir containing 700 ml of the 0.08 M sodium chloride buffer. Elution of the ! column was started and all connections checked so that the reservoir drained into the mixing chamber at the same rate that effluent was collected in test tubes by an automatic fraction collector. When the reservoir emptied, 700 ml of the buffer having the next higher sodium chloride concen tration was added to it, the connections re-checked, and the elution continued. Transferrin was the first component to come off the column, as shown by the correlation between the salmon-pink color and the protein concentration measured by the biuret reaction on aliquots of the collected fraction. The con tents of the test tubes with visible salmon-pink color were pooled in three fractions. The first contained the bulk of the color while the second and third fractions contained the remainder, the tail of the transferrin peak. The first fraction ultimately yielded the transferrin used for the determination of the dissociation constants of the trans ferrin-iron complex, although the second and third frac tions also were purified. Each of the fractions was dialysed against distilled water at *f°C until free of chloride, lyophilized, redissolved in starting buffer and purified by chromatography on a fresh column of DEAE cellu lose, prepared as before, The protein peak corresponded to t the peak in salmon-pink color and yielded over 50 mg of transferrin after dialysis and lyophilization. Starch gel electrophoresis by the method of Smithies (22) yielded a single component which corresponded; to the ^-globulin of a duplicate electropheragram using pooled human serum, fhe gel strip was sliced in half lengthwise before staining and one-half stained for pro tein, fhe other half was stained for iron, fhe protein band on the transferrin gel was the same as the band which was stained with potassium ferrocyanide in dilute hydro chloric acid to give the faint blue color due to iron at the site of the protein, fhe extinction coefficient of a solution of the purified transferrin, calculated for a 1,00 cm path and a 1 per cent solution, was 0,52 at k?0 m/t, fhe first step in the preparation of the trans ferrin for determination of the dissociation constants of the complex was to remove the endogenous iron and replace it with radioactive iron so that the iron concentration co-old "be determined in solutions containing it at far below the chemically detectable limits. All subsequent manipula tions of the transferrin were carried out in solutions made up from iron-free distilled water and at temperatures never; in excess of 5° C. Forty-five milligrams of purified transferrin were dissolved in 12.0 ml of 0.2 M phosphate buffer at pH *f.5 which made the transferrin-bound iron labile. The iron was removed by adding 1.5 ml of 0.002 'M sodium citrate (which binds iron effectively at this pH), allowing the solution to stand ten minutes, and adding 1.95 grams of prepared ion exchange resin (IRA *+01). Preparation of the anion exchange resin consisted of washing it with 2 H hydrochloric acid to convert it to the chloride form, rinsing it with iron-free distilled water, to remove the excess hydrochloric acid and suspend ing it in Teronal buffer (O.OMf M in Veronal and 0.11 M in sodium chloride) at pH 7.5 until time for use. Adding the anion exchange resin effectively removes the iron-citrate complex from solutions of pH *f.O to 8.0 without affecting transferrin. The protein solution was decanted from the resin, the resin washed three times with 0.5 ml of ion-free water, and the washings added to the protein solution. Superflu ous ions were removed and the pH brought to 7.23 by dialysis against bicarbonate buffer until free of phosphate. After 12 ! treatment in this fashion, the protein solution was com pletely colorless. 59 Solutions of radioactive iron (Fe y) with carrier iron were prepared from ferric chloride (FeCl- 3*6^ 0) in 59 0.002 M sodium citrate at pH 7»b by adding 5#1 of Fe Cl 3 solution to 5 ml of the iron-citrate solution, fhe total iron concentration was measured by the method of Peters (23) and the specific activity of the solution, 5,200 counts per minute per /tgm of iron, was determined in a scintilla tion well counter with pulse height analyzer. All solu tions used subsequently were made 0,0b M in bicarbonate and saturated with a mixture of 5 per cent carbon dioxide- 95 per cent oxygen at atmospheric pressure to ensure a pH of 7»b (verified by measurement on a pH meter). fhe iron- citrate solution was added to the protein solution and the salmon-pink color returned immediately. Excess iron- citrate was removed by addition of ion exchange resin as before, but at pH 7»b so that none of the transferrin-bound iron was removed. Four portions (^.0 ml each) of the labelled trans ferrin solution, 0,0b M in sodium bicarbonate, were pipeted into separate beakers, fo portion A, 1 ml of 0.0b- M sodium bicarbonate was added to dilute the transferrin solution to the same final concentration as each of the other portions, fo portion B, 0.5 ml of 0.01 M EDfA, also O.Ob M in sodium bicarbonate, was added to make the final concentration of EDfA 0.001 M and 0.5 ml of 0.0*+ M sodium bicarbonate to bring the total volume to 5.0 ml. fo portion C, 0.6 ml of i 0.0*f M sodium bicarbonate and O.^f ml of 0.125 M EDfA, 0.0*+ M in sodium bicarbonate, were added to make the final con centration of EDfA 0.01 M. fo portion D, 0.9 ml of 0.0M- M sodium bicarbonate, 0.09 ml of 6 I sodium hydroxide (which • contained no added bicarbonate), and 186 mgm of EDfA were | added to yield a final EDfA concentration of 0.1 M at pH 7.*+. All EDfA solutions were prepared from the disodium, dihydrate salt of EDfA. All portions were at a final pH ; of 7.*+ and identical total bicarbonate concentration, fhe molar ionic strengths of the portions were, respectively: /<&, o.o1 * ; /%* o.o5; o.6. Duplicate 2.0 ml aliquots of portions A, B, 0, and D were pipeted into Visking dialysis tubing bags which were carefully tied off, rinsed, and each one suspended in 100 ml of solution containing 0.0b M sodium bicarbonate and the same concentration of EDfA as inside the bags, fhe pH was maintained between 7.3 and 7.b throughout the subse quent equilibration and mixed by constant aeration with a gas mixture containing 5 per cent carbon dioxide and 95 per cent oxygen. Periodically, 3.0 ml samples of the dialysate were removed and the activity determined in the scintillation well counter. At the end of 3& hours, maximum activity in the dialysate was reached and the equilibration was stopped i at the end of *+2 hours. The bags were rinsed, blotted, and 1 weighed. Their contents were removed and an aliquot from each counted. The empty bags were then weighed and the volume of solution calculated from the difference in weights. The protein concentration at the end of the experiment was determined by the biuret reaction. A sum mary of the data is presented in Table 1. TABLE 1 EQUILIBRIUM DIALYSIS OF mGU^-TRANSFERRIM AGAINST INCREASING CONCENTRATIONS OF EDTA AT pH 7 A Components Experimental Condition 1 2 3 EDTA Cone. (M) 0 10-3 o i ro H O 1 H Cpm/ml in bag U>-- -r o o 1750 1280 b65 ^Inside bag 10.6 5.7 M-.O lA Cpm/ml (outside bag) 0 38 bl 59 ^Outside bag ^ 0 0.12 0.13 0.19 (Tr>total ^M) 5.5 5.5 5.5 5.5 THEOEETICAL Equilibrium dialysis was chosen for the determina tion of the dissociation constants of the transferrin-iron complex since it offered the most rapid and accurate means of finding these values, fhe equations relating the bind ing of iron to transferrin and to EDTA are as follows (The abbreviations used are: T for apo-transferrin, T-Fe for half-saturated transferrin, T-E^ for saturated transferrin, L for the EDTA ligand, L-Fe for EDTA-iron chelate, and Fe i for free ferric ion): (1) I-Pe2 = T-Fe + Fe ^ . (2) f-Fe = T + Fe K0 = ^ (T-Fe) _ (L)(Fe) ' ! > (L-Fe) The constants of equations (1) and (2) are valid only for the specific pH and bicarbonate concentration used in these experiments. The involvement of hydrogen ion and bicarbonate ion is included implicitly in the values for Kq and ^2- The total transferrin is restricted to the inside (3) D-Fe = L + Fe Kr = (LHfe) 16 1 7 : of the dialysis bag and occurs as: M (3?)total = (T) + (T-Fe) + (T-Fe 2) The total iron concentration inside the bag is present as: (5) (Fe)bag = C^e) + (L-Fe) + (T-Fe) + 2 (T-Fe2) j The only iron in the solution outside the bag is: (6> ‘^ o u t s i d e = + Combining the data from Table 1 with the relation ships described above and including the effects of pH, it can be shown (see Appendix A for detailed calculations) —P7 that the values of and K2 are, respectively, 5.6 x 10 ' ■ and 2.5 x 10”28. DISCUSSION In order to place this work on transferrin in proper perspective, it is necessary to discuss the overall 1 problem in which transferrin is a significant factor. The ! , important areas to he discussed are regulation of iron absorption, the metabolism of iron after it is absorbed, I and the structural features of transferrin which give it ! such a unique capability for binding iron. ; The transport of the ions K , Fa , Ca , and Mg I across living cell membranes against concentration gradi ents is directly coupled to metabolic energy. This involve1 ment of metabolic processes has given rise to the concept of active transport for these ions. Their uptake is regu lated by inhibition or activation of this energy-dependent mechanism. Another mechanism by which ions can accumulate against an apparent concentration gradient by a passive, diffusion process has been proposed for iron, copper, and zinc (2*+). The uptake of these ions has been shown to be independent of metabolic energy.. The temperature depend ence of the rate of uptake suggests ah energy of activation appropriate for physical diffusion rather than for the 18 chemical reactions of metabolism, fhe mechanism proposed | is a two step process, fhe ions first diffuse passively across the cell membrane to sites on a specific ion-binding entity, fhey then undergo reactions which rapidly bind them to the specific sites, fhese binding reactions take place rapidly as compared with the transport rate. Regulation of absorption by this diffusion process is achieved by control of the concentration of available binding sites, the strength of binding for each ion by the specific sites, and the competition of other ions for the site. In addition, the process of diffusion across the cell membrane may involve a complex of the ion with some diffusible, low molecular weight, binding entity which enables the ion to pass through the membrane. Oxidation- reduction reactions may affect the rate of uptake since the binding of the ion is specific for a given oxidation state. Early proposals by Hahn (25) and Granick (26) gave the first evidence that the intestinal mucosal cells were important regulators of iron absorption. The mechanism proposed by Hahn and elaborated by Granick became known as the mucosal block and made the presence of apo-ferritin a prerequisite for facilitated iron transport. As Crosby points out in his editorial (27), the experiments of Granick actually demonstrate a failure of the mucosal block rather than its existence. A recent series of papers by Crosby and co-workers (28, 29, 30, 31) Has clearly shown : involvement of the intestinal mucosa but in a way not i. anticipated by Granick in his detailed theory of the mucosal block, Crosby postulates two pathways for iron absorption.: The first is by means of a rapid transport system whose capacity is dependent on the state of iron deficiency. i Iron deficiency increases the capacity of this system while j iron loading inhibits it. This rapid transport system works effectively for only about 1-2 hours. Hypotheses suggested to explain the termination of this rapid phase of absorption are: Formation of non absorbable iron complexes in the lumen of the gut (32), movement of the iron from the duodenum (where absorption takes place readily) to the jejunum (where iron is not absorbed readily), and saturation of transferrin so that iron can enter the blood only as it is removed from trans ferrin in the iron storage areas (33). The second stage of iron absorption postulated by Crosby involves a receptor system in the mucosal cells which puts the absorbed iron into a bound, slowly-released form, probably as ferritin. Iron which is not absorbed from the cells into the blood is sloughed off with the mucosal cells as they migrate to the tip of the villi. Control of absorption in the second stage is effected either hy inhibition of the receptor system when iron i stores are deficient or activation when iron stores are ' loaded. Studies by Charley (3*+) on the appearance of radio iron in the blood of rabbits after administration of iron into isolated duodenal loops also indicated failure of the mucosal block. Radio-iron, administered as ferrous sulfate or ferric chelates, appeared in the blood within minutes after being injected into the gut. The same initial phases ; i I of rapid, direct absorption of iron observed by Crosby was , seen in this work. The iron concentration in the blood ' i reached a maximum in a few hours. Investigations of the gut contents after administra^ tion of doses of iron led to the discovery of an endogenous chelate which is specific for ferrous iron (3*0. Ixperi- i ments by Pirzio-Biroli (35) on the effect of intestinal secretions in the absorption of iron have shown that the intestinal mucosal secretions from iron-deficient animals enhance absorption when compared with secretions from normal or iron-loaded animals. Recent work by Sarkar (36) has verified the existence of such endogenous chelates and their specificity for ferrous iron. Studies by Saltman, et al.. (3? 3^, 37) on how iron-chelate structures are related to their ability to facilitate transfer of iron across cellular membranes, both 'in vivo and in vitro, lead to the conclusion that iron can be moved without invoking active transport. A way in which transferrin may play a role in the regulation of iron absorption has been proposed by Crosby (31). The equilibrium distribution of iron among storage : sites, blood, and the intestinal mucosa is disturbed by iron loss or loading. This equilibrium is restored rela tively slowly. Over long periods under normal conditions, however, the dynamic equilibrium is maintained. The dis covery that iron can enter the mueosal cells from the blood (29) as the cells are formed and that the amount of iron picked up is directly proportional to the serum iron level suggests a way for transferrin to act in the control of iron absorption. The most probable role of transferrin is that of an information-carrying messenger, the information : regarding the state of iron stores being reflected by the relative saturation of transferrin with iron. This infor mation is transferred to the newly formed mueosal cells as a proportionate amount of iron incorporated in them. The fact that iron deficiency and iron loading affect the absorption of iron can be explained by proposing that the secretion of endogenous chelates, the activity of the rapid-transport system, and the activity of the receptor system which binds iron in the mucosal cells are all pro portional to the amount of iron taken up from transferrin , . _ by the mucosal cells as they form. The systems which pro- ; mote iron absorption would be inhibited or activated and ;the systems trapping iron in the mucosal cells would be activated or inhibited by the presence or absence of iron incorporated in the cells by uptake from transferrin in the blood at the time of their formation. The lag time between administration of iron or loss :of iron by bleeding and its effect on iron absorption is explained by the fact that those cells currently controlling iron absorption require about 72 hours to migrate to the villi tips and be replaced by cells which reflect the changed need for iron. It is implicit in Crosby*s assumptions that iron is absorbed through the mucosal cell wall by processes of active transport. Passive diffusion of iron in the form of neutral chelates seems more likely in view of the previ ously cited work by Saltman, et al. The major change required by processes of passive diffusion is that control would take place by changes in the relative numbers of mobile binding entities, such as the endogenous chelate, and the fixed binding entities, such as apo-ferritin. Increasing the number of fixed sites relative to mobile sites would favor trapping of the iron within the cell and subsequent loss. The work of Wheby and Jones (33) shows that the relative level of iron-saturation of transferrin has no direct effect on the absorption of iron from the gat. , However, the absorption of iron is a dynamic process in which iron is removed from transferrin in storage sites. The transferrin from which iron was removed would then be free to accept more iron, even if nearly 100 per cent of the transferrin in the blood were iron-saturated. On the other hand, if the acceptor system of the storage sites became saturated, then all of the transferrin would soon become saturated and, being unable to accept more iron, would block further uptake from the mucosal cells. Iron continuing to move into the mucosal cells would then be incorporated into ferritin and lost as the cells sloughed off. The two iron atoms attached to transferrin are not equivalent (38) although the binding sites are identical and separate so that no direct interaction takes place (39, ifO). The first atom bound to transferrin changes the environment of the second site so that the second iron bound is kinetically more labile (*§-0). The normal level of iron-saturation in serum is less than 50 per cent (1*0. These facts suggest that the transferrin-iron complex with one site occupied may play the primary role in activating or inhibiting the various mechanisms in the mucosal cells which are involved in iron absorption while the second iron : 25 ;site, being occupied by a more labile iron ion, may play 1 the primary role in transporting iron from the mucosal . cells to storage sites at those times when iron is present : . in the gut and absorbed. The exceptionally strong affinity of transferrin for iron is reasonable when one considers the following facts. In the presence of air, more than 90 per cent of the iron reaching the gut is present as ferric iron, even in the presence of food (*+1). At the pH found in the duo denum, ferric iron is so insoluble that it is quantita tively precipitated. In order to be successful in solubil izing iron and keep it in a form suitable for entry into the mucosal transport system, endogenous chelating agents would have to bind iron more strongly than the hydroxide gel. Transferrin, as the only iron carrier in the blood (^2), would need to be even more effective in its binding of iron to remove it from the mucosal cells. Another important phase of iron metabolism is the uptake and utilization of iron by the reticulocytes to form, hemoglobin. Transferrin is involved directly and specifi cally in this transfer (^3). Iron is removed from trans ferrin and incorporated into hemoglobin in the reticulo cytes without destruction of the transferrin (Mf). The specificity of this transfer is so exact that transferrins from different species are not interchangeable (*f5). Transfer of iron from transferrin to storage sites in the liver, however, is not species specific. Dialysis ' experiments give evidence for the involvement of chelates as intermediates for iron transfer. Collision-specific interchange directly from transferrin to the liver is not necessary O^). Recent work on the kinetics of iron exchange from chelates to transferrin suggest a chelate- iron-transferrin intermediate, analogous to an enzyme- substrate combination (37)• Studies on the effect of chelating agents on the uptake of iron by reticulocytes, in vitro, during hemoglo bin synthesis show that the powerful affinity of trans ferrin for iron increases the efficiency of iron utiliza tion by reducing the nonselective binding of ionic iron to the membrane structure (^6, *f7). The protein nature of transferrin also may promote iron utilization by its effect on the environment and charge of the metal ion, enabling it to approach transport sites on the reticulocyte membrane surface (*+7). Human transferrin shows a number of genetically determined forms, differentiated by the migration pattern upon gel electrophoresis (*+8, **9, 50). These differences do not change the binding of iron (5l). They are caused by the different number of sialic acid residues attached to transferrin. Removing the sialic acid with neuraminidase alters the mobility of transferrin without affecting its antigenic properties or iron-binding characteristics (53) so that it appears unlikely that sialic acid has any ma;jor ; effect on the roles played by transferrin as discussed above. Transferrin is also inherently interesting because of the strength with which it binds iron. What groups are involved in this unique ability and how are they affected by their presence in the protein structure to so greatly enhance their hold on ferric ions? An early observation by Dutcher (53) that the red color of aspergillic acid with ferric chloride in methanol is due to the binding of iron by the cyclic hydroxamic acid structure led Fiala and Burk (5*0 to compare the iron complexes of transferrin, conalbumin, hydro xylamine, and aspergillic acid. The similarity of wavelengths at which solutions of these compounds absorb light maximally, the common involvement of bicarbonate in the formation of their complexes with iron, and the fact that copper is bound by each, led to the proposal that iron is bound to each by means of the hydroxamate group as follows: C = 0-. I >e N - O'- ' (H) Differences in the ferric complexes among these compounds iwere ascribed to the effect of differing side chains attached to the active site. The discovery by Emery and leilands (55) that the : iron-binding center of the ferrichrome compounds also involved the hydroxamic acid structure strengthened the case for hydroxamate involvement in the iron-binding site : of transferrin and eonalbumin. The ferrichrome complexes bind iron with a strength comparable to transferrin as they also are not decomposed by EDTA. Another similarity exists in the fact that copper is bound by the iron-free ferrichromes. Warner and Weber (56) opposed the hydroxamate structure on the basis of their work with eonalbumin, the iron-binding protein of egg white which is quite similar to transferrin. They proposed that tyrosine phenolic groups were responsible for the binding of iron. The evidence for their proposal was found by examining the ionization of the iron-binding site at a variety of pH*s and by comparing the colors of the eonalbumin and transferrin complexes with iron and copper with those of the hydroxamic acid complexes with iron and copper. Their titration data showed that three phenolic hydroxyls were involved in the active center. The fact that the stability of the ferrichrome-iron com plexes is decreased at more basic pH*s while the stability of the transferrin-iron complex is increased also argues against the involvement of the hydroxamate structure in transferrin. The hydroxamic acid structure was conclusively shown to be unfeasible by Fraenkel-Conrat's work with model compounds (57) and his inability to find the hydroxamate structure in eonalbumin. The specificity and stability of : ; the iron-eonalbumin complex must be due only to the peptide- chain folding of the usual amino acid groups in such a way that their spatial relationships result in the special ; binding properties observed for the protein. Avian conal- • bumln has been shown to be immunologically identical to ' avian transferrin and their amino acid compositions proba- ' bly are identical (58). Many of the electronic properties : of the binding sites of eonalbumin and transferrin have been shown to be similar (39 9 *+0). Since transferrin and eonalbumin share so many properties, the conclusion drawn about the binding site structure of eonalbumin by Fraenkel- Conrat is also valid for transferrin. More detailed studies of the properties of the binding site for iron on eonalbumin and transferrin have been reported recently (59). Equilibrium dialysis studies with citrate in which equilibrium is approached from both sides show that the two sites are independent and equiva lent in their binding affinity. Careful titrations verify that three hydrogen ions are displaced for each iron bound 30; and that the pH dependence of the displacement strongly suggests phenolic hydroxyl groups as first proposed by i Warner and Weber. Electronspin resonance investigations on: \ ! the copper-transferrin eomplex suggest the involvement of two nitrogen ligands. Aasa, et al. (39) propose that the imidazole group of histidine furnishes the nitrogen while Windle, et al. (*f0) claim that the guanidyl group and the €-amino nitrogen of lysine cannot be eliminated as possi- ; bilities. Both Aasa and Windle agree that the binding sites are separated by at least 9 Angstroms and are inde pendent with regard to magnetic interaction. Windle, et a l . ' propose the following structure for the binding of ferric irons hco3 Since the saturated transferrin-iron complex can be crystallized, X-ray diffraction studies would be useful in elucidating the structure of the binding site. Schultze, et al. (60) have determined the amino acid 31 composition of transferrin. Azari and Feeney (6l, 6*f) have reported that the transferrin-iron complex resists pro teolysis. It would be interesting to attempt to partially degrade the protein, without destroying the complex, and then determine the amino acid sequence by the methods of Smith and Margoliash (62, 63) in the portion retaining iron-binding properties. In order to verify the role of transferrin in the regulation of iron absorption, one might be able to compare the absorption in normal animals with animals in which the iron-saturation of transferrin has been maintained by parenteral infusion and the iron stores diminished by bleeding. If one could maintain a transferrin-iron level which would reflect an apparent loading of iron while the animals actually were made iron-deficient, blocked iron absorption in these animals would be strong evidence in favor of the proposal by Crosby. Enhanced iron absorption in spite of the maintenance of high transferrin-iron levels would argue that the role of transferrin is simply as a carrier of iron, with control of absorption lying directly in the storage sites and their acceptor systems. SUMMARY A method of isolating pure, human transferrin has been described, The dissociation constants of the trans ferrin-iron complex have been determined, pK^ being 26.3 and P&2 being 27.6. The involvement of transferrin in the ; control of iron absorption has been described. Control of ; t iron absorption may lie in the storage sites with trans ferrin playing a passive role, merely reflecting the loaded state of acceptor systems in the storage sites by becoming saturated with iron. Transferrin may act as an information- carrying messenger to the mucosal cells which would act directly to control iron absorption. An experiment to dis tinguish between these possibilities has been suggested. The fact that the two iron atoms are not identically bound to transferrin suggests different roles for the half saturated transferrin and the saturated transferrin. The protein nature of transferrin assists the specific transfer of iron to reticulocytes for hemoglobin synthesis. The high affinity of transferrin for iron pre vents nonselective binding of iron to the membrane struc tures of the reticulocytes. Molecular structures proposed for the iron-binding 32 site on transferrin have been discussed. Further experi ments have been suggested to determine the amino acids involved in the iron-binding site and the structural con figuration responsible for the binding of iron. I APPEIBIX APPENDIX CALCULATION OP THE DISSOCIATION CONSTANTS OP THE TRANSFERRIN-IRON COMPLEX The dissociation constants for the transferrin-iron; complex are and K2> defined by equations (1) and (2) on page 16. Equation (3) defines the equilibrium established between the free ferric ion concentration, the free EDTA concentration (as the tetravalent ethylenediaminetetra acetate ion), and the concentration of the negative, singly charged, iron-EDTA complex. Since the iron-EDTA complex undergoes hydrolysis and the form of the EDTA itself is dependent on pH, the free iron concentration is determined by the relationships among the following equilibria (65, 66) : t Ti X n w (L)(Pe) — — — 25.1 (3) L-Pe = L + Pe Kr = — ---- =10 h (L-Pe) (7) H+ + L = HL K7 = — = 10+10,26 / (H ) (L ) (8 ) H* + HL s H2I Kg = = 10+6-16 (9 ) H4" + H2L = HqL Kq = — — = 10+2*67 3 ^ (h+)(h2l) (1 0 ) H+ + H^L = % L K10 = (HtfL) = 10+2.00 (h+)(h3l) 35 Combining equations (7) to (10), the only signifi- : cant forms of "L” (the total EDTA concentration) are H2L ; and EL since the ratio (H^L)t(H2Ij)s(HL)t(L) equals Hl0°sl0“2,86. The amount of HL*» bound to : iron is small compared with the total so the concentration : of H2L and HL can be calculated from the above ratio and the total "L“ concentration. Substituting the calculated value of (HL) and the measured value of (H+) into equation (7), the concentration of free EDTA, (L), can be determined for each experimental condition. The complex of iron with EDTA, L-Fe, also undergoes hydrolysis and the concentration of L-Fe must be found from the relationships1 (11) L-Fe + OH- = L-Fe-OH Klx = (L-Fe-OHl. = 10+6.*+5 (L-Fe)(OH ) (12) (Fe) . .. = (Fe) + (L-Fe) + (L-Fe-OH) 0llbSiCl6 The free iron concentration can be assumed to be negligible compared with the other two forms of iron. At a pH of 7.H-, the concentration of L-Fe at each of the total EDTA concentrations, calculated from the data of -6 Table 1 and equations (11) and (12), is? 0.11 x 10 at 0.1 M EDTA, 0.076 x 10~6 at 0.01 M EDTA, and 0.070 x 10”6 at 0.001 M EDTA. Substituting the values for the concentrations of 37 L and L-Fe into equation (3)? the free ferric ion concen tration, (Fe), is found to be* 6,7 x lO*"2^ at 0.1 M EDTA, b.6 x 10~28 at 0.01 M EDTA, and ^-.2 x 1(T27 at 0.001 M EDTA, verifying that no error is introduced by assuming that the free ferric ion concentration is negligible compared with the bound iron. In order to determine the constants and K2, it is necessary to develop an independent method of relating them. It is convenient to define as follows* a s ) k3 = K i k 2 = (T-Fe2) The fraction of free transferrin, found by combining equa tions (2), (*+), and (13) is: { 1 W m ,---------- 1 ---------- ^'total _ (Fe) (Fe)2 x + + k2 The iron bound to transferrin is the difference between the total iron concentration in the bag and that outside the bag. Subtracting equation (6) from equation (5) yields: (15) (Pel . . m = (T-Fe) + 2 (T-Fe?> bound to T ^ Combining equation (15) with equations (2) and (13)5 the iron bound to transferrin can be expressed as: ^ (Fe)(X) 2 (Fe)2(T) (16) (*e)bOTma t0 T = — --- + ---------- 2 3 Writing the average number, R, of iron atoms bound to transferrin? (17) H = ( 18) Abound to f (Fe)(f) 2 (Fe)2(T) — —1 1 <T>total *2 ' k3 ^^otal (Fe)2(R - 2) *3 R k2 (Fe)(R - 1) K- Using the data developed in the preceding steps and given in fable 1, the set of values listed in fable 2 can be computed. Replacing by ^2^1 en^er^nS the appropriate data from fable 2 into equation (18)? (19) At 0.1 M EDTA, -3.8 x 10“^6 = -^(-2.5 x 10~28) - K1K2 (20) At 0.01 H EDTA, -38*^ x 10~56 = -%(-1.9 x 10~28) Subtracting equation (20) from (19) and solving yields equal to 5.6 x 10 2^. Substituting this value for K-^ and —28 solving for E2 yield K2 equal to 2.5 x 10 . The data at 0.001 M EDfA is incompatible with the results at the other two concentrations. Aasa, et al. (39) indicate that for this condition, equilibrium is not established in the time allowed even though the level of radio-iron outside the bag was no longer increasing measurably. TABLE 2 CALCULATED YALUES FOR COMPUTING K± and K2 Experimental Condition if 3 2 Total EDTA eonc 0.1 0.01 0.001 (L) 1.32 x 10"^ 1.32 x 10~5 1.32 x 10"6 (L-Fe) (x 106) 0.11 0.076 0.070 (Fe) 6.7 x 10"29 if.6 x 10”2^ if.2 x 10“27 R 0.21 0.71 1.02 (R-l) -0.7*+ -0.29 +0.02 (R-2) -1.79 * ”1.29 -0.98 (Ve}{R-l2 (x 1028) R -2.5 -1.88 +0.82 (Fe)2(R-2) (x 1056, R -3.8 -38 • i f 1700 — _ T I LIST OP HBPBIES'CES LIST OF REFERENCES 1* Saltman, P., Fed. Proc., 20, Supp. 10, 156 (1961). 2. Hallberg, L. and Solvell, L., Acta Mediea Scand., 168, Supp. 358, 3 (I960). 3. Charley, P. J., Rosenstein, M., Shore, E., and Saltman, P., Arch. Biochem. Biophys., 88, 222 (I960). *f, Holmberg, C. G. and Laurel-1, C-B., Acta Physiol. Scand., 10, 307 (19*4-5). 5. Schade, A. L. and Caroline, L., Science, lO^f, 3**0 (I960). 6. Charlwood, P. A., Biochem. J., 88, 39*+ (1963)* 7. Oncley, J. L., Scatchard, G., and Brown, A., J. Phys. Coll. Chem., £1, 18*+ (19*+7). 8. Laurell, C-B. and Ingelman, B,, Acta Chem. Scand., 1, 770 (19*4-7). 9. Laurell, C-B., Acta Chem. Scand., 7, 1*1-07 (1953). 10. Surgenor, L. M., Koechlin, B. A., and Strong, L. E., J. Clin. Invest., 28 , 73 (19*+9). 11. Holmherg, C. G. and Laurell, C-B., Acta Chem. Scand., 1, 9Mt (19^7). 12. Ehrenberg, A. and Laurell, C-B., Acta Chem. Scand., 2, 68 (1955). 13. Schade, A. !., Reinhart, R. W., and Levy, H., Arch. Biochem. Biophys., 20, 170 (19*+9). l*f. Laurell, C-B., Pharm. Rev., ^f, 371 (1952). 15. Laurell, C-B., in The Plasma Proteins (Frank; W. Putnam, Ed.), Vol. 1, Chapter 10, p. 3*+5, Academic Press, Hew York (I960). *+1 16. Coton, E. J., in Chemical Specificity in Biological Interactions (1*. R. H. Gurd. Id.), p. 7. Academic Press, New York (195*+). 17. Rubin, M,, Houlihan, J., and Princiotto, J. V., Proc. Soc, Expt'l. Biol. Med., 103. 663 (I960). ; 18. Goton, E. J., Strong, 1. E., Hughes, W. L., Jr., Mulford, D. J., Ashworth, J. N., Melin. M., and Taylor, H. 1., J. Am. Chem. Soc., 08, 459 (1946) 19. Horejsi, J. and Smetana, R., Acta Mediea Scand., 155, 65 (1958). 20. Boettcher, E. W., Kistler. P., and Hitsctomann, Hs., Mature, 181, 490 (1958). 21. Sober, H. A., Gutter, P. J., Wyckoff, M. W., and Peterson, E. A., J. Am. Chem. Soc., 78, 756 (1956). 22. Smithies, 0., Biochem. J., 6l, 629 (1955). 23. Peters, T., Giovanniello, T. J., Apt, 1., and Ross, J. P., J. lab. Clin. Med., 48, 280 (1956). 24. Saltman, P., Proc. Second International Conf., Peaceful Uses of Atomic Energy, 24, 152 (1958). “ 25. Hahn, P. P., Bale, W. P., Lawrence, E. 0., and Whipple, G. H., J. Exp. Med., 6^, 739 (1939). 26. Graniek, S., J. Biol. Chem., 164, 737 (19*+6). 27. Crosby, W. H., Blood, 22, 44l (1963). 28. Hartman, R. S., Conrad, M. E., Jr., Hartman, R. E., Joy, R. J. T., and Crosby, W. H., Blood, 22, 397 (1963). 29. Conrad, M. E., Jr. and Crosby, W. H., Blood, 22, 406 (1963). 30. Wheby, M. S. and Crosby, W. H., Blood, 22, 4l6 (1963). 31. Conrad, M. E., Jr., Weintraub, L. R., and Crosby, W. H. J. Clin. Invest., 4^, 963 (1964). : ' ~ " 43 : 32. Duthie, H. L., Code, D. I1 ., and Owen, C. A,, Jr., Gastroenterology, 42, 599 (1962). 33. Wheby, M. S. and Jones, I. G., J. Clin. Invest., 42, 1007 (1963). 34. Charley, P. J., Doctoral Dissertation, Univ. So. Calif. (I960). 35. Pirzio-Biroli, G., in Report of the 19th Ross Pediatric Research Conference, p. 25~(1956). 36. Sarkar, B., Doctoral Dissertation, Univ. So. Calif. (1964). 37. Saltman, P., Stitt, C., and Bates, G. W., Red. Proc., ■ 22, 135 (1964). 38. Davis, B., Saltman, P., and Benson, S.. Biochem. Biophys. Res. Coma., 8, 56 (19o2). 39* Aasa, R., Malmstrom, B. G., Saltman, P. and Yanngard, I* Biochim. Biophys. Acta, £5, 203 (1963). I 40. Windle, J. J., Wiersema, A. K., Clark, J. R., and Eeeney, R. E., Biochemistry, £, 1341 (1963). 41. Johnston, E. A. and Clark, S. J., Am. J. Clin. Rut., 2, 203 (1959). 42. Wallenius, G., Scand. J. Clin. lab. Invest., 4, 24 (1952). 43. London, I. M., Shemin, D.. and Rittenberg, D., J. Biol. Chem.,182, 7^9 (1950). 44. Paoletti, C., Durand, M., Gosse, Ch., and Boiron, M., Rev. Rranc. d’Etudes Clin, et Biol., 2? 259 (1958). 45. Paoletti, C., Boiron, M., 2hibiana, M., Trohaut, R., and Bernard, J,, Sang., 29. 492 (1958). 46. Jandl, J. H., Inman, J. K., Simmons, R. L., and Allen, D. W., J. Clin. Invest., 2§. l6l (1959). 47. Princiotto, J. Y., Rubin, M., Shashaty, G. C., and Zapolski, E, J., J. Clin. Invest., 43, 825 <196*0. !**8. Smithies, 0., Mature, 181, 1203 (1958). : ^9. Smithies. 0. and Horsfall, W, R., Science, 128. 35 (1958). ' 50. Harris, H., Rohson, E. B.. and Siniscalco, M., Mature, 182. *02 (1958). 51. Poulik, M. B., Clin. Chim. Acta, 6, *+93 (1961). ( ; 52. Parker, W. C. and Beam, A. G., Science, 133. 101*+ (1961). ' 53. Butcher, J. B., J. Biol. Chem., 121, 321 (19*+7). 5*+. Fiala, S. and Burk, B., Arch. Biochem. Biophys., 20, 172 (19*0). 55. Emery, T. and Heilands, J. B., Mature, 18*+, 1632 (1959). 56. Warner, R. C, and Weber, I., J. Am. Chem. Soc., 25, 509^ (1953). 57. Fraehkel-Conrat, H., Arch. Biochem. Biophys., 28, *+52 (1950). 58. Williams, J., Biochem, J., 82, 355 (1962). 59. Malmstrom, B. G. and Heilands, J. B., in Ann. Rev. Biochem. (Snell, 1. E., Luek, J. M., Boyer, P. B., and Maekinney, G., Eds.), p. 331, Annual Reviews, Inc., Palo Alto (196*+)! 60. Schultze, H. E., Heide, K., and Muller, H., Behringwerk-Mitteilungen, ^2. , 1 (1957). 61. Azari, P. R. and Feeney, R. P., J. Biol. Chem., 232. 293 (1958)! 62. Margoliash, E., Smith, E. L., Kreil, G., and Tuppy, G., Mature, 1^2, 1125 (1961). 63. Margoliash, E,, J. Biol. Chem., 2^7, 2l6l (1962). 6*+. Azari, P. R. and Peeney, R. E. . Arch. Biochem. Biophys., ^ (1961). 65. Schwartzenbach, G. and Heller, J., Helv. Chim. Acta, 3!+ , 576 (1951). V5; 66. Chaherek, S. and Martell, A. E., Organic Sequestering ; Agents, p. 57*+} John Wiley and Sons, Inc., lew York (1959). University of Southern California

Asset Metadata

Creator Davis, Ward Benjamin (author)

Core Title Studies of iron binding by the human serum protein, transferrin

Contributor Digitized by ProQuest (provenance)

Degree Master of Science

Degree Program Biochemistry

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

Tag chemistry, biochemistry,OAI-PMH Harvest

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-782351

Unique identifier UC11348066

Identifier EP41355.pdf (filename),usctheses-c17-782351 (legacy record id)

Legacy Identifier EP41355.pdf

Dmrecord 782351

Document Type Thesis

Rights Davis, Ward Benjamin

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