The Weak Binding Reaction Between Folate and Human Serum Proteins

Transcription

1 ANNALS OF CLINICAL AND LABORATORY SCIENCE, Vol. 8, No. 1 Copyright 1978, Institute for Clinical Science The Weak Binding Reaction Between Folate and Human Serum Proteins ALFRED ZETTNER, M.D. and P. E. DULY Division o f Clinical Pathology, Department o f Pathology, School o f Medicine, University o f California, San Diego, CA ABSTRACT New evidence is presented that folic acid in serum binds with low affinity to major serum proteins. This low affinity binding is distinct from the high affinity binding by folate binding protein (FBP), a minor protein which is known to occur in serum with great quantitative variability. These conclusions are based on results obtained by equilibrium dialysis of serum containing only negligible amounts of FBP. At 4, the equilibrium value of the ratio of bound to free folate was approximately 0.81 and rem ained the same even w ith up to 1,000 tim es greater than normal folic acid concentrations; however, w ith higher concentrations the ratio decreased progressively. These results are predicted by the laws of mass equilibria for a binding system w ith low ligand concentration vis-a-vis high binder concentration and low affinity betw een the reactants. A rough estimate of the mean affinity constant K governing this weak folate interaction w ith serum proteins yielded a value of 1.12 ±0.13 x 103 M-1. Introduction Folate appears to be bound by human serum in two different ways: (1) firmly and specifically to special, m inor serum proteins w ith high affinity and (2) weakly and perhaps nonspecifically to major serum proteins with low affinity. These two binding phenom ena up to now have not been clearly distinguished from each other in the literature concerned with serum folate binding, even though that evidence for at least two types of binding has b een accumulating for a num ber of years. The presence of high affinity folate binding proteins (FBPs) in human serum is now firmly established3,4,10,28,31 and their partial purification by various techniques has been reported.14,27,33 The complex of folate with FBPs is characterized by an affinity high enough to withstand such dissociating conditions as charcoal adsorption,12,20,31 chromatography14,20,27,33 and dialysis.20,33 A survey31 of over 1,000 sera from patients and healthy volunteers showed that the serum content of unsaturated FBP is highly variable, the binding capacity ranging from less than 0.1 to 8.5 fig of folate per lite r o f serum. T he p h ysiologic functions or importance of serum FBP are not clear, and th e ele v a tio n of unsaturated FBP in some sera may be incidental to tissue turnover.33 In most sera, how ever, the endogeneous folate

2 58 ZETTN ER AND DULY appears to exceed the binding capacity of the serum FBP considerably, so that the major fraction of the serum folates is not bound by FBP and therefore available for the weaker interactions with the other serum proteins. The weak binding of folate by serum is demonstrable with such non-dissociating m ethods as e q u ilib riu m d ialy sis,1,26 Sephadex gel filtration17,24 and ultrafiltration.6,18 These weak complexes of folate and serum proteins dissociate readily under such treatm ent as DEAE-cellulose colum n chrom atography,33 charcoal adsorption,20,31 e le c tro p h o re sis6 or dialysis.18,20,33 Furtherm ore, this weak binding reaction presented the rather puzzling phenom enon of being seem ingly nonsaturable because the ratio of free and bound ligand w hen using folic acid6,11,18 or methyltetrahydrofolic acid25 remained the same over three orders of magnitude of folate concentrations. The bound folate fraction was variously reported as 50 percent,6,24 46 to 51 percent,18 64 p ercent,11 51 to 61 percent when the data of Alter et al1 are recalculated as the actual bound fraction, and 59 to 76 percent.25 Some theoretical considerations and experim ental data are presented to support the concept that serum folate, in addition to its interaction with FBP, normally forms weak complexes w ith major serum proteins. Furthermore, it is shown that the maintenance of a seemingly constantbinding ratio over large concentration ranges of folate is the behavior predicted in principle by the laws of mass equilibria for a binding system of low affinity when the binder concentration exceeds that of its ligand by orders of m agnitude. On the basis of these theoretical considerations, the present authors predicted and have shown experim entally that higher concentrations of folate than those tested by previous workers6,11,18,25 are needed to demonstrate a decrease of the binding ratios. T heoretical Considerations For the purpose of formulating a model approxim ating the weak folate-serum protein interaction, it shall suffice to make the following sim plifying assum ptions: (1) The major serum proteins capable of forming such weak complexes react identically w ith folates; (2) The concentration of these serum proteins is approximately 7.0 x 10-4 M (based on a protein content in serum of 70 g per 1 and an average mol w t of 100,000); (3) These proteins possess only one binding site for folates; (4) The concentration of folates in serum is approximately 2 x 10"8 M (based on a serum folate level of 9 fig per 1 and an average mol wt of the endogenous folates of 450); and (5) The affinity constant K for the weak binding reaction of folates by serum proteins is ofthe order of 1.43 x 103 M_1.This estim ate is derived from the basic relationship K = [PQ]/[P][Q] (formula #1) where [PQ], [P] and [Q] are the molar concentrations ofthe complex, free folate and free protein, respectively. If it is assumed for the bound folate fraction, based on the reported findings (vide supra), an approximate value of 50 percent, the ratio of [PQ]/[P] then becomes 1 and K = 1/[Q]. Since the protein complexed with folate thus is only a m inute fraction (1/70 000) of total protein, it can set [Q] equal with total protein concentration and thus K = 1/7 x 10~4 M = 1.43 x 103 M"1can be obtained. U nder these restricting assumptions, the equilibrium ratio of bound to free folate is th en given by the quadratic equation5 R2 + R (1 + Kp - Kq) - Kq = O (formula #2) where: R = bound folate/free folate, K = the affinity constant,

3 WEAK BINDING O F FO LA TE BY HUMAN SERUM 5 9 p = the m olar concentration of total folate, and q = the molar concentration of protein. Using equation 2, the R values for six folate concentrations have been calculated ranging from 2 X 10-8 to 2 x 10-3 M and a constant protein concentration of 7.0 x 10 4 M (table I). W hen p < q (folate concentrations from 2 x 10~8 to 2 x 10-5 M), changes of p result only in minute changes of R, differences obviously too small to be detected experimentally. Only w hen the concentration of p approaches (2 X 10_4M) or exceeds (2 x 10~3 M) the order of magnitude of q, the predicted changes of R becom e sufficiently large to be discernable by measurement. In the following are reported equilibrium dialysis e x p erim en ts involv ing whole serum with the folate content raised in ten-fold increm ents to extend over the concentration range shown in table I. The results are in good agreement with the theoretical predictions. Materials Phosphate-citrate buffer: 0.05 M, ph The osmolality of this buffer was raised to 290 milliosmols by the addition of NaCl (approximately 1.5 g per 1) in order to obtain the same osmolality as was found in the serum pool. [%5'-3H] folic acid: ([3H]PGA) 22 Ci per millimol.* Working solutions were made in phosphate-citrate buffer to contain 500 fig per 1 of [3H]PGA. The concentration of [3H]PGA was determ ined through competitive binding assays using unlabeled folic acid as described elsew here.31 Folic acid: (PGA) crystalline.! Working solutions were made in phosphate-citrate buffer and stored at -20. These were stable for several months. Serum pool: P atients sera obtained routinely by the clinical chem istry labora * Lot No: BR-2574 from Schwartz/M ann, Orangeburg, NY f Sigma Chemical Co., St. Louis, MO tory, excluding those that were icteric, lipemic, or hemolyzed, were screened, after being kept at 4 for seven days, for the presence of unsaturated FB P by the addition of [3H]PGA followed by adsorption w ith polyvinylpyrrolidone-coated charcoal.31,32 The purpose of the seven days of aging of the sera was to decrease the endogenous folate through spontaneous decay and thereby to uncover saturated binding sites on the serum FBP. Sera exhibiting a binding capacity of less than 0.2 fig of [3 H]PGA per 1 were selected to make a pool thus containing only negligible amounts of the high affinity FBP and frozen at 20. The endogenous folate content of this pool, apparently due to the aging of the sera, was only 0.6 fig per 1 as determ ined by radioassay.23 T he total protein content of the pool was 70 g per 1 and that of album in 38 g per 1. Scintillant: 5 g of 2,5-diphenyloxazole, 100 ml of Bio-Solv BBS-3 and 1 liter toluene. Dialysis cells: 10 ml capacity TechniLab equilibrium dialysis cells. 1 Dialysis membranes: These were cut from regenerated cellulose dialysis tubing, type D U The tu b in g was soaked in the phosphate-citrate buffer overnight prior to use. The adsorption of [3H]PGA by these membranes amounted to less th an 2 p e rc e n t of the total [3H]PGA present in the chambers over a time period of 52 hours. Shaker: Y ankee V ariable S p eed Rotator.** Syringes: H am ilton Syringes, 250 / I volume, f f M allinckrodt C hem ical Works, St. Louis, MO Beckman Instruments, Anaheim, CA Obtained from Cole-Parmer Instruments Co., Chicago, IL O btained from Scientific Products, Irvine, CA ** O btained from Clay Adams, Parsippany, NJ f t Obtained from Hamilton Co., P. O. Box 10030, Reno, NV

4 6 0 Z E T T N E R AND DULY T A B L E I Equilibrium R Values Calculated for Weak Binding Reaction Between Folates and Serum Proteins Folate Concentration P Protein Concentration q Affinity Constant k R = Bound folate/ free folate 2 x 10~8 M 7 x M 1.43 x 103 M" X ICI-7 M X ICI'6 M x ICI-5 M x lcr1* M X 10"3 M Methods Equilibrium dialysis was carried out in the following way: mixtures of 5 ml of serum, 0.1 m l of w orking solution of [3H]PGA and 1.0 ml of solutions containing 0, 500, 5,000, 50,000, 500,000 and 5,000,000 fig per 1 of unlabeled PGA were placed into one side of the dialysis chambers and mixtures identical to these in [3H]PGA and unlabeled PGA contents but substituting buffer for serum were placed into the other chamber side. In preparing the 5,000,000 fig per 1 solution of unlabeled PGA, the entire solids did not dissolve in the diluting buffer as some fine crystals rem ained after several hours of mixing. However, when 1.0 ml aliquots of these well-mixed suspensions were added to the 5 ml aliquots of serum or buffer, th e crystals disso lv ed com pletely. The cham bers w ere protected from light with black paper, placed onto the shaker and rotated at 90 rpm for 52 hours. The experiments were carried out in a walk-in refrigerator at 4 since at room tem perature or at 37 the [3H]PGA did not rem ain sufficiently stable over the time required for equilibrium dialysis. After periods of 16, 20, 28, 44 and 52 hours of dialysis, aliquots of 100 fil from the serum and the buffer sides were aspirated w ith the H am ilton syringes and placed into 5 ml of scintillant for radioactive counting. The level of radioactivity in these aliquots ranged from approximately 20,000 to 50,000 cpms, and the coefficient of variation of the entire procedure calculated from replicate m easurem ents was 1.4 percent. Equilibrium was reached betw een 18 and 20 hours. The stability of the [3H]PGA was monitored repeatedly throughout the period of dialysis by determ ining the bindability of the [3H]PGA to milk folate binding protein7 in the samples without unlabeled PGA. No reduction in the bindability of [3H]PG A was seen throughout the 52 hours of dialysis. The absorption of the radioactivity by polyvinylpyrrolidonecoated charcoal32 from the dialysis mixtu res was check ed th ro u g h o u t the dialysis period. The blank values re maining in the supernate after charcoal treatm ent did not rise above 1.5 percent of initial total radioactivity in the serum and above 1.0 percent in the buffer, a further indication of the good stability of the [3H]PGA during dialysis. After equilibrium was reached, the m olar concentrations of folic acid in either side of the chambers w ere d e te r m ined at 20, 28, 44 and 52 hours of dialysis from the d istrib u tio n o f the radioactivity minus the charcoal blank values betw een the two chamber sides and the total amount of folic acid (labeled and u n lab eled ) know n to have b e e n added at the beginning of dialysis. The blank values in the serum included the radioactivity bound by FBP. Thus, in effect, four replicate determinations were obtained of the equilibrium conditions spanning a total of 32 hours. The small amount of endogenous folate (0.6 fig per 1) rem aining in the serum after seven days of aging at 4 was considered negligible and therefore not included in these estimations. The concentration of bound folate was calculated from the differences in the folate concentrations betw een the buffer and the serum dialysates.

5 WEAK BINDING O F FO LA TE BY HUMAN SERUM 61 Results and D iscussion The results are summarized in table II. The equilibrium values shown are the means of the values found at 20, 28, 44 and 52 hours. As is readily apparent, during dialysis the folic acid concentrations in the serum increased and those in the buffer decreased. The value of R, the ratio of bound to free folic acid, rem ained constant with folic acid concentrations ranging from 2.03 x 10-8 to 1.86 x 10-5 M. With concentrations of 1.86 x 10-4 M or higher, a decrease of the values of R is seen. The mean affinity constant K calculated from these data was 1.12 ±0.13 x 103 M-1. This compares well with the results published by Soliman and Olesen24 who studied the binding of [3H]PGA to p u re hum an plasm a album in by Sephadex gel filtration and found a K value of 0.9 x 103 M-1. Our experimental data are in principal agreem ent with the theoretical values in table I. The results of the interaction of folate and serum proteins in the equilibrium dialysis experim ents are those exp ected of a bin d in g system w hich is characterized by low ligand concentration and higher binder concentration exceeding that of the ligand by several orders of m agnitude, and w hich is gove rn e d by a low affinity. W hen such a binding system comes to equilibrium, a large fraction of the total ligand will be free even though a huge excess of binder is present. This is reflected in the low R values (e.g. <1) which show little change u n less the lig an d co n cen tratio n ap proaches that of the binder. Association and dissociation of these weak complexes m ust be rather rapid processes since it is possible with charcoal to adsorb folates, that are not bound to FBP, from serum within less than one minute,31 as rapidly and c.ompletely as from buffer solutions free of protein. Similarly, the complexing of PGA by cow s milk folate binder, requiring less than three m inutes for completion at room tem perature,30 is not recognizably slowed down in the presence of serum. Both these processes, the charcoal adsorption and the complexing with milk binder, depend on the dissociation of the folate-serum protein complex. It is also known that association and dissociation of the weak complexes b etw een ligands and serum proteins in general have half-tim es of the order of a few m illiseconds, although there are exceptions to this rule.15,16 The results of our binding studies, as w ell as those o f others, w ith folic acid1,6 11,18 w ere q u antitatively quite similar to those w ith m ethyltetrahydro- TABLE I I Equilibrium Dialysis Experiments Folate Concentrations Before and After Dialysis* and Resulting R Values A t s t a r t o f A t e q u il ib r iu m (m eans o f d e t e r m in a ti o n a t d i a l y s i s i n 2 0, 2 8, 44 and 52 h o u r s ) R = C o n c e n tr a tio n b o th s i d e s P e r c e n t Bound f o l a t e / F a c to r o f ch a m b e rs Serum S id e B u f f e r S id e Bound F o la t e Bound f r e e f o l a t e : 1 ( T 8 10"7 10" lct4 10" *Folic acid concentrations are in m per 1. Values are the means of four replicate determinations obtained after 20» 28, 44 and 52 hours of dialysis.

6 6 2 Z E T T N E R AND DULY folic acid,25 the major endogenous serum folate. Also, various tem peratures of 4 as used by us, N eal and W illiam s18 and Spector et al,25 50,1 23,25 37 n and seem ed to have little effect on the binding ratios obtained with either folate derivative. These observations imply that the endogenous folate in circulating plasm a norm ally is partially bound in these weak complexes and that the in vivo binding ratios are similar to those dem onstrated in vitro. The weak binding of folate by serum proteins raises the question as to what effect such complex formation may have on the rate of folate delivery to the tissues, particularly since an interference with the uptake of folates by tumor cells has been dem onstrated for the strong FBP com plex.29 T he effect of weak folateserum protein complexes on tissue delivery would be expected to be much smaller since, as the experim ental data indicate, at physiologic concentrations approxim ately 50 p e rc e n t of the total plasma folate is not bound by proteins and presum ably readily available to the high affinity folate receptors2,3,13,21,22 and dihydrofolate reductase27 in the cells. Furtherm ore, any w ithdraw al of free folate from the circulating plasma by the tissues w ould be followed by rapid dissociation of the folate protein complexes until equilibrium is re-established. Lastly, the effect that drugs interacting with serum proteins may have on folate binding should be considered. In general, a drug may interfere with the b in d in g o f a n o th er su b stan ce (exogenous or endogenous) to serum proteins either by competition for the same binding sites or by lowering the binding affinity.15,16 Both mechanisms result in a decrease of the bound fraction. The effect of a drug on the binding ratio of folates, when competing for binding sites with the same affinity, can be formulated by the general equation for com petitive binding5 R2 + R (1 + Kpfol + K p ^ - Kq) - Kq = 0 (Formula #3) where p foi is the total molar concentration of folate and p(irug the total molar concentration of drug. Two interesting aspects derive from this formulation: (1) that the effect of the drug on R is simply additive with that of folate and (2) a consequence of 1, that the molar concentration of the drug in serum would have to approximate that of the proteins in order to lower significantly the percentage of bound folate, analogous to th e effect of raising folate c o n c en trations (tables I and II). A lter e t a l1 showed that folate binding by serum was decreased when acetylsalicylic acid was added to th e dialysis m ixtures. T hey suggested that the decrease of folate binding may have been due to the transacetylation of album in by acetylsalicylic acid known to occur in vitro9 as w ell as in vivo,8 and that this structural alteration changed the binding affinity for folates. How ever, the sim ple quantitative relationships o f d rug interferences as outlined suffice to explain this effect of acetylsalicylic acid on the basis of its competition with folates for binding sites on the album in molecule. The in vitro and in vivo concentrations of the drug producing the effect seen by Alter et al1 were given as from 5 to 40 mg per dl, that is from 0.28 to 2.23 x 10-3 M, concentrations ranging just below and above those of serum proteins. These are also the drug concentrations for which form ula # 3 predicts significantly and m easurably lower R values for folates. Since the free folate is in dynam ic equilibrium not only w ith the bound folate in plasma but also with tissue folate, the free folate concentration would be expected to change only insignificantly when drugs displace the bound fraction. C linically, low er serum folate levels should be observed without the signs of folate deficiency. This expectation is w ell

7 bom e out by the in vivo observations of Alter et al.1 WEAK BINDING O F FO LA TE BY HUMAN SERUM 63 References 1. Alter, H. J., Z vaifler, N. J., and Rath, C. E.: Interrelationship of rheumatoid arthritis, folic acid and aspirin. Blood 38: , C orrocher, R., D e Sandre, G., Pacor, M. L., and HOFFBRAND, A. V.: Hepatic protein binding of folate. Clin. Sci. Mol. Med. 46: , D ac osta, M. and Ro th en ber g, S. P.: Appearance of a folate binder in leukocytes and serum of women who are pregnant or taking oral contraceptives. J. Lab. Clin. Med. 83: , E ichn er, E. R., Paine, C. J., D ickson, V. L., and H argrove, M. D., Jr.: Clinical and laboratory observations of serum folate-binding protein. Blood 46: , E kins, R. P., New m an, G. B., and O Riordan, J. L. H.: Theoretical aspects of saturation and radioimmunoassay. Radioisotopes in Medicine: In Vitro Studies. Hayes, R. L., Goswitz, F. A., and PearsonMurphy, B. E., eds. U.S. Atomic Energy Commission, Oak Ridge, TN, p. 59, E lsborg, L.: Binding of folic acid to human plasma proteins. Acta Haemat. 48: , F ord, J. E., Salter, D. N., and Scott, K. J.: The folate-binding protein in milk. J. Dairy Res. 36: , H awkins, D., P inckard, R. N., and F arr, R. S.: Acetylation of human serum albumin by acetylsalicylic acid. Science 260: , H awkins, D., P inckard, R. N., C raw ford, I. P., and F arr, R. S.: Structural changes in human serum albumin induced by ingestion of acetylsalicylic acid. J. Clin. Invest. 48: , H in es, J. D., Kamen, B., and Caston, D.: Abnormal folate binding protein(s) in azotemic patients. Sixteenth Annual Meeting, American Society of Hematology, Chicago, IL, p. 57, Jo h ns, D. G., Sper ti, S., and Burgen, A. S.V.: The metabolism of tritiated folic acid in man. J. Clin. Invest. 40: , Ka m e n, B. A. and C a s t o n, J. D.: D irect radiochemical assay for serum folate: competition between 3H-folic acid and 5-methyl tetrahydrofolic acid for a folate binder. J. Lab. Clin. Med. 83: , Kamen, B. A. and Caston, J. D.: Identification of a folate binder in hog kidney. J. Biol. Chem. 250: , Kamen. B. A. and C aston, J. D.: Purification of folate binding factor in normal umbilical cord serum. Proc. Nat. Acad. Sci. 72: , Koch-Weser, J. and Sell er s, E. M.: Binding of drugs to serum albumin. New Eng. J. Med. 294: , Ko c h-weser, J. and Seller s, E. M.: Binding of drugs to serum albumin. New Eng. J. Med. 294: , M a r k k a n e n, T.: Pteroylglutamic acid (PGA) activity of serum in gel filtration. Life Sci. 7: , N e a l, G. E. and W illia m s, D. C.: The fate of intravenously injected folate in rats. Biochem. Pharmacol. 14: , O Bro in, J. D., T em perley, I. J., Brow n, J. P., and SCOTT, J. M.: Nutritional stability of various naturally occurring monoglutamate derivatives of folic acid. Amer. J. Clin. Nutr. 28: , R e t i e f, F. P., H e y n s, A. d u P., O o s t h u i z e n, M., v a n R e e n e n, O. R., and B a d e n h o r s t, C. J.: In vitro binding of folates by body fluids. Brit. J. Haematol. 32: , R o t h e n b e r g, S. P.: A macromolecular factor in some leukemic cells which binds folic acid. Proc. Soc. Exp. Biol. Med. 233: , R o t h e n b e r g, S. P. and D a C o s t a, M.: Further observations on the folate-binding factor in some leukemic cells. J. Clin. Invest. 50: , R o t h e n b e r g, S. P., D a C o s t a, M., and R o s e n b e r g, Z.: A radioassay for serum folate: use of a two-phase sequential-incubation, ligand-binding system. New Eng. J. Med. 286: , SOLIMAN, H. A. and Olesen, H.: Folic acid binding by hum an plasma albumin. Scand. J. Clin. Lab. Invest. 36: , Spector, R., L o renzo, A. V., and D rum, D. E.: Serum binding o f m ethyltetrahydrofolic acid. Biochem. Pharmacol. 24: , T o e n n ie s, G., F r a n k, H. G., and G a l l a n t, D. L.: On the folic acid activity of human blood. J. Biol. Chem. 200:23-30, WAXMAN, S.: Folate binding proteins. Brit. J. Haematol. 29:23-29, WAXMAN, S. and SCHREIBER, C.: Characteristics of folic acid-binding protein in folatedeficient serum. Blood 42: , W a x m a n, S. and S c h r e i b e r, C.: The role of folic acid binding proteins (FABP) in the cellular uptake of folates. Proc. Soc. Exp. Biol. Med. 247: , Z e t t n e r, A. and D u l y, P. E.: Principles of com petitive binding assays (saturation analyses). II. sequential saturation. Clin. Chem. 20:5-14, Z e t t n e r, A. and D u l y, P. E.: New evidence for a binding principle specific for folates as a normal constituent of human serum. Clin. Chem. 20: , Z e t t n e r, A. and D u l y, P. E.: Relative efficacy of separation of free and bound [3,5'- 3H]pteroylglutamate by charcoal coated with various materials. Clin. Chem. 22 : , Zettn er, A. and Duly, P. E.: Separation of folate binding protein from human serum by DEAE-cellulose column chromatography. Clin. Chem. 22: , 1976.