Cysteine Residues in Factor VIII - 2015 Update


            I have come across some other work which I had forgotten - I found two papers while going through some old files.  Both papers come from Alan Johnson laboratory at New York University.  Alan was a modest man and an excellent physician-scientist; his modesty likely precluded the acquisition of greater fame.  While most of Alan's work on factor VIII and the fibrinolytic system, his early work(a-c) on the use of pancreatic deoxyribonuclease (DNAse) as therapeutic was of considerable.  One of the more interesting studies was the intravenous administration of  two million units of DNAse without adverse reaction(a). DNase in combination with streptokinase was developed as a therapeutic product in 1950 (d) and is still used for wound debridement (e).  
            Harris, Newman, and Johnson (f) found that treatment of fresh citrated plasma with 1mM dithiothreitol followed by gel filtration and chromatography on polyelectrolyte E (copolymer of ethyl maleic anhydride and dimthylaminopropylimide) yield a factor VIII preparation purified approximately 5,000-fold over the starting plasma and essential free of von Willebrand factor.  The purified factor VIII was shown to have a molecular weight of 116,000 by several different techniques and was activated by thrombin and neutralized by antibody to human factor VIII.  This material could not be obtained from frozen plasma.  In the subsequent paper (g),  an activated mixed disulfide (h) was obtained from the reaction of pyridyl disulfide (2'-dipyridyl disulfide) with the partially purified factor VIII obtained from Sepharose Cl-4B.  This product was taken to a thiopropyl-agarose for disulfide-exchange chromatography(i).  The factor VIII obtained from this step was 17,000-fold purified from starting plasma and essentially free of vWF.
            I have no reason to doubt this work. It joins other work which is outside of the mainstream (an increasingly dangerous place to be since "the nail than sticks up will be hammered down."(j). A notable example is the work from the laboratory of my late colleague, John Borden Graham with Emily Barrow with factor VIII obtained from succinylation of plasma(k) and the factor VIII activity isolated from kidney(l).
            I find it difficult to fit all of the various pieces of information into a neat package. It is fortunate that such is not required for the successful use of factor VIII products.  The lack of good protein chemistry on factor, specifically functional group reactivity, makes it difficult to interpret all of the information of factor VIII.   While the site of factor VIII function is known, it is poorly characterized either as active pharmaceutical ingredient or final drug product (m) and such information is not necessary for clinical use.  Consider the rapid rate of application of gene editing technology(n), CRISPR/Cas 9, the hemophilia A market may see some radical change in the next decade.


Cysteine Residue Factor VIII 2013


           
The issue of cysteine residues in blood coagulation factor VIII continues to puzzle me and I had posted most of the below information several years ago. Two years have passed and several events have caused me to again consider this problem. The first was the completion of the book entitled Biotechnology of Plasma Proteins which taught me that there are other plasma proteins which contain cysteine residues without clearly defined function. The most notable is albumin but α1-antitrypsin has one cysteinyl residues and α2-antiplasmin has three or four.  The cysteine residue in albumin is thought to be functional although elucidation of that function has not been obtained.  The second was the publication of work describing the removal of a disulfide bond from the factor VIII protein resulting in increased expression in either COS-1 cells or CHO cells (1).  The expressed protein has activity but it is not clear as whether it has activity comparable to the wild-type protein.  For the purpose of the current discussion, this work confirms the existence of eight disulfide bonds in factor VIII, one of which may be removed without major consequences.   Factor VIII contains 23 cysteine residues of which three have been designated as "free"(2).   This does leave 4 cysteine residues unaccounted for.   This situation is not particularly unusual since cysteine residues in proteins have been designated as freely reacting, sluggish, or unreactive for at least sixty years.   There are a number of proteins where cysteine residues are not available without denaturation; ovalbumin is a leading example and for years the exposure of cysteine residues was an index of protein denaturation.  Nevertheless, this remains an untidy situation which should be clarified for several practical reasons.  First, the presence of cysteine in a protein may contribute to long-term solution stability; long-term solution stability of importance for continuous infusion.  The second is that the free cysteine residues may be important to functional half-life as discussed below for fibroblast growth factor-1.   These is also other work which I consider of importance to our understanding of factor VIII. Bayele and coworkers describe the factor VIII activity of thioredoxin and related oxidoreductases such as protein disulfide isomerase in the factor IXa-catalyzed conversion of factor X to factor Xa (3).  As a corollary to this observations, these investigators also described the ability of factor VIII to function as a protein disulfide isomerase in "unscrambling" of "scrambled" ribonuclease.  The second studies is a report by Mei and coworkers (4) on the engineering of cysteine residues into factor VIII for the preparation of pegylated factor VIII.  The three free cysteine residues intrinsic to B-domain deleted factor VIII are said to be buried and not accessible for chemical modification.  In absence of any contrary information, it is assumed that there are no major conformational differences between B-domain deleted factor VIII and native factor VIII which would affect sulfhydryl group reactivity.  McMullen and coworkers (2) were able to modify the cysteine residues in factor VIII with iodoacetamide and reported differential reactivity of these residues suggesting Cys692 is exposed on the surface while Cys310 was buried and less reactive but not unreactive to iodoacetamide.  Both these cysteine residues are the factor VIII heavy chain. Cys2000 is in the A3 domain and is modified by iodoacetamide.  The modification reactions were performed with isolated heavy and light chains at pH ≈ 8.5; the isolated heavy and light chains were stored in the presence of 2-mercaptoethanol to maintain biological activity. Ngo and coworkers reported on the crystal structure of B-domain deleted factor VIII (5) and found Cys310 and Cys2000 were involved in binding cupric ions in a site also involving two histidine residues.  This coordination is consistent with a classical type 1 copper binding site (6,7). The disposition of Cys692 is not reported but it would be appear to a surface-exposed residue.  Shen and coworkers (8) reported the crystal structure of a recombinant factor VIII heterodimer consisting of the heavy chain (A1 domain-A1 domain) and light chain (A3 domain-C1 domain-C2 domain).  This is similar to the B-domain deleted derivative but lacks the linker peptide.   These investigators reported copper-binding sites identical to those described by Ngo and coworkers (5) where Cys310 is binding cupric ion in coordination with two histidine residues (267 and 315) and Cys2000 is in coordination with His1954 and His2005.  There was no comment on Cys692 but it is assumed to be surface-exposed.  Neither group mentions the participation of a methionine residue.  Earlier modeling studies by Pan and coworkers (9) suggested the presence of a single type 2 copper binding site which would not have a cysteine residue in the coordination complex.    Fay and Smuzdin (10) reported modification of Cys528in the heavy chain of factor VIII obtained from a human plasma concentrate with 7-diethylamino-3-[4'-maleimidophenyl]-4-methylcoumarin and Cys1858 in the light chain with N-[1-pyrenyl]maleimide for use in internal fluorescence energy transfer studies.  McMullen and coworkers (2) reported that these two cysteine were disulfide-bonded in their studies.  As will be discussed in further detail below, the problem of non-specific oxidation of cysteine residues in factor VIII could be an issue.
The following discussion is divided into two sections; the first is a consideration of various studies on the chemical reactivity of the cysteine residues in factor VIII while the second is a brief consideration of the various factors which influence sulfhydryl residues in proteins.

Chemical Modification Cysteine Residues in Factor VIII
The following studies must considered with a certain amount of skepticism in that modified amino acid residues in the protein were not identified and most of the studies were performed with impure proteins.  Thus, it is useful to remember Efraim Racker's admonition "Don't waste clean thinking on dirty enzymes."  However, since there is no reason to be unduly suspicious, the results are interpreted in manner consistent with demonstrated chemical reactivity which is based on accessibility and nucleophilicity (11)  Austen (12) showed that bovine factor VIII was inactivated by iodine, hydrogen peroxide, iodoacetamide, or p-chloromercuribenzoate.  Factor VIII inactivated by p-chloromercuribenzoate (PCMB) could be reactivated by cysteine.  Limited studies showed similar behavior of human factor VIII.  Activity was measured with the classic two-stage assay which is conceptually similar to current "chromogenic" assays (13). The results obtained with PCMB provides the strongest support for the importance of a cysteine residue(s) in factor VIII function. Kaelin (14) modified low-purity human factor VIII with sodium periodate.  Factor VIII activity was increased in the two-stage assay but unchanged in the one-stage assay; there was a decrease in the observed molecular weight consistent with dissociation of the vWF-Factor VIII complex.   Periodate is used most frequently to oxidize vicinal diols in carbohydrates to aldehydes but is also known to oxidize cysteine to cystine and to convert N-terminal threonine or serine to an aldehyde (15,16).  Blombäck and coworkers (17) demonstrated the inactivation of factor VIII in human plasma with iodoacetate after reaction with dithiothreitol.  Savidge and coworkers (18) subsequently reported a modest reduction of factor VIII activity in plasma on reaction thioredoxin followed by reaction with iodoacetate.  These investigators also noted a decrease in the molecular weight of the crude factor VII consistent with dissociation of the vWF-Factor VIII complex as noted by Kaelin above.   Manning and coworkers (19) modified recombinant factor VIII with several homobifunctional maleimide derivatives without loss of activity.  Analysis of the modified from with dithio-bis-nitrobenzoic acid (DTNB, Ellman's Reagent) demonstrated modification of 60% of the available sulfhydryl groups.   Reaction with hydrogen peroxide did result in the loss of activity; the ability of hydrogen peroxide to oxidize cysteine residues is problematic but is useful for the oxidation of methionine.  The various studies on cysteinyl residues in factor VIII suggest that there are three cysteine residues and that modification of one or more of these residues results in loss of activity.  The results are complicated by the possibility of in vivo factor VIII oxidation or modification during purification.  An increase in factor VIII activity has been observed on infusion of N-acetylcysteine (20) and exposure to peroxynitrite (21), known to oxidize cysteine to sulfenic acid (22).  Bayele and coworkers (23) have suggested the importance of redox balance in blood coagulation.  Oxidation also occurs during protein purification (24-27)  and the presence of sulfenic acid in proteins is well documented (28,29).  Thus, some of the variability in experimental results might reflect heterogeneity in the cysteine residues of factor VIII.

Factors which Influence Cysteine Reactivity in Proteins
Cysteine is usually the most reactive nucleophile in protein in its reduced or thiol form (29,30). Reactivity of cysteine is largely dependent on ionization to the thiolate form (28). Historically, the measurement of the pKa of cysteine has been a challenge (29) but techniques are improving (31).  Pace and coworkers (32) have recently compiled a list of pKa values for ionizable groups in proteins; the average value for the thiol pKa of cysteine is 6.8±2.7 with a range from 2.5 to 11.1.  Cysteine thiol groups at enzyme active sites generally have lower pKa values.  For example, in Escherichia coli thioredoxin which catalyzes a redox reaction using thiol-disulfide exchange, one cysteine residue is said to have a pKa of 6-7 while the second thiol has a pKa at or above 10 (31-33). In the study on thiol-disulfide exchange reaction with glutaredoxin and glutathione, Iversen and coworkers (33) obtained a pKa of 4.5 for the cysteine thiol in glutaredoxin and 8.9 for the cysteine thiol in glutathione. Finally, cysteine residues at enzyme active sites react for more rapidly with alkylating agents such as chloroacetate than does free cysteine (34).  We have no information on the ionization constants for the free sulfhydryl groups in factor VIII.  Thus, there are various factors which influence the reactivity of nucleophilic groups in proteins such as the cysteine thiol. In the case of factor VIII, it is argued that the sulfhydryl groups of the B domain deleted factor VIII are buried and thus inaccessible for modification with a maleimide derivative of poly(ethylene) glycol (4).  The concept of "buried" versus "accessible" has been an active area of discussion since early work on tyrosine modification some 40+ years ago (35).  The term "buried" implies lack of exposure to solvent (and, hence, lack of reactivity) although a precise definition of exposure is a bit challenging (36).  Reactivity of a buried residue depends on the reagent used as much as the degree of solvent exposure (37).  The binding of metal ions in a coordination complex by cysteine (38) can also influence thiol reactivity (39) As a final note, Lee and Blaber (40) suggest that the presence of a buried cysteine residue can be destabilizing and mutation of buried cysteine residues can improve the solution stability of a protein.  This work on fibroblast growth factor-1 is worth some serious consideration by the protein engineers.
Conclusions

 References


a.  Johnson, A.J., Geger, P.R., and Tillett, W.S., The intravenous injection of bovine crystalline pancreatic deoxyribonuclease into patients, J.Clin.Invest. 33, 1670-1686, 1954.
b. Ayvazain, J.H., Johnson, A.J., and Tillett, W.S., The use of parenterally administered deoxyribonuclease in the treatment of pulmonary absesses, Ann.Rev.Tuberc. 76, 1-21, 1957.
c.  Johnson, A.J., Ayvazian, J.H., and Tillett, W.S., Crystalline pancreatic deoxyribonuclease as an adjunct to the treatment of pneumococcal meningitis, New Eng.J.Med. 260, 893-900, 1959.
d.  Anonymous, New approved drugs; streptokinase-streptodornase,varidase, Lederle, Med.Bull. 17, 320-327, 1957.
e.   Smith, F., Dryburgh, N., Donaldson,J., and Mitchell, M., Debridement for surgical wounds, Cochrane Database Syst.Rev. 9:CD006214, 2013.
f.  Harris, R.B., Newman, J., and Johnson, A.J., Partial purification of biologically active, low molecular weight, human antihemophilic factor free of von Willebrand factor. I. Partial characterization and evidence for disulfide bond(s) susceptible to limited reduction, Biochim.Biophys.Acta 668, 456-470,. 1981.
g.  Harris, R.B., Johnson, A.J., and Hodgins, L.T., Partial purification of biologically active, low molecular weight, human antihemophilic factor free of von Willebrand factor. II.  Further purification with thiol-disulfide interchange chromatography and additional evidence for disulfide bonds susceptible to limited reduction, Biochim.Biophys.Acta 668, 471-480, 1981.
h.  Faulstich, H. and Heintz,D., Reversible introduction of thiol compounds into proteins by use of activated mixed disulfides, Methods Enzymol. 251, 357-366, 1995.
i.  Lundblad, R.L., Chemical Reagents for Protein Modification, 4th edn., Chapter 7, pps. 217-327, CRC Press/Taylor and Francis, Boca Raton, Florida, USA, 2014.
j.  Adapted from a Japanese proverb.
k.  Barrow, E.M. and Graham, J.B., Factor VIII (AHF) activity of small size produced by succinylating plasma, Am.J.Physiol. 222, 134-141, 1972.
l.   Barrow, E.M. and Graham, J.B., Kidney antihemophilic factor.  Partial purification and some properties, Biochemistry 7, 3917-2925, 1968.
m.  Lubiniecki, A.S. and Federici, M.M., Comparbility is not just analytical equivalence, Biologicals 34, 45-47, 2006.
n.  Cyranoski, D. and Reardon, S., Embryo editing sparks epic debate, Nature 520, 593-594, 2015
References for Earlier Text.
1.  Phelan, S.A., AOP2 (antioxidant protein 2): structure and function of a unique thiol-specific antioxidant, Antioxid.Redox.Signal. 1, 571-584, 1999.
2.  McMullen, B.A., Fujikawa, K., Davie, E.W., et al., Locations of disulfide bonds and free cysteines in the heavy and light chains of recombinant human factor VIII (antihemophilic factor A), Protein Sci. 4, 740-746, 1995.
3. Bayele, H.K., Murdock, P.J., and Pasi, K.J., Residual factor VIII-like cofactor activity of thioredoxin and related oxidoreducatases, Biochim.Biophys.Acta 1800, 398-404, 2010.
4. Mei, B., Pan, C., Jiang, H., et al., Rational design of a fully active long-acting PEGylated factor VIII for hemophilia A treatment, Blood, in press, 2010.
5. Ngo, J.C.K., Huang, M., Roth, D.A., et al., Crystal structure of human factor VIII: implications for the formation of the factor IXa-Factor VIIIa complex, Structure 16, 597-606, 2008.
6. McPherson, I.S. and Murphy, M.E., Type-2 copper-containing enzymes, Cell.Mol.Life Sci. 64, 2887-2899, 2007.
7. Sakuri, T. and Kataoka, K., Structure and function of type 1 copper in multicopper oxidases, Cell.Mol.Life Sci. 64, 2642-2656, 2007.
8. Shen, B.W., Spiegel, P.C., Chang, C.-H., et al., The tertiary structure and domain organization of coagulation factor VIII, Blood 111, 1240-1247, 2008.
9.Pan, Y., DeFay, T., Gitschier, J., and Cohen, F.E., Proposed structure of the A domains of factor VIII by homology modeling, Nat.Struct.Biol. 2, 740-744, 1995.
10. Fay, P.J. and Smudzin, T.M., Intersubunit fluorescence energy transfer in human factor VIII, J.Biol.Chem. 264, 14005-14010, 1989.
11. Lundblad, R.L., Chemical Reagents for Protein Modification, 3rd edn., CRC Press, Boca Raton, FL, USA, 2004.
12. Austen, D.E.G., Thiol groups in the blood clotting action of factor VIII, Brit.J.Haematol. 19, 477-484, 1970.
13. Lundblad, R.L., Kingdon, H.S., Mann, K.G., et al., Issues with the assay of factor VIII activity in plasma and factor VIII concentrates, Thromb.Haemost. 84, 942-948, 2000.
14. Kaelin, A.C., Sodium periodate modification of Factor VIII procoagulant activity, Brit.J.Haematol. 31, 349-359, 1975.
15. Beppu, M., Ochiai, H., and Kikugawa, K., Macrophage recognition of periodate-treated erythrocyte: Involvement of disulfide formation of the erythrocyte membrane proteins, Biochim.Biophys.Acta  979, 35-45, 1989.
16. Chelius, D. and Shaler, T.A., Capture of peptides with N-terminal serine and threonine: a sequence-specific chemical method for peptide mixture simplification, Bioconjug.Chem. 14, 205-211, 2003.
17. Blombäck, B., Hessel, B., Savidge, G., et al., The effect of reducing agents on factor VIII and other coagulation factors, Thromb.Res. 12, 1177-1194, 1978
18.  Savidge, G., Carlebjörk, G., Thorell, L., et al., Reduction of factor VIII and other coagulation factors by the thioredoxin system, Thromb.Res. 16, 587-599, 1979.
19.  Manning, F., Fagain, C.O., O'Kennedy, R., and Woodhams, B., Effects of chemical modifiers on recombinant factor VIII activity, Thromb.Res. 80, 247-254, 1995.
20. Knudsen, T.T., Thorsen, S., Jensen, S.A., et al., Effect of intravenous N-acetylcysteine infusion on haemostatic parameters in healthy subjects, Gut 54, 515-521, 2005.
21. Nielsen, V.G., Crow, J.P., Mogal, A., et al., Peroxynitrite decreases hemostasis in human plasma n vitro, Anesth.Analg. 99, 21-26, 2004.
22.  Turell, L., Botti, H., Carballal, S., et al., Reactivity of sulfenic acid in human serum albumin, Biochemistry 47, 358-367, 2008.
23.  Bayele, H.K., Murdock, P.J., Perry, D.J. and Pasi, K.J., Simple shifts in redox/thiol balance that perturb blood coagulation, FEBS Lett. 510, 67-70, 2002.
24. Cockle, S.A., Epond, R.M., Stollery, J.G., and Moscarello, M.A., Nature of the cysteinyl residues in lipophilin from human myelin, J.Biol.Chem. 255, 9182-9188, 1980.
25. Jenkins, N. Murphy, L., and Tyther, R.,Post-translational modifications of recombinant proteins: significance for biopharmaceuticals, Mol.Biotechnol. 39, 113-118, 2008.
26. Graebski, A.C., Advances in preparation of biological extracts for protein purification, Methods Enzymol. 463, 285-303, 2009.
27. Dewey, J.S., Struck, D.K., and Young, R., Thiol protection in membrane protein purification: A study with phage holins, Analyt.Biochem. 390, 221-223, 2009.
28. Allison, W.S., Formation and reactions of sulfenic acids in proteins, Accounts Chem.Res. 9, 293-299, 1976.
29. Liu, T.-Y., The role of sulfur in proteins, in The Proteins, Vol. 3., 3rd edn., ed. H. Neurath and R.L. Hill, Chapter 3, pps.239-402, Academic Press, New York, NY, USA, 1977.
30. Lundblad, R.L., Chemical Reagents for Protein Modification, 3rd edn., CRC Press, Boca Raton, FL, USA, 2005.
31.  Nelson, K.J., Day, A.E., Zeng, B.-B., et al., Isotope-coded, iodoacetamide-based reagent to determine individual cysteine pKa values by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry, Analyt.Biochem. 375, 187-195, 2008.  
32. Dyson, H.J., Jeng, M.-F., Tennant, L.L., et al., Effects of buried charged groups on cysteine thiol ionization and reactivity in Escherichia coli thioredoxin: Structural and functional characterization of mutants of Asp26 and Lys 57, Biochemistry 36, 2622-2636, 1997.
33. Iversen, R., Andersen, P.A., Jensen, K.S., et al., Thiol-disulfide exchange between glutaredoxin and glutathione, Biochemistry 49, 810-820, 2010.
34. Sluyterman, L.A.A., The rate-limiting reaction in papain as derived from reaction of the enzyme with chloroacetic acid, Biochim.Biophys.Acta 151, 178-187, 1968.
35. Gorbunoff, M.J., Exposure of tyrosine residues in proteins. II. The reactions of cyanuric fluoride and N-acetylimidazole with pepsinogen, soybean trypsin inhibitor, and ovomucoid, Biochemistry 7, 2547-2554, 1968.
36. Kyte, J., Structure in Protein Chemistry, 2nd edn., Chapter 6, pps. 273-274, Garland Science, New York, NY, USA, 2007.
37. Stehlin, C., Heacock, D.H., Liu, H., and Musier-Forsyth, K., Chemical modification and site-directed mutagenesis of the single cysteine in motif 3 of class II Escherichia coli prolyl-tRNA synthetase, Biochemistry 36, 2932-3938, 1997.
38. Krezel, A. and Maret, W., Different redox states of metallothionein/thionein in biological tissues, Biochem.J. 4-2. 551-558, 2007.
39. Wetz, K. and Crichton, R., Chemical modification as a probe of the topography and reactivity of horse-spleen apoferritin, Eur.J.Biochem. 61, 545-550, 1976.
40. Lee, J. and Blaber, M., The interaction between thermodynamic stability and buried free cysteines in regulating the functional half-life of fibroblast growth factor -1, J.Mol.Biol. 393, 113-127, 2009.