Protein Chemistry
Gene, Biosynthesis, and Metabolism
Interaction with Heparin and Dermatan Sulfate
Stimulation by Glycosaminoglycans
Physiologic Function
Selected References

      In 1939, Brinkhous and co-workers showed that the anticoagulant activity of heparin is mediated by an endogenous component(s) of plasma that they termed "heparin cofactor."  The first heparin cofactor purified from plasma in 1968 was antithrombin (or antithrombin III), a protein that inhibits serine proteases of the intrinsic coagulation pathway (particularly thrombin, factor Xa and factor IXa) in the presence of mast cell heparin or vascular heparan sulfate.  A second heparin cofactor, chromatographically distinguishable from antithrombin, was observed in 1974 and subsequently isolated.  This protein, termed heparin cofactor II (HCII), inhibits thrombin but not other coagulation proteases.  Its activity is stimulated by the endogenous glycosaminoglycans heparin, heparan sulfate, and dermatan sulfate, as well as by various other natural and synthetic polyanions.  Although HCII and antithrombin both inhibit their target proteases by forming a stable bimolecular complex in which the protease is inactive, the mechanism by which glycosaminoglycans increase the rate of complex formation differs for the two inhibitors.  The physiologic function of HCII remains obscure despite anecdotal reports of HCII deficiency in some patients with thromboembolic disease.

Protein Chemistry

        Human HCII is a single glycosylated polypeptide chain 480 amino acids in length with an estimated mass of 65,600 Da (GenBank accession #M12849).  It contains 3 cysteine residues but no intramolecular disulfide bonds.  HCII is ~30% identical in sequence to antithrombin and other members of the serpin superfamily.  However, it contains an N-terminal extension of ~80 amino acid residues that is not present in other serpins.  The N-terminal region includes a tandem repeat of two highly acidic sequences, each of which contains a tyrosine residue that becomes O-sulfated during biosynthesis.  This region of HCII is required for rapid inhibition of thrombin in the presence of a glycosaminoglycan (see below).  HCII has been identified in several mammalian species, as well as in frog and chicken, suggesting that its function may be highly conserved  (1) .
        Like other inhibitory serpins, HCII functions as a suicide substrate for its target protease.  Thrombin attacks the reactive site peptide bond of HCII (Leu444-Ser445) and appears to become trapped in a covalent acyl intermediate of the proteolytic reaction  (2) .  The presence of leucine at the P1 position of the reactive site, which is unusual for a thrombin substrate, may explain the fact that HCII inhibits chymotrypsin more rapidly than thrombin in the absence of a glycosaminoglycan and does not inhibit other trypsin-like proteases of the coagulation pathway.  An HCII mutant with Leu444 replaced by arginine inhibits thrombin ~100 times faster than wild-type HCII in the absence of a glycosaminoglycan and inhibits coagulation factor Xa at an appreciable rate, but it does not inhibit chymotrypsin  (3) .  Therefore, the P1 leucine residue is a critical determinant of the protease specificity of HCII.

Gene, Biosynthesis, and Metabolism

        The gene for human HCII is located on chromosomal band 22q11 and contains five exons distributed over ~16 kb of DNA (GenBank accession #J05309).  Human hepatocytes contain a 2.3-kb mRNA encoding HCII, and biosynthesis of the protein has been demonstrated in cultured human hepatoma cells.  HCII mRNA has not been detected by Northern blot analysis in other tissues, including heart, brain, placenta, lung, muscle, kidney, or pancreas.  HCII is secreted by the liver into the bloodstream, where it is present at a concentration of ~1.2 +/- 0.4 µM (mean +/- 2 SD) in human plasma.  The half-life of HCII in the circulation is ~2.5 days, whereas thrombin-HCII complexes are cleared much more rapidly by the low density lipoprotein receptor-related protein on hepatocytes  (4) .  Although turnover studies of labeled HCII in man suggest that ~40% of the protein equilibrates into an extravascular compartment, the tissue distribution of HCII has not yet been determined.
        The plasma concentration of HCII is low in neonates and in some adult patients with liver disease or disseminated intravascular coagulation.  Normal HCII levels are present in plasma during oral anticoagulant use, in the vast majority of patients with venous thrombosis, and in most patients with hereditary antithrombin deficiency.  The HCII concentration may be elevated during acute inflammatory reactions.  The mechanism of this elevation is unknown, since the production of HCII in hepatoma cells does not appear to be regulated by the inflammatory cytokines interleukin-6, interleukin-1beta, or tumor necrosis factor-alpha  (5) .

Interaction with Heparin and Dermatan Sulfate

        Certain glycosaminoglycans, including heparin, heparan sulfate, and dermatan sulfate, increase the rate of inhibition of thrombin by HCII more than 1000-fold.  Thus, addition of dermatan sulfate to plasma decreases the t1/2 of thrombin inhibition from ~1 min to less than 0.05 s.  Binding of the glycosaminoglycan to HCII is required for the stimulatory effect.  Heparin binds to HCII with a lower affinity than to antithrombin; therefore, a 10-fold higher concentration of heparin is required to accelerate thrombin inhibition by HCII.  Whereas antithrombin binds to a specific pentasaccharide structure in heparin that includes a 3-O-sulfated glucosamine residue, HCII binds to most heparin oligosaccharides >4 monosaccharide units in length regardless of their composition.  HCII is unique among serpins in its ability to be activated by dermatan sulfate.  In contrast to the rather non-specific binding of HCII to heparin oligosaccharides, HCII binds to a minority of dermatan sulfate oligosaccharides >6 monosaccharide units in length.  The high-affinity binding site for HCII in porcine skin dermatan sulfate is a tandem repeat of three iduronic acid 2-sulfateÆN-acetylgalactosamine 4-sulfate disaccharide subunits  (6) .  The binding site for HCII in dermatan sulfate from other tissues may also contain iduronic acid?N-acetylgalactosamine 4,6-disulfate subunits.
        Analysis of the natural variant HCII Oslo (Arg189 to His) established that heparin and dermatan sulfate interact with different amino acid residues on the surface of the inhibitor  (7) .  This mutation causes a large decrease (~60-fold) in the affinity of HCII for dermatan sulfate but does not affect the affinity of the inhibitor for heparin.  Arg189 occurs in a cluster of basic amino acid residues in helix D that can be aligned with the heparin-binding site of antithrombin.  Mutations of Lys173, Arg184, and Arg185 in recombinant HCII diminish the binding of heparin and its ability to stimulate the thrombin-HCII reaction, whereas mutations of Arg184, Arg185, Arg189, Arg192, and Arg193 affect the interaction with dermatan sulfate.  Thus, the binding sites for heparin and dermatan sulfate overlap but are not identical.

Stimulation by Glycosaminoglycans

        The N-terminal acidic domain of HCII (residues 54-75) resembles the C-terminal portion of the leech anticoagulant hirudin, which interacts with anion-binding exosite I of thrombin.  A synthetic peptide that corresponds to residues 54-75 of HCII competitively inhibits binding of hirudin to thrombin but does not inhibit thrombin's catalytic activity.  Studies with mutant forms of recombinant HCII suggest that thrombin interacts with the N-terminal acidic domain in the intact inhibitor.  For example, deletion of residues 1-67, which include the first acidic repeat, reduces the maximum rate of thrombin inhibition 2 to 3 orders of magnitude in the presence of heparin or dermatan sulfate  (8) .  Thus, the first acidic repeat is essential for rapid inhibition of thrombin in the presence of a glycosaminoglycan.  In the absence of a glycosaminoglycan, the native and truncated forms of HCII inhibit thrombin at essentially the same (slow) rate.  Deletions or point mutations of the acidic repeats increase the affinity of HCII for heparin, suggesting that the acidic domain interacts with the glycosaminoglycan-binding site.
        These experiments support a model in which binding of a glycosaminoglycan to HCII displaces the N-terminal acidic domain from the glycosaminoglycan-binding site, thereby allowing the acidic domain to interact with exosite I of thrombin (Fig. 1).  Such an interaction could facilitate inhibition by positioning the active site of thrombin next to the reactive site of HCII.  This model is supported by the observation that mutations in the glycosaminoglycan-binding site, which presumably weaken its interaction with the N-terminal acidic domain, increase the rate of inhibition of thrombin in the absence of a glycosaminoglycan ~100-fold  (9) .  Additional support for this model has been obtained from experiments with variants of thrombin defective in exosite I  (10) .  In the absence of a glycosaminoglycan, these variants are inhibited by HCII at rates similar to that of native thrombin, but the maximal rates of inhibition in the presence of a glycosaminoglycan are greatly reduced.  By contrast, mutations of exosite I have little or no effect on the rate of inhibition of thrombin by antithrombin in the presence of heparin.

Figure 1.  Comparison of the proposed mechanisms of inhibition of thrombin by HCII and antithrombin in the presence of a glycosaminoglycan.  AT, antithrombin.  GAG, glycosaminoglycan.  Exo I, thrombin exosite I (hirudin-binding site).  Exo II, thrombin exosite II (glycosaminoglycan-binding site).  L and R represent the reactive site P1 residues of HCII and antithrombin, respectively.  S represents the catalytic serine residue of thrombin.

Physiologic Function

        The presence of thrombin-HCII complexes in normal human plasma suggests that HCII inhibits thrombin in vivo  (13) .  Cultured fibroblasts and vascular smooth muscle cells accelerate inhibition of thrombin by HCII, but endothelial cells do not.  In the case of fibroblasts, which synthesize both heparan sulfate and dermatan sulfate, a small dermatan sulfate proteoglycan is responsible for the stimulatory effect.  Two isolated dermatan sulfate-containing proteoglycans, biglycan and decorin, are capable of stimulating HCII activity in vitro  (14) .  These results suggest that HCII may inhibit thrombin at extravascular sites where dermatan sulfate is present.
        Heterozygous deficiency of HCII is found in ~1% of apparently healthy individuals and in approximately the same percentage of patients with venous thrombosis  (15) .  Thus, associations between HCII deficiency and thrombosis in anecdotal case reports may be coincidental.  Two sisters with homozygous HCII deficiency (10-15% of normal plasma HCII activity) have been reported  (16) .  One of the homozygous individuals had a history of recurrent venous thromboembolism but was also found to have heterozygous antithrombin deficiency.  No history of thrombosis was obtained from the other homozygous individual or from any of the 12 heterozygous HCII-deficient members of this family.
        Circumstantial evidence suggests that the activity of HCII is increased during pregnancy, when both the maternal and fetal plasma contain trace amounts of a dermatan sulfate proteoglycan that stimulates inhibition of thrombin by HCII.  The placenta is rich in dermatan sulfate and may be the source of this proteoglycan  (17) .  Elevated concentrations of HCII have been reported in women who are pregnant or who use oral contraceptives, and thrombin-HCII complexes are elevated ~3- to 6-fold over baseline at term and immediately post-partum  (18) .  Conversely, decreased HCII levels have been reported in patients with pre-eclampsia  (19) .  Thus, HCII could be activated locally to inhibit coagulation of maternal blood in the placenta.

Selected References

1. N.S. Colwell and D.M. Tollefsen, Thromb. Haemost. 80: 784-790 (1998).
2. M.J. Griffith, C.M. Noyes, J.A. Tyndall and F.C. Church, Biochemistry 24: 6777-6782 (1985).
3. V.M. Derechin, M.A. Blinder and D.M. Tollefsen, J. Biol. Chem. 265: 5623-5628 (1990).
4. M.Z. Kounnas, F.C. Church, W.S. Argraves and D.K. Strickland, J. Biol. Chem. 271: 6523-6529 (1996).
5. C. Koike, Y. Hayakawa, K. Niiya, N. Sakuragawa and H. Sasaki, Thromb. Haemost. 75: 298-302 (1996).
6. M.S.G. Pavão, K.R.M. Aiello, C.C. Werneck, L.C.F. Silva, A.-P. Valente, B. Mulloy, N.S. Colwell, D.M. Tollefsen and P.A.S. Mourão, J. Biol. Chem. 273: 27848-27857 (1998).
7. M.A. Blinder, T.R. Andersson, U. Abildgaard and D.M. Tollefsen, J. Biol. Chem. 264: 5128-5133 (1989).
8. V.M.D. Van Deerlin and D.M. Tollefsen, J. Biol. Chem. 266: 20223-20231 (1991).
9. P.C.Y. Liaw, R.C. Austin, J.C. Fredenburgh, A.R. Stafford and J.I. Weitz, J. Biol. Chem. 274: 27597-27604 (1999).
10. T. Myles, F.C. Church, H.C. Whinna, D. Monard and S.R. Stone, J. Biol. Chem. 273: 31203-31208 (1998).
11. J.P. Sheehan, D.M. Tollefsen and J.E. Sadler, J. Biol. Chem. 269: 32747-32751 (1994).
12. J.-H. Han, H.C.F. Côte and D.M. Tollefsen, J. Biol. Chem. 272: 28660-28665 (1997).
13. L. Liu, L. Dewar, Y. Song, M. Kulczycky, M.A. Blajchman, J.W. Fenton, II, M. Andrew, M. Delorme, J. Ginsberg, K.T. Preissner and F.A. Ofosu, Thromb. Haemost. 73: 405-412 (1995).
14. H.C. Whinna and F.C. Church, J. Protein Chem. 12: 677-688 (1993).
15. J. Mateo, A. Oliver, M. Borrell, N. Sala and J. Fontcuberta, Blood Coagul. Fibrinolysis 9: 71-78 (1998).
16. P. Villa, J. Aznar, A. Vaya, F. España, F. Ferrando, Y. Mira and A. Estellés, Thromb. Haemost. 82: 1011-1014 (1999).
17. M.A. Delorme, L. Xu, L. Berry, L. Mitchell and M. Andrew, Thromb. Res. 90: 147-153 (1998).
18. T. Andersson, B. Lorentzen, H. Høgdahl, T. Clausen, M.-C. Mowinckel and U. Abildgaard, Thromb. Res. 82: 109-117 (1996).
19. J. Bellart, R. Gilabert, L. Cabero, J. Fontcuberta, J. Monasterio and R.M. Miralles, Blood Coagul. Fibrinolysis 9: 205-208 (1998).

Suggestions for Further Reading

G.J. Broze, Jr., and D.M. Tollefsen, Regulation of blood coagulation by protease inhibitors, in The Molecular Basis of Blood Diseases, 2nd Edition (G. Stamatoyannopoulos, A.W. Nienhuis, P.W. Majerus and H. Varmus, eds.), W.B. Saunders Co., Philadelphia, 1994, pp. 629-656.

F.C. Church, D.D. Cunningham, D. Ginsburg, M. Hoffman, S.R. Stone, and D.M. Tollefsen (eds.), Chemistry and Biology of Serpins, Plenum Press, New York, 1997.