Post-translational cellular processing of the factor VIII (FVIII or F8) precursor enables O-sulfation of tyrosine residues (Pittman DD et al. 1992; Michnick DA et al. 1994). Biochemical characterization demonstrated that recombinant human FVIII when metabolically labeled with [35S]-sulfate upon expression in Chinese hamster ovary (CHO) cells or monkey kidney tissue COS-1 cells contains six potential tyrosine sulfation sites, ie, four on the heavy chain (at amino acid residues 365, 737, 738, and 742) and two in the a3 subdomain of the light chain (residues 1683 and 1699) (Pittman DD et al. 1992; Michnick DA et al. 1994). The presence of six tyrosine sulfate residues in FVIII was further confirmed by a combination of liquid chromatography and electrospray ionization mass spectrometry (LC/ESI-MS) studies of the recombinant human FVIII protein derived from baby hamster kidney (BHK) cells (Severs JC et al. 1999) or CHO cells (Schmidbauer S et al. 2015). Site-directed mutagenesis of individual or multiple tyrosine residues showed that all the six sulfation sites are required to modulate FVIII activity (Pittman DD et al. 1992; Michnick DA et al. 1994). Further, treatment of CHO cells that express FVIII with sodium chlorate, an inhibitor of ATP sulphurylase involved in the synthesis of PAPS, did not affect FVIII secretion, but reduced the functional activity by 5-fold, indicating that sulfation was not required for FVIII secretion (Pittman DD et al. 1992). In addition, mutagenesis of Tyr1699 to Phe (Y1699F) demonstrated that sulfation at that residue was required for high affinity interaction of FVIII with von Willebrand factor (vWF) (Leyte A et al. 1991). In the absence of tyrosine sulfation at 1699 in FVIII, the affinity for vWF was reduced by 5-fold (Leyte A et al. 1991). The nuclear magnetic resonance (NMR) spectrum studies of the complex between FVIII and vWF showed significantly larger residue-specific chemical shift changes when Y1699 was sulfated further highlighting the importance of FVIII sulfation at Y1699 for the binding affinity to vWF (Dagil L et al. 2019). The significance of FVIII sulfation at Y1699 in vivo is made evident by the presence of a Y1699F mutation that causes a moderate hemophilia A, likely due to reduced interaction with vWF and decreased plasma half-life (Higuchi M et al. 1990; van den Biggelaar M et al. 2011). Sulfation at tyrosine residues 365 and 1683 increased FVIII activity by increasing the rate of thrombin cleavage at the adjacent thrombin cleavage sites 391 and 1708, respectively (Michnick DA et al. 1994). Mutation of tyrosine residues 737, 738, and 742 had no effect on the thromhin activation rate, even though the cleavage rate at Arg759 was slightly reduced (Michnick et al. 1994). Further. lower FXa-generation activity (86% of the wild-type activity) and lower clotting activity (51% of the wild-type activity) was observed for the FVIII triple point mutant at Tyr residues 737, 738, and 742 (Michnick et al. 1994). This result is in contrast to other study in which no functional differences were found between full-length FVIIl lacking sulfation at one or more of these three residues (Y737, Y738, and Y742) and the fully sulfated form of FVIII (Mikkelsen J et al. 1991).
Protein tyrosine O-sulfation is a common post-translational modification that is catalyzed by a tyrosyl protein sulfotransferase (TPST) (Moore KL 2003; Yang YS et al. 2015). In humans, two TPST isoforms, termed TPST1 and TPST2, have been identified (Ouyang Yb et al. 1998; Mishiro E et al. 2006). The enzyme was shown to catalyze the transfer of sulfate from the universal sulfate donor adenosine 3′-phosphate 5′-phosphosulfate (PAPS) to the hydroxyl group of a peptidyltyrosine residue to form a tyrosine O4-sulfate ester and 3′,5′-ADP (Lee RW & Huttner WB 1983). Structural studies showed that human TPSTs share the same catalytic mechanism (Teramoto T et al. 2013; Tanaka S et al. 2017). In mammalian cells, tyrosine O-sulfation of membrane and secretory proteins was found to occur in the trans-Golgi network, and biochemical studies indicated that the enzyme was membrane-bound (Lee RW & Huttner WB 1985; Baeuerle PA & Huttner WB 1987).