Response: Cys¹⁸⁶-Cys²⁰⁹ disulfide–mutated tissue factor does not equal cryptic tissue factor: no impairment in decryption of disulfide mutated tissue factor

The primary focus of our recent report¹  is to examine whether Cys¹⁸⁶-Cys²⁰⁹ disulfide–mutated tissue factor (TF) mimics cryptic TF and the importance of Cys¹⁸⁶-Cys²⁰⁹ disulfide bond formation in TF decryption, 2 key issues that were not examined earlier.² 

In our view, one should be careful not to equate TF mutants that exhibit low procoagulant activity with cryptic TF. This incorrectly broadens the definition of cryptic TF and creates unnecessary confusion and controversy. The classic definition of cryptic TF is that it can form a stable complex with factor VIIa (FVIIa), but the resultant TF-FVIIa complexes fail to activate its macromolecular substrates, factors IX and X.^3,4  Therefore, when probing the cryptic nature of TF, one should evaluate the activity of TF-FVIIa complexes at the cell surface. Despite some inherent differences in their binding properties, both cryptic and decrypted TF form stable high-affinity associations with FVII/FVIIa, well below plasma concentrations of FVII (see references in Bach and Monroe⁵ ). Cryptic TF would remain cryptic even in the presence of supraphysiologic concentrations of FVIIa. TF mutants that exhibit lower procoagulant activity at physiologic concentrations of FVII/FVIIa, but could be rescued by supraphysiologic concentrations of FVIIa should not be equated with cryptic TF. The Cys¹⁸⁶-Cys²⁰⁹ disulfide–mutated TF falls into this category. They exhibit low procoagulant activity because the Cys¹⁸⁶-Cys²⁰⁹ disulfide bond is required for efficient FVIIa binding^1,2  and not because they conform to cryptic TF form.

One of the criticisms of our study raised by Ruf and Versteeg is that in concluding that the allosteric disulfide bond is not essential for TF's procoagulant activity, we took into consideration only data at supraphysiologic concentrations of TF ligands. Because the physiologic concentration of FVIIa was not sufficient to saturate the TF mutants at the cell surface, we had to use higher concentrations of FVIIa. We actually measured the amount of TF-FVIIa complexes formed at the cell surface and used this number to determine the specific activities of wild-type and TF mutants. This experimental design was appropriate and correct to probe the cryptic nature of TF mutants. Nonetheless, to provide a complete picture to the reader, in our report we have also provided data on TF activity measurements in the presence of low concentrations of FVIIa. In their letter, Ruf and Versteeg also suggest that clotting assay is the “gold standard” for TF procoagulant activity. However, the clotting assay alone is not appropriate and useful in investigating the cryptic nature of TF.

Ruf and Versteeg showed that extreme overexpression of TF^C209A yielded cell-surface TF procoagulant activity that was similar to cells expressing very low levels of wild-type TF and decryption induced by cell lysis was much more efficient in cells expressing wild-type TF compared with cells expressing TF^C209A. These data differ from our recent data obtained using TF^C186SC209S mutant.¹  In our studies, we did not observe extreme overexpression of TF mutants, as observed by Ruf and Versteeg, when endothelial cells were transduced with 25- to 50-fold higher concentration of virus encoding TF mutants (25-50 MOI/cell) compared with wild-type TF (1-2 MOI/cell) to obtain similar levels of TF expression at the cell surface. Ruf and Versteeg suggest that cotransduction with PAR2 in their study may be responsible for overexpression of TF^C209A. However, their own data (see Figure 1B of their letter) fail to support this assumption.

Here, we examined TF decryption induced by cell lysis in wild-type TF and TF^C209S mutant (Figure 1). Consistent with the data reported recently¹  for TF^C186SC209S mutant, we found a similar fold-increase in TF activity (∼ 40- to 50-fold) in cells expressing either TF^C209S mutant or wild-type TF upon their lysis (Figure 1D). It is unclear why Ruf and Versteeg failed to obtain similar data. It is possible that extreme overexpression of TF^C209A in their study might have resulted in very low TF–anionic phospholipid ratio upon cell lysis compared with wild-type TF, which was expressed at very low levels. It is also possible that TF^C209A mutant could have behaved differently from that of our TF^C209S mutant. Here it may be pertinent to note that in their earlier studies they found that their TF^C209A and TF^C186A mutants behaved differently.⁶  In our studies, we obtained similar data with all 3 of our disulfide mutants: TF^C186S, TF^C209S, and TF^C186C209S. Limited experiments performed with alanine-substituted mutants (plasmid transfection in Chinese hamster ovary cells) also yielded data similar to that of serine mutants.

In conclusion, our recent reports^1,7,8  demand candid discussion on recently postulated hypotheses that PDI regulates TF procoagulant activity by its chaperone activity⁹  and/or Cys¹⁸⁶-Cys²⁰⁹ disulfide bond isomerization.⁶