Role of Prothrombin Fragment 1 in the Pathway of Regulatory Exosite I Formation during Conversion of Human Prothrombin to Thrombin*

Prothrombin (Pro) activation by factor Xa generates the thrombin catalytic site and exosites I and II. The role of fragment 1 (F1) in the pathway of exosite I expression during Pro activation was characterized in equilibrium binding studies using hirudin54–65 labeled with 6-(N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino)hexanoate ([NBD]Hir54–65(\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{SO}_{3}^{-}\) \end{document})) or 5-(carboxy)fluorescein ([5F]Hir54–65(\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{SO}_{3}^{-}\) \end{document})). [NBD]Hir54–65(\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{SO}_{3}^{-}\) \end{document}) distinguished exosite I environments on Pro, prethrombin 1 (Pre 1), and prethrombin 2 (Pre 2) but bound with the same affinities as [5F]Hir54–65(\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{SO}_{3}^{-}\) \end{document}). Conversion of Pro to Pre 1 caused a 7-fold increase in affinity for the peptides. Conversely, fragment 1.2 (F1.2) decreased the affinity of Pre 2 for [5F]Hir54–65(\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{SO}_{3}^{-}\) \end{document}) by 3-fold. This was correlated with a 16-fold increased affinity of F1.2 for Pre 2 in comparison to thrombin, demonstrating an enhancing effect of F1 on F1.2 binding. The active intermediate, meizothrombin, demonstrated a 50- to 220-fold increase in exosite affinity. Free thrombin and thrombin·F1.2 complex bound [5F]Hir54–65(\batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \(\mathrm{SO}_{3}^{-}\) \end{document}) with indistinguishable affinity, indicating that the effect of F1 on peptide binding was eliminated upon expression of catalytic activity and exosite I. The results demonstrate a new zymogen-specific role for F1 in modulating the affinity of ligands for exosite I. This may reflect a direct interaction between the F1 and Pre 2 domains in Pro that is lost upon folding of the zymogen activation domain. The effect of F1 on (pro)exosite I and the role of (pro)exosite I in factor Va-dependent substrate recognition suggest that the Pro activation pathway may be regulated by (pro)exosite I interactions with factor Va.

) by 3-fold. This was correlated with a 16-fold increased affinity of F1.2 for Pre 2 in comparison to thrombin, demonstrating an enhancing effect of F1 on F1.2 binding. The active intermediate, meizothrombin, demonstrated a 50-to 220-fold increase in exosite affinity. Free thrombin and thrombin⅐F1.2 complex bound [5F]Hir 54 -65 (SO 3 ؊ ) with indistinguishable affinity, indicating that the effect of F1 on peptide binding was eliminated upon expression of catalytic activity and exosite I. The results demonstrate a new zymogen-specific role for F1 in modulating the affinity of ligands for exosite I. This may reflect a direct interaction between the F1 and Pre 2 domains in Pro that is lost upon folding of the zymogen activation domain. The effect of F1 on (pro)exosite I and the role of (pro)exosite I in factor Va-dependent substrate recognition suggest that the Pro activation pathway may be regulated by (pro)exosite I interactions with factor Va.
In the penultimate step of blood coagulation, thrombin is generated by factor Xa cleavage of Pro at Arg 271 -Thr 272 and Arg 320 -Ile 321 . Pro activation is regulated by the protein cofactor, factor Va, phospholipids, and calcium, which together elicit a ϳ300,000-fold increase in the activation rate through formation of the membrane-bound, factor Xa⅐factor Va, prothrombi-nase complex (1)(2)(3)(4). In the absence of factor Va, factor Xa cleavage of Arg 271 -Thr 272 results in accumulation of prethrombin 2 (Pre 2) 1 and activation fragment 1.2 (F1.2) noncovalent complex as an activation intermediate. Factor Va confers a substrate specificity change such that factor Xa cleavage of Arg 320 -Ile 321 is preferred, and the catalytically active intermediate meizothrombin (MzT) is generated predominately (5)(6)(7)(8).
Cleavage of the alternative sites in the intermediates generates the products thrombin and F1.2.
The physiological substrate specificity of thrombin and the localization of thrombin activity are mediated by one or both of two exosites (I and II) distinct from the catalytic site (9,10). Exosite I is in a precursor state on Pro called proexosite I (11,12). Activation of the catalytic site in the formation of thrombin is accompanied by an overall ϳ100-fold increase in affinity of exosite I for hirudin 54 -65 (Hir 54 -65 (SO 3 Ϫ )). The proexosite has been implicated in the mechanism of factor Va rate acceleration of Pro activation and cofactor-mediated Pro substrate recognition (13,14). The role of proexosite I in substrate recognition is supported further by recent site-directed mutagenesis studies demonstrating that mutation of proexosite I residues in prethrombin 1 (Pre 1) results in loss of factor Va cofactor activity and is correlated with loss of affinity of the proexosite for Hir 54 -65 (SO 3 Ϫ ) (15). A natural mutation of Arg 67 to His in proexosite I of Pro isolated from a patient with a severe procoagulant defect and mild bleeding phenotype also showed reduced factor Va acceleration of its activation (16). In the preceding paper, proexosite I on Pre 1, a Pro analog that lacks the fragment 1 (F1) domain, and Pre 2 was shown to be activated by cleavage of Arg 320 -Ile 321 , generating the active intermediate, meizothrombin des-fragment 1 (MzT(-F1)), and the product, thrombin, respectively. Removal of F1 from Pro by thrombin cleavage to yield Pre 1 resulted in a 6-fold increase in affinity for hirudin peptides. This result suggested a new role for F1 in modulating activation of exosite I. Previous studies also suggested a role for F1 in expression of exosite I in the observation that macromolecular exosite I ligands bound to MzT(-F1) but not meizothrombin (MzT) (17).
The roles of F1, F2, and the catalytic domain (Pre 2) in factor Va regulation of Pro activation are not fully understood. Early studies demonstrated a rate-enhancing effect of F2 in factor * This work was supported by National Institutes of Health Grant HL38779 (to P. E. B.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Va-accelerated Pre 2 activation (18), suggesting that F2 mediated Pro binding to factor Va. Subsequent studies demonstrated similarly significant rate-enhancing effects of F2 and F1.2 on Pre 2 activation in solution (19). A kinetic analysis of Pre 2 activation by the membrane-bound factor Xa⅐factor Va complex, however, does not support a significant role for F2 (19) and indicates that Pre 2 substrate recognition is mediated by exosites expressed on factor Xa in the factor Va-assembled complex (20 -22). Productive binding of Pro to factor Va in the prothrombinase complex has also been linked to the Gla domain (23) and kringle domains of both F1 and F2 (24,25). Our studies of thrombin-and Pro-factor Va interactions support a direct role for proexosite I on the Pro catalytic domain in mediating productive binding to the heavy chain subunit of factor Va within the factor Xa⅐factor Va complex (13,26,27). The proexosite I-factor Va interactions are thought to be influenced by changes in exosite I expression during Pro activation.
[NBD]Hir 54 -65 (SO 3 Ϫ ) displayed fluorescence spectral changes upon binding to the Pro activation intermediates that reported different proexosite I environments on Pro, Pre 1, and Pre 2. Expression of exosite I on the Pro activation intermediates, as measured by the increase in affinity for the peptides, displayed a similar pattern as for the Pre 1 activation intermediates, where initial cleavage at Arg 320 -Ile 321 caused simultaneous activation of the active site and full activation of exosite I (28). Binding of F1.2 to Pre 2 decreased the affinity of the fluorescein-labeled peptide for exosite I by 3-fold, whereas no effect of F1.2 was observed for thrombin. The attenuating effect of F1 was correlated with binding of F1.2 to Pre 2 with a 16-fold higher affinity compared with thrombin. The results indicate that F1.2 interacts with the zymogen proteinase domain with greater affinity than the active proteinase and decreases the affinity of exosite I for hirudin peptides only in the zymogen forms. The results characterize a new role of F1 in the expression of exosites I and II, which is disengaged on folding of the proteinase "activation domain" (29) into the catalytically active form, simultaneous with activation of exosite I. These observations suggest that changes in (pro)exosite I interactions mediated by the F1 domain and differential expression of exosite I on the Pro activation intermediates may regulate factor Va-Pro interactions that control the activation pathway.
Direct binding of the labeled peptides to Pro, the Pro activation intermediates, and thrombin was measured by titrating the labeled peptide with each protein. Changes in fluorescence (⌬F/F 0 ϭ F obs Ϫ F 0 /F 0 ) were monitored as described for [5F]Hir 54 -65 (SO 3 Ϫ ) (28). For [NBD]Hir 54 -65 (SO 3 Ϫ ), fluorescence was measured with excitation at 480 nm (16-nm band pass) and emission at 540 nm (8-nm band pass). The data were fit by the quadratic binding equation to obtain the maximal fluorescence change (⌬F max /F 0 ) and the dissociation constant (K D ) for peptide binding. Competitive binding of labeled and unlabeled Hir 54 -65 -(SO 3 Ϫ ) was measured by titrations of fixed concentrations of labeled peptide and protein as a function of competing ligand concentration. The direct and competitive binding data were analyzed simultaneously with the cubic binding equation (28,34,35). Ϫ ) (H) binding to free Pre 2 (P2) and the Pre 2⅐F1.2 (P2F1.2) complex (K P2(H) and K P2F1.2(H) ), and the maximum fluorescence changes for each of the fluorescent species (⌬F max P2(H) /F 0 and ⌬F max P2F1.2(H)/F 0 ) as described previously (28,36). The dissociation constant for F1.2 binding to Pre 2 (K P2(F1.2) ) was fixed at the value determined as described above. The effect of F1.2 on binding of [5F]Hir 54 -65 (SO 3 Ϫ ) to thrombin was determined similarly and analyzed to determine the binding constants for [5F]Hir 54 -65 (SO 3 Ϫ ) binding to free thrombin (T) and thrombin⅐F1.2 complex (K T(H) and K TF1.2(H) ), and the maximum fluorescence changes for each of the fluorescent species (⌬F max T(H) /F 0 and ⌬F max TF1.2(H) /F 0 ). The dissociation constant for F1.2 binding to thrombin (K T(F1.2) ) was fixed at the determined value.
Free Energy Calculations for [5F]Hir 54 -65 (SO 3 Ϫ ) Binding to Pro and Pre 1 Activation Intermediates-The change in free energy of association upon [5F]Hir 54 -65 (SO 3 Ϫ ) binding to Pro and the activation intermediates was calculated from ⌬G ϭ RT ln K D . The change in free energy of peptide binding to each of the activation intermediates relative to Pro was calculated from: ⌬⌬G ϭ RT ln(K Int /K Pro ), where K Int is the dissociation constant for [5F]Hir 54 -65 (SO 3 Ϫ ) binding to one of the intermediates and K Pro is the dissociation constant for [5F]Hir 54 -65 (SO 3 Ϫ ) binding to Pro (37,38). Nonlinear least squares analysis was performed with SCIENTIST (MicroMath). Reported errors represent Ϯ 2 S.D. Ϫ ) bound to Pre 1 with a 7-fold increased affinity in comparison to Pro (data not shown, Table I), confirming the 6-fold increase in affinity observed for [5F]Hir 54 -65 (SO 3 Ϫ ) (28). Pre 2 bound the NBD-labeled peptide with a slightly lower (Ͻ2-fold) affinity compared with Pre 1, which was not considered significant. In contrast to the enhancement seen with Pro, thrombin quenched the fluorescence of [NBD]Hir 54 -65 (SO 3 Ϫ ) by 50 Ϯ 1% and bound with a ϳ250-fold tighter dissociation constant of 15 Ϯ 3 nM, indistinguishable from the affinity determined for the fluorescein-labeled peptide (Fig. 2B and Table I). Comparison of the results for binding of the two peptides to Pro and thrombin showed a 250-fold increase in affinity for the NBDlabeled peptide binding to thrombin, similar to the 130-fold increased affinity for [5F]Hir 54 -65 (SO 3 Ϫ ) seen previously (11). Comparison of Pre 2 and Pre 1 showed that, although the spectral changes were distinct, there was no correlation with a change in exosite affinity. The results suggested that removal of F1 from Pro resulted in either a conformational change in proexosite I on formation of Pre 1 or that F1 interfered directly with the binding of peptides to proexosite I on Pro.

Characterization of Fluorescence
Binding of Fragment 1. Ϫ ) binding to P2⅐F1.2 complex to form the P2⅐F1.2⅐H ternary complex corresponded to a 3-fold weaker affinity (K P2F1.2(H) 1.3 Ϯ 0.2 M) and slightly smaller fluorescence change compared with free Pre 2 ( Fig. 4 and Table I). In view of the previous results demonstrating no effect of F2 on binding of peptides to Pre 2, the decrease in affinity of the peptide for the Pre 2⅐F1.2 complex indicated an effect of the presence of F1 in F1.2 on peptide binding to (pro)exosite I,  Table I). The data were analyzed with the ternary complex model to obtain the dissociation constants for T⅐H (K T(H) 31 Ϯ 4 nM) and T⅐F1.2⅐H (K TF1.2(H) 38 Ϯ 9 nM), which were indistinguishable from the previous results for [5F]Hir 54 -65 (SO 3 Ϫ ) binding (28) ( Table I). The results demonstrated that binding of F1.2 to exosite II on thrombin did not affect binding of labeled hirudin peptides to exosite I, similar to results seen with F2 alone (28,36). The presence of F1 in F1.2 lowered exosite I affinity in the zymogen forms, Pro and the Pre 2⅐F1.2 complex, but not in the active species, MzT and T⅐F1.2 complex. Ϫ ) with MzT displayed dissociation constants for the unlabeled and labeled peptides that were indistinguishable ( Fig. 6B and Table I).

Binding of Hirudin Peptides to MzT(-F1) and MzT-Exosite
[NBD]Hir 54 -65 (SO 3 Ϫ ) bound to MzT with an indistinguishable dissociation constant (data not shown, Table I). These results showed that cleavage of Pro at Arg 320 -Ile 321 altered the conformation of exosite I such that binding of the hirudin peptides was increased by 50-to 250fold. The results confirmed that peptide bond cleavage in the proteinase domain of Pro resulted in simultaneous activation of exosite I and the catalytic site.
Pathway of Exosite I Expression-Changes in free energy of [5F]Hir 54 -65 (SO 3 Ϫ ) association with Pro and the Pro activation intermediates were calculated for each of the species (Table II). Fig. 7 maps the changes in free energy of exosite I binding for the Pre 1 and Pro activation species determined in this study and the companion manuscript (28). The free energy change for [5F]Hir 54 -65 (SO 3 Ϫ ) binding to Pro was Ϫ7.5 kcal/mol (Table II). Removal of F1 from Pro to form Pre 1 increased the free energy   Table I. Titrations were performed and analyzed as described under "Experimental Procedures." change of binding by ϳ1.1 kcal/mol. Peptide binding to Pre 2 in the presence and absence of F2 displayed a free energy change of Ϫ8.7 kcal/mol independent of F2, indicating similar effects of F1 and F1 plus F2 removal on the free energy of peptide binding due to F1. A decrease in the free energy change of association was observed for peptide binding to Pre 2 in the presence of F1.2 of ϳ0.7 kcal/mol, reflecting the decrease in affinity due to the presence of F1. Proteolytic cleavage of Arg 320 -Ile 321 , activating the proteinase domain and yielding the active species, MzT, MzT(-F1), and thrombin, increased the binding free energy change by ϳ3 kcal/mol. These results indicated that F1 had a similar modulating effect on proexosite I of Pro and the Pre 2⅐F1.2 complex. This effect was specific for the zymogen forms and was alleviated upon the conformational change that activates the catalytic site and exosite I. DISCUSSION The pathway of exosite I expression on human Pro activation intermediates and products and the role of F1 in exosite expression were characterized in quantitative equilibrium binding studies for the first time. A new environmentally sensitive NBD derivative of hirudin 54 -65 was used in fluorescence spec-tral studies to probe the environment of proexosite I on the Pro species. The fluorescence emission spectra of [NBD]Hir 54 -65 -(SO 3 Ϫ ) revealed differences between the environments of the probe in proexosite I on the zymogen forms, Pro, Pre 1, and Pre 2, and on the active enzymes, MzT, MzT(-F1), and thrombin. NBD-peptide binding to Pre 1 showed very little change in fluorescence compared with the enhancement seen for Pro, demonstrating a perturbation of exosite I in Pro by the F1 domain. By contrast to Pro and Pre 1, quenching of the NBDpeptide was observed for Pre 2, and larger and spectroscopically indistinguishable quenching was observed for all of the active enzyme forms. The results indicated that the microenvironment around the N terminus of the hirudin 54 -65 peptide was altered in distinctly different ways when each of the activation fragments was successively removed from the proteinase zymogen domain and, independently of the activation fragments on zymogen activation.
[NBD]Hir 54 -65 (SO 3 Ϫ ) bound to Pro and Pre 1 with dissociation constants similar to those seen for the fluorescein-labeled peptide, confirming a 7-fold increase in affinity for hirudin peptides upon removal of F1. On the pathway of exosite I activation, the Pre 2 domain demonstrated no change in affinity for [NBD]Hir 54 -65 (SO 3 Ϫ ) in comparison to Pre 1, confirming the results with [5F]Hir 54 -65 (SO 3 Ϫ ). Although the complexes with Pre 1 and Pre 2 were spectroscopically different, there was no significant difference in affinity for the peptides. MzT, formed by cleavage of Pro at Arg 320 -Ile 321 , displayed a 50-to 150-fold increase in affinity for the fluorescein-labeled and unlabeled peptides in comparison to Pro, similar to the 250-fold increase in affinity seen for the NBD-labeled peptide. These results are consistent with the previous finding that cleavage at Arg 320 -Ile 321 in Pre 1 to activate the catalytic site is accompanied by full activation of proexosite I (28). The results contrast the observations of Liu and colleagues (39) that cleavage of either activation peptide bond results in formation of exosite I. Characterization of exosite I expression in those studies was performed with bovine Pro species and with a different peptide, hirudin 53-64 . A ϳ5-fold lower affinity of hirudin 54 -65 peptides for bovine Pro and thrombin (11) may also translate to differences in the activation intermediates.
Overall, activation of proexosite I affinity for hirudin peptides as model ligands follows a discrete pathway where exosite I and the catalytic site are activated simultaneously. As summarized in Fig. 7, the free energy of peptide binding to each active proteinase species, MzT, MzT(-F1), and thrombin in the presence and absence of either F1.2 or F2 was the same, consistent with the indistinguishable spectroscopic properties of the [NBD]Hir 54 -65 (SO 3 Ϫ ) bound species and a similar conformation of exosite I. The results demonstrate that the presence of the F1 domain decreases the affinity of hirudin 54 -65 peptides for Pro and Pre 2 in a complex with F1.2. The affinity of the hirudin peptides for Pre 2 was decreased 3-fold by F1.2 binding, in comparison to Pre 2 alone, toward a value that was within ϳ2.5-fold of the affinity for Pro. This represented similar 0.7-to 1.1-kcal/mol decreases in binding free energy attributable to the presence of F1. The effect of the F1 domain was alleviated upon proteolytic cleavage of Pro at Arg 320 -Ile 321 . The results indicated that full expression of exosite I accompanied the conformational change of the proteinase activation domain (29) that activates the catalytic site and eliminated the modulating effect of the F1 domain on (pro)exosite I binding of hirudin 54 -65 . The finding of Wu and colleagues (17) that removal of F1 from MzT to form MzT(-F1) resulted in activation of exosite I toward macromolecular ligands was not observed here with hirudin peptides, which had the same affinity for MzT and MzT(-F1).
An effect of F1 on exosite II binding of F1.2 was demon-  Table I. Titrations were performed and analyzed as described under "Experimental Procedures." strated here for the first time. F1.2 bound to thrombin with the same affinity as F2, whereas Pre 2 displayed a 16-fold increased affinity, demonstrating that F1 aids in the binding of F1.2 to Pre 2. The increased affinity of F1.2 for Pre 2 by contrast to thrombin indicates that the cleavage of Arg 320 -Ile 321 that causes the activating conformational change is also linked to reduced affinity of F1.2 for exosite II. This decrease in affinity on formation of thrombin would be expected to facilitate release of thrombin from the prothrombinase product complex. The more favorable interaction of F1.2 with the catalytic domain only in the zymogen forms is correlated with the effect of F1 to reduce proexosite I affinity. The combined results indicate that F1 either interacts directly with the Pre 2 domain in Pro or induces a conformational change affecting the affinity of F1.2 for exosite II and exosite I affinity for hirudin 54 -65 . Other studies support the possibility that F1 is in proximity to the Pre 2 domain and may interact directly to enhance affinity and reduce access to exosite I. In the absence of calcium, recombinant MzT cleaves the Gla domain at Arg 54 -Asp 55 and in the presence of calcium native MzT cleaves Arg 155 -Ser 156 to release F1 in intramolecular reactions (40,41). Analysis of the x-ray crystal structure of bovine MzT(-F1) suggests that the Arg 54 -Asp 55 bond in F1 may be within 15 Å of the catalytic site, in a compact structure for Pro in which the F1 and Pre 2 domains are in close proximity (42).
Conformationally distinct exosite I environments on zymogen and active forms of Pro may have importance in Pro substrate recognition by the factor Xa⅐factor Va complex. Previous studies of Pro domain-deletion mutants and effects of activation fragments on Pro activation have provided evidence that the Gla domain and kringle 1 of F1, and F2 may all participate in Pro-factor Va interactions (23)(24)(25). In the absence of membranes, F2 and F1.2 accelerate Pre 2 activation substantially in a factor Va-dependent manner (18,19), supporting the early hypothesis that F2 mediates Pro-factor Va binding (18). A detailed quantitative analysis of Pre 2 activation by the lipid- assembled factor Xa⅐factor Va complex, however, demonstrated that acceleration of Pre 2 activation by factor Va is largely independent of F2 (19). These studies concluded that factor Va-dependent substrate recognition exosites that mediate productive binding of Pre 2 are expressed on factor Xa within the factor Xa⅐factor Va complex. In this model, exosite binding of substrate results in formation of an initial encounter complex with the factor Xa⅐factor Va complex in which the factor Xa active site is accessible, followed by isomerization of the complex to engage the catalytic site and permit bond cleavage (20 -22, 43). The role of factor Va-substrate interactions in the binding and conformational changes of this model is not clearly understood. Our studies support an important role for sites on the Pre 2 domain in factor Va-mediated substrate recognition. Pro and thrombin both bind to the heavy chain subunit of factor Va, and the thrombin interaction is mediated by exosite I (26, 27, 44 -46). The demonstrated dependence of factor Va-rate acceleration of Pro activation on the low affinity, proexosite I on Pro is postulated to reflect its mediation of productive binding of Pro substrate species to factor Va within the factor Xa⅐factor Va complex (13,14). This interpretation is supported by recent mutagenesis studies demonstrating critical roles of residues in proexosite I on factor Va-dependent Pre 1 substrate recognition (15). On the basis of the available information, substrate recognition exosites expressed on factor Xa upon binding to factor Va and proexosite I-mediated factor Va-Pro interactions both likely play important roles in the recognition mechanism.
The involvement of proexosite I in factor Va interactions suggests that modulation of proexosite I affinity by F1 and full activation of exosite I affect interactions of the Pro activation species with factor Va. Proexosite I interactions may contribute to factor Va regulation of the activation pathway through the differential expression of exosite I on the alternative reaction intermediates. Because the effect of the F1 domain to decrease exosite affinity on the zymogen forms, Pro and Pre 2⅐F1.2, is also lost on cleavage at Arg 320 -Ile 321 , this domain may also alter factor Va-substrate interactions. The interconnectedness of Pro domains and proexosite I expression shown here suggests that domain deletion mutants of Pro are unlikely to have simple additive functional properties. Thus, whether the effects of deletions of the kringles in the F1 and F2 domains truly reflect direct involvement of these domains in factor Va binding or indirect effects due to disruption of domain-domain interactions and changes in exosite affinity remains unclear (24).
Kinetic studies of the four partial reactions of Pro activation have shown that factor Va redirects the activation pathway from one in which Pre 2⅐F1.2 predominates to one in which MzT is the major intermediate (5)(6)(7). This can be explained by differential effects of factor Va on cleavage of Arg 320 -Ile 321 and Arg 271 -Thr 272 , where initial cleavage at Arg 320 is greatly accelerated by factor Va compared with Arg 271 (6,7,47). The result is preferential acceleration by factor Va of MzT formation from Pro and thrombin⅐F1.2 formation from Pre 2⅐F1.2. As shown here, the cleavage reaction most stimulated by factor Va is also that in which the catalytic site and exosite I are activated, and the modulating effect of F1 is lost. Factor Va has much less effect on cleavage of Arg 271 in either Pro conversion to Pre 2⅐F1.2 or MzT to thrombin⅐F1.2, for which there is little or no associated change in exosite affinity. The results suggest that expression of proexosite I may regulate the preferred order of bond cleavage by altering factor Va-substrate interactions. On this basis, we speculate that proexosite I-mediated binding of Pro or Pre 2⅐F1.2 to factor Va may direct cleavage at Arg 320 , whereas the expression of increased affinity of exosite I on MzT may reorient the factor Va-bound substrate to favor cleavage at Arg 271 .