Previous studies from our laboratory and others have demonstrated that zymogen factor XI (FXI) binds to heparin with moderate (KD ∼110 nM) affinity via residues (K252, K253 and K255) located in the Apple 3 (A3) domain. In contrast, the enzyme, FXIa, was shown to bind to heparin (

Biochemistry
40
:
7569
–7580,
2001
) with significantly higher affinity (KD ∼9 nM by SPR and ∼1.5 nM by ELISA) via residues (K529, R530 and R532) within the catalytic domain (CD). This interaction potentiates by ∼10-fold the inhibition of FXIa by protease nexin-2. Also, polyanions heparin and dextran sulfate inhibit the catalytic activity of the enzyme factor XIa. The present study was designed to determine the relative contributions of positively charged residues as well as the dimeric structure of FXI to heparin binding. Mutational analysis of full-length FXI expressed in HEK293 cells was based on the following criteria:

  1. Conservation of the positively charged residues in FXI among various species;

  2. Surface exposure of the residues based on the X-ray crystal structure of FXI (

    Papagrigoriou E, McEwan P, Walsh PN, Emsley J,
    Nat. Struct. & Mol. Biol
    .
    2006
    ;
    13
    :
    557
    –558
    ); and

  3. comparison with human plasma prekallikrein (PK), which does not bind heparin.

Two positively charged residues Arg507 (147, chymotrypsin numbering) and Arg532 (173) are conserved in FXI genes of all species for which sequences are available. In human PK, Arg507 is replaced by lysine, while Arg532 is replaced by a neutral glutamine. We have expressed and purified wtFXI, R507A, R532A as well as monomeric C321S/K331A and C321A/I290A. While wtFXI, R507A and R532A demonstrated normal activity in APTT assays; monomeric FXI mutants retained 60-70% activity. The R532A and R507A mutants demonstrated ∼75% decrease in total number of heparin binding sites based on the solid phase ELISA assay using 5F7 monoclonal antibody. Also, the apparent dissociation constants for R507 (11 nM) and R532A (22 nM) were 7 and 11-fold increased respectively compared with 1.6 nM for the wtFXI. We also characterized monomeric FXI C321S/K331A and C321A/I290A proteins for their ability to bind to heparin compared with wtFXI using surface plasmon resonance (SPR). Surprisingly, the monomeric FXI mutants, C321S/K331A and C321A/I290A, which had no mutations in any heparin-binding regions, displayed major defects in binding to heparin by SPR. Although kinetic analysis is challenging due to complex binding kinetics, while Rmax is about 10-fold lower, the off-rate for the binding of the monomeric FXI mutants is drastically increased when compared to that of wtFXI. These results suggest the possibility that the unique dimeric structure of FXI is required for cooperative binding to heparin. Thus, the dimeric structure of FXI and basic residues R507 and R532 in the catalytic domain of factor XI are both necessary for high-affinity heparin binding.

Author notes

Disclosure: No relevant conflicts of interest to declare.

Sign in via your Institution