SummaryIt is well known that on artificial surfaces, binding and autoactivation of factor XII (FXII) is the initiating event of plasma prekallikrein (PK) activation. We performed investigations to examine whether this mechanism was true for FXII activation on endothelial cells (HUVEC). Activation of PK on HUVEC required an optimal substrate and Zn2+ concentration, the latter of which varied with the buffer’s carrier protein. Maximal PK activation required the addition of 250 μM or 10 μM Zn2+ to buffers containing bovine serum albumin (BSA) or gelatin, respectively. However, the actual free Zn2+ concentration in these buffers was the same at 8 μM. In both BSA- and gelatin-containing buffers and using two different chromogenic substrates for FXII, no autoactivation of FXII on HUVEC was seen when incubated for up to 60 min. Rather, initiation of FXII enzymatic activity required the presence of PK. FXII activation after PK activation contributed to the extent of measured enzymatic activity, but its role was secondary because treatment with corn trypsin inhibitor or a neutralizing antibody to FXIIa did not abolish the measured enzymatic activity. They also reduced the activity to the level seen with PK activation alone. Alternatively, soybean trypsin inhibitor abolished the proteolytic activity associated with PK and FXII activation on HUVEC. Further, only normal human and FXII-deficient plasmas, not PK-deficient plasma, had the ability to generate proteolytic activity when incubated over endothelial cells. In a purified system, maximal PK activation was measured after a 10-15 min incubation depending upon the concentration of reactants. When FXII was added with the PK, maximal activation occurred within 7.5-10 min. In normal human or FXII-deficient plasmas, but not in PK-deficient plasma, maximal activation was seen in 4 min. These data indicate that on HUVEC, unlike artificial surfaces, PK activation when bound to HK is the initiating activation event in this system. FXII activation is secondary to PK activation and contributes to the extent of measured enzymatic activity. These data challenge the accepted dogmas of “contact activation” and suggest that on biologic membranes a new notion as to how this system is activated needs to be considered.