In vitro studies started a few years ago and completed by LM last year showed that the IgG fraction of sera from patients previously shown to have developed high titers of plasminogen-peptide cross-reactive antibodies could inhibit plasmin activity in vitro. This inhibitory effect was synergistically enhanced in the presence of alpha-2-antiplasmin, the physiologic inhibitor of plasmin activity. In contrast, IgG from flavivirus seronegative controls had no such effect. This provided the first evidence that plasmiongen-peptide cross reactive antibodies could actually bind plasmin and suggested that such antibodies might play a role in bleeding in both DF and DHF. A manuscript was written this year and submitted for publication. No new laboratory work was initiated during 1996-97. Infection with dengue viruses often leads to bleeding. LM previously showed that a 21-amino-acid domain in the dengue virus E protein gives rise to antibodies that cross-react strongly in vitro with a peptide representing a 20-amino-acid domain in plasmin, which is a clotting factor responsible for the removal of fibrin from the circulatory system. These antibodies were present in 75% of Thai patients experiencing DF or DHF/DSS. A subsequent study conducted among dengue-infected children in Tahiti showed a correlation between the occurrence of bleeding and the development of plasminogen cross-reactive antibodies. Unpublished in vitro studies conducted by LM show that plasminogen cross-reactive antibodies act synergistically with a-2-antiplasmin, the naturally occurring inhibitor of plasmin, to reduce plasmin activity further. This study also demonstrated that antibodies produced in vivo against E can bind plasmin. This undesireable cross-reactive response could be elicited by any potential dengue vaccine and needs to be considered in the approval process.