1 The beta-secretase BACE1 is one of the enzymes that contribute to the production of the Abeta peptide, in vitro and in vivo. JMV1195 was previously shown to inhibit BACE activity in vitro but was unable to block cellular BACE activity. We have designed a new permeable inhibitor, JMV2764 that corresponds to a derivative of JMV1195 to which a penetratin sequence had been added at its N-terminus. We have assessed the ability of JMV2764 to affect BACE1 activity in vitro, and to modify Abeta production in various cell systems. 2 Endogenous beta-secretase or BACE1 activities were monitored in vitro by means of two distinct fluorimetric substrates in HEK293 extracts of cells expressing either wild-type betaAPP, Swedish mutated betaAPP or SPA4CT constructs. Abeta40 recovery was monitored by immunoprecipitation and Western blot analysis. 3 JMV2764 and JMV1195 inhibited endogenous beta-secretase activity of HEK293 cellular homogenates with IC(50)s of 0.8 and 6.6 microM, respectively. Interestingly, JMV2764 also inhibited beta-secretase activity after preincubation with intact cells while JMV1195 was inactive, indicating that unlike JMV1195, JMV2764 could penetrate into the cells. 4 JMV2764 but not JMV1195 also prevented Abeta production by HEK293 cells overexpressing wild-type and Swedish-mutated betaAPP. However, JMV2764 was unable to affect Abeta production from cells expressing SPA4CT, a betaAPP-derived sequence that does not need beta-secretase to produce Abeta. 5 Altogether, we have designed a new cell-permeable BACE1 inhibitor that allows to envision to prevent Abeta production in vivo. Work is in progress to assess the potential of these compounds to prevent Abeta production in transgenic mice overproducing Abeta.