However, such studies have been impeded by the lack availability of effective PDE8 inhibitors until the recent report of Pfizer��s PF-04957325. To discover PDE4/8 inhibitors as well as novel PDE8 inhibitors, we optimized and conducted a 222,711 4-IBP compound HTS based on PKAregulated growth of an S. pombe strain that expresses murine PDE8A1 as its only PDE, with follow-up screens of candidate compounds against PDE4-expressing strains. This led to the identification of BC8-15, which potently elevates steroidogenesis when used alone in mouse Leydig cells. A screen of known bioactive compounds did not identify an effective PDE8 inhibitor. Although dipyridamole, which nonselectively inhibits PDE8, was in the collection, it showed no activity in this screen. This appears to be due to poor solubility of dipyridamole in the yeast medium or its inability to be taken up by yeast, as we have not detected a growth response to dipyridamole with either PDE5A- or PDE8A-expressing strains. From the bioactive compound collection, epiandrosterone gave the highest composite Z score- indicating a statisticallysignificant response relative to negative control wells; although the increase in optical density was modest. While it is tempting to speculate that epiandrosterone, a metabolite of dehydroepiandrosterone, could be a natural regulator in steroidogenesis, it only weakly inhibits recombinant PDE8A in an in vitro enzyme assay. While we had also identified steroids in our PDE7 inhibitor screen, in this case they were among our most effective hit compounds in contrast to the weak effect of epiandrosterone in this current study. The HTS against other commercial libraries identified 2,266 compounds that promoted growth of the PDE8A-expressing strain to a statistically-significant level relative to DMSO-pinned controls. Fifty-six of the 63 strong hits that produced OD values of.0.6 came from only six libraries that represented 37% of the screened compounds. These libraries had a hit rate of 0.07%, in comparison to 0.005% for the other 19 libraries. Examination of hit compounds revealed few structurallyrelated compounds in the more productive libraries, thus, the tendency of related compounds to be present in a given library cannot account for the widely varying hit rate. Although we have not assessed the compound characteristics among libraries in detail, compounds in a given library may have similar physicochemical properties that result in favorable solubility and stability in 5FOA medium or allow uptake by yeast cells. As such, these libraries might serve as a prioritized collection for future yeast ACPA growth-based screens. Eleven of the 30 compounds acquired based on 5FOA-growth results inhibited PDE8A effectively in in vitro enzyme assays.