Researchers at Wyeth are testing GSI-953, a thiophene sulfonamide derivative. At the SfN meeting, Steven Jacobsen at Wyeth Research in Princeton, New Jersey, reported that the compound preferentially inhibits processing of APP over that of Notch. Also called Begacestat, GSI-953 was described last month in the Journal of Medicinal Chemistry (see Mayer et al., 2008). In cell-free and cell-based assays, the inhibitor blocks production of Aβ42/Aβ40 with an EC50 of about 10-15 nM, which is more potent than the γ-secretase inhibitor DAPT (N-[N-(3,5-difluorophenylacetyl-l-alanyl)]-S-phenylglycine t-butylester), reported Jacobsen. Begacestat is less potent for Aβ reduction than Eli Lilly’s LY411575 (see ARF related news story), but is more potent than LY450139, which is in Phase 3 clinical trials. However, GSI-953 excels in selectivity. Jacobsen showed that the Notch/APP EC50 ratio is about 16.8, which beats the 0.9 and 2.4 for LY411575 and LY450139, respectively. The data “suggests that GSI-953 selectively inhibits cleavage of APP while sparing Notch processing,” said Jacobsen.

Exactly how GSI-953 works is not entirely clear, but it competes with other compounds that bind near the γ-secretase active site. In Tg2576 transgenic mice it reduces Aβ by about 40-60 percent in the CSF and brain. It works rapidly; within 15 minutes reduction in brain Aβ is apparent, said Jacobsen. This, incidentally, demonstrates that Aβ has a very short half-life. The compound also reverses contextual-fear memory deficits. Wyeth has taken the compound into Phase 1 clinical trials (see ClinicalTrials.gov), and preliminary data suggest that it causes a dose-dependent decrease in plasma Aβ in humans, followed by a rebound. This is typical for γ-secretase inhibitors and likely means that the drug has engaged its target in the periphery. “This does not necessarily reflect Aβ levels in the brain, since Aβ levels were not increased in the brain of TgAPP mice,” said Jacobsen.

Speaking at ICAD, Charlie Albright, Bristol-Myers Squibb, Wallingford, Connecticut, reported the same rebound plasma Aβ dynamic in both animal models and in human volunteers with BMS-708163, his company’s new Notch-sparing γ-secretase inhibitors. This compound decreases CSF Aβ40 at a safe and tolerable dose in humans, Albright told the audience. It has an IC50 for APP cleavage of 0.3 nM compared to 58 nM for Notch. The compound dose-dependently lowers Aβ in the brain of rats and dogs and similarly reduces CSF Aβ40. Perhaps in keeping with the substrate specificity, the compound had no gastrointestinal, thymus, or spleen effects in animals treated with 10-fold the dose needed to see reduction in brain Aβ. Gastrointestinal and immune system problems have been documented in γ-secretase inhibitors in the past (see ARF related news story).

According to Albright, Bristol-Myers Squibb has also taken this compound into Phase 1 clinical trials, though ClinicalTrials.gov does not list an active trial at present. (It does list a completed drug interaction study.) At ICAD, Albright reported that up to 150 mg of the compound has been given to healthy volunteers for up to 28 days. Pharmacokinetics shows that the drug disappears from the plasma in a biphasic fashion, with a second phase half-life of about 40 hours, Albright said. The researchers used continuous lumbar catheterization in humans to measure changes in CSF Aβ. They saw a dose-dependent reduction in CSF Aβ after single and multiple doses of the compound. After the 28-day study, trough levels of CSF Aβ40 and CSF Aβ42 were reduced by about 30 percent at 100 mg and 60 percent at 150 mg doses.

When asked, Albright responded that there were no serious adverse events associated with this Phase 1 trial. Of course, further study will be needed to test the efficacy and safety of both these compounds. For an up-to-date review on γ-secretase inhibitors, see Imbimbo, 2008.—Tom Fagan.