28 July 2006. This is part 2 of our 3-part series. Also see part 1 and part 3 or download PDF.
Anti-BACE Drugs Appear on Horizon
On BACE inhibition, Martin Citron of Amgen in Thousands Oaks, California, noted that more than 100 patent applications have been filed, and a growing number of non-peptide inhibitors are being published now that researchers have gained more experience with BACE1. High-throughput screening against BACE was largely unsuccessful as nothing of use stuck to BACE1, so companies switched over to rational design based on the crystal structure. In the patent literature, hydroxyethylamines are a central chemical theme of BACE1 inhibition. In Madrid, scientists for the first time began introducing compounds that appear to work in vivo.
James McCarthy, of Eli Lilly and Co. in Indianapolis, presented the first data on a BACE1 inhibitor that seems to have turned the corner in BACE1 drug development after years of frustration. McCarthy noted that the Eli Lilly team had long worked with a group of sulfone and sulfonamide compounds that are highly potent, but whose physicochemical properties stubbornly kept them outside the blood-brain barrier. The scientists then decided to scrap this class of compounds and instead go after others that started out with a slightly lesser affinity to BACE1 but with more attractive physicochemical properties, such as a lower molecular weight, lower polar surface area, and other parameters.
After describing stereochemical modifications to their initial lead compounds, McCarthy presented an experimental BACE1 inhibitor, LY2434074. This is the first publicly shown BACE1 inhibitor that enters the brain of PDAPP mice and reduces sAPPβ, the product of BACE1 cleavage, in cortex and hippocampus in a dose-dependent manner, McCarthy noted. The product of the alternative α cleavage that processes APP in the absence of BACE1, that is, sAPPα, went up in the brains of the injected mice. Aβ levels decreased in CSF and in plasma, as detailed in a subsequent poster presented by Patrick May of the same group.
Other scientists confirmed that this approach for the first time has demonstrated proof of principle for BACE1 inhibition in brain by a chemical given systemically. They also pointed out that the compound McCarthy presented likely is not the one the company is pursuing for clinical development. It had to be injected in rather large doses, implying problems with its oral availability or possibly its metabolism. Indeed, McCarthy replied in response to a question that Eli Lilly has more suitable compounds in hand. Colleagues from other drug development companies applauded Eli Lilly’s decision to present a potent structure. They added that other firms also have overcome some of the structural challenges posed by BACE1’s unwieldy active site. Indeed, Sethu Sankaranarayanan and colleagues from Merck’s team in West Point, Pennsylvania, presented evidence that intravenous injection of their own inhibitor lowers Aβ in the brain of Bruce Lamb’s human wild-type APP-transgenic mice.
Toward a BACE Vaccine
If anti-Aβ antibodies hold promise, why not hit BACE1 in the same way? Two groups reported progress toward this goal in Madrid. Wan-Pin Chang, in Jordan Tang’s group at the Oklahoma Medical Research Foundation in Oklahoma City, followed in the footsteps of Aβ immunotherapists and injected Tg2576 mice with BACE1. Chang’s prior experiments had detected reduced Aβ production in cultured cells treated with polyclonal anti-BACE1 antibodies, and he had also noticed that a fraction of injected anti-BACE1 antibodies entered the brain of mice. The underlying rationale for his approach, Chang said, would be that antibodies stick to BACE1 on neuronal cell surfaces and prevent its internalization into endosomes, where BACE1 cleavage of APP finds a conducive pH of 4.5.
In Madrid, Chang described a two-pronged study of active immunization with recombinant BACE1. A prevention arm began injecting BACE1 into Tg2576 mice repeatedly at 1 month of age, and a treatment arm began injecting BACE1 at 10 months, when plaques form. The scientists tracked the mice’s behavior and measured Aβ levels at 15 months or 23 months, respectively. In both study arms, but more so in the preventive one, the scientists measured rising antibody titers and waning Aβ40 and Aβ42 levels in serum and brain, as well as a reduced plaque load in brain as the immunization protocol progressed. Immunized mice outdid the untreated mice in negotiating the Morris water maze, Chang added. T cells, microglia, and astrocytes showed no sign of activation.
Michal Arbel, who works with Beka Solomon at Tel-Aviv University in Israel, took a different tack. Because BACE1 cleaves not only APP, Arbel works on devising an immunotherapy that interferes specifically with the BACE1-APP interaction rather than inhibiting or eliminating BACE1 altogether. Arbel develops antibodies directed against the BACE1 cleavage site on APP, which bind to human wild-type APP and human APP carrying the Swedish FAD mutation, but not to Aβ itself (see ARF related news story). (On a broader note: Scientists are realizing, to their surprise, that immunotherapy in mice works quite well across the board. Ajodeji Azuni, working with Einar Sigurdsson at New York University School of Medicine reported initial data of a tau vaccine. A P301L tauopathy mouse model responded to active vaccination with a phospho-tau peptide by mounting a tau-specific antibody response, showing less tau pathology, and performing better on some sensorimotor tasks.)—Gabrielle Strobel.