Could β-secretase go the way of its cousin? Once valued as a target by Alzheimer’s drug developers, γ-secretase fizzled when inhibitors targeting the protease sped up cognitive decline and caused skin cancers in clinical trial volunteers. Researchers blamed the disappointing outcomes on interference with crucial biological processes driven by γ-secretase substrates, notably the cell-surface protein Notch (see Aug 2010 research news). With nearly a dozen pharmaceutical companies poised to develop BACE inhibitors, some researchers have been whispering concern that history may repeat itself. That sentiment boiled to the surface at BACE Proteases in Health and Disease, a meeting organized by Stefan Lichtenthaler, Technical University Munich, Germany. At the meeting, held October 6–8 at Kloster Seeon, a 10th-century Benedictine monastery 90 minutes east of the Bavarian capital city, researchers from academia and industry debated the recent science around BACE, including whether known and yet-to-be-discovered substrates might spell doom for potential BACE drugs.
BACE drug programs got off to a slow start because it was difficult to model the large, wide-open active site of the protease and design brain-penetrant drugs to fit it. Just as researchers solved that problem, another emerged. BACE1 knockouts, once thought to be innocuous, turned out to have subtle, potentially troubling side effects, including poor myelination, loss of dendritic spines, axon guidance defects, neural network defects, seizures, and even memory problems. At the same time, screens began turning up dozens of potential BACE substrates in addition to amyloid precursor protein. With BACE inhibitors already in Phase 1 to 3 clinical trials, some researchers began to stress that more work was needed to predict and prevent potentially dangerous side effects of chronic BACE inhibition.
Is Neuregulin the Notch of BACE Inhibitor Development?
The usually silent Benedictine monks might have wrung their hands over the hubbub 70 passionate scientists generated when they were cloistered on this beautiful island retreat for two days. A panel discussion captured the hot-button issues. “Neuregulin is the Notch of BACE inhibitor development,” was how Bart De Strooper, KU Leuven, Belgium, summed up a sinking feeling among academic researchers. That sent shivers down the spines of some industry representatives, who cautioned against throwing BACE out with the neuregulin bathwater.
Scientists at this meeting agreed that neuregulin is a bona fide BACE substrate, but opinion diverged on how that will affect drug development. Dieder Moechars, Janssen Research and Development, emphasized that it may be possible to reduce Aβ production with a partial BACE inhibition, limiting adverse events. Christian Haass, Ludwig-Maximilians University, Munich, agreed that moderate inhibition of BACE might work, but pleaded for a step-by-step approach to avoid the discouraging fallout that followed γ-secretase inhibitor (GSI) trials. GSI side effects were mechanism-based and predictable, Haass said, and the outcome of the clinical trials shook confidence among patients and funding agencies, casting public doubt over the amyloid hypothesis.
Many of the phenotypes that emerge by knocking out BACE can be traced to problems during development, such as poor myelination, lack of pigmentation, and stunted growth. Even so, BACE also regulates crucial processes in adults, Haass said, including maintenance of muscle spindles and dendritic spines. In addition to neuregulin, tens of potential BACE substrates have emerge from various screening approaches (see part two of this story). “There are still other substrates to be identified, and we have no idea what they do in the brain,” Haass said.
On that point, Bruce Albala, Eisai Inc., the fourth member on the panel, expressed a concern among industry researchers. “We need to know which of these phenotypes is more important,” he said. “We are in a business that is increasingly harder to sustain, and a compound that fails because of safety issues helps no one.” Albala said that a consensus paper from the academic community that outlines the most important safety issues would help all industry navigate the BACE drug-development process. “We can’t wait for academia to exhaustively research and vet all the potential substrates to get the go-ahead for industry to begin development,” he added.
Patrick May, Eli Lilly and Company, agreed. May further stressed that inconsistency between the preclinical models presents a problem for pharma. Phenotypes often are not reproducible from lab to lab. “Furthermore, it is challenging to take controversial preclinical data and apply it to clinical studies,” he said. Ryan Watts, Genentech, echoed that sentiment. “For example, we don’t see the ocular phenotype, and the models may not always predict what will happen in humans treated with BACE inhibitors,” he said (see Cai et al., 2012). Mark Albers, Harvard Medical School, Boston, said it is unfair to look at every drug through the prism of side effects in animals. Albers noted that given what researchers now know about statins, one of the most successful drug classes ever, they might have never brought them to market because statins affect many more targets that the intended one, cholesterol biosynthesis. “Attempts to predict what will happen in humans are often inaccurate,” he said.
On phenotypes, Lichtenthaler concurred that the field needs to reproduce results in different models. “We really need to think about whether we are looking at effects that are due to BACE and its substrates, or due to the model itself,” he said. That some phenotypes cannot be reproduced does not automatically mean the results are wrong; rather, it could reflect modifier effects unique to specific genetic backgrounds. Lichtenthaler cautioned against over-interpreting knockout data. “A knockout is not the same thing as a drug-treated animal,” he said. Genetic knockout can remove a gene entirely from birth, allowing compensatory mechanisms to develop in some instances, whereas drug treatment reduces the target partially during aging.
For their part, pharmaceutical companies seem to be taking basic research discoveries seriously. Matthew Kennedy of Merck described how his company compared knockout phenotypes reported in the literature with their in-house models and with BACE inhibitor-treated animals. He noted that they have never seen seizures, impaired cognition, or retinal neurodegeneration with their inhibitors. Merck’s researchers have, however, noticed hypomyelination, reduced prepulse inhibition, and slower nerve conduction with some compounds. Merck’s lead compound has undergone Phase 1 testing (see Jul 2012 news story) and enrollment for a Phase 2/3 trial is underway (see March 2013 news story). Haass said he was impressed that companies are working on these phenotypes step by step.
De Strooper deplored a general lack of feedback from pharma once drugs enter trials. “There is a black hole. That compromises communication between companies and academia,” he said. De Strooper challenged industry to share samples and publish data, especially when things go wrong. Others objected to singling out industry for failing to publish negative data when the problem is rife among academic labs, as well (see May 2013 news story; Sep 2004 news story). While pharma does not share with everyone, they do share samples with collaborators. Lilly’s May cited the analysis of CSF samples from Lilly trials of ß- and γ-secretases by Erik Portelius and colleagues at the University of Gothenburg, Sweden (see February 2012 news story; Portelius et al., 2012).
One easy point of consensus was that developing drugs for AD has become so expensive that a new approach may be needed. “The way we develop drugs right now suits antibiotics and other drugs with short duration of action, but for drugs that will have to be tested and used long-term it hardly seems feasible,” said De Strooper. “We need to figure out how our respective societies as a whole can facilitate trials by taking on part of the burden.”—Tom Fagan
- Lilly Halts IDENTITY Trials as Patients Worsen on Secretase Inhibitor
- BACE—Substrates, Functions, Developmental Phenotypes
- Wave of New BACE Inhibitors Heading to Phase 2
- BACE Inhibitors Barrel Forward—Next Hurdles: Safety, Efficacy
- Guidelines at Nature Aim to Stem Tide of Irreproducibility
- Not the Usual Suspects: Tracking BACE Inhibition, Axon Role
- Cai J, Qi X, Kociok N, Skosyrski S, Emilio A, Ruan Q, Han S, Liu L, Chen Z, Bowes Rickman C, Golde T, Grant MB, Saftig P, Serneels L, De Strooper B, Joussen AM, Boulton ME. β-Secretase (BACE1) inhibition causes retinal pathology by vascular dysregulation and accumulation of age pigment. EMBO Mol Med. 2012 Sep;4(9):980-91. PubMed.
- Portelius E, Zetterberg H, Dean RA, Marcil A, Bourgeois P, Nutu M, Andreasson U, Siemers E, Mawuenyega KG, Sigurdson WC, May PC, Paul SM, Holtzman DM, Blennow K, Bateman RJ. Amyloid-β1-15/16 as a Marker for γ-Secretase Inhibition in Alzheimer's Disease. J Alzheimers Dis. 2012 Apr 18; PubMed.
- Wave of New BACE Inhibitors Heading to Phase 2
- Barcelona: Out of Left Field—Hit to The Eye Kills BACE Inhibitor
- Paris: Semagacestat Autopsy and Other News of Trial Tribulations
- Drug Company Halts Development of γ-Secretase Inhibitor Avagacestat
- Satori Pharmaceuticals Shuts Down, Abandons γ-Secretase Modulators