This is Part 1 of a two-part story. See Part 2.
29 March 2008. The second annual Drug Discovery for Neurodegeneration meeting, a joint effort of the Alzheimer’s Drug Discovery Foundation (ADDF, formerly the Institute for the Study of Aging), the National Institutes of Aging and the National Institute of Neurological Disorders and Stroke, convened in Washington, DC, on 4-5 February. The goal of the yearly meeting, chaired by ADDF executive director Howard Fillit, is to speed the development of new therapies for neurodegenerative disease by getting academic researchers more involved in the hunt. This year, for the first time, the ADDF and Alzforum are making the slides and audio of the presentations available online. The webcast is organized by session, so to locate a specific talk, consult the meeting program, then select the appropriate session on the ADDF website. Once the recorded slide session has begun, specific talks or slides can be viewed by selecting from a left-hand navigation bar. Viewers will need to go through a brief, one-time registration.
The problem, as Fillit outlined it in his introductory remarks, lies in the gap between how far academic researchers usually go in uncovering potential therapeutic targets, and the stage at which a target is considered sufficiently validated, or a compound sufficiently promising, that pharmaceutical or biotech companies will take a chance on the project. Bridging that gap requires academic researchers to move into unfamiliar territory. How, one might wonder, does a company pick which targets to study? What can academics do to interest pharma in their favorite targets? How can a university researcher most efficiently move a promising project forward?
“More and more, big pharma is getting involved in neurodegenerative disease,” Fillit said. “This meeting is about trying to build an academic base that can meet pharma halfway.” Toward that end, talks from academic, industry, and government researchers covered concepts including target discovery and validation, hits to leads, lead optimization and so on. (If some of those terms sound entirely foreign, an additional resource is the Alzforum/ADDF tutorial on drug development.) For Alzheimer researchers, there were some tidbits of new results and interesting approaches, described in brief below and in Part 2 of our coverage.
Peter Reinhart (find his talk in Session I of the webcast) of Wyeth Research in Princeton, New Jersey, and Peter Seubert (Session II) of Elan Pharmaceuticals in South San Francisco, California, both talked about Aβ immunotherapy, with an eye to the complexities of deploying biologics as therapeutics. Their companies have been working together on both active and passive immunization strategies, with a large trial of passive immunization currently recruiting participants (see ARF related news story). Studies in mice and to a limited extent in humans indicate the antibodies clear plaque and improve cognition. How does that happen? Do antibodies trigger microglial activation to clear plaque, directly dissolve plaque, or lure Aβ out of the brain into the periphery? And how are those effects tied to cognition?
Both Reinhart and Seubert showed data addressing those questions by comparing the in vivo effects of monoclonal antibodies that recognize different epitopes on the Aβ42 peptide. When injected into AD mice, antibodies directed against the N-terminus of Aβ (3D6, 12A11) cleared plaque, but a central epitope antibody (266) did not. Likewise, in the contextual fear conditioning test of memory, some, but not all, N-terminal antibodies improved cognitive function. The effects of 3D6 and 12A11 lasted for 10 days after a single injection and were seen even in old animals with established plaque. In a summary of the effects of a panel of antibodies, Reinhart said that they found cognitive improvement with most N-terminal antibodies, and none of the C-terminal antibodies. Some, but not most of the central epitope antibodies were effective. “It seems simple to say you raise an antibody to Aβ and you will have a therapeutic, but it’s actually more complicated than that,” he concluded.
The ability to improve cognition appeared to correlate with the ability to bind soluble Aβ oligomers, as measured by Western blotting. Out of 10 antibodies tested, those that preferred to bind monomers tended not to show cognitive improvement. However, Reinhart cautioned, these are still the “early days” for this kind of study.
Seubert expanded on this data, showing an extensive comparison between the effects of the 3D6 and 266 antibodies. 3D6 recognizes an N-terminal epitope very similar to the antibodies raised by Aβ immunization in the AN1792 human trial, and binds to both plaques and soluble Aβ. The antibody 266, raised to a central epitope, binds plaques less efficiently but recognizes soluble Aβ. The researchers showed that 3D6, but not 266, prevented plaque formation and reduced existing plaque in PDAPP mice. These data were recently published (Seubert et al., 2008).
Additional data showed that weekly injections of 3D6, but not 266, also reduced vascular amyloid. Seubert said this was the first demonstration of reversal of vascular amyloidosis, and said they are now looking at vessel structure and function to determine if damage is also reversible.
However, 266 was not void of activity. Like 3D6, it did seem to spare synapses as indicated by synaptophysin staining, and improved behavior in the contextual fear-conditioning test. These effects may reflect 266’s ability to recognize toxic oligomers. Seubert concluded that the epitope recognized by Aβ antibodies may determine their mechanism of action. In addition, he said, the different properties of the various antibodies should allow researchers to test specific aspects of the amyloid hypothesis.
Building Better Inhibitors—It’s Academic
As a case study, Jordan Tang (Session II) of the Oklahoma Medical Research Foundation in Oklahoma City traced the design in his lab of β-secretase inhibitors. This effort led to the development of CTS21166 at CoMentis, Inc. in South San Francisco, California. The company has completed a Phase 1 trial, and Tang reported that a single intravenous dose of the compound caused a 70 percent reduction in plasma Aβ level in healthy volunteers. The company is embarking on another Phase 1 trial, he said, this time with oral dosing.
Peter Reinhart mentioned some preclinical data on Wyeth’s β-secretase inhibitor, WAY258131. Treating PSAPP mice for three months with the inhibitor caused a reduction in plaque as detected by immunohistochemistry on brain sections, and in vivo by two-photon microscopy.
Turning to another target not specific for AD, but of interest for stroke and neurodegeneration in general, Richard Silverman (Session II) of Northwestern University, Evanston, Illinois, described the design of inhibitors specific for neuronal nitric oxide synthase. Production of excess nitric oxide has been implicated in neuronal death after ischemia, although some studies point to a protective role in AD (see ARF related news story). Three NOS isoforms exist: neuronal (nNOS) is restricted to neurons, inducible (iNOS) functions in the immune response, and endothelial (eNOS) regulates blood pressure. Both iNOS and nNOS have been implicated in neurodegeneration, raising interest in specific inhibitors of these enzymes, which would not interfere with eNOS. However, all of the enzymes have nearly identical active sites, and extensive screening by several companies did not yield specific inhibitors.
To solve the problem, Silverman’s lab took a different approach, pursuing de novo drug design. They identified a promising start by modifying NOS’s arginine substrate to extend out into a region just outside of the active site that is different in nNOS compared to the other enzymes. To optimize the inhibitor for both potency and selectivity, Silverman and colleagues devised a process they call fragment hopping, where they separately screened for small chemical fragments that bound different parts of the enzyme active site. By including features that add metabolic stability and avoid toxicity, they identified a set of chemical building blocks that they then assembled, Lego-style, into one molecule. They put together inhibitors with nanomolar potency and 1,000-fold selectivity for nNOS (Ji et al., 2008). The inhibitors were active in vivo, protecting against neurodegeneration in a rabbit model of ischemia-induced cerebral palsy, with no effect on blood pressure.
Elias Michaelis (Session V) of the University of Kansas in Lawrence spoke about a new AD drug development program being carried out in an academic setting. Based on the neuroprotective actions of heat shock 90 protein inhibitors (see ARF related news story and Ansar et al., 2007), he and his colleagues are moving forward with development of new inhibitors.—Pat McCaffrey.
This is Part 1 of a two-part story. See Part 2.