This story originally ran on July 13.
By Tony Fong
Name: Tony Wyss-Coray
Postion: Associate professor, department of neurology and neurological sciences, Stanford University School of Medicine, 2005 to present; research career scientist, Veterans Affairs Palo Alto Health Care System, 2007 to present
Background: Assistant professor, department of neurology and neurological sciences, Stanford University School of Medicine, 2002 to 2005; health scientist, GRECC, Veterans Affiars Palo Alto Health Care System, 2002 to 2007
The protein breakdown product beta-amyloid has long been implicated in Alzheimer's disease, and clinical trials are currently being conducted to test whether immunization with beta-amyloid antibodies can reduce amyloid plaque in the brain.
But more than the plaques themselves, it may be smaller aggregations of beta-amyloid molecules called oligomers that are more toxic to neurons, and so, of greater interest in Alzheimer's research. In a study published online July 6 in the Proceedings of the National Academy of Sciences, researchers used peptide microarrays to demonstrate the presence of antibodies naturally occurring in humans against known toxic beta-amyloids and amyloidogenic non-beta-amyloid species in blood and cerebrospinal fluid samples of 250 AD patients and healthy controls.
The antibodies targeted multiple forms and aggregation states of beta-amyloids in both healthy and diseased samples with antibodies against oligomers showing the most immunoreactivity. The researchers also showed, for the first time, that these overall levels of antibodies decline with increasing age.
"Our findings support the concept of conformation-specific, cross-reactive antibodies that may protect against amyloidogenic toxic peptides," the authors said in the PNAS study. "If a therapeutic benefit of beta-amyloids can be confirmed in AD patients, stimulating the production of such neuroprotective antibodies or passively administering them to the elderly population may provide a preventive measure toward AD."
ProteoMonitor spoke with Tony Wyss-Coray, the senior author on the study and an associate professor of neurology and neurological sciences at the Stanford University School of Medicine, last week about the research. Below is an edited transcript of the conversation.
Describe the microarrays that you used for your research.
These are standard epoxy glass slides that we used, and printed different types of peptides that are involved in neurodegenerative diseases, specifically Alzheimer's disease.
We printed peptides on these slides and incubated them with plasma from patients with disease or control. The array is a pretty standard proteomic array, [but] instead of printing antibodies like some people do, we print peptides and then try to find antibodies that bind to these peptides in plasma.
When you customized these arrays, was it more in the content than the actual array itself?
Yes, the customization was the peptides themselves, and … some of them we purchased, some we synthesized, and a lot of them were from one of our collaborators who is an expert in amyloid peptides, so we had a lot of these various different peptides.
And then we either spotted them unmodified, or we incubated them first in certain buffers, so that they would form aggregates. By using different types of buffers, you can steer them toward making more oligomer-type aggregates, or to make fibrils, which are longer or higher order magnitude aggregates.
These different types of preparations were then spotted in an array fashion on the slides.
You have a provisional patent on the technology. What does this cover?
It's actually just an invention disclosure. …This is really [for] the approach of using the different conformations and different types of peptides to monitor disease and monitor maybe efficiency of a vaccine.
The interesting aspect of this work is that one of the most promising treatments for Alzheimer's in the pipeline is a vaccine, or monoclonal antibody that binds to amyloid. So you immunize with amyloid peptides, or you use monoclonal antibodies, humanized antibodies. However, if you use the monoclonal antibody, then of course you know what the specificity is.
[ pagebreak ]
But if you actively immunize people, they may produce all kinds of different antibodies, and some of them may be more beneficial than others. We know that this immunization is quite effective in getting rid of the amyloid in the brain in mice, in monkeys, and in people, but the clinical benefits are not yet well established.
Even though people seem to be able to get rid of this amyloid in their brains, their brains are still not working much better. … However, some people do actually better, but it's not clear who responds and who doesn't.
And our antibody array may provide a way to look [at] whether the type of antibody that different individuals make could be correlated with clinical benefits.
Is that why you chose the microarray platform?
Exactly. First of all [we wanted] to find out whether people already have such antibodies naturally without the vaccine, but then with the idea that this could be used in clinical trials to monitor which patients [have] a good response and which ones [have] a response that correlates with a significant clinical benefit.
How much have you found in terms of this being a natural antibody? Have you hypothesized that everybody manufactures this naturally?
Yes, everybody has this antibody. Actually [everybody has] an amazing repertoire of these antibodies. What we don't know is if they are significant in delaying this disease, or preventing this disease, or if they have any physiological role, [or if] they're just auto-antibodies and the titers are not very high. We estimate that they're about 100-fold to 1,000-fold lower than the doses that are being used in the treatment, in the passive immunization trials.
But [in the clinical trial] you give a bolus of the antibody every four months.
What do you know about this trial that you can share? Any data?
The active immunization trial [conducted by Elan Pharmaceuticals] took place [in 2001 and 2002]. It had to be stopped because some people developed meningomyelitis and cephalomyelitis. They actually had an inflammation in the brain. What's speculated is their immune [system] made more of a T-cell response rather than just antibodies. So the immune system recognized the peptides more like a virus, where you want to have a T-cell response rather than just something that has to be neutralized.
That's why the trials were stopped. But some of the individuals who later died, when their brains were analyzed, it showed that they actually had less amyloid in their brains than control people who were not immunized.
So that tells us the vaccine approach is working in principle in clearing the amyloid from the brain, but the trial was not completed. It was not [possible] to really analyze whether there really are cognitive benefits.
There were some individuals who clearly had some long-lasting benefits. But whether that was [due to] the vaccine or something else, you can't really make that analysis [based on the trial].
That's why companies moved into just using passive immunization, where they use just one specific monoclonal antibody. Those trials are still underway and they look promising, but they're not going to be the final treatment for this disease.
They may have slight benefits, but it's still too early to make a firm conclusion.
Would your research provide a new pathway to develop a vaccine or would it provide something completely different from the clinical trials underway?
I think our approach could help [in] monitoring people who are actively immunized, and retrospectively look at clinical data from trials that have been done to see whether people who showed some clinical benefit had a certain type of antibody.
That's where it could really help — in learning more about what the potential beneficial antibodies are that people produce. If we find an antibody against an oligomer from, let's say, a phosphorylated form of a peptide, then you would probably want this type of peptide for immunization or even make a monoclonal antibody and use it in passive immunization trials.
[ pagebreak ]
Can you describe how your research builds upon what we already knew about Alzheimer's, especially in relation to the beta-amyloids that your paper mentions?
What was previously known was that people do have antibodies against A-beta. It was conflicting — some people said Alzheimer's patients have more [of this peptide], others said they have less. And what was previously shown was for one type of A-beta.
It was not clear whether the conformation, the post-translational modifications, whether any of this made a difference. People didn't look at the whole repertoire of potential antibodies, and that's where we came in.
We looked at a larger set [of patients and controls]. … We also looked at normal aging and found that young people who have no Alzheimer's, who have no exposure to the disease in their brains, already have all kinds of antibodies against this amyloid and other amyloid peptides, suggesting that these antibodies are cross-reactive.
And they seem to decrease as people get older, so the youngest people have the highest titers, and the oldest have the lowest titers, and that would be consistent with the potentially protective role of these antibodies. They might actually neutralize amyloid peptides that we always produce in our bodies, and as we get older, we have less of these antibodies and may become susceptible to this disease.
Now, we didn't prove that, but that's one possible interpretation.
We also found that we have a lot of antibodies against amyloid peptides that we don't produce ourselves, again supporting this concept of cross-reactive antibodies. And that opens the possibility that other than immunizing patients with the A-beta peptide itself — the peptide that accumulates in the brain — that you can immunize potentially with something much more potent.
There was this discussion about this potato virus that has an amyloid structure that another group had previously shown [to be] very immunogenic and that people have antibodies against. That would be an example of a cross-reactive antibody.
And the other is a related dementia that is present only in a very few individuals in the world. It's called familial British dementia. These patients make an amyloid that is different from A-beta, yet we all have antibodies against this British amyloid.
And again, that suggests that it's cross-reactive.
We also show that we can isolate not individual pieces of antibodies, but we can take all the IgG antibodies from plasma and these antibodies can protect neurons against toxic A-beta.
Now whether they do this also in vivo, we can't really prove that, but we can isolate them and show in cell culture they are beneficial.
In the paper, you said that the origin of these beta-amyloids and amyloidogenic peptide-specific antibodies isn't clear. Will it be important to know the origin for developing a therapeutic?
Not necessarily. What is more important is [knowing] which ones are the best in protecting our brains, and once we know that, we can just stimulate them or produce them as monoclonal antibodies in a bioreactor and use them as a vaccine, basically.
[ pagebreak ]
The work described in the paper takes a very therapeutic angle. Is there a diagnostic or prognostic implication, as well?
Not at this point, because we still need to know which types of antibodies correlate with the clinical outcome. Once we know that, then they can be used as a biomarker for efficacy of the vaccine.
But right now, we don't see big differences between Alzheimer's patients and healthy people, so they're probably not going to be very useful for screening people.
Are you pursuing that angle?
What we're trying to pursue is to work with pharma that have clinical trials running and trying to get a hand on their samples so that we could, together with them, find out whether there are antibodies that correlate with clinical benefits. From clinical trials that are ongoing, we would like to get blood samples from patients.
Your research is specifically directed at Alzheimer's but does it have any applications for other neurodegenerative diseases?
We don't really know. There are increasing numbers of studies where people are trying to immunize, for example, Parkinson's mice with a peptide that accumulates in that disease. People have tried it with Huntington's disease. I think that the results are not as clear as they are with Alzheimer's, so at this point, we can't generalize it, but I wouldn't exclude it.
Can you share any ongoing work that you and your colleagues are doing in this area, and any data?
We don't really have any additional data yet, but we're working with one of the co-authors … on a clinical study on monkeys immunized with a beta[-amyloid] and they were then analyzed for cognitive functions with memory tests and all kinds of things.
We're analyzing plasma from these monkeys to see whether we can isolate specific antibodies that correlate with improved memory performance.
The next step is to try to do the same thing in human trials. I would like to look at the samples from this active vaccination trial that I mentioned that was stopped. To look at those samples would be interesting.
And then there [are] a number of Phase 1 studies with new types of vaccines, and again it would be interesting to get a hand on some of the samples from these studies and see whether there are differences in antibodies.