17 December 2010. Scientists are developing better ways to predict who will develop Alzheimer’s disease (see Part 1 of this series), but earlier diagnosis of the disease will be of limited benefit without viable treatments. In a November 16 press conference held at the 2010 Society for Neuroscience annual meeting in San Diego, California, two researchers described possible ways to tackle the Aβ and tau pathology believed to underlie disease progression. In the first talk, Charlie Glabe, of the University of California in Irvine, and also a consultant to biotechnology company Kinexis, Inc., discussed efforts to develop a safe vaccine against Aβ. In theory, vaccination could harness the immune system to clear Aβ deposits and slow the course of the disease. In mice, vaccination has worked to reduce memory loss and learning difficulties (see ARF related news story). In 2002, however, clinical development of an Aβ vaccine by Elan Corporation was halted because some patients developed brain inflammation due to autoimmune effects (see ARF related news story).

Since then, many newer immunotherapy approaches have entered the clinic. For his part, Glabe and colleagues took a different tack to immunization. They focused on Aβ oligomers, rather than fibrillar or monomeric Aβ, as an antigen. Oligomeric Aβ, believed by many to be the most toxic form of the peptide, assembles itself into β-sheet structures, Glabe said. “When they form β-sheet oligomers, the peptides stick together precisely, just like Lego blocks, and the amino acid sequence is in exact alignment with the peptide above and below in the stack,” Glabe said in the press conference. The side chains of each amino acid stick out perpendicularly to the sheet, giving rise to vertical bars on the surface of the sheet that resembles a product bar code. Since the amino acids are lined up, each bar is unique to a specific amino acid. The immune system recognizes the vertical bars, Glabe said, but not their sequence. “This means that when you vaccinate an animal with a β-sheet oligomer of one sequence, you get an immune response that recognizes all β-sheet oligomers, as long as it has the same bar code present somewhere on the surface of the sheet.” In other words, the immune response is not specific to the peptide sequence.

Glabe and colleagues generated several random, 20-amino-acid peptide sequences that form β-sheet oligomers but are not found in the human genome. They screened for the oligomer (3A) that was best recognized by oligomeric Aβ antibodies. To test the effectiveness of this peptide as an antigen, first author Suhail Rasool immunized 3xTg AD mice with 3A oligomers, Aβ oligomers, islet amyloid polypeptide oligomers, or Aβ fibrils, at ages ranging from three to 14 months. Vaccination with 3A oligomer worked as well as the Aβ forms in preventing plaque formation, lowering total tau and hyperphosphorylated tau, and improving cognitive function. Rasool and colleagues found high levels of oligomer-specific antibodies in 3A-vaccinated mice, and these antibodies recognized Aβ oligomers, but not monomers or fibrils. This demonstrated that a random peptide sequence can generate antibodies against amyloid oligomers.

Because the 3A sequence is not in the human genome, it should not create an autoimmune response. Mice vaccinated with 3A oligomers showed less activated microglia and astrocytes than mice receiving Aβ peptides, Rasool said. Glabe suggested an additional safety feature: Because antibodies to 3A oligomers do not bind to fibrillar plaques, they are less likely to lead to microhemorrhages.

Glabe said he wants to move the 3A peptide toward a human trial once he can secure funding. Since oligomer-specific antibodies cannot distinguish between oligomeric forms of different peptides, these antibodies might be equally effective at clearing other kinds of pathogenic proteins, such as those present in Parkinson’s disease and frontotemporal dementias. Glabe said his group is starting to study this vaccine in Parkinson’s animal models.

Several other types of Aβ vaccine are currently in clinical trials. Elan and Wyeth (now part of Pfizer) are conducting a Phase 2 trial of a second-generation vaccine (see ARF related news story). An anti-Aβ antibody known as bapineuzumab has been shown to reduce amyloid plaques in AD patients (see ARF related news story), and Johnson & Johnson is testing bapineuzumab in numerous clinical trials. Eli Lilly has an anti-Aβ antibody in Phase 3 trials (see ARF related news story). Additionally, Novartis is running a Phase 2 trial, and United Biomedical has a vaccine in Phase 1 trials. Swedish pharmaceutical company BioArctic Neuroscience AB, headquartered in Stockholm, and Eisai Co., Ltd. in Tokyo, Japan, announced in September they will begin a Phase 1 clinical trial of an antibody that recognizes soluble Aβ aggregates (see press release).

With the exception of the BioArctic approach, none of the vaccination strategies currently in clinical trials use oligomeric Aβ as an antigen; rather, they use part of the native structure of APP as an epitope, Glabe said in a telephone interview. This means that the antibodies they raise “don’t necessarily distinguish between the pathologically misfolded form and the native form” of Aβ. This could lead to off-target effects, especially if the native form of the peptide is more abundant. Vaccinating with oligomeric forms targets only the pathological species, Glabe said.

Thomas Wisniewski of New York University in New York City is pursuing a similar strategy using a different antigen. Wisniewski and colleagues immunized APP/PS1 AD mice with oligomeric British amyloidosis related peptide, which has no sequence similarity to normal Aβ and therefore should not produce autoimmune effects. Treated mice showed reduced amyloid load and improved cognition (see Goñi et al., 2010).

In the final talk of the press conference, Ottavio Arancio of Columbia University in New York City showed that oligomers of tau cause memory problems and disrupt synaptic plasticity in mice. This suggests that vaccines targeting tau might also be effective against AD (see ARF related SfN story).—Madolyn Bowman Rogers.

This is Part 2 of a two-part series. See also Part 1.