See ARF Live Discussion on Alzheimer Immunotherapy

Howard Weiner, Harvard Medical School, discussed published data (Monsonego et al. 2001) as well as additional data on the immune response to active immunization with Aβ. The paper suggested that AβPP-transgenic mice have a smaller antibody and cellular response to Aβ than nontransgenic mice. He mentioned that in active immunization paradigms, there are different responses to Aβ depending on mouse genetic background. He brought up the question whether in sporadic AD, is it possible that the natural T cell response could influence susceptibility to AD? He presented some data to suggest that some older individuals have a greater T-cell response to Aβ in vitro than do young people.

Ken Ugen, University of South Florida, Tampa, discussed work published previously from the laboratory of David Morgan and colleagues (Morgan et al., 2000) regarding effects of active immunization on improvement in memory-related behaviors. In their newer experiments of actively immunized "old" AβPP-transgenic mice (16.5 to 18 months of age), there was no improvement of memory in a water maze task in actively immunized versus control-treated mice. In their work, they find that "activation of microglia" is associated with reduced amyloid load. This is contrary to what some other researchers find, namely that there are fewer "activated microglia" with decreased Aβ associated with active immunization.

Dale Schenk, Elan Pharmaceuticals. mentioned in regard to the recent discontinuation of the phase II human trials on active immunization with Aβ that meningoencephalitis was rare but evident in some individuals. They do not yet know why it happened. In most patients with the side effect, it occurred after the second immunization but in some after the first or third immunizations. Schenk stated there was no correlation between anti-Aβ antibody titer and the side effect, and that some patients with the side effect had developed no anti-Aβ antibodies. This suggests the side effect is due to a T-cell mediated process. He also said that different immunization approaches (different active and passive immunization) against Aβ still have great potential to treat AD.

Beka Solomon, Tel Aviv University, presented data that certain anti-Ab antibodies are able to dissolve fibrillar Aβ in vitro. Newer data she presented related to their group's recent use of filamentous phage to express the Aβ epitope EFRH (Frenkel et al. 2001). They are using this to develop a peripheral immune response to Aβ in animals, as well as to see if that has an effect on pathology in AβPP-transgenic mice. Solomon has developed an ScFv antibody fragment to Aβ (25 kDa) and was expressing this with filamentous phage. She is trying to inject phage into the olfactory system as a "gate" into the brain. She showed some data suggesting that filamentous phage can cross the BBB via this route. She is trying this approach with the ScFv fragment to see if this will label plaques and be useful as a diagnostic tool to detect plaques ultimately in vivo.

Dennis Selkoe, Harvard Medical School, discussed the active immunization approach via the nose. He discussed data presented in the paper from his group published in Annals of Neurology (Weiner et al. 2000). Nasal immunization with Aβ in PDAPP mice resulted in significantly decreased Aβ deposition in brain in some groups of nasally immunized mice. This effect correlated with a decreased intrinsic brain inflammatory response (decreased microglia, decreased dystrophic neurites).

Einar Sigurdsson, New York University, presented data published in 2001 (see ARF news story). His group is using active immunization with soluble forms of Aβ1-30 modified with N-terminal lysines (K6Ab30). One reason they are doing this is because they feel that Aβ1-42 has the potential to get into the brain and cause toxicity. He hypothesized that the effects of their immunization approach may be working via increasing clearance of Aβ into the blood.

Frederique Bard, Elan Pharmaceuticals, reviewed data shown in previous work (Bard et al., 2000) on effects of passive immunization. She also presented new data showing that of N-terminal anti-Aβ antibodies that reduce plaque, they find that IgG2A subtypes seem to be more effective. In more recent passive immunization experiments, they gave different doses of monoclonal antibodies at 10 mg/kg intraperitoneally to PDAPP mice, starting at 12 months until 18 months of age. Of antibodies that reduced plaque (e.g., 6C5, 10D5, 2C1 and 12B4), 2C1 and 12B4 seemed more effective and were IgG2A. While significant effects were seen, Bard noted that plaque load varies greatly in these mice and that there is a lot of overlap between PBS-treated controls and antibody-treated animals. They also performed a new "active" immunization approach with different Aβ epitopes: 1-5, 3-9, 5-11, 6-12, 15-24, and 5-1 (control). Bard said that most animals developed good antibody titers to most of these epitopes. She also said epitope 1-5-injected mice had the best anti-plaque effects and 3-9 had some effects.

Brad Hyman, Harvard Medical School, presented beautiful images demonstrating the use of multiphoton microscopy in living PDAPP mice to image plaques, as well as other features of brain anatomy (see ARF news story). His group can follow individual plaques and CAA over time in living PDAPP mice. He showed data that not only full-length anti-Aβ antibodies but all Fab fragments result in local plaque clearing. Other new data presented showed the ability to label neurites in vivo with a fluorescently labeled dextran red (see ARF news story) and to examine the relationship of neurites and Aβ-related pathology in vivo. In addition, their group is trying to label free radicals generated in vivo with different fluorescent dyes, such as Amplex red and DCF. Data suggested with these dyes that free radicals were being generated associated with plaques.

David Peretz, University of California, San Francisco, presented data published last year that certain anti-PrPc monoclonal antibodies can clear PrPSc from scrapie-infected neuroblastoma cells in culture (see ARF news story). Interestingly, antibodies to certain epitopes work better than others, and there is no evidence that the effective antibodies ever bind to PrPsc (only the normal cellular prion protein-PrPc). The effective antibodies that he has tested so far in vitro on cells do not have synergistic effects when used together.

Cynthia Lemere, Harvard Medical School, summarized her work on active nasal immunization with Aβ. The group has optimized their protocol with the use of E.coli heat-labile enterotoxin (LT). LT markedly increased the antibody response to nasally delivered Aβ in mice. A non-toxic form, LT(R192G), was an even better adjuvant. Most antibodies generated to Aβ in their experiments saw epitopes between 1-15. No obvious side effects were seen in the mice.

In recent experiments with PSAPP mice, the animals were given both systemic active immunization followed by nasal immunization. In recent experiments with PSAPP mice, the animals were given single-dose systemic Aβ immunization followed by repetitive nasal Aβ immunization. Animals developed high titers of anti-Aβ antibodies. Immunized mice showed a 75 percent decrease in plaque number and a decrease in gliosis and neuritic dystrophy. Lemere et al. found that serum Aβ, measured by ELISA, was increased 33-fold in immunized versus non-immunized mice (53 ng/ml immunized versus 1.6 ng/ml in non-immunized) and that most of the Aβ in serum was complexed to antibodies.

David Holtzman, Washington University, St. Louis, Missouri, presented data published in last year by DeMattos et al. (see ARF news story) demonstrating that administration of the anti-Aβ antibody 266 results in a rapid, massive increase in plasma Aβ. Data gathered in collaboration with Kelly Bales, Jean Dodart, and Steven Paul showed that following administration of 266 in PDAPP mice, positive effects on certain memory tasks can be detected within 1-3 days after administration of the antibody. New data was also presented showing that measurement of plasma Aβ following administration of 266 was highly correlated with and predictive of amyloid load in the brain of PDAPP mice.

David Westaway, University of Toronto, presented published data (Janus et al., 2000), as well as newer behavioral data in their TgCRND8 transgenic mice, which produced very high Aβ levels. Data showed that the mice are behaviorally abnormal in the Morris water maze from five weeks of age on (well before plaque deposition).

Interestingly, when these Mice are repeatedly tested as they get older, their behavior improves to control levels. Then, at about 30-40 weeks of age, they begin to perform less well than control mice again (now they have lots of plaques). When mice were actively immunized against Aβ from 45-59 weeks, transgenic mouse behavior improved, and they almost closed the gap compared to the control mice again.