Out-of-control T cells are blamed for the encephalitis that doomed Elan’s Aβ vaccine trial for Alzheimer disease (AD), and so researchers have turned their efforts toward immunogens that stimulate antibody-producing B cells instead. But T cells might be useful allies in modulating the immune response for the better, according to new work from the lab of Michal Schwartz at the Weizmann Institute of Science (Rehovot, Israel). In a paper in press in PNAS online, Schwartz and colleagues show that immunization of AD mice with the T cell stimulator and multiple sclerosis drug glatiramer acetate (GA) induces plaque clearance, normalization of hippocampal neurogenesis and improved memory and learning.

This is not the first time GA has been shown to be effective in mouse models of AD—Howard Weiner and colleagues demonstrated last year that nasal immunization stimulated a plaque-clearing microglial response (Frenkel et al., 2005). But what the new work reveals is that the activated microglia in immunized animals appear to be “good” microglia, expressing the neurotrophic cytokine insulin-like growth factor-1, rather than tissue-damaging TNFα. By fine-tuning the immune response to endogenous Aβ, the researchers claim, GA promotes a non-toxic, beneficial immune response. They use their data to argue against a role for anti-inflammatory medications in treating AD, instead preferring immune modulators like GA.

But what of selective anti-inflammatory medicines? “Bad”’ microglia flood the brain with neurotoxic TNFα, the very factor targeted in a different and novel therapeutic approach used by Edward Tobinick and colleagues at the University of California in Los Angeles. Earlier this year, they published results of a 6-month pilot trial of the TNFα antagonist etanercept in 15 patients with moderate to severe AD. Etanercept has been a hugely successful treatment for arthritis, and they showed that a once-a-week perispinal injection of the drug significantly improved mental function in the subjects. There results appear in the online journal Medscape General Medicine.

The GA work from Schwartz follows up on observations they made on the activation of microglia by Aβ peptides. Previously, first author Oleg Butovsky showed that in vitro activation of microglia with aggregated Aβ1-14 resulted in cells that produce TNFα but not the neurotrophic cytokine IGF-1. But when they provided the microglia with interleukin-4 (IL-4), the T cell cytokine inhibited the production of TNFα and stimulated IGF-1, a phenotype switch that promoted neuron survival (Butovsky et al., 2005). In the current study, they show that stimulating microglia with Aβ plus IL-4 promotes neurogenesis from mouse adult neural progenitor cells in vitro, consistent with their idea that IL-4 converts Aβ-reactive microglia into neuron nurturers.

IL-4 is a product of T helper type 2 cells (Th2), which led Butovsky and coworkers straight to GA, an immune modulator that has been on the market to treat MS for 10 years. GA is a copolymer of glutamic acid, lysine, alanine and tyrosine (hence the name, GLATiramer) which is thought to mimic a myelin peptide. Even now, the mechanism of GA action is considered unclear, but one theory is that it shifts the immune response away from damaging autoimmunity to a beneficial Th2 response.

In agreement with Weiner’s previous work, Butovsky and colleagues show that immunization of 8-month-old, double transgenic AD mice (APP/PS1, see Borchelt et al., 1997) with subcutaneous GA resulted in many fewer plaques after 7 weeks. The vaccinated mice also had higher levels of hippocampal neurogenesis compared to untreated transgenics and showed better spatial learning and memory in the Morris water maze. In these parameters, the vaccinated mice were very similar to non-transgenic control mice.

Taking a closer look at the immune response in these mice, the investigators found that in untreated mice, plaques were associated with abundant activated (CD11b+) microglia, some of which expressed TNFα. In vaccinated mice, they found far fewer CD11b+ microglia per area, presumably due to the lower plaque burden. The microglia they did find, in contrast to the unvaccinated state, expressed the dendritic cell markers MHCII and CD11c, suggesting they could present antigen. The MHCII+/CD11c+ microglia also expressed IGF-1. In addition, significantly more T cells were associated with plaques in immunized mice than in non-immunized animals, and some T cells appeared to be directly contacting microglia. To ask whether T cell-generated IL-4 could be driving the microglial phenotype, the researchers tested their microglial cultures for CD11c, and found that indeed, IL-4 upregulated this marker, even in microglial cells that had been pretreated with Aβ.

The results “argue in favor of the use of a myelin-related antigen such as GA, but not an Aβ peptide, as a T cell-based therapy for AD,” the authors conclude. T cells activated by weak self-antigens go on to supply cytokines and growth factors to promote the dendritic phenotype in microglial cells, which they conclude has a protective role in the brain. This type of response might be of use in a number of neurodegenerative diseases. Indeed, in addition to its widespread use in MS, GA is currently in clinical development for ALS (Gordon et al., 2006) and has shown some promise in an animal model of Parkinson disease (see ARF related news story).

Schwartz and colleagues wrap up with the provocative statement: “Our results strongly argue against the need for anti-inflammatory treatment for patients with AD. On the contrary, we propose that in fighting off AD, as in combating any other neurodegenerative disease, immune activation, rather than immune suppression, is required.”

But what of anti-inflammatory medications that prevent TNFα from damaging the brain? Drugs that selectively shut down TNFα-mediated inflammation have been spectacularly successful in treating arthritis and other inflammatory diseases. Now, one of those TNF antagonists, etanercept, is being tested for AD, and the data from Tobinick (who holds patents that claim the use of TNFα inhibitors, including etanercept, to treat Alzheimer disease) and colleagues look quite promising. Etanercept is a TNF receptor fusion protein that blocks TNF binding to cellular receptors. Once-a-week treatment with the drug over 6 months resulted in significant improvement in mini-mental state scores, ADAS-Cog, and the severe impairment battery (SIB). One patient who began the study severely impaired (MMSE score of 0) showed an improvement of 4 points on the MMSE, and 35 points in the SIB.

The complicated part of this study was the delivery—the drug was injected weekly by perispinal injection in the back of the neck. The authors speculate this gave better CNS delivery, but that remains to be proven. The trial was limited by its design, with no placebo and a small number of patients, but the results clearly warrant additional study.—Pat McCaffrey

See also:

Tobinick E, Gross H, Weinberger A, Cohen H. TNF-alpha modulation for treatment of Alzheimer’s disease: A 6-month Pilot Study. Medscape General Medicine. 2006; 8:25. Posted 4/26/2006. http://www.medscape.com/viewarticle/529176 (Requires registration)


  1. Butovsky and colleagues have reported that “Glatiramer acetate fights against Alzheimer’s disease by inducing dendritic-like microglia expressing insulin-like growth factor 1.” The authors have not shown that glatiramer fights against AD, per se. They do not know whether it will help, harm or be without benefit, because they have not administered it to AD patients. What the authors have done is to administer 5 subcutaneous doses of glatiramer to doubly transgenic APP/PS1 mice and have shown, compared with untreated littermates, less amyloid deposition and less impairment in water maze testing. Their results are comparable to the earlier findings of Frenkel et al. (2006), who administered glatiramer intranasally rather than subcutaneously to transgenic mice. Glatiramer is a mixture of synthetic polypeptides which is currently in use to treat multiple sclerosis. Its mechanism of action is still unclear.

    The theory of Butovsky et al. is that the vaccination caused a phenotypic shift in microglial expression from production of the complement receptor CD11b to CD11b/CD11c, resulting in improved phagocytosis and increased neurogenesis in the transgenic mice. What needs to be emphasized is that transgenic mouse models of AD are not AD itself, and to assume that they are, especially with respect to engaging the adaptive immune system through vaccination, can have severe consequences. This was the case with Elan’s clinical trial for an Aβ vaccine where immune stimulation induced sterile meningitis and cerebral damage in about 5 percent of the cases despite spectacular results in transgenic mice.

    There are notable differences in the pathology of AD and transgenic mouse models. For example, in AD there is further processing of the Aβ deposits, converting them into a more insoluble state. In humans there is a higher level of inflammation, caused in large part by vigorous activation of the human complement system by Aβ deposits. Since mouse C1q poorly recognizes human Aβ deposits, complement activation in transgenic mice is minimal. In human AD, there is full activation of the complement system resulting in neuronal destruction by the membrane attack complex. The latter may be the most problematical consequence of immune stimulation in AD.

    Butovsky et al. concluded that anti-inflammatory therapy should not be used in AD, and that appropriate immune stimulation should be an effective treatment. If this theory were correct, then individuals taking anti-inflammatory therapy should have a higher risk of developing AD. The opposite is the case. We reviewed 17 epidemiological studies from nine different countries in 1996 (McGeer et al., 1996). All but two showed decreased odds of contracting AD amongst those suffering from arthritis or known to be taking anti-inflammatory drugs. We updated the review in 2006 (McGeer and McGeer, 2006), specifically concentrating on NSAIDs since these are the most widely used anti-inflammatory agents. Twelve of 14 studies showed a decreased risk of developing AD. In addition, eight of eight transgenic animal studies showed a reduction in both Aβ deposits and behavioral deterioration in mice given traditional NSAIDs.

    Butovsky et al. noted that their theory “is in line with studies showing that anti-inflammatory drugs, such as cyclooxygenase 2 inhibitors, do not benefit AD.” This is certainly true, since four clinical trials of selective COX-2 inhibitors have failed. But COX-2 is a questionable target for the brain. It is one of the few organs of the body which constitutively expresses this enzyme, which is most highly concentrated in pyramidal neurons. Presumably, there is a significant physiological function associated with this high level of expression, and blocking prostaglandin production in pyramidal neurons could have negative consequences. Moreover, COX-2 inhibitors have been too recently introduced for any epidemiological evidence to have accumulated showing whether their long-term use increases or reduces the risk of developing AD. However, COX-2 inhibitors have been tried without benefit in transgenic animal studies (see Kukar et al., 2005).

    It is not beyond the realm of possibility that ways can be found in humans of stimulating microglia to phagocytose while blunting the self-destruction they cause by excessive output of oxygen free radicals, prostaglandins, inflammatory cytokines, proteases, complement proteins, and other toxic materials. But whether or how this might be done is still unknown. Butovsky et al. have suggested a possibility which certainly deserves further exploration. We can hope they have set investigators on a promising trail, but direct application of their theory to AD cases should be approached with caution.


    . Arthritis and anti-inflammatory agents as possible protective factors for Alzheimer's disease: a review of 17 epidemiologic studies. Neurology. 1996 Aug;47(2):425-32. PubMed.

    . NSAIDs and Alzheimer disease: epidemiological, animal model and clinical studies. Neurobiol Aging. 2007 May;28(5):639-47. Epub 2006 May 11 PubMed.

    . Diverse compounds mimic Alzheimer disease-causing mutations by augmenting Abeta42 production. Nat Med. 2005 May;11(5):545-50. PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Parkinson's Omnibus—Vaccination, New Rat Model, and More

Paper Citations

  1. . Nasal vaccination with a proteosome-based adjuvant and glatiramer acetate clears beta-amyloid in a mouse model of Alzheimer disease. J Clin Invest. 2005 Sep;115(9):2423-33. PubMed.
  2. . Activation of microglia by aggregated beta-amyloid or lipopolysaccharide impairs MHC-II expression and renders them cytotoxic whereas IFN-gamma and IL-4 render them protective. Mol Cell Neurosci. 2005 Jul;29(3):381-93. PubMed.
  3. . Accelerated amyloid deposition in the brains of transgenic mice coexpressing mutant presenilin 1 and amyloid precursor proteins. Neuron. 1997 Oct;19(4):939-45. PubMed.
  4. . Randomized controlled phase II trial of glatiramer acetate in ALS. Neurology. 2006 Apr 11;66(7):1117-9. PubMed.

Further Reading

Primary Papers

  1. . Glatiramer acetate fights against Alzheimer's disease by inducing dendritic-like microglia expressing insulin-like growth factor 1. Proc Natl Acad Sci U S A. 2006 Aug 1;103(31):11784-9. PubMed.