Immunotherapy for Alzheimer disease

Abstract

Immunotherapy approaches for Alzheimer's disease currently are among the leading therapeutic directions for the disease. Active and passive immunotherapy against the β-amyloid peptides that aggregate and accumulate in the brain of those afflicted by the disease have been shown by numerous groups to reduce plaque pathology and improve behavior in transgenic mouse models of the disease. Several ongoing immunotherapy clinical trials for Alzheimer’s disease are in progress. The background and ongoing challenges for these immunological approaches for the treatment of Alzheimer's disease are discussed.

The publication in Nature on a vaccine approach for Alzheimer disease (AD) by Schenk and colleagues in 1999 initiated a push for treatment for this major disease of aging. AD neuropathology is characterized by the progressive loss of synapses and neurons, and the aberrant accumulation in the brain of β-amyloid peptides in plaques and the microtubule associated protein tau in neurofibrillary tangles. Mutations in familial forms of AD have been associated with elevated β-amyloid levels, whereas mutations in tau have been linked to familial forms of frontotemporal dementia. Remarkably, injection of β-amyloid peptides with Freund's adjuvant into transgenic mice harboring a human AD mutation that develop AD-like neuropathology and progressive cognitive decline led to reduced β-amyloid plaque pathology.1 This study was subsequently confirmed and extended by multiple groups to show also behavioral improvement in AD transgenic mice with active β-amyloid immunization.2,3 Passive immunotherapy with antibodies directed at β-amyloid were similarly effective in reducing plaques and improving behavior in AD transgenic mice.4 A temporary setback occurred when the first clinical trial with β-amyloid vaccination was halted after 6% of patients developed an inflammatory reaction in the brain (chemical meningoencephalitis). A subsequent study supported clinical benefits among patients in this active vaccination trial.5 A more recent postmortem study on a subset of patients who had participated in the aborted trial supported active removal of β-amyloid plaques by inflammatory cells, but also indicated that 7 of the 8 patients who were studied at autopsy continued to have progressive cognitive decline despite the removal of amyloid plaques.6

The critical mechanisms whereby active or passive vaccination against β-amyloid can affect the disease process remain uncertain. Recruitment and activation of microglia, the macrophage of the central nervous system, by β-amyloid antibodies is thought to lead to β-amyloid plaque removal. At the same time, fibrillar β-amyloid containing plaques, formerly viewed as the major toxic entities in AD, are increasingly viewed as potentially only pathological remnants of the disease. Smaller assemblies, particularly of two to twelve β-amyloid peptides (oligomers), are considered pathogenic, although the site of pathogenesis remains controversial. Secreted, extracellular β-amyloid oligomers have been shown to damage synapses.7 Some groups stress the aberrant accumulation of β-amyloid within neurons and synapses leading to subsequent extracellular localization following destruction of neurites and synapses.8 Evidence has been presented that antibodies targeting β-amyloid peptides up to 42–43 amino acids can block the toxic effects of extracellular β-amyloid oligomers on synapses.7 Interestingly, β-amyloid immunotherapy was also shown to clear intraneuronal β-amyloid in an AD transgenic mouse; the intraneuronal variety is a pool of β-amyloid that correlates with the onset of cognitive decline prior to plaques and tangles in these mice.9 Intriguingly, antibodies directed at the β-amyloid domain exposed to the extracellular space within the amyloid precursor protein (APP) were shown to be internalized by neurons, where they reduced the intraneuronal pool of β-amyloid and protected against synaptic damage in neurons cultured from AD transgenic mice.10,11 It is possible that inefficient clearance of the intracellular pool of β-amyloid played a role in the continued cognitive decline in the seven of eight patients in the aborted active vaccination clinical trial studied at autopsy who showed clearance of β-amyloid plaques.

Work on β-amyloid immunotherapy in AD contributed to a reevaluation of the role of the immune system in the brain. Previously, it was considered that the brain was immune privileged, and that antibodies entered the brain only with the breakdown of the blood brain barrier. Rare neuroimmunological disorders had suggested more complex interactions. Pathological antibodies directed at neuronal proteins could be found localizing to specific types of neurons in paraneoplastic diseases linked to diverse systemic cancers12,13 or collagen-vascular diseases such as lupus.14 Such pathological antibodies can be directed at synaptic or even intracellular proteins in selective neurons in the brain, leading to localized neurological symptoms. For paraneoplastic diseases it is hypothesized that antibodies directed at the cancer cells cross-react with neuronal antigens. Since titers of antibodies can be higher in brain than in the blood, intrathecal synthesis of antibodies from sequestration of B cells has been proposed to occur in the brain.15 The interaction between the immune system and the brain is therefore viewed as increasingly complex, with antibodies not only gaining access to the brain but also nerve cells, where they can even alter intracellular biology.10 These findings open up new possibilities for antibody-directed therapies for diseases of the nervous system.

Currently, leading concerns for β-amyloid immunotherapy are the potential development of chemical meningoencephalitis and micro-hemorrhages in the brain. Involvement of T cells in damage to the brain vasculature is considered to contribute to these potential side effects. In addition, the β-amyloid released upon antibody-induced removal of plaques may damage blood vessels as β-amyloid is cleared from the brain via the vasculature.16 Recently, a phase 2 Elan/Wyeth study using passive β-amyloid immunotherapy with a humanized monoclonal antibody described (at the 2008 International Conference on Alzheimer's disease) significant benefits in patients not harboring the apolipoprotein E4 (apoE4) allele genetic risk factor for late onset AD. In contrast, no clear therapeutic benefit and more cases with brain inflammation occurred in those with the apoE4 allele linked with an increased risk for AD. Why apoE4 carriers did not benefit in this β-amyloid immunotherapy trial is unknown, but has prompted separation of patients into E4 negative and positive groups in subsequent clinical trials. The less robust than hoped for effects even in the apoE4 negative patients has further dampened expectations. The reason for why the human studies are not showing the protection seen in the transgenic mouse studies could relate to β-amyloid playing less of a role in the more typical late onset AD than it does in the rare autosomal dominant familial forms used to generate the AD transgenic mice. It is also not clear which β-amyloid epitope(s) should be targeted by antibodies to maximize potential benefits while minimizing side effects in AD patients. Optimizing antibody specificity for immunotherapy is further complicating by the varied conformations of different β-amyloid aggregation states. In addition, β-amyloid immunotherapy may be more challenging in patients with AD because it is not effective in reducing tau tangle pathology.6 Most immunotherapy studies were done on transgenic AD mouse models that deposit β-amyloid plaques, but not tau tangles. In the more recently generated triple transgenic AD mouse that develops both plaques and tangles, β-amyloid antibodies reversed β-amyloid pathology and early pre-tangle tau pathology, but not hyperphosphorylated tau aggregates.8 Recent evidence supports that β-amyloid neurotoxicity acts synergistic with tau,17 and that both pathologies begin at synapses.18 Interestingly, tau immunotherapy was reported to protect against tau pathology in transgenic mice harboring mutant tau.19 Thus, dual immunotherapy targeting of both β-amyloid and tau can be considered. Finally, immunotherapy at earlier stages of the disease process may be more effective.

In summary, the β-amyloid vaccine heralded a new era of therapeutic research for AD and despite some setbacks is actively being pursued in several ongoing clinical trials. It continues to be among the leading hopes in the AD research community. Another major effort to specifically block the generation of β-amyloid is also progressing, although not without setbacks along the way. For example, the protease involved in the final cleavage to liberate β-amyloid was found to be involved in multiple other important activities, such as cleavage of Notch. Antibody approaches are also being applied in efforts to block secretase cleavage to generate β-amyloid.20 Finally, there remains some worry that β-amyloid peptides have an as yet unknown normal biological function, although cumulative immunotherapy and other therapeutic studies in animal models have provided sufficient support for the continued pursuit of β-amyloid lowering as a treatment for AD.

Abbreviations

AD	=	Alzheimer disease
APP	=	amyloid precursor protein
apoE4	=	apolipoprotein E4