Sci. Aging Knowl. Environ., 16 November 2005
Vol. 2005, Issue 46, p. pe35
[DOI: 10.1126/sageke.2005.46.pe35]


Immune Shaping and the Development of Alzheimer's Disease Vaccines

Howard J. Federoff, and William J. Bowers

The authors are at the Departments of Neurology (H.J.F. and W.J.B.) and Microbiology and Immunology (H.J.F.) and the Center for Aging and Developmental Biology (H.J.F. and W.J.B.) at the University of Rochester School of Medicine and Dentistry, Rochester, NY 14642, USA. E-mail: howard_federoff{at}

Key Words: Alzheimer's disease • amyloid beta • vaccine • immune shaping • gene transfer • virus vector


Because age is the greatest risk for Alzheimer's disease (AD), it is anticipated that its prevalence will grow over the next four decades to become the number one cause of death in North America (see Trojanowski Perspective). The economic burden attending this scourge will also grow to unprecedented levels. These considerations compel a systematic societal effort to identify the means to attenuate the emergence of and to slow the natural history of AD.

Late-onset AD, the most common form, is not diagnosed preclinically, and at the time of its presentation it is likely that the patient has already suffered extensive synaptic dysfunction and loss. This problem makes evident the need for improved diagnostic methods to detect disease earlier, as well as therapeutic interventions that are synapse sparing. Work in many laboratories is devoted to achieving these important goals (see, for example, Reddy Perspective, Geerts Perspective, Saito Perspective, and "Plaques Aglow").

One of the central pathophysiological features of AD is the excessive accumulation of amyloid beta (A{beta}1-42), a 42-amino acid peptide, in extracellular senile plaques (see "Detangling Alzheimer's Disease"). Multiple strategies are directed against the production of A{beta}, notably small molecule inhibitors of the two enzymes, {beta}- and {gamma}-secretases, which are responsible for liberating the toxic A{beta} peptide from its transmembrane substrate, the amyloid precursor protein (APP). Among its disease-promoting properties, A{beta} induces inflammation, a state that persists throughout the disease course. This chronic inflammation is defined by activated microglia, reactive astrocytosis, complement activation, and production of proinflammatory cytokines (see McGeer Review). Efforts to attenuate the inflammation once AD has been diagnosed have proven unsuccessful, leaving open the possibility that presymptomatic blockade may be required to have an impact on disease emergence.

Vaccination Against AD

The landmark observation that A{beta}-directed vaccination of a mouse model of AD could prevent disease occurrence triggered a wave of interest in understanding the underlying immunological mechanism(s) (Fig. 1A) (1-3). In general, any substance that is capable of generating an immune response is termed an immunogen, whereas an antigen represents a substance that can be bound by antibody. Not all antigens are effective immunogens. In mouse models of AD and in humans, A{beta} peptide is a self-peptide (that is, a peptide produced by the organism itself) that is not sufficiently immunogenic on its own, and thus this type of vaccination requires coadministration with a substance termed an "adjuvant" to boost the immune response. Adjuvants can consist of a mixture of immunogenic substances or could be composed of a single gene product (termed "molecular adjuvant"). Codelivery of A{beta} with a potent adjuvant leads to uptake by antigen-presenting cells (APCs) such as dendritic cells that process the antigen and adjuvant via the major histocompatibility complex II (MHC II) pathway, leading to the expression of proteins (cytokines and chemokines) that promote the proliferation and attraction of T cells. Depending on the profile of cytokines and chemokines that are expressed as a result of APC and T cell engagement, the adaptive immune response can proceed via T cell helper 1 (TH1)- or T helper 2 (TH2)-dependent pathways. Immune responses driven by TH1 T cells [dependent on interferon-{gamma} (IFN-{gamma}) and interleukin-1 (IL-1) cytokines] lead to activation of B cells that express antibodies (humoral response) of immunoglobulin G2b (IgG2b) or a related isotype, but TH1 T cells can also activate cellular responses that lead to activation of A{beta}-specific cytotoxic T lymphocytes (CTLs). Such responses are inherently more inflammatory and are therefore unwanted activities for an AD vaccine. TH2-dependent responses (IL-4 and IL-10 cytokine-driven) are believed to be the safest for derivation of an A{beta}-specific immune response in the setting of an AD-afflicted individual. TH2-dependent pathways drive A{beta}-specific humoral responses that produce antibodies from isotype classes (IgG1 or similar) that are not as likely to participate in inflammatory reactions (see "Immunity Challenge" and "Wollscheid-Lengeling Perspective" for further discussion of the immune system).

View larger version (23K):
[in this window]
[in a new window]
Fig. 1. Schematic representation of A{beta}-based immunotherapeutic approaches for Alzheimer's disease with predicted immune responses and clinical outcomes. (A) The initial A{beta}1-42 peptide-based vaccine approach for AD involved the prefibrillarization of A{beta} and coadministration of a potent adjuvant. Newer modalities have targeted oligomeric A{beta} structures. In AD mouse models, these approaches resulted in (i) engagement of the TH2 arm of the immune response, (ii) elicitation of antibodies to A{beta}, primarily of the IgG1 isotype, (iii) prevention of and/or dissolution of amyloid plaque-like deposits in the brain, and (iv) correlative improvement of learning/memory behaviors. Translation of this vaccine approach to Phase I clinical trials resulted in a subset of immunized patients developing encephalitic symptoms, likely due to the elicitation of TH1-like responses that ultimately led to robust cellular infiltration and inflammatory conditions within the brains of these individuals. (B) A subsequent vaccine approach employed a characterized B cell epitope that encompassed amino acids 1 to 15 of A{beta} in combination with potent TH2 adjuvants. In AD mouse models, this approach led to the production of potent anti-A{beta} humoral responses and behavioral (or functional) improvement. (C) Other studies have used A{beta} in the context of fused molecular adjuvants such as TtxFC to induce skewed TH2-like responses to the A{beta} antigen. (D) More recent iterations of AD vaccines point to the utility of DNA plasmid- and viral-vector-based vaccines, which provide the means to coexpress immunomodulatory cytokines such as IL-4 along with A{beta} epitopes to expand the pool of TH2 T cells that participate in anti-A{beta} immune responses. This approach, in theory, would subsequently lead to the production of humoral responses that are overwhelmingly skewed toward TH2-related antibody isotypes that are believed to be efficacious but inherently less inflammatory. TCR, T cell receptor; TNF-{beta}, tumor necrosis factor-beta; NK cell, natural killer cell.

Two schools of thought about the mode of action of the A{beta} peptide vaccine, each with supporting data, have evolved. The first holds that a humoral response characterized by production of antibodies to A{beta} promotes the egress of the toxic peptide from the brain to the vascular compartment, where it is bound by antibody and eliminated peripherally (4). The second posits that microglia interact with A{beta}-specific antibodies in the brain by recognition of the Fc portion of the antibody, leading to phagocytosis and eventual clearance of A{beta} (5). These are not mutually exclusive mechanisms and may operate in concert after A{beta} vaccination.

On the strength of robust preclinical data, Elan Pharmaceuticals began clinical trials to evaluate A{beta}1-42 peptide vaccination in mild to moderate AD patients. Because adverse events were not observed in an initial Phase I trial, the company initiated a larger Phase IIa trial. During the vaccination of 300 patients, ~5% (18) of subjects developed an aseptic meningoencephalitis (inflammation of the brain and its surrounding membranes in the absence of a pathogenic microorganism); this result compelled termination of the trial [reviewed in (6)]. There was no correlation between elicited A{beta} antibody responses or titers and the development of meningoencephalitis (7, 8). Whereas some patients recovered, others remained compromised. Several such patients have come to postmortem examination during which brain analysis revealed patchy clearance of A{beta} deposits, infiltration of T cells, and evidence of exacerbated inflammation (9).

The adverse events associated with this trial were unexpected and unfortunate. However, analysis of a subset of the Phase IIa vaccinated subjects suggested that those in whom an antibody response included reactivity against senile plaque-associated aggregated A{beta} exhibited an attenuation of disease progression (10). This observation, among others, portends the therapeutic utility of AD vaccination but underscores the associated challenges (see McGeer Perspective). Succinctly, the question becomes: Can vaccines be developed that protect against A{beta} toxicity without promoting or exacerbating underlying brain inflammation?

Development of a Safer Vaccine

Work in several fields impinges on the development of a safer vaccine strategy: (i) determining the types of immune responses that are elicited in response to different forms of A{beta} vaccination, (ii) developing animal models that are reflective of human AD to test vaccine responses, and (iii) understanding the role of immune shaping in these responses. We discuss each of these issues in more detail below.

Mode of antigen delivery

Most AD vaccine studies, including the human clinical trials, have employed peptide-based means for immunization. This approach involves prior formation of higher order A{beta} aggregates (termed "fibrils") and codelivery with a potent adjuvant. The use of preformed fibrillar A{beta} is believed to better mimic, in the immunogenic sense, the pathogenic form of A{beta} found in plaques in AD brain than does the use of soluble monomeric A{beta}. Resultant immune responses from peptide-based vaccination can be somewhat modulated through the use of different segments (epitopes) of A{beta} and by varying adjuvants. Newer vaccine approaches have sought to exploit gene transfer to engender more controlled immune responses. Genetic sequences encoding A{beta} and molecular adjuvants (DNA sequences that code for immunostimulatory gene products) can be delivered to APCs by two broad classes of vectors: viral and nonviral. Viral vectors exploit the evolutionary achievements of viruses to propagate, package, and transfer genetic material from cell to cell and from organism to organism. Through genetic modification of mammalian viruses, it is possible to specifically target gene expression in desired cellular populations. Adeno-associated virus (AAV) and herpes simplex virus (HSV)-based vectors have been tested as immunotherapeutics in mouse models of AD (described in more detail below). AAV vectors exhibit above-average safety profiles and express transgenes essentially lifelong but have a small transgene size capacity (<4.5 kb). HSV vectors, especially those of the amplicon type (from which much of the viral genome has been removed), have been shown to efficiently transduce dendritic cells, induce strong transgene-specific immune responses, and carry transgene segments up to 130 kb in size. Nonviral or plasmid DNA-based vaccination are less efficient in delivering the gene encoding an A{beta} antigen in vivo but could be employed to prime the adaptive immune response when combined with another of the vaccination modalities.

Animal models

A{beta} vaccination has been extensively evaluated in mice, both in models of AD (transgenic animals that produce variant forms of APP) and in nontransgenic strains. In nontransgenic mice in which A{beta} is a foreign protein, vaccination can elicit different immune responses that are dependent on (i) adjuvant composition, (ii) route of vaccine administration (e.g., intradermal, intraperitoneal, subcutaneous, intranasal, or intramuscular), and (iii) mode of antigen delivery. In addition, the genetic background of the mouse strain also affects the nature of the immune response to vaccination. In only one case out of nearly 100 reports did investigators observe encephalitis in C57BL/6 mice. This occurred when the mice were vaccinated with A{beta} peptide and the adjuvant pertussis toxin, which is known to promote autoimmunity (11). This vaccination paradigm may thus serve as an informative mouse model in which to study A{beta}-vaccine-mediated brain inflammation and encephalitis.

Numerous investigators have evaluated A{beta} vaccine efficacy in mouse models of AD. All such models harbor and express a familial AD APP mutant transgene that results in the accumulation of excess human amyloid in the brain. Most strains do not recapitulate the varied pathologies, such as chronic inflammation, seen in the human AD brain. In these mice, the transgenically expressed human A{beta} is a self-peptide to which tolerance must be broken to elicit an immune response. Immune tolerance refers to the failure of organisms, including humans, to make an immune response to an antigen to which they have been exposed early in life. A{beta}-based vaccine assessments using a recently developed AD mouse model that harbors genes encoding variant forms of human APP, presenilin 1 (a component of the {gamma}-secretase complex), and the microtubule-binding protein tau [3xTg-AD (12)], have yet to be reported.

Multiple vaccination approaches ranging from peptide only to DNA and virus or viruslike particles have proven effective in breaking tolerance to transgenic A{beta} and conferring protection against and/or reduction of accumulated brain amyloid [reviewed in (13)]. In one instance, the use of an HSV amplicon A{beta} vaccine resulted in a marked increase in brain inflammation in the AD model Tg2576 (14). This robust up-regulation of a number of proinflammatory cytokines occurred after several vaccinations. More relevant to vaccine safety was the mortality of more than 50% of the herpes amplicon A{beta}-vaccinated Tg2576 mice. Given the limited number of observations demonstrating A{beta} vaccine-elicited inflammatory effects in AD mouse models, it seems imperative that AD immunotherapeutics be assessed in models that more closely mimic the immune system of an elderly individual. In this regard, aged nonhuman primates may represent more optimal candidates for assessing vaccine safety profiles relating to untoward brain inflammatory responses.

Immune shaping

Immune shaping has become a goal of vaccine strategies. As discussed briefly above, production of an immune response is characterized by a blending of two types of T cell responses: TH1 and TH2. The intent is to direct a skewed response toward one or the other. TH1 responses are associated with the production of the cytokines IFN-{gamma} and IL-1 and a cell-mediated immune response. TH2 responses are accompanied by a different cytokine profile, usually by increases in the concentrations of IL-4 and IL-10. This latter profile favors a humoral immune response. In the setting of neuroinflammation, it would be desirable to promote a TH2-biased response whereby an advantageous IgG1 antibody response is elicited. Several issues are pertinent to this strategy. Selection of the A{beta} epitopes that strongly elicit a B cell response, for example, one encompassing amino acids 1 to 15, may enhance humoral efficacy (Fig. 1B) (15). Similarly, fusion of A{beta} sequences to chimeric peptides that engender strong T cell help can also augment the elicitation of a robust antibody response. Examples of the latter are the "C" fragment of tetanus toxin (TtxFC) (14) and the synthetic Pro-Ala-Asp-Arg-Glu sequence (Fig. 1C) (15). In one study, the A{beta} peptide was fused to the "C" fragment (A{beta}/TtxFC) and delivered with a herpes amplicon platform in the Tg2576 AD mouse model (14). In contrast to the morbidity and mortality seen with the A{beta} herpes vaccine described above, vaccination with the herpes amplicon expressing A{beta}-TtxFC chimera elicited no adverse outcomes, induced the production of A{beta}-specific IgG1 isotypic antibodies, and significantly attenuated A{beta} accumulation, a measure of vaccine efficacy. These studies highlight the utility of the vaccine platform, such as the herpes amplicon, that affords experimental control of antigen type, levels of expression, modulation with codelivered immune-modulating proteins, and efficient transfer to antigen-presenting dendritic cells (Fig. 1D).

Previous studies have shown that A{beta} peptides undergo spontaneous conversion in vitro from soluble monomeric forms to less well-characterized intermediates, referred to as A{beta}-derived diffusible ligands (ADDLs) (16), to protofibrillar forms, and finally to insoluble fibril conformations. Midlevel oligomeric A{beta} forms or ADDLs alter the normal physiology of neuronal cells, often leading to neuronal dysfunction and death (17) (see, for example, Lashuel Perspective). Delivery of mutant A{beta} forms known to more rapidly adopt oligomeric structures and thus achieve the pathogenic protein conformation may result in enhanced vaccine efficacy. Antibodies can be generated specifically against soluble oligomeric A{beta} (18). These polyclonal antibodies not only recognize and neutralize the toxicity of ADDLs but also neutralize soluble oligomers of prion protein and {alpha}-synuclein, indicating the existence of a shared structural epitope(s) among these pathogenic proteins. Antigenic conformation is also crucial for the elicitation of TH1-mediated responses, because iteratively modifying the structure of a vaccine antigen has been demonstrated to influence resultant T cell responses significantly (19).

Characterization of patient population

Evolution of a safe AD vaccine should include efforts to characterize the patient population to be vaccinated. Aged individuals exhibit progressive alterations in humoral and cellular immunity that lead to reduced vaccine responses and increased generation of auto-antibodies. In fact, recent data indicate that a fraction of AD patients harbor A{beta}-reactive T cells and systemic markers consistent with inflammation (20). Moreover, deficiencies in immunoglobulin class switching have also been demonstrated in the elderly, which results in diminished maturation of vaccine-elicited humoral responses to secondary isotypes [reviewed in (21)]. The question about implementation of an AD vaccine then becomes: Does the intended cohort have AD, mild cognitive impairment, central nervous system or systemic inflammation, and/or age-related immunological alterations? As some of these variables can be evaluated, their inclusion in future AD trial design may be warranted. Although immunological profiling may prove impractical in the context of a large-scale clinical trial, such an approach could possibly be streamlined for larger study designs.


The graying of our population and the growth of AD prevalence underscore the vital nature of discovering disease-modifying therapeutics for this scourge. The potential for vaccination of at-risk individuals remains a promising approach, but concerns of exacerbating underlying brain inflammation and aggravating disease progression compel the derivation of safer vaccines. Through the use of vaccine platforms that support immune shaping, it is possible to identify optimal antigens to elicit the most protective responses. Near-term research must focus on the systematic evaluation of parameters and testing in animal models of AD where brain inflammation is present. The successful maturation of these vaccines from preclinical studies to clinical trials warrants examination in the context of differences in patient immune status to A{beta} and the propensity for worsening brain inflammation. Adherence to such a stepwise and deliberate process may portend the arrival of a truly new therapeutic.

November 16, 2005
  1. C. Janus, J. Pearson, J. McLaurin, P. M. Mathews, Y. Jiang, S. D. Schmidt, M. A. Chishti, P. Horne, D. Heslin, J. French et al., A beta peptide immunization reduces behavioural impairment and plaques in a model of Alzheimer's disease. Nature 408, 979-982. (2000).[CrossRef][Medline]
  2. D. Morgan, D. M. Diamond, P. E. Gottschall, K. E. Ugen, C. Dickey, J. Hardy, K. Duff, P. Jantzen, G. DiCarlo, D. Wilcock et al., A beta peptide vaccination prevents memory loss in an animal model of Alzheimer's disease. Nature 408, 982-985 (2000).[CrossRef][Medline]
  3. D. Schenk, R. Barbour, W. Dunn, G. Gordon, H. Grajeda, T. Guido, K. Hu, J. Huang, K. Johnson-Wood, K. Khan et al., Immunization with amyloid-beta attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature 400, 173-177. (1999).[CrossRef][Medline]
  4. R. B. DeMattos, K. R. Bales, D. J. Cummins, J. C. Dodart, S. M. Paul, D. M. Holtzman, Peripheral anti-A beta antibody alters CNS and plasma A beta clearance and decreases brain A beta burden in a mouse model of Alzheimer's disease. Proc. Natl. Acad. Sci. U.S.A. 98, 8850-8855 (2001).[Abstract/Free Full Text]
  5. F. Bard, C. Cannon, R. Barbour, R. L. Burke, D. Games, H. Grajeda, T. Guido, K. Hu, J. Huang, K. Johnson-Wood et al., Peripherally administered antibodies against amyloid beta-peptide enter the central nervous system and reduce pathology in a mouse model of Alzheimer disease. Nat. Med. 6, 916-919. (2000).[CrossRef][Medline]
  6. B. P. Imbimbo, Toxicity of beta-amyloid vaccination in patients with Alzheimer's disease. Ann. Neurol. 51, 794 (2002).[Medline]
  7. J. M. Orgogozo, S. Gilman, J. F. Dartigues, B. Laurent, M. Puel, L. C. Kirby, P. Jouanny, B. Dubois, L. Eisner, S. Flitman et al., Subacute meningoencephalitis in a subset of patients with AD after Abeta42 immunization. Neurology 61, 46-54 (2003).[Abstract/Free Full Text]
  8. C. Hock, U. Konietzko, A. Papassotiropoulos, A. Wollmer, J. Streffer, R. C. von Rotz, G. Davey, E. Moritz, R. M. Nitsch, Generation of antibodies specific for beta-amyloid by vaccination of patients with Alzheimer disease. Nat. Med. 8, 1270-1275 (2002).[CrossRef][Medline]
  9. J. A. Nicoll, D. Wilkinson, C. Holmes, P. Steart, H. Markham, R. O. Weller, Neuropathology of human Alzheimer disease after immunization with amyloid-beta peptide: A case report. Nat. Med. 9, 448-452 (2003).[CrossRef][Medline]
  10. C. Hock, U. Konietzko, J. R. Streffer, J. Tracy, A. Signorell, B. Muller-Tillmanns, U. Lemke, K. Henke, E. Moritz, E. Garcia et al., Antibodies against beta-amyloid slow cognitive decline in Alzheimer's disease. Neuron 38, 547-554 (2003).[CrossRef][Medline]
  11. R. Furlan, E. Brambilla, F. Sanvito, L. Roccatagliata, S. Olivieri, A. Bergami, S. Pluchino, A. Uccelli, G. Comi, G. Martino, Vaccination with amyloid-beta peptide induces autoimmune encephalomyelitis in C57/BL6 mice. Brain 126, 285-291 (2003).[Abstract/Free Full Text]
  12. S. Oddo, A. Caccamo, J. D. Shepherd, M. P. Murphy, T. E. Golde, R. Kayed, R. Metherate, M. P. Mattson, Y. Akbari, F. M. LaFerla, Triple-transgenic model of Alzheimer's disease with plaques and tangles: Intracellular Abeta and synaptic dysfunction. Neuron 39, 409-421 (2003).[CrossRef][Medline]
  13. D. Schenk, M. Hagen, P. Seubert, Current progress in beta-amyloid immunotherapy. Curr. Opin. Immunol. 16, 599-606 (2004).[CrossRef][Medline]
  14. W. J. Bowers, M. A. Mastrangelo, H. A. Stanley, A. E. Casey, L. J. Milo Jr., H. J. Federoff, HSV amplicon-mediated Abeta vaccination in Tg2576 mice: Differential antigen-specific immune responses. Neurobiol. Aging 26, 393-407 (2005).[CrossRef][Medline]
  15. M. G. Agadjanyan, A. Ghochikyan, I. Petrushina, V. Vasilevko, N. Movsesyan, M. Mkrtichyan, T. Saing, D. H. Cribbs, Prototype Alzheimer's disease vaccine using the immunodominant B cell epitope from beta-amyloid and promiscuous T cell epitope pan HLA DR-binding peptide. J. Immunol. 174, 1580-1586 (2005).[Abstract/Free Full Text]
  16. S. G. Younkin, The role of A beta 42 in Alzheimer's disease. J. Physiol. Paris 92, 289-292 (1998).[CrossRef][Medline]
  17. J. P. Cleary, D. M. Walsh, J. J. Hofmeister, G. M. Shankar, M. A. Kuskowski, D. J. Selkoe, K. H. Ashe, Natural oligomers of the amyloid-beta protein specifically disrupt cognitive function. Nat. Neurosci. 8, 79-84 (2005).[CrossRef][Medline]
  18. R. Kayed, E. Head, J. L. Thompson, T. M. McIntire, S. C. Milton, C. W. Cotman, C. G. Glabe, Common structure of soluble amyloid oligomers implies common mechanism of pathogenesis. Science 300, 486-489 (2003).[Abstract/Free Full Text]
  19. W. Paster, M. Kalat, M. Zehetner, T. Schweighoffer, Structural elements of a protein antigen determine immunogenicity of the embedded MHC class I-restricted T cell epitope. J. Immunol. 169, 2937-2946 (2002).[Abstract/Free Full Text]
  20. A. Monsonego, V. Zota, A. Karni, J. I. Krieger, A. Bar-Or, G. Bitan, A. E. Budson, R. Sperling, D. J. Selkoe, H. L. Weiner, Increased T cell reactivity to amyloid beta protein in older humans and patients with Alzheimer disease. J. Clin. Invest. 112, 415-422 (2003).[CrossRef][Medline]
  21. D. Frasca, R. L. Riley, B. B. Blomberg, Effect of age on the immunoglobulin class switch. Crit. Rev. Immunol. 24, 297-320 (2004).[CrossRef][Medline]
Citation: H. J. Federoff, W. J. Bowers, Immune Shaping and the Development of Alzheimer's Disease Vaccines. Sci. Aging Knowl. Environ. 2005 (46), pe35 (2005).

Science of Aging Knowledge Environment. ISSN 1539-6150