It seems increasingly likely that α-synuclein is central to the pathogenesis of Parkinson disease and diffuse Lewy body disease. Therefore, strategies like the one outlined in this paper to decrease α-synuclein protein levels may be viable options for therapeutics in the future. Other labs have reported different approaches to the same problem, such as recombinant intrabodies (Emadi et al., 2004), anti-aggregating peptides (El-Agnaf et al., 2004), or overexpression of β-synuclein (Hashimoto et al., 2004). Where the current paper is especially significant is that the major species decreased by this immunization strategy is aggregated forms of the protein. Although this is very hard to quantify, in figure 5 the amount of monomer is left relatively unchanged where oligomers are decreased. This observation provides us with a way to test the idea that relatively soluble aggregates are the major toxic species. What isn’t included in this paper is a test of whether there is a functional decrease...  Read more

It seems increasingly likely that α-synuclein is central to the pathogenesis of Parkinson disease and diffuse Lewy body disease. Therefore, strategies like the one outlined in this paper to decrease α-synuclein protein levels may be viable options for therapeutics in the future. Other labs have reported different approaches to the same problem, such as recombinant intrabodies (Emadi et al., 2004), anti-aggregating peptides (El-Agnaf et al., 2004), or overexpression of β-synuclein (Hashimoto et al., 2004). Where the current paper is especially significant is that the major species decreased by this immunization strategy is aggregated forms of the protein. Although this is very hard to quantify, in figure 5 the amount of monomer is left relatively unchanged where oligomers are decreased. This observation provides us with a way to test the idea that relatively soluble aggregates are the major toxic species. What isn’t included in this paper is a test of whether there is a functional decrease in damage resulting from synuclein aggregation, which would be the best predictor of efficacy in humans. Some of the viral models of α-synuclein expression (Lo Bianco et al., 2002; Kirik et al., 2003) have additional measures such as TH cell loss. It will be interesting to see if this peripheral immunization with recombinant synuclein can slow TH cell loss or behavioral outcomes (which are mild so far in many of these models).

There will inevitably be questions about how the antibodies are effective against an intracellular protein, which is very different from the previous reports of immunization against extracellular amyloid plaques. We can be certain that there will be several future studies aimed to define the mechanism, but this is more than an academic argument. The various models posited by Masliah et al. seem to require that there is a small amount of the antigen available to antibodies on the surface of the neuron. Whilst this is true for synuclein, to my knowledge it is not true for other intracellular proteins that aggregate. For these proteins, we may need additional approaches to really get at the aggregation within a neuron. This may be a difficult process, but I think it’s worth trying to jump over the hurdles and get to the point where we can really address whether protein aggregation is central to neurodegeneration—and how to prevent it.

View all comments by Mark Cookson

α-synuclein is a soluble protein that is highly enriched in the presynaptic terminals of neurons in the central nervous system (Iwai A et al., 1995), and it may play a central role in the pathogenesis of Parkinson disease (PD) (Polymeropoulos et al., 1997; Singleton et al., 2003). It is not clear what initiates the production of this native protein, nor is it known by which mechanism α-synuclein can induce neurodegenerative disease. Trojanowski and colleagues (Trojanowski and Lee, 1998; Trojanowski et al., 1998) have suggested that accumulation of oligomeric forms of α-synuclein and/or bigger aggregates in the synaptic terminals and axons may be neurotoxic. Using a Drosophila model of PD, Chen et al. recently demonstrated that phosphorylation of α-synuclein controls neurotoxicity and inclusion body...  Read more

α-synuclein is a soluble protein that is highly enriched in the presynaptic terminals of neurons in the central nervous system (Iwai A et al., 1995), and it may play a central role in the pathogenesis of Parkinson disease (PD) (Polymeropoulos et al., 1997; Singleton et al., 2003). It is not clear what initiates the production of this native protein, nor is it known by which mechanism α-synuclein can induce neurodegenerative disease. Trojanowski and colleagues (Trojanowski and Lee, 1998; Trojanowski et al., 1998) have suggested that accumulation of oligomeric forms of α-synuclein and/or bigger aggregates in the synaptic terminals and axons may be neurotoxic. Using a Drosophila model of PD, Chen et al. recently demonstrated that phosphorylation of α-synuclein controls neurotoxicity and inclusion body formation. These authors suggested that inclusion bodies are protective, because their increased numbers correlated with reduction of toxicity.

Several therapeutic strategies are currently being investigated for PD, including the administration of neurotrophic factors and transplantation of dopaminergic cells. In the current issue of Neuron, Masliah et al. have used an immunotherapeutic approach to reduce the levels of human α-synuclein in the brains of transgenic mice or to block the assembly of this protein into potentially pathological forms. For immunization, the authors used 3- and 6-month-old heterozygous transgenic mice (Line D) that express human α-synuclein and display abnormal accumulation of insoluble α-synuclein in the brain. When allowed to age, these animals mimic certain aspects of PD, such as motor and neurodegenerative deficits. Masliah et al. used eight injections of 8 μg each of recombinant bacterial α-synuclein first in complete Freund’s adjuvant (CFA) (first injection) and then in incomplete Freund’s adjuvant (IFA) (second immunization) or in phosphate buffered saline (all subsequent administrations) to induce low titers of anti-α-synuclein antibodies in both groups of experimental mice (detected by ELISA). Such low titers of antibodies are not surprising, since α-synuclein is a self-antigen in the transgenic mice used in these experiments. Using overlapping peptides, the authors next demonstrated that anti-α-synuclein antibodies recognized amino acids 85-99, 109-123, 112-126, and 126-138 of human α-synuclein. Although mice responded quite differently to immunizations, antisera with relatively low and high titers of antibodies have been specific to brain homogenates, neurons, presynaptic terminals, and intraneuronal inclusions. More importantly, the immunizations reduced α-synuclein accumulation, and the antibody titers correlated with reductions in accumulation of neuronal α-synuclein. Successful vaccination also preserved synaptic density without inducing neuroinflammatory responses. However, it should be mentioned that active immunizations with fibrillar Aβ42 did not induce inflammation in any of the mouse models of Alzheimer disease (AD), as well, but inflammation was detected in AD patients that received the AN-1792 vaccine containing human Aβ42.

To further characterize the effects of the immunizations, the authors analyzed the levels of synaptophysin, a 38-kDa calcium-binding glycoprotein that is present in the presynaptic vesicles of neurons and in the neurosecretory granules of neuroendocrine cells in vaccinated and/or CFA injected mice. In the CFA injected transgenic mice, the levels of synaptophysin were reduced, but vaccinated and non-transgenic mice had similar levels of this protein. The authors also analyzed the potential therapeutic mechanism of anti-α-synuclein antibodies in vivo and demonstrated that this antibody bound to α-synuclein associated with neuronal membranes and promoted the degeneration of α-synuclein aggregates. The same data were generated with purified and fluorescein isothiocyanate (FITC)-tagged antibodies. This allowed authors to propose that the antibodies generated after active immunizations recognized aggregates of α-synuclein that are associated with the neuronal membrane and/or bound, for example, via Fc-receptors, to proteins on the cell surface, were internalized, and thus could be targeted, along with the antigen, to the lysosomal pathways. Unfortunately, the authors did not analyze the isotype of the antibodies, so it is difficult to discuss the possibility of their binding to FcR. In addition, it is still not clear how antibodies in lysosomes can promote degradation of α-synuclein in this compartment. In fact, it is difficult to explain the clearance of any intracellular endogenous protein by extracellular antibodies. Further studies are required to understand such mechanisms.

Another important question is whether there is an inflammatory response to α-synuclein immunizations. As mentioned above, there were no reports of any adverse autoimmune or inflammatory responses in peripheral tissues or in the brains of APP/Tg mice, a model for AD, after immunization with fibrillar Aβ42. The only adverse event observed was the increased incidence of cerebral microhemorrhage detected in very old APP/Tg mice injected with high levels of anti-Aβ monoclonal antibodies. However, there were no documented reports of APP/Tg mice developing cerebral hemorrhages in response to active immunization with fibrillar Aβ42. Despite impressive preclinical results, Aβ-immunotherapy of AD patients was halted because a small group of vaccinated patients developed meningoencephalitis. Presently, the actual cause of the adverse effects of active immunization with the AN-1792 vaccine is not known, but importantly, the antibody response to Aβ did not correlate with the presence or severity of the symptoms. In fact, some of the patients that developed meningoencephalitis did not have detectable levels of antibodies specific to Aβ peptide, suggesting that the adverse reaction to Aβ-immunotherapy was not due to the humoral antibody response, but rather to a cell-mediated immune response stimulated by AN-1792. The prominent T cell infiltration documented in two case reports currently available from the clinical trial strongly suggest that the sub-acute meningoencephalitis was caused by autoreactive anti-Aβ CD4+ and/or CD8+ T cells. Unfortunately, information on the cytokine profiles in the affected patients has not yet been published. This is unfortunate, because Th1 cytokines (Il12, IL18, IFNγ) have been implicated in many autoimmune disorders, whereas Th2-type responses (IL-4, IL-10, and TGFβ) attenuate cell-mediated immunity and inhibit autoimmune disease.

In this first α-synuclein-immunotherapy, study analyses of Th1 and Th2, cell responses were not included, though these could help us better understand the potency of this vaccine approach for humans. Even the detection of IgG1 and IgG2a isotypes, for example, could help distinguish between a Th1- or Th2-type of humoral immune response.

However, it is clear that the data generated by scientists from the University of California, San Diego, and Elan Pharmaceuticals, Inc. will stimulate further preclinical trials with α-synuclein immunogens in different animal models and perhaps even clinical trials in PD patients. Such studies will also allow us to better understand the antibody-mediated α-synuclein clearance mechanisms and how these relate to the pathology of PD.

View all comments by Michael G. Agadjanyan

Since 2005, have other investigators been able to replicate those results from Masliah's group where anti-α-synuclein antibodies are able to penetrate the blood-brain barrier and go into the intracellular compartment (intraneuronal) Lewy bodies to bind α-synuclein?

Since 2005, I have not seen any successful anti-α-synuclein immunotherapy or vaccination except from Sierks's group using ScFV intrabodies.

Experts: Do you believe in that kind of approach for intracellular neurodegenerative deposits?

View all comments by Dan Goldy