Date: February 2017
Authors: Howard E. Gendelman, MD, Martin Ingelsson, Vladimir Kostić, MD, PhD, Leonidas Stefanis, MD, PhD
Editors: Michael S. Okun, MD, and Stella M. Papa, MD
The relationship of the immune system to the pathogenesis of Parkinson’s disease (PD) and in particular to alpha-synuclein (α-syn) has been the focus of intense research efforts over many years. This relationship has come recently to the forefront of interest of the PD community, as the first clinical trials in PD using immunization strategies against α-syn have commenced. The apparent encouraging results from similar strategies targeting beta-amyloid (Aβ) in Alzheimer’s disease (AD) have spurred the interest further, and beyond to different strategies for immunologic therapies in PD. We have asked two experts in the field – Drs. Howard Gendelman and Martin Ingelsson – to comment on these issues. Dr. Vladimir Kostić has provided an additional commentary from a clinical perspective.
Given the multiple mediators of the immune responses apparently involved in PD pathogenesis, which components have been identified as specific targets for immunotherapies?
Howard Gendelman: As we all accept, there are no means yet to repair neural damage in PD. Current therapeutic modalities are neither curative nor protective. None halt the progressive nigrostriatal degeneration or the more widespread degeneration that is responsible for many of the debilitating non-motor symptoms of the disease. The major pathological hallmarks of the disease include intraneuronal Lewy bodies (LBs), inflammation and oxidative stress. Increased numbers of activated microglia surround areas of dopaminergic neuronal loss. Whether the immune control of the brain's microenvironment exerts its effects indirectly, via microglia, or directly upon neuronal cells is a subject of debate. In any case, transforming the adaptive immune system, based on B- and T-cell responses represents a viable therapeutic approach.
There are many unanswered questions: How can one improve nigral neuronal function by shifting the functional activities of the innate (based on microglia) and adaptive immune system? Can this occur acutely by altering neurotransmitters, or just shifting the cytokine environment damaged by the release, and then engagement of misfolded aggregated and nitrosylated proteins, such as α-syn? Can misfolded proteins by themselves cause disease pathobiology, and how can such activities be harnessed for treatment? Our own research seeks to investigate how immunity can be harnessed as a neuroprotective strategy, and in so doing work to target specific elements of the immune system for therapeutic gain. Our goal is to develop relevant disease biomarkers that can be altered to improve clinical outcomes; the hope is that such biomarkers can predict disease outcomes and monitor strategic improvements of PD manifestations.
Based on the role of α-syn in PD pathology, immunization against α-syn may be a good strategy, but an effective immunization in patients faces many challenges. What are the main problems that need to be considered with this approach?
Martin Ingelsson: Immuno-based therapies have proven efficacious for a growing number of non-infectious disorders, ranging from various malignancies to rheumatic disorders and osteoporosis. As for CNS disorders, only multiple sclerosis can yet be treated with this type of approach. However, a number of immunotherapy trials on AD - targeting Aβ, the protein accumulating as plaques - have been initiated and some of these are still ongoing. Similarly, we now hope that immunotherapy against α-syn, the protein accumulating as Lewy bodies/Lewy neurites, can be a feasible strategy for PD. The situation for α-syn in PD is in many aspects similar to Aβ in AD. Strong genetic and biochemical evidence indicate that both proteins are truly central to the pathogenesis of the respective disorders. Moreover, several studies have shown that prevention and/or reduction of aggregated Aβ and α-syn on rodent transgenic models by immunotherapy can partially alleviate the phenotype of the mice. Thus, there are good reasons to believe that targeting α-syn by immune-mediated therapies could be effective.
Similar to, and even more important than for Aβ-directed therapies, it may be inappropriate to target α-syn in general. This protein is believed to carry out important synaptic functions and a general reduction of α-syn may thus lead to unwanted effects. Moreover, targeting peripheral (presumably monomeric) α-syn should probably also be avoided. For these reasons, it could be more advantageous to seek a selective reduction of aggregated species. Since several lines of evidence suggest that prefibrillar, still soluble, oligomers/protofibrils have a pronounced neurotoxic effect, such protein forms may be particularly suitable to target, and this is an approach that we have taken in our own work.
Studies in transgenic mice have demonstrated a reduction of α-syn pathology upon both passive and active immunization. From such studies it is not possible to tell whether the effects can be ascribed to intra- or extracellular action (or both). Passage of monoclonal antibodies over the blood brain barrier is limited. However, cell culture studies indicate that the antibodies can readily pass over cellular membranes. Thus, even though only a small fraction of the administered antibodies reach the central nervous system, it is likely that a substantial portion of them could end up inside cells (where most of the pathology can be found). However, the growing realization that protein pathology might be propagated from cell to cell makes it more likely that we could produce therapeutic effects also by targeting α-syn extracellularly.
The completed and ongoing trials targeting Aβ in AD have taught us several things. Firstly, we need to aim at recruiting the patients at an early disease stage. Secondly, we have to ascertain that the diagnosis is correct. Thirdly, we need to demonstrate a treatment effect on one or several parameters related to the pathogenesis. For AD, it has been possible due to the detailed knowledge of the early symptomatology and to the development of novel PET based imaging methods. As for PD, the prodromal symptoms have not yet been as clearly defined and therefore cannot be easily used for patient recruitment. Moreover, we still do not have any method by which we can demonstrate the accumulation of α-syn in the PD brain. Thus, novel diagnostic methods – based either on cerebrospinal fluid analyses or positron emission tomography imaging - need to be developed. Once in place, such techniques may be useful both to identify subjects for clinical trials and to demonstrate target engagement.
Howard Gendelman: LBs containing aggregated and nitrated a-syn released to the extraneuronal environment induce activated microglia and affect the emergence of T cell populations from novel antigens presented in the periphery. Activated microglia produce proinflammatory and neurotoxic mediators that damage surrounding neurons. Such an innate immune response may also affect immunological tolerance resulting in specific T cell activation profiles. For example, our group has demonstrated that in the presence of modified self-protein, nitrated a-syn, specific effector T cells (Teff) are induced in the periphery and traffic to inflamed foci of the brain to mediate neurodestructive activities in animal models of PD, either by direct neurotoxicity or by indirect exacerbation of microglial function. Alternatively, regulatory T cells (Treg) may act in a neuroprotective function by maintaining immunological tolerance and modulating inflammation. In the setting of neurodegeneration, Treg can transform T cells with neurodestructive Th1 and Th17 responses to neuroprotective cells, serving to protect damaged dopaminergic neurons.
Clearly, therapies directed to the inflammation surrounding α-syn pathology may be neuroprotective. We ask both experts to comment on the advantages and the shortcomings of immunomodulation versus immunization.
Martin Ingelsson: A successful development of a future therapy for PD needs to build on the knowledge of the underlying pathophysiological processes. The observation of pathological features on post mortem PD brain tissue does typically not provide us with any information as to whether the changes are primary or secondary. Hence, we cannot be certain that the α-syn -containing LBs in themselves are inducing the potentially damaging activation of the immune system. Alternatively, immune system-related events might occur first and lead to a changed microenvironment that in its turn would trigger the protein pathology. Yet another scenario would be that these processes are influencing and reinforcing each other, meaning that both pathological alpha-synuclein and aberrant inflammatory reactions could be relevant therapeutic targets.
The possibility that immune responses, including the activation of microglia and astrocytes as well as the release of different cytokines and an altered balance between certain T-cell subpopulations, would be causing or triggering neurodegeneration has been explored in various animal models. Several studies have provided encouraging data, for instance suggesting that activation of regulatory T cells or the use of certain Non Steroidal Anti-Inflammatory Drugs (NSAIDs) can ameliorate nigral pathology. However, we need to examine more closely the relationship between α-syn and inflammation - as most of the previous studies have been carried out in models that were not α-syn-related. We also need to investigate if there are certain components of the immune system that are altered in a disease-specific manner to be utilized as biomarkers. No such CSF or plasma markers have been demonstrated to date, but the use of deprenyl as an astrocytosis-related Positron Emission Tomography (PET)-ligand has indicated that inflammatory changes may occur early in the course of AD. Thus, such an imaging method – or one that is based on radiolabeled antibodies against a critical neuroinflammatory marker – might be proven useful for the detection also of prodromal PD.
Whereas epidemiological studies have indicated that individuals using NSAIDs have a reduced risk for both PD and AD, therapeutic intervention with anti-inflammatory strategies have so far failed for both disorders. Taken together, these observations could mean that the immune system responses are appropriate pharmacological targets only at a very early disease stage. Once LB pathology has developed, it might be more fruitful to target the factors that underlie the immune responses, i.e. the α-syn protein itself.
Howard Gendelman: It is unclear whether immunotherapies targeting α-syn are a viable therapeutic option. While removal of α-syn (through active and passive immunotherapy) has been shown to modify disease in animal models, there are significant hurdles left in translating such interventions to human disease. There is the fact that the protein aggregates are intracellular, and the ability of such approaches to target them may be suboptimal. In addition, there is a real potential for autoimmunity and that effector T cell responses that could be elicited in humans would speed microglial inflammatory responses and neural tissue injuries. Another issue is that, despite the design of the antigen, the exclusivity of the immune response to α-syn alone remains uncertain. Exclusivity is essential for any ultimate positive outcome, it is critical for example that the immune response does not cross-react with the closely related β-synuclein. Penetration of the immune response within the brain would be needed for optimal effect, and this would need to affect both microglia and neurons. The limitations in translating vaccination animal studies to human clinical trials in AD need be considered, and notably the emergence of adaptive immune neurodestructive responses in the latter.
Martin Ingelsson: Both immunomodulation and immunization have been successful in mouse models and I therefore consider it worthwhile to also explore these approaches for clinical use. However, the literature describes many examples of studies that have worked excellently well in rodents but where the same approaches have failed in the clinical trial. Nevertheless, I believe there are reasons to be cautiously optimistic regarding the prospect for immune-based interventions against PD. Among the different strategies I would then mainly envision passive immunization with alpha-synuclein monoclonals as the most fruitful way forward. If the recent promising data from the antibody-based AD clinical trials hold up I think that we in a similar fashion can look forward to positive results also from α-syn -based trials. Admittedly, the α-syn pathology in PD will most likely be harder to target than the Aβ pathology in AD. The antibodies should be designed to avoid the normal, physiological protein and they probably need to penetrate not only the BBB, but also the neuronal/glial cell membranes. At the clinical level, one of the greatest challenges will be to identify patients at a very early stage, before a too extensive cell death has occurred. However, I believe that ongoing efforts on the development of appropriate antibodies, BBB/cell membrane transportation vehicles and prodromal diagnostics are greatly increasing the chances for an efficient future immunization therapy against PD and related disorders.
We asked Dr. Kostić to comment on the critical steps forward to advance immunotherapies as disease-modifying strategies as a path to alter the course of PD.
Commentary by Vladimir Kostić: Unfortunately, at this point of time we do not have any defined protective or repairing therapy for Parkinsonʼs disease (PD). Attempts to transfer treatment experiences from experimental animal models to patients have generally been disappointing. Many reasons may be the cause of that, but I would highlight only one. The initial trigger of a neurodegenerative process releases, in a sequence or in parallel, a whole network of pathophysiological mechanisms that maintain pathology progression. A pathological mechanism that is key at a certain point of disease development (e.g. in its early stages) need not necessarily be all that relevant in another stage of the disease. For example, although reduction of risk of PD was suggested in patients receiving NSAIDs, clinical trials exploring effects of NSAIDs on PD progression were disappointing. One may hypothesize that the main reason for such discrepancy would be related to the timing of NSAIDs administration. When a neurodegenerative disease is diagnosed, neuronal loss is already advanced and the molecular processes driving disease evolution are likely different from those favoring the initial development of the disease. Ingelsson points out uncertainty in what precedes what in the pathological sequence of PD: a-syn-containing aggregates or immune system-related changes. Therefore, for neuroprotective approaches, an obvious task for clinicians would be to obtain acceptable criteria and biomarkers for identifying the earliest stages of PD. Those we have today are neither sensitive, nor specific enough.
Immunotherapies in PD rely on two basic strategies: (a) generation of antibodies against α-syn (primarily for removal of α-syn aggregates), and (b) the induction of a particular T cell response to modulate the neuroinflammatory response. What form of α-syn represents the best therapeutic target is still not known. Designing and targeting of antibodies for specific species of α-syn and assessing their relative therapeutic potential in animal models would be a rational first step. Indeed, data provided by Spencer et al. (2016) suggest that selectively targeting different α-syn variants provides differential therapeutic effects in α-syn Tg-mice depending on the timing of administration, i.e. prophylactically or during disease development. Therefore, not only different therapeutic strategies may be most beneficial during different stages of PD, but also, more than one antibody may prove to be useful, depending on the timing of application during the disease course.