Blunt Instruments: α-Synuclein Antibodies Poorly Distinguish Forms
Quick Links
The exquisite specificity of antibodies powers neurodegenerative research, enabling reliable assays, precise diagnostics, and targeted therapeutics. Unless … some antibodies are not as specific as has been thought. In the September 22 Neurobiology of Disease, researchers led by Hilal Lashuel at École Polytechnique Fédérale de Lausanne, Switzerland, present a provocative study of 16 α-synuclein antibodies reported to be selective for oligomeric or fibrillar forms of the protein. In rigorously controlled experiments on recombinant α-synuclein, Lashuel and colleagues found that nearly all the antibodies reacted equally strongly with oligomers and fibrils, and most bound weakly to monomers as well. In other words, none were truly conformation specific. The findings suggest the need to verify previous claims about the role of oligomers in Parkinson’s disease that were based solely on antibody binding, Lashuel told Alzforum.
• Many α-synuclein antibodies are reported to be specific for oligomers or fibrils
• But a survey of 16 such antibodies found they bound all forms of α-synuclein
• The findings raise questions about the accuracy of data obtained with conformation-specific antibodies
Other scientists deemed this study valuable. “This is an extremely useful analysis of the current α-synuclein antibodies used in the field, and quite an eye-opener,” Laura Parkkinen at the University of Oxford, U.K., wrote to Alzforum. Gregory Petsko at Weill Cornell Medical College in New York City noted, “In the neurodegenerative disease field in particular, given the use of antibodies as therapeutic agents in clinical trials, these results should sound a loud alarm bell,” (see full comments below).
Varied Shapes. Electron microscopy reveals a mix of structures among α-synuclein oligomers (upper left), contrasting with amorphous blobs in dopamine-incubated oligomers (bottom left), long chains in HNE-incubated oligomers (bottom right), and filaments in fibril preparations (upper right). [Courtesy of Kumar et al., Neurobiology of Disease.]
As with Aβ in Alzheimer’s disease, α-synuclein oligomers are believed to be more toxic than larger deposits (Dec 2007 conference news; Oct 2010 news). Researchers have tried to develop antibodies selective for both types of aggregate. Of the handful of α-synuclein antibodies in clinical trials, AbbVie’s ABBV-0805 is purported to be oligomer-specific, while Biogen’s cinpanemab and Roche’s prasinezumab recognize any aggregated forms. However, few data are available on how these clinical antibodies, or others used for research, were characterized.
To fill in this knowledge gap, Lashuel and colleagues undertook a survey of 16 α-synuclein antibodies used for research. None are in clinical trials. They tested six antibodies that are reputed to be fibril-specific: three from Biolegend (SYNO2, SYNO3, and SYNO4), two from the University of Pennsylvania (7015 and 9029), and one from Abcam (MJFR-14). Another six were raised against aggregated forms of α-synuclein: four from Biolegend (A17183A, A17183B, A17183E, and A17183G), one from Kovacs (5G4), and one from Agrisera (ASyO5). Finally, with the support of the Michael J. Fox Foundation, the authors generated four antibodies directed against oligomers (24H6, 26B10, 12C6, and 26F1).
Joint first authors Senthil Kumar and Somanath Jagannath prepared standardized synthetic α-synuclein samples against which to test these 16 antibodies. After incubating the protein at high temperature to allow aggregation, they separated different-size molecules via filtration and size exclusion chromatography. This produced pure solutions of monomers, oligomers, and fibrils. The authors created two additional oligomeric mixtures by adding reagents to the monomers during incubation. In one case they added dopamine, which tends to generate disordered oligomers; in the other, the lipid peroxidation product 4-hydroxy-2-nonenal (HNE), which tends to induce β-sheet containing oligomers.
Circular dichroism spectroscopy revealed that the original oligomeric preparations assumed a range of structures, including spheres, tubes, and rings, that all contained β-sheets. The authors did not further characterize potential structural differences among the species in this oligomer solution. On the other hand, when oligomers formed in the presence of dopamine, they clumped into disordered globs that lacked β-sheet structure. In HNE solutions, oligomers assembled into longer chains of β-sheets (see images).
Test Material. Five different preparations of α-synuclein help test antibody specificity. [Courtesy of Kumar et al., Neurobiology of Disease.]
Kumar and colleagues dotted these five α-synuclein preparations onto nitrocellulose membrane and added each antibody to assess binding. In this slot blot paradigm, all antibodies bound with nearly equal affinity to all aggregated forms. Two exceptions were 26F1 and 5G4, which barely reacted with monomers or dopamine-incubated oligomers, suggesting they might be specific for β-sheet structure. Most of the other antibodies bound strongly to all aggregates and weakly to monomers, but the signal increased at high monomer concentration. Some bound all forms of α-synuclein equally.
In a sandwich ELISA assay, which uses low protein concentrations, few of the antibodies bound monomers. However, again, most of the antibodies bound oligomers and fibrils with equal strength. 5G4, A17183E, 24H6, and 26B10 were exceptions, binding to fibrils much better than to oligomers.
Lashuel believes this relative lack of selectivity for putative conformation-specific antibodies points to the need to better define the properties of these reagents. “We need to devote more time, effort, and resources to characterizing and validating the tools and assays that we use to build the foundation of our research and drug discovery programs,” he said.
Tiago Outeiro, University Medical Center Göttingen in Germany, agreed on the need for more precise guidelines and standardization in the field. At the same time, Outeiro noted that the recombinant α-synuclein preparations used here likely do not match the post-translationally modified forms of the protein found in the brain, and so are not the ideal material for this task. Michel Goedert, MRC Laboratory of Molecular Biology, Cambridge, U.K., points out that conformers of α-synuclein aggregates differ between the different synucleinopathies (see full comments below). And Lashuel agrees with that. He plans to isolate α-synuclein oligomers and fibrils from brain tissue and use this material to characterize antibodies in the future.
Tim Bartels at the U.K. Dementia Research Institute at University College London said Lashuel’s study will be a helpful guide for researchers in selecting α-synuclein antibodies. Despite the relative lack of specificity shown here, he believes the antibodies are still useful research tools, since most do bind aggregates more strongly than monomers. Bartels also speculated that the oligomer preparations used in this study might contain some small diffusible fibrils as well, which would further blur the issue of antibody specificity. CD spectroscopy would not distinguish those structural details, Bartels said (see comment below).
In the Alzheimer’s field, antibodies with imperfect selectivity, such as BAN2401 and aducanumab, have made it into late-stage trials and even onto the FDA’s desk for marketing approval.—Madolyn Bowman Rogers
References
News Citations
- San Diego: α-Synuclein Protofibrils Create Sense of Déjà Vu in PD, DLB
- An Oligomer Test for PD—Could One for AD Be Close Behind?
Therapeutics Citations
Other Citations
Further Reading
News
- An Oligomer Test for PD—Could One for AD Be Close Behind?
- Uppsala: Immunotherapy—Not Just for AD Anymore
- α-Synuclein Antibodies Enter Phase 2, Sans Biomarker
- Shape of α-Synuclein Aggregates Influences Pathology
- Electron Microscope Yields Finer Structure of α-Synuclein, Aβ Fibrils
- Parkinson’s Treatments Go After Genetic Targets
Primary Papers
- Kumar ST, Jagannath S, Francois C, Vanderstichele H, Stoops E, Lashuel HA. How specific are the conformation-specific α-synuclein antibodies? Characterization and validation of 16 α-synuclein conformation-specific antibodies using well-characterized preparations of α-synuclein monomers, fibrils and oligomers with distinct struct. Neurobiol Dis. 2020 Dec;146:105086. Epub 2020 Sep 22 PubMed.
Annotate
To make an annotation you must Login or Register.
Comments
University of Oxford
This is an extremely useful analysis of the current α-Syn antibodies used in the field, and quite an ‘eye-opener’.
The study clearly shows that all the antibodies that have claimed to be ‘oligomer/conformation-specific’ also detect fibrils, i.e. the antibodies cannot distinguish between oligomeric and fibrillar forms. Although they may have preferential binding toward aggregated forms, they still do pick up also monomeric forms in a concentration-dependent manner. This is an extremely important point to raise, as it continues to cause confusion in the field even among the experts and misinterpretation of some previous studies.
The pathological heterogeneity of α-Syn-positive structures in the human brain when using different antibodies is well-recognized in the neuropathology field (Beach 2018, Croisier 2006). However, it has not been studied in detail due to lack of well-characterized antibodies (to the level done here), as well as lack of AI-based digital imaging tools which are only now emerging.
Our own preliminary studies have shown that this heterogeneity can be more profound than previously accepted, e.g. α-Syn aggregation in astroglial cells is only detected with α-Syn antibodies that have their epitope within the central NAC region. This suggests that the aggregates in different cell types are most likely formed of different α-Syn proteoforms, and that we really need to use several better-characterized antibodies to capture the diversity of a-Syn species in the brain.
References:
Beach TG, White CL, Hamilton RL, Duda JE, Iwatsubo T, Dickson DW, Leverenz JB, Roncaroli F, Buttini M, Hladik CL, Sue LI, Noorigian JV, Adler CH. Evaluation of alpha-synuclein immunohistochemical methods used by invited experts. Acta Neuropathol. 2008 Sep;116(3):277-88. Epub 2008 Jul 15 PubMed.
Croisier E, MRes DE, Deprez M, Goldring K, Dexter DT, Pearce RK, Graeber MB, Roncaroli F. Comparative study of commercially available anti-alpha-synuclein antibodies. Neuropathol Appl Neurobiol. 2006 Jun;32(3):351-6. PubMed.
Harvard Medical School and Brigham&Women's Hospital
I think most thoughtful people in the field have wished for better antibodies for some time. This work will make it clear how urgent that is. The question is whether it’s possible to get them.
Structural biologists have often been suspicious of ‘conformation-specific’ antibodies in the absence of direct data about what exactly is being detected. Reproducibility of biological experiments has been a topic of much discussion in the past few years, and many of the people involved have come to the conclusion that a lot of irreproducibility is due to, for lack of a better word, crappy antibodies.
In the neurodegenerative disease field in particular, given the use of antibodies as therapeutic agents in clinical trials, these results should sound a loud alarm bell. The authors were able to identify a few antibodies that were somewhat more specific than most, which should be somewhat encouraging, but even those had their issues.
Let me end with a word of praise and thanks to the authors. The painstaking quality control study that you see here might well be almost impossible to do in an academic lab in the United States these days, given the current funding situation and the drive to publish hot new findings. It’s probably not surprising that, even in Europe, this study had participation by industry as well as academia. The authors have done the field a service. Let’s hope the field responds by finding better ways to tackle this problem.
University Medical Center Goettingen
In this important study, Kumar and colleagues conducted an impressive amount of rigorous work on the study of different forms and assemblies of alpha-synuclein, a central protein in synucleinopathies. The study is systematic, carefully executed, and highlights the need for uniformization in terms of protocols and language used in the field.
Recent evidence from numerous studies in cell and animal models of disease suggest that oligomeric species of different proteins might be more toxic that the larger, fibrillar forms. This idea is backed up by a lack of correlation between the presence of the typical pathological hallmark inclusions and disease. E.g. at autopsy, amyloid-beta plaques and Lewy bodies are often found in the brains of people who had no overt signs of Alzheimer’s or Parkinson’s disease, respectively.
The idea that oligomeric forms of proteins are formed “on pathway” to the formation of the mature amyloid-like fibrils is also supported by numerous studies. However, the field suffers from major limitations, including the ambiguity associated with the term “oligomer,” as this is often used to refer to a variety of protein species that is difficult to compare between different laboratories.
The present study assessed the behavior of a panel of 18 antibodies in the context of in vitro-prepared alpha-synuclein species: monomers, oligomers, and fibrils. The findings may appear surprising but, in fact, I believe they were to be expected. The antibodies were developed using alpha-synuclein species produced in different laboratories and, therefore, they may not be directly comparable to those used in this study. In addition, and as the authors point out, a major limitation is to know how the alpha-synuclein species used as reference in the study relate to those accumulating in the human brain. One may expect them to be very different in fact, due to the absence of the posttranslational modifications that take place in any biological context.
Nevertheless, the study has great merit as it highlights the fact that we need to be cautious when assuming the specificity of the numerous antibodies used in the field. The study also highlights the need for more precise guidelines and standardization in the field. This is the only way we might move forward and rationally develop tools and strategies for advancing our understanding of these devastating diseases.
U.K. Dementia Research Institute at University College London
Kumar and colleagues set out in their study to characterize a variety of available ‘conformation- specific antibodies’ for α-Syn to test the antibodies’ specificity toward monomeric, oligomeric and fibrillar forms of α-Syn. The results largely agree with what is known in the field, i.e. most conformation-specific antibodies detect oligomeric and fibrillar α-Syn with higher sensitivity compared to monomers, but are ultimately not completely selective.
The study also finds that there does not seem any distinction between fibrillar and smaller oligomeric material, even though this might be due to the authors’ preparation method of oligomeric samples.
This study will be a helpful guide for people looking to use the described antibodies in their research, and it emphasizes the importance of being mindful of the relative concentrations of monomers in your preparations. The results do imply that most of these antibodies have higher sensitivity toward oligomers and fibrils than monomer. Given that, using these antibodies mostly in comparative studies, in which total α-Syn protein concentration (i.e. monomer + multimers/oligomers + fibrils), independent of conformation, is similar between samples will still deliver meaningful results.
MRC Laboratory of Molecular Biology
Few questions have led to more confusion than that of the relative importance of oligomers and filaments. Part of the problem stems from the fact that filaments are easily observed in human brain, whereas oligomers are largely invisible. It would therefore be valuable to have reagents that are specific for protein oligomers.
One option is to produce antibodies that can distinguish between monomers, oligomers, and filaments. It has been reported that such antibodies exist for α-synuclein. Kumar and colleagues have now studied these and other antibodies using a wide variety of techniques, including single molecule assays, surface plasmon resonance, and immunoblotting. They report that none of the 18 anti-α-synuclein antibodies analysed were specific for oligomers of recombinant α-synuclein.
Although some antibodies did not visualize monomeric alpha-synuclein, they recognised both oligomers and filaments. We do not know how the structures of recombinant α-synuclein oligomers relate to those of oligomers from human brain. However, we do know that the structures of alpha-synuclein filaments from multiple system atrophy (MSA) differ from those of recombinant assembled protein (Schweighauser et al., 2020).
Distinct conformers of assembled alpha-synuclein are probably present in Parkinson’s disease, dementia with Lewy bodies and MSA (Spillantini et al., 1998a, Spillantini et al., 1998b; Crowther et al., 2000; Schweighauser et al., 2020). We are looking forward to the production of antibodies that are specific for the conformational states of assembled alpha-synuclein from human brain.
References:
Crowther RA, Daniel SE, Goedert M. Characterisation of isolated alpha-synuclein filaments from substantia nigra of Parkinson's disease brain. Neurosci Lett. 2000 Oct 6;292(2):128-30. PubMed.
Schweighauser M, Shi Y, Tarutani A, Kametani F, Murzin AG, Ghetti B, Matsubara T, Tomita T, Ando T, Hasegawa K, Murayama S, Yoshida M, Hasegawa M, Scheres SH, Goedert M. Structures of α-synuclein filaments from multiple system atrophy. Nature. 2020 Sep;585(7825):464-469. Epub 2020 May 27 PubMed.
Spillantini MG, Crowther RA, Jakes R, Hasegawa M, Goedert M. alpha-Synuclein in filamentous inclusions of Lewy bodies from Parkinson's disease and dementia with lewy bodies. Proc Natl Acad Sci U S A. 1998 May 26;95(11):6469-73. PubMed.
Spillantini MG, Crowther RA, Jakes R, Cairns NJ, Lantos PL, Goedert M. Filamentous alpha-synuclein inclusions link multiple system atrophy with Parkinson's disease and dementia with Lewy bodies. Neurosci Lett. 1998 Jul 31;251(3):205-8. PubMed.
The Michael J. Fox Foundation for Parkinson's Research
The study from Kumar and colleagues is an important contribution to the field. It attempts to bring some consensus on the validity and characterization of these commonly used "conformation-specific" antibodies. Characterizing and validating antibodies with different orthogonal assays is a significant undertaking, but it will be important for the field to agree on criteria to validate antibodies (and other tools/assays for that matter), so that companies and labs developing these tools can include comprehensive tool characterization in their data packages.
The study shows these antibodies' general inability to distinguish oligomers from fibrils. This was somewhat expected as we have a limited understanding of the structural determinants of oligomers and how they compare to fibrils, and there is a lack of consensus on the defining characteristics of an oligomer. Therefore, these antibodies might have been raised against heterogeneous species ranging from small oligomers to soluble sonicated preformed fibrils.
The ability for most of these antibodies to also bind to monomeric α-Syn suggests that binding preference to aggregated forms of α-Syn is driven by avidity rather than specificity, an important observation that should be taken into consideration in future studies. This will be especially important in studies aimed at measuring synuclein species in human bio-specimens, where aggregated forms of α-Syn are either transient or exist at much lower concentrations than monomeric α-synuclein. There should be an effort to ensure that assay conditions are set up to minimize/avoid binding to α-Syn monomers.
Finally, while this study may be mostly perceived as a negative outcome for the field, it is nevertheless positive that some of these antibodies are able to capture different types of oligomers and fibrils, suggesting the potential to capture the diversity of α-Syn aggregates in human bio-specimens.
Aarhus University
This is a tour de force assessment of different antibodies previously reported to be specific for certain types of α-synuclein aggregate. They are not!
I commend the authors for this thorough, systematic, honest, and self-critical approach to preparing samples. This is exactly what the field needs.
This study contains nice details, such as tests using multiple freeze-thaw rounds followed by painstaking EM analyses to check for sample stability, and a robust assessment of the study’s limitations.
The conclusion is a good starting point for researchers choosing the right antibodies for their studies. This study is a service to the Parkinson’s community, and a reminder to the whole amyloid field to be extra vigilant about the conclusions they draw from antibodies claimed to be specific for certain species or conformations. I thoroughly enjoyed reading it.
Aarhus University
The comprehensive comparison of aggregate-selective α-synuclein antibodies by Kumar et al. has been published after the preprint version earlier was published on BioRxiv. The field’s responses are positive—but then what?
I think the title of this Alzforum piece is misleading. The antibodies are not blunt, and far from those crappy ones that Gregory Petsko mentions have marred the field of neurodegenerative diseases.
One reason for the confusion is that we have been studying in vitro-formed oligomers and fibrils for more than a decade. However, not until the recent Schweighauser 2020 Nature paper did we actually see the true high-resolution cryoEM structures of aggregates from brains affected by MSA. What did we see? Two different folding strains isolated from one brain and evidence that aggregates from DLB brains are different.
The two resolved strains were not identical to the cryoEM structures from in vitro-formed filaments, but did share significant similarity. So how much can we ask for in terms of specific antibodies? What if aggregates vary within the brains of individual patients, and between patients with identical clinically perceived diseases? The currently generated α-synuclein aggregates, amplified from patient samples by in vitro RT-QuIC or PMCA techniques, may help us get an idea of the variability between patient aggregates, although they of course may be artifactually shaped by the in vitro amplification protocol.
Regarding the oligomers we are more in the open. They are hypothesized to exist in patient brains but have not yet shown their coat. We have built our belief on structures generated in vitro, using different α-synuclein aggregation protocols often spiced up with a twist of chemicals like dopamine, lipid aldehydes, etc. Our reason to believe these oligomers resemble what may exist in brain is meager, and our techniques to actually isolate them haven’t yet been blessed by techniques like PMCA or RT-QuIC and still lack reproducible protocols. This will likely remain a controversial field for quite some time.
Back to the antibodies that were tested by Kumar et al. Most were able to distinguish between monomer and aggregated α-syn species. The SRP-based kinetic analysis in table 3 demonstrates the highest difference in KD for monomer/fibril of >4000 for the MJFR-14. This is pretty good.
The study actually demonstrates we have antibodies that not just bind α-synuclein but can distinguish between some species. Of course we shall continue to improve our tools, but this will take time and resources, and another paper might be written in five years that states these new tools are blunt.
I suggest we focus on refining our protocols with the tools available. This may be done by an online assembly of protocols from those who manage to get positive results from the antibodies. Other investigators can then ask questions and challenge their results. This approach should enable us to have a platform representing best practice that, importantly, is constantly curated by researchers in the field. BioRxiv may serve the purpose by allowing us to upload data that can be commented on by readers.
References:
Schweighauser M, Shi Y, Tarutani A, Kametani F, Murzin AG, Ghetti B, Matsubara T, Tomita T, Ando T, Hasegawa K, Murayama S, Yoshida M, Hasegawa M, Scheres SH, Goedert M. Structures of α-synuclein filaments from multiple system atrophy. Nature. 2020 Sep;585(7825):464-469. Epub 2020 May 27 PubMed.
Make a Comment
To make a comment you must login or register.