For the first time, scientists have detected α-synuclein aggregates lurking in the brains of the living. This thanks to 18F-ACI-12589, a new tracer developed by AC Immune in Lausanne, Switzerland. Presented at AD/PD 2022, held March 15-20 in Barcelona, Spain, and online, the first PET scans using the tracer showed uptake in the cerebellar white matter of people with multiple system atrophy (MSA). 18F-ACI-12589 bound specifically to α-synuclein fibrils while shunning the other amyloids that often accompany Lewy bodies—namely amyloid plaques and tau tangles. Alas, PET signals were undetectable in the brains of people with other synucleinopathies, including Parkinson’s disease and dementia with Lewy bodies, despite selective binding of the tracer to postmortem brain samples of people who had died with those disorders.

  • ACI-12589 binds α-synuclein aggregates in postmortem brain samples from different neurodegenerative diseases.
  • The tracer is selective for α-synuclein, but binds weakly.
  • In first PET scans, the tracer worked for multiple system atrophy, not other synucleinopathies.

Oskar Hansson of Lund University in Sweden led clinical studies of the tracer and unveiled the first glimpse of the PET scans at the meeting. He thinks the findings bode well for detection of α-synuclein pathology during life in people with MSA, and will aid in diagnosis as well as monitoring of treatment effects in clinical trials.

Why the tracer fell short in people with other synucleinopathies remains unclear. Hansson and others suspect it is because they have far fewer deposits, and tracer binding may have been too weak to detect them. Differences in α-synuclein fibril conformation, or in where the aggregates are in the brain, could also contribute, Hansson said.

“If verified in larger samples, this would represent a landmark moment, similar to first reports of imaging amyloid with PIB and tau with FTP (at the time T807),” wrote Gil Rabinovici of the University of California, San Francisco, in a comment to Alzforum. “Overall, while preliminary, these are exciting data, and I am cautiously optimistic that the field may have its first foothold into synuclein imaging.”

The road to synuclein imaging has been long and strewn with obstacles. Compared to amyloid plaques and neurofibrillary tangles, α-synuclein fibrils exist in low concentrations in the human brain. This, combined with its intracellular location and myriad structural conformations, makes α-synuclein a tough target for PET tracers. Multiple contenders have been developed over more than a decade of trying. All of them failed, most due to subpar binding affinity, off-target binding to aggregates of Aβ or tau, or both (for review, see Korat et al., 2021; Alzghool et al., 2022).

To meet the challenge, the Michael J. Fox Foundation had gathered consortia of scientists over the years (Eberling et al., 2013; Merchant et al., 2019). According to MJFF’s Jamie Eberling, the foundation is currently funding 10 different groups, including AC Immune. In 2016, it even offered a $2 million prize to the first group to develop an α-synuclein tracer. While AC Immune appears to be the top contender, Eberling said more data is needed to support their tracer's specificity. She noted that other groups are edging close to testing theirs in the clinic. “I would love to award that prize,” Eberling said.

At AD/PD 2022, AC Immune’s Francesca Capotosti presented preclinical data for ACI-12589. The tracer emerged from the company’s Morphomer platform, a library of some 15,000 small molecules designed to cross the blood-brain barrier, enter cells, and bind with high specificity to pathological forms of different amyloidogenic target proteins. Capotosti said more than 2,000 compounds from this library were designed and selected as candidates for α-synuclein tracers. She emphasized that the researchers used α-synuclein aggregates derived from brain samples of people with PD, then further validated using aggregates from other synucleinopathies.

The use of brain-derived α-synuclein to select potential tracers sets AC Immune’s approach apart from other groups, most of which rely upon synthetic α-synuclein fibrils, at least in initial screening steps. Fibrils generated in vitro are structurally distinct from brain-derived fibrils, and Capotosti told Alzforum that AC Immune has avoided using them for this reason. “I think our approach to screen on brain-derived material has been one of key elements that has allowed the discovery of AC-12589,” she wrote.

Previously, the researchers used this approach to develop a tau tracer, PI-2620, which binds 4R forms of the protein that get tangled up in progressive supranuclear palsy and other primary tauopathies (Jul 2020 news). Efforts are underway to develop a TDP-43 tracer as well, and at AD/PD, AC Immune’s Tamara Seredenina reported that lead candidates detect this pathology in brain tissue from people with frontotemporal lobar degeneration who had FTLD-TDP Type A, B, or C pathology.

Autoradiography. In brain sections from people with familial PD or MSA, 3H-ACI-12589 bound α-synuclein aggregates (top panels) with minimal non-specific binding (middle). The distribution matched the immunohistochemistry of phosphorylated α-synuclein (bottom). [Courtesy of AC Immune.] [Courtesy of AC Immune.]

ACI-12589 was the top compound to emerge as a potential α-synuclein tracer. A tritiated version bound Lewy body inclusions and Lewy body neurites in affected brain regions in people who had died with a synucleinopathy, be it familial PD, idiopathic PD, PDD, DLB, or MSA. Its distribution in these brain sections matched immunohistochemistry of phosphorylated α-synuclein, suggesting the tracer was binding its intended target. Capotosti reported similar findings when the tracer was radiolabeled with fluorine-18.

The tracer did not bind to brain sections or homogenates from people without α-synuclein pathology. It also ignored tau tangles or Aβ-plaque-ridden brain sections from people with AD, suggesting specificity for α-synuclein, Capotosti said.

Finally, Capotosti reported that the tracer only weakly bound monoamine oxidase B. This enzyme co-localizes with Lewy bodies and has ensnared other tracers in off-target binding, including some tau tracers.

Would ACI-12589 reach its target in the brain of a living person? Ruben Smith of Lund University presented results from a clinical study that monitored uptake in 25 participants via PET scan. Eight were healthy controls, seven had PD, two DLB, and eight MSA. MSA is broadly divided into cerebellar and parkinsonian subtypes, and this study included six people with MSA-c and two with MSA-p. Initially, Smith monitored uptake of the tracer continuously over 90 minutes. He found a favorable kinetic profile, with the tracer sticking around in the brain long enough to decipher its retention, and then washing away effectively thereafter.

A signal leapt out of the cerebellum in all people with MSA, Smith reported. In agreement with the distribution of α-synuclein pathology in this disorder, Smith saw tracer retention in the cerebellar white matter and a stalk-shaped structure called peduncles, but not in cerebellar gray matter. Though people with either MSA subtype evinced binding in these regions, it was greater, on average, in those with MSA-c. Tracer uptake in cerebellar white matter completely distinguished people with MSA from controls and from those with other synucleinopathies, none of whom had significant uptake in this region.

Smith spotted some unspecific binding in the pons (see image below). This signal was less visible in MSA due to atrophy in the region and to masking by the stronger signal in the cerebellum. It is more easily seen in PET scans overlaid on MRI scans, noted Hansson. He and Smith noted that this unspecific binding was relatively weak, and did not detract from the specific signal emanating from the cerebellar white matter.

Finally, α-Synuclein PET? 18F-ACI-12589 bound cerebellar white matter and cerebellar peduncles (white arrows) in people with MSA (bottom) but not in healthy controls (top). The green blob in controls reflects unspecific binding in the pons, which is masked by the stronger specific signal in MSA cases. [Courtesy of Ruben Smith, Lund University.]

Alas, in people with PD, PDD, or DLB, there was nothing much to see. Smith did spot some uptake in the basal ganglia of a person with PD as well as in a person with MSA-p, but some controls also had this signal, suggesting it was non-specific. Similarly, three MSA-p patients had strong tracer binding in the globus pallidus, but so did one control. In all, the only disease-specific signals arose from the cerebellar white matter and peduncles in people with MSA.

Quantity or Quality? Researchers at AD/PD were excited that a tracer appears to work in MSA, but their enthusiasm was tempered by the lack of a signal in people with the more common synucleinopathies. “Why do you think it didn’t work in PD?” asked session co-chair Tamara Shiner of Tel Aviv University. Smith was unsure, but said that given the small number of people scanned so far, including only seven with PD, it might be too soon to close the book on this tracer’s performance in other synucleinopathies.

Among the main α-synucleinopathies, MSA progresses faster and comes with more Lewy bodies than do the others, suggesting the tracer’s affinity may simply be too low to pick up all deposits.

Conformation of α-synuclein aggregates could also play a role, Hansson and Smith believe, as the fibrils are known to twist into distinctive shapes in each synucleinopathy (Mar 2020 conference news; Tarutani et al., 2018). Curiously, though, the tracer bound well to α-synuclein in brain sections from people with these different disorders, suggesting it was capable of latching onto multiple forms. Reaching a target in the brain of a living person is a far higher bar for a tracer than sticking to it in a brain section.

Robert Mach of the University of Pennsylvania, Philadelphia, heads a massive multi-institutional effort, called the Center without Walls for Imaging Proteinopathies with PET, to develop tracers for α-synuclein and 4R-tau. He favors the idea that differences in structure best explain why the tracer bound in MSA and not in other synucleinopathies. He noted that a wealth of data, from binding of fluorescent probes and tracers in his own work to antibody binding experiments and cryo-EM, support the idea that structural differences in α-synuclein influence the binding of ligands. As to why the tracer bound to multiple forms of α-synuclein in tissue sections, but not in vivo, Mach said that tissue sections are subjected to biochemical procedures that might expose binding sites that are normally cloaked in the human brain.

Chet Mathis of the University of Pittsburgh is a co-developer of PiB and has been searching for an α-synuclein tracer for years. He called the findings a definite advance for the field, noting that while past α-synuclein tracer candidates had bound to α-synuclein in the brains of people with MSA, this is the first one that does so specifically, without off-target binding to Aβ or tau. However, he considers the lack of binding in people with PD or DLB a disappointment that likely boils down to the tracer’s low affinity.

Capotosti reported that the tracer 's dissociation constant was between 8 to 30 nM for α-synuclein aggregates in tissue slices and brain homogenates from different α-synucleinopathies. This seemed shockingly weak to Mathis, who said most researchers agree the sweet spot for tracer affinity is 1 nM or lower. Oddly, the tracer’s affinity for brain-derived α-synuclein in vitro was weaker for MSA than it was for PD. Capotosti reported a Kd of 17 nM for tracer binding to α-synuclein in a frontal cortex section of a person with PD, compared to 30 nM in a cerebellum section from a person with MSA. Similarly, the tracer bound with Kd values of 8 nM and 22 nM in brain homogenates from people with PD and MSA, respectively.

“The surprising thing is that it does work in MSA despite this low affinity,” Mathis said. For whatever reason—be it abundance, location, or conformation—MSA α-synuclein appears to be more forgiving than PD α-synuclein in terms of the tracer affinity required to detect it, Mathis added.

Like Mathis, Mach was also surprised that a tracer with such a low affinity could work. On the one hand, those findings are highly encouraging, he said, because they mean that tracer candidates need not eclipse a 1 nM affinity before being put to the test in clinical studies. On the other hand, he said that the results beg a major question: “What are the properties of the ligand that enabled it to succeed where others have failed?”

Hansson told Alzforum he plans to submit a detailed manuscript soon. Capotosti said efforts are underway to test and optimize 18F-ACI-12589 in people with MSA and other synucleinopathies, and to try other promising compounds from their Morphomer library that may work for PD.

Progress on other α-synuclein tracers was reported at AD/PD, as well. Felix Schmidt of MODAG, a biotech company based in in Wendelsheim, Germany, and funded in part by MJFF, reported specific binding of their compound, MODAG-005, in tissue sections from people with PD and MSA. This tracer bound to brainstem and midbrain sections in people with PD, as well as cerebellar white matter in people with MSA. In mice, C11-MODAG-005 entered the brain and labeled recombinant α-synuclein fibrils that had been previously injected.

Ruiqing Ni, of the University of Zurich also reported promising data from α-synuclein tracer candidates in animal models. Specifically, Ni reported that one was taken up in the striata of M83 mice, a model for PD, but not in models with amyloid or tangle accumulation.—Jessica Shugart


  1. These very exciting pilot data demonstrate proof-of-concept for imaging α-synuclein in MSA. If verified in larger samples, this would represent a landmark moment, similar to first reports of imaging amyloid with PiB and tau with FTP (at the time T807).

    The selectivity of binding to synuclein over Aβ and tau is encouraging and unique compared to other synuclein imaging leads that I am familiar with. It will be important to better understand the patterns of in vivo AC-12589 binding across the spectrum of synucleinopathies. Even though the preclinical data hint at a potential pan-synuclein marker, the in-vivo results imply selectivity for MSA over PD/DLB. It is unclear if these results are due to a higher concentration of pathology in cerebellar white matter in MSA, a lower affinity of the tracer for synuclein deposits in PD and DLB, or a combination of these factors.

    Early cryo-electron microscopy work suggests that synuclein fibrils in MSA may be structurally distinct from DLB fibrils, though the latter have not yet been definitively characterized to my knowledge (Schweighauser et al., 2020). If correct, one would in fact hypothesize that PET ligands would have differential affinity for aggregates in different synuclein disorders, as we have observed across the different tauopathies.

    Overall, while preliminary, these are exciting data, and I am cautiously optimistic that the field may have its first foothold into synuclein imaging.


    . Structures of α-synuclein filaments from multiple system atrophy. Nature. 2020 Sep;585(7825):464-469. Epub 2020 May 27 PubMed.

  2. Finding ways to image α-synuclein pathology in the human brain is a critical need in the neurodegenerative field, as this could allow for both evaluation of disease progression and, potentially, for differential diagnosis between diseases. For example, confirming or refuting ideas about the spread of pathology between different brain regions as diseases such as Parkinson’s progress could be much more informative and detailed if the same set of individuals could be scanned at different time points. Additionally, having an imaging-based approach could be helpful to establish target engagement in the brains of candidate therapeutic agents.

    In this context, the results reported at AD/PD from AC immune on their new tracer are exciting. The data presented suggest that the compound detects pathological synuclein in tissue, with reasonable ability to discriminate against other proteins that may aggregate in the brain, namely tau and Aβ, especially in the context of mixed pathology.

    What is a bit puzzling so far is that the signal in vivo seems stronger for MSA than for PD. Several possible explanations come to mind. An obvious one is that the structure of α-synuclein aggregates in MSA and PD are different, and this is supported by reported differences in the seeding propensity of MSA and PD α-synuclein “strains.” Alternatively, the sensitivity of the probe in vivo may be sufficient to pick out MSA but not PD due to a higher concentration of pathology in the former. Distinguishing between these two possibilities (and others that I have not addressed here) will be important, because development of further generations of tracers will depend on whether we need to aim for discrimination between MSA and PD or to simply have more sensitive probes for synucleinopathy in general.

    Nonetheless, these first signals of synuclein pathology in living humans are potentially useful right now in terms of differential diagnosis for MSA, where we expect the clinical course to be very different from PD, so being able to “see” that very early is a practical advance.

Make a Comment

To make a comment you must login or register.


News Citations

  1. PET Tracer PI-2620 Detects 4R Tau Deposits
  2. Behold the First Human α-Synuclein CryoEM Fibril Structure

Paper Citations

  1. . Alpha-Synuclein PET Tracer Development-An Overview about Current Efforts. Pharmaceuticals (Basel). 2021 Aug 26;14(9) PubMed.
  2. . α-Synuclein Radiotracer Development and In Vivo Imaging: Recent Advancements and New Perspectives. Mov Disord. 2022 May;37(5):936-948. Epub 2022 Mar 15 PubMed.
  3. . α-synuclein imaging: a critical need for Parkinson's disease research. J Parkinsons Dis. 2013;3(4):565-7. PubMed.
  4. . A Proposed Roadmap for Parkinson's Disease Proof of Concept Clinical Trials Investigating Compounds Targeting Alpha-Synuclein. J Parkinsons Dis. 2019;9(1):31-61. PubMed.
  5. . Potent prion-like behaviors of pathogenic α-synuclein and evaluation of inactivation methods. Acta Neuropathol Commun. 2018 Apr 18;6(1):29. PubMed.

Further Reading

No Available Further Reading