12 Oct 2018

A new model of Parkinson’s disease re-creates key features of the disorder, and helps steady one disputed theory of what causes it. Mice carrying mutations that disrupt physiological tetramers of the α-synuclein protein develop brain pathology and neurodegeneration typical of PD, according to work from the lab of Dennis Selkoe at Brigham and Women’s Hospital in Boston. Animals as young as three months had trouble walking and endured spontaneous whole-body tremors. As in people, the symptoms were worse in males and responded partially to treatment with the drug L-DOPA. The mice offer a more complete picture of PD than existing models, said Selkoe, which bodes well for their use in testing potential treatments. The work supports the hypothesis, formulated in Selkoe’s lab, that α-synuclein toxicity stems from destabilization of physiological tetramers and accumulation of aggregation-prone monomers. The work appeared October 10 in Neuron.

“We have been trying to find evidence that the tetramer/monomer ratio matters," said Selkoe. “These mice, with their consistent phenotype of Parkinson features—the sex difference, a resting tremor that gets worse, gait abnormalities, and the L-DOPA response—support the hypothesis that shifting tetramers to monomers is adverse,” he told Alzforum.

“The study is very exciting, and the authors need to be commended for their relentless pursuit of this unconventional (and once-unpopular) idea," wrote Subhojit Roy, University of Wisconsin, Madison.

Other researchers also praised the work, and the new model. “This is an extremely elegant study in which the authors rigorously assess the phenotype of a transgenic mouse model in which the α-synuclein tetrameric form is destabilized. The phenotype seems more striking than observed in other mouse α-synuclein models and the partial response to L-DOPA treatment also suggests that this model may mimic human disease,” wrote Dario Alessi, University of Dundee, Scotland.

Mutations in α-synuclein cause Parkinson’s disease, dementia with Lewy bodies, and other synucleinopathies, which are all characterized by the accumulation of aggregates of the protein in the brain. A cytoplasmic resident that normally regulates trafficking of presynaptic vesicles in neurons, α-synuclein exists in cells in two forms, according to Selkoe’s hypothesis: a stable, helically folded tetramer, and free monomers (Aug 2011 news; Oct 2014 news). Mutations in α-synuclein that cause familial PD destabilize the tetramer structure in cells, and the theory holds that shifting the ratio between tetramers and aggregation-prone monomers kicks off pathological protein accumulation, and ultimately, cell death (Apr 2015 conference news). 

In the new work, first author Silke Nuber parlayed tetramer-destabilizing mutations into a unique mouse model of PD. Previously, the lab had identified a series of six-residue KTKEGV repeats in α-synuclein that hold tetramers together (Dettmer et al., 2015). The PD mutation, E46K, in the middle of one repeat, was sufficient to break up tetramers, but its effects were amplified, they found, by making two additional analogous mutations in adjacent repeats. While one E46K decreased the ratio of tetramers to monomers in cells by 40 percent, the E35K/E46K/E61K triple mutant, which they call 3K, reduced that ratio by more than 90 percent, and triggered more severe α-synuclein aggregation and toxicity in cells.

What would the mutation do in mice? To find out, Nuber generated transgenic mouse strains expressing human α-synuclein with the triple mutant (3K) or wild-type human α-synuclein. To avoid overexpression artifacts, Nuber selected transgenic strains with α-synuclein levels similar to those found in human brain, and compared the animals to an existing strain expressing the E46K mutant (1K) made by Elan Pharmaceuticals. In agreement with studies in cells, the mutations decreased tetramers, while monomers became more plentiful. The ratio of tetramers to monomers in 1K brain dropped by 60 percent, while in the 3K mice, the reduction was greater than 90 percent, and was detected in multiple brain regions. In 3K brain, more monomers ended up in the insoluble fraction, and some of them appeared to be truncated, similar to what has been seen in PD brain.

The reduction in tetramers was accompanied by dramatic motor changes starting at an early age. While most PD models begin to show motor impairment at eight to 12 months, by three months the 3K mice already had developed a spontaneous head and body tremor, which gradually worsened with age. This type of tremor is not seen in other PD models but resembles the resting tremor of PD patients, Selkoe said. Their coordination also suffered: The mice had trouble climbing a pole, or balancing on a rotating rod. By six months, the animals developed a stiff gait and moved around less in their cages. Symptoms appeared at a younger age and were more robust in male mice than female, which is also true of people with Parkinson’s.

These behavioral changes tracked with early signs of brain pathology. At three months, the 3K mice harbored protease-resistant, truncated, and Ser129-phosphorylated α-synuclein deposits. Some of the α-synuclein organized into vesicle-associated, lipid-rich aggregates that grew over time and might be precursors to Lewy bodies, claimed the authors. Indeed, in 16-month-old mice, Nuber did detect rare Lewy body-like inclusions. In the 1K mice, pathological changes accrued more slowly and never reached the same severity, consistent with the milder phenotype reported for these mice.

The 3K mutations also triggered loss of dopaminergic neurons, the hallmark of PD. Six-month-old mice had 27 percent fewer dopaminergic cells than wild-type α-synuclein controls, and produced39 percent less dopamine in the striatum. Boosting the neurotransmitter with a single dose of L-DOPA improved the animals’ performance: They climbed and hung onto a pole more easily, and the fluidity of their walking especially improved. The single shot of L-DOPA did not improve the tremor, or their ability to walk on the rotating rod, which suggested to the authors that these phenotypes might be due to cortical neurodegeneration.

“The model, which recapitulates more cardinal features of PD than published PD models, supports the novel mechanistic insight that interfering with physiological α-synuclein tetramers can lead to PD,” wrote Hanseok Ko, Johns Hopkins University in Baltimore. Ko sees the model as an important step toward better animal models of PD for mechanistic and therapeutic studies. He and Ted Dawson, also at Johns Hopkins, recently reported that tetramer instability contributes to α-synuclein toxicity that comes along with mutations in glucocerebrosidase 1 (GBA1), the most common genetic risk factor for idiopathic PD (Kim et al., 2018). Restoring tetramers by either overexpressing GBA1 or lowering levels of toxic lipids that accumulate in mutant cells prevented α-synuclein toxicity in their cell models.

If the tetramers are critical to α-synuclein function, then compounds that stabilize tetramers might be good treatments, Selkoe said. That’s the idea behind an FDA-approved treatment for familial amyloid polyneuropathy, which promotes tetrameric, non-amyloidogenic assembles of the transthyretin protein (Aug 2011 news). Selkoe hopes to replicate that approach for α-synuclein. He told Alzforum that his lab has identified commercially available small molecules that restore the tetramer/monomer ratio in cells bearing α-synuclein mutations, and is now testing those in the animal model.

Clinically, the E46K mutation associates not only with motor symptoms but also with dementia in Parkinson’s patients or dementia with Lewy bodies (DLB). While the authors highlight PD symptoms in the paper, Selkoe said the 3K mice likely also model DLB. Nuber explained that it was challenging to measure cognition in the mice, because many tests rely on animals walking or swimming well. They are now testing the mice in cognitive tasks, such as object recognition, that don’t rely so much on movement.

“This is impressive work but I don't think yet the model has been phenotyped sufficiently to recommend it as the PD animal model everyone should be using,” wrote Clive Ballard, University of Exeter. “I think we need more information re. transcriptional changes and the long-term progression of pathology and behavioral changes. We also need more understanding of how this impacts on other aspects of the disease that may be important—mitochondrial dysfunction, clearance of synuclein, neuroinflammation, etc.—and the pros and cons compared to other emerging PD models such as GBA mutations.” Different models may be called for, for different types of studies, he wrote.

Selkoe told Alzforum they are planning some of those studies. They will be tracking the progression of α-synuclein pathology over time, and since several lines of evidence point to an important role of lipids in tetramer stabilization, lipidomics approaches are on the table, too. Nuber told Alzforum she is beginning to explore the reasons for the prominent sex differences.

What about other α-synuclein mutations? The lab previously found that, among the five missense mutations that cause PD, G51D, which lies outside of the repeat domain, has the strongest tetramer-abrogating effect in cells. They want to see what that mutation will do in mice, too.

On a technical note, all the studies of 3K mice used heterozygotes, carrying just one copy of the mutated gene. Because of the young onset and exaggerated tremor symptoms, male mice had trouble breeding. With some effort, Nuber got the mice to produce one litter of homozygotes, but they had a lethal motor phenotype. The mice developed whole-body tremor shortly after birth and couldn’t get to food or water on their own. By four weeks, they had severe pSer129-postive α-syn aggregates in their brains.

Instead, Nuber mated female heterozygous mutants with wild-type males, and genotyped the offspring to identify mutation carriers. For ease of future work, she generated a lower-expressing 3K homozygous line, with intermediate pSer129 pathology, and later onset of less-dramatic motor symptoms. These mice will be easier to maintain for testing treatments, for example. Selkoe said they plan to make both of the 3K strains available to researchers through Jackson Labs in Bar Harbor, Maine.—Pat McCaffrey

Comments

1. • Quyen Hoang
     Indiana University School of Medicine
   • Posted: 12 Oct 2018

   The mouse model presented by Nuber et al. is remarkable in that the patho-histology and the motor phenotypes closely resemble all the major hallmarks of Parkinson’s disease. Thus, it appears to be a good animal model for the disease. More broadly, their findings strongly support that the dynamic interconversion between monomeric and tetrameric α-synuclein is important for its function and that losing this ability results in aggregation and toxicity in mice. 

   Studying the tetrameric form of α-synuclein has been difficult due to its dynamic and transient nature, which is commonly observed in small proteins that are highly charged and mostly helical (such as osteocalcin for example, Hoang et al., 2003); however, this mouse model might enable investigations of α-synuclein-associated toxic mechanisms and the monomer-tetramer dynamics without the need to isolate the tetramers.

   Given the results, I wonder what happens when a persistently tetrameric form of α-synuclein is expressed in mice and whether it could rescue the effects observed in the 3K mice.

   References:

   Hoang QQ, Sicheri F, Howard AJ, Yang DS. Bone recognition mechanism of porcine osteocalcin from crystal structure. Nature. 2003 Oct 30;425(6961):977-80. PubMed.

   View all comments by Quyen Hoang
2. • Dario Alessi
     Medical Research Council Protein Phophorylation and Ubiquitylation Unit
   • Posted: 12 Oct 2018

   I think that this is an extremely elegant study in which the authors rigorously assess the phenotype of transgenic mouse model in which the synuclein tetrameric form is destabilized. The phenotype seems more striking than observed in other mouse synuclein models and partial response to L-DOPA treatment also suggests that this model may mimic human disease. There has been a lot of discussion about the interplay of LRRK2 and synuclein biology. It would be fascinating to cross the new synuclein transgenic mouse model with LRRK2 pathogenic knock-in mutations or VPS35[D620N] mice that also display high LRRK2 pathway activity, to see if these exacerbates the phenotype or time that phenotype emerges.

   It would also be fascinating to treat the new synuclein mouse model with LRRK2 inhibitors both before and after the phenotypes are observed to see if there is any delay and/or amelioration in phenotype. It would also be intriguing to test whether LRRK2 pathway is activated in brain tissues of the new synuclein transgenic mice, which could be achieved by studying Rab protein phosphorylation by immunofluorescence. Such studies would help provide further insight into whether LRRK2 contributes to disease effects of synuclein aggregation. I hope that these mice will be easy for researchers to access!

    

   View all comments by Dario Alessi
3. • Subhojit Roy
     University of California, San Diego
   • Posted: 12 Oct 2018

   This study by Silke Nuber, Ulf Dettmer, Dennis Selkoe and colleagues explores a key idea that they have championed—that α-syn normally exists as a multimer/tetramer, but abrogation of these physiologic conformations leads to an increase in monomers, aggregation of these free monomers, and subsequent pathology. Here they generated a new mouse model (“3K line”) where they added three E->K mutations in the protein (the human E46K mutation, and two other E->K mutations that decrease α-syn multimers/tetramers—based on their own previous studies). The authors convincingly show that addition of these three mutants—and consequent decrease in α-syn multimers—leads to an increase in monomers as well as an increase in insoluble α-syn fractions; increase in pathologic signatures like α-syn phosphorylation at defined residues (Serine 129); motor deficits; and degeneration of nigral neurons. Though the 3K mutations are not disease-related, they are an elegant tool to explore the authors’ hypothesis in a true in vivo setting.

   A few things were unclear. Since the E46K mutation is associated with human disease and is itself pathologic, it is important to show that pathology—and associated motor deficits—in the 3K mice is more pronounced than the E46K mutation alone. Though the authors did use mice expressing only the E46K mutation (“1K”) for some experiments, it was not clear if these were used to examine three key features: α-syn insolubility, dopaminergic degeneration, and motor deficits. Regarding the immuno-EM studies, the conclusion is that in the 3K mice, α-syn monomers are preferentially associated with synaptic vesicles. However, a large number of the α-syn gold particles are not on synaptic vesicles (for example, there is only ~ 1 immunogold particle per vesicle in the 3K boutons, whereas a visual inspection suggests that there are numerous scattered particles throughout). So it may be best to interpret these data cautiously. Also, the number of gold particles is higher in 3K to begin with—compared to WT—so it is not clear if the increase in vesicle association is specifically due to an increase in α-syn monomers (following a loss of tetramers). The 1K control is also missing here.

   Finally, it is unclear if tetramers are the only higher-order α-syn species, as advocated by the authors. Though tetramers are certainly the predominant species in the M17D cells, a glance at the literature suggests that the conformations are cell-type dependent and not necessarily tetrameric. Thus, “multimers” might be a better usage until this issue is completely nailed down in neurons and synapses. Despite these comments, the studies by Nuber et al. are very exciting, and the authors need to be commended for their relentless pursuit of this unconventional (and once-unpopular) idea. Further studies may further clarify some of the mechanistic steps involved. 

   View all comments by Subhojit Roy
4. • Hanseok Ko
     Johns Hopkins University
   • Posted: 12 Oct 2018

   This model, which recapitulates more cardinal features of PD than published PD models, supports the novel mechanistic insight that interfering with physiological α-synuclein tetramers can lead to PD. This mechanism is not widely recognized in the PD field, and should be confirmed in sporadic PD brain. Some issues including artificial effects and less degeneration of dopaminergic neurons than expected (30 percent) remain to be resolved. The new model would be an important step toward better animal models of PD and can be used for future mechanistic and therapeutic studies by PD and related α-synucleinopathy researchers.

   View all comments by Hanseok Ko
5. • Clive Ballard
     The University of Exeter
   • Posted: 12 Oct 2018

   This is a fascinating finding which builds on the authors’ previous work highlighting the importance of an increase in the ratio of -synuclein monomers to tetramers as a key mechanism accelerating progressive changes.

   They show in these very careful studies with their mouse model that the ratio of monomer to tetramer is significantly altered and that this is associated with clear toxic effects and leads to a behavioral phenotype—abnormal motor movements that are L-dopa responsive.

   This is impressive work, but I don't think yet the model has been phenotyped sufficiently to recommend it as the PD animal model everyone should be using. I think we need more information re. transcriptional changes and the long-term progression of pathology and behavioral changes. We also need more understanding of how this impacts on other aspects of the disease that may be important—mitochondrial dysfunction, clearance of synuclein, neuroinflammation, etc., and the pros and cons compared to other emerging PD models, such as GBA mutations, which may be different for different types of studies.

   Whilst this may become a key PD model—particularly for specific therapy development re. stabilizing tetramers and probably synuclein immunotherapy—I don't think we're quite there yet.

   View all comments by Clive Ballard

Make a Comment

To make a comment you must login or register.

References

News Citations

1. An α-Synuclein Twist—Native Protein a Helical Tetramer 19 Aug 2011
2. Synuclein Oligomers: Is EnSNAREing Synaptic Vesicles Their True Calling? 3 Oct 2014
3. Form and Function: What Makes α-Synuclein Toxic? 10 Apr 2015
4. Research Brief: Novel Drug Candidates for Amyloid Cardiomyopathy 26 Aug 2011

Paper Citations

1. Dettmer U, Newman AJ, von Saucken VE, Bartels T, Selkoe D. KTKEGV repeat motifs are key mediators of normal α-synuclein tetramerization: Their mutation causes excess monomers and neurotoxicity. Proc Natl Acad Sci U S A. 2015 Aug 4;112(31):9596-601. Epub 2015 Jul 7 PubMed.
2. Kim S, Yun SP, Lee S, Umanah GE, Bandaru VV, Yin X, Rhee P, Karuppagounder SS, Kwon SH, Lee H, Mao X, Kim D, Pandey A, Lee G, Dawson VL, Dawson TM, Ko HS. GBA1 deficiency negatively affects physiological α-synuclein tetramers and related multimers. Proc Natl Acad Sci U S A. 2018 Jan 23;115(4):798-803. Epub 2018 Jan 8 PubMed.

Other Citations

1. PD models

Further Reading

Papers

1. Ruskey JA, Zhou S, Santiago R, Franche LA, Alam A, Roncière L, Spiegelman D, Fon EA, Trempe JF, Kalia LV, Postuma RB, Dupre N, Rivard GE, Assouline S, Amato D, Gan-Or Z. The GBA p.Trp378Gly mutation is a probable French-Canadian founder mutation causing Gaucher disease and synucleinopathies. Clin Genet. 2018 Oct;94(3-4):339-345. Epub 2018 Jul 16 PubMed.