When good proteins go bad in neurodegenerative diseases, the road to ruin often runs through modification and via aggregation to toxicity. The α-synuclein protein, a causative force in Parkinson disease and the main component of Lewy body inclusions, is no different. Scientists share a growing sense that phosphorylation of α-synuclein plays a role in its sinister effects, but they are in the dark about exactly how that works. One residue, serine129, is heavily phosphorylated in Lewy bodies, but studies have disagreed about whether this phosphorylation promotes or prevents neurotoxicity. In the March 3 issue of the Journal of Neuroscience, Hilal Lashuel and colleagues at the Swiss Federal Institute of Technology in Lausanne have identified another serine, at position 87, that is also phosphorylated in Lewy body disease inclusions and in Alzheimer disease. The researchers provide evidence that modifications at serine 87 affect both α-synuclein’s pathological aggregation and its normal interaction with cell membranes.

Along the way, the scientists strengthen a case they have been making for the past two years that, when it comes to α-synuclein, the widely used approach of phosphomimetic mutations may not always paint an accurate picture of the role of protein phosphorylation in vivo. More broadly, the results make clear that phosphorylation is apt to affect α-synuclein’s function on many levels besides of its tendency to aggregate. Because of all that, Lashuel told ARF, it is too early to characterize α-synuclein phosphorylation in terms of good or bad. “If you ask me today how phosphorylation affects α-synuclein toxicity, I will tell you I don’t know,” he said. “We will need the right tools and detailed studies to answer that question.” Both academia and industry are grappling with this issue at present.

To complicate matters further, a second paper in the same journal suggests that not all α-synuclein aggregates are created equal, and the location of oligomer formation in the brain may be important for toxicity. Joseph Mazzulli of the Massachusetts General Hospital, Charlestown, and Harry Ischiropoulos of the Children’s Hospital of Philadelphia, Pennsylvania, report finding soluble α-synuclein assemblies throughout the brain in synuclein transgenic mice. The catch: Only assemblies/oligomers from inclusion-bearing regions had the ability to promote further aggregation and kill neurons. These poisonous oligomers were more oxidized, suggesting that region-specific modifications determine toxicity.

In the first study, the potential for phosphorylation of α-synuclein at serine 87 (S87) interested Lashuel and coworkers for several reasons. First, this serine resides in a hydrophobic stretch in the C-terminus; this place is important for aggregation, and the addition of a charged phosphate group there might dramatically affect protein structure and oligomerization. Second, S87 is one of only five residues that differ between human and rat or mouse, two rodent species in whom the protein forms no aggregates. Finally, the residue was known to be phosphorylated in cells in culture.

To look for the S87 modification in brain tissue, first authors Katerina Paleologou and Abid Oueslati generated a phospho-S87 antibody. With that, they saw elevated reactivity in brain extracts from cases of Alzheimer disease, dementia with Lewy bodies (DLB), multiple system atrophy, and from several mouse models of synucleinopathy. In all cases, most of the modified protein was associated with membranes. Fluorescent immunostaining of fresh human brain revealed S87 phosphorylation in Lewy bodies from DLB cases, though at a much lower level than the predominant phosphorylation at S129.

In-vitro studies with α-synuclein showed that phosphorylation at S87 inhibited aggregation. To translate that finding in vivo, the investigators initially turned to phosphomimic mutants, a standard technique for such studies. This approach involves replacing serine with an acidic glutamate (E) or aspartate (D) residue to simulate phosphorylation, or with alanine to prevent phosphorylation. Phosphomimics have been used to study S129 in vivo, but have yielded conflicting results. Some of the confusion likely stems from species differences: Work in Drosophila indicates the S129D phosphomimic enhances toxicity (Chen and Feany, 2005), while in rats the mutant produces either the same or less neurodegeneration than does wild-type protein (Gorbatyuk et al., 2008; Azeredo da Silveira et al., 2009; McFarland et al., 2009). Moreover, previous work from the Lashuel lab indicated that structural properties of the S129E/D mimetic mutants measured in vitro differ from bona fide phosphorylated protein (Paleologou et al., 2008), raising questions about how well the substitutions truly recapitulate phosphorylation.

With that in mind, the researchers proceeded cautiously with the S87 mutants. They found that in aggregation assays, the non-phosphorylatable S87A behaved like wild-type protein, while S87E was slower. These results suggest that the S to E substitution reproduces the effects of phosphorylation, at least on aggregation, and could be used to probe aggregation in vivo. However, there was a catch. In structural studies, the investigators found that phosphorylation of S87 altered the shape of the α-synuclein monomer—as measured by an increased hydrodynamic radius—much more than did the S to E substitution. In addition, NMR of membrane-bound synuclein showed that phosphorylation at S87 and S129 resulted in changes in protein conformation and membrane association that were not recapitulated by an S87E mutant.

These results highlight the pitfalls of using phosphomimic mutants, Lashuel told ARF. “We see that although the S87E mutant can reproduce the effect of phosphorylation on aggregation, it does not reproduce the effect of phosphorylation on membrane binding. And we are only looking at two phenomena. A problem arises when people take this to mimic phosphorylation. It may reproduce one aspect but not another, and we don’t know which aspect is important.”

The solution, said Lashuel, lies in more detailed studies with the actual phosphorylated proteins to define the effects of phosphorylation on a range of synuclein structural and functional outcomes. “While the mimics can be informative, they should be interpreted with caution. It is not until we can actually significantly perturb the level of synuclein phosphorylation in vivo that we will be able to answer these questions,” he noted. Toward that end, the Lashuel group is working to identify all of α-synuclein’s phosphorylation sites, as well as the kinases and phosphatases that modify them, so that the scientists can build in-vivo models that depend on changing the state of protein phosphorylation. Similar comprehensive characterizations of all phosphorylation sites were done with the protein tau, as well (e.g., Cripps et al., 2006).

Recently, the Lashuel lab confirmed an earlier finding by scientists at Elan Pharmaceuticals (Inglis et al., 2009) that polo-like kinases (PLKs) are the major catalysts for S129 phosphorylation in vivo (Mbefo et al., 2010). Interestingly, last year Susan Lindquist’s lab at MIT reported that, among other things, the yeast ortholog of human PLK2 suppresses α-synuclein toxicity in yeast, as well as in worm and rat neurons (see ARF related news story on Gitler et al., 2009). The paper by Mbefo et al. shows that S129 phosphorylation blocks aggregation in vitro. Nonetheless, Lashuel insisted, phosphorylation is an important regulator of many interactions, so focusing solely on aggregation is likely to miss important insights. For example, a recent proteomics paper indicates that the phosphorylation of S129 or Y125 completely changes the constellation of proteins that associate with α-synuclein compared to the dephosphorylated form (McFarland et al., 2008).

The complexity of the modifications may be another source of confusion. Mel Feany and colleagues at Harvard Medical School recently identified tyrosine 125 (Y125) as an additional site of phosphorylation on α-synuclein in human brain tissue, and showed that phosphorylation there decreases with age (Chen et al., 2009). In this group’s fly model of human synuclein expression, mutating the tyrosine to a non-phosphorylatable phenylalanine residue accelerated neurodegeneration and loss of motor skills. Increasing phosphorylation of the protein by co-expressing the tyrosine kinase called shark reduced toxicity, opposite of the S129 effect. Overall, the results in flies suggest that phosphorylation at S129 promotes toxicity, while modification at Y125 blocks it. In all cases, toxicity correlated with the appearance of soluble α-synuclein oligomers, and not large aggregates. These authors argue that the loss of Y125 phosphorylation with age may predispose to α-synuclein toxicity in human disease.

Companies are taking note of modifications of α-synuclein with an eye to therapy development. This was apparent in a “synuclein summit” hosted in January by the Michael J. Fox Foundation in New York City. At this meeting, Jennifer Johnston of Elan Pharmaceuticals characterized α-synuclein as a validated target and decreasing α-synuclein as a validated pathway to a drug. However, how to get there is not clear. Echoing Lashuel’s caution, Johnston indicated that industry most likely would “wait and see” what comes of more research on modifications of α-synuclein (view her slides).

Martin Ingelsson of the Rudbeck Laboratory in Uppsala, Sweden, who was not involved in the work, compared α-synuclein with tau, another pathogenic protein that undergoes phosphorylation at multiple sites having complex effects on protein function. He pointed out that for tau, the ability to form stable microtubules seems to decrease if certain sites are phosphorylated and increase if other sites are phosphorylated. “At first, it looked as if there were only one or two relevant phosphorylation sites in the α-synuclein molecule, so my hope was that this was not as complicated as with tau,” Ingelsson said. “Now these additional studies are revealing novel phosphorylation sites, so it could be that phosphorylation of certain sites could increase aggregation, whereas other sites are decreasing aggregation. It’s very hard to tell at this point.”

Ingelsson had praise for the Mazzulli and Ischiropoulos study of soluble oligomers in mouse brain. “The work adds evidence to already existing data that oligomers are truly existing in the mice and that they are causing problems,” he told ARF. In that study, first author Elpida Tsika and coworkers characterized oligomers from the brain of transgenic mice expressing the A53T mutant of α-synuclein. They found detergent-soluble oligomers in different brain regions, regardless of whether inclusions were also present or not. The oligomers from different regions did share similar basic biochemical properties (they were SDS, heat, and urea stable, but sensitive to proteinase K digestion), but oligomers that came from regions with inclusions were more modified, as indicated by reaction with a monoclonal antibody that recognizes oxidized or nitrated α-synuclein (syn303, Duda et al., 2002). The oxidized aggregates potently seeded α-synuclein oligomerization, and were more toxic to neurons in culture. Oligomers from regions of the brain that had no inclusions showed fewer oxidative modifications, were non-toxic, and did not seed α-synuclein aggregation as efficiently in vitro. The results suggest to the authors that not all oligomers are on the path to forming inclusions. This, they write, could explain the regional specificity of α-synuclein, if cell-intrinsic factors act in concert with oligomers to bring cells down.—Pat McCaffrey

Comments

  1. I'd like to post a correction to a mistake in this article: the difference in amino acid sequence of alpha-synuclein between human and rat/mouse is 7, not 5. These include A53T, S87N, L100M, N103G, A107Y, D121S/G, and N122S. The sequences were from GenBank: human, NP000336.1; rat, NP062042.1; mice, NP033247.1

  2. I have never found the concept of mimicking phosphoserine with glutamate convincing. Glutamate lacks an oxygen and cannot have a doubly negative charge, unlike phosphoserine. Aspartate, which is often also used to replace phosphoserine, is even worse, as it is one bond shorter. If the protein or peptide is short enough, it can be synthesized using a phosphorylated amino acid.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Microarrays Probe Protein Interactions

Paper Citations

  1. . Alpha-synuclein phosphorylation controls neurotoxicity and inclusion formation in a Drosophila model of Parkinson disease. Nat Neurosci. 2005 May;8(5):657-63. PubMed.
  2. . The phosphorylation state of Ser-129 in human alpha-synuclein determines neurodegeneration in a rat model of Parkinson disease. Proc Natl Acad Sci U S A. 2008 Jan 15;105(2):763-8. PubMed.
  3. . Phosphorylation does not prompt, nor prevent, the formation of alpha-synuclein toxic species in a rat model of Parkinson's disease. Hum Mol Genet. 2009 Mar 1;18(5):872-87. PubMed.
  4. . Alpha-synuclein S129 phosphorylation mutants do not alter nigrostriatal toxicity in a rat model of Parkinson disease. J Neuropathol Exp Neurol. 2009 May;68(5):515-24. PubMed.
  5. . Phosphorylation at Ser-129 but not the phosphomimics S129E/D inhibits the fibrillation of alpha-synuclein. J Biol Chem. 2008 Jun 13;283(24):16895-905. PubMed.
  6. . Alzheimer disease-specific conformation of hyperphosphorylated paired helical filament-Tau is polyubiquitinated through Lys-48, Lys-11, and Lys-6 ubiquitin conjugation. J Biol Chem. 2006 Apr 21;281(16):10825-38. PubMed.
  7. . Polo-like kinase 2 (PLK2) phosphorylates alpha-synuclein at serine 129 in central nervous system. J Biol Chem. 2009 Jan 30;284(5):2598-602. PubMed.
  8. . Phosphorylation of synucleins by members of the Polo-like kinase family. J Biol Chem. 2010 Jan 22;285(4):2807-22. PubMed.
  9. . Alpha-synuclein is part of a diverse and highly conserved interaction network that includes PARK9 and manganese toxicity. Nat Genet. 2009 Mar;41(3):308-15. PubMed.
  10. . Tyrosine and serine phosphorylation of alpha-synuclein have opposing effects on neurotoxicity and soluble oligomer formation. J Clin Invest. 2009 Nov;119(11):3257-65. PubMed.
  11. . Novel antibodies to synuclein show abundant striatal pathology in Lewy body diseases. Ann Neurol. 2002 Aug;52(2):205-10. PubMed.

External Citations

  1. view her slides

Further Reading

Primary Papers

  1. . Distinct region-specific alpha-synuclein oligomers in A53T transgenic mice: implications for neurodegeneration. J Neurosci. 2010 Mar 3;30(9):3409-18. PubMed.
  2. . Phosphorylation at S87 is enhanced in synucleinopathies, inhibits alpha-synuclein oligomerization, and influences synuclein-membrane interactions. J Neurosci. 2010 Mar 3;30(9):3184-98. PubMed.