Dieticians recommend consuming mostly unsaturated fats for cardiovascular health and a slim waistline. Now, in a surprising twist, researchers claim one such fat, oleic acid, may worsen α-synuclein pathology. In the December 4 Molecular Cell, researchers led by Dennis Selkoe, Ulf Dettmer, and Saranna Fanning at Brigham and Women’s Hospital, Boston, report that oleic acid, a monounsaturated fat that happens to be the major constituent of olive oil, promoted toxicity in a handful of cellular and animal models of Parkinson’s disease. Oleic acid caused aggregated and phosphorylated α-synuclein to build up in the cytoplasm, and the cells were more likely to die. But don’t change your dietary habits just yet—this oleic acid was made on the spot in cells, and may have little to do with consumption. In fact, in a vicious pathological circle, its production seems to be promoted by high levels of α-synuclein. Both yeast and neurons that accumulated α-synuclein overproduced the fatty acid.

The enzyme that converts stearic acid to oleic, stearoyl-CoA-desaturase (SCD), might be a promising new drug target, the authors believe. Inhibiting SCD reduced oleic acid levels, lessened α-synuclein aggregates, and protected neurons from degeneration. The researchers have obtained a grant from the Michael J. Fox Foundation to test SCD inhibitors in mouse models.

A paper in the December 4 Cell Reports supports this approach. Researchers at Yumanity Therapeutics in Cambridge, Massachusetts, independently identified SCD inhibitors as protective against α-synuclein toxicity in an unbiased screen. Inhibitors preserved neuronal health and smoothed out vesicle trafficking, which slows in Parkinson’s disease. Yumanity plans to test an SCD inhibitor in clinical studies next year. “We think this is an intriguing new approach, and we’re excited to test this mechanism in patients,” CEO Kenneth Rhodes told Alzforum.

Lipid Inhibitor for Parkinson’s?

Neuroblastoma cells (top) expressing mutant α-synuclein (green) made fewer α-synuclein inclusions (bright dots) when given an SCD inhibitor (bottom). [Courtesy of Molecular Cell, Fanning et al.]

Other scientists agreed the approach has potential, but urged that the research first gather detailed pharmacologic and toxicity data in mice. They note a potential for side effects since oleic acid is abundant in cell membranes throughout the body. “While this new finding is tremendously exciting and highlights novel areas of research for the Parkinson’s community, some caution may be required moving forward toward the clinic,” Simon Stott at The Cure Parkinson’s Trust, a charity based in London, wrote to Alzforum (full comment below). Subhojit Roy at the University of Wisconsin, Madison, said that α-synuclein has been known for some time to interact with lipids, but few therapeutic strategies have exploited this relationship.

Lipidomics Screen: Out Pops Oleic Acid
It’s known that α-synuclein associates with acidic phospholipids in membranes, allowing the protein to hop on and off vesicles (Davidson et al., 1998; Perrin et al., 2000). Selkoe and others reported that α-synuclein directly binds fatty acids, including oleic (Sharon et al., 2001; Kubo et al., 2005). These interactions seemed to nudge α-synuclein to assume a helical shape and promoted its oligomerization (Perrin et al., 2001; Feb 2003 news). 

Fanning wondered how α-synuclein might affect the lipid composition of cells. Working initially in the lab of the late Susan Lindquist at MIT, Fanning analyzed the lipidomic profile of yeast that expressed human α-synuclein driven by an estrogen promoter. Within six hours of α-synuclein induction, these cells massively overproduced oleic acid, up to 60-fold excess. Fatty acids are incorporated into diglycerides, and in keeping with this, Fanning saw a threefold excess of lipid droplets that contain these fats in the yeast as well.

Oleic acid and diglycerides both seemed to harm cells. With the buildup of diglycerides came sluggish vesicle trafficking and cytotoxicity. Adding exogenous oleic acid to yeast cultures that expressed α-synuclein aggravated cell death. Oleic acid did not harm wild-type yeast. Other fatty acids had no effect on cell health, however. Turning off the yeast ortholog of SCD, Ole1, which converts stearic acid to oleic acid, rescued cells.

Would the link between α-synuclein and fat production hold true in neurons? Results from several model systems suggest as much, the scientists report. In a roundworm model of PD, suppressing oleic acid production cut dopaminergic neuron death in half. In rat primary cortical neurons, overexpression of human α-synuclein led to a twofold excess of oleic acid and sped cell death. As in yeast, knocking down SCD1 preserved cells. Mice that express human mutant E46K α-synuclein accumulated excess diglycerides whose levels correlated with motor deficits. Human neurons derived from iPS cells followed the same pattern, whereby overexpressing wild-type α-synuclein in these cells doubled oleic acid levels. Neurons derived from patients with an α-synuclein triplication, and neurons carrying the familial E46K mutation, also made more diglycerides than control lines.

To get an idea of how oleic acid could end up being toxic, the authors used a human neuroblastoma cell line that expresses α-synuclein with E46K mutations in several KTKEGV motifs that repeat in the protein. E46K α-synuclein adopts monomeric form more readily than wild-type, and this causes neuronal inclusions, as seen in a mouse E46K repeat model developed in the Selkoe lab (Oct 2018 news). Knocking down or inhibiting SCD1 suppressed inclusions. Inhibiting SCD also lessened other signs of α-synuclein toxicity in these neuroblastoma cells. The tetramer-to-monomer ratio rose, levels of phosphorylated α-synuclein sank, and more of the protein was soluble rather than membrane-bound. Selkoe and colleagues argue that tetramers represent a physiological, beneficial form of α-synuclein (Aug 2011 news; Feb 2012 news; Apr 2015 news). Exposing these cells to oleic, but not other, fatty acids dose-dependently increased the number of α-synuclein inclusions.

The exact mechanism of how a cell’s lipid composition affects these α-synuclein properties remains unclear. When monounsaturated fatty acids, such as oleic, slide into membranes, the lipid bilayers become more fluid. Fanning thinks this might matter. “The way α-synuclein interacts with those membranes apparently causes problems in some way and increases toxicity to the cell,” she told Alzforum. Likewise, it is a mystery how α-synuclein boosts oleic acid production. Dettmer emphasized the circular relationship. “I found it fascinating that α-synuclein seems to upregulate the exact factor that makes it more toxic,” Dettmer said.

Drug Screen: Out Pops SCD Inhibitor
Yumanity researchers arrived at SCD from a different direction. Joint first authors Daniel Tardiff and Benjamin Vincent screened for compounds that rescued growth in yeast overexpressing α-synuclein. A clutch of compounds containing an oxadiazole core did the best. Their benefit was specific for α-synuclein; it did not reduce other toxicities, such as Aβ. The most potent one, YTX-465, boosted growth more than fourfold, to 40 percent of that in wild-type cells, at a concentration of 50 nM. It also restored vesicle trafficking to normal and reduced the number of α-synuclein foci.

Further experiments revealed the SCD ortholog Ole1 as a target of YTX-465. Ole1 produces both oleic and palmitoleic acid. Adding either to yeast cultures abrogated YTX-465’s effects, supporting these fatty acids as key downstream effectors for the compound. At 500 nM, YTX-465 inhibited the growth of wild-type yeast, indicating that suppressing oleic acid by too much can be toxic.

The authors generated human cortical neurons from iPS cells, transfected them with human mutant A53T α-synuclein, and treated them with an YTX-465 analogue that was active against SCD1. Levels of monounsaturated fatty acids rose, and neuron viability improved. This analogue will not advance to the clinic, however; Rhodes noted that Yumanity has selected a molecule with a different chemical scaffold and distinct properties for human studies.

Overall, despite differences in the model systems, both groups’ findings dovetail closely. “It’s encouraging that the two papers are in agreement on so many levels,” Fanning said. Rhodes agreed, noting, “This lends credence to the importance of this target in α-synuclein biology.”

One discrepancy was that the Yumanity researchers saw no change in fatty acid composition in yeast expressing α-synuclein. Fanning thinks that new fatty acids might have been missed because they were incorporated into triglycerides. The Yumanity researchers did see an increase in triglyceride content. Selkoe noted that Yumanity used a galactose-inducible promoter to express α-synuclein in yeast, rather than an estradiol one. The addition of galactose to yeast cultures by itself changes the fatty acid composition, and could have masked the effects of α-synuclein, Selkoe said. Rhodes agreed this could be the case.

Implications for Health?
Is SCD inhibition promising as a therapeutic approach? David Standaert at the University of Alabama, Birmingham, noted that many cellular factors have been found to modulate α-synuclein aggregation and toxicity. This includes molecules involved in trafficking, lysosomal degradation, and protein folding. “Oleic acid, however, may be more amenable to therapeutic intervention than some of these other pathways,” he wrote to Alzforum (full comment below).

Nonetheless, others wondered about safety. “If oleic acid is important for the toxicity of α-synuclein, it might also be important for its normal function,” said Robert Edwards at the University of California, San Francisco. Stott pointed to preclinical studies where inhibiting SCD exacerbates inflammation in models of colitis, and is detrimental in at least one diabetes model (Chen et al., 2008Flowers et al., 2007). “There could be complications in peripheral tissues,” Stott suggested.

Dettmer agrees. Ideally, SCD inhibition should be directed to the brain, he said. One way to do this might be to target SCD5, which is confined to the central nervous system, rather than SCD1. Selkoe plans to investigate the pros and cons of various existing SCD inhibitors in mouse models of Parkinson’s. He believes the target range for suppressing oleic acid production will be low, perhaps inhibition of less than 30 percent.

So what about dietary fats? Oleic acid is found in many foods considered healthy, such as olive oil, avocadoes, and nuts. Rhodes noted that fatty acids in the bloodstream barely enter the brain. “It looks like most of the lipids we are studying are generated within the CNS,” he said. Still, Selkoe believes the issue is worth studying. He suggested putting a PD mouse model on a diet low in monounsaturated fats and examining the effects, if any, on α-synuclein pathology. In the meantime, it’s probably still safe to enjoy those foods. Adherence to a Mediterranean diet, which is rich in monounsaturated fats, has been linked to a lower risk of Parkinson’s in small epidemiological studies (Alcalay et al., 2012; Maraki et al., 2018; Agarwal et al., 2018).—Madolyn Bowman Rogers

Comments

  1. This is an interesting but complex study by Fanning and colleagues.

    My takeaway is that there are two inter-related themes here: increased α-synuclein leads to changes in lipid metabolism; and alterations in lipids can modulate the toxicity of α-synuclein. While both are of interest, I think the data on the effect of oleic acid on α-synuclein toxicity in Fig. 4 is the most intriguing. They show nicely that the abundance of oleic acid can modulate the aggregation and toxicity of α-synuclein. I think of this as part of the broader class of targets that seems to modulate aggregation and toxicity of α-synuclein, which also includes factors that modulate trafficking, lysosomal function, and chaperone proteins. Oleic acid, however, may be more amenable to therapeutic intervention than some of these other pathways.

    An obvious next step is to take this approach into an intact mammalian model system. I will be interested to see how this turns out.​

  2. Model organisms are increasingly exploited to identify disease mechanisms and suggest novel therapies. Fanning et al. capture the essence of the work of late Susan Lindquist and her legacy, shared with Dennis Selkoe and colleagues, by providing an elegant illustration of how a molecular genetic study in yeast focusing on α-synuclein (αS) uncovers a potential molecular mechanism underlying neurotoxicity associated with Parkinson’s disease (PD) as well as a candidate therapeutic.

    It has been known for a long time that the biology of aS is intimately linked to that of membrane lipids, primarily reflecting the ability of this protein to physically interact with anionic phospholipids present in various organelles, including synaptic vesicles. This interaction is critical to the physiological function of aS at synapses and is severely perturbed in several PD-linked mutants, affecting its ability to oligomerize, aggregate, and undergo degradation in cells through various mechanisms. This new study focuses on a vastly distinct aspect of lipid biology related to aS toxicity and rather focuses on so-called glycerolipids and fatty acids, such as oleic acid. In fact, earlier studies in the Lindquist lab laid the foundation for this study by showing that overexpressing aS in budding yeast is toxic and interferes with the early biosynthetic pathway, a phenomenon reflecting the ability of aS to modulate vesicular traffic along intracellular compartments and rescued by overexpression of Rab1. Additionally, separate studies from the same lab showed that aS overexpression in yeast leads to the formation of lipid droplets (LDs), which are organelles that are generated from the early biosynthetic pathway (i.e., the endoplasmic reticulum) and store neutral lipids such as triacylglycerol (TG), diacylglycerol (DG), and sterol esters. However, the specific molecular mechanisms leading to aS toxicity in yeast were unknown, and whether those mechanisms applied to neurons was unclear as well. 

    Collaborating with world experts in lipid biology, such as Tobias Walther, Robert Farese, Sepp Kohlwein and Supriya Srinivasan, Selkoe, Lindquist, and colleagues were able to identify a rather specific mechanism accounting for the toxicity of aS and it relates to lipid toxicity. It is well known in the field of lipid biology that storing specific lipids, such as fatty acids and sterols in LDs, is a process that prevents lipid toxicity, as excess fatty acids or cholesterol is deleterious to cells. In the case of fatty acids, their storage in the form of glycerolipids, such as DG and TG, also represents a major source of energy that can be mobilized by cytosolic lipases or macrolipophagy (i.e., a form of autophagy that clears lipid droplets in lysosomes). Here, by interfering with various factors controlling glycerolipid and fatty-acid metabolism as well as LD dynamics through a number of genetic manipulations, and combined with comprehensive lipidomic analyses, the authors showed that formation of LDs induced by aS overexpression is protective. In essence, preventing TG synthesis in LDs or LD formation drastically enhances aS-related cytotoxicity in yeast, whereas enhancing lipid storage capacity is protective. The authors were able to narrow down the toxicity mechanism to at least two precursor lipids for TG synthesis: DG and the fatty acid, oleic acid (OA). Specifically, channeling excess DG into another pathway for phospholipid synthesis and decreasing production of OA are both protective. The latter mechanism was further exploited and lays the foundation for a potential therapeutic approach.

    Since OA, a fatty acid with 18 carbons and one unsaturation, was identified as a mediator of αS toxicity, it was logical to assume that the rate-limiting enzyme for its synthesis, the Δ9 FA desaturase encoded by yeast gene Ole1 and homologous to the mammalian stearoyl-CoA desaturase (SCD), would underlie this toxicity. Indeed, ablating Ole1 in yeast abolished αS toxicity and corrected trafficking defects induced by αS overexpression. The next critical question was to determine whether these striking, yeast-based mechanisms could be extrapolated to mammalian neurons with the added challenge that LD biology has not been much studied in neurons, in part based on scant evidence these organelles appear in CNS neurons, at least in vivo. While this potentially questions the pathophysiological significance of observations made in cell culture, an alternate view may be that the reduced ability of neurons to generate abundant LDs may precisely sensitize them to lipid toxicity. With these considerations in mind, the authors provided evidence that their yeast findings generally translate into cultured rat cortical neurons, albeit with a somewhat less striking outcome and similar concerns about αS overexpression. Nevertheless, various molecular genetic manipulations in rat cortical neurons did confirm that αS toxicity can be attenuated by reducing DG and OA levels, the latter being achieved by silencing or pharmacologically inhibiting SCD. Besides validating this neurotoxic pathway in dopaminergic neurons in the worm, the authors further addressed its pathophysiological significance by studying it in human iPSC-derived neurons overexpressing αS, expressing higher αS levels via gene triplication (which causes PD), and expressing endogenous levels of the disease mutant E46K. While the relevant lipid changes were not as striking, they did show alteration in glycerolipid and unsaturated fatty acid metabolism consistent with those reported in yeast, and somewhat consistent with those found in the brains of mice expressing the familial PD-linked αS mutant E46K.

    How does this translate into αS-mediated toxicity and aggregation? The authors provided a potential explanation, focusing on several aspects of αS-related phenotypes observed in a neuroblastoma cell line overexpressing αS, such as formation of inclusions. Pharmacological inhibition and reducing levels of SCD1 were both able to decrease αS cytotoxicity, inclusions, hyperphosphorylation, and the tetramer-monomer ratio (previously shown to reduce aggregate formation). Interestingly, Vincent and colleagues reported independent findings supporting the notion that SCD inhibition may be beneficial in human iPSC-derived neurons expressing various forms of αS, however, starting from a phenotypic screen in yeast. The enrichment of both DG and OA-containing lipids in lipid bilayers is known to destabilize membranes and alter their curvature, likely affecting the ability of αS to interact with them and oligomerize. Additional work will be required to precisely delineate the mechanisms through which lipids affect the behavior of αS, its propensity to oligomerize, aggregate, and exert its neurotoxicity.

    Is SCD a viable drug target? While results from both papers are promising, the road to the clinic is long and intricate, as it is for all candidate targets. Some of the fundamental aspects that will have to be addressed along the way include demonstrating proof of concept that manipulating the SCD pathway shows efficacy in preclinical models of familial PD (expressing mutant forms of αS), with the caveat that at least two related isoforms of SCD are expressed in the brain, SCD1 and SCD5. Are these two isoforms expressed in dopaminergic neuron populations that are vulnerable in PD and other synucleinopathies? Is inhibition of one or both required to achieve efficacy? Use of mouse genetics combined with the development of brain-penetrant, potentially isoform-specific compounds should facilitate this endeavor. Another concern in the evaluation of drug targets is the potential safety liabilities. SCD inhibitors have been considered for peripheral indications, including diabetes and liver steatosis, and appear to be relatively well tolerated, despite early concerns raised in preclinical studies. There may be brain-specific safety concerns related to manipulating this pathway in CNS neurons and potentially other cell types in the CNS, including oligodendrocytes, astrocytes, and microglia. Of note, accumulation of OA-containing neutral lipids in ependymal cells of an Alzheimer’s disease (AD) mouse models has been shown to interfere with the proliferation of neural stem cells in the forebrain, and SCD inhibition rescued these proliferative defects (Hamilton et al., 2015). Before that, another study found that SCD is upregulated in AD patients’ brain (Astarita et al., 2011). Overall, there is a growing interest for the SCD pathway in the area of neurodegenerative disorders, and future work should establish whether it can be manipulated in a beneficial way. In the absence of direct genetic evidence linking the SCD genes to PD, AD, or other neurodegenerative disorders, it will be also essential to establish that this pathway is activated in disease. Finally, given the recently-identified genetic associations of other genes operating in DG or fatty acid metabolism with PD (i.e., DGKQ and ELOVL7) (Chang et al., 2017), mechanistic studies exploring these connections will be informative.

    References:

    . Aberrant Lipid Metabolism in the Forebrain Niche Suppresses Adult Neural Stem Cell Proliferation in an Animal Model of Alzheimer's Disease. Cell Stem Cell. 2015 Oct 1;17(4):397-411. Epub 2015 Aug 27 PubMed.

    . Elevated stearoyl-CoA desaturase in brains of patients with Alzheimer's disease. PLoS One. 2011;6(10):e24777. PubMed.

    . A meta-analysis of genome-wide association studies identifies 17 new Parkinson's disease risk loci. Nat Genet. 2017 Sep 11; PubMed.

  3. This research truly validates the late Prof. Susan Lindquist's novel idea of using yeast to study neurodegeneration—an idea which in the early 2000s was considered "crazy" even by people in her own lab (see New York Times article). And the fact that this research is now resulting in compounds being developed for clinical testing only further cements an amazing legacy for Prof. Lindquist.

    All of that said, the clinical translation of stearoyl-CoA-desaturase (SCD) inhibition may face significant challenges. We know that SCD inhibition can have both beneficial and detrimental effects depending on context. For example, while inhibition of SCD may reduce α-synuclein toxicity in the brain, in models of colitis it exacerbates proinflammatory responses (Chen et al., 2008). In addition, in various preclinical models of diabetes, SCD inhibition has been reported to have completely different effects depending on the model used (Flowers et al., 2007). Thus, in a scenario reminiscent of the early LRRK2 inhibitors, there could be complications in non-CNS tissues that will need to be addressed.

    While this new finding is tremendously exciting and highlights novel areas of research for the Parkinson's community, some caution may be required moving forward toward the clinic.

    References:

    . Metabolomics reveals that hepatic stearoyl-CoA desaturase 1 downregulation exacerbates inflammation and acute colitis. Cell Metab. 2008 Feb;7(2):135-47. PubMed.

    . Loss of stearoyl-CoA desaturase-1 improves insulin sensitivity in lean mice but worsens diabetes in leptin-deficient obese mice. Diabetes. 2007 May;56(5):1228-39. Epub 2007 Mar 16 PubMed.

  4. These are exciting findings, with implications for α-synuclein therapeutics and Parkinson's disease. Researchers have known that α-synuclein interacts with lipids for some time, but few therapeutic strategies have exploited this property.

    The two papers use different approaches but arrive at the same target—the enzyme stearoyl-CoA desaturase (SCD)—which gives confidence that this is a bona fide target. Overall, the work from both groups shows that reducing levels of SCD decreases α-synuclein-induced toxicity, and thus could be a potential therapeutic avenue in Parkinson’s disease. 

    While the data in both studies is convincing, the effects need to be validated in established in vivo animal models of Parkinson’s (so far, it seems the results are validated in cell lines, iPSC-neurons, and a C. elegans model). Given that SCD is likely an important enzyme playing myriad roles in maintaining physiology, detailed pharmacologic and toxicity studies are warranted. Advantages of this approach over other means to lower α-synuclein directly are also unclear at this time.​

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Playing "Gotcha" with the Phantom: α-synuclein Oligos Spotted in Vivo
  2. Sans Synuclein Tetramers, Mice Mimic Parkinson’s Disease
  3. An α-Synuclein Twist—Native Protein a Helical Tetramer
  4. Synuclein—Researchers Out of Sync on Structure
  5. Form and Function: What Makes α-Synuclein Toxic?

Paper Citations

  1. . Stabilization of alpha-synuclein secondary structure upon binding to synthetic membranes. J Biol Chem. 1998 Apr 17;273(16):9443-9. PubMed.
  2. . Interaction of human alpha-Synuclein and Parkinson's disease variants with phospholipids. Structural analysis using site-directed mutagenesis. J Biol Chem. 2000 Nov 3;275(44):34393-8. PubMed.
  3. . alpha-Synuclein occurs in lipid-rich high molecular weight complexes, binds fatty acids, and shows homology to the fatty acid-binding proteins. Proc Natl Acad Sci U S A. 2001 Jul 31;98(16):9110-5. PubMed.
  4. . A combinatorial code for the interaction of alpha-synuclein with membranes. J Biol Chem. 2005 Sep 9;280(36):31664-72. Epub 2005 Jul 14 PubMed.
  5. . Exposure to long chain polyunsaturated fatty acids triggers rapid multimerization of synucleins. J Biol Chem. 2001 Nov 9;276(45):41958-62. PubMed.
  6. . Metabolomics reveals that hepatic stearoyl-CoA desaturase 1 downregulation exacerbates inflammation and acute colitis. Cell Metab. 2008 Feb;7(2):135-47. PubMed.
  7. . Loss of stearoyl-CoA desaturase-1 improves insulin sensitivity in lean mice but worsens diabetes in leptin-deficient obese mice. Diabetes. 2007 May;56(5):1228-39. Epub 2007 Mar 16 PubMed.
  8. . The association between Mediterranean diet adherence and Parkinson's disease. Mov Disord. 2012 Feb 7; PubMed.
  9. . Mediterranean diet adherence is related to reduced probability of prodromal Parkinson's disease. Mov Disord. 2019 Jan;34(1):48-57. Epub 2018 Oct 10 PubMed.
  10. . MIND Diet Associated with Reduced Incidence and Delayed Progression of ParkinsonismA in Old Age. J Nutr Health Aging. 2018;22(10):1211-1215. PubMed.

Further Reading

Primary Papers

  1. . Lipidomic Analysis of α-Synuclein Neurotoxicity Identifies Stearoyl CoA Desaturase as a Target for Parkinson Treatment. Mol Cell. 2019 Mar 7;73(5):1001-1014.e8. Epub 2018 Dec 4 PubMed.
  2. . Inhibiting Stearoyl-CoA Desaturase Ameliorates α-Synuclein Cytotoxicity. Cell Rep. 2018 Dec 4;25(10):2742-2754.e31. PubMed.