Triple-Knockout, Synuclein-Free Mice Display Compromised Lipid Pattern

Recent studies have implicated synucleins in several reactions during the biosynthesis of lipids and fatty acids in addition to their recognised role in membrane lipid binding and synaptic functions. These are among aspects of decreased synuclein functions that are still poorly acknowledged especially in regard to pathogenesis in Parkinson’s disease. Here, we aimed to add to existing knowledge of synuclein deficiency (i.e., the lack of all three family members), with respect to changes in fatty acids and lipids in plasma, liver, and two brain regions in triple synuclein-knockout (TKO) mice. We describe changes of long-chain polyunsaturated fatty acids (LCPUFA) and palmitic acid in liver and plasma, reduced triacylglycerol (TAG) accumulation in liver and non-esterified fatty acids in plasma of synuclein free mice. In midbrain, we observed counterbalanced changes in the relative concentrations of phosphatidylcholine (PC) and cerebrosides (CER). We also recorded a notable reduction in ethanolamine plasmalogens in the midbrain of synuclein free mice, which is an important finding since the abnormal ether lipid metabolism usually associated with neurological disorders. In summary, our data demonstrates that synuclein deficiency results in alterations of the PUFA synthesis, storage lipid accumulation in the liver, and the reduction of plasmalogens and CER, those polar lipids which are principal compounds of lipid rafts in many tissues. An ablation of all three synuclein family members causes more profound changes in lipid metabolism than changes previously shown to be associated with γ-synuclein deficiency alone. Possible mechanisms by which synuclein deficiency may govern the reported modifications of lipid metabolism in TKO mice are proposed and discussed.

Interest in this protein family is first and foremost associated with their special role in human neurodegenerative diseases, collectively termed "synucleinopathies", and largely with a potential involvement of α-synuclein aggregates in Parkinson's disease (PD).
α-Synuclein was originally identified as a protein associated with synaptic vesicles (SV) [6]. It is widely expressed in various brain regions but localises specifically to the nerve terminal where it modulates synaptic functions. Several non-neural tissues, e.g., red blood including many PD cell and animal models [19,[36][37][38][39][40]. Lipids influence the structure and, consequently, the functions of synucleins (and vice versa, of note!) through complex interplaying and overlapping mechanisms. As an interplay, numerous recorded changes in lipids (and lipid metabolism in cells and various synuclein mouse models) were associated to either the synuclein loss-of-function or a toxic gain-of-function hypothesis [40][41][42][43][44]. In summary, they include: uptake of fatty acids into cells; changes in brain cardiolipin and its fatty acid acyl chains [2]; an increased level of brain triacylglycerol (TAG) content in α-synuclein mouse mutant model associated with increased fatty acid synthase expression and acyl-CoA synthase activity, with no variations in TAG lipase activity or in fatty acid β-oxidation [45]. α-Synuclein-null animals showed decreased palmitate uptake and the altered metabolism of this FA in brain, and turnover of arachidonate and docosahexanoate in brain phospholipids [46,47]. Our previous work on γ-synuclein null mice showed significant brain region-specific changes in phosphatidylserine and docosahexaenoic acid when compared to wild-type mice [48]. This member of the family has been demonstrated to be directly involved in lipid metabolism by enhancing TAG lipolysis in brown and white adipose tissues, where γ-synuclein was highly expressed [49,50]. In white fat adipocytes isolated from TKO mice, TAG synthesis was further boosted by high fat diet or reduced by caloric restriction [49,50].
In this work, we studied how deficiency of all three synucleins affected lipid composition in the brain regions of adult triple knockout (TKO) mice that do not display any signs of pathological changes in the nervous system [51,52]. We suggested that triple ablation would eliminate a well-known effect of functional redundancy and compensatory ability for loss of functions of the absent synuclein family member(s). In addition, in order to get a broader view on possible lipid modulations in transgenic mice, lipids and fatty acid composition of plasma and liver were also investigated.

Total Lipids and Lipid Classes in Liver and Plasma from WT and TKO Mice
Synuclein deficiency resulted in a significant reduction of the total lipids in the liver compared to those in WT animals, whereas the concentrations of total lipids in the plasma were unchanged (Figure 1). In the brain, the total lipid concentrations were unchanged: 2.65 ± 0.78 mg of FA/100 mg of fresh weigh (FW) and 2.17 ± 0.47 mg of FA/100 mg of FW in the midbrain of WT and TKO mice, respectively; 2.06 ± 0.33 mg of FA/100 mg of FW and 2.51 ± 0.53 mg of FA/100 mg of FW in the cortex of WT and TKO mice, respectively.
Molecules 2021, 26, x FOR PEER REVIEW 3 of 24 their oxidation products. Insofar, many studies revealed the molecular basis of interactions between α-synuclein and PUFA and their importance in both, neuroprotection and neurodegeneration and in vivo, including many PD cell and animal models [19,[36][37][38][39][40]. Lipids influence the structure and, consequently, the functions of synucleins (and vice versa, of note!) through complex interplaying and overlapping mechanisms. As an interplay, numerous recorded changes in lipids (and lipid metabolism in cells and various synuclein mouse models) were associated to either the synuclein loss-of-function or a toxic gain-of-function hypothesis [40][41][42][43][44]. In summary, they include: uptake of fatty acids into cells; changes in brain cardiolipin and its fatty acid acyl chains [2]; an increased level of brain triacylglycerol (TAG) content in α-synuclein mouse mutant model associated with increased fatty acid synthase expression and acyl-CoA synthase activity, with no variations in TAG lipase activity or in fatty acid β-oxidation [45]. α-Synuclein-null animals showed decreased palmitate uptake and the altered metabolism of this FA in brain, and turnover of arachidonate and docosahexanoate in brain phospholipids [46,47]. Our previous work on γ-synuclein null mice showed significant brain region-specific changes in phosphatidylserine and docosahexaenoic acid when compared to wild-type mice [48]. This member of the family has been demonstrated to be directly involved in lipid metabolism by enhancing TAG lipolysis in brown and white adipose tissues, where γ-synuclein was highly expressed [49,50]. In white fat adipocytes isolated from TKO mice, TAG synthesis was further boosted by high fat diet or reduced by caloric restriction [49,50].
In this work, we studied how deficiency of all three synucleins affected lipid composition in the brain regions of adult triple knockout (TKO) mice that do not display any signs of pathological changes in the nervous system [51,52]. We suggested that triple ablation would eliminate a well-known effect of functional redundancy and compensatory ability for loss of functions of the absent synuclein family member(s). In addition, in order to get a broader view on possible lipid modulations in transgenic mice, lipids and fatty acid composition of plasma and liver were also investigated.

Total Lipids and Lipid Classes in Liver and Plasma from WT and TKO Mice
Synuclein deficiency resulted in a significant reduction of the total lipids in the liver compared to those in WT animals, whereas the concentrations of total lipids in the plasma were unchanged (Figure 1). In the brain, the total lipid concentrations were unchanged: 2.65 ± 0.78 mg of FA/100 mg of fresh weigh (FW) and 2.17 ± 0.47 mg of FA/100 mg of FW in the midbrain of WT and TKO mice, respectively; 2.06 ± 0.33 mg of FA/100 mg of FW and 2.51 ± 0.53 mg of FA/100 mg of FW in the cortex of WT and TKO mice, respectively.  The profiles of lipid classes found in liver and plasma from WT and TKO mice are shown in Figure 2. In the liver, the total polar lipids (TPL) and TAG were the major fractions with SE being a minor class in both, WT and TKO mice. The relative amounts of TPL increased in the liver of TKO animals (from 34 to 48.8%), whereas those of TAG and SE  (Figure 2A). In the plasma, free fatty acids (FFA, or non-esterified FA) were also determined, and the level of this fraction significantly decreased in TKO mice (4.8%) in comparison to WT animals (11.1%) ( Figure 2B). TPL and TAG were unaltered in the plasma of TKO mice; the relative amount of SE was much higher in the plasma than in the liver and increased in TKO mutants up to 21.3% comparing to WT animals (16.4%). The profiles of lipid classes found in liver and plasma from WT and TKO mice are shown in Figure 2. In the liver, the total polar lipids (TPL) and TAG were the major fractions with SE being a minor class in both, WT and TKO mice. The relative amounts of TPL increased in the liver of TKO animals (from 34% to 48.8%), whereas those of TAG and SE decreased (from 63.7% to 50% and from 2.3% to 1.2%, consequently) ( Figure 2A). In the plasma, free fatty acids (FFA, or non-esterified FA) were also determined, and the level of this fraction significantly decreased in TKO mice (4.8%) in comparison to WT animals (11.1%) ( Figure 2B). TPL and TAG were unaltered in the plasma of TKO mice; the relative amount of SE was much higher in the plasma than in the liver and increased in TKO mutants up to 21.3% comparing to WT animals (16.4%).

Fatty Acid Composition of Lipid Classes in Liver and Plasma from WT and TKO Mice
In liver and plasma, all the separated lipid classes were analysed for their fatty acid composition. FA profiles of TAG isolated from liver and plasma are quite similar with the domination of palmitic and oleic acids followed by LA, stearic and isomers of C16:1. Only minor amounts of PUFA (not-exceeding 2%) were present in plasma and liver with the higher percentage of α-linolenic acid (C18:3n3, ALA) in TAG of the plasma of both, WT and TKO animals (Table S2).
In TAG of plasma and liver, the subtle differences in FA percentages were found when compared WT and TKO mice. In the plasma, only the level of stearic acid was affected by synuclein deficiency: its relative amount was lower in the plasma TAG from TKO mice. In the liver TAG, the percentages of both isomers of C16:1 were higher whereas the level of other monoenoic FA, oleic acid, which is predominant in TAG, was lower and accounted for 52.5% and 47.3% in the liver from WT and TKO mice, respectively (Table S2).
FA of SE from the plasma are characterised by the high levels of ARA (about 30% in both, WT and TKO mice) and LA accounted for 34% in the plasma SE from WT and for 42.4% in TKO mice (Table S3). The other PUFA, including ALA, DGLA, EPA and DHA were detected in small amounts (about 1-4%) in WT and TKO mice. Saturated palmitic and stearic acids were found in the relative amounts of not-exceeding 6% and 4%, respectively, in WT mice. In TKO mice, the level of palmitic acid was not changed compared to that in WT mice, whereas the relative concentration of oleic acid decreased from 3.7 to 1.9% under synuclein deficiency. In SE fraction isolated from the liver, oleic acid was dominant in both, WT and TKO mice, with a significant decrease in the latter (from 61.1 to 43.2%). Two other major FA in SE of the liver were LA (no differences in its relative amounts were found between WT and TKO mice) and palmitic acid. The percentage of this acid was significantly higher in SE of the liver from TKO animals. The level of other saturated FA, oleic acid, was also increased in TKO mice up to 9.8% in comparison to 4.7% in WT. Additionally, the higher percentages of C16:1n9 and ALA in SE were detected in the liver of TKO mice. No long-chain PUFA, except ARA (1.9% in both WT and mutant mice), were detected in the liver of the studied animals (Table S3).
FA profiles of the TPL fraction isolated from plasma and liver are present in Table 1. In general, FA compositions of this lipid fraction in both samples from WT mice were characterised by the large amounts of saturated FA (palmitic and stearic acids), monoand dienoic C18 FA (oleic acid and LA) and ARA followed by DHA, DGLA, nervonic acid (C24:1n9), ALA and palmitoleic acid. These parameters were remarkable affected by synuclein deficiency. Thus, in TPL of the plasma from TKO mice, a significant reduction in the level of LA (from 19.2% in WT to 2.7% in TKO) was noted, and only the trace (<0.5%) amounts of ARA, DHA and DGLA were detected. These changes were accompanied by significant increases in the levels of saturated palmitic and monoenic oleic acids (Table 1). In the liver, analysis of FA in TPL fraction also revealed increases in palmitic and oleic acids and a decline in the percentage of ARA. The relative amounts of LA, DGLA and DHA were not altered in TPL fraction in the liver from TKO mice.

Composition of Free Fatty Acids in Plasma
A comparison of the profiles of free fatty acids isolated from the plasma of WT and TKO animals is illustrated in Figure 3. In consonance with the data on FA profiles of the total lipids, SE and TPL (described above), the analysis of FFA showed a reduction of the percentages of LA (from 16.6% in WT to 12.4% in TKO), ARA (from 6.2 to 1.5%) and DHA (from 2.5 to 0.3%) together with increases in C16:1n9, oleic acid and ALA. No differences were noted for saturated FA and palmitoleic acid between WT and TKO mice ( Figure 3).

Composition of Free Fatty Acids in Plasma
A comparison of the profiles of free fatty acids isolated from the plasma of WT and TKO animals is illustrated in Figure 3. In consonance with the data on FA profiles of the total lipids, SE and TPL (described above), the analysis of FFA showed a reduction of the percentages of LA (from 16.6% in WT to 12.4% in TKO), ARA (from 6.2% to 1.5%) and DHA (from 2.5% to 0.3%) together with increases in C16:1n9, oleic acid and ALA. No differences were noted for saturated FA and palmitoleic acid between WT and TKO mice ( Figure 3).  Figure 4 shows the data on fatty acid composition of individual polar lipids (PL) in the liver from WT and TKO mice. PL composition was in keeping with the generally known lipids found in mammalian livers, particularly: phosphatidylcholine (PC) and phosphatidylethanolamine (PE) as predominant phosphoglycerolipids, followed by phosphatidylserine (PS), phosphatidylinositol (PI), cardiolipin (CL) and sphingomyelin (SM) (data not shown).  Figure 4 shows the data on fatty acid composition of individual polar lipids (PL) in the liver from WT and TKO mice. PL composition was in keeping with the generally known lipids found in mammalian livers, particularly: phosphatidylcholine (PC) and phosphatidylethanolamine (PE) as predominant phosphoglycerolipids, followed by phosphatidylserine (PS), phosphatidylinositol (PI), cardiolipin (CL) and sphingomyelin (SM) (data not shown).

Fatty Acid Composition of Individual Phospholipids in Liver from WT and TKO Mice
The analysis of FA in PC revealed six dominant FA, namely, palmitic, stearic, LA, ARA, oleic and DHA varying in their percentages from 18.9% (palmitate) to 8.8% (DHA) in PC from WT mice. Vaccenic, palmitoleic and ALA were found in low or trace amounts in PC of the liver. In synuclein-free mice, the level of ARA was significantly reduced from 14.5 to 8.9% with a reciprocal increase in the relative amounts of LA (from 15 to 18.6%); the reduced percentage of stearate was also noted in PC of the liver of TKO mice ( Figure 4). The liver PE was dominated by the same FA as found in PC, although their relative concentrations were different ( Figure 4). Thus, ARA, stearate and DHA showed the highest percentages accounting for 22%, 19.5% and 17.5%, respectively, followed by palmitate (15.2%), oleate (11.1%) and LA (9.2%) in WT mice. Palmitoleic acid was present in trace amounts; vaccenic, eicosapentaenoic (C20:5n3, EPA) and docosapentaenoic (C22:5n3, DPA) acids were the minor FA in PE of the liver from both, WT and TKO mice. Synuclein-deficiency resulted in a decline in DHA percentage and in an appearance of DGLA and gondoic acid in the liver PE in small amounts (0.6-0.8% of the total FA).
In PS, stearate, ARA and DHA were the dominant acyl chains (43.7%, 25.1% and 12.8%, respectively) with palmitic and oleic acid accounted for 6.2% and 4.5% of the total FA in the liver of WT mice ( Figure 4). The minor percentages of vaccenate, LA, ALA, adrenic acid (C22:4n6) and DPA were detected in PS of the liver. A notable decrease in the level of ARA (from 25.1 to 17.3%) alongside with an increase in palmitate (from 6.2 to 11.1%) was found in PS in the liver of TKO mutants. Two FA, stearate and DHA, were prevalent reaching 46.3% and 41.2% in PI of the liver from the studied animals ( Figure 4). The moderate amounts of palmitate (3.1%), oleate (2.4%), DGLA (2%), eicosadienoic acid (C20:2n6; 1.4%) and LA (1.3%) as well as the small amounts of vaccenate, ALA and DHA were detected in WT mice. In PI of the liver from TKO mutants, a significant reduction in the percentage of DHA (from 41.2 to 29.9%) was noted. A similar trend was shown for PC and PS ( Figure 4). This decline was accompanied by increases in the amounts of palmitate (from 3.1 to 6.6%), oleate (from 2.4 to 4.4%) and LA (from 1.3 to 3.1%) in the liver of synuclein-free mice.
In cardiolipin, LA was accounted for more than 50% of the total fatty acid chains. Other FA with the relative amounts in the range between 4% and 10% were: DHA, DGLA, vaccenate, palmitate, stearate and oleate. The small concentrations of eicosadienoic, arachidic, and trace amounts of ARA and eruric acid (C20:1n9) were also determined. No differences in cardiolipin FA profiles in the liver form WT and TKO mutants were revealed ( Figure 4). Sphingomyelin (SM) (sphingophospholipid) was characterised by the appreciable amounts of saturated fatty acids including palmitic (20.2%), stearic (15.7%), behenic (11.1%) and lignoceric (11.2%) acids, as well as a long-chain monoenoic acid, nervonic acid (C24:1n9, 20.7%) in the liver from WT mice. Another shorter chain monoenoic eruric acid and oleic acid were found in the relative amounts of 4.8% and 4.6%, respectively, in WT mice. The small percentages of DHA, ARA, DGLA, the isomers of C16:1 and vaccenic acid were also determined in these samples. Under synuclein deficiency, a significant increase in the level of behenic acid (from 11.1 to 19.1%) was recorded in the liver SM with no statistically significant changes in other acyl chains of this sphingophospholipid ( Figure 4).  The analysis of FA in PC revealed six dominant FA, namely, palmitic, stearic, LA, ARA, oleic and DHA varying in their percentages from 18.9% (palmitate) to 8.8% (DHA) in PC from WT mice. Vaccenic, palmitoleic and ALA were found in low or trace amounts in PC of the liver. In synuclein-free mice, the level of ARA was significantly reduced from 14.5% to 8.9% with a reciprocal increase in the relative amounts of LA (from 15% to 18.6%); the reduced percentage of stearate was also noted in PC of the liver of TKO mice ( Figure  4). The liver PE was dominated by the same FA as found in PC, although their relative

Fatty Acid Composition of Total Lipids in Midbrain and Cortex from WT and TKO Mice
The midbrain FA were characterised by the high levels of palmitate and oleate (about 20% of total FA), followed by palmitate (17% in WT and 18% in TKO mice) and DHA (12% and 14% in two groups of mice) ( Table S4). The percentages of ARA were 8% in WT animals and 9% in TKO mutants. The levels of adrenic, gondoic and vaccenic acids were in the range of 3-5% of the total FA; small amounts of lignoceric, nervonic, hexacosenoic acid (C26:1n9) and palmitoleic acids were also detected. In this brain region, synuclein deficiency resulted in a slight increase of ARA (from 8.3 ± 0.5% in WT to 9.3 ± 0.4% in TKO mice) in parallel to decreases in the percentages of hexacosenoic (from 2.8 ± 0.2% to 1.8 ± 0.1%), nervonic (from 2.7 ± 0.3% to 1.8 ± 0.2%) and lignoceric (from 0.9 ± 0.1% to 0.6 ± 0.1%) acids (Table S4).
Two major saturated FA, palmitate and stearate, were determined at 21-22% of the total FA in the cortex from WT and TKO animals, followed by DHA (18%), oleate (16%) and ARA (11%). In comparison to FA composition of the midbrain, the lower levels of oleate, vaccenate, gondoic, adrenic, lignoceric, nervonic and hexacosenoic acids alongside with the higher percentages of palmitate, ARA and DHA were found in the cortex (Table S4). No changes in the cortex FA were noted in TKO mice when compared to WT animals.

Composition of Individual Phospholipids in Midbrain and Cortex from WT and TKO Mice
Phosphatidylcholine (PC) and phosphatidylethanolamine (PE) were most abundant lipids in both brain regions (around 30-40% of TPL), followed by phosphatidylserine (PS) accounted for about 15% of PL, phosphatidylinositol (PI), sphingomyelin (SM), cardiolipin (CL), sulfatide (SL) and cerebroside (CER): the latter five lipids were found in the percentages not-exceeding 10% of TPL ( Figure 5). In the midbrain, the amounts of PC and SM were lower, while the levels of SL, CER and CL were higher than the concentrations of these polar lipids in the cortex. An effect of synuclein ablation was seen only in the midbrain, where the concentration of PC was elevated in TKO mice, and the amount of CER was reduced when compared to WT animals ( Figure 5). with the higher percentages of palmitate, ARA and DHA were found in the cortex (Table  S4). No changes in the cortex FA were noted in TKO mice when compared to WT animals.

Composition of Individual Phospholipids in Midbrain and Cortex from WT and TKO Mice
Phosphatidylcholine (PC) and phosphatidylethanolamine (PE) were most abundant lipids in both brain regions (around 30-40% of TPL), followed by phosphatidylserine (PS) accounted for about 15% of PL, phosphatidylinositol (PI), sphingomyelin (SM), cardiolipin (CL), sulfatide (SL) and cerebroside (CER): the latter five lipids were found in the percentages not-exceeding 10% of TPL ( Figure 5). In the midbrain, the amounts of PC and SM were lower, while the levels of SL, CER and CL were higher than the concentrations of these polar lipids in the cortex. An effect of synuclein ablation was seen only in the midbrain, where the concentration of PC was elevated in TKO mice, and the amount of CER was reduced when compared to WT animals ( Figure 5).  Figure 6 shows the data on fatty acid composition of the individual polar lipids in midbrain and cortex from WT and TKO mice.

Fatty Acid Composition of Individual Phospholipids in Midbrain and Cortex from WT and TKO Mice
The FA distribution in these lipids is typical of that which has been established in murine brain tissues including our previous study [48]. Analysis of FA in PI showed a domination of stearic acid and ARA in both brain regions from WT and TKO mice with  Figure 6 shows the data on fatty acid composition of the individual polar lipids in midbrain and cortex from WT and TKO mice.

Fatty Acid Composition of Individual Phospholipids in Midbrain and Cortex from WT and TKO Mice
The FA distribution in these lipids is typical of that which has been established in murine brain tissues including our previous study [48]. Analysis of FA in PI showed a domination of stearic acid and ARA in both brain regions from WT and TKO mice with the higher level of ARA in the cortex in comparison to the midbrain (34% vs. 42%) ( Figure 6). In the cortex, the lower relative amounts of oleic acid and DHA were also noted. The appreciable amounts of palmitic acid and two isomers of C18:1 acid together with a low (about 0.5% of total FA in PI) level of palmitoleic acid were also determined in PI from both brain regions of the studied mice. In the midbrain, in addition to described FA, gondoic (C20:1n9) and adrenic (C22:4n6) acids were found in the relative amounts of about 1% of FA, whereas in the cortex, these acids were present only in the trace amounts (less than 0.5% of total FA). There were no differences in FA composition in PI in the midbrain and cortex between WT and TKO animals.
In PS, stearic acid, the combined isomers of C18:1 (n7 + n9) and DHA were the major acyl chains. In the cortex of both animal groups, the higher levels of stearate (43% vs. 40-41%) and DHA (33% vs. 19%) and the significantly lower (nearly halved) percentages of C18:1 (n7 + n9) (14% vs. 25-26%) were demonstrated in comparison to the midbrain. In addition, the lower relative amounts of gondoic acid and ARA were found in the cortex. Synuclein-deficiency did not course any changes in FA composition of PS in midbrain and cortex ( Figure 6).
Analysis of FA in PC showed the domination of palmitate in the midbrain (34% in WT and 37% in TKO mice) and in the cortex (42% in WT and 44% in TKO mice) of the studied animals. In both brain regions, palmitate and oleate were also determined in the considerable amounts in PC with the slightly lower levels of these acids in the cortex (13% vs. 17%, and 6% vs. 8%, respectively, in WT mice). ARA, DHA and vaccenic acid were present in the range of 4-8% of the total FA in midbrain and cortex of the studied animals. Small amounts of adrenic, gondoic and isomers of C16:1 were determined in this phospholipid from both brain regions. In the midbrain PC of synuclein-free mice, the proportions of DHA, gondoic and stearic were reduced in comparison to WT mice.
In PE of the midbrain (the other major phospholipid of brain tissue), the dominant FA were C18:1 (22% as two combined isomers), DHA (18%), stearate (16% in WT) and ARA (10-11%). In addition, significant amounts of plasmalogen form of PE, estimated as percentages of dimethyl acetals (DMA), were detected in the proportions of 2.1% of C16:0 DMA and 14% C18:0 + C18:1 DMA in the midbrain with the lower amount of the latter in the cortex (8% in WT mice). PE also contains moderate amounts of palmitic, gondoic, and adrenic acids, as well as a minor quantity of two isomers of C16:1 ( Figure 6). A reduction in C18 DMA (from 14.2 ± 1.0% to 11.4 ± 1.3%) in the midbrain as a result of synucleindeficiency was revealed when comparing PE from WT mice with TKO mutants. This decline was accompanied by a little rise (1-2%) in the relative concentrations of saturated palmitic and stearic acids. In the cortex, FA composition of this lipid was unaffected by synuclein ablation in the studied mouse model ( Figure 6).
Cardiolipin isolated from the brain tissues contains oleate as a dominant FA (39% in the midbrain and 34% in the cortex of WT) followed by ARA, DHA and stearate in the range from 17 to 9% in the midbrain and from 18% and 11% in the cortex of WT mice ( Figure 6). Palmitate, isomers of C16:1, vaccenate and gondoic acid were determined in moderate amounts (2-7%) in this lipid from both brain regions. Adrenic acid was present in the relatively small concentrations, 1% and 2% in the midbrain and cortex, respectively. It should be noted that LA, which is a predominant FA in cardiolipin from the liver, was not detected in cardiolipin from the brain samples in this work (as well as in the other analysed polar lipids). These findings are in accordance with the previous studies (see Discussion). Synuclein deficiency affected only FA composition in cardiolipin from the cortex with no changes found in the midbrain. The relative amount of oleic acid was increased in TKO in comparison to WT animals (37.6 ± 0.9% vs. 33.5 ± 0.5%) for the expenses of the level of vaccenic acid.
The FA profile in SM was characterised by the large amounts of saturated FA, namely, stearate (63% in the midbrain and 80% in the cortex from WT mice) and the moderate percentages of palmitic, lignoceric, behenic and arachidic acids (4-7% in the midbrain from WT mice). The relative concentrations of three latter FA were lower in SM of the cortex from these animals. Monoenoic acids, including nervonic acid and isomers of C16:1 and C18:1 acids, were another group of the acyl chains detected in this lipid with nervonic acid being dominant and counted for 13% in the midbrain and for 9% in the cortex from WT mice. The relative concentration of this acid was much lower (9% vs. 13%) in the cortex, whereas the level of stearic acid was significantly higher (80% vs. 63%) in the cortex when compared to the midbrain from WT mice. In the midbrain of synuclein-free mice, the subtle reductions in the levels of C18:1 isomers (from 1.6 ± 0.4% to 0.7 ± 0.2%) and arachidic acid (from 3.7 ± 0.5% to 2.2 ± 0.4%) were noted. No changes in the FA composition of SM were found in the cortex in TKO mutants. The composition of the conventional (non-hydroxylated) FA of two major brain glycosphidgolipids, i.e., sulfatides and cerebrosides, is shown in Figure 6. As indicative for these lipid classes, SL and CER contain saturated and monounsaturated (n-9) FA in the range of carbon chains from C16 to C26. Stearic and nervonic acids were the major FA in SL from midbrain and cortex ranging from 22 to 28%. The substantial proportions of oleic, hexacosenoic and palmitic acids were found in this sphingolipid from both brain regions. Some lower but the appreciable amounts of 9-hexadecenoic, arachidic, gondoic, behenic, eruric and lignoceric acids were detected in SL in the brain of the studied animals. In TKO mice, the relative concentrations of lignoceric (7.5 ± 1.1%), nervonic (21.8 ± 2.0%), and hexacosenoic (5.7 ± 0.1%) acids were reduced in comparison to those in WT mice (10.4 ± 1.1%; 25.4 ± 1.7%; 7.5 ± 0.4%, respectively). Analysis of FA in SL of the cortex revealed only a small decline in the proportion of hexacosenoic acid from 7.0 ± 0.2% in WT mice to 6.2 ± 0.3% in synuclein-free mice.
Three very-long-chain FA, namely, lignoceric (9-13%), nervonic (30-33%) and hexacosenoic (8-16%) acids were predominant in CER isolated from both brain regions of the studied animals ( Figure 6). When comparing the FA profiles of this lipid class between midbrain and cortex, the latter was characterised by the higher levels of 9-hexadecenoicstearic, stearic, lignoceric and tricosanoic (C23:0) acids and the lower percentages of hexacosenoic acid. Synuclein-deficiency resulted in the reduced amounts of hexacosenoic acid in CER isolated from the midbrain from 15.6 ± 1.2% in WT to 12.7 ± 1.1% in TKO mice. No changes in CER acyl chains from the cortex between WT and synuclein-free animals were noted in our study ( Figure 6).

Discussion
Mammalian cells produce a remarkable diversity of fatty acids and lipids, but deeper understanding of molecular complexity of lipids in cells is arising from the technological advances offered by lipidomics [53,54]. Such complexity is the result of many networking processes including the synthesis, trafficking and turnover of lipid compounds which are strongly regulated in different ways in the different cell types and tissues [55].
An involvement of synucleins in the modulation of lipid metabolism has been shown by numerous studies [40][41][42][43][44]48]. Special attention has been paid to the complex interactions between lipids (fatty acids and membrane lipids) and synucleins, which have been briefly outlined in the Introduction. Indeed, the previous studies on synuclein-lipid binding have provided pivotal information on the structural properties of these proteins and membrane interactions, but studies of membrane lipids/proteins, using simplified mixtures of the common lipids and in vitro approaches, may not adequately reveal the role of synucleins in membrane binding, reorganisation or signaling. In particular, a link between dysregulation of lipid metabolism and the pathology of Parkinson's disease has been widely recognised [56,57] and more complex research on the involvement of synucleins in modulation of lipids on the tissue level is needed on the role of synucleins in PD and in the cell in general. Based on this suggestion, we analysed and compared the lipid and fatty acid profiles of plasma, liver and two brain regions in triple-synuclein knockout mice with these parameters in WT animals.
The liver plays a key role in lipid metabolism and, importantly, a special role in supplying FA to the central nervous system. Lipid metabolism in the liver affects the concentration and composition of secreted lipoproteins (LP) in the plasma and, eventually, the lipids and FA in peripheral tissues.
Our study showed that lipid content in the liver from TKO mice was significantly lower than that in WT animals (Figure 1). This reduction was on the account of the decreased TAG and SE accumulation with no changes (only relative, as %) of TPL (Figure 2). It should be noted that triple synuclein knockout mice show no signs of the dysfunction of their liver or cardiovascular system. TAG in hepatocytes, depending on the energy demand, are stored in LD and also packed into lipoproteins which are secreted and hydrolysed by lipoprotein lipases in the vascular system. PL synthesis supplies the lipid substrates for VLDL as well as contributes to the biogenesis and the maintenance of ER/Golgi membranes. This has an impact on both, the cargo and the trafficking machinery for VLDL assembly and secretion [58]. PC, cholesteryl esters, FA and TAG affect the VLDL assembly [59]. In our study, no changes in the circulating PL and TAG were found suggesting that synuclein ablation did not affect the amounts of secreted lipids as VLDL compounds in TKO mice despite of the reduced TAG accumulation in the liver, while the FA compositions of circulating lipids were significantly altered. In our opinion, one of the main findings of this work is a remarkable modification of PUFA synthesis and esterification of the major polar lipids by PUFA acyl chains in the liver that was accompanied by an enormous reduction of these FA in the plasma polar lipids. The observed changes will affect the structure/fluidity of surface monolayer of VLDL, their density, the lipases access to TAG and SE during hydrolysis of VLDL and, consequently and more importantly, the supply of PUFA to the peripheral tissues.
Several mechanisms of synuclein involvement in the modulation of FA and lipid metabolism have been suggested. This protein has structural similarities to the class A2 lipoproteins and some sequence similarity to FABP. In addition, it is present in the large quantities in microsomes, where the synthesis of complex lipids takes place. Since in the brain, but not in the liver, a specific role of α-synuclein in FA uptake and trafficking has been demonstrated, it was hypothesised that α-synuclein may function as FABP in the CNS [60]. For instance, its deletion decreased the palmitic acid and ARA uptake and altered the incorporation of these FA into individual PL classes in mouse brain [60].
Some direct evidence for this hypothesis was provided by a study demonstrating that in α-synuclein-deficient astrocytes, the cellular lipid pool has been altered by reducing the uptake of palmitic acid similarly to the effect of FABP3 in cell cultures [61,62]. However, the binding affinity of α-synuclein to PUFA was two orders of magnitude much less than that for FABP; thereby, the role of α-synuclein in modulation of ER-located long chain-acyl-CoA synthetases (ACSL) has been suggested [46,60]. ACSLs are essential in the complex lipid biosynthesis and in FA targeting for incorporation into the specific lipid pools.
The metabolic fates of various acyl-CoA are determined by a network of proteins that channel them toward or away from the specific pathways to promote FA partitioning [63]. Hepatic FA partitioning requires the interaction of ACSL1 with other specific proteins, such as peroxisomal and lipid droplet proteins, tethering proteins and vesicle proteins that uncovers a dynamic role for ACSL1 in organelle and LD interactions. Proteins involved in lipid metabolism were also identified, including acyl-CoA-binding proteins (ACBP) and ceramide synthase (CerS) isoforms 2 and 5 [64]. Protein complexes composed of several ACSL isoforms have been found in mitochondria. They include ACSL3 and several SNARE proteins, such as SNAP23 (synaptosomal-associated protein 23), VAMP (vesicle-associated membrane protein) and syntaxin 17, suggesting a role for ACSL at membrane contact sites between mitochondria and other organelles or vesicles. In the liver, syntaxin 17 interacts with ACSL3, but not with LD formation-unrelated ACSL1 or ACSL4, through its SNARE domain promoting ACSL3 translocation to the ER to the surface of LD [65]. It should be stressed that more research needed to reveal a role of synuclein deficiency in mitochondrial PUFA synthesis and turnover, since mitochondrial dysfunction is a critical step in the nigral degeneration in PD.
We suggest that in the absence of all synucleins, ACSL4 activity is reduced impairing the synthesis of PUFA-CoA and their incorporation into hepatic phospholipids, namely, ARA to PC, PS and PI and DHA in PE. As to the possible role of synucleins in TAG storage, VLDL synthesis and secretion in blood, it may be connected to the well-known interactions of synucleins with SNARE proteins (which is in turn also interact with ACSL isoforms) that may modulate the formation and trafficking of TAG/LD and phospholipids to/from the ER/Golgi/mitochondria/peroxisomes. The lipoprotein assembly is regulated by proteins by mediating the interactions with receptors, enzymes and lipid transport proteins. The following components of the SNARE play a role in docking and fusion of the VLDL transport vesicles: Sec22b (vesicle SNARE), syntaxin 5, rBet1 and Gos28 (target membrane SNARE) [66]. Thus, it is tenable to suggest that synucleins can be involved in VLDL synthesis via several anticipated mechanisms. The assembling of VLDL with the trace amounts of PUFA in their polar lipids, that ensured by the trace levels of PUFA in the plasma TPL, implies the mechanisms of specific binding and/or preferential transport of phospholipids esterified mainly with palmitic and palmitoleic acids but not with PUFA to the nascent VLDL. It is noteworthy, that the degree of PUFA esterification of steryl esters in the plasma lipoproteins is substantially higher than that in the liver, and it was not altered by synuclein ablation (Table S3). In additional to the de novo synthesis of SE, the mechanism of SE formation via transferring 2-acyl groups from phospholipids (the sn-2 position of glycerol backbones is predominantly esterified by PUFA) to cholesterol also exists. The further (onward the lipoprotein secretion pathway) remarkable reduction of PUFA levels in plasma phospholipids can hereby be explained in TKO mice, where PUFA incorporation into the liver total polar lipids is already impaired. The circulation of highly unsaturated SE in the blood stream of synuclein mutants may ensure a sufficient PUFA supply to peripheral tissues required for the maintaining lipid synthesis and homeostasis despite of the significantly reduced total pool of the circulating PUFA. The increased plasma concentration of SE also contributes to the plasma PUFA level.
Although the plasma lipoproteins are the major transport medium for esterified FA, the appreciable amounts of FA are transported in non-esterified form as bound to albumin. In our study, the total pool of these free FA was significantly reduced in synuclein-ablated animals with a remarkable reduction in the circulating free ARA and DHA (Figures 2 and 3). Passive diffusion across the plasma membrane is well-known to be modulated by FABPs which possess the acyl-CoA synthetase activities that in turn may be regulated by synucleins (as discussed above). Thus, synuclein deficiency markedly affects the hepatic lipid homeostasis as well as the diffusion and transport of PUFA to peripheral tissues. Elucidation of the precise mechanisms of these actions deserves further investigations. In addition, the interaction of α-synuclein with the plasma lipoproteins has been well-documented suggesting an important role the synuclein-related particles in lipid and fatty acid transport in the health conditions and in diseases [67,68]. The brain is highly enriched with PUFAs, especially with ARA and DHA, and their uptake from the circulating lipid pool was previously postulated to be essential to maintain a homeostatic pool of LCPUFA in the nervous tissues. Further studies using rodents demonstrated that the synthesis from the essential precursors, namely, LA and ALA, could provide the sufficient ARA and DHA levels for the estimated adult brain requirement for these LCPUFA [69,70]. The synthesis of these LCPUFA from their precursors takes place primarily in the hepatocytes and then transported, as mentioned above, in both, the esterified and non-esterified forms, to the nervous system.
Besides an uptake of the preformed PUFA from the circulating blood, the appreciable amounts of ARA and DHA can be synthesised from their uptaken precursors (LA, DGLA and ALA) in the microvascular endothelium, oligodendrocytes and astrocytes [71,72]. Neurons themselves are inefficient in the lipid synthesis, although they showed a limited capacity to produce DHA from the omega-3 precursors under the conditions of low dietary supply as well as some elongation of the shorter chain precursors to produce DHA and ARA [71].
PUFA play several physiological roles, including the generally known structural role, an importance in the energy storage and production, inflammation and in cell signaling [73]. They regulate the gene expression by affecting the transcription factors, such as hepatic peroxisome proliferator-activated receptor alpha (PPAR-α), sterol regulatory elementbinding protein 1 (SREBP-1), carbohydrate-responsive element-binding protein (ChREBP) and max-like protein X (MLX) that regulate proteins which are involved in the lipid synthesis and oxidation, and lipoprotein secretion [74]. At the subcellular level, PUFAs modulate the membrane functioning via influencing lipid rafts, where they alter the raft structural and dynamic characteristics, such as domain size and membrane order, most likely by the lateral segregation of LCPUFA-rich domains from cholesterol-rich membrane regions [73]. The neurophysiological outcomes are substantial because lipid rafts participate in neurotransmission, including the regulation of dopamine transporters, dopamine uptake and dopamine D 2 receptor oligomerization [75].
Polar lipids in the brain are characterised by the very high structural complexity due to a substantial number of the different molecular species of glycerophospholipids and sphingolipids. These two main lipid classes play an important structural and functional role in the cell membranes in general and in lipid rafts. In the midbrain of TKO mice, the levels of C24-C26 sulfatides and cerebrosides were reduced indicating the altered activity of CerS2 that catalyses the acylation of sphingolipids with these long acyl chains. The mechanism of the modulation of CerS2 activity may be connected to the similarity of synucleins to the fatty acid binding proteins or ACSLs, discussed above, since the regulation of very-long acyl chain ceramide synthesis by acyl-CoA binding protein has been demonstrated in mice model work [76].
In the midbrain of TKO mice, a significant increase in PC with a concomitant reduction of cerebrosides was noted a result of synuclein deficiency. Not only the accumulation of these lipid classes was altered, but the remarkable alterations of acyl chains, e.g., decreased DHA in PC and C26:1n9 in ceramides, were also demonstrated. Since the level of DHA was not altered in the midbrain in TKO mutants, we suggest that the channeling of DHA to PC, not its synthesis in the brain, was specifically affected decreasing the unsaturation level of this major brain phospholipid by synuclein deficiency. It is generally accepted, that the changes in membrane fluidity can alter the equilibrium of free versus the membrane bound α-synuclein, and therefore impact the aggregation of α-synuclein into amyloid fibrils. Recently, the importance of the unsaturation level of zwitterionic PC (in addition to previously studied anionic phospholipids) for the conformation and aggregation of N-terminally acetylated α-synuclein has been evidenced [24]. We showed that in the brain, the relative amounts of DHA-enriched PS and its fatty acids were not affected by synuclein deficiency. The importance of this phospholipid in the interaction with N-terminal of α-synuclein has been found to facilitate the SNARE complex formation that was critical for vesicle docking using the single-vesicle docking/fusion assays [77]. Moreover, DHA-enriched PC and PS exhibited the different improvements on 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced mice with PD by the conducting behavioral experiments: DHA-PS showed the more pronounced effects on PD symptoms and increased the number of dopaminergic neurons in comparison to DHA-PC [78]. As to the additional role of DHA in the brain, its endogenously synthesised endocannabinoidlike metabolite, N-docosahexaenoylethanolamine (synaptamide) has emerged as a potent mediator of neurotrophic and neuroprotective effects of DHA [69]. It is interesting that N-arachidonylethanolamine (anandamide) synthesised from ARA, being a structural analogue of synaptamide, possesses the distinctive neuro-bioactivity modulating the brain reward circuitry and neuroinflammation [79].
In the liver, PS was enriched with ARA that markedly reduced in TKO mice. This fatty acid is known to stimulate SNARE complex formation and exocytosis, and α-synuclein efficiently blocks ARA-induced SNARE interactions [80].
In spite of the remarkable PUFA modifications in the individual lipid classes in liver and plasma of TKO mice, we can conclude that the total supply of PUFA (both, LCPUFA and their short chain essential precursors) was sufficient to maintain the characteristically high level of LCPUFA in midbrain and cortex of the studied mutants. More research on the synthesis of PUFA in the liver and brain cells/tissues is required to understand the mechanisms of its metabolic regulation in the synuclein-free mouse models.
In the midbrain, a marked reduction in ethanolamine plasmalogens in synuclein free mice can be considered as an important finding, since the abnormal ether lipid levels has been previously implicated in the neurological disorders [81]. Plasmalogens is a class of ether lipids with an ether bond attached to sn-1 position of the glycerol backbone as opposite to an ester bond of the more common diacyl phospholipids [82]. Plasmalogens are present in mammalian plasma and intracellular membranes. Nervous, immune and cardiovascular systems are enriched in these lipids, where plasmalogens have been proposed to influence membrane dynamics and intracellular signaling being important components of lipid rafts and also act as membrane antioxidants and the reservoir of PUFA [82]. The reduced brain plasmalogen level has been demonstrated in various neurodegenerative and metabolic deregulated diseases which were associated with the increased lipid oxidation. Regarding PD, patients with this disease showed a significant decrease in the post-mortem frontal cortex lipid raft plasmalogens level [82,83]. The results of other studies also showed an importance of plasmalogens for brain functions, albeit a contrary to the hypothesis that the synthesis of these lipids predominantly occurred in the brain and less likely supplied by circulation from peripheral tissues [84]. In our study, the level of plasma DMA among fatty acids of the total polar lipids was non-detectable because of the small sample volumes. Lipidomic approach, which is under our consideration for the future work in this direction, will allow to estimate the level of circulating plasmalogens in TKO mice. From the results obtained in the present study and based on the notion that plasmalogen may function as the potential endogenous antioxidants, we suggest that the reduced level of plasmalogen PE in the midbrain of TKO mice makes the cells more susceptible to the oxidative stress. In support, there is some evidence of involving of plasmalogens in scavenging a variety of reactive oxygen species and in protecting of the unsaturated membrane lipids from oxidation by singlet oxygen [82]. Moreover, the plasmalogen supplementation protected striatal dopamine neurons that degenerated in the response to MPTP-treatment in mice, a PD model [85]. Our results once again emphasise that not only pathological accumulation of synucleins but also their ablation has a profound effect on lipid metabolism. The present study underlined several lipid biosynthetic pathways that were modified by synuclein deficiency in the TKO mouse model. Future research in this direction by studying the highlighted "targets" of synuclein actions as the modulators of lipid synthesis will reveal the mechanisms and the physiological relevance of the crosstalk between synucleins and lipids not only in the brain but at the level of organism in general. Using lipidomic approach for studying lipid rafts (despite of some controversy around the existence of such membrane microdomains [86]), plasmalogen and sphingolipid synthesis, lipoprotein formation and the circulating lipids and fatty acids, as well as for the better understanding of lipid synthesis in astrocytes, will undoubtedly integrate our knowledge of synuclein functions in the cells.

Animals
A line of synuclein-free mice that are null mutants for the all three genes coding for α-, b-, and γ-synuclein was described early [4]. The colony was kept on C57Bl6J (Charles River) background and wild-type control mice on the same C57Bl6J (Charles River) background were bred in the Cardiff University Transgenic Animal Unit. Mice were maintained in conventional open-lid cages with ad libitum access to standard chow and water. Animals were fed a special diet DOI 58Y2 with energy from fats (labDiets). Fatty acid composition of the diet is given in our previous publication [48]. It contained oleic acid as a major compound (35% of total diet FA) followed by linoleic (25%), palmitic (20%), and stearic (13%) acids. The relative amount of a-linolenic (an essential n-3 fatty acid) did not exceed 2.5% of total FA. Lipids accounted for 10% of total energy in this diet.
Genotyping was carried out by conventional PCR as described previously [4]. Animals (13-month-old male mice) were sacrificed by cervical dislocation and tissues were collected on ice and processed. All animal work was carried out in accordance with the United'Kingdom Animals (Scientific Procedures) Act (1986).

Lipid Analysis
The following tissues were used for lipid and fatty acid analysis: plasma, liver, and two brain areas (cortex and midbrain). Lipids were extracted immediately by Folch's method [87], and lipid extracts were stored at −20 • C under nitrogen atmosphere prior to further analysis. Non-polar lipids were separated using one-dimensional TLC on 10 × 10 cm 2 silica gel G plates (Merck KGaA, Darmstadt, Germany) in the solvent system hexane/diethyl ether/acetic acid (80:20:1, v/v/v). Polar lipids were separated by two-dimensional TLC on 10 × 10 cm 2 1.2% boric acid-impregnated silica gel plates using chloroform/methanol/ammonium hydroxide (65:25:4, v/v/v) in the first dimension and n-butanol/acetic acid/water (90:20:20, v/v/v) in the second direction. Lipids were identified by reference to authentic standards and by using specific colour reagent [88]. In addition, several lipid identifications were performed using HPLC-MS-MS (details can be found in our previous publication [48]. Plates were sprayed with 0.2% (w/l) 8-anilino-4naphtholenesulphonic acid in dry methanol and viewed under U.V. light to reveal lipids.
Individual lipids were scraped from the plates and their fatty acid compositions and contents were determined by gas chromatography (GC) with heptadecanoate (C17:0) as an internal standard.

Fatty Acid Analysis
Aliquots of the total lipid extracts and individual lipid classes separated by TLC were used for fatty acid methyl esters (FAMEs) preparation. FAMEs were prepared using a solution of 2.5% H 2 SO 4 in dry methanol/toluene (2:1, v/v). Samples were heated to 70 • C for 2 h, with occasional vortexing. FAMEs were extracted by two additions of 3 mL of hexane. The hexane fractions were transferred into clean glass tubes, evaporated under a stream of nitrogen, and reconstituted in 50 µL of hexane. All reagents for fatty acid extraction and derivatisation were of HPLC quality.

Gas Chromatography
FAMEs were analysed using a Clarus 500 GC with a flame ionisation detector (Perkin-Elmer, Norwalk, CT, USA). A Perkin-Elmer Elite 225 column (30 m × 0.32 mm × 0.25 µm) was used with a nitrogen carrier gas constant pressure of 20 psi. The oven temperature program began at 170 • C for 3 min, was ramped to 220 • C at 4 • C/min, and was then held for 30 min. FAMEs were identified by comparison of the peak retention times with those of a GC-411 standard mixture (Nu-Chek Prep. Inc., Elysian, MN, USA). Perkin-Elmer TotalChrom software was used for data acquisition and calculation.

Statistical Analysis
Statistical significance between groups was determined by Mann-Whitney U test. A probability of error less than 5% was considered significant (i.e., p < 0.05). The data are reported as mean ± standard deviation of six independent determinations.

Conclusions
Our data demonstrate that synuclein deficiency changes the PUFA synthesis and the storage lipid accumulation in the liver. It also results in the reduction of plasmalogens and CER that are the principal compounds of lipid rafts in many tissues. An ablation of all three synuclein family members causes more profound changes in lipid metabolism than changes previously shown to be associated with γ-synuclein deficiency alone.
Supplementary Materials: The following are available online, Table S1: Fatty acid composition of total lipids (% of total fatty acids) in plasma and liver from triple-synuclein null mu-tant (TKO) and wild-type (WT) mice, Table S2: Fatty acid composition of triacylglycerols (% of TAG fatty acids) in plasma and liver from triple-synuclein null mutant (TKO) and wild-type (WT) mice, Table S3: Fatty acid composition of steryl esters (% of SE fatty acids) in plasma and liver from triple-synuclein null mu-tant (TKO) and wild-type (WT) mice, Table S4: Fatty acid composition of total lipids (% of total fatty acids) in midbrain and cortex from triple-synuclein null mutant (TKO) and wild-type (WT) mice.