Acyl-CoA synthetase activity links wild-type but not mutant α-synuclein to brain arachidonate metabolism

Abstract

Because α-synuclein (Snca) has a role in brain lipid metabolism, we determined the impact that the loss of α-synuclein had on brain arachidonic acid (20:4n-6) metabolism in vivo using Snca^-/- mice. We measured [1-¹⁴C]20:4n-6 incorporation and turnover kinetics in brain phospholipids using an established steady-state kinetic model. Liver was used as a negative control and no changes were observed between groups. In Snca^-/- brains, there was a marked reduction in 20:4n-6-CoA mass and in microsomal acyl-CoA synthetases (Acsl) activity toward 20:4n-6. Microsomal Acsl activity was completely restored after the addition of exogenous wt mouse or human α-synuclein, but not by A30P, E46K, and A53T forms of α-synuclein. Acsl and acyl-CoA hydrolase expression was not different between groups. The incorporation and turnover of 20:4n-6 into brain phospholipid pools was markedly reduced. The dilution coefficient lambda, which indicates 20:4n-6 recycling between the acyl-CoA pool and brain phospholipids, was increased 3.3-fold, indicating more 20:4n-6 was entering the 20:4n-6-CoA pool from the plasma relative to that being recycled from the phospholipids. This is consistent with the reduction in Acsl activity observed in the Snca^-/- mice. Using titration microcalorimetry, we determined that α-synuclein bound free 20:4n-6 (K_d of 3.7 μM), but did not bind 20:4n-6-CoA. These data suggest α-synuclein is involved in substrate presentation to Acsl rather than product removal. In summary, our data demonstrate that α-synuclein has a major role in brain 20:4n-6 metabolism through its modulation of endoplasmic reticulum localized acyl-CoA synthetase activity, although mutants forms of α-synuclein fail to restore this activity.

α-Synuclein is a 140 amino acid soluble protein that is highly expressed in the central nervous system (1, 2) and is abundant in presynaptic terminals of neurons (1, 3-5). α-Synuclein is also found in other regions of neurons, in astrocytes, and in oligodendroglia (6-11). Overexpression of and mutations in α-synuclein are associated with early onset Parkinson’s disease (12-15) and other neurodegenerative diseases (16-20). Despite this association with neurodegenerative diseases, the physiological function of this protein remains unclear.

Several lines of evidence suggest that α-synuclein can influence brain lipid metabolism. It has structural similarities to class A2 apolipoproteins (21, 22) and to fatty acid binding proteins (23), suggesting that α-synuclein may alter intracellular lipid trafficking, the regulation of lipid metabolism, and may act to stabilize lipid membranes. α-Synuclein binds to small phospholipid vesicles (22, 24, 25) and to brain vesicles (26). Consistent with this binding, the lack of α-synuclein decreases the resting/reserve pool of synaptic vesicles (27, 28). Although the direct binding of fatty acids is controversial (23, 29), recent studies indicate a strong potential for an important role of α-synuclein in fatty acid uptake and metabolism (11, 29, 30). In astrocytes lacking α-synuclein palmitic (16:0) and arachidonic (20:4n-6) acid uptake and metabolism is depressed through an unknown mechanism (11). Lack of α-synuclein in brain depresses 16:0 uptake and significantly alters its metabolism (29), although the impact on brain 20:4n-6 uptake and metabolism is unknown.

In addition, α-synuclein modulates phospholipase C and D activities in vitro (22, 31-33), suggesting a role in lipid-mediated signal transduction. Studies in vivo also demonstrate that α-synuclein affects enzymes involved in lipid metabolism, because α-synuclein overexpression leads to phospholipase D inhibition in yeast (34) and down regulates expression of phospholipase A₂ and of long chain fatty acid CoA synthetase in Drosophila (35). These additional findings suggest that α-synuclein impacts lipid-mediated signal transduction, including 20:4n-6 release. Release of 20:4n-6 during signal transduction is crucial for proper CNS function (36-41) and pathophysiological responses (42-45).

To address the potential role for α-synuclein in brain 20:4n-6 uptake and metabolism, we used steady-state kinetic modeling of [1-¹⁴C]20:4n-6 metabolism in vivo coupled with studies using enzyme assays and mRNA expression of key fatty acid metabolic enzymes. These data show, for the first time, a mechanistic explanation for the impact of α-synuclein deficiency on brain lipid metabolism. Herein, we demonstrate that, in the absence of α-synuclein, brain microsomal acyl-CoA synthetase activity was decreased, which in turn reduced 20:4n-6 uptake and turnover in brain phospholipid pools. . More importantly, addition of wt human or mouse α-synuclein completely restored acyl-CoA synthetase activity in these microsomes, while the A30P, E46K, and A53T mutant forms of α-synuclein failed to restore activity.

MATERIALS and METHODS

Mice

This study was conducted in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals (NIH publication 80-23) and under an animal protocol approved by the IACUC at the University of North Dakota (Protocol #0110-1). α-Synuclein gene-ablated mice (Snca^-/-) were generated from 129/SvEv strain by gene targeted deletion (28). Male mice (25-30 g) were maintained on standard laboratory chow diet and water ad libitum. The ages of the mice used in this study were between 9-11 months in both groups.

Mouse surgery and tracer infusion

The mouse surgery and tracer infusion was performed as previously described (29, 46, 47). Briefly, [1-¹⁴C]20:4n-6 (Moravek Biochemical, Brea, CA) was solubilized in 5 mM HEPES (pH 7.4) buffer containing “essentially fatty acid free” bovine serum albumin (50 mg/ml, Sigma Chemical Co., St. Louis, MO). Awake (3-4 hours post-operative) fasted, male mice (Snca^+/+ or Snca^-/-) were infused with 170 μCi/kg of [1-¹⁴C]20:4n-6 into the femoral vein over 10 min at a constant rate of 30 μL/min. Prior to and during the infusion, arterial blood samples (∼20 μL) were taken to determine plasma radioactivity and confirm steady-state plasma radioactivity. Following infusion, each mouse was killed using pentobarbital (100 mg/kg, i.v.) and immediately subjected to focused microwave irradiation to heat denature enzymes in situ. Brain and liver were rapidly removed, frozen in liquid nitrogen and pulverized under liquid nitrogen temperatures to a fine, homogeneous powder.

Lipid extraction

Lipids from the tissue powder, plasma, and blood samples were extracted using a two-phase extraction procedure (48). The radioactivity in the aqueous and organic fractions was determined by liquid scintillation counting. The extracts were concentrated under a stream of N₂ at 40°C and dissolved in n-hexane : 2-propanol : water (56.7 : 37.8 : 5.5 by vol.).

Thin layer chromatography

Tissue phospholipids (PL) were separated by thin-layer chromatography (TLC) on heat-activated Whatman silica gel-60 plates (20 × 20 cm, 250 μm) and developed in chloroform : methanol : acetic acid : water (55 : 37.5 : 3 : 2 by vol.) (49). Brain, liver, and plasma neutral lipids (NL) were separated by TLC on heat-activated Whatman silica gel-60 plates (20 × 20 cm, 250 μm) and developed in petroleum ether : diethyl ether : acetic acid (75 : 25 : 1.3 by vol.) (50). Once separated, individual PL and NL were isolated and used to quantify individual PL and NL class fatty acid mass and radioactivity.

Gas liquid chromatography

The 20:4n-6 mass in phosphatidylinositol (PtdIns), phosphatidylserine (PtdSer), choline glycerophospholipids (ChoGpl), and ethanolamine glycerophospholipids (EtnGpl) was determined following base-catalyzed transesterification to form the fatty acid methyl esters (51).

Sphingomyelin (CerPCho) individual fatty acid mass, as well as mass of free fatty acids in plasma, brain, and liver, were measured by gas liquid chromatography (GLC) after acid-catalyzed esterification (52). The gas liquid chromatograph (Trace GC, ThermoElectron, Austin, TX) was equipped with a capillary column (SP 2330; 30 m × 0.32 mm i.d., Supelco, Bellefonte, PA) and a flame ionization detector. Fatty acids were quantified using a standard curve from commercially purchased standards (NuChek Prep, Elysian, MN) and 17:0 was the internal standard (29).

Acyl-CoA extraction, separation and quantification

Acyl-CoA from brain and liver were extracted and purified using a solid-phase extraction procedure and separated by HPLC on C-18(2) column (Luna, Phenomenex, Torrance, CA) (53). The system was controlled by Beckman 127 solvent module (Fullerton, CA, USA). The eluent was monitored at 260 nm using a Beckman 166 UV/Vis detector. (53). Mass of 20:4n-6-CoA was determined using a standard curve from commercially purchased standards (Sigma Chemical Co., St. Louis, MO) and 17:0-CoA was the internal standard. Radioactivity of 20:4n-6-CoA was determined by liquid scintillation counting.

Liquid scintillation counting

Samples were placed into 20-ml liquid scintillation vials, and 0.5 ml of H₂O was added, followed by 10 ml of Scintiverse BD (Fisher). After mixing, the samples were quantified by liquid scintillation counting using a Beckman LS5000 CE liquid scintillation counter (Beckman Instruments, Fullerton, CA) at least 1 h after the addition of the liquid scintillation mixture.

Kinetic analysis

Our study was performed at steady-state conditions using a quantitative kinetic model as previously described (29, 36, 54-56). The main aspect of this model is that it accounts for fatty acid recycling between phospholipid pools and acyl-CoA pool. The dilution coefficient, λ, represents this recycling and is defined as the steady-state specific activity of brain 20:4n-6-CoA relative to specific activity of plasma 20:4n-6.

Titration microcalorimetry

Titration microcalorimetry was carried out as previously described (29, 57). Briefly, recombinant α-synuclein (50 μM) was titrated with 20:4n-6 at 27°C in 20 mM potassium phosphate buffer containing 50 mM KCl (pH 7.2). The sample was titrated with 25-30 aliquots of the fatty acid containing solution (1 mM) added at 3 min intervals. Raw data were integrated using the supplied ORIGIN software (MicroCal, Inc.) and isotherms were analyzed as previously described (58, 59).

Northern blot analysis

Northern blot analysis was carried out as previously described (60). Briefly, 20 μg of isolated mRNA from Snca^-/- and Snca^+/+ mouse brains was resolved on a 1.2% agarose-formaldehyde gel and transferred to a Biodyne B membrane. Probes were chosen for each Acsl that were for specific regions of the mRNA that differed significantly from each other. Probes were radioactively labelled by random priming using [α-³²P]dCTP and High Prime solution (Roche). A murine cDNA (position 9-969; accession no. BC011106) of the ubiquitously expressed acidic ribosomal phosphoprotein P0 (36B4) was used as a loading control. To detect murine Acsl3, Acsl4, Acsl6, and Bach mRNA, the following probes were used: mouse Acsl3 fragment of 342 bp (nucleotides 1060-1402 of accession no. NM_028817); mouse Acsl4 cDNA fragment of 329 bp (nucleotides 878-1206 of accession no. NM_019477); mouse Acsl6 fragment of 398 bp (nucleotides 877-1274 of accession no. NM_144823); and mouse Bach cDNA fragment of 386 bp (position 559-944 of accession no. NM_133348). The blot was hybridized using ExpressHyb solution (BD Biosciences, Franklin Lakes, NJ) at 68°C and washed to high stringency according to manufacturer’s instructions.

Changes in Acsl6 expression were confirmed using quantitative real time PCR. For each assay, 4 ng cDNA (generated by GeneAmp RNA PCR Kit, Applied Biosystems) was analyzed using the iCycler iQ real-time PCR detection system (Bio-Rad). Mouse Acsl6 expression was quantified using the primers nt877 5′- atagaggactgtggccgaga -3′ (forward) and nt979 5′- ctttggggttccctgttgta -3′ (reverse) (Acc. No. NM_144823) with Sybr-Green. As a control, hydroxy-phosphoribosyl transferase (Hprt) mRNA levels were quantified using the primers nt503, 5′- ctggtaaaacaatgcaaac -3′ (forward), nt622, 5′- caaagtctggcctgtatc -3′ (reverse), nt569, 5′-FAM-caagcttgctggtgaaaagga -DABCYL-3′ (TaqMan probe). Standard curves were generated by serial dilution of plasmids containing Acsl6 or Hprt cDNA (kindly provided by Dr. D.W. Melton, University of Edinburgh, U.K). The thermocycler was programmed: 95°C for 10 min followed by 50 cycles at 95°C for 20 s and 58°C for 50 s for Hprt and 95°C for 10 min followed by 35 cycles at 95°C for 10 s and 60°C for 50 s for Acsl6. At the end of the Acsl6 Sybr-Green-detection, a melt curve was performed for each well to ensure the absence of unspecific PCR products.

Brain homogenate and brain microsomes associated acyl-CoA synthetase activity

Brain microsomal fractions, from wt and snca^-/- mice, containing acyl-CoA synthetases were prepared by centrifugation and long chain fatty acyl-CoA synthetase activity was measured (61, 62). The protein content in microsomal fraction and whole brain homogenate was measured using a dye-binding assay (63). Acsl activity was measured in three different groups: 1). microsomes from wt mice; 2). microsomes from snca^-/- mice; and 3). microsomes from snca^-/- mice incubated with 1.0% wt mouse α-synuclein for 30 min. In all groups, Acsl activity was measured in whole brain homogenate or in microsomal homogenate by adding 20 μg of protein to the assay cocktail, containing 670 μM Triton X-100, having a final [1-¹⁴C]20:4n-6 concentration between 3.6 and 190 μM. The reactions were terminated after incubation at 37°C for 4 min. After extraction of non-esterified 20:4n-6, the radioactivity of 20:4n-6-CoA was determined by liquid scintillation counting. The initial rates of reaction were calculated based on the specific radioactivity of the reaction substrate. The Vmax and Km of Acsl activity with 20:4n-6 as a substrate was calculated using standard Michaelis-Menton kinetic analysis.

Rescue of acyl-CoA synthetase activity

Purified, recombinant human and mouse wt α-synuclein or its mutant forms (A30P, E46K, and A53T) were purified (64) and used to determine if these different forms of α-synuclein could rescue microsomal Acsl activity. To determine which concentration of wt or mutant α-synuclein was the ideal concentration to restore Acsl activity, we incubated microsomes with 0.01 to 10% α-synuclein relative to total microsomal protein. This corresponds to a final concentration for α-synuclein of 0.6-634 μM in the assay. The α-synuclein, in Tris buffer (pH 7.4), was added 30 min prior to incubating the Snca^-/- brain microsomal fractions (20 μg of protein) with [1-¹⁴C]20:4n-6 (190 μM). Acsl activity was determined as stated above. To assess potential endogenous acyl-CoA synthetase activity of α-synuclein, recombinant α-synuclein (1.2 μg) was added to 50 μl of solubilization buffer, which was then added to the assay cocktail.

Western blot analysis of α-synuclein

Brain microsomal protein (40 μg) from Snca^+/+ mice (n=8) was resolved on a 15% polyacrylamide gel and transferred to PVDF membrane (11, 65). The positive control was the recombinant α-synuclein and the antibody used to detect α-synuclein was a monoclonal anti-synuclein antibody 4D6 (1:1,000 Signet, Dedham, MA. Blots were developed using chemiluminiscence super-signal (Amersham, Piscataway, NJ).

Statistics

Statistical significance was assessed using an unpaired, two-way, Student’s t-test, with a p < 0.05 considered to be statistically significant. For microsomal Acsl assays, statistical significance was assessed using a one-way ANOVA and a Tukey-Kramer post-hoc test, with a p<0.05 considered to be significant. The number of mice used per experiment is indicated in the figure and table legends.

RESULTS

α-Synuclein binds 20:4n-6 but not 20:4n-6-CoA

We have shown that α-synuclein does not bind palmitate or oleate using titration microcalorimetry (29), although others have shown oleate binding using the classical Lipidex assay (23). To assess if α-synuclein bound monomeric 20:4n-6, we again used titration microcalorimetry. We determined that α-synuclein bound 20:4n-6 with a K_d of 3.7 ± 1.7 μM (n=3) (Figure 1a), but did not bind 20:4n-6-CoA (Figure 1b). The binding affinity for 20:4n-6 is 1 to 2 orders of magnitude less than that of a classical FABP (66).

Figure 1.

a). α-Synuclein binding of 20:4n-6 as determined using titration microcalorimetry, n=3.

b). Lack of α-synuclein binding of 20:4n-6-CoA as determined using titration microcalorimetry, n=3.

Absence of α-synuclein reduced brain 20:4n-6-CoA mass

We noted that the mass of the brain 20:4n-6-CoA pool was reduced 58% in the Snca^-/-mice, while the mass in liver 20:4n-6-CoA was unchanged (Figure 2a). Because liver tissue does not contain α-synuclein (29), we used it as a negative control in these experiments. This reduction in brain 20:4n-6-CoA mass was not accounted for by reduced incorporation of 20:4n-6 from plasma into the brain 20:4n-6-CoA pool (k*), as these values were equivalent between groups (data not shown). This indicates that the reduced 20:4n-6-CoA pool was not the result of a reduction in plasma derived 20:4n-6, but rather from a potential reduction in 20:4n-6 recycling from brain phospholipid pools. This was confirmed by an increase in the dilution coefficient, λ, in the Snca^-/- brains, indicating an increase in the amount of fatty acid entering the 20:4n-6-CoA pool via the plasma relative to the amount coming from the recycling of 20:4n-6 from endogenous phospholipid pools (Figure 2b). From these results, we would predict that the markedly reduced 20:4n-6-CoA pool would cause a reduction in the incorporation of 20:4n-6-CoA into Snca^-/- brain phospholipid pools.

Figure 2.

a). Mass of 20:4n-6-CoA in brain and liver isolated from snca^-/- and snca^+/+ mice. Values represent mean ± SD, n=8-10. The * indicates statistical significance from snca^+/+ mice, p<0.05.

c). Dilution coefficient λ for snca^-/- and snca^+/+ mice. Values represent mean ± SD, n=8-10. The * indicates statistical significance from snca^+/+ mice, p<0.05.

Absence of α-synuclein decreased brain 20:4n-6-CoA incorporation into phospholipids

We next examined the impact of α-synuclein deficiency on changes in the kinetics of 20:4n-6 incorporation and turnover in brain phospholipid pools. In brain tissue, we found the incorporation coefficient k* for 20:4n-6 entering PtdIns, PtdSer, ChoGpl, and CerPCho was decreased 25%, 34%, 22%, and 50%, respectively, in Snca^-/- compared to Snca^+/+ mice (Table 1). The incorporation rate (J_FA,i) for 20:4n-6-CoA into EtnGpl, PtdIns, PtdSer, ChoGpl, and CerPCho was also decreased 48%, 46%, 41%, 47%, and 66%, respectively, in Snca^-/- compared to Snca^+/+ mice. As a result, the half-life (T_1/2) for 20:4n-6 in brain EtnGpl, PtdIns, PtdSer, ChoGpl, and CerPCho was increased 2.8-fold, 3.3-fold, 2.4-fold, 2.9-fold, and 3.7-fold, respectively, Snca^-/- compared to Snca^+/+ mice. Importantly, there were no significant differences in liver individual phospholipid kinetics between groups (Table 2), indicating that these changes were brain specific.

Table 1.

Effects of α-synuclein ablation on the unilateral incorporation coefficient (k*), rate of incorporation into individual brain phospholipid pools, and turnover in these individual phospholipid pools

	ki* ×10-5 sec-1		JFA,i nmol/hour		T1/2,i hour		Cbr,i nmol/g ww
	Snca+/+	Snca-/-	Snca+/+	Snca-/-	Snca+/+	Snca-/-	Snca+/+	Snca-/-
EtnGpl	2.7±.7	2.5±0.6	102±37	53±28*	11.2±3.7	31.6±12.1*	1802±619	1892±266
PtdIns	6.8±1.8	5.2±1.3*	252±110	135±72*	1.1±0.3	3.6±0.9*	357±72	348±80
PtdSer	1.2±.4	0.8±0.2*	39±16	23±9*	2.7±1.2	6.6±2.1*	110±21	135±33
ChoGpl	9.3±2.3	7.3±1.1*	350±160	186±93*	1.4±0.3	4.0±1.4*	783±163	901±118
CerPCho	0.8±0.6	0.4±0.2*	32±10	11±4*	0.9±0.3	3.3±1.6*	42±6	38±9

Values represent mean ± SD, n=7-9. The * indicates statistical significance from control mice, p<0.05. Abbreviations are EtnGpl - ethanolamine glycerophospholipids; PtdIns - phosphatidylinositol; PtdSer - phosphatidylserine; ChoGpl - choline glycerophospholipids; CerPCho - sphingomyelin; k_i* - incorporation coefficient of 20:4n-6 from plasma into individual phospholipid; J_FA,i - net rate of incorporation from 20:4n-6-CoA into individual phospholipid; T_1/2,i - half life of 20:4n-6 in individual phospholipid; C_br,i - mass of 20:4n-6 in individual phospholipid.

Table 2.

Effects of α-synuclein ablation on the unilateral incorporation coefficient (k*), rate of incorporation into individual liver phospholipid pools, and turnover in these individual phospholipid pools

	ki* ×10-5 sec-1		JFA,i nmol/hour		T1/2,i hour		Cliv,i nmol/g ww
	Snca+/+	Snca-/-	Snca+/+	Snca-/-	Snca+/+	Snca-/-	Snca+/+	Snca-/-
EtnGpl	42.6±5.8	51.9±14.6	551±207	562±164	1.3±0.5	1.2±0.3	989±148	1045±118
PtdIns	24.2±2.9	23.2±5.4	261±46	233±68	1.7±0.6	1.6±0.5	694±116	627±77
PtdSer	10.0±1.8	9.6±2.9	127±28	107±28	1.1±0.3	1.4±0.5	189±30	180±30
ChoGpl	191.7±15.1	208.3±59.8	2443±817	2271±692	0.6±0.2	0.6±0.2	1935±155	1990±82
CerPCho	4.6±1.0	4.5±1.4	57±18	47±12	1.9±0.5	2.2±0.6	166±14	155±24

Values represent mean ± SD, n=7-9. Abbreviations are as found in Table 1.

While this reduction in 20:4n-6-CoA affected the kinetics of incorporation of 20:4n-6 into brain phospholipid pools, it did not cause a reduction in the absolute mass of 20:4n-6 in individual brain phospholipids (Table 1). These results are consistent with a reduction in the 20:4n-6-CoA available as substrate for phospholipid fatty acid remodeling. We then turned our attention to examining whether this effect was mediated by decreased synthesis of 20:4n-6-CoA and 20:4n-6 uptake.

Absence of α-synuclein reduced brain 20:4n-6 uptake

Previously, we demonstrated that α-synuclein deficiency decreased brain 16:0 uptake in vivo (29) and reduced 16:0 and 20:4n-6 uptake in cultured astrocytes (11). We assessed the effect of α-synuclein on brain 20:4n-6 uptake in vivo by infusing Snca^+/+ and Snca^-/- mice with [1-¹⁴C]20:4n-6 (i. v.). There were no differences in input function (integrated plasma radioactivity) for Snca^+/+ (1394 ± 445 nCi × ml^-1 × min^-1) and Snca^-/- (1391 ± 204 nCi × ml^-1 × min^-1) mice, indicating that there were no overt differences in 20:4n-6 metabolism between groups. In brain extracts, the total and organic (lipid containing) phase incorporation coefficient for Snca^-/- mice was modestly decreased 15% and 12%, respectively, (Figure 3). There was no significant difference in tracer entering the brain aqueous compartment (Figure 3), which represents products of β-oxidation (46, 67-69). The lack of difference in β-oxidation adds additional evidence supporting the idea that the changes observed herein are the result of altered ER based Acsl expression or activity, rather than changes in acyl-CoA pools associated with mitochondrial β-oxidation. However, the lack of a large reduction in brain 20:4n-6 uptake suggests 20:4n-6-CoA formation may account for the reduction observed in 20:4n-6-CoA mass.

Figure 3.

Uptake of [1-¹⁴C] 20:4n-6 into brain from snca^-/- and snca^+/+ mice expressed as total uptake or uptake into the aqueous and organic brain fractions. Values were normalized to the plasma curve radioactivity and expressed as the unidirectional incorporation coefficient k*×10^-5 (s^-1). Values were corrected for radioactivity associated with the residual blood left in brain and represent mean ± SD, n=7-9. The * indicates statistical significance from snca^+/+ mice, p<0.05.

Acyl-CoA synthetase and hydrolase expression is unaffected by the lack of α-synuclein

One explanation for the reduction in 20:4n-6-CoA mass would be a reduction in brain acyl-CoA synthetase expression or an increase in brain acyl-CoA hydrolase expression in the Snca^-/- mice. Because Acsl3, 4, and 6 are expressed in brain (70-75), we measured the mRNA levels of these three synthetases using Northern blot hybridization and real time PCR analyses. All three synthetases were equally expressed in Snca^+/+ and Snca^-/- mice (Figure 4a). Expression of Acsl6 was also assessed by real time PCR relative to a housekeeping mRNA for a housekeeping enzyme, Hprt, and was found to be unchanged in Snca^-/- mice (Figure 4b). We also measured expression of type II Acyl-CoA thioesterase (Bach), which is highly expressed in brain (76, 77) and hydrolyzes 20:4n-6-CoA very efficiently (78). There were no differences in Bach expression (Figure 4a). Although there are other thioesterases (Type-I), which hydrolyze 20:4n-6-CoA, they do so with much less efficiency and it seems unlikely that these could account for the observed reduction in brain 20:4n-6-CoA levels. Hence, while acyl-CoA synthetase expression was not altered by α-synuclein deficiency, such measurements do not assess a reduction in enzymic activity.

Figure 4.

a) Northern blot analysis of brain acyl-CoA synthetase and acyl-CoA hydrolase mRNA expression in snca^-/- and snca^+/+ mice. Abbreviations are: Acsl - acyl-CoA synthetase mRNA; Bach - acyl-CoA hydrolase mRNA; 36B4 - acidic ribosomal phosphoprotein P0 (loading control).

b). Quantitative real-time RT-PCR of Acsl6 mRNA isolated from snca^-/- and snca^+/+ mice.

Absence of α-synuclein decreased brain microsomal acyl-CoA synthetase activity

Because α-synuclein deficiency reduced brain 20:4n-6-CoA mass without altering acyl-CoA synthetase and hydrolase expression, we measured acyl-CoA synthetase activities, using 20:4n-6 as a substrate, in whole brain homogenates and in brain microsomal fractions. No differences in acyl-CoA synthetase activity were found in whole brain homogenate (Snca^+/+ 1.6 ± 0.1 and Snca^-/- 1.5 ± 0.1 pmol × min^-1 × mg^-1 (n=8)), consistent with the similar incorporation coefficients from plasma into the 20:4n-6-CoA pools. These data suggested an alteration in the activity of endoplasmic reticulum (microsomal) expressed acyl-CoA synthetases which was confirmed using 20:4n-6 as a substrate (Figure 5). Addition of wt mouse α-synuclein to brain microsomes isolated from snca^-/- mice clearly demonstrates a complete restoration of activity at all but the lowest concentration of 20:4n-6. Indeed, α-synuclein deficiency reduced Vmax and Km for microsomal acyl-CoA synthetase activity, using 20:4n-6 as a substrate, 32% and 16%, respectively, but these values were completely restored by addition of wt mouse α-synuclein (Table 3). Calculating the ratio of Vmax to Km, a situation that may more accurately reflect activity under physiological conditions (75), we demonstrate a significant 19% reduction in this ratio in the microsomes from snca^-/- mice as compared to control mice. Because ER localized acyl-CoA synthetase activity is crucial for 20:4n-6 recycling from phospholipid pools into the 20:4n-6-CoA pool, this reduction in enzymic activity accounts for the observed increase in the dilution coefficient (λ) (Fig. 2b) and the reduction in 20:4n-6 uptake (Figure 3).

Figure 5.

Acsl activity was measured in brain microsomes as described in the Experimental Procedures using 20:4n-6 as a substrate. Values represent mean ± SD, n=8. Samples were from individually prepared microsomes isolated from 8 snca^-/- and snca^+/+ mice. Inset is a representative double reciprocal plot of these data.

Table 3.

	Vmax (pmol × min-1 × mg-1)	Km (μM)	Vmax/Km × 10-6
Snca+/+	15.8±1.8	33.7±3.4	0.47±0.05
Snca-/-	10.8±0.8*	28.4±1.5*	0.38±0.05*
Snca-/- + WT Snca	17.7±2.1	35.9±4.0	0.49±0.05

The Vmax and Km of Acsl activity with 20:4n-6 as a substrate was calculated using standard Michaelis-Menton kinetic analysis. The Vmax/Km ratio was also calculated for individual groups. All values represent mean ± SD, n=8. The * indicates statistical significance from snca^+/+ and snca^-/- + wt snca.

Wt but not mutant α-synuclein restores acyl-CoA synthetase activity

To further establish a modulatory role for >wt α-synuclein on acyl-CoA synthetase activity in the endoplasmic reticulum, we performed an add back experiment with wt mouse and human α-synuclein and the A30P, E46K, and A53T forms of α-synuclein. Others have shown that α-synuclein is enriched in microsomes (23) and total brain protein is comprised of between 0.1% to 1.0% α-synuclein (79, 80). We confirmed the presence of α-synuclein in our wt brain microsomes using Western blot analysis (Figure 6). Using a dose-response curve, we established that 0.1% to 10% wt mouse and human α-synuclein elicited the same effect on Acsl activity with 20:4n-6 as the substrate (Figure 7a). (Note that this corresponds to a final assay concentration for α-synuclein of 6.3 to 634 μM) However, only wt mouse and human α-synuclein increased Acsl activity in snca^-/- microsomes, while E46K and A53T α-synuclein had no effect on activity (Figure 7b). Surprisingly, A30P α-synuclein depressed enzyme activity. In addition, note that 0.05% α-synuclein had significantly less activity than 0.1% and significantly more activity than 0.01% α-synuclein, indicating that the lower concentration of α-synuclein needed to modulate Acsl activity is quite narrow. Importantly, the mutant forms of α-synuclein did not restore activity, demonstrating an important physiological impact of these mutations on brain 20:4n-6 metabolism. The presence of α-synuclein in microsomes isolated from Snca^+/+ brain and the restoration of acyl-CoA synthetase activity by exogenously supplied α-synuclein, demonstrate a clear mechanistic role for α-synuclein in brain 20:4n-6 metabolism. Whether α-synuclein interacts with the enzyme directly or is involved in substrate presentation remains to be determined.

Figure 6.

Western blot analysis confirmed the presence α-synuclein in brain microsomes from snca^+/+ mice used in 5 . Microsomes were isolated from mouse brains as described in the Experimental Procedures. Lanes 1-8 contain 50 μg of total protein from snca^+/+ brain microsomes. The positive control (+) is 0.3 μg of total protein from α-synuclein expressing HEK-293 cells;

Figure 7.

a). Dose-response curve for microsomal Acsl activity in brain microsomes isolated from snca^-/- mice following incubation for 30 min with 0.01% to 10% of wt (mouse and human), A30P, E46K, and A53T α-synuclein. This corresponds to a final assay concentration of 0.6 to 634 μM for α-synuclein. Values represent activity in pmol × min^-1 × mg protein^-1, n=8 for each group. The * indicates statistical significance from microsomes deficient in α-synuclein.

b).Restoration of Acsl activity was determined by measuring Acsl activity in snca^-/- brain microsomes following the addition of 1.0% of of wt (mouse and human), A30P, E46K, and A53T α-synuclein. This corresponds to a final assay concentration of 63 μM for the α-synuclein. Microsomes were isolated from 8 snca^-/- mouse brains as described in the Experimental Procedures. The substrate was 20:4n-6 (190 μM). All experimental groups were assayed in the same experiment. Values represent mean ± SD, n=8. The * indicates statistical significance from snca^+/+ mice, p<0.05, and the ** indicates significance from snca^-/- mice, p<0.05.

DISCUSSION

There are two major driving forces for cellular fatty acid uptake: binding to proteins (46, 47, 81-83) and metabolic utilization, specifically acyl-CoA formation (84-86). For instance, expression of Acsl6 in PC-12 cells substantially increases fatty acid uptake (84, 86), whereas expression of FABP increases fatty acid uptake into cells (81-83) and the loss of FABP expression reduces tissue fatty acid uptake in vivo (46, 47). We have previously shown in our studies of 16:0 incorporation and turnover in brain phospholipid pools in Snca^-/- mice that α-synuclein does not bind 16:0, yet Snca^-/-mice have decreased 16:0 incorporation and turnover in brain without changing 16:0-CoA mass (29), suggesting that altered 16:0 metabolism was the driving force for this reduction in uptake. Others have shown that α-synuclein binds oleate (18:1n-9) using a classical Lipidex competition assay (23), although we demonstrated no 18:1n-9 binding using titration microcalorimetry (29). Although α-synuclein binds to 20:4n-6, it does so with an affinity considerably lower than for a classical FABP (66, 87), suggesting that the demonstrated effect of α-synuclein on brain fatty acid uptake is more likely linked to acyl-CoA formation rather than fatty acid binding.

Absence of α-synuclein depressed brain 20:4n-6 uptake and altered fatty acid targeting to the brain organic fraction, although targeting to the aqueous fraction was unaltered (Figure 3). These results are consistent with those observed for uptake of 16:0 into brains lacking α-synuclein (29) and with its affect on 20:4n-6 and 16:0 uptake and metabolism in primary astrocytes (11). Brain cytosolic, unesterified polyunsaturated fatty acid mass is reduced when α-synuclein is absent (88), suggesting a connection between α-synuclein expression and polyunsaturated fatty acids. The observed reduction in targeting to the organic fraction is accounted by reduced incorporation coefficients for 20:4n-6 into individual brain phospholipids in Snca^-/-mice. Thus, there is a modest reduction in 20:4n-6 uptake in Snca^-/-mice that is related to reduced incorporation into brain phospholipids pools.

We also observed a marked reduction in 20:4n-6-CoA mass in Snca^-/- mice. This reduction was associated with depressed kinetics of 20:4n-6 incorporation into brain phospholipid pools, as evidenced by the elevation in lambda and prolonged half-life for 20:4n-6 in phospholipid pools. The net rate of 20:4n-6 incorporation from 20:4n-6-CoA into individual phospholipids was decreased 40-65% and brain individual phospholipid half-lives were increased 2.4-fold to 3.7-fold in α-synuclein deficient conditions (Table 1). It is important to note that absolute levels of 20:4n-6 in phospholipids pools was not altered, but rather the kinetics of 20:4n-6 incorporation and the duration that 20:4n-6 remained on the phospholipid was increased. The dilution coefficient, λ, was increased 3.3-fold in the absence of α-synuclein, indicating that there was significantly less fatty acid recycling from the endogenous pools of 20:4n-6 found in brain phospholipids and more fatty acid entered 20:4n-6-CoA pool from plasma (Figure 8). These data are consistent with the interpretation that α-synuclein modulates specific brain acyl-CoA synthetases that target fatty acids to phospholipids.

Figure 8.

A model for the role of α-synuclein in brain 20:4n-6 metabolism is presented. Abbreviations are: 20:4n-6 - arachidonic acid; k* 20:4-CoA - incorporation coefficient of 20:4 from plasma into 20:4-CoA pool; FATP - fatty acid binding protein; [20:4-CoA] - mass of 20:4-CoA in tissue; Lamda - dilution coefficient; and Incorp. Rate 20:4-CoA - net rate of 20:4 incorporation from 20:4-CoA into individual phospholipids.

Several mechanisms could lead to decreased 20:4n-6-CoA mass in the absence of α-synuclein. Because α-synuclein overexpression alters expression of acyl-CoA synthetase in Drosophila (35), we determined if α-synuclein deficiency has an effect on the expression of key enzymes involved in acyl-CoA metabolism (Figure 4a,b). Our data indicate that neither acyl-CoA synthetase expression nor acyl-CoA hydrolase expression was altered, hence this possibility does not account for the reduction in observed 20:4n-6-CoA levels. On the other hand, we demonstrated a 32% reduction in brain microsomal acyl-CoA synthetase activity (Vmax) in preparations from α-synuclein deficient mouse brains (Figure 5 and Table 3). In addition, the Vmax/Km ratio was also significantly depressed in the gene-ablated mice (Table 3), demonstrating that under physiological conditions (75) there would be an effect by α-synuclein on Acsl activity. While we demonstrated the presence of α-synuclen in snca^+/+- microsomes (Figure 6), others have also noted that α-synuclein is particularly enriched in microsomal fractions containing primarily ER (23) as well as colocalized with ER in synaptosomal preparations (21, 80, 89), but not found associated with synaptic vesicles (80).

More importantly, addition of a physiologically relevant amount of recombinant wt mouse or human α-synuclein completely restored acyl-CoA synthetase activity towards 20:4n-6 (Figure 7a,b). These results indicate that α-synuclein can modulate acyl-CoA synthetase activity, while its mutant (A53T and A30P) forms failed to do so. In fact, A30P significantly reduced acyl-CoA synthetase activity when added to α-synuclein deficient microsomes. Although wt mouse α-synuclein has a natural A53T mutation, it has a number of significant differences in the carboxy tail region as compared to human wt α-synuclein or α-synuclein bearing the A53T mutation. We speculate that these differences may account for the ability of the wt mouse α-synuclein but not the human A53T form of α-synuclein to restore Acsl activity in microsomes isolated from snca^-/- mice. These results indicate, for the first time, a significant impact of the mutant forms of α-synuclein on a physiological activity, implying a potentially important reduction in brain 20:4n-6 metabolism in subjects expressing these forms.

To address if α-synuclein may be involved in substrate presentation or product removal, we determined the ability of α-synuclein to bind 20:4n-6 and 20:4n-6-CoA. The lack of 20:4n-6-CoA binding by α-synuclein does not support modulation of acyl-CoA synthetase activity via product removal, while the binding of free 20:4n-6 supports modulation via substrate presentation, but does not rule out a direct protein-to-protein interaction. These results clearly indicate that α-synuclein interacts with the enzyme in a manner that modulates activity, either by a direct protein-protein interaction or through substrate presentation.

Although microsomal Acsl activity was significantly reduced in Snca^-/- brains, total brain acyl-CoA synthetase activity toward 20:4n-6 was not altered. This lack of altered activity for total brain acyl-CoA formation is consistent with the similar incorporation coefficients (k*) from plasma into the 20:4n-6-CoA pools. Taken together, these data indicate that α-synuclein modulates specific acyl-CoA synthetase pools. Acyl-CoA pools are spatially arranged in organelles, with pools ranging from the mitochondrial pool to those in the endoplasmic reticulum (29). The lack of differences in the amount of tracer entering brain aqueous fraction, which represents products of β-oxidation (46, 67-69) (Figure 3), indicates that α-synuclein does not impact the mitochondrial pool relative to the endoplasmic reticulum pool, similar to the effect observed using 16:0 (29). Moreover, fatty acid transport protein (FATP) is associated with plasma membrane and exhibits both fatty acid transport and acyl-CoA synthetase activities (90-92) and FATP-4 is expressed in the brain (93). The potential involvement of this enzyme in acyl-CoA formation may account for the lack of differences between the groups in whole brain homogenate acyl-CoA activity and the similar incorporation coefficients for 20:4n-6 from the plasma into the total brain acyl-CoA pool. This suggests that α-synuclein modulates specific acyl-CoA synthetases localized in the ER, consistent with our data presented above. Taking into account that brain Acsl6 targets fatty acids to phospholipids and neutral lipids (84), but that Ascl1 targets fatty acids only to triglycerides (75), we speculate that α-synuclein selectively affects one or more specific acyl-CoA synthetases (Acsl 3, 4, or 6) associated with phospholipid synthesis. It is important to note that in α-synuclein deficient astrocytes, there is significantly more fatty acid targeted for incorporation into neutral lipids, e.g. triacylglycerols and cholesteryl esters as well as a substantial increase in neutral lipid mass (11). In addition, brain neutral lipid mass was also increased in the Snca^-/- brains (unpublished data), suggesting a similar situation occurs in vivo.

Collectively, our results demonstrate that α-synuclein is critical for maintenance of brain 20:4n-6-CoA levels that are absolutely essential for incorporation of 20:4n-6 into brain phospholipid pools (Figure 6). This effect on metabolism accounts for the small, but significant reduction in 20:4n-6 uptake. There was a large reduction in 20:4n-6-CoA mass in Snca^-/- brains, resulting from a reduction in endoplasmic reticulum (microsomal) acyl-CoA synthetase activity. This reduction in activity decreased the net 20:4n-6-CoA pool size, which caused a substrate limiting reduction in 20:4n-6-CoA resulting in decreased incorporation into individual phospholipids. As a result, the half-life of individual phospholipids was increased, indicating less recycling of 20:4n-6 from endogenous phospholipid stores. These results are important because they are the first demonstration that α-synuclein impacts brain 20:4n-6 metabolism through modulation of an ER based acyl-CoA synthetase rather than through the influence on fatty acid uptake via the direct binding of 20:4n-6 by α-synuclein. Because of the fundamental importance of 20:4n-6 metabolism in brain function, we demonstrate herein a key means of regulating 20:4n-6 metabolism through modulation of brain 20:4n-6-CoA pools; thereby providing a means to channel 20:4n-6 to phospholipids. This newly described physiological function for α-synuclein provides a broader view of how it regulates brain lipid metabolism and how changes in this protein may impact brain function, contributing to the pathophysiology associated with neurodegenerative disease. In addition, we demonstrate the inability of the mutant forms of α-synuclein to restore brain microsomal 20:4n-6-CoA production via microsomal Acsl, suggesting a potentially critical reduction in 20:4n-6 metabolism in subjects expressing these mutant forms of α-synuclein.

ABBREVIATIONS

Snca

alpha-synuclein

Acsl

acyl-CoA synthetase

Bach

acyl-CoA hydrolase mRNA

36B4

acidic ribosomal phosphoprotein P0 (italics gene and non-italics protein)

EtnGpl

ethanolamine glycerophospholipids

PtdIns

phosphatidylinositol

PtdSer

phosphatidylserine

ChoGpl

choline glycerophospholipids

CerPCho

sphingomyelin

20:4n-6

arachidonic acid

k_i*

incorporation coefficient of 20:4n-6 from plasma into individual phospholipid

J_FA,i

net rate of incorporation from 20:4n-6-CoA into individual phospholipid

T_1/2,i

half life of 20:4n-6 in individual phospholipid

k* 20:4-CoA

incorporation coefficient of 20:4 from plasma into 20:4-CoA pool

[20:4-CoA]

mass of 20:4-CoA in tissue

Lambda

dilution coefficient

Incorp. Rate 20:4-CoA

net rate of 20:4 incorporation from 20:4-CoA into individual phospholipids

FATP

fatty acid binding protein