The impact of choline availability on muscle lipid metabolism

Abstract

Consumption of choline-rich foods is essential to ensure membrane integrity, neurotransmission and genomic methylation pathways. Insufficient dietary choline supply can cause choline deficiency (CD) which manifests in the development of non-alcoholic fatty liver disease. There is very limited information regarding the effect of CD on non-hepatic tissues such as muscle. In this study, we induced CD in muscle cells and investigated the effect on choline transport, phosphatidylcholine (PC), fatty acid and triacylglycerol (TAG, fat) metabolism. Choline transport was stable across the plasma membrane of CD cells but significantly impaired in mitochondria. The main choline-transporter SLC44A1 was down-regulated by CD at the mRNA level, and SLC44A1 protein was reduced in total cell lysates and isolated mitochondria. CD significantly reduced PC synthesis but PC degradation was unaffected. PC from CD muscle was modified and contained more monounsaturated fatty acids at the expense of saturated fatty acids. Surprisingly, CD muscle cells also accumulated TAG in the form of large lipid droplets. Those droplets were formed from endogenous fatty acids and by slower TAG metabolism. This study established for the first time that choline availability affects muscle membrane lipid composition and intracellular lipid metabolism, and underlines the significance of choline-rich foods for proper muscle function.

---

Introduction

The essential nutrient choline is prominent in eggs, meats and meat products, liver, fish, soy and wheat germ; beet root, wheat bran, and spinach are rich in choline's oxidation product, betaine.^1,2 The National Academy of Sciences has classified choline as an indispensable dietary component.³ Experimental data has corroborated a vital role for choline in neurotransmission, membrane integrity and methylation pathways.

Phosphatidylcholine (PC) is the major plasma membrane phospholipid of each cell, and therefore the majority of choline in the body is present as the PC headgroup. In the mitochondria of liver and kidney, choline is oxidized to betaine, a methyl group donor in the homocysteine-methionine methylation pathway. Choline is furthermore crucial for neuronal function as component of the neurotransmitter acetylcholine and the membrane lipid sphingomyelin. Foods contain both free and esterified choline compounds and their dietary availability depends on whether they are lipid- or water soluble. Free choline is absorbed in the upper small intestine and enters the portal circulation. PC is partly processed by gut phospholipases, is primarily absorbed as lyso-PC and transported as part of chylomicrons in the lymph.

The adequate intake level for choline is 550 mg day⁻¹ for men, 425 mg day⁻¹ for women, and 450 mg day⁻¹ for pregnant women. Previous studies suggested that the recommended levels were overall met or exceeded by most people,^4,5 however more recent epidemiological data indicates that the individual dietary intake varies and could pose a risk for becoming choline deficient (CD).^6–12 This is especially true for pregnant and lactating women, infants, elderly people, cirrhosis patients and patients depending on parenteral nutrition.¹³

Fatty liver (non-alcoholic steatosis) is the most prominent phenotype of adult CD.¹⁴ While CD has been largely investigated as a liver-centric problem, there is at present a lack of information on the effect of CD on other tissues, such as muscle, which—forming the majority of body tissue mass—is not only rich in PC but also utilizes lipid components as energy source in mitochondria. Even though choline oxidation to betaine is traditionally considered a liver- and kidney-specific process, we have previously demonstrated the presence of the choline transporter SLC44A1 in the mitochondrial membrane of muscle cells.¹⁵ It seems therefore imperative to elucidate the importance of choline metabolism in muscle cells. Previous studies have demonstrated that choline deficiency induces damage to mouse¹⁶ and human^16,17 muscle cells. This damage is attributed to a higher fragility of cell membranes and the induction of apoptosis.¹⁶

Since choline is a positively charged molecule, it cannot freely cross hydrophobic membranes and depends on transport systems to enter the cell¹⁸ and the mitochondria.¹⁹ In the present study we characterize choline transport and utilization during CD in muscle cells, and identify mechanisms for TAG accumulation. We hypothesized that choline transport, membrane composition and TAG metabolic pathways are directly regulated by choline availability.

Results and discussion

Effects of CD on choline transport

When C2C12 muscle cells were grown without choline in the medium, there was no significant change in choline uptake across the plasma membrane (Fig. 1A). There was, however, a ∼40% (p < 0.01, Fig. 2C) reduction in mitochondrial choline uptake. This finding supports our hypothesis that the choline-specific transporter SLC44A1 is particularly important for muscle mitochondrial choline transport, and this is substantiated by a previous tissue panel expression analysis in our laboratory showing that SLC44A1 has the highest expression in skeletal muscle²⁰ and is elevated in muscle myopathies.²¹

Fig. 1 Choline transport and transporter expression A. Choline transport across the plasma membrane was unaffected by CD in cells cultured with or without choline. Cells were incubated with varying concentrations of radiolabelled choline, lysed, and radioactivity determined by liquid scintillation counting. B, C. Expression of the choline transport protein SLC44A1 was down-regulated by CD both at the mRNA (B) and protein level (C) (p < 0.01). SLC44A2 expression did not change significantly; the organic cation transporters OCT1 and OCT2 were not expressed (B.). Total RNA isolated from control or CD cells was reverse transcribed and choline transport expression analyzed using specific primers. Proteins from whole cell lysates were separated by SDS-PAGE and detected with an SLC44A1-specific antibody by Western blotting.

Fig. 2 Mitochondrial choline transporter expression and choline transport A. Immunocytochemistry experiments revealed the disappearance of SLC44A1 from mitochondria in CD cells. Cells were incubated in control or CD media for 72h on coverslips, incubated with MitoTracker Red, fixed, and SLC44A1 detected with specific primary and fluorescent secondary antibodies. Microscopy data shown was previously published in ref. 11. B. Loss of mitochondrial SLC44A1 in CD cells was confirmed by Western blotting experiments with isolated mitochondria. Equal protein loading was ensurred by Ponceau S staining of Western blotting membranes. C. Choline uptake across the mitochondrial membrane assessed with radiolabelled choline was significantly decreased in CD mitochondria (p < 0.01, C).

Choline-specific transporter expression is reduced by CD

C2C12 muscle cells expressed SLC44A1 both at the cell surface and in the mitochondria as determined by RT-PCR, Western blotting and immunostaining (Fig. 1B and C, 2A and B). The expression of total SLC44A1 at the mRNA level was diminished in CD muscle cells by 59% (p < 0.01, Fig. 1B). The SLC44A1 protein amount also decreased in CD muscle cells (Fig. 1C). The putative choline transporter SLC44A2 was also present but its mRNA levels did not change significantly with CD (Fig. 1B). The non-specific choline transporters OCT1 and OCT2 could not be detected and were not further investigated (Fig. 1B). Mitochondrial SLC44A1 protein expression was significantly diminished by CD as well (Fig. 2B and C). We initially anticipated an up-regulation of SLC44A1 during choline deprivation similarly to the observed up-regulation of neuronal choline transporter CHT1²² and other nutrient transporters such as folate²³ and amino acid²⁴ transporters when substrate availability is limited. However, down-regulation of SLC44A1 might reflect a different adaptation of this transport to choline deprivation, such as a prevention of intracellular choline loss from the cells.

Phospholipid PC homeostasis is modified in muscle cells exposed to CD

PC synthesis was followed using 1, 2 and 3h pulse radiolabelling with ³H-glycerol or ¹⁴C-oleate. As shown in Fig. 3A, the rate of PC formation from ³H-glycerol was significantly decreased by the lack of choline (p < 0.001). Labelling muscle cells with ¹⁴C-oleate (Fig. 3B) showed no difference in the rate of PC synthesis. The rate regulatory genes for the de novo Kennedy pathway (CTP:phosphocholine cytidylyltransferase, CT) and for the phosphatidylethanolamine (PE) methylation pathway (phosphatidylethanolamine methyltransferase, PEMT) were analyzed at the mRNA level. In agreement with the glycerol radiolabelling data for PC, the expression of CT mRNA was down-regulated after CD (by ∼38% at 72h, p < 0.01, Fig. 3C). PEMT expression was not affected in CD muscle cells (Fig. 3C).

Fig. 3 Phosphatidylcholine metabolism A. The rate of PC synthesis was significantly decreased when cells were incubated with ³H-glycerol in the respective media, e.g. control media for control cells and CD media for CD cells (p < 0.001 after 3h pulse). B. The rate of PC synthesis was unaltered when cells were incubated with ¹⁴C-oleate. C. Expression of the rate-regulatory enzyme of PC synthesis, CT, was down-regulated after 72h CD (p < 0.01). PEMT mRNA was expressed but unaltered by CD. D. When control and 72h CD cells were labelled with ³H-glycerol in choline-containing media (repletion), the rate of PC synthesis was restored quickly in CD cells (p < 0.05). E. The rate of PC degradation was unaltered in pulse-chase experiments with ³H-glycerol. All radiolabelling experiments were performed by incubating cells with the label diluted in media, and lipids were isolated and analyzed by thin-layer chromatography.

When choline was briefly supplied in the media, e.g. the control media which contained choline was used for pulse labelling of CD cells, the rate of PC synthesis from ³H-glycerol was restored with a significantly increased rate during repletion (p < 0.05) (Fig. 3D). The rate of PC degradation was not changed by CD (Fig. 3E). Specific radiolabelling of the Kennedy pathway with choline was not possible to perform since the intracellular pools of choline are different under CD and choline supplemented conditions and therefore skew the proportion of incorporated label. Taken together, the radiolabelling data indicated that the regulatory mechanisms for the muscle PC synthesis under CD are not at the level of plasma membrane choline transport but at the level of the rate-limiting enzyme of de novo PC synthesis, CT.

The fact that PC synthesis was quickly restored in CD cells when choline was replenished indicates a fast response of the de novo pathway when choline became available. This agrees with numerous studies in the liver cells showing that choline repletion after a period of deprivation can reverse the CD phenotype to normal.^25,26 Even though PEMT was expressed in muscle cells, expression was unaffected by CD. PEMT expression in skeletal muscle has been reported previously in a human tissue panel,²⁷ however PEMT activity is almost exclusively hepatic.²⁸ We analyzed the PE methylation pathway by radiolabelling cells with ¹⁴C-methionine or ¹⁴C-ethanolamine for 1, 2 and 3h and measuring PC synthesis, and found that no radiolabeled PC was synthesized, and hence there was no PEMT activity in C2C12 cells (data not shown).

CD specifically modifies PC fatty acid composition

PC saturated FA side chains decreased in CD muscle cells (from 50.75% to 41.24%, p < 0.05, Fig. 4A), due to a significant decrease in stearic acid (18 : 0). PC oleic acid (18 : 1) was increased by CD (p < 0.05), however the total change in PC unsaturated FAs was not significant. There was a trend towards an increase in omega-6 FAs in PC (from 1.7% to 2.62%; Fig. 4A) but it was not significant. Total phospholipid FA composition was not modified by CD (Fig. 4B). We are unaware of any similar analysis of PC FA composition in muscle.

Fig. 4 Phospholipid fatty acid composition A. Phosphatidylcholine-PC FAs were more unsaturated in CD cells compared to control cells (p < 0.05), which could be attributed to a significant decrease in stearic acid (18 : 0, p < 0.05) and a significant increase in oleate (18 : 1, p < 0.05). There was a trend towards an increase in polyunsaturated FAs which was however not significant. B. There were no significant changes in the FAs of the total phospholipids.

Fat accumulates in CD muscle cells

As determined by staining cells with Oil Red O, TAG (fat) accumulated in CD muscle cells in the form of lipid droplets (Fig. 5A). Extraction and quantification of the total staining confirmed a significant (p < 0.01) increase in TAG accumulation in CD cells. In addition, analysis of TAG content (Fig. 5B) using an enzyme-based kit demonstrated a significant accumulation of TAG in CD cells (p < 0.01). The observed TAG accumulation under CD in muscle cells is to our knowledge a phenomenon as yet unknown. CD has been shown to cause muscle damage due to membrane fragility and induction of apoptosis,¹⁶ but has not been described as a lipid droplet phenotype in muscle. Muscle cells do not generate the lipoprotein particles, and the accumulation of TAG can therefore not be caused by an impaired export of TAG, which was the case with liver CD,^19,20 but must instead reflect a direct impact of CD on TAG metabolic pathways, as we investigated in the following section.

Fig. 5 Lipid droplet accumulation A. CD induced the accumulation of lipid droplets as analyzed by Oil Red O staining. B. TAG content in CD cells was significantly increased (p < 0.01) when measured by both the total TAG content and the total Oil Red O change in absorbance.

Mechanism of TAG accumulation in CD muscle cells

Short-term (1, 2 and 3h) pulse radiolabelling with ³H-glycerol revealed that de novo DAG synthesis from glucose (lipogenesis) was inhibited (p < 0.05) and that TAG synthesis was not significantly affected by CD (Fig. 6A,B). When ³H-glycerol labeling was performed for 24h, to radiolabel the entire glycerolipid pool (DAG, TAG, PC and PE in Fig. 8), the equilibrium TAG showed a trend towards an increase during CD, while DAG, PC and PE become reduced. This was in agreement with an increased total TAG content and the lipid droplet formation as discussed above (Fig. 5), yet the source of elevated TAG in CD was not an increased glucose utilization. To establish if fatty acids could be alternative contributors to the elevated TAG pool in CD cells, additional radiolabelling was performed with ¹⁴C-oleate (Fig. 6C,D); the obtained data clearly demonstrated significantly increased synthesis of both DAG and TAG from oleate (p < 0.05) of CD cells (slopes in Fig. 6C,D). Finally, the pulse-chase experiments established a significant inhibition of TAG degradation (lipolysis) in CD (p < 0.05, Fig. 6E). Therefore, TAG accumulated in CD muscle cells because of both, an increased TAG synthesis from FA and a reduced TAG degradation.

Fig. 6 Diglyceride and triglyceride metabolism A. The rate of DAG synthesis from glycerol was significantly decreased in muscle cells (p < 0.05), while the rate of TAG synthesis (B) was unaltered after short-term (1, 2 and 3h) incubation with ³H-glycerol. C. The rate of DAG synthesis from oleate was unaffected by CD in C2C12 cells, however TAG synthetic rate (D) was significantly increased (p < 0.05). E. Pulse chase studies with ³H-glycerol demonstrated that the rate of TAG degradation was significantly reduced by CD in muscle cells (p < 0.05).

Expression of the key lipogenic gene SREBP1c was significantly down-regulated (54.8% p < 0.01) while the downstream targets were slightly reduced (stearoyl-CoA desaturase 1-SCD1; 91.1%, p < 0.05; Fig. 7) or did not change (fatty acid synthase-FAS) (Fig. 7). Diacylglycerol O-acyltransferases, DGAT1 and DGAT2, were reduced by 57.5% (p < 0.05) or not affected by CD. This data provided further evidence that the observed TAG accumulation in CD muscle was not due to increased lipogenesis (utilization of glucose) but because of increased FA and DAG utilization from membrane phospholipids, as reduced DAG, PE and PC pools (Fig. 8A, C, D) and reduced total cellular content (Fig. 8E, F) in CD muscle strongly support.

Fig. 7 Expression of lipogenic genes The key lipogenic gene SREBP1c showed significant down-regulation. Downstream gene FAS expression was unaltered and SCD1 expression was down-regulated; DGAT1 expression was down-regulated, while DGAT 2 expression was unaltered in CD muscle cells.

Fig. 8 Modification of glycerolipid pools and total phospholipid content The equilibrium radiolabeling of glycerolipids DAG (A), TAG (B), PC (C) and PE (D) with ³H-glycerol was performed for 24h in the respective media, e.g. control media for control cells and choline deficient media for CD cells as in Fig. 6. TAG pool was the only elevated while all other pools, for DAG, PC and PE, were reduced, with the highest reduction observed for PE. Total cell PE (E) and PE (F) content.

Experimental

Cell culture

C2C12 mouse muscle cells were cultured in DMEM medium supplemented with 8% fetal bovine serum (FBS); cells were maintained at 37 °C and 5% CO₂. CD and control media were prepared according to the formulation by HyClone (DMEM). The control media contained 4 mg L⁻¹ choline (=28.6 μM). For CD experiments, cells were cultured for 24h in their normal growth medium. On the next day, the media was replaced with control or CD media for 72h. Media for CD experiments was free of FBS to entirely deplete CD cells of choline. Under such conditions, the C2C12 cells growth was reduced and their mitochondria were intact; at 72h only 10–15% of cells undergoes apoptosis.

Choline uptake

For plasma membrane choline uptake experiments, control or CD cells were incubated with varying concentrations of ³H-choline (0.5, 1, 1.5 or 2 μCi, equals 6, 12, 18 and 24 nM; specific activity 83 Ci/mmol) in control or CD media for 5 min at room temperature. Cells were then washed twice with 1 mM ‘cold’ choline in Krebs-Ringer-Hepes buffer (130 mM NaCl, 1.3 mM KCl, 2.2 mM CaCl₂, 1.2 mM MgSO₄, 1.2 mM KH₂PO₄, 10 mM HEPES, pH 7.4, 10 mM glucose), scraped in lysis buffer (10 mM Tris, 1 mM EDTA, 10 mM NaF) and ³H radioactivity determined with liquid scintillation counting.

Detection of cellular transporters

Total mRNA was extracted from cells using TriZol reagent (Invitrogen) and cDNA was synthesized from 1 μg of RNA using a poly-dT primer and Superscript reverse transcriptase II (Invitrogen). SLC44A1 cDNA was amplified using specific forward and reverse primers (PCR details listed in Table 1). The PCR conditions were: 5 min initial denaturation at 94 °C; 32 cycles of: 94 °C 30 s, 53 °C 30 s, 72 °C 30 s; 5 min final extension at 72 °C. Expression of OCT1, OCT2 and SLC44A2 was analyzed with the same PCR conditions as SLC44A1 with individual adjustments of the annealing temperature. Mouse kidney mRNA was reverse transcribed and used as a positive control for OCT1/2 expression. 18S-RNA served as an internal control; both a poly-dT primer and an 18S-RNA-specific primer were used for cDNA synthesis and 18S-cDNA was amplified using specific primers. PCR conditions for 18S-cDNA amplification were 5 min initial denaturation at 94 °C, 22 cycles of: 94 °C 30 s, 58 °C 30 s, 72 °C 30 s, and a 5 min final extension at 72 °C. SLC44A1 protein expression was determined with non-denaturing PAGE Western blotting and immunocytochemistry using specific antibodies as described in our published work.¹⁹ Briefly, nondenaturing sample buffer (62 mM Tris-HCl, 0.01% bromophenol blue, and 10% glycerol) was added to 15–20 μg of mitochondrial or whole-cell protein and proteins were separated by nondenaturing polyacrylamide gel electrophoresis (10% gel, no SDS) at 120 V for 1.5 h. Proteins were transferred to polyvinylidene difluoride membranes in a semidry transfer system (Bio-Rad, Hercules, CA, USA). Membranes were blocked for 1 h at RT with 5% skim milk in TBS-T. Incubation of membranes with the SLC44A1 antibody was performed overnight at 4 °C (1 : 1000 in 5% skim milk in TBS-T). After several washes with TBS-T, membranes were incubated with a horseradish-peroxidase-conjugated secondary antibody (1 : 10,000) for 1 h at room temperature. Membranes were again washed with TBS-T, and bands visualized with an enhanced chemiluminescence detection kit (Sigma).

Table 1 PCR primers for genes involved in choline transport and lipid metabolism

Gene	Accession No.	Primer sequence	Product size	Annealing temperature
SLC44A1	BC113167	F: ccggtttggctgggattatgc	372 bp	54 °C
		R:ggagagccttgtgcaaacagc
SLC44A2	BC031535	F: ttgctgtgtgttgctcttcc	385 bp	53 °C
		R: ggtgataaccgctggacact
OCT1	NM_009202	F: tgaacttgggcttcttcctg	235 bp	50 °C
		R: agatggctgtcgttctcctg
OCT2	BC015250	F: agaccatcgaggatgctgag	210 bp	52 °C
		R: agctggacacatcagtgcaa
18S RNA	NR_003278	F: taccacatccaaggaaggcagca	180 bp	58 °C
		R: tggaattaccgcggctgctggca
CT	BC018313	F: atgcacagagttcagctaaag	170 bp	50 °C
		R: gggcttactaaagtcaacttcaa
PEMT	BC026796	F: tgtttgtgctgtccagcttc	320 bp	52 °C
		R: ttccaaagatccttcatggc
FAS	BC046513	F: cttcgagatgtgctcccagctgc	279 bp	57 °C
		R: cttagtgataaggtccacggaggc
SCD1	NM_009127	F: cgcatctctatggatatcgcccc	279 bp	54 °C
		R: ctcagctactcttgtgactcccg
SREBP1c	NM_011480	F: tcacaggtccagcaggtccc	197 bp	57 °C
		R: ggtactgtggccaagatggtcc
DGAT1	NM_010046	F: atccagacaacctgacctaccg	257 bp	53 °C
		R: gaccgccagctttaagagacgc
DGAT2	NM_026384	F: ggctggtaacttccggatgcc	233 bp	55 °C
		R: gatcagctccatggcgcaggg

---

For immunocytochemistry, cells grown on glass coverslips in 6-well plates were incubated in medium (± choline) containing 100 nM MitoTrackerRed CMXRos for 30 min at 37 °C and 5% CO₂. After removal of the medium cells were washed with pre-warmed culture medium and fixed with paraformaldehyde (4%w/v in PBS) for 15 min at 37 °C. After rinsing of cells with PBS, the plasma membrane was permeabilized by incubation of cells in 0.2% TritonX-100 in PBS for 5 min at RT. Cells were blocked (5% goat serum in PBS, 1 h at RT) and incubated with the primary antibody (SLC44A1 antibody, 1 : 100 in blocking solution) for 1 h at room temperature. Cells were washed again with PBS, followed by incubation with the secondary antibody (AlexaFluor488, Invitrogen, 1 : 500 in blocking solution) for 30 min at RT. Cells were rinsed again with PBS, mounted onto microscope slides in Permafluor mounting medium, and analyzed using a confocal laser scanning microscope with a Leica TCS SP2 system (Leica Microsystems, Wetzlar, Germany).

Measurements of mitochondria-specific choline transport and transporters

Mitochondria were isolated from control and CD cells using standard protocols for differential centrifugation.²⁹ Briefly, cell pellets were resuspended in RSB buffer (10 mM NaCl, 1.5 mM MgCl₂, 10 mM Tris-HCl) and allowed to swell for 10 min on ice. Cells were homogenized using 15 strokes in a glass/Teflon Potter-Elvehjem grinder, Mannitol-Sucrose isolation buffer (210 mM mannitol, 70mM sucrose, 5 mM Tris-HCl, 1 mM EDTA) was added and the homogenate centrifuged at 2,500 rpm for 5 min. The supernatant was centrifuged at 12,500 rpm to pellet the crude mitochondrial fraction. Intactness of mitochondria was determined with an isolated mitochondria staining kit (Sigma-Aldrich, product # CS0760). Mitochondrial SLC44A1 expression was analyzed by non-denaturing PAGE and Western blotting, and controlled for with a positive mitochondrial loading control (COXIV antibody, Abcam). For choline uptake assays, mitochondria were incubated with 1 μCi ³H-choline in mitochondrial uptake buffer (120 μM KCl, 5 μM HEPES/KOH, 1 μM EGTA, 5 μM KH₂PO₄, 0.5 μM MgCl₂, 5 μM L-glutamate, and 1.2 μM L-malate; pH 7.2) for 5 min at RT, pelleted by centrifugation at 12,500 rpm for 5 min, washed with mitochondrial uptake buffer containing 1 mM ‘cold’ choline and radioactivity determined with liquid scintillation counting.

Pulse and pulse-chase radiolabelling

For pulse radiolabelling experiments, 72 h CD cells grown in 6-well plates were incubated for 1, 2, 3 or 24 h with 5 μCi ³H-glycerol (specific activity: 20 Ci/mmol) or for 1, 2 or 3 h with ¹⁴C-Oleate (specific activity: 27 Ci/mmol) in 2 mL of the respective medium. For oleate pulse labelling, ¹⁴C-oleate was first incubated with BSA-complexed ‘cold’ Na-oleate to facilitate uptake. Cells were washed twice with PBS and total lipids extracted according to the Bligh-Dyer method.³⁰ A fraction of the cells was kept for determination of protein concentration using a BCA assay kit (Pierce). PC and PE were separated from the chloroform phase by thin layer chromatography (TLC) with a solvent system of chloroform–methanol–ammonia (65 : 35 : 5). Diacylglycerols (DAG) and TAG in the chloroform phase were separated by TLC with a solvent system of heptane–isopropyl ether–acetic acid (60 : 40 : 5). Standards and radiolabeled lipids were visualized with dichlorofluorescein under UV light, specific bands scraped and radioactivity determined by liquid scintillation counting. For pulse-chase experiments, cell were incubated with 5 μCi ³H-glycerol in 2 mL of the respective media for 2 h, then the medium was removed and replaced with medium containing an excess of ‘cold’ glycerol (250 μM). Lipids were extracted after a chase period of 1, 2 or 4 h and separated as described above.

Total TAG. PE and PC content and glycerolipid pool determination

C2C12 mouse muscles cells were cultured to confluence in 96-well plates in DMEM supplemented with 10%(v/v) fetal bovine serum (FBS). After 24 h, the media was replaced with control or CD media (both without FBS) and cells were cultured for 48 h before experiments.

For experiments, cells were washed twice with PBS and 10 μl of dimethyl sulfoxide (DMSO) were added to each well for 1 min to extract TAG from the cells according to.³¹ TAG concentration was determined using the enzyme based assay kit from Ambion. The absorbance at 570 nm was measured using a microplate reader and total TG content was calculated based on a standard curve. Total PC and PE content were determined by thin-layer chromatography and densitometry, using the fluorescent probe 1,6-diphenylhexatriene and standard curves generated for each lipid.³² To measure various glycerolipid pools, the cells grown as above were radiolabeled with ³H-glycerol for 24h to reach the equilibrium, and the radiolabeled PC, PE, DAG and TAG determined by TLC as in pulse and pulse-chase experiments.

Phospholipid fatty acid composition

Lipids were extracted from control and CD cells using the standard Bligh-Dyer method. Lipids were transmethylated to determine their FA composition. Briefly an aliquot of lipid extract was dried under a stream of nitrogen and methylated with 6% H₂SO₄ in methanol at 80 °C for 3 h. The methylated FA were collected by the addition of petroleum ether and water followed by centrifugation at 1,000 × g for 15 min at 4 °C. Samples were stored at −20 °C until analysis by gas-liquid chromatography (HP 5890 Series II; Hewlett Packard).

Oil Red-O staining

Control and 72h CD cells were fixed with 3.7% formaldehyde in PBS for 5 min at RT. Cells were incubated in Oil Red-O solution for 15 min at RT, washed with 60% isopropanol, and nuclei were counterstained for 5 min with hematoxylin. Cells were washed with water, coverslips mounted onto microscope slides in Permafluor mounting solution, and analyzed by microscopy. For Oil Red-O stain extraction experiments, the hematoxylin step was omitted, Oil Red O extracted by incubation of cells in 100% isopropanol for 5 min, and absorbance measured at 510 nm.

Lipid gene analysis

Sterol regulatory element binding protein (SREBP1c), fatty acid synthase (FAS), steaoryl-CoA desaturase 1 (SCD1), diacylglycerol O-acyltransferase 1 and 2 (DGAT1, DGAT2), CTP:phosphocholine cytidylyltransferase-alpha (CT-alpha) and phosphatidylethanolamine methyltransferase 1 (PEMT-1) were amplified with specific primers and the same cDNA synthesis and PCR protocol used for SLC44A1 as described above, with individual adjustments of the annealing temperatures (Table 1).

Statistical analysis

CD samples were compared to control samples with an unpaired two-sided t-test at a 95% confidence interval, while rates of lipid synthesis and degradation was determined by linear regression; all statistical tests were performed with GraphPad prism 4 software.

Conclusions

Dissecting the mechanisms for TAG accumulation in muscle cells revealed that TAG synthesis increases during CD, and that this increase can be attributed to an elevated incorporation of preexisting FAs into DAG rather than de novo FA and DAG synthesis. Total DAG and phospholipid pools as well as DAG synthesis from glucose were reduced in CD muscle cells. Therefore FA originates from membrane phospholipids since they were not supplied from the media. PC synthesis from choline was blunted in CD cells, which directly led to an increased availability of DAG and FA for TAG synthesis. The study provides further evidence for a tightly linked regulation of nutrient choline availability with muscle glucose, phospholipid, fatty acid and fat metabolism and underlines the importance of incorporation choline-rich foods in the diet for the prevention of muscle lipotoxicity and related disorder.

Abbreviations

CD	choline deficiency
CT	CTP:phosphocholine cytidylyltransferase;
DAG	diacylglycerol
DGAT 1/2	diacylglycerol acyltransferase 1/2
FA	fatty acid
FAS	fatty acid synthase
OCT	organic cation transporter
PC	phosphatidylcholine
PEMT	phosphatidylethanolamine methyltransferase
SCD	stearoyl-CoA desaturase
SLC44A1	solute carrier 44A1
SREBP	sterol regulatory element binding protein
TAG	triacylglycerol
VLDL	very low density lipoprotein

Acknowledgements

This study was supported by an Ontario Graduate Scholarship in Science and Technology (to V. Michel) and by an operating grant from the Natural Sciences and Engineering Research Council of Canada and the Ontario Ministry of Agriculture, Food and Rural Affairs (to M. Bakovic).

Bibliographic references

1. K. Y. Patterson, S. A. Bhagwat, J. R. Williams, J. C. Howe, J. M. HoldenNutrient Data Laboratory, Agricultural Research Service, U.S. Department of Agriculture 2008.
2. S. H. Zeisel, M. H. Mar, J. C. Howe and J. M. Holden, J Nutr, 2003, 133(5), 1302–7.
3. J. K. Blusztajn, Science, 1998, 281(5378), 794–5.
4. L. M. Fischer, J. A. Scearce, M. H. Mar, J. R. Patel, R. T. Blanchard, B. A. Macintosh, M. G. Busby and S. H. Zeisel, J Nutr, 2005, 135(4), 826–9.
5. I.o. M. a. N. A. o. S. o. t. USA, Dietary Reference Intakes for Folate, Thiamin, Riboflavin, Niacin, Vitamin B12, Panthothenic Acid, Biotin, and Choline.National Academic PressWashington, DC, 1998; Vol. 1, p 390–422.
6. A. Bidulescu, L. E. Chambless, A. M. Siega-Riz, S. H. Zeisel and G. Heiss., BMC Cardiovascular Disorders, 2007, 7, 20.
7. E. Cho, W. C. Willett, G. A. Colditz, C. S. Fuchs, K. Wu, A. T. Chan, S. H. Zeisel and E. L. Giovannucci, J. Natl. Cancer Inst., 2007, 99(16), 1224–31.
8. E. Cho, S. H. Zeisel, P. Jacques, J. Selhub, L. Dougherty, G. A. Colditz and W. C. Willett, Am J Clin Nutr, 2006, 83(4), 905–11.
9. G. W. Dalmeijer, M. R. Olthof, P. Verhoef, M. L. Bots and Y. T. van der Schouw, Eur. J. Clin. Nutr., 2008, 62(3), 386–94.
10. P. Detopoulou, D. B. Panagiotakos, S. Antonopoulou, C. Pitsavos and C. Stefanadis, Am J Clin Nutr, 2008, 87(2), 424–30.
11. H. H. Jensen, S. P. Batres-Marquez, A. Carriquiry and K. L. Schalinske, Faseb J, 2007, 21, lb219.
12. G. M. Shaw, S. L. Carmichael, W. Yang, S. Selvin and D. M. Schaffer, Am. J. Epidemiol., 2004, 160(2), 102–9.
13. S. H. Zeisel, Am J Clin Nutr, 2005, 82(4), 719–20.
14. B. Lombardi, G. Ugazio and A. N. Raick, Am J Physiol, 1966, 210(1), 31–6.
15. V. Michel and M. Bakovic., Faseb J, 2009.
16. K. A. da Costa, M. Badea, L. M. Fischer and S. H. Zeisel., Am J Clin Nutr, 2004, 80(1), 163–70.
17. L. M. Fischer, K. A. daCosta, L. Kwock, P. W. Stewart, T. S. Lu, S. P. Stabler, R. H. Allen and S. H. Zeisel, Am J Clin Nutr, 2007, 85(5), 1275–85.
18. V. Michel, Z. Yuan, S. Ramsubir and M. Bakovic, Exp Biol Med (Maywood), 2006, 231(5), 490–504.
19. V. Michel and M. Bakovic, FASEB J., 2009, 23(8), 2749–58.
20. Z. Yuan, L. Wagner, A. Poloumienko and M. Bakovic, Gene, 2004, 341, 305–12.
21. Z. Yuan, A. Tie, M. Tarnopolsky and M. Bakovic, Physiol. Genomics, 2006, 26(1), 76–90.
22. T. J. Mellott, N. W. Kowall, I. Lopez-Coviella and J. K. Blusztajn, Brain Res., 2007, 1151, 1–11.
23. I. Ifergan and Y. G. Assaraf, Vitam. Horm., 2008, 79, 99–143.
24. H. Kashiwagi, K. Yamazaki, Y. Takekuma, V. Ganapathy and M. Sugawara, Amino Acids, 2009, 36(2), 219–30.
25. A. L. Buchman, M. E. Ament, M. Sohel, M. Dubin, D. J. Jenden, M. Roch, H. Pownall, W. Farley, M. Awal and C. Ahn, JPEN, J. Parenter. Enteral Nutr., 2001, 25(5), 260–8.
26. K. A. Waite, N. R. Cabilio and D. E. Vance, J Nutr, 2002, 132(1), 68–71.
27. D. J. Shields, L. B. Agellon and D. E. Vance, Biochim. Biophys. Acta, Mol. Cell Biol. Lipids, 2001, 1532(1–2), 105–14.
28. Z. Cui and D. E. Vance, J. Biol. Chem., 1996, 271(5), 2839–43.
29. D. L. Spector; R. D. Goldman; L. A. Leinwand, Cold Spring Harbor Laboratory Press: Plainview, NY, 1998.
30. E. G. Bligh and W. J. Dyer, Can J Biochem Physiol, 1959, 37(8), 911–7.
31. C. Zou and Z. Shen, J. Pharmacol. Toxicol. Methods, 2007, 56(1), 63–6.
32. R. A. Igal, M. C. Pallanza de Stringa and I. N. Tacconi de Gómez Dumm., Int. J. Biochromatogr, 1994, 1, 137–142.

---