点击显示 收起
1 From the Department of Nutrition, School of Public Health and School of Medicine, University of North Carolina at Chapel Hill
2 Some of these data were presented in poster form at the Experimental Biology meetings in New Orleans in 2002 (FASEB J 2002:16:A1023) and in San Diego in 2003 (FASEB J 2003;17:A344). 3 Supported by grant DK55865 from the National Institutes of Health and by grants DK56350 and ES10126 from the NIH to the UNC Clinical Nutrition Research Unit and the Center for Environmental Health Susceptibility, respectively. 4 Address reprint requests and correspondence to SH Zeisel, Department of Nutrition, School of Public Health and School of Medicine, University of North Carolina at Chapel Hill, CB# 7461, Chapel Hill, NC 27599. E-mail: steven_zeisel{at}unc.edu.
ABSTRACT
Background: Choline is a required nutrient, and humans deprived of choline develop liver damage.
Objective: This study examined the effect of choline deficiency on muscle cells and the release of creatine phosphokinase (CPK) as a sequela of that deficiency.
Design: Four men were fed diets containing adequate and deficient amounts of choline, and serum was collected at intervals for measurement of CPK. C2C12 mouse myoblasts were cultured in a defined medium containing 0 or 70 µmol choline/L for up to 96 h, and CPK was measured in the media; choline and metabolites were measured in cells. Apoptosis was assessed by using terminal deoxynucleotidyl transferase–mediated dUTP-biotin end labeling and activated caspase-3 immunohistochemistry. Cell fragility in response to hypo-osmotic stress was also assessed.
Results: Three of 4 humans fed a choline-deficient diet had significantly elevated serum CPK activity derived from skeletal muscle (up to 66-fold; P < 0.01) that resolved when choline was restored to their diets. Cells grown in choline-deficient medium for 72 h leaked 3.5-fold more CPK than did cells grown in medium with 70 µmol choline/L (control medium; P < 0.01). Apoptosis was induced in cells grown in choline-deficient medium. Phosphatidylcholine concentrations were diminished in choline-deficient cells (to 43% of concentrations in control cells at 72 h; P < 0.01), as were concentrations of intracellular choline, phosphocholine, and glycerophosphocholine. Cells grown in choline-deficient medium had greater membrane osmotic fragility than did cells grown in control medium.
Conclusions: Choline deficiency results in diminished concentrations of membrane phosphatidylcholine in myocytes, which makes them more fragile and results in increased leakage of CPK from cells. Serum CPK may be a useful clinical marker for choline deficiency in humans.
Key Words: Creatine phosphokinase • choline deficiency • muscle • myoblasts • apoptosis • phosphatidylcholine
INTRODUCTION
Choline, or its metabolites, is needed for the structural integrity and signaling functions of cell membranes; it is the major source of methyl groups in the diet (one of choline's metabolites, betaine, participates in the methylation of homocysteine to form methionine), and it directly affects cholinergic neurotransmission, transmembrane signaling, and lipid transport and metabolism (1). Choline is a required nutrient, and the Institute of Medicine of the National Academy of Sciences set an adequate intake (AI) standard of 550 mg choline/d for men and 425 mg choline/d for women (2).
Many species of animals fed a diet deficient in choline and methyl groups have depleted choline stores and develop liver dysfunction (3–7). Animals fed this deficient diet may also develop growth retardation, renal dysfunction, and hemorrhage or bone abnormalities (6, 8, 9). Healthy male humans with normal methionine, folate, and vitamin B-12 status fed a choline-deficient diet have diminished plasma choline and phosphatidylcholine concentrations and incur liver damage (ie, elevated plasma alanine aminotransferase) (10). Liver cell death occurs in persons with choline deficiency because hepatocytes initiate programmed cell death (apoptosis) when deprived of choline (11, 12). Hepatic steatosis and elevated transaminase activities are also seen in humans fed total parenteral nutrition solutions that are deficient in choline; these abnormalities resolve when patients are supplemented with a choline source (13–15). Although vigorous exercise can lower plasma choline concentrations (16, 17), it has not previously been appreciated that choline deficiency is associated with effects on muscle in the human. We now report that in some humans deprived of choline, creatine phosphokinase (CPK) leaks from muscle, and that this process is reversed with the repletion of choline. We conducted studies in myocytes (C2C12 mouse muscle myoblasts) in culture to ascertain the mechanisms whereby choline nutriture might influence CPK leakage from muscle.
SUBJECTS AND METHODS
Study in humans
Four healthy male volunteers aged 26, 38, 44, and 48 y (3 whites and 1 African American) and with body mass indexes (in kg/m2) of 23, 22, 22, and 25, respectively, were recruited for a protocol approved by the Institutional Review Board at the University of North Carolina at Chapel Hill. Subjects were admitted to the General Clinical Research Center and given various research diets. A description of these diets is presented in Table 1. Specifically, for 10 d, subjects were fed a baseline diet of normal foods containing 550 mg choline · 70 kg body weight–1 · d–1 [approximately the current presumed AI (2); dietary choline content assayed by our laboratory (18)]. Subjects were then switched to a choline-depletion diet containing <50 mg choline · 70 kg body weight–1 · d–1 until they developed abnormal serum CPK (6–31 d; one subject did not develop abnormal CPK and stayed on the deficient diet for 42 d). The 3 subjects with elevated serum CPK were then fed a choline-repletion diet for 10 d containing 138 mg choline · 70 kg body weight–1 · d–1 (25% of the AI; 2), which was sufficient to return serum CPK values to normal. They were then given the 550-mg choline diet made up of normal foods for 3 d before being discharged. In one subject (26-y-old white man), the diet protocol was changed so that he consumed the choline-deficient diet for 42 d, and, because he did not have elevated CPK, we skipped the repletion diet phase and immediately fed him the 550-mg choline diet made up of normal foods for 3 d before discharge. The diets, which were composed of 0.8 g/kg high-biologic-value protein, with 30% kcal from fat and the remaining kcal from carbohydrates, met or exceeded the estimated average requirement for methionine plus cysteine and the daily reference intake for vitamin B-12. The 550-mg choline diet contained 400 dietary folate equivalents (DFE)/d, and, in the depletion and repletion diets, there was 100 DFE/d, an amount at the lower end of that found in unfortified American diets (2).
View this table:
TABLE 1. Nutrient content of the research diets1
Cooked nonfolate–fortified liquid egg whites and choline-free soy protein beverages were used to provide adequate amounts of high-quality protein during the choline-depletion and -repletion phases. Cola or a sugar-sweetened soft drink mix (Kool-Aid; Kraft Foods, White Plains, NY), wheat starch bread, and canned fruit were used to add very-low-choline carbohydrate calories. Fiber was added to the diet in the form of psyllium (Metamucil; Procter and Gamble, Cincinnati) mixed into the soy protein beverage. The soy protein concentrate used in the beverage as the protein source was a spray-dried soy protein concentrate obtained from defatted soybean flakes that were sequentially extracted with the use first of aqueous ethanol and then of HCl, pH 4.5 (The Solae Company, St Louis). Cold-pressed safflower oil (Loriva Brand; nSpired Natural Foods, San Leandro, CA) was used as the fat source and sucrose as the carbohydrate source, and the beverage was flavored with either chocolate or strawberry syrup (Hershey's Syrup; Hershey Foods Corporation, Hershey, PA). Bread containing added lecithin (The Solae Company) at a concentration of 2 mg/g was used to adjust the diets from <50 mg total choline to 550 mg choline/70 kg body weight in the repletion diets. A multivitamin-multimineral supplement (Centrum multivitamin multimineral liquid; Wyeth Consumer Healthcare, Madison, NJ) provided vitamins A, B-6, B-12, C, D, thiamine, riboflavin, and niacin at the recommended dietary allowance, as well as the minerals iron, zinc, manganese, and chromium. Additional mineral supplements to attain recommended dietary allowances included iron, zinc, copper, selenium, calcium, magnesium, and chromium (Caltrate 600 Plus; Wyeth Consumer Healthcare; GNC Ironchel 18 and GNC Selenium 50; General Nutrition Corporation, Pittsburgh). The multivitamin supplement was provided daily throughout the study, but the additional mineral supplements were provided only during the depletion and repletion phases of the study. Diets were well tolerated by human volunteers, and are described in detail elsewhere (19).
Blood was drawn by venipuncture into a serum separator tube and analyzed for serum CPK activity on day 10 of the 550-mg choline diet and then every 3–4 d until the subject was discharged. For plasma folate, choline, and phosphatidylcholine measurements, blood was collected on day 10 of the 550-mg diet, day 42 of the choline-depletion diet (or when abnormal CPK occurred), and at the end of the repletion diet (when CPK returned to normal) into a heparinized tube, which was immediately placed on ice and spun to separate plasma for analysis. Samples for CPK analysis were subjected to centrifugation at 1000 x g for 10 min at 24 °C, and serum was separated and analyzed by using a dry-slide colorimetric method for CPK activity at the McClendon Clinical Laboratories at University of North Carolina Hospitals, which is accredited both under the Clinical Laboratory Improvement Act and by the College of American Pathologists. CPK isoenzyme analyses were conducted by Mayo Medical Laboratories (Rochester, MN) with the use of an agarose electrophoresis assay (20). Plasma folate was measured with the use of a microbiological assay (21). Choline and phosphatidylcholine were measured in plasma with the use of liquid chromatography/electrospray ionization–isotope dilution mass spectrometry after the addition of internal standards labeled with stable isotopes (22).
Cell culture
Mouse muscle myoblasts (C2C12; American Type Culture Collection, Manassas, VA) were initially maintained in DMEM (Gibco Invitrogen, Grand Island, NY) containing 1 g glucose/L, 584 mg L-glutamine/L, 110 mg sodium pyruvate/L, and 4 mg vitamin B-6 hydrochloride/L and supplemented with 10% fetal bovine serum (FBS; Sigma, St Louis) and 1% antimycotic solution (Gibco Invitrogen) at 37 °C in a humidified atmosphere of 5% CO2 in air.
C2C12 cells were gradually acclimated to serum-free medium so that they could be grown in a defined medium consisting of DMEM/F12 (Atlanta Biologicals, Norcross, GA), 100 mg transferrin/L (Intergen, Purchase, NY), 17.24 mg methionine/L (Sigma), 16.6 mg putresceine/L (Sigma), 10 mg insulin/L (Sigma), 0.4 mg L-thyroxine/L (Sigma), 0.378 mg triiodothyronine/L (Sigma), 0.62 mg progesterone/L (Sigma), 0.039 mg sodium selenite/L (Sigma), 11.5 mmol bovine serum albumin/L (Sigma), 20 µmol sodium oleate/L (Sigma), 7.2 µmol sodium linoleate/L (Sigma), 2.2 g sodium bicarbonate/L, and 70 µmol choline/L. Briefly, cells were started in medium containing 10% FBS, and then we reduced the amount of serum incrementally at successive passages of cells (approximately every 4 d) using medium with DMEM containing 5%, 2.5%, 1%, and 0.5% FBS, respectively (cells were seeded in 10% FBS medium for 3 h to allow for attachment). After 2 passages at 0.5% FBS, we replaced the growth medium with the defined medium containing 70 µmol choline/L, as described above. Cell morphology was observed daily with the use of light microscopy.
Expression of muscle-specific proteins -actinin (crosslinks with actin filaments in striated muscle), MyoD (expressed in proliferating myogenic cells; 23), and myosin heavy chain (MHC; the major contractile protein) was assessed with the use of immunohistochemistry and Western blotting. We used goat polyclonal anti--actinin, rabbit polyclonal anti-MyoD, and goat polyclonal anti-MHC (Santa Cruz Biotechnology, Santa Cruz, CA). The anti-MyoD did not cross-react with myogenin or other muscle-specific transcription factors, whereas the anti-MHC reacted with skeletal and cardiac MHC of mouse, rat, and human origin. Cells were stained with a biotinylated secondary antibody, avidin-peroxidase, and the peroxidase substrate diaminobenzidine or fluorescence-labeled secondary antibodies (ABC Staining System; Santa Cruz Biotechnology). The primary antibodies were used in a concentration of 0.6 µg blocking serum/mL, and the specific fluorescence-labeled secondary antibodies were diluted 1:100. We counterstained the nuclei with 0.5% methyl green and 4,6-diamidino-2-phenylidole (1 µg/mL in phosphate-buffered saline). For Western blotting, protein was measured by the method of Lowry before gel electrophoresis was performed (24). Briefly, equal amounts of cytoplasmic extract (15 µg protein) were loaded and separated on 4–20% Tris-HCl Ready Gels (Bio-Rad, Hercules, CA) before being transferred to a Hybond ECL nitrocellulose membrane (Amersham Pharmacia Biotechnology, Little Chalfont, United Kingdom). We used the same primary antibodies (0.4 µg/mL) as listed above and probed the blots with immunoglobulin G secondary antibodies conjugated with horseradish peroxidase (Santa Cruz Biotechnology) diluted 1:1000 in blocking buffer (0.1% Tween, 1.5% bovine serum albumin in phosphate-buffered saline). Proteins were visualized by using chemiluminescence enhanced with luminol reagent (Santa Cruz Biotechnology) and exposed to Kodak Biomax Light film (Eastman Kodak, Rochester, NY). In the defined 70 µmol choline/L medium without serum, cells maintained normal morphology, and they retained expression of the muscle-specific proteins -actinin, MyoD, and MHC much as did cells grown in DMEM + 10% FBS (data not shown).
Experimental design
C2C12 cells were seeded at 5 x 105 per 100-mm dish in medium containing 10% FBS for 3 h. The medium was then replaced with the defined medium containing 70 µmol choline/L for 48 h. At the end of this 48-h time period (time 0), cells were switched to the 70 µmol choline/L (control) or 0 µmol choline/L (choline-deficient) medium for 24, 48, 72, and 96 h. Media were changed every 48 h.
DNA fragmentation
We measured DNA strand breaks by using terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL method) with the use of an in situ apoptosis detection kit (ApopTag; Intergen, Purchase, NY; 25) and by following the manufacturer's directions. Briefly, the cells were fixed in 1% paraformaldehyde and suspended on a silanized microscope slide by using cytospin. After being stained with diaminobenzidine-peroxidase substrate, the nuclei were counterstained with 0.5% methyl green. The slides were mounted with Permount (Fair Lawn, NJ), and the percentage of TUNEL-positive cells was counted by light microscopy. At least 300 total cells were counted for each data point.
Activated caspase 3
Cells were labeled with FAM-DEVD-FMK reagent (CaspaTag kit; Intergen), which specifically binds to activated caspase-3. Nuclei were then stained with diamidino-2-phenylidole, and activated caspase-3 was detected with the use of a band pass filter [excitation wavelength (EX) 490 nm, emission wavelength (EM) 520 nm] on a microscope. An ultraviolet filter (EX 365 nm, EM 460 nm) was used to visualize nuclear staining. We used the National Institutes of Health–based SCION IMAGE software (version 1.61; Scion Corp, Frederick, MD) to quantify the caspase-3 integrated optical density; 6–16 images were used for each data point.
Myocyte steatosis
Cells were fixed in 3.7% formaldehyde for 2 min and then stained with 0.2% Oil Red O (Sigma) in 60% isopropanol for 1 h (26, 27). The nuclei were counterstained with hematoxylin, and the lipid droplets were visualized by using light microscopy.
Choline and metabolites
Choline and its metabolites were extracted from cells by the method of Bligh and Dyer (28). Aqueous and organic compounds were separated, analyzed, and quantified directly by liquid chromatography/electrospray ionization–isotope dilution mass spectrometry after the addition of internal standards labeled with stable isotopes (22). Internal labeled standards were used to correct for recovery. Data were expressed per DNA concentration, which was measured by using a fluorimetric method (29).
Enzyme assays in culture media
Medium (100 µL) was assayed for lactate dehydrogenase (LDH) by using a colorimetric endpoint assay (Sigma; 30). Medium (200 µL) was analyzed for CPK activity by using a colorimetric assay kit (Sigma; 31). Activities are reported per total number of cells in the dish.
Osmotic fragility
Cells were grown in control or choline-deficient medium for 72 h. The media were then removed, and 2 mL of 0.9% NaCl, 0.6% NaCl, 0.3% NaCl, or water was added. After 30-min incubation at room temperature, the media were collected and assayed for LDH activity. Duplicate plates were used to count the number of attached cells before the osmotic challenge. The cell experiments were done in duplicate and repeated at least once.
Statistics
Statistical differences were determined with the use of a paired t test for the human study. For the cell experiments, data were analyzed by two-way analysis of variance with interaction to determine whether treatments differed significantly over time or by percentage of salt. Parametric (unpaired two-sample t tests) and nonparametric (Wilcoxon's two-sample tests) tests were also used to measure statistical differences at each time point (SAS/STAT, version 8; SAS Institute Inc, Cary, NC).
RESULTS
In 3 of the 4 men studied, serum CPK activities increased significantly when dietary choline intake was restricted (Figure 1). Activities increased 66-, 36- and 43-fold, respectively, in these subjects, compared with values on their baseline diet and did not return to normal until 10 d on diets containing choline. In 2 subjects, activities began to decrease on the morning of the first day of the repletion period. The CPK isoform detected in serum was the MM isoform, derived from skeletal muscle and not from brain or heart. In one subject, we observed no increase in CPK activity even after 42 d on the choline-deficient diet (data not shown).
FIGURE 1.. Serum creatine phosphokinase (CPK) activities in men fed choline-deficient diets. Four healthy men were fed a diet of typical foods containing 550 mg choline/70 kg body weight for 10 d and then fed a choline-depletion diet containing 50 mg choline/70 kg body weight. Once CPK activity rose to >5 times the upper limit of normal, subjects were offered a repletion diet containing 138 mg choline/70 kg body weight until CPK activity returned to normal. Blood was drawn at the end of the 550-mg choline diet (day 10) and every 1–4 d thereafter, and serum CPK was measured. Values are depicted for the 3 subjects in whom we observed increases in CPK activity. The dietary daily intake of choline is indicated for each diet period.
In the 3 subjects who had elevated CPK activities, plasma choline concentrations fell from 10.6 ± 0.3 nmol/mL at baseline to 7 ± 0.5 nmol/mL at depletion (P = 0.028, paired t test) before rising to 9.5 ± 0.5 nmol/mL after repletion (different from depletion, P = 0.004). In the subject with no CPK changes, plasma choline concentration fell from 12.6 nmol/mL at baseline to 8.0 nmol/mL after 42 d on depletion diet. In the 3 subjects who had elevated CPK activities, plasma folate concentrations were 28 ± 3 nmol/L at baseline, 22 ± 5 nmol/L at depletion (P = 0.173 by paired t test), and 18 ± 3 nmol/L after repletion (not different from depletion by paired t test, P = 0.218). In the subject with no CPK changes, plasma folate concentration dropped from 36 nmol/L at baseline to 21 nmol/L at the end of the depletion diet. In the 3 subjects who had elevated CPK activities, phosphatidylcholine concentrations in plasma dropped from 2192 ± 179 nmol/mL at baseline to 1641 ± 139 nmol/mL at depletion, but this drop was not significant. After repletion, plasma phosphatidylcholine concentrations increased to 1932 ± 152 nmol/mL. In the subject with no CPK changes, phosphatidylcholine concentration in plasma dropped from 1433 nmol/mL at baseline to 1314 nmol/mL at end of the depletion diet.
After 72 h in choline-deficient medium, myocytes had no detectable concentrations of choline or phosphocholine (P < 0.01 by Student's t test; Table 2), 57% less phosphatidylcholine (P < 0.01), and 47% less glycerophosphocholine (P < 0.01) than did cells grown in 70 µmol choline/L. There were no changes in sphingomyelin concentrations. Myocytes did not contain betaine, another metabolite of choline.
View this table:
TABLE 2. Concentrations of choline and its metabolites in choline-deficient mouse myocytes and in control cells1
Using Oil Red O, we detected intracellular accumulation of lipid droplets in the muscle cells grown in choline-deficient medium that was not observed in the control cells at 72 h (Figure 2). In addition, we observed numerous pyknotic nuclei in the choline-deficient myocytes at 72 h.
FIGURE 2.. Choline-deficient muscle cells accumulated lipid droplets. C2C12 myocytes were grown in the defined medium containing 70 µmol choline/L for 48 h (time 0) and then switched to 70 µmol choline/L (A) or 0 µmol choline/L (B) for 48 h. After the medium was removed, cells were fixed in 3.7% formaldehyde and stained with Oil Red O. Light microscopy was used to visualize the (red) lipid droplets.
There was no significant change in the CPK activity in media from cells grown in 70 µmol choline/L for 96 h (Figure 3). On the other hand, after 48, 72, and 96 h of treatment, CPK activity measured in the media of choline-deficient cells was 2, 3.5, and 16 times greater, respectively, than that measured in the media of control cells at those same time points (Figure 3; each value significantly different from the control group at the same time point, P < 0.01).
FIGURE 3.. Mean (± SE) creatinine phosphokinase (CPK) activity in the media of muscle cells grown in a choline-deficient medium. C2C12 myocytes were grown in the defined medium containing 70 µmol choline/L for 48 h (time 0) and then switched to 70 µmol choline/L (•) or 0 µmol choline/L () for 24, 48, 72, and 96 h. Media were removed and analyzed for CPK activity with the use of a colorimetric assay. n = 3/group. There was a significant (P < 0.0001) time x treatment interaction. *Significantly different from values with 70 µmol choline/L at the same time point, P < 0.01.
At 24 and 48 h of treatment, there was no appreciable difference in DNA strand breaks, as measured by the TUNEL assay, between myocytes grown in the control and choline-deficient media (Figure 4). At 72 and 96 h in cells grown in choline-deficient medium, we detected 3-fold (P = 0.0002) and 9-fold (P = 0.0004) more TUNEL-positive myocytes, respectively, than we detected in cells grown in control medium at the same time points (Figure 4).
FIGURE 4.. Mean (± SE) percentage of cells positive on terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL; choline deficiency–induced apoptosis in C2C12 myocytes). Cells were grown in the defined medium with 70 µmol choline/L for 48 h (time 0) and then switched to 70 µmol choline/L (•) or 0 µmol choline/L () for 24, 48, 72, and 96 h. They were then fixed onto a slide, and cells with single-stranded breaks in DNA were detected by using a TUNEL method. At least 300 cells were counted per data point. n = 5–12/group. There was a significant (P < 0.0001) time x treatment interaction. *Significantly different from values with 70 µmol choline/L at the same time point, P < 0.01.
Amounts of activated caspase 3 did not change over time in cells grown in 70 µmol choline/L (Figure 5). However, activated caspase 3 was twice as high in cells grown in 0 µmol choline/L for 24 h and rose to 2.8 times control at 72 h (Figure 5; each different from control at the same time point, P < 0.01). Caspase 3 activation is an early marker of apoptosis (32) because it occurs significantly earlier than do DNA strand breaks, and thus 72 h was chosen as the time point for caspase 3 assessment.
FIGURE 5.. Mean (± SE) increase in activated caspase 3 in choline-deficient C2C12 myocytes. Cells were grown in the defined medium with 70 µmol choline/L for 48 h (time 0) and then switched to 70 µmol choline/L (•) or 0 µmol choline/L () for 24, 48, and 72 h. Activated caspase 3 was detected with the use of a fluorescent antibody and quantified by using the SCION IMAGE software program (version 1.61; Scion Corp, Frederick, MD). Integrated optical density was calculated from 10–17 images/group. There was a significant (P < 0.0001) time x treatment interaction. *Significantly different from values with 70 µmol choline/L at the same time point, P < 0.01.
LDH (a cytosolic enzyme) activity was low in the media of control cells (1.3 ± 0.1 IU/106 cells;
FIGURE 6.. Mean (± SE) increase in lactate dehydrogenase (LDH) activity as measured with the use of a colorimetric assay. Membrane fragility increased in choline-deficient C2C12 myocytes. C2C12 cells were grown in the defined medium with 70 µmol choline/L for 48 h (time 0) and then switched to 70 µmol choline/L (•) or 0 µmol choline/L () for 72 h. Media were removed, and 0–0.9% NaCl was added. After 30 min incubation at room temperature, the cell-free NaCl supernatant fluid was collected, and LDH activity was measured. n = 4–8/group. There was a significant (P < 0.0001) percentage salt x treatment interaction. *Significantly different from values with 70 µmol choline/L at the same time point, P < 0.01.
DISCUSSION
CPK is an intracellular protein found in skeletal muscle (MM isoform), brain tissue (BB isoform), and heart muscle (MB isoform). Elevated CPK in blood is usually caused by injury to muscle or by muscular dystrophy (33–35). This is the first report of CPK elevations in the serum of humans fed a choline-deficient diet (Figure 1), and it suggests that choline is important in maintaining the integrity of myocytes. In rabbits, choline deficiency was associated with a progressive muscular dystrophy (36–39). In our previous study, we observed that healthy male humans fed a choline-deficient diet developed liver damage (ie, elevated plasma alanine aminotransferase; 10), but we did not observe evidence of muscle damage. When we conducted that study, we did not realize that foods contained significant amounts of phosphocholine (18), and therefore we did not measure it when constructing low-choline diets. In the current study, the choline-deficient diet contains 50 mg total choline moiety, which makes this diet much lower in total choline content than was the previous diet (10). The previous diet also contained more folate (300 DFE/d) than does the current diet (100 DFE/d). Other investigators suggested that diets limited in folate increase the dietary requirement for choline (40, 41). It is possible that marginal folate intake in combination with choline deficiency resulted in the muscle damage observed in the present study. However, although plasma folate concentrations in the 4 subjects decreased during depletion, they never fell below the 6.6 nmol/L concentration that indicates negative folate balance (2). In addition, CPK activities returned to normal with consumption of a repletion diet that contained added choline but no added folate (with no restoration of plasma folate concentrations). Thus, we suggest that, although low folate may enable this effect of choline depletion, very low choline intake is the cause of the observed leakage of CPK that is indicative of muscle damage in choline-deficient males. We do not know why 1 of these 4 subjects did not develop muscle damage while consuming the choline-deficient diet; we speculate that the reason may be the subject's long-term dietary status, age, or genetic differences.
Cell culture model systems have been helpful in elucidating the effects of choline in several tissues. We now use C2C12 mouse myoblasts to study the effects of choline deficiency on muscle. These cells grew as well in a defined medium containing 70 µmol choline/L as they did in an established sera-containing medium, and they similarly expressed muscle-specific proteins. When choline was removed from the medium, C2C12 cells rapidly depleted their choline stores. In particular, choline deficiency was associated with a marked decline in phosphatidylcholine concentrations in the membranes of these myocytes. These changes in choline and its metabolites are similar to those described previously in choline-deficient hepatocytes, neurons, and PC12 cells (11, 42, 43). Choline-deficient myocytes also accumulated intracellular lipid droplets (Figure 2). Steatosis has been observed in choline-deficient liver and hepatocytes (5, 44–47), and its presence is likely due to a specific requirement for phosphatidylcholine for the export of VLDL from liver (46, 47). We do not know the mechanism for lipid accumulation in muscle cells.
In hepatocytes and PC12 neuroblastoma cells, choline deficiency induces apoptosis via signaling systems that involve transforming growth factor ß1, reactive oxygen species, nuclear factor-B, and caspase activation (11, 42, 48–51). Phosphatidylcholine concentrations in critical membranes are likely involved in the induction of choline deficiency–induced apoptosis (43), which results in the generation of reactive oxygen species from mitochondria and endoplasmic reticulum (48). We report that myocytes deprived of choline also undergo apoptosis as measured by DNA-strand breaks (TUNEL; Figure 4) and by activation of caspase 3 (Figure 5). Apoptotic cells could be expected to release CPK into medium, as we observed (Figure 3).
Because phosphatidylcholine is the predominant phospholipid in membranes (1), the integrity of the plasma membrane is compromised in choline deficiency. It is possible that myocytes deprived of choline are less resistant to mechanical stress. We tested this hypothesis by using hypo-osmotic media and found that choline-deficient myocytes did indeed leak intracellular proteins (LD) more readily than did control cells in a time period (30 min) when differences in apoptosis rates were insignificant.
Our studies indicate that choline deficiency depletes muscle cell phosphatidylcholine, inducing apoptosis and increasing membrane fragility, and that this results in the leakage of CPK from intracellular to extracellular space. We observed in some male humans that very significant elevations in CPK activity occur in serum when the men are deprived of dietary choline, and those CPK activities in serum return to normal when subjects are fed diets that contain 25% of the recommended AI of choline. We suggest that serum CPK may be a useful marker for choline deficiency in humans.
ACKNOWLEDGMENTS
We thank Renee Blanchard for help with the human subjects and Henry P. Goodell for his assistance with the osmotic fragility experiments. Marjorie Busby, Tricia Judd and Beth Macintosh designed the experimental human diets. Julie Scearce assisted with data analyses, and Larry Kupper gave invaluable help with our statistical analyses.
KdC and MB performed the myocyte studies described; LF and KdC performed the human studies described; and SZ designed and oversaw all studies. All authors participated in writing the manuscript. None of the authors had any personal or financial conflicts of interest.
REFERENCES