The Journal of Nutrition. First published ahead of print March 16, 2011 as doi: /jn

Transcription

1 The Journal of Nutrition. First published ahead of print March 16, 2011 as doi: /jn The Journal of Nutrition Biochemical, Molecular, and Genetic Mechanisms Glycine-N Methyltransferase Expression in HepG2 Cells Is Involved in Methyl Group Homeostasis by Regulating Transmethylation Kinetics and DNA Methylation 1,2 Yi-Cheng Wang, Feng-Yao Tang, Shih-Yin Chen, Yi-Ming Chen, and En-Pei Isabel Chiang* Department of Food Science and Biotechnology, National Chung Hsing University, Taichung, Taiwan Abstract Glycine-N methyltransferase (GNMT) is a potential tumor suppressor that is commonly inactivated in human hepatoma. We systematically investigated how GNMT regulates methyl group kinetics and global DNA methylation. HepG2 cells (GNMT inactive, GNMT2) and cells transfected with GNMT expressed vector (GNMT+) were cultured in low (10 mmol/l), adequate (100 mmol/l), or high (500 mmol/l) L-methionine, each with 2.27 mmol/l folate. Transmethylation kinetics were studied using stable isotopic tracers and GC-MS. Methylation status was determined by S-adenosylmethionine (SAM) and S-adenosylhomocysteine (SAH) levels, SAM:SAH ratio, DNA methyltransferase (DNMT) activity, and methylated cytidine levels in DNA. Compared with GNMT2 cells, GNMT+ cells had lower homocysteine and greater cysteine concentrations. GNMT expression increased methionine clearance by inducing homocysteine transsulfuration and remethylation metabolic fluxes when cells were cultured in high or adequate L-methionine. In contrast, homocysteine remethylation flux was lower in GNMT+ cells than in GNMT2 cells and homocysteine transsulfuration fluxes did not differ when cells were cultured in low methionine, suggesting that normal GNMT function helps to conserve methyl groups. Furthermore, GNMT expression decreased SAM and increased SAH levels and reduced DNMT activity in high or adequate, but not low, methionine cultures. In low methionine cultures, restoring GNMT in HepG2 cells did not lead to sarcosine synthesis, which would waste methyl groups. Methylated cytidine levels were significantly lower in GNMT2 cells than in GNMT+ cells. In conclusion, we have shown that GNMT affects transmethylation kinetics and SAM synthesis and facilitates the conservation of methyl groups by limiting homocysteine remethylation fluxes. J. Nutr. doi: /jn Introduction Methionine is essential to protein and polyamine synthesis, as well as cellular methylation reactions (1,2). Methyl deficiency can induce site-specific hypomethylation and increase the expression of genes involved in cell growth (3). Thus, methyl group depletion has been identified as a promoter for tumorigenesis and cancer progression (4). Methionine depletion is potentially carcinogenic, yet excessive methionine could be toxic (5 7). The optimal methionine concentration is crucial to normal growth and development in mammals. Glycine-N methyltransferase (GNMT; 3 EC ) is an abundant enzyme that accounts for.1% of the total cytosolic 1 Supported by the National Science Council in Taiwan (NSC B MY3, Chiang EP) and by the Department of Health in Taiwan (DOH 97-TD-D , Chiang EP). 2 Author disclosures: Y-C. Wang, F-Y.Tang, S-Y. Chen, Y-M. Chen, and E-P. I. Chiang, no conflicts of interest. 3 Abbreviations used: DNMT, DNA methyltransferase; GNMT, glycine-n methyltransferase; MAT, S-adenosylmethionine synthase; MTHFR, methylenetetrahydrofolate reductase; SAH, S-adenosylhomocysteine; SAM, S-adenosylmethionine. * To whom correspondence should be addressed. chiangisabel@nchu. edu.tw. protein in the liver (8). GNMT converts S-adenosylmethionine (SAM) to S-adenosylhomocysteine (SAH) while generating sarcosine from glycine. Therefore, GNMT serves as an alternative pathway to regulate the SAM:SAH ratio (9). About 50% of nontumorous liver tissues and 96% of hepatocellular carcinoma tissues do not express GNMT (10). Numerous hepatoblastoma and hepatocellular carcinoma cell lines do not express normal GNMT (10). As a result, it has been suggested that defective GNMT may be an early event in hepatocellular carcinoma development, and GNMT may serve as a new tumor susceptibility gene for hepatocellular carcinoma. On the other hand, recent studies found that sarcosine increased in metastatic prostate cancer cases and exogenous administration of sarcosine could enhance the invasiveness of prostate epithelial cell lines (11). These results suggest that components of the sarcosine pathway could be potential therapeutic targets of human prostate cancer invasion, and sarcosine and its regulatory enzymes may play a role in the modulation of cell invasion and migration. The regulation of GNMT in homocysteine transmethylation and catabolism kinetics under excessive or restricted methionine ã 2011 American Society for Nutrition. Manuscript received November 21, Initial review completed December 26, Revision accepted February 10, doi: /jn Copyright (C) 2011 by the American Society for Nutrition 1of6

2 conditions is still not fully understood. It is possible that GNMT can protect cells from methionine toxicity, because hepatic GNMT is upregulated by excess dietary methionine in vivo in a dose-dependent manner (12). However, it is still unclear whether the expression of GNMT exacerbates methionine depletion in liver cells. Aberrant upregulation of GNMT may lead to wastage of methyl groups due to the incorporation of methyl groups into sarcosine. Such an increase may lead to diminished hepatic levels of both methionine and SAM and possibly results in aberrant DNA methylation. Perturbed GNMT function may also lead to abnormal cellular methionine metabolism and transmethylation. The role of GNMT in hepatic one-carbon metabolic fluxes has not been firmly established. Specifically, the distribution of metabolic flux in methionine metabolism between transmethylation and transsulfuration under different methionine conditions should be further studied. Stable isotope tracer studies have been conducted to investigate how specific enzymes mediate the flux of one-carbon units of SAM, thymidylate, and de novo purine syntheses (13 16). By regulating the intracellular SAM status, GNMT may alter the multiple enzymes involved in the methionine cycle and thus affect homocysteine homeostasis via transmethylation and/or transsulfuration. The objective of this study was to investigate the role of GNMT in the regulation of the hepatic methyl groups and homocysteine kinetics in low, adequate, or high methionine cultures. We hypothesized that restoring GNMT function in GNMT deficient cells would facilitate the clearance of excessive methionine via homocysteine transsulfuration and/or remethylation. Moreover, we hypothesized that when methionine is restricted, GNMT expression does not lead to wastage of methyl groups, and cells with normal GNMT function are less susceptible to methionine depletion-induced DNA hypomethylation. Experimental Materials and Methods Cell lines and cell culture. Human hepatoblastoma cell line HepG2 was used in this study, because it was reported that HepG2 cells retain morphological and biochemical characteristics of normal human hepatocytes (17 19), albeit with diminished GNMT activity. The establishment of stable clones expressing GNMT has been described in detail (20). In the present study, cell line SCG2 1-1 was used as GNMT expressing cells (GNMT+) and SCG2-neg, a stable HepG2 cell line cotransfected with pflag-cmv-5 and ptk-hyg plasmids, was used as a control (GNMT2) (21). The GNMT2 and GNMT+ HepG2 human hepatoma cell lines were grown in MEM ALPHA medium containing 10% (v:v) dialyzed fetal calf serum (14), 0.12% sodium bicarbonate, penicillin (100 ku/l), streptomycin (100 mg/l), amphotericin (0.25 mg/l), and 5% CO 2 in an incubator at 378C. The medium was replaced every h. The treatments, which included high methionine (500 mmol/l L-methionine, 2.27 mmol/l folate), adequate methionine (100 mmol/l L-methionine, 2.27 mmol/l folate), and low methionine (10 mmol/l L-methionine, 2.27 mmol/l folate), took 144 h. Cysteine, homocysteine, SAM, and SAH. Cellular homocysteine and cysteine are maintained at low levels by frequent removal; thus, the concentrations of these thiol compounds in the medium reflect the quantity produced and exported by these cells during the experimental periods. After the treatment, medium and cell pellets were collected, protein was precipitated (22), and the supernatants were stored at 2808C for determination of cysteine and homocysteine (23). For analyses of SAM and SAH, cell pellets were extracted with 0.4 mol/l perchloric acid and the supernatants were filtered through 0.45-mm microspin filters prior to HPLC analysis (24). SAM synthase activity. After the treatment, cells were harvested, washed, pelleted, and homogenized in an ice-cold buffer consisting of mol/l potassium chloride/50 mmol/l Tris- HCl and EDTA (ph 7.4). The S-adenosylmethionine synthase (MAT) activity in cell extracts was determined by quantifying the SAM production by 300 mg cellular protein. The MAT activity was calculated as the SAM production after subtracting the baseline SAM and expressed as nmol SAM formed mg protein min 21 (15). Methionine and sarcosine. The cells were harvested, washed, pelleted, and homogenized in ice-cold 0.1 mol/l HCl. The cell lysate was centrifuged and the supernatant was derivatized with dabsyl-chloride. Intracellular methionine and sarcosine levels were determined by HPLC (Hitachi) using a UV detector (25). The methionine and sarcosine levels were expressed as pmol million cells 21. DNA methyltransferase activity. DNA methyltransferase (DNMT) activity was measured by incubating cell lysates containing 40 mg of protein with 0.5 mg of poly[d(i-c)d(i-c)] template (Amersham Pharmacia Biotech) and 3 mci [ 3 H]-SAM (15 mmol/l) (NEN Life Sciences) for 2 h at 378C as described previously. After the reaction, the DNA template was purified by organic extraction and ethanol precipitation. Pellets were resuspended in 0.3 mol/l sodium hydroxide, incubated at 378C for 1 h, spotted onto GF/C Whatman filter papers, and processed for liquid scintillation counting (15). Intracellular global DNA methylation. DNA was isolated from cells using a standard phenol/chloroform/isoamyl alcohol procedure. The degree of global DNA methylation was determined as the measured content of 5-methyldeoxycytidine in the DNA (15). The isolated DNA was treated with ribonuclease A and hydrolyzed with 2 enzymes, nuclease p1 and alkaline phosphatase at 378C for 5 h. The measurement of 5-methyldeoxycytidine content was achieved on the HPLC with a UV detector. Stable isotope tracer studies. To investigate the elevated GNMT expression influencing homocysteine remethylation fluxes, stable isotope tracer experiments were performed in the following procedures. HepG2 cells were plated in a 60-mm dish at 30 50% confluence in the treatment medium. The media were supplemented with either 13 C 5 -methionine (50% of total methionine) or L-[2, 3, 3-2 H 3 ] serine (50% of total serine) combined with L-[5, 5, 5-2 H 3 ] leucine (50% of total leucine) and were refreshed every 72 h (15). Cellular protein was isolated by lysing whole cell pellets with 5% ice-cold trichloroacetic acid and pelleting the protein by centrifugation at 48C. The supernatant and pellet were stored separately at 2808C for further analysis by GC-MS as previously described (13,15,16). Statistical analyses. Data from each of the biochemical analyses were analyzed for equality of variance. If the variance was heterogeneous, an appropriate transformation of the data was performed. A 2-factor ANOVA was used to analyze the effect of cell line and methionine condition. When the interaction was significant, all 6 means were compared with one another in the post hoc analysis. If the interaction was nonsignificant, a Tukey post hoc analysis was performed to determine whether there were 2 of 6 Wang et al.

3 differences due to cell line within the same methionine condition or if there were differences due to methionine condition within the same cell line. Pearson s correlation test was performed to examine correlations between continuous variables. A difference was considered significant at P, All statistical analyses were performed with Systat 10.0 for Windows (SYSTAT). Results Intracellular homocysteine concentration. GNMT protein expressions were determined in GNMT+ and GNMT2 cells cultured in high (500 mmol/l), adequate (100 mmol/l), or low (10 mmol/l) methionine (Fig. 1). In GNMT+ cells, GNMT protein expression was induced in high methionine and reduced in low methionine compared with the adequate methionine culture. In GNMT2 cells, GNMT protein expression was undetectable in all cases (Fig. 1). Compared with GNMT2 cells, the intracellular homocysteine concentration was significantly reduced in GNMT+ cells regardless of methionine conditions (Table 1). In addition, GNMT expression increased intracellular cysteine levels by 46 and 53% in medium containing adequate and high L-methionine, respectively (Table 1). These results suggest that GNMT expression may have increased the transsulfuration flux when the L-methionine supply was high or adequate, and GNMT has a potential role in sulfur amino acid homeostasis. On the other hand, when the cells were cultured in low methionine, the cysteine level was not greater in GNMT+ cells, suggesting that GNMT expression may have helped to conserve homocysteine for methionine synthesis when methionine was in demand (low methionine condition). Homocysteine reduced by induction of transsulfuration. The effect of GNMT on the transsulfuration flux was studied in subsequent experiments using stable isotopic tracers. In adequate and high methionine cultures, relative cystathionine s from tracer 13 C-serine were significantly increased in GNMT+ cells by 90 and 99%, respectively (Table 2), thereby FIGURE 1 GNMT protein levels in GNMT2 ( ) and GNMT+ (+) HepG2 cells incubated in high, adequate, or low concentrations of L-methionine for 144 h. Values are means + SE, n = 3 (untransformed data are presented). Effects of cell line, methionine, and their interaction were significant, P, Means without a common letter differ, P, TABLE 1 Intracellular homocysteine and cysteine concentrations in GNMT2 and GNMT+ HepG2 cells cultured in high (500 mmol/l), adequate (100 mmol/l), and low (10 mmol/l) methionine concentrations for 144 h 1 supply Cells Homocysteine Cysteine nmol/10 6 cells mmol/10 6 cells High GNMT a bc GNMT c a Adequate GNMT a c GNMT e a Low GNMT b b GNMT d bc Methionine P, P = GNMT P, P, Interaction P, P, Values are means 6 SD, n = 3. Untransformed data are shown. Means in a column without a common letter differ, P, supporting our postulation of induction of transsulfuration by GNMT expression. In contrast, when L-methionine supply was low, neither cysteine levels nor the transsulfuration flux differed between GNMT+ and GNMT2 cells. These results demonstrate that GNMT expression can facilitate the clearance of excessive methionine through homocysteine transsulfuration (Table 2). Homocysteine remethylation flux. Homocysteine remethylation was undetectable in the GNMT2 cells cultured in high or adequate methionine (Table 3). In contrast, the relative cytoplasmic methionine (M+4) from the 13 C 5 -methionine tracer was elevated in GNMT+ cells cultured in high or adequate methionine (Table 3), suggesting that under these conditions, GNMT expression can induce homocysteine remethylation (Table 3). Conversely, when methionine was low, overall remethylation flux was lower in GNMT+ cells than in GNMT2 cells (P = 0.022) (Table 3), suggesting that GNMT expression also helped to conserve the methyl groups by limiting homocysteine remethylation flux when methionine was in demand. SAM synthesis and sarcosine production. The intracellular methionine levels significantly increased when both cell lines were cultured in high methionine but were lower when cultured in low methionine compared with adequate methionine (Table 4). SAM was decreased and SAH increased by GNMT expression under all of the conditions examined (Table 4), presumably due to the accelerated conversion of SAM to SAH by GNMT. SAM concentrations correlated with methionine concentrations in the culture medium in both GNMT2 (r = 0.97; P, ; n = 9) and GNMT+ cells (r = 0.95; P, ; n = 9). The reduction of intracellular SAM by GNMT expression was responsive to methionine conditions in the medium (Table 4). Intracellular SAM was drastically reduced (;39%) in high methionine but only mildly reduced (;4.9%; P = 0.029) in low methionine (Table 4). MAT activity from cell extracts was correlated with the methionine concentration in the culture medium in both GNMT2 (r = 0.97; P, ) and GNMT+ cells (r = 0.93; P = 0.001). MAT activity was significantly elevated in GNMT+ cells compared with GNMT2 cells (Table 4). Intracellular sarcosine was undetectable in GNMT2 cells under all conditions examined (Table 4). In contrast, GNMT+ cells produced substantial levels of sarcosine Glycine-N methyltransferase regulates methyl group kinetics 3 of 6

4 TABLE 2 Cytoplasmic homocysteine transsulfuration in GNMT2 and GNMT+ HepG2 cells cultured in high (500 mmol/l), adequate (100 mmol/l), and low (10 mmol/l) methionine concentrations for 144 h 1 supply Cells Leucine+3 Serine+1 Cystathionine+1 Transsulfuration 2 High GNMT a c b GNMT a a a * Adequate GNMT a b b GNMT a a a * Low GNMT b b b GNMT b c b Methionine P, P, P = NS 3 GNMT NS 3 P, P, P = Interaction NS 3 P, P, NS 3 1 Values are means 6 SD, n = 3. Untransformed data are shown. Means in a column without a common letter differ, P, *Different from GNMT2, P, Relative s in cystathionine+1 from L-[ 13 C]-serine tracer. 3 NS, not significant, P $ that reflected the L-methionine supply in the medium (Table 4). These results suggest that GNMT expression in HepG2 cells could facilitate methyl donor homeostasis. Genomic DNA methyl cytidine content. GNMT expression significantly altered cellular methylation potential (Table 5). Furthermore, DNA methyltransferase activity in the GNMT+ cells was significantly lower than in the GNMT2 cells cultured in adequate or high methionine (Table 5). In high or adequate methionine cultures, methyl cytidine content in DNA was significantly reduced by GNMT expression (Table 5). Conversely, when cells were cultured in low methionine, despite the reduced SAM and increased SAH in GNMT+ cells, neither DNMT activity nor methyl cytidine content in the DNA was reduced (Table 5). The relatively elevated DNA methyl cytidine in GNMT+ cells cultured in low methionine indicated that restoring GNMT in these cells may alleviate hypomethylation due to the low methionine concentration. In sum, these results suggest that GNMTexpression may help to maintain methyl group homeostasis and methylation status via regulation of intracellular homocysteine transsulfuration and transmethylation kinetics. Discussion This study has demonstrated the role of GNMT in the regulation of hepatic methyl groups and homocysteine kinetics under various methionine conditions. The intracellular SAM concentration reflected the methionine level in cells with or without GNMT, and the percentage reduction of SAM due to the expression of GNMT was in proportion to the L-methionine concentrations in the culture medium. GNMTexpression enhanced the conversion of SAM to SAH, resulting in lower SAM levels compared with those in the GNMT2 cells. Furthermore, GNMT altered the transsulfuration enzyme that is allosterically regulated by SAM via its effect on SAM homeostasis. Our data demonstrated that GNMT expression can enhance homocysteine transsulfuration and, moreover, homocysteine remethylation fluxes when methionine is high or adequate. At present, it is unclear whether there is a secondary effect through allosteric regulation of cystathionine b synthase by SAM or there is a direct interaction between GNMT and cystathionine b synthase, or possibly with additional players. Yet our study clearly demonstrated that GNMT protects cells from both methionine excess and depletion, and it facilitates homocysteine clearance when needed (cells exposed to high methionine). Future studies on the mechanism by which GNMT facilitates homocysteine transsulfuration are warranted. Our study also demonstrated that normal GNMT function could be essential in optimizing methyl group supply and subsequent methylation reactions in humans with abnormal methionine or homocysteine metabolism. Because GNMT expression enhanced the conversion of SAM to SAH, it was necessary to examine whether upregulation of GNMT resulted in wastage of methyl groups when L-methionine supply was limited. Importantly, our data demonstrated that SAM concentrations were similar in GNMT2 and GNMT+ cells cultured in low methionine, indicating that GNMT expression does not decrease the SAM content when the L-methionine supply is low. This crucial observation demonstrated that GNMT expression does not exacerbate the methyl group deficiency when L-methionine supply is low in this model. Furthermore, GNMT+ cells had TABLE 3 supply Homocysteine remethylation fluxes in GNMT2 and GNMT+ HepG2 cells cultured in high (500 mmol/l), adequate (100 mmol/l) and low (10 mmol/l) methionine concentrations for 144 h 1 Cells Methionine+4 Methionine+5 Homocysteine remethylation 2 High GNMT ab a 0 d3 GNMT c b c Adequate GNMT b a 0 d3 GNMT bc b c Low GNMT a c a GNMT b c b Methionine P, P, P, GNMT P, P, P, Interaction P = P, P, Values are means 6 SD, n = 3. Untransformed data are shown. Means in a column without a common letter differ, P, The relative s in methionine+4 from methionine+5. 3 No homocysteine remethylation flux was detected under these conditions. 4 of 6 Wang et al.

5 TABLE 4 Intracellular methionine, MAT activity, SAM, SAH, and sarcosine productions in GNMT2 and GNMT+ HepG2 cells cultured in high, adequate, and low L-methionine for 144 h 1 supply Cells Methionine MAT SAM SAH Sarcosine 2 pmol/10 6 cells nmol/(mg protein h) nmol/(mg protein h) pmol/(mg protein h) pmol/10 6 cells High GNMT a c a c,0.039 d GNMT a a b a a Adequate GNMT b d b d,0.039 d GNMT b b c a b Low GNMT c e d c,0.039 d GNMT c d e b c Methionine P, P, P, P, P, GNMT NS 3 P, P, P, P, Interaction NS 3 P = P, P, P, Values are means 6 SD, n = 3. Untransformed data are shown. Means in a column without a common letter differ, P, For SAM, analysis was performed on untransformed data. 2 The concentration of sarcosine was below the detection limit of the assay (,0.039 mmol/l). 3 NS, not significant, P $ significantly elevated MAT activity, which could have contributed to the SAM supply to compensate for the accelerated transmethylation reactions in these cells. We suggest that GNMT+ cells maintain SAM homeostasis by increased MAT activity. Perturbed homocysteine metabolism has been observed clinically in some patients with cardiovascular diseases (26), Alzheimer s disease (27), renal dysfunction (28), diabetes (29), and chronic inflammation (30 32). A less active GNMT could be associated with increased plasma homocysteine concentrations. GNMT 1289TT carriers have higher plasma homocysteine compared with CC or CT genotypes when they are restricted in folate, possibly due to increased SAM, which can inhibit methylenetetrahydrofolate reductase (MTHFR) and decrease 5-methyl-tetrahydrofolate. As a potential regulatory mechanism for homocysteine homeostasis, normal GNMT function is essential for homocysteine clearance in humans with pathological defects in homocysteine metabolism. Results from the present study also suggest that normal GNMT function is essential for homocysteine removal via remethylation and/or transsulfuration, especially in patients with impaired homocysteine metabolism when dietary methionine consumption is excessive. In GNMT2 cells, SAM concentrations were closely regulated by the L-methionine concentrations in the culture medium. In contrast, GNMT+ cells were less sensitive (or resistant) to the high methionine levels in the culture medium. These observations suggest that restoring GNMT in GNMT-disrupted or mutant cells could help maintain methyltransferase function and DNA methylation. The presence of GNMT appeared to be essential to the actions of methyltransferase in low methionine cultures and the expression of GNMT would not lead to unnecessary SAM depletion when methionine is low. We previously demonstrated that transformed human lymphoblasts with lower MTHFR activity has advantages in de novo purine synthesis when folate is adequate but that they are more susceptible to SAM depletion when folate is restricted (16). Furthermore, in human colon and breast cancer cell models, the MTHFR 677T mutation can induce cell-specific changes in genomic DNA methylation and alter uracil misincorporations (33), which are both possible molecular bases for site-specific cancer risk modification (34). MTHFR, the enzyme that controls the utilization of different folate cofactors, is inhibited by SAM. Therefore, GNMT may indirectly alter the competition of folate cofactors between folatedependent reactions in liver by regulating intracellular SAM homeostasis. On the other hand, 5-methyl-tetrahydrofolate, the product of the MTHFR-catalyzing reaction and the primary methyl donor for homocysteine remethylation, directly inhibited GNMT (35,36). When the SAM level is high, MTHFR is inhibited and 5-methyl-tetrahydrofolate formation is reduced, resulting in a more active GNMT to reduce excess SAM levels. When SAM is low, a more active MTHFR leads to increased 5-methyltetrahydrofolate, which inhibits GNMT and spares methyl groups for transmethylation reactions. The interactions between these folate-dependent pathways are currently being investigated. In conclusion, this study has shown how GNMT expression affects homocysteine remethylation, MAT activity, and SAM synthesis. Specifically, GNMT protects cells from both methionine toxicity and deficiency. GNMT also facilitates homocysteine clearance when needed. This study presented new evidence TABLE 5 supply Methylation potential, DNMT activity and global DNA methylation in GNMT2 and GNMT+ HepG2 cells in GNMT2 and GNMT+ HepG2 cells cultured in high, adequate, and low L-methionine for 144 h 1 Cells Methylation potential 2 DNMT activity DNA methylation 3 MBq/mg protein % High GNMT a a b GNMT b b bc Adequate GNMT a a a GNMT c b c Low GNMT b b e GNMT c b d Methionine P, P = P, GNMT P, P, P = Interaction P, P, P, Values are means 6 SD, n = 3. Untransformed data are shown. Means in a column without a common letter differ, P, For DNA methylation, analysis was performed on untransformed data. 2 The SAM:SAH ratio. 3 The DNA methylation was calculated as percent of methylated cytidine (methylated +unmethylated cytidine) 21. Glycine-N methyltransferase regulates methyl group kinetics 5 of 6

6 that GNMT expression does not exacerbate methyl group deficiency when the L-methionine supply is limited and GNMT protein can protect against cellular methionine depletion and DNA hypomethylation. More studies are needed to determine whether these mechanisms may in part account for the protective role of GNMT against liver tumorigenesis. Studies of the mechanism by which GNMT facilitates homocysteine transsulfuration are underway. Acknowledgments We thank Yan-Jun Lin for technical support. E-P.I.C. designed research (project conception, development of overall research plan, and study oversight); Y-C.W. and F-Y.T. conducted research (hands-on conduct of the experiments and data collection); S-Y.C. and Y-M.C. provided cell lines; Y-C.W. and E-P.I.C. analyzed the data and performed statistical analysis; Y-C.W. and E-P.I.C. wrote the paper; Y-C.W. and E-P.I.C. had primary responsibility for final content. All authors read and approved the final manuscript. Literature Cited 1. Bartosinski B. Biosynthesis of methyl group of methionine. Postepy Biochem. 1964;10: Smith TK, Hyvonen T, Pajula RL, Eloranta TO. Effect of dietary methionine, arginine and ornithine on the metabolism and accumulation of polyamines, S-adenosylmethionine and macromolecules in rat liver and skeletal muscle. Ann Nutr Metab. 1987;31: Dizik M, Christman JK, Wainfan E. Alterations in expression and methylation of specific genes in livers of rats fed a cancer promoting methyl-deficient diet. Carcinogenesis. 1991;12: Liu SP, Li YS, Chen YJ, Chiang EP, Li AF, Lee YH, Tsai TF, Hsiao M, Huang SF, et al. Glycine N-methyltransferase2/2 mice develop chronic hepatitis and glycogen storage disease in the liver. Hepatology. 2007;46: Kato H, Saito A, Meguro K. Amino acidtoxicity and congenital amino acid metabolic disorders. Nippon Shonika Gakkai Zasshi. 1964;68: Satter LD, Lang RL, Van Loo JW, Carlson ME, Kepler RW. Adverse effect of excess methionine or methionine hydroxy analog on feed consumption in cattle. J Dairy Sci. 1975;58: Wainfan E, Poirier LA. Methyl groups in carcinogenesis: effects on DNA methylation and gene expression. Cancer Res. 1992;52:2071s 7s. 8. Heady JE, Kerr SJ. Purification and characterization of glycine N-methyltransferase. J Biol Chem. 1973;248: Cook RJ, Wagner C. Glycine N-methyltransferase is a folate binding protein of rat liver cytosol. Proc Natl Acad Sci USA. 1984;81: Liu HH, Chen KH, Shih YP, Lui WY, Wong FH, Chen YM. Characterization of reduced expression of glycine N-methyltransferase in cancerous hepatic tissues using two newly developed monoclonal antibodies. J Biomed Sci. 2003;10: Sreekumar A, Poisson LM, Rajendiran TM, Khan AP, Cao Q, Yu J, Laxman B, Mehra R, Lonigro RJ, et al. Metabolomic profiles delineate potential role for sarcosine in prostate cancer progression. Nature. 2009;457: Rowling MJ, McMullen MH, Chipman DC, Schalinske KL. Hepatic glycine N-methyltransferase is up-regulated by excess dietary methionine in rats. J Nutr. 2002;132: Herbig K, Chiang EP, Lee LR, Hills J, Shane B, Stover PJ. Cytoplasmic serine hydroxymethyltransferase mediates competition between folatedependent deoxyribonucleotide and S-adenosylmethionine biosyntheses. J Biol Chem. 2002;277: Anguera MC, Field MS, Perry C, Ghandour H, Chiang EP, Selhub J, Shane B, Stover PJ. Regulation of folate-mediated one-carbon metabolism by 10-formyltetrahydrofolate dehydrogenase. J Biol Chem. 2006; 281: Chiang EP, Wang YC, Chen WW, Tang FY. Effects of insulin and glucose on cellular metabolic fluxes in homocysteine transsulfuration, remethylation, S-adenosylmethionine synthesis, and global deoxyribonucleic acid methylation. J Clin Endocrinol Metab. 2009;94: Chiang EP, Wang YC, Tang FY. Folate restriction and methylenetetrahydrofolate reductase 677T polymorphism decreases adomet synthesis via folate-dependent remethylation in human-transformed lymphoblasts. Leukemia. 2007;21: Fukuda Y, Ishida N, Noguchi T, Kappas A, Sassa S. Interleukin-6 down regulates the expression of transcripts encoding cytochrome P450 IA1, IA2 and IIIA3 in human hepatoma cells. Biochem Biophys Res Commun. 1992;184: Aden DP, Fogel A, Plotkin S, Damjanov I, Knowles BB. Controlled synthesis of HBsAg in a differentiated human liver carcinoma-derived cell line. Nature. 1979;282: Knowles BB, Howe CC, Aden DP. Human hepatocellular carcinoma cell lines secrete the major plasma proteins and hepatitis B surface antigen. Science. 1980;209: Chen SY, Lin JR, Darbha R, Lin P, Liu TY, Chen YM. Glycine N-methyltransferase tumor susceptibility gene in the benzo(a)pyrenedetoxification pathway. Cancer Res. 2004;64: Wang YC, Chen YM, Lin YJ, Liu SP, Chiang EP. GNMT expression increases hepatic folate contents and folate-dependent methionine synthase-mediated homocysteine remethylation. Mol Med. Epub 2011 Jan Chiang EP, Bagley PJ, Roubenoff R, Nadeau M, Selhub J. Plasma pyridoxal 59-phosphate concentration is correlated with functional vitamin B-6 indices in patients with rheumatoid arthritis and marginal vitamin B-6 status. J Nutr. 2003;133: Araki A, Sako Y. Determination of free and total homocysteine in human plasma by high-performance liquid chromatography with fluorescence detection. J Chromatogr. 1987;422: Fell D, Benjamin LE, Steele RD. Determination of adenosine and S-adenosyl derivatives of sulfur amino acids in rat liver by highperformance liquid chromatography. J Chromatogr. 1985;345: Lin JK, Lai CC. High performance liquid chromatographic determination of naturally occurring primary and secondary amines with dabsyl chloride. Anal Chem. 1980;52: Clarke R, Daly L, Robinson K, Naughten E, Cahalane S, Fowler B, Graham I. Hyperhomocysteinemia: an independent risk factor for vascular disease. N Engl J Med. 1991;324: Joosten E, Lesaffre E, Riezler R, Ghekiere V, Dereymaeker L, Pelemans W, Dejaeger E. Is metabolic evidence for vitamin B-12 and folate deficiency more frequent in elderly patients with Alzheimer s disease? J Gerontol A Biol Sci Med Sci. 1997;52:M Robins AJ, Milewczyk BK, Booth EM, Mallick NP. Plasma amino acid abnormalities in chronic renal failure. Clin Chim Acta. 1972;42: Hultberg B, Agardh E, Andersson A, Brattstrom L, Isaksson A, Israelsson B, Agardh CD. Increased levels of plasma homocysteine are associated with nephropathy, but not severe retinopathy in type 1 diabetes mellitus. Scand J Clin Lab Invest. 1991;51: Chiang EP, Selhub J, Bagley PJ, Dallal G, Roubenoff R. Pyridoxine supplementation corrects vitamin B6 deficiency but does not improve inflammation in patients with rheumatoid arthritis. Arthritis Res Ther. 2005;7:R Chiang EP, Smith DE, Selhub J, Dallal G, Wang YC, Roubenoff R. Inflammation causes tissue-specific depletion of vitamin B6. Arthritis Res Ther. 2005;7:R Chiang EP, Bagley PJ, Selhub J, Nadeau M, Roubenoff R. Abnormal vitamin B(6) status is associated with severity of symptoms in patients with rheumatoid arthritis. Am J Med. 2003;114: Sohn KJ, Jang H, Campan M, Weisenberger DJ, Dickhout J, Wang YC, Cho RC, Yates Z, Lucock M, et al. The methylenetetrahydrofolate reductase C677T mutation induces cell-specific changes in genomic DNA methylation and uracil misincorporation: a possible molecular basis for the site-specific cancer risk modification. Int J Cancer. 2009; 124: Stempak JM, Sohn KJ, Chiang EP, Shane B, Kim YI. Cell and stage of transformation-specific effects of folate deficiency on methionine cycle intermediates and DNA methylation in an in vitro model. Carcinogenesis. 2005;26: Yeo EJ, Briggs WT, Wagner C. Inhibition of glycine N-methyltransferase by 5-methyltetrahydrofolate pentaglutamate. J Biol Chem. 1999;274: Wagner C, Briggs WT, Cook RJ. Inhibition of glycine N-methyltransferase activity by folate derivatives: implications for regulation of methyl group metabolism. Biochem Biophys Res Commun. 1985;127: of 6 Wang et al.