The Journal of Nutrition Vol. 128 No. 6 June 1998, pp. 973-976

Histidine Availability Alters Glucagon Gene Expression in Murine TC6 Cells^1,2

Gregory L. Paul, Aparna Waegner, H. Rex Gaskins^*, and Neil F. Shay³

Department of Food Science and Human Nutrition, ^* Department of Animal Sciences, and Division of Nutritional Sciences, University of Illinois, Urbana-Champaign, IL 61801

	ABSTRACT

Abstract Introduction Results Discussion References

Because individual amino acids (AA) stimulate glucagon release from pancreatic -cells, the purpose of this study was to determine if individual AA could influence glucagon gene expression. Preproglucagon mRNA levels were 67% lower (P < 0.05) in mouse TC6 cells incubated for 12 h in amino acid-free medium compared with cells incubated in complete medium containing all 20 AA. A time-course study indicated that TC6 cells incubated in amino acid-free medium ± 1 µmol/L puromycin or amino acid-containing medium plus puromycin exhibited similar preproglucagon mRNA decreases over 12 h. When 1 µmol/L actinomycin was added to medium with or without AA, ppG mRNA concentrations decreased (P < 0.05) for 3 h; however, values at 12 h were not different than those at 3 h. Deletions of single AA from complete medium demonstrated that only histidine removal or depletion reproduced the decrease in ppG mRNA observed in amino acid-free medium. We conclude that histidine is involved in the regulation of preproglucagon mRNA levels in TC6 cells and that this regulation may be operative during both transcriptional and post-transcriptional events.

	INTRODUCTION

Abstract Introduction Results Discussion References

Glucagon, a 29 amino acid (AA)⁴ polypeptide hormone secreted by pancreatic -cells, plays an essential role in macronutrient metabolism as a regulator of liver glycogenolysis, ketogenesis and gluconeogenesis (Unger and Orci 1981). Acting in concert with insulin, glucagon mediates a regulated flow of fuels to metabolically active tissues over a wide range of physiologic conditions (Unger and Orci 1981). Glucagon is translated as a prohormone in -cells from preproglucagon mRNA (Habener et al. 1991). Preproglucagon mRNA is also found in intestinal L-cells and selected neurons of the hypothalamus and brain stem. Post-translational processing of proglucagon produces tissue-cell-specific peptides (Habener et al. 1991) that are secreted differentially in response to glucagon secretagogues. For example, glucose enhances glucagon peptide secretion from intestinal L-cells (Kreymann et al. 1987) while inhibiting pancreatic -cell peptide secretion (Habener et al. 1991, Pipeleers et al. 1985).

Amino acids also influence -cell glucagon secretion. Glucagon secretion by purified rat -cells is increased in medium containing glutamine, alanine and arginine (Pipeleers et al. 1985). Perfusion studies with isolated rat pancreas indicate that arginine, acting in conjunction with thyrotropin-releasing hormone, stimulates glucagon secretion (Ebiou et al. 1992).

The potentiating effect of AA on glucagon secretion has also been observed in vivo. Alanine and glycine, infused individually into sheep, increased plasma glucagon levels 300% over baseline values, whereas infusions of serine, arginine, asparagine or glutamine produced 100-200% increases in plasma glucagon (Kuhara et al. 1991). Hyperaminoacidemia increased plasma glucagon concentrations in healthy human volunteers during insulin-induced hypoglycemia (Caprio et al. 1993), a finding supporting earlier in vivo work (Nair et al. 1990), and consistent with the in vitro work of Pipeleers and others (1985).

If glucagon secretion is related to preproglucagon mRNA levels, it follows that AA availability may regulate glucagon gene expression in -cells. Amino acid availability regulates the expression of several mammalian genes including asparagine synthetase (Gong et al. 1991), ribosomal protein L17 (Laine et al. 1994, Shay et al. 1990), insulin-like growth factor I (IGF-I; Thissen et al. 1994), and c-myc (Yokota et al. 1995). Thus, the purpose of this study was to characterize how individual AA affect glucagon gene expression in TC6 cells, an islet-cell line transformed by the simian virus 40 large tumor antigen (Powers et al. 1990). The study represents the initial steps towards understanding the role specific nutrients play at the molecular level in regulating plasma glucagon concentrations. This knowledge may ultimately prove useful for designing therapies to help normalize the aberrant insulin:glucagon ratios implicated in numerous pathologies associated with diabetes mellitus.

	EXPERIMENTAL PROCEDURES

Cell culture.  TC6 cells were grown in Dulbecco's modified Eagle's medium (25 mmol/L glucose) supplemented with 15% horse serum, 2.5% fetal clone II (HyClone Laboratories, Logan, UT), 50 mg/L gentamicin (GIBCO, Gaithersburg, MD), and 25 mg/L Fungizone (GIBCO). Cells were maintained in an atmosphere of 95% air/5% CO₂ at 37°C. For each experiment, cells were seeded in 24-well plates (Corning, Corning, NY) at 2 × 10⁸ cells/L. At 70% confluence, cells were incubated in serum-free medium for 24 h, then washed three times with PBS. Cells were then incubated for 3-12 h in serum-free RPMI 1640 medium (GIBCO) containing 12.5 mmol/L glucose. Preproglucagon mRNA abundance was not affected by changing the medium glucose concentration from 25 to 12.5 mmol/L. Amino acids were added to the medium individually (Select-Amine kit, GIBCO) at concentrations (5-300 mg/L) specified in the RPMI formulation. When used in time-course studies, metabolic inhibitors actinomycin D (transcription inhibitor, Sigma Chemical, St. Louis, MO) or puromycin (translation inhibitor, Sigma) were added to the medium at a final concentration of 1 µmol/L to examine contributions of transcription or translation to changes in ppG gene expression. Cell viability in all treatments up to 24 h was confirmed at >95% by using trypan blue exclusion. Unless specifically indicated otherwise, n = 3-4 for each treatment in all cell experiments. Each study was replicated with independent groups of cells.

Mouse preproglucagon cDNA synthesis.  Two oligodeoxyribonucleotide primers were designed from the rat preproglucagon cDNA sequence (Heinrich et al. 1984) as follows: 5'-ATGAAGACCGTTTACATCGTGGCTGGATTG-3' (5' end, bases 60-89) and 5'-AATCAGCCAGTTGATGAAGTCTCTGGTGGC-3' (3' end, bases 547-576); they were synthesized by the University of Illinois DNA Synthesis Core Facility. Total RNA (4 µg) from TC9 cells was reverse transcribed in a 20 µL reaction mixture by using a first-strand cDNA reaction kit and the manufacturer's protocol (GIBCO). The cDNA product was amplified by polymerase chain reaction (PCR) in a 40-µL reaction containing 6 µL of the reverse transcriptase reaction mixture, 20 mmol/L tris-HCl, 50 mmol/L KCl, 1.5 mmol/L MgCl₂, 2.5 U Taq DNA polymerase (GIBCO) and 10 µmol/L of each primer. The reaction mixture was incubated for five cycles of 1 min at 94°C, 1 min at 40°C and 1 min at 72°C, followed by 25 cycles of 1 min at 94°C and 2 min at 55°C. A 516-bp PCR product was identified by electrophoresis and cloned into pCR II vector by using a TA Cloning kit (Invitrogen, San Diego, CA). The cloned cDNA insert was sequenced by the dideoxy method and compared with the rat preproglucagon sequence to confirm its identity.

Ribonuclease protection assay.  The pCR II vector containing the 516-bp glucagon cDNA sequence was cut at two sites with Rca I (GIBCO) to produce fragments of 1428 and 3020 bp. Transcription of the 1428-bp fragment in the presence of SP6 polymerase (MAXIscript, Ambion, Austin, TX) and ³²P-CTP (800 Ci/mmol, Amersham International, Arlington Heights, IL) generated a 341-bp preproglucagon antisense riboprobe with a specific activity of ~36 GBq/µg (~6 × 10⁸ cpm/µg). The ribonuclease protection assay was performed according to Arkins et al. (1994) by using 0.25-0.5 µg total RNA isolated from TC6 cells (passage 17-26) with Ultraspec Reagent (Biotecx Laboratories, Houston, TX). Cells were rinsed twice with PBS to remove poorly adhering cells before isolation of RNA. Hybridization was performed at 45°C for 18 h. Twenty units RNase T1 and 1 µg RNase A (both from Sigma) were added to digest unprotected RNA. This generated a 244-bp protected RNA fragment from the original 341-bp riboprobe, which was resolved on a 5% polyacrylamide native gel. The gel was then dried and the protected fragments were quantitated by phosphorimaging (Molecular Dynamics, Sunnyvale, CA). For purposes of normalization, 18S rRNA levels were measured by using a commercially available 97-bp riboprobe (Ambion) that yielded an 80-bp protected fragment after hybridization and digestion. Because 18S rRNA represents nearly 20% of total rRNA, an 18S riboprobe with a specific activity of ~18 MBq/µg (3 × 10⁵ cpm/µg) was required to produce a band density similar to the preproglucagon riboprobe. Separation of fragments was performed using a 5% polyacrylamide gel containing 8 mol/L urea to ensure complete denaturation of the unprotected riboprobes. The 18S protected fragment runs as a doublet when separated on a denaturing gel, but is observed as a single band after nondenaturing gel electrophoresis. Levels of the 18S rRNA did not vary more than 10% between conditions. Hence, changes in preproglucagon mRNA rather than the 18S abundance represent treatment differences in preproglucagon mRNA:18S rRNA ratios.

Statistics.  Data for the 0 and 20 AA time-course experiments were analyzed by using a three-way ANOVA with time, AA concentration, and inhibitor concentration as independent variables (SigmaStat, Sausalito, CA). A two-way ANOVA was used to analyze the histidine deletion time-course data. A one-way ANOVA was used to analyze data from all other experiments. Tukey's test (Montgomery 1984) was performed to identify individual mean differences when significant interactions or main effects were observed. The level of significance was set at P < 0.05.

Format of studies.  All studies used time courses of up to 12 h, which we found to be an acceptable length with which to study changes in ppG mRNA expression without compromising cell viability (data not shown). Study 1: Preproglucagon:18S rRNA levels were measured in 0 AA- and 20 AA-containing media. Study 2: To help identify whether the observed preproglucagon mRNA differences between the 0 and 20 AA treatments resulted from transcriptional or post-transcriptional events, a time-course study was conducted utilizing 1 µmol/L actinomycin D (ACT), an inhibitor of transcription, or 1 µmol/L puromycin (PUR), an inhibitor of translation, in medium containing either 0 or 20 AA. Preproglucagon mRNA levels were measured when the maintenance medium was replaced with treatment medium (time 0) and 3 and 12 h later. Study 2 was also repeated with the translation inhibitor cycloheximide and transcription inhibitor -amanitin (data not shown). Study 3: Amino acids were individually deleted from complete medium for up to 12 h. A 0 AA treatment was also used (as a second control), and cells were incubated for 6 h. Study 4: cells were incubated for 3, 6 or 12 h in His medium with or without ACT or PUR. A 20 AA and a 0 AA treatment were included as controls.

	RESULTS

Abstract Introduction Results Discussion References

Preproglucagon gene expression was stable for up to 24 h when TC6 cells were switched from normal culture glucose concentrations (25 mmol/L) to lower glucose concentrations (data not shown). This range included 12.5 mmol/L glucose, the glucose concentration of RPMI medium, which was used in all subsequent experiments. Preproglucagon gene expression was 67% lower (P < 0.05) in TC6 cells incubated for 12 h in amino acid-free medium (0 AA) compared with medium containing 20 amino acids (20 AA, Fig. 1). A significant interaction was found among time, AA concentration and inhibitor concentration. After 3 h of incubation, preproglucagon mRNA abundance was significantly lower in TC6 cells incubated in 0 AA + ACT and 20 AA + PUR compared with 0 AA or 0 AA + PUR; however, all treatments were significantly lower than 20 AA. Preproglucagon mRNA levels decreased significantly between 3 and 12 h in cells incubated in 0 AA, 0 AA + PUR and 20 AA + PUR, but did not change in cells incubated in ACT with either 0 or 20 AA. After 12 h of incubation, preproglucagon mRNA abundance did not differ in TC6 cells incubated in 0 AA, 0 AA + PUR, 20 AA + PUR and 0 AA + ACT; however, these treatments were all significantly lower than 20 AA and 20 AA + ACT. Additional data obtained using cycloheximide and -amanitin (results not shown) resembled the data obtained using puromycin and actinomycin D, respectively, indicating that the results described above were not dependent on the specific inhibitor used. Both inhibitors of translation produced similar results, and both inhibitors of transcription produced similar results.

View larger version (18K): [in this window] [in a new window]   Fig 1. Change over time of the preproglucagon mRNA:18S rRNA ratio in the presence or absence of amino acids and inhibitors of transcription and translation. Murine TC6 cells were seeded in a 24-well plate at 200,000 cells/well. At 70% confluence, maintenance medium was removed and cells were incubated in RPMI-1640 medium containing no amino acids (0), 20 amino acids (20), 0 + 1 µmol/L puromycin (0/P), 20 + 1 µmol/L puromycin (20/P), 0 + 1 µmol/L actinomycin D (0/A) or 20 + 1 µmol/L actinomycin D (20/A) for 3 or 12 h. A significant interaction among time, amino acid concentration, and inhibitor concentration was found (three-way ANOVA). Means with the same letter at the same time do not differ significantly as determined by Tukey's test. Each data point represents the mean ± SEM, n = 3.

View larger version (86K): [in this window] [in a new window]   Fig 2. Effect of individual amino acid deletions on the preproglucagon mRNA:18S rRNA ratio in murine TC6 cells. TC6 cells were seeded in 24-well plates at 200,000 cells/well. At 70% confluence, maintenance medium was removed and cells were incubated for 12 h in RPMI-1640 medium containing 0 amino acids, 20 amino acids or 20 amino acids minus a specific amino acid. Cells incubated in 0 amino acids and 20 amino acids minus histidine had significantly lower preproglucagon mRNA:18S rRNA levels than all others as determined by a one-way ANOVA and Tukey's test. Data from two identical experiments, which elicited the same results with four samples per treatment, were combined. Thus, each preproglucagon mRNA:18S rRNA represents the mean ± SEM, n = 8. Asterisks indicate values that are significantly different (P < 0.05) from all unmarked values.

To determine whether one or more AA accounted for the significantly lower preproglucagon gene expression in TC6 cells incubated in 0 AA compared with 20 AA, an AA deletion experiment was performed. As shown in Figure 2, only deletion of histidine from medium (His) containing 20 AA reproduced the decrease in preproglucagon mRNA abundance observed when cells were incubated in 0 AA. Figure 3 is a representative phosphorimage of the protected preproglucagon mRNA and 18S rRNA from cells incubated 6 h in medium containing either 0 AA, 20 AA or His.

View larger version (70K): [in this window] [in a new window]   Fig 3. Phosphorimage of the protected preproglucagon mRNA and 18S rRNA in murine TC6 cells incubated in amino acid (AA) free, AA-containing and AA-containing medium without histidine. The protected fragments were separated in a 5% polyacrylamide, non-denaturing gel. TC6 cells were seeded in a 24-well plate at 200,000 cells/well. At 70% confluence, maintenance medium was removed and cells were incubated in RPMI-1640 medium containing no AA (0), 20 AA (20), and 20 AA minus histidine (His) for 12 h.

We next sought to determine if preproglucagon gene expression is regulated similarly when TC6 cells are incubated in medium containing either His or 0 AA. As depicted in Figure 4, the preproglucagon mRNA decay time courses were similar for cells incubated in 0 AA, His, or His + PUR. Cells incubated in His + ACT exhibited an initial decrease in preproglucagon mRNA, but then levels stabilized and did not change between 3 and 12 h. This decay pattern is similar to the pattern observed when cells were incubated in either 0 AA + ACT or 20 AA + ACT (Fig. 1).

View larger version (22K): [in this window] [in a new window]   Fig 4. The effect of histidine deletion on the preproglucagon mRNA:18S rRNA time course in murine TC6 cells. TC6 cells were seeded in a 24-well plate at 200,000 cells/well. At 70% confluence, maintenance medium was removed and cells were incubated in RPMI-1640 medium containing no amino acids (0), 20 amino acids (20), 20 minus histidine (His), His + 1 µmol/L actinomycin D (His/A), and His + 1 µmol/L puromycin (His/P) for 3, 6 or 12 h. A significant interaction among treatment and time was found (three-way ANOVA). Means with the same letter at the same time do not differ significantly as determined by Tukey's test. Each data point represents the mean ± SEM, n = 4.

	DISCUSSION

Abstract Introduction Results Discussion References

These results provide the first documentation that AA availability influences glucagon gene expression. After 3 h of incubation in amino acid-free medium, preproglucagon gene expression by TC6 cells was significantly depressed compared with cells incubated in medium containing 20 AA. Only the deletion of histidine from complete media mimicked results observed in amino acid-free medium. AA deprivation decreases the expression of other mammalian genes. IGF-I gene expression decreases nearly 60% after 24 h in cultures of rat primary hepatocytes incubated in medium containing 0.2× AA compared with medium containing 1× amino acids (Pao et al. 1993, Thissen et al. 1994). Others (Laine et al. 1994, Yokota et al. 1995) have reported occurrences of AA deprivation increasing the expression of specific genes.

Our results demonstrate that the regulation of glucagon gene expression by AA also involves factors dependent upon both transcription and translation. TC6 cells, incubated in the presence of puromycin, exhibited a similar decay in preproglucagon mRNA regardless of medium AA content, and this decay was essentially identical to that observed in cells incubated in amino acid-free medium without puromycin (Fig. 3). We suggest that these observations are consistent with the mRNA of a preproglucagon mRNA stabilizing factor being translated when cells are incubated in amino acid-complete medium, thus maintaining a normal preproglucagon mRNA turnover rate and normal preproglucagon mRNA levels. However, when amino acid-starved cells were incubated in the presence of actinomycin D, preproglucagon mRNA levels dropped during the initial 3 h, then remain unchanged. If translation of the preproglucagon mRNA stabilizing factor mRNA requires the presence of all 20 AA, then the preproglucagon mRNA should have decayed further rather than remain unchanged in cells incubated in amino acid-free medium with actinomycin D.

Previous in vitro studies in mammalian systems have shown that several genes are responsive to single AA deletions. Gong and colleagues (1991) found that asparagine synthetase gene expression increased in ts11 cells incubating in asparagine-free medium and medium lacking leucine, isoleucine or glutamine. Upon repletion with asparagine, asparagine synthetase gene expression was suppressed. By inhibiting transcription and translation, it was shown that changes in synthesis of RNA and protein were involved in regulating asparagine synthetase gene expression in response to AA starvation. Recently, Marten et al. (1994) found striking differences in mRNA abundance for 19 genes in rat hepatoma cells limited for phenylalanine, methionine, leucine or tryptophan. Genes that decreased expression in response to AA limitation were liver specific and defined as Class I genes. Inhibitor studies indicated that the decline in mRNA abundance was due to decreased gene transcription as well as increased mRNA turnover.

Previous in vivo studies demonstrated that individual AA can stimulate pancreatic glucagon secretion. Kaneto and Kosaka (1971) reported that either arginine or histidine infused into the pancreaticoduodenal artery of dogs increased pancreatic glucagon secretion. Kuhara et al. (1991) found that alanine, glycine, serine, arginine and asparagine, but not histidine, stimulated glucagon secretion after each was intravenously infused into sheep. Because we did not measure glucagon gene expression in response to the addition of each individual AA, it remains to be determined whether glucagon secretion stimulated by individual AA relates to gene expression.

Although it is not possible to infer an in vivo role for histidine in regulating glucagon gene expression, the observations do provide clear evidence that histidine is required to maintain preproglucagon mRNA levels in TC6 cells. Future investigation of the mechanism involved may reveal how preproglucagon message levels are controlled by AA supply. This might be advantageous in conditions such as insulin dependent and noninsulin-dependent diabetes mellitus in which disregulation of glucagon gene expression is thought to contribute to the elevated release of glucagon from pancreatic -cells (Dobbs et al. 1975).

	FOOTNOTES

	LITERATURE CITED

Abstract Introduction Results Discussion References

• Arkins S., Liu Q., Kelley K. W. Theoretical and functional aspects of measuring insulin-like growth factor-I mRNA expression in myeloid cells. Immunomethods 1994; 5:8-20[Medline]
• Caprio S., Tabmorlane W. V., Zych K., Gerow K., Sherwin R. S. Loss of potentiating effect of hypoglycemia on the glucagon response to hyperaminoacidemia in IDDM. Diabetes 1993; 42:550-555[Abstract]
• Dobbs, R., Sakurai, H., Sasaki, H., Faloona, G., Valverde, I., Baetens, D., Orci, L. & Unger, R. (1975) Glucagon: role in the hyperglycemia of diabetes mellitus. Science (Washington, DC) 187: 544-547.
• Ebiou J.-C., Grouselle D., Aratan-Spire S. Antithyrotopin-releasing hormone serum inhibits secretion of glucagon from isolated perfused rat pancreas: an experimental model for positive feedback regulation of glucagon secretion. Endocrinology 1992; 131:765-771[Abstract/Free Full Text]
• Gong S. S., Guerrini L., Basilico C. Regulation of asparagine synthetase gene expression by amino acid starvation. Mol. Cell. Biol. 1991; 11:6059-6066[Abstract/Free Full Text]
• Habener, J. F., Drucker, D. J., Mojsov, S., Knepel, W. & Philippe, J. (1991) Biosynthesis of glucagon. In: The Endocrine Pancreas (Samols, E., ed.), pp. 53-71. Raven Press, New York, NY.
• Heinrich G., Gros P., Habener J. F. Glucagon gene sequence. Four of six exons encode separate functional domains of rat pre-proglucagon. J. Biol. Chem. 1984; 259:14082-14087[Abstract/Free Full Text]
• Kaneto A., Kosaka K. Effect of leucine and isoleucine infused intrapancreatically on glucagon and insulin secretion. Endocrinology 1971; 88:1239-1245[Abstract/Free Full Text]
• Kreymann B., Williams G., Ghatei M. A., Bloom S. R. Glucagon-like peptide-I 7-36 amide: a physiological incretin in man. Lancet 1987; 11:1300-1303
• Kuhara T., Ikeda S., Ohneda A., Sasaki Y. Effects of intravenous infusion of 17 amino acids on the secretion of GH, glucagon, and insulin in sheep. Am. J. Physiol. 1991; 260:E21-E26[Abstract/Free Full Text]
• Laine R. O., Shay N. F., Kilberg M. S. Amino acid-dependent transcriptional regulation of mammalian ribosomal proteins L17 and S25: retention of the induced mRNA. J. Biol. Chem. 1994; 269:9693-9697[Abstract/Free Full Text]
• Marten N. W., Burke E. J., Hayden J. M., Straus D. S. Effect of amino acid limitation on the expression of 19 genes in rat hepatoma cells. FASEB J. 1994; 8:538-544[Abstract]
• Montgomery, D. C. (1984) Design and Analysis of Experiments, 2nd ed. John Wiley & Sons, New York, NY.
• Nair S. K., Wille S. L., Tito J. Effect of plasma amino acid replacement on glucagon and substrate responses to insulin-induced hypoglycemia in humans. Diabetes 1990; 39:376-382[Abstract]
• Pao C.-I., Farmer P. K., Begovic S., Villafuerte B. C., Wu G., Robertson D. G., Phillips L. S. Regulation of insulin-like growth factor-I (IGF-I) and IGF-binding protein I gene transcription by hormones and provision of amino acids in rat hepatocytes. Endocrinology 1993; 7:1561-1568
• Pipeleers D. G., Schuit F. C., Van Schravendijk C.F.H., Van De Winkel M. Interplay of nutrients and hormones in the regulation of glucagon release. Endocrinology 1985; 117:817-823[Abstract/Free Full Text]
• Powers A. C., Efrat S., Mojsov S., Spector D., Habener J. F., Hanahan D. Proglucagon processing similar to normal islets in pancreatic -like cell line derived from transgenic mouse tumor. Diabetes 1990; 39:406-414[Abstract]
• Shay N. F., Nick H. S., Kilberg M. S. Molecular cloning of an amino acid-regulated mRNA (amino acid starvation-induced) in rat hepatoma cells. J. Biol. Chem. 1990; 265:17844-17848[Abstract/Free Full Text]
• Thissen J.-P., Pucilowska J. B., Underwood L. E. Differential regulation of insulin-like growth factor I (IGF-I) and IGF binding protein-I messenger ribonucleic acids by amino acid availability and growth hormone in rat hepatocyte primary culture. Endocrinology 1994; 134:1570-1576[Abstract/Free Full Text]
• Unger R. H., Orci L. Glucagon and the A-cell: physiology and pathophysiology. N. Engl. J. Med. 1981; 304:1518-1524[Medline]
• Yokota T., Kanamoto R., Hayashi S. c-myc mRNA is stabilized by deprivation of amino acids in primary cultured rat hepatocytes. J. Nutr. Sci. Vitaminol. 1995; 41:455-463

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences