Differential effect of thioacetamide on hepatic methionine

Differential Effect of Thioacetamide on Hepatic Methionine Adenosyltransferase Expression in the Rat ZONG-ZHI HUANG,1 JOSE´ M. MATO,2 GARY KANEL,3

Liver-specific and non–liver-specific methionine adenosyltransferase (MAT) are products of two genes, MAT1A and MAT2A, respectively, that catalyze the formation of S-adenosylmethionine (SAM), the principal methyl donor. Mature liver expresses mainly MAT1A. We showed a switch from MAT1A to MAT2A gene expression in human liver cancer cells that may offer a growth advantage. To gain a better understanding of the chronology and significance of the change in MAT expression, we examined changes in hepatic MAT expression after acute treatment of rats with a hepatocarcinogen, thioacetamide (TAA). TAA treatment for 3 weeks did not change the MAT1A mRNA level but reduced the liver-specific MAT protein level to below 30% of control. TAA also acutely reduced the activity of liverspecific MAT when added to normal liver homogenates. In contrast, both the mRNA and protein levels of non–liverspecific MAT were induced. Because liver-specific MAT exhibits a much higher Km for methionine (mmol/L) than non–liver-specific MAT (⬇10 ␮mol/L), MAT activity was decreased at 5 mmol/L but increased at 20 ␮mol/L methionine concentration. The SAM level, SAM-to–S-adenosylhomocysteine (SAH) ratio, and DNA methylation all fell during treatment. In summary, TAA treatment induced differential changes in hepatic MAT expression. The reduction in liver-specific MAT protein level represents a novel mechanism of inactivation of liver-specific MAT. This along with induction in MAT2A contributed to a fall in the SAM-to-SAH ratio. The resulting DNA hypomethylation may be important in the process of hepatocarcinogenesis. (HEPATOLOGY 1999;29:1471-1478.) Methionine adenosyltransferase (MAT) is a critical cellular enzyme that catalyzes the formation of S-adenosylmethionine (SAM), the principal biological methyl donor and the ulti-

Abbreviations: MAT, methionine adenosyltransferase; SAM, S-adenosylmethionine; DMSO, dimethyl sulfoxide; TAA, thioacetamide; SAH, S-adenosylhomocysteine; 5-mC, 5-methylcytosine; pCNA, proliferating cell nuclear antigen. From the 1Division of Gastroenterology and Liver Diseases, USC Liver Disease Research Center, USC School of Medicine, Los Angeles, CA; 2Division of Hepatology and Gene Therapy, University of Navarra School of Medicine, Pamplona, Spain; and 3Department of Pathology, Rancho Los Amigos, USC School of Medicine, Los Angeles, CA. Received December 10, 1998; accepted February 10, 1999. Supported by NIH grant DK-51719 (S. C. Lu), Professional Staff Association grant 6-268-0-0, USC School of Medicine (S. C. Lu), Plan Nacional de I⫹D SAF 98/0132 ( J. M. Mato), Europharma (J. M. Mato), and Knoll (J. M. Mato). Address reprint request to: Shelly C. Lu, M.D., HMR Bldg., Room 415, USC School of Medicine, 2011 Zonal Ave., Los Angeles, CA 90033. E-mail: [email protected]; fax: (323) 442-3234. Copyright r 1999 by the American Association for the Study of Liver Diseases. 0270-9139/99/2905-0018$3.00/0

AND

SHELLY C. LU1

mate source of the propylamine moiety used in polyamine biosynthesis.1,2 In mammals, two different genes, MAT1A and MAT2A, encode for two homologous MAT catalytic subunits, ␣1 and ␣2.3-5 MAT1A is expressed only in liver, and it encodes the ␣1 subunit found in two native MAT isozymes, which are either a dimer (MAT III) or tetramer (MAT I) of this single subunit.5 MAT2A is widely distributed.3-5 MAT2A also predominates in the fetal liver and is progressively replaced by MAT1A during development.6,7 MAT2A encodes for a catalytic subunit (␣2) found in a native MAT isozyme (MAT II), which is associated with a catalytically inactive regulatory subunit (␤) in lymphocytes encoded by yet a third unknown gene.5,8 This regulatory ␤ subunit may be unique to lymphocytes, because it is not found in the kidney.8 Different isoforms of MAT differ in kinetic and regulatory properties and responsiveness to sulfhydryl agents and dimethyl sulfoxide (DMSO).2,9-13 The kinetic parameters varied in different studies depending on the purification procedure used and the purity of the enzyme. The Km for methionine is lowest for MAT II (⬇4-10 µmol/L), intermediate for MAT I (23 µmol/L-1 mmol/L), and highest for MAT III (215 µmol/L-7 mmol/L), with different studies reporting different absolute values.9-13 The activity of MAT is also modulated by SAM, the product of the reaction it catalyzes. SAM strongly inhibits MAT II (IC50 ⫽ 60 µmol/L), whereas it minimally inhibits MAT I (IC50 ⫽ 400 µmol/L) and stimulates MAT III (up to eightfold at 500 µmol/L SAM).10 In terms of responsiveness to sulfhydryl-modifying agents, the activity of liverspecific MAT is inhibited when certain critical cysteine residues of the enzyme are covalently modified.1,11,14-16 Modifications of these critical cysteine residues can inactivate the enzyme by direct interference with the substrate binding site(s) or by causing dissociation of the oligomers.1,11,14-16 In contrast, non–liver-specific MAT appears to be insensitive to sulfhydryl-modifying agents.9,16 Finally, DMSO activates liverspecific MAT, mainly MAT III, whereas it inhibits non–liverspecific MAT.9,17 We recently reported a switch in gene expression from MAT1A to MAT2A in various human-derived liver cancer cell lines and hepatoma tissues resected from patients.18 To see if this switch in MAT gene expression in liver cancer is pathogenetically important, we established a cell line that differs only in the type of MAT expressed by stable transfection and antisense treatment strategies. Using this model, we found that the type of MAT expression greatly influenced the rate of growth and DNA synthesis.19 Specifically, expression of MAT2A was associated with increased rates of growth and DNA synthesis, and the converse was true of MAT1A expression.19 The type of MAT enzyme expressed had a profound influence on the steady-state SAM level and methylation

1471

1472 HUANG ET AL.

status, which may have affected cell growth.19 In addition to liver cancer, MAT2A gene expression is induced during liver regeneration20,21 and during T-lymphocyte activation.22 In the case of liver regeneration after two-thirds partial hepatectomy the induction in MAT2A gene expression occurred during periods of rapid liver growth, and the expression was back to a normal low level by the fourth day.20 This contrasts with hepatocellular carcinoma, in which the MAT2A gene remained turned on.18 To gain insight into the chronology and significance of changes in MAT expression in hepatocarcinogenesis, we examined early changes in MAT expression in rats treated with a commonly used hepatocarcinogen, thioacetamide (TAA), before development of any cancerous lesion. MATERIALS AND METHODS Materials. TAA, SAM, S-adenosylhomocysteine (SAH), 5-methylcytosine (5-mC), cytosine, and HEPES were purchased from Sigma Chemical Co. (St. Louis, MO). 32P-dCTP (3,000 Ci/mmol) and methyl-L-[3H]methionine (214 mCi/mmol) were purchased from New England Nuclear (DuPont, Boston, MA). The total RNA isolation kit was obtained from Promega (Madison, WI). All other reagents were of analytical grade and were obtained from commercial sources. Animal Experiments. Male Sprague-Dawley rats (Harlan Laboratory Animals, Inc., San Diego, CA), weighing 200 to 300 g were housed with a 12:12-hour light-dark cycle and fed Purina rodent chow (Ralston Purina Co., St. Louis, MO) and water ad libitum. All animals received humane care in compliance with the National Research Council’s criteria for humane care as outlined in the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and published by the National Institutes of Health (NIH publication no. 86-23, revised 1985). Animals were treated with TAA (50 mg/kg/d, dissolved in phosphate-buffered saline) or vehicle control (phosphate-buffered saline) intraperitoneally (1 mL/kg) for 10 days, 2 weeks, or 3 weeks. All animals were killed in the morning, between 9 and 10 AM, about 20 hours after the last dose of TAA. Rat livers were removed, aliquoted into smaller pieces, and snap-frozen in liquid nitrogen for subsequent measurement of SAM, SAH, and % 5-mC levels, MAT activity, MAT1A, and MAT2A steady-state mRNA and protein levels as described below. Liver specimens were also processed for routine histological examination and Western blot analysis using commercially available antibody specific against proliferating cell nuclear antigen (pCNA) (see below). During exsanguination, blood samples were collected for serum alanine transaminase levels measured using a diagnostic kit (Sigma 505-P). MAT Activity. MAT activity was measured in liver cytosols as described.18,20 Livers were homogenized in 4 vol of 10 mmol/L Tris/HCl (pH 7.5) containing 0.3 mol/L sucrose, 0.1% ␤-mercaptoethanol, 1 mmol/L benzamidine, and 0.1 mmol/L phenylmethylsulfonyl fluoride. The homogenate was centrigued at 12,000g for 20 minutes, and the supernatant was subsequently centrifuged at 100,000g for 65 minutes to obtain liver cytosol. The reaction mixture contained 80 mmol/L Tes (pH 7.4), 50 mmol/L KCl, 40 mmol/L MgCl2, 5 mmol/L adenosine triphosphate, 10 mmol/L dithiothreitol, 0.5 mmol/L ethylenediaminetetraacetic acid, 20 µmol/L or 5 mmol/L methionine, and 0.3 to 0.5 µCi L-[methyl-3H]methionine. Liver cytosol containing 250 to 400 µg of protein as determined by the method of Bradford23 was then added to the above reaction mixture (final volume, 100 µL) for 40 minutes at 37°C. At the end of the incubation, the mixture was applied to a phosphocellulose paper disc (HA 0.45-µm, Millipore) and placed on a filtering system for washing. Adequacy of washing was first determined, and no radioactivity was recovered after washing with 5 mL of ice-cold distilled water. The disc was added to 10 mL of ScintiVerse E for scintillation counting (Beckman model LS6000TA Liquid Scintilla-

HEPATOLOGY May 1999

tion Counter). A 4°C blank was included for each condition and subtracted from 37°C values. The activity of MAT was temperaturedependent and varied linearly relative to time between 0 and 50 minutes and to the amount of protein extract between 125 and 500 µg. MAT activity is reported in units of nanomoles of SAM formed per milligram of protein per 40 minutes. Nucleic Acids Extraction. RNA was isolated from frozen liver specimens according to the method of Chomczynski and Sacchi.24 RNA concentration was determined spectrophotometrically before use, and the integrity was checked by electrophoresis with subsequent ethidium bromide staining. Northern Hybridization Analysis. Northern hybridization analysis was performed on total RNA using standard procedures as described using specific MAT1A and MAT2A cDNA probes.20,25 Both probes were labeled with [32P]dCTP using a random-primer kit (Primer-It II Kit; Stratagene, La Jolla, CA). To ensure equal loading of RNA samples and transfer in each of the lanes, before hybridization, membranes were rinsed with ethidium bromide and photographed, and the same membranes were also rehybridized with 32P-labeled 18S cDNA probe comprised of a 303–base pair fragment corresponding to nucleotides 1003 to 1305 of the published rat 18S sequence.26 Autoradiography and densitometry (Gel Documentation System, Scientific Technologies, Carlsbad, CA, and NIH Image 1.60 software program) were used to quantitate relative RNA. Results of the Northern blot analysis were normalized to 18S. DNA Methylation. Hepatic 5-mC DNA content was measured as described.20,27 DNA was extracted from livers and hydrolyzed with 88% formic acid at 180°C for 25 minutes. The DNA hydrolysate was evaporated under nitrogen and resuspended in an equal volume of 100 mmol/L HCl. The DNA bases released by hydrolysis were analyzed by high-performance liquid chromatography (Series 410 LC pump, Perkin Elmer) with a LC-90 UV detector and a LC-100 integrator (Perkin Elmer). Samples (100-200 µg) were injected into a Partisil SCX 10-µm column (25 ⫻ 0.44-cm inner diameter; Whatman Chem. Sep. Maidstone, Cleveland, OH) at room temperature, at a flow rate of 2.5 mL/min, and cytosine and 5-mC were eluted with 0.02 mol/L ammonium phosphate/HCl buffer (pH 2.3). The DNA bases were identified by spiking with standards, with detection at 280 nm at a sensitivity scale of 0.01 on the UV detector. The amounts of 5-mC and cytosine were calculated by comparing the areas of the peaks with known standards. The percent of methylation was calculated by the ratio of nanomoles of 5-mC to nanomoles of 5-mC plus nanomoles of cytosine multiplied by 100. SAM and SAH Measurement. SAM and SAH levels were measured as described with slight modification.20 Liver specimens were homogenized in 0.5 mol/L perchloric acid and centrifuged at 1,000g (Beckman GPR centrifuge) for 15 minutes, and the aqueous layer was quantitatively removed and neutralized with 3 mol/L KOH. SAM and SAH were determined in the neutralized perchloric acid extracts by high-performance liquid chromatography (Series 410 LC pump, Perkin Elmer) with a LC-90 UV detector and a LC-100 integrator (Perkin Elmer) using a Partisil SCX 10-µm column (25 ⫻ 0.44-cm inner diameter; Whatman Chem. Sep. Maidstone). SAM and SAH were eluted isocratically at 1 mL/min with 0.03 mol/L NH4H2PO4 containing 2% acetonitrile (vol/vol) and adjusted to pH 2.6 with 2 mol/L H3PO4. SAM and SAH were identified by measuring absorbance at 254 nm at a sensitivity scale of 0.01. The amount of SAM and SAH in each sample was calculated from standard curves of SAM and SAH. The identity of SAM and SAH peaks was also confirmed by spiking the sample with known standards. SAM and SAH levels are reported as nanomoles per gram of liver weight, and the ratio of SAM to SAH was determined for each sample. Western Blot Analysis. Antibodies used for Western blot analysis were rabbit polyclonal antibodies raised against purified rat liverspecific MAT7 and antibodies raised against a synthetic peptide of rat MAT2A, NGQLNGFHEAFIEEGT, which corresponds to amino acid residues 2-17 of rat MAT2A.3 This is the same peptide sequence that Horikawa et al. used to raise anti–non–liver-specific MAT antibodies.6 Both peptide synthesis and antibody generation were per-

HEPATOLOGY Vol. 29, No. 5, 1999

formed by Multiple Peptide Systems (San Diego, CA). Anti-pCNA antibodies were obtained from Zymed Laboratories Inc. (San Francisco, CA). The steady-state protein level of liver-specific MAT, non–liverspecific MAT, and pCNA was determined by Western blot analysis as described with slight modifications.28 For comparison, rat kidney and liver-specific MAT were partially purified according to published methods10 and included in the Western blots for non–liverspecific MAT. Protein concentration was determined by the method of Bradford.23 Liver homogenate (for pCNA Western blots) or cytosol (for MAT Western blots) containing 50 µg protein were solubilized in equal volumes of Sample buffer (285 mmol/L Tris [pH 6.8], 30% glycerol, 6% sodium dodecyl sulfate, 1.5% 2-mercaptoethanol, and 0.01% bromphenol blue) and subjected to sodium dodecyl sulfate–10% polyacrylamide gel electrophoresis29 and electrotransferred to NitroCellulose membranes using a Semidry Transfer Cell (Bio-Rad, Hercules, CA).30 The NitroCellulose membranes were subsequently subjected to Western blot analysis using the ECL Western blotting kit (Cat. #RPN 2209, Amersham Life Science, Arlington Heights, IL) or the Amplified Alkaline Phosphatase Immun-Blot Assay (Cat. #170-6412, Bio-Rad) according to procedures described in the kits. The first antibody was diluted 1:250 in TBST (10 mmol/L Tris [pH 8.0], 150 mmol/L NaCl, containing 0.05% Tween 20 and 0.1% dry milk). Statistical Analysis. For comparison between control and TAA treatment, ANOVA followed by Fisher’s t test was used for multiple groups (⬎2), and the two-tailed nonpaired Student t test was used for two groups. The two-tailed paired Student t test was used for comparing control and treatment using the same liver or kidney homogenates. For changes in mRNA and protein levels, ratios of MAT1A or MAT2A to 18S or Coomasie blue staining densitometric values were compared between control and TAA-treated groups by ANOVA, followed by Fisher’s t test. Significance was defined by P ⬍ .05. RESULTS Effect of TAA on MAT Gene Expression. We first examined the effect of TAA treatment on MAT gene expression. As shown in Fig. 1, TAA treatment for up to 3 weeks had no effect on the steady-state mRNA level of MAT1A, but it significantly induced the mRNA level of MAT2A (⬇450% of control). The increase in MAT2A mRNA level was observed 10 days after treatment, the earliest time point examined (data not shown). Changes in the mRNA levels are summarized in Table 1. Effect of TAA on MAT Protein Levels. We next examined the effect of TAA on the steady-state MAT protein levels. To our surprise, although the steady-state MAT1A mRNA level did not change with TAA treatment, its protein level fell progressively with duration of treatment (Fig. 2). In contrast, the protein level of non–liver-specific MAT was increased in proportion to the increase in its mRNA level (Fig. 3). Table 1 summarizes these changes. Changes in Hepatic MAT Activity After TAA Treatment. Liverspecific MAT exhibits a much higher Km for methionine (mmol/L range) than non–liver-specific MAT (⬇10 µmol/L).9-13 To see if the TAA-induced changes in the hepatic relative MAT isozyme levels are reflected in the MAT activities, we measured MAT activity using two different methionine concentrations: 5 mmol/L and 20 µmol/L. Table 2 shows that MAT activity at 5-mmol/L methionine concentration, representing liver-specific MAT, was reduced by ⬇70%, whereas MAT activity at a 20-µmol/L methionine concentration, representing largely non–liver-specific MAT, increased by ⬇70%. In another series of experiments, the effect of TAA on liver-specific MAT activity was examined by adding TAA (0.1

HUANG ET AL.

1473

FIG. 1. Effect of TAA treatment on the hepatic steady-state MAT1A and MAT2A levels. Liver RNA (20 µg each lane) samples obtained from two control and two TAA-treated rats for 2 and 3 weeks were analyzed by Northern blot hybridization with a 32P-labeled MAT1A cDNA probe as described in Materials and Methods. The same membrane was then sequentially rehybridized with 32P-labeled MAT2A and 18S cDNA probes. A representative Northern blot is shown.

mg/mL to 1 mg/mL final concentration) to liver homogenates obtained from normal rats. After incubating normal liver homogenates for 1 hour at 37°C with varying doses of TAA, MAT activity was determined under 5-mmol/L methionine concentration as described in Materials and Methods. Treatment of liver homogenates with TAA resulted in an acute reduction of MAT activity (control ⫽ 1.15 ⫾ 0.07, with 0.1 mg/mL TAA ⫽ 0.56 ⫾ 0.01 nmol of SAM formed per milligram of protein per 40 minutes; results represent mean ⫾ SE from three different normal liver homogenates; P ⬍ .001 by paired t test). Higher doses of TAA or a longer incubation period did not exert an additional reduction (not shown). This acute treatment (up to 4 hours) had no effect on the steady-state liver-specific MAT protein level (not shown). To see if TAA also inactivates non–liver-specific MAT, MAT activity was determined using 20 µmol/L methionine after treating normal kidney homogenates with TAA (0.1 mg/mL TABLE 1. Effect of TAA Treatment on Hepatic MAT1A and MAT2A Steady-State mRNA and Protein Levels % of Control mRNA Level

MAT1A MAT2A

Protein Level

2 wk

3 wk

2 wk

3 wk

116 ⫾ 17 339 ⫾ 67*

118 ⫾ 12 449 ⫾ 93*

65 ⫾ 6* 535 ⫾ 56*

29 ⫾ 4* 630 ⫾ 100*

NOTE. Results represent mean ⫾ SE from 4 to 9 animals for each condition. Rats were treated with TAA (50 mg/kg ip per day) for 2 or 3 weeks, and liver specimens were processed for determination of steady-state MAT1A and MAT2A mRNA and their protein levels as described in Materials and Methods. MAT mRNA levels were normalized to 18S, and MAT protein levels were normalized to protein loading as determined by Coomasie blue staining. Ratios of these arbitrary densitometric units were compared with that of vehicle treated controls as described in Materials and Methods. *P ⬍ .05 vs. control by ANOVA, followed by Fisher’s test.

1474 HUANG ET AL.

HEPATOLOGY May 1999

FIG. 2. Effect of TAA treatment on the hepatic steady-state liverspecific MAT protein level. Liver cytosol (50 µg/lane) obtained from three control and three TAA-treated rats for 2 and 3 weeks were analyzed by Western blot analysis using antirat liver-specific MAT antibodies as described in Materials and Methods. Equivalent protein loading was assured by Coomasie Blue staining of gels after transblotting (bottom panel). A representative Western blot analysis is shown.

final concentration) for 1 hour at 37°C. TAA exerted no inhibitory influence on MAT activity in the kidney homogenates (control ⫽ 0.01 ⫾ 0.001, with 0.1 mg/mL TAA ⫽ 0.01 ⫾ 0.0003 nmol of SAM formed per mg protein per 40 minutes; results represent mean ⫾ SE from three different normal kidney homogenates).

Effect of TAA on Liver Histology and Liver Proliferation. TAA exerted striking nuclear changes. The hepatocyte nuclei were enlarged, the vast majority of which showed intranuclear inclusions (Fig. 4). This was observed in hepatocytes of all three zones of the liver (data not shown). The bile duct epithelium and endothelium did not exhibit these changes.

FIG. 3. Effect of TAA treatment on the hepatic steady-state non–liver-specific MAT protein level. Liver cytosol (50 µg/lane) obtained from two control and two TAA-treated rats for 2 and 3 weeks were analyzed by Western blot analysis using pre- or postimmune sera obtained from rabbit immunized against a synthetic peptide of MAT2A as described in Materials and Methods. For comparison, partially purified liver-specific and kidney (non–liver-specific) MAT (25 µg/lane) were also probed by the same antibodies. Protein loading was determined by Coomasie Blue staining of gels after transblotting (bottom panel). A representative Western blot analysis is shown.


HUANG ET AL.

1475

TABLE 2. Effect of TAA Treatment on Hepatic MAT Activity MAT Activity (nmol SAM formed/ mg protein/40 min)

Condition

Methionine, 20 ␮mol/L

Methionine, 5 mmol/L

Control TAA

0.024 ⫾ 0.0004 0.040 ⫾ 0.003*

1.33 ⫾ 0.22 0.41 ⫾ 0.14*

NOTE. Results represent mean ⫾ SE from 4 to 5 animals for each condition. Rats were treated with TAA (50 mg/kg ip per day) for 3 weeks, and hepatic MAT activity was determined as described in Materials and Methods. *P ⬍ .05 vs. control by unpaired Student t test.

Spotty necrosis was present, but there was little inflammation. This is reflected in the lack of any significant change in the plasma alanine transaminase level (control ⫽ 34 ⫾ 1.4 U/mL, TAA ⫽ 38 ⫾ 6.7 U/mL; n ⫽ 4-5 for each group). A previous study showed increased liver proliferation after a

FIG. 5. Effect of TAA treatment on the hepatic steady-state pCNA level. Liver homogenates (50 µg/lane) obtained from three control and three TAA-treated rats for 3 weeks were analyzed by Western blot analysis using monoclonal anti-pCNA antibodies as described in Materials and Methods. Equivalent protein loading was assured by Coomasie Blue staining of gels after transblotting (not shown). A representative Western blot analysis is shown.

single dose of TAA (also 50 mg/kg intraperitoneally) peaking at 36 hours after treatment.31 To see if there is increased liver proliferation, Western blot analysis of pCNA was performed. Figure 5 shows that by 3 weeks after daily TAA treatment, there remained a small increase in the pCNA expression (135% ⫾ 6% of control level; results represent mean ⫾ SE from three animals each; P ⬍ .05 vs. control by unpaired t test). Changes in Hepatic SAM, SAH Levels, and DNA Methylation After TAA Treatment. We previously showed that the cellular levels

of SAM, SAH, and DNA methylation are significantly influenced by the relative expression of the MAT isozymes.19,20 Because TAA treatment resulted in an increase in the hepatic expression of non–liver-specific MAT and a decrease in liver-specific MAT, we determined how this change influenced hepatic SAM, SAH, and DNA methylation. Table 3 shows that the hepatic SAM level fell by 60% after 3 weeks of TAA treatment, while the SAH level remained unchanged. The SAM/SAH ratio also fell in parallel, and as expected, global DNA methylation was significantly lower in the treated group. Global DNA methylation was already reduced significantly after 2 weeks of TAA treatment (control ⫽ 4.23% ⫾ 0.13%, TAA-treated ⫽ 2.96% ⫾ 0.22%; results represent mean ⫾ SE from three animals for each condition; P ⬍ .05 by unpaired t test). DISCUSSION

MAT catalyzes the formation of SAM from methionine and adenosine triphosphate. In liver, this represents the major catabolic pathway of methionine, because nearly half of the TABLE 3. Effect of TAA Treatment on Hepatic SAM, SAH, and DNA Methylation Levels

FIG. 4. Effect of TAA on liver histology. (A) Control rat liver. The liver cell nuclei contain small to indistinct nucleoli. (B) Liver from a TAA-treated rat. The nuclei are enlarged compared with controls and exhibit prominent inclusions that most likely represent nucleoli. Representative histological changes are shown from four animals for each condition. (Hematoxylineosin; [A, B] ⫻561.)

Condition

SAM (nmol/g liver)

SAH (nmol/g liver)

SAM/SAH

DNA Methylation (% cytosine methylated)

Control TAA

123 ⫾ 9 48 ⫾ 7*

26 ⫾ 2 29 ⫾ 2

4.8 ⫾ 0.6 1.7 ⫾ 0.3*

4.1 ⫾ 0.5 2.4 ⫾ 0.2†

NOTE. Results represent mean ⫾ SE from 4 to 5 animals for each condition. Rats were treated with TAA (50 mg/kg ip per day) for 3 weeks, and hepatic SAM, SAH, and DNA methylation levels were determined as described in Materials and Methods. *P ⬍ .005 vs. control, †P ⬍ .05 vs. control by unpaired Student t test.

1476 HUANG ET AL.

daily intake of methionine is converted to SAM.1 Under normal conditions, SAM can be decarboxylated to be used as a precursor for the synthesis of polyamines, but most of it is used in transmethylation reactions in which methyl groups are added to compounds and SAM is converted to SAH.1 SAH is a potent competitive inhibitor of transmethylation reactions; both an increase in SAH level as well as a decrease in the SAM-to-SAH ratio are known to inhibit transmethylation reactions.1 Abnormalities in hepatic MAT have long been realized in patients with various causes of cirrhosis and in several models of liver injury.15,32-37 Inhibition of the liver-specific MAT by 50% to 60% was felt to be the explanation for hypermethioninemia and delayed plasma clearance of methionine after intravenous injection in cirrhotic patients.32,33,38,39 With the exception of hypoxia, in which a decreased MAT1A mRNA level was found,35,37 all of the other examples of decreased liver-specific MAT activity reflected inactivation of the enzyme without alteration in the protein15,16 or mRNA levels.1,34,40 The molecular mechanism of the inactivation of liver-specific MAT has been recently elucidated. Modification of critical cysteine residues inactivated the enzyme by direct interference with the substrate binding site(s) or by causing dissociation of the oligomers.1,11,14-16,37 This can occur under conditions of increased oxidative stress, nitric oxide formation, or decreased intracellular glutathione level, which are common in many of these liver injuries. Although inactivation of liver-specific MAT is well known in various causes of liver disease, induction of non–liverspecific MAT had only been described in two animal models of hepatocarcinogenesis until recently.41,42 In these earlier works, the consequences of the induction in MAT2A expression (i.e., levels of SAM, SAH, and DNA methylation) were not examined.41,42 We reported a switch in MAT gene expression from MAT1A to MAT2A in several human-derived liver cancer cell lines as well as from hepatocellular carcinoma specimens resected from patients.18 We and others also found induction of MAT2A during periods of rapid liver growth, as in liver regeneration after partial hepatectomy.20,21 As a result of the differences in the MAT isozyme regulatory properties, we speculated that changes in MAT expression would impact on the steady-state SAM, SAH, and DNA methylation levels. Thus, because SAM inhibits non–liverspecific MAT but either has no effect or stimulates liverspecific MAT,10 the SAM level in cells that express only the non–liver-specific MAT isoform should be relatively unaffected by fluctuations in methionine availability because of the negative feedback inhibition. This is in contrast to cells that express mostly liver-specific MAT in which the rate of SAM synthesis and SAM level increased with increasing methionine availability.43 Indeed, we found the type of MAT expressed to influence these variables and, more importantly, to impact on the rate of cell growth.19 Specifically, cells that express MAT1A exhibited the highest SAM level, SAM-toSAH ratios, and DNA methylation, and the lowest rates of cell growth, whereas cells that express MAT2A exhibited the exact opposite.19 Thus, relative expression of MAT isozymes in liver is likely to impact significantly on the rate of liver growth and possibly hepatocarginogenesis. To achieve a better understanding of the chronology and significance of changes in MAT expression, we studied early changes in MAT expression after treatment of animals with a commonly used hepatocarcinogen, TAA. This agent was

HEPATOLOGY May 1999

chosen because an earlier work from Okada et al. showed a fall in the DMSO-stimulatable MAT activity after only 9 days of treatment of rats with TAA.17 This suggests the possibility of a fall in liver-specific MAT activity, because DMSO stimulates only liver-specific MAT.9,17 Using the same dose and interval of TAA treatment, we found a profound inhibition of liver-specific MAT activity. This occurred as the result of a dramatic, progressive fall in the liver-specific MAT protein level despite the lack of any change in the MAT1A mRNA level. This is the first example of a fall in liver-specific MAT protein level in any known condition of liver injury. TAA is rapidly metabolized in vivo in the rat, and an obligate intermediate metabolite of TAA that binds to proteins with the formation of acetylimidolysine derivatives appears to be responsible for TAA-induced hepatotoxic effects.31 Others have also shown that a reactive electrophilic metabolite of TAA forms adducts with the lysine, and possibly cysteine, residues of proteins.44 To explain the dissociation between MAT1A mRNA and protein levels, either there is decreased protein translation or increased protein breakdown, or both. We were not able to examine the possibility of decreased protein translation, because the anti–liver-specific antibody currently available was not able to immunoprecipitate the protein after overnight radiolabeling (Mato, JM and Lu, SC, unpublished observations, October 1998). However, the fact that TAA was able to acutely reduce liver-specific MAT activity when added to liver homogenate suggests that TAA or its metabolite may have formed adducts with liver-specific MAT and led to its inactivation. We speculate that these adducts may be degraded more rapidly, thus explaining the fall in liver-specific MAT protein level. Our previous studies suggest that liver-specific MAT has a very long half-life20,37; thus, it is not surprising that no change in the liver-specific MAT protein level was observed after incubating liver homogenates with TAA for only 4 hours. Furthermore, the proteolytic enzymes required to degrade the protein are generally disrupted in the homogenate. More work will be required to elucidate the exact mechanism(s). In contrast to the decreased expression of liver-specific MAT, both the mRNA and protein levels of MAT2A were induced early on. This induction was not caused by the increased presence of inflammatory cells, because inflammation was minimal. The only known conditions in which hepatic MAT2A expression is induced are liver cancer and liver regeneration.18,20,21,42 In liver regeneration, the induction in MAT2A was transient and returned to baseline by the fourth day after hepatectomy.20 After 3 weeks of TAA treatment, there was significant hepatocyte nuclear enlargement, which has been well described,45 and a minimal increase in liver proliferation as measured by the pCNA protein level. A recent report also found no significant cell proliferation after the fourth repeated dose of TAA given in 96-hour intervals.46 Thus, proliferation cannot explain the persistent induction of MAT2A in this model, at least not at this time point. Future studies are planned to elucidate the molecular mechanism(s) of how TAA treatment induces MAT2A expression. Although the mRNA and protein levels of MAT2A increased ⬇500%, the increase in MAT activity at 20 µmol/L was very modest. This may be explained by the profound inhibition in liver-specific MAT activity, which normally contributes to the total MAT activity measured with a 20-µmol/L methionine concentration. TAA treatment resulted in a significant fall in hepatic SAM level, SAM-to-SAH ratio, and DNA hypomethylation. This


can be partly explained by the change in the relative expression of the MAT isozymes as we had observed with our HuH-7 cell model.19 Similar to our findings, Feo et al. reported a fall in liver SAM level, SAM-to-SAH ratio (SAH levels remained unchanged), and overall DNA hypomethylation and increased expression of proto-oncogenes such as c-myc, c-Ha-ras, and c-Ki-ras during the early stages of diethylnitrosamine-induced rat liver carcinogenesis.45,46 All of these changes were prevented with exogenous SAM treatment.47,48 The protective effect of SAM appears to be related to DNA methylation, because the protection was reversed by an inhibitor of DNA methylation, 5-azacytidine.49 In their work, the fall in hepatic SAM level was attributed to increased SAM utilization for polyamine synthesis, which was also induced in the early stages of hepatocarcinogenesis, although whether MAT2A was induced was not examined.47,48,50 Thus, the fall in SAM is likely the result of a combination of increased utilization and a switch in the MAT expression. The question is whether genomic hypomethylation is pathogenetically related to carcinogenesis. The relationship between DNA methylation status and cancer development is complex, because both global hypomethylation and regional hypermethylation may occur simultaneously during various stages of tumor progression.51,52 One common hypothesis is that hypomethylation of growth-promoting proto-oncogenes and/or hypermethylation of tumor suppressor genes will alter transcription factor binding and expression of these genes to promote a selective growth advantage for the initiated cell.51 Alternatively, DNA hypomethylation may promote malignant transformation by inducing regional alterations in DNA conformation and chromatin structure, rendering affected regions more accessible to DNA-damaging agents.51 In support of the latter hypothesis, Pogribny et al. showed an inverse correlation between DNA methylation status and the incidence of DNA strand breaks.51 Specifically, nuclei isolated from livers of methyl-deficient rats exhibited global DNA hypomethylation and increased sensitivity to enzyme- and oxidant-induced DNA strand breaks.51 Thus, DNA hypomethylation observed early in the course of TAA treatment is likely to play an important pathogenetic role in the subsequent development of cancer. In summary, acute treatment of rats with TAA induced differential changes in hepatic MAT expression. While it had no effect on the steady-state mRNA level of MAT1A, it induced a profound reduction in its protein level. This represents a novel mechanism of inactivation of the liverspecific MAT. On the other hand, TAA induced the expression of MAT2A. Together, these changes in MAT expression contributed to the fall in hepatic SAM level, SAM-to-SAH ratio, and DNA methylation. These changes are likely to be important in the process of hepatocarcinogenesis. REFERENCES 1. Mato JM, Alvarez L, Corrales FJ, Pajares MA. S-adenosylmethionine and the liver. In: Arias IM, Boyer JL, Fausto N, Jakoby WB, Schachter DA, Shafritz DA, eds, The Liver: Biology and Pathobiology, Third Edition, New York: Raven, 1994:461-470. 2. Surfin JR, Lombardini JB. Differences in the active-site region of tumor versus normal isozymes of mammalian ATP:L-methionine S-adenosyltransferase. Mol Pharmacol 1982;22:752-759. 3. Horikawa S, Tsukada K. Molecular cloning and developmental expression of a human kidney S-adenosylmethionine synthetase. FEBS Lett 1992;312:37-41.

HUANG ET AL.

1477

4. Horikawa S, Tsukada K. Molecular cloning and nucleotide sequence of cDNA encoding the human liver S-adenosylmethionine synthetase. Biochem Int 1991;25:81-90. 5. Kotb M, Mudd SH, Mato JM, Geller AM, Kredich NM, Chou JY, Cantoni GL. Consensus nomenclature for the mammalian methionine adenosyltransferase genes and gene products. Trends Genet 1997;13:51-52. 6. Horikawa S, Ozasa H, Ota K, Tsukada K. Immunohistochemical analysis of rat S-adenosylmethionine synthetase isozymes in developmental liver. FEBS Lett 1993;330:307-311. 7. Gil B, Casado M, Pajares M, Bosca´ L, Mato JM, Martıń-Sanz P, Alvarez L. Differential expression pattern of methionine adenosyltransferase isoenzymes during rat liver development. HEPATOLOGY 1996;24:876-881. 8. De La Rosa J, Ostrowski J, Hryneiwicz MM, Kredich NM, Kotb M, LeGros Jr HL, Valentine M, et al. Chromosomal localization and catalytic properties of the recombinant a subunit of human lymphocyte methionine adenosyltransferase. J Biol Chem 1995;270:21860-21868. 9. Okada G, Teraoka H, Tsukada K. Multiple species of mammalian S-adenosylmethionine synthetase. Partial purification and characterization. Biochemistry 1981;20:934-940. 10. Sullivan DM, Hoffman J. Fractionation and kinetic properties of rat liver and kidney methionine adenosyltransferase isozymes. Biochemistry 1983;22:1636-1641. 11. Pajares MA, Duran C, Corrales F, Pliego M, Mato JM. Modulation of rat liver S-adnosylmethionine synthetase activity by glutathione. J Biol Chem 1992;267:17598-17605. 12. Liau MC, Chang CF, Belanger L, Grenier A. Correlation of isozyme patterns of S-adenosylmethionine synthetase with fetal stages and pathologic states of the liver. Cancer Res 1979;39:162-169. 13. Cabrero C, Puerta J, Alemany S. Purification and comparison of two forms of S-adenoysl-L-methionine synthetase from rat liver. Eur J Biochem 1987;170:299-304. 14. Mingorance J, Alvarez L, Sańchez-Gońgora E, Mato JM, Pajares MA. Site-directed mutagenesis of rat liver S-adenosylmethionine synthetase. Biochem J 1996;315:761-766. 15. Avila MA, Mingorance J, Martıńez-Chantar ML, Casado M, Martıń-Sanz P, Basca´ L, Mato JM. Regulation of rat liver S-adenosylmethionine synthetase during septic shock: role of nitric oxide. HEPATOLOGY 1997;25:391-396. 16. Sańchez-Gońgora E, Ruiz F, Mingorance J, An W, Corrales FJ, Mato JM. Interaction of liver methionine adenosyltransferase with hydroxyl radical. FASEB J 1997;11:1013-1019. 17. Okada G, Sawai Y, Teraoka H, Tsukada K. Differential effects of dimethylsulfoxide on S-adenosylmethionine synthetase from rat liver and hepatoma. FEBS Lett 1979;106:25-28. 18. Cai J, Sun WM, Hwang JJ, Stain S, Lu SC. Changes in S-adenosylmethionine synthetase in human liver cancer: molecular characterization and significance. HEPATOLOGY 1996;24:1090-1097. 19. Cai J, Mao Z, Hwang JJ, Lu SC. Differential expression of methionine adenosyltransferase genes influences the rate of growth of human hepatocellular carcinoma cells. Cancer Res 1998;58:1444-1450. 20. Huang Z, Mao Z, Cai J, Lu SC. Changes in methionine adenosyltransferase during liver regeneration in the rat. Am J Physiol 1998;38:G14-G21. 21. Horikawa S, Ozasa H, Ito K, Katsuyama I, Tsukada K, Sugiyama T. Expression of S-adenosylmethionine synthetase isozyme genes in regenerating rat liver after partial hepatectomy. Biochem Mol Biol Int 1996;40:807-814. 22. Tobenã R, Horikawa S, Calvo V, Alemany S. Interleukin-2 induces ␥-S-adenosyl-L-methionine synthetase gene expression during Tlymphocyte activation. Biochem J 1996;319:929-933. 23. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the priniciple of protein-dye binding. Anal Biochem 1976;72:248-254. 24. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987;162:156-159. 25. Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning: A Laboratory Manual, 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, 1989:13.7-13.8, 16.3-16.4. 26. Torczynski RM, Bollon AP, Fuke M. The complete nucleotide sequence of the rat 18S ribosomal RNA gene and comparison with the respective yeast and frog genes. Nucleic Acids Res 1983;11:4879-4890. 27. Diala ES, Cheah MSC, Rowitch D, Hoffman RM. Extent of DNA methylation in human tumor cells. J Natl Cancer Inst 1983;71:755-764. 28. Cai J, Sun WM, Lu SC. Hormonal and cell density regulation of hepatic ␥-glutamylcysteine synthetase gene expression. Mol Pharmacol 1995;48: 212-218.

1478 HUANG ET AL. 29. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bactriophage T4. Nature 1970;227:680-685. 30. Towbin H, Staehein T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitro-cellulose sheets: procedure and some applications. Proc Natl Acad Sci U S A 1989;76:4350-4354. 31. Mangipudy RS, Chanda S, Mehendale HM. Tissue repair response as a function of dose in thioacetamide hepatotoxicity. Environ Health Perspect 1995;103:260-267. 32. Cabrero C, Duce AM, Ortiz P, Alemany A, Mato JM. Specific loss of the high-molecular weight form of S-adenosyl-L-methionine synthetase in human liver cirrhosis. HEPATOLOGY 1988;8:1530-1534. 33. Duce AM, Ortiz P, Cabrero C, Mato JM. S-Adenosyl-L-methionine synthetase and phospholipid methyltransferase are inhibited in human cirrhosis. HEPATOLOGY 1988;8:65-68. 34. Corrales F, Gimeńez A, Alvarez L, Caballerıá J, Pajares MA, Andreau H, Pare´s A, et al. S-Adenosylmethionine treatment prevents carbon tetrachloride–induced S-adenosylmethionine synthetase inactivation and attenuates liver injury. HEPATOLOGY 1992;16:1022-1027. 35. Chawla RK, Jones DP. Abnormal metabolism of S-adenosyl-L-methionine in hypoxia rat liver. Similarities to its abnormal metabolism in alcoholic cirrhosis. Biochim Biophys Acta 1994;1199:45-51. 36. Shimizu K, Abe M, Yokoyama S, Takahashi H, Sawada N, Mori M, Tsukada K. Decreased activities of S-adenosylmethionine synthetase isozymes in hereditary hepatitis in Long-Evans rats. Life Sci 1990;46: 1837-1842. 37. Avila MA, Carretero MV, Rodriguez EN, Mato JM. Regulation by hypoxia of methionine adenosyltransferase activity and gene expression in rat hepatocytes. Gastroenterology 1998;114:364-371. 38. Kinsell LW, Harper HA, Barton HC, Michaels GD, Weiss HA. Rate of disappearance from plasma of intravenously administered methionine in patients with liver damage. Science 1947;106:589-594. 39. Horowitz JH, Rypins EB, Henderson JM, Heymsfield SB, Moffitt SD, Bain RP, Chawla RK, et al. Evidence for impairment of transsulfuration pathway in cirrhosis. Gastroenterology 1981;81:668-675. 40. Alvarez L, Corrales F, Martin-Duce A, Mato JM. Characterization of a full-length cDNA encoding human liver S-adenosylmethionine synthetase: tissue-specific gene expression and mRNA levels in hepatopathies. Biochem J 1993;293:481-486. 41. Tsukada K, Okada G. S-adenosylmethionine synthetase isozyme pat-

HEPATOLOGY May 1999

42.

43. 44. 45. 46. 47.

48.

49.

50.

51.

52.

terns from rat hepatoma induced by N-2-fluorenlacetamide. Biochem Biophys Res Commun 1980;94:1078-1082. Horikawa S, Kobayashi Y, Sugiyama T, Terashima H, Wada K, Tsukada K. Expression of non-hepatic-type S-adenosylmethionine synthetase isozyme in rat hepatomas induced by 38-methyl-4-dimethylaminoazobenzene. FEBS Lett 1993;334:69-71. Finkelstein JD. Methionine metabolism in mammals. J Nutr Biochem 1990;1:228-237. Dyroff MC, Neal RA. Identification of the major protein adduct formed in rat liver after thioacetamide administration. Cancer Res 1981;41:3430-3435. Olason D, Smuckler EA. Changes in hepatic nuclei induced by acetamide and thioacetamide. Arch Pathol Lab Med 1976;100:415-418. Mangipudy RS, Mehendale HM. Temporal changes in tissue repair upon repeated exposure to thioacetamide. Toxicol Appl Pharmacol 1998;149: 254-257. Pascale RM, Simile MM, Satta G, Seddaiu MA, Daino L, Pinna G, Vinci MA, et al. Comparative effects of L-methionine, S-adenosyl-L-methionine and 58-methylthioadenosine on the growth of preneoplastic lesions and DNA methylation in rat liver during the early stages of hepatocarcinogenesis. Anticancer Res 1991;11:1617-1624. Simile MM, Pascale R, De Miglio MR, Nufris A, Daino L, Seddaiu MA, Gaspa L, et al. Correlation between S-adenosyl-L-methionine content and production of c-myc, c-Ha-ras, and c-Ki-ras mRNA transcripts in the early stages of rat liver carcinogenesis. Cancer Lett 1994;79:9-16. Pascale R, Simile MM, Ruggiu ME, Seddaiu MA, Satta G, Sequenza MJ, Daino L, et al. Reversal by 5-azacytidine of the S-adenosyl-L-methionineinduced inhibition of the development of putative preneoplastic foci in rat liver carcinogenesis. Cancer Lett 1991;56:259-265. Garcea R, Pascale R, Daino L, Frassetto S, Cozzolina P, Ruggiu ME, Vannini MG, et al. Variations of ornithine decarboxylase activity and S-adenosyl-L-methionine and 58-methylthioadenosine contents during the development of diethylnitrosamine-induced liver hyperplastic nodules and hepatocellular carcinoma. Carcinogenesis 1987;8:653-658. Pogribny IP, Basnakian AG, Miller BJ, Lopatina NG, Poirier LA, James SJ. Breaks in genomic DNA and within the p53 gene are associated with hypomethylation in livers of folate/methyl-deficient rats. Cancer Res 1995;55:1894-1901. Laird PW, Jaenisch R. DNA methylation and cancer. Hum Mol Genet 1994;3:1487-1495.