1198
Chem. Res. Toxicol. 1997, 10, 1198-1204
Dietary Glycine Prevents Increases in Hepatocyte Proliferation Caused by the Peroxisome Proliferator WY-14,643 Michelle L. Rose,† Dori Germolec,‡ Gavin E. Arteel,† Robert Schoonhoven,§ and Ronald G. Thurman*,† Laboratory of Hepatobiology and Toxicology, Curriculum in Toxicology, Departments of Pharmacology and Environmental Sciences, University of North Carolina, Chapel Hill, North Carolina, and National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina Received May 9, 1997X
Peroxisome proliferators are a group of nongenotoxic carcinogens which include a number of hypolipidemic drugs, solvents, and industrial plasticizers. Although the mechanism by which they cause cancer remains unknown, one likely possibility is that they act as tumor promoters by increasing cell proliferation. Hepatic Kupffer cells represent a rich source of mitogenic cytokines (e.g., tumor necrosis factor R, TNFR) and are stimulated by peroxisome proliferators. Since glycine prevents activation of Kupffer cells, these experiments were designed to test the hypothesis that a diet containing glycine could block the mitogenic effect of the peroxisome proliferator [[4-chloro-6-(2,3-xylidino)pyrimidinyl]thio]acetic acid (WY-14,643). The effects of a glycine-enriched diet on WY-14,643-induced increases in cell proliferation after a single dose or after feeding WY-14,643 in the diet for 3 weeks were assessed. As expected, 24 h after a single dose of WY-14,643, rates of cell proliferation increased from basal values of 0.7 ( 0.3% to 5.1 ( 0.5%. Glycine largely prevented the increase caused by WY-14,643 with proliferation only reaching 1.9 ( 0.4% (p < 0.05). Acyl CoA oxidase increased from 1.4 ( 0.1 to 3.5 ( 0.6 nmol of H2O2 min-1 (mg of protein)-1 (p < 0.05) indicating that peroxisome-specific enzyme activity was induced about 2-fold in livers of WY-14,643-treated rats after 24 h. Unlike cell proliferation, however, acyl CoA oxidase was not affected by dietary glycine, consistent with the hypothesis that cell proliferation and peroxisome proliferation occur via different mechanisms. After 3 weeks, dietary glycine reduced basal rates of cell proliferation by about 50% and completely prevented the sustained 5-fold increase in cell proliferation caused by feeding WY-14,643. Moreover, the 3-fold increase in TNFR mRNA caused by WY-14,643 was blocked completely by the glycine-enriched diet. Similarly, immunohistochemical staining for TNFR was increased 6-fold by WY-14,643, an increase which was prevented by dietary glycine. However, the 6-fold increase in acyl CoA oxidase activity was unaffected by glycine under similar conditions demonstrating that a diet containing 5% glycine prevents increased hepatocyte proliferation caused by a potent peroxisome proliferator without affecting induction of peroxisomes. These data demonstrate that a glycine-enriched diet prevents stimulated cell proliferation most likely by inhibiting TNFR production and raise the possibility that dietary glycine will be effective in preventing cancer caused by nongenotoxic carcinogens such as WY-14,643.
Introduction WY-14,6431
is a member of the class of compounds known collectively as peroxisome proliferators, which increase both the number and size of peroxisomes as well as associated peroxisomal enzymes in hepatocytes (1). Chronic exposure to WY-14,643 leads to hepatocellular adenoma and carcinoma in rodents, with 100% of animals exhibiting tumors in 1 year making this one of the most active carcinogens in this class of chemicals (2). The risk * Send all correspondence to Dr. Ronald G. Thurman, Laboratory of Hepatobiology and Toxicology, Department of Pharmacology, CB# 7365, FLOB, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7365. Phone: (919) 966-4745. Fax: (919) 966-1893. E-mail:
[email protected]. † Department of Pharmacology. ‡ National Institute of Environmental Health Sciences. § Department of Environmental Sciences. X Abstract published in Advance ACS Abstracts, September 1, 1997. 1 Abbreviations: [[4-chloro-6-(2,3-xylidino)pyrimidinyl]thio]acetic acid, WY-14,643; tumor necrosis factor R, TNFR; diethylhexyl phthalate, DEHP; reverse transcription-polymerase chain reaction, RT-PCR; 5-bromo-2′-deoxyuridine, BrdU; 3,3′-diaminobenzidine, DAB; trichloroacetic acid, TCA; glyceraldehyde-3-phosphate dehydrogenase, G3PDH; lipopolysaccharide, LPS.
S0893-228x(97)00079-9 CCC: $14.00
of human exposure to peroxisome proliferators including hypolipidemic drugs such as clofibrate is unclear. A World Health Organization study found a higher mortality rate in high-cholesterol patients treated with clofibrate than those that were untreated (3). While the clofibrate-treated group had more deaths from cancer, one disease alone (e.g., cancer, ischemic heart disease, stroke) did not account for the overall excess death in clofibrate patients (3). In fact, cancer risk to humans remains controversial since the mechanism by which peroxisome proliferators cause cancer remains unknown (4, 5); however, several studies suggest that they act via nongenotoxic mechanisms involving increased cell replication (1, 2, 6, 7). Increased hepatocyte proliferation is likely a key factor in the genesis of liver cancer by these chemicals. For example, their potency as carcinogens has been associated with their ability to sustain cell proliferation; WY14,643 increased cell proliferation for as long as the compound was administered, while diethylhexyl phthalate (DEHP), a much less potent carcinogen, did not © 1997 American Chemical Society
Glycine Inhibits Cell Proliferation
sustain proliferation even at doses 12-fold higher (2). Since WY-14,643 and nafenopin cause a greater number of preneoplastic lesions in livers from older than younger rats, elevated rates of hepatocyte replication may be important in the promotion of previously initiated cells, which are more numerous in older rats (8, 9). Thus, it has been proposed that peroxisome proliferators act as tumor promoters by stimulating cell proliferation of spontaneously initiated cells (8, 9). Unfortunately, the mechanism by which these chemicals stimulate mitogenesis is unknown. However, Kupffer cells, the resident hepatic macrophages, are a rich source of mitogenic stimuli (e.g., TNFR, hepatocyte growth factor, prostaglandin E2) in the liver (10, 11). It was hypothesized, therefore, that peroxisome proliferators activate Kupffer cells to release cytokines which stimulate cell replication in the liver. Indeed, in vivo treatment with WY-14,643 and nafenopin activated Kupffer cell phagocytosis, and inactivation of Kupffer cells with methyl palmitate, a nonhydrolyzable fatty acid, completely prevented the increase in cell proliferation caused by WY-14,643 (12, 13). Moreover, antibodies to tumor necrosis factor R (TNFR) inhibited WY-14,643-induced cell proliferation demonstrating that TNFR was responsible for stimulating hepatocyte replication. TNFR was localized immunohistochemically in sinusoidal lining cells (most likely Kupffer cells) 24 h following treatment with WY-14,643 (14). Taken together, these data are consistent with the hypothesis that TNFR of Kupffer cell origin is involved in mechanisms of WY-14,643-stimulated hepatocyte replication. Since inactivation of Kupffer cells with methyl palmitate requires daily iv injections, long-term studies are not practical. Therefore, modulation of Kupffer cell activity with dietary agents that limit or prevent cytokine production in response to mitogenic stimuli would provide an ideal approach to studying the role of Kupffer cellderived TNFR in sustained increases in hepatocyte proliferation caused by WY-14,643. An increase in TNFR in the serum of rats treated with endotoxin was both delayed and blunted in rats fed dietary glycine (15). Since TNFR is involved in the stimulation of cell proliferation by WY-14,643 and glycine diminishes TNFR production, these experiments were designed to test the hypothesis that a glycine-enriched diet would prevent the increase in hepatocyte replication caused by the liver carcinogen WY-14,643. Preliminary accounts of this work have appeared elsewhere (16).
Materials and Methods Animals and Treatment. Male Sprague-Dawley rats were housed 2-3/cage and given semisynthetic modified AIN 76 diets (Table 1) and water ad libitum. Two different experimental protocols were used for this study. First, a single dose of WY14,643 was given to study the initial burst in cell replication; second, feeding WY-14,643 for 3 weeks was used to evaluate effects on sustained cell proliferation. For the single-dose study, animals were fed either control or 5% glycine diet for 3 days. On the third day, rats were treated with either 100 mg/kg WY14,643 or an equal volume of olive oil control vehicle (ig). Twenty-four hours later all rats were given 100 mg/kg 5-bromo2′-deoxyuridine (BrdU; Sigma) ip to label proliferating hepatocytes and sacrificed 1 h later. For the 3 week feeding study, rats were fed either control or 5% glycine diet for 3 days; then one-half of the rats in each group were switched to a diet containing 0.1% WY-14,643 for 3 weeks. Four groups (control, glycine, WY-14,643, and WY-14,643 + glycine) were then compared. During the last 3 days of the experiment, all rats were given drinking water containing BrdU (80 mg/100 mL) to
Chem. Res. Toxicol., Vol. 10, No. 10, 1997 1199 Table 1. Dietary Content by Percenta control glycine WY-14,643 casein sucrose corn oil R-cellulose mineral mix vitamin mix D,L-methionine choline bitartrate corn starch glycine WY-14,643
20 50 5 5 3.5 1 0.3 0.2 15 0 0
15 50 5 5 3.5 1 0.3 0.2 15 5 0
20 49.9 5 5 3.5 1 0.3 0.2 15. 0 0.1
WY-14,643 + glycine 15 49.9 5 5 3.5 1 0.3 0.2 15 5 0.1
a The contribution of each dietary component is reported as a percent (w/w).
label proliferating hepatocytes. Rats were sacrificed under pentobarbital anesthesia between 5:00 and 7:00 p.m. since apoptosis is highest at the end of the rats’ light cycle when they have not eaten for nearly 12 h (20). Cell Proliferation. Livers were perfused with KrebsHenseleit buffer (pH 7.6) to remove blood and fixed with 4% paraformaldehyde. A section of duodenum, a tissue which proliferates rapidly, was collected as a positive control for BrdU incorporation. Tissue sections were deparaffinized, rehydrated, and hydrolyzed in 4 N HCl for 20 min at 37 °C. Endogenous peroxidase was quenched with 0.03% hydrogen peroxide containing sodium azide (DAKO Envision System; peroxidase). A primary monoclonal antibody to BrdU was diluted 1:200 (DAKO; clone Bu20a) and allowed to incubate at room temperature for 10 min. A peroxidase-labeled polymer conjugated to goat antirabbit and goat anti-mouse was incubated for 10 min at room temperature followed by addition of 3,3′-diaminobenzidine (DAB) in a buffered solution containing hydrogen peroxide for 8 min (DAKO Envision System; peroxidase). Sections were rinsed twice with phosphate-buffered saline containing 1% Tween 20 after each incubation. Slides were counterstained with hematoxylin, and proliferating cells were identified from brown staining of precipitated polymerized DAB. Cell proliferation was quantitated by determining the percentage of BrdUpositive hepatocytes in 10 random high-power fields/slide (1000 hepatocytes/slide). Acyl CoA Oxidase Activity. Acyl CoA oxidase is localized in peroxisomes, and its activity is an accepted measure of induction of peroxisomes (17). It was measured as formaldehyde formed from hydrogen peroxide generated by peroxisomal β-oxidation. Liver samples were homogenized in 10 volumes of 0.25 M sucrose buffer. A reaction mixture (pH 8.3) containing 22 mg of palmitate, 400 µL of methanol, 8 mg of CoA, 138 mg of ATP, 41 mg of MgCl2, 10 µL of Triton X-100, 13 mg of NAD+, 90 mg of fatty acid free bovine serum albumin, and 403 mg of niacinamide (per 100 mL of buffer) was warmed to 37 °C. The reaction was started by adding 200 µL of homogenate to 1.4 mL of reaction buffer and was terminated after 10 min with 40% trichloroacetic acid (TCA). TCA was added before homogenate to the blanks. The solution was centrifuged to pellet protein, and 0.5 mL of supernatant was added to 0.2 mL of Nash Reagent to measure formaldehyde (18). After 60 min of incubation at 37 °C, absorbance was read at 405 nm. Protein concentration in the homogenate was determined by the method of Lowry (19). Apoptotic Index. Apoptotic bodies were identified based on morphological features of hematoxylin and eosin-stained liver sections as described by Schulte-Hermann since the popular method of detecting fragmented DNA in situ (TUNEL) identifies both necrotic and apoptotic cells (20, 21). The features used to determine apoptotic cells were condensation of cytoplasm, accumulation of condensed chromatin at the nuclear membrane, and fragmentation of the nucleus and cell, giving rise to apoptotic bodies. The apoptotic rate is expressed as the percentage of apoptotic cells counted in 10 random high-power fields (4000 hepatocytes counted per slide). Reverse Transcription-Polymerase Chain Reaction (RTPCR) Amplification of TNFr mRNA. Approximately 100 mg samples of liver tissue were homogenized in 2 mL of RNAzol B
1200 Chem. Res. Toxicol., Vol. 10, No. 10, 1997 solution (Biotecx Laboratories, Inc., Houston, TX), and total cellular RNA was extracted according to the manufacturer’s instructions. RNA was dissolved in Tris-EDTA buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0). For the synthesis of cDNA, 3 µg total RNA from each sample was resuspended in a 20 µL final volume of the reaction buffer consisting of 25 mM TrisHCl, pH 8.3, 37.5 mM KCl, 10 mM dithiothreitol, 1.5 mM MgCl2, 10 mM of each dNTP (Perkin Elmer, Cetus, Norwalk, CT), and 0.5 ng of oligo d(T) 12-18 primer (GibcoBRL, Gaithersburg, MD). After the reaction mixture reached 42 °C, 400 U of SuperScript reverse transcriptase (GibcoBRL) was added to each tube, the mixture was incubated for 35 min at 42 °C, and the reaction was stopped by denaturing the enzyme at 99 °C for 5 min. The reaction mixture was diluted with distilled water to a volume of 50 µL. PCR primers for rat TNFR and glyceraldehyde-3-phosphate dehydrogenase (G3PDH) were purchased from Clontech Laboratories, Inc. (Palo Alto, CA) and contained the following sequences: sense, 5′-TACTGAACTTCGGGGTGATTGGTCC-3′, and antisense, 5′-CAGCCTTGTCCCTTGAAGAGAACC3′ for TNFR; and sense, 5′-TGAAGGTCGGTGTCAACGGATTTGGC-3′, and antisense, 5′-CATGTAGGCATGAGGTCCACCAC3′ for G3PDH. The amplified products contained 692 base pairs for TNFR and 983 for G3PDH. Five-microliter aliquots of the synthesized cDNA (corresponding to 10 ng of mRNA) were added to 45 µL of mixture containing 5 µL of 10× PCR buffer, 1 µL of deoxynucleotides (1 mM each), 0.5 µL of sense and antisense primers (0.15 mM), and 0.25 µL of DNA polymerase (GeneAmp PCR kit, Perkin Elmer Cetus). The reaction mixture was covered with a wax tablet (Perkin Elmer Cetus), and amplification was initiated by 1 min of denaturation at 95 °C for 1 cycle, then by 25, 30, or 35 cycles at 94 °C for 15 s, 55 °C for 30 s, and 72 °C for 30 s using a DNA thermal cycler (Perkin Elmer Cetus). The samples were incubated for 7 min at 72 °C after the last cycle of amplification. When necessary, the concentrations of cDNA were normalized to G3PDH and the PCR process was repeated. For each set of primers, dilutions of cDNA were amplified for 20, 23, 25, 28, 30, 33, and 35 cycles to define optimal conditions for linearity and to permit semiquantitative analysis of signal strength. When appropriate, the specificity of the PCR bands was confirmed by restriction site analysis of the amplified cDNA which generated restriction fragments of the expected size (not shown). The amplified PCR products were separated by electrophoresis through 2.0% agarose gel (UltraPure, GibcoBRL) at 60 V for 90 min. Bands of cDNA were visualized by ultraviolet illumination after staining with 0.5 µg/mL ethidium bromide. Gels were photographed with type 55 positive/negative film (Polaroid, Cambridge, MA), and films were analyzed using the Eagle Eye II Image Analysis System (Stratagene, La Jolla, CA) and NIH Image 1.54 software. TNFr Immunohistochemistry and Image Analysis. Samples of liver from each rat were quick-frozen and 6 µm sections prepared for immunohistochemistry. The sections were fixed in acetone for 10 min, allowed to air-dry, and blocked with normal goat serum (1:67 dilution in undiluted automation buffer; Biomedia Corp., Foster City, CA; and 1% nonfat dry milk) for 30 min. All further incubations were made in a solution of automation buffer containing 1% bovine serum albumin. Endogenous peroxidase activity was quenched by incubation for 10 min in 3% hydrogen peroxide. Slides were incubated for 1 h at 4 °C with a 1:75 dilution of polyclonal goat anti-mouse TNFR antibody (R&D Systems, Minneapolis, MN) or normal goat serum as a control, washed in automation buffer, and then incubated for 1 h with a 1:200 dilution of rabbit anti-goat IgG biotinylated antibody (Vector Lab Inc., Burlingame, CA). After washing, the antibodies were labeled using the Vectastain Elite kit (Vector Laboratories) and localized with the peroxidase substrate DAB enhanced with NiCl2. Finally, slides were counterstained with modified Harris’ hematoxylin and dehydrated for mounting. A Universal Imaging Corp. Image-1/AT image acquisition and analysis system (Chester, PA) incorporating an Axioskop 50 microscope (Carl Zeiss, Inc., Thornwood, NY) was used to photograph and analyze the immunostained tissue sections at 100× magnification. Color detection ranges were set for the red-brown color of the DAB chromogen based on an
Rose et al.
Figure 1. Hepatocyte proliferation 24 h after treatment with WY-14,643. Proliferating cells were identified and quantitated as described in Materials and Methods. Proliferation rates are reported as means ( SEM for rats treated as described in Materials and Methods for the following groups: control (CON), glycine diet (GLY), WY-14,643 (WY), glycine diet with WY-14,643 treatment (WY + GLY). Asterisks denote statistical differences from the control, glycine, and WY + glycine groups (p < 0.05; n ) 4 for all groups). intensely labeled Kupffer cell. The extent of TNFR expression in the liver is defined as the percent of the field area within the default color range determined by the software. Data from each tissue section (5 random fields/section) were pooled to determine means. Statistics. Results are reported as means ( SEM with n ) 4 or 5 in each group. Treatment groups were compared using one-way ANOVA and Student-Newman-Keuls post hoc tests for all possible comparisons between groups; p < 0.05 was selected to determine statistical differences between groups.
Results Single-Dose Study. WY-14,643 (100 mg/kg) increased cell proliferation almost 8-fold 24 h after administration from 0.7 ( 0.3% to 5.1 ( 0.3% replicating hepatocytes (p < 0.05 compared to control, Figure 1). While feeding a diet containing 5% glycine for 3 days did not affect basal rates of hepatocyte proliferation under these conditions (Gly, Figure 1), it largely prevented increases in cell replication due to WY-14,643, with values only reaching 1.9 ( 0.4% (WY + Gly, Figure 1). This value was significantly less than the 7-fold increase in hepatocyte replication characteristically caused by WY-14,643 and did not differ from controls. A defining characteristic of all peroxisome proliferators is an induction of peroxisome-specific enzymes. Twentyfour hours after treatment with WY-14,643, acyl CoA oxidase was increased 2.5-fold (p < 0.05, Figure 2A), an increase which is known to plateau in 1-2 weeks (2). The glycine-enriched diet had no effect on basal acyl CoA oxidase activity and surprisingly did not prevent the increase due to WY-14,643 (Figure 2A). Peroxisomespecific enzyme activity after 3 weeks of treatment with WY-14,643 was even more extensive (Figure 2B). WY-
Glycine Inhibits Cell Proliferation
Chem. Res. Toxicol., Vol. 10, No. 10, 1997 1201
Figure 3. Effect of WY-14,643 on body weight. Body weights were measured every 5 days and on the first and final days of a 3 week feeding protocol. Asterisks denote statistical differences from control and glycine (p < 0.05, n ) 5), and # denotes statistical differences from control, glycine, and WY groups (p < 0.05, n ) 5). Table 2. Liver and Body Weights after Three Weeks of WY-14,643a
Figure 2. Acyl CoA oxidase activity after treatment with WY14,643. (A) Acyl CoA oxidase activity was determined 24 h after treatment with a single dose of WY-14,643 as described in Materials and Methods and is reported as means ( SEM as described in Figure 1 (n ) 4 for all groups). (B) Acyl CoA oxidase activity after 3 weeks of treatment with WY-14,643. Asterisks denote statistical differences from control and glycine groups (p < 0.05, n ) 5 for all groups).
14,643 caused a 6-fold increase in acyl CoA oxidase activity, which also was not prevented by glycine. Marsman et al. hypothesized that peroxisome proliferation and cell proliferation are not tightly linked events (2), which is supported by the interesting finding that glycine prevents WY-14,643-stimulated increases in cell proliferation without affecting peroxisome proliferation. Three Week WY-14,643 Feeding Study. To determine if a glycine-enriched diet could prevent the sustained increase in cell proliferation caused by chronic feeding of WY-14,643, rats were fed 0.1% WY-14,643 with or without 5% glycine for 3 weeks. Diets (see Table 1 for composition) were provided ad libitum, and consumption was monitored daily. Importantly, average diet consumption did not differ between the treatment groups studied (data not shown); however, weight gain varied considerably (Figure 3). Body weights of control and glycine-fed rats did not differ at any time during the feeding period. In contrast, the WY-14,643 diet caused an average weight gain of only about 60% of control values in agreement with results from a previous chronic feeding study (2). Rats fed WY-14,643 or WY-14,643 + glycine were significantly smaller than controls by day 15. After 20 days, the WY-14,643 + glycine rats weighed about 35% less than controls and the WY-14,643 group about 24% less than controls.
diet
body weight (g)
liver weight (g)
relative liver weight (liver/ body weight)
control glycine WY-14,643 WY-14,643 + glycine
365 ( 8 358 ( 16 284 ( 12* 232 ( 20*,#
18.5 ( 0.9 18.1 ( 1.4 24.5 ( 0.8* 18.3 ( 1.0#
5.0 ( 0.3 5.0 ( 0.3 8.7 ( 0.2* 7.9 ( 0.5*
a Weights are reported in grams (g) for animals fed semisynthetic diets for 3 weeks. Asterisks (*) denote statistical differences from control and glycine groups; number sign (#) denotes statistical difference from WY-14,643 (p < 0.05, n ) 5 for all groups).
The relative liver weight for rats fed the control semisynthetic diet was 5.0 ( 0.3% and was not affected by glycine (Table 2). The WY-14,643 diet caused a significant increase in relative liver weight of about 70% as expected (8.7 ( 0.2%, p < 0.05 compared to control), a phenomenon not prevented by glycine (7.9 ( 0.5%). However, the addition of 5% glycine to the WY-14,643 diet prevented the increase in absolute liver size caused by WY-14,643 (p
