- Email: [email protected]
Chemoprevention by amiloride against experimental hepatocarcinogenesis induced by N-nitrosomorpholine in Sprague-Dawley rats
I_-
Cancer Letters I 19 ( 1997) IO9- I 13
Chemoprevention by amiloride against experimental hepatocarcinogenesis induced by Wnitrosomorpholine in Sprague-Dawley rats Masaharu Tatsuta*, Hiroyasu Iishi, Miyako Baba, Hiroyuki Yano, Noriko Sakai, Hiroyuki Uehara, Akihiko Nakaizumi Deppartment oj Gastrointestinal
Oncology, Osaka Medical Center Nakumichi I-chome, Higashinari-ku,
.for Cancer
and Cardiovascular Osaka 537. .Iupan
Dwow~,
i- :
Received 12 March lYY7; received in revised form 25 April 1997: accepted 25 April 199)
--.-- ..__... -l..-..-- ..-Abstract
The effects of prolonged administration of the diuretic amiloride on hepatocarcinogenesis induced by ~-nitrosomr,rpholine
(NNM) and the labeling index of the liver were investigated in male Sprague-Dawley rats. Rats were given &inking water containing NNM for 8 weeks and received S.C.injections of 2.5 mg/kg or 5.0 mg/kg body weight of amiloride every other day until the end of the experiment at week 16. Preneoplastic and neoplastic lesions staining positively for glutathione-Stransferase, placental type (GST-P) were examined immunohistochemically. In week 16, administration of amiloride at 2.5 mg/kg body weight significantly reduced the size (as mean area) of GST-P-positive lesions, and amiloride at 5.0 mglkg body weight significantly reduced the number (as no./cm*) and sizes (as mean area and percent of parenchyma) of GST-P-positive lesions. Amiloride at both dosagesalso significantly decreasedthe labeling index of enzyme-alteredlesions and amiloride at
higher dosage significantly decreasedthe labeling index of adjacent hepatocytes.These findings indicate that amiioride inhibits hepatocarcinogenesis and suggest that this effect may be closely related to amiloride’s inhibition of celi proliferation in
enzyme-alteredlesions and/or adjacent liver. 0 1997 Elsevier Science Ireland Ltd.
Keywords:
Amiloride: Hepatocarcinogenesis;Chemoprevention;N-Nitrosomorpholine
1. Introduction The diuretic drug the Na’/H+ antiport inhibit cell growth [ 3,4]. Previously we
amiloride is a potent inhibitor of [ 1.21 and has been reported to in some extrahepatic tumors found that prolonged administra-
*Corresponding author. Tel.: +81 6 9721181; fax: +81 6 98 14067.
03043835/97/$17.00 PII
SO304-3805(97)00262-O
___-- .. -._
_.__ - ..-_-.
tion of amiloride inhibits the development of cancers in the stomach [5], colon [6], and pancreas [7] in rats. Amiloride strongly inhibits [‘Hlthymidine incorporation in a well-differentiated hepat.oma cell line HepGl [S]. These findings suggest that amiloride may inhibit hepatocarcinogenesis. To investigate this possibility, in the present work we examined amiloride‘s effects on hepatocarcinogenesis by treating rats with amiloride during and after treatment with N-nitrosomorpholine (NNM).
0 1997 Elsevier Science Ireland Ltd. All rights reserved
110
2. Materials
M. Tatsuta et al. /Cancer Letters 119 (1997) 109-113
and methods
2.1. Animals Sixty inbred 6-week-old male Sprague-Dawley rats were purchased from Japan SLC (Shizuoka, Japan). The animals were housed in suspended metal cages in animal quarters with controlled temperature (21-22”C), humidity (30-50%), and light (12-h cycle) and were given free access to regular chow pellets (Nihon-Nosan, Yokohama, Japan). 2.2. Experimental design The animals were randomly divided into three groups of 20 rats each and from the beginning of the experiment were given drinking water containing 175 mg/l of NNM (Sigma Chemical Co., St. Louis, MO) for 8 weeks. The NNM was dissolved in distilled water at a concentration of 50 g/l and stored in a cool place. The stock solution was diluted to 175 mg/l with tap water just before use and supplied to rats from bottles, which were replaced every other day. From week 9, rats were given normal tap water only until the end of the experiment. From the beginning of the experiment, the animals also received S.C. injections of the vehicle, plain olive oil only (group l), or 2.5 mg/kg (group 2) or 5.0 mg/kg (group 3) body weight of amiloride. Amiloride (Sigma) was injected as a suspension in olive oil to prolong the effective duration of its activity. Injections were given S.C. every other day at a volume of 1 ml/kg body weight, between 1400 and 1500 h, various sites of injections being chosen. The rats in group 1 received 1 ml&g of plain olive oil, administered as for groups 2 and 3. 2.3. Histochemical observations All surviving animals were killed by cervical dislocation under light ether anesthesia at the end of the experiment at week 16. The liver was immediately excised and 2- or 3-mm-thick liver sections obtained from the left and middle lobes were fixed in cold acetone (0-4°C) for 6 h and embedded in paraffin. Serial sections of 3-pm thickness were stained with hematoxylin and eosin and with an anti-rat glutathione-S-transferase, placental type (GST-P) rabbit
serum (Bio Prep Medlabs, Stillorgan, Ireland) and the peroxidase-antiperoxidase method for immunohistochemical examination of GST-P [9]. 2.4. Measurement of GST-P-positive lesions Sections were scored for GST-P-positive lesions without knowledge of their group of origin. Only preneoplastic lesions of 0.2 mm or more in greatest diameter in the plane of transection were counted, since reproducible evaluation of lesions of less than 0.2 mm in diameter was impossible. The transectional area of the lesions in the plane of the tissue section and the area of the entire liver section were measured with an LA-500 Personal Image Analyzer System (Pias, Tokyo, Japan). 2.5. Measurement of labeling index The labeling indices of enzyme-altered lesions and their adjacent liver were examined in week 16. The labeling index was measured with an immunohistochemical analysis kit (Becton-Dickinson, Mountain View, CA) for assaying bromodeoxyuridine (BrdU) incorporation [lO,ll]. For this purpose, five rats in each of groups received 1 ml/kg body weight of olive oil (group l), or 2.5 mg/kg (group 2) or 5.0 mg/kg (group 3) body weight of amiloride. One hour later the animals received an intraperitoneal injection of 20 mg/kg body weight of BrdU (Sigma) and were killed with ether after a further 1 h. Sections obtained from the left liver lobe were promptly fixed in 70% ethanol (0-4’C) for 4 h and embedded in paraffin. Serial sections of 3-pm thickness were immersed in 2 N HCl solution for 30 min at room temperature and then in 0.1 M NazB407 to neutralize the acid. The reactions were stained with anti-BrdU monoclonal antibody (1:25) for 2 h at room temperature. Then sections were washed, stained with biotinconjugated horse anti-mouse antibody (diluted 1:200) for 30 min, and stained with avidin-biotin-peroxidase complex for 30 min. The reaction product was located with 3,3’-diaminobenzidine tetrahydrochloride, and cells containing BrdU were identified by the presence of dark pigment over their nuclei. For determination of the labeling index, the number of BrdU-labeled cells was determined among a total of 1000 hepatocytes or 500 cells in hepatic lesions of 0.7-3.0 mm in
ill
M. Tatsuta et al. /Cancer Letters 119 (1997) 109-113 Table 1 Body weight and liver weight of NNM-treated rats Group no.
Treatment”
Effective no. of rats
Body weight (g)
Liver weight (g)
Relative liver weight (g/l00 g hndy weight i
-I 2 3
Control Amiloride 2.5 mg/kg Amiloride 5.0 mg/kg
Initial
Week 16
161 f 1 173 * 2 168 * 3
363 + I 395 + 10 388 IL 8
20 20 20
15.1 + 0.4 15.5 * 0.5 16.2 f 0.5
---__..l_l-4.1’7 2 0.12 3.94 1- 0.09 4.?9 zk0.18 --.____I___
“Rats were given drinking water containing 175 mg/l of NNM for 8 weeks. From the beginning of the experiment. 2.5 mgJkg (group 2) or 5.0 mg/kg (group 3) body weight of amiloride was injected S.C.every other day. Controls received S.C.injections of olive oil onI>,
greatest diameter. The labeling index was expressed as the percentage of labeled cells among the cells examined. 2.6. Statistical analysis Results were analyzed by the chi-square test, Fisher’s exact probability test, or one-way analysis with Dunn’s multiple comparison [12]. Data are given as means 5 SE. Differences were considered significant when the calculated P value was less than 0.05.
3. Results 3. I. Body and liver weights In week 16, the body weights of rats that had received amiloride were slightly greater than that of control rats, but the difference was not statistically significant (Table 1). Treatment with amiloride had no significant effect on gross liver weight or liver weight relative to body weight.
3.2. Number and size of GST-P-positivr !esions of the liver The size (as mean area) of GST-P-pasitive lesions in group 2 (amiloride at 2.5 mg/kg body weight) and the number (lesions per cm2) and size (mean area and percentage of parenchyma) in group 3 (amiloride at 5.0 mg/kg body weight) were significantly less than in group 1 (control; Table 2). 3.3. Labeling index The labeling index of enzyme-altered lesions was significantly lower in groups 2 (amiloride at 2.5 mg/ kg body weight) and 3 (amiloride at 5.0 mgtkg body weight) than in group 1 (control; Table 3). The labeling index of adjacent hepatocytes was also significantly lower in group 3, but not in group 2, than in group 1.
4. Discussion Our present results indicate that prolonged admin-
Table 2 Number and size of GST-P-positive lesions in the livers of NNM-treated rats Group no.
I 2 3
Treatment”
Control Amiloride 2.5 mg/kg Amiloride 5.0 mg/kg
No. of rats examined
20 20 20
GST-P-positive lesions No./cm’
Mean area (mm’)
97.6 f 5.4 91.6 F 4.0 80.6 * 0.4h
I .09 f. 0.06 0.35 zk 0.02’ 0.36 It: 0.04’
‘For explanation of treatments, see Table I. %gnificantly different from the value for group 1: bP < 0.05, ‘P < 0.001
--__-__5%area 36’7 .5 3.9 32.0 L 2.3 2-4.6 i: 2.0”
112
M. Tatsuta et al. /Cancer L.etter.s 119 (1997) 109-113
Table 3 Labeling indices of enzyme-altered lesions and adjacent liver in NNM-treated rats Group no. Treatment”
1 2 3
Control Amiloride 2.5 mgikg Amiloride 5.0 mg/kg
Labeling index (%) Enzymealtered lesion
Adjacent liver
5.5 f 0.2 3.8 2~ 0.4b 2.6 f 0.1'
1.6 f 0.1 1.3 zk0.1 0.8 f 0.1’
“For explanation of treatments, see Table 1. b.cSignificantly different from the value for group 1: bP < 0.01, cP < 0.001.
istration of amiloride during and after treatment with NNM significantly inhibits the development of GSTP-positive hepatic lesions. Our results also suggest that an amiloride-induced decrease in cell proliferation in enzyme-altered lesions and/or adjacent hepatocytes may be related to suppression of hepatocarcinogenesis. The intracellular mechanisms by which amiloride decreased cell proliferation in GST-P-positive hepatic lesions and/or adjacent hepatocytes are not fully understood, but at least two possible explanations may be considered. One possibility involves the effect of amiloride on intracellular pH (pHi). The pHi plays an important role in the metabolic activation of quiescent cells after a proliferative stimulus [8]. Benedetti et al. [ 131 examined pHi regulation in isolated periportal and perivenular rat hepatocytes and found that pHi recovery from an acid load was inhibited by amiloride by 36% to 7%. Strazzabosco et al. [S] also examined pHi regulation in a well-differentiated hepatoma cell line HepGz in resting conditions and found that amiloride decreased baseline pHi in HepG2 cells. To investigate the signaling pathways to Na+/H+ exchanger activation with epidermal growth factor in hepatocytes, Tanaka et al. [14] measured changes in primary cultured hepatocytes and found that epidermal growth factor induced a pHi increase, but this effect was completely abolished by amiloride. These findings suggest that in HepG2 cells, pHi, which is mainly regulated by Na+/H+ exchange activation, may play a critical role in the growth of some hepatic tumors. A second possible explanations may involve an effect of amiloride on intracellular Ca” ([Ca”]i). Ele-
vated ]Ca2’], is thought to be closely related to cell proliferation [ 151. Ogawa [ 161 examined the effect of amiloride on the [Ca2+]i spike of rat ascites hepatoma cells AH130 and found that amiloride suppressed the [Ca2’], spike induced by addition of serum. The present results indicate that amiloride inhibits hepatocarcinogenesis and this inhibition may be closely related to amiloride’s inhibition of cell proliferation.
Acknowledgements This work was supported in part by a Grant-in-Aid from the Ministry of Health and Welfare for the New Comprehensive lo-Year Strategy for Cancer Control.
References [II C. Frelin, P. Vigne, A. Ladoux, M. Lezdunski, The regulation of the intracellular pH in cells from vertebrates, Eur. J. Biochem. 174 (1988) 3-4. PI S. Grinstein, D. Rotin, J.M. Mason, Naf/Hf exchange and growth factor-induced cytosolic pH changes. Role in cellular proliferation, Biochim. Biophys. Acta 988 (1989) 73-97. [31 R.L. Spark, T.B. Pool, N.K.R. Smith, I.L. Cameron, Effects of amiloride on tumor growth and intracellular element content of tumor cells in viva, Cancer Res. 43 (1993) 73-77. [41 E. Szolgay-Daniel, J. Carlsson, K. Zierold, G. Holtermann, E. Dufan, H. Acker, Effects of amiloride treatment on U-118 MG and U-25 1 MG human glioma and HT-29 human colon carcinoma cells, Cancer Res. 51 (1991) 1039-1044. 151M. Tatsuta, H. Iishi, M. Baba, H. Uehara, A. Nakaizumi, H. Taniguchi, Inhibition by amiloride of gastric carcinogenesis induced by N-methyl-hr-nitro-N-nitrosoguanidine in Wistar rats, Br. J. Cancer 67 (1993) 101 l-1014. &I M. Tatsuta, H. Iishi, M. Baba, H. Uehara, A. Nakaizumi, Chemoprevention by amiloride of experimental carcinogenesis in rat colon induced by azoxymethane, Carcinogenesis 16 (1995) 941-942. I71 M. Tatsuta, H. Iishi, M. Baba, H. Yano, K. Iseki, H. Uehara, A. Nakaizumi, Inhibition by amiloride of experimental carcinogenesis induced by azaserine in rat pancreas. Cancer Lett. ( 1996) in press. @I M. Strazzabosco, C. Poci, C. Spirli, A. Zsembery, A. Granato, M.L. Massimino, G. Crepaldi, Intracellular pH regulation in HepGz cells: effects of epidermal growth factor, transforming growth factor-alpha, and insulin-like growth factor-II on Na+/ H+ exchange activity, Hepatology 22 (1995) 588-597. r91 L.A. Stemberg, P.H. Hardy, J.J. Cuculis, H.G. Mayer, The unlabeled antibody enzyme method of immunochemistry. Preparation and properties of soluble antigen-antibody com-
M. Tatsuta et al. /Cancer plcx (horseradish peroxidase-antihorseradish peroxidase) and its use in identification of spirochaetes, J. Histochem. Cytochem. 8 (1970) 315-333. [ 101 H.G. Gratzner, Monoclonal antibody to 5-bromo- and 5-iododeoxyuridine: a new reagent for detection of DNA replication, Science 2 I8 ( 1982) 474-475. I I 11 G. Morstyn, S.M. Hsu, T. Kinsella, H. Gratzner, A. Russo, J.B. Mitchell, Bromodeoxyuridine in tumors and chromosomes detected with monoclonal antibody, J. Clin. Invest. 72 (1983) 1844-1850. [ 121 C.W. Snedecor, W.C. Cochran, Statistical Methods. Iowa State University Press, Ames, IA, 1967. I I3! A Benedetti, G. Svegliati-Baroni, L. Marucci, R. Mancini,
Letters I19 (IYY7) IOO--~II.?
113
A.M. Jezequel, F. Orlandi, Regulation of intr~tcellular pH in isolated periportal and perivenular rar hepatocvtcs. Gastroenterology 105 (1993) 1797-1805. 1141 Y. Tanaka, N. Hayashi, A. Kaneko, 7‘. ho. M. Horimoto. Y. Sasaki, A. Kasahara. H, Fusamoto, I-. Kam&, C‘haracterifldtion of signaling pathways to Na’ /H cxchangcr activation with epidermal growth factor in hcpatocvh? IIepatology 70 (1994) 966-974. 1151 D.D. Gimy, E.R. Seidal, Polyaminc-dependcn: growth and calmodulin-regulated induction of ornithinc decarboxylase. Am. J. Physiol. 256 (1989) G3422G34X. [16] H. Ogawa, Inhibition of tumor cell invasitrn b> guamdinc analogues. Med. J. Osaka Univ. 46 : 10931 I i ‘i
Cancer Letters I 19 ( 1997) IO9- I 13
Chemoprevention by amiloride against experimental hepatocarcinogenesis induced by Wnitrosomorpholine in Sprague-Dawley rats Masaharu Tatsuta*, Hiroyasu Iishi, Miyako Baba, Hiroyuki Yano, Noriko Sakai, Hiroyuki Uehara, Akihiko Nakaizumi Deppartment oj Gastrointestinal
Oncology, Osaka Medical Center Nakumichi I-chome, Higashinari-ku,
.for Cancer
and Cardiovascular Osaka 537. .Iupan
Dwow~,
i- :
Received 12 March lYY7; received in revised form 25 April 1997: accepted 25 April 199)
--.-- ..__... -l..-..-- ..-Abstract
The effects of prolonged administration of the diuretic amiloride on hepatocarcinogenesis induced by ~-nitrosomr,rpholine
(NNM) and the labeling index of the liver were investigated in male Sprague-Dawley rats. Rats were given &inking water containing NNM for 8 weeks and received S.C.injections of 2.5 mg/kg or 5.0 mg/kg body weight of amiloride every other day until the end of the experiment at week 16. Preneoplastic and neoplastic lesions staining positively for glutathione-Stransferase, placental type (GST-P) were examined immunohistochemically. In week 16, administration of amiloride at 2.5 mg/kg body weight significantly reduced the size (as mean area) of GST-P-positive lesions, and amiloride at 5.0 mglkg body weight significantly reduced the number (as no./cm*) and sizes (as mean area and percent of parenchyma) of GST-P-positive lesions. Amiloride at both dosagesalso significantly decreasedthe labeling index of enzyme-alteredlesions and amiloride at
higher dosage significantly decreasedthe labeling index of adjacent hepatocytes.These findings indicate that amiioride inhibits hepatocarcinogenesis and suggest that this effect may be closely related to amiloride’s inhibition of celi proliferation in
enzyme-alteredlesions and/or adjacent liver. 0 1997 Elsevier Science Ireland Ltd.
Keywords:
Amiloride: Hepatocarcinogenesis;Chemoprevention;N-Nitrosomorpholine
1. Introduction The diuretic drug the Na’/H+ antiport inhibit cell growth [ 3,4]. Previously we
amiloride is a potent inhibitor of [ 1.21 and has been reported to in some extrahepatic tumors found that prolonged administra-
*Corresponding author. Tel.: +81 6 9721181; fax: +81 6 98 14067.
03043835/97/$17.00 PII
SO304-3805(97)00262-O
___-- .. -._
_.__ - ..-_-.
tion of amiloride inhibits the development of cancers in the stomach [5], colon [6], and pancreas [7] in rats. Amiloride strongly inhibits [‘Hlthymidine incorporation in a well-differentiated hepat.oma cell line HepGl [S]. These findings suggest that amiloride may inhibit hepatocarcinogenesis. To investigate this possibility, in the present work we examined amiloride‘s effects on hepatocarcinogenesis by treating rats with amiloride during and after treatment with N-nitrosomorpholine (NNM).
0 1997 Elsevier Science Ireland Ltd. All rights reserved
110
2. Materials
M. Tatsuta et al. /Cancer Letters 119 (1997) 109-113
and methods
2.1. Animals Sixty inbred 6-week-old male Sprague-Dawley rats were purchased from Japan SLC (Shizuoka, Japan). The animals were housed in suspended metal cages in animal quarters with controlled temperature (21-22”C), humidity (30-50%), and light (12-h cycle) and were given free access to regular chow pellets (Nihon-Nosan, Yokohama, Japan). 2.2. Experimental design The animals were randomly divided into three groups of 20 rats each and from the beginning of the experiment were given drinking water containing 175 mg/l of NNM (Sigma Chemical Co., St. Louis, MO) for 8 weeks. The NNM was dissolved in distilled water at a concentration of 50 g/l and stored in a cool place. The stock solution was diluted to 175 mg/l with tap water just before use and supplied to rats from bottles, which were replaced every other day. From week 9, rats were given normal tap water only until the end of the experiment. From the beginning of the experiment, the animals also received S.C. injections of the vehicle, plain olive oil only (group l), or 2.5 mg/kg (group 2) or 5.0 mg/kg (group 3) body weight of amiloride. Amiloride (Sigma) was injected as a suspension in olive oil to prolong the effective duration of its activity. Injections were given S.C. every other day at a volume of 1 ml/kg body weight, between 1400 and 1500 h, various sites of injections being chosen. The rats in group 1 received 1 ml&g of plain olive oil, administered as for groups 2 and 3. 2.3. Histochemical observations All surviving animals were killed by cervical dislocation under light ether anesthesia at the end of the experiment at week 16. The liver was immediately excised and 2- or 3-mm-thick liver sections obtained from the left and middle lobes were fixed in cold acetone (0-4°C) for 6 h and embedded in paraffin. Serial sections of 3-pm thickness were stained with hematoxylin and eosin and with an anti-rat glutathione-S-transferase, placental type (GST-P) rabbit
serum (Bio Prep Medlabs, Stillorgan, Ireland) and the peroxidase-antiperoxidase method for immunohistochemical examination of GST-P [9]. 2.4. Measurement of GST-P-positive lesions Sections were scored for GST-P-positive lesions without knowledge of their group of origin. Only preneoplastic lesions of 0.2 mm or more in greatest diameter in the plane of transection were counted, since reproducible evaluation of lesions of less than 0.2 mm in diameter was impossible. The transectional area of the lesions in the plane of the tissue section and the area of the entire liver section were measured with an LA-500 Personal Image Analyzer System (Pias, Tokyo, Japan). 2.5. Measurement of labeling index The labeling indices of enzyme-altered lesions and their adjacent liver were examined in week 16. The labeling index was measured with an immunohistochemical analysis kit (Becton-Dickinson, Mountain View, CA) for assaying bromodeoxyuridine (BrdU) incorporation [lO,ll]. For this purpose, five rats in each of groups received 1 ml/kg body weight of olive oil (group l), or 2.5 mg/kg (group 2) or 5.0 mg/kg (group 3) body weight of amiloride. One hour later the animals received an intraperitoneal injection of 20 mg/kg body weight of BrdU (Sigma) and were killed with ether after a further 1 h. Sections obtained from the left liver lobe were promptly fixed in 70% ethanol (0-4’C) for 4 h and embedded in paraffin. Serial sections of 3-pm thickness were immersed in 2 N HCl solution for 30 min at room temperature and then in 0.1 M NazB407 to neutralize the acid. The reactions were stained with anti-BrdU monoclonal antibody (1:25) for 2 h at room temperature. Then sections were washed, stained with biotinconjugated horse anti-mouse antibody (diluted 1:200) for 30 min, and stained with avidin-biotin-peroxidase complex for 30 min. The reaction product was located with 3,3’-diaminobenzidine tetrahydrochloride, and cells containing BrdU were identified by the presence of dark pigment over their nuclei. For determination of the labeling index, the number of BrdU-labeled cells was determined among a total of 1000 hepatocytes or 500 cells in hepatic lesions of 0.7-3.0 mm in
ill
M. Tatsuta et al. /Cancer Letters 119 (1997) 109-113 Table 1 Body weight and liver weight of NNM-treated rats Group no.
Treatment”
Effective no. of rats
Body weight (g)
Liver weight (g)
Relative liver weight (g/l00 g hndy weight i
-I 2 3
Control Amiloride 2.5 mg/kg Amiloride 5.0 mg/kg
Initial
Week 16
161 f 1 173 * 2 168 * 3
363 + I 395 + 10 388 IL 8
20 20 20
15.1 + 0.4 15.5 * 0.5 16.2 f 0.5
---__..l_l-4.1’7 2 0.12 3.94 1- 0.09 4.?9 zk0.18 --.____I___
“Rats were given drinking water containing 175 mg/l of NNM for 8 weeks. From the beginning of the experiment. 2.5 mgJkg (group 2) or 5.0 mg/kg (group 3) body weight of amiloride was injected S.C.every other day. Controls received S.C.injections of olive oil onI>,
greatest diameter. The labeling index was expressed as the percentage of labeled cells among the cells examined. 2.6. Statistical analysis Results were analyzed by the chi-square test, Fisher’s exact probability test, or one-way analysis with Dunn’s multiple comparison [12]. Data are given as means 5 SE. Differences were considered significant when the calculated P value was less than 0.05.
3. Results 3. I. Body and liver weights In week 16, the body weights of rats that had received amiloride were slightly greater than that of control rats, but the difference was not statistically significant (Table 1). Treatment with amiloride had no significant effect on gross liver weight or liver weight relative to body weight.
3.2. Number and size of GST-P-positivr !esions of the liver The size (as mean area) of GST-P-pasitive lesions in group 2 (amiloride at 2.5 mg/kg body weight) and the number (lesions per cm2) and size (mean area and percentage of parenchyma) in group 3 (amiloride at 5.0 mg/kg body weight) were significantly less than in group 1 (control; Table 2). 3.3. Labeling index The labeling index of enzyme-altered lesions was significantly lower in groups 2 (amiloride at 2.5 mg/ kg body weight) and 3 (amiloride at 5.0 mgtkg body weight) than in group 1 (control; Table 3). The labeling index of adjacent hepatocytes was also significantly lower in group 3, but not in group 2, than in group 1.
4. Discussion Our present results indicate that prolonged admin-
Table 2 Number and size of GST-P-positive lesions in the livers of NNM-treated rats Group no.
I 2 3
Treatment”
Control Amiloride 2.5 mg/kg Amiloride 5.0 mg/kg
No. of rats examined
20 20 20
GST-P-positive lesions No./cm’
Mean area (mm’)
97.6 f 5.4 91.6 F 4.0 80.6 * 0.4h
I .09 f. 0.06 0.35 zk 0.02’ 0.36 It: 0.04’
‘For explanation of treatments, see Table I. %gnificantly different from the value for group 1: bP < 0.05, ‘P < 0.001
--__-__5%area 36’7 .5 3.9 32.0 L 2.3 2-4.6 i: 2.0”
112
M. Tatsuta et al. /Cancer L.etter.s 119 (1997) 109-113
Table 3 Labeling indices of enzyme-altered lesions and adjacent liver in NNM-treated rats Group no. Treatment”
1 2 3
Control Amiloride 2.5 mgikg Amiloride 5.0 mg/kg
Labeling index (%) Enzymealtered lesion
Adjacent liver
5.5 f 0.2 3.8 2~ 0.4b 2.6 f 0.1'
1.6 f 0.1 1.3 zk0.1 0.8 f 0.1’
“For explanation of treatments, see Table 1. b.cSignificantly different from the value for group 1: bP < 0.01, cP < 0.001.
istration of amiloride during and after treatment with NNM significantly inhibits the development of GSTP-positive hepatic lesions. Our results also suggest that an amiloride-induced decrease in cell proliferation in enzyme-altered lesions and/or adjacent hepatocytes may be related to suppression of hepatocarcinogenesis. The intracellular mechanisms by which amiloride decreased cell proliferation in GST-P-positive hepatic lesions and/or adjacent hepatocytes are not fully understood, but at least two possible explanations may be considered. One possibility involves the effect of amiloride on intracellular pH (pHi). The pHi plays an important role in the metabolic activation of quiescent cells after a proliferative stimulus [8]. Benedetti et al. [ 131 examined pHi regulation in isolated periportal and perivenular rat hepatocytes and found that pHi recovery from an acid load was inhibited by amiloride by 36% to 7%. Strazzabosco et al. [S] also examined pHi regulation in a well-differentiated hepatoma cell line HepGz in resting conditions and found that amiloride decreased baseline pHi in HepG2 cells. To investigate the signaling pathways to Na+/H+ exchanger activation with epidermal growth factor in hepatocytes, Tanaka et al. [14] measured changes in primary cultured hepatocytes and found that epidermal growth factor induced a pHi increase, but this effect was completely abolished by amiloride. These findings suggest that in HepG2 cells, pHi, which is mainly regulated by Na+/H+ exchange activation, may play a critical role in the growth of some hepatic tumors. A second possible explanations may involve an effect of amiloride on intracellular Ca” ([Ca”]i). Ele-
vated ]Ca2’], is thought to be closely related to cell proliferation [ 151. Ogawa [ 161 examined the effect of amiloride on the [Ca2+]i spike of rat ascites hepatoma cells AH130 and found that amiloride suppressed the [Ca2’], spike induced by addition of serum. The present results indicate that amiloride inhibits hepatocarcinogenesis and this inhibition may be closely related to amiloride’s inhibition of cell proliferation.
Acknowledgements This work was supported in part by a Grant-in-Aid from the Ministry of Health and Welfare for the New Comprehensive lo-Year Strategy for Cancer Control.
References [II C. Frelin, P. Vigne, A. Ladoux, M. Lezdunski, The regulation of the intracellular pH in cells from vertebrates, Eur. J. Biochem. 174 (1988) 3-4. PI S. Grinstein, D. Rotin, J.M. Mason, Naf/Hf exchange and growth factor-induced cytosolic pH changes. Role in cellular proliferation, Biochim. Biophys. Acta 988 (1989) 73-97. [31 R.L. Spark, T.B. Pool, N.K.R. Smith, I.L. Cameron, Effects of amiloride on tumor growth and intracellular element content of tumor cells in viva, Cancer Res. 43 (1993) 73-77. [41 E. Szolgay-Daniel, J. Carlsson, K. Zierold, G. Holtermann, E. Dufan, H. Acker, Effects of amiloride treatment on U-118 MG and U-25 1 MG human glioma and HT-29 human colon carcinoma cells, Cancer Res. 51 (1991) 1039-1044. 151M. Tatsuta, H. Iishi, M. Baba, H. Uehara, A. Nakaizumi, H. Taniguchi, Inhibition by amiloride of gastric carcinogenesis induced by N-methyl-hr-nitro-N-nitrosoguanidine in Wistar rats, Br. J. Cancer 67 (1993) 101 l-1014. &I M. Tatsuta, H. Iishi, M. Baba, H. Uehara, A. Nakaizumi, Chemoprevention by amiloride of experimental carcinogenesis in rat colon induced by azoxymethane, Carcinogenesis 16 (1995) 941-942. I71 M. Tatsuta, H. Iishi, M. Baba, H. Yano, K. Iseki, H. Uehara, A. Nakaizumi, Inhibition by amiloride of experimental carcinogenesis induced by azaserine in rat pancreas. Cancer Lett. ( 1996) in press. @I M. Strazzabosco, C. Poci, C. Spirli, A. Zsembery, A. Granato, M.L. Massimino, G. Crepaldi, Intracellular pH regulation in HepGz cells: effects of epidermal growth factor, transforming growth factor-alpha, and insulin-like growth factor-II on Na+/ H+ exchange activity, Hepatology 22 (1995) 588-597. r91 L.A. Stemberg, P.H. Hardy, J.J. Cuculis, H.G. Mayer, The unlabeled antibody enzyme method of immunochemistry. Preparation and properties of soluble antigen-antibody com-
M. Tatsuta et al. /Cancer plcx (horseradish peroxidase-antihorseradish peroxidase) and its use in identification of spirochaetes, J. Histochem. Cytochem. 8 (1970) 315-333. [ 101 H.G. Gratzner, Monoclonal antibody to 5-bromo- and 5-iododeoxyuridine: a new reagent for detection of DNA replication, Science 2 I8 ( 1982) 474-475. I I 11 G. Morstyn, S.M. Hsu, T. Kinsella, H. Gratzner, A. Russo, J.B. Mitchell, Bromodeoxyuridine in tumors and chromosomes detected with monoclonal antibody, J. Clin. Invest. 72 (1983) 1844-1850. [ 121 C.W. Snedecor, W.C. Cochran, Statistical Methods. Iowa State University Press, Ames, IA, 1967. I I3! A Benedetti, G. Svegliati-Baroni, L. Marucci, R. Mancini,
Letters I19 (IYY7) IOO--~II.?
113
A.M. Jezequel, F. Orlandi, Regulation of intr~tcellular pH in isolated periportal and perivenular rar hepatocvtcs. Gastroenterology 105 (1993) 1797-1805. 1141 Y. Tanaka, N. Hayashi, A. Kaneko, 7‘. ho. M. Horimoto. Y. Sasaki, A. Kasahara. H, Fusamoto, I-. Kam&, C‘haracterifldtion of signaling pathways to Na’ /H cxchangcr activation with epidermal growth factor in hcpatocvh? IIepatology 70 (1994) 966-974. 1151 D.D. Gimy, E.R. Seidal, Polyaminc-dependcn: growth and calmodulin-regulated induction of ornithinc decarboxylase. Am. J. Physiol. 256 (1989) G3422G34X. [16] H. Ogawa, Inhibition of tumor cell invasitrn b> guamdinc analogues. Med. J. Osaka Univ. 46 : 10931 I i ‘i