Wu Q et al / Acta Pharmacol Sin 2002 Sep; 23 (9): 835-841

Effects of retinoic acid on metastasis and its related proteins in gastric cancer cells in vivo and in vitro¹

WU Qiao², CHEN Yu-Qiang³, CHEN Zheng-Ming, CHEN Fu, SU Wen-Jin

Key Laboratory of the Ministry of Education for Cell Biology and Tumor Cell Engineering, School of Life Sciences, Xiamen University, Xiamen 361005, China

¹ Supported by the National Outstanding Youth Science Foundation of China (B type, ¡í 39825502), the National Natural Science Foundation of China (¡í 39880015), and Key Project of Science & Technology of the Ministry of Education (00073).

² Correspondence to WU Qiao, PhD. Phn 86-592-218-2542. Fax 86-592-208-6630. E-mail xgwu@xmu.edu.cn

³ Now at Hospital 174, PLA, Xiamen 361003, China.

Received 2001-09-14 Accepted 2002-06-20

KEY WORDS tretinoin; neoplasm metastasis; stomach neoplasms; nude mice

ABSTRACT

AIM: To investigate the effects of all-trans retinoic acid (ATRA) on metastasis and its related proteins in human gastric cancer cells in vivo and in vitro. METHODS: Gastric cancer cells, MGC80-3 and SGC-7901, were inoculated into spleen subcapsule of nude mice, respectively. Nude mice were administered with ATRA (0.7 mg/kg, ig) every other day. Six weeks later, nude mice were sacrificed. All the tumors formed in spleen and in liver were removed. Some of them were fixed, and then embedded. Others were kept in liquid nitrogen for further use. Expression level of proteins in tumor and in cell was analyzed by Western blot. Microvessel in tumor section was shown by immunohistochemistry and adhesive ability of cell to amnion was measured by adhesion assay. RESULTS: When inoculated nude mice were treated with ATRA, the xenograft tumors in spleen and metastatic tumors in liver were suppressed by 50 % respectively, and inhibition of microvessel formation in xenograft and metastatic tumors was also observed obviously. Although ATRA regulated expression of nm23 and mts1/p16 proteins at different patterns in vivo and in vitro, high ratio of nm23:mts1/p16 was in association with low adhesive activity of cells. In addition, ATRA induced ICAM-1 protein expression in vivo and in vitro. CONCLUSION: ATRA inhibits the growth of xenograft tumors and their metastasis to liver. This process may be associated with regulation of metastatic related proteins, including nm23, mts1/p16, and ICAM-1 in vivo and in vitro.

INTRODUCTION

Although significant progress has been made in the surgical treatment of tumors, as well as in diagnostic techniques, more than 50 % patients still die from malignant tumors. It is partially due to distant metastasis at the time of diagnosis. Metastasis is known to consist of a cascade of multiple steps. Tumor cell migrates from the primary lesion, invades the tumor border and hosts tissues, initiates angiogenesis, and colonizes at distal sites^[1]. Thus, blocking one of these steps is of clinical significance in prevention of tumor metastasis.

Metastatic process is regulated by activation and inactivation of several specific genes and proteins. The nm23 gene, an anti-metastatic gene, was found originally to be correlated with tumor metastatic potential in murine melanoma cell^[2] and was associated with a good prognosis in breast cancer^[3]. However, the relationship between expression of nm23 protein and prognostic value was not found in colon cancer, neuroblastoma, or pulmonary carcinoma^[4-6]. Another gene mts1/p16, together with its protein, was also found to be related to high metastatic potential and growth of some human tumors^[7]. It was reported recently that there was a close relationship between nm23 and mts1/p16 in regulation of metastatic behavior^[8]. Since mts1/p16 mutations occur preferentially in metastatic tumors, they might be important events in late phases of tumor progression and could represent useful markers of tumor aggressiveness in non-small cell lung carcinomas^[9].

Angiogenesis has been shown to be critical for tumor growth. It facilitates the infiltration of reparative cells, enhances the delivery of oxygen, nutrients, growth factors, and cytokines, and allows the removal of waste products^[10]. Primary and metastatic tumors can not grow beyond 2 mm in diameter without an enhanced vascularity. The newly formed blood vessels are more penetrative to tumor mass and may contribute to metastasis^[11]. Thus, anti-angiogenesis has been studied as one of the major approaches for tumor treatment.

Retinoids play an important role in proliferation and differentiation of a variety of cell types. In vitro and animal studies have suggested that retinoids are promising agents in the prevention and treatment of human cancers. Recently, we have developed an in vivo model for estimating the ability of tumor metastasis^[12], in which tumor cells were inoculated into spleen subcapsule of nude mice, and xenograft tumors formed in spleen metastasized to liver. In the present study, we investigated the effects of all-trans retinoic acid (ATRA) on metastasis and its related proteins in human gastric cancer cells in vivo and in vitro.

MATERIALS AND METHODS

Cell lines and culture condition Human gastric cancer cell line SGC-7901 was purchased from the Institute of Cell Biology, Shanghai, China. MGC80-3 cell line was established by Cancer Research Center in Xiamen University. Both of cell lines were maintained in RPMI-1640 medium supplemented with 10 % FCS, glutamine 1 mmol/L, and benzylpenicillin 100 U/L.

Inoculation of nude mice with cancer cells Athymic nude mice (BALB/c, SPF Grade, 18-21 g, 7 weeks old, Certificate ¡í 23-007 approved by Administration Committee of Medical Experimental Animal, Fujian Province) were housed in a laminar flow under sterilized condition. The temperature was maintained at 28 ¡æ. The light was controlled at 12 h light and 12 h dark. Mice were fed with autoclaved mouse chow.

Cells were trypsinized and washed twice with phosphate buffered saline (PBS). The inoculation dosage was 0.05 mL cell suspension per mouse (cell density was 2.5×10¹⁰/L). Before inoculation, nude mouse was injected with 0.1 mL of 0.5 % CCl₄, then 24 h later, anesthetized with 0.625 % sodium pentothal (0.2 mL, ip) and subjected to a routine surgical operation. Cells were inoculated into spleen subcapsule of nude mice. The inoculated nude mice were randomly separated into experimental and control groups (6 mice for each group, 3 male and 3 female). For experimental group, nude mice were administered with ATRA 0.7 mg/kg every other day (ATRA was dissolved in ethanol and Me₂SO with a ratio of 1:1). For control group, nude mice were administered with the corresponding dissolvent without ATRA. After 6 weeks, nude mice were sacrificed and all the tumors formed in spleen and in liver were removed. Some of them were fixed and embedded for immunohistochemistry and others were kept in liquid nitrogen for Western blot.

Immunohistochemistry Sections from spleen tumor and liver tumor were deparaffinized. Endogenous peroxidase was blocked by 0.3 % hydrogen peroxide in methanol for 30 min. After washed with PBS, sections were incubated with 10 % goat serum in PBS for 20 min at room temperature to block non-specific binding of the second antibody, then incubated overnight at
4 ¡æ with rabbit anti-human factor-VIII-related-antigen antibody (Zymed) 1:100 in PBS containing 1 % bovine serum albumin. Rinsed three times in PBS, sections were treated with biotinylated anti-rabbit immunoglobulin for 1 h at room temperature, then washed again and reacted with streptoavidin-biotin system using 0.04 % 3,3'-diaminobenzidene tetrahydrochloride for 1 min as chromogen. To calculate inhibitory rate of microvessel density (MVD), at least four sections, including those from spleen tumors and liver tumors, were observed under microscope and microvessels were counted within 20 hot spots randomly.

Western blot and quantitative analysis Cells or tumors kept in liquid nitrogen were lysed in NaCl 150 mmol/L, Tris-HCl 10 mmol/L (pH 7.4), edetic acid 5 mmol/L, 1 % TritonX-100, and phenylmethylsulphonyl fluoride. Total protein (50 mg) was separated by 10 % SDS-PAGE and transferred to nitrocellulose. Protein was visualized using ECL (enhanced chemilumine-scence) detection system (Amersham), after incubation for 2-4 h at room temperature with primary antibodies, including ICAM-1, nm23, and mts1/p16 (Santa Cruz), and their corresponding secondary antibodies, respectively. To quantify relative expression level of nm23 or mts1/p16 protein, each band revealed by ECL was scanned by densitometer.

Adhesion assay Amnion, obtained from fresh placenta, was washed twice with PBS and immersed in NH₄OH 0.25 mol/L for 2 h to remove the epithelial cells. When treated with ATRA for 30 d and 80 d respectively, cells were trypsinized and suspended in RPMI-1640 medium to a final density of 5×10⁷ /L. Cell suspension 2 mL was added to the upper level of Boyden chamber, and RPMI-1640 medium was filled to the bottom level of chamber. After incubated for 4 h and 6 h respectively at 37 ¡æ in CO₂ incubator, the upper chamber was removed and the cells attached to amnion were trypsinized and counted under microscope. Result represented the mean of two independent experiments.

Statistical analysis Tukey's procedure was applied to estimate the difference among groups. Statistical significance was defined as P<0.05.

RESULTS

Effects of ATRA on growth of xenograft tumor in spleen and its metastasis to liver Gastric cancer cells were inoculated into spleen subcapsule of nude mice. Inoculated nude mice in experimental group were administered with ATRA every other day. After 6 weeks, nude mice were sacrificed. Xenograft tumors formed in spleen were found in all of mice except one mouse in experimental group. However, number of spleen tumors in experimental group was fewer than that in control group, the mean weight of spleen tumors in experimental group was also lighter than that in control group significantly (P<0.05, Tab 1). Metastatic tumors found in liver were multiple and noncapsuled in both experimental and control groups. However, the occurrence of metastatic tumors was obviously decreased by 50 % in experimental groups, number of metastatic tumors reduced much more and the mean weight of liver was also lighter (Tab 1), compared with the control group.

Tab 1. Effect of ATRA (0.7 mg/kg, ig every other day for 6 weeks) on xenograft and metastatic tumors in nude mice. n=6. Mean±SD. ^bP<0.05, ^cP<0.01 vs control.

Cell line	Group		Spleen	Weight/g		Liver	Weight/g
		Rate of tumor	Number		Rate of tumors	Number
		occurrence/%	Of tumors		occurrence/%	Of tumor
MGC1)	Control	100(5/5)2)	13	1.6¡À1.0	100(5/5)	134	3.2¡À1.0
	ATRA	100(6/6)	5	0.4¡À0.5b	50(3/6)	8	1.7¡À0.3c
SGC	Control	100(6/6)	7	0.6¡À0.2	100(6/6)	48	1.8¡À0.4
	ATRA	83.3(5/6)	6	0.3¡À0.1b	50(3/6)	18	1.3¡À0.4c

¹⁾ Some data of MGC80-3 cell were cited from reference 13.
²⁾ One of the MGC80-3 inoculated mice in control group died at the 40th day due to the development of tumor.

Effect of ATRA on microvessel formation Immunohistochemistry analysis indicated a strong staining for microvessel in all of the tumor sections, including those in spleen tumors and liver tumors. However, microvessel density (MVD) in experimental group was lower than that in control group by statistical analysis. The inhibitory rate of MVD showed significant decrease in MGC80-3 experimental group (P<0.05, Tab 2) and more significant decrease in SGC-7901 experimental group (P<0.01, Tab 2).

Tab 2. Effect of ATRA (0.7 mg/kg, ig every other day for 6 weeks) on microvessel density (MVD) in vivo. n=4. Mean±SD. ^bP<0.05, ^cP<0.01 vs control.

Cell	Tumor	MVD		Inhibitory rate
line		Control	ATRA	ofMVD/%
MGC	Spleen	11.2¡À3.0	8.0¡À2.9b	28.6
	Liver	10.9¡À2.4	8.4¡À2.4b	22.9
SGC	Spleen	8.6¡À2.0	6.6¡À1.5b	23.7
	Liver	13.3¡À2.3	8.3¡À1.6c	37.6

Effects of ATRA on regulation of nm23, mts1/p16, and ICAM-1 proteins in vivo and in vitro To investigate the molecular mechanism of ATRA in affecting metastatic progress, the expression level of nm23, mts1/p16, and ICAM-1 proteins correlated with metastasis were detected by Western blot. Protein in vivo was extracted from liver tumors that represented metastatic tumor lesion. Protein in vitro was obtained from cells. The results showed that regulation of ATRA in these two types of protein (in vivo and in vitro) differed obviously. Expression of nm23 and ICAM-1 proteins in vivo was up-regulated by ATRA, and expression of mts1/p16 protein was not changed obviously (Fig 1). In contrast to the result shown in vivo, after treatment of cells with ATRA for 24 h in vitro, expression of nm23 protein was decreased in MGC80-3 cells and increased in SGC-7901 cells. Expression of mts1/p16 protein was down-regulated by ATRA in both MGC80-3 and SGC-7901 cells. However, ICAM-1 showed the similar trend as that in vivo (Fig 1). The expression level of nm23 and mts1/p16 proteins was further quantified by densitometer, respectively. The values of relative abundances were shown in Tab 3. Although ATRA treatment led to increase or decrease in nm23 and mts1/p16 protein levels in different groups respectively, nm23:mts1/p16 ratio always kept higher in experimental group than that in control group in vivo and in vitro.

Fig 1. Expressions of nm23, mts1/p16, and ICAM-1 proteins shown by Western blot in vivo and in vitro. a-Tubulin was used to quantify the amount of protein used. Protein in vivo was extracted from several liver tumors, and protein in vitro was extracted from cells treated with ATRA 1.0 mmol/L for 24 h.

Tab 3. Quantitative analysis of nm23 and mts1/p16 protein level by ATRA in vivo (0.7 mg/kg, ig every other day for 6 weeks) and in vitro (10 mmol/L, for 24 h).

To further determine the possible link between nm23 and mts1/p16, various concentrations of ATRA were used to treat cells for 24 h in vitro. Although expression of nm23 and p16 proteins, revealed by Western blot, was down-regulated in an ATRA concentration-dependent manner in both MGC80-3 and SGC-7901 cells (Fig 2), nm23:mts1/p16 ratio displayed an increase also in an ATRA concentration-dependent manner after quantitative analysis by densitometer (Tab 4).

Fig 2. Western blot was performed to examine the levels of nm23 and mts1/p16 protein expression. a-Tubulin was used to quantify the amount of protein used. Protein was extracted from cells treated with various concentrations of ATRA for 24 h.

Tab 4. Quantitative analysis of nm23 and mts1/p16 protein level at various concentrations of ATRA (mmol/L, for 24 h).

Effect of ATRA on cell adhesion To elucidate effect of ATRA on cell adhesion, the adhesive ability of cell to amnion was measured by adhesion assay. As shown in Tab 5, the adhesive ability of cell to amnion was suppressed remarkably with extension of ATRA treatment. When cells were treated with ATRA for 30 d and 80 d respectively, the rate of SGC-7901 cell adhesion (4 h and 6 h attachment) and that of MGC80-3 cell adhesion (6 h attachment) showed more significant decrease (P<0.01, Tab 5), compared with the control group.

Tab 5. Effect of ATRA (1.0 mmol/L) on adhesive capability of cells. n=2. Mean±SD. ^bP<0.05, ^cP<0.01 vs control.

¹⁾ Some data of MGC80-3 cells were cited from reference 13.

DISCUSSION

In this animal model, nude mice were injected with CCl₄ and some growth factors were secreted to facilitate tumor cell proliferation and mass formation in liver. Autopsy and statistics showed that when gastric cancer cells were inoculated into spleen subcapsule of nude mice, xenograft tumors were formed in spleen, and then metastasized to liver. However, after nude mice were administered with ATRA every other day for 6 weeks, it apparently caused the suppression in growth of xenograft tumors in spleen. More importantly, tumor metastasis was inhibited obviously in experimental group. Only 50 % of the splenic tumor-bearing individuals was developed into the metastasized tumors in liver, as compared with 100 % metastasis occurred in the control group. These results suggested that ATRA might be a promising agent for tumor prevention and treatment in vivo. In addition, the fact that ATRA inhibited angiogenesis formation in xenograft splenic tumor and in metastasized liver tumor implied that ATRA might be one of the valuable candidates clinically used for anti-angiogenesis. This result was in accordance with other reports from breast and vagina cancer cells^[14].

Low expression of nm23 protein has been found to be correlated with high metastatic potential in several tumors^[2,3]. However, some data revealed the contrast results in other tumors, suggesting that the function of nm23 in cancer cells remained unclear. More recently, it was reported that expression of nm23 gene alone was not related to metastatic behavior, but the relative ratio of nm23:mts1/p16 protein correlated with metastatic potential^[15]. This evidence implied that they might act as a co-regulator on tumor metastasis. Consistent with this observation, our data showed that not only the ratio of nm23:mts1/p16 protein in experimental groups was increased obviously in an ATRA concentration-
dependent manner, but always kept higher than that in control group, no matter whether expression of nm23 and mts1/p16 proteins was decreased or increased by ATRA in vivo or in vitro. According to the reports, SGC-7901 cell line was established from the metastatic lesion of lymph node in gastric cancer^[16], and MGC80-3 cell line from the mass of gastric adenocarcinoma with lower metastasis^[17]. Thus some of special characteristics associated with metastasis should be shown in SGC-7901 cells markedly after treatment of ATRA. They were displayed in vivo and in vitro: highest inhibitory effect on tumor microvessel density, more significant suppression rate of adhesion, as well as induction of nm23 protein expression. These features might explain partially why ATRA regulation on expression of nm23 and mts1/p16 proteins displayed distinct patterns in vivo and in vitro between MGC80-3 and SGC-7901 cells. It was likely that in gastric cancer cells, expression of nm23 protein along in vivo was associated with metastatic behavior, and the relative ratio of nm23:mts1/p16 expression in vitro correlated with metastatic potential. Accordingly, it was quite possible that (1) phenomena observed in vitro were not relevant in vivo, since ATRA might be quickly metabolized and what was observed in vivo might be mediated by a metabolite and differed significantly from effects by ATRA in vitro. (2) Role of nm23 with regard to prognosis was cell-type specific, which was associated with metastatic capability of cells. Thus inhibitory effect of ATRA on metastasis might lie in its concurrent regulation on expression of two proteins, nm23 and mts1/p16. Of course, it should be necessary to further address this possibility by using different types of cell, including high and low metastatic cancer cell lines.

Lakshimi et al pointed out that one possible link between nm23 and mts1/p16 and their target effects might be microtubule assembly^[8]. Parket et al reported that nm23 and mts1/p16 expressions affected the invasive properties of tumor cells^[18]. Present results, combining with our other results^[13], further confirmed that high ratio of nm23:mts1/p16 was in association with low adhesive activity, such as decrease of adhesion, disappearance of microvilli at surface of tumor cell, re-assembly of microfilament^[13].

In cancer therapy, retinoic acid (RA) has been used to increase ICAM-1 expression to render tumor cell more susceptible to immune attack^[19]. Thus, it was suggested that up-regulation of ICAM-1 by RA played a major role in cell adhesion, cell movement, and even tissue development. Induction of ICAM-1 expression by ATRA depended on retinoic acid response element (RARE) located at 266 bp upstream of the 5' translation start site in the ICAM-1 promoter^[20]. Our previous study has shown that effect of RA was mainly mediated by its receptors, which bound to RARE and regulated positively or negatively transcriptional activities of target genes^[21]. Thus, it was implied that in gastric cancer cells, ATRA might up-regulate ICAM-1 expression directly through the mediation of retinoic acid receptors that bound to RARE located in the ICAM-1 promoter. Surely, the detail, especially on function of ICAM-1 in gastric cancer cells, should be studied further.

REFERENCES

• 1 Aznavoorian S, Murphy AN, Stetler-Stevenson WG, Liotta LA. Molecular aspects of tumor cell invasion and metastasis. Cancer 1993; 71: 1368-83.
• 2 Steeg PS, Bevilacqua G, Kopper L, Thorgeirsson UP, Talmadge JE, Liotta LA, et al. Evidence for a novel gene associated with low tumor metastatic potential. J Natl Cancer Inst 1988; 80: 200-4.
• 3 Hennessy C, Henry JA, May FE, Westley BR, Angus B, Lennard TWJ. Expression of the antimetastatic gene nm23 in human breast cancer: an association with good prognosis. J Natl Cancer Inst 1991; 83: 281-5.
• 4 Haut M, Steeg PS, Willson JK, Markowitz SD. Induction of nm23 gene expression in human colonic and equal expression in colon tumor of high and low metastatic potential. J Natl Cancer Inst 1991; 83: 712-6.
• 5 Hailat N, Keim DR, Reynolds CP, Seeger RC, Lottspeich F, Strahler J, et al. High levels p19/nm23 protein in neuroblastoma are associated with advanced stage disease and with Nmyc gene amplification. J Clin Invest 1991; 88: 341-5.
• 6 Higashiyama M, Doi O, Yokouchi H, Kodama K, Nakamori S, Tateishi R, et al. Immunohistochemical analysis of nm23 gene product/NDP kinase expression in pulmonary adenocarcinoma: lack of prognostic value. Br J Cancer 1992; 66: 533-6.
• 7 Okamoto A, Demetrick DJ, Spillare EA, Hagiwara K, Hussain SP, Bennett WP, et al. Mutations and altered expression of p16INK4 in human cancer. Proc Natl Acad Sci USA 1994; 91: 11045-9.
• 8 Lakshimi MS, Parker C, Sherbet GV. Metastasis associated MTS1 and nm23 genes affect tubulin polymerisation in B16 melanomas: a possible mechanism of their regulation of metastatic behaviour of tumor. Anticancer Res 1993; 13: 299-304.
• 9 Antonio M, Fiamma B, Silvia P, Gloria B, Antonio C, Vittoria C, et al. Alterations of p16 (mts1) in node-positive non-small cell lung carcinomas. J Pathol 1997; 181: 178-82.
• 10 Diaz-Flores L, Gutierrez R, Varela H. Angiogenesis: an uptate. Histol Histopathol 1994; 49: 807-83.
• 11 Woodhouse EC, Chuaqui RF, Liotta LA. General mechanisms of metastasis. Cancer (supple) 1997; 80: 1529-37.
• 12 Yang SM, Chen F, Zhang CG, Gu XX. Establishment of a model of tumor gastric cancer with liver metastasis in nude mice. Chin J Oncol 1993; 15: 195-7.
• 13 Chen YQ, Wu Q, Chen ZM, Su WJ, Chen F. Effect of retinoic acid on metastatic ability of gastric cancer cells MGC80-3 in vivo and in vitro. World Chin J Digestol 1998; 6: 869-72.
• 14 Majewski S, Marczak M, Szmurlo A. Retinoids, interferon alpha, 1,25-dihydroxyvitamin D-3 and their combination inhibit angiogenesis induced by non-HPV-harboring tumor cell lines. Cancer Lett 1995; 89: 117-24.
• 15 Hsu JW, Hsu SL, Chu JJ, Liu MC, Chiang CD. Increased nm23:mts1 ratio inversely correlated with metastasis behaviour in human lung squamous cell carcinoma. Anticancer Res 1997; 17: 407-12.
• 16 Lin CH, Fu ZM, Liu YL. The establishment of human gastric cancer cell line, SGC-7901. Tumor 1981; 1: 1-3.
• 17 Wang Kaihua. An in vitro cell line (MGC80-3) of a poorly differentiated mucoid adenocarcinoma of human stomach. Acta Biol Exp Sin 1983; 16: 257-67.
• 18 Parket C, Whittaker PA, Weeks RJ, Thody AJ, Sherbet GV. Modulator of metastatic behaviour alter the expression of metastasis-associated genes mts1 and nm23 in metastatic variants of the B16 murine melanoma. Clin Biotech 1991; 3:217-22.
• 19 Webb DSA, Mostowski HS, Gerrard TL. Cytokine-induced enhamcement of ICAM-1 expression result in increased vulnerability of tumor cells to monocyte-mediated lesis. J Immunol 1991; 146: 3682-6.
• 20 Fawzi A, Marc B, Christian S, Gunter V, Marie A. Regulation of intercellular adhesion molecule-1 expression by retinoic acid:analysis of the 5' regulatory region of the gene. Int J Cancer 1994; 58: 543-9.
• 21 Wu Q, Dawson MI, Zheng Y, Hobbs PD, Agadir A, Jong L. Inhibition of trans-retinoic acid-resistant human breast cancer cell growth by retinoid X receptor-selective retinoids. Mol Cell Biol 1997; 17: 6598-608.