The Study On TGF-β/Smads Signal Pathway Genes In Gastric Cardia Adenocarinoma

Transforming growth factor-β(TGF-β) plays important role in the regulation of numerous cellular functions including cell proliferation, differentiation, adherence, apoptosis, extracellular matrix metabolism, inflammatory reaction, interstitial fibrosis, immunity and cancer development. Abnormality of any genes in TGF-β/Smads signal pathway may lead to disorder of signal transmission and subsequently influence the biological effects of TGF-β. However, the mechanisms of genes inactivation in the TGF-β/Smads signal pathway may be different for the different tumor types. A series of epidemiological studies have demonstrated that the incidence of gastric cardia adenocarcinoma (GCA) kept increasing in very recent years, however, the exact mechanisms of the occurrence of GCA still remains unclear for the moment. It is essential to study the effect of the genes in TGF-β/Smads signal pathway for revealling the mechanism of GCA occurrence and development.In our study, RT-PCR and immunohistochemistry methods were used respectively to detect the mRNA and protein expression of TGF-β1, TGFBR1, TGFBR2, Smad2, Smad3, Smad4 and Smad7 in GCA; the TGF-β1 C-509T, G-800A, C-988A, T869C, G915C and C788T polymorphism sites, TGFBR1 Int7G24A and *6A sites, and TGFBR2 G-875A site were analyzed respectively in GCA and healty controls to show the relationship of these single nucleotide polymorphisms (SNP) with susceptibility of GCA; the methylation frequencies of CpG islands of TGFBR1, TGFBR2 and Smad4 were detected in GCA and dysplasia and the association of these genes methylation with mRNA and protein expression was further analyzed. The main research contents and results were shown as follows:PartⅠ:Expression of critical genes of TGF-β/Smads signal pathway in GCAObjective:To investigate the variation of critical genes of TGF-β/Smads signal pathway in transcription and protein level, including TGF-β1 and receptors (TGFBR1 and TGFBR2), Smads (Smad2, Smad3, Smad4 and Smad7).Methods:RT-PCR method was used to detect the mRNA expression of TGF-β1, TGFBR1, TGFBR2, Smad2, Smad3, Smad4 and Smad7 in GCA, immunohistochemistry method was used to detect the protein expression of TGF-β1, TGFBR1, TGFBR2, p-Smad2/3, Smad4 and Smad7 in GCA.Results:1 Expression of TGF-β1 mRNA in GCA tumor tissues was significantly increased with comparison to the paired normal tissues (0.8494±0.3024 vs. 0.3879±0.1754, P<0.01). Positive protein expression of TGF-β1 in GCA tumor tissues was also significantly increased with comparison to the paired normal tissues (P<0.01). TGF-β1 mRNA and protein expression in stage III andⅣGCA patients was significantly higher than that in stage I and II GCA patients, and TGF-β1 mRNA and protein expression in poor differentiation group was significantly higher than that in well and moderate differentiation groups (P<0.05).2 Expression of TGFBR1 mRNA in GCA tumor tissues was significantly reduced with comparison to the paired normal tissues (0.4015±0.1758 vs. 0.9064±0.2925, P<0.01). Positive protein expression of TGF-β1 in GCA tumor tissues was also significantly decreased with comparison to the paired normal tissues (P<0.01). Expression of TGFBR1 mRNA and protein in stageⅢandⅣGCA patients was significantly lower than that in stage I and II GCA patients (P<0.05).3 Expression of TGFBR2 mRNA in GCA tumor tissues was significantly reduced with comparison to the paired normal tissues (0.3879±0.1205 vs. 0.8917±0.2765, P<0.01). Positive protein expression of TGF-β1 in GCA tumor tissues was also significantly decreased with comparison to the paired normal tissues (P<0.01). Expression of TGFBR2 mRNA and protein in stage ⅢandⅣGCA patients was significantly lower than that in stageⅠandⅡGCA patients (P<0.05). The protein expression of TGFBR2 among the three differentiation groups also showed significant difference (P<0.05).4 Expression of Smad2 and Smad3 mRNA in GCA tumor tissues was significantly reduced with comparison to the paired normal tissues (0.4956±0.1862 vs.0.8611±0.2914 and 0.4713±0.1712 vs.0.8314±0.2811, respectively, P<0.01). The positive protein expression of p-Smad2/3 in GCA tumor tissues (42.7%) was significantly lower than that in paired normal tissues (93.6%) (P<0.01). Protein expression of p-Smad2/3 in stage III and IV tumor tissues was significantly lower than that in stage I and II tumor tissues. The protein expression of p-Smad2/3 among the three differentiation groups also showed significant difference (P<0.05).5 Expression of Smad4 mRNA in GCA tumor tissues was significantly reduced with comparison to the paired normal tissues (0.5145±0.1987 vs. 0.8954±0.2856, P<0.01). The positive protein expression of Smad4 in GCA tumor tissues was significantly lower than that in paired normal tissues (P<0.01). Protein expression of Smad4 in stageⅢandⅣtumor tissues was significantly lower than that in stageⅠandⅡtumor tissues, and the protein expression of Smad4 among the three differentiation groups also showed significant difference (P<0.05).6 Expression of Smad7 mRNA in GCA tumor tissues was significantly increased with comparison to the paired normal tissues (0.5114±0.1962 vs. 0.2012±0.1006, P<0.01). The positive protein expression of Smad7 in GCA tumor tissues was significantly higher than that in paired normal tissues (P<0.01). Protein expression of Smad7 among the three differentiation groups showed significant difference (P<0.05).7 Association analysis of the protein expression of critical genes of TGF-β/Smads pathway in GCA showed that TGFBR1 and TGFBR2, p-Smad2/3 and Smad4, TGF-β1 and Smad7, TGFBR1 and p-Smad2/3 were positively correlated (r=0.74,0.84,0.24 and 0.38, respectively, P<0.05); TGF-β1 and TGFBR1, TGF-β1 and p-Smad2/3, TGFBR1 and Smad7 were inversely correlated (r=-0.54,-0.49 and -0.22, respectively, P<0.05); p-Smad2/3 and Smad7 did not show any correlation (r=-0.09, P>0.05).Conclusions:1 Overexpression of TGF-β1 and Smad7 mRNA and protein in GCA may be associated with infiltration and metastasis of GCA.2 Decreased expression of TGFBR1, TGFBR2, Smad2, Smad3 (p-Smad2/3) and Smad4 mRNA and protein in GCA may be associated with the occurrence of GCA.3 Association analysis showed that protein expression of TGFBR1 and TGFBR2, p-Smad2/3 and Smad4, TGF-β1 and Smad7, TGFBR1 and p-Smad2/3 in GCA were positively correlated; TGF-β1 and TGFBR1, TGF-β1 and p-Smad2/3, TGFBR1 and Smad7 were inversely correlated; p-Smad2/3 and Smad7 did not show any correlation, suggesting that any genes variation of the pathway may lead to the lost of sensitivity to TGF-β1 in GCA.PartⅡ:Polymorphism of TGF-β1 and its receptor 1 and 2 in GCA Objective:To investigate the relationship of TGF-β1 C-509T, G-800A, C-988A, T869C, G915C and C788T, TGFBR1 Int7G24A and *6A, TGFBR2 G-875A SNP with susceptibility to GCA in north China. Methods:Polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP), polymerase chain reaction-primer introduced restriction analysis (PCR-PIRA), polymerase chain reaction-amplification refractory mutation system (PCR-ARMS) and PCR product length analytical methods were adopted respectively to investigate the genotype of these genes; enzyme linked immunosorbent assay (ELISA) was used to detect the level of TGF-β1 in GCA and healthy controls; and terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) method was used to determine the apoptotic rate of tumor-infiltrating lymphocytes (TILs) in GCA tumor tissues.Results: 1 The proportion of positive family history of UGIC in GCA patients (36.8%) was significantly higher than that in healthy controls (5.1%) (P<0.01) which indicated that family history of UGIC significantly increased the risk of developing GCA (the age and sex adjusted OR=10.61,95% CI=7.64-12.28).2 T allelotype of C-509T in GCA patients (52.4%) was significantly higher than that in control group (44.0%) (P<0.01). Compared with CC genotype, CT or TT genotype significantly increased the risk of GCA (adjusted OR=1.67 and 1.89,95%CI=1.17-2.45 and 1.25-2.71, respectively). Compared with CC genotype, CT or TT genotype increased the risk of GCA in non-smoker group (adjusted OR=1.97 and 2.33,95%CI=1.28-2.87 and 1.61-3.01, respectively), in negative family history of UGIC group (adjusted OR=1.76 and 1.97, 95%CI=1.25-2.67 and 1.10-2.89, respectively) and stage III and IV GCA (adjusted OR=1.94 and 2.35,95%CI=1.40-2.57 and 1.67-2.89, respectively).3 C allelotype of T869C in GCA patients (53.3%) was significantly higher than that in control group (45.5%) (P<0.01). Compared with TT genotype, TC or CC genotype significantly increased the risk of GCA (adjusted OR=1.65 and 1.85,95%CI=1.15-2.41 and 1.21-2.69, respectively). Compared with TT genotype, TC or CC genotype increased the risk of GCA in non-smoker group (adjusted OR=1.98 and 2.07,95%CI=1.31-2.89 and 1.45-2.91, respectively), in negative family history of UGIC group (adjusted OR=1.72 and 1.87, 95%CI=1.23-2.61 and 1.19-2.78, respectively) and stageⅢandⅣGCA (adjusted OR=1.90 and 2.07,95%CI=1.47-2.56 and 1.54-2.69, respectively).4 The genotype and allelotype distributions of TGF-β1 G-800A, C-988A, G915C and C788T in GCA patients were not significantly different from that in healthy controls (P>0.05), which indicated that the four SNP sites of TGF-β1 were not correlated with the risk of GCA in north China.5 The C-509T and T869C SNP are in strong linkage disequilibrium (D'=0.94). Haplotype analyses showed that compared with -509C/869T haplotype,-509T/869C haplotype significantly increased the risk of GCA (adjusted OR=1.92,95%CI=1.38-2.59).6 The level and positive protein expression of TGF-β1 in GCA patients with-509CC genotype was significantly lower than that in GCA patients with TT or CT+TT genotype. The level and positive protein expression of TGF-β1 in GCA patients with 869TT genotype was significantly lower than that in GCA patients with CC or TC+CC genotype. The mean percentage (±SD) of apoptotic TILs among total CD3+ lymphocytes (apoptotic index) in GCA tumors with TGF-β1 -509TT and 869CC genotype was significantly higher than that in GCA tumors with -509CC and 869TT genotype (29.8±1.8% vs. 17.3%±1.7%, P<0.05), whereas the values were not significantly different between the -509TT,869CC genotype and -509CT,869TC genotype (25.4±1.5%).7 A allelotype of TGFBR1 Int7G24A significantly elevated the risk of developing GCA (adjusted OR=1.34,95%CI=1.03-1.87). Compared with GG genotype, the GA and AA genotype significantly elevated the risk of developing GCA (adjusted OR=1.34,95%CI=1.02-1.73). Compared with GG genotype, GA and AA genotype significantly increased the risk of stageⅢandⅣGCA (adjusted OR=1.41,95%CI=1.05-1.98). The genotype and allelotype distributions of TGFBR1 *6A polymorphism in GCA patients were not significantly different from that in healthy controls (P>0.05). The TGFBR1 protein expression was not correlated with genotypes of Int7G24A and *6A in GCA.8 A allelotype of TGFBR2 G-875A significantly decreased the risk of developing GCA (adjusted OR=0.73,95%CI=0.49-0.92). Compared with GG genotype, GA and AA genotype significantly decreased the risk of GCA (adjusted OR=0.71,95%CI=0.52-0.95). Compared with GG genotype, GA and AA genotype significantly decreased the risk of stageⅢandⅣGCA (adjusted OR=0.70,95%CI=0.47-0.92). The TGFBR2 protein expression was not correlated with genotype of G-875A in GCA.9 Compared with protective genotypic combination -509CC -Int7GG--875AA, only one site carrying predisposing gene (heterozygote or homozygote) has 1.31-fold higher risk of developing GCA (adjusted OR=1.31, 95%CI=1.02-1.58), when two sites simultaneous existing predisposing genes, the risk of developing GCA will increase 1.54 times (adjusted OR=1.54, 95%CI=1.21-1.85), three sites simultaneous existing predisposing genes, the risk of developing GCA will increase 2.05 times (adjusted OR=2.05, 95%CI=1.52-2.61). Tendency analysis showed that the more predisposing allele, the higher risk of GCA development (χ2=21.843, P<0.01).Conclusions:1 UGIC family history significantly increased the risk of developing GCA.2 Compared with -509CC genotype, CT or TT genotype significantly increased the risk of GCA and was more evident in stageⅢandⅣGCA. Compared with 869TT genotype, TC or CC genotype significantly increased the risk of GCA and was more evident in stageⅢandⅣGCA. GCA patients with -509CC or 869TT genotype showed decreased TGF-β1 level, protein expression and apoptotic TILs. The C-509T and T869C SNP are in strong linkage disequilibrium. Compared with -509C/869T haplotype,-509T/869C haplotype significantly increased the risk of GCA.3 The genotype and allelotype distributions of TGF-β1 G-800A, C-988A, G915C and C788T in GCA patients were not significantly different from that in healthy controls,4 Compared with TGFBR1 Int7G24A GG genotype, the GA and AA genotype significantly elevated the risk of developing GCA and was more evident in stageⅢandⅣGCA. The genotype and allelotype distributions of TGFBR1 *6A polymorphism in GCA patients were not significantly different from that in healthy controls. The protein expression of TGFBR1 was not correlated with genotypes of Int7G24A and *6A in GCA.5 Compared with TGFBR2 G-875A GG genotype, GA and AA genotype significantly decreased the risk of GCA, however, the protein expression of TGFBR2 was not correlated with genotypes of G-875A in GCA.6 Compared with protective genotypic combination -509CC -Int7GG--875AA, the more predisposing allele, the higher risk of GCA development.PartⅢ:Methylation status of TGF-βreceptor 1,2 and Smad4 in GCAObjective:To detect methylation status of TGFBR1, TGFBR2 and Smad4 in GCA and dysplasia, and provide possible mechanism of the three genes inactivation in GCA occurrence and development. Methods:MSP method was adopted to investigate the methylation status of TGFBR1, TGFBR2 and Smad4 in GCA and dysplasia tissues.Results:1 The frequencies of TGFBR1 methylation in the tissues of normal, low grade dysplasia, high grade dysplasia and GCA were 1.8%(2/110),3.8% (1/26),11.4%(4/35) and 30.9%(34/110), respectively. The frequencies of TGFBR1 methylation in high grade dysplasia and GCA tumor tissues were significantly higher than that in normal tissues (P<0.05). Methylation frequency of TGFBR1 in stageⅢandⅣtumor tissues was significantly higher than that in stageⅠandⅡtumor tissues (P<0.05).2 Expression of TGFBR1 mRNA in GCA tumor tissues with positive methylation of the gene was significantly lower than that in GCA tumor tissues with negative methylation of the gene (0.1816±0.0816 vs.0.6514±0.2187, P<0.01). All of the GCA tumor tissues with TGFBR1 methylation demonstrated negative protein expression of the gene. A close correlation was noted between TGFBR1 methylation and the loss of protein expression in GCA(P<0.01).3 The methylation frequency of -25 and -140 sites in TGFBR2 was identical. The frequencies of TGFBR2 methylation in the tissues of normal, low grade dysplasia, high grade dysplasia and GCA were 3.6%(4/110),7.7%(2/26), 25.7%(9/35) and 47.3%(52/110), respectively. The frequencies of TGFBR2 methylation in high grade dysplasia and GCA tumor tissues were significantly higher than that in normal tissues (P<0.01). Methylation frequency of TGFBR2 in stage III and IV tumor tissues was significantly higher than that in stage I and II tumor tissues (P<0.05). The TGFBR2 methylation among three differentiation groups also showed significant difference (P<0.05), and methylation frequency of TGFBR2 in poor differentiation group was higher than that in moderate and well differentiation groups.4 Expression of TGFBR2 mRNA in GCA tumor tissues with positive methylation of the gene was significantly lower than that in GCA tumor tissues with negative methylation of the gene (0.1745±0.0756 vs.0.5894± 0.2061, P<0.01). Forty-seven of the 52 TGFBR2 methylated GCA tumor tissues showed negative protein expression of the gene. A close correlation was noted between TGFBR2 gene methylation and the loss of protein expression in GCA (P<0.01).5 Six of 110 GCA tissues (5.5%) were detected methylation of Smad4 in the-248～26 sites, while no dysplasia and paired normal tissues were detected Smad4 methylation in this region. The methylation frequency of Smad4 in GCA tumor tissues was significantly higher than that in normal tissues (P<0.05). For the 30～171 sites, the methylation frequencies of Smad4 in the tissues of high grade dysplasia and GCA were 8.6%(3/35) and 24.5% (27/110), respectively. No low grade dysplasia and paired normal tissues were detected Smad4 methylation in this region. The frequencies of Smad4 methylation in high grade dysplasia and GCA tumor tissues were significantly higher than that in normal tissues (P<0.05). Methylation frequency of Smad4 in stageⅢandⅣtumor tissues was significantly higher than that in stageⅠandⅡtumor tissues (P<0.05). The Smad4 methylation among three differentiation groups also showed significant difference (P<0.05), and methylation frequency of Smad4 in poor differentiation group was higher than that in moderate and well differentiation groups.6 For -248～26 sites, mRNA and protein expression of Smad4 in the GCA tumor tissues with positive methylation of Smad4 was not significantly different from that in the GCA tumor tissues with negative methylation of Smad4 (P>0.05). For 30～171 sites, expression of Smad4 mRNA in the GCA tumor tissues with positive methylation of Smad4 was significantly lower than that in GCA tumor tissues with negative methylation of the gene (0.2987±0.0925 vs.0.6415±0.2814, P<0.01), and 24 of the 27 methylated GCA tumor tissues of the 30～171 sites showed negative protein expression. A close correlation was noted between 30～171 sites of Smad4 methylation and the loss of protein expression in GCA (P<0.01).7 When TGFBR1, TGFBR2 and Smad4 genes were combined to analyze, the frequencies of at least one gene methylation, simultaneous methylation of at least two genes and simultaneous methylation of three genes were progressively increased along with the increasing of TNM stage (P<0.05). Frequency of simultaneous methylation of at least two genes and simultaneous methylation of three genes was progressively increased along with the decreasing of histological differentiation (p<0.05).Conclusions:1 The frequencies of TGFBR1 methylation in high grade dysplasia and GCA tumor tissues were significantly higher than that in normal tissues, and the methylation status of TGFBR1 in GCA was associated with TNM stages. A close correlation was noted between TGFBR1 methylation and the loss expression of TGFBR1 mRNA and protein in GCA.2 The methylation frequency of -25 and -140 sites in TGFBR2 was identical. The frequencies of TGFBR2 methylation in high grade dysplasia and GCA tumor tissues were significantly higher than that in normal tissues, and the methylation status of TGFBR2 in GCA was associated with TNM stages and histological differentiation. A close correlation was noted between TGFBR2 methylation and the loss expression of TGFBR2 of mRNA and protein in GCA.3 The frequency of Smad4 methylation of the 30～171 sites in GCA was higher than that of -248～26 sites, and was associated with TNM stages and histological differentiation. A close correlation was noted between 30～171 sites methylation and the loss expression of Smad4 mRNA and protein in GCA.4 The combined analysis of the three genes suggested that simultaneous methylation of two or three genes of TGFBR1, TGFBR2, and Smad4 may be associated with progression and poor prognosis of GCA.