The results revealed that there were 141 males and 37 females, with a median age of 62 years (range 34–38 years). The number of patients with the primary lesion site localized to the cardiac region, gastric fundus, gastric body and gastric antrum was 61, 14, 49 and 54, respectively. The histological types of tumor cells were divided into 4 categories based on the degree of tumor differentiation: well or moderately differentiated (82 cases), poorly differentiated (75 cases), mucinous adenocarcinoma (17 cases) and signet ring cell carcinoma (2 cases). Tumor stages for all patients were redefined according to pathological outcomes, which were based on the AJCC/UICC TNM classification revised in 2010. Five patients were staged I, 98 patients were stage II and 75 patients were staged III cancer.
Immunohistochemical analyses were performed to detect protein expression levels of Apelin in 53 of 178 normal gastric tissues adjacent to tumor tissues, while 88 of 178 tumor tissues were analyzed by immunohistochemical staining (29.78% vs 49.44%, P < 0.001, Table 1.1). Additionally, the immunoreactivity of Apelin exhibited a diffuse cytoplasmic immunostaining in tumor tissues (Fig. 1.1). Similarly, the percentage of VEGF-positive tumor cells was significantly higher in tumor tissues than in adjacent normal tissues, which was consistent with previous studies (55.62% vs 39.89%, P = 0.003, Table 1.1). Furthermore, immunoexpression of VEGF protein exhibited a cytoplasmic pattern consistent with Apelin (Fig. 1.2).
Table 1.1
Expression of Apelin and VEGF in gastric cancer tissues and adjacent tissues
| | cancer tissues | adjacent tissues | χ2 value | P value |
Apelin | positive | 88 | 53 | 14.386 | ༜0.001 |
negative | 90 | 125 |
VEGF | positive | 99 | 71 | 8.827 | 0.003 |
negative | 79 | 107 |
A chi-square test was used to investigate the correlations between clinicopathological factors and Apelin and VEGF expression. Apelin expression was significantly positively correlated with vessel invasion and lymph node metastasis (P < 0.01), (P < 0.01) (Table 1.2). In addition, significant positive correlations were observed between Apelin expression and late-staged tumor (T) status, pathological type, and nerve invasion (P < 0.05). Conversely, no significant associations were observed between Apelin expression and gender, age, and the site of primary lesion (P > 0.05). VEGF expression was significantly high during late T stage (P < 0.01), vessel invasion and positive N stage (P < 0.05); however, no statistically significant difference was observed between VEGF expression and gender, age, the site of primary lesion, pathological type and nerve invasion (P > 0.05).
Table 1.2
The relationship between Apelin and VEGF and clinicopathologic factors
Characteristics | Apelin | χ2 value | P value | VEGF | χ2 value | P value |
+(n) | -(n) | +(n) | -(n) |
Gender | | | 1.001 | 0.317 | | | 0.810 | 0.368 |
Male | 67 | 74 | | | 76 | 65 | | |
Female | 21 | 16 | 23 | 14 |
Age(years) | | | 0.840 | 0.772 | | | 1.585 | 0.208 |
≤ 62 | 46 | 49 | | | 57 | 38 | | |
༞62 | 42 | 41 | 42 | 41 |
site of primary lesion | | | 3.240 | 0.356 | | | 6.420 | 0.093 |
Cardia Fundus | 28 | 33 | | | 26 | 35 | | |
7 | 7 | 9 | 5 |
Body | 21 | 28 | 31 | 18 |
Antrum | 32 | 22 | 33 | 21 |
Pathological differentiation | | | 11.240 | 0.024 | | | 5.103 | 0.277 |
Well or moderately differentiated | 33 | 49 | | | 43 | 39 | | |
Poorly differentiated | 42 | 33 | 42 | 33 |
Mucinous adenocarcinoma | 11 | 6 | 10 | 7 |
Signet ring cell carcinoma | 0 | 2 | 2 | 0 |
Other | 2 | 0 | 2 | 0 |
Nerve invasion | | | 5.731 | 0.017 | | | 3.116 | 0.078 |
Yes | 49 | 34 | | | 52 | 31 | | |
No | 39 | 56 | 47 | 48 |
Vessel invasion | | | 7.910 | 0.005 | | | 5.807 | 0.016 |
Yes | 38 | 21 | | | 40 | 19 | | |
No | 50 | 69 | 59 | 60 |
T stage | | | 9.845 | 0.020 | | | 11.597 | 0.009 |
T1 | 1 | 1 | | | 1 | 1 | | |
T2 | 2 | 11 | 3 | 10 |
T3 | 58 | 63 | 64 | 57 |
T4 | 27 | 15 | 31 | 11 |
N stage | | | 14.433 | 0.002 | | | 8.653 | 0.034 |
N0 | 18 | 23 | | | 21 | 20 | | |
N1 | 13 | 26 | 16 | 23 |
N2 | 22 | 27 | 35 | 14 |
N3 | 35 | 14 | 27 | 22 |
A statistically significant difference was observed in the Apelin-positive subgroup with high MVD expression (33.086 ± 7.862; P < 0.05) compared to the Apelin-negative tumor tissues with low MVD expression (21.071 ± 6.320). Similarly, a statistically significant difference was observed between the VEGF-positive subgroup with high MVD expression (29.075 ± 8.193) and VEGF-negative subgroup with low MVD expression (19.638 ± 5.614; P < 0.05) (Table 1.3 and Fig. 1.3).
Table 1.3
The relationship between the expression of Apelin, VEGF and MVD
Characteristics | MVD | t | P |
Apelin | positive | 33.086 ± 7.862 | 2.917 | 0.013 |
negative | 21.071 ± 6.320 |
VEGF | positive | 29.075 ± 8.193 | 2.514 | 0.025 |
negative | 19.638 ± 5.614 |
The results of immunohistochemical analyses further revealed that the percentage of VEGF-positive expression, 72.7% (64/88) was significantly higher in Apelin-positive tumor cells than in Apelin-negative tumor cells, 38.9% (35/90). Furthermore, Apelin expression was positive in 64 (64.6%) of 99 VEGF-positive patients and 24 (30.3%) of 79 VEGF-negative patients. The results implied a significant positive correlation between Apelin protein expression and VEGF protein expression with relative risks of 0.856 (P < 0.01, Table 1.4).
Table 1.4
The association between Apelin and VEGF expression
Characteristics | Apelin | rs | P |
| | positive | negative | 0.856 | 0.004 |
VEGF | positive | 64 | 35 |
negative | 24 | 55 |
The OS rates of patients with positive expression of Apelin at 1 year, 3 years and 5 years were 97.73%, 62.50% and 28.41%, respectively (Fig. 1.4), whereas the OS rates of patients with negative expression of Apelin at 1 year, 3 years and 5 years were 97.78%, 76.67% and 53.33%, respectively. The median OS time for patients with positive expression of Apelin was 40.7 months and 60.2 months for patients with negative expression of Apelin and a statistically significant difference was observed between the two subgroups (χ2 = 12.549, P < 0.001). The PFS rates between patients with negative expression of Apelin and patients with positive expression of Apelin at 1 year was 94.44% vs 96.60%, at 3 years was 65.56% vs 44.32%, and at 5 years was 40.00% vs 19.32% (Fig. 1.5). A statistically significant difference in the median PFS (49.8 months vs 30.5 months, χ2 = 9.537, P = 0.002) was observed between the two subgroups. Overall, patients in the Apelin-negative subgroups had significantly longer OS and PFS intervals than patients in the Apelin-positive subgroups.
Subsequently, univariate Cox proportional hazards model was used to analyze the effects of various variables on OS. The analysis revealed that Apelin expression, VEGF expression, T stage, lymph node (N) stage, vessel invasion and histological type (all P < 0.05) excluding gender, age, the site of primary tumor or nerve invasion (all P > 0.05) were significantly associated with poor prognosis (Table 1.5). Multivariate analysis revealed that lymph node metastasis (P < 0.001), late-staged status (P = 0.008), poor differentiation (P = 0.027) and other histological types (including mucinous adenocarcinoma, signet ring cell carcinoma and small cell carcinoma) were independent prognostic factors of poor OS (P = 0.012) (Table 1.6).
Table 1.5
COX univariate analysis of factors associated with survival in 178 patients with gastric cancer
Characteristics | 5 years overall survival(OS) |
HR | 95%CI | P value |
Age(years) | | | 0.451 |
≤ 62 | 1.000 | Ref. | |
> 62 | 1.284 | 0.786–1.852 |
Gender | | | 0.423 |
Male | 1.000 | Ref. | |
Female | 0.936 | 0.605–1.633 |
Position | | | |
Cardia | 1.000 | Ref. | — |
Fundus | 0.972 | 0.112–6.668 | 0.503 |
Corpora | 0.774 | 0.276–1.942 | 0.137 |
Sinuses | 0.865 | 0.413–1.826 | 0.257 |
Pathological differentiation | | | |
Well/moderately differentiated | 1.000 | Ref. | — |
Poorly differentiated | 2.336 | 1.324–3.735 | 0.030 |
Other a | 4.413 | 1.041–12.546 | ༜0.001 |
Nerve invasion | | | 0.068 |
No | 1.000 | Ref. | |
Yes | 1.258 | 0.646–2.425 |
Vessel invasion | | | 0.025 |
No | 1.000 | Ref. | |
Yes | 2.369 | 0.983–4.337 |
T stages | | | 0.016 |
T1 + T2 | 1.000 | Ref. | |
T3 + T4 | 3.007 | 1.353–5.724 |
N stages | | | ༜0.001 |
N0 | 1.000 | Ref. | |
N + b | 4.591 | 2.081–8.852 |
Apelin | | | 0.001 |
Negative | 1.000 | Ref. | |
Positive | 3.984 | 1.276–6.824 |
VEGF | | | 0.027 |
Negative | 1.000 | Ref. | |
Positive | 2.322 | 1.020–5.316 |
a: Mucinous adenocarcinoma, signet ring cell carcinoma and small cell carcinoma. b: N1、N2、N3 |
Table 1.6
Multivariable analysis of the factors associated with survival in 178 patients with gastric cancer
Characteristics | 5 years overall survival(OS) |
HR | 95%CI | P value |
Apelin | | | 0.095 |
Negative | 1.000 | Ref. | |
Positive | 1.226 | 0.559–2.673 |
VEGF | | | 0.139 |
Negative | 1.000 | Ref. | |
Positive | 1.178 | 0.420–2.569 |
Pathological differentiation | | | |
Well/ moderately differentiated | 1.000 | Ref. | — |
Poorly differentiated | 2.605 | 1.228–4.947 | 0.027 |
Othera | 3.239 | 1.153–10.345 | 0.012 |
Vessel invasion | | | 0.069 |
No | 1.000 | Ref. | |
Yes | 1.427 | 0.698–3.566 |
T stages | | | 0.008 |
T1 + T2 | 1.000 | Ref. | |
T3 + T4 | 3.225 | 1.547–6.035 |
N stages | | | ༜0.001 |
N0 | 1.000 | Ref. | |
N + b | 3.987 | 1.836–8.148 |
a: Mucinous adenocarcinoma, signet ring cell carcinoma and small cell carcinoma. b: N1、N2、N3 |
Three cell lines including MGC-803, SGC-7901 and HGC-27 were used to evaluate the level of mRNA expression of Apelin in different GC cells in vitro. The results of RT-PCR analysis indicated that different cell lines presented varying Apelin expression levels, and a statistically significant difference was observed between each two groups (P < 0.05, Fig. 2.1). MGC-803 cells (Ct = 1.000 ± 0.344) exhibited the lowest Apelin expression levels, which represented approximately a 30-fold lower expression than in HGC-27 cells (Ct = 29.755 ± 5.067).
Western blotting was performed to assess the expression of Apelin at the protein level in MGC-803, SGC-7901 and HGC-27 cell lines. The lysates of MGC-803 cells had the lowest levels of immunoreactive Apelin based on the results of mRNA expression of Apelin. A significant difference was observed between the two groups (P < 0.05).
ELISA was performed to detect the concentration of cellular secretion of Apelin that was secreted by the examined cell lines into the culture medium. The results were similar to those observed in the protein and mRNA expressions of Apelin, accounting for the lowest concentration in MGC-803 cells (132.00 ± 31.97 pg/mL). In addition, a statistically significant difference was observed between each two cell lines (P < 0.05, Fig. 2.3).
The human Apelin expression vector, pcDNA3.1-Apelin vector was successfully transfected into MGC-803 cells, which were selected because they contain the lowest amounts of Apelin. Analyses revealed that the levels of Apelin mRNA, Apelin-encoding protein and the concentration of Apelin protein secreted in the culture medium considerably were increased in stably transfected MGC-803 cells compared to cells transfected with an empty vector (control). Moreover, the changes in Apelin expression levels exhibited a significant difference after transfection with pcDNA3.1-Apelin (Fig. 2.4, 2.5 and 2.6).
Cells were incubated with different concentrations of Apelin-13 for 24 hours, and the MTT assay subsequently performed to evaluate the effect of exogenous Apelin-13 on proliferation of MGC-803 cells. Exogenous Apelin-13 significantly ameliorated cell proliferation rate in a concentration-dependent manner, and the optimum concentration was 2.5 µmol/L (Fig. 2.7). The proliferation of MGC-803 cells that were treated with Apelin-13 (concentration of 2.5 µmol/L) for 48 hours exhibited a time-dependent relationship, presenting a significant and steady increase from 12 hours to 48 hours compared to the control group (Fig. 2.8).
The proliferation rate of MGC-803 cells transfected with Apelin-encoding pcDNA3.1-Apelin vector and incubated for 48 hours increased remarkably from 24 hours to 48 hours compared to cells transfected with empty vectors (control), with the absorbance detected at 570 nm by MTT analysis at different time intervals (Fig. 2.9)
Four types of cells were selected based on the results obtained from MTT assay as our experimental objects to evaluate migration and invasion abilities of MGC-803 cells, including untreated cells, Apelin-13 induced cells (incubated with a concentration of 2.5 µmol/L for 24 hours), mock-vector transfected cells (incubated for 24 hours after transfection with pcDNA3.1 vector) and pcDNA3.1-Apelin transfected cells (incubated for 24 hours after transfection with pcDNA3.1-Apelin vector ). Apelin-13 induced cells exhibited a significant improvement on the capacity to transfer to the membranes compared to untreated cells (t=-9.577, P = 0.001) (Fig. 2.10). A similar observation was also made between mock transfected cells and cells transfected with pcDNA3.1-Apelin (t=-5.142, P = 0.007).
The rate of apoptosis induced by SAHA was substantially increased in treated cells compared to the untreated cells (control) ((15.100 ± 3.997) % vs (7.300 ± 1.745) %, P < 0.05) (Fig. 2.11). Strikingly, the proportion of apoptotic cells ((14.000 ± 3.899) %) treated with both SAHA and Apelin-13, and the ratio of apoptotic cells ((14.425 ± 4.393) %) treated with both SAHA and pcNDA3.1-Apelin exhibited similar apoptotic ratios in the subgroups treated with SAHA. There was no significant correlation observed in groups treated with SAHA and SAHA + Apelin-13, SAHA and SAHA + pcNDA3.1-Apelin, respectively (t = 0.342, P = 0.774; t = 0.816, P = 0.446).
Cytokines were assessed using western blot assays to explore the underlying molecular mechanisms of Apelin-13 induction including pERK1/2, pAkt, cyclin D1 and MMP-9, and Apelin overexpression that are associated with biological behavior of MGC-803 cells. As previously mentioned, there were four subgroups, untreated cells, Apelin-13 induced cells (incubated with a concentration of 2.5 µmol/L Apelin for 24 hours), mock-vector transfected cells (incubated for 24 hours after transfection with pcDNA3.1 vector) and pcDNA3.1-Apelin transfected cells (incubated for 24 hours after transfection with pcDNA3.1-Apelin vector). By contrast, no significant correlation was observed in the protein levels of pAkt between the control and treated groups (P > 0.05), but the protein levels of other three cytokines, namely pERK1/2, cyclin D1 and MMP-9, were significantly increased relative to control cells after Apelin-13-induction and transfection with pcDNA3.1-Apelin (P < 0.05, Fig. 2.12).
We successfully established pLV-puro-Apelin and pLV-puro stable transformants through lentiviral mediation and puromycin screening with 100% transfection efficiency, namely pLV-puro-Apelin and pLV-puro. Real time RT-PCR analyses indicated that the mRNA expression level of Apelin increased considerably after transfecting pLV-puro-Apelin into MGC-803 cells compared to the control group (t=-4.236, P = 0.013, Fig. 3.1). Western blot analysis revealed that Apelin-encoding protein expression level was substantially enhanced by treatment with pLV-puro-Apelin transfected into MGC-803 cells in comparison to the control group (t=-3.833, P = 0.019) (Fig. 3.2).
We established a total of 14 gastric cancer xenografts formed in nude mice by subcutaneously injecting mice with MGC-803 cells transfected with pLV-puro-Apelin or MGC-803 cells transfected with pLV-puro. One nude mouse died three days after receiving a subcutaneous injection of MGC-803 cells transfected with pLV-puro-Apelin, and tumorigenesis was not observed in one nude mouse in the control group. The nude mice in the Apelin-overexpression group gradually exhibited decreased activity and poor nutritional status after injection, and the observation was terminated at 24 days after injection. Variables including (a) long diameter, (b) short diameter, and (c) weights of tumor of the nude mice were monitored twice a week by the same investigator. Furthermore, the tumor volumes (mm3) were calculated using the formula, V = ab2/2 and a tumor growth curve plotted. The neoplastic growth velocity in mice subcutaneously injected with pLV-puro-Apelin significantly increased from the fourteenth day (Fig. 3.3 and Table 3.1). A maximum body weight and tumor size of 2.96 ± 0.61 g was observed in the mice injected with MGC-803 cells of overexpressed-Apelin, whereas a minimum body weight and tumor size of 1.70 ± 0.43 g was observed in mice of the control group (Fig. 3.4). The results revealed the function of Apelin in promoting proliferation of gastric cancer cells via in vitro and in vivo experiments.
Table 3.1
The volume of gastric cancer xenografts in nude mice and statistical analysis
Time(day) | Control(± s) | Overexpression of Apelin(± s) | t value | P value |
7 | 22.656 ± 8.702 | 26.775 ± 11.745 | -0.630 | 0.543 |
10 | 61.517 ± 23.151 | 80.836 ± 26.352 | -1.232 | 0.246 |
14 | 106.115 ± 36.515 | 196.010 ± 66.854 | -2.639 | 0.025 |
17 | 306.789 ± 105.315 | 567.346 ± 198.854 | -2.598 | 0.027 |
21 | 748.919 ± 214.059 | 1399.063 ± 499.872 | -2.673 | 0.023 |
24 | 1367.348 ± 444.701 | 2557.645 ± 684.409 | -3.261 | 0.009 |
MVD assay by CD34 immunolabeling was performed to explore the mechanism underlying the differences in neoplastic growth velocity between over-expressed Apelin and control, and to establish the relationship between Apelin overexpression and angiogenesis. MVD was significantly higher in xenograft tumors with Apelin overexpression (168.833 ± 35.078) than in the control group, which had low MVD (112.333 ± 29.859; t=-2.734, P = 0.021) (Fig. 3.5).