Evaluation of the role of soluble B7-H3 in association with membrane B7-H3 expression in gastric adenocarcinoma

Abstract

BACKGROUND and OBJECTIVE:

Gastric adenocarcinoma (GAC) is one of the most common malignancies. Increasing data have indicated a correlation between soluble B7-H3 (sB7-H3) levels and tumor malignancies. In this study, we aim to investigate the level of soluble B7-H3 in serum of GAC patients. Further, we analyze the correlation between sB7-H3 level and tissue B7-H3 expression and explore the clinical evaluation value of sB7-H3 associated with pathological characteristics and prognosis of GAC patients.

METHODS:

One hundred and twenty-eight serum and tissue samples of GAC, 20 serum and tissue samples of gastritis patients and 77 serum, 5 tissue samples of healthy controls were collected. The serum levels of sB7-H3 were detected by Enzyme-linked immunosorbent assay (ELISA), while the expression of membrane B7-H3 (mB7-H3) and Ki67 were evaluated by immunohistochemistry. The correlation between sB7-H3 and mB7-H3, sB7-H3 and Ki67, sB7-H3 or mB7-H3 and clinical features were analyzed by Pearson’s Chi-square test.

RESULTS:

Both serum level of sB7-H3 and tissue B7-H3 of GAC patients were significantly higher than those of gastritis patients and healthy controls. sB7-H3 level was correlated with total B7-H3 expression in tissues ( $r=$ 0.2801, $P=$ 0.0014). Notably, the concentration of sB7-H3 was correlated with its expression of membrane form in tumor cells ( $r=$ 0.3251, $P=$ 0.002) while not in stromal cells ( $r=$ 0.07676, $P=$ 0.3891). Moreover, the levels of sB7-H3 in patients with TNM stage III/IV or with infiltration depth T3/T4 or with lymph node metastasis were significantly higher than those of patients with TNM stage I/II ( $P=$ 0.0020) or with Infiltration depth T1/T2 ( $P=$ 0.0169) or with no lymph node metastasis ( $P=$ 0.0086). Tumor B7-H3 score, but not stromal B7-H3 score, in patients with TNM stage III/IV or with lymph node metastasis was significantly higher than those with TNM stage I/II ( $P=$ 0.0150) or with no lymph node metastasis ( $P=$ 0.182).

CONCLUSIONS:

Soluble B7-H3 level may reflect the tissue B7-H3 expression on tumor cells of GAC tissues. Elevated level of sB7-H3 in serum suggests poor clinical pathological characteristics of GAC patients.

1. Introduction

Gastric adenocarcinoma (GAC) is one of the most fatal malignancies, and was the third leading cause of cancer death worldwide [1]. Once infiltration is present, the survival rate would reduce to 31%, even 5% when transferred to distant organs. Early detection is critical for improving the prognosis of GAC patients. The 5-year survival rate is up to 68% if GAC patients are diagnosed and treated before the spread of cancer cells [2]. Thus, a potential indicator is urgent for clinical pathological characteristics and further prognosis judgments, which may be valuable for clinical evaluation.

B7-H3 is a member of the B7 family of costimulatory molecules, which shares 20–27% amino acid identity with other B7 family ligands [3]. The mRNA of human B7-H3 has been found to be overexpressed in various human tumor tissues but relatively limited in normal tissues [4]. A vast majority of studies have demonstrated that B7-H3 could induce tumor immune escape as well as enhance tumor metastasis [5, 6]. Aberrant expression of B7-H3 was correlated with poorer prognosis in several cancers such as hepatocellular cancer, breast cancer, non-small cell lung cancer, craniopharyngioma, renal cell carcinoma and gastric cancer [7, 8, 9, 10, 11, 12]. In gastric cancer, B7-H3 was expressed in both tumor and stromal cells. Tumor B7-H3 could promote gastric tumor cell migration, invasion, radiotherapy resistance and metastasis [13, 14]. In addition, stromal B7-H3 expression was related to tumor progression and lower intratumoral CD8+ T cell density [15, 16]. Several researchers believe that B7-H3 is a potential biomarker as well as a novel and valuable target for immunotherapy. Chimeric antigen receptor (CAR) T cells and bispecific antibodies targeting B7-H3 have been developed and applied in clinical trials in patients with B7-H3 positive cancers [4, 17, 18, 19, 20, 21, 22]. Therefore, it is valuable to investigate the expression pattern of B7-H3 for judging tumor progression as well as picking suitable treatment strategies.

Soluble B7-H3 lacks the transmembrane and cytoplasmic regions and could be detected in serum and plasma in some tumors and infectious diseases [4, 23]. High levels of soluble B7-H3 have also been detected in body fluids of several tumors patients and culture supernatants of tumor cell lines, and its high level expression was significantly correlated with the invasion and metastasis of tumors [24, 25, 26]. Early in 2008, Our group discovered the existence of soluble form of B7-H3 in human serum for the first time [23]. We further implicated the diagnosis value of sB7-H3 in non-small lung cancer [27]. However, the relationship between sB7-H3 level and corresponding tissue B7-H3 expression in GAC remains unknown. In the present study, we investigated the levels of sB7-H3 in the serum by ELISA in GAC patients and evaluated its correlation with B7-H3 expressed in tumor tissues. Further, we analyzed the significance of sB7-H3 with clinical features in GAC patients.

2. Materials and methods

2.1 Patients and plasma samples

In this study, 128 patients diagnosed with GAC were recruited as experiment group. Ninety-seven cases matched for sex and age were recruited as control group including 20 patients diagnosed with chronic atrophic gastritis via gastroscopy and 77 healthy volunteers. All samples of patients and healthy volunteers were obtained from the First Affiliated Hospital of Soochow University (Suzhou, Jiangsu, China). Patients’ characteristics were summarized in Table 1. One hundred and twenty-eight GAC tissues were derived from surgery resection. At the same time, the clinical and pathological data including gender, age, differentiation, tumor stage, tumor depth, lymph node metastasis were retrieved. None of the patients received pre-operative chemotherapy or radiotherapy. Twenty chronic atrophic gastritis and 5 healthy control tissues which diagnosed by pathologists were derived from gastroscopy. The tissue and blood samples from the subjects were provided according to a protocol approved by the local ethics committee of Soochow University.

Table 1

Characteristics of cases of GAC, gastritis and healthy control

Characteristic	Patients number (%)
	GAC	Gastritis	Healthy control
Total cases	128	20	77
Age (years)
Median	59	53	58
Range	38–88	30–77	39–86
Gender
Male	89 (69.5%)	14 (70%)	52 (67.5%)
Female	39 (30.5%)	6 (30%)	25 (32.5%)
Differentiation
Well	3 (2.3%)	–	–
Moderately	59 (46.1%)	–	–
Poorly	66 (51.6%)	–	–
Tumor stage
I $+$ II	44 (34.4%)	–	–
III $+$ IV	84 (65.6%)	–	–
Tumor depth
T1 $+$ T2	30 (23.4%)	–	–
T3 $+$ T4	98 (76.6%)	–	–
LN involvement
N0	37 (28.9%)	–	–
N1	91 (71.1%)	–	–
sB7-H3 (ng/ml)	9.52 $\pm$ 4.10	7.47 $\pm$ 2.33	6.96 $\pm$ 1.83

2.2 Enzyme-linked Immunosorbent Assay (ELISA)

The levels of sB7-H3 in serum were measured using human B7-H3 quantikine ELISA kit (R&D systems) according to the manufacturer’s instruction. In brief, the serum samples were transferred to a 96 well plate and incubated for 2 hours at room temperature. After washed 3 times, 100 $\mu$ L of human B7-H3 conjugate were added and incubated for 2 hours at room temperature. 50 $\mu$ L stop solutions were then added to end the reaction and the absorbance was measured at 450 nm.

2.3 Immunohistochemistry

Immunohistochemistry was performed by the two-step EnVision method (Dako, Glostrup, Denmark). In brief, the 4 $\mu$ m paraffin-embedded sections were dried at 65 ${}^{\circ }$ C for 2 hours. The tissues were deparaffinized with xylene and rehydrated though graded ethanol solution. Endogenous peroxidase was blocked with 0.3% hydrogen peroxide solution for 15 minutes at room temperature. The sections were pretreated with citrate buffer (pH 6.0) by boiling at 110 ${}^{\circ }$ C for 3 minutes. After being washed three times with phosphate buffered saline (PBS) for 5 minutes each time, the section was blocked with 10% normal horse serum followed by incubation with goat anti-human B7-H3 antibody (Cat.No: AF1027 10 $\mu$ g/mL; R&D systems, USA) or 1:200 dilution of primary Ki67 (Cat.No: ab237863, Abcam, UK) overnight at 4 ${}^{\circ }$ C. Then the slides were washed three times with PBS, followed by incubation with horseradish peroxidase (HRP)-conjugated secondary antibody (GP016029 of anti-goat or GP016129 of anti-mouse, Genetech, China) for 30 minutes. The samples were washed twice with PBS followed by immunodetection using the Dako EnVision detection system. The negative control was performed by replacing the primary antibody with mouse or goat IgG.

2.4 Assessment of tissue staining

All immunohistochemical studies were independently and blindly reviewed by two pathologists and scanned with a Dmetrix imaging system (Dmetrix, Tucson, AZ, USA). The staining histoscore of the slides was classified as 0–3 (0, no staining; 1, $<$ 10% staining; 2, 10–50% staining; and 3, $>$ 50% staining). The histoscore of staining quantification were assessed according to the formula (3 $+$ percent cells) $\times$ 3 $+$ (2 $+$ percent cells) $\times$ 2 $+$ (1 $+$ percent cells) $\times$ 1. This yielded the immunohistochemical score ranging from 0 to 300.

2.5 Statistical analysis

Statistical analysis was performed using SPSS (v22.0; IBM Corporation, USA). $T$ -test was applied to identify the statistical differences between sB7-H3 in GAC patients and gastritis or healthy volunteers. The correlation of sB7-H3 with mB7-H3 or Ki67 was analyzed using Pearson correlation. The difference of sB7-H3 level or mB7-H3 score with clinicopathological factors was analyzed by unpaired $t$ test. Statistical tests were two-sides, and $P$ value $<$ 0.05 was considered as statistically significant.

2.6 Ethics statement

This study was carried out in accordance with the recommendations of Clinical Research Ethics Committee of the First Affiliated Hospital of Soochow University. The protocol was approved by the Clinical Research Ethics Committee of the First Affiliated Hospital of Soochow University (approval number: 2018121). All subjects gave written informed consent in accordance with the Declaration of Helsinki.

Figure 1.

Comparison of serum sB7-H3 in healthy controls group, gastritis group and GAC group. Error bar is Mean $\pm$ SEM. ( ${}^{***}$ , $P<$ 0.001, ${}^{*}$ , $P<$ 0.05, NS, no significant).

3. Results

3.1 Enhancement of serum levels of sB7-H3 and mB7-H3 in GAC patients

We first investigated the levels of sB7-H3 in serum of 128 GAC patients 20 gastritis patients and 77 healthy controls (Fig. 1). The mean concentration of sB7-H3 in GAC patients was significantly higher than that in gastritis (9.524 $\pm$ 0.3622 ng/mL vs. 7472 $\pm$ 0.5213 ng/mL, $P=$ 0.038) and healthy controls (9.524 $\pm$ 0.3622 ng/mL vs. 6.959 $\pm$ 0.2088 ng/mL, $P<$ 0.001) We then assayed the expression of B7-H3 in above three groups by immunohistochemistry (Fig. 2A). B7-H3 was highly expressed in GAC tissue, but hardly in gastritis and control tissues (Fig. 2B). The difference was statistically significant between GAC and gastritis ( $P<$ 0.001) or between GAC and control ( $P<$ 0.001).

Figure 2.

Expression of mB7-H3 in healthy control group, gastritis group and GAC group. (A) Immunohistochemstry stain of each group. (B) Staining histoscore of each group. Error bar is Mean $\pm$ SEM. ( ${}^{***}$ , $P<$ 0.001). The scale bar represents 40 $\mu$ m.

3.2 Significant correlations between sB7-H3 in serum and mB7-H3 in tumor cell

As the expression of B7-H3 in tissues is closely related to tumor progression, we then analyzed the expression of B7-H3 by immunohistochemistry in GAC patients. Our results showed that B7-H3 was either expressed in tumor cells or stromal components, or both (Fig. 3A–D). The expression of B7-H3 was detected with a frequency of 69.5% (89/128) in GAC patients. Among them, 91.0% (81/89) had stromal B7-H3 expression while 49.4% (44/89) was expressed in tumor cells as well as 41.6% (37/89) was expressed in both tumor and stromal cells according to our previous methodology [15] (Fig. 3E). In order to explore the potential source and the clinical valuable of sB7-H3 the correlation between sB7-H3 and mB7-H3 was analyzed by chi-squared test. The results showed that the levels of sB7-H3 in serum was significantly correlated with the expression of total mB7-H3 in tissues of GAC patients ( $r=$ 0.2801, $P=$ 0.0014, Fig. 4A). Surprisingly, sB7-H3 was significantly correlated with mB7-H3 expression in tumor cells ( $r=$ 0.3265, $P=$ 0.002, Fig. 4B) but not in stromal components ( $r=$ 0.07676, $P=$ 0.3891, Fig. 4C). To understand the relationship between sB7-H3 and tumor cell proliferation, we analyzed the correlation between sB7-H3 level and Ki-67 expression. There was no significant correlation between them in GAC patients ( $r=$ 0.1453, $P=$ 0.1019, Fig. 4D).

Figure 3.

B7-H3 expression in GAC tissues. (A) B7-H3 negative expression, (B) tumor B7-H3 expression, (C) stromal B7-H3 expression and (D) both tumor and stromal B7-H3 expression, (E) Patient cases in various B7-H3 expression. The scale bar represents 20 $\mu$ m.

Figure 4.

The correlation between sB7-H3 level and mB7-H3 score or Ki67 expression. (A) sB7-H3 level and total B7-H3 score. (B) sB7-H3 level and tumor B7-H3 score. (C) sB7-H3 level and stromal B7-H3 score. (D) sB7-H3 level and Ki67 expression.

Figure 5.

sB7-H3 level of GAC serum and various mB7-H3 expression of GAC tissue in association with cancer progression. (A) Serum sB7-H3 level (B) Total B7-H3 score (C) Tumor B7-H3 score and (D) Stromal B7-H3 score in samples with various pathological features. ( ${}^{**}$ , $P<$ 0.01, ${}^{*}$ , $P<$ 0.05, NS, no significant).

3.3 Prognostic value of soluble B7-H3 in GAC

To further determine the value of increased soluble form of B7-H3 in serum of GAC patients, the correlation between sB7-H3 concentration and clinical features, including tumor grade, infiltration depth of tumor and lymph metastasis were assessed. As shown in Fig. 5A, the level of sB7-H3 in serum of GAC patients was detected to be upregulated during the neoplastic stage. The sB7-H3 levels from patients with stage III/IV or with infiltration depth T3/T4 or with lymph node metastasis were distinctly higher than that from patients with TNM stage I/II (10.32 $\pm$ 0.5020 ng/mL vs 8.000 $\pm$ 0.3406 ng/mL, $P=$ 0.0020) or with infiltration depth T1/T2 (10.00 $\pm$ 0.4451 ng/mL vs 7.968 $\pm$ 0.4210 ng/mL, $P=$ 0.0169) or with no lymph node metastasis (10.11 $\pm$ 0.4653 ng/mL vs 8.014 $\pm$ 0.4035 ng/mL, $P=$ 0.0086). We further analyzed the mB7-H3 expression score in association with clinical features. Total B7-H3 score was higher in stage III/IV samples than in stage I/II samples ( $P=$ 0.0325, Fig. 5B). What’s more, tumor B7-H3 score was significantly higher in stage III/IV samples as well as in lymph node metastasis samples than in stage I/II samples ( $P=$ 0.0150, Fig. 5C) or in no lymph node metastasis samples ( $P=$ 0.0182, Fig. 5C). However, stromal B7-H3 score was no significantly changes during any clinical features (Fig. 5D). Thus, these results indicated that either high level of sB7-H3 in GAC serum or high tumor B7-H3 score in GAC tissue may lead to a poor progression of GAC.

4. Discussion

B7-H3, identified in 2001, is a cell surface molecule in B7 immunoglobulin super family which plays an important role in the initiation and termination of immune cell responses and cancer development. B7-H3 was overexpressed in gastric cancer and was regarded as a promising biomarker of diagnosis as well as a novel target for anti-tumor therapeutics [12, 14, 28]. Herein, we demonstrated that the soluble form of B7-H3 in serum was significantly upregulated in GAC than in gastritis and healthy controls. Furthermore, the serum levels of sB7-H3 were associated with membrane form of B7-H3 which was expressed in tumor cells but not in stromal cells. In addition, sB7-H3 levels were significantly correlated with clinical features including patients’ tumor grade, infiltration depth and lymph node metastasis. Thus, our results indicated that sB7-H3 could be a potential biomarker for clinical diagnosis and prognosis of GAC.

Prior literature have demonstrated the presence of B7-H3 in various stromal cells including endothelial cells, activated macrophages, fibroblasts, dendritic cells etc. In this study, we showed that the serum levels of sB7-H3 were positively correlated with mB7-H3 in tumor cells rather than stromal cells in GAC ( $r=$ 0.3265, $P=$ 0.002 vs $r=$ 0.07676, $P=$ 0.3891), suggesting that increased sB7-H3 in serum was mostly originated from tumor cells but not stromal components. These results also revealed that B7-H3 played a distinct role in tumor cells from that in stromal cells. Recently, we have reported that compared with stromal B7-H3, B7-H3 expressed in tumor cells could present higher correlation with poorer outcome of GAC [15]. B7-H3 expressed in cancer associated fibroblast promotes tumor progression by regulating the secretion of multiple cytokines as well as the polarization of T cells to Tregs [15, 29]. Given the different role of B7-H3 in tumor and stromal cells, the distinction may due to the expression of sB7-H3. We hypothesized that the form or modification of B7-H3 protein may result in a diversity of functions. Although the exact role of sB7-H3 remains unclear, its soluble form may lead to a more complicated and multiple function of the B7-H3. Further studies on biological functions of sB7-H3 may shed light into a new perspective to define the mechanism of the tumorigenesis and progression.

In addition, the level of sB7-H3 in serum was not statistically correlated with the expression of Ki67 but with worse prognosis, indicating that sB7-H3 had referential significance to the evaluation of tumor malignant degree but not proliferative activity of tumor cells in GAC. Increasing evidences also showed that B7-H3 could enhance tumor stem cell population and promote aggression and metastasis by targeting epithelial-mesenchymal transition in several cancers such as breast cancer, colorectal cancer and hepatocellular cancer [30, 31, 32]. Aberrant B7-H3 expression in tumor cells can also promote tumor progress through inhibiting anti-tumor immune response and several signaling pathways [4, 33, 34]. Consistent with that, sB7-H3 inhibited T cell proliferation and reduced the secretion of IL-2 and IFN- $\gamma$ in hepatocellular carcinoma [35]. Soluble B7-H3 could also induced angiogenesis via promoting the expression of VEGF and IL-8 [26]. Our previous study also demonstrated that sB7-H3 could bind to the B7-H3 receptor on activated T cells [23]. Therefore, the soluble form of B7-H3 may play as an active form in tumors and be critical in tumor progression.

As the overall prognosis in advanced-stage or with metastasis was poor, dynamic and timely monitor of tumor progression would improve the prognosis of patients with GAC. This study showed that sB7-H3 expression was positively correlated with TNM grade, infiltration depth and lymph node metastasis in GAC patients. Similar results were demonstrated in osteosarcoma, early-stage hepatocellular carcinoma and non-small-cell lung cancer [24, 27, 36, 37]. Soluble B7-H3 level was significantly correlated with lymph node metastasis in this study. A recent study showed that sB7-H3 promoted metastasis of pancreatic carcinoma cells through the TLR4/NF- $\kappa$ B pathway [26]. Further, blood specimens from patients with gastric cancer contained significantly more copies of B7-H3 mRNA than healthy controls [38]. Considering the positive correlation of serum sB7-H3 and B7-H3 in tumor cells, the tumor cells with high B7-H3 expression may preferentially transfer or drop from the primary site to circulate tumor cells. Therefore, soluble form of B7-H3 in serum, which is more convenient to obtain, might provide application for prediction of invasion and metastasis in GAC.

In general, our present findings suggested that the levels of sB7-H3 in serum were enhanced in GAC patients. Serum sB7-H3 levels were positively correlated with B7-H3 in tumor cells. We also found that the expression of sB7-H3 was correlated with tumor TNM grade, infiltration depth and lymph node metastasis. Last, sB7-H3 in serum could serve as a potential biomarker for tumor diagnosis and prognosis, as well as provide a new strategy for GAC clinical treatment.