Expression of immune checkpoints in T cells of esophageal cancer patients

Inhibition of immune checkpoint proteins (checkpoints) has become a promising anti-esophageal cancer strategy. We here tested expressions of immune checkpoints in human esophageal cancers. Our results showed the expressions of many immune checkpoints, including CD28, CD27, CD137L, programmed death 1 (PD-1), T cell immunoglobulin mucin-3 (TIM-3), T cell Ig and ITIM domain (TIGIT), CD160, cytotoxic T lymphocyte antigen 4 (CTLA-4), CD200, CD137 and CD158, were dysregulated in peripheral T cells of esophageal cancer patients. Further, the expressions of PD-1, TIM-3 and TIGIT were upregulated in tumor infiltrating lymphocytes (TILs), which might be associated with TILs exhaustion. Meanwhile, the expressions of PD-1 and TIM-3 on CD4+ T cells were closely associated with clinic pathological features of esophageal cancer patients. These results indicate that co-inhibitory receptors PD-1, TIM-3 and TIGIT may be potential therapeutic oncotargets for esophageal cancer.


INTRODUCTION
The prognosis of esophageal cancer patient is still poor [1]. Immune checkpoint proteins (or checkpoints) are many inhibitory immune signalings that are vital for maintaining self-tolerance and dictating the duration or the amplitude of immune responses in peripheral tissues [2]. They are extremely important to minimize collateral tissue damages [2]. Existing evidences have shown that blockade of immune checkpoints could potently activate therapeutic anti-cancer immunity, which has become a promising anticancer (i.e. esophageal cancer) strategy [2].
Preclinical studies have demonstrated that antibodies that block the checkpoints could enhance antigen-specific T cell responses [2,3]. A fully humanized anti-cytotoxic T lymphocyte antigen 4 (CTLA-4) antibody, ipilimumab, has displayed a long-term survival benefit in patients with advanced melanoma [4]. Blocking programmed death 1 (PD-1) pathway could overcome immune resistance, and induce clinical responses in patients with melanoma, renal cell carcinoma and nonsmall cell lung cancer [5]. Therefore, immune checkpoints are promising immunotherapeutic targets for esophageal cancer treatment. The PD-1 pathway has been evaluated in esophageal cancer by immunohistochemistry (IHC), yet immune checkpoints have not been systematically tested in esophageal cancers [6].
In the current study, we showed that, as compared to normal donors, the expressions of several co-inhibitory receptors (or checkpoints), including PD-1, T cell immunoglobulin mucin-3 (TIM-3), T cell Ig and ITIM domain (TIGIT) and CD160, were significantly increased in a fraction of peripheral T cells of esophageal cancer patients. Meanwhile, the expression levels of PD-1, TIM-3 and TIGIT were significantly higher in tumor-infiltrating lymphocytes (TILs) than that in peripheral T cells in the cancer patients.  We first tested expressions of a set of immune  checkpoints, including CD200, PD-1, CD137L, CD273,  CD274, CTLA-4, TIM-3, CD137, CD158, CD160, B-and  T-lymphocyte attenuator (BTLA), CD28, CD27, TIGIT and CD278, in circulating CD4 + and CD8 + T cells that were derived from normal donors (n = 10) or esophageal cancer patients (n = 10, Figure 1). Results showed that expressions of PD-1, TIM-3 and CD158 in CD4 + T cells of esophageal cancer patients were significantly higher than that in healthy donors ( Figure 2). Similar results were also observed when analyzing PD-1, TIM-3, CD200 and CD27 expressions on CD4 + T cells through mean fluorescence intensity (MFI, Figure 2). No significant differences were observed when analyzing CD158, CD28 and TIGIT expressions on CD4 + T cells from cancer patients and healthy donors ( Figure 2). Intriguingly, MFIs of BTLA or CTLA-4 on CD4 + T cells from cancer patients were significantly higher than that from normal donors ( Figure 2).

Expressions of PD-1, TIM-3, TIGIT and BTLA in tumor-infiltrating lymphocytes (TILs) of esophageal cancer patients
Above results showed that PD-1, TIM-3, TIGIT and BTLA expressions were dysregulated on a fraction of peripheral blood T cells of esophageal cancer patients. We next assessed the expressions of PD-1, TIM-3, TIGIT and BTLA on CD4 + and CD8 + T cells isolated from esophageal cancer tissues, adjacent esophageal mucosa (AEM), and peripheral blood mononuclear cell (PBMC) from esophageal cancer patients ( Figure 3).
We showed that the expressions of PD-1 and TIM-3 on circulating CD4 + and CD8 + T cells from esophageal cancer patients (n = 35) were significantly higher than that from normal donors (n = 10, Figure  4A). The expression of TIGIT on cancer patients' (n = 35) circulating CD8 + T cells (not circulating CD4 + T cells) was significantly lower than that of normal donors (n = 10) ( Figure 4A). However, the difference in BTLA expression in circulating T cells between patients (n = 35) and normal donors (n = 10) was not statistically significant ( Figure 4A).

DISCUSSION AND CONCLUSIONS
Esophageal cancer is a highly lethal disease that lacks effective systemic treatment, making exploration of immunotherapy targets is extremely important. To our best knowledge, this study represents the first systematic effort  The error bars indicate SEM. "*" represents p < 0.05, "**" represents p < 0.01, "***" represents p < 0.001, "ns" represents no significantly different.
to characterize expression pattern of immune checkpoints in esophageal cancers.
We show that the expressions of co-stimulatory molecules CD28, CD27 and CD137L were downregulated on a fraction circulating T cells from esophageal cancer patients ( Figure 2). Yet, the expressions of co-inhibitory receptor PD-1, TIM-3, CD160 and CTLA-4 were upregulated (Figure 2). These observations indicated that circulating T cells may present with an immune suppressive phenotype in esophageal cancer patients. However, the expressions of coinhibitory receptor CD200 and TIGIT were downregulated, and the expressions of co-stimulatory receptor CD137 and CD158 were upregulated on a fraction of circulating T cells from esophageal cancer patients (Figure 2).
Our results demonstrated that both PD-1 and TIM-3 were up-regulated on peripheral T cells and TILs from esophageal cancer patients, and that approximately half of TILs were PD-1 + TIM-3 + , a 26 to 100-fold increase compared to peripheral T cells of normal donors ( Figure 6). The level of PD-1 + TIM-3 + peripheral T cells of esophageal cancer patients was increased 18 to 40-fold compared to those from normal donors ( Figure 6). Moreover, our observations showed statistically significant positive correlations between PD-1 and TIM-3 expression in T cells in PBMC and tumor tissue ( Figure 5A and 5B). Similar results were also observed when analyzing the correlations between PD-1 and TIGIT expressions on T cells from esophageal cancer patients ( Figure 5C). It has been demonstrated that TIGIT directly inhibits T cell activation [7,8]. Therefore, our results suggest that PD-1, TIM-3 and TIGIT expressions on T cells from esophageal cancer patients would be in a co-expression pattern, which may cause T cell exhaustion.
In recent clinical cancer studies, the concurrent blockage of PD-1 and TIM-3 pathways exhibited better anti-cancer efficiency [9,10]. Studies have shown that TIGIT is over-expressed on human and murine TILs. Dual blockade of TIGIT and PD-1 pathway additively increased proliferation, cytokine production, and de-granulation of tumor antigenspecific CD8 + T cells and CD8 + TILs [11,12]. Our observations exhibited that the expression of PD-1 on circulating CD4 + T cells was positively correlated with histological grade of patients, and negatively correlated with tumor size and lymph node status of patients (Table 1). Moreover, the expressions of PD-1 and TIM-3 on CD4 + TILs were significantly associated with TNM stage of patients (Table 1). Studies have indicated that how PD-1 expression could affect T cell functions [13]. Recent data has suggested that TIM-3 positive CD4 + T cells could represent as functional regulatory T cells in human tumors [14]. However, due to the limitation of this small sample study, further studies will be needed to evaluate potential capacity of PD-1 and TIM-3 as prognostic factors, and to dissect the mechanism of PD-1 and TIM-3 pathways in esophageal cancer. In conclusion, our observations suggest that the coinhibitory receptors PD-1, TIM-3 and TIGIT could be the immunotherapeutic targets in esophageal cancer.

Study subjects
Thirty esophageal cancer patients, hospitalized at Hai'an County People's Hospital (Hai'an, China), were enrolled in this study (Table 1). Detailed clinic pathologic data were summarized in Table 1. Patients received no immunotherapy or chemotherapy prior surgery. Ten informed consent healthy normal donors were recruited from Hai'an County People's Hospital. The study protocol was approved by the Ethics Committee at Hai'an County People's Hospital. Written-informed consent was obtained from each participant. All investigations were conducted according to the principles expressed in the Declaration of Helsinki.

Isolation of PBMC and TILs
A four mL peripheral blood sample was drawn from each health donor or esophageal cancer patient before surgery. The blood samples were centrifuged through a Ficoll-Paque Plus (GE, Shanghai, China) gradient. TILs were isolated by dissociating tumor tissue with the plunger portion of syringes on mesh sieves before centrifugation on a Ficoll-Paque Plus (GE) gradient. All investigations were conducted according to approval by the ethics committee and the principles expressed in the Declaration of Helsinki.

Statistics
Data were collected and were utilized to calculate the mean ± standard error (SEM). Statistical differences were analyzed by one-way ANOVA followed by multiple comparisons performed with post hoc Bonferroni test (GraphPad Prism software). A multivariable analysis was performed using a logistic regression model in order to explore the association between expression of different  represents p < 0.05, "**" represents p < 0.01, "***" represents p < 0.001, "ns" represents no significantly different. checkpoint proteins, or the association between checkpoint protein expression and patients' pathological features [15]. A step-down procedure method was selected. The criterion for variable removal was the likelihood ratio statistic based on the maximum partial likelihood estimates. Values of p < 0.05 were considered statistically significant.