Pancreatic cancer is a disease that seriously endangers human health, and methods for its early diagnosis and treatment are limited, resulting in poor prognosis[23]. Recent research on pancreatic tumor biology has focused on the TME[24–26], and an increasing number of studies have shown that the TME plays an important role in the occurrence and development of pancreatic cancer and tumor immunotherapy. Immunotherapy has achieved great results, but its efficacy is limited. Moreover, the molecular pathways and potential mechanisms of the TME in pancreatic cancer have not been fully elucidated[27], and new comprehensive treatment strategies are urgently needed.
In our study, transcriptome expression data (ESTIMATE algorithm) of PAAD patients were used to evaluate the immune score, stromal score, and ESTIMATE score for each tumor sample. The immune score, stromal score and comprehensive score were closely related to the clinicopathological characteristics of patients with pancreatic cancer. Subsequently, we evaluated the DEGs related to the immune score and the stromal score and determined the prognosis-related genes from the clinicopathological data, and we identified CXCL10 and CXCL11 as prognostic factors in the pancreatic cancer TME. Finally, the molecular mechanism by which CXCL10 and CXCL11 are involved in immune cell infiltration in pancreatic cancer was analyzed, and these results are anticipated to facilitate pancreatic cancer immunotherapy.
CXCL10 and CXCL11 are ELR-negative CXC chemokines that participate in the chemotaxis, differentiation and activation of peripheral immune cells by binding to the CXCR3 receptor[28, 29]. Some studies have shown that CXCL10 and CXCL11 can inhibit angiogenesis and exert antitumor effects[30]. However, CXCL10 and CXCL11 can also increase tumor proliferation and metastasis[31]. Recent studies have explored the role of CXCL10 and CXCL11 in pancreatic cancer. CXCL10 is overexpressed in human pancreatic cancer and is related to the poor survival of patients with PAAD[32]. The serum CXCL10 level in pancreatic cancer patients with lymph node metastasis is significantly higher than that in pancreatic cancer patients without lymph node metastasis[33]. CXCL10 can promote the migration of pancreatic cancer cells to sensory neurons and mediate the pain response in patients[34]. Interferon-γ (IFNγ) is the key cytokine in antitumor immunity and can affect a variety of cells in the TME of pancreatic cancer[32], Although the study of Huimin Huang demonstrated the role of CXCL10 in the immune infiltration of PAAD[35], it was not comprehensive enough in the prognostic biomarkers and immune infiltration mechanism, and the research content was relatively limited. Our study has detailly demonstrated the synergistic mechanism of CXCL10 and CXCL11 in the ELR-negative CXC chemokines gene family in the tumor microenvironment and immune infiltration of PAAD, which is more convincing and has potential research value, and further studies are needed. Furthermore, the research on CXCL11 in pancreatic cancer is still in the initial stage, related studies have only shown that CXCL11 is highly expressed in the serum of patients with pancreatic cancer and has a protumor function[36, 37]. Nevertheless, our research confirmed that CXCL10 and CXCL11 are important genes for the prognostic evaluation of pancreatic cancer. Most importantly, our study explored the mechanism of action of CXCL10 and CXCL11 in the TME and immune cell infiltration of PAAD and provided information anticipated to facilitate pancreatic cancer immunotherapy.
The composition and function of tumor-infiltrating immune cells are changed according to the immune status of the host[38, 39]. Tumor-associated macrophages (TAMs) are the most abundant immune cells infiltrating the TME. In the early stage of tumor development, tumor-infiltrating M1-polarized macrophages usually show a phenotype of high IL-12 and low IL-10 expression, promoting the immune response and causing lysis of tumor cells. During the development of advanced tumors, TAMs are usually polarized toward the M2 phenotype, promote tumor invasion and metastasis, and create a favorable microenvironment promoting tumor survival, growth and angiogenesis[40, 41]. The results of this study showed that the expression levels of CXCL10 and CXCL11 were negatively correlated with the level of M0 macrophages, positively correlated with the level of M1 macrophages and not correlated with the level of M2 macrophages, possibly because most of the PAAD samples included in this study were from tumors in the early stage of development, and M0 macrophages inhibit tumor-specific T cell immunity and enhance tumor growth. Or there is immune escape of M1 macrophages, which leads to the failure of M1 macrophages to exert their inhibition of tumor growth and tumor-induced angiogenesis. Therefore, further study of M1 macrophage immune escape mechanism is helpful for the treatment of PAAD.
Some studies have shown that B cells can induce and maintain beneficial antitumor activity, while others have found that B cells may play a protumor role due to their different immunosuppressive subtypes[42]. The results of this study showed that the content of memory B cells was inversely proportional to the expression levels of CXCL10 and CXCL11, indicating that the antitumor activity of B cells was inhibited in PAAD. CD8 + T cells, as one of the main antitumor effector immune cells, are activated by signaling from specific dendritic cells with the help of CD4 + T cells to CD8 + T cells in order to optimize the scale and quality of the cytotoxic T lymphocyte (CTL) response[43], but this study showed that the contents of these cells are proportional to the CXCL10 and CXCL11 expression levels. This relationship might be explained by the observation that when CD8 + T cells infiltrate into tumor tissue, they are usually in a dysfunctional state characterized by impaired activation and proliferation abilities; thus, the apoptosis rate is increased and the effect of cytokine production is reduced[44]. These dysfunctional CD8 + T cells are a barrier to successful cancer elimination. As immune suppressor cells, Tregs play a negative immunoregulatory role in the immune response and can inhibit the proliferation and activation of T cells[45]. However, in the current study, we found that the CXCL10 and CXCL11 expression level was negatively correlated with the Treg content, Lunardi S et al 's study show CXCL10 could recruit CD4+/CD8 + effector T cells and FoxP3C- Tregs, however, due to the large increase in circulating Tregs compared with effector T cells, CXCR3C Tregs may be preferentially recruited to inhibit the adaptive immune response (via effector T cells and NK cells)[46], thus establishing an immunosuppressive and protumor microenvironment that promotes poor prognosis in cancerous person[47], which contradicted with our results, suggesting that Tregs may have more complex immunomodulatory mechanisms that may be explained by the combination of specific immune-related signaling pathways of CXCL10 and CXCL11. Further study is needed to identify approaches to maintain or restore the tumor infiltration efficiency of CD8 + T cells to assist the development of tumor-specific CTLs in lymphoid organs, establish effective and durable antitumor immunity, or combine targeted Treg cells with the activation of tumor-specific effector T cells in order to develop novel therapeutic strategies to increase the efficacy of immunotherapy in PAAD.
In conclusion, our study showed that the immune score, stromal score and ESTIMATE score were significantly correlated with the sex and grade of pancreatic cancer patients. The results of the PPI network and Cox regression analyses showed that CXCL10 and CXCL11 are valuable factors involved in the TME of pancreatic cancer and are correlated with the prognosis and clinicopathological characteristics of pancreatic cancer patients. In addition, we revealed the relationship of CXCL10 and CXCL11 in immune cell infiltration landscape of pancreatic cancer. Our findings indicate that CXCL10 and CXCL11 may be new targets for pancreatic cancer immunotherapy. However, our study has some limitations; it was limited to bioinformatic predictions and the mechanism of immune infiltration is not well defined, further studies are needed.