Pancreatic adenocarcinoma, commonly referred to as pancreatic cancer, is a type of cancer that begins in the pancreas, an organ located deep within the abdomen (1, 2). The pancreas plays a crucial role in digestion and blood sugar regulation by producing enzymes and hormones, such as insulin and glucagon (3, 4). PAAD is a highly aggressive malignancy, exhibiting the poorest clinical outcome of all solid tumors. Generally, when patients come to the hospital, they are diagnosed with advanced PAAD. It is characterized by lower immunogenicity, with a lower number of predicted neoantigens, low immune infiltration, and poor response rates to ICI therapy, with a median survival of 6 months and growing incidence rates worldwide (5–8). The low tumor-infiltrating immune cell microenvironment features of PAAD consist of cancer cells, fibroblasts, myofibroblasts, endothelial cells, tumor-infiltrating lymphocytes (TILs), and extracellular matrix.

The majority of pancreatic cancers are adenocarcinomas, which start in the cells that line the ducts of the pancreas (9, 10). There are other less common types of pancreatic cancer, such as neuroendocrine tumors and acinar cell carcinomas, leading to drug resistance and great inertia in the response rates to immunotherapy (11–13). Therefore, it is particularly difficult to treat PAAD. Specific targeted therapies such as EGFR-targeted therapy confer a significant survival benefit in a minority of PAAD patients who are wild-type KRAS patients (14, 15). Approximately 90% of pancreatic cancers harbor oncogenic KRAS mutations (16, 17). With the improvement in many advanced medical diagnostic technologies, such as endoscopic ultrasound, the early diagnosis rate of PAAD has been greatly improved (18–20). However, given the limited effectiveness of current treatments for PAAD, new promising biomarkers are urgently needed to improve the prognosis and treatment of PAAD.

As a fundamental and prominent post-translational modification of proteins, the involvement of glycosylation in cellular functions includes transformation, adhesion, cell growth, differentiation, and tumor immune surveillance (21, 22). Abnormal glycosylation has been implicated as a crucial mechanism that triggers tumor progression (23, 24). Dysregulation in glycosylation can affect the function, sensitivity and expression of cell-surface receptors. These features make glycosylation a critical molecular event in cancer pathology (25). Glycoproteins are currently the most commonly used clinical tumor biomarkers (e.g., the AFP test for liver cancer, CA125 test for ovarian cancer, CEA, CA199 test for colon cancer, and PSA test for prostate cancer) (26, 27). Subsequently, the role of aberrant glycosylation influences tumorigenesis, which affects growth, immune surveillance, and immunotherapy.

Many recent publications have indicated that glycosylation can lead to an improvement in the immunotherapeutic effect and confer better treatment outcomes in cancer (28–30). As such, further elucidating the molecular mechanisms and consequently promoting cancer immune evasion is urgently required (31). As one of the important components of the tumor microenvironment, proteoglycans promote the progression of cancers, such as lung cancer, pancreatic cancer, and GBM (32–34). Multiple family members are implicated in tumor occurrence and development (35). The current evidence suggests that carbohydrate sulfotransferase 12 (CHST12) is a cancer-related enzyme and is a potential biomarker in some tumor subtypes. For instance, GBMs with high CHST12 expression have a reduced survival rate and positive associations with KI67 expression (36). Qiang Ren et al. found that CHST12 was expressed at a significantly higher level in pancreatic tumors, highlighting the potential diagnostic utility of the CHST12 expression profile in pancreatic cancer (37, 38). However, the specific functions and role of CHST12 in the prognosis of pancreatic cancer and the regulation of TME remain a mystery.

Here, we explored CHST12-related biological processes and CHST12-enhanced pathways to understand its functions in PAAD. The prognostic model based on univariate and multivariate Cox regression analysis revealed an independent prognosticator characteristic. Additionally, we used the TIDE algorithm to indicate the characteristics of CHST12 as a predictor for ICI therapy efficacy. The present study may help develop CHST12 as a prognostic biomarker and predict the efficacy of the immunotherapy treatment of PAAD patients.

Carbohydrate sulfotransferases are enzymes that transfer sulfate to carbohydrate groups in glycoproteins, which have been implicated in the metastatic spread of tumor cells through the modulation of the TME (46–49). The deregulated expression of CHST family members has been reported in a wide range of human cancer types (36). Here, according to our findings, CHST12 mRNA expression was upregulated in PAAD, and low CHST12 expression predicted a favorable outcome for PAAD patients.

Through transcriptome analysis, we found that CHST12 mRNA was highly expressed in tumors, but CHST12 protein showed lower expression in PAAD tumor tissues than in normal tissues. The contradiction between the upregulation of CHST12 mRNA and downregulation of the protein may involve potential post-transcriptional modification mechanisms, such as RNA splicing or protein degradation. This result highly suggests that CHST12 has experienced post-transcriptional modification. Moreover, pancreatic cancer patients with low CHST12 protein expression have a worse prognosis, which is consistent with the findings of Han Nie et al. (50). Through multivariate regression analysis, we identified CHST12 and N staging as independent prognostic risk factors for OS in PAAD patients, which indicates that CHST12 may be useful in clinical settings as a prognostic biomarker and possesses potential clinical utility in PAAD patients. The ROC curve shows the diagnostic performance of CHST12 in discriminating PAAD diagnosis with an AUC of 0.911, suggesting that CHST12 is a high-performance biomarker for PAAD diagnosis. These results agree with previously reported results that other tumor patients with high CHST12 mRNA levels have shorter survival times (51, 52). Considering the key value of CHST12 in clinical diagnosis, we further investigated the potential functions and molecular mechanisms of CHST12 in PAAD. In our study, functional enrichment analysis of CHST12-associated DEGs showed that they are significantly enriched in tumor progression and immune-related signaling pathways. CHST12 has shown a similar effect in hepatocellular carcinoma (53). All the above analyses indicate that CHST12 may be involved in the regulation of the TME in PAAD.

Glycosylation plays a significant role in immunity and endows unique properties to glycoproteins (54, 55). Moreover, our functional studies have found that CHST12 plays a vital role in the TME. Hence, the effects of CHST12 expression on the TME and immunotherapy of pancreatic cancer is a key aspect of our study. According to our findings, CHST12 expression is positively correlated with the infiltration of immune cells (CD4+ T cells, macrophages, neutrophils, and dendritic cells), which contributes to immunosurveillance, the elimination of tumor cells, and the slowing of immune evasion in the PAAD microenvironment (56, 57). Additionally, we proved that CHST12 mRNA expression was associated with several immune checkpoint distributions, including PD-1, CTLA-4, and LAG-3, in PAAD. Using the gene set prioritization module, we found that low CHST12 was related to the increase in the clinical efficacy of immunotherapy in melanoma and bladder cancer. These results suggest that CHST12 may be a potential candidate predictor for immunotherapy, especially ICIs. To test the prediction, we will next conduct a verification experiment clinicopathologically. Combined with the results of the previous discovery of the post-transcriptional modification of CHST12, our study supports the notion that the decreased expression of CHST12 protein in the TME may produce an anti-tumor effect by acting on the recruitment of immune cells to the TME. Notably, in terms of CTLA4, LAG3, and PDCD1, the low expression CHST12 group differed significantly from the high expression CHST12 group. CTLA4, LAG3, and PDCD1 are important targets for cancer ICI therapy. This finding prompted us to investigate whether CHST12 expression impacts clinical outcomes when treated with ICIs. Using the TIDE algorithm, we predicted the response of patients to immunotherapy. We found that high CHST12 mRNA expression is related to the increased efficacy of immunotherapy in melanoma and bladder cancer. These data imply that CHST12 may be a predictor for the efficacy of immunotherapy.

Our findings uncover CHST12 as a new independent predictor of PAAD prognosis and immunotherapeutic efficacy, but it has its limitations. First, CHST12 expression and its prognostic significance were identified through bioinformatic analysis. Hence, further study with clinical samples for such applications is needed to verify the above findings. In addition, in vivo and in vitro experiments are needed to provide additional information about how CHST12 impacts immune infiltration in the PAAD TME.

Conclusively, in this study, we have shown that CHST12 mRNA expression is upregulated in PAAD, whereas CHST12 protein expression is downregulated. The expression of CHST12 is an independent predictor of a good prognosis in PAAD. In addition, CHST12 may affect the infiltration of immune cells in the microenvironment of PAAD. Our data highlight the potential of CHST12 in predicting ICI efficacy. In the future, we hope to further investigate the regulatory mechanisms of CHST12 through animal trials to test its validation. Additionally, we hope to conduct clinical trials as well. Overall, CHST12 may be a novel prognosis biomarker and a potential predictor of the response to ICI treatment in PAAD patients.

The data sets used and/or analyzed during the current study arec available from the following public data sources: the TCGA-PAAD (https://xenabrowser.net/datapages/) and GSE21501 (https://ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE21501) data sets. Publicly available datasets were analyzed in this study. These data can be obtained from the following sites: http://www.xiantao.love; https://www.proteinatlas.org/; https://metascape.org/; http://tide.dfci.harvard.edu/; http://cis.hku.hk/TISIDB/.

The studies involving humans were approved by Shengjing Hospital of China Medical University. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

Interpretation or analysis of the data: KL and LL; Preparation of the manuscript: KL and LL; Revision for important intellectual content: GH; Supervision: GH. All authors contributed to the article and approved the submitted version.