Crohn’s disease (CD) is a major type of inflammatory bowel disease (IBD), affecting approximately 0.5% of the worldwide population, and the incidence rate is increasing in China (1, 2). CD is an immune-mediated disorder characterized by chronic inflammation in the gastrointestinal tract. The exact etiology of CD remains largely unknown, but genetic and environmental factors are believed to be the main contributors that trigger inflammatory processes in the intestinal mucosa (3, 4). To date, no curative medical approach is available for CD (5); therefore, exploring the role of immune cells and the related cytokines is crucial for the development of effective therapeutic strategies for patients with CD.

Previous studies have shown that abnormal changes in immune cells and their effective cytokines exaggerate immune responses in the intestinal mucosa of patients with CD. For instance, CD14⁺ macrophages participate in CD by inducing activation-induced cell death resistance in CD4⁺ T cells (6). CK2α contributes to the pathogenesis of CD by promoting CD4⁺ T cell proliferation and Th1 and Th17 response (7). Human leukocyte antigen class II molecules, IL-1, IL-6, and IL-8, are also overexpressed in the intestinal epithelial cells of patients with active CD (8). This evidence demonstrates that chronic inflammation is induced by immune cells in patients with CD.

With the advancement in genomic sequencing technology, an increasing number of microarray analyses of disease-related datasets have been conducted, and these datasets have become an ideal source to quantify immune cell infiltration in various diseases. Recently, Chen et al. (9) provided an overview of immune cell alterations in CD using Gene Expression Omnibus (GEO) datasets and found that CXCL8 and IL-1B have a good diagnostic value in CD. However, relevant cytokines still need to be elucidated. Therefore, in this study, we used single-sample gene set enrichment analysis (ssGSEA) aiming to estimate immune cell infiltration in CD, and assess the role of immune-related cytokines in CD, which will highlight the relationship between immune infiltration at the molecular level and provide potential therapeutic targets for patients with CD.

Evidences have confirmed that CD is a chronic inflammatory intestinal disease, which is caused by an unrestrainable immune response against luminal bacterial antigens. The role of inflammatory cells infiltrating in maintaining an active stage is well established and most of the therapeutic strategies are aiming to block the cascade of inflammatory reaction and release of pro-inflammatory cytokines (14, 15). In the present study, by analyzing the data for CD, non-inflammatory tissues, and normal intestinal tissues using ssGSEA, we described the landscape of immune infiltration in CD and found that the fraction of 16 immune cells was significantly different among these three kinds of tissues. In addition, we found more immune cells that were significantly different between CD tissues and non-inflammatory tissues, and most of the immune cells showed higher infiltration in CD tissues than in non-inflammatory tissues. These results were consistent with previous studies (9), which further confirmed the important role of abnormal immune cells in the pathogenesis and progression of CD.

The abnormal immune system is a critical contributor to the pathogenesis and progression of CD (16). During the last decades, many studies have confirmed that CD tissues often present high immune infiltration, and immune cells and immune-related genes are attractive targets for modulating CD progression (17). Importantly, innate and adaptive immune cell infiltration is closely associated with patient response to treatments and clinical outcomes. Therefore, this evidence indicates that the application of immune-based features in CD is a reasonable approach to estimate the pathogenesis and clinical outcome (18). By analyzing the transcriptome data of CD using ssGSEA (19), xCell (20) or CIBERSORTx (9), researchers could provide insights into the regulatory network between CD and immune cells, which could deepen our knowledge on the association of CD with its immune environment.

Generally, immune cells perform their role by secreting immune-related cytokines. We screened the potential immune-related genes by overlapping the immune cell markers and DEGs and using a PPI network to identify the key immune-related genes. The PPI network and the sub-cluster analysis method have been widely applied to identify key genes in the pathogenesis of diseases (21, 22). Using the same method, we firstly established a PPI for the immune-related genes in CD, and then screened the key genes from the sub-cluster of PPI. The results showed that AQP9, CD27, and HVCN1 ranked the highest score in each sub-cluster, and the correlation analysis demonstrated that these three genes were significantly correlated with several immune cells. It is noteworthy that AQP9, CD27, and HVCN1 are associated with CD, as shown in previous studies. For example, AQP9 in the peripheral blood could discriminate patients with CD from patients with chronic inflammation and healthy controls (23). In rat myocardial infarction model, the mRNA and protein expression levels of AQP9 was significantly increased, and silencing AQP9 gene can inhibit the activation of ERK1/2 signaling pathway, attenuate the inflammatory response in rats with myocardial infarction, inhibit apoptosis of myocardial cells, and improve cardiac function (24). In addition, AQP9 was showed to play a role in regulating tissue-specific physiological properties in tight junctions in ulcerative colitis (25). Regarding the CD27 in CD, reduced CD27^- IgD^- B cells in the blood and raised CD27^- IgD^- B cells was found in the gut-associated lymphoid tissue in IBD (26), study also observed that patients with CD were characterized by a reduced proportion of B cells of the memory CD27+ phenotype compared to the non-IBD controls (27). In term of HVCN1, expression of HVCN1 was found to significantly positively correlate with the clinical activity of CD (28). Moreover, in cystic fibrosis patients who underwent ivacaftor therapy, the HVCN1 mRNA expression was significantly higher than baseline at 1-3 months and decreased after 6 months of treatment (29). The similar results were found in cystic fibrosis patients who underwent antibiotic therapy (30). All the above evidences demonstrated that the three genes play critical roles in the pathogenesis of inflammatory diseases or CD. Therefore, we select these three genes in the following analysis.

Since AQP9, CD27, and HVCN1 are immune-related genes overexpressed in CD tissues, we assessed the diagnostic value of these three genes. As expected, AQP9, CD27, and HVCN1 had a mild to moderate diagnostic value in CD, especially when using normal intestinal tissues as controls, and the diagnostic value of AQP9 was much better than that of CD27 and HVCN1. Although the diagnostic value of CD27 has been reported in multiple myeloma (31), its diagnostic value in CD has not been described previously. It should be noted that Chen et al. (9) showed that CXCL8, encoding IL-1B, together with immune cells has a good diagnostic value in CD. Therefore, our results provided other candidate diagnostic indicators for diagnosis.

Regarding the treatment of CD, infliximab, adalimumab, and ustekinumab are the crucial agents of anti-TNFα therapy, which have shown high efficacy in many patients with CD (32, 33). However, early identification of non-responders to anti-TNFα therapy remains a challenge in the clinical setting. Currently, some indicators or methods have been reported to predict the response of patients with CD who underwent anti-TNFα therapy, such as serum oncostatin M (34), small bowel ultrasound (35), and SMAD7 (36). However, more indicators still need to be explored. In this study, we examined the predictive value of AQP9, CD27, and HVCN1 in patients who underwent anti-TNFα therapy. Our results indicated that only AQP9 had a good predictive value in those responders. More importantly, the expression of AQP9 did not significantly decrease in non-responders, suggesting a higher specificity of AQP9. To date, although some inflammatory and immune biomarkers are used to diagnose the CD, considering the difficulty of diagnosis CD in the clinical setting, this study provided other alternatives to the diagnosis of CD, and these biomarkers also help to assess the anti-TNFα therapy on CD patients. In addition, our results imply that the expression of AQP9 and CD27, was significantly correlated to the activity of CD, and AQP9 could be served as a valuable indicator to monitor the effect of anti-TNFα treatment.

Although the present study employed many GEO datasets with a larger number of tissues to explore the immune cells and immune-related genes in CD and identified three genes, with promising diagnostic and predictive values in patients with CD, some limitations remain to be addressed. First, the fraction of immune cells in CD tissues was analyzed using ssGSEA; therefore, the results need to be verified using tissue-based flow cytometry. Second, we applied IHC to validate the expression of AQP9, CD27, and HVCN1 in CD tissues; however, due to lack of samples after anti-TNFα therapy, the predictive ability of these genes still needs to be validated in future studies. Third, due to the lack of data, including disease phenotype and disease activity status, the associations between immune cells and related genes, and disease severity, need to be further estimated.

The present study provided insight into the landscape of immune cells underlying CD and identified AQP9, CD27, and HVCN1 as auxiliary diagnostic indicators for CD. We also identified AQP9 as a promising predictive indicator in patients who underwent anti-TNF therapy. However, future studies with larger samples and clinical information using flow cytometry analyses are warranted to validate these findings.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by People’s Hospital of Guangxi Zhuang Autonomous Region. The patients/participants provided their written informed consent to participate in this study.

Study concept and design: BY, X-wG, and B-lH. Collection and assembly of data: Y-xY, Z-gP, and Y-pT. Data analysis and interpretation: BY, Y-xY, Z-gP, and Y-pT. Immunohistochemical assay: BY, K-lW, and X-wG. Manuscript revised: B-lH, K-ZL, and BY. Manuscript writing and review: All authors. All authors contributed to the article and approved the submitted version.

This study was partially supported by research funding from Cultivation project of NSFC (no. GJPY2019005) and the National Natural Science Foundation of Guangxi (no. 2020GXNSFAA297116).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.