Crohn’s disease (CD) is a type of inflammatory bowel disease (IBD) in which the lesion is mostly located in the terminal ileum and right colon, and its inflammation is typically segmental, asymmetrical, and transmural. Most patients exhibit symptoms such as chronic diarrhea, abdominal pain, fatigue, and weight loss initially. Half of them develop complications (including fibrotic strictures, enteric fistulae, and intestinal neoplasia) with time, which often affects patients’ quality of life and leads to surgery (1–3). With the rapid growth of population, urbanization, industrialization, and westernization of lifestyle, the CD became a global disease at the turn of the 21st century (4, 5). The rapid changes in CD epidemiology have placed a heavy burden on healthcare delivery worldwide (6).

The etiology of CD is unknown and may be related to impaired intestinal barrier function, dysregulated innate and adaptive immune responses, disturbances in the intestinal microbiota, and genetic factors (2). Many studies have explored the immune mechanism of CD. Some studies reported that T lymphocytes played a dominant role in the chronic inflammation process of CD (7, 8). Some studies pointed out that the imbalance of innate lymphoid cells (ILCs) in the gut led to CD (9, 10). However, immune regulation was a complex process, and immune cells did not exist independently but interacted with each other. Therefore, a multidimensional and in-depth explanation of the pathogenesis of CD will be of great significance.

Currently, the activity of CD is assessed based on standard clinical scores, which might not adequately reflect the immune mechanism of disease activity (11). To reveal the immunological characterization of the active phase and inactive phase for CD was momentous. However, few articles were performed to study immunological characteristics of the difference between CD patients with active phase (CDa) and those with inactive phase (CDin). Martin, Chang et al. conducted a study associated with resistance to anti-TNF therapy of CD by identifying an active cellular module (12), which inspired us that the determination of the immunological characteristics of CDa and CDin was of great significance for therapy.

Both mass cytometry (CyTOF) and single-cell RNA sequencing (scRNA-seq) were used to explore the immune characteristics of different stages of the disease. CyTOF is a high-throughput technique for measuring the abundance of proteins on the cell surface or within cells (13). CyTOF was applied to various functional assays, such as phenotypic characterization, intracellular cytokine determination, intracellular signaling state, and cell viability identification. Thus, the use of CyTOF offers new opportunities to address the complexity of cellular immunology (14). scRNA-seq is a new technology for amplification and sequencing the whole transcriptome at the single-cell level (15). Through scRNA-seq, various cell types can be acquired by learning the expression level and mutated structure of genes in individual cells. scRNA-seq plays a key role in the classification of cell types, identification of cellular immune microenvironment, and cell lineage evolution (16).

Mitsialis et al. performed a single-cell analysis of colon and blood samples from patients with IBD and identified different immune cell signatures of UC and CD, respectively (17). However, the immune characteristics of different stages of CD were unknown. This study intends to explore the immune cell characteristic of patients with CDa and CDin using CyTOF and scRNA-seq and to compare differences in immune cells among different disease periods of CD.

CyTOF analysis was performed on the data of intestinal mucosa from 37 CD (CDin = 25 and CDa = 12) and 18 HC. The characteristics of demographic data were as follows. There were no significant differences between the three groups in factors such as gender, age at collection, age at diagnosis, and duration of disease among others (Supplementary Table 1).

Single-cell RNA sequencing data contained six HC and 12 patients with CD (CDin = 6 and CDa = 6). scRNA-seq was conducted to verify the results of CyTOF. Furthermore, CellChat analysis was applied to deduce the cell communication network among subsets by 22 CD samples (CDin = 11 and CDa = 11). There was no significant difference for factors of gender (Supplementary Table 1).

Subsets with significant abundance differences among CDa, CDin, and HC were identified using CyTOF and scRNA-seq. The high abundance of CD16⁺CD38⁺CD4⁺CXCR3⁺CCR6⁺ naive T cells, HLA-DR⁺CD38⁺IFNγ⁺TNF⁺ EM T cells, HLA-DR⁺IFNγ⁺ naive B cells, and CD14⁺⁺CD11C⁺IFNγ⁺IL1B⁺ monocytes were considered to be cellular immune characteristics for the patients with CDa. In addition, the expansion of CD16⁺CD56^dimCD44⁺HLA-DR⁺IL22⁺ NK cells was considered to be a special subset for CDin.

The high abundance of CD16⁺CD38⁺CD4⁺CXCR3⁺ CCR6⁺ naive T cells was first found to be associated with CDa. According to our result, CD16⁺CD38⁺CD4⁺CXCR3⁺CCR6⁺ naive T cells were obviously increased in the patients with CDa compared with HC and CDin. The expansion of CD16⁺CD38⁺CD4⁺CXCR3⁺CCR6⁺ naive T cells might be related to the formation of T-helper 1 (Th1), which was relevant to the pathogenesis of CD. In IBD, naive T cells differentiate into various T-helper (Th) cells mediated by relevant antigen-presenting cells (23). Furthermore, an increase in CD16, CXCR3, and CCR6 was associated with inflammation in CDa. CD16 is a low-affinity Fc receptor that mediates phagocytosis and cytotoxicity and is a pharmacogenetic biomarker of anti-TNF therapy (24). CXCR3 was a chemokine expression receptor that determined the different functional states of CD4⁺ T cells and gave impetus to the formation of Th1 (25). In addition, CXCR3 could activate chemokines, such as CXCL9, CXCL10, and CXCL11, to recruit immune cells at the site of inflammation, causing local inflammation amplification and thus inducing a deterioration of clinical manifestations (26).

HLA-DR⁺CD38⁺IFNγ⁺TNF⁺ EM T cells may be a subset associated with immune dysregulation in CDa. This subset had an increment in the patients with CDa than those with CDin. The increase of activated EM T cells in CD realized multiple pro-inflammatory functions by the release of more cytokines, which might further enhance and perpetuate inflammation (27–29). Furthermore, activated EM T cells were promoted to secrete HLA-DR, CD38, TNF, IFNγ, and other cytokines, and these cytokines were strongly associated with further enhancement of the active phase of IBD (30). In addition, the results of CellChat showed that this subset was regulated as a receptor in the CD80/CD86 signaling pathway, which was inextricably linked to the activation of T cells (31, 32). The increase of EM T cells in CDa made the activation of effector cells increase, leading to immune dysregulation and persistent inflammation (27, 29). Therefore, HLA-DR⁺CD38⁺IFNγ⁺TNF⁺ EM T cells may be cellular immune characteristics for CDa.

CD14⁺⁺CD11C⁺IFNγ⁺IL1B⁺ monocytes may be an influential subset affecting the transition between active and inactive phases of CD. Our results pointed out that CD14⁺⁺CD11C⁺IFNγ⁺IL1B⁺ monocytes enriched in the patients with CDa. Coincidentally, Zhang et al. applied scRNA-seq and CyTOF to define unique cell populations in rheumatoid arthritis (RA) and found that monocytes with high expression of IFNγ and IL1B were significantly enriched in RA (33). (1) This might be the result of pro-inflammatory signals from the site of infection (including pro-inflammatory cytokines and microbial molecules), inducing monocytes migration, such as the high expression of adhesion molecules in the endothelium, allowing monocytes to migrate across the endothelium when inflammation occurs (34–36). It could be guessed that under pathological conditions, monocytes acquired inflammatory effector functions as a result of diverse combinations of cytokines in the microenvironment. (2) Moreover, CellChat results showed that CD14⁺⁺CD11C⁺IFNγ⁺IL1B⁺ monocytes were a paramount signal transmitter in TNF signaling pathway, and its major receptors were endothelial cells and T cells, which corresponded with preceding findings (37). Sampaio EP et al. also mentioned that the interaction between T cells and monocytes enhanced TNF production, which might contribute to the secretion of pro-inflammatory cytokines in disease states (38). In addition, T cell–monocyte interactions modulated epithelial dysfunction (39).

CD16⁺CD56^dimCD44⁺HLA-DR⁺IL22⁺ NK cells were considered to be a special subset for CDin. CyTOF results showed that CD16⁺CD56^dimCD44⁺HLA-DR⁺IL22⁺ NK cells were markedly increased in the patients with CDin compared with CDa, which was not mentioned before. Berahovich et al. reported that CD16⁺CD56^dim NK cells had strong toxicity (40), which might be more toxic when combined with the high expression of CD44 (41, 42). Previous studies reported that the killing capacity of peripheral circulating NK cells in patients with IBD was abated, which might lead to an increased risk of secondary infection for patients with IBD (43). In addition, the high expression of CD44 indicated that this subset had strong cytotoxicity (41, 42). The high expression of IL22 protected the integrity of the intestinal mucosal barrier (44, 45).

Our study pointed out that IFNγ played a momentous role in the pathogenesis of CD. IFNγ was highly expressed in four subsets (CD16⁺CD38⁺CD4⁺CXCR3⁺CCR6⁺ naive T cells, HLA-DR⁺CD38⁺IFNγ⁺TNF⁺ EM T cells, HLA-DR⁺IFNγ⁺ naive B cells, and CD14⁺⁺CD11C⁺IFNγ⁺IL1B⁺ monocytes). Moreover, the four subsets were notably expanded in CDa compared with CDin. The results were confirmed by other research (46). Studies showed that IFNγ could enhance innate immune cell activation and recruitment to infected tissues. It could directly or indirectly induce the secretion of pro-inflammatory cytokines, antigen presentation, effector molecule expression, and the differentiation of each innate cell subpopulation (47). The high expression of IFNγ might keep pace with the increase of these cell subsets in the active phase of CD.

Although there are some interesting discoveries in the study, further mechanistic studies or experimental validation are essential to identify the pathogenic factors or biomarkers associated with a different phase of CD.

In conclusion, the high abundance of CD16⁺CD38⁺CD4⁺CXCR3⁺CCR6⁺ naive T cells, HLA-DR⁺CD38⁺IFNγ⁺TNF⁺ EM T cells, HLA-DR⁺IFNγ⁺ naive B cells, and CD14⁺⁺CD11C⁺IFNγ⁺IL1B⁺ monocytes were considered to be cellular immune characteristics for CDa. The expansion of CD16⁺CD56^dimCD44⁺HLA-DR⁺IL22⁺ NK cells was considered to be a special subset for CDin. These cells may be related to the changes in the active and inactive phases of CD. We hope that the results can provide assistance in disease diagnosis and treatment for patients with CD in the future.

The original contributions presented in this study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

X-LC and WL were responsible for the conception and design of the study. WL was responsible for the data processing and analysis. X-LC, WL, SL, and JZ wrote the manuscript. TL and ZX modified the figures. X-LC, WL, HWL, ZH, YXL, YL, YWL, PL, QX, and HBL had provided professional suggestions. All authors contributed to the writing of this manuscript and approved the final version of the manuscript.