Innate immune cells with pattern recognition receptors (PRRs) recognise pathogen-associated molecular pattern molecules (PAMPs) and damage-associated molecular pattern molecules (DAMPs) and induce an immune response involving phagocytosis, antigen presentation, and cytokine production. In UC, the breakdown of the epithelial barrier promotes the influx of intestinal antigens. Consequently, the increased PAMP and DAMP levels in intestinal tissues stimulate innate immune cells, leading to sustained inflammation (Boyapati et al., 2016[17]; Khor et al., 2011[57]). This section explains the roles of various cells involved in the innate immunity in UC (Figure 1(Fig. 1)).

T and B cells are the main cells that play central roles in acquired immunity. Naïve T cells are activated and differentiated by antigen-presenting cells in the gut-associated lymphoid tissues (GALTs) and mesenteric lymph nodes. Differentiated T cells express homing receptors and migrate to the intestinal mucosa by interacting with adhesion molecules on vascular endothelial cells (Arseneau and Cominelli, 2015[5]). The homing receptor α4β7 integrin interacts with mucosal addressin cell-adhesion molecule (MAdCAM)-1, α4β1 integrin interacts with vascular adhesion molecule-1 and fibronectin, and αLβ2 integrin interacts with intercellular adhesion molecule (ICAM)-1 (Habtezion et al., 2016[43]). T cells that migrate to the intestinal mucosa perform their respective functions. This section describes the roles of various cells involved in adaptive immunity in UC (Figure 1(Fig. 1)).

TNF-α exists in two forms: membrane-bound TNF-α (mTNF-α) and soluble TNF-α (sTNF-α) (Parameswaran and Patial, 2010[88]). sTNF-α binds preferentially to TNF-α receptor (TNFR) 1, whereas mTNF-α binds preferentially to TNFR2 (Wajant et al., 2003[124]). TNFR1 is ubiquitous and involved in apoptosis, cell proliferation, and secretion of inflammatory cytokines. TNFR2 is mainly expressed in immune cells and is involved in resistance to apoptosis (Wajant et al., 2003[124]). Anti-TNF-α antibodies inhibit the production of inflammatory cytokines and the proliferation of inflammatory cells by inactivating TNF-α and inducing apoptosis of activated cells (Billmeier et al., 2016[15]).

Anti-TNF-α antibodies, such as infliximab, adalimumab, and golimumab, which inhibit both mTNF-α and sTNF-α, have been effective in the treatment of UC in several controlled trials (Rutgeerts et al., 2005[99]; Sandborn et al., 2012[108], 2014[105]). In contrast, etanercept and onercept, which act mainly on sTNF-α, failed in clinical trials for Crohn's disease (CD) (Sandborn et al., 2001[106]; Van den Brande et al., 2003[123]; Rutgeerts et al., 2006[100]). Perrier et al. reported that neutralisation of mTNF-α, but not sTNF-α, was crucial for the treatment of experimental colitis (Perrier et al., 2013[89]). This highlights the importance of targeting mTNF-α in IBD treatment.

Mayer et al. reported that in patients with UC who responded to anti-TNF-α antibodies, T and B cells decreased and epithelial cells increased, but the number of innate immune cells did not change, using the spatial atlas of colon biopsy tissue using multiplexed immunofluorescence imaging technology (Mayer et al., 2023[75]). Wang et al. identified eight cell populations that were upregulated in the TNF-unresponsive group (inflammatory fibroblasts, post-capillary venules, inflammatory monocytes, macrophages, DCs, and cycling B cells) using a deep learning model based on single-cell ribonucleic acid sequence (scRNA-seq) data (Wang et al., 2023[127]). Smillie et al. identified inflammatory fibroblasts, monocytes, and cDC2s as drug-resistant cell populations (Smillie et al., 2019[116]). In their study, the three genes (IL-13 receptor alpha 2 (IL13RA2), TNF receptor superfamily member 11b (TNFRSF11B), and IL11) that showed the highest correlation with anti-TNF-α antibody resistance were specifically expressed in inflammatory fibroblasts. The IL-13Rα2 is a decoy IL-13 receptor that interferes with IL-13 function and is involved in intestinal homeostasis. Using scRNA-seq analysis, Thomas et al. reported that after adalimumab treatment, the non-responder group showed an increase in pDCs, the main producers of type I IFN, and an increase in IFN responses in epithelial and immune cells afterwards (Thomas et al., 2024[121]). Their study showed an increase in the expression of oncostatin M receptor (OSMR) and neutrophil chemotactic factors, such as CXCL1 and CXCL6, in submucosal and lamina propria fibroblasts. In addition, upregulation of CCL19, a T-cell attractant, was observed in fibroblasts close to the intestinal stem cell niche. The OSMR promotes the secretion of inflammatory cytokines by fibroblasts by binding to its ligand, OSM. It is thought that the innate immune cells and fibroblasts that were increased in anti-TNF-α antibody non-responders represent the main axis of residual inflammation after anti-TNF-α antibody treatment. Furthermore, the fact that these cells increase after treatment suggests that they may also have a mechanism to escape anti-TNF-α antibodies. Therefore, it is believed that a treatment that can stop the cross-talk between innate immune cells and characteristic fibroblasts would be preferable for treating anti-TNF-α antibody non-responders.

Anti-IL-12/23 p40 antibody binds to IL-12 p40 and IL-23 p40, preventing their respective cytokines from binding to their receptors (Nakase et al., 2022[83]). IL-12 is an important cytokine that induces the differentiation of naïve T cells into Th1 cells (Zhu et al., 2010[140]). By inhibiting IL-12, Th1 cell differentiation is inhibited, and the production of Th1 cytokines, such as TNF-α and IFN-γ, is reduced. IL-23 plays an important role in maintaining and proliferating Th17 cells (Bettelli et al., 2007[13]). By inhibiting IL-23, the activation and proliferation of Th17 cells are suppressed, and the production of Th17 cytokines, such as IL-17 and IL-21, is reduced. Ustekinumab is effective and safe in patients with UC compared with a placebo (Sands et al., 2019[110]). Anti-IL-23 p19 antibody binds to IL-23 p19 and inhibits receptor binding of IL-23 to its receptor. Because it inhibits IL-23 without inhibiting IL-12, it suppresses the Th17 pathway and the production of Th17 cytokines and reduces mucosal inflammation. Mirikizumab, risankizumab, and guselkumab are more effective and safe than placebo in patients with UC (D'Haens et al., 2023[29]; Louis et al., 2024[68]; Rubin et al., 2024[97]). A head-to-head comparison of ustekinumab and risankizumab showed the non-inferiority of risankizumab in terms of the rate of clinical remission at 24 weeks and the superiority of risankizumab in the rate of endoscopic remission at 48 weeks in patients with CD (Peyrin-Biroulet et al., 2024[90]). Zhou et al. reported that the affinity and inhibitory effects of risankizumab on IL-23 were higher than those of ustekinumab (Zhou et al., 2021[138]). Meyaard et al. reported that the immune response to IL-12 increased the production of IL-10 via negative feedback (Meyaard et al., 1996[77]). These findings explain the superiority of risankizumab. In addition, because anti-IL-23 p19 antibodies do not suppress IL-12, the Th1 immune response to infections and malignancies is maintained, and these antibodies may be safer than anti-IL-12/23 p40 antibodies (Deepak and Sandborn, 2017[33]).

Imazu et al. reported that flow cytometry of peripheral blood from patients with UC who responded to ustekinumab showed a decrease in the number of Th17 cells (Imazu et al., 2024[49]). In this study, patients with a high percentage of Th17 cells before ustekinumab treatment showed a marked improvement in clinical symptoms compared with patients with a low percentage of Th17 cells. However, in patients with CD, anti-IL-23 p19 antibodies, which inhibit the Th17 pathway more selectively, may be more effective. A head-to-head study on anti-IL-12/23 p40 and anti-IL-23 p19 antibodies is also desirable, due to the lack of existing data on UC. Wang et al. analysed gene expression in patients with CD who were resistant to anti-TNFα antibody treatment and were treated with risankizumab, based on a scRNA-seq dataset of anti-TNFα antibody treatment-resistant cases (Wang et al., 2023[127]). In this study, several modules that were upregulated in anti-TNFα antibody treatment-resistant cases were all reduced in the risankizumab response group. The innate immunity module showed the greatest reduction. In addition, analysis of the scRNA-seq data showed that the cell fractions of inflammatory fibroblasts, inflammatory monocytes, macrophages, circulating B cells, and immature intestinal epithelial cells were significantly reduced in the treatment response group. Inflammatory monocytes were the most reduced cell types. This finding supports the use of risankizumab as a next option when anti-TNF-α antibody therapy is ineffective.

JAKs include four types of intracellular tyrosine kinases: JAK1, JAK2, JAK3, and tyrosine kinase 2 (Clark et al., 2014[25]). When a cytokine binds to its receptor, JAKs bound to the intracellular domain of the receptor move away from each other, and their constant inhibition is released, causing them to become active. Consequently, the transcription factor STAT is phosphorylated, and phosphorylated STAT moves into the nucleus to regulate gene expression (Banerjee et al., 2017[8]). The pairing of JAK and STAT is determined by the ligand and receptor (Figure 3(Fig. 3)). JAK inhibitors exert anti-inflammatory effects by inhibiting the signalling of inflammatory cytokines and cytokines involved in the differentiation of inflammatory cells. Currently, three types of drugs are available for the treatment of UC: tofacitinib, filgotinib, and upadacitinib. Tofacitinib acts on all JAKs, filgotinib acts specifically on JAK1, and upadacitinib acts on JAK1 and partially on JAK2 (Nakase, 2023[81]). These drugs target different JAKs at different doses; therefore, there are differences in their efficacy and safety. Lasa et al. performed a systematic review and network meta-analysis to compare the relative efficacy and safety of targeted therapies in patients with moderate-to-severe UC (Lasa et al., 2022[60]). The results showed that upadacitinib was significantly more effective than all other JAK inhibitors in inducing clinical remission but also had the most adverse effects. Nakase et al. identified potential early mediators of the effect of filgotinib treatment using patient samples from a phase 2b/3 large-scale clinical trial, the SELECTION study (Nakase et al., 2024[82]). In this study, lower levels of systemic inflammatory biomarkers (C-reactive protein, serum amyloid A, and IL-6), neutrophil-associated biomarkers (calprotectin, OSM, and neutrophil gelatinase-associated lipocalin), and Th17 cytokines (IL-17A and IL-22) at week 4 were positively associated with subsequent clinical responses at week 10. In addition, filgotinib did not significantly reduce the levels of anti-inflammatory cytokines (IL-10 and TGFβ-1). Filgotinib may reduce the severity of UC by simultaneously reducing systemic inflammation, Th17 cytokines, and neutrophil inflammation while maintaining some anti-inflammatory pathways. Massimino et al. used UC-derived intestinal microvascular endothelial cell lines to analyse the differences in gene expression between tofacitinib-responding and non-responding cells (Massimino et al., 2022[71]). In this study, gene ontology analysis revealed that responding cell lines were enriched in biological processes related to JAK-STAT signalling, negative regulation of leukocyte adhesion, and epithelial barrier formation compared with non-responding cell lines. In addition, transcriptome analysis showed that tofacitinib administration reduced the population of ICAM-1 and chemokines involved in leukocyte migration in reactive cell lines. In contrast, transcripts encoding tight junction proteins were not altered by tofacitinib. This study suggests that drugs other than anti-integrin antibodies inhibit blood cell trafficking. Clarifying the effects of each drug on the vascular endothelium is thought to be one of the important factors in making treatment choices based on the pathology of UC.

Wang et al. analysed gene expression in patients with CD who were resistant to anti-TNFα antibody treatment and were treated with tofacitinib (Wang et al., 2023[127]). In this study, the analysis of scRNA-seq data showed that the cell fractions of inflammatory fibroblasts, inflammatory monocytes, macrophages, circulating B cells, and immature intestinal epithelial cells were significantly reduced in the upadacitinib response group. Inflammatory fibroblasts were the most affected cell type. Upadacitinib may be effective against populations of cells, such as inflammatory fibroblasts, that are resistant to therapy with anti-TNF-α antibodies.

Anti-α4β7 and anti-α4 integrin antibody preparations are used in the treatment of UC. Both exert their anti-inflammatory effects in the treatment of UC by inhibiting the migration of lymphocytes into the intestinal tract and the infiltration of inflammatory cells into the intestinal tract. Vedolizumab, an anti-α4β7 integrin antibody, and carotegrast methyl, an anti-α4 integrin antibody, are effective in patients with UC (Feagan et al., 2017[37]; Yoshimura et al., 2015[134]). Carotegrast methyl, an anti-α4 integrin antibody, has also been used in limited clinical practice because it cannot be excluded that it is a risk factor for progressive multifocal leukoencephalopathy (Yoshimura et al., 2015[134]). However, anti-α4β7 integrin antibody drugs are relatively safe, with a low risk of side effects, because they act specifically on the intestinal tract (Lasa et al., 2022[60]; Loftus et al., 2020[65]).

Boden et al., assessed immune cell localisation after vedolizumab administration using flow cytometry (Boden et al., 2024[16]). This study showed a decrease in the numbers of naïve B cells, naïve T cells, and IgM⁺ memory B cells in the colon. Hsu et al. reported that the proportions of intestinal Th17 cells, terminal effector CD4⁺ T cells, terminal effector CD8⁺ T cells, and ILC/NK cells were reduced in vedolizumab responders using cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq) analysis (Hsu et al., 2023[48]). Canales-Herrerias et al. reported that the size of GALTs and follicular tissues was reduced, and intestinal IgG⁺ plasma cells were reduced in vedolizumab responders (Canales-Herrerias et al., 2023[21]). Zeissig et al. reported that vedolizumab did not affect the number of T cells in the lamina propria, T-cell activation, or the T-cell receptor repertoire (Zeissig et al., 2019[136]). This report also demonstrated the effect of vedolizumab on innate immunity, with changes in the expression of pattern recognition receptors, chemokines, and NK molecules and a shift in macrophages from inflammatory to regulatory. Mennillo et al. reported that CITE-seq profiling of colon tissues from patients responding to vedolizumab showed a significant decrease in tissue mononuclear phagocytes, expansion of some epithelial subsets, and a trend towards a decrease in activated fibroblasts (Mennillo et al., 2024[76]). Boden et al. reported that the most significant change following vedolizumab administration was a significant decrease in cDC2 in the intestinal tract of patients who responded to vedolizumab (Boden et al., 2024[16]). Vedolizumab treatment may broadly control T cells, innate immune cells, acquired immune cells, and stromal cells. Vedolizumab is highly safe and widely controllable and may be suitable for maintenance treatment, which requires long-term stability.

In contrast, Hsu et al., reported that Th17 cells interacted with inflammatory monocytes/macrophages expressing IL1A, IL1B, IL1RN, OSM, and CCL20 and with bone marrow DCs expressing OSM and were thought to be involved in amplifying and controlling inflammation in vedolizumab non-responders (Hsu et al., 2023[48]). Rath et al. reported that TNF-dependent signalling was highly activated in vedolizumab non-responders (Rath et al., 2018[94]). As interactions between Th17 cells and innate immune cells and activation of TNF-dependent signalling were observed in the vedolizumab non-responder group, treatment with IL23 p19 and anti-TNFα antibodies would be worth trying.

S1P is a bioactive lipid molecule that is mainly produced by erythrocytes and endothelial cells (Hla and Brinkmann, 2011[46]). There are five types of S1P receptor (S1PR), each with different signalling properties. S1PR1 is expressed in lymphocytes and mediates their migration from the lymph nodes (Matloubian et al., 2004[72]). S1PR1 is also expressed in atrial cardiomyocytes and cardiovascular endothelial cells and is involved in blood pressure and heart rate regulation (Hla and Brinkmann, 2011[46]). S1PR2 and S1PR3 are expressed in the central nervous system, vascular endothelial cells, and smooth muscle cells and are involved in vascular endothelial barrier function, neuronal migration, and vascular tension (Groves et al., 2013[42]; Shida et al., 2008[114]). S1PR4 is mainly expressed in lymphocytes and haematopoietic cells and is involved in lymphocyte migration and the regulation of DCs and Th17 cells (Brinkmann 2007[18]; Schulze et al., 2011[112]). S1PR5 is involved in the maturation of oligodendrocytes in the central nervous system and the migration of NK cells in the spleen (Comi et al., 2017[27]; Jenne et al., 2009[50]).

S1PR modulators such as ozanimod and etrasimod have recently become available for treating UC. Ozanimod is an orally administered selective S1PR1 and S1PR5 modulator that maintains peripheral lymphocytes in the lymph nodes by inducing intracellular translocation and degradation (Scott et al., 2016[113]). This inhibits lymphocyte migration into the inflamed tissue. In clinical trials, ozanimod has shown better results than placebo in both the induction and maintenance of remission (Sandborn et al., 2021[104]). Etrasimod is an oral S1PR1, S1PR4, and S1PR5 modulator that exerts anti-inflammatory effects by controlling lymphocyte trafficking. In addition, etrasimod has inhibited inflammatory cytokines, such as TNF-α, IL-1β, IL-6 and IL-17A, and increased the expression of the anti-inflammatory cytokine IL-10 in a mouse model of colitis (Al-Shamma et al., 2019[3]). In clinical trials, both the induction and maintenance of remission have been shown to be superior to those of a placebo (Sandborn et al., 2023[109]). Ozanimod and etrasimod cause bradycardia because they suppress S1PR1, which is involved in the regulation of heart rate (Becher et al., 2022[10]; Sandborn et al., 2023[109]). Because S1PR modulators are a new class of drugs, they have not yet been analysed at the molecular level, and further analysis is expected in the future.

None.

The authors declare that they have no conflicts of interest.

YS, YY, HK, KA, TK, YH, KM, and HN contributed to methodology. YS, YY, amd HN contributed to conceptualization, data curation, and formal analysis. YS and YY contributed to investigation and visualization. YS contributed to writing - original draft. HN contributed to supervision, writing - review & editing.

No artificial intelligence (AI)-assisted technologies (e.g., Large Language Models [LLMs], chatbots, or image creations) were used in the submission process to complete the manuscript.