Type 17 immunity is an immune response against certain types of pathogens that are not effectively controlled by classical type 1 or type 2 immunity. While type 1 immunity is involved primarily in combating intracellular pathogens, such as viruses and some bacteria, and type 2 immunity is required for defense against extracellular parasites, such as helminths, type 17 immunity is crucial for controlling extracellular bacteria and fungi that can cause mucosal tissue damage. Therefore, type 17 immunity is a specialized arm of the immune system that is mediated primarily by specific subsets of innate and T lymphocytes, including group 3 innate lymphoid cells (ILC3s) and IL-17-producing T helper 17 (Th17) cells. This type of immunity is characterized by the production of the signature cytokine IL-17A, along with other related cytokines such as IL-17F, IL-22, IL-23, and GM-CSF. These proinflammatory molecules mediate and sustain tissue inflammation by recruiting neutrophils and other immune cells to sites of infection or tissue injury.

Interleukin (IL)-23 is a heterodimeric cytokine composed of the proteins p19 and p40, which are encoded by Il23a and Il12b, respectively, and is essential for sustaining and amplifying type 17 immunity. While p19 is the unique subunit of IL-23, p40 is the common subunit shared with IL-12 [1]. A subset of mononuclear phagocytes (MNPs) is considered a key cellular source of IL-23 in response to various physiological and pathological stimuli. IL-23 signaling in both innate and adaptive immune cells is vital for mediating type 17 immunity, resulting in the upregulation of several type 17 cytokines [2]. These responses confer gut-protective effects but have also been implicated in the pathogenesis of inflammatory and autoimmune disorders. IL-23 promotes the activation of effector functions in ILC3s and Th17 cells, which both express high levels of the IL-23 receptor. Notably, downstream targets of IL-23, namely, IL-17A/F and IL-22, act on gut epithelial cells to promote wound healing and enhance gut barrier integrity by increasing the expression of mucus, antimicrobial peptides, chemokines, and tight junction proteins [3,4,5].

MNPs, which include monocytes, macrophages, and conventional dendritic cells (cDCs), play crucial roles in regulating the immune system under physiological and pathological conditions. In response to microbial infection, MNPs orchestrate innate and adaptive immune responses by sensing pathogens, upregulating costimulatory molecules and various inflammatory cytokines, including IL-23, mobilizing, and presenting foreign antigens to T cells, thereby driving the activation of innate immune cells and the differentiation of T helper (Th) cells [6].

In this review, we first describe the roles of type 17 immunity in governing host defense mechanisms and inflammatory conditions in the gut. We subsequently extended our scope to multiorgan networks, with a particular emphasis on T-cell-mediated autoimmune diseases.

Type 17 immunity in maintaining intestinal homeostasis and controlling gut pathogensIL-23 in gut homeostasis

Although the gastrointestinal tract harbors a myriad of commensal microorganisms, gut homeostasis is primarily sustained by basal immune responses to these microbes and the maintenance of physical barrier functions, effectively mitigating both local and systemic inflammation. IL-23 is constitutively detected in the gut but not in extraintestinal tissues such as the spleen, lung, liver, kidney, or skin [7]. Under physiological conditions, IL-23 is expressed primarily by cDCs within gut-associated lymphoid tissues such as mesenteric lymph nodes (mLNs), Peyer’s patches, and tertiary lymphoid organs, including isolated lymphoid follicles (ILFs) and cryptopatches (CPs), thereby maintaining gut homeostasis [8,9,10]. Activation of the intestinal IL-23‒IL-22 pathway regulates the abundance of specific commensal populations, such as segmented filamentous bacteria (SFB), capable of inducing nonpathogenic Th17 cells [11, 12].

IL-23‒IL-22 axis in the host defense against mucosal infections The role of IL-22 and IL-17A/F in gut infection

The IL-23‒IL-22 axis is essential for the elimination of gut invasive pathogens, such as attached and effiving bacteria, by orchestrating the recruitment of neutrophils and inflammatory monocytes via the induction of chemokines, including CXCL1, CXCL2, CCL1, and CCL2, which are secreted by intestinal epithelial cells (IECs) in response to IL-22 stimulation. In addition, IL-22 signaling enhances the expression of RegIII, an antimicrobial peptide, and fucosyltransferase 2 (Fut2), which promotes the fucosylation of the luminal side of the IEC membrane [3, 13,14,15]. Il23a−/− and Il22−/− mice succumb to Citrobacter. rodentium (C. rodentium) infection, a model organism for human enteropathogenic E. coli infection [13, 16, 17]. The signaling pathway downstream of the IL-22 receptor in IECs has been well characterized [18]. The binding of IL-22 to its receptor induces the phosphorylation of JAK1 and TYK2 and subsequently phosphorylates STAT3. The dimerized pSTAT3 then translocates to the nucleus, where it induces the expression of chemokines and antimicrobial peptides. Additionally, IL-22 signaling activates the PI3K and MAPK pathways, which synergistically promote the expression of these genes together with pSTAT3.

IECs can be subdivided into two major branches: absorptive cells and secretory cells, the latter comprising goblet cells, Paneth cells, enteroendocrine cells, and tuft cells. Recent studies have highlighted the role of IL-22 signaling in different IEC subtypes. One report suggested that IL-22 signaling in absorptive cells is involved in the elimination of C. rodentium infections [19], whereas another study showed that IL-22 signaling in secretory cells, especially in goblet cells, is important for maintaining gut barrier integrity [20]. IL-22 signaling has also been implicated in the differentiation and maturation of Paneth cells, as shown by studies using Il22−/− mice and human gut organoid models [15, 21]. However, further study is needed to elucidate the precise mechanisms by which IL-22 signaling regulates the specific functions of each IEC subset during the clearance of intestinal infections.

While IL-23 is known for its ability to stimulate IL-17A/F production from Th17 cells alongside IL-22, the efficacy of IL-17A/F in clearing C. rodentium is relatively modest in comparison to that of IL-22. Mice deficient in IL-17A and/or IL-17F signaling exhibit an increased bacterial burden during C. rodentium infection; however, all infected mice survive and maintain a normal body weight [13, 22]. Interestingly, the upregulation of IL-17A during C. rodentium infection occurs independently of IL-23 [13, 16], indicating the functional capacity of Th17 cell-derived IL-17A production machinery in the absence of IL-23, potentially under the regulation of alternative Th17-associated cytokines such as IL-6, TGF-β, and IL-1β. Conversely, the induction of IL-22 production by both ILC3s and Th17 cells is completely dependent on IL-23 [13]. Mechanistically, IL-22 induces the production of antimicrobial peptides, including RegIII and S100A proteins, from the intestinal epithelium, whereas IL-17A/F induces β-defensin production, which appears to be insufficient for controlling C. rodentium infection [22]. Thus, the IL-23‒IL-22 axis plays a more critical role in combating C. rodentium compared to the IL-23‒IL-17A/Faxis, which has distinct functions in maintaining gut homeostasis.

IL-23-producing cells in C. rodentium infection

The susceptibility of Il23a−/− mice to C. rodentium infection has sparked considerable interest in identifying the specific intestinal cell subsets responsible for IL-23 production [16]. While IL-23 from bone marrow-derived CD11c+ MNPs, comprising cDCs and macrophages, is crucial for C. rodentium clearance in a CD11c-diphteria toxin receptor (DTR) mouse model [23], the dominant IL-23-producing subset within MNPs remains elusive because of the heterogeneity of cDCs and macrophages.

Gut cDCs can be subdivided into four subpopulations on the basis of the surface markers CD103 and CD11b. While CD103+ CD11b- cDCs are classified as cDC1s, CD103+ CD11b+, CD103- CD11b+, and CD103- CD11b- cDCs are classified as cDC2s [6]. The development of cDC1s and cDC2s relies on the transcription factors Batf3 and IRF4, respectively [24,25,26]. Ly6Chi monocytes differentiate into Ly6C- MHCII+ macrophages through the coexpression of Ly6C and MHCII [27, 28]. Although both cDCs and macrophages express high levels of CD11c and MHCII, they can be segregated using cDC-specific markers such as Zbtb46 and CD26 and macrophage-specific markers such as CD64 and CD88 [6, 29,30,31,32].

It was initially proposed that CD103+ CD11b+ cDC2s are the key source of IL-23 in the gut [33, 34]. CD11cCreNotch2flox/flox mice lacking CD103+ CD11b+ cDC2s were shown to be highly susceptible to C. rodentium infection. Notably, mixed bone marrow chimeric mice generated from CD11cCreNotch2flox/flox and Il23a−/− bone marrow (BM) cells succumbed to C. rodentium infection, similar to those receiving only Il23a−/− BM cells. These findings collectively underscore the indispensable role of IL-23 derived from Notch2-dependent CD11c+ MNPs, primarily CD103+ CD11b+ cDC2s, in conferring protection against C. rodentium infection. Conversely, another study investigated the role of CD103+ CD11b+ cDC2s as the primary source of IL-23 in the gut [35]. This study utilized bacterial artificial chromosome transgenic mice, called huLangerin-diphteria toxin (DTA) mice, which also lack gut CD103+ CD11b+ cDC2s similar to those of CD11cCreNotch2flox/flox mice. Intriguingly, despite the depletion of gut CD103+ CD11b+ cDC2s in these mice, huLangerin-DTA mice were resistant to C. rodentium infection compared with WT mice, suggesting that alternative cellular sources capable of producing IL-23 distinct from CD103+ CD11b+ cDC2s.

To further delineate the cellular source of IL-23, we generated Il23aVenus mice, which allow direct visualization of IL-23-producing cells [8]. This strain showed that the majority of IL-23-producing cells were EpCAM+ DCIR2+ cDC2s, which can be divided into two subpopulations: CD103+ CD11b+ and CD103- CD11b-. Notably, EpCAM+ DCIR2+ CD103- CD11b- cDC2s, termed cDCIL-23, exhibited greater potency in terms of IL-23 production than did the CD103+ CD11b+ cDC2 fraction during C. rodentium infection. Furthermore, we demonstrated that the development of the EpCAM+ DCIR2+ cDC2 subset depends on Notch2 signaling (Fig. 1). CD11cCreNotch2flox/flox mice indeed lack IL-23 expression not only in CD103+ CD11b+ cDC2s but also in EpCAM+ DCIR2+ CD103− CD11b− cDC2s (cDCIL-23). IL-23 expression in cDCIL-23 is likely intact in Langerin-DTA mice [35]. The absence of the cDCIL-23 subset likely accounts for the high susceptibility of CD11cCreNotch2flox/flox mice to C. rodentium infection [8, 34]. Consequently, our findings concerning cDCIL-23 provide insights that reconcile the discrepancy observed in the phenotypes of CD11cCreNotch2flox/flox and huLangerin-DTA mice during C. rodentium infection.

Fig. 1

Three-step developmental model of IL-23-producing cDCs. The activation of Notch2 signaling serves as the initial trigger for the differentiation process of EpCAM+ DCIR+ cDCs. This is followed by the induction of CIA-DC signature genes, such as LyzM, in cDCs through lymphotoxins from ILC3s and their recruitment into tertiary lymphoid organs. Ultimately, the terminal differentiation of these cells necessitates the involvement of retinoic acid signaling, which is pivotal for endowing them with robust ability to produce IL-23

IL-23-producing cDCs are specifically localized in mLNs and tertiary lymphoid organs, such as ILFs and CPs, which harbor abundant numbers of IL-22-producing ILC3s [8]. Recently, a distinct cDC subpopulation called “CIA-DCs” was shown to be confined to ILFs and CPs [10]. These cDCs are defined as LyzMhi Plet1+ CD103- cDCs. Single-cell RNA-seq analysis of SILP cDCs revealed that LyzMhi cDCs highly express Il23a, Epcam, Clec4a4 (DCIR2), and other gene signatures akin to Il23a-Venus+ cDCs, indicating that LyzMhi CIA-DCs likely include the majority of the cDCIL-23 population [10] (Fig. 2A, B). In addition, ILC3-derived lymphotoxin is essential for the development of LyzMhi Plet1+ CD103− cDCs, as CD11cCreLtbrflox/flox mice lack these cDCs alongside some CD103+ CD11b+ cDC2s, resulting in decreased Il23a expression in the colon and increased bacterial burdens in the feces, liver, and spleen during C. rodentium infection compared with those in WT mice [36]. Taken together, these findings underscore the importance of cDCIL-23, which is located within gut-associated lymphoid tissues, in orchestrating mucosal host defense against C. rodentium.

Fig. 2

Similar gene expression profiles between Il23a-Venus+ EpCAM+ DCIR2+ CD103- CD11b- cDCs (cDCIL-23) and LyzMhi CIA-DCs. A Reanalysis of the scRNA-seq atlas of SILP cDCs (E-MTAB-9522) [10]. Marker gene expression associated with CIA-DC and Il23a-Venus+ cDCs was projected via UMAP analysis of SILP cDCs. B The gene set score of the Il23a-Venus+ cDC signature was calculated via the scanpy.tl.score_genes function and projected via UMAP analysis [190]. The top 100 upregulated genes sorted by adjusted p value in bulk RNA-seq data of Il23a-Venus+ cDCs compared with CD103+ CD11b+ cDCs were used as the signature genes of Il23a-Venus+ cDCs (DRA016070) [8]

Additionally, CX3CR1+ macrophages were proposed as a potential cellular source of IL-23 during C. rodentium infection via CX3CR1-DTR generated by crossing CX3CR1-flox-STOP-flox-DTR with CD11cCre or Rosa26-flox-STOP-flox-DTR mice with CX3CR1Cre [23, 37]. The administration of diphtheria toxin to these mice enables the selective depletion of CX3CR1+ macrophages while preserving CD103+ CD11b+ cDC2s, increasing the susceptibility of the mice to C. rodentium infection. Furthermore, mixed bone marrow chimeric mice generated from CX3CR1-DTR and Il23a−/− BM cells also succumbed to C. rodentium infection. However, CX3CR1 is also expressed by CD103- cDC2s, suggesting potential depletion of IL-23-producing cDC2s in CX3CR1-DTR mice [38, 39]. Indeed, CX3CR1-Cre is expressed in CD103- cDC2s from CX3CR1Cre mice [23]. Therefore, it is necessary to consider that not only CX3CR1+ macrophages but also cDCIL-23 could be affected in CX3CR1-DTR mice. Given the absence of IL-23 reporter expression in gut macrophages [8], it is reasonable to assert that cDCIL-23, rather than CX3CR1+ macrophages, represents a key cellular subset crucial for IL-23 production in C. rodentium infection.

Spatiotemporal regulation of the IL-23‒IL-22 axis in C. rodentium infection

There seems to be complex regulation of the cellular origins of IL-23 and IL-22 during the early and late phases of C. rodentium infection. As described earlier, IL-23-producing cDCs in tertiary lymphoid organs detect C. rodentium and produce high amounts of IL-23 during the initial phase of infection. Subsequently, ILC3s in close proximity to such cDCs upregulate IL-22 expression in response to IL-23 (Fig. 3A). This cascade triggers the activation of IECs, instigating early host defense mechanisms characterized by the production of antimicrobial peptides and mucus, alongside the recruitment of inflammatory monocytes and neutrophils via chemokine signaling [40]. Notably, ~60% of PlzfCreIl22flox/flox mice, which lack Il22 in innate-type lymphoid cells such as ILCs, NKs, and γδTs but not in CD4+ T cells, died during the early stages of C. rodentium infection. Like Il23a−/− and PlzfCreIl22flox/flox mice [34, 41], mice with deficiencies in IL-23 production from cDCs, such as Zbtb46-DTR or CD11cCreNotch2flox/flox mice, also succumb during the early phase of C. rodentium infection [34].

Fig. 3

Spatiotemporal regulation of the IL-23‒IL-22 axis during C. rodentium infection. A During the early phases of C. rodentium infection, the IL-23‒IL-22 axis in tertiary lymphoid organs, such as ILFs and CPs, is critical. Specifically, IL-23-producing EpCAM+ DCIR2+ CD103- CD11b- cDCs (cDCIL-23) activated by TLR signaling via pathogen-associated molecular patterns play a pivotal role in stimulating IL-22 expression by ILC3s. B During its later phase, in addition to the interaction between cDCIL-23 and ILC3s in tertiary lymphoid organs, the engagement of macrophages and IECs in the activation of Th17 and/or ILC3s within the lamina propria emerges as an additional significant aspect in combating C. rodentium infection. Notably, CX3CR1+ macrophages and IECs facilitate IL-22 expression by Th17 cells through the presentation of antigens, costimulatory molecules, and cytokines such as IL-1β, thereby fostering sustained activation of gut epithelial cells and eventual elimination of C. rodentium infection

However, at later stages of infection, CD4+ T cells in the crypts emerge as the major sources of IL-22, as demonstrated by studies using CD4CreIl22flox/flox mice [40], which specifically lack Il22 in T cells. The absence of IL-22 in CD4+ T cells results in an inability to restrain the proliferation of C. rodentium during the late phase of infection, leading to ~40% mortality. In these mice, C. rodentium invasion extends into deep crypts as the infection progresses. Histologically, IL-23-producing cDCs are mainly localized within tertiary lymphoid structures, colocalizing with abundant ILC3s [8, 40], whereas macrophages are predominantly situated within the crypts [42]. These macrophages can regulate the effector functions of IL-22-producing CD4+ T cells during the late phase of infection. Intriguingly, compared with Il23a−/− and Zbtb46-DTR mice, mice lacking macrophages, such as CX3CR1-DTR or MM-DTR mice, tend to succumb during the later phase of infection [23, 37, 41], underscoring the essential roles of both Th17 cells and CX3CR1+ macrophages in controlling C. rodentium infection during the late phase. It is tempting to speculate that differentiated Th17 cells in secondary lymphoid organs could interact with CX3CR1+ macrophages within crypts during the late phase of infection.

Additionally, it has been reported that IECs in crypts play a role in pathogen-derived antigen presentation via MHCII molecules, which in turn induces IL-22 production from