Type 2 conventional dendritic cell functional heterogeneity: Ontogenically committed or environmentally plastic?

Abstract

Conventional dendritic cells (cDCs) are sentinels of the mammalian immune system, sensing a wide range of danger and homeostatic signals to induce appropriately targeted T cell immune responses. Traditionally classified into two main subsets, cDC1 and cDC2, recent research shows that cDC2 exhibits significant heterogeneity and can be further subdivided. Studies in mice and humans show that, beyond their ontogeny, cDC2s acquire dynamic and tissue-specific characteristics influenced by local environmental signals, which impact their functions during homeostasis, inflammation, and infection. We propose the novel concept that tissue-derived signals and tissue plasticity can override preestablished developmental programming, thereby redefining developmental trajectories and cDC2 functionality. Ultimately, understanding cDC2 heterogeneity and plasticity has important implications for modulating T cell immunity in health and disease.

Functional heterogeneity of cDCs: the roles of ontogeny and plasticity.

Conventional dendritic cells (cDCs) are essential for inducing various T cell responses under different contexts, underscoring the heterogeneity of their functions (Box 1). The functional heterogeneity of cDCs is shaped by their ontogeny, characterized by distinct independent lineages derived from committed precursors that define different ‘cell subsets”, as well as by their plasticity—the ability to acquire functions in response to environmental cues that influence their “cell activation/polarization state or profile”. Two primary subsets—cDC1 and cDC2—originate from distinct lineage-restricted, bone marrow-derived progenitor cells. [1–3]. Each subset possesses specialized functions. cDC1s are particularly effective at eliciting CD8⁺ T cell responses and driving Th1 immunity, which are crucial for combating viral infections and tumors. The lineage-defining transcription factor for cDC1s is IRF8, which collaborates with other transcription factors such as Batf3, ID2, and NFIL3 to regulate their development and function. In contrast, cDC2s display broader functional versatility, influencing a wide array of T cell responses. They are essential for initiating Th2 and Th17 responses, maintaining inducible T regulatory cells (Treg), and even promoting Th1 and CD8⁺ T cell responses under certain conditions. The primary transcription factor for cDC2 differentiation is IRF4; however, conditional deletion of IRF4 only partially affects cDC2 development in mice, suggesting the involvement of additional regulatory mechanisms.

Box 1. Functional heterogeneity of conventional dendritic cells (cDCs) in orchestrating T cell responses.

cDCs serve as a crucial bridge between innate and adaptive immunity by sensing signals generated in the tissue and translating these signals to T cells in lymphoid tissues, thereby initiating specific adaptive T cell responses. In particular, cDCs provide naïve T cells with the essential signals for full activation, expansion, and polarization, allowing CD4⁺ T cells to differentiate into distinct functional subtypes, including T-bet⁺ Th1, RORγt⁺ Th17, Foxp3⁺ regulatory T (Treg), and GATA3⁺ Th2 cells [35]. Additionally, cDCs guide polarized CD4⁺ T cells in acquiring T follicular helper, effector, and memory capabilities [30, 47], or assist naïve CD8⁺ T cells in acquiring cytotoxic, effector, and memory functions [84, 86–88, 90]. cDCs accomplish these functions in the context of exposure to antigens from invasive microorganisms, allergens, tumor antigens, or auto/self-antigens. Through these interactions, cDCs can promote protective immunity to pathogens, enhance tumor immunity, or maintain immune homeostasis. However, they can also contribute to the development of allergies and autoimmune diseases.

Recent studies have identified additional subdivisions within cDC2s. These sub-classified categories are referred to as cDC2A and cDC2B [4–6] or, alternatively, as cDC2 and cDC3 by other groups [7–12], in mice and humans (refer to Table 1 for a summary of cDC2 subclassifications). While some researchers speculate that cDC2B and cDC3 may represent overlapping populations [7, 8] others have demonstrated that DC3s originate from LY6C⁺ monocyte-dendritic cell precursors (MDPs) and, therefore, should not be classified as cDCs, which are defined as deriving from common dendritic progenitors (CDPs) [6, 9]. CDPs originate from MDPs, which in turn derive from common myeloid progenitors (CMPs), progressively narrowing their potential for differentiation as they advance through hierarchical specification. CDPs give rise to cDC2s through a pre-cDC2 intermediate [1, 6]. However, alternative pathways for cDC2 development from lymphoid-derived pDC-like precursors have also been described, uncovering convergent pathways from lymphoid as well as myeloid origins [6, 13, 14]. Revisiting the classification of cDC2s and their precursors is necessary to integrate these diverse populations into a unified and non-overlapping framework. Additionally, it remains unclear whether the specification of cDC2s into various populations, including cDC2A and cDC2B, is driven by ontogenetic programming or influenced by signals from the tissue microenvironment. The latter would suggest that these categories may reflect activation or polarization states rather than distinct subsets. Some evidence points to environmental signals, such as that implicating microbial products from commensals or pathogens in playing a role in lineage specification [4, 15, 16]; by contrast, other studies propose that these subsets arise from distinct committed progenitors in the bone marrow [5]. Thus, further research is needed to determine whether cDC2A and cDC2B represent discrete subsets or cell states shaped by external stimuli. Moreover, functional cDCs studies have predominantly focused on migratory cDCs (including migratory cDC1 and cDC2), which travel from tissues to regional lymph nodes after acquiring a “mature” state, where they promote distinct CD4⁺ and CD8⁺ T cell responses [15–22]. However, secondary lymphoid organs (SLO), such as lymph nodes and spleen, also harbor resident populations of cDC1 and cDC2 that locally differentiate within the lymphoid tissue, and which exhibit immature and semi-mature states that are distinct from their migratory counterparts. Indeed, dimensionality reduction techniques such as UMAP, applied in single-cell transcriptomics, consistently segregate migratory from resident cDCs, independently of their cDC1 or cDC2 identity. This suggests substantial transcriptional differences that reflect functional and biological distinctions between migratory and resident cDCs [4, 12, 23]. The role of resident cDCs in T cell immunity remains less understood. Notably, the aforementioned cDC2 subsets, categorized as cDC2A and cDC2B, have been preferentially characterized within the resident cDC2 population of the spleen [4–6, 9]. Future research should investigate whether similar classifications apply to migratory cDCs, particularly migratory cDC2s, under homeostatic or pathogenic conditions across different tissues. It is also essential to determine whether these classifications result from distinct committed progenitors or reflect progenitor adaptations to environmental cues.

Table 1.

Summary of reports describing distinct type 2 conventional dendritic cell (cDC2) subsets in humans and mice. Ref, references.

Subsets	Species	Location	Characteristics	Ref.
ESAM+ cDC2 ESAM− cDC2	Mouse	Spleen (resident cDC2s)	- Notch2-dependent - Involved in the development of T follicular helper T cells to local antigens	[121,122,124]
			- Active LyzM promoter - Pro-inflammatory after microbial stimulation (e.g., TNFα, IL-12)
cDC2A cDC2B	Mouse	Spleen (resident cDC2s)	T-bet+, ESAM+ - Non-inflammatory	[4]
			RORγT+ CLEC12A+ CLEC10A+ - Pro-inflammatory after microbial stimulation (e.g. TNFα, IL-6)
cDC2A cDC2B	Mouse	Spleen (resident cDC2s)	T-bet+, ESAM+ - Non-inflammatory	[5]
			- CLEC12A+ - Active LyzM promoter - KLF4-dependent
CD11b− cDC2 CD11b+ cDC2	Mouse	Peripheral lymph node (migratory cDC2s)	- KLF4-dependent - Involved in Th2 responses	[17]
CD11b− cDC2 CD11b+ cDC2	Mouse	Peripheral lymph node (migratory cDC2s)	- IL-13-dependent, STAT6-dependent, KLF4-dependent - Involved in Th2 responses	[19]
ESAM+ cDC2 (spleen) cDC2 and DC3 (other tissues)	Mouse	Multiple tissues	- Only present in spleen, unrepresentative for cross-tissue network	[12]
			- cDC2s can be subcategorized by CLEC12A and CD301b expression - DC3 are FcγRIIB/III+
CD1C+_A cDC2 CD1C+_B cDC2	Human	Blood	- CD32B+ - Non-inflammatory	[7]
			- CD14+, CD163+, CD36+ - Pro-inflammatory
DC2 DC3	Human	Blood	- CD5+ - Non-inflammatory	[8]
			- CD14+, CD163+ - Pro-inflammatory
DC3	Human	Blood	- CD1c+ CD163+ CD88−, CD14+, S100A8+, S100A9+ - GM-CSF-dependent - Develop independently of cDC and monocyte progenitors	[11]

Although the subdivision of cDC2 into an increasing number of distinct subpopulations may reflect predetermined developmental pathways, emerging evidence introduces a novel perspective: environmental signals—such as microbe-derived signals, cytokines, alarmins, metabolites, and cell interactions— can dynamically influence their transcriptional profiles and refine their post-developmental trajectories, or “cell states”, along with their functional capabilities. Our viewpoint is that, rather than being rigidly dictated by developmental ontogeny, this dynamic adaptation enables cDC2s to modulate their cytokine production and surface marker expression by activating specific transcriptional programs, facilitating effective T cell immune responses across various contexts. We discuss how diverse environmental signals can shape cDC2 function and influence their ability to instruct T cell immunity. This plasticity of cDC2s allows them to tailor their functions in response to stimuli from pathogens, danger signals, or homeostatic conditions, fine-tuning immunity to meet environmental demands. This duality—rooted both in lineage and significantly influenced by local signals—highlights the importance of understanding the heterogeneity of cDC2s and their role in shaping T cell immunity in health and disease. Furthermore, insights into the plasticity and functional specialization of cDC2s mediated by environmental cues can enhance the development of targeted immunotherapies and improve strategies for treating cancers, autoimmune diseases, and allergies.

Plasticity and tissue adaptation of cDC2s: environmental influences and functional implications

IL-13-STAT6 and TSLP-STAT5 signaling programs cDC2s to induce Th2 cell responses

cDC2s are well known to induce Th2 cell responses against parasites or allergens [24, 25]. Pivotal studies in mice using DC-specific Cre/FL systems have established that the expression of the transcription factors IRF4 [26, 27] and KLF4 [17] are required for the Th2 cell-inducing function in cDC2s in models of parasite infection and allergen-induced airway inflammation. Recent research indicates that the cDC2B subset in the mouse spleen is KLF4-dependent [5, 14] and may follow an ontogenically determined developmental path [5]. However, KLF4-expressing cDC2s can also emerge in response to tissue-induced factors, specifically STAT6-activating cytokines such as IL-4 and IL-13. In mouse and human skin, IL-13 signaling, along with the functional activation of STAT6, IRF4, and KLF4, is necessary to support a gene program in cDC2s for the development of Th2 cell responses [19]. This was demonstrated through transcriptome analysis of cDC2s from mouse and human skin and the use of DC-specific Cre/FL systems to disrupt cytokine signaling in DCs. Furthermore, IL-13-dependent cDC2s with acquired Th2 cell functions can be phenotypically identified by low surface expression of CD11b [17, 19]. In summary, these studies suggest that IL-13 signaling activates a transcriptional program in cDC2s that enhances their functional specialization for Th2 cell induction. In tissues such as the skin, lungs, and gut, type-2 innate lymphoid cells (ILC2s) are a principal source of IL-13 [28]. Indeed, ILC2-derived IL-13 is crucial for the Th2 cell-inducing function of cDC2s in mice [19, 29, 30]. IL-13 in ILC2s is induced by cytokines IL-25, IL-33, and thymic stromal lymphopoietin (TSLP) [28]. These so-called alarmins are released by epithelial or other cells in response to damage, such as allergen exposure or helminth infection. Additionally, other mediators, such as neuropeptides and lipid mediators, can activate ILC2s [28]. While ILC2s are an early source of IL-13 in response to external stimuli that promote Th2 cell immunity, they can also produce homeostatic IL-13 in specific tissues, such as the mouse and human skin [19]. Thus, both homeostatic and induced IL-13 signaling from ILC2s can direct the functional programming of cDC2s for Th2 cell induction.

TSLP signaling also stimulates mouse and human cDC2s to prime Th2 cells [31–33] through STAT5 activation [33]. The loss of STAT5, achieved using a DC-specific Cre/FL system, abrogates TSLP signaling in cDC2s and their ability to induce Th2 responses in in vivo mouse models of Th2-driven skin contact hypersensitivity or lung rechallenge in ovalbumin (OVA)/alum-sensitized mice [33]. TSLP is produced by keratinocytes in atopic dermatitis lesions [31] and by non-hematopoietic cells in the lung, including epithelial cells, particularly during recall responses as in OVA/alum-sensitized mice [34]. Thus, TSLP signaling originating from barrier tissues can direct the functional programming of cDC2s, representing a potentially relevant mechanism for conditioning cDC2s in Th2-mediated skin disorders or chronic lung Th2 responses.

One aspect that remains unclear is the specific signals cDC2s provide to naïve CD4⁺ T cells to drive Th2 cell responses and how IL-13 and TSLP-driven programming—along with the activation of STAT6, IRF4, KLF4, and STAT5—promote these signals in cDC2s. It has been proposed that Th2 cell differentiation can occur through a default pathway when key cDC-derived signals that induce other Th pathways (i.e., the polarizing cytokines IL-12, interferon (IFN)-γ, IL-6, IL-10, TGF-β) are absent [24, 35]. In this context, STAT6, IRF4, and KLF4 might suppress the production of polarizing cytokines by various mechanisms. IL-13-mediated activation of STAT6, along with the expression of IRF4 and KLF4 by murine cDC2s, reduces their ability to detect bacterial and fungal components by repressing the expression of Toll-like receptors (TLRs) and C-type lectin receptors (CLRs) [19, 26]. TLR activation triggers the induction of pro-inflammatory and polarizing cytokines, wherein the TLR-interacting adapter protein MyD88 recruits several signaling molecules and transcription factors. Notably, IRF4 interacts with MyD88 and acts as a negative regulator of TLR signaling [36], while KLF4 collaborates with STAT6 to sequester coactivators required for NF-kB activation downstream of TLR signaling [37], as shown by in vitro assays and in vivo studies using IRF4- and KLF4-deficient mice treated with TLR ligands. Conversely, in the TSLP-STAT5-dependent programming, cDC2s upregulate OX40L, which limits IL-10 and IFN-γ expression in T cells [31, 32], potentially facilitating a default Th2 pathway. Additionally, TSLP stimulates cDC2s to express TARC (CCL17) and MDC (CCL22) [31, 33], which preferentially attract CCR4-expressing Th2 cells, enhancing the likelihood of cDC2-T cell interactions and promoting the development and maintenance of Th2 responses. Despite this, IL-13 and TSLP signaling enables cDC2s to upregulate costimulatory molecules and migrate to draining lymph nodes [19, 29, 31, 33, 38]. Consequently, these cDC2s can still provide adequate signals to T cells for autocrine IL-2 production and subsequent STAT5-mediated signaling, which is essential for the optimal expansion of newly activated T cells and the differentiation of Th2 cells [35, 39–44]. Additionally, cDC2s effectively prime Th2 cell responses in specific regions of secondary lymphoid tissues, particularly at the T-B cell border zone and in the perifollicular region, rather than in the T cell zone [42, 43, 45–48]. In vivo mouse models have shown that altering the location of cDC-T cell interactions can suppress Th2 responses to helminth parasites, highlighting the significance of immune cell spatial organization in Th2 polarization [48]. Since the T cell area concentrates cellular sources of polarizing cytokines such as IL-12 [24], and restricts IL-2 signaling through Treg cell-dependent consumption [49], compartmentalizing Th2 cell induction away from the T cell zone might be an immune strategy to favor Th2 differentiation by decreasing signals from polarizing cytokines while increasing IL-2 signaling [24]. An open question remains as to whether IL-13 signaling influences the ability of cDC2s to migrate to different regions of secondary lymphoid tissues, in addition to controlling their capacity to provide costimulatory and cytokine-derived signals to T cells.

Overall, recent evidence suggests that the IL-13-STAT6 and TSLP-STAT5 pathways activate transcriptional programs in cDC2s that suppresses their ability to produce polarizing cytokines in response to microbial or danger signals while promoting their capacity to stimulate Th2 cells (Figure 1). It is plausible that IL-4-dependent signaling plays a similar role to IL-13 signaling in cDC2s since both signal through related heterodimeric receptors, activating STAT6 [50]. IL-4 can be produced in tissues by eosinophils and basophils, especially during parasitic infections and allergies [51–55]. Like IL-13, IL-4 suppresses the production of various pro-inflammatory and polarizing mediators, particularly in monocytes and macrophages from mice infected with Th2-skewing pathogens. [37, 51, 56]. Supporting the idea that IL-4 signaling in cDC2s plays a role similar to IL-13 in supporting Th2-skewing functions, recent studies indicate that IL-4 is produced by basophils in the skin of both humans and mice during allergen-induced inflammation, where it signals in skin cDC2s to enhance their capacity for Th2 polarization and drive allergic skin inflammation [55]. Additionally, IL-33 may function like TSLP in cDC2s, because IL-33 signaling in lung murine cDC2s induces OX40L expression, suppresses polarizing cytokine production, and promotes Th2 cell skewing in response to inhaled allergens [57].

Figure 1. Functional specialization of type 2 conventional dendritic cells (cDC2s) for Th2 cell induction.

Activation of STAT6 by IL-13 or IL-4 upregulates transcription factors KLF4 and IRF4 in cDC2s, driving their functional specialization to support Th2 differentiation [17, 19, 26, 27, 55]. This is achieved by at least three independent mechanisms that ultimately reduce the ability of cDC2s to produce polarizing cytokines, such as IL-12, IL-6, and IL-23: (i) repression of Toll-like receptor (TLR) and C-type lectin receptor (CLR) expression via STAT6 signaling and IRF4/KLF4 activation [19, 26], (ii) negative regulation of TLR signaling by IRF4 interaction with MyD88 [36], and (iii) sequestration of coactivators necessary for NF-kB activation downstream of TLR/CLR activation through KLF4 cooperation with STAT6 [37]. Thymic stromal lymphopoietin (TSLP) activates STAT5 signaling in cDC2s, promoting Th2 differentiation by upregulating surface OX40L, which interacts with OX40 on T cells [31–33]. IL-33 can also induce OX40L expression in cDC2s [57]. Furthermore, IL-4/IL-13-driven STAT6 and TSLP-driven STAT5 signaling upregulate CD80 and CD86 expression on cDC2s [19, 31, 33], required for engaging CD28 on CD4⁺ T cells, induce autocrine IL-2 production and signaling, and promote initial Th2 lineage commitment [35, 39–42]. Type-2 innate lymphoid cells (ILC2s) are a key local non-T cell source of IL-13 [19, 28–30]. Basophils and eosinophils are the primary IL-4-producing non-T cells in tissues [51–54]. Alarmins TSLP, IL-33, and IL-25, released from epithelial cells during homeostasis or stress, activate ILC2s, basophils, eosinophils, and cDC2s [28, 31]. IL-1AcP, IL-1 accessory protein; IL-4Rα, IL-4 receptor alpha; IL-7Rα, IL-7 receptor alpha; IL-13Rα, IL-13 receptor alpha; ST2, suppression of tumorigenicity 2; TRAF6, TNF receptor-associated factor 6; TSLPR, TSLP receptor; γc, Common gamma chain. Figure created with Biorender.com.

TNFα- and IFN-driven signaling cooperate to program cDC2s to acquire Th1-like functions

cDC1s are an important source of IL-12. As such, cDC1s are commonly associated with the Th1-inducing cDC subset [58]. However, migratory cDC2 can also acquire the capacity to produce IL-12 thereby gaining Th1 cell-inducing functions [15, 16, 59–64]. Furthermore, IL-12-producing cDC2s play a significant role in preventing Th2 cell responses in the lung upon host detection of microbial contaminants in aeroallergens, as demonstrated in mice [15, 16]. Recent evidence in mouse models supports the concept that the ability of migratory cDC2s to produce IL-12 depends on the upregulation and activation of the transcription factors T-bet [15], IRF1 [65], and IRF8 [61].

T-bet expression is associated with the mouse spleen cDC2A subset [4], which is proposed to differentiate along an ontogenically determined developmental pathway [5]. However, in mice, T-bet expression in migratory cDC2s can be strongly induced in response to in vivo exposure to bacterial endotoxins (lipopolysaccharide [LPS]) and host activation of TLR4 [15, 16]. Additionally, other microbial-derived products or microbiota-dependent signals [4] likely exert similar effects, given the common signaling pathways that mediate various TLR-induced responses [66]. While cDC2s express TLR4 and can respond to LPS, in vivo LPS exposure induces T-bet expression and IL-12 production in murine migratory cDC2s, even when these cells lack TLR4, suggesting that direct LPS recognition is not required [15, 16]. Instead, these functions in cDC2s are regulated by inflammatory mediators produced by classical Ly6C^hi monocytes after they activate the differentiation program into monocyte-derived DCs (moDCs) [16]. In LPS-treated mice, blockade of the CCL2-CCR2 axis, which prevents monocyte recruitment, inhibits T-bet and IL-12 expression in cDC2s, highlighting the key role of monocyte-derived cells in this response [16]. moDCs are superior to other DCs in their detection of, and response to LPS due to their increased expression of LPS receptors (i.e., TLR4 and CD14) and as members of the MyD88-dependent signaling pathway. Their increased sensitivity to LPS is controlled by GM-CSF signaling, a cytokine produced locally in inflammatory settings and crucial for moDC differentiation [16, 67]. In vivo neutralization of GM-CSF, and thus, inhibition of moDC differentiation, indirectly inhibits cDC2 programming [16]. Thus, GM-CSF can guide the development of moDCs and enhance their sensitivity to LPS (and likely other microbial products), promoting the production of pro-inflammatory cytokines that indirectly activate cDC2s for subsequent upregulation of T-bet and acquisition of IL-12 production capabilities.

Among the pro-inflammatory cytokines released by moDCs following in vivo TLR4 engagement in mice, TNFα is essential for the indirect activation of cDC2s and the induction of T-bet and IL-12, as shown by the absence of these responses when TNF receptor expression on cDCs is lacking. TNFα signaling activates NF-κB and activator protein-1 (AP-1) transcription factors in cDC2s, which are necessary for upregulating T-bet [15]. However, TNFα alone is insufficient to induce T-bet in cDC2s in vitro, suggesting that additional factors are needed [4, 15]. In this context, cytokines that regulate T-bet in other cells, such as IL-12, IL-18, and IFNγ, emerge as potential candidates for future investigation. Indeed, in vitro culture with IFNγ strongly upregulates T-bet in virtually all mouse spleen cDC2s; importantly, this effect occurs regardless of their initial T-bet expression or their classification as cDC2A (T-bet⁺) or cDC2B (T-bet⁻) [4]. This finding suggests that IFNγ can induce T-bet independently of baseline T-bet expression or subset identity. However, DC2A cells tend to exhibit higher T-bet expression [4], suggesting that they may be better prepared to respond to these signals. This observation raises the possibility of a dynamic interaction between ontogenetic origin and tissue-driven adaptation, which warrants further investigation. Additionally, further studies are needed to identify the cellular sources involved in cDC2 activation. For instance, sustained in vivo IL-12 production by cDCs can be licensed by IFNγ-producing T cells in certain murine tumor models, suggesting that T cell crosstalk is crucial for efficient IL-12 production and effective Th1 cell responses [68]. Therefore, this axis might enhance the upregulation of T-bet in cDC2s and Th1 cells. Furthermore, NK cells and group 1 innate immune cells (ILC1s) are potent producers of IFNγ in tissues [69], while monocytes and macrophages highly secrete IL-12 and IL-18 upon microbial stimulation [70]. Another important question is how T-bet controls IL-12 production in cDC2s. In T cells, T-bet induces transcription by recruiting chromatin remodeling complexes to its target genes [71]. Thus, in cDC2s, T-bet may be necessary to promote epigenetic modifications of the genes encoding IL-12 p40 and p35 to enhance their expression.

Similar to the inability to produce IL-12 that has been observed in T-bet-deficient cDC2s [15], defects in the expression of the transcription factor IRF8 also impair the capacity of murine and human migratory cDC2s to secrete IL-12 and stimulate IFNγ responses in T cells [61, 72]. In murine monocyte and macrophage-like cell lines, gain- and loss-of-function assays have shown that IRF8 is essential for activating the promoter encoding IL-12 p40/IL-12B [73, 74]. Meanwhile, IRF8, in cooperation with IRF1, activates the promoter encoding IL-12 p35/IL-12A [74, 75]. It can be speculated that similar regulatory mechanisms occur in cDC2s as well. In murine cDC2, IRF8 can be induced by type 1 IFN [61], which can be induced at tonic amounts by microbiota [76] or further enhanced following viral infection and innate sensing of viral nucleic acids by intracellular TLRs such as TLR3, TLR7, TLR8, and TLR9 and intracellular nucleic acid sensors [77]. However, in other cells, such as murine macrophages, IRF8 can also be induced by IFNγ [78], which is generally elicited following broad TLR activation during microbial infection or tissue damage, along with subsequent production of innate cytokines (primarily IL-12 and IL-18) [79].

Overall, exposure to type 1 IFN or IFNγ induces and activates T-bet and IRF8, and most likely IRF1, in cDC2s. The joint function of these transcription factors appears to be crucial for promoting IL-12 synthesis in cDC2s when combined with stimuli that further activate NF-kB- and AP-1-dependent signaling (Figure 2) [15,16]. Microbial stimuli such as LPS or nucleic acids can activate NF-κB and AP-1 through direct activation of TLRs in cDC2s, at least in vitro. However, cDC2s are less potent than other innate cells, such as monocytes and, in particular, inflammatory moDCs, in detecting and responding to exogenous TLR ligands. Therefore, during in vivo microbial exposure, inflammatory monocytes or moDCs likely play a key role in microbial sensing and amplify the response by producing inflammatory mediators such as TNFα [16]. TNFα signaling in cDC2s efficiently activates NF-κB and AP-1 signaling pathways [15]. Current evidence suggests that the combined activities of TLR/TNFα-driven activation of NF-κB/AP-1 and IFN-mediated activation of T-bet, IRF1, and IRF8 synergistically stimulate the production of IL-12 in cDC2s during microbial exposure or infection (Figure 2) [15,16].

Figure 2. Functional specialization of type 2 conventional dendritic cells (cDC2s) for Th1 cell induction.

Combined activation triggered by TNFα, IL-12, and type 1 and type 2 interferon (IFN) drives the functional specialization of cDC2s to support Th1 cell differentiation by programming an enhanced capacity to produce IL-12. This is achieved by (i) induced expression of T-bet through TNFα-driven activation of NF-kB/AP-1 [15, 16] and IL-12/IFNγ-mediated signaling [4, 68] (ii) induced expression of IRF8 and IRF1 through IFN-mediated signaling and STAT1 activation [61, 72–74, 78]. Monocytes that differentiate into inflammatory monocyte-derived DCs (moDCs) in the presence of GM-CSF have a superior capacity to sense microbial products and indirectly initiate local cellular activation by releasing cytokines such as TNFα, IL-12, and IL-18 [16]. NK cells and type-1 innate lymphoid cells (ILC1) are the earliest sources of IFNγ in the tissues after exposure to IL-12 and IL-18 [28], while T cells are the primary source of IFNγ later on [79]. Plasmacytoid dendritic cells (pDC) are the main source of type 1 IFN upon bacterial and viral nucleic acid recognition [76]. IFNAR, type 1 IFN receptor; IFNγR, type 2 IFN receptor; IL-12R, IL-12 receptor; TLR, toll-like receptor; TNFR, TNF receptor. Figure created with Biorender.com.

CD40 licensing and IFN signaling can program cDC2s to promote CD8⁺ T cell responses.

Cross-presentation of exogenous antigens is essential for activating CD8⁺ T cell responses and, thus, for effective adaptive immunity against viruses and tumors. There is ample in vivo evidence demonstrating that the cDC1 subset is a key player in this process, because cDC1s possess a constitutively superior ability to cDC2s to engulf dead cell material and cross-present cell-associated antigens [80–83]. However, in murine models, cDC2s can also cross-present exogenous or cell-associated antigens in vivo, particularly during inflammatory conditions or infection [18, 61, 84–86]. For instance, during influenza virus infection, both murine lung-derived cDC1s and cDC2s can cross-present antigens and activate CD8⁺ T cells, with a notable division of labor between the two subsets: cDC1s support the generation of terminal effector CD8⁺ T cells that rapidly migrate to the lung to control viral replication [86] or facilitate the differentiation of tissue-resident memory T (Trm) cells during skin viral infections [87]. In contrast, cDC2s promote the generation and programming of fully functional central memory CD8⁺ T cells, which can undergo vigorous secondary expansion and exert effector functions upon antigen re-encounter [18, 84, 86, 88]. In murine infection models, the ability of migratory cDC2s to cross-present antigens to CD8⁺ T cells is controlled by CD40-dependent licensing [84], provided by activated CD4⁺ T cells [89]— a response that is absent in CD40-deficient cDC2s [84] or in the absence of CD4⁺ T cell help [89]. CD40 expression is upregulated in cDCs upon sensing pathogen molecules or inflammatory mediators. Consequently, the licensing of cross-presentation ability in cDC2s involves a two-step process: first, cDC2s upregulate CD40 after activation, followed by CD40 interactions that license this function.

Murine migratory cDC2s are capable of cross-presenting antigens during the replication stage of influenza virus infection [18, 84]. Notably, they continue to cross-present antigens to CD8⁺ T cells for extended periods of time beyond the elimination of the infectious pathogen during the contraction phase of the response [90]. This prolonged antigen cross-presentation by cDC2s is essential for programming long-lived central memory CD8⁺ T cells that provide efficient protection to the host upon secondary challenges [90]. In the mouse model of influenza virus infection, the ability of cDC2s to sustain antigen cross-presentation is dependent on specific antibodies that form immune complexes (IC) and target Fc-gamma receptors (FcγR) expressed on cDC2s, thereby enhancing the capture of exogenous antigens and directing them toward the cross-presentation pathway [90, 91]. Under resting conditions, cDC2s lowly express the high-affinity type I FcγR (CD64, encoded by Fcgr1a), resulting in a reduced capacity to capture ICs; however, CD64 expression can be induced on murine migratory cDC2s upon exposure to type I IFN [61] or IFNγ [92]. Thus, the detection of pathogens via TLRs and other pattern recognition receptors (PRRs), leading to IFN production, enhances the ability of cDC2s to cross-present antigens by upregulating FcγR-mediated uptake of exogenous antigens. This function is particularly relevant when antigens become scarce after infection. Furthermore, transcriptional and epigenetic analysis of cDCs from germ-free and IFN-I receptor (IFNAR)-deficient (Ifnar1^−/−) mice shows that prior tonic IFN type I signaling induced by the microbiota, is necessary to prepare cDC2s for cross-presentation and CD8⁺ T cell activation during future immune challenges [76]. Overall, the data support the notion that cross-presentation by migratory cDC2s is licensed by commensal microbiota and further enhanced by external signals produced during inflammation and infection, especially via CD40 interactions and IFN signaling. This regulation could ultimately influence the quality and functionality of antiviral CD8⁺ T cell memory responses (Figure 3). However, because this work was conducted in mouse models, further research is necessary to investigate these processes in humans.

Figure 3. Functional specialization of type 2 conventional dendritic cells (cDC2s) for CD8⁺ T cell induction.

Cross-presentation by cDC2s, and the ability to efficiently present exogenous antigens to CD8⁺ T cells, is licensed by CD40:CD40L interactions and interferon (IFN) signaling. Specifically, two steps are required: (i) IFN-mediated signaling and STAT1 activation induces expression of high-affinity Fc-gamma receptors (FcγR) for IgG (CD64), which increases the uptake of immune-complexed antigens [61, 90]. (ii) Engagement of CD40 on activated cDC2s by activated CD40L-expressing CD4⁺ T cells induces positive signaling that promotes antigen processing and cross-presentation [18, 84, 89]. IFNAR, type 1 IFN receptor; IFNγR, type 2 IFN receptor; ILC1, type 1 innate lymphoid cell; MHC class I, major histocompatibility complex class I; pDC, plasmacytoid dendritic cells; IFN-I, type I IFN. Figure created with Biorender.com.

TGFβ signaling programs cDC2s towards Th17 and Treg stimulatory functions.

Although Foxp3⁺ Treg and RORγt⁺ Th17 cells play important roles in many different tissues and organs, both T cell subsets play essential roles in intestinal homeostasis (reviewed in [93]). Tregs make up approximately 10% of CD4⁺ T cells in most mouse tissues, but their frequency rises to 20–30% in the intestinal lamina propria [94]. Th17 cells typically develop in response to pathogen exposure in barrier tissues such as skin or respiratory tract, but they are enriched in the gut under steady conditions [93]. Homeostatic conditions in the gut favor Treg and Th17 cell development, influenced by commensal microbiota and food antigens [93]. In vivo studies in mouse models have concluded that the differentiation of gut homing Treg cells that maintain tolerance to commensal bacteria and food antigens requires signaling mediated by TGFβ and all-trans retinoic acid (ATRA) produced from dietary vitamin A [94–99]. Additionally, tryptophan metabolite Kynurenine can also promote Treg cell differentiation [100]. Intestinal Th17 responses are crucial for mucosal barrier integrity and depend on TGFβ and IL-6 signaling [99, 101]. Thus, TGFβ signaling regulates the balance between Th17 and Treg cell differentiation through integrated signals. This applies to the gut, as discussed below, though distinctions in other tissues and contexts will not be addressed here.

Intestinal cDCs drive both Treg and Th17 cells during steady state [20, 97, 98, 101–103]. Recent studies in mice indicate that the outcomes are influenced by gut-specific homeostatic signals modulating cDC function. As in other tissues, cDCs in the intestinal lamina propria consist of cDC1 and cDC2 subsets. While CD103 typically identifies cDC1, about 50% of lamina propria cDC2s also express CD103, resulting in two phenotypes: CD103⁺ and CD103⁻ cDC2s. cDC1s promote gut-homing Treg differentiation by producing TGFβ and ATRA [96–98]. Additionally, cDC1s express indoleamine 2,3-dioxygenase 1 (IDO1), which generates the tryptophan metabolite kynurenine [104]. Kynurenine binds to aryl hydrocarbon receptor (AhR) in T cells, potentiating the role of TGFβ and enhancing Treg cell differentiation [100]. Thus, unlike cDC1s in other tissues, intestinal cDC1s are strongly influenced by the microenvironment to promote Treg differentiation, making them a tolerogenic subset in the gut. In contrast, CD103⁺ cDC2s induce both Treg and Th17 cell responses [20, 98, 101, 104]. In short, cDC1s and CD103⁺ cDC2s cooperate and likely play a redundant role in the induction of Treg cells in the gut [98, 104], but CD103⁺ cDC2s also uniquely facilitate Th17 cell responses [20, 103].

Recent studies show that local cues dictate the ability of gut cDC2s to promote Th17 versus Treg cell differentiation. Treg and Th17-inducing functions by cDC2s require TGFβ signaling, because deletion of TGFβ receptor activity in DCs using DC-specific Cre/FL systems in mice impairs CD103⁺ cDC2 differentiation, leading to decreased generation of Treg cells that promote tolerance to food antigens and reduced frequencies of lamina propria Th17 cells [102]. Lack of TGFβ signaling in gut DCs correlates with increased local IFNγ production [105], suggesting that TGFβ stimulation suppresses the ability of cDC2s to promote Th1 cell responses while poising their Treg and Th17-inducing functions. The mechanism underlying these functions in cDC2s remains to be elucidated, but one possibility is that TGFβR signaling in CD103⁺ cDC2s could promote sustained TGFβ production via a positive feedback circuit [106]. Still, the initial cellular source and the defined niche in the intestinal mucosa that promotes the selective conditioning effect of TGFβ on cDC2 remains unknow.

Although TGFβ receptor activity in murine cDC2 is required for Treg and Th17 cell induction, the generation of CD103⁺ cDC2 with solely pro-Treg cell/tolerogenic capacity additionally requires intra-epithelial location/interactions [107], β-catenin signaling [108], local production and signaling by ATRA [107, 109, 110], and the tryptophan metabolite kynurenine [104]. The intra-epithelial location within the intestinal mucosa [107] and constitutive β-catenin signaling [108] are both associated with the acquisition of the tolerogenic phenotype in CD103⁺ cDC2s, while suppressing their ability to induce Th17 cell responses. TGFβ signaling upregulates CD103 [111] and the expression of CD103 and E-cadherin in intestinal cDC2s promotes adhesion to epithelia through interaction with E-cadherin [111]. Since E-cadherin-mediated cell adhesion can promote β-catenin signaling [112], it can be speculated that intra-epithelial location and E-cadherin-mediated interactions may regulate β-catenin activation in CD103⁺ cDC2s. In any case, using DC-specific Cre/FL systems in mice, β-catenin signaling is required to induce retinal dehydrogenase 2 (RALDH2) encoded by Aldh1a2. This enzyme is crucial for enabling intestinal DCs to produce ATRA and directly present it to interacting CD4⁺ T cells, thereby promoting Treg cell differentiation. [108]. Additionally, β-catenin–mediated signaling suppresses the ability to produce the Th17 polarizing cytokines IL-23 and IL-6 in cDC2s [108]. In a subsequent step, ATRA signaling in cDCs further induce the expression of RALDH2, thus promoting sustained ATRA production in a positive feedback loop [110].

Additionally, studies in vitamin A-deficient mice and in vitro cultures with retinoic acid show that ATRA induces a transcriptional programming in intestinal cDC2s that suppresses proinflammatory NF-κB-dependent gene expression [109]. Thus, β-catenin signaling initiates an ATRA-driven positive feedback loop in intestinal CD103⁺ cDC2s, which is crucial for the induction of Treg cells and suppression of Th17 responses (Figure 4). Furthermore, cDC2s can activate the IDO1 pathway and acquire tolerogenic functions through the action of kynurenine signaling via AhR [104] (Figure 4). Kynurenine is highly produced by IDO⁺ cDC1s that have regulatory functions [104], a phenotype that has been shown in tumor microenvironments, and is induced via β-catenin–mediated signaling [113]. Thus, kynurenine/IDO⁺ cDC1 and cDC2 can establish communications in which cDC1 cells extend their immune capacity to cDC2s. This has been demonstrated in the context of tryptophan metabolism [104], but more research is needed to test whether this communication and transfer of functions might also be a regulatory mechanism in other contexts. Recent studies identified RORγt-expressing antigen-presenting cells (APCs) in the gut of mice as being essential for inducing specific subsets of inducible Tregs that also express RORγt and promote tolerance to dietary antigens [114] and gut microbiota [114–117]. Although these APCs express canonical markers of migratory cDCs, including CD11c, Zbtb46, CCR7, and high amounts of MHCII, their precise lineage remains under debate. Initially, they were thought to belong to the ILC lineage [116], but recent evidence seems to rule out this possibility [114, 115]. Additionally, Clec9a-Cre manipulations in mice, which selectively target the intestinal cDC1 lineage [118], show no impact on the functionality of RORγt⁺ APCs [115], suggesting they are unrelated to cDC1s. Some authors instead propose that they may belong to the cDC2 lineage [119] or represent a distinct category of APCs with as-yet-undefined characteristics [120]. Further research is necessary to clarify their lineage and define the requirements for their development.

Figure 4. Functional specialization of type 2 conventional dendritic cells (cDC2s) for regulatory T (Treg) cells and Th17 cell induction.

TGFβ-driven signaling induces the functional specialization of cDC2s to support Th17 and Treg cell differentiation [102, 105] by, in turn, programming their ability to produce TGFβ [106]. Integration of additional signals specifies their differential ability to induce Treg or Th17 cell responses. The acquired capacity to produce all-trans retinoic acid (ATRA) from vitamin A [107, 109, 110] and kynurenine from tryptophan [104] enhances the ability of cDC2s to promote Treg cell differentiation. Production of ATRA requires expression of retinal dehydrogenase 2 (RALDH2; encoded by Aldh1a2), which is upregulated in cDC2s by β-catenin [108] and ATRA signaling [107, 110]. Production of kynurenine requires expression of indoleamine 2,3-dioxygenase 1 (IDO1), which is upregulated in cDC2s by kynurenine signaling through the aryl hydrocarbon receptor (AhR) [104]. Also, the production of IL-6 and IL-23 (together with TGFβ) by cDC2s after microbial recognition promotes Th17 cell differentiation [102, 121, 122]. Activation of Notch2 receptor signaling in cDC2s has been implicated in mediating optimal IL-6 and IL-23 production for Th17 cell differentiation [121, 122]. NICD, notch intracellular domain; RAR/RXR, retinoic acid receptor/retinoid X receptor; TGFβR, TGFβ receptor. Figure created with Biorender.com.

Although the integration of β-catenin and ATRA- and kynurenine-driven signaling with TGFβ signaling governs the tolerogenic properties of cDC2s, Notch2 receptor signaling instead appears to enable cDC2s to induce Th17 cell responses in the gut, as demonstrated in mice using DC-specific Cre/FL systems and infection models (Figure 4) [121, 122]. When Notch receptors interact with their ligands, Delta or Jagged, on neighboring cells, this leads to receptor cleavage and the release of the Notch intracellular domain (NICD), which subsequently translocates to the nucleus to regulate transcription [123]. Notably, Notch2-expressing cDC2s have been associated with the cDC2A subset in the mouse spleen [4, 5, 121, 122]. This subset is described by some as an ontogenically determined lineage [5], while others suggest it is exclusive to the spleen microenvironment [12], where splenic fibroblasts expressing the Notch2 ligand Delta-like 1 (DL1) support their development [121,122,124]. Further research is needed to elucidate the regulation of Notch receptor activity in cDC2s, particularly concerning the relative influence of ontogeny versus tissue-specific signals.

Concluding remarks.

While cDC2s arise from a common progenitor pool, their differentiation is influenced by a complex interplay of ontogenetically determined constraints and environmental signals. Intrinsic factors, including transcriptional regulation and epigenetic modifications, guide the differentiation of cDC2 progenitors and establish initial functional potential of cDC2s. However, as cDC2 progenitors seed the tissues, they encounter a myriad local signals that further refine their developmental trajectory and functional capabilities. Further research should seek to understand how developmental constraints interact with environmental influences, leading to the emergence of tissue-specific cDC2 subsets with distinct phenotypic and functional properties (see Outstanding Questions). For instance, tissue-resident cDC2s may develop traits that enhance their ability to respond to local homeostatic signals, maintaining a symbiotic relationship with commensal organisms. Conversely, these cells can alter their developmental trajectory, acquiring distinct transcriptional properties and functionalities in response to invasive pathogens, thereby inducing protective T cell responses. This dynamic nature suggests that the developmental ontogeny that defines distinct cDC2 ‘subsets’ might not rigidly dictate their fate; rather, cDC2s and their progenitors may exhibit significant tissue plasticity, allowing them to adapt their functions to specific environmental cues and delineate transcriptional programs that reflect distinct activation or polarization ‘states.’ This opinion emphasizes the role of tissue-derived factors in controlling the functional diversification of cDC2s, which must meet the demands of the prevailing tissue conditions. Thus, our view is that the function of the cDC2 pool is not predetermined but is dynamic and adaptable to changing local conditions. A deeper understanding of this area could have therapeutic implications, allowing the manipulation of cDC plasticity to design targeted treatments for T cell-mediated diseases or cancer. Limitations will include how to direct therapies to specific tissues and design precise strategies to selectively modulate cDC2 functions, avoiding unintended targets.

Outstanding questions:

• What specific environmental signals are vital for the differentiation and functional adaptation of cDC2s in various tissues? How do these signals interact with intrinsic developmental programming to shape T cell responses?

• Can cDC2s, similar to CD4⁺ T cell subsets, be categorized and identified by the upregulation of specific transcriptional pathways associated with different functionalities tailored to develop corresponding T cell responses? If so, are these transcriptional programs regulated by intrinsic ontogeny or modulated by dynamic environmental cues, allowing cDC2s progenitors to take different paths and adopt distinct transcriptional programs depending on external signals? These questions can be addressed by studying cDC2 trajectories under homeostatic versus non-homeostatic conditions, i.e. in response to microbial or inflammatory stimuli. This may help identify key environmental signals and precursor contributions to cDC2 function.

• Can we define ‘subsets’ versus ‘cell polarization states’ in cDC nomenclature to distinguish between processes that are ontogenically programmed by committed, defined precursors and those driven by external stimuli? ‘Subsets’ could refer to stable populations with core, intrinsic characteristics, while ‘polarization states’ could represent transitional conditions shaped by changing environmental cues, where the same cDC precursors can adopt different functional directions based on external signals, influencing their T cell-inducing capacities.

Highlights.

• Conventional dendritic cells (cDCs) are the professional antigen-presenting cells of the mammalian immune system, classified into two primary subsets: cDC1 and cDC2. These subsets arise from distinct bone marrow-derived precursors. cDC1 and cDC2 have specialized functions in driving T cell responses: cDC1s are involved in type-1 immunity to intracellular pathogens, while cDC2s exhibit diverse functions linked to their heterogeneity.

• Recent studies have identified additional subdivisions within cDC2s, highlighting distinct ontogeny. Other studies show another new aspect: cDC2s possess remarkable plasticity, enabling them to adapt to environmental cues and acquire specific transcriptional programming.

• Current debates focus on whether the functional heterogeneity of cDC2s is shaped by ontogenically distinct subsets or by the adaptability of precursor differentiation in response to signals from the tissue. This discussion is highlighted by the contrast between ontogenetic studies, which primarily examine spleen-resident DCs in steady-state conditions, and functional studies that generally investigate migratory cDCs responding to signals from their tissues.

• Future research should investigate whether cDC2 subclassifications are consistent across different tissues and how environmental stimuli influence their development and function. Understanding these factors is crucial. We propose that the adaptability and plasticity of cDCs significantly shape their ability to instruct a variety of T cell responses.

Significance box.

Unlike other dendritic cell subsets, type 2 conventional dendritic cells (cDC2s) drive a diverse range of T cell responses due to their broad functional versatility. Understanding the ontogeny and plasticity of cDC2s is crucial for comprehending their functional heterogeneity. This knowledge is essential for unraveling their complex roles in T cell responses and developing innovative, targeted therapeutic strategies to modulate immune functions.

Glossary:

Central memory CD8⁺ T cells

Circulate through blood and secondary lymphoid organs, providing long-term, rapid systemic (body-wide) immunity by responding to previously encountered antigens

Conventional dendritic cells (cDCs)

subset of dendritic cells (DCs) primarily responsible for capturing, processing, and presenting antigens to T cells, playing a central role in initiating and regulating adaptive immunity. They differ ontogenically and functionally from monocyte-derived DCs (moDCs), which arise from monocytes during inflammation, and plasmacytoid DCs (pDCs), which specialize in antiviral responses and type I interferon production. cDCs are often divided into two main subtypes: cDC1 and cDC2

C-type lectin receptors (CLRs)

family of pattern recognition receptors (PRRs) that detect carbohydrate structures on pathogens and self-antigens. CLRs trigger phagocytosis, cytokine production, and adaptive immune modulation. Examples: Dectin-1, Dectin-2, DC-SIGN, langerin, mannose receptor, and Mincle

DC-specific Cre/FL system

A genetic tool using Cre-loxP recombination to selectively modify genes in DCs. Cre, driven by a DC-specific promoter, recombines FL (floxed) alleles — genes flanked by loxP sites — to induce gene deletion, activation, or modification, enabling DC-specific functional studies

Dimensionality reduction techniques

Computational methods for simplifying high-dimensional data while preserving key patterns. UMAP is widely used in single-cell transcriptomics to visualize gene expression in 2D or 3D, clustering cells based on their gene expression profiles

Migratory cDCs

express CCR7 and travel from peripheral tissues to draining lymph nodes via afferent lymphatics, guided by the chemokines CCL19 and CCL21. Subtypes include migratory cDC1 and cDC2, crucial for activating T cells

Ontogeny (of the immune cells)

developmental or differentiation process through which hematopoietic stem cells give rise to various immune cell lineages, population types, or subsets, each acquiring a distinct phenotypic, transcriptional, and functional identity

Plasticity

The ability of cell populations or subsets to adapt and differentiate into various functional states in response to environmental signals, beyond their original developmental lineage. This adaptability allows them to acquire specialized functions based on cues from their surroundings, under either homeostatic or pathogenic conditions

Resident cDCs

differentiate and remain in secondary lymphoid tissues like the spleen and lymph nodes. They lowly express CCR7 and exhibit an immature or semi-mature phenotype. They can present antigens to local T cells but their role in T cell responses is less defined compared to migratory cDCs. Subtypes include resident cDC1 and cDC2

Terminal effector CD8⁺ T cells

Fully differentiated, with potent cytotoxic and cytokine-producing abilities. They provide immediate defense by killing infected or malignant cells but have limited proliferative capacity and tend to undergo apoptosis post-function

Tissue-resident memory T (Trm) cells

Memory T cells residing in non-lymphoid tissues (e.g., skin, mucosa), offering immediate, long-term local protection by responding to reinfection at the initial pathogen entry site

Toll-like receptors (TLRs)

PRRs recognizing pathogen-associated (PAMPs) or damage-associated molecular patterns (DAMPs), activating pathways for cytokine and chemokine production. Surface TLRs (1, 2, 4, 5, 6) detect a variety of microbial components (e.g., LPS, lipoproteins, flagellin); endosomal TLRs (3, 7, 8, 9) detect nucleic acids