Dendritic cells (DCs), the most potent professional antigen presenting cells (APCs), serve as the sentinels in immune responses that integrate a wide array of incoming signals and convey them to lymphocytes, directing the appropriate immune responses (reviewed in []). DCs comprise a network of subsets with distinct developmental origins, surface phenotypes, and functions, which add further layers of complexity in the coordination of immune responses. Due to differences in the distinguishing markers of DC subsets between mice and humans, it has been extremely difficult to address whether there are functional equivalents between mouse and human DC subsets.
Mouse DCs were initially broadly classified into two major DC types: plasmacytoid DCs (pDCs) and conventional DCs (cDCs; earlier termed as myeloid DCs) (reviewed in []). cDCs have dendritic form and functions in steady state, and include migratory DCs and lymphoid DCs. Migratory DCs survey peripheral tissues, and have the ability to acquire antigen and subsequently migrate to the draining lymph nodes. Based on their expression of CD103, migratory DC subsets can be further separated into CD103+ and CD103− cDCs, which have been found in the skin, intestinal tract, liver, lung, and kidney. CD103+ DCs are primarily involved in cross-presentation to specific CD8+ T cells, induction of tissue-specific homing molecules on T cells and generation of regulatory T cells, while CD103− DCs show a higher capacity of antigen clearance and chemoattraction of leukocytes (reviewed in []). Lymphoid DCs reside in the lymphoid organs such as lymph nodes, spleen, and thymus and have been further divided into CD8−cDCs, encompassing CD4+ and CD4− subsets, and CD8+ cDCs (reviewed in []). CD8− DCs appear to preferentially induce type 2 responses and induce antigen-specific CD4+ T-cell immunity via the MHC class II pathway. In contrast, CD8+ DCs have been shown to have the unique ability to cross-present exogenous antigens on MHC class I molecules to CD8+ T cells to generate CTL responses (reviewed in []). Furthermore on activation, CD8+ DCs are major producers of IL-12 and induce vigorous inflammatory responses [], while in steady state, they have immune regulatory properties and play a crucial role in the maintenance of self-tolerance [[7, 8]]. In addition, under inflammatory or steady-state conditions, monocytes can give rise to different subsets of DCs such as TNF-α-iNOS-producing DCs. Monocytes-derived DCs can be found in the skin and the lamina propria of the gastrointestinal, respiratory, and urogenital tracts, and appear to fulfill an essential role in adaptive and innate immunity against pathogens (reviewed in []).
As described in mice, human DCs are also a heterogeneous cell population []. Because peripheral blood is the only readily available source, human DCs are frequently studied with cells derived in vitro from monocytes or from CD34+ hematopoietic progenitors. Human DCs have been extensively phenotyped and, like the situation in the mouse, are subdivided into CD123hi CD303 (BDCA-2)+ CD304 (BDCA-4)+ pDCs and into CD1c (BDCA-1)+, and CD141 (BDCA-3)+ mDC subsets (reviewed in []). However, monocyte-derived and CD34+ hematopoietic progenitor cell-derived DCs may differ considerably from the naturally occurring DCs present in vivo and information on the function of human primary DCs remains very rare, apparently because of their scarcity in blood and the difficulties in isolating them from other human tissues (reviewed in []). In three related articles, the studies undertaken by Radford and colleagues have now filled much of this knowledge gap [[12-14]]. Early studies investigating the function of distinct DC subpopulations were often confounded by lineage cell contamination and partial DC activation by the immunoselecting antibody. Radford and colleagues established a new protocol for the isolation of highly purified individual human blood DC subsets [], thus opening a new window for exploiting their biological functions. DCs can be isolated from peripheral blood by an initial immunomagnetic depletion of lineage-positive mononuclear cells. Subsequent flow cytometry-based cell sorting enables precision in selecting the desired DC subset resulting in sufficient yields of even the rarest subpopulations for most functional studies.
Continuing their studies in isolation methods, in this issue of the European Journal of Immunology, Radford and colleagues further examined the phenotype and function of CD1c+ DCs in response to Escherichia coli, characterized by low-level production of TNF, IL-6, and IL-12 but high production of the anti-inflammatory cytokine IL-10, and expression of the regulatory molecules indoleamine dioxygenase (IDO) and soluble CD25 []. Moreover, E. coli-activated CD1c+ DCs suppressed allogeneic T-cell proliferation via IL-10 []. It has been well established that E. coli and its cell wall component, LPS, are potent inducers of pro-inflammatory cytokine production and Th1 induction by DC subsets derived from monocytes in humans (reviewed in []). However, the manner in which naturally human CD1c+ DC subsets would respond to E. coli is poorly understood. Thus, these new results of Radford and colleagues [] add new elements and perspectives to the current understanding of human CD1c+ DCs for severe bacterial infections.
The big surprise of this study [] is the regulatory properties of human CD1c+ DCs after initial encounter with E. coli. The authors found that low levels of inflammatory cytokines but high levels of IL-10 production by CD1c+ DCs in response to E. coli, which is reminiscent of a tolerogenic DC phenotype []. Moreover, E. coli-activated CD1c+ DCs suppressed T-cell proliferation in an IL-10-dependent manner. These data are consistent with a previous study showing that liver CD1c+ DCs activated by LPS produce substantial amounts of IL-10, but comparatively low levels of inflammatory cytokines []. In combination with an increase in numbers of regulatory T (Treg) cells and IL-4-producing Th2 cells via IL-10, liver CD1c+ DCs induced less allo-MLR T-cell proliferation and promoted T-cell hyporesponsiveness []. The essential role of the circulating or organ-specific IL-10-producing DC subsets in protecting against systemic infections is also evident in several mouse models [[17, 18]].
Recent studies showed that tolerogenic DCs promote central or peripheral tolerance through various mechanisms, including T-cell deletion, induction of T-cell anergy and induction of Treg cells, expression of immunomodulatory molecules (e.g. PD-L1/2; heme oxygenase-1 (HO-1); HLA-G; CD95L; TNF-related apoptosis-inducing ligand (TRAIL); galectin-1 and DC-SIGN), and production of immunosuppressive factors (e.g. IL-10; TGF-β; IDO; IL-27, and NO) [[19-21]]. Several stimuli may influence the “decision” of mouse DCs to become tolerogenic, including recognition of apoptotic cells, interaction with splenic, liver, or pulmonary stromal cells, and exposure to soluble factors such as IL-10, TGF-β, vasoactive intestinal peptide (VIP), and vitamin D3 (reviewed in [[19, 22]]). In humans, evidence of regulatory DCs has been limited to work using monocytes-derived DCs that are prepared ex vivo in the presence of GM-CSF and IL-4. A recent report showed that, in the tumor environment, human Foxo3-expressing pDCs enforce conversion of CD8+ T cells into Treg cells []. However, the factors that modulate the plasticity of circulating or organ-specific human DCs, as well as the detailed mechanisms for the negative regulation of the T-cell response by these regulatory cells need to be fully investigated.
The immune response to a given pathogen is tightly regulated by multiple DC subsets that are equipped with unique PRRs to detect sophisticated microbial products. In mice, pDCs predominantly use TLR7 and TLR9 for nucleic acid sensing []. In contrast to CD8− cDCs, CD8+ cDCs highly express TLR3 but hardly express TLR7 or RIG-I-like helicases (RLHs), and so detect dsRNAs rather than ssRNAs []. Correlation of the human and mouse DC subsets has been hampered by differences in their defining markers. An earlier report by Robbins et al. indicates that CD141+ and CD1c+ myeloid DCs are the human counterparts of mouse CD8− and CD8+ cDCs analyzed by computational genome-wide expression profiling, respectively []. The new information reported in the study by Radford and colleagues [] highlights the marked differences in how human DC subsets respond to various pathogen infections. Earlier studies had shown that human CD1c+ DCs produce only limited amounts of IL-12p70, CXCL10, IL-10, and IFN-β, but induce TNF, IL-6, and IL-1-β production as well as augment Th1 responses after poly-I:C activation []. Similarly, individual triggering of TLR2, TLR4, TLR7, TLR8, or Dectin-1 is not sufficient to induce high levels of IL-10 by CD1c+ DCs [[14, 27]]. However, in the paper by Radford and colleagues [], intact E. coli was found to induce IL-10 production by CD1c+ DCs, even though TLR4 and Dectin-1 were only weakly expressed on these cells []. One possible explanation for the difference seen in the IL-10 production of CD1c+ DCs to E. coli might be the engagement of an unidentified PRR or cooperation of more than one receptor (such as TLR2 and NOD2). Given that innate recognition pathways are key controllers of DC function and act as a determinant of their properties and the findings in the paper by Radford and colleagues such as the role of Syk signaling in IL-10 secretion by CD1c+ DCs after E. coli activation [], the importance of the identification of the cellular and molecular mechanisms that modulate this plasticity of human DC subsets is emphasized.
In addition to the emerging data regarding CD1c+ DCs, much attention has also been turned toward the investigation of the functional capacity of the CD141+ DC subset (Fig. 1). Radford and colleagues reported the first detailed functional analysis of the human CD141+ DC subset []. They found that, in contrast to CD1c+ DCs, the CD141+ DC subset expresses high levels of TLR3, produces IFN-β and IL-12p70, and induces superior Th1 responses and CTL responses by cross-presenting necrotic cell antigens []. As pointed out by the authors, the CD141+ DC subset is the human functional homologue of the mouse CD8+ DC subset. Interestingly, in the spleens of primary human and humanized mice, Poulin et al. characterized a population of human DNGR-1 (CLEC9 A)+ CD141hi DCs resembling mouse CD8+ DCs in phenotype and function []. Furthermore, Bachem et al. showed that CD141+ DCs are the only cells in human blood that express the chemokine receptor XCR1 and react to the specific ligand XCL1 by mobilization of intracellular Ca2+ and by strong chemotaxis in vitro. More importantly, primary CD141+ DCs correspond to mouse CD8+ DCs and excel in cross-presentation of antigen when directly compared with other mDC subsets []. These findings suggest that human CD141+ DCs might be equivalent to the mouse CD8+ DC subset and constitute a homogeneous population. However, human CD141+ DCs have not been fully characterized and this will be an important area for future investigation.
Altogether, these recently published reports, including the latest [] in this issue of the European Journal of Immunology, highlight the importance of the description of DC subsets with distinct functions, as well as their plasticity in responding to extrinsic signals. The identification of the human DC subsets with similar functions to define mouse DC subsets should help fundamental immunological discoveries generated in mouse models to be more effectively translated into clinical practice []. Furthermore, understanding the principles by which DCs control immunity and tolerance provides an opportunity for the design of new therapeutic approaches to treat cancers and infectious diseases.