Volume 102, Issue 4 p. 1003-1016
Review
Free Access

Granulocytes as modulators of dendritic cell function

Annelot Breedveld

Annelot Breedveld

Department of Molecular Cell Biology and Immunology, VU University Medical Center, Amsterdam, The Netherlands;

Search for more papers by this author
Tom Groot Kormelink

Tom Groot Kormelink

Department of Experimental Immunology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands; and

Search for more papers by this author
Marjolein van Egmond

Corresponding Author

Marjolein van Egmond

Department of Molecular Cell Biology and Immunology, VU University Medical Center, Amsterdam, The Netherlands;

Department of Surgery, VU University Medical Center, Amsterdam, The Netherlands

VU University Medical Center, De Boelelaan 1108, 1081 BT, Amsterdam, The Netherlands. E-mail: [email protected]

Search for more papers by this author
Esther C. de Jong

Esther C. de Jong

Department of Experimental Immunology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands; and

Search for more papers by this author
First published: 01 October 2017
Citations: 42

These authors contributed equally to this work.

These authors contributed equally to this work.

Abstract

Effector T cell development is directly driven by APCs, in particular, by antigen-primed dendritic cells (DCs). Depending on the pathogenic stimulus and the microenvironment, DCs induce proliferation and polarization of naive CD4+ T cells into different effector subsets, such as Th1, Th2, Th17, or regulatory T cells (Tregs). During inflammation, DCs are found in close proximity to other innate immune cells, including all granulocyte subtypes, which potentially influence the immunomodulatory capacities of DCs. Neutrophils, eosinophils, and basophils are rapidly recruited into infected tissues where their main function is to eliminate invading pathogens. Mast cells are tissue-resident granulocytes that also contribute to host defense against pathogens but have, thus far, primarily been associated with their detrimental roles in allergic diseases. Although granulocytes have always been considered essential in innate immunity, their ability to influence the development of adaptive immunity has long been overlooked. This view is now changing, as multiple studies showed significant modulating effects of granulocytes on key players of adaptive immunity, including DCs and lymphocytes. Neutrophils, eosinophils, basophils, and mast cells regulate recruitment and activation of DCs through the release of mediators or via direct cell–cell contact, thereby influencing antigen-specific T cell responses. In this review, we will summarize the current knowledge on the impact of granulocytes on DC functioning and the subsequent putative consequences of this cross-talk on T cell proliferation and polarization. Together, this overview underscores the importance of granulocyte–DC communication to establish optimal immune responses.

Abbreviations

  • BALF
  • bronchoalveolar lavage fluid
  • BCG
  • bacillus Calmette-Guérin
  • BMDC
  • bone marrow-derived dendritic cell
  • CD
  • cluster of differentiation
  • cDC
  • conventional dendritic cell
  • CEACAM1
  • carcinoembryonic antigen-related cell adhesion molecule 1
  • DC
  • dendritic cell
  • DC-SIGN
  • dendritic cell-specific ICAM-3-grabbing nonintegrin
  • EDN
  • eosinophil-derived neurotoxin
  • EPX
  • eosinophil peroxidase
  • EV
  • extracellular vesicle
  • FoxP3
  • forkhead box P3
  • LC
  • Langerhans cell
  • LT
  • leukotriene
  • Mac-1
  • macrophage 1 antigen
  • MBP
  • major basic protein
  • MHC-II
  • MHC class II
  • MoDC
  • monocyte-derived dendritic cell
  • MPO
  • myeloperoxidase
  • NE
  • neutrophil elastase
  • NET
  • neutrophil extracellular trap
  • OX40L
  • OX40 ligand
  • PAF
  • platelet-activating factor
  • PD-1
  • programmed cell death 1
  • PD-L1
  • programmed cell death ligand 1
  • pDC
  • plasmacytoid dendritic cell
  • slanDC
  • 6-sulfo LacNAc+ dendritic cell
  • Treg
  • regulatory T cell
  • TSLP
  • thymic stromal lymphopoietin
  • WT
  • wild-type
  • Introduction

    DCs are specialized APCs that play a key role in the initiation of antigen-specific immune responses, thereby forming an important link between innate and adaptive immunity [1, 2]. Multiple DC subsets have been described in both men and mice. In humans, pDCs (CD11cCD123+) and distinct subsets of cDCs (CD11c+CD123) are identified, such as CD1c+, cross-presenting CD141+, CD1cCD141, and recently described siglec6+ DCs [3, 4]. In addition, specific DC subsets are present in various tissues, such as LCs in the epidermis, CD16+LacNAc+ (slan) DCs in the dermis and blood, and CD103+ DCs at mucosal areas [5]. At inflammatory sites, moDCs may be present [68]. Although similar subsets exist in mice, discriminatory markers used to define human DC subsets are not found on murine DC subsets and vice versa, complicating extrapolation of data obtained in murine (DC) models to humans. Subset similarities have been described, however; e.g., human CD1c+ cDCs are equivalent to murine CD11b cDCs, and human CD141+ cDCs are identified in mice as CD8+/CD103+ cDC [9, 10]. The distinct DC populations differ in their ontogeny, localization, and function but are not further outlined in detail here (extensively reviewed in refs. [3, 9, 11, 12]).

    In an immature state, DCs serve as sentinels for invading pathogens [1]. After recognition and uptake of pathogens, DCs migrate to draining lymph nodes, where pathogenic antigens are processed, and DCs will mature [characterized by the up-regulation of MHC molecules and costimulatory molecules (e.g., CD80 and CD86)]. Mature DCs have the ultimate capacity to activate antigen-specific naive T cells in the draining lymph nodes to initiate adequate adaptive immune responses [1]. Importantly, the type of DC-activating stimulus is decisive for the polarization of naive CD4+ Th cells into a final effector phenotype. Different Th cell subsets are involved in the clearance of selective sets of pathogenic microorganisms (or are associated with distinct chronic inflammatory disorders). For example, Th1 cells develop in response to intracellular pathogens (such as viruses and mycobacteria), whereas extracellular pathogens mostly initiate DC-driven Th2 (e.g., parasites) or Th17 cell development (e.g., bacteria and fungi) [13]. In contrast, Tregs suppress unnecessary, persistent immune activation or pathogen-specific immune responses once the pathogen is cleared [14].

    Interestingly, in recent years, it is increasingly recognized that the presence of accessory cells facilitates DC-driven induction of early and appropriate adaptive immune responses. Granulocytes are important examples of such accessory cells that can significantly influence DC function. Generally, granulocytes have long been exclusively considered important players during innate immune responses, but recently, their potential role in contributing to adaptive immune responses has been acknowledged [15, 16]. Circulating granulocytes have a relatively short lifespan of up to several days. However, prolonged lifespan is observed when these cells are rapidly recruited to sites of infection [17]. Neutrophils are the most abundant circulating granulocytes, whereas eosinophils and basophils represent minor populations. Mast cells are long-lived tissue-resident cells that are physiologically found in virtually all vascularized tissues. Moreover, all granulocyte subsets have been shown to migrate to peripheral and lymphoid tissues during inflammation [1822]. This tissue distribution enables granulocytes to encounter, interact, and communicate with DCs and in such a way, act as important initiators, as well as regulators, of innate and adaptive immune responses [17, 23]. Here, we will review how the different granulocyte populations contribute to the recruitment, maturation, and/or modulation of DCs and subsequent induction of adaptive immune responses.

    NEUTROPHILS

    Neutrophils are the most abundant type of leukocytes, comprising 50–70% of the total circulating human WBC count [24, 25]. They are continuously generated in the bone marrow, achieving a production of 1–2 × 1011 cells/d in an adult human [26]. Mature neutrophils, characterized by a segmented nucleus and a cytoplasm enriched with secretory granules and vesicles, are released from the bone marrow into the blood, where they are readily available for rapid recruitment into peripheral and lymphoid tissues in case of infection [18, 27, 28]. Traditionally, neutrophils have been considered as short-lived cells with a lifespan of 8 h in the human circulation. This dogma was challenged a few years ago [29], as labeling of human neutrophils with deuterium oxide in vivo supported that neutrophils may have a t1/2 of ∼5 d [17, 24, 29]. Additionally, a longer lifespan was observed at sites of inflammation, where neutrophils get activated by cytokines and bacterial components, which delays apoptosis [30]. This ensures the prolonged presence of activated neutrophils at inflammatory sites to perform complex activities to eliminate pathogens and to contribute to resolution of infection [24].

    Neutrophils can phagocytose and kill intracellular pathogens, mediated through granular antimicrobial proteins, such as MPO, NE, cathepsin-G and lactoferrin, and oxidative stress [24, 31]. Moreover, degranulation of neutrophils facilitates extracellular killing of nearby pathogens through the release of granular antimicrobial proteins into the environment [24, 25]. A more recently discovered mechanism to trap and kill pathogens is the formation of NETs, which consist of DNA to which histones and granular proteins are attached [32]. Unfortunately, NETs can also cause major tissue damage [24], and they have been associated with the pathogenesis of a wide range of noninfectious pathophysiological conditions, such as autoimmune diseases (rheumatoid arthritis, systemic lupus erythematosus, and diabetes), vascular diseases (atherosclerosis and vasculitis), cancer, and acute injuries [3337].

    In addition to this first line of innate immune defense, activated neutrophils have been shown to release chemotactic factors that aid in the migration of both innate and adaptive immune cells, including DCs, to sites of infection and nearby lymph nodes (Fig. 1) [3841]. Neutrophil-derived CCL3, CCL4, CCL5, and CCL20 induced the migration of murine BMDCs in vitro [39, 40]. Correspondingly, infection with Leishmania major led to a neutrophil-derived, CCL3-dependent recruitment of moDCs, dermal DCs, and LCs to the dermis of Ccl3−/− mice [39]. Other murine in vivo studies further implicated a role for neutrophils in DC migration upon infection with pathogens [38, 41, 42]. Both Mycobacterium tuberculosis and Aspergillus fumigatus enhanced migration of CD11b+CD11c+ DCs from lungs to mediastinal lymph nodes in WT mice compared with mice that had been treated with neutrophil-depleting antibodies, anti-Gr-1 or anti-Ly6G [38, 41], whereas Escherichia coli-infected neutrophils induced migration of dermal DCs containing neutrophil-derived material from the footpad to the draining lymph nodes [42]. Additionally, in a contact hypersensitivity model, CD11c+ DC migration from the skin to draining lymph nodes was enhanced in WT mice compared with neutrophil-deficient (Mcl-1ΔMyelo) mice during the initiation of inflammation [40]. Surprisingly, infection with Leishmania mexicana in the dermis of WT mice impaired the recruitment of CD45+CD11c+Ly6C+CD11b+ DCs to the site of infection and subsequent draining lymph nodes, whereas DC recruitment was not abrogated in neutrophil-depleted mice (treated with anti-Ly6G antibodies) [43], indicating that in case of L. Mexicana infection, neutrophils hamper DC migration. Lactoferrin may be one of the mediators responsible for this inhibitory effect, as both murine and human studies demonstrated inhibited migration to and reduced accumulation of DCs in the draining lymph nodes after injection of lactoferrin in the skin [4446]. Additionally, MPO was shown to inhibit CCR7 expression on DCs [47].

    Details are in the caption following the image

    Schematic representation of neutrophils, their major preformed mediators, cellular markers, and putative influence on DCs.

    Adhesion molecules and mediators involved in the communication with DCs, and modulation of DC migration, maturation, and cytokine/chemokine release are indicated. Depending on the mode of activation, neutrophils can influence DC-driven development of Th1, Th2, Th17, or Tregs.

    Neutrophils may not only affect the recruitment and migration of DCs but also, their capacity to activate and polarize T cells. Both in vitro and in vivo studies have implicated that neutrophils and DCs cluster together, which depends on various molecules, including the C-type lectin DC-SIGN (CD209) and ICAM-1 on DCs and neutrophil-specific glycosylated Mac-1 (CD11b/CD18) and CEACAM1 on neutrophils [28, 4852]. Resting and LPS-, TNF-α-, or N-fMLP-activated human neutrophils interacted and clustered with moDCs either in a DC-SIGN-dependent [49] or -independent manner [53]. Moreover, neutrophils enhanced maturation of moDCs and slanDCs, which were partially dependent on CD18, as well as survival and IL-12 production by slanDCs [48, 51, 52]. DC-SIGN was essential for A. fumigatus-induced maturation of lung CD11b+ DCs (up-regulation of CD86, CD40, and MHC-II). Maturation was reduced when neutrophils had been depleted using anti-Gr-1 antibodies [38]. Moreover, in vivo OVA/LPS administration in the skin induced neutrophil infiltration of skin draining lymph nodes in C57BL/6 mice, resulting in close interactions with CD11c+ DCs [47].

    Human moDCs and murine BMDCs that had been cocultured with LPS-, E. coli, Toxoplasma gondii, M. tuberculosis, or BCG vaccine-stimulated neutrophils showed an up-regulated expression of CD80, CD83, CD86, and HLA-DR, reduced IL-10 release, and increased IL-12 production. Moreover, enhanced capacity to induce DC-dependent CD8+ T cell proliferation and concomitant T cell-derived IFN-γ production was promoted by BCG-, Candida albicans-, LPS-, E. coli-, T. gondii-, or M. tuberculosis-activated neutrophils [42, 48, 5356]. It was shown that the presence of neutrophil-derived TNF-α was an essential driver for up-regulation of MHC-II and costimulatory molecules on murine DCs [53, 54], but direct cell–cell contact was crucial for Th1 polarization [41, 56].

    It has been demonstrated that both human and murine DCs have the capacity to internalize neutrophils. In vitro, internalization of human neutrophils by moDCs depended partially on CD18 [42], and in vivo studies demonstrated that injection of OVA-activated neutrophils into the footpad of C57BL/6 mice resulted in OT-I T cell proliferation in the skin draining lymph node [42], suggesting uptake of neutrophil-derived antigens into DCs. Likewise, L. major-infected neutrophils that had been injected in ears of WT C57BL/6 mice were internalized by DCs. In contrast, however, this resulted in an unexpected decrease of CD40, CD86, and MHC-II on DCs and reduced T cell proliferation and Th1 polarization [57], suggesting that L. major, when internalized in neutrophils, induced immune suppression.

    Activated neutrophils can release multiple antimicrobial mediators from their granules that have been shown to modulate DC activation, such as MPO, cathepsin-G, NE, LL-37, and lactoferrin (Fig. 1) [5861]. These antimicrobial peptides can either be released as soluble molecules via degranulation or bound to NETs. The exact role of NETs in modulating DCs is still unclear, and conflicting results have been reported. Short-term (30 min) exposure to NETs decreased the expression of HLA-DR, CD40, CD80, and CD86 on—and the LPS-induced production of TNF-α, IL-6, IL-8, IL-10, IL-12, and IL-23 by—human moDCs in one study [62]. In contrast, NETs induced CD80 and CD86 expression on human moDCs in another study [63]. Although unclear, this discrepancy may be explained by the differences in NETs introduced by variations in the methods used for their release and isolation. Furthermore, CD4+ T cell proliferation and IFN-γ, IL-17, and IL-10 production by T cells were partially inhibited by NET-stimulated moDCs; however, IL-5 and IL-13 production increased, suggesting that NETs block the onset of Th1/Th17 responses (as established by LPS) and favor Th2 induction instead [62]. Neutrophil DNA mixed with secretory leukocyte protease inhibitor and NE [64], LL-37 [60], or cathepsin-G [61] (all resembling NET-like structures) induced type 1 IFN release by human pDCs via TLR-9 activation [60]. Additionally, neutrophils from systemic lupus erythematosus patients containing high levels of LL-37 released more NETs after activation, resulting in a higher potential to trigger pDC activation (CD80, CD83, and CD86) and release of IFN-α, TNF-α, and IL-6 by pDCs [60, 65, 66]. Likewise, results regarding the modulatory effect of lactoferrin on DCs are contradicting as well. Maturation of human moDCs and CD1c+ DCs (up-regulation of CD80, CD83, CD86, and HLA-DR) [58, 67]; release of IL-6, TNF-α, IL-1β, and IL-12(p70); and proliferation of T cells, as well as Th1 induction by DCs, were induced by lactoferrin [58, 68]. However, unstimulated and BCG vaccine-stimulated murine BMDCs released IL-6 and IL-12(p40), which were reduced when lactoferrin was present [68], suggesting that lactoferrin has the capacity to suppress the primary induced Th1 polarization. LL-37 changed the in vitro morphology of immature human moDCs (bigger, adherent, and more dendrites); enhanced their endocytic capacity; induced expression of CD86, CD11c, CD11b, and CD18; but decreased the expression of DC-SIGN, CD16, and CD32, thereby abrogating receptor-mediated phagocytic capacity. Furthermore, LL-37-stimulated mature moDCs produced high levels of IL-12 and IL-6 and induced IFN-γ production by T cells [59]. In contrast, neutrophil-derived MPO interacted with CD11c+ DCs in draining inguinal lymph nodes, resulting in reduced CD4+ T cell activity and response to antigens. Additionally, enzymatically active MPO decreased maturation of murine BMDCs and human moDCs (down-regulated CD86 and HLA-DR expression) and reduced IL-12 production by both DC subsets in vitro [47]. However, in the presence of an anti-Mac-1 antibody, IL-12 production increased, indicating a modulatory role of MPO on DCs via Mac-1 [47]. Human moDCs treated with NE produced high levels of TGF-β and favored the generation of CD4+FoxP3+ T cells [69, 70].

    Next to soluble mediators and NETs, neutrophils also rapidly release membrane vesicles after activation, also referred to as ectosomes or EVs [71]. EVs are submicron vesicles that are composed of a selective set of lipids, proteins, and nucleic acids and play a role in intercellular communication [72]. Neutrophil-derived EVs were shown to contain a repertoire of different (phospo)lipids and proteins, including the “eat me” signal phosphatidylserine, MPO, lactoferrin, and NE [73, 74]. EVs released by fMLP-activated neutrophils modulated morphology and endocytic capacity of human moDCs [75]. Like EVs, apoptotic and necrotic neutrophils exposed phosphatidylserine on the outer membranes [48, 56] and thereby, partially suppressed human moDC and pDC maturation (down-regulated CD40, CD80, CD83, CD86, and HLA-DR) [65, 75, 76]. In the presence of Annexin V (competitive binding for phosphatidylserine), moDC maturation was enhanced, supporting that phosphatidylserine was essential for decreased maturation of moDCs [75]. Production of IL-10, IL-12, TNF-α, IL-8, and IFN-γ by moDCs was significantly reduced in the presence of fMLP-activated, neutrophil-derived EVs and apoptotic or necrotic neutrophils, whereas TGF-β release was induced by fMLP-activated, neutrophil-derived EVs [75, 76]. In either condition, subsequent DC-induced lymphocyte proliferation was diminished significantly [75, 76].

    To summarize, neutrophils have multiple immunomodulatory effects on DC function that may be mediated via cell–cell contact or the release of soluble factors, NETs, and EVs (Fig. 1). Depending on the activation trigger, neutrophils can either induce or inhibit DC maturation and activation and modulate their specific T cell polarizing capacity. Research performed over the last decades emphasizes the important and complex role of neutrophils in the onset and efficient regulation of adaptive immunity.

    EOSINOPHILS

    Eosinophils only represent a small number of circulating leukocytes (1–6%) and are short-lived cells having a t1/2 of 8–18 h in the circulation [77, 78]. Expansion and differentiation of eosinophils in the bone marrow are regulated by IL-3, GM-CSF, and primarily IL-5, which is also the main trigger for eosinophil release into the circulation [79]. IL-5 is produced by many immune cells, including eosinophils themselves, NK cells, mast cells, activated Th2 cells, and innate lymphoid cells [80, 81]. In healthy individuals, >90% of the eosinophils reside (in resting state) in adipose tissues, lung, mammary glands, thymus, uterus, and especially in the lamina propria of the gastrointestinal tract [82, 83]. Their extravasation is mainly regulated by the constitutive expression of eotaxins (subfamily of eosinophil chemotactic proteins) in tissues, having its highest expression in the gut [80, 83].

    Eosinophils are classically known for their role in fighting parasitic infections and their detrimental effects in allergic diseases. Their granule-enriched cytoplasm consists of antimicrobial proteins, such as EDN, EPX, and MBP, which can be released in response to invading pathogens [82]. Additionally, eosinophils can become activated by a variety of Th2 cytokines, including IL-5, and the epithelial-derived cytokines TSLP, IL-33, and IL-25, which prevent apoptosis and extend the eosinophil lifespan up to 2–5 d [17, 78]. A prolonged presence of activated eosinophils at inflamed sites suggests that they have an important role in the development of adaptive immunity. An increased expression of MHC-II and costimulatory molecules on human eosinophils and their potential to polarize and recruit Th2 effector cells through release of cytokines and chemokines (e.g., IL-4, IL-13, CCL17, and CCL22) were the first indicators that eosinophils may indeed be potential regulators of adaptive immunity [82, 8486].

    Although information about the interplay between DCs and eosinophils is still limited, recent studies suggest that eosinophils and their active compounds play a role in the recruitment and activation of DCs (Fig. 2), thereby supporting the generation of optimal adaptive immune responses. It was shown that EDN is an endogenous TLR-2 ligand that enhanced the migration of human CD34+ cell-derived DCs, moDCs, and murine BMDCs in vitro [87]. Likewise, subcutaneous injection of eosinophil-associated RNase 2—a murine equivalent of human EDN—in Balb/c mice resulted in local recruitment of MHC-II+CD11c+ DCs [87]. By acting as an endogenous TLR-2 ligand, EDN stimulated the MyD88 signaling pathway, including activation of MAPKs and NF-κB in human moDCs, resulting in their maturation (up-regulated CD83, CD86, and HLA-DR) and increased release of IL-6, TNF-α, IL-8 and IL-12(p70) [88, 89]. However, another study implicated that MBP protein released by TLR-9 ligand (CpG)-activated eosinophils, rather than EDN, played an essential role in the modification of human moDCs. TLR-9 ligand-activated eosinophils induced maturation of human moDCs (up-regulated CD80, CD83, and CD86), which partly depended on direct cell–cell contact and correlated positively with the eosinophil:DC ratio [90]. The capacity of human moDCs to promote T cell proliferation was enhanced by EDN as well [88]. In addition, in vivo experiments demonstrated a role for EDN and eosinophils on DC modulation, as C57BL/6 mice immunized with OVA together with EDN, or EDN-treated, OVA-pulsed DCs led to the release of Th2 cytokines (IL-4, IL-5, IL-6, IL-10 and IL-13) by splenocytes in a TLR-2-dependent mechanism [88]. Furthermore, OVA-immunized, eosinophil-deficient PHIL mice challenged with aerosolized OVA had reduced numbers of CD11b+CD11c+ cDCs in the lung and draining lymph nodes, and T cell accumulation in the lungs was abrogated compared with WT mice [91]. Moreover, adoptive transfer of OVA-pulsed BMDCs obtained from WT mice into the lungs of eosinophil-deficient mice resulted in significantly increased concentrations of IL-17, IFN-γ, and IL-13 in the BALF after challenge. Instead, when eosinophils were cotransferred, IL-17 and IFN-γ levels decreased, but IL-13 levels remained stable, indicating that eosinophils suppressed Th17 and Th1 responses, and favored Th2 polarization (Fig. 2) [91].

    Details are in the caption following the image

    Schematic representation of eosinophils, their major preformed mediators, cellular markers, and putative influence on DCs.

    Mediators involved in the modulation of DC migration, maturation, and cytokine/chemokine release are depicted. The ability to influence Th2 cell polarization, either via DCs or by direct effects on T cells, is shown as well. EMR-1, epidermal growth factor-like module containing mucin-like hormone receptor 1.

    As stated above, the gastrointestinal tract lamina propria contains markedly high numbers of eosinophil under homeostatic conditions [83], which were shown to be involved in modulation of DC function upon inflammation. Eosinophils were essential for the activation (CD86 and CCR7) and migration of intestinal mucosal CD103+ DCs to mesenteric lymph nodes in response to intragastric exposure to peanut and the Th2-inducing adjuvant cholera toxin in a murine food allergy model [92]. These effects were, at least partly, induced by the antimicrobial enzyme EPX, which was released upon eosinophil degranulation after exposure to peanut and cholera toxin in vitro and in vivo [92]. Activation of human moDCs and murine BMDCs (up-regulation of CD86, CD80, CD83, OX40L, and CCR7) and enhanced release of IL-6 and TNF by both subsets were induced by the peroxidase enzymatic activity of EPX [92]. Furthermore, extraintestinal DC activation and adaptive immunity were induced by EPX. By acting as a Th2-inducing adjuvant, EPX induced activation of DCs (up-regulated CD80 and CD86), migration to inguinal or thoracic lymph nodes, and IL-5 and IL-13 release in the BALF of intranasally OVA-challenged mice that had been immunized with a combination of OVA with EPX [92, 93].

    Overall, it can be concluded that although the evidence supporting crosstalk between eosinophils and DCs is still limited, some data suggest a modulatory role for eosinophil-derived granular proteins on DC function and their capacity to promote Th2-balanced immune responses (Fig. 2). Furthermore, direct cell–cell contact might be involved in promoting well-established adaptive immune responses.

    BASOPHILS

    Basophils are circulating cells characterized by cytoplasmic secretory granules containing preformed mediators, including histamine. Basophils are released as mature cells from the bone marrow, a process that is generally considered highly dependent on activated T cell-derived IL-3, although TSLP has recently also been associated with the elicitation of phenotypically and functionally distinct basophil subsets in mice and humans [9497]. Basophils represent only a minor cell population in the circulation, usually accounting for <1% of total blood leukocytes, and their lifespan is, on average, 1–3 d [98].

    Conditions associated with Th2-type immunity, such as helminth infections and allergic pulmonary or atopic skin inflammation, enhance the development and release of basophils into the periphery [99, 100]. In these conditions, basophils play important and nonredundant roles, not only as effector cells but also as promotors of Th2 cell differentiation (reviewed in refs. [101104]). For example, basophils are essential for inducing secondary IgE-dependent immune responses necessary for acquired resistance to parasitic reinfections [105108]. In addition to the FcεRI, basophils express multiple other immune receptors (e.g., TLRs and complement receptors). Ligation of these receptors leads to mediator release and subsequent immediate responses, such as leukocyte recruitment, supporting their role in innate immunity. Moreover, basophils are associated with various chronic inflammatory disorders, including autoimmunity and cancer, although their exact contributions are still poorly defined [94, 101, 109111].

    Although the current knowledge on basophil effector and immunomodulatory functions is limited, their involvement in inflammatory responses is supported by their recruitment to peripheral inflammatory sites (e.g., skin and lung) [21, 100, 105, 107, 112, 113]. Additionally, basophils can be recruited to draining lymph nodes in both mice and human, typically in a T cell-dependent manner [21, 111, 114117]. In murine lymph nodes, colocalization of basophils and CD11c+ DCs was shown in the T cell paracortex [116, 117], although functional consequences of this interaction are not yet clear. It was recently demonstrated that OX40L expression on BMDCs and IL-4 production by bone marrow-derived basophils were both enhanced as a result of cell–cell contact [118]. Basophils are well known for their ability to selectively release their mediators, such as the cytokines IL-4 and IL-13, histamine, and LTC4, depending on the activating stimulus. For example, both the cysteine protease papain and IgE cross-linking induced the release of IL-4, whereas only IgE cross-linking additionally resulted in histamine release [117]. Profound effects on both murine and human DCs (e.g., BMDCs, moDCs, myeloid DCs, LCs) have been described after histamine exposure via activation of distinct histamine receptors (reviewed in refs. [119, 120]). Histamine-mediated effects include enhanced expression of MHC-II and costimulatory molecules; increased endocytosis of soluble proteins; modulation of DC migration, cytokine, and chemokine release; and T cell differentiation. Together, these effects aid in inducing Th2 cell responses by several mechanisms, such as inhibition of IL-12(p70) and IFN-α release, up-regulation of Th2-attracting chemokines CCL17 and CCL22, and down-regulation of the Th1-associated chemokine CXCL10 [119124]. Furthermore, LTC4 was shown to stimulate human moDC migratory properties and enhanced their capacity to induce cytotoxic T lymphocytes and Th1 CD4+ T cell responses [125].

    In addition to histamine and LTs, PAF [126, 127] and TSLP [117] have been shown to be released by basophils, although it is not yet clear whether TSLP is also produced by human basophils. TSLP activates human CD11c+ DCs and murine BMDCs, as indicated by up-regulation of MHC-II and costimulatory molecules (e.g., CD80 and CD86), as well as prolonged DC survival [128131]. Moreover, TSLP-activated DCs were shown to increase OX40L expression, provide sustained proliferation and survival signals for naive CD4+ T cells, favor Th2-permissive microenvironments, and promote Th2 memory cell responses [117, 132, 133]. PAF receptor expression was shown on human moDCs, and its activation was associated with directional DC migration and IL-12 and IL-18 release [134, 135]. As such, basophil-derived soluble factors may profoundly influence DC function, although ample evidence for direct DC–basophil cross-talk is still lacking.

    Nonetheless, it is now increasingly accepted that basophils may contribute to the initiation and amplification of DC-driven Th2 cell responses, which generally involves their key ability to produce rapidly large amounts of IL-4 (Fig. 3). This was first demonstrated in murine cell cultures in which the presence of basophils was required for induction of IL-4-producing CD4+ T cells by DCs. IL-4+ T cell generation was attenuated in the presence of IL-4-neutralizing antibodies or when Il4−/− basophils were used instead of WT basophils [136, 137]. These data were recently corroborated using murine BMDCs, bone marrow-derived basophils, and OVA-specific OT-II cells [118]. Additionally, IL-4-producing basophils were essential for promoting epicutaneous Th2 sensitization to OVA and subsequent IgE-mediated experimental food allergy (using Il4–3′-untranslated mice) [118]. Others demonstrated that in vivo-elicited Th2 cell responses after antigen (cysteine protease papain plus OVA) injection into the mouse skin required the presence of and cooperation between CD8αDEC-205+ dermal DCs and IL-4+ basophils in the draining lymph nodes [138]. Depletion of basophils by targeting FcεRI significantly reduced Th2 immunity, although this effect could (partly) be mediated by the depletion of FcεRI+ DCs, as has been reported [116]. In this same model, isolation of CD11c+ DCs and basophils from draining lymph nodes after immunization and subsequent coculture with naive CD4+ OT II cells in vitro resulted in the induction of IL-4+ T cells, only when both DCs and basophils were present. Moreover, activated migratory dermal DCs were shown to increase CCL7 production, which was involved in basophil recruitment to draining lymph nodes [138]. The interdependence of DCs and basophils for inducing Th2 responses was recently further supported in a TSLP-mediated skin inflammation model [139]. Different basophil depletion strategies (targeting FcεRI, CD200R3, and use of Mcpt8DTR mice) did not inhibit CD4+ T cell expansion but markedly reduced Th2 cell differentiation. However, basophil depletion did not completely abolish IL-4 and IL-13 levels, indicating that the presence of basophils was not an absolute prerequisite. Indeed, it was demonstrated that migratory inflammatory CD11b+Ly-6C+FcεRI+ DCs (but not lymphoid tissue-resident steady-state CD8α+ and CD8α cDCs) were necessary and sufficient for Th2 priming to inhaled house dust mite antigen [116]. Nevertheless, basophils contributed to the strength of this response, as their depletion partially reduced Th2 immunity [116]. Additionally, basophils significantly increased Th2 cell development to protein antigens in in vitro assays using murine BMDCs or lung-derived I-E+CD11b+CD11c+ DCs and OVA-specific (DO11.10) T cells [140, 141].

    Details are in the caption following the image

    Schematic representation of basophils, their major preformed mediators, cellular markers, and putative influence on DCs.

    Mediators that may be involved in the modulation of DC migration, maturation, and cytokine/chemokine release are indicated. The ability to influence Th2 cell polarization, either via DCs or by direct effects on T cells, is depicted.

    The relatively low number of publications and certain discrepancies in the above-described basophil functions indicate that their effects on DCs (or DC subsets) and their importance in establishing Th2 cell responses are still not fully understood. The effects likely depend on multiple parameters, such as disease model systems and the nature of applied antigens. Additionally, the described in vivo functions are primarily based on data obtained in murine model systems. Still, little is known about the contributions of human basophils in shaping DC functions and DC-driven Th2 cell development. Nonetheless, current data indicate that the cooperation between basophils and DCs may not be critical for the induction of Th2 responses but support that basophils (by producing IL-4 but possibly also histamine and TSLP) can act as important accessory cells during CD4+ T cell activation (Fig. 3). Hereby, they likely contribute to the optimal induction of DC-mediated Th2 responses in vivo. Basophil recruitment to peripheral sites, including lymph nodes, during inflammatory responses may corroborate an important immunomodulatory role for basophils.

    MAST CELLS

    Mature mast cells are highly granular and long-lived cells (weeks to months). In contrast to other granulocytes, they are strategically located throughout the body, closely surrounding blood and lymph vessels at sites contacting the external environment (e.g., in the skin, lungs, and gastrointestinal tract). Mast cell progenitor cells enter the circulation from the bone marrow and start their maturation at final target tissues [142144]. The tissue distribution, together with the ability of activated mast cells to migrate to draining lymph nodes and the spleen [22, 145, 146], supports a role for mast cells as regulators of immune responses. Although mast cells are mostly known for their detrimental effector function in Th2 cell-mediated disorders, evidence that they play prominent roles in innate and adaptive immunity required for host defense against parasites, bacteria, virus, and animal venoms is accumulating. Moreover, mast cells are associated with the initiation and progression of multiple immune disorders and paradoxically, with immune tolerance as well (extensively reviewed in refs. [144, 147151]).

    The ability of mast cells to mediate such diverse and sometimes opposing effector and immunomodulatory functions is likely explained by several factors. For example, in both mice and humans, distinct mast cell phenotypes (based on composition and transcriptional profiles) are present at different anatomic sites [142, 152]. Additionally, mast cells express multiple activating (e.g., FcR; pathogen recognition receptors, such as TLRs; and complement receptors) and inhibitory (e.g., FcγRIIB, CD200R, and CD300a) receptors and have a unique capacity to release a large range of mediators, including EVs, after stimulation [144, 153156]. Depending on the type and strength of a particular stimulus, different cytokines can be released, not necessarily coinciding with degranulation [153, 157, 158]. As distinct mast cell phenotypes vary in their receptor expression, granule composition, and mediator release, diversity in responses can be expected following interaction with activating or inhibitory stimuli.

    Mast cells communicate with a multitude of different cell types, including tissue-resident cells (e.g., endothelial and epithelial cells), innate immune cells (e.g., recruitment of neutrophils and eosinophils to sites of infection [159161]), and adaptive immune cells (influence B cell, CD4+ and CD8+ T cell, and Treg functions and recruitment of T cells to lymph nodes [145, 162165]). Moreover, mast cells and DCs have been shown to localize in close proximity in tissues [166169], and a substantial body of evidence indicates that mast cells communicate and collaborate with DCs to influence the initiation, modulation, and/or maintenance of optimal adaptive immune responses (Fig. 4).

    Details are in the caption following the image

    Schematic representation of mast cells, their major preformed mediators, cellular markers, and putative influence on DCs.

    Adhesion molecules and mediators involved in the communication with DCs and modulation of DC migration, maturation, and cytokine/chemokine release are shown. Depending on the mode of activation, mast cells can influence DC-driven development of Th1, Th2, or Th17 cells or Tregs. tm, transmembrane.

    Enhancement of DC migration to sites of infection and to draining lymph nodes is one of the most described effects induced by mast cells. In several murine models, mast cell deficiency resulted in significantly attenuated CD11c+MHC-II+ DC migration [170173]. Correspondingly, enhanced CD11c+MHC-IIhi DC migration and antigen uptake were observed in an airway hypersensitivity model in mice having hyperactive mast cells as a result of the specific ablation of the NF-κB negative-feedback regulator A20 [174]. The exact mediators involved in these attenuated or enhanced effects were not further defined in these studies. However, histamine and TNF-α are the 2 most prominent mast cell mediators that have been associated with mast cell-induced DC migration (LCs, CD11c+ DCs, CD11c+MHC-II+ DCs, or CD11c+MHC-II+CD8+ DCs) from skin or nasal tissue toward draining lymph nodes [165, 175178]. In line with earlier observations [179, 180], mast cell-derived TNF-α enhanced CCR7 expression on mouse CD11b+CD11c+MHC-II+ DCs in the dermis, as well as CCL21 expression in draining lymph nodes, thereby promoting migration of activated DCs to draining lymph nodes [177]. These observations support the previously reported requirement for mast cells for optimal induction of CCR7 expression in vivo on mouse CD11c+MHC-II+ DCs [165]. Moreover, it was shown that the recruitment of DCs from blood into inflamed dermis was facilitated by enhanced expression of E-selectin on local blood vessels that had been induced by mast cell-derived TNF-α [177]. Other mast cell-derived products, such as CD40 ligand, PGE2, and LTB4 (but not histamine), are potentially involved in regulating CCR7 expression on DCs as well [181184] and can thereby promote chemotaxis of murine BMDCs toward CCL21 in vitro [185]. A critical role for TNF-α in inducing the migration of specifically CD8+ DCs toward draining lymph nodes was shown in a proinflammatory hapten (2,4-dinitrofluorobenzene)-induced, contact-hypersensitivity model [178] and also in a tolerogenic skin allograft model [186]. In this latter model system, IL-9-activated mast cells (no degranulation) also released GM-CSF, which increased the survival of migrated tolerant CD8+ DCs, a well-known effect of GM-CSF [187].

    The finding that activated mast cells, without signs of overt degranulation, release mediators that affect DC migration is supported by several additional studies. For example, it was shown in mice that bacterial peptidoglycan-mediated mast cell activation induced the influx of pDCs and CD11c+CD11bCD8+ DCs into inflamed draining lymph nodes, which relied on histamine and IL-6, respectively [188]. Furthermore, mobilization of LCs in response to bacterial peptidoglycan-induced complement C3 production was dependent on complement-activated mast cells that released GM-CSF, LTC4, and CCL3 [176]. Another study indicated that IL-1β derived from TLR-7-activated mast cells was partly involved in the initiation of LC migration in mice [189].

    Concomitant with enhanced DC migration, DC maturation can be affected by mast cells as well (Fig. 4). Enhanced maturation, often represented by increased expression of the costimulatory molecules CD80, CD86, and/or CD40, was induced by histamine [166, 175], IL-6 [190], and TNF-α [169, 178, 190]. Additionally, it was shown that especially mast cell-derived EVs enhanced the expression of the abovementioned maturation markers on murine BMDCs, possibly as a result of their heat shock protein 60 and heat shock cognate protein 70 content [155]. However, as EV isolation procedures were not optimal for proving a true vesicular mechanism, and findings have, so far, not been supported by others, the exact role for mast cell-derived EV is not clear yet. Additional studies indicated that direct cellular interactions of human and murine mast cells (cord blood-derived, bone marrow-derived, and peritoneal mast cells) with moDCs or BMDCs can lead to enhanced DC maturation [166, 169, 191]. Possible molecules involved in this contact are CD11c; PECAM-1; ICAM-1; ICAM-2; LFA-1; and α4, β1, and β7 integrins [169, 191].

    By influencing DC migration and maturation, mast cells can significantly modulate DC-driven immune responses (Fig. 4). For example, the development of pronounced contact hypersensitivity responses to several haptens (e.g., 2,4-dinitrofluorobenzene, FITC, and oxazolone) was highly dependent on mast cell-mediated effects on DCs [169, 170, 173, 192]. In line with the selective effect on CD8+ DCs described above, a preferential enhancement of CD8+ DC function by mast cell-derived TNF-α, leading to CD8+ T cell-dominated adaptive immunity, may be essential in these contact hypersensitivity models [178]. Notably, a prominent role for membrane-bound TNF-α on mast cells (not soluble TNF-α) was suggested by 2 groups [169, 192]. Moreover, a selective enhancement of cytotoxic T lymphocyte responses was also induced by LCs in mice under the influence of TLR-7-activated dermal mast cells [189].

    Mast cells have also been implicated in controlling DC-mediated host defense against several microorganisms in mice. Mast cell-derived TNF-α was important to maximize the induction of a protective primary humoral response to E. coli [177]. Furthermore, uric acid-induced mast cell degranulation led to release of membrane-bound Flt3 ligand in a murine Plasmodium chabaudi infection model. Subsequently, Flt3 ligand again preferentially stimulated the expansion of murine CD8α+ DCs, which substantially impacted the magnitude of CD8+ T cell activation. Flt3 ligand-dependent expansion of human CD141+ DCs (equivalent to murine CD8α+ DCs) was demonstrated in this study as well [193]. Adequate systemic protection of mice against L. major infections relied on the regulation of CD11c+ DC migration to inflamed skin lesions by mast cells. Additionally, mast cells increased the release of the T cell modulating cytokines IL-12, IFN-γ, IL-6, and TGF-β by matured BMDCs in a cell–cell contact-dependent manner, which were likely involved in shaping the Th1 (IL-12, IFN-γ)- and Th17 (IL-6, TGF-β)-dependent immunity that was required to control parasitic infection [194]. Mast cell–DC contact further enhanced the CD4+ T cell-expanding capacity of murine BMDCs [172, 191]. Moreover, degranulating mast cells were shown to contribute to DC-mediated T cell polarization toward Th2 cells in the presence of DC-activating compounds. PGD2 and histamine are likely involved in this process by reducing IL-12 release from murine BMDCs or human moDCs, respectively [166, 195, 196]. The DC-mediated, Th2-skewing capacities of histamine and PGD2 have been suggested in earlier studies as well [122, 197199]. Furthermore, it was suggested that contact between activated human cord blood-derived mast cells and moDC (in the presence of LPS or TNF-α/IL-1β, mimicking an inflammatory milieu) was equally important for optimal induction of Th2-promoting DCs [166].

    Mast cells may contribute to the induction of adaptive immunity by transferring protein antigens to DCs as well. It was demonstrated that this may occur via VLA-4- and LFA-1-dependent cell–cell contact of degranulating mast cells with DCs [168]. Alternatively, DCs were suggested to take up antigen-loaded, mast cell-derived EVs, resulting in antigen processing, presentation, and T cell activation. However, as indicated above, a true EV-mediated mechanism still needs to be adequately proven [155].

    In contrast to the proinflammatory effects of mast cells on DCs and DC-driven immune responses, it was recently shown that cellular contact between human immature moDCs and CD34+ stem cell-derived mast cells induced a tolerant DC phenotype when mast cells were not activated/degranulated. These tolerant-like DCs exhibited reduced HLA-DR and CD80 expression and increased PD-L1 expression. PD-1/PD-L1-mediated mast cell–DC contact resulted in STAT3-dependent expression of IDO by DCs, and the induction of TGF-β- and IL-10-producing FoxP3+ Tregs [200]. These findings may complement the earlier-described immunosuppressive effects of nondegranulating mast cells on DCs in a murine allograft tolerance model [186, 201]. Others similarly showed immunosuppressive effects of murine mast cells, among others, via the release of IL-10, potentially acting on DCs, although this was not demonstrated [146, 202, 203].

    Taken together, mast cells are increasingly recognized for their immunomodulatory capacities, including their significant ability to influence DC functions (Fig. 4). Additionally, this influence is presumably DC-subset specific. The reported mechanisms and mediators involved and the described functional consequences are diverse and sometimes contradicting. This likely is a consequence of the different mast cell-deficient mouse strains that have been tested [204], the multiple in vivo disease model systems that were analyzed (e.g., hapten-induced contact hypersensitivity, bacterial or parasitic infections), and the various in vitro investigated culture conditions (e.g., different human/mouse mast cell and DC phenotypes, with or without mast cell degranulation, and/or direct mast cell–DC interactions). Therefore, the reported data indicate that depending on the (patho)physiologic context, mast cells significantly aid in the generation and progression of optimal adaptive pro- or anti-inflammatory immune responses.

    CONCLUDING REMARKS

    In this review, we described the current knowledge on cross-talk between granulocytes and DCs, focusing on the functional consequences of this interaction for adaptive immune responses. Collectively, it is recognized that neutrophils, eosinophils, basophils, and mast cells are all able to facilitate early and/or optimal induction of DC-mediated immune responses. All granulocyte subtypes can modulate DC migration and maturation and change cytokine release patterns by activated DCs, thereby altering adaptive T cell responses.

    It should be noted that most of the described data were obtained using in vitro or mouse model systems. Confirmation in human (model) systems is required, because of known differences in DC subsets and granulocytes between mice and men. For example, some murine DC subsets are thought to be orthologous with human DCs, whereas other subsets are considered phenotypically and functionally distinct. Moreover, granulocyte numbers and tissue distribution can significantly differ, and mouse and human granulocytes can respond differently to certain stimuli by releasing distinct protein patterns upon activation.

    Nevertheless, by taking these considerations into account, the data reviewed here clearly implicate an important task for granulocytes in modulating DC functions and shaping DC-mediated adaptive immune responses.

    AUTHORSHIP

    A.B. wrote the paper and designed the figures. T.G.K. wrote the paper. M.v.E. and E.d.J. supervised and revised the paper.

    ACKNOWLEDGMENTS

    M.v.E. is supported by Netherlands Organisation for Scientific Research (VICI 91814650). All authors contributed equally to this work.

      DISCLOSURES

      The authors declare no conflicts of interest.