Biological role of granulocyte macrophage colony-stimulating factor (GM-CSF) and macrophage colony-stimulating factor (M-CSF) on cells of the myeloid lineage

Review on M-CSF and GM-CSF as homeostatic vs. inflammatory cytokine.

Abstract

M-CSF and GM-CSF are 2 important cytokines that regulate macrophage numbers and function. Here, we review their known effects on cells of the macrophage-monocyte lineage. Important clues to their function come from their expression patterns. M-CSF exhibits a mostly homeostatic expression pattern, whereas GM-CSF is a product of cells activated during inflammatory or pathologic conditions. Accordingly, M-CSF regulates the numbers of various tissue macrophage and monocyte populations without altering their "activation" status. Conversely, GM-CSF induces activation of monocytes/macrophages and also mediates differentiation to other states that participate in immune responses [i.e., dendritic cells (DCs)]. Further insights into their function have come from analyses of mice deficient in either cytokine. M-CSF signals through its receptor (CSF-1R). Interestingly, mice deficient in CSF-1R expression exhibit a more significant phenotype than mice deficient in M-CSF. This observation was explained by the discovery of a novel cytokine (IL-34) that represents a second ligand of CSF-1R. Information about the function of these ligands/receptor system is still developing, but its complexity is intriguing and strongly suggests that more interesting biology remains to be elucidated. Based on our current knowledge, several therapeutic molecules targeting either the M-CSF or the GM-CSF pathways have been developed and are currently being tested in clinical trials targeting either autoimmune diseases or cancer. It is intriguing to consider how evolution has directed these pathways to develop; their complexity likely mirrors the multiple functions in which cells of the monocyte/macrophage system are involved.

Introduction

M-CSF (or CSF-1) and GM-CSF (or CSF-2) were first identified as hematopoietic growth factors but have since been shown to play important roles in regulating mature myeloid cell populations under both homeostatic and inflammatory conditions [1, 2]. GM-CSF and M-CSF have a wide range of effects on myeloid populations, including survival, activation, differentiation, and mobilization. Studies using gene-deficient mice have revealed a pivotal role for M-CSF in the maintenance of multiple myeloid lineage populations, whereas GM-CSF is required for maturation of alveolar macrophages and effector differentiation of iNKT cells [3–5]. Additionally, both M-CSF and GM-CSF play a role in the development and function of certain DC populations [6, 7].

M-CSF and GM-CSF have different patterns of expression. M-CSF is produced ubiquitously and is constitutively expressed under homeostatic conditions by many cells and tissues of the body, whereas GM-CSF is mainly produced by activated leukocytes, which appear as part of host responses to infection or injury [2]. M-CSF and GM-CSF also induce opposite responses in macrophages, with M-CSF polarizing macrophages toward an anti-inflammatory phenotype, whereas GM-CSF promotes a proinflammatory phenotype. These opposing effects of M-CSF and GM-CSF on macrophages have been linked to “M2 and M1 macrophages,” respectively, although such terminology should be applied with caution, as these states may not represent alternative activation endpoints, and many other macrophage activation states are likely to exist [8–11]. Given the wide range of functions induced by CSFs under both homeostatic and inflammatory states, several clinical trials are currently being conducted, aimed at exploring the potential of these CSFs as therapeutics for several autoimmune and inflammatory diseases. Here, we review some of the roles that GM-CSF and M-CSF play in myeloid cell polarization and functions.

M-CSF (CSF-1)

M-CSF is constitutively produced by variety of cells, including endothelial cells, fibroblasts, osteoblasts, smooth muscle, and macrophages, and can be detected in plasma at ∼10 ng/ml [2, 12, 13]. Increases in circulating M-CSF levels have been reported in numerous diseases, including cancer, inflammation, and autoimmune disorders [14–17]. The levels of M-CSF in circulation also increase during pregnancy (where it contributes to placental development) [18, 19]. Under homeostatic conditions, the level of M-CSF is regulated via CSF-1R-mediated endocytosis and subsequent intracellular degradation [20]. In addition to myeloid cells, CSF-1R has been shown to be expressed in neurons and capillary endothelial cells of the CNS, where it provides neuroprotective and survival signals in response to brain injury and neurodegeneration [21, 22]. M-CSF exists in 3 biologically active isoforms: sgCSF-1, spCSF-1, and csCSF-1. sgCSF-1 and spCSF-1 are produced by endothelial cells and can exert their functions at a distance, whereas csCSF-1 plays a role in local regulation [23]. Studies using transgenic mice have shown that all 3 isoforms share some biologic effects but also mediate specific functions [24, 25].

Functions

Early in vitro studies demonstrated that M-CSF-cultured bone marrow precursor cells generate macrophage colonies. Because of these observations, it was initially thought that M-CSF drives a macrophage differentiation program and is essential for in vivo generation of macrophages from bone marrow precursors. Surprisingly, in the osteopetrotic Csf11^op/Csf^op mouse (which has a naturally occurring mutation in the CSF-1 locus), the number of blood monocytes is normal [25]. Instead, the functions and numbers of several tissue macrophage populations are altered [26–28], and Csf1^op/op mice have drastically reduced macrophage numbers in the peritoneal cavity, kidney, liver, and dermis [29]. Furthermore, in vivo administration of recombinant M-CSF results in expansion of the mononuclear phagocyte system, causing a substantial increase (up to 10-fold) in blood monocyte numbers, as well as a parallel, large increase in resident macrophage numbers [30]. M-CSF also has been shown to play an unexpected role in controlling the differentiation of several DC subsets, including pDCs, CD103⁻CX₃CR1⁺ lamina propria DCs, and CD11b⁺ cDCs in the dermis, lung, and kidney [7, 31–33]. In vivo administration of M-CSF results in an increase in pDC and cDC numbers, a result that reflects the high expression of CSF-1R in these cells [34]. In addition, M-CSF deficiency has a number of pleiotropic consequences, including osteopetrosis (as a result of deficient production of bone-resorbing osteoclasts), abnormalities in the central nervous system (CNS), and defects in reproductive functions [29, 35–39].

Cytokines/phenotype

M-CSF is commonly used in vitro to derive macrophages from bone marrow cells, which are called BMMs (Fig. 1). These macrophages exhibit high phagocytic activity and low antigen-presentation capacity [40]. M-CSF induces macrophage polarization toward an immunosuppressive phenotype, and these macrophages express high levels of the anti-inflammatory cytokine IL-10 and a CCL2/MCP-1, which preferentially recruits macrophages through its receptor (CCR2) [2]. We have recently identified another cytokine selectively produced by BMMs, which is encoded by a gene called Metrnl and for which we have suggested the name IL-39 to reflect its association with immune system functions [41]. The production of Metrnl/IL-39 is induced in macrophages by IL-4 and inhibited by IFN-γ, suggesting that it may have a role in type II immune responses, although the immune physiology of this cytokine remains to be established [41].

Figure 1. Functional phenotype of GM-BMMs (GM-BMDM) and M-BMMs (BMM).

(A) GM-BMMs are characterized by their proinflammatory phenotype and good antigen-presentation capacities. GM-BMMs produce proinflammatory cytokines, including TNF, IL-1β, IL-6, and IL-12, as well as chemokines, such as CCL5, CCL22, and CCL24, which recruit CD4⁺ T cells, NK cells, eosinophils, and basophils. (B) M-BMMs are characterized by their anti-inflammatory phenotype. They produce anti-inflammatory cytokine IL-10, as well as CCL2, which result in monocyte recruitment.

Receptor

The biologic activities of M-CSF are mediated by signaling through the type III tyrosine kinase transmembrane receptor colony stimulating factor receptor 1 (CSF-1R), which is encoded by the c-fms proto-oncogene [23, 29]. The binding of M-CSF to CSF-R1 leads to receptor dimerization, autophosphorylation, activation of PI3K, ERK, phospholipase C, and subsequent nuclear localization of the transcription factor Sp1 [9, 23]. In humans, mutations in CSF-1R have been associated with hereditary diffuse leukoencephalopathy, development of acute myeloid leukemia, and myelodysplastic syndromes [42, 43].

IL-34

Studies with gene-deficient mice, however, revealed that CSFR1^−/− mice exhibit a more severe phenotype than mice deficient in functional M-CSF, known as osteopetrotic (CSF^op/op) mice [44]. This discrepancy was elegantly explained by the subsequent identification of a second ligand called IL-34 [45], which is a dimeric glycoprotein that resembles the dimeric M-CSF glycoprotein but does not exhibit significant sequence homology with M-CSF (or any other known cytokine) [45]. Like M-CSF, IL-34 also induces in vitro generation of macrophages with potent immunosuppressive properties from bone marrow precursors by triggering CSF-1R signaling and promoting macrophage proliferation and differentiation [44–46]. Moreover, transgenic expression of IL-34 under the control of the CSF1 promoter in Csf1^op/op mice was able to rescue bone, tissue macrophage, osteoclasts, and fertility defects in these mice, indicating redundant functions of 2 ligands [44].

Unlike M-CSF, which has a very broad expression pattern, IL-34 expression is mainly observed in brain and skin [47–49]. Studies with IL-34-deficient reporter mice (IL34^LacZ/LacZ) indicate that in the brain, IL-34 is produced by neurons, whereas in the skin, it is produced by keratinocytes [47]. In addition, IL-34 is expressed by several small cell subsets in lymph nodes, spleen, and kidney [48, 49]. Consistent with this expression pattern, IL-34-deficient mice exhibit defects in microglia and LCs, which are 2 major myeloid cell populations in the brain and skin, respectively [47]. Interestingly, whereas IL-34 is essential for the development and homeostatic control of LCs, it seems to be dispensable for regeneration of LCs in response to tissue injury [50]. This suggests that during inflammation, M-CSF plays a predominant role in driving differentiation of LCs from blood monocyte progenitors.

In addition to different spatiotemporal expression of M-CSF and IL-34, these 2 ligands might have differing biologic effects as a result of their unique interactions with distinct regions of the CSF-1R [51]. A recent study has also reported that at least in the brain, IL-34 can also signal though an alternative receptor, namely, protein-tyrosine phosphatase ζ [52]. These complex biologic networks aimed at maintenance of cells of the monocyte-macrophage lineage likely reflect specialized functions of macrophage subsets that remain to be defined. In another exciting development, IL-34 has recently been reported to be produced by regulatory T cells and to be an important mediator of the regulatory function of these cells [53]. These observations strongly suggest that IL-34 may be involved in tolerance induction. If this is correct, then it may be exercising this function through effects on specific populations of "regulatory" macrophages. Given the clinical importance of macrophages, the IL-34/CSF-1R axis raised a lot of interest as a potential target in several human diseases. This prediction is likely correct, as IL-34 has now been identified as an important player in various pathologic processes, including obesity and insulin resistance [54], hepatitis and liver fibrosis [55], tumor progression and metastasis [56], and inflammatory bowel disease [57].

M-CSF and disease

M-CSF has been studied in a wide range of diseases, and several correlations have been reported. Increased levels of circulating M-CSF expression have been observed in several autoimmune diseases, including arthritis, kidney inflammation, pulmonary fibrosis, obesity, inflammatory bowel disease, and cancer metastasis [12, 58–60]. Consistent with this, neutralization of M-CSF and/or blockade CSF-1R showed improvements in various murine models of inflammation, autoimmunity, and cancer metastasis models [61–66]. Multiple studies, for example, have shown that in a mouse model of lupus nephritis [MRL-Fas (lpr) mice], deletion of M-CSF leads to kidney inflammation, whereas an increase in systemic M-CSF levels leads to accelerated disease onset [59, 67].

Mounting evidence indicates a central role for macrophages in tumor progression and metastasis. M-CSF promotes recruitment of macrophages through induction of CCL2 production, a chemokine that along with its receptor CCR2 recruits monocytes to sites of inflammation. Furthermore, M-CSF supports the survival and polarization of tumor-associated macrophages [68, 69], which promote tumor development through secretion of growth and proangiogenic factors, production of immunosuppressive cytokines, and inhibition of T cell effector functions [69, 70]. High expression of M-CSF and the presence of CSF-1R-positive macrophages in tumors have been observed in multiple tumor types and shown to correlate with poor prognosis [71, 72]. These observations strongly suggest that the M-CSF/CSF-1R axis, therefore, may be an attractive therapeutic target for many cancers. Accordingly, CSF-1R inhibition showed promising results in several murine models of cancer, including glioma, pancreatic, cervical, and mammary tumor models [73–75].

Several clinical trials using inhibitors of the M-CSF/CSF-1R axis have been conducted with mixed results in various cancers and in inflammatory/autoimmune diseases. The blockade of CSF-1R has not shown efficacy in some autoimmune indications, such as RA, despite evidence of an adequate exposure and effective peripheral target engagement [76]. Likewise, in cancers, such as glioblastoma, the blockade of the M-CSF/CSF-1R axis using inhibitors showed only a marginal therapeutic benefit and at best, only resulted in a delay in cancer growth [77, 78]. On the other hand, in a Phase II clinical trial, treatment of patients affected with the orphan indication TGCT with a CSF-1R inhibitor (PLX3397) showed prolonged regression in tumor volume, leading to the advancement of this drug to Phase III clinical trials [79]. Several companies have begun testing M-CSF/CSF-1R antagonists in combination with other therapeutic candidates in various cancers. Currently, several clinical trials testing M-CSF/CSF-1R inhibitors (either as single agents or as part of combinational therapy) are underway. Table 1 provides a list of some of the clinical trials being conducted (from clinicaltrials.gov).

TABLE 1.

Clinical trials targeting GM-CSF/GM-CSFR axis and M-CSF/CSF-1R axis

Target	Drug	Type	Indication	Phase	Reference
GM-CSFR	Mavrilimimab (CAM-3001)	mAb	RA	II	NCT01050998
GM-CSF	MOR103	mAb	RA MS	I/II Ib	NCT01023256 NCT01517282
CM-CSF	Namilumab (MT203)	mAb	Plaque psoriasis RA	II I	NCT02129777 NCT01317797
GM-CSF	KB003	mAb	Asthma	II	NCT01603277
GM-CSF	MORab-022	mAb	RA	I	NCT01357759
CSF-1R	PLX3397	Small molecule	Solid tumors PVNS/GCT-TS Prostate cancer	I III II	NCT01525602 (combination with Paclitaxel) NCT02371369 NCT01499043
CSF-1R	PLX7486	Small molecule	Solid tumors	I	NCT01804530 (combination with Gemcitabine Plus Nab-Paclitaxel)
CSF-1R	RO5509554	mAb	Solid tumors	I	NCT01494688 (combination with Paclitaxel)
CSF-1R	FPA008	mAb	PVNS/Dt-TGCT	I/II	NCT02471716
CSF-1R	IMC-CS4	mAb	Solid tumors	I	NCT01346358

Dt-TGCT, diffuse-type TGCT; GCT-TS, giant cell tumor of the tendon sheath; PVNS, Pigmented villonodular synovitis. For specific information, refer to clinicaltrials.gov, and enter the reference number of each trial.

GM-CSF (CSF-2)

Unlike M-CSF, GM-CSF has low basal circulating levels under homeostatic conditions, but its levels can quickly become elevated during infection or inflammation [80, 81]. GM-CSF is produced during inflammatory/autoimmune reactions by a variety of cells, including fibroblasts, endothelial cells, macrophages, DCs, T cells, neutrophils, eosinophils, resident tissue cells, and cancer cells [82–84]. Expression of GM-CSF is induced by proinflammatory cytokines, such as IL-1α, IL-1β, TNF-α, and IL-12, whereas IL-4, IFN-γ, and IL-10 suppress it [85–89]. At sites of inflammation, GM-CSF has proinflammatory effects through recruitment of myeloid cells and by enhancing their survival and activation [90–92]. In several murine models of autoimmunity/inflammation, GM-CSF blockade led to reduced levels of monocyte and neutrophil recruitment with corresponding alleviation of disease severity, whereas in vivo administration of GM-CSF resulted in mobilization of monocytes from the bone marrow into the bloodstream [93–95]. Although the in vivo kinetics of GM-CSF-mediated chemokine production responsible for immune cell recruitment remains to be elucidated, there is evidence indicating that GM-CSF stimulates the production of the chemokines CCL3 and CCL2 by neutrophils and macrophages [96]. These observations contrast with the more "homeostatic" role of M-CSF, as GM-CSF not only promotes the survival of macrophages but also supports their differentiation toward a proinflammatory phenotype.

Cytokines

As discussed above, GM-CSF shifts the phenotype of macrophages in a proinflammatory direction. It promotes this through induction of proinflammatory cytokines, such as TNF-α, IL-6, IL-12p70, IL-23, and IL-1β, and chemokines, such as CCL22, CCL24, CCL5, and CCL1, which promote leukocyte recruitment [9, 97–99]. GM-CSF-treated monocytes/macrophages have been linked to the M1 activation state of macrophages, as many of the cytokines induced by GM-CSF are also induced by IFN-γ. However, as mentioned above, such terminology should be applied with caution. Because of their excellent antigen-presenting capacity, GM-CSF-treated bone marrow cells are often referred to as DCs. It is important to note, however, that GM-CSF-cultured bone marrow cells express a mixture of macrophage and DC-specific markers. For example, both GM-BMMs and M-BMMs express common macrophage markers, such as CD11b (Mac-1), F4/80, and c-Fms (M-CSFR), whereas only GM-BMM expresses the DC-associated marker CD11c [99, 100]. Unlike DCs, GM-BMMs are good osteoclast precursors and are phagocytic [40, 97, 100]. Additionally, despite significant differences in their cytokine and chemokine production repertoire, meta analysis of mouse microarray data indicate that GM-CSF- and M-CSF-stimulated cells show more similarities than differences [40]. Because of this, the molecular nature of the mediators that define the proinflammatory phenotype of GM-BMMs has long been elusive. The proinflammatory nature of GM-CSF-treated myeloid cells may be explained by the observation that IFN regulatory factor 5 plays a key role in GM-CSF-mediated macrophage polarization driving the expression of IL-12p70 and IL-23 cytokines, while suppressing expression of IL-10 [101]. Additionally, activin A has been identified recently as an emerging player in the regulation of M- and GM-BMM polarization. One study has shown that treating M-BMMs with activin A prevents LPS-induced production of IL-10, and conversely, blockade of activin A suppressed the acquisition of proinflammatory markers by human monocytes cultured with GM-CSF [102]. Thus, the latter study suggests that activin A may be a key molecule inducing the proinflammatory phenotype in macrophages.

Role of GM-CSF in DC development

Despite the ability of GM-CSF to generate cells with DC characteristics from monocytes or bone marrow cells in vitro, the deficiency of GM-CSF or GM-CSFR in mice resulted in only minor defects in DC development in lymphoid organs. Instead, the phenotype of these mice indicates that GM-CSF plays an important role in the maturation of alveolar macrophages [3]. In humans, mutations in the GM-CSFR/IL-3R/IL-5R common β chain lead to alveolar macrophage defects, which are associated with proteinosis [103, 104]. GM-CSF is also involved in steady-state development of dermal CD103⁺CD11b⁺ DCs and a subset of intestinal lamina propria DCs but suppresses the development of resident CD8⁺ DCs [105–107]. GM-CSF has also been shown to be involved in the development of monocyte-derived DCs and inflammatory DCs, which accumulate in tissues in response to infection or tissue injury [108–110].

Receptor

GM-CSF signals through its receptor CSF-2R, which is composed of 2 subunits: a specific ligand-binding subunit (CSF-2Rα) and a common signal-transduction subunit (CSF-2Rβ). The latter is shared with the IL-5R and IL-3R [111–113]. In addition to sharing a common receptor subunit, genes encoding IL-3, IL-5, and GM-CSF are located in close proximity on human chromosome 5, reflecting a common evolutionary origin of these cytokines [114]. The binding of GM-CSF to CSF-2R triggers stimulation of multiple downstream signaling pathways, including JAK2/STAT5, the MAPK pathway, and the PI3K pathway [115–118]. Additionally, a recently described pathway, mediated though transcription factor 4 and β-catenin, has also been shown to play a role in the GM-CSF-driven differentiation of myeloid cells [119].

GM-CSF in inflammatory/autoimmune disease

Although GM-CSF is dispensable for the development and maintenance of the most major hematopoietic cell types in spleen, blood, and bone marrow, this cytokine plays an important role in the induction of a wide range of effector functions in many cells of the immune lineage, including stimulation of proliferation and activation of monocytes/macrophages, DCs, T cells, neutrophils, and B cells [120]. Mice lacking GM-CSF are more susceptible to pulmonary [121, 122] and intestinal [123] infections, indicating an important role for this cytokine against many pathogens.

Although GM-CSF production in T cells is often associated with Th17 cells, other T cell subsets, including Th1, Th2, and CD8⁺ T cells also produce GM-CSF upon activation [124–128]. The expression of GM-CSF in T cells is regulated by multiple cytokines. For example, inflammatory cytokines, such as IL-6, TNF-α, and IL-23, induce production of GM-CSF, whereas conversely, IL-10, IL-4, and IL-27 inhibit it [129, 130]. CD4⁺ T and CD8⁺ T cells from GM-CSF-deficient mice exhibit impaired, proliferative responses and effector functions [129, 131]. Furthermore, several studies have demonstrated that GM-CSF-producing T cells play an essential role in the induction of immune responses against Mycobacterium tuberculosis, EBV, and HIV-1 [132–134]. It is unlikely though that GM-CSF acts directly on T cells, as these cells do not express the GM-CSFR [135, 136]. Instead, T cell-derived GM-CSF promotes maturation and activation of APCs, which in turn, potentiate T cell functions [131, 137]. The complexity behind the production of T cell-derived GM-CSF suggests that it represents a powerful immune system mediator in both inflammation and infections.

Moreover, other non-T cell mechanisms against infection, where GM-CSF also participates, have been described. For example, it has been implicated against pneumonia through its ability to activate IgM production in B1a cells [138]. GM-CSF production by iNKT cells may be involved in resistance to tuberculosis [139]. The increase of evidence points to a central role of GM-CSF in a proinflammatory cytokine "positive feedback" loop, which often underlies the chronic nature of many inflammatory and autoimmune disorders. Th17 cells have been identified as potent inducers of many inflammatory/autoimmune diseases, including CIA, EAE, and arthritis in the SKG mouse [86, 125, 140–142]. Th17 cells have been shown to have high "plasticity," which allows their differentiation into highly pathogenic Th1/Th17 cells (particularly in humans). These cells are characterized by their coproduction of IL-17A, IFN-γ, and GM-CSF, coexpression of retinoid acid receptor-related orphan receptor γt and T-bet, as well as coexpression of the Th1 and Th17 signature chemokine receptors CXCR3 and CCR6 [143–145]. Multiple reports have shown that it is not IL-17A but rather Th cell-derived GM-CSF that mediates the pathogenic effects observed [86, 142]. Accordingly, studies using GM-CSF-deficient mice have demonstrated that the inability to produce GM-CSF protects these mice from developing multiple autoimmune diseases, including: EAE, CIA, and autoimmune myocarditis [146–148]. Moreover, in vivo administration of an anti-GM-CSF neutralizing mAb during the early stages of clinical disease completely prevented or strongly ameliorated autoimmune-associated inflammation [147, 149].

The success of blockade/neutralization of GM-CSF in several mouse models of autoimmune/inflammatory diseases suggested that neutralizing the GM-CSF/CSF-2R axis could be a useful therapeutic strategy for patients with MS, arthritis, and other indications, such as psoriasis. In clinical trials, an anti-GM-CSF mAb (MOR103) has shown significant improvements in patients with active RA. Its clinical efficacy supports further advancement into clinical development [150]. Likewise, Phase II clinical trials with mavrilimimab (an anti-GM-CSFRα mAb) showed rapid improvement in patients with RA with no significant adverse effects [151]. Table 1 shows a list of some ongoing/completed clinical trials targeting GM-CSF or its receptor (more information available at clinicaltrials.gov).

CONCLUDING REMARKS

M-CSF and GM-CSF were described initially as cytokines that regulate myeloid cell proliferation and development. Both have been shown to be critical to the functions of monocytes, macrophages, and DCs. However, each plays nonredundant, functional roles in the physiology of these cells. M-CSF shows more characteristics of a homeostatic cytokine that normally regulates the production and homing of many myeloid cells. Therefore, there are "normal levels" of circulating M-CSF in plasma. In contrast, GM-CSF is produced in large amounts under inflammatory conditions by activated cells of the immune system, and its functions on myeloid cells reflect this proinflammatory role. It is involved in the activation of myeloid cells and therefore, may be a more promising target for the development of future drugs aimed at controlling inflammatory autoimmune diseases. Conversely, M-CSF may have promise as a biomarker for prognostic or diagnostic purposes in diseases, such as cancer or autoimmune diseases. The identification of IL-34 as a second ligand of CSF-R1 is further evidence of the biologic complexity that lies behind these macrophage-associated cytokines. The encouraging news is that given the wide breadth of biologic functions, in which macrophages are involved, the understanding of these complex interactions will very likely lead to the identification of targets for the development of novel therapeutics suitable for various human diseases.

AUTHORSHIP

I.U. and A.Z. researched and wrote this article.

Glossary

BMM

bone marrow-derived macrophage

cDC

conventional dendritic cell

CIA

collagen-induced arthritis

csCSF-1

membrane-spanning cell surface glycoprotein 1 CSF-1

DC

dendritic cell

EAE

experimental autoimmune encephalomyelitis

GM-BMM

GM-CSF-derived bone marrow macrophage

iNK

invariant NKT

LC

Langerhans cell

M-BMM

macrophage-derived bone marrow macrophage

Metrnl

Meteorin-like

MS

multiple sclerosis

pDC

plasmacytoid dendritic cell

RA

rheumatoid arthritis

sgCSF-1

secreted glycoprotein CSF-1

spCSF-1

secreted proteoglycan CSF-1

TGCT

tenosynovial giant cell tumor

DISCLOSURES

The authors declare no conflicts of interest.