Molecular control of monocyte development

Abstract

Monocyte development is a tightly regulated and multi-staged process, occurring through several defined progenitor cell intermediates. The key transcription factors, including PU.1, IRF8 and KLF4, growth factors, such as M-CSF and IL-34 and cytokines that drive monocyte development from hematopoietic progenitor cells are well defined. However, the molecular controls that direct differentiation into the Ly6C^hi inflammatory and Ly6C^lo monocyte subsets are yet to be completely elucidated. This review will provide a summary of the transcriptional regulation of monocyte development. We will also discuss how these molecular controls are also critical for microglial development despite their distinct haematopoetic origins. Furthermore, we will examine recent breakthroughs in defining mechanisms that promote differentiation of specific monocyte subpopulations.

1. Introduction

Monocytes are a heterogeneous population of circulating phagocytes that give rise to tissue macrophage and dendritic cell (DC) populations under homeostatic and inflammatory conditions [1, 2]. In the mouse, monocytes can be divided into two primary subsets based on phenotype and function. Inflammatory monocytes express high levels of CCR2 and Ly6C and are rapidly recruited to sites of infection or injury, where they give rise to proinflammatory macrophages and DC [3–5]. Inflammatory monocytes are critical for the control of a number of pathogens, including bacteria such as Listeria and Mycobacterium [6–8], fungi including Cryptococcus and Candida [9–11], parasites such as Toxoplasma [12, 13] and viruses including Herpes simplex and Murine hepatitis virus [14–16]. However, these cells also significantly contribute to immunopathology and autoimmunity in models of atherosclerosis and cardiac infarction [3, 17–19], rheumatoid arthritis [20], multiple sclerosis [21], inflammatory bowel disease [22], stroke [23] and encephalitis [5, 24–28]. Ly6C^lo monocytes, identified by high expression of the chemokine receptor CX₃CR1 and low expression of CCR2, patrol blood vessels and mediate early responses against insult [29, 30]. These cells have also been shown to promote wound healing and angiogenesis in models of atherosclerosis and cardiac infarction [31, 32].

The developmental relationship between the Ly6C^hi and Ly6C^lo subsets is yet to be completely elucidated, however a large body of evidence suggests that circulating Ly6C^lo monocytes derive from Ly6C^hi monocytes [33–36]. Following clodronate liposome depletion, Ly6C^hi monocytes are the first subset to repopulate the blood, followed by Ly6C^lo monocytes, suggesting that the latter derive from Ly6C^hi cells [33, 37]. Supporting this, experiments tracing circulating Ly6C^hi monocytes with fluorescent beads or Dil-labeled liposomes after clodronate depletion showed that these cells convert into a Ly6C^lo phenotype after two days in the circulation [33, 37]. Furthermore, adoptive transfer studies have shown that Ly6C^hi monocytes can shuttle between the blood and bone marrow and downregulate Ly6C expression [34, 36]. However, there is some evidence to suggest that circulating Ly6C^lo monocytes may derive directly from hematopoietic precursors in the absence of their Ly6C^hi counterparts, highlighting the potential plasticity of monocyte differentiation [1].

Although significant efforts have been made to define the developmental relationship between Ly6C^hi and Ly6C^lo monocytes, there is distinct lack of data examining the molecular controls i.e. transcription factors, growth factors and cytokine/chemokine signaling that directs differentiation into these two phenotypically and functionally distinct subsets. This review will provide a current overview of the development of Ly6C^hi and Ly6C^lo monocyte subsets and the transcriptional regulation of this process. A better understanding of the molecular signals that mediate the differentiation of these subsets will be beneficial in understanding the development and function of these cells under homeostatic conditions and during disease.

2. The monocyte development pathway

The development of monocytes from hematopoietic stem cells (HSC) in the bone marrow and spleen occurs via several myeloid-committed progenitors (Figure 1.). The earliest precursors derived from HSC are the common myeloid precursors (CMP), which express the surface glycoprotein CD34, but not stem cell antigen-1 (SCA-1) [38–41]. These cells in turn give rise to a multipotent pool of precursors known as the granulocyte/macrophage precursors (GMP), which are identified by expression of the Fcγ receptors CD16 and CD32 [38]. Included within this subset are the more recently defined macrophage/DC precursors (MDP), which express the cytokine receptors CD115 (CSF-1R/M-CSFR), CX₃CR1 and Flt-3 (CD135) [42–47]. These cells have no significant granulocytic potential, but give rise to monocytes and tissue macrophage subsets. MDP also differentiate into some conventional DC (cDC) and plasmacytoid DC (pDC) subsets directly, without a monocytic intermediate, whereas monocytes themselves give rise to some DC subsets including inflammatory DC and mucosal DC [1, 36, 47]. Recently, a DC-restricted progenitor, known as the common DC precursor (CDP), was described [47–49]. The CDP shares the phenotypic markers of MDP and has been shown to give rise to cDC and pDC, but not monocytes. There is some debate in regards to whether CDP represent a distinct GMP precursor population, a population overlapping with MDP, or a DC-restricted population derived from MDP [47–49].

Figure 1.

The monocyte development pathway. Monocytes are derived from HSC in the bone marrow and spleen via several myeloid-restricted progenitors. Sca-1⁺ HSC (a) give rise to CD34⁺ CMP (b). These cells in turn give rise to CD34⁺CD16/32⁺ GMP (c). A population of these precursors also expresses CD115, CX₃CR1 and Flt-3, known as the MDP (d). A recently identified monocyte precursor downstream of the MDP loses expression of Flt-3 and upregulates Ly6C, known as the cMoP (e). This cell gives rise to Ly6C^hi monocytes (f), which have also been shown to differentiate into Ly6C^lo monocytes (g).

Recently, a monocyte and macrophage-restricted clonogenic progenitor downstream of the MDP was identified (Figure 1.). The common monocyte progenitor (cMoP) loses expression of Flt-3, gains expression of Ly6C and maintains expression of CD115 and CX₃CR1 [50]. These cells have been shown to give rise to monocytes and macrophages in vitro and in vivo, but not DC. Further supporting the hypothesis that Ly6C^lo monocytes are derived from Ly6C^hi monocytes, cMoP were shown to first give rise to Ly6C^hi monocytes in vivo 1–2 days post transfer, which then differentiated into Ly6C^lo monocytes 3–4 days post transfer. Similar but delayed developmental kinetics were also shown following the adoptive transfer of MDP in this study [50]. Furthermore, genomic analysis comparing the expression of transcription factors by cMOP, Ly6C^hi and Ly6C^lo monocytes indicated a closer relationship between cMoP and Ly6C^hi monocytes compared to cMoP and Ly6C^lo monocytes. This data further supports the hypothesis that Ly6C^hi monocytes give rise to Ly6C^lo monocytes [50].

3. Molecular control of the monocyte development pathway

Transcription Factors

Monopoeisis is controlled by the expression and suppression of specific transcription factors at defined timepoints in the developmental pathway. Targeted gene knock-out and knock-in technologies have provided significant insight into the temporal roles of specific transcription factors, as well as cytokines, growth factors and other molecules which play a critical role in this developmental process [49, 51]. The transcription factor PU.1 plays a prominent role in monocyte differentiation at various stages of commitment. As homozygous PU.1-deficient mice are embryonic lethal or die within a few days after birth, fetal liver stem cells or conditional knockouts have been used to investigate the role of this transcription factor in myeloid cell development [52]. High expression of PU.1 antagonizes key regulators of other developmental pathways, such as GATA-1, GATA-2 and C/EBPα [53–55] and activates myeloid-specific factors such as Interferon regulatory factor-8 (IRF8), Kruppel-like Factor 4 (KLF-4) and Erg1 [56–59]. Conversely, low expression of PU.1 favors the development of granulocyte lineage cells. In the absence of PU.1, HSC fail to give rise to myeloid progenitors, resulting in the loss of monocytes and most DC subsets. PU.1 deficiency also favors the development of the granulocyte lineage, resulting in excess proliferation of neutrophils [52, 60–62].

IRF8, also known as interferon consensus sequence-binding protein (ICSBP), is an IFN-γ-regulated transcription factor that is also critical for monocyte differentiation. Interestingly, although PU.1-and KLF4-deficient mice are homozygous lethal, IRF8-deficient mice are viable. HSC express low levels of IRF8, which is upregulated as these cells differentiate into CMP and GMP [63]. In conjunction with PU.1, IRF8 promotes the differentiation of monocytes from GMP and inhibits the differentiation of granulocytes [64]. IRF8-deficient animals show accumulation of GMP, significant expansion of neutrophils and defective monocyte development [65–68]. Differentiation into macrophage and DC subsets is also defective [67, 69–72]. Recent studies indicate that IRF8 is critical for both the differentiation of Ly6C^hi and Ly6C^lo monocytes. Mice lacking IRF8 have significantly reduced numbers of Ly6C^hi and Ly6C^lo monocytes in the blood, bone marrow and spleen [71, 73]. We have also shown that Ly6C^hi monocytes fail to express high levels of CCR2 in IRF8-deficient mice and as a consequence these cells show defective migration to sites of inflammation (Terry et al, unpublished data).

The PU.1 and IRF8-activated transcription factor KLF4 also plays an important role in monocyte development [73]. As homozygous KLF4-deficient mice die within a few days after birth, fetal liver stem cells have been used to investigate the role of this transcription factor in vitro and in vivo [74]. KLF4-deficient hematopoietic cells fail to differentiate into Ly6C^hi monocytes and have very few Ly6C^lo monocytes in the bone marrow, blood and spleen. These cells also showed increased apoptosis and failed to express key trafficking molecules such as CD62L [74]. Increased numbers of neutrophils are also produced by KLF4-deficient HSC [56]. Overexpression of KLF4 in HSC or CMP restricts these progenitors to the monocytic lineage [56]. Furthermore, reintroduction of KLF4 into PU.1 or IRF8-deficient progenitor cells was sufficient to rescue monocyte development [56, 73]. These data indicate that KLF4 is a major downstream target of IRF8 and PU.1 that is critical for development of the monocyte lineage.

Growth Factors

PU.1 is critical for the expression of the receptor tyrosine kinase CD115 (CSF-1R/M-CSFR) in early myeloid progenitors and the upregulation of this receptor during the later stages of differentiation [42–45]. Expression of this growth factor receptor is indispensable for monocyte development as it is a key regulator of survival, proliferation and differentiation [75]. Mice that lack functional expression of CD115 show deficiencies in most monocyte and macrophage populations, as well as some DC subsets. The two known ligands of CD115, M-CSF (CSF-1) and IL-34, are both important for development of monocytes from their hematopoietic progenitors, as M-CSF- and IL-34-deficient mice exhibit a milder phenotype than animals lacking CD115 expression [76–79].

Cytokine signaling

The proinflammatory cytokine Interferon-γ (IFN-γ) plays an import role in driving monocyte development under inflammatory conditions. During infection, the expression of PU.1 and IRF8 are upregulated in GMP derived from wild-type (WT) but not IFN-γ-deficient mice. Furthermore, direct stimulation of GMP with IFN-γ results in the upregulation of these transcription factors critical for monocyte development [80]. In vivo, WT mice showed significant expansion of monocytes at the expense of neutrophils in response to lymphocytic choriomeningitis virus (LCMV) infection, resulting in increased numbers of these cells in the blood and bone marrow. In comparison, IFN-γ-deficient mice showed significantly increased numbers of neutrophils, highlighting the role of IFN-γ in directing monocyte development during inflammation [80].

Type I interferon (IFN-I) signaling also plays an important role in monocyte development under inflammatory conditions. Mice lacking the IFN-I receptor show increased migration of neutrophils and decreased recruitment of monocytes to the influenza-infected lung, and as a result show increased mortality [81]. Interestingly, IFN-I receptor-deficient HSC fail to give rise to Ly6C^hi monocytes. A significant Ly6C^int monocyte population was instead observed in the bone marrow and infected lung, which shows altered morphology and chemokine production compared to WT Ly6C^hi monocytes [81]. Culture of WT and IFN-I receptor-deficient bone marrow cells with bronchiolar lavage fluid or influenza directly, revealed that WT bone marrow cells were able to differentiate into Ly6C^hi monocytes, IFN-I receptor-deficient cells were not. The role of the IFN-I receptor was confirmed by culturing infected WT bone marrow cells in the presence or absence of IFN-I receptor-blocking antibodies, in which IFN-I receptor blockade inhibited differentiation into Ly6Chi monocytes. These findings confirm the role of type I IFN signaling in promoting inflammatory monocyte development during infection [81].

4. Transcriptional control of monocyte subset differentiation

The molecular controls that promote the differentiation of specific monocyte subsets remain for the most part unknown. However, a significant breakthrough has been made in recent years in identifying the role of transcription factor nuclear receptor subfamily 4, group A, member 1 (NR4A1) in Ly6C^lo monocyte development. Hanna and colleagues (2011) showed that NR4A1 plays a critical role in mediating the differentiation and survival of Ly6C^lo but not Ly6C^hi monocytes [82]. MDP in NR4A1-deficent mice fail to give rise to a significant Ly6C^lo population and these animals do not have mature Ly6C^lo monocytes circulating in the blood, spleen, or patrolling the endothelium. The authors show that the few Ly6C^lo cells that remain in the bone marrow are unable to differentiate into mature monocytes and rapidly undergo apoptosis [82]. This is a particularly significant finding as there is a distinct lack of knowledge in regards to the molecular controls determining the differentiation of Ly6C^hi or Ly6C^lo monocytes. A recent study has also confirmed that NR4A1 is not expressed by MDP, cMoP or Ly6C^hi monocytes, but is exclusively expressed by Ly6C^lo monocytes, further supporting the hypothesis that this transcription factor plays a critical role in directing the differentiation of Ly6C^lo monocytes from Ly6C^hi precursors [50].

5. Monocytes and microglia arise via distinct haematopoetic pathways but are dependent on the same molecular controls

Microglia, the resident macrophages of the brain, play an important role in protecting the central nervous system (CNS) from infection. However, these cells also contribute to the pathogenesis of some inflammatory diseases [83, 84]. Recent studies have confirmed that microglia arise from myeloid progenitors in the yolk sac during embryogenesis, through a developmental pathway distinct to that of bone marrow monocytes [85–88]. However, a major question that is still under contention is whether circulating monocytes can replenish microglial populations in the adult. While it is clear that in some models of infection and injury, Ly6C^hi inflammatory monocytes migrate to the CNS and exhibit a microglial phenotype, there is little evidence to suggest that this process occurs under homeostatic conditions in unperturbed animals [27, 73, 87, 89, 90].

Despite their distinct haematopoetic origins, the development of both monocytes and microglia is critically dependent on the transcription factor PU.1. Mouse embryos lacking PU.1 expression are completely devoid of microglia in the CNS [88]. The PU.1-regulated chemokine receptor CD115 is also critical for normal microgliogenesis. Mice deficient in CD115 or its ligand IL-34 lack microglia in the CNS [79, 86, 88]. It is also clear that the transcription factor IRF8 plays an important role in microglial development, however there is some dispute in regards to its precise function. While a recent study has shown that numbers of microglia are significantly reduced in the brain of IRF8^−/− embryonic mice compared to the WT, in the adult CNS, normal or even slightly increased microglial cell numbers were reported in several other studies [73, 91, 92]. Furthermore, IRF8 deficiency was also shown to have a role in the downstream morphology and function of microglia [73, 91, 92]. The discrepancies between these studies may be primarily due to changes in microglial populations that occur with the aging process, or the result of different experimental techniques used to enumerate microglial populations. Nevertheless, these data indicate that further investigation of the role of IRF8 in the embryonic vs. adult brain is warranted.

6. Summary

The molecular control of monocyte differentiation, specifically transcription factors, growth factors and cytokine/chemokine signaling critical for the development of these cells, have been well described in recent years [93–95]. However, the factors that promote the differentiation of specific monocyte subsets remain for the most part unknown. Recent studies have identified some key molecular signals that are critical for driving subset differentiation, of which the transcription factor NR4A1 represents a significant breakthrough. NR4A1 is not only the first transcription factor to be identified that is primarily critical for the development of Ly6C^lo monocytes, but furthermore supports the current model of monocyte development in which Ly6C^lo precursors arise via a Ly6C^hi monocyte precursor. Future studies which identify other transcription factors, growth factors and cytokine/chemokine signaling pathways critical for Ly6C^hi and/or Ly6C^lo monocyte differentiation will be undoubtedly be key in broadening our understanding of myeloid cell development and the role that these cells play in infections, cancer and autoimmune diseases.

Highlights.

• Monocyte development is a tightly regulated molecular process

• Transcription factors PU.1, IRF8 and KLF4 are critical for monocyte development

• NR4A1 is critical for Ly6Clo monocyte development from Ly6Chi progenitors

Abbreviations

cDC

conventional dendritic cells

CDP

common dendritic cell precursors

cMoP

common monocyte progenitor

CMP

common myeloid precursors

CNS

central nervous system

DC

dendritic cells

GMP

granulocyte-macrophage precursors

HSC

hematopoietic stem cells

IRF8

interferon regulatory factor-8

KLF

Kruppel-like factor

MDP

macrophage/dendritic cell precursors

pDC

plasmacytoid dendritic cells

WT

wild-type