Macrophages: Key Regulators of Steady State and Demand-Adapted Hematopoiesis

Abstract

Hematopoietic stem cell (HSC) function is required for balanced blood production throughout life, thus it is essential to understand the mechanisms regulating this highly dynamic process. Bone marrow-resident macrophages (Mϕs) have recently emerged as an important component of the HSC niche where they contribute to regulating HSC and progenitor cell (HSPC) mobilization and function. Here we review the role of Mϕs on immune cell production, HSPC pool size, and mobilization at steady state and under inflammatory conditions. Inflammation induces marked changes in hematopoiesis to restrict or promote generation of specific cell lineages, and this often has a negative impact on hematopoietic stem cell (HSC) function. Cytokines and growth factors induced during inflammation influence hematopoiesis by acting directly on HSPCs and/or by modulating niche cell function. We focus particular attention on the opposing effects of two key inflammatory proteins, interferon gamma (IFNγ) and granulocyte-colony stimulating factor (G-CSF), in regulating bone marrow-resident Mϕs and HSPCs. Mϕs are essential for tissue homeostasis, and here we highlight their emerging role as a central regulator of both steady state and demand-adapted hematopoiesis.

I. Introduction

Tissue-resident macrophages (Mϕs) play a critical role in maintaining homeostasis [1–4], modulating immune responses, and initiating tissue repair [5, 6]. Originally thought to be of monocytic origin [7], resident Mϕs primarily rely on self-renewal for replenishment in most adult tissues [8]. Moreover, fate-mapping studies have demonstrated that resident Mϕs arise from pre- or postnatal HSCs and progenitors, with the exception of microglia and some Langerhans cells, which arise from the yolk sac [9–16]. The specific origin of BM Mϕs is not yet clear, as this tissue was not examined in recent lineage-tracing studies. However, one study demonstrated that BM cells were able to replenish BM Mϕs after radiation treatment [8]. These findings suggest that although BM Mϕs rely on self-renewal for their replenishment, adult hematopoietic cells have the capability of replenishing BM Mϕs upon radiation-induced impairment of self-renewal [8]. While questions still remain on the precise origin of BM Mϕs, recent studies have revealed new information on their function.

Mϕs in the BM regulate erythropoiesis [1–4, 17] and more recently have been reported to influence other branches of the hematopoietic system. Within the BM compartment, Mϕs regulate HSPC location [18, 19], BM niche function [18–21], and steady state granulopoiesis [22–25], yet the precise mechanisms by which Mϕs regulate these processes are not clear. Moreover, the impact that inflammatory cytokines have on BM resident Mϕs, and the consequence on HSCs, is still in the early stages of investigation. Here we review the literature on Mϕs in both medullary and extramedullary hematopoietic tissues, as regulators of demand-adapted hematopoiesis. This topic is particularly relevant to mitigating HSC exhaustion during infection and other chronic inflammatory diseases, and improving HSPC mobilization for transplantation.

II. Macrophage Characterization in the Bone Marrow

BM-resident Mϕs were first characterized in the adult mouse as supportive “nurse” cells present in erythroblast islands [26–28], where they play a critical role in red blood cell (RBC) formation, adhesion, and clearance of expelled nuclei (reviewed in [2, 4, 17, 29]). BM-resident Mϕs can be distinguished from monocytes and monocyte-derived Mϕs by a reduced respiratory burst and a difference in developmental origin [8, 28]. BM Mϕs are located both in the central marrow and on the bone lining, where they support osteoblast (Ob) function [19–21]. Unlike osteoclasts, however, resident Mϕs are not multinucleated, do not express tartrate-resistant acid phosphatase (TRAP), and promote bone deposition, rather than bone resorption [30]. Multiple resident Mϕ populations have been described in the BM based on surface antigen expression, location, and function (Table 1). However, significant overlap remains in Mϕ identification and depletion methods, thus making it difficult to determine the distinct contribution of each of these Mϕ populations to hematopoietic processes.

Table 1.

BM Mϕ Characterization by Surface Antigens, Function, and Location

Mϕ Markers	Alternative names	Depletion/Reduction Method	Function	Location	Ref.
F4/80hi, CD11b−, mannosyl/fucosyl receptor+, CD169+, FcR+ (IgG2a, IgG2b)	N/A	N/A	Highly phagocytic, absent respiratory burst (with zymosan)	Immature hematopoietic islands	28, 31–32
F4/80+, CD68+, Mac-2lo, Mac-3+	Osteomacs	Mafia mice, Clod-lip, G-CSF	Support Ob function, bone healing and collagen Type 1 deposition, matrix mineralization	Ob lining	19–21, 30
F4/80+, CD11b+, Ly6G+, VCAM-1+, CD169+	CD11b+ Mϕs	MaFIA mice, Clod-lip, G-CSF, CD169-DTR	Support Ob function, support expression of HSC retention factors, restrict HSC numbers/dormancy	Erythroblast Islands	19, 38–39
F4/80+, CD169+, CD11bint./lo, CD68+, Gr-1−, CD115int., VCAM-1+, MHCII+	CD169+ Mϕs, CD11bint./lo Mϕs	CD169-DTR, G-CSF, Clod-lip	Support Nestin+ MSC function, support expression of HSC retention factors, support steady-state and pathological erythropoiesis, neutrophil engulfment and HSPC mobilization, restrict HSC numbers/dormancy	Proximal to Nestin+ MSCs, Erythroblast Islands	1, 3, 18, 22, 39
CX3CR1+, GR-1lo/−	Resident mono/Mϕs	G-CSF	Mobilize HSPCs in response to direct G-CSF signaling	N/A	69

Broadly, BM Mϕs express the surface proteins F4/80, CD68 (microsialin), vascular-adhesion molecule-1 (VCAM-1), CD169 or sialic acid binding immunoglobulin-like lectin-1 (Siglec-1), and express variable amounts of CD11b [1, 18, 28, 31–33]. These proteins reflect Mϕ-specific function as CD11b, VCAM-1, and CD169 participate in efferocytosis, adhesion, and endocytosis, respectively [34–37]. Osteomacs were identified as a stromal supportive population of F4/80⁺ cells that form a canopy on bone-lining osteoblasts in vivo and are detectable in osteoblast cultures in vitro [21]. More recently, CD169⁺ Mϕs and Ly6G⁺ Mϕs were identified as critical stromal niche supportive cells [18–20] that regulate erythropoiesis [1, 3, 38], HSC cycling and HSC pool size [39]. It is not clear if CD169⁺ and Ly6G⁺ Mϕs comprise the osteomac population, however, as both Mϕ types are reduced with clodronate-liposomes and G-CSF [3, 18, 19, 39] and the transgenic mouse model CD169-DTR [18, 38]. However, CD169⁺ Mϕs express low to intermediate CD11b while Ly6G⁺ Mϕs exhibit high CD11b expression [18, 19, 39]. CD11b is an integrin involved in efferocytosis and adhesion [35, 40], therefore, differential expression may reflect functional and locational differences [41]. Moreover, these Mϕ populations may differ in their developmental origin as CD11b^hi Mϕs in multiple tissues were reported to arise from adult BM, while CD11b^lo/int Mϕs are derived prenatally [15]. As the BM was not examined in the recent Mϕ-origin study, further investigation into BM Mϕ origin may aid in the development of tools to distinguish between various BM Mϕ subtypes and allow study of their respective functions in regulating hematopoiesis and HSC biology.

III. Macrophages Regulate Hematopoiesis

Developmental hematopoiesis

During embryonic life, Mϕs appear in the yolk sac and are present in the embryo prior to the emergence of definitive HSPCs [11, 42, 43], which is thought to be important in development and immune protection of the fetus [44–46]. Mϕs also participate in HSPC emergence within the embryo. Primitive Mϕs in zebrafish produce inflammatory factors that are associated with the appearance of HSPCs in the caudal hematopoietic tissue [47]. In addition, Mϕs produce matrix metalloproteinases that degrade extracellular matrix (ECM) around HSPCs in the aorta-gonad mesonephros, leading to HSPC release into circulation where they subsequently seed hematopoietic tissues [48]. ECM components contribute to adhesion and cell fate within adult stem cell niches [49] and Mϕs control the production and degradation of ECM [50]. An important and unanswered question is whether adult Mϕs contribute to hematopoiesis via ECM-dependent processes in a manner similar to their primitive counterparts.

Adult medullary hematopoiesis

In adult hematopoiesis, the BM microenvironment or “niche” plays a critical role in influencing HSC location and function. Various non-hematopoietic cell types, including osteoblasts Obs, mesenchymal stromal cells (MSCs), and endothelial cells (ECs), contribute to HSC maintenance and function and have been studied extensively. In brief, these cells express soluble and contact-dependent factors critical for anchoring HSPCs in the BM and directing cell fate decisions [51, 52]. Mϕs support BM stromal cell function within the HSPC niche [18, 19, 21, 53], yet the Mϕ-dependent factor(s) involved in this process have not been clearly identified. Moreover, little is known regarding the impact that Mϕs have on HSC function.

Macrophage regulation of circulating HSPCs

HSPC homing and anchorage in the BM occurs via binding of their receptor CXCR4 to CXCL12, a chemokine that is highly expressed in the BM [54, 55]. Administration of granulocyte-colony stimulating factor (G-CSF) reduces CXCL12 production in the BM and is routinely used to mobilize HSPCs into circulation for autologous transplantations. However, some patients are poor mobilizers due to underlying disease, age, and prior treatment [56, 57], thus limiting the yield of transplantable HSPCs. Low BM HSPC numbers and aberrant HSC retention in the niche are thought to contribute to poor mobilization [57]. Thus, to enhance HSPC release from the BM, the CXCR4 antagonist, AMD3100 (Plerixafor), is used in combination with G-CSF in poor mobilizers. However, mobilization efficiency remains low in some patients, thus, strategies that safely augment HSPC mobilization are still needed. Mϕs regulate HSPC mobilization as their depletion using the Mϕ Fas-induced apoptosis (MAFIA) mouse model, clodronate-encapsulated liposomes [18, 19], and CD169-DTR mice [18], increases circulating HSPCs in the blood. Mϕ depletion correlates with reduced gene expression of HSPC retention factors including CXCL12, angiopoietin-1, Kit ligand, and VCAM-1 in osteolineage cells, suggesting that Mϕ depletion increases egress of HSPCs from the BM. Surprisingly, upon examining HSCs in the BM after Mϕ depletion, an accumulation of both proliferative and dormant HSCs was observed [39], which is inconsistent with the idea that Mϕs retain HSPCs in the BM. In fact, it suggests that Mϕs negatively regulate HSC proliferation and pool size. While retention factors anchor HSCs in the BM, many also function to maintain HSCs in a quiescent state [58, 59]. Therefore, Mϕs may promote HSC quiescence through such factors. Alternatively, Mϕs may promote HSC quiescence directly, as one report suggested that a rare population of α-smooth muscle actin positive monocytic cells maintain HSC quiescence during stress conditions in a contact-dependent manner [60]. Thus, Mϕ depletion at steady state may increase HSC numbers in both the BM and in circulation by permitting proliferation.

HSPCs are enriched in the BM, but they are also able to traffic and seed peripheral tissues including lymph nodes, liver, kidney, and spleen even at steady state [61]. As current methods of Mϕ depletion are not tissue-specific, depletion of Mϕs in other tissues, in addition to the BM, may also contribute to the accumulation of HSPCs in circulation. In fact, VCAM-1+ Mϕs retain HSPCs in the spleen and Mϕ depletion, via silencing the M-CSF receptor, releases HSPCs into the circulation [62]. Furthermore, the mobilizing agent AMD3100 is thought to release a splenic reservoir of HSPCs into circulation, in addition to its action on HSPC release from the BM [63]. Therefore, systemic Mϕ depletion may increase circulating HSPCs, in part, via release from extramedullary tissues. Understanding the mechanisms whereby Mϕs regulate HSPC location and circulation is important for optimizing clinical mobilization protocols.

Effects of G-CSF on bone marrow macrophages and HSPC mobilization

G-CSF is elicited during many bacterial and fungal infections to boost granulocyte production, a process termed ‘emergency granulopoiesis’ [64]. Interestingly, increased granulopoiesis correlates positively with HSPC pool size and mobilization. Administration of G-CSF causes mobilization of BM HSPCs in the blood and an increase in phenotypically-defined dormant HSCs in the BM [65]. Concurrent to HSPC mobilization, G-CSF alters the BM microenvironment, leading to decreased osteoblast function and reduced expression of key HSC retention signals such as CXCL12 and VCAM1 [66–68]. Lineage-specific transgenic mouse models have revealed that G-CSF signaling specifically in CD68-expressing Mϕ lineage cells is required for G-CSF-induced HSPC mobilization [69]. As multiple Mϕ subtypes, including monocytes and some myeloid progenitors, express CD68 [70], it remains unclear if G-CSF acts directly in resident Mϕs to drive mobilization or if its action is indirect. Interestingly, G-CSF also reduces the number of Mϕs in the BM [19, 69, 71], thus, it is possible that G-CSF-induced Mϕ loss drives HSPC mobilization. Mϕ depletion was also found to synergize with G-CSF and AMD3100 treatment to increase circulating HSPCs [18], underscoring the potential clinical relevance of targeting Mϕs for mobilization. These data also suggest non-redundant roles for G-CSF and Mϕ-depletion during HSPC mobilization. As Mϕ depletion alone expands the BM HSC pool, it is possible that the G-CSF-dependent reduction in Mϕs induces an expansion of the HSPC pool prior to their mobilization. Alternatively, as Mϕs retain HSPCs in the spleen [62], Mϕ depletion may release a splenic reservoir of HSPCs, thus enhancing G-CSF induced mobilization. Indeed, G-CSF-dependent mobilization is increased in splenectomized mice [63], thus supporting the idea that splenic Mϕs may play a role in sequestering circulating HSPCs. The mechanism whereby G-CSF-mediated signaling reduces BM-resident Mϕs is still unclear, however G-CSF causes Mϕ relocation from the bone lining to the sinusoids, suggesting that G-CSF may induce Mϕ emigration from the BM [19]. G-CSF enhances splenic hematopoiesis at the expense of BM hematopoiesis [72], thus it is tempting to speculate that BM Mϕs seed peripheral sites after G-CSF treatment to support extramedullary hematopoiesis (EMH).

Macrophage regulation of steady state granulopoiesis

Neutrophils are short-lived cells produced in large quantities in the BM, and their clearance, or efferocytosis, by Mϕs in the BM, liver, and spleen is critical in maintaining blood homeostasis [23]. Efferocytosis also regulates steady state granulopoiesis and HSPC trafficking. Mice lacking the antiapoptotic gene, cellular FLICE-like inhibitory protein (c-FLIP), under the LyzM-Cre promoter have a severe Mϕ deficiency and, as a consequence, exhibit neutrophilia, which is partially G-CSF dependent [24]. Mϕ and DC-dependent efferocytosis of senescent neutrophils suppresses IL-23 production, a cytokine that is important for IL-17- and G-CSF-dependent neutrophil production, survival, and recruitment [25]. Thus, Mϕs participate in a negative feedback loop to suppress neutrophil production via their uptake. Furthermore, Mϕ efferocytosis of aged neutrophils induces HSPC mobilization by reducing niche-dependent CXCL12 via LXR signaling and circadian rhythms [22]. Therefore, Mϕs are central regulators of both granulopoiesis and HSPC trafficking at steady state.

IV. Macrophages Regulate Stress-Hematopoiesis

Inflammation impacts hematopoiesis

Hematopoiesis can be markedly changed during infection or in response to inflammation. These changes are mediated by production of cytokines, chemokines, and growth factors, as a response for generation of specific cell types under unique stress conditions [73]. Sensing of pathogen associated molecular patterns (PAMPs), and cytokine and chemokine signaling can occur by both HSPCs directly [61, 74–76] as well as via cells of the HSC niche [77, 78]. Mϕs are abundant cells in most tissues, including the BM, and are poised to respond to systemic cues via their expression of multiple cytokine and chemokine receptors. The contribution of Mϕs in regulating demand-adapted hematopoiesis in the BM, however, has not yet been explored. Among other inflammatory mediators, the growth factor, G-CSF and the cytokine IFNγ significantly impact the hematopoietic compartment. Moreover, they both directly modulate Mϕ numbers and function, suggesting that Mϕs may represent a BM niche component central to inflammation-induced hematopoiesis (Figure 1).

Figure 1. G-CSF and IFNγ differentially modulate BM Mϕ numbers, HSPC pool size, and lineage fate.

Mϕs reside within the HSPC niche, where they efferocytose senescent neutrophils, suppress granulopoiesis, and mobilize HSPCs via LXR signaling and circadian rhythms. Circulating HSPC levels are also influenced by their retention in the spleen via Mϕ-dependent VCAM-1 binding to VLA-4. G-CSF (Left column): reduces BM Mϕ numbers which correlates with HSPC expansion in the BM, blood, and spleen. Reduced Mϕ numbers also correlates with a lift on the suppression of granulopoiesis. IFNγ (Right column): increases BM Mϕs and signaling in Mϕs restricts the pool of HSPC, limits circulating HSPCs, and increases monopoiesis. Splenic Mϕs and IFNγ also enhance EMH in the spleen which may also sequester circulating HSPCs.

Effects of G-CSF on BM macrophages and emergency granulopoiesis

HSPCs and neutrophils are retained within the BM via similar mechanisms (i.e. CXCL12 and VLA4), therefore, G-CSF is also a robust inducer of neutrophil mobilization and production during stress conditions [64, 79]. Loss of neutrophils from the BM, due to infection, administration of the adjuvant Alum, or antibody-mediated depletion, results in HSPC proliferation and production of neutrophils in a process that requires G-CSF, IL-1R-, and TLR4-mediated signaling [80, 81]. This is thought to occur through a density feedback mechanism that detects cell loss, yet the sensors of this cell loss have not been elucidated [81]. Notably, antibody-mediated neutrophil depletion, which enhances emergency granulopoiesis, requires Mϕ-dependent efferocytosis [82]. Thus, while efferocytosis suppresses granulopoiesis at steady state [24, 25, 83], this process may enhance granulopoiesis under stress conditions where rapid neutrophil turnover and clearance occur. As such, Mϕ-dependent efferocytosis may differentially regulate granulopoiesis at homeostasis and during inflammation. Although Mϕs can produce G-CSF [23, 84], it was recently established that endothelial cells are key sensors of a high dose of LPS, reflecting a systemic infection, and produce G-CSF in a MyD88-dependent manner inducing emergency granulopoiesis [85]. Thus, it is possible that endothelial-specific G-CSF reduces BM Mϕs thereby enhancing granulopoietic responses. At the same time, Mϕ-dependent production of G-CSF may act locally within the BM to enhance granulopoiesis under less extreme, infectious conditions. Much of what is understood about Mϕs and granulopoiesis comes from steady state conditions, yet G-CSF signaling in Mϕs and efferocytosis during stress conditions likely impact the initial stages of granulopoiesis, and warrants further investigation.

Effects of IFNγ on BM macrophages

Interferon gamma (IFNγ) is a Th1 type pro-inflammatory cytokine that directs innate and adaptive immune responses against intracellular pathogens. In concert with other factors, IFNγ promotes classical Mϕ activation, which is critical for cytokine production, intracellular pathogen killing, and antigen presentation to T cells [86, 87]. While IFNγ directs mature effector cell function, it also modulates immune responses by influencing HSPC activity and differentiation. IFNγ activates HSC proliferation [88] and subsequently boosts myelopoiesis [89]. Monocyte production correlates with pathogen clearance and IFNγ specifically promotes monopoiesis during many intracellular infections [90, 91].

While increased immune cell production benefits the host during acute insults, HSC exhaustion and bone marrow failure can occur if inflammation is left unchecked. IFNγ negatively regulates HSC function both in vitro and in vivo and is often associated with BM failure [92–94]. Therefore, understanding how IFNγ impacts HSC numbers and function is relevant to the treatment of hematologic conditions. Depending on the nature of the experiment or disease model used, IFNγ can influence HSC activity directly [74, 88] or indirectly, via neighboring cells in the BM niche [39, 95]. In mixed BM chimeric mice infected with the intracellular bacteria E. muris, IFNγ did not act directly on HSCs to drive their loss but rather, acted through an indirect pathway. By using Mϕ-depletion and transgenic mice with Mϕ-lineage cells that are insensitive to IFNγ (MIIG mice), IFNγ was shown to act on Mϕs to drive the loss in HSC numbers and function during infection. IFNγ increased CD11b⁺ Mϕs and maintained CD11b^low Mϕs in the BM, yet, the mechanism by which IFNγ signaling in Mϕs restricts HSC numbers is unknown [39]. During LCMV infection, IFNγ causes severe anemia by inducing Mϕ-dependent phagocytosis of RBCs, or hemophagocytosis [96]. Thus, aberrant IFNγ signaling in Mϕs may promote hemophagocytosis of HSCs in a similar manner. Alternatively, in response to IFNγ, Mϕs may produce additional inflammatory factors (i.e. TNFα, IL-6, NO) that either act directly in HSCs, or their niche, to rapidly drive differentiation or cell death.

IFNγ also influences HSPC localization in circulation and in extramedullary tissues. IFNγ restricts blood and splenic HSPC numbers during bacterial infection, and this correlated with increased BM Mϕs [39, 97]. While consistent with the finding that BM Mϕ numbers inversely correlate with circulating HSPC numbers [18, 19, 69], HSPCs were also reduced in the BM. Therefore, IFNγ likely restricts the entire HSPC pool rather than increasing retention of HSPCs in the BM during infection. Notably, reduced circulating HSPCs during infection [39] was accompanied by enhanced splenomegaly [98], thus it is possible that IFNγ limits detection of mobilized HSPCs due to peripheral tissue sequestration and rapid differentiation in situ. In this regard, HSC pool restriction may depend upon where IFNγ is produced as BM-derived IFNγ may restrict the HSC pool locally, whereas systemic IFNγ could enhance EMH in peripheral tissues. Similarly, BM and splenic Mϕs may create very different niches that affect how HSCs respond to IFNγ and other inflammatory mediators.

Mϕs can promote splenomegaly during inflammatory conditions [3, 99] and Mϕ depletion alone is sufficient to reduce splenic cellularity [100] and abrogate IFNγ-dependent splenomegaly during E. muris infection (Figure 2). Understanding how IFNγ and Mϕs regulate EMH is clinically relevant as EMH contributes to host defense, but when left unchecked, can promote pathology during autoimmune disorders [101]. Splenic VCAM-1+ Mϕs, in particular, contribute to EMH and pathogenesis during atherosclerosis through HSPC retention, which facilitates myeloid cell production [62]. Additionally, IFNγ induces splenic EMH to drive pathogenesis during colitis [102] and allergic inflammation [103], but promotes the clearance of malaria-infected erythrocytes [104]. In light of the IFNγ- and Mϕ-dependent consequences on host defense, disease pathology, and HSPC function, it will be important to understand how tissue-specific Mϕ subtypes alter hematopoiesis under different inflammatory conditions.

Figure 2. Mϕ depletion reduces splenic cellularity and abrogates infection-induced splenomegaly.

Mice were infected with 5e4 copies Ehrlichia muris and administered PBS- or clodronate-encapsulated liposomes (Clod) on day 4 and 6 post-infection (250ul i.v.). On day 11 post-infection, spleens were homogenized with frosted slides and cells were enumerated. A) Graph represents the number of spleen cells in control uninfected (white bar) and Em-infected (black bar) mice. B) Photograph of spleens from Em-infected mice treated with PBS or clodronate-encapsulated liposomes. The mean ± SEM is shown. Two-tailed student’s t-test was used to compare between groups. ^*p<0.05, ^**p<0.01, ^***p<0.001.

Opposing roles of G-CSF and IFNγ on macrophages

IFNγ and G-CSF rise concurrently during many inflammatory conditions, however, they often act in an opposing manner when shaping the immune response. IFNγ is essential for clearance of many intracellular pathogens, while G-CSF is a critical mediator of extracellular bacterial and fungal clearance. While both IFNγ and G-CSF impact mature effector cells to combat infection, they also influence the production of immune cells in opposing ways. IFNγ can induce monopoiesis over granulopoiesis, by inhibiting G-CSF dependent signaling through STAT3 [90]. This was demonstrated in myeloid progenitors in vitro, however, IFNγ may also act on Mϕs in a similar manner. IFNγ stimulation of Mϕs in vitro can suppress an LPS-induced gene program and neutrophil chemotaxis [105]. Moreover, immature granulocytes are robustly increased in the BM of E. muris-infected mice when IFNγ signaling is abrogated in Mϕs (MIIG mice) or after Mϕ depletion (Figure 3A), suggesting that IFNγ signaling in Mϕs restricts granulocyte production. Additionally, while IFNγ favors increased circulating monocytes over neutrophils during infection [89, 90], a shift toward circulating neutrophils was observed in MIIG mice during E. muris infection (Figure 3B). Thus, IFNγ promotes monopoiesis at the expense of neutrophil production in a process that involves Mϕ-lineage cells.

Figure 3. IFNγ signaling in Mϕ-lineage cells restricts immature granulocytes in the BM and mature neutrophils in circulation.

Mice were infected with 5e⁴ copies Ehrlichia muris via intraperitoneal injection and tissues were collected on day 11 post-E. muris infection. Blood was collected in 0.5M EDTA and BM was flushed from one femur and tibia, RBCs were lysed with ammonium chloride Tris buffer, and single cell suspensions were surfaced stained with fluorescently conjugated antibodies and analyzed by flow cytometry on an LSR II and analyzed using FlowJo software. A) Graph represent the absolute frequency of immature granulocytes (CD11b+Ly6Cint.Ly6Glow) in the BM of control uninfected (white bar) and Em-infected (black bar) MIIG and LC mice (left panel) or 7 days post PBS- or Clodronate-encapsulated liposome administration (right panel; 250ul i.v.). The mean ± SEM is shown. Two-tailed student’s t-test was used to compare between groups. ^*p<0.05, ^**p<0.01, ^***p<0.001, ^****p<0.0001. B) Graph represents complete blood cell count (CBC) for white blood cells in Em-infected MIIG or LC mice. CBCs were determined on an automated hematology analyzer (Advia 120; Bayer Corporation, Norwood, MA).

IFNγ and G-CSF also affect BM and peripheral HSPC numbers in an opposite manner. While IFNγ drives BM HSC loss during infection, HSC numbers are rescued upon G-CSF administration [39]. Additionally, G-CSF- and AMD3100-induced mobilization is impaired during bacterial infection where IFNγ production is robust [39]. Whether mobilization is less efficient due to a reduced overall HSPC pool or increased HSPC sequestration in peripheral tissues is not clear. However, this finding is directly relevant to improving the efficiency of HSPC mobilization in the clinic as aberrant IFNγ signaling and Mϕ function may reduce the availability of circulating HSPCs obtained for transplantation. Modulating Mϕ numbers impacts the hematopoietic compartment [1, 18, 19, 24, 39, 69], thus it is possible that differential Mϕ responses to these cytokines is central to tailoring blood cell production by impacting HSPC activity.

V. Conclusions and Future Directions

In conclusion, recent studies reveal a role for BM and splenic Mϕs in regulating HSPC location, pool size/quiescence, and lineage commitment, both at steady state and during inflammation. IFNγ and G-CSF differentially impact Mϕ numbers and this correlates with changes in BM and circulating HSPCs. It will be important to more fully understand the underlying mechanisms whereby Mϕs regulate HSPCs. The origin of tissue resident Mϕs has been reported to change under inflammatory conditions and impact disease outcome in the liver and heart [9, 106]. Therefore, it is possible that BM Mϕ-dependent changes in the hematopoietic compartment may also be due to a change in Mϕ origin and function during disease. Thus, further investigation into the precise mechanisms by which IFNγ and G-CSF affect Mϕs, in physiologically relevant settings, is necessary to our understanding of inflammation-induced hematopoiesis.

An obstacle that remains when studying inflammation-induced hematopoiesis is examining the specific contribution of Mϕs in the BM when compared to other vascularized tissues such as the spleen and liver. This will require more stringent definitions between tissue resident Mϕ subtypes and the development of methods to target Mϕs in specific tissues. Although not discussed in this review, changes in Mϕ origin and function during disease have made comparative studies more challenging, thus, distinguishing the origin of tissue Mϕs during inflammation (self-renewal vs. monocyte precursors) will also provide insight into how Mϕ function is altered during stress conditions. Ultimately, defining the Mϕ-dependent factor(s) that support and adversely affect niche cell function and HSPC pool size will be invaluable for improving mobilization and transplantation protocols and treating hematologic disorders.

• An emerging role for Mϕs in regulating HSC pool size and HSPC mobilization is discussed.

• Evidence that Mϕs influence demand-adapted myelopoiesis is presented.

• The opposing effect of IFNγ and G-CSF on Mϕ numbers and HSC pool size is highlighted.