Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

In 1908 the Nobel Prize in Physiology and Medicine was awarded jointly to Paul Ehrlich and Elie Metchnikoff “in recognition of their work on immunity,” for their contribution in defining what is today known as adaptive and innate immunity, respectively. In the following century, studies on adaptive immunity boomed, leading to paramount discoveries in the field, while innate immunity was almost forgotten. It is only in the last few years that new highlights into macrophage biology have reverted the trend leading to the recognition of the important role of macrophages not only within the immune response but also in several physiological and pathological processes within the body. An increasing amount of evidence redefines macrophages not only as cells of the innate immune system designated to eliminate invading pathogens but also as polyfunctional and plastic cells that define and preserve the tissue integrity at steady state and restore homeostasis upon disturbances. This evidence encompasses their origin (not only from adult bone marrow but also from embryonic hematopoiesis) (Hoeffel and Ginhoux 2015), their heterogeneity in different tissues (they have specific specializations and roles depending on the tissue where they are located) (Gordon and Taylor 2005; Gordon et al. 2014), their capacity to acquire different functional phenotypes (M1/M2 polarization: they are responsible for triggering an inflammatory reaction and afterward of driving its resolution, tissue repair, and remodeling) (Gordon and Taylor 2005; Biswas and Mantovani 2010), and their ability to react differently upon a second challenge (tolerant and trained responses, also known as innate memory) (Netea et al. 2015). Thus, the role of macrophages has changed from sentinels of the body or “trash disposal units” serving at the bequest of T and B cells, to main “conductors” and effectors of the inflammatory response, and to immune memory cells, able to remember (albeit not specifically) previously encountered stimuli. All these new concepts on monocytes/macrophages are revolutionary and will force the researchers to rewrite the textbooks of immunology, finally giving to macrophages the attention they deserve.

What was defined as the reticuloendothelial system (that included all cells with phagocytic capacity) by K.A.L. Aschoff in 1924 (Aschoff 1924) in the 1970s was redubbed the “mononuclear phagocyte system (MPS),” excluding fibroblasts and endothelial cells and including monocytes, all types of macrophages, and dendritic cells (DCs). After a long and hot discussion on the origin of macrophages, the position of Ralph van Furth prevailed that all tissue-resident macrophages are derived from bone marrow or bone marrow-derived blood monocytes (van Furth et al. 1972). Thus, despite a number of studies challenging this position, until few years ago, the common knowledge was that tissue macrophages derive from monocytes, which terminally differentiate in the tissue and are unable to proliferate. The discovery of an embryonic origin of tissue-resident macrophages and of their capacity of in situ self-renewal without loss of their differentiated cellular identity is now well established and generally accepted (Jenkins et al. 2011; Gentek et al. 2014). Amidst new evidence on macrophage biology, this is the one that raises the need of understanding of the characteristics of tissue macrophages as compared to monocytes (that are no longer simple precursors of tissue macrophages), the intrinsic differences between these cell types, and their relative involvement in building and maintaining tissue homeostasis in normal conditions and during an inflammatory reaction.

Without going in detail of the aforementioned aspects (ontogeny, heterogeneity, polarization, memory), which are excellently reviewed elsewhere (Netea et al. 2015; Gentek et al. 2014; Gomez Perdiguero and Geissmann 2016; Murray 2017), this review will summarize the recent advances in our understanding of the origin of tissue macrophages and of their differences from monocytes and monocytes-derived macrophages.

Blood circulating monocytes can be recruited into a tissue in different circumstances. In physiological conditions, they can enter a tissue for replenishing the pool of resident macrophages that maintain homeostasis. In pathological events, they are recruited for fighting an infection, for restricting the growth of a tumor, or for repairing damage. Once monocytes migrate into a tissue, then, by definition, they become macrophages. Although some investigators use the term tissue monocytes, we support the opinion that the term monocyte should be restricted to cells in the blood compartment and also in the bone marrow and spleen, given that these are reservoirs that can replenish the blood monocyte pool (Ziegler-Heitbrock 2015). Monocytes recruited into the tissue are often referred to as bone marrow-derived macrophages, monocyte-derived macrophages (moMϕ), inflammatory monocytes/macrophages, nonresident macrophages, peripherally derived inflammatory macrophages (newly differentiated macrophages within inflammatory lesions), and so on. Throughout the chapter when we mention monocytes, we will refer to the bone marrow-derived and circulating monocytes (not yet activated and differentiated into macrophages in the tissue), while when we mention moMϕ, we will refer to the bone marrow- and monocyte-derived cells newly recruited in the tissue, in order to distinguish them from tissue-resident macrophages (trMϕ) originating in the embryo and also known as embryonic-derived macrophages.

2 Heterogeneity of Circulating Monocytes and Tissue-Resident Macrophages

Historically, the MPS includes monocytes, DCs, and macrophages, distinguished on the basis of functional and phenotypic features (van Furth et al. 1972). A number of issues have complicated the picture leading to much confusion regarding the definition of these cell types and their related subsets. First, the phenotypic markers proposed for the identification often are not unique, and many functions are shared/overlapping between cell types (Hume 2008; dos Anjos Cassado paper 2017). Second, the markers used for defining a particular cell or cell subset are not always consistent between mice and humans (Reynolds and Haniffa 2015; Ziegler-Heitbrock 2014). Third, many trMϕ are of embryonal origin (see below) as opposed to monocytes and DCs that originate from adult hematopoietic stem cell (HSC)-derived progenitors, although there are no markers that can discriminate them uniquely. In order to understand which phenotypical and functional differences distinguish these cell types and to avoid confusion due to the aforementioned issues, a brilliant and fully sharable unified nomenclature has been recently proposed, based primarily on the ontogeny of myeloid cells and secondarily on their location (Guilliams et al. 2014).

Without going into DCs and the hot debate about their origin and relation to monocytes and macrophages (Hume 2008; Geissmann et al. 2010), the focus of this section will be on monocytes and trMϕ. Monocytes are the cells of the myeloid lineage that circulate in the blood stream, while macrophages are the myeloid cells that colonize the tissues throughout the body. Monocytes are generated during adult hematopoiesis from HSC progenitors in the bone marrow (Ginhoux and Jung 2014), while macrophages originate predominantly during embryonic hematopoiesis, from progenitors in the yolk sac (YS) and fetal liver (FL) (Epelman et al. 2014a; Wynn et al. 2013). The ontogeny/origin of macrophages will be described in the next section. Both monocytes and macrophages are involved in host defense against infections, in maintaining tissue homeostasis and development (Gordon et al. 2014; Wynn et al. 2013), in initiation and resolution of the inflammatory reaction (Italiani and Boraschi 2014), in sensing tissue damage and orchestrating tissue repair and remodeling (Mantovani et al. 2004), and in disease progression (e.g., cancer; Noy and Pollard 2014). While macrophages are already present in tissues, blood monocytes are recruited into the tissue mainly in response to a dangerous event (e.g., damage, infection, or disease) and are the main effectors of the inflammatory reaction. In some cases, monocytes can enter a tissue for homeostatic reasons, acting as cell reservoir for steady-state maintenance of the resident macrophage population (e.g., for dermis and intestine macrophage repopulation - Tamoutounour et al. 2013; Varol et al. 2015). When the challenges (infections or damages) are limited and weak, trMϕ can directly exert defensive functions. However, in order to successfully handle a vast and significant danger, trMϕ can locally proliferate, to increase their number, but their main role is that of initiating the recruitment of inflammatory monocytes from blood.

The advent of polychromatic flow cytometry has enabled the characterization/identification of distinct monocyte and macrophage subsets, demonstrating that both monocytes and macrophages are actually heterogeneous populations. In mice, two different monocyte subsets have been identified, based on the expression of a GPI-anchored surface glycoprotein (Ly6C) and two chemokine receptors (CX3CR1 and CCL2), the Ly6ChighCCR2+ and the Ly6ClowCX3CR1+ subsets. Both monocyte subsets can infiltrate peripheral tissues under appropriate stimuli. Ly6Chigh cells are essential for the inflammatory response, while Ly6Clow cells are able to patrol the vascular endothelium and survey its integrity (Auffray et al. 2007). They also seem able to extravasate rapidly into tissues before the arrival of Ly6Chigh inflammatory monocytes (Sumagin et al. 2010), although their extravasation into tissue is apparently a rare event (Auffray et al. 2007). A number of recent reviews describe the functions and relationship of these two populations (Ziegler-Heitbrock 2015; Ginhoux and Jung 2014; Gutknecht and Bouton 2014; Sprangers et al. 2016). Evidence suggests that Ly6Chigh monocytes give rise to Ly6Clow monocytes, both in the bone marrow and in the circulation (Yona et al. 2013; Hettinger et al. 2013). In humans, three different subsets of blood monocytes are identified based on the relative expression of the surface markers CD14 and CD16: CD14++CD16− (classical), CD14++CD16+ (intermediate), and CD14+CD16++ (nonclassical) monocytes. The functions and the differences between these three subsets are still controversial (Ziegler-Heitbrock 2007; Wong et al. 2012). The classical and nonclassical human monocytes are homologous to their counterparts in mice, Ly6Chigh and Ly6Clow, respectively. It seems that classical monocytes are inflammatory and phagocytic, even if it seems that the intermediate are mainly involved in various inflammatory diseases (Wong et al. 2012; Fingerle et al. 1993) and have been shown to be of prognostic relevance in cardiovascular diseases (Rogacev et al. 2012).

Macrophages are present in all tissues, and their different phenotypes and functional specializations in specific tissue microenvironments explain their heterogeneity (Gordon and Taylor 2005; Epelman et al. 2014a, b; Haldar and Murphy 2014). The heterogeneity of trMϕ is the result of the combination of tissue-identity signals, which promote their local differentiation, and the availability of functional demand signals, which induce different functional phenotypes of the macrophages. In this way, the overall phenotype of trMϕ is perfectly tailored to the needs of the tissue in which they reside, as an integration of stable and irreversible differentiation and stable and reversible polarization programs. Recently, this matter has been described in detail in an excellent review (Okabe and Medzhitov 2016). Thus, macrophages take different names according to their tissue location and specialized functions. For example, bone macrophages are known as osteoclasts and are highly specialized in bone resorption (Boyle et al. 2003). In the lung alveoli, they are known as alveolar macrophages and are responsible for recycling of surfactant molecules (Hussell and Bell 2014). In the brain, trMϕ are known as microglia and sustain the brain development through trophic functions (synaptic pruning, regulation of neuronal precursor numbers, and production of neurotrophic factors) (Schafer et al. 2012; Parkhurst et al. 2013). Liver macrophages are known as Kupffer cells and support the tolerogenic milieu existing within the liver, ensure protection during infections, and maintain iron homeostasis (Eckert et al. 2015). In the skin, they are known as Langerhans cells (LCs) and are involved in immune functions and epidermis/skin homeostasis (Collin and Milne 2016). In the spleen, red pulp macrophages are involved in the processing of heme and iron from senescent erythrocytes (Kohyama et al. 2009), while macrophages in the splenic marginal zone are important for the capture of blood-borne antigens (Aichele et al. 2003) and for the proper positioning of B cells (Karlsson et al. 2003). Macrophages in the bone marrow have a crucial role in the retention of hematopoietic stem cells (HSCs) in the bone marrow (Chow et al. 2011). Intestinal macrophages in the muscularis mucosa engage in crosstalk with surrounding muscle cells and enteric neurons to ensuring normal intestinal peristalsis (Muller et al. 2014).

Different gene expression profiles, transcriptional regulatory pathways, and enhancer activities, all driven by the local microenvironment, underlie tissue-specific macrophage identities and specialized functions. All these aspects linked to transcriptional and epigenetic programs are excellently reviewed elsewhere (Romanoski et al. 2015; Glass and Natoli 2016; Amit et al. 2016; Lavin et al. 2014; Lavin et al. 2015; Gosselin et al. 2014).

3 Origin of Macrophages

Hematopoiesis is the process by which lymphoid and myeloid lineage cells of the blood are formed, through a tightly regulated stepwise process involving several progenitor cells (De Kleer et al. 2014). Hematopoiesis occurs in several temporally and spatially regulated waves during the embryonic development (“primitive hematopoiesis”) in the YS and culminates with the generation of HSCs in the FL before birth and in the bone marrow (BM) after birth and during adulthood (“definitive hematopoiesis”). An exhaustive description of hematopoiesis and of the ontogeny of monocytes and trMϕ is beyond the aim of this review. However, the section below will summarize the key steps of the development of mononuclear phagocytes referring to definitive/adult hematopoiesis as the main source of circulating monocytes and to primitive/definitive/fetal hematopoiesis as the main source of trMϕ.

3.1 Monocytes and moMϕ Development: Definitive/Adult Hematopoiesis

Lineage-committed mononuclear phagocytes, i.e., circulating monocytes and moMϕ, are generated primarily in the bone marrow from tightly regulated differentiation of HSCs to progressively more committed cell population. High levels of the transcription factor c-Myc induce the differentiation of HSCs toward more committed progenitors (Wilson et al. 2004). At the top of monocyte/macrophage development, there is the lineage-determining master regulator PU.1 that, upon activation by CSF-1 (also known as M-CSF) and IL-34 (Mossadegh-Keller et al. 2013), controls the gene expression module common to all monocytes/macrophages and is the crucial regulator of myeloid lineage commitment at the fork in the road between myeloid and lymphoid fates (McKercher et al. 1996; Nerlov and Graf 1998). The transcription factor MafB represses the activity of CSF-1 (Sarrazin et al. 2009), and its expression is reduced in myeloid-committed cells in respect to HSCs. CSF-1 not only directs lineage commitment of HSCs toward CMP (common myeloid progenitors—lymphoid potential is lost) but also of GMP (granulocyte-macrophage progenitors) toward MDP (monocyte-macrophage DC progenitors—granulocyte potential is lost). Recently, the purification of the more restricted MDP population in mouse allowed to define a precursor from which monocytes and DCs are derived, that is, the distinction between CDP (common DC progenitors), a DC-restricted precursor in the BM, and cMoP (common monocyte progenitors), a monocyte-restricted BM precursor that represents the more differentiated, committed monocyte population (Hettinger et al. 2013). cMoP give rise to both monocyte subsets Ly6Chigh and Ly6Clow. A cMoP monoblast type of cell has not been identified yet in humans. Figure 2.1A summarizes the sequential steps of monocyte development.

Fig. 2.1
figure 1

Origin of monocyte-derived macrophages (moMϕ) and tissue-resident macrophages (trMϕ). (A) moMϕ are generated in the bone marrow (BM) through the tightly regulated differentiation of HSCs to progressively more committed cells up to monocytes. This hematopoiesis is defined as “definitive hematopoiesis” and occurs in fetal liver (FL) during the late phase of the embryonic development (before birth) and in BM during adulthood (after birth). Monocytes are released to the bloodstream and can be recruited into the tissue where they differentiate into macrophages. HSCs hematopoietic stem cells, CMPs common myeloid progenitors, GMPs granulocyte-macrophage progenitors, MDPs monocyte-macrophage DC progenitors, CDPs common dendritic cell progenitors, cMoPs common monocyte progenitors. (B) trMϕ arise from multiple sources during the embryonic development with a sequential timing: from yolk sac (YS), FL and aorta-gonads-mesonephros (AGM) regions according to three different routes. These routes are a simplification of the three main successive waves of the embryonic hematopoiesis and also of the three proposed models for macrophage ontogeny. (1) The first wave arises directly from YS, which produces “early” erythro-myeloid progenitors (EMPs) (“primitive hematopoiesis”) from which YS macrophages are generated. To date, the progenitors giving rise to macrophages are poorly characterized. The first model proposes that these YS-derived macrophages represent the main precursors for the most trMϕ and not exclusively for microglia. (2) The second wave generates “late” EMPs that could migrate from YS into FL and could represent transient definitive progenitors. The second model proposes that these late EMPs represent the main precursors for most trMϕ, with the exception of microglia, through a monocytic intermediate (late EMP-derived fetal monocytes). (3) The third wave starts with the generation of immature HSCs in the AGM that colonize the FL where they establish a “definitive hematopoiesis” and maybe seed the fetal bone marrow (BM) generating HSC-derived fetal monocytes. Then, these cells will finally lead to generation of HSCs in the BM during adulthood. The third model hypothesizes that trMϕ (except microglia) arise from HSC-derived fetal monocytes, and these cells, rather than late EMPs, might generate FL monocytes. PL placenta

Full size image

3.2 trMϕ Ontogeny: Primitive/Fetal Hematopoiesis

Until few years ago, it was believed that tissue macrophages derive entirely from circulating blood monocytes, through adult hematopoiesis. Given that there are some tissues which require blood-borne precursors to replenish the pool of resident macrophages, such as the dermis (Tamoutounour et al. 2013), gut (Bain et al. 2014), mammary gland (Franklin et al. 2014), and heart (Epelman et al. 2014a, b), the resident macrophage pool of most tissues derives from embryonic precursors that colonize these tissues prior to birth and is maintained locally through in situ proliferation in adulthood (Gentek et al. 2014; Sieweke and Allen 2013). In the early 2000s, a series of elegant fate-mapping experiments, experiments in parabiotic mice, and genetically engineered mouse models (Tamoutounour et al. 2013; Yona et al. 2013; Bain et al. 2014; Epelman et al. 2014a, b; Ajami et al. 2007; Ginhoux et al. 2010; Guilliams et al. 2013; Hashimoto et al. 2013) demonstrated once and for all that trMϕ (e.g., microglia, LCs, alveolar macrophages, Kupffer cells) originate from early embryonic precursors prior to birth and that the extent to which they can originate from adult HSCs depends on the context and tissue (Ginhoux and Jung 2014). Assuming that trMϕ have an embryonic origin, the next step was to understand the embryonal hematopoiesis and from which precursor tissue macrophages derive. Mammalian embryonic hematopoiesis is a complex process that makes particularly challenging the goal of determining the exact ontogeny of fetal macrophages (for review, see Hoeffel and Ginhoux 2015). In the mouse, embryonic hematopoiesis is characterized by distinct waves, occurring in different districts of the embryo and in a sequential way. The first wave arises from the blood island of YS around E7–7.5 (embryonic day) and gives rise to the so-called erythro-myeloid progenitors (EMPs). This phase is termed “primitive hematopoiesis” and generates macrophages without going through a monocytic progenitor (myb-independent hematopoiesis) (Gomez Perdiguero and Geissmann 2013). Actually, EMPs have been renamed “early EMPs,” to distinguish them from the “late EMPs” that arise from the YS hemogenic endothelium at E8–8.5. This phase represents the second wave of hematopoiesis and is characterized by the emergence of lympho-myeloid progenitors (Li et al. 2014). This wave is called “transient hematopoiesis” because it does not persist upon transplant in immune-compromised mice (Hoeffel and Ginhoux 2015; Hoeffel et al. 2015). At E8.5 the blood circulation is established, and EMPs are able to seed the FL, where they expand and generate fetal monocytes. Concomitantly with “late EMPs” at E8.5, at third wave arises from the intraembryonic hemogenic endothelium, which generates immature HSCs in the para-aortic splanchnopleura (P-Sh) region and proceeds to give rise to fetal HSCs in the aorta, gonads, and mesonephros (AGM) region at E10.5. Then, these precursors colonize both the FL, where they establish a “definitive hematopoiesis” (myb-dependent) (Hoeffel et al. 2015), and fetal BM, where they finally will generate adult BM HSCs. From E12.5 onward, the FL becomes the main hematopoietic organ within the embryo. While the embryonic origin of certain tissue macrophages is now accepted, the exact identity of progenitors, the exact pathway of differentiation to mature cells, and the transcription factors involved are still unknown. Three different models of the macrophage embryonic ontogeny have been proposed: (1) YS-derived macrophages represent the main precursors for most trMϕ (Gomez Perdiguero et al. 2015a), (2) “late EMPs” could represent the main precursors for most trMϕ through a monocytic intermediate (Hoeffel et al. 2015), (3) fetal HSCs are the main precursors of FL monocytes, and adult macrophages (with the exception of microglia and partially LCs) arise from these definitive fetal HSCs (Sheng et al. 2015a). Which of the three hypotheses is valid remains still a matter of heated and constructive debate (Sheng et al. 2015b; Gomez Perdiguero et al. 2015b). For an exhaustive description of the three models and for an overview of the experiments on the ontogeny of macrophages, we refer the reader to the recent wonderful review by Ginhoux and Guilliams (Ginhoux and Guilliams 2016). Figure 2.1B summarizes the sequential steps of macrophage development.

4 trMϕ Versus moMϕ

According to the above evidence, we will consider trMϕ predominantly as macrophages derived from embryonic hematopoiesis that have colonized different tissues before birth and moMϕ as the mononuclear phagocyte population originating through definitive hematopoiesis after birth. This different origin has contributed not only to a primary distinction in terms of ontogeny but also to a secondary one in terms of functions. Generally, a conventional view implies that at steady-state the embryonic-derived trMϕ serve to maintain tissue homeostasis via surveillance of local tissue microenvironment. On the other hand, during inflammation (due to infection, disease, or trauma), it is generally accepted that circulating monocytes are predominantly recruited in the tissue, where they mediate various effector functions and differentiate into macrophages (moMϕ). To date it is not fully understood how much these two populations differ from each other and in their contribution to the initiation and resolution of the inflammatory response. Likewise, it is not known how much their functions differ due to diverse development origin (“nature”) or how much overlap due to influence of the environmental cues (“nurture”). The next sections will address these issues.

4.1 In Situ Macrophage Expansion/Proliferation Versus Monocyte Recruitment

In healthy tissues and in steady-state conditions, trMϕ are maintained by self-renewal with a variable contribution by circulating monocytes (Yona et al. 2013; Hashimoto et al. 2013; Gomez Perdiguero et al. 2015a, b). It has been observed that in situ proliferation versus peripheral BM-derived monocyte recruitment for the maintenance of the tissue macrophages can change based on the tissue. Thus, on the one hand, we observed that the brain microglia have the potential for efficient self-renewal without the contribution of BM-derived precursors (Bruttger et al. 2015), and the same happens for resident macrophages in the liver, epidermis, lung, pancreas, and splenic red pulp (Hashimoto et al. 2013; Chorro et al. 2009; Calderon et al. 2015). On the other hand, a crucial role for adult monocytes in replenishing the resident macrophage pool has been identified in the gut (Bain et al. 2014), dermis (Jenkins et al. 2011), heart (Epelman et al. 2014a, b; Molawi et al. 2014), and recently in the peritoneum (Bain et al. 2016).

During inflammatory conditions (upon infection, damage, or pathology), a large influx of BM-derived monocytes occurs, which differentiate into macrophage-like cells (moMϕ), and this differentiation occurs in parallel with the expansion of trMϕ within the tissue. The fate of these moMϕ is not known. For instance, it is not clear if they mostly die (as it has been postulated) or if some of them survive becoming memory macrophages (Italiani and Boraschi 2014) (see Fig. 2.2).

Fig. 2.2
figure 2

Fate of monocyte-derived macrophages (moMϕ) in steady state and inflammation. In the steady state, tissue-resident macrophages (trMϕ) derived from embryonic progenitors populate various tissues (i.e., the brain, epidermis, lung, liver) and maintain their mass by self-renewing proliferation. In the tissues with high rate of cell turnover (i.e., the gut, dermis, heart, pancreas, mammary glands), trMϕ can originate from bone marrow (BM)-derived circulating monocytes. During inflammation, a large number of BM-derived circulating monocytes enter all tissues where they differentiate into macrophage-like cells (moMϕ), which might proliferate. The main fates of moMϕ are highlighted: (1) they die during the effector phase of the inflammatory reaction, (2) they become part of the trMϕ, and (3) they preserve the “memory” of the inflammatory challenge

Full size image

Currently, the most relevant question is whether moMϕ are able to do self-maintenance in the tissue, i.e., the extent by which they survive and populate the inflamed tissue after recruitment, thereby contributing to replenishing trMϕ after the resolution of inflammation. Recent studies revealed that BM-derived monocytes display limited expression of proliferation genes as compared to FL monocytes and YS macrophages (Hoeffel et al. 2015; van de Laar et al. 2016) and often fail to stably persist in a tissue once inflammation resolves (Ajami et al. 2007; Hashimoto et al. 2013). Thus, they can be considered as “passenger myeloid cells” with respect to “resident counterparts of embryonic origin” (Gomez Perdiguero and Geissmann 2016). However, new evidence, obtained by using neonatal Csf2rb−/− mice with empty alveolar macrophage niche and a model of diphtheria toxin-mediated depletion of liver-resident Kupffer cells, demonstrates that the BM origin per se does not preclude the development of self-maintaining trMϕ (van de Laar et al. 2016, Scott et al. 2016).

Conversely, in the lung and brain, BM-derived monocytes do not seem to substantially contribute to the resident macrophage population after the resolution of infection or injury (Hashimoto et al. 2013; Ajami et al. 2011). In the epidermis, a small subset of moMϕ has been observed that may be remnant of monocytes recruited upon tissue damage (Sere et al. 2012), despite LCs also demonstrate a local proliferation (Chorro et al. 2009). In the liver, the nature of the damage determines the contribution of moMϕ to the pool of Kupffer cell. Upon bacterial infection, moMϕ repopulate trMϕ (Blériot et al. 2015), whereas upon paracetamol-induced injury, Kupffer cells proliferate and expand in the liver (Zigmond et al. 2014). Upon viral infection, Kupffer cells have a beneficial antiviral effect in the early phase after infection but seem to suppress the antiviral immunity during chronic infection. Moreover, it is difficult to distinguish the contribution of Kupffer cells versus infiltrating monocytes/macrophages because of the lack of distinctive phenotypical markers (Ju and Tacke 2016). Likewise, local proliferation of the resident macrophage population has been observed in atopic dermatitis (Chorro et al. 2009), while a contribution from moMϕ has been observed in ultraviolet irradiation-induced skin damage (Ginhoux et al. 2006). It has been observed that not only trMϕ but also BM-derived moMϕ proliferate in the zymosan-induced inflammatory reaction in the mouse peritoneal cavity (Davies et al. 2013). This proliferation depends on CSF-1 but is independent of IL-4, suggesting that the in vivo proliferation of trMϕ mediated by IL-4 upon nematode infection (Jenkins et al. 2011; Jenkins et al. 2013) may be restricted to type 2 inflammatory reactions. Recently, multiple fate-mapping approaches demonstrated that arterial macrophages arise from CX3CR1+ embryonic precursors and postnatally from BM-derived monocytes. Arterial macrophages are maintained by self-renewal and local proliferation, without a substantial contribution from blood monocytes in adulthood and after severe depletion during polymicrobial sepsis (Ensan et al. 2016). On the other hand, moMϕ are able to proliferate during pancreatic injury (Van Gassen et al. 2015) and during hemolysis and erythrocyte damage in the red pulp spleen (Haldar et al. 2014). Monocytes contribute to the resident macrophage population also during inflammation due to pathological conditions such as atherosclerosis (Tacke et al. 2007) (where an important local proliferation of trMϕ is also observed - Robbins et al. 2013), cardiac inflammation (Epelman et al. 2014b), or aged heart (Molawi et al. 2014), in spinal cord injury (Shechter et al. 2009) but not in brain injury and neurodegeneration (Ajami et al. 2007).

4.2 Are moMϕ and trMϕ Phenotypically Distinguishable?

Upon inflammation, monocytes may infiltrate the tissue and differentiate into macrophages (moMϕ). This raises the crucial questions how to distinguish them from trMϕ, how much they are similar or differ in terms of gene expression/phenotype and in terms of their function in the inflammatory reaction, and the role of the environment in modulating their functional programming. These aspects are still not fully established. Using a genotoxic irradiation model, in which embryonic (host)- and postnatal (donor)-derived macrophages coexist in the tissue, and comparing the transcriptome between embryonic macrophages and BM moMϕ, more than 90% identity of gene expression has been observed in the lung, peritoneal cavity, and liver (Lavin et al. 2014; Scott et al. 2016; Gibbings et al. 2015; Beattie et al. 2016), although a few phenotypic markers (such as MARCO or Tim4) have been identified that could be used to phenotypically separate the two types of macrophages (Gibbings et al. 2015; Beattie et al. 2016). These findings highlight that the environment largely dictates the transcriptional programming of macrophages. However, by using the same model of macrophages depletion coupled with genotoxic irradiation, it was shown that the monocyte-derived microglia possess more than 2000 genes differentially expressed as compared to embryonic microglia (Bruttger et al. 2015).

4.3 Are moMϕ and trMϕ Functionally Interchangeable?

We will briefly review the functional contribution of trMϕ versus moMϕ during local inflammatory events, with a focus in the brain, gut, lung, and liver (Table 2.1). For the role of resident versus incoming macrophages in heart diseases (including stoke/ischemic damage) and chronic inflammatory conditions such as rheumatoid arthritis and cancer, we refer the reader to recent excellent reviews (Udalova et al. 2016; Lahmar et al. 2016; Mirò-Mur et al. 2016).

Table 2.1 The main functions of trMϕ versus moMϕ in some tissues of the body
Full size table

Recent fate-mapping studies have established that microglia are of embryonic origin (Ginhoux et al. 2010), persist in the brain during adulthood and in the healthy organism, and are maintained independently on BM-derived monocytes by a limited self-renewal capacity (Ginhoux et al. 2013). However, moMϕ are massively recruited in the brain during an inflammatory event, although they do not contribute to replenishing microglia once homeostasis is restored (Ajami et al. 2011). Recently, the advantages and disadvantages of the various microglial ablation models and the origin of the “new” repopulating microglia have been discussed and reviewed (Waisman et al. 2015). At present, it is difficult to discriminate resident microglia from infiltrating myeloid cells using currently known markers and current tools (Greter et al. 2015). However, it is evident that both microglia and moMϕ play an important role in brain pathologies, as observed in experimental autoimmune encephalomyelitis (EAE), the mouse model of multiple sclerosis (Shemer and Jung 2015; Wlodarczyk et al. 2015). While infiltrating monocytes are harmful and critical in the effector phase of the disease, are highly phagocytic and inflammatory, are associated with nodes of Ranvier, and initiate demyelination, on the other side microglia remain rather inert during the early stage of EAE development, demonstrating globally suppressed cellular metabolism, but it is able to clear debris and its activation precedes the massive monocyte infiltration (Yamasaki et al. 2014). Indeed, microglia secrete numerous chemokines believed to play a role in EAE induction as responsible of the massive monocyte recruitment (Jiang et al. 2014). Therefore, an inhibition of microglia could slow down disease progression (Shemer and Jung 2015).

In the intestine, the situation is opposite, as this tissue is practically devoid of embryonic macrophages. The gut contains the largest pool of functionally specialized macrophages in the body (Gordon et al. 2014), which are essential for the tight crosstalk with microbiota, ensuring a symbiotic relationship and tolerogenic environment. Continual exposure to environmental challenge warrants constant replenishment by blood monocytes. Experiments with fate-mapping and parabiotic mouse models have demonstrated that embryonic precursors populate the intestinal mucosa during neonatal period, but they do not persist in the intestine of adult mice, and they are constantly replaced by circulating monocytes (Ly6Chigh in mice and CD14++ in humans), which differentiate in situ into mature anti-inflammatory macrophages (Bain et al. 2014), favoring the constant need for epithelial renewal (and tissue remodeling). For this reason, in the gut, the distinction is not between trMϕ and moMϕ but rather between moMϕ (which are the resident tissue macrophages) and newly recruited monocytes. Under steady-state conditions, monocytes recruited from the blood differentiate locally into anti-inflammatory moMϕ. They are positioned immediately beneath the epithelial barrier, where they contribute to its integrity. These macrophages are able to survey the tissue sensing and sampling the luminal content by extending processes between epithelial cells, and they produce IL-10 that facilitates the expansion of regulatory T cells. In the mouse, Ly6Chigh monocytes are recruited during inflammation and mount an inflammatory reaction, while the resident moMϕ retain their anti-inflammatory signature. All these events are described in detail elsewhere (Gordon et al. 2014).

The role of macrophages is also essential in the lung, constantly exposed to airborne irritants and microbes. About 90% of the pulmonary macrophage population is represented by alveolar macrophages, located in the alveolar spaces. Alveolar macrophages originate from fetal liver monocytes (Thomas et al. 1976), and in steady state, they are sustained by self-renewal through local proliferation (Tarling et al. 1987). In inflammatory conditions, the repopulation of alveolar macrophages is context specific. In fact, it has been observed that during lethal irradiation, they are replenished by BM monocytes (Duan et al. 2012), while upon inoculation with influenza virus, they are replenished by self-renewal proliferation (Hashimoto et al. 2013) and, upon LPS stimulation, by both incoming monocytes and self-renewal (Upham et al. 1995). Alveolar macrophages seem to have an immunosuppressive function, as they can suppress antigen-induced cell proliferation (Holt et al. 1993) and downregulate antigen presentation by lung DCs (Careau and Bissonnette 2004). Also, alveolar macrophages seem to be protective against airway hyper-responsiveness (Guilliams et al. 2013), and, although exhibiting microbicidal and tumoricidal activities, they are less responsive than macrophages resident in other lung compartments (Hoidal et al. 1981). In the case of inflammatory diseases, such as COPD, monocytes are recruited in the lung, but their contribution to the alveolar macrophage pool remains to be determined (Vlahos and Bozinovski 2014). Using hyperreactivity mouse models with house dust mite and OVA, it has been demonstrated that alveolar macrophages dampen, whereas circulating monocytes promote, early events in allergic lung inflammation (Zaslona et al. 2014).

The liver trMϕ, Kupffer cells, represent the hematopoietic cell population among non-parenchymal cells within the liver. They arise from YS during fetal development (Schulz et al. 2012) and self-renew their population number at steady state throughout adult life with minimal contribution of blood monocytes (Hashimoto et al. 2013). Kupffer cells mainly support the tolerogenic milieu within the liver (Thomson and Knolle 2010), but their presence ensures the protection of the liver during infections (Lee et al. 2010). Recently, it has been observed that Kupffer cell death is a key signal orchestrating type 1 microbicidal inflammation and type 2-mediated liver repair upon infection (Blériot et al. 2015). Indeed, infection by Listeria monocytogenes induces the early necroptotic death of Kupffer cells, which is followed by monocyte recruitment and an antibacterial type 1 inflammatory response. Kupffer cell death also triggers a type 2 response that involves the hepatocyte-derived alarmin IL-33 and IL-4. This leads to the alternative activation of the moMϕ recruited to the liver, which replace ablated Kupffer cells and restore liver homeostasis (Blériot et al. 2015). Both Kupffer cells and moMϕ are involved in liver fibrosis, a common endpoint of many chronic liver diseases such as viral hepatitis, primary biliary cirrhosis, alcoholic and NASH, or autoimmune liver disorder (Eckert et al. 2015). Generally, Kupffer cells are involved in the initiation and moMϕ in the progression phase of the fibrosis through the production of inflammatory cytokines. During disease progression, Ly6Chi cells seem to develop into Ly6Clo restorative macrophages, able to express MMPs and phagocytosis-related genes. These cells, if the harmful agent is eliminated, lead to resolution and can restore normal tissue architecture (Eckert et al. 2015). Otherwise, with the persistence of the initiating agent, they are responsible of anomalous repairing activity, thereby inducing fibrosis.

5 Conclusions

The current knowledge on the origin and role of trMϕ suggests that these cells, either coming from YS/FL precursors before birth or from blood monocytes in adulthood, have a central role in defining and maintaining tissue architecture, function, and homeostasis. The specific role of these cells obviously changes from tissue to tissue, as it is shaped (by the tissue microenvironment) to support the specific tissue requirements. It appears that the origin of these cells makes little difference in the eventual role they have within a tissue, as this is dictated by the tissue itself. Likewise, it is in most cases difficult to distinguish phenotypically, within the trMϕ, between moMϕ and YS-derived cells.

In inflammatory/disease conditions, tissue macrophages mostly act as alarmins, which do not directly exert a potent reaction against the dangerous event but that recall the specialized effector cells, the monocytes, from the blood to the affected tissue. The fate of these inflammatory monocytes, once entering the tissue to eliminate the danger, is not clear. While most of them probably die during the inflammatory reaction, it is possible that some of them survive and persist in the tissue, taking part to the phase of resolution of inflammation and tissue reconstruction/remodeling. Alternatively, or in parallel, it is possible that moMϕ, which enter the inflamed tissue from blood in successive waves, may become highly inflammatory effector cells in the initial phases of inflammation, and “healing” cells in the final phases, being differently polarized by the different tissue microenvironmental conditions. Eventually, it is possible that some of these moMϕ become part of the tissue-resident macrophage pool and develop the capacity of self-renewal. If these macrophages, should they really exist, preserve the “memory” of the past experience, or how this memory influences their response to subsequent dangerous events, i.e., how memory can influence macrophage polarization, is one of the most exciting questions in macrophage biology.