The Biology of Macrophages – An Online Review
In 2008, Siamon Gordon wrote a historical essay on the life of Elie Metchnikoff, who received his Nobel prize a century earlier. The father of natural immunity, as Siamon described him, is credited with the discovery that antimicrobial defence requires the specific recruitment of specialized cells, the phagocytes, which are able to kill and eat potential pathogens. The larger of these specialized cells in mammals, the macrophages (or big eaters) are the subject of this review.
My interest in macrophages started during my PhD, which included a year spent at the Max Planck Institute for Immunobiology in Freiburg in the late 1970s. This Institute was directed by its founder, Professor Otto Westphal, a pioneer in understanding the biochemistry of bacterial lipopolysaccharides (LPS), also known as endotoxins, which initiate much of the pathology of septicaemia. We now know that LPS acts on macrophages to initiate a cascade of inflammatory processes that are essential for innate immunity. I was subsequently fortunate to spend a critical time in the earli 1980s with Siamon Gordon at the Sir William Dunn School of Pathology in Oxford, another organization with an important history at the centre of the development of antibiotics by Florey, Chain and colleagues. I have worked on the biology of macrophages ever since, and in fact, after some 40 years, I am still working on monoclonal antibodies against the CSF1 receptor and other surface markers of macrophages, a project that I undertook in Siamon’s laboratory.
This review is purely a reflection of my personal views and is a synthesis of multiple reviews of macrophage biology that I have written over the past 30 years. With apologies to many colleagues in the field no direct references are provided. At the end of the review, I provide a list of relevant reviews that I have written previously on specific topics. This list will be updated periodically.
Macrophages as recognized by Metchnikoff can be found in all tissues, and as circulating cells called blood monocytes. Blood monocytes make up around 20% of the peripheral blood mononuclear cell (PBMC) fraction in mice and humans, considerably less numerous than the other major phagocyte population, the polymorphonuclear cells, or neutrophilic granulocytes. Granulocytes and monocytes/macrophages are collectively referred to as myeloid, contrasting to lymphoid. During inflammation, monocytes are recruited from the blood into tissues in response to a very wide range of stimuli, with slower and rather distinct kinetics from the short-lived granulocytes. The older literature considered that macrophages were derived from mesenchymal cells, and shared biology with endothelial cells, leading to the widespread use of the term reticuloendothelial system (RES).
It was only in the 1960s that it became clear that inflammatory and tissue macrophages are haematopoietic in origin and derived ultimately from progenitors shared with other blood cells. The concept of the mononuclear phagocyte system (MPS) was developed by van Furth, Cohn and colleagues in the late 1960s and early 1970s. The MPS was defined as a family of cells that includes committed precursors in the bone marrow, circulating blood monocytes and tissue macrophages in every organ in the body.
During embryonic development in birds and mammals, professional phagocytic cells appear first in the yolk sac in a process that is distinct from classical haematopoiesis. These early macrophages migrate within (and perhaps ahead of) the developing vasculature to eventually populate the entire embryo where they actively engage in clearance of dying cells. The yolk sac also gives rise to a committed erythromyeloid progenitor cell population that migrates to the foetal liver and gives rise to monocytes and definitive erythrocytes. The earliest detectable haematopoietic cells in the foetal liver are macrophages that form the centre of islands of proliferating red cells. At this time, definitive red blood cells are formed and the macrophages engulf the nuclei expelled during red cell maturation. At around the same time, definitive haematopoietic stem cells that will eventually populate the classical bone marrow niche develop in a specialised location called the aorta-gonad-mesonephros (AGM).
In recent literature the MPS concept has been portrayed as a dogma in which tissue macrophages are constantly replaced by blood monocytes in the steady state but this is a misrepresentation. The current alternative view based on studies in inbred mice is that most tissue macrophage populations maintained in the adult are seeded from the yolk sac and/or by monocytes produced by the foetal liver and thereafter maintained by self-renewal. In fact, the MPS concept proposed by van Furth and Cohn never excluded foetal origins and self-renewal as complementary mechanisms underlying homeostasis of resident macrophages. In both mice and humans with mutations in key transcriptional regulators, resident macrophages can be maintained in the complete absence of blood monocytes. However, the current literature on monocyte-macrophage homeostasis is based upon the interpretation of complex genetic models using a single inbred mouse strain in highly-controlled pathogen-free animal facility environments. My personal view is that many of the assumptions underlying the interpretation of so-called lineage trace models are flawed and invalid and the contribution of foetal-derived cells to resident macrophgae populations is over-estimated. It is certainly the case that blood monocytes and tissue macrophages are closely-related cell types, and monocytes can and do leave the circulation to replace resident tissue macrophages in the normal steady state. The extent of this contribution in outbred animals, in different tissues and species and in response to challenge is the subject of ongoing studies.
Identification of macrophages in tissues
Until the early 1980s, macrophages resident within tissues, generally referred to as histiocytes, were recognized largely based upon morphology and location, notably the presence within them of the evidence of previous bouts of phagocytosis. With the advent of monoclonal antibody technologies, numerous anti-macrophage antibodies were produced that bound selectively to surface antigens on macrophages of multiple species. When those antibodies were used in combination, it became clear that they are not perfectly correlated with each other and can be used to define subpopulations (discussed below).
Despite the levels of heterogeneity there are a small number of surface proteins that are found on the large majority of macrophages and can be detected using monoclonal antibodies. A particularly useful monoclonal antibody for the mouse generated in Siamon Gordon’s laboratory was called F4/80 (the 80th hybridoma in the 4th attempted fusion). I started my career in macrophage biology localising F4/80 antigen in sections of every tissue I could examine in the mouse. The image database elsewhere on this website contains some colour images from the early characterization of F4/80 localisation by immunohistochemistry on fixed mouse tissue sections. F4/80 binds to a protein encoded by the ADGRE1 gene, a G protein coupled receptor with a very large extracellular domain likely involved in adhesion to extracellular matrix. Both the expression pattern, and the protein sequence of ADGRE1 are highly divergent amongst mammalian species suggesting strong evolutionary selection by pathogens.
More recently, immunohistochemistry using monoclonal antibodies has been supplemented by the generation of transgenic reporters expressed solely in cells of the mononuclear phagocyte lineage. These are commonly multi-copy transgenes where a defined promoter element derived from a macrophage-specific gene is place upstream of a fluorescent protein reporter gene such as green fluorescent protein (GFP) or mApple so that all macrophages are fluorescent. Increasingly, new reporters are being generated by insert the reporter gene into the locus of interest so that it more actually reflects the normal gene expression. Development of fluorescent reporters has enabled live imaging of tissues to monitor macrophage motility and cell-cell interactions. The image database elsewhere on this website also contains many of the fluorescent images we have generated in mouse and rat transgenic models. The top level of the site shows an image of a cyan-fluorescent protein reporter in a mouse embryo, where all the macrophages are bright blue.
One of the most striking things that emerged from immunohistochemical and reporter transgene studies was the very large numbers of macrophages and the extensive ramification of their processes that extend throughout the tissues. Macrophages may well be the most numerous single nucleated cell type in the body; 10-15% of total cells in almost all organs and even more in connective tissues, membranes and capsules. Macrophages have a particularly intimate relationship with epithelial and endothelial cells. In simple epithelia, and throughout the capillary and lymphatic circulation, tissue macrophages spread along the ablumenal side of basement membranes; in stratified and pseudostratified epithelia such as skin, trachea and cervix and in secretory ducts in exocrine organs, they are integrated within the epithelium. Sinusoidal macrophages, such as those of liver, spleen and some endocrine organs have direct contact with the blood. But the separation by endothelium does not prevent pericapillary macrophages from extending processes into the lumen and sampling the blood contents. The ability of macrophages to extend processes across epithelia and into lymphatic vessels has also been recognized.
We commonly think of macrophages as cells of the immune system and indeed they are essential to both innate immunity and to the development of specific acquired immunity. As innate immune effectors, macrophages engulf microorganisms, kill and digest them within phagocytic vesicles. They recognise microorganisms through an array of cell surface receptors that bind to generic structures shared by microorganisms. The recognition of potential pathogens may be augmented by coating with antibody or complement, a process called opsonisation. The microbicidal activity of macrophages is augmented in response to products of activated T lymphocytes so they also act as effectors of acquired cell-mediated immunity.
In addition to their role in innate immunity macrophages have been attributed numerous functions in embryonic and postnatal development, homeostasis and wound repair. As an example, the macrophages of the epidermis, known as Langerhans cells, form the centre of so-called epidermal proliferative units and contribute to the control of proliferation and differentiation of keratinocytes. As noted above, macrophages form the centre of haematopoietic islands in foetal liver and also in the adult bone marrow. Those lining the surfaces of bone (referred to as osteomacs) are able to control osteoblast differentiation and calcification, and those in the embryo appear to control development and nephron endowment in the kidney.
Surprisingly, given the abundance of macrophages in the embryo and their active engagement in the removal of apoptotic cells and tissue remodelling, the complete absence of embryonic macrophages in various mutant mouse and rat lines does not appear to have a major impact on prenatal development. The central importance of macrophages in postnatal development is highlighted by the many systems affected by macrophage depletion in animals that lack macrophages through mutations in the major growth factor that controls their development (macrophage colony-stimulating factor, CSF1) or the lineage-restricted receptor (CSF1R). Mutations in the CSF1R gene have also been reported in humans. The impacts of the absence of macrophages include severe somatic growth retardation and multi-organ developmental delay, osteopetrosis and male and female infertility. Macrophage deficient mice and rats are indistinguishable from their littermates at birth. Furthermore, the effects of the loss of macrophages can be prevented by the postnatal transfer of wild-type bone marrow cells indicating that any subtle defect that does occur is reversible. In the absence of professional phagocytes, it seems that amateurs (e.g fibroblasts and astrocytes) can acquire phagocytic activity and adapt to take their place.
In the postnatal period, the population of macrophages in each organ expands rapidly. Aside from the separation from the mother, the neonate must adapt to breathing air and, perhaps most importantly, to the population of the gut with the so-called microbiota. As organs mature, the resident MPS cells become adapted to perform particular functions in different organs; so that brain macrophages (microglia) are very different from alveolar macrophages of the lung, Kupffer cells of the liver, or the largest tissue macrophage population, those lining the wall of the gut. The macrophages of the liver and spleen, for example, are adapted to internalise senescent red cells and recycle the haem iron back to the bone marrow. Specialised multinucleated cells called osteoclasts are required for physiological turnover of bone, and as mentioned above bone surfaces are lined with a separate macrophage population called osteomacs.
Foetal liver and adult bone marrow each contain pluripotent hematopoietic progenitor cells that can replenish cells of the MPS and other blood cells following transplantation into an irradiated host. In the process of lineage commitment, MPS cells progress through a series of defined morphologically-distinct stages; a common myeloid progenitor shared with granulocytes giving rise to monoblasts, promonocytes and then monocytes which migrate into tissues. The production of mononuclear phagocytes from progenitor cells is directed by combinations of colony-stimulating factors (CSF), to some extent lineage restricted and hierarchical in their actions. CSF are so-named because they promote the formation of colonies of progeny when added to bone marrow cultures. They include interleukin 3 (IL3), stem cell factor (KITL, SCF), macrophage colony-stimulating factor (CSF1, M-CSF), interleukin 34, granulocyte macrophage colony-stimulating factor (CSF2, GM-CSF) and fms-like tyrosine kinase 3 ligand (FLT3L). There is an ongoing discussion in the literature as to whether these agonists are instructive (driving differentiation) or selective (promoting the proliferation of cells that have become lineage committed through intrinsic processes). In my view, the latter model is more consistent with available data and understanding of mechanisms of transcriptional regulation. Whatever the mechanism, several of these colony-stimulating factors can expand circulating myeloid cells (granulocytes and macrophages) when they are administered in vivo and are used clinically to treat myelodeficiency. Many studies of macrophage biology in cell culture in vitro are based upon the use of recombinant CSF1 or CSF1-containing cell culture supernatants to promote proliferation and differentiation of bone marrow progenitors.
Natural mutations in the CSF1 gene in mice and rats provided the evidence that this factor is absolutely required for the production of the large majority of tissue macrophages in both species, at the same highlighting the importance of tissue macrophages in many aspects of normal development. More recently, IL34 was identified as a second macrophage growth factor that shares a receptor with CSF1, the CSF1R. The two ligands are found in all vertebrates and the identification of IL34 explained why CSF1R mutations have a more severe impact on tissue macrophage populations and postnatal development than mutations in CSF1 alone.
Monocytes are the circulating arm of the mononuclear phagocyte system, distinguished from the more abundant polymorphonuclear leukocytes or neutrophils. They are continuously produced by the bone marrow and enter the circulation in response to specific chemoattractants, notably a peptide that resident macrophages themselves can produce called monocyte chemotactic factor, encoded by the CCL2 gene. The production and release of monocytes into the blood is regulated and greatly enhanced in response to increased demand, for example in response to inflammation, infection or injury. Monocytes can also enter tissues continuously to replace tissue macrophages, especially in organs where the turnover is relatively rapid. Monocyte extravasation is also controlled by chemoattractants including CCL2. Mutation of the receptor for CCL2, CCR2, is commonly used to study the relative importance of monocyte recruitment in various disease models in mice.
Blood monocytes leave the circulation probabilistically, with a half-life of around 4-7 days in most species. However the subset of monocytes that remains in the circulation for longer (probably by chance, there is little evidence that this is predetermined) differentiates in response to CSF1/CSF1R signals to generate a distinct population of more long-lived monocytes that have been dubbed non-classical. These cells have been called a “subset” but an intermediate population can also be identified and really the entire population of monocytes is just a differentiation series. As they differentiate, monocytes change their gene expression profiles, notably down-regulating CCR2 and up-regulating a distinct chemotactic factor receptor CX3CR1. The mature non-classical monocytes may be considered the resident macrophages of the blood, and there is evidence from live imaging that they patrol the endothelial surfaces of the vasculature. However, the more mature monocytes may also be recruited into inflamed tissues so the distinction between inflammatory and resident cells is not clear-cut.
The absolute number and relative proportions of monocyte “subsets” differs greatly amongst mammalian species. Small ruminants (sheep and goats) have few detectable monocytes. In mice, around 50% of monocytes are non-classical, in humans this percentage is much lower (around 10%), in rats much higher (around 90%) and in pigs there is no clear non-classical population. These differences presumably reflect the balance between relative rates of release from the marrow and extravasation in each species and whether individual monocytes remain in the circulation for sufficent time to differentiate in response to CSF1.
Macrophage heterogeneity within tissues
Although macrophages in tissues have many features in common, including extensive lysosomes, stellate morphology and location relative to epithelia and vasculature, they are nevertheless extremely heterogeneous in terms of function and surface marker expression as discussed above. Our knowledge of this plasticity is most extensive for the mouse.
Molecules expressed on the cell surface are of particular functional interest because they determine the ability of MPS cells to interact with pathogens, and with other cell types, to generate an appropriate innate and acquired immune response. With the exception of CSF1R, I am not aware of any surface marker that is expressed specifically and ubiquitously on all MPS cells. The abundant cytoplasmic calcium binding protein IBA1 (encoded by the AIF1 gene) appears to be present in the large majority of tissue macrophages and can be a useful marker but is also present in several unrelated cell types including spermatocytes.
Surface markers of mouse MPS cells can be divided into two categories; those that are heterogeneously-expressed on individual cells within any one location and those that are enriched on defined functional subpopulations of cells in specific organs or locations within organs. Considering the latter category, within the mouse spleen the distributions of F4/80, mutliple integrins, scavenger receptors and C-type lectins differ between the T cell, B cell and germinal centre regions of the white pulp, the red pulp and the marginal zone. If one considers the full set of surface markers that display heterogeneous expression in macrophages, which also includes the Fc receptors, CD36, CD14, SIRPalpha, TLRs, integrins and other cell adhesion molecules, EGF-TM7 proteins (of which F4/80/ADGRE1 is one), other Ig superfamily receptors (Siglecs) and multiple C-type lectins the set of combinations and subpopulations is essentially infinite. This is especially true if one identifies subsets of cells on a FACS profile that express high or low levels of a marker. At least some of this heterogeneity arises because of the stochastic nature of transcriptional regulation, so that there is a genuinely random assortment of surface markers. One could take the view that this presents potential pathogens with a formidable arsenal of potential host defense in which every macrophage is unique, especially as we know that some receptors which have functional polymorphisms, such as TLR4, are expressed mono-allelically in individual cells.
In other cases, a particular combination of surface markers determines potential function (e.g. the antigen uptake, presenting and co-stimulatory molecules such as CD80, CD86 and CD40) but even in this case, different combinations may provide different T cell subsets with distinct signals. Surface markers such as the chemokine receptors and integrins must act in concert to determine recruitment and location in a tissue. There must be a degree of determination that ensures that certain sets of genes are co-expressed at the right time and place, and that in turn determines a cellular function. The appropriate timing is an important issue, because much of the heterogeneity we see in tissue reflects the fact that macrophages within the tissue are in different stages of their cycle of life and death, migration, development and function, responding to an infinite combination of signals. Many apparent subsets are clearly interconvertible and derived from a common progenitor. So, surface marker expression cannot be taken as the sole indication of lineage, function or destiny amongst macrophages.
Studies of transcriptional regulation in cells of the mononuclear phagocyte system began in the early 1980s. Such studies were significantly constrained in the first instance by the fact that macrophages recognize and respond to bacterial DNA. Classical promoter transfection studies were difficult to achieve, because primary macrophages undergo apoptosis in response to DNA introduced into the cytoplasm. Only recently, the receptor responsible for the response, AIM2, was identified by a number of groups including mine. In the mouse at least, macrophages are also activated by non-mammalian DNA through the recognition of unmethylated CG dinucleotides by TLR9.
Understanding of the process of macrophage lineage commitment gained major impetus with the cloning of the transcription factor, PU.1 (aka Spi-1, or Sfpi1) by Richard Maki and colleagues. A member of the Ets transcription factor family, and named for its binding to purine-rich sequence motifs, PU.1 was found to be restricted in its expression to macrophages and B lymphocytes. The level of both PU.1 mRNA and nuclear PU.1 protein are much higher in macrophages than B cells, and later studies confirmed high expression of PU.1 was essential to macrophage lineage commitment and macrophage-specific gene expression.
PU.1 has two apparent functions in macrophage transcriptional regulation. Firstly, a specific subset of promoters that is active in macrophages, exemplified by that of the Csf1r gene, lacks either a TATA box or a CpG island, and instead contains repeats of a purine-rich motif that binds PU.1. Multimerised PU.1 sites alone can generate transcription initiation in macrophages; but PU.1 must act in concert with another member of the Ets transcription factor family. To initiate transcription on such promoters, and to specify the transcription start site, PU.1 cooperates with Ewing sarcoma protein. The other major function of PU.1 appears to be as a pioneer factor to generate open chromatin around enhancers that are either constitutively, or potentially, activated in cells of the macrophage lineage and which can subsequently be occupied by other transcription factors.
PU.1 is certainly not the only factor one needs to consider to understand macrophage differentiation. The phenotype of the PU.1 knockout in mice, which certainly greatly reduces the numbers of all mature cells of the mononuclear phagocyte lineage, macrophages, DC and osteoclasts as well as other myeloid cells actually depends upon mouse genetic background. The PU.1 locus itself has a purine-rich TATA-less promoter, and at some point in myeloid lineage commitment, it must itself be transcriptionally activated. The numerous factors that interact with PU.1 include members of the RUNX, ETS, CEBP, MITF, MYB, KLF, AP1/ATF, IRF and nuclear hormone receptor families. In fact, around 2/3rds of all transcription factors encoded in the mammalian genome can be expressed in macrophages in some state of differentiation or activation.
Macrophage recruitment and chemotaxis
As discussed above, monocytes are recruited into tissues in response to a very wide range of different stimuli. Where a pathogen is involved, they are commonly preceded by neutrophils, which release a range of toxic agents designed to kill extracellular pathogens. The macrophage then has the task of clearing both the dead pathogens and the dead neutrophils. To enter a tissue, the monocyte in peripheral blood must adhere to the vessel wall, cross the endothelial cell barrier, and then migrate towards the stimulus; a process known as chemotaxis. The process of recruitment of neutrophils and macrophages involves the resident macrophages which act as sentinels. They responds to local stimuli by producing cytokines that make the endothelial cells more sticky (through the increased expression of cell adhesion molecules such as P-selectin) and so-called chemokines (such as CCL2), that promote the directed migration of inflammatory cells. Monocytes may also migrate towards increasing concentrations of molecules that produced by microorganisms themselves, by damaged tissues, or by the activation of the complement or clotting cascades which release bioactive peptides such as C5a. One example of a microbial chemoattractant is N-formyl-methionyl peptides; which are unique to bacteria because this is the initiating amino acid at the N terminus of all bacterial proteins.
Chemokines are a family of small peptides, subdivided into classes based upon the core cysteine motifs that form disulphide bonds to fold the molecule; CC chemokines has two adjacent cysteines whilst in CXC chemokines, there is an intervening amino acid. A single chemokine, CX3CL1 has three intervening amino acids. Chemokine receptors are classified accordingly, CCR, CXCR and CX3CR families. The large majority of receptor that control chemotaxis fall into the class of G protein coupled receptors (GPCRs). Agonist occupation of GPCRs stimulates a change in conformation of the receptor, which couples the receptor to a so-called G-protein and promotes the exchange of GDP for GTP on the α-subunit. The GTP-bound α-subunit dissociates from the βγ-subunit; the free subunits then regulate effector enzymes positively or negatively, ultimately leading to a biological response, in this case, directed cell migration. Monocytes and macrophages are remarkably adapted to respond to a wide range of different signals coupled to GPCRs, with very high levels of expression of many of the downstream signaling and feedback control mechanisms. Expression of specific chemokine receptors (e.g. CCR2 and CX3CR1) on different populations of monocytes provides a mechanism for their differential recruitment in response to different signals.
Endocytosis and phagocytosis
All mammalian cells are able to take up macromolecules and particles from the extracellular environment through fluid-phase or receptor-mediated uptake processes. Monocytes and macrophages are “professional” phagocytes. They are exceptional compared to other cells in the scale of membrane movement devoted to endocytosis, the diversity of receptors used for receptor-mediated uptake, the specific adaptation to internalize larger particles rapidly and the focus on degradation of the internalized materials. Phagocytosis, generally defined as the uptake of particles around 1µ or greater in diameter, is a function that is conserved from the simplest eukaryotic organisms such a slime moulds which feed on bacteria. Quite apart from the functions in immunity in adult life, phagocytosis is used during mammalian embryonic development to eliminate cells undergoing programmed cell death and key molecules involved have been identified in model organisms such as C.elegans and D.melanogaster based upon the persistence of cellular corpses.
The precise process of phagocytosis depends upon the particle being internalized, its size and whether it controls its own fate. In broad terms, the uptake process usually requires receptor-mediated contact around the full circumference of the particle, trans-membrane signaling to promote membrane extension and polymerization of the underlying actin cytoskeleton, and subsequent maturation of the internalized vacuole (the phagosome) to acidify (through the vacuolar ATPase which drives transport of H+ into the lumen of the internalised vacuole) and finally fuse with lysosomes to initiate particle degradation through the activity of lysosomal hydrolases.
Phagocytosis is a front-line defense against pathogen attack, so almost by definition, a pathogen is an infectious agent that avoids being killed by phagocytosis. Some microorganisms produce anti-phagocytic capsules, other produce toxins that are specifically toxic to macrophages. Many pathogens (e.g mycobacteria) exploit macrophages as their preferred site of replication, promoting their own internalization and either preventing phagosome-lysosome fusion, permitting such fusion and resisting destruction, or escaping into the cytoplasm through rupture of the uptake vesicle.
Phagocytosis is a process that requires a mechanism for self-nonself discrimination (or in the case of recognition of dead cells and debris, a mechanism for distinguishing that material has passed its use-by date). Macrophages possess numerous receptors that allow direct recognition of particles based upon novel sugars, lipids, protein sequences and concentrations of charge that are unique to pathogens (so-called pathogen-associated molecular pathogens). Particles may also be recognized indirectly if they are coated with opsonins such as specific antibodies or complement components.
The process of phagocytosis and degradation requires the concerted actions of hundreds of gene products. With the advent of gene expression profiling and access to very large datasets comparing cell types and tissues, it became possible to identify the genes that make a phagocyte distinctive. They include all of the digestive enzymes found in lysosomes (a mini-stomach within the cell), the components of the proton pump that acidifies the lysosome, and the apparatus needed to make and move the lysosome and internalize particles.
The innate immune system and the acquired immune system mediated by T and B lymphocytes are linked by the fact that T lymphocytes do not respond directly to soluble or particulate antigens. The fundamental finding, now generally-accepted, is that antigens derived from extracellular sources must be taken up, processed, and presented to T lymphocytes, bound into the cleft formed by class II MHC which is bound by CD4 on helper T cells. The process of uptake, processing and presentation via a specialized endosomal compartment is now well understood. The cells responsible for this activity are referred to as antigen presenting cells (APC). The early literature on antigen presentation considered that macrophages were the main cells responsible. Most tissue macrophages express class II MHC on their surface (especially in species other than mice), it is further inducible by T cell products (notably interferon gamma) and is expressed at highest levels on the macrophages recruited in response to an immune stimulus. Recognition of the antigen-MHC II complex by the T cell receptor is not sufficient to trigger T cell activation; this requires a second co-stimulatory signal from the APC in the form of specific cytokines and coreceptors, each of which can also be expressed inducibly in phagocytic cells.
In the early 1970s Steinman and Cohn described a distinct cell type they referred to as the dendritic cell (DC). The DC as originally defined was a non-phagocytic cell that was particularly active at recognising cells from unrelated individuals of the same species (allo-antigens) in the allogeneic mixed lymphocyte reaction. The original DCs are most likely derived from the network of interdigitating cells within T cell areas of lymphoid tissues. Purification of these cells from spleen or lymph node led to substantial enrichment of APC activity; and the contrast was made with the phagocytic “macrophage fraction” which apparently lacked APC activity in typical in vitro assays. However, the purification of DCs simultaneously removes the suppressive class II MHC-negative macrophages of the red pulp, which exert a dominant effect in the unphysiological context of an in vitro assay in a round bottom tissue culture well.
DCs were originally identified in lymphoid tissues, but were subsequently isolated from peripheral organs such as the gut lamina propria again clearly distinguished from active phagocytes/macrophages isolated from the same tissue. Subsequently, it was found that monocytes could differentiate into “DCs” and that APC activity could be elicited in active phagocytes grown from bone marrow or peripheral blood monocytes in the growth factor, GM-CSF (CSF2).
The current view of APC in the mouse recognises three separate populations, monocyte-derived APC and two subsets of classical DC, cDC1 and cDC2. The cDC1 cells isolated from spleen and lymph nodes most clearly resemble the original cells described by Steinman and Cohn. Each of the candidate APC apparently arises from a shared monocyte-DC progenitor in the bone marrow (at least in mice). It is not clear to me that cDC2 in particular can be defined as a separate entity based on gene expression; in terms of the transcriptional profiles they are no more distinctive than liver macrophages compared to microglia. They have no distinct cell surface markers that distinguish them from macrophages and their gene expression profiles overlaps almost entirely. The alternative view is that DC can be defined based upon derivation from a distinct committed DC progenitor (CDP) and/or dependence upon FLT3 ligand. Like alternative models of the ontogeny of resident macrophages, the ontogeny model for DC relies on lineage trace experiments in a single inbred mouse strain that depend upon multiple assumptions.
In most non-lymphoid tissues cells identified as DC based upon surface markers are CSF1/CSF1R dependent and depleted by anti-CSF1R treatment. The only clear functional and differentiation dichotomy is between phagocytic (macrophages) and non (or much less)-phagocytic APC (cDC1). A substantial proportion of the actively phagocytic, resident tissue macrophages associated with epithelia/mucosal surfaces express class II MHC, and are able to stimulate naïve T cells, sometimes more directed towards tolerance/suppression. In my view, antigen presentation is a regulated process that is not restricted to any particular cell type.
Recruited monocytes and macrophages alter their function to deal with the nature of the challenge that promoted their recruitment. This has been referred to as macrophage activation, and it can be mimicked to some extent by exposing monocytes and macrophages to specific stimuli in vitro. Classical macrophage activation refers specifically to the broad class of pro-inflammatory activation observed in response to challenge by microorganisms. Classically-activated macrophages are strongly positive for class II-MHC,and adapted to kill microorganisms and tumour cells and present antigen to T lymphocytes. The classical macrophage activating factor, which is produced by stimulated Th1 lymphocytes and NK cells, is interferon-gamma. In this respect, classically activated macrophages are key downstream effectors of T cell-mediated immunity against microorganisms.
IL4 is a major product of stimulated Th2 lymphocytes activated during parasite infections. Macrophages respond to IL4 with a distinct set of gene expression changes to produce secretory products that drive a coordinated cellular response that is more effective against this class of pathogens. Recent evidence indicates that in parasite infection, IL4 can also promote proliferation of resident tissue macrophages above the homeostatic levels imposed by CSF1 availability.
The T cell products are, of course, only part of the story of macrophage activation. Macrophages respond directly to pathogen-associated molecular patterns (PAMPs), broad classes of sugars, lipid, peptides and nucleic acids that are shared by many micoorganisms and distinguish them from mammalian cells. Macrophages recognize PAMPs through diverse plasma membrane and cytoplasmic receptors such as the Toll-like receptors, C-type lectins and intracellular receptors of the NLR family. The pattern recognition receptors themselves are under strong evolutionary selection and are commonly associated with variation in disease susceptibility between individuals. Classical macrophage activation, involving a synergistic interaction between interferon-gamma and a pathogen molecule such as lipopolysaccharide (LPS) is just one of the numerous interactions that occurs between distinct stimuli.
Based upon the Th1/Th2 dichotomy, many authors have adopted a binary view of macrophage phenotypes and defined M1 and M2 markers. However, this simplification is not supported by large scale gene expression profiling where proposed markers are not correlated with other and where many so-called M2 markers are expressed by resident tissue macrophages. Furthermore, the sets of co-expressed genes in inflammatory macrophages are highly divergent amongst mammalian species. For example, the ability of classically-activated macrophages to make nitric oxide as an anti-microbial effector is unique to rodents.
A more sustainable view of macrophage biology could consider the range of macrophage phenotypes in inflammation as resembling the spectrum of colours on a colour wheel. Macrophages could be classified as “red”, “yellow” and “blue”, but every combination and shade is possible and they can be inter-converted. Indeed during its lifetime in an inflammatory lesion an individual macrophage probably transitions through multiple activation states. This kind of plasticity means that the number of “subsets” than can be defined is infinite: like the subdivision based on cell surface markers and flow cytometry, it is really just a function of the number of markers evaluated. In recent times, the spectrum of macrophage activation states has been dissected using single cell RNA-seq, where the number of subsets that can be defined using standard dimensionality reduction approaches is really just a matter of researcher preference and arbitrary imposition of boundaries.
The activation of macrophages leads to the production of a wide-range of hormone-like molecules called cytokines, that are important for orchestration of an appropriate inflammatory and acquired immune response, but can also initiate much of the pathology of disease. These molecules include tumor necrosis factor, interleukin-1, interleukin-6 and interleukin 10. Together they contribute to initiation of systemic responses including fever, known collectively as the acute-phase response. These responses contrive to initiate sickness behaviour and to make the body less conducive to pathogen replication. In extremus, they are themselves the cause of life-threatening complications seen in septic shock; multiple organ failure and disseminated intravascular coagulation. For this reason, macrophage-derived cytokines have been attractive and effective targets for therapeutic interventions in chronic inflammatory diseases.
Dynamic networks in macrophage activation.
The process of macrophage activation, regardless of the nature of the stimulus, involves a radical change in gene expression, with numerous genes increasing or decreasing their expression with time. To understand how transcriptional regulation is achieved in macrophages, or any other system, it is worthwhile to start to look at the entire system. Looking at one gene at a time can be rather like gazing at the individual brush strokes of an impressionist painting. System-wide views of transcriptional networks have been greatly expedited by the advent of genome-scale technologies and network analysis tools that extract sets of co-regulated transcripts and identify the underlying shared trascriptional regulators. There are a number of features of the response of macrophages to a stimulus such as bacterial lipopolysaccharide (LPS) that are shared by many other biological systems
- There is a sequential cascade of gene regulation. In the case of LPS-stimulated macrophages, the early response genes, which include immediate early transcription factors, a number of classical inflammatory cytokines such as TNF and interferon beta 1, have been shown to be subject to regulation primarily at the level of transcription elongation from poised RNA pol II complexes. Later response genes are regulated by autocrine factors, including TNF and interferon, and inducible transcription factors.
- The numbers and magnitudes of regulation/expression of induced genes are almost precisely balanced by the numbers of repressed genes. In a sense, this is obvious, because the total amount of mRNA per cell does not change radically. Amongst the repressed genes are numerous transcriptional repressors that would otherwise block the response to LPS and genes involved in other pathways.
- Many of the early response genes, both induced and repressed, return to the pre-stimulation level with time; the response is in some measure self-limiting even in the continued presence of the agonist. Amongst the LPS-inducible genes are numerous additional feedback-regulators, which we have referred to as “inflammation suppressor genes”.
In broad terms, we might think of the control of macrophage activation like an automatic vehicle, in which forward progress requires the release of the brake as well as the application of the accelerator, and in which the accelerator links back to the reapplication of the brake.
The signaling pathway from the receptor, TLR4, leading to transcriptional activation, has been described in considerable detail; the major focus has been on the transcription factor complex NF-kappaB, and on the interferon-regulated factors (IRFs) which are induced in part by autocrine interferon-beta. Genome-scale analysis by the FANTOM consortium revealed that some 2/3rds of annotated transcription factors are expressed in primary macrophages. A relatively small number of expressed transcription factors are very highly-connected to others, whereas others occupy peripheral niches within the network with relatively few direct connections. It is important to recognize that because this is a single network, every transcription factor in the network is connected in some way to every other factor via multiple paths. So, the system, in this case macrophage activation, will respond in some way to perturbation of any of the components of the network.
The set of genes expressed by macrophages responding to a microbial challenge evolves rapidly across species. Comparative analysis of the underlying regulatory networks of human and mouse macrophages was expedited by the availability in both species of data that precisely defined transcription start sites, so that we were able to look sequences of the promoters of regulated genes in both species, including the numerous genes that were induced in one species and not the other. The list of transcription factor binding sites over-represented in the promoters, and their relative over-representation, was identical between the two species, suggesting that in both species the promoters sample a common transcriptional milieu. In essence, the network architecture is conserved. However, a pairwise comparison of individual promoters, revealed that there was very little conservation of individual elements between the species, even though we did not specify direct alignment. This is consistent with a model in which gain and loss of individual motifs, including the TATA box, is a significant driver of evolution, few individual motifs/binding sites have indispensable functions.
One of the most striking examples of evolutionary divergence is in the response to glucocorticoids. Glucocorticoids are a major component of feedback regulation of inflammation and accordingly amongst the most widely-applied anti-inflammatory drugs. They act on macrophages by inducing transcription of feedback inhibitors of the response to activators such as LPS and interferon-gamma. In a direct comparison of the responses of human and mouse macrophages to dexamethasone, only around 10% of inducible genes were shared and the gain and loss of regulation was associated with the gain and loss of DNA sequences that bound the glucocorticoid receptor.
A great deal of our current knowledge of macrophage biology is based upon the inbred mouse as a model. Whilst much of the basic cell biology is shared across verterbrate species and more so amongst mammals, there are also quite fundamental differences. Many papers in the immunology literature have definitive statement titles claiming that a particular cell type or process is essential for a particular immune response. The species or strain is seldom included in the title and often not mentioned even in the text. Unfortunately, many such definitive statements are only true in a particular strain or species. In my view, models of monocyte-macrophage ontogeny are true only in the mouse strain in which they were studied (if at all). The most commonly-studied inbred mouse strain, C57BL/6J, has a mutation in the NNT gene that in humans is associated with hereditary glucocorticoid deficiency. In my own group, I have focussed in the rat as an alternative model whilst also recognising the importance of mouse strain variation. Innate immunity is intrinsically subject to strong evolutionary selection by pathogens. If our objective is to treat human disease, we need to recognise that inbred mice may not be great models for mouse macrophage biology, let alone humans.
- Geissmann F, Gordon S, Hume DA, Mowat AM, Randolph GJ. Unravelling mononuclear phagocyte heterogeneity. Nat Rev Immunol. 2010;10(6):453-60. Epub 2010/05/15. doi: 10.1038/nri2784. PubMed PMID: 20467425; PubMed Central PMCID: PMCPMC3032581.
- Gow DJ, Sester DP, Hume DA. CSF-1, IGF-1, and the control of postnatal growth and development. J Leukoc Biol. 2010;88(3):475-81. Epub 2010/06/04. doi: 10.1189/jlb.0310158. PubMed PMID: 20519640.
- Hume DA. Differentiation and heterogeneity in the mononuclear phagocyte system. Mucosal Immunol. 2008;1(6):432-41. Epub 2008/12/17. doi: 10.1038/mi.2008.36. PubMed PMID: 19079210.
- Hume DA. Macrophages as APC and the dendritic cell myth. J Immunol. 2008;181(9):5829-35. Epub 2008/10/23. doi: 10.4049/jimmunol.181.9.5829. PubMed PMID: 18941170.
- Hume DA. Plenary perspective: the complexity of constitutive and inducible gene expression in mononuclear phagocytes. J Leukoc Biol. 2012;92(3):433-44. Epub 2012/07/10. doi: 10.1189/jlb.0312166. PubMed PMID: 22773680; PubMed Central PMCID: PMCPMC3427611.
- Hume DA. The Many Alternative Faces of Macrophage Activation. Front Immunol. 2015;6:370. Epub 2015/08/11. doi: 10.3389/fimmu.2015.00370. PubMed PMID: 26257737; PubMed Central PMCID: PMCPMC4510422.
- Hume DA, Caruso M, Ferrari-Cestari M, Summers KM, Pridans C, Irvine KM. Phenotypic impacts of CSF1R deficiencies in humans and model organisms. J Leukoc Biol. 2020;107(2):205-19. Epub 2019/07/23. doi: 10.1002/JLB.MR0519-143R. PubMed PMID: 31330095.
- Hume DA, Caruso M, Keshvari S, Patkar OL, Sehgal A, Bush SJ, et al. The Mononuclear Phagocyte System of the Rat. J Immunol. 2021;206(10):2251-63. Epub 2021/05/10. doi: 10.4049/jimmunol.2100136. PubMed PMID: 33965905.
- Hume DA, Freeman TC. Transcriptomic analysis of mononuclear phagocyte differentiation and activation. Immunol Rev. 2014;262(1):74-84. Epub 2014/10/17. doi: 10.1111/imr.12211. PubMed PMID: 25319328.
- Hume DA, Irvine KM, Pridans C. The Mononuclear Phagocyte System: The Relationship between Monocytes and Macrophages. Trends Immunol. 2019;40(2):98-112. Epub 2018/12/24. doi: 10.1016/j.it.2018.11.007. PubMed PMID: 30579704.
- Jenkins SJ, Hume DA. Homeostasis in the mononuclear phagocyte system. Trends Immunol. 2014;35(8):358-67. Epub 2014/07/23. doi: 10.1016/j.it.2014.06.006. PubMed PMID: 25047416.
- Joshi A, Pooley C, Freeman TC, Lennartsson A, Babina M, Schmidl C, et al. Technical Advance: Transcription factor, promoter, and enhancer utilization in human myeloid cells. J Leukoc Biol. 2015;97(5):985-95. Epub 2015/02/27. doi: 10.1189/jlb.6TA1014-477RR. PubMed PMID: 25717144; PubMed Central PMCID: PMCPMC4398258.
- Kaur S, Raggatt LJ, Batoon L, Hume DA, Levesque JP, Pettit AR. Role of bone marrow macrophages in controlling homeostasis and repair in bone and bone marrow niches. Semin Cell Dev Biol. 2017;61:12-21. Epub 2016/08/16. doi: 10.1016/j.semcdb.2016.08.009. PubMed PMID: 27521519.
- Mabbott NA, Baillie JK, Brown H, Freeman TC, Hume DA. An expression atlas of human primary cells: inference of gene function from coexpression networks. BMC Genomics. 2013;14:632. Epub 2013/09/24. doi: 10.1186/1471-2164-14-632. PubMed PMID: 24053356; PubMed Central PMCID: PMCPMC3849585.
- Pettit AR, Chang MK, Hume DA, Raggatt LJ. Osteal macrophages: a new twist on coupling during bone dynamics. Bone. 2008;43(6):976-82. Epub 2008/10/07. doi: 10.1016/j.bone.2008.08.128. PubMed PMID: 18835590.
- Rojo R, Pridans C, Langlais D, Hume DA. Transcriptional mechanisms that control expression of the macrophage colony-stimulating factor receptor locus. Clin Sci (Lond). 2017;131(16):2161-82. Epub 2017/08/02. doi: 10.1042/CS20170238. PubMed PMID: 28760770.
- Schultze JL, Freeman T, Hume DA, Latz E. A transcriptional perspective on human macrophage biology. Semin Immunol. 2015;27(1):44-50. Epub 2015/04/07. doi: 10.1016/j.smim.2015.02.001. PubMed PMID: 25843246.
- Summers KM, Bush SJ, Hume DA. Network analysis of transcriptomic diversity amongst resident tissue macrophages and dendritic cells in the mouse mononuclear phagocyte system. PLoS Biol. 2020;18(10):e3000859. Epub 2020/10/09. doi: 10.1371/journal.pbio.3000859. PubMed PMID: 33031383; PubMed Central PMCID: PMCPMC7575120.