Origin and ontogeny of lung macrophages: from mice to humans

Summary

Macrophages are tissue‐resident myeloid cells with essential roles in host defense, tissue repair, and organ homeostasis. The lung harbors a large number of macrophages that reside in alveoli. As a result of their strategic location, alveolar macrophages are critical sentinels of healthy lung function and barrier immunity. They phagocytose inhaled material and initiate protective immune responses to pathogens, while preventing excessive inflammatory responses and tissue damage. Apart from alveolar macrophages, other macrophage populations are found in the lung and recent single‐cell RNA‐sequencing studies indicate that lung macrophage heterogeneity is greater than previously appreciated. The cellular origin and development of mouse lung macrophages has been extensively studied, but little is known about the ontogeny of their human counterparts, despite the importance of macrophages for lung health. In this context, humanized mice (mice with a human immune system) can give new insights into the biology of human lung macrophages by allowing in vivo studies that are not possible in humans. In particular, we have created humanized mouse models that support the development of human lung macrophages in vivo. In this review, we will discuss the heterogeneity, development, and homeostasis of lung macrophages. Moreover, we will highlight the impact of age, the microbiota, and pathogen exposure on lung macrophage function. Altered macrophage function has been implicated in respiratory infections as well as in common allergic and inflammatory lung diseases. Therefore, understanding the functional heterogeneity and ontogeny of lung macrophages should help to develop future macrophage‐based therapies for important lung diseases in humans.

The lung harbors several types of macrophages that are critical for barrier immunity to airborne challenges and lung homeostasis. In this review, we discuss topical issues and concepts related to lung macrophages, covering both mouse and human studies.

Abbreviations

BAL

bronchoalveolar lavage

BPD

bronchopulmonary dysplasia

GM‐CSF

granulocyte–macrophage colony‐stimulating factor

HSPCs

hematopoietic stem and progenitor cells

IL

interleukin

LPS

polysaccharide

M‐CSF

macrophage colony‐stimulating factor

MAF

musculoaponeurotic fibrosarcoma oncogene homolog

PAP

pulmonary alveolar proteinosis

PPARγ

peroxisome proliferator‐activated receptor γ

TGF‐β

transforming growth factor β

Alveolar macrophages

The lung has two main types of macrophages that reside in different anatomical compartments, namely interstitial and alveolar macrophages.1, 2, 3, 4, 5, 6 In addition, intravascular macrophages have been described in humans and other species, but they remain poorly characterized.7 Macrophage localization is related to specific tasks, thereby achieving a division of labor to best serve parenchymal organ function (Fig. 1). The tissue specialization of macrophages is shaped by instructive signals from the local environment (such as cytokines and metabolites), which induce the expression of distinct transcription factors imprinting tissue‐specific macrophage function.8, 9, 10, 11

Figure 1.

Heterogeneity of lung macrophages. Macrophages occupy distinct locations in the lung, which corresponds to a division of labor. Alveolar macrophages reside in the airways, where they promote barrier immunity and surfactant clearance. Alveolar macrophages depend on the cytokines granulocyte–macrophage colony‐stimulating factor (GM‐CSF) and transforming growth factor β (TGF‐β). Interstitial macrophages are located within the lung tissue. Studies in mice have revealed two different populations of interstitial macrophages: (i) LYVE‐1^high MHC Class II^low interstitial macrophages adjacent to blood vessels that are involved in wound healing and tissue repair. (ii) LYVE‐1^low MHC Class II^high interstitial macrophages found near neurons that are specialized in antigen presentation. Lung macrophages can derive either from embryonic precursors or from blood monocytes. Parts of the figure have been adapted from Servier Medical Art.

Alveolar macrophages express the master transcription factor peroxisome proliferator‐activated receptor γ (PPARγ), a key regulator of lipid metabolism, which is induced by the cytokine granulocyte–macrophage colony‐stimulating factor (GM‐CSF).12 Mouse studies have shown that transforming growth factor β (TGF‐β) is another cytokine required for alveolar macrophage development and homeostasis.13 In contrast to GM‐CSF, which is secreted by alveolar type II cells, TGF‐β produced by alveolar macrophages themselves supports their homeostasis.13 As their name suggests, alveolar macrophages are located in the lumen of alveoli, where the gas exchange takes place at the alveolar–capillary membrane. Their strategic location within the distal lung allows alveolar macrophages to clear the airways of microbes, dead cells, and other airborne particles through phagocytosis, which is essential to maintain the vital oxygen uptake, e.g. during respiratory infection in mice.14, 15 The lung environment confers an anti‐inflammatory phenotype on alveolar macrophages (see review by Hussell and Bell2). Alveolar macrophages express several factors that promote immune tolerance, such as TGF‐β, as well as inhibitory receptors, restraining their pro‐inflammatory activity under steady‐state conditions.2 An important tissue‐specific function of alveolar macrophages is their ability to catabolize lung surfactant. Lack of GM‐CSF or PPARγ, and hence the lack of alveolar macrophages, results in pulmonary alveolar proteinosis (PAP),16 an inflammatory lung syndrome caused by the defective clearance of surfactant. In addition, alveolar macrophages promote regeneration of the injured mouse lung through the production of repair factors like amphiregulin.17

Alveolar macrophages are the most‐studied macrophage population in the lung because they can be relatively easily obtained by bronchoalveolar lavage (BAL). However, it is noteworthy that this technique likely samples mainly ‘motile’ alveolar macrophages, whereas their ‘sessile’ counterparts (attached to the alveolar epithelium) may be under‐represented. Interestingly, mouse studies have shown that sessile macrophages within different alveoli communicate with each other and the alveolar epithelium through connexin 43 channels.18 Another unexplored possibility is that mouse alveolar macrophages may move between adjacent alveoli through pores of Kohn,19 as not all alveoli contain macrophages at a given time. Clearly, the potential link between alveolar macrophage motility and function deserves further study. Also, there are currently no surface markers to distinguish sessile from motile alveolar macrophages. Finally, BAL macrophages may comprise airway macrophages that not only reside in alveoli, but also in larger airways, such as bronchi.

Interstitial macrophages

Interstitial lung macrophages are less studied because their isolation requires digestion of lung tissue, which is less accessible in humans than BAL fluid. They are located in the space between the lung epithelium and capillaries, where they are in contact with other immune cells, such as lymphocytes and dendritic cells. However, in most studies their exact localization is unclear, which may be in the alveolar interstitium, the submucosa, or the perivascular adventitia.4 Interstitial macrophages likely contribute to barrier immunity in the lung together with alveolar macrophages. Furthermore, they perform antigen presentation and tissue remodeling in the lung.20 A characteristic feature of both mouse and human interstitial lung macrophages seems to be their ability to produce the immunosuppressive cytokine interleukin‐10 (IL‐10).5 Interstitial lung macrophages are heterogeneous and recently two main distinct lineages have been identified in mice by the groups of Ginhoux and Jakubzick (Fig. 1).20, 21 These are LYVE‐1^low MHC Class II^high interstitial macrophages, found adjacent to neurons, that specialize in antigen presentation; and LYVE‐1^high MHC Class II^low perivascular macrophages performing wound healing and tissue repair.20 Interestingly, the latter population regulates the permeability of blood vessels and thereby the influx of inflammatory cells in the lung, e.g. during fibrosis.20 These two interstitial macrophage subsets seem to be conserved between mice and humans, although the human LYVE‐1^high MHC Class II^low subset shares a CD206⁺ CD169⁺ MARCO⁺ phenotype with human alveolar macrophages (see below).20

Heterogeneity of human lung macrophages

Different monocyte–macrophage populations in the human lung have only recently been rigorously defined according to characteristic cell surface phenotypes.22, 23, 24, 25, 26, 27 Common surface markers for human lung macrophages (and monocytes) are HLA‐DR, CD11b, CD11c, and CD64. Human alveolar macrophages are large and highly autofluorescent CD14^low CD16⁺ cells defined by the expression of the mannose receptor CD206 and the sialoadhesin CD169, as well as the scavenger receptor MARCO. Interestingly, one study suggested that two populations of CD206^high CD169⁺ macrophages may be distinguished in lung tissue by intermediate versus high expression of CD163, the scavenger receptor for hemoglobin–haptoglobin.22 In contrast, most macrophages in BAL were CD163^high.22 It remains to be explored whether CD163 expression relates to macrophage origin, localization, activation state, or motility. Human interstitial macrophages are smaller than alveolar macrophages and have the surface phenotype CD206⁺ CD169^low CD14⁺ CD16⁺, whereas lung monocytes are CD206⁻ CD169⁻. Differences between human interstitial and alveolar macrophages in terms of phagocytosis and cytokine production in response to lipopolysaccharide (LPS) have been reported.28, 29

The advent of single‐cell RNA‐sequencing has led to deeper insights into the heterogeneity of human lung macrophages. Recent studies have revealed great macrophage diversity in healthy, malignant, and fibrotic lung tissue.30, 31, 32, 33 In particular, chemokines distinguished different clusters of human lung macrophages.34 This raises the idea that different types of macrophages may act as gate keepers of immune cell recruitment to the lung (e.g. neutrophils versus T lymphocytes), thereby ensuring an optimal response to environmental challenges. Furthermore, inter‐individual variation in the relative abundance of lung macrophage clusters was observed.31 Finally, unique macrophage clusters were found to be present in human lung disease and associated with poor survival, e.g. in lung cancer.30, 31, 34 Interestingly, although the main monocyte subsets are well‐conserved between humans and mice,34, 35 the lung macrophage compartment seems to be more diverse in humans.34 Further studies are needed to clarify whether the observed macrophage clusters represent different activation states or truly independent lineages. The former possibility may be more likely because macrophage clusters were rather poorly separated. It will also be interesting to investigate whether the different populations of human macrophages correspond to macrophages occupying distinct niches in the lung.

Macrophage origin and development in the steady‐state lung

In mice, seeding of tissues with macrophage progenitors occurs in three developmental waves from yolk sac, fetal liver, and adult bone marrow.11, 36, 37, 38, 39 This also seems to apply to the mouse lung.40 Therefore, macrophages are classified according to their origin as either derived from embryonic/fetal precursors that largely self‐renew locally or from adult blood monocytes that develop from hematopoietic stem and progenitor cells (HSPCs) in the bone marrow (Fig. 1). Mouse studies have demonstrated that alveolar macrophages are predominantly of embryonic origin in steady state, as they are maintained independently of circulating monocytes.41, 42, 43, 44 Specifically, mouse alveolar macrophages originate from fetal monocytes that seed the lung and differentiate after birth into mature alveolar macrophages under the influence of GM‐CSF, TGF‐β, and PPARγ.12, 13, 43 The critical role of GM‐CSF and PPARγ in alveolar macrophage development and homeostasis is conserved between mice12, 43, 45, 46, 47, 48, 49 and humans.50, 51, 52, 53, 54, 55 Moreover, l‐plastin, a protein regulating the actin cytoskeleton, is required for the migration of mouse macrophage precursors into the alveolar space.56 It has been suggested that continued lung growth after birth might require a further contribution of bone marrow‐derived macrophages to the alveolar macrophage pool.3 This idea is supported by the finding that the contribution of HSPC‐derived monocytes to the alveolar macrophage compartment in mice steadily increases with age.57, 58 Furthermore, alveolar macrophages may be depleted with ageing, leading to a decrease in their number in old mice.59 In contrast, interstitial macrophages have a mixed origin,40 originating from blood and lung monocytes in mice60 with a minor early contribution from yolk sac macrophages. The mixed origin may relate to distinct populations of interstitial macrophages that occupy specific niches in the lung.40 In mice, both the LYVE‐1^low MHC Class II^high and the LYVE‐1^high MHC Class II^low subset of interstitial lung macrophages are slowly replaced by circulating monocytes.20, 21 Finally, the monocytic origin of interstitial lung macrophages may explain why their energy metabolism relies on glycolysis, whereas alveolar macrophages mainly use fatty acid oxidation.12, 61

Homeostasis of lung macrophages

Macrophage development and proliferation depends on colony‐stimulating factor receptor 1 (CSFR1; CD115), the receptor for macrophage colony‐stimulating factor (M‐CSF) and IL‐34.8, 36, 62 Pioneering work by Sieweke and colleagues revealed that M‐CSF promotes the self‐renewal of mouse macrophages through a set of transcription factors (c‐MYC, KLF2, KLF4) that are shared with pluripotent stem cells.63, 64, 65 Musculoaponeurotic fibrosarcoma oncogene homolog (MAF) transcription factors inhibit M‐CSF‐induced macrophage proliferation through the suppression of c‐MYC and other self‐renewal genes.64, 65 Mouse studies demonstrated that alveolar macrophages are maintained in steady state through their stem cell‐like capacity for self‐renewal.41, 43, 44, 63, 66 Furthermore, it has been shown that human alveolar macrophages have the ability to proliferate in vitro 65, 67 and in vivo.68 Consistent with their self‐renewal ability, mouse alveolar macrophages have low expression of c‐MAF and MAFB.65 The ageing‐related reduction of alveolar macrophages in mice may be due to higher c‐MAF and lower c‐MYC expression.59 Based on mouse studies, GM‐CSF and TGF‐β are not only required for the development of alveolar macrophages, but also for their steady‐state homeostasis,13, 43 whereas M‐CSF is largely dispensable.69 Accordingly, mouse alveolar macrophages express GM‐CSF receptors (Csfr2a, Csfr2b), but little Csfr1.21 Moreover, alveolar macrophage recovery after genotoxic stress in mice is mostly dependent on GM‐CSF, whereas other tissue macrophages rely more on M‐CSF.41 Interleukin‐36γ is another cytokine that has been implicated in promoting the survival of mouse alveolar macrophages after influenza infection.70 The molecular mediators of alveolar macrophage self‐renewal are being identified in mice, such as sirtuin protein SIRT171 and BHLHE40/BHLHE41 transcription factors.72

Interstitial lung macrophages have a shorter lifespan than alveolar macrophages in rhesus macaques.73 Furthermore, mouse interstitial lung macrophages express more Csfr1, but less Csfr2a and Csfr2b, than alveolar macrophages.21 Despite their responsiveness to M‐CSF, the impaired self‐renewal capacity of interstitial lung macrophages may be explained by the observation that, in contrast to their alveolar counterparts, mouse interstitial lung macrophages express MAF transcription factors,21 likely blocking proliferation. The differential dependence on M‐CSF and GM‐CSF could allow separate regulation of the interstitial and alveolar macrophage compartments in the lung. After its depletion, the alveolar macrophage pool in mice recovers more slowly than the pool of interstitial macrophages.20 This might be due to the rapid recruitment of circulating monocytes that preferentially differentiate into interstitial lung macrophages in response to M‐CSF.

Macrophage origin in lung injury and inflammation

Mouse studies have shown that resident alveolar macrophages are depleted during severe tissue injury, e.g. after ionizing radiation, viral infection, and LPS‐induced lung injury.41, 74, 75, 76 It has been proposed that damage to the lung epithelium may lead to the loss of integrin‐dependent TGF‐β activation, thereby causing a reduction in alveolar macrophages.13, 77 Depending on the degree of injury, the alveolar macrophage pool in mice may be restored either by local proliferation of remaining resident macrophages41 or, if the injury is more severe, by the recruitment of monocytes into the alveolar niche that then differentiate into monocyte‐derived macrophages.75, 76, 78 The latter leads to a situation where alveolar macrophages of different origin (embryonic and monocytic) co‐exist in the lung. According to the ‘niche competition model’, the degree of monocytic macrophage origin is determined by accessibility of the niche and whether it is occupied.79 Furthermore, the type and timing of injury likely determine whether monocyte‐derived alveolar macrophages persist or not. Consistent with this idea, it has been reported that, in the acute LPS lung injury model, monocyte‐derived lung macrophages die after resolution of tissue damage,75 whereas they persist long‐term in mouse models of lung fibrosis or herpesvirus infection.76, 80

Resident lung macrophages can also expand after tissue damage, e.g. interstitial macrophages in the mouse lung after administration of bacterial products, such as LPS and CpG.60 This expansion was mediated by monocytes differentiating into interstitial lung macrophages, either through the CCR2‐dependent recruitment of blood monocytes (in response to LPS) or in a CCR2‐independent manner through local lung monocytes and splenic monocyte reservoirs (in response to CpG).60

The molecular pathways regulating the differentiation of monocytes into lung macrophages remain poorly defined, with the β‐catenin pathway potentially being involved, at least in mice.81 Furthermore, MAF transcription factors, highly expressed in monocytes, are progressively down‐regulated upon the differentiation of mouse monocytes into alveolar macrophages.80, 82

Impact of origin on lung macrophage function: nature versus nurture

The relative importance of cellular origin versus tissue environment9 in determining macrophage function is a topical question in the field. Mouse studies support the notion that the transcriptional identity of alveolar macrophages in steady state is largely enforced by the lung microenvironment.83, 84, 85, 86 Moreover, yolk sac macrophages, fetal monocytes, and bone marrow monocytes are all capable of reconstituting an empty alveolar niche and preventing PAP in mice.86 However, fetal monocytes repopulated the empty alveolar niche faster because of their greater proliferative ability.86 Furthermore, one mouse study found that the scavenger receptor MARCO is more highly expressed in resident embryonic‐derived than in monocyte‐derived alveolar macrophages,85 which may affect their clearance function. Therefore, it is worth testing whether fetally derived lung macrophages have greater tissue‐repair function than adult monocyte‐derived macrophages that may be more prone to support pro‐inflammatory responses in the injured lung.

The monocytic origin of lung macrophages likely matters more in the context of lung injury and inflammation than in steady‐state conditions. Specifically, epigenetic changes may result in either enhanced or reduced pro‐inflammatory activity.87 The former is called trained immunity, a form of innate memory that ensures enhanced protection against, for example, bacterial infection.87 However, it could also predispose to chronic inflammation after subsequent, repeated lung insults. This idea is supported by mouse studies of lung fibrosis. In the bleomycin‐induced lung fibrosis model, monocyte‐derived alveolar macrophages persist after the resolution of fibrosis and over time become transcriptionally and phenotypically similar to resident alveolar macrophages.80 However, during the acute phase of lung injury, monocyte‐derived alveolar macrophages have a distinct gene expression profile and they drive lung fibrosis in mice.80, 82 Similarly, single‐cell RNA‐sequencing has revealed a transitional pro‐fibrotic macrophage population in the same mouse model of lung fibrosis.88 Furthermore, a specific population of pro‐fibrotic alveolar macrophages, presumably of blood monocyte origin, is present in humans with lung fibrosis.32, 80 Finally, the gene expression profile of monocyte‐derived mouse alveolar macrophages differs from their resident counterparts during the acute inflammatory phase in the LPS‐induced lung injury model.89 On the other hand, monocyte‐derived alveolar macrophages can also acquire tolerogenic/regulatory properties that are beneficial for the host. For example, herpesvirus infection in mice causes the replacement of resident alveolar macrophages by monocyte‐derived macrophages, which protects against allergic airway inflammation.76

Ontogeny and homeostasis of human lung macrophages

In contrast, the origin and ontogeny of human lung macrophages, especially in lung diseases, is poorly understood because invasive in vivo experiments are impossible. However, useful information has been obtained from bone marrow and lung transplantation studies. In the context of allogeneic bone marrow transplantation, profound depletion of host alveolar macrophages occurs as a result of the toxic myelo‐ablative conditioning before transplantation. This causes rapid alveolar macrophage turnover, resulting in mixed host–donor chimerism, where alveolar macrophages mostly originate from donor hematopoiesis.90, 91, 92 This is consistent with the notion that blood monocytes derived from donor HSPCs replace resident alveolar macrophages. In contrast, lung transplantation is not associated with depletion of lung macrophages and therefore alveolar macrophages in the donor lung persist, although the extent of donor alveolar macrophage persistence differed markedly between studies.68, 93, 94, 95 These observations support the concept that, in the steady‐state lung, human alveolar macrophages are largely maintained by local self‐renewal, whereas during lung injury associated with macrophage depletion, alveolar macrophages are replaced by monocytes. The notion that human alveolar macrophages may be maintained independently of circulating monocytes during steady state is further supported by the finding that alveolar macrophages are present in certain leukemias, where blood monocytes are severely depleted.96 However, these studies do not determine conclusively whether human alveolar macrophages in the developing lung are of embryonic origin. Furthermore, the specific embryonic and adult progenitor of human alveolar macrophages is unknown.

Stillborn infants lack CD169⁺ alveolar macrophages,22 suggesting that alveolar macrophages develop postnatally in humans, which is similar to findings in mice.43 This is likely because, while GM‐CSF is already produced from gestational week ≤16, the alveolar niche is only established at birth, when the lungs are inflated, with subsequent continued maturation of the alveolar tree well into childhood. Furthermore, alveolar macrophages are present in neonates ≥48 hr after birth,97 suggesting that oxygen intake may drive their maturation.

Lung transplantation studies indicate that human interstitial macrophages (and/or lung monocytes) have a more rapid turnover than alveolar macrophages.94 Again, this does not conclusively establish the origin of human interstitial macrophages, but their origin is likely monocytic. Moreover, CD206⁺ CD169⁻ macrophages are present in the lung interstitium of stillborn infants22 and therefore develop earlier than alveolar macrophages, likely because the alveolar niche is not fully established at birth. However, it is unclear whether these interstitial macrophages give rise to CD169⁺ alveolar macrophages or whether alveolar macrophages are derived from a distinct precursor, independently of interstitial macrophages.

A model to study human lung macrophages in vivo

The biology of lung macrophages has been difficult to investigate in humans because of the lack of in vivo experimental systems. Humanized mouse models that support the development of human lung macrophages98 offer a solution to this limitation (Fig. 2). Human immune system mice are generated by the transplantation of immunodeficient mice with human CD34⁺ HSPCs.99, 100 We have improved human hematopoiesis in the mouse host by providing critical human factors, such as cytokines, through gene knock‐in.101 To create an in vivo model for human lung macrophages, we engineered human gene knock‐in mice expressing GM‐CSF, which allows the development of human alveolar macrophages.102 Subsequently, by providing multiple human cytokines, we created a humanized mouse strain named, ‘MISTRG’ that expresses the human proteins M‐CSF, IL‐3/GM‐CSF, signal‐regulatory protein α, and thrombopoietin in the Rag2 ^−/− Il2rg ^−/− background.103 Human lung macrophages develop in MISTRG mice after transplantation with fetal, neonatal, and adult CD34⁺ cells,102, 103, 104, 105 reinforcing the concept that human GM‐CSF supports human macrophage reconstitution of an empty alveolar niche. These results show that human alveolar macrophages can derive from HSPCs of different developmental age and can therefore have a monocytic origin. Importantly, human alveolar macrophages of monocytic origin are functional because they are able to prevent PAP.102

Figure 2.

A model to study human lung macrophages in vivo. Humanized ‘MISTRG’ mice express the human proteins macrophage colony‐stimulating factor (M‐CSF), interleukin‐3 (IL‐3)/granulocyte–macrophage colony‐stimulating factor (GM‐CSF), signal‐regulatory protein α(SIRPα), and thrombopoietin (TPO) in the Rag2 ^−/− Il2rg ^−/− background. MISTRG mice support the development of human blood monocytes as well as human interstitial and alveolar macrophages after transplantation with CD34⁺ hematopoietic stem and progenitor cells. Parts of figure have been adapted from Servier Medical Art.

Another advantage of this model is that human lung macrophages (both interstitial and alveolar) develop in the absence of pre‐conditioning (own observation and ref. 105), which allows the study of human lung macrophage development without irradiation‐induced tissue injury and inflammation. Furthermore, all types of blood monocytes that have been described in humans106 are found in MISTRG mice after transplantation with human CD34⁺ cells.103 Interestingly, in the human lung CD14⁺ CD16⁺ ‘intermediate’ monocytes are the predominant subset in the airways, whereas the CD14⁺ CD16⁻ ‘classical’ subset is more abundant in lung tissue.23, 25 The latter are also found in the lungs of MISTRG mice and may represent an extravascular reservoir of tissue monocytes in the lung (own observation and ref. 24), similar to what has been reported in mouse studies.60, 107 Limitations of humanized mouse models for studying human lung macrophages are the absence of human cells at birth, the interaction of human macrophages with mouse lung epithelium, and the limited lifespan of MISTRG mice after transplantation with human cells. Despite these limitations, the MISTRG model is an excellent tool to dissect the origin, development, and function of human lung macrophages in vivo.

Lung macrophage function: from cradle to grave

Macrophage function is impacted by the age of the host (Fig. 3). Alveolar macrophage adaptation to increased oxygen concentration at birth influences their maturation and function. In fact, mice lacking the oxygen‐sensing Von Hippel–Lindau protein have alveolar macrophages with an immature phenotype and impaired self‐renewal capacity.108 Furthermore, these immature alveolar macrophages are not able to prevent PAP, which is consistent with defective surfactant clearance.108 Alveolar macrophages in the newborn may also be hypo‐responsive to certain respiratory pathogens, e.g. to Pneumocystis, as has been shown in mice.109 Another study showed that exposure of neonatal mice to microbial extracts stimulates the phagocytic function of alveolar macrophages, which promotes resistance to respiratory pathogens, such as Streptococcus pneumoniae.110 Therefore, exposure to the external environment after birth likely promotes the maturation of alveolar macrophages.

Figure 3.

Factors regulating lung macrophage function. Environmental factors such as the gut and local lung microbiota as well as pathogen exposure over the lifetime of the host affect lung macrophage function. In addition, differences have been observed in macrophage function depending on the age and sex of the individual. By regulating the function of lung macrophages, these factors likely influence the susceptibility to lung diseases. Parts of figure have been adapted from Servier Medical Art.

On the other hand, neonatal lung inflammation, surfactant dysregulation, and altered alveolar macrophage maturity are factors involved in respiratory distress syndrome.111 Respiratory distress syndrome can develop into bronchopulmonary dysplasia (BPD), a common pulmonary complication that comes with extreme prematurity and prolonged mechanical ventilation.112 BPD is characterized by altered lung morphogenesis and increased inflammation, where the arrested development results in a reduced capacity for gas exchange that is due to a decrease in the number of alveoli and capillary formation.113 The very premature birth and the level of BPD severity in early age will have a knock‐on effect on lung function throughout adolescent life.114 Genetic impairment of alveolarization in mice rapidly increases macrophages in the neonatal lung, causing altered lung morphogenesis. Macrophage depletion rescued arrested lung development, suggesting that macrophages drive BPD pathology.115 Furthermore, macrophages in the fetal mouse lung mediate the inflammatory response to LPS by activation of the nuclear factor‐κB pathway that in turn disrupts airway morphogenesis.116 Accordingly, immature macrophage phenotype has been linked to disease severity in preterm infants with chronic lung disease/BPD in humans.117

Similarly, resident lung macrophages from old mice are phenotypically, transcriptionally, and functionally distinct from their younger counterparts with increased basal activation,118 but impaired cytokine production in response to infection.119, 120 The proliferative capacity of alveolar macrophages in old mice is also reduced as revealed in a recent single‐cell RNA‐seqencing study.121 Furthermore, the clearance of apoptotic neutrophils by mouse lung macrophages becomes impaired with age.59 Finally, alveolar macrophages from old mice are refractory to interferon‐γ, resulting in impaired killing of phagocytosed bacteria.118 Therefore, age‐related changes in lung macrophages likely favor a state of impaired host defense coupled with neutrophilic inflammation, causing excessive tissue damage in the lung. This may contribute to increased mortality after infection with important respiratory pathogens, such as S. pneumoniae 122 and influenza virus,59 as shown in mouse studies.

Environmental factors regulating lung macrophage function: the microbiota and pathogen exposure

Among environmental factors, the microbiota plays a dominant role in shaping immune activity (Fig. 3). The normal gut microbiota promotes host defense in the lung against viruses and bacteria. For example, butyrate produced by the gut microbiota enhances protection against influenza infection in mice, partly through stimulating the production of Ly6C^lo monocytes that differentiate into anti‐inflammatory lung macrophages.123 Furthermore, depletion of the gut microbiota by antibiotics causes impaired resistance to respiratory infection with S. pneumoniae in mice.124 The microbiota promotes resistance against respiratory bacterial infection through enhanced production of GM‐CSF, which stimulates bacterial killing by mouse alveolar macrophages.125 These mouse studies suggest that the ‘healthy’ microbiota stimulates the phagocytic function of alveolar macrophages and their responsiveness to bacterial products, such as LPS. On the other hand, systemic dysbiosis (reduced diversity of the gut microbiota), as it occurs in human immunodeficiency virus‐infected humans, triggers lung macrophage dysfunction.126

The local microbiota also has beneficial effects on lung immunity.127, 128 The microbiota in the lower airways in humans is established within the first 2 months after birth.129 Several studies support the notion that a ‘balanced’ lung microbiota supports the anti‐inflammatory function of lung macrophages and tolerance to airborne material, which has implications for lung‐tissue remodeling130, 131 and allergic responses132 in humans, as well as for immunosurveillance of cancer in mice.133 Furthermore, bacterial components stimulate IL‐10 production by interstitial lung macrophages, which protects against allergic airway inflammation in mice.60 This finding supports the concept that the local lung microbiota contributes to the hygiene hypothesis through effects on lung macrophages.5 Conversely, lung disease states are often associated with local dysbiosis that is due to abnormal airway structure, altered mucus production, and change in pH.127, 128 This local dysbiosis could trigger a dysregulated inflammatory response, involving lung macrophages, thereby predisposing to disease exacerbations.

Studies in mice housed in specific‐pathogen‐free conditions do not reflect the situation in humans who are exposed to a multitude of airborne microbes during their lifetime. This raises the idea that continued microbial exposure in humans may lead to a lung macrophage pool that is increasingly blood‐monocyte‐derived. Consistent with this possibility, it has been shown in mice that herpesviruses, which are very common in humans, promote an alveolar macrophage compartment mainly consisting of blood monocyte‐derived macrophages.76 Accordingly, herpesvirus infection of mice can have beneficial (protection from allergic airway inflammation)76 as well as deleterious (pro‐fibrotic responses) effects134 through modulating lung macrophages. Previous pathogen exposure may also result in trained immunity to respiratory infection with viruses and parasites through effects on alveolar macrophages, at least in mice.135, 136 Conversely, lack of mouse alveolar macrophages causes increased susceptibility to secondary bacterial infection.74 Moreover, it was shown that T‐cell‐produced interferon‐γ inhibits the phagocytic function of alveolar macrophages after influenza virus infection, which impairs the resistance against secondary bacterial infection in mice.137

Concluding remarks

Macrophages are currently one of the most topical cell types in immunology and great progress has been made recently in understanding their biology in mice. As far as human lung macrophages are concerned, several interesting open questions remain that merit further investigation (Table 1). With the emergence of powerful novel technologies, such as single‐cell RNA‐sequencing, CRISPR/Cas9 genome engineering, and advanced in vivo models, we expect rapid progress in the field of lung macrophages.

Table 1.

Open questions related to human lung macrophages

What is the embryonic and adult progenitor of human lung macrophages? Do lung macrophages of embryonic origin have greater tissue‐repair function?

What is the origin of lung macrophages in human lung disease? Are there surface markers that distinguish embryo‐derived from adult‐derived lung macrophages to track macrophage origin in human lung disease?

Does continued pathogen exposure over the lifetime lead to a human lung macrophage compartment that is mainly of monocytic origin? Which molecular pathways drive the differentiation of monocytes into human lung macrophages?

How can we distinguish sessile from motile human alveolar macrophages by flow cytometry? Do human macrophage clusters described in single‐cell RNA‐sequencing studies represent distinct lineages and/or occupy specific niches in the lung?

Do sex‐specific differences in lung macrophage function underlie gender bias in lung disease?

Disclosures

The authors have no competing interests to declare.