Mouse Tissue-Resident Peritoneal Macrophages in Homeostasis, Repair, Infection, and Tumor Metastasis.

Peritoneal tumor metastasis occurs after detachment of tumor cells from primary tumors exposed to the peritoneal cavity, being ovarian, gastric, colorectal, pancreatic, and appendicular cancers the most prone to peritoneal metastasis. [ 72 , 73 ] Patients with peritoneal metastasis have a poor prognosis and suffer from excruciating symptoms like intestinal obstruction, accumulation of malignant ascites, and severe incurable pain syndromes that severely compromise their quality of life in the terminal stages of disease. Peritoneal metastasis requires an initial interaction of detached tumor cells with the mesothelial lining that covers the peritoneal wall and the organs located in the peritoneal cavity. [ 74 ] This initial step is followed by the invasion of the submesothelial space by tumor cells, leading to the remodeling of the peritoneal stroma, promoting the adhesion of tumor cells to the extracellular matrix, that supports their proliferation and thus, progression of metastasis. [ 72 ] During peritoneal tumor metastasis, tumor cells display a preferential tropism to the omentum, [ 75 ] that has been correlated to the hypoxic status of omental milky spots, possessing a unique vascular system that favors VEGF‐mediated angiogenesis, [ 76 ] and to the production of fatty‐acid binding proteins, that promote lipid transfer from adipocytes to tumor cells, enhancing β ‐oxidation metabolism and consequently tumor cell proliferation. [ 77 ] LPMs are the first potential line of defense against peritoneal metastasis, yet recent experimental evidence supports that tumors can subvert peritoneal macrophage metabolism and/or function, leading to the acquisition of a pro‐tumor phenotype. Although tumor promoting functions of macrophages were initially associated to monocyte‐derived macrophages, [ 78 ] embryonic tissue‐resident macrophages were recently demonstrated to be crucial role for tumor progression in mouse models of glioblastoma, [ 79 ] pancreatic ductal adenocarcinoma, [ 80 ] colon adenoma, [ 81 ] and lung carcinoma. [ 82 ] Peritoneal macrophages have been reported to promote metastatic tumor growth in mouse models of peritoneal cancer metastasis, [ 34 , 83 , 84 , 85 , 86 , 87 ] yet, as discussed below, the conclusions drawn from these studies are particularly divergent regarding the specific contribution of LPMs to peritoneal metastasis and the mechanistic basis of their tumor‐promoting role remains controversial ( Table   1 ). Experimental evidence of the tumor promoting function of peritoneal macrophages Weiss et al, 2018 ‐ Intraperitoneal injection of ID8 ovarian tumor cells or B16 melanoma tumor cells. LPMs characterized in tumor‐bearing mice as F4/80 high GATA6 + cells that proliferated during tumor progression. No expression of prototypical protumoral genes ( IL10, TGFB and Retnla) , but of proinflammatory genes ( IL6, TNFA and IL1B) . Irg‐1/itaconate‐induced ROS production by LPMs promoted tumor progression through MAPK activation in tumor cells. LPM depletion by administration of clodronate‐loaded liposomes led to delayed tumor progression. Irg‐1 silencing in peritoneal macrophages by lentiviral shRNA led to delayed tumor progression. Xia et al, 2020 ‐ Intraperitoneal injection of ID8 ovarian tumor cells. During tumor progression LPMs proliferated actively and generated Tim4 + TAMs. Tim4 + TAMs produced high levels of mitochondrial ROS production, and developed a high resistance to oxidative stress, through increased mitophagy. Not addressed. Autophagy deficiency in myeloid cells induced apoptosis in Tim4 + TAMs, due to accumulation of mitochondrial ROS, and led to delayed tumor progression. Chow et al, 2021 ‐ Intraperitoneal injection of MC38 colon carcinoma cells. Not addressed. Tim4‐mediated sequestration of phosphatidylserine‐expressing anti‐tumor cytotoxic CD8 + T cells, impairing anti‐tumor immunity. Antibody‐mediated Tim4 blockade, or Tim4 deficiency, enhanced the efficacy of anti‐PD‐1 immunotherapy. Goossens et al, 2019 ‐ Intraperitoneal injection of ID8 ovarian tumor cells. Monocyte‐derived F4/80 int MHCII int macrophages gradually replaced LPMs. Expression of genes related to cholesterol metabolism and reverse cholesterol efflux. Macrophage reprogramming associated to a tumor promoting phenotype, involving increased arginine metabolism, and inhibition of IFN γ ‐induced gene expression. IL‐4‐mediated, STAT6/PI3K‐dependent, macrophage reprogramming, triggered by cholesterol efflux, led to increased arginine metabolism, that promoted immunosuppression and tumor growth. Blocking IL‐4 signaling by anti‐IL‐4R delayed tumor growth. Prevention of cholesterol efflux by genetic deletion of ABC membrane cholesterol efflux transporters delayed tumor growth. Etzerodt et al, 2020 ‐ Intraperitoneal injection of ID8 ovarian tumor cells. Lyve‐1 + CD163 + Tim4 + tissue‐resident omental macrophages that expressed genes associated with positive regulation of JAK‐STAT signaling, linked with macrophage self‐renewal and genes associated with tumor‐promoting functions. Lyve‐1 + CD163 + Tim4 ‐ monocyte‐derived omental macrophages expressed genes associated with tumor‐promoting functions. CD163 + Tim4 + macrophages promoted the acquisition of cancer stem cell and epithelial‐to‐mesenchymal transition characteristics by ovarian cancer cells. Depletion of CD163 + Tim4 + macrophages by diphtheria toxin‐mediated specific ablation prevented metastatic spread and invasive disease. Depletion of CD163 + macrophages by CD163‐targeted, doxorubicin‐loaded, lipid nanoparticles, delayed omental tumor progression. Zhang et al, 2021 ‐ Intraperitoneal injection of ID8 ovarian tumor cells. Lyve‐1 high MHCII low mesenteric and peritoneal wall macrophages displaying an alternatively‐activated expression profile. Not addressed. Genetic ablation of Lyve‐1 high macrophages and omentectomy delayed tumor progression within ascites. Yin et al, 2016 ‐ Intraperitoneal injection of ID8 ovarian tumor cells. During tumor progression peritoneal cavity macrophages proliferated and switched gradually from expressing M1 ( Ccr2, Ifnar, iNOS ) to M2 genes ( Cd206, arginase 1, Cd163 ). Whether tumor‐associated macrophages resulted from proliferation of resident LPMs, and/or differentiated from monocytes, was not addressed. Formation of spheroids composed of macrophages and tumor cells through β 2 integrin‐ICAM‐1 interactions. Production of EGF by macrophages promoted tumor growth by EFGR signaling in tumor cells, triggering VEGF release by tumor cells and autocrine VEGFR signaling. Blockage of spheroid formation by clodronate‐liposome‐mediated macrophage depletion or anti‐ICAM‐1 antibodies delayed tumor growth. Pharmacological blockade of EGFR delayed tumor growth. LPMs were reported to infiltrate primary ovarian tumors, in an orthotopic syngeneic ovarian cancer model, in which Upk10 ovarian tumor cells were injected in the ovarian bursa. The number of tumor‐infiltrating LPMs was markedly reduced in mice deficient in RXR αβ in myeloid cells, displaying a disrupted LPM homeostasis, and this was associated with a delayed tumor growth, suggesting that LPMs promoted tumor progression, yet the mechanistic basis of this pro‐tumor effect was not addressed. [ 34 ] Interestingly, ovarian and melanoma peritoneal tumors were reported to subvert peritoneal macrophage metabolism, resulting in Irg1‐induced production of itaconic acid by LPMs from tumor‐bearing mice, that led to a fatty acid oxidation‐mediated increase in oxidative phosphorylation and glycolysis. [ 86 ] This caused an enhanced mitochondrial ROS production by LPMs, that promoted tumor progression, through ROS‐mediated MAPK activation in tumor cells. Tumor growth was reduced after peritoneal macrophage depletion by administration of clodronate‐loaded liposomes, or Irg‐1 silencing in peritoneal macrophages by lentiviral shRNA leading to decreased ROS production, supporting that LPMs had a tumor‐promoting function. [ 86 ] Interestingly, in this experimental setting, the protumor phenotype of LPMs was not related to the expression of prototypical protumor genes, such as IL10 , TGFB , and Retnla , but proinflammatory genes, such as IL6 , TNFA , and IL1B . In line with this report, during ovarian peritoneal metastasis, LPMs were demonstrated to proliferate and generate a population of Tim4 + tumor‐associated macrophages (TAMs), that displayed increased mitochondrial activity and mitochondria‐related ROS production, and developed a high resistance to oxidative stress through increased mitophagy, that led to the elimination of damaged mitochondria. [ 88 ] Interestingly, mitophagic activity was claimed to correlate with a high arginase‐1 activity increasing arginine metabolism in Tim4 + TAMs, resulting in low levels of arginine, that caused mTORC1‐mediated mitophagy inhibition. [ 88 ] Mice in which myeloid cells were deficient in FIP200, a protein essential for the induction of autophagy, [ 89 ] displayed a significant reduction in Tim4 + TAMs, due to apoptosis caused by accumulation of damaged mitochondria and mitochondrial ROS, and a delayed peritoneal ovarian tumor progression, supporting that Tim4 + TAMs promoted tumor progression. [ 88 ] An alternative tumor‐promoting function of LPMs has been reported in a recent study, based on a model of colon carcinoma peritoneal metastasis, claiming that LPMs supported tumor progression by impairing anti‐tumor CD8 + T cell immunity, through Tim4‐mediated sequestration of phosphatidylserine‐expressing anti‐tumor cytotoxic CD8 + T cells, preventing their proliferation. [ 85 ] Indeed, antibody‐mediated blockade of Tim4, or Tim4 deficiency, enhanced the efficacy of anti‐PD‐1 immunotherapy, that was associated with an increase in CD8 + T cells in the peritoneal cavity. However, an in vivo imaging of Tim4‐mediated macrophage‐CD8 + T cell interactions, demonstrating the sequestration hypothesis, was not provided in this report. A recent study from Dr. T. Lawrence's lab has demonstrated, in a peritoneal ovarian cancer model, that LPMs were progressively replaced by monocyte‐derived macrophages, that gradually upregulated genes related to cholesterol metabolism and reverse cholesterol efflux. [ 87 ] Indeed, tumor cells promoted cholesterol efflux in monocyte‐derived macrophages that drove IL‐4‐mediated, STAT6/PI3K‐dependent, macrophage reprogramming, involving increased arginine metabolism, and inhibition of IFN γ ‐induced gene expression, that promoted tumor progression. Genetic deletion of ABC membrane cholesterol efflux transporters prevented cholesterol efflux and reverted the tumor promoting functions in peritoneal monocyte‐derived macrophages. [ 87 ] Whether these monocyte‐derived macrophages acquired, in the long‐term a resident LPM identity was not addressed in this study. In a second report from the same group, based on the same peritoneal metastasis model, two populations of Lyve‐1 + omental macrophages, with different protumor functions, were described. Lyve‐1 + CD163 + Tim4 + tissue‐resident embryonic‐derived macrophages were required for metastatic spread of ovarian cancer cells, whereas both Lyve‐1 + CD163 + Tim4 + , and Lyve‐1 + CD163 + Tim4 − monocyte‐derived omental macrophages, contributed to tumor progression in the omentum. [ 83 ] Transcriptomic analyses revealed that CD163 + Tim4 + macrophages expressed genes associated with positive regulation of JAK‐STAT signaling, linked with macrophage self‐renewal, [ 90 ] while both CD163 + Tim4 + and CD163 + Tim4 − macrophages expressed genes associated with tumor‐promoting functions, such as angiogenesis, blood vessel development, and tissue remodeling. [ 91 ] Specific depletion of CD163 + Tim4 + macrophages did not affect omentum metastasis seeding, but prevented the development of invasive disease, that correlated with the ability of CD163 + Tim4 + omental macrophages to promote the acquisition of cancer stem cell and epithelial‐to‐mesenchymal transition characteristics by ovarian cancer cells. [ 83 ] The developmental and functional link between LPMs and omental CD163 + Tim4 + macrophages remains to be established. In line with these results, CCR1‐deficiency in ovarian tumor cells led to a reduction in their omentum seeding, that was claimed to depend on the expression of the CCR1‐ligand CCL6 by omental macrophages. [ 92 ] Interestingly, G. Randolph's lab identified recently two populations of F4/80 high ICAM‐2 − CD206 + tissue‐resident macrophages, mainly located in the mesentery and peritoneal wall, characterized as Lyve‐1 low MHCII high and Lyve‐1 high MHCII low cells, that were not dependent on the GATA6 transcription factor. [ 37 ] Lyve‐1 high MHCII low macrophages were shown to be of embryonic origin, depend on CFS1 and display an alternatively‐activated macrophage phenotype and, based on their transcriptional profile, claimed to be related to omental, tumor‐promoting, Lyve‐1 + CD163 + Tim4 + macrophages, recently described. [ 83 ] The potential protumor function of Lyve‐1 high macrophages was assessed in mice with genetic ablation of Live‐1 + cells, that were omentectomized, in order to exclude the tumor promoting function of omental Lyve‐1 + CD163 + Tim4 + macrophages. Ovarian peritoneal tumor growth was delayed in these mice, particularly within ascites, supporting that mesenteric and peritoneal wall Lyve‐1 high MHCII low macrophages promoted tumor progression. [ 37 ] Interestingly, in a mouse model of peritoneal ovarian cancer, based on the intraperitoneal injection of ID8 tumor cells, spheroids formed by peritoneal macrophages and tumor cells, through  β 2 integrin‐ICAM‐1 interactions, were detectable in the ascites from 3 weeks after ID8 injection. [ 84 ] Macrophages present in the peritoneal cavity increased in number almost tenfold along the first 8 weeks after ID8 injection, and switched gradually from expressing M1 ( Ccr2 , Ifnar , iNOS ) to M2 genes ( Cd206 , arginase 1 , Cd163 ). Production of EGF by macrophages promoted tumor growth by EFGR signaling in tumor cells, triggering VEGF release by tumor cells and autocrine VEGFR signaling. Prevention of spheroid formation by macrophage depletion or anti‐ICAM‐1 antibodies, or pharmacological blockade of EGFR, significantly delayed tumor growth, [ 84 ] supporting that peritoneal macrophages are critical for ovarian cancer metastatic progression by driving spheroid formation. Whether the TAM population involved in spheroid formation resulted from proliferation of resident LPMs, and/or differentiated of ii‐moLPMs, was not addressed in this report. In conclusion, research developed over the last years has demonstrated that, during peritoneal cancer metastasis, LPMs contribute to tumor progression, most likely reflecting both an intrinsic protumor potential, and the acquisition of tumor‐promoting functions, through tumor‐induced changes in their metabolism. Different molecular mechanisms have been proposed to explain the protumor function of LPMs, summarized in Table 1, but additional experimental work is needed to integrate these data and achieve an in‐depth and comprehensive understanding of the role of LPMs in peritoneal tumor progression. Importantly, in addition to LPMs, different peritoneal macrophage subpopulations with protumor potential have been recently identified, including Lyve‐1 + CD163 + omental macrophages, and Lyve‐1 high MHCII low mesenteric and peritoneal wall macrophages that, consequently, have to be taken into account in the design of experiments aiming at exploring the tumor promoting function of peritoneal macrophages, and in the development of immunotherapeutical antitumor strategies.

Inflammatory reactions in the peritoneal cavity induced by sterile inflammatory stimuli, [ 5 , 39 , 42 , 45 , 46 ] abdominal surgery, [ 39 ] or bacterial infection [ 47 ] were reported to cause LPM cell death leading to a reduction in the number of resident LPMs (including resident embryonic LPMs and resident moLPMs), whose extent correlates with the severity of inflammation. [ 42 , 46 ] Recovery of the original LPM pool occurs by proliferation of the remaining resident LPMs [ 45 ] and replacement by LPMs derived from inflammatory monocytes (hereafter ii‐moLPMs for inflammation‐induced moLPMs) as demonstrated using different experimental strategies, based on fate‐mapping models, [ 42 ] tissue‐protected bone marrow chimeric mice and adoptive transfer experiments. [ 39 , 46 ] Using an experimental model based on the induction of mild inflammation, caused by low‐dose zymosan (10 µg per mouse), or severe inflammation caused by high‐dose zymosan (1000 µg per mouse), and adoptive transfer experiments to track ii‐moLPMs and assess how the inflammatory environment controls their differentiation, Jenkins and colleagues proposed that the degree of replacement of resident LPMs by ii‐moLPMs, and the extent to which the later acquire the identity and function of resident LPMs is determined by the severity of the inflammatory process and the magnitude of LPM death [ 46 ] (Figure  2 ). ii‐moLPMs formed after mild inflammation co‐existed long‐term with remaining resident LPMs, but competition with resident LPMs and alterations in peritoneal environment retained them in an aberrant state of activation, and blocked the acquisition of a resident LPM phenotype. In contrast, severe inflammation can lead to the total ablation of resident LPMs, which are ultimately replaced by ii‐moLPMs, that acquired a resident LPM identity, but maintained transcriptionally and functionally divergent features, determined by their origin, peritoneal inflammation, and time‐of‐residency. [ 46 ] The phenotype of ii‐moLPMs was proposed to comprise intrinsic markers determined by their origin, such as CD62L and Semaphorin 4a, markers whose expression is controlled by competition with resident LPMs but is reprogrammed with time, such as GATA6, MHCII, and CCR5, and markers related to time‐of‐residency, independent of competition with resident LPMs, such as Tim4, CD209b, and VSIG4. A significant proportion of genes differentially expressed by resident LPMs and ii‐moLPMs appear to be controlled by differences in retinoic acid signaling, either directly or in a GATA6 dependent manner. [ 46 ] ii‐moLPMs exhibit a higher proliferative activity than resident LPMs, [ 38 , 46 ] that was suggested to correlate with differences in the enhanced ability of the former to proliferate in response to CSF1 produced by mesothelial cells. [ 36 ] In addition, ii‐moLPMs displayed a lower ability to phagocytose bacteria and uptake dying cells, and failed to produce CXCL13. [ 46 ] While the number of peritoneal B1 cells increase with age in homeostasis, peritoneal inflammation led to a defective accumulation of B1 cells [ 46 ] since, as pointed out above, CXCL13 production by LPMs control B1 cell homing to the peritoneal cavity. [ 48 ] Therefore, the fact that developmental and functional heterogeneity of the LPM population depends on sex and age has important implications when addressing the role of LPMs in repair, defense, and implication in peritoneal tumor metastasis, that need to be taken into account in futures studies. It is important to note that monocytes recruited to the peritoneal cavity during inflammatory reactions, related to non‐infectious peritoneal damage, infection, or metastatic tumor growth, can potentially differentiate into monocyte‐derived cells that fulfill specific repair, defense, or tumor‐promoting functions, but might not acquire phenotypic or functional LPM characteristics, and thus should not be considered ii‐moLPMs. However, defining the identity of cells differentiated from monocytes recruited to the inflamed peritoneum can be controversial since in most reports focusing on the functional relevance of peritoneal monocyte‐derived cells, the time‐of‐persistency, and/or acquisition of LPMs features by these monocyte‐derived cells was not addressed and, inversely, in reports on resident LPM replacement during inflammation, the function of ii‐moLPMs was not explored in‐depth. In line with the hypothesis that competition for a particular physical niche, defined by cellular and molecular microenvironmental factors, determines the contribution of monocytes to tissue resident macrophages, [ 43 ] the existence of a biochemical niche for peritoneal resident macrophages was proposed. [ 46 ] Accordingly, competition for signals and cell‐to‐cell interactions controlling survival, proliferation and function of LPMs would control the balance between resident LPMs and ii‐moLPMs, as well as, the acquisition of mature resident LPM identity by ii‐moLPMs.

Research developed over the last few years has significantly broadened our understanding of LPM biology by defining the dynamics of the replacement of resident embryonic LPMs by resident moLPMs, leading to sexually dimorphic phenotype and function, and explaining how LPMs, that move passively in the fluidic environment of the peritoneal cavity in the steady state, form mesothelium‐bound LPM aggregates to fulfill a crucial role in repairing peritoneal injuries and controlling microbial and parasitic infections. On the other hand, a number of recent reports have demonstrated that, during peritoneal cancer metastasis, LPMs contribute to tumor progression, most likely reflecting an intrinsic protumor potential and/or the acquisition of tumor‐promoting functions, through tumor‐induced changes in their metabolism. Different molecular mechanisms have been proposed to explain the protumor function of LPMs, summarized in Table 1, but additional experimental work is needed to integrate these data and achieve an in‐depth and comprehensive understanding of the role of LPMs in peritoneal tumor progression. Importantly, in addition to LPMs, different peritoneal macrophage subpopulations with protumor potential have been recently identified, including Lyve‐1 + CD163 + omental macrophages, and Lyve‐1 high MHCII low mesenteric and peritoneal wall macrophages that, consequently, have to be taken into account in the design of experiments aiming at exploring the tumor promoting function of peritoneal macrophages, and in the development of immunotherapeutical antitumor strategies. Importantly, these studies have revealed that the protumor potential of LPMs can be reverted by strategies blocking tumor‐induced subversion of LPM metabolism, providing the basis for the development of novel immunotherapeutic approaches against peritoneal tumor metastasis based on peritoneal macrophage reprogramming.

The peritoneal cavity, as well as the pleural and pericardial cavities, are generated, from the embryonic coelome, a cavity resulting from the formation of the embryonic body wall, comprising the parietal plate mesoderm and the ectoderm, and the gut wall, comprising the visceral plate mesoderm and the endoderm. [ 1 ] The process by which the embryonic coelome is formed has been conserved from the primitive superphyla Protostomia and Deuterostomia, so that invertebrates of the phyla Annelida, Mollusca, Echinodermata, and Tunicata possess a coelomic cavity anatomically and developmentally equivalent to the embryonic coelome, [ 2 ] that generates the peritoneal, pleural, and pericardial cavities during the embryonic development of mammals. The peritoneal cavity is covered by the peritoneum, the largest serous membrane of the body, with a surface area comparable to that of the skin, composed by the mesothelium, an epithelium of mesodermal origin, a basal membrane, and a submesothelial connective tissue. [ 3 ] The parietal peritoneum lines the inner surface of the abdominal wall, whereas the visceral peritoneum integrates with the serosal layers of intra‐abdominal organs. A double fold of the peritoneum forms the mesentery, that connects abdominal digestive organs to the abdominal wall, and serves as a conduit for vessels, nerves, and lymphatics. A small volume of peritoneal fluid secreted by mesothelial cells serves as a lubricant in the peritoneal cavity, and prevents mechanical friction between abdominal organs. In mice, total peritoneal fluid volume was estimated in two recent reports to be around 50–100 µL in the steady state, [ 4 , 5 ] and was claimed to differ between males and females (≈20 µL vs ≈100 µL) and, in the latter, to change during the estrous cycle. [ 6 ] Drainage of the peritoneal fluid into the lymphatic system allows peritoneal fluid recirculation, [ 7 ] and is achieved through openings in the mesothelium, called stomata, that are mainly located in the diaphram and omentum. [ 3 ] The omentum is a visceral adipose tissue that develops by overgrowth of the mesentery and harbors a specialized vascular system and an organized lymphoid tissue, claimed to play an important role in defense against peritoneal infection. [ 8 ] Peritoneal fluid draining through the diaphragm collects into the subperitoneal lymphatic lacunae to reach the diaphragm collecting lymphatics, that drain into the mediastinal lymph nodes, whereas peritoneal fluid draining through the omentum collects in the omental lymphatics, that in turn collect into the intestinal lymphatic trunk that connects to the thoracic duct through the cisterna chyli. [ 3 ] Drainage of the peritoneal cavity allows control of peritoneal homeostasis and leukocyte recirculation, but increases the risk of pathogen and metastatic tumor cell dissemination. The peritoneal cavity is exposed to two major pathologies, infection and tumor metastasis, generally associated with a high mortality, due to the easy spreading of pathogens or tumor cells throughout intra‐abdominal organs, and to the anatomical features of the peritoneal cavity that greatly hinders the development of efficient treatments against these diseases. Despite the fact that the peritoneal cavity is a confined space, not readily exposed to invading pathogens such as those penetrating the skin, the lungs or the gut, peritoneal infections can arise due to the loss of intestinal wall integrity (caused by ulcers, strangulation of hernias, appendicitis, or tumor growth), liver cirrhosis, accidental abdominal injuries, abdominal surgery, or peritoneal dialysis. The peritoneal cavity is also exposed to injuries in the parietal or visceral peritoneum caused by trauma, infection, or abdominal surgery, which can lead to peritoneal adhesions. Additional pathologies of the peritoneal cavity include—peritoneal endometriosis, involving the formation of ectopic vascularized endometrial tissue in the peritoneum associated with chronic inflammation—peritoneal autoimmune serositis, a chronic inflammation of the peritoneum caused by autoimmune diseases, such as Crohn's disease and—post‐surgical peritoneal adhesions. [ 3 , 9 , 10 ] Immune defense against peritoneal infection and tumor metastasis relies on a first line of local defense supported by resident peritoneal immune cells, present in the peritoneal cavity in the steady state, with innate immunity sensing and responding properties. The second line of immune defense in the peritoneal cavity is provided by functional units of lymphoid tissue, associated to adipose tissue located in the omentum, mesentery or gonadal fat, called fat‐associated lymphoid clusters (FALCs), or milky spots for omental FALCs. [ 8 ] FALCs harbor an structural organization similar to that found in secondary lymphoid organs, including a reticular cell‐based stroma, B and T cell compartments, and specialized blood and lymphatic vessels, allowing leukocyte migration to and from the peritoneal cavity. [ 8 ] Resident peritoneal immune cells include tissue‐resident peritoneal macrophages, generally named large peritoneal macrophages (LPMs) and B1 cells. Recent experimental evidence has unveiled that, apart from their primary phagocytic function, LPMs fulfill different homeostatic, repair and immunological defense functions, that reflect a previously unexpected functional plasticity. [ 11 ] Peritoneal B1 cells are considered as innate‐like B cells, that constitutively produce natural IgM, providing local immune protection against a wide variety of pathogens. In addition, B1 cells actively produce IgM in response to viruses, bacteria, fungi, and parasites. [ 12 ] The first line of immunity in the peritoneal cavity in mammals, relying on phagocytic and antibody‐mediated defense mechanisms supported by LPMs and B1 cells, is reminiscent of the primitive defense mechanisms sustained by different populations of coelomocytes present in the coelomic cavity of invertebrates. [ 13 , 14 , 15 ] Immune defense strategies in coelomic cavities have been therefore highly conserved throughout evolution from invertebrates to higher vertebrates. In this review, we discuss recent evidence that has widened our knowledge on the biology of LPMs by describing the mechanisms of resident embryonic LPM replacement by resident bone marrow monocyte‐derived LPMs (moLPMs), that result in phenotypic and functional LPM sexual dimorphism, and unveiling how LPMs, free in a fluidic environment in the steady state, perform repair and immune defense functions, by forming thrombus‐like structures in response to peritoneal injury, and mesothelium‐bound dynamic LPM aggregates after bacterial infection. Moreover, recent experimental evidence support that peritoneal tumors can subvert LPM metabolism, leading to the acquisition of tumor‐promoting functions that, nevertheless, might be reverted by experimental strategies blocking tumor‐induced subversion of LPM function, that could be the basis for the development of novel immunotherapeutic approaches against peritoneal tumor metastasis based on macrophage reprogramming.

The authors declare no conflict of interest.