Macrophage-stem cell crosstalk: regulation of the stem cell niche

The immune system comprises a vast diversity of cell types and functions, which often makes the process of defining the roles of specific cells and signals during immune-regulated processes difficult. Most often, the immune system is defined for its function in pathogen sensing and clearance; however, it also has larger roles in regulating self by preventing autoimmunity, screening for cellular abnormalities within tissues and directing repair (reviewed by Paludan et al., 2020). Immune cells have been ascribed differing roles during tissue regeneration, many of which have been previously extensively reviewed (Alshoubaki et al., 2022). However, immune cells are now emerging as surprisingly direct regulators of the niche environment, across a number of tissues, that control both the formation and quiescence of adult stem cell populations.

A key role for stem cell niches is to maintain a state of stem cell quiescence or activate stem cell proliferation in response to injury (Cheung and Rando, 2013; Hicks and Pyle, 2023). The specific components of the niche differ depending on the tissue type, which comprises both the cellular and acellular components (see Box 1). As the central and consistent factor across different tissue and regeneration models, how immune cells and inflammation directly dictate stem cell fate has been a focus in recent years.

Macrophages have also emerged as a key innate immune cell type driving injury responsiveness and, despite their historical role as mainly phagocytic cells responsible for debris clearance and regulation of inflammation, are orchestraters during the entirety of the repair process (reviewed by Di Pietro et al., 2023). Macrophages are a highly heterogeneous cell type with complex mechanisms driving diverse cellular phenotypes (Sullivan et al., 2011; Link et al., 2018). As macrophages are found in a variety of states, there has been much contention over nomenclature and classification practices, with an increasing preference towards naming based on activation driver or specific markers (Murray et al., 2014; Sanin et al., 2022). However, they have been grossly characterized as either predominantly pro-inflammatory, and key drivers of the early inflammatory response, or predominantly pro-regenerative and more focused on stem cell activation, differentiation and wound resolution, concepts which will be discussed throughout this Review.

Although the idea of pro-regenerative macrophages directing aspects of stem cell behaviour is not new (Keightley et al., 2014), exactly how this is regulated at the cellular and molecular level has now come into sharper focus. Consequently, this Review will focus specifically on the new and emerging roles of macrophages in stem cell modulation, with a specific focus on macrophage-stem cell crosstalk and direct cell-cell contact requirements.

Aligning with the requirement for an appropriate inflammatory response, the control and balance of pro-inflammatory/pro-regenerative macrophages is a crucial component of the niche. The plasticity of macrophage isotype switching from pro-inflammatory in early injury to pro-regenerative during regeneration is a crucial component to their response, with a wide range of mechanisms including chromatin and epigenetic changes to microRNA-based priming (reviewed by Locati et al., 2020). Simplistically, naïve macrophages stimulated by inflammatory factors, such as interferon γ (IFNγ) and lipopolysaccharide (LPS), trigger nuclear factor kappa B (NF-κB) signalling and metabolic reprogramming to polarize towards a pro-inflammatory phenotype. These pro-inflammatory macrophages then release inflammatory cytokines including IL1β and tumour necrosis factor α (TNFα) to influence the niche (Takashiba et al., 1999; Su et al., 2015) (Fig. 1). As mentioned above, this inflammatory environment typically prevents stem cell activation and proliferation and instead can result in damage. Pro-regenerative macrophages have historically been classified in their role of being anti-inflammatory and can be polarized by a number of different stimuli, such as key anti-inflammatory cytokines interleukin (IL) 4, IL10, IL13 and transforming growth factor-β (TGF-β), and have diverse roles and impacts on stem cell behaviour, including promoting and co-ordinating activation, proliferation and differentiation (Cantini et al., 1994; Arnold et al., 2007; Saclier et al., 2013). However, as discussed below, the use of these broad definitions does not represent the heterogeneity or nuance of pro-regenerative macrophage populations nor adequately describe their actions towards stem cells during homeostasis and regeneration.

Along with the shifting of this macrophage polarization phenotype, there is also an observed shift requirement between different types of pro-regenerative macrophages, with differing roles for tissue-resident and infiltrating macrophages (Ginhoux and Guilliams, 2016; Gordon and Plüddemann, 2017). In adulthood, macrophages are derived from haematopoietic stem cells (HSCs) in the bone marrow (BM); however, colonization of tissues by primitive macrophages from erythromyeloid progenitors (EMPs) in the yolk sac generate long lived tissue-resident macrophage populations with wide transcriptional and functional heterogeneity, including microglia in the brain, alveolar macrophages in the lung and Kupffer cells in the kidney (Ajami et al., 2007; Klein et al., 2007; Ginhoux et al., 2010; Schulz et al., 2012; Guilliams et al., 2013) (Fig. 1). In addition, some tissue-resident populations are able to be replenished by circulating monocytes, as is observed in intestinal and renal macrophages (Varol et al., 2009; De Schepper et al., 2018; Liu et al., 2020) (Fig. 1). Tissue-resident macrophages are able to take on their tissue-specific identify via signals from the peripheral tissue in which they reside (Bonnardel et al., 2019; Sakai et al., 2019), and the ability for macrophages to take on and maintain this phenotype through their own macrophage niche has been nicely reviewed (Guilliams et al., 2020). It has also been shown in juvenile zebrafish that primitive macrophages are able to bypass typical HSC development processes and form typical infiltrating macrophages (Herbomel et al., 1999), suggesting that these early progenitors have the broad potential observed in adult macrophages.

The shift in macrophage requirement is sometimes observed as a decrease in the number of tissue-specific resident macrophages and increase in infiltrating wound-responsive macrophages for appropriate regeneration (Chakrabarti et al., 2018; Choi et al., 2020), indicating the balance of macrophage subpopulations may modulate stem cell responsiveness. This macrophage-specific balance has been observed in other tissues to the opposite effect. For example, mammary gland macrophages, but not peritoneal macrophages, stimulate proliferation of mammary gland stem cells (MaSCs) (Chakrabarti et al., 2018). Within the skin, it has also been observed that a fetal-derived and transcriptionally distinct resident macrophage population is required for surveillance, axon sprouting and regeneration of local neurons in response to injury (Kolter et al., 2019). Similarly, a transcriptionally distinct aortic macrophage population has been shown to develop an osteoclast-like phenotype in atherosclerosis, and therefore may contribute to calcification in aortic lesions (Cochain et al., 2018). In addition, tissue-resident macrophage population changes appear to be important during cancer progression, where folate receptor 2⁺ (FOLR2⁺) macrophages in the mammary gland, which are usually found localized to the perivasculature, are found in the tumour stroma and recruit T cells through release of chemoattractants, leading to increased survival rates (Nalio Ramos et al., 2022).

In skeletal muscle, a highly transcriptionally distinct macrophage population drives a seemingly tissue-specific stem cell proliferation response (Ratnayake et al., 2021). Recently, a T-cell immunoglobulin and mucin domain containing 4 (TIM4)⁺ long-lived, self-renewing resident macrophage (SRRM) population was identified (Babaeijandaghi et al., 2022). TIM4⁺ macrophages drive clearance of damaged cells and an ability to modulate the metabolic landscape of the tissue by driving a fibre-type switch from fast-twitch glycolytic fibres to more resistant, oxidative slow-twitch fibres (Babaeijandaghi et al., 2022). In addition, a tissue-resident TIM4-macrophage population was observed with potential to be replenished by circulating monocytes (Fig. 1), indicating a further example of circulating monocytes developing tissue-resident qualities.

Given the ability of macrophages to develop unique tissue-specific phenotypes and undergo subpopulation specification, as well as extensive heterogeneity across activated populations, there is an increasing focus on classifying macrophage responses based on either transcriptional subtype or effector molecule(s), rather than a more typical pro- and anti-inflammatory dichotomy. As such, a new framework to characterize macrophages based on different ‘activation paths’ has been suggested (Sanin et al., 2022). This uses predictive modelling to integrate multi-tissue data and identify four main activation paths, namely phagocytic, inflammatory/cytokine-producing, oxidative stress and remodelling paths. Despite this increasing shift towards new classifications and evidence of diversity of functional states, there remains merit in loosely identifying macrophages as being either pro-inflammatory or pro-regenerative, as these qualifiers remain good predictors of outcome, as changing proportions of each is often considered to be a factor contributing towards (or a consequence of) disease. However, how these unique tissue-specific pro-regenerative macrophage subtypes are individually generated and maintained during tissue repair remains largely unknown. This knowledge will be crucial for manipulating macrophage subtypes for delivering therapeutically relevant tissue regeneration and also for the concept of pro-regenerative macrophage cell therapy, as has been observed in many injury models including skeletal muscle (Dumont and Frenette, 2013; Martins et al., 2020; Stepien et al., 2020; Ratnayake et al., 2021; Babaeijandaghi et al., 2022), and holds much promise for directed tissue repair strategies (reviewed by Brown et al., 2014; Spiller and Koh, 2017; Poltavets et al., 2020).

Although macrophages are known to influence the development of specific tissue architectures, such as lymphatics (Gordon et al., 2010; Cahill et al., 2021), recent studies have also suggested a direct role for individual macrophage populations in the generation of tissue resident stem cell compartments, including macrophage-directed guidance of stem cells to permanent niches (Li et al., 2018), quality assurance of stem cells to direct clonality (Wattrus et al., 2022) and generation of transient niches to support proliferation during regeneration (Ratnayake et al., 2021).

Localization of stem cells to their niche is most studied from the perspective of haematopoietic stem and progenitor cells (HSPCs), which undergo a migration from the aorta-gonad-mesonephros (AGM) to their site of maturation in the fetal liver (mammals) or caudal haematopoietic tissue (CHT; zebrafish) before finally migrating to either the BM or kidney marrow, respectively (Medvinsky and Dzierzak, 1996; Amatruda and Zon, 1999; Murayama et al., 2006; Jin et al., 2007). Zebrafish live imaging has identified that primitive macrophages direct HSPC mobilization from the AGM and thus establishment of the mature HSC niche, and that the process of both HSPC and primitive macrophage mobilization from the AGM is dependent on the breakdown of extracellular matrix (ECM) by metalloproteinases (MMPs) (Travnickova et al., 2015; 2021). This process is also highly dependent on vascular cell adhesion molecule 1 (VCAM1)- and integrin subunit alpha 4 (ITGα4)-mediated cell-cell adhesion in both mouse and zebrafish (Papayannopoulou et al., 1995; Scott et al., 2003; Li et al., 2018). In the mouse, VCAM1 expression has also been observed in conferring immune tolerance of HSCs during their entry into the BM, where VCAM1 interacts with major histocompatibility complex II (MHC II) on phagocytes as a quality control mechanism (Pinho et al., 2022). Through live imaging and cell tracking in the larval zebrafish, it has been shown that the homing process is directed by a specific macrophage population that guides HSPCs to the CHT through VCAM1/ITGα4 adhesion (Li et al., 2018). Using a macrophage-ablation model, it was also demonstrated that macrophages are specifically required for HSPC retention in the CHT and that this also relies on their cell-cell adhesion capabilities. In global vcam1-knockout zebrafish, overexpression of vcam1 specifically in the macrophage lineage results in HPSCs being retained at VCAM1⁺ macrophage hotspot sites for up to 30 min, suggesting that macrophages may be involved beyond the homing stage (Li et al., 2018). Indeed, during this connection in the CHT, macrophages have been observed to pinch and sample contents of the HSPCs, as identified through uptake of fluorescently-labelled cytoplasm (Fig. 2A) (Wattrus et al., 2022). This was characterized as a macrophage-driven ‘grooming’ response, where ‘groomed’ HSPCs undergo cell division preceding this cytoplasmic sampling event. Here, a ‘dooming’ response, where some HSPCs were directly engulfed and phagocytosed, was observed and this was the predominant mechanism of HSPC death in the CHT. Overall, this implicates macrophages in the homing and fate of early HSPCs, thus providing evidence of direct immune regulation of stem cell development.

In addition to their role in directing HSPC homing, macrophages have been observed to support their maintenance, with CD196⁺ macrophages indirectly promoting the retention of HSCs in the BM via crosstalk with mesenchymal stem cells (MSCs) (Chow et al., 2011). This has also been observed with endosteal macrophages (which reside in the connective tissue of long bones), where they support the retention and homeostasis of HSCs, and the loss of endosteal macrophages leads to mobilization of HSCs (Winkler et al., 2010).

Beyond HSPCs, there is evidence to suggest that macrophages play a direct role in stem cell niche generation within the gastrointestinal tract (GIT) (Kim et al., 2022). Appropriate microbiota colonization during early development is additionally required for the generation of long-lived CD206⁺ tissue-resident macrophages, which are localized with mesenchymal niche cells (MNCs) during early gut development (Kim et al., 2022). Given the known role of macrophages in regulation of Paneth cell differentiation, this further suggests a role of macrophages in the development and promotion of Paneth cells of the intestinal stem cell niche during niche establishment (Huynh et al., 2009; Saha et al., 2016; Sehgal et al., 2018; Kim et al., 2022). Early developmental issues with appropriate formation of these intestinal crypt niches – and subsequent pathologies – can be partially rescued with the introduction of Lactobacillus, which promotes polarization of pro-regenerative macrophages and subsequently drives Paneth cell differentiation (Kim et al., 2022). Thus, commensal bacterial colonization during early development directs the generation of the local macrophage populations and generation of stem cell niches.

Currently, the greatest understanding of the role that macrophages play in directing stem cell behaviours has been illustrated in the context of adult stem cell populations, specifically the ability of macrophages to regulate the inflammatory environment. As such, the next sections will address the role of inflammation in directing stem cell migration and directing stem cell responses.

Different stem cell populations appear to be able to respond to inflammatory signals directly, often expressing immune receptors that sense the inflammatory environment, most often observed as homing and migration of HSPCs and MSCs to sites of inflammation (reviewed by Rustad and Gurtner, 2012; Ho and Takizawa, 2022). HSPC expansion and migration from the BM can be regulated by remote sites; in gastrointestinal tissue repair, commensal bacteria Bacteroides are able to drive an HSPC response, possibly through toll-like receptor (TLR) signalling (Sezaki et al., 2022). These HSPCs then contribute to gastrointestinal repair by differentiating into myeloid cells, including macrophages, and coordinating local inflammation. Alongside HSC expansion, there was an observed skew in differentiation towards the myeloid lineage, including macrophages, and migration of these progenitors towards the GIT to coordinate local inflammation (Sezaki et al., 2022). This is evidence that local inflammation can drive a multi-system approach to tissue and stem cell self-renewal. Interestingly, TLR expression by epithelial cells lining the GIT is not dictated by commensal bacteria, but a spatially-distinct expression pattern of TLRs determined by expression within stem cells during development (Kayisoglu et al., 2021). However, the requirement of these immune signalling pathways in localization of embryonic or adult stem cells in the GIT is unclear. Although stem cells can respond to inflammatory signals themselves, macrophages remain the interpreter and vector of inflammatory signals to the stem cells.

One of the central roles for macrophages in stem cell activation is modulating the inflammatory environment, which has been extensively reviewed (Chazaud, 2020; Ross et al., 2021). These pro-inflammatory signals – specifically IL1 – have been identified to drive activation of HSCs and bias their differentiation towards myeloid lineage specification (Pietras et al., 2016), while the pro-inflammatory regulator prostaglandin E2 (PGE2) has been identified as a factor involved in HSC expansion and homeostasis (North et al., 2007). Intriguingly, in a series of experiments investigating the clonal fitness of mutant HSPC progeny, it was identified that some dominant mutant clones can develop inflammatory resistance mechanisms to promote their survival, thus suggesting an interplay between inflammation and HSPC clonal selection (Avagyan et al., 2021). In addition, the pro-inflammatory environment can have unintended consequences of providing DNA stress and damage to stem cells, with muscle stem cells (MuSCs) upregulating stress genes in response to injury-induced activation (Machado et al., 2021). As such, the specific induction and regulation of this inflammatory environment is crucial. In order to combat this, intrinsic protection of MuSCs against inflammatory damage during regeneration has been observed, with MuSCs displaying protection from tumour necrosis factor (TNF)-driven cell death signals via the expression of Met and Cxcr4 (Lahmann et al., 2021). Inflammation has also been observed to hinder the ability of stem cells to exit quiescence. For example, MuSCs are able to bypass this block and enter the cell cycle via promoting epigenetic changes and modulating the ECM (Nakka et al., 2022). Here, in response to a pro-inflammatory environment, MuSCs undergo demethylation of has2, which encodes the enzyme that synthesises hyaluronic acid (HA), resulting in the secretion of HA. This presence of HA in the ECM then permits MuSCs to receive pro-regenerative signals. Therefore, the response of stem cells to the inflammatory environment is multi-layered and stem cells appear to possess, in some contexts, an intrinsic ability to protect themselves and guide their fate by modulating the niche environment directly. As a key inducer of the inflammatory environment, the appropriate polarization of macrophage phenotype is an important contributor to this stem cell response.

There is also increasing evidence to suggest that stem cells directly contribute to macrophage polarization. A direct immunomodulatory function of stem cells has previously only been ascribed to MSCs, which, in the context of macrophages, have been shown to drive activation (Jin et al., 2022), pro-inflammatory to pro-regenerative phenotype switching (Li et al., 2019) and switching between different pro-regenerative subtypes (Holthaus et al., 2022). Mechanistically, MSCs are thought to achieve this through the release of extracellular vesicles (EVs), which act as carriers to allow the direct transfer of factors, including protein and genetic material, to drive cell responses (Sicco et al., 2017; Ren et al., 2019; Nakazaki et al., 2021; Jin et al., 2022; Kang et al., 2022; Liu et al., 2022). EVs isolated from damage-induced MSCs have been shown to promote pro-inflammatory-to-pro-regenerative macrophage switching in order to drive an anti-inflammatory injury environment (Sicco et al., 2017). Furthermore, MSC-derived EVs isolated from fat deposits after hypoxia-induced injury are able to promote skeletal muscle healing by driving an anti-inflammatory response, despite arising from both a different tissue source and injury paradigm (Sicco et al., 2017). This suggests that these mechanisms driving niche control of the immune cell response may not be specific to an individual tissue microenvironment or to each tissue stem cell type.

There is increasing evidence that other stem cells are also able to influence this switching directly, with one study showing periodontal ligament stem cells (PDLSCs) can drive pro-regenerative macrophage polarization and an anti-inflammatory environment post-transplantation, thus promoting periodontal regeneration (Liu et al., 2019). This has also been observed in neural stem cells (NSCs), which release IL4 to simultaneously drive pro-regenerative macrophage polarization while inhibiting NF-κB signalling and thus suppressing polarization to a pro-inflammatory phenotype (Ji et al., 2020). NSC-conditioned media has also been shown to reduce inflammatory cytokine release by macrophages and decrease systemic inflammation through the downregulation of inducible nitric oxide synthase (iNOS) (Cheng et al., 2017).

Together, the emerging literature indicates that there are several stem cell types that are able to influence macrophage phenotypes, and that this is done in a largely non-specific manner via modulation of cytokine levels in the niche. Whether this represents a specific, intentional and targeted involvement of stem cells in driving macrophage polarization, or whether this is only a small aspect of the overall orchestration of inflammation, remains unclear. In addition, these events have largely been observed in transplantation models and in vitro, indicating that, although stem cells have an ability to influence the immune response within the niche, this may not be recapitulated in vivo. Whether this is a process used by endogenous stem cells to shape their own niche remains to be further explored.

Although immune modulation remains a crucial mechanism of macrophage action in the stem cell niche, more recent evidence indicates that macrophages also provide signals that directly coordinate stem cell behaviours during repair. It has been observed that macrophages modulate highly used pathways, such as Notch signalling, to direct stem cell fate in multiple tissue types (Du et al., 2017; Chakrabarti et al., 2018; Sultan et al., 2021). This requirement is highly established in muscle regeneration, where mature muscle progenitor cells express Delta-like canonical Notch ligand 1 (Dll1) and signal to MuSCs to induce Notch signalling and drive activation, proliferation and self-renewal (reviewed by Gioftsidi et al., 2022). Macrophages have been observed to release stem cells from quiescence by secreting the metalloproteinase ADAMTS1, which cleaves Notch intracellular domain (NICD), activating downstream Notch signalling and resulting in the activation of MuSCs (Du et al., 2017). Trajectory analysis of single-cell RNA-sequencing suggests that regulation of Notch signalling is also required for stem cell differentiation, with extracellular signal-regulated kinase 1/2 (ERK1/2) signalling initiating proliferation, followed by downregulation of NICD to initiate the myogenic process, together maintaining quiescence (Machado et al., 2021). Importantly, Notch signalling has been identified as a crosstalk mechanism in regenerative niches, with Dll1 enrichment in MaSCs driving Notch signalling in mammary gland macrophages (Chakrabarti et al., 2018). Here, Notch signalling between macrophages subsequently drives Wnt signalling to further support Dll1⁺ MaSCs and allow their repopulation, revealing a complex feedback mechanism of niche support.

Recent studies have identified more targeted stimulation of stem cell activation and proliferation by macrophages. This includes the release of the cytokine nicotinamide phosphoribosyltransferase (NAMPT) to directly stimulate MuSC proliferation (Fig. 2B) (Ratnayake et al., 2021). This was confirmed to be a specifically macrophage-driven and direct proliferative response, with lineage-specific ablation of macrophages preventing appropriate proliferation of MuSCs and their failure to appropriately respond to injury. NAMPT is often associated with its role in metabolism, as it is the rate-limiting enzyme responsible for nicotinamide adenine dinucleotide (NAD) biosynthesis (Revollo et al., 2004; Garten et al., 2015). NAMPT has been shown to be upregulated in tumour-associated macrophages (TAMs) in response to IFNγ-driven inflammation to drive their metabolic shifts (Huffaker et al., 2021). Although there remains potential for NAMPT-driven metabolic responses to play a role in the regenerative milieu, this MuSC proliferative response is not dependent on its enzymatic function, but rather dependent on a specific interaction with the C-C motif chemokine receptor 5 (CCR5) receptor on MuSCs (Ratnayake et al., 2021). Interestingly, myoblast proliferation preceding CCR5 activation has been observed previously, and also suggested as being ERK1/2 signalling-dependent (Yahiaoui et al., 2008). Therefore, despite being part of larger cell regulatory processes, NAMPT/CCR5 signalling appears to be a specific mechanism to drive MuSC proliferation.

These examples show how contained, specific signals and more complex signalling networks, such as Notch signalling and commensal bacteria, are used to stimulate stem cell response to injury. Modification of these pathways emphasizes the complexity of the system and that many of pathways may act in coordination with the niche microenvironment, rather than as direct stem cell control mechanisms.

The role of macrophages in injury and homeostasis is often viewed through the lens of their secretome, as well as their role in phagocytosis; however, these activities do not fully encompass their roles. More recently, with the development of in vivo visualization tools, this focus has broadened to include the role of macrophages in direct contact of stem cells (Fig. 2). Electron microscopy images of satellite cells in mouse skeletal muscle have shown that macrophages form prolonged connections during stem cell activation and differentiation, through which there appears to be direct membrane interaction (Ceafalan et al., 2018). This is suggested because of the close membrane proximities and possibility of hemifusion events between macrophages and MuSCs, which has also been observed in zebrafish (Ratnayake et al., 2021). How this works mechanistically, however, and whether it is a required mechanism for regeneration, remains unknown. Macrophages in regenerating skeletal muscle display extended and dynamic interactions with MuSCs preceding proliferation (Ratnayake et al., 2021). This represents what the authors propose as a transcriptionally distinct and tissue-specific macrophage population with transient involvement in forming the proliferative stem cell niche in muscle. Coupled together, these observations suggest that direct interactions drive muscle precursor cells from stem cell to maturing fibre, and that direct interactions may influence stem/progenitor cell maturation. In addition, previously described work in zebrafish HSPC maturation demonstrates that stem cells in niches contacted by macrophages undergo long-term direct contact facilitated by calreticulin, a chaperone protein involved in cellular adhesion, the expression of which can both modulate and be modulated by other adhesion molecules (Goicoechea et al., 2000; Papp et al., 2007; Liu et al., 2016). These calreticulin-mediated interactions often result in the direct sampling of cytoplasmic contents of the stem cell (Wattrus et al., 2022). This, coupled with evidence of possible hemifusion between macrophages and stem cells (Ceafalan et al., 2018; Ratnayake et al., 2021), indicates a possible mechanism for quality control or communication, although this phenomenon has not been widely observed and still remains unexplored. Overall, cell-cell contact appears to be a driver of quality assurance and subsequent directed clonality of stem cells, which may also be a mechanism used during regeneration.

A role for cell-cell contact in stem cell reactivity to macrophages has also been suggested in vitro; whereas MSC-macrophage co-cultures using transwell chamber seeding (in which there is no direct cell-cell contact) are able to increase the expression of the anti-inflammatory factor TNF stimulated gene 6 (TSG6) in MSCs, cell-cell contact seeding systems lead to a greater increase in TSG6 expression facilitated by CD200-CD200R adhesion between MSCs and macrophages, respectively (Fig. 2C) (Li et al., 2019). TSG6 has broad roles in protecting tissues against the effects of inflammation and injury, including altering HA dynamics and orchestration of the immunomodulatory effects of MSCs (Day and Milner, 2019). In addition, increases in TSG6 promotes pro-inflammatory-to-pro-regenerative type switching, as well as subsequent inhibition of CD4⁺ T cell proliferation, thus driving the stem cell niche towards an anti-inflammatory state. TSG6 release by MSCs has also been implicated in driving the pro-regenerative response by decreasing nuclear localization of NF-κB in macrophages (Choi et al., 2011). Thus, direct cell-cell contact between macrophages and stem cells further exemplifies their bidirectional communication, driving greater stem cell activation while simultaneously allowing stem cells to have increased control over their niche. Similarly, MSCs in a cell-cell contact co-culture system with macrophages show enhanced MSC-driven immunosuppression by modifying the balance of CD4⁺ T helper cells (Th cells) and shifting away from a pro-inflammatory environment (Espagnolle et al., 2017). However, this immunosuppressive capability is lost in MSC-macrophage transwell cultures, highlighting the need for direct cell-cell contact. In addition, increased intercellular adhesion molecule 1 (ICAM1) accumulation at the MSC-macrophage interaction site is observed, similar to what is observed at typical immunogenic synapses, where a combination of adhesion molecule interactions and receptor/ligand interactions are required to enhance cell-cell communication (Goldstein et al., 2000; Bromley and Dustin, 2002) (Fig. 2C). How these direct interactions trigger downstream events remains unclear. Nevertheless, close interactions may exist as a mechanism for macrophages to increase the likelihood of signals reaching the cell of interest and forming communication synapses to direct stem cell fate.

It has been observed that MSCs can interact and influence macrophages through the process of mitochondrial transfer (reviewed by Mohammadalipour et al., 2020). Mitochondrial transfer as a cell-cell interactive mechanism was first observed between endothelial progenitor cells and cardiomyocytes as a potential contributor to cell fate determination (Koyanagi et al., 2005), but has subsequently been observed as a mechanism highly used by MSCs to support damaged cells (Mohammadalipour et al., 2020). This can occur through the formation of transient, intercellular tunnelling nanotubes via which organelle transfer can occur (Koyanagi et al., 2005; Jackson et al., 2016), or via the release of mitochondria into microvesicles that are subsequently engulfed (Islam et al., 2012; Morrison et al., 2017; D'Souza et al., 2021; Peruzzotti-Jametti et al., 2021) (Fig. 2C). This is an observed mechanism used to alter macrophage phenotype, whereby MSCs undergo mitochondrial transfer with alveolar macrophages in order to increase their phagocytic behaviour and aid in bacterial clearance (Jackson et al., 2016), as well as promote polarization towards a pro-regenerative phenotype (Morrison et al., 2017). Mitochondrial transfer-mediated polarization of pro-regenerative macrophages by MSCs has also been observed in response to diabetic nephropathy (Yuan et al., 2021). This was observed to require the function of peroxisome proliferator-activated receptor gamma coactivator 1 α (PGC1α), the master regulator of mitochondrial biogenesis (Puigserver et al., 1998), whereupon stimulated mitochondrial transfer in cultured macrophages (RAW 264.7 cells) there was an increase in PGC1α, an increase in biogenesis and autophagy of damaged mitochondria (Yuan et al., 2021). In addition, mitochondria transfer has been identified as a mechanism in white adipose tissue (WAT) homeostasis, where transfer occurs from adipocytes to transcriptionally-distinct macrophages in order to maintain metabolic health (Brestoff et al., 2021). Interestingly, this transfer was heparin sulphate (HS)-dependent (Brestoff et al., 2021), with HS being a key modulator of macrophage polarization cues and known to display dysregulated synthesis in chronic inflammation models (Swart and Troeberg, 2019). Similarly, this dysregulation was observed in obesity models that had decreased HS levels and decreased mitochondrial transfer rates (Brestoff et al., 2021). Dietary lipids, specifically long chain fatty acids, inhibit mitochondrial transfer into macrophages in WAT, instead increasing the level of circulating adipocyte mitochondria in the blood and leading to a systemic distribution of oxidatively-stressed WAT mitochondria in distal tissues such as the heart (Borcherding et al., 2022). Here, the presence of reactive oxygen species (ROS) in the damaged mitochondria triggers the upregulation of antioxidants in the heart to protect against acute oxidative stress (Crewe et al., 2021). This suggests that in healthy metabolic homeostasis, macrophages take up mitochondria from adipocytes as a means to sample and identify metabolic health, and upon nutrient stress these mitochondria are released systemically as a protective mechanism. Overall, intercellular organelle transfer between resident tissue cells or MSCs and macrophages appears to be a mechanism to influence macrophage phenotype directly and is observed as a cell-cell interactive response, as well as a mechanism used for systemic metabolic homeostasis.

Niche stem cells have been observed to change their metabolic state as a result of the metabolic state of neighbouring cells, through either competition for metabolites or through direct cell-cell contact (Wenes et al., 2016; Golan et al., 2020). There has recently been an extensive review of metabolic changes that occur in stem cells as they mature, indicating that these metabolic shifts not only occur during but are also required for stem cell responses (Meacham et al., 2022). This has been shown in various adult stem cell populations, including follicle hair cells, where increasing glycolysis is required for their proliferation and this reliance on glycolysis allows rapid shifts from quiescence to activation (Flores et al., 2017). However, these metabolic shifts have mostly been investigated in HSCs, where increased mitochondrial activity (due to an increased reliance on glycolysis) is correlated with increased ROS levels (Liang et al., 2020). In addition to regulating HSC activation, ROS levels have been shown to dictate HSC fate, with low or medium ROS levels required for maintenance of the stem cell pool and excessive levels leading to apoptosis (Ludin et al., 2014). Recently, however, the function of ROS in determining HSC fate has been widened, as high ROS levels within HSCs have been correlated with the ‘dooming’ response by macrophages, suggesting that macrophages sense and select against HSCs with potential DNA damage as a result of high ROS levels (Wattrus et al., 2022). This suggests a mechanism by which stem cells can direct the macrophage response directly based on their own metabolic state.

Conversely, stem cells are able to directly modulate the metabolic state of macrophages by inhibiting glycolysis or driving oxidative phosphorylation (OXPHOS) (Peruzzotti-Jametti et al., 2018; Deng et al., 2020), which may represent a further mechanism of macrophage-stem cell crosstalk (Fig. 2D). Metabolic changes in macrophages have been well documented, with a shift from a reliance on glycolysis towards OXPHOS necessary for the pro-inflammatory-to-pro-regenerative transition (reviewed by Liu et al., 2021). In addition, the activity of glutamine synthetase, which is required for the conversion of glutamine to glutamate, is necessary for the appropriate pro-inflammatory-to-pro-regenerative switch (Palmieri et al., 2017). Glutamine has been observed to drive proliferation and increase release of anti-inflammatory cytokines in MSCs (dos Santos et al., 2017). In low glutamine environments, macrophages secrete glutamine rather than feeding it into the tricarboxylic acid (TCA) cycle and, in turn, drive the proliferation and differentiation of satellite cells (Shang et al., 2020) (Fig. 2D). This process is heavily reliant on increased activity of glutamine synthetase and is at the expense of the TCA cycle. Thus, macrophages are implicated in sensing and regulating the regenerative milieu, possibly at the expense of its own increased metabolic needs. Similarly, macrophages have been observed to release succinate, which is taken up by NSCs to drive release of PGE2, which in turn suppresses the release of inflammatory cytokines by macrophages (Peruzzotti-Jametti et al., 2018). Succinate, an intermediate of the TCA cycle, has historically been considered a pro-inflammatory factor by both directing autocrine signalling of macrophages to increase IL1β expression, but also as a result of the accumulation of ROS (Mills et al., 2016), indicating that the NSC-macrophage crosstalk is specific in driving this phenotypic switch (Fig. 2D). Although the authors did not investigate the influence of increased intracellular succinate levels on metabolism in the NSC, there is the potential that this acts as an additional point of macrophage-stem cell communication to drive stem cell maturation.

Macrophages are emerging as an essential modulator of stem cell niches, with evidence showing their involvement in development and regeneration. There is also an increasing understanding of the involvement of crosstalk in these processes, with modulation of macrophage responses and inflammatory environments by stem cells themselves. In addition, the advent of both single cell RNA-sequencing technology and improvements with in vivo live imaging capabilities have driven greater understanding in the heterogeneity of these populations both transcriptionally and behaviourally. There is a developing understanding of the role of macrophages in direct cell-cell interaction with stem cells that might hint to a greater regulation within the niche, including direct sampling of stem cells to drive quality assurance. By viewing macrophages through a lens beyond their role in inflammation, we are now beginning to understand the dynamic control mechanisms macrophages deploy with the stem cell niche at both a molecular and cellular level. It is clear that the multifunctional role that macrophages play in tissue repair make them not only the ‘gatekeepers’ but also the ‘conductors’ of regeneration. Through this lens, and by increasing our understanding of both macrophage heterogeneity and plasticity, as well as what drives specificity, we can garner a greater understanding of how to specifically target stem cell responses in vivo to support regeneration and repair. However, given the increasing evidence suggesting macrophage contact with stem cells enhances responses and trophic effects, this presents challenges with in vivo targeting and treatment. By asking what behavioural and specific contact roles are required in these processes we can improve our understanding of how to modulate these effects to increase targeted stem cell responses by macrophages.

We thank Graham Lieschke, Tracy Heng and Mikael Martino for expert comments and critique of the manuscript.