Key Role of Mesenchymal Stromal Cell Interaction with Macrophages in Promoting Repair of Lung Injury

Lung macrophages (Mφs) are essential for pulmonary innate immunity and host defense due to their dynamic polarization and phenotype shifts. Mesenchymal stromal cells (MSCs) have secretory, immunomodulatory, and tissue-reparative properties and have shown promise in acute and chronic inflammatory lung diseases and in COVID-19. Many beneficial effects of MSCs are mediated through their interaction with resident alveolar and pulmonary interstitial Mφs. Bidirectional MSC-Mφ communication is achieved through direct contact, soluble factor secretion/activation, and organelle transfer. The lung microenvironment facilitates MSC secretion of factors that result in Mφ polarization towards an immunosuppressive M2-like phenotype for the restoration of tissue homeostasis. M2-like Mφ in turn can affect the MSC immune regulatory function in MSC engraftment and tissue reparatory effects. This review article highlights the mechanisms of crosstalk between MSCs and Mφs and the potential role of their interaction in lung repair in inflammatory lung diseases.


Introduction
Mesenchymal stromal cells (MSCs) are now known to enhance the repair of injured tissues and are emerging as possible therapeutic agents in acute and chronic inflammatory lung diseases and in COVID-19. The interaction between MSCs and Mϕs has been shown to be a crucial mechanism of such beneficial action in lung injury. Our review focuses on this fascinating crosstalk. We describe the current understanding of the mechanisms and complexity of the MSC-Mϕ interplay. We also point out the numerous gaps in our knowledge that still hinder exploiting this mechanism for therapeutic purposes. The final outcome of MSC-Mϕ interaction depends on multiple factors, and a better understanding of these is essential for designing strategies to improve the efficacy of MSC treatment and maximize positive and avoid possible negative outcomes. We also provide an overview of the potential relevance of the MSC-Mϕ interplay, as emerging from preclinical and clinical studies. Since the therapeutic efficacy of MSCs is still clinically unproven, this knowledge is crucial for personalized therapy for lung injury patients.

Pulmonary Mϕs as a Host Defense Mechanism
The lung is not only vital for gas exchange but also serves as a major immune organ that protects the host from inhaled pathogens, allergens, and toxins [1]. Since pulmonary Mϕs are fundamental for the regulation of tissue homeostasis, modulation of their function could be used to prevent injury and promote repair both in acute and chronic lung injury.

Role of Mϕ Plasticity in Lung Injury and Repair
During acute inflammation, damage-associated molecular patterns (DAMPs), cytokines, and growth factors (GM-CSF and M-CSF) [18] induce the recruitment and differentiation of residual yolk-sac-derived pulmonary Mϕs and promote their interaction with surrounding cells. Bone-marrow-derived monocytes are also recruited to the lung and differentiate into AMϕs upon inflammatory stimuli [19]. Cytokines and chemokines, especially CCL2 (also referred to as monocyte chemoattractant protein 1-MCP1) and its receptor CCR2, play a crucial role in Mϕ recruitment and polarization [20].
The activation and polarization of Mϕs into two extreme phenotypes, M1 (proinflammatory) and M2 (involved in the resolution of inflammation and tissue repair), was described in the early 1990s [21]. M1-like Mϕs start and sustain inflammatory responses, while M2-like Mϕs promote the resolution of inflammation and coordinate the restoration of tissue integrity ( Figure 1).
Importantly, Mϕ responses to environmental challenges are complex, and their functional polarization into M1 and M2 types is an over-simplification. Rather than being an "on-off" process, current data clearly indicate that Mϕ polarization is a continuum along the whole spectrum between M1 and M2 phenotypes. Therefore, terms such as M1-like and M2-like Mϕs more appropriately describe the phenotypes [28].
Of note, neither AMϕs nor PIntMϕs can be defined exclusively by M1 or M2 markers, and in the healthy lung both populations co-express markers historically considered as M1or M2-specific [12,29,30]. istered to patients with respiratory diseases, in part because they would likely elici immune response. Rather, therapeutic strategies mainly focus on harnessing the plasti of existing pulmonary Mφs.
In this respect, mesenchymal stromal cell (MSC) treatment offers an opportunit interact with host macrophage populations and modulate their function and vice ve as described in the next sections. Lung tissue macrophages (Mφs), alveolar (AMφs) and interstitial (PIntMφs), are sential for host defense. AMφs eliminate the small daily load of aspirated microbes regulate pulmonary surfactant. PIntMφs have anti-inflammatory and regulatory r [1,2].
Inflammation attracts circulating Mφs (by CCL-2) [20] and causes Mφ prolifera (induced by GM-CSF) [18,19]. Pathogen sensors, TLR-4, and other TLRs cause NF-κB The above-described plasticity renders Mϕs attractive candidates for therapeutic purposes, as environmental influences can dynamically and reversibly alter their phenotype [23,28]. For example, our group has successfully used pluripotent stem cell-derived Mϕs in rats to treat systemic sepsis with multi-organ involvement that includes lung injury [31]. Indeed, Mϕ transplantation therapy may ultimately be used as a therapeutic approach for lung injury, but its feasibility and effects in humans have yet to be investigated. So far, Mϕs have not typically been considered as cells that can be directly administered to patients with respiratory diseases, in part because they would likely elicit an immune response. Rather, therapeutic strategies mainly focus on harnessing the plasticity of existing pulmonary Mϕs.
In this respect, mesenchymal stromal cell (MSC) treatment offers an opportunity to interact with host macrophage populations and modulate their function and vice versa, as described in the next sections.

Preclinical Studies-MSC Action, Licensing, and Genetic Modifications
Acute respiratory distress syndrome (ARDS) is the leading cause of morbidity and mortality (30-50%) in the critically ill receiving supportive treatment. Despite some improvements, there is still no causal therapy available [41,42]. The hallmarks of ARDS include alveolar epithelial-capillary barrier disruption, consequent edema formation, and widespread uncontrolled lung inflammation. In preclinical studies (Table 1), the treatment of ARDS rodent models with MSCs [43] or with MSC-derived extracellular vesicles (EVs) significantly mitigated lung injury [44]. Interestingly, EVs isolated from young MSCs were more effective [45]. Studies also demonstrated that MSCs or MSC-EVs acted by decreasing the production of pro-inflammatory cytokines by AMϕs [44] and/or by inducing alveolarendothelial barrier restoration partially via mitochondrial transfer [46]. Administration of MSCs protected rodents from ventilator-induced lung injury [47,48], with MSCs being equally [47] or more effective than their secretome [49]. MSCs were also found to be beneficial in mice with sepsis-induced lung injury. In these models, the improvement was ascribed to the downregulation of miR-27a-5p microRNA [50] and microRNA (miR)-193b-5p [51] in the septic lung and upregulation of their respective target genes, i.e., VAV3 and the tight junctional protein occludin.
Several studies demonstrated that modification of MSCs can augment their beneficial effects. The immune-modulatory and tissue-reparatory properties of MSCs can be enhanced by pre-stimulation, i.e., licensing or by their transfection with targeted genes prior to the treatment. Cytokine-induced pre-activation (with IL-1β, TNF-α, IFN-γ) augmented MSCinduced repair and resolution of ventilator-induced lung injury [52]. EVs from interferon (IFN)-γ-primed human umbilical cord (hUC)-MSCs were more beneficial in E. coli-induced lung injury in rats than EVs from naïve MSCs [53]. MSCs cultured in hypoxic conditions displayed more protection in radiation-induced lung injury in mice by promoting MSC viability and improving their antioxidant capacity [54]. A hypoxic environment promoted EV release by MSCs and enhanced their potency in suppressing airway inflammation in asthmatic mice [55]. Hyperthermia increased the efficacy of MSC-driven immune suppression [56].
Genetic modification of MSCs can be achieved by overexpression or silencing of specific genes using different knock-in and knock-out technologies including CRISPR (clustered regularly interspaced short palindromic repeats)/Cas9 gene-editing system, RNA interference technology, etc. These manipulations could be used to control native MSC gene expression or introduce foreign genes for specific therapeutic applications (as reviewed in detail in Varkouhi et al., 2020 [57]).
Our group has shown that overexpression of IL-10 enhanced the efficacy of hUC-MSCs in E. coli pneumosepsis in rats [58]. However, of note, in acid-primed lung injury associated with the development of fibrosis, MSC treatment was harmful. Correction of the microenvironment after acid-primed lung injury or treatment with MSCs carrying the human IL-10 gene or hepatocyte growth factor (HGF) reversed the detrimental effects of naïve MSCs [48]. Therefore, the lung microenvironment, disease type, and severity have to be taken into account as they affect MSC activation, function, and therapeutic effectiveness in both acute and chronic lung diseases [59,60].

Preclinical Model Intervention (MSCs/EVs) Outcome Mechanism References
Radiation-induced pneumonia and late fibrosis in mice.
Murine BM-MSCs cultured in normoxic and hypoxic environment.
Therapeutic effect of MSCs exposed to hypoxia was more pronounced compared to MSCs exposed to normoxia.
Hypoxia-treated MSCs were more viable and resistant to hypoxia decreasing oxidative stress in lungs by HIF1-α. hUC-MSCs-derived EVs from MSCs cultured in normoxic (Nor-EVs) and hypoxic (Hypo-EVs) conditions.
Hypo-EVs were more effective than Nor-EVs in attenuation of chronic asthma. Regardless of whether MSCs or their derivatives were used in the studies, their interplay with monocytes/macrophages is a key mechanism of therapeutic benefits [66,67]. As summarized in Table 1, most studies found Mϕ involvement in the beneficial effects of MSCs in lung injury or described a role for alveolar IL-10 increase in the effect. It should be noted that IL-10 secretion mainly originates from Mϕs or neutrophils and is involved in Mϕ polarization towards an M2-like phenotype [68,69].
Co-culture experiments provide direct evidence for a role of MSC/Mϕ interaction. MSCs were shown to enhance Mϕ phagocytic [70,71] and anti-inflammatory capacity [72]. Moreover, MSC-derived exosomes inhibited M1 and promoted M2 polarization in LPSstimulated Mϕs [73], and MSCs primed with Mϕ-derived conditioned media exhibited enhanced immunomodulatory potential [74]. Further evidence for a key role of the MSC/Mϕ interaction was obtained in in vivo lung disease models. Specifically, systemic Mϕ de-pletion experiments showed that both monocyte recruitment to the lungs and MSC/Mϕ interaction are crucial for the favorable MSC effects in pulmonary diseases. Selective AMϕs depletion reversed the therapeutic benefits of MSC treatment in a mouse model of allergic asthma [75] and in an E. coli ARDS mouse model [70]. Treatment of mice with BM-MSCs prevented the development of obliterative bronchiolitis after tracheal allografts, and this effect was eliminated by systemic depletion of Mϕs [76]. Systematic depletion of Mϕs also weakened the therapeutic effect of MSC-derived exosomes or MSCs in mouse models of severe asthma [73] and allergic airway inflammation [77], respectively.
While the general role of the MSC/Mϕ interaction is now well documented, the details of this interplay remain poorly defined. A better understanding of this crucial interaction is essential since the therapeutic efficacy of MSCs in ARDS and other lung diseases remains unproven. More mechanistic insights into the MSC/pulmonary Mϕ interaction will be needed for the optimization and personalization of MSCs in respiratory diseases.

Crosstalk between MSCs and Mϕs-Mechanisms of Action
The MSC-Mϕ crosstalk is now known to affect both cell types. Several mechanisms have been implicated as a vital mediator of these effects (see Figure 2). Most studies point out that administering live MSCs is much more effective therapeutically, although some studies have shown immunomodulatory effects and benefits of apoptotic MSCs instilled into septic animals [78,79]. Thus, improving MSC viability was seen as an important factor in augmenting cell efficacy in clinical trials for moderate to severe ARDS [80] precisely because the MSC-Mϕ dyad is the key for the full spectrum of MSC action [66,[81][82][83].
Combined, these studies provide evidence that EVs and their cargo are indeed key mediators of the effects of MSCs. An improved understanding of the role of the EV cargo will be a key aspect of improving the design of MSC therapy.

Role of Autophagy, Mitophagy, and Oxidative Stress in MSC-Mφ Interplay
Autophagy, the main cellular mechanism for degrading and recycling intracellular proteins and organelles, plays an important role in maintaining bioenergetic homeostasis in health and disease and may contribute to the therapeutic action of MSCs [158]. The modulation of autophagy may also change Mφ efficiency and polarization. Indeed, BM-MSCs were shown to exert beneficial effects in a mouse model of sepsis acting primarily by enhancing mitophagy in Mφs and decreasing mitochondrial ROS, thus inhibiting NLRP3 inflammasome activation [159]. The activation of autophagy, HO-1, and mitochondrial biogenesis occurs after MSC exposure to Mt isolated from somatic cells. During tissue injury, MSCs are prompted by Mt released from damaged cells to donate Mt to injured cells, thereby enhancing tissue reparation [129].
Oxidative stress, frequently present in inflamed and damaged tissue, is also known to activate autophagic processes [160,161]. BM-MSCs are able to modulate autophagy in Mφs through the PI3K/Akt/HO-1 signaling pathway [162] and protect rats against liver I/R injury via the promotion of HO-1-mediated autophagy [136].
In an MSC-Mφ co-culture system and mouse model of silicosis [163], MSCs reduced intracellular oxidative stress. This was attributed to targeting depolarized Mt and releasing MVs containing entire Mt selected for mitophagy, lysosomes, and several miRNAs (especially miR451a). The vesicles formed by this process are then engulfed and reutilized by Mφs causing the repression of TLR and NF-κB signaling in Mφs and decreasing the production of inflammatory and pro-fibrotic mediators.  Many studies found a bidirectional effect of direct MSC-Mϕ contact, affecting both cell types, that is important for the beneficial effects. Direct contact between MSCs and pro-inflammatory Mϕs has been shown to reinforce tumor necrosis factor-stimulated gene-6 (TSG-6) production by MSCs, further promoting Mϕs switching to an M2-like phenotype and suppressing T cell proliferation [84], and to lead to upregulation of CD200. Interaction with CD200R on Mϕs in turn facilitated the reprogramming of Mϕs towards an anti-inflammatory phenotype [84].

Microtubular Network
Recent studies have shown that the transfer of MSC extracellular vesicles (EVs) and mitochondria (Mt) to Mϕs could happen via a tunneling nanotubule formation that established direct MSC-Mϕ contact [70,85].
Mitochondrial transfer to injured cells [86] or Mϕs through nanotubes seems to be facilitated by mitochondrial Rho-GTPases [87,88] enhancing cell oxidative phosphorylation and Mϕ phagocytosis in in vitro and in vivo models of ARDS [70,85,89].

MSCs and Mϕ Secretome-Paracrine-Mediated Mechanisms
The majority of evidence to date supports the notion that the effects of MSCs on Mϕs or donor cells are largely paracrine, i.e., through secreted factors [90].

COX/PGE2/EP4 Axis
The secretion of prostaglandin (PG)E2 by MSCs serves as a crucial regulator of MSC/Mϕ interaction controlling inflammation and tissue homeostasis, repair, and regeneration [124].
PGE2 is derived from arachidonic acid by cyclooxygenase synthases (constitutively active COX1 and inducible COX2) and PGE synthases [140]. The studies by Németh and colleagues [123] have revealed that BM-MSCs, especially if activated by LPS or TNF-α, release PGE2 that acts on the Mϕs through PG receptors (EP2 and EP4) triggering IL-10 secretion. The role of PGE2 and the mechanism involving IL-6 and IL-10 was later confirmed and elaborated by many other groups in a variety of injury and disease models [122,133,[141][142][143].
Naïve UC-MSCs increase HO-1 expression and phagocytic capabilities in hMϕs. This effect is augmented by transfection of MSCs with IL-10 and is abolished by PGE2 and lipoxygenase A4 blockade. This mechanism was also verified in a co-culture of MSCs and Mϕs isolated from rat lungs with induced E. coli pneumosepsis [58] and from the peritoneal cavity of septic rats [34].
Vascular growth factor (VEGF) and angiopoietin-1 (Ang-1), secreted by MSCs or present on MSC-EVs in the form of mRNA, have been found to be important not only for the restoration of vascular stability but also for the repair of acute lung injury [109] or resolution of allergic asthma in mice [108] through Mϕ immunomodulation.

Role of MSC-Derived Extracellular Vesicles (EVs), mRNA, MicroRNA, and Mitochondrial Transfer in Immunomodulation through Mϕs
MSCs have been known to release EVs that contain a variety of cargos including endosomal and plasma membrane, intracellular organelles (e.g., mitochondria-Mt), cytokines, growth factors, signaling lipids, mRNAs, and regulatory miRNAs [146]. EVs released by MSCs might be as effective in therapy as whole MSCs. Thus, MSC-EVs represent an appealing option for cell-free regenerative medicine [147] as their content could be delivered to immune cells present in an inflammatory environment (M1-like Mϕs, dendritic, CD4+Th1, and Th17 cells), promoting their phenotypic switch into immunosuppressive M2-like Mϕs, tolerogenic DCs, and regulatory T cells [91]. Moreover, EVs could be exploited as attractive tools for diagnostic and therapeutic agent delivery [148]. A meta-analysis screening 52 articles [149] demonstrates the clear potential of MSC-EVs as a therapeutic tool for acute and chronic lung diseases in particular. MSC-EVs or exosomes are already being used in a few dozen clinical trials in [150] and [151], including in some COVID-19 pneumonia trials.
Several studies demonstrated that the mechanism of the beneficial effects of MSC-EVs are executed, at least in part, through their interaction with Mϕ.
An in vitro co-culture of MSCs with Mϕs enhanced their M2-like polarization mainly through MSC-derived exosomes [126,152]. Infusing MSCs lacking exosomes led to a lower number of M2-like Mϕs in vivo. Further, the hyperoxia-induced reduction in CD163positive Mϕs was prevented by MSC-EVs in a rat model of bronchopulmonary dysplasia [64]. In an ALI E. coli endotoxin mouse model, instilled MSC-EVs reduced pulmonary edema and lung inflammation by decreasing Mϕ inflammatory protein-2 levels in the BAL fluid [153].
Exciting data suggest that mitochondrial transfer from MSC-EVs to recipient cells, including Mϕs, is an important mechanism for enhancing Mϕ anti-inflammatory and regenerative capacity in injury and inflammation. Using a co-culture system consisting of MSCs and Mϕs, Yuan et al. [155] showed that MSC-derived Mt were transferred into Mϕs, which contributed to their M2 polarization. The beneficial outcome of ALI in mice subjected to the adoptive transfer of AMϕs pretreated with MSC-EVs was dependent on EV-mediated mitochondrial transfer [44].
In addition to the above-described cargo, different MSC-MV-derived microRNAs were also found to be important mediators in Mϕ switching towards an M2-like phenotype, including miR-223 [126] and miR-182 [152]. The alleviation of ARDS in mice was found to be mediated by miR181 [156] while miR-466 contributed to MSC-EV-induced improvement in a multidrug-resistant pseudomonas aeruginosa pneumonia mouse model [157].
Combined, these studies provide evidence that EVs and their cargo are indeed key mediators of the effects of MSCs. An improved understanding of the role of the EV cargo will be a key aspect of improving the design of MSC therapy.

Role of Autophagy, Mitophagy, and Oxidative Stress in MSC-Mϕ Interplay
Autophagy, the main cellular mechanism for degrading and recycling intracellular proteins and organelles, plays an important role in maintaining bioenergetic homeostasis in health and disease and may contribute to the therapeutic action of MSCs [158]. The modulation of autophagy may also change Mϕ efficiency and polarization. Indeed, BM-MSCs were shown to exert beneficial effects in a mouse model of sepsis acting primarily by enhancing mitophagy in Mϕs and decreasing mitochondrial ROS, thus inhibiting NLRP3 inflammasome activation [159]. The activation of autophagy, HO-1, and mitochondrial biogenesis occurs after MSC exposure to Mt isolated from somatic cells. During tissue injury, MSCs are prompted by Mt released from damaged cells to donate Mt to injured cells, thereby enhancing tissue reparation [129].
Oxidative stress, frequently present in inflamed and damaged tissue, is also known to activate autophagic processes [160,161]. BM-MSCs are able to modulate autophagy in Mϕs through the PI3K/Akt/HO-1 signaling pathway [162] and protect rats against liver I/R injury via the promotion of HO-1-mediated autophagy [136].
In an MSC-Mϕ co-culture system and mouse model of silicosis [163], MSCs reduced intracellular oxidative stress. This was attributed to targeting depolarized Mt and releasing MVs containing entire Mt selected for mitophagy, lysosomes, and several miRNAs (especially miR451a). The vesicles formed by this process are then engulfed and reutilized by Mϕs causing the repression of TLR and NF-κB signaling in Mϕs and decreasing the production of inflammatory and pro-fibrotic mediators.

Therapeutic Potential of MSC-Mϕ Interaction and Lung Injury Resolution
The MSC-Mϕ interplay has a pivotal role in lung injury combating inflammation and promoting injury resolution [66,71,72,95,111,167]. For tissue recovery, crucial effects include the dampening of inflammation, release of IL-10 by Mϕs, and favourable effects of eicosapentaenoic acid (EPA) and its derivatives, resolvins (Rv-D1, E1, and E2) and protectins secreted by MSCs [125]. EPA also gives rise to eicosanoids and their metabolites PGE2 and lipoxin A4 in MSCs [58,97,123]. These are key for IL-10 induction and enhanced Mϕ phagocytosis, contributing to the resolution of inflammation and tissue recovery. EPA pre-conditioning of MSCs further reduces sepsis-induced lung injury and leads to faster recovery [125].

Reparatory Potential of MSC-Mϕ Interaction in Chronic Lung Diseases
The reparative effect of MSCs and MSC-Mϕ interaction allowed their use not only in acute but also in chronic lung injury [61,62,149,168,169], as mentioned before. MSCs or their secreted factors and EVs were found to be beneficial in many models of chronic lung injury including in allergic, ragweed, or ovalbumin-induced asthma models [55,170,171]. In asthmatic mice, TGF-β production by MSCs increased the presence of regulatory Mϕs and T cells, which helped restore cytokine balance and prevented harmful allergic responses [172,173]. In bronchopulmonary dysplasia (BPD) mouse models, MSC-EVs reduced lung injuries partly by increasing M2-like interstitial/alveolar Mϕ polarization and their anti-inflammatory and anti-proliferative action [64,130]. In COPD, MSCs act by attenuating the airway infiltration of neutrophils and Mϕs, leading to decreased production of IL-1β and IL-6 while increasing IL-10 and the levels of growth factors (VEGF, HGF, EGF, TGF-β), therefore boosting tissue repair [63]. Similarly, in models for idiopathic pulmonary fibrosis (IPF), MSCs or MSC-derived EVs could attenuate lung fibrosis by acting on Mϕs by promoting ATII cell proliferation and by inhibiting lung fibroblast proliferation [65,131]. MSC-derived EVs were also able to prevent or reverse lung fibrosis in bleomycin-treated mice by modulating pulmonary Mϕ phenotypes, shifting them to an immunoregulatory and anti-inflammatory phenotype [131].
However, collecting more data on the mechanisms of MSC action and MSC-Mϕ interaction in clinical studies and particularly in chronic lung disease is imperative. Randomized controlled trials with large cohorts of patients and with a mechanistic approach are needed. Clinical trials in IPF, asthma, silicosis, and COPD conducted so far are mainly oriented to assessing the safety of the treatment, with single or two MSC doses and mostly focused on the short-term effects of therapy [62,[174][175][176][177].
Despite the encouraging results from these studies, special attention needs to be paid to the possibility of worsening lung function and fibrosis caused by the administration of MSC in the chronic stage of IPF [178].

Clinical Studies
MSCs are being tested in clinical trials in critically ill patients with sepsis, ARDS, COVID-19 pneumonia, ARDS with malignances, pulmonary emphysema, pulmonary arterial hypertension, aplastic anemia, cancer, and liver cirrhosis, as well as for the prevention of graft vs. host disease (GVHD), etc. The treatments assessed involve autologous or allogeneic BM-MSCs, umbilical cord MSCs, or exosomes derived from MSCs. Moreover, allogeneic MSCs are also safely used in some clinical studies including trials for pulmonary emphysema and patients with COVID-19 ARDS.

MSC-Mϕ Interaction in COVID-19-More Studies Are Needed
In addition to the above described effects, antiviral properties have also been attributed to MSCs and MSC-EVs. MSCs could inhibit virus replication and shedding through IDO and LL37 secretion and improve influenza-induced viral pneumonia and ARDS due to antiinflammatory and reparative potential, partly through the interaction with Mϕ [182,183].
SARS-CoV2 enters cells, including alveolar epithelial and lung capillary endothelial cells, through the widely expressed angiotensin-converting enzyme 2 (ACE2) receptor [184]. MSCs are ACE2-negative, and therefore they are resistant to SARS-CoV-2 infection and retain their immunomodulatory activities when encountering the virus [185,186]. This important feature favors MSC therapeutic use in COVID-19 [187]. Interaction with innate immune cells, especially Mϕs, and molecules secreted by MSCs and Mϕs, described above, could dampen immune hyper-activation in SARS-CoV-2 infection. Further, MSCs also have anti-fibrotic and lung reparatory effects [188]. Thus, MSCs or MSC-derived EVs could be used in a potential supportive and/or curative strategy for COVID-19, especially in critically ill patients [127,186,189]. So far, dozens of clinical trials have been performed in COVID-19 patients with severe disease [190]. Since most of these are still ongoing, limited preliminary data are available [189]. The results so far from studies registered on the US ClinicalTrials.gov page (Table 3) are promising, showing the absence of adverse effects in COVID-19 patients treated with MSCs or MSC-EVs [191][192][193], reduced expression of pro-inflammatory cytokines [194], and improved recovery time [195]. Some studies from non-US goverment pages (China, Iran) showed an increase in IL-10 expression with improved lung function outcomes [196] and, mortality, as well as positive effects on chest imaging results [197] and oxygen saturation in the MSC-treated COVID-19 group [195,198].

Discussion and Conclusions
The overall objective of this review was to provide an overview of up-to-date knowledge of MSC-Mϕ crosstalk and its importance in acute and chronic pulmonary disease treatment and resolution of lung injury. The beneficial effects are thought to be due to bidirectional MSC-Mϕ communication, which is attributed to direct contact, soluble factor secretion/activation, and organelle transfer. Further, MSC-derived EVs represent an appealing option for cell-free regenerative medicine. Their delivery in a pro-inflammatory environment dominated by M1-like Mϕs could promote Mϕ transition towards immunosuppressive and pro-reparatory M2-like cells. Moreover, EVs could be used as an attractive tool for diagnostic and therapeutic agent delivery.
It is important to note that the outcome of the MSC-Mϕ interaction depends on many factors, and the final outcome is not always beneficial. The environment, disease stage, and other components of innate and adaptive immunity could all influence the MSC-Mϕ interplay and its consequences. More research is clearly needed to define these factors in specific diseases and disease stages and to make this interaction predictable and modifiable, assuring favorable outcomes.
Importantly, the clinical use of MSCs is still restricted, and clinical trials conducted so far have mainly investigated the safety of MSC-based therapy in acute and chronic lung diseases and in patients with diseases resistant to other therapeutic options. This clearly precludes the immediate use of MSCs in lung diseases, even though novel therapeutic approaches are desperately needed. More critically, there is no effective treatment for ARDS, a condition that develops and progresses rapidly and has high mortality [200]. For this reason, most of the clinical studies exploring the use of MSCs or their derivatives are conducted in ARDS patients [200,201]. However, these studies were carried out with a relatively small number of enrolled patients. Therefore, large-scale clinical trials are lacking. The manipulation of MSC-Mϕ interaction described in this review represents an attractive option for the improvement of MSC efficacy in the treatment of lung diseases. However, more detailed studies are needed. These should include cytokine measurements and the isolation of monocytes and Mϕ from plasma, BALF, and patient lung biopsies to characterize Mϕ phenotype, phagocytic properties, and secretion profiles. Moreover, since the effect of MSCs depends on the microenvironment, a better understanding of the composition of the microenvironment in lung diseases is paramount for developing appropriate strategies to control MSC action and MSC-Mϕ crosstalk. Our increasing knowledge and understanding of the mechanisms of action of MSCs and the MSC-Mϕ interaction in particular, together with use of cell-free MSC derivatives (conditioned media, EVs, Mt), could help develop strategies for the effective treatment of lung injury. Ideally, this should happen in the imminent future as there is a substantial unmet medical need for efficacious treatment options.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the writing of the manuscript or in the decision to publish the results.