Potential application of mesenchymal stem cells and their exosomes in lung injury: an emerging therapeutic option for COVID-19 patients

The COVID-19 pandemic has negatively impacted the global public health and the international economy; therefore, there is an urgent need for an effective therapy to treat COVID-19 patients. Mesenchymal stem cells (MSCs) have been proposed as an emerging therapeutic option for the SARS-CoV-2 infection. Recently, numerous clinical trials have been registered to examine the safety and efficacy of different types of MSCs and their exosomes for treating COVID-19 patients, with less published data on the mechanism of action. Although there is no approved effective therapy for COVID-19 as of yet, MSC therapies showed an improvement in the treatment of some COVID-19 patients. MSC’s therapeutic effect is displayed in their ability to reduce the cytokine storm, enhance alveolar fluid clearance, and promote epithelial and endothelial recovery; however, the safest and most effective route of MSC delivery remains unclear. The use of poorly characterized MSC products remains one of the most significant drawbacks of MSC-based therapy, which could theoretically promote the risk for thromboembolism. Optimizing the clinical-grade production of MSCs and establishing a consensus on registered clinical trials based on cell-product characterization and mode of delivery would aid in laying the foundation for a safe and effective therapy in COVID-19. In this review, we shed light on the mechanistic view of MSC therapeutic role based on preclinical and clinical studies on acute lung injury and ARDS; therefore, offering a unique correlation and applicability in COVID-19 patients. We further highlight the challenges and opportunities in the use of MSC-based therapy.

tract symptoms to progressive life-threatening viral pneumonia and progressive hypoxemia requiring mechanical ventilatory support. The leading cause of mortality in COVID-19 patients is hypoxemic respiratory failure most frequently resulting in ARDS, characterized by diffuse lung damage with edema, hemorrhage, and intraalveolar fibrin deposition [6,7]. More interestingly, laboratory findings indicate a hyperactivated nature of the immune system, specifically high levels of circulating CD4 + and CD8 + lymphocytes. Looking at the hyperactive immune response detected in COVID-19 patients, several potential treatments relating to key immunoregulators have been proposed. Another important factor influencing the prognosis of COVID-19 patients is having a state of hyperinflammation, where several immunosuppression modalities have provided a tool to decrease the mortality in patients with severe condition [8]. Understanding the pathogenesis of SARS-CoV-2 in association with the host immune response will help elucidate some key targeted treatment options. Repurposing of previously approved medications, such as the anti-malarial drug hydroxychloroquine, anti-rheumatic drugs, such as tocilizumab (interleukin [IL]-6 receptor inhibitor), baricitinib (Janus kinase [JAK] inhibitor), and anakinra (IL-1 receptor antagonist), have been employed to treat COVID-19, largely attributed to their known pharmacokinetic and safety profiles [9].
Mesenchymal stromal/stem cells (MSCs) offer a promising emerging therapeutic approach toward modifying the adverse effects of the infection in SARS-CoV-2 patients. This therapy has been found to decrease the cytokine storm and exert anti-inflammatory, immunomodulatory, and regenerative functions by altering the expression of pro-inflammatory cytokines, and aid in repairing the damaged tissues in COVID-19 patients. Several clinical trials have already provided a proof of concept showing that intravenous (IV) infusion of MSCs is a safe option and could lead to clinical and immunological improvement in some patients with severe COVID-19 pneumonia [10]. Such findings support employing phase 2 randomized controlled trial, where other randomized trials with a control arm consisting of standard treatment, will help to elucidate the mechanistic potential of MSC-based therapeutic strategy. This review summarizes the immunopathogenesis of the SARS-CoV-2 and the therapeutic potentials of MSCs for treating lung injuries associated with COVID-19. Furthermore, we highlight the current clinical trials using MSCs for treating COVID-19 patients and discuss limitations of the existing MSC-based treatment strategies.

Pathogenesis of SARS-COV-2
The lung alveoli are lined with the alveolar epithelium consisting of a monolayer of alveolar type I (AT1) cells and alveolar type II (AT2) cells. Under normal condition, the AT2 cells secrete surfactant covering all the lining epithelium to facilitate alveolus expansion. AT1 and AT2 are tightly connected with tight junctions, which control the transfer of ions and fluid across the epithelium. The endothelial cells of the blood capillaries are connected by intercellular junctions and control the influx of inflammatory cells and fluid into the interstitial space between the aveoli. Initially, the spike glycoprotein (S protein) expressed on viral envelopes binds to the angiotensin-converting enzyme 2 (ACE2) receptor [11], a very similar structure to that of SARS; however, with a 10-20 times much higher binding affinity when compared to the SARS S protein [12]. This binding capability partially explains the high transmission of SARS-CoV-2 [12]. The main target cells for SARS-CoV-2 infection are AT2 cells and resident alveolar macrophages, because they are expressing ACE2. SARS-CoV-2 utilizes ACE2 for entry and the serine protease TMPRSS2, which is also expressed by the alveolar cells, for S protein priming [13]. This activation induces chemokine and cytokine secretion that recruits inflammatory and immune cells into the infected alveoli, followed by other waves of cytokine release. Activated macrophages have a significant role in hemophagocytic lymphohistiocytosis (HLH)-like cytokine storm during COVID-19 [14]. Secondary HLH could be precipitated by a genetic defect in cytolytic pathways or observed in during infection, malignancy, and rheumatic disease. HLH is characterized by a predominance of inflammatory cytokines and expansion of tissue macrophages displaying hemophagocytic activity [15]. Cytopenias, a state of elevated inflammatory cytokines or hypercytokinaemia, unremitting fever, elevated ferritin level, and multi-organ damage, are among the key characteristics of HLH seen in seriously ill COVID-19 patients [8]. Type I interferons (IFN) and natural killer (NK) cells result in cytolytic immune responses, following a successful recognition of pathogenassociated molecular pattern. This serves as a first line of defense against SARS-CoV-2 infection through the innate immune system. Activated cytotoxic T cells and B cells are key players of the adaptive immunity helping with viral clearance via destruction of virusinfected cells and antibody production, respectively. However, when the anti-viral immune response remains active, an aberrant and uncontrolled production of inflammatory cytokines occurs, causing what is known as the "cytokine storm", leading to damage in the pulmonary tissue [16,17].
Among the histological profiles of COVID-19 are the significant alterations in the morphology of the endothelial cells, which also express ACE2. These changes include damage of the intercellular junctions, a loss of attachment to the basement membrane, and cell swelling [20] (Fig. 1). The migrated neutrophils and monocytederived macrophages release toxic mediators, causing endothelial and epithelial injuries ( Figs. 1 and 2). The intercellular junctions are disrupted leading to formation of spaces between the alveolar cells as well as between the endothelial cells, resulting in an increase in the permeability of the epithelial and endothelial cells (Figs. 1, 2, and 3). The increase in the permeability facilitates the migration of inflammatory cells and allows the influx of RBCs and fluid from the blood capillary. Large volume of fluid (alveolar edema) fills the airspace leading to a difficulty in the breathing. Also, the inflammatory reactions may lead to alveolar cell death, fibrin deposition, and hyaline membrane formation. These findings support an important role of endothelial cells in the vascular phase of COVID-19. Furthermore, pulmonary intussusceptive angiogenesis and other pulmonary vascular lesions have been observed in autopsy specimen of COVID-19 patients [20].
Severe respiratory illness could be a major symptom of SARS-CoV-2 infection, because the ACE2 receptor is expressed in the lung AT2 cells, alveolar macrophage, and capillary endothelial cells [11] (Fig. 1). The expression of the ACE2 has been detected in other tissues, such as the cardiovascular, hepatic, renal, pancreatic, and the gastrointestinal tissues. This expression profile partially explains why some infected patients not only develop ARDS, but also develop other complications, such as myocardial injury (MI), arrhythmia, acute kidney injury (AKI), shock, multi organ failure, diabetes, and ultimately death [21].
IL-6 has an essential part in inflammatory cytokine storm in COVID-19. IL-6-producing CD14 + CD16 + inflammatory monocytes are significantly high [22]; therefore, the rationale for using tocilizumab has been used in COVID-19 patients. Tocilizumab, which is a recombinant humanized monoclonal antibody against the IL-6 receptor, is likely to induce its antagonistic effect on IL-6producing monocytes following activated Th1 cells in the lung. Tocilizumab is a first drug for the treatment of cytokine storm in COVID-19, especially in patients with multiple comorbidities. Despite the numerous ongoing trials assessing the safety and efficacy of tocilizumab in COVID-19 patients, IL-6 play a role in controlling the lung inflammation and is important for the clearance of viruses [23]. Therefore, inhibiting IL-6 raises the possibility of impaired viral clearance or exacerbation of lung inflammation [9].
Interestingly, an abnormal coagulation profile has been shown in COVID-19 patients during the late stage of the disease; specifically, increased concentrations of D-dimer and other fibrin degradation products are mainly associated with poor prognosis [24]. The HScore is a recommended evaluation as well as prognostic tool used in patients with secondary HLH at high risk of hyperinflammation. The score combines both critical laboratory as well as clinical parameters, assessing for an underlying of immunosuppression and cytopenias, measuring serum aspartate aminotransferase (AST), triglycerides, fibrinogen, ferritin, body temperature, organomegaly, and hemophagocytosis on bone marrow aspirate [8]. The HScores generate a probability for the presence of secondary HLH; a score more than 169 is 93% sensitive and 86% specific for HLH [8].
Finally, since the anti-viral immunity is needed to recover from COVID-19, the use of immunosuppressants on these patients should be used with caution. One strategy to avoid the inhibition of anti-viral immunity is to use targeted instead of broad immunosuppressive medications. Unfortunately, we still lack consensus on the optimal timing of treatment administration to decrease the harmful effects of immunosuppression, as well as the routes of their administration.

Mesenchymal stem cells (MSCs): characteristics and types
MSCs are a heterogeneous cell population propagating in vitro as plastic-adherent cells, have fibroblastlike morphology, and form colonies in vitro [25]. The International Society for Cellular Therapy (ISCT) defining criteria for MSCs is that they adhere to plastic, express the surface markers CD90, CD73, and CD105, are negative for the hematopoietic markers CD14, CD34, CD45, CD19, and HLA-DR, and should express a multilineage differentiation capability into adipogenic, osteogenic, and chondrogenic lineages [26].
Bone marrow MSCs (BM-MSCs) are considered the most widely used and investigated type of MSCs, which was first isolated from the bone marrow by Friedenstein Fig. 1 Immunopathogenesis of the SARS-CoV-2. The alveolar epithelium consists of a monolayer of alveolar type I (AT1) cells and alveolar type II (AT2) cells. Under normal condition, the AT2 cells secrete surfactant covering all the lining epithelium to facilitate alveolus expansion. AT1 and AT2 are tightly connected with tight junctions, which control the transfer of ions and fluid across the epithelium. The endothelial cells of the blood capillaries are connected by intercellular junctions and control the influx of inflammatory cells and fluid into the interstitial space. SARS-CoV-2 infects AT2 cells and resident alveolar macrophages that express ACE2. This activation induces chemokine secretion that recruits inflammatory and immune cells into the infected alveoli. The increased inflammatory cells in the lung lead to secretion of large amounts of pro-inflammatory cytokines "cytokine storm" that lead to damages in the lung. The migrated neutrophils and monocytes release toxic mediators, causing endothelial and epithelial injuries. The intercellular junctions are disrupted leading to formation of gaps between the alveolar cells as well as between the endothelial cells, resulting in an increase in the permeability of the epithelial and endothelial cells. The increase in the permeability facilitates the migration of inflammatory cells and allows the influx of RBCs and fluid from the blood capillary. Large volume of fluid (alveolar edema) fills the airspace leading to a difficulty in the breathing. Also, the inflammatory reactions may lead to alveolar cell death, fibrin deposition, and hyaline membrane formation and colleagues in 1974 [27]. Later, MSCs were identified and successfully produced from other sources, such as the perivasculature [28], adipose [29], dental pulp [30], muscle [31], dermis [31], and fetal tissue [32]. The abundance of adipose-derived stem cells (ASCs), their ease of isolation using a minimally invasive procedure [33], and their expansion as well as their differentiation ability into multiple lineages make ASCs a promising less-invasive alternative to BM-MSCs for therapeutic applications [34,35]. The most commonly used adult sources for human MSCs are bone marrow [36] and the adipose tissue stromal vascular fraction [29,34,35]. The highly harvestable bone marrow or unwanted/waste product of adipose sources forms the foundation for most of the data in the field of MSC-based therapeutics. The umbilical cord (UC) tissue [37] and the placenta [38,39] and their associated tissue Wharton jelly (WJ), and amniotic fluid (AF), are among the other young "adult" tissues, that are also considered good sources of human MSCs, where they are normally discarded after birth. MSCs are generally recognized as immune evasive making them safe when used in allogeneic settings [40]. Allogeneic MSCs are able to bypass the immune system due to low expression of the major histocompatibility complex-1 (MHC-I) and -II proteins. MSCs are often referred to as being "immunoprivileged" due to lack of the T cell costimulatory molecules, CD80 and CD86 [41]. Previous studies reported that fetal MSCs, adult BM-MSCs, and ASCs express HLA-I and do not express HLA-II [42][43][44][45][46]; however, these MSCs start to express HLA-II after stimulation with IFN-γ [44,45,47,48]. A recent study demonstrated that iPSC-derived MSCs do not express HLA-II and costimulatory molecules [49]. Interestingly, induced pluripotent stem cell (iPSC)-derived MSCs express a very low level of HLA-II in comparison to MSCs derived from fetuses and adult sources after their stimulation with interferon-γ (IFN-γ) [49]. These findings present iPSC-MSCs as an efficient source for allogenic transplantation without the risk of immune rejection due to the lower immunogenicity compared to adult MSCs.

Therapeutic potentials of MSCs
MSCs have been extensively studied over the past~30 years for their wide clinical applications and regenerative capacity. MSCs have made their way over the past 25 years into now over 950 registered clinical trials listed with the FDA, exhibiting an excellent safety profile. With over a 10,000 patients treated with MSCs in a controlled clinical setting, and upon successful completion of phase 1 or phase 2 trials, several tens of MSCs-based studies have advanced to phase 3 clinical trials (www. clinicaltrials.gov).
A fundamental clinical decision remains to choose among whether to use autologous vs allogeneic sources of MSCs, where both have displayed successful production of large numbers of MSCs [50,51]. MSC replacement in the large numbers is needed to treat significant tissue injury, a process that further requires orchestrated steps involving successful engraftment and cell differentiation [52]. A target dose of 100-150 million MSCs can be obtained from cell culturing and expansion of 25 ml of BM-MSC aspirate. In about 3 weeks duration, a volume of about 0.4-0.5 ml of packed cells can be generated [53]. The MSC isolation from different tissues, such as BM and adipose tissue and their re-implantation at other sites highlight their ability to repair tissues in vivo. However, this process clearly diminishes in aging population compared to younger adults [54]. In addition, MSCs could be generated in vitro in large number from human pluripotent stem cells (hPSCs) [55,56], which showed a lower immunogenicity in comparison to adult sources [49].
MSCs have been extensively examined for their therapeutic capacity in regenerative medicine, because of their ability to home to sites of inflammation and damaged tissue, ultimately serving as a source of growth and trophic factors and regenerative molecules. The potential therapeutic effect of MSCs is based on their low immunogenicity, their immunomodulatory characteristics, and their ability to secrete growth factors, as well as anti-microbial peptides [57]. MSCs administered systemically tend to migrate to the injury region to promote functional recovery [58]. MSCs can also extravasate from the blood vessels, just like immune cells, via the expression of cell surface adhesion molecules. Migration of MSCs occur in response to chemokines binding to cognate receptors present on their cell surface [59] and result in the stimulation of matrix metalloproteinases degrading the basal membrane and allowing subsequent extravasation [60]. By displaying a coordinated rolling, MSCs contact the endothelial cells in a P-selectin-and vascular cell-adhesion molecule 1 (VCAM1)-dependent manner [61]. Guided by chemotactic signals, MSCs migrate through the interstitium to the injured area. An increase in the MSC migration capacity toward chemokines is achieved via the upregulation of their receptors, CCR2, CCR3, and CCR4. Also, interleukin (IL)-8, an inflammatory chemokine, may induce migration of MSCs to injured areas [62,63].

Immunoregulatory functions of MSCs
One of the major therapeutic characteristics of MSCs is their immunomodulatory role, including a network of cytokines and cell-cell interactions. Interestingly, MSCs only exert its immunoregulatory capacity after receiving the activation signals from the inflammatory milieu; therefore, MSC's immunoregulatory capacity is not constitutive, rather is driven by "licensing" process [64]. Previous studies showed that the macrophages play an essential role during wound healing; thus, they have emerged as key candidate targets in therapeutic tissue regeneration approaches [64,65]. Macrophages exhibit functional repolarization as tissue repair progresses, shifting from the proinflammatory or M1-phenotype to an antiinflammatory or M2-phenotype. M1 macrophages secrete high levels of pro-inflammatory cytokines, while M2 macrophages secrete lower levels of proinflammatory cytokines, exhibit tissue repair, and enhance the resolution of inflammation [66]. Imbalance between M1-and M2-activities can lead to continuous inflammation and hinders the normal repair process, both contributing to impaired tissue repair [67]. MSCs enhance tissue repair and regeneration by modulating the immune response, acting as sensors and switchers of inflammation, rather than by replacing damaged cells. This is largely attributed to the secretion of growth factors; among the immunoregulatory factors are prostaglandin E2 (PGE2) and IL-6 that help in transitioning macrophages toward M2 phenotype [68,69] (Fig. 2). Further, the classical proinflammatory cytokines produced at the acute stage of inflammation, such as IFN-γ, TNF-α, or IL-1β enhances the paracrine effects of MSCs exerted on macrophages [70,71].
In order to stimulate the MSC immunosuppressive effect, threshold levels of inflammatory factors are required. Insufficient MSC activation can lead to an increase in the inflammation [72]. Recently, it has been shown that IL-10 alone is insufficient to enhance MSC immunomodulation, rather enhances the priming influence of TNF-α, indicating that MSC activation by IL-10 is dependent on TNF-α [64]. MSCs further decrease TNF-α secretion via PGE2 but not IL-6, supporting the concept that MSC immunomodulatory potential is highly correlated to the release of PGE2 [64].

Anti-inflammatory and antiproliferative effects of MSCs
MSCs reduce the pro-inflammatory effect of DCs by suppressing their secretion of TNF [82]. Also, plasmacytoid DCs (pDCs), a set of specific cells for the secretion of high levels of type I IFN, increase the production of IL-10 following the incubation with MSCs [82]. MSCs can further inhibit the cytotoxic activity of resting NK cells by reducing the production of natural cytotoxicity receptor 3 (NKp30) and natural-killer group 2, member D (NKG2D), involved in the activation of NK cells and target cell killing [83]. Therefore, MSCs inhibit NK cell proliferation and IFN production [84,85]. Also, neutrophils are important cells of innate immunity, undergoing a process known as the respiratory burst when binding to an antigen. MSCs have been reported to eliminate the respiratory burst and to prevent the neutrophil cell death by an IL-6-dependent mechanism [86]. Also, MSCs play a key role in the adaptive immune system, where it inhibits the proliferation of T cells activated with antigens [76]. This leads to a reduction in the IFN production and an increase in IL-4 production by T helper 2 (T2) cells, indicating a change in T cells from a pro-inflammatory (IFN-producing) to an antiinflammatory (IL-4-producing) state [82]. Furthermore, MSCs have been shown to downregulate CD8 + cytotoxic T lymphocytes (CTL)-mediated cytotoxicity [87] and further inhibit B cell expansion in vitro. Also, MSCs can suppress B cell differentiation and the constitutive secretion of chemokine receptors, affected by the MSC-mediated suppression of T cell functions [88]. Furthermore, MSC-derived IDO has been shown to be required in the inhibition of the expansion of IFNsecreting Th1 cells and, together with PGE2, to stop NK cell activity [89].
Finally, interleukin-1 receptor antagonist (IL-1Ra) and programmed death-1 (PDL1) are among the other secreted regulators of MSCs. IL-1Ra is among the antiinflammatory cytokine produced by MSCs, which can inhibit Th17 polarization. IL-1Ra expression tends to increase in MSCs exposed to IL-1β, TNF-α, and IFN-γ. Th17 cells induce the upregulation of PDL1, playing a major role in activating the MSC immunosuppressive effect [98]. PDL1 further support the cell-cell contact through MSC-mediated inhibition on Th17 cells [98]. MSC enhanced PDL1 ligand secretion suppress the activation of CD4 + T cells and downregulate IL-2 secretion [99].

MSC-derived exosomes
Extracellular vesicle (EV) is a term including both exosomes and microvesicles (MVs). The exosome diameter is less than 200 nm, while MV diameter can reach up to 1000 nm. The secretomes of MSCs and their vesicles offer a powerful tool for cell-free therapy due to their paracrine and/or endocrine effects [100]. This strategy bypasses most of the safety concerns related to cellbased therapy, such as contamination with oncogenic cells and continuous cell proliferation [101]. The key features of MSC-derived EVs are (1) non-proliferative, which reduce the risk of tumor formation; (2) negative for HLA-I and HLA-II, which can be induced, and therefore, they can be used from other individuals without any risk of immune response; (3) small in size allowing them to pass from the small blood capillaries; and (4) stored without using DMSO, which may change their characteristics [102]. EVs bind to a receptor on the cell membrane of the targeted cells, where they merge with the membrane to secrete the EV contents inside the cell or enter into the cytoplasm in the form of endocytic vesicles [103].
Among the targeted proteins, Sema3A has been found to induce sepsis-triggered cytokine storm through an interaction with Plexin-A4 and Toll-like receptors (TLRs) [116]. Stat3 is another targeted protein, a key upstream stimulator of inflammatory pathways during sepsis [117]. Finally, EVs act as biological regulators that can promote changes in their targets through targeted pathways. The cargo of the EVs is enriched with miR-NAs and other transcripts that act as regulators of the immune system [118,119]. Therefore, EVs are attractive tools for clinical applications as immunosuppressants, vaccines, or activators of differentiation and repair processes [120].

MSCs and their exosomes as potential therapies for COVID-19
MSCs have been well described in ALI and ARDS. It exerts its function via targeting both infectious, inflammatory, and endothelial factors. MSCs can release KGF2, PGE2, GM-CSF, IL-6, and IL-13 to facilitate phagocytosis (Figs. 2 and 3). In addition, multiple clinical studies [121][122][123][124][125] investigated the effect and mechanism of MSCs and MSC-EVs on lung injuries caused by different reasons (Table 2). MSCs and their secreted secretome exert an immunomodulatory, anti-inflammatory, antiapoptotic, and anti-fibrotic functions in ALI and ARDS.
MSC-derived exosomes are a multitargeted biologic agent, which can reduce the cytokine storm and reverse the inhibition of host anti-viral defenses associated with COVID-19 [154]. The functions of the MSC-derived exosomes have been studied in in vitro and in vivo models. EVs collected from the conditioned medium of BM-MSCs have been used to treat ARDS in a mouse model. The results showed that EVs have a similar effect as MSCs in reducing the inflammation and edema in the lung [126]. The effect of MSCs on macrophage modulation in ARDS mouse models has been mainly found due to the effect of EVs [127]. Treating mouse alveolar Anti-apoptotic and antiinflammatory [108] miR-122 Cyclin G1, IGF1R, and P4HA1 Anti-proliferative and antifibrotic [109] miR-223 NLRP3 Anti-inflammatory: decrease pytoptosis and IL-1β [110] miR-181 Bcl-2 and Stat3 Anti-fibrotic and activated autophagy [111] miR-133 RhoA and connective tissue growth factor Enhanced plasticity [112] miR-17-92 PTEN Cell survival signaling pathway [113] IGF1R insulin-like growth factor receptor 1, P4HA1 prolyl 4-hydroxylase alpha 1, NLRP3 NLR pyrin domain-containing 3, RhoA homolog gene family member A, BCL-2 B cell lymphoma 2 family  macrophages with MSC-EVs has a protective effect in the mouse model of lipopolysaccharide (LPS)-induced lung injury [127]. It has been found that the transferred active mitochondria through EVs leads to a change in the polarization of the macrophage from M1 (pro-inflammatory) to M2 (anti-inflammatory) phenotype as a result of an increase in the oxidative phosphorylation [127]. Also, it has been reported that functional mitochondria transferred through MSC-EVs enhances mitochondrial function in primary human alveolar cells and enhances their ability to repair lung injuries [155]. In addition, the effect of MSC-EVs has been examined in pneumonia mouse model induced by E. coli [129]. The data showed that EVs could decrease lung inflammation by reducing neutrophil and macrophage recruitment as well as MIP-2 level [129]. It has been found that the EVs reduce lung edema and endothelial permeability and the expression of CD44 on the target cells are required for the binding and uptake of EVs into cells [129]. BM-MSC-EVs have been used in another ARDS mouse  [151]. Furthermore, EVs exert an immunomodulatory function in the macrophage by inhibiting the secretion of TNF-α and enhancing the secretion of IL-10 [151]. In a pig model, the influence of MSC-EVs on influenza virus-induced ARDS has been investigated [155]. Administration of EVs has been found to decrease the influenza virus replication, pro-inflammatory cytokines, and alveolar cell death in pigs through the transfer of RNA [155]. UC-MSC-EVs have been also used in a rodent model. The study found that the UC-EVs are effective in reducing ALI and the EVs primed with INF-γ are more efficient than normal EVs in improving ALI [156]. This indicates that EVs isolated from different source could be used for lung injury. Interestingly, the primed EVs have been found to be larger in size than normal EVs; however, the mechanism of this size increase remains unclear [156]. Previous studies used in vitro human injured lungs to investigate the effect of MSC-EVs. MSC-EVs restore fluid clearance and reduce edema in human injured lungs in vitro [143,157]. Another study examined the effect of EVs on human lungs with pneumonia induced by E. coli found that EVs reduce the permeability of lung protein and enhance alveolar fluid clearance [158]. Barrier properties of the of human lung endothelial cells injured with TNF-α, IFN-γ, and IL-1β are restored with EVs. This improvement is due to an increase in the levels of Ang-1 in the injured endothelium, treated with EVs [159]. Although there are promising results obtained from using MSC-EVs in lung injury, more mechanistic studies are needed to improve our understanding on the molecular mechanisms involved in EV effect. MSCs could act upon two ways in the novel COVID19 treatment, namely via its immunomodulatory effects and differentiation ability. MSCs display numerous advantages of relevance to ALI and ARDS. Although progress in the management of ALI/ARDS depends on improvements in supportive measures, ultimately decreasing the mortality rates [160], the failure of pharmacologic treatments indicate the need to consider new strategies for ALI/ARDS. MSC possible therapeutic potential is attributed to their accessible derivation from several adult tissues, their low immunogenicity, indicating that they could be given allogeneically [161], and their relative ease of isolation and expansion ability in culture. In case of COVID-19 patients, autologous and allogenic MSC transplantation could be applied, because MSCs do not express ACE2 and TMPRSS2; therefore, patient's own MSCs cannot be infected by SARS-CoV2 [162].
However, the negative effects caused by the SARS-CoV2 infection on the blood cells and different organs may influence the ability to isolate autologous MSCs of high quality and sufficient number to treat the same patient. Taken together with the low immunogenicity of MSCs and the complications associated with the SARS-CoV2 infection, using allogenic MSCs is the method of choice for COVID-19 patients.
Resolution of ALI/ARDS in COVID19 is hindered by the disruption of the epithelial barrier that suppresses alveolar fluid clearance and depletes surfactant [163]. MSC capacity to aid in restoring epithelial and endothelial function by differentiating MSCs into these cell types or by secreting paracrine and trophic factors to increase restoration of the lung tissue offers a promise for treatment of ALI/ARDS in COVID19. MSCs have been widely studies in other inflammatory conditions, where they demonstrated a reduction in injury and/or enhanced restoration of function in the kidney [164,165], liver [166], and heart [167]. MSC immunomodulatory properties exhibit a promise for treating ALI/ARDS in COVID19 via their ability to 'reprogramme' the immune response to decrease the destructive inflammatory components, while maintaining the host response to infections, in addition to enhancing the repair and resolution of lung injury by acting an effector for tissue regeneration.  [168]. The patient received IV infusion of MSCs three doses (5 × 10 7 cells/dose, every 3 days), in which following the second dose, clinical improvement has been observed. Furthermore, the number of the neutrophils and inflammatory cells in the patient reduced to a normal level, while the number of lymphocytes elevated to their normal levels [168]. Also, a recent study reported that IV injection of clinical-grade MSCs (1 × 10 6 /kg) into seven patients with ARS-CoV-2 leads to an improvement in the functional outcomes and a recovery enhancement [162]. Among the beneficial outcomes of MSC treatment are an observed increase in the number of peripheral lymphocytes, a decline in the C-reactive protein (CRP), a decrease of overactivated cytokine-producing immune cells (CXCR3 + CD8 + T cells, CXCR3 + CD4 + T cells, and CXCR3 + NK cells) and TNF-α, and increase in IL-10,                mainly attributed to the anti-inflammatory and immunomodulatory functions of MSCs [162]. Furthermore, an increase in CD14 + CD11c + CD11b mid regulatory DC population has been observed [162]. Leng [162].

MSC clinical trials for COVID-19 patients
Another recently published study has tested the safety and efficacy of allogeneic AT-MSCs in 13 COVID-19 adult patients under invasive mechanical ventilation [169].
The patients had received previous anti-inflammatory and/or antiviral drugs, including lopinavir/ritonavir, steroids, tocilizumab and/or hydroxychloroquine, among others. Thirteen patients have been enrolled in the trial and have received two IV doses of AT-MSC 0.98 × 10 6 per kg of body weight, 3 days apart. The treatment has been followed by a reduction in inflammatory parameters (CRP, ferritin, LDH, IL-6) as well as increase in B-lymphocytes (67%) and CD4 + and CD8 + (100%) T lymphocytes). Remarkably, a reduction of D-dimer and fibrinogen 5 days after the first dose of AT-MSCs has been observed in most patients, and the patients do not develop a thromboembolic event [169,170].
Furthermore, a recent clinical trial has been employed using MSC-derived exosomes. A prospective nonrandomized open-label cohort study by Sengupta et al. has evaluated the safety and efficacy of exosomes (ExoFlo™) obtained from allogeneic BM-MSCs as treatment for severe COVID-19. A total of 24 patients have received 15 ml of ExoFlo™. The treatment resulted in significant  measurement of key  cytokines (IL1, IL6, IL8, TNF-alpha, IL10, TGF-beta, sRAGE, Ang-2) at baseline and at defined time frames following therapy is crucial.
One of the key points in designing the clinical trial is the MSC dose and route of administration. IV route is the most commonly used method for systemic delivery of MSCs in the majority of clinical trials, with much fewer trials using intra-arterial (IA) injection [172]. The most frequently utilized route of MSC administration in ARDS and COVID-19 is also via IV infusion. Of interest, IV route remains the most well studied route used for MSC delivery in pulmonary diseases [172]. Regardless of the site of inflammation and tissue injury, and opposing to the old concept that MSCs only migrate to the site of injury following IV administration, MSCs are mostly trapped in lungs and undergo phagocytosis within 24 h [173]. In the majority of clinical indications, human MSCs are frequently transfused IV at doses ranging from 1 to 2 million cells/kg and never exceeding a dose of 12 million cells/kg [172,174]. The median dose for IV route is 1 × 10 8 MSCs per patient per total dose. Analysis of MSC trials using IV route indicated minimal effective doses (MEDs). Efficacy dose-response outcome data indicated a narrower MED range of MSCs ranging from 100 to 150 million, where either higher or lower has been less efficient [172].

Challenges in treating COVID-19 using MSCs and their exosomes
One of the most significant challenges for MSC therapies is to optimize MSC homing efficiency. IV administration of MSCs shows low homing efficiency where cells get trapped in the pulmonary capillaries [175], a process that has been partially explained by insufficient production of homing factors, such as CXCR4, on MSCs [176,177]. It has been reported that the in vitro propagation of MSCs gradually leads to dramatic reduction in the expression of homing factors [178,179]. Several strategies have been used to improve MSC homing capacity, including targeted administration, genetic modification, magnetic guidance, in vitro priming, cell surface modification, and radiotherapeutic techniques [180,181].
Further studies are needed to define the optimal source and dose of MSCs, administration route, the time window of MSC administration, and dose frequency (single vs. multiple-dose regimen). Due to MSC expression of tissue factor (TF/CD142), which triggers the coagulation, a pro-coagulation can be triggered ultimately leading to thromboembolic events following infusion. Thus, the use of anti-coagulant during MSC administration can be considered during administration guidelines and protocols [182]. Due to the trapping effect of MSCs occurring in the lung, a justification for the high doses of MSCs has been proposed [183]. Therefore, MSCs could be genetically modified to overexpress selected genes in order to increase in their engraftment. Pretreatment with a series of preconditioning approaches could also promote MSCs therapeutic effects and enhance their survival in the lung. Of note, challenges with autologous MSC transplantation in ARDS are demonstrated by their impaired potential due to the immunomodulatory effects of bone marrow MSCs [184].
As extracorporeal membrane oxygenation (ECMO) is among the salvage therapy for refractory respiratory failure in the context of acute respiratory compromise associated with SARS-CoV-2, it is critical to learn how MSCs would act in such setting [185]. IV administration of MSCs has been found to attach to membrane oxygenator fibers during ECMO in an in vitro ARDS model, leading to a significant decrease in the flow through the circuit [186]. Intratracheal and IV infusions of MSCs before ECMO or during a pause in the flow are some suggested strategies to overcome this limitation. Interestingly, a more recent study demonstrated an enhanced endogenous MSC mobilization in patients with ARDS undergoing ECMO [187]. Therefore, intratracheal administration of MSCs might be an option in ARDS requiring continuous high-flow ECMO.
The fate of MSCs after infusion also remains to be investigated. Cell migration and distribution studies have shown that the majority of MSCs localize to the lungs after IV infusion [188,189]. Intravascular arrest of MSCs is due to MSC's diameter ranging from 10 to 20 μm, bigger than the width of the pulmonary micro-capillaries [190]. However, following IV administration, MSCs still tend to migrate to sites of injury and move from the lungs to other organs, such as the liver and spleen [191,192]. Nevertheless, challenges remain in interpreting this data as this tracking could be detecting phagocytosed MSCs. This is in line with the evidence showed that most of MSCs become apoptotic after administration [193]. To avoid poor cell survival following MSC transplantation, several preconditioning strategies have been proposed. Ang1-preconditioned cell survival was significantly increased via increased Akt phosphorylation. This has further reduced the apoptotic rate in vitro via increased expression of B cell lymphoma protein 2 (Bcl-2) and the ratio of Bcl-2/Bcl-associated X (Bax) [193]. Several priming strategies with pharmacological agents, inflammatory cytokines or mediators, hypoxia, and biomaterial have been shown to enhance the therapeutic efficacy of MSC transplantation. MSCs enhanced trafficking and homing to sites of injury is demonstrated in the high expression of chemokine receptors, such as CXCR4, CXCR7, and CX3CR1 [194]. Finally, although several in vitro culturing strategies have been developed to mimic the natural MSC niche, preserving the function and quality of a scalable clinical-grade cell expansion remain a challenge.
The adult sources of MSCs include painful and invasive procedures with possible donor site morbidity [195]. Regardless the origin, the MSCs display heterogeneity in their abilities to propagate and differentiate [196]. The differences in the properties of MSCs are associated with the variations in the age of the donor, method of MSC isolation, and in vitro culturing approaches. As an alternative source, hPSCs could differentiate into unlimited number of MSCs, displaying MSC characteristics [55,197,198]. Further studies are needed to extensively examine the differences between MSCs isolated from different tissues and those derived from hPSCs.
A systemic procoagulant state has been observed in severely ill COVID-19 patients, which tends to result in poor outcome. Such patients are at high risk of disseminated intravascular coagulation (DIC) and thromboembolism. Looking at the risk of hypercoaguable state in COVID-19 patients, the safety profile of MSCs could be further challenged [199]. DIC and thromboembolism taking place after the administration of TF/CD142-expressing MSC products have been reported [182]. The production of the highly procoagulant tissue factor TF/ CD142 between products could vary. BM-MSCs have the lowest TF/CD142 expression, whereas ASC display the highest expression profile [182]. Finally, the dynamics of the current pandemic and the rising global demand highlight the need for scalable manufacturing required to provide enough doses of MSC product of high quality in a reproducible and timely manner.

Conclusion and future perspectives
MSCs have a potential therapeutic function in COVID-19, which is displayed in their ability to enhance alveolar fluid clearance and promote epithelial and endothelial recovery through transfer of EV components together with the cell-cell contact as well as their secreted soluble factors. As ACE2 is widely expressed in other tissue types in addition to lungs, it is intuitive to consider MSC effect on the other organs as well. MSC treatment may reduce the progression of ARDS in severely ill COVID-19 patients with multiple organ failure.
Although several clinical trials have been recently registered to examine the safety and efficacy of MSCs as an emerging therapeutic option for COVID-19-induced disease, fewer studies have been published. Learning from these clinical trials, MSCs could exert its immunomodulatory and regenerative capacity in COVID-19 patients. Although there is no approved treatment for COVID- 19 as of yet, MSC therapies continue to show improvement in the treatment of some of the leading causes of mortality in COVID-19 patients, namely acute ARDS, pneumonia, inflammation, and sepsis. The majority of these clinical trials are based on IV infusion of MSCs and their derived exosomes. Despite the clinical improvement witnessed, the safest and most effective route of MSC delivery into COVID-19 patients remains unclear, especially in the context of the heterogeneity of MSC-based products, intravascular arrest, and poor cell survival. IV infusions of poorly characterized MSC products remain one of the most significant drawbacks of MSC cell-based therapy, which could theoretically promote the risk for thromboembolism. The best delivery route for MSCs giving the highest positive effects with minimum toxic effects remains to be resolved. Whether any difference between IV, intratracheal, and intraperitoneal administration routes exists also remains unclear. Furthermore, elucidating the molecular mechanisms of MSCs during lung injury is crucial to understand their role in ARDS. The premature marketing of unproven stem cell therapy to the public resulted in the unfortunate increase of unregulated stem cell clinics; therefore, we cannot recommend any MSC-based treatment, which does not use characterized cell product, perform functional mechanisms, define variability in donor and tissue source, measure intermediate parameters, and define final patient endpoints(s), key steps that are common practice in FDA registered trials. Lack of consensus underlines major challenges to the clinical translation of MSC-based therapy.
Future studies should focus on developing genetically modified MSCs, generating significantly large number of EVs that could safely transfer different potent and effective therapeutic factors [114]. Finally, optimizing the clinical-grade production of MSCs as well as establishing a consensus on registered clinical trials based on cellproduct characterization and mode of delivery would aid in laying the foundation for a safe and effective MSCbased therapy in COVID19.