**3. Mesenchmal stem cells (MSCs)**

#### **3.1 Origins of MSCs**

In 1968, Friedenstein et al. isolated stem cells from the bone marrow (BMSCs) of mice [40]. The study showed that BM contained clonogenic progenitor cells and adherent cells similar to fibroblasts, termed as a colony forming unit-fibroblast [40]. These cells were found to have the ability to differentiate into chondrocytes, osteocytes, osteoblasts and adipocytes*in vitro* [40]. In 1991, Arnold Caplan changed the terminology to "Mesenchymal Stem Cell", due to their similarities with stem cells from mesodermal origins in embryonic tissues [41]. Later in 2017 Caplan suggested that the name MSC be alerted to "medicinal signaling cells" to accurately reflect their*in vivo* abilities of acting as an in situ medication [42]. Currently, "Mesenchymal Stem Cell" is the most common nomenclature, however Caplan did manage to emphasize their function.

With the variations in nomenclature as well as controversy surrounding their characteristics, the need for an official and concise criterion was needed. In 2006, The International Society of Cellular Therapy (ISCT) established parameters with four minimum criteria should be used to define MSCs. The criteria were quickly accepted by the medical community and are the status quo currently [43, 44].

The ISCT criteria for MSCs: 1) Plastic adherence in standard culture condition, 2)Positive expression (≥95%) of CD105, CD90, CD73 cell surface antigens, 3) Low expression (≤ 2%) of CD45, CD34, CD14, CD11b, CD79, CD19 and HLA-DR cell surface antigens, 4)Potential to differentiate into osteoblasts, adipocytes and chondrocytes in vitro.

#### **3.2 MSC sources**

MSCs can be differentiated by either being totipotent, pluripotent, multipotent, or unipotent [45, 46]. Totipotent MSCs for example, can form both embryonic and extraembryonic structures and proliferate indefinitely into cell types from all three embryonic germ layers [45]. Multipotent MSCs or adult stem cell are the most widely used and can differentiate into cell types from their respective source tissue [46]. MSCs can then be further subdivided by their source tissue. Two of the most common sources of MSCs are BM and adipose tissue. These are autologous sources that have been studied substantially and have the most associated data. Both of these sources require the patient to undergo an invasive procedure and are considered the first and second most reputable sources respectively for MSCs [47, 48]. Allogenic birth derived tissues such as umbilical cord (UC), UC-derived Wharton's jelly, amniotic fluid and placenta are also viable sources for MSCs. These sources have advantages in relation to their availability, lack of invasiveness, and presence of more pluripotent cells [12, 49, 50]. However, these sources have less data and do not have such an extensive history of use in comparison with allogenic sources.

#### **3.3 MSC's mechanisms of action**

MSCs have a long history of use in the treatment of viral lung infections, pneumonia, ALI and ARDS [6, 12]. This prior literature has been used to support

their current use in COVID-19. Studies have showed that when IV administered, MSCs have specific and optimal mechanisms of action for the treatment of COVID-19. MSCs are able to evade the body's immune system and accumulate within the lung microvasculature enabling them to act locally [51, 52]. They have direct antiviral activity, as well as anti-inflammatory, anti-apoptotic, and anti-fibrotic properties [47]. MSCs have also been touted for their ability to induce tissue regeneration, transdifferentiate into cells and produce EVs [53].

IV infusion is the one of the most commonly used route for MSC delivery with hundreds of clinical trials showing evidence of its safety [54]. A systematic review and meta-analysis by Lalu et al. summarized the results of IV administered MSCs in over 1000 patients [55]. The review indicated that there were no associated adverse events within any of the studies and no patient developed any organ system complications, infusion related toxicity, infections nor death [55]. In a study by Hwa Lee et al. IV infused MSCs were shown to accumulate into emboli within the lungs with no negative physiological effects [51]. In fact, the cells were noted to secrete TSG-6, a potent anti-inflammatory, the effects of which were amplified due to the sequestration within the lung [51]. As immune privileged cells, MSCs can be used either allogenically or autologously, due to their low levels of class I major histocompatibility complex (MHC) and class II MHC [52]. MSCs have also been shown to lack the associated co-stimulatory molecules (B7–1, B7–2, CD40, CD80 and CD86) needed to activate antigen presenting cells and the inflammatory process [52]. With these factors in mind MSCs are primed to act locally within the lungs to effectively and efficiently carry out their functions.

#### **3.4 MSCs and immunomodulation**

#### *3.4.1 Innate immune response*

In addition to the therapeutic potential of MSCs in regenerative medicine, for which they been most known for, they have also shown promising results in the regulation of immune responses [47]. MSCs through their ability to secrete various soluble factors are able to suppress both the innate and adaptive immune responses [47].

NOs and MOs both play a vital role in the innate immune response with DCs being the gate keeper to the adaptive response [56]. MOs can be subdivided into M1 or M2 subtypes each with their own distinct functions [57]. The M1 subtype are well known to be classically activated and responsible for phagocytosis, antigen presentation to DCs and secretion of pro-inflammatory cytokines such as TNF-α, ΙL-1α, IL-β, IL-6, IL-12 ultimately promoting a Th1 response [57]. The M2 subtype are known for their high secretion of IL-10 promoting an anti-inflammatory Treg and Th2 response along with inducing tissue remodeling and wound repair [57]. MSCs have been shown to secrete prostaglandin E2 (PGE2) and induce a switch in the MO population into an M2 subtype as well as substantially decreasing levels of IL-1β and IL-6 [58]. Wahnon et al. further elucidated this anti- inflammatory switch by reporting that the transcription factor signal transducer activators of transcription-3 (STAT3) activated in MSCs through cell to cell interactions between MOs produced IL-10 and promoted an M2 phenotypic switch [59]. NO activation and function have also been shown to be inhibited by MSCs. NOs are known to be a key component of the innate immune response and in pathophysiology of ARDS. NOs when activated release harmful reactive oxygen species, superoxide anions, peroxidases and proteases that lead to diffuse alveolar damage, and accumulation of alveolar fluid that underlie ARDS [60]. MSCs have been shown to secrete a potent antioxidant enzyme, SOD3 that has been shown to decrease the release of peroxidases,

proteases and the oxidative burst of NOs [61]. They have also able to directly engulf dead NOs through ICAM-1 thereby further inhibiting release of their toxic contents [61]. Secretion of tumor necrosis factor-inducible gene 6 protein (TSG-6) via MSCs has also been shown to bind to IL-8 and CXCL8, inhibiting further migration, extravasation and recruitment of NOs [62].

Immature DCs patrol peripheral tissues for foreign antigens and are activated by cytokines (TNF-α, IL-1β, and IL-6) from M1 MOs [63]. Once immature DCs are activated they mature into conventional DCs and present their cleaved epitopes on their HLA complexes, inducing a pro-inflammatory Th1 and Th17 response [63]. PGE2 from MSCs has been shown to decrease CD38, CD80, CD86, IL-6, and IL-12 thereby decreasing DC function and pro-inflammatory T cell responses [64]. Preventing the maturation of these conventional DCs is vital in order to prevent this T cell response and the associated pro inflammatory state. Furthermore, DC maturation was inhibited by the inactivation of MAPK and NF-κB signaling cascades via the secretion of the TSG-6 [65]. In a study by Chen et al. DC maturation was induced from a conventional (pro-inflammatory) DC into a plasmacytoid DC population by PGE2 from MSCs, shifting the T cell population into a Th2 (antiinflammatory) subset [66]. In addition, specific miRNAs (miR-21-5p, miR-142-3p, miR-223-3p, miR-126-3p) within EVs of MSCs have shown to further attenuate the DC maturation process [67].

#### *3.4.2 Adaptive immune response*

MSCs role in modulating T and B cell responses begins with their attenuation of MO and DC functions and continues with PGE2 from MSCs. PGE2 has been proven to increase the production of cAMP in T cells down regulating IL-2, and the IL-2 receptor as well as inhibiting the release of intracellular Ca2+ resulting in the direct inhibition of T cell activation [68]. PGE2 has also been shown to inactivate T cells via the hydrolysis of phosphatidylinositol, diacylglycerol and inositol phosphate [68]. In addition, PGE2 promotes a Th2 and a T reg shift in the T cell population overall influencing immunosuppression and an anti-inflammatory response [13, 69]. MSCs through their secretion of IDO, PGE2, TGF-β1, and Hepatocyte growth factor (HGF) have also been shown to induce G0/G1 cell cycle arrest in T cells and B cells [70, 71]. Nitric oxide (NO) from MSCs has shown to play a role in this by suppressing the phosphorylation of signal transducer and activator of transcription 5, thereby inhibiting TCR activated T cell proliferation and production of cytokines [72]. Studies have also suggested that MSCs can induce T cell and B cell apoptosis through direct cell to cell contact. Utilizing their interactions with the Fas/ Fas ligand, TNF-related apoptosis-inducing ligand/death receptor signaling and programed death ligand-1/programmed death-1 pathways have shown to promote T and B cell apoptosis [73, 74]. This process was especially seen in CD4+, CD8+ and Th17 cells with a synergistic increase in T reg cells [75]. The down regulation of CXCR4, and CXCR5 via MSCs has shown further evidence of inhibiting B cell migratory abilities towards chemoattractant agents such as CXCL12 and CXClL13 [74]. Lastly, GM-CSF from MSCs have been recognized as having inhibitory actions on the production of CXCR4, CXR5, IL-6, and IL-7 while having no negative effects on IL-4 and IL-10 from B cells with a net anti-inflammatory affect [74].

#### **3.5 MSC's additional mechanisms of action**

Studies have shown that MSCs have been effective in inhibiting the viral replication of influenza, hepatitis B, herpes simplex, cytomegalovirus and the measles

virus [76–79]. In a study by Khatri et al. MSCs had the ability to inhibit viral replication, shedding and lung damage in a porcine model with influenza induced pneumonia [76]. MSC-derived EVs were shown to be the key players in this process via their transfer of RNAs to virus infected epithelial cells. Lung epithelial cell apoptosis, hemagglutination and viral shedding were all significantly reduced in the study [76]. MSC-derived EVs have also demonstrated to decrease pro-inflammatory cytokine while increasing IL-10 and increase T regs [76]. IDO via MSCs has also been shown to directly decrease viral replication in most of the viruses that have been studied [76–79].

The secretion of various CKs, GFs and EVs have been reported to promote tissue regeneration and inhibit apoptosis, tissue fibrosis and alveolar fluid accumulation. As previously elucidated, M2 MOs promote anti-inflammatory Treg and Th2 responses while inducing tissue remodeling and wound repair [57]. Direct tissue regeneration from MSCs has been attributed to keratinocyte growth factor (KGF), vascular endothelial growth factor (VEGF), and hepatocyte growth factor (HGF) all of which have also been known to contribute to the in decrease collagen build up and fibrosis [80, 81]. In an in vivo bleomycin-induced pulmonary fibrosis model, Aguilar et al. noted that KGF was the key factor in the inhibition of collagen accumulation, promoting endogenous type II pneumocyte proliferation and overall attenuation of lung damage [82]. Previous studies have also further characterized KGF as being a potent factor in lung epithelial cell proliferation, while simultaneously being capable to increase matrix metalloproteinase-9 (MMP-9), IL-1RA and promoting clearance of apoptotic cells and inhibiting fibrosis [82, 83]. Gazdhar et al. used an in vivo bleomycin induced lung injury model in which he found that MSCderived HGF was able to inhibit lung fibrosis and induce alveolar epithelial repair by decreasing TGF-B and α-smooth muscle actin expression [84]. The positive effects of HGF was further studied by Wang et al. who showed that MSC-derived HGF was responsible for increasing endothelial cell proliferation, intercellular junction proteins (VE-cadherin and occludin), and IL-10 while decreasing IL-6 and overall apoptosis [85]. MSC-derived VEGF and HGF have also shown to be able to stabilize Bcl-2 and inhibit pro-apoptotic factors hypoxia-inducible factor-1α protein, Bnip3 and CHOP contributing to their anti-apoptotic and anti-fibrotic effects [86]. In addition to the intracellular stabilization via these aforementioned GFs, factors such as MSC derived aniopoietin-1, and EVs have shown to induce alveolar fluid clearance within the lungs adding in their therapeutic benefits in ARDS [87]. In a study by Zhu et al. using an E.coli endotoxin-induce ALI model, MSC-derived EVs showcased their ability to transfer mRNA encoding for KGF inhibiting NOs, pulmonary edema and lung permeability [88].
