Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which is responsible for the COVID-19 pandemic, belongs to the family Coronaviridae. The Coronaviridae family has 2 subfamilies, Coronavirinae and Torovirinae. Coronavirinae includes seven coronavirus-caused respiratory diseases in human, which are divided into two groups: normal coronaviruses (HCoV-229E, HCoV-NL63, HCoV-HKU1, HCoV-OC43) and novel coronaviruses (MERS-CoV, SARS-CoV and SARS-CoV-2)5. The normal coronaviruses usually causes acute mild upper tract respiratory diseases (except NL63), while the novel coronaviruses can lead to pneumonia and severe acute respiratory syndrome (SARS). Among the three novel coronaviruses, SARS-CoV-2 has had the lowest case fatality rate (CFR)—under 5% — but the highest level of transmissibility.

Coronaviruses (CoVs) have large single-stranded positive-sense RNA genomes—around 30 kilobases, containing a 5’-cap and 3’ poly (A) tail structure6. CoVs have a lipid bilayer membrane and include four main structural proteins: nucleocapsid (N) protein, membrane (M) protein, envelope (E) protein and, most importantly, spike (S) protein6, 7. S protein plays a pivotal role in the initial attachment and penetration of the virus to the host cell by binding to the angiotensin-converting enzyme 2 receptor (ACE-2). The spike protein is composed of two subunits, S1 and S2. S1 mediates the binding of the virus to the host cell receptor, while S2 is responsible for the fusion of virion and cell membrane7. Inside the host cell, the virus replicates and performs exocytosis, after which it is ready to invade adjacent cells.

ACE-2 has been determined to be a main functional receptor for SARS-CoV, HCoV-NL63, and SARS-CoV-2. ACE-2 has primarily been expressed in type II pulmonary epithelial cells, so the respiratory system is the easiest method for viruses to infect patients. However, ACE-2 has also been observed in nasal goblet cells, gastrointestinal epithelial cells, pancreatic β cells, and renal podocytes6, 8, leading to digestive issues, myocarditis, kidney diseases and, ultimately, multiple organ dysfunction syndromes. Compared to SARS-CoV, SARS-CoV-2 has a 10–20 times higher affinity for ACE-2 receptors9, which is why there have been more COVID-19-related myocarditis cases (even among those who are young and have no comorbidities) than in the 2002-2004 SARS outbreak in Hong Kong.

The renin-angiotensin system (RAS) plays an essential role in regulating blood pressure and homeostasis. In the lungs, angiotensin I (Ang I) metabolizes to angiotensin II (Ang II) by the ACE. Ang II acts through angiotensin type I and type II receptors (AT1R and AT2R), leading to the constriction of vascular, water, and sodium retention, stimulation of chronic inflammation, fibrosis, and lung damage. Conversely, the ACE-2 converts Ang II into Ang 1-7, which exerts vasodilation through binding Mas receptor, thus inhibiting inflammation and preventing pulmonary diseases and edema10. Therefore, the ACE2/Ang1-7/Mas axis is antagonized to the ACE/AngII/AT1R pathway (classic RAS). When SARS-CoV-2 binds to cells with a high affinity, the bond between the virus and ACE-2 is formed, leading to the down regulation of the ACE-2 axis. This imbalance is partly responsible for the acute and chronic inflammation (due to overactivity of RAS), resulting in medium to severe COVID-19 symptoms and long-term complications.

Many researchers have concluded that there is a correlation between the increasing level of pro-inflammatory cytokines and the outcome of COVID-19 patients; cytokine storm syndrome is the main factor that causes severe and fatal complications in patients11, 12, 13, 14. Cytokines are small protein molecules secreted through the interaction between cells, including pro-inflammatory and anti-inflammatory cytokines15. The virus invades the nasal epithelial cell, replicates, and immigrates to the upper and lower respiratory cells, then to lymphocytes. With COVID-19, cytokines are produced not only through PAMPs (pathogen-associated molecular pattern molecules) and DAMPs (damage-associated molecular pattern molecules) pathways but also during the crosstalk between epithelial cells and immune cells via the trigger of macrophages. In mild to moderate cases of COVID-19, interferon-1 (IFN-1, which includes IFN-α and IFN-ß) is efficiently and productively secreted, functioning as a lung protector. Otherwise, the inflammation is regulated by the balancing between pro-inflammatory (TNF-α, IL-1, IL-6) and anti-inflammatory (IL-10, IL-1) cytokines. This facilitates quicker and more effective virus clearance without any serious complications.

On the other hand, in patients with severe to critical disease, there is not only impaired INF-1 production and activity accelerating DAMPs and PAMPs pathways but also exacerbated inflammation, with an excess of pro-inflammatory cytokines (TNF-α, IL-6), leading to the cytokines auto-amplification and cytokine storm, resulting in ARDS, multi-organ failure or even death12, 14. Moreover, the imbalance of the ACE and ACE-2 axis that leads to the ACE/AngII/AT1R over-activity will exaggerate the releasing soluble TNF-α, HB-EGF, and IL-6Rα to adjacent cells and activate NK-kB and STAT3 to produce more IL-68. Serum IL-6 and TNF-α levels can be used to reliably and independently predict poorer prognoses16. Therefore, ACE/AngII blockers and anti-IL-6 have recently been considered viable medication for treating disease of medium to critical severity.

Besides local immune activity (innate cell immunopathology and plasma cytokines signature), understanding the specific immune response is a vital part of controlling the virus invasion, with three major components: antibodies (produced by B cells), CD4⁺ T cells, and CD8⁺ T cells. The adaptive immune response works slower than innate immunity; it takes time to proliferate and differentiate naive cells into effector cells. When the virions, ingested by antigen-presenting cells (APC), are recognized by CD4⁺ T cells, the T helper cells activate both cytotoxic T cells to destroy infected cells and virus, and B cells produce antibodies. If the adaptive immune response works fast and effectively, eliminating the circulating virions, the disease can be controlled; patients usually have no symptoms. Conversely, T cells and antibodies may not be activated soon enough, in which case the local immune activity can lead to over-activity in cytokine production. In the case of SARS-CoV-2 and COVID-19, the virus can avoid or delay innate immunity related to INF-1 efficiently, leading to the overproduction of pro-inflammatory cytokines17, 18, 19, 20.

To date, there have been more than 20 reports on the use of MSCs for PF in animal models. Most of the studies used MSCs from bone marrow or adipose tissues. Other kinds of MSCs used to treat PF in animal models include umbilical cord MSCs, amniotic membrane MSCs, human embryonic MSCs, and human menstrual blood–derived MSCs. In a 2021 review by Li et al.34, 12 animal studies showed that animals treated with MSCs had significantly improved survival rates compared to the control group (p < 0.001)34. Moreover, in this meta-analysis review, Li et al. (2021) showed that MSC transplantation clearly reduced pulmonary fibrosis scores compared to the control group (p < 0.01)34.

In a 2013 study, Tzouvelekis et al. used adipose-derived stromal vascular fraction (SVF) to treat 14 PF patients35. In this study, the researchers used a low dose of SVF (0.5 million cells per kg of body weight per infusion). Although no cases of serious adverse events were reported, treatment efficacy was not significantly improved in patients who received transplants35.

Averyanov et al. (2019) used a high dose of MSCs to treat PF in 20 human patients. This clinical trial showed that the cumulative dose of 2 billion cells is safe and tolerable in PF patients. After 52 weeks of follow-up, the authors confirmed that MSC transplantation with a cumulative high dose is effective for treating rapidly progressing PF36.

A study by Campo et al. (2021) transplanted autologous bone marrow MSCs (BM-MSCs) for 13 patients37. The transplantation of BM-MSCs was determined to be safe, and the mean forced vital capacity declined by 8.1% at 3 months37.

Figure 1

Table 1

ClinicalTrials.gov Identifier	Title	Phase/Patients	Kind of cells	Methods	Country
NCT04909892		Phase II (60 patients)	Allogeneic culture-expanded adipose-derived mesenchymal stem cells	MSCs will be administered intravenously on Day 0, Day 2, and Day 4 with 2 groups of 2 doses: one vial, ~18.5 million cells per infusion and ~37 million cells per infusion	USA
NCT05126563		Phase II (80 patients)	Allogeneic culture-expanded adipose-derived mesenchymal stem cells	Dose: 200 million Route: Intravenous Regimen: Weeks 0, 2, 6, and 10.	USA
NCT04798066		Phase I (5 patients)	Autologous adipose derived mesenchymal stem cells	Dose: 200 million HB-adMSCs Route: intravenous infusion only, Regimen: a treatment duration of 14 weeks	USA
NCT05116761		Phase I/II (60 patients)	Bone Marrow Mesenchymal Stem Cell Derived Extracellular Vesicles	Dose: Normal saline 85 mL and ExoFlo 15 mL, which is 10.5 x 108 EV Route: Intravenous	USA

Effects of MSCs on COVID-19 treatment, including (1) anti-viral effects, (2) immune modulation, and (3) lung regeration, have been comprehensively discussed in some previous publications38, 39. Unlike COVID-19 treatment, post-COVID-19 treatment can be effective if the fibrosis process is reversed. Ideally, the treatment for fibrosis should achieve three steps of fibrotic tissue regeneration, namely, the inhibition or reduction in fibrosis progression (via inhibiting inflammation, reducing myofibroblast activity, and preventing further collagen synthesis), the degradation of the accumulated/existing collagen, and the triggering of lung regeneration (Figure 1). MSCs can target three therapeutic mechanisms of fibrotic lung regeneration.

MSCs can display immune modulation that significantly reduces inflammation by suppressing T lymphocytes, B lymphocytes, dendritic cells, and NK cells and stimulating regulatory T cells. MSCs also secrete an array of anti-inflammatory cytokines (reviewed in Pham and Vu, 202040). In a model of radiation-induced pulmonary fibrosis in animals, Dong et al. (2015) showed that MSCs can reduce fibrosis in PF in animals by stimulating the endogenous secretion of HGF and PGE241. MSCs also secrete HGF that inhibits epithelial-to-mesenchymal cell transition and promotes myofibroblast apoptosis42. The role of MSC-derived HGF in protection from bleomycin-induced PF was confirmed by Cahill et al. (2016). Cahill et al. (2016) compared the effects of MSCs and HGF-knockdown MSCs on bleomycin-induced PF and showed that HGF knockdown MSCs could not protect against fibrosis in vivo43.

The degradation of existing cumulated collagen in fibrotic lungs is an essential step in fibrotic lung regeneration. Transplantation of MSCs can activate the host’s macrophages, especially M2 macrophage, to synthesize the MMP-9 degrading the collagen44. Indeed, in animal models of PF, UCMSC transplantation reduced collagen deposition (determined by Sirius red staining)44. Indeed, overexpression of MMP9 in macrophages can attenuate bleomycin-induced PF45. In vitro, Khan et al. (2017) also showed that MSCs could decrease Col1A1 deposition by increasing the ratio of MMP-1 to TIMP-146.

The last step of anti-fibrosis in PF is tissue regeneration. Previous studies have confirmed that MSCs could produce and secrete some growth factors that can activate the local stem cells for tissue homeostasis47, 48. Chu et al. (2020) determined the indirect way that MSC transplantation triggers tissue regeneration in PF through TLR-444. Transplanted MSCs will promote the expression of TLR-4 in the host’s alveolar epithelial cells. TLR-4 promotes alveolar progenitor cell renewal and prevents severe pulmonary fibrosis49.

Xiao et al. (2020) reported that MSC-derived exosomes could reverse the progression of LPS-induced lung injury and fibrosis50. Exosomes from MSCs transmitted miR-23a-3p and miR-182-5p to the MLE-12 cells and inhibited the NF-kB and Hedgehog pathways by silencing lkbkb and destabilizing IKKbeta50. Zhang et al. (2021) also demonstrated that exosomes from BMMSCs can reverse the epithelial-mesenchymal transition to alleviate silica-induced pulmonary fibrosis51. Indeed, exosomes from BMMSCs caused an increase in epithelial markers (including E-cadherin and cytokeratin 19) and reduced the expression of fibrosis marker proteins (including alpha-SMA). The authors proposed that a mechanism for the anti-fibrotic effects of exosomes could be the attenuation of the Wnt/beta-catenin signaling pathway51.