Human mesenchymal stromal cells do not express ACE2 and TMPRSS2 and are not permissive to SARS‐CoV‐2 infection

Abstract Anti‐inflammatory and immune‐modulatory therapies have been proposed for the treatment of COVID‐19 and its most serious complications. Among others, the use of mesenchymal stromal cells (MSCs) is under investigation given their well‐documented anti‐inflammatory and immunomodulatory properties. However, some critical issues regarding the possibility that MSCs could be infected by the virus have been raised. Angiotensin‐converting enzyme 2 (ACE2) and type II transmembrane serine protease (TMPRSS2) are the main host cell factors for the severe acute respiratory syndrome‐coronavirus 2 (SARS‐CoV‐2), entry, but so far it is unclear if human MSCs do or do not express these two proteins. To elucidate these important aspects, we evaluated if human MSCs from both fetal and adult tissues constitutively express ACE2 and TMPRSS2 and, most importantly, if they can be infected by SARS‐CoV‐2. We evaluated human MSCs derived from amnios, cord blood, cord tissue, adipose tissue, and bone marrow. ACE2 and TMPRSS2 were expressed by the SARS‐CoV‐2‐permissive human pulmonary Calu‐3 cell line but not by all the MSCs tested. MSCs were then exposed to SARS‐CoV‐2 wild strain without evidence of cytopathic effect. Moreover, we also excluded that the MSCs could be infected without showing lytic effects since their conditioned medium after SARS‐CoV‐2 exposure did not contain viral particles. Our data, demonstrating that MSCs derived from different human tissues are not permissive to SARS‐CoV‐2 infection, support the safety of MSCs as potential therapy for COVID‐19.


| INTRODUCTION
In December 2019, an outbreak caused by a novel coronavirus, later named SARS-CoV-2, occurred in China and rapidly spread throughout several other countries, becoming pandemic. 1  , the disease caused by SARS-CoV-2, mainly affects the respiratory system and can progress to respiratory distress syndrome (ARDS), a fatal condition in more than 50% of the cases. 1 Up to 30% of COVID-19 patients may develop cardiac damage due to acute coronary syndrome, septic heart, or acute myocarditis. [2][3][4] As there are no specific therapeutics for treating COVID-19, in particular for the most severe cases complicated by ARDS or acute fulminant myocarditis, new innovative therapeutic approaches are urgently needed. Anti-inflammatory drugs have been proposed as possible approaches and it has been suggested that immunosuppressive therapy may mitigate the manifestations of COVID-19. 5,6 Mesenchymal stem cells (MSCs) possess immunomodulatory

Significance statement
Human mesenchymal stromal cells (hMSCs) are currently under investigation for the treatment of COVID-19. However, the potential safety profile of hMSCs in this context has never been defined since none has described if they express ACE2 and TMPRSS2, the main host cell factors for SARS-CoV-2 entry, and if they can be infected by SARS-CoV-2. This study provides the first evidence that ACE2 and TMPRSS2 are not expressed in hMSCs derived from both adult and fetal human tissues and, most importantly, that hMSCs are not permissive to SARS-CoV-2 infection. These results support the safety of MSCs as potential therapy for COVID-19.
F I G U R E 1 hACE2 and TMPRSS2 expression in human MSCs from fetal and adult tissues. A, Relative quantification by quantitative reverse transcription PCR (RT-qPCR) and the 2 −ΔΔCt method of ACE2 (left) and TMPRSS2 (right) mRNA levels in MSCs from cord blood (CB-MSC, n = 2 lines), cord tissue (CT-MSC, n = 2 lines), amnios (A-MSC, n = 4 lines), bone marrow (BM-MSC, n = 4 lines), and adipose tissue (AT-MSC, n = 2 RNA samples from the same line), compared with the Calu-3 cell line. Each sample was loaded in triplicate. Columns represent mean 2 −ΔΔCt values and SD. The glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as a reference gene. *P < .05, **P < .01, ***P < .001 vs Calu-3. B, Representative Western blot of hACE2 and TMPRSS2 in cell lysates (left) and serum free 48 hours-conditioned media (right) of the different MSC types, and the positive control cell line Calu-3. Actin was used as loading control for cell lysates. The amount of conditioned media loaded for each sample was obtained from 1 × 10 5 cells. C, ACE2 protein quantification by enzyme-linked immunoadsorbent assay (ELISA) in serum-free 48 hours-conditioned media generated by 1 × 10 5 MSCs. The bars represent the mean and SD of triplicates obtained from a single line of Calu-3 and 4 distinct lines of both A-MSCs and BM-MSCs. *P < .001 vs Calu-3. MSCs, mesenchymal stromal cells properties as demonstrated by numerous in vitro and animal model studies. These effects are mediated by cytokines and soluble factors able to modulate the systemic but also the tissue inflammatory response. 7 In particular, when administered intravenously, most MSCs lodge in the pulmonary vascular bed where they survive for at least a few days. 8  Accordingly, the aim of the present work was to evaluate whether human MSCs from various sources express ACE2 and TMPRSS2 and, most importantly, if they are permissive to SARS-CoV-2 infection.

| MATERIALS AND METHODS
An expanded method section is available as Supporting Information.

| Cell culture
The MSC lines were previously isolated from amniotic membrane of human placenta, cord blood, cord tissue, bone marrow, and adipose tissue, expanded and characterized. [13][14][15][16][17] All the MSCs fulfill the criteria set by the International Society for Cell & Molecular Therapy (ISCT). 18 To perform the experiments, we used passage 3 to 5 MSCs. Conditioned media were generated as described 13,19 with some minor modification described in the Supporting Information Methods, in order to evaluate the soluble amount of ACE2. The human lung Calu-3 (ATCC HTB-52) and the African green monkey kidney VERO E6 (VERO C1008; ATCC CRL-1586) cell lines were purchased and maintained as indicated by ATCC (www.lgcstandards-atcc.org).
A, Experimental design. First, MSCs from CB, CT, A, BM and AT, and the permissive cell lines VERO E6 and Calu-3 were incubated with SARS-CoV-2 wild-type strain. CPE defined as cell rounding, detachment, degeneration, and/or syncytium formation was assessed by microscopic analysis, and the cells maintained in culture for 7 days, changing the medium every three 3 days. Then, supernatants were collected and tested in a reinoculation experiment on VERO E6. B, Representative phase contrast images of VERO-E6, Calu-3, and hMSCs before (not infected-upper panels) and after exposure to SARS-CoV-2 (infected-lower panels). Scale bar = 200 μm. hMSCs and Calu-3 showed no signs of CPE and maintained their typical spindle-shaped and epithelial morphology, respectively, whereas VERO E6 displayed clear signs of CPE, since almost all the cells were round or fused into syncytia or detached. C, Representative phase contrast images of VERO E6 cells after exposure to supernatant generated by the infected cell lines respectively indicated on top of each frame. Scale bar = 200 μm. CPE, cytopathic effect; hMSCs, human mesenchymal stromal cells; MSCs, mesenchymal stromal cells glycoprotein (VSV-G) as a positive control. MSCs were incubated with Spike or VSV-G pseudotyped viral particles for 24 hours.

| Infection with SARS-CoV-2 wild strain
MSCs, Calu-3, and VERO E6 were infected with 100 μL (100 TCID 50 / mL) of a previously titrated SARS-CoV-2 wild strain, isolated from an infected patient. The virus was incubated for 1 hour and then removed; the medium was changed every 3 days. Cells were scored every other day and for 1 week using a light microscope to detect the appearance of cell rounding, detachment, degeneration, and/or syncytium formation, called hereafter cytopathic effect (CPE). To verify if the cells tested can be infected and allow SARS-CoV-2 replication even in the absence of CPE, at day 7 from infection, supernatants from each MSC, Calu-3, and VERO E6 culture were collected and inoculated into VERO E6. CPE occurrence was monitored for 1 week.

| Statistical analysis
All results are reported as mean ± SD and the data were analyzed with a one-way or two-way analysis of variance followed by Bonferroni all pair-wise multiple comparison test using the InStat software (GraphPad Software, Inc., San Diego, California; http://www. graphpad.com). P values less than .05 were considered statistically significant.

| RESULTS
We assessed ACE2 and TMPRSS2 expression on human fetal MSCs derived from amniotic membrane of placenta (A, n = 4), cord blood (CB, n = 2) or cord tissue (CT, n = 2), and human adult MSCs derived from bone marrow (BM, n = 4) or adipose tissue (AT, n = 1). The lung epithelial cell line Calu-3, which expresses high levels of both ACE2 and TMPRSS2 and is permissive to SARS-CoV-2 infection and nonlytic replication, 21 was used as positive control. Compared with Calu-3, the levels of ACE2 and TMPRSS2 mRNAs in all MSC lines considered were around 100-fold and 200-fold lower, respectively ( Figure 1A). ACE2 and TMPRSS2 protein expression in MSC lysates was undetectable by Western blot ( Figure 1B). Finally, we were unable to detect any soluble amount of ACE2 in MSC-derived conditioned media by both Western blot and enzyme-linked immunoadsorbent assay (ELISA) (Figure 1B,C).
To substantiate the absence of ACE2 and TMPRSS2 in fetal MSCs infected in adhesion showed the typical spindle shape morphology with no signs of CPE ( Figure 3B and Table 1). Also, MSCs infected in suspension and seeded in 24 well plates were found adherent to plastic and showed no CPE starting from the day postinoculum up to 7 days (data not shown). Conversely, a 100% CPE was detected in the control VERO E6 cell line ( Figure 3B and Table 1). As expected, 21 we did not observe CPE in Calu-3 cells for the entire observation period ( Figure 3B and Table 1 experiment in VERO E6 cells ( Figure 3A). As expected, a typical CPE was evident in all the wells inoculated with supernatants collected from infected VERO-E6 and Calu-3 cultures ( Figure 3C and Table 1).
On the contrary, none of the supernatant collected from the different MSC cell lines induced CPE, demonstrating the absence of viral replication inside the MSC lines and, consequently, the absence of SARS-CoV-2 infection ( Figure 3C and Table 1).

| DISCUSSION
The

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author.