The human mesothelial cell line MeT5A (ATCC, Rockville, MD) was cultured in Earle’s M199 supplemented with 10% FBS (Gibco-Life Technologies), 2 mM L-glutamine (EuroClone), 100 U/ml penicillin, 100 µg/ml streptomycin (Gibco-Life Technologies). This cell line was isolated from pleural fluids obtained from a non-cancerous individual. The human adenocarcinoma cell line Calu-3 was grown in RPMI 1640 (Euroclone, Milan, Italy; cat. n°ECB9006L) supplemented with 10% FBS, 2 mM L-glutamine, (EuroClone), 100 U/ml penicillin, 100 µg/ml streptomycin (Gibco-Life Technologies).

Cell lines were grown at 37°C, in a humidified atmosphere with 5% CO₂. MeT5A or Calu-3 cells cells were treated with DMSO or MS-275 in a dose-dependent experimental model ranging between 1 to 0.06 μM or 5 to 0.06 μM, respectively. MeT5A cells were treated with the epigenetics drugs MS-275, MC-2468, MC-4448, Tubastatin A, PCI34051, MC-2500 at 0.25μM and MGCD0103 at 0.125μM. Pharmacological treatment was repeated every 48 hours during the experimental procedure.

The primary antibodies for western blotting (WB) experiments: mAbs anti-GAPDH (sc-32233) and -Histone H4, were from Santa Cruz Biotechnology (Dallas, TX). pAbs anti–ACE2 (AB_2792286) was from Invitrogen (Waltham, MA). pAbs anti-TMPRSS2 (ab109131) was from Abcam (Cambridge, UK). Ab anti-acetyl-H3 (06599) was from Millipore (Merk, Kenilworth, NJ). Ab anti-acetyl-H4 (06866) was from Millipore (Merk, Kenilworth, NJ); control rabbit IgG (NB810-56910) was from Novus Bio, Minneapolis, MN.

Antibodies for Immunofluorescence: anti-ACE2 (Thermo Fisher Scientific, Waltham, MA, USA, AB_2792286). Cy3-conjugated anti-rat secondary antibodies (Jackson ImmunoResearch, 112-165-003). DRAQ5 staining solution (#130-117-343) was from Milteny. MS-275, MGCD0103, MC-2468, MC-4448, Tubastatin A, PCI34051, MC-2500 were from Prof. Mai laboratory. The VSV-spike-GFP construct was a kind gift from Sean Whelan, Washington University St. Louis, USA.

Subconfluent MeT5A (2.0×10⁵ cells/well; 24 wells) or Calu-3 (3.75x10⁵ cells/well; 24 wells) were pretreated with different doses of either MS-275 or DMSO as control, 24 hours (h) before infection.

Cells were then incubated with SARS-CoV-2 (isolate SARS-CoV-2/Human/ITA/PAVIA1073 4/2020, clade G, D614G (S) obtained from Dr. Fausto Baldanti, Policlinico San Matteo, Pavia, Italy) in serum-free Eagle’s Minimum Essential Medium at a multiplicity of infection (MOI) of 1 for MeT5A cells and MOI 0.001 for Calu-3 cells for 1 h at 37°C, 5% CO₂. Then, cells were washed with 1X PBS to remove viral inoculum, and complete culture medium supplemented by DMSO or MS-275 was added. Fresh DMSO or MS-275 was added 48 h post-infection. Culture supernatants and cell lysates were collected at 72 h post-infection (p.i.) (MeT5A cells) or 24 h and 48 h p.i. (Calu-3 cells).

To estimate the production of infectious SARS-CoV-2, serial dilutions of either MeT5A or Calu-3 cell culture supernatants, treated either with DMSO or MS-275, were measured by back-titrating assay in Vero E6 cells, as we recently described (Alonzi et al., 2021; Alonzi et al., 2022). Readout of the virus back-titration was based on detection of cytopathic effect (CPE), and infectious titer was expressed as 50% tissue-culture infective dose (TCID50) values, calculated according to the Reed–Muench method.

MeT5A cells were pretreated with MS-275 or DMSO as control for 48 hours before infection. Cells were then incubated for 1 hour with VSV-spike-GFP construct at MOI=1 to evaluate Spike-dependent viral entry. Then, cells were washed three times with 1X PBS to remove the viral inoculum and fixed with 4% paraformaldehyde (Merck Life Science, Milan, Italy) in 1X PBS to perform an immunofluorescence assay.

siGENOME SMARTpool siRNAs HDAC1(3065), siHDAC2 (3066) and siRNA control were purchased from Dharmacon. 200×10³ MeT5A cells were seeded on 12-well plates 24 h prior to transfection. Cells were transfected with 25 pmol of siRNAs, and 3.5 μl Lipofectamine^® RNAiMAX Reagent (Thermo Fisher Scientific) were diluted in two different tubes with 200 μl Opti-MEM (Gibco-Life Technologies). The two solutions were mixed gently and were incubated for 20 minutes at room temperature. Transfection solutions were added to cells with 0.6 ml of supplemented medium. Efficiency transfection was evaluated using RT-PCR.

Cellular RNA was extracted from cell cultures using TRIzol reagent (Life Technologies, Carlsbad, CA), according to the manufacturer’s instructions. cDNA synthesis was generated using a reverse transcription kit (A3500) from Promega (Madison, WI), according to the manufacturer’s recommendations. cDNAs were amplified by qPCR reaction using Maxima SYBR Green/ROX qPCR Master Mix (K0253) from Thermo Fisher Scientific (Waltham, MA). qPCR reactions were performed with the Rotor-Gene 6000 thermocycler (Corbett Research, Cambridge, United Kingdom). The primer sequences used in this study are shown in Table 1 . Relative amounts obtained with 2 (−ΔCt) method were normalized with respect to the housekeeping gene L34. Statistical significance was determined with a t-test with Prism version 8.0. Differences were considered significant at p < 0.05.

To evaluate cell viability upon MS-275 treatment Met5A cells (5.000 cells/well) and Calu-3 cells (20.000 cells/well) were stimulated with MS-275 for 48 hours and then tested with CyQUANT™ MTT Cell Viability Assay (V13154) from Invitrogen (Waltham, MA, USA) following manufacturer instruction. Absorbance were measured at 480 nm with Microplate Reader from Biorad (Hercules, CA, USA).

MeT5A cells were lysed in CelLytic™ MT Cell Lysis Reagent supplemented with 1 mM PMSF; 1 μg/ml each of aprotinin, leupeptin, and pepstatin; and 25 mM NaF (all from Merck Life Science). For histone extraction, MeT5A cells were lysed in Triton Extraction Buffer (TEB: PBS containing 0.5% Triton 100x, 2mM PMSF and 0.2% NaN₃). Pellet was suspended in 0.2N HCl. Equal amounts of protein were resolved by SDS-PAGE. Proteins were transferred to nitrocellulose membranes (Amersham Life Sciences, Little Chalfont, United Kingdom) and probed with primary antibodies using standard procedures. Peroxidase–conjugated secondary antibodies anti-rabbit (711-036-152) and anti-mouse (715-036-150) were from Jackson Immuno Research Laboratories (West Grove, PA, United States). Nitrocellulose-bound antibodies were detected by enhanced chemiluminescence (ECL) Immobilon Classico WBLUC0500 and Immobilon Crescendo Western HRP substrate WBLUR0500 from Millipore (Burlington, MA, United States).

After pharmacological treatments or infections, cells were washed in cold PBS, fixed with 4% paraformaldehyde (Merck Life Science) in 1X PBS, and permeabilized with 0.2% Triton X-100 (Merck Life Science) in 1X PBS. Coverslips were mounted in Prolong Gold antifade (Life Technologies) and examined under confocal microscopes (Leica TCS SP2, Wetzlar, Germany and Celldiscoverer 7, Carl Zeiss AG, Oberkochen, Baden-Württemberg, Germany). Digital images were acquired with the Leica software, and the image adjustments and merging were performed using the appropriate tools of ImageJ software. A minimum of 4 fields per sample (at least 150 total cells per total) from two independent experiments was analyzed.

Chromatin crosslinking and sonication were performed as in (Battistelli et al., 2017). After determining the DNA concentration, 50 μg of chromatin were used for each sample. It was diluted 10 folds with dilution buffer (TE 1x, sodium dodecyl sulfate (SDS) 0.5%, and protease and phosphatase inhibitors) and was incubated overnight with 5 μg of acetyl-H3, acetyl-H4 or control rabbit IgG and 20 μl magnetic beads. Immuno-precipitated samples were put on a magnetic rack, and Input sample was recovered by the supernatant of the IgG control. The samples were washed with four successive buffers: I) Low salt Buffer, II) High salt Buffer, III) LiCl Buffer, IV) TE Buffer. During the washes, the samples were in rotation, and at the end of each wash, magnetic beads were recovered. Then, the samples were dissolved in 300 μl of elution buffer (1% SDS, 0.1M NaHCO₃) and 10 mg of RNase A; 0.5% SDS and RNase were added to the Input sample for 10 min at RT. 30 μl of Proteinase K (10 mg/ml) were added to all samples (included input). Elution, proteinase K and reversal crosslink steps were performed at 62°C for 5 hours while vortexing. DNA extraction was performed like conventional ChIP, see above. Finally, DNA was resuspended in 50 μl of nuclease-free water.

To analyze the impact of HDAC in SARS-CoV-2 infection, we first analyzed the expression of the main receptors ACE2 ( Figure 1A ) and TMPRSS2 ( Figure 1B ) upon treatment of MeT5A cells with seven specific pharmacological inhibitors with different class/isoform selectivity. Treatment with the HDAC1/2 inhibitor Mocetinostat (MGCD0103) (Fournel et al., 2008; Khan et al., 2008), the HDAC1-3 inhibitor Entinostat (MS-275) (Saito et al., 1999; Suzuki et al., 1999; Rossi et al., 2018), the HDAC3 inhibitor MC4448, the class I/IIb HDAC inhibitor MC2468 (Zwergel et al., 2021), the HDAC6 inhibitor Tubastatin A (Butler et al., 2010), and the HDAC8 inhibitor PCI-34051 (PCI) (Balasubramanian et al., 2008) differently modulated the expression of these receptors. Interestingly, MGCD0103 and MS-275 induced a significant increase of both ACE2 and TMPRSS2, whereas treatment with compound MC2500 (Suzuki et al., 1999; Rossi et al., 2018), an inactive meta isomer of MS-275, did not modify the expression of these receptors. Of note, the HDAC3-selective inhibitor MC4448 induced the expression of both ACE2 and TMPRSS2. On the other hand, treatment with the HDAC6-selective inhibitor Tubastatin A significantly downregulated ACE2, while not affecting TMPRSS2 expression. Treatment with the HDAC8-selective inhibitor PCI-34051, as well as the class I/IIb HDAC inhibitor MC2468, was not effective. These results highlight the different activity of isoform-specific HDAC inhibitors regarding ACE2 and TMPRSS2 expression, emphasizing the role of HDAC1-3 inhibition in this context.

We then focused on MS-275, which was the most potent HDAC inhibitor in ACE2 modulation. Dose-response experiments confirmed the induction of ACE2 and TMPRSS2 gene expression, with a peak of induction at 0.25 µM and 0.5 µM, respectively ( Figures 2A, B ). The decreased expression of ACE2 and TMPRSS2 observed at higher doses of MS-275, are likely due to a significant reduced cellular viability ( Supplementary Figure 1A ). ACE2 and TMPRSS2 expression was confirmed at protein level by Western Blot analysis ( Figure 2C ) and by confocal microscopy ( Figure 2D ). The specific genetic silencing of HDAC1 and HDAC2 isoforms confirmed that both HDAC1 and HDAC2 are involved in the expression of ACE2 and TMPRSS2 ( Figures 2E, F ).

Since MS-275 induced an upregulation of SARS-CoV-2 receptors, we asked whether this could lead to an increased cell entry by the virus. Therefore, we used a VSV-spike-GFP pseudovirus (Case et al., 2020) and measure its entry in Met5A cells. As shown in Figure 3A , MS-275 treatment (0.25 μM) significantly enhanced viral entry.

We previously demonstrated the ability of MCs to support SARS-CoV2 infection and replication (Matusali et al., 2021). To analyze the role of HDAC1-3 inhibition in this process, MeT5A cells were treated with different concentrations of MS-275 or DMSO as a control for 24 hours before SARS-CoV-2 infection (MOI=1). As shown in Figure 3B , MS-275 treatment induced a significant increase of viral particles production in a dose-dependent manner. Accordingly, MS-275-treated cells (0.25 µM) have a higher expression of viral ENV RNA with respect to those treated with vehicle alone ( Figure 3C ).

These results demonstrated that HDAC1-3 pharmacological inhibition by MS-275 favors SARS-CoV-2, cell entry, replication and virion production in MeT5A cells.

To mechanistically demonstrate an effect of HDAC1-3 inhibition in ACE2 and TMPRRS2 expression, the acetylation status of ACE2 and TMPRSS2 promoters was investigated.

The study of human ACE2 and TMPRSS2 enhancers/promoters from ENCODE revealed the presence of multiple acetylation peaks overlapping in 7 cell lines analyzed ( Figures 4A, B ). We then studied the acetylation status of histone H3 (Ac-H3) and H4 (Ac-H4) by specific chromatin immunoprecipitation (ChIP) assay.

Ac-H3 ChIP assay demonstrated that treatment with MS-275 at the concentration of 250 nM significantly enhances H3 acetylation on specific peak regions found on ACE2 and TMPRSS2 promoters ( Figures 4C, D ), but not on the promoter of hRLP30 ( Figure 4E ), a gene encoding a constitutively expressed ribosomal protein. Using the same experimental setting, an increase of H4 acetylation on specific sequences at ACE2 and TMPRSS2 promoters was found ( Figures 4F-H ).

Accordingly, MS-275 treatment promoted the acetylation on Histone H3 and H4, as shown by western blot analysis ( Figures 4I, J ). Overall, these results suggest that MS-275 promotes ACE2 and TMPRSS2 mRNAs expression by enhancing H3 and H4 acetylation, thus favoring an opened chromatin conformation in the promoter regions.

The analysis of the effect of HDAC1-3 inhibition was extended to Calu-3 cells, a lung adenocarcinoma cell line often used as an experimental system in SARS-CoV-2-related studies. Calu-3 cells express high levels of ACE2 and TMPRSS2 with respect to MeT5A cells ( Supplementary Figure 2 ) (Liao et al., 2013; Najafi Fard et al., 2021; Alonzi et al., 2022). Similarly to the results obtained with MeT5A cells, treatment with MS-275 promoted a significant induction of both ACE2 and TMPRSS2 with a peak of induction at 1.0 µM and 2.5 µM, respectively ( Figures 5A, B ). The decreased levels of ACE2 and TMPRSS2 observed at higher doses of MS-275, are likely due to reduced cellular viability ( Supplementary Figure 1B ). According to the increased expression of receptors, pretreatment with MS-275 significantly increased both the TCID value, although at a less extent respect to that observed in MeT5A cells, and the expression of SARS-CoV-2 nucleoprotein ( Figure 5C, D ). These results demonstrate that besides cells from pleura, HDAC1-3 pharmacological inhibition by MS-275 favors SARS-CoV-2 replication and virion production also in human lung epithelial cells.