A549 and MDCK cells (Husain and Cheung, 2014; Nagesh and Husain, 2016) were grown and maintained in minimum essential medium (MEM; Life Technologies) supplemented with 10% fetal bovine serum (FBS; Sigma–Aldrich), penicillin-streptomycin, and L-glutamine (Life Technologies) at 37°C and under 5% CO₂ atmosphere. Influenza virus A/PR/8/34 (H1N1), A/New Caledonia/20/1999 (H1N1), and A/WSN/34 (H1N1) strains (Husain and Cheung, 2014; Nagesh and Husain, 2016), were propagated in 10-days old embryonated chicken eggs and titrated on MDCK cells.

The cell monolayers were washed twice with serum-free MEM and the virus inoculum prepared in serum-free MEM was added to the cells, which were then incubated at 35°C for 1 h. After removing the virus inoculum, cells were washed once with serum-free MEM. Finally, fresh serum-free MEM was added to the cells and cells were incubated back at 35°C. In some experiments, NH₄Cl (Sigma–Aldrich), MG132 (Calbiochem) or caspase 3 inhibitor (Calbiochem) were included in serum-free MEM. To inactivate IAV, virus inoculum was irradiated under 30W UV bulb for 5 min before adding to the cells. Generally, cells were infected with IAV at the multiplicity of infection (MOI) of 0.5 to 5.0 plaque-forming units (PFUs) per cell.

The total RNA was extracted from cells using PureLink RNA isolation kit (Life Technologies). The RNA integrity was confirmed on Bioanalyzer 2100 (Agilent) using RNA 6000 Nano Chip and purity and quantity were measured on NanoDrop 1000 (ThermoFisher). The SuperScript III First-Strand Synthesis System (Life Technologies) and total RNA template was used to synthesize the cDNA. The quantitative real-time PCR was performed on ViiA 7 Real-Time PCR System (Applied Biosystems) by using the cDNA, SYBR green select master mix (Life Technologies) and pre-designed KiCqStart primers (Sigma–Aldrich). Primer sequence – HDAC2 forward: 5′-GGTCATGCTAATGTGTAGAAG-3′, HDAC2 reverse: 5′-GTCGGTCCAAAATACTCAAG-3′, 18S RNA forward: 5′-ATCGGGGATTGCAATTATTC-3′, 18S RNA reverse: 5′-CTCACTAAACCATCCAATCG-3′, GAPDH forward: 5′-CTTTTGCGTCGCCAG-3′, GAPDH reverse: 5′-TTGATGGCAACAATATCCAC-3′, beta-actin forward: 5′-GACGACATGGAGAAAATCTG-3′, beta-actin reverse: 5′-ATGATCTGGGTCATCTTCTC-3′, viperin forward: 5′-CTTTTGCTGGGAAGCTCTTG-3′, viperin reverse: 5′-CAGCTGCTGCTTTTCTCCTCT-3′. Beta-actin, 18S RNA, and GAPDH were used as reference genes for normalizing and determining the fold- or percent change in the expression of HDAC2 and viperin mRNAs, which was calculated using 2(-Delta Delta C(T)) method as described elsewhere (Livak and Schmittgen, 2001).

The cells were harvested and lysed in lysis buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 0.5% sodium dodecyl sulfate [SDS], 0.5% sodium deoxycholate, 1% TritonX-100, and 1X protease-inhibitor cocktail [Roche]), and total protein was estimated using BCA kit (ThermoFisher). Equal amount of proteins were resolved on 10 or 15% Tris-glycine SDS-PAGE under reducing conditions. Proteins were transferred onto Protran Premium nitrocellulose membrane (GE Healthcare), which was then probed with either mouse anti-HDAC2 (clone 3F3) (1:1,000, Cell Signaling), rabbit anti-viperin (1:1,000; Clone D5T2X, Cell Signaling), mouse anti-STAT1 (1:1,000; Clone 42/Stat1, BD biosciences), mouse anti-STAT1 (pY701) (1:1,000; Clone, 14/P-STAT1, BD biosciences), mouse anti-ubiquitin (1:500; Clone, P4D1, Santa Cruz), mouse anti-NP (1:1,000; NR-4282, obtained through BEI Resources, NIAID, NIH), goat anti-NP (1:1,000; kindly provided by St Jude Children’s Research Hospital), rabbit anti-actin (1:5,000; Abcam) or rabbit anti-protein disulfide isomerase (PDI) (1:5,000; Sigma–Aldrich) antibody followed by IRDye 680LT- or IRDye 800LT-conjugated (1:10,000; Li-COR) or horseradish peroxidase-conjugated anti-mouse, anti-goat, or anti-rabbit IgG antibody (1:2,000–5,000; Life Technologies). Protein bands were visualized by either fluorescence or using a chemiluminescent ECL substrate (GE Healthcare). Images were acquired on Odyssey Fc imaging system (Li-COR) and exported as TIFF files, which were then cropped and compiled in Adobe Photoshop CC 2015.

Pre-designed small-interfering RNA (siRNA) targeting HDAC2 (ID: SASI_Hs02_00332058) and non-targeting MISSION^® universal negative control obtained from Sigma–Aldrich were delivered to cells using Lipofectamine RNAiMax reagent (Life Technologies) by following the manufacturer’s guidelines. Briefly, siRNA oligonucleotides and RNAiMax were separately diluted in OptiMEM I medium (Life Technologies), then mixed together and incubated for 20–30 min at room temperature. Subsequently, the siRNA-RNAiMax complex was mixed with the cells and transferred to a cell culture plate. Cells were incubated at 37°C for 72 h before infecting or further processing them.

The viability of cells was ascertained by performing the MTT (3-[4, 5-dimethylthiazol-2-yl]-2, 5 diphenyl tetrazolium bromide) assay (van Meerloo et al., 2011). Cells transfected with siRNA were washed twice with the phosphate-buffered saline (PBS), then treated with 1 ml MTT reagent (Sigma–Aldrich) and incubated at 37°C for 1 h. Subsequently, 1 ml dimethyl sulfoxide (Calbiochem) was added to the cells, which were further incubated for 15–20 min at room temperature with shaking. Finally, the absorbance was measured at 595 nm on Multiskan Ascent plate reader (LabSystems).

The culture medium harvested from infected cells was cleared off of cell-debris by low-speed centrifugation, supplemented with 0.3% BSA (Sigma–Aldrich) and titrated on MDCK cells to perform the microplaque assay. For this, confluent MDCK cell monolayers were infected with the 10-fold serial dilutions of culture medium as above. After removing the virus inoculum, cells were overlaid with serum-free MEM containing 1 μg/ml TPCK-trypsin (Sigma–Aldrich) and 0.8% Avicel (RC-581; FMC Biopolymer). The overlay was removed after 18–20 h incubation and cells were fixed with 4% formalin (Sigma–Aldrich) and subsequently permeabilized with 0.5% TritonX-100 in 20 mM glycine solution. Cells were then stained with mouse anti-NP antibody (1:1,000) followed by horseradish peroxidase-conjugated anti-mouse IgG antibody (1:1,000). The plaques were visualized by adding the True blue substrate (KPL biosciences).

The statistical analyses were performed using Prism 6 (GraphPad). The P-values were calculated using unpaired t-tests for pairwise data comparisons and one-way analysis of variance (ANOVA) or two-way ANOVA for multiple data set comparisons. A P-value of ≤ 0.05 was considered significant.

Recently, we described an antiviral role of host HDAC1 during IAV infection (Nagesh and Husain, 2016). The HDAC1 and HDAC2 are very similar proteins and share almost 82% amino acid identity as well as have some common domains and motifs (Brunmeir et al., 2009). It is believed that HDAC1 and HDAC2 originated from a common ancestor through a gene duplication event. Although HDAC1 and HDAC2 acquired some non-redundant and distinct biological functions during their evolution, they do retain some functional redundancy and complementarity (Brunmeir et al., 2009). Hence, we hypothesized that, like HDAC1, host HDAC2 also has an anti-IAV function and IAV dysregulates HDAC2 to undermine its antiviral function. To test this hypothesis, we first determined the HDAC2 transcript level in response to the IAV infection. Human lung epithelial A549 cells were infected with the influenza virus A/PR/8/34(H1N1) strain (hereafter referred to as PR8) or A/WSN/34(H1N1) strain (hereafter referred to as WSN) at an MOI of 0.5 and 5.0. After 24 h of infection, cells were processed to measure the HDAC2 mRNA level by quantitative real-time PCR (qPCR). In addition, the mRNA levels of three reference genes – 18S ribosomal RNA (18S RNA), glyceraldehyde 3-phosphate dehydrogenase (GAPDH), and actin were measured as the normalizing controls. Furthermore, the mRNA level of interferon-stimulated gene (ISG), viperin was measured as a positive control for infection as well as qPCR. We found that both PR8 (Figure 1A) and WSN (Figure 1B) downregulated the HDAC2 mRNA level in A549 cells in a statistically significant and dose-dependent manner. However, the extent of HDAC2 mRNA downregulation was dependent on the reference gene used for normalization. We found that when normalized to GAPDH, the infection of A549 cells with PR8 at an MOI of 0.5 and 5.0 caused a significant 40% (P = 0.0001) and 61.3% (P = 0.0001) reduction in HDAC2 mRNA level, respectively (Figure 1A). Likewise, when normalized to GAPDH, the infection with WSN at an MOI of 0.5 and 5.0 caused a significant 40.6% (P = 0.0001) and 60.6% (P = 0.0001) reduction in HDAC2 mRNA level, respectively (Figure 1B). Similarly, when normalized to actin, the HDAC2 mRNA levels in A549 cells were found to be reduced by a significant 36% (P = 0.0001) and 57.3% (P = 0.0001) when infected with the PR8 at an MOI of 0.5 and 5.0, respectively (Figure 1A), and a significant 35% (P = 0.0003) and 57.6% (P = 0.0001) when infected with the WSN at an MOI of 0.5 and 5.0, respectively (Figure 1B). However, when normalized to 18S RNA, the reduction in HDAC2 mRNA levels in A549 cells was found to be 72% (P = 0.0001) and 84.6% (P = 0.0001) after infection with PR8 at an MOI of 0.5 and 5.0, respectively (Figure 1A), and a profound 88% (P = 0.0001) and 95% (P = 0.0001) after infection with WSN at an MOI of 0.5 and 5.0, respectively (Figure 1B). The expression level of viperin is known to increase in response to virus infection (Fitzgerald, 2011). Consistent with this, there was a significant and dose-dependent increase in viperin mRNA levels in response to infection with both PR8 and WSN when normalized to 18S RNA (Figure 1C) as well as GAPDH and actin (Supplementary Figure 1). These data indicated that IAV downregulates HDAC2 mRNA level in H1N1 strain-independent manner.

After detecting a significant reduction in HDAC2 mRNA level in IAV-infected cells, we next sought to determine whether HDAC2 polypeptide level is also downregulated in IAV-infected cells. A549 cells were infected with WSN at an MOI of 0.5, and the infected and corresponding uninfected cells were harvested after 0, 6, 12, and 24 h of infection (‘0 h’ is the time point when virus inoculum is removed after 1 h incubation – see Materials and Methods). The total cell lysates were prepared and equal protein amounts were resolved on SDS-PAGE to detect the HDAC2, viral nucleoprotein (NP), and PDI polypeptides by western blotting (WB). The viral NP was detected as an infection marker and the PDI was detected as a control. We found that the level of HDAC2 polypeptide (∼60 kDa) in WSN-infected cells decreased with the progress of infection, peaking at 24 h; whereas, it remained largely unchanged throughout in uninfected cells (Figure 2A). To quantitate this decrease in HDAC2 polypeptide level, the intensity of all HDAC2 and PDI bands was quantified using the Image Studio Lite software (Version 5.0, LI-COR), and the amounts of HDAC2 was normalized with the corresponding PDI amounts. Subsequently, the amounts of HDAC2 at 0 h in uninfected and infected cells were considered 100% to compare its amount in respective samples at subsequent time points. We found that there was a statistically non-significant 24.7% and a significant 47.3% (P = 0.02) decrease in HDAC2 polypeptide levels in infected cells after 12 and 24 h of infection, respectively (Figure 2B). However, in uninfected cells, there was no significant change in the HDAC2 polypeptide levels between 0 and 24 h (Figure 2B). To determine whether, like mRNA, the reduction in HDAC2 polypeptide level was independent of IAV strain, A549 cells were infected with PR8 at an MOI of 0.5 or 5.0. Cells were then harvested after 24 h and processed to detect HDAC2, NP and actin polypeptides by WB. Here, actin was detected as a control. We found that, like WSN, infection with PR8 also reduced the HDAC2 polypeptide level in A549 cells, and in a dose-dependent manner (Figure 2C). Both PR8 and WSN are considered as the prototypic IAV lab strains and have been used extensively to study IAV biology in the host (Coombs et al., 2010; Tse et al., 2013). Furthermore, because of its pathogenic nature, WSN is also considered by some as a model for highly pathogenic influenza viruses (Tse et al., 2013). Nevertheless, we next infected A549 cells with influenza virus A/New Caledonia/20/1999(H1N1) strain – a more clinically relevant IAV strain (Hartmann et al., 2015), at an MOI of 0.5 or 5.0 for 24 h to assess the downregulation of HDAC2 polypeptide. Consistent with the results obtained with WSN and PR8, New Caledonia strain also reduced the level of HDAC2 polypeptide in A549 cells in a dose-dependent manner (Figure 2D). Finally, we infected A549 cells with the UV-irradiated WSN inoculum (0.5 MOI) to confirm that a replication-competent IAV was needed to downregulate the HDAC2 polypeptide levels (Figure 2E).

Above data indicated that IAV downregulates host HDAC2 expression, which potentially occurs mainly at the polypeptide level. Because, unlike mRNA levels, the HDAC2 polypeptide levels were reduced by all three IAV strains used in a similar manner. This prompted us to investigate whether IAV downregulates HDAC2 polypeptide level by promoting its degradation in infected cells. Primarily, two pathways, one governed by the proteasome and the other by lysosome, mediate the degradation of eukaryotic proteins. To identify the pathway employed by IAV to promote the degradation of HDAC2 polypeptide in infected cells, A549 cells were infected with WSN as above and treated with proteasome-inhibitor MG132 or lysosome-inhibitor NH₄Cl using previously reported and tested concentrations of 10 μM and 20 mM, respectively (Wang et al., 2014; Nagesh and Husain, 2016). The levels of HDAC2 polypeptide were then analyzed, quantified and normalized with the corresponding actin levels as described above. We found that MG132 treatment rescued the downregulation of HDAC2 polypeptide in infected cells almost to the level of uninfected cells (Figure 3A). However, NH₄Cl treatment did not have any effect on the downregulation of HDAC2 polypeptide in infected cells (Figure 3A). Compared to the corresponding uninfected cells, there was a significant 40.2% (P = 0.003) and 48% (P = 0.001) decrease in HDAC2 polypeptide levels in untreated and NH₄Cl-treated infected cells, respectively (Figure 3B), which was consistent with the data presented in Figure 2 above. In contrast, there was basically no decrease in the level of HDAC2 polypeptide in MG132-treated infected cells compared to the corresponding uninfected cells (Figure 3B). An increase in the level of ubiquitinated proteins detected in MG132-treated cells (Supplementary Figure 2B) and a reduced level of NP cleavage product (Supplementary Figure 2A, asterisks) detected in NH₄Cl-treated infected cells confirmed the potency of MG132 and NH₄Cl, respectively. The NP is known to be cleaved by caspases (Zhirnov et al., 1999), activation of which is known to be induced and inhibited by the lysosomes and NH₄Cl, respectively (Appelqvist et al., 2013). Furthermore, the treatment with caspase 3 inhibitor also did not rescue the HDAC2 polypeptide level in infected cells (Supplementary Figure 3). Therefore, these data indicated that IAV promotes the degradation of HDAC2 polypeptide via host proteasomal pathway.

Above data demonstrated that IAV downregulates the level of HDAC2 polypeptide in infected cells by at least 40%. This indicated a potential anti-IAV role for HDAC2 and by downregulating its expression, IAV is potentially minimizing the antiviral effect of HDAC2. If this is true, then IAV should grow to higher titers in the cells that are further depleted with HDAC2 expression. To assess this, IAV growth kinetics was analyzed in A549 cells with depleted expression of HDAC2, and RNA interference was employed to achieve this. First, to obtain a significant knockdown of HDAC2 expression, 10, 25, 50, and 100 nM of a siRNA targeting human HDAC2 and 50 nM of a non-targeting control siRNA were delivered to A549 cells, and subsequently the knockdown of HDAC2 expression was analyzed by WB. A significant knockdown of HDAC2 expression was achieved with all siRNA concentrations used (Figure 4A). However, the 10 nM concentration was chosen for subsequent experiments as it was adequate to knockdown the HDAC2 expression by 90% (Figure 4B), and with no cytotoxicity compared to the control siRNA (Figure 4C). Next, A549 cells transfected with either non-targeting control siRNA or HDAC2-targeting siRNA were infected with WSN at an MOI of 0.5, and the culture medium and the infected cells were separately harvested after 2, 6, 12, and 24 h of infection. The cells were processed and subjected to WB to confirm the knockdown of HDAC2 expression; whereas, the culture media were titrated by microplaque assay to measure the amount of released infectious IAV progeny. Indeed, we found that IAV grows to higher titers and exhibits a faster growth characteristic in HDAC2-depleted cells (Figure 4D). After 12 and 24 h of infection, the cells with depleted expression of HDAC2 released a statistically non-significant 2.6-fold and a significant 4.2-fold (P < 0.0001) more infectious progeny, respectively, than the control cells at the corresponding times (Figure 4D). The WB analysis of cells confirmed the significant knockdown of HDAC2 expression at each time point (Figure 4E). In addition, the downregulation of HDAC2 polypeptide in control siRNA-transfected cells after 24 h of infection was consistent with the data presented in Figures 2, 3. Furthermore, a significant 1.9-fold (P = 0.02) increase in intracellular NP level in HDAC2-depleted cells compared to control cells was also observed after 24 h infection by WB (Figure 4E). A fluorescence microscopy experiment was carried out to determine whether such increase was due to an increase in the NP level per cell or overall increase in the number of infected cells. The data obtained seem to indicate the latter (Supplementary Figure 4); however, further investigations involving the single-cell analyses are needed to ascertain this.

The data presented so far indicated that endogenous HDAC2 possesses anti-IAV properties. Next, we aimed to understand the antiviral mechanism of HDAC2. Hosts have evolved complex and sophisticated innate and adaptive mechanisms to counter virus infections. An antiviral state is created in the host cells as soon as they encounter an IAV particle and results in the secretion of type I IFNs (Garcia-Sastre, 2011). Secreted type I IFNs then engage the type I IFN receptor on host cells in an autocrine and paracrine manner and activate the cytoplasmic Janus kinases (JAKs), which, in turn, phosphorylate the cytoplasmic transcription factor Signal Transducer and Activator of Transcription I (STAT1). The phosphorylated STAT1 is then translocated to the nucleus, where it forms a transcription complex called as interferon-stimulated gene factor 3 (ISGF3). Subsequently, ISFG3 binds to interferon-stimulated response element (ISRE) of the ISGs in a sequence-specific manner and induces the expression of plethora (over 300) of ISGs. Many of these ISGs have been shown to inhibit IAV infection by targeting various steps of its life cycle (Garcia-Sastre, 2011). Previously, by mostly using the purified IFN-alpha or infection with viruses other than IAV, it has been demonstrated that class I and class II HDACs activity is required to create an antiviral state and activation of the type I IFN-mediated response and expression of ISGs (Nusinzon and Horvath, 2003; Chang et al., 2004; Sakamoto et al., 2004). However, a precise role of individual class I and class II HDACs in IFN response and subsequent expression of their effector proteins is yet to be elucidated. We recently demonstrated a critical role of class I and class II HDACs activity in the induction of IAV-mediated phosphorylation of STAT1 and expression of ISGs such as IFITM3, ISG15 and viperin (Nagesh and Husain, 2016). Furthermore, we also demonstrated that HDAC1 is important for the IAV-induced expression of viperin (Nagesh and Husain, 2016). Therefore, to get insight into the antiviral mechanism of HDAC2, we first investigated its role in the induction of IAV-mediated STAT1 phosphorylation and expression of viperin. To accomplish this, A549 cells transfected with the non-targeting control siRNA or HDAC2-targeting siRNA were infected with WSN at an MOI of 0.5. Cells were then harvested after 2, 6, 12, and 24 h of infection, processed, and levels of phosphorylated STAT1 (pSTAT1) and viperin were analyzed by WB. The phosphorylation of STAT1 was first detected in control siRNA-transfected cells within 2 h of infection and peaked at 12 h of infection, then subsided at 24 h of infection almost to the level of 2 h (Figure 5A); potentially due to the antagonism by IAV or caspase-mediated cleavage of total STAT1 (King and Goodbourn, 1998). However, the kinetics of the extent of STAT1 phosphorylation was slower in the absence of HDAC2. A noticeably decreased levels of pSTAT1 were detected in the cells transfected with HDAC2 siRNA compared to the cells transfected with control siRNA after 6 and 12 h of infection (Figure 5A). A parallel time-course experiment with UV-irradiated WSN inoculum was carried to rule out the presence of contaminating interferons in inoculum inducing the STAT1 phosphorylation (Supplementary Figure 5). On the other hand, the expression of viperin was first detected in control siRNA-transfected cells after 12 h of infection and peaked at 24 h of infection. However, similar to pSTAT1, the expression kinetics of viperin was slower in the absence of HDAC2. Viperin was barely detectable in HDAC2 siRNA-transfected cells after 12 h of infection, and after 24 h of infection, it was detected at a reduced level compared to the control siRNA-transfected cells (Figure 5A). To quantify the decrease in pSTAT1 and viperin levels in the absence of HDAC2, we repeated this experiment at least two more times and quantitated their levels at 12 h of infection. We found that there was a significant 41% (P < 0.01)) decrease in pSTAT1 level in HDAC2-depleted cells compared to the control cells (Figures 5B,C). Similarly, about 53% less (P = 0.001) viperin was detected in HDAC2-depleted cells than in the control cells (Figures 5D,E). These data indicated that HDAC2 is an important component of IAV-induced host innate antiviral response.