Downregulation of A20 Expression Increases the Immune Response and Apoptosis and Reduces Virus Production in Cells Infected by the Human Respiratory Syncytial Virus

Human respiratory syncytial virus (HRSV) causes severe lower respiratory tract infections in infants, the elderly, and immunocompromised adults. Regulation of the immune response against HRSV is crucial to limiting virus replication and immunopathology. The A20/TNFAIP3 protein is a negative regulator of nuclear factor kappa B (NF-κB) and interferon regulatory factors 3/7 (IRF3/7), which are key transcription factors involved in the inflammatory/antiviral response of epithelial cells to virus infection. Here, we investigated the impact of A20 downregulation or knockout on HRSV growth and the induction of the immune response in those cells. Cellular infections in which the expression of A20 was silenced by siRNAs or eliminated by gene knockout showed increased inflammatory/antiviral response and reduced virus production. Similar results were obtained when the expression of A20-interacting proteins, such as TAX1BP1 and ABIN1, was silenced. Additionally, downregulation of A20, TAX1BP1, and ABIN1 increased cell apoptosis in HRSV-infected cells. These results show that the downregulation of A20 expression might contribute in the control of HRSV infections by potentiating the early innate immune response and increasing apoptosis in infected cells.


Introduction
Human respiratory syncytial virus (HRSV), which belongs to the Pneumoviridae family, is an important respiratory pathogen that affects the general population. However, severe infections, including bronchiolitis and pneumonia, occur mainly in infants, the elderly, and immunocompromised adults [1,2]. In children under five, HRSV produces more than 33 million lower respiratory tract infections globally each year, of which approximately 3 million require hospitalization, and 0.2% die [3,4]. Additionally, severe pediatric HRSV infections are associated with the development of long-term sequelae, such as wheezing and asthma [5]. HRSV also contributes to the exacerbation of asthma and chronic obstructive pulmonary disease in adults [6,7].
Despite its significant impact on global health, there is no vaccine or effective treatment against HRSV. For a long time it has been thought that an exuberant inflammatory response was behind

Viral Infections and Plaque Assays
A549 subconfluent monolayers were infected with HRSV at a multiplicity of infection (MOI) of 3 plaque-forming units (pfu) per cell in DMEM with 2% FBS (DMEM2) and incubated for 90 min at 37 • C. After this time, the inoculum was removed, and fresh DMEM2 was added. Samples (culture supernatants and cell pellets) were collected at different times post-infection.
HRSV titers were determined in HEp-2 cell monolayers inoculated with serial dilutions of the culture supernatants for 90 min at 37 • C and then overlaid with 0.7% agarose in DMEM2. Five days post-infection (dpi), the cell monolayers were fixed with 4% formaldehyde and permeabilized with methanol. Plaques were visualized by one-hour incubation with a mixture of monoclonal antibodies previously obtained in our laboratory against the HRSV glycoprotein G, glycoprotein F, and phosphoprotein [29], followed by one-hour incubation with an anti-mouse IgG horseradish peroxidase linked whole antibody (Abcam, Cambridge, UK), and 3-amino-9-ethylcarbazole (AEC, Alfa Aesar, Ward Hill, MA, USA). Virus plaques, which were visible to the naked eye, were counted and HRSV titers calculated.
Gene expression was normalized to ACTB, and relative quantifications were determined by the comparative CT (∆∆CT) method.

siRNA Silencing
A549 cells were seeded 24 h before transfection at a density of 4.5 × 10 4 cells per well in 24-well plates. Twenty-four hours later, cells were transfected with 6 pmol of control small interfering RNAs (siRNAs) (negative control #2) or specific siRNAs against TNFAIP3 (A20, ID # s14259), TAX1BP1 (ID # s16984), ABIN1 (TNIP1, ID # s20174), ITCH (ID # s38163) and RNF11 (ID # s25671) (all siRNAs were purchased from Ambion-Thermo Fisher, Rockford, IL, USA) and 1 µl of Lipofectamine RNAiMAx reagent (Invitrogen) per well. Twenty-four hours after transfection, cells were infected with HRSV at a MOI of 3. Cell supernatants for viral titration and cell pellets for RNA and protein extraction were harvested at different hours post-infection (hpi), as indicated in the figure legends.

A20 Knockout A549 Cells
Two clones of A20 knockout (A20 -/-) A549 cells were produced using the Transcription Activator-Like Effector Nucleases (TALENs) technology [33] and the plasmids Human-H27583_TNFAIP3_TALEN-L and Human-H27583_TNFAIP3_TALEN-R (Talen Library Resource, Seoul National University), as previously described [31]. In brief, A549 cells (parental cells) were transfected with those plasmids encoding sequence-specific DNA-cleaving nucleases against A20. Three days after transfection, the cells were trypsinized and cloned by limiting dilution in 96-well plates at a density of 1 cell per well. Single-cell clones were selected, cloned a second time by limiting dilution, and expanded to generate stocks. Screening for A20 -/clones was done by PCR amplification and DNA sequencing using the following primers: forward (5 -CCTTTGCAACATCCTCAGAAG-3 ) and reverse (5 -ACTAACCAAGCAAGTCACAGAAC-3 ). A20 -/clones were checked by Western blotting using an A20 specific antibody (Ab167154, Abcam, Cambridge, UK). Two A20 -/clones (KO-1 and KO-2) were selected. In addition, one wild-type (A20 +/+ ) clone (WT-1) that underwent the same process of transfection and cloning, but in which the A20 gene was not disrupted, was selected as control (Supplementary Materials, Figure S1).
For overexpression assays, 1.6 × 10 5 A549 cells were plated in each well of a 12-well plate and incubated for 24 h before transfection. Cells were then transfected with 1 µg of the purified plasmid (EndoFree Plasmid Maxi Kit, Qiagen, Hilden, Germany) and 3 µl of Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA) per well. Twenty-four hours after transfection, the cells were infected with HRSV at a MOI of 3. Cell supernatants for viral titration and cell pellets for RNA and protein extraction were collected at different hpi.

Apoptosis Assay
Cell apoptosis was measured by the Single Channel Dead Cell Apoptosis Kit with Annexin V-Alexa-fluor488 and Sytox green dyes from Thermo Fisher (cat. # V13240) following the manufacturer's instructions. Fluorescence was detected in a BD FACSCanto TM Flow Cytometer (BD Biosciences, San José, CA, USA) and analyzed by the FlowJo TM software (BD Biosciences).

Stranded mRNA-seq Library Preparation and Sequencing
Parental A549 cells and KO-1 and KO-2 cell lines were infected with HRSV at a MOI of 3. Twenty-four hours later, total RNA was extracted from infected cells with the ReliaPrep RNA Cell Miniprep System (Promega, Madison, WI, USA) following the manufacturer's instructions. RNA was quantified with a BioPhotometer Plus (Eppendorf, Hamburg, Germany) and RNA quality was assessed with the 2100 Bioanalyzer RNA NANO assay (Agilent Technologies, Santa Clara, CA, USA). All samples had an RNA integrity number (RIN) higher than nine (Supplementary Materials, Figure S2). Libraries were prepared using the TruSeq Stranded mRNA Sample Prep Kit v2 (ref. RS-122-2101/2, Illumina) according to the manufacturer's protocol. Briefly, 500 ng of total RNA was used for poly(A)-mRNA selection using Oligo (dT) magnetic beads and were subsequently fragmented to approximately 300bp. cDNA was synthesized using reverse transcriptase (SuperScript II, ref. 18064-014, Invitrogen) and random primers. The second strand of the cDNA incorporated dUTP in place of dTTP. Double-stranded DNA was further used for library preparation. dsDNA was subjected to A-tailing and ligation of the barcoded Truseq adapters. All purification steps were performed using AMPure XP Beads (ref. A63881, Beckman Coulter). Library amplification was performed by PCR using the primer cocktail supplied in the kit.
Final libraries were analyzed using Agilent DNA 1000 chip to estimate the quantity and check size distribution and were then quantified by qPCR using the KAPA Library Quantification Kit (ref. KK4835, KapaBiosystems) before amplification with Illumina's cBot. Libraries were sequenced using 50 base read lengths in single-end mode (1 × 50) on Illumina's HiSeq 2500.

Data Analysis
The obtained RNA-Seq data were analyzed by the Bioinformatics Facility of the Instituto de Salud Carlos III (ISCIII). First, quality control analysis involving FastQC (v0.11.3) [38] was carried out, and any adapter sequences, as well as low-quality 3 ends were removed using Trimmomatic v0.36 [39]. A sliding window quality filtering approach was performed, scanning from the 5 end of the read, and removing the 3 end of the read when the average quality dropped below a Q score of 15. Reads lower than 50 nucleotides in length were removed.
Transcriptome prediction and gene/isoform quantification were calculated using Cufflinks (v2.2.1) based on Hg38 Ref-Seq reference genes [44,45]. Finally, differential expression analysis was carried out using Cuffdiff, which uses Benjamini-Hochberg correction to compute the False Discovery Rates (FDR) (q-value). Differentially expressed genes (DEGs) were considered when FDR values were ≤0.05. The CummeRbund package (v2.14.0) was used for quality control and visualization of the results.

Functional Annotation of Candidate Genes
KO cells and parental A549 cells were infected at a MOI of 3. At 24 hpi, RNA was extracted and processed for sequencing. Data from three independent biological experiments were used to generate two lists of differentially expressed genes (FC ≥ 1.5, q ≤ 0.05) by comparing KO-1 or KO-2 cells vs. the A549 parental cells. These two lists were compared and the genes that were upregulated in both KO lines were selected (Supplementary Materials, File S1). Functional annotation of selected genes was performed using DAVID (Database for Annotation, Visualization and Integrated Discovery) Bioinformatics Resources 6.8 [46,47], applying an EASE score threshold of 0.1 and a count threshold of 2.

Statistical Analysis
Pairwise comparisons between control and test conditions were done by using a Student's t-test to determine statistically significant differences. Statistics were calculated with the Graphpad Prism 8 software. From the p values obtained, q-values (FDR) were calculated using the Benjamini-Hochberg method and indicated in the figures as * (q-values < 0.05) and ** (q-values < 0.01). Means and standard deviations from at least three independent experiments are represented in the figures.

HRSV Infection Induces High Levels of A20 Expression
To explore if A20 expression was upregulated during HRSV infection, A549 cells were infected, and the levels of A20 and HRSV nucleoprotein (N) RNAs were measured by qRT-PCR at different times post-infection. A large amount of A20 mRNA was produced in infected cells at later times post-infection (36-48 h post-infection) ( Figure 1A). The expression of A20 was delayed with respect to HRSV N. While levels of HRSV N started to increase after 3-6 hpi, the accumulation of A20 mRNA was only detected following 16-24 hpi, reaching levels of about 500 times over uninfected cells by 48 hpi ( Figure 1A). To confirm that the upregulation of A20 mRNA correlated with an increase at the protein level, the amounts of A20 protein were analyzed by Western blotting at different times post-infection. The results confirmed that A20 protein was accumulated in HRSV infections after 24 hpi ( Figure 1B).

Downregulation of A20, TAX1BP1, or ABIN1 Enhances Intracellular Immune Response and Decreases HRSV Virus Production
Since A20 is a negative regulator of NF-κB and IRF3/7 transcription factors, which are key factors in the inflammatory and antiviral responses, we analyzed the impact of A20 downregulation on the expression of different cytokines and chemokines in cells infected with HRSV. Also, downregulation of some A20-interacting proteins that modulate A20 activity and specificity was tested. Gene silencing was confirmed by RT-PCR and the results are shown in Supplementary Materials, Figure S3.
siRNA-mediated silencing of A20 led to an increase in the expression at 24 hpi of all immune genes tested, except CCL5 ( Figure 2). Similar results were obtained when TAX1BP1 or ABIN1 were downregulated (Figure 2), in accordance with previous reports indicating that A20, TAX1BP1, and ABIN1 collaborate to inhibit antiviral signaling [48]. However, only ISG15, TNF-α and IL-1β were overexpressed in ITCH-silenced cells, and none of the immune genes tested was upregulated in RNF11-silenced cells (Figure 2). In contrast, the expression of the majority of the immune genes tested was slightly downregulated in all cases at 48 hpi ( Figure 2).
To test if the downregulation of these mRNAs had any effect on HRSV growth, virus titers in the supernatant of silenced cells were measured. Again, A20, TAX1BP1, and ABIN1 silencing reduced virus titers significantly (Figure 3 and Supplementary Materials, Table S1). RNF11 downregulation did not affect HRSV titers, but ITCH downregulation reduced virus titers by approximately ten times ( Figure 3 and Supplementary Materials, Table S1).

HRSV Production is Impaired in A20 Knockout Cells While the Antiviral/Inflammatory Response Is Augmented in These Cells
To confirm that A20 is involved in the regulation of the early innate immune response against HRSV, two A20 knockout lines were generated from A549 cells (KO-1 and KO-2). As a control, one wild-type cell line was generated via the same process of cloning (WT-1). The lack of expression of A20 protein in KO cells was checked by Western blotting (Supplementary Materials, Figure S1).
Cell lines were infected with HRSV, and the expression of inflammatory/antiviral genes was compared to infected uncloned A549 wild-type cells. Similar to what happened in the A20 silenced cells, most of the immune genes tested were significantly upregulated in both KO lines at 24 hpi ( Figure 4). Only IL-1β was downregulated at this time in the KO-2 line (Figure 4). At 48 hpi, however, the majority of genes not only were not upregulated, but were even downregulated (Figure 4), also resembling the results observed in A20 silenced cells (Figure 2). Only IL-6 and IL-8 were significantly upregulated at 48 hpi in KO-2 ( Figure 4). Unlike the KO lines, the wild-type line (WT-1) showed decreased expression of the immune genes at 24 hpi ( Figure 4) and variable expression at 48 hpi ( Figure 4). From the p values obtained, FDRs were calculated: *, q-value < 0.05 and **, q-value < 0.01. KO-1 and KO-2, A549 cell lines that do not express A20. WT-1, wild-type A549 cell line obtained following the same process as with KO-1 and KO-2. Fold change = 1 means that there is no expression change with respect to the control (A549 cells) and is represented by a dotted line.
To further confirm the role of A20 in the regulation of the early innate immune response following HRSV infection, we restored the expression of A20 in the KO cell lines by transfection with plasmids encoding this protein. As expected, the restoration of A20 expression in KO-2 cells decreased the expression of most of the antiviral/inflammatory genes at 24 hpi when compared to KO-2 cells transfected with a control plasmid (empty vector) (Supplementary Materials, Figure S4C). However, this effect was not observed in the KO-1 cells (Supplementary Materials, Figure S4B), probably because, for unknown reasons, the expression of A20 in KO-1 cells was much lower than in KO-2 cells (Supplementary Materials, Figure S4A).
Since A20 silencing led to a decrease in HRSV titers, the production of infectious HRSV was analyzed in the KO lines. Like with A20 silenced cells, virus titers were reduced in both KO lines (from two to three times) when compared to A549 cells and the WT-1 line at 48 hpi ( Figure 5 and Supplementary Materials, Table S2).

Immune Response and Infection-Related Signaling Pathways Are Upregulated in A20 Knockout Cells Following HRSV Infection
To further confirm that A20 is involved in the regulation of the intracellular immune response against HRSV, uncloned wild-type A549 cells and the A20 KO lines were infected with the virus, and the expression of cellular genes was analyzed by RNAseq. Data from three independent biological experiments (infections) were analyzed to obtain a list of genes that were upregulated (FC ≥ 1.5; q ≤ 0.05) in both KO-1 and KO-2 cell lines (Supplementary Materials, File S1). This gene list was analyzed using the DAVID Functional Annotation Tool (see Material and Methods). Gene Ontology Biological Processes (GO-BP) related to immune response were over-represented in both KO lines when compared to A549 cells (Table 1). Similarly, pathways related to immune response and anti-microbial response were also over-represented in the KO lines (Table 2).  Table 2. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways over-represented in the genes that are up-regulated in both the infected KO-1 and KO-2 cells.
Gene silencing showed that downregulation of A20, TAX1BP1, ABIN1, and ITCH but not RNF11, increased apoptosis in cells infected with HRSV at later times post-infection ( Figure 6A). Similarly, enhanced apoptosis was observed in the A20 KO-1 cell line following virus infection ( Figure 6B). However, no differences were detected between the KO-2 cell line and wild-type cells ( Figure 6B).

Discussion
In this study, we show that A20 is involved in the regulation of the innate immune response and apoptosis in HRSV-infected epithelial cells. Downregulation of A20 by specific siRNAs or the lack of A20 expression in KO cells led to increased inflammatory/antiviral response and apoptosis after HRSV infection. A decrease in virus titers was also observed. At least two A20-interacting proteins, TAX1BP1 and ABIN1, seem to cooperate in that regulation since similar effects were observed in cells in which the expression of those proteins was downregulated by siRNA treatment.
The induction of the inflammatory/antiviral response in HRSV-infected epithelial cells mostly depends on the activation of the retinoic acid-inducible gene I (RIG-I) receptor by viral RNA and subsequent activation of the NF-κB and IRF3 transcription factors [55]. For example, the activation of both NF-κB and IRF3 by HRSV infection, as well as the expression of several cytokines and antiviral genes, appear to be inhibited when RIG-I is downregulated by siRNAs [55] (our unpublished results). A20 regulates the activity of NF-κB and IRF3/7 in response to multiple stimuli, including RIG-I activation by viral RNAs [22]. A20 interrupts RIG-I signaling at several junctures by ubiquitination/deubiquitination of some proteins that participate in the signaling pathway [24,56]. Non-catalytic mechanisms, such as A20 competing with other ubiquitin-binding proteins like NEMO to limit IKK phosphorylation and downstream NF-κB activation have also been observed [57][58][59][60]. Some of these roles of A20 require that the interactions with partner proteins are functional. For instance, A20 cooperates with TAX1BP1 and ABIN1 to disrupt the TRAF3-TBK1-IKKε complex, thereby inhibiting IRF3 activation [48,61]. In agreement with this, we observed the upregulation of the innate immune genes in HRSV-infected cells silenced for A20, TAX1BP1 or ABIN1, showing the importance of those proteins in regulating the inflammatory/antiviral immunity against this virus [24]. In contrast, RNF11 silencing did not result in increased expression of immune genes, apoptosis, or HRSV growth, indicating that this protein does not collaborate with A20 in the regulation of these processes, despite the fact that RNF11 seems to form a complex with A20, TAX1BP1, and ITCH to inhibit TNF and LPS-induced NF-κB signaling [62]. Results from ITCH silencing are less easy to interpret since its impact on the innate immune response against HRSV appeared to be less than that of A20, TAX1BP1 or ABIN-1 silencing, while the reduction in viral titers was similar or even higher. It has been shown that TAX1BP1 recruits ITCH to the mitochondrial adaptor MAVS to promote its ubiquitination and degradation and thus restricts virus-induced apoptosis [50]. The increased apoptosis seen in ITCH-silenced cells may explain, at least in part, the reduced HRSV replication observed in our study, as discussed below.
Overall, the results obtained on A20 silencing were confirmed in the studies on HRSV infections of A20 KO cells. However, the two A20 KO lines differed significantly from each other and from A20-silenced wild-type cells in regards to the innate immune response after HRSV infection. It should be taken into account that the KO lines were established from a single cell after multiple rounds of replication. In this process, different epigenetic changes may have occurred in the two KO lines to compensate, to some extent, for the A20 deficiency.
Overexpression experiments were conducted to confirm the results obtained in KO cells and via siRNA silencing. Although we transfected cells with single, or triple combinations of plasmids expressing A20, TAX1BP1, or ABIN1, we were unable to see differences from cells transfected with control plasmids in regards to the expression of immune genes or HRSV replication (Supplementary Materials, Figure S5). Overexpression experiments are more challenging to perform than siRNA silencing, especially when several genes have to be expressed in a single cell to reconstruct a protein complex. Furthermore, since A20 can interact with multiple partners to carry out different functions [34,36,62,63], other unidentified proteins, in addition to A20, TAX1BP1, and ABIN1, may form part of the complex required for the regulation of the intracellular innate immune response against HRSV.
In the present study, downregulation of A20, TAX1BP1, and ABIN1 expression led to an increased immune response in HRSV-infected cells at 24 hpi, but not at 48 hpi. There are at least two possible explanations for this result. First, although A20 negatively regulates the canonical NF-κB pathway, it has been reported that it also activates the non-canonical pathway by a non-catalytic mechanism [64]. In that study, the authors propose that A20 may be involved in the molecular switch from canonical to non-canonical activation of NF-κB pathways. The non-canonical route is activated in HRSV-infected cells [65], and we have shown that genes involved in this pathway, including NFKB2, RELB, and NIK, are upregulated at later times post-infection [29]. It is possible that at 48 hpi, the activation of the canonical pathway due to a lack of A20-related activity in silenced cells may be compensated by the inhibition of the non-canonical route. However, it has not been elucidated if the activation of the non-canonical pathway by A20 requires the cooperation of TAX1BP1 and ABIN1. To add to the complexity, Liu et al. [66] showed that in addition to the canonical and non-canonical pathways, a "cross-talk pathway" operates in HRSV-infected cells downstream of RIG-I, leading to the activation of the NIK-IKKα complex and RelA release. Second, downregulation of A20, TAX1BP1 and ABIN1 decreased virus production at later times post-infection, and this may have contributed to reducing the intracellular immune response at this time. Inhibition of HRSV replication has been shown to decrease the inflammatory/antiviral response in infected cells [67].
We have shown that A20 is highly expressed at later times after HRSV infection of epithelial cells. This is in line with the idea that one of the central functions of A20 is to contribute to the termination of the immune response to avoid immune-mediated tissue damage. However, the upregulation of A20 may also be beneficial for the virus. Thus, an interesting observation of our study is that viral replication was reduced at 48 hpi in cells treated with siRNAs against A20 or A20-interacting proteins, as well as in A20 KO cells. This may be a result of the increased antiviral response observed in those cells. In line with this, it has been described that A20 promotes influenza virus replication by suppressing the antiviral response of infected cells [68]. Similarly, it has been shown that A20 promotes hepatitis C virus and human cytomegalovirus replication [69][70][71]. Additionally, the increased apoptosis observed in silenced and KO cells may have also contributed to reducing virus production. Apoptosis is a mechanism of programmed cell death that can function to restrict microbial growth [72]. In agreement with this, it has been reported that inhibition of apoptosis promotes HRSV replication [73,74]. Therefore, the strong upregulation of A20 in HRSV-infected cells may facilitate virus replication and persistence by decreasing antiviral response and apoptosis [75].

Conclusions
In conclusion, our data indicate that A20, probably with the collaboration of other protein partners, modulates the intracellular innate immune response and apoptosis in HRSV-infected epithelial cells, which affects virus production. Therefore, A20 and/or A20-interacting proteins may be potential targets to regulate two crucial aspects of HRSV infections, namely, virus growth and the early inflammatory/antiviral response.

Supplementary Materials:
The following are available online at http://www.mdpi.com/2076-393X/8/1/100/s1, Figure S1: Western blotting of A20-knockout cell lines, Figure S2: RNA quality parameters, Figure S3: Gene silencing of A20, TAX1BP1, ABIN1, ITCH, and RNF11 genes, Figure S4: Overexpression of A20 in A20-knockout cell lines, Figure S5: HRSV titers and immune gene expression in cells transfected with individual plasmids or a mixture of plasmids encoding A20, TAX1BP1 and ABIN1, Table S1: Raw values of virus titers corresponding to Figure 3. Data from three independent experiments, mean and standard deviation (SD) are shown, Table S2: Raw values of virus titers corresponding to Figure 5. Data from three independent experiments, mean and standard deviation (SD) are shown. File S1: List of genes upregulated in KO cells. Funding: This work was supported by grant PI15CIII/0024 from Instituto de Salud Carlos III to IM.