Truncation of C-terminal 20 amino acids in PA-X contributes to adaptation of swine influenza virus in pigs

The PA-X protein is a fusion protein incorporating the N-terminal 191 amino acids of the PA protein with a short C-terminal sequence encoded by an overlapping ORF (X-ORF) in segment 3 that is accessed by + 1 ribosomal frameshifting, and this X-ORF exists in either full length or a truncated form (either 61-or 41-condons). Genetic evolution analysis indicates that all swine influenza viruses (SIVs) possessed full-length PA-X prior to 1985, but since then SIVs with truncated PA-X have gradually increased and become dominant, implying that truncation of this protein may contribute to the adaptation of influenza virus in pigs. To verify this hypothesis, we constructed PA-X extended viruses in the background of a “triple-reassortment” H1N2 SIV with truncated PA-X, and evaluated their biological characteristics in vitro and in vivo. Compared with full-length PA-X, SIV with truncated PA-X had increased viral replication in porcine cells and swine respiratory tissues, along with enhanced pathogenicity, replication and transmissibility in pigs. Furthermore, we found that truncation of PA-X improved the inhibition of IFN-I mRNA expression. Hereby, our results imply that truncation of PA-X may contribute to the adaptation of SIV in pigs.

Scientific RepoRts | 6:21845 | DOI: 10.1038/srep21845 To determine the effects of the truncation of PA-X for adaptation of SIV in pigs, we used reverse genetics to construct two TR swine influenza viruses, only differing in PA-X. These included truncated PA-X and full-length PA-X, and we compared their biological characteristics in vitro and in vivo.

Results
Prevalence of SIVs with truncated or full-length PA-X. To explore the relationship between truncation of PA-X and viral adaptation in swine, we analyzed the frequency of the PA-X truncations in SIV by year based on all SIV sequences available in the National Center for Biotechnology Information (NCBI) Influenza Viruses Resource, regardless of subtype and lineage. As shown in Fig. 1, the PA-X proteins in all SIV cases were full-length before 1985, following which the number of SIVs containing truncated PA-X gradually increased. By 2015, more than 90% SIVs possessed truncated forms of PA-X. These data suggested that the truncation of PA-X protein may play an important role in the adaptation of SIV in pigs.

Generation of recombinant SIVs with truncated or full-length PA-X.
To investigate the effect of the length of PA-X on these biological properties, a reverse genetics system of wild type TR H1N2 SIV A/swine/ Guangdong/1222/2006 [Sw-41X(WT)] possessing 41 amino acids in the X domain of PA-X was established as previously described 10 . Meanwhile, the X-ORF of Sw-41X(WT) virus was extended to the full-length 61 amino acids X domain by substituting the stop codon to tryptophan at the 42nd amino acid position in X-ORF without altering PA ORF (Fig. 2). This PA-X-extended H1N2 virus was named Sw-61X.
In vitro properties of SIVs with truncated or full-length PA-X. The influence of the length of PA-X on the replicative ability of H1N2 SIV in vitro was evaluated by inoculating porcine kidney (PK15) cells with each virus at a multiplicity of infection (MOI) of 0.01 in the presence of 1 μ g/ml tosylsulfonyl phenylalanyl chloromethyl ketone-treated trypsin. As shown in Fig. 3A, the virus titers of Sw-41X(WT) were significantly higher than Sw-61X at 24 and 36 hours post-inoculation (hpi) (P < 0.05).
An ex vivo organ-culture model of the pig respiratory tract maintained at an air-liquid interface as a biologically relevant in vitro system was also used to study the replication of PA-X mutants as previously described 11 . Briefly, porcine nasal turbinate, trachea, and lung explants cultured in 12-well plates were inoculated with 10 6  The PA-X encodes a common N-terminal domain (amino acid residues 1 to 191) of PA fused to a unique C-terminal region (41 or 61 amino acids) produced by a + 1 reading frame of PA mRNA via ribosomal frameshifting. The red letters represent mutation sites in PA-X to produce a full-length PA-X.
Scientific RepoRts | 6:21845 | DOI: 10.1038/srep21845 TCID 50 of each virus, and virus titers in supernatants were subsequently tested at 24, 48, and 72 hpi. Sw-41X(WT) replicated more efficiently than Sw-61X in the explants of nasal turbinate at all three time points, and of trachea and lung at 48 hpi (P < 0.05) ( Fig. 3B-D). Collected, these data indicated that truncation of PA-X promoted viral production in porcine cells or swine respiratory tissues infected ex vivo.
Pathogenicity and transmissibility of SIVs with truncated or full-length PA-X in pigs. To determine whether truncation of PA-X influenced the pathogenicity and transmissibility of SIV in pigs, nine four-week-old landrace specific-pathogen-free piglets obtained from Beijing Center for SPF Swine Breeding & Management of each group were inoculated intranasally with 10 6 TCID 50 of each virus, and three infected pigs from each group were removed to a separate room with three naïve pigs at 24 hpi. The nasal turbinates, tracheas, and bronchoalveolar lavage fluids (BALFs) from three pigs in each inoculated group were collected for virus detection at 3 and 5 dpi. Clinical signs of the remaining three pigs were monitored daily throughout the duration of the study, and the nasal washes were collected at 3, 5, 7, and 9 days post-inoculation (dpi). Five naïve pigs were maintained as a control group and were mock-inoculated with 2 ml PBS, per nostril.
As shown in Fig. 4A, the mean rectal temperature in pigs inoculated with Sw-41X(WT) was significantly higher than that in Sw-61X infected pigs during 1 to 3 dpi (P < 0.05). Sw-41X(WT) caused lethargy and lack of appetite in two of three inoculated animals during 1 to 3 dpi, and nasal secretion and coughing were observed in one of them at 2 dpi. By contrast, infection with Sw-61X only produced slight lethargy and lack of appetite at 2 dpi in one infected animal. The virus titers of Sw-41X(WT) in nasal washes were significantly higher than Sw-61X at 3, 5 and 7 dpi (P < 0.05) (Fig. 4B,C). Furthermore, pigs infected with Sw-41X(WT) virus shed the virus for much longer than Sw-61X virus. At 7 dpi, Sw-41X(WT) was still detectable in two of three pigs with virus titers of 10 4.5 TCID 50 /ml and 10 3.5 TCID 50 /ml, respectively, but none in Sw-61X group had detectable virus by that time (Fig. 4B,C). For the contacted animals, viral shedding was detectable in all the pigs co-housed with Sw-41X(WT) with titers of 3.5 ± 0.3 TCID 50 /ml and 1.9 ± 0.3 TCID 50 /ml at 2 and 4 days post-contact (dpc), respectively; while virus was only detectable in two of three pigs in Sw-61X group at 2 dpc with titers of 10 1.8 TCID 50 /ml (Fig. 4B,C). Seroconversion was only detected in animals with viral shedding. Overall, SIV with truncated PA-X had increased pathogenicity and transmissibility compared with that with full-length PA-X.
Histopathology and immunohistochemistry of swine lungs infected by PA-X mutant viruses. To further examine viral pathology, three pigs from the infection group and one pig from the control group were anesthetized and their lungs were collected for evaluation of histopathology (H&E) and immunohistochemistry (IHC) at 5 dpi. Histopathological analysis showed that the Sw-41X(WT) virus caused moderate bronchopneumonia and consolidation as characterized by local edema, dropout of epithelial cells in bronchia, and infiltration of inflammatory cells; while only mild bronchopneumonia was observed for pigs infected by Sw-61X (Fig. 5A). Compared to Sw-61X, pigs exposed to Sw-41X(WT) showed increased SIV-positive signals in lungs (Fig. 5A). The average histopathological and immunohistochemical scores in the Sw-41X(WT) infected group were significantly higher than that of Sw-61X infected groups at 5 dpi (P < 0.05) (Fig. 5B,C).

Figure 4. Body temperatures, virus shedding and transmission in pigs.
Nine Pigs were intranasally inoculated with 10 6 TCID 50 of each virus. Five naïve pigs were maintained as a control group and were mockinoculated with PBS. After 24 h, the inoculated animals were housed together with three contact pigs. Tissues from three pigs in each inoculated group were collected for virus detection at 3 and 5 dpi, respectively. (A) Rectal temperature of infected piglets was taken at fixed time-point daily from 0 to 9 dpi. (B,C) Viral titers in nasal washes of three infected or co-housed piglets from each group. Each bar represents the virus titer from an individual animal. The horizontal dashed line corresponds to the TCID 50 assay detection limit. Data are presented as means ± SD of three pigs. *the value of Sw-41X(WT) was significantly different from Sw-61X (P < 0.05, ANOVA).
Scientific RepoRts | 6:21845 | DOI: 10.1038/srep21845 No microscopic lesions and SIV-positive signals were observed in the lungs of control pigs (Fig. 5B,C). Taken together, the above results further demonstrated that truncated PA-X increased the pathogenicity of SIV in pigs.

Replication of PA-X mutant SIVs in pigs.
Pathogenicity experiments in pigs demonstrated that SIV with truncated PA-X presented higher pathogenicity than full-length PA-X SIV. To determine if the high pathogenicity was consistent with enhanced replication in tissues, virus titers in nasal turbinate, trachea, and BALFs were evaluated from three pigs in each group at 3 and 5 dpi. For the viral replication in the tissues of inoculated pigs, the replication titers of Sw-41X(WT) in nasal turbinates, tracheas and BALFs were significantly higher than Sw-61X at 3 dpi and/or 5 dpi (P < 0.05) (Fig. 6A-C). These results suggested that PA-X with 41 amino acids in X-ORF lead to increased viral replication in vivo.
Inhibition of type I interferons expression by truncated or full-length PA-X. We next explored the potential mechanism behind the enhanced replicatibility and transmissibility virulence of SIV caused by PA-X truncation. The induction of type I interferon (IFN-I) establishes an antiviral state which impedes viral replication and spread in host cells 12 . Recently, the PA-X protein of the influenza virus was demonstrated to be involved in the function of host shutoff which could inhibit IFN-I expression 9 . To assess whether the increased virulence of SIV with truncated PA-X was related to the increased inhibition of IFN-I expression, the expression levels of IFN-α and IFN-β mRNA in porcine alveolar macrophages (PAMs) and lungs of pigs infected with each virus were analyzed by quantitative real-time PCR as previously described. In PAMs, the expression levels of IFN-α and IFN-β mRNA from Sw-41X(WT)-infected cells were significantly lower than Sw-61X-infected cells at 12 hpi (P < 0.05) (Fig. 7A). In the lungs of infected pigs, Sw-41X(WT) virus induced a lower level of IFN-I expression than Sw-61X at 3 dpi (P < 0.05) (Fig. 7B).
To further confirm the direct effect of PA-X on IFN-I production, we expressed the truncated or full-length PA-X proteins ectopically in 293T cells and then infected them with Sendai virus to stimulate an innate immune response. Activation of IFN-β promoter was determined by a luciferase mediated reporter assay 13 . We found that IFN-β levels in cells expressing full-length PA-X was two-fold higher than cells expressing truncated PA-X (P < 0.05) (Fig. 7C). In general, these findings indicated that suppressing IFN-I response may contribute to the enhanced virulence of SIV with truncated PA-X.

Discussion
Adaptation is thought to be the driving force in evolution, during which beneficial mutations are selected in nature because of increased replication and transmission in the new host. Pigs are proposed to be "mixing vessels" or opportune intermediate hosts for the generation of novel influenza viruses with pandemic potential 14,15 . Despite the pivotal role of pigs in the ecology of influenza A viruses, few virulence markers of influenza virus in pig are identified. Based on the fact that SIVs with truncated PA-X have been gradually increased and became dominan 5 , we investigate whether truncation of PA-X could adapt SIV for enhanced growth and transmission in pigs. We observed that truncation of PA-X promoted viral growth in PK15 cells and swine respiratory tissues infected ex vivo, increased the viral replication, pathogenicity and transmission in swine, and we further demonstrated that the altered ability of PA-X in suppressing IFN-I expression might responsible for these changes. These results suggested that SIV with truncated PA-X seemed to present a selective advantage in pigs, which is in accordance with the fact that the proportion of SIVs possessing truncated PA-X has continually increased and have become predominant in nature. In contrast, our previous study found that the 2009 H1N1, H5N1, and H9N2 influenza viruses with full length PA-X showed higher replication levels in A549 cells than those with truncated PA-X, and that virus with full length PA-X enhanced viral replication and pathogenicity in mice 16 . Coincidently, all of the avian and human influenza viruses possessed full-length PA-X in nature. Therefore, we speculate that the role of length of PA-X is associated with host specificity.
IFN-I consist of several structurally related IFN-α proteins and a single IFN-β protein, together with interferon-stimulated genes (ISGs), they establish an antiviral state in infected cells that function to inhibit viral replication and restrict viral spread 17 . To escape the antiviral response of the host, several viral proteins have been shown to suppress IFN-I expression, including vhs from herpes simplex virus (HSV) and nsp1 from severe acute respiratory syndrome-related coronavirus (SARS-CoV) or mouse hepatitis virus (MHV), that is necessary for efficient virus growth and virulence [18][19][20][21][22][23][24][25][26] . For the influenza virus, the newly identified PA-X protein, along with NS1, PB2, PB1, PA and PB1-F2 have been demonstrated to be involved in the IFN-I response modulation 27 . The most prominent IFN-I antagonist in influenza virus is the NS1 protein which can suppress the IFN-I response by interacting with RIG-I and TRIM25 [28][29][30] . Graef et al. demonstrated that the PB2 subunit of the influenza virus RNA polymerase complex interacted with MAVS and inhibited MAVS-mediated IFN-β expression 31 . Additionally, the other polymerase proteins PB1 and PA were identified as part of the IFN-I inhibitory strategy evolved by influenza virus and suggested to interact with RIG-I, subsequently inhibiting RIG-I mediated IFN-I signaling 32 . Effects outlined of the PB1-F2 are strain specific, however interferon suppression seems to be a common property, as A/ Hong Kong/156/1997 (H5N1) as well as PR8 (H1N1) strains both exhibit IFN-β antagonism 27 . More recently, Hayashi et al. reported that the presence of PA-X in A/California/04/2009 (H1N1) also resulted in a suppression of IFN-β mRNA production through the shutoff activity of PA-X, leading to more efficient viral replication and more severe lung pathology in infected cells and mice 9 . Here, we clearly showed that truncated PA-X was more effective in inhibiting IFN-I mRNA expression in PAMs and pigs through the shutoff activity of PA-X, which is consisted with the enhanced replication, pathogenicity and transmissibility of PA-X mutants in pigs. Therefore, our study provides evidence that the adaptation of SIV in pigs may be partly due to the shutoff activity of PA-X.
In summary, we found that truncation of PA-X played an important role in the adaptation of SIV in pigs. Compared with that of full length PA-X, it conferred increased viral pathogenicity, replication and transmissibility in pigs, promoted viral growth in PK15 cells and swine respiratory tissues infected ex vivo, and greater inhibition of host IFN-I expression, which was consistent with the increasing occurrence of SIV with truncated PA-X.  Generation of recombinant viruses by reverse genetics. All eight gene segments were amplified by reverse transcription (RT)-PCR from Sw-41X(WT) virus, and cloned into the dual-promoter plasmid pHW2000. The mutations were introduced into the PA gene using a Site-directed QuikChange Mutagenesis kit (Agilent) according to the manufacturer's instructions. PCR primer sequences are available upon request. The TR H1N2 SIV with full-length PA-X (61aa), Sw-61X, was constructed by extending PA-X from 41aa to 61aa on the backbone of Sw-41X(WT) virus. Substitution from stop codon (UAG) to tryptophan (UGG) codon at position 42 in the X-ORF did not change the PA ORF (Fig. 2). Rescued viruses were detected using haemagglutination assays. Viral RNA was extracted and analysed by RT-PCR, and each viral segment was sequenced to confirm sequence identity. Viral titration and growth kinetics. The 50% tissue culture infectious dose (TCID 50 ) was determined in MDCK cells with 10-fold serially diluted viruses inoculated at 37 °C for 72 h, as previous studies with Reed-Muench method 33,34 . Multistep replication kinetics was determined by inoculating PK15 cells with a MOI of 0.01. Supernatants were sampled at 6,12,24,36,48,60, and 72 hours post-inoculation (hpi) and titrated by inoculating MDCK cells in 96-well plates. Three independent experiments were performed.
Isolation, culture, and infection of porcine respiratory explants. Porcine nasal turbinate, trachea, and lung explants were prepared as described previously 11,35,36 . All respiratory explants were cultured at an air-liquid interface in 12-well plates at 37 °C in 5% CO 2 atmosphere. At 24 h of culture, explants were washed with PBS, and 10 6 TCID 50 of each virus diluted in 500 μ l explant medium were deposited in the upper compartment of Transwells for 1 h at 37 °C. Subsequently, explants were washed three times with PBS and the culture was replenished with 500 μ l of explant media. At 24, 48, and 72 hpi, 100 μ l supernatant was collected to assess virus yields.

Infection and transmission of SIV in pigs.
Four-week-old, landrace piglets that were free of influenza virus, porcine reproductive and respiratory syndrome virus (PRRSV), classic swine fever virus (CSFV), pseudorabies virus (PRV), porcine circovirus type 2 (PCV2) and M. hyopneumoniae infections were obtained from Beijing Center of SPF Swine Breeding & Management. Nine pigs from each group were inoculated intranasally with 10 6 TCID 50 of tested virus, delivered in a final volume of 2 ml per nostril, using a mucosal atomization device (MAD, Wolfe Tory Medical, Inc.) to mimic aerogenic infection. Five uninfected pigs were maintained as a control group and were mock-inoculated with 2 ml PBS per nostril, using the same method as the infected pigs. Three infected pigs from each group were removed to a separate room with three naïve pigs at 2 days post-inoculation (dpi). Clinical signs were observed by veterinary inspections at a fixed time daily throughout the duration of the study (14 days). Nasal washes were collected from all pigs at 3, 5, 7, and 9 dpi. At 3 and 5 dpi, three pigs from each infection group and one pig from control group were anesthetized and their nasal turbinates, tracheas, and lungs were collected for virus titration.
Microscopical examination of lungs. The lung tissues collected at 5dpi were fixed in 10% phosphate-buffered formalin for histopathological examination which was performed as described previously [37][38][39] . Briefly, samples were processed for paraffin embedding and cut into 5-μ m-thick sections. One section from each tissue sample was stained with hematoxylin and eosin (H&E) stain. Another section was processed for immunohistological staining with a rabbit polyclonal antibody (orb176401; Biorbyt, USA) against influenza A virus NP. Specific antigen-antibody reactions were visualized by using DAB Detection Kit (Polymer) (ORIGENE, Wuxi, China) and counterstained with hematoxylin.
The severity of microscopic lung lesions was scored by the distribution or by the extent of lesions within the sections examined as follows: 0, no visible changes; 1, mild focal or multifocal change; 2, moderate multifocal change; 3, moderate diffuse change; and 4, severe diffuse change 40 . The detection of swine influenza antigen in lung was executed through a ranked sore of 0-4, which was used to evaluate the number of positive cells per section taken from each block, as previously described 41 . The indication of the scores was as follows: 0, no SIV-antigen-positive cells; 1, 1-10 positive cells; 2, 11-30 positive cells; 3, 31-100 positive cells; and 4, = or > 100 positive cells. Two independent pathologists scored all lungs and slides from blinded experimental groups.

Quantitative real-time PCR (qRT-PCR).
Total RNA in PAM cells or lung homogenates were extracted using TRIzol reagent at specified time points according to the manufacturer's instructions (Invitrogen). Oligo-(dT) primed cDNA was generated from 1 μ g of total RNA per sample using Superscript III First-Strand Synthesis SuperMix (Invitrogen). The qRT-PCR mixture for each sample consisted of 10 μ l 2 × SYBR green PCR master mix (Applied Biosystems), 7 μ l of nuclease-free water, 0.5 μ l of each primer, and 2 μ l of cDNA template. Amplification was performed using the 7500 real-time PCR system (Applied Biosystems) with the following program: 1 cycle at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Expression values for each gene, relative to RPL4, were calculated using the 2 −△△CT method. Each experiment contained three technical replicates for each sample, with two experimental replicates performed. Primers for qRT-PCR analysis of pig IFN-α /β are listed as following: IFN-α , Forward, 5′ -GCCTCCTGCACCAGTTCTACA-3′ , Reverse, 5′ -TGCATGACACAGGCTTCCA-3′ ; IFN-β , Forward, 5′ -TGCAACCACCACAATTCC-3′ , Reverse, 5′ -CTGAGAATGCCGAAGATCTG-3′ .
Interferon-β reporter assay. To measure the IFN-I shutoff function of PA-X, a luciferase based, Sendai virus-mediated IFN-β promoter activation assay was conducted as previously described 13 . Briefly, 293T cells in 24-well plates were transfected with 500 ng of empty vector or different PA-X proteins respectively. Additionally, 200 ng of an IFN-β -promoter-luciferase reporter plasmid (pIFNbLuc) and 20 ng of a plasmid constitutively expressing Renilla luciferase (pRL-TK from Promega) were transfected. At 18 hours post transfection, cells were infected with Sendai virus to induce the IFN-β promoter at a MOI of 5.0. At 18 hour post virus inoculation, cell Scientific RepoRts | 6:21845 | DOI: 10.1038/srep21845 lysate was prepared with Dual-Luciferase Reporter Assay System (Promega, USA) and luciferase activity was measured using GloMax 96 microplate luminometer (Promega, USA). The relative luciferase activity of the group with the empty vector was set as 100%, and the other groups were presented relative to that. Statistical analyses. All statistical analyses were performed using GraphPad Prism Software Version 5.00 (GraphPad Software Inc., San Diego, CA, USA). Statistically significant differences between experimental groups were determined using the analysis of variance (ANOVA) method. Differences were considered statistically significant at P < 0.05.