RACK1 is indispensable for porcine reproductive and respiratory syndrome virus replication and NF-κB activation in Marc-145 cells

Porcine reproductive and respiratory syndrome virus (PRRSV) causes porcine reproductive and respiratory syndrome (PRRS), which is currently insufficiently controlled. RACK1 (receptor of activated protein C kinase 1) was first identified as a receptor for protein kinase C, with increasing evidence showing that the functionally conserved RACK1 plays important roles in cancer development, NF-κB activation and various virus infections. However, the roles of RACK1 during PRRSV infection in Marc-145 cells have not been described yet. Here we demonstrated that infection of Marc-145 cells with the highly pathogenic PRRSV strain YN-1 from our lab led to activation of NF-κB and upregulation of RACK1 expression. The siRNA knockdown of RACK1 inhibited PRRSV replication in Marc-145 cells, abrogated NF-κB activation induced by PRRSV infection and reduced the viral titer. Furthermore, knockdown of RACK1 could inhibit an ongoing PRRSV infection. We found that RACK1 is highly conserved across different species based on the phylogenetic analysis of mRNA and deduced amino acid sequences. Taken together, RACK1 plays an indispensable role for PRRSV replication in Marc-145 cells and NF-κB activation. The results would advance our further understanding of the molecular mechanisms underlying PRRSV infection in swine and indicate RACK1 as a promising potential therapeutic target.

nonstructural protein of white spot syndrome virus (WSSV) 18 . Walleye dermal sarcoma virus Orf B functions through RACK1 and protein kinase C 19 . Mumps virus V protein has the ability to interact strongly with RACK1 20 . In addition, RACK1 plays an anti-apoptotic role during infectious bursal disease virus (IBDV) infection via interaction with VDAC2 and VP5. It suggests that VP5 sequesters RACK1 and VDAC2 in the apoptosis-inducing process 21 . Furthermore, together with integrin beta, RACK1 is involved in the cell entry of Bombyx mori cypovirus 22 . Additionally, RACK1 was discovered as one of the 16 interacting cellular proteins by the yeast two-hybrid system during a screen for classic swine fever virus (CSFV) NS5A interactive proteins in the cDNA library of the swine umbilical vein endothelial cell (SUVEC) 23 . RACK1 was also screened with CSFV E2 as bait protein by yeast two-hybrid from porcine alveolar macrophages (PAM cells) expression library, which was further confirmed by co-transformation, GST pull-down and laser confocal assays 24 .
However, there are no studies that define the role of RACK1 during PRRSV infection. Here we analyzed the sequence similarity of RACK1 (at both mRNA and amino acid levels) across diverse species, and then investigated how PRRSV infection in Marc-145 cells affects the expression level of RACK1. We further show that siRNA knockdown of RACK1 in Marc-145 cells downregulates PRRSV replication, abrogates NF-κB activation induced by PRRSV infection and reduces the viral titer.

Results
Phylogenetic analysis and multiple sequence alignment. RACK1 mRNA sequence from Marc-145 cells (Macaca mulatta, a member of Rhesus monkey) is 954 nucleotides long (Genbank accession number KT751174.1) and encodes a protein sequence with a length of 317 amino acids. RACK1 orthologous mRNA sequences and amino acid sequences from 26 species, including the one sequenced at our lab from Marc-145 cells, were downloaded from The National Center for Biotechnology Information (NCBI) database. The phylogenetic analysis and multiple sequence alignment based on these data were performed using online tool Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/) with the default parameters. The result showed that the RACK1 mRNA sequence of Marc-145 cells has a strong evolutionary connection with human and chimpanzee (indicated by the red square in Fig. 1A). Furthermore, the RACK1 amino acid sequence is also highly conserved between primates and additional species such as pig (highlighted in the red square in Fig. 1B,C).

PRRSV infection upregulated cellular RACK1 expression. RACK1 expression is stimulus-and
cell-type-specific. For instance, RACK1 is downregulated in EBV-infected monocytes 25 , while it is highly up-regulated through nerve growth factor (NGF) induced NF-κB activation 26 or by TGF-β1 in mice 27 . Hence, in this study, we investigated the effect of PRRSV infection on the RACK1 expression. Marc-145 cells in 6-well plates were analysed at different time points post infection with 25 TCID 50 of YN-1 strain or without virus challenge. RNA and proteins were extracted from the cells for virus copy determination, RACK1 mRNA and protein expression analysis. Copy number of PRRSV ORF7 mRNA significantly increased after 36 hours post infection ( Fig. 2A) and viral N protein encoded by ORF7 was detected by western blot 48 hours post infection (Fig. 2C), indicating efficient virus replication in Marc-145 cells. During the whole experimental time period (from 1 hour to 60 hours), substantial increase of cellular RACK1 mRNA (Fig. 2B, uninduced levels highlighted by the red line) and protein (Fig. 2C) was observed, compared with the non-infection control. Even at the very beginning of PRRSV infection (such as 1-12 hours post virus inoculation), striking upregulation of RACK1 expression was recorded (Fig. 2C). However, we also observed an early increase of RACK1 mRNA production following PRRSV infection (Fig. 2B, 1 hpi), while RACK1 protein levels remain constant (Fig. 2C) throughout the period of infection. We reason that the mRNA expression data can not always correlate with the protein expression data in an exact manner. siRNA knockdown of RACK1 inhibited PRRSV replication and abrogated NF-κB activation induced by PRRSV infection. The transcription factor NF-κB plays a pivotal role in innate immunity in response to a variety of stimuli. When stimulated, the IκB proteins are phosphorylated by IκB kinases (IKKs) and degraded by proteasomes. Thus the IκB proteins allow the release and translocation of NF-κB into the nucleus to activate the transcription of genes involved in innate and adaptive immunity. The coordinated regulation of this pathway determines the proper host responses to extracellular signals [28][29][30] . Therefore, the authors investigated whether PRRSV infection in Marc-145 cells can activate NF-κB and whether RACK1 plays any role in the NF-κB activation.
First, efficient knockdown was achieved using two siRNAs against RACK1, as demonstrated by RT-qPCR over 72 hours (Fig. 3A) and by western blot (Fig. 3B), when normalized to the internal GAPDH control and compared with the non-targeting siRNA transfection. Then two days post siRNA transfection, the Marc-145 cells cultured  Figure 1) and led to the abrogation of phosphorylation of IκBa and p65 (Fig. 3D,F), while no influence on the protein level of total p65 (Fig. 3E) was observed, suggesting that RACK1 downregulation blocked the NF-κB activation induced by PRRSV infection. We also ruled out the possibility that the anti-virus effects might come from the enhanced interferon alpha level resulted from the siRNA treatment (Supplementary Figure 2).
In addition, immunofluorescent staining analysis was applied to further investigate the effect of RACK1 knockdown on PRRSV replication in Marc-145 cells. The cells were transfected with siRNAs, challenged with PRRSV 48 hours post transfection, fixed 24 hours or 48 hours post infection and immunofluorescently stained to visualize the nuclei (blue channel) and the expression level of RACK1 (red channel) and viral N protein (green channel). Reduction of RACK1 protein level occurred after siRNA knockdown with the two siRNA sequences     Table 1) used in this study targeting different RACK1 mRNA sequences. After SDS-PAGE and protein transferring, the membranes were cut based on the molecular weights of the target proteins and the markers, put together along the same film for exposure with same exposure time for each sub-figure. The histograms and blots shown here are representative data from three independent experiments. were analyzed at different time points after transfection (48, 72 and 84 hours post transfection). The data in Fig. 6 clearly showed that compared with the non-targeting siRNA control (siScramble), RACK1 siRNA knockdown significantly reduced the mRNA and protein expression level of cellular RACK1 and viral ORF7, suggesting that knockdown of RACK1 could inhibit an ongoing PRRSV infection.   10.5 µl of water. The amplification program was 94 °C for 5 min and 35 cycles of 94 °C for 30 s, 60 °C for 30 s and 72 °C for 50 s, followed by additional 5 mins extension at 72 °C. The primer pairs used in this study are designed based on the reference sequence XM_011745042 and listed in Table 1. PCR amplicon was electrophoresed using 1.2% agarose gel, purified using the SanPrep column DNA gel purification kit from Sangon (Shanghai, China, Cat. #B518131) and subjected to ligation at 4 °C for overnight. The ligation was carried out in 10 µl reaction, containing 5 µl of Solution I, 50 ng of pMD-18T Vector (pMD-18T Vector cloning kit, Takara, Cat. #3270) and 200-1000 ng cDNA template. The ligation reaction was transformed into competent DH5α cells (Takara, Cat. #9057) which were cultured according to the instruction. Subsequently, the plasmid was extracted using SanPrep Column Plasmid DNA Extraction Kit (Sangon, Cat. #B518191) and confirmed by BamH I (Takara, Cat. #1010A) digestion. Then the RACK1 mRNA was sequenced by Sangon (Shanghai, China), analyzed using Megalign and SeqMan modules under the DNAStar package (version 7.1) and submitted to GenBank (access number: KT751174.1, https://www.ncbi.nlm.nih.gov/nuccore/975883998).

siRNA transfection and virus infection. Cell seeding, siRNAs transfection and virus infection were
performed according to our previous study 32 . In brief, Marc-145 cells were seeded into 96-well plates (10 4 cells per well) or 6-well plates (3 × 10 5 cells per well) during reverse transfection with 40 nM siRNA (against RACK1 or with random sequence) and Lipofectamine 3000 (Thermofisher Scientific, Cat. # L3000008) (1.5 µl/well for 96-well plate or 6 µl/well for 6-well plate). Forty-eight hours post transfection, the cells were inoculated with PRRSV YN-1 strain (25 TCID 50 /well, equivalent to an approximate MOI of 0.02,) till further analysis. Each treatment was performed in triplicate. The siRNAs used in this study are listed in Table 1 and purchased from Sangon (Shanghai, China). Transfection with siScramble serves as neutral control for normalization.
Indirect immunofluorescence staining. According to our previous study 32 , 24 or 48 hours post PRRSV infection, with or without siRNA transfection, the Marc-145 cells were washed with PBS (0.05 M, pH7.4) and fixed with 4% paraformaldehyde (PFA) at room temperature for 10-15 minutes. Then the cells were washed with PBS for three times, permeabilizated with PBS containing 0.3% Triton X-100 for 15 minutes and blocked with PBS containing 1% BSA for 2 hours at 4 °C. The nuclei staining with 5 µg/ml of Hoechst 33342 (Life Technology) was carried out for 20 minutes at room temperature. The cells were subsequently co-incubated with PRRSV antibody against N protein (encoded by ORF7) (VMRD, Cat. #080728-004, mouse origin, 5 µg/ml) and anti-RACK1 antibody (Cell Signaling Technology, Cat. #5432 s, rabbit origin, 2.7 µg/ml) at 4 °C overnight, washed three times with PBS and then co-incubated with Alexa Fluor 488 conjugated goat anti-mouse IgG (H + L) antibody (Proteintech, Cat. #861163) and Alexa Fluor 546 conjugated goat anti-rabbit IgG (H + L) antibody (Thermofisher Scientific, Cat. #11010) at 5 µg/ml for each for 1 h at 37 °C. Staining with Hoechst 33342 (Sigma-Aldrich, Cat. #B2261) at room temperature for 15 minutes was used to visualize the nuclei. After three times PBS wash, the cells were subjected to image analysis by fluorescence microscopy (Olympus).
SCIentIfIC REPORts | (2018) 8:2985 | DOI:10.1038/s41598-018-21460-4 Virus titration. As described in our previous study 32 , Marc-145 cells were seeded into 96-well plates (10 4 cells/well in 100 µl) during siRNA transfection. Forty-eight hours post transfection, a 10× serial dilution of PRRSV YN-1 strain was prepared and each dilution was added into six wells (100 µl/well). CPE was recorded using the inverted microscope over a period of 3 days post virus challenge. Cell number was counted and the 50% tissue culture infected dose (TCID 50 ) was determined by Reed-Muench method.

Discussion
The PRRSV shows a restricted tropism for subsets of porcine macrophages in vivo. To date, at least seven cellular molecules have been described as putative receptors for PRRSV, including heparan sulfate, vimentin, CD151, sialoadhesin (CD169), dendritic cell-specific intercellular adhesion melecule-3-grabbing non-integrin (DC-SIGN; CD209), vimentin and CD163 34,35 . Incubation of macrophages at 37 °C with both sialoadhesin-and CD163-specific antibodies completely blocked the PRRSV infection 36 . However, most of the PRRSV receptors were identified on porcine alveolar macrophages (PAMs) with specific functions for PRRSV pathogenesis, for instance, heparan sulphate for binding 37,38 and sialoadhesin for binding and internalization 39 . Only CD163 was described to be essential in PRRSV infection of non-permissive cells (for example Marc-145 cells, which were used in this study) 36,40 . Therefore, we speculate that additional host factors are needed for PRRSV to complete the viral life cycle in PAMs or Marc-145 cells.
RACK1 is highly conserved across species 41,42 . Therefore, in this study, degenerate polymerase chain reaction (PCR) primers, based on highly conserved regions of RACK1 from the 26 species used to construct the phylogenetic tree, were used in a reverse transcriptase polymerase chain reaction (RT-PCR) reaction to amplify RACK1 from Marc-145 cells. The deduced protein sequence of RACK1 cDNA with a full length of 954 bp from Marc-145 cells shows that it contains 317 amino acid residues, and shares 100% identity with human and porcine (Fig. 1C). Therefore an antibody against human RACK1 could be applied in this study.
A number of pathogenic viruses, such as Epstein-Barr virus 43 and hepatitis C virus 44 , exploit the NF-κB system for their own profit. PRRSV is recognized as one of the most important viruses for the swine industry, mainly due to its persistence in pigs for quite a long time after initial infection. NF-κB activation has been believed to be one of the key pathogenic mechanisms. Several studies showed that PRRSV infection could trigger NF-κB signal pathway in MARC-145 cells or porcine alveolar macrophages 31,[45][46][47] . The NF-κB activating viral proteins include nucleocapsid (N) protein 47 and NSP2 protein 48 . However, there are also some controversial data on the interactions between PRRSV and the NF-κB pathway. Some studies reported that the NF-κB deactivating viral proteins NSP2 49 and NSP1a 45,46,50 could inhibit the NF-κB signaling pathway by interfering with the poly-ubiquitination process of IκBα. PRRSV may have developed sophisticated strategies to either activate or inhibit NF-κB for its own benefit at different stages of its life cycle. Here our data showed that infection of Marc-145 cells with YN-1 strain which belongs to the PRRSV genotype 2 resulted in phosphorylation of IκBα and p65, 1 hpi and 4 hpi, respectively, signifying NF-κB activation (Fig. 3D,F). Our results are consistent with some previous studies reporting that phosphorylation of IκBα was observed already 30 minutes post infection or 24 hours post transfection in PAMs challenged with PRRSV or in CRL2843 cells transfected with N protein expression vector 31,45-47 . The slight difference may stem from the specific cell type, PRRSV strain and multiplicity of infection. We speculate that sampling from more time points very early in in vitro infection (i.e., from 30 mins to 4 hrs) would aid in profiling NF-κB activation.
RACK1 was identified as a novel negative regulator of NF-κB signaling physically associated with the IKK complex in a TNF-triggered manner in 293 T cells 13 . This interaction interferes with the recruitment of the IKK complex to TRAF2, which is a critical step for IKK phosphorylation and subsequent activation triggered by TNF. However, in this study we demonstrated RACK1 as a positive regulator of NF-κB signaling induced by PRRSV infection in Marc-145 cells (Fig. 3). The complete opposite functionality of RACK1 in NF-κB signaling probably stem from the differing stimuli and cell types. The elaborate mechanisms by which PRRSV regulates NF-κB activation and how RACK1 plays its roles in NF-κB and PRRSV replication require further studies.
There was a report showing that viral replication of PRRSV strain CH-1a was not obviously affected when bone marrow-derived macrophages (BMDMs) were treated with of NF-κB inhibitor BAY11-7082 51 . Hence we reasoned that inhibition of NF-κB signaling may not interfere with the replication of PRRSV strain YN-1 in Marc-145 cells, and that the inhibition of YN-1 replication (Fig. 4) and reduction of viral titer (Fig. 5) by RACK1 siRNA knockdown was not through deactivation of NF-κB. This also needs further investigation.
Fighting viral infections is hampered by the scarcity of viral targets and their variability. Viruses depend on cellular molecules, which are attractive alternative targets for drug development, provided that they are dispensable for normal cell function. Inhibition of RACK1 does not affect Drosophila or human cell viability and proliferation, and RACK1-silenced adult flies are viable, indicating that this protein is not essential for general translation 14 . A similar phenotype was observed in this study, as shown in Figs 3, 4 and 5 in terms of cell number.
Collectively, in spite of the controversy about whether PRRSV infection activates or deactivates the NF-κB pathway, our results demonstrate that the highly pathogenic PRRSV strain YN-1 isolated by our lab did activate the NF-κB signal pathway and upregulate RACK1 expression over the investigation time period in Marc-145 cells. The authors suggested that RACK1 is very conservative across different species based on the phylogenetic analysis of mRNA and deduced amino acid sequences. For the first time we demonstrated that siRNA knockdown of RACK1 inhibited PRRSV replication in Marc-145 cells, abrogated the NF-κB activation induced by PRRSV infection and eventually reduced the viral titer. Furthermore, we demonstrated that knockdown of RACK1 could inhibit an ongoing PRRSV infection. The conclusion can be drawn that RACK1 is an indispensable cellular factor for PRRSV replication in Marc-145 cells and thus NF-κB activation, while dispensable for cell viability. This study may provide some insights into the molecular mechanisms of PRRSV infection in swine and RACK1 was uncovered as a promising potential therapeutic target for PRRSV intervention. The therapeutic potential of RACK1 could be realised by small molecules which can selectively target RACK1or/and its interacting proteins 52  molecule screen is currently carried out in our lab to identify chemical compounds for inhibition of PRRSV replication by downregulating the RACK1 expression level.