Proteolytic processing, deubiquitinase and interferon antagonist activities of Middle East respiratory syndrome coronavirus papain-like protease

The emerging Middle East respiratory syndrome coronavirus (MERS-CoV) causes severe pulmonary disease in humans and represents the second example of a highly pathogenic coronavirus (CoV) following severe acute respiratory syndrome coronavirus (SARS-CoV). Genomic studies revealed that two viral proteases, papain-like protease (PLpro) and 3C-like protease (3CLpro), process the polyproteins encoded by the MERS-CoV genomic RNA. We previously reported that SARS-CoV PLpro acts as both deubiquitinase (DUB) and IFN antagonist, but the function of the MERS-CoV PLpro was poorly understood. In this study, we characterized MERS-CoV PLpro, which is a protease and can recognize and process the cleavage sites (CS) of nsp1-2, nsp2-3 and nsp3-4. The LXGG consensus cleavage sites in the N terminus of pp1a/1ab, which is generally essential for CoV PLpro-mediated processing, were also characterized in MERS-CoV. MERS-CoV PLpro, like human SARS-CoV PLpro and NL63-CoV PLP2, is a viral deubiquitinating enzyme. It acts on both K48- and K63-linked ubiquitination and ISG15-linked ISGylation. We confirmed that MERS-CoV PLpro acts as an IFN antagonist through blocking the phosphorylation and nuclear translocation of IFN regulatory factor 3 (IRF3). These findings indicate that MERS-CoV PLpro acts as a viral DUB and suppresses production of IFN-b by an interfering IRF3-mediated signalling pathway, in addition to recognizing and processing the CS at the N terminus of replicase polyprotein to release the non-structural proteins. The characterization of proteolytic processing, DUB and IFN antagonist activities of MERS-CoV PLpro would reveal the interactions between MERS-CoV and its host, and be applicable to develop strategies targeting PLpro for the effective control of MERS-CoV infection.

INTRODUCTION

In September 2012, a new human coronavirus (CoV) emerged in Jeddah and Saudi Arabia and quickly spread to some European countries (Anderson & Baric, 2012; Chan et al., 2012; Zaki et al., 2012). This new CoV, named Middle East respiratory syndrome coronavirus (MERS- CoV), was formerly called human coronavirus-Erasmus Medical Center (HCoV-EMC) (de Groot et al., 2013). It can cause severe acute respiratory syndrome and renal failure. From September 2012 to 1 August 2013, MERS- CoV infected at least 94 people with a fatality rate of about 50 % globally (novel coronavirus infection update; http:// www.who.int/csr/disease/coronavirus_infections/en/index. html). The epidemic pattern and clinical symptoms caused by MERS-CoV are very similar to the severe acute respira- tory syndrome coronavirus (SARS-CoV) (Booth & Stewart, 2005; Chan-Yeung & Xu, 2003). Person-to-person trans- mission of MERS-CoV was recently confirmed; the virus can spread globally and cause considerable morbidity en- dangering world health and the economy (Assiri et al., 2013).
Several isolates of MERS-CoV have been fully sequenced and confirmed to belong to group C b-coronaviruses (Chan et al., 2012; Khan, 2013; Lau et al., 2013; van Boheemen et al., 2012). The analysis of genome size,organization and sequence revealed that MERS-CoV is most closely related to bat CoVs, BtCoV-HKU4 and BtCoV- HKU5 (Chan et al., 2012; Lau et al., 2013; van Boheemen et al., 2012). The MERS-CoV genome is just over 30 kb and contains at least 10 predicted ORFs including polypro- tein pp1a/1ab, S protein, E protein, N protein and some canonical structural proteins (van Boheemen et al., 2012). The ORF1a sequence of MERS-CoV encodes two protease domains that are conserved in all other CoVs, a papain-like protease (PLpro) in non-structural protein (nsp) 3 and a 3C- like protease (3CLpro; also known as the ‘main protease’) in nsp5 (van Boheemen et al., 2012). Sequence comparison of MERS-CoV with other CoVs allowed us to predict the putative pp1a/ pp1ab cleavage sites (CS) and annotate the processing products of nsp1-16 (van Boheemen et al., 2012).

Analysis of proteolytic processing of several CoVs has re- vealed that the replicase polyproteins pp1a and pp1ab are cotranslationally and post-translationally processed by the viral proteases PLpro and 3CLpro to generate as many as 16 nsps. These nsps contribute to subgenomic RNA syn- thesis by forming the replicase complex with the endo- plasmic reticulum membranes (Chen et al., 2007; Gosert et al., 2002; Prentice et al., 2004; Snijder et al., 2006). Several studies revealed that papain-like protease processes the N terminus of pp1a and pp1ab by processing the CS 1–3 to release nsp1-3, whereas 3CLpro processes the sub- sequent CS to release nsp4-16 (Barretto et al., 2005; Chen et al., 2007; Oostra et al., 2008). Proteolytic processing of the CoV proteases is essential for correct localization and function of the replicase proteins, ongoing viral RNA synthesis and virus replication (Chen et al., 2007; Kim et al., 1995). In addition to the proteolytic processing, CoV PLpro has deubiquitination activity and serves as an IFN antagonist (Barretto et al., 2005; Clementz et al., 2010; Frieman et al., 2009; Sulea et al., 2005), indicating that PLpro is a multifunctional protease that is involved in CoV replication and modulates the interaction between the virus and its host (Sun et al., 2010, 2012). A very recent study reported the expression and activity of MERS-CoV PLpro and 3CLpro in cell-based biosensor assays, which allows for rapid evaluation of viral protease activity and the identification of protease inhibitors (Kilianski et al., 2013). Thus, CoV proteases are attractive targets for the develop- ment of antiviral drugs for reduction of viral replication and pathogenicity. However, the characteristics and func- tion of MERS-CoV PLpro are little known.

In this study, we first identified the MERS-CoV PLpro activity and the putative catalytic triad that is required for cleavage of pp1a/1ab. We found that the membrane-anchored form of PLpro (PLpro-TM) alone is sufficient to process the predicted nsp1/2, nsp2/3 and nsp3/4 sites. The putative cleavage sites between replicase products were also characterized by site- directed mutagenesis of predicted critical residues. Further- more, we confirmed that, like SARS-CoV PLpro and HCoV- NL63 (Barretto et al., 2005; Chen et al., 2007), MERS-CoV PLpro has deubiquitinase (DUB) activity that recognizes and processes both ubiquitin (Ub) and ISG15-linked conjugations. Finally, MERS-CoV PLpro acts as a potential IFN antagonist through disrupting the phosphorylation and nuclear trans- location of IFN regulatory factor 3 (IRF3). The identification of PLpro activity and IFN antagonist activity, and character- ization of MERS-CoV PLpro DUB would facilitate the comparative studies of CoV protease activity and the development of novel antiviral reagents specifically against HCoV, including MERS-CoV, HCoV-NL63 and SARS-CoV.

RESULTS

Identification of protease activity and the catalytic sites of MERS-CoV PLpro

MERS-CoV has been predicted to encode replicase poly- proteins that possess about 35 % sequence homology to those of SARS-CoV (van Boheemen et al., 2012). The MERS-CoV polyproteins were predicted to produce 16 nsps by two distinct proteases, PLpro and 3CLpro (Fig. 1a). Previous studies revealed that expression of an extended region of PLpro/PLP2 that included the downstream trans- membrane (TM) domains was required for processing at the predicted N terminal CS in the polyprotein (Harcourt et al., 2004; Kanjanahaluethai et al., 2007). Firstly, in order to identify the protease activity of PLpro, we generated PLpro-TM (Fig. 1a), and three substrate constructs con- taining the predicted cleavage sites at the N terminus of pp1a/1ab (nsp1-2, nsp2-3 and nsp3-4) (Fig. 1b). A cell-based trans-cleavage system was established by coex- pressing PLpro-TM with the truncated substrates of MERS-CoV in HEK293T cells. The protease activity was probed by detection of released processing products using anti-Flag antibody or anti-V5 antibody (Fig. 1b, c). The construct PLpro-TM was cotransfected with plasmid DNAs that expressed the substrates of nsp1-2, nsp2-3 or nsp3-4 in HEK293T cells. The cells were harvested and subjected to Western blotting using the indicated antibodies. As shown in Fig. 1c, the WT PLpro completely cleaved the substrate nsp2-3 to produce the cleavage products (Fig. 1c, lane 3). PLpro-TM partially processed the substrates nsp1-2 and nsp3-4, although the processed products were detectable (Fig. 1c, lanes 2 and 4). We note that PLpro-TM is detected as two closely migrating bands, perhaps due to glycosylation of the transmembrane domains, which was confirmed in SARS-CoV and murine hepatitis virus (MHV) (Harcourt et al., 2004; Kanjanahaluethai et al., 2007). These findings indicated that MERS-CoV PLpro has protease activities to process pp1a/1ab at the N terminus to release product nsp1-3, which is consistent with other human CoVs, such as SARS-CoV and NL63-CoV (Barretto et al., 2005; Chen et al., 2007; Harcourt et al., 2004).

To identify which critical amino acid residues within the MERS-CoV PLpro core domain were important for pro- tease activity, we aligned the amino acids of the PLpro core domain (aa1476 to aa1804) of MERS-CoV with PLpro/ PLP2 of representative members from all coronaviral genera (Fig. 2a). We found that the predicted triad catalytic cysteine, histidine and aspartic acid residues (C1592–H1759–D1774), and the cysteine/histidine residues essential for binding zinc were highly conserved; these are marked with red and blue boxes, respectively in Fig. 2(a). The three-dimensional (3D) structure of MERS-CoV PLpro core domain (aa1476 to aa1804) was modelled using bioin- formatic analysis, and the results showed that the topology of MERS-CoV PLpro is remarkably similar to that of SARS- CoV PLpro (Fig. 2b). It indicated that MERS-CoV PLpro consists of four distinctive domains: an N-terminal Ubl domain, a palm, a thumb and a finger domain, which form an extended right-hand architecture that closely resembles the structure of SARS-CoV PLpro (Ratia et al., 2006). As PLpro exhibited DUB activity on ubiquitinated proteins without any specificity for either K48 or K63 linkage.

It has been reported that both SARS-CoV PLpro and NL63 PLP2 have deISGylation activities (Clementz et al., 2010; Lindner et al., 2007; Sun et al., 2010). To examine whether MERS-CoV PLpro could influence the ISGylation of cellular proteins, a combination of Ube1L, UbcH8 and Myc-ISG15 together with either WT PLpro-TM or the catalytic mutants was transfected into HEK293T cells. The cell lysates were subjected to Western blotting using anti-Myc antibody to detect the ISGylation of the cellular proteins. The ISGylation of substrate proteins was markedly decreased by MERS-CoV PLpro-TM (Fig. 5d, lane 3), indicating that MERS-CoV PLpro acted as a deISGylation enzyme. Interestingly, introducing mutations into the catalytic site D1774 did not have any effect on deISGy- lation enzyme activity (Fig. 5d, lane 6), indicating that D1774 is not required for deISGylation, which is similar to DUB activity against Ub (Fig. 5a, lane 6). However, in contrast to PLpro DUB, the catalytic mutants C1592A and H1759A of PLpro lost almost all deISGylation activity (Fig. 5d, lanes 4 and 5), indicating that the catalytic sites C1592 and H1759 are required for deISGylation. These results definitely indicated that MERS-CoV PLpro had potent DUB activity of both Ub and ISG15-conjugated proteins in host cells, which may be associated with its protease activity.

MERS-CoV PLpro is an IFN antagonist

We previously reported that PLpros of human CoVs, SARS- CoV and NL63-CoV, acted as IFN antagonists (Clementz et al., 2010; Devaraj et al., 2007; Sun et al., 2010, 2012). To determine if MERS-CoV PLpro was an IFN antagonist, which may contribute to inhibiting the expression of type I IFN, HEK293T cells were transfected with the plasmids expressing WT PLpro-TM or its catalytic mutants sepa- rately, along with an IFN-b promoter-driven luciferase reporter and Flag-RIG-IN (N-terminal helicase domain of RIG-I as its constitutively active mutant) that was used to activate the IFN-b promoter activity. As shown in Fig. 6a, the IFN-b promoter was activated 100-fold when the cells were transfected with RIG-IN, whereas the IFN-b promoter activated by RIG-IN was completely inhibited in the presence of MERS-CoV PLpro-TM, which was similar to the previously reported IFN antagonist activity of SARS- CoV PLpro and NL63-CoV PLP2 (Clementz et al., 2010; Devaraj et al., 2007; Sun et al., 2010). To determine if MERS-CoV PLpro catalytic activity is essential for inhibi- tion of IFN expression, the catalytic mutants of MERS-CoV PLpro-TM (C1592A, H1759A and D1774A) were trans- fected together with the IFN-b-Luc and RIG-IN plasmids into HEK293T cells; we found that all catalytic mutants of PLpro-TM had inhibition effects on RIG-IN-activated IFN-b promoter-driven luciferase activity (Fig. 6a). These data demonstrated that MERS-CoV PLpro-TM was an IFN antagonist, which was independent of its protease activity.

Cellular transcription factors, NF-kB, IRF3 and AP-1 play critical roles in the transcriptional control of the IFN-b promoter (Barber, 2011; Dev et al., 2011; Marsili et al., 2012). To determine whether MERS-CoV PLpro inhibition of the activation of the IFN-b promoter was related to its suppressive effect on NF-kB, IRF3 or AP-1 activation, we performed a luciferase reporter assay to analyse the trans- criptional activity of NF-kB, IRF3 and AP-1 in the presence of MERS-CoV PLpro. As shown in Fig. 6(b, c), we found that the activation of IRF3 and NF-kB was significantly inhibited by WT MERS-CoV PLpro-TM. However, the AP-1-dependent IFN-b expression was comparable to the control vector when it was cotransfected with MERS-CoV PLpro-TM (Fig. 6d). These findings suggested that MERS- CoV PLpro-TM suppressed IFN-b transcription by inter- fering with the IRF3- and NF-kB-mediated IFN expression signalling pathway.

To determine how MERS-CoV PLpro-TM inhibited the activation of IFN production signalling, Western blotting and immunofluorescence assays were conducted to detect the phosphorylation status and subcellular localization of IRF3 in PLpro-TM-transfected cells. We found that MERS- CoV PLpro-TM significantly inhibited the phosphorylation of IRF3, which was activated by RIG-IN (Fig. 6e, lane 4). Compared with the control, it had no inhibitory effect on the RIG-IN-mediated phosphorylation of IRF3, which was mock-transfected with PLpro-TM (Fig. 6e, lane 2). Interes- tingly, consistent with the inhibition on IRF3 phosphor- ylation, we observed that RIG-IN-induced IRF3 nuclear translocation (Fig. 6f) was markedly blocked in PLpro- TM-expressing cells. In contrast, IRF3 was exclusively located in the nuclei of the cells when the cells were trans- fected with RIG-IN only. Taken together, the data suggested that MERS-CoV PLpro inhibited IFN-b transcription and expression through blocking the phosphorylation and nuclear translocation of IRF3.

DISCUSSION

It is noteworthy that there were six human CoVs identified before 2012, but of these only SARS-CoV causes severe pulmonary disease with a high mortality rate in humans. In June 2012, MERS-CoV was isolated from a 60-year-old Saudi Arabian patient who died from acute respiratory distress syndrome with multiple organ failure (Zaki et al., 2012). Person-to-person transmission of MERS-CoV could spread the disease globally and become a serious health issue (Assiri et al., 2013). Currently, there are no specific anti-viral drugs to prevent the transmission of MERS-CoV or provide a cure for the disease.

In this study, we firstly identified the proteolytic processing activity of PLpro of MERS-CoV, the protease that can recognize and process the cleavage sites of nsp1-2, nsp2-3 and nsp3-4, and the catalytic residues required for MERS-CoV PLpro. Secondly, the LXGG consensus cleav- age sites at the N terminus of pp1a/1ab, which is generally essential for MERS-CoV PLpro-mediated processing, were characterized. We also demonstrated that MERS-CoV PLpro has DUB enzyme activity as observed in other human and animal CoVs (Barretto et al., 2005; Chen et al., 2007; Clementz et al., 2010; Frieman et al., 2009; Harcourt et al., 2004; Sun et al., 2010, 2012; Ziebuhr et al., 2007). Similar to human SARS-CoV PLpro and NL63-CoV PLP2, MERS-CoV PLpro is a viral deubiquitinating enzyme, which acts on both K48- and K63-linked ubiquitination and ISG15-linked ISGylation. Finally, we confirmed that MERS-CoV PLpro acts as an IFN antagonist through blocking the phosphorylation and nuclear translocation of IRF3. The characterization of protease, DUB and IFN antagonist activities of MERS-CoV PLpro would reveal the multilayered interactions between virus and host, and might contribute to the development of new strategies targeting PLpro for an effective prevention of MERS-CoV infection.

One of the intriguing findings in this study is that MERS- CoV PLpro is sufficient to process the substrates. By comparing the sequences with other CoVs, we found the cleavage site sequences at the N terminus of pp1a/1ab are highly conserved. MERS-CoV PLpro processed the polyproteins nsp1-2, nsp2-3 and nsp3-4 at the consensus cleavage site LXGG, which is also the target sequence recognized by many cellular DUB enzymes (Barretto et al., 2005; Sulea et al., 2005). Our results demonstrate that LXGG is an important determinant element for PLpro recognition and processing, which is in agreement with previous studies (Barretto et al., 2005; Chen et al., 2007). To our surprise, the consensus cleavage site LKGG in nsp2-3 is not absolutely required for PLpro recognition and processing, suggesting that other residues around the site would also be involved in PLpro recognition and processing. More studies with more mutations around the consensus cleavage site are warranted. Proteolytic proces- sing of the replicase polyprotein, which is essential for the proper assembly of the replication complex, plays a critical role in mediating RNA synthesis (Gosert et al., 2002; Kim et al., 1995; Sawicki et al., 2005). A previous study reported that MHV-CoV PLP1-mediated processing at CS1 and CS2 was important for efficient viral replication (Kanjanaha- luethai et al., 2003). Based on the findings of this study, our future research will aim to explore whether PLpro activity and processing of the substrates are necessary for MERS- CoV replication.

Another interesting result from this study is the demon- stration of MERS-CoV PLpro DUB activity, which is independent of the catalytic triad. Our results revealed that MERS-CoV PLpro is a new DUB enzyme that is encoded by the viral genome and may play important roles in CoV replication and pathogenesis. Cellular DUBs alter the fate of proteins in specific signalling pathways of the cell (Katz et al., 2010; Reyes-Turcu et al., 2009), for example deubi- quitination may protect the ubiquitinated proteins from degradation, and affect the proteins which are modified by Ub-like proteins, such as ISGs (van der Veen & Ploegh, 2012). For instance, a central gatekeeper in inflammation and immunity, A20, has DUB activity and removes K63- linked polyubiquitin lines from RIP1, TRAF6, RIP2 and NEMO, which results in the negative regulation of the innate immune response (Coornaert et al., 2009; Ma & Malynn, 2012). Cellular proteins, DUBA and CYLD, also negatively regulate the innate immune responses (Kayagaki et al., 2007; Sun, 2008, 2010). Furthermore, a lot of viral DUBs have similar structures to known cellular DUBs, such as USP14 and HAUSP, which have been identified as negative regulators of innate immunity (Chen et al., 2010; Kim et al., 2009; Ratia et al., 2006; Wang et al., 2011). For example, human CoV NL63, SARS-CoV and MERS- CoV, reported in this study, have contributed to encoding DUBs, and likely modulating the innate immune res- ponses. We previously reported that human CoV NL63 PLP2 reduced the ubiquitinated forms of STING, RIG-I, TBK1 and IRF3 (Sun et al., 2012). However, the underlying mechanisms by which MERS-CoV PLpro affects viral rep- lication and regulation of host innate immunity remain largely unknown. Here, for the first time, we report that MERS-CoV PLpro inhibits the conjugation of Ub, which is independent of the catalytic triad (C1592–H1759–D1774), and PLpro may target the Ub-proteasome pathway to promote virus replication and block the innate immunity. There is another possibility that PLpro may affect the proteins that are post-translationally modified by the Ub-like modifier, such as ISG15. We found MERS-CoV PLpro has deISGylation activity as a large number of ISG15- conjugated proteins are inhibited when PLpro is coex- pressed with ISG15. These results suggested that MERS- CoV PLpro may affect protein subcellular localization, protein activity and stability as well as signal transduction through these modifications and plays a critical role in the innate immune responses.
MERS-CoV causes a highly lethal infectious disease in human that is characterized by an aberrant immune response. The production of type I IFN, the major com- ponent of antiviral innate immunity, is firstly inhibited by MERS-CoV infection (Zielecki et al., 2013). Further- more, MERS-CoV fails to elicit strong type I or III IFN or pro-inflammatory innate immune responses (Chan et al., 2013; Kindler et al., 2013). IRF3 and NF-kB are the key
transcription factors for IFN-b expression (Barber, 2011; Dev et al., 2011; Marsili et al., 2012). The results of this study indicated that MERS-CoV PLpro markedly reduces the activation of the IFN-b, IRF3 and NF-kB promoters, which are induced by RIG-IN, indicating that MERS-CoV PLpro acts as an IFN antagonist. The immunofluorescence analysis shown in Fig. 6(f) indicates that IRF3 nuclear translocation is disrupted and remains located in the cytoplasm following cotransfection with PLpro and RIG- IN. This is consistent with previous studies of other CoVs (Devaraj et al., 2007; Xing et al., 2013). In line with this, Western blotting indicates that the phosphoryaltion of IRF3 is also blocked by PLpro. Thus, MERS-CoV may possess the ability to dampen human innate immune responses by PLpro through blocking the activation of IRF3 and regulation of the IFN response. While the underlying mechanisms currently remain unclear, further experiments are needed to explore the detailed mechanisms of how MERS-CoV PLpro regulates host antiviral innate immune responses.

In summary, the characterization of proteolytic processing, DUB and IFN antagonist activities of MERS-CoV PLpro would reveal the multilayered interactions between MERS- CoV and its host. These findings might be applicable for developing new strategies targeting PLpro for the effective control of MERS-CoV infection and investigating the roles of a viral DUB activity in CoV replication and pathogenesis.

METHODS

Cell and plasmid DNA. HEK293T and HeLa cells were cultured in Dulbecco’s modified Eagle’s medium containing 10% (v/v) FCS supplemented with penicillin (100 U ml21) and streptomycin (100 mg ml21). The reporter plasmids IFN-b-Luc, PRD (III-I)4-Luc, NF- kB-Luc and pAP-1-Luc, constructs HA-tagged Ub, HA-tagged Ub-K63 and HA-tagged Ub-K48 (all other K residues were mutated), and plasmids Ube1L, UbcH8, Myc-ISG15 and Flag-RIG-IN were described previously (Clementz et al., 2010; Devaraj et al., 2007; Sun et al., 2012).

Generating MERS-CoV PLpro-TM, substrate constructs and site-directed mutants. MERS-CoV PLpro-TM (Fig. 1a; amino acids 1474–2359 of pp1a of MERS-CoV EMC/2012 strain; GenBank accession number JX869059.1) and its three substrate constructs nsp1-2 (amino acids 1–330 of pp1a), nsp2-3 (amino acids 676– 1005 of pp1a) and nsp3-4 (amino acids 2581–2908 of pp1a) described in Fig. 1(b) were generated in this study. To obtain high- level expression in eukaryotic cells, we optimized the codon usage of these constructs based on the human codon usage frequency and the potential splicing sites and removed polyadenylation signal sequences from the coding regions. The optimized DNA was synthesized and cloned into pcDNA3.1-V5/HisB (Invitrogen) between the BamHI and EcoRI sites with in-frame fusion with a C-terminus V5 tag for PLpro-TM (Fig. 1a) or in-frame fusion with a Flag tag at the N terminus and a V5 tag at the C terminus for substrate constructs (Fig. 1b). To generate the catalytically inactive mutants of MERS-CoV PLpro-TM (C1592A, H1759A and D1774A) in Fig. 1(a) or specific mutants at the cleavage sites of substrates (Fig. 4a), we designed mutagenic primers (sequences are available on request) according to the methods previously reported (Zheng et al., 2004) and incorporated them into newly synthesized DNA using the QuikChange site-directed mutagenesis protocol (Stratagene). All introduced mutations were confirmed by DNA sequencing.

Cell-based trans-cleavage assay of MERS-CoV PLpro-TM. PLpro-TM or the catalytic mutants (4.5 mg) and the indicated sub- strates (0.5 mg) were transfected into HEK293T cells using Lipofec- tamine 2000 (Invitrogen) according to the manufacturer’s protocol. Twenty-four hours post-transfection, the cells were lysed for Western blotting to detect the cleavage products, as previously described (Sun et al., 2012).

Luciferase detection assay. The reporter plasmids (IFN-b-Luc, PRD(III-I)4-Luc, NF-kB-Luc or pAP-1-Luc) and Flag-RIG-IN, were cotransfected together with MERS-CoV PLpro-TM or the corres- ponding catalytic mutants into HEK293T cells using Lipofectamine 2000 according to the manufacturer’s protocol. Luciferase assay was performed as described previously (Sun et al., 2012; Xing et al., 2013). SARS-CoV PLpro-TM was used as the positive control of inhibitor for IFN-b expression (Clementz et al., 2010). Data between Vector and PLpros were subjected to unpaired, two-tailed Student’s t test using the Microsoft SPSS 12.0 software. A P value of ,0.05 was considered statistically significant (Kiefer et al., 2009; Vaux et al., 2012).

Assay of DUB and deISGylation activities. The assay of MERS- CoV PLpro-TM DUB was performed as described previously (Clementz et al., 2010; Xing et al., 2013). The effect of MERS-CoV PLpro-TM on ISG15-conjugated proteins in cultured cells was assessed as previously described (Evans et al., 2004; Zhao et al., 2005).

IRF3 phosphorylation detection. To assess IRF3 phosphorylation, we transfected Flag-RIG-IN (1.2 mg) in the presence or absence of V5-tagged PLpro-TM (3 mg) into HEK293T cells. IRF3 phosphoryla- tion detection assay was performed as described previously (Xing et al., 2013). Immunoflurescence assay. HeLa cells were grown to confluence in a six-well plate. Plasmid DNA expressing Flag-RIG-IN (1.2 mg per well) was transfected in the presence or absence of MERS-CoV PLpro-TM. Immunoflurescence assay was performed as described previously (Xing et al., 2013).

Sequence comparison and structure analysis of MERS-CoV PLpro. The protocol for sequence comparison and structure analysis of MERS-CoV PLpro was previously described (Barretto et al., 2005; Ratia et al., 2006). Briefly, the amino acid sequences of the PLpros from representative CoVs (Fig. 2a) were aligned with the MERS- CoV PLpro catalytic core domain (amino acid sequences 1476 to 1804) as the reference sequence using the program CLUSTAL 2.1. The 3D structure of MERS-CoV PLpro domains was modelled based on the structure of human SARS-CoV PLpro (PDB code, 2FE8) (Ratia et al., 2006) using SWISS-MODEL (Kiefer et al., 2009),USP25/28 inhibitor AZ1a an automated protein structure homology-modelling server (http://beta.swissmodel. expasy.org/workspace).