key: cord-336093-ic6q6ke8
authors: Sun, Ying; Wang, Zidao; Tao, Jiali; Wang, Yi; Wu, Andong; Yang, Ziwen; Wang, Kaimei; Shi, Liqiao; Chen, Yu; Guo, Deyin
title: Yeast-based assays for the high-throughput screening of inhibitors of coronavirus RNA cap guanine-N7-methyltransferase
date: 2014-02-11
journal: Antiviral Res
DOI: 10.1016/j.antiviral.2014.02.002
sha: 
doc_id: 336093
cord_uid: ic6q6ke8

The 5′-cap structure is a distinct feature of eukaryotic mRNAs and is important for RNA stability and protein translation by providing a molecular signature for the distinction of self or non-self mRNA. Eukaryotic viruses generally modify the 5′-end of their RNAs to mimic the cellular mRNA structure, thereby facilitating viral replication in host cells. However, the molecular organization and biochemical mechanisms of the viral capping apparatus typically differ from its cellular counterpart, which makes viral capping enzymes attractive targets for drug discovery. Our previous work showed that SARS coronavirus (SARS-CoV) non-structural protein 14 represents a structurally novel and unique guanine-N7-methyltransferase (N7-MTase) that is able to functionally complement yeast cellular N7-MTase. In the present study, we developed a yeast-based system for identifying and screening inhibitors against coronavirus N7-MTase using both 96-well and 384-well microtiter plates. The MTase inhibitors previously identified by in vitro biochemical assays were tested, and some, such as sinefungin, effectively suppressed N7-MTase in the yeast system. However, other compounds, such as ATA and AdoHcy, did not exert an inhibitory effect within a cellular context. These results validated the yeast assay system for inhibitor screening yet also demonstrated the difference between cell-based and in vitro biochemical assays. The yeast system was applied to the screening of 3000 natural product extracts, and three were observed to more potently inhibit the activity of coronavirus than human N7-MTase.

The cellular messenger RNAs (mRNAs) of higher eukaryotes and many viral RNAs are sequentially methylated at the N-7 and 2 0 -O positions of the 5 0 -guanosine cap. The cap structure has several important biological roles, such as protecting mRNA from degradation by 5 0 -3 0 exoribonucleases (Schwer et al., 1998) and directing pre-mRNA splicing and mRNA export from the nucleus (Darnell, 1979) . In addition, the cap structure confers stability to mRNAs and ensures their efficient recognition by translation initiation factor 4F for translation (Filipowicz et al., 1976; Schibler and Perry, 1977) . In contrast, host and virus RNA molecules with unprotected 5 0 -ends are degraded in cytoplasmic compartments (Liu and Kiledjian, 2006) . Uncapped RNAs, such as nascent viral transcripts, may also be detected as ''non-self'' by the RNA sensors RIG-I, Mda-5, and IFIT in host cells (Abbas et al., 2013; Bowzard et al., 2013; Hornung et al., 2006; Zust et al., 2011) , triggering antiviral innate immune responses through the production of interferon or IFIT (interferon-induced protein)-mediated antiviral activity (Daffis et al., 2010; Nallagatla et al., 2008; Rehwinkel et al., 2010) .

The following four cap-forming enzymes are involved in the formation of the cap-1 structure. (i) RNA triphosphatase hydrolyzes the 5 0 -triphosphate of nascent pre-mRNA to a 5 0 -diphosphate. (ii) RNA guanylyltransferase caps the diphosphate RNA with GMP. (iii) RNA guanine-N7-methyltransferase (N7-MTase) methylates the GpppN cap at the N7 position of guanine, resulting in the cap-0 structure (m7GpppN) (Shuman, 2001) . (iv) Ribose 2 0 -O-MTase further methylates the first nucleotide of higher eukaryotic cellular and viral mRNAs at the ribose 2 0 -OH position to form cap-1 (m7GpppNm) structures (Furuichi and Shatkin, 2000) . Both N7 and 2 0 -O-MTases utilize S-adenosyl-L-methionine (AdoMet) as a methyl http://dx.doi.org/10.1016/j.antiviral.2014.02.002 0166-3542/Ó 2014 Elsevier B.V. All rights reserved.

Abbreviations: SARS, severe acute respiratory syndrome; SARS-CoV, SARS coronavirus; nsp, nonstructural protein; N7-MTase, guanine-N7-methyltransferase; 2 0 -O-MTase, 2 0 -O-methyltransferase; AdoMet, S-adenosyl-L-methionine; AdoHcy, S-adenosyl-L-homocysteine; ATA, aurintricarboxylic acid; IC 50 , inhibitory concentration at 50% activity. donor and generate S-adenosyl-L-homocysteine (AdoHcy) as a byproduct.

Eukaryotic viruses that replicate in the cytoplasm encode their own capping apparatus, and the structure and mechanisms of the viral RNA capping apparatus are different from those of host cells (Furuichi and Shatkin, 2000) , which could be useful for the development of antiviral drugs. Indeed, a number of biochemical and functional studies have previously addressed methyltransferases as potential inhibitor targets (Chrebet et al., 2005; Schwer et al., 2001; Woyciniuk et al., 1995) . Moreover, the enzymes involved in the coronavirus capping pathway are increasingly considered to be promising targets for potential anti-coronavirus drugs .

Coronaviruses (CoVs), infect many species of animals, including humans, and cause acute or chronic respiratory diseases (e.g., severe acute respiratory syndrome coronavirus [SARS-CoV] , Middle East respiratory syndrome coronavirus [MERS-CoV] (de Groot et al., 2013; Zaki et al., 2012) , and infectious bronchitis virus [IBV] , enteric diseases (e.g., transmissible gastroenteritis virus [TGEV]), and central nervous system (CNS) diseases (murine hepatitis virus [MHV] ) (Weiss and Navas-Martin, 2005) . CoVs are the largest RNA viruses, are enveloped, and contain a single-stranded, positive-sense RNA genome ranging from 27 to 31.5 kb in length. The genome of SARS-CoV contains 14 open reading frames (ORFs) (Snijder et al., 2003) and generates 16 nonstructural proteins (nsps) produced by the autocatalytic processing of the polyprotein by 2 viral proteases (Ziebuhr et al., 2000) . CoVs replicate in the host cytoplasm and encode their own capping enzymes. Among the four capping enzymes involved in coronavirus m7GpppAm-cap formation, guanine-N7-methyltransferase (N7-MTase) was identified as nsp14 in our previous work using a yeast genetic system (Chen et al., 2009) , and 2 0 -O-methyltransferase (2 0 -O-MTase) is formed by nsp16 with nsp10 as a cofactor (Chang et al., 2011; Chen et al., 2011; Decroly et al., 2011; Lugari et al., 2010) . SARS-coronavirus nsp14 was previously characterized as a 3 0 -to 5 0 -exoribonuclease (ExoN) (Chen et al., 2007; Minskaia et al., 2006) , and the N7-MTase domain was mapped to the carboxy-terminal part of the protein (Chen et al., 2009) . Interestingly, the ExoN active site is dispensable, though the ExoN domain is required for N7-MTase activity. The combination of the two functional domains (Chen et al., 2009 (Chen et al., , 2013 , a unique feature among all N7-MTases, indicates that SARS-CoV N7-MTase is a novel form of RNA-processing enzyme and thus suggests it as an attractive target for the development of antiviral drugs.

It has been shown that the capping functions in yeast cells can be replaced by the cap-forming enzymes of mammals or DNA viruses (Ho et al., 2000; Saha et al., 1999 Saha et al., , 2003 , and we previously found that coronavirus nsp14 could replace yeast cap N7-MTase in vivo (Chen et al., 2009 (Chen et al., , 2013 . In the present study, we further developed the yeast genetic system as a high-throughput enzymatic activity assay platform of various N7-MTases to identify coronavirus N7-MTase inhibitors. The system was validated using MTase inhibitors previously identified by in vitro biochemical assays and then used for the screening of 3000 natural product extracts.

The coding sequences of nonstructural protein 14 of SARS-CoV, MHV, TGEV, and IBV were PCR amplified from cDNAs of the SARS-CoV WHU strain (GenBank accession No. AY394850) (Hussain et al., 2005) , MHV-A59 (GenBank accession No. AY700211.1), TGEV (GenBank accession No. FJ755618.2), and IBV (GenBank accession No. AJ311317.1) and inserted into the BamHI and XhoI sites of the yeast vector pMceK294A (2 lm, TRP1) (Chen et al., 2009) .

The cDNA of MERS-CoV nsp14 was chemically synthesized according to the deposited sequence of MERS-CoV (GenBank accession No. KF192507.1) and cloned into pMceK294A. For expression in E. coli cells, the coding sequences of SARS and MHV nsp14 were inserted into the NdeI and SalI sites of pET30a, and TGEV nsp14 (kindly provided by Dr. Luis. Enjuanes) was cloned into the BamHI and SalI sites of pETduet-1. The plasmid for IBV expression (pDest14-IBV-nsp14) was a kind gift from Dr. Eric J. Snijder.

E. coli BL21 (DE3) cells (Novagen) were separately transformed with the pET30a-SARS-nsp14, pET30a-MHV-nsp14, pET-duet1-TGEV-nsp14, and pDest14-IBV-nsp14 plasmids. The cells were cultured at 37°C in 1 L of LB medium supplemented with kanamycin (50 lg/ml) or ampicillin (100 lg/ml) until the culture density (A 600 ) reached 0.6-0.8 and then induced with 0.5 mM isopropylb-D-1-thiogalactopyranoside (IPTG) for 20 h at 16°C. The cells were harvested, and the recombinant proteins were purified by affinity chromatography using previously described protocols (Chen et al., 2007) . The protein was concentrated by ultra-filtration (Millipore), and the buffer was changed to 20 mM Tris-HCl (pH 8.0), 150 mM NaCl, 20% glycerol, and 0.2 mM DTT. The protein was stored at -70°C until use.

N7-MTase activity assays were carried out in a 30-lL reaction mixture (40 mM Tris-HCl [pH 7.5], 2 mM MgCl 2 , 2 mM DTT, 40 units RNase inhibitor, 0.01 mM SAM) with 1 lCi of S-adenosyl [methyl-3 H] methionine (67.3 Ci/mmol, 0.5 mCi/ml). The purified coronavirus nsp14 proteins were added at a final concentration of 200 nM using 0.5 mM GpppA cap analog as the substrate. After incubation at 37°C for 1.5 h, the reaction was stopped by adding an equal volume of stop solution (0.2% SDS and 20 mM EDTA). The samples were purified using DEAE-Sephadex columns, and the methylation of the RNA substrates was quantitated in counts per minute (cpm) using a Scintillation Counter (Beckman Coulter LS 6500).

To measure the inhibitory effects of coronaviral N7-MTase activity, AdoHcy, Aurintricarboxylic acid (ATA), ribavirin, and sinefungin were used at a maximum concentration of 100 lM in the assays. Typically, the enzyme and GpppA substrate were pre-incubated with each compound at room temperature (RT) for 20 min, and the reactions were initiated by the addition of S-adenosyl [methyl-3 H] methionine. A control reaction was performed in the presence of 1% DMSO instead of each tested compound. The IC 50 (concentration of the drug to repress 50% of the N7-MTase activity) value of sinefungin against SARS-CoV nsp14 was determined using GraphPad Prism5. The data were adjusted to a logistic doseresponse function.

The yeast N7-MTase gene Abd1 has been shown to be essential for yeast cell growth (Mao et al., 1995) . Saccharomyces cerevisiae strain YBS40 (MATa leu2 ade2 trp1 his3 ura3 can1 abd1::hisGp360-ABD1) (Mao et al., 1995) was transformed with yeast expression plasmids carrying coronavirus nsp14 and human N7-MTase (pYX232-HCM1) as a positive control (Chen et al., 2009; Saha et al., 1999) . Trp+ transformants were selected at 30°C on an agar medium lacking tryptophan. The cells were then streaked onto an agar medium lacking tryptophan and containing 0.75 mg/ml of 5-fluoroorotic acid (5-FOA) (Sikorski and Boeke, 1991) at 30°C for up to 5 days to counter-select the URA3+ plasmid carrying the yeast Abd1 gene. The formation of FOA-resistant colonies indicated that the transformation of the mutants could replace or complement the endogenous cap N7-methyltransferase function residing in the URA3 plasmid (Chen et al., 2009 (Chen et al., , 2013 .

A solution of sinefungin (adenosylornithine, CALBioCHEM 5670-51), ribavirin (1-b-D-ribofuranosyl-1,2,4-triazole-3-carboxamide, Sigma R-9644), and AdoHcy (adenosine-homocysteine, Sigma A-9384) was prepared in H 2 O or DMSO (Decroly et al., 2008; Luzhkov et al., 2007; Milani et al., 2009) . ATA (aurintricarboxylic acid, Sigma A-1895) was dissolved in 0.1 mM NaOH (Decroly et al., 2008) . Concentrations of the stock solutions were set to 10 mM, and the compounds were stored at À20°C.

Microtiter plates (96-or 384-well plates) were used for the yeast growth suppression test and inhibitor screening. A single transformed colony of the YBS40 strain containing plasmids expressing human N7-MTase (MT-Human), SARS-CoV N7-MTase (MT-SARS), N7-MTases of other coronaviruses (MT-MHV, MT-TGEV, and MT-IBV), and the pMceK294A vector as control (representing the yeast N7-MTase [MT-Yeast]), were inoculated separately into 5 ml of a basic medium (Min SD Base) lacking tryptophan and incubated at 30°C for 21-24 h until reaching a similar final cell density in the stationary phase (0.5-1.0 Â 10 8 cells/ml) (Chrebet et al., 2005) . The starting cell density of the yeast-based assay was A 595 = 0.01 and contained different inhibitor candidates or 1% DMSO as a negative control. For high-throughput screening, 30 lg/ml or 15 lg/ml natural product extract was added into the 96-well or 384-well microtiter plates. The starting volumes for the 96-well and 384-well microtiter plates were 100 lL and 50 lL, respectively, and the final cell density (A 595 ) was measured using a Multifunctional Microplate Reader (Tecan GENios) after incubation for 20 h at 30°C in a shaker.

After incubation, yeast cells were collected and then ground in liquid nitrogen. Total RNA was isolated from cells using Trizol reagent (Invitrogen) and subjected to real-time PCR analysis to measure mRNA expression. Real-time PCR was performed using Faststart Universal SYBR Green Mastermix (Roche) and analyzed on 7500 Real Time PCR System (Applied Biosystems). Gene-specific primer sequences are as follows: N7-MTase of the MT-SARS, forward: 5 0 -GATCATTACTGGTCTTCATCCTACA-3 0 and reverse: 5 0 -AATC CACGCACGAACGTGACGAATA-3 0 ; N7-MTase of the MT-Human, forward: 5 0 -AGCTCAACAGTGGCTGCCCATTACA-3 0 and reverse: 5 0 -ATC GGCAATATCAGTACAAACTAGC-3 0 ; N7-MTase of the MT-Yeast and YBS40, forward: 5 0 -TGTCACAAGAAGACTATGACCGTCA-3 0 and reverse: 5 0 -ATGGATCTATTTGCAGTCATCTCTA-3 0 ; Beta-actin, forward: 5 0 -AAGACACCAAGGTATCATGGTCGG-3 0 and reverse: 5 0 -CGGAAG AGTACAAGGACAAAACGGC-3 0 .

The microbial natural product extracts were obtained from the microbial natural product library at Hubei Biopesticide Engineering Research Centre (HBERC) and consisted of purified secondary metabolites, semi-purified fractions and extracts from actinomycetes and fungi isolated from soil, lichen, fresh leaves, organic samples, and mushrooms using classical isolation methods. For the soil or lichen, 1 mg was quantified and placed into a tube containing 10 ml of sterilized sodium lauryl sulfate solution and shaken for 1 min. The other samples were seeded aseptically and chopped into small pieces using a blender. The samples were then plated on two different agar media and incubated at 28°C for 7-9 days. The actinomyces and fungi were isolated to pure culture and incubated at 28°C. The microbial isolates were morphologically characterized, stored and archived. The ratio of actinomycetes/fungi in the collection of microorganisms at HBERC was around 2.5:1.

The microbial identity will be further characterized by classical and molecular methodology when active compounds are identified from the corresponding microbial extracts. For preparation of the microbial extracts, each microbial isolate was fermented in 200 ml fermentation medium and the resulting cultures were extracted with ethyl acetate. For making the working solution of the microbial extracts, the extracts were dissolved in ethanol to a concentration of 1 mg/ml (w/v). In total, 3000 microbial culture extracts were used for MTase inhibitor screening in this study, which were prepared from actinomycetes and fungi.

Our previous study showed that SARS-CoV nsp14 possesses N7-MTase activity that could substitute for cellular N7-MTase in yeast (Chen et al., 2009 (Chen et al., , 2013 . In this system, the gene Abd1 encoding yeast N7-MTase is knocked out from the yeast genome and complemented by expression of Abd1 from a plasmid with a URA3 selective marker, resulting in yeast strain YBS40, whereas the exogenous N7-MTase is expressed from another plasmid with a Trp selective marker. The expression of a functional URA3 gene encoding orotine-5 0 -monophosphate decarboxylase results in the conversion of the nontoxic 5-FOA compound to toxic 5-flurouracil in yeast. In a medium containing 5-FOA, which counter-selects URA3-expressing cells, the yeast cells can grow normally only when the exogenous protein can replace the function of the cellular N7-MTase (Fig. 1A) . In this study, we confirmed that SARS-CoV and human N7-MTase could rescue the growth of yeast cells that were deficient in N7-MTase. We further tested whether this N7-MTase activity is universal for nsp14 from other coronaviruses. As shown in Fig. 1B , nsp14 from other coronaviruses, including TGEV (group 1), MHV (group 2a), SARS-CoV (group 2b), MERS-CoV (group 2c), and IBV (group 3), showed N7-MTase activity that was functional in vivo by complementing the endogenous yeast enzyme. These results indicate that N7-MTase is well conserved among different coronaviruses and that nsp14 appears to be a target for the development of a universal inhibitor against coronaviruses. We then generated three yeast strains that grew dependently on the N7-MTase activity from SARS-CoV (MT-SARS), humans (MT-human), and yeast (MT-yeast).

To compare the expression levels of these N7-MTases, we analyzed the mRNA levels of the N7-MTases using real-time PCR. As shown in Fig. 1C , there are no significant differences between the mRNA levels of MT-SARS and MT-human, which are under the control of the same yeast TPI1 promoter. The mRNA of MTyeast, which was transformed with pMceK294A vector as control, is similar with yeast train YBS40, because the mRNAs of yeast N7-MTases are transcribed from the complementary plasmids (URA3) bearing the yeast-MT. In the strains with MT-SARS and MT-human, the expression of yeast N7-MTase could not be detected (data not shown). As the expression of MT-SARS and MT-human is at the similar level, these yeast strains could be used to identify the inhibitors that inhibit SARS-CoV MTase more potently than human MTase. However, the mRNA levels of MT-SARS and MT-human are four times higher than that of MT-yeast, the latter being expressed from a low-copy-number plasmid with centromeric element (Cen), the inhibitory activity on SARS-CoV and yeast MTases cannot be directly compared.

To establish a drug screening system using these yeast strains, we first tested whether the expression of viral or human N7-MTase influenced yeast growth. Individual colonies of MT-Yeast, MT-Human, and MT-SARS were cultured in a medium lacking tryptophan, and the pre-cultures were diluted with fresh medium to an A 595 = 0.01. A 100-ll aliquot of the culture was incubated in each well of two 96-well plates, one of which was used for measuring yeast propagation by spectrophotometry, whereas the other plate was used for the direct counting of the cell number. Measurements were performed at 4-h intervals over 48 h. As shown in Fig. 1D and E, all three yeast strains (MT-SARS, MT-human, and MT-yeast) showed similar growth curves within 48 h, indicating that both the SARS-CoV and human MTases could support the normal growth of yeast cells. At such a growth condition, starting with a liquid culture at A 595 = 0.01, the time period between 4 and 24 h corresponded to the logarithmic growth phase (Fig. 1D and E) . During this growth phase, the optical density (A 595 ) was well correlated with the cell number as demonstrated for the yeast strain MT-SARS by linear regression analysis (Fig. 1F ). In the linear regression equation, y = 8.60359x À 0.67266 (R 2 = 0.924, R 2 is regressive coefficient), it shows that there is a correlativity between the A 595 of culture liquid and cell number in the logarithmic phase (4-24 h) of MT-SARS when The least square method was used to evaluate the interrelationships between the two variables (X and Y), where X represents the spectrophotometric measurement (A 595 ) and Y represents the corresponding cell number in the logarithmic phase (4-24 h) of MT-SARS. The linear regression equation for the MT-SARS strain, y = 8.60359x À 0.67266 (R 2 = 0.924, R 2 is regressive coefficient), was obtained through normative analysis in least square method with Excel program. In this linear regression equation, there is a correlativity between the A 595 of culture liquid and cell number in the logarithmic phase (4-24 h) of MT-SARS when the R 2 > 0.9. the R 2 > 0.9. Similarly, there was also a high correlativity between the two optical absorbance and the cell number for strains MT-Yeast and MT-Human, with linear regression equations y = 8.379x À 1.0337 (R 2 = 0.926) and y = 8.514x À 0.6286 (R 2 = 0.922), respectively. As spectrophotometric measurements can facilitate rapid and high-throughput evaluation of microtiter plates, the optical density of the cell culture was applied to measure yeast growth in the ensuing experiments.

To validate the yeast-based system for identifying and screening inhibitors against viral N7-MTases, we tested previously reported MTase inhibitors, including AdoHcy, sinefungin, aurintricarboxylic acid (ATA), and ribavirin . We first detected the N7-MTase activities of coronaviruses nsp14 in vitro when incubated with different compounds at 100 lM using biochemical assays. As shown in Fig. 2A , sinefungin effectively inhibits the activities of all four N7-MTases and AdoHcy and ATA weakly inhibit the N7-MTase activities of SARS, MHV and TGEV with approximately 45-70% inhibition at 100 lM, except for IBV with 10-30% inhibition.

In contrast, ribavirin, which is a guanosine (ribonucleic) analog used to interfere with viral RNA synthesis and viral mRNA capping of many different viruses (Kim and Lee, 2013; Scholtissek, 1976; Zhao et al., 2012) , did not inhibited the coronavirus N7-MTases. The doseresponse curve of sinefungin, which was the best inhibitor of the three effective compounds, showed an IC 50 value of sinefungin against SARS-CoV nsp14 of 383.4 ± 7.54 nM (Fig. 2B) . Sinefungin could widely inhibit the activities of MTases from diverse sources including viruses and yeast. For example, the IC 50 values of sinefungin for N7-MTase and 2 0 O-MTases of Vaccinia virus in vitro are 12.0 and 39.5 nM, respectively (Pugh et al., 1978) , and the IC 50 values for 2 0 O-MTase (NS5MTase) activity of Dengue virus in vitro are 420 nM (Li et al., 2007) and 630 nM . The IC 50 values of sinefungin for yeast N7-MTase are 55 nM in vivo (Chrebet et al., 2005) and 24 nM in vitro (Zheng et al., 2007) , respectively. It was reported previously that the IC 50 values of AdoHcy are 1 and 0.5 mM for N7-MTase and 2 0 O-MTases activities of Vaccinia virus in vitro, respectively (Pugh and Borchardt, 1982) . In this study, the inhibitory effects of sinefungin, AdoHcy and ATA on MTases were consistent with that of previous reports.

As sinefungin, AdoHcy and ATA were effective inhibitors of coronaviruses N7-MTases in biochemical assays, they were used to validate the yeast-based assay system established in this study (Fig. 3) . The yeast strains MT-Yeast, MT-Human, and MT-SARS were incubated with 100 lM of each compound, and the cell growth was measured at 4-h intervals over a 48-h period. Remarkably, sinefungin significantly repressed the growth of all three yeast strains (Fig. 3) , confirming the previous report of sinefungin as a potent inhibitor (Chrebet et al., 2005; Li et al., 2007; Pugh et al., 1978) . In contrast, ATA, AdoHcy, and ribavirin did not exhibit inhibitory activities against the three yeast strains. Because Ado-Hcy and ATA could inhibit the activity of coronavirus N7-MTase in biochemical assays, the results showed a clear difference between in vitro biochemical assays and in vivo cell-based assays, indicating that N7-MTase inhibitors identified using in vitro biochemical assays may not necessarily be effective within a cellular context. Such discrepancy may be attributed to the low membrane permeability to AdoHcy and ATA or that the IC 50 of AdoHcy and ATA are far greater than 100 lM. Indeed, sinefungin has been reported to be actively transported across yeast membranes, whilst AdoHcy and ATA are not (Zheng et al., 2006 (Zheng et al., , 2007 .

Sinefungin, which is produced by Streptomyces, is a natural sulfur-modified analog of AdoHcy, a well-characterized, nonspecific MTase inhibitor. As shown in Fig. 3 , sinefungin exhibited a similar inhibitory effect on all three yeast strains at high concentration (100 lM). However, stronger inhibitory effect on MT-Yeast was observed than that on MT-Human and the strains harboring the MTases of other coronaviruses at low concentration (0.1 lM) (Fig. 4A and B). This could be explained by intrinsic difference of inhibitory effect to different MTases or the difference of the expression levels of the MTases as we showed that higher mRNA levels of MT-Human and MTases of coronaviruses were expressed than that of MT-yeast (Fig. 1C) . To further characterize sinefungin as an MTase inhibitor in the yeast-based system, yeast strain YBS40, expressing different N7-MTases, was incubated with sinefungin at different concentration for 20 h (Fig. 4C and D) . The IC 50 (n = 3, mean values ± SD) of MT-Yeast, MT-Human and MTases of coronaviruses were calculated and shown in Table 1 . These results indicate that sinefungin is a broad-spectrum inhibitor against various species of N7-MTases, including those from yeasts, humans, and coronaviruses and demonstrate that the yeast-based system could be used for analysis of N7-MTase inhibitors. However, sinefungin is not an ideal antiviral inhibitor due to its lack of specificity on coronavirus N7-MTases compared to yeast and human MTases. Therefore, it is still necessary to screen for specific viral N7-MTases inhibitors, an effort that will benefit the functional study of viral MTases and clinical drug development.

3.3. Screening for effective microbial natural products using the yeastbased assay

As the three yeast strains, MT-Yeast, MT-Human, and MT-SARS, have the same genetic background and their growth depends on yeast, human, and SARS-CoV MTases, respectively, the yeast-based system described above can be adopted to screen specific inhibitors against coronavirus N7-MTase. It is of note that the mRNA levels of MT-SARS and MT-Human are similar but they are higher than that of MT-yeast (Fig. 1C) . However, this would not influence the screening for inhibitors with more specific inhibition on SARS-CoV MTase over human MTase.

For the screening of inhibitors to SARS-CoV N7-MTase, over 3000 microbial natural product extracts were tested in the yeastbased system. The microbial extracts were prepared from liquid culture of actinomycetes and fungi as described in the Materials and Methods section. The overnight cultures of MT-Yeast, MT-Human, and MT-SARS were diluted to A 595 = 0.01, and a 100-lL aliquot of each cell culture was seeded into 96-well microtiter plates in the presence of 30 lg/ml of microbial natural product extract. Twenty hours post-incubation, the cell density was monitored by measuring the optical density at A 595 . In the first round, the extracts that did not repress MT-Human were selected for further testing. The primary candidate extracts were tested twice for reproducibility and inhibitory effects using independent samples of each extract. In summary, we obtained three candidates extracts, PF35468 (natural product extract from a fungus species, preliminarily characterized as Penicillium spp.), PA48202 and PA48523 (natural product extracts from actinomycetes, preliminarily characterized as Streptomyces spp.), which could potently inhibit the growth of MT-SARS and MT-yeast but significantly less effectively inhibit MT-Human (Fig. 5) . The inhibition ratios of the 3 extracts on MT-SARS to MT-Human were 3.5, 3, and 9, respectively. In contrast, sinefungin suppressed the growth of all 3 yeast strains at a final concentration of 10 lM (3.81 lg/ml). The fungal and actinomyces natural product extracts used in this study were composed of complex ingredients, and these three extracts could be used for the isolation of the active ingredients in future work. Taken together, the yeast-based assay system could provide a platform for screening N7-MTase inhibitors that are more effective against viral MTases and less effective against human MTases.

Efficient drug development often depends on high-throughput screening, which offers rapid and sensitive data acquisition coupled with high-content analyses. To explore the possibility of miniaturizing the yeast-based assay and to increase its throughput scale, yeast cells were seeded in 384-well microtiter plates, and the growth of the MT-SARS yeast cells was monitored at 4-h intervals over 48 h at 30°C in a shaker ( Fig. 6A and B) . During the logarithmic growth phase (4-24 h), the optical density of the culture at A 595 was well correlated with the cell number ( Fig. 6A-C) . The following linear regression equation was established for the MT-SARS strain: y = 8.82419x À 0.12315 (R 2 = 0.992), indicating that they are more correlative in 384-well microtiter plates than 96-well microtiter plates (Fig. 6C) . To validate the yeast-based system for the highthroughput screening of inhibitors against N7-MTases, we performed a growth suppression assay and measured the IC 50 of sinefungin against the 6 yeast strains that had been tested in the 96-well format ( Fig. 6D and E) . The IC 50 values (n = 3, mean values ± SD) for sinefungin against MT-Yeast, MT-Human, MT-SARS, MT-MHV, MT-TGEV, and MT-IBV were 39.18 ± 12.13 nM, 623 ± 1.36 nM, 1621 ± 2.14 nM, 1349 ± 3.33 nM, 1398 ± 1.34 nM, and 1430 ± 4.87 nM, respectively (Table 1) , which were highly consistent with that of the 96-well microtiter plates (Figs. 4 and 6 and Table 1 ). These results indicate that the yeast-based high-throughput screening system could be used for the discovery of specific antiviral inhibitors. In comparison to the 96-well microplate assays, the 384-well assay system was more reliable, and cost effective. 

Many coronaviruses cause severe human and animal diseases. Recently, a novel SARS-like coronavirus, the Middle East respiratory syndrome coronavirus (MERS-CoV) belonging to the genus Betacoronavirus, was identified in Saudi Arabia and subsequently spread to the United Kingdom (UK), Germany, and France (Al-Tawfiq, 2013; Lim et al., 2013; Wiwanitkit, 2014) . As there are no clinically approved antiviral therapies, including effective antiviral drugs, available to date for the treatment of coronavirusassociated diseases, the establishment of a high-throughput antiviral inhibitor screening system is important for the development of novel anti-coronavirus drugs. a The numbers represent the mean ± standard deviations (SD) with n = 6. Fig. 5 . The inhibitory effects of microbial natural product extracts screened using the yeast-based system. Sinefungin as an N7-MTase inhibitor was used as a positive control, and yeast culture in the absence of extracted compounds was used as the mock control. PF35468, PA48202, and PA48523 represent the three microbial natural product extracts. The growth suppression in percentage was calculated as [(X À Y)/X] Â 100% where X is the A 595 value of the mock culture and Y is the A 595 value in the presence of sinefungin or microbial natural product extracts. The experiments were repeated three times. The results are shown as the percent suppression of cell growth of yeast strains MT-SARS, MT-human, and MT-Yeast.

Although AdoHcy, ATA and sinefungin, were previously reported to be inhibitors of coronavirus RNA MTases in vitro , only sinefungin significantly suppressed the growth of the MT-yeast, MT-human, and MT-SARS yeast cells (Fig. 3 ). Among these three strains, sinefungin was more effective against MT-Yeast than MT-SARS (Fig. 4) . Thus, the observed differences between in vitro biochemical assays and yeast-based assays indicate that the inhibitors identified by an in vitro biochemical assay may not be effective in cells. As the yeast cell-based assay system is more similar to host cells with regard to the complexity of chemical composition than an in vitro biochemical assay, the former may have advantages over the latter for the identification of antiviral inhibitors that can function effectively in living cells.

Although the yeast-based N7-MTase assay system was validated by testing previously identified inhibitors, no specific inhibitors against coronavirus N7-MTases were identified. As the yeast-based system developed in this study included three yeast strains carrying yeast, human, or coronavirus N7-MTase activity, this method could be used to screen and identify inhibitors that are specific against coronaviruses and not (or significantly less effective) to human N7-MTase. In addition, the structural and mechanistic differences among coronavirus, yeast, and human MTases also provide the rationale for developing specific inhibitors. Accordingly, we exploited these differences and screened over 3000 natural product extracts. Interestingly, three extracts showed more potent inhibitory effects on SARS-CoV and yeast N7-MTase than human N7-MTase (Fig. 5) .

A useful drug screening system should be able to be operated in a high-throughput manner. Therefore, the yeast cell-based system of this study was optimized for 384-well microtiter plates. In this system, the yeast cells grew well and did not sediment (Fig. 6) . Currently, only the AdoMet and AdoHcy backbones are used to synthesize analogs for the development of inhibitors that are specific for viral N7-MTases (Lim et al., 2011) . The high-throughput yeastbased platform makes it possible to screen a large number of natural product extracts. The yeast-based system can be further used to screen a chemically synthesized compound library. The compounds obtained from such screenings may contribute to the drug design and development needed for the control of coronaviruses. 

Structural basis for viral 5 0 -PPP-RNA recognition by human IFIT proteins

Middle east respiratory syndrome-coronavirus infection: An overview

In vitro reconstitution of SARS-coronavirus mRNA cap methylation

RIG-I goes beyond naked recognition

Construction and genetic analysis of murine hepatitis virus strain A59 Nsp16 temperature sensitive mutant and the revertant virus

Biochemical characterization of exoribonuclease encoded by SARS coronavirus

Functional screen reveals SARS coronavirus nonstructural protein nsp14 as a novel cap N7 methyltransferase

Biochemical and structural insights into the mechanisms of SARS coronavirus RNA ribose 2 0 -O-methylation by nsp16/nsp10 protein complex

Structure-function analysis of severe acute respiratory syndrome coronavirus RNA cap guanine-N7-methyltransferase

Cell-based assays to detect inhibitors of fungal mRNA capping enzymes and characterization of sinefungin as a cap methyltransferase inhibitor

0 -O methylation of the viral mRNA cap evades host restriction by IFIT family members

Transcription units for mRNA production in eukaryotic cells and their DNA viruses

Middle east respiratory syndrome coronavirus (MERS-CoV): announcement of the coronavirus study group

Crystal structure and functional analysis of the SARS-coronavirus RNA cap 2 0 -O-methyltransferase nsp10/nsp16 complex

Coronavirus nonstructural protein 16 is a cap-0 binding enzyme possessing (nucleoside-2 0 O)-methyltransferase activity

A protein binding the methylated 5 0 -terminal sequence, m7G pppN, of eukaryotic messenger RNA

Viral and cellular mRNA capping: past and prospects

A yeast-based genetic system for functional analysis of viral mRNA capping enzymes

0 -Triphosphate RNA is the ligand for RIG-I

Identification of novel subgenomic RNAs and noncanonical transcription initiation signals of severe acute respiratory syndrome coronavirus

Ribavirin efficiently suppresses porcine nidovirus replication

Vesicular stomatitis viruses resistant to the methylase inhibitor sinefungin upregulate RNA synthesis and reveal mutations that affect mRNA cap methylation

Middle east respiratory syndrome coronavirus (MERS CoV): update 2013

Small molecule inhibitors that selectively block dengue virus methyltransferase

Decapping the message: a beginning or an end

Molecular mapping of the RNA Cap 2 0 -O-methyltransferase activation interface between severe acute respiratory syndrome coronavirus nsp10 and nsp16

Virtual screening and bioassay study of novel inhibitors for dengue virus mRNA cap (nucleoside-2 0 O)-methyltransferase

Yeast mRNA cap methyltransferase is a 50-kilodalton protein encoded by an essential gene

Flaviviral methyltransferase/RNA interaction: structural basis for enzyme inhibition

Discovery of an RNA virus 3 0 ? 5 0 exoribonuclease that is critically involved in coronavirus RNA synthesis

A brilliant disguise for self RNA: 5 0 -end and internal modifications of primary transcripts suppress elements of innate immunity

Effects of S-adenosylhomocysteine analogues on vaccinia viral messenger ribonucleic acid synthesis and methylation

Sinefungin, a potent inhibitor of virion mRNA(guanine-7-)-methyltransferase, mRNA(nucleoside-2 0 -)-methyltransferase, and viral multiplication

RIG-I detects viral genomic RNA during negative-strand RNA virus infection

Characterization of human, Schizosaccharomyces pombe, and Candida albicans mRNA cap methyltransferases and complete replacement of the yeast capping apparatus by mammalian enzymes

Yeast-based genetic system for functional analysis of poxvirus mRNA cap methyltransferase

The 5 0 -termini of heterogeneous nuclear RNA: a comparison among molecules of different sizes and ages

Inhibition of influenza RNA synthesis by virazole (ribavirin)

Characterization of the mRNA capping apparatus of Candida albicans

Accelerated mRNA decay in conditional mutants of yeast mRNA capping enzyme

Biochemical characterization of the (nucleoside-2 0 O)-methyltransferase activity of dengue virus protein NS5 using purified capped RNA oligonucleotides (7Me)GpppAC(n) and GpppAC(n)

Structure, mechanism, and evolution of the mRNA capping apparatus

In vitro mutagenesis and plasmid shuffling: from cloned gene to mutant yeast

Unique and conserved features of genome and proteome of SARS-coronavirus, an early split-off from the coronavirus group 2 lineage

Coronavirus pathogenesis and the emerging pathogen severe acute respiratory syndrome coronavirus. Microbiol

Novel middle east respiratory syndrome coronavirus

The methyltransferase inhibitor neplanocin A interferes with influenza virus replication by a mechanism different from that of 3-deazaadenosine

Isolation of a novel coronavirus from a man with pneumonia in Saudi Arabia

Interference with ERK and STAT signaling pathways and inhibition of hepatitis C virus replication by ribavirin

Mutational analysis of Encephalitozoon cuniculi mRNA cap (guanine-N7) methyltransferase, structure of the enzyme bound to sinefungin, and evidence that cap methyltransferase is the target of sinefungin's antifungal activity

Sinefungin resistance of Saccharomyces cerevisiae arising from Sam3 mutations that inactivate the AdoMet transporter or from increased expression of AdoMet synthase plus mRNA cap guanine-N7 methyltransferase

Virus-encoded proteinases and proteolytic processing in the Nidovirales

Ribose 2 0 -O-methylation provides a molecular signature for the distinction of self and non-self mRNA dependent on the RNA sensor Mda5

We thank Dr. S. Shuman for kindly providing the yeast plasmids and strains and Dr. Eric J. Snijder for the IBV-nsp14 E. coli expression plasmid. We are grateful to Dr. Luis. Enjuanes for providing the TGEV cDNA. This study was supported by the China ''973'' Basic Research Program (2010CB911800 and 2013CB911101), China NSFC grants (81130083, 31170152, 81271817, and 31221061) and China Specialized Research Fund for the Doctoral Program of Higher Education (20120141130008). D.G. is supported by Hubei Province' s Outstanding Medical Academic Leader Program.