key: cord-0274741-5xd97edj
authors: Yuan, Xuye; Kadowaki, Tatsuhiko
title: DWV 3C protease uncovers the diverse catalytic triad in insect RNA viruses
date: 2022-01-06
journal: bioRxiv
DOI: 10.1101/2022.01.06.475182
sha: fd4dbc05604c433d6a21f26ec88a005f71b4e7f7
doc_id: 274741
cord_uid: 5xd97edj

Deformed wing virus (DWV) is the most prevalent Iflavirus that is infecting honey bees worldwide. However, the mechanisms of its infection and replication in host cells are poorly understood. In this study, we analyzed the structure and function of DWV 3C protease (3Cpro), which is necessary for the cleavage of the polyprotein to synthesize mature viral proteins. We found that the 3Cpros of DWV and picornaviruses share common enzymatic properties, including sensitivity to the same inhibitors, such as rupintrivir. The predicted structure of DWV 3Cpro by AlphaFold2, the predicted rupintrivir binding domain, and the protease activities of mutant proteins revealed that it has a Cys-His-Asn catalytic triad. Moreover, 3Cpros of other Iflaviruses and Dicistrovirus appear to contain Asn, Ser, Asp, or Glu as the third residue of the catalytic triad, suggesting diversity in insect RNA viruses. Both precursor 3Cpro with RNA-dependent RNA polymerase and mature 3Cpro are present in DWV-infected cells, suggesting that they may have different enzymatic properties and functions. DWV 3Cpro is the first 3Cpro characterized among insect RNA viruses, and our study uncovered both the common and unique characteristics among 3Cpros of Picornavirales. Furthermore, the specific inhibitors of DWV 3Cpro could be used to control DWV infection in honey bees.

Winter colony loss in honey bees is strongly correlated with the presence of deformed wing virus (DWV) and the ectoparasitic mites, Varroa destructor and Tropilaelaps mercedesae (Highfield et al., 2009; Nazzi and Le Conte, 2016) . The mites transmit DWV to the host (de Miranda and Genersch, 2010; Rosenkranz, Aumeier and Ziegelmann, 2010; Chantawannakul et al., 2018; Wu et al., 2020) and increase viral loads (Shen et al., 2005; Wu, Dong and Kadowaki, 2017; Forsgren et al., 2009; Khongphinitbunjong et al., 2016; Wu et al., 2020) . With the spread of V. destructor, DWV has become the most prevalent virus that is infecting honey bees worldwide (Wilfert et al., 2016) . Honey bees often show multiple symptoms due to high DWV levels, including the death of pupae, deformed wings, shortened abdomen, and reduced lifespan (Yue et al., 2007; Tentcheva et al., 2004; de Miranda and Genersch, 2010; Rosenkranz, Aumeier and Ziegelmann, 2010) .

Picornavirales. It has a non-enveloped virion approximately 30 nm in diameter and contains a positive-strand RNA genome. The RNA genome is translated into a polyprotein that is processed by the 3C protease (3C pro ) to produce mature viral proteins (Lanzi et al., 2006) . The crucial 3C pro s have been well characterized in picornaviruses and share similar structures and functions (Matthews et al., 1994; Bergmann et al., 1997; Mosimann et al., 1997; Birtley et al., 2005; Lee et al., 2009; Ohlenschläger et al., 2004; Cui et al., 2011) . Indeed, all of them have a Cys-His-Asp/Glu (Asp or Glu) catalytic triad. 3C pro plays important roles in the viral life cycle and host-virus interactions. For example, 3C pro not only processes a viral polyprotein but also cleaves the specific host proteins necessary for transcription, translation, and nucleocytoplasmic trafficking to modify the cell physiology for viral replication. Furthermore, 3C pro and its precursor with RNA-dependent RNA polymerase (RdRP) bind to the 5'UTR region of the viral genome RNA to initiate replication (Yi et al., 2021; Sun et al., 2016) . Based on structural conservation studies, inhibitors of picornavirus 3C pro s have been identified and characterized to develop potential antiviral treatments (Lall, Jain and Vederas, 2004) .

DWV is the best characterized virus among honey bee viruses; however, very little is known about its mechanism of infection and replication in host cells. In this study, we expressed and purified DWV 3C pro as a glutathione-S-transferase (GST) fusion protein to characterize the protease activity as well as the inhibitors. We also studied the structure-function relationship of DWV 3C pro based on the structure predicted by AlphaFold2 as well as the protease activities of mutant proteins compared to that of the 3C pro s of other Iflaviruses and Dicistrovirus. We also discussed the mechanism of synthesis and maturation of DWV 3C pro in infected cells.

We expressed DWV 3C pro as a GST-fusion protein and further included the potential VPg region (2094-2353 of DWV polyprotein) to make it soluble in E. coli. Protease activity of DWV 3C pro was measured by a fluorescence resonance energy transfer (FRET) assay using a 15 amino acid peptide with the potential cleavage site, AKPEMD, as the substrate. This six amino acid sequence is present at three sites in the DWV polyprotein. The optimum pH for protease activity was 6 ( Fig. 1A) , and the optimum temperature was between 25-35 ℃ (Fig. 1B) . Dithiothreitol was required for maximum activity as a cysteine protease ( Fig. 1C ) (Wilkesman, 2017) . The protease activity (RFU/min) was proportional to the amount of GST-DWV 3C pro added in the reaction mixture ( Fig. 1D ). Based on the Michaelis-Menten equation, Km, Vmax, and Kcat/Km values were 4.233 ± 0.256 μM, 0.2538 ± 0.0067 μM/min, and 59.96 ± 2.032 mM -1 min -1 , respectively (Fig. 1E ).

We next tested the effects of known protease inhibitors of picornaviruses and other viruses on DWV 3C pro . Eight compounds inhibited enzyme activity with various IC50 values, as shown in Table 1 . Figure 2 shows the dose-response curves for rupintrivir (Patick et al., 1999) , ebselen (Amporndanai et al., 2021) , disulfiram (Ma et al., 2020) , GC376 (Kim et al., 2012 ), carmofur (Jin et al., 2020b Ma et al., 2020 ), 6,7-dichloroquinoline-5,8-dione (Jung et al., 2018 , quercetin (Yao et al., 2018) , and zinc. Rupintrivir was the most effective compound and the IC50 value (0.36 μM) was lower than that of enterovirus 3C pro (1.65-7.3 μM) but higher than that of human rhinovirus 3C pro (5 nM) (Tan et al., 2016; Wang et al., 2011; Dragovich et al., 1999) .

Ebselen, disulfiram, and carmofur were only effective without dithiothreitol, as they modified cysteine in the catalytic triad (Ma et al., 2020) . DWV 3C pro is sensitive to some protease inhibitors of mammalian viruses, suggesting that their catalytic domains and substrate-binding domains share similar structures.

We previously showed that rupintrivir suppresses DWV replication in honey bee pupal cells based on the synthesis of RdRP (Wu et al., 2021) . Thus, we tested the effect of ebselen on DWV replication. Ebselen decreased the synthesis of the 97 kDa precursor of RdRP and 3C pro in a concentration-dependent manner ( Fig. 3A and B ).

However, cell viability was reduced at concentrations of 100 and 250 μM (Fig. 3C ),

suggesting that ebselen affects the functions of host proteins as well.

We used Alphafold2 to predict the structure of DWV 3C pro (polyprotein 2199-2353) (Jumper et al., 2021; Mirdita et al., 2021) . We obtained multiple models that are very similar to each other, and one model with the overall structure is shown in Fig. 4A .

Based on the IDDT values per position, the structure was predicted with high confidence, except for the N-and C-termini as well as single loop (polyprotein 2330-2332) (Supplementary figure 1). DWV 3C pro has a trypsin-like structure consisting of six β-sheets, which are folded into two β-barrel domains packed perpendicularly to each other. The catalytic center appears to be present between the two domains. This is similar to those of other 3C pro s in addition to the position of β-ribbon (polyprotein 2281-2293) which constitutes the substrate-binding domain (Matthews et al., 1994; Bergmann et al., 1997; Mosimann et al., 1997; Birtley et al., 2005; Lee et al., 2009; Ohlenschläger et al., 2004; Cui et al., 2011) (Fig. 4A ). To identify the amino acids critical for protease activity, we aligned 3C pro sequences of DWV and 23 other Iflaviruses (Supplementary figure 2). Among the conserved 12 amino acids, the positions of C2307 and H2170 together with non-conserved N2227 in the DWV 3C pro structure corresponded to those of the known 3C pro catalytic triads (Matthews et al., 1994; Bergmann et al., 1997; Mosimann et al., 1997; Birtley et al., 2005; Lee et al., 2009; Ohlenschläger et al., 2004; Cui et al., 2011) . Although Y2171 was well conserved, its position relative to H2170 did not appear to be compatible with the third residue of the catalytic triad. There is also D2225, but the side chain with a carboxyl group was oriented in the opposite direction ( Fig. 4C To prove that the structure of DWV 3C pro predicted by Alphafold2 is correct, we prepared mutant proteins in which the potentially critical amino acid was substituted with alanine and measured the protease activities. The activities were dramatically reduced with the mutant proteins for the catalytic triad (C2307, H2170, and N2227) and

substrate binding domains, including conserved Y2299 and H2324 domains ( Fig. 4C and F). N2134 of DWV is also well conserved between the 3C pro s of Iflaviruses, but it localizes far from the catalytic center (Fig. 4D ). Nevertheless, we found that the protease activity of the N2134A mutant was reduced to approximately one-third of that of the wild type protein (Fig. 4F) . A previous study suggested that H2190 and D2225

constitute the catalytic triad together with C2307 (Lanzi et al., 2006) . However, H2190

was far from C2307 ( Fig. 4E) , and the H2190A mutant and wild type protein showed comparable protease activity (Fig. 4F ). D2225 is close to N2227 as described above, but the protease activity of D2225A was reduced to only half of that of the wild type protein ( Fig. 4F ). Thus, H2190 and D2225 are not part of the catalytic triad of DWV 3C pro .

D2304A and E2329A mutants have the same activity as the wild type protein, indicating that they are not the third residue of the catalytic triad ( Fig. 4E and F). The above results demonstrate that the Alphafold2 predicted structure of DWV 3C pro should be correct.

We next determined how rupintrivir binds DWV 3C pro to inhibit its activity using a molecular docking tool. Rupintrivir was predicted to interact with 16 amino acids in the catalytic triad (C2307, H2170, and N2227) as well as a substrate-binding domain. Nitrogen atoms in the side chains of N2227 and H2324 appeared to form hydrogen bonds with rupintrivir. Five amino acids (I2283, N2284, A2285, L2288, and Y2289) in β-ribbon also appeared to interact with rupintrivir ( Fig. 5A , B, and E). Thus, 3C pro s of DWV and picornaviruses appear to bind rupintrivir in a similar manner (Wang et al., 2011; Matthews et al., 1994; Binford et al., 2005) . Analysis of rupintrivir binding with DWV 3C pro revealed a precise map of the amino acids critical for protease activity at the protein surface. The catalytic triad (C2307, H2170, and N2227) are adjacent to each other; R2156 and Y2171 are adjacent to C2307 and H2170. The largely reduced protease activities of R2156A and Y2171A mutants were consistent with their localization (Fig. 4F ). The substrate-binding domain is composed of two grooves on the left and right sides, and V2325 and A2326 are in the middle to connect them. The left groove was constructed using H2302, G2303, D2304, G2305, H2324, G2327, and E2329. Among them, H2302, G2303, D2304, and E2329 create a porous structure. The deep right groove is made by R2169, E2173, I2283, N2284, A2285, L2288, Y2289, and V2291. Most of the amino acids were in the β-ribbon structure (Fig. 5 C and D).

When we purified the 56 kDa wild-type GST-DWV 3C pro protein, a 30 kDa small band was co-purified. This band was absent in the mutant protein lacking protease activity, for example, C2307A (Fig. 6A ). These results suggest that the 30 kDa band was generated by the self-cleavage of GST-DWV 3C pro either by cis or trans. Our DWV 3C pro fusion protein contains two potential cleavage sites, Q2118 and E2180 (Lanzi et al., 2006) . To determine the cleavage site, we prepared mutant proteins in which one of the above two amino acids was substituted with alanine. As shown in Figure 6A , the 30 kDa band was absent only in the Q2118A mutant, suggesting that GST-DWV 3C pro is self-cleaved at Q2118 in E. coli. We also tested whether the purified GST-DWV 3C pro was self-cleaved during incubation at 33 °C and found that it remained stable without self-cleavage for 18 h (Fig. 6B ).

To characterize the 3C pro synthesized in DWV-infected honey bee cells, we raised antibodies against DWV 3C pro . We found that a 97 kDa protein specifically present in DWV-infected cells is recognized by both anti-3C pro and anti-RdRP antibodies, confirming that this is a 3C pro precursor with RdRP. There was also a 42 kDa band specifically recognized by the anti-3C pro antibody, suggesting that this corresponds to the mature 3C pro in DWV-infected cells (Fig. 6C) . Thus, DWV 3C pro is present as both a precursor and a mature protein in the infected cells.

Compared to the Km values of the best peptide substrate for 3C pro s of human rhinovirus (250 μM), poliovirus (7 μM), and enterovirus (30 and 43 μM) (Cordingley et al., 1989; Weidner and Dunn, 1991; Shang et al., 2015; Cui et al., 2011) , DWV 3C pro has a higher affinity for the peptide substrate used in this study (Km: 4.233 μM). Nevertheless, DWV polyprotein has other potential cleavage sites by 3C pro so that the efficiency of cleaving a peptide substrate containing these sites would be varied. The Kcat/Km value of DWV 3C pro was 59.96 mM -1 min -1 and comparable to those of hepatitis A virus (126 mM -1 min -1 ), enterovirus (11.8 and 0.71 mM -1 min -1 ), and foot-and-mouth disease virus (59.4 mM -1 min -1 ) 3C pro s with their best substrates (Jewell et al., 1992; Shang et al., 2015; Cui et al., 2011; Sweeney et al., 2007) . These results suggest that the presence of Asn instead of Asp/Glu in the catalytic triad of DWV 3C pro does not dramatically affect protease activity. In fact, substituting Asp in the catalytic triad of hepatitis A virus 3C pro with Asn only resulted in slower processing (Gosert, Dollenmaier and Weitz, 1997) .

Ebselen, disulfiram, and carmofur have been shown to inhibit the main protease of SARS-CoV-2 by covalently modifying cysteine in the catalytic diad (Jin et al., 2020b; Ma et al., 2020; Amporndanai et al., 2021 ). Although we were not able to predict the binding site of ebselen in DWV 3C pro using a molecular docking approach, it is likely to target C2307 in the catalytic triad to inhibit protease activity. Ebselen inhibited DWV replication in honey bee cells; however, it was also toxic to the cells (Fig. 3) . Thus, ebselen (MW: 274.18) could be a potential lead compound for fragment-based drug discovery approaches to increase the specificity and potency to DWV 3C pro . Whether the 3C pro inhibitors we identified could be effective in controlling DWV in honey bees remains to be tested.

In contrast to picornaviruses, DWV 3C pro contains Asn (N2227) as the third residue of the catalytic triad instead of Asp/Glu ( Fig. 4A and B) . Moreover, Asn is not (Wang et al., 2011) . Meanwhile, Cys and His in the catalytic triad are under strong negative selection, suggesting that they are the most critical amino acids for protease reaction. This is consistent with the fact that the main protease of coronavirus has the Cys-His catalytic diad (Jin et al., 2020a) . Because the N2227A mutant completely lost its protease activity (Fig. 4F) , and the same was observed with other 3C pro s (Cheah, Leong and Porter, 1990; Hämmerle, Hellen and Wimmer, 1991; Miyashita et al., 1993) , the catalytic triad seems to be necessary for Picornavirales. In many Iflaviruses, the Tyr residue next to the His residue of the catalytic triad (Y2171 for DWV) is conserved and is substituted with Phe in the Dicistrovirus, CrPV. Thus, Tyr appears to be important to position the imidazole side chain of His close to the side chain of Cys by stacking interactions.

The pores made by H2302, G2303, D2304, and E2329 in the left groove of the substrate-binding domain did not appear to be essential because the protease activities of the D2304A and E2329A mutants were not affected (Fig. 4F) . β-ribbon is the major component of the deep right groove of the substrate-binding domain. In picornavirus 3C pro s, the β-ribbon adopts an open conformation to increase the substrate accessibility to its binding domain, and the interaction between the β-ribbon and the N-terminal end of the substrate stabilizes the closed conformation to form an ES complex (Matthews et al., 1994; Bergmann et al., 1997; Mosimann et al., 1997; Birtley et al., 2005; Lee et al., 2009; Ohlenschläger et al., 2004; Cui et al., 2011) . Comparing the surface views of the substrate-binding domains of DWV, BrBV, LsPVI2, SeIV-1, SBV, and CrPV 3C pro ( Fig.   5C and D; Supplementary figure 4) , the overall conformation of the catalytic triad and β-ribbon was similar, but the shapes of substrate binding domains were quite different.

Thus, the structure of the substrate-binding domain of each 3C pro is shaped by a preferred cleavage site with a different amino acid sequence.

We found that GST-DWV 3C pro was self-cleaved at Q2118. However, this seems to only occur during the synthesis in E. coli because the mature folded protein was not self-cleaved at 33 °C for 18 h (Fig. 6B) . These results suggest that only the fraction of unfolded GST-DWV 3C pro is cleaved in trans at 15 °C. It is likely that DWV polyprotein is cleaved at Q2118 to release the 97 kDa precursor of 3C pro and RdRP.

However, we need to assume that the 97 kDa precursor is active as a protease similar to picornavirus 3CD precursor (Winston and Boehr, 2021) and that cleavage should occur in cis, at least during the early stage of polyprotein synthesis until a sufficient 97 kDa precursor and 42 kDa mature 3C pro s accumulate in DWV-infected cells. Because the 97 kDa precursor remains abundant in the virus-infected honey bee cells (Fig. 6C) , processing between 3C pro and RdRP appears to be slower than the cleavage at Q2118.

These characteristics appear to be shared among picornaviruses as well (Jiang et al., 2014) . It will be interesting to test whether the 97 kDa precursor and 42 kDa mature 3C pro s have different protease activities and functions for viral RNA replication in future studies.

The 3C protease cDNA corresponding to amino acid 2094-2353 of DWV polyprotein was amplified by PCR using two primers, GST-3C pro -5-BamHI and GST-3C pro -3-NotI (Supplementary Table 1 ). The amplified PCR product was digested by BamHI (NEB) and NotI (NEB), and then subcloned to pGEX-4T-1vector (Cytiva) followed by transformation to BL21. The transformed BL21 was grown in 1 L of LB medium containing 1 % glucose and 0.1 mg/mL Ampillicin at 37 ℃ until A600 reached to 0.5.

The cell suspension was cooled down, and then IPTG was added at 0.1 mM to induce the protein expression at 15 ℃ for 16 h. E. coli was collected by centrifugation and resuspended in 100 mL of ice-cold TNED buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 2 mM EDTA, 1 mM DTT) with 0.5 % TX-100 and protease inhibitor cocktail (Beyotime). The cell lysate was prepared by sonication using Q700 Sonicator (Qsonica) at amplitude 100 on ice for 45 min (30 sec pulse with 3 min-interval). 1 mL of BeyoGold™ GST-tag Purification Resin (Beyotime) was added to the supernatant collected after centrifugation. After gently rotating at 4 ℃ for 2 h, the resin was washed five times with 10 mL TNED buffer. The bound protein was eluted with 1. 5 mL TNED buffer containing 10 mM reduced glutathione. The two-thirds of eluted protein was dialyzed against 2L of TNED buffer twice at 4 ℃ for 24 h. The rest of one-third was dialyzed against 2L of TNE buffer without DTT. We measured the protein concentation of purified protein using Enhanced BCA Protein Assay Kit (Beyotime).

GST-3C pro mutant proteins were generated by fusion PCR of two PCR products amplified using pGEX 5' primer as well as the reverse primer for each mutant and pGEX 3' primer as well as the forward primer for each mutant (Supplementary Table 1 ).

The plasmid DNA to express GST-DWV 3C pro was used as the template for the 1st PCR.

The fusion PCR products were cloned in pGEX-4T-1vector and the mutant proteins were expressed and purified as above. All plasmid DNAs were sequenced to verify the intended mutations.

Dabcyl-PVQAKPEMDNPNPGE-Edans derived from the cleavage site between L-protein and VP2 (underlined, 205-218 of polyprotein) was used as a peptide substrate to measure the protease activity. E was added to the C-terminus to link Edans. FRET experiments were performed with Varioskan™ LUX multimode microplate reader (Thermo Fisher). We measured 100 μL of reaction mixture with 50 mM Citrate buffer, To identify the inhibitors for DWV 3C pro , various concentrations of the tested compounds were added. To determine the inhibitory effects of the various compounds, we preincubated the compound and enzyme on ice for 10 min prior to addition of the peptide substrate. To analyze the effects of ebselen, disulfiram, and carmofur, we used the enzyme dialyzed against TNE buffer and carried out the reaction with or without DTT. We determined the initial velocities of the enzymatic reactions and fitted to a sigmoidal dose-response equation with nonlinear regression analysis using GraphPad Prism 9. The data from three independent assays were used to determine IC50 value of each compound.

Honey bee pupae with pale/pink eyes were collected from a mite-free colony. They were surface sterilized by washing with 10 % bleach followed by sterile PBS three times (5 min for each wash). Heads from the pupae were dissected and homogenized seven times with 1 mL Grace culture medium using Dounce homogenizer (Loose fitting). The homogenate was then filtered through a cell strainer (Falcon) and the number of cells was counted. 10 6 cells were suspended in 100 μL of Grace medium containing 10 % FBS, antibiotics (penicillin and streptomycin), and ebselen at the indicated concentration with or without DWV (MOI: 10) in 24-well plate at 33 ℃ for 1 h. Fresh culture medium (400 μL) with ebselen at the indicated concentration was then added (the final cell density at 2 × 10 6 /mL) and incubated at 33 ℃ for 16 h.

After DWV infection, the pupal head cells were collected by centrifugation, and then homogenized with 150 μL of RIPA buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 % NP-40, 0.5 % sodium deoxycholate, 0.1 % SDS) containing protease inhibitor cocktail. After centrifugation of the homogenate, the protein concentration of supernatant was measured using Enhanced BCA protein assay kit. The cell lysate with 20 μg of protein was analyzed by western blot for each sample. The protein samples in SDS-PAGE sample buffer (2 % SDS, 10 % glycerol, 10 % β-mercaptoethanol, 0.25 % bromophenol blue, 50 mM Tris-HCl, pH 6.8) were heated at 95 ℃ for 5 min. After centrifugation, the supernatants were applied to 10 % SDS-PAGE and the proteins were transferred to a nitrocellulose membrane (Pall ® Life Sciences). Another 10 % SDS-PAGE gel with the same samples was stained by Instant Blue. The membrane was then blocked with PBST (PBS with 0.1 % Tween-20) containing 5 % BSA at room temperature for 30 min followed by incubating with 1000-fold diluted anti-RdRP antibody (Wu et al., 2020) or 500-fold diluted anti-3C pro antibody at 4 ℃ overnight.

The membrane was washed five times with PBST (5 min each), and then incubated with 10,000-fold diluted IRDye® 680RD donkey anti-rabbit IgG (H+L) (LI-COR Biosciences) in PBST containing 5 % skim milk at room temperature for 2 h. The membrane was washed as above, and then visualized using Odyssey Imaging System (LI-COR Biosciences). Band intensity of 97 kDa RdRP precursor was measured by image-J.

Honey bee pupal head cells were prepared as above and 100 μL cell suspension (2 × 10 6 /mL) was inoculated to each well in 96-well plate. The cells were cultured in the presence of either DMSO or ebselen at the indicated concentrations at 33 ℃ for 16 h.

The cultured plate was first incubated at room temperature for 10 min followed by adding 100 μL of reagent solution (CellTiter-Lumi™ Luminescent Cell Viability Assay Kit, Beyotime). The plate was shaken at room temperature for 2 min to promote cell lysis, and then further incubated for 10 min to stabilize the chemiluminescence signal. The signal was detected using Varioskan™ LUX multimode microplate reader and depends on intracellular ATP level, thus the relative viability of cells.

We determined the structures of 3C proteases of DWV, BrBV, LsPVI2, SeIV-1, SBV, and CrPV using ColabFold (Mirdita et al., 2021) based on AlphaFold v2.1.0 (Jumper et al., 2021) . The generated structures were analyzed and displayed using UCSF ChimeraX (Pettersen et al., 2021; Goddard et al., 2018) . The molecular structure of rupintrivir was collected from PubChem (CID: 6440352). We used AutoDock Tools package (Morris et al., 2009) to prepare the ligand and receptor by removing water and other heteroatoms and assigning partial charges. The size of the grid box for docking was set at 20 Å in each direction. We provided explicit coordinate to find all possible binding sites. After preparing the ligand, receptor, definition of binding site, AutoDock Vina (version 1.1.2) (Trott and Olson, 2010) program was used for the molecular docking simulation.

10 μg of wild type, C2307A, Q2118A, and E2180A GST-DWV 3C pro protein was analyzed by 10 % SDS-PAGE followed by Instant Blue staining. 20 μg of wild type GST-DWV 3C pro was incubated in the enzyme reaction buffer without the substrate at 33 ℃ and the 4 μg protein was sampled every 3 h for 12 h and at 18 h. All samples including the one without the incubation (0 h) were analyzed by SDS-PAGE as above.

DWV 3C pro insert in pGEX-4T-1 was transferred to pGEX-6P-3 followed by transformation to BL21. The protein was expressed and purified as above except 5 mL TNED buffer containing 100 units of PreScission Protease (Beyotime) was added to the resin after the final wash. The resin was incubated at 4 ℃ overnight and the supernatant with the released DWV 3C pro was collected. The remaining resin was eluted once with 5 mL TNED buffer and the eluate was combined with the above supernatant. The total eluate was dialyzed against 2L of PBS with 1 mM EDTA, 0.5 mM DTT, and 0.1 % sarcosyl twice at 4 ℃ for 24 h. The purified protein was delivered to GeneScript-Nanjing to raise the anti-rabbit polyclonal antibody.

3C protease sequences of various Iflaviruses were searched in NCBI by BLASTP using DWV 3C pro sequence as a query. Twenty three sequences were picked up and aligned together with DWV 3C pro using MUSCLE program (EMBL-EBI).

All data presented were from representative independent experiments. Statistical analyses were performed with Bell Curve for Excel (Social Survey Research Information Co., Ltd.) and no data point was excluded. The applied statistical tests and P-values are described in figure legends.

The authors declare no conflict of interest.

TK conceived and designed research strategy and wrote the paper. XY performed the experiments. XY and TK analyzed data. Cricket paralysis virus (E). The helix, strand, and coil structures are indicated by orchid, pale green, and wheat colors, respectively. β-ribbons are colored by cyan. The catalytic triad with amino acid residues is indicated by a square with a white dotted line.

Inhibition mechanism of SARS-CoV-2 main protease by ebselen and its derivatives

The refined crystal structure of the 3C gene product from hepatitis A virus: specific proteinase activity and RNA recognition

Conservation of amino acids in human rhinovirus 3C protease correlates with broad-spectrum antiviral activity of rupintrivir, a novel human rhinovirus 3C protease inhibitor

Crystal structure of foot-and-mouth disease virus 3C protease. New insights into catalytic mechanism and cleavage specificity

Tropilaelaps mite: an emerging threat to European honey bee

Site-directed mutagenesis suggests close functional relationship between a human rhinovirus 3C cysteine protease and cellular trypsin-like serine proteases

Cleavage of small peptides in vitro by human rhinovirus 14 3C protease expressed in Escherichia coli

Crystal structure of human enterovirus 71 3C protease

Deformed wing virus

Structure-based design, synthesis, and biological evaluation of irreversible human rhinovirus 3C protease inhibitors. 4. Incorporation of P1 lactam moieties as L-glutamine replacements

Deformed wing virus associated with Tropilaelaps mercedesae infesting European honey bees (Apis mellifera

UCSF ChimeraX: Meeting modern challenges in visualization and analysis

3C-like proteases of picornaviruses, noroviruses, and coronaviruses

Inhibitors of 3C cysteine proteinases from Picornaviridae

Molecular and biological characterization of deformed wing virus of honeybees (Apis mellifera L.)

Structural basis of inhibition specificities of 3C and 3C-like proteases by zinc-coordinating and peptidomimetic compounds

Ebselen, Disulfiram, Carmofur, PX-12, Tideglusib, and Shikonin Are Nonspecific Promiscuous SARS-CoV-2 Main Protease Inhibitors

Structure of human rhinovirus 3C protease reveals a trypsin-like polypeptide fold, RNA-binding site, and means for cleaving precursor polyprotein

ColabFold -Making protein folding accessible to all

Site-directed mutagenesis of the putative active site residues of 3C proteinase of coxsackievirus B3: evidence of a functional relationship with trypsin-like serine proteinases

AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility

Refined X-ray crystallographic structure of the poliovirus 3C gene product

Ecology of Varroa destructor, the Major Ectoparasite of the Western Honey Bee, Apis mellifera

The structure of the stemloop D subdomain of coxsackievirus B3 cloverleaf RNA and its interaction with the proteinase 3C

In vitro antiviral activity of AG7088, a potent inhibitor of human rhinovirus 3C protease

UCSF ChimeraX: Structure visualization for researchers, educators, and developers

Biology and control of Varroa destructor

Biochemical characterization of recombinant Enterovirus 71 3C protease with fluorogenic model peptide substrates and development of a biochemical assay

The role of varroa mites in infections of Kashmir bee virus (KBV) and deformed wing virus (DWV) in honey bees

Roles of the Picornaviral 3C

Proteinase in the Viral Life Cycle and Host Cells', Viruses

Structural and mutagenic analysis of foot-and-mouth disease virus 3C protease reveals the role of the beta-ribbon in proteolysis

Antiviral activities of peptide-based covalent inhibitors of the Enterovirus 71 3C protease

Prevalence and seasonal variations of six bee viruses in Apis mellifera L. and Varroa destructor mite populations in France

AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading

Crystal structures of enterovirus 71 3C protease complexed with rupintrivir reveal the roles of catalytically important residues

Development of synthetic peptide substrates for the poliovirus 3C proteinase

Deformed wing virus is a recent global epidemic in honeybees driven by Varroa mites

Cysteine Protease Zymography: Brief Review'

The Picornavirus Precursor 3CD Has Different Conformational Dynamics Compared to 3C

Characterization of the Copy Number and Variants of Deformed Wing Virus (DWV) in the Pairs of Honey Bee Pupa and Infesting

Honey Bee Suppresses the Parasitic Mite Vitellogenin by Antimicrobial Peptide

DWV Infection

Inhibition of enterovirus 71 replication and viral 3C protease by quercetin

Picornavirus 3C -a protease ensuring virus replication and subverting host responses

Vertical-transmission routes for deformed wing virus of honeybees (Apis mellifera)

We thank Yixiang Tang, Shiqi Xu, Jiarui Li, Zhuoran Leng, and Shangqing Li for their contributions to conduct the experiments. We are grateful to Dr. David Ruiz-Carrillo for his comments on the paper.