Characterization of Proteinase Cleavage Sites in the N-Terminal

CORE Metadata, citation and similar papers at core.ac.uk

Provided by Elsevier - Publisher Connector

Available online at www.sciencedirect.com

Virology 375 (2008) 159–169 www.elsevier.com/locate/yviro

Characterization of proteinase cleavage sites in the N-terminal region of the RNA1-encoded polyprotein from Arabis mosaic virus (subgroup A nepovirus) ⁎ Thierry Wetzel a, Joan Chisholm b, Alexandra Bassler a, Hélène Sanfaçon b,

a RLP Agroscience, AlPlanta Institute for Plant Research, Breitenweg 71, 67435, Neustadt an der Weinstrasse, Germany b Pacific Agri-Food Research Centre, Agriculture and Agri-Food Canada, P.O. Box 5000, 4200 Highway 97, Summerland, B.C., Canada V0H 1Z0 Received 19 November 2007; returned to author for revision 20 December 2007; accepted 30 January 2008 Available online 4 March 2008

Abstract

Arabis mosaic virus is a subgroup A nepovirus. The RNA1-encoded polyprotein (P1) contains the domains for the NTP-binding protein (NTB), VPg, proteinase (Pro) and polymerase at its C-terminus. Putative cleavage sites delineating these domains have been proposed. However, the number and location of cleavage sites upstream of the NTB domain are not known. Using in vitro processing assays, we have confirmed proteolytic cleavage at the NTB-VPg and VPg-Pro sites. In addition, we have identified two cleavage sites in the N-terminal region of P1. Site- directed mutagenesis and immunoprecipitation experiments using inserted peptide tags confirmed that the position of these cleavage sites corresponds to that of cleavage sites delineating the X1 and X2 domains in Tomato ringspot virus (subgroup C nepovirus). Amino acid alignments implied the presence of similar cleavage sites in the P1 polyprotein of other nepoviruses. Our results suggest that the presence of two protein domains upstream of NTB is a common feature of nepoviruses. © 2008 Elsevier Inc. All rights reserved.

Keywords: Positive-strand RNA virus; Picorna-like virus; Nepovirus; Comoviridae; Protease; Cleavage site; Genomic organization

Introduction Each nepovirus genomic RNA encodes one large polyprotein that is cleaved into mature proteins and intermediate precursors by The genus Nepovirus is a group of plant-infecting picorna- the viral proteinase (Pro). Nepovirus proteinases are related to the like viruses. They have a bipartite genome and share sequence 3C Pro of picornaviruses (Gorbalenya et al., 1989b). A conserved identity, common genome expression strategies and biological histidine in the substrate-binding pocket of the 3C Pro of properties (Le Gall et al., 2005, 2007). Nepoviruses belong to the picornaviruses is responsible for the specific recognition of a family Comoviridae along with members of the genera Como- glutamine or glutamate at the −1 position of the cleavage sites virus and Fabavirus. Among picorna-like viruses, nepoviruses (Gorbalenya et al., 1989a; Seipelt et al., 1999). This histidine is have the unique property of encapsidating their genome using a present in the substrate-binding pocket of proteinases of subgroup single large coat protein (CP) subunit (Le Gall et al., 2005, C nepoviruses and polyprotein cleavage was shown to occur after 2007). In spite of common characteristics, nepoviruses are a glutamine, asparginine or aspartate (Bacher et al., 1994; Carrier diverse and the genus has been divided into three subgroups et al., 1999; Latvala et al., 1998). The proteinases of subgroup A based on the length of RNA2, immunological properties of the and B nepoviruses have a leucine instead of a histidine in their virions, sequence identities in the CP and cleavage site speci- substrate-binding pockets (Margis and Pinck, 1992). Subgroup B ficity of the viral proteinase (Le Gall et al., 2005; Mayo and nepovirus cleavage sites have a lysine or arginine at the −1 Robinson, 1996). position (Demangeat et al., 1991; Hemmer et al., 1995). Subgroup A nepovirus proteinases cleave after a cysteine, arginine or ⁎ Corresponding author. Fax: +1 250 494 0755. glycine (Blok et al., 1992; Buckley et al., 1993; Margis et al., E-mail address: [email protected] (H. Sanfaçon). 1993; Pinck et al., 1991; Serghini et al., 1990).

Nepoviruses of different subgroups have similar but distinct terized. In the case of ToRSV, in vitro processing studies have genomic organizations. The RNA2-encoded polyprotein (P2) of identified two cleavage sites upstream of the 66 kDa NTB all nepoviruses contains the domains for the movement protein domain (Wang and Sanfacon, 2000). These cleavage sites define (MP) and CP at its C-terminus. However, while the N-terminal the 45 kDa X1 protein and the 22 kDa X2 protein. The ToRSV region of subgroup A and B nepovirus P2 polyproteins contains X2 protein shares many similarities with the comovirus 32 kDa a single protein domain upstream of the MP (Demangeat et al., protein. Both proteins are highly hydrophobic and possess a 1991; Margis et al., 1993), the P2 polyprotein of Tomato conserved amino acid motif [F-x(28)-W-x(11)-L-x(23)-E] ringspot virus (ToRSV, a subgroup C nepovirus), contains two (Zhang and Sanfacon, 2006). The ToRSV X1 protein does not protein domains in addition to the MP and CP (Carrier et al., have a functional equivalent in the comovirus genome. Amino 2001). The RNA1-encoded polyprotein (P1) of all nepoviruses acid sequences highly similar to the ToRSV X2 protein and includes in its C-terminal region the domains for the putative comovirus 32 kDa protein are also found in the N-terminal nucleotide triphosphate-binding protein (NTB), the small viral region of the P1 polyproteins of subgroup A and B nepoviruses protein linked to the genome (VPg), the Pro and the RNA- (Zhang and Sanfacon, 2006) (see Fig. 1). A cleavage site was dependent RNA polymerase (Pol) (see Fig. 1). Furthermore, the identified upstream of the conserved NTB motif in the P1 poly- arrangement of these protein domains within the polyprotein is proteins from Beet ringspot virus (BRSV, a subgroup B virus conserved in all picorna-like viruses (Le Gall et al., 2007). The previously referred to as Tomato black ring virus)(Demangeat N-terminal region of nepovirus P1 polyproteins is less charac- et al., 1990; Le Gall et al., 2005) and Grapevine fanleaf virus (GFLV, a subgroup A virus) (Margis et al., 1994). However, the position of these cleavages relative to the conserved NTB and X2 motifs differs (Fig. 1). Cleavage in the BRSV P1 polyprotein was detected between the X2 and NTB motifs at a position corresponding to the ToRSV X2-NTB site (Demangeat et al., 1990). In the case of GFLV, the observed cleavage was upstream of the X2 motif, at a position corresponding to the ToRSV X1– X2 site (Margis et al., 1994). The possibility that additional cleavage sites exist in the N-terminal region of subgroup A and B nepovirus P1 polyproteins has not been rigorously examined. In this study, we have examined the proteolytic processing of the P1 polyprotein of Arabis mosaic virus (ArMV), a subgroup A nepovirus closely related to GFLV (Wetzel et al., 2004, 2001). Using in vitro processing assays, we have identified two cleavage sites upstream of the NTB domain at positions equivalent to the ToRSV X1–X2 and X2-NTB cleavage sites. Amino acid sequence alignments also implied similar cleavage sites in the polyproteins of other nepoviruses, suggesting that the presence of two protein domains upstream of NTB is a common feature of nepoviruses.

Results Fig. 1. Cleavage sites in the P1 polyproteins of como- and nepoviruses. The P1 polyproteins of Cowpea mosaic virus (CPMV, a comovirus), Tomato ringspot virus (ToRSV, a subgroup C nepovirus), Beet ringspot virus (BRSV, a subgroup Detection of processing events at the NTB-VPg and VPg-Pro B nepovirus), Grapevine fanleaf virus and Arabis mosaic virus (GFLV and cleavage sites ArMV, two subgroup A nepoviruses) are represented by the boxes. The grey triangles, black diamonds and open ovals represent conserved amino acid motifs To study processing events in the ArMV P1 polyprotein, we in the RNA-dependent RNA polymerase domain (Pol), 3C-like proteinase (Pro) and NTP-binding protein (NTB) domains, respectively (Gorbalenya et al., used previously described in vitro processing assays (Wang 1989a, 1990; Koonin and Dolja, 1993). The asterisks indicate an amino acid et al., 1999; Wang and Sanfacon, 2000). We first tested a small motif conserved in the putative X2 domain of nepoviruses and analogous to the truncated P1 polyprotein to confirm processing at the NTB-VPg proteinase co-factor motif of comoviruses (Ritzenthaler et al., 1991; Rott et al., and VPg-Pro sites. The precursor was designed to include the 1995). Thick vertical lines represent cleavage sites previously identified by putative 66 kDa NTB domain, as well as the VPg and Pro in vitro processing assays. Dipeptides, proposed or shown to be recognized by the proteinase, are indicated above each identified cleavage site. Dipeptides domains (NTB-Pro precursor, see Materials and methods and shown in parenthesis were suggested based on analysis of the deduced amino Fig. 2A). As a negative control, a mutation was introduced in the acid sequence but have not been tested experimentally (Demangeat et al., 1990; proteinase domain of the NTB-Pro precursor to replace the Margis et al., 1994). Other dipeptides were either identified by direct amino acid conserved histidine (H1283) of the catalytic triad for a serine (HS sequencing of the cleaved product or supported by in vitro mutagenesis of the mutation, numbering from the N-terminus of the deduced amino predicted dipeptide (Hemmer et al., 1995; Pinck et al., 1991; Wang et al., 1999; Wang and Sanfacon, 2000; Wellink et al., 1986). In the case of ArMV, the acid sequence for the P1 polyprotein according to Wetzel et al., suggested location of putative cleavage sites is shown by the vertical dashed 2004). Previous studies with the GFLV and ToRSV proteinases lines (Wetzel et al., 2004 and this study). have shown that mutation of this conserved histidine eliminates T. Wetzel et al. / Virology 375 (2008) 159–169 161 their proteolytic activity (Hans and Sanfacon, 1995; Margis and included the full-length polyprotein precursor. Several back- Pinck, 1992). ground proteins were also concentrated in this fraction sug- After in vitro translation, the NTB-ProHS mutant polypro- gesting that they included the S-tag and were probably produced tein appeared as a protein of approximately 95 kDa in SDS- by premature termination of translation. The 69 kDa and 66 kDa polyacrylamide gels (SDS-PAGE, see Fig. 2B, lane 1). The cleavage products were enriched in the pull-down fraction while apparent molecular mass of this protein is close to that expected the 32 kDa and 31 kDa cleavage products remained in the for the full-length NTB-Pro polyprotein (100 kDa including an unbound fraction (Fig. 2B, lanes 5 and 8). Because the S-tag S-tag at the N-terminus and a His-tag at the C-terminus). Minor was fused to the N-terminus of the NTB domain, this result proteins of lower apparent molecular masses were also confirmed that the 69 kDa and 66 kDa proteins contain this observed. These proteins were likely produced by translation domain and provided experimental support for the suggestion initiation at an internal AUG or premature termination of the that they correspond to the NTB-VPg intermediate and to the translation. The unprocessed wild-type (WT) NTB-Pro premature NTB protein, respectively. cursor was also 100 kDa, but after overnight incubation at The above results suggested that the ArMV proteinase 16 °C, additional proteins appeared with apparent molecular recognized both the NTB-VPg and the VPg-Pro cleavage sites. masses of 69 kDa, 66 kDa, 32 kDa and 31 kDa that likely By analogy with characterized GFLV cleavage sites, putative corresponded to the NTB-VPg, NTB, VPg-Pro and Pro cleavage sites between the NTB, VPg and Pro domains were cleavage products, respectively (Fig. 2B, lane 2 and summary previously proposed to correspond to dipeptide C1216/S (for table in C). These proteins were not produced after overnight NTB-VPg) and either G1240/D or R1239/G (for VPg-Pro) (Wetzel incubation of the NTB-ProHS mutant polyprotein (Fig. 2B, et al., 2004). After processing of the WT NTB-Pro precursor, the lane 1). To verify the nature of the cleavage products, we intensity of the bands corresponding to the 66 kDa (NTB) and took advantage of the S-tag, which was fused in frame with the 32 kDa (putative VPg-Pro) cleavage products was two to three N-terminus of the NTB-Pro polyprotein, and conducted pull- times stronger than that of bands corresponding to the 69 kDa down assays (Fig. 2B). As expected, the pull-down fraction (NTB-VPg) and 31 kDa (putative Pro) products as judged by quantification of the radioactivity contained in each band, taking into account the number of radiolabeled methionine incorporated in each protein (Fig. 2C and Materials and methods). This suggested that cleavage occurred predominantly between the NTB and VPg domains. To verify this, we deleted the C/S

Fig. 2. In vitro processing of the NTB-Pro polyprotein and identification of the NTB-VPg cleavage site. (A) Schematic representation of the processing of the ArMV NTB-Pro polyprotein. The polyprotein is represented at the top. An S-tag epitope (grey box) and a histidine tail (black circle) fused in frame with the N- and C-termini of the polyprotein are shown. Vertical lines represent cleavage sites present in the precursor with the predicted C/S dipeptide indicated above the NTB- VPg cleavage site. Cleavage products observed after processing of the polyprotein are shown below the arrow. The expected molecular masses of the precursor polyprotein and cleavage products are shown. These molecular masses were calculated taking into account the non-viral sequences fused in frame with the proteins. (B) In vitro processing assays. The wild-type (WT) NTB-Pro protein and mutant derivatives NTB-ProHS (mutation in the catalytic triad of the proteinase) and NTB-ProΔCS (deletion of the C/S dipeptide at the junction between the NTB and VPg domains) were synthesized in vitro and incubated at 16 °C overnight to allow for proteolytic processing of the polyprotein. After processing, S-tag pull-down assays were conducted as described in Materials and methods and proteins present in the starting reactions (in vitro processing), the immunoprecipitated fraction (S-tag pull-down) or the supernatant after immunoprecipitation (unbound) were separated by SDS-PAGE as indicated above each lane. The precursor polyprotein and cleavage products were separated by SDS-PAGE (12% polyacrylamide). Bands corresponding to cleavage products are identified by the symbols on the left side of each lane. The migration of molecular mass markers (in kDa) is indicated at the left of the gel. (C) Summary of the results and deduced identity of the cleavage products. Presence (+) or absence (−)ofeachcleavage product in the wild-type (WT) or mutated (ΔCS) NTB-Pro processing reactions is indicated. The ability of each cleavage product to be immunoprecipitated by the

S-tag antibody (S-Tag column) is also shown. The number of methionines (Nmet) present in each cleavage product is indicated. Relative intensity of each cleavage product present in the processing products of the wild-type polyprotein was estimated by calculating the number of digital light units (DLU) present in each band, corrected for the number of methionines (see Materials and methods) and expressed as follows: 7–10 (+++), 3–4 (++), 0.5–1 (+), and less than 0.05 (+/−). 162 T. Wetzel et al. / Virology 375 (2008) 159–169 dipeptide at the predicted NTB-VPg cleavage site. In vitro P1 polyprotein at a position immediately upstream of the puta- processing of the NTB-ProΔCS mutant resulted in the release of tive X2 domain, cleavage at the junction between the putative the 69 kDa and 31 kDa cleavage products (Fig. 2B, lane 3). The X2 and NTB domains had not been tested (Margis et al., 1994). 66 kDa and 32 kDa products were not detected. This result As above, the H1283 to S mutation was introduced in the suggested that cleavage at the junction between the NTB and proteinase domain of the X2-Pro polyprotein as a negative VPg domains occurs at or near the predicted C1216/S dipeptide. control. The X2-ProHS full-length polyprotein migrated as a However, our experiments do not exclude the possibility that protein with an apparent molecular mass of 115 kDa in SDS- deletion of dipeptide C1216/S induces conformational changes in PAGE (expected molecular mass is 120 kDa, Fig. 3B, lane 1). the polyprotein that could affect cleavage at a different dipeptide. As with the NTB-ProHS polyprotein, several background bands were also observed in addition to the full-length polyprotein. Identification of a sub-optimal cleavage event at the junction After overnight processing of the WT X2-Pro precursor at between the X2 and NTB domains 16 °C, several additional bands appeared that were not present after incubation of the X2-ProHS mutant polyprotein under the To identify cleavage events upstream of the NTB domain, we same conditions (compare lane 2 to lane 1). Two major proteins designed additional truncated P1 polyproteins of increasing of approximately 85 kDa and 32 kDa and three minor proteins length. The X2-Pro precursor was designed to include sequences of approximately 95 kDa, 62 kDa and 23 kDa were observed. corresponding to the ToRSV X2 and NTB domains (Fig. 3A). The 95 kDa and 66 kDa proteins were difficult to resolve due to Although a cleavage site was previously identified in the GFLV the presence of many background bands of similar size but were more easily visualized when the cleavage products were separated by SDS-PAGE using a lower concentration of acrylamide (8%, Fig. 3B lower panel). Based on their apparent molecular masses, we speculated that the predominant 85 kDa and 32 kDa proteins corresponded to the X2-NTB and VPg-Pro cleavage products (see Fig. 2A for the calculated molecular mass of each cleavage product). The 95 kDa, 62 kDa and 23 kDa proteins may correspond to the NTB-Pro intermediate, the mature NTB protein and the mature X2 protein, respectively. S-tag pull-down experiments resulted in the concentration of the 23 kDa and 85 kDa cleavage products in the pull-down fraction while the 95 kDa, 66 kDa and 32 kDa cleavage products remained in the unbound fraction (Fig. 2B, lanes 7 and 12). Because the S-tag is fused to the N-terminus of the X2 domain, this result provided experimental support for the suggestion that the 85 kDa and 23 kDa proteins correspond to the X2-NTB intermediate and to the mature X2, respectively.

Fig. 3. In vitro processing of the X2-Pro polyprotein and identification of the X2-NTB cleavage site. (A) Schematic representation of the processing of the ArMV X2-Pro polyprotein. The polyprotein is shown at the top. As in Fig. 2,an S-tag epitope (grey box) and a histidine tail (black circle) are indicated. Vertical lines represent cleavage sites. The C/S dipeptide predicted as the NTB-VPg cleavage site is shown. Two possible cleavage sites (dipeptides G/Vand R/V) are indicated at the junction between the X2 and NTB domains. Cleavage products observed after processing of the polyprotein are represented below the arrow. The calculated molecular masses of the precursor polyprotein and cleavage products are shown. (B) In vitro processing assays. The wild-type (WT) X2-Pro protein and mutant derivatives X2-ProHS (mutation in the catalytic triad of the proteinase), X2-ProΔCS (deletion of the C/S dipeptide), X2-ProΔRV and X2- ProΔGV (deletion of the R/V or G/V dipeptides at the junction between the X2 and NTB domains) were tested using in vitro processing assays as described in the legend of Fig. 2. After processing, S-tag pull-down assays were conducted as described in Materials and methods and proteins present in the starting reactions (in vitro processing), the immunoprecipitated fraction (S-tag pull-down) or the supernatant after immunoprecipitation (unbound) were separated by SDS-PAGE as indicated above each lane. Smaller cleavage products were visualized after separation on a 12% polyacrylamide gel (top panel). Larger intermediate polyproteins were also separated on an 8% polyacrylamide gel (bottom panel). Bands corresponding to cleavage products are identified by the symbols on the left side of each lane. The migration of molecular mass markers (in kDa) is indicated at the left of the gel. (C) Summary of the results and deduced identity of the cleavage products are as explained in the legend of Fig. 2. T. Wetzel et al. / Virology 375 (2008) 159–169 163

To further confirm the identity of the cleavage products, we Detection of the X2, putative NTB and putative NTB-Pro conducted site-directed mutagenesis of predicted cleavage sites cleavage products (23 kDa, 62 kDa and 95 kDa proteins, respec- as above. Deletion of the C1216/S dipeptide between the NTB tively) suggested that a cleavage occurred between the predicted and VPg domains prevented the release of the 85 kDa and X2 and NTB domains. Examination of the deduced amino acid 32 kDa products (see X2-ProΔCS mutant, Fig. 3B, lane 3, and sequence did not reveal the presence of dipeptides previously C). This result confirmed that the 85 kDa and 32 kDa proteins identified or suggested for ArMVor other subgroup A nepovirus correspond to the X2-NTB and VPg-Pro cleavage products after cleavage sites (R/G, C/A, C/S, G/E). However, two alternative cleavage at the NTB-VPg site. The 62 kDa cleavage product dipeptides were identified that may function as possible cleavage was also not observed in the cleavage products of the mutated sites (G604/V and R618/V). Deletion of R618/V did not alter the X2-ProΔCS polyprotein (this band appears as a faint band below processing of the X2-Pro precursor (Fig. 3B, lanes 4 and 9). In one of the background proteins; see Fig. 3B, lower panel). This contrast, deletion of G604/V prevented the release of the 23 kDa result was in agreement with the suggestion that this protein cleavage product (X2, lanes 5 and 10). The 62 kDa (putative NTB) corresponds to the mature NTB domain, which could not be and 95 kDa (putative NTB-VPg-Pro) cleavage products were not released after mutation of the NTB-VPg cleavage site. detected, although it should be noted that the presence of many background bands obscured somewhat the resolution of these proteins. As expected, release of the 85 kDa (X2-NTB) and 32 kDa (putative VPg-Pro) proteins was not affected by this mutation. This result confirms that a proteolytic cleavage occurs between the X2 and NTB domains. The amount of the 23 kDa (X2) and 62 kDa (putative NTB) proteins in the processing products was ten times less than that of the 85 kDa (X2-NTB) and 32 kDa (putative VPg-Pro) proteins as determined by quantification of the radioactivity in each band, see Fig. 2C. This suggested that cleavage at the X2-NTB site is inefficient. Based on the result from the site-directed mutagenesis experiments, we tentatively propose dipeptide G604/V as the X2-NTB cleavage site.

Identification of a cleavage site between the X1 and X2 domains

A cleavage site was previously identified immediately upstream of the putative X2 domain in the N-terminal region of the GFLV P1 polyprotein (Margis et al., 1994). Dipeptide C416/Awas proposed as a putative cleavage site (Margis et al., 1994). In ArMV, the corresponding dipeptide C414/G was suggested as a putative cleavage site (Wetzel et al., 2004). To determine whether the ArMV proteinase could recognize this proposed cleavage site,

Fig. 4. In vitro processing of the X1-Pro polyprotein and identification of the X1–X2 cleavage site. (A) Schematic representation of the processing of the ArMV X1-Pro polyprotein. The polyprotein is shown at the top. As in Fig. 2,an S-tag epitope (grey box) and a histidine tail (black circle) are indicated. Vertical lines represent cleavage sites. The C/S and C/G dipeptides predicted as the NTB- VPg and X1–X2 cleavage sites are shown. Cleavage products observed after processing of the polyprotein are shown below the arrow with their calculated molecular masses. (B) In vitro processing assays. The wild-type (WT) X1-Pro polyprotein and mutant derivatives X1-ProHS (mutation in the catalytic triad of the proteinase), X1-ProΔCS (deletion of the C/S dipeptide), X1-ProΔCG (deletion of the C/G dipeptide) and X1-ProΔCS + CG (deletion of both the C/S and C/G dipeptides) polyproteins were tested using in vitro processing assays as described in the legend of Fig. 2 After processing, S-tag pull-down assays were conducted as described in Materials and methods and proteins present in the starting reactions (in vitro processing), the immunoprecipitated fraction (S-tag pull-down) or the supernatant after immunoprecipitation (unbound) were separated by SDS-PAGE as indicated above each lane. Cleavage products were visualized after separation on a 12% polyacrylamide gel (top panel). The 20 kDa cleavage product was also separated using a 15% polyacrylamide gel (bottom panel). Bands corresponding to cleavage products are identified by the symbols on the left side of each lane. The migration of molecular mass markers (in kDa) is indicated at the left of the gel. (C) Summary of the results and deduced identity of the cleavage products are as explained in the legend to Fig. 2. NT: not tested. 164 T. Wetzel et al. / Virology 375 (2008) 159–169 we designed a new precursor that included the entire N-terminal region of the ArMV P1 polyprotein (precursor X1-Pro, Fig. 4A). As above, the HS mutation was introduced in the proteinase domain to create mutant X1-ProHS. The results are shown in Fig. 4B and summarized in Fig. 4C. Processing of the WT X1-Pro precursor resulted in the release of multiple cleavage products with apparent molecular masses of 116 kDa, 95 kDa, 85 kDa, 64 kDa, 55 kDa, 32 kDa and 20 kDa (Fig. 4B, lane 2). The apparent molecular masses of these proteins were close to those calculated for the X2-Pro (116 kDa), NTB-Pro (94 kDa), X2-NTB (88 kDa), NTB (66 kDa), X1 (50 kDa) and VPg-Pro (28 kDa) cleavage products. S-tag pull-down assays resulted in the concentration of the 55 kDa protein, confirming that it corresponds to the N-terminal X1 cleavage product (Fig. 4B, lane 9). The 20 kDa protein (putative X2) was difficult to detect in the cleavage products from the WT X1-Pro polyprotein. To improve the detection of this protein and confirm its identity, an HA tag was inserted at the N-terminus of the predicted X2 domain; i.e., three amino acids downstream of the proposed X1–X2 cleavage site (mutant X1-ProN3-X2-HA, Fig. 5A). After processing of the HA-tagged polyprotein, immunoprecipitation experiments were conducted using commercially available anti- HA antibodies (see Materials and methods). The 20 kDa protein (which migrates as a 21 kDa protein after fusion to the HA tag) was concentrated in the HA pull-down fraction (Fig. 5B, lane 5), confirming that it is the mature X2. The 116 kDa and 85 kDa proteins were also present in the HA pull-down fraction. This result was in agreement with the suggestion that these proteins correspond to the X2-Pro and X2-NTB cleavage products, respectively. We also inserted an HA tag between the G604/V and R618/V dipeptides at the junction between the X2 and NTB domains (mutant X1-ProN4-NTB-HA, Fig. 5A). Based on the suggestion that G604/V corresponds to the X2-NTB cleavage site (see above, Fig. 3), the tag was predicted to be inserted at the N-terminus of the NTB domain. Indeed, the 64 kDa protein (putative NTB) was concentrated in the pull-down fraction after Fig. 5. HA-tagging of the predicted X2 and NTB domain within the X1-Pro N4-NTB-HA immunoprecipitation of the cleavage products of X1-Pro polyprotein. (A) Site of insertion of the HA-tags in the ArMV X1-Pro poly- (Fig. 5B, lane 6). The 116 kDa, 95 kDa and 85 kDa proteins were protein. The site of insertion of two HA-tags is indicated by the arrow-heads over also found in the HA pull-down fraction, validating their deduced the schematic representation of the polyprotein. The precise site of insertion is shown by the asterisks in the amino acid sequence below the polyprotein. (B) In identity as the X2-Pro, NTB-Pro and X2-NTB intermediates, vitro processing assays and HA pull-downs. The wild-type (WT) X1-Pro respectively. This result also confirms that cleavage between polyprotein and mutant derivatives X1-ProHS (mutation in the catalytic triad of 604 the X2 and NTB domains occurs at or near the predicted G /V the proteinase), X1-ProN3-X2-HA and X1-ProN4-NTB-HA were synthesized and dipeptide. processed as described in the legend to Fig. 2. After processing, HA pull-down To further analyze the processing of the X1-Pro precursor, assays were conducted as described in Materials and methods and proteins present in the starting reactions (in vitro processing), the immunoprecipitated specific deletions of dipeptides C1216/S (proposed NTB-VPg 414 fraction (HA pull-down) or the supernatant after immunoprecipitation (unbound) cleavage site) or C /G (putative X1–X2 cleavage site) were were separated by SDS-PAGE as indicated above each lane. Cleavage products introduced either individually (mutants ΔCS and ΔCG) or in were visualized after separation on a 12% (top panel) or 8% (lower panel) combination (mutant ΔCS+CG). Deletion of C1216/S prevented polyacrylamide gel. Bands corresponding to cleavage products are identified the release of the 20 kDa (X2), 32 kDa (putative VPg-Pro), 64 kDa with symbols as in Fig. 4. The migration of molecular mass markers (in kDa) is indicated at the left of the gel. (C) Summary of the results and deduced identity of (NTB) and 82 kDa (X2-NTB) cleavage products, while deletion the cleavage products. of C414/G eliminated the production of the 116 kDa (X2-Pro), 85 kDa (X2-NTB), 55 kDa (X1) and 20 kDa (X2) proteins (Fig. 4B, lanes 1–2). After simultaneous deletion of both proposed (Fig. 4C) and supported the suggestion that cleavage between cleavage sites the 95 kDa protein (NTB-Pro) was the only the X1 and X2 domains occurs at or near dipeptide C414/G. cleavage product detected (Fig. 4B, lane 3). These results were Taken together, our results suggest that the X1-Pro poly- consistent with the deduced identity of each cleavage product protein is predominantly cleaved at the X1–X2 and NTB-VPg T. Wetzel et al. / Virology 375 (2008) 159–169 165 sites. The accumulation of large amounts of the X2-NTB intermediate polyprotein and the low concentration of the X2 and NTB cleavage products as determined by quantification of the radioactivity present in each cleavage product (Fig. 4C) also confirmed that the X2-NTB cleavage site is a sub-optimal cleavage site.

Ability of various P1 cleavage sites to be processed intermolecularly

The X1–X2 cleavage site of GFLV was previously shown to be recognized in trans by the VPg-Pro precursor form of the proteinase (Margis et al., 1994). In contrast, the GFLV Pro-Pol site is cleaved only in cis (Margis et al., 1991). In ToRSV, none of the P1 truncated precursors tested (including precursors that contained the X1–X2 site) were recognized in trans by the ToRSV Pro or VPg-Pro (Chisholm et al., 2001; Wang and Sanfacon, 2000). To determine if ArMV P1 cleavage sites could be recognized in trans by the proteinase, the radiolabeled mutated version of the X1-Pro polyprotein precursor containing an inactive proteinase domain (HS mutant) was supplemented with processing products derived after processing of in vitro translated non-labeled NTB- Pro polyprotein and incubated at 16 °C overnight to allow for processing to occur. In addition to the proteinase domain present in NTB-Pro, two other forms of the proteinase (VPg-Pro and Pro) Fig. 6. Trans-processing assays of the X1-Pro precursor. (A) Schematic are released from this polyprotein (see Fig. 2). Thus, this assay representation of the trans-processing assays. The X1-Pro precursor containing a mutation in the catalytic triad of the proteinase (HS mutant) is shown at the top. would allow us to simultaneously test if the proteinase domain After incubation with the processing products of NTB-Pro, only two cleavage contained in NTB-Pro, VPg-Pro or Pro could recognize in trans products are detected. (B) Trans-processing assays. The radiolabeled X1-ProHS HS the cleavage sites present in the X1-Pro polyprotein. As a (HS) was synthesized in vitro and incubated at 16 °C overnight in the absence negative control, we used a mutated version of NTB-Pro con- (lane 2) or the presence (lanes 3–5) of the unlabeled processing products from taining the inactive proteinase domain (HS mutant). As expected, wild-type or mutated versions of the NTB-Pro polyprotein as indicated above HS each lane. As a control, the wild-type (WT) X1-Pro processing products were proteolytic processing of the X1-Pro polyprotein was not examined on the same gel (lane 1). Proteins were separated by SDS-PAGE using observed when it was incubated alone or in combination with the 8% (top panel) or 12% (lower panel) polyacrylamide gels. Only the relevant HS NTB-Pro polyprotein (Fig. 6, lanes 2 and 5). In contrast, the X1 portion of the gels is shown. Bands corresponding to cleavage products are and X2-Pro cleavage products were released from the X1-ProHS identified with symbols as in Fig. 4. The migration of molecular mass markers polyprotein after incubation with the processing products derived (in kDa) is indicated at the right of the gel. from the wild-type NTB-Pro (lane 3). Other possible cleavage products of the X1-ProHS polyprotein (e.g. X2-NTB or VPg-Pro) was similar to those previously reported for the ToRSV P1 were not detected in the trans-processing assays. The X1-ProHS polyprotein (Wang and Sanfacon, 2000) and delineates two polyprotein was also incubated with the processing products of protein domains upstream of NTB. Using the ClustalW the NTB-ProΔCS mutant polyprotein. The efficiency of release of program, we aligned the deduced amino acid sequences of the the X1 and X2-Pro cleavage products was similar after overnight entire P1 polyprotein of nepoviruses currently available in the incubation of X1-ProHS with the processing products derived database. The quality of the alignment was verified by ensuring from the WT or mutated (ΔCS) versions of the NTB-Pro that previously identified conserved amino acids in the X2, polyprotein. We have shown above that VPg-Pro is not released NTB, Pro and Pol domains were properly aligned (Fig. 7 and from NTB-ProΔCS. These results suggest that either the mature data not shown). The portions of the alignment corresponding to proteinase (released in small amounts from the NTB-ProΔCS and regions flanking the putative X1–X2 and X2-NTB cleavage NTB-Pro polyproteins) and/or the proteinase domain present in sites are shown in Fig. 7. Cleavage sites were deduced based on the unprocessed NTB-Pro are responsible for trans-cleavage of the known specificity of the proteinases from viruses of the X1-ProHS at the X1–X2 cleavage site. different subgroups. The previously identified or suggested ToRSV X1–X2 and X2-NTB sites, GFLV X1–X2 site, BRSV Prediction of proteinase cleavage sites in the N-terminal region X2-NTB site and ArMV X1–X2 and X2-NTB sites (Deman- of the P1 polyprotein from other nepoviruses geat et al., 1990; Margis et al., 1994; Wang and Sanfacon, 2000 and this study) were also used as a guide to predict other The results presented above suggest that the ArMV pro- cleavage sites. teinase recognizes two cleavage sites upstream of the NTB Dipeptides R/A or R/S were predicted as putative cleavage domain, at least in vitro. The position of these cleavage sites sites in the P1 polyprotein of BRSV, Tomato black ring virus, 166 T. Wetzel et al. / Virology 375 (2008) 159–169

Fig. 7. Amino acid alignments of the X2 region of the P1 polyprotein of nepoviruses and prediction of X1–X2 and X2-NTB cleavage sites. Amino acid sequences from the P1 polyprotein of nepoviruses were aligned using the ClustalW program as described in Materials and methods. Two regions of the alignment are shown. The upper portion of the figure includes the N-terminal region of X2 and the putative X1–X2 cleavage site. The lower portion of the figure includes the N-terminal region of NTB and the putative X2-NTB cleavage site. A boxed phenylalanine (F) within the X2 domain corresponds to the first consensus amino acid of the motif conserved in nepoviruses and comoviruses (Zhang and Sanfacon, 2006). Predicted cleavage sites are underlined. Cleavage sites supported by experimental evidence (mutagenesis) are also shown in bold italics. In the case of the TRSV X2-NTB cleavage site, two possible dipeptides are underlined.

Grapevine chrome mosaic virus or Cycas necrosis stunt virus Based on site-directed mutagenesis of putative cleavage sites (subgroup B nepoviruses, see Fig. 7), which is in agreement and HA-tagging experiments, we tentatively propose dipeptides with previously identified subgroup B nepovirus cleavage sites C414/G, G604/V and C1216/S as the three N-terminal cleavage that have a lysine or arginine at the −1 position (Demangeat sites of the polyprotein. Dipeptide C1216/S separates the NTB and et al., 1991; Hemmer et al., 1995). In the case of Blackcurrant VPg domains. Based on amino acid alignments of the entire reversion virus and Peach rosette mosaic virus, two subgroup P1 polyproteins from various nepoviruses, the proposed ArMV C nepoviruses, dipeptides N/G, D/G, N/A or D/L were pre- NTB-VPg cleavage site aligns well with previously identified dicted as possible cleavage sites. Processing after a glutamine, NTB-VPg cleavage sites from other nepovirus P1 polypro- asparginine or aspartate was previously reported for this sub- teins (data not shown). The positions of dipeptides C414/G group of nepoviruses (Bacher et al., 1994; Carrier et al., 1999; and G604/V correspond to those of cleavage sites identified at the Latvala et al., 1998). Finally, for GFLV, Tobacco ringspot virus N-terminus of the ToRSV polyprotein (Wang and Sanfacon, and Raspberry ringspot virus, three subgroup A nepoviruses, 2000) and delineate two protein domains upstream of the NTB dipeptides C/G, C/A, C/S, G/V, G/A or G/L were predicted as domain. Possible cleavage sites were found at equivalent posi- possible cleavage sites based on the previous identification of tions in the P1 polyproteins of all other nepoviruses examined. subgroup A nepovirus cleavage sites with a cysteine, arginine or Although further experiments will be required to confirm that the glycine at the −1 position (Blok et al., 1992; Buckley et al., proposed cleavage sites are recognized by the corresponding 1993; Margis et al., 1993; Pinck et al., 1991; Serghini et al., viral proteinases, this analysis suggests that the presence of two 1990). In all cases, putative cleavage sites consistent with the protein domains (termed X1 and X2 by analogy with the specificity of the corresponding proteinase were found at posi- proposed genome organization of ToRSV) in the N-terminal tions equivalent to that of the X1–X2 and X2-NTB cleavage region of the P1 polyprotein upstream of the NTB domain is a sites from the ToRSV and ArMV P1 polyproteins, suggesting conserved feature among nepoviruses. that the presence of two cleavage sites upstream of the NTB The results presented here suggest that the ArMV X2-NTB domain is conserved among nepoviruses. cleavage site is a sub-optimal cleavage site and the X2-NTB precursor was shown to accumulate in the in vitro cleavage Discussion products of the X2-Pro and X1-Pro precursors. In the case of ToRSV, in vitro cleavage between X2-NTB is readily detected In this study, we have analyzed the processing of the although the X2-NTB intermediate polyprotein also accumu- N-terminal region of the ArMV P1 polyprotein. Four cleavage lates in the cleavage products of truncated P1 polyprotein sites were shown to be recognized by the ArMV proteinase. precursors (Wang and Sanfacon, 2000). In ToRSV-infected T. Wetzel et al. / Virology 375 (2008) 159–169 167 plants, the mature NTB protein and NTB-VPg intermediate et al., 2001; Wang and Sanfacon, 2000). One possible expla- polyprotein are found in association with endoplasmic reticu- nation for these differences is that the absence of the N-terminal lum-derived membranes, confirming that cleavage between the portion of X1 on the truncated ToRSV cX1-Pro precursor X2 and NTB domains occurs in vivo (Han and Sanfacon, 2003). influences the folding of the polyprotein and prevents optimal However, small amounts of an intermediate precursor of the presentation of the X1–X2 cleavage site for recognition by the expected molecular mass for the X2-NTB-VPg protein are also proteinase. Alternatively, it is also possible that regulation of consistently detected in association with ER membranes in cleavage at the X1–X2 site differs among nepoviruses. infected cells, suggesting that cleavage at the junction between While X2 is related to the comovirus 32 kDa protein, the the X2 and NTB domains is incomplete (Chisholm et al., 2007; X1 protein domain is unique to nepoviruses within the family Han and Sanfacon, 2003). Further experiments will be Comoviridae. This protein domain is not conserved among nepo- necessary to study the efficiency of cleavage at the ArMV viruses and amino acid alignments of this region of the poly- X2-NTB site in vivo and to compare the relative accumulation protein are difficult to validate because of the lack of conserved of the X2-NTB, X2 and NTB proteins in infected plants. amino acid sequence motifs (data not shown). Further studies will Nepovirus X2 proteins share many common properties with be necessary to decipher the biological function of this protein in the 32 kDa protein of Cowpea mosaic virus (CPMV,a comovirus) the replication cycle of nepoviruses and determine why it is absent including having a very hydrophobic nature and a signature from the genome of the related comoviruses and fabaviruses. amino acid motif (Zhang and Sanfacon, 2006). The ToRSV X2 In conclusion, our results suggest that the presence of two protein and the CPMV 32 kDa protein both have the ability to protein domains upstream of the NTB domain is a distinguish- interact directly with endoplasmic reticulum (ER) membranes and ing common feature of nepoviruses, although the specific regu- have been suggested to play a role in anchoring the viral lation of proteolytic cleavage at these sites may differ from one replication complex into the membranes (Carette et al., 2002; nepovirus to another. The genus Nepovirus constitutes an in- Zhang and Sanfacon, 2006). The NTB proteins of ToRSV and teresting group of viruses that share many similar properties but CPMV are also associated with ER-derived membranes and are are at the same time very diverse (Le Gall et al., 2005). probably also involved in the assembly of the replication complex Nepoviruses have been separated into three subgroups based on (Carette et al., 2002; Han and Sanfacon, 2003; Wang et al., 2004; the length and genomic organization of RNA 2, the cleavage Zhang et al., 2005). The accumulation of intermediate poly- site specificity of the proteinase and phylogenetic analysis using proteins that contain both the X2 and NTB domains may allow the coat protein sequence (Le Gall et al., 2005, 2007). Based on coordination of the function of these proteins in viral replication these distinct characteristics, it could be argued that nepoviruses complex assembly. It is interesting to note that in the case of should be separated into three distinct genera. However, the CPMV, although cleavage between the 32 kDa protein and the three nepovirus subgroups do not separate in distinct clades in NTB domain is very efficient in vivo and in vitro, the two proteins phylogenetic studies using the Pro-Pol region (Le Gall et al., interact with each other and remain associated after proteolytic 2007). A distinct and unifying property of members of the cleavage (Peters et al., 1992). genus Nepovirus is the presence of a single large CP subunit. The CPMV 32 kDa protein has been shown to modulate the The results presented here suggest that the genomic organiza- activity of the proteinase (Peters et al., 1992). It has been tion of RNA 1 is another common property of nepoviruses proposed that an interaction between the 32 kDa protein and the which is shared by members of all three subgroups. NTB domain present within the NTB-VPg-Pro-Pol polyprotein alters the conformation of this polyprotein and renders cleavage Materials and methods sites less accessible to recognition by the proteinase. However, there is no evidence that the X2 protein of nepoviruses play a Construction of plasmids similar role. Indeed, the presence or absence of the X2 domain on truncated P1 polyproteins from ArMV or ToRSV does not Plasmids pCITE-NTB-Pro, pCITE-X2-Pro and pCITE-X1- significantly influence the efficiency of processing at the NTB- Pro were constructed by amplifying the corresponding region of VPg site (Wang and Sanfacon, 2000 and this study). the ArMV genome from a full-length cDNA clone of a grapevine Our results indicate that the ArMV X1–X2 cleavage site is isolate of ArMV (Wetzel et al., 2004 and unpublished) using the recognized very efficiently by the proteinase, at least in vitro. Phusion High-Fidelity PCR kit (Finnzymes, distributed by New Indeed, precursors that include both the X1 and X2 domains are England Biolabs). The following primer pairs were used for the not detected in significant amounts in the cleavage products of amplification: for plasmid pCITE-NTB-Pro, primer Pro-as the X1-Pro polyprotein. In contrast, ToRSV intermediate poly- [5′ acgcgtcgactctcacaaatgaggagctataac 3′, complementary to proteins containing the X1 and X2 domains were consistently nts 4064–3982 of ArMV RNA1 and including a SalI restriction detected in the cleavage products of a truncated P1 precursor that site (underlined)] and primer NTB-s [5′ acgcgatatcgtcggtcagag- included the C-terminal portion of X1 (cX1-Pro precursor, Wang catgattagc 3′, corresponding to nts 2082–2102 of ArMV RNA1 and Sanfacon, 2000). Another notable difference is that while and including an EcoRVrestriction site]; for plasmid pCITE-X2- the ArMV and GFLV X1–X2 cleavage sites can be recognized Pro, primers Pro-as and X2-s [5′ acgcgatatcgggcaatttaatgattgg- in trans by one or several forms of the proteinase (Margis gcttc 3′, corresponding to nts 1470–1492 of ArMV RNA1 and et al., 1994 and this study), trans-cleavage at the ToRSV X1–X2 including an EcoRVrestriction site]; and for plasmid pCITE-X1- cleavage site by Pro or VPg-Pro was not detected (Chisholm Pro, primers Pro-as and X1-s [5′ acgcgatatcatgtggcaaatttctgaggg 168 T. Wetzel et al. / Virology 375 (2008) 159–169

3′, corresponding to nts 228–247 of ArMV RNA1 and including Amino acid alignments an EcoRV restriction site]. The amplified fragments were di- gested by EcoRV and SalI and inserted into the corresponding Amino acid sequences from the entire P1 polyprotein of sites of plasmid pCITE-4a(+) (Novagen). Mutation, deletion or nepoviruses were aligned using the ClustalW program (Chenna insertion of amino acids was produced by site-directed muta- et al., 2003) and were manually readjusted to ensure alignments genesis using specific primers, as described (Urban et al., 1997). of previously identified conserved amino acids. The following All plasmids were sequenced prior to analysis to verify that the accession numbers were used for the alignment: Peach rosette desired mutations were correctly introduced. mosaic virus (PRMV, AAB69867), Blackcurrant reversion virus (BRV, AF3682772), ToRSV (DQ469829 for the X2 domain In vitro processing assays and immunoprecipitations and L19655 for the rest of the P1 polyprotein), Tomato black ring virus (TBRV, AY157993), Grapevine chrome mosaic Coupled in vitro transcription and translation reactions and virus (GCMV, X15346), Cycas necrotic stunt virus (CNSV, in vitro processing assays were conducted as described previously AB073147), BRSV (D00322), Tobacco ringspot virus (TRSV, for the processing of ToRSV polyproteins (Wang et al., 1999). U50869), Raspberry ringspot virus (RpRSV, AY303787), ArMV Briefly, in vitro translations were conducted for 90 min at 30 °C in (AY303786), and GFLV (D00915). the presence of [35S]-methionine using the TNT coupled rabbit reticulocyte system (Promega). After translation, RNase ‘A’ and Acknowledgments cold methionine were added and the samples were incubated at room temperature for 10 min. The translation products were then This research was supported by a grant from Agriculture and diluted 1:2 in 2× processing buffer (200 mM Tris–HCl pH 8.0, Agri-Food Canada and by a grant from the Klaus Tschira Stiftung 2 mM DTT, 20% (v/v) glycerol) and incubated at 16 °C overnight in Germany, and was facilitated by a collaborative grant from the to allow for proteolytic processing. For trans-processing assays, German International Bureau of the federal Ministry of Education wild-type or mutated versions of the NTB-Pro polyprotein were and Research (Internationales Büro des Bundesministerium für first synthesized in vitro and processed as above but in the Bildung und Forschung, CAN06/A06). presence of cold methionine rather than radiolabeled methionine. The cold NTB-Pro processing products were then added to the References radiolabeled X1-ProHS polyprotein produced by in vitro translation in the presence of [35S]-methionine. The mixture was then Bacher, J.W., Warkentin, D., Ramsdell, D., Hancock, J.F., 1994. Sequence ′ incubated at 16 °C overnight to allow for processing of X1-ProHS. analysis of the 3 termini of RNA1 and RNA2 of Blueberry leaf mottle virus. Virus Res. 33, 145–156. Labeled proteins were separated by SDS-PAGE (Laemmli, 1970) Blok, V.C., Wardell, J., Jolly, C.A., Manoukian, A., Robinson, D.J., Edwards, and visualized by autoradiography. For quantification of the M.L., Mayo, M.A., 1992. The nucleotide sequence of RNA-2 of raspberry relative concentration of each cleavage product, gels were ex- ringspot nepovirus. J. Gen. Virol. 73, 2189–2194. posed to a phosphorimager screen and the radioactivity present Buckley, B., Silva, S., Singh, S., 1993. Nucleotide sequence and in vitro expression – into each band was estimated using the OptiQuant program and of the capsid protein gene of Tobacco ringspot virus. Virus Res. 30, 335 349. Carette, J.E., van Lent, J., MacFarlane, S.A., Wellink, J., van Kammen, A., a phosphorimager (Cyclone Plus, Perkin Elmer). The values 2002. Cowpea mosaic virus 32- and 60-kilodalton replication proteins target (digital light units) corresponding to each band were readjusted and change the morphology of endoplasmic reticulum membranes. J. Virol. taking into consideration the number of radiolabeled methionine 76, 6293–6301. incorporated in each protein. Carrier, K., Hans, F., Sanfacon, H., 1999. Mutagenesis of amino acids at two S-tag pull-down assays were conducted essentially as de- tomato ringspot nepovirus cleavage sites: effect on proteolytic processing in cis and in trans by the 3C-like protease. Virology 258, 161–175. scribed in the manufacturer's instructions (Novagen). Translation Carrier, K., Xiang, Y., Sanfacon, H., 2001. Genomic organization of RNA2 of products were incubated with S-protein agarose under 1× binding Tomato ringspot virus: processing at a third cleavage site in the N-terminal buffer conditions (20 mM Tris, pH 7.5; 150 mM NaCl; 0.1% region of the polyprotein in vitro. J. Gen. Virol. 82, 1785–1790. Triton X-100) and in the presence of Complete protease inhibitor Chenna, R., Sugawara, H., Koike, T., Lopez, R., Gibson, T.J., Higgins, D.G., (Roche). The agarose was washed 3 times in 1× binding buffer Thompson, J.D., 2003. Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Res. 31, 3497–3500. and then finally resuspended in 2× Laemmli SDS-PAGE buffer Chisholm, J., Wieczorek, A., Sanfacon, H., 2001. Expression and partial puri- (Laemmli, 1970). Samples were heat treated and the products fication of recombinant tomato ringspot nepovirus 3C-like proteinase: were separated by SDS-PAGE and visualized by autoradiography. comparison of the activity of the mature proteinase and the VPg-proteinase For HA-pull-down experiments, the anti-HA affinity matrix precursor. Virus Res. 79, 153–164. (Roche) was pre-treated with 1% BSA (w/v) in PBS-TDS [PBS Chisholm, J., Zhang, G., Wang, A., Sanfacon, H., 2007. Peripheral association of a polyprotein precursor form of the RNA-dependent RNA polymerase of containing 1% (v/v) Triton X-100 and 0.05% (w/v) sodium Tomato ringspot virus with the membrane-bound viral replication complex. deoxycholate] for 30 min. The matrix was then added to Virology 368, 133–144. translations that were adjusted to 50 mM Tris, pH 7.5; 0.15 M Demangeat, G., Greif, C., Hemmer, O., Fritsch, C., 1990. Analysis of the in vitro NaCl; 1% NP-40 and incubations were allowed to proceed for 1 h cleavage products of the Tomato black ring virus RNA-1-encoded 250K – in the presence of Complete protease inhibitor (Roche). Samples polyprotein. J. Gen. Virol. 71, 1649 1654. Demangeat, G., Hemmer, O., Fritsch, C., Le Gall, O., Candresse, T., 1991. In vitro were washed 7 times in PBS-TDS and finally, the matrix was processing of the RNA-2-encoded polyprotein of two nepoviruses: Tomato resuspended in 2× Laemmli buffer, heat treated and the products black ring virus and Grapevine chrome mosaic virus. J. Gen. Virol. 72, were analyzed by SDS-PAGE followed by autoradiography. 247–252. T. Wetzel et al. / Virology 375 (2008) 159–169 169

Gorbalenya, A.E., Donchenko, A.P., Blinov, V.M., Koonin, E.V., 1989a. Mayo, M.A., Robinson, D.J., 1996. Nepoviruses: molecular biology and Cysteine proteases of positive strand RNA viruses and chymotrypsin-like replication. In: Harrison, B.D., Murrant, A.F. (Eds.), The Plant Viruses, serine proteases. A distinct protein superfamily with a common structural Volume 5: Polyhedral Virions and Bipartite RNA Genomes. Plenum Press, fold. FEBS Lett. 243, 103–114. New York, pp. 139–185. Gorbalenya, A.E., Koonin, E.V., Donchenko, A.P., Blinov, V.M., 1989b. Peters, S.A., Voorhorst, W.G., Wery, J., Wellink, J., van Kammen, A., 1992. A Coronavirus genome: prediction of putative functional domains in the non- regulatory role for the 32K protein in proteolytic processing of Cowpea structural polyprotein by comparative amino acid sequence analysis. Nucleic mosaic virus polyproteins. Virology 191, 81–89. Acids Res. 17, 4847–4861. Pinck, M., Reinbolt, J., Loudes, A.M., Le Ret, M., Pinck, L., 1991. Primary Gorbalenya, A.E., Koonin, E.V., Wolf, Y.I., 1990. A new superfamily of putative structure and location of the genome-linked protein (VPg) of Grapevine NTP-binding domains encoded by genomes of small DNA and RNA fanleaf nepovirus. FEBS Lett. 284, 117–119. viruses. FEBS Lett. 262, 145–148. Ritzenthaler, C., Viry, M., Pinck, M., Margis, R., Fuchs, M., Pinck, L., 1991. Han, S., Sanfacon, H., 2003. Tomato ringspot virus proteins containing the Complete nucleotide sequence and genetic organization of Grapevine nucleoside triphosphate binding domain are transmembrane proteins that fanleaf nepovirus RNA1. J. Gen. Virol. 72, 2357–2365. associate with the endoplasmic reticulum and cofractionate with replication Rott, M.E., Gilchrist, A., Lee, L., Rochon, D., 1995. Nucleotide sequence of complexes. J. Virol. 77, 523–534. Tomato ringspot virus RNA1. J. Gen. Virol. 76, 465–473. Hans, F., Sanfacon, H., 1995. Tomato ringspot nepovirus protease: character- Seipelt, J., Guarne, A., Bergmann, E., James, M., Sommergruber, W., Fita, I., ization and cleavage site specificity. J. Gen. Virol. 76, 917–927. Skern, T., 1999. The structures of picornaviral proteinases. Virus Res. 62, Hemmer, O., Greif, C., Dufourcq, P., Reinbolt, J., Fritsch, C., 1995. Functional 159–168. characterization of the proteolytic activity of the tomato black ring nepovirus Serghini, M.A., Fuchs, M., Pinck, M., Reinbolt, J., Walter, B., Pinck, L., 1990. RNA-1-encoded polyprotein. Virology 206, 362–371. RNA2 of Grapevine fanleaf virus: sequence analysis and coat protein Koonin, E.V., Dolja, V.V., 1993. Evolution and taxonomy of positive-strand cistron location. J. Gen. Virol. 71, 1433–1441. RNA viruses: implications of comparative analysis of amino acid sequences. Urban, A., Neukirchen, S., Jaeger, K.E., 1997. A rapid and efficient method for Crit. Rev. Biochem. Mol. Biol. 28, 375–430. site-directed mutagenesis using one-step overlap extension PCR. Nucleic Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the Acids Res. 25, 2227–2228. head of bacteriophage T4. Nature 227, 680–685. Wang, A., Carrier, K., Chisholm, J., Wieczorek, A., Huguenot, C., Sanfacon, H., Latvala, S., Susi, P., Kalkkinen, N., Lehto, K., 1998. Characterization of the coat 1999. Proteolytic processing of Tomato ringspot nepovirus 3C-like protease protein gene of mite-transmitted blackcurrant reversion associated nepo- precursors: definition of the domains for the VPg, protease and putative virus. Virus Res. 53, 1–11. RNA-dependent RNA polymerase. J. Gen. Virol. 80, 799–809. Le Gall, O., Iwanami, T., Jones, A.T., Lehto, K., Sanfacon, H., Wellink, J., Wang, A., Sanfacon, H., 2000. Proteolytic processing at a novel cleavage site in Wetzel, T., Yoshikawa, N., 2005. Comoviridae. In: Fauquet, C.M., Mayo, the N-terminal region of the tomato ringspot nepovirus RNA-1-encoded M.A., Maniloff, J., Desselberger, U., Ball, L.A. (Eds.), Virus Taxonomy, polyprotein in vitro. J. Gen. Virol. 81, 2771–2781. Eighth Report of the International Committee on the Taxonomy of Viruses. Wang, A., Han, S., Sanfacon, H., 2004. Topogenesis in membranes of the NTB- Elsevier/Academic Press, London, pp. 8078–8818. VPg protein of Tomato ringspot nepovirus: definition of the C-terminal Le Gall, O., Sanfacon, H., Ikegami, M., Iwanami, T., Jones, T., Karasev, A., transmembrane domain. J. Gen. Virol. 85, 535–545. Lehto, K., Wellink, J., Wetzel, T., Yoshikawa, N., 2007. Cheravirus and Wellink, J., Rezelman, G., Goldbach, R., Beyreuther, K., 1986. Determination of Sadwavirus: two unassigned genera of plant positive-sense single-stranded the proteolytic processing sites in the polyprotein encoded by the bottom- RNA viruses formerly considered atypical members of the genus Nepovirus component RNA of Cowpea mosaic virus. J. Virol. 59, 50–58. (family Comoviridae). Arch. Virol. 159, 1767–1774. Wetzel, T., Beck, A., Wegener, U., Krczal, G., 2004. Complete nucleotide Margis, R., Pinck, L., 1992. Effects of site-directed mutagenesis on the sequence of the RNA 1 of a grapevine isolate of Arabis mosaic virus. Arch. presumed catalytic triad and substrate-binding pocket of grapevine fanleaf Virol. 149, 989–995. nepovirus 24-kDa proteinase. Virology 190, 884–888. Wetzel, T., Meunier, L., Jaeger, U., Reustle, G.M., Krczal, G., 2001. Complete Margis, R., Viry, M., Pinck, M., Pinck, L., 1991. Cloning and in vitro char- nucleotide sequences of the RNAs 2 of German isolates of grapevine fanleaf acterization of the Grapevine fanleaf virus proteinase cistron. Virology 185, and Arabis mosaic nepoviruses. Virus Res. 75, 139–145. 779–787. Zhang, G., Sanfacon, H., 2006. Characterization of membrane-association domains Margis, R., Ritzenthaler, C., Reinbolt, J., Pinck, M., Pinck, L., 1993. Genome within the Tomato ringspot nepovirus X2 protein, an endoplasmic reticulum- organization of Grapevine fanleaf nepovirus RNA2 deduced from the 122K targeted polytopic membrane protein. J. Virol. 80, 10847–10857. polyprotein P2 in vitro cleavage products. J. Gen. Virol. 74, 1919–1926. Zhang, S.C., Zhang, G., Yang, L., Chisholm, J., Sanfacon, H., 2005. Evidence that Margis, R., Viry, M., Pinck, M., Bardonnet, N., Pinck, L., 1994. Differential insertion of Tomato ringspot nepovirus NTB-VPg protein in endoplasmic proteolytic activities of precursor and mature forms of the 24K proteinase of reticulum membranes is directed by two domains: a C-terminal transmembrane grapevine fanleaf nepovirus. Virology 200, 79–86. helix and an N-terminal amphipathic helix. J. Virol. 79, 11752–11765.