Archives of Virology

, Volume 153, Issue 10, pp 1955–1960

Cleavage sites of the “P3 region” in the nonstructural polyprotein precursor of a dicistrovirus


    • National Institute of Agrobiological Sciences
  • Yuri Nakamura
    • National Institute of Agrobiological Sciences
Brief Report

DOI: 10.1007/s00705-008-0208-5

Cite this article as:
Nakashima, N. & Nakamura, Y. Arch Virol (2008) 153: 1955. doi:10.1007/s00705-008-0208-5


Plautia stali intestine virus (PSIV) is a member of the family Dicistroviridae. Dicistroviruses, like picornaviruses, are thought to encode a 3C-like protease; however, the cleavage sites of dicistroviral nonstructural polyprotein precursors are unknown except for those in genome-linked virus protein (VPg) regions. We expressed part of the PSIV polyprotein in Escherichia coli and detected autoprocessed viral proteins. N-terminal sequence analysis of the autoprocessed proteins showed that Q/GWLSW and Q/NGVFD in the PSIV sequence correspond to 2C/3A and 3C/3D cleavage sites in picornaviruses. Alignments of deduced amino acid sequences of the “P3 region” suggest that these cleavage sites can be predicted.

Picornaviruses have a positive-stranded RNA genome with a single large open reading frame (ORF). The ORF is translated by internal ribosome entry site (IRES)-mediated initiation to produce a large polyprotein precursor that is cleaved to form functional viral proteins. This cleavage is performed by virally encoded proteases, which are named L, 2A, and 3C proteases under the systematic L434 nomenclature system [11]. In contrast, dicistroviruses have two nonoverlapping ORFs, both of which possess their own IRES elements [4, 10]. ORF1 is located in the 5′ part of the genome and encodes motifs for helicase, protease, and RNA-dependent RNA polymerase, which are homologous to those of viruses in the recently created order of Picornavirales [6]. However, the cleavage sites in the nonstructural protein precursors of dicistroviruses have not been mapped, except for those in the genome-linked virus protein (VPg) regions [9]. Characterization of the cleavage sites in ORF1 products would facilitate analyses of how translation of host protein synthesis is controlled by viral infection.

Here, we have determined cleavage sites in the nonstructural protein precursor of PSIV by carrying out direct N-terminal sequencing of 3C-like protease-mediated cleavage products obtained from bacterially expressed proteins. Because nonstructural proteins of dicistroviruses, such as the 3C-like protease and 3D-like polymerase, are thought to have significant similarity to those of picornaviruses, we have tentatively applied the L434 system for naming nonstructural proteins of dicistroviruses. Although the concepts employed for the P1, P2, and P3 precursors may not be applicable to dicistroviruses because the dicistroviral capsid precursor is encoded in a separate ORF (Fig. 1a), previous studies have shown that capsid proteins are primarily observed in experiments using dicistroviruses [8, 17]. We therefore considered that designation of the capsid precursor as P1 and the replicases as P2 and P3 would be appropriate.
Fig. 1

a Comparison of genomic organization between a picornavirus (foot-and-mouth disease virus, FMDV) and a dicistrovirus (PSIV). b Alignment of amino acid sequences deduced from dicistroviral 3C-like sequences. Amino acid positions at the terminal residues are shown in parentheses. Highly conserved amino acids are shown on a black background. The cysteine residue that constitutes part of the catalytic triad is indicated by an asterisk above the sequence. Accession numbers for viral sequences: PSIV AB006531; TrV AF178440; BQCV AF183905; HiPV AB017037; HoCV-1 DQ288865; RhPV AF022937; ALPV AF536531; IAPV EF219380; KBV AY275710; ABPV AF150629; SINV-1 AY634314; DCV AF014388; CrPV AF218039; TSV AF277675. c, d Detection of histidine- or Flag-tagged proteins in bacterial lysates by western blot. Molecular masses of marker proteins are shown at the left of the panel. Protein products from PSIV cDNA are marked at the right of the panel. e Diagrams showing expressed PSIV cDNAs and cleavage sites in the nonstructural polyprotein precursor of PSIV ORF1

Alignment of the dicistroviral amino acid sequences deduced from ORF1 suggests that these viruses encode the core motif, GXCG, of a 3C-like protease (Fig. 1b) [2]. Because 3C proteases have autoprocessing activity, we analyzed N-terminal amino acid sequences of histidine-tagged proteins expressed from pET vectors that produce a C-terminal histidine tag.

We first detected proteins from a cDNA region containing nucleotides 2656–4380, which correspond to deduce amino acid sequence positions 696–1270 in PSIV ORF1. The cDNA fragment was amplified from the virus by RT-PCR with forward primer 5′-taccatggagaacttgcaatcgaaagt-3′ and reverse primer 5′-ttgcggccgcttgaattgtcgggttcagat-3′. Bold type in these primer sequences indicates recognition sequences for NcoI and NotI. The amplified fragment was cloned into pT7Blue T-vector (Novagen), sequenced, digested with NcoI and NotI, and ligated with sites of pET20b+ (Novagen), generating pET-c1(His) (Fig. 1e). We assumed that this construct would contain the cleavage site of 2C/3A that corresponds to the picornaviral cleavage site in precursors P2 and P3. Because PSIV ORF1 encodes the conserved core motif of the 3C-like protease at positions 1205–1242 (Fig. 1b), this cDNA was estimated to stretch from the C-terminal part of the 2C-helicase-like region to the 3C protease region. Therefore, we designated the expected protein product 2C3ABC. When the cloned cDNA was expressed by using pET-c1(His), the bacterial lysate contained several proteins detectable with anti-His-tag antibody (Fig. 1c, lane 2). To identify proteins cleaved by the virally encoded protease, we mutated the cysteine residue at position 1219 (Fig. 1b) to alanine because this cysteine is likely to constitute part of the catalytic triad in the 3C-cysteine protease [2]. After bacterial expression of the C1219A vector, the His-tag antibody did not detect the 50- and 30-kDa proteins, and the band intensity of the 65-kDa protein increased as a result of the C1219A mutation (Fig. 1c, lane 3), indicating that 1219C is indeed involved in the catalytic activity. This observation means that partial cleavage of the 65-kDa protein produced the 50- and 30-kDa proteins observed in Fig. 1c, lane 2. A faint band was observed at approximately 50 kDa in the alanine mutant (Fig. 1c, lane 3), but this was one of a few background proteins that were visible throughout the lane. N-terminal Edman sequencing showed that the 50-kDa protein was produced by cleavage at 823Q/824G. This QG pair was aligned very closely to the positions of the 2C/3A cleavage sites of picornaviruses when the deduced amino acid sequences of PSIV ORF1 and picornaviral P2-P3 regions were analyzed by ClustalX (data not shown); thus, we concluded that 823Q/824G represents the 2C/3A cleavage site in PSIV. Results for the 30-kDa protein are described later.

Next, we analyzed His-tag proteins expressed from pET-101(His) containing cDNA of PSIV nucleotide positions 2656–4845, corresponding to deduced amino acid sequence positions 696–1425, to identify the 3C/3D cleavage site. N-terminal sequence analyses, however, detected unexpected proteins, such as bacterial ribosomal protein rpL13, which contains repeated histidine residues, and a protein produced from the internal AUG triplet of PSIV cDNA positions 4402–4405, which is preceded by an AG-rich Shine-Dalgarno-like sequence (data not shown). To prevent these problems, we replaced the C-terminal His-tag sequence with the Flag-tag sequence by PCR-based mutagenesis, and the AUG triplet at positions 4402–4404 was changed to GCG. The modified plasmid produced a Flag-tagged protein at 18 kDa (Fig. 1d, lane 2). This protein was not produced when 1219C was mutated to alanine (Fig. 1d, lane 3), indicating that the 18-kDa protein is a 3C-cleaved product. The N-terminal sequence of the 18-kDa protein was NGVFD. This sequence is found between the conserved 3C protease and 3D polymerase motifs in the deduced amino acid sequence of PSIV. In addition, alignment of 3CD regions including those of PSIV and picornaviruses by ClustalX located the 1280Q/1281N cleavage site at a region corresponding to the 3C/3D cleavage site of picornaviruses (data not shown). Thus, we conclude that the 3C/3D cleavage site in PSIV is 1280Q/1281N.

The plasmid pET-c101(Flag) produced the protein product 3ABCD (Fig. 1d, lane 2), which is likely to correspond to the 50-kDa protein from pET-c1(His) (Fig. 1c, lane 2), because the pET-c101(Flag) vector was assumed to produce a protein approximately 17 kDa larger than that expressed from pET-c1(His) (Fig. 1e). However, a protein thought to correspond to the 30-kDa protein expressed from pET-c1(His) was not expressed from pET-c101(Flag) (Fig. 1d, lane 2). The N-terminal sequence of the 30-kDa protein observed in Fig. 1c, lane 2, indicates that this protein was cleaved at 1009Q/1010N. This site could not be identified as the 3B/3C junction because 1009Q is located 24 amino acids downstream from the C-terminal end of the repeated VPg coding region [9]. The sequence of the 24 amino acids 986SVFNATEDSVNMQCNGNVENLALQ does not resemble the triplicate and loosely conserved VPg sequence 943, 957, 971SQEKXGXXXXXXXE [9]. Because the 1009Q/1010 N cleavage was not detected in the expression assay using the pET-c101(Flag) plasmid, the cleavage forming the 30-kDa protein from pET-c1(His) was indicated to be artificial. The presence of a stable 3CD precursor is known in picornaviruses; however, the polyprotein expressed from pET-c1(His) did not contain the 3D region. This truncation might lead to the unusual cleavage observed for the 30-kDa protein. Also, cleavage in the 3B region was not observed in any of our bacterial expression analyses. Because picornaviruses form stable processing intermediates during their replication cycles [5], 3B regions would be difficult to cleave in our bacterial experiment.

Phylogenetic analysis based on the deduced amino acid sequences of the RNA-dependent RNA polymerase of dicistroviruses has been published [3]. Dicistroviruses seem to be composed of several groups: (1) ABPV, IAPV, KBV and SINV-1; (2) CrPV and DCV; (3) TSV; (4) BQCV, HiPV, TrV, PSIV, and HoCV-1; and (5) ALPV and RhPV. In each group of viruses, residues flanking the cleavage sites between the capsid protein VP2/VP4 and VP3/VP1 are conserved in comparison to those flanking these sites in other groups of viruses (Table 1). This sequence preference for amino acids flanking cleavage sites in 3C-mediated cleavage may be explained by analysis using picornaviruses. For example, the 3C protease of poliovirus 1 has a narrow catalytic space and cannot accommodate residues larger than glycine at the N-terminus of the residue to be cleaved, whereas that of hepatitis A virus has a wider catalytic space and can accommodate larger amino acids [12]. In the case of PSIV, LXLQ/SG is conserved in the capsid cleavage sites (Table 1); however, the cleavage sites for 2C/3A and 3C/3D are LATQ/GW (Fig. 1c) and KMIQ/NG (Fig. 1d), suggesting that a dicistroviral 3C can accommodate several types of amino acid at the cleavage site. On the basis of these observations, we next carried out sequence alignment to estimate the cleavage site in the P3 region of dicistroviruses.
Table 1

3C-like protease-mediated cleavage sites in dicistroviral capsid protein precursor

Virus (abbreviation)



Acute bee paralysis virus (ABPV)



Kashmir bee virus (KBV)



Israeli acute paralysis virus (IAPV)



Solenopsis invicta virus-1 (SINV-1)



Cricket paralysis virus (CrPV)



Drosophila C virus (DCV)



Aphid lethal paralysis virus (ALPV)



Rhopalosiphum padi virus (RhPV)



Plautia stali intestine virus (PSIV)



Himetobi P virus (HiPV)



Triatoma virus (TrV)



Black queen-cell virus (BQCV)



Homalodisca coagulata virus-1 (HoCV-1)



Taura syndrome virus (TSV)



aCleavage sites shown in parentheses were not experimentally determined but estimated by alignment of deduced amino acid sequences from ORF2 of dicistroviruses

IAPV, KBV, ABPV, and SINV-1 are isolated from hymenopteran insects, and their capsid proteins show preference for cleavage at MQ/I sequences (Table 1). Sequence alignment of the 3C/3D regions of dicistroviruses shows a conserved MQI in the four viruses, suggesting that these are the cleavage sites (Fig. 2a). DCV and CrPV have a single Q residue in the nonconserved region. ALPV and RhPV are isolated from aphids, and these two viruses also have a single Q in the nonconserved region. PSIV and TrV are isolated from pentatomid insects and have Q at a position differing by a single amino acid. BQCV is isolated from the honey bee, but its phylogenetic position suggests that it is more closely related to HiPV rather than to ABPV; indeed, the glutamine residues of BQCV and HiPV are aligned at the same place. TSV seems to prefer cleavage at an H/AG sequence (Table 1). In Fig. 2a, TSV has a single HAG sequence in the nonconserved region. Although the actual 3C/3D cleavage site in some dicistroviruses may be shifted to one or two neighboring Q residues as compared with our estimate, this alignment analysis can pinpoint candidates for the 3C/3D cleavage sites.
Fig. 2

Alignments of deduced dicistroviral amino acid sequences corresponding to the 3C/3D (a) and 2C/3A (b) regions. The position of the first amino acid displayed is indicated in parentheses. Highly conserved amino acids are shown on a black background. Loosely conserved amino acids are shown on a gray background. GXF sequences conserved in CrPV, DCV, ALPV, RhPV, HoCV-1, PSIV, and TrV are marked with dots. The PSIV 3D and 3A regions identified by N-terminal sequence are shown in bold and the estimated N-terminal 3D and 3A regions in other dicistroviruses are shown in gray

Figure 2b shows alignment of the C-terminal region of the 2C protein and the N-terminal region of 3A protein of six dicistroviruses. HiPV has an EQ/VN sequence at the VP2/VP4 cleavage site, and another EQVN sequence appears at the 13-amino-acid C-terminal position of the PSIV 2C/3A cleavage site. The VP3/VP1 cleavage site of TrV is AQVG, and another AQVG is also observed 8 amino acids upstream in comparison to the 2C/3A cleavage site in PSIV. DCV and CrPV use AQ/V or AQ/A sequences for their capsid cleavage. AQVG and AQGG, which are likely to be 3C-mediated cleavage sites, are close to each other. In addition, in BQCV, an AQ sequence is preferred for cleavage, and AQGG is found in the downstream part of the 2C conserved region (Fig. 2b).

Because ABPV, KBV, SINV-1, IAPV and TSV have longer sequences in the C-terminal part of the 2C region, they are not included in Fig. 2b. As mentioned above, however, the four viruses ABPV, KBV, SINV-1 and IAPV have aligned MQ/I or MQ/V sequences in regions corresponding to the 2C/3A cleavage regions. These sequences are 801MQ/I in ABPV, 814MQ/I in KBV, 266MQ/V in SINV-1, and 763MQ/V in IAPV. In the amino acid sequence deduced from TSV ORF1, two occurrences of an HAG sequence, which is used in capsid cleavage (Table 1), are found at amino acid positions 993–995 and 1572–1574. We found that the 1572HAG sequence was aligned by ClustalX [13] to the region of the 3C/3D cleavage sites (Fig. 2a); thus, 993HAG would be the cleavage site for 2C/3A in TSV. Indeed, 993HAG was also aligned at a position close to the 2C/3A cleavage sites of other dicistroviruses (data not shown). Because ALPV, RhPV, and HoCV-1 are unlikely to have the repeated VPg (3B) coding sequences [9], we could not estimate the 3A region of these viruses and therefore they were not included in Fig. 2b.

It has been shown that DCV, CrPV, ABPV, IAPV, and KBV have an NPGP motif, which is known to be associated with 2A/2B cleavage in cardioviruses and aphthoviruses [1]. However, the other dicistroviruses do not have an NPGP sequence. In the case of TSV, sequence similarity to inhibitor of apoptosis proteins was found in the 5′ part of ORF1 [7]; however, such motifs are not included in other dicistroviruses. DCV has a dsRNA-binding motif in the N-terminal region of the P2 precursor [15]. In the case of CrPV, viral suppressors of the host RNAi system have been identified within a region lying 150 codons from the 5′ end of ORF1 [16]. These differences in the N-terminal region of the P2 precursor imply that dicistroviruses have distinct mechanisms for escaping the innate immunity of the host.

IGR-IRES-mediated translation is known to occur during synthesis of the capsid protein, and this pathway is activated under eIF2-phosphorylated conditions [14]. However, the factors that induce shutdown of host protein synthesis in dicistroviruses are unknown. To reveal this mechanism, knowledge of the cleavage pattern of ORF1 products is required. The present results show the similarity of 3C-mediated cleavage sites in groups of dicistroviruses and highlight the dissimilarity of the N-terminal region of the putative P2 precursors in dicistroviruses. Further analyses of the cleavage pattern of the P2 regions in dicistroviruses would contribute to resolution of this issue.

Copyright information

© Springer-Verlag 2008