Rotavirus Viroplasm Proteins Interact with the Cellular SUMOylation System: Implications for Viroplasm-Like Structure Formation

Abstract

Posttranslational modification by SUMO provides functional flexibility to target proteins. Viruses interact extensively with the cellular SUMO modification system in order to improve their replication, and there are numerous examples of viral proteins that are SUMOylated. However, thus far the relevance of SUMOylation for rotavirus replication remains unexplored. In this study, we report that SUMOylation positively regulates rotavirus replication and viral protein production. We show that SUMO can be covalently conjugated to the viroplasm proteins VP1, VP2, NSP2, VP6, and NSP5. In addition, VP1, VP2, and NSP2 can also interact with SUMO in a noncovalent manner. We observed that an NSP5 SUMOylation mutant protein retains most of its activities, such as its interaction with VP1 and NSP2, the formation of viroplasm-like structures after the coexpression with NSP2, and the ability to complement in trans the lack of NSP5 in infected cells. However, this mutant is characterized by a high degree of phosphorylation and is impaired in the formation of viroplasm-like structures when coexpressed with VP2. These results reveal for the first time a positive role for SUMO modification in rotavirus replication, describe the SUMOylation of several viroplasm resident rotavirus proteins, and demonstrate a requirement for NSP5 SUMOylation in the production of viroplasm-like structures.

INTRODUCTION

Rotavirus, a member of the Reoviridae family, is the major etiological cause of severe gastroenteritis of viral origin in infants and young children. The infective virion consists of a nonenveloped triple-layered particle (TLP). Inside the inner layer, composed by pentamers of the structural protein VP2, are contained the 11 double-stranded RNA (dsRNA) segments of the viral genome, the RNA-dependent RNA polymerase VP1, and the RNA capping enzyme VP3, altogether forming the core of the virus. Around the core is present a second intermediate layer, composed by the structural protein VP6, forming a double-layered particle (DLP) that is surrounded by the third outermost layer composed by the proteins VP7 and VP4 forming the fully assembled infectious TLP.

Upon virus entry in the host cell, the outermost layer of the virus is lost and DLPs become active in transcribing the viral mRNA from the dsRNA genome, acting VP1 also as a transcriptase. Even though it has been shown in vitro that the minimal requirement for viral replication is represented by VP1 and VP2 (1, 2), in vivo replication and packaging occur in viral factories, called viroplasms (3). These structures are formed, apart from VP1 and VP2, also by the other structural proteins necessary for the formation of the DLPs, VP3 and VP6, and two nonstructural proteins, NSP2 and NSP5. Both nonstructural proteins are essential for viroplasm formation and virus replication (4–6), but while NSP2 has been proposed to be the molecular motor responsible of the packaging of rotavirus genome in newly synthesized cores (7, 8), the role for NSP5 is less clear. The NSP5 protein, synthesized by the smallest segment of rotavirus genome, has a molecular mass of 26 kDa, a very high content of serine and threonine (25%), and a large number of lysines at its C terminus. NSP5 is posttranslationally modified by O-GlcNAc glycosylation (9) and by extensive phosphorylation that causes, in infected cells, the appearance of a smear of bands that span up to 34 kDa (10, 11). NSP5 is able to interact with the polymerase VP1 and NSP2 both in infected cells and in cotransfection experiments (10, 12). In addition, cotransfection of mammalian cells with NSP5, together with NSP2 or VP2, causes the formation of spherical structures, resembling viroplasms, called viroplasm-like structures (VLS) known as VLS-NSP2i when NSP2-induced and VLS-VP2i when VP2-induced, respectively (13, 14). A recent study has shown that NSP5 is the only viral protein necessary for the formation of VLS and the recruitment of all other viroplasmic proteins to these structures, suggesting a fundamental role for NSP5 in viroplasms formation (13). However, the mechanism through which NSP5 induces the formation of viroplasms has still to be clearly elucidated.

The small ubiquitin-like modifier (SUMO) is a molecule of 11.5 kDa that is covalently bound to lysine residues of target proteins. Usually the target lysine is located in the consensus sequence ψKxE (where ψ is a hydrophobic residue, and x is any residue) (15, 16). However, SUMO can be also conjugated to lysine residues located in nonconsensus sequences. To date, four SUMO isoforms have been discovered in mammals: SUMO1, the most similar to the yeast Smt3; SUMO2 and SUMO3, very similar to one another and characterized by an internal SUMOylation site that allows the formation of SUMO chains; and SUMO4, which has been correlated to diabetes (17–19). SUMOylation regulates a wide range of processes, such as protein stability or nucleocytoplasm transport, but its main function is to regulate protein-protein interactions (20). In addition, an increasing number of SUMOylated proteins can also interact with SUMO in a noncovalent manner, through a SUMO-interacting-motif (SIM) (21).

Viral proteins were among the first substrates shown to be modified by SUMO and SUMOylation seems to facilitate viral infection in cells (22, 23). Although the list of viruses able to exploit the SUMOylation machinery has considerably increased in the last years, the role of SUMO in the replication of members of the Reoviridae family has not been reported thus far. We show here that a change in the levels of SUMOylation machinery components in the cells alters both rotavirus replication and rotavirus protein production. In addition, we demonstrate that rotavirus proteins that localize in viroplasms are SUMOylated and interact with SUMO in a noncovalent manner. Finally, we show that the expression of NSP5 mutated in the SUMOylation sites abolishes the formation of VLS-VP2i. To our knowledge, this is the first demonstration of exploitation of the cellular SUMOylation machinery by a member of the Reoviridae family.

MATERIALS AND METHODS

Cell lines, transfections, and virus.

MA104, HeLa, COS7, and HEK-293 cells were maintained in Dulbecco modified Eagle medium supplemented with 10% fetal calf serum (Gibco), 5 mmol of l-glutamine (Invitrogen)/liter, and penicillin-streptomycin (Invitrogen). HEK-293 cells were transfected using FuGene (Roche), and COS7, HeLa, and MA104 cells were transfected with Lipofectamine 2000, according to the manufacturer's instructions. When stated, cells were infected with T7-recombinant vaccinia virus (strain vTF7.3) at an multiplicity of infection (MOI) of 5 and 1 h later transfected with plasmids DNA for 16 h, as previously described (24). Infection with T7-recombinant vaccinia virus was used to increase the expression level of proteins encoded by the transfected plasmids.

The OSU strain of rotavirus was propagated and titrated in MA104 cells as described previously (25). Rotaviruses were activated by incubation with 10 μg of trypsin per ml at 37°C for 30 min.

Plasmids, siRNAs, and reagents.

Plasmids pcDNA3-NSP5wt, pcDNA3-NSP2, pcDNA3-SV5-VP1, and pcDNA3-VP2 were previously described (12, 14). Lysine-to-arginine mutations were carried out using the QuikChange PCR-based site-directed mutagenesis kit (Stratagene, La Jolla, CA) according to the manufacturer's instructions, pcDNA3-NSP5 plasmid DNA as a template, and the oligonucleotides listed in Table 1. Plasmids pcDNA-His6-SUMO1, pcDNA-His6-SUMO2, and pcDNA-Ubc9 were previously described (26, 27). siNSP5 has been previously described (4). Smart-pool small interfering RNAs (siRNAs) against Ubc9 (siUbc9) and scramble siRNA (siIRR) were purchased from Dharmacon. GST-SENP1 was purchased from Biomol.

Table 1.

Oligonucleotides used in site-directed mutagenesis

Amino acid position(s)	Oligonucleotide
	Orientationa	Sequence (5′–3′)
24 (NSP5-wt siRES)	F	TAAAAATGAATCGTCTTCAACAACGTCAACTCTTTCTGG
	R	CCAGAAAGAGTTGACGTTGTTGAAGACGATTCATTTTTA
24 (NSP5-SUMOmut siRES)	F	TAGAAATGAATCGTCTTCAACAACGTCAACTCTTTCTGG
	R	CCAGAAAGAGTTGACGTTGTTGAAGACGATTCATTTCTA
19	F	CTTCTAGTATCTTTAGAAATGAATCGTCTTCTACAACG
	R	CGTTGTAGAAGACGATTCATTTCTAAAGATACTAGAAG
82	F	CGAATGCAGTTAGGACAAATGCAGACGC
	R	GCGTCTGCATTTGTCCTAACTGCATTCG
133	F	CAATCTCAACTGATAATAGAAAGGAGAAATCCAAG
	R	CTTGGATTTCTCCTTTCTATTATCAGTTGAGATTG
138	F	GGAGAAATCCAGGAAAGATAAAAGTAGG
	R	CCTACTTTTATCTTTCCTGGATTTCTCC
134, 136	F	CAATCTCAACTGATAATAGAAGGGAGAGATCCAGG
	R	CCTGGATCTCTCCCTTCTATTATCAGTTGAGATTG
139, 141, 144	F	CCAGGAGAGATAGAAGTAGGAGACACTACCCG
	R	CGGGTAGTGTCTCCTACTTCTATCTCTCCTGG

^aF, forward; R, reverse.

In vitro SUMO conjugation assay.

In vitro SUMO conjugation assays were performed on [³⁵S]methionine-labeled in vitro-transcribed/translated proteins as described previously (28). Briefly, [³⁵S]methionine-labeled proteins were incubated with E1 in a 10-μl reaction including an ATP regenerating system (50 mM Tris [pH 7.6], 5 mM MgCl₂, 2 mM ATP, 10 mM creatine phosphate, 3.5 U of creatine kinase/ml, and 0.6 U of inorganic pyrophosphatase/ml), 10 μg of SUMO1 or SUMO2, and 600 ng of Ubc9. The reactions were incubated at 30°C for 45 min. After terminating the reactions with sodium dodecyl sulfate (SDS) sample buffer containing β-mercaptoethanol, the reaction products were fractionated by SDS-PAGE and detected by fluorography. The in vitro transcription/translation of proteins was performed by using 1 μg of plasmid DNA and a rabbit reticulocyte-coupled transcription/translation system according to the instructions provided by the manufacturer (Promega).

In vitro deSUMOylation assay.

NSP5-SUMO1 obtained in an in vitro SUMOylation reaction as described above was incubated with 2 μg of GST-SENP1 (Biomol) in 30 μl of reaction buffer containing 50 mM Tris (pH 7.5), 2 mM MgCl₂, and 5 mM β-mercaptoethanol. Reactions were incubated at 37°C for 1 h and terminated with SDS sample buffer containing mercaptoethanol. Reactions products were then fractionated on a 12% SDS-polyacrylamide gel, dried for 1 h, and exposed to X-ray film.

Immunoprecipitation assay.

Cells were lysed in TNN buffer (100 mM Tris-HCl [pH 8], 250 mM NaCl, 0.5% NP-40) at 4°C, centrifuged at 15,800 × g for 5 min, and immunoprecipitated overnight at 4°C after addition of 1 μl of the specified antibody and 50 μl of 50% protein A-Sepharose CL4B beads (GE Healthcare). The beads were then washed four times with TNN buffer and resuspended in 30 μl of SDS-PAGE loading buffer.

Lambda phosphatase treatment.

[³⁵S]methionine-labeled in vitro transcribed/translated NSP5 SUMOmut was incubated with 1 μl of λ-phosphatase (BioLabs) in buffer for lambda-phosphatase treatment (50 mM Tris-HCl [pH 7.5], 100 mM NaCl, 0.1 mM EGTA, 2 mM dithiothreitol, 0.01% Brij 35; New England BioLabs) supplemented with 2 mM MnCl₂. The reaction was incubated for 30 min at 30°C and was stopped with SDS-PAGE loading buffer.

Western blot analysis and antibodies.

Cells were washed in phosphate-buffered saline (PBS), scraped into SDS-PAGE loading buffer, and boiled for 5 min. Proteins of total extracts were separated by SDS-PAGE and transferred onto a nitrocellulose membrane. The membranes were incubated with the following antibodies: anti-NSP2 mouse serum (1:3,000), anti-NSP5 guinea pig serum (1:10,000), anti-VP2 guinea pig serum (1:5,000), and anti-SV5 mouse monoclonal antibody (1:10,000). Signals were detected by using chemiluminescence. Quantification of band intensities was performed by using ImageJ software and normalized by the actin densitometry values.

Immunofluorescence staining.

Cells cultured on coverslips were infected with T7-recombinant vaccinia virus and 1 h later transfected with the indicated plasmids. At 16 h posttransfection, the cells were fixed with 2% paraformaldehyde, permeabilized with 0.2% Triton X-100 in PBS, and blocked with 2% bovine serum albumin (BSA). Cells were incubated with primary antibodies overnight in a moist chamber at 4°C. Coverslips were washed extensively with PBS and further incubated with appropriate Alexa-conjugated secondary antibodies for 1 h at room temperature. Nuclei were stained with DAPI (4′,6′-diamidino-2-phenylindole), and coverslips were mounted with ProLong (Molecular Probes). Mouse anti-NSP5 antibody was used at a dilution of 1:3,000, and anti-VP2 guinea pig serum was used a dilution of 1:3,000. Rabbit anti-SUMO2 antibody (1:200) was obtained from Zymed Laboratories. Secondary Alexa 488-conjugated, Alexa 594-conjugated, and Alexa 555-conjugated antibodies were obtained from Molecular Probes. Analysis of the samples was carried out on a Leica TCS SP5 confocal laser microscope using simultaneous scans to avoid shift between the optical channels. Colocalization analysis was done by calculating the overlap coefficient using the LAS-AF software version 2.0.2 (Leica). Particle size analysis was done by using the ImageJ software. Images were exported by use of Adobe Photoshop version 9.0.2.

Purification of His-tagged conjugates.

The purification of His-tagged conjugates using Ni²⁺-NTA-agarose beads allowing the purification of proteins that are covalently conjugated to SUMO was performed as described previously (29). Briefly, cells were lysed in 4 ml of 6 M guanidinium-HCl, 0.1 M Na₂HPO₄/NaH₂PO₄, and 0.01 M Tris-HCl (pH 8.0), plus 5 mM imidazole and 10 mM β-mercaptoethanol per 75-cm³ flask. Then, lysates were mixed with 50 μl of Ni²⁺-nitrilotriacetic acid-agarose beads prewashed with lysis buffer and incubated for 2 h at room temperature. Beads were successively washed with the following: 6 M guanidinium-HCl, 0.1 M Na₂HPO₄/NaH₂PO₄, and 0.01 M Tris-HCl (pH 8.0), plus 10 mM β-mercaptoethanol; 8 M urea, 0.1 M Na₂HPO₄/NaH₂PO₄, 0.01 M Tris-HCl (pH 8.0), 10 mM β-mercaptoethanol; 8 M urea, 0.1 M Na₂HPO₄/NaH₂PO₄, 0.01 M Tris-HCl (pH 6.3), and 10 mM β-mercaptoethanol (buffer A) plus 0.2% Triton X-100; buffer A; and finally buffer A with 0.1% Triton X-100. After the last wash with buffer A, the beads were eluted with 200 mM imidazole in sample buffer. The eluates were subjected to SDS-PAGE and Western blotting as indicated above.

GST pulldown.

Glutathione S-transferase (GST) pulldown experiments, allowing the identification of noncovalent interaction between the target proteins and SUMO, were performed using [³⁵S]methionine-labeled in vitro-transcribed/translated rotavirus proteins and GST or GST-SUMO1 as described previously (29). Briefly, in vitro-translated [³⁵S]methionine-labeled rotavirus proteins were incubated with GST or GST-SUMO1 bound onto glutathione-Sepharose 4B overnight at 4°C in binding buffer containing 50 mM Tris-HCl (pH 7.8), 150 mM NaCl, 0.5 mM EDTA, 0.1% (vol/vol) Triton X-100, 0.1% (vol/vol) Nonidet P-40, 5 mM MgCl₂, 10% (vol/vol) glycerol, 50 μM ZnCl₂, and protease inhibitor cocktail. The resin was then washed four times with 1 ml of binding buffer, and bound proteins were eluted with sample buffer and heated at 95°C for 5 min. Proteins were separated on SDS-PAGE gels and detected by fluorography.

RESULTS

Modulation of SUMO or Ubc9 levels regulates rotavirus replication.

In order to evaluate whether SUMOylation influenced rotavirus replication, as observed for other viruses, we analyzed the effect of Ubc9 and SUMO1 or Ubc9 and SUMO2 overexpression on rotavirus protein synthesis. HeLa cells were cotransfected with plasmids encoding Ubc9 and His6-SUMO1, Ubc9, and His6-SUMO2 or a control empty vector. At 48 h after transfection cells were infected with OSU strain at an MOI of 5 and 6 h after infection, protein extracts were analyzed by Western blotting with the indicated antibodies. As shown in Fig. 1A, overexpression of SUMO1 or SUMO2 caused an increase in the production of the rotavirus proteins NSP5, NSP2, and VP2. Similar results were observed in COS7 or HEK-293 cells. We then decided to evaluate if overexpression of SUMO had also an effect in rotavirus replication. HeLa cells were transfected with Ubc9 and His6-SUMO1, Ubc9, and His6-SUMO2 or a control vector as described above and, at 48 h after transfection, the cells were infected with OSU at an MOI of 1. At 24 h after infection, the production of infective virus was determined. Overexpression of SUMO1 or SUMO2 caused an increase in the virus production, indicating that SUMO has a positive role in rotavirus replication (Fig. 1B). An increase in virus production in COS7 cells transfected with Ubc9 and SUMO2 was also detected. A positive effect of SUMOylation machinery components on rotavirus replication was indeed confirmed, downmodulating the levels of the E2 SUMO conjugating enzyme Ubc9 using an Ubc9 specific siRNA. HeLa cells were transfected with siRNAs against Ubc9; at 48 h after transfection, the cells were infected with OSU at an MOI of 5 and, at 6 h after infection, the protein extracts were analyzed by Western blotting with the indicated antibodies. Transfection of siRNAs against Ubc9 (siUbc9) caused a clear decrease in the Ubc9 levels, as expected (Fig. 1C). In addition, we observed a concomitant decrease in the production of the structural and nonstructural viral proteins NSP5, NSP2, and VP2 (Fig. 1C). A decrease in rotavirus protein synthesis in cells with reduced Ubc9 levels was also observed in COS7 cells. Moreover, we also evaluated the production of infective particles in HeLa cells transfected with siUbc9 at 24 h after infection. Downmodulation of the Ubc9 levels produced a significant decrease in the yield of infective particles, indicating a positive role for Ubc9 in rotavirus replication (Fig. 1D). Taken together, these results indicate that SUMOylation increases the replication of rotavirus.

Fig 1.

Modulation of SUMO or Ubc9 levels regulates rotavirus replication. (A) Overexpression of Ubc9 and SUMO1 or SUMO2 increases viral protein production. HeLa cells were cotransfected with Ubc9 and His6-SUMO1, Ubc9 and His6-SUMO2, or an empty vector and, at 48 h after transfection, the cells were infected with the OSU strain. At 6 h after infection, the total protein extracts were analyzed with the indicated antibodies. The numbers indicate the band intensities. (B) HeLa cells were cotransfected with Ubc9 and His6-SUMO1, Ubc9 and His6-SUMO2, or an empty vector and, at 48 h after transfection, the cells were infected with OSU at an MOI of 1. At 24 h after infection, quantification of virus titer was assessed. The data represent means ± the standard errors (SE) of three independent experiments. *, P < 0.05 (as determined by the Student t test) compared to pcDNA-transfected cells. (C) HeLa cells were transfected with Ubc9 siRNA or random control siRNA and, 48 h after transfection, infected with OSU at an MOI of 5. At 6 h after infection, total protein extracts were analyzed with the indicated antibodies. (D) HeLa cells were transiently transfected with Ubc9 siRNA or random siRNA (siC) and, at 48 h after transfection, the cells were infected with OSU at an MOI of 1 as described above. At 24 h after infection, quantification of the virus titer was assessed. The data represent means ± the SE of three independent experiments. *, P < 0.05 (as determined by the Student t test) compared to siC-transfected cells.

NSP5 is covalently modified by SUMO.

Rotavirus replication takes place in viroplasms. It has been previously demonstrated that NSP5 is essential for the formation of viroplasms in rotavirus-infected cells, and it is the only protein necessary for the formation of VLS in cotransfection experiments (4, 5, 13). For this reason, to evaluate whether SUMOylation has a role in viroplasm formation, we first investigated whether NSP5 was SUMOylated. We carried out an in vitro SUMOylation assay using [³⁵S]methionine-labeled in vitro-translated NSP5 as a substrate. As expected, NSP5 was detected as a band of around 26 kDa (Fig. 2A, first lane). The addition of SUMO1 to the SUMOylation reaction led to the appearance of a smear of additional higher-molecular-mass bands that correspond to SUMOylated NSP5 (Fig. 2A, second lane). Similarly, when SUMO2 was added to the SUMOylation reaction, we observed a ladder of discreet higher-molecular-mass bands that corresponds to NSP5-SUMO2 (Fig. 2A, third lane). To further demonstrate that the bands observed corresponded to SUMOylated-NSP5, NSP5-SUMO1 obtained as described above, was deSUMOylated by incubation with the catalytic domain of the SUMO-specific protease SENP1 fused to GST (GST-SENP1). The incubation of NSP5-SUMO1 in the presence of SENP1 led to the disappearance of the higher-molecular-mass bands corresponding to NSP5-SUMO1 (Fig. 2B). These results indicate that NSP5 can be SUMOylated in vitro. In order to analyze whether the same is true in transfected cells, HEK-293 cells were infected with T7-vaccinia virus and cotransfected with a plasmid encoding NSP5, Ubc9, His6-SUMO1, or His6-SUMO2, and whole protein extracts or nickel column-purified histidine-tagged proteins were analyzed by Western blotting with an anti-NSP5 antibody. Analysis of the whole-cell extracts revealed that while a single band of the expected molecular mass of 26 kDa corresponding to NSP5 was observed in control cells, an additional band of ∼37 kDa was detected in cells where His6-SUMO1 or His6-SUMO2 were cotransfected with NSP5. Histidine-purified extracts revealed the presence of the additional bands only in cells cotransfected with His6-SUMO1 or His6-SUMO2 (Fig. 2C), confirming that NSP5 is SUMOylated in transfected cells. To further investigate NSP5 SUMOylation, we then evaluated whether NSP5 is also SUMOylated when produced by the virus. HEK-293 cells transfected with Ubc9 and either His6-SUMO1 or His6-SUMO2 were infected with rotavirus at an MOI of 5 and, at 6 h postinfection, we analyzed the whole-cell extracts and histidine-tagged proteins purified on nickel beads by Western blotting with an anti-NSP5 antibody. Analysis of the whole-cell extracts revealed the appearance of a broad NSP5 band in cells transfected with pcDNA, as expected (Fig. 2D). We only detected additional higher-molecular-mass bands in both the whole cells extracts or in histidine-purified extracts of cells cotransfected with His6-SUMO1 or His6-SUMO2 (Fig. 2D). These results demonstrate that NSP5 can be also SUMOylated when it is expressed in the context of viral infection. We also detected SUMOylation of NSP5 in HeLa or COS7 infected cells but to a lesser extent compared to HEK-293 cells. Altogether, these data demonstrate that the rotavirus protein NSP5 is SUMOylated by SUMO1 and SUMO2 in vitro, in transfected cells, and in virus-infected cells.

Fig 2.

Covalent modification of NSP5 by SUMO1 or SUMO2 in vitro, in transfected cells, and in virus-infected cells. (A) [³⁵S]methionine-labeled in vitro-translated NSP5 was used as a substrate in an in vitro SUMOylation assay in the presence of SUMO1 or SUMO2. (B) Deconjugation of SUMO1 from NSP5 by SENP1. (C) HEK-293 cells were infected with T7-vaccinia virus and, at 1 h after infection, cotransfected with NSP5 and an empty vector, with NSP5, Ubc9, and His6-SUMO1, or with NSP5, Ubc9, and His6-SUMO2 as indicated. At 16 h after transfection, the total protein extracts or histidine-purified proteins were analyzed by Western blotting (WB) with an anti-NSP5 antibody. (D) HEK-293 cells were transfected with empty vector, Ubc9 and His6-SUMO1, or Ubc9 and His6-SUMO2 and then infected with OSU at an MOI of 5. At 6 h after infection, total protein extracts or histidine-purified proteins were analyzed by Western blotting with an anti-NSP5 antibody.

VP1, VP2, VP6, and NSP2 are covalently conjugated to SUMO.

To evaluate whether the other viroplasm-located proteins VP1, VP2, VP6, and NSP2 are also able to conjugate SUMO, we carried out an in vitro SUMOylation assay with [³⁵S]methionine-labeled in vitro-translated VP1, VP2, VP6, or NSP2 as substrates. As expected, VP1, VP2, VP6, and NSP2 were detected as bands of around 125, 102, 43, and 35 kDa, respectively (Fig. 3A). The addition of SUMO1 or SUMO2 to the SUMOylation reaction caused the appearance of additional higher-molecular-mass bands that correspond to VP1-SUMO, VP2-SUMO, VP6-SUMO, and NSP2-SUMO, respectively (Fig. 3A). These results indicated that four other components of viroplasms (VP1, VP2, VP6, and NSP2) can be SUMOylated in vitro.

Fig 3.

VP1, VP2, VP6, and NSP2 interact covalently with SUMO1 and SUMO2 in vitro and in cells. (A) [³⁵S]methionine-labeled in vitro-translated VP1, VP2, VP6, or NSP2 were used as substrates in an in vitro SUMOylation assay in the presence of SUMO1 or SUMO2. An asterisk (*) indicates an unspecific band. (B) HEK-293 cells were infected with T7-vaccinia virus and, 1 h after infection, the cells were cotransfected with SV5-VP1, VP2, VP6, or NSP2, together with Ubc9 and His6-SUMO1, Ubc9 and His6-SUMO2, or an empty vector. At 16 h after transfection, total protein extracts and histidine-purified proteins were analyzed by Western blotting with the indicated antibodies. (C) HEK-293 cells were transfected with Ubc9 and His6-SUMO1, Ubc9 and His6-SUMO2, or an empty vector, and at 48 h after transfection the cells were infected with OSU at an MOI of 5. At 6 h after infection, total protein extracts or histidine-purified proteins were then analyzed by Western blotting with the indicated antibodies.

SUMOylation of these proteins in transfected cells was investigated in T7-vaccinia virus-infected HEK-293 cells cotransfected with the plasmid encoding SV5-VP1, VP2, VP6, or NSP2 with a control empty plasmid or with Ubc9 and His6-SUMO1 or Ubc9 and His6-SUMO2. At 16 h after transfection, we analyzed both the whole-protein extracts and the histidine-purified extracts by Western blotting. Cells cotransfected with SV5-VP1 and pcDNA showed one band of ∼125 kDa, as expected (Fig. 3B). When His6-SUMO1 or His6-SUMO2 were cotransfected with SV5-VP1, higher-molecular-weight bands of the expected size corresponding to SUMOylated VP1 were detected in the purified extracts (Fig. 3B). Similar results were obtained with VP2, VP6, and NSP2 (Fig. 3B). In order to evaluate whether VP2 and NSP2 are also SUMOylated when expressed from the virus, we transfected HEK-293 cells with a control empty vector or Ubc9 with His6-SUMO1 or with His6-SUMO2 and, 48 h after transfection, the cells were infected with rotavirus at an MOI of 5. At 6 h after infection, we analyzed the whole-cell extracts and the histidine-purified proteins by Western blotting with anti-VP2 or anti-NSP2 antibodies. As shown in Fig. 3C, upon Ni-agarose purification, one major VP2-SUMO1 conjugate and several VP2-SUMO2 bands were detected in cells transfected with SUMO1 or SUMO2 (Fig. 3C), while for NSP2 we detected only one NSP2-SUMO2 band (Fig. 3C). Altogether, these results demonstrate that relevant components of viroplasms can be SUMOylated.

VP1, VP2, and NSP2 can interact with SUMO in a noncovalent manner.

It has been recently demonstrated that some proteins susceptible to be SUMOylated can also interact with SUMO in a noncovalent manner through a SUMO interacting motif (SIM) (30). In silico analysis of the amino acid sequence of VP1, VP2, and NSP2 proteins revealed the existence of several putative SIMs in their sequences. To verify whether these proteins can interact with SUMO1 in a noncovalent manner, we performed a GST-pulldown assay using [³⁵S]methionine-labeled in vitro-translated VP1, VP2, and NSP2 and either GST or GST-SUMO1. As shown in Fig. 4, the VP1, VP2, and NSP2 proteins were able to interact noncovalently with GST-SUMO1 but not with GST. Altogether, these data demonstrate that the relevant viroplasmic proteins VP1, VP2, and NSP2 also interact with SUMO in a noncovalent manner.

Fig 4.

VP1, VP2, and NSP2 interact with SUMO1 in a noncovalent manner. A pulldown assay of [³⁵S]methionine-labeled in vitro-translated VP1, VP2, or NSP2 with GST-SUMO1 was performed. Inputs (first lane) and GST- or GST-SUMO1-bound extracts (second and third lanes) were loaded on an SDS-PAGE gel and detected by fluorography.

An NSP5 mutant for SUMO conjugation is hyperphosphorylated.

One of the main functions of SUMOylation is represented by regulation of the interaction between proteins (20). It has been recently shown that NSP5 is the protein necessary for the recruitment of the other viroplasmic proteins to form VLS (13). In order to evaluate whether NSP5 SUMOylation could have a role in this process, we first tried to identify the lysine residues that conjugate to SUMO in NSP5. In silico analysis of the amino acid sequence of NSP5 revealed several lysine residues as putative SUMO conjugation sites. We then generated a series of mutants in these residues and evaluated their SUMOylation in vitro. We observed a clear reduction in NSP5 SUMOylation only after mutating to arginine, lysines 19, 82, and all seven present from amino acid residues 133 to 144 in NSP5 (NSP5 SUMOmut). As observed in Fig. 5A, when we performed the in vitro SUMOylation assay with this mutant, we did not detect any band corresponding to NSP5 SUMOmut protein conjugated to SUMO1 or SUMO2. To note, in vitro-translated NSP5 SUMOmut was observed as a band of the expected molecular mass of ∼26 kDa and a smear of higher-molecular-mass bands that suggested the protein was hyperphosphorylated (Fig. 5A). Treatment with lambda phosphatase caused the disappearance of these higher-molecular-mass bands demonstrating that the NSP5 SUMOmut is in fact hyperphosphorylated (Fig. 5B).

Fig 5.

Mutation of the lysine residues in NSP5 that conjugate SUMO leads to the hyperphosphorylation of the viral protein. (A) [³⁵S]methionine-labeled in vitro-translated NSP5 wt or NSP5 SUMOmut were used as substrates in an in vitro SUMOylation assay in the presence of SUMO1 or SUMO2 as indicated. (B) [³⁵S]methionine-labeled in vitro-translated NSP5 SUMOmut was treated with lambda (λ) phosphatase for 30 min. (C) HEK-293 cells were infected with the T7-vaccinia virus and subsequently cotransfected with NSP5 wt or NSP5 SUMOmut and empty vector, Ubc9 and His6-SUMO1, or Ubc9 and His6-SUMO2. At 16 h after infection, total protein extracts and histidine-purified proteins were analyzed by Western blotting with an anti-NSP5 antibody.

To determine whether these lysine residues are also implicated in NSP5 SUMOylation in cells, HEK-293 cells were infected with the T7-vaccinia virus and, 1 h after infection, the cells were cotransfected with NSP5 SUMOmut or the control NSP5 wild type (NSP5 wt), together with Ubc9 and His6-SUMO1 or with Ubc9 and His6-SUMO2. At 16 h after transfection, both whole-cell extracts and histidine-purified proteins were analyzed by Western blotting with anti-NSP5 antibody. A clear reduction in the SUMOylation of NSP5 SUMOmut with both SUMO1 and SUMO2 compared to the NSP5 wt protein was observed (Fig. 5C). Similar results were observed in HeLa and MA104.

Mutation of NSP5 SUMOylation sites alone is not sufficient to modulate rotavirus replication.

Covalent SUMO conjugation may regulate the subcellular localization or stability of the target proteins. We then analyzed whether any of these functions were affected by NSP5 SUMOylation. We did not detect any difference between the subcellular localization or stability of the NSP5 SUMOmut compared to NSP5 wt. We then analyzed the effect of mutating NSP5 SUMOylation sites on rotavirus replication. It has been previously shown that depletion of NSP5 with a specific siRNA completely inhibits viroplasms assembly and virus replication in infected cells. This depletion can be complemented in trans by a construct expressing a NSP5 resistant to the siRNA (4). In order to evaluate whether NSP5 SUMOmut is able to complement in trans NSP5 depletion by RNAi in infected cells, we constructed a plasmid encoding NSP5 SUMOmut (OSU strain) with a point mutation in the sequence recognized by the siRNA sufficient to result in resistance to the RNA interference. We cotransfected HeLa cells with the specific siRNA for OSU-NSP5 (siNSP5), together with the siRNA-resistant NSP5 wt (NSP5 wt siRES) or NSP5 SUMOmut (NSP5 SUMOmut siRES), and at 48 h after transfection, we infected cells with the OSU strain of rotavirus at an MOI of 5. At 12 h after infection, the synthesis of viral proteins was determined. As shown in Fig. 6A, left panel, the plasmids encoding NSP5 wt siRES and SUMOmut siRES were expressed at similar levels. As expected, the siNSP5 caused a strong decrease in the levels of the viral proteins NSP5 and VP2 (Fig. 6A, right panel, second lane). However, both NSP5 wt siRES and NSP5 SUMOmut siRES were capable of complementing the synthesis of the viral proteins at similar levels (Fig. 6A, right panel, third and fourth lanes). In order to evaluate whether the SUMOylation-deficient mutant of NSP5 was able to complement also virus replication, HeLa cells were transfected and subsequently infected with OSU as described above described and, 24 h after infection, the viral supernatants were collected and titrated in MA104 cells. As shown in Fig. 6B, NSP5 wt siRES and NSP5 SUMOmut siRES were equally capable of complementing virus replication. These data indicate that mutation of NSP5 SUMOylation sites alone is not sufficient to modulate rotavirus replication.

Fig 6.

Mutation of NSP5 SUMOylation sites alone is not sufficient to modulate rotavirus replication. (A) HeLa cells were transfected with NSP5 wt siRES or NSP5 SUMOmut siRES and, 48 h after transfection, the protein extracts were analyzed by Western blotting with an anti-NSP5 antibody (left panel). HeLa cells were cotransfected with NSP5 siRNA or control siRNA and with the plasmids pcDNA3-NSP5wt siRES or pcDNA3-NSP5 SUMOmut siRES, as indicated. At 48 h after transfection, the cells were infected with OSU at an MOI of 5 and, 12 h after infection, the total protein extracts were analyzed by Western blotting with the indicated antibodies (right panel). (B) HeLa cells were cotransfected with NSP5 siRNA or control siRNA and with the plasmids pcDNA3-NSP5wt siRES or pcDNA3-NSP5 SUMOmut siRES, as indicated. At 48 h after transfection, the cells were infected with OSU at an MOI of 1 and, at 12 h after infection, quantification of the virus titer was performed. The data represent means ± the SE for a representative experiment.

Formation of VLS-VP2i requires the SUMOylation of NSP5.

Since NSP5 is able to specifically interact with VP1 and NSP2 and to form VLS when coexpressed with NSP2 or with VP2, we decided to evaluate whether NSP5 SUMOylation was involved in these processes. We infected HEK-293 cells with the T7 recombinant vaccinia virus and cotransfected them with SV5-VP1 and NSP5 SUMOmut or NSP5 wt. As shown in Fig. 7A, NSP5 SUMOmut was coimmunoprecipitated by the anti-SV5 antibody with the same efficiency than NSP5 wt, suggesting that the SUMOylation status of NSP5 does not affect its interaction with VP1. Similar results were obtained in coimmunoprecipitation experiments in cells coexpressing NSP2 with NSP5 wt of NSP5 SUMOmut (Fig. 7B). In addition, we evaluated the capability of NSP5 SUMOmut to form VLS-NSP2i or VLS-VP2i. HeLa cells were infected with the T7-vaccinia virus and, 1 h after infection, cotransfected with NSP5 SUMOmut or the control NSP5 wt, together with NSP2 or VP2. At 16 h after transfection, VLS formation was analyzed by immunostaining with the indicated antibodies. VLS produced after transfection of NSP2 and NSP5 SUMOmut were similar to those observed after transfection of NSP5 wt (Fig. 7C). In contrast, the ability of NSP5 SUMOmut to form VLS when coexpressed with VP2 (VLS-VP2i) compared to NSP5 wt was significantly altered, as shown by the smaller and more diffused punctuated structures observed (Fig. 7D). Similar results were observed in MA104 cells. Particle size analysis of at least 300 VLS-VP2i in each transfection revealed that only 7% of the VLS-VP2i detected in MA104 cells transfected with NSP5 SUMOmut had a size larger than 1 μm² in comparison to more than 32% of the NSP5 wt-transfected cells (Fig. 7D). A similar alteration in the formation of VLS-VP2i was also detected in cells transfected with siUbc9 in comparison to those observed in siC-transfected cells (Fig. 7E). In agreement with these data, immunofluorescence analysis of VLS-VP2i and VLS-NSP2i detected in HeLa cells using anti-SUMO2 and anti-NSP5 antibodies revealed a clear colocalization between SUMO2 and NSP5 in VLS-VP2i but not in VLS-NSP2i (Fig. 7F). In addition, analysis of the colocalization between SUMO1 and NSP5 in four images containing at least 80 VLS-NSP2i or VLS-VP2i formed in MA104 cells as described above revealed that SUMO1 and NSP5 colocalized in VLS-VP2i, as shown by an overlap coefficient of 0.78 ± 0.076, but not in VLS-NSP2. Taken together, these data indicate that NSP5 SUMOylation is not required for the interaction with VP1 and NSP2 but plays an important role in the formation of VLS-VP2i.

Fig 7.

NSP5 SUMOylation is required for the formation of VLS-VP2i. (A) HEK-293 cells were infected with the T7-vaccinia virus and, 1 h after infection, the cells were cotransfected with NSP5 wt or NSP5 SUMOmut, together with SV5-VP1. At 16 h after transfection, the protein extracts were immunoprecipitated (IP) with an anti-SV5 monoclonal antibody or an irrelevant control antibody (lanes c). Western blot analysis of the immunoprecipitated proteins with anti-NSP5 or anti-SV5 antibodies was then carried out, as indicated. (B) HEK-293 cells were infected with the T7-vaccinia virus and, 1 h after infection, cotransfected with NSP2, NSP5 wt, or NSP5 SUMOmut. At 16 h after transfection, the protein extracts were cross-linked with DSP and immunoprecipitated with anti-NSP5 antibody or an irrelevant control antibody (lanes c) and analyzed by Western blotting with an anti-NSP2 antibody. (C) HeLa cells were infected with the T7-vaccinia virus and, 1 h after infection, cotransfected with NSP2 and NSP5 wt or NSP5 SUMOmut. At 16 h after transfection, the cells were fixed and stained with anti-NSP5 antibodies and DAPI. The images were analyzed by confocal microscopy. (D) Both HeLa and MA104 cells were infected with the T7-vaccinia virus and, 1 h after infection, cotransfected with VP2 and NSP5 wt or NSP5 SUMOmut. At 16 h after transfection, the cells were fixed and stained with anti-VP2 and anti-NSP5 antibodies. The images were analyzed by confocal microscopy. The top panel shows VLS-VP2i formed in HeLa cells transfected with NSP5 wt or NSP5 SUMOmut. The lower panel shows the size distribution of VLS-VP2i detected in MA104 cells enumerated by percentage of particles in each size interval, obtained after analysis of at least 300 VLS-VP2i per condition using confocal analysis and ImageJ software. (E) HeLa cells were transfected with siC or siUbc9 and, 48 h after transfection, the cells were infected with the T7-vaccinia virus. At 1 h after infection, the cells were cotransfected with NSP5 wt and VP2 and, 16 h after transfection, the cells were fixed and stained with anti-Ubc9 and anti-VP2 antibodies. The images were analyzed by confocal microscopy. (F) HeLa cells were infected with the T7-vaccinia virus and, 1 h after infection, cotransfected with NSP5 wt and NSP2 or VP2. At 16 h after transfection, the cells were fixed and stained with anti-SUMO2 and anti-NSP5 antibodies. The images were analyzed by confocal microscopy.

DISCUSSION

Since its discovery in the mid-1990s, posttranslational modification by SUMO has proven to be a key regulator of protein functions. Viruses interact extensively with SUMO in order to regulate the activity of either cellular or viral proteins. Although most of the studies have been carried out for DNA viruses, RNA viruses have been also shown to interact at various levels with the SUMO pathway (23).

In this study we show that SUMOylation plays a positive role in rotavirus replication. Upregulation of SUMO protein levels had a positive effect on the production of viral proteins and increased rotavirus replication. Similarly, RNA interference of E2 SUMOylation enzyme Ubc9 caused a marked decrease in the production of viral proteins and the virus titer, demonstrating a relevant effect of the SUMOylation system on rotavirus replication. These results are similar to those observed with other RNA viruses where SUMO system showed to be important for virus replication (31–33). To our knowledge, this is the first demonstration of exploitation of the cellular SUMOylation machinery by a member of the Reoviridae family.

Rotavirus replication occurs within highly specialized entities called viroplasms, containing four structural (VP1, VP2, VP3, and VP6) and two nonstructural proteins (NSP2 ad NSP5). Bioinformatics analysis revealed that all of the viroplasm-resident proteins contain one or more putative SUMOylation sites (34) and VP1, VP2, and NSP2 contain one or more putative SIMs. NSP5 has been demonstrated to play a fundamental role in architectural assembly of viroplasms and in recruitment of viroplasmic proteins (4, 5, 13). In addition, NSP5 is thus far the only rotavirus protein that has been shown to cause formation of VLS when coexpressed with either NSP2 or VP2 (13, 14). Since the important role of SUMO in protein-protein interaction, we then speculated that SUMO might have a role in the interaction between viroplasmic resident proteins.

In the present study we demonstrated that rotavirus proteins VP1, VP2, VP6, NSP2, and NSP5 can be SUMOylated both in vitro and in cells and that VP1, VP2, and NSP2 interact with SUMO in a noncovalent manner. Our results demonstrated that the conjugation of SUMO1 to NSP5 led to the appearance of bands with different molecular masses, indicating that NSP5 is modified by SUMO in more than one lysine residue. In silico analysis of NSP5 showed one putative SUMOylation consensus site in lysine 19 and one high-score site in lysine 133. It has been described that lysine selection can be promiscuous, since the mutation of lysine to arginine at preferred SUMO modification sites may result in modification at secondary SUMOylation sites (35, 36). In the case of NSP5, only after the mutation of lysines 19 and 82 and seven lysine residues in the segment 133-144 (i.e., lysines 133, 134, 136, 138, 139, 141, and 144) could we observe a clear decrease in the SUMO conjugation to NSP5 in vitro and in transfected cells.

SUMOmut protein showed a diffuse localization when expressed in cells in the absence of other viral proteins, led to the formation of VLS-NSP2i after cotransfection with NSP2, interacted with VP1 or NSP2 in coimmunoprecipitation experiments, and it is able to complement in trans the NSP5 depletion in infected cells, similar to the NSP5 wt protein. Interestingly, NSP5 SUMOmut showed increased phosphorylation in vitro and in transfected cells, suggesting the possible existence of an interplay between NSP5 SUMOylation and phosphorylation. Extensive studies have been carried out on the function of NSP5 phosphorylation. NSP5 is hyperphosphorylated when produced upon virus infection, and blocking of this phosphorylation causes the formation of viroplasms with altered shape (10, 11, 37). In addition, coexpression of NSP5 with VP2 or NSP2 causes hyperphosphorylation of the protein in the absence of virus infection (13, 14). We did not detect a change in the level of NSP5 phosphorylation after overexpression of SUMO or downmodulation of Ubc9, nor did we detect a change in the levels of NSP5-SUMO conjugation after the coexpression of NSP2, VP2, or VP1. Based on these observations, we cannot conclude that there is a relationship between these two posttranslational modifications.

In a recent report, the polybasic region 132-146 of NSP5 has been shown to be the responsible for its interaction with NSP2 (38). In our study, however, NSP5 SUMOylation was not necessary for the NSP5-NSP2 interaction since NSP5 SUMOmut (with several lysines mutated in the 133-144 region) retained interaction with NSP2. The lack of effect observed can be due to the fact that we did not change the polybasic nature of this region since lysines were mutated to arginines. This result is also confirmed by the fact that no alteration in the shape and number between VLS-NSP2i induced by NSP5 wt or NSP5 SUMOmut could be observed in immunofluorescence.

VLS can be considered a valid model for viroplasm assembly. It has been shown that the viroplasm-resident proteins VP1 and VP6 can be recruited to VLS-VP2i in the absence of viral infection (13). Our results demonstrate that NSP5 SUMOmut is defective in the formation of VLS-VP2i, demonstrating that SUMOylation of NSP5 modulates the interaction with VP2. The notion that SUMO modification plays an essential role in the formation of VLS-VP2i suggests a structural role for this protein in viroplasm assembly. However, our results indicate that SUMOylation of NSP5 is not required for rotavirus replication, probably because of the extensive SUMO interaction of the other viroplasm components. Further studies to evaluate whether inhibiting the SUMO interaction with the other rotavirus proteins leads to a more dramatic effect on rotavirus replication are required.

In summary, these results demonstrate for the first time that rotavirus highjack the SUMOylation machinery of the cell for efficient replication probably through modification of several components of the viroplasms, structures where rotavirus replication take place, providing potential new therapeutic targets for rotavirus infection.