Interaction of Moloney Murine Leukemia Virus Capsi
http://www.100md.com
病菌学杂志 2006年第1期
Howard Hughes Medical Institute, Department of Biochemistry and Molecular Biophysics, College of Physicians and Surgeons, Columbia University, New York, New York 10032
Integrated Program in Cellular, Molecular, and Biophysical Studies, College of Physicians and Surgeons, Columbia University, New York, New York 10032
National Health Research Institutes, Division of Biotechnology and Pharmaceutical Research, Taipai, Taiwan, Republic of China
ABSTRACT
Yeast two-hybrid screens led to the identification of Ubc9 and PIASy, the E2 and E3 small ubiquitin-like modifier (SUMO)-conjugating enzymes, as proteins interacting with the capsid (CA) protein of the Moloney murine leukemia virus. The binding site in CA for Ubc9 was mapped by deletion and alanine-scanning mutagenesis to a consensus motif for SUMOylation at residues 202 to 220, and the binding site for PIASy was mapped to residues 114 to 176, directly centered on the major homology region. Expression of CA and a tagged SUMO-1 protein resulted in covalent transfer of SUMO-1 to CA in vivo. Mutations of lysine residues to arginines near the Ubc9 binding site and mutations at the PIASy binding site reduced or eliminated CA SUMOylation. Introduction of these mutations into the complete viral genome blocked virus replication. The mutants exhibited no defects in the late stages of viral gene expression or virion assembly. Upon infection, the mutant viruses were able to carry out reverse transcription to synthesize normal levels of linear viral DNA but were unable to produce the circular viral DNAs or integrated provirus normally found in the nucleus. The results suggest that the SUMOylation of CA mediated by an interaction with Ubc9 and PIASy is required for early events of infection, after reverse transcription and before nuclear entry and viral DNA integration.
INTRODUCTION
The early events of the retroviral life cycle, especially the intracellular trafficking of the virus, are poorly understood (see references 11, 14, 18, and 51 for reviews). Soon after entry into the cytoplasm of an infected cell, the viral RNA genome is reverse transcribed to give rise to a linear double-stranded DNA. This DNA is present in a large structure termed the preintegration complex (PIC): the composition and structure of the PIC are poorly characterized, but it is likely to contain Gag proteins, reverse transcriptase (RT), integrase (IN), and several host proteins. There is good evidence that the Gag proteins, and especially capsid (CA), are important early in infection. The PIC of the Moloney murine leukemia virus (Mo-MuLV) notably retains CA for many hours (5, 15). The presence of CA in the PIC is consistent with the effects of many gag mutations in the CA domain on early stages of infection (1, 2, 30). CA is even more strongly implicated in early events of infection by its role as a target of the Fv1 gene, a dominant-acting restriction system present in many mouse strains (41, 55). Fv1 encodes a Gag-like protein (4) most closely related to the HERV-L family of endogenous retrovirus proteins and somehow blocks the incoming viral PIC after reverse transcription. The susceptibility of various MuLV isolates to Fv1 restriction is largely determined by a single amino acid, residue 110, of the CA protein (34). Very recently, human cells have been shown to exhibit a similar restriction, first dubbed Ref1 activity, which blocks infection by MuLVs very early after infection, even before reverse transcription (29, 36). TRIM5, a member of the so-called tripartite motif family, has been identified as the protein responsible for this activity (25, 33, 53, 67, 75). Remarkably, sensitivity of the MuLVs to TRIM5 is determined by CA and even by the same residue of the CA protein as for Fv1 (53). These findings strongly suggest that the CA protein remains bound to the incoming viral genome and plays an important role in the early events of infection.
CA may be involved directly in nuclear entry. The simple retroviruses typically require cell division and the associated breakdown of the nuclear membrane for nuclear entry (47, 57), suggesting that the PIC might find its way into the nucleus only by allowing the nuclear membrane to reform around it in the newly formed daughter cells. The lentiviruses, in contrast, can infect nondividing cells (40, 70) and thus have acquired a mechanism for entry into an intact nucleus. The viral proteins and the cellular machinery responsible for both routes of entry are uncertain. Recently, studies with chimeric viruses containing different regions of the viral gag genes of Moloney MuLV and human immunodeficiency virus type 1 (HIV-1) have suggested that the capsid protein may be a key determinant of the difference between these viruses (74).
Protein trafficking often involves covalent modification of the cargo. For example, the covalent attachment of ubiquitin or polyubiquitin chains to lysine residues of selected proteins is an important posttranslational modification that can lead to the protein's degradation or targeted transport to particular intracellular locations. A number of related molecules, including the small ubiquitin-like modifier 1 (SUMO-1; also known as sentrin), are similarly transferred to substrate lysine residues by conjugating enzymes (for review, see references 26, 46, and 49). SUMO-1 addition involves its activation by an E1 enzyme and transfer to an E2-conjugating enzyme, Ubc9, before attachment to the substrate (12, 62) and is often facilitated by one of several E3 ligases (RanBP2 and PIAS1, -3, -x, and -y), which recognize the substrate and confer specificity to Ubc9 (28, 31, 60). SUMO-1 is usually transferred to lysines in a Ubc9 binding site motif of consensus sequence KxE (where is a hydrophobic residue) (48, 61), though lysines in other contexts can sometimes be modified. The functions of SUMO-1 conjugation are not all known but can include nuclear localization (54), intranuclear movement (50), and activation of transcription factors (68). A huge number of proteins have been found to be modified by SUMO-1 addition, including many viral transcriptional regulators and the Mason-Pfizer monkey virus and HIV-1 Gag proteins (19, 71). In this report, we present evidence that the MuLV CA protein interacts with both Ubc9 and PIASy, the E2 and E3 enzymes for SUMO-1 transfer; that CA is SUMOylated in vivo; and that these steps are required for formation of the nuclear viral DNA forms and viral replication.
MATERIALS AND METHODS
Yeast plasmids. The yeast expression vector pNLexA (Origene Technologies, Rockville, MD) and pSH2-1 (20) encode an N-terminal LexA DNA-binding domain (LexADB) and carry the HIS3 marker; the pGADNOT vector encodes a C-terminal Gal4 activation domain (Gal4AD) and carries the LEU2 marker (43). The CA region of the wild-type Mo-MuLV was synthesized by PCR on plasmid pNCS (8) with primers 5'-ACGAATTCAAAATGAAGCCCCTCCGCGCAGGAGGAA-3' and 5'-AGTCGGATCCGCAATAGCTTGCTCATCTCTCTATG-3'. The amplified CA fragment was digested with EcoRI and BamHI and inserted into pNLexA and pSH2-1 to create pMuLVCA-LexADB and pLexADB-MuLVCA constructs, respectively. The plasmid pMuLVGAG-LexADB was prepared by fusion of the Gag coding region of Mo-MuLV to pNLexA. The Gag fragment was amplified by PCR with primers 5'-ATCCGAATTCATGGGCCAGACTGTTACCACTC-3' and 5'-ATCCGGATCCAGTCATCTAGGGTCAGGAGGGA-3'.
The full-length Ubc9 DNA fragment was amplified by PCR with primers 5'-ATCTAGGATCCAAATGTCGGGGATCGCCCTCAGC-3' and 5'-ATCTCTAGTCGACTTATGAGGGGGCAAACTTCTT-3'. The resulting PCR fragment was digested with BamHI and SalI and ligated with pGADNOT DNA cut with the same enzymes to form pGal4AD-Ubc9.
The N-terminal deletion mutants of CA were created by PCR with the fixed primer 5'-AGTCGGATCCGCAATAGCTTGCTCATCTCTCTATG-3' and one of the following primers: dlN1 (5'-CGAATTCAAAATGCTAGTCCACTATCGCCAGTTG-3'), dlN2 (5'-CGAATTCAAAATGAATGTGTCTATGTCTTTCATT-3'), or dlN3 (5'-CGAATTCAAAATGACCCCGGAAGAAAGAGAG-3').
The C-terminal deletion mutants of CA were created by PCR with the fixed primer 5'-AACGAATTCAAAATGAAGCCCCTCCGCGCAGGAGGA-3' and one of the following primers: dlC1 (5'-TCGGATCCGTTCTCGTTTATTAAAGATCTT-3'), dlC2 (5'-TCGGATCCGTTTTTTAAATCTTCTAACCT-3'), dlC3 (5'-TCGGATCCGAGTTTCTTGCCCTGGGTCCTC-3'), dlC4 (5'-TCGGATCCGGCTTCTGCCCGCGTTTTGGAG-3'), dlC5 (5'-TCGGATCCGTTGAGTGGGGCGCCCATCATC-3'), or dlC6 (5'-TCGGATCCGAACAGACTCGATCAGAGCTGT-3').
The various resulting PCR products were digested with EcoRI and BamHI and ligated with the pNLexA vector DNA digested with the same restriction enzymes.
Substitution mutant CA plasmids were constructed by introducing point mutations into the CA region of pMuLVCA-LexADB. Triple-alanine mutations were created by overlap extension PCR with pMuLVCA-LexADB as a template by using outside primers (forward primer in the N terminus of CA, 5'-AACGAATTCAAAATGAAGCCCCTCCGCGCAGGAGGA-3'; and reverse primer in the C terminus of CA, 5'-AGTCGGATCCGCAATAGCTTGCTCATCTCTCTATG-3').
The sense-strand primers utilized for creating mutations were as follows, and the antisense primers were their complements: CA/T1 (5'-AAGTTAGAGAGGGCAGCTGCATTAAAAAACAAGACG-3'), CA/T2 (5'-AGGTTAGAAGATGCAGCTGCAAAGACGCTTGGAGAT-3'), CA/T3 (5'-GATTTAAAAAACGCAGCTGCAGGAGATTTGGTTAGA-3'), CA/T4 (5'-AACAAGACGCTTGCAGCTGCCGTTAGAGAGGCAG-3'), CA/T5 (5'-CTTGGAGATTTGGCAGCTGCCGCAGAAAAGATCTTT-3'), CA/T6 (5'-TTGGTTAGAGAGGCAGCTGCAATCTTTAATAAACGAGA-3'), CA/T7 (5'-GAGGCAGAAAAGGCAGCTGCAAAACGAGAAACCCCGG-3'), CA/T8 (5'-AAGATCTTTAATGCAGCTGCAACCCCGGAAGAAAGAG-3'), CA/T9 (5'-AATAAACGAGAAGCAGCTGCCGAAAGAGAGGAACGT-3'), CA/T10 (5'-GAAACCCCGGAAGCAGCTGCCGAACGTATCAGGA-3'), CA/T11 (5'-GAAGAAAGAGAGGCAGCTGCCAGGAGAGAAACAG-3'), CA/T12 (5'-CTCCAAAACGCGGCAGCTGCCCCCACCAATTTGG-3'), CA/T13 (5'-GCGGGCAGAAGCGCAGCTGCCTTGGCCAAGGTA-3'), CA/T14 (5'-AGCCCCACCAATGCAGCTGCCGTAAAAGGAATAACA-3'), CA/T15 (5'-AATTTGGCCAAGGCAGCTGCCATAACACAAGGGCCC-3'), CA/T16 (5'-GCCAAGGTAAAAGCAGCTGCCCAAGGGCCCAATGAG-3'), CA/T17 (5'-AAAGGAATAACAGCAGCTGCCAATGAGTCTCCCTCG-3'), CA/T18 (5'-CAAGGGCCCAATGCAGCTGCCCCCTCGGCCTTCCTA-3'), CA/T19 (5'-CCCAATGAGTCTGCAGCTGCCTTCCTAGAGAGACTT-3'), CA/T20 (5'-TCTCCCTCGGCCGCAGCTGCCAGACTTAAGGAAGCC-3'), CA/T21 (5'-GCCTTCCTAGAGGCAGCTGCCGAAGCCTATCGCAGG-3'), CA/T22 (5'-AAGGAAGCCTATGCAGCTGCCACTCCTTATGACCCT-3'), CA/T23 (5'-TACACTCCTTATGCAGCTGCCGACCCAGGGCAAGAA-3'), CA/T24 (5'-TATGACCCTGAGGCAGCTGCCCAAGAAACTAATGTG-3'), CA/T25 (5'-GAGGACCCAGGGGCAGCTGCCAATGTGTCTATGTCT-3'), and CA/T26 (5'-GGGCAAGAAACTGCAGCTCTATGTCTTTCATTTGG-3).
All PCR-amplified products were verified by DNA sequence analysis.
The complete open reading frame for mouse PIASy was synthesized by PCR using primers mPIASy/5'BamHI (5'-ACTACTGGATCCACATGGCGGCAGAGCTGGTGGAGGCCAAAAAC-3') and mPIASy/3'Sal (5'-ATCTACTAGTCGACTCAGCACGCGGGCACCAGGCCTTTCTGGAA-3'). The product was cleaved with BamHI plus SalI and cloned into pGADNOT cleaved with the same enzymes.
Yeast strains and two-hybrid library screening. Saccharomyces cerevisiae strain CTY10-5d (MATa ade2 trp1-901 leu2-3,112 his3-200 gal4 gal80 URA3::lexA-lacZ) contains an integrated GAL1-LacZ gene with LexA promoter. The yeast two-hybrid library of mouse cDNAs from WEHI-3B was described previously (3). Transformation of yeast strains and scoring for LacZ expression by 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal) assay on nitrocellulose lifts were performed as described previously (6). Blue transformants were restreaked and retested, and cDNA was recovered and analyzed by sequencing.
The strengths of protein-protein interactions in yeast were qualitatively assessed by monitoring the development of blue color after staining of colony lifts over time. The induction of reporter gene expression was measured quantitatively using a -galactosidase assay of permeabilized yeast grown in liquid cultures with ortho-nitrophenyl--D-galactopyranoside (ONPG) as a substrate (58). Assays were performed with cultures from three independent transformants for each pair of constructs to be tested.
Expression of recombinant Ubc9, PIASy, and CA proteins in bacteria. A mouse Ubc9 cDNA fragment was amplified by PCR from a murine cDNA preparation using the primers 5'-ATCTAGGATCCGCCATGAGTGGGATCGCCCTCAGC-3' and 5'-ATCTAGAATTCTTATGAGGGGGCAAACTTCTT-3'. The resulting Ubc9 DNA was digested with BamHI and EcoRI and cloned in vector pGEX-5X-1 (Pharmacia) to form plasmid pGEX-Ubc9. A GEX-PIASy plasmid encoding a glutathione S-transferase (GST)-PIASy fusion protein was a kind gift of Rudolf Grosschedl at University of Munich. Escherichia coli strain BL21 was transformed with pGEX-Ubc9, pGEX-PIASy, and the empty vector pGEX-6P-1 (as control) and induced by the addition of 1 mM isopropyl--D-thiogalactopyranoside (IPTG) for 4 h. Cells were pelleted and resuspended in 10 ml of lysis buffer (10 mM Tris HCl, pH 8, 150 mM NaCl, 0.1% Triton X-100, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride [PMSF], 1 μg/ml aprotinin, 1 μg/ml leupeptin, 1 μg/ml pepstatin A) containing 1 mg/ml of lysozyme. After incubation on ice for 1 h and brief sonication, cell debris was removed by centrifugation at 30,000 x g for 1 h at 4°C. Five hundred microliters of glutathione-agarose beads (G-beads; 50% slurry) was added, and the mixture was incubated for 1 h at 4°C and extensively washed in the same buffer. The recovery and purity of the bound protein were examined by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).
For CA protein preparation, a DNA fragment containing the CA coding region was synthesized from the wild-type viral genome of pNCA by PCR with primers 5'-ATCCACTACCATGGCTCCCCTCCGCGCAGGAGGAAAC-3' and 5'-GACTACAAGTCGACCGACAATAGCTTGCTCATCTCTCTATG-3', digested with NcoI and SalI, and cloned into the pet21d vector (Novagene) to yield plasmid pet21d-CA. E. coli strain BL21 was transformed with pet21d-CA and induced and lysed by the same method as used for preparation of GST-Ubc9. One milliliter of Ni-nitrilotriacetic acid (NTA) beads (QIAGEN) was used to bind CA protein. The bound CA proteins were eluted with 100 mM imidazole. The purified CA protein was examined by 12% SDS-PAGE.
In vitro binding of CA by GST-Ubc9 and GST-PIASy beads. In a standard protocol for binding of CA by beads containing GST fusion proteins, 20 μl of a 50% (vol/vol) slurry of beads with bound GST, GST-Ubc9, or GST-PIASy (typically containing 2 to 4 μg of protein) was incubated with 100 ng of purified CA protein in 500 μl of TN buffer (50 mM Tris, pH 8, 150 mM NaCl, 0.1% Triton X-100) containing 1 mg/ml of bovine serum albumin (Boehringer). Following incubation for 1 h at 4°C, beads were washed three times in TN and the bound proteins were eluted by 40 mM reduced glutathione and subjected to SDS-PAGE and Western blotting.
Cell culture. 293 cells are human embryonic kidney (HEK) cells that express the E1 region of adenovirus 5. 293T cells are 293 cells that stably express SV40 large T antigen. Rat2-2 cells are rat embryonic fibroblast cells (16). 293, 293T, and Rat2-2 cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum.
Mammalian expression constructs. Plasmid pNCA contains an infectious copy of Mo-MuLV proviral DNA (8); pNCS is a version of pNCA that carries a simian virus 40 (SV40) replication origin in the vector to allow high-level expression in 293T cells.
Mutant plasmids were constructed by introducing point mutations into pNCS. A silent mutation was introduced to create an EcoRI site in the C terminus of CA of pNCS by QuickChange site-directed mutagenesis kit (Stratagene) with the primer 5'-GAAGAAAGAGAGGAACGAATTCGGAGAGAAACAGAG-3'. The resulting construct was named pNCS/Eco.
Point mutations were created by overlap extension PCR with pNCS/Eco template DNA by using outside primers (forward primer in CA, 5'-CGCTTTTCCCCTCGAGCGCCCAGACTGGGAT-3'; reverse primer in CA, 5'-CTCCGAATTCGTTCCTCTCTTTCTTCCGG-3'). The sense-strand primers utilized for creating mutations were as follows, and the antisense primers were their complement. The primers utilized for creating mutations were as follows: M1 (5'-GCATTTGCAAAACGAGAAACCCCGGAAGAA-3'), M2 (5'-ATCGCTGCAAAACGAGAAACCCCGGAAGAA-3', M3 (5'-ATCGCTGCAGCACGAGAAACCCCGGAAGAA-3'), K193R (5'-ATTGGGAGACGATTAGAGAGGTTAGAAGATTTA-3'), K201R (5'-TTAGAAGATTTACGAAACAAGACGCTTGGAGAT-3'), K218R (5'-AAGATCTTTAATCGACGAGAAACCCCGGAAGAA-3'), K3R (5'-CCAGACATTGGGAGAAGATTAGAGAGGTTAGAAGATTTAAGAAACAGAACGCTTGGAGATTTG-3'), and K5R (5'-GACGCTTGGAGATTTGGTTAGAGAGGCAGAACGGATCTTTAATCGA-3').
To create the pCMV2/CA construct, a CA DNA fragment was amplified by PCR with pNCS/Eco template DNA by using primers 5'-AACGAATTCAAAAATGAAGCCCCTCCGCGC-3' and 5'-AGTCGGATCCCTACAATAGCTTGCTCATCTCTCTATG-3'. The PCR products were digested by BamHI and EcoRI and inserted into the pFlag/CMV2 vector (Sigma) cut with the same restriction enzymes. To create mutants K193R, K201R, K218R, K3R, and K5R in the context of pCMV2/CA, two synthetic primers, 5'-AACGAATTCAAAAATGAAGCCC CTCCGCGC-3' and 5'-CTCTGTTTCTCTCCGAATTCGTTCCTCTCTTTCT-3', were used to amplify a portion of CA by PCR with variant proviral KR mutant template DNA. The amplified DNA fragment was digested by EcoRI and inserted into the pCMV2/CA vector cut with EcoRI. All mutants created by PCR were verified by sequencing.
Mammalian cell transfection, viral infections, and virion protein analysis. To examine virus viability, 293T cells were transiently transfected with wild-type (pNCS) or mutant proviral DNAs. Virus was harvested and used to infect naive Rat2-2 cells. Culture supernatants were collected on successive days and monitored for virus production by RT assay. To examine virion proteins, viral particles were harvested from transfected 293T cells and purified by ultracentrifugation, and the proteins were analyzed by Western blotting as described previously (77).
In vivo SUMO conjugation assay. The plasmid for expression of the His-tagged SUMO-1 (pSG5/His-SUMO-1) was the kind gift of Anne Dejean of the Institut Pasteur. 293T cells were grown and cotransfected with wild-type or CA mutant pCMV2/CA and pSG5/His-SUMO-1 plasmids in 140-mm-diameter dishes. Forty-eight hours posttransfection, the transfected cells were harvested and lysed in 1 ml of lysis buffer (6 M guanidinium HCl, 100 mM NaH2PO4, and 10 mM Tris-HCl, pH 7.8). After sonication, 50% of the lysate was incubated with 100 μl of Ni-NTA agarose beads (QIAGEN). The beads were washed twice with washing buffer (pH 7.8) containing 8 M urea, followed by washing with acid buffer (pH 6.3) also containing 8 M urea. The bound proteins were eluted twice with 8 M urea (pH 4.5) containing 20 mM imidazole. Finally, the remaining proteins were eluted with SDS-PAGE sample buffer once. The combination of all eluates from beads was used for SDS-PAGE. A portion of the crude cell lysate was subjected to trichloroacetic acid precipitation and used as a whole-cell extract. The proteins were analyzed by Western blotting using anti-CA antibody.
Analysis of viral DNA synthesized in vivo. Virus was prepared in 293T cells. To minimize contamination of the virus with transfecting DNA, the cells were transfected by Fugene 6 (Roche Applied Science, Indianapolis, IN) using only 1.5 μg of DNA per dish. Forty-eight hours after transfection, the culture supernatant was harvested, assayed for RT as described previously (17, 77), and normalized to equal RT activity. The preparations were treated with DNase I (2 μg/ml; Roche) at 37°C for 1 h to remove residual input DNA and were used to infect nave Rat2-2 cells. Preintegrative viral DNAs were isolated from Rat2-2 cells 18 h postinfection (27) and analyzed by Southern blotting, using a 32P-labeled viral DNA probe. PCR was used to detect circular viral DNA containing two long terminal repeats (LTRs). The primers used to amplify the LTR-LTR junction were MR5784 (5'-AGTCCTCCGATTGACTGAG-3') and MR4091 (5'-CTCTTTTATTGAGCTCGGG-3') (63). The PCR conditions were 94°C for 1 min, 50°C for 1 min, and 72°C for 2 min, repeated for 30 cycles. The primers used to amplify the CA region of all viral DNAs were 5'-CCCCTCCGCGCAGGAGGAAAC-3' and 5'-CCCAGTCTGGGCGCTCGAGGGG-3'. The PCR conditions were 94°C for 1 min, 50°C for 1 min, and 72°C for 45 s, repeated for 24 cycles.
RESULTS
The Moloney MuLV CA protein interacts with Ubc9. Both genetic and biochemical experiments suggest that the MuLV CA protein is involved in the trafficking or nuclear entry of the viral DNA in the early phase of the life cycle. To identify cellular proteins that might interact with CA to mediate these processes, we performed a yeast two-hybrid screen. Initial tests showed that conventional "bait" yeast plasmids expressing LexADB-CA fusions were not functional, as judged by their inability to activate the reporter gene by homodimerization with a Gal4 activation domain fusion to CA. Plasmids expressing a CA-LexADB fusion with an N-terminal CA protein and a C-terminal LexADB, however, were able to strongly dimerize with Gal4AD-CA. This CA-LexADB "bait" was used to screen more than 1 x 106 yeast transformants of a large library of cDNAs from WEHI-3B, a macrophage/monocytic cell line from BALB/c mice (3). All of the three positive clones contained the complete open reading frame of Ubc9, encoding the major SUMO-1 E2-conjugating enzyme (12, 62). The Gal4AD sequences were joined at various positions in the 5' untranslated portion of the Ubc9 sequences, such that translation would proceed in each case into the Ubc9 coding region in the correct reading frame.
Yeast strains were transformed with various pairs of DNAs to determine the specificity of the interaction of Ubc9 and CA (Table 1). Yeast strains expressing Gal4AD-Ubc9 plus CA-LexADB showed strong activation, and those expressing Gal4AD-Ubc9 plus the LexADB-CA fusion showed none, confirming that the C-terminal CA fusion proteins were not functional. Gal4AD-Ubc9 also showed no interaction with a Gag-LexADB fusion protein containing the entire Gag precursor protein, even though the Gag fusion protein could interact with other partners (data not shown). These results suggest that Ubc9 could only interact with mature CA and not with CA in the context of the immature Gag.
To test whether the interaction of Ubc9 with CA could be detected in vitro, recombinant CA was expressed as a histidine-tagged CA protein in bacteria and purified from lysates by binding and elution from nickel-NTA beads. Ubc9 was expressed in bacteria as a GST fusion protein and was recovered from lysates by binding to glutathione-agarose beads. The beads were then used as an affinity matrix for the binding of CA under various conditions, and the bound proteins were eluted and examined by SDS-PAGE followed by Western blotting developed with anticapsid antiserum (Fig. 1A). While GST control beads showed no detectable background binding of CA, the GST-Ubc9 beads efficiently bound CA under physiological salt concentrations (150 mM NaCl) (Fig. 1A). Experiments performed in higher salt concentrations (500 mM) showed no binding, indicating that the interaction is moderately salt sensitive. These results suggest that the interaction is not an artifact of the yeast system and that Ubc9 can bind CA in vitro in the absence of any other components of the transferase machinery.
Identification of binding site for Ubc9 on CA. To localize the portion of the CA protein required for the interaction with Ubc9, a series of deletion mutants of CA were generated, encoding truncated proteins lacking different portions of either the N terminus or C terminus (Fig. 2A). The proteins were expressed as CA-LexADB fusions and tested for reporter activation in concert with the Gal4AD-Ubc9 fusion. CA mutants containing amino acids 202 to 220 showed strong activation, comparable to that of the intact CA protein, while mutants lacking this region showed little or no signal. To define the binding sequences further, a series of substitution mutants of CA were constructed, each encoding a protein with a block of three adjacent amino acids changed to alanine, spanning the essential region. The mutants were expressed as CA-LexADB fusion proteins and tested for interaction with Gal4AD-Ubc9 as before. Most of the mutants still interacted strongly, but mutant T7 (with changes at amino acids 215 to 217) showed significantly weaker reporter activation, and mutant T8 (with changes at amino acids 218 to 220) showed barely detectable activation (Fig. 2B). These results suggest that residues 215 to 220 (sequence IFNKRE) were important for Ubc9 binding. This sequence constitutes a near-perfect match to the known consensus site for Ubc9 binding and SUMOylation of its major targets (sequence KxE).
To extend the qualitative tests for strength of interaction, quantitative assays of the LacZ reporter gene activation for the affected mutants were performed (Fig. 2C). The results confirm the filter lift assays and document the strong effects of the consensus site mutations.
CA mutations altering the Ubc9 binding site and nearby lysines block or reduce virus replication. The above mutational studies suggest that Ubc9 binds to CA residues 215 to 220 in yeast. If this binding takes place in mammalian cells, Ubc9 would be expected to transfer SUMO-1 to one or more nearby lysine residues. Examination of the CA sequence revealed the presence of five lysines close to the binding site, including one within the site itself. To determine if either the binding site or any of these lysines were important for virus replication, a series of substitution mutations were introduced into CA. Mutants M1, M2, and M3 contained alanine substitutions within the binding site, and eight more mutants contained changes of various of the lysines to arginines—either one, two, three, or all five lysines (Fig. 3A).
The various mutations were introduced into CA in the context of a complete infectious proviral DNA clone (pNCS), and the mutant DNAs were then used to transform 293T cells. Culture supernatants were collected 48 h posttransfection, and the yield of virus was estimated by assays for virion-associated RT. All of the mutants produced levels of virus roughly similar to those of the wild type (Fig. 3B). The supernatants were then normalized by RT levels and used to infect permissive Rat2 cells. The culture medium of these cells was collected daily, and the production of virus was monitored by RT assay (Fig. 3B). While wild-type virus was detected quickly, the binding site mutants were all completely replication defective, with no detectable release of virus at any time point. Mutants K193R and K201R and the double mutant K(193, 201)R replicated with kinetics indistinguishable from those of the wild type. Mutant K218R and the double mutants K(193, 218)R and K(201, 218)R all showed a significant delay in replication, with the first detectable virus appearing at day 4 rather than day 1. The identical phenotype shown by these three mutants suggests that K218R was the significant mutation for all three. Finally, the K3R and K5R mutants were almost completely replication defective, showing only very weak signals at the latest time points. These mutants suggest that K218 was important for replication, but that in the absence of K218 the other lysines contributed partial function for virus replication.
CA mutant viruses assemble and release normal levels of mature virion particles. The CA mutations in the Ubc9 binding site and nearby lysine residues could have effects on the expression of the Gag proteins, Gag processing by the viral protease, or the infectivity of the virions. To test for effects on Gag expression and processing, wild-type or mutant proviral DNAs were used to transform 293T cells as before; a plasmid DNA encoding -galactosidase was included as a control for transfection efficiency. After 72 h, both the cells and culture supernatants were collected. The cells were lysed, and -galactosidase assays were used to correct for variation in transfection efficiency. The virions were purified by ultracentrifugation through sucrose and lysed, and the virion proteins were subjected to SDS-PAGE. The Gag proteins were visualized by a Western blot developed with anti-CA antisera (Fig. 3C). The mutants all produced levels of virion-associated Gag proteins comparable to those of the wild-type virus and showed similar extents of proteolytic processing. In each case, the mature p30 CA was the major species, with less Pr65Gag and traces of a 40-kDa intermediate also present. These results suggest that the mutations in and around the SUMOylation consensus motif did not significantly affect virus production.
CA lysine mutant viruses synthesize linear viral DNA but little or no circular DNAs. The ability of the mutant viruses to induce the formation of normal levels of virions suggests that the replication-defective mutants might be blocked during early steps of infection. To determine the position of the block, we tested whether several of the mutants could direct the synthesis of the various viral DNA forms. Virions were harvested from 293T cells after transfection as before and were then used to infect Rat2 cells. At 18 h postinfection, the low-molecular-weight DNA was harvested, separated by agarose gel electrophoresis, and blotted, and the viral DNAs were detected by hybridization with a 32P-labeled Mo-MuLV-specific probe followed by autoradiography. The wild-type virus directed the formation of large amounts of the full-length 8.8-kb linear DNA and much smaller amounts of the two circular DNAs containing one and two copies of the viral LTR. All the mutants synthesized levels of the linear viral DNA comparable to those of the wild type (Fig. 4A). Mutants K193R, K201R, and K218R also made normal levels of the circular DNAs. Mutants K3R and K5R, however, made no detectable circles. Thus, all the mutants were able to enter the cell and carry out reverse transcription normally. The replication-defective viruses were blocked after reverse transcription and before the formation of the nuclear DNA forms. These results suggest that the lysines around the SUMO-1 addition consensus site are critical for a step after reverse transcription.
To confirm these findings, a more sensitive PCR-based assay of viral DNAs was used. The low-molecular-weight DNAs from the infected Rat2 cells were used as templates for PCR with primers specific either for the LTR-LTR junction unique to the two-LTR circular DNA, or for the CA region present on all the viral DNAs. The reactions were initiated with two amounts of input DNA, differing by 10-fold. The products were then analyzed by agarose gel electrophoresis and detected by ethidium stain (Fig. 4B). The wild type showed high levels of the product of the circular DNA, mutant K218R showed lower levels, and mutants K3R and K5R showed no detectable circular DNA. The wild-type and the mutant samples all exhibited similar levels of the linear DNA, with the signal responsive to input DNA amounts. These data confirm the results of the Southern blotting that the lysine mutants are blocked after formation of linear DNA but before circular DNAs.
CA protein is SUMOylated in vivo. The analysis of the Ubc9 binding site and lysine substitution mutants suggests that this portion of CA plays a role early in infection and raises the possibility that SUMO-1 modification of these CA lysines may be required. Direct examination of CA for SUMO-1 addition in the context of the infecting virus, unfortunately, is not feasible due to the very low levels of CA present, even after high-multiplicity infection. To determine if CA can be SUMOylated under any circumstance, we overexpressed CA along with a poly-histidine-tagged SUMO-1 (His-SUMO-1) in 293T cells. After 24 h, lysates were prepared under strongly denaturing conditions and proteins containing His-SUMO-1 were isolated by binding to nickel-agarose beads. The bound proteins were eluted with histidine, heated in sample buffer, and separated by SDS-PAGE, and the presence of CA protein was detected by Western blotting with polyclonal anti-CA antiserum (Fig. 5). Cells expressing both CA and His-SUMO-1 contained high levels of a protein detected by the anti-CA antiserum at 48 kDa, the size expected for CA modified by the addition of a single SUMO-1 moiety (lane 3). Cells expressing His-SUMO-1 alone or CA alone did not contain this product. The formation of this species, binding to Ni-agarose and reactive to CA antiserum, only upon coexpression of the two proteins, strongly suggests CA modification. In all samples, small amounts of free CA were detected, perhaps due to nonspecific binding to the agarose or to cleavage of the His-SUMO-1 after elution from the agarose.
To determine which of the lysine residues of CA were required for addition of His-SUMO-1, various CA mutants were expressed along with His-SUMO-1, and the proteins were recovered on Ni-agarose and analyzed by Western blotting as before (Fig. 5). Single-lysine mutants K193R, K201R, and K218R all showed the formation of wild-type levels of the CA-His-SUMO-1 product. Thus, no single one of these lysines was essential. Mutant K3R, however, showed drastically reduced levels of the addition product, and mutant K5R showed no detectable product. These results suggest that any one of the three to five lysines in the vicinity of the Ubc9 binding site could serve as an addition site, but other lysines outside this region could not. We cannot rule out the possibility, however, that these mutations have effects on the overall folding of CA and indirectly block the modification. Tests for the levels of unmodified CA protein in the crude lysates showed good expression from all mutants, though some (notably those containing K193R) contained significantly less CA protein than the wild type (Fig. 5, bottom). Even in these mutants, CA levels were more than adequate for detection of the addition product (e.g., in K193R).
CA interacts with PIASy, an E3 ligase for SUMO-1 addition. While in some cases the Ubc9 E2-conjugating enzyme can mediate SUMO-1 addition in the absence of an E3 factor, in most cases the reaction utilizes one or another of the SUMO-specific E3 ligases. We have identified one of the known E3 ligases, PIASy (60), as interacting with another Gag-like molecule, the restriction factor Fv1 (manuscript in preparation). We therefore tested whether CA could similarly interact with PIASy in the yeast two-hybrid system. Yeasts expressing Gal4AD-PIASy and CA-LexADB did indeed show strong LacZ reporter activation, while negative controls did not (Table 1). As for Ubc9, only bait constructs containing N-terminal CA, and neither C-terminal CA nor Pr65Gag, could interact with PIASy.
The interactions between the SUMO ligases and their target proteins are often weak and transient, but specific binding can sometimes be observed in vitro. To test if the interaction of PIASy with CA in vitro could be detected, we performed experiments similar to those used for Ubc9 above. Recombinant CA was prepared as before, and PIASy was expressed as a GST fusion protein and recovered by binding to glutathione-agarose beads. The beads were incubated with the purified CA under various conditions, and the bound proteins were eluted and analyzed by SDS-PAGE followed by Western blotting (Fig. 1B). The GST-PIASy beads bound CA protein under physiological salt concentrations, though less strongly than did GST-Ubc9 beads. The binding was not detectable in higher salt concentrations. As before, CA did not bind to the control GST beads. These results suggest that PIASy does indeed bind CA in the absence of other factors in vitro.
To determine the region of CA required for the interaction with PIASy, the series of N- and C-terminal fragments of CA were expressed in yeast as LexADB fusion proteins along with Gal4AD-PIASy, and the yeast strains were scored for reporter activation (Fig. 6A). A single region of CA required for binding was mapped to residues 132 to 173. This site is nonoverlapping with the Ubc9 site, though adjacent and immediately upstream from it, suggesting that Ubc9 and PIASy could bind to CA simultaneously. The PIASy binding site is almost perfectly centered on the major homology region (MHR), the most highly conserved sequence in Gag proteins. To identify individual residues required for the binding, a new series of alanine-scanning substitution mutants were generated, each encoding a protein with a cluster of three adjacent residues changed to alanine. These mutants were tested as CA-LexADB fusion proteins for interaction with PIASy in the yeast two-hybrid assay (Fig. 6B). Seven of the mutants (T13, T14, T16, T18, T22, T24, and T25) showed dramatically reduced binding relative to the wild type; one of these, T16, showed no detectable binding. The disruptive mutations were scattered along the region, but included two (T16 and T18) within the MHR itself. The remaining eight mutants showed normal or even elevated reporter activity. The results suggest that PIASy interacts with CA and that the binding site completely overlaps with the MHR. Importantly, all the constructs interacted normally with Ubc9, indicating that the mutant proteins were stable and functional in yeast and were not grossly misfolded or degraded in response to the mutations (Fig. 6B).
To confirm the results of the qualitative filter lift assays, the level of reporter gene expression was measured by quantitative assays for -galactosidase activity in yeast expressing various of the interaction-defective mutants (Fig. 6C). In all cases, these assays closely supported the results of the filter lift assays.
CA mutants that fail to interact with PIASy show reduced SUMOylation in vivo. If the SUMOylation of CA by Ubc9 is an E3-dependent reaction, there should be a dependence on the binding of the E3 enzyme to CA. To test this possibility, various of the nonbinding alanine substitution mutants were expressed along with His-SUMO-1 in 293T cells as before, and the levels of modified CA-His-SUMO-1 were determined by SDS-PAGE and Western blotting (Fig. 7). Mutants T16 and T24 showed strongly reduced levels of the modified CA product, and mutant T25 showed essentially none. The total levels of CA protein of these mutants, judged from Western blots of the crude lysates, were comparable to the wild type. Several others of the mutant CA proteins, though stable in yeast, were poorly expressed in 293T cells and could not be tested for acceptor activity. Nevertheless, these results with the stable mutant proteins suggest that the interaction of CA with PIASy is important for efficient SUMOylation of CA by Ubc9 and that this is likely an E3-dependent reaction.
DISCUSSION
The experiments described above demonstrate that the MuLV CA protein interacts with the SUMO transferase machinery both in yeast and in vitro and is SUMOylated in vivo and that the residues important for the interaction and modification are required early in infection. The interaction of CA with Ubc9 requires a canonical binding motif (KxE) in the C-terminal third of the molecule. The interaction with PIASy, remarkably, seems to require a sequence directly overlapping with the MHR, a highly conserved motif found on Gag proteins from nearly all the retroviral genera. The relative positions of these two binding sites suggest that Ubc9 and PIASy could bind simultaneously to CA and cooperate to promote SUMO transfer. There seem to be strict requirements beyond the primary binding sites for the interaction of CA with both proteins. Neither interacts with conventional two-hybrid proteins in which the N terminus of CA is blocked by fusion to the LexA DNA binding protein. This may reflect difficulties in folding of the CA protein in this context; structural studies suggest that a free N terminus of CA might well be critical for proper folding, as the amino terminus is bound back into a pocket in the mature CA proteins of several retroviruses (69). The Ubc9 and PIASy proteins also failed to interact with LexA-Gag containing the complete Gag precursor, even though these constructs are fully functional for the readout of Gag-Gag dimerization and interactions of Gag with other partners. Thus, these failures may reflect a genuine inability of Ubc9 and PIASy to access or recognize their CA binding sites when CA is present in the context of the precursor. The ability of Ubc9 and PIASy to bind CA only in its mature, processed, and properly folded form would be consistent with their role in modifying CA in the context of the PIC early in infection.
The binding of PIASy directly on the MHR is particularly intriguing and raises the possibility that this is a critical function for this highly conserved sequence element. The MHR may have several roles in replication. Some mutants of the Rous sarcoma virus and the Mason-Pfizer monkey virus with substitutions at the MHR are affected in virion assembly, and others are specifically affected in early events, perhaps in the timing of removal of CA from the incoming nucleic acid (10, 66). Deletion of the entire MHR in HIV-1 causes subtle effects on Gag-Gag interactions and disrupts virus assembly, but these effects may be due to global changes in Gag structure (56); substitution mutations in the MHR can also affect assembly but not Gag-Gag binding (45). For many retroviruses, the early functions of the MHR might involve its recognition by the E3 ligases for SUMO-1 transfer. It is also not known whether the MHR of other simple viruses, or of the complex viruses, directly interacts with any E3 ligase, nor whether SUMOylation is important for their replication. The extreme conservation of the MHR suggests that the process might be very widespread among the retroviruses. It is not clear whether PIASy is the only E3 ligase that can recognize the Moloney MuLV CA or whether the other known members of the family (PIAS1, -3, and -x) (7, 42, 60) could also contribute. Recently, knockout mice lacking PIASy were generated (59, 73), and the homozygous mutants were found to be able to support at least basic normal murine virus replication (73). Thus, if an E3 activity is required by the virus, another PIAS family member must be able to perform this role.
The residues required for the interaction of CA with Ubc9 in yeast were found to be critical for virus replication. Furthermore, a number of lysine residues in the vicinity of the Ubc9 consensus site were also important for virus viability. Mutation of lysine 218 to arginine in the consensus site alone significantly impaired replication, and mutation of three or five lysines in the region including this residue dramatically or completely abolished viability. The mutations had no apparent effect on virion assembly and release, and the resulting viruses entered cells normally and resulted in the synthesis of normal levels of linear viral DNA. However, these mutants made significantly less or no detectable circular viral DNAs. Thus, Ubc9 binding seemed to be critical for a step temporally close to nuclear entry. This step could be transport of the PIC toward the nucleus; targeting to chromatin or nuclear components upon mitosis; retention of the PIC within the nucleus upon membrane reformation; intranuclear localization of the PIC to particular sites; or release of the DNA ends from the structure of the PIC to permit DNA circularization by host ligases. SUMOylation has been associated with nuclear targeting of several proteins (54, 72), and especially with subnuclear localization of proteins such as the promyelocytic leukemia protein PML to intranuclear bodies (13, 60). Perhaps SUMOylation of CA is similarly involved in localization of the PIC to critical sites where the viral DNA is released from the PIC, such that both circularization and integration can occur.
The results presented here also show that CA expressed in 293T cells can indeed be modified by SUMO-1 addition in vivo. The coexpression of CA and a tagged SUMO-1 resulted in readily detectable levels of a conjugate that contained the tag, was recognized with anti-CA antisera, migrated at the expected size, and survived boiling in SDS. As seen for most other modified proteins, the steady-state levels of the CA-SUMO-1 adduct were low, consistent with the presence of highly active proteases specific for removal of SUMO-1 in most cells (21). The highest level of modification required the PIASy binding site, indicating that this is an E3-dependent reaction. Mutation of three or five lysines near the Ubc9 binding site reduced or eliminated SUMO-1 modification, indicating that one or more of these lysines were absolutely required in this setting and likely were serving as the addition sites. Mutation of single lysines had no effect on the levels of the modification, suggesting that no one lysine served as a unique target of SUMO-1 addition. Mutation K218R in particular had no effect on SUMO-1 addition, even though this single change did reduce virus replication. This discrepancy may be due to the forced overexpression of CA. CA is expressed at much higher levels in these experiments than in a normal infection, and it may not recreate the oligomeric structure of the virion core or the PIC. Thus, the sensitivity of its substrate activity for SUMOylation to mutation may not be identical to that seen in the PIC.
Taken together, these experiments suggest that the binding and modification of CA by Ubc9 and PIASy or related E3 ligases are required for the early events of infection by Moloney MuLV. Although Ubc9 mediates effects on some targets without covalent modification (35), because mutation of lysine residues at the binding site in this case affects replication, it is likely that the actual transfer of SUMO-1 to CA is required. The step affected is close in time to nuclear entry, consistent with the role of SUMOylation of many other substrates. This time is identical to that affected by mutations in another Gag protein, p12, encoded immediately upstream of CA in the Gag precursor (76, 77). The two Gag proteins, p12 and CA, seem to both be required to properly ensure that the PIC enters the nucleus in a functional way early in infection, and indeed these two proteins often need to be derived from the same virus to function properly (39). There is no indication that p12 itself is modified by SUMO-1 or regulates CA modification.
The mechanism of nuclear entry for the MuLVs remains obscure. These viruses, like most other simple retroviruses, do not efficiently infect nondividing cells, and the major block is at the time of nuclear entry (47, 57). In dividing cells, the viruses depend on the breakdown of the nuclear envelope associated with mitosis and are presumably localized inside the nuclei of the daughter cells as their nuclear envelopes reform. The MuLV PICs are thus probably not imported into the nucleus through the activity of nuclear pore complexes, but rather through association with cellular DNA, chromatin, or chromatin-associated proteins. The key step may be the targeting of the PIC to nuclear components or PIC retention in the nucleus. It is also possible that the key step mediated by SUMOylation is after nuclear entry and involves intranuclear targeting to critical sites or bodies within the nucleus. The addition of SUMO-1 to many transcription factors and other target proteins for their intranuclear localization into nuclear bodies is fully consistent with this proposed role for SUMOylation of CA in MuLV infection.
A number of host restriction factors that block early events have been found to be strongly dependent on the particular CA protein present on the incoming virus. The Fv1 gene of mice, which provides a dominant block to infection after reverse transcription but before appearance of the nuclear DNA forms, is a prime example (64). There are two major Fv1 alleles: Fv1n, which blocks so-called B-tropic viruses with glutamic acid at CA residue 110; and Fv1b, which blocks N-tropic viruses with arginine at this position. It is possible that the target of Fv1 action is not only CA but actually the SUMOylation of CA. Preliminary studies from our laboratory indicate that Fv1, like CA, interacts with Ubc9 and PIASy; and thus an attractive model for its mode of action is that it might prevent the required modification of CA. The block by Fv1 is precisely at the same time in the life cycle as the block induced by mutation of the SUMO-1 modification sites, and so inhibition of CA SUMOylation by Fv1 would correctly account for its time of action. In primates, a different restriction factor, the TRIM5 gene product, has recently been shown to block many retroviruses (25, 33, 67, 75) and again its activity depends on the CA protein of the incoming virus (9, 22-24, 52, 53; see references 38 and 65 for reviews). The block by TRIM5 is earlier, before reverse transcription, however, and so is not straightforwardly consistent with affecting SUMOylation.
The involvement of viral proteins with the SUMO conjugation system during infection is not likely to be limited to the MuLVs or even the retroviruses. The Mason-Pfizer monkey virus CA protein has recently been reported to interact with Ubc9, and coexpression of the proteins results in their colocalization to foci in the perinuclear region (71). Overexpression of Ubc9 was found to drive a fraction of the Gag protein into the nucleus. The p6 Gag of HIV-1 has recently been shown to be modified by SUMO-1 addition to lysine K27, though mutation of this residue had no apparent effect on virus replication (19). The nucleocapsid protein (NP) of the Seoul, Hantaan, and Tula hantaviruses, negative-strand RNA viruses, also interacts with Ubc9 and SUMO-1, and this interaction is required for localization of NP to the perinuclear region (32, 37, 44). Thus, many viruses may use the SUMO-1 system for localization near or inside the nuclear membrane. These findings suggest that SUMOylation might be a prime target for intervention by antiviral inhibitors. Such new approaches may significantly expand the range and mechanisms of action of drugs available for antiviral therapy.
ACKNOWLEDGMENTS
We are grateful to Valmik K. Vyas for helpful discussions and critical readings of early drafts. We thank Matthew Evans, Gilda Tachedjian, Scott Hughes, and Barbara Studamire for technical support and advice.
This work was supported by PHS grant R37 CA 30488 from the National Cancer Institute. A.Y. and S.B. are Associates and S.P.G. is an Investigator of the Howard Hughes Medical Institute.
REFERENCES
Alin, K., and S. P. Goff. 1996. Amino acid substitutions in the CA protein of Moloney murine leukemia virus that block early events in infection. Virology 222:339-351.
Auerbach, M. R., C. Shu, A. Kaplan, and I. R. Singh. 2003. Functional characterization of a portion of the Moloney murine leukemia virus gag gene by genetic footprinting. Proc. Natl. Acad. Sci. USA 100:11678-11683.
Bacharach, E., J. Gonsky, K. Alin, M. Orlova, and S. P. Goff. 2000. The carboxy-terminal fragment of nucleolin interacts with the nucleocapsid domain of retroviral Gag proteins and inhibits virion assembly. J. Virol. 74:11027-11039.
Best, S., P. Le Tissier, G. Towers, and J. P. Stoye. 1996. Positional cloning of the mouse retrovirus restriction gene Fv1. Nature 382:826-829.
Bowerman, B., P. O. Brown, J. M. Bishop, and H. E. Varmus. 1989. A nucleoprotein complex mediates the integration of retroviral DNA. Genes Dev. 3:469-478.
Breedon, L., and K. Nasmyth. 1985. Regulation of the HO gene. Cold Spring Harbor Symp. Quant. Biol. 50:643-650.
Chung, C. D., J. Liao, B. Liu, X. Rao, P. Jay, P. Berta, and K. Shuai. 1997. Specific inhibition of Stat3 signal transduction by PIAS3. Science 278:1803-1805.
Colicelli, J., and S. P. Goff. 1988. Sequence and spacing requirements of a retrovirus integration site. J. Mol. Biol. 199:47-59.
Cowan, S., T. Hatziioannou, T. Cunningham, M. A. Muesing, H. G. Gottlinger, and P. D. Bieniasz. 2002. Cellular inhibitors with Fv1-like activity restrict human and simian immunodeficiency virus tropism. Proc. Natl. Acad. Sci. USA 99:11914-11919.
Craven, R. C., A. E. Leure-duPree, R. A. Weldon, Jr., and J. W. Wills. 1995. Genetic analysis of the major homology region of the Rous sarcoma virus Gag protein. J. Virol. 69:4213-4227.
Cullen, B. R. 2001. Journey to the center of the cell. Cell 105:697-700.
Desterro, J. M., J. Thomson, and R. T. Hay. 1997. Ubch9 conjugates SUMO but not ubiquitin. FEBS Lett. 417:297-300.
Duprez, E., A. J. Saurin, J. M. Desterro, V. Lallemand-Breitenbach, K. Howe, M. N. Boddy, E. Solomon, H. de The, R. T. Hay, and P. S. Freemont. 1999. SUMO-1 modification of the acute promyelocytic leukaemia protein PML: implications for nuclear localisation. J. Cell Sci. 112:381-393.
Dvorin, J. D., and M. H. Malim. 2003. Intracellular trafficking of HIV-1 cores: journey to the center of the cell. Curr. Top. Microbiol. Immunol. 281:179-208.
Fassati, A., and S. P. Goff. 1999. Characterization of intracellular reverse transcription complexes of Moloney murine leukemia virus. J. Virol. 73:8919-8925.
Gao, G., and S. P. Goff. 1999. Somatic cell mutants resistant to retrovirus replication: intracellular blocks during the early stages of infection. Mol. Biol. Cell 10:1705-1717.
Goff, S., P. Traktman, and D. Baltimore. 1981. Isolation and properties of Moloney murine leukemia virus mutants: use of a rapid assay for release of virion reverse transcriptase. J. Virol. 38:239-248.
Goff, S. P. 2001. Intracellular trafficking of retroviral genomes during the early phase of infection: viral exploitation of cellular pathways. J. Gene Med. 3:517-528.
Gurer, C., L. Berthoux, and J. Luban. 2005. Covalent modification of human immunodeficiency virus type 1 p6 by SUMO-1. J. Virol. 79:910-917.
Hanes, S. D., and R. Brent. 1989. DNA specificity of the bicoid activator protein is determined by homeodomain recognition helix residue 9. Cell 57:1275-1283.
Hang, J., and M. Dasso. 2002. Association of the human SUMO-1 protease SENP2 with the nuclear pore. J. Biol. Chem. 277:19961-19966.
Hatziioannou, T., S. Cowan, and P. D. Bieniasz. 2004. Capsid-dependent and -independent postentry restriction of primate lentivirus tropism in rodent cells. J. Virol. 78:1006-1011.
Hatziioannou, T., S. Cowan, S. P. Goff, P. D. Bieniasz, and G. J. Towers. 2003. Restriction of multiple divergent retroviruses by Lv1 and Ref1. EMBO J. 22:385-394.
Hatziioannou, T., S. Cowan, U. K. von Schwedler, W. I. Sundquist, and P. D. Bieniasz. 2004. Species-specific tropism determinants in the human immunodeficiency virus type 1 capsid. J. Virol. 78:6005-6012.
Hatziioannou, T., D. Perez-Caballero, A. Yang, S. Cowan, and P. D. Bieniasz. 2004. Retrovirus resistance factors Ref1 and Lv1 are species-specific variants of TRIM5alpha. Proc. Natl. Acad. Sci. USA 101:10774-10779.
Hay, R. T. 2005. SUMO: a history of modification. Mol. Cell 18:1-12.
Hirt, B. 1967. Selective extraction of polyoma DNA from infected mouse cell cultures. J. Mol. Biol. 26:365-369.
Hochstrasser, M. 2001. SP-RING for SUMO: new functions bloom for a ubiquitin-like protein. Cell 107:5-8.
Hofmann, W., D. Schubert, J. LaBonte, L. Munson, S. Gibson, J. Scammell, P. Ferrigno, and J. Sodroski. 1999. Species-specific, postentry barriers to primate immunodeficiency virus infection. J. Virol. 73:10020-10028.
Hsu, H. W., P. Schwartzberg, and S. P. Goff. 1985. Point mutations in the P30 domain of the gag gene of Moloney murine leukemia virus. Virology 142:211-214.
Jackson, P. K. 2001. A new RING for SUMO: wrestling transcriptional responses into nuclear bodies with PIAS family E3 SUMO ligases. Genes Dev. 15:3053-3058.
Kaukinen, P., A. Vaheri, and A. Plyusnin. 2003. Non-covalent interaction between nucleocapsid protein of Tula hantavirus and small ubiquitin-related modifier-1, SUMO-1. Virus Res. 92:37-45.
Keckesova, Z., L. M. Ylinen, and G. J. Towers. 2004. The human and African green monkey TRIM5alpha genes encode Ref1 and Lv1 retroviral restriction factor activities. Proc. Natl. Acad. Sci. USA 101:10780-10785.
Kozak, C. A., and A. Chakraborti. 1996. Single amino acid changes in the murine leukemia virus capsid protein gene define the target of Fv1 resistance. Virology 225:300-305.
Kurtzman, A. L., and N. Schechter. 2001. Ubc9 interacts with a nuclear localization signal and mediates nuclear localization of the paired-like homeobox protein Vsx-1 independent of SUMO-1 modification. Proc. Natl. Acad. Sci. USA 98:5602-5607.
LaBonte, J. A., G. J. Babcock, T. Patel, and J. Sodroski. 2002. Blockade of HIV-1 infection of New World monkey cells occurs primarily at the stage of virus entry. J. Exp. Med. 196:431-445.
Lee, B. H., K. Yoshimatsu, A. Maeda, K. Ochiai, M. Morimatsu, K. Araki, M. Ogino, S. Morikawa, and J. Arikawa. 2003. Association of the nucleocapsid protein of the Seoul and Hantaan hantaviruses with small ubiquitin-like modifier-1-related molecules. Virus Res. 98:83-91.
Lee, K., and V. N. KewalRamani. 2004. In defense of the cell: TRIM5alpha interception of mammalian retroviruses. Proc. Natl. Acad. Sci. USA 101:10496-10497.
Lee, S.-K., K. Nagashima, and W.-S. Hu. 2005. Cooperative effect of Gag proteins p12 and capsid during early events of murine leukemia virus replication. J. Virol. 79:4159-4169.
Lewis, P. F., and M. Emerman. 1994. Passage through mitosis is required for oncoretroviruses but not for the human immunodeficiency virus. J. Virol. 68:510-516.
Lilly, F. 1967. Susceptibility to two strains of Friend leukemia virus in mice. Science 155:461-462.
Liu, B., J. Liao, X. Rao, S. A. Kushner, C. D. Chung, D. D. Chang, and K. Shuai. 1998. Inhibition of Stat1-mediated gene activation by PIAS1. Proc. Natl. Acad. Sci. USA 95:10626-10631.
Luban, J., K. L. Bossolt, E. K. Franke, G. V. Kalpana, and S. P. Goff. 1993. Human immunodeficiency virus type 1 Gag protein binds to cyclophilins A and B. Cell 73:1067-1078.
Maeda, A., B. H. Lee, K. Yoshimatsu, M. Saijo, I. Kurane, J. Arikawa, and S. Morikawa. 2003. The intracellular association of the nucleocapsid protein (NP) of Hantaan virus (HTNV) with small ubiquitin-like modifier-1 (SUMO-1) conjugating enzyme 9 (Ubc9). Virology 305:288-297.
Mammano, F., . hagen, S. Hglund, and H. G. Gttlinger. 1994. Role of the major homology region of human immunodeficiency virus type 1 in virion morphogenesis. J. Virol. 68:4927-4936.
Melchior, F. 2000. SUMO—nonclassical ubiquitin. Annu. Rev. Cell Dev. Biol. 16:591-626.
Miller, D. G., M. A. Adam, and A. D. Miller. 1990. Gene transfer by retrovirus vectors occurs only in cells that are actively replicating at the time of infection. Mol. Cell. Biol. 10:4239-4242.
Minty, A., X. Dumont, M. Kaghad, and D. Caput. 2000. Covalent modification of p73alpha by SUMO-1. Two-hybrid screening with p73 identifies novel SUMO-1-interacting proteins and a SUMO-1 interaction motif. J. Biol. Chem. 275:36316-36323.
Muller, S., C. Hoege, G. Pyrowolakis, and S. Jentsch. 2001. SUMO, ubiquitin's mysterious cousin. Nat. Rev. Mol. Cell Biol. 2:202-210.
Muller, S., M. J. Matunis, and A. Dejean. 1998. Conjugation with the ubiquitin-related modifier SUMO-1 regulates the partitioning of PML within the nucleus. EMBO J. 17:61-70.
Nisole, S., and A. Saib. 2004. Early steps of retrovirus replicative cycle. Retrovirology 1:9.
Owens, C. M., P. C. Yang, H. Gttlinger, and J. Sodroski. 2003. Human and simian immunodeficiency virus capsid proteins are major viral determinants of early, postentry replication blocks in simian cells. J. Virol. 77:726-731.
Perron, M. J., M. Stremlau, B. Song, W. Ulm, R. C. Mulligan, and J. Sodroski. 2004. TRIM5alpha mediates the postentry block to N-tropic murine leukemia viruses in human cells. Proc. Natl. Acad. Sci. USA 101:11827-11832.
Pichler, A., and F. Melchior. 2002. Ubiquitin-related modifier SUMO1 and nucleocytoplasmic transport. Traffic 3:381-387.
Pincus, T., W. P. Rowe, and F. Lilly. 1971. A major genetic locus affecting resistance to infection with murine leukemia viruses. II. Apparent identity to a major locus described for resistance to Friend murine leukemia virus. J. Exp. Med. 133:1234-1241.
Provitera, P., A. Goff, A. Harenberg, F. Bouamr, C. Carter, and S. Scarlata. 2001. Role of the major homology region in assembly of HIV-1 Gag. Biochemistry 40:5565-5572.
Roe, T., T. C. Reynolds, G. Yu, and P. O. Brown. 1993. Integration of murine leukemia virus DNA depends on mitosis. EMBO J. 12:2099-2108.
Rose, M. D., F. Winston, and P. Hieter. 1990. Methods in yeast genetics: a laboratory course manual. Cold Spring Harbor Laboratory Press, Plainview, N.Y.
Roth, W., C. Sustmann, M. Kieslinger, A. Gilmozzi, D. Irmer, E. Kremmer, C. Turck, and R. Grosschedl. 2004. PIASy-deficient mice display modest defects in IFN and Wnt signaling. J. Immunol. 173:6189-6199.
Sachdev, S., L. Bruhn, H. Sieber, A. Pichler, F. Melchior, and R. Grosschedl. 2001. PIASy, a nuclear matrix-associated SUMO E3 ligase, represses LEF1 activity by sequestration into nuclear bodies. Genes Dev. 15:3088-3103.
Sampson, D. A., M. Wang, and M. J. Matunis. 2001. The small ubiquitin-like modifier-1 (SUMO-1) consensus sequence mediates Ubc9 binding and is essential for SUMO-1 modification. J. Biol. Chem. 276:21664-21669.
Schwarz, S. E., K. Matuschewski, D. Liakopoulos, M. Scheffner, and S. Jentsch. 1998. The ubiquitin-like proteins SMT3 and SUMO-1 are conjugated by the UBC9 E2 enzyme. Proc. Natl. Acad. Sci. USA 95:560-564.
Smith, C. M., W. B. Potts III, J. S. Smith, and M. J. Roth. 1997. RNase H cleavage of tRNAPro mediated by M-MuLV and HIV-1 reverse transcriptases. Virology 229:437-446.
Stoye, J. P. 1998. Fv1, the mouse retrovirus resistance gene. Rev. Sci. Tech. 17:269-277.
Stoye, J. P. 2002. An intracellular block to primate lentivirus replication. Proc. Natl. Acad. Sci. USA 99:11549-11551.
Strambio-de Castillia, C., and E. Hunter. 1992. Mutational analysis of the major homology region of Mason-Pfizer monkey virus by use of saturation mutagenesis. J. Virol. 66:7021-7032.
Stremlau, M., C. M. Owens, M. J. Perron, M. Kiessling, P. Autissier, and J. Sodroski. 2004. The cytoplasmic body component TRIM5alpha restricts HIV-1 infection in Old World monkeys. Nature 427:848-853.
Verger, A., J. Perdomo, and M. Crossley. 2003. Modification with SUMO. A role in transcriptional regulation. EMBO Rep. 4:137-142.
von Schwedler, U. K., T. L. Stemmler, V. Y. Klishko, S. Li, K. H. Albertine, D. R. Davis, and W. I. Sundquist. 1998. Proteolytic refolding of the HIV-1 capsid protein amino-terminus facilitates viral core assembly. EMBO J. 17:1555-1568.
Weinberg, J. B., T. J. Matthews, B. R. Cullen, and M. H. Malim. 1991. Productive human immunodeficiency virus type 1 (HIV-1) infection of nonproliferating human monocytes. J. Exp. Med. 174:1477-1482.
Weldon, R. A., Jr., P. Sarkar, S. M. Brown, and S. K. Weldon. 2003. Mason-Pfizer monkey virus Gag proteins interact with the human sumo conjugating enzyme, hUbc9. Virology 314:62-73.
Wilson, V. G., and D. Rangasamy. 2001. Intracellular targeting of proteins by sumoylation. Exp. Cell Res. 271:57-65.
Wong, K. A., R. Kim, H. Christofk, J. Gao, G. Lawson, and H. Wu. 2004. Protein inhibitor of activated STAT Y (PIASy) and a splice variant lacking exon 6 enhance sumoylation but are not essential for embryogenesis and adult life. Mol. Cell. Biol. 24:5577-5586.
Yamashita, M., and M. Emerman. 2004. Capsid is a dominant determinant of retrovirus infectivity in nondividing cells. J. Virol. 78:5670-5678.
Yap, M. W., S. Nisole, C. Lynch, and J. P. Stoye. 2004. Trim5alpha protein restricts both HIV-1 and murine leukemia virus. Proc. Natl. Acad. Sci. USA 101:10786-10791.
Yuan, B., A. Fassati, A. Yueh, and S. P. Goff. 2002. Characterization of Moloney murine leukemia virus p12 mutants blocked during early events of infection. J. Virol. 76:10801-10810.
Yuan, B., X. Li, and S. P. Goff. 1999. Mutations altering the Moloney murine leukemia virus p12 Gag protein affect virion production and early events of the virus life cycle. EMBO J. 18:4700-4710.(Andrew Yueh, Juliana Leun)
Integrated Program in Cellular, Molecular, and Biophysical Studies, College of Physicians and Surgeons, Columbia University, New York, New York 10032
National Health Research Institutes, Division of Biotechnology and Pharmaceutical Research, Taipai, Taiwan, Republic of China
ABSTRACT
Yeast two-hybrid screens led to the identification of Ubc9 and PIASy, the E2 and E3 small ubiquitin-like modifier (SUMO)-conjugating enzymes, as proteins interacting with the capsid (CA) protein of the Moloney murine leukemia virus. The binding site in CA for Ubc9 was mapped by deletion and alanine-scanning mutagenesis to a consensus motif for SUMOylation at residues 202 to 220, and the binding site for PIASy was mapped to residues 114 to 176, directly centered on the major homology region. Expression of CA and a tagged SUMO-1 protein resulted in covalent transfer of SUMO-1 to CA in vivo. Mutations of lysine residues to arginines near the Ubc9 binding site and mutations at the PIASy binding site reduced or eliminated CA SUMOylation. Introduction of these mutations into the complete viral genome blocked virus replication. The mutants exhibited no defects in the late stages of viral gene expression or virion assembly. Upon infection, the mutant viruses were able to carry out reverse transcription to synthesize normal levels of linear viral DNA but were unable to produce the circular viral DNAs or integrated provirus normally found in the nucleus. The results suggest that the SUMOylation of CA mediated by an interaction with Ubc9 and PIASy is required for early events of infection, after reverse transcription and before nuclear entry and viral DNA integration.
INTRODUCTION
The early events of the retroviral life cycle, especially the intracellular trafficking of the virus, are poorly understood (see references 11, 14, 18, and 51 for reviews). Soon after entry into the cytoplasm of an infected cell, the viral RNA genome is reverse transcribed to give rise to a linear double-stranded DNA. This DNA is present in a large structure termed the preintegration complex (PIC): the composition and structure of the PIC are poorly characterized, but it is likely to contain Gag proteins, reverse transcriptase (RT), integrase (IN), and several host proteins. There is good evidence that the Gag proteins, and especially capsid (CA), are important early in infection. The PIC of the Moloney murine leukemia virus (Mo-MuLV) notably retains CA for many hours (5, 15). The presence of CA in the PIC is consistent with the effects of many gag mutations in the CA domain on early stages of infection (1, 2, 30). CA is even more strongly implicated in early events of infection by its role as a target of the Fv1 gene, a dominant-acting restriction system present in many mouse strains (41, 55). Fv1 encodes a Gag-like protein (4) most closely related to the HERV-L family of endogenous retrovirus proteins and somehow blocks the incoming viral PIC after reverse transcription. The susceptibility of various MuLV isolates to Fv1 restriction is largely determined by a single amino acid, residue 110, of the CA protein (34). Very recently, human cells have been shown to exhibit a similar restriction, first dubbed Ref1 activity, which blocks infection by MuLVs very early after infection, even before reverse transcription (29, 36). TRIM5, a member of the so-called tripartite motif family, has been identified as the protein responsible for this activity (25, 33, 53, 67, 75). Remarkably, sensitivity of the MuLVs to TRIM5 is determined by CA and even by the same residue of the CA protein as for Fv1 (53). These findings strongly suggest that the CA protein remains bound to the incoming viral genome and plays an important role in the early events of infection.
CA may be involved directly in nuclear entry. The simple retroviruses typically require cell division and the associated breakdown of the nuclear membrane for nuclear entry (47, 57), suggesting that the PIC might find its way into the nucleus only by allowing the nuclear membrane to reform around it in the newly formed daughter cells. The lentiviruses, in contrast, can infect nondividing cells (40, 70) and thus have acquired a mechanism for entry into an intact nucleus. The viral proteins and the cellular machinery responsible for both routes of entry are uncertain. Recently, studies with chimeric viruses containing different regions of the viral gag genes of Moloney MuLV and human immunodeficiency virus type 1 (HIV-1) have suggested that the capsid protein may be a key determinant of the difference between these viruses (74).
Protein trafficking often involves covalent modification of the cargo. For example, the covalent attachment of ubiquitin or polyubiquitin chains to lysine residues of selected proteins is an important posttranslational modification that can lead to the protein's degradation or targeted transport to particular intracellular locations. A number of related molecules, including the small ubiquitin-like modifier 1 (SUMO-1; also known as sentrin), are similarly transferred to substrate lysine residues by conjugating enzymes (for review, see references 26, 46, and 49). SUMO-1 addition involves its activation by an E1 enzyme and transfer to an E2-conjugating enzyme, Ubc9, before attachment to the substrate (12, 62) and is often facilitated by one of several E3 ligases (RanBP2 and PIAS1, -3, -x, and -y), which recognize the substrate and confer specificity to Ubc9 (28, 31, 60). SUMO-1 is usually transferred to lysines in a Ubc9 binding site motif of consensus sequence KxE (where is a hydrophobic residue) (48, 61), though lysines in other contexts can sometimes be modified. The functions of SUMO-1 conjugation are not all known but can include nuclear localization (54), intranuclear movement (50), and activation of transcription factors (68). A huge number of proteins have been found to be modified by SUMO-1 addition, including many viral transcriptional regulators and the Mason-Pfizer monkey virus and HIV-1 Gag proteins (19, 71). In this report, we present evidence that the MuLV CA protein interacts with both Ubc9 and PIASy, the E2 and E3 enzymes for SUMO-1 transfer; that CA is SUMOylated in vivo; and that these steps are required for formation of the nuclear viral DNA forms and viral replication.
MATERIALS AND METHODS
Yeast plasmids. The yeast expression vector pNLexA (Origene Technologies, Rockville, MD) and pSH2-1 (20) encode an N-terminal LexA DNA-binding domain (LexADB) and carry the HIS3 marker; the pGADNOT vector encodes a C-terminal Gal4 activation domain (Gal4AD) and carries the LEU2 marker (43). The CA region of the wild-type Mo-MuLV was synthesized by PCR on plasmid pNCS (8) with primers 5'-ACGAATTCAAAATGAAGCCCCTCCGCGCAGGAGGAA-3' and 5'-AGTCGGATCCGCAATAGCTTGCTCATCTCTCTATG-3'. The amplified CA fragment was digested with EcoRI and BamHI and inserted into pNLexA and pSH2-1 to create pMuLVCA-LexADB and pLexADB-MuLVCA constructs, respectively. The plasmid pMuLVGAG-LexADB was prepared by fusion of the Gag coding region of Mo-MuLV to pNLexA. The Gag fragment was amplified by PCR with primers 5'-ATCCGAATTCATGGGCCAGACTGTTACCACTC-3' and 5'-ATCCGGATCCAGTCATCTAGGGTCAGGAGGGA-3'.
The full-length Ubc9 DNA fragment was amplified by PCR with primers 5'-ATCTAGGATCCAAATGTCGGGGATCGCCCTCAGC-3' and 5'-ATCTCTAGTCGACTTATGAGGGGGCAAACTTCTT-3'. The resulting PCR fragment was digested with BamHI and SalI and ligated with pGADNOT DNA cut with the same enzymes to form pGal4AD-Ubc9.
The N-terminal deletion mutants of CA were created by PCR with the fixed primer 5'-AGTCGGATCCGCAATAGCTTGCTCATCTCTCTATG-3' and one of the following primers: dlN1 (5'-CGAATTCAAAATGCTAGTCCACTATCGCCAGTTG-3'), dlN2 (5'-CGAATTCAAAATGAATGTGTCTATGTCTTTCATT-3'), or dlN3 (5'-CGAATTCAAAATGACCCCGGAAGAAAGAGAG-3').
The C-terminal deletion mutants of CA were created by PCR with the fixed primer 5'-AACGAATTCAAAATGAAGCCCCTCCGCGCAGGAGGA-3' and one of the following primers: dlC1 (5'-TCGGATCCGTTCTCGTTTATTAAAGATCTT-3'), dlC2 (5'-TCGGATCCGTTTTTTAAATCTTCTAACCT-3'), dlC3 (5'-TCGGATCCGAGTTTCTTGCCCTGGGTCCTC-3'), dlC4 (5'-TCGGATCCGGCTTCTGCCCGCGTTTTGGAG-3'), dlC5 (5'-TCGGATCCGTTGAGTGGGGCGCCCATCATC-3'), or dlC6 (5'-TCGGATCCGAACAGACTCGATCAGAGCTGT-3').
The various resulting PCR products were digested with EcoRI and BamHI and ligated with the pNLexA vector DNA digested with the same restriction enzymes.
Substitution mutant CA plasmids were constructed by introducing point mutations into the CA region of pMuLVCA-LexADB. Triple-alanine mutations were created by overlap extension PCR with pMuLVCA-LexADB as a template by using outside primers (forward primer in the N terminus of CA, 5'-AACGAATTCAAAATGAAGCCCCTCCGCGCAGGAGGA-3'; and reverse primer in the C terminus of CA, 5'-AGTCGGATCCGCAATAGCTTGCTCATCTCTCTATG-3').
The sense-strand primers utilized for creating mutations were as follows, and the antisense primers were their complements: CA/T1 (5'-AAGTTAGAGAGGGCAGCTGCATTAAAAAACAAGACG-3'), CA/T2 (5'-AGGTTAGAAGATGCAGCTGCAAAGACGCTTGGAGAT-3'), CA/T3 (5'-GATTTAAAAAACGCAGCTGCAGGAGATTTGGTTAGA-3'), CA/T4 (5'-AACAAGACGCTTGCAGCTGCCGTTAGAGAGGCAG-3'), CA/T5 (5'-CTTGGAGATTTGGCAGCTGCCGCAGAAAAGATCTTT-3'), CA/T6 (5'-TTGGTTAGAGAGGCAGCTGCAATCTTTAATAAACGAGA-3'), CA/T7 (5'-GAGGCAGAAAAGGCAGCTGCAAAACGAGAAACCCCGG-3'), CA/T8 (5'-AAGATCTTTAATGCAGCTGCAACCCCGGAAGAAAGAG-3'), CA/T9 (5'-AATAAACGAGAAGCAGCTGCCGAAAGAGAGGAACGT-3'), CA/T10 (5'-GAAACCCCGGAAGCAGCTGCCGAACGTATCAGGA-3'), CA/T11 (5'-GAAGAAAGAGAGGCAGCTGCCAGGAGAGAAACAG-3'), CA/T12 (5'-CTCCAAAACGCGGCAGCTGCCCCCACCAATTTGG-3'), CA/T13 (5'-GCGGGCAGAAGCGCAGCTGCCTTGGCCAAGGTA-3'), CA/T14 (5'-AGCCCCACCAATGCAGCTGCCGTAAAAGGAATAACA-3'), CA/T15 (5'-AATTTGGCCAAGGCAGCTGCCATAACACAAGGGCCC-3'), CA/T16 (5'-GCCAAGGTAAAAGCAGCTGCCCAAGGGCCCAATGAG-3'), CA/T17 (5'-AAAGGAATAACAGCAGCTGCCAATGAGTCTCCCTCG-3'), CA/T18 (5'-CAAGGGCCCAATGCAGCTGCCCCCTCGGCCTTCCTA-3'), CA/T19 (5'-CCCAATGAGTCTGCAGCTGCCTTCCTAGAGAGACTT-3'), CA/T20 (5'-TCTCCCTCGGCCGCAGCTGCCAGACTTAAGGAAGCC-3'), CA/T21 (5'-GCCTTCCTAGAGGCAGCTGCCGAAGCCTATCGCAGG-3'), CA/T22 (5'-AAGGAAGCCTATGCAGCTGCCACTCCTTATGACCCT-3'), CA/T23 (5'-TACACTCCTTATGCAGCTGCCGACCCAGGGCAAGAA-3'), CA/T24 (5'-TATGACCCTGAGGCAGCTGCCCAAGAAACTAATGTG-3'), CA/T25 (5'-GAGGACCCAGGGGCAGCTGCCAATGTGTCTATGTCT-3'), and CA/T26 (5'-GGGCAAGAAACTGCAGCTCTATGTCTTTCATTTGG-3).
All PCR-amplified products were verified by DNA sequence analysis.
The complete open reading frame for mouse PIASy was synthesized by PCR using primers mPIASy/5'BamHI (5'-ACTACTGGATCCACATGGCGGCAGAGCTGGTGGAGGCCAAAAAC-3') and mPIASy/3'Sal (5'-ATCTACTAGTCGACTCAGCACGCGGGCACCAGGCCTTTCTGGAA-3'). The product was cleaved with BamHI plus SalI and cloned into pGADNOT cleaved with the same enzymes.
Yeast strains and two-hybrid library screening. Saccharomyces cerevisiae strain CTY10-5d (MATa ade2 trp1-901 leu2-3,112 his3-200 gal4 gal80 URA3::lexA-lacZ) contains an integrated GAL1-LacZ gene with LexA promoter. The yeast two-hybrid library of mouse cDNAs from WEHI-3B was described previously (3). Transformation of yeast strains and scoring for LacZ expression by 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal) assay on nitrocellulose lifts were performed as described previously (6). Blue transformants were restreaked and retested, and cDNA was recovered and analyzed by sequencing.
The strengths of protein-protein interactions in yeast were qualitatively assessed by monitoring the development of blue color after staining of colony lifts over time. The induction of reporter gene expression was measured quantitatively using a -galactosidase assay of permeabilized yeast grown in liquid cultures with ortho-nitrophenyl--D-galactopyranoside (ONPG) as a substrate (58). Assays were performed with cultures from three independent transformants for each pair of constructs to be tested.
Expression of recombinant Ubc9, PIASy, and CA proteins in bacteria. A mouse Ubc9 cDNA fragment was amplified by PCR from a murine cDNA preparation using the primers 5'-ATCTAGGATCCGCCATGAGTGGGATCGCCCTCAGC-3' and 5'-ATCTAGAATTCTTATGAGGGGGCAAACTTCTT-3'. The resulting Ubc9 DNA was digested with BamHI and EcoRI and cloned in vector pGEX-5X-1 (Pharmacia) to form plasmid pGEX-Ubc9. A GEX-PIASy plasmid encoding a glutathione S-transferase (GST)-PIASy fusion protein was a kind gift of Rudolf Grosschedl at University of Munich. Escherichia coli strain BL21 was transformed with pGEX-Ubc9, pGEX-PIASy, and the empty vector pGEX-6P-1 (as control) and induced by the addition of 1 mM isopropyl--D-thiogalactopyranoside (IPTG) for 4 h. Cells were pelleted and resuspended in 10 ml of lysis buffer (10 mM Tris HCl, pH 8, 150 mM NaCl, 0.1% Triton X-100, 5 mM EDTA, 1 mM phenylmethylsulfonyl fluoride [PMSF], 1 μg/ml aprotinin, 1 μg/ml leupeptin, 1 μg/ml pepstatin A) containing 1 mg/ml of lysozyme. After incubation on ice for 1 h and brief sonication, cell debris was removed by centrifugation at 30,000 x g for 1 h at 4°C. Five hundred microliters of glutathione-agarose beads (G-beads; 50% slurry) was added, and the mixture was incubated for 1 h at 4°C and extensively washed in the same buffer. The recovery and purity of the bound protein were examined by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).
For CA protein preparation, a DNA fragment containing the CA coding region was synthesized from the wild-type viral genome of pNCA by PCR with primers 5'-ATCCACTACCATGGCTCCCCTCCGCGCAGGAGGAAAC-3' and 5'-GACTACAAGTCGACCGACAATAGCTTGCTCATCTCTCTATG-3', digested with NcoI and SalI, and cloned into the pet21d vector (Novagene) to yield plasmid pet21d-CA. E. coli strain BL21 was transformed with pet21d-CA and induced and lysed by the same method as used for preparation of GST-Ubc9. One milliliter of Ni-nitrilotriacetic acid (NTA) beads (QIAGEN) was used to bind CA protein. The bound CA proteins were eluted with 100 mM imidazole. The purified CA protein was examined by 12% SDS-PAGE.
In vitro binding of CA by GST-Ubc9 and GST-PIASy beads. In a standard protocol for binding of CA by beads containing GST fusion proteins, 20 μl of a 50% (vol/vol) slurry of beads with bound GST, GST-Ubc9, or GST-PIASy (typically containing 2 to 4 μg of protein) was incubated with 100 ng of purified CA protein in 500 μl of TN buffer (50 mM Tris, pH 8, 150 mM NaCl, 0.1% Triton X-100) containing 1 mg/ml of bovine serum albumin (Boehringer). Following incubation for 1 h at 4°C, beads were washed three times in TN and the bound proteins were eluted by 40 mM reduced glutathione and subjected to SDS-PAGE and Western blotting.
Cell culture. 293 cells are human embryonic kidney (HEK) cells that express the E1 region of adenovirus 5. 293T cells are 293 cells that stably express SV40 large T antigen. Rat2-2 cells are rat embryonic fibroblast cells (16). 293, 293T, and Rat2-2 cells were cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum.
Mammalian expression constructs. Plasmid pNCA contains an infectious copy of Mo-MuLV proviral DNA (8); pNCS is a version of pNCA that carries a simian virus 40 (SV40) replication origin in the vector to allow high-level expression in 293T cells.
Mutant plasmids were constructed by introducing point mutations into pNCS. A silent mutation was introduced to create an EcoRI site in the C terminus of CA of pNCS by QuickChange site-directed mutagenesis kit (Stratagene) with the primer 5'-GAAGAAAGAGAGGAACGAATTCGGAGAGAAACAGAG-3'. The resulting construct was named pNCS/Eco.
Point mutations were created by overlap extension PCR with pNCS/Eco template DNA by using outside primers (forward primer in CA, 5'-CGCTTTTCCCCTCGAGCGCCCAGACTGGGAT-3'; reverse primer in CA, 5'-CTCCGAATTCGTTCCTCTCTTTCTTCCGG-3'). The sense-strand primers utilized for creating mutations were as follows, and the antisense primers were their complement. The primers utilized for creating mutations were as follows: M1 (5'-GCATTTGCAAAACGAGAAACCCCGGAAGAA-3'), M2 (5'-ATCGCTGCAAAACGAGAAACCCCGGAAGAA-3', M3 (5'-ATCGCTGCAGCACGAGAAACCCCGGAAGAA-3'), K193R (5'-ATTGGGAGACGATTAGAGAGGTTAGAAGATTTA-3'), K201R (5'-TTAGAAGATTTACGAAACAAGACGCTTGGAGAT-3'), K218R (5'-AAGATCTTTAATCGACGAGAAACCCCGGAAGAA-3'), K3R (5'-CCAGACATTGGGAGAAGATTAGAGAGGTTAGAAGATTTAAGAAACAGAACGCTTGGAGATTTG-3'), and K5R (5'-GACGCTTGGAGATTTGGTTAGAGAGGCAGAACGGATCTTTAATCGA-3').
To create the pCMV2/CA construct, a CA DNA fragment was amplified by PCR with pNCS/Eco template DNA by using primers 5'-AACGAATTCAAAAATGAAGCCCCTCCGCGC-3' and 5'-AGTCGGATCCCTACAATAGCTTGCTCATCTCTCTATG-3'. The PCR products were digested by BamHI and EcoRI and inserted into the pFlag/CMV2 vector (Sigma) cut with the same restriction enzymes. To create mutants K193R, K201R, K218R, K3R, and K5R in the context of pCMV2/CA, two synthetic primers, 5'-AACGAATTCAAAAATGAAGCCC CTCCGCGC-3' and 5'-CTCTGTTTCTCTCCGAATTCGTTCCTCTCTTTCT-3', were used to amplify a portion of CA by PCR with variant proviral KR mutant template DNA. The amplified DNA fragment was digested by EcoRI and inserted into the pCMV2/CA vector cut with EcoRI. All mutants created by PCR were verified by sequencing.
Mammalian cell transfection, viral infections, and virion protein analysis. To examine virus viability, 293T cells were transiently transfected with wild-type (pNCS) or mutant proviral DNAs. Virus was harvested and used to infect naive Rat2-2 cells. Culture supernatants were collected on successive days and monitored for virus production by RT assay. To examine virion proteins, viral particles were harvested from transfected 293T cells and purified by ultracentrifugation, and the proteins were analyzed by Western blotting as described previously (77).
In vivo SUMO conjugation assay. The plasmid for expression of the His-tagged SUMO-1 (pSG5/His-SUMO-1) was the kind gift of Anne Dejean of the Institut Pasteur. 293T cells were grown and cotransfected with wild-type or CA mutant pCMV2/CA and pSG5/His-SUMO-1 plasmids in 140-mm-diameter dishes. Forty-eight hours posttransfection, the transfected cells were harvested and lysed in 1 ml of lysis buffer (6 M guanidinium HCl, 100 mM NaH2PO4, and 10 mM Tris-HCl, pH 7.8). After sonication, 50% of the lysate was incubated with 100 μl of Ni-NTA agarose beads (QIAGEN). The beads were washed twice with washing buffer (pH 7.8) containing 8 M urea, followed by washing with acid buffer (pH 6.3) also containing 8 M urea. The bound proteins were eluted twice with 8 M urea (pH 4.5) containing 20 mM imidazole. Finally, the remaining proteins were eluted with SDS-PAGE sample buffer once. The combination of all eluates from beads was used for SDS-PAGE. A portion of the crude cell lysate was subjected to trichloroacetic acid precipitation and used as a whole-cell extract. The proteins were analyzed by Western blotting using anti-CA antibody.
Analysis of viral DNA synthesized in vivo. Virus was prepared in 293T cells. To minimize contamination of the virus with transfecting DNA, the cells were transfected by Fugene 6 (Roche Applied Science, Indianapolis, IN) using only 1.5 μg of DNA per dish. Forty-eight hours after transfection, the culture supernatant was harvested, assayed for RT as described previously (17, 77), and normalized to equal RT activity. The preparations were treated with DNase I (2 μg/ml; Roche) at 37°C for 1 h to remove residual input DNA and were used to infect nave Rat2-2 cells. Preintegrative viral DNAs were isolated from Rat2-2 cells 18 h postinfection (27) and analyzed by Southern blotting, using a 32P-labeled viral DNA probe. PCR was used to detect circular viral DNA containing two long terminal repeats (LTRs). The primers used to amplify the LTR-LTR junction were MR5784 (5'-AGTCCTCCGATTGACTGAG-3') and MR4091 (5'-CTCTTTTATTGAGCTCGGG-3') (63). The PCR conditions were 94°C for 1 min, 50°C for 1 min, and 72°C for 2 min, repeated for 30 cycles. The primers used to amplify the CA region of all viral DNAs were 5'-CCCCTCCGCGCAGGAGGAAAC-3' and 5'-CCCAGTCTGGGCGCTCGAGGGG-3'. The PCR conditions were 94°C for 1 min, 50°C for 1 min, and 72°C for 45 s, repeated for 24 cycles.
RESULTS
The Moloney MuLV CA protein interacts with Ubc9. Both genetic and biochemical experiments suggest that the MuLV CA protein is involved in the trafficking or nuclear entry of the viral DNA in the early phase of the life cycle. To identify cellular proteins that might interact with CA to mediate these processes, we performed a yeast two-hybrid screen. Initial tests showed that conventional "bait" yeast plasmids expressing LexADB-CA fusions were not functional, as judged by their inability to activate the reporter gene by homodimerization with a Gal4 activation domain fusion to CA. Plasmids expressing a CA-LexADB fusion with an N-terminal CA protein and a C-terminal LexADB, however, were able to strongly dimerize with Gal4AD-CA. This CA-LexADB "bait" was used to screen more than 1 x 106 yeast transformants of a large library of cDNAs from WEHI-3B, a macrophage/monocytic cell line from BALB/c mice (3). All of the three positive clones contained the complete open reading frame of Ubc9, encoding the major SUMO-1 E2-conjugating enzyme (12, 62). The Gal4AD sequences were joined at various positions in the 5' untranslated portion of the Ubc9 sequences, such that translation would proceed in each case into the Ubc9 coding region in the correct reading frame.
Yeast strains were transformed with various pairs of DNAs to determine the specificity of the interaction of Ubc9 and CA (Table 1). Yeast strains expressing Gal4AD-Ubc9 plus CA-LexADB showed strong activation, and those expressing Gal4AD-Ubc9 plus the LexADB-CA fusion showed none, confirming that the C-terminal CA fusion proteins were not functional. Gal4AD-Ubc9 also showed no interaction with a Gag-LexADB fusion protein containing the entire Gag precursor protein, even though the Gag fusion protein could interact with other partners (data not shown). These results suggest that Ubc9 could only interact with mature CA and not with CA in the context of the immature Gag.
To test whether the interaction of Ubc9 with CA could be detected in vitro, recombinant CA was expressed as a histidine-tagged CA protein in bacteria and purified from lysates by binding and elution from nickel-NTA beads. Ubc9 was expressed in bacteria as a GST fusion protein and was recovered from lysates by binding to glutathione-agarose beads. The beads were then used as an affinity matrix for the binding of CA under various conditions, and the bound proteins were eluted and examined by SDS-PAGE followed by Western blotting developed with anticapsid antiserum (Fig. 1A). While GST control beads showed no detectable background binding of CA, the GST-Ubc9 beads efficiently bound CA under physiological salt concentrations (150 mM NaCl) (Fig. 1A). Experiments performed in higher salt concentrations (500 mM) showed no binding, indicating that the interaction is moderately salt sensitive. These results suggest that the interaction is not an artifact of the yeast system and that Ubc9 can bind CA in vitro in the absence of any other components of the transferase machinery.
Identification of binding site for Ubc9 on CA. To localize the portion of the CA protein required for the interaction with Ubc9, a series of deletion mutants of CA were generated, encoding truncated proteins lacking different portions of either the N terminus or C terminus (Fig. 2A). The proteins were expressed as CA-LexADB fusions and tested for reporter activation in concert with the Gal4AD-Ubc9 fusion. CA mutants containing amino acids 202 to 220 showed strong activation, comparable to that of the intact CA protein, while mutants lacking this region showed little or no signal. To define the binding sequences further, a series of substitution mutants of CA were constructed, each encoding a protein with a block of three adjacent amino acids changed to alanine, spanning the essential region. The mutants were expressed as CA-LexADB fusion proteins and tested for interaction with Gal4AD-Ubc9 as before. Most of the mutants still interacted strongly, but mutant T7 (with changes at amino acids 215 to 217) showed significantly weaker reporter activation, and mutant T8 (with changes at amino acids 218 to 220) showed barely detectable activation (Fig. 2B). These results suggest that residues 215 to 220 (sequence IFNKRE) were important for Ubc9 binding. This sequence constitutes a near-perfect match to the known consensus site for Ubc9 binding and SUMOylation of its major targets (sequence KxE).
To extend the qualitative tests for strength of interaction, quantitative assays of the LacZ reporter gene activation for the affected mutants were performed (Fig. 2C). The results confirm the filter lift assays and document the strong effects of the consensus site mutations.
CA mutations altering the Ubc9 binding site and nearby lysines block or reduce virus replication. The above mutational studies suggest that Ubc9 binds to CA residues 215 to 220 in yeast. If this binding takes place in mammalian cells, Ubc9 would be expected to transfer SUMO-1 to one or more nearby lysine residues. Examination of the CA sequence revealed the presence of five lysines close to the binding site, including one within the site itself. To determine if either the binding site or any of these lysines were important for virus replication, a series of substitution mutations were introduced into CA. Mutants M1, M2, and M3 contained alanine substitutions within the binding site, and eight more mutants contained changes of various of the lysines to arginines—either one, two, three, or all five lysines (Fig. 3A).
The various mutations were introduced into CA in the context of a complete infectious proviral DNA clone (pNCS), and the mutant DNAs were then used to transform 293T cells. Culture supernatants were collected 48 h posttransfection, and the yield of virus was estimated by assays for virion-associated RT. All of the mutants produced levels of virus roughly similar to those of the wild type (Fig. 3B). The supernatants were then normalized by RT levels and used to infect permissive Rat2 cells. The culture medium of these cells was collected daily, and the production of virus was monitored by RT assay (Fig. 3B). While wild-type virus was detected quickly, the binding site mutants were all completely replication defective, with no detectable release of virus at any time point. Mutants K193R and K201R and the double mutant K(193, 201)R replicated with kinetics indistinguishable from those of the wild type. Mutant K218R and the double mutants K(193, 218)R and K(201, 218)R all showed a significant delay in replication, with the first detectable virus appearing at day 4 rather than day 1. The identical phenotype shown by these three mutants suggests that K218R was the significant mutation for all three. Finally, the K3R and K5R mutants were almost completely replication defective, showing only very weak signals at the latest time points. These mutants suggest that K218 was important for replication, but that in the absence of K218 the other lysines contributed partial function for virus replication.
CA mutant viruses assemble and release normal levels of mature virion particles. The CA mutations in the Ubc9 binding site and nearby lysine residues could have effects on the expression of the Gag proteins, Gag processing by the viral protease, or the infectivity of the virions. To test for effects on Gag expression and processing, wild-type or mutant proviral DNAs were used to transform 293T cells as before; a plasmid DNA encoding -galactosidase was included as a control for transfection efficiency. After 72 h, both the cells and culture supernatants were collected. The cells were lysed, and -galactosidase assays were used to correct for variation in transfection efficiency. The virions were purified by ultracentrifugation through sucrose and lysed, and the virion proteins were subjected to SDS-PAGE. The Gag proteins were visualized by a Western blot developed with anti-CA antisera (Fig. 3C). The mutants all produced levels of virion-associated Gag proteins comparable to those of the wild-type virus and showed similar extents of proteolytic processing. In each case, the mature p30 CA was the major species, with less Pr65Gag and traces of a 40-kDa intermediate also present. These results suggest that the mutations in and around the SUMOylation consensus motif did not significantly affect virus production.
CA lysine mutant viruses synthesize linear viral DNA but little or no circular DNAs. The ability of the mutant viruses to induce the formation of normal levels of virions suggests that the replication-defective mutants might be blocked during early steps of infection. To determine the position of the block, we tested whether several of the mutants could direct the synthesis of the various viral DNA forms. Virions were harvested from 293T cells after transfection as before and were then used to infect Rat2 cells. At 18 h postinfection, the low-molecular-weight DNA was harvested, separated by agarose gel electrophoresis, and blotted, and the viral DNAs were detected by hybridization with a 32P-labeled Mo-MuLV-specific probe followed by autoradiography. The wild-type virus directed the formation of large amounts of the full-length 8.8-kb linear DNA and much smaller amounts of the two circular DNAs containing one and two copies of the viral LTR. All the mutants synthesized levels of the linear viral DNA comparable to those of the wild type (Fig. 4A). Mutants K193R, K201R, and K218R also made normal levels of the circular DNAs. Mutants K3R and K5R, however, made no detectable circles. Thus, all the mutants were able to enter the cell and carry out reverse transcription normally. The replication-defective viruses were blocked after reverse transcription and before the formation of the nuclear DNA forms. These results suggest that the lysines around the SUMO-1 addition consensus site are critical for a step after reverse transcription.
To confirm these findings, a more sensitive PCR-based assay of viral DNAs was used. The low-molecular-weight DNAs from the infected Rat2 cells were used as templates for PCR with primers specific either for the LTR-LTR junction unique to the two-LTR circular DNA, or for the CA region present on all the viral DNAs. The reactions were initiated with two amounts of input DNA, differing by 10-fold. The products were then analyzed by agarose gel electrophoresis and detected by ethidium stain (Fig. 4B). The wild type showed high levels of the product of the circular DNA, mutant K218R showed lower levels, and mutants K3R and K5R showed no detectable circular DNA. The wild-type and the mutant samples all exhibited similar levels of the linear DNA, with the signal responsive to input DNA amounts. These data confirm the results of the Southern blotting that the lysine mutants are blocked after formation of linear DNA but before circular DNAs.
CA protein is SUMOylated in vivo. The analysis of the Ubc9 binding site and lysine substitution mutants suggests that this portion of CA plays a role early in infection and raises the possibility that SUMO-1 modification of these CA lysines may be required. Direct examination of CA for SUMO-1 addition in the context of the infecting virus, unfortunately, is not feasible due to the very low levels of CA present, even after high-multiplicity infection. To determine if CA can be SUMOylated under any circumstance, we overexpressed CA along with a poly-histidine-tagged SUMO-1 (His-SUMO-1) in 293T cells. After 24 h, lysates were prepared under strongly denaturing conditions and proteins containing His-SUMO-1 were isolated by binding to nickel-agarose beads. The bound proteins were eluted with histidine, heated in sample buffer, and separated by SDS-PAGE, and the presence of CA protein was detected by Western blotting with polyclonal anti-CA antiserum (Fig. 5). Cells expressing both CA and His-SUMO-1 contained high levels of a protein detected by the anti-CA antiserum at 48 kDa, the size expected for CA modified by the addition of a single SUMO-1 moiety (lane 3). Cells expressing His-SUMO-1 alone or CA alone did not contain this product. The formation of this species, binding to Ni-agarose and reactive to CA antiserum, only upon coexpression of the two proteins, strongly suggests CA modification. In all samples, small amounts of free CA were detected, perhaps due to nonspecific binding to the agarose or to cleavage of the His-SUMO-1 after elution from the agarose.
To determine which of the lysine residues of CA were required for addition of His-SUMO-1, various CA mutants were expressed along with His-SUMO-1, and the proteins were recovered on Ni-agarose and analyzed by Western blotting as before (Fig. 5). Single-lysine mutants K193R, K201R, and K218R all showed the formation of wild-type levels of the CA-His-SUMO-1 product. Thus, no single one of these lysines was essential. Mutant K3R, however, showed drastically reduced levels of the addition product, and mutant K5R showed no detectable product. These results suggest that any one of the three to five lysines in the vicinity of the Ubc9 binding site could serve as an addition site, but other lysines outside this region could not. We cannot rule out the possibility, however, that these mutations have effects on the overall folding of CA and indirectly block the modification. Tests for the levels of unmodified CA protein in the crude lysates showed good expression from all mutants, though some (notably those containing K193R) contained significantly less CA protein than the wild type (Fig. 5, bottom). Even in these mutants, CA levels were more than adequate for detection of the addition product (e.g., in K193R).
CA interacts with PIASy, an E3 ligase for SUMO-1 addition. While in some cases the Ubc9 E2-conjugating enzyme can mediate SUMO-1 addition in the absence of an E3 factor, in most cases the reaction utilizes one or another of the SUMO-specific E3 ligases. We have identified one of the known E3 ligases, PIASy (60), as interacting with another Gag-like molecule, the restriction factor Fv1 (manuscript in preparation). We therefore tested whether CA could similarly interact with PIASy in the yeast two-hybrid system. Yeasts expressing Gal4AD-PIASy and CA-LexADB did indeed show strong LacZ reporter activation, while negative controls did not (Table 1). As for Ubc9, only bait constructs containing N-terminal CA, and neither C-terminal CA nor Pr65Gag, could interact with PIASy.
The interactions between the SUMO ligases and their target proteins are often weak and transient, but specific binding can sometimes be observed in vitro. To test if the interaction of PIASy with CA in vitro could be detected, we performed experiments similar to those used for Ubc9 above. Recombinant CA was prepared as before, and PIASy was expressed as a GST fusion protein and recovered by binding to glutathione-agarose beads. The beads were incubated with the purified CA under various conditions, and the bound proteins were eluted and analyzed by SDS-PAGE followed by Western blotting (Fig. 1B). The GST-PIASy beads bound CA protein under physiological salt concentrations, though less strongly than did GST-Ubc9 beads. The binding was not detectable in higher salt concentrations. As before, CA did not bind to the control GST beads. These results suggest that PIASy does indeed bind CA in the absence of other factors in vitro.
To determine the region of CA required for the interaction with PIASy, the series of N- and C-terminal fragments of CA were expressed in yeast as LexADB fusion proteins along with Gal4AD-PIASy, and the yeast strains were scored for reporter activation (Fig. 6A). A single region of CA required for binding was mapped to residues 132 to 173. This site is nonoverlapping with the Ubc9 site, though adjacent and immediately upstream from it, suggesting that Ubc9 and PIASy could bind to CA simultaneously. The PIASy binding site is almost perfectly centered on the major homology region (MHR), the most highly conserved sequence in Gag proteins. To identify individual residues required for the binding, a new series of alanine-scanning substitution mutants were generated, each encoding a protein with a cluster of three adjacent residues changed to alanine. These mutants were tested as CA-LexADB fusion proteins for interaction with PIASy in the yeast two-hybrid assay (Fig. 6B). Seven of the mutants (T13, T14, T16, T18, T22, T24, and T25) showed dramatically reduced binding relative to the wild type; one of these, T16, showed no detectable binding. The disruptive mutations were scattered along the region, but included two (T16 and T18) within the MHR itself. The remaining eight mutants showed normal or even elevated reporter activity. The results suggest that PIASy interacts with CA and that the binding site completely overlaps with the MHR. Importantly, all the constructs interacted normally with Ubc9, indicating that the mutant proteins were stable and functional in yeast and were not grossly misfolded or degraded in response to the mutations (Fig. 6B).
To confirm the results of the qualitative filter lift assays, the level of reporter gene expression was measured by quantitative assays for -galactosidase activity in yeast expressing various of the interaction-defective mutants (Fig. 6C). In all cases, these assays closely supported the results of the filter lift assays.
CA mutants that fail to interact with PIASy show reduced SUMOylation in vivo. If the SUMOylation of CA by Ubc9 is an E3-dependent reaction, there should be a dependence on the binding of the E3 enzyme to CA. To test this possibility, various of the nonbinding alanine substitution mutants were expressed along with His-SUMO-1 in 293T cells as before, and the levels of modified CA-His-SUMO-1 were determined by SDS-PAGE and Western blotting (Fig. 7). Mutants T16 and T24 showed strongly reduced levels of the modified CA product, and mutant T25 showed essentially none. The total levels of CA protein of these mutants, judged from Western blots of the crude lysates, were comparable to the wild type. Several others of the mutant CA proteins, though stable in yeast, were poorly expressed in 293T cells and could not be tested for acceptor activity. Nevertheless, these results with the stable mutant proteins suggest that the interaction of CA with PIASy is important for efficient SUMOylation of CA by Ubc9 and that this is likely an E3-dependent reaction.
DISCUSSION
The experiments described above demonstrate that the MuLV CA protein interacts with the SUMO transferase machinery both in yeast and in vitro and is SUMOylated in vivo and that the residues important for the interaction and modification are required early in infection. The interaction of CA with Ubc9 requires a canonical binding motif (KxE) in the C-terminal third of the molecule. The interaction with PIASy, remarkably, seems to require a sequence directly overlapping with the MHR, a highly conserved motif found on Gag proteins from nearly all the retroviral genera. The relative positions of these two binding sites suggest that Ubc9 and PIASy could bind simultaneously to CA and cooperate to promote SUMO transfer. There seem to be strict requirements beyond the primary binding sites for the interaction of CA with both proteins. Neither interacts with conventional two-hybrid proteins in which the N terminus of CA is blocked by fusion to the LexA DNA binding protein. This may reflect difficulties in folding of the CA protein in this context; structural studies suggest that a free N terminus of CA might well be critical for proper folding, as the amino terminus is bound back into a pocket in the mature CA proteins of several retroviruses (69). The Ubc9 and PIASy proteins also failed to interact with LexA-Gag containing the complete Gag precursor, even though these constructs are fully functional for the readout of Gag-Gag dimerization and interactions of Gag with other partners. Thus, these failures may reflect a genuine inability of Ubc9 and PIASy to access or recognize their CA binding sites when CA is present in the context of the precursor. The ability of Ubc9 and PIASy to bind CA only in its mature, processed, and properly folded form would be consistent with their role in modifying CA in the context of the PIC early in infection.
The binding of PIASy directly on the MHR is particularly intriguing and raises the possibility that this is a critical function for this highly conserved sequence element. The MHR may have several roles in replication. Some mutants of the Rous sarcoma virus and the Mason-Pfizer monkey virus with substitutions at the MHR are affected in virion assembly, and others are specifically affected in early events, perhaps in the timing of removal of CA from the incoming nucleic acid (10, 66). Deletion of the entire MHR in HIV-1 causes subtle effects on Gag-Gag interactions and disrupts virus assembly, but these effects may be due to global changes in Gag structure (56); substitution mutations in the MHR can also affect assembly but not Gag-Gag binding (45). For many retroviruses, the early functions of the MHR might involve its recognition by the E3 ligases for SUMO-1 transfer. It is also not known whether the MHR of other simple viruses, or of the complex viruses, directly interacts with any E3 ligase, nor whether SUMOylation is important for their replication. The extreme conservation of the MHR suggests that the process might be very widespread among the retroviruses. It is not clear whether PIASy is the only E3 ligase that can recognize the Moloney MuLV CA or whether the other known members of the family (PIAS1, -3, and -x) (7, 42, 60) could also contribute. Recently, knockout mice lacking PIASy were generated (59, 73), and the homozygous mutants were found to be able to support at least basic normal murine virus replication (73). Thus, if an E3 activity is required by the virus, another PIAS family member must be able to perform this role.
The residues required for the interaction of CA with Ubc9 in yeast were found to be critical for virus replication. Furthermore, a number of lysine residues in the vicinity of the Ubc9 consensus site were also important for virus viability. Mutation of lysine 218 to arginine in the consensus site alone significantly impaired replication, and mutation of three or five lysines in the region including this residue dramatically or completely abolished viability. The mutations had no apparent effect on virion assembly and release, and the resulting viruses entered cells normally and resulted in the synthesis of normal levels of linear viral DNA. However, these mutants made significantly less or no detectable circular viral DNAs. Thus, Ubc9 binding seemed to be critical for a step temporally close to nuclear entry. This step could be transport of the PIC toward the nucleus; targeting to chromatin or nuclear components upon mitosis; retention of the PIC within the nucleus upon membrane reformation; intranuclear localization of the PIC to particular sites; or release of the DNA ends from the structure of the PIC to permit DNA circularization by host ligases. SUMOylation has been associated with nuclear targeting of several proteins (54, 72), and especially with subnuclear localization of proteins such as the promyelocytic leukemia protein PML to intranuclear bodies (13, 60). Perhaps SUMOylation of CA is similarly involved in localization of the PIC to critical sites where the viral DNA is released from the PIC, such that both circularization and integration can occur.
The results presented here also show that CA expressed in 293T cells can indeed be modified by SUMO-1 addition in vivo. The coexpression of CA and a tagged SUMO-1 resulted in readily detectable levels of a conjugate that contained the tag, was recognized with anti-CA antisera, migrated at the expected size, and survived boiling in SDS. As seen for most other modified proteins, the steady-state levels of the CA-SUMO-1 adduct were low, consistent with the presence of highly active proteases specific for removal of SUMO-1 in most cells (21). The highest level of modification required the PIASy binding site, indicating that this is an E3-dependent reaction. Mutation of three or five lysines near the Ubc9 binding site reduced or eliminated SUMO-1 modification, indicating that one or more of these lysines were absolutely required in this setting and likely were serving as the addition sites. Mutation of single lysines had no effect on the levels of the modification, suggesting that no one lysine served as a unique target of SUMO-1 addition. Mutation K218R in particular had no effect on SUMO-1 addition, even though this single change did reduce virus replication. This discrepancy may be due to the forced overexpression of CA. CA is expressed at much higher levels in these experiments than in a normal infection, and it may not recreate the oligomeric structure of the virion core or the PIC. Thus, the sensitivity of its substrate activity for SUMOylation to mutation may not be identical to that seen in the PIC.
Taken together, these experiments suggest that the binding and modification of CA by Ubc9 and PIASy or related E3 ligases are required for the early events of infection by Moloney MuLV. Although Ubc9 mediates effects on some targets without covalent modification (35), because mutation of lysine residues at the binding site in this case affects replication, it is likely that the actual transfer of SUMO-1 to CA is required. The step affected is close in time to nuclear entry, consistent with the role of SUMOylation of many other substrates. This time is identical to that affected by mutations in another Gag protein, p12, encoded immediately upstream of CA in the Gag precursor (76, 77). The two Gag proteins, p12 and CA, seem to both be required to properly ensure that the PIC enters the nucleus in a functional way early in infection, and indeed these two proteins often need to be derived from the same virus to function properly (39). There is no indication that p12 itself is modified by SUMO-1 or regulates CA modification.
The mechanism of nuclear entry for the MuLVs remains obscure. These viruses, like most other simple retroviruses, do not efficiently infect nondividing cells, and the major block is at the time of nuclear entry (47, 57). In dividing cells, the viruses depend on the breakdown of the nuclear envelope associated with mitosis and are presumably localized inside the nuclei of the daughter cells as their nuclear envelopes reform. The MuLV PICs are thus probably not imported into the nucleus through the activity of nuclear pore complexes, but rather through association with cellular DNA, chromatin, or chromatin-associated proteins. The key step may be the targeting of the PIC to nuclear components or PIC retention in the nucleus. It is also possible that the key step mediated by SUMOylation is after nuclear entry and involves intranuclear targeting to critical sites or bodies within the nucleus. The addition of SUMO-1 to many transcription factors and other target proteins for their intranuclear localization into nuclear bodies is fully consistent with this proposed role for SUMOylation of CA in MuLV infection.
A number of host restriction factors that block early events have been found to be strongly dependent on the particular CA protein present on the incoming virus. The Fv1 gene of mice, which provides a dominant block to infection after reverse transcription but before appearance of the nuclear DNA forms, is a prime example (64). There are two major Fv1 alleles: Fv1n, which blocks so-called B-tropic viruses with glutamic acid at CA residue 110; and Fv1b, which blocks N-tropic viruses with arginine at this position. It is possible that the target of Fv1 action is not only CA but actually the SUMOylation of CA. Preliminary studies from our laboratory indicate that Fv1, like CA, interacts with Ubc9 and PIASy; and thus an attractive model for its mode of action is that it might prevent the required modification of CA. The block by Fv1 is precisely at the same time in the life cycle as the block induced by mutation of the SUMO-1 modification sites, and so inhibition of CA SUMOylation by Fv1 would correctly account for its time of action. In primates, a different restriction factor, the TRIM5 gene product, has recently been shown to block many retroviruses (25, 33, 67, 75) and again its activity depends on the CA protein of the incoming virus (9, 22-24, 52, 53; see references 38 and 65 for reviews). The block by TRIM5 is earlier, before reverse transcription, however, and so is not straightforwardly consistent with affecting SUMOylation.
The involvement of viral proteins with the SUMO conjugation system during infection is not likely to be limited to the MuLVs or even the retroviruses. The Mason-Pfizer monkey virus CA protein has recently been reported to interact with Ubc9, and coexpression of the proteins results in their colocalization to foci in the perinuclear region (71). Overexpression of Ubc9 was found to drive a fraction of the Gag protein into the nucleus. The p6 Gag of HIV-1 has recently been shown to be modified by SUMO-1 addition to lysine K27, though mutation of this residue had no apparent effect on virus replication (19). The nucleocapsid protein (NP) of the Seoul, Hantaan, and Tula hantaviruses, negative-strand RNA viruses, also interacts with Ubc9 and SUMO-1, and this interaction is required for localization of NP to the perinuclear region (32, 37, 44). Thus, many viruses may use the SUMO-1 system for localization near or inside the nuclear membrane. These findings suggest that SUMOylation might be a prime target for intervention by antiviral inhibitors. Such new approaches may significantly expand the range and mechanisms of action of drugs available for antiviral therapy.
ACKNOWLEDGMENTS
We are grateful to Valmik K. Vyas for helpful discussions and critical readings of early drafts. We thank Matthew Evans, Gilda Tachedjian, Scott Hughes, and Barbara Studamire for technical support and advice.
This work was supported by PHS grant R37 CA 30488 from the National Cancer Institute. A.Y. and S.B. are Associates and S.P.G. is an Investigator of the Howard Hughes Medical Institute.
REFERENCES
Alin, K., and S. P. Goff. 1996. Amino acid substitutions in the CA protein of Moloney murine leukemia virus that block early events in infection. Virology 222:339-351.
Auerbach, M. R., C. Shu, A. Kaplan, and I. R. Singh. 2003. Functional characterization of a portion of the Moloney murine leukemia virus gag gene by genetic footprinting. Proc. Natl. Acad. Sci. USA 100:11678-11683.
Bacharach, E., J. Gonsky, K. Alin, M. Orlova, and S. P. Goff. 2000. The carboxy-terminal fragment of nucleolin interacts with the nucleocapsid domain of retroviral Gag proteins and inhibits virion assembly. J. Virol. 74:11027-11039.
Best, S., P. Le Tissier, G. Towers, and J. P. Stoye. 1996. Positional cloning of the mouse retrovirus restriction gene Fv1. Nature 382:826-829.
Bowerman, B., P. O. Brown, J. M. Bishop, and H. E. Varmus. 1989. A nucleoprotein complex mediates the integration of retroviral DNA. Genes Dev. 3:469-478.
Breedon, L., and K. Nasmyth. 1985. Regulation of the HO gene. Cold Spring Harbor Symp. Quant. Biol. 50:643-650.
Chung, C. D., J. Liao, B. Liu, X. Rao, P. Jay, P. Berta, and K. Shuai. 1997. Specific inhibition of Stat3 signal transduction by PIAS3. Science 278:1803-1805.
Colicelli, J., and S. P. Goff. 1988. Sequence and spacing requirements of a retrovirus integration site. J. Mol. Biol. 199:47-59.
Cowan, S., T. Hatziioannou, T. Cunningham, M. A. Muesing, H. G. Gottlinger, and P. D. Bieniasz. 2002. Cellular inhibitors with Fv1-like activity restrict human and simian immunodeficiency virus tropism. Proc. Natl. Acad. Sci. USA 99:11914-11919.
Craven, R. C., A. E. Leure-duPree, R. A. Weldon, Jr., and J. W. Wills. 1995. Genetic analysis of the major homology region of the Rous sarcoma virus Gag protein. J. Virol. 69:4213-4227.
Cullen, B. R. 2001. Journey to the center of the cell. Cell 105:697-700.
Desterro, J. M., J. Thomson, and R. T. Hay. 1997. Ubch9 conjugates SUMO but not ubiquitin. FEBS Lett. 417:297-300.
Duprez, E., A. J. Saurin, J. M. Desterro, V. Lallemand-Breitenbach, K. Howe, M. N. Boddy, E. Solomon, H. de The, R. T. Hay, and P. S. Freemont. 1999. SUMO-1 modification of the acute promyelocytic leukaemia protein PML: implications for nuclear localisation. J. Cell Sci. 112:381-393.
Dvorin, J. D., and M. H. Malim. 2003. Intracellular trafficking of HIV-1 cores: journey to the center of the cell. Curr. Top. Microbiol. Immunol. 281:179-208.
Fassati, A., and S. P. Goff. 1999. Characterization of intracellular reverse transcription complexes of Moloney murine leukemia virus. J. Virol. 73:8919-8925.
Gao, G., and S. P. Goff. 1999. Somatic cell mutants resistant to retrovirus replication: intracellular blocks during the early stages of infection. Mol. Biol. Cell 10:1705-1717.
Goff, S., P. Traktman, and D. Baltimore. 1981. Isolation and properties of Moloney murine leukemia virus mutants: use of a rapid assay for release of virion reverse transcriptase. J. Virol. 38:239-248.
Goff, S. P. 2001. Intracellular trafficking of retroviral genomes during the early phase of infection: viral exploitation of cellular pathways. J. Gene Med. 3:517-528.
Gurer, C., L. Berthoux, and J. Luban. 2005. Covalent modification of human immunodeficiency virus type 1 p6 by SUMO-1. J. Virol. 79:910-917.
Hanes, S. D., and R. Brent. 1989. DNA specificity of the bicoid activator protein is determined by homeodomain recognition helix residue 9. Cell 57:1275-1283.
Hang, J., and M. Dasso. 2002. Association of the human SUMO-1 protease SENP2 with the nuclear pore. J. Biol. Chem. 277:19961-19966.
Hatziioannou, T., S. Cowan, and P. D. Bieniasz. 2004. Capsid-dependent and -independent postentry restriction of primate lentivirus tropism in rodent cells. J. Virol. 78:1006-1011.
Hatziioannou, T., S. Cowan, S. P. Goff, P. D. Bieniasz, and G. J. Towers. 2003. Restriction of multiple divergent retroviruses by Lv1 and Ref1. EMBO J. 22:385-394.
Hatziioannou, T., S. Cowan, U. K. von Schwedler, W. I. Sundquist, and P. D. Bieniasz. 2004. Species-specific tropism determinants in the human immunodeficiency virus type 1 capsid. J. Virol. 78:6005-6012.
Hatziioannou, T., D. Perez-Caballero, A. Yang, S. Cowan, and P. D. Bieniasz. 2004. Retrovirus resistance factors Ref1 and Lv1 are species-specific variants of TRIM5alpha. Proc. Natl. Acad. Sci. USA 101:10774-10779.
Hay, R. T. 2005. SUMO: a history of modification. Mol. Cell 18:1-12.
Hirt, B. 1967. Selective extraction of polyoma DNA from infected mouse cell cultures. J. Mol. Biol. 26:365-369.
Hochstrasser, M. 2001. SP-RING for SUMO: new functions bloom for a ubiquitin-like protein. Cell 107:5-8.
Hofmann, W., D. Schubert, J. LaBonte, L. Munson, S. Gibson, J. Scammell, P. Ferrigno, and J. Sodroski. 1999. Species-specific, postentry barriers to primate immunodeficiency virus infection. J. Virol. 73:10020-10028.
Hsu, H. W., P. Schwartzberg, and S. P. Goff. 1985. Point mutations in the P30 domain of the gag gene of Moloney murine leukemia virus. Virology 142:211-214.
Jackson, P. K. 2001. A new RING for SUMO: wrestling transcriptional responses into nuclear bodies with PIAS family E3 SUMO ligases. Genes Dev. 15:3053-3058.
Kaukinen, P., A. Vaheri, and A. Plyusnin. 2003. Non-covalent interaction between nucleocapsid protein of Tula hantavirus and small ubiquitin-related modifier-1, SUMO-1. Virus Res. 92:37-45.
Keckesova, Z., L. M. Ylinen, and G. J. Towers. 2004. The human and African green monkey TRIM5alpha genes encode Ref1 and Lv1 retroviral restriction factor activities. Proc. Natl. Acad. Sci. USA 101:10780-10785.
Kozak, C. A., and A. Chakraborti. 1996. Single amino acid changes in the murine leukemia virus capsid protein gene define the target of Fv1 resistance. Virology 225:300-305.
Kurtzman, A. L., and N. Schechter. 2001. Ubc9 interacts with a nuclear localization signal and mediates nuclear localization of the paired-like homeobox protein Vsx-1 independent of SUMO-1 modification. Proc. Natl. Acad. Sci. USA 98:5602-5607.
LaBonte, J. A., G. J. Babcock, T. Patel, and J. Sodroski. 2002. Blockade of HIV-1 infection of New World monkey cells occurs primarily at the stage of virus entry. J. Exp. Med. 196:431-445.
Lee, B. H., K. Yoshimatsu, A. Maeda, K. Ochiai, M. Morimatsu, K. Araki, M. Ogino, S. Morikawa, and J. Arikawa. 2003. Association of the nucleocapsid protein of the Seoul and Hantaan hantaviruses with small ubiquitin-like modifier-1-related molecules. Virus Res. 98:83-91.
Lee, K., and V. N. KewalRamani. 2004. In defense of the cell: TRIM5alpha interception of mammalian retroviruses. Proc. Natl. Acad. Sci. USA 101:10496-10497.
Lee, S.-K., K. Nagashima, and W.-S. Hu. 2005. Cooperative effect of Gag proteins p12 and capsid during early events of murine leukemia virus replication. J. Virol. 79:4159-4169.
Lewis, P. F., and M. Emerman. 1994. Passage through mitosis is required for oncoretroviruses but not for the human immunodeficiency virus. J. Virol. 68:510-516.
Lilly, F. 1967. Susceptibility to two strains of Friend leukemia virus in mice. Science 155:461-462.
Liu, B., J. Liao, X. Rao, S. A. Kushner, C. D. Chung, D. D. Chang, and K. Shuai. 1998. Inhibition of Stat1-mediated gene activation by PIAS1. Proc. Natl. Acad. Sci. USA 95:10626-10631.
Luban, J., K. L. Bossolt, E. K. Franke, G. V. Kalpana, and S. P. Goff. 1993. Human immunodeficiency virus type 1 Gag protein binds to cyclophilins A and B. Cell 73:1067-1078.
Maeda, A., B. H. Lee, K. Yoshimatsu, M. Saijo, I. Kurane, J. Arikawa, and S. Morikawa. 2003. The intracellular association of the nucleocapsid protein (NP) of Hantaan virus (HTNV) with small ubiquitin-like modifier-1 (SUMO-1) conjugating enzyme 9 (Ubc9). Virology 305:288-297.
Mammano, F., . hagen, S. Hglund, and H. G. Gttlinger. 1994. Role of the major homology region of human immunodeficiency virus type 1 in virion morphogenesis. J. Virol. 68:4927-4936.
Melchior, F. 2000. SUMO—nonclassical ubiquitin. Annu. Rev. Cell Dev. Biol. 16:591-626.
Miller, D. G., M. A. Adam, and A. D. Miller. 1990. Gene transfer by retrovirus vectors occurs only in cells that are actively replicating at the time of infection. Mol. Cell. Biol. 10:4239-4242.
Minty, A., X. Dumont, M. Kaghad, and D. Caput. 2000. Covalent modification of p73alpha by SUMO-1. Two-hybrid screening with p73 identifies novel SUMO-1-interacting proteins and a SUMO-1 interaction motif. J. Biol. Chem. 275:36316-36323.
Muller, S., C. Hoege, G. Pyrowolakis, and S. Jentsch. 2001. SUMO, ubiquitin's mysterious cousin. Nat. Rev. Mol. Cell Biol. 2:202-210.
Muller, S., M. J. Matunis, and A. Dejean. 1998. Conjugation with the ubiquitin-related modifier SUMO-1 regulates the partitioning of PML within the nucleus. EMBO J. 17:61-70.
Nisole, S., and A. Saib. 2004. Early steps of retrovirus replicative cycle. Retrovirology 1:9.
Owens, C. M., P. C. Yang, H. Gttlinger, and J. Sodroski. 2003. Human and simian immunodeficiency virus capsid proteins are major viral determinants of early, postentry replication blocks in simian cells. J. Virol. 77:726-731.
Perron, M. J., M. Stremlau, B. Song, W. Ulm, R. C. Mulligan, and J. Sodroski. 2004. TRIM5alpha mediates the postentry block to N-tropic murine leukemia viruses in human cells. Proc. Natl. Acad. Sci. USA 101:11827-11832.
Pichler, A., and F. Melchior. 2002. Ubiquitin-related modifier SUMO1 and nucleocytoplasmic transport. Traffic 3:381-387.
Pincus, T., W. P. Rowe, and F. Lilly. 1971. A major genetic locus affecting resistance to infection with murine leukemia viruses. II. Apparent identity to a major locus described for resistance to Friend murine leukemia virus. J. Exp. Med. 133:1234-1241.
Provitera, P., A. Goff, A. Harenberg, F. Bouamr, C. Carter, and S. Scarlata. 2001. Role of the major homology region in assembly of HIV-1 Gag. Biochemistry 40:5565-5572.
Roe, T., T. C. Reynolds, G. Yu, and P. O. Brown. 1993. Integration of murine leukemia virus DNA depends on mitosis. EMBO J. 12:2099-2108.
Rose, M. D., F. Winston, and P. Hieter. 1990. Methods in yeast genetics: a laboratory course manual. Cold Spring Harbor Laboratory Press, Plainview, N.Y.
Roth, W., C. Sustmann, M. Kieslinger, A. Gilmozzi, D. Irmer, E. Kremmer, C. Turck, and R. Grosschedl. 2004. PIASy-deficient mice display modest defects in IFN and Wnt signaling. J. Immunol. 173:6189-6199.
Sachdev, S., L. Bruhn, H. Sieber, A. Pichler, F. Melchior, and R. Grosschedl. 2001. PIASy, a nuclear matrix-associated SUMO E3 ligase, represses LEF1 activity by sequestration into nuclear bodies. Genes Dev. 15:3088-3103.
Sampson, D. A., M. Wang, and M. J. Matunis. 2001. The small ubiquitin-like modifier-1 (SUMO-1) consensus sequence mediates Ubc9 binding and is essential for SUMO-1 modification. J. Biol. Chem. 276:21664-21669.
Schwarz, S. E., K. Matuschewski, D. Liakopoulos, M. Scheffner, and S. Jentsch. 1998. The ubiquitin-like proteins SMT3 and SUMO-1 are conjugated by the UBC9 E2 enzyme. Proc. Natl. Acad. Sci. USA 95:560-564.
Smith, C. M., W. B. Potts III, J. S. Smith, and M. J. Roth. 1997. RNase H cleavage of tRNAPro mediated by M-MuLV and HIV-1 reverse transcriptases. Virology 229:437-446.
Stoye, J. P. 1998. Fv1, the mouse retrovirus resistance gene. Rev. Sci. Tech. 17:269-277.
Stoye, J. P. 2002. An intracellular block to primate lentivirus replication. Proc. Natl. Acad. Sci. USA 99:11549-11551.
Strambio-de Castillia, C., and E. Hunter. 1992. Mutational analysis of the major homology region of Mason-Pfizer monkey virus by use of saturation mutagenesis. J. Virol. 66:7021-7032.
Stremlau, M., C. M. Owens, M. J. Perron, M. Kiessling, P. Autissier, and J. Sodroski. 2004. The cytoplasmic body component TRIM5alpha restricts HIV-1 infection in Old World monkeys. Nature 427:848-853.
Verger, A., J. Perdomo, and M. Crossley. 2003. Modification with SUMO. A role in transcriptional regulation. EMBO Rep. 4:137-142.
von Schwedler, U. K., T. L. Stemmler, V. Y. Klishko, S. Li, K. H. Albertine, D. R. Davis, and W. I. Sundquist. 1998. Proteolytic refolding of the HIV-1 capsid protein amino-terminus facilitates viral core assembly. EMBO J. 17:1555-1568.
Weinberg, J. B., T. J. Matthews, B. R. Cullen, and M. H. Malim. 1991. Productive human immunodeficiency virus type 1 (HIV-1) infection of nonproliferating human monocytes. J. Exp. Med. 174:1477-1482.
Weldon, R. A., Jr., P. Sarkar, S. M. Brown, and S. K. Weldon. 2003. Mason-Pfizer monkey virus Gag proteins interact with the human sumo conjugating enzyme, hUbc9. Virology 314:62-73.
Wilson, V. G., and D. Rangasamy. 2001. Intracellular targeting of proteins by sumoylation. Exp. Cell Res. 271:57-65.
Wong, K. A., R. Kim, H. Christofk, J. Gao, G. Lawson, and H. Wu. 2004. Protein inhibitor of activated STAT Y (PIASy) and a splice variant lacking exon 6 enhance sumoylation but are not essential for embryogenesis and adult life. Mol. Cell. Biol. 24:5577-5586.
Yamashita, M., and M. Emerman. 2004. Capsid is a dominant determinant of retrovirus infectivity in nondividing cells. J. Virol. 78:5670-5678.
Yap, M. W., S. Nisole, C. Lynch, and J. P. Stoye. 2004. Trim5alpha protein restricts both HIV-1 and murine leukemia virus. Proc. Natl. Acad. Sci. USA 101:10786-10791.
Yuan, B., A. Fassati, A. Yueh, and S. P. Goff. 2002. Characterization of Moloney murine leukemia virus p12 mutants blocked during early events of infection. J. Virol. 76:10801-10810.
Yuan, B., X. Li, and S. P. Goff. 1999. Mutations altering the Moloney murine leukemia virus p12 Gag protein affect virion production and early events of the virus life cycle. EMBO J. 18:4700-4710.(Andrew Yueh, Juliana Leun)