Role of Simian Virus 40 Vp1 Cysteines in Virion Infectivity

Abstract

We have developed a new nonoverlapping infectious viral genome (NO-SV40) in order to facilitate structure-based analysis of the simian virus 40 (SV40) life cycle. We first tested the role of cysteine residues in the formation of infectious virions by individually mutating the seven cysteines in the major capsid protein, Vp1. All seven cysteine mutants—C9A, C49A, C87A, C104A, C207S, C254A, and C267L—retained viability. In the crystal structure of SV40, disulfide bridges are formed between certain Cys104 residues on neighboring pentamers. However, our results show that none of these disulfide bonds are required for virion infectivity in culture. We also introduced five different mutations into Cys254, the most strictly conserved cysteine across the polyomavirus family. We found that C254L, C254S, C254G, C254Q, and C254R mutants all showed greatly reduced (around 100,000-fold) plaque-forming ability. These mutants had no apparent defect in viral DNA replication. Mutant Vp1's, as well as wild-type Vp2/3, were mostly localized in the nucleus. Further analysis of the C254L mutant revealed that the mutant Vp1 was able to form pentamers in vitro. DNase I-resistant virion-like particles were present in NO-SV40-C254L-transfected cell lysate, but at about 1/18 the amount in wild-type-transfected lysate. An examination of the three-dimensional structure reveals that Cys254 is buried near the surface of Vp1, so that it cannot form disulfide bonds, and is not involved in intrapentamer interactions, consistent with the normal pentamer formation by the C254L mutant. It is, however, located at a critical junction between three pentamers, on a conserved loop (G2H) that packs against the dual interpentamer Ca²⁺-binding sites and the invading C-terminal helix of an adjacent pentamer. The substitution by the larger side chains is predicted to cause a localized shift in the G2H loop, which may disrupt Ca²⁺ ion coordination and the packing of the invading helix, consistent with the defect in virion assembly. Our experimental system thus allows dissection of structure-function relationships during the distinct steps of the SV40 life cycle.

Proper virion assembly is important for the spread of a DNA tumor virus such as simian virus 40 (SV40), a member of the papovavirus family. The icosahedral capsid of SV40, which houses the viral minichromosome along with the minor structural proteins Vp2 and Vp3, is made from 72 pentamers of the major structural protein Vp1. These Vp1 pentamers form the building blocks which interact via the five long carboxy-terminal arms extending from each pentamer into neighboring pentamers (13). The crystal structure of SV40 shows bound Gd³⁺ (which acts as a Ca²⁺ mimetic) at two neighboring sites per Vp1 subunit, forming bridges between the pentamer core and the invading C-terminal arms of another pentamer (13, 19). Structural refinement (19) indicates the presence of interpentamer, but not intrapentamer or intramonomer, disulfide linkages among some of the Vp1 cysteine (Cys104) residues. For mouse polyomavirus, intrapentamer disulfide bridges have been observed between Vp1 cysteines 19 and 114 (18). In addition, evidence indicates that calcium ion chelation and disulfide linkage may further stabilize the interaction between Vp1 pentamers on the capsid. Disruption of both SV40 and mouse polyomavirus capsids requires the reducing agent dithiothreitol (DTT) (3, 20) and can be accomplished with the combination of DTT and the calcium chelator EGTA (2). Bacterially produced polyomavirus Vp1 can be induced to self-assemble in vitro into a capsid-like structure by Ca²⁺ addition at physiologic salt concentration (15). Recently reported in vitro studies have hinted that the ways in which Vp1 cysteines contribute to assembly and/or capsid disassembly may be different for SV40 and polyomavirus. Cysteine-free mutant SV40 Vp1 synthesized in vitro forms pentamers but not postpentameric complexes, and the conversion into these complexes has been suggested to involve Vp1 disulfide bonding (10). In contrast, bacterially made cysteine-free mouse polyomavirus Vp1 forms virus-like particles, though at 50% of the wild-type level (17). The greater stability of the wild-type capsids is apparently conferred by the intrapentamer C19–C114 disulfides (17). These results collectively point to calcium ions and disulfide bonds as integral parts of the virion structure (13, 19). However, it is not known whether the observed Vp1 disulfide formation, such as that through cysteines 104 in SV40, is essential for infectious virion formation. We wished to determine which, if any, of the seven SV40 Vp1 cysteines are essential for infectivity, and to define the roles of the essential residues, using a mutagenesis approach. We report here that neither the C104 disulfide nor any other single disulfide bridge is required for virion infectivity. We also provide evidence for the importance of Cys254, a cysteine that is conserved across the polyomavirus family (see references within reference 14), for virion infectivity.

First, seven single-cysteine Vp1 mutants were constructed. Except for the previously created C9A mutation in pSV-Vp1 (9), the systematic mutation of the remaining cysteines began with the construction of plasmid pBS-Vp1 (Fig. 1), which not only serves as an intermediate for cloning the Vp1 mutant viral genomes but also allows for the in vitro synthesis of Vp1. Within the Vp1 coding sequence of pBS-Vp1, several silent base pair substitutions have been made to introduce three additional unique restriction sites—BstBI at SV40 nucleotide 2143, SpeI at nucleotide 2324, and MluI at nucleotide 2449—which facilitated the creation of individual cysteine mutations and of combination mutations for future studies. Individual mutations at cysteines 49, 87, 104, 207, 254, or 267 were introduced into pBS-Vp1 by oligonucleotide-directed mutagenesis via PCR or by ligation of short mutant coding segments (Fig. 1); then suitable fragments of the resulting plasmids were transferred to the nonoverlapping, infectious viral genome (NO-SV40) (8). In this manner, a series of single-cysteine NO-SV40 mutants were created: the alanine-substituted C9A, C49A, C87A, C104A, and C254A mutants; the serine-substituted C207S and C254S mutants; the leucine-substituted C254L and C267L mutants; the glycine-substituted C254G mutant; the glutamine-substituted C254Q mutant; and the arginine-substituted C254R mutant.

FIG. 1.

Construction of Vp1 cysteine mutant plasmid DNAs. Individual Vp1 cysteine residues except cysteine 9 were mutagenized within the plasmid pBS-Vp1, which encodes wild-type Vp1 amino acid sequence and whose Vp1-encoding and immediately flanking regions are shown as a bold horizontal line. Locations of unique restriction sites XbaI (Xb), AccI (Ac), AflII (Af), EcoRI (E), PstI (P), ApaI (Ap), BamHI (Ba), SacI (Sa), and XhoI (Xh) are indicated by vertical lines. Three other unique sites indicated by diamonds and stars—BstBI (Bs), SpeI (Sp), and MluI (M)—are not present in the natural Vp1 coding sequence [Vp1 (natural)] but were introduced into pBS-Vp1 (Vp1-BSM) at the SV40 nucleotide positions noted above the respective sites via silent base pair substitutions. Dots indicate the locations of the seven cysteines, with their respective amino acid numbers noted underneath. Truncated pBS-Vp1 derivatives with the carboxy-terminal 58 amino acids deleted, the pBS-Vp1-ΔC58 (Vp1-ΔC58) series, were also created.

All subcloning techniques were performed as described previously (16). Oligonucleotides for PCR, for linkers, and for sequencing were synthesized by Genosys (The Woodlands, Texas) or by the Oligonucleotide Preparation Laboratory of the University of California—Los Angeles (UCLA) Molecular Biology Institute. All mutations were confirmed by double-stranded DNA sequencing using the ABI PRISM Dye Terminator Cycle Sequencing Ready Reaction Kit (Perkin-Elmer). All DNA sequences below are given in uppercase letters except for mutated SV40 Vp1 nucleotides, which are lowercased. Relevant restriction sites are underlined.

To create pBS-Vp1, pBluescript II KS(+) (Stratagene, La Jolla, Calif.) was first inserted with a linker between the SacI and ApaI sites to introduce unique XbaI, PstI, BstBI, SpeI, MluI, SacI, and XhoI sites and to knock out the plasmid's original SacI and ApaI sites via this insertion. The resulting pBS plasmid was sequentially inserted with five Vp1-encoding fragments: the 495-bp XbaI-to-PstI (Vp1 amino acids 1 to 162) fragment from pSV-Vp1 (8); the 155-bp PstI-to-BstBI (Vp1 amino acids 162 to 214) PCR fragment with template pSV-Vp1 and the 5′-TTTGGAGCTGCAGGGTGTG sense and 5′-GTTTTCATTTTTcgaaGGATCAGGAACCCAGCACTC CACTGGATAAGC antisense primers; the 181-bp BstBI-to-SpeI (Vp1 amino acids 214 to 274) PCR fragment using the 5′-GTTCCTGATCCttcgAAAAATGAAAACACTAGATAT TTTGGAACCTACACAGGTGG sense and 5′-CTGTAAACACCCGACAAATGGTTGTGAtcACCTTGTGTC antisense primers; the 125-bp SpeI-to-MluI (Vp1 amino acids 274 to 316) PCR fragment using the 5′-TTACCAACACTagTGGAACACAGCAGTGGAAGGGACTT sense and 5′-GACCAATACgCgtTGTGTCCTCCTGTTAATTAGGTCAC antisense primers; and the 172-bp MluI-to-XhoI (Vp1 amino acids 316 to 362) PCR fragment using the 5′-GAGGACACAacGcGTGGATGGGCAGCCTATGATTGGA sense and 5′-ATTCCAAACTC GAGGCGCGCTGAGCTCTAAGCACCGCGGCCGCTCT GCATTCTAGTTGTGGTTTGTCC-3′ antisense primers.

Single-cysteine mutant pBS-Vp1's were then derived by replacing Vp1-coding fragments of wild-type pBS-Vp1 with respective mutant fragments as follows. pBS-Vp1-C49A was made by substituting the 630-bp AccI-to-ApaI PCR fragment derived using template pSV-Vp1 and the 5′-GGAGTAGACA GCTTCACTGAGGTGGAGgcCTTTTTAAATCCTCAA ATG sense and 5′-CAAGGGCCCAACACCCTGCTC antisense primers. pBS-Vp1-C87A was made by substituting the 289-bp XbaI-to-EcoRI PCR fragment derived with the 5′-TGGTCTAGATGAAGATGGCCC sense and 5′-AAGGAATTC TAGCCACACTGTAGgcAGGCAGTTGTTCTTTGTCTGG antisense primers. pBS-Vp1-C104A was made by substituting the 849-bp EcoRI-to-EcoRI PCR fragment derived with the 5′-CTAGAATTCCTTTGCCTAATTTAAATGAGGACTTA ACCgcTGGAAATATTTTGATGTGGG sense and 5′-CTTC AAGAATTCGAGCTCGCC antisense primers. pBS-Vp1- C207S was constructed by substituting the 155-bp PstI-to-BstBI PCR fragment derived with the 5′-TTTGGAGCTGCAGGGTGTGTTAG sense and 5′-TTCATTTTTCGAAGGATCAGGAACCCAagACTCCACTGGATAAGC antisense primers. pBS-Vp1-C254A, -C254S, -C254L, -C254G, -C254Q, and -C254R were made by substituting the 66-bp ApaI-to-SpeI linker in which the Cys254 codon was converted into GCT, TCT, CTT, GGC, CAA, and CGC, respectively. pBS-Vp1-C267L was made via a similar linker in which the Cys267 codon was converted into CTT.

pBS-Vp1-ΔC58 and its C207S and C254L mutants were created by replacing the SpeI-to-XhoI region of pBS-Vp1 and its C207S and C254L mutants with a linker that begins with the same sequence as pBS-Vp1 from the SpeI site to the 304th Vp1 codon (TTT) and ends with a stop codon and the XhoI site.

To create viral genomes containing single-cysteine Vp1 mutations, the 1,179-bp XbaI-to-SacI region of the wild-type NO-pSV40 (8) was replaced with the corresponding fragment from pSV-Vp1-C10A (containing the C9A mutation) (9) to yield NO-pSV40-C9A or was replaced with the corresponding fragment from pBS-Vp1-C49A, -C87A, -C104A, -C207S, -C254A, -C254S, -C254L, -C254G, -C254Q, -C254R, or C267L to produce the respective NO-pSV40 mutant. Mutant NO-SV40 viral genomes were prepared from their respective NO-pSV40 plasmids by digestion with BamHI and recircularizing with T4 DNA ligase as described elsewhere (8).

Individual cysteine mutations and viability.

Each mutant NO-SV40 was tested for viability by one or both of two types of plaque-forming assays. In the first assay, cells were microinjected with each NO-SV40 DNA by a previously described method (21). In the second assay, cells were infected with serial dilutions of the NO-SV40 DNA-transfected cell lysate, which was prepared by transfecting CV-1 cells on a 60-mm dish with 1 μg of NO-SV40 DNA using the Effectene transfection kit (Qiagen), harvesting the cells at 72 h posttransfection in 800 μl of serum-free culture medium, and freeze-thawing the cell suspension three times to release potential virions or virion-like particles. For either assay, infected cells or microinjected cells were incubated for 21 days under an agar medium and the plaques formed were visualized as before (21). The viability results are summarized in Table 1. The microinjection type assays gave a plaquing or nonplaquing phenotype, and the ensuing infection type assays permitted measurement of the number and sizes of the plaques. NO-SV40-BSM, which contains the three additional restriction sites of pBS-Vp1 but no amino acid mutations, formed plaques at nearly the wild-type NO-SV40 level (1.8 × 10⁸ PFU/ml), or 1.6 × 10⁸ PFU per ml of transfected cell lysate. Seven of the NO-SV40 single-cysteine mutants, the C9A, C49A, C87A, C104A, C207S, C254A, and C267L mutants, had PFUs that were either similar to or no less than one-fifth that of the wild type. The average plaque diameters of the mutants ranged from 5.2 to 1.6 mm, which are equivalent to, or somewhat smaller than, that of the wild type, 5.0 mm. In contrast, five different C254 substitution mutants, the C254S, C254L, C254G, C254Q, and C254R mutants, showed dramatically reduced viability, either producing no plaques at all (C254R) or producing 20,000- to 400,000-fold fewer plaques than the wild type. Mutant C254L plaques were also exceptionally small (about 0.5 mm). Nonetheless, these results indicate that individual Vp1 cysteines are not required for SV40 infectivity. In particular, the highly viable C104A mutant suggests that the crystallographically observed interpentamer disulfide linkages among some of the Cys104 residues on the CD loops (19) are not essential for infectious virion formation, at least in culture. They presumably formed after particle release in an extracellular oxidizing environment.

TABLE 1.

Viability of Vp1 single-cysteine mutants

NO-SV40 mutant	Formation of plaquesa	Titer in transfected cell lysate (PFU/ml)b	Plaque diam (mm)b
Wild type	Yes	1.8 × 108	5.0 ± 1.2
BSM	Yes	1.6 × 108	4.0 ± 1.0
C9A	Yes	1.2 × 108	3.3 ± 1.5
C49A	Yes	1.4 × 108	4.1 ± 1.3
C87A	Yes	3.5 × 107	3.2 ± 1.2
C104A	Yes	8.0 × 107	5.2 ± 0.3
C207S	Yes	3.9 × 107	1.6 ± 0.5
C254A	ND	2.2 × 108	4.0 ± 0.9
C254S	ND	1.6 × 103	2.4 ± 0.2
C254L	No	4.0 × 103	0.5c
C254G	ND	8.0 × 103	2.6 ± 0.8
C254Q	ND	4.4 × 102	2.0 ± 0.6
C254R	ND	0d
C267L	Yes	1.5 × 108	2.7 ± 0.6

^aIn microinjection assay. ND, not done. 

^bDetermined from an infection type plaque assay. Cells were infected with serially diluted, freeze-thawed cell lysates obtained from transfection with respective viral DNAs, as described in the text. 

^cTiny, barely visible plaques. 

^dNo plaques detected in 0.1 ml of transfected cell lysate. 

Capsid protein subcellular localization and DNA replication of mutants.

The mutant viral genomes were tested for the structural proteins' subcellular localization as described previously (4). Vp1-BSM localized to the nucleus as well as wild-type Vp1. The Vp1's of all Vp1 single-cysteine mutants more or less localized to the nucleus, indicating conservation of the nuclear targeting function of each mutant Vp1. We noted that though nuclearly localized, C254R mutant Vp1 frequently accumulated in several large aggregates in the nucleus (data not shown). The amount and stability of the mutant proteins synthesized are not addressed at this time. As expected from our previously reported result (8), the wild-type Vp2/3 encoded in all the NO-SV40 constructs was similarly found in the nucleus (data not shown).

The mutants were also tested for viral DNA replication. CV-1 cells on a 150-mm dish were transfected with 4 μg of wild-type or cysteine mutant NO-SV40 DNA (prepared by restriction enzyme digestion and ligation of bacterially propagated NO-pSV40 [8]) and harvested at 72 h posttransfection in TD buffer (25 mM Tris-Cl [pH 7.4], 140 mM NaCl, 5 mM KCl, 1 mM NaP_i). The cells were collected by centrifugation and resuspended and sonicated in 0.5 ml of hypotonic buffer (25 mM Tris-Cl [pH 7.6], 1 mM MgCl₂, 0.4 mM CaCl₂, 0.5 mM DTT). It is known that transfected wild-type viral DNA has replicated to a readily detectable amount by this time point and that some of the progeny DNA is already packaged into virion particles. To quantitate the replicated DNA, total DNA in 12.5 μl of the cell lysate was purified by digestion with 0.75 μg of proteinase K/μl in 25 mM Tris-Cl (pH 7.6)–10 mM EDTA at 50°C for 3 h, extraction with phenol-chloroform, and precipitation with ethanol. The resulting DNA was then digested with either KpnI alone or the combination of KpnI and DpnI and was detected by Southern blotting with a ³²P-labeled SV40 DNA probe (Fig. 2). For both the C207S and C254L mutants as well as the wild type, a vast majority of the total viral DNA was resistant to DpnI digestion (Fig. 2) and hence represented host cell-replicated viral DNAs rather than bacterially propagated input DNAs. For all other single-cysteine mutants, intracellular episomal DNA was extracted by the Hirt method (7) from 72-h-transfected cells, similarly analyzed with DpnI, and detected by Southern blotting or ethidium bromide staining following agarose gel separation. We found similar extents of replication (data not shown). Thus, there is no apparent defect in viral genome replication, and no substantial alteration in the mutant proteins' nuclear localization, for all of the single-cysteine mutants. The observation that five of the Cys254 mutants have greatly reduced viability is intriguing. The following experiments were thus carried out with the C254L mutant to examine the stages of the life cycle that are affected by the cysteine mutations.

FIG. 2.

DNA replication and packaging by C207S and C254L mutants. To analyze viral DNA replication, total DNA was prepared from the sonicated lysate of cells transfected with wild-type or mutant (C207S or C254L) NO-SV40 as described in the text, and identical aliquots thereof were digested with KpnI alone or with both KpnI and DpnI, followed by separation on a 0.9% agarose gel, Southern transfer, and hybridization with a nick-translated ³²P-labeled SV40 DNA probe (0.1 μg, or 10⁷ cpm, in 10 ml). The radioactivity of the linearized 5.2-kbp viral DNA band for the KpnI-DpnI double reaction, quantitated by a phosphorimager, was expressed as a percentage of the radioactivity of the same band for the KpnI-only reaction. To analyze viral DNA packaging, identical aliquots of each transfected cell lysate were either treated or not treated with DNase I as described in the text, and the remaining DNA in the lysate was purified and digested with KpnI and analyzed by Southern blotting as above. The radioactivity of the 5.2-kbp DNase I-resistant viral DNA band was expressed as a percentage of the radioactivity of the same band for the non-DNase I-treated sample.

C254L mutant Vp1 forms pentamers in vitro.

We looked at the ability of C254L mutant Vp1 to form pentamers. The carboxy-terminal-truncated wild-type, C207S mutant, and C254L mutant Vp1 proteins were in vitro synthesized from the T7 promoter-based pBS-Vp1-ΔC58 plasmid series (Fig. 1), and the oligomerization states of the resulting [³⁵S]methionine-labeled Vp1's were analyzed by velocity sedimentation in sucrose gradients (Fig. 3).

FIG. 3.

C207S and C254L mutant Vp1's form pentamers in vitro. In vitro-transcribed and -translated wild-type (Wt) or mutant (C207S or C254L) Vp1-ΔC58 proteins were analyzed by sucrose gradient sedimentation as described in the text.For each fraction, the radioactivity of the Vp1 band in the gel lane was quantitated using a phosphorimager. Each quantitated value was expressed as a decimal fraction of the summed Vp1 radioactivity values for all 17 fractions (taken to be 1.0), and was plotted against the fraction number. Arrows indicate the peak positions for three sedimentation markers, from the bottom to the top of the gradient: catalase B (11.3S), IgG (7S), and bovine serum albumin (4.5S).

Four micrograms of pBS-Vp1-ΔC58, pBS-Vp1-C207S-ΔC58, or pBS-Vp1-C254L-ΔC58 DNA was transcribed and translated in a 125-μl reaction mixture containing 50 μCi of [³⁵S]methionine (>1,000 Ci/mmol; Amersham) and 100 μl of TNT Coupled Reticulocyte Lysate Quick Master Mix (Promega) at 30°C for 90 min. The resulting reaction was treated with 78 Worthington units of RNase A and 24 Kunitz units of DNase I in the presence of protease inhibitors (1,000 U of aprotinin/ml, 1 μg of pepstatin A/ml, 1 μg of leupeptin/ml, 10 nM phenylmethylsulfonyl fluoride) at 37°C for 1 h, diluted twofold with distilled water, layered onto a 10.5-ml, 5 to 20% continuous sucrose gradient in 50 mM HEPES (pH 7.5)–140 mM NaCl, and centrifuged at 35,000 rpm for 23 h at 4°C in an SW41 rotor. Seventeen fractions were collected from the bottom of the gradient, and each entire fraction was reacted overnight at 4°C with 10 μl of a 50% (vol/vol) slurry of protein A-Sepharose (Pharmacia) that had been freshly complexed with the immunoglobulin G (IgG) fraction of rabbit anti-Vp1 serum. The Vp1 immunoprecipitates were collected by centrifugation, washed twice with 10 mM Tris-Cl (pH 8.0)–140 mM NaCl–0.1% Triton X-100–0.25% gelatin–protease inhibitors and twice more with the same buffer without the Triton and gelatin, and then analyzed by sodium dodecyl sulfate–10% polyacrylamide gel electrophoresis (SDS–10% PAGE) (12) and fluorography using the Amplify reagent (Amersham).

The 58-amino-acid deletion at the carboxy-terminal end of SV40 Vp1 is analogous to the 57-amino-acid deletion at the carboxy-terminal end of polyomavirus Vp1, which preserves the protein's ability to form pentamers but abolishes its interpentameric interaction (6). If the C254L mutation affected pentamer formation, we would expect to observe a reduced presence of Vp1 in the expected pentamer peak. The results in Fig. 3 showed no obvious effect of C254L or C207S mutations on Vp1 pentamer formation. About 30, 40, and 20% of the Vp1 radiolabel was found in the 7S-to-8S pentamer region for Vp1-ΔC58, Vp1-C254L-ΔC58, and Vp1-C207S-ΔC58, respectively, with the remaining Vp1 label being in the monomeric form. Thus, the ability of C254L mutant Vp1 to form pentamers was not compromised.

The C254L mutant forms virion-like particles but at a reduced amount.

Since the C254L mutant replicated its DNA, and its mutant Vp1 could form pentamers and localize to the nucleus along with Vp2/3, this mutant may be able to form virion particles. So we next examined the amount of virion or virion-like particles formed by the C254L mutant. Sonicated, transfected cell lysates, prepared as described above for the replication analysis, were digested with 500 U of DNase I/ml in 20 mM Tris-Cl (pH 7.6)–2 mM MgCl₂ to degrade unpackaged DNA. Under such conditions, protein-free SV40 DNA, but not DNA in intact virions, was effectively digested and not detected by Southern blotting (data not shown). It must be noted, however, that the nuclease could penetrate the particles and cleave the viral chromosome if the integrity of the particles was somehow compromised by mutations. After the nuclease digestion, the remaining DNA was purified as described earlier and linearized with KpnI before being analyzed by Southern blotting. As shown in Fig. 2, about 70 and 50% of intracellular wild-type and C207S mutant viral DNAs, respectively, were resistant to DNase I, whereas only a small proportion, about 4%, of the C254L mutant viral DNA was protected from DNase I digestion. Either the mutant particles had a structure different from that of wild-type particles or the mutant formed much fewer particles than the wild type, or both. To confirm that the C254L mutant did package a fraction of the viral DNA into virion particles, an aliquot of wild-type- or C254L mutant-transfected cell lysate that was DNase I treated and contained the same amount of remaining viral DNAs was analyzed by velocity sedimentation in sucrose gradients, followed by Southern blotting. Although the viral DNA peak was found in fraction 4 of both wild-type and C254L mutant samples, as expected for the sedimentation of mature virions (Fig. 4), the C254L mutant distribution curve trailed on the righthand (lighter) side of the peak. It is possible that a small portion of mutant particles had become structurally altered during the centrifugation. We noted in the gradients three viral DNA species: a singly nicked DNA appearing near the 7.3-kbp marker, broad bands of linear DNA centering at the expected 5.2-kbp position, and a covalently closed circular DNA. The last species was present in a reduced amount in the mutant lysate relative to that in the wild-type lysate. These particle-derived DNA species were not investigated further at this time, but they may have arisen from the heterogeneous populations of particles that were in various stages of particle maturation in the nucleus (1, 5, 11). Some of these particles may be accessible to the nuclease, leading to the DNA band patterns observed. Also, the variation could in part reflect artifacts of sonication or fractionation that helped release immature, cell-associated particles. Nonetheless, these results indicate that NO-SV40-C254L is capable of forming virion-like particles in the transfected cells, though at an approximately 18-fold reduced level compared to wild-type NO-SV40. Since the reduction in the physical particles does not fully account for the 45,000-fold reduction in plaque number for this mutant (Table 1), we conclude that the low viability of the C254L mutant is due to the sum of the reduced production of mutant virion particles in the nucleus and the apparently poor infectivity of the mutant particles that are produced.

FIG. 4.

The C254L mutant forms virion-like particles. Sonicated lysate alignots of wild-type (Wt) or C254L mutant NO-SV40-transfected cells, estimated to contain similar amounts of DNase I-resistant viral DNA (200 μl for the wild type and 1 ml for the C254L mutant), were centrifuged at 350 × g for 5 min at 4°C to pellet cellular debris. The supernatant was treated with DNase I as described in the text and was sedimented through a 5 to 32% sucrose gradient in 50 mM HEPES (pH 7.5) at 37,000 rpm at 4°C for 80 min in an SW41 rotor. Eighteen fractions were collected from the bottom of the gradient, and the DNA in each fraction was extracted following proteinase K treatment. One-half of each DNA sample was separated on a 0.9% agarose gel, Southern transferred, and hybridized with a nick-translated ³²P-labeled SV40 DNA probe. To the left of each panel, upper and lower lines indicate the positions of 7,554- and 5,243-bp marker DNA fragments, respectively, and an arrow points to the covalently closed circular DNA band. An arrowhead above fraction 4 indicates the location of control wild-type virions sedimented in a parallel gradient.

Structural environment of C254.

In this report, we examined effects on the viral life cycle of mutating each of the seven SV40 Vp1 cysteines and showed that (i) each cysteine could be mutated without significant loss of viability and (ii) certain mutations of Cys254 resulted in a dramatic reduction in virus viability. Cys254, along with the Vp1 G2H loop on which it lies, is highly conserved across the polyomavirus family. Inspection of the three-dimensional crystal structure suggests that most mutations cause subtle effects that interfere with the ability to form capsids. As illustrated in Fig. 5, Cys254 lies on a short loop connecting strands G2 and H, and it is not involved in Vp1–Vp1 interactions that stabilize the pentamer. Its side chain points into the hydrophobic core of Vp1, so that it cannot participate in disulfide bonding. Substitution by the longer side chains, leucine, glutamine, and arginine, would be expected to create a local perturbation of the structure of this loop to accommodate the longer side chain (e.g., the loop might move as a rigid body by 1 to 2 Å away from the core), while substitution by the smaller alanine would not. We cannot readily explain why the C254S and C254G mutants lost viability. The G2H loop lies at a critical junction between three pentamers within the capsid, such that subtle alterations in structure might be expected to affect capsid assembly significantly. On one side of the G2H loop are the twin Ca²⁺-binding sites (Fig. 5) (13, 19), in which the Ca²⁺ ions are coordinated each by three to four acidic residues from different pentamers that are brought into close apposition as a result of capsid assembly. Residues from the G2H loop form direct hydrogen bonds and salt bridges to the Ca²⁺-coordinating residues, including a salt bridge between the side chains of Lys255 and Glu157 and the interaction of the main-chain C⩵O of Leu253 with the side chain of Asn217; thus, subtle changes in the structure or position of the G2H loop might well disrupt Ca²⁺ binding. The G2H loop also makes close contact with a third pentamer, making hydrophobic packing contacts with an invading C-helix (the “αC**” depicted in Fig. 5). Additionally, a shift of the G2H loop could alter the angle of C-helix packing. The C-helix interaction is likely to be important in defining the curvature of the assembling capsid, and alterations in the angle of packing could lead to some of the aberrant assembly products, such as T=1 particles and tubes, observed in in vitro assembly experiments (15).

FIG. 5.

Structural environment of the Vp1 cysteines. (Left) Ribbon structure of an SV40 Vp1 pentamer, with one of the Vp1 monomers shown in white. An invading arm from a neighboring pentamer is pinkish red. The locations of six Vp1 cysteines mutated in this study are shown by yellow or red (Cys254) circles. Cys9 lies within an amino-terminal arm that is disordered in the crystal structure, and is not shown. Two blue circles represent the twin Ca²⁺ sites. (right) An enlarged view of the Ca²⁺-binding sites shows the location of C254 in the G2H loop and its proximity to the Ca²⁺ ions, which are coordinated by glutamate and aspartate side chains from three Vp1 monomers on two different pentamers. Glu48 (E48, marked with a star) is from a neighboring monomer in the same pentamer. Glu330 and Asp345 (E330 and D345, marked with double stars) are from a second pentamer, and the helix αC** is from a third pentamer. Thus, Cys254 lies close to the junction of three pentamers, and mutation to the longer leucine residue is expected to disrupt these interactions and hence capsid assembly. The image on the left is adapted from Structure (19) with permission of the publisher.

In conclusion, we have developed an experimental system that allows dissection of structure-function relationships during distinct steps of the SV40 life cycle. For all polyomavirus family Vp1's, amino acids within the G2H loop region are well conserved and may hold a key to capsid formation in the nucleus. By assessing the overall viability and the cell entry capability of alternative Cys254 mutants, as well as of mutants in which the metal binding residues are altered, we may shed light on the importance of the G2H loop in the virus life cycle.