• No results found

Essential role of the Vp2 and Vp3 DNA-binding domain in simian virus 40 morphogenesis.


Academic year: 2019

Share "Essential role of the Vp2 and Vp3 DNA-binding domain in simian virus 40 morphogenesis."


Loading.... (view fulltext now)

Full text



Copyrightq1995, American Society for Microbiology

Essential Role of the Vp2 and Vp3 DNA-Binding Domain in

Simian Virus 40 Morphogenesis


Department of Biology and Molecular Biology Institute, University of California, Los Angeles, California 90024

Received 26 May 1994/Accepted 2 November 1994

Both a DNA-binding domain and a Vp1 interactive determinant have been mapped to the carboxy-terminal 40 residues of the simian virus 40 (SV40) minor capsid proteins, Vp2 and Vp3 (Vp2/3), with the last 13 residues being necessary for these activities. The role of this DNA-binding domain in SV40 morphogenesis and the ability to separate these two signals were investigated by mutagenesis and assessment of the activity and viability of the mutants. The carboxy-terminal 40 residues of Vp2/3 were expressed as a polyhistidine fusion protein, and five basic residues at the extreme carboxy terminus (Vp3 residues K226, R227, R228, R230, and R233) were mutagenized. The wild-type fusion protein bound DNA with aKdof 33102


M, identical to that of the full-length Vp3. Mutant proteins containing either one to three or four amino acid substitutions bound DNA 4- to 7-fold or 20- to 30-fold less well, respectively, than the wild-type protein did. The most severe point mutants showed residual DNA binding similar to that of a truncated protein which lacks the entire 13 carboxy-terminal residues. All of the point mutants were able to interact with Vp1, indicating that the two signals within this region are mediated by different residues. When the mutations were placed into the context of the viral DNA and introduced into cells, all the structural proteins were expressed and localized correctly. Not all, however, were viable: mutant genomes whose Vp2/3 bound DNA with intermediate affinities formed plaques just as well as wild-type SV40 DNA did, but three mutants showing greatly reduced DNA binding failed to form plaques at all. These results are consistent with the hypothesis that Vp2/3 plays an essential role in SV40 virion assembly in the nucleus.

The structural simplicity of simian virus 40 (SV40) makes it ideal for studying the complex processes of virion assembly. SV40 is composed of a circular 5,243-bp DNA genome, which is enclosed within an icosahedral capsid made up of three virally encoded proteins (32). Seventy-two Vp1 pentamers form the outer shell of the virus, while the minor coat proteins, Vp2 and Vp3, of which there are about 72 copies, lie within this capsid (1, 26, 27). Vp2 and Vp3 are inferred to form prongs which radiate from a minichromosome core into the central cavities of the Vp1 pentamers and, as such, would place the minor coat proteins in contact with the viral DNA as well as Vp1 (19). Since Vp2 is identical in sequence to Vp3 but has an additional 118 amino acids at its amino terminus, the interac-tions of the proteins with Vp1 and DNA are presumed to be the same, occurring through common residues. The location of the Vp2 unique region in the virus structure is not known. The mature virion particle also contains the host histones H2A, H2B, H3, and H4, complexed with the DNA to form a minichromosome (6, 9).

How viral assembly proceeds in the nucleus remains unclear. Several studies with both polyomavirus and SV40 have sug-gested that Vp1 and either Vp2 or Vp3 form ‘‘subvirion’’ complexes immediately following the synthesis of the capsid proteins in the cytoplasm and that these complexes are then transported into the nucleus, where virion assembly proceeds (12, 14, 22, 24, 27). Once the structural proteins enter the nucleus, the interaction of these proteins with the progeny viral

genome has been shown to occur in a sequential manner, with the capsid proteins arranging on the viral minichromosome to form a virion particle (3, 4, 9, 13, 15, 23). However, the order in which the proteins associate with the DNA or whether they do so independently is unknown. This assembly pathway pre-dicts that signals within the structural proteins mediate pro-tein-DNA and protein-protein interactions. Indeed, all three SV40 structural proteins bind DNA but with no apparent se-quence specificity, and they also interact with each other both in vivo and in vitro (7, 17, 24, 31). Studies also have demon-strated that Vp1 from murine polyomavirus binds DNA in a non-sequence-specific manner and can self-assemble to form pentamers and capsid structures (28, 29).

The recently identified SV40 Vp3 DNA-binding domain (DBD) resides within the carboxy-terminal 40 residues of Vp2 and Vp3 (Vp2/3) in a region that is rich in functional domains (7). The majority of the DNA-binding activity is present in the last 13 residues of the protein, since removal of these residues results in a 93% reduction in DNA binding. Removal of these amino acids also results in the loss of interaction of Vp2/3 with Vp1 (16). Thus, this region also encodes, at least partially, a Vp1-interactive determinant. Also present in the last 40 resi-dues of the protein is a nuclear transport signal (Vp3 resiresi-dues 198 to 206), responsible for targeting the newly synthesized Vp2 and Vp3 to the nucleus (8, 11, 16).

Since the Vp2/3 DBD appears to compact DNA, we previ-ously hypothesized that one role of Vp2/3 may be to condense the DNA for packaging during assembly (7). In this study we have analyzed the Vp3 DBD to determine which residues are important for the interaction with DNA and whether these residues also play a role in the interaction with Vp1. We also have asked whether this domain plays a role in virion morpho-genesis.

* Corresponding author. Mailing address: Department of Biology and Molecular Biology Institute, University of California, 405 Hilgard Ave., Los Angeles, CA 90024.


on November 9, 2019 by guest




Construction of plasmids.The expression vector ptrcHisB (Invitrogen, San Diego, Calif.), which contains a protein-coding sequence containing a run of six histidyl residues under the control of the tac promoter, was used to express fusion proteins of the carboxyl terminus of Vp2/3. The plasmid ptrcHis-Vp3C40 was made by inserting a 215-bp HindIII fragment of SV40 DNA (nucleotides [nt] 1493 to 1708) into the HindIII site present in the vector polylinker. The plasmid ptrcHis-Vp3C40DC13, which encodes a fusion protein containing Vp3 residues 195 to 221, was made by inserting a HindIII (nt 1493)-PvuI fragment from pGEX-Vp3DC13 (7) into ptrcHisB digested with HindIII and PvuI. To construct ptrcHis-Vp3C18, which encodes a fusion protein containing the carboxy-terminal 18 residues of Vp3, SV40 was digested with NlaIV, BamHI linkers (CGGAT CCG) were ligated to the 415-bp fragment (nt 1560 to 1975), and the resulting fragment was digested with BamHI and EcoRI (nt 1782) and inserted into the corresponding sites within the ptrcHisB vector.

After the DBD of ptrcHis-Vp3C40 was mutagenized (see below), the DBD mutations were moved into the SV40 genome with pSVsmall, a 5,211-bp plasmid resulting from the digestion of pSV40 (7) with SphI and religation to remove SV40 nt 5411 to 200, as an intermediary. First, the wild-type SV40 AvrII-AflII fragment (nt 1078 to 1628) of pSVsmall was replaced with a 415-bp AvrII-HindIII fragment (nt 1078 to 1493) from SV40 and a HindIII-AflII fragment (nt 1493 to 1628), containing the DBD mutations, from the mutant ptrcHis-Vp3C40 plasmids. A 1,278-bp KpnI-AccI fragment (nt 294 to 1572) from the resulting pSVsmall mutant plasmids was then used to replace the corresponding fragment in pSV40 to generate pSV40-DBD1 to pSV40-DBD12.

Plasmid pGEX-Vp1, which codes for the glutathione-S-transferase (GST)– Vp1 fusion protein, was constructed by subcloning the 1.2-kb XbaI-SacI restric-tion fragment from pSVP1A, which contains the entire SV40 Vp1 coding se-quence (22), into unique XbaI and SacI sites in pGEX-3XS. pGEX-3XS was made by inserting an XbaI-SacI linker (59-CCGGTTCTAGAGTCGACGA GCTC-39; 59-CCGGGAGCTCGTCGACTCTAGAA-39) into the unique XmaI site of pGEX-3X (Pharmacia LKB Biotechnology Inc.). All DNA constructs were confirmed by dideoxynucleotide, double-stranded plasmid sequencing with Sequenase 2.0 (U.S. Biochemicals).

Mutagenesis.Mutations within the DBD of Vp2/3 were made by PCR with a degenerate oligonucleotide primer [59-tacgtatcttaagGTCTACTCCAGTTTTAA CTCC(T/g)AGAACTT(C/g)TATTCC(T/g)CC(T/g)T(T/c)TATGACGAGCTT T39] to mutate Vp2/3 (nt 1633 to 1579, capital letters) and a second primer (59-CCAGACACCCATCAACAGTATTATTT-39) corresponding to sequence within the vector ptrcHisB (nt 3742 to 3767). PCR was performed for 40 cycles as described previously (30) with ptrcHis-Vp3C40 as the template, and the resulting fragment was purified, digested with AflII and ApaI, and used to replace the wild-type fragment from the parent plasmid ptrcHis-Vp3C40.

Expression and purification of fusion proteins.Escherichia coli RR1 cells harboring plasmids encoding the His6-fusion proteins were grown in

Luria-Bertani medium (30) and induced for 4 h with 1 mM isopropyl-b-D -thiogalacto-pyranoside. The cells were harvested and washed twice with 1/10 volume of 0.5 M NaCl in 20 mM KPi(pH 7.8), resuspended in 1/100 volume of the same

solution, and disrupted by sonication on ice. The fusion proteins were purified from the soluble sonic extracts by passing the extracts over a 1-ml Ni21chelating

column (Qiagen, San Diego, Calif.), washing the column with 30 column volumes of 0.5 M NaCl in 20 mM KPi(pH 7.8) followed by 30 column volumes of 0.5 M

NaCl in 20 mM KPi(pH 6.3), and eluting the fusion proteins with 2 column

volumes of 0.3 M imidazole in 0.5 M NaCl–20 mM KPi(pH 6.3). The fusion

proteins were dialyzed against 25 mM Tris (pH 7.6) to remove the imidazole, concentrated to 1 to 2 mg/ml, and stored at2708C. Typically, 1 liter of induced cells yielded between 1 and 4 mg of fusion protein, which accounted for 90 to 95% of the protein present in the affinity column eluate as judged from Coo-massie brilliant blue-stained polyacrylamide gels (data not shown).

GST-Vp3 and GST-Vp1 were affinity purified from the soluble fractions of induced cells harboring the pGEX-Vp3 or pGEX-Vp1 expression vector, respec-tively, as described previously for the purification of GST-Vp3 (7) but with phosphate-buffered saline as the buffer. The purified GST-Vp1 was dialyzed against 25 mM Tris-Cl (pH 7.5) to remove free glutathione, concentrated to 1 to 5 mg/ml, and stored at2708C. Typically, 1 liter of induced cells produced 35 mg of GST-Vp1, of which 20 to 25% was soluble.

Cleavage of GST-Vp1.Affinity-purified GST-Vp1 was digested for 16 to 18 h at room temperature with factor Xa(50:1 [wt/wt]; Pierce) in 50 mM Tris-Cl (pH

7.6) containing 150 mM NaCl and 1 mM CaCl2. This usually resulted in a 50%

total digestion to yield Vp1, GST, and GST-Vp1. The digestion products were passed over a cross-linked glutathione-agarose column to enrich for the cleaved Vp1 in the unbound flowthrough fraction. This Vp1-enriched fraction was con-centrated, iodinated, and stored at2708C.

Radiolabeling.Fusion proteins were iodinated with Na125

I to specific activities of around 5mCi/mg as described previously (11) and stored at2708C.32

P-labeled SV40 DNA was prepared by nick translation as described previously (7).

DNA-binding assays.Filter-binding assays were performed in duplicate with affinity-purified fusion proteins and radiolabeled nick-translated SV40 DNA (1 ng; specific activity, 53107

cpm/mg) as described previously (7). Southwestern

(DNA-protein) blots were performed as previously described (7) with32

P-la-beled SV40 or pBR322 DNA.

Vp1 interaction assays.SV40 empty particles were purified from infected cells as described previously (25) and used as a source of Vp1. The particles were dissociated by incubating empty particle protein (0.25mg/ml) in 1 mM ethylene glycol-bis(b-aminoethyl ether)-N,N,N9,N9-tetraacetic acid (EGTA)–10 mM di-thiothreitol–50 mM Tris (pH 8.0) at 378C for 1 h followed by the addition of KCl to 850 mM and incubation for 30 min (10). Wild-type and mutant125I-His


-Vp3C40 proteins (200,000 cpm; approximately 0.3mg) were added to the capsid proteins (1mg), and the mixtures were then diluted with 50 mM Tris (pH 8.0)–1 mM EGTA to a final salt concentration of 150 mM in 0.1 ml and incubated for 2 h at room temperature. The reaction mixtures were diluted with 0.4 ml of 150 mM NaCl–50 mM Tris (pH 7.5)–1 mM EDTA–0.5% Triton X-100, and 2.5ml of anti-Vp1 antiserum was added. Immunoprecipitations were performed as de-scribed previously (17). The immune complexes were dissociated and separated on 15% polyacrylamide gels, dried, and exposed to film.

For nondenaturing dot blots, 0.5 to 2.5 mg of His6 fusion proteins were

immobilized on nitrocellulose by incubation at room temperature for 16 h in 25 mM Tris-Cl (pH 7.6) containing 50 mM NaCl in a 96-well dot blot manifold. The blots were blocked for 3 to 5 h in 50 mM Tris-Cl (pH 7.6)–25 mM KCl–2.5 mM MgCl2–0.03% NaN3–1 mM phenylmethylsulfonyl fluoride–50 mg of bovine

se-rum albumin per ml and subsequently incubated for 16 to 18 h with 6 to 20 ng of


I-Vp1 per ml as part of the Vp1-enriched fraction described above (specific activity, 53107

cpm/mg) in 20 ml of blocking buffer at room temperature. The blots were washed extensively with 50 mM Tris-Cl (pH 7.6)–25 mM KCl–2.5 mM MgCl2, wrapped in Saran Wrap, and exposed to film.

Preparation of viral DNAs for microinjection.Plasmids pSV40 and pSV40-DBD1 to pSV40-pSV40-DBD12 (10mg) were digested with BamHI, extracted with phenol-chloroform, and religated with 25 U of T4 DNA ligase (Promega) in 1.5 ml for 16 h at 168C. The products were concentrated to 0.2 ml with Centricon-30 microconcentrators (Amicon), extracted with phenol-chloroform, ethanol pre-cipitated, and resuspended in Tris-EDTA. The concentration of ligated variant viral DNAs was determined by agarose gel electrophoresis and adjusted to 50 ng/ml with phosphate-buffered saline.

Cells, microinjection, and plaque assays.TC7 cells, a subline of African green monkey kidney epithelia, were grown as described previously (8). For determi-nation of virally encoded protein expression and localization, cells were grown on coverslips and viral DNAs were injected into their nuclei. The cells were fixed 24 h postmicroinjection and immunofluorescent detection of proteins was per-formed as described previously (8). Typically, 100 cells were used for each DNA injection.

For plaque assays, TC7 cells were grown in 60-mm dishes to early confluency, and 20 cells within 1 mm2were given cytoplasmic injections of viral DNAs at

marked areas on the dish. Plaques were allowed to form and were visualized as described previously (34) with two modifications. First, the cells received an additional agar overlay on day 20 and were visualized on day 24, and second, portions of the agar from plates showing plaques were removed and subsequently used to infect new cells to obtain viral DNA by Hirt lysis in order to sequence the mutant viral DNA (20).


DNA binding by Vp3-containing fusion proteins. We re-cently identified a DBD, present within the carboxy-terminal 35 residues of Vp2/3, which enables Vp2/3 to bind DNA in a non-sequence-specific manner (7). To characterize this domain

further, we constructed a new fusion protein, His6-Vp3C40,

that contained the last 40 residues of Vp2/3 fused to the car-boxy terminus of a small polypeptide containing a run of six

histidyl residues. His6-Vp3C40 bound DNA with a Kdof 33

1028M, identical to that of the full-length Vp3 expressed as a

GST fusion protein (Fig. 1), confirming that the last 40 resi-dues of Vp2/3 constitute the entire DNA-binding activity of

Vp3. To ensure that His6-Vp3C40 was the only protein

con-tributing to the DNA-binding activity in the preparations, we performed a Southwestern blot with radiolabeled SV40 DNA

and affinity-purified His6-Vp3C40, immobilized on

nitrocellu-lose after gel electrophoresis, and found that only the fusion

protein bound to DNA (results not shown). The unfused His6

polypeptide produced by the ptrcHisB vector did not bind DNA either on Southwestern blots or in filter-binding assays

(data not shown). Similarly to GST-Vp3, His6-Vp3C40 showed

no apparent sequence specificity, since on Southwestern blots it bound radiolabeled pBR322 DNA as well as it bound SV40 DNA (7; data not shown). A truncated fusion protein lacking

the last 13 residues of Vp2/3, His6-Vp3C40DC13, bound to

on November 9, 2019 by guest



DNA with a 30-fold-higher Kd (9 3 102

7 M; Fig. 1), also

confirming our previous finding that the removal of these res-idues greatly lowers the Vp3 DNA-binding activity (7). To determine whether the extreme carboxy terminus is

responsi-ble for the binding activity, we constructed and expressed His6

-Vp3C18, a fusion protein containing the last 18 residues of

Vp3. This protein bound DNA with a Kdof 1.731027M (Fig.

1). This result suggests that these 18 residues confer DNA-binding activity but not to the full extent that the carboxy-terminal 40 residues do. Since Vp3 proteins containing multi-ple mutations in the nuclear transport signal (residues 198 to 206) bind DNA similarly to the wild-type protein and are probably not involved in DNA binding (7), it is likely that the remaining residues between the nuclear transport signal and the carboxy-terminal 18 residues (i.e., 207 to 217) contribute to the total Vp3 DNA-binding activity.

Mutants in the Vp2/3 DBD.Since the removal of the

car-boxy-terminal 13 residues of His6-Vp3C40 resulted in a 95%

loss of DNA-binding activity, we mutagenized this region to identify the residues responsible for DNA binding. Five basic residues (K226, R227, R228, R230, and R233) of Vp3 were mutagenized, and 12 independent mutants were isolated (Ta-ble 1). These proteins were tested for their abilities to bind DNA in a filter-binding assay with SV40 DNA as a probe (Fig. 2). All of the mutants displayed altered DNA-binding affinities. Those with one, two, or three mutations bound DNA with 4- to 7-fold-reduced affinities, while mutants with four mutations showed 20- to 30-fold-lower affinities (Fig. 2). One exception to this trend was DBD9, which contained only three substitutions

but bound DNA with a Kdslightly higher than that of all the

other mutants and almost that of the truncated Vp3 mutant,


Interaction of wild-type and mutant His6-Vp3C40 proteins

with Vp1. A Vp1-interactive determinant also has been mapped to the last 13 residues of Vp2/3 and is required for the interaction of SV40 Vp2/3 with Vp1. The last 40 residues of

Vp2/3, expressed as ab-galactosidase fusion protein, are

suf-ficient to confer the Vp1 interaction (16, 17). To determine what effect the mutations that reduce DNA binding have on the interaction between Vp2/3 and Vp1, the ability of the

His6-DBD proteins to interact with Vp1 was tested. We used

SV40 empty particles as a source of Vp1. Empty virion parti-cles were dissociated and incubated with wild-type and mutant


6-Vp3C40 proteins to promote association, and

com-plexes were immunoprecipitated with anti-Vp1 antibody. Im-munoprecipitations with anti-Vp1 antibody were also per-formed in the absence of Vp1 to determine the level of nonspecific precipitation of the fusion proteins. Wild-type Vp3C40 coprecipitated with Vp1 (Fig. 3, lane 1), while

Vp3C40DC13 did not, as expected (lane 3) (16). All of the

DBD mutants were able to interact with Vp1, but only those which bound DNA with the least affinity are shown (Fig. 3, lanes 5 through 12, and data not shown). The ability of the fusion proteins which showed reduced DNA binding to inter-act with Vp1 was also evident from the finter-act that they were viable (see below).

Vp1 binding by the mutant His6-Vp3C40 proteins was

con-firmed by performing dot blots in which the fusion proteins were immobilized on nitrocellulose and probed with radiola-beled Vp1. For these experiments, we expressed and purified a GST-Vp1 fusion protein (Fig. 4A to C, lanes 1) that reacted with rabbit anti-Vp1 antibody (Fig. 4B, lane 1) as did

virion-FIG. 1. DNA binding by Vp2/3-containing fusion proteins. Filter-binding assays were performed by incubating purified GST-Vp3 (E), His6-Vp3C40 (å),

His6-Vp3C18 (F), or His6-Vp3C40DC13 (Ç) at the concentrations indicated with 32

P-labeled SV40 DNA (specific activity, 53107


cpm/mg) as described in Ma-terials and Methods. The assays were performed in duplicate, and the results shown are representative of three different experiments.

FIG. 2. DNA binding by mutant His6-Vp3C40 proteins. The relative

DNA-binding affinities of the mutant His6-Vp3C40 proteins were determined by

filter-binding assays as described in Materials and Methods. The Kds from two or more

independent experiments, each involving at least five protein concentrations, were averaged, and their reciprocals were made relative to that of the wild-type fusion protein, which was taken as 100%.

TABLE 1. Vp2/3 DBD mutants

Mutant Sequencea
















The sequence shown is from Vp3 amino acyl residue 222 to the carboxyl terminus (residue 234). The residues that were mutated are underlined in the wild-type sequence, and only the amino acid substitutions are shown for the mutants.

bwt, wild type. cHis

6-Vp3C40DC13, containing Vp3 residues 195 to 221, is a truncated His6

-Vp3C40 fusion protein which lacks the last 13 residues.

on November 9, 2019 by guest



derived Vp1 (Fig. 4A and B, lanes 2). The fusion protein was

digested with factor Xa, and the Vp1-enriched fraction (Fig.

4C, lane 2) was used to test for binding by the His6-Vp3C40

fusion proteins (Fig. 4D). As in the immunoprecipitation

ex-periments, all of the DBD mutants bound to Vp1 while the

truncated mutant, His6-Vp3C40DC13, did not. DBD2, DBD3,

DBD4, DBD5, DBD6, DBD10, and DBD12 bound Vp1 as well

as the wild-type His6-Vp3C40 protein, while DBD1, DBD7,

DBD8, DBD9, and DBD11 bound Vp1 less well, albeit still more strongly than the truncated mutant protein did. We also performed ligand blots with immobilized virion proteins that had been separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and probed with

radiola-beled His6-DBD proteins. As in the immunoprecipitations and

dot blots, the wild type and all the mutant DBD fusion proteins

bound to Vp1 but His6-Vp3C40DC13 did not (data not shown).

Thus, the mutations that affect DNA binding do not greatly affect the interaction of Vp3 with Vp1, suggesting that the two signals are functionally distinct.

Biological activity of DBD mutations in mutant viral ge-nomes. To determine what effect these DBD mutations in Vp2/3 have on the biological activity, we placed some of the mutations into the context of the viral DNA by exchanging a fragment containing the DBD mutations with that of the wild-type SV40 DNA counterpart. The variant viral genomes were microinjected into TC7 cells, and the subcellular distribution of viral proteins and plaque formation were examined. Since the carboxy-terminal 35 residues of Vp2/3 and the amino ter-minus of Vp1 are expressed in different reading frames of the same overlapping region of DNA (32), when mutations are made in one of the genes, mutations may also be introduced into the reading frame of the other gene. Figure 5 shows the total potential mutations in both proteins as a result of our mutagenesis, and the specific mutations introduced into Vp1 are shown in Table 2. One mutant, SV40-Vp3DBD9, did not cause any changes in the Vp1 protein. In other mutants,

resi-FIG. 3. Interaction of DBD mutant proteins with Vp1.125


fusion proteins were incubated with (1) or without (2) dissociated SV40 empty particles, and complexes were immunoprecipitated with anti-Vp1 sera and ana-lyzed on SDS–15% polyacrylamide gels. Lanes: 1 and 2, His6-Vp3C40 (wild

type); 3 and 4, His6-Vp3C40DC13; 5 and 6, His6-Vp3C40-DBD9; 7 and 8,

His6-Vp3C40-DBD10; 9 and 10, His6-Vp3C40-DBD11; 11 and 12, His6


-Vp3C40-DBD12. Molecular weight markers are indicated in thousands.

FIG. 4. Interaction of Vp1 with DBD mutant proteins. (A and B) A 0.5-mg sample of affinity-purified GST-Vp1 (lane 1) and 0.8mg of SV40 virion proteins (lane 2) were separated on 10% polyacrylamide gels and either stained with Coomassie brilliant blue (A) or transferred to nitrocellulose, reacted sequentially with anti-Vp1 antisera and125

I-protein A, and subjected to autoradiography to detect anti-Vp1-reactive proteins (B). (C) Affinity-purified GST-Vp1 (1.0mg; lane 1) and GST-Vp1 digested with factor Xaand enriched for Vp1 (lane 2) were

separated on 10% polyacrylamide gels and stained with Coomassie brilliant blue. The cleaved products, after enrichment for Vp1 (see Materials and Methods), were iodinated and used for the ligand dot blots. (D) His6-Vp3C40, His6

-Vp3C40DC13, and His6-DBD mutant fusion proteins were immobilized on

ni-trocellulose and reacted with iodinated proteins enriched for Vp1, as described in Materials and Methods. A 2.5-mg sample of the His6fusion proteins was

immobilized for each dot, except for DBD2, DBD6, and DBD7 (0.5mg each) and DBD4, DBD8, and DBD10 (1.0 mg each). wt, His6-Vp3C40; D, His6


-Vp3C40DC13; 1 to 12, His6-Vp3C40DBD1 to His6-Vp3C40DBD12.

FIG. 5. Potential mutations in Vp1 caused by alterations in Vp2/3. The nucleotide sequences of part of the overlapping region of SV40 Vp1 and Vp2/3 are shown for the wild type and a mutant containing all potential introduced mutations. The amino acid sequences of Vp1 (residues 25 to 38) and Vp3 (residues 222 to 234), expressed in different reading frames, are also shown. The mutated residues are marked by asterisks in the wild-type sequence, and only the substitutions are shown for the mutant proteins.

TABLE 2. Plaque-forming abilities of SV40-Vp2/3 DBD mutants


Change in

Vp1 residue: plaques/No. of

no. of trials

Size (mm) of plaques

30 34

SV40 Ka Ea 9/9 6.761.9

SV40-Vp3DBD1 Q 10/10 5.961.1

SV40-Vp3DBD2 Q 9/9 8.362.7

SV40-Vp3DBD5 Q 10/10 6.461.6

SV40-Vp3DBD6 Q 9/10 7.160.9

SV40-Vp3DBD7 Q Q 9/10 3.760.5

SV40-Vp3DBD9 0/15

SV40-Vp3DBD10 Q Q 0/10

SV40-Vp3DBD11 Q 0/10

SV40-Vp3DBD12 Q Q 10/15 3.260.9


The amino acid residues for the wild-type Vp1 are shown.

on November 9, 2019 by guest


[image:4.612.109.250.407.596.2] [image:4.612.315.555.583.718.2]

dues 30 and/or 34 of Vp1 were altered to glutamine residues. Figure 6 shows the staining pattern of Vp1 and Vp2/3 as detected by indirect immunofluorescence. Both Vp1 and mu-tant Vp2/3 expressed from SV40-DBD10 localized to the nu-cleus (Fig. 6C and D), as did the structural proteins of all of the other mutants (data not shown), and the pattern of staining was indistinguishable from that of the wild-type SV40 DNA-injected cells (Fig. 6A and B). All cells that expressed one of the coat proteins also expressed the other. Nuclear large-T antigen was also observed in cells injected with all variant DNAs (data not shown).

It has been suggested that Vp2/3 may mediate the viral assembly process through its DBD (7). We thus tested the capability of the mutants with altered DNA-binding properties to form plaques. While the wild-type SV40 DNA and mutant DNAs with intermediate affinities for DNA, i.e.,

SV40-Vp3DBD1, SV40-Vp3DBD2, SV40-Vp3DBD5,

SV40-Vp3DBD6, and SV40-Vp3DBD7, were able to form plaques effectively, mutants that showed greatly reduced DNA-binding affinities, i.e., Vp3DBD9, Vp3DBD10, and SV40-Vp3DBD11, were unable to do so (Table 2). One exception

was SV40-Vp3DBD12, which bound DNA with a Kdof 6.53

1027 M and yet formed plaques in some cases. The

SV40-Vp3DBD12 plaques, however, were much smaller than those of wild-type SV40 or the other mutants (Table 2). DNA se-quencing of viral DNA obtained from the plaques of all viable mutants confirmed that the initial mutations were maintained during the infection.

The following observations about the relationship between the inability of the mutants to form plaques and the ability of Vp2/3 to bind DNA were made. First, the alterations in the Vp1 coding sequence do not appear to be responsible for the

lack of plaque-forming ability. SV40-Vp3DBD9, whose Vp3

bound DNA with a Kdthat was 30-fold lower than the

wild-type Vp2/3, contained an unaltered Vp1 coding sequence. SV40-Vp3DBD11, containing a K30Q mutation in Vp1, did not form plaques, whereas Vp3DBD1 and SV40-Vp3DBD5, containing the same Vp1 mutation, did. Similarly, Vp3DBD10, containing identical Vp1 changes to SV40-Vp3DBD7 and SV40-Vp3DBD12, which were viable, was also unable to form plaques. Second, all of the mutant proteins were able to interact with Vp1, indicating that the mutations affect primarily DNA binding. While we do not know whether the interaction between Vp2/3 and Vp1 is quantitatively nor-mal in all of the mutants or occurs with kinetics equal to those of the wild type in vivo, our results are consistent with the conclusion that altered Vp2/3 DNA binding is responsible for the loss of viability.


The analysis of a series of mutations in the carboxy-terminal 13 residues of Vp2/3 indicates that the Vp2/3 DBD plays an important role in virion morphogenesis. We have shown that alteration of basic residues in the carboxy terminus of Vp2/3, expressed as fusion proteins, greatly reduced the affinity of the proteins for DNA while not greatly affecting their ability to interact with the major capsid protein, Vp1. This is the first evidence that the DBD and the Vp1-interactive determinant within Vp2/3 are functionally separable. Furthermore, intro-duction into the viral genome of mutations that severely re-duced Vp2/3 DNA binding resulted in the inability of the genome to direct a productive infectious cycle, consistent with

FIG. 6. Nuclear localization of virally encoded Vp1 and Vp2/3. TC7 cell nuclei were given injections of either circularized wild-type SV40 DNA (A and B) or SV40-Vp3DBD10 DNA (C and D), grown for 24 h at 378C, fixed, and doubly stained with guinea pig anti-Vp1 (A and C) or rabbit anti-Vp3 (B and D) antibodies followed by rhodamine-labeled (A and C) or fluorescein-labeled (B and D) secondary antibodies.

on November 9, 2019 by guest



the hypothesis that DNA binding by Vp2 and Vp3 plays an important role in virion assembly.

The carboxy terminus of SV40 Vp2/3 contains at least three functional domains: a nuclear transport signal, a Vp1-interac-tive determinant, and a DBD (7, 8, 11, 16–18). The ability of SV40 Vp2/3 to interact with Vp1 has been shown to be medi-ated by a determinant that resides entirely within the carboxy-terminal 40 residues of Vp2/3 and also, at least in part, in the

last 13 residues (16, 17). That the His6-Vp3C40 protein bound

to Vp1 confirms that these 40 residues are sufficient for the two proteins to interact. That many of the mutants could form plaques further demonstrates that the mutations do not se-verely impair the Vp1-Vp2/3 interactions. We believe that the rest of the carboxy-terminal 13 residues contribute to the in-teraction; however, how they do so remains to be seen.

The mutations in Vp2/3 in the variant viral DNAs are not within the nuclear transport signal and thus did not affect nuclear targeting of the protein, confirming our previous find-ings that the nuclear transport signal and DBD are functionally distinct and separable (7). The concomitant mutations in Vp1, caused by the overlapping Vp2/3 mutations, are outside of the Vp1 nuclear localization signal (33), and all Vp1 proteins lo-calized to the nucleus, as expected.

That the fusion protein containing the last 18 residues of Vp3 bound DNA with a lower affinity than did the protein containing the last 40 residues, coupled with our previous find-ings that nuclear transport signal mutants show no altered DNA binding, suggests that while the carboxy-terminal 18 res-idues can confer DNA binding, the intervening resres-idues (ami-no acids 207 to 217) may contribute to the Vp3 DNA-binding activity. These intervening residues contain several basic amino acids and share homology with histone H1 (18), further sug-gesting that they may be involved in DNA binding.

Alterna-tively, the 18 residues in the His6-Vp3C18 fusion protein may

confer a protein structure that affects DNA binding. For this reason, in the subsequent DNA-binding studies we used the mutant fusion proteins with amino acid substitutions (see be-low).

It is interesting that the Vp2/3 proteins from murine poly-omavirus terminate immediately following the nuclear trans-port signal and consequently do not contain the last 28 residues that are present in the SV40 counterparts. It is therefore not surprising that polyomavirus Vp2/3 does not bind DNA (5). However, the fact that Vp1 and Vp2/3 are presumed to form cytoplasmic complexes which are then transported to the nu-cleus for assembly (12, 14) implies that the polyomavirus Vp2/3 protein must contain a Vp1-interactive determinant different from that of the SV40 protein. It was recently demonstrated that residues 140 to 184 of polyomavirus Vp2/3 are necessary for the interaction with Vp1 (2). The polyomavirus Vp1-inter-active region is hydrophobic, and within it a homology was found with residues 174 to 186 (2, 19, 26). Since the last 40 residues of SV40 Vp2/3, which are mostly absent in the poly-omavirus proteins, are sufficient for the capsid proteins to interact with SV40 Vp1, as has been shown previously (17; also see above), it is possible that the Vp2/3 proteins from these related yet distinct viruses interact differently with their respec-tive counterpart Vp1 molecules. Alternarespec-tively, multiple active domains may exist in Vp2/3 and contribute to the inter-action with Vp1 to form the virion.

Two human papovaviruses, JC and BK, encode Vp2/3 pro-teins that contain carboxy termini homologous to that of SV40 (32). Within the last 13 residues, 5 basic amino acids are conserved among the viruses, suggesting their importance in DNA binding, and thus were selected for mutagenesis in this study. Analysis of the mutants described in this report revealed

that even slight alterations in the Vp2/3 coding sequence re-sulted in proteins that showed reduced DNA binding and that these five basic residues dominate the interaction between Vp2/3 and DNA. Thus, conservation of these basic residues among related viruses is not unexpected. One mutant, DBD9, contained substitutions at only three positions yet bound DNA with the lowest affinity of all the mutants isolated. Why DBD9 bound DNA less well than did DBD10, which contained all the mutations in DBD9 with an additional substitution at residue 226, is unclear, but the difference is minimal. These three residues, R227, R228, and R233, may be the most important in the protein-DNA interaction. Mutants that showed greatly re-duced DNA binding were not viable, except for DBD12, which was somewhat surprising since it, too, showed reduced DNA binding. The reason that this mutant is viable is unclear. The nonviable mutants all contained three or four mutations, in-cluding R227G and R228G substitutions. DBD12 also con-tained four mutations but did not carry the substitution at R228. Thus, it is possible that there is something specific to these two residues, rather than or in addition to a general defect in DNA binding, that renders the virus nonviable.

Our present result that impaired Vp3 DNA binding corre-lates reasonably well with the inability to form plaques is con-sistent with the hypothesis that Vp2/3 plays a role in virion assembly (6, 7, 21). SV40 Vp1 also has been shown to bind DNA (7, 31), and the Vp1 DBD has been mapped within the amino-terminal 8 residues in the closely related murine poly-omavirus (28), although the corresponding domain in the SV40 protein has not been determined. Since Vp1 and Vp2/3 are targeted to the nucleus and can still interact despite the pres-ence of the Vp2/3 mutations, they presumably form Vp1

pen-tamers and (Vp2/3)(Vp1)5 complexes (17, 22). Except for

SV40-Vp3DBD9, all other Vp2/3 mutations changed either Vp1 lysine 30 or glutamic acid 34, or both, to glutamine. These

alterations are within and near the first b-sheet of the Vp1

structure and are in the region involved in the pentamer-pentamer interaction through the carboxyl residues of a neigh-boring Vp1 molecule (26). Indeed, lysine 30, but not glutamic acid 34, is involved in the pentamer-pentamer interaction by forming a salt bridge with glutamic acid 331. Its mutation into glutamine is expected to alter the interaction but was tolerated in other DBD mutants. If the defect in plaque-forming ability of these mutants is a result of decreased DNA binding, the effect of the mutations would appear to be dominant to the effect of the five Vp1 DBDs present in the Vp1 pentamer and

in the putative (Vp2/3)(Vp1)5complex. One role of Vp2/3 may

be to mediate the compaction and packaging of the viral ge-nome through this DBD as proposed previously (7), and the reduction in the affinity of Vp2/3 for DNA could thus lead to a DNA-packaging deficiency. A study of whether the proteins may form empty particles that contain no packaged DNA, which would be the case for SV40-Vp3DBD9, or remain as unassembled complexes may shed light on the process of virion assembly.


We thank Taylor Wang for help in constructing pSVsmall, Daniel Simmons and Russell Doolittle for stimulating and helpful discussions on protein-DNA interactions and mutagenesis, Jared Clever for estab-lishing the DNA injection and retrieval techniques used in the plaque assays, and Timothy P. Halloran for discussions on expression and purification of the fusion proteins. We also are grateful to Robert Liddington for providing structural analysis of Vp1.

This work was supported by Public Health Service grant CA50574 and a grant from the UCLA Academic Senate. D.A.D. was supported by National Institutes of Health Postdoctoral Fellowship CA09148.

on November 9, 2019 by guest




1. Baker, T. S., J. Drak, and M. Bina. 1988. Reconstruction of the three-dimensional structure of simian virus 40 and visualization of the chromatin core. Proc. Natl. Acad. Sci. USA 85:422–426.

2. Barouch, D. H., and S. C. Harrison. 1994. Interactions among the major and minor coat proteins of polyomavirus. J. Virol. 68:3982–3989.

3. Baumgartner, I., C. Kuhn, and E. Fanning. 1979. Identification and charac-terization of fast-sedimenting SV40 nucleoprotein complexes. Virology 96: 54–63.

4. Bina, M., V. Blasquez, S.-C. Ng, and S. Beecher. 1983. SV40 morphogenesis. Cold Spring Harbor Symp. Quant. Biol. 48:565–569.

5. Chang, D., X. Cai, and R. A. Consigli. 1993. Characterization of the DNA binding properties of polyomavirus capsid proteins. J. Virol. 67:6327–6331. 6. Christiansen, G. T., T. Landers, J. Griffith, and P. Berg. 1977. Character-ization of components released by alkali disruption from simian virus 40. J. Virol. 21:1079–1084.

7. Clever, J. L., D. A. Dean, and H. Kasamatsu. 1993. Identification of a DNA-binding domain in simian virus 40 capsid proteins Vp2 and Vp3. J. Biol. Chem. 268:20877–20883.

8. Clever, J. L., and H. Kasamatsu. 1991. Simian virus 40 Vp2/3 small structural proteins harbor their own nuclear transport signal. Virology 181:78–90. 9. Coco-Prados, M., and M.-T. Hsu. 1979. Intracellular forms of simian virus 40

nucleoprotein complexes. II. Biochemical and electron microscopic analysis of simian virus 40 virion assembly. J. Virol. 31:199–208.

10. Colomar, M. C., C. Degoumois-Sahli, and P. Beard. 1993. Opening and refolding of simian virus 40 and in vitro packaging of foreign DNA. J. Virol.


11. Dean, D. A., and H. Kasamatsu. 1994. Signal- and energy-dependent nuclear transport of SV40 Vp3 by isolated nuclei. Establishment of a filtration assay for nuclear protein import. J. Biol. Chem. 269:4910–4916.

12. Delos, S. E., L. Montross, R. B. Moreland, and R. L. Garcea. 1993. Expres-sion of the polyomavirus Vp2 and Vp3 proteins in insect cells: coexpresExpres-sion with the major capsid protein Vp1 alters Vp2/3 localization. Virology 194: 393–398.

13. Fernandez-Munoz, R., M. Coco-Prados, and M.-T. Hsu. 1979. Intracellular forms of simian virus 40 nucleoprotein complexes. I. Methods of isolation and characterization in CV-1 cells. J. Virol. 29:612–623.

14. Forstova´, J., N. Krauzewicz, S. Wallace, A. J. Street, S. M. Dilworth, S. Beard, and B. E. Griffin.1993. Cooperation of structural proteins during late events in the life cycle of polyomavirus. J. Virol. 67:1405–1413.

15. Garber, E. A., M. M. Seidman, and A. J. Levine. 1980. Intracellular SV40 nucleoprotein complexes: synthesis to encapsidation. Virology 107:389–401. 16. Gharakhanian, E., and H. Kasamatsu. 1990. Two independent signals, a nuclear localization signal and a Vp1-interactive signal, reside within the carboxy-35 amino acids of SV40 Vp3. Virology 178:62–71.

17. Gharakhanian, E., J. Takahashi, J. Clever, and H. Kasamatsu. 1988. In vitro assay for protein-protein interaction: carboxyl-terminal 40 residues of simian

virus 40 structural protein Vp3 contain a determinant for interaction with Vp1. Proc. Natl. Acad. Sci. USA 85:6607–6611.

18. Gharakhanian, E., J. Takahashi, and H. Kasamatsu. 1987. The carboxyl 35 amino acids of SV40 Vp3 are essential for its nuclear accumulation. Virology


19. Griffith, J. P., D. L. Griffith, I. Rayment, W. T. Murakami, and D. L. D.

Caspar.1992. Inside polyomavirus at 25-Å resolution. Nature (London)


20. Hirt, B. 1967. Selective extraction of polyoma DNA from infected mouse cell cultures. J. Mol. Biol. 26:365–369.

21. Huang, E., M. K. Estes, and J. S. Pagano. 1972. Structure and function of the polypeptides in simian virus 40. Existence of subviral deoxynucleoprotein complexes. J. Virol. 9:923–929.

22. Ishii, N., A. Nakanishi, M. Yamada, M. H. Macalalad, and H. Kasamatsu. Functional complementation of nuclear targeting-defective mutants of sim-ian virus 40 structural proteins. J. Virol. 68:8209–8216.

23. Jakobovits, E. B., and Y. Aloni. 1980. Isolation and characterization of various forms of simian virus 40 DNA-protein complexes. Virology 102:107– 118.

24. Kasamatsu, H., and A. Nehorayan. 1979. Vp1 affects intracellular localiza-tion of Vp3 polypeptide during simian virus 40 infeclocaliza-tion. Proc. Natl. Acad. Sci. USA 76:2808–2812.

25. Kasamatsu, H., and M. Wu. 1976. Protein-SV40 DNA complex stable in high salt and sodium dodecyl sulfate. Biochem. Biophys. Res. Commun. 68:927– 936.

26. Liddington, R. C., Y. Yan, J. Moulai, R. Sahli, T. L. Benjamin, and S. C.

Harrison.1991. Structure of simian virus 40 at 3.8-Å resolution. Nature (London) 354:278–284.

27. Lin, W., T. Hata, and H. Kasamatsu. 1984. Subcellular distribution of viral structural proteins during simian virus 40 infection. J. Virol. 50:363–371. 28. Moreland, R. B., L. Montross, and R. L. Gracea. 1991. Characterization of

the DNA-binding properties of the polyomavirus capsid protein Vp1. J. Virol. 65:1168–1176.

29. Salunke, D. M., D. L. D. Caspar, and R. L. Garcea. 1986. Self-assembly of purified polyomavirus capsid protein Vp1. Cell 46:895–904.

30. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

31. Soussi, T. 1986. DNA-binding properties of the major structural protein of simian virus 40. J. Virol. 59:740–742.

32. Tooze, J. 1980. Molecular biology of tumor viruses, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

33. Wychowski, C., D. Benichou, and M. Girard. 1986. A domain of SV40 capsid polypeptide Vp1 that specifies migration into the cell nucleus. EMBO J.


34. Yamada, M., and H. Kasamatsu. 1993. Role of nuclear pore complex in simian virus 40 nuclear targeting. J. Virol. 67:119–130.

on November 9, 2019 by guest



TABLE 1. Vp2/3 DBD mutants
FIG. 5. Potential mutations in Vp1 caused by alterations in Vp2/3. Thenucleotide sequences of part of the overlapping region of SV40 Vp1 and Vp2/3
Figure 6 shows the staining pattern of Vp1 and Vp2/3 asdetected by indirect immunofluorescence


Related documents

Following the educational intervention there was an increase in knowledge of the link between AMD and smoking from 77 to 95%, respondents were also more likely to take a

La Crosse virus infection of BHK cells leads to a dramatic shutoff of not only host protein synthesis but also viral protein synthesis later in infection.. This shutoff can be

CBP complex activity, determined in the presence of eIF3 from poliovirus-infected cells, is present in uninfected cells only and comigrates on sucrose gradient with an activity

This doctoral thesis by prospective publication aims to provide pragmatic, evidence-based guidance for the development and evaluation of physiological breech skills and

The realization of a marketing minimizer type of strategy produces higher performance when accompanied by: (a) high levels of centralization, (b) high levels of formalization, (c)

In vitro, Hi DNA was not restricted by BsuR under standard conditions (200 mM salt), although the expected frequency of -GGCC- cleavage sites was approxi- mately 250.. However,

research confirms that bullying also takes place among university students, including.. undergraduates, post-graduates and doctoral

In this paper, a Monte Carlo simulation based model has been developed to trace the path of photons within a highly scattering medium with low absorbance (optically