Formation of covalently modified folding intermediates of simian virus 40 Vp1 in large T antigen-expressing cells.
Journal: 2013/July - Journal of Virology
ISSN: 1098-5514
Abstract:
The folding and pentamer assembly of the simian virus 40 (SV40) major capsid protein Vp1, which take place in the infected cytoplasm, have been shown to progress through disulfide-bonded Vp1 folding intermediates. In this report, we further demonstrate the existence of another category of Vp1 folding or assembly intermediates: the nonreducible, covalently modified mdVp1s. These species were present in COS-7 cells that expressed a recombinant SV40 Vp1, Vp1ΔC, through plasmid transfection. The mdVp1s persisted under cell and lysate treatment and SDS-PAGE conditions that are expected to have suppressed the formation of artifactual disulfide cross-links. As shown through a pulse-chase analysis, the mdVp1s were derived from the newly synthesized Vp1ΔC in the same time frame as Vp1's folding and oligomerization. The apparent covalent modifications occurred in the cytoplasm within the core region of Vp1 and depended on the coexpression of the SV40 large T antigen (LT) in the cells. Analogous covalently modified species were found with the expression of recombinant polyomavirus Vp1s and human papillomavirus L1s in COS-7 cells. Furthermore, the mdVp1s formed multiprotein complexes with LT, Hsp70, and Hsp40, and a fraction of the largest mdVp1, md4, was disulfide linked to the unmodified Vp1ΔC. Both mdVp1 formation and most of the multiprotein complex formation were blocked by a Vp1 folding mutation, C87A-C254A. Our observations are consistent with a role for LT in facilitating the folding process of SV40 Vp1 by stimulating certain covalent modifications of Vp1 or by recruiting certain cellular proteins.
Relations:
Content
Citations
(1)
References
(49)
Grants
(11)
Diseases
(1)
Chemicals
(2)
Organisms
(4)
Processes
(1)
Anatomy
(2)
Affiliates
(1)
Similar articles
Articles by the same authors
Discussion board
J Virol 87(9): 5053-5064

Formation of Covalently Modified Folding Intermediates of Simian Virus 40 Vp1 in Large T Antigen-Expressing Cells

INTRODUCTION

The icosahedral capsid of simian virus 40 (SV40), a polyomavirus, has an intricate structure known at an atomic resolution. The capsid is composed of pentamers of the major capsid protein, Vp1, along with two embedded, internally situated minor capsid proteins, Vp2 and Vp3. Each Vp1 monomer has, at its core, a β-barrel domain structure of jelly roll topology, whose secondary-structural elements interdigitate between adjacent monomers within the pentamer (1, 2). The C-terminal arm of Vp1, along with disulfide bridges, ties the pentamers together on the capsid (1, 2), while the N-terminal arm contains Vp1's nuclear localization signal (NLS) and DNA-binding domain (3, 4). The three capsid proteins have separate and distinct functions in the viral life cycle (57). Vp2 and Vp3 are required for the transport of the infecting viral DNA to the cell nucleus (5, 7). Vp1 is necessary for the packaging of the viral minichromosome and assembly of the capsid and mediates cell attachment and entry (5, 6). Thus, the formation of infectious SV40 virions depends on the proper folding of the newly synthesized Vp1 into the functional building block of the capsid, namely, the Vp1 pentamer.

Our previous studies have shown that the folding of Vp1 requires specific molecular determinants within Vp1, involves the participation of certain other proteins, and proceeds through distinct Vp1 intermediates. The Vp1 pentamer is formed during or soon after the monomer's synthesis in the SV40-infected cytoplasm (8, 9). This pentamer formation is accompanied by the sequential appearance of transitory, disulfide-bonded Vp1 intermediates, beginning with an intramolecularly disulfide-bonded monomer (8), which converts into the disulfide-free Vp1 chain before giving rise to intermolecularly disulfide-bonded Vp1 dimers through pentamers (8). These disulfide redox exchanges are expected to involve certain pairs of Vp1 cysteine residues. In fact, the mutation of two Vp1 cysteine pairs (C49A-C87A and C87A-C254A) leads to defective Vp1 folding in the cytoplasm and the loss of viral viability (10, 11). The mutant Vp1s, despite harboring a normal NLS, are largely blocked in their movement to the cell nucleus and induce an aberrant subcellular localization of the heat shock proteins Hsp70 and Hsc70 (which are referred to as HSP70 here) (10, 12).

Furthermore, we recently discovered a new class of Vp1 intermediates, which we termed the modified Vp1s (mdVp1s). These species, found in the cytoplasm of COS-7 cells expressing a recombinant Vp1 lacking the C-terminal arm, Vp1ΔC (10), have SDS-PAGE mobilities slower than that of the Vp1 monomer but, unlike the disulfide-containing intermediates, are nonreducible (12). Hence, the mdVp1s appear to harbor unidentified covalent modifications. We hypothesize that these mdVp1 species are Vp1 folding intermediates and play a key role in the SV40 life cycle. In support of this idea, we previously identified an analogous nonreducible, 78-kDa Vp1 species in the cytoplasm of SV40-infected TC7 cells (8).

The HSP70 molecular chaperones, which are known to assist in the folding of nascent or unfolded proteins (1317), have been implicated in the life cycles of polyomaviruses (1721). HSP70 couples its binding and release of protein substrates to ATP hydrolysis. Cochaperones of the Hsp40 family, which interact with HSP70 via their J domains, regulate the ATPase activity and substrate selection of HSP70 (16, 22, 23). The SV40-encoded oncoproteins, the large T (LT) and small t (ST) antigens, are also J domain proteins (2428) and interact with Hsc70 (2933). The roles of LT/ST in viral DNA replication, transcriptional regulation, transformation, and virion maturation are well known (reviewed in references 34 and 35). Our recent study has further implicated LT/ST, in addition to or instead of Hsp40, in mediating the interaction of HSP70 with the core region of Vp1. This finding suggests that LT/ST might orchestrate the chaperone-assisted folding of Vp1 pentamers in the cytoplasm, perhaps by recruiting specific cellular proteins to the multiprotein complexes involved.

In this report, we asked whether the nonreducible, modified mdVp1s represent genuine residents in the cytoplasm of Vp1ΔC-expressing COS-7 cells and whether they are intermediates of the normal Vp1 folding process. We also asked whether mdVp1s interact with cellular and viral chaperones and cochaperones. Finally, we asked whether our observation of nonreducible covalent modification for SV40 Vp1 can be extended to the major capsid proteins of human polyomaviruses and human papillomaviruses.

MATERIALS AND METHODS

Plasmid construction and recombinant protein preparation.

The construction of pCI-Vp1-ΔC58-H6 (pCI-Vp1ΔC) and its mutant derivatives pCI-Vp1-ΔN20ΔC58-H6, pCI-Vp1ΔC58-H6-C49A-C87A, and pCI-Vp1ΔC58-H6-C87A-C254A was described previously (10, 12). pCI-Vp1ΔC encodes SV40 Vp1ΔC, which contains Vp1 amino acids (aa) 1 to 298 plus additional amino acids before (MKM) and after it (in order, GPAS, three repeats of the flexible linker GGGGS, and EFESGR) and a His tag (HHHHHH) at the very C terminus. SV40 Vp1ΔC has an expected mass of 35.6 kDa.

pCI-FLAG-SV40-Vp1ΔC58-H6 (pCI-FLAG-Vp1ΔC), encoding SV40 FLAG-Vp1ΔC (containing Vp1 aa 2 to 198), was constructed by inserting the coding sequence for the FLAG epitope tag (FLAG tag; DYKDDDDK) at the N-terminal end of the SV40 Vp1ΔC coding sequence in pCI-Vp1ΔC. This construction was accomplished by first generating two PCR fragments, with pCI-Vp1ΔC as a template, in two separate rounds of amplification. Note that we used silent unique restriction sites in pCI-Vp1ΔC: XbaI, AvrII, and XmaI sites located upstream of the Vp1 initiating methionine; an NheI site located within the Vp1 coding sequence (aa 303); and a BamHI site located downstream of H6 (12). The first round of amplification (PCR1) generated an XbaI-to-AvrII fragment that includes the corresponding region in pCI-Vp1ΔC but adds a BsrGI site and the FLAG coding sequence downstream of AvrII (sense primer, 5′-GGT GGG AGG TCT ATA TAA GCA GA; antisense primer, 5′-CTT GTC ATC GTC GTC CTT GTA GTC CAT TGT ACA CCT AGG CGG CCG CGT GCA). The PCR1 product contains, in order, XbaI, AvrII, and BsrGI sites and the FLAG coding sequence. The second round of amplification (PCR2) generated an XmaI-to-BamHI fragment that includes the corresponding region in pCI-Vp1ΔC but adds the FLAG coding sequence and an AgeI site upstream and downstream of XmaI, respectively (sense primer, 5′-GAC TAC AAG GAC GAC GAT GAC AAG CCC GGG ACC GGT ATG AAG ATG GCC CCA ACA; antisense primer, 5′-GC ATT TTT TTC ACT GCA TTC GGA TCC). The PCR2 product contains, in order, the FLAG coding sequence, XmaI and AgeI sites, the SV40 Vp1ΔC58-H6 coding sequence, and a BamHI site. A heteroduplex of the two PCR products, XbaI-AvrII-BsrGI-FLAG and FLAG-XmaI-AgeI-Vp1ΔC58-H6-BamH I, was made through denaturation and neutralization. Further amplification was performed using the hybridized product as a template, along with the sense primer of PCR1 and the antisense primer of PCR2. The product was digested with XbaI and BamHI and used to replace the corresponding fragment in pCI-Vp1ΔC.

pCI-FLAG-SV40-Vp1ΔNΔC58-H6 (pCI-FLAG-Vp1ΔNΔC), encoding SV40 FLAG-Vp1ΔNΔC (containing Vp1 aa 22 to 298), was constructed in a similar manner, except that 5′-AAG GAC GAC GAT GAC AAG CCC GGG ACC GGT ATG AAG ATG GCC CAA GTG CCA AAG CTC GTC-3′ was used as the sense primer for PCR2 amplification to obtain a fragment containing FLAG-XmaI-AgeI-Vp1Δ20NΔC58-H6-BamHI.

pCI-FLAG-SV40-Vp1ΔC100-H6, encoding SV40 FLAG-Vp1ΔC100 (containing Vp1 aa 1 to 256), was constructed by inserting an NheI site at the 3′ end of the Vp1ΔC100 coding segment and replacing the Vp1ΔC58 coding segment in pCI-FLAG-Vp1ΔC with the Vp1ΔC100 coding segment. Using pCI-FLAG-Vp1 as a template, a fragment containing an XbaI site, the Vp1ΔC100 coding segment, and an NheI site was amplified using a sense primer that hybridizes 100 bp upstream of the XbaI site of pCI-FLAG-SV40-Vp1ΔC100-H6 (5′-GGC GTG TAC GGT GGG AGG TCT ATA) and an antisense primer that includes an NheI site (5′-TGA ACC GCC TCC ACC GCT AGC GGG CCC AAC ACC CTG CTC ATC) (PCR1). Then the fragment of pCI-FLAG-Vp1ΔC that includes an NheI site, GPAS; a flexible linker, EFESGR; a His tag; and a BamHI site was amplified using a sense primer containing an NheI site (5′-GCT AGC GGT GGA GGC GGT TCA) and an antisense primer containing a BamHI site (5′-GC ATT TTT TTC ACT GCA TTC GGA TCC) (PCR2). The two PCR products were hybridized, and the hybridized product was used as the template to amplify an XbaI-BamHI fragment. The PCR product was digested and used to replace the XbaI-BamHI fragment (containing the Vp1ΔC58-H6 coding sequence) of pCI-FLAG-Vp1ΔC.

Plasmids encoding N-terminally FLAG-tagged and C-terminally His-tagged BK virus (BKV) Vp1ΔC (Vp1 aa 1 to 301; pCI-FLAG-BKV-Vp1ΔC-H6), JC virus (JCV) Vp1ΔC (Vp1 aa 1 to 293; pCI-FLAG-JCV-Vp1ΔC-H6), Merkel cell polyomavirus (MCV) Vp1ΔC (Vp1 aa 1 to 325; pCI-FLAG-MCV-Vp1ΔC-H6), human papillomavirus type 16 (HPV16) L1ΔC (L1 aa 1 to 404; pCI-FLAG-HPV16-L1ΔC-H6), and HPV18 L1ΔC (L1 aa 1 to 404; pCI-FLAG-HPV18-L1ΔC-H6) were constructed by replacing the SV40 Vp1 coding region of pCI-FLAG-Vp1ΔC with the respective protein-encoding fragments derived from PCR amplifications. The capsid proteins encoded have the expected masses, 37.0, 36.4, and 35.9 kDa, for the FLAG-Vp1ΔCs of BKV, JCV, and MCV, respectively, and 45.0 kDa for the FLAG-L1ΔCs of HPV16 and HPV18. The last three plasmids were constructed by DNA Express (Quebec, Canada). The PCR methods used in creating all subcloned fragments are as follows.

For DNA fragments encoding BKV Vp1ΔC and JCV Vp1ΔC, PCRs were performed using pGEM-BKV-Vp1 (36) and pCXSN-JCV-Vp1 (37), gifts from Michael Imperiale and Hirofumi Sawa, respectively, as templates, using an XmaI site-containing sense primer for both (5′-GGA TAA AAA AAA CCC GGG ATG GCC CCA ACC AAA AGA AAA GG) and NheI site-containing antisense primers (5′-GGA TAT CGA CTG GCT AGC TGG GTA AGG ATT CTT TAC AG and 5′-GGT AAT TCT CTA GCT AGC TGG GTA GGG GTT TTT AAC CC, respectively). The products were subcloned as XmaI-NheI fragments. For DNA fragments encoding MCV Vp1ΔC, HPV16 L1ΔC, and HPV18 L1ΔC, PCRs were performed using pwM, p16L1-GFP, and p18L1-GFP, gifts from Chris Buck (3840), as the respective templates. The sense primers used were XmaI site-containing 5′-GGT GAC AAG CCC GGG ACT AAG ATG GCC CCG AAG CGC AAG GCC and 5′-GGT GAC AAG CCC GGG ACT AAG ATG AGC CTG TGG CTG CCC AGC G and the AgeI site-containing 5′-CCT GTG AAA ACC GGT CCT AGA ATG GCC CTC TGG AGA CCA TCC GAT AAC, respectively; the antisense primers, all NheI site containing, were 5′-TGA TCC CTT GCT AGC GAC CAC GGG ATA TGG ATT CTT GAC, 5′-TGA TCC CTT GCT AGC GAA GTT CCA GTC CTC CAG GAT GGT GCT, and 5′-TGA TCC CTT GCT AGC GAA ATT CCA GTC TTC CAG GAT GGA GGA, respectively. The PCR products were subcloned as XmaI-NheI or AgeI-NheI fragments.

The SV40 recombinant proteins, His-tagged Vp1ΔC and GST-LT83-708, were expressed in Escherichia coli transformed with, respectively, pQE-Vp1-2CA-ΔC58 (4) and pGEX-LT83-708 (41), which was a gift from Ellen Fanning. The proteins were purified by using previously described methods for His tag and glutathione S-transferase (GST) fusion proteins (4).

Cell culture, DNA transfection, metabolic labeling, and immunofluorescence.

The culture conditions for 293H, 293T, 293TT (42), CV-1, and COS-7 (43) cells have been described (42, 43). U2OS and U2OS/T cells were cultured under conditions identical to those for COS-7 cells. DNA transfection was performed as described previously (12), using 10 μg of plasmid DNA per 10-cm plate containing 4 × 10 CV1, COS-7, U2OS, or U2OS/T cells or 3 × 10 293H, 293T, or 293TT cells. For metabolic radiolabeling, cells transfected with pCI-Vp1ΔC were labeled with 0.4 mCi of [S]methionine per 10-cm plate at 20 h posttransfection (p.t.), as described previously (8, 9, 12). For immunofluorescence microscopy, TC-7 cells grown on coverslips were transfected with pCI-FLAG-Vp1ΔC or pCI-FLAG-Vp1ΔNΔC and fixed with acetone-methanol at 24 h p.t. (12). The subcellular localization of Vp1 was detected with mouse anti-FLAG antibody (Sigma-Aldrich), followed by Alexa Fluor 488-conjugated anti-mouse antibody. Upon observation under an epifluorescence microscope, DAPI (4′,6-diamidino-2-phenylindole) was used to indicate cell nuclei.

Cell lysate preparation and treatment with N-ethylmaleimide (NEM).

The preparation of whole-cell lysates has been described previously (12). Briefly, cells were washed with phosphate-buffered saline (PBS) and extracted with 500 μl per 10-cm dish of modified radioimmunoprecipitation assay (RIPA) buffer {50 mM PIPES [piperazine-N,N′-bis(2-ethanesulfonic acid)], pH 7.2, 1% Nonidet P-40 [NP-40], 0.1% sodium dodecyl sulfate [SDS], 1% sodium deoxycholate [NaDOC], 150 mM NaCl, 40 U DNase I per ml, and 1 mM phenylmethylsulfonyl fluoride [PMSF]}, and the lysate was clarified by centrifugation and stored in aliquots at −70°C.

Two cytoplasmic subfractions, termed cytoplasmic soluble (Sol) and cytoskeletal (Csk), were prepared as previously described (12), with modifications. Briefly, the Sol fraction was extracted with PSCM buffer (10 mM PIPES, pH 7.2, 1 mM CaCl2, 1 mM MgCl2, 250 mM sucrose, 1% NP-40, and 1 mM PMSF), and the Csk fraction was extracted from the PSCM-insoluble material with RSB buffer (10 mM HEPES, pH 7.2, 10 mM NaCl, 1 mM MgCl2, and 1 mM PMSF) containing the added detergents 1% Tween 40 and 0.5% NaDOC. Each 10-cm culture yielded 500 μl of Sol and 300 μl of Csk.

For the prelysis treatment of cells with the membrane-permeable sulfhydryl alkylating agent NEM, cells were incubated with 2.5 mM NEM in PBS for 2 min, and this NEM buffer was removed before the addition of lysis buffers. For the treatment of lysates with NEM under denaturing conditions (postlysis NEM treatment), lysates were brought up to 0.5% SDS, supplemented with 1 mM NEM, and allowed to incubate in the dark for 20 min before further analysis.

Slot blot quantification.

The amounts of Vp1 and LT were measured for lysates prepared from various cell lines with or without endogenous expression of LT and transfected with various Vp1 and LT expression plasmids. Aliquots of whole-cell lysates were serially diluted with a buffer containing 20 mM dithiothreitol (DTT), 20 mM EDTA, and 1% SDS; brought up to 80 μl with TM buffer (20 mM Tris-HCl, pH 7.6, and 20% methanol); applied to 0.1 μm nitrocellulose membrane secured in the slot blot apparatus; and allowed to incubate for 15 min. The membrane was then exposed to UV light for 2 min to cross-link the adhered proteins to the membrane. Serving as standards, purified recombinant proteins, His-tagged Vp1ΔC and GST-LT83-708, of known concentrations were also serially diluted and processed on the same membrane. Vp1 Western blot analysis (Vp1-WB) and LT-WB were subsequently performed using affinity-purified rabbit anti-Vp1 IgG and mouse monoclonal antibody against an LT unique region (PAb423) as the primary antibodies. The quantification of each lysate type was performed with three lysates prepared from three independent transfection experiments. The amount of Vp1 or LT was determined by visually comparing the band intensities of individual lysate samples with those of purified Vp1 or LT standards. The initial values in nanograms per microliter lysate were then converted into micrograms per million cells.

Sucrose sedimentation analysis.

The procedures for fractioning lysates through 5 to 20% or 5 to 32% continuous sucrose gradients have been described previously (8, 12). Briefly, 250 μl of RIPA lysate prepared from 3 × 10 U2OS/T cells was applied to a gradient and sedimented at 37 krpm for 90 min in an SW41 rotor, and the gradient was then collected from the bottom in 15 or 16 fractions. Two different sedimentation conditions were used in the present study: neutral and denaturing. Sedimentation under neutral conditions used a gradient buffer of 10 mM HEPES, pH 7.5, and 150 mM NaCl. For sedimentation under denaturing conditions, the gradient buffer contained additional 0.1% SDS, and the lysate sample was supplemented with 2 mM NEM and 0.5% SDS (final concentration) and incubated at 50°C for 30 min before loading onto the gradient.

IP.

The procedures for immunoprecipitation (IP) and the elution of proteins from IP complexes under either nonreducing or reducing conditions have been described previously (8, 9, 12). Briefly, 100 to 400 μl of whole-cell lysates or 100 to 300 μl of gradient-separated fractions was incubated with an appropriate antibody overnight at 4°C with rotation and then incubated for 2 h at 4°C with 30 μl of an affinity resin. The resin-bound proteins were eluted by boiling for 10 min in a reducing sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 0.1% bromophenol blue, 50 mM DTT, and 20 mM EDTA) or in the counterpart nonreducing buffer (without the DTT). For most types of IP performed—Vp1-IP, Hsp70-IP, Hsp40-IP, and β-galactosidase (β-Gal)–WB (against β-Gal)—the immunoprecipitating antibody was the same as the primary antibody for the corresponding types of Western blotting (see below) and the immunoprecipitating resin was protein G-Sepharose beads (GE Healthcare). For LT-IP (against unique epitopes of LT), the mouse monoclonal antibody SV40-Ab-2 (Oncogene Research Products) was used, followed by TrueBlot anti-mouse Ig beads (eBioscience).

Isolation of His-tagged proteins from cytoplasmic subfractions.

The isolation of His-tagged Vp1ΔC species from Sol and Csk fractions was performed under denaturing conditions as previously described (12), with modifications. An aliquot of the Sol or Csk fraction was first denatured with added SDS (1% final concentration) and DTT (2 mM final concentration) by incubating at 50°C for 10 min and then adjusted to the conditions of the denaturing binding buffer (6 M urea, 40 mM Tris-Cl, 150 mM NaCl, and 10% glycerol) at pH 7.6. This mixture was cleared of debris by centrifugation and then incubated at 4°C for 1 h with 30 μl of Ni-nitrilotriacetic acid (Ni-NTA) resin (50% slurry; Qiagen) that had been equilibrated with the denaturing binding buffer at pH 8.0. The reacted resin was collected and washed three times with the denaturing binding buffer at pH 8.0, except that the second wash contained a higher concentration of NaCl (1 M) and the third wash contained added 30 mM imidazole. The resin-bound proteins were eluted by boiling in reducing Laemmli sample buffer for 10 min and analyzed by SDS-PAGE and fluorography.

Diagonal SDS-PAGE.

Diagonal, or two-dimensional (2D), SDS-PAGE was performed as previously described (8), with modifications. For SDS-PAGE in the first dimension, protein separation was carried out under either reducing (1D-R) or nonreducing (1D-NR) conditions, with differing sample preparation and gel electrophoresis conditions, as described below. For SDS-PAGE in the second dimension (2D), individually excised 1D-R or 1D-NR lanes were incubated with reducing solution (62.5 mM Tris-Cl, pH 6.8, 0.5% SDS, 1 mM EDTA, 5% glycerol, and 50 mM DTT) at 45°C for 30 min with periodic agitation and then embedded on top of 12.5% polyacrylamide gels in reducing agarose (62.5 mM Tris-Cl, pH 6.8, 0.5% SDS, 1 mM EDTA, 50 mM DTT, and 1% agarose). The 2D gels were resolved at 50 mA for 3.5 to 4.5 h and then electroblotted onto a nitrocellulose membrane for the subsequent detection of Vp1 species by Western blotting. The molecular weights of the detected Vp1 species were estimated via the inclusion of the BenchMark prestained protein marker ladder (Life Technologies) in the diagonal SDS-PAGE procedure. For 1D-R, two different methods were used. In the first method, lysate samples were denatured and reduced by boiling in the Laemmli sample loading buffer (SLB) supplemented with 20 mM EDTA and 50 mM DTT (regular SLB) and then resolved in precast 4 to 15% Tris-glycine gradient gels (Bio-Rad). In the second method, NuPAGE LDS Sample Buffer (Life Technologies), which contains lithium dodecyl sulfate and has a pH of 8.4, was used. Lysate samples were boiled in this sample buffer at 1× strength with added 50 mM DTT (high-pH SLB) and loaded onto NuPAGE Novex 4 to 12% Bis-Tris gels (Life Technologies), and electrophoresis was conducted with running buffer containing 1× antioxidant (Life Technologies). For 1D-NR, the samples consisting of gradient fractions from denaturing sucrose sedimentation were precipitated with trichloroacetic acid (TCA), resuspended in a nonreducing sample buffer (67 mM HEPES, pH 6.8, 3% SDS, and 10% glycerol), incubated at 80°C for 10 min, and resolved in 4 to 15% Tris-glycine gradient gels (Bio-Rad).

WB and fluorography.

Lysate samples for SDS-PAGE and WB were prepared in Laemmli sample loading buffer containing 50 mM DTT as the reducing agent and additional 20 mM EDTA, except where different sample preparation conditions are noted for diagonal SDS-PAGE experiments. WB was otherwise performed as previously described (12, 44). The primary antibodies included the following: for Vp1-WB, affinity-purified rabbit polyclonal anti-Vp1 IgG (45); for LT-WB, the mouse monoclonal IgG PAb419 (against the LT/ST J domain) (46) or PAb423 (against a unique C-terminal epitope of LT) (46); for Hsp70-WB, the mouse monoclonal antibody SPA-810 (Assay Designs); for Hsp40-WB, the rabbit polyclonal antibody SPA-400 (Assay Designs); for FLAG-WB, the mouse monoclonal antibody F3165 (clone M2; Sigma-Aldrich); and for β-Gal–WB, a mouse monoclonal antibody against β-galactosidase (Promega). The secondary antibody was either horseradish peroxidase (HRP)-conjugated anti-rabbit or anti-mouse antibody (MP Biomedicals) or TrueBlot HRP-conjugated anti-rabbit or anti-mouse antibody (eBioscience).

For fluorography, [S]methionine-labeled samples were separated by SDS-PAGE and transferred onto a 0.2-μm nitrocellulose membrane, and the membrane was treated with Autofluor (National Diagnostics) for 1 h with shaking. The membrane was dried and exposed to X-ray film at −70°C.

Plasmid construction and recombinant protein preparation.

The construction of pCI-Vp1-ΔC58-H6 (pCI-Vp1ΔC) and its mutant derivatives pCI-Vp1-ΔN20ΔC58-H6, pCI-Vp1ΔC58-H6-C49A-C87A, and pCI-Vp1ΔC58-H6-C87A-C254A was described previously (10, 12). pCI-Vp1ΔC encodes SV40 Vp1ΔC, which contains Vp1 amino acids (aa) 1 to 298 plus additional amino acids before (MKM) and after it (in order, GPAS, three repeats of the flexible linker GGGGS, and EFESGR) and a His tag (HHHHHH) at the very C terminus. SV40 Vp1ΔC has an expected mass of 35.6 kDa.

pCI-FLAG-SV40-Vp1ΔC58-H6 (pCI-FLAG-Vp1ΔC), encoding SV40 FLAG-Vp1ΔC (containing Vp1 aa 2 to 198), was constructed by inserting the coding sequence for the FLAG epitope tag (FLAG tag; DYKDDDDK) at the N-terminal end of the SV40 Vp1ΔC coding sequence in pCI-Vp1ΔC. This construction was accomplished by first generating two PCR fragments, with pCI-Vp1ΔC as a template, in two separate rounds of amplification. Note that we used silent unique restriction sites in pCI-Vp1ΔC: XbaI, AvrII, and XmaI sites located upstream of the Vp1 initiating methionine; an NheI site located within the Vp1 coding sequence (aa 303); and a BamHI site located downstream of H6 (12). The first round of amplification (PCR1) generated an XbaI-to-AvrII fragment that includes the corresponding region in pCI-Vp1ΔC but adds a BsrGI site and the FLAG coding sequence downstream of AvrII (sense primer, 5′-GGT GGG AGG TCT ATA TAA GCA GA; antisense primer, 5′-CTT GTC ATC GTC GTC CTT GTA GTC CAT TGT ACA CCT AGG CGG CCG CGT GCA). The PCR1 product contains, in order, XbaI, AvrII, and BsrGI sites and the FLAG coding sequence. The second round of amplification (PCR2) generated an XmaI-to-BamHI fragment that includes the corresponding region in pCI-Vp1ΔC but adds the FLAG coding sequence and an AgeI site upstream and downstream of XmaI, respectively (sense primer, 5′-GAC TAC AAG GAC GAC GAT GAC AAG CCC GGG ACC GGT ATG AAG ATG GCC CCA ACA; antisense primer, 5′-GC ATT TTT TTC ACT GCA TTC GGA TCC). The PCR2 product contains, in order, the FLAG coding sequence, XmaI and AgeI sites, the SV40 Vp1ΔC58-H6 coding sequence, and a BamHI site. A heteroduplex of the two PCR products, XbaI-AvrII-BsrGI-FLAG and FLAG-XmaI-AgeI-Vp1ΔC58-H6-BamH I, was made through denaturation and neutralization. Further amplification was performed using the hybridized product as a template, along with the sense primer of PCR1 and the antisense primer of PCR2. The product was digested with XbaI and BamHI and used to replace the corresponding fragment in pCI-Vp1ΔC.

pCI-FLAG-SV40-Vp1ΔNΔC58-H6 (pCI-FLAG-Vp1ΔNΔC), encoding SV40 FLAG-Vp1ΔNΔC (containing Vp1 aa 22 to 298), was constructed in a similar manner, except that 5′-AAG GAC GAC GAT GAC AAG CCC GGG ACC GGT ATG AAG ATG GCC CAA GTG CCA AAG CTC GTC-3′ was used as the sense primer for PCR2 amplification to obtain a fragment containing FLAG-XmaI-AgeI-Vp1Δ20NΔC58-H6-BamHI.

pCI-FLAG-SV40-Vp1ΔC100-H6, encoding SV40 FLAG-Vp1ΔC100 (containing Vp1 aa 1 to 256), was constructed by inserting an NheI site at the 3′ end of the Vp1ΔC100 coding segment and replacing the Vp1ΔC58 coding segment in pCI-FLAG-Vp1ΔC with the Vp1ΔC100 coding segment. Using pCI-FLAG-Vp1 as a template, a fragment containing an XbaI site, the Vp1ΔC100 coding segment, and an NheI site was amplified using a sense primer that hybridizes 100 bp upstream of the XbaI site of pCI-FLAG-SV40-Vp1ΔC100-H6 (5′-GGC GTG TAC GGT GGG AGG TCT ATA) and an antisense primer that includes an NheI site (5′-TGA ACC GCC TCC ACC GCT AGC GGG CCC AAC ACC CTG CTC ATC) (PCR1). Then the fragment of pCI-FLAG-Vp1ΔC that includes an NheI site, GPAS; a flexible linker, EFESGR; a His tag; and a BamHI site was amplified using a sense primer containing an NheI site (5′-GCT AGC GGT GGA GGC GGT TCA) and an antisense primer containing a BamHI site (5′-GC ATT TTT TTC ACT GCA TTC GGA TCC) (PCR2). The two PCR products were hybridized, and the hybridized product was used as the template to amplify an XbaI-BamHI fragment. The PCR product was digested and used to replace the XbaI-BamHI fragment (containing the Vp1ΔC58-H6 coding sequence) of pCI-FLAG-Vp1ΔC.

Plasmids encoding N-terminally FLAG-tagged and C-terminally His-tagged BK virus (BKV) Vp1ΔC (Vp1 aa 1 to 301; pCI-FLAG-BKV-Vp1ΔC-H6), JC virus (JCV) Vp1ΔC (Vp1 aa 1 to 293; pCI-FLAG-JCV-Vp1ΔC-H6), Merkel cell polyomavirus (MCV) Vp1ΔC (Vp1 aa 1 to 325; pCI-FLAG-MCV-Vp1ΔC-H6), human papillomavirus type 16 (HPV16) L1ΔC (L1 aa 1 to 404; pCI-FLAG-HPV16-L1ΔC-H6), and HPV18 L1ΔC (L1 aa 1 to 404; pCI-FLAG-HPV18-L1ΔC-H6) were constructed by replacing the SV40 Vp1 coding region of pCI-FLAG-Vp1ΔC with the respective protein-encoding fragments derived from PCR amplifications. The capsid proteins encoded have the expected masses, 37.0, 36.4, and 35.9 kDa, for the FLAG-Vp1ΔCs of BKV, JCV, and MCV, respectively, and 45.0 kDa for the FLAG-L1ΔCs of HPV16 and HPV18. The last three plasmids were constructed by DNA Express (Quebec, Canada). The PCR methods used in creating all subcloned fragments are as follows.

For DNA fragments encoding BKV Vp1ΔC and JCV Vp1ΔC, PCRs were performed using pGEM-BKV-Vp1 (36) and pCXSN-JCV-Vp1 (37), gifts from Michael Imperiale and Hirofumi Sawa, respectively, as templates, using an XmaI site-containing sense primer for both (5′-GGA TAA AAA AAA CCC GGG ATG GCC CCA ACC AAA AGA AAA GG) and NheI site-containing antisense primers (5′-GGA TAT CGA CTG GCT AGC TGG GTA AGG ATT CTT TAC AG and 5′-GGT AAT TCT CTA GCT AGC TGG GTA GGG GTT TTT AAC CC, respectively). The products were subcloned as XmaI-NheI fragments. For DNA fragments encoding MCV Vp1ΔC, HPV16 L1ΔC, and HPV18 L1ΔC, PCRs were performed using pwM, p16L1-GFP, and p18L1-GFP, gifts from Chris Buck (3840), as the respective templates. The sense primers used were XmaI site-containing 5′-GGT GAC AAG CCC GGG ACT AAG ATG GCC CCG AAG CGC AAG GCC and 5′-GGT GAC AAG CCC GGG ACT AAG ATG AGC CTG TGG CTG CCC AGC G and the AgeI site-containing 5′-CCT GTG AAA ACC GGT CCT AGA ATG GCC CTC TGG AGA CCA TCC GAT AAC, respectively; the antisense primers, all NheI site containing, were 5′-TGA TCC CTT GCT AGC GAC CAC GGG ATA TGG ATT CTT GAC, 5′-TGA TCC CTT GCT AGC GAA GTT CCA GTC CTC CAG GAT GGT GCT, and 5′-TGA TCC CTT GCT AGC GAA ATT CCA GTC TTC CAG GAT GGA GGA, respectively. The PCR products were subcloned as XmaI-NheI or AgeI-NheI fragments.

The SV40 recombinant proteins, His-tagged Vp1ΔC and GST-LT83-708, were expressed in Escherichia coli transformed with, respectively, pQE-Vp1-2CA-ΔC58 (4) and pGEX-LT83-708 (41), which was a gift from Ellen Fanning. The proteins were purified by using previously described methods for His tag and glutathione S-transferase (GST) fusion proteins (4).

Cell culture, DNA transfection, metabolic labeling, and immunofluorescence.

The culture conditions for 293H, 293T, 293TT (42), CV-1, and COS-7 (43) cells have been described (42, 43). U2OS and U2OS/T cells were cultured under conditions identical to those for COS-7 cells. DNA transfection was performed as described previously (12), using 10 μg of plasmid DNA per 10-cm plate containing 4 × 10 CV1, COS-7, U2OS, or U2OS/T cells or 3 × 10 293H, 293T, or 293TT cells. For metabolic radiolabeling, cells transfected with pCI-Vp1ΔC were labeled with 0.4 mCi of [S]methionine per 10-cm plate at 20 h posttransfection (p.t.), as described previously (8, 9, 12). For immunofluorescence microscopy, TC-7 cells grown on coverslips were transfected with pCI-FLAG-Vp1ΔC or pCI-FLAG-Vp1ΔNΔC and fixed with acetone-methanol at 24 h p.t. (12). The subcellular localization of Vp1 was detected with mouse anti-FLAG antibody (Sigma-Aldrich), followed by Alexa Fluor 488-conjugated anti-mouse antibody. Upon observation under an epifluorescence microscope, DAPI (4′,6-diamidino-2-phenylindole) was used to indicate cell nuclei.

Cell lysate preparation and treatment with N-ethylmaleimide (NEM).

The preparation of whole-cell lysates has been described previously (12). Briefly, cells were washed with phosphate-buffered saline (PBS) and extracted with 500 μl per 10-cm dish of modified radioimmunoprecipitation assay (RIPA) buffer {50 mM PIPES [piperazine-N,N′-bis(2-ethanesulfonic acid)], pH 7.2, 1% Nonidet P-40 [NP-40], 0.1% sodium dodecyl sulfate [SDS], 1% sodium deoxycholate [NaDOC], 150 mM NaCl, 40 U DNase I per ml, and 1 mM phenylmethylsulfonyl fluoride [PMSF]}, and the lysate was clarified by centrifugation and stored in aliquots at −70°C.

Two cytoplasmic subfractions, termed cytoplasmic soluble (Sol) and cytoskeletal (Csk), were prepared as previously described (12), with modifications. Briefly, the Sol fraction was extracted with PSCM buffer (10 mM PIPES, pH 7.2, 1 mM CaCl2, 1 mM MgCl2, 250 mM sucrose, 1% NP-40, and 1 mM PMSF), and the Csk fraction was extracted from the PSCM-insoluble material with RSB buffer (10 mM HEPES, pH 7.2, 10 mM NaCl, 1 mM MgCl2, and 1 mM PMSF) containing the added detergents 1% Tween 40 and 0.5% NaDOC. Each 10-cm culture yielded 500 μl of Sol and 300 μl of Csk.

For the prelysis treatment of cells with the membrane-permeable sulfhydryl alkylating agent NEM, cells were incubated with 2.5 mM NEM in PBS for 2 min, and this NEM buffer was removed before the addition of lysis buffers. For the treatment of lysates with NEM under denaturing conditions (postlysis NEM treatment), lysates were brought up to 0.5% SDS, supplemented with 1 mM NEM, and allowed to incubate in the dark for 20 min before further analysis.

Slot blot quantification.

The amounts of Vp1 and LT were measured for lysates prepared from various cell lines with or without endogenous expression of LT and transfected with various Vp1 and LT expression plasmids. Aliquots of whole-cell lysates were serially diluted with a buffer containing 20 mM dithiothreitol (DTT), 20 mM EDTA, and 1% SDS; brought up to 80 μl with TM buffer (20 mM Tris-HCl, pH 7.6, and 20% methanol); applied to 0.1 μm nitrocellulose membrane secured in the slot blot apparatus; and allowed to incubate for 15 min. The membrane was then exposed to UV light for 2 min to cross-link the adhered proteins to the membrane. Serving as standards, purified recombinant proteins, His-tagged Vp1ΔC and GST-LT83-708, of known concentrations were also serially diluted and processed on the same membrane. Vp1 Western blot analysis (Vp1-WB) and LT-WB were subsequently performed using affinity-purified rabbit anti-Vp1 IgG and mouse monoclonal antibody against an LT unique region (PAb423) as the primary antibodies. The quantification of each lysate type was performed with three lysates prepared from three independent transfection experiments. The amount of Vp1 or LT was determined by visually comparing the band intensities of individual lysate samples with those of purified Vp1 or LT standards. The initial values in nanograms per microliter lysate were then converted into micrograms per million cells.

Sucrose sedimentation analysis.

The procedures for fractioning lysates through 5 to 20% or 5 to 32% continuous sucrose gradients have been described previously (8, 12). Briefly, 250 μl of RIPA lysate prepared from 3 × 10 U2OS/T cells was applied to a gradient and sedimented at 37 krpm for 90 min in an SW41 rotor, and the gradient was then collected from the bottom in 15 or 16 fractions. Two different sedimentation conditions were used in the present study: neutral and denaturing. Sedimentation under neutral conditions used a gradient buffer of 10 mM HEPES, pH 7.5, and 150 mM NaCl. For sedimentation under denaturing conditions, the gradient buffer contained additional 0.1% SDS, and the lysate sample was supplemented with 2 mM NEM and 0.5% SDS (final concentration) and incubated at 50°C for 30 min before loading onto the gradient.

IP.

The procedures for immunoprecipitation (IP) and the elution of proteins from IP complexes under either nonreducing or reducing conditions have been described previously (8, 9, 12). Briefly, 100 to 400 μl of whole-cell lysates or 100 to 300 μl of gradient-separated fractions was incubated with an appropriate antibody overnight at 4°C with rotation and then incubated for 2 h at 4°C with 30 μl of an affinity resin. The resin-bound proteins were eluted by boiling for 10 min in a reducing sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 0.1% bromophenol blue, 50 mM DTT, and 20 mM EDTA) or in the counterpart nonreducing buffer (without the DTT). For most types of IP performed—Vp1-IP, Hsp70-IP, Hsp40-IP, and β-galactosidase (β-Gal)–WB (against β-Gal)—the immunoprecipitating antibody was the same as the primary antibody for the corresponding types of Western blotting (see below) and the immunoprecipitating resin was protein G-Sepharose beads (GE Healthcare). For LT-IP (against unique epitopes of LT), the mouse monoclonal antibody SV40-Ab-2 (Oncogene Research Products) was used, followed by TrueBlot anti-mouse Ig beads (eBioscience).

Isolation of His-tagged proteins from cytoplasmic subfractions.

The isolation of His-tagged Vp1ΔC species from Sol and Csk fractions was performed under denaturing conditions as previously described (12), with modifications. An aliquot of the Sol or Csk fraction was first denatured with added SDS (1% final concentration) and DTT (2 mM final concentration) by incubating at 50°C for 10 min and then adjusted to the conditions of the denaturing binding buffer (6 M urea, 40 mM Tris-Cl, 150 mM NaCl, and 10% glycerol) at pH 7.6. This mixture was cleared of debris by centrifugation and then incubated at 4°C for 1 h with 30 μl of Ni-nitrilotriacetic acid (Ni-NTA) resin (50% slurry; Qiagen) that had been equilibrated with the denaturing binding buffer at pH 8.0. The reacted resin was collected and washed three times with the denaturing binding buffer at pH 8.0, except that the second wash contained a higher concentration of NaCl (1 M) and the third wash contained added 30 mM imidazole. The resin-bound proteins were eluted by boiling in reducing Laemmli sample buffer for 10 min and analyzed by SDS-PAGE and fluorography.

Diagonal SDS-PAGE.

Diagonal, or two-dimensional (2D), SDS-PAGE was performed as previously described (8), with modifications. For SDS-PAGE in the first dimension, protein separation was carried out under either reducing (1D-R) or nonreducing (1D-NR) conditions, with differing sample preparation and gel electrophoresis conditions, as described below. For SDS-PAGE in the second dimension (2D), individually excised 1D-R or 1D-NR lanes were incubated with reducing solution (62.5 mM Tris-Cl, pH 6.8, 0.5% SDS, 1 mM EDTA, 5% glycerol, and 50 mM DTT) at 45°C for 30 min with periodic agitation and then embedded on top of 12.5% polyacrylamide gels in reducing agarose (62.5 mM Tris-Cl, pH 6.8, 0.5% SDS, 1 mM EDTA, 50 mM DTT, and 1% agarose). The 2D gels were resolved at 50 mA for 3.5 to 4.5 h and then electroblotted onto a nitrocellulose membrane for the subsequent detection of Vp1 species by Western blotting. The molecular weights of the detected Vp1 species were estimated via the inclusion of the BenchMark prestained protein marker ladder (Life Technologies) in the diagonal SDS-PAGE procedure. For 1D-R, two different methods were used. In the first method, lysate samples were denatured and reduced by boiling in the Laemmli sample loading buffer (SLB) supplemented with 20 mM EDTA and 50 mM DTT (regular SLB) and then resolved in precast 4 to 15% Tris-glycine gradient gels (Bio-Rad). In the second method, NuPAGE LDS Sample Buffer (Life Technologies), which contains lithium dodecyl sulfate and has a pH of 8.4, was used. Lysate samples were boiled in this sample buffer at 1× strength with added 50 mM DTT (high-pH SLB) and loaded onto NuPAGE Novex 4 to 12% Bis-Tris gels (Life Technologies), and electrophoresis was conducted with running buffer containing 1× antioxidant (Life Technologies). For 1D-NR, the samples consisting of gradient fractions from denaturing sucrose sedimentation were precipitated with trichloroacetic acid (TCA), resuspended in a nonreducing sample buffer (67 mM HEPES, pH 6.8, 3% SDS, and 10% glycerol), incubated at 80°C for 10 min, and resolved in 4 to 15% Tris-glycine gradient gels (Bio-Rad).

WB and fluorography.

Lysate samples for SDS-PAGE and WB were prepared in Laemmli sample loading buffer containing 50 mM DTT as the reducing agent and additional 20 mM EDTA, except where different sample preparation conditions are noted for diagonal SDS-PAGE experiments. WB was otherwise performed as previously described (12, 44). The primary antibodies included the following: for Vp1-WB, affinity-purified rabbit polyclonal anti-Vp1 IgG (45); for LT-WB, the mouse monoclonal IgG PAb419 (against the LT/ST J domain) (46) or PAb423 (against a unique C-terminal epitope of LT) (46); for Hsp70-WB, the mouse monoclonal antibody SPA-810 (Assay Designs); for Hsp40-WB, the rabbit polyclonal antibody SPA-400 (Assay Designs); for FLAG-WB, the mouse monoclonal antibody F3165 (clone M2; Sigma-Aldrich); and for β-Gal–WB, a mouse monoclonal antibody against β-galactosidase (Promega). The secondary antibody was either horseradish peroxidase (HRP)-conjugated anti-rabbit or anti-mouse antibody (MP Biomedicals) or TrueBlot HRP-conjugated anti-rabbit or anti-mouse antibody (eBioscience).

For fluorography, [S]methionine-labeled samples were separated by SDS-PAGE and transferred onto a 0.2-μm nitrocellulose membrane, and the membrane was treated with Autofluor (National Diagnostics) for 1 h with shaking. The membrane was dried and exposed to X-ray film at −70°C.

RESULTS

Nonreducibly modified species of SV40 Vp1, mdVp1s, are present in Vp1ΔC-expressing COS-7 cells.

In previous studies, we have made observations that suggest the formation of nonreducible, covalently modified Vp1 intermediates during Vp1's folding and oligomerization in the cytoplasm. First, we have noted in SV40-infected TC7 cells the presence of a nonreducible, metabolically radiolabeled 78-kDa protein, which was immunoprecipitated by anti-Vp1 antibody (8). This Vp1-related protein was a minor species relative to the monomeric Vp1 (45 kDa) but increased during the chase period following the pulse-labeling as Vp1 oligomerization began (8). It was not a product of Vp1 oxidation or dimerization during or after cell lysis: the protein notably persisted even when the cells were exposed to the cell-permeable alkylating agent NEM during the 40-min chase period (8). Second, we have also reported the presence of three nonreducible minor Vp1 species, md2, md3, and md4 (collectively named the mdVp1s), associated with the expression of a Vp1 lacking the last 58 amino acids (Vp1ΔC) in COS-7 cells (12). These proteins migrated in SDS-PAGE gels with apparent mobilities of 45, 50, and 80 kDa, or more slowly than the major Vp1ΔC monomer of 40 kDa (denoted md1) (12). We speculate that the mdVp1s we observed in the Vp1ΔC expression system could be related to the 78-kDa species made during SV40 infection, though further analysis is required to confirm this.

It is conceivable that the mdVp1s—especially the 80-kDa md4, which has the expected size of a Vp1ΔC dimer—might in fact be disulfide-linked Vp1 species, formed either naturally in the cells or artifactually during isolation procedures. An incomplete reduction of their disulfide bridges during SDS-PAGE processing could have led to the detection of these apparently nonreducible species. To test these possibilities, we developed a diagonal SDS-PAGE procedure, which is illustrated in Fig. 1A. The lysate sample and the prestained protein marker were loaded onto adjacent lanes of the first-dimension gel and resolved under reducing conditions (1D-R). The resolved gel lanes were excised, re-treated with the reducing agent dithiothreitol, loaded together onto the second-dimension gel, and again resolved under reducing conditions (2D). This procedure allowed us to subject the same sample to reducing SDS-PAGE twice in a row; any disulfide-linked protein that failed to be reduced in the 1D-R analysis would be able to resolve into its component proteins in the 2D analysis, thereby confirming the completeness of disulfide reduction for the observed protein species. We performed this diagonal SDS-PAGE procedure for the whole-cell RIPA lysate of pCI-Vp1ΔC-transfected COS-7 cells and then examined the disappearance or persistence of the mdVp1s by Western blotting for Vp1 (Vp1-WB) (Fig. 1B and C, a, and and2a).2a). As seen in the Vp1-WB profile, md2, md3, and md4 all resolved along the diagonal of the 2D gel: none resolved into the md1 monomer or other components (Fig. 1B and C, a, and and2a).2a). We next wished to determine if mdVp1 species similar to what we have observed in the whole-cell RIPA lysate are present in cytoplasmic fractions. Two cytoplasmic subfractions—the cytoplasmic soluble fraction (Sol) and the cytoskeletal fraction (Csk)—were sequentially prepared by extraction with different nondenaturing detergents. When the same diagonal SDS-PAGE analysis was performed with the Sol and Csk of pCI-Vp1ΔC-transfected cells, all the mdVp1s were also present exclusively on the diagonal line (Fig. 1C, ,ee and andi,i, and and2e2e and andi).i). Note that in the procedure, the prestained marker proteins, visible as colored spots forming a diagonal line on each of the membranes, served to mark or document the migration patterns of variously sized proteins (Fig. 1C, ,a′,a′, e′, and i′). These results indicate that our 1D-R procedure achieved effective reduction of the mdVp1s.

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500001.jpg

Diagonal SDS-PAGE analysis of mdVp1s. (A) Scheme of sample loading. For diagonal SDS-PAGE (for details, see Materials and Methods), the lysate sample and the Benchmark prestained protein marker ladder were loaded onto adjacent lanes of the first-dimension gel under reducing conditions (1D-R), and the resolved gel lanes were excised and loaded together onto the second-dimension gel under reducing conditions (2D). The prestained marker proteins (represented by blue spots) are shown resolved in the 2D gel on a diagonal line. The symbols that are placed at the left of the blue spots identify the markers' molecular masses: ♢, 120 kDa; ▽, 85 kDa; ○, 50 kDa; and △, 40 kDa). (B) Detection of mdVp1s. RIPA lysate prepared from 1.2 × 10 COS-7 cells 48 h after transfection with pCI-Vp1ΔC was resolved by diagonal SDS-PAGE under reducing conditions in both the first (1D-R) and the second (2D) dimensions, followed by Vp1-WB. The mdVp1 species detected, md1, md2, md3, and md4, are marked above the 1D gel strip and on the right of the 2D image. (C) Estimation of molecular masses. Vp1ΔC-expressing COS-7 cells were either pretreated (+/− columns) or not pretreated (−/− and −/+ columns) with NEM before the preparation of lysates (RIPA, Sol, and Csk). The lysates were either treated (−/+ column) or not treated (−/− and −/+ columns) with NEM under denaturing conditions. The lysates were then resolved, along with Benchmark, under reducing conditions (1D-R), and the resulting lanes were treated with reducing agent and analyzed on new gels under reducing conditions (2D). The 2D gels were transferred onto nitrocellulose membranes and subjected to Vp1-WB to reveal the patterns of the mdVp1 species (a, b, e, f, i, and j: same as in Fig. 2). The membranes also documented the migration patterns of the marker protein bands as colored spots forming a diagonal line (a′, b′, e′, f′, i′, and j′). The apparent molecular masses of the mdVp1s were estimated by aligning the Vp1-WB profiles with their corresponding internal marker profiles. The sizes of the marker spots are designated by the same symbols as in panel A.

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500002.jpg

Nonreducible Vp1 species are formed in COS-7 cells and persist after NEM treatments and high-pH SDS-PAGE. Whole-cell lysate (RIPA) (a and b) or detergent-fractionated Sol and Csk fractions were prepared at 48 h after pCI-Vp1ΔC transfection from COS-7 cells. The cells were either pretreated (+/− columns) or not pretreated (−/− and −/+ columns) with NEM before the addition of lysis buffers, and the lysates were either treated (−/+ column) or not treated (−/− and +/− columns) with NEM under denaturing conditions. Aliquots of the lysate samples equivalent to 1.2 × 10 cells each were resolved by diagonal SDS-PAGE under reducing conditions in both dimensions. Some of the samples for the first-dimension SDS-PAGE were prepared in the NuPAGE LDS sample loading buffer with a pH of 8.4 and 50 mM DTT (high-pH), while others were prepared in the regular Laemmli sample buffer, also with 50 mM DTT (regular). The high-pH-loaded samples were run with NuPAGE 1× antioxidant in the running buffer. (a, b, e, f, i, and j) Western blot images. (c, d, g, h, k, and l) Pictograms prepared from the respective Western blots. The arrows indicate the directions of protein separation from high to low molecular masses. The positions of molecular mass standards, determined by the procedure described in the legend to Fig. 1C, are marked as a series of bars in the center, corresponding to 120, 85, 50, and 40 kDa from the top.

To further confirm that the mdVp1s were genuine nonreducible Vp1 species present in cells, we tested the effects of NEM treatment and alternative SDS-PAGE conditions on the detection of the mdVp1s. COS-7 cells transfected with pCI-Vp1ΔC were treated or not treated with NEM (2.5 mM for 2 min) before lysis, and the resulting lysates or subcellular fractions were either treated or not treated with NEM under denaturing (0.5% SDS) conditions. The lysates were then analyzed for the presence of mdVp1s by diagonal SDS-PAGE under reducing conditions, which was performed with the use of either a standard Laemmli-style neutral-pH sample loading buffer (regular SLB) or the alkaline sample loading buffer of the Invitrogen NuPAGE system (high-pH SLB), which is designed to completely reduce disulfide bonds and prevent their re-formation during electrophoresis. Overall, the profiles for the RIPA, Sol, and Csk extracts continued to include all the mdVp1 species, and the species continued to lie along the diagonal under all conditions tested (Fig. 1C and and2).2). The profiles obtained from the use of regular SLB (Fig. 1C, b, f, and j, and 2b, f, and j) and high-pH SLB (Fig. 1C, d, h, and l) were highly similar. We note that some mdVp1s had shifted electrophoretic mobilities when the cells were treated with NEM before lysis (Fig. 1C, b, f, and j, and 2b, f, and j or Fig. 1C, d, h, and l) or when the lysates were treated with NEM after lysis (Fig. 1C, c, g, and k, and 2c, g, and k), relative to the mobilities observed without either NEM treatment (Fig. 1C, a, e, and i, and 2a, e, and i). We assume that the differences arose from the addition of NEM's alkyl group to the mdVp1s. The fact that none of the mdVp1s from 1D-R gels resolved into md1 in the 2D analysis regardless of the SDS-PAGE buffer system used assured us of the effectiveness of our protein reduction and 1D-R and 2D gel methods. The fact that the mdVp1 species persisted in our analyses despite prelysis or postlysis NEM treatment shows that they are genuine nonreducible Vp1 species present in Vp1ΔC-expressing COS-7 cells and that their signal intensities as detected by Vp1-WB after 1D-R or 2D analysis can be taken to represent the species' relative abundances in the cells. We note, however, that this extent of information as conveyed by the diagonal patterns may apply only to the four specific mdVp1 species md1 through md4. We saw a few more putative mdVp1 spots in the Vp1-WB profiles shown in Fig. 1C and and22 than the four we have described; we have not attempted to address what might have brought about the differences.

MdVp1s accumulate over time after synthesis.

Having shown the genuine presence of mdVp1s in Vp1ΔC-expressing COS-7 cells, we wished to test the idea that these proteins play a role in the folding of Vp1. To do so, we examined whether mdVp1s are derived from the newly synthesized Vp1ΔC monomer by performing a pulse-chase analysis. At 20 h p.t. with pCI-Vp1ΔC, COS-7 cells were pulse-labeled with [S]methionine for 5 min and then either harvested immediately or chased in the presence of excess unlabeled methionine for 10 or 40 min before being processed for fractionation into Sol and Csk. All radiolabeled Vp1 species in the subcellular fractions were isolated by His tag affinity purification under denaturing conditions, resolved by reducing SDS-PAGE, and detected by fluorography (Fig. 3).

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500003.jpg

Synthesis of mdVp1s. Twenty hours after transfection with pCI-Vp1ΔC, COS-7 cells were pulse-labeled with [S]methionine for 5 min and either harvested immediately (P5′) or further incubated in medium with excess unlabeled methionine for 10 min (C10′) or for 40 min (C40′). As a control, untransfected COS-7 cells (Cont) were harvested after pulse-labeling. The Sol (lanes 1 to 4) and Csk (lanes 5 to 8) subcellular fractions were extracted from all sets of cells (6 × 10 cells per condition), and the radiolabeled Vp1s in an aliquot of each fraction (1/5 of Sol or 1/3 of Csk) were isolated by His tag affinity purification with Ni-NTA resin under denaturing conditions, resolved by SDS-PAGE, and visualized by autoradiography. The positions of the mdVp1s (md1, md1′, md2, md3, and md4) are marked on the right of each panel. On the left of the panels, host protein bands appearing in the control lanes are marked with asterisks, and molecular mass markers of 40 and 80 kDa are indicated with bars.

We observed the formation of all mdVp1 species in the course of the pulse-chase. In the Sol fraction, unmodified md1 was the major species that appeared immediately after pulse-labeling, though md4 was faintly detectable. The intensity of the md4 band increased over the chase periods (Fig. 2, lanes 2 to 4), whereas md2 and md3 were detected beginning from 10 min of chase (lanes 3 and 4). In the Csk fraction, md1 was also the major species after pulse-labeling (Fig. 3, lane 6). md2 and md4 were detected during the chase times (Fig. 3, lanes 7 to 8), but md3 was either absent in the Csk or was masked by copurified host proteins (bands marked by the lower two asterisks) in the same region of the gel. We detected in the Sol and Csk, particularly during chase times, an additional 42-kDa Vp1 species (md1′), which was not well resolved from md1 (Fig. 3, lanes 3, 4, 7, and 8). This species, rarely detectable in steady-state Vp1-WB analysis, is not addressed further in this study. We do not know the precise order in which individual mdVp1 species were formed. Nevertheless, the mdVp1s emerged and increased in the same time frame following the synthesis of Vp1ΔC as the folding and oligomerization of Vp1 during SV40 infection (8). At least some of the mdVp1s became associated with the Csk fraction of the cytoplasm, as do the previously characterized disulfide-bonded Vp1 folding intermediates (8). These results support the idea that the newly synthesized md1 Vp1 monomer undergoes posttranslational covalent protein modifications to form the md2, md3, and md4 species in the cytoplasm.

Levels of mdVp1s are elevated in cells that express SV40 LT.

The coexpression of Vp1 and LT/ST from the SV40-derived SV-Vp1 DNA is sufficient for the formation of minichromosome-containing virion-like particles in TC7 cells (5). However, it is difficult to discern a direct role of LT or ST in Vp1 folding and assembly in this system, because LT is expressed together with ST and is required to activate the Vp1 promoter in SV-Vp1 and to replicate the minichromosome. The possible participation of the SV40 oncoproteins in Vp1 folding is hinted at by our recent finding that the Vp1 core, Vp1ΔNΔC (aa 21 to 261), binds HSP70 in vitro only when ATP and LT/ST are present (12). Our cytomegalovirus (CMV) promoter-based Vp1ΔC expression plasmid, pCI-Vp1ΔC, could permit a more focused examination of LT/ST's role in Vp1 folding in the context of a living cell.

We first asked if the presence or absence of LT/ST in cells influences the production and accumulation of mdVp1s. Our approach was to examine the levels of mdVp1s from the expression of Vp1ΔC in cell lines that differ in their endogenous production of LT/ST: CV-1, U2OS, and 293H, which do not express LT or ST; COS-7, 293T, and 293TT, which express both LT and ST, with 293TT expressing LT in excess; and U2OS/T, which expresses LT but not ST. These cells were transfected with pCI-Vp1ΔC and extracted with RIPA buffer at 48 h p.t. The amounts of Vp1 and LT present in the lysates were quantified by dot blot analysis (Table 1). Based on the results of the quantification, aliquots of the cell lysates containing equal amounts of Vp1 were detected for their mdVp1 profiles by Vp1-WB (Fig. 4).

Table 1

Concentrations of Vp1 and LT in whole-cell lysates prepared 48 h after transfection with pCI-Vp1ΔC

Lysate cell typeaVp1 concn
LT concn
ng/μlbμg/10 cellscng/μlbμg/10 cellsc
CV-12.500.416
COS-716.02.6025.04.12
U2OS5.000.833
U2OS/T50.08.3350.08.33
293H0.220.005
293T2.220.05015.00.350d
293TT50.01.1975.01.78d
Transfection with pCI-Vp1ΔC and whole-cell lysate preparation were performed as described in Materials and Methods.
Concentrations were determined from lysate aliquots by slot blot analysis as described in Materials and Methods. Each value represents the average of three independent transfection experiments, with a standard error of less than ±0.25 ng/μl.
Values in micrograms per million cells were converted from the corresponding values in nanograms per microliter.
About 4.2 × 10 293T and 293TT cells and 6 × 10 COS-7 and U20S cells are present per 10-cm plate at 100% confluence. The lower amounts of LT per 10 cells for 293T and 293TT in the table reflect the smaller sizes of these cells.
An external file that holds a picture, illustration, etc.
Object name is zjv9990975500004.jpg

Higher levels of mdVp1s are found in LT-expressing cell lines. Whole-cell lysates were prepared from CV-1 (lanes 2 to 4), COS-7 (lanes 5 to 6), U2OS (lanes 8 to 10), and U2OS/T (lanes 11 to 13) cells 48 h after transfection with pCI-Vp1ΔC (WT) or with one of its mutant counterparts, C47A-C87A (mt1) or C87A-254A (mt2). For each cell type, lysate aliquots containing equal amounts of Vp1 were taken according to the quantitation results presented in Table 1 and subjected to Vp1-WB for the detection of mdVp1s. The positions of molecular mass standards in kilodaltons and of mdVp1 species md1 through md4 are marked on the left and on the right, respectively. As a control, untransfected CV-1 lysate (Cont) (lane 1) prepared from the same number of cells as the pCI-Vp1ΔC-transfected CV-1 lysate (lane 2) was analyzed.

The levels of mdVp1s detected in the pCI-Vp1ΔC-transfected cells clearly correlated with the absence or presence of LT in those cells. All samples contained unmodified monomer, md1, except for the control untransfected cell lysate (Fig. 3, lane 1), which lacked detectable Vp1 species. The lysates of transfected CV-1 and U2OS cells, which do not express LT or ST, had no detectable md2 and md3 and only a very low level of md4 (Fig. 4, WT, lanes 2 and 8). The lysate of 293H showed a similar scarcity of mdVp1s (not shown). Strikingly, the lysates COS-7 and U2OS/T, both of which express LT, contained substantially higher levels of md2, md3, and md4 (Fig. 4, lanes 5 and 11) than their LT-lacking counterparts (lanes 2 and 8). Also striking was the difference in total Vp1 content between cell types. Whereas only 0.42 μg and 0.83 μg of Vp1 per 10 cells were found for the non-LT-expressing CV-1 and U2OS, respectively, 2.6 μg and 8.3 μg, or 6 and 10 times the amounts, were found for the LT-expressing COS-7 and U2OS/T, respectively (Table 1). Likewise, the Vp1 content of the lysate increased from 293H to 293T to 293TT, in direct correlation with their LT contents (Table 1). That the LT-expressing U2OS/T cells supported levels of mdVp1 formation relative to the total amount of Vp1 to those of the dual LT/ST-expressing COS-7 cells suggests that the presence of LT, not ST, is responsible for promoting the formation of the mdVp1s, as well as stimulating the overall production of Vp1.

We next tested the effects on mdVp1 formation of two Vp1 cysteine pair mutations, mt1 (C49A-C87A) and mt2 (C49A-C254A), which cause defective Vp1 folding in the cytoplasm (10). mt1 and mt2 pCI-Vp1ΔC DNAs were transfected into CV-1, COS-7, U2OS, and U2OS/T cells, and their whole-cell lysates were then analyzed by Vp1-WB. Because the steady-state levels of the two mutant Vp1ΔCs were consistently lower than that of the wild-type Vp1ΔC for each cell type, analysis was performed using lysate aliquots with identical amounts of total Vp1, which was measured by the same dot blot procedure as before. The results show differing patterns of mdVp1 accumulation for the two mutants (Fig. 4). mt1 followed a pattern of Vp1 modification similar to that of wild-type Vp1ΔC, accumulating md2 through md4 in LT-expressing COS-7 and U2OS/T cells (Fig, 4, mt1, lanes 6 and 12) but not in LT-lacking CV-1 and U2OS cells (lanes 3 and 9). The fast-migrating 35-kDa Vp1 species (Fig. 4, lanes 3 and 9) is presumably a truncated form of mt1 Vp1ΔC.

mt2, however, displayed a very different pattern from those of the wild type and mt1. The presence of all modified forms was drastically reduced in both LT-lacking (Fig. 4, mt2, lanes 4 and 10) and LT-expressing (lanes 7 and 13) cells, while the prominence of the 35-kDa truncated species was greatly increased, suggesting a greater tendency of the mt2 Vp1ΔC to undergo degradation. Analysis performed with 293H, 293T, and 293TT cells confirmed the differing patterns of LT correlation with Vp1 modification for wild-type, mt1, and mt2 Vp1ΔCs (data not shown). These results imply that different steps in normal Vp1 folding are perturbed for the two mutants, resulting in mt2, but not mt1, either losing the ability to produce md2, md3, and md4 or making highly unstable forms of those species.

Covalent modification occurs within the core of Vp1, amino acids 21 to 261, in the cytoplasm.

To determine the region(s) of Vp1 in which covalent modifications occur, we expressed in COS-7 cells a series of N-terminally FLAG-tagged Vp1s with different deletions and examined their abilities to form mdVp1 species. Whole-cell lysates were prepared from the expression plasmid-transfected cells and analyzed by anti-FLAG Western blotting (FLAG-WB), which is expected to detect all tagged species equally. For the control Vp1ΔC lysate (ΔC58), the modified forms md2 and md4 were detected, along with the major unmodified md1 (Fig. 5A, lane 1). The removal of 20 residues from the N terminus (ΔNΔC58) or of an additional 42 residues from the C terminus (ΔC100) did not affect the formation of the modified species, and the shifts in banding pattern compared with the Vp1ΔC profile were consistent with the sizes of the respective deletions (Fig. 5A, lanes 2 and 3). The continued presence of the modified forms indicates that the modifications occur within the core region of Vp1 (aa 21 to 261). We note that the Vp1ΔC100 sample included a fragment much smaller than the unmodified Vp1ΔC100 monomer (Fig. 5A, lane 3). This fragment is presumed to be a proteolytic product and may have occurred because of the altered conformation of the truncated Vp1.

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500005.jpg

Existence of covalently modified species of polyomavirus Vp1s and HPV L1s. (A) Covalent modifications are located in the core of SV40 Vp1. Whole-cell lysates, each equivalent to 6 × 10 cells, were prepared from COS-7 cells 48 h after transfection with pCI-FLAG-Vp1ΔC and its derivative plasmids encoding FLAG-Vp1ΔC (ΔC58) (lane 1), FLAG-Vp1ΔNΔC (ΔNΔC58, further missing the N-terminal 20 residues) (lane 2), and FLAG-Vp1ΔC100 (ΔC100) (lane 3). The lysates were subjected to FLAG-WB for the detection of covalently modified Vp1s. (B) Subcellular localization of FLAG-tagged SV40 Vp1s. TC-7 cells were transfected with pCI-FLAG-Vp1ΔC plasmids encoding FLAG-Vp1ΔC (i and ii) and FLAG-Vp1ΔNΔC (iii and iv), fixed, and subjected to immunofluorescence microscopy for FLAG (i and iii) and DAPI staining (ii and iv). (C and D) Covalent modifications in polyomavirus and HPV major capsid proteins are detected in COS-7 cells, but not in CV-1 cells. Whole-cell lysates, equivalent to 6 × 10 cells each, were prepared from COS-7 (C) or CV-1 (D) cells transfected with the pCI plasmids expressing the FLAG-tagged Vp1ΔCs of SV40 (lanes 1), JCV (lanes 2), BKV (lanes 3), and MCV (lanes 4) or the FLAG-tagged L1ΔCs of HPV16 (lanes 5) and HPV18 (lanes 6). The lysates were examined for the presence of FLAG-tagged, covalently modified species as in panel A.

The N-terminal 20 residues of Vp1 constitute a bipartite NLS (3). Consistent with this, FLAG-tagged Vp1ΔC primarily localized to the nucleus (Fig. 5B, ,ii and ii), while FLAG-tagged Vp1ΔNΔC, lacking the NLS, was mainly cytoplasmic (iii and iv). The fact that the cytoplasmically localized Vp1ΔNΔC underwent nonreducible covalent modifications indicates that the modifications occur in the cytoplasm, consistent with the detection of the modified species in cytoplasmic subfractions (Fig. 1, ,2,2, and and33).

Polyomavirus Vp1s and HPV L1s form nonreducible modified species in LT-expressing cells.

We next asked if the Vp1s of other members of the polyomavirus family or the L1s of the HPV family can make covalently modified species when expressed in COS-7 cells. The various Vp1s and L1s, with their C termini removed to limit their assembly beyond pentamers, were expressed as FLAG-tagged proteins from plasmids that were converted through coding fragment exchanges from the plasmid that expresses SV40 Vp1ΔC, pCI-FLAG-Vp1ΔC. The various plasmids were transfected into COS-7 or CV-1 cells, and RIPA lysates were prepared and examined for the pattern of Vp1 or L1 species by FLAG-WB.

Besides the main Vp1 or L1 monomers, more slowly migrating FLAG-tagged species were generally detected only in COS-7 cells (Fig. 5C). In CV-1 cells (Fig. 5D), the only modified species detected was the 50-kDa derivative of JCV Vp1ΔC (lane 2). In COS-7 cells, various modified species were found for each virus: for JCV Vp1ΔC (aa 1 to 293), 50-, 57- (minor), and 90-kDa species were seen, in addition to the 40-kDa monomer (Fig. 5C, lane 2); for BKV Vp1ΔC (aa 1 to 301), 55-kDa and 88-kDa species, in addition to the 44-kDa monomer (lane 3); and for MCV Vp1ΔC (aa 1 to 325), 47- and 80-kDa species, in addition to the 42-kDa monomer (lane 4). The mobility shifts from the monomer to the modified species in the JCV and BKV profiles were very similar to those found in the SV40 profile. The modified species in the MCV profile resemble md2 and md4 of SV40 Vp1. For HPV16 and HPV18 L1ΔCs (aa 1 to 404), we found a shifted 50-kDa species and a minor 100-kDa species besides the 48-kDa monomer (lanes 5 and 6). Thus, our observations correlate the presence of SV40 LT with the presence of covalently modified Vp1 or L1 species among the human polyomaviruses and papillomaviruses, suggesting the intriguing possibility that all these covalent modifications arise from common mechanisms involving LT and some shared molecular features of these different major capsid proteins.

mdVp1s form multiprotein complexes with LT, Hsp70, and Hsp40.

In SV40 infection, Vp1 interacts with LT, Hsp70, and Hsp40 in the cytoplasm (12). We tested whether these proteins associate as multiprotein complexes in U2OS/T cells transfected with pCI-Vp1ΔC. The whole-cell lysate was fractionated by sedimentation through neutral sucrose gradients and examined for the presence and types of mdVp1s by Vp1-WB and for the presence of LT by anti-LT Western blotting (LT-WB). We found all types of mdVp1s throughout the gradient, fractions 4 through 8 of which contained protein complexes larger than 1,000 kDa (Fig. 6A, wt). Thus, a substantial portion of all mdVp1s were in association either among themselves, with many cellular proteins, or with LT. LT was present throughout the gradient whether Vp1ΔC was coexpressed (Fig. 6B, wt) or not (Fig. 6B, Cont; control U2OS/T). We estimated through anti-Vp1 immunoprecipitation analysis that about 1 to 2% of the total LT in the unfractionated whole-cell lysate of pCI-Vp1ΔC-transfected COS-7 cells was in complex with Vp1ΔC (data not shown).

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500006.jpg

MdVp1s form noncovalent multiprotein complexes that include LT and cellular chaperones. Whole-cell lysate was prepared from 3 × 10 U2OS/T cells 48 h after transfection with pCI-Vp1ΔC (wt) or with the mutant counterpart C49A-C87A (mt1) or C87A-C254A (mt2) plasmid. Control lysate was prepared from the same number of untransfected U2OS/T cells (Cont). Each lysate was fractionated by sedimentation through a neutral 5 to 32% sucrose gradient and collected in 15 fractions. (A) Reduced noncovalent complex formation by mutant Vp1ΔCs. For wt, mt1, and mt2 gradients, 1/30 of each even fraction was analyzed by Vp1-WB. The positions of md1 through md4 are marked on the right of each panel. (B) LT cofractionation with Vp1ΔC. For control (Cont) and wt gradients, 1/6 of each even fraction was analyzed by LT-WB using PAb419. (C) Reduced mt2 Vp1ΔC interaction with LT. For wt or mutant (mt1 and mt2) gradients, 1/12 or 1/8, respectively, of select even fractions was mixed and subjected to LT-IP using Ab-2, followed by LT-WB (LT) or Vp1-WB (Vp1). The analyzed samples included the fraction 4-fraction 6 mixtures (lanes 2, 5, and 8), the fraction 8-fraction 10 mixtures (lanes 3, 6, and 9), the fraction 12-fraction 14 mixtures (lanes 4, 7, and 10), and the control whole-cell lysate (lane 1). (D) Reduced association of mutant Vp1ΔCs with cellular chaperones. For wt, mt1, and mt2 gradients, 1/3 of fractions 3, 5, and 7 were pooled and subjected to Vp1-IP (Vp1) (lane 1), Hsp70-IP (Hsp70) (lane 2), Hsp40-IP (Hsp40) (lane 3), or β-Gal–IP (β-gal) (lane 4), followed by Vp1-WB. The Vp1 present in 1/10 of each input lysate is shown in lane 5. (E) Association of Vp1ΔC with cellular chaperones. For the wt gradient, the same pooled fraction was analyzed by the same IP types as in panel D, followed by Hsp40-WB (Hsp40) or Hsp70-WB (Hsp70). The Hsp40 or Hsp70 present in 1/2 of the input lysate is shown in lane 5.

We next probed for the presence of Hsp70, Hsp40, and LT in the multiprotein complexes by analyzing different portions of the sucrose gradient by a series of IPs. Wild-type Vp1ΔC coprecipitated with LT from pooled fractions 4 and 6, 8 and 10, and 12 and 14 (Fig. 6C, wt Vp1, lanes 2 to 4) and also with Hsp40 and Hsp70 from pooled fractions 3, 5, and 7 (Fig. 6D, wt, lanes 2 and 3), but not with anti-β-Gal antibody (lane 4). The specificity of the coprecipitations can be seen from the absence of Hsp40 in the Hsp70-IP complex (Fig. 6E, lane 2) and vice versa (lane 3) and from the absence of both Hsp70 and Hsp40 in the β-Gal–IP complex (lane 4). Wild-type Vp1ΔC formed noncovalent multiprotein complexes that included LT, Hsp70, and Hsp40.

The mt1 and mt2 mutants of Vp1ΔC exhibited different profiles of association with various proteins. mt1 Vp1ΔC, sedimenting mainly in fractions 8 through 14 (Fig. 6A, mt1), evidently formed some 1,000-kDa-plus protein complexes, though to a lesser extent than wild-type Vp1ΔC (Fig. 6A, wt). By using more of the mt1 pooled fractions than the wild-type ones for the LT-IP analysis, we observed coprecipitation of mt1 Vp1ΔC with LT in all three fraction pools (Fig. 6C, Vp1, lanes 5 to 7) and with Hsp70 and Hsp40 in the combined fractions 3, 5, and 7 (Fig. 6D, mt1, lanes 2 and 3).

mt2 Vp1, on the other hand, sedimented only in fractions 10 through 14, suggesting a severely reduced extent of protein association and an inability to form large multiprotein complexes (Fig. 6A, mt2). Little mt2 Vp1 coprecipitated with LT (Fig. 6C, Vp1, lanes 8 and 9) other than from fractions 12 and 14 (lane 10), which are expected to contain proteins of less than 150 kDa. Mt2 Vp1ΔC was not detected in the Hsp70-IP or Hsp40-IP complexes (Fig. 6D, mt2, lanes 2 and 3) derived from pooled lanes 3, 5, and 7, as can be expected from the lack of mt2 Vp1 in these fractions (Fig. 6A, mt2). None of the Vp1ΔC proteins were found in the β-Gal–IP complex (Fig. 6D, lane 4). The fact that mt1, but not mt2, Vp1ΔC was in large multiprotein complexes with LT, Hsp70, and Hsp40 suggests a role for LT-Vp1 interaction in the formation of multiprotein complexes that include cellular chaperones. LT may well be able to associate with mt2 Vp1 to some extent, but the mt2-LT complex apparently cannot further associate with cellular proteins that normally assist in the folding or oligomerization of Vp1.

A fraction of md4 is covalently linked to md1.

Given that the mdVp1s form multiprotein complexes either among themselves or with other proteins in the cell (Fig. 6), we explored the possibility that some of mdVp1's protein associations involve disulfide linkages. First, we examined the whole-cell lysate of pCI-Vp1ΔC transfected U2OS/T by denaturing through sucrose gradient sedimentation, followed by Vp1-WB under reducing conditions (Fig. 7A). In contrast to the broad mdVp1 distribution seen in a nondenaturing gradient, all mdVp1s were found in fractions 14 through 16 after sedimentation under denaturing conditions (compare Fig. 7A with Fig. 6A, wt). Hence, the multiprotein complexes found in fractions 4 through 8 of the nondenaturing gradient (Fig. 6A) must have involved largely noncovalent associations.

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500007.jpg

Formation of covalent complexes by mdVp1s. (A) Sedimentation profile of covalently linked Vp1 complexes. Whole-cell lysate was prepared from 3 × 10 U2OS/T cells transfected for 48 h with pCI-Vp1ΔC. The lysate was incubated with NEM and SDS, fractionated through a denaturing 5 to 20% sucrose gradient, and collected in 16 fractions. One-thirtieth of the even fractions was subjected to Vp1-WB. (B) A fraction of md4 is covalently linked to md1. One-sixth of fractions 9 and 11 from the denaturing gradient in panel A was combined, precipitated with TCA, and resuspended in 20 μl of 10 mM Tris-Cl, pH 8.0. Then, diagonal SDS-PAGE was performed under nonreducing conditions in the first dimension (1D-NR) (upper gel strip) and under reducing conditions in the second dimension (2D) (bottom). Identical gel strips from the 1D-NR stage and after 2D processing were subjected to Vp1-WB for the detection of disulfide-linked and fully reduced mdVp1 species, respectively. a, b, and c represent md1, while d and e represent md4. The positions of molecular mass standards are indicated for the 1D-NR strip.

Next, we looked for the presence of disulfide-linked mdVp1 complexes in the Vp1ΔC-expressing U2OS/T lysate. We have previously detected disulfide-linked Vp1 oligomers in fractions 8 through 12 of a denaturing, but nonreducing, 5 to 20% sucrose gradient (8). Under the same conditions, few mdVp1s were detected in the corresponding fractions (Fig. 7A, lanes 4 to 6). We concentrated the combined fractions 9 and 11, in which we barely detected Vp1, and analyzed the sample by a modified diagonal SDS-PAGE, followed by Vp1-WB detection for the mdVp1s. The proteins here were first separated by nonreducing SDS-PAGE (1D-NR) to dissociate noncovalently associated proteins while preserving disulfide-bonded complexes. The way 1D-NR is run is different from 1D-R of the diagonal SDS-PAGE analysis described in Fig. 1. The disulfide-linked protein complexes were then resolved by reducing SDS-PAGE (2D) into component proteins. Proteins resolved from disulfide-linked complexes would be recognized as off-diagonal spots on the 2D gel. While cellular proteins linked to mdVp1s would escape detection at this stage, Vp1-WB would reveal mdVp1s linked to themselves. As seen positioned on the diagonal (Fig. 7B), the most prominent noncomplexed mdVp1s were the 40-kDa md1 monomer (species a) and the 80-kDa md4 (species d). Md1 was also found off the diagonal as a component of two disulfide-linked complexes with mobilities of about 100 kDa (Fig. 7B, species b) and 220 kDa (species c). The other component of the 100-kDa complex appears to be the off-diagonal md4 (species e). The 220-kDa complex either could be a multimer of md1 or could consist of one or more other components not detected by Vp1-WB. We did not observe the disulfide association of either md2 or md3 with other mdVp1s. Thus, our observations indicate that a fraction of md4 is disulfide linked to md1.

Nonreducibly modified species of SV40 Vp1, mdVp1s, are present in Vp1ΔC-expressing COS-7 cells.

In previous studies, we have made observations that suggest the formation of nonreducible, covalently modified Vp1 intermediates during Vp1's folding and oligomerization in the cytoplasm. First, we have noted in SV40-infected TC7 cells the presence of a nonreducible, metabolically radiolabeled 78-kDa protein, which was immunoprecipitated by anti-Vp1 antibody (8). This Vp1-related protein was a minor species relative to the monomeric Vp1 (45 kDa) but increased during the chase period following the pulse-labeling as Vp1 oligomerization began (8). It was not a product of Vp1 oxidation or dimerization during or after cell lysis: the protein notably persisted even when the cells were exposed to the cell-permeable alkylating agent NEM during the 40-min chase period (8). Second, we have also reported the presence of three nonreducible minor Vp1 species, md2, md3, and md4 (collectively named the mdVp1s), associated with the expression of a Vp1 lacking the last 58 amino acids (Vp1ΔC) in COS-7 cells (12). These proteins migrated in SDS-PAGE gels with apparent mobilities of 45, 50, and 80 kDa, or more slowly than the major Vp1ΔC monomer of 40 kDa (denoted md1) (12). We speculate that the mdVp1s we observed in the Vp1ΔC expression system could be related to the 78-kDa species made during SV40 infection, though further analysis is required to confirm this.

It is conceivable that the mdVp1s—especially the 80-kDa md4, which has the expected size of a Vp1ΔC dimer—might in fact be disulfide-linked Vp1 species, formed either naturally in the cells or artifactually during isolation procedures. An incomplete reduction of their disulfide bridges during SDS-PAGE processing could have led to the detection of these apparently nonreducible species. To test these possibilities, we developed a diagonal SDS-PAGE procedure, which is illustrated in Fig. 1A. The lysate sample and the prestained protein marker were loaded onto adjacent lanes of the first-dimension gel and resolved under reducing conditions (1D-R). The resolved gel lanes were excised, re-treated with the reducing agent dithiothreitol, loaded together onto the second-dimension gel, and again resolved under reducing conditions (2D). This procedure allowed us to subject the same sample to reducing SDS-PAGE twice in a row; any disulfide-linked protein that failed to be reduced in the 1D-R analysis would be able to resolve into its component proteins in the 2D analysis, thereby confirming the completeness of disulfide reduction for the observed protein species. We performed this diagonal SDS-PAGE procedure for the whole-cell RIPA lysate of pCI-Vp1ΔC-transfected COS-7 cells and then examined the disappearance or persistence of the mdVp1s by Western blotting for Vp1 (Vp1-WB) (Fig. 1B and C, a, and and2a).2a). As seen in the Vp1-WB profile, md2, md3, and md4 all resolved along the diagonal of the 2D gel: none resolved into the md1 monomer or other components (Fig. 1B and C, a, and and2a).2a). We next wished to determine if mdVp1 species similar to what we have observed in the whole-cell RIPA lysate are present in cytoplasmic fractions. Two cytoplasmic subfractions—the cytoplasmic soluble fraction (Sol) and the cytoskeletal fraction (Csk)—were sequentially prepared by extraction with different nondenaturing detergents. When the same diagonal SDS-PAGE analysis was performed with the Sol and Csk of pCI-Vp1ΔC-transfected cells, all the mdVp1s were also present exclusively on the diagonal line (Fig. 1C, ,ee and andi,i, and and2e2e and andi).i). Note that in the procedure, the prestained marker proteins, visible as colored spots forming a diagonal line on each of the membranes, served to mark or document the migration patterns of variously sized proteins (Fig. 1C, ,a′,a′, e′, and i′). These results indicate that our 1D-R procedure achieved effective reduction of the mdVp1s.

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500001.jpg

Diagonal SDS-PAGE analysis of mdVp1s. (A) Scheme of sample loading. For diagonal SDS-PAGE (for details, see Materials and Methods), the lysate sample and the Benchmark prestained protein marker ladder were loaded onto adjacent lanes of the first-dimension gel under reducing conditions (1D-R), and the resolved gel lanes were excised and loaded together onto the second-dimension gel under reducing conditions (2D). The prestained marker proteins (represented by blue spots) are shown resolved in the 2D gel on a diagonal line. The symbols that are placed at the left of the blue spots identify the markers' molecular masses: ♢, 120 kDa; ▽, 85 kDa; ○, 50 kDa; and △, 40 kDa). (B) Detection of mdVp1s. RIPA lysate prepared from 1.2 × 10 COS-7 cells 48 h after transfection with pCI-Vp1ΔC was resolved by diagonal SDS-PAGE under reducing conditions in both the first (1D-R) and the second (2D) dimensions, followed by Vp1-WB. The mdVp1 species detected, md1, md2, md3, and md4, are marked above the 1D gel strip and on the right of the 2D image. (C) Estimation of molecular masses. Vp1ΔC-expressing COS-7 cells were either pretreated (+/− columns) or not pretreated (−/− and −/+ columns) with NEM before the preparation of lysates (RIPA, Sol, and Csk). The lysates were either treated (−/+ column) or not treated (−/− and −/+ columns) with NEM under denaturing conditions. The lysates were then resolved, along with Benchmark, under reducing conditions (1D-R), and the resulting lanes were treated with reducing agent and analyzed on new gels under reducing conditions (2D). The 2D gels were transferred onto nitrocellulose membranes and subjected to Vp1-WB to reveal the patterns of the mdVp1 species (a, b, e, f, i, and j: same as in Fig. 2). The membranes also documented the migration patterns of the marker protein bands as colored spots forming a diagonal line (a′, b′, e′, f′, i′, and j′). The apparent molecular masses of the mdVp1s were estimated by aligning the Vp1-WB profiles with their corresponding internal marker profiles. The sizes of the marker spots are designated by the same symbols as in panel A.

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500002.jpg

Nonreducible Vp1 species are formed in COS-7 cells and persist after NEM treatments and high-pH SDS-PAGE. Whole-cell lysate (RIPA) (a and b) or detergent-fractionated Sol and Csk fractions were prepared at 48 h after pCI-Vp1ΔC transfection from COS-7 cells. The cells were either pretreated (+/− columns) or not pretreated (−/− and −/+ columns) with NEM before the addition of lysis buffers, and the lysates were either treated (−/+ column) or not treated (−/− and +/− columns) with NEM under denaturing conditions. Aliquots of the lysate samples equivalent to 1.2 × 10 cells each were resolved by diagonal SDS-PAGE under reducing conditions in both dimensions. Some of the samples for the first-dimension SDS-PAGE were prepared in the NuPAGE LDS sample loading buffer with a pH of 8.4 and 50 mM DTT (high-pH), while others were prepared in the regular Laemmli sample buffer, also with 50 mM DTT (regular). The high-pH-loaded samples were run with NuPAGE 1× antioxidant in the running buffer. (a, b, e, f, i, and j) Western blot images. (c, d, g, h, k, and l) Pictograms prepared from the respective Western blots. The arrows indicate the directions of protein separation from high to low molecular masses. The positions of molecular mass standards, determined by the procedure described in the legend to Fig. 1C, are marked as a series of bars in the center, corresponding to 120, 85, 50, and 40 kDa from the top.

To further confirm that the mdVp1s were genuine nonreducible Vp1 species present in cells, we tested the effects of NEM treatment and alternative SDS-PAGE conditions on the detection of the mdVp1s. COS-7 cells transfected with pCI-Vp1ΔC were treated or not treated with NEM (2.5 mM for 2 min) before lysis, and the resulting lysates or subcellular fractions were either treated or not treated with NEM under denaturing (0.5% SDS) conditions. The lysates were then analyzed for the presence of mdVp1s by diagonal SDS-PAGE under reducing conditions, which was performed with the use of either a standard Laemmli-style neutral-pH sample loading buffer (regular SLB) or the alkaline sample loading buffer of the Invitrogen NuPAGE system (high-pH SLB), which is designed to completely reduce disulfide bonds and prevent their re-formation during electrophoresis. Overall, the profiles for the RIPA, Sol, and Csk extracts continued to include all the mdVp1 species, and the species continued to lie along the diagonal under all conditions tested (Fig. 1C and and2).2). The profiles obtained from the use of regular SLB (Fig. 1C, b, f, and j, and 2b, f, and j) and high-pH SLB (Fig. 1C, d, h, and l) were highly similar. We note that some mdVp1s had shifted electrophoretic mobilities when the cells were treated with NEM before lysis (Fig. 1C, b, f, and j, and 2b, f, and j or Fig. 1C, d, h, and l) or when the lysates were treated with NEM after lysis (Fig. 1C, c, g, and k, and 2c, g, and k), relative to the mobilities observed without either NEM treatment (Fig. 1C, a, e, and i, and 2a, e, and i). We assume that the differences arose from the addition of NEM's alkyl group to the mdVp1s. The fact that none of the mdVp1s from 1D-R gels resolved into md1 in the 2D analysis regardless of the SDS-PAGE buffer system used assured us of the effectiveness of our protein reduction and 1D-R and 2D gel methods. The fact that the mdVp1 species persisted in our analyses despite prelysis or postlysis NEM treatment shows that they are genuine nonreducible Vp1 species present in Vp1ΔC-expressing COS-7 cells and that their signal intensities as detected by Vp1-WB after 1D-R or 2D analysis can be taken to represent the species' relative abundances in the cells. We note, however, that this extent of information as conveyed by the diagonal patterns may apply only to the four specific mdVp1 species md1 through md4. We saw a few more putative mdVp1 spots in the Vp1-WB profiles shown in Fig. 1C and and22 than the four we have described; we have not attempted to address what might have brought about the differences.

MdVp1s accumulate over time after synthesis.

Having shown the genuine presence of mdVp1s in Vp1ΔC-expressing COS-7 cells, we wished to test the idea that these proteins play a role in the folding of Vp1. To do so, we examined whether mdVp1s are derived from the newly synthesized Vp1ΔC monomer by performing a pulse-chase analysis. At 20 h p.t. with pCI-Vp1ΔC, COS-7 cells were pulse-labeled with [S]methionine for 5 min and then either harvested immediately or chased in the presence of excess unlabeled methionine for 10 or 40 min before being processed for fractionation into Sol and Csk. All radiolabeled Vp1 species in the subcellular fractions were isolated by His tag affinity purification under denaturing conditions, resolved by reducing SDS-PAGE, and detected by fluorography (Fig. 3).

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500003.jpg

Synthesis of mdVp1s. Twenty hours after transfection with pCI-Vp1ΔC, COS-7 cells were pulse-labeled with [S]methionine for 5 min and either harvested immediately (P5′) or further incubated in medium with excess unlabeled methionine for 10 min (C10′) or for 40 min (C40′). As a control, untransfected COS-7 cells (Cont) were harvested after pulse-labeling. The Sol (lanes 1 to 4) and Csk (lanes 5 to 8) subcellular fractions were extracted from all sets of cells (6 × 10 cells per condition), and the radiolabeled Vp1s in an aliquot of each fraction (1/5 of Sol or 1/3 of Csk) were isolated by His tag affinity purification with Ni-NTA resin under denaturing conditions, resolved by SDS-PAGE, and visualized by autoradiography. The positions of the mdVp1s (md1, md1′, md2, md3, and md4) are marked on the right of each panel. On the left of the panels, host protein bands appearing in the control lanes are marked with asterisks, and molecular mass markers of 40 and 80 kDa are indicated with bars.

We observed the formation of all mdVp1 species in the course of the pulse-chase. In the Sol fraction, unmodified md1 was the major species that appeared immediately after pulse-labeling, though md4 was faintly detectable. The intensity of the md4 band increased over the chase periods (Fig. 2, lanes 2 to 4), whereas md2 and md3 were detected beginning from 10 min of chase (lanes 3 and 4). In the Csk fraction, md1 was also the major species after pulse-labeling (Fig. 3, lane 6). md2 and md4 were detected during the chase times (Fig. 3, lanes 7 to 8), but md3 was either absent in the Csk or was masked by copurified host proteins (bands marked by the lower two asterisks) in the same region of the gel. We detected in the Sol and Csk, particularly during chase times, an additional 42-kDa Vp1 species (md1′), which was not well resolved from md1 (Fig. 3, lanes 3, 4, 7, and 8). This species, rarely detectable in steady-state Vp1-WB analysis, is not addressed further in this study. We do not know the precise order in which individual mdVp1 species were formed. Nevertheless, the mdVp1s emerged and increased in the same time frame following the synthesis of Vp1ΔC as the folding and oligomerization of Vp1 during SV40 infection (8). At least some of the mdVp1s became associated with the Csk fraction of the cytoplasm, as do the previously characterized disulfide-bonded Vp1 folding intermediates (8). These results support the idea that the newly synthesized md1 Vp1 monomer undergoes posttranslational covalent protein modifications to form the md2, md3, and md4 species in the cytoplasm.

Levels of mdVp1s are elevated in cells that express SV40 LT.

The coexpression of Vp1 and LT/ST from the SV40-derived SV-Vp1 DNA is sufficient for the formation of minichromosome-containing virion-like particles in TC7 cells (5). However, it is difficult to discern a direct role of LT or ST in Vp1 folding and assembly in this system, because LT is expressed together with ST and is required to activate the Vp1 promoter in SV-Vp1 and to replicate the minichromosome. The possible participation of the SV40 oncoproteins in Vp1 folding is hinted at by our recent finding that the Vp1 core, Vp1ΔNΔC (aa 21 to 261), binds HSP70 in vitro only when ATP and LT/ST are present (12). Our cytomegalovirus (CMV) promoter-based Vp1ΔC expression plasmid, pCI-Vp1ΔC, could permit a more focused examination of LT/ST's role in Vp1 folding in the context of a living cell.

We first asked if the presence or absence of LT/ST in cells influences the production and accumulation of mdVp1s. Our approach was to examine the levels of mdVp1s from the expression of Vp1ΔC in cell lines that differ in their endogenous production of LT/ST: CV-1, U2OS, and 293H, which do not express LT or ST; COS-7, 293T, and 293TT, which express both LT and ST, with 293TT expressing LT in excess; and U2OS/T, which expresses LT but not ST. These cells were transfected with pCI-Vp1ΔC and extracted with RIPA buffer at 48 h p.t. The amounts of Vp1 and LT present in the lysates were quantified by dot blot analysis (Table 1). Based on the results of the quantification, aliquots of the cell lysates containing equal amounts of Vp1 were detected for their mdVp1 profiles by Vp1-WB (Fig. 4).

Table 1

Concentrations of Vp1 and LT in whole-cell lysates prepared 48 h after transfection with pCI-Vp1ΔC

Lysate cell typeaVp1 concn
LT concn
ng/μlbμg/10 cellscng/μlbμg/10 cellsc
CV-12.500.416
COS-716.02.6025.04.12
U2OS5.000.833
U2OS/T50.08.3350.08.33
293H0.220.005
293T2.220.05015.00.350d
293TT50.01.1975.01.78d
Transfection with pCI-Vp1ΔC and whole-cell lysate preparation were performed as described in Materials and Methods.
Concentrations were determined from lysate aliquots by slot blot analysis as described in Materials and Methods. Each value represents the average of three independent transfection experiments, with a standard error of less than ±0.25 ng/μl.
Values in micrograms per million cells were converted from the corresponding values in nanograms per microliter.
About 4.2 × 10 293T and 293TT cells and 6 × 10 COS-7 and U20S cells are present per 10-cm plate at 100% confluence. The lower amounts of LT per 10 cells for 293T and 293TT in the table reflect the smaller sizes of these cells.
An external file that holds a picture, illustration, etc.
Object name is zjv9990975500004.jpg

Higher levels of mdVp1s are found in LT-expressing cell lines. Whole-cell lysates were prepared from CV-1 (lanes 2 to 4), COS-7 (lanes 5 to 6), U2OS (lanes 8 to 10), and U2OS/T (lanes 11 to 13) cells 48 h after transfection with pCI-Vp1ΔC (WT) or with one of its mutant counterparts, C47A-C87A (mt1) or C87A-254A (mt2). For each cell type, lysate aliquots containing equal amounts of Vp1 were taken according to the quantitation results presented in Table 1 and subjected to Vp1-WB for the detection of mdVp1s. The positions of molecular mass standards in kilodaltons and of mdVp1 species md1 through md4 are marked on the left and on the right, respectively. As a control, untransfected CV-1 lysate (Cont) (lane 1) prepared from the same number of cells as the pCI-Vp1ΔC-transfected CV-1 lysate (lane 2) was analyzed.

The levels of mdVp1s detected in the pCI-Vp1ΔC-transfected cells clearly correlated with the absence or presence of LT in those cells. All samples contained unmodified monomer, md1, except for the control untransfected cell lysate (Fig. 3, lane 1), which lacked detectable Vp1 species. The lysates of transfected CV-1 and U2OS cells, which do not express LT or ST, had no detectable md2 and md3 and only a very low level of md4 (Fig. 4, WT, lanes 2 and 8). The lysate of 293H showed a similar scarcity of mdVp1s (not shown). Strikingly, the lysates COS-7 and U2OS/T, both of which express LT, contained substantially higher levels of md2, md3, and md4 (Fig. 4, lanes 5 and 11) than their LT-lacking counterparts (lanes 2 and 8). Also striking was the difference in total Vp1 content between cell types. Whereas only 0.42 μg and 0.83 μg of Vp1 per 10 cells were found for the non-LT-expressing CV-1 and U2OS, respectively, 2.6 μg and 8.3 μg, or 6 and 10 times the amounts, were found for the LT-expressing COS-7 and U2OS/T, respectively (Table 1). Likewise, the Vp1 content of the lysate increased from 293H to 293T to 293TT, in direct correlation with their LT contents (Table 1). That the LT-expressing U2OS/T cells supported levels of mdVp1 formation relative to the total amount of Vp1 to those of the dual LT/ST-expressing COS-7 cells suggests that the presence of LT, not ST, is responsible for promoting the formation of the mdVp1s, as well as stimulating the overall production of Vp1.

We next tested the effects on mdVp1 formation of two Vp1 cysteine pair mutations, mt1 (C49A-C87A) and mt2 (C49A-C254A), which cause defective Vp1 folding in the cytoplasm (10). mt1 and mt2 pCI-Vp1ΔC DNAs were transfected into CV-1, COS-7, U2OS, and U2OS/T cells, and their whole-cell lysates were then analyzed by Vp1-WB. Because the steady-state levels of the two mutant Vp1ΔCs were consistently lower than that of the wild-type Vp1ΔC for each cell type, analysis was performed using lysate aliquots with identical amounts of total Vp1, which was measured by the same dot blot procedure as before. The results show differing patterns of mdVp1 accumulation for the two mutants (Fig. 4). mt1 followed a pattern of Vp1 modification similar to that of wild-type Vp1ΔC, accumulating md2 through md4 in LT-expressing COS-7 and U2OS/T cells (Fig, 4, mt1, lanes 6 and 12) but not in LT-lacking CV-1 and U2OS cells (lanes 3 and 9). The fast-migrating 35-kDa Vp1 species (Fig. 4, lanes 3 and 9) is presumably a truncated form of mt1 Vp1ΔC.

mt2, however, displayed a very different pattern from those of the wild type and mt1. The presence of all modified forms was drastically reduced in both LT-lacking (Fig. 4, mt2, lanes 4 and 10) and LT-expressing (lanes 7 and 13) cells, while the prominence of the 35-kDa truncated species was greatly increased, suggesting a greater tendency of the mt2 Vp1ΔC to undergo degradation. Analysis performed with 293H, 293T, and 293TT cells confirmed the differing patterns of LT correlation with Vp1 modification for wild-type, mt1, and mt2 Vp1ΔCs (data not shown). These results imply that different steps in normal Vp1 folding are perturbed for the two mutants, resulting in mt2, but not mt1, either losing the ability to produce md2, md3, and md4 or making highly unstable forms of those species.

Covalent modification occurs within the core of Vp1, amino acids 21 to 261, in the cytoplasm.

To determine the region(s) of Vp1 in which covalent modifications occur, we expressed in COS-7 cells a series of N-terminally FLAG-tagged Vp1s with different deletions and examined their abilities to form mdVp1 species. Whole-cell lysates were prepared from the expression plasmid-transfected cells and analyzed by anti-FLAG Western blotting (FLAG-WB), which is expected to detect all tagged species equally. For the control Vp1ΔC lysate (ΔC58), the modified forms md2 and md4 were detected, along with the major unmodified md1 (Fig. 5A, lane 1). The removal of 20 residues from the N terminus (ΔNΔC58) or of an additional 42 residues from the C terminus (ΔC100) did not affect the formation of the modified species, and the shifts in banding pattern compared with the Vp1ΔC profile were consistent with the sizes of the respective deletions (Fig. 5A, lanes 2 and 3). The continued presence of the modified forms indicates that the modifications occur within the core region of Vp1 (aa 21 to 261). We note that the Vp1ΔC100 sample included a fragment much smaller than the unmodified Vp1ΔC100 monomer (Fig. 5A, lane 3). This fragment is presumed to be a proteolytic product and may have occurred because of the altered conformation of the truncated Vp1.

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500005.jpg

Existence of covalently modified species of polyomavirus Vp1s and HPV L1s. (A) Covalent modifications are located in the core of SV40 Vp1. Whole-cell lysates, each equivalent to 6 × 10 cells, were prepared from COS-7 cells 48 h after transfection with pCI-FLAG-Vp1ΔC and its derivative plasmids encoding FLAG-Vp1ΔC (ΔC58) (lane 1), FLAG-Vp1ΔNΔC (ΔNΔC58, further missing the N-terminal 20 residues) (lane 2), and FLAG-Vp1ΔC100 (ΔC100) (lane 3). The lysates were subjected to FLAG-WB for the detection of covalently modified Vp1s. (B) Subcellular localization of FLAG-tagged SV40 Vp1s. TC-7 cells were transfected with pCI-FLAG-Vp1ΔC plasmids encoding FLAG-Vp1ΔC (i and ii) and FLAG-Vp1ΔNΔC (iii and iv), fixed, and subjected to immunofluorescence microscopy for FLAG (i and iii) and DAPI staining (ii and iv). (C and D) Covalent modifications in polyomavirus and HPV major capsid proteins are detected in COS-7 cells, but not in CV-1 cells. Whole-cell lysates, equivalent to 6 × 10 cells each, were prepared from COS-7 (C) or CV-1 (D) cells transfected with the pCI plasmids expressing the FLAG-tagged Vp1ΔCs of SV40 (lanes 1), JCV (lanes 2), BKV (lanes 3), and MCV (lanes 4) or the FLAG-tagged L1ΔCs of HPV16 (lanes 5) and HPV18 (lanes 6). The lysates were examined for the presence of FLAG-tagged, covalently modified species as in panel A.

The N-terminal 20 residues of Vp1 constitute a bipartite NLS (3). Consistent with this, FLAG-tagged Vp1ΔC primarily localized to the nucleus (Fig. 5B, ,ii and ii), while FLAG-tagged Vp1ΔNΔC, lacking the NLS, was mainly cytoplasmic (iii and iv). The fact that the cytoplasmically localized Vp1ΔNΔC underwent nonreducible covalent modifications indicates that the modifications occur in the cytoplasm, consistent with the detection of the modified species in cytoplasmic subfractions (Fig. 1, ,2,2, and and33).

Polyomavirus Vp1s and HPV L1s form nonreducible modified species in LT-expressing cells.

We next asked if the Vp1s of other members of the polyomavirus family or the L1s of the HPV family can make covalently modified species when expressed in COS-7 cells. The various Vp1s and L1s, with their C termini removed to limit their assembly beyond pentamers, were expressed as FLAG-tagged proteins from plasmids that were converted through coding fragment exchanges from the plasmid that expresses SV40 Vp1ΔC, pCI-FLAG-Vp1ΔC. The various plasmids were transfected into COS-7 or CV-1 cells, and RIPA lysates were prepared and examined for the pattern of Vp1 or L1 species by FLAG-WB.

Besides the main Vp1 or L1 monomers, more slowly migrating FLAG-tagged species were generally detected only in COS-7 cells (Fig. 5C). In CV-1 cells (Fig. 5D), the only modified species detected was the 50-kDa derivative of JCV Vp1ΔC (lane 2). In COS-7 cells, various modified species were found for each virus: for JCV Vp1ΔC (aa 1 to 293), 50-, 57- (minor), and 90-kDa species were seen, in addition to the 40-kDa monomer (Fig. 5C, lane 2); for BKV Vp1ΔC (aa 1 to 301), 55-kDa and 88-kDa species, in addition to the 44-kDa monomer (lane 3); and for MCV Vp1ΔC (aa 1 to 325), 47- and 80-kDa species, in addition to the 42-kDa monomer (lane 4). The mobility shifts from the monomer to the modified species in the JCV and BKV profiles were very similar to those found in the SV40 profile. The modified species in the MCV profile resemble md2 and md4 of SV40 Vp1. For HPV16 and HPV18 L1ΔCs (aa 1 to 404), we found a shifted 50-kDa species and a minor 100-kDa species besides the 48-kDa monomer (lanes 5 and 6). Thus, our observations correlate the presence of SV40 LT with the presence of covalently modified Vp1 or L1 species among the human polyomaviruses and papillomaviruses, suggesting the intriguing possibility that all these covalent modifications arise from common mechanisms involving LT and some shared molecular features of these different major capsid proteins.

mdVp1s form multiprotein complexes with LT, Hsp70, and Hsp40.

In SV40 infection, Vp1 interacts with LT, Hsp70, and Hsp40 in the cytoplasm (12). We tested whether these proteins associate as multiprotein complexes in U2OS/T cells transfected with pCI-Vp1ΔC. The whole-cell lysate was fractionated by sedimentation through neutral sucrose gradients and examined for the presence and types of mdVp1s by Vp1-WB and for the presence of LT by anti-LT Western blotting (LT-WB). We found all types of mdVp1s throughout the gradient, fractions 4 through 8 of which contained protein complexes larger than 1,000 kDa (Fig. 6A, wt). Thus, a substantial portion of all mdVp1s were in association either among themselves, with many cellular proteins, or with LT. LT was present throughout the gradient whether Vp1ΔC was coexpressed (Fig. 6B, wt) or not (Fig. 6B, Cont; control U2OS/T). We estimated through anti-Vp1 immunoprecipitation analysis that about 1 to 2% of the total LT in the unfractionated whole-cell lysate of pCI-Vp1ΔC-transfected COS-7 cells was in complex with Vp1ΔC (data not shown).

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500006.jpg

MdVp1s form noncovalent multiprotein complexes that include LT and cellular chaperones. Whole-cell lysate was prepared from 3 × 10 U2OS/T cells 48 h after transfection with pCI-Vp1ΔC (wt) or with the mutant counterpart C49A-C87A (mt1) or C87A-C254A (mt2) plasmid. Control lysate was prepared from the same number of untransfected U2OS/T cells (Cont). Each lysate was fractionated by sedimentation through a neutral 5 to 32% sucrose gradient and collected in 15 fractions. (A) Reduced noncovalent complex formation by mutant Vp1ΔCs. For wt, mt1, and mt2 gradients, 1/30 of each even fraction was analyzed by Vp1-WB. The positions of md1 through md4 are marked on the right of each panel. (B) LT cofractionation with Vp1ΔC. For control (Cont) and wt gradients, 1/6 of each even fraction was analyzed by LT-WB using PAb419. (C) Reduced mt2 Vp1ΔC interaction with LT. For wt or mutant (mt1 and mt2) gradients, 1/12 or 1/8, respectively, of select even fractions was mixed and subjected to LT-IP using Ab-2, followed by LT-WB (LT) or Vp1-WB (Vp1). The analyzed samples included the fraction 4-fraction 6 mixtures (lanes 2, 5, and 8), the fraction 8-fraction 10 mixtures (lanes 3, 6, and 9), the fraction 12-fraction 14 mixtures (lanes 4, 7, and 10), and the control whole-cell lysate (lane 1). (D) Reduced association of mutant Vp1ΔCs with cellular chaperones. For wt, mt1, and mt2 gradients, 1/3 of fractions 3, 5, and 7 were pooled and subjected to Vp1-IP (Vp1) (lane 1), Hsp70-IP (Hsp70) (lane 2), Hsp40-IP (Hsp40) (lane 3), or β-Gal–IP (β-gal) (lane 4), followed by Vp1-WB. The Vp1 present in 1/10 of each input lysate is shown in lane 5. (E) Association of Vp1ΔC with cellular chaperones. For the wt gradient, the same pooled fraction was analyzed by the same IP types as in panel D, followed by Hsp40-WB (Hsp40) or Hsp70-WB (Hsp70). The Hsp40 or Hsp70 present in 1/2 of the input lysate is shown in lane 5.

We next probed for the presence of Hsp70, Hsp40, and LT in the multiprotein complexes by analyzing different portions of the sucrose gradient by a series of IPs. Wild-type Vp1ΔC coprecipitated with LT from pooled fractions 4 and 6, 8 and 10, and 12 and 14 (Fig. 6C, wt Vp1, lanes 2 to 4) and also with Hsp40 and Hsp70 from pooled fractions 3, 5, and 7 (Fig. 6D, wt, lanes 2 and 3), but not with anti-β-Gal antibody (lane 4). The specificity of the coprecipitations can be seen from the absence of Hsp40 in the Hsp70-IP complex (Fig. 6E, lane 2) and vice versa (lane 3) and from the absence of both Hsp70 and Hsp40 in the β-Gal–IP complex (lane 4). Wild-type Vp1ΔC formed noncovalent multiprotein complexes that included LT, Hsp70, and Hsp40.

The mt1 and mt2 mutants of Vp1ΔC exhibited different profiles of association with various proteins. mt1 Vp1ΔC, sedimenting mainly in fractions 8 through 14 (Fig. 6A, mt1), evidently formed some 1,000-kDa-plus protein complexes, though to a lesser extent than wild-type Vp1ΔC (Fig. 6A, wt). By using more of the mt1 pooled fractions than the wild-type ones for the LT-IP analysis, we observed coprecipitation of mt1 Vp1ΔC with LT in all three fraction pools (Fig. 6C, Vp1, lanes 5 to 7) and with Hsp70 and Hsp40 in the combined fractions 3, 5, and 7 (Fig. 6D, mt1, lanes 2 and 3).

mt2 Vp1, on the other hand, sedimented only in fractions 10 through 14, suggesting a severely reduced extent of protein association and an inability to form large multiprotein complexes (Fig. 6A, mt2). Little mt2 Vp1 coprecipitated with LT (Fig. 6C, Vp1, lanes 8 and 9) other than from fractions 12 and 14 (lane 10), which are expected to contain proteins of less than 150 kDa. Mt2 Vp1ΔC was not detected in the Hsp70-IP or Hsp40-IP complexes (Fig. 6D, mt2, lanes 2 and 3) derived from pooled lanes 3, 5, and 7, as can be expected from the lack of mt2 Vp1 in these fractions (Fig. 6A, mt2). None of the Vp1ΔC proteins were found in the β-Gal–IP complex (Fig. 6D, lane 4). The fact that mt1, but not mt2, Vp1ΔC was in large multiprotein complexes with LT, Hsp70, and Hsp40 suggests a role for LT-Vp1 interaction in the formation of multiprotein complexes that include cellular chaperones. LT may well be able to associate with mt2 Vp1 to some extent, but the mt2-LT complex apparently cannot further associate with cellular proteins that normally assist in the folding or oligomerization of Vp1.

A fraction of md4 is covalently linked to md1.

Given that the mdVp1s form multiprotein complexes either among themselves or with other proteins in the cell (Fig. 6), we explored the possibility that some of mdVp1's protein associations involve disulfide linkages. First, we examined the whole-cell lysate of pCI-Vp1ΔC transfected U2OS/T by denaturing through sucrose gradient sedimentation, followed by Vp1-WB under reducing conditions (Fig. 7A). In contrast to the broad mdVp1 distribution seen in a nondenaturing gradient, all mdVp1s were found in fractions 14 through 16 after sedimentation under denaturing conditions (compare Fig. 7A with Fig. 6A, wt). Hence, the multiprotein complexes found in fractions 4 through 8 of the nondenaturing gradient (Fig. 6A) must have involved largely noncovalent associations.

An external file that holds a picture, illustration, etc.
Object name is zjv9990975500007.jpg

Formation of covalent complexes by mdVp1s. (A) Sedimentation profile of covalently linked Vp1 complexes. Whole-cell lysate was prepared from 3 × 10 U2OS/T cells transfected for 48 h with pCI-Vp1ΔC. The lysate was incubated with NEM and SDS, fractionated through a denaturing 5 to 20% sucrose gradient, and collected in 16 fractions. One-thirtieth of the even fractions was subjected to Vp1-WB. (B) A fraction of md4 is covalently linked to md1. One-sixth of fractions 9 and 11 from the denaturing gradient in panel A was combined, precipitated with TCA, and resuspended in 20 μl of 10 mM Tris-Cl, pH 8.0. Then, diagonal SDS-PAGE was performed under nonreducing conditions in the first dimension (1D-NR) (upper gel strip) and under reducing conditions in the second dimension (2D) (bottom). Identical gel strips from the 1D-NR stage and after 2D processing were subjected to Vp1-WB for the detection of disulfide-linked and fully reduced mdVp1 species, respectively. a, b, and c represent md1, while d and e represent md4. The positions of molecular mass standards are indicated for the 1D-NR strip.

Next, we looked for the presence of disulfide-linked mdVp1 complexes in the Vp1ΔC-expressing U2OS/T lysate. We have previously detected disulfide-linked Vp1 oligomers in fractions 8 through 12 of a denaturing, but nonreducing, 5 to 20% sucrose gradient (8). Under the same conditions, few mdVp1s were detected in the corresponding fractions (Fig. 7A, lanes 4 to 6). We concentrated the combined fractions 9 and 11, in which we barely detected Vp1, and analyzed the sample by a modified diagonal SDS-PAGE, followed by Vp1-WB detection for the mdVp1s. The proteins here were first separated by nonreducing SDS-PAGE (1D-NR) to dissociate noncovalently associated proteins while preserving disulfide-bonded complexes. The way 1D-NR is run is different from 1D-R of the diagonal SDS-PAGE analysis described in Fig. 1. The disulfide-linked protein complexes were then resolved by reducing SDS-PAGE (2D) into component proteins. Proteins resolved from disulfide-linked complexes would be recognized as off-diagonal spots on the 2D gel. While cellular proteins linked to mdVp1s would escape detection at this stage, Vp1-WB would reveal mdVp1s linked to themselves. As seen positioned on the diagonal (Fig. 7B), the most prominent noncomplexed mdVp1s were the 40-kDa md1 monomer (species a) and the 80-kDa md4 (species d). Md1 was also found off the diagonal as a component of two disulfide-linked complexes with mobilities of about 100 kDa (Fig. 7B, species b) and 220 kDa (species c). The other component of the 100-kDa complex appears to be the off-diagonal md4 (species e). The 220-kDa complex either could be a multimer of md1 or could consist of one or more other components not detected by Vp1-WB. We did not observe the disulfide association of either md2 or md3 with other mdVp1s. Thus, our observations indicate that a fraction of md4 is disulfide linked to md1.

DISCUSSION

One of the fundamental processes that enable the formation of infectious SV40 virions is the proper folding of the major capsid protein Vp1 in the cytoplasm. To date, our investigations have uncovered two distinct, naturally occurring sets of Vp1 folding intermediates in host cells. The first set comprises transitory disulfide-bonded intermediates, including an intramolecularly disulfide-bonded Vp1 monomer that gives rise to intermolecularly disulfide-bonded Vp1 oligomers ranging from dimers to pentamers (8). The other set, the mdVp1s, comprises three nonreducible, covalently modified species, md2, md3, and md4. The unidentified covalent modifications are associated with the mdVp1s' reduced SDS-PAGE mobilities relative to that of the unmodified monomer md1 (12). How the disulfide intermediates and the mdVp1s relate to each other in the Vp1 folding pathway is not known. In this study, we have unambiguously shown that the mdVp1s are genuine resident products in cells and have linked them to the normal Vp1 folding process. In addition, we have implicated the SV40 LT in the formation of the mdVp1s.

In order to rule out the mdVp1s' being noncovalent or disulfide-bonded Vp1 complexes, we developed a system of new procedures and used them to demonstrate that these Vp1 species satisfy many experimental criteria for nonreducible, covalently modified proteins. First, we treated cells and cell extracts with the sulfhydryl alkylating agent NEM, which blocks the formation of artifactual disulfide bonds during and after cell lysis, respectively (Fig. 1C and and2).2). Second, we performed diagonal SDS-PAGE to confirm the profile of fully denatured and reduced proteins in a second-dimensional gel analysis (Fig. 1 and and2).2). We also checked the profile obtained under Laemmli SDS-PAGE conditions against that obtained by using the NuPAGE system, which has additional safeguards against incomplete reduction and artifactual disulfide formation during sample preparation and gel electrophoresis (Fig. 1C and and2).2). Third, we isolated the Vp1 species from cell extracts by His tag affinity purification under denaturing conditions (6 M urea), which more stringently excludes noncovalent complexes than IP, before SDS-PAGE and fluorography analysis (Fig. 3). The fact that the mdVp1s persisted under all the above-described experimental conditions indicates that they are indeed covalently modified Vp1s. Cell lysis and fractionation procedures (e.g., RIPA lysis), coupled with immunoprecipitation, are widely used to identify cellular and viral proteins that interact. In their basic forms, these procedures permit the detection of not only covalent protein forms, but also noncovalently associated, even artifactually disulfide cross-linked, protein complexes. As summarized above, the methodology we developed for the analysis of mdVp1s in this study offers an important, useful approach to the identification of nonreducibly covalently modified proteins in general.

The results of the present study link the mdVp1s to the normal Vp1 folding process in terms of the kinetics and subcellular location of their production and the Vp1 determinants involved in their covalent modification. The mdVp1s—md2, md3, and md4—emerged within minutes from the newly synthesized Vp1ΔC monomer md1 (Fig. 3), consistent with their being folding intermediates. The alternative possibility that some of mdVp1s are dead-end products has not been ruled out. A fraction of md4 was covalently linked to md1 (Fig. 7), implicating md4 in mediating the formation of disulfide-linked Vp1 oligomers. The covalent modifications occurred in the cytoplasm and involved the Vp1 core (aa 21 to 261) (Fig. 5), requiring in particular the Vp1 cysteine pair Cys87-Cys254. Although two cysteine pair Vp1 mutants, C49A-C87A (mt1) and C87A-C254A (mt2), are both defective in folding in the cytoplasm, only mt2 accumulated dramatically less md2, md3, and md4 than total Vp1 (Fig. 4). It remains to be learned how mt1 and mt2 differ in the specific aspects of the Vp1 folding process that are perturbed.

An intriguing observation from our study is that the presence of LT increased the total level of Vp1 and especially the proportion of mdVp1s in expression hosts. The levels of Vp1 detected correlated with the concentrations of LT in cell lines (Table 1). LT-expressing cells, COS-7 and U2OS/T, accumulated elevated levels of mdVp1s, while non-LT-expressing cells, CV-1 and U2OS, did not, even after the levels were normalized by that of total Vp1 (Fig. 4). We do not know whether an enhanced mdVp1 level and an enhanced total Vp1 level are two independent effects of LT or are related by cause and effect. Furthermore, the presence of SV40 LT stimulated analogous nonreducible covalent modifications in the major capsid proteins of other polyomaviruses and of distantly related human papillomaviruses. When expressed in COS-7 cells, a small fraction of BKV, JCV, and MCV Vp1s, as well as of HPV16 and HPV18 L1s, was present as distinct species that migrated more slowly than the unmodified monomer, though the number of modified species or the magnitudes of protein mass shifts were not identical to those of SV40 mdVp1s (Fig. 5C). Since the conserved residues of polyomavirus Vp1s fall within SV40 Vp1 aa 21 to 262, these residues may harbor sites for covalent modifications of the respective Vp1s. The mechanism by which LT induces these covalent modifications remains to be clarified.

Our observations so far suggest LT may promote the formation of the mdVp1s or the folding of Vp1 at least in part by promoting the association of certain cellular proteins with Vp1. We have found that LT was a component of noncovalent multiprotein complexes that included mdVp1s and the cellular chaperone proteins Hsp70 and Hsp40 (Fig. 6). The Vp1ΔC mutant mt1, which made some mdVp1s, appeared to be able to form noncovalent association with LT to an extent that enabled the formation of multiprotein complexes that included Hsp70 and Hsp40, even though the complexes were smaller on average than those of wild-type Vp1ΔC (Fig. 6C and andD).D). In contrast, mt2, which made very little mdVp1, was simultaneously deficient in noncovalent interaction with LT and multiprotein complex assembly involving Hsp70 and Hsp40 (Fig. 6C and andD),D), consistent with the interpretation that an inability to interact with LT leads to a failure to initiate the cascade of events in the Vp1 folding process. We speculate that the noncovalent association between LT and the newly synthesized Vp1 may be a key event that promotes the assembly of the cellular folding machinery to facilitate the folding of Vp1. The HSP70 chaperones and the Hsp40 cochaperone would be among the recruited cellular proteins and would provide further protein-protein interactions and catalysis. During this process, specific posttranslational protein modifications could occur, producing the Vp1 intermediates md2 through md4, which may associate with additional cellular or viral proteins. The differing characteristics of the two folding-defective mutant Vp1s support the idea that the Vp1 folding process consists of discrete steps and a large, heterogeneous multiprotein network. An elevated level of mdVp1s in LT-expressing cells could also suggest that LT induces a unique set of gene products that promote the production of infectious SV40 particles, including those products that generate the Vp1 modifications. The observation that the appearance of mdVp1s was blocked by the Vp1-folding mutation of mt2 supports the idea that mdVp1s are functional intermediates for proper Vp1 folding and oligomerization. How LT promotes the formation of mdVp1s during Vp1 folding remains to be elucidated.

In order for the precise functional role of the transient Vp1 covalent modifications to be elucidated, their precise chemical nature must be identified. In this regard, many possibilities exist, though none have been verified at this point. The first methionine of Vp1 is cleaved and the penultimate alanine is acetylated (47). Methylation of arginine and lysine residues has been shown to occur in murine polyomavirus Vp1, a homolog of SV40 Vp1 (48). Phosphorylations of Vp1 have been known for some time (49); however, we did not see a change in mdVp1 patterns after treating the lysates with potato acid phosphatase or with λ phosphatase (preliminary observation). SUMOylation and ubiquitination of Vp1 are also possible, as their potential target sites are present within Vp1 aa 21 to 262. Our preliminary result from the coexpression analysis of Vp1 and hemagglutinin (HA)-tagged ubiquitin or SUMO-1 has not shown the presence of ubiquitin or SUMO-1 moieties in any of the mdVp1s; we have, however, detected the presence of very small quantities of distinct SUMOylated Vp1 species. ISGylation, another candidate for the posttranslational modification of Vp1, is enhanced in the presence of SV40 LT (50). Our preliminary analysis has indicated the presence of an ISG15 moiety in md4: mdVp1s purified through His tag affinity are detected by ISG15 Western blotting, and the anti-ISG15 immunoprecipitate contains mdVp1s. Further spectrometry analysis is needed to detail the nature of the Vp1 modifications that generate md2, md3, and md4.

In summary, we have shown that the nonreducible, covalently modified Vp1s are genuine resident Vp1 species in SV40 LT-expressing cells. These species represent folding intermediates of Vp1, and their abundance is greatly increased in the presence of LT. Vp1 forms multiprotein complexes with LT and with cellular chaperone proteins in the cytoplasm, but a folding-defective mutant Vp1, mt2, cannot do so and cannot produce the folding intermediates. These findings have potentially uncovered the importance of certain chemical modifications of the SV40 major capsid protein in guiding its chaperone-assisted proper folding in the cytoplasm. The presence of nonreducible, covalently modified species of human polyomavirus Vp1s and human papillomavirus L1s hints at a universal role of nonreducible covalent modifications in the folding of the respective major capsid proteins in their host cells. Further investigation is needed to clarify the mechanism by which individual cellular and viral proteins participate in the folding of the major capsid proteins of small DNA tumor viruses.

Department of Molecular, Cell and Developmental Biology and Molecular Biology Institute, University of California at Los Angeles, Los Angeles, California, USAa
Section of Gene Therapy, Department of Aging Intervention, National Center for Geriatrics and Gerontology, Obu, Aichi, Japanb
Department of Microbiology-Immunology and Robert H. Lurie Comprehensive Cancer Center, Northwestern University, Chicago, Ilinois, USAc
Molecular Oncology Research Institute, Tufts Medical Center, Boston, Massachusetts, USAd
Corresponding author.
Address correspondence to Harumi Kasamatsu, ude.alcu.ibm@k_imurah.
Address correspondence to Harumi Kasamatsu, ude.alcu.ibm@k_imurah.
Received 2012 Apr 17; Accepted 2012 Feb 12.

Abstract

The folding and pentamer assembly of the simian virus 40 (SV40) major capsid protein Vp1, which take place in the infected cytoplasm, have been shown to progress through disulfide-bonded Vp1 folding intermediates. In this report, we further demonstrate the existence of another category of Vp1 folding or assembly intermediates: the nonreducible, covalently modified mdVp1s. These species were present in COS-7 cells that expressed a recombinant SV40 Vp1, Vp1ΔC, through plasmid transfection. The mdVp1s persisted under cell and lysate treatment and SDS-PAGE conditions that are expected to have suppressed the formation of artifactual disulfide cross-links. As shown through a pulse-chase analysis, the mdVp1s were derived from the newly synthesized Vp1ΔC in the same time frame as Vp1's folding and oligomerization. The apparent covalent modifications occurred in the cytoplasm within the core region of Vp1 and depended on the coexpression of the SV40 large T antigen (LT) in the cells. Analogous covalently modified species were found with the expression of recombinant polyomavirus Vp1s and human papillomavirus L1s in COS-7 cells. Furthermore, the mdVp1s formed multiprotein complexes with LT, Hsp70, and Hsp40, and a fraction of the largest mdVp1, md4, was disulfide linked to the unmodified Vp1ΔC. Both mdVp1 formation and most of the multiprotein complex formation were blocked by a Vp1 folding mutation, C87A-C254A. Our observations are consistent with a role for LT in facilitating the folding process of SV40 Vp1 by stimulating certain covalent modifications of Vp1 or by recruiting certain cellular proteins.

Abstract

ACKNOWLEDGMENTS

We are grateful to those who generously provided us with plasmid constructs: Ellen Fanning for pGEX-LT83-708, Michael Imperiale for pGEM-BKV-Vp1, Hirofumi Sawa for pCXSN-JCV-Vp1, and Chris Buck for pwM, p16L1-GFP, and p18L1-GFP. We thank Patricia Johnson and Maria Delgadillo-Correa for providing Autofluor. We thank Peggy Li for in-depth editing and helpful comments. We also thank Russell Doolittle and Jun-Ichi Tomizawa for encouragement and Hirofumi Sawa and Takuji Kasamatsu for discussions and suggestions.

This work was supported by gifts from the University of California to H.K. and in part by the National Institutes of Health (NIH) (CA50574); awards to A.N. from the Ministry of Education, Science, Sports and Culture of Japan, the Japan Foundation for Aging and Health, and the Ministry of Health, Labor and Welfare of Japan; and awards from the NIH to K.R. (CA21327;) and to O.G. (AI078926).

ACKNOWLEDGMENTS

Footnotes

Published ahead of print 20 February 2013

Footnotes

REFERENCES

REFERENCES

References

  • 1. Liddington RC, Yan Y, Moulai J, Sahli R, Benjamin TL, Harrison SC. 1991 Structure of simian virus 40 at 3.8-A resolution. Nature354:278–284 [[PubMed][Google Scholar]
  • 2. Stehle T, Gamblin SJ, Yan Y, Harrison SC. 1996 The structure of simian virus 40 refined at 3.1 A resolution. Structure4:165–182 [[PubMed][Google Scholar]
  • 3. Ishii N, Minami N, Chen EY, Medina AL, Chico MM, Kasamatsu H. 1996 Analysis of a nuclear localization signal of simian virus 40 major capsid protein Vp1. J. Virol.70:1317–1322 [Google Scholar]
  • 4. Li PP, Nakanishi A, Shum D, Sun PC, Salazar AM, Fernandez CF, Chan SW, Kasamatsu H. 2001 Simian virus 40 Vp1 DNA-binding domain is functionally separable from the overlapping nuclear localization signal and is required for effective virion formation and full viability. J. Virol.75:7321–7329 [Google Scholar]
  • 5. Nakanishi A, Itoh N, Li PP, Handa H, Liddington RC, Kasamatsu H. 2007 Minor capsid proteins of simian virus 40 are dispensable for nucleocapsid assembly and cell entry but are required for nuclear entry of the viral genome. J. Virol.81:3778–3785 [Google Scholar]
  • 6. Nakanishi A, Nakamura A, Liddington R, Kasamatsu H. 2006 Identification of amino acid residues within simian virus 40 capsid proteins Vp1, Vp2, and Vp3 that are required for their interaction and for viral infection. J. Virol.80:8891–8898 [Google Scholar]
  • 7. Nakanishi A, Shum D, Morioka H, Otsuka E, Kasamatsu H. 2002 Interaction of the Vp3 nuclear localization signal with the importin alpha 2/beta heterodimer directs nuclear entry of infecting simian virus 40. J. Virol.76:9368–9377 [Google Scholar]
  • 8. Li PP, Nakanishi A, Clark SW, Kasamatsu H. 2002 Formation of transitory intrachain and interchain disulfide bonds accompanies the folding and oligomerization of simian virus 40 Vp1 in the cytoplasm. Proc. Natl. Acad. Sci. U. S. A.99:1353–1358 [Google Scholar]
  • 9. Lin W, Shurgot JL, Kasamatsu H. 1986 The synthesis and transport of SV40 structural proteins. Virology154:108–120 [[PubMed][Google Scholar]
  • 10. Li PP, Nakanishi A, Fontanes V, Kasamatsu H. 2005 Pairs of Vp1 cysteine residues essential for simian virus 40 infection. J. Virol.79:3859–3864 [Google Scholar]
  • 11. Li PP, Nakanishi A, Tran MA, Salazar AM, Liddington RC, Kasamatsu H. 2000 Role of simian virus 40 Vp1 cysteines in virion infectivity. J. Virol.74:11388–11393 [Google Scholar]
  • 12. Li PP, Itoh N, Watanabe M, Shi Y, Liu P, Yang HJ, Kasamatsu H. 2009 Association of simian virus 40 vp1 with 70-kilodalton heat shock proteins and viral tumor antigens. J. Virol.83:37–46 [Google Scholar]
  • 13. Deuerling E, Bukau B. 2004 Chaperone-assisted folding of newly synthesized proteins in the cytosol. Crit. Rev. Biochem. Mol. Biol.39:261–277 [[PubMed][Google Scholar]
  • 14. Gething MJ, Sambrook J. 1992 Protein folding in the cell. Nature355:33–45 [[PubMed][Google Scholar]
  • 15. Hartl FU. 1996 Molecular chaperones in cellular protein folding. Nature381:571–579 [[PubMed][Google Scholar]
  • 16. Hartl FU, Hayer-Hartl M. 2002 Molecular chaperones in the cytosol: from nascent chain to folded protein. Science295:1852–1858 [[PubMed][Google Scholar]
  • 17. Mayer MP. 2005 Recruitment of Hsp70 chaperones: a crucial part of viral survival strategies. Rev. Physiol. Biochem. Pharmacol.153:1–46 [[PubMed][Google Scholar]
  • 18. Cripe TP, Delos SE, Estes PA, Garcea RL. 1995 In vivo and in vitro association of hsc70 with polyomavirus capsid proteins. J. Virol.69:7807–7813 [Google Scholar]
  • 19. Khandjian EW, Turler H. 1983 Simian virus 40 and polyoma virus induce synthesis of heat shock proteins in permissive cells. Mol. Cell. Biol.3:1–8 [Google Scholar]
  • 20. Phillips B, Abravaya K, Morimoto RI. 1991 Analysis of the specificity and mechanism of transcriptional activation of the human hsp70 gene during infection by DNA viruses. J. Virol.65:5680–5692 [Google Scholar]
  • 21. Sainis I, Angelidis C, Pagoulatos G, Lazaridis I. 1994 The hsc70 gene which is slightly induced by heat is the main virus inducible member of the hsp70 gene family. FEBS Lett.355:282–286 [[PubMed][Google Scholar]
  • 22. Ohtsuka K, Hata M. 2000 Mammalian HSP40/DNAJ homologs: cloning of novel cDNAs and a proposal for their classification and nomenclature. Cell Stress Chaperones5:98–112 [Google Scholar]
  • 23. Young JC, Agashe VR, Siegers K, Hartl FU. 2004 Pathways of chaperone-mediated protein folding in the cytosol. Nat. Rev. Mol. Cell Biol.5:781–791 [[PubMed][Google Scholar]
  • 24. Campbell KS, Mullane KP, Aksoy IA, Stubdal H, Zalvide J, Pipas JM, Silver PA, Roberts TM, Schaffhausen BS, DeCaprio JA. 1997 DnaJ/hsp40 chaperone domain of SV40 large T antigen promotes efficient viral DNA replication. Genes Dev.11:1098–1110 [[PubMed][Google Scholar]
  • 25. Fewell SW, Pipas JM, Brodsky JL. 2002 Mutagenesis of a functional chimeric gene in yeast identifies mutations in the simian virus 40 large T antigen J domain. Proc. Natl. Acad. Sci. U. S. A.99:2002–2007 [Google Scholar]
  • 26. Genevaux P, Lang F, Schwager F, Vartikar JV, Rundell K, Pipas JM, Georgopoulos C, Kelley WL. 2003 Simian virus 40 T antigens and J domains: analysis of Hsp40 cochaperone functions in Escherichia coli. J. Virol.77:10706–10713 [Google Scholar]
  • 27. Kelley WL, Georgopoulos C. 1997 The T/t common exon of simian virus 40, JC, and BK polyomavirus T antigens can functionally replace the J-domain of the Escherichia coli DnaJ molecular chaperone. Proc. Natl. Acad. Sci. U. S. A.94:3679–3684 [Google Scholar]
  • 28. Sock E, Enderich J, Wegner M. 1999 The J domain of papovaviral large tumor antigen is required for synergistic interaction with the POU-domain protein Tst-1/Oct6/SCIP. Mol. Cell. Biol.19:2455–2464 [Google Scholar]
  • 29. Sawai ET, Butel JS. 1989 Association of a cellular heat shock protein with simian virus 40 large T antigen in transformed cells. J. Virol.63:3961–3973 [Google Scholar]
  • 30. Sawai ET, Rasmussen G, Butel JS. 1994 Construction of SV40 deletion mutants and delimitation of the binding domain for heat shock protein to the amino terminus of large T-antigen. Virus Res.31:367–378 [[PubMed][Google Scholar]
  • 31. Srinivasan A, McClellan AJ, Vartikar J, Marks I, Cantalupo P, Li Y, Whyte P, Rundell K, Brodsky JL, Pipas JM. 1997 The amino-terminal transforming region of simian virus 40 large T and small t antigens functions as a J domain. Mol. Cell. Biol.17:4761–4773 [Google Scholar]
  • 32. Sullivan CS, Gilbert SP, Pipas JM. 2001 ATP-dependent simian virus 40 T-antigen-Hsc70 complex formation. J. Virol.75:1601–1610 [Google Scholar]
  • 33. Whalen KA, de Jesus R, Kean JA, Schaffhausen BS. 2005 Genetic analysis of the polyomavirus DnaJ domain. J. Virol.79:9982–9990 [Google Scholar]
  • 34. Fanning E, Zhao X, Jiang X. 2009 Polyomavirus life cycle, p 1–24 In Pipas JM, Damania B. (ed), DNA tumor viruses. Springer Science and Business Media, New York, NY [PubMed][Google Scholar]
  • 35. Sullivan CS, Pipas JM. 2002 T antigens of simian virus 40: molecular chaperones for viral replication and tumorigenesis. Microbiol. Mol. Biol. Rev.66:179–202 [Google Scholar]
  • 36. Jiang M, Abend JR, Tsai B, Imperiale MJ. 2009 Early events during BK virus entry and disassembly. J. Virol.83:1350–1358 [Google Scholar]
  • 37. Qu Q, Sawa H, Suzuki T, Semba S, Henmi C, Okada Y, Tsuda M, Tanaka S, Atwood WJ, Nagashima K. 2004 Nuclear entry mechanism of the human polyomavirus JC virus-like particle: role of importins and the nuclear pore complex. J. Biol. Chem.279:27735–27742 [[PubMed][Google Scholar]
  • 38. Buck CB, Cheng N, Thompson CD, Lowy DR, Steven AC, Schiller JT, Trus BL. 2008 Arrangement of L2 within the papillomavirus capsid. J. Virol.82:5190–5197 [Google Scholar]
  • 39. Buck CB, Pastrana DV, Lowy DR, Schiller JT. 2005 Generation of HPV pseudovirions using transfection and their use in neutralization assays. Methods Mol. Med.119:445–462 [[PubMed][Google Scholar]
  • 40. Tolstov YL, Pastrana DV, Feng H, Becker JC, Jenkins FJ, Moschos S, Chang Y, Buck CB, Moore PS. 2009 Human Merkel cell polyomavirus infection. II. MCV is a common human infection that can be detected by conformational capsid epitope immunoassays. Int. J. Cancer125:1250–1256 [Google Scholar]
  • 41. Li D, Zhao R, Lilyestrom W, Gai D, Zhang R, DeCaprio JA, Fanning E, Jochimiak A, Szakonyi G, Chen XS. 2003 Structure of the replicative helicase of the oncoprotein SV40 large tumour antigen. Nature423:512–518 [[PubMed][Google Scholar]
  • 42. Buck CB, Pastrana DV, Lowy DR, Schiller JT. 2004 Efficient intracellular assembly of papillomaviral vectors. J. Virol.78:751–757 [Google Scholar]
  • 43. Gluzman Y. 1981 SV40-transformed simian cells support the replication of early SV40 mutants. Cell23:175–182 [[PubMed][Google Scholar]
  • 44. Lin W, Hata T, Kasamatsu H. 1984 Subcellular distribution of viral structural proteins during simian virus 40 infection. J. Virol.50:363–371 [Google Scholar]
  • 45. Nakanishi A, Clever J, Yamada M, Li PP, Kasamatsu H. 1996 Association with capsid proteins promotes nuclear targeting of simian virus 40 DNA. Proc. Natl. Acad. Sci. U. S. A.93:96–100 [Google Scholar]
  • 46. Harlow E, Crawford LV, Pim DC, Williamson NM. 1981 Monoclonal antibodies specific for simian virus 40 tumor antigens. J. Virol.39:861–869 [Google Scholar]
  • 47. Lazarides E, Files JG, Weber K. 1974 Simian virus 40 structural proteins: amino-terminal sequence of the major capsid protein. Virology60:584–587 [[PubMed][Google Scholar]
  • 48. Burton KS, Consigli RA. 1996 Methylation of the polyomavirus major capsid protein VP1. Virus Res.40:141–147 [[PubMed][Google Scholar]
  • 49. O'Farrell PZ, Goodman HM. 1976 Resolution of simian virus 40 proteins in whole cell extracts by two-dimensional electrophoresis: heterogeneity of the major capsid protein. Cell9:289–298 [[PubMed][Google Scholar]
  • 50. Rathi AV, Cantalupo PG, Sarkar SN, Pipas JM. 2010 Induction of interferon-stimulated genes by simian virus 40 T antigens. Virology406:202–211 [Google Scholar]
Collaboration tool especially designed for Life Science professionals.Drag-and-drop any entity to your messages.