VP2 Cleavage and the Leucine Ring at the Base of t
http://www.100md.com
病菌学杂志 2006年第1期
Departments of Laboratory Medicine Genetics
Graduate Program in Microbiology, Yale University Medical School, 333 Cedar Street, New Haven, Connecticut 06510
ABSTRACT
Cylindrical projections surrounding the fivefold-symmetry axes in minute virus of mice (MVM) harbor central pores that penetrate through the virion shell. In newly released DNA-containing particles, these pores contain residues 28 to 38 belonging to a single copy of VP2, disposed so that its extreme N-terminal domain projects outside the particle. Virions are metastable, initially sequestering internally the N termini of all copies of the minor capsid protein, VP1, that is essential for entry. This VP1 domain can be externalized in vitro in response to limited heating, and we show here that the efficiency of this transition is greatly enhanced by proteolysis of VP2 N termini to yield VP3. This step also renders the VP1 rearrangement pH dependent, indicating that VP2 cleavage is a maturation step required to prime subsequent emergence of the VP1 "entry" domain. The tightest constriction within the cylinder is created by VP2 leucine 172, the five symmetry-related copies of which form a portal that resembles an iris diaphragm across the base of the pore. In MVMp, threonine substitution at this position, L172T, yields infectious particles following transfection at 37°C, but these can initiate infection only at 32°C, and this process can be blocked by exposing virions to a cellular factor(s) at 37°C during the first 8 h after entry. At 32°C, the mutant particle is highly infectious, and it remains stable prior to VP2 cleavage or following cleavage at pH 5.5 or below. However, upon exposure to neutral pH following VP2 cleavage, its VP1-specific sequences and genome are extruded even at room temperature, underscoring the significance of the VP2 cleavage step for MVM particle dynamics.
INTRODUCTION
Viral particles function as entry "nanomachines" that not only protect the genome from environmental assaults encountered during transmission, but also recognize and respond to a succession of specific cellular signals that allow them to navigate complex entry portals in their host cell and ultimately deliver their nucleic acid to the appropriate cellular replication compartment. How nonenveloped viruses achieve this odyssey is generally poorly understood, but in several virus families, a virion protein that is known to be essential for membrane penetration is subject to a specific proteolytic cleavage (6, 8). This allows the particle to exist in a "metastable" state, where the lowest energy form of the cleaved product is sequestered by the energy barrier between the two forms (19). During entry, these viruses encounter some form of catalyst, such as low pH or an interaction with a specific receptor, which releases the metastable configuration, allowing the de novo exposure of sequences required for membrane penetration. A variety of "penetration proteins" have been identified, including the N-myristoylated VP4 protein of poliovirus (3, 4, 17), the membrane lytic μ1 protein of reovirus (7), and protein VI of adenovirus (42). Here, we provide evidence for an activation cleavage step in the maturation of virions of the murine parvovirus minute virus of mice (MVM) that predisposes the particle for exposure of its specific "entry" peptide.
MVM packages a small, 5-kilobase genome of single-stranded linear DNA into a nonenveloped T = 1 icosahedral capsid. These have a maximum external diameter of just 280 (2, 23) and are constructed from 60 equivalent polypeptide subunits, of which VP2 is the major species, while VP1 is present at about 10 copies per particle (34). VP1 is an N-terminal extension of VP2, containing a sequence of 142 additional amino acids termed the VP1-specific region (VP1SR). In the crystal structure of the MVM virion, 547 amino acids from the C-terminal region common to all of the component VP polypeptides are ordered, leaving signal-rich N-terminal extensions of 38 residues for VP2 and 180 residues for VP1, which resist 60-fold averaging. These N-terminal peptides appear to serve as major mediators of cellular exit and reentry pathways, becoming sequentially externalized by the particle to mediate successive interactions with its host. Since parvoviral virions lack any accessory proteins, the capsid proteins are the sole mediators of these processes.
A cylindrical projection surrounds each of the 12 fivefold-symmetry axes in the MVM particle and is itself encircled by a 15--deep exterior depression, of unknown function, called the canyon. The cylinder, which is created by the juxtaposition of antiparallel -ribbons from each of the fivefold related capsid proteins, contains a central pore that penetrates through the virion shell to the particle interior (reviewed in references 1 and 9). Despite its apparently constrained diameter, this pore is believed to serve as an extrusion portal for the externalization of first VP2 and then VP1 amino-terminal peptides at appropriate times during virion trafficking. Further genetic evidence, including data presented in this report, suggests that one of these pores may also mediate the encapsidation of the viral genome late in infection and, subsequently, its controlled release in the next infectious cycle (5, 11, 16, 22, 24, 38, 39). In the crystal structure determined for MVM (2), ordered structure begins at VP2 residue 40, which lies on the inside of the shell at the base of the pore. In mature virions, but not in empty particles, the pore contains additional weak density, into which has been modeled a single copy of a conserved glycine-rich peptide spanning VP2 residues 28 to 38, as illustrated in Fig. 1A. Additional density, corresponding to residues 36 to 38 from the remaining capsid proteins, extends back into the particle interior (2). Since in the crystal structure, each pore can only accommodate one glycine-rich peptide, a maximum of one of the five locally available VP N termini can be externalized at once. However, almost all of these peptides can become surface exposed during entry, or during proteolytic digestion in vitro, suggesting that dynamic fluctuations in pore structure are possible.
Both the VP1 and VP2 N termini are completely sequestered within empty particles, but copies of the VP2 N termini first begin to emerge at the virion surface during the genome encapsidation process, revealing export signals that allow packaged virions to be trafficked out of the nucleus, in a cell-type-specific manner, prior to cell lysis (14, 25). DNA-containing particles are thus released with all their VP2 N termini intact, but a third structural protein, VP3, is subsequently generated from most VP2 molecules by proteolytic cleavage that removes 22 amino acids from its the N terminus. VP2-to-VP3 cleavage can occur in the extracellular environment following release, but if not, it invariably occurs during entry into a new host cell (10, 26, 29). This cleavage can be mimicked in vitro by incubating virions with a broad variety of proteases, but the cleavage site appears flexible and is highly exposed, so that it has been essentially impossible to totally ablate cleavage by mutagenesis or to stop it from occurring in vivo using combinations of protease inhibitors (10, 37; G. A. Farr, S. F. Cotmore, and P. Tattersall, unpublished results). In this study, we show that intact VP2 N termini help to protect virion integrity, since their cleavage predisposes the particle to a structural transition that exposes the VP1SR entry peptide.
Unlike VP2, VP1 is dispensable for both capsid assembly and DNA packaging but is absolutely required for the production of infectious virions (36). Although it contains the same proteolytic site as VP2, this is not accessible to trypsin digestion even in mature virions (35) and remains totally sequestered within the particle during the early stages of entry. However, the particle is metastable, allowing the VP1SR to be exposed experimentally at the surfaces of a variable proportion of particles following heating to temperatures in excess of 52°C (11). When antibodies directed against intact capsid particles or against the VP1SR were microinjected into the cytoplasm of cells infected with the closely related virus canine parvovirus (CPV), they blocked infection, indicating that intact particles traverse into the cytosol and that the VP1SR becomes exposed in vivo prior to nuclear entry (40, 41). Since the structural transition required for VP1SR exposure at the virion surface in vitro initiates at around 45°C, and total disruption of full particles requires temperatures in excess of 80°C (11), energy considerations suggest that virions likely remain intact throughout this cytoplasmic phase, although this remains to be demonstrated directly in vivo. However, in support of this interpretation, CPV virions microinjected into the cytoplasm were found to translocate into the nucleus intact, as demonstrated by their reactivity with structure-specific antibodies, although whether this was VP1SR driven remains uncertain (40). The unique VP1SR contains both nuclear localization signals (24, 41) and an active phospholipase A2 (PLA2) enzymatic core (15, 18, 43), which are both essential for particle infectivity. Mutation of PLA2 catalytic residues does not inhibit the accumulation of virions in perinuclear compartments, apparently late endosomes/lysosomes. We have recently shown that this is because PLA2 activity is required for escape through the endosomal membrane into the cytosol, since PLA2 knockout mutants can be complemented in trans by a variety of endosomolytic reagents (G. A. Farr, L.-G. Zhang, and P. Tattersall, submitted for publication).
The tightest constriction point in the fivefold pore, with a diameter of just 8 , is created near its base by a leucine residue at position 172 in the VP2 polypeptide, as shown in Fig. 1A and B. Previous analysis of MVM mutants carrying a full range of substitutions at this position revealed that some modifications render capsids temperature sensitive for assembly or influence DNA packaging. However, with the exception of two viable isoleucine and valine substitutions, all of the remaining mutants proved defective in establishing infection (16). In this study, we examine the entry defect in the threonine mutant L172T (modeled in Fig. 1B) in greater detail. Transferring this mutation into the cold-resistant genetic background of the prototype MVM strain (MVMp) has allowed analysis of the productive entry process at 32°C. Under these conditions, the mutant particle is highly infectious and remains stable prior to VP2 cleavage, or following cleavage under acidic conditions, but becomes unable to protect both its VP1SR and its genome following VP2 cleavage at neutral pH, thus underscoring the significance of the VP2-to-VP3 cleavage step for MVM particle dynamics during entry.
MATERIALS AND METHODS
Mutant infectious-clone construction. The pdBMVp-SacI clone harbors silent mutations between nucleotides 2782 and 2786 that introduce a unique SacI site into the MVMp genome in pdBMVp (21), just upstream of the VP2 start codon. Unique AflII and NsiI sites were engineered in pdBMVp-SacI at nucleotides 3248 and 3316, respectively, to flank VP2 residue L172, creating pdBMVp-pore. The threonine substitution, L172T, was constructed by cloning complementary mutant oligonucleotides between the AflII and NsiI sites of pdBMV-pore. All constructs were validated by DNA sequencing.
Cells and viral expansion assays. A9 cells are the A9 ouabr11 ouabain-resistant derivative of the HGPRT– cell line A9 and are permissive host cells for MVMp (33). Approximately 3,500 A9 cells were seeded on glass spots on Teflon-coated "spot slides" (Cell-Line Associates, Inc., Newfield, NJ) and incubated at 37°C for 18 to 20 h. They were transfected in duplicate with 200 ng/spot of infectious plasmid DNA using Superfect transfection reagent (QIAGEN, Inc. Valencia, CA). After 3 h, the cells were washed and incubation continued in the presence or absence of 1% neutralizing antiserum for a total of 50 to 96 h at the desired temperature. The cells were fixed with 2.5% paraformaldehyde and analyzed by indirect immunofluorescence for viral-gene products, as previously described (13). Total cell numbers per field were determined by counterstaining nuclei with DAPI (4',6'-diamidino-2-phenylindole). Color images were acquired using standardized exposures on a Nikon OptiPhot epifluorescence microscope fitted with a Kodak digital camera driven by MDS 290 software. Multiple images were quantified by single-blinded analysis in Adobe Photoshop 6.0.
Generation of purified virions for infection assays. Viral stocks were generated by transfecting plasmid DNA into A9 cells grown at 37°C, using Superfect transfection reagent (QIAGEN, Inc., Valencia, CA), and then incubating the cells at 32°C until cytopathic effects appeared. Virus was extracted by freezing and thawing the cells in 50 mM Tris-HCl, 0.5 mM EDTA, pH 8.7 (TE 8.7) and cleared by centrifugation at 15,000 x g for 30 min at 4°C. Cleared samples (5.8 ml) were floated on top of a 6-ml iodixanol (OptiPrep; Axis-Shield, Oslo, Norway) step gradient (1 ml 55% and 2 ml 45% in TE 8.7, followed by 2 ml 35% and 1 ml 15% in phosphate-buffered saline [PBS] plus 1 mM MgCl2 and 2.5 mM KCl). Samples were centrifuged at 35,000 rpm for 18 h at 18°C in a Beckman SW41 rotor; fractions were collected from the bottom of the gradient, particle concentrations were assessed by hemagglutination, and full virions were pooled for analysis. Encapsidated DNA was analyzed by denaturing agarose gel electrophoresis and Southern blotting following micrococcal-nuclease digestion (16), and genomes were quantitated with a Molecular Dynamics PhosphorImager SI.
Viral-infectivity assays. Semiconfluent monolayers of A9 cells growing on spot slides were infected at various genome equivalents per cell in Dulbecco's modified Eagle's medium (DMEM) containing 1% fetal bovine serum (FBS) and 25 mM HEPES, pH 7.0. Unless otherwise stated, infections were carried out for 3 h at either 32°C or 37°C. After infection, the cultures were washed with PBS to remove the remaining inoculum and incubated in DMEM containing 5% FBS and Clostridium perfringens neuraminidase (Sigma Chemical Co., St. Louis, MO) at 0.1 mg/ml. The cells were fixed 24 h postinfection and stained with a mouse monoclonal antibody directed against the viral NS1 protein.
Production and purification of 35S-labeled virions and VP1 transition assays. A9 fibroblasts were synchronized, infected with wild-type MVMp, and labeled with EXPRE35S protein labeling mix (NEN Life Sciences, Inc., Boston, MA), and virions containing only VP1 and VP2 were collected from the culture medium and purified, as previously described (11). Half of the purified virions were then digested to completion with trypsin, to yield VP1/VP3 virions, and repurified by centrifugation. Purified 35S-labeled virions were subjected to controlled heating in vitro in buffers containing 135 mM NaCl, 50 mM KCl, 0.5 mM Na2HPO4, 25 mM Tris-HCl and titrated to the specified pHs with HCl, and the pHs of all reaction mixtures were checked following preliminary reagent-mixing experiments that mimicked the experimental protocol. Shifts from pH 7.5 to 4.5 and back were accomplished by sequential addition of similar "adjustment" buffers, whose volumes and pHs had been predetermined to achieve the required transition. Finally, the pHs, ionic strengths, and volumes of all samples were normalized prior to immunoprecipitation. After cooling on ice, samples were immunoprecipitated using rabbit antibodies directed against intact MVM particles or against a recombinant peptide expressing the amino-terminal 141 amino acids of MVM VP1 and then analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and fluorography using 1 M salicylic acid, as previously described (11, 12, 16).
Trypsin digestion, Western blot analysis, virion repurification, and packaging assays. For trypsin digestion, virions were diluted into 100 mM NH4Cl (pH 5.5 to 7.5), 100 mM NaCl, 2.5 μM CaCl. TPCK-Trypsin (Cooper Biomedical, Malvern, PA) was added to a 2-ng/μl final concentration and incubated at the specified temperature for 1 h. The trypsin was inactivated by incubation in 4.5 mM phenylmethylsulfonyl fluoride and 20 mM EDTA for 10 min at room temperature prior to analysis by SDS-PAGE. The separated proteins were electroblotted onto polyvinylidene difluoride membranes using a Mini trans-blot transfer cell (Bio-Rad, Hercules, CA). The membranes were blocked and developed using the ECL system, according to the manufacturer's instructions (Amersham, Uppsala, Sweden).
For further analysis of trypsinized virions, samples were digested as described above, and the trypsin was inactivated with 15 mM EDTA and addition of soybean trypsin inhibitor to 10 times the predicted effective dose (Sigma Chemical Co., St. Louis, MO). Samples were immediately loaded onto a 4-ml iodixanol (OptiPrep; Axis-Shield, Oslo, Norway) step gradient (55% [500 μl], 45% [1 ml], 35% [1 ml], 25% [1 ml], and 15% [500 μl]). Gradient solutions were prepared either in PBS (pH 7.5) plus 1 mM MgCl2 and 2.5 mM KCl or in 50 mM MES (morpholineethanesulfonic acid) (pH 5.5) containing 120 mM NaCl, 1 mM MgCl2, and 2.5 mM KCl. Samples were centrifuged at 40,000 rpm for 20 h at 18°C in a Beckman SW50.1 rotor, and the fractions were collected and analyzed as described above. The micrococcal-nuclease sensitivities of particle-associated genomes were determined for each gradient fraction using the previously described packaging assays (16), following adjustment of the sample to pH 8.0.
RESULTS
The L172T mutant is temperature sensitive in the MVMp background. We have previously shown that substituting threonine for leucine 172 in the VP2 polypeptide of the immunosuppressive strain of MVM, MVMi, created a mutant virus that could assemble capsids and package progeny genomes effectively but that was defective for cell entry at 37°C (16). However, attempts to explore its entry phenotype further, at reduced temperatures, were hindered by the observation that when wild-type MVMi was assayed for viability at 32°C by expansion in iD5 cells, the virus strain proved to be cold sensitive (data not shown). However, when wild-type virus of the prototype strain of MVM, MVMp, was assayed in the same manner, its ability to initiate infection, as assessed by the appearance of the viral NS1 protein, was enhanced approximately 25-fold during the course of 3 days, indicating that MVMp could replicate productively at 32°C (Fig. 1C and D). Accordingly, we transferred the threonine mutation into the MVMp background, but to facilitate this switch, we first engineered a new wild-type infectious plasmid clone of MVMp called pdBMVp-pore, which contains silent mutations creating unique AflII and NsiI sites flanking L172. As illustrated in Fig. 1D, virions derived from pdBMVp and pdBMVp-pore behaved identically in viral expansion assays, indicating that the silent mutations had no appreciable effect on virus propagation. Throughout this report, therefore, the threonine mutation cloned into this MVMp background will be referred to simply as L172T, and pdBMVp-pore will be referred to as pdBMVp. Like its MVMi counterpart, when assayed for viral expansion at 37°C, L172T showed an 50-fold reduction in growth relative to wild-type virus (Fig. 1D). However, when cultured at 32°C for various periods posttransfection, viruses carrying this mutation expanded through the culture to levels that were within twofold of those seen for wild-type virus (Fig. 1D).
L172T is specifically temperature sensitive for the establishment of infection. To assess whether the growth defect exhibited by L172T at 37°C was due to the production of defective virions or to its inability to infect cells at this temperature, we purified virus that had been produced, following transfection, at either 32°C or 37°C and used it to infect A9 cells growing at 37°C. As seen in Fig. 2, wild-type virions produced at either temperature infected A9 cells efficiently, while even those L172T virions produced at 32°C failed to infect cells at 37°C, indicating that the defect operated during the establishment of infection. L172T inoculated at 48,000 genomes per cell gave the same number of infected cells as seen with wild-type virus at only 750 genomes per cell, suggesting that, at 37°C, L172T virions were approximately 64-fold less efficient at initiating infection than pdBMVp.
When cells were infected at 32°C, however, a strikingly different pattern was observed, with L172T virions establishing infection as efficiently as pdBMVp, regardless of the temperature at which they were produced (Fig. 2). At this lower temperature, infections with pdBMVp were somewhat less efficient than at 37°C, giving initiation efficiencies that were approximately eightfold lower throughout the lower range of input multiplicities (187.5 to 3,000 genomes per cell), while at higher multiplicities, the difference in infectivity at 32°C and 37°C appeared negligible. These experiments therefore indicate that L172T virions are intrinsically stable once assembled and packaged but are temperature sensitive for the establishment of a subsequent cycle of infection.
L172T virions are rapidly inactivated following infection at 37°C and must traffic to a late compartment before they are protected. To investigate whether L172T virions became inactivated at 37°C in a specific compartment in the entry pathway, we performed the temperature shift experiment documented in Fig. 3A. A9 cells were infected for 2 h at 32°C with 12,000 genome-containing virions per cell, washed, and transferred to medium containing neuraminidase, which blocks further infection. Immediately thereafter, and at subsequent 2-hour intervals, samples were shifted to 37°C. Remarkably, when compared to samples incubated at 37°C for the whole of the experiment, incubation at 32°C for periods of less than 8 h had little protective effect, and it was not until the shift occurred after 8 or 10 h at 32°C that infection rates began to resemble those seen with cultures maintained constantly at 32°C. This suggests either that inactivation occurs at a late point during entry or that it can occur at any time during that process.
Figure 3B documents the inverse experiment, where cells were initially infected at 37°C and then shifted down to 32°C at different times postinfection. If L172T virions were specifically inactivated in a late entry compartment, short incubations at 37°C during early time points should have had no effect on infection. However, L172T was rapidly inactivated during entry at 37°C. Compared to control samples maintained at 32°C throughout, infectivity levels dropped 2-, 10-, and 20-fold following incubation at 37°C for 2, 4, or 8 h, respectively. Thus, incubating cells at 37°C for even a few hours during the first 8 to 10 h of an L172 infection markedly impaired infection rates.
Since L172T virions produced at 37°C are infectious at 32°C, we did not expect that simply incubating virions at 37°C would destroy their infectivity. However, we verified this by incubating purified virions in vitro for 2 h at 37°C before using them to infect cells growing at 32°C. As shown in Fig. 3C, in vitro incubation at 37°C had no effect on subsequent infection rates. Similarly, L172T virions incubated in vitro at 37°C for 2 h with medium supplemented with 10% A9-conditioned medium prior to infection at 32°C remained fully infectious (Fig. 3D), indicating that the cells do not produce a soluble inactivating factor but that actual contact with the host cell at 37°C is required for inactivation. Infectious entry of wild-type MVMp at 37°C has been shown to require exposure to low endosomal pH (29). Similarly, exposing cultures to the vacuolar proton pump inhibitor bafilomycin A1 for 1 hour prior to and for 3 h during incubation with pdBMVp or L172T at 32°C reduced the infectivity of both viruses approximately sixfold, as did incubation of pdBMVp infections at 37°C, while incubation of L172T-infected cells with inhibitor at this temperature further impaired its residual infectivity (data not shown). Thus, entry of both pdBMVp and L172T MVMp appears to require passage through an acidified endosomal compartment at either 32°C or 37°C.
VP2-to-VP3 cleavage lowers the energy required to expose VP1 N termini at neutral pH but makes this transition highly dependent upon pH. Before pursuing the basis of the L172T defect further, we sought to clarify our understanding of the structural transition that MVM particles are thought to undergo during entry, which results in the exposure of VP1 N termini at the virion surface. These termini are sequestered within infectious virions, but wild-type MVMi and MVMp and MVMi L172T particles have been shown to be metastable in vitro, undergoing a structural rearrangement in response to limited heating in which the particles remain intact but at least some of the VP1 N termini become externalized (11, 16). Specifically, we wished to ask whether prior cleavage of VP2 to VP3 significantly influenced the energy required for this transition and whether the low-pH conditions encountered during entry would impinge on the transition process. To this end, we prepared [35S]methionine/[35S]cysteine-labeled wild-type MVMp virions in which the VP2 N termini were intact (hereafter called VP1/VP2 virions), digested half of them exhaustively with trypsin, and rebanded them, thus generating a population in which the great majority of VP2 N termini had been cleaved to VP3 (called VP1/VP3 virions). These two forms of wild-type MVMp were then incubated for 10 min at various temperatures and pHs, as indicated in Fig. 4, before the buffer concentrations and pHs were normalized, and their structures were assessed by immunoprecipitation. Following exposure to neutral pH at 4°C, very few VP1/VP2 virions had accessible VP1 N termini, but heating them to increasing temperatures at pH 7.5 induced progressive exposure of these peptides (Fig. 4A, lanes 2 to 5), although only a relatively small proportion of particles (<10%) underwent this transition even after being heated to 65°C in the experiment shown here.
In contrast, exposure to pH 4.5 caused VP1 N termini to become accessible in a minor, but detectable, proportion of VP1/VP2 particles, but subsequent heating to 45 or 55°C had no further effect on VP1 accessibility, although surface-exposed VP1 N termini were detected in <10% of virions after heating them to 65°C (Fig. 4A, lanes 6 to 9). Since MVMp has a particle-to-infectivity ratio in A9 cells of around 300:1 (32), it is formally possible that the small proportion of VP1/VP2 virions in which VP1SR becomes exposed at pH 4.5 constitute essential entry intermediates. However, this is unlikely because, as discussed below, trypsin digestion at low pH, which would preclude both the formation and integrity of such an intermediate, fails to inactivate wild-type MVMp.
N-terminal peptides were fully protected in VP1/VP3 virion populations at 4°C and neutral pH (Fig. 4B, lane 2), but exposing these particles to heat had a much more profound effect on particle structure, causing VP1 N termini to become surface accessible on most virions after being heated to 55°C and on all virions after being heated to 65°C (Fig. 4B, compare lanes 1 and 5). To confirm that particles remained intact throughout this procedure, we also precipitated samples that had been heated to 65°C at pH 4.5 and 7.5 with antibodies that detect only intact particles (lanes 10 in Fig. 4A and 4B, respectively). Thus, the energy required to induce particle transition is markedly reduced following VP2-to-VP3 cleavage, so that particles universally undergo this rearrangement at temperatures that are easily compatible with particle integrity. We suggest, therefore, that during cell entry, VP2 cleavage may be an important and enabling prelude to VP1 extrusion.
Shifting VP1/VP3 virions to pH 4.5, however, effectively inhibited all such rearrangements (Fig. 4B, lanes 6 to 9). Thus, following cleavage of the VP2 N termini, wild-type virions become completely refractory to the conformational shift required for VP1 exposure. However, this change is not permanent, since it can be reversed simply by shifting particles back from acidic to neutral pH, as shown in the experiment illustrated in Fig. 4C, where VP1SR exposure following heating to 55°C for 10 min was compared in samples that had been either maintained at pH 7.5 throughout the experiment (Fig. 4C, lane 3) or shifted from pH 7.5 to pH 4.5 for 20 min (lane 5) and then shifted back to neutral pH (lane 10). Thus, cleavage of VP2 to VP3 profoundly increases the susceptibility of MVMp virions to a critical structural rearrangement, but this shift becomes markedly pH dependent.
Trypsin digestion of L172T VP2 at pH 7.5 leads to exposure of VP1-specific sequences, even at room temperature, except at low pH. The immunoprecipitation approach shown in Fig. 4 to assess the effects of VP2 cleavage proved unsuitable for analysis of the L172T mutant because it was not possible to purify homogeneous L172T populations following tryptic digestion. Instead, we analyzed VP1 exposure in the mutant by its susceptibility to cleavage with trypsin. Using MVMi-based virions carrying this mutation, we had previously shown that, while VP1 was sequestered from antibodies in newly purified virions, it could be partially degraded when these were exposed to trypsin at 37°C, whereas the peptide remained protected in wild-type particles (16). To analyze whether this susceptibility persisted in the cold-insensitive MVMp genetic background, stocks of MVMp wild-type and L172T virions were harvested and purified by standard procedures, which typically gave a mixture of cleaved and intact forms of the VP2 N termini, and then were further digested under experimental conditions with trypsin, and the products were analyzed by SDS-PAGE. Digestion of wild-type virions at neutral pH and either 37°C or 30°C resulted in cleavage of most of the remaining VP2 polypeptides to VP3, but VP1 was not affected (Fig. 5A). In contrast, but consistent with the MVMi mutant, trypsin digestion of MVMp L172T at neutral pH and either 30°C or 37°C led to cleavage of both VP2 and a large proportion of the VP1 N termini, indicating that these peptides become exposed at the surfaces of the VP2-cleaved L172T particles, irrespective of virion tropism. Since the L172T mutant is temperature sensitive for entry, we had suspected that digestion at lower temperatures might prevent exposure of VP1 N termini, but we conclude from this experiment that at neutral pH, cleavage of VP2 N termini is linked to exposure of VP1 N termini, even at 30°C. However, because in vivo exposure at 32°C might not be adequately mimicked by cleavage at similar temperatures under these somewhat artificial in vitro conditions, we next digested virions at 23°C. As seen in Fig. 5B, cleavage at pH 7.5 again resulted in the loss of some L172T VP1 N termini. Thus, we conclude that even at room temperature VP2 cleavage allows L172T virions to transition to a VP1-exposed configuration.
In the same experiment (Fig. 5B), mutant L172T particles were also incubated with trypsin at pH 6.5. When digested at 23°C, some VP1 loss was still apparent at pH 6.5, but the phenotype was partially rescued, while at 30°C or 37°C (Fig. 5C), VP1 degradation was comparable to that seen at pH 7.5 in Fig. 5A. In contrast, when the pH of the reaction mixture was lowered further to 5.5, degradation of VP1 was completely abrogated at either 30°C or 37°C (Fig. 5D). Thus, as in the immunoprecipitation experiments with wild-type virions presented in Fig. 4, incubation at low pH following VP2 cleavage prevented L172T virions from transitioning into a VP1-exposed configuration. Since during infection, VP2-to-VP3 cleavage routinely occurs during passage of the virus through degradative compartments within the cytosol, this suggests that cleavage in compartments where the pH is 6.5 or above might lead to loss of VP1-specific functions, whereas VP2 cleavage at pHs of 5.5 or below, as might be encountered in late endosomes, would allow the particle to retain its VP1 N termini, and hence its endosomal escape function and nuclear trafficking signals.
Infectivity of L172T, but not of wild-type MVMp, is abrogated by prior cleavage of VP2 N termini. During entry, MVM would be expected to encounter neutral or somewhat acidic cellular compartments before it reached a protective low-pH environment. If so, cleaving the VP2 N termini of L172 virions in vitro at pH 5.5, which leaves their VP1 N termini intact, might nonetheless effectively inactivate the virus, since such virions would transition at an early stage of entry, exposing their previously sequestered sequences to potential degradation. To explore this possibility, wild-type and L172T virions were digested with trypsin at 30°C, at either pH 5.5 or 7.5. The trypsin was then inactivated by addition of 10% fetal bovine serum, and the resulting viruses were used to infect cells at 32°C. Although the addition of 10% serum inhibited pdBMVp infections slightly, trypsin digestion had little effect on the infectivity of wild-type virus relative to untreated controls, as seen in Fig. 6. However, prior digestion at either pH 5.5 or 7.5 totally abrogated the infectivity of L172T. Thus, provided the majority of their VP2 N termini are intact when they are first taken up by the cell, L172T virions are infectious, but prior cleavage of VP2 to VP3 is effectively lethal.
Loss of VP1 integrity leads to exposure of the viral genome. To explore the structural integrity of L172T virions after proteolytic cleavage, we again digested virions with trypsin at 30°C at either pH 5.5 or 7.5, inactivated the trypsin with EDTA and soy bean trypsin inhibitor, and sedimented the digested virions to equilibrium in iodixanol gradients buffered at the appropriate pH. Fractions from these gradients were analyzed by Western blotting, using antibodies directed against MVM capsid proteins. The gradient profile of wild-type virions after tryptic cleavage at either pH 5.5 or 7.5 was essentially unaltered (Fig. 7A), with virions reaching equilibrium in fractions that contained approximately 50% iodixanol. Similarly when L172T virions were banded at pH 5.5 following digestion at pH 5.5, they reached equilibrium at approximately this density (Fig. 7B). However, virions digested and centrifuged at pH 7.5 were dispersed more broadly through the gradient, reaching a peak in fraction 9, indicating that they had become substantially less dense (Fig. 7C).
Fractions from the pH 5.5 gradient of L172T virions were then analyzed by denaturing gel electrophoresis, before and after nuclease digestion, as shown in Fig. 7D, left. The leading edge of this peak was seen to contain predominantly full-length genomes that remained resistant to nuclease digestion, while trailing fractions tended to contain more subgenomic DNA, much of which nevertheless remained protected from nuclease and thus sequestered. In contrast, while full-length copies of the viral genome still sedimented with capsids after digestion at pH 7.5, they now peaked in fraction 9, and the genomes were not protected from nuclease digestion (Fig. 7D, right). Most of these genomes were extensively degraded and could not be detected with our probe, while some remained as a smear of subgenomic protected species. Thus, proteolysis of L172T at pH 5.5 causes cleavage of VP2-specific sequences, but the particle remains intact and is able to protect its genome. On the other hand, proteolysis of L172T at pH 7.5 causes loss of both VP2 and VP1 N termini and a substantial drop in particle density. Since this density shift reflects exposure of the viral DNA, which remains attached to, and partially protected by, the particle, this suggests that the fivefold pores of L172T are unable to contain their VP1 N termini or, if these are cleaved, prevent the subsequent release of the genome. In the experiments presented here, virions cleaved and centrifuged at pH 5.5 were digested with nuclease for 30 min at 37°C and neutral pH to ensure that the nuclease was fully functional. Thus, VP2-cleaved but VP1-intact L172T virions retain their integrity at neutral pH, at least for short periods of time.
DISCUSSION
VP2-to-VP3 cleavage is a priming event for VP1 extrusion. We show here that the hollow cylinder surrounding each of the 12 fivefold-symmetry axes in the MVM virion plays an essential role in controlling particle structure and integrity. Leucine residues from each of the fivefold related polypeptides create the narrowest (8-) constriction point at VP2 residue 172, situated near the base of these pores, as shown in Fig. 1A, but when the diameter of the channel is somewhat expanded at this position by the relatively conservative substitution of aliphatic threonine hydroxyl groups for the somewhat more bulky hydrophobic leucine side chains, the virion becomes conditionally stable, remaining intact and infectious provided its VP2 N termini are uncleaved. However, it rapidly unravels to expose its essential, but vulnerable, VP1 N termini when 22 amino acids from the N termini of VP2 are removed. These exposed VP2 N termini, which project to the particle exterior through the fivefold channels, thus have a major constraining effect on particle dynamics in the L172T mutant, but this also appears to be true in wild-type virus. MVM virions are known to be metastable, initially sequestering their VP1-specific "entry" peptide, which carries a PLA2 enzymatic core required for endosome escape, but extruding this domain to the outside surface of the otherwise-intact virion in response to externally applied energy. However, virions that have intact VP2 N termini undergo this transition somewhat inefficiently and following exposure to relatively high temperatures, possibly because VP1 N termini must compete with the VP2 N termini for their normal egress portal. In contrast, once these VP2 tethers have been cleaved, particles transition quantitatively to expose at least one VP1 N terminus per particle and at lower temperatures. This suggests that VP2-to-VP3 cleavage is a natural component of virion maturation, priming it for the subsequent emergence of VP1-specific sequences. Thus, generation of VP3 polypeptides corresponds to the preprotein cleavage step characterized in some other nonenveloped viruses (6, 8), priming the virion for an upcoming transition to a lower free-energy state in which the VP1 N-terminal entry peptide is exposed.
While the importance of the VP2 cleavage has long been suspected for viruses in the genus Parvovirus, it has proven remarkably hard to evaluate experimentally. In part, this is because cleavage appears to occur in a particular region of the peptide, where it emerges from the mouth of the fivefold cylinder, rather than at a specific residue, and the relevant proline-rich N-terminal peptide is highly exposed. Thus, it can be cleaved in vitro by a wide variety of both endopeptidases and exopeptidases, while the rest of the particle remains untouched, and we have been unable to prevent cleavage from occurring during infection in cell culture using a broad range of class-specific protease inhibitors, both alone and in various combinations (S. F. Cotmore and P. Tattersall, unpublished observations; 16a), suggesting that there are multiple proteases resident in the viral entry pathway that can effect this truncation. This sequence is still cleaved effectively in vivo, albeit rather less efficiently, even following ablation of the potential furin-like protease target motif, N-RVER-C, in this peptide, which is also the site of tryptic cleavage (37), or if this region is replaced with an unrelated 7-amino-acid cleavage site for a plant virus protease (Cotmore and Tattersall, unpublished). Finally, since even wild-type MVMp has a particle-to-infectivity ratio of around 300, structural analysis of virions retrieved from infected cells, for VP2-intact or VP2-cleaved forms, are hard to interpret, since it is difficult to ascertain whether these are part of the infectious process or the products of misrouting through nonproductive cellular compartments. Three observations relevant to this problem are documented in this study. First, VP2 cleavage predisposes the virion to VP1 extrusion; second, this cleavage makes particle dynamics highly responsive to changes in ambient pH; and third, L172 mutants, which have compromised fivefold cylinders, transition at very low temperatures but can use their intact VP2 N termini to maintain particle integrity. Together, these observations provide the first clear evidence of a central role for VP2 cleavage in the infectious process.
Exactly how the constraining effects of low pH on particle structure operate following this cleavage will be of great interest to explore structurally, but presumably cleavage allows the particle to adopt a lower free-energy state in which charged groups effectively control pore dynamics. The biological implications of these findings are significant, since they suggest that cleaved forms of the virion that are present in low-pH compartments within the cell would be unlikely to undergo the structural rearrangements that appear to be required for traversing the endosomal bilayer. In consequence, we suggest that such particles must either be trafficked further, to a suitable neutral compartment or microdomain, or be involved in some other structural interaction, perhaps with the viral receptor, that effects this rearrangement. In this respect, it is of particular interest that for the related virus CPV, in which entry is known to be mediated via interactions with the transferrin receptor (27), antibodies directed against the cytosolic domain of this receptor can still inhibit infection when injected into the cytoplasm 4 hours after viral uptake, suggesting that virions may remain in contact with their receptor for several hours (20). However, for CPV, treatment of particles in vitro at pH 4 to 6 has been reported to result in an increase in PLA2 activity, which may indicate that the CPV VP1 N terminus becomes exposed at low pH (31), perhaps indicating a striking difference between the chemistries of MVM and CPV particles, and perhaps also in their entry strategies. It should also be noted, however, that exposure of VP1-specific peptides in an endosomal compartment several hours after viral uptake has been demonstrated for both MVM and CPV by immunofluorescence microscopy (30, 31), although whether these particles are engaged in infectious entry or have transitioned prematurely and are being degraded remains equivocal.
What is the L172T defect and how is it rescued at low temperature Purified L172T virions are resistant to in vitro incubation at 37°C, and cells transfected at this temperature produce infectious virus, suggesting that a cellular factor to which virions are exposed during entry is responsible for their temperature-sensitive inactivation. However, evidence presented here for MVMp virions bearing the L172T substitution, as well as previous data for this mutant in the MVMi strain, suggests that these viruses are stable and appear structurally normal until their VP2 N termini are cleaved, at which point they are no longer able to protect their VP1-specific peptides, at least at neutral pH. Supporting this argument, cleavage of VP2 to VP3 in vitro prior to infection completely inactivated L172, even if this cleavage was carried out at pH 5.5 in order to ensure that its VP1 N termini were intact at the time it was added to the cells. Extrusion of the VP1 N termini was also tightly linked to exposure of the viral genome, although at present we do not understand the exact phasing of these two processes. In vitro, heat-induced transitions also lead to exposure of much of the DNA without VP1 cleavage, so that while in the experiments presented here such cleavage apparently preceded genome exposure, this is unlikely to be absolutely required. However, if either VP1 N termini or the genome were to become accessible during passage through early entry endosomes, it would likely be exposed to hydrolytic enzymes or the depurinating effects of low pH, so that one would expect the virus to be inactivated. Thus, we suggest that L172T is conditionally viable predominantly because lower temperatures substantially reduce the likelihood of the VP2-to-VP3 cleavage occurring in an early entry compartment, where the ambient pH is higher than 5.5. Our data also support the idea that parvoviral entry is a protracted procedure, requiring passage through acidified compartments, and that VP2 cleavage can be effected by a variety of cellular proteases in multiple compartments during this process. Since VP2 cleavage primes the virion for subsequent rearrangement, we further suggest that, for L172T, such cleavage must reliably occur in a low-pH environment, where the particle is stabilized, since proteolysis could then proceed without causing the particle to unravel. Thus, we suggest that L172T may be infectious at 32°C because cellular proteases are less active at this temperature, allowing the particle to access a low-pH compartment prior to VP2 cleavage.
While L172T virions can be inactivated within the first 2 hours of infection at 37°C, a shift up to 37°C even 8 h into infection at 32°C still influences the success of this process, suggesting that allowing particles to traffic deep into the endocytic pathway does not totally protect them. Indeed, by 10 h postinfection at 32°C, when L172T infection finally becomes completely resistant to shift-up, many of the infecting virions may well have entered the cytosol. This suggests that the L172T mutant could also have a second defect, encountered in a late entry compartment, which causes infection to proceed more reliably at lower temperature. While the nature of this defect remains obscure, we do know that in coinfection experiments at 37°C, the L172T mutant can complement an endosomal penetration defect seen in PLA2-null viruses as well as wild-type virus can (Farr et al., submitted), but without itself establishing infection in the same cells. In light of this experiment, we conclude that either proteolysed VP1 N termini can still support endosomolysis or, perhaps more probably, it is damage to the viral genome that most commonly inactivates L172T mutants at 37°C. Because this premature extrusion defect may mimic a normal event that for the wild type occurs later in infection, we are currently exploring the mechanism of L172T genome exposure in greater detail, as a paradigm for parvoviral uncoating.
Role of the fivefold cylinder. VP2 N termini are sequestered within empty MVM particles, but at least some of these termini are extruded to the particle exterior during genome encapsidation. A glycine-rich sequence connects these extruded N termini to the rest of the VP2 polypeptide. However, in the crystal structure, the fivefold pore is only wide enough to accommodate the connector sequences, so that it is not clear how the bulkier N-terminal peptide gets to the outside of the capsid shell. Moreover, at least three more VP2 N termini are initially positioned at the base of each cylinder, and in the mature virion, most of these can be cleaved by externally applied protease. Since the X-ray structures clearly indicate that these pores cannot accommodate more than one of these connector sequences at a time, this suggests a model in which, following removal of its external N-terminal domain, the resident connector somehow retracts, allowing another, intact VP2 N terminus to take its place. Since 22 N-terminal amino acids of this VP2 molecule must also be threaded through the pore, it seems probable that the cylinder opens or is otherwise somehow enlarged, perhaps transiently, but repeatedly, to permit the sequential egress of these domains. If so, this makes it easier to imagine how leucine 172, positioned at the base of the cylinder, might prove critical for particle integrity. Thus, if much of the cylinder wall can flex, the annulus formed by leucine at its base could serve as the ultimate retainer for structures that still need to be sequestered within the capsid. In support of such a model, we have previously observed that in MVMi-based L172 mutants, the VP1 N-terminal peptides become susceptible to trypsin degradation not only when this residue is replaced by amino acids with somewhat smaller side groups, such as threonine or glycine, but also when it is replaced by those with bulkier, aromatic side groups, such as phenylalanine (16), suggesting that leucine 172 is somehow uniquely suited to maintaining the integrity of this molecular "iris diaphragm."
It seems likely, moreover, that these cylinders not only support the egress of VP2 N-terminal peptides, but that one of them may also be involved in the presumably much more extensive rearrangements needed for it to serve as the portal for viral DNA. Thus, when leucine 172 in MVMi was replaced by tryptophan, empty capsids assembled effectively at 32°C but were unable to package DNA, while substitution of aspartic or glutamic acid at this position seriously impaired, but did not totally prevent, packaging. Somewhat similar observations were made for adeno-associated virus type 2 (AAV2) by Bleker and colleagues (5), who showed that packaging efficiency could be impeded by specific mutations in the funnel-shaped structure that forms the base of the AAV pore and by other mutations positioned in nearby regions of the cylinder wall. However, in AAV, mutation of "funnel" residues at the base of the pore did not lead to premature exposure of VP1 N termini when the particle was heated in vitro but rather to their aberrant retention. Remarkably, substitution of alanine for MVM residue F526, V40, N149, S43, or D263, which are positioned around the base of the fivefold pore, or for N170, positioned in the wall, had a similarly negative effect on pore flexibility, at least as assessed by tryptophan exposure in baculovirus-expressed virus-like empty particles assembled from VP2 molecules (28). In these studies conformational shifts were assessed by monitoring changes in fluorescence as a function of temperature, revealing an abrupt shift for wild-type sequences at around 46°C, which was shown to coincide with extrusion of some VP2 N termini, possibly through the fivefold pores. In the alanine mutants mentioned above, this shift was ablated, and in some cases extrusion of VP2 N termini was shown to be impaired, suggesting that, like the AAV substitutions, mutation of some of the flanking sequences in MVM may depress pore flexibility. Thus, it will be of great interest to see if these alanine substitutions exert a similar phenotype when expressed as part of an MVM virion. However, the structures of the MVM and AAV pores are substantially different, and AAV virions are constructed with only five copies each of their VP2 and VP1 N-terminal peptides, so that their particle dynamics may be markedly different from those of members of the genus Parvovirus. Nevertheless, in both cases, the fivefold pore appears to be involved in extrusion of the PLA2-bearing VP1 N termini, since AAV pore mutants are unable to effectively expose this essential VP1 domain, whereas the MVM VP2 L172 mutants are unable to effectively protect it from degradation at neutral pH.
ACKNOWLEDGMENTS
We thank colleagues at the HHMI Biopolymer Laboratory and W. M. Keck Foundation Biotechnology Resource Laboratory at Yale University for DNA sequencing and synthesis of oligonucleotides used in this project.
G.A.F. was supported, in part, by T32 AI07640. This work was supported by Public Health Service grant CA29303 from the National Cancer Institute.
REFERENCES
Agbandge-McKenna, M., and M. S. Chapman. 2005. Correlating structure with function in the viral capsid, p. 125-139. In J. R. Kerr, S. F. Cotmore, M. E. Bloom, R. M. Linden, and C. R. Parrish (ed.), Parvoviruses. Edward Arnold, London, United Kingdom.
Agbandje-McKenna, M., A. L. Llamas-Saiz, F. Wang, P. Tattersall, and M. G. Rossmann. 1998. Functional implications of the structure of the murine parvovirus, minute virus of mice. Structure 6:1369-1381.
Belnap, D. M., D. J. Filman, B. L. Trus, N. Cheng, F. P. Booy, J. F. Conway, S. Curry, C. N. Hiremath, S. K. Tsang, A. C. Steven, and J. M. Hogle. 2000. Molecular tectonic model of virus structural transitions: the putative cell entry states of poliovirus. J. Virol. 74:1342-1354.
Belnap, D. M., B. M. McDermott, Jr., D. J. Filman, N. Cheng, B. L. Trus, H. J. Zuccola, V. R. Racaniello, J. M. Hogle, and A. C. Steven. 2000. Three-dimensional structure of poliovirus receptor bound to poliovirus. Proc. Natl. Acad. Sci. USA 97:73-78.
Bleker, S., F. Sonntag, and J. A. Kleinschmidt. 2005. Mutational analysis of narrow pores at the fivefold symmetry axes of adeno-associated virus type 2 capsids reveals a dual role in genome packaging and activation of phospholipase A2 activity. J. Virol. 79:2528-2540.
Bubeck, D., D. J. Filman, N. Cheng, A. C. Steven, J. M. Hogle, and D. M. Belnap. 2005. The structure of the poliovirus 135S cell entry intermediate at 10-angstrom resolution reveals the location of an externalized polypeptide that binds to membranes. J. Virol. 79:7745-7755.
Chandran, K., D. L. Farsetta, and M. L. Nibert. 2002. Strategy for nonenveloped virus entry: a hydrophobic conformer of the reovirus membrane penetration protein micro 1 mediates membrane disruption. J. Virol. 76:9920-9933.
Chandran, K., J. S. Parker, M. Ehrlich, T. Kirchhausen, and M. L. Nibert. 2003. The delta region of outer-capsid protein micro 1 undergoes conformational change and release from reovirus particles during cell entry. J. Virol. 77:13361-13375.
Chapman, M. S., and M. Agbandje-McKenna. 2005. Atomic structure of viral particles, p. 107-123. In J. R. Kerr, S. F. Cotmore, M. E. Bloom, R. M. Linden, and C. R. Parrish (ed.), Parvoviruses. Edward Arnold, London, United Kingdom.
Clinton, G. M., and M. Hayashi. 1975. The parovivirus MVM: particles with altered structural proteins. Virology 66:261.
Cotmore, S. F., A. M. D'Abramo, Jr., C. M. Ticknor, and P. Tattersall. 1999. Controlled conformational transitions in the MVM virion expose the VP1 N-terminus and viral genome without particle disassembly. Virology 254:169-181.
Cotmore, S. F., A. M. D'Abramo, Jr., L. F. Carbonell, J. Bratton, and P. Tattersall. 1997. The NS2 polypeptide of parvovirus MVM is required for capsid assembly in murine cells. Virology 231:267-280.
Cotmore, S. F., and P. Tattersall. 1990. Alternate splicing in a parvoviral nonstructural gene links a common amino-terminal sequence to downstream domains which confer radically different localization and turnover characteristics. Virology 177:477-487.
Cotmore, S. F., and P. Tattersall. 2005. Encapsidation of minute virus of mice DNA: aspects of the translocation mechanism revealed by the structure of partially packaged genomes. Virology 336:100-112.
Dorsch, S., G. Liebisch, B. Kaufmann, P. von Landenberg, J. H. Hoffmann, W. Drobnik, and S. Modrow. 2002. The VP1 unique region of parvovirus B19 and its constituent phospholipase A2-like activity. J. Virol. 76:2014-2018.
Farr, G. A., and P. Tattersall. 2004. A conserved leucine that constricts the pore through the capsid fivefold cylinder plays a central role in parvoviral infection. Virology 323:243-256.
Farr, G. A. 2005. Ph.D. thesis. Yale University, New Haven, Conn.
Fricks, C. E., and J. M. Hogle. 1990. Cell-induced conformational change in poliovirus: externalization of the amino terminus of VP1 is responsible for liposome binding. J. Virol. 64:1934-1945.
Girod, A., C. E. Wobus, Z. Zadori, M. Ried, K. Leike, P. Tijssen, J. A. Kleinschmidt, and M. Hallek. 2002. The VP1 capsid protein of adeno-associated virus type 2 is carrying a phospholipase A2 domain required for virus infectivity. J. Gen. Virol. 83:973-978.
Hogle, J. M. 2002. Poliovirus cell entry: common structural themes in viral cell entry pathways. Annu. Rev. Microbiol. 56:677-702.
Hueffer, K., L. M. Palermo, and C. R. Parrish. 2004. Parvovirus infection of cells by using variants of the feline transferrin receptor altering clathrin-mediated endocytosis, membrane domain localization, and capsid-binding domains. J. Virol. 78:5601-5611.
Kestler, J., B. Neeb, S. Struyf, J. Van Damme, S. F. Cotmore, A. D'Abramo, P. Tattersall, J. Rommelaere, C. Dinsart, and J. J. Cornelis. 1999. cis requirements for the efficient production of recombinant DNA vectors based on autonomous parvoviruses. Hum. Gene Ther. 10:1619-1632.
Kronenberg, S., B. Bottcher, C. W. von der Lieth, S. Bleker, and J. A. Kleinschmidt. 2005. A conformational change in the adeno-associated virus type 2 capsid leads to the exposure of hidden VP1 N termini. J. Virol. 79:5296-5303.
Llamas-Saiz, A. L., M. Agbandje-McKenna, W. R. Wikoff, J. Bratton, P. Tattersall, and M. G. Rossmann. 1997. Structure determination of minute virus of mice. Acta Crystallogr. D 53:93-102.
Lombardo, E., J. C. Ramirez, J. Garcia, and J. M. Almendral. 2002. Complementary roles of multiple nuclear targeting signals in the capsid proteins of the parvovirus minute virus of mice during assembly and onset of infection. J. Virol. 76:7049-7059.
Maroto, B., N. Valle, R. Saffrich, and J. M. Almendral. 2004. Nuclear export of the nonenveloped parvovirus virion is directed by an unordered protein signal exposed on the capsid surface. J. Virol. 78:10685-10694.
Paradiso, P. R., K. R. Williams, and R. L. Costantino. 1984. Mapping of the amino terminus of the H-1 parvovirus major capsid protein. J. Virol. 52:77-81.
Parker, J. S., W. J. Murphy, D. Wang, S. J. O'Brien, and C. R. Parrish. 2001. Canine and feline parvoviruses can use human or feline transferrin receptors to bind, enter, and infect cells. J. Virol. 75:3896-3902.
Reguera, J., A. Carreira, L. Riolobos, J. M. Almendral, and M. G. Mateu. 2004. Role of interfacial amino acid residues in assembly, stability, and conformation of a spherical virus capsid. Proc. Natl. Acad. Sci. USA 101:2724-2729.
Ros, C., C. J. Burckhardt, and C. Kempf. 2002. Cytoplasmic trafficking of minute virus of mice: low-pH requirement, routing to late endosomes, and proteasome interaction. J. Virol. 76:12634-12645.
Ros, C., and C. Kempf. 2004. The ubiquitin-proteasome machinery is essential for nuclear translocation of incoming minute virus of mice. Virology 324:350-360.
Suikkanen, S., M. Antila, A. Jaatinen, M. Vihinen-Ranta, and M. Vuento. 2003. Release of canine parvovirus from endocytic vesicles. Virology 316:267-280.
Tattersall, P. 1972. Replication of the parvovirus MVM. I. Dependence of virus multiplication and plaque formation on cell growth. J. Virol. 10:586-590.
Tattersall, P., and J. Bratton. 1983. Reciprocal productive and restrictive virus-cell interactions of immunosuppressive and prototype strains of minute virus of mice. J. Virol. 46:944-955.
Tattersall, P., P. J. Cawte, A. J. Shatkin, and D. C. Ward. 1976. Three structural polypeptides coded for by minute virus of mice, a parvovirus. J. Virol. 20:273-289.
Tattersall, P., A. J. Shatkin, and D. C. Ward. 1977. Sequence homology between the structural polypeptides of minute virus of mice. J. Mol. Biol. 111:375-394.
Tullis, G. E., L. R. Burger, and D. J. Pintel. 1993. The minor capsid protein VP1 of the autonomous parvovirus minute virus of mice is dispensable for encapsidation of progeny single-stranded DNA but is required for infectivity. J. Virol. 67:131-141.
Tullis, G. E., L. R. Burger, and D. J. Pintel. 1992. The trypsin-sensitive RVER domain in the capsid proteins of minute virus of mice is required for efficient cell binding and viral infection but not for proteolytic processing in vivo. Virology 191:846-857.
Valle, N., L. Riolobos, and J. M. Almendral. 2005. Synthesis, post-translational modification and trafficking of the parvovirus structural polypeptides, p. 291-304. In J. R. Kerr, S. F. Cotmore, M. E. Bloom, R. M. Linden, and C. R. Parrish (ed.), Parvoviruses. Edward Arnold, London, United Kingdom.
Vihinen-Ranta, M., and C. R. Parrish. 2005. Cell infection processes of autonomous parvoviruses, p. 157-163. In J. R. Kerr, S. F. Cotmore, M. E. Bloom, R. M. Linden, and C. R. Parrish (ed.), Parvoviruses. Edward Arnold, London, United Kingdom.
Vihinen-Ranta, M., W. Yuan, and C. R. Parrish. 2000. Cytoplasmic trafficking of the canine parvovirus capsid and its role in infection and nuclear transport. J. Virol. 74:4853-4859.
Vihinen-Ranta, M., D. Wang, W. S. Weichert, and C. R. Parrish. 2002. The VP1 N-terminal sequence of canine parvovirus affects nuclear transport of capsids and efficient cell infection. J. Virol. 76:1884-1891.
Wiethoff, C. M., H. Wodrich, L. Gerace, and G. R. Nemerow. 2005. Adenovirus protein VI mediates membrane disruption following capsid disassembly. J. Virol. 79:1992-2000.
Zadori, Z., J. Szelei, M. C. Lacoste, Y. Li, S. Gariepy, P. Raymond, M. Allaire, I. R. Nabi, and P. Tijssen. 2001. A viral phospholipase A2 is required for parvovirus infectivity. Dev. Cell. 1:291-302.(Glen A. Farr, Susan F. Co)
Graduate Program in Microbiology, Yale University Medical School, 333 Cedar Street, New Haven, Connecticut 06510
ABSTRACT
Cylindrical projections surrounding the fivefold-symmetry axes in minute virus of mice (MVM) harbor central pores that penetrate through the virion shell. In newly released DNA-containing particles, these pores contain residues 28 to 38 belonging to a single copy of VP2, disposed so that its extreme N-terminal domain projects outside the particle. Virions are metastable, initially sequestering internally the N termini of all copies of the minor capsid protein, VP1, that is essential for entry. This VP1 domain can be externalized in vitro in response to limited heating, and we show here that the efficiency of this transition is greatly enhanced by proteolysis of VP2 N termini to yield VP3. This step also renders the VP1 rearrangement pH dependent, indicating that VP2 cleavage is a maturation step required to prime subsequent emergence of the VP1 "entry" domain. The tightest constriction within the cylinder is created by VP2 leucine 172, the five symmetry-related copies of which form a portal that resembles an iris diaphragm across the base of the pore. In MVMp, threonine substitution at this position, L172T, yields infectious particles following transfection at 37°C, but these can initiate infection only at 32°C, and this process can be blocked by exposing virions to a cellular factor(s) at 37°C during the first 8 h after entry. At 32°C, the mutant particle is highly infectious, and it remains stable prior to VP2 cleavage or following cleavage at pH 5.5 or below. However, upon exposure to neutral pH following VP2 cleavage, its VP1-specific sequences and genome are extruded even at room temperature, underscoring the significance of the VP2 cleavage step for MVM particle dynamics.
INTRODUCTION
Viral particles function as entry "nanomachines" that not only protect the genome from environmental assaults encountered during transmission, but also recognize and respond to a succession of specific cellular signals that allow them to navigate complex entry portals in their host cell and ultimately deliver their nucleic acid to the appropriate cellular replication compartment. How nonenveloped viruses achieve this odyssey is generally poorly understood, but in several virus families, a virion protein that is known to be essential for membrane penetration is subject to a specific proteolytic cleavage (6, 8). This allows the particle to exist in a "metastable" state, where the lowest energy form of the cleaved product is sequestered by the energy barrier between the two forms (19). During entry, these viruses encounter some form of catalyst, such as low pH or an interaction with a specific receptor, which releases the metastable configuration, allowing the de novo exposure of sequences required for membrane penetration. A variety of "penetration proteins" have been identified, including the N-myristoylated VP4 protein of poliovirus (3, 4, 17), the membrane lytic μ1 protein of reovirus (7), and protein VI of adenovirus (42). Here, we provide evidence for an activation cleavage step in the maturation of virions of the murine parvovirus minute virus of mice (MVM) that predisposes the particle for exposure of its specific "entry" peptide.
MVM packages a small, 5-kilobase genome of single-stranded linear DNA into a nonenveloped T = 1 icosahedral capsid. These have a maximum external diameter of just 280 (2, 23) and are constructed from 60 equivalent polypeptide subunits, of which VP2 is the major species, while VP1 is present at about 10 copies per particle (34). VP1 is an N-terminal extension of VP2, containing a sequence of 142 additional amino acids termed the VP1-specific region (VP1SR). In the crystal structure of the MVM virion, 547 amino acids from the C-terminal region common to all of the component VP polypeptides are ordered, leaving signal-rich N-terminal extensions of 38 residues for VP2 and 180 residues for VP1, which resist 60-fold averaging. These N-terminal peptides appear to serve as major mediators of cellular exit and reentry pathways, becoming sequentially externalized by the particle to mediate successive interactions with its host. Since parvoviral virions lack any accessory proteins, the capsid proteins are the sole mediators of these processes.
A cylindrical projection surrounds each of the 12 fivefold-symmetry axes in the MVM particle and is itself encircled by a 15--deep exterior depression, of unknown function, called the canyon. The cylinder, which is created by the juxtaposition of antiparallel -ribbons from each of the fivefold related capsid proteins, contains a central pore that penetrates through the virion shell to the particle interior (reviewed in references 1 and 9). Despite its apparently constrained diameter, this pore is believed to serve as an extrusion portal for the externalization of first VP2 and then VP1 amino-terminal peptides at appropriate times during virion trafficking. Further genetic evidence, including data presented in this report, suggests that one of these pores may also mediate the encapsidation of the viral genome late in infection and, subsequently, its controlled release in the next infectious cycle (5, 11, 16, 22, 24, 38, 39). In the crystal structure determined for MVM (2), ordered structure begins at VP2 residue 40, which lies on the inside of the shell at the base of the pore. In mature virions, but not in empty particles, the pore contains additional weak density, into which has been modeled a single copy of a conserved glycine-rich peptide spanning VP2 residues 28 to 38, as illustrated in Fig. 1A. Additional density, corresponding to residues 36 to 38 from the remaining capsid proteins, extends back into the particle interior (2). Since in the crystal structure, each pore can only accommodate one glycine-rich peptide, a maximum of one of the five locally available VP N termini can be externalized at once. However, almost all of these peptides can become surface exposed during entry, or during proteolytic digestion in vitro, suggesting that dynamic fluctuations in pore structure are possible.
Both the VP1 and VP2 N termini are completely sequestered within empty particles, but copies of the VP2 N termini first begin to emerge at the virion surface during the genome encapsidation process, revealing export signals that allow packaged virions to be trafficked out of the nucleus, in a cell-type-specific manner, prior to cell lysis (14, 25). DNA-containing particles are thus released with all their VP2 N termini intact, but a third structural protein, VP3, is subsequently generated from most VP2 molecules by proteolytic cleavage that removes 22 amino acids from its the N terminus. VP2-to-VP3 cleavage can occur in the extracellular environment following release, but if not, it invariably occurs during entry into a new host cell (10, 26, 29). This cleavage can be mimicked in vitro by incubating virions with a broad variety of proteases, but the cleavage site appears flexible and is highly exposed, so that it has been essentially impossible to totally ablate cleavage by mutagenesis or to stop it from occurring in vivo using combinations of protease inhibitors (10, 37; G. A. Farr, S. F. Cotmore, and P. Tattersall, unpublished results). In this study, we show that intact VP2 N termini help to protect virion integrity, since their cleavage predisposes the particle to a structural transition that exposes the VP1SR entry peptide.
Unlike VP2, VP1 is dispensable for both capsid assembly and DNA packaging but is absolutely required for the production of infectious virions (36). Although it contains the same proteolytic site as VP2, this is not accessible to trypsin digestion even in mature virions (35) and remains totally sequestered within the particle during the early stages of entry. However, the particle is metastable, allowing the VP1SR to be exposed experimentally at the surfaces of a variable proportion of particles following heating to temperatures in excess of 52°C (11). When antibodies directed against intact capsid particles or against the VP1SR were microinjected into the cytoplasm of cells infected with the closely related virus canine parvovirus (CPV), they blocked infection, indicating that intact particles traverse into the cytosol and that the VP1SR becomes exposed in vivo prior to nuclear entry (40, 41). Since the structural transition required for VP1SR exposure at the virion surface in vitro initiates at around 45°C, and total disruption of full particles requires temperatures in excess of 80°C (11), energy considerations suggest that virions likely remain intact throughout this cytoplasmic phase, although this remains to be demonstrated directly in vivo. However, in support of this interpretation, CPV virions microinjected into the cytoplasm were found to translocate into the nucleus intact, as demonstrated by their reactivity with structure-specific antibodies, although whether this was VP1SR driven remains uncertain (40). The unique VP1SR contains both nuclear localization signals (24, 41) and an active phospholipase A2 (PLA2) enzymatic core (15, 18, 43), which are both essential for particle infectivity. Mutation of PLA2 catalytic residues does not inhibit the accumulation of virions in perinuclear compartments, apparently late endosomes/lysosomes. We have recently shown that this is because PLA2 activity is required for escape through the endosomal membrane into the cytosol, since PLA2 knockout mutants can be complemented in trans by a variety of endosomolytic reagents (G. A. Farr, L.-G. Zhang, and P. Tattersall, submitted for publication).
The tightest constriction point in the fivefold pore, with a diameter of just 8 , is created near its base by a leucine residue at position 172 in the VP2 polypeptide, as shown in Fig. 1A and B. Previous analysis of MVM mutants carrying a full range of substitutions at this position revealed that some modifications render capsids temperature sensitive for assembly or influence DNA packaging. However, with the exception of two viable isoleucine and valine substitutions, all of the remaining mutants proved defective in establishing infection (16). In this study, we examine the entry defect in the threonine mutant L172T (modeled in Fig. 1B) in greater detail. Transferring this mutation into the cold-resistant genetic background of the prototype MVM strain (MVMp) has allowed analysis of the productive entry process at 32°C. Under these conditions, the mutant particle is highly infectious and remains stable prior to VP2 cleavage, or following cleavage under acidic conditions, but becomes unable to protect both its VP1SR and its genome following VP2 cleavage at neutral pH, thus underscoring the significance of the VP2-to-VP3 cleavage step for MVM particle dynamics during entry.
MATERIALS AND METHODS
Mutant infectious-clone construction. The pdBMVp-SacI clone harbors silent mutations between nucleotides 2782 and 2786 that introduce a unique SacI site into the MVMp genome in pdBMVp (21), just upstream of the VP2 start codon. Unique AflII and NsiI sites were engineered in pdBMVp-SacI at nucleotides 3248 and 3316, respectively, to flank VP2 residue L172, creating pdBMVp-pore. The threonine substitution, L172T, was constructed by cloning complementary mutant oligonucleotides between the AflII and NsiI sites of pdBMV-pore. All constructs were validated by DNA sequencing.
Cells and viral expansion assays. A9 cells are the A9 ouabr11 ouabain-resistant derivative of the HGPRT– cell line A9 and are permissive host cells for MVMp (33). Approximately 3,500 A9 cells were seeded on glass spots on Teflon-coated "spot slides" (Cell-Line Associates, Inc., Newfield, NJ) and incubated at 37°C for 18 to 20 h. They were transfected in duplicate with 200 ng/spot of infectious plasmid DNA using Superfect transfection reagent (QIAGEN, Inc. Valencia, CA). After 3 h, the cells were washed and incubation continued in the presence or absence of 1% neutralizing antiserum for a total of 50 to 96 h at the desired temperature. The cells were fixed with 2.5% paraformaldehyde and analyzed by indirect immunofluorescence for viral-gene products, as previously described (13). Total cell numbers per field were determined by counterstaining nuclei with DAPI (4',6'-diamidino-2-phenylindole). Color images were acquired using standardized exposures on a Nikon OptiPhot epifluorescence microscope fitted with a Kodak digital camera driven by MDS 290 software. Multiple images were quantified by single-blinded analysis in Adobe Photoshop 6.0.
Generation of purified virions for infection assays. Viral stocks were generated by transfecting plasmid DNA into A9 cells grown at 37°C, using Superfect transfection reagent (QIAGEN, Inc., Valencia, CA), and then incubating the cells at 32°C until cytopathic effects appeared. Virus was extracted by freezing and thawing the cells in 50 mM Tris-HCl, 0.5 mM EDTA, pH 8.7 (TE 8.7) and cleared by centrifugation at 15,000 x g for 30 min at 4°C. Cleared samples (5.8 ml) were floated on top of a 6-ml iodixanol (OptiPrep; Axis-Shield, Oslo, Norway) step gradient (1 ml 55% and 2 ml 45% in TE 8.7, followed by 2 ml 35% and 1 ml 15% in phosphate-buffered saline [PBS] plus 1 mM MgCl2 and 2.5 mM KCl). Samples were centrifuged at 35,000 rpm for 18 h at 18°C in a Beckman SW41 rotor; fractions were collected from the bottom of the gradient, particle concentrations were assessed by hemagglutination, and full virions were pooled for analysis. Encapsidated DNA was analyzed by denaturing agarose gel electrophoresis and Southern blotting following micrococcal-nuclease digestion (16), and genomes were quantitated with a Molecular Dynamics PhosphorImager SI.
Viral-infectivity assays. Semiconfluent monolayers of A9 cells growing on spot slides were infected at various genome equivalents per cell in Dulbecco's modified Eagle's medium (DMEM) containing 1% fetal bovine serum (FBS) and 25 mM HEPES, pH 7.0. Unless otherwise stated, infections were carried out for 3 h at either 32°C or 37°C. After infection, the cultures were washed with PBS to remove the remaining inoculum and incubated in DMEM containing 5% FBS and Clostridium perfringens neuraminidase (Sigma Chemical Co., St. Louis, MO) at 0.1 mg/ml. The cells were fixed 24 h postinfection and stained with a mouse monoclonal antibody directed against the viral NS1 protein.
Production and purification of 35S-labeled virions and VP1 transition assays. A9 fibroblasts were synchronized, infected with wild-type MVMp, and labeled with EXPRE35S protein labeling mix (NEN Life Sciences, Inc., Boston, MA), and virions containing only VP1 and VP2 were collected from the culture medium and purified, as previously described (11). Half of the purified virions were then digested to completion with trypsin, to yield VP1/VP3 virions, and repurified by centrifugation. Purified 35S-labeled virions were subjected to controlled heating in vitro in buffers containing 135 mM NaCl, 50 mM KCl, 0.5 mM Na2HPO4, 25 mM Tris-HCl and titrated to the specified pHs with HCl, and the pHs of all reaction mixtures were checked following preliminary reagent-mixing experiments that mimicked the experimental protocol. Shifts from pH 7.5 to 4.5 and back were accomplished by sequential addition of similar "adjustment" buffers, whose volumes and pHs had been predetermined to achieve the required transition. Finally, the pHs, ionic strengths, and volumes of all samples were normalized prior to immunoprecipitation. After cooling on ice, samples were immunoprecipitated using rabbit antibodies directed against intact MVM particles or against a recombinant peptide expressing the amino-terminal 141 amino acids of MVM VP1 and then analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and fluorography using 1 M salicylic acid, as previously described (11, 12, 16).
Trypsin digestion, Western blot analysis, virion repurification, and packaging assays. For trypsin digestion, virions were diluted into 100 mM NH4Cl (pH 5.5 to 7.5), 100 mM NaCl, 2.5 μM CaCl. TPCK-Trypsin (Cooper Biomedical, Malvern, PA) was added to a 2-ng/μl final concentration and incubated at the specified temperature for 1 h. The trypsin was inactivated by incubation in 4.5 mM phenylmethylsulfonyl fluoride and 20 mM EDTA for 10 min at room temperature prior to analysis by SDS-PAGE. The separated proteins were electroblotted onto polyvinylidene difluoride membranes using a Mini trans-blot transfer cell (Bio-Rad, Hercules, CA). The membranes were blocked and developed using the ECL system, according to the manufacturer's instructions (Amersham, Uppsala, Sweden).
For further analysis of trypsinized virions, samples were digested as described above, and the trypsin was inactivated with 15 mM EDTA and addition of soybean trypsin inhibitor to 10 times the predicted effective dose (Sigma Chemical Co., St. Louis, MO). Samples were immediately loaded onto a 4-ml iodixanol (OptiPrep; Axis-Shield, Oslo, Norway) step gradient (55% [500 μl], 45% [1 ml], 35% [1 ml], 25% [1 ml], and 15% [500 μl]). Gradient solutions were prepared either in PBS (pH 7.5) plus 1 mM MgCl2 and 2.5 mM KCl or in 50 mM MES (morpholineethanesulfonic acid) (pH 5.5) containing 120 mM NaCl, 1 mM MgCl2, and 2.5 mM KCl. Samples were centrifuged at 40,000 rpm for 20 h at 18°C in a Beckman SW50.1 rotor, and the fractions were collected and analyzed as described above. The micrococcal-nuclease sensitivities of particle-associated genomes were determined for each gradient fraction using the previously described packaging assays (16), following adjustment of the sample to pH 8.0.
RESULTS
The L172T mutant is temperature sensitive in the MVMp background. We have previously shown that substituting threonine for leucine 172 in the VP2 polypeptide of the immunosuppressive strain of MVM, MVMi, created a mutant virus that could assemble capsids and package progeny genomes effectively but that was defective for cell entry at 37°C (16). However, attempts to explore its entry phenotype further, at reduced temperatures, were hindered by the observation that when wild-type MVMi was assayed for viability at 32°C by expansion in iD5 cells, the virus strain proved to be cold sensitive (data not shown). However, when wild-type virus of the prototype strain of MVM, MVMp, was assayed in the same manner, its ability to initiate infection, as assessed by the appearance of the viral NS1 protein, was enhanced approximately 25-fold during the course of 3 days, indicating that MVMp could replicate productively at 32°C (Fig. 1C and D). Accordingly, we transferred the threonine mutation into the MVMp background, but to facilitate this switch, we first engineered a new wild-type infectious plasmid clone of MVMp called pdBMVp-pore, which contains silent mutations creating unique AflII and NsiI sites flanking L172. As illustrated in Fig. 1D, virions derived from pdBMVp and pdBMVp-pore behaved identically in viral expansion assays, indicating that the silent mutations had no appreciable effect on virus propagation. Throughout this report, therefore, the threonine mutation cloned into this MVMp background will be referred to simply as L172T, and pdBMVp-pore will be referred to as pdBMVp. Like its MVMi counterpart, when assayed for viral expansion at 37°C, L172T showed an 50-fold reduction in growth relative to wild-type virus (Fig. 1D). However, when cultured at 32°C for various periods posttransfection, viruses carrying this mutation expanded through the culture to levels that were within twofold of those seen for wild-type virus (Fig. 1D).
L172T is specifically temperature sensitive for the establishment of infection. To assess whether the growth defect exhibited by L172T at 37°C was due to the production of defective virions or to its inability to infect cells at this temperature, we purified virus that had been produced, following transfection, at either 32°C or 37°C and used it to infect A9 cells growing at 37°C. As seen in Fig. 2, wild-type virions produced at either temperature infected A9 cells efficiently, while even those L172T virions produced at 32°C failed to infect cells at 37°C, indicating that the defect operated during the establishment of infection. L172T inoculated at 48,000 genomes per cell gave the same number of infected cells as seen with wild-type virus at only 750 genomes per cell, suggesting that, at 37°C, L172T virions were approximately 64-fold less efficient at initiating infection than pdBMVp.
When cells were infected at 32°C, however, a strikingly different pattern was observed, with L172T virions establishing infection as efficiently as pdBMVp, regardless of the temperature at which they were produced (Fig. 2). At this lower temperature, infections with pdBMVp were somewhat less efficient than at 37°C, giving initiation efficiencies that were approximately eightfold lower throughout the lower range of input multiplicities (187.5 to 3,000 genomes per cell), while at higher multiplicities, the difference in infectivity at 32°C and 37°C appeared negligible. These experiments therefore indicate that L172T virions are intrinsically stable once assembled and packaged but are temperature sensitive for the establishment of a subsequent cycle of infection.
L172T virions are rapidly inactivated following infection at 37°C and must traffic to a late compartment before they are protected. To investigate whether L172T virions became inactivated at 37°C in a specific compartment in the entry pathway, we performed the temperature shift experiment documented in Fig. 3A. A9 cells were infected for 2 h at 32°C with 12,000 genome-containing virions per cell, washed, and transferred to medium containing neuraminidase, which blocks further infection. Immediately thereafter, and at subsequent 2-hour intervals, samples were shifted to 37°C. Remarkably, when compared to samples incubated at 37°C for the whole of the experiment, incubation at 32°C for periods of less than 8 h had little protective effect, and it was not until the shift occurred after 8 or 10 h at 32°C that infection rates began to resemble those seen with cultures maintained constantly at 32°C. This suggests either that inactivation occurs at a late point during entry or that it can occur at any time during that process.
Figure 3B documents the inverse experiment, where cells were initially infected at 37°C and then shifted down to 32°C at different times postinfection. If L172T virions were specifically inactivated in a late entry compartment, short incubations at 37°C during early time points should have had no effect on infection. However, L172T was rapidly inactivated during entry at 37°C. Compared to control samples maintained at 32°C throughout, infectivity levels dropped 2-, 10-, and 20-fold following incubation at 37°C for 2, 4, or 8 h, respectively. Thus, incubating cells at 37°C for even a few hours during the first 8 to 10 h of an L172 infection markedly impaired infection rates.
Since L172T virions produced at 37°C are infectious at 32°C, we did not expect that simply incubating virions at 37°C would destroy their infectivity. However, we verified this by incubating purified virions in vitro for 2 h at 37°C before using them to infect cells growing at 32°C. As shown in Fig. 3C, in vitro incubation at 37°C had no effect on subsequent infection rates. Similarly, L172T virions incubated in vitro at 37°C for 2 h with medium supplemented with 10% A9-conditioned medium prior to infection at 32°C remained fully infectious (Fig. 3D), indicating that the cells do not produce a soluble inactivating factor but that actual contact with the host cell at 37°C is required for inactivation. Infectious entry of wild-type MVMp at 37°C has been shown to require exposure to low endosomal pH (29). Similarly, exposing cultures to the vacuolar proton pump inhibitor bafilomycin A1 for 1 hour prior to and for 3 h during incubation with pdBMVp or L172T at 32°C reduced the infectivity of both viruses approximately sixfold, as did incubation of pdBMVp infections at 37°C, while incubation of L172T-infected cells with inhibitor at this temperature further impaired its residual infectivity (data not shown). Thus, entry of both pdBMVp and L172T MVMp appears to require passage through an acidified endosomal compartment at either 32°C or 37°C.
VP2-to-VP3 cleavage lowers the energy required to expose VP1 N termini at neutral pH but makes this transition highly dependent upon pH. Before pursuing the basis of the L172T defect further, we sought to clarify our understanding of the structural transition that MVM particles are thought to undergo during entry, which results in the exposure of VP1 N termini at the virion surface. These termini are sequestered within infectious virions, but wild-type MVMi and MVMp and MVMi L172T particles have been shown to be metastable in vitro, undergoing a structural rearrangement in response to limited heating in which the particles remain intact but at least some of the VP1 N termini become externalized (11, 16). Specifically, we wished to ask whether prior cleavage of VP2 to VP3 significantly influenced the energy required for this transition and whether the low-pH conditions encountered during entry would impinge on the transition process. To this end, we prepared [35S]methionine/[35S]cysteine-labeled wild-type MVMp virions in which the VP2 N termini were intact (hereafter called VP1/VP2 virions), digested half of them exhaustively with trypsin, and rebanded them, thus generating a population in which the great majority of VP2 N termini had been cleaved to VP3 (called VP1/VP3 virions). These two forms of wild-type MVMp were then incubated for 10 min at various temperatures and pHs, as indicated in Fig. 4, before the buffer concentrations and pHs were normalized, and their structures were assessed by immunoprecipitation. Following exposure to neutral pH at 4°C, very few VP1/VP2 virions had accessible VP1 N termini, but heating them to increasing temperatures at pH 7.5 induced progressive exposure of these peptides (Fig. 4A, lanes 2 to 5), although only a relatively small proportion of particles (<10%) underwent this transition even after being heated to 65°C in the experiment shown here.
In contrast, exposure to pH 4.5 caused VP1 N termini to become accessible in a minor, but detectable, proportion of VP1/VP2 particles, but subsequent heating to 45 or 55°C had no further effect on VP1 accessibility, although surface-exposed VP1 N termini were detected in <10% of virions after heating them to 65°C (Fig. 4A, lanes 6 to 9). Since MVMp has a particle-to-infectivity ratio in A9 cells of around 300:1 (32), it is formally possible that the small proportion of VP1/VP2 virions in which VP1SR becomes exposed at pH 4.5 constitute essential entry intermediates. However, this is unlikely because, as discussed below, trypsin digestion at low pH, which would preclude both the formation and integrity of such an intermediate, fails to inactivate wild-type MVMp.
N-terminal peptides were fully protected in VP1/VP3 virion populations at 4°C and neutral pH (Fig. 4B, lane 2), but exposing these particles to heat had a much more profound effect on particle structure, causing VP1 N termini to become surface accessible on most virions after being heated to 55°C and on all virions after being heated to 65°C (Fig. 4B, compare lanes 1 and 5). To confirm that particles remained intact throughout this procedure, we also precipitated samples that had been heated to 65°C at pH 4.5 and 7.5 with antibodies that detect only intact particles (lanes 10 in Fig. 4A and 4B, respectively). Thus, the energy required to induce particle transition is markedly reduced following VP2-to-VP3 cleavage, so that particles universally undergo this rearrangement at temperatures that are easily compatible with particle integrity. We suggest, therefore, that during cell entry, VP2 cleavage may be an important and enabling prelude to VP1 extrusion.
Shifting VP1/VP3 virions to pH 4.5, however, effectively inhibited all such rearrangements (Fig. 4B, lanes 6 to 9). Thus, following cleavage of the VP2 N termini, wild-type virions become completely refractory to the conformational shift required for VP1 exposure. However, this change is not permanent, since it can be reversed simply by shifting particles back from acidic to neutral pH, as shown in the experiment illustrated in Fig. 4C, where VP1SR exposure following heating to 55°C for 10 min was compared in samples that had been either maintained at pH 7.5 throughout the experiment (Fig. 4C, lane 3) or shifted from pH 7.5 to pH 4.5 for 20 min (lane 5) and then shifted back to neutral pH (lane 10). Thus, cleavage of VP2 to VP3 profoundly increases the susceptibility of MVMp virions to a critical structural rearrangement, but this shift becomes markedly pH dependent.
Trypsin digestion of L172T VP2 at pH 7.5 leads to exposure of VP1-specific sequences, even at room temperature, except at low pH. The immunoprecipitation approach shown in Fig. 4 to assess the effects of VP2 cleavage proved unsuitable for analysis of the L172T mutant because it was not possible to purify homogeneous L172T populations following tryptic digestion. Instead, we analyzed VP1 exposure in the mutant by its susceptibility to cleavage with trypsin. Using MVMi-based virions carrying this mutation, we had previously shown that, while VP1 was sequestered from antibodies in newly purified virions, it could be partially degraded when these were exposed to trypsin at 37°C, whereas the peptide remained protected in wild-type particles (16). To analyze whether this susceptibility persisted in the cold-insensitive MVMp genetic background, stocks of MVMp wild-type and L172T virions were harvested and purified by standard procedures, which typically gave a mixture of cleaved and intact forms of the VP2 N termini, and then were further digested under experimental conditions with trypsin, and the products were analyzed by SDS-PAGE. Digestion of wild-type virions at neutral pH and either 37°C or 30°C resulted in cleavage of most of the remaining VP2 polypeptides to VP3, but VP1 was not affected (Fig. 5A). In contrast, but consistent with the MVMi mutant, trypsin digestion of MVMp L172T at neutral pH and either 30°C or 37°C led to cleavage of both VP2 and a large proportion of the VP1 N termini, indicating that these peptides become exposed at the surfaces of the VP2-cleaved L172T particles, irrespective of virion tropism. Since the L172T mutant is temperature sensitive for entry, we had suspected that digestion at lower temperatures might prevent exposure of VP1 N termini, but we conclude from this experiment that at neutral pH, cleavage of VP2 N termini is linked to exposure of VP1 N termini, even at 30°C. However, because in vivo exposure at 32°C might not be adequately mimicked by cleavage at similar temperatures under these somewhat artificial in vitro conditions, we next digested virions at 23°C. As seen in Fig. 5B, cleavage at pH 7.5 again resulted in the loss of some L172T VP1 N termini. Thus, we conclude that even at room temperature VP2 cleavage allows L172T virions to transition to a VP1-exposed configuration.
In the same experiment (Fig. 5B), mutant L172T particles were also incubated with trypsin at pH 6.5. When digested at 23°C, some VP1 loss was still apparent at pH 6.5, but the phenotype was partially rescued, while at 30°C or 37°C (Fig. 5C), VP1 degradation was comparable to that seen at pH 7.5 in Fig. 5A. In contrast, when the pH of the reaction mixture was lowered further to 5.5, degradation of VP1 was completely abrogated at either 30°C or 37°C (Fig. 5D). Thus, as in the immunoprecipitation experiments with wild-type virions presented in Fig. 4, incubation at low pH following VP2 cleavage prevented L172T virions from transitioning into a VP1-exposed configuration. Since during infection, VP2-to-VP3 cleavage routinely occurs during passage of the virus through degradative compartments within the cytosol, this suggests that cleavage in compartments where the pH is 6.5 or above might lead to loss of VP1-specific functions, whereas VP2 cleavage at pHs of 5.5 or below, as might be encountered in late endosomes, would allow the particle to retain its VP1 N termini, and hence its endosomal escape function and nuclear trafficking signals.
Infectivity of L172T, but not of wild-type MVMp, is abrogated by prior cleavage of VP2 N termini. During entry, MVM would be expected to encounter neutral or somewhat acidic cellular compartments before it reached a protective low-pH environment. If so, cleaving the VP2 N termini of L172 virions in vitro at pH 5.5, which leaves their VP1 N termini intact, might nonetheless effectively inactivate the virus, since such virions would transition at an early stage of entry, exposing their previously sequestered sequences to potential degradation. To explore this possibility, wild-type and L172T virions were digested with trypsin at 30°C, at either pH 5.5 or 7.5. The trypsin was then inactivated by addition of 10% fetal bovine serum, and the resulting viruses were used to infect cells at 32°C. Although the addition of 10% serum inhibited pdBMVp infections slightly, trypsin digestion had little effect on the infectivity of wild-type virus relative to untreated controls, as seen in Fig. 6. However, prior digestion at either pH 5.5 or 7.5 totally abrogated the infectivity of L172T. Thus, provided the majority of their VP2 N termini are intact when they are first taken up by the cell, L172T virions are infectious, but prior cleavage of VP2 to VP3 is effectively lethal.
Loss of VP1 integrity leads to exposure of the viral genome. To explore the structural integrity of L172T virions after proteolytic cleavage, we again digested virions with trypsin at 30°C at either pH 5.5 or 7.5, inactivated the trypsin with EDTA and soy bean trypsin inhibitor, and sedimented the digested virions to equilibrium in iodixanol gradients buffered at the appropriate pH. Fractions from these gradients were analyzed by Western blotting, using antibodies directed against MVM capsid proteins. The gradient profile of wild-type virions after tryptic cleavage at either pH 5.5 or 7.5 was essentially unaltered (Fig. 7A), with virions reaching equilibrium in fractions that contained approximately 50% iodixanol. Similarly when L172T virions were banded at pH 5.5 following digestion at pH 5.5, they reached equilibrium at approximately this density (Fig. 7B). However, virions digested and centrifuged at pH 7.5 were dispersed more broadly through the gradient, reaching a peak in fraction 9, indicating that they had become substantially less dense (Fig. 7C).
Fractions from the pH 5.5 gradient of L172T virions were then analyzed by denaturing gel electrophoresis, before and after nuclease digestion, as shown in Fig. 7D, left. The leading edge of this peak was seen to contain predominantly full-length genomes that remained resistant to nuclease digestion, while trailing fractions tended to contain more subgenomic DNA, much of which nevertheless remained protected from nuclease and thus sequestered. In contrast, while full-length copies of the viral genome still sedimented with capsids after digestion at pH 7.5, they now peaked in fraction 9, and the genomes were not protected from nuclease digestion (Fig. 7D, right). Most of these genomes were extensively degraded and could not be detected with our probe, while some remained as a smear of subgenomic protected species. Thus, proteolysis of L172T at pH 5.5 causes cleavage of VP2-specific sequences, but the particle remains intact and is able to protect its genome. On the other hand, proteolysis of L172T at pH 7.5 causes loss of both VP2 and VP1 N termini and a substantial drop in particle density. Since this density shift reflects exposure of the viral DNA, which remains attached to, and partially protected by, the particle, this suggests that the fivefold pores of L172T are unable to contain their VP1 N termini or, if these are cleaved, prevent the subsequent release of the genome. In the experiments presented here, virions cleaved and centrifuged at pH 5.5 were digested with nuclease for 30 min at 37°C and neutral pH to ensure that the nuclease was fully functional. Thus, VP2-cleaved but VP1-intact L172T virions retain their integrity at neutral pH, at least for short periods of time.
DISCUSSION
VP2-to-VP3 cleavage is a priming event for VP1 extrusion. We show here that the hollow cylinder surrounding each of the 12 fivefold-symmetry axes in the MVM virion plays an essential role in controlling particle structure and integrity. Leucine residues from each of the fivefold related polypeptides create the narrowest (8-) constriction point at VP2 residue 172, situated near the base of these pores, as shown in Fig. 1A, but when the diameter of the channel is somewhat expanded at this position by the relatively conservative substitution of aliphatic threonine hydroxyl groups for the somewhat more bulky hydrophobic leucine side chains, the virion becomes conditionally stable, remaining intact and infectious provided its VP2 N termini are uncleaved. However, it rapidly unravels to expose its essential, but vulnerable, VP1 N termini when 22 amino acids from the N termini of VP2 are removed. These exposed VP2 N termini, which project to the particle exterior through the fivefold channels, thus have a major constraining effect on particle dynamics in the L172T mutant, but this also appears to be true in wild-type virus. MVM virions are known to be metastable, initially sequestering their VP1-specific "entry" peptide, which carries a PLA2 enzymatic core required for endosome escape, but extruding this domain to the outside surface of the otherwise-intact virion in response to externally applied energy. However, virions that have intact VP2 N termini undergo this transition somewhat inefficiently and following exposure to relatively high temperatures, possibly because VP1 N termini must compete with the VP2 N termini for their normal egress portal. In contrast, once these VP2 tethers have been cleaved, particles transition quantitatively to expose at least one VP1 N terminus per particle and at lower temperatures. This suggests that VP2-to-VP3 cleavage is a natural component of virion maturation, priming it for the subsequent emergence of VP1-specific sequences. Thus, generation of VP3 polypeptides corresponds to the preprotein cleavage step characterized in some other nonenveloped viruses (6, 8), priming the virion for an upcoming transition to a lower free-energy state in which the VP1 N-terminal entry peptide is exposed.
While the importance of the VP2 cleavage has long been suspected for viruses in the genus Parvovirus, it has proven remarkably hard to evaluate experimentally. In part, this is because cleavage appears to occur in a particular region of the peptide, where it emerges from the mouth of the fivefold cylinder, rather than at a specific residue, and the relevant proline-rich N-terminal peptide is highly exposed. Thus, it can be cleaved in vitro by a wide variety of both endopeptidases and exopeptidases, while the rest of the particle remains untouched, and we have been unable to prevent cleavage from occurring during infection in cell culture using a broad range of class-specific protease inhibitors, both alone and in various combinations (S. F. Cotmore and P. Tattersall, unpublished observations; 16a), suggesting that there are multiple proteases resident in the viral entry pathway that can effect this truncation. This sequence is still cleaved effectively in vivo, albeit rather less efficiently, even following ablation of the potential furin-like protease target motif, N-RVER-C, in this peptide, which is also the site of tryptic cleavage (37), or if this region is replaced with an unrelated 7-amino-acid cleavage site for a plant virus protease (Cotmore and Tattersall, unpublished). Finally, since even wild-type MVMp has a particle-to-infectivity ratio of around 300, structural analysis of virions retrieved from infected cells, for VP2-intact or VP2-cleaved forms, are hard to interpret, since it is difficult to ascertain whether these are part of the infectious process or the products of misrouting through nonproductive cellular compartments. Three observations relevant to this problem are documented in this study. First, VP2 cleavage predisposes the virion to VP1 extrusion; second, this cleavage makes particle dynamics highly responsive to changes in ambient pH; and third, L172 mutants, which have compromised fivefold cylinders, transition at very low temperatures but can use their intact VP2 N termini to maintain particle integrity. Together, these observations provide the first clear evidence of a central role for VP2 cleavage in the infectious process.
Exactly how the constraining effects of low pH on particle structure operate following this cleavage will be of great interest to explore structurally, but presumably cleavage allows the particle to adopt a lower free-energy state in which charged groups effectively control pore dynamics. The biological implications of these findings are significant, since they suggest that cleaved forms of the virion that are present in low-pH compartments within the cell would be unlikely to undergo the structural rearrangements that appear to be required for traversing the endosomal bilayer. In consequence, we suggest that such particles must either be trafficked further, to a suitable neutral compartment or microdomain, or be involved in some other structural interaction, perhaps with the viral receptor, that effects this rearrangement. In this respect, it is of particular interest that for the related virus CPV, in which entry is known to be mediated via interactions with the transferrin receptor (27), antibodies directed against the cytosolic domain of this receptor can still inhibit infection when injected into the cytoplasm 4 hours after viral uptake, suggesting that virions may remain in contact with their receptor for several hours (20). However, for CPV, treatment of particles in vitro at pH 4 to 6 has been reported to result in an increase in PLA2 activity, which may indicate that the CPV VP1 N terminus becomes exposed at low pH (31), perhaps indicating a striking difference between the chemistries of MVM and CPV particles, and perhaps also in their entry strategies. It should also be noted, however, that exposure of VP1-specific peptides in an endosomal compartment several hours after viral uptake has been demonstrated for both MVM and CPV by immunofluorescence microscopy (30, 31), although whether these particles are engaged in infectious entry or have transitioned prematurely and are being degraded remains equivocal.
What is the L172T defect and how is it rescued at low temperature Purified L172T virions are resistant to in vitro incubation at 37°C, and cells transfected at this temperature produce infectious virus, suggesting that a cellular factor to which virions are exposed during entry is responsible for their temperature-sensitive inactivation. However, evidence presented here for MVMp virions bearing the L172T substitution, as well as previous data for this mutant in the MVMi strain, suggests that these viruses are stable and appear structurally normal until their VP2 N termini are cleaved, at which point they are no longer able to protect their VP1-specific peptides, at least at neutral pH. Supporting this argument, cleavage of VP2 to VP3 in vitro prior to infection completely inactivated L172, even if this cleavage was carried out at pH 5.5 in order to ensure that its VP1 N termini were intact at the time it was added to the cells. Extrusion of the VP1 N termini was also tightly linked to exposure of the viral genome, although at present we do not understand the exact phasing of these two processes. In vitro, heat-induced transitions also lead to exposure of much of the DNA without VP1 cleavage, so that while in the experiments presented here such cleavage apparently preceded genome exposure, this is unlikely to be absolutely required. However, if either VP1 N termini or the genome were to become accessible during passage through early entry endosomes, it would likely be exposed to hydrolytic enzymes or the depurinating effects of low pH, so that one would expect the virus to be inactivated. Thus, we suggest that L172T is conditionally viable predominantly because lower temperatures substantially reduce the likelihood of the VP2-to-VP3 cleavage occurring in an early entry compartment, where the ambient pH is higher than 5.5. Our data also support the idea that parvoviral entry is a protracted procedure, requiring passage through acidified compartments, and that VP2 cleavage can be effected by a variety of cellular proteases in multiple compartments during this process. Since VP2 cleavage primes the virion for subsequent rearrangement, we further suggest that, for L172T, such cleavage must reliably occur in a low-pH environment, where the particle is stabilized, since proteolysis could then proceed without causing the particle to unravel. Thus, we suggest that L172T may be infectious at 32°C because cellular proteases are less active at this temperature, allowing the particle to access a low-pH compartment prior to VP2 cleavage.
While L172T virions can be inactivated within the first 2 hours of infection at 37°C, a shift up to 37°C even 8 h into infection at 32°C still influences the success of this process, suggesting that allowing particles to traffic deep into the endocytic pathway does not totally protect them. Indeed, by 10 h postinfection at 32°C, when L172T infection finally becomes completely resistant to shift-up, many of the infecting virions may well have entered the cytosol. This suggests that the L172T mutant could also have a second defect, encountered in a late entry compartment, which causes infection to proceed more reliably at lower temperature. While the nature of this defect remains obscure, we do know that in coinfection experiments at 37°C, the L172T mutant can complement an endosomal penetration defect seen in PLA2-null viruses as well as wild-type virus can (Farr et al., submitted), but without itself establishing infection in the same cells. In light of this experiment, we conclude that either proteolysed VP1 N termini can still support endosomolysis or, perhaps more probably, it is damage to the viral genome that most commonly inactivates L172T mutants at 37°C. Because this premature extrusion defect may mimic a normal event that for the wild type occurs later in infection, we are currently exploring the mechanism of L172T genome exposure in greater detail, as a paradigm for parvoviral uncoating.
Role of the fivefold cylinder. VP2 N termini are sequestered within empty MVM particles, but at least some of these termini are extruded to the particle exterior during genome encapsidation. A glycine-rich sequence connects these extruded N termini to the rest of the VP2 polypeptide. However, in the crystal structure, the fivefold pore is only wide enough to accommodate the connector sequences, so that it is not clear how the bulkier N-terminal peptide gets to the outside of the capsid shell. Moreover, at least three more VP2 N termini are initially positioned at the base of each cylinder, and in the mature virion, most of these can be cleaved by externally applied protease. Since the X-ray structures clearly indicate that these pores cannot accommodate more than one of these connector sequences at a time, this suggests a model in which, following removal of its external N-terminal domain, the resident connector somehow retracts, allowing another, intact VP2 N terminus to take its place. Since 22 N-terminal amino acids of this VP2 molecule must also be threaded through the pore, it seems probable that the cylinder opens or is otherwise somehow enlarged, perhaps transiently, but repeatedly, to permit the sequential egress of these domains. If so, this makes it easier to imagine how leucine 172, positioned at the base of the cylinder, might prove critical for particle integrity. Thus, if much of the cylinder wall can flex, the annulus formed by leucine at its base could serve as the ultimate retainer for structures that still need to be sequestered within the capsid. In support of such a model, we have previously observed that in MVMi-based L172 mutants, the VP1 N-terminal peptides become susceptible to trypsin degradation not only when this residue is replaced by amino acids with somewhat smaller side groups, such as threonine or glycine, but also when it is replaced by those with bulkier, aromatic side groups, such as phenylalanine (16), suggesting that leucine 172 is somehow uniquely suited to maintaining the integrity of this molecular "iris diaphragm."
It seems likely, moreover, that these cylinders not only support the egress of VP2 N-terminal peptides, but that one of them may also be involved in the presumably much more extensive rearrangements needed for it to serve as the portal for viral DNA. Thus, when leucine 172 in MVMi was replaced by tryptophan, empty capsids assembled effectively at 32°C but were unable to package DNA, while substitution of aspartic or glutamic acid at this position seriously impaired, but did not totally prevent, packaging. Somewhat similar observations were made for adeno-associated virus type 2 (AAV2) by Bleker and colleagues (5), who showed that packaging efficiency could be impeded by specific mutations in the funnel-shaped structure that forms the base of the AAV pore and by other mutations positioned in nearby regions of the cylinder wall. However, in AAV, mutation of "funnel" residues at the base of the pore did not lead to premature exposure of VP1 N termini when the particle was heated in vitro but rather to their aberrant retention. Remarkably, substitution of alanine for MVM residue F526, V40, N149, S43, or D263, which are positioned around the base of the fivefold pore, or for N170, positioned in the wall, had a similarly negative effect on pore flexibility, at least as assessed by tryptophan exposure in baculovirus-expressed virus-like empty particles assembled from VP2 molecules (28). In these studies conformational shifts were assessed by monitoring changes in fluorescence as a function of temperature, revealing an abrupt shift for wild-type sequences at around 46°C, which was shown to coincide with extrusion of some VP2 N termini, possibly through the fivefold pores. In the alanine mutants mentioned above, this shift was ablated, and in some cases extrusion of VP2 N termini was shown to be impaired, suggesting that, like the AAV substitutions, mutation of some of the flanking sequences in MVM may depress pore flexibility. Thus, it will be of great interest to see if these alanine substitutions exert a similar phenotype when expressed as part of an MVM virion. However, the structures of the MVM and AAV pores are substantially different, and AAV virions are constructed with only five copies each of their VP2 and VP1 N-terminal peptides, so that their particle dynamics may be markedly different from those of members of the genus Parvovirus. Nevertheless, in both cases, the fivefold pore appears to be involved in extrusion of the PLA2-bearing VP1 N termini, since AAV pore mutants are unable to effectively expose this essential VP1 domain, whereas the MVM VP2 L172 mutants are unable to effectively protect it from degradation at neutral pH.
ACKNOWLEDGMENTS
We thank colleagues at the HHMI Biopolymer Laboratory and W. M. Keck Foundation Biotechnology Resource Laboratory at Yale University for DNA sequencing and synthesis of oligonucleotides used in this project.
G.A.F. was supported, in part, by T32 AI07640. This work was supported by Public Health Service grant CA29303 from the National Cancer Institute.
REFERENCES
Agbandge-McKenna, M., and M. S. Chapman. 2005. Correlating structure with function in the viral capsid, p. 125-139. In J. R. Kerr, S. F. Cotmore, M. E. Bloom, R. M. Linden, and C. R. Parrish (ed.), Parvoviruses. Edward Arnold, London, United Kingdom.
Agbandje-McKenna, M., A. L. Llamas-Saiz, F. Wang, P. Tattersall, and M. G. Rossmann. 1998. Functional implications of the structure of the murine parvovirus, minute virus of mice. Structure 6:1369-1381.
Belnap, D. M., D. J. Filman, B. L. Trus, N. Cheng, F. P. Booy, J. F. Conway, S. Curry, C. N. Hiremath, S. K. Tsang, A. C. Steven, and J. M. Hogle. 2000. Molecular tectonic model of virus structural transitions: the putative cell entry states of poliovirus. J. Virol. 74:1342-1354.
Belnap, D. M., B. M. McDermott, Jr., D. J. Filman, N. Cheng, B. L. Trus, H. J. Zuccola, V. R. Racaniello, J. M. Hogle, and A. C. Steven. 2000. Three-dimensional structure of poliovirus receptor bound to poliovirus. Proc. Natl. Acad. Sci. USA 97:73-78.
Bleker, S., F. Sonntag, and J. A. Kleinschmidt. 2005. Mutational analysis of narrow pores at the fivefold symmetry axes of adeno-associated virus type 2 capsids reveals a dual role in genome packaging and activation of phospholipase A2 activity. J. Virol. 79:2528-2540.
Bubeck, D., D. J. Filman, N. Cheng, A. C. Steven, J. M. Hogle, and D. M. Belnap. 2005. The structure of the poliovirus 135S cell entry intermediate at 10-angstrom resolution reveals the location of an externalized polypeptide that binds to membranes. J. Virol. 79:7745-7755.
Chandran, K., D. L. Farsetta, and M. L. Nibert. 2002. Strategy for nonenveloped virus entry: a hydrophobic conformer of the reovirus membrane penetration protein micro 1 mediates membrane disruption. J. Virol. 76:9920-9933.
Chandran, K., J. S. Parker, M. Ehrlich, T. Kirchhausen, and M. L. Nibert. 2003. The delta region of outer-capsid protein micro 1 undergoes conformational change and release from reovirus particles during cell entry. J. Virol. 77:13361-13375.
Chapman, M. S., and M. Agbandje-McKenna. 2005. Atomic structure of viral particles, p. 107-123. In J. R. Kerr, S. F. Cotmore, M. E. Bloom, R. M. Linden, and C. R. Parrish (ed.), Parvoviruses. Edward Arnold, London, United Kingdom.
Clinton, G. M., and M. Hayashi. 1975. The parovivirus MVM: particles with altered structural proteins. Virology 66:261.
Cotmore, S. F., A. M. D'Abramo, Jr., C. M. Ticknor, and P. Tattersall. 1999. Controlled conformational transitions in the MVM virion expose the VP1 N-terminus and viral genome without particle disassembly. Virology 254:169-181.
Cotmore, S. F., A. M. D'Abramo, Jr., L. F. Carbonell, J. Bratton, and P. Tattersall. 1997. The NS2 polypeptide of parvovirus MVM is required for capsid assembly in murine cells. Virology 231:267-280.
Cotmore, S. F., and P. Tattersall. 1990. Alternate splicing in a parvoviral nonstructural gene links a common amino-terminal sequence to downstream domains which confer radically different localization and turnover characteristics. Virology 177:477-487.
Cotmore, S. F., and P. Tattersall. 2005. Encapsidation of minute virus of mice DNA: aspects of the translocation mechanism revealed by the structure of partially packaged genomes. Virology 336:100-112.
Dorsch, S., G. Liebisch, B. Kaufmann, P. von Landenberg, J. H. Hoffmann, W. Drobnik, and S. Modrow. 2002. The VP1 unique region of parvovirus B19 and its constituent phospholipase A2-like activity. J. Virol. 76:2014-2018.
Farr, G. A., and P. Tattersall. 2004. A conserved leucine that constricts the pore through the capsid fivefold cylinder plays a central role in parvoviral infection. Virology 323:243-256.
Farr, G. A. 2005. Ph.D. thesis. Yale University, New Haven, Conn.
Fricks, C. E., and J. M. Hogle. 1990. Cell-induced conformational change in poliovirus: externalization of the amino terminus of VP1 is responsible for liposome binding. J. Virol. 64:1934-1945.
Girod, A., C. E. Wobus, Z. Zadori, M. Ried, K. Leike, P. Tijssen, J. A. Kleinschmidt, and M. Hallek. 2002. The VP1 capsid protein of adeno-associated virus type 2 is carrying a phospholipase A2 domain required for virus infectivity. J. Gen. Virol. 83:973-978.
Hogle, J. M. 2002. Poliovirus cell entry: common structural themes in viral cell entry pathways. Annu. Rev. Microbiol. 56:677-702.
Hueffer, K., L. M. Palermo, and C. R. Parrish. 2004. Parvovirus infection of cells by using variants of the feline transferrin receptor altering clathrin-mediated endocytosis, membrane domain localization, and capsid-binding domains. J. Virol. 78:5601-5611.
Kestler, J., B. Neeb, S. Struyf, J. Van Damme, S. F. Cotmore, A. D'Abramo, P. Tattersall, J. Rommelaere, C. Dinsart, and J. J. Cornelis. 1999. cis requirements for the efficient production of recombinant DNA vectors based on autonomous parvoviruses. Hum. Gene Ther. 10:1619-1632.
Kronenberg, S., B. Bottcher, C. W. von der Lieth, S. Bleker, and J. A. Kleinschmidt. 2005. A conformational change in the adeno-associated virus type 2 capsid leads to the exposure of hidden VP1 N termini. J. Virol. 79:5296-5303.
Llamas-Saiz, A. L., M. Agbandje-McKenna, W. R. Wikoff, J. Bratton, P. Tattersall, and M. G. Rossmann. 1997. Structure determination of minute virus of mice. Acta Crystallogr. D 53:93-102.
Lombardo, E., J. C. Ramirez, J. Garcia, and J. M. Almendral. 2002. Complementary roles of multiple nuclear targeting signals in the capsid proteins of the parvovirus minute virus of mice during assembly and onset of infection. J. Virol. 76:7049-7059.
Maroto, B., N. Valle, R. Saffrich, and J. M. Almendral. 2004. Nuclear export of the nonenveloped parvovirus virion is directed by an unordered protein signal exposed on the capsid surface. J. Virol. 78:10685-10694.
Paradiso, P. R., K. R. Williams, and R. L. Costantino. 1984. Mapping of the amino terminus of the H-1 parvovirus major capsid protein. J. Virol. 52:77-81.
Parker, J. S., W. J. Murphy, D. Wang, S. J. O'Brien, and C. R. Parrish. 2001. Canine and feline parvoviruses can use human or feline transferrin receptors to bind, enter, and infect cells. J. Virol. 75:3896-3902.
Reguera, J., A. Carreira, L. Riolobos, J. M. Almendral, and M. G. Mateu. 2004. Role of interfacial amino acid residues in assembly, stability, and conformation of a spherical virus capsid. Proc. Natl. Acad. Sci. USA 101:2724-2729.
Ros, C., C. J. Burckhardt, and C. Kempf. 2002. Cytoplasmic trafficking of minute virus of mice: low-pH requirement, routing to late endosomes, and proteasome interaction. J. Virol. 76:12634-12645.
Ros, C., and C. Kempf. 2004. The ubiquitin-proteasome machinery is essential for nuclear translocation of incoming minute virus of mice. Virology 324:350-360.
Suikkanen, S., M. Antila, A. Jaatinen, M. Vihinen-Ranta, and M. Vuento. 2003. Release of canine parvovirus from endocytic vesicles. Virology 316:267-280.
Tattersall, P. 1972. Replication of the parvovirus MVM. I. Dependence of virus multiplication and plaque formation on cell growth. J. Virol. 10:586-590.
Tattersall, P., and J. Bratton. 1983. Reciprocal productive and restrictive virus-cell interactions of immunosuppressive and prototype strains of minute virus of mice. J. Virol. 46:944-955.
Tattersall, P., P. J. Cawte, A. J. Shatkin, and D. C. Ward. 1976. Three structural polypeptides coded for by minute virus of mice, a parvovirus. J. Virol. 20:273-289.
Tattersall, P., A. J. Shatkin, and D. C. Ward. 1977. Sequence homology between the structural polypeptides of minute virus of mice. J. Mol. Biol. 111:375-394.
Tullis, G. E., L. R. Burger, and D. J. Pintel. 1993. The minor capsid protein VP1 of the autonomous parvovirus minute virus of mice is dispensable for encapsidation of progeny single-stranded DNA but is required for infectivity. J. Virol. 67:131-141.
Tullis, G. E., L. R. Burger, and D. J. Pintel. 1992. The trypsin-sensitive RVER domain in the capsid proteins of minute virus of mice is required for efficient cell binding and viral infection but not for proteolytic processing in vivo. Virology 191:846-857.
Valle, N., L. Riolobos, and J. M. Almendral. 2005. Synthesis, post-translational modification and trafficking of the parvovirus structural polypeptides, p. 291-304. In J. R. Kerr, S. F. Cotmore, M. E. Bloom, R. M. Linden, and C. R. Parrish (ed.), Parvoviruses. Edward Arnold, London, United Kingdom.
Vihinen-Ranta, M., and C. R. Parrish. 2005. Cell infection processes of autonomous parvoviruses, p. 157-163. In J. R. Kerr, S. F. Cotmore, M. E. Bloom, R. M. Linden, and C. R. Parrish (ed.), Parvoviruses. Edward Arnold, London, United Kingdom.
Vihinen-Ranta, M., W. Yuan, and C. R. Parrish. 2000. Cytoplasmic trafficking of the canine parvovirus capsid and its role in infection and nuclear transport. J. Virol. 74:4853-4859.
Vihinen-Ranta, M., D. Wang, W. S. Weichert, and C. R. Parrish. 2002. The VP1 N-terminal sequence of canine parvovirus affects nuclear transport of capsids and efficient cell infection. J. Virol. 76:1884-1891.
Wiethoff, C. M., H. Wodrich, L. Gerace, and G. R. Nemerow. 2005. Adenovirus protein VI mediates membrane disruption following capsid disassembly. J. Virol. 79:1992-2000.
Zadori, Z., J. Szelei, M. C. Lacoste, Y. Li, S. Gariepy, P. Raymond, M. Allaire, I. R. Nabi, and P. Tijssen. 2001. A viral phospholipase A2 is required for parvovirus infectivity. Dev. Cell. 1:291-302.(Glen A. Farr, Susan F. Co)