Stable Ubiquitination of the ICP0R Protein of Herpes Simplex Virus Type 1 during Productive Infection

Virology 253, 288–298 (1999) Article ID viro.1998.9502, available online at http://www.idealibrary.com on

Stable Ubiquitination of the ICP0R Protein of Herpes Simplex Virus Type 1 during Productive Infection Peter C. Weber,1 Stephen J. Spatz,2 and Eric C. Nordby Infectious Diseases Section, Parke–Davis Pharmaceutical Research, Division of Warner–Lambert Company, 2800 Plymouth Road, Ann Arbor, Michigan 48105 Received August 20, 1998; returned to author for revision October 6, 1998; accepted November 2, 1998 ICP0R is the polypeptide product of an alternatively spliced transcript of the gene encoding the transactivator protein ICP0 of herpes simplex virus type 1 (HSV-1). Although it has been shown to act as a transrepressor of gene expression in transfection assays, overexpression of the ICP0R protein in the recombinant virus HSV–KST was previously found to have no detectable effect on virus replication, so that the role it plays in HSV-1 infection remains unclear. Analysis of HSV–KSTinfected cell lysates by Western blotting revealed the presence of not only the 41-kDa ICP0R polypeptide but also a 64-kDa processed form of the protein. This processing event was the result of ubiquination of the ICP0R protein, as demonstrated by the reactivity of the 64-kDa species with antibody specific for the influenza virus hemagglutinin (HA) protein epitope in experiments where the gene encoding ICP0R was coexpressed with a gene encoding HA-tagged ubiquitin. Surprisingly, the 64-kDa form of ICP0R was found to be remarkably stable and persisted in infected cells for many hours after processing, despite the fact that ubiquitination normally functions as a means of tagging proteins for rapid degradation. Analyses of mutant polypeptides containing arginine substitutions at each of the lysine residues of ICP0R, which represent potential ubiquitin conjugation sites, revealed that a single lysine residue at codon 248 was both necessary and sufficient for the appearance of the 64-kDa processed form. However, a number of ICP0R mutants that retained the ubiquitination site at lysine 248 but contained disruptions of sequences at distant sites also lacked detectable 64-kDa protein, indicating that the integrity of the overall structure of ICP0R was an additional determinant for ubiquitination. © 1999 Academic Press

exhibit remarkably different properties when analyzed in transient expression assays. ICP0 behaves as a transactivator of gene expression (Everett, 1984; Gelman and Silverstein, 1985; Mosca et al., 1987; O’Hare and Hayward, 1985; Ostrove et al., 1987; Quinlan and Knipe, 1985; Sekulovich et al., 1988; Shapira et al., 1987) while ICP0R acts as a transrepressor (Weber et al., 1992; Weber and Wigdahl, 1992), but both function on a wide variety of viral and cellular promoters in a manner that is independent of specific promoter sequences. It has been postulated that the phenomenon of transrepression is the result of nonproductive interactions between ICP0R and a host cell factor that is required for transactivation by ICP0, so that ICP0R technically functions as a dominant negative mutant (Weber et al., 1992; Weber and Wigdahl, 1992). Consistent with results from transfection assays, recombinant HSV-1 carrying deletions of both copies of the ICP0 gene are impaired for both transcriptional activation of viral genes (Robert and Schaffer, 1997) and replication in low-multiplicity infections in cell culture (Sacks and Schaffer, 1987; Stow and Stow, 1986), in vivo infections (Cai et al., 1993; Clements and Stow, 1989; Leib et al., 1989), and in vitro latency models (Harris et al., 1989; Zhu et al., 1990). In contrast, a recombinant HSV-1 that was engineered to express high levels of the ICP0R protein instead of ICP0 was found to replicate no worse than an ICP0 null mutant (Spatz et al., 1997). This obser-

INTRODUCTION Infected cell polypeptide 0 (ICP0) and ICP0R are proteins encoded by a diploid gene present in the long repeat sequences of the herpes simplex virus type 1 (HSV-1) genome (Fig. 1). ICP0 is a 775-amino-acid protein that is encoded by all three exons of the gene (Perry et al., 1986), while ICP0R represents a truncated form of ICP0 that is created through alternative splicing of intron 2 (Everett et al., 1993; Weber et al., 1992). As a result of this differential splicing, ICP0R contains all 241 amino acids encoded by exons 1 and 2 of the ICP0 gene, but acquires a unique 21-amino-acid tail derived from unspliced intron 2 sequences instead of the 534 amino acids from exon 3 (Fig. 1). ICP0 is one of five immediate early polypeptides encoded by HSV-1 and is expressed at abundant levels throughout infection. In contrast, ICP0R accumulates to detectable levels only at later times in infection (Everett et al., 1993; Weber, Spatz, and Nordby, unpublished observations), since its synthesis depends on the failure to use splicing signals that are normally very efficiently utilized by the host cell splicing machinery (Perry et al., 1986). ICP0 and ICP0R also 1 To whom reprint requests and correspondence should be addressed. Fax: (734) 622–3318. E-mail: [email protected]. 2 Present address: Origin, Inc., East Lansing, MI 48909.

0042-6822/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

288

UBIQUITINATION OF HSV-1 ICP0R PROTEIN

289

FIG. 1. Expression of the ICP0 and ICP0R proteins in HSV-1. A map of the HSV-1 genome, shown at the top, identifies the long (L) and short (S) components; the inverted repeat sequences a, b, and c; and the location of the two copies of the ICP0 gene in the b sequences. The primary mRNA transcripts and the intron–exon structures of the ICP0 genes in wild-type HSV-1 and the HSV–KST mutant are shown below. In wild-type HSV-1, complete splicing of the primary transcript and subsequent translation of all three exons normally occur, resulting in high-level expression of the ICP0 protein. However, inefficient alternative splicing of intron 2 can also occur during infection, resulting in translation of the glycine-rich codons of the unspliced intron 2 sequences (“G”) instead of exon 3, and subsequent low-level expression of the ICP0R protein. In HSV–KST, splicing of the intron 2-encoded carboxy-terminal sequences of ICP0R is prevented by a 1.7-kb deletion encompassing the splice acceptor site of intron 2 and the entire coding sequences of exon 3 (striped box). This manipulation forces this mutant to express high levels of ICP0R instead of ICP0.

vation, coupled with the findings that a recombinant HSV-1 whose ICP0 genes lack introns and therefore can synthesize only ICP0 and not ICP0R replicated normally in cell culture and in vivo (Everett, 1991; Natarajan et al.,

1991), indicates that the role of the ICP0R protein in the downregulation of viral gene expression is unclear at this time. Despite its apparent role in transcriptional activation,

290

WEBER, SPATZ, AND NORDBY

the ICP0 protein differs from other HSV-1 transactivator proteins like ICP4 and VP16 in that it does not bind to specific DNA sequences, nor has it been found to interact with known components of the host cell transcriptional machinery. In contrast, the ability of ICP0 to associate with and disrupt the nuclear structure ND10 appears to play a key role in its ability to activate gene expression (Everett and Maul, 1994; Maul and Everett, 1994; Maul et al., 1993), as does the recruitment of herpesvirus-associated ubiquitin-specific protease (HAUSP) to this site through specific interactions with an exon 3-encoded domain of ICP0 (Everett et al., 1997). Recent evidence indicates that ND10 disruption events are linked to the ICP0-induced degradation of PIC1-modified isoforms of a specific ND10 component, the PML protein (Everett et al., 1998). ICP0 has also been shown to associate with the G1 cell cycle regulator cyclin D3 (Kawaguchi et al., 1997), although the significance of this observation is less clear. This interaction occurs through a domain that is encoded by exon 2, which contains a RING finger metal binding domain (Freemont et al., 1991; Lovering et al., 1993) and which includes the only sequences of ICP0 that are conserved among the alpha herpesvirus family (Spatz et al., 1996). Moreover, exon 2 has been shown in several studies to represent the most mutationally sensitive region of the ICP0 molecule (Cai and Schaffer, 1989; Chen et al., 1991; Everett, 1987; 1988; Spatz et al., 1996). As it lacks exon 3-encoded sequences (Fig. 1), ICP0R is unable to associate with HAUSP and fails to disrupt ND10 (Everett and Maul, 1994); however, it does contain the entire cyclin D3 binding domain in exon 2 (Kawaguchi et al., 1997). Since ICP0 promotes the degradation of PIC1-modified proteins (Everett et al., 1998), associates with the deubiquitinating enzyme HAUSP (Everett et al., 1997), and can interact with cyclin D3, a protein known to be ubiquitinated prior to its degradation during the cell cycle (Hershko, 1991; Murray, 1995), it seems likely that ICP0 function somehow involves interactions with ubiquitindependent degradation pathways. Ubiquitin is a highly conserved 76-amino-acid polypeptide that can be covalently linked to target proteins through a multienzyme pathway (Ciechanover, 1994; Hershko, 1991). In most proteins, ubiquitin conjugation involves the creation of a peptide bond between the carboxyl group of the carboxyterminal glycine residue of ubiquitin and the e-amino group of a lysine residue in the substrate protein. Branched polyubiquitination structures can then occur on the substrate protein through continued rounds of peptide bond formation between the carboxy-terminal glycine residue of a new ubiquitin molecule and the amino group of the lysine 48-residue of a ubiquitin that has already been conjugated to the substrate protein. Polyubiquitinated proteins are commonly recognized and degraded by a large multienzyme complex called the 26S proteasome, so that this form of processing typically

serves as a means of tagging proteins for proteolytic destruction. However, rare exceptions of ubiquitination serving as a form of functional posttranslational modification rather than a signal for protein recycling have been described, particularly in the histone family (Nickel and Davie, 1989; Thorne et al., 1987). In this study, evidence is presented that ICP0R of HSV-1 represents one of these unusual instances of a polypeptide that is modified by stable ubiquitination. RESULTS Identification of a processed form of ICP0R in HSV–KST-infected cells Since ICP0R is normally synthesized at low levels in wild-type HSV-1 infections through an inefficient alternative splicing mechanism, the recombinant virus HSV–KST was constructed in a previous study as a means of exploring the effects of overexpression of this protein on viral gene expression (Spatz et al., 1997). HSV–KST contains a deletion of the splice acceptor site of intron 2 and the entire exon 3 coding sequences from both genomic copies of the ICP0 gene; this prevents splicing of the intron 2 sequences from the primary transcript of ICP0 and forces this virus to express high levels of ICP0R instead of ICP0 (Fig. 1). Overexpression of the ICP0R protein in this manner was shown to have little negative effect on virus replication, since the growth properties of HSV–KST were identical to those of dl1403, a mutant virus that expressed neither ICP0 nor ICP0R (Spatz et al., 1997). Western blot analysis of HSV–KST-infected cell lysates using the monoclonal antibody 11060, which is specific for the exon 2-derived sequences of both ICP0 and ICP0R (Everett et al., 1993), revealed the presence of an abundant 41-kDa polypeptide whose molecular weight was consistent with that predicted for ICP0R (Fig. 2B). However, a less abundant 64-kDa protein was also unexpectedly observed in these lysates (Fig. 2B). Evidence that this 64-kDa species was of viral origin and specific for HSV–KST infections came from the observation that it was absent from lysates of either mock- or dl1403-infected cells (Fig. 2B). Moreover, the demonstration that both the 41- and 64-kDa proteins could be detected in lysates from cells transfected with pKST, a plasmid that expresses the ICP0R gene (Fig. 2A), confirmed that the 64-kDa species was related to ICP0R and could be synthesized in the complete absence of other viral genes (Fig. 2B). One obvious mechanism to account for the existence of the 64-kDa protein would be a posttranslational processing event that occurs in the 41-kDa ICP0R protein; however, the structure of the ICP0R gene allows for other possible explanations. Since the ICP0R gene contains the large first intron of the ICP0 gene (Fig. 1), it is possible that a previously unidenti-

UBIQUITINATION OF HSV-1 ICP0R PROTEIN

291

FIG. 2. Detection of ICP0R and a 64-kDa processed form of ICP0R in both infected and transfected cells. (A) Maps of ICP0R genes present in the HSV-1 recombinant HSV–KST and the plasmids pKST and pKST-C1. Although their coding sequences are identical, these genes can be distinguished by the presence or absence of intron 1 sequences as well as the utilization of polyadenylation signals derived from either the ICP0 gene or the SV40 early transcription unit. (B) Western blot analysis of infected and transfected cells. Vero cells were mock infected, infected with either HSV–KST or the ICP0/ICP0R null mutant virus dl1403, or transfected with either pKST or pKST-C1. Extracts were then prepared from these cells at 6 h postinfection or 48 h posttransfection and analyzed in Western blots using the ICP0 exon 2-specific monoclonal antibody 11060 (Everett et al., 1993). A molecular weight scale is included at the left, and the native and processed forms of ICP0R (41 and 64 kDa, respectively) are indicated by arrows at the right.

fied exon could be spliced onto exon 2 instead of exon 1, or that fortuitous promoter elements mapping within the intron could initiate transcription of sequences within the intron that would continue into exon 2. Either of these scenarios could generate a protein that would still be recognized by the 11060 antibody that was used in this study, since the epitope that it recognizes has been mapped to exon 2 (Everett et al., 1993). To address these possibilities, pKST-C1, a derivative that lacks the intron sequences but retains all of the other protein-coding sequences of the ICP0R gene (Fig. 2A), was constructed. Western blot analysis of lysates of cells transfected with pKST-C1 revealed the presence of both the 41- and 64-kDa species (Fig. 2B), indicating that the latter protein was not derived through the use of alternative coding sequences within the ICP0 gene. The significant reduction in the levels of both proteins in lysates of pKST-C1transfected cells, compared with pKST-transfected cells, was most likely due to a debilitating effect of the loss of the intron sequences on stable expression of the transfected ICP0R gene, as has been described previously for ICP0 (Everett, 1991). These findings, coupled with the results of pulse–chase labeling experiments described below, demonstrate that the 64kDa species arises from posttranslational modification of the 41-kDa ICP0R protein.

Identification of the 64-kDa protein as a ubiquitinated form of ICP0R The inability to demonstrate that the 64-kDa protein was derived through a common means of posttranslational modification such as phosphorylation, glycosylation, or poly(ADP)-ribosylation (Weber, Spatz, and Nordby, unpublished observations) indicated that it was generated by an unusual processing mechanism. The apparent 23-kDa molecular weight increase in ICP0R after processing was consistent with a monoubiquitination event, given the molecular weight of ubiquitin (10 kDa) and the compounding effect that conjugation of a ubiquitin moity to an internal lysine residue of ICP0R would have on electrophoretic mobility. However, all attempts to demonstrate a reactivity of the 64-kDa protein with antiubiquitin sera that were generated in either mice or rabbits failed (Weber, Spatz, and Nordby, unpublished observations). Since the highly conserved nature of ubiquitin makes it only weakly immunogenic in animals, it was likely that the negative results obtained could be explained by the limited sensitivity of these antisera. Thus, an alternative approach had to be employed to demonstrate ubiquitination of the ICP0R protein. This involved the use of the plasmid pMT123 (Treier et al., 1994), which encodes a polyubiquitin precursor whose subunits are each tagged with the influenza virus hemagglutinin (HA) protein epitope. When transfected into

292

WEBER, SPATZ, AND NORDBY

FIG. 3. Detection of ubiquitinated forms of the ICP0R protein. Extracts of Vero cells that had been transfected with the plasmids indicated in the grid at the top were prepared at 48 h transfection. ICP0R protein was then immunoprecipitated from these extracts and analyzed in Western blots using either the ICP0 exon 2-specific monoclonal antibody 11060 (left) or an antibody specific for the influenza virus hemagglutinin (HA) protein epitope that is tagged onto the polyubiquitin protein encoded by pMT123 (Treier et al., 1994) (right). A molecular weight scale is included at the right, and the identities of unprocessed ICP0R (ICP0R), ICP0R that has been conjugated to endogenous ubiquitin (ICP0R-Ub), and ICP0R that has been conjugated to HA-tagged ubiquitin (ICP0R-HA-Ub) are indicated by arrows at the left. The band corresponding to the last species is also identified by a black dot in either panel.

cells, the polyprotein synthesized from this construct is processed by available cellular ubiquitin-specific proteases into HA-tagged ubiquitin molecules which can then be conjugated onto a protein expressed from a cotransfected gene. In experiments that used the HA-tagged ubiquitinexpressing construct, Vero cells were transfected with either the ICP0R-expressing plasmid pKST alone, a mixture of pMT123 and pKST, or a mixture of pMT123 and pKST-F1I. The pKST-F1I construct encodes an ICP0R mutant that is defective for processing to the 64-kDa form and therefore serves as a negative control in this experiment; this class of ICP0R mutants is described in more detail below. ICP0R protein was immunoprecipitated from the transfected cell lysates using the 11060 antibody and then analyzed on Western blots which employed either the 11060 antibody or an antibody specific for the HA epitope. In either transfection employing pKST, both the 41- and 64-kDa forms of ICP0R were immunoprecipitated; however, in the pKST transfection in which pMT123 was also present, an additional 66-kDa form of ICP0R was also isolated (Fig. 3). Since the amino acid sequences corresponding to the HA epitope were pre-

dicted to increase the molecular weight of ubiquitin by 2 kDa (Treier et al., 1994), an increase in the molecular weight of the 64-kDa form of ICP0R to 66 kDa would be expected if ubiquitination was indeed responsible for the processing event and HA-tagged ubiquitin was now being used as a ligand instead of endogenous ubiquitin. Furthermore, both the 64- and 66-kDa forms were absent from the immunoprecipitate derived from transfections of pMT123 and pKST-F1I (Fig. 3), indicating that synthesis of the 66-kDa species had the same requirement for normal processing of the ICP0R protein as the 64-kDa protein. Finally, Western blot analysis of the three immunoprecipitates using an antibody specific for the HA epitope confirmed that the 66-kDa species which was unique to the transfection employing pMT123 and pKST did contain HA-tagged ubiquitin (Fig. 3), thereby confirming that processing of ICP0R by ubiquitination does occur. Interestingly, low levels of a 72-kDa form could be detected in addition to the 66 kDa form in this analysis (Fig. 3); this species likely represents an ICP0R derivative that has undergone a second ubiquitination event. Determination of the kinetics of ICP0R ubiquitination during HSV–KST infection Since ubiquitination normally functions as a means of tagging proteins for rapid degradation, yet significant levels of ubiquitinated ICP0R accumulate in both transfected and infected cells (Fig. 2), it was of interest to determine the rates of appearance and disappearance of this protein species. The kinetics of 64-kDa protein synthesis were initially examined in pulse–chase experiments employing immunoprecipitation of ICP0R from 35 S-labeled HSV–KST-infected cells at 5 h postinfection. The results of these studies indicated that ubiquitination of ICP0R initiated within the first few minutes after the protein was synthesized, and that one-half of the protein could be converted to the 64-kDa form after a 1-h chase (Fig. 4A). However, little additional processing could be detected when the length of the chase was extended past 1 h, which corresponded to 6 h postinfection and thereafter (Fig. 4A). Remarkably, the ubiquinated ICP0R was found to accumulate as a stable species and did not undergo any detectable degradation or polyubiquitination events, even after chase incubations as long as 17 h (Fig. 4A). These findings indicated that ICP0R represents a rare instance of a protein that undergoes ubiquitination with no apparent loss in stability. The kinetics of 64-kDa protein synthesis were also examined by assessing the accumulation of ICP0R protein in HSV–KST-infected Vero cells over time. Lysates were prepared from identically infected cells at 3-h increments postinfection and analyzed on Western blots (Fig. 4B). At 3 h postinfection, the majority of the ICP0R protein existed as the 64-kDa ubiquitinated form, but by 6 h postinfection this trend had reversed and a majority

UBIQUITINATION OF HSV-1 ICP0R PROTEIN

293

FIG. 4. Kinetics of ubiquitination of the ICP0R protein during HSV–KST infection. (A) Pulse–chase labeling of ICP0R protein. HSV–KST-infected Vero cells were pulse labeled for 5 min at 5 h postinfection and subjected to chase incubations for the lengths of time indicated at the top. ICP0R was then immunoprecipitated from extracts of these cells, electrophoresed on polyacrylamide gels, and analyzed by autoradiography. The unprocessed and ubiquitinated forms of ICP0R are indicated at the right. (B) Stable accumulation of the ICP0R protein. Extracts were prepared of Vero cells that had been infected with HSV–KST for the lengths of time indicated at the top and analyzed in Western blots using the ICP0 exon 2-specific monoclonal antibody 11060. The unprocessed and ubiquitinated forms of ICP0R are indicated at the right.

of the ICP0R protein remained unprocessed. By 9 h postinfection, the level of 64-kDa protein remained virtually unchanged while the level of the unprocessed protein continued to increase through 24 h postinfection (Fig. 4B). Interestingly, a portion of the accumulated 64kDa protein underwent an increase in electrophoretic mobility starting at 12 h postinfection and continuing through 24 h postinfection; this was found to be due to an apparent dephosphorylation event that had occurred within the 64-kDa protein that had already accumulated (Weber, Spatz, and Nordby, unpublished observations). The source of this dephosphorylation is unclear, but appears to be specific for ICP0R synthesized in HSV–KST infections because it was not observed in pKST transfections (Fig. 2). Together, the results of the pulse–chase and time course experiments demonstrate that the 64-kDa ubiquitinated form of ICP0R is a highly stable species and that most of the 64-kDa protein that will accumulate during an infection has already been generated by 6 h postinfection. These findings indicate that although the ICP0R protein is synthesized throughout infection by HSV–KST, its modification by ubiquitination is limited to the immediate early and early stages of the replication cycle. Nevertheless, the ubiquitinated form of ICP0R can be detected well into the late stages of infection by virtue of its unexpectedly high level of stability. Analysis of the amino acid sequences required for ubiquitination of ICP0R As further confirmation that the 64-kDa species was derived from ubiquitination of the ICP0R protein, an attempt was made to identify the specific lysine residue at

which the ubiquitin moiety was conjugated. Three lysine residues exist within the ICP0R polypeptide, each of which is located within an interesting region of the molecule. The lysine residues at codons 144 and 159 map within the RING finger, which represents the most conserved and mutationally sensitive sequences of the ICP0R protein (Spatz et al., 1996). In contrast, the lysine residue at codon 248 maps within a glycine-rich region derived from the unspliced intron 2 sequences, and therefore lies within the only segment of the ICP0R protein that is not shared with ICP0. Site-directed mutagenesis was carried out on the ICP0R gene which converted each of the three lysine residues at codons 144, 159, and 248 to arginine residues. The use of lysine-to-arginine substitutions representing conserved alterations in the primary amino acid sequence of ICP0R allowed the role of each lysine residue in ubiquitination of ICP0R to be addressed with minimum disruption to the overall protein structure. Lysates were prepared of Vero cells that had been transfected with each of the lysine-to-arginine substitution mutants of ICP0R and analyzed in Western blots using the 11060 antibody (Fig. 5). The lysine residues at positions 144 and 159 were dispensable for ubiquitination either individually or in combination, since the 64kDa processed form of ICP0R was readily detected in cells transfected with the pKST-K144R, pKST-K159R, and pKST-K144R/K159R mutants (Fig. 5). The observation that processing still occurred in the K144R/K159R double mutant, which contained only the lysine at codon 248, indicated that the latter residue represented the actual site of ubiquitin conjugation in the ICP0R protein. This was confirmed in the analysis of cells transfected with either

294

WEBER, SPATZ, AND NORDBY

DISCUSSION

FIG. 5. Identification of the specific lysine residue in the ICP0R protein at which ubiquitin conjugation occurs. Vero cells were transfected with pKST derivatives containing the mutations indicated in the grid at the bottom. Extracts were then prepared from these cells at 48 h posttransfection and analyzed in Western blots using the ICP0 exon 2-specific monoclonal antibody 11060. The unprocessed and ubiquitinated forms of ICP0R are indicated by arrows at left. Lane 1, pKST; lane 2, pKST–K144R; lane 3, pKST–K159R; lane 4, pKST–K144R/K159R; lane 5, pKST–K144R/K159R/K248R; lane 6, pKST–K248R; lane 7, pKST–F1I.

the pKST-K248R or pKST-K144R/K159R/K248R mutants, both of which lacked the lysine residue at position 248 and both of which failed to generate detectable levels of the 64-kDa form of ICP0R (Fig. 5). Taken together, the results of this mutational analysis demonstrated that the single lysine residue at codon 248 was both necessary and sufficient for ubiquitination of the ICP0R protein. Although this analysis identified the actual site of ubiquitin conjugation in the ICP0R polypeptide, other sequences within the protein may also be required for processing to occur. To explore this possibility, a large panel of ICP0R mutants that had been used to map the amino acid sequences required for transrepression in a previous study (Spatz et al., 1996) were examined by Western blot analyses for the presence or absence of the 64-kDa processed form. Without exception, the entire panel of mutants could be divided into two phenotypic classes: one that was transrepression competent and underwent ubiquitination to produce the 64-kDa form, and one that was transrepression incompetent and failed to produce detectable levels of the 64-kDa form. This is clearly illustrated in a set of 10 linker insertion mutants of ICP0R, where the R1I, E25I, E3I, and A2I mutants fall into the former group, while the F1I, R2I, E8I, E13I, R3I, and E32I mutants fall into the latter group (Fig. 6). Comparable groupings could be obtained in the analyses of additional sets of internal deletion and truncation mutants of ICP0R (Weber, Spatz, and Nordby, unpublished observations). The finding that so many mutants that retained the lysine residue at codon 248 failed to generate the 64-kDa protein was consistent with a requirement for additional sequences apart from the site of ubiquitin conjugation for detectable levels of ICP0R processing to occur.

Posttranslational modification of the HSV-1 ICP0R protein by ubiquitination was demonstrated in this study by the detection of a conjugated epitope-tagged ubiquitin moiety on processed ICP0R (Fig. 3) and the identification of the lysine residue at codon 248 as the actual site of ubiquitin conjugation in ICP0R (Fig. 5). While ubiquitination is hardly an unusual protein processing event, as polyubiquitination represents the signal most frequently used by cells in tagging polypeptides for recycling in proteolytic degradation pathways, proteins like ICP0R which undergo predominantly monoubiquitination events (Fig. 3) and which remain stable over extended periods (Fig. 4) are relatively rare. It should be noted that stable ubiquitination has also been described in one other HSV-1 protein to date, the product of the US9 open reading frame (Brandimarti and Roizman, 1997). However, unlike ICP0R, the UL9 gene product is a lysineless protein that appears to be modified through an alternate pathway of ubiquitin conjugation and exists primarily as a group of polyubiquitinated proteins. The lysine 248 residue which serves as the site of ubiquitin conjugation in the ICP0R protein maps near the carboxy-terminal end of this 262-amino-acid polypeptide. Since this residue is encoded by the unspliced sequences of intron 2 of the ICP0 gene, it is absent from the ICP0 protein; this explained the inability to detect any stable ubiquitinated forms of ICP0 in experiments that employed the pMT123 construct expressing HA-tagged ubiquitin (Weber, Spatz, and Nordby, unpublished observations). Interestingly, the only other proteins for which stable ubiquitination has been described and for which the site of ubiquitin conjugation has been determined also undergo processing near their carboxy-terminal ends. In the 129-amino-acid histone H2A protein, lysine 119 is used for ubiquitin conjugation, while in the 125-aminoacid histone H2B protein, lysine 120 is used (Nickel and Davie, 1989; Thorne et al., 1987) (Fig. 7). The latter proteins are remarkable in that they both contain multiple lysine residues within their carboxy-terminal sequences, yet ubiquitin is conjugated onto only a specific single lysine residue in either polypeptide. Although it possesses only a single lysine residue within its carboxyterminal sequences, a similar level of selectivity appears to exist in ICP0R, since transfer of the lysine at codon 248 to the final codon of the protein results in the loss of detectable ubiquitination (Weber, Spatz, and Nordby, unpublished observations). Together, these results indicate that the stable ubiquitination of these three proteins requires not only the lysine residue that is being processed, but additional flanking amino acid sequences as well. However, there is no obvious sequence homology between the ubiquitination sites of the ICP0R, histone H2A, and histone H2B proteins (Fig. 7). Moreover, the secondary structures of these sequences are likely to be

UBIQUITINATION OF HSV-1 ICP0R PROTEIN

295

FIG. 6. Relationship between ubiquitination and transrepression competence in mutant derivatives of ICP0R. Vero cells were transfected with pKST derivatives containing linker insertion mutations that resulted in four or five amino acid insertions within the ICP0R protein. The designations of these linker insertions as well as the sites at which the amino acid insertions occur within the ICP0R protein are indicated at the top and bottom, respectively. Extracts were prepared from these cells at 48 h posttransfection and analyzed in Western blots using the ICP0 exon 2-specific monoclonal antibody 11060. The unprocessed and ubiquitinated forms of ICP0R are indicated by arrows at the left. The ability of each mutant to mediate transrepression as determined in a previous study (Spatz et al., 1996) is indicated at the bottom (1, .75% of the transrepression observed with wild-type ICP0R; 2, ,30% of the transrepression observed with wild-type ICP0R).

dissimilar as well, since the carboxy-terminal ends of the histone H2A and histone H2B proteins are known to be randomly structured and a-helical, respectively (Luger et al., 1997), while polyglycine tracts similar to the one in the carboxy-terminal end of ICP0R typically form b structures (Munoz et al., 1983). Thus, the structural features that envelop an individual lysine residue within a protein and allow it to be stably ubiquitinated remain obscure. In addition to the contribution of amino acids within the immediate vicinity of lysine 248 to ICP0R ubiquitination, more distant regions of the polypeptide also appeared to play an important role in mediating this processing event. ICP0R derivatives containing mutations that mapped over a broad region of the molecule were found in many cases to be defective for ubiquitination, despite the fact that they still retained the lysine 248 residue (Fig. 6). Since this amino acid maps within a glycine-rich region at the extreme carboxy-terminal region of ICP0R (Fig. 7) which is predicted to form a stable b structure (Munoz et al., 1983), and many of the mutations responsible for the processing-deficient phenotype map far upstream, it is unlikely that the ubiquitination site of the

FIG. 7. Comparison of the sites of ubiquitination in ICP0R with histones H2A and H2B. The amino acid sequences that flank the lysine residue to which ubiquitin is conjugated (black “K” residue) in ICP0R, histone H2A (Nickel and Davie, 1989), and histone H2B (Thorne et al., 1987) are shown. The codon number of the first residue listed is designated at the left; the final residue listed represents the carboxyterminal codon for each of the three proteins.

protein could have been conformationally sequestered in all of these mutants. The simplest interpretation of these results is that ubiquitination of ICP0R requires not only the lysine residue at codon 248, but also the integrity of a large part of the overall structure of the protein. This would explain the remarkable association between transrepression competence and ubiquitination that was observed in this panel of mutants (Fig. 6), since any disruption in ICP0R structure that was significant enough to abolish transrepression capability also appeared to prevent recognition by host cell ubiquitination enzymes. However, it cannot be ruled out that the ubiquitination event is actually a consequence of the unknown molecular interactions that result in ICP0R-mediated transrepression. It was noted in a previous study (Spatz et al., 1996) and was again apparent in this work (Fig. 6) that transrepression-deficient mutants of ICP0R accumulate to significantly higher levels than the wild-type protein. While the most obvious explanation of this result is that these mutants simply fail to autorepress their own synthesis, other evidence has suggested that this is more likely due to an increase in the stability of the mutant proteins (Spatz et al., 1996). The inability of these same transrepression-deficient derivatives of ICP0R to undergo ubiquitination (Fig. 6) would seem to be at least partially responsible for any increase in polypeptide stability. However, the results of this study indicate that the ubiquitination event that occurs at lysine 248 surprisingly contributes little if anything to ICP0R turnover. This was clearly demonstrated in the pKST-K144R/K159R/K248R mutant, which lacked all three lysine residues and therefore all potential sites for initiation of ubiquitin-dependent degradation. As expected, this mutant failed to undergo detectable ubiquitination, yet surprisingly still possessed

296

WEBER, SPATZ, AND NORDBY

the low levels of accumulation associated with the wildtype protein (Fig. 5). This was in marked contrast to the transrepression-deficient pKST-F1I mutant, which also failed to undergo ubiquitination (Fig. 3) but which accumulated to very high levels in transfected cells (Figs. 5 and 6) (Spatz et al., 1996). Moreover, the 64-kDa form of ICP0R did not appear to serve as a precursor to further polyubiquitination events and therefore an intermediate in the degradation of the protein in HSV–KST-infected cells (Fig. 4). Although these results would seem to suggest that the instability of the ICP0R protein may be somehow mediated by the molecular interactions that result in transrepression, there is no direct evidence for this at the present time. However, it is clear that there appears to be an unusual separation of protein instability from ubiquitination in this polypeptide. Despite the ability of ICP0R to mediate transrepression of viral promoters in transfection assays, the absence of a phenotype in the ICP0R-overexpressing virus HSV–KST (Spatz et al., 1997) indicates that the contribution of this protein to the regulation of viral gene expression has yet to be demonstrated. The role of ubiquitination in ICP0R function is even less clear. Each of the lysine residue mutants that were generated in this study, including the lysineless pKST-K144R/K159R/K248R mutant (Fig. 5), were fully able to mediate transrepression events when tested in transfection assays (Weber, Spatz, and Nordby, unpublished observations). This indicated that the ubiquitination of the ICP0R protein and the 64kDa form of ICP0R were each dispensable for transrepression. Moreover, in wild-type HSV-1, ICP0R accumulates to levels that are readily detectable only at later times in infection (Everett et al., 1993; Weber, Spatz, and Nordby, unpublished observations). Since the analysis of the kinetics of ICP0R processing during HSV–KST infection in this study (Fig. 4) indicates that modification by ubiquitination is limited to the immediate early and early stages of the replication cycle, it is likely that only a small portion of the ICP0R synthesized during a wild-type virus infection becomes ubiquitinated. This is consistent with the inability to detect ubiquitinated ICP0R in lysates of wild-type virus-infected cells using the reagents that are currently available at this time (Weber, Spatz, and Nordby, unpublished observations). Further studies are therefore necessary to define the functions of both ICP0R and its unusual form of processing in the HSV-1 life cycle. MATERIALS AND METHODS Cells and viruses Vero (African green monkey kidney) cells were used in all experiments and were maintained in Dulbecco’s minimum essential medium (GIBCO-BRL) supplemented with 10% defined fetal bovine serum, HEPES buffer, penicillin, streptomycin, and glutamine. HSV-1 (strain 17) was

the wild type HSV-1 strain employed in all experiments. HSV–KST is a derivative of HSV-1 (strain 17) that has been engineered to express high levels of the ICP0R protein instead of ICP0; its construction and characterization are described elsewhere (Spatz et al., 1997). dl1403 is an HSV-1 (strain 17) mutant that expresses neither ICP0 nor ICP0R (Stow and Stow, 1986). Both HSV-1 (strain 17) and dl1403 were generously provided by R. Everett (MRC Virology Unit, Glasgow, U.K.). Plasmid constructions pKST contains the gene encoding the wild-type ICP0R protein (Weber et al., 1992). pKST-C1 contains a gene encoding the wild-type ICP0R protein from which the intron 1 sequences have been removed. It was constructed by replacing the 1.4-kb intron 1-containing NcoI/ KpnI fragment of the ICP0R gene in pKST with the 0.6-kb NcoI/KpnI fragment of the intronless ICP0 gene of p111-C1 (Everett, 1991); the latter plasmid was generously provided by R. Everett (MRC Virology Unit, Glasgow, U.K.). pKST-R1I, pKST-E25I, pKST-E3I, pKST-A2I, pKST-F1I, pKST-R2I, pKST-E8I, pKST-E13I, pKST-R3I, and pKST-E32I are derivatives of pKST which contain linker insertion mutations that result in four or five amino acid insertions within the ICP0R protein; their construction and characterization have been described previously (Spatz et al., 1996). pMT123 expresses a polyubiquitin precursor that has been tagged with an epitope of the influenza virus HA protein (Treier et al., 1994); it was generously provided by D. Bohmann (EMBL, Heidelberg, Germany). Site-directed mutagenesis to replace each of the lysine residues of the ICP0R protein with arginine residues was carried out using the QuickChange system (Stratagene) and PAGE-purified oligonucleotide primers (GIBCO-BRL). The K144R mutation was generated with the following complementary oligonucleotides, which also created unique NsiI and FspI sites in the ICP0R gene: 59-CCGCTTCTGCATCCCATGCATGCGCACCTGGATGC-39 and 59-GCATCCAGGTGCGCATGCATGGGATGCAGAAGCGG-39. The K159R mutation was generated with the following complementary oligonucleotides, which also created a unique BssHII site in the ICP0R gene: 59-CCGCTGTGCAACGCGCGCCTGGTGTACCTGATAGTGGG-39 and 59-CCCACTATCAGGTACACCAGGCGCGCGTTGCACAGCGG-39. The K248R mutation was generated with the following complementary oligonucleotides, which also created a unique NruI site in the ICP0R gene: 59-GGCGGGGGGCGGTCGCGACCCTGGGGGAGG39 and 59-CCTCCCCCAGGGTCGCGACCGCCCCCCGCC39. The K144R and K159R mutations were introduced individually or in combination into a subcloned 0.3-kb XhoI/KpnI fragment from pKST, while the K248R mutation was introduced into a subcloned 0.2-kb KpnI/XbaI fragment from pKST. The predicted mutations in these sub-

UBIQUITINATION OF HSV-1 ICP0R PROTEIN

fragments were then verified by dye terminator DNA cycle sequencing (ABI PRISM) before being used to replace the 0.3-kb XhoI/KpnI or 0.2-kb KpnI/XbaI fragments of wild-type pKST. Transfection, Western blot, and immunoprecipitation analyses. Transfection of pKST or its derivatives into Vero cells for the purpose of Western blot analysis of the ICP0R protein was carried out using 2 3 105 cells seeded in six-well plates, 4 mg of plasmid DNA, 12 ml of Pfx2 liposome (Invitrogen), and a 48-h incubation period according to the manufacturer’s recommendations. Lysates were prepared from transfected or infected cells and analyzed on Western blots using procedures described previously (Spatz et al., 1996). Primary mouse monoclonal antibodies used in the Western blots included the 11060 antibody specific for the exon 2 sequences of ICP0 and ICP0R (Everett et al., 1993), which was generously provided by R. Everett (MRC Virology Unit, Glasgow, U.K.), and an anti-influenza virus HA protein antibody conjugated to peroxidase (Boehringer-Mannheim). Immunoprecipitations of 35S-labeled or unlabeled ICP0R proteins were carried out using the 11060 antibody and the procedure of Meredith et al. (1995). Pulse–chase labeling of HSV–KST-infected cell proteins was performed at 5 h postinfection using a 5-min pulse incubation with 35Strans-labeled amino acids (ICN) followed by variablelength chase incubations and lysis in immunoprecipitation buffer on ice. ACKNOWLEDGMENTS We thank R. Everett for plasmids, viruses, and the generous gift of 11060 antibody; D. Bohmann for plasmids; and M. Vartanian, C. Jacobs, and M. Tumilo for advice and assistance in automated DNA sequencing.

REFERENCES Brandimarti, R., and Roizman, B. (1997). U(s)9, a stable lysine-less herpes simplex virus 1 protein, is ubiquitinated before packaging into virions and associates with proteasomes. Proc. Natl. Acad. Sci. USA 94, 13973–13978. Cai, W., and Schaffer, P. A. (1989). Herpes simplex virus type 1 ICP0 plays a critical role in the de novo synthesis of infectious virus following transfection of viral DNA. J. Virol. 63, 4579–4589. Cai, W. H., Astor, T. L., Liptak, L. M., Cho, C., Coen, D. M., and Schaffer, P. A. (1993). The herpes simplex virus type-1 regulatory protein ICP0 enhances virus replication during acute infection and reactivation from latency. J. Virol. 67, 7501–7512. Chen, J., Zhu, X., and Silverstein, S. (1991). Mutational analysis of the sequence encoding ICP0 from herpes simplex virus type 1. Virology 180, 207–220. Ciechanover, A. (1994). The ubiquitin–proteasome proteolytic pathway. Cell 79, 13–21. Clements, G. B., and Stow, N. D. (1989). A herpes simplex virus type 1 mutant containing a deletion within immediate early gene 1 is latency-competent in mice. J. Gen. Virol. 70, 2501–2506. Everett, R. D. (1984). Transactivation of transcription by herpes virus

297

products: Requirement for two HSV-1 immediate early polypeptides for maximum activity. EMBO J. 3, 3135–3141. Everett, R. D. (1987). A detailed mutational analysis of Vmw 110, a trans-acting transcriptional activator encoded by herpes simplex virus type 1. EMBO J. 6, 2069–2076. Everett, R. D. (1988). Analysis of functional domains of herpes simplex virus type 1 immediate early polypeptide Vmw 110. J. Mol. Biol. 202, 87–96. Everett, R. D. (1991). Construction and characterization of herpes simplex virus type 1 viruses without introns in immediate early gene 1. J. Gen. Virol. 72, 651–659. Everett, R. D., and Maul, G. G. (1994). HSV-1 IE protein Vmw 110 causes redistribution of PML. EMBO J. 13, 5062–5069. Everett, R. D., Cross, A., and Orr, A. (1993). A truncated form of herpes simplex virus type-1 immediate-early protein Vmw 110 is expressed in a cell type dependent manner. Virology 197, 751–756. Everett, R. D., Freemont, P., Saitoh, H., Dasso, M., Orr, A., Kathoria, M., and Parkinson, J. (1998). The disruption of ND10 during herpes simplex virus infection correlates with the Vmw 110- and proteasome-dependent loss of several PML isoforms. J. Virol. 72, 6581– 6591. Everett, R. D., Meredith, M., Orr, A., Cross, A., Kathoria, M., and Parkinson, J. (1997). A novel ubiquitin-specific protease is dynamically associated with the PML nuclear domain and binds to a herpesvirus regulatory protein. EMBO J. 16, 566–577. Freemont, P. S., Hanson, I. M., and Trowsdale, J. (1991). A novel cysteine-rich sequence motif. Cell 64, 483–484. Gelman, I. H., and Silverstein, S. (1985). Identification of immediate early genes from herpes simplex virus that transactivate the virus thymidine kinase gene. Proc. Natl. Acad. Sci. USA 82, 5265–5269. Harris, R. A., Everett, R. D., Zhu, X., Silverstein, S., and Preston, C. M. (1989). Herpes simplex virus type 1 immediate early protein Vmw 110 reactivates latent herpes simplex virus type 2 in an in vitro latency system. J. Virol. 63, 3513–3515. Hershko, A. (1991). The ubiquitin pathway for protein degradation. Trends Biochem. Sci. 16, 265–268. Kawaguchi, Y., VanSant, C., and Roizman, B. (1997). Herpes simplex virus 1 alpha regulatory protein ICP0 interacts with and stabilizes the cell cycle regulator cyclin D3. J. Virol. 71, 7328–7336. Leib, D. A., Coen, D. M., Bogard, C. L., Hicks, K. A., Yager, D. R., Knipe, D. M., Tyler, K. L., and Schaffer, P. A. (1989). Immediate early regulatory gene mutants define different stages in the establishment and reactivation of herpes simplex virus latency. J. Virol. 63, 759–768. Lovering, R., Hanson, I. M., Borden, K. L. B., Martin, S., Oreilly, N. J., Evan, G. I., Rahman, D., Pappin, D. J. C., Trowsdale, J., and Freemont, P. S. (1993). Identification and preliminary characterization of a protein motif related to the zinc finger. Proc. Natl. Acad. Sci. USA 90, 2112–2116. Luger, K., Mader, A. W., Richmond, R. K., Sargent, D. F., and Richmond, T. J. (1997). Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389, 251–260. Maul, G. G., and Everett, R. D. (1994). the nuclear location of PML, a cellular member of the C3HC4 zinc-binding domain protein family, is rearranged during herpes simplex virus infection by the C3HC4 viral protein ICP0. J. Gen. Virol. 75, 1223–1233. Maul, G. G., Guldner, H. H., and Spivack, J. G. (1993). Modification of discrete nuclear domains induced by herpes simplex virus type-1 immediate early gene-1 product (ICP0). J. Gen. Virol. 74, 2679–2690. Meredith, M., Orr, A., Elliott, M., and Everett, R. (1995). Separation of sequence requirements for HSV-1 Vmw 110 multimerisation and interaction with a 135-kDa cellular protein. Virology 209, 174–187. Mosca, J. D., Bednarik, D. P., Raj, N. B. K., Rosen, C. A., Sodroski, J. G., Haseltine, W. A., Hayward, G. S., and Pitha, P. M. (1987). Activation of human immunodeficiency virus by herpesvirus infection: Identification of a region within the long terminal repeat that responds to a trans-acting factor encoded by herpes simplex virus 1. Proc. Natl. Acad. Sci. USA 84, 7408–7412.

298

WEBER, SPATZ, AND NORDBY

Munoz, G. S., Puiggali, J., Rodriguez, A., and Subirana, J. A. (1983). Crystals of polyglycine in the beta form. J. Mol. Biol. 167, 223–225. Murray, A. (1995). Cyclin ubiquitination: The destructive end of mitosis. Cell 81, 149–152. Natarajan, R., Deshmane, S., Valyi-Nagy, T., Everett, R., and Fraser, N. W. (1991). A herpes simplex virus type 1 mutant lacking the ICP0 introns reactivates with normal efficiency. J. Virol. 65, 5569–5573. Nickel, B. E., and Davie, J. R. (1989). Structure of polyubiquitinated histone H2A. Biochemistry 28, 964–968. O’Hare, P., and Hayward, G. S. (1985). Evidence for a direct role for both the 175,000- and 110,000-molecular weight immediate early proteins of herpes simplex virus in the transactivation of delayed early promoters. J. Virol. 53, 751–760. Ostrove, J. M., Leonard, J., Weck, K. E., Rabson, A. B., and Gendelman, H. E. (1987). Activation of the human immunodeficiency virus by herpes simplex virus type 1. J. Virol. 61, 3726–3732. Perry, L. J., Rixon, F. J., Everett, R. D., Frame, M. C., and McGeoch, D. J. (1986). Characterization of the IE110 gene of herpes simplex virus type 1. J. Gen. Virol. 67, 2365–2380. Quinlan, M. P., and Knipe, D. M. (1985). Stimulation of expression of a herpes simplex virus DNA-binding protein by two viral functions. Mol. Cell. Biol. 5, 957–963. Robert, J. D., and Schaffer, P. A. (1997). Activation of gene expression by herpes simplex virus type 1 ICP0 occurs at the level of mRNA synthesis. J. Virol. 71, 6850–6862. Sacks, W. R., and Schaffer, P. A. (1987). Deletion mutants in the gene encoding the herpes simplex virus type 1 immediate early protein ICP0 exhibit impaired growth in cell culture. J. Virol. 61, 829–839. Sekulovich, R. E., Leary, K., and Sandri-Goldin, R. M. (1988). The herpes simplex virus type 1 alpha protein ICP27 can act as a trans-repressor or trans-activator in combination with ICP4 and ICP0. J. Virol. 62, 4510–4522.

Shapira, M., Homa, F. L., Glorioso, J. C., and Levine, M. (1987). Regulation of the herpes simplex virus type 1 late (gamma2) glycoprotein C gene: Sequences between base pairs 234 to 129 control transient expression and responsiveness to transactivation by the products of the immediate early (alpha) 4 and 0 genes. Nucleic Acids Res. 15, 3097–3111. Spatz, S. J., Nordby, E. C., and Weber, P. C. (1996). Mutational analysis of ICP0R, a transrepressor protein created by alternative splicing of the ICP0 gene of herpes simplex virus type 1. J. Virol. 70, 7360–7370. Spatz, S. J., Nordby, E. C., and Weber, P. C. (1997). Construction and characterization of a recombinant herpes simplex virus type 1 which overexpresses the transrepressor protein ICP0R. Virology 228, 218– 228. Stow, N. D., and Stow, E. C. (1986). Isolation and characterization of a herpes simplex virus type 1 mutant containing a deletion within the gene encoding the immediate early polypeptide Vmw 110. J. Gen. Virol. 67, 2571–2585. Thorne, A. W., Sautiere, P., Briand, G., and Crane, R. C. (1987). The structure of ubiquitinated histone H2B. EMBO J. 6, 1005–1010. Treier, M., Staszewski, L. M., and Bohmann, D. (1994). Ubiquitin-dependent c-jun degradation in vivo is mediated by the delta domain. Cell 78, 787–798. Weber, P. C., and Wigdahl, B. (1992). Identification of dominant-negative mutants of the herpes simplex virus type 1 immediate-early protein ICP0. J. Virol. 66, 2261–2267. Weber, P. C., Kenny, J. J., and Wigdahl, B. (1992). Antiviral properties of a dominant negative mutant of the herpes simplex virus type-1 regulatory protein-ICP0. J. Gen. Virol. 73, 2955–2961. Zhu, X., Chen, J., Young, C. S. H., and Silverstein, S. (1990). Reactivation of latent herpes simplex virus by adenovirus recombinants encoding mutant IE-0 gene products. J. Virol. 64, 4489–4498.