Herpesvirus mRNAs Are Sorted for Export via ... - Journal of Virology

JOURNAL OF VIROLOGY, Mar. 2000, p. 2814–2825 0022-538X/00/$04.00⫹0 Copyright © 2000, American Society for Microbiology. All Rights Reserved.

Vol. 74, No. 6

Herpesvirus mRNAs Are Sorted for Export via Crm1Dependent and -Independent Pathways TARIK M. SOLIMAN

AND

SAUL J. SILVERSTEIN*

Department of Microbiology, College of Physicians and Surgeons, Columbia University, New York, New York 10032 Received 20 September 1999/Accepted 20 December 1999

Cellular pre-mRNA splicing is inhibited by ICP27, a herpes simplex virus regulatory protein, resulting in the shutoff of host protein synthesis. Here we reveal that ICP27 also mediates the export of some virus RNAs via a Crm1-dependent pathway and present evidence that independent domains are required for these functions. Sorting of some viral mRNAs for nuclear export requires Crm1, while other virus mRNAs are exported via another pathway.

Herpes simplex virus type 1 (HSV-1) is a large doublestranded DNA virus that encodes at least 75 proteins. Various HSV proteins function to shut off host protein synthesis, evade the immune system, activate virus gene expression, replicate virus DNA, and generate new infectious particles (reviewed in reference 62). During a productive infection, HSV-1 gene expression proceeds in a tightly regulated cascade (20, 21). Based on their temporal expression, HSV-1 genes are classified into three kinetic classes; immediate-early, early, and late. The IE gene products ICP4, ICP0, ICP27, and ICP22 cooperatively regulate the expression of all kinetic classes of virus genes (8, 9, 13, 14, 33, 47, 48, 60, 62, 64). ICP27 is an RNA binding protein that is required throughout a productive infection (64). ICP27 performs multiple functions that affect processing of pre-mRNA, including effects on splicing, polyadenylation, and transport (34–36). ICP27 inhibits splicing of cellular transcripts, which effectively shuts off host protein synthesis (15, 17, 50, 54, 66–68, 75). Because HSV transcripts are predominately intronless, this may provide a selective advantage for virus gene expression (75). ICP27 may also have a direct role in virus gene expression, as it binds RNAs and mediates their nucleocytoplasmic export (4, 39, 40, 65, 75). Currently, the mechanism of RNA recognition by ICP27 is unclear, because neither the RNAs it binds to nor the sites at which ICP27 binds RNA have been determined (4, 23, 40, 65). Domains identified within the amino terminus of ICP27 include a nuclear export signal (NES), a nuclear localization signal (NLS), and an RGG box-type RNA binding motif (38, 40, 65). The carboxy terminus of ICP27 is required for both its activator and repressor functions as defined in reporter assays (16, 37, 68, 74). However, the only domain described for this region is a potential zinc finger motif (78). Here, we propose that the carboxy terminus of ICP27 contains three K homology (KH)-like RNA binding motifs as well as an SM protein-protein interaction motif. The KH motifs were first identified in the human heterogeneous nuclear ribonucleoprotein (hnRNP) K protein as a triple repeat (73). Subsequently, a number of proteins, such as Fmr-1, Nova-1, and ribosomal S3 proteins, were also found to contain KH motifs (5). It is now clear that KH domains bind

single-stranded RNA either singly or collectively and often nonspecifically. A single point mutation in the KH motif of Fmr-1, which disrupts RNA binding in vitro, results in a severe form of fragile X syndrome (72). The structures of a wild-type KH domain as well as one carrying the equivalent Fmr mutation have been determined by nuclear magnetic resonance (42). The point mutation disrupts the structure of the KH domain, which explains the loss of function. Here, we introduced an equivalent mutation into the potential KH domains in ICP27 to study their function. SM domains were identified as a common motif found in a set of evolutionarily conserved SM proteins, which are essential for the biogenesis of snRNP particles that mediate premRNA splicing (18, 70). The SM domain is utilized by these proteins to form a higher-order complex through protein-protein interactions (26). Intragenic suppressors of a conditional lethal temperaturesensitive (ts) mutant were identified, and they map to the carboxy terminus of ICP27 (75). The suppressors restored the defects in virus growth and shuttling which were disrupted by the ts mutation. While these mutations affect shuttling, they do not map to the amino-terminal NES of ICP27. We have now probed the basis for these observations and propose an explanation for this phenotype. In the past, ICP27 null viruses were used to characterize ICP27-dependent effects on virus gene expression. However, these results may be misleading, as ICP27 performs multiple functions that regulate virus gene expression. To examine the individual contributions of two functions of ICP27, RNA export and host shutoff, on HSV gene expression, we disrupted each of them either metabolically or genetically. MATERIALS AND METHODS Cells and viruses. Vero cells were grown and maintained in Dulbecco’s modified Eagle’s medium (Gibco BRL, Grand Island, N.Y.) containing 5% bovine calf serum (HyClone Laboratories, Inc., Logan, Utah) and supplemented with 100 U of penicillin and 100 ␮g of streptomycin (Gibco BRL) per ml. The strain of wild-type HSV-1 used in this study was KOS 1.1A. HSV-1 strains vBSLG4 (tsR480H) and vBS⌬27 were described previously. Virus hr114 has a deletion of the UL52 gene and was provided by Sandra Weller, University of Connecticut. The D448A virus was generated by marker rescue as described previously (75) with a BamHI-linearized fragment containing the ␣27 allele from the plasmid pBSD448A. Plasmids. All point mutations were generated using the QuickChange sitedirected mutagenesis kit (Stratagene, La Jolla, Calif.) and oligonucleotides as indicated below. To verify that these mutations were introduced, plasmids were sequenced by the Columbia University Cancer Center DNA Sequencing Facility using oligonucleotides specific for the ICP27 open reading frame. The template

* Corresponding author. Mailing address: Department of Microbiology, College of Physicians and Surgeons, Columbia University, 701 W. 168th St., New York, NY 10032. Phone: (212) 305-8149. Fax: (212) 305-5106. E-mail: [email protected]. 2814

VOL. 74, 2000 used to generate these mutations was pBS27. It contains a wild-type copy of the ␣27 gene as described previously (75). pBSD448A contains a point mutation in the ICP27 gene that changes amino acid 448 from D to A. The oligonucleotides used were 27D448A5⬘ (5⬘ GCGGAGATCGCCTACGCGACC) and 27D448A3⬘ (5⬘ GGTCGCGTAGGCGATCTCCGC). pBS27nes contains two point mutations in the ICP27 gene that change amino acids 13 and 15 from L to A. The oligonucleotides used were 27NES5⬘ (5⬘ GCTAATTGACGCCGGCGCGGACCTCTCCG) and 27NES3⬘ (5⬘ CGGAGA GGTCCGCGCCGGCGTCAATTAGC). pBSR3A contains a triple mutation in the ICP27 gene that changes amino acids 144 to 146 from R to A. The oligonucleotides used were 27R3A5⬘ (5⬘ GACGCCGTGGGGCTGCCGCGGGTCGGGGTCG) and 27R3A3⬘ (5⬘ CGA CCCCGACCCGCGGCAGCCCCACGGCGTC). pBSF303N contains a point mutation in the ICP27 gene that changes amino acid 303 from F to N. The oligonucleotides used were 27F303N5⬘ (5⬘ GGAGG GCCCAATGACGCCGAG) and 27F303N3⬘ (5⬘ CTCGGCGTCATTGGGCCC TCC). pBSL387N contains a point mutation in the ICP27 gene that changes amino acid 387 from L to N. The oligonucleotides used were 27L387N5⬘ (5⬘ CCGCG GCGGTGAACGATAACCTCGCC) and 27L387N3⬘ (5⬘ GGCGAGGTTATCG TTCACCGCCGCGG). Protein synthesis assay. Labeling and analysis were done as described previously (75). Immunoblotting. Cells were infected at a multiplicity of infection (MOI) of 10 and harvested at various times. Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with 7.5% acrylamide and transferred to nitrocellulose. Primary antibodiies used were as follows: for ICP4, mouse monoclonal 58S (71); for ICP0, Clu7 (51); for ICP27, Clu38 (51); for UL9, R-anti-UL9 (provided by M. Challberg, National Institutes of Health); for VP5, NC-1 (provided by G. Cohen, Department of Microbiology, University of Pennsylvania); anti-VP16 (Clontech); for gB, R69 (provided by G. Cohen); and for gC, R47 (provided by G. Cohen). Secondary antibodies were horseradish peroxidase-conjugated goat anti-mouse and anti-rabbit antibodies (Kirkegaard and Perry, Gaithersburg, Md.). Detection utilized the chemiluminescent substrate LumiGLO (Kirkegaard and Perry) followed by exposure to Biomar blue film (Marsh Biochemical, Rochester, N.Y.). Immunofluorescence. Vero cells were infected with 10 PFU of HSV-1 per cell for 6 h and then treated with 100 ␮g of cycloheximide per ml for 2 h and harvested. Where indicated below, cells were treated with 300 ␮g of phosphonoacetic acid (PAA) per ml and 25 ng of Leptomycin B (LMB) (provided by Minoru Yoshida, University of Tokyo [28–30, 44]) per ml. Cells were fixed with 3.7% formaldehyde, permeabilized with acetone, and incubated with the polyclonal antibody Clu38. ICP27 was visualized by addition of fluorescein isothiocyanate-conjugated goat anti-rabbit immunoglobulin G (Kirkegaard and Perry). Preparations were viewed using a Leitz Dialux microscope with optical systems for the selective visualization of fluorescein. Representative fields of cells were photographed using EPH 1600 film (Eastman Kodak, Rochester, N.Y.) and a Nikon UFX-DXII photographic system. Images of the developed slides were secured using a Polaroid ES Plus scanner and Adobe Photoshop software. Digitized images were assembled using Canvas on a Macintosh computer. UV cross-linking, RNA extraction, and Northern blot analysis. UV crosslinking and RNA isolation of cytoplasmic and nuclear poly(A) RNA were done as described previously (55). Briefly, Vero cells were infected with 10 PFU of HSV-1 or tsR480H per cell for 9 h at 39.5°C for cross-linking experiments and at 37°C for Northern blot analysis. LMB was used as described above. UV cross-linking. Medium was removed from cell monolayers on a 150-mm tissue culture dish, the monolayers were washed twice with PBS(⫹) (phosphatebuffered saline containing 1 mM CaCl2 and 0.5 mM MgCl2), and medium was replaced with 2 ml of PBS(⫹). To induce covalently cross-linked RNA-protein complexes, a 15-W germicidal light was placed 4.5 cm from the uncovered 150-mm tissue culture plate for 4 min. Cell fractionation. Cells were lysed directly on the tissue culture plate by adding 3.5 ml of RSB (10 mM Tris [pH 7.4], 10 mM NaCl, and 1.5 mM MgCl2) supplemented with 0.5% Triton X-100, 10 mM Vanadyl ribonucleoside complex (Gibco BRL), and complete protease inhibitor (Boehringer Mannheim Biochemicals, Indianapolis, Ind.). Two hundred microliters of 10⫻ MAGIK (10% Tween 40 and 5% deoxycholate) was added after 2 min. Cells were scraped with a rubber policeman, passed four times through a 25-gauge needle, and then centrifuged at 3,000 ⫻ g. The cytoplasmic fraction was saved and the nuclear fraction was resuspended in the same volume of RSB following treatment with RQ1 DNase (Promega, Madison, Wis.) for 15 min at 37°C. Poly(A) RNA isolation. Nuclear and cytoplasmic RNA fractions were supplemented with SDS to 1%, 2-mercaptoethanol to 1%, and EDTA to 10 mM and then heated to 68°C for 10 min. Samples were quickly cooled in an ice bath and adjusted with LiCl to 500 mM. Poly(A) RNA was isolated using oligo(dT)cellulose (Sigma, St. Louis, Mo.) by allowing binding at room temperature, performing three washes with binding buffer (10 mM Tris-HCl, 1 mM EDTA, 1% SDS, and 0.5 M LiCl), and eluting the samples at 68°C with elution buffer (10 mM Tris-HCl, 1 mM EDTA, and 0.05% SDS). Poly(A) RNA was ethanol precipitated with LiCl. For UV cross-linking experiments, samples were treated with RNase for 30 min and subjected to SDS-PAGE (see above). For Northern blot analysis, RNA samples were resuspended in RNase-free water.

ICP27 DOMAINS AND FUNCTIONS

2815

TABLE 1. Rescue of an HSV ICP27 null by plasmid DNA containing point mutations in the gene encoding ICP27a ICP27 domain with mutation

Alteration(s)

None (wild type) None NES L13A, L15A RGG R144–146A KH1 F303N KH2 L387N SM D448A SM-KH3 R480H

Marker rescue

Yield

Phenotype

Yes No Yes No Yes Yes Yes

109 ND 107 ND 108 107 109

Wild type Lethal ND Lethal ND Host shutoff⫺ ts lethal, shuttling⫺, host shutoff⫺

a The mutations in specific domains of ICP27 used in this study are presented with the results of marker rescue experiments using an ICP27 deletion virus as the target. Comparative yields are given for recombinant viruses. The phenotype of each recombinant virus is described. ND, not done.

Northern blot analysis. RNA was subjected to electrophoresis in 1.5% agarose–6% formaldehyde gels and transferred to GeneScreen Plus (NEN-Dupont, Boston, Mass.). The blot was probed, stripped, and reprobed several times utilizing 32P-labeled DNA probes prepared by random priming of gel-isolated DNA fragments using the Strip-EZ DNA kit (Ambion, Austin, Tex.). DNA fragments for the HSV-specific probes were isolated as follows: for ICP4, a BamHI fragment from plasmid pGX58 (a gift from Chris Preston, MRC Virology Unit, Glasgow, United Kingdom); for VP16, a SalI fragment from plasmid pRab14 (2); for gC, an EcoRI/HindIII fragment from plasmid gC18; and for gH, an EcoRI/BamHI fragment from plasmid gH-18 (K. Wen and S. J. Silverstein, unpublished results). The fragment for UL18 was generated by PCR using oligonucleotides UL18 U (5⬘ GAGGTTGGTCGCCCGTCTCTGCTAC) and UL18 L (3⬘ TCCGCACACATTCGCTCCTATCACAC). The fragment for UL35 was generated by PCR using oligonucleotides UL35 U (5⬘ AAAAAAGG ACGCACCGCCGCCCTAATC) and UL35 L (3⬘ CGCCGTGCTGACCAGCC TACATCAC).

RESULTS Functional analysis of ICP27. Previous results demonstrated that three independent intragenic suppressors of a conditional lethal ts mutation in ICP27 restore shuttling and virus growth (75). These results suggested that ICP27’s RNA export function may be critical for virus growth. To address this hypothesis, we asked what phenotype amino acid substitutions in the NES and RGG box RNA binding motifs would have in a marker rescue assay. Amino acid substitutions in the NES of ICP27 were chosen based on loss-of-function mutations in NES sequences from other proteins (80). The highly conserved leucines at positions 13 and 15 were changed to alanines. The resulting plasmid (pBS27nes) was used in marker rescue experiments to attempt to complement an ICP27 null mutant HSV for virus growth. Failure to rescue virus growth suggested that the NES of ICP27 was essential for its biological activity (Table 1). Sandri-Goldin reported a similar result when she attempted to complement an ICP27 null virus with an ICP27 plasmid that had a deletion in its NES. Consistent with the function of an NES, the mutant protein was confined to the nucleus (65). To determine if the known RNA binding motif was essential, an arginine triad in the center of the ICP27 RGG box was changed to alanines. Marker rescue experiments revealed that this mutation was not lethal (Table 1). However, the yield from this virus was reduced compared to that from wild-type HSV-1, and ICP27 appeared to overaccumulate in the cytoplasm of infected cells (data not shown). Previously, a virus with a deletion of the RGG box was reported to also have a reduced yield in a single-step growth curve (38). As ICP27 is a known RNA binding protein, these results suggest that other essential RNA binding domains are present and unidentified. Amino acid alignment reveals potential RNA binding domains in ICP27. To identify other potential RNA binding

2816

SOLIMAN AND SILVERSTEIN

domains, the amino acid sequence of ICP27 was compared to known RNA binding domains. Utilizing a visual alignment of amino acids, three regions of ICP27 (KH1, KH2, and KH3) with potential homology to a known RNA binding motif (KH) from the cellular protein hnRNP K were identified (Fig. 1A). The size and spacing of these motifs in ICP27 are consistent with those found in other proteins containing KH domains. Alignment of the amino acid sequences of ICP27 homologs from other alphaherpesviruses revealed that these motifs are highly conserved (Fig. 1B). Embedded between and overlapping with the KH3 motif in ICP27 is an SM motif that is also conserved among the homologs (Fig. 1A and C). SM motifs are found in specific splicing factors (SM proteins) and are used to mediate protein-protein interactions. The three C-terminal KH motifs map to the activation region and the SM motif maps to the overlapping repressor region of ICP27 as defined by reporter assays (Fig. 1A) (16, 37, 61). The KH motif was first identified in hnRNP K as a triple repeating motif (73). The identification of KH domains in other proteins is difficult because the motif consists of conserved hydrophobic amino acids that are precisely spaced. The alignment shown in Fig. 1B highlights the conserved hydrophobic amino acids and the defined spacing. The KH domains in ICP27 fit the consensus with the same degree of flexibility as other KH domains (42). For example, KH3 of ICP27 has 14 of 18 positions conserved, which is identical to the number of positions conserved in the KH1 domain of Fmr-1. The KH1 and KH2 domains of ICP27 are slightly less conserved; they have 11 and 13, respectively, of the 18 possible conserved amino acids. Similar criteria were used in the alignment of the SM motifs; 7 of the 8 conserved amino acids were found in the putative SM motif in ICP27. While fewer amino acids are highly conserved among SM motifs, the putative SM motif in ICP27 contains a high proportion of these residues that are properly spaced. The alignment of these putative KH and SM motifs may provide a framework to investigate how ICP27 functions. In the following experiments we assessed if these motifs constitute functional domains of ICP27. The KH-like motifs in ICP27 overlap precisely with what was functionally defined as the activation domain (16, 61). Numerous insertion and deletion mutations have been generated in this region of ICP27, and they identify it as essential to the protein’s biological function. To determine whether the KHlike motifs are the functional elements of the activation domain, single amino acid substitutions, resembling a previously described mutation in the KH2 domain of Fmr-1, were introduced separately into the KH1 and KH2 motifs of ICP27. This change, when present in KH motifs in other proteins, disrupts their structure and function (42, 72). Marker rescue experiments revealed that a mutation (F303N) in the KH1 motif of ICP27 was lethal (Table 1). In contrast, the corresponding amino acid substitution in the KH2 motif (L387N) was able to rescue virus growth (Table 1). These results suggest that the KH1 motif is a functional element of the activation domain that is essential for virus growth. UV cross-linking of RNA-protein complexes in vivo. The mutation R480H in ICP27 from the ts mutant LG4 resides in the putative KH-like motif KH3 and results in a lethal phenotype at the restrictive temperature (75). To determine if this KH-like motif is required for RNA binding, we asked whether the ts mutation affects RNA binding by ICP27. UV crosslinking experiments to assay for RNA binding by ICP27 were performed with cells infected with wild-type HSV-1 and tsR480H at the restrictive temperature. UV light was used to covalently cross-link RNA-protein complexes in vivo. Crosslinked poly(A) RNA-protein complexes were purified under

J. VIROL.

protein-denaturing conditions by oligo(dT) chromatography from nuclear and cytoplasmic cellular fractions, digested with RNase, and subjected to SDS-PAGE analysis. The results, assayed by Western blot analysis, show that ICP27 copurified with poly(A) RNA from cells infected with wild-type HSV-1 but not from cells infected with tsR480H (Fig. 2A). This finding demonstrates that the ts mutation abrogates RNA binding by ICP27 in vivo and supports a critical role for the KH3 motif in RNA binding. The control extract from cells infected with wild-type HSV-1 was not UV cross-linked. Under these conditions ICP27 still copurified with poly(A) RNA. To determine whether ICP27 bound to the RNA or the oligo(dT) column, identical samples were either treated with RNase A for 30 min or left untreated prior to oligo(dT) chromatography. Under these conditions, ICP27 was unable to bind to the column after RNase treatment, demonstrating the specificity of binding (Fig. 2C). This result reveals that ICP27 is bound tightly to poly(A) RNA in both the nucleus and cytoplasm, because protein-denaturing conditions did not dissociate the complex. Although there appears to be more ICP27 in lanes that were not cross-linked, these results varied and are not indicative of a significant difference (Fig. 2A). Western blots of samples taken prior to oligo(dT) chromatography reveal the intracellular distribution of ICP27 and confirm the presence of the protein in those extracts. In cells infected with HSV-1, ICP27 is distributed between the nuclear and cytoplasmic compartments, whereas ICP27 from tsR480H-infected cells is predominantly in the nuclear compartment (Fig. 2B). This confirms previous results obtained by immunofluorescence analysis which showed that the ts mutation disrupts shuttling between the nucleus and the cytoplasm (75). These data suggest that loss of RNA binding activity may limit or prevent the nuclear export of ICP27. Trafficking of ICP27. We previously proposed that RNA binding by ICP27 was a prerequisite for nuclear export (75). In those experiments, cytoplasmic accumulation of transiently expressed ICP27 required coexpression of HSV late RNA. Moreover, shuttling of ICP27 in infected cells occurred only at late times postinfection, correlating with the expression of late RNA. To strengthen this correlation, we examined the effect of limiting HSV late gene transcription on the nuclear export of ICP27. Because true late gene transcription is dependent on viral DNA replication, inhibition of late gene transcription was accomplished by inhibiting viral DNA replication (19, 25, 32). Inhibition of viral DNA replication was accomplished in two ways. The first method involved using the drug PAA, a potent inhibitor of herpesvirus DNA replication (1, 22, 57). In the second approach, hr114, a DNA replication-defective mutant of HSV-1, was used. Localization of ICP27 under these conditions was assayed by indirect immunofluorescence of cells infected for 8 h with HSV-1 at a high MOI. ICP27 appeared in the cytoplasm of 50 to 80% of cells infected with wild-type virus (Fig. 3A). Cytoplasmic accumulation results from shuttling of ICP27, because the mutant tsR480H does not accumulate ICP27 in the cytoplasm under the same conditions (Fig. 3B). Limiting late gene transcription by metabolic inhibition with PAA or by genetic measures with the mutant hr114 results in reduced accumulation of ICP27 in the cytoplasm (Fig. 3C and E). When the complementing cell line 2D6 is infected with hr114, the accumulation of ICP27 in the cytoplasm is restored (Fig. 3F). Thus, limiting HSV late gene transcription or the presence of a ts mutation in the KH3 motif affects the accumulation of ICP27 in the cytoplasm. While the ts mutation in KH3 affects RNA binding by ICP27 and expression of late proteins, transcription of late RNAs is not affected (68). This suggests that the defect in the expression of late proteins re-

VOL. 74, 2000


2817

FIG. 1. Schematic diagram of known and proposed functional domains in ICP27. (A) The characterized domains include an NES, an NLS, and an RGG box-type RNA binding motif. Domains proposed in this study include three KH motifs (KH1, KH2, and KH3) and an SM motif. Also shown are the sequence of the NES and RGG box, with the amino acids that were mutated to alanine underlined; the position of the ts mutation; the locations of the D448A mutation and three independent ts suppressor mutations (sup 3-3, 4-3, and 5-3); and the sites of the Fmr-like mutations (F303N and L387N). (B and C) Amino acid alignments of ICP27 KH and SM motifs from alphaherpesviruses. (B) Alignments of the KH-like motifs in ICP27 homologs from HSV, bovine herpesvirus (BHV), varicella-zoster virus (VZV), pseudorabies herpesvirus (PRV), and equine herpesvirus (EHV) with hnRNP K, Fmr-1, and Nova-1. (C) Alignment of the putative SM motifs in ICP27 homologs with the SM motifs in the B, D2, D1, D3, E, F, and G core proteins from Homo sapiens (H. Sap). Dark gray highlights the conserved amino acids and the proper spacing that define the KH and SM motifs. Below each alignment is the consensus for the motifs, given as the amino acids utilized at the defined positions. Underlined are residues referred to in the text that are altered. Overlined residues are the amino acids common to the KH3 and SM motifs in ICP27 from HSV-1.

2818


FIG. 2. UV cross-linking of ICP27 to poly(A) RNA. Vero cells were infected with wild-type HSV-1 or tsR480H at the restrictive temperature for 8 h and then either exposed to UV light (⫹) to cross-link (Xlink) RNA-protein complexes or left untreated (⫺). (A) Poly(A) RNA was purified by oligo(dT) chromatography under protein-denaturing conditions. Eluted samples were treated with RNase and subjected to SDS-PAGE. To determine if ICP27 copurified with the RNA, Western blot analysis was performed using a rabbit polyclonal antiserum specific for ICP27 (Clu38). (B) Samples taken prior to oligo(dT) chromatography were treated with RNase and subjected to SDS-PAGE and Western blot analysis as described above. (C) Samples were treated with RNase A prior to chromatography and then eluted and subjected to SDS-PAGE and Western blot analysis as described above.

sults from the failure of ICP27 to export RNA. Furthermore, temperature shift-down experiments with tsR480H revealed that ICP27 can readily resume shuttling (75), suggesting that the loss of RNA binding is the reason nuclear export is affected. These results demonstrate that the shuttling protein ICP27 appears to exit the nucleus only after binding to late RNA. Heterokaryon assays have also been used to demonstrate that ICP27 shuttles between the nucleus and cytoplasm (39). However, this experimental approach is not applicable to the above-mentioned analysis because ICP27 is already present in the cytoplasm of cells infected with wild-type HSV. Thus, the heterokaryon assay would not demonstrate shuttling under these conditions, because the cytoplasmic ICP27 is free to move into the nucleus of the heterokaryon. Likewise, shuttling mutants of ICP27 that limit the accumulation of ICP27 in the cytoplasm of infected cells also do not score positive in a heterokaryon shuttling assay (39). Thus, the heterokaryon assay is useful for determining whether predominantly nuclear proteins are capable of shuttling, while proteins whose subcellular localization is both cytoplasmic and nuclear are not suitable substrates. Some RNA export proteins use the exportin Crm1, a RanGTP-dependent transporter, to shuttle their cargo from the nucleus to the cytoplasm (56). To determine if ICP27 utilized Crm1 as a cofactor for nuclear export we utilized LMB, an inhibitor of Crm1 (11, 12, 49). Infected cells were treated with LMB after virus adsorption, and the localization of ICP27 was examined at 8 h postinfection by indirect immunofluorescence. Figure 3D demonstrates that LMB blocks the cytoplasmic accumulation of ICP27. This is consistent with ICP27 containing a leucine-rich NES that uses Crm1 to exit the nucleus. Cytoplasmic RNA accumulation as a measure of RNA export by ICP27. ICP27 is reported to export all HSV intronless RNAs (65). However, an ICP27 deletion virus was used in those studies, and thus the loss of multiple ICP27 functions

J. VIROL.

may not accurately reflect the dependence of RNAs on ICP27’s RNA export function. To identify the RNAs that are dependent on ICP27 for their export and not for other functions, the effect of inhibiting the nuclear export of ICP27 on the cytoplasmic accumulation of HSV-1 RNAs was examined. LMB was used to block the nuclear export of ICP27 as described above. Poly(A) RNA was isolated from the nucleus and cytoplasm of cells infected with HSV-1 in either the presence or absence of LMB. RNAs were displayed on denaturing agarose gels, and following transfer to GeneScreen Plus, the filters were hybridized with a series of HSV-1 gene-specific probes and the relative intensities of the hybridization signals were quantitated. The cytoplasmic accumulation of RNA encoding the immediate-early protein ICP4 and the late protein VP22 was unaffected by the presence of LMB. In contrast, the amount of several late RNAs (those encoding VP16, UL17, UL18, UL35, gC, and gH) showed reduced cytoplasmic accumulation in the presence of LMB. These results reveal that virus-specified late RNAs, to various degrees, are dependent on the nuclear export function of ICP27 (Fig. 4). To a first approximation, the levels of these transcripts in the nucleus of infected cells were equivalent in the presence and absence of LMB. This suggests that transcription and polyadenylation are not determinants of nuclear retention in the presence of LMB. While one might expect increased nuclear accumulation of these transcripts, it is conceivable that transcripts which overaccumulate in the nucleus are subject to degradation. Thus, the appearance of similar levels of these transcripts in the nucleus may simply reflect the steady-state level of these RNAs rather than a decrease in their transcription rate. Support for this hypothesis comes from experiments with the ts mutant LG4, in which ICP27 was shown to act a posttranscriptional level (68, 74, 75).

FIG. 3. Shuttling of ICP27. Vero cells (A to E) or 2D6 cells (F) were infected at an MOI of 10 with wild-type HSV-1 (A, C, and D), tsR480H (B), or hr114 (E and F). After 1 h of adsorption, HSV-infected cells were treated with PAA (C) or LMB (D). Infected cells were incubated at 39.5°C and processed for immunofluorescence at 8 h postinfection following a 2-h treatment with cycloheximide. The intracellular distribution of ICP27 was examined with an antibody specific for ICP27 and a fluorescein isothiocyanate-labeled secondary antibody.

VOL. 74, 2000

FIG. 4. Compartmentalization of HSV RNAs. Nuclear (N) and cytoplasmic (C) poly(A) RNA was purified from cells infected with HSV in either the presence or absence of LMB as indicated. Representative Northern blots were hybridized with HSV gene-specific probes for RNAs encoding ICP4, VP16, UL18, UL35, gC, and gH. Because of the overlaps in HSV transcripts, VP22 RNA was also detected with the VP16 probe and UL17 RNA was also detected with the UL18 probe. Shown are the autoradiograms. Signals were detected and quantified separately using a PhosphorImager and ImageQuant software (Molecular Dynamics). Results (% of HSV-1) represent the percentage of signal detected in lanes where RNA was isolated from HSV-infected cells in the presence of LMB (⫹LMB) compared to lanes where RNA was isolated from HSV-infected cells in the absence of LMB (HSV-1). ND, no data were available because the signals detected were not significantly above background levels. In some cases, the autoradiograms were not exposed long enough to show intense bands; therefore, it is the quantitative data collected from the PhosphorImager that are presented.

The most notable results were with the VP16 and VP22 RNAs. These transcripts are both detected with the VP16specific probe. This occurs because the VP22 transcript starts upstream of the VP16 transcript and continues through it to utilize a common polyadenylation site. Analysis of the hybridization signals generated by these overlapping transcripts shows that the accumulation of VP22 RNA is unaffected by LMB, while the VP16 RNA is reduced 42%. Thus, HSV-1 transcripts can utilize independent pathways for their nuclear export, since LMB differentially affects the cytoplasmic accumulation of two late intronless transcripts that utilize the same polyadenylation site. Role of the putative SM motif in host shutoff. Shortly after infection with HSV-1, host cell protein synthesis is shut off as a result of the introduction of a virion-associated protein (vhs) and the expression of ICP27 (10, 45, 46, 76). ICP27 inhibits host cell pre-mRNA splicing, resulting in the shutoff of host cell protein synthesis (15, 17, 66, 67, 75). The identification of an SM-like motif in the region of ICP27 required for suppressing splicing led us to hypothesize that this domain may mediate ICP27’s effect on host cell splicing. SM domains are found in cellular SM proteins that are involved in splicing. These pro-


2819

FIG. 5. Protein synthesis in cells infected with HSV. Vero cells were either mock infected (M) or infected with wild-type HSV-1, vBS⌬27, or the D448A mutant at an MOI of 10 and incubated at 37°C. The cells were then pulse-labeled with [35S]methionine for 30 min at 4, 8, or 12 h. Total-cell extracts were harvested and subjected to SDS-PAGE analysis. Labeled proteins were visualized by autoradiography. Solid circles identify host proteins, and arrows indicate virusspecified proteins.

teins utilize SM domains to mediate protein-protein interactions to form a higher-order complex containing all seven members of the SM family. Critical residues required for complex formation within the SM domains have been identified (6, 26). The corresponding amino acids in the SM domain of ICP27 are D448 and R480. The ts mutation R480H disrupts host cell shutoff. This phenotype may reflect the location of this critical amino acid in an SM motif. However, this mutation also maps to the overlapping KH3 motif and perturbs RNA binding, as shown above. To segregate these phenotypes, a change to another conserved amino acid (D448A) in the ICP27 SM domain was made. This mutation exclusively maps to the SM motif between the KH2 and KH3 motifs and thus should not affect RNA binding as the ts mutation does. A mutant HSV-1 carrying this allele of ICP27 (D448A) was generated by transplacement as described previously (Table 1). (75). A kinetic analysis of polypeptide synthesis in cells infected with D448A, HSV-1, or an ICP27 null virus was performed. The pattern of polypeptides synthesized in infected cells pulselabeled at 4-h intervals was examined by SDS-PAGE. The autoradiogram in Fig. 5 shows that there was a gradual shutoff of host protein synthesis in cells infected with HSV-1, and virus proteins were readily detected by 4 h postinfection. By contrast, in cells infected with the ICP27 deletion virus, host proteins continued to be synthesized and only a few virus proteins were detected. In cells infected with the D448A mutant the level of virus proteins synthesized was lower than that in cells infected with the wild type, while host proteins continued to be synthesized as late as 9 h postinfection. This delay in host shutoff differs from the complete loss of host shutoff observed in cells infected with the ICP27 null virus. However, it is consistent with a defective host shutoff phenotype. These results

2820


FIG. 6. Analysis of virus growth kinetics. Vero cells were infected at an MOI of 0.1 with wild-type HSV-1 or the D448A mutant and incubated at 37°C. The infections were stopped at the indicated times, and the virus yields were determined by titration on Vero cells. Data points represent the averages of three independent infections, each titrated in duplicate.

confirm the hypothesis that the SM-like motif in ICP27 is a functional domain which is involved in host cell shutoff. To determine if the growth rate and/or virus yield was affected by the loss of ICP27’s host shutoff function, the kinetics of virus production by the D448A mutant was compared to that of wild-type HSV-1. Figure 6 shows that while the D448A mutant is able to grow on Vero cells, its rate of growth and the

J. VIROL.

yield are lower than those of wild-type HSV-1. These results are identical to the phenotype of the intragenic suppressors of LG4 that were described previously. The suppressor mutations which restore shuttling and virus growth do not restore ICP27’s host cell shutoff function (75). Thus, ICP27’s host shutoff function, while not essential, provides for a more robust infection. Separation of ICP27’s host shutoff function from its RNA export function. To study the individual contributions of ICP27’s functions in RNA export and host cell shutoff to HSV-1 gene expression, the accumulation of viral proteins over the course of infection was examined while limiting one or both of these functions. To examine the effect of inhibition of shuttling and RNA export by ICP27, cells were infected in the presence or absence of LMB. To examine the influence of ICP27’s host cell shutoff function on HSV gene expression, cells were infected with the D448A mutant or wild-type HSV-1. Finally, cells were infected with tsR480H at the nonpermissive temperature to examine the effect of limiting both RNA export and host cell shutoff, and these results were compared to those for wild-type HSV-1 infection under the same conditions. Cell extracts were harvested at 3, 6, and 9 h postinfection and subjected to SDS-PAGE. Western blots were performed using antibodies to immediate-early proteins ICP4, ICP0, and ICP27; early protein UL9; and late proteins VP5, VP16, gB, and gC. Inhibition of RNA export by ICP27 reveals that immediate-early and early protein products accumulate to wild-type levels whereas accumulation of late proteins is markedly decreased, with the effect on gC being the most severe (Fig. 7). This result correlates with the decrease in the cytoplasmic accumulation of several late RNAs, including those for VP16 and gC (Fig. 4). When ICP27’s host shutoff function is inhibited (D448A),

FIG. 7. Analysis of virus-specified protein accumulation. Vero cells were either mock infected (M) or infected with wild-type HSV-1, the D448A mutant, or tsR480H at an MOI of 10. Where indicated, HSV-1-infected cells were treated with LMB to block Crm1-mediated export of ICP27. For the last panel, wild-type- and tsR480H-infected cells were incubated at 39.5°C. For the other panels, infected cells were incubated at 37°C. At the times indicated, infected cells were harvested and subjected to SDS-PAGE. Shown are representative Western blots using a series of antibodies specific for HSV-1 proteins ICP4, ICP0, ICP27, UL9, VP16, VP5, gB, and gC.

VOL. 74, 2000


there is a global decrease in the accumulation of immediateearly and early proteins and a delay in the appearance of late proteins. The delay in the appearance of late proteins may reflect the decrease in immediate-early proteins, as they are required to regulate expression of late genes. Note that the levels of ICP0, which is encoded by a spliced RNA, are also reduced in the absence of host shutoff. This suggests that the effect on host shutoff is not limited to increasing the expression of intronless genes. In cells infected with tsR480H at 39.5°C, these effects appear to be compounded compared to the effects in cells infected with wild-type HSV-1. The accumulation of both immediate-early and early proteins is reduced and gC is virtually undetectable. This phenotype differs from that seen in cells infected with the D448A mutant, which is the host shutoff mutant. Although the effect on late gene expression in the absence of host shutoff is complicated by the reduction in immediate-early and early gene expression, suppressor analysis has revealed that the defect in RNA export is lethal, while the absence of host shutoff results in only a moderate growth defect. These experiments demonstrate that ICP27 functions in at least two distinct ways to increase the expression of viral proteins: through the export of HSV RNA and the shutoff of host protein synthesis. DISCUSSION ICP27 is a multifunctional protein that binds RNA and shuttles between the nucleus and cytoplasm. Functions performed by ICP27 include inhibition of cellular pre-mRNA splicing, modulation of ICP4’s DNA binding activity, activation of premRNA polyadenylation, and RNA export (4, 15–17, 38–40, 51–54, 58–61, 65–68, 74, 75, 77). Domains previously identified within ICP27 include an NES, an NLS, and an RGG box-type RNA binding domain. Here, we provide evidence for additional functional domains in ICP27 that mediate RNA binding and inhibit host cell pre-mRNA splicing. The experiments described here demonstrate that RNA binding is a component of ICP27’s essential function in the export of HSV late RNAs and that inhibition of splicing enhances viral protein synthesis, leading to increased virus titers. Analysis of ICP27’s carboxy terminus revealed three KHlike RNA binding motifs. Sequence alignment showed that these motifs are conserved among ICP27 homologs from other alphaherpesviruses (Fig. 1), providing further confirmation of their identity and potential importance. Several insertion and deletion mutations have been generated in what we define as the KH domains in ICP27, and they uniformly resulted in lethal phenotypes (16, 37, 59, 61). Here, single point mutations that are known to disrupt the structure of a KH domain were introduced into the sequences encoding the KH1 and KH2 motifs of ICP27 to determine if these motifs are the functional elements of the activation domain. This analysis revealed that substitution of a single highly conserved amino acid in the KH1 motif disrupts virus growth, supporting our hypothesis that the KH domains contribute to the function of the activation domain. In an analysis of whether the KH motifs in ICP27 contribute to RNA binding, we demonstrated that ICP27 from LG4 (which has a conditional lethal ts mutation [R480H] in the KH3 motif) does not bind RNA in vivo (Fig. 2). Interestingly, three intragenic suppressor mutations that rescue the ts phenotype map to the KH1 and KH3 motifs. Since the RGG box has previously been shown to be required for RNA binding (4, 23, 40, 65), these data suggest a cooperative role for the KH motifs together with the N-terminal RGG box in RNA binding. ICP27 is reported to export HSV intronless RNA (65). RNA export by ICP27 would require two activities of the protein,

2821

RNA binding and shuttling. The finding that amino acid changes in either the NES, KH1, or KH3 motif in ICP27 are lethal suggests that RNA export may be an essential function in virus gene expression. This is supported by previous data that showed a correlation between shuttling and late gene expression with virus growth (75). Identification of RNAs exported by ICP27 was facilitated by the use of LMB, an inhibitor of Crm1. Crm1 binds to proteins that contain a leucine-rich NES and mediates their export from the nucleus. ICP27 contains this type of NES, and its nuclear export was inhibited by LMB, consistent with a Crm1-dependent export pathway. Northern blot analysis revealed that LMB blocked, to various degrees, the accumulation of HSV late RNAs in the cytoplasm. Therefore, ICP27 must mediate, in part, their transport. These effects were independent of transcription and polyadenylation, as accumulation of late polyadenylated nuclear RNAs was not reduced by LMB (Fig. 4). This confirms that some intronless HSV RNAs are exported by ICP27 (65). Other HSV intronless RNAs may be exported by a cellular pathway, since the RNAs for ICP4 and VP22 accumulated to wild-type levels in the cytoplasm in the presence of LMB, as was shown for the thymidine kinase (TK) RNA (50). Therefore, this study revealed that the export of only some late RNAs is dependent on ICP27. Through competition experiments and utilization of LMB, it is becoming clear that Crm1 is not routinely utilized for mRNA export in higher eukaryotes or in Saccharomyces cerevisiae (11, 43). In contrast, Rev from human immunodeficiency virus and ICP27 from HSV utilize Crm1 for the export of virus-specified RNAs (56). While other export pathways exist, and several Crm1 homologs can be found in the databases, these pathways are not yet well defined. Thus, it is not clear why Crm1 appears to be the adapter of choice for these viral RNA export proteins. Cellular RNA transport is a complex process requiring many proteins to bind RNA and transit through the nuclear pore, perhaps as a large complex (5, 7, 24). Many transcripts, after undergoing splicing, appear to access a cellular export pathway that appears inaccessible to intronless transcripts. Recent work on RNA export of intronless transcripts, such as the TK mRNA from HSV-1, has identified cis-acting elements that can functionally substitute for introns in mediating RNA export (31, 50). Transcripts encoded by HSV are largely intronless and may be exported by both cellular and virus-encoded proteins. ICP27 appears to mediate the nucleocytoplasmic export of some intronless herpesvirus RNAs. Other HSV RNAs are independent of ICP27 for their export. They may contain elements similar to those found in the TK gene that function to increase their cytoplasmic accumulation. Here, we have also identified an SM motif in ICP27. We investigated whether this motif was involved in the host shutoff function of ICP27 because the SM motif maps to a region of ICP27 required for this activity (16, 66, 67). These motifs are utilized by SM proteins, whose seven members along with the U snRNAs form the core spliceosome (26). The identification of an SM motif in ICP27 may help elucidate the molecular mechanism responsible for the inhibition of cellular splicing. A critical residue in the SM motif of ICP27, thought to be required for core assembly of SM proteins, was changed. In the resulting mutant, host shutoff was impaired. Furthermore, the ts mutation alters another critical amino acid in the SM motif, and a virus with this mutation is also defective in host shutoff (75). We propose that ICP27 requires its SM motif to bind to the SM complex and inactivate its function in splicing. We are currently testing whether ICP27 binds to one or more of the SM proteins directly. Our model predicts that the SM domain of ICP27 is re-

2822


J. VIROL.

FIG. 8. Structural model of KH domains in ICP27. Ribbon diagrams of the KH1 and KH3 motifs from ICP27 (blue) were superimposed on the KH1 domain of Fmr-1 (yellow) using PrISM. Side chains are shown for F303, the site of the Fmr-1 mutation in KH1 of ICP27 (A); S334, the site of the suppressor mutation in KH1 (B); R480, the site of the ts mutation in KH3 (C); and V487 and V496, the sites of the two suppressor mutations in KH3 (D and E, respectively). Also shown is a side view of KH3 with the side chains of the same sites highlighted. The structure is based on the amino acid alignment in Fig. 1B, using the structure of the KH1 from Fmr-1 as a template.

quired to interact with the spliceosome. However, the ICP27 mutant tsR480H, which is defective in host shutoff, was reported to colocalize with splicing factors in infected cells (67). We propose that ICP27 may interact with more than one SM protein. Thus, a mutation in any critical amino acid of the SM motif may have the effect of disrupting the interaction with one SM protein without disrupting the interaction with another SM protein. This hypothesis is supported by data on the interactions between SM proteins presented by Camasses et al. (6) and can be explained by the fact that a single SM protein can interact simultaneously with two SM proteins (26). Each SM motif creates two ␤ strands that are used to interact with two SM partners. Because each SM protein contains these strands, a daisy chain of the seven SM proteins forms. This structure is thought to create a doughnut with a positively charged center that interacts with RNA. The mutation in tsR480H, which disrupts one of the critical amino acids in the SM motif of ICP27, may only disrupt the interaction of ICP27 with one SM protein, leaving the other ␤ strand available for interaction. Thus, ICP27 interacting with only one SM protein may still colocalize with splicing factors, but this may not be sufficient to disrupt the function of the SM proteins in pre-mRNA splicing. Molecular dissection of ICP27’s functions reveals distinct roles for host shutoff and RNA export in regulating HSV gene expression. To determine their respective contributions, we analyzed accumulation of viral proteins in the absence of ICP27’s functions in host shutoff, RNA export, or both. Infected cells which continue to synthesize host proteins show reduced accumulation of immediate-early and early proteins and a delay in the appearance of late proteins. In contrast, only the expression of late genes is affected by blocking RNA export by ICP27, supporting our hypothesis that these functions are distinct. In cells infected with tsR480H, the profile of viral proteins synthesized appears consistent with a defect in both functions of ICP27. Thus, ICP27 increases virus protein synthesis by performing two functions, the export of late HSV RNA and the shutoff of host protein synthesis. However, these results do not fully explain all of ICP27’s functions. While tsR480H is unable to mediate host shutoff or export HSV RNA, replication of HSV DNA is not affected at the restrictive temperature (data not shown). This suggests that, as demonstrated for the UL9 protein, early intronless RNAs coding for DNA replication proteins may not require ICP27 for their export. In contrast, ICP27 deletion viruses or mutants in which

the acidic amino terminus of ICP27 is deleted are defective in HSV DNA replication (33, 60). Thus, other functions of ICP27 have yet to be elucidated. Virus gene expression is enhanced via two functions of ICP27, RNA export and host shutoff. These activities benefit HSV because most virus-specified RNAs are intronless. When splicing of cellular pre-mRNA is inhibited, intronless RNAs are afforded an advantage. However, transport of intronless RNAs is inherently less efficient (31). ICP27 compensates for this inefficiency by serving as an exporter of viral intronless RNAs. Conservation of the KH and SM motifs among the various ICP27 homologs from other alphaherpesviruses suggests that these two activities, inhibition of splicing and RNA export, may be an important regulatory feature of the homologs (Fig. 1). These functions are also reported for the more distant ICP27 homolog, the SM protein from Epstein-Barr virus (3, 27, 63, 69). The virus carrying the ts mutation, tsR480H, is defective in RNA binding, shuttling, and host shutoff. This complex phenotype results from the single amino acid change in the overlapping SM and KH3 motifs that affects the function of both domains. To explain the observation that this mutation blocks the nuclear export of ICP27, we propose that RNA binding is a prerequisite for nuclear export of ICP27. We show that limiting the transcription of late RNA, by blocking DNA replication in two distinct ways, severely reduces the accumulation of ICP27 in the cytoplasm (Fig. 3). These studies reveal a direct correlation between the nuclear export of ICP27, transcription of late RNA, and RNA binding. In support of this model, three independent suppressors of the ts mutation, which restore shuttling of ICP27, were found to map to the KH-like domains (75). We previously hypothesized that ICP27’s shuttling must be regulated because the protein shuttles only at late times postinfection (75). Based on that finding, we proposed that the signals that mediate import and export of ICP27 may be sensitive to whether it is bound to RNA. The model predicts that the NES of ICP27 is masked in the nucleus until the protein binds its RNA cargo. Binding of RNA results in exposure of the NES and sequestering of the NLS. This permits nuclear export of the RNA-protein complex. Once in the cytoplasm, ICP27 releases its cargo, resulting in a protein with an accessible NLS. This process allows ICP27 to return to the nucleus for further rounds of RNA export. Consistent with this hypothesis are our

VOL. 74, 2000


2823

FIG. 9. Model of ICP27 functions. The model shows three pathways for mRNA export. In the first pathway, cellular transcripts require splicing before export. ICP27 binds to the spliceosome through its SM domain and inhibits splicing, resulting in nuclear retention of those transcripts. In the second pathway, cellular intronless transcripts and some virus-specified intronless transcripts access an unidentified cellular pathway to undergo export. In the third pathway, ICP27 mediates the export of other virus-specified intronless transcripts. In binding RNA through its KH domains, it undergoes a conformational change that exposes its NES for binding by the nuclear export adapter Crm1 and is exported via Ran-GTP.

observations that correlate RNA binding with ICP27’s trafficking. While we have found that ICP27 exits the nucleus only after transcription of HSV late RNA, others have reported different results. Our initial observation demonstrated that transiently expressed ICP27 required coexpression of a late RNA for nuclear export (75). It has since been reported that in some assays, transiently expressed ICP27 can be exported from the nucleus without the requirement for other viral factors (39, 65). We have investigated this apparent discrepancy and found that ICP27, when overexpressed in transfected cells, no longer requires a cofactor to shuttle. While it is clear that a mutation in the RNA binding domain of ICP27 blocks nuclear export of ICP27, it is likely that wild-type ICP27 in the absence of late RNA is unable to shuttle because of the lack of cargo. It is conceivable that outside the context of infection some cellular RNAs may serve as cargo for ICP27, and this may explain the ability of ICP27 to shuttle when transiently expressed. An alternative explanation of these differences requires asking why ICP27 is retained in the nucleus in the absence of cargo. ICP27 might be retained in the nucleus by a protein that masks its NES. The binding of RNA by ICP27 may release the inhibitory protein and allow nuclear export. In cells where ICP27 is overexpressed, the inhibitory molecule may be limiting; thus, some ICP27 molecules are allowed to exit the nucleus. While this hypothesis resolves the discrepancies, further analysis is required to determine how ICP27’s shuttling is regulated. To determine where the point mutations may lie in the three-dimensional structure of the KH motifs in ICP27, we generated a structural model using the sequence-to-structure alignment program PrISM (protein informatics system for modeling) (81). The resulting structure is based on the KH

sequence alignment (Fig. 1B) and on the nuclear magnetic resonance structure for the KH1 motif of Fmr-1 (41). The positioning of the ts and suppressor mutations in KH3 is along one face of this domain (Fig. 8). From the predicted structure, we posit that there is an interaction interface in KH3 that could be weakened by the ts mutation at the restrictive temperature and stabilized by the suppressor mutations in this domain. The hydrophobic amino acid sequence at the putative interface suggests a protein-protein rather than a protein-nucleic acid interaction. The position of the suppressor mutation in KH1 is also shown. This mutation may lie along the corresponding interface occupied by the position of the ts and suppressor mutations in KH3. One interpretation of this model is that the KH1 and KH3 domains interact with each other. These interactions could be important to form a structure that can interact with RNA. Alternatively, the KH domains might interact with another protein. Consistent with this hypothesis, ICP27 has recently been shown to interact with hnRNP K, a cellular protein that also contains three KH domains (79). It is tempting to speculate that the overlapping KH and SM motifs, involved in two independent functions of ICP27, dictate the conformation of the protein. Perhaps in one conformation, ICP27 can bind RNA and expose the NES to permit export via Crm1. In this conformation, the SM domain may be masked. In the alternative conformation, the SM domain is available to inhibit pre-mRNA splicing and shut off host protein synthesis. The separate activities of these overlapping domains became evident with the characterization of the suppressors of the ts mutation. The suppressor mutations map to the KH domains but do not restore the host shutoff function of ICP27. The identification of overlapping motifs with independent functions begins to explain the molecular basis for the ts phenotype.

2824


The rationale behind the emergence of these overlapping domains may stem from the requirement to regulate protein function through alternative conformations. Figure 9 presents a model for two functions of ICP27 and the two pathways for export of intronless viral RNAs. For the expression of most cellular genes and four HSV genes, splicing is required. ICP27 may bind to the core of the spliceosome through its SM motif or substitute for an SM protein in the core and inhibit splicing. The majority of HSV genes, and some cellular genes, are intronless and thus cannot access the RNA export pathway utilized by spliced RNAs. Some intronless virus-specified RNAs may be exported by a pathway accessed by intronless cellular RNAs; other HSV RNAs require ICP27 for their export. ACKNOWLEDGMENTS We thank Minoru Yoshida for generously providing LMB, Steve Goff for helpful discussions, and An-Suei Yang for the structural modeling. This study was supported by a grant from the Public Health Service (AI-33952) to Saul J. Silverstein. REFERENCES 1. Bird, R. M., A. V. Broadhurst, I. B. Duncan, M. J. Hall, R. W. Lambert, and P. Wong-Kai-In. 1986. Antiviral activity of 5⬘-PAA and 5⬘-PFA phosphate esters of 2⬘-deoxyuridines. J. Antimicrob. Chemother. 18(Suppl. B):201–205. 2. Bohenzky, R. A., A. G. Papavassiliou, I. H. Gelman, and S. Silverstein. 1993. Identification of a promoter mapping within the reiterated sequences that flank the herpes simplex virus type 1 UL region. J. Virol. 67:632–642. 3. Boyle, S. M., V. Ruvolo, A. K. Gupta, and S. Swaminathan. 1999. Association with the cellular export receptor CRM 1 mediates function and intracellular localization of Epstein-Barr virus SM protein, a regulator of gene expression. J. Virol. 73:6872–6881. 4. Brown, C. R., M. S. Nakamura, J. D. Mosca, G. S. Hayward, S. E. Straus, and L. P. Perera. 1995. Herpes simplex virus trans-regulatory protein ICP27 stabilizes and binds to 3⬘ ends of labile mRNA. J. Virol. 69:7187–7195. 5. Burd, C. G., and G. Dreyfuss. 1994. Conserved structures and diversity of functions of RNA-binding proteins. Science 265:615–621. 6. Camasses, A., E. Bragado-Nilsson, R. Martin, B. Se´raphin, and R. Bordonne´. 1998. Interactions within the yeast Sm core complex: from proteins to amino acids. Mol. Cell. Biol. 18:1956–1966. 7. Dreyfuss, G., M. Hentze, and A. I. Lamond. 1996. From transcript to protein. Cell 85:963–972. 8. Everett, R. D. 1986. The products of herpes simplex virus type 1 (HSV-1) immediate early genes 1, 2, and 3 can activate HSV-1 gene expression in trans. J. Gen. Virol. 68:2507–2513. 9. Everett, R. D. 1984. Trans-activation of transcription by herpes virus products: requirement for two HSV-1 immediate-early polypeptides for maximum activity. EMBO J. 3:3135–3141. 10. Fenwick, M. L., and M. J. Walker. 1982. Early and delayed shut-off of host protein synthesis in cells infected with HSV. J. Gen. Virol. 61:121–125. 11. Fornerod, M., M. Ohno, M. Yoshida, and I. W. Mattaj. 1997. CRM1 is an export receptor for leucine-rich nuclear export signals. Cell 90:1051–1060. 12. Fukuda, M., S. Asano, T. Nakamura, M. Adachi, M. Yoshida, M. Yanagida, and E. Nishida. 1997. CRM1 is responsible for intracellular transport mediated by the nuclear export signal. Nature 390:308–311. 13. Gelman, I. H., and S. Silverstein. 1986. Coordinate regulation of herpes simplex virus gene expression is mediated by the functional interaction of two immediate early gene products. J. Mol. Biol. 191:395–409. 14. Gelman, I. H., and S. Silverstein. 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. 15. Hardwicke, M. A., and R. M. Sandri-Goldin. 1994. The herpes simplex virus regulatory protein ICP27 contributes to the decrease in cellular mRNA levels during infection. J. Virol. 68:4797–4810. 16. Hardwicke, M. A., P. J. Vaughan, R. E. Sekulovich, R. O’Conner, and R. M. Sandri-Goldin. 1989. The regions important for the activator and repressor functions of herpes simplex virus type 1 ␣ protein ICP27 map to the Cterminal half of the molecule. J. Virol. 63:4590–4602. 17. Hardy, W. R., and R. M. Sandri-Goldin. 1994. Herpes simplex virus inhibits host cell splicing, and regulatory protein ICP27 is required for this effect. J. Virol. 68:7790–7799. 18. Hermann, H., P. Fabrizio, V. A. Raker, K. Foulaki, H. Hornig, H. Brahms, and R. Luhrmann. 1995. snRNP Sm proteins share two evolutionarily conserved sequence motifs which are involved in Sm protein-protein interactions. EMBO J. 14:2076–2088.

J. VIROL. 19. Holland, L. E., K. P. Anderson, C. Shipman, Jr., and E. K. Wagner. 1980. Viral DNA synthesis is required for the efficient expression of specific herpes simplex virus type 1 mRNA species. Virology 101:10–24. 20. Honess, R. W., and B. Roizman. 1974. Regulation of herpesvirus macromolecular synthesis. I. Cascade regulation of the synthesis of three groups of viral proteins. J. Virol. 14:8–19. 21. Honess, R. W., and B. Roizman. 1975. Regulation of herpesvirus macromolecular synthesis: sequential transition of polypeptide synthesis requires functional viral polypeptides. Proc. Natl. Acad. Sci. USA 72:1276–1280. 22. Honess, R. W., and D. H. Watson. 1977. Herpes simplex virus resistance and sensitivity to phosphonoacetic acid. J. Virol. 21:584–600. 23. Ingram, A., A. Phelan, J. Dunlop, and J. B. Clements. 1996. Immediate early protein IE63 of herpes simplex virus type 1 binds RNA directly. J. Gen. Virol. 77:1847–1851. 24. Izaurralde, E., and I. W. Mattaj. 1995. RNA export. Cell 81:153–159. 25. Johnson, P. A., and R. D. Everett. 1986. DNA replication is required for abundant expression of a plasmid-borne late US11 gene of herpes simplex virus type 1. Nucleic Acids Res. 14:3609–3625. 26. Kambach, C., S. Walke, R. Young, J. M. Avis, E. de la Fortelle, V. A. Raker, R. Luhrmann, J. Li, and K. Nagai. 1999. Crystal structures of two Sm protein complexes and their implications for the assembly of the spliceosomal snRNPs. Cell 96:375–387. 27. Key, S. C., T. Yoshizaki, and J. S. Pagano. 1998. The Epstein-Barr virus (EBV) SM protein enhances pre-mRNA processing of the EBV DNA polymerase transcript. J. Virol. 72:8485–8492. 28. Kudo, N., S. Khochbin, K. Nishi, K. Kitano, M. Yanagida, M. Yoshida, and S. Horinouchi. 1997. Molecular cloning and cell cycle-dependent expression of mammalian CRM1, a protein involved in nuclear export of proteins. J. Biol. Chem. 272:29742–29751. 29. Kudo, N., H. Taoka, T. Toda, M. Yoshida, and S. Horinouchi. 1999. A novel nuclear export signal sensitive to oxidative stress in the fission yeast transcription factor pap1. J. Biol. Chem. 274:15151–15158. 30. Kudo, N., B. Wolff, T. Sekimoto, E. P. Schreiner, Y. Yoneda, M. Yanagida, S. Horinouchi, and M. Yoshida. 1998. Leptomycin B inhibition of signal-mediated nuclear export by direct binding to CRM1. Exp. Cell Res. 242:540– 547. 31. Liu, X., and J. E. Mertz. 1995. HnRNP L binds a cis-acting RNA sequence element that enables intron-independent gene expression. Genes Dev. 9:1766–1780. 32. Mavromara-Nazos, P., and B. Roizman. 1987. Activation of herpes simplex virus 1 gamma 2 genes by viral DNA replication. Virology 161:593–598. 33. McCarthy, A. M., L. McMahan, and P. A. Schaffer. 1989. Herpes simplex virus type 1 ICP27 deletion mutants exhibit altered patterns of transcription and are DNA deficient. J. Virol. 63:18–27. 34. McGregor, F., A. Phelan, J. Dunlop, and J. B. Clements. 1996. Regulation of herpes simplex virus poly(A) site usage and the action of immediate-early protein IE63 in the early-late switch. J. Virol. 70:1931–1940. 35. McLauchlan, J., A. Phelan, C. Loney, R. M. Sandri-Goldin, and J. B. Clements. 1992. Herpes simplex virus IE63 acts at the posttranscriptional level to stimulate viral mRNA 3⬘ processing. J. Virol. 66:6939–6945. 36. McLauchlan, J., S. Simpson, and J. B. Clements. 1989. Herpes simplex virus induces a processing factor that stimulates poly(A) site usage. Cell 59:1093– 1105. 37. McMahan, L., and P. A. Schaffer. 1990. Repressing and enhancing functions of the herpes simplex virus regulatory protein ICP27 map to C-terminal regions and are required to modulate viral gene expression very early in infection. J. Virol. 64:3471–3485. 38. Mears, W. E., V. Lam, and S. A. Rice. 1995. Identification of nuclear and nucleolar localization signals in the herpes simplex virus regulatory protein ICP27. J. Virol. 69:935–947. 39. Mears, W. E., and S. A. Rice. 1998. The herpes simplex virus immediate-early protein ICP27 shuttles between nucleus and cytoplasm. Virology 242:128– 137. 40. Mears, W. E., and S. A. Rice. 1996. The RGG box motif of the herpes simplex virus ICP27 protein mediates an RNA-binding activity and determines in vivo methylation. J. Virol. 70:7445–7453. 41. Musco, G., A. Kharrat, G. Stier, F. Fraternali, T. J. Gibson, M. Nilges, and A. Pastore. 1997. The solution structure of the first KH domain of FMR1, the protein responsible for the fragile X syndrome. Nat. Struct. Biol. 4:712–716. 42. Musco, G., G. Stier, C. Joseph, M. A. Castiglione Morelli, M. Nilges, T. J. Gibson, and A. Pastore. 1996. Three-dimensional structure and stability of the KH domain: molecular insights into the fragile X syndrome. Cell 85: 237–245. 43. Neville, M., and M. Rosbash. 1999. The NES-Crm1p export pathway is not a major mRNA export route in Saccharomyces cerevisiae. EMBO J. 18: 3746–3756. 44. Nishi, K., M. Yoshida, D. Fujiwara, M. Nishikawa, S. Horinouchi, and T. Beppu. 1994. Leptomycin B targets a regulatory cascade of crm1, a fission yeast nuclear protein, involved in control of higher order chromosome structure and gene expression. J. Biol. Chem. 269:6320–6324. 45. Nishioka, Y., and S. Silverstein. 1977. Degradation of cellular mRNA during

VOL. 74, 2000 infection by herpes simplex virus. Proc. Natl. Acad. Sci. USA 74:2370–2374. 46. Nishioka, Y., and S. Silverstein. 1978. Requirement of protein synthesis for the degradation of host mRNA in Friend erythroleukemia cells infected with herpes simplex virus type 1. J. Virol. 27:619–627. 47. O’Hare, P., and G. S. Hayward. 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. 48. O’Hare, P., and G. S. Hayward. 1985. Three trans-acting regulatory proteins of herpes simplex virus modulate immediate-early gene expression in a pathway involving positive and negative feedback regulation. J. Virol. 56: 723–733. 49. Ossareh-Nazari, B., F. Bachelerie, and C. Dargemont. 1997. Evidence for a role of CRM1 in signal-mediated nuclear protein export. Science 278:141– 144. 50. Otero, G. C., and T. J. Hope. 1998. Splicing-independent expression of the herpes simplex virus type 1 thymidine kinase gene is mediated by three cis-acting RNA subelements. J. Virol. 72:9889–9896. 51. Panagiotidis, C. A., E. K. Lium, and S. J. Silverstein. 1997. Physical and functional interactions between herpes simplex virus immediate-early proteins ICP4 and ICP27. J. Virol. 71:1547–1557. 52. Phelan, A., M. Carmo-Fonseca, J. McLauchlan, A. I. Lamond, and J. B. Clements. 1993. A herpes simplex virus type 1 immediate-early gene product, IE63, regulates small nuclear ribonucleoprotein distribution. Proc. Natl. Acad. Sci. USA 90:9056–9060. 53. Phelan, A., and J. B. Clements. 1997. Herpes simplex virus type 1 immediate early protein IE63 shuttles between nuclear compartments and the cytoplasm. J. Gen. Virol. 78:3327–3331. 54. Phelan, A., J. Dunlop, and J. B. Clements. 1996. Herpes simplex virus type 1 protein IE63 affects the nuclear export of virus intron-containing transcripts. J. Virol. 70:5255–5265. 55. Pinol-Roma, S., S. A. Adam, Y. D. Choi, and G. Dreyfuss. 1989. Ultravioletinduced cross-linking of RNA to proteins in vivo. Methods Enzymol. 180: 410–418. 56. Pollard, V. W., and M. H. Malim. 1998. The HIV-1 Rev protein. Annu. Rev. Microbiol. 52:491–532. 57. Purifoy, D. J., and K. L. Powell. 1977. Herpes simplex virus DNA polymerase as the site of phosphonoacetate sensitivity: temperature-sensitive mutants. J. Virol. 24:470–477. 58. Rice, S. A., and D. M. Knipe. 1990. Genetic evidence for two distinct transactivation functions of the herpes simplex virus ␣ protein ICP27. J. Virol. 64:1704–1715. 59. Rice, S. A., and V. Lam. 1994. Amino acid substitution mutations in the herpes simplex virus ICP27 protein define an essential gene regulation function. J. Virol. 68:823–833. 60. Rice, S. A., V. Lam, and D. M. Knipe. 1993. The acidic amino-terminal region of herpes simplex virus type 1 alpha protein ICP27 is required for an essential lytic function. J. Virol. 67:1778–1787. 61. Rice, S. A., S. Su, and D. M. Knipe. 1989. Herpes simplex virus alpha protein ICP27 possesses separable positive and negative regulatory activities. J. Virol. 63:3399–3407. 62. Roizman, B., and A. Sears. 1993. The human herpes viruses: biology, pathogenesis, and treatment. Raven Press, New York, N.Y. 63. Ruvolo, V., E. Wang, S. Boyle, and S. Swaminathan. 1998. The Epstein-Barr virus nuclear protein SM is both a post-transcriptional inhibitor and activator of gene expression. Proc. Natl. Acad. Sci. USA 95:8852–8857. 64. Sacks, W. R., C. C. Greene, D. P. Aschman, and P. A. Schaffer. 1985. Herpes simplex virus type 1 ICP27 is an essential regulatory protein. J. Virol. 55: 796–805.


2825

65. Sandri-Goldin, R. M. 1998. ICP27 mediates HSV RNA export by shuttling through a leucine-rich nuclear export signal and binding viral intronless RNAs through an RGG motif. Genes Dev. 12:868–879. 66. Sandri-Goldin, R. M., and M. K. Hibbard. 1996. The herpes simplex virus type 1 regulatory protein ICP27 coimmunoprecipitates with anti-Sm antiserum, and the C terminus appears to be required for this interaction. J. Virol. 70:108–118. 67. Sandri-Goldin, R. M., M. K. Hibbard, and M. A. Hardwicke. 1995. The C-terminal repressor region of herpes simplex virus type 1 ICP27 is required for the redistribution of small nuclear ribonucleoprotein particles and splicing factor SC35; however, these alterations are not sufficient to inhibit host cell splicing. J. Virol. 69:6063–6076. 68. Sandri-Goldin, R. M., and G. E. Mendoza. 1992. A herpesvirus regulatory protein appears to act post-transcriptionally by affecting mRNA processing. Genes Dev. 6:848–863. 69. Semmes, O. J., L. Chen, R. T. Sarisky, Z. Gao, L. Zhong, and S. D. Hayward. 1998. Mta has properties of an RNA export protein and increases cytoplasmic accumulation of Epstein-Barr virus replication gene mRNA. J. Virol. 72:9526–9534. 70. Seraphin, B. 1995. Sm and Sm-like proteins belong to a large family: identification of proteins of the U6 as well as the U1, U2, U4 and U5 snRNPs. EMBO J. 14:2089–2098. 71. Showalter, S. D., M. Zweig, and B. Hampar. 1981. Monoclonal antibodies to herpes simplex virus type 1 proteins, including the immediate-early protein ICP 4. Infect. Immun. 34:684–692. 72. Siomi, H., M. Choi, M. C. Siomi, R. L. Nussbaum, and G. Dreyfuss. 1994. Essential role for KH domains in RNA binding: impaired RNA binding by a mutation in the KH domain of FMR1 that causes fragile X syndrome. Cell 77:33–39. 73. Siomi, H., M. J. Matunis, W. M. Michael, and G. Dreyfuss. 1993. The pre-mRNA binding K protein contains a novel evolutionarily conserved motif. Nucleic Acids Res. 21:1193–1198. 74. Smith, I. L., M. A. Hardwicke, and R. M. Sandri-Goldin. 1992. Evidence that the herpes simplex virus immediate early protein ICP27 acts post-transcriptionally during infection to regulate gene expression. Virology 186:74–86. 75. Soliman, T. M., R. M. Sandri-Goldin, and S. J. Silverstein. 1997. Shuttling of the herpes simplex virus type 1 regulatory protein ICP27 between the nucleus and cytoplasm mediates the expression of late proteins. J. Virol. 71:9188–9197. 76. Sydiskis, R. J., and B. Roizman. 1966. Polysomes and protein synthesis in cells infected with a DNA virus. Science 153:76–78. 77. Uprichard, S. L., and D. M. Knipe. 1996. Herpes simplex ICP27 mutant viruses exhibit reduced expression of specific DNA replication genes. J. Virol. 70:1969–1980. 78. Vaughan, P. J., K. J. Thibault, M. A. Hardwicke, and R. M. Sandri-Goldin. 1992. The herpes simplex virus immediate early protein ICP27 encodes a potential metal binding domain and binds zinc in vitro. Virology 189:377– 384. 79. Wadd, S., H. Bryant, O. Filhol, J. E. Scott, T. Y. Hsieh, R. D. Everett, and J. B. Clements. 1999. The multifunctional herpes simplex virus IE63 protein interacts with heterogeneous ribonucleoprotein K and with casein kinase 2. J. Biol. Chem. 274:28991–28998. 80. Wen, W., J. L. Meinkoth, R. Y. Tsien, and S. S. Taylor. 1995. Identification of signal for rapid nuclear export of proteins from the nucleus. Cell 82:463– 473. 81. Yang, A. S., and B. Honig. 1999. Sequence to structure alignment in comparative modeling using PrISM. Proteins 37:66–72.