Specific Interaction of Heterogeneous Nuclear Ribonucleoprotein Particle U with the Leader RNA Sequence of Vesicular Stomatitis Virus

Abstract

The 3′ ends of the genome and antigenome RNA of vesicular stomatitis virus (VSV) serve as the promoter sites for the RNA-dependent RNA polymerase in the initiation of transcription and replication, respectively. The leader RNA, the first transcript synthesized during the RNA synthetic step, contains sequences to initiate encapsidation with the nucleocapsid protein, which is a prerequisite for replication. It also plays a role in the inhibition of cellular RNA synthesis. To search for a specific cellular factor(s) which may interact with the leader RNA sequences and regulate these processes, we used a gel mobility shift assay to identify such a protein(s). By using nuclear extract, it was found that in addition to the previously reported La protein, a 120-kDa nuclear protein specifically interacts with the leader RNA. Biochemical and immunological studies identified the 120-kDa protein as heterogeneous nuclear ribonucleoprotein particle U (hnRNP U), which is involved in pre-mRNA processing. We also demonstrate that hnRNP U is associated with the leader RNA in the nuclei of VSV-infected cells and also packaged within the purified virions. By double immunofluorescence labeling and confocal microscopy, hnRNP U appears to colocalize with the virus in the cytoplasm of infected cells. These results strongly suggest that hnRNP U plays an important role in the life cycle of VSV.

When a virus infects a cell, one of the hallmarks of the process is the recruitment by the virus of specific cellular proteins for its replicative advantage. Viruses interact with such cellular proteins primarily to aid their own multiplication. Viruses also shut off cellular functions by sequestering or inhibiting synthesis of vital cellular proteins for their own replicative advantage. Vesicular stomatitis virus (VSV), a prototype rhabdovirus, is a paradigm for studying such host-virus interactions. VSV contains a negative-strand RNA genome 11,161 nucleotides (nt) long which, when transcribed by a virion-associated RNA polymerase, synthesizes in vitro or in vivo five monocistronic messages in the following order: 3′ nucleocapsid protein (N), phosphoprotein (P), glycoprotein (G), matrix protein (M), and the RNA polymerase (L) 5′ (1). The RNA-dependent RNA polymerase consists of two subunits, L and P. It first synthesizes a 47-nucleotide leader RNA and then sequentially synthesizes five mRNAs that are capped and polyadenylated (1, 2). During replication, however, the RNA polymerase first synthesizes the full-length plus-sense antigenome which is enwrapped with the N protein, forming the N-RNA complex; this complex then serves as the template for the synthesis of the negative-sense progeny genome RNA (1, 2). It is envisaged that the N protein complexes with the nascent leader RNA transcript to initiate encapsidation (1, 3–5, 12) of the growing RNA chains, leading to the replicative reaction. It still remains unclear how the RNA polymerase switches its transcription mode and enters the replicative mode. Several recent studies suggest that the L protein may associate with the N-P complex, a prerequisite entity for the replicative event, and the resulting tripartite complex along with a specific host protein(s) may initiate the replicative reaction on the N-RNA template (6, 13).

It is generally believed that the 3′-terminal RNA sequence of the genome RNA is the binding site of the VSV RNA polymerase (2, 14, 15) to initiate transcription. Thus, the 3′-terminal domain of the genome RNA and its complement (leader-sense [LS]) RNA are the two important cis-acting RNA sequences that are potential targets for cellular proteins to bind and promote the functions of the transcriptase and the replicase, respectively. Keene et al. (18, 19, 25) have shown previously that both plus-strand and minus-strand leader RNA (the complement of the 3′-terminal sequence of the plus-strand genome RNA) interact specifically with the nuclear autoantigen, La protein, in infected-cell cytoplasm, raising the possibility that this interaction may have some specific role in the replicative pathway of the virus. Moreover, in view of the similarity in the sequences of RNA polymerase III products and the 3′ end of the leader RNA, it seems that the interaction of La protein with the leader RNA may be mediated by a sequence motif which regulates VSV transcription and replication. In a separate series of studies, the leader RNA of VSV was implicated in inhibiting cellular RNA synthesis by its transient localization inside the nucleus following infection (19, 24), suggesting that it may interact with specific nuclear proteins involved in RNA synthesis. To test that directly, McGowan et al. (20), using a soluble cell extract as the source of the RNA polymerase, showed that purified leader RNA indeed inhibits DNA-dependent transcription of adenovirus and simian virus 40 genes in vitro. Since leader RNA specifically interacts with the nuclear autoantigen La, the possibility that this interaction may have some role in the inhibition of RNA polymerase II activity exists, although such interaction within the nucleus has not yet been reported. So far, no other cellular proteins involved in transcription or posttranscription steps of mRNA synthesis have been shown to interact with the VSV leader RNA.

In the present study, we searched for an additional cellular factor(s) which specifically interacts with VSV leader RNA sequences. Using a gel mobility shift assay with nuclear extract, we showed that in addition to La protein, a nuclear protein, heterogeneous nuclear ribonucleoprotein particle U (hnRNP U), specifically binds to the VSV leader RNA both in vitro and in vivo. Moreover, colocalization of hnRNP U with VSV in the infected-cell cytoplasm, coupled with its packaging within the purified virions, raises an interesting possibility of its direct involvement in the life cycle of VSV and in virus-induced pathogenesis.

MATERIALS AND METHODS

Materials.

All enzymes and biochemicals were obtained either from Boehringer Mannheim or from Sigma. Radionucleotides were purchased from Amersham. Tissue culture reagents and media were obtained from Life Technologies, Inc.

Cell cultures and virus.

HeLa cells were grown in Eagle’s minimal medium containing 5% fetal calf serum. VSV, Indiana serotype, Mudd Summers strain, was used for infecting the cells.

Cloning of the LS RNA construct and in vitro transcription.

The LS RNA construct was made in the pUC19 vector, which contains the oligodeoxynucleotides corresponding to the first 60 nt from the 5′ end of plus-sense RNA, under the control of T7 RNA polymerase. The construct was made essentially by reverse transcriptase-PCR amplification of the genome RNA isolated from purified virions of VSV, Indiana serotype. The primer used for the reverse transcriptase reaction for making LS RNA was 5′CCGGAATTCTAATACGACTCACTATAGGACGAAGACAAACCCATTA3′, which in addition to viral sequences contains an EcoRI site and T7 RNA polymerase promoter sequences. PCR amplification was carried out with the addition of a second primer, 5′TGCACTGCAGATTACTGTTAAAGTTTCTCC3′, which contains a PstI site. Digestion of the recombinant pUC19 vector with PstI followed by a Klenow reaction in the absence of deoxynucleotides would generate the exact 3′ end for LS RNA synthesized by T7 RNA polymerase. In vitro transcription reaction was carried out with T7 RNA polymerase and [α-³²P]UTP according to the manufacturer’s protocol (Boehringer Mannheim). The radiolabeled RNA was analyzed in a 10% polyacrylamide-urea gel, and the RNA band was excised and eluted in a buffer containing 0.5 M ammonium acetate, 1 mM EDTA, and 0.1% sodium dodecyl sulfate (SDS) and purified by phenol-chloroform extraction followed by ethanol precipitation.

Gel mobility shift assay.

The binding of radiolabeled leader RNA with the cellular proteins was carried out in a 20-μl binding buffer containing 15 mM HEPES (pH 7.5), 15 mM KCl, 0.25 mM EDTA, 0.25 mM dithiothreitol, 5 mM MgCl₂, 1 mM phenylmethylsulfonyl fluoride, 200 μg of yeast tRNA per ml, 10% glycerol, 0.1 ng of labeled RNA, and, unless otherwise mentioned, cell extract containing 2.5 μg of protein. Incubations were done at room temperature for 30 min, and the samples were analyzed in a 5% native polyacrylamide gel. The gels were run at 130 V, dried, and subjected to autoradiography (7).

UV cross-linking.

Binding of the radiolabeled RNA with the cellular proteins was carried out in binding buffer at room temperature for 30 min. The reaction mixtures were then exposed to short-wavelength UV light at a 4-cm distance for 1 h in ice. The UV-cross-linked mixture was then treated with RNase A (0.1 μg) for 15 min at 37°C. Samples were run in a 10% SDS-polyacrylamide gel, followed by staining, drying, and autoradiography.

Immunoprecipitation of VSV RNAs.

In vitro, radiolabeled LS RNA was allowed to bind with either nuclear extract from HeLa cells or the bacterially expressed hnRNP U (U protein). Monoclonal antibody to U protein (3G6) was then added to the reaction mixture and incubated for 30 min followed by protein A-Sepharose binding. Immunoprecipitates were washed in HEPES buffer (pH 7.5) containing 0.1% Nonidet P-40 and 250 mM NaCl, followed by suspension in 100 μl of Tris-EDTA buffer (pH 8.0), phenol-chloroform extraction, and ethanol precipitation. Products were analyzed in a urea–10% polyacrylamide gel and subjected to autoradiography. Nucleoplasm from VSV-infected HeLa cells was isolated according to the method of Pinol-Roma et al. (22). Immunoprecipitation of U-protein-bound nucleic acid was carried out with the 3G6 monoclonal antibody. Immunoprecipitates were then washed and purified as described above. The immunoprecipitated RNAs were annealed with radiolabeled LS RNA complementary probe and subjected to RNase T2 digestion followed by phenol-chloroform extraction and ethanol precipitation. The RNA samples were run in a urea–10% polyacrylamide gel followed by autoradiography.

Immunofluorescent labeling.

HeLa cells were grown on coverslips and infected with VSV at 10 PFU/cell. At various times postinfection, the cells were washed with phosphate-buffered saline followed by fixation with 3.6% paraformaldehyde and permeabilization with 1% Triton X-100. The cells were then incubated with anti-VSV P protein (rabbit polyclonal) and monoclonal antibody to hnRNP U, either separately or together. For labeling, the coverslips were washed and incubated with rhodamine-conjugated anti-rabbit immunoglobulin G (IgG) and fluorescein isothiocyanate (FITC)-tagged anti-mouse IgG secondary antibodies for virus and hnRNP U, respectively. For double staining, both the secondary antibodies were added together. The coverslips were finally washed, mounted, and then examined with a Leica confocal laser scanning microscope.

RESULTS

Interaction of cellular proteins with VSV LS RNA.

To identify the cellular proteins that bind to VSV LS RNA, we inserted the cDNA copies of the first 60 nt complementary to the 3′ end of the genome RNA (LS RNA) (Fig. 1A) in a pUC19 vector along with a T7 RNA polymerase promoter sequence, as described in Materials and Methods. It is important to note that two nonviral 5′ G residues shown in a box are incorporated in the LS RNA during transcription from the T7 promoter. Thus, the LS RNA (Fig. 1A) includes sequences for a nucleation site(s) for binding of N protein at its 5′ end. In addition, the LS RNA includes the intergenic sequence between the leader RNA and the start site of the N gene spanning 13 nt. The latter sequence may interact with presumptive viral and cellular proteins required for replication.

FIG. 1.

Binding of nuclear proteins to LS RNA of VSV. (A) The sequence of LS RNA is shown. The two Gs shown in the box represent the nonviral sequence present at the 5′ end of the T7 transcript. The arrow in the sequence denotes the 47th nucleotide position from the 5′ end, which is the exact size of the LS RNA. (B) Radiolabeled LS RNA was used in a gel mobility shift assay with (lane 2) and without (lane 1) the nuclear extract of HeLa cells. I and II are the two LS RNA-protein complexes. (C) UV cross-linking of LS RNA was carried out with semipurified nuclear extract. Lane 1, probe alone; lanes 2 and 3, UV-cross-linked products, respectively, of complex I- and II-forming fractions. UV cross-linking was followed by RNase digestion. Numbers on the right indicate the migration positions of standard molecular weight markers (weights are in thousands).

A T7 RNA polymerase-transcribed, ³²P-labeled LS RNA probe was used in a gel mobility shift assay following incubation with HeLa cell nuclear extracts. As shown in Fig. 1B, LS RNA formed two distinct and identifiable complexes (I and II) when incubated with the nuclear extract. These two bands were not present when cytoplasmic extract was used; instead, a new band whose identity has not been confirmed appeared (data not shown). In UV cross-linking experiments, the nuclear proteins associated with complexes I and II migrated in SDS-polyacrylamide gel electrophoresis consistent with molecular weights of 50,000 and 120,000, respectively. The characterization of the protein associated with complexes I and II is described below.

Analysis of the nuclear protein present in complex I.

The nuclear protein complexed with leader RNA in complex I appeared to be a 50-kDa protein as shown by UV cross-linking (Fig. 1C). Based on the fact that the La protein, also a 50-kDa protein, has previously been shown to interact specifically with VSV leader RNA (18), we used bacterially expressed La protein to establish the identity of the protein present in complex I. The La protein expressed in Escherichia coli and subsequently purified was mixed with ³²P-labeled VSV leader RNA, and the complex was analyzed in a gel mobility shift assay, as described in the legend to Fig. 1B. A distinct RNA-La complex migrated in the same position as complex I (Fig. 1B) in the gel mobility shift assay and when cross-linked by UV irradiation (Fig. 2A and B). Furthermore, immunoprecipitation of complex I with anti-La antibody resulted in the recovery of the ³²P LS RNA (Fig. 2C, lane 3). No ³²P LS RNA was precipitated when the probe was treated only with anti-La antibody and protein A-Sepharose as a control (Fig. 2C, lane 2). These results strongly suggest that the protein present in complex I is indeed the autoantigen La and confirm the previous observations made by Kurilla and Keene (18).

FIG. 2.

Binding of bacterially expressed La protein to LS RNA. Bacterially expressed La protein was incubated with the radiolabeled LS RNA, as described in Materials and Methods. (A) Gel mobility shift assays were done with (lane 2) and without (lane 1) the La protein. (B) UV cross-linking of the bacterially expressed La protein to LS RNA was followed by RNase I digestion. Lane 1, probe alone; lane 2, probe with La protein. Numbers on the right indicate the migration positions of molecular weight markers (numbers are in thousands). (C) Immunoprecipitation of LS RNA by anti-La antibody after incubation with (lane 3) or without (lane 2) La protein. Lane 1, LS RNA alone without binding and immunoprecipitation.

Analysis of the nuclear protein present in complex II.

In UV cross-linking experiments, the protein present in complex II was determined to have a molecular weight of 120,000 (Fig. 1C). The first step in characterizing the protein was to carry out a computer search for the RNA binding proteins from the protein data bank to obtain information on its estimated molecular weight and localization in the nucleus. This search yielded two proteins, namely, hnRNP U and nucleolin. Nucleolin was eliminated because antibody against nucleolin failed to immunoprecipitate LS RNA from complex II (data not shown). To test whether hnRNP U was the observed 120-kDa RNA binding protein, we first carried out competition experiments with ribonucleotide homopolymers to determine whether their properties matched the characteristic properties of hnRNP U. As shown in Fig. 3, poly(A) had no effect even in a 500-fold molar excess, whereas poly(U) in a 500-fold molar excess reduced the binding of 120-kDa protein to LS RNA by 80%. However, poly(G) competed out almost completely the binding of the 120-kDa protein to LS RNA even in a 250-fold molar excess. These findings are consistent with the known properties of hnRNP U observed previously (16, 17). As expected, in a control experiment, T7 RNA polymerase-transcribed, unlabeled LS RNA completely competed out the binding of this RNA to the protein. As shown in Fig. 3B, a heterologous RNA transcribed from the pGEM4A vector failed to compete with the binding of the 120-kDa protein with LS RNA, indicating specificity of this interaction.

FIG. 3.

Interaction of the 120-kDa protein by competition with different ribonucleotide homopolymer competitors and homologous and heterologous RNA probes. Binding of the 120-kDa protein to LS RNA in the presence of competitors was assayed by UV cross-linking followed by polyacrylamide gel electrophoresis. (A) The ribonucleotide homopolymer competitor used is shown at the top of each lane. The number at the top of each lane corresponds to the molar excess (fold) of competitor used in the reaction. (B) The number at the top of each lane corresponds to the molar excess (fold) of heterologous RNA (HL RNA) competitor used. Heterologous RNA was synthesized by T7 RNA polymerase from an EcoRI-digested pGEM4Z-vector DNA. In panels A and B, numbers on the sides denote the positions of migration of molecular weight markers (weights are in thousands). Lane 1, control (probe alone); lane 2, probe cross-linked with the 120-kDa protein in the absence of any competitor.

To confirm more definitively the identity of the 120-kDa protein, a cDNA clone of hnRNP U was expressed in E. coli, and the recombinant hnRNP U protein was tested for its ability to bind the LS RNA probe. As shown in Fig. 4A, bacterially expressed hnRNP U protein had a gel mobility shift pattern identical to that of the LS RNA bound to the 120-kDa protein from nuclear extract. Furthermore, this RNA-protein complex immunoprecipitated the bound leader RNA when a monoclonal antibody (3G6) raised against hnRNP U protein was used (10, 16). As shown in Fig. 4B, radiolabeled LS RNA was effectively immunoprecipitated when incubated in vitro either with nuclear extract (lane 3) or with recombinant hnRNP U (lane 5). No LS RNA was recovered in the control lanes where nuclear extract or antibody was omitted (lanes 2 and 4, respectively). Since hnRNP U protein is also known as a scaffold attachment factor A which binds to the scaffold-attachment-region element of genomic DNAs for its bending (11, 23), we investigated the competition of binding of hnRNP U protein to LS RNA with the MII fragment of the human topoisomerase I gene containing the specific DNA sequences which bind to scaffold attachment factor A. In a Northwestern blot experiment with purified HeLa nuclear extract, the binding of the 120-kDa protein to the LS RNA was competed out to about 80% when unlabeled MII DNA was added along with the probe in an 80-fold molar excess (data not shown). These results further established that the 120-kDa protein in the nuclear extract of HeLa cells is indeed hnRNP U.

FIG. 4.

Characterization of the 120-kDa protein. (A) LS RNA was incubated without (lane 1) or with (lanes 2 to 5) total nuclear extract made from HeLa cells, or with bacterially expressed recombinant hnRNP U protein (lanes 6 to 9), and subjected to a gel retardation assay. The concentrations of the nuclear extract used were 0.2, 0.4, 0.8, and 1.6 μg for lanes 2, 3, 4, and 5, respectively. The concentrations of recombinant hnRNP U were 25, 50, 100, and 200 ng, respectively for lanes 6, 7, 8, and 9. The arrow indicates the position of the complex formed by the 120-kDa protein. (B) Immunoprecipitation of the LS RNA with monoclonal antibody of hnRNP U (3G6) was done as described in Materials and Methods. NE, nuclear extract; Pr.A, protein A-Sepharose.

Association of hnRNP U with leader RNA in nuclei of VSV-infected cells.

Since it has been shown that following VSV infection leader RNA transiently enters the nucleus (19), we investigated whether hnRNP U also binds to LS RNA in the nuclei of VSV-infected cells. Monoclonal antibody to hnRNP U protein (3G6) was used to immunoprecipitate the nucleic acid bound to hnRNP U protein in the nuclear extract prepared from VSV-infected cells. Immunoprecipitates were then subjected to RNase protection assay with radiolabeled LS RNA complement as the probe. It can be seen in Fig. 5 that only in the immune complex isolated from VSV-infected nuclear extract (lane 5) was a 47-nt-long RNA fragment protected by the LS RNA complementary probe. We conclude from this experiment that (i) the hnRNP U-associated RNA within the nucleus of the cell is VSV specific and (ii) since the protected length of the RNA species is 47 nt and not 60 nt (the length of the LS RNA probe used), the LS RNA seems to be degraded to 47 nt. This result prompted us to determine whether the 47-nt-long leader RNA synthesized in vitro is sufficient to form a complex with hnRNP U. In addition, we carried out deletion analyses to determine which part of the LS RNA binds to the hnRNP U protein. Figure 6A shows the sequences of LS RNA and the deletion mutants used in our studies. It is apparent from Fig. 6B that bacterially expressed hnRNP U protein bound as effectively to LS 3′Δ13 RNA, which is the authentic 47-nt-long leader RNA, as the 60-nt-long LS RNA. Deletion of an additional 11 nt from the 3′ end of the leader RNA did not affect the binding efficiency. However, deletion of 15 nt from the 5′ end of LS RNA markedly decreased the efficiency of binding to the hnRNP U protein. Taken together, these results demonstrate that 47-nt-long leader RNA is sufficient to bind to hnRNP U and that the sequences at the 5′ end of the leader RNA are crucial for binding to the protein.

FIG. 5.

Association of hnRNP U with leader RNA in the nuclei of VSV-infected cells. Nucleoplasm from the VSV-infected HeLa cells was isolated 3 h postinfection. Monoclonal antibody against hnRNP U protein (3G6) was used to immunoprecipitate the protein-bound nucleic acids. Radiolabeled complement of LS RNA was used for RNase protection as described in Materials and Methods. NE, nuclear extract. Arrows denote the positions of 60- and 47-nt-long RNAs.

FIG. 6.

Determination of the region of hnRNP U binding to LS RNA. (A) The sequences of the wild-type LS RNA (LSwt) and the deletion mutants derived from it are shown. The two Gs present in a box at the 5′ end of each RNA sequence represent the nonviral sequence derived from T7 transcription. (B) The wild type and each deletion mutant were used for gel retardation either alone or with mock bacterial extract or extract expressing hnRNP U protein. Probes: lanes 1, wild type; lanes 2, 3′Δ24; lanes 3, 3′Δ13; lanes 4, 5′Δ15; and lanes 5, 5′Δ30.

Intracellular distribution of hnRNP U and VSV RNP.

We used indirect double immunofluorescence labeling and confocal microscopy to study colocalization, if any, of hnRNP U with the viral RNP in the infected cells. As described in detail in Materials and Methods, HeLa cells were infected with VSV on coverslips. At different times postinfection, the fixed cells were immunostained either singly or doubly with FITC-conjugated secondary antibody for hnRNP U and/or rhodamine-conjugated secondary antibody for VSV P protein. Colocalization of virus and hnRNP U was examined by confocal microscopy. As shown in Fig. 7A, hnRNP U is exclusively localized in the nucleus of uninfected cells. However, at 1.5 h postinfection, a small but distinct staining of hnRNP U was evidence in the cytoplasm (Fig. 7D) and, with double staining (Fig. 7F), there was the appearance of yellow color, signifying colocalization. At 3 h postinfection, increased staining of hnRNP U was observed in the cytoplasm with a concomitant increase in yellow color (Fig. 7G and I). Finally, at 6 h postinfection, a distinct green halo was seen surrounding the nucleus, suggesting further accumulation of hnRNP U in the cytoplasm (Fig. 7J). The strong yellow color followed the same pattern as that observed for virus and hnRNP U individually (Fig. 7J through L). It is important to note that at 6 h postinfection several uninfected cells did not release hnRNP U in the cytoplasm, suggesting that a distinct and discernible amount of hnRNP U either is retained in the cytoplasm following its synthesis or extrudes from the nucleus following VSV infection. Most importantly, the virus particles appear to remain always associated with hnRNP U in the cytoplasm.

FIG. 7.

Intracellular distribution of hnRNP U and viral RNP. HeLa cells were infected with VSV at 10 PFU/cell. At 1.5 h (D through F), 3 h (G through I), and 6 h (J through L) postinfection, cells were fixed and permeabilized as described in Materials and Methods. Coverslips were treated with anti-VSV-P polyclonal antibody and anti-hnRNP U monoclonal antibody, either separately or together. As a secondary antibody, FITC-conjugated anti-mouse antibodies (for hnRNP U) or rhodamine-conjugated anti-rabbit antibodies (for VSV-P protein) were used. (A) Uninfected cells treated with 3G6 antibody and FITC-tagged secondary antibody; (B) infected cells treated with anti-P antibody and rhodamine-conjugated secondary antibody; (C) uninfected cells treated with anti-P antibody followed by addition of secondary antibody for P protein and hnRNP U protein; (D to L) infected cells treated with both anti-P and anti-hnRNP U antibody followed by FITC- and rhodamine-tagged secondary antibody, respectively. A, D, G, and J show hnRNP U staining, and B, E, H, and K show virus staining. C, F, I, and L show colocalization of virus and hnRNP U staining.

Finally, we wanted to study whether such an interaction between hnRNP U and virus particles leads to packaging of hnRNP U in the mature virions. Figure 8 shows the results of Western blot analysis of purified RNP: hnRNP U was clearly discernible when blotted against 3G6 monoclonal antibody, indicating that hnRNP U interacts and remains associated with the RNP during the virus’s replicative cycle.

FIG. 8.

Western blot analysis of hnRNP U in purified virions. Total HeLa cell nuclear extract and purified virions of VSV (50 μg) were run in an SDS–10% polyacrylamide gel and transferred onto nitrocellulose membrane. The blot was developed with monoclonal antibody against hnRNP U followed by peroxidase-conjugated goat anti-mouse IgG. The antigen-antibody complex was detected by ECL reagent (Amersham Corp.). The numbers at the left denote the migration positions of molecular weight markers (weights are in thousands).

DISCUSSION

Leader RNA, the RNA product synthesized during initial transcription of VSV genome RNA, plays an important role in vivo in the synthesis of the full-length positive-strand genome RNA by providing the necessary cis-regulatory sequences for signaling encapsidation with the N protein (1). In addition, leader RNA appears to have a role in the inhibition of cellular RNA synthesis by its entry into the nucleus (19) and subsequent downregulation of polymerase II activity. To gain insight into these presumed activities of the leader RNA and the possible involvement of host proteins in these processes, we wanted to study the specific interactions, if any, of the cytoplasmic and nuclear proteins with the leader RNA.

By using a gel mobility shift assay with cell extracts, we were able to detect specific binding of host proteins with the leader RNA. One of the proteins was characterized as an autoantigen, La protein, which bound to both cytoplasmic extracts (data not shown) and nuclear extracts (Fig. 1 and 2). La protein was previously shown to bind with similar cis-acting RNA sequences of human parainfluenza virus type 3 (7), suggesting that both viruses contain common RNA sequences or structures at the 3′ end of the genome RNA and its complement which mediate specific binding with the La protein. The La protein was previously shown to bind leader RNA in VSV-infected cells (18). Based on the observation that leader RNA enters the nucleus during the early phase of infection (19), it was suggested that leader RNA may interact with the La protein, a protein associated with RNA-synthetic machinery in the nucleus, and may have a negative effect on cellular RNA synthesis (18, 19). It is important to note that the La protein also binds specifically to the 5′ untranslated region of poliovirus to regulate RNA translation (21). Thus, it seems that La protein may have pleotropic functions in the life cycle of certain RNA viruses.

In the present study, we have characterized an additional unique protein, identified as hnRNP U, which specifically bound to the leader RNA of VSV, both in vitro and in vivo (Fig. 3B and 5). Initial characterization of this protein was done by computer search based on its molecular weight (120,000), nuclear localization, and competition with oligonucleotides. This protein was precisely identified by its binding to bacterially expressed hnRNP U protein and immunoprecipitation of bound RNA with a monoclonal antibody specific for hnRNP U (Fig. 4). Competition of this binding with a specific cellular substrate, the MII fragment of the human topoisomerase I gene, further confirmed its identity (data not shown). Remarkably, LS RNA is competed out by poly(G) most efficiently, although LS RNA is rich in A and U (Fig. 3). This observation is similar to that previously reported regarding the binding of hnRNP U to ribohomopolymers (16, 17). Since the binding motif of hnRNP U is not known, it appears that it recognizes a domain within the secondary structure of the cis-acting RNAs rather than its sequences. This observation was further confirmed by mutational analysis, in which deletion of 15 nucleotides from the 5′ end of LS RNA (67% A content) reduced binding by 90% (Fig. 6A), although poly(A) by itself failed to compete binding of LS RNA to hnRNP U (Fig. 3). Further studies are needed to determine the precise structural motif in LS RNA needed for hnRNP U binding.

hnRNP U is an abundant nucleoplasmic phosphoprotein, the largest of the major hnRNP proteins (16, 17) in eucaryotic cells. These molecules have been shown to associate with nascent RNA polymerase II transcripts in the nucleoplasm to form hnRNP complexes which are implicated in pre-mRNA processing (8, 9, 16). The finding that leader RNA can be immunoprecipitated from the nuclei of VSV-infected cells with the monoclonal antibody against the hnRNP U protein (Fig. 5) indicates that upon infection with VSV, LS RNA could be a natural substrate for hnRNP U and may have some role in disrupting the machinery of cellular RNA synthesis, as postulated previously (18–20).

Indirect double immunofluorescence labeling and confocal microscopic analyses provided evidence for the localization of both hnRNP U and viral RNP during infection (Fig. 7). It is apparent from these studies that hnRNP U, which is predominantly a resident protein in the nucleus, diffuses out in the cytoplasm following VSV infection and colocalizes with VSV RNP in the same region. These results can be interpreted to suggest that VSV either blocks the entry of newly synthesized hnRNP U into the nucleus or facilitates its exit from the nucleus. At present, it is not clear which event is actually facilitated by VSV. Whatever the role of VSV in this process, the immunofluorescence studies strongly suggest that hnRNP U and VSV RNP colocalize in the cytoplasm. Coupled with the finding that hnRNP U is also packaged within the virions (Fig. 8), this strongly suggests that hnRNP U plays an important role in the life cycle of VSV. It is tempting to speculate that leader RNA, by its binding to hnRNP U, may be involved in VSV-mediated shutoff of host DNA and RNA metabolism. To the virus’s advantage, the binding of hnRNP U to leader RNA, possibly in association with the La protein, may cause structural alteration of the leader-N junction, enabling the RNA polymerase to read through the junction region and leading to the synthesis of the full-length antigenome. Future experiments along these lines would certainly shed light on the role of hnRNP U in the gene expression of VSV.