INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The study of transcriptional regulation of viral genes by cellular proteins provides important information on the interaction between the virus and host cell during steps involved in reactivation of the virus and during the course of its lytic infection. Previously, we and others demonstrated that the human neurotropic JC virus (JCV) contains an enhancer-promoter region with the ability to interact with several regulatory proteins from glial and nonglial cells (38). JCV is a polyomavirus which infects greater than 70% of the human population during childhood, and its reactivation in immunocompromised individuals causes the fatal demyelinating disease progressive multifocal leukoencephalopathy (PML) (8, 17, 32). Although the JCV genome has been detected in several tissues (14, 15, 22, 33), replication of the viral genome is seen predominantly in glial cells of the central nervous system (CNS) (8). Several lines of studies, including in vitro transcription of the viral genome (1, 2), transfection of various cells with recombinant plasmids containing the viral regulatory sequence (16, 25, 46), and creation of transgenic animals containing the JCV promoter sequence (16, 23, 41), have established that the tissue-specific replication of the viral genome is due to the transcriptional activity of the viral promoter, which is more efficient in CNS cells. Therefore, much effort has been directed toward analysis of the JCV regulatory sequence and identification of host proteins which, upon interaction with their target sequence within the promoter, confer specificity to viral gene transcription. The typical regulatory sequence of JCV is comprised of two 98-bp tandem repeats, each containing a TATA box in juxtaposition with a pentanucleotide repeat, AGGGAAGGGA, and an NF-1 motif. This strain of JCV is called JCV_Mad-1 (18). Results from earlier studies led us to believe that the pentanucleotide repeat sequence, also referred to as the lytic control element (LCE), has the ability to modulate JCV early and late promoters (44, 45) and contributes to viral DNA replication (10, 30). In subsequent studies, two cellular single-stranded DNA binding proteins named Pur and YB-1, which recognize purine-rich and C/T-rich sequences, respectively, of the LCE, were identified (12, 26, 44). Results from subsequent studies indicated that in JCV_Mad-1, binding of YB-1 to its DNA target within the LCE is increased by Pur. In contrast, interaction of Pur with the LCE motif is diminished once YB-1 is included in the reaction mixture (12). Functionally, Pur was able to induce JCV early gene transcription, whereas YB-1 was the activator for the JCV late promoter (11). Results from DNA binding and transfection studies suggested a functional role for these proteins via the LCE motif in transcription of the JCV promoters in glial cells (11, 12).

Another peculiarity in the JCV genome is the hypervariability of the structural organization of the viral control sequences (17). For example, the control region of a newly isolated JCV from brain tissue of a PML patient (51) contains a 23-bp sequence element within the pentanucleotide repeat, disrupting this important regulatory motif. A similar DNA sequence with an identical nucleotide composition is present within the JCV regulatory region of the JCV archetype, JCV_CY (Fig. 1). JCV_CY was first isolated from the urine of nonimmunosuppressed patients as well as healthy individuals (49). According to one model, JCV_Mad-1, which is routinely detected in brains of PML patients, may evolve from JCV_CY by DNA rearrangements whereby the 23-bp sequence element is deleted. Although the importance of the 23-bp sequence in the regulation of the JCV genome remains to be investigated, early studies have indicated that despite disruption of the LCE motif, the presence of the 23-bp sequence element does not alter the level of viral gene transcription (51). These observations suggest (i) that the presence of the 23-bp sequence element within the JCV enhancer-promoter region does not interfere with the ability of the LCE binding proteins, i.e., YB-1 and Pur, to modulate viral gene transcription and/or (ii) that the 23-bp DNA sequence provides a new target for interaction of Pur and YB-1, thus ensuring efficient transcription of the viral genome in the infected cells. Here, we have performed a series of DNA binding and protein-protein interaction experiments to investigate the association between Pur, YB-1, and the 23-bp DNA sequence and their effect on transcription of the viral promoter in glial cells.

View larger version (10K): [in this window] [in a new window]   FIG. 1.   Structural organization of the regulatory region of JCVCY. NF-B and ori, regulatory sequences for binding of the NF-B transcription protein and the origin of viral DNA replication, respectively. The regulatory region of JCVCY contains one 98-bp repeat with two additional motifs of 23 and 66 bp. The 23-bp sequence element disrupts the pentanucleotide repeat sequence (AGGGAAGGGA). The nucleotide composition of the 23-bp insertion is shown at the bottom. The directions of transcription of the viral early and late promoters are indicated by the arrows.

We demonstrate that Pur and YB-1 have the ability to interact with the 23-bp sequence element. Evidently, the interplay between YB-1 and Pur determines their ability to bind to this element. Moreover, we demonstrate that ectopic expression of YB-1 and Pur causes synergistic activation of the JCV minimal promoter and that this event requires an intact 23-bp sequence element within the viral genome.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cell culture. U-87MG cells were maintained in Dulbecco's modified Eagle medium (GIBCO, Grand Island, N.Y.) supplemented with 10% fetal calf serum and penicillin-streptomycin. Cells were grown at 37°C in 7% CO₂.

Construction of plasmids. Reporter plasmids, designated pBLCAT₃-CY_E and pBLCAT₃-CY_L, contain the minimal promoter region from JCV_CY, which spans nucleotides 5077 to 63, in the early and late gene orientations, respectively. These constructs were created by PCR amplification of the JCV_CY regulatory region with the MS-1 (5'-TGAGCTCATGCTTGGCTGGCAA-3') and MS-2 (5'-CCTAAAAAGCCTCCACGCCCT-3') primers. The PCR-amplified DNA fragments were first cloned into the SmaI site of Bluescript SK (Stratagene, La Jolla, Calif.) and then transferred into the BamHI/PstI site of pBLCAT₃ (29). Each clone was examined for the orientations of their inserts by direct DNA sequencing. pGEX2T-YB-1 was provided by Gene MacDonald (31). Glutathione S-transferase (GST)-YB-1 C-terminal deletion mutants were created as follows. pGEX2T-YB-1 was first digested with SalI/EcoRI, AlwNI/EcoRI, StyI/EcoRI, and BglI/EcoRI enzymes, and the appropriate DNA fragments were removed by gel purification. The 5' and 3' ends of the DNA fragments were blunt ended upon treatment with Klenow enzyme and religated with T4 DNA ligase. The plasmids obtained from these manipulations were designated pGEX2T-YB-1(1-203), -YB-1(1-125), -YB-1(1-75), and -YB-1(1-37).

EMSA. For electrophoretic mobility shift assays (EMSAs), single-stranded DNA fragments representing the early strand (3'-ATCCCTCCTCGACCGATTTTGAC-5') or late strand (5'-TAGGGAGGAGCTGGCTAAAACTG-3') of the JCV_CY 23-bp sequence were 5' end labeled with [-³²P]ATP by using T4 polynucleotide kinase. Bacterially expressed highly purified fusion proteins, i.e., maltose binding protein (MBP) fused to YB-1 (MBP-YB-1) or to Pur (MBP-Pur) (12), were mixed with the single-stranded probes (30,000 cpm/reaction) in a binding buffer containing 0.1 µg of poly(dI-dC)/µl, 12 mM HEPES (pH 7.9), 4 mM Tris-HCl (pH 7.5), 60 mM KCl, 5 mM MgCl₂, and 0.1 mM dithiothreitol and incubated for 30 min at 4°C. The DNA-protein complexes were resolved on 6% polyacrylamide gels in 0.5× TBE (1× TBE is 89 mM Tris-HCl [pH 8.0], 89 mM boric acid, and 2 mM EDTA [pH 8.0]). For Fig. 2F and G, EMSAs were performed as described above with GST-Pur and GST-YB-1. The gels were dried, and complexes were detected by autoradiography at 70°C with an intensifying screen.

In vitro transcription and translation assay. To synthesize YB-1 and Pur proteins in vitro, we utilized a TNT coupled in vitro transcription-translation system (Promega, Madison, Wis.). The proteins were labeled with [³⁵S]methionine as recommended by the supplier.

GST fusion protein affinity chromatography. U-87MG cells expressing Pur or YB-1 were lysed in LB150 buffer, containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 5 mM EDTA, 0.1% Nonidet P-40, 50 mM sodium fluoride, 1 mM phenylmethylsulfonyl fluoride, 1 µg of leupeptin per ml, and 1 µg of aprotinin per ml, for 30 min on ice. Lysates were scraped, collected into microcentrifuge tubes, vortexed briefly, and microcentrifuged for 30 min at 4°C. Five hundred micrograms of whole-cell extract prepared from U-87MG cells either untransfected or transfected with HA-tagged Pur (HA-Pur) was incubated with 1 µg of GST or GST-YB-1 fusion protein coupled to glutathione-Sepharose beads in 300 µl of LB 150 for 2 h at 4°C with continuous rocking. Alternatively, 500 µg of whole-cell extract prepared from U-87MG cells either untransfected or transfected with histidine-tagged YB-1 (His-YB-1) was incubated with 1 µg of GST or GST-Pur under similar conditions. After incubation, the beads were pelleted and washed five times with LB 150 buffer. Bound proteins were eluted with Laemmli sample buffer and heated at 95°C for 10 min. Complexes were resolved on sodium dodecyl sulfate (SDS)-10% polyacrylamide gel and analyzed for HA-Pur or His-YB-1 with either anti-HA or anti-T7 antibodies, respectively. Thirty micrograms of crude extracts was loaded on the gels as migration controls.

Expression and purification of recombinant fusion proteins. YB-1 and Pur fusion proteins used in band shift studies were prepared by methods described previously (12). Expression of GST-YB-1, GST-Pur, and the mutant variants was carried out by the procedure described previously (19). Briefly, 100 ml of overnight cultures of Escherichia coli DH5, transformed with either pGEX2T-YB-1, pGEX2T-Pur, or their respective deletion mutant plasmids, was diluted 1:10 in fresh Luria-Bertani medium supplemented with ampicillin (100 µg/ml). Cultures were induced with 0.4 M isopropyl--D thiogalactopyranoside (IPTG) at an optical density of 0.6 and incubated for an additional 1.5 h at 37°C. Cells were collected by centrifugation and resuspended in 10 ml of lysis buffer containing 20 mM Tris-HCl (pH 8.0), 100 mM NaCl, 1 mM EDTA, and 0.5% Nonidet P-40 supplemented with 1 mM phenylmethylsulfonyl fluoride, 2 µM pepstatin A, 0.6 µM leupeptin, and 2 mM benzamidine. After sonication, clear cell lysates were prepared and incubated with 100 µl of 50% GST-Sepharose beads (Pharmacia, Piscataway, N.J.) overnight at 4°C. GST fusion proteins were washed three times with 50 ml of lysis buffer, and protein quality was analyzed by SDS-PAGE followed by Coomassie blue staining. Additionally, GST-Pur and GST-YB-1 proteins were eluted from the Sepharose beads with 10 mM glutathione in elution buffer containing 20 mM Tris-HCl (pH 8.0), 150 mM NaCl, and 5 mM dithiothreitol and subsequently dialyzed in 20 mM Tris-HCl (pH 7.4)-150 mM NaCl-10% glycerol.

Transient-transfection assays. U-87MG cells were seeded in 60-mm-diameter culture dishes the day before transfection. Cells were transfected with plasmid DNA by calcium phosphate coprecipitation (21). The total amounts of DNA in the transfection mixtures were equalized to 30 µg with empty vector DNA. At 48 h posttransfection, cells were lysed by three freeze-thaw cycles, and the protein contents were determined by the method of Bradford (9). Chloramphenicol acetyltransferase (CAT) activity was determined by a previously described method (20). Results are expressed as fold activation. Transfections were performed at least three times with different plasmid preparations. In all transfection studies, 1 µg of plasmid expressing -galactosidase (-gal) (RSV--gal) was included in the transfection mixture, and the level of -gal activity was used as the baseline to normalize the CAT levels from the various experiments.

Western blot analysis. Whole-cell extracts were prepared from either untransfected or transfected U-87MG cells (4), and 30 µg of protein extract was separated by SDS-10% PAGE prior to transfer to nitrocellulose membranes (Schleicher and Schuell, Keene, N.H.). The filters were probed with specific antibodies (1:200 dilution) and developed with an ECL detection kit according to the recommendations of the manufacturer (Amersham, Arlington Heights, Ill.).

Coimmunoprecipitation assay. EBV-His-YB-1 and CMV-HA-Pur expression plasmids were introduced into U-87MG cells by the calcium phosphate precipitation method (21). At 48 h posttransfection, cells were lysed in LB 150 buffer, and approximately 0.5 mg of whole-cell extract in a total volume of 0.5 ml of lysis buffer was incubated with 0.5 µg of either anti-His (Novagen, Madison, Wis.) or anti-HA (Berkeley Antibody, Richmond, Calif.) antibodies for 16 h at 4°C. Immunocomplexes were precipitated with the addition of protein A-Sepharose beads (Pharmacia) for an additional 2 h at 4°C. After four washes in lysis buffer, immunocomplexes were resolved by SDS-10% PAGE and analyzed by Western blotting with an ECL detection kit.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Interaction of Pur and YB-1 with the JCV_CY 23-bp sequence. To examine the ability of Pur and YB-1 to interact with the 23-bp DNA sequence, synthetic single-stranded DNA probes representing the early (23E) and late (23L) strands of the 23-bp sequence element were incubated with highly purified bacterially produced Pur and YB-1. The protein-DNA complexes were analyzed by native PAGE. As shown in Fig. 2A, lane 2, a major band corresponding to the YB-1:23E complex was observed. Inclusion of Pur in the binding reaction mixture at a low concentration enhanced the intensity of the band corresponding to the YB-1:23E complex and at higher concentrations showed an inhibitory effect on the association of YB-1 with the 23E probe (Fig. 2A, compare lane 2 to lanes 3, 4, and 5). The band corresponding to the Pur:23E complex showed slower electrophoretic mobility and migrated above the YB-1:23E complex. Of interest is that the intensity of the Pur:23E band was not enhanced upon increasing the concentration of Pur in the binding reaction mixture. These experiments were performed under conditions where the probe was in excess, indicating that the observed inhibition of YB-1:23E assembly by Pur may not be attributed to the limitation of the DNA probe.

View larger version (159K): [in this window] [in a new window]   FIG. 2.   Pur and YB-1 modulate each other's binding to 23-bp single strands. (A to C) Band shift assay. (A) A 23-bp single-stranded DNA probe from the early strand (23E) was incubated with 30 ng of MBP-YB-1 alone (lane 2) or with 200, 400, or 600 ng of MBP-Pur (lanes 3 to 5, respectively). Lanes 1 and 6 contain 600 ng of MBP and 600 ng of Pur, respectively. The positions of YB-1 and Pur complexes are shown by the arrows. (B) The 23-bp early DNA probe (23E) was incubated with 200 ng of Pur alone (lane 2) or with 30, 60, or 120 ng of YB-1 (lanes 3 to 5, respectively). (C) The 23-bp single-stranded DNA probe for the late strand (23L) was incubated with YB-1 and Pur fusion proteins as described for panel B. (D and E) Competitive band shift assays. (D) The 23E probe was mixed with YB-1 protein in the absence or presence of 50- and 250-fold molar excesses of unlabeled competitor DNA as indicated. (E) The 23L probe was incubated with Pur protein alone (lane 2) or in the presence of 50- and 250-fold molar excesses of competitor DNAs. The positions of the Pur and YB-1 complexes are shown by the arrows. (F and G) Antibody (Ab) supershift assays. (F) GST-YB-1 fusion protein (150 ng) was incubated with the 23E probe (lanes 1 to 3), and the binding reaction mixture included either 1 µg of preimmune (pre) (lane 2) or 1 µg of anti-GST (lane 3) antibodies. The positions of the GST-YB-1:23E complexes are indicated by a bracket. (G) Two hundred nanograms of GST-Pur fusion protein was incubated with the 23L probe. The reaction mixture included preimmune (lane 2) and anti-GST (lane 3) antibodies as described for panel F. The positions of the GST-Pur:23L complexes are indicated by arrows. Supershifted complexes are depicted by an arrowhead.

Interaction of YB-1 and Pur in the absence of the 23-bp sequence. Results from DNA binding studies implied that Pur and YB-1 may interact with each other and that this interaction may determine the levels of their binding to the 23-bp motif. To investigate the direct interaction of YB-1 and Pur in glial cells, in vitro GST pull-down experiments were performed. To this end, U-87MG cells were transfected with an HA-tagged Pur expression plasmid (pCMV-HA-Pur) or with the histidine-tagged YB-1 expression plasmid (pEBV-His-YB-1). Forty-eight hours after transfection, protein extracts were prepared and incubated with bacterially produced GST and GST-Pur and GST-YB-1 fusion proteins immobilized on glutathione-Sepharose beads. After washing, proteins retained on the beads were analyzed by immunoblot analysis with an HA antibody which recognizes the HA tag on Pur or an anti-His antibody which recognizes the histidine tag on YB-1. As shown in Fig. 3A, incubation of extract derived from cells transfected with HA-Pur and Sepharose columns containing either GST or GST-YB-1 resulted in the specific retention of HA-Pur on the column containing GST-YB-1 (lane 6) but not on the column containing GST alone (lane 5). As expected, no band corresponding to the HA-Pur protein was detected in lanes corresponding to protein from untransfected cells (Fig. 3A, lane 1) or in the lanes containing eluates obtained after incubation with GST or GST-YB-1 (Fig. 3A, lanes 3 and 4). A band corresponding to HA-Pur was also detected in the unfractionated protein extract from the pCMV-HA-Pur-transfected cells (Fig. 3A, lane 2). In reciprocal experiments, protein extracts from the control untransfected cells and the cells transfected with pEBV-His-YB-1 were examined for production of His-YB-1 and for the ability of the ectopically produced YB-1 to bind to Pur. As shown in Fig. 3B, anti-His antibody was able to detect the YB-1 fusion protein in the extract from transfected cells (lane 2) and in the fractions that bound to GST-Pur but not GST alone (compare lane 6 to lane 5). Together, these observations point to the in vitro association of Pur and YB-1.

View larger version (31K): [in this window] [in a new window]   FIG. 3.   In vitro interaction of Pur and YB-1. (A) Bacterially produced GST or GST-YB-1 was immobilized on GST-Sepharose beads and incubated with whole-cell extracts prepared from untransfected U-87MG cells or U-87MG cells transfected with CMV-HA-Pur expression plasmid for 2 h at 4°C. The Sepharose beads were washed extensively with lysis buffer, and proteins interacting with GST or GST-YB-1 were analyzed by SDS-PAGE followed by immunoblotting with anti-HA antibody, which detected the HA-Pur fusion protein. (B) Whole-cell extracts prepared from untransfected U-87MG cells or cells transfected with EBV-His-YB-1 expression plasmid were incubated with either GST alone or GST-Pur. Proteins interacting with GST or GST-Pur were harvested and analyzed by immunoblotting with anti-His antibody for detection of His-tagged YB-1. Arrowheads indicate nonspecific bands detected with the anti-His antibody. Numbers on the left are molecular masses in kilodaltons.

View larger version (21K): [in this window] [in a new window]   FIG. 4.   Coimmunoprecipitation of YB-1 and Pur. (A) Approximately 0.5 mg of whole-cell extracts prepared from untransfected U-87MG cells or cells transfected with EBV-His-YB-1 and CMV-HA-Pur expression plasmids were immunoprecipitated (IP) with 0.5 µg of monoclonal anti-His (-His) or preimmune (mouse) serum (-pre). Immunocomplexes were analyzed by SDS-PAGE followed by immunoblotting with an antibody directed against the HA tag for detection of HA-Pur fusion protein. An asterisk indicates the position of immunoglobulin G detected by the secondary antibody. (B) Whole-cell extracts (0.5 mg) were immunoprecipitated with an antibody directed against HA tag or preimmune mouse serum. The immunocomplexes were analyzed by SDS-PAGE followed by immunoblotting with anti-His antibody for detection of His-YB-1. Arrowheads indicate nonspecific bands.

Functional interaction of Pur and YB-1 on the JCV_CY minimal promoter containing the 23-bp sequence. To investigate the functional importance of the interplay between Pur and YB-1 for the transcriptional activity of the 23-bp sequence, we performed cotransfection experiments. In these studies the human glial cell line U-87MG was transfected with the JCV minimal promoter construct containing the 23-bp motif, alone or together with the recombinants expressing Pur and YB-1. As shown in Fig. 5A, in the presence of a small amount of YB-1, no significant changes in the activity of the JCV_E promoter were detected (compare lane 1 to lane 2). Under similar experimental conditions, the addition of Pur expression plasmids to the transfection mixture, in particular at the higher concentration, significantly increased the activity of the JCV_E promoter in these cells (compare lane 2 to lane 4). The level of activation was substantially higher than the sum of the transcriptional activities obtained with YB-1 (lane 2) and Pur (lane 5) alone, suggesting that YB-1 and Pur may function synergistically to stimulate the JCV early promoter. Figure 5A also shows that at a lower concentration, Pur has no effect on the JCV_E promoter (lane 7), whereas in the presence of YB-1, a drastic increase in viral promoter activity is observed (compare lane 7 with lanes 8 and 9). Again, the level of activation mediated by Pur plus YB-1 was much higher than the activity observed for the sum of YB-1 and Pur, pointing to the synergistic activity of these two proteins on the JCV_E promoter.

View larger version (33K): [in this window] [in a new window]   FIG. 5.   Functional cooperation between Pur and YB-1 in transcriptional activation of the JCVCY minimal promoter in U-87MG cells. (A) pBLCAT3-CYE reporter plasmid containing the JCVCY minimal promoter (7.5 µg), as shown (top), was introduced into U-87MG cells alone or together with YB-1 and Pur expression plasmids. In cotransfections, as the plasmid concentration for one transactivator was kept constant at 10 µg for each lane (lanes 2 to 4 for YB-1 and lanes 7 to 9 for Pur), the plasmid concentration for the other transactivator was varied (5, 10, and 10 µg of plasmid DNA of Pur for lanes 3 to 5 respectively, and 5, 10, and 10 µg of plasmid DNA of YB-1 for lanes 8 to 10, respectively). One microgram of RSV--gal plasmid was added to each transfection mixture. The DNA concentrations for each lane were normalized with the addition of empty expression vector DNA. CAT activity was measured and normalized for -gal activity. The transfection experiments were repeated at least three times. The data obtained from CAT assays were quantitated and are presented as fold activation relative to the basal expression of the minimal promoter (bottom). Error bars indicate standard deviations. Results of a representative CAT assay are shown in the middle. (B) Experiments similar to those detailed for panel A were performed with the JCV late promoter, pBLCAT3-CYL.

Protein domains involved in formation of the Pur:YB-1 complex. In the next series of experiments, we attempted to identify the region(s) within Pur which is important for its interaction with YB-1. In this study, in vitro-translated ³⁵S-labeled YB-1 protein was prepared and used in GST pull-down assays with GST-Pur deletions encompassing various regions of Pur. As shown in Fig. 6A, binding of YB-1 to Pur was gradually decreased upon removal of amino acid residues 54 to 166 and was completely abrogated by further amino-terminal deletions. These data suggest that the region located between residues 167 and 216 is the minimal sequence which is important for Pur interaction with YB-1. Results with C-terminal deletion mutants further indicated that residues 174 to 215 are critical for complex formation between YB-1 and Pur. These results are summarized in Fig. 6C.

View larger version (29K): [in this window] [in a new window]   FIG. 6.   Mapping of the domain(s) of Pur involved in interaction with YB-1. (A and B) Mapping of the interaction domain of Pur with YB-1. GST-Pur or its N-terminal (A) or its C-terminal (B) deletion mutants were incubated with in vitro-translated [35S]methionine-labeled YB-1. Sepharose beads were washed three times with lysis buffer, and bound proteins were resolved by SDS-10% PAGE. One-tenth of the input YB-1 used in each reaction was loaded for migration controls (lane 1 in each panel). The labeled arrow marks the position of in vitro-translated [35S]methionine-labeled YB-1, and the arrowhead indicates a degradation product of YB-1. Numbers on the left are molecular masses in kilodaltons. (C) Summary of the results obtained from in vitro mapping assays. A schematic representation of the Pur protein is shown at the top (not shown to scale). The relative strengths of the interactions between GST-Pur and its deletion mutants with in vitro-translated [35S]methionine-labeled YB-1 are depicted.

View larger version (37K): [in this window] [in a new window]   FIG. 7.   Mapping the domain(s) of YB-1 involved in the interaction with Pur. (A) In vitro-translated 35S-labeled Pur was incubated with either GST alone or C-terminal deletion mutants of GST-YB-1 fusion proteins. Bound proteins were analyzed as described for Fig. 6. The labeled arrow marks the position of in vitro-translated 35S-labeled Pur. (B) Three different 35S-labeled in vitro-translated amino-terminal mutants of YB-1, i.e., YB-1(126-318), YB-1(204-318), and YB-1(250-318), were incubated with GST (lanes 2, 5, and 8) or GST-Pur (lanes 3, 6, and 9). Bound proteins were analyzed as described for Fig. 6. Lanes 1, 4, and 7 contain 1/10 of the amount used in the pull-down experiments with YB-1(126-318), YB-1(204-318), and YB-1(250-318), respectively. The arrowhead designates the position of the in vitro-translated amino-terminal deletion mutants. The asterisk denotes a nonspecific product present in the in vitro transcription-translation reactions. (C) A schematic representation of full-length YB-1 is shown at the top. CSD, the cold shock domain of YB-1. The relative strengths of the interactions observed between YB-1 and Pur are depicted on the right.

Interaction of Pur and YB-1 is important for their regulatory action on the JCV_CY promoter. To further assess the functional interaction between YB-1 and Pur, transient-transfection studies utilizing deletion mutants of each protein were performed. Expression of these mutants was verified by Western blot analysis (data not shown). In these studies, U-87MG cells were transfected with the minimal JCV_L promoter containing the 23-bp motif along with a small amount of YB-1 in the presence of the Pur mutant which binds YB-1 [Pur(85-322)] or its variant with no ability to bind YB-1 [Pur(216-322)]. As shown in Fig. 8A, ectopic expression of YB-1 and Pur(85-322) resulted in a synergistic activation of the JCV promoter. This is similar to the activity of full-length Pur and YB-1 (Fig. 5). Under similar conditions, no synergism between YB-1 and Pur(216-322) was detected (Fig. 8B). These observations suggest that the region within Pur which is important for its association with YB-1 is required for its functional interaction with YB-1 with the JCV promoter sequence. In a different series of experiments, cells were transfected with the JCV_L promoter together with a plasmid expressing a mutant YB-1, YB-1(1-125), which binds to Pur, or its variant, YB-1(1-37), with no ability to bind Pur. As shown in Fig. 8C, while Pur or YB-1(1-125) alone did not show a significant stimulatory effect on the JCV_L promoter (lanes 2 and 5, respectively), coproduction of these two proteins enhanced the activity of the JCV_L promoter in the transfected cells. Under similar conditions, coproduction of the full-length Pur with YB-1(1-37) had no drastic effect on the transcription of the JCV promoter (Fig. 8D). These observations along with binding results indicate that a cooperative interaction between YB-1 and Pur may determine their transcriptional ability with the JCV promoter. In the next series of transfection experiments, we utilized a JCV promoter construct with a cluster mutation within the 23-bp sequence that converts GGA to AAA. As shown in Fig. 8E, ectopic expression of YB-1 in the absence and presence of Pur had no stimulatory effect on the mutant viral promoter (compare lane 1 to lanes 2 to 5). In fact, these proteins, alone or in combination, exert a negative effect on the basal promoter activity and decrease the level of transcription from the JCV minimal promoter. Figure 8E (bottom) shows a summary of the results from these experiments. These observations together suggest that association of Pur and YB-1 with each other and the viral 23-bp regulatory sequence is functionally important for transcription of the viral genome.

View larger version (31K): [in this window] [in a new window]   FIG. 8.   Functional interaction of YB-1 and Pur deletion mutants on the JCVCY late minimal promoter. (A and B). A 7.5-µg amount of JCVL minimal promoter reporter construct (shown above the panels) was introduced into U-87MG cells alone (lanes 1) or in combination with YB-1 and Pur deletion mutants. In cotransfections, as the plasmid concentration for YB-1 was kept constant at 10 µg (lanes 2 to 4), the plasmid concentrations for Pur deletion mutants were varied (5, 10, and 10 µg of plasmid in lanes 3 to 5, respectively). CAT activity was measured and normalized as described for Fig. 5. The lower panels show the quantitative analysis of the results from transfection experiments. Error bars indicate standard deviations. (C and D) Transfections were performed with 7.5 µg of JCVL promoter as described above and with 10 µg of Pur-expressing plasmid alone or with 5 or 10 µg of YB-1 mutants as indicated. (E) Approximately 7.5 µg of the JCVL minimal promoter reporter construct with a mutation in the 23-bp motif (shown at the top) was introduced into U-87MG cells alone or together with YB-1 and Pur expression plasmids. The experimental design was virtually identical to that described in the legend to Fig. 5. The bottom panel shows the quantitative analysis of representative experiments.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

A mechanism, so-called cross-family interaction, in which one transcription factor modulates the function of another one is an important event in the control of eukaryotic gene transcription. Examples of this cross-family interaction include association of NF-B family members with Jun/Fos or C/EBP family members (42, 43) and the interaction of Jun/Fos with MyoD (7), which can determine the regulatory activity of these factors. We have utilized the human neurotropic virus JCV as a model to unravel the regulatory pathways which include DNA-protein and protein-protein interactions in mediating transcription of eukaryotic genes in CNS cells. In earlier studies, we demonstrated that YB-1, a DNA binding protein which recognizes the inverted CCAAT sequence, may exert its regulatory action upon communication with other cellular proteins, including NF-B subunits via the D-regulatory motif (39). Moreover, we demonstrated that YB-1 could bind to another viral regulatory motif, the LCE (26). As the 23-bp sequence, by disrupting the LCE motif, may affect its regulatory action on the JCV genome, we performed a series of structural and functional studies to analyze the functional importance of the 23-bp sequence for JCV gene transcription. Here, we demonstrate that while YB-1 binds to the 23-bp DNA sequence of JCV, its interaction with the DNA can be regulated by Pur, a single-stranded DNA binding protein that recognizes the GC/GA-rich motif. This regulatory event is mutual, as YB-1 was able to influence the association of Pur with its target GC/GA nucleotide within the 23-bp sequence. Of interest is that results from our band shift studies showed no evidence for formation of a ternary complex indicative of simultaneous association of the 23-bp sequence with YB-1 and Pur. Thus, according to one model, Pur, by transient interaction with YB-1, may induce a conformational change in YB-1, which inhibits its association with the DNA molecule. This event could be bidirectional, as interaction of Pur with the DNA is also affected. Such a mechanism has been previously reported, where the activity of MyoD1 is enhanced in the presence of heat shock protein 90 (40).

Alternatively, it is possible that Pur, by binding to the YB-1:23-bp complex, forms an unstable ternary complex, YB-1:23-bp:Pur, which may dissociate during gel electrophoresis. Results from our in vitro and in vivo protein-protein interaction studies indicated that these two proteins can form a heterodimer in the absence of DNA and that certain domains within each protein are critical for their heterodimerization. Results from functional studies demonstrated that overexpression of YB-1 and Pur synergistically stimulate transcription of the JCV genome. It was evident that the 23-bp sequence is important for such a regulatory effect. Thus, while Pur and YB-1 can form a complex off their target DNA sequences, their association with the DNA sequence is critical for their transcriptional activity. Taken together, these observations demonstrate that cooperative interaction of two cellular proteins with distinct DNA recognition sites positioned within the 23-bp motif modulates viral gene transcription. Of note is that in earlier studies we demonstrated that in the absence of the 23-bp sequence, the intact LCE may provide a target for the interplay of YB-1 and Pur and for their regulatory action on JCV early and late gene transcription (11, 12). While the 23-bp motif interrupts the pentanucleotide AGGGAAGGGA repeat of the LCE motif, it contains nucleotide sequences which provide targets for YB-1 and Pur binding. However, the important difference is that the early strand of LCE bound exclusively to YB-1, whereas its late strand interacts only with Pur. Here, we demonstrate that while YB-1 does not interact with the late 23-bp sequence element, both YB-1 and Pur form complexes with the early strand of the 23-bp sequence element. Interestingly, results from functional studies revealed that while both Pur and YB-1 alone can stimulate viral early and late gene transcription, together their activities are significantly increased. This observation differs from previous results for JCV_Mad-1, containing an intact LCE but not the 23-bp sequence element, in which Pur and YB-1 had more effect on early and late gene transcription, respectively. Thus, while the 23-bp sequence element provides an alternative site for binding of YB-1 and Pur, differential activities of these regulatory proteins on the viral genome may exist, suggesting that the structural organization of the viral promoter can dictate the activities of the participant regulatory proteins.

Both YB-1 and Pur represent cellular regulatory proteins with the ability to control expression of a broad range of cellular and viral genes. For example, earlier studies demonstrated that YB-1 modulates transcription of the human MDR1 gene (3), the chicken -2(1) collagen gene (6), the grp 78 gene (28), the matrix metalloproteinase 2 gene (34), the major histocompatibility complex class II HLA-DR- gene (35), the thyrotropin receptor gene (37), the gamma interferon gene (48), and the myosin light chain 2 V gene (52). On the other hand, earlier reports demonstrated that Pur stimulates expression of the myelin basic protein gene (23), human immunodeficiency virus type 1 (13), the transforming growth factor 1 gene (47), and the neuron-specific FE65 gene (50). As our results presented here suggest that the activities of YB-1 and Pur on the JCV genome could be regulated by their mutual interaction, one can postulate a similar mechanism for control of the other responsive viral and cellular genes. The common feature of YB-1 and Pur is their ability to interact in a sequence-specific manner with the single-stranded DNAs. Interestingly, while Pur recognizes the purine-rich sequences, YB-1 binds to the C/T-rich DNA elements. The ability of these two proteins to recognize complementary single-stranded elements leads to interesting speculation about their control over transcription. Furthermore, our data presented here demonstrate the physical interaction of these two proteins off the DNA molecule. This information, along with the earlier observation pointing to the ability of YB-1 to promote single-stranded DNA regions (31), suggests that YB-1 may function as a chaperone and, by promoting the single-stranded region within the duplex DNA, facilitate binding of Pur to its purine-rich target. Furthermore, in recent studies it has been shown that expression of YB-1 may be regulated by environmental stress, such as UV-irradiation or drug treatment, etc. (5). Thus, the functional role of cooperative interaction of YB-1 and Pur in stimulation of cellular genes whose products are important for cell survival during stress may become an important regulatory event. Moreover, as both Pur and YB-1 are expressed in a variety of cells and tissues, future studies will be directed toward the understanding of the interaction with the JCV genome in other tissues, such as B cells, where JCV has been detected.