Kaposi's Sarcoma-Associated Herpesvirus K8? Is Der
http://www.100md.com
病菌学杂志 2005年第22期
HIV and AIDS Malignancy Branch, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892
ABSTRACT
Kaposi's sarcoma-associated herpesvirus (KSHV) is a lymphotropic DNA tumor virus that induces Kaposi's sarcoma and AIDS-related primary effusion lymphoma. KSHV open reading frame 50 and K8 genes in early viral lytic infection express, respectively, a tricistronic and a bicistronic pre-mRNA, which undergo alternative splicing to create two major spliced mRNA isoforms, and ?, by inclusion (?) or exclusion () of an intron at nucleotides 75563 to 75645. This intron contains some suboptimal features, which cause the intron 5' splice site (ss) to interact weakly with U1 snRNA and the 3' ss to bind a U2 auxiliary factor, U2AF, with low affinity. Optimization of this intron in K8 (K8 intron 2) promoted the interaction of the 5' ss with U1 and the 3' ss with U2AF, resulting in a substantial increase in intron splicing. Splicing of K8 intron 2 has also been shown to be stimulated by the splicing of a downstream intron. This was confirmed by the insertion of a human ?-globin intron into the K8? exon 3-exon 4 splice junction, which promoted splicing of K8? intron 2 and conversion of the K8? mRNA to the K8 mRNA that encodes a K-bZIP protein. Intron 2 contains a premature termination codon, yet the K8? mRNA is insensitive to nonsense-mediated mRNA decay, suggesting that the truncated K8? protein may have a biological function. Indeed, although the truncated K8? protein is missing only a C-terminal leucine zipper domain from the K-bZIP, its expression antagonizes the ability of the K-bZIP to induce p53 and p21 and blocks K-bZIP-CDK2 interaction through interfering K8 mRNA production.
INTRODUCTION
Kaposi's sarcoma-associated herpesvirus (KSHV), or human herpesvirus 8, is a human gammaherpesvirus that closely resembles human Epstein-Barr virus (7). As with other gammaherpesviruses, infection with KSHV is known to be strongly associated with lymphoproliferative disorders and malignancies. It is present in all forms of Kaposi's sarcoma (16, 47) and in two other diseases, body cavity-based B-cell lymphoma (also known as primary effusion lymphoma [PEL]) and multicentric Castleman's disease (40, 41).
Like infection with other herpesviruses, KSHV infection displays two viral life cycles. Latent KSHV infection in Kaposi's sarcoma tissues and B-cell lines features the highly restricted expression of only five viral genes (9, 62). The lytic KSHV life cycle is characterized by the production of progeny virus from infected cells and can be induced by chemicals such as tetradecanoyl phorbol acetate (TPA) and n-butyrate in PEL-derived B cells with latent KSHV infection (26, 27, 34). Chemical induction in PEL-derived B cells initiates the expression of a KSHV transactivator, open reading frame 50 (ORF50), which is absolutely required for the lytic switch from KSHV latency (24, 43).
KSHV ORF50 is an immediate-early gene transcribed as a polycistronic RNA with five exons and four introns. It encodes a pivotal transactivator that mediates transcription of all early genes (24, 43). ORF50 exon 1 and exon 2 contain the coding region of ORF50, whereas exons 3, 4, and 5 of ORF50 are noncoding exons that overlap with the K8 and K8.1 coding regions (43, 46). This overlap enables ORF50 to use a common poly(A) site at nucleotide (nt) 76714, downstream of the K8.1 coding region. This unique gene structure means that KSHV ORF50, K8, and K8.1 are positioned side by side in the virus genome, and the transcripts of the three genes overlap each other (56). However, each gene initiates its transcription at different stages of the viral life cycle (23, 36, 45) and undergoes extensive RNA splicing during gene expression (56). KSHV K8 is an early gene transcribed as a bicistronic RNA with four exons and three introns that expresses a K-bZIP protein (K8) involved in viral DNA replication (1, 22, 51) and transcription (19, 21, 28, 49) as well as induction of cell cycle arrest (20, 52, 53). KSHV K8.1 is a late gene transcribed as a monocistronic RNA with two exons and one intron that encodes an envelope glycoprotein (6, 32, 45).
We have profiled the splicing patterns of KSHV ORF50, K8, and K8.1 through multiple approaches and have reported that splicing of these pre-mRNAs is extremely complex and often results in inclusion of an intron (intron 3 of ORF50, intron 2 of K8) at nt 75563 to 75645 (Fig. 1A), presumably due to suboptimal features of the intron splice sites (46, 56). mRNAs with this intron inclusion are referred to as the ? isoform; those without the intron are referred to as the isoform. We have also shown by using KSHV+ cell lines that exon 3 of the K8 pre-mRNA (exon 4 of ORF50) includes most of a previously defined intron 3 and has three 5' splice sites (ss), at nt 75838, 76155, and 76338 (46). Selection of the nt 75838 5' ss dictates the production of K8 mRNA and overwhelms the RNA processing. The alternative selection of the other two 5' ss is feasible and leads to the production of two additional bicistronic mRNAs: K8/K8.1 and -?. However, only a little K8 and no detectable K8.1 protein are translated from the novel bicistronic K8/K8.1 mRNAs in 293 cells, suggesting that production of the K8/K8.1 mRNAs may be an essential way to limit K8 mRNAs, especially K8, to a threshold at the RNA processing level (46). Altogether, we have concluded that a substantial proportion of pre-mRNAs transcribed from the ORF50, K8, and K8.1 genes undergo alternative RNA splicing, leading to the production of at least 19 species of spliced and unspliced transcripts (56). In this regard, the inclusion or exclusion of the intron from nt 75563 to 75645 plays an important role in the alternative RNA splicing of both ORF50 and K8 pre-mRNA. However, what controls this intron splicing and how it affects the biology of ORF50 and K8 remains to be understood.
In this report, we have characterized K8 intron 2 (intron 3 of ORF50) and demonstrated that the suboptimal features of intron 2 determine its inefficient interaction with the cellular splicing machinery. Further experiments showed that the inclusion of this intron in K8? mRNA results in the production of a truncated protein that plays a dominant negative role in fully spliced K8 mRNA expression, K8-mediated induction of p21 and p53, and K8-CDK2 interaction.
MATERIALS AND METHODS
Cells. The KSHV-positive B-cell line JSC-1 was a generous gift from Richard Ambinder of the Johns Hopkins University. BCBL-1 cells were obtained from Michael McGrath and Don Ganem of the University of California at San Francisco through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, National Institutes of Health. Both PEL-derived B-cell lines were maintained in RPMI 1640 medium (Invitrogen, Carlsbad, CA). Human 293 cells and HeLa cells were obtained from the American Type Culture Collection (Manassas, VA) and were maintained in Dulbecco's modified Eagle's medium (Invitrogen). Both culture media were supplemented with 10% fetal bovine serum (HyClone, Logan, UT) and with 2 mM L-glutamine, 100 U penicillin/ml, and 100 μg streptomycin/ml. The cells were incubated in the complete culture medium at 37°C in the presence of 5% CO2.
KSHV activation by chemical induction. To activate KSHV lytic gene expression in JSC-1 cells, 1 x 107 cells were treated with sodium butyrate (Aldrich, Milwaukee, WI) at a final concentration of 3 mM for 24 h. BCBL-1 cells (1 x 107) were induced with 20 ng/ml TPA (Sigma, St. Louis, MO) for 48 h. Total cell RNA and total protein samples were then prepared for further analysis of K8 expression. To analyze nonsense-mediated RNA decay (NMD) of KSHV K8? RNA, a protein synthesis inhibitor, cycloheximide (CHX; Sigma, St. Louis, MO), was added at a concentration of 50 μg/ml to the JSC-1 cells after 4 h of butyrate induction (63). The cells were further cultured for 24 h in the presence of both CHX and butyrate, after which total cell RNA was prepared for RNase protection assays (RPA).
Cell transfection. For transfection of 293 cells and HeLa cells, plasmid DNA was transfected into 1 x 107 cells using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Total cell RNA or protein was prepared 24 or 48 h after transfection, as indicated in the figures. The RNA was used for reverse transcription (RT)-PCR and RPA, and the protein was used for Western blot analysis. To study the NMD of the KSHV K8?, 293 cells were transfected with plasmid pST1, a K8 gene expression vector (46); 8 h later, the cells were cultured in the presence of CHX (28 μg/ml) (5) for 24 h prior to total cell RNA preparation.
Electroporation was used for transfection of KSHV+ BCBL-1 cells with plasmid pORF50, pST1, pST3, and pcDNA 3.0 vector. Briefly, BCBL-1 cells were washed twice with cold phosphate-buffered saline (PBS) (calcium and magnesium free) and resuspended at 1 x 107 cells/ml in RPMI 1640 without any supplements (GIBCO). Approximate 400 μl of the cell suspension was mixed with 20 μg of each plasmid and transfected in a prechilled 0.4-cm cuvette (Bio-Rad, Hercules, CA) by electroporation with an Electro Cell Manipulator BTX ECM630 (Genetronics, Inc., San Diego, CA) at the following conditions: voltage, 350 V; capacitance, 960 μF; and resistance, 50 . Under such electroporation conditions, over 50% of the cells remained viable, and half of those viable cells could be transfected with a green fluorescent protein fusion plasmid as a transfection control. The cell suspensions after electroporation were then incubated at 37°C in 20 ml of complete RPMI 1640 medium for 72 h using regular cell culture conditions prior to total cell RNA extraction and protein sample preparation.
Plasmid construction. The plasmids pKY1 and pKY2 are KSHV K8 expression plasmids based on the wild-type K8 gene expression vector pST1 (46), but with an optimized intron 2. pKY1 has an optimized (mutated) nt 75645 3' ss, and pKY2 contains an optimized nt 75563 5' ss in K8 intron 2. These plasmids were constructed using an overlapping PCR technique (60). Basically, two primer sets were used to introduce mutations into each splice site. To amplify PCR fragments for the construction of plasmid pKY1, the primer set Pr74850 (oST41, 5'-CACC/ATGCCCAGAATGAAGGAC-3') and Pr75644 (oZMZ320, 5'-CTGAAAAGAAAGAAAGGTGGCGCCATTGTTCCCATTTGAG-3') and the primer set Pr75625 (oZMZ321, 5'-CCACCTTTCTTTCTTTTCAGGCATTAGAAGAAAAGGATGC-3') and Pr76687 (oST42, 5'-TCC/AGGGTTTCTTAC GCCG-3') were used. The amplified PCR products from each set of primers were mixed, heated at 95°C for 5 min, annealed at room temperature for 30 min, and reamplified using the primer Pr74850 in combination with the primer Pr76687. Following gel purification, the amplified DNA fragment was inserted into an expression vector, pcDNA3.1/V5-His-TOPO (Invitrogen). The same strategy was used to construct plasmid pKY2 with two different primer sets, Pr74850 in combination with Pr75589 (oZMZ 332, 5'-GATAGACACCTGTTT CCCGTACTTACCTGCTGCAGCTGTCT-3') and Pr75549 (oZMZ 331, 5'-AGACAGCTGCAGCAGGTAAGTACGGGAAACAGGTGTCTATC-3') in combination with Pr76687. The plasmid pKY3 is optimized at both the 75563 5' ss and the 75645 3' ss and was derived from pKY1 using the same primer sets as for the pKY2 construction. All optimization mutations described for pKY1, pKY2, and pKY3 were verified by sequencing.
The plasmid pKY4 was derived from plasmid pST3 (46), a full-length KSHV K8? cDNA expression vector, and has a human ?-globin intron inserted at the K8? exon 3-exon 4 junction. To introduce human ?-globin into the exon 3-exon 4 junction of K8?, a two-step overlapping PCR approach was taken. Briefly, the first step of the overlapping PCR was carried out with two separate PCR products: the K8 exon 4 product and a human ?-globin intron product obtained from two separate PCRs. K8 exon 4 was amplified from plasmid pST3 using a sense primer, Pr76433 (oZMZ 364, 5'-GATCATATTCATCTGGGGAACC-3'), in combination with an antisense primer, Pr76687. The human ?-globin intron was amplified from plasmid pSP64-H?6 (Promega, Madison, WI) using a K8 exon 3-?-globin chimeric sense primer, Pr146 (oZMZ 365, 5'-ACCATGCCAGACTTTGTGTG/GTTGGTATCAAGGTTACAAG-3'), in combination with a K8 exon 4-?-globin chimeric antisense primer, Pr275 (oZMZ 366, 5'-GGTTCCCCAGATGAATATGATC/CTAAGGGTGGGAAAATAGACC-3'). The two separate PCR products were then mixed for overlapping PCR using the sense primer Pr146 combined with the antisense primer Pr76687; the resulting PCR product had part of exon 3, a ?-globin intron, and exon 4. The first-step overlapping PCR was then mixed with another PCR product obtained from plasmid pST3 using the sense primer Pr74850 in combination with an antisense primer, Pr75838 (oZMZ 367, 5'-CACACAAAGTCTGGCATGGTTC-3'), which contained all K8? sequences except exon 4. The second step of the overlapping PCR was performed on the mixture using the primer pair Pr74850 and Pr76687, followed by gel purification. The amplified product was then cloned into the expression vector pcDNA3.1/V5-His-TOPO. The plasmid inserts were verified by sequencing.
RT-PCR. Total cell RNA was extracted from cells by using TRIzol reagent (Invitrogen) according to the manufacturer's instructions. Following DNase I digestion, reverse transcription of 1 μg of total RNA was performed at 42°C for 1 h using random hexamers and then amplified by PCR with AmpliTaq DNA polymerase (Roche, Indianapolis, IN) using different primer sets as described in each figure. The following list includes the primer sequences, which are named by the locations of their 5' ends. Pr74259 (oZMZ317, 5'-GACCATCGG AGGCACAAC-3'), Pr74565 (oST43, 5'-ACGAGGTACAGGAGTCCG-3'), Pr75182-T7 (oST1, 5'-TAATACGACTCACTATAGGG/ACCACCAAGA GGACCACACATTTC-3'), and Pr75354 (oST23, 5'-GAACTGACAAGGGGT GACGC-3') are forward primers; Pr75504 (oST4, 5'-GCGTGGGAGAATGTG ACT-3'), Pr75813 (oST32, 5'-CGGACGACGGCACAACCA-3'), Pr75838 (oST3, 5'-GTACTCACCCC/CACACAAAGTCTGGCATGGTTCTCCC-3'), and Pr 76576 (oST31, 5'-TGTAGTGCGCGTCTCTTCCTC-3') are reverse primers. The final PCR products were detected using agarose gel electrophoresis and ethidium bromide staining and analyzed with a ChemiDoc system (Bio-Rad, Hercules, CA). All RT-PCR products were gel purified and confirmed by sequencing through the splicing junction.
Human GAPDH RNA in each RNA sample was amplified by RT-PCR using a sense primer, Pr6732 (oZMZ269, 5'-GTCATCAATGGAAATCCCATCACC-3'), in combination with an antisense primer, Pr7207 (oZMZ257, 5'-TAATACGACTCACTATAGGGA/CCGTTCAGCTCAGGGATG-3').
In vitro RNA splicing assay. DNA templates were obtained separately from plasmids pST1, pKY1, pKY2, and pKY3 by PCR using a sense chimeric SP6/KSHV primer, Pr75484 (oZMZ335, 5'-ATTTAGGTGACACTATAG/ATC AGTCACATTCTCCCACGC-3'), combined with the antisense primer Pr75813. In vitro transcription using SP6 RNA polymerase was performed on each template DNA in the presence of a cap analog (m7GpppG) and [-32P] GTP. An in vitro RNA splicing assay was then conducted on each transcript using a HeLa cell nuclear extract in an optimized splicing condition. The spliced products were resolved in an 8% denaturing polyacrylamide gel electrophoresis (PAGE) gel and analyzed as described previously (58).
RPA. Antisense RNA probes were prepared as described above by in vitro transcription from individual DNA templates obtained from plasmids pST1, pST3, pKY1, pKY2, pKY3, and pKY4 by PCR using a sense primer, Pr75605 (oST36, 5'-CTCAAATGGGAACAATGGCG-3'), combined with an antisense chimeric T7/KSHV primer, Pr75935 (oST33, 5'-TAATACGACTCACTATAGGG/CGAATCTGTGTGGAACTCA-3'). A human cyclophilin probe was transcribed from a commercially available DNA template (Ambion, Austin, TX) and used as an internal control. For RPA, 100 μg of total cell RNA was hybridized with 4 ng of an antisense RNA probe in hybridization buffer at 42°C overnight, followed by RNase (RNase A and T1) digestion as instructed by the manufacturer of the RPA III kit (Ambion). Protected RNA fragments were separated in an 8% denaturing PAGE gel. A gel image was captured using a Molecular Dynamic PhosphorImager Storm 860 and analyzed with ImageQuant software.
RNase H cleavage protection assay and U1 and U2 depletion. Individual single-strand DNA oligonucleotides complementary to the exon-intron junctions of K8 intron 2 were used for oligonucleotide-directed RNase H digestion assays. The following antisense DNA oligonucleotides were used: oligonucleotide oKY2 (5'-TCTATACCTGCTG-3'), complementary to the exon 2-intron 2 junction of wild-type (wt) K8 RNA transcribed from plasmid pST1, and oKY4 (5'-TAC TTACCTGCTG-3'), complementary to the same region of mutant (mt) K8 RNA transcribed from plasmid pKY3; and oKY3 (5'-AAGACAGCAAGGT-3'), complementary to the intron 2-exon 3 junction of wt K8 RNA transcribed from plasmid pST1, and oKY5 (5'-AAGAAAGAAAGGT-3'), complementary to the same region of mt K8 RNA transcribed from plasmid pKY3. A RNase H digestion reaction was carried out in 20 μl of reaction mixture containing 40% HeLa nuclear extract, 2.5 mM MgCl2, 1 U of RNasin, and 4 ng of 32P-labeled pre-mRNA as described previously (61). Each reaction mixture was incubated for 10 min at 30°C in the absence of a DNA oligonucleotide and then incubated for another 10 min in the presence of an oligonucleotide (2 μM) before digestion for 10 min with 2 or 4 U of RNase H (Promega). The digestion was terminated by adding 3 volumes of proteinase K solution (5 mg/ml proteinase K, 50 mM EDTA, 0.5% sodium dodecyl sulfate [SDS]) (10), incubated at 75°C for 30 min, and extracted with phenol-chloroform. The RNA cleavage products were then resolved in an 8% polyacrylamide gel containing 8 M urea. The RNase H digestion efficiency for each pre-mRNA was calculated as the total amount of digestion products divided by the sum of the total digestion products plus the remaining pre-mRNA.
The oligonucleotide-directed RNase H digestion described above was also used to deplete small nuclear RNA U1 and U2 from the HeLa cell nuclear extract. Briefly, a human U1 (oKY6, 5'-TGCCAGGTAAGTA-3') or U2 (oKY7, 5'-AGGCCGAGAAGCGA-3') oligonucleotide was incubated at 30°C for 10 min with the RNase H reaction mixture described above in the absence of a 32P-labeled pre-mRNA. After the addition of RNase H (2 U) to the reaction, the incubation was extended for another 10 min. A 32P-labeled pre-mRNA and a specific K8 oligonucleotide (oKY2 or oKY3 for pST1; oKY4 or oKY5 for pKY3) were added, and the mixture was incubated for 10 min at 30°C before proteinase K digestion.
Western blot analysis. 293 cells or HeLa cells with or without transfection and JSC-1 or BCBL-1 cells with or without butyrate or TPA induction were washed twice with PBS and lysed in 2x SDS protein gel loading solution (Quality Biological, Inc., Gaithersburg, MD) containing 10% ?-mercaptoethanol or lysed in radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) (Boston BioProducts, Ashland, MA). If RIPA buffer was used, the cell lysates were first incubated on ice for 30 min and then sonicated several times (5 s each) before centrifugation for 10 min at 10,000 rpm. The supernatant was mixed with 2x SDS sample buffer containing 10% ?-mercaptoethanol. Protein samples were separated on a 4% to 12% Bis-Tris gel (Invitrogen) and transferred onto a nitrocellulose membrane. The membrane was blocked for 2 h at room temperature in a blocking buffer containing 5% skim milk in Tris-buffered saline (TBS) (10 mM Tris, 150 mM NaCl [pH 7.4]) and then incubated for 90 min with a 1:2,000 dilution of rabbit polyclonal anti-K-bZIP antibody (a gift from Don Ganem at the University of California, San Francisco), a 1:500 dilution of mouse monoclonal anti-human p21 antibody (BD Pharmingen, San Diego, CA), or a 1:100 dilution of mouse monoclonal anti-human p53 antibody in TTBS buffer (10 mM Tris, 150 mM NaCl, 1% Tween 20) containing 5% skim milk, after which the membrane was washed with TTBS three times (10 min each time). The membrane was then incubated for 60 min at room temperature with a 1:20,000 dilution of horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin G antibody for K-bZIP detection or with a 1:10,000 dilution of horseradish peroxidase-conjugated goat anti-mouse immunoglobulin G antibody for p21 and p53 detection and visualized with a SuperSignal West Pico chemiluminescence detection system (Pierce, Rockford, IL). The same membrane was stripped in a stripping buffer (Pierce) and reprobed using a polyclonal goat anti-CDK2 antibody (BD PharMingen) for immunoprecipitated CDK2 detection or using a monoclonal anti-?-tubulin antibody (BD PharMingen) for sample loading.
p53 reporter assays. HeLa cells (5 x 105) were transfected with 2 μg of pST1 (mainly expressing K8) or pST3 (K8?) in the presence of vector pcDNA3 (2 μg) or cotransfected with both pST1 (2 μg) and pST3 (2 μg) using Lipofectamine 2000 (Invitrogen). After 48 h of incubation, the transfected HeLa cells were transfected again with 2 μg of a p53 reporter plasmid (PG13-Luc, a gift from John Brady of the National Cancer Institute) (31) and incubated for additional 24 h at 37°C and 5% CO2. The cells were then collected and washed twice with calcium- and magnesium-free PBS. Part of the cell pellets were resuspended in 0.2 ml of Promega reporter lysis buffer at room temperature for 15 min. After centrifugation, 100 μl of the supernatant was collected for luciferase assay (Promega). Relative promoter activity was calculated by the light unit readings obtained from each plasmid-transfected cell lysate divided by the light unit readings obtained from pcDNA3-transfected cell lysate. The other part of the cell pellet was then lysed in RIPA buffer on ice for 30 min with sonication (several times, for 5 s each time). After centrifugation for 10 min at 10,000 rpm, the supernatant was mixed with 2x SDS sample buffer containing 10% ?-mercaptoethanol for the p53 Western blot assay.
CDK2 immunoprecipitation. One hundred microliters of a total of 700 μl of cell lysate in RIPA buffer obtained from HeLa cells (1 x 107 cells) transfected with pST1 or pST3 or pST1 and pST3 in combination was precleaned with 20 μl of a 50% slurry of protein A-Sepharose beads (Upstate, Waltham, MA) for 3 h at 4°C and washed twice with 1x immunoprecipitation (IP) buffer (59). After centrifugation at 13,000 rpm for 3 min, 20 μl of the supernatant was collected and incubated with 10 μl of anti-CDK2-coated beads (BD PharMingen) in 500 μl of 1x IP buffer overnight at 4°C. The beads were then gently washed four times with 1x IP buffer, and the proteins bound on the beads were eluted with 2x SDS sample buffer containing 10% ?-mercaptoethanol. Protein samples were separated on a 4 to 12% Bis-Tris gel (Invitrogen) and transferred onto a nitrocellulose membrane for Western blotting as described above.
RESULTS
Alternative splicing of ORF50 and K8 pre-mRNAs produces and ?, but not , mRNAs during KSHV lytic gene expression. In our previous study, we demonstrated that alternative splicing of K8 pre-mRNA produces two major forms of the spliced products ( and ?). The form differs from the ? by the exclusion of intron 2. That is, the K8 mRNA is a fully spliced mRNA, whereas the K8? is a partially spliced mRNA that retains intron 2. Since the 3' half of the ORF50 transcript overlaps the K8 transcript and has the same sequence composition as the entire K8 transcript (Fig. 1A), the part of the and ? isoforms of the K8 transcripts detected from KSHV+ cells must be spliced products of ORF50. However, exon 2 of ORF50 pre-mRNA has a size of 2,751 nt, which is too large to be recognized as an internal exon by the cellular splicing machinery (42). Whether such an exon structure of ORF50 could permit cross talking between intron 1 and intron 2 over exon 2 and therefore restrain ORF50 pre-mRNA from splicing has been questioned. In this study, RT-PCR was performed with total cell RNA purified from butyrate-stimulated JSC-1 cells with a primer from ORF50 exon 2 in combination with the primer Pr75504 (positioned in ORF50 exon 3 or K8 exon 2). The RT-PCR analysis demonstrated that splicing of ORF50 pre-mRNAs was efficient and did not retain intron 2 (K8 intron 1) (Fig. 1B), suggesting that a unknown function residing in ORF50 exon 2 could promote recognition of an oversized internal exon by cellular splicing machinery.
Previous reports from other laboratories have suggested that there is a isoform of the ORF50 and K8 message due to alternative RNA splicing in lytic KSHV gene expression (23, 63). In these reports, the RNA is depicted as retaining two introns, one from nt 75323 to 75471 and another from nt 75563 to 75645, but lacking the intron from nt 75838 to 76433 due to RNA splicing. A computer analysis of all three of these introns in our laboratory suggested that both the intron from nt 75323 to 75471 and the intron from nt 75838 to 76433 contain consensus splice sites and should be efficiently recognized by the cellular splicing machinery; it would be unusual for the cellular splicing machinery to miss recognizing an intron with consensus splice site sequences. In fact, we found that splicing of the intron from nt 75323 to 75471 is efficient. To further test this observation, we carried out an RT-PCR strategy to detect a RNA in butyrate-stimulated JSC-1 cells and TPA-stimulated BCBL-1 cells using an intron primer, Pr75354, from the intron region of nt 75323 to 75471, in combination with an antisense exon primer, Pr75813 or Pr76576 (Fig. 1C). The predicted RNA would be 636 nt in size when primer Pr76576 was paired with primer Pr75354 for the amplification. However, total cell RNA prepared from butyrate-stimulated JSC-1 cells did not have a band of this size (Fig. 1C, lane 6), despite the existence of RNAs containing the intron from nt 75323 to 75471 and the intron from nt 75563 to 75645 (Fig. 1C, lanes 3, 5, and 8) as amplified with a primer pair of Pr75813 and Pr75354. The same was true for BCBL-1 cells (data not shown). The RNA containing these two introns but lacking a 3' terminal exon sequences may more likely be a partially transcribed, unspliced pre-mRNA of ORF50 and/or K8.
KSHV K8? RNA is not subject to NMD and encodes a K8? protein during lytic infection. K8 intron 2 (positioned from nt 75563 to 75645) contains a premature termination codon (PTC), and its retention in the K8? mRNA introduces a PTC into the mRNA. K8? should therefore be subject to NMD. As currently understood, NMD serves as a posttranscriptional control system by which cells recognize and degrade mRNAs carrying a PTC positioned >50 nt upstream of a splice junction to prevent the possible toxic effects of truncated peptides (25). It has been shown that translation or a translation-like process is a prerequisite for NMD, and inhibition of translation by CHX results in the abolishment of NMD and, subsequently, the accumulation of the PTC+ mRNA (5, 33). We therefore compared the quality and quantity of the K8? RNA by RPA in 293 cells transfected with a K8 expression vector and in butyrate-activated KSHV+ JSC-1 cells with or without CHX treatment (Fig. 2). The expected RPA products are shown in Fig. 2A. We postulated that cells treated with CHX would have more K8? mRNA due to a minimized NMD response. However, K8 and K8? were present in very similar amounts in the presence or absence of CHX in 293 cells transfected with the K8 expression vector (pST1, lanes 6 to 9), suggesting no accumulation of K8? and therefore no NMD-mediated K8? mRNA degradation. In the butyrate-activated JSC-1 cells, CHX treatment inhibited both K8 and K8? gene expression (Fig. 2B, lanes 3 to 5). This was not surprising, since several studies have reported that expression of K8 in KSHV+ BCBL-1, BC-1, and JSC-1 cells is sensitive to CHX treatment (44, 63). Thus, it was impossible to draw any conclusion on NMD-mediated K8? degradation at the RNA level in JSC-1 cells. However, we were able to detect K8? protein from both JSC-1 and BCBL-1 cells after butyrate and TPA stimulation (Fig. 2C), respectively, suggesting that K8? RNA resists NMD-mediated RNA degradation in those cells and is capable of being translated.
Optimization of K8 intron 2 splice sites promotes splicing of K8 pre-mRNA. As proposed in our previous report (46), the inclusion of intron 2 in K8? mRNA could result from its suboptimal nature, since this intron contains a nonconsensus 5' ss and a weak 3' ss. The sequence of the 5' ss of intron 2 misses pairing a few nucleotides from the U1 5' end. Most importantly, the 5' ss misses two nucleotides, at intron positions 4 and 5, for base pairing to the six core nucleotides of the U1 5' end (Fig. 3A, top). A missing base pairing of a U1 snRNA 5' core nucleotide to the RNA 5' ss at intron position 5 has been reported to inhibit U1 snRNA binding and splicing (15, 61). The sequences of the intron 3' ss contain a few purine Gs interspersed in its putative polypyrimidine tract (PPT), which lies between the branch point sequence (BPS) and the AG dinucleotide (Fig. 3A, bottom). Recognition of the BPS by U2 snRNA initiates the first step of RNA splicing. The PPT is a run of 15 to 40 pyrimidines (usually U's) and contains binding sites for U2AF, a heterodimer of U2AF65 and U2AF35. Binding of U2AF to the PPT strengthens U2 recognition of the BPS (57). It has been reported that any purines interspersed in the PPT will weaken the binding of U2AF and affect U2 interaction with the BPS. Moreover, the nature of the nucleotide preceding the AG dinucleotide has a profound effect on splicing (39), and CAG is twice as common as UAG in mammalian 3' ss (55). Together, these features of K8 intron 2 would be expected to decrease the inherent strength of its recognition by U1 and U2 snRNAs.
To test our hypothesis, we constructed three K8 expression vectors. pKY1 has an optimized 3' ss (Fig. 3A, bottom, mt 3' ss), pKY2 has an optimized 5' ss (Fig. 3A, middle), and pKY3 has optimized 3' ss and 5' ss. Following transient transfection of 293 cells with these expression vectors, we used RT-PCR and RPA to examine the expression of K8 and K8? in parallel with the expression of a wild-type K8 expression vector, pST1. As shown in Fig. 3B and C, the optimization of K8 intron 2 increased the splicing efficiency of the K8 pre-mRNA and resulted in at least twofold reduction of K8? production in both assays. The data suggest that strengthening either the 3' ss or 5' ss in K8 intron 2 improved the recognition of the intron by the cellular splicing machinery and led to the splicing of intron 2.
In vitro splicing of pre-mRNAs transcribed from the optimized and wt plasmids showed similar results. The pre-mRNA transcribed from wt K8 had a poor splicing efficiency (only 2%) (Fig. 3D, pre-mRNA 1), while the optimized pre-mRNAs exhibited an increased (>9-fold) splicing efficiency (Fig. 3D, pre-mRNAs 2 to 4). Notably, the pre-mRNA with both the 5' ss and 3' ss optimized showed the highest splicing efficiency (13-fold greater than the wild type) (Fig. 3D).
The escape of K8 intron 2 from splicing is due to its inefficient recognition by U1 snRNA and U2AF. To understand further how intron 2 escapes RNA splicing, the RNAs transcribed from pST1 (wt K8) and pKY3 (K8 with both of the intron 2 splice sites optimized) were also used for a DNA oligonucleotide-directed RNase H digestion assay. As mentioned above, we would expect that the suboptimal features of intron 2 do not favor efficient interactions between U1 and the 5' ss or between U2AF and the 3' ss. If this prediction is right, the optimization of both splice sites should lead to increased U1-5' ss interactions and U2AF-3' ss interactions and, therefore, to increased splicing efficiency of intron 2.
RNase H digests only RNA-DNA hybrids in which a DNA oligonucleotide hybridizes to a specific RNA sequence through base pairing (8, 14). An RNA-RNA hybrid in the same region or a region of RNA with bound protein will prevent a DNA oligonucleotide from base pairing and will consequently prevent RNase H digestion. Based on this theory, we synthesized antisense DNA oligonucleotides (12 or 13 nt in length) specific to the 5' ss or the 3' ss of wt or mt K8 intron 2, as diagrammed in Fig. 4A. Using this approach, we were able to quantify a U1- 5' ss interaction and a U2AF-3' ss interaction in the presence of HeLa cell nuclear extracts containing enriched U snRNAs and splicing factors. As shown in Fig. 4B, the wt 5' ss and 3' ss of the K8 intron 2 have a weak binding affinity for U1 and U2AF, allowing more DNA oligonucleotides to bind to their corresponding splice sites; as a result, approximately 90% and 61% of the RNA was digested by RNase H directed by a 5' ss-specific DNA oligonucleotide and a 3' ss-specific DNA oligonucleotide, respectively. In contrast, RNase H digestion was much less effective when both the intron 5' ss and 3' ss were optimized: only about 38% of the RNA was subjected to the oligonucleotide-directed digestion of the RNA 5' ss, and approximately 3% of the RNA was subjected to the oligonucleotide-directed digestion of the RNA 3' ss.
To further confirm that the reduction in digestion of the RNA intron 2 splice sites by RNase H was specifically due to the improvement in the U1-5' ss and U2AF-3' ss interactions, we depleted either U1 or U2 snRNAs from HeLa nuclear extracts using RNase H digestion directed by an antisense U1 or U2 oligonucleotide. The U1- or U2-depleted nuclear extracts were then used in DNA oligonucleotide-directed digestion of wt or mt K8 intron 2. Since U2AF is abundant in HeLa nuclear extracts and complete depletion of U2AF would be difficult by antibody affinity or poly(U) approaches, we depleted U2 instead of U2AF from HeLa nuclear extracts on the assumption, based on the U2 recruitment model, that the RNA-U2AF interaction is dynamic and less U2AF would be fixed on the RNA in the absence of U2 snRNA (11, 12). Depletion of U1 from HeLa nuclear extracts did increase the oligonucleotide-directed RNase H digestion of both the wt (Fig. 4C, compare lanes 3 and 4) and the mt (Fig. 4D, compare lanes 3 and 4) K8 intron 2, indicating a specific U1-5' ss interaction. In addition, an increase in oligonucleotide-directed RNase digestion of the wt K8 intron 2 was observed when U2-depleted HeLa nuclear extracts were used (Fig. 4C, compare lanes 5 and 6). However, we were unable to see a similar result from the oligonucleotide-directed RNase H digestion of the 3' ss of mt K8 intron 2 (Fig. 4D, compare lanes 7 and 8), suggesting that optimization of the 3' ss might also create an additional binding site (UCUU) for another protein, most likely a PPT-binding protein (PTB) (29), which is unrelated to U2 recruitment.
Insertion of a human ?-globin intron into the K8? exon 3-exon 4 splice junction stimulates splicing of K8 intron 2 and converts K8? to K8. We have reported that K8 exon 3 contains three alternative 5' ss, at nt 75838, 76155, and 76338. Selection of the nt 75838 5' ss is subject to exon definition and predominates over the other two 5' ss during RNA splicing, even though all three 5' ss are optimal splice sites with consensus U1 binding sequences. In vivo experiments also suggest that utilization of the nt 75838 5' ss stimulates splicing of an upstream nt 75645 3' ss, crossing over exon 3 (46). Disruption of the nt 75838 5' ss promotes either the inclusion of intron 2 or the exclusion of exon 3 due to alternative splicing of exon 2 to exon 4 (data not shown). This implies that intron 3 5' ss recognition may play an important role in the inclusion and exclusion of K8 intron 2 and the production of K8? mRNA.
To test this role of downstream intron recognition in upstream intron splicing, we inserted a human ?-globin intron into the exon 3-exon 4 splice junction of a K8? cDNA construct, pST3, to create plasmid pKY4 (Fig. 5A). Since the human ?-globin intron has optimal splice sites, the splicing of this inserted intron was expected to stimulate removal of the K8? intron upstream. When pKY4 was used to transfect 293 cells, the splicing of the ?-globin intron in the pKY4-derived pre-mRNA did stimulate splicing of the upstream K8? intron, and much of the K8? RNA was converted into K8 mRNA (Fig. 5B).
Western blot analysis was performed with a polyclonal K-bZIP antibody to examine K8 and K8? protein expression simultaneously in the transfected 293 cells. As shown in Fig. 5C, a K8 genomic DNA construct, pST1, expressed mainly K8 protein in 293 cells, accompanied by a lower amount of K8? expression. In contrast, the K8? cDNA construct pST3 expressed K8? protein but very little K8 protein. The pKY4 construct, which had the ?-globin intron inserted into the pST3 exon 3-exon 4 splice junction, expressed almost equal amounts of K8 and K8? protein, as predicted (Fig. 5C).
K8? regulates K8 function. As described above, the inclusion of K8 intron 2 also creates a PTC in the K8? mRNA and results in the production of the truncated protein K8? in KSHV+ JSC-1 cells and BCBL-1 cells as well as in transient transfected epithelial cells. Thus, both K8 protein and K8? protein have the same N-terminal structure, but they differ at their C-terminal ends, where a leucine zipper domain in K8 is missing in K8?. Recent reports indicate that one of the functions of KSHV K8 during viral lytic replication in PEL cells is to induce expression of the universal CDK inhibitor p21 by stabilizing p21 from proteasome-mediated degradation (20, 52, 53). This observation prompted us to wonder whether the expression of K8? might antagonize the activity of K8 if this truncated protein cannot similarly induce p21.
To examine the biological functions of K8?, we investigated the effect of KSHV K8 and K8? by transient transfection on the expression of p21 initially in HeLa cells. We transiently transfected HeLa cells with a K8 expression vector that predominantly expresses K8 (pST1) or with the K8? expression vector (pST3) and found that only the pST1-transfected HeLa cells strongly induced p21 (Fig. 6A, lane 2). Although pST3 itself had no effect on p21 induction in HeLa cells (Fig. 6B), pST3 did suppress the pST1-induced p21 expression in a dose-dependent manner (Fig. 6A, compare lanes 3 to 5 and lane 2; Fig. 6C). At a 1:1 ratio of pST1 and pST3, the expression of both K8 and K8? was reduced (Fig. 6D, lane 5). Interestingly, this reciprocal suppression of K8 and K8? appeared only in the HeLa cells cotransfected with pST1 and pST3, not in cells cotransfected with pST1 or with pST3 and an empty vector, pcDNA3 (Fig. 6E and F, lanes 4). Thus, the expression of K8? antagonizes the function of K8.
Since p21 expression is usually induced by p53, it is conceivable that K8-mediated p21 accumulation might occur because of an accumulated p53 expression in cells overexpressing K8. To address this question, we examined p53 expression by using HeLa cells transfected with pST1 (K8), pST3 (K8?), or pST1 and pST3 in combination. HeLa cells were selected for the study, since they contain only a residual amount of p53. As shown in Fig. 7A and B, the overexpression of K8 (pST1) in HeLa cells was found to induce >2.5-fold activation activity of p53 in p53 reporter assays, which could be translated into a 71% higher p53 protein accumulation in the cells by Western blotting. Although the overexpression of K8? (pST3) had no effect on p53 accumulation, HeLa cells cotransfected with both K8 (pST1) and K8? (pST3) expression vectors showed no p53 accumulation or activation activity but, rather, p53 reduction. Data suggest that the K8-mediated p21 accumulation and its suppression by K8? may be attributable partially to K8-mediated p53 expression.
A recent study also suggests that the inhibition of CDK2 activity by K-bZIP through direct CDK2-K-bZIP interaction is partially responsible for the increased p21 protein level (20), since phosphorylation by CDK2 targets p21 for degradation (38). We wondered whether K8? could also affect the interaction of K8 and CDK2 and examined this interaction by using HeLa cells transfected with pST1 (K8) or pST3 (K8?) or pST1 plus pST3 in combination. The cell lysates were immunoprecipitated with anti-CDK2-coated beads, and the precipitates were immunoblotted with anti-K-bZIP. As shown in Fig. 7C, K8 (pST1) did interact with CDK2 (lane 2), but this interaction could be completely eliminated in the presence of K8? (pST3, lane 3) despite the fact that the latter had no detectable interaction with CDK2 (lane 4). These data, together with the p53 results described above, suggest that K8 mediates the increase in p21 expression, presumably through several pathways.
K8? affects K8 expression at the mRNA level. As stated above, K8? differs from K8 by lacking a C-terminal leucine zipper motif which is required for K8 dimerization and functions. Thus, it is very unlikely that K8? could form a dimer with K8 as a mechanism of the K8?-mediated suppression. Considering the fact that K8? also reduces K8 protein expression in HeLa cell cotransfection assays (Fig. 6D, lane 5), we were interested to know if the K8?-mediated K8 protein reduction is a result of K8?-mediated suppression of K8 RNA expression, since K8? has the same exon structures in K8 mRNA and differs from K8 only by the inclusion of a K8 intron 2 as an extra piece in its mRNA. As demonstrated in Fig. 8, the cotransfection of HeLa cells with a K8? expression vector (pST3) resulted in an approximately fivefold reduction of K8 mRNA (pST1) in RT-PCR assays (Fig. 8A, compare lanes 7 and 8). To further confirm our RT-PCR results, total cell RNA obtained from separate experiments was analyzed by RPA with an antisense K8 probe and showed a sixfold reduction of K8 mRNA (pST1) in the presence of K8? (pST3) (Fig. 8B, compare lanes 6 and 8). Although K8 mRNA expression was reduced greatly in the presence of K8?, the K8? mRNA level remained relatively constant in the presence of K8 (compare lanes 6 and 9). Together, the results from two different assays lead us to conclude that K8?-mediated suppression of K8 function is most likely a consequence of K8?-mediated inhibition of K8 RNA expression.
Since KSHV+ B cells express both K8 and K8? via alternative RNA splicing during lytic infection, it is difficult to dissect the function of K8 from K8? during lytic infection in those cells. More importantly, the K8 gene overlaps the ORF50 gene in the virus genome despite the fact that each gene has its own ORF, as shown in Fig. 1A. Accordingly, it is not feasible to study K8 function in its native context without disruption of the overlapped ORF50. To duplicate our observation with HeLa cells, we analyzed K8? on K8 function with BCBL-1 cells by transient transfection. As shown in Fig. 9A, TPA activation of BCBL-1 cells (lane 2) or expression of KSHV ORF50 (lane 4) or K8 (lane 5, pST1) alone in BCBL-1 cells promotes p21 expression. The expression of K8? alone (lane 7, pST3) in BCBL-1 cells did not induce p21 expression, as was done by its vector control (pcDNA, lane 3). As expected, we were able to demonstrate that the expression of K8? could antagonize K8-mediated induction of p21 (Fig. 9A, compare lanes 5 and 6) in BCBL-1 cells. In parallel, BCBL-1 cotransfection with pST1 (K8) and pST3 (K8?) had much less production of K8 mRNAs (Fig. 9B, lane 7) than the cells transfected with pST1 alone (Fig. 9B, lanes 6), implying that K8? diminishes K8-mediated function most likely at the RNA level.
DISCUSSION
In this report, we have analyzed alternative splicing of two overlapping primary transcripts expressed during the KSHV lytic cycle, the ORF50 and K8 pre-mRNAs. As we (46) and others (13, 23, 63) previously reported, this alternative RNA splicing leads to the production of two major spliced isoforms ( and ?) of mRNAs due to the inclusion (?) or exclusion () of an intron from nt 75563 to 75645 (intron 3 of the ORF50 pre-mRNA and intron 2 of the K8 pre-mRNA). However, we were unable to detect a previously reported isoform message (23, 63).
The intron from nt 75563 to 75645 (K8 intron 2 or ORF50 intron 3) has suboptimal features, including a nonconsensus 5' ss and a weak 3' ss with purines interspersed in its putative polypyrimidine tract. As a result, K8 intron 2 interacts with U1 snRNA and U2AF protein with low affinity. The optimization of these splice sites was found to improve the interactions with U1 snRNA and U2AF and, consequently, to promote the splicing of K8 intron 2 in vitro and in vivo. Our data provide strong experimental evidence that the escape of K8 intron 2 (ORF50 intron 3) from RNA splicing can most likely be attributed to its suboptimal features. The mRNA that results from the retention of intron 2, K8?, contains a PTC in the intron at a distance of >50 nt upstream of an exon-exon junction. Despite having this PTC, the K8? mRNA bypasses an mRNA surveillance mechanism that initiates NMD to rapidly degrade mRNAs carrying PTCs. This was surprising, since a major function of NMD is to block the synthesis of truncated proteins that could have dominant negative effects (25). Although there are many other characteristics of mammalian transcripts that determine the efficiency of NMD, including a nonsense codon positioned close to the initiation AUG (4, 18, 30, 54), the absence of an NMD-promoting element (3), and the presence of NMD-mediated up-regulation of RNA splicing (48), we do not know how the K8? mRNA avoids NMD. Nevertheless, in PEL cells and transiently transfected mammalian cells, the K8? mRNA produces a K8? protein that is 48 amino acids smaller than the K-bZIP protein encoded by fully spliced K8 mRNA. This led us to speculate that the PTC-containing K8? might avoid the NMD pathway because the encoded protein has a function in viral gene expression.
Despite the suboptimal features of the intron from nt 75563 to 75645, the majority of the K8 pre-mRNAs completely remove this intron to produce a fully spliced K8 message, which encodes a K-bZIP protein important for viral gene transcription (19, 21, 28, 49), DNA replication (1, 22, 51), and cell cycle arrest (20, 52, 53). To investigate what promotes the splicing of this intron, we demonstrated that a cis element, a nt 75838 5' ss crossing over exon 3, could trigger the splicing. This finding is consistent with a well-established exon definition proposing that cross talk between a 3' ss and a 5' ss over the exon promotes splicing of an upstream intron (2, 17, 42, 57). Based on this principle, we inserted a human ?-globin intron into the K8? exon 3-exon 4 splice junction and demonstrated that this intron insertion stimulated efficient splicing of the K8? intron and resulted in production of K8 mRNA from a K8? cDNA. Together, our RNA and protein expression data provide conclusive evidence that the splicing of K8 intron 2 is enhanced by the exon definition machinery crossing over exon 3 and that the K8? RNA is derived from an RNA splicing intermediate during viral RNA processing.
The production of K8? and its escape from NMD-mediated degradation in early viral lytic infection suggest that the expression of K8? may play a role in viral gene expression and pathogenesis. The K8? protein has exactly the same amino acid sequences as the K-bZIP (K8) protein in its N-terminal region, including a basic domain close to the leucine zipper domain, but is missing a C-terminal leucine zipper domain encoded by the K8 exon 3 region (23). Recent studies have shown that the KSHV K-bZIP protein increases p21 expression and causes p21-mediated G1 cell cycle arrest through CCAAT/enhancer-binding protein- (C/EBP) (52, 53) and/or inhibition of CDK2 (20). The K-bZIP protein binds very efficiently to both C/EBP and p21 and stabilizes them from proteasome-mediated protein degradation. Although the only region of the K-bZIP protein that interacts with and stabilizes C/EBP is the basic leucine zipper region in the C terminus, both the N-terminal and C-terminal domains of the K-bZIP protein are required to stimulate C/EBP-induced expression from both the C/EBP and p21 promoters in cotransfected cells (53). In contrast, only the basic region (residues 123 to 189), not the leucine zipper region, of K-bZIP was found to play a functional role in CDK2 binding and down-modulation of CDK2 activity (20). Consistent with these reports, we also demonstrated in our transient transfection assays that expression of the K-bZIP protein could cause both p53 and p21 induction. However, coexpression of K-bZIP and K8? in the cells reduced the K-bZIP-mediated p53 and p21 induction, suggesting that K8? acts as an antagonist. Moreover, we further demonstrated that the expression of K8? also blocks the interaction of K-bZIP with CDK2. This is the first report of a biological function for K8?.
Although enhancement of p21 expression by K-bZIP can be induced in a p53-independent pathway (53), the finding of K-bZIP interaction with p53 (20, 28) in other laboratories also provides some convincing evidence for interpreting our results, indicating that K-bZIP-p53 interaction might stabilize p53 from degradation and result in accumulation of the p53 that activates transcription of the p21 gene. More interestingly, we found a significant reduction of p21 expression, along with a reciprocal suppression of K-bZIP and K8? expression, when equal amounts of expression vector pST1 (mainly for K-bZIP expression) and pST3 (for K8?) were cotransfected. In this case, the expression levels of the K8 mRNAs in the cotransfected cells were greatly reduced along with an unrestricted expression of K8? mRNA, suggesting that overexpression of K8? RNA might affect splicing of K8 pre-mRNA. However, how K-bZIP suppresses K8? protein expression remains to be further understood. Altogether, these observations clearly indicate that expression of K8 and K8? must remain at a constant ratio. Production of K8? over a threshold is detrimental to K8 expression and consequently is subtly regulated by K8. Supporting this is the fact that it has been found consistently that K8 and K8? mRNA are expressed at approximately 5-6 to 1 (this study and Tang and Zheng [46]). It will be interesting to know how the virus keeps this ratio of K8 and K8? mRNAs in check during viral gene expression and what controls this regulation, in addition to the exon definition we described in this study.
In summary, we have investigated in detail the features of an intron from KSHV nt 75563 to 75645 and its inclusion in a ? isoform mRNA during KSHV ORF50 and K8 expression. Because the 3' UTR of the ORF50 transcript overlaps the entire K8 transcript, we simplified our focus on understanding alternative splicing of this intron in processing of the K8 pre-mRNA to exemplify the way the inclusion or exclusion of this intron may take place during the splicing of the ORF50 pre-mRNA. Both the and ? isoforms of the ORF50 mRNA presumably encode the same protein (replication and transcription activator; RTA) with the same size, since the intron from nt 75563 to 75645 in the ORF50 transcript is positioned in its 3' UTR; however, the inclusion of this intron in the K8? mRNA leads to the production of a truncated protein that plays a dominant negative role in K8 mRNA expression and K8 function.
ACKNOWLEDGMENTS
This study was supported by the Intramural Research Program of the NIH, National Cancer Institute, Center for Cancer Research. K.Y. was supported by a research fellowship from the Japan Society for the Promotion of Science at NIH.
We thank Richard Ambinder for providing JSC-1 cells and the NIAID AIDS Research and Reference Reagent Program, Division of AIDS, for providing BCBL-1 cells originally obtained from Michael McGrath and Don Ganem. We also thank Don Ganem for providing the rabbit polyclonal anti-K-bZIP antibody, John Brady for p53 reporter plasmid, and Yan Yuan for pORF50 expression vector.
REFERENCES
AuCoin, D. P., K. S. Colletti, S. A. Cei, I. Papouskova, M. Tarrant, and G. S. Pari. 2004. Amplification of the Kaposi's sarcoma-associated herpesvirus/human herpesvirus 8 lytic origin of DNA replication is dependent upon a cis-acting AT-rich region and an ORF50 response element and the trans-acting factors ORF50 (K-Rta) and K8 (K-bZIP). Virology 318:542-555.
Berget, S. M. 1995. Exon recognition in vertebrate splicing. J. Biol. Chem. 270:2411-2414.
Bühler, M., A. Paillusson, and O. Mühlemann. 2004. Efficient downregulation of immunoglobulin μ mRNA with premature translation-termination codons requires the 5'-half of the VDJ exon. Nucleic Acids Res. 32: 3304-3315.
Buzina, A., and M. J. Shulman. 1999. Infrequent translation of a nonsense codon is sufficient to decrease mRNA level. Mol. Biol. Cell 10:515-524.
Carter, M. S., J. Doskow, P. Morris, S. Li, R. P. Nhim, S. Sandstedt, and M. F. Wilkinson. 1995. A regulatory mechanism that detects premature nonsense codons in T-cell receptor transcripts in vivo is reversed by protein synthesis inhibitors in vitro. J. Biol. Chem. 270:28995-29003.
Chandran, B., C. Bloomer, S. R. Chan, L. Zhu, E. Goldstein, and R. Horvat. 1998. Human herpesvirus-8 ORF K8.1 gene encodes immunogenic glycoproteins generated by spliced transcripts. Virology 249:140-149.
Chang, Y., E. Cesarman, M. S. Pessin, F. Lee, J. Culpepper, D. M. Knowles, and P. S. Moore. 1994. Identification of herpesvirus-like DNA sequences in AIDS-associated Kaposi's sarcoma. Science 266:1865-1869.
Donis-Keller, H. 1979. Site specific enzymatic cleavage of RNA. Nucleic Acids Res. 7:179-192.
Fakhari, F. D., and D. P. Dittmer. 2002. Charting latency transcripts in Kaposi's sarcoma-associated herpesvirus by whole-genome real-time quantitative PCR. J. Virol. 76:6213-6223.
Frilander, M. J., and J. A. Steitz. 1999. Initial recognition of U12-dependent introns requires both U11/5' splice-site and U12/branchpoint interactions. Genes Dev. 13:851-863.
Graveley, B. R. 2000. Sorting out the complexity of SR protein functions. RNA 6:1197-1211.
Graveley, B. R., K. J. Hertel, and T. Maniatis. 2001. The role of U2AF35 and U2AF65 in enhancer-dependent splicing. RNA 7:806-818.
Gruffat, H., S. Portes-Sentis, A. Sergeant, and E. Manet. 1999. Kaposi's sarcoma-associated herpesvirus (human herpesvirus-8) encodes a homologue of the Epstein-Barr virus bZip protein EB1. J. Gen. Virol. 80:557-561.
Gunzl, A., Z. Palfi, and A. Bindereif. 2002. Analysis of RNA-protein complexes by oligonucleotide-targeted RNase H digestion. Methods 26:162-169.
Hamm, J., and I. W. Mattaj. 1990. Monomethylated cap structures facilitate RNA export from the nucleus. Cell 63:109-118.
Hayward, G. S. 2003. Initiation of angiogenic Kaposi's sarcoma lesions. Cancer Cell 3:1-3.
Hwang, D. Y., and J. B. Cohen. 1996. Base pairing at the 5' splice site with U1 small nuclear RNA promotes splicing of the upstream intron but may be dispensable for slicing of the downstream intron. Mol. Cell. Biol. 16: 3012-3022.
Inacio, A., A. L. Silva, J. Pinto, X. Ji, A. Morgado, F. Almeida, P. Faustino, J. Lavinha, S. A. Liebhaber, and L. Romao. 2004. Nonsense mutations in close proximity to the initiation codon fail to trigger full nonsense-mediated mRNA decay. J. Biol. Chem. 279:32170-32180.
Izumiya, Y., S. F. Lin, T. Ellison, L. Y. Chen, C. Izumiya, P. Luciw, and H. J. Kung. 2003. Kaposi's sarcoma-associated herpesvirus K-bZIP is a coregulator of K-Rta: physical association and promoter-dependent transcriptional repression. J. Virol. 77:1441-1451.
Izumiya, Y., S. F. Lin, T. J. Ellison, A. M. Levy, G. L. Mayeur, C. Izumiya, and H. J. Kung. 2003. Cell cycle regulation by Kaposi's sarcoma-associated herpesvirus K-bZIP: direct interaction with cyclin-CDK2 and induction of G1 growth arrest. J. Virol. 77:9652-9661.
Liao, W., Y. Tang, S. F. Lin, H. J. Kung, and C. Z. Giam. 2003. K-bZIP of Kaposi's sarcoma-associated herpesvirus/human herpesvirus 8 (KSHV/HHV-8) binds KSHV/HHV-8 Rta and represses Rta-mediated transactivation. J. Virol. 77:3809-3815.
Lin, C. L., H. Li, Y. Wang, F. X. Zhu, S. Kudchodkar, and Y. Yuan. 2003. Kaposi's sarcoma-associated herpesvirus lytic origin (ori-Lyt)-dependent DNA replication: identification of the ori-Lyt and association of K8 bZip protein with the origin. J. Virol. 77:5578-5588.
Lin, S. F., D. R. Robinson, G. Miller, and H. J. Kung. 1999. Kaposi's sarcoma-associated herpesvirus encodes a bZIP protein with homology to BZLF1 of Epstein-Barr virus. J. Virol. 73:1909-1917.
Lukac, D. M., J. R. Kirshner, and D. Ganem. 1999. Transcriptional activation by the product of open reading frame 50 of Kaposi's sarcoma-associated herpesvirus is required for lytic viral reactivation in B cells. J. Virol. 73: 9348-9361.
Maquat, L. E. 2004. Nonsense-mediated mRNA decay: splicing, translation and mRNP dynamics. Nat. Rev. Mol. Cell Biol. 5:89-99.
Miller, G., M. O. Rigsby, L. Heston, E. Grogan, R. Sun, C. Metroka, J. A. Levy, S. J. Gao, Y. Chang, and P. Moore. 1996. Antibodies to butyrate-inducible antigens of Kaposi's sarcoma-associated herpesvirus in patients with HIV-1 infection. N. Engl. J. Med. 334:1292-1297.
Moore, P. S., S. J. Gao, G. Dominguez, E. Cesarman, O. Lungu, D. M. Knowles, R. Garber, P. E. Pellett, D. J. McGeoch, and Y. Chang. 1996. Primary characterization of a herpesvirus agent associated with Kaposi's sarcomae. J. Virol. 70:549-558.
Park, J., T. Seo, S. Hwang, D. Lee, Y. Gwack, and J. Choe. 2000. The K-bZIP protein from Kaposi's sarcoma-associated herpesvirus interacts with p53 and represses its transcriptional activity. J. Virol. 74:11977-11982.
Perez, I., C. H. Lin, J. G. McAfee, and J. G. Patton. 1997. Mutation of PTB binding sites causes misregulation of alternative 3' splice site selection in vivo. RNA 3:764-778.
Perrin-Vidoz, L., O. M. Sinilnikova, D. Stoppa-Lyonnet, G. M. Lenoir, and S. Mazoyer. 2002. The nonsense-mediated mRNA decay pathway triggers degradation of most BRCA1 mRNAs bearing premature termination codons. Hum. Mol. Genet. 11:2805-2814.
Pise-Masison, C. A., R. Mahieux, M. Radonovich, H. Jiang, and J. N. Brady. 2001. Human T-lymphotropic virus type I Tax protein utilizes distinct pathways for p53 inhibition that are cell type-dependent. J. Biol. Chem. 276: 200-205.
Raab, M. S., J. C. Albrecht, A. Birkmann, S. Yaguboglu, D. Lang, B. Fleckenstein, and F. Neipel. 1998. The immunogenic glycoprotein gp35-37 of human herpesvirus 8 is encoded by open reading frame K8.1. J. Virol. 72:6725-6731.
Rajavel, K. S., and E. F. Neufeld. 2001. Nonsense-mediated decay of human HEXA mRNA. Mol. Cell. Biol. 21:5512-5519.
Renne, R., W. Zhong, B. Herndier, M. McGrath, N. Abbey, D. Kedes, and D. Ganem. 1996. Lytic growth of Kaposi's sarcoma-associated herpesvirus (human herpesvirus 8) in culture. Nat. Med. 2:342-346.
Russo, J. J., R. A. Bohenzky, M. C. Chien, J. Chen, M. Yan, D. Maddalena, J. P. Parry, D. Peruzzi, I. S. Edelman, Y. Chang, and P. S. Moore. 1996. Nucleotide sequence of the Kaposi sarcoma-associated herpesvirus (HHV8). Proc. Natl. Acad. Sci. USA 93:14862-14867.
Sakakibara, S., K. Ueda, J. Chen, T. Okuno, and K. Yamanishi. 2001. Octamer-binding sequence is a key element for the autoregulation of Kaposi's sarcoma-associated herpesvirus ORF50/Lyta gene expression. J. Virol. 75:6894-6900.
Reference deleted.
Sherr, C. J., and J. M. Roberts. 1999. CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev. 13:1501-1512.
Smith, C. W., E. B. Porro, J. G. Patton, and B. Nadal-Ginard. 1989. Scanning from an independently specified branch point defines the 3' splice site of mammalian introns. Nature 342:243-247.
Soulier, J., L. Grollet, E. Oksenhendler, P. Cacoub, D. Cazals-Hatem, P. Babinet, M. F. d'Agay, J. P. Clauvel, M. Raphael, L. Degos, and F. Sigaux. 1995. Kaposi's sarcoma-associated herpesvirus-like DNA sequences in multicentric Castleman's disease. Blood 86:1276-1280.
Staskus, K. A., R. Sun, G. Miller, P. Racz, A. Jaslowski, C. Metroka, H. Brett-Smith, and A. T. Haase. 1999. Cellular tropism and viral interleukin-6 expression distinguish human herpesvirus 8 involvement in Kaposi's sarcoma, primary effusion lymphoma, and multicentric Castleman's disease. J. Virol. 73:4181-4187.
Sterner, D. A., T. Carlo, and S. M. Berget. 1996. Architectural limits on split genes. Proc. Natl. Acad. Sci. USA 93:15081-15085.
Sun, R., S. F. Lin, L. Gradoville, Y. Yuan, F. Zhu, and G. Miller. 1998. A viral gene that activates lytic cycle expression of Kaposi's sarcoma-associated herpesvirus. Proc. Natl. Acad. Sci. USA 95:10866-10871.
Sun, R., S. F. Lin, K. Staskus, L. Gradoville, E. Grogan, A. Haase, and G. Miller. 1999. Kinetics of Kaposi's sarcoma-associated herpesvirus gene expression. J. Virol. 73:2232-2242.
Tang, S., K. Yamanegi, and Z. M. Zheng. 2004. Requirement of a 12-base-pair TATT-containing sequence and viral lytic DNA replication in activation of the Kaposi's sarcoma-associated herpesvirus K8.1 late promoter. J. Virol. 78:2609-2614.
Tang, S., and Z. M. Zheng. 2002. Kaposi's sarcoma-associated herpesvirus K8 exon 3 contains three 5'-splice sites and harbors a K8.1 transcription start site. J. Biol. Chem. 277:14547-14556.
Viejo-Borbolla, A., M. Ottinger, and T. F. Schulz. 2003. Human herpesvirus 8: biology and role in the pathogenesis of Kaposi's sarcoma and other AIDS-related malignancies. Curr. Infect. Dis. Rep. 5:169-175.
Wang, J., Y. F. Chang, J. I. Hamilton, and M. F. Wilkinson. 2002. Nonsense-associated altered splicing: a frame-dependent response distinct from nonsense-mediated decay. Mol. Cell 10:951-957.
Wang, S. E., F. Y. Wu, Y. Yu, and G. S. Hayward. 2003. CCAAT/enhancer-binding protein- is induced during the early stages of Kaposi's sarcoma-associated herpesvirus (KSHV) lytic cycle reactivation and together with the KSHV replication and transcription activator (RTA) cooperatively stimulates the viral RTA, MTA, and PAN promoters. J. Virol. 77:9590-9612.
Reference deleted.
Wu, F. Y., J. H. Ahn, D. J. Alcendor, W. J. Jang, J. Xiao, S. D. Hayward, and G. S. Hayward. 2001. Origin-independent assembly of Kaposi's sarcoma-associated herpesvirus DNA replication compartments in transient cotransfection assays and association with the ORF-K8 protein and cellular PML. J. Virol. 75:1487-1506.
Wu, F. Y., Q. Q. Tang, H. Chen, C. ApRhys, C. Farrell, J. Chen, M. Fujimuro, M. D. Lane, and G. S. Hayward. 2002. Lytic replication-associated protein (RAP) encoded by Kaposi sarcoma-associated herpesvirus causes p21CIP-1-mediated G1 cell cycle arrest through CCAAT/enhancer-binding protein-. Proc. Natl. Acad. Sci. USA 99:10683-10688.
Wu, F. Y., S. E. Wang, Q. Q. Tang, M. Fujimuro, C. J. Chiou, Q. Zheng, H. Chen, S. D. Hayward, M. D. Lane, and G. S. Hayward. 2003. Cell cycle arrest by Kaposi's sarcoma-associated herpesvirus replication-associated protein is mediated at both the transcriptional and posttranslational levels by binding to CCAAT/enhancer-binding protein alpha and p21(CIP-1). J. Virol. 77:8893-8914.
Zhang, J., and L. E. Maquat. 1997. Evidence that translation reinitiation abrogates nonsense-mediated mRNA decay in mammalian cells. EMBO J. 16:826-833.
Zhang, M. Q. 1998. Statistical features of human exons and their flanking regions. Hum. Mol. Genet. 7:919-932.
Zheng, Z. M. 2003. Split genes and their expression in Kaposi's sarcoma-associated herpesvirus. Rev. Med. Virol. 13:173-184.
Zheng, Z. M. 2004. Regulation of alternative RNA splicing by exon definition and exon sequences in viral and Mammalian gene expression. J. Biomed. Sci. 11:278-294.
Zheng, Z. M., and C. C. Baker. 2000. Parameters that affect in vitro splicing of bovine papillomavirus type 1 late pre-mRNAs. J. Virol. Methods 85: 203-214.
Zheng, Z. M., P. J. He, and C. C. Baker. 1997. Structural, functional, and protein binding analyses of bovine papillomavirus type 1 exonic splicing enhancers. J. Virol. 71:9096-9107.
Zheng, Z. M., E. S. Reid, and C. C. Baker. 2000. Utilization of the bovine papillomavirus type 1 late-stage-specific nucleotide 3605 3' splice site is modulated by a novel exonic bipartite regulator but not by an intronic purine-rich element. J. Virol. 74:10612-10622.
Zheng, Z. M., M. Tao, K. Yamanegi, S. Bodaghi, and W. Xiao. 2004. Splicing of a cap-proximal human papillomavirus 16 E6E7 intron promotes E7 expression, but can be restrained by distance of the intron from its RNA 5' cap. J. Mol. Biol. 337:1091-1108.
Zhong, W., H. Wang, B. Herndier, and D. Ganem. 1996. Restricted expression of Kaposi sarcoma-associated herpesvirus (human herpesvirus 8) genes in Kaposi sarcoma. Proc. Natl. Acad. Sci. USA 93:6641-6646.
Zhu, F. X., T. Cusano, and Y. Yuan. 1999. Identification of the immediate-early transcripts of Kaposi's sarcoma-associated herpesvirus. J. Virol. 73:5556-5567.(Koji Yamanegi, Shuang Tan)
ABSTRACT
Kaposi's sarcoma-associated herpesvirus (KSHV) is a lymphotropic DNA tumor virus that induces Kaposi's sarcoma and AIDS-related primary effusion lymphoma. KSHV open reading frame 50 and K8 genes in early viral lytic infection express, respectively, a tricistronic and a bicistronic pre-mRNA, which undergo alternative splicing to create two major spliced mRNA isoforms, and ?, by inclusion (?) or exclusion () of an intron at nucleotides 75563 to 75645. This intron contains some suboptimal features, which cause the intron 5' splice site (ss) to interact weakly with U1 snRNA and the 3' ss to bind a U2 auxiliary factor, U2AF, with low affinity. Optimization of this intron in K8 (K8 intron 2) promoted the interaction of the 5' ss with U1 and the 3' ss with U2AF, resulting in a substantial increase in intron splicing. Splicing of K8 intron 2 has also been shown to be stimulated by the splicing of a downstream intron. This was confirmed by the insertion of a human ?-globin intron into the K8? exon 3-exon 4 splice junction, which promoted splicing of K8? intron 2 and conversion of the K8? mRNA to the K8 mRNA that encodes a K-bZIP protein. Intron 2 contains a premature termination codon, yet the K8? mRNA is insensitive to nonsense-mediated mRNA decay, suggesting that the truncated K8? protein may have a biological function. Indeed, although the truncated K8? protein is missing only a C-terminal leucine zipper domain from the K-bZIP, its expression antagonizes the ability of the K-bZIP to induce p53 and p21 and blocks K-bZIP-CDK2 interaction through interfering K8 mRNA production.
INTRODUCTION
Kaposi's sarcoma-associated herpesvirus (KSHV), or human herpesvirus 8, is a human gammaherpesvirus that closely resembles human Epstein-Barr virus (7). As with other gammaherpesviruses, infection with KSHV is known to be strongly associated with lymphoproliferative disorders and malignancies. It is present in all forms of Kaposi's sarcoma (16, 47) and in two other diseases, body cavity-based B-cell lymphoma (also known as primary effusion lymphoma [PEL]) and multicentric Castleman's disease (40, 41).
Like infection with other herpesviruses, KSHV infection displays two viral life cycles. Latent KSHV infection in Kaposi's sarcoma tissues and B-cell lines features the highly restricted expression of only five viral genes (9, 62). The lytic KSHV life cycle is characterized by the production of progeny virus from infected cells and can be induced by chemicals such as tetradecanoyl phorbol acetate (TPA) and n-butyrate in PEL-derived B cells with latent KSHV infection (26, 27, 34). Chemical induction in PEL-derived B cells initiates the expression of a KSHV transactivator, open reading frame 50 (ORF50), which is absolutely required for the lytic switch from KSHV latency (24, 43).
KSHV ORF50 is an immediate-early gene transcribed as a polycistronic RNA with five exons and four introns. It encodes a pivotal transactivator that mediates transcription of all early genes (24, 43). ORF50 exon 1 and exon 2 contain the coding region of ORF50, whereas exons 3, 4, and 5 of ORF50 are noncoding exons that overlap with the K8 and K8.1 coding regions (43, 46). This overlap enables ORF50 to use a common poly(A) site at nucleotide (nt) 76714, downstream of the K8.1 coding region. This unique gene structure means that KSHV ORF50, K8, and K8.1 are positioned side by side in the virus genome, and the transcripts of the three genes overlap each other (56). However, each gene initiates its transcription at different stages of the viral life cycle (23, 36, 45) and undergoes extensive RNA splicing during gene expression (56). KSHV K8 is an early gene transcribed as a bicistronic RNA with four exons and three introns that expresses a K-bZIP protein (K8) involved in viral DNA replication (1, 22, 51) and transcription (19, 21, 28, 49) as well as induction of cell cycle arrest (20, 52, 53). KSHV K8.1 is a late gene transcribed as a monocistronic RNA with two exons and one intron that encodes an envelope glycoprotein (6, 32, 45).
We have profiled the splicing patterns of KSHV ORF50, K8, and K8.1 through multiple approaches and have reported that splicing of these pre-mRNAs is extremely complex and often results in inclusion of an intron (intron 3 of ORF50, intron 2 of K8) at nt 75563 to 75645 (Fig. 1A), presumably due to suboptimal features of the intron splice sites (46, 56). mRNAs with this intron inclusion are referred to as the ? isoform; those without the intron are referred to as the isoform. We have also shown by using KSHV+ cell lines that exon 3 of the K8 pre-mRNA (exon 4 of ORF50) includes most of a previously defined intron 3 and has three 5' splice sites (ss), at nt 75838, 76155, and 76338 (46). Selection of the nt 75838 5' ss dictates the production of K8 mRNA and overwhelms the RNA processing. The alternative selection of the other two 5' ss is feasible and leads to the production of two additional bicistronic mRNAs: K8/K8.1 and -?. However, only a little K8 and no detectable K8.1 protein are translated from the novel bicistronic K8/K8.1 mRNAs in 293 cells, suggesting that production of the K8/K8.1 mRNAs may be an essential way to limit K8 mRNAs, especially K8, to a threshold at the RNA processing level (46). Altogether, we have concluded that a substantial proportion of pre-mRNAs transcribed from the ORF50, K8, and K8.1 genes undergo alternative RNA splicing, leading to the production of at least 19 species of spliced and unspliced transcripts (56). In this regard, the inclusion or exclusion of the intron from nt 75563 to 75645 plays an important role in the alternative RNA splicing of both ORF50 and K8 pre-mRNA. However, what controls this intron splicing and how it affects the biology of ORF50 and K8 remains to be understood.
In this report, we have characterized K8 intron 2 (intron 3 of ORF50) and demonstrated that the suboptimal features of intron 2 determine its inefficient interaction with the cellular splicing machinery. Further experiments showed that the inclusion of this intron in K8? mRNA results in the production of a truncated protein that plays a dominant negative role in fully spliced K8 mRNA expression, K8-mediated induction of p21 and p53, and K8-CDK2 interaction.
MATERIALS AND METHODS
Cells. The KSHV-positive B-cell line JSC-1 was a generous gift from Richard Ambinder of the Johns Hopkins University. BCBL-1 cells were obtained from Michael McGrath and Don Ganem of the University of California at San Francisco through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, National Institutes of Health. Both PEL-derived B-cell lines were maintained in RPMI 1640 medium (Invitrogen, Carlsbad, CA). Human 293 cells and HeLa cells were obtained from the American Type Culture Collection (Manassas, VA) and were maintained in Dulbecco's modified Eagle's medium (Invitrogen). Both culture media were supplemented with 10% fetal bovine serum (HyClone, Logan, UT) and with 2 mM L-glutamine, 100 U penicillin/ml, and 100 μg streptomycin/ml. The cells were incubated in the complete culture medium at 37°C in the presence of 5% CO2.
KSHV activation by chemical induction. To activate KSHV lytic gene expression in JSC-1 cells, 1 x 107 cells were treated with sodium butyrate (Aldrich, Milwaukee, WI) at a final concentration of 3 mM for 24 h. BCBL-1 cells (1 x 107) were induced with 20 ng/ml TPA (Sigma, St. Louis, MO) for 48 h. Total cell RNA and total protein samples were then prepared for further analysis of K8 expression. To analyze nonsense-mediated RNA decay (NMD) of KSHV K8? RNA, a protein synthesis inhibitor, cycloheximide (CHX; Sigma, St. Louis, MO), was added at a concentration of 50 μg/ml to the JSC-1 cells after 4 h of butyrate induction (63). The cells were further cultured for 24 h in the presence of both CHX and butyrate, after which total cell RNA was prepared for RNase protection assays (RPA).
Cell transfection. For transfection of 293 cells and HeLa cells, plasmid DNA was transfected into 1 x 107 cells using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Total cell RNA or protein was prepared 24 or 48 h after transfection, as indicated in the figures. The RNA was used for reverse transcription (RT)-PCR and RPA, and the protein was used for Western blot analysis. To study the NMD of the KSHV K8?, 293 cells were transfected with plasmid pST1, a K8 gene expression vector (46); 8 h later, the cells were cultured in the presence of CHX (28 μg/ml) (5) for 24 h prior to total cell RNA preparation.
Electroporation was used for transfection of KSHV+ BCBL-1 cells with plasmid pORF50, pST1, pST3, and pcDNA 3.0 vector. Briefly, BCBL-1 cells were washed twice with cold phosphate-buffered saline (PBS) (calcium and magnesium free) and resuspended at 1 x 107 cells/ml in RPMI 1640 without any supplements (GIBCO). Approximate 400 μl of the cell suspension was mixed with 20 μg of each plasmid and transfected in a prechilled 0.4-cm cuvette (Bio-Rad, Hercules, CA) by electroporation with an Electro Cell Manipulator BTX ECM630 (Genetronics, Inc., San Diego, CA) at the following conditions: voltage, 350 V; capacitance, 960 μF; and resistance, 50 . Under such electroporation conditions, over 50% of the cells remained viable, and half of those viable cells could be transfected with a green fluorescent protein fusion plasmid as a transfection control. The cell suspensions after electroporation were then incubated at 37°C in 20 ml of complete RPMI 1640 medium for 72 h using regular cell culture conditions prior to total cell RNA extraction and protein sample preparation.
Plasmid construction. The plasmids pKY1 and pKY2 are KSHV K8 expression plasmids based on the wild-type K8 gene expression vector pST1 (46), but with an optimized intron 2. pKY1 has an optimized (mutated) nt 75645 3' ss, and pKY2 contains an optimized nt 75563 5' ss in K8 intron 2. These plasmids were constructed using an overlapping PCR technique (60). Basically, two primer sets were used to introduce mutations into each splice site. To amplify PCR fragments for the construction of plasmid pKY1, the primer set Pr74850 (oST41, 5'-CACC/ATGCCCAGAATGAAGGAC-3') and Pr75644 (oZMZ320, 5'-CTGAAAAGAAAGAAAGGTGGCGCCATTGTTCCCATTTGAG-3') and the primer set Pr75625 (oZMZ321, 5'-CCACCTTTCTTTCTTTTCAGGCATTAGAAGAAAAGGATGC-3') and Pr76687 (oST42, 5'-TCC/AGGGTTTCTTAC GCCG-3') were used. The amplified PCR products from each set of primers were mixed, heated at 95°C for 5 min, annealed at room temperature for 30 min, and reamplified using the primer Pr74850 in combination with the primer Pr76687. Following gel purification, the amplified DNA fragment was inserted into an expression vector, pcDNA3.1/V5-His-TOPO (Invitrogen). The same strategy was used to construct plasmid pKY2 with two different primer sets, Pr74850 in combination with Pr75589 (oZMZ 332, 5'-GATAGACACCTGTTT CCCGTACTTACCTGCTGCAGCTGTCT-3') and Pr75549 (oZMZ 331, 5'-AGACAGCTGCAGCAGGTAAGTACGGGAAACAGGTGTCTATC-3') in combination with Pr76687. The plasmid pKY3 is optimized at both the 75563 5' ss and the 75645 3' ss and was derived from pKY1 using the same primer sets as for the pKY2 construction. All optimization mutations described for pKY1, pKY2, and pKY3 were verified by sequencing.
The plasmid pKY4 was derived from plasmid pST3 (46), a full-length KSHV K8? cDNA expression vector, and has a human ?-globin intron inserted at the K8? exon 3-exon 4 junction. To introduce human ?-globin into the exon 3-exon 4 junction of K8?, a two-step overlapping PCR approach was taken. Briefly, the first step of the overlapping PCR was carried out with two separate PCR products: the K8 exon 4 product and a human ?-globin intron product obtained from two separate PCRs. K8 exon 4 was amplified from plasmid pST3 using a sense primer, Pr76433 (oZMZ 364, 5'-GATCATATTCATCTGGGGAACC-3'), in combination with an antisense primer, Pr76687. The human ?-globin intron was amplified from plasmid pSP64-H?6 (Promega, Madison, WI) using a K8 exon 3-?-globin chimeric sense primer, Pr146 (oZMZ 365, 5'-ACCATGCCAGACTTTGTGTG/GTTGGTATCAAGGTTACAAG-3'), in combination with a K8 exon 4-?-globin chimeric antisense primer, Pr275 (oZMZ 366, 5'-GGTTCCCCAGATGAATATGATC/CTAAGGGTGGGAAAATAGACC-3'). The two separate PCR products were then mixed for overlapping PCR using the sense primer Pr146 combined with the antisense primer Pr76687; the resulting PCR product had part of exon 3, a ?-globin intron, and exon 4. The first-step overlapping PCR was then mixed with another PCR product obtained from plasmid pST3 using the sense primer Pr74850 in combination with an antisense primer, Pr75838 (oZMZ 367, 5'-CACACAAAGTCTGGCATGGTTC-3'), which contained all K8? sequences except exon 4. The second step of the overlapping PCR was performed on the mixture using the primer pair Pr74850 and Pr76687, followed by gel purification. The amplified product was then cloned into the expression vector pcDNA3.1/V5-His-TOPO. The plasmid inserts were verified by sequencing.
RT-PCR. Total cell RNA was extracted from cells by using TRIzol reagent (Invitrogen) according to the manufacturer's instructions. Following DNase I digestion, reverse transcription of 1 μg of total RNA was performed at 42°C for 1 h using random hexamers and then amplified by PCR with AmpliTaq DNA polymerase (Roche, Indianapolis, IN) using different primer sets as described in each figure. The following list includes the primer sequences, which are named by the locations of their 5' ends. Pr74259 (oZMZ317, 5'-GACCATCGG AGGCACAAC-3'), Pr74565 (oST43, 5'-ACGAGGTACAGGAGTCCG-3'), Pr75182-T7 (oST1, 5'-TAATACGACTCACTATAGGG/ACCACCAAGA GGACCACACATTTC-3'), and Pr75354 (oST23, 5'-GAACTGACAAGGGGT GACGC-3') are forward primers; Pr75504 (oST4, 5'-GCGTGGGAGAATGTG ACT-3'), Pr75813 (oST32, 5'-CGGACGACGGCACAACCA-3'), Pr75838 (oST3, 5'-GTACTCACCCC/CACACAAAGTCTGGCATGGTTCTCCC-3'), and Pr 76576 (oST31, 5'-TGTAGTGCGCGTCTCTTCCTC-3') are reverse primers. The final PCR products were detected using agarose gel electrophoresis and ethidium bromide staining and analyzed with a ChemiDoc system (Bio-Rad, Hercules, CA). All RT-PCR products were gel purified and confirmed by sequencing through the splicing junction.
Human GAPDH RNA in each RNA sample was amplified by RT-PCR using a sense primer, Pr6732 (oZMZ269, 5'-GTCATCAATGGAAATCCCATCACC-3'), in combination with an antisense primer, Pr7207 (oZMZ257, 5'-TAATACGACTCACTATAGGGA/CCGTTCAGCTCAGGGATG-3').
In vitro RNA splicing assay. DNA templates were obtained separately from plasmids pST1, pKY1, pKY2, and pKY3 by PCR using a sense chimeric SP6/KSHV primer, Pr75484 (oZMZ335, 5'-ATTTAGGTGACACTATAG/ATC AGTCACATTCTCCCACGC-3'), combined with the antisense primer Pr75813. In vitro transcription using SP6 RNA polymerase was performed on each template DNA in the presence of a cap analog (m7GpppG) and [-32P] GTP. An in vitro RNA splicing assay was then conducted on each transcript using a HeLa cell nuclear extract in an optimized splicing condition. The spliced products were resolved in an 8% denaturing polyacrylamide gel electrophoresis (PAGE) gel and analyzed as described previously (58).
RPA. Antisense RNA probes were prepared as described above by in vitro transcription from individual DNA templates obtained from plasmids pST1, pST3, pKY1, pKY2, pKY3, and pKY4 by PCR using a sense primer, Pr75605 (oST36, 5'-CTCAAATGGGAACAATGGCG-3'), combined with an antisense chimeric T7/KSHV primer, Pr75935 (oST33, 5'-TAATACGACTCACTATAGGG/CGAATCTGTGTGGAACTCA-3'). A human cyclophilin probe was transcribed from a commercially available DNA template (Ambion, Austin, TX) and used as an internal control. For RPA, 100 μg of total cell RNA was hybridized with 4 ng of an antisense RNA probe in hybridization buffer at 42°C overnight, followed by RNase (RNase A and T1) digestion as instructed by the manufacturer of the RPA III kit (Ambion). Protected RNA fragments were separated in an 8% denaturing PAGE gel. A gel image was captured using a Molecular Dynamic PhosphorImager Storm 860 and analyzed with ImageQuant software.
RNase H cleavage protection assay and U1 and U2 depletion. Individual single-strand DNA oligonucleotides complementary to the exon-intron junctions of K8 intron 2 were used for oligonucleotide-directed RNase H digestion assays. The following antisense DNA oligonucleotides were used: oligonucleotide oKY2 (5'-TCTATACCTGCTG-3'), complementary to the exon 2-intron 2 junction of wild-type (wt) K8 RNA transcribed from plasmid pST1, and oKY4 (5'-TAC TTACCTGCTG-3'), complementary to the same region of mutant (mt) K8 RNA transcribed from plasmid pKY3; and oKY3 (5'-AAGACAGCAAGGT-3'), complementary to the intron 2-exon 3 junction of wt K8 RNA transcribed from plasmid pST1, and oKY5 (5'-AAGAAAGAAAGGT-3'), complementary to the same region of mt K8 RNA transcribed from plasmid pKY3. A RNase H digestion reaction was carried out in 20 μl of reaction mixture containing 40% HeLa nuclear extract, 2.5 mM MgCl2, 1 U of RNasin, and 4 ng of 32P-labeled pre-mRNA as described previously (61). Each reaction mixture was incubated for 10 min at 30°C in the absence of a DNA oligonucleotide and then incubated for another 10 min in the presence of an oligonucleotide (2 μM) before digestion for 10 min with 2 or 4 U of RNase H (Promega). The digestion was terminated by adding 3 volumes of proteinase K solution (5 mg/ml proteinase K, 50 mM EDTA, 0.5% sodium dodecyl sulfate [SDS]) (10), incubated at 75°C for 30 min, and extracted with phenol-chloroform. The RNA cleavage products were then resolved in an 8% polyacrylamide gel containing 8 M urea. The RNase H digestion efficiency for each pre-mRNA was calculated as the total amount of digestion products divided by the sum of the total digestion products plus the remaining pre-mRNA.
The oligonucleotide-directed RNase H digestion described above was also used to deplete small nuclear RNA U1 and U2 from the HeLa cell nuclear extract. Briefly, a human U1 (oKY6, 5'-TGCCAGGTAAGTA-3') or U2 (oKY7, 5'-AGGCCGAGAAGCGA-3') oligonucleotide was incubated at 30°C for 10 min with the RNase H reaction mixture described above in the absence of a 32P-labeled pre-mRNA. After the addition of RNase H (2 U) to the reaction, the incubation was extended for another 10 min. A 32P-labeled pre-mRNA and a specific K8 oligonucleotide (oKY2 or oKY3 for pST1; oKY4 or oKY5 for pKY3) were added, and the mixture was incubated for 10 min at 30°C before proteinase K digestion.
Western blot analysis. 293 cells or HeLa cells with or without transfection and JSC-1 or BCBL-1 cells with or without butyrate or TPA induction were washed twice with PBS and lysed in 2x SDS protein gel loading solution (Quality Biological, Inc., Gaithersburg, MD) containing 10% ?-mercaptoethanol or lysed in radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS) (Boston BioProducts, Ashland, MA). If RIPA buffer was used, the cell lysates were first incubated on ice for 30 min and then sonicated several times (5 s each) before centrifugation for 10 min at 10,000 rpm. The supernatant was mixed with 2x SDS sample buffer containing 10% ?-mercaptoethanol. Protein samples were separated on a 4% to 12% Bis-Tris gel (Invitrogen) and transferred onto a nitrocellulose membrane. The membrane was blocked for 2 h at room temperature in a blocking buffer containing 5% skim milk in Tris-buffered saline (TBS) (10 mM Tris, 150 mM NaCl [pH 7.4]) and then incubated for 90 min with a 1:2,000 dilution of rabbit polyclonal anti-K-bZIP antibody (a gift from Don Ganem at the University of California, San Francisco), a 1:500 dilution of mouse monoclonal anti-human p21 antibody (BD Pharmingen, San Diego, CA), or a 1:100 dilution of mouse monoclonal anti-human p53 antibody in TTBS buffer (10 mM Tris, 150 mM NaCl, 1% Tween 20) containing 5% skim milk, after which the membrane was washed with TTBS three times (10 min each time). The membrane was then incubated for 60 min at room temperature with a 1:20,000 dilution of horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin G antibody for K-bZIP detection or with a 1:10,000 dilution of horseradish peroxidase-conjugated goat anti-mouse immunoglobulin G antibody for p21 and p53 detection and visualized with a SuperSignal West Pico chemiluminescence detection system (Pierce, Rockford, IL). The same membrane was stripped in a stripping buffer (Pierce) and reprobed using a polyclonal goat anti-CDK2 antibody (BD PharMingen) for immunoprecipitated CDK2 detection or using a monoclonal anti-?-tubulin antibody (BD PharMingen) for sample loading.
p53 reporter assays. HeLa cells (5 x 105) were transfected with 2 μg of pST1 (mainly expressing K8) or pST3 (K8?) in the presence of vector pcDNA3 (2 μg) or cotransfected with both pST1 (2 μg) and pST3 (2 μg) using Lipofectamine 2000 (Invitrogen). After 48 h of incubation, the transfected HeLa cells were transfected again with 2 μg of a p53 reporter plasmid (PG13-Luc, a gift from John Brady of the National Cancer Institute) (31) and incubated for additional 24 h at 37°C and 5% CO2. The cells were then collected and washed twice with calcium- and magnesium-free PBS. Part of the cell pellets were resuspended in 0.2 ml of Promega reporter lysis buffer at room temperature for 15 min. After centrifugation, 100 μl of the supernatant was collected for luciferase assay (Promega). Relative promoter activity was calculated by the light unit readings obtained from each plasmid-transfected cell lysate divided by the light unit readings obtained from pcDNA3-transfected cell lysate. The other part of the cell pellet was then lysed in RIPA buffer on ice for 30 min with sonication (several times, for 5 s each time). After centrifugation for 10 min at 10,000 rpm, the supernatant was mixed with 2x SDS sample buffer containing 10% ?-mercaptoethanol for the p53 Western blot assay.
CDK2 immunoprecipitation. One hundred microliters of a total of 700 μl of cell lysate in RIPA buffer obtained from HeLa cells (1 x 107 cells) transfected with pST1 or pST3 or pST1 and pST3 in combination was precleaned with 20 μl of a 50% slurry of protein A-Sepharose beads (Upstate, Waltham, MA) for 3 h at 4°C and washed twice with 1x immunoprecipitation (IP) buffer (59). After centrifugation at 13,000 rpm for 3 min, 20 μl of the supernatant was collected and incubated with 10 μl of anti-CDK2-coated beads (BD PharMingen) in 500 μl of 1x IP buffer overnight at 4°C. The beads were then gently washed four times with 1x IP buffer, and the proteins bound on the beads were eluted with 2x SDS sample buffer containing 10% ?-mercaptoethanol. Protein samples were separated on a 4 to 12% Bis-Tris gel (Invitrogen) and transferred onto a nitrocellulose membrane for Western blotting as described above.
RESULTS
Alternative splicing of ORF50 and K8 pre-mRNAs produces and ?, but not , mRNAs during KSHV lytic gene expression. In our previous study, we demonstrated that alternative splicing of K8 pre-mRNA produces two major forms of the spliced products ( and ?). The form differs from the ? by the exclusion of intron 2. That is, the K8 mRNA is a fully spliced mRNA, whereas the K8? is a partially spliced mRNA that retains intron 2. Since the 3' half of the ORF50 transcript overlaps the K8 transcript and has the same sequence composition as the entire K8 transcript (Fig. 1A), the part of the and ? isoforms of the K8 transcripts detected from KSHV+ cells must be spliced products of ORF50. However, exon 2 of ORF50 pre-mRNA has a size of 2,751 nt, which is too large to be recognized as an internal exon by the cellular splicing machinery (42). Whether such an exon structure of ORF50 could permit cross talking between intron 1 and intron 2 over exon 2 and therefore restrain ORF50 pre-mRNA from splicing has been questioned. In this study, RT-PCR was performed with total cell RNA purified from butyrate-stimulated JSC-1 cells with a primer from ORF50 exon 2 in combination with the primer Pr75504 (positioned in ORF50 exon 3 or K8 exon 2). The RT-PCR analysis demonstrated that splicing of ORF50 pre-mRNAs was efficient and did not retain intron 2 (K8 intron 1) (Fig. 1B), suggesting that a unknown function residing in ORF50 exon 2 could promote recognition of an oversized internal exon by cellular splicing machinery.
Previous reports from other laboratories have suggested that there is a isoform of the ORF50 and K8 message due to alternative RNA splicing in lytic KSHV gene expression (23, 63). In these reports, the RNA is depicted as retaining two introns, one from nt 75323 to 75471 and another from nt 75563 to 75645, but lacking the intron from nt 75838 to 76433 due to RNA splicing. A computer analysis of all three of these introns in our laboratory suggested that both the intron from nt 75323 to 75471 and the intron from nt 75838 to 76433 contain consensus splice sites and should be efficiently recognized by the cellular splicing machinery; it would be unusual for the cellular splicing machinery to miss recognizing an intron with consensus splice site sequences. In fact, we found that splicing of the intron from nt 75323 to 75471 is efficient. To further test this observation, we carried out an RT-PCR strategy to detect a RNA in butyrate-stimulated JSC-1 cells and TPA-stimulated BCBL-1 cells using an intron primer, Pr75354, from the intron region of nt 75323 to 75471, in combination with an antisense exon primer, Pr75813 or Pr76576 (Fig. 1C). The predicted RNA would be 636 nt in size when primer Pr76576 was paired with primer Pr75354 for the amplification. However, total cell RNA prepared from butyrate-stimulated JSC-1 cells did not have a band of this size (Fig. 1C, lane 6), despite the existence of RNAs containing the intron from nt 75323 to 75471 and the intron from nt 75563 to 75645 (Fig. 1C, lanes 3, 5, and 8) as amplified with a primer pair of Pr75813 and Pr75354. The same was true for BCBL-1 cells (data not shown). The RNA containing these two introns but lacking a 3' terminal exon sequences may more likely be a partially transcribed, unspliced pre-mRNA of ORF50 and/or K8.
KSHV K8? RNA is not subject to NMD and encodes a K8? protein during lytic infection. K8 intron 2 (positioned from nt 75563 to 75645) contains a premature termination codon (PTC), and its retention in the K8? mRNA introduces a PTC into the mRNA. K8? should therefore be subject to NMD. As currently understood, NMD serves as a posttranscriptional control system by which cells recognize and degrade mRNAs carrying a PTC positioned >50 nt upstream of a splice junction to prevent the possible toxic effects of truncated peptides (25). It has been shown that translation or a translation-like process is a prerequisite for NMD, and inhibition of translation by CHX results in the abolishment of NMD and, subsequently, the accumulation of the PTC+ mRNA (5, 33). We therefore compared the quality and quantity of the K8? RNA by RPA in 293 cells transfected with a K8 expression vector and in butyrate-activated KSHV+ JSC-1 cells with or without CHX treatment (Fig. 2). The expected RPA products are shown in Fig. 2A. We postulated that cells treated with CHX would have more K8? mRNA due to a minimized NMD response. However, K8 and K8? were present in very similar amounts in the presence or absence of CHX in 293 cells transfected with the K8 expression vector (pST1, lanes 6 to 9), suggesting no accumulation of K8? and therefore no NMD-mediated K8? mRNA degradation. In the butyrate-activated JSC-1 cells, CHX treatment inhibited both K8 and K8? gene expression (Fig. 2B, lanes 3 to 5). This was not surprising, since several studies have reported that expression of K8 in KSHV+ BCBL-1, BC-1, and JSC-1 cells is sensitive to CHX treatment (44, 63). Thus, it was impossible to draw any conclusion on NMD-mediated K8? degradation at the RNA level in JSC-1 cells. However, we were able to detect K8? protein from both JSC-1 and BCBL-1 cells after butyrate and TPA stimulation (Fig. 2C), respectively, suggesting that K8? RNA resists NMD-mediated RNA degradation in those cells and is capable of being translated.
Optimization of K8 intron 2 splice sites promotes splicing of K8 pre-mRNA. As proposed in our previous report (46), the inclusion of intron 2 in K8? mRNA could result from its suboptimal nature, since this intron contains a nonconsensus 5' ss and a weak 3' ss. The sequence of the 5' ss of intron 2 misses pairing a few nucleotides from the U1 5' end. Most importantly, the 5' ss misses two nucleotides, at intron positions 4 and 5, for base pairing to the six core nucleotides of the U1 5' end (Fig. 3A, top). A missing base pairing of a U1 snRNA 5' core nucleotide to the RNA 5' ss at intron position 5 has been reported to inhibit U1 snRNA binding and splicing (15, 61). The sequences of the intron 3' ss contain a few purine Gs interspersed in its putative polypyrimidine tract (PPT), which lies between the branch point sequence (BPS) and the AG dinucleotide (Fig. 3A, bottom). Recognition of the BPS by U2 snRNA initiates the first step of RNA splicing. The PPT is a run of 15 to 40 pyrimidines (usually U's) and contains binding sites for U2AF, a heterodimer of U2AF65 and U2AF35. Binding of U2AF to the PPT strengthens U2 recognition of the BPS (57). It has been reported that any purines interspersed in the PPT will weaken the binding of U2AF and affect U2 interaction with the BPS. Moreover, the nature of the nucleotide preceding the AG dinucleotide has a profound effect on splicing (39), and CAG is twice as common as UAG in mammalian 3' ss (55). Together, these features of K8 intron 2 would be expected to decrease the inherent strength of its recognition by U1 and U2 snRNAs.
To test our hypothesis, we constructed three K8 expression vectors. pKY1 has an optimized 3' ss (Fig. 3A, bottom, mt 3' ss), pKY2 has an optimized 5' ss (Fig. 3A, middle), and pKY3 has optimized 3' ss and 5' ss. Following transient transfection of 293 cells with these expression vectors, we used RT-PCR and RPA to examine the expression of K8 and K8? in parallel with the expression of a wild-type K8 expression vector, pST1. As shown in Fig. 3B and C, the optimization of K8 intron 2 increased the splicing efficiency of the K8 pre-mRNA and resulted in at least twofold reduction of K8? production in both assays. The data suggest that strengthening either the 3' ss or 5' ss in K8 intron 2 improved the recognition of the intron by the cellular splicing machinery and led to the splicing of intron 2.
In vitro splicing of pre-mRNAs transcribed from the optimized and wt plasmids showed similar results. The pre-mRNA transcribed from wt K8 had a poor splicing efficiency (only 2%) (Fig. 3D, pre-mRNA 1), while the optimized pre-mRNAs exhibited an increased (>9-fold) splicing efficiency (Fig. 3D, pre-mRNAs 2 to 4). Notably, the pre-mRNA with both the 5' ss and 3' ss optimized showed the highest splicing efficiency (13-fold greater than the wild type) (Fig. 3D).
The escape of K8 intron 2 from splicing is due to its inefficient recognition by U1 snRNA and U2AF. To understand further how intron 2 escapes RNA splicing, the RNAs transcribed from pST1 (wt K8) and pKY3 (K8 with both of the intron 2 splice sites optimized) were also used for a DNA oligonucleotide-directed RNase H digestion assay. As mentioned above, we would expect that the suboptimal features of intron 2 do not favor efficient interactions between U1 and the 5' ss or between U2AF and the 3' ss. If this prediction is right, the optimization of both splice sites should lead to increased U1-5' ss interactions and U2AF-3' ss interactions and, therefore, to increased splicing efficiency of intron 2.
RNase H digests only RNA-DNA hybrids in which a DNA oligonucleotide hybridizes to a specific RNA sequence through base pairing (8, 14). An RNA-RNA hybrid in the same region or a region of RNA with bound protein will prevent a DNA oligonucleotide from base pairing and will consequently prevent RNase H digestion. Based on this theory, we synthesized antisense DNA oligonucleotides (12 or 13 nt in length) specific to the 5' ss or the 3' ss of wt or mt K8 intron 2, as diagrammed in Fig. 4A. Using this approach, we were able to quantify a U1- 5' ss interaction and a U2AF-3' ss interaction in the presence of HeLa cell nuclear extracts containing enriched U snRNAs and splicing factors. As shown in Fig. 4B, the wt 5' ss and 3' ss of the K8 intron 2 have a weak binding affinity for U1 and U2AF, allowing more DNA oligonucleotides to bind to their corresponding splice sites; as a result, approximately 90% and 61% of the RNA was digested by RNase H directed by a 5' ss-specific DNA oligonucleotide and a 3' ss-specific DNA oligonucleotide, respectively. In contrast, RNase H digestion was much less effective when both the intron 5' ss and 3' ss were optimized: only about 38% of the RNA was subjected to the oligonucleotide-directed digestion of the RNA 5' ss, and approximately 3% of the RNA was subjected to the oligonucleotide-directed digestion of the RNA 3' ss.
To further confirm that the reduction in digestion of the RNA intron 2 splice sites by RNase H was specifically due to the improvement in the U1-5' ss and U2AF-3' ss interactions, we depleted either U1 or U2 snRNAs from HeLa nuclear extracts using RNase H digestion directed by an antisense U1 or U2 oligonucleotide. The U1- or U2-depleted nuclear extracts were then used in DNA oligonucleotide-directed digestion of wt or mt K8 intron 2. Since U2AF is abundant in HeLa nuclear extracts and complete depletion of U2AF would be difficult by antibody affinity or poly(U) approaches, we depleted U2 instead of U2AF from HeLa nuclear extracts on the assumption, based on the U2 recruitment model, that the RNA-U2AF interaction is dynamic and less U2AF would be fixed on the RNA in the absence of U2 snRNA (11, 12). Depletion of U1 from HeLa nuclear extracts did increase the oligonucleotide-directed RNase H digestion of both the wt (Fig. 4C, compare lanes 3 and 4) and the mt (Fig. 4D, compare lanes 3 and 4) K8 intron 2, indicating a specific U1-5' ss interaction. In addition, an increase in oligonucleotide-directed RNase digestion of the wt K8 intron 2 was observed when U2-depleted HeLa nuclear extracts were used (Fig. 4C, compare lanes 5 and 6). However, we were unable to see a similar result from the oligonucleotide-directed RNase H digestion of the 3' ss of mt K8 intron 2 (Fig. 4D, compare lanes 7 and 8), suggesting that optimization of the 3' ss might also create an additional binding site (UCUU) for another protein, most likely a PPT-binding protein (PTB) (29), which is unrelated to U2 recruitment.
Insertion of a human ?-globin intron into the K8? exon 3-exon 4 splice junction stimulates splicing of K8 intron 2 and converts K8? to K8. We have reported that K8 exon 3 contains three alternative 5' ss, at nt 75838, 76155, and 76338. Selection of the nt 75838 5' ss is subject to exon definition and predominates over the other two 5' ss during RNA splicing, even though all three 5' ss are optimal splice sites with consensus U1 binding sequences. In vivo experiments also suggest that utilization of the nt 75838 5' ss stimulates splicing of an upstream nt 75645 3' ss, crossing over exon 3 (46). Disruption of the nt 75838 5' ss promotes either the inclusion of intron 2 or the exclusion of exon 3 due to alternative splicing of exon 2 to exon 4 (data not shown). This implies that intron 3 5' ss recognition may play an important role in the inclusion and exclusion of K8 intron 2 and the production of K8? mRNA.
To test this role of downstream intron recognition in upstream intron splicing, we inserted a human ?-globin intron into the exon 3-exon 4 splice junction of a K8? cDNA construct, pST3, to create plasmid pKY4 (Fig. 5A). Since the human ?-globin intron has optimal splice sites, the splicing of this inserted intron was expected to stimulate removal of the K8? intron upstream. When pKY4 was used to transfect 293 cells, the splicing of the ?-globin intron in the pKY4-derived pre-mRNA did stimulate splicing of the upstream K8? intron, and much of the K8? RNA was converted into K8 mRNA (Fig. 5B).
Western blot analysis was performed with a polyclonal K-bZIP antibody to examine K8 and K8? protein expression simultaneously in the transfected 293 cells. As shown in Fig. 5C, a K8 genomic DNA construct, pST1, expressed mainly K8 protein in 293 cells, accompanied by a lower amount of K8? expression. In contrast, the K8? cDNA construct pST3 expressed K8? protein but very little K8 protein. The pKY4 construct, which had the ?-globin intron inserted into the pST3 exon 3-exon 4 splice junction, expressed almost equal amounts of K8 and K8? protein, as predicted (Fig. 5C).
K8? regulates K8 function. As described above, the inclusion of K8 intron 2 also creates a PTC in the K8? mRNA and results in the production of the truncated protein K8? in KSHV+ JSC-1 cells and BCBL-1 cells as well as in transient transfected epithelial cells. Thus, both K8 protein and K8? protein have the same N-terminal structure, but they differ at their C-terminal ends, where a leucine zipper domain in K8 is missing in K8?. Recent reports indicate that one of the functions of KSHV K8 during viral lytic replication in PEL cells is to induce expression of the universal CDK inhibitor p21 by stabilizing p21 from proteasome-mediated degradation (20, 52, 53). This observation prompted us to wonder whether the expression of K8? might antagonize the activity of K8 if this truncated protein cannot similarly induce p21.
To examine the biological functions of K8?, we investigated the effect of KSHV K8 and K8? by transient transfection on the expression of p21 initially in HeLa cells. We transiently transfected HeLa cells with a K8 expression vector that predominantly expresses K8 (pST1) or with the K8? expression vector (pST3) and found that only the pST1-transfected HeLa cells strongly induced p21 (Fig. 6A, lane 2). Although pST3 itself had no effect on p21 induction in HeLa cells (Fig. 6B), pST3 did suppress the pST1-induced p21 expression in a dose-dependent manner (Fig. 6A, compare lanes 3 to 5 and lane 2; Fig. 6C). At a 1:1 ratio of pST1 and pST3, the expression of both K8 and K8? was reduced (Fig. 6D, lane 5). Interestingly, this reciprocal suppression of K8 and K8? appeared only in the HeLa cells cotransfected with pST1 and pST3, not in cells cotransfected with pST1 or with pST3 and an empty vector, pcDNA3 (Fig. 6E and F, lanes 4). Thus, the expression of K8? antagonizes the function of K8.
Since p21 expression is usually induced by p53, it is conceivable that K8-mediated p21 accumulation might occur because of an accumulated p53 expression in cells overexpressing K8. To address this question, we examined p53 expression by using HeLa cells transfected with pST1 (K8), pST3 (K8?), or pST1 and pST3 in combination. HeLa cells were selected for the study, since they contain only a residual amount of p53. As shown in Fig. 7A and B, the overexpression of K8 (pST1) in HeLa cells was found to induce >2.5-fold activation activity of p53 in p53 reporter assays, which could be translated into a 71% higher p53 protein accumulation in the cells by Western blotting. Although the overexpression of K8? (pST3) had no effect on p53 accumulation, HeLa cells cotransfected with both K8 (pST1) and K8? (pST3) expression vectors showed no p53 accumulation or activation activity but, rather, p53 reduction. Data suggest that the K8-mediated p21 accumulation and its suppression by K8? may be attributable partially to K8-mediated p53 expression.
A recent study also suggests that the inhibition of CDK2 activity by K-bZIP through direct CDK2-K-bZIP interaction is partially responsible for the increased p21 protein level (20), since phosphorylation by CDK2 targets p21 for degradation (38). We wondered whether K8? could also affect the interaction of K8 and CDK2 and examined this interaction by using HeLa cells transfected with pST1 (K8) or pST3 (K8?) or pST1 plus pST3 in combination. The cell lysates were immunoprecipitated with anti-CDK2-coated beads, and the precipitates were immunoblotted with anti-K-bZIP. As shown in Fig. 7C, K8 (pST1) did interact with CDK2 (lane 2), but this interaction could be completely eliminated in the presence of K8? (pST3, lane 3) despite the fact that the latter had no detectable interaction with CDK2 (lane 4). These data, together with the p53 results described above, suggest that K8 mediates the increase in p21 expression, presumably through several pathways.
K8? affects K8 expression at the mRNA level. As stated above, K8? differs from K8 by lacking a C-terminal leucine zipper motif which is required for K8 dimerization and functions. Thus, it is very unlikely that K8? could form a dimer with K8 as a mechanism of the K8?-mediated suppression. Considering the fact that K8? also reduces K8 protein expression in HeLa cell cotransfection assays (Fig. 6D, lane 5), we were interested to know if the K8?-mediated K8 protein reduction is a result of K8?-mediated suppression of K8 RNA expression, since K8? has the same exon structures in K8 mRNA and differs from K8 only by the inclusion of a K8 intron 2 as an extra piece in its mRNA. As demonstrated in Fig. 8, the cotransfection of HeLa cells with a K8? expression vector (pST3) resulted in an approximately fivefold reduction of K8 mRNA (pST1) in RT-PCR assays (Fig. 8A, compare lanes 7 and 8). To further confirm our RT-PCR results, total cell RNA obtained from separate experiments was analyzed by RPA with an antisense K8 probe and showed a sixfold reduction of K8 mRNA (pST1) in the presence of K8? (pST3) (Fig. 8B, compare lanes 6 and 8). Although K8 mRNA expression was reduced greatly in the presence of K8?, the K8? mRNA level remained relatively constant in the presence of K8 (compare lanes 6 and 9). Together, the results from two different assays lead us to conclude that K8?-mediated suppression of K8 function is most likely a consequence of K8?-mediated inhibition of K8 RNA expression.
Since KSHV+ B cells express both K8 and K8? via alternative RNA splicing during lytic infection, it is difficult to dissect the function of K8 from K8? during lytic infection in those cells. More importantly, the K8 gene overlaps the ORF50 gene in the virus genome despite the fact that each gene has its own ORF, as shown in Fig. 1A. Accordingly, it is not feasible to study K8 function in its native context without disruption of the overlapped ORF50. To duplicate our observation with HeLa cells, we analyzed K8? on K8 function with BCBL-1 cells by transient transfection. As shown in Fig. 9A, TPA activation of BCBL-1 cells (lane 2) or expression of KSHV ORF50 (lane 4) or K8 (lane 5, pST1) alone in BCBL-1 cells promotes p21 expression. The expression of K8? alone (lane 7, pST3) in BCBL-1 cells did not induce p21 expression, as was done by its vector control (pcDNA, lane 3). As expected, we were able to demonstrate that the expression of K8? could antagonize K8-mediated induction of p21 (Fig. 9A, compare lanes 5 and 6) in BCBL-1 cells. In parallel, BCBL-1 cotransfection with pST1 (K8) and pST3 (K8?) had much less production of K8 mRNAs (Fig. 9B, lane 7) than the cells transfected with pST1 alone (Fig. 9B, lanes 6), implying that K8? diminishes K8-mediated function most likely at the RNA level.
DISCUSSION
In this report, we have analyzed alternative splicing of two overlapping primary transcripts expressed during the KSHV lytic cycle, the ORF50 and K8 pre-mRNAs. As we (46) and others (13, 23, 63) previously reported, this alternative RNA splicing leads to the production of two major spliced isoforms ( and ?) of mRNAs due to the inclusion (?) or exclusion () of an intron from nt 75563 to 75645 (intron 3 of the ORF50 pre-mRNA and intron 2 of the K8 pre-mRNA). However, we were unable to detect a previously reported isoform message (23, 63).
The intron from nt 75563 to 75645 (K8 intron 2 or ORF50 intron 3) has suboptimal features, including a nonconsensus 5' ss and a weak 3' ss with purines interspersed in its putative polypyrimidine tract. As a result, K8 intron 2 interacts with U1 snRNA and U2AF protein with low affinity. The optimization of these splice sites was found to improve the interactions with U1 snRNA and U2AF and, consequently, to promote the splicing of K8 intron 2 in vitro and in vivo. Our data provide strong experimental evidence that the escape of K8 intron 2 (ORF50 intron 3) from RNA splicing can most likely be attributed to its suboptimal features. The mRNA that results from the retention of intron 2, K8?, contains a PTC in the intron at a distance of >50 nt upstream of an exon-exon junction. Despite having this PTC, the K8? mRNA bypasses an mRNA surveillance mechanism that initiates NMD to rapidly degrade mRNAs carrying PTCs. This was surprising, since a major function of NMD is to block the synthesis of truncated proteins that could have dominant negative effects (25). Although there are many other characteristics of mammalian transcripts that determine the efficiency of NMD, including a nonsense codon positioned close to the initiation AUG (4, 18, 30, 54), the absence of an NMD-promoting element (3), and the presence of NMD-mediated up-regulation of RNA splicing (48), we do not know how the K8? mRNA avoids NMD. Nevertheless, in PEL cells and transiently transfected mammalian cells, the K8? mRNA produces a K8? protein that is 48 amino acids smaller than the K-bZIP protein encoded by fully spliced K8 mRNA. This led us to speculate that the PTC-containing K8? might avoid the NMD pathway because the encoded protein has a function in viral gene expression.
Despite the suboptimal features of the intron from nt 75563 to 75645, the majority of the K8 pre-mRNAs completely remove this intron to produce a fully spliced K8 message, which encodes a K-bZIP protein important for viral gene transcription (19, 21, 28, 49), DNA replication (1, 22, 51), and cell cycle arrest (20, 52, 53). To investigate what promotes the splicing of this intron, we demonstrated that a cis element, a nt 75838 5' ss crossing over exon 3, could trigger the splicing. This finding is consistent with a well-established exon definition proposing that cross talk between a 3' ss and a 5' ss over the exon promotes splicing of an upstream intron (2, 17, 42, 57). Based on this principle, we inserted a human ?-globin intron into the K8? exon 3-exon 4 splice junction and demonstrated that this intron insertion stimulated efficient splicing of the K8? intron and resulted in production of K8 mRNA from a K8? cDNA. Together, our RNA and protein expression data provide conclusive evidence that the splicing of K8 intron 2 is enhanced by the exon definition machinery crossing over exon 3 and that the K8? RNA is derived from an RNA splicing intermediate during viral RNA processing.
The production of K8? and its escape from NMD-mediated degradation in early viral lytic infection suggest that the expression of K8? may play a role in viral gene expression and pathogenesis. The K8? protein has exactly the same amino acid sequences as the K-bZIP (K8) protein in its N-terminal region, including a basic domain close to the leucine zipper domain, but is missing a C-terminal leucine zipper domain encoded by the K8 exon 3 region (23). Recent studies have shown that the KSHV K-bZIP protein increases p21 expression and causes p21-mediated G1 cell cycle arrest through CCAAT/enhancer-binding protein- (C/EBP) (52, 53) and/or inhibition of CDK2 (20). The K-bZIP protein binds very efficiently to both C/EBP and p21 and stabilizes them from proteasome-mediated protein degradation. Although the only region of the K-bZIP protein that interacts with and stabilizes C/EBP is the basic leucine zipper region in the C terminus, both the N-terminal and C-terminal domains of the K-bZIP protein are required to stimulate C/EBP-induced expression from both the C/EBP and p21 promoters in cotransfected cells (53). In contrast, only the basic region (residues 123 to 189), not the leucine zipper region, of K-bZIP was found to play a functional role in CDK2 binding and down-modulation of CDK2 activity (20). Consistent with these reports, we also demonstrated in our transient transfection assays that expression of the K-bZIP protein could cause both p53 and p21 induction. However, coexpression of K-bZIP and K8? in the cells reduced the K-bZIP-mediated p53 and p21 induction, suggesting that K8? acts as an antagonist. Moreover, we further demonstrated that the expression of K8? also blocks the interaction of K-bZIP with CDK2. This is the first report of a biological function for K8?.
Although enhancement of p21 expression by K-bZIP can be induced in a p53-independent pathway (53), the finding of K-bZIP interaction with p53 (20, 28) in other laboratories also provides some convincing evidence for interpreting our results, indicating that K-bZIP-p53 interaction might stabilize p53 from degradation and result in accumulation of the p53 that activates transcription of the p21 gene. More interestingly, we found a significant reduction of p21 expression, along with a reciprocal suppression of K-bZIP and K8? expression, when equal amounts of expression vector pST1 (mainly for K-bZIP expression) and pST3 (for K8?) were cotransfected. In this case, the expression levels of the K8 mRNAs in the cotransfected cells were greatly reduced along with an unrestricted expression of K8? mRNA, suggesting that overexpression of K8? RNA might affect splicing of K8 pre-mRNA. However, how K-bZIP suppresses K8? protein expression remains to be further understood. Altogether, these observations clearly indicate that expression of K8 and K8? must remain at a constant ratio. Production of K8? over a threshold is detrimental to K8 expression and consequently is subtly regulated by K8. Supporting this is the fact that it has been found consistently that K8 and K8? mRNA are expressed at approximately 5-6 to 1 (this study and Tang and Zheng [46]). It will be interesting to know how the virus keeps this ratio of K8 and K8? mRNAs in check during viral gene expression and what controls this regulation, in addition to the exon definition we described in this study.
In summary, we have investigated in detail the features of an intron from KSHV nt 75563 to 75645 and its inclusion in a ? isoform mRNA during KSHV ORF50 and K8 expression. Because the 3' UTR of the ORF50 transcript overlaps the entire K8 transcript, we simplified our focus on understanding alternative splicing of this intron in processing of the K8 pre-mRNA to exemplify the way the inclusion or exclusion of this intron may take place during the splicing of the ORF50 pre-mRNA. Both the and ? isoforms of the ORF50 mRNA presumably encode the same protein (replication and transcription activator; RTA) with the same size, since the intron from nt 75563 to 75645 in the ORF50 transcript is positioned in its 3' UTR; however, the inclusion of this intron in the K8? mRNA leads to the production of a truncated protein that plays a dominant negative role in K8 mRNA expression and K8 function.
ACKNOWLEDGMENTS
This study was supported by the Intramural Research Program of the NIH, National Cancer Institute, Center for Cancer Research. K.Y. was supported by a research fellowship from the Japan Society for the Promotion of Science at NIH.
We thank Richard Ambinder for providing JSC-1 cells and the NIAID AIDS Research and Reference Reagent Program, Division of AIDS, for providing BCBL-1 cells originally obtained from Michael McGrath and Don Ganem. We also thank Don Ganem for providing the rabbit polyclonal anti-K-bZIP antibody, John Brady for p53 reporter plasmid, and Yan Yuan for pORF50 expression vector.
REFERENCES
AuCoin, D. P., K. S. Colletti, S. A. Cei, I. Papouskova, M. Tarrant, and G. S. Pari. 2004. Amplification of the Kaposi's sarcoma-associated herpesvirus/human herpesvirus 8 lytic origin of DNA replication is dependent upon a cis-acting AT-rich region and an ORF50 response element and the trans-acting factors ORF50 (K-Rta) and K8 (K-bZIP). Virology 318:542-555.
Berget, S. M. 1995. Exon recognition in vertebrate splicing. J. Biol. Chem. 270:2411-2414.
Bühler, M., A. Paillusson, and O. Mühlemann. 2004. Efficient downregulation of immunoglobulin μ mRNA with premature translation-termination codons requires the 5'-half of the VDJ exon. Nucleic Acids Res. 32: 3304-3315.
Buzina, A., and M. J. Shulman. 1999. Infrequent translation of a nonsense codon is sufficient to decrease mRNA level. Mol. Biol. Cell 10:515-524.
Carter, M. S., J. Doskow, P. Morris, S. Li, R. P. Nhim, S. Sandstedt, and M. F. Wilkinson. 1995. A regulatory mechanism that detects premature nonsense codons in T-cell receptor transcripts in vivo is reversed by protein synthesis inhibitors in vitro. J. Biol. Chem. 270:28995-29003.
Chandran, B., C. Bloomer, S. R. Chan, L. Zhu, E. Goldstein, and R. Horvat. 1998. Human herpesvirus-8 ORF K8.1 gene encodes immunogenic glycoproteins generated by spliced transcripts. Virology 249:140-149.
Chang, Y., E. Cesarman, M. S. Pessin, F. Lee, J. Culpepper, D. M. Knowles, and P. S. Moore. 1994. Identification of herpesvirus-like DNA sequences in AIDS-associated Kaposi's sarcoma. Science 266:1865-1869.
Donis-Keller, H. 1979. Site specific enzymatic cleavage of RNA. Nucleic Acids Res. 7:179-192.
Fakhari, F. D., and D. P. Dittmer. 2002. Charting latency transcripts in Kaposi's sarcoma-associated herpesvirus by whole-genome real-time quantitative PCR. J. Virol. 76:6213-6223.
Frilander, M. J., and J. A. Steitz. 1999. Initial recognition of U12-dependent introns requires both U11/5' splice-site and U12/branchpoint interactions. Genes Dev. 13:851-863.
Graveley, B. R. 2000. Sorting out the complexity of SR protein functions. RNA 6:1197-1211.
Graveley, B. R., K. J. Hertel, and T. Maniatis. 2001. The role of U2AF35 and U2AF65 in enhancer-dependent splicing. RNA 7:806-818.
Gruffat, H., S. Portes-Sentis, A. Sergeant, and E. Manet. 1999. Kaposi's sarcoma-associated herpesvirus (human herpesvirus-8) encodes a homologue of the Epstein-Barr virus bZip protein EB1. J. Gen. Virol. 80:557-561.
Gunzl, A., Z. Palfi, and A. Bindereif. 2002. Analysis of RNA-protein complexes by oligonucleotide-targeted RNase H digestion. Methods 26:162-169.
Hamm, J., and I. W. Mattaj. 1990. Monomethylated cap structures facilitate RNA export from the nucleus. Cell 63:109-118.
Hayward, G. S. 2003. Initiation of angiogenic Kaposi's sarcoma lesions. Cancer Cell 3:1-3.
Hwang, D. Y., and J. B. Cohen. 1996. Base pairing at the 5' splice site with U1 small nuclear RNA promotes splicing of the upstream intron but may be dispensable for slicing of the downstream intron. Mol. Cell. Biol. 16: 3012-3022.
Inacio, A., A. L. Silva, J. Pinto, X. Ji, A. Morgado, F. Almeida, P. Faustino, J. Lavinha, S. A. Liebhaber, and L. Romao. 2004. Nonsense mutations in close proximity to the initiation codon fail to trigger full nonsense-mediated mRNA decay. J. Biol. Chem. 279:32170-32180.
Izumiya, Y., S. F. Lin, T. Ellison, L. Y. Chen, C. Izumiya, P. Luciw, and H. J. Kung. 2003. Kaposi's sarcoma-associated herpesvirus K-bZIP is a coregulator of K-Rta: physical association and promoter-dependent transcriptional repression. J. Virol. 77:1441-1451.
Izumiya, Y., S. F. Lin, T. J. Ellison, A. M. Levy, G. L. Mayeur, C. Izumiya, and H. J. Kung. 2003. Cell cycle regulation by Kaposi's sarcoma-associated herpesvirus K-bZIP: direct interaction with cyclin-CDK2 and induction of G1 growth arrest. J. Virol. 77:9652-9661.
Liao, W., Y. Tang, S. F. Lin, H. J. Kung, and C. Z. Giam. 2003. K-bZIP of Kaposi's sarcoma-associated herpesvirus/human herpesvirus 8 (KSHV/HHV-8) binds KSHV/HHV-8 Rta and represses Rta-mediated transactivation. J. Virol. 77:3809-3815.
Lin, C. L., H. Li, Y. Wang, F. X. Zhu, S. Kudchodkar, and Y. Yuan. 2003. Kaposi's sarcoma-associated herpesvirus lytic origin (ori-Lyt)-dependent DNA replication: identification of the ori-Lyt and association of K8 bZip protein with the origin. J. Virol. 77:5578-5588.
Lin, S. F., D. R. Robinson, G. Miller, and H. J. Kung. 1999. Kaposi's sarcoma-associated herpesvirus encodes a bZIP protein with homology to BZLF1 of Epstein-Barr virus. J. Virol. 73:1909-1917.
Lukac, D. M., J. R. Kirshner, and D. Ganem. 1999. Transcriptional activation by the product of open reading frame 50 of Kaposi's sarcoma-associated herpesvirus is required for lytic viral reactivation in B cells. J. Virol. 73: 9348-9361.
Maquat, L. E. 2004. Nonsense-mediated mRNA decay: splicing, translation and mRNP dynamics. Nat. Rev. Mol. Cell Biol. 5:89-99.
Miller, G., M. O. Rigsby, L. Heston, E. Grogan, R. Sun, C. Metroka, J. A. Levy, S. J. Gao, Y. Chang, and P. Moore. 1996. Antibodies to butyrate-inducible antigens of Kaposi's sarcoma-associated herpesvirus in patients with HIV-1 infection. N. Engl. J. Med. 334:1292-1297.
Moore, P. S., S. J. Gao, G. Dominguez, E. Cesarman, O. Lungu, D. M. Knowles, R. Garber, P. E. Pellett, D. J. McGeoch, and Y. Chang. 1996. Primary characterization of a herpesvirus agent associated with Kaposi's sarcomae. J. Virol. 70:549-558.
Park, J., T. Seo, S. Hwang, D. Lee, Y. Gwack, and J. Choe. 2000. The K-bZIP protein from Kaposi's sarcoma-associated herpesvirus interacts with p53 and represses its transcriptional activity. J. Virol. 74:11977-11982.
Perez, I., C. H. Lin, J. G. McAfee, and J. G. Patton. 1997. Mutation of PTB binding sites causes misregulation of alternative 3' splice site selection in vivo. RNA 3:764-778.
Perrin-Vidoz, L., O. M. Sinilnikova, D. Stoppa-Lyonnet, G. M. Lenoir, and S. Mazoyer. 2002. The nonsense-mediated mRNA decay pathway triggers degradation of most BRCA1 mRNAs bearing premature termination codons. Hum. Mol. Genet. 11:2805-2814.
Pise-Masison, C. A., R. Mahieux, M. Radonovich, H. Jiang, and J. N. Brady. 2001. Human T-lymphotropic virus type I Tax protein utilizes distinct pathways for p53 inhibition that are cell type-dependent. J. Biol. Chem. 276: 200-205.
Raab, M. S., J. C. Albrecht, A. Birkmann, S. Yaguboglu, D. Lang, B. Fleckenstein, and F. Neipel. 1998. The immunogenic glycoprotein gp35-37 of human herpesvirus 8 is encoded by open reading frame K8.1. J. Virol. 72:6725-6731.
Rajavel, K. S., and E. F. Neufeld. 2001. Nonsense-mediated decay of human HEXA mRNA. Mol. Cell. Biol. 21:5512-5519.
Renne, R., W. Zhong, B. Herndier, M. McGrath, N. Abbey, D. Kedes, and D. Ganem. 1996. Lytic growth of Kaposi's sarcoma-associated herpesvirus (human herpesvirus 8) in culture. Nat. Med. 2:342-346.
Russo, J. J., R. A. Bohenzky, M. C. Chien, J. Chen, M. Yan, D. Maddalena, J. P. Parry, D. Peruzzi, I. S. Edelman, Y. Chang, and P. S. Moore. 1996. Nucleotide sequence of the Kaposi sarcoma-associated herpesvirus (HHV8). Proc. Natl. Acad. Sci. USA 93:14862-14867.
Sakakibara, S., K. Ueda, J. Chen, T. Okuno, and K. Yamanishi. 2001. Octamer-binding sequence is a key element for the autoregulation of Kaposi's sarcoma-associated herpesvirus ORF50/Lyta gene expression. J. Virol. 75:6894-6900.
Reference deleted.
Sherr, C. J., and J. M. Roberts. 1999. CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev. 13:1501-1512.
Smith, C. W., E. B. Porro, J. G. Patton, and B. Nadal-Ginard. 1989. Scanning from an independently specified branch point defines the 3' splice site of mammalian introns. Nature 342:243-247.
Soulier, J., L. Grollet, E. Oksenhendler, P. Cacoub, D. Cazals-Hatem, P. Babinet, M. F. d'Agay, J. P. Clauvel, M. Raphael, L. Degos, and F. Sigaux. 1995. Kaposi's sarcoma-associated herpesvirus-like DNA sequences in multicentric Castleman's disease. Blood 86:1276-1280.
Staskus, K. A., R. Sun, G. Miller, P. Racz, A. Jaslowski, C. Metroka, H. Brett-Smith, and A. T. Haase. 1999. Cellular tropism and viral interleukin-6 expression distinguish human herpesvirus 8 involvement in Kaposi's sarcoma, primary effusion lymphoma, and multicentric Castleman's disease. J. Virol. 73:4181-4187.
Sterner, D. A., T. Carlo, and S. M. Berget. 1996. Architectural limits on split genes. Proc. Natl. Acad. Sci. USA 93:15081-15085.
Sun, R., S. F. Lin, L. Gradoville, Y. Yuan, F. Zhu, and G. Miller. 1998. A viral gene that activates lytic cycle expression of Kaposi's sarcoma-associated herpesvirus. Proc. Natl. Acad. Sci. USA 95:10866-10871.
Sun, R., S. F. Lin, K. Staskus, L. Gradoville, E. Grogan, A. Haase, and G. Miller. 1999. Kinetics of Kaposi's sarcoma-associated herpesvirus gene expression. J. Virol. 73:2232-2242.
Tang, S., K. Yamanegi, and Z. M. Zheng. 2004. Requirement of a 12-base-pair TATT-containing sequence and viral lytic DNA replication in activation of the Kaposi's sarcoma-associated herpesvirus K8.1 late promoter. J. Virol. 78:2609-2614.
Tang, S., and Z. M. Zheng. 2002. Kaposi's sarcoma-associated herpesvirus K8 exon 3 contains three 5'-splice sites and harbors a K8.1 transcription start site. J. Biol. Chem. 277:14547-14556.
Viejo-Borbolla, A., M. Ottinger, and T. F. Schulz. 2003. Human herpesvirus 8: biology and role in the pathogenesis of Kaposi's sarcoma and other AIDS-related malignancies. Curr. Infect. Dis. Rep. 5:169-175.
Wang, J., Y. F. Chang, J. I. Hamilton, and M. F. Wilkinson. 2002. Nonsense-associated altered splicing: a frame-dependent response distinct from nonsense-mediated decay. Mol. Cell 10:951-957.
Wang, S. E., F. Y. Wu, Y. Yu, and G. S. Hayward. 2003. CCAAT/enhancer-binding protein- is induced during the early stages of Kaposi's sarcoma-associated herpesvirus (KSHV) lytic cycle reactivation and together with the KSHV replication and transcription activator (RTA) cooperatively stimulates the viral RTA, MTA, and PAN promoters. J. Virol. 77:9590-9612.
Reference deleted.
Wu, F. Y., J. H. Ahn, D. J. Alcendor, W. J. Jang, J. Xiao, S. D. Hayward, and G. S. Hayward. 2001. Origin-independent assembly of Kaposi's sarcoma-associated herpesvirus DNA replication compartments in transient cotransfection assays and association with the ORF-K8 protein and cellular PML. J. Virol. 75:1487-1506.
Wu, F. Y., Q. Q. Tang, H. Chen, C. ApRhys, C. Farrell, J. Chen, M. Fujimuro, M. D. Lane, and G. S. Hayward. 2002. Lytic replication-associated protein (RAP) encoded by Kaposi sarcoma-associated herpesvirus causes p21CIP-1-mediated G1 cell cycle arrest through CCAAT/enhancer-binding protein-. Proc. Natl. Acad. Sci. USA 99:10683-10688.
Wu, F. Y., S. E. Wang, Q. Q. Tang, M. Fujimuro, C. J. Chiou, Q. Zheng, H. Chen, S. D. Hayward, M. D. Lane, and G. S. Hayward. 2003. Cell cycle arrest by Kaposi's sarcoma-associated herpesvirus replication-associated protein is mediated at both the transcriptional and posttranslational levels by binding to CCAAT/enhancer-binding protein alpha and p21(CIP-1). J. Virol. 77:8893-8914.
Zhang, J., and L. E. Maquat. 1997. Evidence that translation reinitiation abrogates nonsense-mediated mRNA decay in mammalian cells. EMBO J. 16:826-833.
Zhang, M. Q. 1998. Statistical features of human exons and their flanking regions. Hum. Mol. Genet. 7:919-932.
Zheng, Z. M. 2003. Split genes and their expression in Kaposi's sarcoma-associated herpesvirus. Rev. Med. Virol. 13:173-184.
Zheng, Z. M. 2004. Regulation of alternative RNA splicing by exon definition and exon sequences in viral and Mammalian gene expression. J. Biomed. Sci. 11:278-294.
Zheng, Z. M., and C. C. Baker. 2000. Parameters that affect in vitro splicing of bovine papillomavirus type 1 late pre-mRNAs. J. Virol. Methods 85: 203-214.
Zheng, Z. M., P. J. He, and C. C. Baker. 1997. Structural, functional, and protein binding analyses of bovine papillomavirus type 1 exonic splicing enhancers. J. Virol. 71:9096-9107.
Zheng, Z. M., E. S. Reid, and C. C. Baker. 2000. Utilization of the bovine papillomavirus type 1 late-stage-specific nucleotide 3605 3' splice site is modulated by a novel exonic bipartite regulator but not by an intronic purine-rich element. J. Virol. 74:10612-10622.
Zheng, Z. M., M. Tao, K. Yamanegi, S. Bodaghi, and W. Xiao. 2004. Splicing of a cap-proximal human papillomavirus 16 E6E7 intron promotes E7 expression, but can be restrained by distance of the intron from its RNA 5' cap. J. Mol. Biol. 337:1091-1108.
Zhong, W., H. Wang, B. Herndier, and D. Ganem. 1996. Restricted expression of Kaposi sarcoma-associated herpesvirus (human herpesvirus 8) genes in Kaposi sarcoma. Proc. Natl. Acad. Sci. USA 93:6641-6646.
Zhu, F. X., T. Cusano, and Y. Yuan. 1999. Identification of the immediate-early transcripts of Kaposi's sarcoma-associated herpesvirus. J. Virol. 73:5556-5567.(Koji Yamanegi, Shuang Tan)