Specific Inhibition of Serine- and Arginine-rich ... - Cancer Research

[CANCER RESEARCH 61; 6876 – 6884, September 15, 2001]

Specific Inhibition of Serine- and Arginine-rich Splicing Factors Phosphorylation, Spliceosome Assembly, and Splicing by the Antitumor Drug NB-5061 Bartosz Pilch,2 Eric Allemand,2 Michael Facompre´, Christian Bailly, Jean-François Riou, Johann Soret, and Jamal Tazi3 Institut de Geńe´tique Molećulaire, UMR 5535, Centre National de la Recherche´ Scientifique, IFR 24, Universite´ de Montpellier II, 34293 Montpellier Cedex [B. P., E. A., J. S, J. T.]; Unite´ MEDIAN, Centre National de la Recherche Scientifique, FRE 2141, UFR de Pharmacie, Universite´ de Reims Champagne-Ardenne, 51096 Reims Cedex [J. F. R.]; and Laboratoire de Pharmacologie Molećulaire Antitumorale du Centre Oscar Lambret et U-124 Institut National de la Sante´ et de la Recherche´ Me´dicale, 59045 Lille Cedex [C. B., M. F.], France

ABSTRACT Specific phosphorylation of serine- and arginine-rich pre-mRNA splicing factors (SR proteins) is one of the key determinants regulating splicing events. Several kinases involved in SR protein phosphorylation have been identified and characterized, among which human DNA topoisomerase I is known to have DNA-relaxing activity. In this study, we have investigated the mechanism of splicing inhibition by a glycosylated indolocarbazole derivative (NB-506), a potent inhibitor of both kinase and relaxing activities of topoisomerase I. NB-506 completely inhibits the capacity of topoisomerase I to phosphorylate, in vitro, the human splicing factor 2/alternative splicing factor (SF2/ASF). This inhibition is specific, because NB-506 does not demonstrate activity against other kinases known to phosphorylate SF2/ASF such as SR protein kinase 1 and cdc2 kinase. Importantly, HeLa nuclear extracts competent in splicing but not splicingdeficient cytoplasmic S100 extracts treated with the drug fail to phosphorylate SF2/ASF and to support splicing of pre-mRNA substrates containing SF2/ASF-target sequences. Native gel analysis of splicing complexes revealed that the drug affects the formation of the spliceosome, a dynamic ribonucleoprotein structure where splicing takes place. In the presence of the drug, neither pre-spliceosome nor spliceosome is formed, demonstrating that splicing inhibition occurs at early steps of spliceosome assembly. Splicing inhibition can be relieved by adding phosphorylated SF2/ASF, showing that extracts treated with NB-506 lack a phosphorylating activity required for splicing. Moreover, NB-506 has a cytotoxic effect on murine P388 leukemia cells but not on P388CPT5 camptothecin-resistant cells that carry two point mutations in conserved regions of topoisomerase I gene (Gly361Val and Asp709Tyr). After drug treatment, P388 cells accumulated hypophosphorylated forms of SR proteins and polyadenylated RNA in the nucleus. In contrast, neither SR protein phosphorylation nor polyadenylated mRNA distribution was affected in P388 CPT5-treated cells. Consistently, NB506 treatment altered the mRNA levels and/or splicing pattern of several tested genes (Bcl-X, CD 44, SC35, and Sty) in P388 cells but not in P388 CPT5 cells. The study shows for the first time that indolocarbazole drugs targeting topoisomerase I can affect gene expression by modulating pre-mRNA splicing through inhibition of SR proteins phosphorylation.

INTRODUCTION Pre-mRNA splicing is an essential step in the expression of most metazoan protein-coding genes that can lead from a single array of coding sequences (exons) to the production of different proteins or, alternatively, to a lack of protein expression by introduction of a Received 3/14/01; accepted 7/16/01. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Supported by grants (to J. T.) from the Association pour la Recherche´ sur le Cancer. E. A. was supported by graduate fellowships from the Ministe`re de l’Education Nationale, de la Recherche et de la Technologie and benefited from a graduate training fellowship from the Association pour la Recherche contre le Cancer and Fondation pour la Recherche Me´dicale. B. P. has benefited from an undergraduate training fellowship from the ERASMUS program. 2 Both authors contributed equally to this work. 3 To whom requests for reprints should be addressed, at IGM-CNRS, 1919 Route de Mende, 34293 Montpellier, France. Phone: (33) 04-67-61-36-85; Fax: (33) 04-67-04-0245; E-mail: [email protected].

premature stop codon (1). Determining how these sometimes subtle changes in mRNA sequences affect protein function is a crucial question in many aspects of developmental and cell biology, including control of apoptosis (2) and tumor progression (3). The accurate excision of intervening sequences (introns) from premRNAs by two transesterification reactions occurs in a large ribonucleoprotein complex called spliceosome (4). The spliceosome formation results from a dynamic series of interactions between the four major snRNPs4 (U1, U2, U4/U6, and U5), non-snRNP proteins, and the pre-mRNA (4). Several metazoan splicing factors involved in splice site selection at earliest steps of the spliceosome assembly pathway are characterized by RNA recognition motif and RS domains that are essential for their function (1, 5). These include members of SR protein family (SRp20, SF2/ASF or SRp30a, SC35 or SRp30b, 9G8, SRp30c, SRp40, SRp55, and SRp75; Refs. 5, 6), U1 snRNPspecific protein U1-70K, and the splicing factor U2AF, which comprises two subunits (35K and 65K; Ref. 7). All of the these factors can mediate a network of protein-protein interactions through their RS domains, resulting in the stimulation and/or stabilization of complexes assembled at the 5⬘ or 3⬘ splice sites (7, 8). These interactions can directly be affected by phosphorylation of the serines at the RS domains (9 –14). Although regulation of SR protein phosphorylation and the functional consequences associated with it are just beginning to be understood, it is established that SR proteins can be phosphorylated by several kinases (15). We have established recently that DNA topoisomerase I, a constitutively expressed nuclear phospho-protein that localizes to active transcription sites and belongs to a super family of DNA topoisomerases and tyrosine recombinases (16, 17), exhibits a SR protein-phosphorylating activity in addition to its DNA-relaxing activity (18). Although topoisomerase I lacks sequence motifs homologous to known protein kinases (e.g., ATP-binding site), it efficiently binds ATP (18, 19). Photoaffinity labeling combined with mutational analysis showed that the COOH terminal is required for ATP binding (19), whereas the NH2-terminal 174 amino acids are essential for the interaction with the RS domain of SF2/ASF protein in vitro (20). DNA topoisomerase I is also the nuclear target for a number of anticancer agents derived from the plant alkaloid CPT and for indolocarbazole derivatives (16) such as the antibiotic rebeccamycin, the antitumor agent NB-506, which is undergoing clinical trials (21), and its regio-isomer J-109,382 (22). Although NB-506 bears a structural analogy with the specific PKC inhibitor staurosporine, it has no significant effect on PKC (23) but inhibits both relaxation and kinase activities of topoisomerase I (shown here). In this study, we show that NB-506 has the capacity to block, in vitro, spliceosome assembly and splicing through inhibition of SR protein phosphorylation. NB-506 leads also to specific inhibition of SR protein phosphorylation in 4 The abbreviations used are: snRNP, small nuclear ribonucleoprotein; RS, arginineserine rich; SF2/ASF, splicing factor 2/alternative splicing factor; CPT, camptothecin; PKC, protein kinase C; RT-PCR, reverse transcription-PCR; SRPK1, SR protein kinase 1; NE, nuclear extract; oligo(dT), oligodeoxythymidylic acid.

6876

SPLICING INHIBITION BY NB-506 DRUG

cultured cells and modulates gene expression by changing the splicing pattern of protein-encoding genes. To our knowledge, NB-506 is the first antitumor drug shown to interfere with the spliceosome assembly pathway.

MATERIALS AND METHODS

to a 12% SDS-polyacrylamide gel. Radioactivity incorporated into SF2/ASF was detected by autoradiography of the dried gel and quantitated with a Molecular Dynamics imaging analyzer. The relative amount of [␥-32P]ATP incorporated in each assay was quantitated by densitometry scanning of the gel using ImageQuant software version 3.22. The kinetics of phosphorylation in HeLa NEs or S100 extracts were performed under standard splicing conditions (27), except that polyvinyl alcohol was omitted. Reaction mixtures in a total volume of 20 ␮l contained 200 ng of indicated recombinant proteins and 3 ␮l of extracts. Aliquots (3 ␮l) from each time point mixed with 5 ␮l of (3⫻) Laemmli loading buffer were run on 10% SDS-polyacrylamide gel and transferred to nitrocellulose by electroblotting for 90 min in 10 mM 3-(cyclohexylamino)propanesulfonic acid (pH 11.0) transfer buffer containing 10% methanol. Blots were probed by X-press antibody (Invitrogen) and visualized using a chemiluminescence kit (Amersham). Splicing and Spliceosome Assembly Assays. The single-intron human ␤-globin construct pSPH␤-3S was described (28). Radiolabeled ␤-3S RNA was synthesized by in vitro transcription in the presence of 20 units of SP6 RNA polymerase (Boehringer), 1 ␮g of pSPH␤-3S plasmid linearized with BamHI, and 5 ␮M [␣-32P]UTP (800 Ci/mmol) in 25-␮l reactions according to the manufacturer’s conditions. The splicing reactions were done under standard conditions for 1 h in a total volume of 20 ␮l containing 50 fmol labeled pre-mRNA and 6, 8, or 10 ␮l of HeLa NE complemented with buffer D (27). Splicing products were analyzed by electrophoresis on denaturing 6% polyacrylamide gels and revealed by autoradiography. Kinetics of appearance of splicing complexes were performed as described (29). Aliquots (5 ␮l) from the various reactions treated with heparin (2 mg/ml) were mixed with 1 ␮l of 97% glycerol, 1% bromophenol blue and resolved directly on a 4% nondenaturing polyacrylamide gel (acrylamide:bis-acrylamide weight ratio of 80:1) in 50 mM tris-glycine (pH 8.3). Determination of Splicing Patterns Using RT-PCR Assay. Cytoplasmic mRNAs were purified from P388 and CPT5 cells before and after a 24-h treatment with 1.0 ␮M NB-506. First-strand cDNA was synthesized from 5 ␮g of DNase-treated RNA with the Amersham cDNA synthesis kit. For analysis of SC35, Clk/Sty, CD44, and Bcl-X alternatively spliced mRNA isoforms, one tenth of the reaction was amplified with Gold Taq polymerase (Hybaid) with the following forward and reverse primers (name, sequence is given 5⬘ to 3⬘): SC35 (forward), CAAGTCTCCAGAAGAAGAGG; SC35 (reverse), GCATTACTCAACTGCTACAC; Clk/Sty (forward), GCATAGTAGCAAGTCCTCTG; Clk/Sty (reverse), TACTGCTACACGTCTACCTC; CD44 (forward), CTATTGTCAACCGTGATGGTAC; CD44 (reverse), GCCAGGAGAGATGCCAAGATG; Bcl-X (forward), ATGTCTCAGAGCAACCGGGA; and Bcl-X (reverse), TCACTTCCGACTGAAGAGTG. The PCR cycle number was kept to a minimum to maintain linearity. After denaturation for 2 min at 94°C, the PCR regimes were 1 min at 94°C, 1 min at 56°C (SC35 and Clk/Sty), 58°C (CD44), or 54°C (Bcl-X), and 1 min (SC35, Clk/Sty, and Bcl-X) or 1 min 30 s (CD44) at 72°C for 25 cycles, followed by a final 10 min at 72°C in a Hybaid PCR Sprint machine. PCR products separated on 1.5% (SC35, Clk/Sty, and Bcl-X) or 1.0% (CD44) agarose gels containing ethidium bromide were visualized under UV light.

Drugs. NB-506 and J-109,382 were kindly provided by Dr. Tomoko Yoshinari (Banyu Pharmaceutical Co., Ltd., Tokyo, Japan). Their chemical synthesis has been reported (21, 24). Rebeccamycin, purified as described (23), was provided by Dr. Michelle Prudhomme (SEESIB, Aubie`re, France). Drugs were dissolved in DMSO at 10 mg/ml. Fresh dilutions (in 10% DMSO) were prepared immediately before use. The final DMSO concentration never exceeded 0.5% (v/v), a concentration at which DMSO (also present in the controls) does not affect the topoisomerase I activity. Cell Cultures, Drug Treatment, and mRNA Visualization. P388 murine leukemia cell line was obtained from the tumor bank of the National Cancer Institute (Bethesda, MD). P388CPT5 cell line resistant to CPT was derived from a stable clone of the P388CPT0.3 cell line obtained at the 42nd passage (25). Both cell lines were grown in RPMI 1640 medium containing 0.01 mM 2-mercaptoethanol, 10 mM L-glutamine, 10% (v/v) FCS, 100 IU/ml penicillin, 2 ␮g/ml streptomycin, 50 ␮g/ml gentamicin, and 50 ␮g/ml nystatin at 37°C in a humidified atmosphere containing 5% CO2. To localize polyadenylated mRNA, exponentially growing cells, either untreated or treated for 24 h with 1 ␮M NB-506, were harvested by centrifugation at 600 rpm/min on coverslips coated with gelatin. The cells were fixed and hybridized to CY3 (Amersham, Arlington Heights, IL) fluorophoreconjugated oligo(dT) as described (26). Coverslips were mounted in 50% glycerol containing 1 ␮g/ml 4⬘,6-diamidino-2-phenylindole to stain the cell nucleus. Immunofluorescence was performed using confocal scanning microscopy (Leica). Labeling, Purification, and Two-dimensional Gel Analysis of SR Proteins. To achieve in vivo efficient labeling of SR proteins, P388 and P388CPT5 cells were seeded at 5 ⫻ 105 cells/ml in medium supplemented with 5% (v/v) FCS and exposed to 2.5 mCi/liter carrier-free [H3]32PO4 (Amersham) for 12 h. Treated cells were grown under the same conditions except that 0.5 ␮M or 1 ␮M NB-506 was added after 6 h of labeling. SR proteins were immediately prepared from 107 freshly labeled cells as described (18). One fourth of the MgCl2 pellets of 65–90% ammonium sulfate extracts were analyzed by SDS-PAGE. The remaining samples were subjected to two-dimensional analysis with Immobiline Drystrip (110 mm; pH 3–10) using a Multiphor II electrophoresis unit (Amersham Pharmacia Biotech) according to manufacturer’s instructions. Identification of The Mouse Topoisomerase I Mutations. RT-PCR was used to amplify overlapping segments of the mouse topoisomerase I cDNA from positions 1 to 400, 270 to 1060, 601 to 1060, 901 to1300, 601 to 1300, 1201 to 1600, 1501 to 2304, 1800 to 2304, and 2100 to 2304, taking the first nucleotide of the initiating methionine as position 1. The different fragments were gel-purified and sequenced with both reverse and forward primers. Compared with existing sequences in the GenBank database (D10061), cDNAs from both P388 and P388CPT5 have A instead of C at position 363. Two other RESULTS AND DISCUSSION mutations corresponding to T instead of G at position 1082 and T instead of G Specific Inhibition of The Kinase Activity of Topoisomerase I at position 2125 were detected in the cDNA from P388CPT5 but not from P388. Sequences of the oligonucleotides used for PCR amplifications and by NB-506. We have shown previously (18) that recombinant human DNA topoisomerase I, overexpressed in insect cells using the bacusequencing are available upon request. Purification of Recombinant Topoisomerase I, SF2/ASF Proteins, and lovirus system, has the same characteristics as the native HeLa enKinase Assays. Recombinant topoisomerase I and SF2/ASF proteins were zyme. It binds ATP efficiently and specifically phosphorylates in vitro produced and purified from baculovirus-infected Sf9 cells as described (18). SR protein-splicing factors at short RS domain regions (19, 20). To The hexahistidine-tagged SF2/ASF protein was also produced and purified investigate the inhibition of kinase activity by the indolocarbazole from Escherichia coli as described (20). derivative NB-506, the purified recombinant topoisomerase I was The reaction mixtures for protein kinase activity contained 100 ng of used in kinase assays containing increasing concentrations of drug, recombinant topoisomerase I or equivalent kinase activity of GST-SRPK1 (a and kinase activity was detected by its ability to phosphorylate bacgift from Dr. Thomas Giannakouros) or p34 cdc2 (provided by M. Dore´ terially expressed recombinant SF2/ASF with [␥-32P]ATP. Analysis Laboratory), 300 ng of recombinant SF2/ASF protein, 1 ␮M ATP, 3 ␮Ci of [␥-32P]ATP (3000 Ci/mmol), and 1 ␮l of compound NB-506 or DMSO in a of the phosphorylated products by SDS-PAGE and detection by final volume of 20 ␮l of buffer A [50 mM HEPES (pH 7.0), 10 mM MgCl2, 3 autoradiography (Fig. 1) revealed that, under conditions where the 32 mM MnCl2, 50 mM KCl, and 0.5 mM DTT]. The samples were incubated at concentration of [␥- P]ATP was constant, the phosphorylation of 30°C for 30 min and then mixed with 6 ␮l of (4⫻) loading buffer and applied SF2/ASF is markedly inhibited in the presence of NB-506 (Fig. 1, 6877


Fig. 1. Inhibition of topoisomerase I kinase activity by NB-506. Phosphorylation of SF2/ASF by the kinase activity of topoisomerase I (panel TOPO I), SRPK1 (panel SRPK1), and p34 cdc2 kinase (panel Cdc2) is as described in “Materials and Methods” with increasing concentration of the drug: 1 ␮M, 10 ␮M, and 50 ␮M (Lanes 1–3, 4 – 6, or 7–9, respectively). Ctr, the control reaction with 10% DMSO alone instead of NB-506. The labeled SF2/ASF was analyzed on a 12% SDS-polyacrylamide gel and revealed by autoradiography. The position of SF2/ASF was localized by staining the gel with Coomassie Blue.

panel TOPO I). The kinase activity of topoisomerase I is reduced by 40% in the presence of 1 ␮M drug concentration (Fig. 1, panel TOPO I, Lane 3), whereas a drug concentration of 50 ␮M is required to block the phosphorylation of SF2/ASF to more than 90% (Lane 5). IC50 for inhibition of SF2/ASF phosphorylation by NB-506 was found equal to 1.5 ␮M. The specificity of the NB-506 inhibitory effect was further confirmed by using two other kinases, SRPK1 (Fig. 1, panel SRPK1) and cdc2 kinase (Fig. 1, panel Cdc2 kinase), which have been shown previously (20) to phosphorylate in vitro the RS domain of SF2/ASF. No inhibition of the phosphorylation of SF2/ASF by either kinase was observed with NB-506, even when concentrations as high as 50 ␮M were used (Fig. 1, both panels, Lane 5). Given that NB-506 is also inactive toward PKC (23) and that indolocarbazole compounds like staurosporine usually interact with the ATP-binding site, it can be concluded that inhibition of the kinase activity of topoisomerase I does not involve the binding of NB-506 to its ATP-binding site. Consistent with this conclusion, staurosporine does not inhibit the kinase activity of topoisomerase I, and binding studies performed with fluorescently labeled ATP revealed that NB-506 does not compete with the binding of ATP (data not shown). NB-506 Prevents SF2/ASF Complete Phosphorylation from HeLa NE But Not from S100 Extract. To address whether NB-506mediated kinase inhibition affects pre-mRNA splicing, we determined whether inhibition of SF2/ASF phosphorylation can be reproduced using bacterially expressed SF2/ASF and HeLa NE or cytoplasmic S100 extracts as sources of protein kinases. The recombinant SF2/ ASF was designed to contain a NH2-terminal His-tag to facilitate purification over an immobilized metal ion affinity column (His-bind resin) and detection with an anti-his tag antibody. A time course analysis of SF2/ASF phosphorylation was performed under splicing conditions in extracts from HeLa cells, and SF2/ASF was revealed by Western blots with the anti-his tag antibody (Fig. 2). In this assay,

cold ATP and creatine phosphate to regenerate ATP are included, and changes in the phosphorylation status are qualitatively reflected by changes in electrophoretic mobility. After 120 min of incubation in NE (Fig. 2A, Lane 7) or S100 (Fig. 2B, Lane 7), the majority of SF2/ASF demonstrated a lower mobility (P-SF2/ASF) compared with unphosphorylated SF2/ASF (Fig. 2, A and B; compare Lane 1 and Lane 7), implying that kinases from either extracts can phosphorylate SF2/ASF on multiple sites. The difference in the observed mobility was entirely attributable to phosphorylation of SF2/ASF, as judged by phosphatase treatment of samples from 120 min of incubation (data not shown). Treatment of S100 extracts with concentrations of NB506 as high as 100 ␮M did not affect the mobility shift of SF2/ASF at any time points (Fig. 2B, Lanes 8 –17) showing that kinases contained in S100 extracts were refractory to inhibition by the drug. In sharp contrast, the drug caused an accumulation of both unphosphorylated and partially phosphorylated forms when NB-506 (50 ␮M and 100 ␮M) was added to NE (Fig. 2A, Lanes 8 –17). Inhibitory concentrations of NE phosphorylation also correlate with those measured with the kinase activity of purified topoisomerase I. However, phosphorylation of SF2/ASF in NE was not affected by 100 ␮M of CPT (Fig. 2A, Lanes 19 –22), which does not directly inhibit the kinase activity of topoisomerase I. Thus, the data indicate that NB-506-mediated inhibition is selective for a kinase activity of NEs, which could correspond to that of topoisomerase I, a major nuclear phospho-protein. NB-506 Inhibits Splicing. Accurate phosphorylation and dephosphorylation within the RS domain of SR proteins is important for their activities in splicing (9 –15). Given that both hyperphosphorylation and hypophosphorylation of SR proteins inhibit splicing (13) and that NB-506 prevents SF2/ASF complete phosphorylation in NE, we were particularly interested to know whether the drug would affect the splicing activity of NE. To assess the effect of NB-506 on pre-mRNA splicing, we used a model ␤-globin pre-mRNA substrate (␤-3S1) that has three copies of a high-affinity binding site for SF2/ASF established by systematic evolution of ligands by exponential enrichment (SELEX) analysis (28), inserted 20 bases downstream of the first ␤-globin intron. Because of these sequences, the ␤-3S1 substrate was efficiently spliced with low levels of HeLa NE (Fig. 3A, Lanes 1 and 9). When this preparation of HeLa NE was supplemented with increasing amounts of NB-506 (25–100 ␮M), splicing was readily inhibited in a dose-dependent manner (Fig. 3A, Lanes 2– 4). Consistently, the inhibitory effect of the drug was reduced when higher concentrations of NE were used (10 ␮l of NE; Fig. 3A, Lanes 6 – 8).

Fig. 2. Kinetics of phosphorylation of recombinant SF2/ASF expressed and purified from E. coli (eSF2/ASF) in either NEs (A) or S100 cytoplasmic extracts (B) from HeLa cells. Recombinant protein was incubated under splicing conditions in the indicated extracts in the absence (Lanes 3–7) or presence of either 50 ␮M (Lanes 8 –12) and 100 ␮M (Lanes 13–17) of NB-506 or 100 ␮M of CPT (Lanes 18 –22). Aliquots of these reactions terminated at the indicated times were fractionated by SDS-PAGE and analyzed by Western blotting with an anti-His tag antibody. Lane 1, recombinant proteins without added extracts. Lane 2, extracts without added recombinant proteins.

6878


Fig. 3. The effect of NB-506 on pre-mRNA splicing in vitro. A, NB-506 inhibits splicing of ␤-3S in HeLa NEs. 32P-labeled ␤-3S pre-mRNA (50 pmol) was incubated with either 6 ␮l (Lanes 1– 4, 9 –12, and 17–20) or 10 ␮l of HeLa cell NEs (Lanes 5– 8, 13–16, and 21–24) under splicing conditions without (Lanes 1, 5, 9, 13, 17, and 21) or with 25 ␮M (Lanes 2 and 6), 50 ␮M (Lanes 3 and 7), and 100 ␮M (Lanes 4 and 8) of NB-506 with 25 ␮M (Lanes 10 and 14), 50 ␮M (Lanes 11 and 15), and 100 ␮M (Lanes 12 and 16) of rebeccamycin or with 25 ␮M (Lanes 18 and 22), 50 ␮M (Lanes 19 and 23), and 100 ␮M (Lanes 20 and 24) of CPT. B, phosphorylated SF2/ASF relieves splicing inhibition mediated by NB-506. HeLa NEs were complemented with 30 pmol of recombinant SF2/ASF expressed and purified either from E. coli (eSF2/ASF; Lanes 2, 5, and 9) or from baculovirus-infected SF9 cells (bSF2/ASF; Lanes 3, 7, and 10) in the absence (Lanes 1– 4) or the presence of 50 ␮M (Lanes 5– 8) and 100 ␮M (Lanes 9 –11) of NB-506. Splicing products (depicted on the right of the panel) were analyzed on a 7% denaturing polyacrylamide gel and revealed by autoradiography.

No inhibition was observed with either the antibiotic rebeccamycin, a naturally occurring indolocarbazole structurally similar to NB-506 and the regio-isomer J-109,382, or CPT, which does not inhibit phosphorylation of SF2/ASF by topoisomerase I (Fig. 3A, Lanes 9 –16 and 17–24, respectively). However, splicing was also inhibited by J-109,382 (22), which was as efficient as NB-506 at inhibiting the kinase activity of topoisomerase I (data not shown), demonstrating a requirement for a specific structure of the indolocarbazole for the inhibition of splicing. Furthermore, J-109,382 did not bind RNA (data not shown) and interacts only weakly with DNA in vitro (22), indicating that splicing inhibition is not attributable to mere nonspecific binding to pre-mRNA. Interestingly, purified recombinant SF2/ASF expressed in a baculovirus system (bSF2/ASF), where the phosphorylation of recombinant proteins is expected to take place, but not unphosphorylated version expressed in bacteria (eSF2/ASF), restored full splicing activity of NB-506-treated HeLa NEs (Fig. 3B; compare Lanes 5 and 9 with Lanes 7 and 10). The amount of purified bSF2/ASF added was of the same order of magnitude of that contained in our preparation of HeLa NE, as judged from Western blot analysis performed with an antibody specific for SF2/ASF (data not shown). A similar quantity of unphosphorylated eSF2/ASF was rather detrimental for splicing (Fig. 3B, Lanes 2 and 5). Thus, the data support the notion that NB-506 mediates splicing inhibition by preventing complete phosphorylation of the SF2/ASF SR protein. NB-506 Inhibits Spliceosome Assembly. SR proteins are known to promote the earliest stages of the spliceosome assembly (5, 7). Therefore, we tested whether NB-506 could interfere with the spliceosome formation. 32P-labeled ␤-3S1 was incubated with HeLa NE treated with various concentrations of NB-506, and assembled ribonucleoprotein complexes were analyzed by native gel electrophoresis (29). As shown in Fig. 4A, the untreated control extracts show the characteristic pattern of spliceosome assembly pathway. Two heparin-resistant complexes corresponding to complex A and complex B are formed in an ATP-dependent fashion at early time points, in addition to a fast migrating nonspecific complex H (Fig. 4A, Lane 3). The complex A formation involves

the stable binding of U2 snRNP to the pre-mRNA sequences near to the 3⬘ splice site (pre-spliceosome). Subsequent binding of a tri-snRNP particle U4/U5/U6 rather than that of the individual snRNPs gives rise to the B complex (spliceosome). Therefore, at later time points (10 –30 min), the level of complex A gradually decreased with parallel increase in the amount of complex B (Fig. 4A, Lanes 4 – 6). A conformational change probably involving disruption of base-paired U4 and U6 snRNAs appears to be a critical step in the initiation of splicing reactions by the assembled B complex (4, 15). Thus, upon longer incubation (60 min), the B form was converted to a faster migrating complex containing the products of the splicing reaction (complex X; Fig. 4A, Lane 6). In sharp contrast, neither complex A nor complex B was formed after 5-min incubation in extracts treated with NB-506 concentrations varying from 25 to 100 ␮M (Fig. 4A, Lanes 8, 14, and 19). At low concentrations of the drug, only a low level of complex B is still formed after 30-min incubation (Lane 10), whereas at higher concentrations, the formation of both A and B complexes was completely abolished at any time points (Fig. 4A, Lanes 14 –17 and Lanes 19 –22). The complexes assembled in the presence of the drug were similar to the complexes formed in the absence of ATP (compare Lanes 2, 7, 13 and 18 with Lanes 14 –17 and 19 –22) and persisted even after long incubation times (Lanes 17 and 22), indicating that the inhibitory effect does not reflect a reduction in the kinetics of spliceosome assembly. Thus, NB-506 inhibits an early step in the spliceosome assembly pathway. If correct, this conclusion predicts that the drug will not have any effect in splicing if the spliceosome is allowed to form. Results shown in Fig. 4B are consistent with this prediction. When NB-506 was added to the splicing assay after a 15-min (Lanes 8, 11, and 14) or 30-min (Lanes 9, 12, and 15) preincubation, splicing products were still observed like in the control reaction (Lane 4). Parallel analysis confirmed that, at 15-min time point, spliceosome assembly (data not shown) and the first catalytic step but not the second catalytic step of splicing (Lane 2) have occurred. In summary, the data show that NB-506, a potent inhibitor of

6879


Fig. 4. The effect of NB-506 on spliceosome assembly. A, splicing complexes formed on 32P-labeled ␤-3S pre-mRNA at the indicated times in HeLa extracts alone (Lanes 3– 6) or containing either 25 ␮M (Lanes 8 –11), 50 ␮M (Lanes 14 –17), or 100 ␮M (Lanes 19 –22) of NB-506 were separated on a 4% nondenaturing polyacrylamide gel as described in “Materials and Methods.” Positions of splicing complexes H, A, and B are indicated. Lanes 2, 7, 13, and 18 are splicing reactions without ATP and creatine phosphate. Lanes 1 and 12 correspond to RNA without extracts. B, the kinetics of appearance of splicing intermediates and products under spliceosome assembly conditions in the absence of the drug are shown (Lanes 2– 4). The incubation times are indicated on the top of the panel. Splicing reactions incubated for 60 min to which DMSO (Lanes 4 – 6), 25 ␮M (Lanes 7–9), 50 ␮M (Lanes 10 –12), or 100 ␮M (Lanes 13–15) of NB-506 were added at 0 time point (Lanes 4, 7, 10, and 13), after 15 min (Lanes 5, 8, 11, and 14) or 30 min (Lane 6, 9, 12, and 15) of preincubation. The preincubation times without drug are indicated on the top of the panel. Splicing products (depicted on the left of the panel) were analyzed on a 7% denaturing polyacrylamide gel and revealed by autoradiography.

topoisomerase I, blocks pre-mRNA splicing in vitro by specifically preventing spliceosome assembly. NB-506 Differentially Affects The Phosphorylation Status of SR Proteins in P388 Leukemia Cells and P388CPT5 Cells Resistant to CPT. To investigate the in vivo effects of NB-506, P388 and P388CPT5 murine leukemia cell lines sensitive and resistant to CPT, respectively, were used. NB-506 proved to have a very toxic effect on P388 cells (30), whereas it exhibited little or no effect on the growth of P388CPT5 cells. Similar results (30) were found for other indolocarbazole derivatives of rebeccamycin. P388CPT5 cells showed marked cross-resistance to these derivatives with relative resistance indexes included between 8.5 and 58, to be compared with a 87.5 relative resistance index for CPT, the drug used to select the resistant cells (31). A recent report (30) indicated that different CPT-resistant cell lines present partial and variable resistance to NB-506 and suggested that NB-506 may have additional targets besides topoisomerase I. A basis for such differences in NB-506 resistance between these

cell lines and P388CPT5 may be related to the location of the different point mutations in the topoisomerase I sequences. Sequencing of the entire topoisomerase I cDNA with specific primers revealed that topoisomerase I polypeptide from P388CPT5 cells harbors two point mutations in highly conserved residues (Fig. 5A): a mutation in the core domain changing codon 361 from a GGG Gly-codon to a GTG Val-codon and another mutation in the linker domain changing codon 709 from a GAT Asp-codon to a TAT Tyr-codon. Although the mechanism by which these two mutations confer drug resistance was not addressed here, it is important to recall that residues in both the core and the linker region are important for CPT action (Fig. 5B and Refs. 16, 32). Topoisomerase I extracted from P388CPT5 cells was found to be biochemically resistant to the inhibitory effect of CPT (25).5 To address the effect of NB506 on the kinase activity of the mutant topoisomerase I, topoisomerase I proteins were isolated (18) from P388 and P388CPT5 cells and their abilities to phosphorylate SF2/ ASF were examined. Surprisingly, the specific activity of the kinase was five times higher (10 nmol of phosphate transferred min⫺1, mg⫺1) in P388CPT5 cells as compared with P388 cells (2 nmol of phosphate transferred min⫺1, mg⫺1), whereas both enzymes were equally sensitive to inhibition by NB506 (data not shown). This finding suggested that the mutant topoisomerase I from P388CPT might adopt an altered conformation that increases the kinase activity of the enzyme and reduces its capacity to form cleavable complexes in the presence of either CPT or NB506. If the kinase activity of topoisomerase I is brought into play to phosphorylate SR proteins in vivo, it can be expected that NB-506 be more active at inhibiting SR protein phosphorylation in P388 than in P388CPT5 cells. We attempted to test this possibility by comparing the levels of phosphorylation of SR proteins purified from NB-506treated cells versus those from untreated cells. To evaluate the level of SR protein phosphorylation, they were metabolically labeled in vivo with [32P]inorganic phosphate and isolated as described (6). Comparison of 32P-labeled SR proteins purified from P388 and P388CPT5 untreated cells and analyzed in SDS PAGE showed no discernible difference between protein patterns (Fig. 6A; compare Lanes 1 and 4). In agreement with previous results (6, 18), five proteins with sizes consistent with those of SR proteins were detected by autoradiography (Fig. 6A, right panel, Lanes 1 and 4) and Coomassie Blue staining (Fig. 6A, left panel, Lanes 1– 6). They reacted with mab 104 antibody (data not shown) and correspond, therefore, to SRp20, SRp30, SRp40, SRp55, and SRp75 characterized previously (6). After 6 h of drug treatment, P388 cells showed a different pattern. Although SRp30 and SRp55 bands are still detected (Fig. 6A, right panel, Lanes 2 and 3), they have lower intensity than those from untreated cells (compare with Lanes 5 and 6). Also, at higher NB-506 concentrations, the labeling of SRp20 and SRp40 was abolished and that of SRp75 was highly reduced (Lane 3). In contrast, only slight changes, including a reduction in the intensity of the SRp40 band, were observed in the pattern of SR proteins from P388CPT5-treated cells (Lanes 5 and 6), implying that NB-506 affects only weakly the overall phosphorylation levels of SR proteins in these cells. By separating the 32P-labeled SR proteins from untreated (Lanes 1 and 4) and treated (Lanes 3 and 6) samples on two-dimensional gels, it became apparent that each SR protein was made of several isoelectric variants that can be detected by autoradiography (Fig. 6B, left and right top panels). Although untreated samples have similar profiles (Fig. 6B, left and right top panels), NB-506 treatment enhanced radioactivity contained in the less phosphorylated variants of SRp75, SRp55, and SRp30 from P388

6880

5

J. F. Riou, unpublished data.


Fig. 5. A, schematic representation of the topoisomerase I. The different domains as well as the region corresponding to the SF2/ASF-binding domain (20) are depicted. Mutations identified in the murine P388CPT5 cells are shown (Mu), and their correspondence within the human topoisomerase I is indicated (Hu). B, comparison with different mutations identified in topoisomerase I mutants expressed in other CPT-resistant cell lines (16).

cells (Fig. 6B, left bottom panel) compared with the same proteins from P388CPT5 cells (right bottom panel). With the exception of SRp40, the detection of which in P388 cells was completely abolished, inhibition of the other SR protein phosphorylation by NB-506 was not complete, implying that other protein kinase(s) not affected by NB-506 are also active in vivo. In keeping with this finding, SR proteins were less phosphorylated in HeLa cells treated with topotecan, a tumor-active analogue of CPT used in cancer chemotherapy (18). Because topoisomerase I appears to be a privileged cellular target of these drugs, it can be assumed that the observed hypophosphorylation of SR proteins in P388-treated cells is a consequence of inactivation of the kinase activity of topoisomerase I. Cellular Effects Associated with NB-506 Treatment. In living cells, spliceosome assembly and splicing are known to occur during ongoing transcription elongation, and inefficient splicing impairs release of RNA from the site of transcription (33). To evidence the effect of NB-506 on pre-mRNA splicing in P388 and P388CPT5 cells, fluorescence in situ hybridization was used to detect the distribution of polyadenylated RNA (34). Treated and untreated cells were hybridized under nondenaturing conditions with fluorescent oligo(dT) (Fig. 7). As expected, polyadenylated RNA species were detected both in the cytoplasm and the nucleus of P388 (panel b) and P388CPT5 (panel j) untreated cells. After 24 h of NB-506 treatment, the cytoplasm of P388 cells was devoid of any hybridization signal (panel f), whereas the nuclear signal was still detected. This result is consistent with the finding that unprocessed RNA species fail to be exported from the nucleus. In contrast, both cytoplasmic and nuclear staining were still visible in the P388CPT5-treated cells (panel n), indicating that a significant proportion of mRNAs in these cells was still correctly processed and exported. To determine whether NB-506 induces changes in the splicing profile and/or mRNA levels of proteins that play key roles in apoptosis and tumor progression, RNA recovered from treated and untreated cells were analyzed by RT-PCR using oligo(dT) for priming the first-strand DNA synthesis and specific primers to either Bcl-X or CD 44 mRNAs (Fig. 8). Bcl-X is a member of the bcl-2 gene family that can either promote or prevent apoptosis and exists in several isoforms generated by alternative splicing (Ref. 35 and Fig. 8, Bcl-X, left panel). Only the large isoform, Bcl-XL, which protects cells against apoptosis, but not short isoform Bcl-Xs was detected in both P388 and P388CPT5 cells (Fig. 8, Bcl-X, right panel, Lanes 1 and 3).

Fig. 6. NB-506 influences the level of SR protein phosphorylation in vivo. A, one-fourth of the total amount of metabolically 32P-labeled SR proteins isolated from P388 cells (Lanes 1–3) or P388CPT5 cells (Lanes 4 – 6) untreated (Lanes 1 and 4) or treated with 0.5 ␮M (Lanes 2 and 5) and 1 ␮M (Lanes 3 and 6) of NB-506 was analyzed in 10% SDS-PAGE and detected by autoradiography (right panel) or by Coomassie Blue staining (left panel). B, three times the amount of total proteins contained in the samples described in A, Lanes 1 (top left panel), 3 (bottom left panel), 3 (top right panel), and 6 (bottom right panel) were analyzed by two-dimensional gel electrophoresis using Immobiline DryStrip (Amersham Pharmacia Biotech). 32P-labeled SR protein variants were detected by autoradiography. The electrophoretic mobility of each variant from each SR protein was determined in comparison with mobilities of cold hnRNP proteins that were analyzed in the same gel and stained with Coomassie Blue.

6881


Fig. 7. NB-506 affects cellular localization of polyadenylated RNA in P388 (panels e-h) but not in P388CPT5 (panels m-p) cells. In situ hybridization of fluorescently labeled oligo(dT) probe to endogenous RNAs was carried out as described in “Materials and Methods” with untreated (panel b) and treated P388 (panel f) or untreated (panel j) and treated (panel n) P388CPT5. Polyadenylated RNA species were detected with the red fluorophore CY3. Nuclei of the same untreated (panel c) and treated (panel g) P388 cells or untreated (panel k) and treated (panel o) P388CPT5 cells were counterstained with 4⬘,6-diamidino-2-phenylindole. The Normarski interference-contrast of the same cells are shown (panel a) for untreated and (panel e) for treated P388 cells and (panel i) for untreated and (panel m) for treated P388CPT5 cells.

In both cases, treatment with NB-506 did not change the splicing profile of Bcl-X gene (Fig. 8, Bcl-X, right panel, Lanes 2 and 4) but induced a large decrease of Bcl-XL mRNA levels in P388 cells (Lane 2). By contrast, the abundance of Bcl-XL mRNA was not significantly affected in P388CPT5-treated cells (Fig. 8, Bcl-X, right panel, Lane 4). This finding is in complete agreement with the observation that, after drug treatment, P388 cells were more sensitive to apoptosis than P388CPT5 cells (data not shown). In P388-treated cells, NB-506 also led to a decrease in mRNA levels of three related isoforms of the CD44 cell surface glycoprotein family (Fig. 8, CD44, right panel, Lane 2) that has been implicated in extracellular matrix attachment and tumor metastasis (36). Not all of the isoforms that derive from the differential splicing of a single CD44 gene (Fig. 8, CD44, left panel) were affected to the same extent by NB-506 treatment. The larger isoform carrying several alternative exons (I1) disappeared after treatment of P388 cells with NB-506, whereas the two smaller isoforms (I2 and I3) were readily detected (compare Lanes 1 and 2). Both untreated and treated P388CPT cells showed identical splicing patterns, and the various isoforms accumulated to similar levels (Fig. 8, CD44, right panel; compare Lanes 3 and 4), implying that NB-506 did not affect the overall CD 44 gene expression in P388CPT5 cells. This finding strongly suggests that NB-506-mediated inhibition of topoisomerase I

may be responsible for alterations of processing events that are crucial for the accumulation of specific splice variants. To test this hypothesis more directly, we considered other genes that are regulated by alternative splicing. The SC35 gene encodes the SR protein SC35, which autoregulates its expression by promoting splicing events that destabilize its mRNAs (37). To detect the various splicing isoforms of SC35, the 3⬘ untranslated region of SC35 transcripts, which is responsible for SC35 mRNAs destabilization, was amplified by RT-PCR using specific primers (Fig. 8, SC35, left panel). In agreement with previous results, PCR products corresponding to the major (2.0-kb mRNA) and the minor (1.6-kb mRNA) transcripts were detected in both P388 and P388CPT5 cells (Fig. 8, SC35, right panel, Lanes 1 and 3, respectively). Interestingly, after NB-506 treatment, a novel PCR product corresponding to the 1.7-kb mRNA was amplified from P388 but not from P388CPT5 cells, implying that inhibition of topoisomerase I led to changes in the splicing pattern of the SC35 primary transcripts. Treatment of P388 with NB-506 also affected the splicing pattern of the Clk/Sty mRNA, encoding a member of the dual specificity kinase family that has the ability to autophosphorylate on serine, threonine, and tyrosine residues (38). Alternative splicing of the primary sty transcripts (Fig. 8, Clk/Sty, left panel) generates mRNAs encoding full-length catalyti-

6882


Fig. 8. NB-506 affects alternative splicing events in P388 but not in P388CPT5 cells. Alternative splicing patterns reported for Bcl-X (35), CD44 (36), SC35 (37), and Clk/Sty (38) primary transcripts are depicted on the left. Because of multiple CD44 alternative splice variants, splicing patterns are depicted by ----. Coding and alternative exons are indicated by o and 0, respectively. Alternative intronic sequences within the SC35 3⬘ untranslated region are shown (u). The position of primers and expected sizes (in nucleotides) of the RT-PCR products corresponding to the different mRNA isoforms are indicated. Right, representative agarose gels of RT-PCR products corresponding to Bcl-X, CD44, SC35, and Clk/Sty splicing variants expressed in P388 (Lanes 1 and 2) and P388CPT5 (Lanes 3 and 4) cells before (Lanes 1 and 3) and after a 24-h exposure to 1.0 ␮M NB-506 (Lanes 2 and 4). The position of size markers (in bp) is shown on the right, and the various splicing isoforms are named on the left of each panel.

cally active enzyme (Sty) and truncated, kinase-deficient polypeptides (tSty). Both Sty and tSty splice variants can be detected in P388 and P388CPT5 cells (Fig. 8, Clk/Sty, right panel, Lanes 1 and 3), but Sty was the predominant transcript. Although Sty expression did not change after NB-506 treatment of P388, the production of tSty relative to that of Sty was increased (Lane 2). Again, NB-506 fails to induce changes in the splicing pattern of the sty gene in P388CPT5 (Lane 4). Conclusion. In the present studies, we have established that the antitumor drug NB-506 causes a hypophosphorylation of SR proteinsplicing factors, both in NEs and in living cells. NB-506 blocks spliceosome assembly, consistent with previous observations showing that phosphorylation of SR proteins is required for spliceosome assembly, whereas dephosphorylation is critical for the splicing reaction to occur (12, 15). Although the antitumor activity of NB-506 has been attributed to its effect on topoisomerase I, this is the first report where NB-506 is shown to interfere with the kinase activity of the enzyme. Failure of NB-506 to inhibit the SR phosphorylating activity of S100 extracts, SRPK1, cdc2 kinase (the present study), and protein kinases affected by indolocarbazoles of the staurosporine family such as PKA, PKC, and PTK (23) makes the kinase activity of topoisomerase I a likely candidate inhibited by the drug in vivo. Given that SR proteins are known to influence alternative splicing decisions in a dosedependent manner (1, 6) and that phosphorylation is required for their recruitment from storage/assembly sites, “the speckles,” to transcription sites (39), the inhibition of SR protein phosphorylation by NB506 was expected to lead to pronounced changes in alternative splicing. Consistently, drug treatment altered the splicing pattern of several genes, among which Bcl-X and CD 44 gene products are known to affect apoptosis and tumor progression. Such alterations of regulated splicing may well be a key step accounting for the remarkable antineoplastic activities exhibited by the drug both in cell culture systems and in xenograft models of human tumor (16, 21). ACKNOWLEDGMENTS We thank Marcel Dore´ e (CRBM, Montpellier, France) for purified cdc2 kinase, James Manley (Columbia University, New York, NY) for SF2/ASF

cDNA expression plasmid, and Thomas Giannakouros (Thessaloniki, Greece) for purified SRPK1. We also thank Marie-Claire Sant (IGM, Montpellier, France) for excellent technical assistance.

REFERENCES 1. Wang, J., and Manley, J. L. Regulation of pre-mRNA splicing in metazoa. Curr. Opin. Genet. Dev., 7: 205–211, 1997. 2. Jiang, Z. H., and Wu, J. Y. Alternative splicing and programmed cell death. Proc. Soc. Exp. Biol. Med., 220: 64 –72, 1999. 3. Cooper, T. A., and Mattox, W. The regulation of splice-site selection, and its role in human disease. Am. J. Hum. Genet., 61: 259 –266, 1997. 4. Staley, J. P., and Guthrie, C. Mechanical devices of the spliceosome: motors, clocks, springs, and things. Cell, 92: 315–326, 1998. 5. Fu, X. D. The superfamily of arginine/serine-rich splicing factors. RNA (N Y), 1: 663– 680, 1995. 6. Zahler, A. M., Lane, W. S., Stolk, J. A., and Roth, M. B. SR proteins: a conserved family of pre-mRNA splicing factors. Genes Dev., 6: 837– 847, 1992. 7. Reed, R. Initial splice-site recognition and pairing during pre-mRNA splicing. Curr. Opin. Genet. Dev., 6: 215–220, 1996. 8. Wu, J. Y., and Maniatis, T. Specific interactions between proteins implicated in splice site selection and regulated alternative splicing. Cell, 75: 1061–1070, 1993. 9. Yeakley, J. M., Tronchere, H., Olesen, J., Dyck, J. A., Wang, H. Y., and Fu, X. D. Phosphorylation regulates in vivo interaction and molecular targeting of serine/ arginine-rich pre-mRNA splicing factors. J. Cell Biol., 145: 447– 455, 1999. 10. Xiao, S. H., and Manley, J. L. Phosphorylation of the ASF/SF2 RS domain affects both protein-protein and protein-RNA interactions and is necessary for splicing. Genes Dev., 11: 334 –344, 1997. 11. Xiao, S. H., and Manley, J. L. Phosphorylation-dephosphorylation differentially affects activities of splicing factor ASF/SF2. EMBO J., 17: 6359 – 6367, 1998. 12. Tazi, J., Kornstadt, U., Rossi, F., Jeanteur, P., Cathala, G., Brunel, C., and Luhrmann, R. Thiophosphorylation of U1–70K protein inhibits pre-mRNA splicing. Nature (Lond.), 363: 283–286, 1993. 13. Prasad, J., Colwill, K., Pawson, T., and Manley, J. L. The protein kinase Clk/Sty directly modulates SR protein activity: both hyper- and hypophosphorylation inhibit splicing. Mol. Cell. Biol., 19: 6991–7000, 1999. 14. Cao, W., Jamison, S. F., and Garcia-Blanco, M. A. Both phosphorylation and dephosphorylation of ASF/SF2 are required for pre-mRNA splicing in vitro. RNA (N Y), 3: 1456 –1467, 1997. 15. Tazi, J., Rossi, F., Labourier, E., Gallouzi, I., Brunel, C., and Antoine, E. DNA topoisomerase I: customs officer at the border between DNA and RNA worlds? J. Mol. Med., 75: 786 – 800, 1997. 16. Pommier, Y., Pourquier, P., Fan, Y., and Strumberg, D. Mechanism of action of eukaryotic DNA topoisomerase I and drugs targeted to the enzyme. Biochim. Biophys. Acta, 1400: 83–105, 1998. 17. Wang, J. C. DNA topoisomerases. Annu. Rev. Biochem., 65: 635– 692, 1996. 18. Rossi, F., Labourier, E., Forne, T., Divita, G., Derancourt, J., Riou, J. F., Antoine, E., Cathala, G., Brunel, C., and Tazi, J. Specific phosphorylation of SR proteins by mammalian DNA topoisomerase I. Nature (Lond.), 381: 80 – 82, 1996.

6883


19. Rossi, F., Labourier, E., Gallouzi, I. E., Derancourt, J., Allemand, E., Divita, G., and Tazi, J. The C-terminal domain but not the tyrosine 723 of human DNA topoisomerase I active site contributes to kinase activity. Nucleic Acids Res., 26: 2963–2970, 1998. 20. Labourier, E., Rossi, F., Gallouzi, I. E., Allemand, E., Divita, G., and Tazi, J. Interaction between the N-terminal domain of human DNA topoisomerase I and the arginine-serine domain of its substrate determines phosphorylation of SF2/ASF splicing factor. Nucleic Acids Res., 26: 2955–2962, 1998. 21. Arakawa, H., Iguchi, T., Morita, M., Yoshinari, T., Kojiri, K., Suda, H., Okura, A., and Nishimura, S. Novel indolocarbazole compound 6-N-formylamino-12,13-dihydro-1,11-dihydroxy-13-(␤-D-glucopyranosyl)-5H-indolo[2,3-a]pyrrolo-[3,4-c]carbazole-5,7(6H)-dione (NB-506): its potent antitumor activities in mice. Cancer Res., 55: 1316 –1320, 1995. 22. Bailly, C., Dassonneville, L., Colson, P., Houssier, C., Fukasawa, K., Nishimura, S., and Yoshinari, T. Intercalation into DNA is not required for inhibition of topoisomerase I by indolocarbazole antitumor agents. Cancer Res., 59: 2853–2860, 1999. 23. Anizon, F., Moreau, P., Sancelme, M., Voldoire, A., Prudhomme, M., Ollier, M., Severe, D., Riou, J. F., Bailly, C., Fabbro, D., Meyer, T., and Aubertin, A. M. Syntheses, biochemical and biological evaluation of staurosporine analogues from the microbial metabolite rebeccamycin. Bioorg. Med. Chem., 6: 1597–1604, 1998. 24. Ohkubo, M., Nishimura, T., Honma, T., Nishimura, I., Ito, S., Yoshinari, T., Suda, H. A., Morishima, H., and Nishimura, S. Synthesis and biological activities of NB-506 analogues: effects of the positions of two hydroxyl groups at the indole rings. Bioorg. Med. Chem. Lett., 9: 3307–3312, 1999. 25. Madelaine, I., Prost, S., Naudin, A., Riou, G., Lavelle, F., and Riou, J. F. Sequential modifications of topoisomerase I activity in a camptothecin-resistant cell line established by progressive adaptation. Biochem. Pharmacol., 45: 339 –348, 1993. 26. Chartrand, P., Bertrand, E., Singer, R. H., and Long, R. M. Sensitive and highresolution detection of RNA in situ. Methods Enzymol., 318: 493–506, 2000. 27. Tazi, J., Alibert, C., Temsamani, J., Reveillaud, I., Cathala, G., Brunel, C., and Jeanteur, P. A protein that specifically recognizes the 3⬘ splice site of mammalian pre-mRNA introns is associated with a small nuclear ribonucleoprotein. Cell, 47: 755–766, 1986. 28. Labourier, E., Allemand, E., Brand, S., Fostier, M., Tazi, J., and Bourbon, H. M. Recognition of exonic splicing enhancer sequences by the Drosophila splicing repressor RSF1. Nucleic Acids Res., 27: 2377–2386, 1999.

29. Konarska, M. M., and Sharp, P. A. Interactions between small nuclear ribonucleoprotein particles in formation of spliceosomes. Cell, 49: 763–774, 1987. 30. Urasaki, Y., Laco, G., Takebayashi, Y., Bailly, C., Kohlhagen, G., and Pommier, Y. Use of camptothecin-resistant mammalian cell lines to evaluate the role of topoisomerase I in the antiproliferative activity of the indolocarbazole, NB-506, and its topoisomerase I binding site. Cancer Res., 61: 504 –508, 2001. 31. Labourier, E., Riou, J. F., Prudhomme, M., Carrasco, C., Bailly, C., and Tazi, J. Poisoning of topoisomerase I by an antitumor indolocarbazole drug: stabilization of topoisomerase I-DNA covalent complexes and specific inhibition of the protein kinase activity. Cancer Res., 59: 52–55, 1999. 32. Reid, R. J., Benedetti, P., and Bjornsti, M. A. Yeast as a model organism for studying the actions of DNA topoisomerase-targeted drugs. Biochim. Biophys. Acta, 1400: 289 –300, 1998. 33. Custodio, N., Carmo-Fonseca, M., Geraghty, F., Pereira, H. S., Grosveld, F., and Antoniou, M. Inefficient processing impairs release of RNA from the site of transcription. EMBO J., 18: 2855–2866, 1999. 34. Taneja, K. L., Lifshitz, L. M., Fay, F. S., and Singer, R. H. Poly(A) RNA codistribution with microfilaments: evaluation by in situ hybridization and quantitative digital imaging microscopy. J. Cell Biol., 119: 1245–1260, 1992. 35. Boise, L. H., Gonzalez-Garcia, M., Postema, C. E., Ding, L., Lindsten, T., Turka, L. A., Mao, X., Nunez, G., and Thompson, C. B. bcl-x, a bcl-2-related gene that functions as a dominant regulator of apoptotic cell death. Cell, 74: 597– 608, 1993. 36. Screaton, G. R., Bell, M. V., Bell, J. I., and Jackson, D. G. The identification of a new alternative exon with highly restricted tissue expression in transcripts encoding the mouse Pgp-1 (CD44) homing receptor. Comparison of all 10 variable exons between mouse, human, and rat. J. Biol. Chem., 268: 12235–12238, 1993. 37. Sureau, A., Gattoni, R., Dooghe, Y., Stevenin, J., and Soret, J. SC35 autoregulates its expression by promoting splicing events that destabilize its mRNAs. EMBO J., 20: 1785–1796, 2001. 38. Duncan, P. I., Howell, B. W., Marius, R. M., Drmanic, S., Douville, E. M., and Bell, J. C. Alternative splicing of STY, a nuclear dual specificity kinase. J. Biol. Chem., 270: 21524 –21531, 1995. 39. Misteli, T., Caceres, J. F., Clement, J. Q., Krainer, A. R., Wilkinson, M. F., and Spector, D. L. Serine phosphorylation of SR proteins is required for their recruitment to sites of transcription in vivo. J. Cell Biol., 143: 297–307, 1998.

6884