The ethylene-inducible PK12 kinase mediates the phosphorylation of SR splicing factors

Transcription

1 The Plant Journal (2000) 21(1), 91±96 SHORT COMMUNICATION The ethylene-inducible PK12 kinase mediates the phosphorylation of SR splicing factors Sigal Savaldi-Goldstein, Guido Sessa ² and Robert Fluhr* Weizmann Institute of Science, PO Box 26, Rehovot, Israel Received 7 October 1999; revised 24 November 1999; accepted 30 November *For correspondence (fax ; Robert.Fluhr@weizmann.ac.il). ² Present address: Boyce Thompson Institute for Plant Research, Cornell University, Tower Road, Ithaca, NY , USA. Summary The tobacco PK12 is induced by the plant hormone ethylene and is a member of the LAMMER family of protein kinases. Members of this family contain in their C-terminus a unique `EHLAMMERI/VLGPLP' motif of unknown function, and are related to cyclin- and mitogen-activated protein (MAP)-dependent kinases. The animal members of this class play a role in differentiation. They phosphorylate and physically interact with serine/arginine-rich (SR) splicing factors in vivo to alter their activity and the splicing of target mrnas. SR proteins have been recently described in plants. The capability of PK12 LAMMER kinase to bind and phosphorylate SR proteins was tested in vitro by kinase and binding assays. The tobacco PK12 phosphorylated both animal and plant SR proteins and speci cally interacted with the plant splicing factor atsrp34/sr1. In addition, by site-directed mutagenesis, the LAMMER motif was found to be required for PK12 kinase activity but was not necessary for substrate binding. Consistent with a role in phosphorylation of splicing factors, PK12 was found to localize to the nucleus when transiently over-expressed in suspension cells. Introduction Protein kinases of the LAMMER family are present in yeast, plants and animals and show sequence similarity to cyclindependent kinases and mitogen-activated protein (MAP) kinases. Nevertheless, they form a distinct group due to the presence in their amino acid sequence of a highly conserved motif, `EHLAMMERI/VLGPLP', in subdomain X of the kinase catalytic domain (Yun et al., 1994). The function of the LAMMER motif is unknown. By analogy to three-dimensional structures available for other kinases, the LAMMER motif is predicted to be positioned at the a-helix below the substrate-binding cleft, which suggests a possible role for this sequence in substrate recognition (Yun et al., 1994). A common feature of LAMMER protein kinases is their ability to autophosphorylate with dual speci city (Duncan et al., 1995; Lee et al., 1996; Sessa et al., 1996). In addition, the analysis of mammalian Clk/Sty and Drosophila DOA, reveals that they are involved in phosphorylation and regulation of serine/arginine (SR) splicing factors. In fact, all Clk/Sty homologues, CLK/STY1±4, phosphorylate SR splicing factors in vivo and affect their distribution in the nucleus (Colwill et al., 1996; Duncan et al., 1997). Moreover, the Drosophila DOA, which plays a role in cell structure maintenance, differentiation and sex determination (Yun et al., 1994), phosphorylates and interacts with SR splicing factors and related regulators, such as RBP1, TRA and TRA2 (Du et al., 1998). For the animal LAMMER Clk/Sty, the association with SR proteins has been shown to occur through its N-terminal serine/arginine-rich (RS) domain (Colwill et al., 1996). SR splicing factors play a key role in both constitutive and alternative splicing (Manley and Tacke, 1996). They include about 10 proteins with a broad size range of 20± 75 kda. Each of them possesses one or two ribonucleoprotein-type RNA binding domains at the N-terminus, and a region of repeated arginine/serine dipeptides, known as the RS domain, at the C-terminus. The latter is subjected to regulation by phosphorylation and dephosphorylation which affects its protein±protein and protein±rna interactions (Xiao and Manley, 1997). Among the protein ã 2000 Blackwell Science Ltd 91

2 92 Sigal Savaldi-Goldstein et al. Figure 1. PK12 phosphorylates mammalian and plant SR splicing factors. (a) Western blot analysis of animal SR splicing factors. SR splicing factors extracted from HeLa cells were treated with alkaline phosphatase (AP; 5 U) in the absence or presence of phosphatase inhibitors (lanes 2 and 3, respectively) or without AP (lane 1). The treated SR proteins were then fractionated and subjected to Western blot analysis using mab 104 as primary antibody. (b) In vitro phosphorylation of puri ed SR splicing factors from Hela cells by PK12. SR splicing factors were treated as described in (a) and were then used as substrates for in vitro kinase assay in the presence or absence of PK12 (lanes 2±4 and 5±7, respectively). Autophosphorylation of PK12 is shown in lane 1. (c) Phosphorylation of recombinant atsrp34/sr1 and MBP by PK12. Each substrate (1 mg) was incubated in the presence of PK12 and fractionated on SDS± PAGE. (d) Stained and Western blot analysis of recombinant atsrp34/sr1 and MBP upon phosphorylation by PK12. Left panel, recombinant plant SR and MBP substrate were treated as indicated and the fractionated proteins were stained by Coomassie blue. Right panel, recombinant plant SR and MBP substrate were treated as indicated and the fractionated proteins were reacted with mab 104. kinases able to phosphorylate SR splicing factors, in addition to Clk/Sty, are the cell cycle-regulated SRPK1 (Gui et al., 1994) and SRPK2 (Wang et al., 1998). Members of the LAMMER family have been isolated from plants (Bender and Fink, 1994; Sessa et al., 1996). The Arabidopsis AFC1 was isolated for its ability to complement a yeast MAP kinase mutant (Bender and Fink, 1994). The tobacco homologue PK12 was identi ed by differential screening of ethylene-induced transcripts (Sessa et al., 1996). Exogenous application of ethylene to tobacco leaves was shown to induce PK12 transcript accumulation as well as PK12 kinase activity. A characteristic that distinguishes plant LAMMER kinases from mammalian components of the same family is the absence of an RS domain that in Clk/Sty allows interaction with SR splicing factors. Very little is known about splicing mechanisms in plants. Nevertheless, similarities as well as distinct characteristics can be attributed to splicing in plants compared to splicing in animals (Brown and Simpson, 1998). Only recently have SR splicing factors been identi ed in plants. They are less than 55 kda in size and can complement in vitro splicing activity in HeLa cell extracts (Lopato et al., 1996a). Moreover, the atsrp34/sr1 and atsrp30 genes isolated from Arabidopsis encode proteins with high similarity to the mammalian SR factors ASF/SF2 (Lazar et al., 1995; Lopato et al., 1999). Interestingly, their transcripts appear to be alternatively spliced in different organs and developmental stages. Over-expression of atsrp30 changes the alternative splicing pattern of its cognate transcript and of several other transcripts. Here we examine functional characteristics of the tobacco PK12 LAMMER kinase related to phosphorylation and binding of SR splicing factors. Despite the absence of RS domains in its sequence, PK12 showed functional characteristics common to animal LAMMER kinases. Consistent with a possible function in splicing regulation,

3 Phosphorylation of SR splicing factors 93 phosphorylation of SRp30±35 (identi ed in mammals as SC35 and ASF/SF2) and SRp20 are speci cally increased, suggesting that they are preferred substrates (Figure 1b, compare lanes 2 and 3). Few SR splicing factors have been isolated in plants. One of them, the Arabidopsis atsrp34/sr1, is highly homologous to ASF/SF2 (Lazar et al., 1995; Lopato et al., 1999). We examined recombinant atsrp34/sr1 as a potential substrate for PK12. atsrp34/sr1 was phosphorylated by PK12 with a similar ef ciency to myelin basic protein (MBP; Figure 1c). In contrast to the MBP substrate, phosphorylation of atsrp34/sr1 by PK12 affected both its migration in SDS±PAGE fractionation and also enabled its detection by mab 104 (Figure 1d). PK12 physically interacts with the Arabidopsis splicing factor atsrp34/sr1 Figure 2. PK12 and PK12 RAQ bind to atsrp34/sr1. In vitro 35 S-methionine-labelled PK12 and PK12 RAQ were used as probes for Western blots containing fractionated atsrp34/sr1 and MBP (1 mg each). PK12 was found to localize to the nucleus in tobacco cells. In addition, it phosphorylated both animal and plant SR splicing factors and speci cally interacted with the Arabidopsis splicing factor atsrp34/sr1. Furthermore, we provide the rst evidence that the LAMMER signature is essential for kinase activity but is not required for substrate recognition. Results PK12 phosphorylates SR splicing factors Animal components of the LAMMER family have been shown to phosphorylate SR splicing factors. We therefore examined the ability of tobacco PK12 to phosphorylate these proteins. The complete repertoire of SR splicing factors was puri ed from HeLa cells and were identi ed by mab 104, a monoclonal antibody that speci cally recognizes phosphorylated epitopes in SR splicing factors (Roth et al., 1990). The immunoblot in Figure 1(a) (lane 1) shows that phosphorylated epitopes are present. PK12-dependent phosphorylation of the splicing factors was readily detected (Figure 1b; compare lanes 2±4 with lanes 5±7). In order to examine the complete potential of PK12 to phosphorylate SR factors, the extract was treated with phosphatase to reduce prephosphorylated sites. This resulted in an altered migration of SR proteins as re ected in the pattern of phosphatase-dependent mab 104 recognition (compare lanes 1±3 in Figure 1a). In this case, PK12-dependent The substrate speci city of PK12 for SRp34/SR1-type SR proteins raised the possibility that PK12 could bind atsrp34/sr1. Binding of Clk/Sty to SR substrate was shown by two-hybrid analysis (Colwill et al., 1996). We tested this possibility by a far-western technique, using as a probe a labelled PK12 translated in vitro in the presence of 35 S-methionine. As shown in Figure 2 (left panel), in vitro translated PK12 ef ciently bound to atsrp34/sr1 but not MBP. The predicted position below the binding cleft of the LAMMER motif raised the question of whether it was involved in substrate recognition. To address this question, site-directed mutagenesis was performed. The residues MME within the motif were replaced by RAQ, respectively, generating the PK12 RAQ mutant protein. In vitro translated PK12 RAQ showed similar binding capability to atsrp34/sr1 as the wild-type kinase, suggesting that the LAMMER motif is not essential for the physical interaction of PK12 with its substrate atsrp34/sr1 (Figure 2, right panel). The LAMMER motif is required for PK12 kinase activity We next asked whether the LAMMER motif is an essential part of the kinase catalytic domain by examining PK12 RAQ phosphorylation activity. As a negative control for phosphorylation, another independent mutant was created by substituting the invariant lysine present in subdomain II by arginine, resulting in the PK12 K125R mutant protein. Mutations in this position are known to abolish the activity of most kinases. The recombinant wild-type and mutant proteins were readily recognized by antibodies to PK12 (Figure 3a, left panel). However, PK12 RAQ, as opposed to PK12 K125R, was not recognized by antibodies that are speci c for the LAMMER peptide (Figure 3a, right panel). The in vitro kinase activity was examined directly using MBP as substrate. PK12 K125R retains 20% of wild-type

4 94 Sigal Savaldi-Goldstein et al. reported to be localized to the nucleus (Colwill et al., 1996). Examination of the PK12 amino acid sequence revealed a stretch of basic residues upstream of the kinase catalytic domain, which is reminiscent of a nuclear localization signal (NLS). To test PK12 subcellular localization, its coding region was fused to the C- terminal region of a b-glucoronidase (GUS) reporter gene under the control of the 35S promoter. The resultant fusion construct 35S:GUS-PK12 and a 35S:GUS control construct were transformed into tobacco suspension cells by particle-gun bombardment. Suspension cells were analysed for GUS expression by a histochemical assay 48 h after bombardment. Under these conditions, GUS expression was detected in the nucleus of cells transformed with 35S:GUS-PK12 and in the cytoplasm of cells transformed with the control construct 35S:GUS (Figure 4, left and right panels, respectively). These results indicate that PK12 contains, in its sequence, a functional nuclear localization signal able to drive GUS expression to the nucleus, and strongly suggests a nuclear localization for PK12. Discussion Figure 3. Western blot and activity analysis of the recombinant wild-type and mutant PK12 proteins. (a) Western blot analysis using antibodies speci c for wild-type PK12 protein (agst-pk12, left panel) or the LAMMER motif (alammer, right panel). The arrows indicate different migration forms of PK12 in SDS± PAGE. The star (*) indicates a cross-reacting bacterial protein artefact. (b, c) Quanti cation of PK12, PK12 K125R and PK12 RAQ kinase activity. The proteins in the upper panel were assayed under conditions of autophosphorylation (1 mg each) or in the presence of MBP substrate (50 ng). Quanti cation of the protein activity is presented as the percentage of the maximum activity (lower panels). autophosphorylation activity and 0.75% of the wild-type MBP phosphorylation (Figure 3b,c). By contrast, as shown in Figure 3, recombinant PK12 RAQ is devoid of both kinase activities. Larger amounts (30 mg) were equally inactive (data not shown), suggesting that the LAMMER motif is essential for enzyme function. The lack of phosphorylation activity explains the faster migration of mutant polypeptides observed in Figure 3(a). PK12 localizes to the nucleus Consistent with a function in splicing regulation, the LAMMER kinases Clk/Sty from mammals have been LAMMER kinases are a gene family with distinct motifs that play a role in development, cell signalling and modulation of splicing activity, yet structure±function relationships are unknown (Bender and Fink, 1994; Colwill et al., 1996; Duncan et al., 1995; Sessa et al., 1996; Yun et al., 1994). The PK12 RAQ mutant demonstrates that the LAMMER motif is critical for PK12 kinase activity. It is unlikely that the amino acid substitutions introduced in the mutant abrogated kinase activity due to perturbation in global structure of the protein. In fact, PK12 RAQ bound atsrp34/sr1 in a similar fashion to the wild-type protein. Additional sitedirected mutations in LAMMER-type kinases have been reported. Over-expression of an inactive Clk/Sty that resulted from arginine substitution of the invariant lysine interferes with alternative splicing and has altered subnuclear localization compared to the wild-type protein (Colwill et al., 1996; Duncan et al., 1997). Mutation at serine 141, a major CLK2 autophosphorylation site that is conserved only among the mammalian LAMMER homologues, also in uences sub-nuclear localization (Nayler et al., 1998). Animal and plant LAMMER kinases differ mainly in their N-terminal portion (Sessa et al., 1996). In animal LAMMER kinases, the N-terminal region is rich in arginine and serine residues. Similar RS motifs characterize the C-terminus of SR splicing factors. Interaction between Clk/Sty and ASF/ SF2 requires the presence of this motif (Colwill et al., 1996), while its absence in a truncated recombinant DOA kinase results in weaker binding to SR proteins (Du et al., 1998). By contrast, tobacco PK12 as well as its Arabidopsis

5 Phosphorylation of SR splicing factors 95 plant SR splicing factor kinase together with the recent isolation of plant SR proteins presents an opportunity to investigate the regulation of alternative and constitutive splicing in plants. Experimental procedures Construction and expression of GUS and GUS:PK12 in tobacco Figure 4. Subcellular localization of PK12:GUS fusion in tobacco suspension cells. Left panel; histochemical examination for GUS expression in tobacco suspension cells bombarded with 35S-GUS:PK12 fusion 48 h after transformation. Right panel; histochemical examination for GUS expression in tobacco suspension cells bombarded with control 35S-GUS 48 h after bombardment. Similar results were obtained in three independent bombardment experiments. Bar = 30 mm. counterparts AFC1±3, have shorter N-terminus regions that lack RS-rich domains. Nevertheless, we show that PK12 binds atsrp34/sr1, suggesting an alternate mode of protein±protein interactions as has been shown for the splicing factor U2AF 35 that displays RS-independent binding (Wu and Maniatis, 1993). A unique characteristic of PK12 is its responsiveness to ethylene. Ethylene is involved in various aspects of plant growth and development as well as in stress and pathogenesis responses (Fluhr and Mattoo, 1996). Phosphorylation of SR splicing factors in nematodes can be developmentally regulated (Sanford and Bruzik, 1999). This raises the possibility that ethylene could exert control of splicing through PK12. In this regard, we note that the SR-related splicing factor Tra2 homologue is induced by exogenous pathogen-derived elicitor (Petitot et al., 1997). Other roles have been shown for LAMMER-type kinases. In Drosophila, Doa is critical for embryonic segmentation, normal development of the eye and the larval stages, and differentiation of the nervous system (Yun et al., 1994). PK12 is the rst plant kinase shown to directly phosphorylate and interact with SR splicing factors. The ability of PK12 to phosphorylate animal SR proteins re ects evolutionary conservation, consistent with reports that plant SR splicing factors are active when assayed in HeLa cell extract (Lazar et al., 1995; Lopato et al., 1996a,b). The activity of plant SR proteins and their regulation by phosphorylation/dephosphorylation remains to be shown. The discovery of a For subcellular localization experiments, the full-length coding region of PK12 was fused to the C-terminus of the GUS reporter gene in prtl2 under control of the 35S promoter (Carrington et al., 1991). The plasmid was then delivered to Nicotiana sylvestris suspension cells using a Model PDS-1000/He Biolistic particle delivery system (BioRad), according to manufacturer protocols. Expression of the reporter gene was analysed 48 h after bombardment by histochemical assay (Stomp, 1992) and was visualized by Nomarskii optics in a Zeiss Axioplan microscope Expression of recombinant PK12, atsrp43/sr1 and in vitro mutagenesis Recombinant PK12 was constructed with a six-histidine tag immediately before the stop codon and subcloned in the pgex1 expression vector (Pharmacia). Recombinant wild-type and mutant proteins were puri ed on an Ni 2+ column (Probond, Invitrogen) in the presence of phosphatase inhibitors (10 mm NaF; 100 mm NaVO 3 and 50 mm b-glycerophosphate) and proteinase inhibitors (1 mm phenylmethylsulphonyl uoride; 10 mg ml ±1 leupeptin and 10 mg ml ±1 aprotinin). To express atsrp34/sr1 in bacteria, its cdna was subcloned into the pqe-30 expression system (Qiagen) at the BamHI/SacI sites, and puri ed from inclusion bodies on an Ni±NTA column as described by Lazar et al. (1995). Site-directed mutagenesis of PK12 (MUTA-GENE kit; BioRad Laboratories) was performed on a single-strand pbluescript vector containing PK12 cdna using the following oligonucleotides: for PK12 K125R, 5 -GATTTGTTGCCATCAGGA- TTATCCGAAGTATC-3 ; for PK12 RAQ, 5 ± GAGCACCTGGCG- AGGGCGCAAAGAGTATTGGG-3 (bold nucleotides represent mutated nucleotides). Kinase activity assays In vitro kinase assays were performed as described by Sessa et al. (1996). Phosphatase treatment of autophosphorylated PK12 and HeLa SR splicing factors was performed with 5 U of CIP alkaline phosphatase (AP, Boehringer) for 40 min at 30 C. Western blot and in vitro binding Proteins were resolved in SDS±PAGE and transferred to a nitrocellulose membrane (Schleicher & Schuell). Conditions for analysis of a-gst-pk12, a-lammer (Sessa et al., 1996) and for mab 104 (Zahler et al., 1992) were as described previously. For in vitro binding analysis, probes PK12 and PK12 RAQ were translated in the presence of 35 S-methionine using the T7 promoter with the TNT Quick Coupled transcription/translated kit (Promega). Nitrocellulose blots containing 1 mg of atsrp34/sr1 and MBP were prepared and the transfer ef ciency of polypeptides was

6 96 Sigal Savaldi-Goldstein et al. assayed by Ponceau S stain. The blots were incubated with equal amounts of radioactive probe, washed as for Western blots and autoradiographed. Acknowledgements Our thanks to Dr Ron Vunsh for technical assistance. AtSRp34/ SR1 cdna was kindly provided by Dr Howard M. Goodman. Antibody to LAMMER sequence was kindly provided by Dr Leonard Rabinow and was af nity-puri ed on a nitrocellulose strip containing PK12. mab 104 and SR splicing factors from HeLa were kindly provided by Dr Joseph Sperling. This research was supported by the Harry and Jeanette Weinberg Center for Plant Molecular Genetics Research. References Bender, J. and Fink, G.R. (1994) AFC1, a LAMMER kinase from Arabidopsis thaliana, activates STE12-dependent processes in yeast. Proc. Natl Acad. Sci. USA, 91, 12105± Brown, J.W.S. and Simpson, C.G. (1998) Splice site selection in plant pre-mrna splicing. Annu. Rev. Plant Physiol. Plant Mol. Biol. 49, 77±95. Carrington, J.C., Freed, D.D. and Leinicke, A.J. (1991) Bipartite signal sequence mediates nuclear translocation of the plant potyviral Nla. Plant Cell, 3, 953±962. Colwill, K., Pawson, T., Andrews, B., Prasad, J., Manley, J.L., Bell, J.C. and Duncan, P.I. (1996) The Clk/Sty protein kinase phosphorylates SR splicing factors and regulates their intranuclear distribution. EMBO J. 15, 265±275. Du, C., McGuf n, M.E., Dauwalder, B., Rabinow, L. and Mattox, W. (1998) Protein phosphorylation plays an essential role in the regulation of alternative splicing and sex determination in Drosophila. Mol. Cell. 2, 741±750. Duncan, P.I., Howell, B.W., Marius, R.M., Drmanic, S., Douville, E.M.J. and Bell, J.C. (1995) Alternative splicing of Sty, a nuclear dual-speci city kinase. J. Biol. Chem. 270, 21524± Duncan, P.I., Stojdl, D.F., Marius, R.M. and Bell, J.C. (1997) In vivo regulation of alternative pre-mrna splicing by the Clk1 protein kinase. Mol. Cell. Biol. 17, 5996±6001. Fluhr, R. and Mattoo, A.K. (1996) Ethylene ± biosynthesis and perception. Crit. Rev. Plant Sci. 15, 479±523. Gui, J.F., Lane, W.S. and Fu, X.D. (1994) A serine kinase regulates intracellular-localization of splicing factors in the cell cycle. Nature. 369, 678±682. Lazar, G., Schaal, T., Maniatis, T. and Goodman, H.M. (1995) Identi cation of a plant serine±arginine-rich protein similar to the mammalian splicing factor Sf2/Asf. Proc. Natl Acad. Sci. USA, 92, 7672±7676. Lee, K., Du, C., Horn, M. and Rabinow, L. (1996) Activity and autophosphorylation of LAMMER protein kinases. J. Biol. Chem. 271, 27299± Lopato, S., Mayeda, A., Krainer, A.R. and Barta, A. (1996a) PremRNA splicing in plants: characterization of Ser Arg splicing factors. Proc. Natl Acad. Sci. USA, 93, 3074±3079. Lopato, S., Waigmann, E. and Barta, A. (1996b) Characterization of a novel arginine/serine-rich splicing factor in Arabidopsis. Plant Cell. 8, 2255±2264. Lopato, S., Kalyna, M., Dorner, S., Kobayashi, R., Krainer, A.R. and Barta, A. (1999) atsrp30, one of two SF2/ASF-like proteins from Arabidopsis thaliana, regulates splicing of speci c plant genes. Genes Dev. 13, 987±1001. Manley, J.L. and Tacke, R. (1996) SR proteins and splicing control. Genes Dev. 10, 1569±1579. Nayler, O., Schnorrer, F., Stamm, S. and Ullrich, A. (1998) The cellular localization of the murine serine/arginine-rich protein kinase CLK2 is regulated by serine 141 autophosphorylation. J. Biol. Chem. 273, 34341± Petitot, A.S., Blein, J.P., Pugin, A. and Suty, L. (1997) Cloning of two plant cdnas encoding a beta-type proteasome subunit and a transformer-2-like SR-related protein: early induction of the corresponding genes in tobacco cells treated with cryptogein. Plant Mol. Biol. 35, 261±269. Roth, M.B., Murphy, C. and Gall, J.G. (1990) A monoclonal antibody that recognizes a phosphorylated epitope stains lampbrush chromosome loops and small granules in the amphibian germinal vesicle. J. Cell Biol. 111, 2217±2223. Sanford, J.R. and Bruzik, J.P. (1999) Developmental regulation of SR protein phosphorylation and activity. Genes Dev. 13, 1513± Sessa, G., Raz, V., Savaldi, S. and Fluhr, R. (1996) PK12, a plant dual-speci city protein kinase of the LAMMER family, is regulated by the hormone ethylene. Plant Cell. 8, 2223±2234. Stomp, A.-M. (1992) Histochemical localization of b- glucuronidase. In GUS Protocols (Gallagher, S.R., ed.) San Diego: Academic Press, pp. 103±113. Wang, H.Y., Lin, W., Dyck, J.A., Yeakley, J.M., Zhou, S.Y., Cantley, L.C. and Fu, X.D. (1998) SRPK2: a differentially expressed SR protein-speci c kinase involved in mediating the interaction and localization of pre-mrna splicing factors in mammalian cells. J. Cell Biol. 140, 737±750. Wu, J.Y. and Maniatis, T. (1993) Speci c interactions between proteins implicated in splice-site selection and regulated alternative splicing. Cell, 75, 1061±1070. Xiao, S.H. and Manley, J.L. (1997) 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. Yun, B., Farkas, R., Lee, K. and Rabinow, L. (1994) The Doa locus encodes a member of a new protein kinase family and is essential for eye and embryonic-development in Drosophila melanogaster. Genes Dev. 8, 1160±1173. Zahler, A.M., Lane, W.S., Stolk, J.A. and Roth, M.B. (1992) SR proteins ± a conserved family of pre-messenger-rna splicing factors. Genes Dev. 6, 837±847.