This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 18/1/146    most recent tpc.105.037069v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (16) Citing Articles via Google Scholar Google Scholar Articles by Isshiki, M. Articles by Shimamoto, K. Search for Related Content PubMed PubMed Citation Articles by Isshiki, M. Articles by Shimamoto, K. Agricola Articles by Isshiki, M. Articles by Shimamoto, K.

The Serine/Arginine-Rich Protein Family in Rice Plays Important Roles in Constitutive and Alternative Splicing of Pre-mRNA^[W]

Laboratory of Plant Molecular Genetics, Nara Institute of Science and Technology, 8916-5 Takayama, Ikoma 630-0101, Japan

¹ To whom correspondence should be addressed. E-mail simamoto{at}bs.naist.jp; fax 81-743-72-5509.

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Ser/Arg-rich (SR) proteins play important roles in the constitutive and alternative splicing of pre-mRNA. We isolated 20 rice (Oryza sativa) genes encoding SR proteins, of which six contain plant-specific characteristics. To determine whether SR proteins modulate splicing efficiency and alternative splicing of pre-mRNA in rice, we used transient assays in rice protoplasts by cotransformation of SR protein genes with the rice Waxy^b (Wx^b)-ß-glucuronidase fusion gene. The results showed that plant-specific RSp29 and RSZp23, an SR protein homologous to human 9G8, enhanced splicing and altered the alternative 5' splice sites of Wx^b intron 1. The resulting splicing pattern was unique to each SR protein; RSp29 stimulated splicing at the distal site, and RSZp23 enhanced splicing at the proximal site. Results of domain-swapping experiments between plant-specific RSp29 and SCL26, which is a homolog of human SC35, showed the importance of RNA recognition motif 1 and the Arg/Ser-rich (RS) domain for the enhancement of splicing efficiencies. Overexpression of plant-specific RSZ36 and SRp33b, a homolog of human ASF/SF2, in transgenic rice changed the alternative splicing patterns of their own pre-mRNAs and those of other SR proteins. These results show that SR proteins play important roles in constitutive and alternative splicing of rice pre-mRNA.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Alternative splicing is an important mechanism in the regulation of gene expression in eukaryotes. It enables the generation of proteins with different functions and structures through variations in the splicing patterns of pre-mRNA from one gene (Cartegni et al., 2002; Maniatis and Tasic, 2002; Black, 2003). In the human genome, 74% of transcripts are alternatively spliced (Johnson et al., 2003). The selection of alternative splice sites is determined by the assembly of the spliceosome, a large complex containing five small nuclear ribonucleoproteins (snRNPs) and non-snRNP proteins. Ser/Arg-rich (SR) proteins can bind to specific RNA sequences and assemble the spliceosome at weak splice sites in alternative splicing. The SR proteins have one or two RNA recognition motifs (RRMs) in the N terminus and an Arg/Ser-rich (RS) domain for protein–protein interaction in the C terminus (Graveley, 2000; Cartegni et al., 2002; Maniatis and Tasic, 2002). Moreover, SR proteins are essential splicing factors for constitutive splicing and are highly conserved in metazoans and plants. Ten members of the SR protein family were identified in humans (Graveley, 2000), and Schizosaccharomyces pombe contains at least two SR proteins (Gross et al., 1998; Lutzelberger et al., 1999).

The activity of SR proteins for alternative splicing depends on specific RNA sequences called exonic splicing enhancers (ESEs). SR proteins bind to ESEs to stimulate U2AF binding to the upstream weak 3' splice site or U1 snRNP binding to the weak 5' splice site (Graveley, 2000; Cartegni et al., 2002; Maniatis and Tasic, 2002). The target sequences of SR proteins have been identified by SELEX, an in vitro selection method, to determine a high-affinity binding site for the RRMs of SR proteins. It was also shown that SR proteins may have distinct functions in constitutive splicing and in the splicing enhancer-dependent function and that these functions can be uncoupled (Graveley and Maniatis, 1998).

The functions of SR proteins in mRNA metabolism, however, are not only in constitutive and alternative splicing but also in nuclear export, mRNA stability, and translation. Three human SR proteins, ASF/SF2, SRp20, and 9G8, shuttle between the nucleus and cytoplasm depending on the phosphorylation of the RS domain (Caceres et al., 1998) and could promote the export of intronless mRNAs (Huang and Steitz, 2001). ASF/SF2 also decreases mRNA stability by binding to a 3' untranslated region sequence (Lemaire et al., 2002). Moreover, ASF/SF2 has been shown to stimulate the translation of a luciferase reporter in an enhancer-dependent manner (Sanford et al., 2004).

The basic mechanism of splicing in higher plants is similar to the one observed in vertebrates. Specifically, the consensus sequences of the 5' and 3' splice sites, AG/GT and AG/ (the slash denotes the exon–intron boundary), and the branch point are conserved in plants (Liu and Filipowicz, 1996). However, the polypyrimidine tract, which is essential for the recognition of the 3' splice site in vertebrates, is not present in the plant intron. Plant introns contain more AU-rich sequences than exons, and AU- or U-rich sequences in the plant intron are required for its recognition (Goodall and Filipowicz, 1989; Ko et al., 1998). Therefore, these results suggest that the splicing factors or mechanisms for the recognition of a plant intron are different from those in vertebrates. UBP1, a U-rich RNA binding protein, was previously identified in Nicotiana plumbaginifolia. Overexpression of UBP1 in protoplasts increased the splicing efficiency (Lambermon et al., 2000), suggesting that UBP1 binding to the U-rich intron in plants recruits splicing factors to pre-mRNA.

Recent analysis of the Arabidopsis thaliana genome has revealed at least 19 SR protein genes (Lorkovic and Barta, 2002; Reddy, 2004). Among those, some are homologs of human SR proteins, ASF/SF2, SC35, and 9G8, but some are plant specific. Novel plant-specific SR proteins identified in Arabidopsis have characteristic structures (Lopato et al., 1996, 2002; Golovkin and Reddy, 1999). Some SR proteins identified in Arabidopsis can complement the HeLa cell S100 extract (Lopato et al., 1996, 1999a), interact with the Arabidopsis U1-70K protein (Golovkin and Reddy, 1998, 1999), and bind to the specific RNA sequences (Lopato et al., 1999a); in addition, they are phosphorylated by Arabidopsis Clk/Sty protein kinase, AFC2 (Golovkin and Reddy, 1999). In addition, the green fluorescent protein (GFP) fusion localizes in the nucleus in a speckled pattern (Lopato et al., 2002; Ali et al., 2003; Docquier et al., 2004; Tillemans et al., 2005). These results suggest that Arabidopsis SR proteins have similar functions to human SR proteins.

Recently, it was demonstrated that the regulation of alternative splicing in higher plants is important in physiology and development. Many nucleotide binding site–leucine-rich repeat (NBS-LRR)-type plant disease resistance genes exhibit alternative splicing (Jordan et al., 2002). Both full-length and alternative transcripts of the tobacco (Nicotiana tabacum) N and the Arabidopsis RPS4 genes are required for disease resistance (Dinesh-Kumar and Baker, 2000; Zhang and Gassmann, 2003). In addition, the Arabidopsis FCA gene, which encodes an RNA binding protein, negatively regulates its own expression by alternative splicing (Quesada et al., 2003). The spinach (Spinacia oleracea) chloroplast ascorbate peroxidase gene has four mRNA variants that differ at the C terminus transit peptide to produce a stromal or thylakoid-bound type protein (Yoshimura et al., 2002). In rice (Oryza sativa), the Waxy (Wx) gene, which encodes a granule-bound starch synthase, has two alleles, Wx^a and Wx^b. Wx^b has a G-to-T mutation at the 5' splice site of intron 1 (Cai et al., 1998; Frances et al., 1998; Hirano et al., 1998; Isshiki et al., 1998). This mutation creates two cryptic splice sites in exon 1 and reduces the splicing efficiency. The alternative splicing pattern of Wx^b mRNA is affected by temperature (Larkin and Park, 1999), resulting in poor quality of rice grains if seed maturation occurs at low temperature.

In this study, we identified 20 genes encoding SR proteins from the rice EST, full-length cDNA, and genome databases and examined their structures, expression, and functions in RNA splicing. The results showed that two SR proteins, RSp29 and RSZp23, enhanced the splicing efficiency and changed the alternative 5' splice sites of Wx^b intron 1. Rice SR proteins homologous to human ASF/SF2 and SC35 reduced the splicing efficiency of Wx^b intron 1. Furthermore, we identified sequences similar to mammalian ESEs in exon 1 of Wx^b and found that their mutations enhanced ß-glucuronidase (GUS) reporter activity when cotransformed with RSp29. Finally, overexpression of some rice SR genes changed the alternative splicing pattern of its own pre-mRNAs and those of other SR proteins in transgenic rice plants.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Structure of Rice SR Protein Genes
We have found 20 genes encoding putative SR proteins in rice (see Supplemental Figure 1 online), and their structures are indicated in Figure 1. Human ASF/SF2 has a conserved RRM with two submotifs, RNP-1 and RNP-2, and a domain homologous to RRM, which is called RRM or HRRM. Rice has four ASF/SF2-type SR proteins that display 75 to 85% identity with each other. Although mRNA for SR20 was detected by RT-PCR, it is not likely to function because of the premature termination codon in its RRM region. Seven genes encode the human SC35-type protein, which has one N-terminal RRM and one C-terminal RS domain. Rice SC35-type proteins are classified into three groups based on the N-terminal amino acid sequences upstream of RRM. SC35a, SC35b, and SC35c are the most similar to human SC35. The N-terminal regions of SCL25, SCL26, SCL30a, and SCL30b are longer than that of the human SC35, and the N-terminal amino acid sequences of SCL25 and SCL26 are different from those of SCL30a and SCL30b. In addition, three human 9G8-type proteins, which have one zinc knuckle of the CCHC-type between the RRM and the RS domain, are present in rice. Plant-specific SR proteins were identified in rice. RSp29 and RSp33 have two tandem RRMs, one of which is an intermediate region (RRM2) different from that of human ASF/SF2. Moreover, RSZ36, RSZ37a, RSZ37b, and RSZ39 contain two CCHC-type zinc knuckles. In addition to the typical SR protein homologs, rice has two proteins similar to Arabidopsis SR45, which contains two RS domains at both the N terminus and the C terminus (Reddy, 2004). A phylogenetic analysis of SR proteins in rice, Arabidopsis, and humans revealed that each family was clustered in accordance with the domain structure (see Supplemental Figure 2 online).

View larger version (35K): [in this window] [in a new window]   Figure 1. Putative Structure of Rice SR Proteins. Only major types of domains are shown (not to scale). SR protein families were classified according to the structure of the domains. SR20 may be a truncated protein because of the presence of a premature stop codon. All putative protein sequences are shown in Supplemental Figure 1 online. RRM, purple; RS, blue; RBD, pseudo RRM, green; Zn, zinc knuckle, yellow.

View larger version (97K): [in this window] [in a new window]   Figure 2. Expression of SR Protein mRNA in Different Tissues of Rice. Total RNA from different tissues of rice cv Kinmaze was analyzed. RT-PCR was performed using specific primers for each SR protein mRNA. R, root; L, leaf; S, immature seeds at 15 d after pollination; C, cell culture.

View larger version (19K): [in this window] [in a new window]   Figure 3. Alternative Splicing Patterns of Rice SR Protein Genes. Schematic diagram of rice SR protein genes having alternative splicing form and their mRNA isoforms. Constitutive exons are shown as white boxes, and alternative exons are shown as gray boxes. Asterisks indicate premature stop codons.

Functional Analysis of Rice SR Proteins in Transient Assays Using Cell Culture Protoplasts
To investigate the effects of SR proteins on the efficiency of pre-mRNA splicing, we used transient assays using protoplasts isolated from rice Oc cell cultures (Isshiki et al., 1998, 2001). For the analysis of splicing, we used the Wx^b-gus gene. The Wx^b gene has a G-to-T mutation at the 5' splice site of intron 1 in the untranslated leader region and stimulates two cryptic splice sites. One is 93 nucleotides upstream of the authentic site (site 2), and the second is one base upstream of the authentic site (Isshiki et al., 1998). When spliced mRNA is analyzed by RT-PCR, the authentic splice site and the second cryptic splice site cannot be distinguished; therefore, they are collectively called site 1 (Figure 4A).

View larger version (31K): [in this window] [in a new window]   Figure 4. Effects of Various Rice SR Proteins on the Expression of the Wxb-gus Reporter Gene. The assays were performed with transiently transformed protoplasts derived from the Oc cell culture. The GUS activities of the extracts obtained from the protoplasts were measured. (A) Constructs used to analyze the effects of SR protein genes on the reporter gene expression. The promoter-ATG region of Wxb was translationally fused to the gus reporter gene (Isshiki et al., 1998). Arrows show the relative positions of primers for RT-PCR. The cDNAs of RSZp23, RSp29, SRp32, SRp33a, SCL25, SCL26, RSp33, RSZ36, and RSZ37a were fused with the cauliflower mosaic virus 35S promoter. (B) to (D) GUS activity in cotransfected protoplasts. The indicated SR plasmids were added to protoplasts together with a Wxb-gus fusion gene. Ten micrograms each of Wxb-gus and various amounts of 35S-SR plasmid DNA (1, 2, 5, and 10 µg) were added to 5 x 106 protoplasts. The mean and standard error for three independent experiments are shown. (B) RSp29 and RSZp23, which enhanced GUS activity. (C) SRp32, SRp33a, SCL25, and SCL26, which reduced GUS activity. (D) RSp33, RSZ36, and RSZ37a, which did not affect GUS activity when 1 to 2 µg of SR protein genes was added.

RSp29, which is a plant-specific SR protein, increased the GUS activity approximately twofold relative to the control (Figure 4B). Results of experiments to test the dosage effect of the plasmid DNA showed that the GUS activity was the highest when 1 to 2 µg of the RSp29 plasmid was used. To examine the splicing efficiency and possible changes in splice site selection, RNAs from these protoplasts were analyzed by RT-PCR, and the splice sites of the Wx^b intron 1 were determined by sequence analysis (Figure 5A). In protoplasts transformed with the Wx^b-gus gene alone, the original splice site (site 1) and site 2, which was 93 nucleotides upstream from site 1, were used with similar efficiency. By contrast, when Wx^b-gus was cotransformed with 35S-RSp29, the 5' splice site was shifted to site 2 (Figure 5A).

View larger version (58K): [in this window] [in a new window]   Figure 5. Analysis of RNA to Examine Splice Site Selection in Transfection Assays. Total RNA isolated from transfected protoplasts shown in Figure 4 was subjected to RT-PCR analysis. PCR conditions were the same for spliced and unspliced transcripts. The RT-PCR products were sequenced to confirm the splice sites. The top loading gels show the spliced transcripts of Wxb-gus. The spliced products are indicated at right: the top fragment shows the site 1–spliced transcript of Wxb-gus, and the bottom fragment shows the site 2–spliced transcript. The bottom PCR product is 93 bp shorter than the top one. E1 and E2 show exon 1 and exon 2 of the Wxb gene, respectively. The bottom loading gels show unspliced transcripts of Wxb-gus. (A) Effects of 35S-RSp29 on the splicing of Wxb intron 1. (B) Effects of 35S-RSZp23 on the splicing of Wxb intron 1. (C) Effects of eight 35S-SR plasmids on splicing patterns. The extracts obtained from transfected protoplasts were used for RNA isolation for RT-PCR analysis. Ten micrograms of Wxb-gus and 2 µg of 35S-SR DNAs were cotransfected into 5 x 106 protoplasts.

View larger version (35K): [in this window] [in a new window]   Figure 6. Splicing Efficiencies of RSp29-SCL26 Chimeric Genes Tested in Rice Protoplasts. (A) Diagram of RSp29 and SCL26 proteins and the chimeric proteins used in the experiments. Domains of RSp29 and SCL26 proteins are shown by white and gray boxes (not to scale), respectively. (B) GUS activity and splicing pattern. The Wxb-gus plasmid (10 µg) and each 35S-SR plasmid (2 µg) were transfected into 5 x 106 protoplasts. The GUS activities of the extracts obtained from transfected protoplasts are shown. The mean and standard error for three independent experiments are shown. The aliquots of the extracts were used for RNA isolation for RT-PCR analysis. The top loading gels show the spliced transcripts of Wxb-gus. The bottom loading gels show unspliced transcripts of Wxb-gus. E1 and E2 show exons 1 and 2 of the Wxb gene, respectively.

The RRM1 Motif of Plant-Specific RSp29 Is Essential for the Increased Efficiency of Splicing, but the RS Domain Is Exchangeable with Other SR Proteins
To test which domain of RSp29 is important for the enhancement of the splicing of Wx^b intron 1, we performed transient assays using constructs having different domains between RSp29 and SCL26 (Figure 6). SCL26 slightly reduced the GUS activity but did not clearly affect the splicing of Wx^b-gus RNA (Figures 4C and 5C). The DS2 construct, which had two RRM motifs derived from RSp29 plus the RS domain of SCL26, showed higher GUS activity than RSp29. Therefore, the results showed that the RS domain was exchangeable with other SR proteins. They also showed that the RS domain of SCL26 may be more efficient that that of RSp29. Next, the requirement of the second RRM (RRM2) of RSp29 was tested using the RRM2 construct in which the RRM2 domain was lacking. The results showed that RRM2 exhibited the same activity as RSp29, indicating that RRM2 is dispensable for the enhancement of splicing by RSp29. Selection of 5' splice site by RRM2 was not changed from RSp29, suggesting that the RRM2 domain of RSp29 is not involved in 5' splice site selection. As expected, RS, which lacked the RS domain, completely lost the ability to enhance GUS activity or splicing but did not change 5' splicing site selection. These results indicate that the RS domain is essential for splicing activity but not for splice site selection. Together, these results indicate that the RRM1 motif of RSp29 is important for the increased efficiency of splicing and splice site selection; however, molecular mechanisms for these findings remain to be studied.

ESE-Like Sequences in Wx^b Pre-mRNA
Sequences within exons have been shown to affect the splice site selection and splicing efficiency. These sequences are called the ESEs. SR proteins can activate in vitro splicing when these specific sequences are introduced into model RNA substrates. High-affinity RNA sequences with SR proteins were determined using the SELEX method. Although SR proteins show distinct RNA binding specificities, these consensus sequences are degenerate. For instance, human ASF/SF2 interacts with the consensus sequences (A/G)GAAGAAC and AGGAC(A/G)(A/G)AGC (Tacke and Manley, 1995). In plants, however, there have been no reports showing that SR proteins activate the splicing of pre-mRNA containing ESEs.

Therefore, to investigate whether specific RNA sequences of Wx^b affect the splicing efficiency, we searched for known ESE sequences in Wx^b pre-mRNA. We found GGAAGAAC (Figure 7A, double underline), which is known as the binding site of human ASF/SF2 (Tacke and Manley, 1995), and UGCAGTC (Figure 7A, single underline), which was reported to be the site for human SC35 (Schaal and Maniatis, 1999), in Wx^b exon 1. No other ESE-like sequences were found in the Wx^b gene. We then made mutant constructs having mutations in these ESE-like sequences in Wx^b (Figure 7B). Because ASF/SF2 interacts with purine-rich sequences (Liu et al., 1998), the Wx^b(mutB)-gus construct was produced by changing the ASF/SF2 binding sequence GGAAGAAC to the pyrimidine-rich sequence CCTTCTTG.

View larger version (27K): [in this window] [in a new window]   Figure 7. Effects of the ESE-Like Sequences in Wxb Exon 1 on Splicing by RSp29 and RSZp23. (A) Sequences of Wxb exon 1, which does not contain the start codon. It is present in exon 2. Sequences that are similar to the human SC35 and ASF/SF2 binding sequences are shown by a single underline (A) and double underline (B), respectively. The 5' splice sites 1 and 2 are shown by two open triangles and a closed triangle, respectively. The SC35 binding sequences overlaps with site 2. (B) Mutant vector constructs used for splicing assays. Wxb(mutA)-gus was changed to ACCAGTA from TGCAGTC. Wxb(mutB)-gus was changed to a pyrimidine-rich sequence (CCTTCTTG) from a purine-rich sequence (GGAAGAAC). (C) GUS activity and splicing in cotransfected protoplasts. Protoplasts (5 x 106) were transfected by Wxb-gus or the mutant Wxb-gus plasmid (10 µg) with 35S-RSp29 or 35S-RSZp23 plasmids (2 µg). The mean and standard error for three independent experiments are shown. RT-PCR analysis of RNA isolated from protoplasts transfected with mutant Wxb-gus constructs. The top loading gels show spliced transcripts of Wxb-gus. The bottom loading gels show unspliced transcripts of Wxb-gus. E1 and E2 show exons 1 and 2 of the Wxb gene, respectively.

The Wx^b(mutA)-gus was also constructed by changing three nucleotides (ACCAGTA). Because this region overlaps with the splice site 2, the core sequence could not be changed. Introduction of mutA did not cause reduction of GUS activity or reduction of splicing efficiency, suggesting that as sequence B, sequence A is unlikely to function as a general ESE in the absence of additional SR proteins. In contrast with Wx^b(mutB)-gus, however, Wx^b(mutA)-gus showed slightly reduced GUS activity in the presence of RSZp23 (Figure 7C). Results of RT-PCR analysis showed that splice sites were changed, but splicing efficiency was not reduced. These results may suggest the presence of a complex relationship between splicing efficiency and GUS activity in these protoplast assays and that sequence A may play a role in other steps of RNA metabolism regulated by RSZp23. By contrast, mutA clearly enhanced the GUS activity and splicing efficiency in the presence of RSp29, which is a plant-specific SR protein, and its binding site is unknown. However, no changes in splice site selection were found. These results may suggest that the mutA sequence may interact with RSp29. Alternatively, other proteins with inhibitory effects on splicing may not be able to bind the mutA sequence. Taken together, these results indicate that ESE-like sequences found in Wx^b exon 1 may not directly affect splicing efficiency but other steps of mRNA metabolism, such as mRNA export or translation, as was recently reported in animals (Sanford et al., 2004).

The GUS activity of Wx^b(mutA)-gus and Wx^b(mutB)-gus was higher than that of the wild-type construct in the absence of additional SR proteins. Although reasons for this observation are not clear at the moment, it is possible that the introduction of these mutations may have altered the structure of the mRNA, which results in increased splicing and changes in other steps of RNA metabolisms.

Overexpression of RSZ36 Causes Changes in the Splicing Pattern of Its Own RNA
To analyze the roles of the SR protein in alternative splicing, rice SR protein cDNAs, SRp32, SRp33a, SRp33b, SCL26, RSZp23, RSp29, RSp33, RSZ36, and RSZ37a, were fused with the 35S promoter and transformed into rice, and the expression of various mRNAs was examined for possible changes in splicing patterns by RT-PCR. Clear changes in the splicing pattern of RSZ36 RNA were observed in rice plants transformed with 35S-RSZ36 (Figure 8B). The RSZ36 transcripts, as shown in Figure 8A, were identified in transgenic plants by sequencing analysis. The RSZ36 gene has three transcripts: one mRNA encodes the full-length protein (mRNA-a), and the second transcript has 219 nucleotides of intron 2 because of the shift of the 3' splice site in intron 2 (mRNA-c). In addition, mRNA-b has a 219-nucleotide sequence derived from intron 2. In mRNA-c, a 111-nucleotide sequence was spliced from exon 5. We could not identify mRNA-b in wild-type plants (Figure 3) because of low expression. The mRNA-b and mRNA-c are likely to produce truncated proteins because of the presence of premature stop codons in the newly produced exon. In untransformed control plants, mainly mRNA-a was detected. In overexpressors of RSZ36, however, mRNA-b and mRNA-c were increased, and mRNA-a was decreased (Figure 8B). These results suggest the possibility that the amount of functional RSZ36 mRNA is regulated by a feedback mechanism so that an appropriate quantity is always present.

View larger version (41K): [in this window] [in a new window]   Figure 8. Overexpression of RSZ36 Affects the Alternative Splicing of Its Own RNAs in Transgenic Rice Plants. (A) Schematic representation of the RSZ36 gene structure and its mRNA isoforms detected in RSZ36-overexpressing plants. Exons are shown as boxes and introns as lines. A gray box indicates an alternative exon in intron 2; a hatched box shows an alternative intron in exon 5. Positions of primers used for RT-PCR analyses are shown with arrowheads. Asterisks indicate premature stop codons. (B) RT-PCR analysis of RSZ36 transcripts in the T0 generation of RSZ36-overexpressing plants. Lanes 3-1, 3-3, and 3-4 were regenerated from the same transformed calli. Lanes 3-1, 4, 12, 13, and 15 were independent transgenic lines. The top panel displays the expression of the endogene and transgene amplified using the F-2 and R primers. This PCR was performed with 25 cycles. The bottom panel shows only the expression of the endogenous RSZ36 gene amplified by F-1 and R primers. This PCR was performed with 30 cycles for the detection of endogenous RSZ36 transcripts. a, b, and c correspond to mRNA-a, -b, and -c, respectively.

View larger version (32K): [in this window] [in a new window]   Figure 9. Overexpression of SRp33b Affects the Alternative Splicing of SRp33a and SRp32 in Transgenic Rice Plants. (A) Schematic representations of SRp33a and SRp32 gene structures and their mRNA isoforms detected in SRp33b-overexpressing plants. Gray boxes indicate alternative exons. Asterisks indicate premature stop codons. (B) RT-PCR analysis in the T0 generation of SRp33b-overexpressing plants. The lane number shows independent transgenic lines. SRp33a, SRp32, and SRp33b transcripts were amplified by each specific primer designed from 5' and 3' untranslated region sequences. The RT-PCR data of SRp33b display the expression of the endogene and transgene. a and c show transcripts encoding the full-length protein. b, d, and e show alternative splicing products. Total RNA was isolated from the leaf of transgenic rice plants.

To examine whether overexpression of rice SR proteins, SRp32, SRp33a, SRp33b, SCL26, RSZp23, RSZ36, and RSZ37a, influences the splicing pattern of endogenous genes, we analyzed the splicing patterns of 38 rice genes (see Supplemental Table 1 online) by RT-PCR analysis. These 38 genes show alternative splicing and are expressed in leaves. They were randomly selected from the full-length cDNA database (Rice Full-Length cDNA Consortium, 2003). However, the results of RT-PCR using leaf RNA did not show any change in the splicing pattern (data not shown). No clear changes in the gross morphology of these overexpressing plants were observed. Transgenic rice plants overexpressing RSp29 and RSp33 could not be recovered by repeated transformation experiments (data not shown), suggesting that these plant-specific SR proteins are involved in the splicing, export, or translation of mRNAs that are essential for growth and development.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Rice SR Proteins
The completion of the rice genome sequencing project (Feng et al., 2002; Goff et al., 2002; Sasaki et al., 2002; Yu et al., 2002; Rice Chromosome 10 Sequencing Consortium, 2003) and rice full-length cDNA data (Rice Full-Length cDNA Consortium, 2003) allow the identification of all members of a gene family, such as the family of SR proteins in rice. Here, we identified 20 genes encoding putative SR proteins in rice. The structure of these putative proteins is essentially similar to those of human SR proteins and plant-specific SR proteins from Arabidopsis (Lorkovic and Barta, 2002; Reddy, 2004). In Arabidopsis, differential expression patterns of SR protein genes were found in different tissues (Lopato et al., 1996, 1999b; Golovkin and Reddy, 1998, 1999; Lazar and Goodman, 2000; Kalyna et al., 2003). Most SR protein genes were highly expressed in root and flower regardless of the SR protein family. Investigations by the use of promoter-GUS fusions revealed that Arabidopsis RSZ33 is mainly expressed in the lateral root and elongation zone of root, style, stigma, and mature pollen (Kalyna et al., 2003). These tissue-specific-expressed SR proteins might regulate the alternative splicing of transcripts in these tissues. In this study, however, the expression of rice SR protein genes was not clearly different in the root, leaf, and other tissues (Figure 2). In Arabidopsis, ectopic expression of SRp30 or RSZ33 showed morphological and developmental changes (Lopato et al., 1999b; Kalyna et al., 2003). By contrast, overexpression of rice SR proteins did not influence plant growth in general, but more detailed observations during various stages of development may be required. These results revealed the possibility that the transcriptional regulation of rice SR protein genes makes a limited contribution to their functions in the splicing machinery.

Each type of rice SR protein has multiple members. Therefore, it is expected that rice SR proteins have certain levels of redundant function in splicing. For example, we identified a Tos17-induced mutant of RSZ36; however, no obvious effect on the morphology and growth was detected in this mutant (our unpublished results). Similar examples have been found in other organisms. In Caenorhabditis elegans, inhibition of the ASF/SF2-like function by RNA interference resulted in embryonic lethality (Longman et al., 2000). However, no visible phenotype was observed in individual RNA interference with other SR genes, and the embryonic lethal phenotype was only found when five SR protein genes, but not ASF/SF2, were simultaneously suppressed. These results suggest that some SR proteins have redundant functions and that the absence of a particular SR protein can be rescued by the presence of other SR proteins.

The expression level of some SR protein genes appears to be autoregulated by changing alternative splicing of its own pre-mRNA, as shown in RSZ36 (Figure 8). This type of negative feedback regulation is known for human ASF/SF2 and SRp20 (Wang et al., 1996; Jumaa and Nielsen, 1997) and Arabidopsis SRp30 and RSZ33 (Lopato et al., 1999b; Kalyna et al., 2003). These results show that feedback regulation could be a common mechanism to control the level of SR proteins in animals and plants. However, overexpression of RSp29, which has no alternative splicing form (Figure 2), may not autoregulate its own pre-mRNA; therefore, it is not possible to regenerate plants from a transgenic callus.

Splice Site Selection by Rice SR Proteins
We demonstrated that two rice SR proteins, RSp29 and RSZp23, could change the alternative 5' splice sites of Wx^b intron 1 in transient assays using rice cell culture plotoplasts (Figure 5). RSp29 and RSZp23 stimulated site 2 (distal site) and site 1 (proximal site), respectively, when they were introduced into protoplasts. In humans, specific SR proteins have distinct functions in promoting alternative splice site selection. For instance, SV40 pre-mRNA, which has two alternative 5' splice sites, is spliced at the proximal 5' splice site by SC35, while it is spliced at the distal 5' splice site by SRp40 or SRp55 (Zahler et al., 1993). These results show how tissue- and/or development-specific alternative splicing is regulated by SR proteins. It was previously shown that a plant SR protein could activate the proximal 5' splice site in a human in vitro alternative splicing system (Lazar et al., 1995; Lopato et al., 1996). In the analysis of transgenic Arabidopsis overexpressing SR proteins, it was reported that SRp30 and RSZ33 could influence alternative 5' splice site selection in intron 10 of SRp34/SR1; SRp30 stimulated splicing at an intronic splice site (proximal site), and RSZ33 increased the transcripts at an exon-terminal site (distal site) (Lopato et al., 1999b; Kalyna et al., 2003). Therefore, the regulation of splice site selection is operated by a complex regulatory mechanism.

Other types of regulation, such as the concentration of splicing factors, phosphorylation of interacting domains, and competition with inhibitor proteins, are involved in alternative splicing. Recently, the dynamics of two Arabidopsis SR proteins, RSp31 and RSZp22, which are homologous to rice RSp29 and RSZp23 (see Supplemental Figure 1 online), was investigated using GFP fusion proteins (Tillemans et al., 2005). These GFP fusion proteins colocalized in the same nuclear territories in plant cells. Treatment with a kinase inhibitor caused the RSZp22 protein to concentrate in the nucleolus, while RSp31 and RSZp22 still colocalized within the nucleoplasm. Phosphorylation of the RS domain of RSZp22 plays a role in shuttling between the nucleoplasm and nucleolus (Tillemans et al., 2005). If two rice SR proteins, RSp29 and RSZp23, are also regulated by phosphorylation of the RS domain, this regulation might influence their difference in the 5' splice site selection of rice Wx^b intron 1.

Functions of the RRM Domain and the RS Domain
The RNA binding domain recognizes specific RNA sequences. These data were obtained by binding assays with enhancer sequences on interacting mRNA or in vitro selection, called SELEX (Graveley, 2000; Cartegni et al., 2002; Maniatis and Tasic, 2002). Although each SR protein recognizes a distinct enhancer sequence, the sequences identified as binding sites for one SR protein can be recognized by other SR proteins (Chandler et al., 1997; Liu et al., 1998). RSp29 could increase the splicing efficiency of the Wx^b intron 1, but RSp33, which belongs to the same family as RSp29, did not affect its splicing (Figure 5C). As compared with the RRMs of RSp29 and those of RSp33, the RNP-1 and RNP-2 submotifs in those two RRMs are similar except for three amino acids (see Supplemental Figure 1 online). Therefore, our results suggest that the splicing activity of RSp29 for Wx^b pre-mRNA is tightly determined by the specificity of its RRMs on RSp29. The exon is recognized by the binding of U1 and U2 snRNPs in the early step of pre-mRNA recognition. SR proteins bound to the ESE have been proposed to promote this exon recognition step by interacting with U2AF³⁵ bound to the upstream 3' splice site and U1-70K bound to the downstream 5' splice site. The results in Figure 7 suggest the possibility that the RSp29 and RSZp23 bound to specific sequences in exon 1 interact with U1-70K and activate the splicing of Wx^b intron 1. It was reported that some plant SR proteins interact with U1-70K (Golovkin and Reddy, 1998, 1999; Lopato et al., 2002; Lorkovic et al., 2004).

In general, it is accepted that the RS domain of SR proteins functions as a protein–protein interacting domain. The phosphorylation of the RS domain influences the localization of SR proteins and interaction with another splicing factor involving SR proteins. In Arabidopsis, the shuttling of RSZp22 in the nucleus changed after treatment with a kinase inhibitor (Tillemans et al., 2005). Phosphorylated SCL33 by AFC2, a Clk/Sty protein kinase, could interact with AFC2 (Golovkin and Reddy, 1999). The interaction of SRp34/SR1 with CypRS64, a cyclophilin isolated in yeast two-hybrid screening with SRp30, is phosphorylation dependent (Lorkovic et al., 2004). Moreover, the deletion of the RS domain of GFP fusion SR proteins failed to exhibit the typical speckled pattern in the nucleus (Tillemans et al., 2005). We showed that RSp29 without the RS domain could not increase the GUS activity of Wx^b-gus. However, the RS domain of SCL26 activated the splicing of Wx^b intron 1 when exchanged with that of RSp29 (Figure 6). These results demonstrate that the RS domain is required for the splicing regulation in rice as well as in animals.

Alternative splicing can generate various protein molecules from a single gene. For example, RSZ36 mRNA3 (Figure 8) could encode a protein of only 47 amino acids having no RNP-1 motif; therefore, it cannot function as an RNA binding protein. On the other hand, SRp33a has two transcripts by alternative splicing on intron 11: one encoding a full-length protein and the other a truncated protein lacking 50 amino acids on the C terminus (data not shown). The truncated SRp33a protein has an RS domain of only 60 amino acids. However, since only 10 RS repeats can function as an RS domain (Zhu and Krainer, 2000; Cazalla et al., 2002), the truncated SRp33a may be functionally equal to the full-length SRp33a. These results suggest that alternative splicing products of SR pre-mRNA may not greatly affect the splicing machinery other than by decreasing the amount of the full-length protein. In Arabidopsis, truncated SR proteins similar to SRp33a and RSZ36 have been reported (Lopato et al., 1999a, 2002). However, Kalyna et al. (2003) reported a minor truncated SR protein carrying a single RRM without other domains. This protein may function as a negative regulator by binding to specific sequences. In rice, such a truncated form of the SR protein has been found in RSZp21b transcripts.

Alternative Splicing in Plants
It has been reported that pre-mRNAs producing various transcripts by alternative splicing were 2.8% of the analyzed 18,933 transcription units in rice (Rice Full-Length cDNA Consortium, 2003). The alternative splicing is regulated by tissue specificity or external stimulations and results in the creation of protein diversity from the genome. Alternatively, spliced transcripts have been described for many plant genes, but their functions are unclear.

The tobacco N gene, which encodes an NBS-LRR–type disease resistance gene, has two transcripts by alternative splicing (Whitham et al., 1994). In resistance reaction, the mRNA encoding the full-length protein was more prevalent for 3 h after virus infection, but another transcript encoding a truncated protein was more abundant 4 to 8 h after infection. Moreover, the results of analysis using transgenic plants showed that both transcripts were necessary to confer complete resistance (Dinesh-Kumar and Baker, 2000). In addition, many NBS-LRR–type plant disease resistance genes show alternative splicing (Jordan et al., 2002). Although disease resistance is an important characteristic of plants, the regulatory mechanisms or factors involved in the alternative splicing of RNA derived from these resistance genes are not known.

To test changes in the splicing patterns of plant endogenous genes by SR proteins, we used transgenic rice plants overexpressing SR protein genes. Overexpression of RSZ36 and SRp33b in transgenic rice changed the alternative splicing patterns of their own pre-mRNAs or those of other SR proteins. In Arabidopsis, overexpression of SRp30 and RSZ33 also influences the alternative splicing patterns of their own pre-mRNAs and those of splicing factors, including SR proteins (Lopato et al., 1999b; Kalyna et al., 2003). However, no changes in alternative splicing of other endogenous genes were found in transgenic rice or Arabidopsis so far. In the transient assays using isolated rice protoplasts, however, we were able to detect the change in 5' splice sites of Wx^b-gus transcripts by addition of rice SR proteins. Therefore, SR proteins are able to change splice site selection in rice cells. These results may suggest that in vivo alternative splicing in intact plant tissues may be regulated by complex systems, including many SR proteins with redundant functions, so that it is not easily altered by overexpression of a single SR protein. In any case, we need to analyze many more alternatively spliced transcripts in SR protein–overexpressing plants to gain insights into regulation of alternative splicing in plants.

We analyzed the function of rice SR proteins using the Wx^b gene. This is an agriculturally important gene that affects the eating quality of cooked rice. Two mutants, du1 and du2, are known to decrease the splicing of Wx^b pre-mRNA (Isshiki et al., 2000). RSp29 and RSZp23 were shown to increase the splicing efficiency of Wx^b pre-mRNA. Therefore, we sequenced the RSp29 gene in these mutants, but we found no alterations in their sequences. The alternative splicing pattern of Wx^b mRNA is also affected by the temperature (Larkin and Park, 1999), resulting in poor quality of rice grains because of the high amylose content when seed maturation occurs at low temperature. If RSp29 and RSZp23 function can be suppressed at low temperature, the splicing efficiency of Wx^b will decrease, resulting in the production of rice grains with less amylose even in a cool summer.

	METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
cDNA Clone
We identified SR protein cDNAs (SRp32 and SRp33a) by screening the cDNA library. EST clones (SCL25, SCL26, RSZp23, RSp29, RSp33, RSZ36, and RSZ37a) were obtained from the DNA bank at the National Institute of Agrobiological Sciences (NIAS), and RT-PCR amplification was based on information from the Rice Full-Length cDNA Database (http://cdna01.dna.affrc.go.jp/cDNA/).

RT-PCR and Sequencing
Total RNA was extracted from root, leaf blade, immature seeds, and cell culture (Chomczynski and Sacchi, 1987). Total RNA from transformed Oc cell lines was extracted using EASYPrep RNA (Takara) to minimize contamination with plasmid DNA. One microgram of DNase-treated total RNA from different organs was used to synthesize first-strand cDNA with an oligo(dT) primer in a reaction volume of 20 µL using SuperScript II transcriptase (Invitrogen). PCR amplifications were performed using each SR protein cDNA-specific primer (see Supplemental Table 2 online) or Wx and gus primers (for spliced transcripts, Wx1F 5'-ACCATTCCTTCAGTTCTTTG-3' and GUS323R 5'-TGATGCTCCATCACTTCCTG-3'; for unspliced transcript, Wx1001F 5'-TGCTCCTTAAGTCCTTATAAGCAC-3' and GUS323R) to analyze transformed cell lines in a final volume of 20 µL using Ex Taq (Takara) or Blend Taq (Toyobo). Both unspliced and spliced Wx transcripts could be amplified using Wx1F and GUS323R primers in the same reaction. However, it was difficult to obtain accurate results of unspliced products because they are much longer than spliced transcripts. This affected the amplification efficiency during PCR. Therefore, we used separate PCR analysis for spliced and unspliced RNAs. PCR cycle conditions for SR protein genes were followed by optional cycles (25 to 35) of 98°C for 15 s, 55°C for 30 s, and 72°C for 1.5 min. To detect the amplified products from transformed Oc cell lines, PCR cycle conditions were followed by 42 cycles of 94°C for 30 s, 57°C for 30 s, and 72°C for 30 s. RT-PCR products of the SR protein and Wx^b-gus transcripts were cloned into a pGEM-T vector (Promega), and the sequence was analyzed using ABI BigDye terminator and the ABI Prism 31 genetic analyzer (Applied Biosystems).

Plasmid Construction
The 35S-SR gene carries the coding region of each SR protein provided by the DNA bank at NIAS, Japan, ligated on a pSN221 vector. 35S-SR had the following structure, with the restriction sites used for ligation: 35S promoter-SalI-SR protein cDNA-NotI-NOS terminator.

We used the following primers for the construction of chimeric proteins: 1, TCTAGAGTCGACATGGGAAGAGGCTATG; 2, 5'-AGAAAGAAACCATCTGAGCCTGTGAATACTAAACCC-3'; 3, 5'-GGGTTTAGTATTCACAGGCTCAGATGGTTTCTTTCT-3'; 4, 5'-CAGCTCGCGGCCGCTCAATCATAAGCTCTTCTG-3'; 5, 5'-TCTAGAGTCGACATGAGGCCCGTGTTCGTGGG-3'; 6, 5'-TTCAGGGATGATACAGAAATGAGGGCTAGGGAAAG-3'; 7, 5'-CTTTCCCTAGCCCTCATTTCTGTATCATCCCTGAA-3'; 8, 5'-CAGCTCGCGGCCGCTCACCGGCTAACAGAAAG-3'; 9, 5'-GATGGCTATAGCAAACCACGAGGTGACAGGTATGAT-3'; 10, 5'-ATCATACCTGTCACCTCGTGGTTTGCTATAGCCATC-3'; 11, 5'-CAGCTCGCGGCCGCTATTCTGTATCATCCCTGAAG-3'.

DS1
SCL26 and RSp29 were independently amplified with primer pairs 1/3 and 2/4 from 35S-SCL26 and 35S-RSp29, respectively. The two PCR products were used for a second PCR amplification with primers 1 and 4. The resulting PCR product was digested with SalI and NotI and subcloned into the corresponding sites of the pSN221 vector.

DS2
SCL26 and RSp29 were independently amplified with primer pairs 5/7 and 6/8 from 35S-SCL26 and 35S-RSp29, respectively. The two PCR products were used for a second PCR amplification with primers 5 and 8. The resulting PCR product was digested with SalI and NotI and subcloned into the corresponding sites of the pSN221 vector.

DS3
SCL26 and chimeric RRMs were independently amplified with primer pairs 6/8 and 1/7 from 35S-SCL26 and DS1, respectively. The two PCR products were used for a second PCR amplification with primers 1 and 8. The resulting PCR product was digested with SalI and NotI and subcloned into the corresponding sites of the pSN221 vector.

RRM2
RSp29 was independently amplified with primer pairs 4/9 and 5/10 from 35S-RSp29. The two PCR products were used for a second PCR amplification with primers 4 and 5. The resulting PCR product was digested with SalI and NotI and subcloned into the corresponding sites of the pSN221 vector.

RS
RSp29 was amplified with primer pairs 5 and 11 from 35S-RSp29. The resulting PCR product was digested with SalI and NotI and subcloned into the corresponding sites of the pSN221 vector.

To produce Wx^b(mutA)-gus and Wx^b(mutB)-gus, Wx^b-gus DNA was mutagenized using one of two mutagenic antisense primers, each of which created the mutations indicated in Figure 7B.

Transient Assays in Rice Protoplasts
Protoplasts (5 x 10⁶) were prepared from the rice (Oryza sativa) Oc suspension cultures, mixed with 10 µg of gus plasmid DNAs and 2 µg or various volumes of 35S-SR plasmid DNAs, and electroporated using a Gene Pulser (Bio-Rad Laboratories). After incubating overnight twice at 30°C, cells were assayed for GUS activity. RNA was isolated from protoplasts for RT-PCR.

Transgenic Rice
To overexpress various SR protein genes in rice cv Kinmaze, the coding regions of SR protein cDNA were cloned in pMSH1 having the cauliflower mosaic virus 35S promoter. This construct was then mobilized into Agrobacterium tumefaciens strain EHA101 by electroporation. Calli induced from seeds of cv Kinmaze were cocultivated with A. tumefaciens, and transgenic calli were selected by hygromycin. Plants were regenerated by transferring the resistant calli to a plant regeneration medium (Hiei et al., 1994).

Phylogenetic Analysis
Phylogenetic analysis of rice SR proteins was based on putative full-length protein sequences. The amino acid sequences of rice, Arabidopsis thaliana, and human SR proteins were aligned with the ClustalW program (http://www.ddbj.nig.ac.jp/search/clustalw-j.html). The phylogenetic tree was displayed with TreeView (http://taxonomy.zoology.gla.ac.uk/rod/treeview.html). Bootstrap values with 1000 repetitions were indicated at each branch point.

Accession Numbers
Sequence data from this article can be found in the GenBank/EMBL data libraries under accession numbers AK061903 (SRp32), AK106176 (SRp33a), AK071503 (SRp33b), AK103534 (SR20), AK103676 (SC35a), AK070292 (SC35b), AK073057 (SC35c), AK073451 (SCL25), AK062577 (SCL30a), AK065531 (SCL30b), AK063879 (RSZp21a), AK073210 (RSZp21b), AK060088 (RSZp23), AK102071 (RSp29), AK121372 (RSp33), AK103897 (RSZ36), AK060171 (RSZ37a), AK073348 (RSZ37b), and AK063518 (RSZ39).

Supplemental Data
The following materials are available in the online version of this article.

Supplemental Table 1. Primers Used for RT-PCR Analysis of Splicing Patterns in 38 Rice Genes.

Supplemental Table 2. Primers Used for RT-PCR Analysis of SR Protein Transcripts.

Supplemental Figure 1. Protein Sequences of Rice SR Proteins.

Supplemental Figure 2. Phylogenetic Tree for SR Proteins.

	Acknowledgments

 
We thank Kunio Inoue for offering advice with our experiments; Shoshi Kikuchi and Hitomi Yamada for providing information on full-length cDNA; Ryuji Ishikawa for providing information of RMu1 transcripts; and Masako Kanda, Sawako Kohashi, and Tomoko Aoi for assistance with the experiments. This work was partially supported by the program "Functional analysis of genes relevant to agriculturally important traits in the rice genome (IP1012)" from the Ministry of Agriculture, Forestry, and Fisheries of Japan, and the "Academic Frontier" Project for Private Universities: matching fund subsidy from Ministry of Education, Culture, Sports, Science, and Technology of Japan, 2004–2008.

	Footnotes

 
The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantcell.org) is: Ko Shimamoto (simamoto{at}bs.naist.jp).

^[W] Online version contains Web-only data.

Article, publication date, and citation information can be found at www.plantcell.org/cgi/doi/10.1105/tpc.105.037069.

Received August 13, 2005; Revision received October 12, 2005. accepted November 9, 2005.

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Ali, G.S., Golovkin, M., and Reddy, A.S. (2003). Nuclear localization and in vivo dynamics of a plant-specific serine/arginine-rich protein. Plant J. 36, 883–893.[CrossRef][ISI][Medline]

Black, D.L. (2003). Mechanisms of alternative pre-messenger RNA splicing. Annu. Rev. Biochem. 72, 291–336.[CrossRef][ISI][Medline]

Caceres, J.F., Screaton, G.R., and Krainer, A.R. (1998). A specific subset of SR proteins shuttles continuously between the nucleus and the cytoplasm. Genes Dev. 12, 55–66.[Abstract/Free Full Text]

Cai, X.L., Wang, Z.Y., Xing, Y.Y., Zhang, J.L., and Hong, M.M. (1998). Aberrant splicing of intron 1 leads to the heterogeneous 5' UTR and decreased expression of waxy gene in rice cultivars of intermediate amylose content. Plant J. 14, 459–465.[CrossRef][ISI][Medline]

Cartegni, L., Chew, S.L., and Krainer, A.R. (2002). Listening to silence and understanding nonsense: Exonic mutations that affect splicing. Nat. Rev. Genet. 3, 285–298.[CrossRef][ISI][Medline]

Cazalla, D., Zhu, J., Manche, L., Huber, E., Krainer, A.R., and Caceres, J.F. (2002). Nuclear export and retention signals in the RS domain of SR proteins. Mol. Cell. Biol. 22, 6871–6882.[Abstract/Free Full Text]

Chandler, S.D., Mayeda, A., Yeakley, J.M., Krainer, A., and Fu, X.D. (1997). RNA splicing specificity determined by the coordinated action of RNA recognition motifs in SR proteins. Proc. Natl. Acad. Sci. USA 94, 3596–3601.[Abstract/Free Full Text]

Chomczynski, P., and Sacchi, N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156–159.[ISI][Medline]

Dinesh-Kumar, S.P., and Baker, B.J. (2000). Alternatively spliced N resistance gene transcripts: Their possible role in tobacco mosaic virus resistance. Proc. Natl. Acad. Sci. USA 97, 1908–1913.[Abstract/Free Full Text]

Docquier, S., Tillemans, V., Deltour, R., and Motte, P. (2004). Nuclear bodies and compartmentalization of pre-mRNA splicing factors in higher plants. Chromosoma 112, 255–266.[CrossRef][ISI][Medline]

Feng, Q., et al. (2002). Sequence and analysis of rice chromosome 4. Nature 420, 316–320.[CrossRef][Medline]

Frances, H., Bligh, J., Larkin, P.D., Roach, P.S., Jones, C.A., Fu, H., and Park, W.D. (1998). Use of alternate splice sites in granule-bound starch synthase mRNA from low-amylose rice varieties. Plant Mol. Biol. 38, 407–415.