Article

• The EMBO Journal (1998) 17, 6049 - 6060
• doi:10.1093/emboj/17.20.6049

Analysis of the functional specificity of RS domains in vivo

Brigitte Dauwalder¹ and William Mattox¹

1. Department of Molecular Genetics, University of Texas, MD Anderson Cancer Center, 1515 Holcombe Blvd, Box 45, Houston, TX 77030, USA

Correspondence to:

William Mattox, E-mail: wmattox@mdacc.tmc.edu

Received 9 April 1998; Accepted 14 August 1998; Revised 13 August 1998

• Keywords:

  • alternative splicing,
  • Drosophila sex determination,
  • nuclear localization,
  • spermatogenesis,
  • SR proteins

Introduction

Alternative pre-mRNA splicing is an important regulatory mechanism that is widely used among eukaryotic genes, but relatively little is known about the factors responsible for tissue-specific regulation of splicing. Studies on several alternatively spliced pre-mRNAs have demonstrated that proteins from the SR superfamily of splicing factors play important roles in controlling the recognition of specific splice sites (Fu, 1995; Manley and Tacke, 1996). During splicing of constitutive introns, several SR proteins are known to facilitate both the initial recognition of splice sites and the assembly of pre-spliceosomal complexes (Ruskin et al., 1988; Kohtz et al., 1994; Staknis and Reed, 1994). In Drosophila, SR proteins are known to affect alternative splicing when they form developmentally specific splicing enhancer complexes with TRANSFORMER (TRA) and TRANSFORMER-2 (TRA-2), two regulatory proteins that are themselves closely related to the SR proteins. These complexes form on splicing enhancers in pre-mRNAs from the sex determination gene doublesex (dsx) and direct its sex-specific processing (Tian and Maniatis, 1993; Lynch and Maniatis, 1995). In vitro studies indicate that TRA and TRA-2 directly facilitate binding of the SR protein RBP1 as well as several other SR proteins to specific regions of the dsx pre-mRNA (Lynch and Maniatis, 1996). These complexes function to recruit the U2AF protein and other components of the general splicing machinery to the female-specific branchpoint/polypyrimidine-tract region of the dsx RNA (Zuo and Maniatis, 1996). The local recruitment of these factors favors use of the female-specific dsx 3' splice site to form an mRNA encoding the female-specific dsx protein. In the absence of TRA or TRA-2, an alternative 3' splice site is used by default, giving rise to mRNA encoding the male specific dsx protein (Nagoshi et al., 1988; Ryner and Baker, 1991).

Both SR proteins and a number of related splicing regulators contain regions rich in arginine and serine (RS domains) that are necessary for interactions between these factors (Wu and Maniatis, 1993; Kohtz et al., 1994). However, it has remained unclear whether the RS domains from individual proteins differ significantly in function or whether they play a generic, non-specific role. In vitro studies have suggested that RS domains can be interchanged between splicing factors or replaced by artificially synthesized RS rich sequences without substantial effects on protein function (Valcarcel et al., 1996; Chandler et al., 1997b). While this suggests that RS domains play a generic role, only a few studies have yet tested how RS domain replacements affect protein function in vivo (Li and Bingham, 1991; Caceres et al., 1997).

The Drosophila sex determination system provides an opportunity to study the contribution of RS domains to the function of splicing regulators in vivo. Because the interactions of TRA, TRA-2 and SR proteins are required for formation of a functional splicing enhancer complex on dsx pre-mRNA, mutations affecting these interactions are predicted to result in both a switch in the dsx splicing pattern and in dramatic changes in the sexual phenotypes of adult flies. Here we have used this system to examine whether functional specificity is conferred by the essential RS domain of the TRA-2 protein. TRA-2 is of particular interest because it functions as a specific regulator in dsx splicing, has no effect on viability and is organized in a manner that closely resembles core SR proteins (Amrein et al., 1988; Goralski et al., 1989). Moreover, in addition to dsx pre-mRNA, TRA-2 also affects alternative splicing of several other specific pre-mRNAs. For instance, in the nervous system TRA-2 is required for sex-specific splicing of fruitless (fru) pre-mRNA (Ito et al., 1996; Ryner et al., 1996; Heinrichs et al., 1998), while in the male germline it affects sex-specific processing of pre-mRNAs from exuperantia (Hazelrigg and Tu, 1994), att (Madigan et al., 1996) and the tra-2 gene itself (Mattox and Baker, 1991; Mattox et al., 1996). This system thus offers the opportunity to study the effects of RS domains on regulation of multiple pre-mRNA targets in vivo.

To assess the ability of RS domains from other proteins to replace the TRA-2 RS2 domain we have generated transgenic fly strains expressing fusion proteins. We find that while some RS domains are functionally interchangeable, others confer only partial or target-specific function. These data argue that RS domains found in different SR proteins are not all functionally equivalent to the TRA-2 RS2 domain in vivo.

In order to determine whether the RS domains from other RS proteins can substitute for the TRA-2 RS2 domain, we first deleted RS2 [amino acids (aa) 193–264] to create TRA-2RS2. The RS domains of the Drosophila TRA, SRp55 (dSRp55), U2AF⁵⁰ (dU2AF⁵⁰) and human SC35 proteins were then inserted in place of RS2 to generate the fusion genes TRA-2/TRA RS, TRA-2/dSRp55 RS, TRA-2/dU2AF⁵⁰ RS and TRA-2/SC35 RS (Figure 1). In addition, we also inserted the C-terminal part of the TRA-2 RS2 domain which has been shown to be more conserved between Drosophila melanogaster and Drosophila virilis than other parts of the RS2 domain (Chandler et al., 1997a), generating TRA-2/RS_carb. All changes were made in the context of the genomic sequences containing the native tra-2 promoter, introns and flanking sequences. The constructs were inserted into the CaSpeR Drosophila transformation vector and used to generate transgenic flies by P-element mediated germline transformation of the tra-2 mutant strain w¹¹¹⁸/B^SY; tra-2^B/CyO. The transgenic lines were examined to determine the effects of the transgenes on sexual differentiation and RNA splicing patterns in tra-2 mutant backgrounds.

Figure 1.

Structure of the fusion constructs. The organization of the tra-2 genomic fragment used to generate transgenic flies is shown in (A). Exons are indicated by open boxes. The autoregulated M1 intron is shown by a horizontal line, all other introns are indicated by blank regions between the exons. The gray areas in exons 6 and 7 indicate the coding region for RS2. ATG start codons for the three different protein isoforms and the stop codon used by all isoforms are shown. The M1 intron splits the ATG start codon for the TRA-2²²⁶ isoform. For N-FLAG tagged proteins, sequences coding for the FLAG epitope were inserted five amino acids downstream of the ATG for the TRA-2²²⁶ isoform. C-terminally tagged proteins were fused to lacZ or to FLAG sequences immediately upstream of the stop codon. TRA-2 protein isoforms and location of epitope tags are shown in (B). The gray areas labeled RS1 and RS2 indicate RS domains. The black area labeled RRM indicates the RNA binding domain and the open box labeled L indicates the linker region. Note that the N-terminal FLAG epitope labels the TRA-2²⁶⁴ and TRA-2²²⁶ isoforms only, whereas the C-terminal tags label all three protein isoforms. RS2 was deleted in TRA-2RS. In fusion proteins RS2 (aa 193–264) was precisely replaced by RS domains of other proteins.

View full figure (38 KB)

Epitope-tag fusions at the C-terminus, but not the N-terminus of the TRA-2 proteins specifically disrupt male germline functions

To follow the expression of RS domain fusion proteins, we added sequences coding for foreign epitopes to them. To identify an epitope tagging strategy that would allow detection of the expressed proteins while not affecting their function in vivo, we initially tested epitopes fused at both the N- and C-terminus of the full-length TRA-2 protein. In two of these, the FLAG epitope tag (9 aa) was fused either at the C-terminus (C-FLAG) or just inside the N-terminus (N-FLAG) of the TRA-2²²⁶ isoform. In the third construct the Escherichia coli -galactosidase protein was fused to the C-terminus of TRA-2 (C-lacZ). It should be noted that the C-terminal epitope fusions mark all three TRA-2 isoforms, while the N-terminal epitope tag labels only the two major isoforms TRA-2²⁶⁴ and TRA-2²²⁶, omitting a minor isoform TRA-2¹⁷⁹ (Figure 1).

All three epitope fusion constructs were able to rescue somatic sex determination functions when introduced into tra-2 mutant flies as judged by several cuticular features that are sexually dimorphic in males and females (described in detail below). When germline functions were examined, however, we found that only N-FLAG restored male fertility to normal levels (Table I). The fact that both C-FLAG and C-lacZ failed to function only in the germline demonstrates that it is possible to inactivate the germline function of the gene without inactivating somatic function. This difference is unlikely to be due to the presence of the untagged TRA-2¹⁷⁹ present in N-FLAG bearing flies since this isoform has previously been shown to be neither necessary nor sufficient for germline function (Mattox et al., 1996). Although tra-2 transgenes with all three types of epitope tags conferred rescue of dsx splicing, the C-terminal FLAG epitopes could not be detected in immunostaining experiments. We therefore used both the N-FLAG and C-lacZ fusions for the analysis of somatic functions. In the germline we focused on N-FLAG fusion proteins for the analysis of male germline function.

Table 1: Comparison of phenotypic rescue using different TRA-2 epitope fusions

View full table

RS domains from other SR proteins can substitute for the TRA-2 RS2 domain in somatic female sexual differentiation

Chromosomally female (XX) individuals lacking TRA-2 function are sexually transformed in somatic tissues and develop into adults that are identical in appearance to males (Watanabe, 1975). Our initial assay for the function of the different transgenes was to determine whether they could rescue this somatic phenotype. Several visible cuticular features that are sexually dimorphic in males and females were examined for each line. These include abdominal pigmentation, the number and position of sex comb bristles (a group of bristles on the basitarsus of the first leg), the number of sixth sternite bristles and the genitalia. For each fusion construct we generated a number of independent transgenic lines to control for variations due to the insertion site of the transposon. The phenotypic analysis for all lines of XX; tra-2 flies with one dose of the transgene is summarized in Table II. As expected, most TRA-2RS transgenic lines do not show rescue of the tra-2 phenotype and are still male in appearance. In a few lines we observed a very slight rotation of the otherwise male genitalia which is an indicator of intersexuality. This suggests a small amount of residual function in these lines. The full-length TRA-2 constructs that were used as a positive control provided complete rescue in a majority of the lines examined, resulting in XX; tra-2 animals with wild-type female appearance. Among the RS fusion constructs tested, the most consistent strong rescue to fully differentiated females was observed with TRA-2/dU2AF⁵⁰ RS. Intermediate rescue producing mostly female-like intersexes was observed for both TRA-2/TRA RS and TRA-2/SC35 RS, suggesting that these RS domains function less well than RS2. Although the above results are qualitatively the same for constructs tagged with the FLAG epitopes and with the lacZ transgene, variable results were observed for TRA-2/dSRp55 RS fusions which did not confer rescue in the FLAG-tagged constructs, but did give rise to somatically rescued flies as a lacZ fusion gene. The cause for this discrepancy is unknown, but it has previously been observed that lacZ fusions stabilize some proteins during Drosophila embryogenesis and it is possible that this is responsible for the observed difference. Rescue similar to that of the full-length construct was observed with the TRA-2/RS_carb transgene which contains only the C-terminal part of the TRA-2 RS2 domain. A few lines (from otherwise rescuing constructs) that did not rescue most likely reflect insertions of the P-element into chromosomal positions where the gene is not sufficiently expressed at the appropriate time in sexual development.

Table 2: Somatic sexual differentiation in transgenic lines

View full table

Two representative lines carrying each construct were chosen for detailed phenotypic analysis, as summarized in Table III and Figure 2. This confirmed the above conclusions and revealed that individual external characteristics follow the same general trend. For example, both lines carrying TRA-2/dU2AF⁵⁰ RS rescued all aspects of cuticular differentiation examined, including abdominal pigmentation, number of sternite bristles, the number of sex combs and the morphology of genitalia. Intermediate rescue (as in the case of TRA-2/SC35 RS and TRA-2/TRA RS) is evidenced by intermediate morphology of sex combs, number of sternite bristles, pigmentation and genital morphology. Together, the above results indicate that RS domains from different SR proteins substitute for TRA-2 RS2 to varying degrees.

Figure 2.

Differentiation of sex combs in transgenic lines. Male sex combs and analogous female bristles from the basitarsus of the first leg are shown in various control and transgenic fly strains. The genotype for all transgenic lines is (XX) w¹¹¹⁸/w¹¹¹⁸; tra-2^B/tra-2^B. The identity of the RS domains replacing the TRA-2 RS2 domain is indicated. Unless otherwise indicated, fusion proteins in the lines shown were epitope tagged at the N-terminus with FLAG.

View full figure (71 KB)

Table 3: Detailed phenotypic analysis of chromosomally female tra-2 individuals from representative transgenic lines

View full table

TRA-2 RS domain substitutions can regulate dsx and fru splicing

Female sex determination requires TRA-2-dependent female specific splicing of the dsx pre-mRNA. The observed somatic phenotypes should therefore be reflected in the ability of the transgenes to direct female specific splicing of the dsx mRNA. We used RT–PCR on RNA from XX; tra-2 animals carrying one dose of the transgenes to examine splicing of the dsx RNA (Figure 3). Each lane corresponds to RNA from a representative independent line. Figure 3, lanes 1 and 2, shows the RT–PCR products from control wild-type females and males, respectively. As shown in Figure 3, lane 3, XX; tra-2 mutant animals produce predominantly the male specific splice form. Figure 3, lanes 4 to 21, shows the dsx splicing patterns in tra-2 mutant animals carrying the different transgenes. TRA-2RS lines produce predominantly male RNA (Figure 3, lanes 7–9) and a smaller amount of the dsx female specific transcript, consistent with the relatively small amount of female differentiation that occurs in these flies. TRA-2/dU2AF⁵⁰ RS fusions (Figure 3, lanes 17–19) and TRA-2/RS_carb (Figure 3, lanes 21–24) produce exclusively the female specific splice form, indistinguishable from the control transgene with the native TRA-2 RS2 domain (Figure 3, lanes 4–6). Fusions that give intermediate phenotypic rescue show both male- and female-specific dsx transcripts (TRA-2/SC35 RS, TRA-2/TRA RS). Splicing in TRA-2/dSRp55 RS varied again depending on the epitope used. In all the RS fusions, the amount of female specific dsx splicing corresponds well with the phenotypic sexual development we observed; we find the expected direct correlation between the amount of female specific dsx splicing and the rescue of female morphology in the different transgenic lines.

Figure 3.

Alternative splicing of dsx pre-mRNAs in transgenic lines. RT–PCR was performed using primers specifically amplifying male- and female-specific dsx mRNAs. The PCR products were visualized by Southern blot hybridization with primers that span the male- and female-specific splice junctions. The genotype of control flies in lanes 1–3 is as indicated, the genotype for all transgenic lines is XX; tra-2 (lanes 4–25). The identity of the RS domains replacing the TRA-2 RS2 domain is as indicated. Proteins were epitope tagged at the C-terminus (C) with -gal or at the N-terminus (N) with FLAG. The identity of individual transgenic lines assayed is indicated by their line number.

View full figure (35 KB)

Female-specific, alternative splicing of the fru pre-mRNA in parts of the nervous system has been shown to be TRA-2 dependent. We therefore used a subset of the same lines to examine how deletion of the RS2 domain and its replacement with other RS domains affected fru splicing. RNA was isolated from XX; tra-2 animals carrying the transgenes and analyzed by RT–PCR (Figure 4). fru mRNA from XX; tra-2/Df(tra-2) mutant animals undergoes male specific splicing that can be reversed completely to female by full-length TRA-2 transgenes with an epitope tag either at the C-terminus (C) or N-terminus (N). Transgenes lacking the RS2 domain failed to rescue fru splicing. TRA-2/dU2AF⁵⁰ RS on the other hand directed female-specific splicing, indistinguishable from the full-length TRA-2 protein, whereas TRA-2/SC35 RS gave rise to both male- and female-specific splice products. These results parallel our findings on dsx splicing, showing that the RS2 domain of TRA-2 is required for fru splicing in the nervous system and can be functionally replaced by RS domains from other proteins.

Figure 4.

Alternative splicing of fru pre-mRNA in transgenic lines. RT–PCR was performed using primers that specifically amplify male- and female-specific fru mRNAs. PCR products were visualized by Southern blot hybridization with sex specific fru probes generated by PCR using a second set of primers that amplify regions of the male and female fru exon. The genotypes of flies and the identity of the RS domains replacing the TRA-2 RS2 domain are indicated. Proteins were epitope tagged at the C-terminus with -gal or at the N-terminus with FLAG. The identity of individual transgenic lines assayed is indicated by their line number below the lanes. RT–PCR reactions were performed in either the presence (+) or absence (-) of reverse transcriptase (RT).

View full figure (46 KB)

In summary, while our results demonstrate that there are differences in the extent of somatic rescue with various RS substitutions, we find that in general the RS domains tested were able to at least partially substitute for RS2. At least one RS domain (that from dU2AF⁵⁰) substituted fully for the TRA-2 RS2 domain.

The fusion proteins function differently in the male germline

In addition to its role in somatic sex determination, TRA-2 affects splicing of multiple targets in the male germline and is required for male fertility. We therefore examined the ability of various transgenes to rescue male fertility in tra-2 mutant males. Table IV summarizes the results obtained for all transgenic N-FLAG lines, testing three to five males from each line. These results show that TRA-2/dU2AF⁵⁰ RS and TRA-2/RS_carb frequently restored fertility whereas TRA-2/SC35 RS and TRA-2/dSRp55 RS did so only in a few lines. Comparison of fertile lines from various N-FLAG constructs indicated that they differed in the frequency at which males were fertile whereas the number of progeny was the same for all fertile males (data not shown). A large number of unfertilized eggs was observed in all vials with infertile males, indicating that the males had mated successfully. These results suggest that there are large differences in the ability of various RS fusions to perform male germline functions.

Table 4: Male fertility in transgenic lines

View full table

One of the targets of the TRA-2 protein in the male germline is the tra-2 pre-mRNA itself. As a part of autoregulation, TRA-2 functions to repress splicing of the M1 intron. Accordingly, the appearance of M1-containing tra-2 RNA is indicative of the presence of functional TRA-2 protein able to regulate M1 splicing. In order to examine TRA-2 autoregulation in the transgenic lines RT–PCR was performed on RNA from XY; tra-2^B mutant males carrying the various transgenes (Figure 5). Since the endogenous tra-2^B allele results from a nonsense mutation it still produces RNA. Our assay detects transcripts from both the endogenous gene and the transgene. Figure 5, lane 1, shows products from the two RNA forms present in the wild type. In tra-2 mutants, the upper band representing RNA with the retained M1 intron is almost completely missing. The very low residual amount derives from unregulated residual retained intron in tra-2 mRNA precursors found in the soma. TRA-2RS constructs were not capable of providing autoregulatory function (Figure 5, lanes 13 and 15). Addition of the RS domain of dU2AF⁵⁰ (Figure 5, lanes 17 and 19) restored splicing repression in one of the two lines examined. The TRA-2/dSRp55 RS (Figure 5, lanes 25 and 27) trangene also partially restored splicing repression in at least one of the lines tested. TRA-2/SC35 RS (Figure 5, lanes 29 and 31) did not produce retained M1 intron significantly above background in either line tested. Although there is clear line-to-line variation in autoregulatory function, these results are roughly consistent with the effects of the transgenes on male fertility and suggest that some but not all RS domain substitutions function in the male germline.

Figure 5.

Alternative splicing of the tra-2 M1 intron in the germline of transgenic males. RT–PCR was performed to amplify the M1 intron and adjacent sequences and products were visualized by Southern blot analysis. PCR fragments originating from tra-2 RNAs with spliced or retained (unspliced) M1 intron sequences are indicated. The identity of the RS domains replacing the TRA-2 RS2 domain are indicated. Wild-type (wt) and tra-2 (mutant) controls are shown. Proteins were epitope tagged at the C-terminus with -gal or at the N-terminus with FLAG. The transgenic lines assayed are indicated by their line number. RT–PCRs were performed either in the presence (+) or absence (-) of reverse transcriptase (RT).

View full figure (83 KB)

Effects of RS domains on subnuclear localization of TRA-2 in primary spermatocytes

In previous studies it has been shown that Drosophila TRA-2 and other RS domain proteins localize in nuclear speckles when expressed in COS-1 and other mammalian cells. Our epitope-tagged versions of the TRA-2 protein allowed us to observe the distribution of the protein in fly cells and to examine the effect of changes in the RS domain on the localization of the protein. In primary spermatocytes, we found that the tagged TRA-2 proteins were highly localized to specific sites within the nucleus. These sites did not appear to be distributed randomly, but rather appeared to be associated with chromosomes in a pattern that changes dynamically as primary spermatocytes mature and bulk transcription ceases (B.Dauwalder and W.Mattox, in preparation). Figure 6A shows a schematic of the organization of chromosomes in primary spermatocytes. The two major autosomes and the X-chromosome are clearly distinguishable and situated close to the periphery of the nucleus (Cenci et al., 1994). Figure 6B and C shows the distribution of the wild-type TRA-2 protein in growth stage primary spermatocytes. The protein is highly localized at sites associated with chromosomes that appear to differ from the speckles seen in mammalian localization studies. In contrast, TRA-2RS protein is confined to the nucleus but fails to localize subnuclearly (Figure 6D and E). This suggests that RS2 is essential for subnuclear localization. We therefore examined whether the RS domains of other proteins were capable of restoring subnuclear localization. When substituted for RS2, the dSRp55 RS domain does not confer normal subnuclear localization, the protein is distributed homogeneously in the nucleus and rarely did we observe a slight concentration around the chromosomes (Figure 6F and G). Similar results were obtained when the C-lacZ version of TRA-2/dSRp55 was tested (data not shown). In contrast to the above, the dU2AF⁵⁰ and SC35 RS domains (Figure 6H–K) as well as RScarb (Figure 6L and M) restored the wild-type pattern of TRA-2 localization in all lines examined. We conclude from these experiments that RS domains can differ substantially in their ability to confer subnuclear localization.

Figure 6.

Subnuclear distribution of fusion proteins in primary spermatocytes. (A) A schematic of a primary spermatocyte nucleus, showing the three chromosome bivalents. (B–M) N-FLAG tagged proteins are detected with anti FLAG antibody (red), chromosomes are stained with Hoechst 33258 (blue). (B and C) Wild-type TRA-2; (D and E) TRA-2RS; (F and G) TRA-2/dSRp55 RS; (H and I) TRA-2/dU2AF⁵⁰ RS; (J and K) TRA-2/SC35 RS; (L and M) TRA-2 RS_carb. (N and O) Double staining of flies carrying both N-FLAG-TRA-2 [red, (N)] and TRA-2/dU2AF⁵⁰-lacZ [green, detected with anti--gal antibody, (O)]. (P and Q) Localization of endogenous dU2AF⁵⁰, detected with anti-dU2AF⁵⁰ antibody (red), chromosomes are stained with Hoechst. All transgenes are in a wild-type genetic background. The magnification in all panels is 630, except for (N) and (O), where it is 400.

View full figure (60 KB)

To examine whether the localization of the TRA-2 wild-type and the TRA-2/dU2AF⁵⁰ proteins is identical, we generated flies simultaneously carrying both the N-FLAG TRA-2 transgene and the TRA-2/dU2AF⁵⁰ -gal fusion transgene, and performed double stainings on spermatocytes. Figure 6N and O shows such a double staining experiment on spermatocytes. The two proteins colocalize, suggesting that the TRA-2/dU2AF⁵⁰ pattern is like that of the full-length TRA-2 protein. If the RS domain of dU2AF⁵⁰ is itself sufficient to confer this pattern we might expect that the dU2AF⁵⁰ protein localizes in a similar way. However, when we stained testes with an antibody against Drosophila dU2AF⁵⁰ (kindly provided by D.Rudner and D.Rio), we observed no subnuclear localization but rather homogeneous distribution in the nucleus (Figure 6P and Q). This is clearly different from the localization we observe in TRA-2/dU2AF⁵⁰ RS flies. Thus, the dU2AF⁵⁰ RS domain when fused to the TRA-2RS protein restores localization to the TRA-2-specific and not the dU2AF⁵⁰-specific pattern, suggesting that although a functional RS domain is necessary for TRA-2 protein localization, it is not alone sufficient for protein specific localization. This implies that other regions of the protein must also be required.

The inability of dSRp55 fusions to substitute for RS2 is not due simply to its length

It is possible that the differences in function observed above for different RS domains are related to their length. The RS domain of dU2AF⁵⁰ functions well and is the shortest of the foreign RS domains we have tested (31 aa compared with 69 aa in RS2), while that of dSRp55 functions poorly and is the largest of the domains tested (135 aa). We therefore generated additional fusions tagged with the N-terminal FLAG epitope to test whether shorter segments of dSRp55 were able to functionally substitute for RS2. Two non-overlapping subregions of sizes like the RS domains of dU2AF⁵⁰ and RS2 itself were tested. One of these subregions (fragment A) encompasses the first 69 aa of the RS domain of dSRp55 (residues 198–266) while the other subregion (fragment B) spans a 31 aa region downstream of fragment A (residues 268–298). As with other fusions, a number of independent transgene insertions were identified and tested for function in a tra-2 mutant background. Somatic sexual phenotype (Tables II and III), male fertility (Table IV), dsx splicing (Figure 3, lanes 21–24) and subnuclear localization (Figure 7) of both the fragment A and B protein fusions were similar to that observed with the original full-length TRA-2/dSRp55 fusion lines. These results demonstrate that RS regions of similar size deriving from SRp55 and other SR proteins differ in their ability to functionally replace RS2. This suggests that it is the specific sequences of these RS regions, rather than their size, that are responsible for the differences in function.

Figure 7.

Subnuclear distribution of fusions between TRA-2 and subsegments of the dSRp55 RS domain. Immunostaining of TRA-2 with anti-FLAG antibody (A, C and E) and DNA with Hoechst 33258 (B, D and F) is shown. The cells in (A) and (B) are from N-FLAG-TRA-2 males, (C) and (D) from N-FLAG TRA-2/Fragment A, and (E) and (F) from N-FLAG TRA-2/Fragment B.

View full figure (62 KB)

RS domains from other splicing factors differ in their ability to functionally replace TRA-2 RS2

RS domains are found in a large number of splicing factors and have been shown to be necessary for protein–protein interactions that facilitate formation of both spliceosomal and regulatory complexes (Wu and Maniatis, 1993; Amrein et al., 1994; Kohtz et al., 1994). While it is probable that interactions between SR proteins are highly specific in vivo, the contributions of RS domains to the functional specificity of individual SR proteins has remained unclear. To address this issue we have examined how replacement or deletion of an RS domain in the splicing regulator TRA-2 affects its defined genetic functions in transgenic fly strains.

There are two RS domains within the TRA-2 protein (Amrein et al., 1988; Goralski et al., 1989). RS1 is located near the N-terminus and RS2 near the C-terminus. In previous genetic studies we have shown that RS1 is essential for functions carried out by TRA2 in the male germline, but is less important for functions in somatic tissues (Mattox et al., 1996). In agreement with earlier studies (Amrein et al., 1994), we find that a deletion completely removing RS2 but not the upstream linker region almost completely abolishes TRA-2 somatic functions (dsx and fru splicing, and female sexual differentiation of cuticular structures) and eliminates all functions tested in the male germline (male fertility, autoregulation of TRA-2 splicing and subnuclear localization of TRA-2).

For somatic functions of TRA-2, the RS domains from several other SR proteins were able to at least partly replace RS2. This strongly indicates that, at least for some functions, the RS2 domain performs a generic role that can be fulfilled by other Arg–Ser rich sequences. However, it is important to note that RS domains were not equivalent in their ability to replace RS2. Variations were observed in somatic tissues, and more pronounced differences in function were found in the male germline where TRA-2 is required for the completion of normal spermatogenesis and male fertility. For instance, we found that fusion genes carrying the SC35 and dSRp55 RS domains functioned poorly in the germline as reflected by the fact that mutant males carrying these transgenes were mostly sterile. In contrast, substitution of the dU2AF⁵⁰ RS domain for RS2, resulted in a protein that functioned normally by all criteria examined. These results suggest that the TRA-2 RS2 domain and the dU2AF⁵⁰ RS domain share at least one functional capability that the SC35 and dSRp55 RS domains lack.

Studies on the RS domain of the human U2AF large subunit (U2AF⁶⁵) have suggested that the specific arrangement of positively charged residues within it is responsible for its ability to facilitate base pairing between U2 snRNA and the branchpoint when U2AF is bound to the nearby polypyrimidine tract (Lee et al., 1993; Valcarcel et al., 1996). The TRA-2 RS2 domain has yet to be tested for an RNA annealing activity, but has been shown to facilitate protein–protein interactions with SR proteins that assemble into the dsx splicing enhancer complex (Wu and Maniatis, 1993; Amrein et al., 1994). Although it has been suggested that the RS domain of the U2AF large subunit is functionally distinct from that of other SR proteins (Valcarcel et al., 1996), our findings indicate that, within the context of the TRA-2 protein, the U2AF⁵⁰ RS domain has the capability of carrying out functions similar to those of TRA-2 RS2.

At this point the specific features of various RS domains giving rise to functional differences and similarities are difficult to define in terms of primary sequence. Size of the RS domain alone does not seem to be a critical factor since RS domains of identical size were observed to differ significantly in function (i.e. dU2AF and dSRp55 fragment B). Because all the RS domains tested have arrays of repeating RS dipeptides, we believe it is most likely that functional specificity arises from sequences interdigitated between these arrays. One clue to the identity of these sequences comes from the observation that elements sufficient to confer functionality are duplicated within the RS2 domain itself. We have shown here that the conserved C-terminal 26 aa of this domain are able to functionally substitute for the entire RS2 region. It has previously been shown (Amrein et al., 1994) that protein produced from a transgene in which this same region is deleted is almost fully functional. Thus, at least two functionally redundant sequences exist within RS2 that are each sufficient to perform its essential functions. One of these sequences is located between amino acids 193 and 235, and the other between amino acids 238 and 264. Comparison of these regions for repeated sequences that might correspond to such a functional motif revealed a 12–13 aa region of similarity (Figure 8). Interestingly, the dU2AF⁵⁰ RS domain contains a sequence that is also similar to part of this repeat while the dSRp55 RS domain does not contain similar sequences. Thus, there is a correlation between the presence of this loosely-defined sequence and the functionality of fusion proteins that have been tested.

Figure 8.

Amino acid region of similarity in the RS domains of TRA-2 and dU2AF⁵⁰. Alignment is shown comparing sequences from the redundant segments of the TRA-2 RS2 domain and from dU2AF⁵⁰ RS. Amino acid identities are shaded black, gray boxes indicate conservative substitutions.

View full figure (32 KB)

Somatic and germline functions of TRA-2 can be separated genetically

The requirement for a specific RS domain in TRA-2 appears to be more dramatic in the male germline than in the somatic tissues. This suggests that the functions of RS2 may differ in these two cell types. The SR proteins, or other factors, that interact with TRA-2 in the male germline are not known but several pieces of evidence suggest that the way TRA-2 functions in the germline differs from that which has been described in the regulation of dsx and fru splicing in the soma. Not only do the target pre-mRNAs affected by TRA-2 differ in these cells, but in addition it is known that the TRA protein, an important cofactor of TRA-2 in the soma, is not expressed in males (McKeown et al., 1987). Thus, TRA-2 must act in a TRA-independent manner in the male germline. Here we have found that fusion of additional protein sequences to the C-terminus of TRA-2 (without deletion of RS2), causes a strong disruption of germline function but has little, if any, effect on function of the protein in the soma. This demonstrates that the TRA-2 germline and somatic functions can be separated from each other genetically and further suggests that the mechanisms by which this factor affects RNA processing differs in these two tissue types.

RS2 is required for discrete subnuclear localization of TRA-2 in primary spermatocytes

When Drosophila splicing factors are expressed in mammalian cells, they behave much like native mammalian splicing factors, localizing to discrete nuclear speckles associated with interchromatin granules in a pattern identical to that of many mammalian splicing factors (Li and Bingham, 1991; Hedley et al., 1995). In the case of the Drosophila splicing regulator suppressor of white apricot (SWAP) it was demonstrated that localization to speckles in COS cells was dependent upon an RS domain in SWAP that could be replaced by an RS domain from the TRA protein (Li and Bingham, 1991). Thus RS domains have been implicated in the subnuclear localization of Drosophila splicing factors when they are expressed in mammalian cells.

Studies on the nuclear distribution of splicing factors in Drosophila cells, however, have not in general revealed distribution patterns like that of the nuclear speckles seen in mammalian cells (Bopp et al., 1991; Kraus and Lis, 1994; Yannoni and White, 1997). In the polytene nuclei of larval salivary gland cells, several Drosophila splicing factors have been shown to localize at discrete chromosomal sites where they appear to be associated with nascent pre-mRNAs (Bopp et al., 1991; Champlin et al., 1991; Amero et al., 1992; Zu et al., 1996; M.E.McGuffin and W.Mattox, unpublished results).

In the course of these studies we have examined the distribution of TRA-2 within the nuclei of growth stage primary spermatocytes and found that it appears to be localized in association with partially condensed pre-meiotic chromosomes. Although the precise nature of these localized sites has yet to be established, we believe that they are likely to correspond to the association of the TRA-2 protein with a subset of nascent pre-mRNAs that are being transcribed in these cells. Localization of TRA-2 to these sites depends upon RS2. A deletion removing this domain results in protein that is diffusely distributed in the nuclei of spermatocytes. Given that RS2 has been implicated in both protein–protein and RNA–protein interactions (Amrein et al., 1994), it seems likely that TRA-2 protein missing this region cannot assemble into specific complexes that normally form at localized sites.

RS domains from other proteins varied in their ability to restore appropriate subnuclear localization to the fusion proteins. Again the most notable difference observed was between the RS domains of dU2AF⁵⁰ and dSRp55. The dU2AF⁵⁰ RS domain conferred localization like that of the full-length TRA2 protein, while protein with sequences from the dSRp55 RS domain was distributed in a diffuse pattern with only traces of localization in a subset of cells. These results indicate that the ability of TRA-2 to associate with specific sites depends on sequences in the RS2 domain. The ability of individual fusion proteins to localize was well correlated with male fertility, and thus it is possible that the inability of fusions proteins to function is a consequence of their failure to localize. However, in strains carrying the TRA-2/SC35 transgene the fusion protein localized normally, but failed to rescue male fertility in mutants. This latter result suggests that localization is not the only germline function affected by the RS domain.

With respect to the specificity of RS domain function, it is interesting to note that the N-terminal RS1 region is an invariant feature of the TRA-2 fusion proteins visualized here. Because deletion of RS2 results in diffuse nuclear localization in these strains, RS1 is not by itself sufficient to direct TRA-2 to the subnuclear sites. Instead RS1 appears to be necessary for nuclear/cytoplasmic localization of TRA-2 in primary spermatocytes (B.Dauwalder and W.Mattox, unpublished material) and contains within it sequences that are similar to a nuclear localization signal identified in the TRA protein (Hedley et al., 1995).

While specific RS domains are needed for correct subnuclear localization of TRA-2, RS2 alone is not likely to be sufficient. The RS domain of dU2AF⁵⁰ supports subnuclear localization of TRA-2 to appropriate sites when substituted for RS2, but does not direct similar localization within the context of the dU2AF⁵⁰ protein itself which is distributed diffusely in the nucleus. We suggest that sequences in the TRA-2 protein that are outside the RS2 domain must also play a necessary role in determining the subnuclear localization pattern.

RS domains and functional specificity of SR proteins

The results presented here suggest that the TRA-2 RS2 domain performs specific functions that are replaced by only a subset of the RS domains tested. Functional specificity as observed in RS2 may not be a universal characteristic of RS domains in other SR proteins. In support of this, several RS domains that have been subjected to similar analysis both in vivo and in vitro, appeared to function generically in that they could be functionally substituted by either artificially generated RS rich sequences or the RS domains of other splicing factors (Chandler et al., 1997b). In contrast, it has recently been demonstrated that the RS domains of several mammalian SR proteins differ in their ability to target the protein to the nucleus and to nuclear speckles and it has been suggested that the determinants for localization differ between proteins with two or only one RRM (Caceres et al., 1997). The observation that some RS domains perform specific functions might explain the fact that these regions in some cases exhibit striking evolutionary conservation. For example, the RS domain of human SC35 is 98% identical to that of chicken SC35 (Fu and Maniatis, 1992; Vellard et al., 1992). Likewise, the RS domains of human and rat TRA-2 are 100% identical (Matsuo et al., 1995; Dauwalder et al., 1996). The conserved sequences in these domains include not only the arrays of RS dipeptides, but also other amino acid residues that are interspersed throughout the RS domain. These observations are consistent with our results on TRA-2 RS2 and further support the idea that the RS domains of some proteins perform specific functions in vivo that differ from those of other SR proteins.

Generation of fusion constructs

Protein fusion constructs were generated in the context of a 3.9 kb EcoRI genomic fragment containing the entire tra-2 transcribed region and all of the flanking sequences necessary for its expression (Goralski et al., 1989). For C-terminal FLAG epitope tags, a plasmid, pTRA2RS, missing the RS2 domain, was first generated by oligonucleotide directed mutagenesis. This replaced sequences from amino acid 193 to C-terminus with sequences coding for the FLAG epitope (GACTACAAGGACGATGACGATAAGGGC) and a TAA stopcodon, flanked by BamHI and XbaI sites. The RS domains of various SR proteins were then PCR-amplified and inserted as BamHI or BamHI–BglII restriction fragments immediately upstream of the FLAG sequences in pTRA2RS. The wild-type control (pTRA2 C-FLAG) was made by oligonucleotide directed mutagenesis (Zoller and Smith, 1987) to introduce a BamHI site, the FLAG epitope, a stop codon and an XbaI site in place of the natural stop codon. For all C-FLAG constructs the fragment from the 5' EcoRI site to the newly created XbaI site was then introduced into pCASPA (Mattox et al., 1990) which contains the tra-2 poly(A) signal and downstream flanking sequences. C-terminal lacZ fusion constructs were generated by introduction of the same EcoRI–XbaI fragments (see above) into pCaZPA (Mattox et al., 1990).

N-terminal FLAG fusion constructs were generated similarly but the FLAG epitope was introduced 5 aa downstream from the start codon of the TRA2²²⁶ isoform. The TRA-2RS and the RS fusion proteins were created in analogous fashion to the C-FLAG constructs except that no epitope tag sequences were added at the C-terminal fusion point and the entire region 3' of the natural stop codon site was retained (as opposed to the more limited 3' UTR sequences present in pCaSPA). The entire mutagenized 3.9 kb fragment was then introduced into the pCaSPeR P-element transformation vector. The RS domains and protein fusion sites in all constructs were confirmed by DNA sequencing.

RS domains were amplified from cDNAs coding for dSRp55, SC35 and TRA. The dU2AF⁵⁰ RS domain was amplifed from Drosophila CS genomic DNA. The regions amplified were from amino acids 198–333, 198–266 and 268–298 in dSRp55, 117–206 in SC35, 19–119 in TRA and 10–41 in dU2AF⁵⁰. The primers used were:

tra RS-1: CCGGGATCCAGGTCTCGATCCCGGCGAGA;

tra RS-2: GGAAGATCTTCGTTCACTGCTGCGACTTC;

dSRp55 RS-1: CCGGGATCCAGCGGACGTGGACGCTCCCG;

dSRp55 RS-2: GGAAGATCTCGAGGCCGAACGGGAGCGTG;

dSRp55 RS-3: GGAAGATCTGGACTTGGACACGTCACGCG;

dSRp55 RS-4: CCGGGATCCTCAAAGTCCCACTCCCGCAC;

SC35 RS-1: CCGGGATCCCGCCGGAGCCGCAGCCCTAG;

SC35 RS-2: CGCGGATCCCGATCGCGACCTGGATTTGG;

dU2AF⁵⁰ RS-1: CCGGGATCCCGCGAGAGACGCCGACATCG;

dU2AF⁵⁰ RS-2: GGAAGATCTCGGCTTGCGCCTCGAGTTCC;

tra-2 RScarb-1: CCGGGATCCAGTCGCAACCGTTACACTCG;

tra-2 RScarb-2: GGAAGATCTATAGCGCGATGAAGTTCGAC.

Generation of transgenic fly lines and crosses

Transgenic flies were obtained by P-element mediated germline transformation of w¹¹¹⁸/B^sY; tra-2^B/CyO embryos. The tra-2^B allele is a nonsense muation in exon 5 (Mattox et al., 1991). Transformants were identified by increased eye pigmentation conferred by the w⁺ gene which was included in the P transposon. All phenotypic and molecular results presented were obtained using flies that carry one copy of the respective transposons.

For the fru RT–PCR experiments the transposons were introduced into the strain w¹¹¹⁸/B^sY; Bl-tra-2/CyO. Flies were then crossed to w¹¹¹⁸/B^sY; Df(2R) trix/CyO (Mattox et al., 1991) flies and the tra-2 heteroallelic transgenic progeny identified by the absence of CyO and presence of the w⁺ marker.

Male fertility was tested by culturing a single male with three w¹¹¹⁸ virgins. Control crosses with wild-type males or mutant tra-2 males were set up simultaneously. All fertile males produced adult progeny.

RNA isolation and low cycle RT–PCR

Drosophila RNA was isolated and dsx RT–PCR was performed as described (Dauwalder et al., 1996). RT–PCR to detect splicing of the M1 intron from tra-2 was carried out in essentially the same way except that cDNA was produced using an oligo dT(12–18 nt) primer and PCR amplifications were carried out with 1 mM MgCl₂ with an annealing step of 60°C. Amplification primers derived from tra-2 exon 3 (5'-CTCAGCCGATTCAGCTGGTGCTCTTG-3') and the exon 5/6 junction (5'- CGCTGTGT/TTGTGCGTCAATCA-3'). Amplified fragments were electrophoresed, blotted and detected with a ³²P-labeled fragment extending from exon 3 to exon 4.

RT–PCR on fru RNA was carried out using random primers for cDNA synthesis and amplifications were done with 1.5 mM MgCl₂ using the primers described by Ryner et al. (1996). Male/common region and female-specific region hybridization probes were PCR amplified, using the same sex specific primers used for RT–PCR, in combination with a downstream sex-specific primer. The downstream primers used were: fru female antisense: 5'-CGCGCCAGTTGGTGGGGATTTG-3' and fru male antisense: 5'-GGGGATGCGGGCGGAAGCGGA-3'. The PCR fragments were labeled with ³²P following the protocol of Mertz and Rashtchian (1994).

Immunohistochemistry

Flies were dissected in Testis Dissection Buffer (TDB) [0.183 M KCl, 47 mM NaCl, 10 mM Tris–HCl pH 6.8; (Bonaccorsi et al., 1988)]. Testes were placed on a microscope slide in a drop of TDB and covered with a silanized coverslip. They were then gently squashed until cells started flowing out of the testes. The slides were placed onto a Kimwipe with the coverslip side down and superfluous liquid allowed to drain. Slides were then dropped into liquid nitrogen. Fixation was done according to Pisano et al. (1993). Slides were washed three times in phosphate-buffered saline (PBS) for 15 min, in PBX (PBS + 0.2% Triton X-100) for 15 min and blocked in 1% bovine serum albumin (BSA) in PBX (BBX) for 20 min. Primary antibody was added in PBX + 2% normal goat serum, the slide covered with a coverslip and incubated at 4°C overnight in a humid chamber. The slides were washed three times in PBX for 15 min, incubated with the secondary antibody (in PBX + 10% normal goat serum) for 5 h at room temperature (RT) and washed three times in PBX. For double staining, the slides were again blocked for 20 min in BBX before application of the second primary antibody. DNA was stained with bis-Benzimide (Hoechst No. 33258, 0.5 g/ml in PBX). The slides were then rinsed in 1 PBS and mounted in Aqua poly/mount (Polysciences) and viewed using a Zeiss Axiophot Microscope.

Antibodies. M2-anti-FLAG (Kodak) was used at dilutions ranging from 1:400 to 1:800 and detected with TR- or FITC-coupled goat anti-mouse IgG at 1:200 or 1:100 dilution (Jackson Laboratories). Affinity-purified anti-dU2AF⁵⁰ was a gift from D.Rudner and D.Rio, and was used at 1:500 dilution and detected with TR-coupled goat anti-mouse IgG at 1:200 dilution (Jackson Laboratories). Rabbit anti--gal (Cappel) was used at 1:500 to 1:700 and detected by TR- or FITC-coupled goat anti-rabbit IgG (Jackson Laboratories) at 1:500 dilution. All antibodies (except for anti dU2AF⁵⁰) were pre-absorbed against Drosophila w¹¹¹⁸ embryos before use.

We thank Mitzi Kuroda, Tom Cooper and the members of the Mattox laboratory, especially Dawn S.Chandler, for helpful discussions and comments. We are grateful to M.Roth and Jo Yeakley for providing plasmid DNAs, and to D.Rudner and D.Rio for the dU2AF⁵⁰ antibody. This work was supported by research grants from the American Cancer Society (NP876), the National Institutes of Health (GM50825) and by a Pew Scholars Award to W.M. Sequencing was done by the DNA Core Sequencing Facility, which is supported by NIH grant CA16672.

Amero SA, Raychaudhuri G, Cass CL, van Venrooij WJ, Habets WJ, Krainer AR and Beyer AL (1992) Independent deposition of heterogeneous nuclear ribonucleoproteins and small nuclear ribonucleoprotein particles at sites of transcription. Proc Natl Acad Sci USA, 89, 8409–8413. | Article | PubMed | ChemPort |

Amrein H, Gorman M and Nöthiger R (1988) The sex-determining gene tra-2 of Drosophila encodes a putative RNA binding protein. Cell, 55, 1025–1035. | Article | PubMed | ISI | ChemPort |

Amrein H, Maniatis T and Nöthiger R (1990) Alternatively spliced transcripts of the sex-determining gene tra-2 of Drosophila encode functional proteins of different size. EMBO J, 9, 3619–3629. | PubMed | ISI | ChemPort |

Amrein H, Hedley ML and Maniatis T (1994) The role of specific protein–RNA and protein–protein interactions in positive and negative control of pre-mRNA splicing by Transformer 2. Cell, 76, 735–746. | Article | PubMed | ISI | ChemPort |

Birney E, Kumar S and Krainer AR (1993) Analysis of the RNA-recognition motif and RS and RGG domains: conservation in metazoan pre-mRNA splicing factors. Nucleic Acids Res, 21, 5803–5816. | Article | PubMed | ISI | ChemPort |

Bonaccorsi S, Pisano C, Puoti F and Gatti M (1988) Y chromosome loops in Drosophila melanogaster. Genetics, 120, 1015–1034. | PubMed | ISI | ChemPort |

Bopp D, Bell LR, Cline TW and Schedl P (1991) Developmental distribution of female-specific Sex-lethal proteins in Drosophila melanogaster. Genes Dev, 5, 403–415. | PubMed | ISI | ChemPort |

Caceres JF, Misteli T, Screaton GR, Spector DL and Krainer AR (1997) Role of the modular domains of SR proteins in subnuclear localization and alternative splicing specificity. J Cell Biol, 138, 225–238. | Article | PubMed | ISI | ChemPort |

Cenci G, Bonaccorsi S, Pisano C, Verni F and Gatti M (1994) Chromatin and microtubule organization during premeiotic, meiotic and early postmeiotic stages of Drosophila melanogaster spermatogenesis. J Cell Sci, 107, 3521–3534. | PubMed | ISI | ChemPort |

Champlin DT, Frasch M, Saumweber H and Lis JT (1991) Characterization of a Drosophila protein associated with boundaries of transcriptionally active chromatin. Genes Dev, 5, 1611–1621. | Article | PubMed | ISI | ChemPort |

Chandler D, McGuffin ME, Piskur J, Yao J, Baker BS and Mattox W (1997a) Evolutionary conservation of regulatory strategies for the sex determination factor transformer-2. Mol Cell Biol, 17, 2908–2919. | PubMed | ISI | ChemPort |

Chandler SD, Mayeda A, Yeakley JM, Krainer AR and Fu XD (1997b) RNA splicing specificity determined by the coordinated action of RNA recognition motifs in SR proteins. Proc Natl Acad Sci USA, 94, 3596–3601. | Article | PubMed | ChemPort |

Dauwalder B, Amaya-Manzanares F and Mattox W (1996) A human homologue of the Drosophila sex determination factor transformer-2 has conserved splicing regulatory functions. Proc Natl Acad Sci USA, 93, 9004–9009. | Article | PubMed | ChemPort |

Fu XD (1995) The superfamily of arginine/serine-rich splicing factors. RNA, 1, 663–680. | PubMed | ISI | ChemPort |

Fu XD and Maniatis T (1992) Isolation of a complementary DNA that encodes the mammalian splicing factor SC35. Science, 256, 535–538. | Article | PubMed | ISI | ChemPort |

Goralski TJ, Edstrom JE and Baker BS (1989) The sex determination locus transformer-2 of Drosophila encodes a polypeptide with similarity to RNA binding proteins. Cell, 56, 1011–1018. | Article | PubMed | ISI | ChemPort |

Hazelrigg T and Tu C (1994) Sex-specific processing of the Drosophila exuperantia transcript is regulated in male germ cells by the tra-2 gene. Proc Natl Acad Sci USA, 91, 10752–10756. | Article | PubMed | ChemPort |

Hedley ML, Amrein H and Maniatis T (1995) An amino acid sequence motif sufficient for subnuclear localization of an arginine/serine-rich splicing factor. Proc Natl Acad Sci USA, 92, 11524–11528. | Article | PubMed | ChemPort |

Heinrichs V and Baker BS (1995) The Drosophila SR protein RBP1 contributes to the regulation of doublesex alternative splicing by recognizing RBP1 RNA target sequences. EMBO J, 14, 3987–4000. | PubMed | ISI | ChemPort |

Heinrichs V, Ryner LC and Baker BS (1998) Regulation of sex-specific selection of fruitless 5' splice sites by transformer and transformer-2. Mol Cell Biol, 18, 450–458. | PubMed | ISI | ChemPort |

Ito H, Fujitani K, Usui K, Shimizunishikawa K, Tanaka S and Yamamoto D (1996) Sexual orientation in Drosophila is altered by the satori mutation in the sex-determination gene fruitless that encodes a zinc finger protein with a BTB domain. Proc Natl Acad Sci USA, 93, 9687–9692. | Article | PubMed | ChemPort |

Kohtz JD, Jamison SF, Will CL, Zuo P, Lührmann R, Garcia-Blanco MA and Manley JL (1994) Protein–protein interactions and 5'-splice-site recognition in mammalian mRNA precursors. Nature, 368, 119–124. | Article | PubMed | ISI | ChemPort |

Kraus ME and Lis JT (1994) The concentration of B52, an essential splicing factor and regulator of splice site choice in vitro, is critical for Drosophila development. Mol Cell Biol, 14, 5360–5370. | PubMed | ISI | ChemPort |

Lee C, Zamore PD, Green MR and Hurwitz J (1993) RNA annealing activity is intrinsically associated with U2AF. J Biol Chem, 18, 13472–13478.

Li H and Bingham PM (1991) Arginine/serine-rich domains of the su(w^a) and tra RNA processing regulators target proteins to a subnuclear compartment implicated in splicing. Cell, 67, 335–342. | Article | PubMed | ISI | ChemPort |

Lynch KW and Maniatis T (1995) Synergistic interactions between two distinct elements of a regulated splicing enhancer. Genes Dev, 9, 284–93. | PubMed | ISI | ChemPort |

Lynch KW and Maniatis T (1996) Assembly of specific SR protein complexes on distinct regulatory elements of the Drosophila doublesex splicing enhancer. Genes Dev, 10, 2089–2101. | Article | PubMed | ISI | ChemPort |

Madigan SJ, Edeen P, Esnayra J and McKeown M (1996) att, a target for regulation by tra2 in the testes of Drosophila melanogaster, encodes alternative RNAs and alternative proteins. Mol Cell Biol, 16, 4222–4230. | PubMed | ISI | ChemPort |

Manley JL and Tacke R (1996) SR proteins and splicing control. Genes Dev, 10, 1569–1579. | Article | PubMed | ISI | ChemPort |

Matsuo N, Ogawa S, Imai Y, Takagi T, Tohyama M, Stern D and Wanaka A (1995) Cloning of a novel RNA binding polypeptide (RA301) induced by hypoxia/reoxygenation. J Biol Chem, 270, 28216–28222. | Article | PubMed | ISI | ChemPort |

Mattox W and Baker BS (1991) Autoregulation of the splicing of transcripts from the transformer-2 gene of Drosophila. Genes Dev, 5, 786–796. | PubMed | ISI | ChemPort |

Mattox W, Palmer MJ and Baker BS (1990) Alternative splicing of the sex determination gene transformer-2 is sex-specific in the germ line but not in the soma. Genes Dev, 4, 789–805. | PubMed | ISI | ChemPort |

Mattox W, McGuffin ME and Baker BS (1996) A negative feedback mechanism revealed by functional analysis of the alternative isoforms of the Drosophila splicing regulator transformer-2. Genetics, 143, 303–314. | PubMed | ISI | ChemPort |

McKeown M, Belote JM and Baker BS (1987) A molecular analysis of transformer, a gene in Drosophila melanogaster that controls female sexual differentiation. Cell, 48, 489–499. | Article | PubMed | ISI | ChemPort |

Mertz LM and Rashtchian A (1994) Nucleotide imbalance and polymerase chain reaction: Effects on DNA amplification and synthesis of high specific activity radiolabeled DNA probes. Anal Biochem, 221, 160–165. | Article | PubMed | ISI | ChemPort |

Pisano C, Bonaccorsi S and Gatti M (1993) The kl-3 loop of the Y chromosome of Drosophila melanogaster binds a tektin-like protein. Genetics, 133, 569–579. | PubMed | ISI | ChemPort |

Ruskin B, Zamore PD and Green MR (1988) A factor, U2AF, is required for U2 snRNP binding and splicing complex assembly. Cell, 52, 207–219. | Article | PubMed | ISI | ChemPort |

Ryner LC, Goodwin SF, Castrillon DH, Anand A, Villella A, Baker BS, Hall JC, Taylor BJ and Wasserman SA (1996) Control of male sexual behavior and sexual orientation in Drosophila by the fruitless gene. Cell, 87, 1079–1089. | Article | PubMed | ISI | ChemPort |

Staknis D and Reed R (1994) SR proteins promote the first specific recognition of pre-mRNA and are present together with the U1 small nuclear ribonucleoprotein particle in a general splicing enhancer complex. Mol Cell Biol, 14, 7670–7682. | PubMed | ISI | ChemPort |

Tian M and Maniatis T (1993) A splicing enhancer complex controls alternative splicing of doublesex pre-mRNA. Cell, 74, 105–114. | Article | PubMed | ISI | ChemPort |

Valcarcel J, Gaur RK, Singh R and Green MR (1996) Interaction Of U2AF⁶⁵ RS region with pre-mRNA of branch point and promotion base pairing with U2 snRNA. Science, 273, 1706–1709. | Article | PubMed | ISI | ChemPort |

Vellard M, Sureau A, Soret J, Martinerie C and Perbal B (1992) A potential splicing factor is encoded by the opposite strand of the trans-spliced c-myb exon. Proc Natl Acad Sci USA, 89, 2511–2515. | Article | PubMed | ChemPort |

Watanabe TK (1975) A new sex transforming gene on the second chromosome of Drosophila melanogaster. Jpn J Genet, 50, 269–271.

Wu JY and Maniatis T (1993) Specific interactions between proteins implicated in splice site selection and regulated alternative splicing. Cell, 75, 1061–1070. | Article | PubMed | ISI | ChemPort |

Yannoni YM and White K (1997) Association of the neuron-specific RNA binding domain-containing protein ELAV with the coiled body in Drosophila neurons. Chromosoma, 105, 332–341. | Article | PubMed | ISI | ChemPort |

Zahler AM, Lane WS, Stolk JA and Roth MB (1992) SR proteins: a conserved family of pre-mRNA splicing factors. Genes Dev, 6, 837–847. | PubMed | ISI | ChemPort |

Zoller MJ and Smith M (1987) Oligonucleotide-directed mutagenesis: a simple method using two oligonucleotide primers and a single-stranded DNA template. Methods Enzymol, 154, 329–350. | PubMed | ISI | ChemPort |

Zu K, Sikes ML, Haynes SR and Beyer AL (1996) Altered levels of the Drosophila HRB87F/hrp36 hnRNP protein have limited effects on alternative splicing in vivo. Mol Biol Cell, 7, 1059–1073. | PubMed | ISI | ChemPort |

Zuo P and Maniatis T (1996) The splicing factor U2AF³⁵ mediates critical protein–protein interactions in constitutive and enhancer-dependent splicing. Genes Dev, 10, 1356–1368. | Article | PubMed | ISI | ChemPort |

Main navigation

• Search EMBO Journal

  Search

• Advanced search

• The EMBO Journal

  • Home
  • Aims and scope
  • Current issue
  • Advance online publication
  • Web focuses
  • Archive:
  • Browse by issue
  • Browse by subject
  • Browse by category
  • Free online sample issue
  • Press releases

• Authors and Referees

  • Editorial process
  • Guide for authors
  • Submit an article
  • Guide for referees
  • Editors and Editorial Board
  • Contact editorial office

• Customer Service

  • Subscribe
  • Order sample copy
  • Purchase articles
  • Reprints and permissions
  • Contact NPG
  • Advertising