Sequences within Bmdsx exon 4 determine the inclusion of exons 3 and 4. (A) The genomic sequences in the Bmdsx gene included in the minigene construct are shown at the top. Schematic diagrams of the wild-type (pWTE2.5) and mutant (pMutE2, pMutE3, pMutE4, and pMutE5) Bmdsx minigene constructs used in panel B are also indicated. The open boxes represent the Bmdsx exons, whereas the exons replaced with the Bombyx silk protein P25 exons are indicated by filled boxes (see Materials and Methods). The numbers within each box represents exon labels. The arrows indicate the approximate locations of the primers that were used for RT-PCR in panel B. BmA3 pro, Bombyx actin 3 promoter sequence; SV40 pA, simian virus 40 polyadenylation sequence; lines, intron sequences. (B) The upper panel indicates RT-PCR analysis results upon transient transfection of the minigene constructs shown in panel A into NIAS-Bm-F1 cells (lanes F1) and NIAS-Bm-M1 cells (lanes M1). To identify the transcript splicing pattern from each minigene construct, PCR products were cloned and sequenced. The lower panel shows the RT-PCR analysis of endogenous Bmdsx transcripts using primers as indicated in Fig. 1D. PCR products were separated on 2% agarose gels and stained with ethidium bromide. The structure of the RT-PCR products is indicated schematically to the left of each panel.

Complex formation between NIAS-Bm-M1 cell nuclear proteins and RNA sequences in exon 4 that are essential for male-specific splicing. (A) Schematics of the ³²P-labeled probes and competitor RNAs. (B) Radiolabeled transcripts were incubated for 15 min on ice in the presence (+) or absence (-) of a NIAS-Bm-M1 cell nuclear extract (M1 cell NE). Complexes were separated on a nondenaturing 4.5% acrylamide gel. The positions of the free RNA and complex are indicated on the left. Asterisks denote an unspecific complex. NC, control RNA containing 140 nt from pBluescript. (C) The RNA-protein interaction between a labeled E4 probe and the NIAS-Bm-M1 cell nuclear extract in the presence of increasing amounts of CE1, CE2, CE3a, CE3b, and CE3c unlabeled competitors (80, 800, and 2,400 ng). Complexes were separated on a nondenaturing 4.5% acrylamide gel. (D) Radiolabeled probes were incubated for 15 min on ice in the presence of increasing amounts of a NIAS-Bm-M1 cell nuclear extract (2.5, 5, and 10 μg). Other experimental procedures were the same as those for panels B and C.

Binding of the nuclear factor to CE1 correlates with repression activity. (A) Sequence of the CE1 probe and the mutated sequences of two types of CE1 mutants (CE1mut1 and CE1mut2). Dots represent identical nucleotides in CE1 and the mutated CE1 sequences. UAA repeats are underlined. CE1s consists of the first 15 nt of CE1. (B) Radiolabeled probes were incubated for 15 min on ice in the presence of increasing amounts of a NIAS-Bm-M1 cell nuclear extract (2.5, 5, and 10 μg). Complexes were separated on a nondenaturing 4.5% acrylamide gel. The positions of the free RNA and complex are indicated on the left. (C) The upper panel shows a schematic diagram of the pWTE2.5 derivative mutant minigenes in which the CE1 sequence was replaced with the CE1mut1 or CE1mut2 mutant sequence. The lower panel indicates the quantification of the RT-PCR analysis shown in panel D. Results are expressed as the percentage of RT-PCR products representing female splicing relative to total RT-PCR products. Data represent the means of four independent experiments. (D) RT-PCR analysis of NIAS-Bm-M1 cells transiently transfected with the mutant minigenes shown in panel C, using primers as indicated in Fig. 2A. As a control, RT-PCR amplification was also performed from cytoplasmic RNA of NIAS-Bm-F1 cells (lane cF) and NIAS-Bm-M1 cells (lane cM) transiently transfected with the wild-type minigene construct pWTE2.5. Labels to the left of the gel refer to female splicing and male splicing.

Down-regulation of BmPSI by RNAi increases female-specific splicing. (A) Schematic diagram of three dsRNAs (a, b, and c) specific for BmPSI. (B) NIAS-Bm-M1 cells were transfected with dsRNAs specific for DsRed (control), BmSqd2, BmELAV, or BmPSI. The upper panel shows the Bmdsx splicing pattern as detected by RT-PCR. Labels to the left of the gel refer to female splicing (Bmdsx F) and male splicing (Bmdsx M). The lower panel shows the specific down-regulation of BmSqd2, BmELAV, and BmPSI expression in dsRNA-treated cells. The RNA level of each gene was detected by RT-PCR using primers specific for each gene. (C) Quantification of the results derived from three independent experiments. Results are expressed as the percentage of RT-PCR products representing female or male splicing relative to total RT-PCR products.

Characterization of NIAS-Bm-M1 and NIAS-Bm-F1 cell lines. (A) Photograph of fertilized eggs, laid by Sex-limited black egg strain females, that show a male-specific defect in egg pigmentation. Primrose yellow eggs are males, and mauve eggs are females. Bar, 5 mm. (B and C) NIAS-Bm-M1 (B) and NIAS-Bm-F1 (C) cells were photographed using phase-contrast microscopy. Bars, 100 μm. (D) Schematic diagram of alternative splicing in Bmdsx pre-mRNA. Boxes represent exons. The gray region indicates the female-specific open reading frame (ORF). The black region indicates the male-specific ORF. The numbers above the diagram represent exon labels. V-shaped lines above (skipping of alternative exons) and below (inclusion of alternative exons) the diagram represent the endogenous Bmdsx splice variants observed in males and females. Stop codons are indicated. The arrows indicate the approximate location of the primers that were used for the RT-PCR in panel E. Skipping of exons 3 and 4 results in an 880-bp PCR product, while inclusion results in a 1,134-bp product. (E) The left panel shows RT-PCR analysis of endogenous Bmdsx transcripts in adult females (lane 1), NIAS-Bm-F1 cells (lane 2), adult males (lane 3), and NIAS-Bm-M1 cells (lane 4) using primers as indicated in panel D. The right panel shows molecular sexing of NIAS-Bm-F1 cells (lane 1) and NIAS-Bm-M1 cells (lane 2) by PCR using oligonucleotide primers derived from the W chromosome-specific retrotransposon. As a control reaction, a genomic fragment of the Bmdsx gene (mapped on chromosome 25) was amplified from NIAS-Bm-F1 cells (lane 3) and NIAS-Bm-M1 cells (lane 4). PCR products were separated on 2% agarose gels and stained with ethidium bromide. Molecular sizes in base pairs are indicated to the left of each panel.

Identification of negatively acting regulatory sequences within Bmdsx exon 4 by linker scan mutagenesis. (A) Sequence of Bmdsx exon 4, indicating the positions of the introduced linker scan mutants (ls1 to ls16). Segments flanked by vertical lines were replaced consecutively by a 10-bp NotI linker sequence (boxed). Uppercase lettering, exon sequence; lowercase lettering, intron sequences. (B) RT-PCR analysis of NIAS-Bm-M1 cells transiently transfected with the ls mutants using primers as indicated in Fig. 2A. As a control, RT-PCR amplification was also performed from cytoplasmic RNA of NIAS-Bm-F1 cells (lane cF) and NIAS-Bm-M1 cells (lane cM) transiently transfected with the wild-type minigene construct pWTE2.5. Labels to the left of the gel refer to female splicing and male splicing. (C) Quantification of the RT-PCR analysis shown in panel B. Results are expressed as the percentage of RT-PCR products representing female splicing relative to total RT-PCR products.

Nuclear factors interacting with CE1 and CE2. Radiolabeled CE1 and CE2 RNAs were incubated with the indicated amounts of NIAS-Bm-M1 (M1) and NIAS-Bm-F1 (F1) cell nuclear extracts. Complexes were separated on a nondenaturing 4.5% acrylamide gel. The position of the free RNA and complex are indicated on the left. NE, nuclear extract.

Identification of BmPSI as a nuclear factor binding specifically to CE1. (A) The left panel shows UV cross-linking of NIAS-Bm-M1 cell nuclear extract to radiolabeled CE1 and CE1mut1 RNAs. Proteins cross-linked to RNase A-digested samples were electrophoresed on an SDS-10% polyacrylamide gel. The arrow indicates an approximately 80-kDa protein band that is present in the CE1 lane as opposed to the CE1mut1 lane. Approximate locations of molecular mass markers are shown on the right. The right panel shows the result of affinity chromatography using adipic acid dehydrazide beads derivatized with CE1 and CE1mut1 RNAs following incubation with a NIAS-Bm-M1 cell nuclear extract. The arrow indicates the protein band with a molecular mass corresponding to the molecular mass of the protein band at 80 kDa that was UV cross-linked to the labeled CE1 RNA. Lane MW, molecular size markers. (B) The full amino acid sequence of BmPSI, with the open boxes corresponding to the sequenced peptides from the excised 80-kDa band. Underlined sequences highlight the four KH-domains. Sequences underlined with dots indicate A and B motifs. (C) The purified recombinant protein (left panel) and its reactivity with labeled CE1 and CE1mut1 RNAs in a UV cross-linking assay (right panel). (D) Radiolabeled CE1 and CE1mut1 RNAs were incubated in the presence of 1.25, 2.5, 5, and 10 pmol of recombinant BmPSI. Complexes were separated on a nondenaturing 4.5% acrylamide gel.

5′ splice site-like sequence upstream of CE1. The sequence around the 3′ splice site of Bmdsx exon 4 is shown. The 5′ splice site-like sequence similar to the 5′ splice site consensus sequence (GURAGU) is underlined. The 3′ splice site of Bmdsx exon 4 is indicated by an arrow. The CE1 sequence is boxed.