PPS, a large multidomain protein, functions with sex-lethal to regulate alternative splicing in Drosophila

Abstract

Alternative splicing controls the expression of many genes, including the Drosophila sex determination gene Sex-lethal (Sxl). Sxl expression is controlled via a negative regulatory mechanism where inclusion of the translation-terminating male exon is blocked in females. Previous studies have shown that the mechanism leading to exon skipping is autoregulatory and requires the SXL protein to antagonize exon inclusion by interacting with core spliceosomal proteins, including the U1 snRNP protein Sans-fille (SNF). In studies begun by screening for proteins that interact with SNF, we identified PPS, a previously uncharacterized protein, as a novel component of the machinery required for Sxl male exon skipping. PPS encodes a large protein with four signature motifs, PHD, BRK, TFS2M, and SPOC, typically found in proteins involved in transcription. We demonstrate that PPS has a direct role in Sxl male exon skipping by showing first that loss of function mutations have phenotypes indicative of Sxl misregulation and second that the PPS protein forms a complex with SXL and the unspliced Sxl RNA. In addition, we mapped the recruitment of PPS, SXL, and SNF along the Sxl gene using chromatin immunoprecipitation (ChIP), which revealed that, like many other splicing factors, these proteins bind their RNA targets while in close proximity to the DNA. Interestingly, while SNF and SXL are specifically recruited to their predicted binding sites, PPS has a distinct pattern of accumulation along the Sxl gene, associating with a region that includes, but is not limited to, the SxlPm promoter. Together, these data indicate that PPS is different from other splicing factors involved in male-exon skipping and suggest, for the first time, a functional link between transcription and SXL-mediated alternative splicing. Loss of zygotic PPS function, however, is lethal to both sexes, indicating that its role may be of broad significance.

Figure 1. PPS, a large multidomain protein, is a SNF interacting protein.

(A) Yeast two-hybrid interactions between PPS and SNF. Positive interactions were tested by assaying the ability of the transformed yeast to grow on selective media after 3 days. (B) PPS/SNF complex assembly tested by GST pull-down assays in whole cell extracts. The lane marked 10% input is a control in which the amount of extract corresponds to 10% of the material applied to the glutathione affinity beads. (C) Diagram of the 2016 amino acid PPS protein. PPS contains 4 conserved motifs, which are drawn approximately to scale. The line above the diagram is the region of the protein used for the yeast two hybrid experiments in (A), in the GST pulls experiments in (B) and for production of the PPS antibody. (D) Genomic organization of PPS and its neighboring genes. Solid boxes represent exons. The position of the insertions used to generate pps¹ is indicated above the diagram. The genomic DNA deleted in pps¹ and the genomic DNA used in the rescue constructs is indicated by a solid line below the diagram.

Figure 2. Sxl splicing is disrupted in the ovaries of incompletely rescued pps¹ mutant females.

(A) DAPI-stained ovariole from a wild female (WT) and a P{pps+}/+; pps¹/Df(3R)Exel7316 female. (B) Diagram of the alternative splicing event that produces sex-specific Sxl transcripts. The arrows below the diagram indicate the position of the PCR primer pairs used for RT–PCR. (C) The tumor phenotype is correlated with Sxl splicing defects. Splicing was assayed by RT–PCR using RNA isolated from ovaries dissected from P{pps+}/+; pps¹/Df(3R)Exel7316 females (pps tumors). Controls include splicing in ovaries isolated from wild type (WT) females and splicing in adult males.

Figure 3. pps is a dosage-sensitive maternal modifier of Sxl.

Synergistic genetic interactions lead to female lethality. In these assays, females of the indicated genotype were mated to Sxl^f1/Y and the resulting progeny scored. On the assumption that an equal number of male and female progeny will be generated from each cross, the percent female viability was calculated by comparing the number of females recovered with the number of males recovered.

Figure 4. PPS associates with SNF and U1-70K in embryonic extracts.

(A) pps¹ is a protein-null allele. Western blot of extracts made from wild type and pps¹ mutant animals probed with antibodies against PPS. SNF is used here as a loading control. (B) PPS interacts with SNF in a RNA–independent manner. Western blots of PPS and U2A' immunoprecipitations (Co-IP) in nuclear extracts made from embryos probed with an antibody against SNF. The RNase sensitivity of this association was tested by pretreating the extract with a combination of RNase A and RNase T1. Controls include the previously described RNase sensitive SNF/U2A' association. (C) PPS associates with U1-70K. Western blot of PPS and SNF immunoprecipitations (Co-IP) in nuclear extracts made from embryos probed with an antibody against U1-70K. Controls include the previously described SNF/U1-70K association. The lanes marked 2.5% input are controls in which the amount of extract corresponds to 2.5% of the material used in each Co-IP experiment.

Figure 5. PPS associates with the SXL protein and the Sxl pre–mRNA.

(A) PPS interacts with SXL in a RNA-dependent manner. Western blots of PPS immunoprecipitations (Co-IP) in nuclear extracts made from embryos probed with an antibody against SXL. The RNase sensitivity of this association was tested by pretreating the extract with a combination of RNase A and RNase T1. (B,C) PPS associates with the unspliced Sxl pre-mRNA in a SXL-independent manner. RNA-immunoprecipitation assays (RIP) were carried out in nuclear extracts made from wild type embryos (WT), or embryos from da¹/da¹ mothers. The presence of unspliced Sxl RNA in the IP pellet was detected by RT-PCR using an intron 3/exon 4 primer pair. Immunoprecipitations with SXL or SNF were included as positive controls. Negative controls included precipitations with no antibody, pre-immune serum and Polycomb (PC). The lanes marked input are controls in which the amount of extract corresponds to a percentage of the material used in each Co-IP experiment.

Figure 6. Accumulation of SXL, SNF, and PPS along the body of the Sxl gene in embryos.

ChIP using SXL, SNF, and PPS specific antibodies. After ChIP, the extracted DNA was analyzed by PCR using primer pairs positioned along the Sxl gene as diagramed. Specificity controls include ChIPs using antibodies directed against the Heat Shock Factor (HSF) and RNA Pol IIa (8WG16), as a well as a no antibody control (no Ab). ChIPs were carried out, from left to right, in 8–12 hour old wild-type embryos, female embryos (embryos from C(1)DX mothers crossed to Sxl^7BO males) and embryos from da¹/da¹ mothers. To ensure that the PCR reactions of the antibody enriched DNA fell within a linear range of amplification, PCR reactions were carried out on serially diluted input DNA, ranging from 1% to 10% of total DNA. The PCR data shown here are representative of three independent ChIP experiments.

Figure 7. Accumulation of PPS near the SxlPm promoter in embryos.

ChIP assays using SNF–and PPS–specific antibodies were carried out using the same population of embryos as described in Figure 6. After ChIP the extracted DNA was analyzed by PCR using primer pairs positioned around the SxlPm promoter as diagramed. The 8WG16 antibody, which detects the hypophosphorylated Pol II (Pol IIa), is used here to mark the promoter. Consistent with published studies, Pol IIa was largely detected at the promoter whereas SNF was only detected by a primer set designed to detect the beginning of the transcription unit.

Figure 8. tra is a PPS target gene.

(A) RIP assays demonstrating that PPS associates with the tra pre-mRNA, but not the snf pre–mRNA or the intronless U2A transcript. The presence of unspliced RNA in the IP pellet was detected by RT–PCR. (B) ChIP assays demonstrating that PPS is detected at the tra promoter (identified by Pol IIa accumulation), but not at the snf or the U2A promoter. The exact position of the primers used in the RIP and ChIP assays are described in the Materials and Methods section.

Figure 9. Co-transcriptional model for Sxl splicing autoregulation.

PPS associates with Pol II during transcription (Pol II, oval) to help recruit the U1 snRNP (U1, blue circle) and SXL (grey circle) to the appropriate locations on the nascent transcript. In addition, PPS may help nucleate the interaction between the U1 snRNP and SXL. Splicing could be blocked immediately (insert) or spliceosome assembly could continue, stalling only later in the pathway. The end result is a dead-end complex that guarantees that the male exon will be skipped, and that exon 2 is spliced to exon 4. In males, where there is no SXL protein, the U1 snRNP is free to assemble into a functional spliceosome and exon 3 is included in the mature transcript (not shown).