NineTeen Complex-subunit Salsa is required for efficient splicing of a subset of introns and dorsal–ventral patterning

Splicing of the first intron of gurken is required for Gurken expression and function. Females homozygous for grk^deltaFRT are viable but sterile, with egg laying severely compromised (Lan et al. 2010). grk^deltaFRT corresponds to an amorphic allele of gurken (deletion mutant) (Lan et al. 2010), whereas grk^HF contains a single base pair change that introduces a precocious stop codon that severely truncates Gurken protein (Thio et al. 2000). (A) Expression of one copy of a genomic construct of gurken with its own endogenous promoter, with (grk^WT) or without (grk^{no 1st intron}) the transcript first intron, rescued fertility (egg hatching) of grk null (grk^deltaFRT/grk^HF) females. Expression of one copy of a genomic construct of gurken with mutated first intron splice sites (grk^{splice site mutant}) only partially rescued egg hatching and a proportion of the rescued L1 larvae died after eclosion. Number of eggs scored for each experiment are indicated above the bar plot and are the sum of two independent experiments. t-test analysis shows that expression of grk^WT (P = 0.0144) or grk^{no 1st intron} (P = 0.0128) significantly increased egg hatching rate when compared to the expression of grk^{splice site mutant}. (B) Expression of one copy of a genomic construct of gurken with its own endogenous promoter, with (grk^WT) or without (grk^{no 1st intron}) the transcript first intron, rescued eggshell dorsal appendages defects (eggshell ventralization) of eggs laid by grk null (grk^deltaFRT/grk^HF) females. Expression of one copy of a genomic construct of gurken with mutated first intron splice sites (grk^{splice site mutant}) only weakly rescued eggshell ventralization. Number of eggs scored for each experiment are indicated above the bar plot and are the sum of two independent experiments. (C) Expression of two copies of a genomic construct of gurken with its own endogenous promoter, with (grk^WT) or without (grk^{no 1st intron}) the transcript first intron, fully rescued eggshell dorsal appendages defects (eggshell ventralization) of eggs laid by grk null (grk^deltaFRT/ grk^deltaFRT) females. Number of eggs scored for each experiment are indicated above the bar plot and are the sum of two independent experiments. (B,C) For quantification of eggshell dorsal appendages defects (eggshell ventralization), observed phenotypes were categorized in four different phenotypic classes based on the eggshell dorsal appendages: class 0: wild-type dorsal appendages—two individualized dorsal appendages); class 1: dorsal appendages only fused at bottom; class 2: fused dorsal appendages—spindle phenotype; and class 3: short eggs (dumpless-like phenotype) without or with extremely short dorsal appendages. Examples of scored phenotypes are shown in Supplemental Figure S10. (D,E) Eggs from grk null (grk^deltaFRT/grk^HF) females lack detectable Gurken protein. Expression of one copy of a wild-type genomic construct of gurken (grk^WT) rescued dorsal anterior expression of Gurken. Retention of the first intron due to splice site mutations (grk^{splice site mutant}) significantly decreased Gurken dorsal anterior expression. First intron 5′ and 3′ splice sites (respectively, GT and AG) were both mutated to TT. Gurken was detected with an anti-Gurken monoclonal antibody and DNA was visualized with DAPI staining. DNA (Blue) and Gurken (green). Scale bars, 30 and 15 µm (respectively, first two and last two images). (E) Relative oocyte Gurken dorsal anterior signal (arbitrary units [a.u.]) corresponds to Gurken signal pixel intensity (average of three different measurements taken from the oocyte dorsal-anterior region with the highest perinuclear signal) divided by the respective background signal (average of three different measurements from the respective oocyte cytoplasm). To minimize sample variation all measurements were obtained from maximum intensity projections obtained from confocal Z stacks of stage 8/9/10 egg chambers. Each dot represents an individual egg chamber. Horizontal lines specify mean values and asterisks indicate significant difference (two-tailed unpaired t-test; P < 0.0001). (F) Expression of a genomic construct of gurken, without or with first intron splice site mutations (respectively, grk^WT and grk^{splice site mutant}), in a grk null (grk^deltaFRT/grk^deltaFRT) background, generated equivalent levels of gurken transcripts. Respectively, RT-ddPCR (see Materials and Methods) measured 460.9 ± 48.6 and 410.6 ± 51 copies of gurken mRNA (exon3) per ng of total RNA. (G) Splicing of the first intron of gurken was completely abrogated by splice site mutations (grk^{splice site mutant}). Measurement of the gurken first intron spliced and unspliced isoforms relative to exon 3. Splice site mutant shows total retention, while in wild-type it was only a 0.03 ± 0.006 for unspliced ratio. Wild-type constructs show 1.26 ± 0.044 for first intron spliced transcripts, whereas the splice site mutant shows 1.29 ± 0.031 for unspliced transcripts. The percentage of unspliced first intron was calculated as the ratio unspliced/(spliced + unspliced). For grk^WT, the unspliced ratio is 2.5 ± 0.41%, whereas grk^{splice site mutant} completely abrogates first intron splicing (grk^{splice site mutant} = 100 ± 0.002%). At least two biological replicates were done for each experiment. Error bars indicate standard deviation.