The SR family proteins B52 and dASF/SF2 modulate development of the Drosophila visual system by regulating specific RNA targets

Abstract

Deciphering the role of alternative splicing in developmental processes relies on the identification of key genes whose expression is controlled by splicing regulators throughout the growth of a whole organism. Modulating the expression levels of five SR proteins in the developing eye of Drosophila melanogaster revealed that these splicing factors induce various phenotypic alterations in eye organogenesis and also affect viability. Although the SR proteins dASF/SF2 and B52 caused defects in ommatidia structure, only B52 impaired normal axonal projections of photoreceptors and neurogenesis in visual ganglia. Microarray analyses revealed that many transcripts involved in brain organogenesis have altered splicing profiles upon both loss and gain of B52 function. Conversely, a large proportion of transcripts regulated by dASF/SF2 are involved in eye development. These differential and specific effects of SR proteins indicate that they function to confer accuracy to developmental gene expression programs by facilitating the cell lineage decisions that underline the generation of tissue identities.

FIG. 1.

Expression of distinct SR proteins in the Drosophila developing eye induces phenotypes differing in their severity. (A) Stereomicroscopic views of adult compound eyes from representative male and female individuals of the indicated transgenic lines. Very few adult females and no males were obtained when males from the UAS-GFP-B52#6 were mated to virgin GMR-GAL4 females. (B) Emergence rate and sex ratio analyses of the progenies obtained following mating of males from the indicated transgenic lines with virgin GMR-GAL4 females. (C) Western blot analysis of the GFP-SR fusion proteins present in extracts purified from anterior quarters of third-instar transgenic larvae. Equal amounts of total proteins (30 μg per lane) were loaded onto the gel and probed with an anti-GFP monoclonal antibody following transfer onto nitrocellulose. The positions of molecular mass markers are indicated on the right. (D) The same nitrocellulose filter was reprobed with an anti-β-tubulin antibody to normalize for loading and transfer levels. (E) Western blot analysis with monoclonal antibody 104, revealing both the endogenous and exogenous B52 proteins in the GFP-NLS and GFP-B52 larval extracts (as described for panel C). (F) Western blot analysis with a polyclonal anti-dASF antibody, revealing both the endogenous and exogenous dASF/SF2 proteins in the GFP-NLS and GFP-dASF larval extracts.

FIG. 2.

GFP-dASF and GFP-B52 induce disorganization of R and cone cells in eye imaginal discs of transgenic third-instar larvae. (A) Direct fluorescence analysis of the GFP-NLS fusion protein expressed in the larval eye reveals the clusters of developing R cells that emerge in a rectangular array. (B and C) The R-cell organization is slightly altered in the GMR/GFP-dASF#1 larval eye (B) and strongly impaired in GMR/GFP-B52#6 transgenics (C). Magnification, ×20 or ×100 (insets). (D to F) Direct fluorescence analysis and DAPI (4′,6′-diamidino-2-phenylindole) staining of the whole developing eye confirm the differential alterations of R-cell organization in GMR/GFP-dASF#1 (E) and GMR/GFP-B52#6 (F) larval eyes compared to that observed in the control (D). The position of the MF is indicated by arrowheads. Magnification, ×20 or ×100 (insets). (G to I) Anti-Cut immunostaining of cone cells indicates that the number of these differentiated cells is significantly reduced in the developing eyes of GMR/GFP-dASF#1 (H) and GMR/GFP-B52#6 (I) third-instar transgenic larvae compared to that of the control (G). Magnification, ×40. (J to L) Anti-caspase 3 immunostaining of the control larval eye (J) reveals apoptotic cells located on the posterior side of the MF (arrowheads). In GMR/GFP-dASF#1 transgenics (K), numerous apoptotic cells are also disseminated throughout the posterior part of the eye imaginal disc, whereas in GMR/GFP-B52#6 larvae, apoptotic cells concentrate at the most posterior part of the eye (L). Magnification, ×20.

FIG. 3.

R-cell axons project aberrantly in optic lobes from GMR/GFP-B52#6 larvae and impair correct development of visual ganglia. Direct fluorescence analyses and DAPI staining indicate that GFP-SR fusion proteins are expressed in R-cell axons that extend from the optic stalk (os) and project into the lamina (la), and medulla (me) visual ganglia in both the control (A and B) and GMR/GFP-dASF#1 (E and F), but not in GMR/GFP-B52#6 (I and J), transgenic larvae. Direct fluorescence analysis and anti-fasciclin II immunostaining reveals R-cell axons as well as lamina monopolar cells that form arborizations in the distal medulla in both the control (C) and GMR/GFP-dASF#1 (H) transgenic larval optic lobes. In GMR/GFP-B52#6 larvae (K), lamina monopolar cells are not detected. The mushroom bodies also revealed by the anti-fasciclin II antibody are indicated (mb). Direct fluorescence analysis and anti-caspase 3 immunostaining indicate that optic lobes from GMR/GFP-B52#6 larvae contain major apoptotic foci (L) compared to those of control (D) and GMR/GFP-dASF#1 (G) transgenics. Magnifications, ×40 (A, B, E, F, I, and J) and ×25 (C, D, G, H, K, and L).

FIG. 4.

Coimmunoprecipitation experiments linked to microarray analyses identify potential dASF/SF2 and B52 target genes. (A) Western blot analysis of the GFP-SR fusion proteins present in the imput (I), supernatant (S), and pellet (P) fractions of the immunoprecipitation step. Protein extracts purified from anterior quarters of the indicated transgenic third-instar larvae were processed as described in Materials and Methods. Samples of the different fractions were probed with an anti-GFP monoclonal antibody. The three bands systematically observed in the pellet fractions correspond to mouse immunoglobulins revealed by the secondary antibody used for enhanced chemiluminescence detection. The positions of the molecular mass markers are indicated on the right, and the positions of the different GFP-tagged proteins are indicated on the left. (B) Sizes of the sets of dASF/SF2 and B52 candidate target genes belonging to three major functional classes (upper part) and subjected to alternative splicing (lower part). (C) Sizes of the sets of dASF/SF2 and B52 candidate target genes regulated by alternative splicing and belonging to three major functional classes (upper part) and histogram representation of the percentages of dASF/SF2 and B52 alternatively spliced candidate target genes included in these functional classes (lower part). (D) Histogram representation of the percentages of dASF/SF2 and B52 candidate target genes involved in eye and brain development (left) and frequencies of alternatively spliced (AS) target genes within the same subclasses (right). To categorize the functions of identified genes, QIAGEN Operon annotations have been cross-checked with those provided by the Flybase and PubMed databases, including mutant phenotypes, expression patterns, and gene functions.

FIG. 5.

RT-PCR validation of selected targets. (A) Validation of dASF/SF2 and B52 common target genes. (B) Validation of dASF/SF2 target genes. Alterations of splicing events annotated in Flybase (http://flybase.bio.indiana.edu) for the indicated genes were assessed in GMR/GFP-NLS, GMR/GFP-dASF#1, and GMR/GFP-B52#6 transgenic larvae by semiquantitative RT-PCR experiments. The schematic genomic organization of the gene region subjected to alternative splicing is represented on the right, and the different mRNA isoforms are designated according to the Drosophila melanogaster Exon Database (DEDB). Constitutive and alternative exons are represented by white and shaded boxes, respectively. The correspondence between mRNA isoforms and amplified products is shown on the right, and oligonucleotide pairs used to amplify the different isoforms are indicated by arrows. Quantitative ratios of the different mRNA isoforms were determined by densitometric analyses.

FIG. 6.

RT-PCR validation of selected B52 targets. Alterations of splicing events annotated for the indicated genes were assessed in GMR/GFP-NLS, GMR/GFP-dASF#1, GMR/GFP-B52#6, and UAS-BBS transgenic larvae as described for Fig. 5. The schematic genomic organization of the gene region subjected to alternative splicing is represented on the right. The correspondence between mRNA isoforms and amplified products is shown on the right, and oligonucleotide pairs used to amplify the different isoforms are indicated by arrows. Quantitative ratios of the different mRNA isoforms were determined by densitometric analyses.