Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2013 Nov;19(11):1510-6.
doi: 10.1261/rna.038919.113. Epub 2013 Sep 4.

Developmental Arrest of Drosophila Survival Motor Neuron (Smn) Mutants Accounts for Differences in Expression of Minor Intron-Containing Genes

Free PMC article

Developmental Arrest of Drosophila Survival Motor Neuron (Smn) Mutants Accounts for Differences in Expression of Minor Intron-Containing Genes

Eric L Garcia et al. RNA. .
Free PMC article

Abstract

Reduced levels of survival motor neuron (SMN) protein lead to a neuromuscular disease called spinal muscular atrophy (SMA). Animal models of SMA recapitulate many aspects of the human disease, including locomotion and viability defects, but have thus far failed to uncover the causative link between a lack of SMN protein and neuromuscular dysfunction. While SMN is known to assemble small nuclear ribonucleoproteins (snRNPs) that catalyze pre-mRNA splicing, it remains unclear whether disruptions in splicing are etiologic for SMA. To investigate this issue, we carried out RNA deep-sequencing (RNA-seq) on age-matched Drosophila Smn-null and wild-type larvae. Comparison of genome-wide mRNA expression profiles with publicly available data sets revealed the timing of a developmental arrest in the Smn mutants. Furthermore, genome-wide differences in splicing between wild-type and Smn animals did not correlate with changes in mRNA levels. Specifically, we found that mRNA levels of genes that contain minor introns vary more over developmental time than they do between wild-type and Smn mutants. An analysis of reads mapping to minor-class intron-exon junctions revealed only small changes in the splicing of minor introns in Smn larvae, within the normal fluctuations that occur throughout development. In contrast, Smn mutants displayed a prominent increase in levels of stress-responsive transcripts, indicating a systemic response to the developmental arrest induced by loss of SMN protein. These findings not only provide important mechanistic insight into the developmental arrest displayed by Smn mutants, but also argue against a minor-intron-dependent etiology for SMA.

Keywords: Drosophila insulin-like peptide 8; RNA-sequencing; dILP8; minor intron; snRNP; spinal muscular atrophy; splicing; stress signaling pathways; survival motor neuron; transcriptome.

Figures

FIGURE 1.
FIGURE 1.
Developmental arrest phenotype of Smn mutants. (A) University of California Santa Cruz (UCSC) genome browser snapshot of mapped RNA-seq reads at the Smn locus from wild-type and Smn mutant animals. Track counts in the wild-type (Ore-R) bedGraph file were normalized using the median of track counts from the Smn mutant bedGraph files. (B) Dendrogram of wild-type (Ore-R) and Smn mutants compared with the developmental transcriptome data from second instar (L2), third instar 12 h post-molt (L312hr), third instar puffstage 1 to 2 with dark blue, bromophenol blue fed guts (L3P1_2), third instar puffstage 3 to 6 with light blue guts (L3P3_6), and the last larval puffstage with clear guts (L3P7_9) (Graveley et al. 2011). Wild-type and Smn mutant RNA-seq data are each the average of two biological samples. Dendrogram line segments correspond to Jensen-Shannon distances (Trapnell et al. 2010). (C) Venn diagrams of DEXSeq, MISO, and Cuffdiff analysis of differences between wild-type and Smn mutant RNA-seq results. For comparison, DEXSeq and MISO results were converted into a nonredundant list of genes that overlap the noted differences in exon usage and alternative splicing events, respectively.
FIGURE 2.
FIGURE 2.
Relatively unaltered minor-intron-containing mRNA levels in Smn mutants. (A) Pairwise comparison of Cuffdiff FPKM levels between wild type and the Smn mutant. FPKM levels were normalized to the wild type, and error bars indicate standard deviation. (B) Heatmap comparison of minor-intron-containing mRNA FPKM levels over development and between wild type and the Smn mutant. Heatmap colors were rescaled for each row, and rows were manually reordered based on trends in gene expression over development. CG33108 was excluded here because Cufflinks’ FPMK analysis did not distinguish between CG33108 and LSm3.
FIGURE 3.
FIGURE 3.
Prohibitin 2 protein levels unchanged in Smn mutants. (A) UCSC browser snapshot of normalized RNA-seq tracks at the Prohibitin 2 gene locus. (B) Western blot of Prohibitin 2 (Phb2). Phb2 levels were analyzed in protein lysates from age-matched wild-type (Ore-R), Smn mutant (Smn), and U6atac mutant (U6atac) larvae as well as a later-stage wandering wild-type larva (Ore-R late L3). The anti-prohibitin antibody recognizes the minor-intron-containing PHB2, upper band, and PHB1, lower band, which doesn't contain a known U12-type minor intron. The Drosophila mitochondrial ATP synthase bellwether (BLW) served as a loading control.
FIGURE 4.
FIGURE 4.
Increase in Stasimon mRNA minor intron retention in Smn mutants and early in the third instar larval stage. (A) UCSC browser snapshot of normalized RNA-seq tracks at the Stasimon gene locus and diagram of junction spanning reads used to measure intron retention. (Gray lines and numbers) Exon–intron junction reads; (black lines and numbers) exon–exon junction reads. (Gray brackets and arrows in gene model) Minor intron locations. (B) Stasimon mRNA minor intron retention levels as a percentage of exon–intron (intron) junction spanning reads over total junction spanning reads between wild type, mutant, and over the developmental time from L2 to L3P7_9. Actual junction-spanning read counts and calculations are also provided in Supplemental Table S7. For comparison with our paired-end data, we used the Celniker paired-end modENCODE developmental data for this analysis (Supplemental Table S7).
FIGURE 5.
FIGURE 5.
Increased stress-responsive target genes in Smn mutants. (A) Stress-responsive transcripts that were significantly increased in the Smn mutants expressed here as fold increase above wild-type levels (from average Cufdiff FPKM values) (Supplemental Table S1). (B) UCSC browser snapshot of normalized RNA-seq tracks at the dILP8 gene locus.

Similar articles

See all similar articles

Cited by 24 articles

See all "Cited by" articles

Publication types

MeSH terms

Substances

Associated data

LinkOut - more resources

Feedback