DGC Stress Transcriptome.xlsx

Deficiencies in the human dystrophin glycoprotein complex (DGC), which links the extracellular matrix with the intracellular cytoskeleton, cause muscular dystrophies, a group of incurable disorders associated with heterogeneous muscle, brain and eye abnormalities. Stresses such as nutrient deprivation and aging cause muscle wasting, which can be exacerbated by reduced levels of the DGC in membranes, the integrity of which is vital for muscle health and function. Moreover, the DGC operates in multiple signaling pathways, demonstrating an important function in gene expression regulation. To advance disease diagnostics and treatment strategies, we strive to understand the genetic pathways that are perturbed by DGC mutations. Here, we utilize a Drosophila model to investigate the transcriptomic changes in mutants of four DGC components under temperature and metabolic stress. We identify DGC-dependent genes, stress-dependent genes and genes dependent on the DGC for a proper stress response, confirming a novel function of the DGC in stress-response signaling. This perspective yields new insights into the etiology of muscular dystrophy symptoms, possible treatment directions and a better understanding of DGC signaling and regulation under normal and stress conditions.  

RNA-sequencing Analyses

Samples were prepared for RNA sequencing as follows. One-week-old flies were used with the following genotypes: Canton S/w1118 (Control), Dg086/ Dg055 (Dg LOF), Df(3R)Exel6184/ Df(3R)Exel6184 (Dys LOF), Df(3 L)BSC450/ΔSyn15-2 (Syn1 LOF) and Df(2 L)BSC230/NosΔ15 (Nos LOF). Flies were kept on standard food for 5 days at 25°C in a 12 h/12 h light/dark cycle for the unstressed condition. In order to induce heat stress, flies were kept in an incubator at 33°C on standard food for 5 days. Metabolic stress was induced by feeding the flies on yeast paste only (yeast paste made from dry yeast mixed with 5% propionic acid in water) at 25°C for 4 days. For RNA extraction, ∼10 male flies per genotype and condition were homogenized together in Trizol (Ambion) and RNAwas extracted using the Direct-zol RNA mini-prep kit with an additional on-column DNAse digestion step (Zymo Research). The quality of the purified RNA was assessed with a Nanodrop ND-1000 spectrophotometer measuring the A260/A280 and A260/A230 ratios. Only RNA with A260/A280>2.0 and A260/A230>1.7 was used for further RNA-seq application. For each sample, 5-10 μg of total RNA was sent to GATC Biotech (Konstanz, Germany) for library preparation and subsequent transcriptome sequencing. In summary, RNA-seq libraries were prepared by RNA poly-A purification, fragmentation, random-primed cDNA synthesis, linker ligation and PCR enrichment. The samples were used to make a random-primed cDNA library, and the run was performed on an Illumina HiSeq platform with single-end, 100 bp reads. The transcriptome sequencing experiment resulted in a sample average of ∼7.5 million reads that could be mapped to a unique transcript. Genome index was generated from genome FASTA files of individual chromosomes (BDGP6 version) and the transcript annotation GTF file Drosophila_melanogaster. BDGP6.84 (ensemble.org database; Yates et al., 2016). Subsequently, the reads were mapped to the reference genome using STAR: ultrafast universal RNA-seq aligner (Dobin et al., 2013). Subsequent analyses were performed in R statistical software (https://www.r-project.org/) using packages from the Bioconductor project (Gentleman et al., 2004). The resulting BAM alignment files were used to generate counts on individual transcripts using the transcript database from the same GTF file [via the Rsamtools (https://bioconductor.org/packages/Rsamtools) and GenomicFeatures (Lawrence et al., 2013) packages] and the ‘summarizeOverlaps’ function in ‘Union’ mode via GenomicAlignments package (Lawrence et al., 2013). In total, 15,930 unique transcripts were detected with at least one count. Next, the counts were analyzed and the genotype and condition comparisons were done through built-in statistical models in the DESeq2 package (Love et al., 2014). For data visualization, gplots (https://cran.rproject. org/package=gplots) and RColorBrewer (Neuwirth, 2011) packages were used. For differential gene expression analysis, twofold difference and P-values smaller than 0.1 were considered significant and filtered by the ‘results’ function with ‘lfcthreshold=1’ and subsetting the genes with P <0.1. The resulting gene lists were subjected to gene interaction and ontology term analysis using STRING and DAVID databases, and ClueGO (Bindea et al., 2009) and CluePedia applications via Cytoscape software (Shannon et al., 2003).