Gene expression profiling of brakeless mutant Drosophila embryos

The transcriptional co-regulator Brakeless performs many important functions during Drosophila development, but few target genes have been identified. Here we use Affymetrix microarrays to identify Brakeless-regulated genes in 2–4 h old Drosophila embryos. Robust multi-array analysis (RMA) and statistical tests revealed 240 genes that changed their expression more than 1.5 fold. We find that up- and down-regulated genes fall into distinct gene ontology categories. In our associated study [2] we demonstrate that both up- and down-regulated genes can be direct Brakeless targets. Our results indicate that the co-repressor and co-activator activities of Brakeless may result in distinct biological responses. The microarray data complies with MIAME guidelines and is deposited in GEO under accession number GSE60048.


a b s t r a c t
The transcriptional co-regulator Brakeless performs many important functions during Drosophila development, but few target genes have been identified. Here we use Affymetrix microarrays to identify Brakeless-regulated genes in 2-4 h old Drosophila embryos. Robust multi-array analysis (RMA) and statistical tests revealed 240 genes that changed their expression more than 1.5 fold. We find that up-and down-regulated genes fall into distinct gene ontology categories. In our associated study [2] we demonstrate that both up-and down-regulated genes can be direct Brakeless targets. Our results indicate that the co-repressor and co-activator activities of Brakeless may result in distinct biological responses. The microarray data complies with MIAME guidelines and is deposited in GEO under accession number GSE60048. &

Data
To identify Brakeless target genes in the early embryo, RNA was isolated from embryos derived from brakeless (bks 278 ) germline clones, which lack the maternal contribution of Brakeless. This was compared to RNA from germline clone embryos generated with the unmutagenized FRT chromosome on which the bks 278 allele was induced [8]. The RNA was converted to cDNA and hybridized to an Affymetrix array. Misregulated genes that change their expression more than 1.5 fold were identified (supplementary material Table 1). We compared our gene list to an RNA-seq dataset that distinguishes maternal from zygotic transcripts in Drosophila embryos using polymorphisms [7]. The Brakeless-regulated genes were categorized as being maternally, zygotically, or maternally and zygotically (matzyg) derived (supplementary material Table 1). They were also subjected to functional annotation analysis using DAVID [3], which groups genes into clusters based on co-association with gene ontology (GO) terms (supplementary material Table 1). As shown in Fig. 1, up-regulated and down-regulated gene fall into distinct GO clusters.

Germline clones
The FLP-FRT dominant female sterile technique previously described in Ref. [1] was used to generate brakeless germline clones. The bks 278 allele was used, which has a 345 bp deletion that causes a frame shift at aa 741 resulting in addition of 79 novel amino acids [5]. It was outcrossed with a w 1118 strain to remove potential second-site mutations. FRT 2R G13 bks 278 /CyO females were crossed with males of the genotype hs-FLP/Y; FRT 2R G13 ovo D1 /CyO, derived from FRT 2R G13 ovo D1 /T(1;2)OR64/CyO (Bloomington stock #4344). Offspring larvae were heat-shocked for 3 h at 37°C on days 3, 4 and 5 after egg-laying to induce expression of the FLP recombinase. Cyþ females were crossed to FRT 2R G13 bks 278 /CyO males and embryos were collected, dechorionated using bleach, and directly frozen in Trizol at À 80°C for RNA isolation. Corresponding crosses, embryo collection and RNA isolation was performed with an unmutagenized FRT 2R G13 chromosome.

Microarray analysis
Staged 2-4 h old embryos were collected and immediately frozen at À 80°C prior to RNA extraction. Total RNA was isolated using TRIzol (Invitrogen) and purified using an RNeasy kit (Qiagen) according to the manufacturer's protocols. Forty ul of embryos were used for each of three biological replicates of embryos derived from FRT 2R G13 bks 278 or FRT 2R G13 c px sp control germline clones. cDNA probes were hybridized to an Affymetrix Drosophila gene chip (version 2). The intensity values were normalized and summarized with the robust multi-array analysis (RMA) method [6], using R (www. R-project.org) and the Bioconductor package [4].

Statistical analysis
After RMA normalization, 640 probes passed a 41.5 fold difference in median expression levels between the conditions investigated. They were subjected to a two tailed unpaired Student´s t-test with a 95% confidence level, followed by Bonferroni´s correction for multiple tests, resulting in a P-value cut-off at 0.000078125. After removal of probe-sets targeting duplicates and pseudogenes, 240 genes remained.

GO analysis
The lists of genes with significantly changed expression levels containing 174 down-and 66 upregulated identifiers were used as input lists for the DAVID Functional Annotation Clustering tool [3]. The tool provides analysis of annotation content and gene ontology term enrichments, to highlight the most relevant GO terms associated with a gene list. The enrichment score is a geometric mean of the member´s P-values in a -log scale within an annotation cluster.