The D. melanogaster strains studied were recently collected in France and in Brazil. Two novel methods were developed in the host laboratory: chromatin immunoprecipitation (ChIP) and RAMPAGE. In Drosophila, active chromatin is often associated with trimethylation of lysine 4 in histone 3 (H3K4me3), while repressive chromatin and more importantly, repressed TE sequences, are associated with trimethylation of lysine 9 of histone 3 (H3K9me3). Cross link chromatin immunoprecipitation was performed in drosophila embryos and the genome-wide sequencing data should be available in the following months. RAMPAGE is an assay that produces genome-wide libraries of transcripts with transcription start sites (TSS) intact, originally performed in Drosophila samples. I have troubleshooted RAMPAGE in our strains and in order to have matched epigenetic libraries (CHIP-seq) and RAMPAGE libraries, I have waited to perform RAMPAGE in the same pool of tissues where the chromatin was extracted for the ChIP analysis. Hence, the RAMPAGE libraries are currently being made and will be sent to sequence in the next weeks.
During the troubleshooting period of the ChIP and the delays imposed by technical problems, I have developed two in silico projects: the first one aims at deciphering the role of TE sequences in differential expression of genes between Drosophila strains but as opposed to the previous proposal, this project will detect TSSs, exonizations and truncations. The second project aims at uncovering gene expression differences between strains of Drosophila simulans that are due to differential piRNA production. PiRNAs are well known regulators of TEs. Both projects are novel and took advantage of genome-wide data produced in the host laboratory allowing immediate use. We found that less than 2% of genes contain a TE sequence within its transcripts. Interestingly, "chimeric genes" present higher expression than genes lacking any TE sequences and in addition an over representation of chimeric genes is found in upregulated compared to stable or downregulated genes between drosophila strains. We are currently confirming such results in candidate genes that are upregulated and harbor TE sequences within their transcripts. In sum, this first in silico project suggests that in Drosophila, TEs could potentially increase the expression of host genes. Concerning the second in silico project, we have shown that small RNAs of 23-29bp with a ping pong signature are potentially able of targeting host genes. Interestingly, these piRNAs do not necessarily map to a TE sequence, suggesting either that the TE sequences are too degraded or that genes are able to produce their own piRNAs. But the most curious finding is that most genes are targeted by such piRNAs, with or without TE similarities. These piRNAs target coding domain sequences and are depleted from 5' untranslated regions, where most regulatory sequences are found. We are currently mining for gene expression between populations differences due to these piRNAs.
