Computational protocol: Translational Regulation of Specific mRNAs Controls Feedback Inhibition and Survival during Macrophage Activation

Similar protocols

Protocol publication

[…] Cytoplasmic RNA and RNA from polysome fractions was quantified with GeneChip Mouse Gene 1.0 ST Arrays. Labeling, hybridization and scanning were performed by the GeneCore Genomics Core Facility at EMBL, Heidelberg. Random primers were used for cDNA synthesis with the Ambion WT Expression Kit to avoid any bias due to poly(A) tail length. Labeling was performed with the Affymetrix GeneChip WT Terminal Labeling Kit. Probe sequences of all perfect match probes were retrieved using the Bioconductor package oligo (version 1.22.0) . Probes were mapped to the mouse RefSeq transcriptome as downloaded from the UCSC Genome Browser mm10 refGene table on February 5, 2013. Probes with perfect complementarity to transcripts of more than one gene (as defined by a common gene symbol) were excluded. For mapping and further processing of probe information, the R packages seqinr and Biostrings were used together with in-house developed Perl scripts. Expression values were quantile normalized and summarized at the gene level with the basicRMA() function of the Bioconductor package oligo and the target gene symbols as probe set names. The different pools (cytoplasmic, free, 40S-associated, light and heavy) were pre-processed as separate groups (6 samples per group), because their signal distributions might differ due to biological and not technical reasons and therefore should not be quantile normalized together. To obtain the proportion of each mRNA in a specific pool, we had to take into account how much of each pool was used for quantification. After pre-processing, the signals were corrected for the different average proportions of each pool that were used for cDNA synthesis. For example, on average 14.2% (volume) of the free RNA pool (F, control condition) was used for cDNA synthesis, but only 0.4% of the heavy polysome pool (H, control condition). The corrected signal of an individual mRNA in a specific pool was then divided by the sum of its signal in all four pools. Only protein-coding genes with at least four specific probes and well detectable expression values in the cytoplasmic samples of treated and untreated cells were included into our analysis. Pre-processed expression values and the distribution over the four pools are represented in . [...] For quantification of RNA by RNASeq, RNA was purified with the EURx GeneMATRIX universal RNA purification kit, including a DNase on-column digestion. RNA libraries were prepared for sequencing using the NEBNext Ultra Directional RNA Library Prep Kit after ribosomal RNA was removed with the Ribo-Zero Magnetic Kit (Epicentre). Library preparation and sequencing was performed by the CellNetworks Deep Sequencing Core Facility at the University of Heidelberg. As spike-in controls, in vitro transcripts (rabbit HBB2 and firefly luciferase) were added at a concentration of 0.4 fmol per 1 µg RNA. Reads were mapped to the mouse RefSeq transcriptome as downloaded from the UCSC Genome Browser mm10 refGene table on February 5, 2013. The sequences of the in vitro transcripts were included in the index. For mapping, Bowtie was used allowing a maximum of two mismatches and reporting all alignments in the best stratum (settings: -a –best –stratum –v 2). With an in-house developed Perl script, read counts were summarized at the gene level discarding all reads that map to transcript isoforms of more than one gene (as defined by a common gene symbol). To calculate fold changes relative to the control condition, library size factors were estimated with the DESeq package .Expression patterns were obtained as follows: A maximum was defined as a time point with a significant increase compared to the last significant minimum (or the control). Unless the maximum is the last time point, it has to be followed by time points with a smaller or not significant fold change compared to the last significant minimum (or the control), until the end of the time course or the next significant minimum is reached. A minimum was defined in an analogous way. Significance was defined as a log2-transformed fold change of >0.5 for maxima or <0.5 for minima and a p-value of <0.05 (see Statistical Procedures). The group g0 contains all mRNAs without any significant changes compared to the control. G1 mRNAs have the first maximum at or after 1 h of stimulation, and no minimum before the first maximum. G2 mRNAs have the first minimum at or after 1 h of stimulation, and no maximum before the first minimum. G3 mRNAs have the first maximum before 1 h of stimulation and the first minimum at or after 1 h of stimulation. G4 mRNAs have their first minimum before 1 h of stimulation and the first maximum at or after 1 h of stimulation.Rpkm values were calculated with the following equation: The number of 58-mers (the read length) that are unique to the transcript isoforms of one gene was obtained with an in-house developed Perl script. Read counts and the number of unique 58-mers are represented in . […]

Pipeline specifications

Software tools seqinr, Biostrings, Bowtie, DESeq
Databases RefGene UCSC Genome Browser
Applications RNA-seq analysis, Genome data visualization
Organisms Mus musculus