Macrophage inflammation resolution requires CPEB4-directed offsetting of mRNA degradation

Wildtype and Cpeb4 KO BMDMs were stimulated with lipopolysaccharide (LPS) and mRNA levels were quantified by RNAseq (n = 4). Differential expression between genotype conditions was performed with DESeq2 R package. Wildtype and Cpeb4 KO samples were compared for each time point independently. Sample information, alignment statistics, and principal component analysis are shown.

Wildtype and Cpeb4 KO bone marrow-derived macrophages (BMDMs) were left untreated or stimulated with lipopolysaccharide (LPS) for 9 hr. Immunoprecipitation (IP) with anti-CPEB4 antibody was then performed, and RNA was extracted and analyzed by RNAseq. CPEB4 targets were defined based on the enrichment between wildtype and Cpeb4 KO IPs.

For each gene, the reference sequence of the longest 3′-UTR was selected. The number of AREs and CPEs was calculated by scanning the corresponding 3′-UTR and counting the number of occurrences of each motif. The ARE/CPE score was calculated as the log2 transformed ratio between the number of ARE and CPE motifs.

Data were obtained from TTP and HuR PAR-iCLIP experiments in lipopolysaccharide (LPS)-stimulated bone marrow-derived macrophages (BMDMs) (Sedlyarov et al., 2016). Only mRNAs with HuR/TTP binding in the 3′-UTR were considered. For TTP, mRNAs bound after 3 hr or 6 hr of LPS stimulation were considered. For HuR-bound mRNAs, PAR-iCLIP data corresponded to 6 hr of LPS stimulation.

Wildtype BMDMs were stimulated with lipopolysaccharide (LPS), and mRNA levels were quantified by RNAseq (n = 4). Wildtype and Cpeb4 KO samples were compared for each time point independently. The expression pattern over time for wildtype samples was analyzed (differential expression results against consecutive time points and Rlog data normalized by maximum).

ARE- and CPE-containing mRNAs were defined as all mRNAs regulated by CPEB4, tristetraprolin (TTP), or Hu-antigen R (HuR) in lipopolysaccharide (LPS)-stimulated bone marrow-derived macrophages (BMDMs) (see also Supplementary files 2 and 4).

To analyze CPE-A-containing mRNAs, the script developed by Piqué et al. was run over mm10 3′-UTR reference sequences (Biomart ENSEMBL archive February 2014). mRNAs containing a putative 3′-UTR with a cytoplasmic polyadenylation element (CPE)-mediated repression and/or activation prediction were considered CPE-containing mRNAs. For CPE-G-containing mRNAs, the same script was adapted to mRNAs containing a putative 3′-UTR with the TTTTGT motif within the optimal distances to the polyadenylation signal (PAS) established by Piqué et al., 2008.The original and modified script, as well as the input and output files of the CPE-A and CPE-G analysis, are included.