Wenxue Zhao, Joshua L. Pollack, Denitza P. Blagev, Noah Zaitlen, Michael T. McManus et al.
Functional characterization of noncoding sequences is crucial for understanding the human genome and learning how genetic variation contributes to disease. 3' untranslated regions (UTRs) are an important class of noncoding sequences but their functions remain largely uncharacterized. We developed a method for massively parallel functional annotation of sequences from 3' UTRs (fast-UTR) and used this approach to measure effects of a total of >450 kb of 3' UTR sequences from >2000 human genes on steady-state mRNA abundance, mRNA stability and protein production. We found widespread regulatory effects on mRNA levels that were coupled to effects on mRNA stability and protein production. Furthermore, we discovered 87 novel cis-regulatory elements and measured the effects of genetic variation within known and novel 3' UTR motifs. This work shows how massively parallel approaches can improve functional annotation of noncoding sequences, advance our understanding of cis-regulatory mechanisms and quantify effects of human genetic variation.
Wenxue Zhao, Joshua L. Pollack, Denitza P. Blagev, Noah Zaitlen, Michael T. McManus, David J. Erle. Massively parallel functional annotation of 3′ untranslated regions. 2014-03-16. https://doi.org/10.1038/nbt.2851
A pooled library of overlapping 128-nt segments spanning the CXCL2 3′-UTR was used in BEAS-2B cells to measure the effect of spontaneous mutations in 8-nt windows on reporter steady-state mRNA. The table contains the published genome-browser track values for each window center.
An error-prone-PCR fast-UTR library introduced random mutations into conserved 160-nt human 3′-UTR segments, and doxycycline transcriptional shutoff time courses in BEAS-2B cells were used to estimate reporter mRNA half-lives.
Random mutations introduced by error-prone PCR into conserved 160-nt 3′-UTR reporter inserts were analyzed in BEAS-2B cells to identify 6–20-nt intervals whose mutations altered reporter mRNA stability. The table contains the paper’s FDR-filtered element-level results and motif annotations.
The high-fidelity conserved 3′-UTR fast-UTR library was transduced at low multiplicity into BEAS-2B cells, and cells in the top or bottom 15% of GFP/LNGFR reporter-protein ratios were sorted. The table combines the published segment lists enriched in the high- and low-protein gates.
The high-fidelity conserved fast-UTR library tested 160-nt segments from conserved human 3′-UTRs in WiDr-tTA cells. Reporter RNA/cDNA-to-genomic-DNA ratios were summarized as segment-level steady-state mRNA activity.
The high-fidelity conserved fast-UTR library tested 160-nt segments from conserved human 3′-UTRs in BEAS-2B-tTA cells. Reporter RNA/cDNA-to-genomic-DNA ratios were summarized as segment-level steady-state mRNA activity.
The high-fidelity conserved fast-UTR library tested 160-nt segments from conserved human 3′-UTRs in Jurkat-tTA cells. Reporter RNA/cDNA-to-genomic-DNA ratios were summarized as segment-level steady-state mRNA activity.
A 67-nt CXCL2 3′-UTR segment containing ARE1 and flanking sequence was tested with all 201 possible single-nucleotide substitutions in BEAS-2B-tTA cells. Reporter steady-state mRNA effects were reported as variant-versus-reference delta mRNA from pooled RNA/DNA clone measurements.
Original supplemental and deposited inputs retained for this study. Download files individually or together as a ZIP; nested folders are preserved. Source reuse terms apply, and sequencing reads may be omitted.
Download all 9 files (ZIP)source_notes.txtsupplementary_data_1_cxcl2_fast_utr_tracks.txtsupplementary_data_2_conserved_3utr_sequences.xlssupplementary_data_3_3utr_mrna_effects_stability.xlssupplementary_data_4_3utr_protein_sort_enrichment.xlssupplementary_data_5_mirna_profile_beas2b.xlssupplementary_data_6_fast_utr_cis_elements.xlssupplementary_data_7_conserved_3utr_fast_utr_tracks.txtsupplementary_text_figures_notes.pdf