ProtocolsPlasmids and molecular cloning
The PolyA-Pull plasmid was constructed adapting the Gateway pDONR221 (Invitrogen) as follows. The pab-1 gene was amplified from N2 genomic DNA using a forward specific primer containing a SacII site and a reverse specific primer containing BamHI and EcoRI sites. The amplicon was then ligated in-frame with GFP (Marco Mangone, unpublished) using T4 DNA Ligase (NEB, Ipswich, MA) and SacII and EcoRI sites. The 3xFLAG epitope DNA sequence was obtained from the DNASU Plasmid Repository (DNASU clone ID: HsCD00298297), and extracted using PCR amplification using a forward primer containing a BamHI site and reverse primer containing an EcoRI site. The amplicon was then ligated into pDONR221, (Invitrogen) downstream and in-frame with the pab-1 gene using T4 DNA Ligase (NEB, Ipswich, MA). The pab-1-Pull plasmid (GFP::Δpab-1::3xFLAG), which does not contain the pab-1 sequence and cannot bind polyA+ mRNAs, was prepared from the PolyA-Pull plasmid using the Stratagene QuikChange® Site-Directed Mutagenesis Kit following the manufacturer’s guidelines (Stratagene, La Jolla, CA).
The 3'UTR of the unc-54 gene cloned in Gateway pDONR P2R-P3 entry vectors was used as an unspecific 3'UTR in all of the destination vectors in this study. The tissue-specific promoters were selected as the genomic sequence of DNA upstream of their transcription start site up to 2kb. We have designed the primers using the UCSC Genome Browser and cloned the resultant amplicons from N2 genomic DNA into the Gateway pDONR P4-P1R entry plasmid (Invitrogen, Carlsbad, CA).
We have used Multisite recombination reactions (LR Clonase plus II, Invitrogen) to join the tissue specific promoters, the PolyA-Pull vector and the unc-54 3'UTRs into the Gateway Compatible MosSCI destination plasmid pCFJ150, (Addgene plasmid #19329), and used these vectors for the preparation of the transgenic strains.
Nematode strains and preparation of transgenic animals
Wild-type strain N2 worms were obtained from the CGC (University of Minnesota), which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). Worm strain EG4322 (to prepare MosSCI transgenics) were maintained at 16°C on HB101 containing NGM agar plates prior to microinjection (as described in wormbuilder.org). Stable transgenic worm strains were prepared using the MosSCI technology as described (PMID 18953339). Microinjection mixes consisting of pJL43.1(50ng/µl), pCFJ90(1ng/µl), pGH8(10ng/µl), pCFJ104(5ng/µl), pCFJ150::TissuePromoter::GFP::pab-1::3xFlag::unc-54 (25ng/µl) were microinjected into worm strain EG4322 (ttTi5605; unc-119(ed9) III), each of which was kindly provided by Priscilla Van Wynsberghe (Colgate University).
Microinjection was carried out using a Leica DMI3000B microscope according to that described previously. Injected worms were plated on NGM growth media plates containing OP51 bacteria and plates containing unc-119 rescued (mobile) worms were chunked onto four new NGM plates and left to starve for at least 30 days at 25°C. Single dauer worms were plated onto small NGM plates, propagated for approximately 2 weeks and verified for GFP expression using a Leica DMI3000B. DIC and fluorescent images were captured using a Leica DFC345FX mounted camera.
Worm gDNA extraction and MosSCI insertion verification
Genomic DNA was phenol-chloroform extracted from one full 60mm NGM plate from each transgenic worms strain, precipitated with sodium acetate and washed in ethanol. To confirm the MosSCI integration of transgenes into the ttTi5605 intergenic region, we performed PCR using Standard Taq Polymerase (NEB, Ipswich, MA) using a forward primer annealing outside of the homologous flanking region (5'- CCTCTGAACTGGTACCTCA -3') and a reverse primer annealing within the unc-119 rescue cassette (5'- GGAAGAAGGAAAAGAGTGTGG -3'), both of which were provided by Priscilla Van Wynsberghe (Colgate University).
The mRNA tagging technique was adapted from past studies. Mixed-stage liquid worm cultures were grown as described at 25°C. Approximately 106 pab-1::3xFLAG transgenic worms were harvested from liquid culture after 3 to 4 days, crosslinked for one hour in 0.5% paraformaldehyde in M9 solution, and flash frozen in ethanol-dry ice bath. Frozen pellets were crushed using a mortar and pestle in liquid nitrogen and the resulting frozen powder was transferred directly into lysis buffer (150mM NaCl, 25mM HEPES, pH 7.5, 0.2mM DTT, 10% glycerol, 0.0625% RNAsin, 1% Triton X-100). Total RNA was extracted from worm lysates using Trizol® Reagent (Life Technologies, Carlsbad, CA) and precipitated with isopropanol. An amount of lysate corresponding to 90µg of total RNA was added to 100µl of Anti-FLAG M2 Magnetic Beads (Sigma-Aldrich, St. Louis, MO) and incubated overnight at 4°C.
Each reaction was washed 3X in 200µl TBS and then 3X in 200µl Proteinase-K buffer with 1000 RPM mixing. Proteinase-K (4mg/ml) was added to the beads and incubated at 37°C for 30 minutes with 1000 RPM mixing. 7M Urea was added to beads and incubated at 37°C at 1000 RPM before RNA was extracted with Trizol® Reagent and precipitated with isopropanol and GlycoBlue (Ambion, Austin, TX). Precipitated RNA was treated with DNAse I (NEB, Ipswich, MA) for ten minutes and extracted again with Trizol® Reagent and isopropanol. RNA was resuspended in nuclease-free water and quantified using a Nanodrop® 2000c spectrophotometer (Thermo-Fisher Scientific, Waltham, MA).
Bioinformatics analysis of RNA-Seq data
Raw Reads Mapping: Paired raw reads were demultiplexed by their unique tissue-specific barcodes and converted individually to FASTQ files by the CASAVA software (Illumina). Unique datasets were then mapped to the C. elegans gene model WS190 using the Burrows-Wheeler Aligner software (BWA) with default parameters.
Expression levels of individual transcripts were estimated from the bam files by using Cufflinks software. The fragment per kilobase per million base (FPKM) number was used to indicate the gene expression levels, and FPKM value >=1 was used as a threshold across all tissues profiled for defining expressed genes. The gene expression levels obtained in each tissue dataset were compared pairwise with other tissues using the Cuffdiff algorithm. Cuffdiff algorithm detected 389 isoforms shared between pharynx and intestine, 286 between body muscle and intestine, and 175 between the two muscle tissues (p-value<0.05).
PolyA cluster preparation and polyA mapping
To map polyA-sites to WS190 worm annotations, raw sequence reads were filtered using custom made Perl scripts. We extracted reads containing greater than or equal to 30 consecutive adenine nucleotides at their 3'end. We obtained 14,472 total reads from intestine, 7,532 for pharynx and 5,185 for body muscle. The polyA elements were then removed and the reads were converted to FASTA format and aligned to the WS190 annotation using the Burrows-Wheeler Aligner with standard parameters. Reads mapping to genomic regions containing >=65% adenosines in either direction and /or with less than 18 consecutively mapped nucleotides were discarded.
The reads produced approximately ~27,000 high-quality PAS clusters mapped through the C. elegans genome. Each of these clusters was then bioinformatically attached to the closest gene within a 1,600nt range in the same orientation. To increase the stringency of our analysis we ignored clusters with less than 5% of the total number of polyA reads detected for a given gene, and PAS clusters that map genomic regions with >40% adenosines, to eliminate as much background as possible. Each cluster has a median length of ~70nt with ~5x coverage, and maps 3'UTRs of genes detected in the corresponding tissue with a FPKM>=1.
Mapped polyA sites were compared with Mangone et al. and Jan et al. to map common 3'UTR isoforms between these datasets. We assigned common PolyA sites if the overlap is between +-10nt. PAS usage was calculated as in Mangone et al. PAS position and PAS nucleotide composition for 3'UTR isoforms in each dataset was extracted from Mangone et al.
Recent 3' UTR Papers
Comparative RNA-Seq analysis reveals pervasive tissue-specific alternative polyadenylation in Caenorhabditis elegans intestine and muscles. BMC Biology 2015 Jan 20
Analysis of C. elegans intestinal gene expression and polyadenylation by fluorescence-activated nuclei sorting and 3'-end-seq. Nucleic Acids Res. 2012 Jul 40(13):6304-18
Regulation of mRNA translation and stability by microRNAs Annu Rev Biochem. 2010;79:351-79.
- March 18 2016
The 3'UTRome website has been moved to ASU
The UTRome v1 was initially developed at NYU in 2011 by Marco Mangone, Kris Gunsalus and Fabio Piano for the modENCODE Consortium. The server was recently moved to the School of Life Sciences at Arizona State University, and is now maintained by Marco Mangone's group at ASU
The UTRome Project is currently funded by the School of Life Sciences at Arizona State University