A synthetic A tail rescues yeast nuclear accumulation of a ribozyme-terminated transcript

Abstract

To investigate the role of 3' end formation in yeast mRNA export, we replaced the mRNA cleavage and polyadenylation signal with a self-cleaving hammerhead ribozyme element. The resulting RNA is unadenylated and accumulates near its site of synthesis. Nonetheless, a significant fraction of this RNA reaches the cytoplasm. Nuclear accumulation was relieved by insertion of a stretch of DNA-encoded adenosine residues immediately upstream of the ribozyme element (a synthetic A tail). This indicates that a 3' stretch of adenosines can promote export, independently of cleavage and polyadenylation. We further show that a synthetic A tail-containing RNA is unaffected in 3' end formation mutant strains, in which a normally cleaved and polyadenylated RNA accumulates within nuclei. Our results support a model in which a polyA tail contributes to efficient mRNA progression away from the gene, most likely through the action of the yeast polyA-tail binding protein Pab1p.

FIGURE 1.

Ribozyme-terminated RNA is predominantly unadenylated. (A) Schematic of the TDH3 promoter-driven GFP constructs: (Txn) transcription start site, (pA) TDH3 polyadenylation signal, (RZ) hammerhead ribozyme element, (A75 RZ) 75 DNA encoded adenosine residues. The start and stop codons of GFP are indicated by ATG and TAA, respectively. All constructs were introduced into yeast on 2μ high-copy plasmids. (B) Primer extension analysis of total (T) RNA, and of flowthrough (FT) and bound (B) RNAs to oligo dT. RNA analysis for total and flowthrough are of 1/10th the amount analyzed for the bound fraction. Primer extension for U2 snRNA is included as an internal control. Construct names have shortened here and in subsequent figures for simplicity. (C) Northern blot analysis for GFP 3′ end fragments. Total RNA from cells was treated with RNase H and an oligonucleotide to internally cleave GFP RNA, in the absence (−) and presence (+) of oligo dT to remove polyA stretches. GFP pA migrates more slowly due to the TDH3 3′ UTR in this RNA.

FIGURE 2.

Synthetic A tail rescue of transcription site-associated RNA accumulation. (A) FISH analysis for GFP RNA in cells transformed with empty vector, GFP pA, GFP RZ, or GFP A75 RZ. (B) Combined FISH for GFP RZ (Cy3 channel) and indirect immunofluorescence for lacI-GFP (FITC channel). LacI-GFP expressing cells were transformed with a 2μ high-copy plasmid harboring both the GFP RZ reporter and 256 tandem repeats of lacO sequence. A merged image is shown to the right. Images in B have been additionally software enhanced for clarity.

FIGURE 3.

Rescue of nuclear accumulation by a synthetic A tail of 48 adenosines. (A) FISH for GFP RNA in cells transformed with constructs containing various synthetic tails. In addition to GFP pA, GFP RZ and GFP A75 RZ are constructs in which, immediately upstream of the ribozyme, is either 50 bp of TDH3 3′ UTR sequence (GFP 50 RZ), DNA encoded stretches of 12, 24, 36, 48, 60, or 150 adenosines (GFP A12 RZ, GFP A24 RZ, etc.), or 75 DNA encoded uridine residues (GFP T75 RZ). (B) Primer extension for GFP RNA and U2 snRNA in strains transformed with these synthetic tail constructs. (C) α-GFP Western blot for GFP protein generated from these constructs.

FIGURE 4.

FISH for RNAs stabilized by an XRN1 deletion or destabilized by NMD. (A) Primer extension analysis for GFP pA, GFP RZ, and GFP A75 RZ in the wild-type W303 and ΔXRN1 strains. Primer extension for U2 snRNA is also shown. (B) Northern blot analysis for GFP RNAs containing a premature termination codon (+PTC series). (C) FISH for GFP RNA in the W303 and ΔXRN1 strains and for the PTC-containing RNAs. The FISH analyses shown were performed at the same time. The panels for the +PTC series have been additionally software enhanced to better visualize the faint nuclear foci in GFP pA +PTC and GFP A75 RZ +PTC.

FIGURE 5.

Polysome association of ribozyme-terminated RNA. (Top) Representative profile of the absorbance at 254 nm (A254) across the 10%–50% sucrose gradient. Peaks for 40S small ribosomal subunits, 60S large ribosomal subunits, and 80S monosomes are indicated. Twelve fractions were collected. (Bottom) Northern blot analysis for GFP pA, GFP RZ, and GFP 50 RZ across sucrose gradients. The amount of each RNA in the polysome fractions (fractions 7–12) as a percentage across the entire gradient are given to the right of each panel.

FIGURE 6.

Transcriptional induction of ribozyme-terminated RNA and synthetic A tail rescue. (A) Primer extension analysis for GFP RNA in strains transformed with GAL1 promoter-driven GFP at time points after transcriptional induction. 3′ ends were directed either by a GAL1 polyadenylation signal (GALGFP pA), ribozyme (GALGFP RZ), or synthetic A tail (GALGFP A75 RZ). Constructs were introduced into yeast on 2μ high-copy plasmids. Primer extension for U2 snRNA is also shown. Time 0 represents uninduced cells. (B) FISH analysis for GFP RNA at time points after transcriptional induction.

FIGURE 7.

Export of a synthetic A tail-containing RNA in 3′ end formation mutant strains. (A) FISH for GFP RNA in strains temperature sensitive for polyA polymerase (pap1-1) or a cleavage factor (rna15-2) transformed with the GAL1 promoter-driven constructs described in Figure 6. Cultures at 25°C were temperature shifted to 37°C with superheated media containing galactose for 40 min prior to fixation and FISH analysis. Uninduced cells (time 0) for GALGFP pA are also shown. (B) FISH analysis for GFP RNA in the wild-type W303 strain and a strain deleted for the nucleoporin Rip1p using the TDH3 promoter-driven constructs. Cultures at 30°C were temperature shifted to 42°C for 30 min prior to fixation and FISH analysis. The exposure time in B is one-eighth typical exposure times to accommodate the strong nuclear accumulation in the ΔRip1 strain.

FIGURE 8.

Pab1p association is restored by a synthetic A tail. (A) Primer extension analysis for GFP RNA after α-V5 immunoprecipitation (IP) from extracts of transformed CBP20 or PAB1 V5 epitope-tagged strains. Primer extension for U2 snRNA is also shown. Input RNA from 1/100th of the extract used for IP is shown on the left. Only a typical input is shown as the levels of GFP pA, GFP RZ, and GFP A75 RZ were unaffected in these strains (data not shown). Primer extension after IP from the tagged extracts, and from untagged W303 extract from GFP pA transformed cells, is shown on the right. (B) Quantitation of the amounts of GFP pA, GFP RZ, and GFP A75 RZ coimmunoprecipitated with Cbp20V5p and Pab1V5p. The GFP/U2 ratios in the IPs were divided by the GFP/U2 ratios in the inputs, and the value for GFP pA was normalized to 1.0. Results from three independent experiments (expt) for CBP20V5 and two independent experiments for PAB1V5 are shown, as indicated.