Identification of a quality-control mechanism for mRNA 5'-end capping

Abstract

The 7-methylguanosine cap structure at the 5' end of eukaryotic messenger RNAs is a critical determinant of their stability and translational efficiency. It is generally believed that 5'-end capping is a constitutive process that occurs during mRNA maturation and lacks the need for a quality-control mechanism to ensure its fidelity. We recently reported that the yeast Rai1 protein has pyrophosphohydrolase activity towards mRNAs lacking a 5'-end cap. Here we show that, in vitro as well as in yeast cells, Rai1 possesses a novel decapping endonuclease activity that can also remove the entire cap structure dinucleotide from an mRNA. This activity is targeted preferentially towards mRNAs with unmethylated caps in contrast to the canonical decapping enzyme, Dcp2, which targets mRNAs with a methylated cap. Capped but unmethylated mRNAs generated in yeast cells with a defect in the methyltransferase gene are more stable in a rai1-gene-disrupted background. Moreover, rai1Δ yeast cells with wild-type capping enzymes show significant accumulation of mRNAs with 5'-end capping defects under nutritional stress conditions of glucose starvation or amino acid starvation. These findings provide evidence that 5'-end capping is not a constitutive process that necessarily always proceeds to completion and demonstrates that Rai1 has an essential role in clearing mRNAs with aberrant 5'-end caps. We propose that Rai1 is involved in an as yet uncharacterized quality control process that ensures mRNA 5'-end integrity by an aberrant-cap-mediated mRNA decay mechanism.

Figure 1. Rai1 preferentially hydrolyzes unmethylated capped RNA

(a) In vitro transcribed ³²P-cap-labeled or uniform-labeled pcP RNAs with a methylated or unmethylated cap were subjected to Rat1 or Rai1 proteins and the decay of the RNAs followed at the indicated times. The RNAs used are denoted on the right and the asterisk represents the position of the ³²P labeling. Quantitation of the amount of RNA remaining in the assays in (a) following Rai1 or Rai1 + Rat1 treatment are graphed in (b). Data are from three independent experiments normalized to a ³²P-labeled DNA oligonucleotide loading control included in the stop buffer and presented relative to time zero. Error bars represent +/− standard deviation (SD). (c) In vitro decapping assays were carried out at 37°C for 15 min as in (a) with the indicated proteins and decapping products were resolved by PEI-TLC developed in 0.45 M (NH₄)₂SO₄. Migration of the cap analog markers are shown on the right. Human Dcp2 was used as a methyl capped RNA decapping positive control (panels 2, 7 and panels 12, 17). (d) In vitro decapping assay using ³²P-cap-labeled methylated or unmethylated 5’ capped pcP RNAs were carried out with 50nM Rat1 and Rai1 recombinant proteins for the indicated times and resolved as in (c) above. Percent decapping of three independent experiments are presented on the bottom.

Figure 2. Rai1 preferentially hydrolyzes unmethylated 5’ end capped mRNA in yeast cells

Yeast strains harboring the temperature-sensitive ABD1 methyltransferase mutant allele, abd1-5, or the abd1–5 rai1Δ double mutation were grown at the 37°C nonpermissive temperature for 45 min prior to transcriptional block with 5 µg/ml thiolutin. RNA was isolated from cells at the indicated time points following thiolutin addition and levels of RNA remaining determined by Northern blot analysis (PGK1 and ACT1) or RT-qPCR (CYH2). Half-lives (t_1/2) of the mRNAs were determined relative to the 18S rRNA and were derived from three independent experiments. The range of half-lives obtained are consistent with previously reported thiolutin-directed transcriptional arrest measurements.

Figure 3. Rai1 functions to clear mRNAs in cells subjected to glucose or amino acid starvation

Wild-type (WT) or rai1Δ yeast strains were grown in complete medium at 30°C to an OD₆₀₀ of 0.6 and subsequently cultured in complete medium (a), glucose minus medium (b) or amino acid minus medium (c) for 30 min followed by addition of 5 µg/ml thiolutin. Total RNAs were isolated and detected at the indicated times following thiolutin addition by Northern Blot analysis and RNA half-lives (t_1/2) quantified from three independent experiments normalized relative to the 18S rRNA are presented.

Figure 4. Aberrantly capped mRNA levels increase in cells exposed to nutrient starvation

(a) Amino acid starvation shifts mRNAs into the soluble mRNP fraction. The mid-log phase yeast strains were shifted to the indicated medium and grown for 45 min prior to fractionation. RNA was isolated from polysome-containing fractions sedimenting at 130,000 × g (P130) and the supernatant (S130) fraction, which contained the soluble mRNP. Distribution of the CYH2 mRNA from each fraction was determined by quantitative RT-PCR. The abd1–5 rai1Δ double mutant strain grown at the permissive 25°C (methylated capped mRNA) or non-permissive 37°C (unmethylated capped mRNA) were used as a positive control. Results of three independent experiments are presented with error bars denoting +/− SD. (b) Aberrantly capped mRNAs are minimally affected by the Dcp2 decapping enzyme. The indicated strains were grown in complete medium at 22°C to an OD₆₀₀ of 0.6 and subsequently cultured in the same medium or amino acid minus medium for 45 min followed by the addition of thiolutin. The levels of CYH2 mRNA were determined by quantitative RT-PCR as in (a) above. (c) Methylated capped RNA was immunopurified utilizing monoclonal anti-trimethylguanosine antibody column from cells grown at the denoted culture conditions for 45 min and were detected by Northern Blot analysis. Quantitations for the mRNA cap methylation were normalized to total input RNA and ³²P-labeled methylated capped pcP RNA internal control (I.C.) and derived from three independent experiments. The error bars represent +/− SD.