Saccharomyces cerevisiae Ngl3p is an active 3'-5' exonuclease with a specificity towards poly-A RNA reminiscent of cellular deadenylases

Abstract

Deadenylation is the first and rate-limiting step during turnover of mRNAs in eukaryotes. In the yeast, Saccharomyces cerevisiae, two distinct 3'-5' exonucleases, Pop2p and Ccr4p, have been identified within the Ccr4-NOT deadenylase complex, belonging to the DEDD and Exonuclease-Endonuclease-Phosphatase (EEP) families, respectively. Ngl3p has been identified as a new member of the EEP family of exonucleases based on sequence homology, but its activity and biological roles are presently unknown. Here, we show using in vitro deadenylation assays on defined RNA species mimicking poly-A containing mRNAs that yeast Ngl3p is a functional 3'-5' exonuclease most active at slightly acidic conditions. We further show that the enzyme depends on divalent metal ions for activity and possesses specificity towards poly-A RNA similar to what has been observed for cellular deadenylases. The results suggest that Ngl3p is naturally involved in processing of poly-adenylated RNA and provide insights into the mechanistic variations observed among the redundant set of EEP enzymes found in yeast and higher eukaryotes.

Figure 1.

Ngl3p is a 3′–5′ exonuclease active at low pH. (A) Fluorescently labelled, mRNA-like substrates with either 10 or 21 adenosine residues at the 3′-end, where (F) represents the 5′-carboxyfluorescein label. (B) pH optimum of Ngl3p shown as the percentage of degraded RNA at different pH values. The pH optimum is pH 6.2 (dashed line). (C) Metal ion requirements. Gel showing degradation of RNA-polyA₁₀ in the presence of 5 mM Mg²⁺, 5 mM Mn²⁺, 1 mM Zn²⁺ and combinations of these. In lane 9, the reaction contained 7.1 mM Mg²⁺, 75 µM Mn²⁺ and 220 µM Zn²⁺(33), and in lane 10, 25 mM EDTA was added. The transiently stabilized fragment (21 nt) is indicated. (D) Degradation of RNA-polyA₁₀ by Ngl3p^wt over time. L is an RNA ladder made by partial alkaline hydrolysis. The RNA ladder sizes (30–20 nt) and the stabilized fragment (21 nt) are indicated and for the lane ‘H485A’, RNA-polyA₁₀ was incubated with the active site mutant, Ngl3p^H485A. (E) Degradation of RNA-polyA₂₁. The points of pausing at 10 and 8 nt are indicated.

Figure 2.

Ngl3p requires a 3′ poly-A tail and does not degrade dsRNA. (A) Oligos used in this figure, where (F) represents the 5′-carboxyfluorescein (FAM) label. (B) Degradation of RNA-intA₁₀, a substrate with an internal oligo-A tract. The sequence of the oligo is shown along the side. (C) Degradation of an RNA containing a strong secondary structure element followed by a 10 nt poly-A tail (RNA-stemloopA₁₀). The shortest product is 22 nt, corresponding to the stem structure plus two overhanging 3′ residues. (D) Degradation of a deoxynucleotide substrate having the same sequence as RNA-polyA₁₀.

Figure 3.

Ngl3p prefers poly-A RNA. (A) Parallel time-course degradation experiments with 10 nM RNA-polyA₁₀, RNA-polyU₁₀, RNA-polyC₁₀ and RNA-poly-G₁₀. The constant diffuse bands seen for RNA-polyC₁₀ around 12, 15 and 20 nt arise from absorption of the bromophenol blue dye at the TAMRA excitation wavelength. (B) Quantification of the degradation experiments in A converted to nanomolar NMP produced at each time point. Error bars represent the standard error of the mean (SEM) of three independent experiments. Errors are relatively large at longer time points due to weakening of the fluorescent label. (C) Measured affinities towards RNA and DNA substrates with different 3′ end sequences. The K_D values measured are 226 ± 63 nM (RNA-polyA₁₀), 195 nM ± 95 (RNA-polyU₁₀), 96 ± 22 nM (RNA-polyG₁₀) and 161 ± 37 nM (DNA-polyA₁₀). The error bars represent 95% confidence intervals.

Figure 4.

Ngl3p degrades a 3′ blocked but not a circular RNA substrate. (A) Degradation of RNA-endoA₁₀ by Ngl3p^wt and Rrp6p^12-536 (5′-end fragments). The lane marked ‘polyA’ is Rrp6p^12–536 incubated with RNA-polyA₁₀. (B) Incubation of Ngl3p^wt with a circular RNA substrate labelled internally with a TAMRA label. The lane marked ‘linear RNA’ is the phosphorylated RNA oligo prior to circularization.

Figure 5.

Deletion of ngl3 does not affect the growth of a ccr4Δ yeast strain. Growth phenotypes of the S. cerevisiae wt, ccr4Δ, ngl3Δ and ccr4Δ/ngl3Δ double deletion mutants. Ten-fold serial dilutions of cells were spotted onto agar plates and grown at the indicated temperatures.