Interaction between Ski7p and Upf1p is required for nonsense-mediated 3'-to-5' mRNA decay in yeast

Abstract

Aberrant mRNAs containing premature termination codons (PTC-mRNAs) are degraded by a conserved surveillance system, referred to as the nonsense- mediated decay (NMD) pathway. Although NMD is reported to operate on the decapping and 5'-to-3' exonucleolytic decay of PTC-mRNAs without affecting deadenylation, a role for an opposite 3'-to-5' decay pathway remains largely unexplored. In this study, we have characterized the 3'-to-5' directed mRNA degradation in the yeast NMD pathway. PTC-mRNAs are stabilized in yeast cells lacking the components of 3'-to-5' mRNA-decay machinery. The 3'-to-5' directed degradation of PTC-mRNAs proceeds more rapidly than that of the PTC-free transcript, in a manner dependent on the cytoplasmic exosome and Upf proteins. Moreover, Upf1p, but not Upf2p, interacts physically with an N-terminal domain of Ski7p, although the interaction requires Upf2p. The efficiency of 3'-to-5' directed degradation of PTC-mRNAs is impaired by overexpression of Ski7p N-domain fragments that contain a sequence of the Upf1p-interaction region. These data suggest that the activation of 3'-to-5' directed NMD is mediated through the interaction between Upf1p and the Ski7p N domain.

Fig. 1. Yeast strains lacking the components of 3′-to-5′ mRNA-decay machinery accumulate an aberrant mRNA containing PTC, PGK1_N103pG. (A) Yeast strains, wild type (WT, YS001), ski7Δ (YS002), ski7/CΔ (YS063), ski2Δ (YS060), ski3Δ (YS061), ski8Δ (YS014), rrp4-1 (YS094) and upf1Δ (YS065), which carry the PGK1_N103pG reporter (pRP611), were grown at 26°C in a galactose-containing medium and harvested at mid-log phase. The aberrant mRNA was detected by northern blot analysis with end-labeled oligonucleotides specific for the PGK1_N103pG (upper panel). The signals were quantified using a PhosphorImager and corrected for loading using the SRP RNA, scR1 (lower panel). The values under PGK1_N103pG northern blot show relative intensity to the amount of mRNA obtained in the WT strain. (B) Yeast strains, wild type (WT, YS001), ski7Δ (YS002) and xrn1Δ (YS065), which carry the PGK1_N103pG reporter (pRP611), were grown to mid-log phase in the galactose medium and shifted to a glucose medium to repress transcription of the reporter mRNA. Time points represent minutes after transcriptional repression. The half-lives (t_1/2; min) of PGK1_N103pG mRNA are shown as means ± standard deviations (SD), which were obtained from at least three independent experiments.

Fig. 2. 3′-to-5′ mRNA degradation is accelerated in the aberrant mRNA containing PTC. (A) Yeast dcp1Δ strain (YS055) carrying PGK1pG (top panel and open circles) or PGK1_N103pG (bottom panel and closed circles) reporters was grown to mid-log phase in the galactose medium and shifted to a glucose medium to repress transcription of the reporter mRNAs. After incubation for the indicated times, northern blot analysis was performed as described in Figure 1, and the amounts of transcript are illustrated as the function of the times. (B) Wild-type strain (YS001) that carries PGK1 _N103pG (top panel) or PGK1pG (bottom panel) reporter was grown in a minimal medium containing galactose and shifted to the medium supplemented with glucose and cycloheximide to inhibit the transcription and decapping. After incubation for the indicated times, northern blot analysis was performed as described in Figure 1. The sample from the 0-min time that had been cleaved by RNaseH with oligo dT was also applied to the upper left lane (dT), which indicates the position of poly(G)–3′-end fragments lacking poly(A). The half-lives (t_1/2; min) of mRNAs [full-length forms and fragments lacking poly(A) tail for A and B, respectively] are shown as means ± SD, which were obtained from three independent experiments.

Fig. 3. 3′-to-5′ directed NMD is mediated through the recognition of PTC and requires both the exosome and the Ski proteins. (A) Yeast xrn1Δ gst1-1 (YS106) strains that carry PGK1_N103pG reporter and either YCplac22-GST1 (bottom panel) or the empty vector (top panel) were grown at 25°C in the galactose medium to mid-log phase and incubated at 37°C for 1.5 h. (B) Yeast strains lacking the DCP1 gene (dcp1Δ, YS055; dcp1Δ rrp4-1, YS095) or containing its temperature-sensitive allele (dcp1-2, YS088; dcp11-2 ski2Δ, YS089; dcp1-2 ski3Δ, YS090; dcp1-2 ski7Δ, YS091; dcp1-2 ski8Δ, YS092) which carry the PGK1_N103pG reporter were grown at 25°C in the galactose medium to mid-log phase and incubated at 37°C for 1.5 h. Transcription was repressed by the addition of glucose. After incubation for the indicated times, northern blot analysis was performed as described in Figure 1. The half-lives (t_1/2; min) of mRNAs are shown as means ± SD, which were obtained from at least three independent experiments.

Fig. 4. Acceleration of 3′-to-5′ directed NMD is mediated through Upf proteins. Yeast strains lacking XRN1 and/or UPF genes (xrn1Δ, YS064; xrn1Δ upf1Δ, YS070; xrn1Δ upf2Δ YS071; xrn1Δ upf3Δ, YS072), which carry PGK1_N103pG (A) and PGK1pG (B) reporters, were grown at 25°C in the galactose medium to mid-log phase. Transcriptional repression and hybridization were performed as described for Figure 1. The half-lives (t_1/2; min) of mRNAs are shown as means ± SD, which were obtained from at least three independent experiments.

Fig. 5. The N-terminal domain of Ski7p associates with Upf1p in a manner dependent on Upf2p. Cell extracts were prepared from SKI7- and/or UPF1/2-deletion strains that carry Upf1p-Myc (A; lanes 1–4, 9–12 for YS037–YS040 and lanes 5–8 for YS045–YS048) or Upf2p-Myc (B; lanes 1–4, 9–12 for YS041–YS044 and lanes 5–8 for YS049–YS052). The yeast strains also expressed the indicated forms of FLAG-tagged Ski7p. The extracts (lanes 1–8) were directly subjected to immunoprecipitation (IP) with an anti-FLAG. In some experiments, the extracts that had been treated with RNase were also subjected to the immunoprecipitation assay (lanes 9–12, RNase-treated). The precipitated proteins (bottom panels), together with the whole-cell extracts (top panels), were immunoblotted (IB) with an anti-Myc (9E10) antibody to detect Upf1p-Myc and Upf2p-Myc.

Fig. 6. Mapping of Upf1p-binding sites on the N-terminal domain of Ski7p. Cell extracts were prepared from yeast strains expressing Upf1p-Myc and the indicated forms of FLAG-tagged Ski7p (A, lanes 1–6 for YS037, YS101, YS103, YS104, YS038, YS039; and B, lanes 1–7 for YS037, YS038, YS123, YS124, YS125, YS126, YS127). The extracts were immunoprecipitated (IP) with the anti-FLAG antibody. The precipitated proteins (bottom panel) and whole-cell extracts (top panel) were immunoblotted (IB) with the anti-Myc antibody to detect Upf1p-Myc. The asterisks in (B) indicate FLAG-tagged Ski7p mutants expressed in the various strains.

Fig. 7. Interaction of the N-terminal domain of Ski7p with Upf1p is required for 3′-to-5′ directed NMD. The xrn1Δ strains expressing PGK1_N103pG (A) and PGK1pG (B) reporters were transformed with plasmids carrying the indicated deletion mutants of Ski7p/N domain (vector, YS077; Ski7p/1–96, YS082; Ski7p/80–264, YS083; Ski7p/80–184, YS085; Ski7p/168–264, YS084; Ski7p/101–184, YS128; Ski7p/80–163; YS132). The transformants were grown in the galactose- containing medium and harvested at mid-log phase. Transcriptional repression and hybridization were performed as described for Figure 1. The half-lives (t_1/2; min) of mRNAs are shown as means ± SD, which were obtained from at least three independent experiments.

Fig. 8. The N domain of Ski7p interacting with Upf1p. Deletion mutants of Ski7p used in this study are shown, together with the summary of present results obtained in Figures 6 and 7.