A guard protein mediated quality control mechanism monitors 5'-capping of pre-mRNAs

Abstract

Efficient gene expression requires properly matured mRNAs for functional transcript translation. Several factors including the guard proteins monitor maturation and act as nuclear retention factors for unprocessed pre-mRNAs. Here we show that the guard protein Npl3 monitors 5'-capping. In its absence, uncapped transcripts resist degradation, because the Rat1-Rai1 5'-end degradation factors are not efficiently recruited to these faulty transcripts. Importantly, in npl3Δ, these improperly capped transcripts escape this quality control checkpoint and leak into the cytoplasm. Our data suggest a model in which Npl3 associates with the Rai1 bound pre-mRNAs. In case the transcript was properly capped and is thus CBC (cap binding complex) bound, Rai1 dissociates from Npl3 allowing the export factor Mex67 to interact with this guard protein and support nuclear export. In case Npl3 does not detect proper capping through CBC attachment, Rai1 binding persists and Rat1 can join this 5'-complex to degrade the faulty transcript.

Figure 1.

Npl3 binds pre-mRNAs with defective 5’-caps. (A) Deletion of NPL3, but not of GBP2 and HRB1, is synthetically lethal with mutations in 5’-capping factors. Ten-fold serial dilutions of the indicated strains were spotted onto –URA plates to select for the covering plasmid and onto FOA plates to assay the growth upon the loss of the covering plasmid. The plates were incubated for three days. n = 3. (B) Npl3 does not interact with the capping enzymes Cet1 and Ceg1. Western blot analysis of co-immunoprecipitation (co-IP) experiments with GFP-tagged Cet1 or Ceg1 and Npl3 are shown. Hem15 served as a negative control. n = 3. (C) The cet1-2 mutation produces uncapped mRNAs. Log phase cells were shifted for 1 h to 37°C. The RNA was isolated and incubated in vitro with the recombinant 5’-3’ degrading enzyme Xrn1. The remaining RNA was subsequently analyzed in qPCRs. n = 3. (D) Npl3 nuclear export is inhibited in mutants of the 5’-capping machinery. An at steady state cytoplasmic version of Npl3 (GFP-Npl3c) that is slower in nuclear re-import, was localized in the indicated strains after a 2 h temperature shift to 37°C. (E) The binding of Npl3 to mRNA is increased in mutants of the capping enzyme. RNA co-immunoprecipitation (RIP) experiments with Npl3 were carried out with wild type and cet1-2 mutants after a 1 h incubation at 37°C. Subsequent qPCRs analyzed the indicated transcripts. n = 3; *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 2.

Npl3 mediates the nuclear retention of pre-mRNAs with defective 5’-caps. (A) A leakage assay reveals mRNA leakage of usually retained false 5’-capped pre-mRNAs into the cytoplasm in cells deleted for NPL3, but not for GBP2 and HRB1. In situ hybridization (FISH) experiments of bulk mRNAs are shown in the indicated strains that were shifted for 3 h to 37°C. The poly(A)⁺RNA was visualized with a Cy3 labeled oligo d(T)₅₀ probe. The DNA was stained with Hoechst (blue). (B) Quantification of the measured nuclear and cytoplasmic signals shown in (A). n= 5. (C) FISH experiment with a single RNA species, the GFP RNA, expressed from the strong ADH1 promoter is shown in the indicated strains, shifted for 2 h to 37°C. The GFP transcript was visualized with a gene specific Cy3 labeled probe (red). (D) Quantification of the experiment shown in (C). n = 3. (E) Nucleo-cytoplasmic fractionation experiments reveal the leakage of transcripts to the cytoplasm in cells deleted for NPL3. The indicated strains were grown to log phase before they were shifted to 37°C for 1h. The cytoplasm was isolated and the containing mRNAs were analyzed in qPCRs. n = 4. (F) A deletion of NPL3 in cet1-2 shows RNA levels close to wild type. Cells were shifted for 1 h to 37°C. The RNA was isolated and incubated in vitro with the recombinant 5’-3’ degrading enzyme Xrn1. The remaining RNA was subsequently analyzed in qPCRs. G, The CBC is not a retention factor for mRNAs. A leakage assay as carried out in (Figure 2A) with cbp80Δ is shown. H, Quantification of the measured nuclear and cytoplasmic signals shown in (F). n = 3, *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 3.

Npl3 recruits the 5’-degradation machinery to uncapped transcripts. (A) NPL3 genetically interacts with mutants of RAT1 and RAI1. Serial dilutions of the indicated strain were spotted onto FOA plates and incubated at the indicated temperatures. n= 5. (B and C) Physical interaction of Npl3 with the Rat1–Rai1 complex. Western blot analysis of co-IPs with GFP-tagged Rat1 (B) or Rai1 (C) and Npl3 are shown. n = 3. (D) Npl3 is retained in the nuclei of mutants defective in the 5’ directed RNA-degradation. GFP-Npl3c was localized in the indicated strains after a 2 h temperature shift to 37°C. n = 5. (E) Increased mRNA binding of Npl3 is detected in rat1-1. Cells were shifted for 1 h to 37°C before the Npl3-bound RNAs were isolated. The indicated mRNAs were analyzed in qPCRs. n = 3. (F and G) Efficient interaction of Rat1 but not Rai1 with the mRNAs requires Npl3. qPCRs are shown from RIP experiments of Rat1 (F) or Rai1 (G) for the indicated transcripts in wild type and npl3Δ. RIP experiments were performed with Rat1-GFP and Rai1-GFP, respectively. n = 4. (F) n = 4 (G). (H) A nuclear mRNA export block does not change the reduced binding of Rat1 to faulty transcripts in npl3Δ. qPCRs are shown from RIP experiments of Rat1 for the indicated transcripts in mex67-5 and mex67-5 npl3Δ. n = 4; *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 4.

Proper 5’-capping is prerequisite for the interaction of CBC and Npl3. (A) Npl3 and Cbp80 physically interact. Co-IPs of Npl3 with GFP-tagged Cbp80 are shown with and without the addition of RNase Dre2 served as a negative control. n = 4. (B) The transcript binding of Cbp80 is reduced in cet1-2 mutants shifted to 37°C for 1 h. RIP-experiments with Cbp80 and subsequent qPCRs were carried out. n = 3. (C) The interaction of Npl3 and Cbp80 is disturbed in the presence of uncapped transcripts. Western blot analysis of co-IPs with Cbp80 and Npl3 was compared in the indicated strains. Hem15 served as a negative control. (D) Quantification of three different experiments one of which is shown in (C). n = 5. (E) Genetic interactions of cbp80Δ with the indicated strains are shown. Serial dilutions of the indicated strains are shown on FOA plates. n = 3; *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 5.

Npl3 binding to Mex67 or Rat1–Rai1 is mutually exclusive. (A) Mex67 binding to Npl3 is reduced in 5’-capping and 5’-degradation mutants. Western blot analysis of Mex67 co-IPs with Npl3 is shown. The indicated strains were shifted to 37°C for 2 h before lysis. (B) Quantification of experiments, one of which is shown in (A). n =8 for wild type, n =8 for cet1-2, n =6 for rat1-1. (C) Mex67-bound Npl3 does not interact with Rat1 and Rai1. A western blot of Npl3, Rat1 and Rai1 co-IPs with Mex67 is shown. (D) Rat1–Rai1-bound Npl3 shows no interaction with Cbp80. A western blot of Npl3 and Rat1 or Rai1 co-IPs with Npl3 and Cbp80 is shown. (E) Rat1- and Rai1-bound Npl3 is not Mex67 bound. A western blot of a Mex67 and Npl3 co-IPs with Rat1 and Rai1 is shown. (F and G) In the absence of CBP80, the Rat binding to transcripts is decreased while Rai1 binding is increased. RIP experiments with Rai1-GFP and subsequent qPCRs are shown for the indicated transcripts. n = 3 for (F), n = 4 for (G); *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 6.

Model for the Npl3-mediated nuclear mRNA quality control at 5’-caps. RNA-polymerase II (RNAP II) synthesized transcripts are capped by the capping machinery Cet1 and Ceg1. Upon correct capping, CBC associates. Npl3 and Rai1 bind to the 5’-end of the emerging transcript. The interaction of Npl3 with CBC results in the dissociation of Rai1 and allows the recruitment of Mex67. Upon completion of splicing and polyadenylation, signalled by the other guard proteins Gbp2, Hrb1 and Nab2, the mature mRNA is exported to the cytoplasm. Proper Mex67 coverage of the guard proteins is monitored by Mlp1 at the nuclear pore complex (NPC) and passage is allowed. On transcripts that have a defective 5’-cap, CBC cannot bind. The missing interaction of Npl3 and CBC prevents Mex67 binding and Rai1 can interact with Rat1. Npl3 dissociates and the Rat1–Rai1 complex degrades the pre-mRNA. In the absence of Npl3, incorrect 5’-capping is not detected and transcripts are not retained and degraded in the nucleus.