Identification and functional analysis of novel phosphorylation sites in the RNA surveillance protein Upf1

Abstract

One third of inherited genetic diseases are caused by mRNAs harboring premature termination codons as a result of nonsense mutations. These aberrant mRNAs are degraded by the Nonsense-Mediated mRNA Decay (NMD) pathway. A central component of the NMD pathway is Upf1, an RNA-dependent ATPase and helicase. Upf1 is a known phosphorylated protein, but only portions of this large protein have been examined for phosphorylation sites and the functional relevance of its phosphorylation has not been elucidated in Saccharomyces cerevisiae. Using tandem mass spectrometry analyses, we report the identification of 11 putative phosphorylated sites in S. cerevisiae Upf1. Five of these phosphorylated residues are located within the ATPase and helicase domains and are conserved in higher eukaryotes, suggesting a biological significance for their phosphorylation. Indeed, functional analysis demonstrated that a small carboxy-terminal motif harboring at least three phosphorylated amino acids is important for three Upf1 functions: ATPase activity, NMD activity and the ability to promote translation termination efficiency. We provide evidence that two tyrosines within this phospho-motif (Y-738 and Y-742) act redundantly to promote ATP hydrolysis, NMD efficiency and translation termination fidelity.

Figure 1.

Identification of novel phosphorylation sites in Upf1. (A) Coomassie blue-stained 10% SDS-PAGE of immunopurified Flag-Upf1 protein. FT = flowthrough, W1–W5 = washes 1–5. (B) Immunopurified Flag-Upf1 protein was in-gel digested with trypsin. The peptides produced were resolved by LTQ linear ion trap mass spectrometry (Proteome X LTQ Workstation,Thermo) and analysed as described in Materials and Methods section. Asterisk (*) shows the Upf1 phosphorylated residues. (C) The arrow on the MS spectrum points toward the charged precursor ion corresponding to phosphopeptide 4. (D) This precursor ion was selected and subjected to fragmentation (MS2), generating b and y product ions that represent specific fragments used for identification of the peptide sequence and phosphorylation sites. The asterisks (*) illustrate phosphorylated amino acids.

Figure 2.

Eleven phosphorylation sites were identified in Upf1. (A) Schematic representation of the Upf1 protein depicting the locations of the cysteine- and histidine-rich (CH-rich) domain, ATPase domain and helicase domain. The localization of the 11 phosphorylation sites identified by MS/MS in Upf1 protein is represented. (B) Amino acid sequence of Upf1 protein in S. cerevisiae. Sixty-seven percent (67%) sequence coverage was achieved by tandem MS/MS. Sequence coverage is depicted in gray, and phosphorylated residues are underlined.

Figure 3.

Phosphorylated residue within the ATPase and helicase domain of Upf1 is completely conserved. Sequence alignment of S. cerevisiae (NCBI NP_013797), H. sapiens (NCBI AAC26788), M. musculus (NCBI EDL28813), A. thaliana (NCBI AAL92018), D. melanogaster (NCBI AAF48115.2) and C. elegans (NCBI AAC26789) shows complete conservation of one phosphorylated residue, Y754F. This alignment was constructed with CLUSTALW using T-Coffee program (version 9.02) (63). Completely conserved sequences are shown in gray.

Figure 4.

The phosphorylated motif-4 of Upf1 is essential for NMD. (A) Schematic representation of Upf1 mutants. Phosphorylated residues are represented as gray rectangles. In motif 1 amino acids T50–N60 were deleted, in motif 2 amino acids L192–L201 were deleted, in motif 3 V490–S495 were eliminated and in motif 4 residues R736–K756 were eliminated. (B) Western blot analysis of cytoplasmic extracts demonstrating Upf1 protein expression. Poly(A) binding protein (Pab1) was used as a loading control. (C) NMD activity of deletions of Upf1 phosphorylated motifs were determined by Northern blot analysis of total cellular RNA. (D) Mean value ± standard deviation of pre-CYH2 mRNA accumulation from Northern blot in (C). (E) Dual luciferase assay was conducted to determine the efficiency of UGA stop codon recognition. Two-tailed t-tests were used for statistical analysis. The asterisk indicates a statistical significant (P < 0.05) result when compared to the WT strain.

Figure 5.

Upf1 region encompassing amino acids 736–745 is essential for NMD. (A) Schematic representation of Upf1 indicating the deleted amino acids within the motif-4. Each deletion contains a phosphorylated residue (underlined). (B) Western blot analysis of cytoplasmic extracts demonstrating Upf1 protein expression. Poly(A) binding protein (Pab1) was used as a loading control. (C) NMD activity of Upf1 phosphorylated motif-4 deletions determined using Northern blot analysis of total cellular RNA. (D) Mean value ± standard deviation of pre-CYH2 mRNA accumulation from Northern blot in (C). (E) Dual luciferase assay was conducted to determine the efficiency of UGA stop codon recognition. Two-tailed t-tests were used for statistical analysis. The asterisk indicates a statistical significant (P < 0.05) result when compared to the WT strain.

Figure 6.

The Upf1 phosphorylated Y-738 residue and the adjacent Y-742 are important for NMD activity. (A) Schematic representation of Upf1 indicating the deleted amino acids in the 736–745 region. The Y738 phosphorylated residue is underlined in gray, and the potential phosphorylated residue, Y742, is underlined in black. (B) Western blot analysis of cytoplasmic extracts demonstrating Upf1 protein expression. Poly(A) binding protein (Pab1) was used as a loading control. (C) (Left Panel) NMD activity of Upf1 phosphorylated residues within 736–745 region measured by pre-CYH2 to mature mRNA ratio and determined by Northern blot analysis of total cellular RNA. (Right Panel) Mean value ± standard deviation of pre-CYH2 mRNA accumulation from Northern blot. (D) (Left Panel) NMD activity of Upf1 phosphorylated residues within 736–745 region measured by can1-100 mRNA levels and determined by Northern blot analysis. (Right Panel) Mean value ± standard deviation of can1-100 mRNA accumulation from Northern blot. (E) Dual luciferase assay was conducted to determine the efficiency of UGA stop codon recognition. Two-tailed t-tests were used for statistical analysis. The asterisk indicates a statistical significant (P < 0.05) result when compared to the WT strain.

Figure 7.

The ATP hydrolysis of Upf1 is inhibited in phosphorylation mutants. WT and mutant form of the Upf1 protein was purified as described in Materials and Methods section. (A) Purified proteins were analysed by SDS-PAGE and analysed by Western blotting using monoclonal antibody against Flag epitope (Sigma). (B) The ATPase activities of the WT and mutant form of Upf1 were determined using 5 ng of purified proteins (see Materials and Methods section). The Upf1 ATPase mutant DE572AA was used as a control.