Structure of the Nmd4-Upf1 complex supports conservation of the nonsense-mediated mRNA decay pathway between yeast and humans

Abstract

The nonsense-mediated mRNA decay (NMD) pathway clears eukaryotic cells of mRNAs containing premature termination codons (PTCs) or normal stop codons located in specific contexts. It therefore plays an important role in gene expression regulation. The precise molecular mechanism of the NMD pathway has long been considered to differ substantially from yeast to metazoa, despite the involvement of universally conserved factors such as the central ATP-dependent RNA-helicase Upf1. Here, we describe the crystal structure of the yeast Upf1 bound to its recently identified but yet uncharacterized partner Nmd4, show that Nmd4 stimulates Upf1 ATPase activity and that this interaction contributes to the elimination of NMD substrates. We also demonstrate that a region of Nmd4 critical for the interaction with Upf1 in yeast is conserved in the metazoan SMG6 protein, another major NMD factor. We show that this conserved region is involved in the interaction of SMG6 with UPF1 and that mutations in this region affect the levels of endogenous human NMD substrates. Our results support the universal conservation of the NMD mechanism in eukaryotes.

Fig 1. The Nmd4 protein wraps around Upf1 RecA1 domain.

(A) Schematic representations of S. cerevisiae Upf1 helicase domain and Nmd4 protein with domain boundaries indicated. The different protein domains are shown with a color code, which is used in all the figures. (B) Cartoon representation of the structure of the Nmd4/Upf1-HD complex. The left and right orientations are related by a 90° rotation along the horizontal axis. (C–F) Zoom-in on 4 regions (boxed in panel B) of the Nmd4/Upf1-HD interface. Hydrogen bonds and salt bridges are depicted as black dashed lines. HD, helicase domain.

Fig 2. Nmd4 influences Upf1 ATPase activity.

(A) Characterization of the interaction between His₆-ZZ-Nmd4 full-length protein (upper left panel), His₆-ZZ-Nmd4 « arm » (upper right panel), His₆-ZZ-Nmd4 PIN domain (lower left panel) or the His₆-ZZ tag (lower right panel), and Upf1-HD by ITC. Upper panel: ITC data obtained by injecting the different Nmd4 constructs or the His₆-ZZ tag into Upf1-HD. Lower panel: Fitting of the binding curves using a single binding site model. Three biologically independent experiments were performed and a representative image is shown. The data underlying this figure can be found in S1 Data. (B) Nmd4 stimulates Upf1-HD ATPase activity. Purified Upf1-HD alone or in combination with full-length or fragments of Nmd4 or His-ZZ (control) was used in a ³²P release assay, as a readout of the ATPase activity. The relative activity (indicated in percent) corresponds to the activity measured in a given condition (different proteins) normalized to the activity of Upf1-HD alone after 30 min of reaction at 30°C. The relative activity for each condition after 30 min is indicated on the right of the different curves. The data underlying this figure can be found in S2 Data. HD, helicase domain; ITC, isothermal titration calorimetry.

Fig 3. The interaction of Nmd4 with Upf1 is important for NMD.

(A) Zoom-in on the interaction network formed by the side chain of Arg210 from Nmd4. Hydrogen bonds are depicted as black dashed lines. The Cα atom of Gly377 from Upf1-HD is shown as a sphere. (B) Zoom-in on the side chain of Trp216 from Nmd4 lying on helix α1 from Upf1-HD stalk region. The Cα atom of Gly243 from Upf1-HD is shown as a sphere. (C) CBP-pull-down experiment showing that the Upf1-HD single mutants (G243R or G377R) or the double mutant (DM; G243R/G377R) affect the binding to CBP-tagged Nmd4. Input and eluate from calmodulin sepharose beads (IP) samples were analyzed on 12% SDS/PAGE and Coomassie blue staining. As controls, each Upf1-HD variant was incubated with the CBP-His₆-ZZ tag, revealing some weak nonspecific retention of the various Upf1-HD proteins on the calmodulin sepharose beads. Degradation products enriched on calmodulin sepharose beads are indicated by an asterisk (*). Three biologically independent experiments were performed and a representative image is shown. (D) Single point mutations in full-length Upf1 disrupt its interaction with Nmd4 in vivo. Co-purification of Nmd4-HA with TAP-Upf1-FL, TAP-Upf1-FL-G243R, and TAP-Upf1-FL-G377R was evaluated by immunoblot. A negative control experiment used the N-terminal region of Upf1 (TAP-Upf1-CH, encompassing residues 1 to 208) known to be unable to interact with Nmd4 [45]. (E) Upf1 deficient for binding to Nmd4 is impaired for its NMD role. A upf1Δ strain was transformed with plasmids expressing N-terminal tagged full-length Upf1 (pUpf1), to restore NMD, a truncated WT version of the protein lacking the CH domain (UPF1-HD-Ct), which can partially complement NMD, or UPF1-HD-Ct single point mutants (G243R or G377R) that are unable to interact with Nmd4. A double mutant upf1Δ/nmd4Δ strain shows an exacerbated NMD defect phenotype, which is not rescued by the expression of UPF1-HD-Ct single point mutants (G243R or G377R) that are unable to interact with Nmd4. Two natural NMD substrates, the uORF containing DAL7 transcript (left) and the unspliced pre-mRNA for the ribosomal protein Rpl28 (right) were used to estimate NMD efficiency and their levels were measured by RT-qPCR. The obtained values were normalized for total RNA amounts using PGK1 and spliced RPL28 transcripts and all the results were adjusted to the situation in which full-length Upf1 was expressed (marked “pUpf1”). The experiments were repeated at least 3 times and the individual results are indicated as dots. Error bars represent standard deviation. The p-values of t tests (Welch variant with continuity correction) for comparisons between results are indicated. The data underlying this figure can be found in S3 Data. (F) Pull-down experiment showing that the His₆-ZZ-Nmd4 single point mutants (R210E or W216A) or the double point mutant (M2; R210E/W216A) affect the binding to Upf1. Input and eluate from NiNTA magnetic beads samples were analyzed on 12% SDS/PAGE and Coomassie blue staining. As control, Upf1-HD was incubated with the His₆-ZZ tag. Three biologically independent experiments were performed and a representative image is shown. (G) The Nmd4 mutants are less potent activators of Upf1-HD ATPase activity than WT Nmd4. The assay was similar to that presented in Fig 2B. The data underlying this figure can be found in S4 Data. HD, helicase domain; NMD, nonsense-mediated mRNA decay; uORF, upstream open reading frame.

Fig 4. An evolutionary conserved « arm » in metazoan SMG6 is critical for the interaction with UPF1-HD and influences the levels of NMD substrates.

(A) Zoom-in on the sequence alignments of the C-terminal end of fungal Nmd4 proteins and of the metazoan SMG6 region encompassing residues 445–465 from human SMG6. Strictly conserved residues are in white on a black background. Strongly conserved residues are in bold. Residues mutated in this study are highlighted by red boxes. This figure was generated using the ENDscript server [76]. (B) The SMG6 R448 and W456 residues corresponding respectively to R210 and W216 from Nmd4 are important for the phospho-independent interaction between human UPF1-HD and SMG6-[207–580] domain in cellulo. Co-immunoprecipitation experiments showing the reduced interaction of the 3 human SMG6-[207–580] point mutants (R448E, W456A, M2) with human UPF1-HD (residues 295–914) as compared to human SMG6-[207–580] WT. The human UPF1-HD-3xFlag and the different human SMG6-[207–580]-HA variants were expressed in HEK293T cells and purified using anti-Flag beads. Co-immunoprecipitated proteins were detected by western blot using the indicated antibodies. Co-IP experiments were performed in the presence of benzonase to exclude a potential role of nucleic acids in these interactions. Three biologically independent experiments were performed and a representative image is shown. (C) The phospho-independent SMG6-UPF1 interaction contributes to NMD in human cells. Top: Schematic representation of human SMG6 protein with domain boundaries indicated. EBM: EJC-binding motif. 14.3.3 (also known as TPR for Tetratrico Peptide Repeat) is a domain involved in the recognition of phosphorylated Ser or Thr. The different protein domains are shown with a color code. Mutants investigated in this study are indicated below the diagram. Bottom: Box plot representation of the effect of SMG6 mutations on the stability of endogenous NMD substrates. Human cells were depleted for endogenous SMG6 using shRNAs and transformed using either empty plasmid, or various SMG6 constructs resistant to the shRNAs. Four previously described endogenous NMD substrates (highlighted in dark red) and 2 NMD insensitive transcripts (labeled in black) were quantified using RT-qPCR. The RQ value of 1 observed for each transcript in parental cell lines is depicted by an horizontal black dashed line. Statistical differences were determined from 9 data points (3 technical replicates for each of the 3 biological replicates) using an independent two-tailed Student’s t test to compare RQ values for the « arm » mutants (M2 and M7) with those of the corresponding controls (WT and M5, respectively). *** p < 0.001; * 0.01 < p < 0.05; #: not statistically significant. The data underlying this figure can be found in S5 Data. (D) Proposed models for the interaction of Upf1/UPF1 with Nmd4/SMG6 and Ebs1/SMG5-7 proteins in fungi (left) and metazoa (right). Phosphorylated Ser/Thr are depicted as red squares. Putative phosphorylation sites in yeast Upf1 are indicated with a question mark (?). EJC, exon junction complex; NMD, nonsense-mediated mRNA decay; RQ, relative quantification.