The hierarchy of exon-junction complex assembly by the spliceosome explains key features of mammalian nonsense-mediated mRNA decay

Abstract

Exon junction complexes (EJCs) link nuclear splicing to key features of mRNA function including mRNA stability, translation, and localization. We analyzed the formation of EJCs by the spliceosome, the physiological EJC assembly machinery. We studied a comprehensive set of eIF4A3, MAGOH, and BTZ mutants in complete or C-complex-arrested splicing reactions and identified essential interactions of EJC proteins during and after EJC assembly. These data establish that EJC deposition proceeds through a defined intermediate, the pre-EJC, as an ordered, sequential process that is coordinated by splicing. The pre-EJC consists of eIF4A3 and MAGOH-Y14, is formed before exon ligation, and provides a binding platform for peripheral EJC components that join after release from the spliceosome and connect the core structure with function. Specifically, we identified BTZ to bridge the EJC to the nonsense-mediated messenger RNA (mRNA) decay protein UPF1, uncovering a critical link between mRNP architecture and mRNA stability. Based on this systematic analysis of EJC assembly by the spliceosome, we propose a model of how a functional EJC is assembled in a strictly sequential and hierarchical fashion, including nuclear splicing-dependent and cytoplasmic steps.

Figure 1. Ordered, spliceosome-mediated assembly of the EJC.

(A) Splicing reactions using MINX as substrate RNA were supplemented with extracts expressing the indicated FLAG-tagged EJC proteins or unfused FLAG-tag as negative control. Reactions were immunoprecipitated with FLAG affinity gel. Positions of the unspliced transcript and the spliced product are displayed schematically. Twelve percent of the total input material was loaded in the input panel. (B) Expression of the FLAG-proteins used in (A) determined by immunoblot analysis with a FLAG antibody. (C) Splicing reactions and immunoprecipitations were performed as in (A) with an intronless MINX transcript (MINX Δi). (D) Splicing reactions and immunoprecipitations with FLAG-CBP80 or unfused FLAG employing the intronless MINX transcript (MINX Δi) used in (C). (E) Splicing reactions and immunoprecipitations were performed as in (A) with a mutated MINX transcript (MINX GG) that does not undergo exon ligation. Positions of the unspliced transcript and the splicing intermediates (exon 1; lariat and exon 2) are displayed schematically. 8% of the total input material was loaded in the input panel. (F) Splicing reactions were performed in the presence (lanes 2, 4, 6, 8, 10, 13, 15, 17, 19, and 21) or absence (lanes 3, 5, 7, 9, 11, 14, 16, 18, 20, and 22) of FLAG-protein extracts or with unfused FLAG as negative control (lanes 1 and 12). Reactions without added FLAG-proteins were supplemented after the completion of splicing with FLAG-protein extracts and incubated on ice (30 min). Immunoprecipitations were done as in (A), except that 15% of the total input material was loaded in the input panel.

Figure 2. MAGOH-Y14 are recruited to the EJC by eIF4A3.

(A) Scheme of the positions of the MAGOH mutants in the EJC. The interactions with UPF3b and PYM are indicated by arrows. BTZ residues present in the crystal structure are displayed in red, an arbitrary shape of full length BTZ is suggested in light red. The EJC structure was rendered using PyMOL with structural data deposited in the Protein Data Bank (http://www.rcsb.org/pdb/home/home.do; ID: 2j0s). (B) Immunoprecipitations were done from RNAse A-treated lysates of HeLa cells that were transfected with FLAG-MAGOH and FLAG-MAGOH mutants or unfused FLAG as negative control together with V5-tagged BTZ, UPF3b, eIF4A3, PYM, and Y14. Co-precipitated proteins were detected by immunoblotting using an anti V5 antibody. (C) Northern blot analysis of RNA from HeLa cells that were transfected with expression plasmids for λNV5-tagged MAGOH or mutants of MAGOH together with the 4boxB reporter plasmid and the transfection control plasmid. Percentages (%) represent the mean of four independent experiments±standard deviations (SD). Bottom panel: expression levels of the tethered MAGOH mutant proteins were detected by immunoblot analysis with a V5-specific antibody. GFP served as internal loading control. (D) Splicing reactions using MINX as substrate RNA were supplemented with extracts expressing the indicated FLAG-tagged MAGOH mutants or unfused FLAG-tag as negative control. Reactions were immunoprecipitated with FLAG affinity gel. Twelve percent of the total input material was loaded in the input panel. (E) Splicing reactions and immunoprecipitations were performed as in (D) with the mutated MINX GG transcript.

Figure 3. Functional analysis of eIF4A3 in EJC assembly and NMD.

(A) Positions of the eIF4A3 mutants within the EJC as shown in Figure 2. Mutants in the ATP binding pocket on the inside of the protein and are not shown. (B) Immunoprecipitations were done from RNAse A-treated lysates of HeLa cells, which were transfected with FLAG-eIF4A3 and FLAG-eIF4A3 mutants or unfused FLAG as negative control together with V5-tagged BTZ, UPF3b, Y14, and MAGOH. Co-precipitated proteins were detected by immunoblotting using an anti V5 antibody. (C) Northern blot analysis of total cytoplasmic RNA from HeLa cells that were transfected with expression plasmids for MS2-tagged eIF4A3 or mutants of eIF4A3 together with the 6MS2 reporter plasmid and the transfection control plasmid. Percentages (%) represent the mean of 3 independent experiments±standard deviations (SD). Bottom panel: expression levels of the tethered eIF4A3 mutants were detected by immunoblot analysis with a MS2-specific antibody. GFP served as internal loading control. (D) Splicing reactions using MINX as substrate RNA were supplemented with extracts expressing the indicated FLAG-tagged eIF4A3 mutants or unfused FLAG-tag as negative control. Reactions were immunoprecipitated with FLAG affinity gel. Twelve percent of the total input material was loaded in the input panel. (E) Splicing reactions and immunoprecipitations were performed as in (C) with the mutated MINX GG transcript. (F) The amounts of MINX or MINX GG RNA immunoprecipitated by eIF4A3 mutants as shown in (C) and (D) were normalized to the amount precipitated by the eIF4A3 wild-type. Mutants that precipitated a lower relative amount of MINX compared to MINX GG were considered to destabilize the EJC during or after completion of splicing (red columns). Conversely, mutants that precipitated higher amounts of MINX compared to MINX GG were considered to increase the stability of the mature EJC (green columns).

Figure 4. Successive formation of a trimeric pre-EJC.

(A) Splicing reactions using MINX or MINX (15) as substrate RNAs were supplemented with FLAG-eIF4A3, FLAG-Y14, or FLAG-expressing extracts. Immunoprecipitations were done with FLAG affinity gel. Lanes with transcripts are labeled T. Twelve percent of the total input material was loaded in the input lanes. (B) Splicing reactions using MINX GG or MINX GG (15) as substrate RNAs were supplemented with FLAG-eIF4A3, FLAG-Y14, or FLAG-expressing extracts. Immunoprecipitations were done with FLAG affinity gel. Twelve percent of the total input material was loaded in the input lanes. (C) Splicing reactions using MINX, MINX GG, or MINX GG (15) as substrate RNAs were supplemented with FLAG-eIF4A3, FLAG-Y14, or FLAG-expressing extracts. Immunoprecipitations were done with FLAG affinity gel in the presence of 1% Empigen BB. Twelve percent of the total input material was loaded in the input lanes. (D) Schematic representation of the transcripts used in (A–C). Exons and introns are displayed as boxes and lines, respectively.

Figure 5. eIF4A3-binding deficient mutants of BTZ do not interact with the EJC, but elicit NMD when tethered to the RNA.

(A) Positions of the BTZ mutations in the EJC as shown in Figure 2. (B) Immunoprecipitations were done from RNAse A-treated lysates of HeLa cells, which were transfected with FLAG-BTZ and FLAG-BTZ mutants or unfused FLAG as negative control together with V5-tagged UPF3b, eIF4A3, Y14, and MAGOH. Co-precipitated proteins were detected by immunoblotting using an anti V5 antibody. (C) Splicing reactions using MINX as substrate RNA were supplemented with extracts expressing the indicated FLAG-tagged BTZ mutants or unfused FLAG-tag as negative control. Reactions were immunoprecipitated with FLAG affinity gel. Twelve percent of the total input material was loaded in the input panel. (D) Northern blot analysis of total cytoplasmic RNA from HeLa cells that were transfected with expression plasmids for λNV5-tagged BTZ or mutants of BTZ together with the 4boxB reporter plasmid and the transfection control plasmid. Percentages (%) represent the mean of three independent experiments±standard deviations (SD). Bottom panel: expression levels of the tethered BTZ mutants were detected by immunoblot analysis with a V5 antibody. GFP served as an internal loading control.

Figure 6. BTZ links the EJC with UPF1-dependent NMD.

(A) Schematic representation of the BTZ deletion mutants used in (B). The SELOR domain is displayed as a gray box, the size of the respective N- or C-terminal deletion is indicated. (B) Northern blot analysis of total cytoplasmic RNA from HeLa cells that were transfected with expression plasmids for λNV5-tagged BTZ or deletion mutants of BTZ together with the 4boxB reporter plasmid and the transfection control plasmid. The presence of the 218 mutation in each mutant is indicated. Percentages (%) represent the mean of four independent experiments±standard deviations (SD). Bottom panel: expression levels of the tethered BTZ mutants were detected by immunoblot analysis with an anti-V5 antibody. GFP served as internal loading control. (C) Northern blot analysis of total cytoplasmic RNA from HeLa cells that were transfected with siRNAs targeting Luciferase (negative control) or UPF1. Thirty hours later, the cells were transfected with the expression plasmid for λNV5-tagged BTZ 218 together with the 4boxB reporter plasmid and the transfection control plasmid. The siRNA depletion was rescued by transfecting a siRNA-insensitive expression plasmid for UPF1. Percentages (%) represent the mean of four independent experiments±standard deviations (SD). (D) Upper panel: immunoblot analysis of UPF1 expression in lysates from cells used in (C) with a UPF1-specific antibody. Tubulin served as control for comparable loading. Lower panel: the expression of tethered BTZ 218 was detected by immunoblot analysis with an anti-V5 antibody. GFP served as internal loading control.

Figure 7. BTZ is required for efficient NMD and interacts with UPF1.

(A) Immunoblot analysis for BTZ of protein lysates from HeLa cells transfected with Luciferase siRNA (negative control; lanes 1–4) or BTZ siRNAs (lanes 5–7). Dilutions corresponding to 50%, 20%, or 10% (lanes 2–4) of the initial protein amount (lane 1) of negative control, siRNA-transfected cells were loaded to assess the efficiency of the BTZ depletion. Reprobing with a tubulin-specific antibody was performed to control for loading. (B) Northern blot analysis of RNA from HeLa cells transfected with the indicated siRNAs and subsequently with the NMD reporter plasmids TCR-β wt or NS98 and a plasmid controlling for transfection efficiency. The numbers indicate changes in mRNA abundance±SD determined by analysis of four independent experiments. (C) Immunoblot analysis of protein lysates from HeLa cells transfected with siRNA targeting Luciferase (lane 1) or siRNA (2) targeting BTZ (lanes 2–3) as described in (A). To rescue the siRNA depletion, a siRNA-insensitive mutant of BTZ was co-transfected (lane 3). (D) Northern blot analysis of RNA from HeLa cells that were transfected with siRNA (lane 2). The NMD reporter plasmids TCR-β wt or NS98 and a transfection efficiency control were transfected together with a plasmid expressing a siRNA-insensitive variant of BTZ (lanes 5 and 6). The numbers indicate changes in mRNA abundance±SD determined by the analysis of four independent experiments. (E) Immunoprecipitations were done from RNAse A-treated lysates of HeLa cells that were transfected with FLAG-UPF1 and FLAG-UPF1 mutants together with V5-tagged BTZ. Co-precipitated proteins were detected by immunoblotting using UPF2, UPF3b, or anti V5 antibodies. (F) Immunoprecipitations were done from RNAse A-treated lysates of HeLa cells that were treated with siRNA and transfected with FLAG-BTZ. Co-precipitated proteins were detected by immunoblotting using an anti UPF1 antibodies. The depletion of UPF2 was assessed in the cell lysates with an UPF2 antibody. The membrane was reprobed with a tubulin antibody to control for loading.

Figure 8. Ordered assembly of the EJC by the spliceosome.

(A) Assembly pathway of the EJC. For details, see Discussion. (B) The exon junction complex recruits NMD activating proteins and combines different complexes leading to UPF1-dependent NMD. The model of the BTZ-dependent pathway is derived from the data presented in this work, whereas the model of the UPF3b-UPF2-UPF1-dependent pathway is derived from collective data of earlier publications ,,,. Details are explained in the Discussion.