PABP/purine-rich motif as an initiation module for cap-independent translation in pattern-triggered immunity

Abstract

Upon stress, eukaryotes typically reprogram their translatome through GCN2-mediated phosphorylation of the eukaryotic translation initiation factor, eIF2α, to inhibit general translation initiation while selectively translating essential stress regulators. Unexpectedly, in plants, pattern-triggered immunity (PTI) and response to other environmental stresses occur independently of the GCN2/eIF2α pathway. Here, we show that while PTI induces mRNA decapping to inhibit general translation, defense mRNAs with a purine-rich element ("R-motif") are selectively translated using R-motif as an internal ribosome entry site (IRES). R-motif-dependent translation is executed by poly(A)-binding proteins (PABPs) through preferential association with the PTI-activating eIFiso4G over the repressive eIF4G. Phosphorylation by PTI regulators mitogen-activated protein kinase 3 and 6 (MPK3/6) inhibits eIF4G's activity while enhancing PABP binding to the R-motif and promoting eIFiso4G-mediated defense mRNA translation, establishing a link between PTI signaling and protein synthesis. Given its prevalence in both plants and animals, the PABP/R-motif translation initiation module may have a broader role in reprogramming the stress translatome.

Keywords: IRES; MPK3/MPK6; PABP; R-motif; RACK1; cap-independent translation; eIF4G; eIFiso4G; pattern-triggered immunity; translational reprogramming.

Figure 1.. Elf18-induced mRNA decapping by the DCP complex positively affects PTI

(A) Independent transgenic plants expressing 35S:DCP2-HA (WT)(3, 13) and 35S:DCP2^E158Q-HA (E158Q)(23, 28) in dcp2-1 (top panel) with similar expression levels (middle panel) and Rubisco as a loading control (bottom panel). Representative seedlings were photographed after 10 days on ½ MS. (B) Confocal images of P-body colocalization of DCP2-mCherry and DCP2^E158Q-YFP transiently expressed in Nb plants. Histone H2B (HTB)-mCherry, negative control. Colocalization coefficient was based on 9 images (bottom graph). Data are presented as box-and-whisker plot. Scale bars, 10 μm. (C and D) Elf18-induced decapping of UBQ10 (C) and TBF1 (D) mRNAs was measured using eight-day-old seedlings after mock or 10 μM elf18 for 1 h and exposed to exonuclease before qPCR. Values are means ± SDs after normalizing to the unexposed control. (E) Elf18-induced growth inhibition. Five-day-old seedlings were treated with or without 50 μM estradiol overnight, followed by mock or 100 nM elf18 for 3–4 days before fresh weight was measured. amiR-DCP2–3, −4, −10, and DCP2OE-4, −10, independent transgenic lines expressing estradiol-inducible artificial microRNA against DCP2 and overexpressing DCP2, respectively. Values are means ± SDs. Each dot represents a biological replicate. Data were analyzed via two-way ANOVA (C and D) and two-way ANOVA with Dunnett multiple comparisons (E), *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. ns, not significant.

Figure 2.. PABPs are required for R-motif-mediated cap-independent protein translation

(A) In planta elf18-induced translation of TBF1 dual luciferase reporters with WT or mutated R-motifs (mRs). 3–4-week-old independent Arabidopsis transgenic lines carrying the reporters were sprayed with 2 mM luciferin overnight before infiltrating with mock (water) or 10 μM elf18 for 1 h. Luciferase activity was measured using a CCD camera with 20 min exposure time. Values are means ± SDs. (B) In vitro translation of uncapped TBF1 reporter mRNAs with WT or mutated R-motifs. Top, a schematic of mRNA carrying the TBF1 5′ leader sequence [3 R-motifs and 2 uORFs (arrows)] and the N-terminal 12 amino acid coding sequence of TBF1 fused with the firefly luciferase gene (FLUC). Bottom, translation activities (FLUC) of the mRNAs measured using the wheat-germ translation system. Values are means ± SDs. (C) In planta translation of TBF1 dual luciferase reporters with mutated uORFs (uorf) and with or without a 5′ translation inhibitory element (TIE). The reporters were transiently expressed in Nb plants for 20 h before the luciferase activities were measured. Values are means ± SDs. (D) In planta translation of TBF1 bicistronic reporters. The 35S:RLUC-uorf/R_TBF1-FLUC (Bicistronic-uorf) or the 35S:RLUC-uorf/mR123_TBF1-FLUC mutant (Bicistronic-uorf/mR123) reporter was transiently coexpressed with the elf18 receptor EFR-GFP in Nb plants for 20 h and infiltrated with 10 μM elf18 or water for 2 h before the luciferase activities were measured. Values are means ± SDs. (E) Pulldown of PABPs by biotinylated TBF1 R-motifs or polyA sequence. PAB8-FLAG protein was purified from protoplasts treated with 1 μM elf18 for the indicated time and the protein was quantified by immunoblotting using an anti-FLAG antibody. Numbers under each blot are the ratios to time 0 after normalized to the input (bottom blot). (F) The effect of PAB8-HA on the translational activities of TBF1 dual luciferase reporters. Reporter activities were determined 2 days after transient coexpression in Nb plants. Values are means ± SDs. (G) The effect of PABP mutation on the TBF1 dual luciferase reporter translation. The reporters were expressed overnight in protoplasts made from WT and the pab2 pab8 pab4^+/− mutant (pab2.5) and treated with 1 μM elf18 for 45 min before luciferase activities were recorded. Values are means ± SDs. Each dot represents a biological replicate. Data were analyzed via two-way ANOVA with Dunnett multiple comparisons (A), one-way ANOVA (Tukey) (B), t-test (left panel C) and two-way ANOVA (right panel C, D, F and G), *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Different letters indicate statistical significance, P < 0.05. ns, not significant.

Figure 3.. PABPs regulate R-motif-containing defense mRNA translation through differential association with eIF4G and eIFiso4G

(A) STRING analysis of PABP-interacting proteins identified using LC-MS/MS (related to Table S3). Text-mining, coexpression and experiments were chosen as active interaction sources, interaction score = 0.700. (B) The dynamics of the interaction between PAB8 and eIF4G (4G) or eIFiso4G1 (I4G1) upon elf18 treatment were determined using the split luciferase assay, in which Cluc-tagged PAB8 and Nluc-tagged eIF4G (4G) or eIFiso4G1 (I4G1) were transiently coexpressed in Nb plants for 2 days. Data are values (elf18/mock) normalized to time zero. (C and D) Elf18-induced translation of R-motif-containing mRNAs, TBF1, ZIK3 and ATG8E, was measured using TBF1, ZIK3 and ATG8E 5’ leader sequences in dual luciferase reporters after overnight expression in protoplasts made from WT, eif4g eif4e1 (4g/e1) (C) and eifiso4g1 eifiso4g2 (i4g1/2) (D) plants. The resulting protoplasts were treated with 1 μM elf18 for 45 min before luciferase activities were measured. Values are means ± SDs. Each dot represents a biological replicate. Data were analyzed via two-way ANOVA with Dunnett multiple comparisons (B) and two-way ANOVA (C and D), *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. ns, not significant.

Figure 4.. PABP and RACK1 regulate basal resistance and PTI through eIF4G and eIFiso4G

(A-C) Elf18-induced resistance. 4-week-old Arabidopsis plants were infiltrated with Psm ES4326 (OD_600nm = 0.001) 24 h after 1 μM elf18 or water treatment and bacterial growth was assessed 2 days later in the mutants of PABP (A), eIF4G and 4E1 (B), eIFiso4G1 and eIFiso4G2 (C). pab2.5, pab2 pab8 pab4^+/−; 4g, eif4g; 4e1, eif4e1; 4g/e1, eif4g eif4e1; i4g1, eifiso4g1; i4g2, eifiso4g2; i4g1/2, eifiso4g1 eifiso4g2; efr, the elf18 receptor mutant. (D) Basal resistance. Bacterial infection was performed as in (A-C) without the elf18 pre-treatment. pab2/8, pab2 pab8; pab2/8/4g/e1, pab2 pab8 eif4g eif4e1. (E and F) Elf18-induced resistance. Bacterial infection was performed as in (A-C) in the pab2/4/i4g1/2 mutant (E) and the estradiol-inducible (50 μM estradiol) RACK1 knockdown mutant lines rack1-es1 and rack1-es2 (F). pab2/4/i4g1/2, pab2 pab4 eifiso4g1 eifiso4g2. Photos of plants were taken before infection (A, D and E). Values are means of log colony-forming units per leaf area (CFU/cm²) ± SDs. Each dot represents a biological replicate. Data were analyzed by t-test (D) and two-way ANOVA (A-C, E and F), *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. ns, not significant.

Figure 5.. MPK3/6 phosphorylate PABP during PTI to enhance its binding to R-motif.

(A) Elf18-induced PAB8 protein mobility shift. PAB8-FLAG was extracted from protoplasts after treatment with 1 μM elf18 for the indicated time. Protein mobility was analyzed using a phos-tag gel immunoblotted with an anti-FLAG antibody. (B) Ser566 as the phospho-site in PAB8. PAB8-HA or the PAB8^S566A-HA mutant protein was produced and analyzed as in (A) with an anti-HA antibody. (C) Elf18-induced phosphorylation of PAB8 by MPK3/6. PAB8-HA was expressed in WT, mpk6SR (inhibitor-sensitized MPK3 variant-rescued mpk3 mpk6 double mutant) and rack1-es2 protoplasts in the presence of 2 μM NA-PP1 inhibitor (mpk6SR) and 50 μM estradiol (rack1-es2), respectively, and analyzed as in (A) with an anti-HA antibody. (D) Elf18-associated growth inhibition. Five-day-old seedlings were pre-treated with or without 2 μM NA-PP1 overnight, followed by water or 100 nM elf18 treatment for 3–4 days before fresh weight was measured. pab2/8, pab2 pab8. Values are means ± SEMs. (E) In vitro PAB8^S566 phosphorylation by MPK6. Protein phosphorylation was detected by antip-S566 antibodies. CBB: Coomassie Brilliant Blue G-250. (F) The interaction between R-motif and PAB8 phosphorylation variants. FLAG-tagged PAB8 and phospho-site variants purified from protoplasts and pulled down by TBF1 R-motif 3 were quantified by immunoblotting using an anti-FLAG antibody. Numbers under each blot are the ratios to PAB8-FLAG normalized to the input (bottom blot). (G) The effect of PAB8^S566 phosphorylation on translation. The TBF1 dual luciferase reporter translational activity was determined 2 days after transient coexpression with PAB8-HA (WT), PAB8^S566A-HA (566A) or PAB8^S556D-HA (566D) in Nb plants. Values are means ± SDs. (H) Elf18-associated growth inhibition. Five-day-old independent Arabidopsis transgenic lines (in pab2 pab8) expressing NP:PAB8-HA (WT), NP:PAB8^S556A-HA (556A) or NP:PAB8^S556D-HA (556D) were treated with mock or 100 nM elf18 for 3–4 days before fresh weight was measured. NP, PAB8 native promoter. Values are means ± SDs. Each dot represents a biological replicate. Data were analyzed by one-way ANOVA (Tukey) (G) and two-way ANOVA with Dunnett multiple comparisons (D and H), different letters indicate statistical significance, P < 0.05.

Figure 6.. MPK3/6-mediated phosphorylation inhibits eIF4G while activating eIFiso4G to reprogram translation during PTI.

(A) Elf18-induced eIF4G protein mobility shift. eIF4G-HA (4G-HA) was extracted from protoplasts after treatment with 1 μM elf18 for the indicated time. Mobility change of the protein was examined using a phos-tag gel immunoblotted with an anti-HA antibody. (B) In vitro eIF4G^S1066/T1069 phosphorylation by MPK6. Protein phosphorylation was detected by anti-pS1066/T1069 antibodies. GST-eIF4G, GST-4G; GST-4G^8A, alanine mutant of the eight MPK3/6 target sites. CBB: Coomassie Brilliant Blue G-250. (C) The effect of eIF4G phosphorylation on translation. The TBF1 dual luciferase reporter translational activity was determined 2 days after transient coexpression with eIF4G-HA (WT), eIF4G^8A (8A) or eIF4G^8D (8D) in Nb plants. 8D, aspartic acid mutant of the eight MPK3/6 target sites. Values are means ± SDs. (D) Elf18-induced eIFiso4G1 protein mobility shift. eIFiso4G1-FLAG (I4G1-FLAG) was extracted from protoplasts after treatment with 1 μM elf18 for the indicated time. Mobility change of the protein was examined using a phos-tag gel immunoblotted with an anti-FLAG antibody. (E) In vitro eIFiso4G1^S487/S542 phosphorylation by MPK6. Protein phosphorylation was detected by anti-pSer487 antibodies (left) and anti-pSer542 antibodies (right). GST-eIFiso4G1, GST-I4G1; GST-eIFiso4G1^S487A/S542A, GST-I4G1^2A. CBB: Coomassie Brilliant Blue G-250. (F) The effect of eIFiso4G phosphorylation on translation. The TBF1 dual luciferase reporter translational activity was determined 2 days after transient coexpression with eIFiso4G-HA (WT), eIFiso4G1^S487A/S542A (2A) or eIF4isoG1^S487D/S542D (2D) in Nb plants. Values are means ± SDs. (G) Elf18-induced resistance. Four-week-old independent Arabidopsis transgenic lines in eifiso4g1 eifiso4g2 (i4g1/2) expressing NP:eIFiso4G1-HA (WT), NP:eIFiso4G1^S487A/S542A-HA (2A) or NP:eIFiso4G1^S487D/S542D-HA (2D) were infiltrated with Psm ES4326 (OD_600nm = 0.001) 24 h after 1 μM elf18 or water treatment and bacterial growth was assessed 2 days later. NP, eIFiso4G1 native promoter. Values are means ± SDs (H) PABP/R-motif-mediated cap-independent translation reprogramming upon PTI induction. In the absence of pathogen challenge, growth mRNAs are translated through the canonical cap-dependent mechanism whereas for defense mRNAs with an R-motif (R), translation is inhibited by PABP through an unknown mechanism (?). Upon perception of pathogen challenge by PRRs, MPK3/6 are activated to induce mRNA decapping and repress eIF4G activity by phosphorylation to inhibit translation of growth mRNAs. For defense mRNAs carrying an R-motif, translation is initiated using R-motif as an IRES through PABP-mediated recruitment of eIFiso4G after phosphorylation by MPK3/6 on the scaffold protein RACK1. Each dot represents a biological replicate. Data were analyzed by one-way ANOVA (Tukey) (C and F) and two-way ANOVA (G), different letters indicate statistical significance, P < 0.05; **P < 0.01, ***P < 0.001, ****P < 0.0001. ns, not significant.