2026/03/11 by Lu Xu, Mingsong Wu, Yanan Liu +2 · 1 voice
Agricultural and Biological Sciences · Biochemistry, Genetics and Molecular Biology · #Plant Gene Expression Analysis #Plant Molecular Biology Research #Plant-Microbe Interactions and Immunity
paper · doi:10.1111/jipb.70214
openalex publication_date 2026/03/11 · openalex created_date 2026/03/13 · openalex updated_date 2026/07/28
Unlike Arabidopsis, Nicotiana benthamiana uses the SENESCENCE ASSOCIATED GENE 101– N REQUIREMENT GENE 1 branch of TIR signaling to activate salicylic acid biosynthesis, illustrating evolutionary divergence in how conserved immune signaling modules are deployed to control salicylic acid production across plant lineages. Salicylic acid (SA) is a central defense hormone in plants. In the model plant Arabidopsis, SA is primarily synthesized via the isochorismate synthase (ICS) pathway (Peng et al., 2021). However, recent studies in Nicotiana benthamiana and rice revealed a more prevalent phenylalanine (Phe)-derived SA biosynthesis pathway in plant species outside the Brassicaceae, where SA is produced from benzoyl coenzyme A (benzoyl-CoA) in three steps (Liu et al., 2025; Wang et al., 2025; Zhu et al., 2025). First, the benzoyl-CoA:benzyl alcohol benzoyl transferase (BEBT) conjugates benzoyl-CoA with benzyl alcohol to form benzylbenzoate. The benzyl benzoate oxidase/benzyl benzoate hydroxylase (BBO/BBH) then catalyzes the oxidation of benzyl benzoate to produce benzyl salicylate. Finally, the benzyl salicylate hydrolase/benzoic salicylate esterase (BSH/BSE) cleaves benzyl salicylate to release SA. The identification of this newly discovered pathway leads to the next important question on how SA biosynthesis is regulated in these plants. From our previously described forward genetic screen searching for N. benthamiana mutants deficient in SA accumulation following infection with the avirulent pathogen Pseudomonas syringae pv. tomato DC3000 (Pst DC3000), we isolated an ethyl methane sulfonate (EMS)-induced mutant line #23-184 (Liu et al., 2025). The SA levels were restored in F1 progeny from two separate crosses with wild-type (WT) plants, suggesting that the causal mutation is recessive (Figure S1A). Whole-genome sequencing (WGS) of a pooled set of 39 F2 individuals displaying SA deficiency post Pst DC3000 infiltration (39 out of 160 F2 plants) revealed a linkage region on chromosome 7 based on single-nucleotide polymorphism (SNP) frequency analysis (Figure 1A). Within this interval, we identified a nonsynonymous mutation (cDNA: G406A; protein: D136N) in a previously identified Resistance (R) gene Roq1 (NbL07g16290.1), which encodes a TIR–NB–LRR (TNL) protein (Figure 1B). HopQ1 recognition by Roq1 triggers EDS1-SAG101b-NRG1-dependent SA accumulation in N. benthamiana (A) SNP frequency along chromosome 7 from bulked segregant analysis of SA-deficient F2 plants derived from mutant #23-184. The red dot marks the Roq1 mutation (SNP% = 100%). (B) Roq1 gene structure showing the EMS mutation and CRISPR-Cas9 gRNA target sites. Orange triangles indicate the protospacer adjacent motif (PAM) sites. (C) Sanger sequencing of two independent roq1 CRISPR mutants. roq1 #1 carries an inversion and roq1 #2 carries a 222-bp deletion between sgRNA sites B and C. (D–G) Free SA levels in 5-week-old WT, roq1, and EDS1-family mutants at 8 hpi after Pst DC3000 or its derivatives (OD600 = 0.02) (D, E, G), or at 8 and 24 hpi with flg22 and chitin (1 μM) (F). (H–L) Gene expression levels of SA biosynthetic genes BEBT (H), BBO1 (I), BBO2 (J), BSH1 (K), and BSH2 (L) in the respective genotypes at 6 hpi with Pst DC3000 (OD600 = 0.02), as measured by qRT-PCR analysis. (M) Domain architecture of Arabidopsis TIR-containing proteins analyzed. (N–Q) Free SA levels in WT, roq1, and EDS1-family mutants at 24 h post transient expression of Arabidopsis TIR proteins AtSNC1 (N), AT4G11170 (O), AT3G04220 (P), and AT2G32140 (Q) (OD600 = 0.4). (R) Model of TIR signaling pathways controlling SA biosynthesis in N. benthamiana versus Arabidopsis. In N. benthamiana, PTI triggered by flg22, chitin, or Pst DC3000-D36E is insufficient to activate the Phe-derived SA biosynthesis pathway. Instead, SA production requires ETI via recognition of HopQ1 by endogenous Roq1 during Pst DC3000 infection, with downstream signaling channeled through the EDS1–SAG101–NRG1 branch. This reliance on EDS1–SAG101–NRG1 extends beyond Roq1 to several heterologously expressed Arabidopsis TIR receptors. (D–G, N–Q) Free SA levels were measured by HPLC. Error bars represent means ± SD (n = 4 independent plants). One-way ANOVA with Tukey's test (D, F–L, N–Q) or two-tailed Student's t-tests (E) were used. Different letters or asterisks indicate significant differences (P < 0.05). ns, not significant. To validate the role of Roq1 in inducing SA biosynthesis, we generated two independent knockout lines by CRISPR-Cas9 (Figure 1B, C). Sanger sequencing revealed that roq1 #1 harbors an inversion between sgRNA target sites B and C, whereas roq1 #2 carries a 222-bp deletion between the same sites (Figure 1C). While these lines showed no visible growth or morphological defects compared to WT (Figure S1B), both lines abolished Pst DC3000-induced SA accumulation (Figure 1D). To further investigate how Roq1 regulates SA biosynthesis, we analyzed the expression of recently identified SA biosynthetic genes, including BEBT, BBO1, BBO2, BSH1, and BSH2 (Liu et al., 2025), at 6 h post-infection (hpi) with Pst DC3000. Consistently, induction of all five genes was dramatically reduced in the roq1 mutant compared to the WT (Figure S2). These findings demonstrate that Roq1 is essential for activating the transcription of key SA biosynthetic genes and for promoting SA accumulation in response to Pst DC3000 infection in N. benthamiana. Pst DC3000 delivers a diverse repertoire of type III-secreted effectors into host cells to manipulate cellular processes and suppress immunity. Among them, HopQ1 is specifically recognized by Roq1 to trigger ETI (Schultink et al., 2017). To test whether HopQ1 recognition by Roq1 alone is sufficient to induce SA biosynthesis, we inoculated WT and the roq1 mutant with Pst DC3000 strains with different effector compositions. Neither DC3000-D36E, a strain lacking all 36 characterized type III effectors, nor DC3000-∆hopQ1-1, which only lacks HopQ1, induced detectable SA induction (Figure 1E). These results demonstrate that HopQ1 recognition by Roq1 alone robustly activates SA biosynthesis, suggesting that SA accumulation in response to Pst DC3000 in N. benthamiana is specifically induced through the HopQ1-triggered activation of Roq1. Given the lack of SA induction by DC3000-D36E in N. benthamiana, we further examined whether treatment with the elicitor flg22, a 22-amino acid peptide derived from bacterial flagellin, can induce SA biosynthesis. As shown in Figure 1F, flg22 did not induce detectable SA accumulation in N. benthamiana at either 8 or 24 h post treatment. Consistent with this observation, treatment with another elicitor, chitin, likewise failed to induce SA accumulation (Figure 1F). The NADase and ADP-ribosylation activities of the TIR domains in plant immune receptors produce danger signals such as pRib-AMP/ADP and ADPr-ATP/di-ADPR, which bind to EDS1 (ENHANCED DISEASE SUSCEPTIBILITY 1)-PAD4 (PHYTOALEXIN DEFICIENT 4) and EDS1-SAG101 (SENESCENCE ASSOCIATED GENE 101) heterodimers, respectively (Huang et al., 2022). Ligand binding induces the association of EDS1-PAD4 with the helper NLR ADR1 (ACTIVATED DISEASE RESISTANCE 1) and the association of EDS1-SAG101 with the helper NLR NRG1 (N REQUIREMENT GENE 1), thereby activating distinct branches of downstream immune signaling (Lapin et al., 2019; Huang et al., 2022). In N. benthamiana, previous studies have shown that the EDS1–SAG101–NRG1 module is essential for TNL immunity, including Roq1-dependent cell death, restriction of bacterial growth, and defense gene activation, with PAD4 and ADR1 generally dispensable for these outputs (Qi et al., 2018; Gantner et al., 2019; Lapin et al., 2019). In contrast, recent work has demonstrated that the EDS1–PAD4–ADR1 module contributes to stomatal immunity (Wang et al., 2024). However, the role of these two branches of TIR signaling in Phe-derived SA biosynthesis was unclear. To determine the contribution of EDS1–PAD4–ADR1 and EDS1–SAG101–NRG1 signaling modules to Roq1-induced SA accumulation in N. benthamiana, we tested CRISPR mutants targeting components of both branches, including EDS1a, PAD4, SAG101b, ADR1, and NRG1. EDS1b and SAG101a were not analyzed because they were reported to be non-functional paralogs in N. benthamiana (Gantner et al., 2019). SA levels were quantified at 8 hpi with Pst DC3000. As shown in Figure 1G, while pad4 and adr1 single mutants retained WT levels of SA, the induction of SA accumulation was almost completely blocked in eds1a, sag101b, and nrg1 single mutants. When the transcript levels of SA biosynthetic genes were analyzed across the mutant lines during Pst DC3000 infection, the expression levels of BEBT, BBO1, BBO2, BSH1, and BSH2 were unchanged in pad4 and adr1 lines but dramatically reduced in eds1a, sag101b, and nrg1 mutants compared to the WT (Figure 1H–L). These findings suggest that SA biosynthesis in N. benthamiana is turned on by the EDS1–SAG101–NRG1 branch rather than the PAD4–ADR1 branch of TIR signaling. To determine whether the EDS1–SAG101–NRG1 module operates broadly in TIR-triggered SA biosynthesis beyond Roq1, we examined additional TIR proteins that were known to activate defense responses when overexpressed in N. benthamiana (Figure 1M). In Arabidopsis, SNC1 activates immune responses largely via the EDS1–PAD4–ADR1 module (Dong et al., 2016). However, when expressed in N. benthamiana, SNC1-mediated SA accumulation occurred via the Phe-derived SA biosynthesis pathway (Figure S3A) and required EDS1, SAG101, and NRG1, whereas PAD4 and ADR1 were dispensable (Figures 1M, N, S3B). For the TNL genes AT4G11170 and AT3G04220 (Figure 1M), SA induction remained intact in pad4, adr1, and roq1 single mutants, but was abolished in eds1a, sag101b, and nrg1 single mutants (Figure 1O, P). A similar pattern was observed for the TIR-only gene AT2G32140 (Figure 1Q). Further analysis of the expression levels of FLAG-tagged SNC1 and HA-tagged AT3G04220 and AT2G32140 by western blot showed that they accumulated to comparable levels across different genotypes (Figure S4A–C). As AT4G11170 was not tagged, we assessed its transcript levels by RT-qPCR and found that its expression levels were similar in WT and the mutants (Figure S4D). Together, these findings reveal that the EDS1–SAG101–NRG1 module acts as a key upstream SA biosynthesis inducer downstream of different TIR proteins. Previous studies have shown that CC–NB–LRR proteins (CNLs) can induce hypersensitive cell death when overexpressed in N. benthamiana (Zhang et al., 2017; Yu et al., 2025). To test whether CNL activation is also sufficient to promote SA accumulation, we transiently expressed representative CNLs in WT N. benthamiana, including AtRPS2 and the autoactive variants NbZAR1D481V and NbNRC4D478V. Under our conditions, expression of NbZAR1D481V or NbNRC4D478V did not result in detectable SA induction, whereas AtRPS2 elicited an increase in SA levels (Figure S5). This is consistent with previous reports that stable overexpression of the typical CNL NtRPP13 is associated with elevated SA accumulation (Yu et al., 2025) and that the NBS-LRR AhRRS5 correlates with increased expression of biosynthetic genes in the Phe-derived SA pathway (Zhang et al., 2017). Together, these findings suggest that CNL activation can, at least in some cases, contribute to SA accumulation, but that the magnitude of this effect may be weaker than the robust SA biosynthesis mediated by TIR signaling. In summary, our forward genetic analysis of SA-deficient mutants in N. benthamiana revealed that activation of TIR signaling through HopQ1 recognition by Roq1 is required for induction of SA biosynthesis upon Pst DC3000 infection. Strikingly, PTI activation alone by flg22 or Pst DC3000-D36E was unable to induce SA production in N. benthamiana. This contrasts with Arabidopsis, where SA biosynthesis is activated by PTI and contributes to enhanced disease resistance (Tsuda et al., 2008). The decoupling of SA production from PRR-triggered signaling in N. benthamiana suggests that the Phe-derived SA biosynthesis pathway may be more specifically regulated than the ICS pathway. Our findings refine the current framework of TNL–EDS1-mediated immunity by demonstrating that Roq1-induced SA accumulation in N. benthamiana depends predominantly on the EDS1–SAG101–NRG1 module. A recent study reported reduced SA accumulation in the Nbadr1 mutant following spray inoculation with Pst DC3000 (Wang et al., 2024), but the difference in SA levels between WT and the mutant appeared to be quite small, suggesting that ADR1 may contribute quantitatively in certain contexts such as stomatal immunity but is not a primary driver of pathogen-induced SA accumulation. A similar pattern was observed for SA induction in N. benthamiana by several other TIRs, including SNC1, which primarily activates immune responses through the EDS1–PAD4–ADR1 module in Arabidopsis. This is in clear contrast to the ICS pathway in Arabidopsis that is mostly activated by EDS1–PAD4–ADR1 signaling (Cui et al., 2017), highlighting evolutionary divergence in the usage of conserved immune signaling modules to control SA production across plant lineages. With the recent elucidation of the Phe-derived SA biosynthesis pathway used beyond Brassicaceae, our study provides a timely framework on the regulation of SA production via this pathway. A striking divergence in the regulatory mechanisms of SA biosynthesis between N. benthamiana and Arabidopsis is revealed (Figure 1R). Whether TIR signaling is similarly used to regulate SA biosynthesis in other plant lineages, such as monocots, awaits further investigation. We would like to thank Dr. Johannes Stuttmann (University of Helle) and Dr. Tiancong Qi (Tsinghua University) for EDS1-family mutants, Dr. Zhonglin Mou (University of Florida) for the Acinetobacter sp. ADPWH-lux strain, Dr. Daniel Voytas (University of Minnesota) for the N. benthamiana Cas9 transgenic line, Dr. Hainan Tian (Sichuan University) for Arabidopsis TIR constructs, and Sophien Kamoun (The Sainsbury Laboratory) for NbZAR1D481V and NbNRC4D478V. This work was supported by funds from the Canadian Natural Sciences and Engineering Research Council (NSERC) Discovery program (X.L.; Y.Z.) and NSERC-CREATE-PRoTECT (X.L.). L.X. is partially supported by a scholarship from the China Scholarship Council (CSC). The authors declare no conflicts of interest. L.X., X.L., and Y.Z. conceived the study. M.W. contributed to WGS data analysis. L.X. established the EMS screen and performed all other experiments. L.X. drafted the manuscript. Y.L., X.L., and Y.Z. revised the manuscript. All authors have read and approved the final manuscript. Additional Supporting Information may be found online in the supporting information tab for this article: http://onlinelibrary.wiley.com/doi/10.1111/jipb.70214/suppinfo Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.