Next-generation mapping of the salicylic acid signaling hub and transcriptional cascade

Abstract

For over 60 years, salicylic acid (SA) has been known as a plant immune signal required for basal and systemic acquired resistance. SA activates these immune responses by reprogramming ∼20% of the transcriptome through NPR1. However, components in the NPR1 signaling hub, which appears as nuclear condensates, and the NPR1 signaling cascade have remained elusive due to difficulties in studying this transcriptional cofactor, whose chromatin association is indirect and likely transient. To overcome this challenge, we applied TurboID to divulge the NPR1 proxiome, which detected almost all known NPR1 interactors as well as new components of transcription-related complexes. Testing of new components showed that chromatin remodeling and histone demethylation contribute to SA-induced resistance. Globally, the NPR1 proxiome has a striking similarity to the proxiome of GBPL3 that is involved in SA synthesis, except for associated transcription factors (TFs), suggesting that common regulatory modules are recruited to reprogram specific transcriptomes by transcriptional cofactors, like NPR1, through binding to unique TFs. Stepwise green fluorescent protein-tagged factor cleavage under target and release using nuclease (greenCUT&RUN) analyses showed that, upon SA induction, NPR1 initiates the transcriptional cascade primarily through association with TGACG-binding TFs to induce expression of secondary TFs, predominantly WRKYs. Further, WRKY54 and WRKY70 were identified to play a major role in inducing immune-output genes without interacting with NPR1 at the chromatin. Moreover, loss of condensate formation function of NPR1 decreases its chromatin association and transcriptional activity, indicating the importance of condensates in organizing the NPR1 signaling hub and initiating the transcriptional cascade. Collectively, this study demonstrates how combinatorial applications of TurboID and stepwise greenCUT&RUN transcend traditional genetic methods to globally map signaling hubs and transcriptional cascades for in-depth explorations.

Keywords: NPR1; TGA; TurboID; WRKY TFs; greenCUT&RUN; salicylic acid-induced transcription.

Figure 1.. The NPR1 proxiome contains transcriptional machineries and chromatin remodelers shared by the GBPL3-proxiome.

(A) Volcano plot of NPR1 proximal proteins 4 h after SA treatment, detected through TurboID biotin affinity purification followed by label-free quantification (LFQ) MS processed under mild conditions (Methods). Red points represent proteins that have an NPR1_LFQ/YFP_LFQ ≥ 2 and p < 0.1 in both mild and harsh washing conditions (Methods) or p < 0.01 in at least one washing condition. The single blue point (on the right) represents NPR1. (B) Log₂(maximum LFQ intensity) of NPR1–3 × HA-TbID, YFP-YFP-TbID (BAIT), and known NPR1 interactors in NPR1–3 × HA-TbID (NPR1) vs. YFP-YFP-TbID (YFP) samples. (C) Venn diagram comparing NPR1 proximal proteins identified in the current TurboID experiment with those identified in the cytoplasmic SA-induced NPR1 condensates (cSINCs) (Zavaliev et al., 2020). (D) Enriched molecular functions (MFs) of the 234 NPR1 proximal proteins. (E) Proximity between NPR1 and LDL3 or BRM. N-terminal YFP-fused NPR1 (NPR1-YN) and the C-terminal YFP-fused BRM C terminus (amino acids 953–2193) (BRM.C-YC) or cYFP-fused LDL3 (LDL3-YC) were co-expressed with the nuclear marker protein histone 2B fused to mCherry (HTB1-mCherry) in N. benthamiana. Plants were imaged after treatment with water (mock) or 1 mM SA for 8 h, and the BiFC intensities were measured from multiple nuclei; values were plotted on a box-and-whisker plot. Different letters indicate statistical significance based on an ordinary one-way ANOVA with Tukey’s multiple-comparisons tests (a single pooled variance). Asterisks indicate statistical significance tested by two-tailed unpaired Student’s t-test (****p < 0.0001). Scale bar, 10 μm. (F and G) WT, npr1–2, brm-3, BRM:BRM-GFP/brm-1 (F), ldl3–1, ldl3–2, and LDL3-FLAG/ldl3–1 (G) treated with H₂O (mock) or 1 mM SA for 24 h prior to inoculation with Psm ES4326 at OD₆₀₀ = 0.001. Bacterial colony-forming units (CFUs) were measured 3 days post inoculation (n = 8; error bars represent SEM; two-sided t-test and two-way ANOVA were used for comparisons within and between genotypes, respectively). (H) Search Tool for the Retrival of Interacting Genes/Proteins (STRING) network analysis (Szklarczyk et al., 2019) of NPR1 proximal proteins relating to chromatin remodeling and transcriptional regulation. Blue shade, proteins shared with the GBPL3 proxiome (Tang et al., 2022).

Figure 2.. NPR1 targets TF gene promoters through association with TGA TFs.

(A) Pearson’s correlation of the greenCUT&RUN data from plants expressing NPR1-GFP (NPR1) and GFP with and without 1 mM SA treatment for 4 h. (B) Integrative Genomics Viewer (IGV) view of the PR1 promoter, showing normalized NPR1-GFP and GFP binding before and after SA treatment. (C) Mean profile of reads per genomic content (RPGCs) of NPR1-GFP reads before and after SA treatment at NPR1 target genes. TSS, transcription start site. (D) Motifs enriched under NPR1-GFP peaks 4 h after 1 mM SA treatment. (E) Cut frequency of all as-1 elements (TGACG) by the GFP nanobody MNase in the overall NPR1-GFP peaks 4 h after 1 mM SA treatment. (F) The enriched biological processes (BPs) and MFs of NPR1 target genes. (G) Venn diagram displaying shared NPR1 loci 1, 2, 4, 8, and 24 h after 1 mM SA treatment. (H) Heatmaps and mean profile of normalized (RPGC) NPR1 binding at shared loci (outlined in blue) and unique loci (outlined in green) 1, 2, 4, 8, and 24 h after 1 mM SA treatment. (I) Mean profile of RPGC of NPR1-GFP and npr1^sim3-GFP (sim3) before and after SA treatment at NPR1 target genes.

Figure 3.. WRKY54/70 are major TFs downstream of NPR1-TGA that positively regulate SA-mediated gene expression and resistance.

(A) Mean profile of RPGCs of WRKY70-GFP (WRKY70) and GFP reads of WRKY70 target genes. (B) Motifs enriched under WRKY70-GFP peaks. (C) Enriched BPs and MFs of WRKY70 target genes. (D) Venn diagram illustrating the overlap between NPR1 and WRKY70 target genes. (E and F) IGV view of normalized NPR1 and WRKY70 binding at the promoters of their shared target genes WRKY63 (E) and PCR1 (F). (G) RPGCs of NPR1-GFP and WRKY70-GFP at 116 shared target genes 1 kb upstream and downstream of NPR1 peaks. (H) RPGCs of all NPR1-GFP and WRKY70-GFP target genes centered on their respective peaks. (I) Correlation between SA-induced transcription and NPR1 dependency in WRKY70 target genes. r, Pearson correlation coefficient. (J) Correlation between SA-induced transcription and WRKY54/70 dependency in NPR1-target genes. (K) Bacterial CFUs in WT, wrky54/70, and npr1–2. Plants were treated with H₂O (mock) or 1 mM SA for 24 h before being inoculated with Psm ES4326 at OD₆₀₀ = 0.001. CFUs were measured 2 days post inoculation (n = 8; error bars represent SEM; two-sided t-test and two-way ANOVA were used for comparison within and between genotypes, respectively).

Figure 4.. Biomolecular condensate formation stabilizes NPR1 association with the TGA TF and enhances its transcriptional activity.

(A) Mean profile of RPGC of NPR1-GFP (NPR1) and npr1^rdr3-GFP (rdr3) at NPR1 target genes before and after 4 h of 1 mM SA treatment. (B and C) CoIP between TGA3 (B) or TGA5 (C) and NPR1 or rdr3 transiently overexpressed in N. benthamiana. The value under the IP blot represents band intensities normalized to TGA TF input. (D) Quantification of normalized coIP TGA3-HA or TGA5-HA band intensity from three independent replicates (n = 3, error bars represent SEM; two-sided t-test was used for comparison between NPR1 and rdr3). (E) Interaction between TGA3 or TGA5 fused to the activator domain (AD) and NPR1 or rdr3 fused to the DNA-binding domain (BD) in the yeast two-hybrid assay. Yeast strains were mated for 24 h, normalized to OD₆₀₀ = 1.0, serially diluted, and plated on the indicated synthetic defined (SD) medium without leucine and tryptophan (LW) or without leucine, tryptophan, histidine, and adenine (LWHA) and incubated at 30°C. Photos were taken 2 days after plating. (F) Protein level of NPR1-GFP and rdr3-GFP in stable transgenic Arabidopsis plants. (G–J) Transcript levels of NPR1 or rdr3 (G) and the target genes PR1 (H), WRKY18 (I), and WRKY70 (J) in 35S:NPR1-GFP/npr1–2, 35S:npr1^rdr3-GFP/npr1–2, and npr1–2 plants measured using qPCR 8 h after SA induction (n = 3, error bars represent standard deviation). (K–N) Mean profile of RPGCs of BRM-GFP reads at BRM target genes (K), NPR1 target genes (L), WRKY70 target genes (M), and SA-induced genes (N) 4 h after treatment with H₂O or 1 mM SA. (O) Working model of the SA/NPR1 signaling hub and transcriptional cascade. Overlapped rectangular shades show that NPR1 and GBPL3 condensates share general transcriptional regulatory machineries (e.g., Mediator, SWI/SNF, and histone modifiers) but target different genes through association with unique TFs. An increase in SA level triggers the transcriptional cascade by first activating NPR1 to induce TGA-mediated expression of WRKY, MYB, NAC, and ERFTFs, which, in turn, activate the subsequent gene expression.