A dominant‐interfering camta3 mutation compromises primary transcriptional outputs mediated by both cell surface and intracellular immune receptors in Arabidopsis thaliana

Summary Pattern recognition receptors (PRRs) and nucleotide‐binding domain and leucine‐rich repeat (LRR)‐containing proteins (NLRs) initiate pattern‐triggered immunity (PTI) and effector‐triggered immunity (ETI), respectively, each associated with the activation of an overlapping set of defence genes. The regulatory mechanism behind this convergence of PTI‐ and ETI‐mediated defence gene induction remains elusive. We generated transgenic Arabidopsis plants that enable conditional NLR activation without pathogen infection to dissect NLR‐ and PRR‐mediated transcriptional signals. A comparative analysis of over 40 transcriptome datasets linked calmodulin‐binding transcription activators (CAMTAs) to the activation of overlapping defence genes in PTI and ETI. We used a dominant camta3 mutant (camta3‐D) to assess CAMTA functions in the corresponding transcriptional regulation. Transcriptional regulation by NLRs, although highly similar to PTI responses, can be established independently of pathogen‐associated molecular pattern (PAMP) perception, defence phytohormones and host cell death. Conditional expression of the N‐terminal coiled‐coil domain of the barley MLA (Mildew resistance locus A) NLR is sufficient to trigger similar transcriptional reprogramming as full‐length NLRs. CAMTA‐binding motifs are overrepresented in the 5′ regulatory regions of the identified primary immune response genes, consistent with their altered expression and disease resistance responses in camta3‐D plants. We propose that CAMTA‐mediated transcriptional regulation defines an early convergence point in NLR‐ and PRR‐mediated signalling.

Cistrome data for CAMTA proteins and ABA--responsive element--binding proteins [submitted as a separate file] Methods S1 Methods related to RNA--seq data acquisition including pathogen inoculation and transcriptomic analysis References related to Supporting Information   A plant line that conditionally expresses monomeric YFP (mYFP) was used as a negative control. MLA CC variants are C--terminally fused with mYFP and MLA FL variants are C--terminally fused with 3xTY1 epitope tags. (a) DEX--inducible expression of the autoactive MLA variants MLA FL (D502V) and MLA CC leads to growth retardation in Arabidopsis stable transgenic lines. The transgenic lines used in this assay are indicated in the left panels. The pictures were taken at 16 days after sowing on medium without DEX (--DEX) or with 10 µM DEX (+ DEX). (b)--(g) Samples were collected from leaves of four--week--old plants after infiltration of 1 µM DEX at the indicated time points. (b) Steady state levels of the MLA variants/truncated forms after DEX infiltration. Samples were examined by immunoblotting anti--GFP or anti--TY1. Ponceau S staining was used to monitor equal loading. (c) Quantification of cell death by ion leakage measurement. The data represent the mean and standard deviation of at least nine biological replicates obtained from at least three independent experiments with at least three biological replicates each. Statistical differences between the genotypes were assessed using ANOVA followed by Tukey HSD (P<0.05). The arrows indicate the earliest time point where statistical differences between the autoactive and inactive variants can be detected. (d) Temporal analysis of protein steady state levels by immunoblotting with anti--GFP antibodies in the lines expressing MLA CC --mYFP and mYFP under a DEX--inducible promoter. Ponceau S staining was used to monitor equal loading.      log 2 FC (treatment/control) values between each of the other treatments and the conditional MLA CC expression at 2 hours post induction; and percentage of genes in this cluster with more than 2--fold induction (%up) in the indicated condition. The position of the seven genes selected for RT--qPCR and induced upon MLA CC expression at 2 hours post induction is indicated by red ticks: AT1G30370 (1), AT4G39670 (2), AT5G41730 (3), AT1G08860 (4), AT5G42380 (5), AT2G24850 (6), and SARD1 (7). A representative subset of the data presented here is shown in Fig.  2. Information for the datasets used in this study is provided in Table  S1. NA, not available (expression not detected).

Methods S1 Methods related to RNA--seq data acquisition including pathogen inoculation and transcriptomic analysis RNA--seq
Leaf samples from at least five independent plants were collected at the indicated time point after treatment. Total RNA was extracted using the RNeasy Plant Mini Kit (Qiagen, Venlo, Netherlands). Total RNA was DNase--treated and purified using RNase--Free DNase Set (Qiagen) and the RNeasy MinElute kit (Qiagen). Ribosomal RNAs were depleted from 2.5 µg RNA using the RiboMinus Plant Kit (Thermo Fisher Scientific, Waltham, MA, USA) for RNA--seq. For the RNA--seq data generated in this study, mRNA sequencing libraries were prepared with barcoding using the TruSeq RNA Sample Preparation Kit (Illumina, San Diego, CA, USA). All procedures were performed according to the manufacturer's instructions. The barcoded libraries were pooled together and sequenced by Illumina HiSeq2000 or Illumina HiSeq2500, yielding between 18 and 40 Mio 100 bp reads per sample. Unless otherwise indicated, three independent biological replicates were processed and analysed for each experimental condition.

RNA--seq data analysis
Raw RNA--seq data were collected from public datasets or obtained in this study (Table S1). Total reads were mapped onto the A. thaliana genome (TAIR10) using TopHat2 (Kim et al., 2013). Read counts per gene were calculated from the mapped RNA--seq reads using HTseq-count function, apart from our previously published 35S:MLA1 pps dataset for which the CoverageBed (bedTools suite, Quinlan & Hall, 2010) had been used. Genes with less than 100 reads within one dataset were discarded, and the log 2 --transformed count data was normalized using the function voom from the R package limma (www.r--project.org, Smyth, 2005) resulting in log 2 counts per million. Differential gene expression between genotypes and/or treatments and/or times, was analysed by fitting a linear model with the appropriate explanatory variables using the function lmFit (R package limma). Log 2 --transformed fold change (log 2 FC) values between sample and control at the same time point were used for most of the comparative transcriptomic analysis conducted in this study. Statistical analysis was performed using moderated t--tests over the contrasts of interests and the resulting p--values were corrected for multiple testing using the Benjamini--Hochberg method. The criteria for significant differential expression were: FDR < 0.01 and |log 2 FC| > 1. MDS plots were created in R from the TMM normalized log2 counts per million using the function plotMDS (R package lima) with default parameter settings, i.e. for each pair of samples the distance is calculated from the top 500 genes that best distinguish these two samples (unless the analyzed gene set contains less than 500 genes, in which case all genes will be included).

Microarray data analysis
Publicly available experiments using the Affymetrix ATH1--121501 platform were obtained from several data sources (Table  S1). Only experiments including at least three biological replicates were selected. The raw .cel files were downloaded and normalized with Robust Multi--array Average (RMA) normalization as implemented in BioConductor (Irizarry et al., 2003;Gentleman et al., 2004). Probe annotation was performed using the ath1121501ACCNUM from the ath1121501.db annotation data. Probes with no or ambiguous annotations were removed. For each dataset the log 2 --base fold changes (log 2 FC) of treatment versus control were computed by fitting a linear model with the appropriate explanatory variables using the function lmFit (R package limma). This log2FC was used for the comparative transcriptomic analysis described in this study. When necessary, differentially expressed genes were extracted using the R/Bioconductor package limma (Ritchie et al., 2015) with the criteria |log 2 FC|>1 and FDR<0.05.

Transcript quantification by RT--qPCR
Leaf samples from at least five independent plants were collected at the indicated time points after treatment. Total RNA was extracted with the RNeasy Plant Mini Kit (Qiagen) according to the manufacturer instructions. Reverse transcription was performed using the Quantitec Reverse Transcription kit (Qiagen) according to the manufacturer's instructions. The cDNA was subjected to qPCR with gene--specific primers using the IQ SYBR Green reagent (Bio--Rad, Hercules, CA) on the iQ5 Multicolor Real--Time PCR Detection System (Bio--Rad) with three technical replicates for each of the three biological replicates unless otherwise stated. The relative expression was normalized to AT4G26410, which was previously described as highly constant under varying stress conditions (Czechowski et al., 2005). For validation of the RNA-seq experiment and further characterization of the CAMTA3--dependent gene regulation, we selected 6 genes among the 10 most induced IE genes at 2 hours post MLA CC induction. CBP60g (AT5G26920) and SARD1 (AT1G73805) were also included due to their known function in the transcriptional regulation of immune responses (Wang et al., 2011;Sun et al., 2015). To visualise the RT--qPCR--derived expression profiles of selected IE genes, the means of the normalized relative expression values from at least two independent experiments were plotted using the pheatmap function (R package pheatmap). For the corresponding heatmaps, genes and samples were clustered according to their averaged relative expression levels, using complete linkage hierarchical clustering with the Euclidean distance as distance measure.

Motif discovery and enrichment analysis
Promoter element enrichment analysis was performed using several tools: MEME--Lab (Brown et al., 2013) was used to find the eight most enriched 6--14 bp elements in promoters of 600 bp length. Scope motif finder was used to find enriched elements in promoters of 1000 bp length, based on BEAM, PRISM, and SPACER programs (http://genie.dartmouth.edu/scope, (Carlson et al., 2007). Weeder1.4 and Weeder2 were run with default settings to identify enriched elements in promoters of 1000 bp length (Pavesi et al., 2006;Zambelli et al., 2014). Pscan was used to assess the enrichment of already known transcription factor binding sites in promoters of 1000 bp length, based on the Jaspar 2016 database (Zambelli et al., 2009;Mathelier et al., 2015). The different outputs were compared to identify motifs consistently and independently picked up by different methods. The RSAT DNA pattern matching tool was used to find occurrences of the selected motifs in the 500--bp regions directly upstream of the transcription start site for the gene set of interest and for the complete set of all A. thaliana promoter sequences used as background reference (TAIR10_upstream_1000_20101104.fas or TAIR10_upstream_500_20101028.fas) (Medina--Rivera et al., 2015). The resulting occurrence data was used to calculate the enrichment false discovery rate (FDR) by applying a cumulative hypergeometric distribution with Benjamini--Hochberg correction for multiple testing in R (p.hyper and p.adjust functions). For representation of the motif frequency along the 5' regulatory sequences, the function seqPattern from the seqPattern package in R was used, based on sequences extracted from TAIR10_upstream_1000_20101104.fas and TAIR10_seq_20110103_representative_gene_model_updated.fas. GO term enrichment analysis Gene ontology (GO) term enrichment analysis was performed using the agriGO (Du et al., 2010) web tool with default settings for Arabidopsis thaliana.

Pathogen infections
Conidiospores of Blumeria graminis f. sp. hordei (isolate K1 or A6) were inoculated onto the abaxial side of detached leaves placed on agar plates containing 1 mM benzimidazole. The agar plates sealed with surgical tape were placed in a phytochamber (10 h : 14 h, light : dark cycle at 22°C, 60% relative humidity) until sample collection. The Tape--Arabidopsis Sandwich method (Wu et al., 2009) was applied to obtain the abaxial epidermis from the leaf. Approximately 20 leaves per condition were used. The abaxial epidermis with tape was ground in a mortar using a pestle with liquid nitrogen and total RNA was extracted with the Plant RNeasy kit (Qiagen) with the following modifications: the ground powder was mixed with 1.8 ml of the lysis buffer by vortex. Immediately after vortex, the lysate was centrifuged to separate the tape from the lysate. The remaining lysate was used for total RNA extraction according to the manufacturer instructions. Preparation of and inoculation with Pseudomonas syringae pv. DC3000 (Pst) expressing AvrRpm1, AvrRps4 or neither of these effectors was performed as described before (Liu et al., 2015). Young fully expanded leaves were syringe--infiltrated with a bacterial solution in 10 mM MgCl 2 after adjusting the OD 600 to 0.05, 0.001, and 0.0001 for ion leakage assay, RT--qPCR, and bacterial growth assays, respectively. Pst growth assays were performed as described previously (Liu et al., 2015) with following modifications: NYGA medium (5 g /l bactopeptone, 3 g/l yeast extract, 20 ml/l glycerol, with/without 15 g /l agar) with appropriated antibiotics was used for bacterial culture. Further humidity control by covering the plants was omitted. The infiltrated plants were kept in a phytochamber (10 h : 14 h, light : dark cycle at 22°C, 60% relative humidity) until sample collection.

Syringe infiltration of dexamethasone or flg22 into leaves
Unless otherwise mentioned, expression of transgenes under the dexamethasone--inducible promoter was performed by syringe infiltration of 1 µM dexamethasone (DEX, D1756--1G, Sigma--Aldrich, St. Louis, MO, USA). Treatment with flg22 was performed by syringe infiltration of 100 nM and 1 µM of the synthetic peptides for the protein and bacterial growth experiments, respectively. In both cases, each solution was infiltrated into the youngest fully expanded leaves with a 1 ml needle--less plastic syringe. Temperature shift RNA--seq experiments Temperature--shift induction of TNL RPS4 immunity in 35S:RPS4--HS transgenic plants was as described in (Heidrich et al., 2013). Seeds were sown directly onto pots (10 seeds--pot) covered with a lid, placed for three days at 4 and subsequently transferred to 28°C under short day conditions (10 h : 14 h, light : dark cycle, 60% relative humidity). The lid was removed one week after germination. At 3.5 weeks after germination, plants were transferred to 19°C (10 h : 14 h, light : dark cycle, 55% relative humidity) and sampled at the indicated time points after shift. The temperature shift was conducted at different times of the day to allow simultaneous sampling between 16:00 h and 17:00 h for all time points. Leaf material from 4--6 individual plants was pooled for each sample and used subsequently for RNA extraction and RNA--seq. The experiment was repeated independently three times. Data for the 2 h time point was obtained from an experiment distinct from the later time points (Table S1).