Hypersensitivity to triforine in lettuce is triggered by a TNL gene through the disease‐resistance pathway

The majority of cloned disease resistance genes (R-genes) encode proteins with nucleotide-binding leucine-rich repeat domains (NLRs). R-genes tend to be physically clustered, and the structure of the cluster often facilitates expansion and sequence exchange among R-gene homologs (Luo et al. 2012). NLR proteins interact directly or indirectly with pathogens effectors, often triggering programmed cell death, also known as hypersensitive response (HR) at the infected sites (Dangl and Jones 2001). HR may be triggered by pesticide molecules rather than pathogen effectors. For example, some tomato cultivars are sensitive to fenthion, developing toxic lesions after exposure to fenthion (Martin et al. 1994). Similarly, some lettuce germplasms are highly sensitive to triforine, an active ingredient in some commercial fungicides, with leaves showing wilting and necrosis 24 hours after exposure to triforine (Figure 1a). Sensitivity to triforine in lettuce is controlled by a single locus (Tr) (Simko et al. 2011), however, the causal gene and its molecular mechanism remain unknown.

The majority of cloned disease-resistance genes (R-genes) encode proteins with nucleotide-binding leucine-rich repeat domains (NLRs). R-genes tend to be physically clustered, and the structure of the cluster often facilitates expansion and sequence exchange amongst R-gene homologues (Luo et al., 2012). NLR proteins interact directly or indirectly with pathogens effectors, often triggering programmed cell death, also known as hypersensitive response (HR) at the infected sites (Dangl and Jones, 2001). HR may be triggered by pesticide molecules rather than pathogen effectors. For example, some tomato cultivars are sensitive to fenthion, developing toxic lesions after exposure to fenthion (Martin et al., 1994). Similarly, some lettuce germplasms are highly sensitive to triforine, an active ingredient in some commercial fungicides, with leaves showing wilting and necrosis 24 h after exposure to triforine ( Figure 1a). Sensitivity to triforine in lettuce is controlled by a single locus (Tr) (Simko et al., 2011); however, the causal gene and its molecular mechanism remain unknown.
In this study, we used genome-wide SNPs to perform genomewide association studies (GWAS) on sensitivity to triforine. The results showed a significant signal on chromosome 1 (Chr1) (Figure 1b). The candidate region spans 5087 kb and the prominent candidate genes include a TNL-encoding family. To confirm the GWAS results and to clone the Tr gene, we constructed an F 2 population by crossing a sensitive genotype (PI344074, Romaine) with an insensitive genotype (PI536839, Crisphead). Using Bulk Segregant Analysis + RNA-seq, the Tr gene was mapped to Chr1 ( Figure 1c). Then, a total of 4639 individuals from an F 3 family were first screened using two farend flanking markers and recombinants were further genotyped using markers in the candidate region. The Tr gene was ultimately fine mapped between markers AGH372 (F-primer: AACTTGACATTCTTCGGTG/R-primer:CTTCTGTTTAGTACAACA TT) and AGH371 (F-primer:TTTAGATACCTATGACAACTT/Rprimer:GTATATGTATCTATGTCTATGT), with an interval of approximately 140 kb (Figure 1d). This region of the reference genome (Lactuca sativa cv Salinas V8) has five genes, and all of them belong to a TIR type NLR-encoding family (TNL). Thus, we hypothesize that the Tr gene in sensitive parents was a homologue of this R-gene family.
To obtain the Tr gene, we used conserved primers to PCR amplify homologues of the R-gene family from the two parental genotypes, PCR products were cloned, and individual colonies were sequenced. Twenty-one and nine distinct Tr homologues were obtained from the sensitive and insensitive parents, respectively. Markers specific to each homologue were designed and used to screen the recombinants. The genetic analysis showed that only one (Tr-like109) homologue from the sensitive parent co-segregated with sensitivity to triforine. We also used the same pair of conserved primers to amplify homologues of the R-gene family from 29 sensitive accessions, PCR products were pooled and sequenced using Illumina Hiseq2500 platform. Similarly, PCR products from 46 insensitive individuals were also sequenced. Nine SNPs specific to the sensitive pool were found, which all present in the homologue Tr-like109. Therefore, the Tr-like109 gene is very likely the candidate for the Tr. Indeed, transformation of the Tr-like109 gene into the insensitive accession changed its reaction to triforine from insensitive to sensitive in the transformants ( Figure 1e). On the other hand, three CRISPR knock-out lines of the Tr-like109 gene in the sensitive accession changed its sensitive reaction to insensitive, confirming that Tr-like109 is the Tr gene encoding sensitivity (susceptibility) to triforine (Figure 1f).
Tr transcripts were detected at multiple developmental stages in sensitive individuals. The expression of the Tr gene increased with leaf age in the first month after germination, then the increase slowed and the expression peaked in the second month, and maintained a high expression level for at least one more month (Figure 1h). We also analysed the expression of genes associated with disease resistance in parents, complementary and knock-out line after treatment with triforine. The Tr gene was rapidly up-regulated after treatment with triforine, and similar upregulation was also found for some genes associated with disease-resistance response (Figure 1i-j).
To verify that the sensitive response to triforine followed the same pathway as the HR in disease-resistance, we knocked out the EDS1 (LG1_140621) gene in the sensitive accession, which is required in the resistance pathway for TNL proteins. The homozygous eds1 mutants were insensitive to triforine (Figure 1g). We, therefore, conclude that the Tr gene triggers the HR response through the disease-resistance pathway. Next, we investigated whether the accumulation of ROS was associated with sensitivity to triforine. DAB, NBT as well as H 2 O 2 content showed that the accumulation of ROS in the leaves originating from triforine-sensitive individuals were higher than leaves This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. originating from triforine-insensitive individuals after triforine treatment (Figure 1k-l). The enzymatic activities of superoxide dismutase (SOD), peroxidase (POD) and catalase (CAT) were considerably higher in the sensitive genotype than in the insensitive genotypes (Figure 1m). Sensitivity to triforine was alleviated if the sensitive plants were pretreated with the ROSinhibitor diphenyleneiodonium chloride (Figure 1k). We used primers specific to the Tr gene to amplify PCR products from 817 lettuce accessions (203 cultivated lettuce (Lactuca sativa) and 614 wild lettuce (Lactuca serriola)). As expected, PCR products were obtained from all sensitive accessions, including 26 cultivated and six wild accessions. PCR products were also obtained from two insensitive accessions. The Tr sequences from all sensitive lettuce cultivars and two sensitive L. serriola accessions (CGN17427, CGN21383) were identical (Figure 1n). Therefore, the Tr gene likely originated from L. serriola and underwent at least five sequence exchanges (SE) during domestication or introgression (Figure 1o).
In this study, the Tr gene was identified through GWAS and map-based cloning. The candidate region contains a large R-gene family, which made the identification of the candidate gene challenging. We exhaustively sequenced the R-gene family in the two parents, and a large segregating population was used to narrow down the candidate gene, which facilitated the following process of gene verification. The Tr gene, encoding extreme sensitivity to triforine, has potential applications in plant biotechnology. For example, the Tr gene, if included in a transformation vector, can be used for larger-scale selection for marker-free individuals. It can be also used as a selection marker in mutagenesis to study the HR signalling transduction pathway.

Accession numbers
The data sets are available in the NCBI (PRJNA689977). The sequences of the Tr gene and its homologues are available in GenBank (MW451218-MW451224). Figure 1 Tr triggered hypersensitivity to triforine in lettuce. (a) Detached leaves from the one-month-old, sensitive and insensitive cultivars before/after triforine treatment. (b) Genome-wide association study of the sensitivity to triforine in lettuce. An R script was used to generate quantile-quantile plots. A significant signal is shown on linkage group 1. (c) BSR-seq analysis of the Tr gene. The Dvalue was plotted along with the nine linkage groups of lettuce (Xaxis). The red and green curves refer to confidence intervals of P = 0.05 and P = 0.01, respectively. (d) The Tr gene was fine mapped to 140 kb region. The numbers below the horizontal line refer to the number of recombinants between two markers from 4639 progenies. (e) The complementation test changed plants insensitive to triforine to sensitive. All Tr gene (f) and EDS1 gene (g) knock-out lines caused frame-shift deletions, resulting in the phenotypic change from sensitive to insensitive. (h-j) The expression level of the Tr gene and other genes associated with disease resistance. (k-m) Sensitive genotypes have stronger ROS than insensitive genotypes when treated with triforine. 'Sensitive + DPI' refers to sensitive genotypes pretreated with diphenyleneiodonium chloride. Mean and standard deviation values were calculated using three biological replicates. (n-o) Reconstruction of sequence exchange events amongst Tr and its six homologues. Rectangle with the same colour represents identical sequences.