The pivotal ripening gene SlDML2 participates in regulating disease resistance in tomato

Abstract

Fruit ripening and disease resistance are two essential biological processes for quality formation and maintenance. DNA methylation, in the form of 5-methylcytosine (5mC), has been elucidated to modulate fruit ripening, but its role in regulating fruit disease resistance remains poorly understood. In this study, we show that mutation of SlDML2, the DNA demethylase gene essential for fruit ripening, affects multiple developmental processes of tomato besides fruit ripening, including seed germination, leaf length and width and flower branching. Intriguingly, loss of SlDML2 function decreased the resistance of tomato fruits against the necrotrophic fungal pathogen Botrytis cinerea. Comparative transcriptomic analysis revealed an obvious transcriptome reprogramming caused by SlDML2 mutation during B. cinerea invasion. Among the thousands of differentially expressed genes, SlβCA3 encoding a β-carbonic anhydrase and SlFAD3 encoding a ω-3 fatty acid desaturase were demonstrated to be transcriptionally activated by SlDML2-mediated DNA demethylation and positively regulate tomato resistance to B. cinerea probably in the same genetic pathway with SlDML2. We further show that the pericarp tissue surrounding B. cinerea infection exhibited a delay in ripening with singnificant decrease in expression of ripening genes that are targeted by SlDML2 and increase in expression of SlβCA3 and SlFAD3. Taken together, our results uncover an essential layer of gene regulation mediated by DNA methylation upon B. cinerea infection and raise the possible that the DNA demethylase gene SlDML2, as a multifunctional gene, participates in modulating the trade-off between fruit ripening and disease resistance.

Keywords: Botrytis cinerea; DNA methylation; SlDML2; defence response; tomato; transcriptome reprogramming.

Figure 1

SlDML2 participates in modulating multiple developmental processes. (a) Genotyping of sldml2‐3, sldml2‐4, and sldml2‐5 mutants generated by CRISPR/Cas9‐mediated gene editing. Three single guide RNAs (sgRNAs) containing different target sequences (T1, T2, and T3) were designed to specifically target the first exon of SlDML2. The red letters indicate the protospacer adjacent motif (PAM). The sldml2‐3 mutant contains a homozygous 2‐bp deletion, and the sldml2‐4 and sldml2‐5 mutants contain a homozygous 5‐bp deletion caused by target T1. (b) Phenotypes of fruit ripening in sldml2 mutants. Representative photographs of the wild‐type (WT), sldml2‐3, sldml2‐4, and sldml2‐5 fruits at 39, 42, 47, and 52 dpa are shown. dpa, days post‐anthesis. (c) Representative photographs and (d) germination rates of the WT, sldml2‐3, and sldml2‐4 seeds. For each germination experiment, one‐hundred mature seeds were cultured with deionized water under long‐day conditions (16 h light/8 h dark, 25 °C). The gemmiparous seeds were photographed at the fourth day after imbibition. Data are presented as mean ± standard deviation (n = 3). (e) Representative photographs and (f) length, width, and weight of the WT, sldml2‐3, and sldml2‐4 leaves. Six‐week‐old leaves were collected, photographed, and measured. (g) Representative photographs and (h) flower number of the WT, sldml2‐3, and sldml2‐4 inflorescences. The first inflorescences were collected and photographed, and the flower number in each inflorescence was measured. In (c), (e), and (g), scale bar = 1 cm. In (f and h), asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001; Student's t test).

Figure 2

The sldml2 mutants exhibited decreased resistance to fungal pathogens B. cinerea and V. dahliae. (a) Disease symptoms and (b) lesion diameters on the detached wild‐type (WT), sldml2‐3, and sldml2‐4 fruits inoculated with B. cinerea at 39 days post‐anthesis (dpa). The disease symptom was observed after 60 h. (c) Disease symptoms and (d) lesion diameters on the detached WT, sldml2‐3, and sldml2‐4 fruits inoculated with B. cinerea at 90 dpa. The disease symptom was observed after 48 h. (e) Disease symptoms and (f) lesion diameters on 4‐week‐old detached leaves of WT, sldml2‐3, and sldml2‐4 inoculated with B. cinerea. The disease symptom was observed after 48 h. (g) Representative photographs and (h) stem lengths of the WT, sldml2‐3, and sldml2‐4 seedlings that were either inoculated with V. dahliae or mock‐inoculated for 18 days. (i) Relative V. dahliae biomass in stems of the WT, sldml2‐3, and sldml2‐4 seedlings that were inoculated with V. dahliae for 18 days. The amounts of V. dahliae GADPH gene were determined by quantitative RT‐PCR analysis, and the tomato RuBisCo gene was used as an internal control. (j) Disease symptoms and trypan blue staining and (k) bacterial biomass in WT, sldml2‐3, and sldml2‐4 leaves that were inoculated with Pst DC3000 for 3 days. In (a), (c), (e), (g), and (j), scale bar = 1 cm. In (b, d, f, h, i and k), asterisks indicate significant differences (**P < 0.01, ***P < 0.001; Student's t test). NS, no significance.

Figure 3

SlDML2 mutation causes transcriptome reprogramming during B. cinerea infection. (a) Volcano plot showing up‐regulated (red) and down‐regulated (blue) genes in the sldml2‐3 mutant compared to the wild‐type (WT). (b) Heat map of differentially expressed genes in the sldml2‐3 mutant compared with the wild‐type. FPKM, fragments per kilobase of exon per million mapped fragments. (c) Gene ontology (GO) enrichment for the differentially expressed genes in the sldml2‐3 mutant compared with the WT. The top 20 catalogues in biological process with the most significant P value were shown. (d) KEGG analysis for the differentially expressed genes in the sldml2‐3 mutant compared with the WT. The top 20 pathways with the most significant P value were shown. The differentially expressed genes were analysed by three independent RNA‐seq experiments in fruits of the WT and sldml2‐3 mutant after B. cinerea inoculation for 48 h.

Figure 4

SlβCA3 is involved in the defence response to B. cinerea. (a) Transcription levels of the tomato β‐type carbonic anhydrase gene SlβCA3 in fruits of the wild‐type (WT) and sldml2 mutants after B. cinerea inoculation for 48 h. (b) Transcription levels of the SlβCA3 gene in leaves of the WT and sldml2 mutants after B. cinerea inoculation for 48 h. (c) Transcription levels of the SlβCA3 gene in WT fruits with or without B. cinerea inoculation at the indicated time. (d) Transcription levels of the SlβCA3 gene in WT leaves with or without B. cinerea inoculation at the indicated time. In (a–d), the SlβCA3 transcription levels were determined by quantitative RT‐PCR analysis, and the tomato SlUBI3 gene was used as an internal control. Asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001; Student's t test). NS, no significance. (e) Transient expression of SlβCA3‐HA protein in N. benthamiana leaves. Total protein was extracted at indicated time after agroinfiltration and then submitted to immunoblot with anti‐HA antibody. Equal loading was confirmed by using the tomato Actin as an internal control. (f) B. cinerea disease symptoms and lesion diameters on detached N. benthamiana leaves with or without the co‐expression of SlβCA3‐HA. Asterisks indicate significant differences (**P < 0.01; Student's t test). Scale bar = 1 cm.

Figure 5

SlβCA3 is transcriptionally regulated by SlDML2‐mediated DNA demethylation. (a) DNA methylation (5mC) level of the SlβCA3 gene body in fruits of the wild‐type (WT) and sldml2 mutant at indicated developmental stages. The 5mC level was analysed by using the published DNA methylome database (Lang et al., 2017). Each vertical bar represents an 5mC and the bar height indicates 5mC level. The differentially methylated region (DMR) in the first intron of SlβCA3 gene was indicated by a shadow box. 5mC levels of representative cytosines in the DMR were shown with pie charts, and the numbers indicate the positions relative to the start codon. dpa, days post‐anthesis. (b) Bisulphite sequencing showing 5mC levels of the representative cytosines shown in (a) in fruits of the wild‐type and sldml2 mutants (sldml2‐3 and sldml2‐4) at 46 dpa. (c) 5mC level of the DMR of SlβCA3 gene in fruits of the WT and sldml2 mutants after B. cinerea inoculation for 48 h. (d) ChIP‐qPCR assay showing that SlDML2 binds to the DMR in the first intron of SlBCA3 gene. The wild‐type fruits at 42 dpa were subjected to the immunoprecipitation with the anti‐SlDML2 polyclonal antibody, and the pre‐immune IgG was used as a control. The first intron structure of SlβCA3 is shown, and the red line indicates the region for PCR. (e) Transcription levels of the SlβCA3 gene in N. benthamiana leaves with or without the co‐expression of SlDML2‐HA. The sequence of SlβCA3 gene body containing all exons and introns was cloned into the dual‐luciferase reporter vector for transcription activity assay under the driving of its native promoter. The renilla luciferase gene RLUC was used as an internal control. (f) 5mC level of the DMR of SlβCA3 gene in N. benthamiana leaves with or without the co‐expression of SlDML2‐HA. (g) Transcription levels of the mutated SlβCA3 gene under the drive of its native promoter in N. benthamiana leaves with or without the co‐expression of SlDML2‐HA. The representative cytosines shown in (a) were mutated to adenine or thymine. NS, no significance. In (c) and (f), 5mC level was revealed by McrBC‐PCR assay. A total of 0.4 μg genomic DNA was digested by McrBC enzyme with GTP (+GTP), or without GTP (−GTP) as a negative control. An unmethylated region in the promoter of tomato PME gene was used as an internal control. In (d and e), asterisks indicate significant differences (**P < 0.01; Student's t test).

Figure 6

SlFAD3 is transcriptionally regulated by SlDML2‐mediated DNA demethylation. (a) Transcription levels of the tomato ω‐3 fatty acid desaturase gene SlFAD3 in fruits of the wild‐type (WT) and sldml2 mutants after B. cinerea inoculation for 48 h. (b) Transcription levels of the SlFAD3 gene in leaves of the WT and sldml2 mutants after B. cinerea inoculation for 48 h. (c) Transient expression of SlFAD3‐HA protein in N. benthamiana leaves. Total protein was extracted at indicated time after agroinfiltration and then submitted to immunoblot with anti‐HA antibody. Equal loading was confirmed by using the tomato Actin as an internal control. (d) B. cinerea disease symptoms and lesion diameters on detached N. benthamiana leaves with or without the co‐expression of SlFAD3‐HA. Scale bar = 1 cm. (e) DNA methylation (5mC) levels of the SlFAD3 promoter in fruits of the WT and sldml2 mutant at indicated developmental stages. The 5mC level was analysed by using the published DNA methylome database (Lang et al., 2017). Each vertical bar represents a 5mC and the bar height indicates 5mC level. The differentially methylated region (DMR) in the SlFAD3 promoter was indicated by a shadow box. 5mC levels of the representative cytosines in the DMR were shown with pie charts, and the numbers indicate the positions relative to the start codon. dpa, days post‐anthesis. (f) Bisulphite sequencing showing 5mC levels of the representative cytosines shown in (e) in fruits of the wild‐type and sldml2 mutants (sldml2‐3 and sldml2‐4) at 46 dpa. (g) 5mC level of the DMR of SlFAD3 promoter in fruits of the WT and sldml2 mutants after B. cinerea inoculation for 48 h. (h) ChIP‐qPCR assay showing that SlDML2 binds to the DMR in the promoter of SlFAD3. The wild‐type fruits at 42 dpa were subjected to the immunoprecipitation with the anti‐SlDML2 polyclonal antibody, and the pre‐immune IgG was used as a control. The promoter structure of SlFAD3 is shown, and the red line indicates the region for PCR. (i) Relative firefly luciferase (FLUC) activity derived by the SlFAD3 promoter in N. benthamiana leaves with or without the co‐expression of SlDML2‐HA. The representative image was shown. The FLUC activity was normalized against the renilla luciferase (RLUC) activity, followed by normalization against the control. (j) Transcription levels of the FLUC gene under the drive of SlFAD3 promoter in N. benthamiana leaves with or without the co‐expression of SlDML2‐HA. The RLUC gene was used as an internal control. (k) 5mC level of the DMR of SlFAD3 promoter in N. benthamiana leaves with or without the co‐expression of SlDML2‐HA. (l) Transcription levels of the FLUC gene under the drive of mutated SlFAD3 promoter in N. benthamiana leaves with or without the co‐expression of SlDML2‐HA. The representative cytosines shown in (e) were mutated to adenine or thymine. Ns, no significance. In (g) and (k), total 5mC level was revealed by McrBC‐PCR assay. A total of 0.4 μg genomic DNA was digested by McrBC enzyme with GTP (+GTP), or without GTP (−GTP) as a negative control. An unmethylated region in the promoter of tomato PME gene was used as an internal control. In (a, b, d, h–j), asterisks indicate significant differences (*P < 0.05, ***P < 0.001; Student's t test).

Figure 7

SlDML2 functions in the same genetic pathway with SlβCA3 and SlFAD3 to regulate tomato resistance to B. cinerea. (a) Transcription levels of SlβCA3 and SlFAD3 in leaves of the sldml2 mutant after virus‐induced gene silencing (VIGS), as determined by quantitative RT‐PCR. The sldml2 mutant infiltrated with the empty vector pTRV2 was used as the control group. The tomato Actin gene was used as an internal control. (b) PCR amplification showing that the virus vectors pTRV1, pTRV2, pTRV2‐SlβCA3, and pTRV2‐SlFAD3 were successfully expressed in leaves of the wild‐type (WT) or sldml2 mutant. (c) Disease symptoms and (d) lesion diameters on leaves of the WT/TRV2, sldml2/TRV2, sldml2/TRV2‐SlβCA3, and sldml2/TRV2‐SlFAD3 inoculated with B. cinerea for 40 h. (e) Transcription levels of SlβCA3 and SlFAD3 in leaves of the sldml2 mutant after overexpression, as determined by quantitative RT‐PCR. The sldml2 mutant infiltrated with the empty vector pCambia1302‐HA was used as the control groups. The tomato Actin gene was used as an internal control. (f) Western blot showing that SlβCA3‐HA and SlFAD3‐HA fusion proteins were successfully expressed in leaves of sldml2/35S _pro :SlβCA3 And sldml2/35S _pro :SlFAD3. Total protein was extracted and submitted to immunoblot with anti‐HA antibody. Equal loading was confirmed by using the tomato Actin as an internal control. (g) Disease symptoms and (h) lesion diameters on leaves of the WT, sldml2, sldml2/35S _pro :SlβCA3, And sldml2/35S _pro :SlFAD3 Inoculated with B. cinerea for 40 h. In (a, d, e, and h), asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001; Student's t test). NS, no significance. In (c and g), scale bar = 1 cm.

Figure 8

B. cinerea invasion causes a delay in ripening of pericarp tissues around the disease region. (a) Representative photograph of wild‐type tomato fruits after B. cinerea inoculation for 4 days. The pericarp tissue with a 5‐mm thickness surrounding disease regions was collected as “response region”, and the pericarp tissue with a 5‐mm thickness around the response region was collected as “control region”. (b) Lycopene content of the response region and control region. (c) Transcription levels of ripening‐related genes in the response region and control region as determined by quantitative RT‐PCR analysis. The tomato SlUBI3 gene was used as an internal control. PG2a, polygalacturonase 2a; EXP1, expansion 1; RIN, ripening inhibitor; CNR, colourless nonripening; NOR, nonripening; NR, never ripe; PSY1, phytoene synthase 1; PDS, phytoene desaturase. (d) Transcription levels of SlβCA3 and SlFAD3 in the response region and control region as determined by quantitative RT‐PCR analysis. The tomato SlUBI3 gene was used as an internal control. In (b–d), asterisks indicate significant differences (*P < 0.05, **P < 0.01, ***P < 0.001; Student's t test). (e) Model for SlDML2‐mediated regulation of tomato resistance to B. cinerea. SlDML2 transcriptionally activates the expression of defence‐related genes SlβCA3 and SlFAD3 through active DNA demethylation in the gene body and promoter region, respectively, thus positively regulating tomato fruit resistance to B. cinerea.