https://orcid.org/0009-0004-2022-7234
Abstract
Cytospora canker, caused by Cytospora mali, is the most destructive disease in production of apples (Malus domestica). Adding potassium (K) to apple trees can effectively control this disease. However, the underlying mechanisms of apple resistance to C. mali under high-K (HK) status remain unknown. Here, we found that HK (9.30 g/kg) apple tissues exhibited high disease resistance. The resistance was impeded when blocking K channels, leading to susceptibility even under HK conditions. We detected a suite of resistance events in HK apple tissues, including upregulation of resistance genes, callose deposition, and formation of ligno-suberized tissues. Further multiomics revealed that the phenylpropanoid pathway was reprogrammed by increasing K content from low-K (LK, 4.30 g/kg) status, leading to increases of 18 antifungal chemicals. Among them, the physiological concentration of coumarin (1,2-benzopyrone) became sufficient to inhibit C. mali growth in HK tissues, and exogenous application could improve the C. mali resistance of LK apple branches. Transgenic apple calli overexpressing beta-glucosidase 40 (MdBGLU40), which encodes the enzyme for coumarin synthesis, contained higher levels of coumarin and exhibited high resistance to C. mali even under LK conditions. Conversely, the suppression of MdBGLU40 through RNAi reduced coumarin content and resistance in HK apple calli, supporting the importance of coumarin accumulation in vivo for apple resistance. Moreover, we found that the upregulation of transcription factor MdMYB1r1 directly activated MdBGLU40 and the binding affinity of MdMYB1r1 to the MdBGLU40 promoter increased in HK apple tissue, leading to high levels of coumarin and resistance in HK apple. Overall, we found that the accumulation of defensive metabolites strengthened resistance in apple when raising K from insufficient to optimal status, and these results highlight the optimization of K content in fertilization practices as a disease management strategy.
Introduction
Cytospora canker, also known as Valsa canker, caused by the necrotrophic fungus Cytospora mali Grove (synonym: Valsa mali Miyabe & G. Yamada), is one of the most destructive diseases of apple (Malus domestica) in East Asia, resulting in severe yield losses and decline in fruit quality (Cao et al. 2009; Fan et al. 2020). The pathogen causes extensive necrotic cankers on branches and even kills entire apple trees when the cankers girdle the trunks (Peng et al. 2016). Extensive surveys have shown that the incidence of Cytospora canker averaged about 53% in the major apple production regions of China and was as high as 80% in severely affected orchards (Cao et al. 2009). During the past decade, increasing numbers of apple orchards in the Loess Plateau region in north–central China have been cut down because of the high incidence of dead trees caused by Cytospora canker, leading to billions of Chinese Yuan lost per year. Traditionally, excising cankers followed by coating the affected areas with fungicide has been the main strategy for combating this disease (Cao et al. 2009). However, since the pathogen penetrates extensively into the host phloem and xylem, chemical treatment cannot effectively cure or control Cytospora canker, allowing for canker reinfection and recurrence year after year (Cao et al. 2009; Ke et al. 2014). Previous studies have shown that C. mali exhibits weak parasitism (He and Fang 1996; Ke et al. 2014), and successful infection and expansion is closely associated with apple tree vigor (He and Fang 1996). Considering the role of nutrient elements in determining tree vigor, we investigated the influence of nutrient status in apple trees on the occurrence of Cytospora canker and revealed that low potassium (K) levels in apple trees were consistently correlated with severe Cytospora canker in China and that raising K content of apple leaves to 1.3% provided efficient control of Cytospora canker (Peng et al. 2016). Foliar K content of apple trees from 1.3% to 1.8% is optimal for vigorous tree growth and high yields of fruit in the field (Marschner 2012). This meant that even when foliar K content was at the lower limit of optimal K (1.3% to 1.8%) (Marschner 2012), resistance to the pathogen was high. However, the mechanisms of strong resistance in apple tree to C. mali influenced by K status have not been elucidated.
Plants grow in highly diverse environmental conditions with differential availability of various nutrient elements (Cheng et al. 2019). As the most abundant cationic nutrient in plants, K is involved in numerous processes in plant cells, including gene transcription, enzyme activation, and cellular homeostasis (Shi et al. 2018). Potassium nutrient also profoundly affects plant resistance (Shi et al. 2018). For example, elevated K concentrations in rice (Oryza sativa) tissues enhance resistance to the rice blast fungus, Magnaporthe oryzae (Shi et al. 2018). The addition of K facilitates the development of stronger cell walls, upregulates resistance-related genes, and speeds up closure of stomata to bolster plant defenses (Shi et al. 2018; Wang et al. 2013). An adequate K supply alleviates lipid peroxidation and maintains lipid homeostasis to enhance rice resistance to sheath rot (Zhang et al. 2021). Despite research advances, the mechanisms of high resistance in plants with high-potassium status are largely unknown. Continuous encounters with pathogens have promoted the evolution of robust defense systems in plants, which requires various active defense events to fend off pathogens, including activation of pathogenesis-related proteins, cell wall reinforcement, and especially synthesis of defensive metabolites (Meng and Zhang 2013; Qiu et al. 2022; Szatmari et al. 2006). Defensive secondary metabolites, including preformed phytoanticipins and pathogen-induced phytoalexins, are derived mainly from the phenylpropanoid pathway (Piasecka et al. 2015). Coumarins are hydroxycinnamic phenolic acids synthesized from the phenylpropanoid pathway, and they mediate physiological processes at various growth and developmental stages of plants in a dose-dependent manner (Parvin et al. 2021). Coumarins are iron-mobilizing compounds secreted by plant roots to aid in direct iron uptake from iron-deprived soils (Harbort et al. 2020; Voges et al. 2019). Coumarins may improve plant tolerance to salinity by enhancing antioxidant defense, glyoxalase system, and ion homeostasis (Pergo et al. 2008). Coumarins are well-known defensive metabolite that performs important roles in plant disease resistance through their direct antifungal and antibacterial activities (Perkowska et al. 2021; Stringlis et al. 2018). Coumarins are also found displaying pharmacological activities that range from antimicrobial, antiviral, anticancer, antioxidant, and anti-inflammatory (Borges et al. 2005). Increasing evidence suggests that the production of defensive metabolites is closely associated with plant K level (Ramezani et al. 2018; Wang et al. 2013; Zhang et al. 2021), implying that defensive metabolites may be involved in high resistance in apple under high-K status.
Recent advances in metabolomics and analysis of large data sets based on multiomics procedures have facilitated the investigation of complex plant biochemical responses (Ranjan et al. 2019; Zhu et al. 2018). In this study, a combination of transcriptomics, metabolomics, and molecular functional tests was performed to reveal detailed mechanisms of enhanced resistance in M. domestica against C. mali under high-K status. It was found that many resistance compounds and events were higher under high-K status in apple tissues. Antimicrobial phytoanticipins, especially coumarin, were found to play essential roles in the resistance of apple to Cytospora canker under high-K status. Moreover, we found coumarin synthesis was regulated by the transcription factor MdMYB1r1 whose affinity to MdBGLU40 involved in coumarin synthesis was increased with K concentration. Overall, the high level of disease resistance in plants associated with high-K status principle highlights a plant disease management strategy.
Results
Potassium status of apple branches strongly affects Cytospora canker development
In fertilizer treatments on apple trees, we generated 5 K levels for apple branches based on their branch K content: 4.30 ± 0.45 g/kg (low-K, LK), 6.06 ± 0.31 g/kg (intermediate-K, IMK1), 7.11 ± 0.30 g/kg (IMK2), 8.10 ± 0.33 g/kg (IMK3), and 9.30 ± 1.01 g/kg (high-K, HK). Levels of other elements such as nitrogen (N) and phosphorus (P) were normal among these groups (Supplemental Table S1). To investigate the impact of low K on apple growth and development, comparisons were made between low- and high-K conditions, and there were no significant differences for root lengths, leaf lengths, leaf widths, leaf amounts, and plant heights (Supplemental Fig. S1, A to F). Fourteen days post inoculation (dpi), lesion sizes on branches gradually decreased with increased K content (Fig. 1A). Apple branches with HK were highly resistant (HR) with almost no canker development, while LK branches were highly susceptible (HS), possessing the largest lesions (Figs. 1B and S1G). Based on histological observation by wheat germ agglutinin (WGA) staining, we found abundant hyphae in the infected LK tissue, whereas only sparse hyphae were seen in HK tissue sections (Fig. 1C), and lignin content in LK was lower than that in HK (Fig. 1, D and E). At 7 dpi, ligno-suberized tissue was observed in HK tissues, but not in LK. At 14 dpi, although ligno-suberized tissue could be observed in LK, it was obviously thinner than that in HK (Fig. 1E). To investigate the resistance-related events, we further observed the callose deposition in HK and LK tissues following pathogen infection. It was found that the deposition of callose was largely induced in HK tissue (Fig. 1, F and G). These results demonstrated that C. mali expansion was significantly inhibited under HK conditions by chemical and physical barriers, suggesting that strong resistance might be induced in apple to Cytospora canker under HK status.
Apple branches exhibit high resistance to C. mali under HK status. A) Lesion sizes on apple branches inoculated with C. mali under different K status. Mock, inoculated with PDA plug; HS, highly susceptible; MS, moderately susceptible; MR, moderately resistant; HR, highly resistant; LK, low-K content (4.3 ± 0.5 g/kg); IMK, intermediate-K content (7.1 ± 0.9 g/kg); HK, high-K content (9.3 ± 1.0 g/kg). The individual images were digitally extracted for comparison. B) Data of lesion diameter on apple branches are presented as means ± Sd; n = 6. C) Hyphal density is reduced in apple tissue under HK status and observed by laser scanning confocal microscopy (LSCM). Hyphae of C. mali is excited green fluorescence, while healthy apple tissue is excited red fluorescence. Scale bars = 0.1 cm. D) Heavy lignin (red brown) accumulation in HK apple tissue. Scale bars = 50 mm. E) Formation of ligno-suberized tissue (blue) in HK (right) than in LK (left) apple tissue. Scale bars = 0.1 cm. IT, infected tissue; HT, health tissue. F) Callose deposits in LK and HK tissue following pathogen challenges. Scale bars = 50 μm. G) The histogram shows the quantification of callose deposits. Data of callose deposits are means ± Sd; n = 5. Statistical analysis is carried out with 1-way ANOVA followed by post hoc Tukey test in B) and Student's t tests in G) (**P < 0.01).
Weakened Cytospora canker development in HK apple tissue is host dependent
To determine whether the changes in Cytospora canker development under HK and LK conditions are dependent on the apple plants or C. mali, the growth of C. mali was observed at HK and LK conditions. Four days post inoculation on potato dextrose agar (PDA), we found that there were no significant differences in C. mali growth on PDA amended with 9.3 g/kg or 4.0 g/kg of K (Fig. 2A), suggesting that the changes of in planta K content did not directly affect the C. mali growth. Then, we analyzed the transcriptome profiles of C. mali when inoculated onto apple tissues under HK and LK status. The results showed that the genes that involved in growth, respiratory chain complex, sporulation, and conidiation of C. mali ranged from suppressed to downregulated in apple tissue under HK status compared with that under LK status (Fig. 2, B and C). Cytospora canker development might be inhibited in HK apple tissue. To test this possibility, we further investigate the expression of genes relevant to growth of C. mali in HK or LK tissue. The functions of genes encoding histone deacetylase 4 (Hst4), WD-40 catabolite repression protein (CreC), P-type ATPase, and cyclin-dependent kinase 6 (CDK6) have been documented to be involved in mycelial growth and development of pathogens (Lin et al. 2021; Li et al. 2016; Matar et al. 2017; Qu et al. 2020). PCR and gel electrophoresis showed that primers for these genes could successfully amplify DNA fragments in apple calli colonized by C. mali at 3 dpi, while no fragments were amplified from noninoculated apple calli (Supplemental Fig. S17). Further RT-qPCR assay showed the lower expression levels of Hst4, CreC, P-type ATPase, and CDK6 in inoculated HK tissue compared with that in LK (Fig. 2D). However, these genes were not suppressed or downregulated by K in C. mali when cultured on PDA amended with 9.3 g/kg or 4.0 g/kg of K (Fig. 2D). Overall, these results supported the host dependence of weakened Cytospora canker development under HK conditions and not a direct effect by the K on the pathogen.
Genes relevant to C. mali growth are suppressed in apple branches under HK status. A) Colony diameter of C. mali on amended PDA plates with high (9.3 g/kg) and low (4.0 g/kg) concentrations of element K at 4 dpi; PDA, potato dextrose agar; n = 3. B) Expression levels of genes relevant to C. mali growth when infecting HK and LK apple tissue. HK, high-K; LK, low-K. C) GO enrichment analysis on downregulated genes of C. mali in HK vs. LK. D) RT-qPCR displays the expression levels of genes relevant to C. mali growth in vivo or in vitro; n = 3. Data in A) and D) were presented as means ± Sd; statistical analysis is carried out with 1-way ANOVA followed by post hoc Tukey test in A) and Student's t tests in D) (**P < 0.01; ns, no significance).
Genes responsible for resistance and secondary metabolisms were highly expressed in apple with HK status
To investigate the biological changes in apple tissues following different K treatments and pathogen inoculation, global transcriptome analysis was conducted on LK, HK, inoculated LK (ILK), and inoculated HK (IHK) branches. Depending on the treatment, 13.9 to 18.0 million raw reads were generated per sample. On average, 3 quarters of the reads mapped to the apple reference genome (https://iris.angers.inra.fr/gddh13; Daccord et al. 2017). Consistent with the observation of low hyphal density in HK tissue, the proportion of C. mali reads in IHK (1.3%) was significantly lower than that in ILK (5%; Supplemental Table S2). Unsupervised hierarchical clustering (HC) and principal component analysis (PCA) suggested that the variations were generated mainly by differences in K, whereas the pathogen-induced changes were evident mainly in LK tissue (Figs. 3A and S2A). The numbers of differentially expressed genes (DEGs) were obtained during the comparisons as follows: 1,857 DEGs (1,187 up- and 670 downregulated genes) in ILK vs. LK, 513 DEGs (205 up- and 308 downregulated genes) in IHK vs. HK, 1,300 DEGs (619 up- and 681 downregulated genes) in HK vs. LK, and 863 DEGs (384 up- and 479 downregulated genes) in IHK vs. ILK (Supplemental Fig. S2B). The 694 upregulated genes from HK vs. LK and the 271 upregulated genes from IHK vs. ILK were combined as Set A (Fig. 3B). Further RT-qPCR results showed a strong correlation with transcriptomic profiles (in each of the 4 pairwise comparisons, R2 > 0.82; Supplemental Fig. S2C), which underlined the biological differences reflected by transcriptomic profiles.
Genes of basal resistance are highly triggered in apple branches under HK status. A) PCA plots separate the transcriptome samples into 4 groups. HK, high-K apple samples; LK, low-K apple samples; IHK, infected high-K apple samples; ILK, infected low-K apple samples. B) The UpSet diagram shows the overlapped genes shared in comparison with different combinations. Horizontal rows (red) are DEG counts in group comparison; DEGs, differentially expressed genes; vertical columns are counts of upregulated genes in intersection comparison. Set A is labeled in the frame. C) RT-qPCR shows differentially expressions of resistance-related genes in HK and LK apple tissue; n = 3. D) Enrichment analysis of upregulated genes in HK vs. LK comparison, showing secondary metabolisms are active in HK apple. E) Landscape of gene expression in secondary metabolisms in HK apple tissue compared with LK. F) RT-qPCR detects the expressions of MYB transcription factors involved in regulating phenylpropanoid biosynthesis in HK and LK apple tissue; n = 3. Data in C) and F) are presented as means ± Sd; statistical analysis in C) and F) is carried out with Student's t tests (**P < 0.01; ns, no significance).
CERK1 encoding transmembrane receptors perceives extracellular molecules including microbe-associated molecular patterns (Albert et al. 2020). Calcium-dependent protein kinases (CDPKs) perceive the rapid Ca2+ influx and trigger downstream defense responses (Zhou et al. 2020). A notable characteristic of the HK apple transcriptome was the prevalence of transcripts related to basal resistance genes, including MAPKs, NBSLRRs, transcription factors (TFs), receptor-like kinases and receptor-like proteins, and pathogenesis-related (PR) proteins (Supplemental Fig. S3A). We also tested the expression patterns of putative resistance-related genes under different K conditions. MdCERK1 and MdCPK5 were highly expressed in HK apple tissues (Fig. 3C), implying more effective pathogen recognition and activation of downstream resistance events in HK apple tissue. As downstream resistance events of various signaling pathways, some genes encoding pathogenesis-related proteins (PRs) including MdPR2, MdPR3, and MdPR4 were also obviously induced under HK conditions (Fig. 3C). Nicotiana benthamiana leaves with overexpression of MdPR2, MdPR3, or MdPR4 exhibited high levels of resistance even under LK conditions (Supplemental Fig. S3, B and C).
We further analyzed the enrichment significance of each plant metabolisms under HK and LK status. The results showed that some secondary metabolic pathways were more active in apple tissue under HK status compared with that in LK tissue (Fig. 3D). Individual genes involved in secondary metabolisms were then displayed in MapMan visualization (Fig. 3E). Among these secondary metabolic pathways, biosyntheses of terpenes, flavonoids, and especially phenylpropanoids were the most active (Fig. 3E), and the expression levels of most genes relevant to these pathways were higher in HK apple tissue compared with LK (Fig. 3E). In addition, we noted the higher expression levels of several TFs (MdMYB72 and MdMYB1r1) in HK (Fig. 3F). MYB TFs have been characterized as regulating phenylpropanoid biosynthesis in plants (Liu et al. 2015). Overall, our results revealed that genes relevant to plant defense responses and especially those involved in secondary metabolic pathways such as phenylpropanoid biosynthesis were more active in HK tissue.
Phenylpropanoid-related biosynthesis was reprogrammed in plants under HK status
To investigate the function of genes associated with changes in K content, various functional analyses were performed on Set A genes. Ontological function and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of upregulated genes (Set A) from HK vs. LK showed that biosynthesis of lignins and phenylpropanoids were the most active processes associated with increased K content (Fig. 4A and Supplemental Table S3). Four modules composed of a total of 1,105 genes and having high correlation with the HK trait were identified and termed as Set B using the weighted gene coexpression network analysis (WGCNA) to verify the results of pathway activity (Supplemental Fig. S4A). Pathway analysis of Set B genes also showed overrepresentation of phenylpropanoid biosynthesis in HK and reflected similar reprogrammed patterns of phenylpropanoid biosynthesis in HK (Supplemental Figs. S4B and S5A). Further Gene Set Enrichment Analysis (GSEA) of all 45,117 genes demonstrated the highest correlation between phenylpropanoid biosynthesis and K content (Supplemental Fig. S4C and Table S4). However, pathway analysis of the 314 hub genes associated with the LK trait did not show significant enrichment of phenylpropanoid biosynthesis (Supplemental Fig. S4D). In accordance with gene expression data in HK, we examined related genes and observed a significant upregulation of most genes in the phenylpropanoid pathway, including BGLU12/40, CAD, CHS/STS, CoMeF, COMT, CYP73A, PAL, and 4CL. We then constructed the synthesis flux pattern of the phenylpropanoid pathway in HK apple tissue (Figs. 4, A and B and S5B). In support, further proteomic analysis on upregulated proteins in HK vs. LK comparison also showed significant enrichment of the phenylpropanoid pathway (Supplemental Fig. S6A). Typically, phenylpropanoid biosynthesis was identified in both transcriptome and proteome profiles of the HK vs. LK comparison (Supplemental Fig. S6, B and C). The representative proteins encoded by genes that were upregulated in HK tissue included BGLU12/40, CAD, CHS/STS, CoMeF, COMT, CYP73A, PAL, and 4CL, leading to a similar active synthesis pattern of phenylpropanoid pathway in apple with HK status at the proteome level (Figs. 4D, S5A, and S6D). Related genes were significantly upregulated as early as 6 hpi in pathogen-challenged HK tissue, including BIS, COMT, PAL, FPS, BGLU42, and BGLU40, whereas most of these genes were induced later in LK tissue at 24 hpi (Fig. 4C and Supplemental Table S5).
Phenylpropanoid pathway is the most active biosynthesis in apple tissue under HK status. A) Scatter plots display the pathway enrichment of DEGs from Set A and show that phenylpropanoid biosynthesis is the most active pathway in HK apple tissue. DEGs, differentially expressed genes. B) Heat map visualizing expression levels of genes involved in phenylpropanoid biosynthesis in response to changes in K content and pathogen inoculation. Asterisks, FDR < 0.05; FDR, false discovery rate; HK, high-K apple samples; LK, low-K apple samples; ILK, infected low-K apple samples; IHK, infected high-K apple samples. C) Gene expression levels (P < 0.05) in phenylpropanoid biosynthesis at time points (6 h, 12 h, 24 h, and 48 h) in HK and LK apple branches upon pathogen attack; n = 3. D) Protein expression pattern based on proteome analysis. Proteins involved in the integrated pathway in Fig. 5E are labeled as black dots. E) K channel inhibitor TEA suppresses 2K calli resistance to C. mali; n = 6. F) RT-qPCR detects the expression levels (P < 0.05) of genes related with phenylpropanoid biosynthesis in calli under LK, 2K + TEA, and 2K treatments; n = 3. Data in E) and F) are presented as means ± Sd; statistical analysis in E) and F) is performed with 1-way ANOVA followed by post hoc Tukey test (**P < 0.01; ns, no significance).
To strictly control K levels, we used the potassium channel inhibitor, tetraethylammonium (TEA), on calli cultured on 2K moderately susceptible (MS) medium or 2K + TEA. TEA treatment effectively reduced the K content of calli on 2K medium (Supplemental Table S6), while TEA-induced K decreases also resulted in loss of resistance (Fig. 4E). Furthermore, the expression levels of representative genes in HK, such as COMT, PAL, BGLU42, BGLU47, and BGLU40, were significantly higher than that in K-deficient and TEA-treated calli (Fig. 4F). Significant similarities were observed in the expression levels of these genes in K-deficient and TEA-treated calli (Fig. 4F). Moreover, there were significant variations in metabolic compositions of apple calli from 2K treatment compared with apple calli from 2K + TEA treatment (Supplemental Table S7 and Fig. S7, A and B), and the metabolic profiles of apple calli from 2K or 2K + TEA treatments showed that phenylpropanoids accounted for the largest proportion of the K-associated metabolic variations (Supplemental Fig. S7C and Table S1). Several phenylpropanoids increased in 2K calli, and the changes in these metabolite contents could be mapped from transcriptome data to the expression changes of genes encoding their synthesizing enzymes (Supplemental Figs. S5B and S7D). These results supported the strengthened apple resistance through reprogramming the phenylpropanoid biosynthesis pathways under HK.
Coumarin dramatically inhibited C. mali at its physiological concentration in HK tissue
Constitutive phytoanticipins and induced phytoalexins derived from phenylpropanoid biosynthesis play important roles in plant basal resistance (Piasecka et al. 2015). Considering the possible antifungal roles of phenylpropanoids in plant resistance, we examined the activity of apple branch biochemicals against C. mali. Four days post inoculation on PDA plates with branch extracts, the results showed that colony diameters were significantly smaller (1.3 ± 0.2 cm) in the presence of the HK branch extracts than those of LK (7.4 ± 0.3 cm), suggesting that the extracts from HK apple tissue had higher toxicity to C. mali than those from lower-K tissues (Fig. 5A). Targeted metabolomics was then performed on LK, HK, ILK, and IHK 2-yr-old apple branches. PCA analysis of the metabolic data showed tight clustering of replicate samples, and samples were obviously segregated by K treatment, confirming reproducibility of the results (Fig. 5B and Supplemental Table S8). Further pathway enrichment analysis found that the main metabolic variations were induced by increases of K content (Supplemental Fig. S8A). A total of 39 constitutive phytoanticipins and 6 phytoalexins were highly increased under HK conditions (Supplemental Fig. S8B). Although the stress of C. mali infection also promoted increases in constitutive phenylpropanoids, the magnitude of change was significantly less than that caused by different K nutrient levels (Supplemental Fig. S8B). Next, by assessing the antifungal ability of 45 acquirable candidate metabolites in amended agar assays, we found 18 of them inhibited C. mali growth, including 12 phytoanticipins (coumarin/1,2-benzopyrone, ferulic acid, 4-coumaric acid, 7-methoxycoumarin, caffeic acid, glycitein, naringenin, phloretin, quercetin, gallic acid, catechol, and sinapic acid) and 6 phytoalexins (4-hydroxycoumarin, biochanin A, 7-hydroxycoumarin, chlorogenic acid, resveratrol, and capsaicin) (Fig. 5, B to D and Supplemental Table S9).
Screening of key antifungal metabolites based on their physiological concentration in apple tissue. A) Colony size of C. mali on PDA amended with total extracts from HK and LK branches at 4 dpi; data are presented as means ± Sd; n = 3 (**P < 0.01; ns, no significance based on 1-way ANOVA followed by post hoc Tukey test); HK, high-K apple samples; LK, low-K apple samples. B) PCA plots of the metabolic profiles from 4 treatments; ILK, infected low-K apple samples; IHK, infected high-K apple samples. C, D) The AIV coefficient of phytoalexins C) and phytoanticipins D). E) The integration of reprogrammed phenylpropanoid pathways in apple tissue under HK status. The effective and important antifungal metabolites in the pathway are shown in purple and blue box, respectively.
To assess the effective contributions of 18 antifungal metabolites to apple resistance, their physiological concentrations in low- and high-K branches were measured by LC–MS (Supplemental Table S10), and the activity in vivo (AIV) coefficients of each compound were evaluated (Fig. 5, C and D). Remarkably, the coumarin AIV coefficient was 0.97 for HK and increased to 1.24 under IHK, indicating that coumarin could fully inhibit C. mali growth in HK apple tissues (Fig. 5D). The antifungal ability analysis showed that coumarin almost completely inhibited mycelial growth at HK and IHK physiological concentrations (191.2 and 245.0 µg/g) at 4 dpi (Fig. 6, A and B and Supplemental Table S10). Compared with LK (0.25), the coumarin AIV coefficient was induced to 0.41 in ILK, but it only inhibited C. mali growth by about 26% in amended PDA plates with its concentration in ILK tissue at 4 dpi (82.0 µg/g; Fig. 6, A and B and Supplemental Table S10). Moreover, after exogenous application of coumarin (200 μg/g) to LK branches, the lesion expansion was significantly reduced (Fig. 6C). In parallel, we further exogenously applied 200 μg/g coumarin to HK apple branches. The results showed that lesion expansion was similar to that on HK apple branches (Supplemental Fig. S9A), implying that the coumarin accumulation in HK branches was high enough to inhibit pathogen expansion.
Blocking coumarin synthesis leads to a decrease in the resistance of HK apple calli to C. mali. A, B) Coumarin shows significant inhibition of C. mali growth at HK (high-K apple tissue) and IHK (infected high-K apple tissue) physiological concentrations at 4 dpi; n = 5; the individual images were digitally extracted for comparison; LK, physiological content of coumarin in low-K apple tissue; ILK, infected low-K apple tissue. C) Exogenous application of coumarin inhibits the extension of C. mali on low-K branches; n = 6; the individual images were digitally extracted for comparison. D) Overexpression of MdBGLU40 (OEMdBGLU40) increases coumarin content and strengthens resistance of low-K apple calli; interference with MdBGLU40 (RiMdBGLU40) suppresses resistance and reduces contents of coumarin in high-K calli to C. mali; n = 6. E) RT-qPCR displays the inhibitory effect of coumarin on the genes involved in C. mali hypha growth and development; n = 3. Data in B to E) are presented as means ± Sd; statistical analysis in B to D) is carried out with 1-way ANOVA followed by post hoc Tukey test and Student's t tests in E) (**P < 0.01; ns, no significance).
Apart from coumarin, 4-coumaric acid, caffeic acid, ferulic acid, and 7-methoxycoumarin, phytoalexins resveratrol and 4-hydroxycoumarin showed higher AIV coefficients under IHK conditions (Figs. 5, C and D, S9B, and Supplemental Table S10). When these compounds were mixed together at their IHK physiological concentrations and amended to agar, C. mali growth was inhibited by about 90% at 4 dpi. In contrast, when they were used at their ILK concentrations, C. mali growth was only inhibited by 23% at 4 dpi (Supplemental Fig. S9C). These results showed that coumarin was a potential key antifungal metabolite for inhibiting C. mali growth in HK apple tissue.
Enhanced coumarin accumulation in vivo leads to high level of resistance in apple to C. mali
Genes encoding β-glucosidase (BGLU) and glycosyltransferase (UGT) are responsible for coumarin synthesis in plants, while β-glucosyl-2-coumarinate catalysis into coumarin by BGLU is the hub step (Huang et al. 2021; Stringlis et al. 2018). Here, we identified 2 UGT genes (MdUGT88a1 and MdUGT2) and 4 BGLU genes (MdBGLU12, MdBGLU40, MdBGLU42, and MdBGLU47) in the apple genome, and KEGG annotation suggested that these genes may be involved in coumarin synthesis. Then, we analyzed the differences in expression levels of these 6 genes in apple under HK and LK conditions using RT-qPCR. The results showed that expression levels of MdUGT88a1, MdBGLU12, MdBGLU40, MdBGLU42, and MdBGLU47 in HK apple tissues were higher than that in LK apple tissues (Supplemental Fig. S10A). MdBGLU40 showed the highest level of upregulation in HK apple tissue compared with that in LK at a 13.1-fold increase (Supplemental Fig. S10A). Then, we analyzed the expression of these genes in different apple tissues. MdBGLU40 mainly expressed in apple branches and calli, followed by apple leaves and fruit (Supplemental Fig. S10B). Additionally, MdBGLU12 and MdBGLU47 mainly functioned in apple calli and branches, while MdBGLU42 and MdUGT88a1 were highly expressed in apple roots (Supplemental Fig. S10B). It was found that BGLU42 of Arabidopsis thaliana is important in synthesizing coumarin in roots to promote plant health (Stringlis et al. 2018). The results suggested that the expression of these coumarin-related genes was tissue specific (Supplemental Fig. S10B).
To further investigate the role of coumarin in apple resistance under HK conditions, we cloned the MdBGLU40 encoding the β-glucosidase 40 involved in coumarin synthesis to obtain its overexpression transgenic calli (OEMdBGLU40; Fig. 6D). RT-qPCR and western blotting showed overexpression of MdBGLU40 in representative calli (Supplemental Fig. S11, A and B). The transgenic OEMdBGLU40 calli was subcultured on different media to obtain LK and HK transgenic calli (Fig. 6D and Supplemental Table S11) and then inoculated with C. mali. The concentration of coumarin in LK transgenic calli (172.6 ± 10.5 μg/g) was then increased greatly to reach levels similar to those in HK-Mock tissue (healthy wild type; 198.3 ± 13.2 μg/g) and almost 3 times higher than that in the LK-Mock (60.4 ± 9.5 μg/g). At 4 dpi, the LK-Mock showed visible lesions, with large average lesions of 4.8 ± 0.2 cm. Remarkably, OEMdBGLU40 calli showed restriction of C. mali even under LK condition (Fig. 6D), suggesting that increasing the content of coumarin could restore the loss of C. mali resistance of LK apple tissues. Furthermore, we modified MdBGLU40 expression through RNA interference (RNAi). Its expression in positive RiMdBGLU40 calli lines was reduced by almost 68% compared with wild-type calli (Supplemental Fig. S11A). Interference with MdBGLU40 reduced coumarin content to 35.3 ± 8.9 μg/g in RiBGLU40 HK-calli which was significantly lower than that in HK-Mock calli. At 4 dpi, the lesions on RiMdBGLU40 HK-calli (2.0 ± 0.1 cm) were larger than those on HK-Mock calli (0.8 ± 0.2 cm), but smaller than those on LK-Mock calli (Fig. 6D). These results supported the important role of coumarin in high resistance of apple to C. mali under HK conditions.
To investigate the inhibitory mechanisms of coumarin on C. mali growth, we measured the expression of hyphal growth-related genes (Hst4, CreC, P-type ATPase, and CDK6) in inoculated OEMdBGLU40 LK-calli. The results showed that the expression of these genes was significantly downregulated in C. mali, demonstrating an inhibitory effect of coumarin on C. mali in vivo. We further evaluated the expression of genes involved in hyphal growth of C. mali after coumarin treatment. RT-qPCR results showed that related genes were suppressed by coumarin treatment (Fig. 6E), supporting the hypothesis that high resistance in HK apple was caused by the accumulation of coumarin which directly inhibited C. mali. Overall, high resistance in HK apple tissue is associated with high level of coumarin that directly inhibited C. mali growth.
MdMYB1r1 was responsible for initiating coumarin synthesis to confer apple resistance under HK conditions
Phenylpropanoid biosynthesis is modulated by numerous TFs, especially MYBs in plants (Liu et al. 2015). To uncover TFs that can manipulate coumarin synthesis, we constructed a coexpression network of genes relevant to coumarin synthesis using WGCNA based on transcriptome profiles with 40 apple branch samples (Fig. 7A). Seven modules were identified based on average hierarchical clustering and dynamic tree clipping, and the MEturquoise module had the highest association with these coumarin-related traits (Fig. 7B). In the MEturquoise module, 9 genes encoding MYBs with correlation values > 0.7 were identified and were selected as candidate TFs responsible for regulating coumarin synthesis (Fig. 7, A and B). The hub coexpression network showed that MdMYB1r1 was the hub gene in the regulating coumarin synthesis and highly correlated (P < 0.05) with the expression of MdBGLU40 (Fig. 7B). Subcellular location assay showed that MdMYB1r1 performed its function in the nucleus, whereas GFP alone (control) was located throughout the cytosol (Supplemental Fig. S12). The expression level of MdMYB1r1 in HK apple tissue was about 13.0-fold higher than that in LK (Fig. 3E). Meanwhile, we noted that the expression characteristic of MdMYB1r1 was similar with MdBGLU40 and that MdMYB1r1 was also highly expressed in apple branches, calli, and fruit (Supplemental Fig. S10B). Moreover, transient overexpression of MdMYB1r1 strengthened C. mali resistance and activated MdBGLU40 expression in apple tissue (Figs. 7, C and D and S11C). Further silencing of MdMYB1r1 via VIGS system showed that MdMYB1r1 silencing resulted in downregulation of MdBGLU40 and reduced resistance in apple to C. mali (Supplemental Figs. S11D and S13, A and B). Thus, these results suggested that MdBGLU40 was transcriptionally regulated by MdMYB1r1.
MdMYB1r1 is responsible for regulating coumarin synthesis in apple tissue under HK status. A) Heat map displays the module–trait relationships. Each row represents a module eigengene, and the column represents a trait. Each cell contains the corresponding correlation value between modules and traits. B) Coexpression network of genes relevant to coumarin synthesis and transcription factors in apple tissue under HK status. C) Transient overexpression of MdMYB1r1 strengthened resistance apple to C. mali; n = 7. D) Expression level of MdBGLU40 in apple tissue with transient overexpression of MdMYB1r1; n = 3. E) Luciferase reporter assay in tobacco showing that MdMYB1r1 binds with the promoter of MdBGLU40. F) The cis-elements in promoter region of MdBGLU40 gene predicted by the PlantCARE online tool. G) Y1H assay shows that MdMYB1r1 may bind to the MdBGLU40 promoter. H) MdMYB1r1 binds to promoter regions of MdBGLU40 in vivo. ChIP assay is performed with Anti-Myc antibody using the transgenic apple calli of OEMdMYB1r1. The qPCR shows that the MdMYB1r1-Myc is enriched at the region covered by ChIP-qPCR primers 1 (P1), P2, and P5 in the promoter region of MdBGLU40. Different letters show significant differences at P < 0.01; n = 3. I) EMSA is performed in the presence (+) or absence (−) of MdMYB1r1 and specific probe and/or unlabeled probes. And the cis-element CAACGG in P1 probe is replaced with AAAAAA and used as the negative probe. J) ChIP-qPCR assay detects the relative enrichment abundance of P1 by MdMYB1r1 in apple tissue under HK and LK status; n = 3. HK, high-K; LK, low-K. K) Luciferase activity assay shows the increases in binding activity of MdMYB1r1 with MdBGLU40 promoter in planta under HK conditions; n = 3. L) Transgenic apple calli with overexpression of MdMYB1r1 exhibited high level of resistance to C. mali even under LK status and interference of MdMYB1r1 reduced resistance in HK apple to C. mali; the individual images were digitally extracted for comparison; n = 4. Data in C), D), H), J), K), and L) are presented as means ± Sd; statistical analysis in C) and D) is performed with Student's t test and 1-way ANOVA followed by post hoc Tukey test in H), J), K), and L) (**P < 0.01; ns, no significance).
To investigate the binding activity of MdMYB1r1 on the MdBGLU40 promoter, the LUC gene was fused with the downstream of the MdBGLU40 promoter, and the open reading frame (ORF) of MdMYB1r1 was incorporated into the pGreenII 62-SK vector for luciferase assay. The transient expression assays demonstrated that pSK-MdMYB1r1 and MdBGLU40Pro-pLUC activated fluorescence by increasing the luminescence density upon coinjection, showing the binding of MdMYB1r1 with the promoter of MdBGLU40 in vivo (Fig. 7E). Meanwhile, 6 cis-elements of MYB binding sites were identified in the MdBGLU40 promoter using the online tool PlantCARE and termed as P1 to P6 (Fig. 7F and Supplemental Table S12), supporting regulation of MdBGLU40 expression by MYB TFs. Then, we performed yeast 1-hybrid (Y1H) assay to investigate the detailed binding sites of MdMYB1r1 on the MdBGLU40 promoter. The results showed that cotransformed yeast cells containing MdBGLU40P1/P2/P5-pHIS2 and MdMYB1r1-AD grew on SD-Trp/-His/-Leu medium, indicating that the MdBGLU40 promoter regions containing P1, P2, and P5 were potential binding sites of MdMYB1r1 (Fig. 7G). A further chromatin immunoprecipitation (ChIP)-qPCR assay was performed with the Anti-Myc antibody using the transgenic apple calli of MdMYB1r1-Myc plants. The results showed that MdMYB1r1-Myc was enriched at the region covered by ChIP-qPCR primers 2 (P2), 5 (P5), and especially 1 (P1) in the promoter region of MdBGLU40 (Fig. 7H). A subsequent electrophoretic mobility shift assay (EMSA) assay indicated that MdMYB1r1 activated the MdBGLU40 expression by binding on the P1 probe which was adjacent to the transcription start site in the entire promoter of MdBGLU40 (Fig. 7I). Meanwhile, we modified the CAACGG in P1 probe to AAAAAA as the negative probe, and the results showed that the negative probe did not affect the binding of MdMYB1r1 on the P1 probe (Fig. 7I), suggesting that MdMYB1r1 specifically bound the P1 site containing CAACGG element in the MdBGLU40 promoter. Moreover, an isothermal titration calorimetry (ITC) assay validated the direct binding of MdMYB1r1 in the region of MdBGLU40 promoter–containing P1 site in vitro, with a dissociation constant of Kd1 = 8.1 × 105 nM. The binding of MdMYB1r1 with the MdBGLU40 promoter was caused by entropy-driven adsorption processes with endothermic processes (ΔH > 0 and ΔS > 0) (Supplemental Fig. S14). The ChIP-qPCR assay showed that the enrichment of the P1 fragment by MdMYB1r1 in apple tissue under HK status was significantly higher than that under LK status (Fig. 7J), implying that the binding activity of MdMYB1r1 on P1 promoter was higher in apple tissue under HK compared with LK status. A subsequent luciferase activity assay supported this, that the binding activity of MdMYB1r1 on the MdBGLU40 promoter in HK tissue was significantly higher than that in LK tissue (Fig. 7K), suggesting that MdMYB1r1 could activate coumarin synthesis more effectively in HK tissue.
Subsequently, we obtained the transgenic apple calli with overexpression and silencing of MdMYB1r1 (Figs. 7L and S11, E to G). The results showed that MdMYB1r1 overexpression strengthened the resistance of apple to C. mali even with LK status (Figs. 7L and S11, E and G). Moreover, MdMYB1r1 overexpression effectively triggered the upregulation of MdBGLU40 in apple and increased the coumarin content to nearly that of HK-Mock tissue (Supplemental Fig. S13, C and D). Additionally, silencing of MdMYB1r1 through RNAi system reduced the resistance in HK apple to C. mali (Figs. 7L and S11F). Meanwhile, the stable silencing of MdMYB1r1 resulted in downregulation of MdBGLU40 and reduced coumarin content in HK apple nearly to that of LK apple tissue (Supplemental Fig. S13, C and D). Thus, these results suggested that MdMYB1r1 was a transcription factor of the coumarin synthesis gene MdBGLU40 and a positive regulator for apple resistance under HK status.
The disease resistance mediated by HK-induced antifungal phenylpropanoids is also conserved in N. benthamiana
To investigate the resistance of other plants under HK status, the resistance of N. benthamiana to Phytophthora parasitica was analyzed under low- and high-K conditions. N. benthamiana seedlings were grown under LK and HK conditions resulting in different potassium contents for LK (23.9 ± 1.3 g/kg) and for HK (40.8 ± 2.7 g/kg) after 30-d cultivation (Supplemental Table S13). At 3 d after inoculation, compared with LK N. benthamiana, P. parasitica–induced lesions on the HK N. benthamiana were effectively reduced by almost 40% (Fig. 8, A and G). HK and LK N. benthamiana leaves were also used for untargeted metabolomics (Supplemental Table S14). PCA analysis revealed that K exerted a strong effect on the metabolome of N. benthamiana, accounting for 67.4% of the observed metabolic variance (PC 1; Fig. 8B).
High level of coumarin confers resistance to P. parasitica in tobacco under high-K status. A) Lesion size increases on the tobacco leaves under different K status; n = 6; LK, low-K; HK, high-K. B) PCA plots of the metabolic profiles of HK and LK tobacco. C) Phenylpropanoid biosynthesis is more active in HK tobacco. D) Seventeen antifungal metabolites that inhibited C. mali growth are found to increase in HK tobacco. E) Coumarin shows significant inhibition of mycelium growth of P. parasitica at 200 μg/g. F) Exogenous application of coumarin to LK tobacco strengthens the resistance to P. parasitica; n = 6. G)NbBGLU40 silencing in N. benthamiana significantly reduces the resistance of HK tobacco to P. parasitica. And NbBGLU40 overexpression increases the resistance in LK tobacco; the individual images were digitally extracted for comparison; n = 5. Data in A), F), and G) are presented as means ± Sd; statistical analysis in A) and F) is performed with 1-way ANOVA followed by post hoc Tukey test and Student's t tests in G) (*P < 0.05 and **P < 0.01).
Pathway analysis of 212 dominant metabolites of HK showed that most upregulated metabolites in HK vs. LK comparison were overrepresented in phenylpropanoid biosynthesis (Fig. 8C), suggesting that phenylpropanoid biosynthesis was also closely associated with increases of K in N. benthamiana. Meanwhile, 17 defensive metabolites that highly increased in HK apple were also upregulated in HK N. benthamiana and inhibited the growth of the phytopathogenic oomycete P. parasitica (Figs. 8D and S15). Among these, coumarin also exhibited significant inhibition to the growth of P. parasitica (Figs. 8E and S15). Exogenous application of coumarin (100 μg/mL) to LK N. benthamiana sharply decreased the size of P. parasitica–caused lesions (Fig. 8F). To confirm the roles of defensive metabolites in conferring P. parasitica resistance in HK N. benthamiana, we used VIGS to silence NbBGLU40 which encodes the synthesizing enzymes β-glucosidase 40 of coumarin (Figs. 8G and S11H). The gene expression level was significantly reduced by almost 70% (Supplemental Fig. S11H). The P. parasitica–induced lesions on NbBGLU40-silenced N. benthamiana were noticeably larger than HK controls (Fig. 8G), suggesting that silencing of NbBGLU40 caused loss of resistance. On the other hand, transient overexpression of NbBGLU40 in LK N. benthamiana reduced the lesion sizes to about 73% of control (Figs. 8G and S11H), indicating that overexpression of NbBGLU40 caused resistance increase. These results supported the substantial involvement of coumarin in high level of N. benthamiana resistance under HK conditions.
Overall, these results demonstrated that the conservation of phenylpropanoid biosynthesis conferred high resistance in plants under HK status.
Discussion
Systematically uncovering the mechanisms in expression of plant resistance to pathogens under different K status is important for developing plant disease management strategies. Here, our study showed that an array of phytoanticipins and phytoalexins enhanced plant resistance under high-K status (within the optimal range for apple). In particular, the phytoanticipin coumarin was found as an important compound, making the major contributions to resistance in apple to Cytospora canker. Genetic analysis showed that genes for synthesizing coumarin augmented apple resistance even under low-K conditions, supporting the critical role of coumarin in strengthening resistance in apple (Fig. 9). We also found similar mechanisms in the N. benthamiana–P. parasitica pathosystem, supporting the evolutionary conservation of high resistance under HK conditions in diverse plants.
The mechanism model of high resistance in apple with HK status. In K-deficient conditions (left), apple branches are susceptible. In high-K (or optimal K) conditions (right), the defensive metabolites including 4-coumaric acid, ferulic acid, resveratrol, and especially coumarin involve in high resistance in apple tissue to C. mali. Typically, MdBGLU40 is modulated by MdMYB1r1 to produce coumarin in sufficient large quantities to make apple branch HR to C. mali under HK status; HK, high-K; LK, low-K.
Defensive secondary metabolites including phytoanticipins and phytoalexins important for high resistance of HK apple
The capacity of plants to deploy phytoanticipins to defend against pathogens is evolutionarily ancient (Piasecka et al. 2015). Various phytoanticipins have been documented in plant resistance, such as sulforaphane in A. thaliana (Wang et al. 2020a) and saponins in Avena spp. (Piasecka et al. 2015). Phytoanticipins are preformed antimicrobial compounds that directly inhibit pathogen invasion and expansion (Naoumkina et al. 2010). They also can be induced to increase following attempted microbial infection (VanEtten et al. 1944). Our results showed that the levels of 12 phytoanticipins which could inhibit C. mali growth were higher in HK apple tissues. Moreover, they also could be induced to increase following C. mali challenge. Although many antifungal compounds from a large variety of plants have been reported, their effects against diseases in vivo are often not clear or not investigated. Thus, we proposed the term AIV coefficient to express quantitatively the actual antimicrobial effect of metabolite in plant tissues. With a criterion of AIV coefficient > 0.25, 5 phytoanticipins (ferulic acid, 7-methoxycoumarin, 4-coumaric acid, caffeic acid, and coumarin) were identified as potentially effective inhibitory compounds. Among them, coumarin (AIV coefficient = 1.24) alone could fully inhibit C. mali growth both in apple tissues and in growth media at its HK physiological concentration, so it should be considered an important resistance chemical against Cytospora canker. Our results demonstrated that preformed phytoanticipins and their induced products promoted by HK were sufficient to defend against pathogen invasion in the apple–C. mali pathosystem (Fig. 9).
Although Cytospora canker resistance varies among apple species and varieties (Abe et al. 2007), no HR variety has been identified to date. BGLU42 of A. thaliana encoding β-glucosidase is important in producing coumarin (Stringlis et al. 2018). We identified that its homologous protein MdBGLU40 was highly upregulated in HK apple tissue. Its overexpression transgenic apple calli could produce higher levels of coumarin, and these calli exhibited strong resistance even under low-K status, suggesting that MdBGLU40 could be used as a key resistance gene for apple cultivar improvement to defend against Cytospora canker in the future.
In our research, 6 phytoalexins, 4-hydroxycoumarin, 7-hydroxycoumarin, biochanin A, capsaicin, chlorogenic acid, and resveratrol (Abe et al. 2007; Liu et al. 2010; Stringlis et al. 2018; Tewksbury et al. 2008; Tocci et al. 2018; Zhang et al. 2015), were identified as increasing and contributing to C. mali resistance. In particular, the resveratrol AIV coefficient was induced to 0.41 in high-K apple tissue, meaning it could inhibit C. mali growth by 40% in apple tissues. Phytoalexins are induced and delivered to the site of pathogen attack to exert their antifungal activity (Lu et al. 2015). Since the concentration of phytoalexins might be higher at the local sites of pathogen attack than that measured in the common tissue extraction in this study, they may play more important roles in high level of resistance in apple under HK status. The spatial distribution patterns and functions of resveratrol and other phytoalexins in apple tissues should be further studied, and this can be aided by spatial metabolomic techniques (Tocci et al. 2018).
Basal resistance–related responses might also confer high resistance of apple tissue under HK status
Plants maintain resistance to restrict pathogen invasion. The immune mechanisms include multiple layer responses, including pathogen recognition by pattern recognition receptors (RLK/RLP), MAPK cascade signaling, activation of transcription factors, and defensive metabolite synthesis (Jones and Dangl 2006; Lu et al. 2015). Increasing evidence has demonstrated the importance of defense in the ultimate outcome of plant resistance (Ahuja et al. 2012). Here, we found high levels of expression of many genes in apple tissue related to plant basal resistance under HK status, especially RLK/RLPs, MdCERK1, and MdFLS2 which were highly expressed in HK without pathogen challenge, indicating that pathogen recognition genes might be involved in the induction of high resistance in apple tissue under HK status. Efficient pathogen recognition can induce synthesis of defensive metabolites more rapidly in resistant host plant lines (Bjornson et al. 2021; Ranjan et al. 2019). Of note, the genes encoding synthesis of defensive metabolites were promptly activated in apple tissues under HK, showing the advantages of timely activation of defense responses. Additionally, abundant research has implied that adding K might enhance plant resistance through cell wall reinforcement (Wang et al. 2013) or activation of PR proteins (Ramezani et al. 2018). We also observed that some related downstream resistance events, such as reactive oxygen species (ROS) accumulation and callose deposition, were more active in HK, which implied that extending our understanding of the full scope of the mechanisms of high resistance in plants under HK status will require additional research.
Insufficient potassium causes loss of plant resistance
Plants require optimal quantities of macroelements and microelements for growth and development. Excessive levels of nutrients may lead to phytotoxicity symptoms, while nutrient shortages cause deficiency symptoms. As an essential macroelement, K exerts profound impacts on plant health. K deficiency in apple leads to dry and necrotic leaf margins. K-deficient plants are susceptible to a broad spectrum of pathogens, and remediating K deficiency content could enhance plant resistance (Wang et al. 2013). For vigorous tree growth and high yields of marketable apple fruit, optimal K content of foliage ranges from 1.3% to 1.8% (Marschner 2012). In this research, we found that when apple foliar K content was at 1.3%, or the lower limit of optimal foliar K, the branch K content was about 0.9%. This meant the amount of the “high K” in our disease system was in the optimal K range for apple. Optimal levels of K as a nutrient vary among different plants. For N. benthamiana, the foliar K content was 4.0% (the lower limit of tobacco optimal K) (Chouteau and Fauconnier 1988; Munson and Sims 1985). This implied that what is considered an optimal level of K nutrient for disease resistance may vary among plants and may even vary by pathogen.
Plant potassium status affects growth and resistance. Generally, at a K-deficient level, plants may exhibit physiological K deficiency symptoms and show greater susceptibility to pathogens. In this research, we found that apples showed no apparent stress or abnormal growth at the lower K level, but showed susceptibility to pathogen. We also found that when K was at an optimal level for plant growth, apple tissues exhibited high resistance to C. mali. This finding supports the view that insufficient potassium is associated with loss of plant disease resistance and adjusting K level from insufficient to optimal status could elevate plant resistance. It is essential to determine the thresholds for insufficient to optimal K levels affecting disease resistance since this is not clear for many crops, especially because lower K levels have no apparent effect on growth and yield in the absence of disease pressure.
Disease resistance of various plants might be higher under optimal potassium status
K is the most abundant cation in plants and performs essential roles in plant physiological processes and disease resistance (Feng et al. 2020). Application of K fertilizer has been reported to strengthen resistance of a wide range of plants in agricultural production (Wang et al. 2013). We demonstrated that various defensive secondary metabolites, such as coumarin, 4-hydroxycoumarin, 7-methoxycoumarin, chlorogenic acid, catechol, and resveratrol, improved the disease resistance in apple under optimal K conditions. Apart from effects on C. mali, coumarin, 4-hydroxycoumarin, and 7-methoxycoumarin also directly inhibited Botryosphaeria dothidea (Supplemental Fig. S16), the causal agent of apple Botryosphaeria canker. This implied that high resistance under optimal K status may work for Botryosphaeria canker on apple. Although chlorogenic acid did not show direct inhibitory effects on C. mali in our study, it has wide antifungal activity against other pathogens, such as Fusarium graminearum, the pathogen causing wheat head blight and corn stalk and ear rot (Gauthier et al. 2016). This implied that chlorogenic acid may promote disease resistance in other plants under optimal K status. Although defensive metabolites vary in different plant disease systems, their production is a common phenomenon. Overall, our study provides evidence that a high level of disease resistance under optimal K status works strongly in apple against C. mali, but may be common in other plants against a variety of diseases.
Environmental conditions profoundly affect plant–pathogen interactions, and understanding their influences is crucial in designing new eco-friendly and long-lasting strategies for integrated pest management. Our study revealed that the accumulation of phytoanticipins and phytoalexins promoted resistance in apple under high-K condition, especially coumarin in the Malus–Cytospora interaction (Fig. 9). Thus, we propose that an optimized K fertilizer regime could become an important modifying factor for new disease management strategies.
Materials and methods
Plant materials
Fertilization trials were conducted during the growing season (from April to September) in 2019 and repeated in 2020, in an experimental apple (M. domestica) orchard (“Fuji” trees on M.9 rootstock, 10 yr old in 2019) at Yangling, Shaanxi Province, China. Particles of 98% monopotassium phosphate (KH2PO4) were applied via groove fertilization at 30-d intervals from early May to June. LK: no addition of KH2PO4; 0.5, 1, and 1.5 kg KH2PO4 was used in IMK treatments including IMK1, IMK2, and IMK3, respectively; HK: 2 kg KH2PO4 consisted of 2 applications. For HK treatments, when foliage reached the optimal K level (1.3% to 1.7%) (Marschner 2012), branches were inoculated. There were 3 sets of 5 trees for each treatment. The treatments were applied in a randomized complete block design. Tissue-cultured calli of the apple cultivar ‘Orin’ were subcultured as described (He et al. 2018). Then, we used potassium channel inhibitor TEA on calli to investigate the effect of potassium as described (Schachtman and Schroeder 1994). The wild-type calli of ‘Orin’ apple were subcultured on 2K, K-deficient, and 2K + TEA (800 μmol/L) MS media for 15 d, after which tissue levels of N, P, and K were evaluated. Each treatment contained at least 6 replicates.
N. benthamiana seedlings were transplanted at the 3-leaf stage as described (Krügel et al. 2002). Then, seedlings were cultured on modified Hoagland's nutrient solution containing 2-fold (0.758 g/L KNO3) or 0 K content (0 g/L KNO3) to obtain high-K (4% to 6%) (Chouteau and Fauconnier, 1988) and low-K (<2%) tobacco plants. To ensure the same nitrogen level in both 0K and 2K Hoagland's solutions, 0.3 g/L NH4NO3 was added to the 0 K nutrient solution. After 3 weeks, the N, P, and K contents were measured, and the N. benthamiana leaves with high-K and low-K content were collected for metabolomics. Additionally, 1-wk-old N. benthamiana seedlings were transplanted into plastic pots with sterile mixtures of vermiculite and substrate at 1:1 (v/v), and then, modified 2K Hoagland's nutrient solution supplemented with 800 μmol/L TEA was added as described (Schachtman and Schroeder 1994). Seedlings subjected to 2K nutrient solution were used as controls. Each treatment contained at least 3 replicates, and the experiment was repeated 3 times.
Pathogen inoculation and Cytospora canker severity
A strain of C. mali (obtained originally from apple and stored at Northwest A&F University) was cultivated at 25°C on PDA. A strain of P. parasitica var. nicotianae provided by Prof. Xili Liu (Northwest A&F University) was cultivated at 25°C on V8 agar. Wound inoculations were done on attached 2-yr-old apple branches as described (Peng et al. 2016). The extension of hyphae in branch tissues was observed by laser scanning confocal microscopy as described (Sun et al. 2014). WGA-AF488 excitation was at 488 nm and detection at 500 to 540 nm; PI excitation was at 561 nm and detection at 600 to 700 nm. The detailed setup for WGA was as follows: emission wavelengths, 510 nm; detection wavelengths, 510 to 550 nm; laser wavelength, 488 nm; laser transmissivity, 1.08%; AOTF/AOM transmissivity, 10.8%; laser ND filter, 10%; and PMT voltage, 604 V; and for PI detection: emission wavelengths, 617 nm; detection wavelengths, 570 to 610 nm; laser wavelength, 561 nm; laser transmissivity, 0.88%; AOTF/AOM transmissivity, 8.8%; laser ND filter, 10%; and PMT voltage, 628 V. Nonwounded detached leaves from N. benthamiana seedlings were inoculated with 5-mm-diameter mycelial plugs of P. parasitica. At 3 dpi, inoculum plugs were removed, and representative leaves were photographed for assessment. Calli were inoculated by placing mycelial plugs of C. mali on their surfaces. Each inoculation experiment contained at least 3 replicates, and the experiment was repeated 3 times.
The resistance ratings were used as follows: HR, lesion length < 0.5 cm; moderately resistant (MR), 1.8 cm > lesion length > 0.6 cm; MS, 3.1 cm > lesion length > 1.9 cm; and HS, lesion length > 3.2 cm.
Extraction of total RNA and RT-qPCR analysis
Total RNA of apple branches, N. benthamiana leaves, and C. mali were extracted using the Omega Plant RNA Kit (Omega Bio-Tek), and then, total RNA was reverse-transcribed to cDNA using an EasyScript cDNA Synthesis Kit (Transgene, China). The detection of gene expression levels was performed in triplicate using gene-specific primers with SYBR Green staining. The expression of Actin was used as an internal standard for RNA samples.
The species specificity of primers used for analyzing the gene expression of C. mali in apple calli was determined by PCR and gel electrophoresis (Supplemental Fig. S17). The RNA of infected apple branches and calli with high- and low-K status were extracted and reverse-transcribed to cDNA. The pathogen gene expression was determined in triplicate using these gene-specific primers by RT-qPCR. The expression of Actin was used as an internal standard for RNA samples. The data of RT-qPCR were analyzed through the 2−△△CT method.
RNA-seq data processing and analysis
The phloem tissues were peeled off from 2-yr-old branches with either low- or high-K content that had been subjected to inoculation and control treatments and bulk processed for transcriptome analysis. In total, 3 apple trees were involved in each treatment, and 5 individual branches from each tree were cut off for phloem collection. RNA sequencing was performed by Peking Biomarker Technologies on an Illumina HiSeq 4000 platform with paired-end 150 bp reads. Raw reads in FASTQ format were processed to obtain clean reads by removing adapter sequences, poly-N reads, and low-quality reads. And then, all clean reads were mapped onto the apple reference genome available at https://iris.angers.inra.fr/gddh13 (GDDH13 version 1.1. genome) and the C. mali genome using Tophat (version 2.1.0). Assembly and quantification of the transcripts were accomplished with Cufflinks software (version 1.3.1) using the apple genome annotation GFF file as the reference (available at the Apple Genome and Epigenome website). Fragments Per Kilobase of transcript per Million mapped reads (FPKM) was used to represent the relative abundances of the transcripts and generate differential expression values [|log2 fold change| > 1; false discovery rate (FDR) < 0.05]. RNA sequencing data have been submitted to the National Genomics Data Center (NGDC) under the accession number CRA009344.
All apple genes were assigned to BIN annotation codes (BINs) using the Mercator automated annotation pipeline (http://mapman.gabipd.org/web/guest/mercator), and the functions of differentially expressed genes (DEGs) were further visualized by MapMan. KEGG pathways and Gene Ontology (GO) enrichment analyses were implemented by the R package clusterProfiler based on the Benjamin–Hochberg correction. The R package WGCNA was used to identify modules of highly correlated genes based on FPKM data.
Inhibition of C. mali by apple branch tissue extracts
Compounds from HK and LK apple branches were extracted following Ranjan et al. (2019). The tissues were ground by mortar and pestle into powder using liquid nitrogen. Approximately 100 g of uninfected HK and LK apple branches was mixed 1:3 w/v in 80% ethanol [ethanol/water (v:v) = 4:1] and incubated at 80°C for 3 h, and this was done 3 times. The liquid above the solid residue was removed each time and combined and freeze-dried by condensation under vacuum. All residues were resuspended in 1.5 mL of ethanol and added to 40 g PDA media for the C. mali inhibition assay. PDA with an equivalent amount of ethanol served as control. Each experiment contained 3 replicates and was repeated 3 times.
Sample preparation and extraction for metabolomic analysis
Untargeted metabolomics was conducted on high- and low-K N. benthamiana and apple calli. Samples of 100 mg were minced and transferred into 10-mL tubes, and 6 mL precold extraction mixture [methanol/water (v:v) = 3:1] was added per tube for metabolite extraction as previously described (Zhang et al. 2015). The supernatant was evaporated in a vacuum concentrator, and then 1 mL methanol was added to redissolve the extracts. Six N. benthamiana plants or apple calli were used for each treatment.
Targeted metabolomic analysis was conducted on 2-yr-old apple branches with low-/high-K content and inoculation/noninoculation treatments. Phloem tissues of 5 2-yr-old branches from 1 tree were peeled off and mixed as 1 sample. In total, 6 apple trees were involved in each treatment. The tissues were ground by mortar and pestle into powder using liquid nitrogen, and the tissue powder was extracted following as described (Zhang et al. 2015). The supernatant of each sample from the 3-fold extractions was combined and freeze-dried by condensation under vacuum. The residues were redissolved with 10 mL methanol and passed through a 0.45-μm filter before metabolite detection based on LC–MS.
LC–MS/MS analysis for metabolomics
The UHPLC separation was carried out using a Nexera UHPLC LC-30A system, equipped with an ACQUITY UPLC HSS T3 1.8 μm column (2.1 × 100 mm, Waters). For untargeted metabolomics, the mobile phase used in positive mode consisted of pure water (A) and acetonitrile (B), whereas the negative mode consisted of 0.1% formic acid [formic acid/water (v:v) = 1:1000] (A) and 0.1% formic acid in methanol (B). The analyses were carried with an elution gradient as follows: 0 to 2 min, 30% to 45% B; 2 to 4.5 min, 45% to 95% B; 4.5 to 6.0 min, 95% B; 6.0 to 10.0 min, 100% B; and 10 to 13 min, 100% to 30% B. For targeted metabolomics, the mobile phases for both negative and positive modes were the same as above. The negative analyses were carried out with elution gradient as follows: 0 to 15 min, 15% to 25% B; 15 to 25 min, 25% B; 25 to 70 min, 25% to 55% B; 70 to 75 min, 55% to 70% B; 75 to 85 min, 70% to 100% B; 85 to 95 min, 100% B; and 95 to 97 min, 100% to 15% B. The column temperature was 40°C, the flow rate was 0.3 mL/min, and the injection volume was 10 μL in both negative and positive modes.
A TripleTOF 5600+ Mass Spectrometer (AB Sciex) was used to acquire MS/MS spectra on an information-dependent basis (IDA) during the LC/MS experiment. The acquisition software (Analyst TF 1.7.1, AB Sciex) continuously evaluates the full scan survey MS data as it collects and triggers the acquisition of MS/MS spectra. In each data collection cycle, the molecular ions with intensity above 1 cps were chosen for MS/MS at collision energy (CE) of 35 ± 15 eV. Electrospray ionization (ESI) source conditions were as follows: gas 1 at 50 psi, gas 2 at 50 psi, curtain gas at 35 psi, source temperature at 550°C, and ion spray voltage floating (ISVF) at 5,500 V (positive) or −4,500 V (negative).
The MS raw data (.wiff) files were converted to the mzXML format using ProteoWizard and processed using the R package XCMS (version 3.2). The process included peak deconvolution, alignment, and integration. Minfrac and cutoff were set as 0.6 and 0.3, respectively. The minimum number of samples necessary in 1 group was set as 3 for its validation. An in-house database and the online (https://metlin.scripps.edu) MS2 database were used for metabolite identification by software PeakView and XCMS. Normalized peak areas of all metabolites were used for further analysis in R package MetaboAnalystR (www.r-project.org/). All metabolites were annotated against the KEGG database for enrichment analysis based on hypergeometric distribution.
Determination of physiological concentration of candidate metabolites and their antifungal activity
The inhibitory activity of candidate metabolites to C. mali and P. parasitica was tested by commercial pure compounds on PDA media. Candidate compounds were obtained from commercial sources, dissolved in ethanol, and added to PDA at final concentrations of 0, 10, 50, 100, 150, and 200 µg/g to calculate their concentrations for 50% (EC50) and 95% (EC95) of inhibitory effects. The activity of effective compounds was further tested at their physiological concentrations. PDA with an equivalent ethanol concentration served as control. Each experiment contained 3 replicates and was repeated 3 times.
The concentration curves of target metabolites were constructed using standard compounds from Sigma and Shanghai Yuanye Bio-Technology to calculate the physiological concentrations. To express the actual effect of a metabolite in plant tissues, the AIV coefficient was calculated as AIV coefficient = PC/EC95 (PC, physiological concentration of a metabolite in vivo; EC95, the effective concentration to cause 95% inhibition of pathogen growth in vitro).
Exogenous application of antifungal metabolites on apple branches and N. benthamiana leaves
Two-year-old low-K apple branches were treated by exogenous coumarin. Both ends of excised branches were wrapped in wet degreased cotton for moisture retention. Coumarin was dissolved in DMSO and further diluted with distilled water to low (50 µg/mL) and high concentrations (200 µg/mL). Branches were submerged in each coumarin solution overnight and then wound inoculated with C. mali. Ten days post inoculation (dpi), the lesion sizes were recorded. Coumarin was exogenously applied to low-K N. benthamiana leaves as described (Wang et al. 2020b). Twenty-four hours after application, P. parasitica was inoculated on the representative leaves. Each treatment was replicated 3 times, and the experiment was repeated 3 times.
Vector construction and genetic transformation in N. benthamiana
To construct TRV vectors, 350-bp DNA fragments of NbBGLU40 (Niben101Scf11389g01020) of N. benthamiana were cloned using a primer pair containing EcoRI and BamHI sites (Supplemental Table S15) and then ligated to pTRV2. The correct recombinant vectors were transformed into Agrobacterium GV3101 and then injected into LK and HK N. benthamiana as described (Senthil-Kumar and Mysore 2014). The empty vector–containing Agrobacterium solution was used as control. At 14 dpi, transcript levels of NbBGLU40 in plants were assessed using RT-qPCR. The silenced N. benthamiana were inoculated with P. parasitica. Additionally, full-length cDNA of NbBGLU40 was amplified using a primer pair containing BamHI sites and then cloned into the pBIN-eGFP overexpression vector as described (An et al. 2017). The overexpression of NbBGLU40 in N. benthamiana was assessed after agro-infiltration as described (Senthil-Kumar and Mysore 2014), using RT-qPCR. There were 3 replicates per treatment, and the experiment was repeated 3 times.
Overexpression and suppression of MdBGLU40 and MdMYB1r1 in apple calli
To obtain MdBGLU40 and MdMYB1r1 (accession numbers: MD16G1062800 and MD14G1189300) overexpression calli, the recombinant 35S:BGLU40 or 35S:MYB1r1 plasmid was introduced into Agrobacterium strain LBA4404 and transformed into ‘Orin’ apple calli as described (Hajdukiewicz et al. 1994). Positive transgenic calli were screened based on kanamycin resistance and further identified by RT-qPCR. Transgenic and WT calli were subcultured on 2K and K-deficient MS media. Fifteen days after culturing, the calli on plates were inoculated with C. mali. Subsequently, 450-bp fragments of MdBGLU40 and MdMYB1r1 were transferred into the pFGC5941 vector to generate MdBGLU40- or MdMYB1r1-silenced calli. A. tumefaciens LBA4404 carrying the MdBGLU40-RNAi or and MdMYB1r1-RNAi vector was transformed into apple calli as described (He et al. 2018). The calli were then transferred to medium containing kanamycin and carbenicillin to screen for transgenic calli.
Firefly luciferase reporter and luciferase transient transcriptional activity assays
The promoter regions at the region 1 (P1) of MdBGLU40 were amplified and cloned into the pGreenII 0800-LUC (pLUC) vector to obtain MdBGLU40P1:LUC reporter constructs. The ORF of MdMYB1r1 was amplified and cloned into the pGreenII 62-SK (pSK) vector to obtain the 35S:MdMYB1r1 construct, which was verified by restriction endonuclease digestion and sequencing. Then, both positive vectors were introduced into A. tumefaciens strain GV3101. At 48 h post infiltration, the transformed leaves were sprayed with 0.1 M luciferin and incubated in the dark for 5 min before luminescence examination. A low-light cooled CCD imaging apparatus (CHEMIPROHT 1300B/LND, 16 bits; Roper Scientific) was used to capture the LUC images. The images were processed by ImageJ software.
For the luciferase activity assay, both A. tumefaciens strains harboring MdBGLU40P1:LUC and 35S:MdMYB1r1 vectors were transformed into N. benthamiana leaves with LK and HK contents. Then, the luciferase activity assay was performed as described (Zhao et al. 2020).
Yeast 1-hybrid assays
The yeast 1-hybrid assay was performed following as described (Liu et al. 2021). The promoter region of MdBGLU40 was amplified from the apple genomic DNA and fused with the pHIS2 vector. Meanwhile, the MdMYB1r1 cDNA sequence was inserted into the pGADT7 vector. The resultant plasmids were cotransformed into Y1H yeast and cultured on SD/-Leu with 150 ng/mL AbA at 30°C for 3 to 7 d.
Protein purification and ITC
The coding sequence of apple MdMYB1r1 was cloned into the GST-tagged PGEX-4T vector for protein production (Transheep Biotechnology, Shanghai, China). The recombinant 4T vector with MdMYB1r1 was converted into BL21 (DE3) to express MdMYB1r1 protein. Positive clones were tested for in-frame insertion using PCR and sequenced. After the bacterial culture had grown to an optical density (OD600) of around 0.6 at 600 nm, isopropyl-D-thiogalactopyranoside (IPTG) was added to promote the production of the MYB1r1-GST protein for 4 h at 37°C. The recombinant MYB1r1 protein was purified according to the manufacturer's procedure using a glutathione S-transferase (GST) tag protein purification kit (Beyotime Biotechnology, Shanghai, China). Then, enterokinase digestion cleaved the upstream 43-residue N-terminal fusion sequence from purified fusion proteins. The proteins that flowed through were collected and dialyzed against TBS glycerol buffer. The BCA technique was used to determine the concentration of pure MYB1r1 protein.
On an iTC200, ITC tests were performed (Nano ITC). As a ligand, the MdMYB1r1 protein solution (8 mM in PBS, 50 L) was titrated into the MdBGLU40 promoter (0.06 mM) in PBS with a volume of 2.5 mL for each titration, and PBS was used as a control.
ChIP-qPCR
The ChIP assay was performed using the EpiQuik Plant ChIP Kit (Epigentek). Briefly, chromatin was isolated from 2 to 4 g leaves of MdMYC1r1:Myc plants and then fragmented to 0.2 to 1.0 kb by sonication. The DNA/protein complex was immune-precipitated with ChIP-grade antibody against Myc (TransGen Biotech). After reverse crosslinking and protease K treatment, the immune-precipitated DNA was purified. Then, the immune-precipitated and input DNA samples were as template for quantitative PCR using the region-specific primers from MdBGLU40 (Supplemental Table S15). The enrichment fold of each fragment was calculated based on normalizing against the promoter of β-Actin (as the internal control) and then in comparison to wild-type plants. Additionally, the leaves of MdMYC1r1:Myc plants under HK and LK status were also collected to perform the ChIP-qPCR assay to further investigate the binding activity of MdMYB1r1 on the regions containing P1 cis-element of the MdBGLU40 promoter under LK and HK conditions. The ChIP-qPCR was performed with triple biological replicates.
Electrophoretic mobility shift assay
For EMSA, cis-elements of MdBGLU40 promoter, which may be bound by MdMYB1r1, were first analyzed by using the PlantCARE database to design EMSA probes (as listed in Supplemental Table S15). Probes P1, P2, and P5 containing cis-elements in MdBGLU40 promoter were commercially synthesized by Shanghai Sangon Biotech and then labeled with biotin using EMSA Probe Biotin Labelling Kit (Beyotime). Nonlabeled DNA oligonucleotides were used as competitors, while the CAACGG in P1 site was modified to AAAAAA as the negative probe, as detailed in Supplemental Table S15. EMSAs were performed using a LightShift Chemiluminescent EMSA Kit (Thermo) according to the manufacturer's instructions. The experiment was independently repeated 3 times.
Statistical analysis
All data were analyzed using 1-way ANOVA at P = 0.05 and Student's t tests (P = 0.05) with SPSS 26.0. Data are shown as mean ± standard deviation (Sd). And biological replicates are shown in green spots in each histogram. Statistical significance of the observed differences was marked in the figures with asterisks according to following categories: **P < 0.01; *P < 0.05; and ns, not significant.
Accession numbers
Sequence data from this article can be found in the database for apple reference genome (https://iris.angers.inra.fr/gddh13), database for N. benthamiana genome (https://solgenomics.net/), and GenBank/EMBL data libraries under accession numbers MdMYB1r1, MD14G1189300; MdBGLU40, MD16G1062800; MdBGLU12, MD11G1023900; MdBGLU47, MD10G1316200; MdBGLU42, MD06G1229100; MdUGT88a1, MD15G1407600; MdPR2, MD06G1120700; MdPR3, MD02G1011100; MdPR4, MD04G1225400; NbBGLU40, Niben101Scf11389g01020; CmHst4, KUI66397.1; CmCreC, KUI70025.1; CmP-type ATPase, KUI73682.1; and CmCDK6, KUI66996.1.
Acknowledgments
We are very grateful to Prof. Yunzhong Guo, Dr. Guangli Liu, Dr. Pengbo Dai, Xueyan Ma, Yangbo Duan, and Yiming Lu (College of Plant Protection, Northwest A&F University) for taking part in field experiments. Prof. Jun Fan and Junfeng Liu (China Agricultural University) provided many constructive suggestions on the trial design and in manuscript organization. The P. parasitica var. nicotianae strain was provided by Prof. Xili Liu (Northwest A&F University). We thank Life Science Research Core Services, NWAFU, for technical support in the LC–TOFMS/MS assay. We thank Luqi Li for LC–TOFMS data acquisition and sample preparation.
Author contributions
G.S. and R.Z. designed the study and revised the manuscript; Y.D. performed the research, analyzed data, and wrote the paper; H.J., Z.Y., and S.W. constructed the transgenic plants; H.M. and Y.L. performed the RT-qPCR; M.Z. performed the filed investigation; B.W. analyzed the transcriptome profiles; M.G. and T.H. refined the ideas and revised the manuscript; S.N., Y.M., and X.L. discussed the results.
Supplemental data
The following materials are available in the online version of this article.
Supplemental Figure S1. Effects of K on the resistance and growth of apple seedlings.
Supplemental Figure S2. Overview of transcriptome data of apple tissue with high- and low-K content.
Supplemental Figure S3. Expression patterns of genes relevant to basal resistance in HK and LK apple tissue.
Supplemental Figure S4. Alterations in transcriptome of apple tissue under HK status are mainly upregulations of genes relevant to phenylpropanoid biosynthesis.
Supplemental Figure S5. Transcription pattern of phenylpropanoid pathway in HK apple tissue.
Supplemental Figure S6. Expression levels of proteins involved in phenylpropanoid biosynthesis are upregulated in HK apple branch tissues.
Supplemental Figure S7. Overview of the metabolic variations between 2K- and 2K + TEA-treated apple callus.
Supplemental Figure S8. Metabolomic analysis of apple tissues with low-K and high-K content.
Supplemental Figure S9. Growth inhibition of C. mali by individual and mixed metabolites.
Supplemental Figure S10. Relative expression levels of genes involved in coumarin synthesis in apple tissue under HK and LK status.
Supplemental Figure S11. Verification of transgenic plants.
Supplemental Figure S12. Subcellular location of MdMYB1r1 in vivo.
Supplemental Figure S13. Effects of MdMYB1r1 on MdBGLU40 expression and resistance in apple tissue.
Supplemental Figure S14. ITC enthalpograms of MdMYB1r1 protein binding to the MdBGLU40 promoter.
Supplemental Figure S15. Toxicity of upregulated metabolites on the mycelium growth of P. parasitica.
Supplemental Figure S16. Growth inhibition to Cytospora leucosperma and B. dothidea of 4-hydroxycoumarin, 7-methoxycoumarin, and coumarin.
Supplemental Figure S17. PCR analysis of the species specificity of primers for determining gene expression levels in C. mali–infected apple tissue.
Supplemental Table S1. The contents of N, P, and K nutrient elements in the apple branches.
Supplemental Table S2. Sequencing statistics of RNA-seq libraries.
Supplemental Table S3. Significant phenylpropanoid-related Go terms in high-K apple tissue (HK vs. LK).
Supplemental Table S4. GSEA analysis of KEGG biosynthesis pathways using transcripts from HK vs. LK.
Supplemental Table S5. Primers and origin time course expression matrix of genes involved in phenylpropanoid and terpenoid biosynthesis.
Supplemental Table S6. The contents of N, P, and K nutrient elements in apple calli for metabolomics.
Supplemental Table S7. Origin data matrix of untargeted metabolic profiles of apple calli from 2K and 2K + TEA treatments.
Supplemental Table S8. Original data of targeted metabolomics of apple branch tissues from LK, ILK, HK, and IHK treatments.
Supplemental Table S9. The standard curve, EC50, and EC95 of all the metabolites positively correlated with the increasing K content.
Supplemental Table S10. Physiological concentration and AIV coefficient of antifungal metabolites in apple tissue under each condition.
Supplemental Table S11. The contents of N, P, and K nutrient elements in transgenic apple calli.
Supplemental Table S12. PlantCARE results of cis-elements of MYB binding sites on the MdBGLU40 promoter.
Supplemental Table S13. The contents of N, P, and K nutrient elements in the N. benthamiana.
Supplemental Table S14. Origin data matrix of untargeted metabolic profiles of N. benthamiana under HK and LK status.
Supplemental Table S15. Information of primers and probes used in EMSA, ChIP, RT-qPCR, and genetic transformation.
Funding
This research was funded by the National Key R&D Program of China (2019YFD1002002) and China Agriculture Research System (CARS27).
References
Author notes
The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plphys/pages/General-Instructions) is: Guangyu Sun ([email protected]).
Conflict of interest statement. The authors have no conflict of interest in this study.
