Leaf dynamics in response to P. aphanis
Changes in fluorescence parameters and morphology were determined in a greenhouse environment following inoculation with P. aphanis. In October, natural day light was only 8–9 h/day. Temperatures were approximately 18–28 °C during the day and 7–14 °C at night, while the relative humidity were approximately 39–58 °C during the day and 91–95 °C at night throughout the experiment (Figure S1). According to Dodgson et al., (2007) 97–100% relative humidity is optimum for germination of condia, while optimum temperature for germination is 15–25 °C and optimum temperature for sporulation is 20 °C [27]. Therefore, the temperature and humidity in Greenhouse are suitable for P. aphianis germination.
To investigate leaf resistance in response to P. aphanis, the conidial germination percentage of P. aphanis was recorded over the entire experimental period (Fig. 1). Visual symptoms (white mycelium) began appearing on leaves at 3 dpi in both groups, and obvious differences were observed between the two groups (there was a larger disease area in ddH2O treatment than SA treatment). Whereas extensive colonisation occurred along the leaf in the ddH2O group at 7 dpi, only small, restricted colonisation was observed in the SA group (Fig. 2b). The developmental stages of P. aphanis were also observed using a microscope (Fig. 2c).
To further ascertain the optimum sampling time, chlorophyll fluorescence parameters in both groups infected with P. aphanis were detected (Fig. 2d). Fv/Fm is the state of the initial part of photosynthesis and reflects the degree of plants under stress. ΦPSII is the actual PSII efficiency and reflects the current actual light energy conversion efficiency of the photosynthetic mechanism. ΦNPQ is the energy dissipated as heat through a regulated photoprotection mechanism, if the value is higher, indicating plant has a higher photoprotection ability. ΦNO is the passive dissipation of heat and fluorescent energy, suggesting the self-protection ability under excessive light is lost. Both groups showed decrease in the Fv/Fm ratio at 1 dpi and an increase at 3 dpi, followed by a continuous decline in Fv/Fm values from 3 to 7 dpi. The Fv/Fm ratios in the SA group were higher than those in the ddH2O group from 1 to 7 dpi, and coincided with visual symptoms at the corresponding time points (Fig. 2d). ΦPSII showed a continuous decline over the course of the experiment; ΦPSII and ΦNPQ were higher in the SA group than ddH2O group from 3 to 7 dpi. In contrast to Fv/Fm, ΦPSII, and ΦNPQ, ΦNO in the ddH2O group was higher than that in the SA group from 5 to 7 dpi. Lower Fv/Fm values usually corresponded to higher ΦNPQ values, indicating dissipative processes that lowered the PS photochemical efficiency. Fv/Fm values was lower in ddH2OInfected than in SAInfected, suggesting that there was damage to photosystem II caused by P. aphains at 3 dpi. However, there was no significant difference in the value of ΦNO between the two treatments expect 5 dpi (Fig. 2d). Infected with P. aphanis, reductions in Fv/Fm, ΦPSII, and ΦNPQ were detected at 3 days, which was consistent with the appearance of visible symptoms. Therefore, these divergent phenotypes (3 dpi) were used in the transcriptome analyses.
Changes in the transcriptome in strawberry leaves infected with P. aphanis
To determine the transcriptome profile of strawberry in response to P. aphanis, RNA sequencing (RNA-Seq) analyses were performed on 12 samples (ddH2OInfected, ddH2OUnifected, SAInfected, and SAUnifected, three replicates per treatment) (Fig. 2a). Approximately 600 million raw reads were obtained in total, the clean reads were mapped to the F. ananassa_Camarosa genome (Table S1). FPKM values of DEGs were used to calculate the fold changes of ddH2OInfected/ddH2OUnifected and SAInfected/SAUnifected. Principle component analysis (PCA) showed that PC1 and PC2 could explain 64.40% of the total transcript expression level variance, in which PC1 explained 50.96% of the total detected variation according to genotype, while the PC2 separated samples according to treatment and explained 12% of the variance (Figure S2). There was greater separation on PC2 for samples from the SAInfected and SAUninfected genotype, indicating a stronger transcriptional differentiation during SA-pretreatment.
Among the two groups, 48,020 transcripts and 45,896 transcript genes were expressed in the strawberry leaves from ddH2OInfected/ddH2OUnifected and SAInfected/SAUnifected leaves, 43,103 genes were commonly expressed in ddH2OInfected/ddH2OUnifected leaves, 2770 genes were expressed only in ddH2OUnifected leaves, and 2147 genes were expressed only in ddH2OInfected leaves. While SAInfected/SAUnifected had 41,939 commonly expressed genes, 1755 genes were expressed only in SAUnifected and 2202 in SAInfected leaves (Figure S3a). Based on p-adjust < 0.05 and |log2FC| ≥ 1, a total of 4417 and 3754 DEGs were detected in ddH2OInfected/ddH2OUnifected and SAInfected/SAUnifected leaves, respectively. As indicated in Figure S2ab, 2224 genes were upregulated in SAInfected/SAUnifected leaves compared to 2110 in ddH2OInfected/ddH2OUnifected (log2FC ≥ 1 for upregulated). By contrast, a greater number of genes were downregulated genes in ddH2OInfected/ddH2OUnifected (2088) compared to SAInfected/SAUnifected leaves (1530) (log2FC ≤ − 1 for downregulated). However, there were commonly 921 genes expressed at two time-points. A great number of DEGs (upregulated of 1476 genes in ddH2O treatment, upregulated of 1565 genes in SA treatment, downregulated of 1801 genes in ddH2O treatment, downregulated of 1268 genes in SA treatment) were noticeable in both groups, indicating that DEGs in response to P. aphanis varied greatly (Fig. S3d). Based on BIRCH clustering, a total of 20 clusters of expression profiles were identified with distinguishable expression patterns during strawberry-P. aphanis interaction in ddH2O and SA treatment (Figure S3e). The powdery mildew infection lead to the rapid up- or down-regualtion of transcripts in both ddH2O treatement (cluster 1, 4, 6, 3, and 5) and SA treatment (cluster1, 4, 6, 2, 3, and 5), indicating transcripts involved in fungus responses. Cluster7, cluster8, cluster9, and cluster10 showed irregular changes in both treatments. In general, the results demonstrated that DEGs in the ddH2O treatemnt were delayed compared with the SA treatment.
Twenty enriched GO terms were mainly categorised as biological process and molecular function (Figure S4ab). In the ddH2O-treated group, most genes were involved in oxidation-reduction process, metal ion binding, cation binding, and oxidoreductase activity. By contrast, only single-organism metabolic process, oxidation-reduction process, and oxidoreductase activity were associated with the SA-treated group. KEGG enrichment analysis showed some of the same pathways, such as phenylpropanoid biosynthesis, phenylalanine metabolism, and flavonoid biosynthesis (Figure S4cd). Enrichment pathway under both groups was plant hormone signal transduction (Figure S5), indicating that SA and JA signalling pathway were mainly involved in the response to P. aphanis. We concluded that the genes for phenylpropanoid and flavonoid biosynthesis, and TFs involved in hormone signal transduction played an important role in strawberry defence aganist P. aphanis.
The flavonoid biosynthesis pathway participates in resistance against P. aphanis
To determine whether exogenous SA could trigger PA accumulation upon P. aphains attack, TFC and PA metabolites were measured (Fig. 3a). The SA concentration in the SA-treated group was significant higher than that in the ddH2O-treated group at two infection points. Moreover, there is no significant difference in TFC content between two groups, although the value was higher in the SA-treated group than in the ddH2O-treated group. Additionally, PA levels were significant higher in the SAInfected than in the ddH2OInfected. To further clarify the regulatory mechanism of SA-triggered the accumulation of PAs, the regulation of DEGs associated with phenylpropanoid and flavonoid pathways was investigated (Fig. 3c). RNA-Seq showed the upregulated expression of key genes involved in the flavonoid pathway. Transcript levels of 4-coumarate-CoA ligase 2 (4CL) encoding genes, chalcone synthase (CHS) encoding genes, chalcone isomerase (CHI) encoding genes, flavanone 3-hydrolase (F3H) encoding genes, dihydroflavonol reductase (DFR) encoding genes, leucoanthocyanidin reductase (LAR) encoding genes, anthocyanidin synthase (ANS) encoding genes, and anthocyanidin reductase (ANR) encoding genes were increased at 3 dpi compared with 0 dpi in both groups. Moreover, the expression levels of these genes were also higher in the SA-treated group than that in the ddH2O-treated group (Table S4). Compared with increased expression of UDP-glucose: anthocyanidin: flavonoid glucosyltransferase (UFGT) encoding genes in the ddH2OInfected, the expression of UFGT in the SAInfected was markedly downregulated, indicating that SA could suppress UFGT to produce more PAs. Overall, we suggest a potential role of PAs in enhancing resistance against P. aphanis. Therefore, we propose that TFC and PAs are potentially important antifungal compounds in defence against P. aphanis.
The MBW complexes involved in PA biosynthesis induced upon P. aphains
Previous study showed that FaMYB9/FaMYB11, FabHLH3 and FaTTG1 functionally interact to regulate PA synthesis during strawberry fruit development [28]. Phylogenetic analysis showed that FaMYB5–1, FaMYB5–2, and FaMYB5–3 are most similar to AtMYB5. R2R3-MYB motif analysis showed that motif 1 and motif 2 were R2 conserved domain, while motif 3 was R3 conserved domain (Fig. 4a). In addition, another two R2R3-MYB, FaMYB9 and FaMYB11 are homologues of MdMYB9 and MdMYB11. Interestingly, phylogenetic analysis shows that the identified proteins were similar in length to their homologues, expect for FabHLH33 (FxaC_27g23560). Motif analysis also shows that there is missing motif 5 (the bHLH domain) and motif 6 (the ACT-like dimerization domain) in FabHLH33, which is more closely related to MdbHLH33 and VvMYCA1 than to AtEGL1/AtEGL3 (Fig. 4b). Besides, FabHLH3 is similar to VvMYC1, MdbHLH, PhAN1 and AtTT8. FaTTG1 showed the high similarity to MdTTG1 and all identified proteins had two WD40 domains (motif 8 and motif 9) (Fig. 4c).
The expression pattern analysis of key genes (FaMYB9, FaMYB11, FabHLH3, FabHLH33, and FaMYB5) encoding the MBW complex were performed to understand the differences in TFC and PAs in two treatments (Fig. 4d). The expression pattern of the FaMYB5, FaMYB9, FaMYB11, FabHLH3 and FaTGG1 had same trend, with the higher expression level in SAInfected compared to ddH2OInfected (Fig. 4d), which may be related to PAs accumulation pattern in the same point (Fig. 2a). In turn, FabHLH3 does not show significant expression level at two point under two groups, although SA treatment also increased its level at 3 dpi compared to the inhibition of its level at 0 dpi. The significant differences in FabHLH33 between treatments were observed at 0 dpi, and then SA treatment further decreased its level at 3dpi. In the case of FaTTG1, a high expression was observed in SAUnifected.
The SA biosynthesis and SA signalling pathway contributes to enhanced resistance
SA biosynthesis occurs in two distinct pathways; the phenylalanine pathway and the ICS1 pathway, from which nearly 95% of SA is produced [29]. The ISC1 pathway involves two steps: the conversion of chorismate to isochorismate, which is first catalyzed by ICS1, followed by the conversion of isochorismate to SA by an unknown enzyme [30]. For the phylogenetic analysis, highly homologous proteins of strawberry were identified (Figure S5a-e). In this study, gene encoding ICS1 were downregulated in both groups (Fig. 3d and Figure S6), while two PAL genes showed strong upregulation. Additionally, expression analysis showed FaICS2 with very low expression but high expression in FaPAL. In general, these results suggested that the SA biosynthesis of strawberry might be mainly from PAL pathway. Moreover, EDS1 and PAD4 act upstream of SA accumulation at the infection site, while expression of the EDS1/PAD4 complex can be increased by exogenous SA [31]. Compared with 0 dpi, FaPAD4 showed upregulation in both groups at 3 dpi (Figure S5e). No obvious change in FaEDS1 was observed between ddH2OInfected and ddH2OUninfected, whereas SA induced strong expression of FaEDS1 at two infection points in SA-treated group, which suggests P. aphanis might suppress EDS1 to inhibit SA biosynthesis. Interestingly, no significant difference in FaPAD4 between two groups was observed, indicating that EDS1 might paly crucial role in mediating SA accumulation in strawberry.
In the transcriptional reprogramming outline of transcripts associated to strawberry resistance to P. aphains and the accumulation of SA content indicates that genes for SA biosynthesis and TFs involved in SA signaling, play a pivotal importance. In this study, they were further characterised in terms of phylogeny analysis and expression pattern aiming to dissect their functional roles in the P. aphains response in strawberry. There are DEGs across the two treatments, such as NPR, TGA, WRKY, MYC2 and JAZ, which are involved in signal transduction (Figure S5). As expected, as SA and JA pathways are activated in both groups, require large scale transcriptional reprogramming, especially transcription factors. To further identify a possible role for these DEGs in the regulation of defence responses, using protein sequence of these DEGs, unrooted phylogenetic trees were constracted to show evolutionary relationships between F. ananssna and A. thaliana (Fig. 5).
Non-expressor of pathogenesis-related gene 1 (NPR1) is positive regulator of the SA-dependent signaling pathway, and then mediates the binding of TGA factors to the as-1 motif found in the pathogenesis-related PR-1 gene. FaNRP3-like was phylogenetically closer to FvNPRL-1, which was more similar to AtNPR3/AtNPR4 than AtNPR1. FvNPRL-1 likely functions similar to Arabidopsis NPR3/NPR4 as a negative regulator of the SA-mediated defense [32]. As shown in Fig. 5, the strong induction of FaNPR3-like was detected as early as 4 h after SA treatment, then decreased rapidly in both groups, suggesting that FaNPR3-like (similar to FvNPRL-1) was a repressor to quickly balance the effect caused by excessive SA. Therefore, FaNPR3-like likely acts as negative regulators of SA-mediated defence pathway. As an important co-activator of NPR1, FaTGA showed a contrasting expression profile between ddH2OInfected/ddH2OUninfected and SAInfected/SAUninfected in this study (Fig. 5). FaTGA shared high similarity to FaTGA1, which might interact with FaNPR1 and play a crucial role in the response to powdery mildew disease [33]. Furthermore, FaTGA also showed closly evolutionary relationships with AtTGA, which was shown to regulated SA biosynthesis by modulating SARD1 [29]. Thus, FaTGA1 likely function as positive regualtors of SA biosynthesis and SA-mediated defence. It is clear that WRKY70 is a downstream regulator of NPR1 and COI1 in positively mediating SA-dependent genes, compared to negatively regulate JA-responsive genes [30]. Although upregulation trend of FaWRKY70 (similar to SlWRKY70 and AtWRKY70) expression was observed in both groups, the transcript level of FaWRKY70 was significantly higher in SA-treated group than in ddH2O-treated (Fig. 5), indicating that SA mediates FaWRKY70 accumulation, which is consistent with previous studies [31, 34]. Because AtWRKY70 is a negative regulator of JA-responsive genes, we hypothesized that FaWRKY70 orthologs maybe a positive regulator of SA-responsive genes.
DEGs involved in JA-dependent defence pathways
It is well known that the JA-SA crosstalk results in fine-tune plant’s defence against different pathogens [35]. The JAZ (JASMONATE-ZIM DOMAIN) proteins are key transcriptional repressors regulating various biological processes. As key repressors in JA responses, when JA-Ile levels are low, JAZ binds to MYC2, leading to repression of JA-responsive genes [36]. The JA signalling pathway could trigger PAs biosynthesis at early stages of strawberry fruit development, especially FaMYC2, FaJAZ1 and FaJAZ8.1, which are JA-responsive genes and correlated with the activation of JA-Ile biosynthesis [37]. Although FaJAZ1 (similar to FvJAZ1) displayed a reduction pattern during infection in both groups, a pronounced expression of FaJAZ1 induced by SA at two points was observed (Fig. 5), according to PAs accumulation during SA-treated leaves response to P. aphains, indicating that exogenous SA could induce expression of JA biosynthesis genes at the early stage of infection, leading to accumulation of PAs to defence against fungus. In contrast, the expression of FaJAZ10–1 and FaJAZ10–2 showed a similar increment pattern from 0 to 3 dpi in both groups. The evolutionary relationship showed that FaMYC2-like had the highest identify with FvMYC2-like, FaMYC2–1 and FaMYC2–2 were clustered closely to FvMYC2 [37]. FaMYC2–1 and FaMYC2–2 showed an expression increment in SA-treatment during infection in a similar way to that observed for FaJAZ10–1 and FaJAZ10–2 (Fig. 5). Collectively, a significant greater expression level was observed for FaJAZ1 and FaMYC2-like in SA-treatment than that in ddH2O-treatment. Moreover, higher relative expression levels were observed for FaMYC2-like and FaJAZ1 in SAUnifected. The expression pattern of genes encoding for key enzymes involved in JA biosynthesis such as FaJAR, FaAOS, and FaLOX2 were also analyzed (Figure S6). At 0 dpi, SA-treated leaves exhibited higher transcript levels of these genes, which decreased from 1 to 3 dpi, expect FaLOX2 (Figure S6). Interestingly, expression of FaLOX2 in SAUninfected was higher than in ddH2OUninfected; however, at 3 dpi, no significant difference for FaLOX2 between two groups was observed. Expression of FaAOS and FaJAR1 showed similar patterns across all time points. Furthermore, these genes were highly homologous to those in previously studied [38, 39]. Taken together, key genes involved in JA biosynthesis followed the same downregulation pattern, which is in agreement with an antagonistic relationship between SA and JA pathway. Overall, JAZ1, JAZ10–2, MYC2-like, MYC2–1, and MYC2–2 and other JA-signalling related genes (FaLOX2, FaAOS, and FaJAR) are downregulated during P. aphanis infection, compared to SA induced higher expression.
The expression pattern of pathogen-related resistance genes
To investigate the defence response involved in SA-induced resistance, the major transcriptional changes (log2 fold change > 3) in response to P. aphanis infection between the SAInfected and SAUninfected were monitored (Table S5), including expression of PR1, PR2 (endo-1,3-glucanase), PR3 (chitinase), PR5 (thaumatin-like protein), PR9 (plant peroxidase), and PR10. Expression patterns of the PR genes in both groups were further examined by RT-qPCR (Fig. 6). The expression of these nine PR genes in strawberry leaves in SA treatment showed a similar trend as that observed with RNA-Seq. SA had a direct influence on FaPR1, FaPR2, FaPR3, FaPR5, and FaPR10, as evidenced by the significant difference in expression in the SA-treated group compared with the ddH2O-treated group. Moreover, during the experimental period, the expression of all PR genes continued for 5 days. In response to P. aphanis, the SA treatment exhibited significant higher expression of FaPR1, FaPR2–1, FaPR2–2, FaPR3, FaPR5–1, and PR5–2 expression than that in the ddH2O treatment. These genes were also significant upregultated (1 dpi) prior to P. aphanis infection (white mycelium visible to the naked eye, 3dpi). These results suggested that FaPR1, FaPR2–1, FaPR2–2, FaPR3, FaPR5–1, and PR5–2 might play direct defence against P. aphanis.