Ubiquitination of the PpMADS2 transcription factor controls linalool production during UV-B irradiation in detached peach fruit

Abstract

Plant secondary metabolites undergo changes in response to UV-B irradiation. Although UV-B irradiation reduces flavor-associated volatile compounds in detached peach (Prunus persica L. Batsch) fruit, the underlying regulatory mechanisms remain unclear. By integrating proteomic, transcriptomic, and metabolomic data from peach fruit following UV-B irradiation, we discovered that the detached fruit responds to UV-B by suppressing the biosynthesis of the flavor-related monoterpene linalool. We identified PpMADS2, a transcription factor that regulates linalool biosynthesis by activating terpene synthase 1 (PpTPS1) expression. PpMADS2 overexpression in peach and tomato fruits significantly increased linalool levels compared with the controls. Proteomic data and immunoblots revealed a decrease in PpMADS2 abundance following exposure to UV-B. Moreover, our results demonstrated that PpMADS2 interacts with the E3 ubiquitin ligase PpCOP1 both in vitro and in vivo. The UV-B-induced 26S-proteasome-mediated degradation of PpMADS2 is largely PpCOP1-dependent. Taken together, our findings demonstrate that linalool biosynthesis in detached peach fruit exposed to UV-B radiation is governed by the PpCOP1–PpMADS2–PpTPS1 module. This study enhances our understanding of the interplay between light signaling and fruit flavor quality. Multiomics approaches offer valuable resources for investigating the mechanisms underlying how light influences metabolism in fruit crops.

Introduction

Light, ranging from ultraviolet-B (UV-B, 280 to 315 nm) to far-red (FR) (∼750 nm) wavelengths, is essential for plant life, and perception of the light environment determines plant growth, morphology, and developmental changes (Lau and Deng 2010; de Wit et al. 2016). Plants are able to respond to light radiation by producing various metabolites, including fatty acids, terpenoids, phenolic compounds, alkaloids, and flavonoids, which in turn affect crop quality and human health (Jaakola and Hohtola 2010; Liu et al. 2023b).

Plants perceive different wavelengths of light through an array of photoreceptors, including UV RESISTANCE LOCUS 8 (UVR8), which detects and transmits signals for UV-B perception (Jenkins 2014; Pham et al. 2018; Wang and Lin 2020; Yadav et al. 2020). The monomeric UVR8 activated by UV-B interacts with CONSTITUTIVE PHOTOMORPHOGENIC 1 (COP1), a core light signaling modulator in response to UV-B processes (Podolec and Ulm 2018; Han et al. 2020). The pivotal role of COP1 lies in its ability to target and degrade various transcription factors (TFs) via the ubiquitin/26S proteasome system. For example, ELONGATED HYPOCOTYL 5 (HY5), a TF regulating expression of genes responding to UV-B irradiation, can be degraded by COP1 in plants (Podolec et al. 2021). Additionally, in the absence of light, many fruits are unable to accumulate anthocyanins due to the ubiquitination of key TFs, including MYBs in apple and pear, by COP1 through the 26S proteasome degradation pathway (Li et al. 2012; Tao et al. 2020). The impact of UV-B irradiation on plant secondary metabolites has been extensively investigated (Jaakola and Hohtola 2010; Eichholz et al. 2011; Song et al. 2015; Liu et al. 2017; Liu et al. 2023a; Liang et al. 2024; Wang et al. 2024). However, compared with progress in our understanding of regulation of flavonoids, including flavonols and anthocyanins, by UV-B, there is a lack of understanding of regulation of terpenoid synthesis in response to UV-B.

Terpenoids are the most abundant class of secondary metabolites in plants, and have enormous roles in growth and development. Terpenes are synthesized from the 5-carbon precursors isopentenyl pyrophosphate (IPP) and dimethylallyl pyrophosphate, which are derived from the plastid located methylerythritol-4-phosphate (MEP) and the cytoplasmic mevalonate (MVA) pathways (Vranova et al. 2012, 2013). Terpenes are classified based on the number of 5-carbon isoprene units in their skeletons, ranging from hemi- (C₅) to tetra- (C₄₀) terpenoids. Terpene synthases (TPSs) are the key enzymes responsible for synthesizing the basic terpenoid backbone structures (Vranova et al. 2012, 2013). Studies on light regulation of terpenoids have focused on the oxygenated sesquiterpene (C₁₅) artemisinin and the tetraterpenoid (C₄₀) carotenoids. AaHY5 positively regulates artemisinin accumulation by directly inducing TF GLANDULAR TRICHOME-SPECIFIC WRKY1 (AaGSW1), AaWRKY9, and AaWRKY14 in the artemisinin synthesis pathway (Hao et al. 2019; Fu et al. 2021; Zhou et al. 2021). Moreover, TF AaMYB108, which interacts with AaGSW1 to regulates artemisinin synthesis, can be ubiquitinated by AaCOP1, revealing the complexity of the transcriptional regulatory network controlling artemisinin biosynthesis in response to light (Liu et al. 2023a). HY5 and PHYTOCHROME INTERACTING FACTORS (PIFs) act antagonistically to regulate expression of phytoene synthase (PSY) in carotenoid synthesis. PIF inhibits PSY expression, whereas HY5 activates its expression in response to shade (Liu et al. 2004; Toledo-Ortiz et al. 2010; Llorente et al. 2016; Wang et al. 2021).

In addition to providing pigmentation and essential nutrients, terpenoids with volatile properties serve as crucial signaling molecules for plants, playing a pivotal role in attracting pollinators, facilitating fruit consumption, and aiding in seed dispersal (Pichersky and Raguso 2018). The linear monoterpene linalool (C₁₀), known for its floral and sweet flavor, is a widely distributed volatile terpenoid within the plant kingdom. It plays a crucial role in influencing consumer preference towards plant-based foods. It has been reported that light exposure leads to a decrease in linalool content, thus reducing flavor quality of fruit and wines (Song et al. 2015). Linalool contents in wines bottled in transparent glass were reduced by 20% to 30% after 3 wk and by 30% to 50% after 50 d on shelves (Carlin et al. 2022). In contrast, for wines in colored glass bottles, linalool was protected after 50 d of storage. Similar light-induced damage to linalool has also been observed in fruit crops, including peach. As an important economic crop, peach (Prunus persica L. Batsch) has annual production exceeding 24 million tons with a remarkable 20% increase in the past decade. Over 100 volatile chemicals have been identified in peach fruit, among which linalool plays a critical role as a key odorant influencing fruit aroma and consumer preference (Cao et al. 2024). Exposure to UV-B irradiation resulted in a decrease in linalool content in peach fruits of various cultivars, consequently compromising the overall quality of fruit flavor (Liu et al. 2017; Wei et al. 2021). However, the mechanism underlying UV-B-mediated regulation of linalool synthesis remains undefined. Our previous investigations have identified peach TPS genes involved in linalool production, as well as TFs, including PpbHLH1 and PpERF61, that modulate linalool biosynthesis (Liu et al. 2017; Wei et al. 2021, 2022).

These observations provide a framework for analysis of the molecular mechanisms underlying the loss of peach flavor induced by UV-B radiation. In addition to the economic implications associated with the deterioration in flavor quality, peaches possess a genome that offers an ideal system for investigating the impact of environmental stress on a genome scale. In this study, we present a comprehensive analysis of the impact of UV-B radiation on the proteome, transcriptome, and flavor metabolome. Our findings reveal that UV-B negatively regulates monoterpene formation by PpCOP1 ubiquitination of the PpMADS2 TF that activates PpTPS1 expression by binding to its promoter. These results demonstrate that a PpCOP1–PpMADS2–PpTPS1 module governs linalool synthesis in peach fruit following exposure to UV-B irradiation.

Results

Proteomic data unveil substantial changes in protein content of detached peach fruit upon exposure UV-B irradiation

To identify and quantify UVB-regulated proteins in detached peach fruits, peel samples were collected from fruits irradiated with UV-B for 6 and 48 h, followed by protein extraction and quantitative proteomic analysis. TMT labeling-based proteome quantification was conducted for each group (6 h_CK, 6 h_UVB, 48 h_CK, and 48 h_UVB) with 3 biological repeats. A total of 9,215 unique proteins were identified at a false discovery rate (FDR) value of 1% (Supplementary Table S1). For quantification, only proteins with intensity values across all 12 samples were included, resulting in the quantification of 7,360 unique proteins across all samples (Supplementary Table S2). Principal component analysis (PCA) demonstrated that the proteins within each group cluster together and that there was a distinct separation between the treatment and control groups at each time point (Fig. 1A). Correlation analysis of proteomic data from 3 biological replicates of UV-B treatment and control indicates a high level of reproducibility among the biological replicates (Fig. 1B). Integrated PCA and correlation analyses revealed that proteins in peach fruits exhibit significant alternations in response to UV-B exposure, particularly after 48 h.

To further screen differentially expressed proteins (DEPs) between treatments, we compared the abundance values of the samples from the control and UV-B-treated groups. A fold change >1.3 with P-values <0.05 was used as the cutoff threshold to indicate significant changes in DEPs (Supplementary Tables S3 to S4). Using these criteria, a total of 122 and 799 DEPs were identified after 6 and 48 h of UV-B irradiation, respectively (Fig. 1C). Notably, the number of DEPs increased more than 5-fold after 48 h of UV-B irradiation. Furthermore, among these DEPs, there was a higher proportion of proteins induced compared with those inhibited by UV-B.

The DEPs were subjected to gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses to investigate the impact of UV-B irradiation on protein regulation in peach fruit (Fig. 1, D and E; Supplementary Table S5 to S8). After 6 h of UV-B irradiation, significant enrichment was observed in terms related to “DNA-binding TF activity,” “Abscisic acid-activated signaling pathway,” and “Response to reactive oxygen species” (Fig. 1D; Supplementary Table S5). In contrast, after 48 h of UV-B irradiation, GO analysis revealed enrichment in terms associated with “Abscisic acid binding,” “IPP biosynthetic process,” “Isoprenoid biosynthetic process,” and “Response to UV-B” (Fig. 1D; Supplementary Table S6). KEGG analysis indicated that at 6 h, the DEPs were enriched in pathways such as “Sesquiterpenoid and triterpenoid biosynthesis,” “MAPK signaling pathway,” and “Plant hormone signal transduction” (Fig. 1E; Supplementary Table S7), while at 48 h they were enriched in pathways including “Biosynthesis of unsaturated fatty acids,” “Terpenoid backbone biosynthesis,” “Monoterpenoid biosynthesis,” and “Glutathione metabolism” (Supplementary Table S8). These findings suggest that UV-B regulates multiple pathways and metabolic processes at the protein level in peach fruit.

It is noteworthy that multiple pathways related to terpene metabolism are significantly enriched in the GO and KEGG analysis of DEPs following UV-B treatment. Specifically, pathways such as “Terpenoid backbone biosynthesis,” “Monoterpenoid biosynthesis,” “Sesquiterpenoid and triterpenoid biosynthesis,” “IPP biosynthetic process,” “Mevalonate pathway,” “Isoprenoid biosynthetic process,” and “Carotenoid biosynthesis” were enriched following 6 and 48 h of UV-B irradiation (Fig. 1, D and E; Supplementary Table S5 to S8). Given our previous findings on the regulatory impact of UV-B irradiation on volatile terpenoids in peach fruit (Liu et al. 2017; Wei et al. 2021), it is imperative to elucidate the underlying molecular mechanisms at the protein level.

Exposure to UV-B irradiation induces significant alternations in fruit metabolite profiles

KEGG analysis revealed significant enrichment of DEPs in multiple pathways, including “Biosynthesis of unsaturated fatty acids,” “Fatty acid elongation,” and “alpha-Linolenic acid metabolism” associated with fatty acids (Fig. 1E; Supplementary Table S8). The results from the fatty acid analysis demonstrated a significant decrease in the content of the saturated fatty acid stearic acid (18:0), as well as the unsaturated oleic acid (18:1), and linoleic acid (18:2) after 48 h UV-B treatment, while there was no significant change observed in linolenic acid (18:3) content (Supplementary Fig. S1A). Fatty acid desaturase (FAD) plays a crucial role in removing hydrogen from carbon chains during the biosynthesis of unsaturated fatty acids to form C=C bonds (Shanklin and Cahoon 1998). Protein levels of PpFAD2 and stearoyl ACP desaturase (SAD), PpSAD1 and PpSAD2, were significantly reduced following 48 h of UV-B irradiation exposure (Supplementary Fig. S1B, Supplementary Table S9). In addition, the protein contents of the 3-Ketoacyl-CoA synthases (PpKCSs, Prupe.1G401500 and Prupe.1G310500), responsible for the synthesis of very long-chain fatty acids (VLCFAs), along with very long-chain hydroxyacyl-CoA dehydratase (Prupe.6G201300), involved in VLCFA elongation process, were also notably decreased (Supplementary Table S9). These findings suggest that under UV-B irradiation conditions it is possible for VLCFAs content to be reduced.

Elevated levels of UV-B irradiation can induce oxidative stress in plant cells, resulting the generation of reactive oxygen species (ROS). Glutathione (GSH), a soluble peptide abundantly present in cells, plays a crucial role in scavenging intracellular free radicals and hydrogen peroxide (H₂O₂). KEGG analysis revealed a significant enrichment of DEPs in the “Glutathione metabolism” pathway (Fig. 1E). Following UV-B irradiation, the entire glutathione metabolic pathway was activated (Supplementary Table S9). The protein abundance of multiple glutathione S-transferases (GSTs), glutathione reductase (GR, Prupe.1G004800), and glutathione peroxidase (GP, Prupe.5G098400) significantly increased after 48 h of UV-B exposure (Supplementary Table S9). Additionally, 2 H₂O₂-scavenging catalase enzymes (Prupe.5G011300 and Prupe.5G011400) were significantly induced by UV-B irradiation (Supplementary Table S9). These results suggest that the metabolic process involved in scavenging peroxides is activated by UV-B irradiation in peach fruits through the participation of glutathione. These results are consistent with observations from other species (Loyall et al. 2000; Apel and Hirt 2004; Jia et al. 2023).

Production of MEP pathway-derived terpenoids is repressed by UV-B irradiation in fruit

Given the significant enrichment of multiple pathways associated with terpenoid metabolism (Fig. 1, D and E), we analyzed the changes in both terpenoid content and proteins involved in terpenoid synthesis pathways under UV-B treatment. A total of 13 monoterpenes, 3 sesquiterpenes and 3 carotenoid-derived terpene volatile compounds were detected (Fig. 2A). The sesquiterpene farnesene, synthesized by the MVA pathway (Supplementary Fig. S2A), was significantly induced by UV-B (Liu et al. 2017), coinciding with a significant increase in protein abundance of several key enzymes (PpHMGS, PpHMGR, PpMVK, PpPMK, PpMPDC, PpIDI, and PpFPPS) within the MVA pathway (Fig. 2A; Supplementary Fig. S2). PpTPS2 has been identified as the enzyme responsible for farnesene synthesis in peach (Liu et al. 2017; Wei et al. 2024). Notably, however, the PpTPS2 protein was under detection limits in the proteomic data (Supplementary Table S2). On the other hand, β-ionone and 7,10-dihydro-β-ionone, which are derived from carotenoids in peach, exhibited significantly reductions upon UV-B treatment (Supplementary Fig. S3A). The degradation of β-carotene into β-ionone catalyzed by carotenoid cleavage dioxygenase (CCD), and the proteomic analysis revealed a significant suppression of PpCCD4 protein in the 48 h UV-B treatment samples (Supplementary Fig. S3B).

The monoterpene linalool, which is associated with flavor, constitutes the largest proportion of terpenoid volatiles in peach fruits (Fig. 2A; Wei et al. 2021). Exposure to UV-B irradiation resulted in a significant reduction in linalool levels (Fig. 2A). After 48 h of UV-B irradiation, the linalool content decreased by ∼40%, from 1,333 ng g⁻¹ fresh weight to 800 ng g⁻¹ fresh weight. To comprehensively investigate the regulation of gene transcription levels in peach fruits by UV-B treatment, transcriptome analysis was performed on samples exposed to UV-B (Supplementary Tables S10 and S11). The transcriptome data were quality controlled to determine the genome-wide expression profiles following UV-B irradiation (Supplementary Table S12). A heatmap of the transcript and protein levels of the MEP pathway was presented in Fig. 2B. The transcript and protein levels of key enzymes involved in the MEP pathway, including PpDXS1 and PpDXS2 as rate-limiting enzymes, along with PpCMK, PpHDS, and PpHDR were reduced following UV-B irradiation (Fig. 2B; Supplementary Table S13). PpTPS1 and PpTPS3 are 2 enzymes involved in the biosynthesis of linalool in peach fruit (Liu et al. 2017; Wei et al. 2021, 2022, 2024). However, the PpTPS3 protein was under detected limits through proteomic sequencing (Supplementary Table S2). As shown in Fig. 2B, both the transcript level and protein abundance of PpTPS1 were observed to decrease following UV-B treatment. This result suggests that the decrease in linalool may be attributed to suppression of transcription and subsequent protein abundance of these enzymes caused by UV-B exposure.

PpMADS2 positively regulates production of linalool in fruit

TPSs are crucial terminal enzymes that catalyze the formation of monoterpenes such as linalool using geranyl diphosphate as a substrate (Vranova et al. 2012, Fig. 2B). Multiple TFs play pivotal roles in regulating terpene biosynthesis (Wei et al. 2021). To further investigate the mechanism underlying linalool loss caused by UV-B irradiation, we conducted a TF screening targeting the PpTPS1 gene. Based on genome annotation and information from the PlantTFDB (https://planttfdb.gao-lab.org/), a total of 35 TFs were identified from DEPs at 6 and 48 h (Supplementary Table S14). Among these TFs, 19 showed significant decreases in protein abundance after UV-B irradiation, suggesting potential roles as positive regulatory factors for linalool biosynthesis. Dual luciferase results demonstrated that PpMADS2 (Prupe.5G208400) exhibited the highest activation effect (8.5-fold) on PpTPS1 promoter activity, indicating that PpMADS2 likely acts as a vital TF governing linalool synthesis and accumulation (Fig. 3A).

PpMADS2, a member of the MADS-box TF family, typically binds to a consensus recognition sequence known as CArG-box [CC(A/T)₆GG] (Riechmann et al. 1996). Therefore, we conducted an electrophoretic mobility shift assay (EMSA) to validate the activation of PpTPS1 transcription by PpMADS2 through binding to the CArG-box. EMSA results demonstrated that GST-PpMADS2 effectively bound to the biotin probe containing the CArG-box element within the PpTPS1 promoter (Fig. 3B). Furthermore, when the predicted binding site was mutated, binding was completely eliminated, indicating that PpMADS2 activates expression of PpTPS1 by directly binding to the CArG-box element in its promoter.

To further investigate the function of PpMADS2 in planta, we examined the transcript levels of PpMADS2 and PpTPS1 in different tissues and developmental stages of peach fruits using transcriptome data (Supplementary Fig. S4) (Wei et al. 2022). Heatmap analysis revealed that both PpMADS2 and PpTPS1 exhibit fruit-specific expression patterns, with significantly higher transcript levels observed at the S3 stage (the second fast growth, 94 DAB) of fruit. These observations suggest that they share similar expression profiles (Supplementary Fig. S4). Subsequently, we conducted transient overexpression experiments in peach fruit and stable transgenesis in tomato. Overexpression of PpMADS2 in peach fruit resulted in a significant increase (1.97-fold) in PpTPS1 expression compared with the empty vector control, accompanied by an ∼2-fold increase of linalool (Fig. 3C). For transgenic overexpression of PpMADS2 in tomato, we constructed the PpMADS2-pBIN19-E8 vector where expression is driven by the fruit-specific E8 promoter. The T₁ generation of transgenic tomato plants along with wild-type (WT) controls were sampled at breaker (Br) + 7 d (red ripe stage) with 3 biological replicates for transcript level analysis to confirm successful overexpression of PpMADS2 (Fig. 3D). Compared with WT tomato fruits, overexpressing lines showed significantly enhanced linalool contents in 2 independent lines (Line 35 and Line 80). Collectively, these results provide strong evidence supporting the conclusion that PpMADS2 acts as a positive regulator for linalool synthesis.

PpMADS2 protein content is suppressed by UV-B irradiation in fruit

Proteomic data revealed a significant inhibition of PpMADS2 protein level, the content reduced to 0.84 relative to the control after 6 h of UV-B irradiation, and 0.71 at 48 h of irradiation (Fig. 4A). To validate the regulatory impact of UV-B irradiation on PpMADS2 protein, a polyclonal rabbit antibody against PpMADS2 protein was generated for protein blot analysis to determine its abundance in fruit (Fig. 4B). The relative abundance of PpMADS2 was quantified through densitometric analysis using ImageJ software based on immunoblotting results (Fig. 4C). The results demonstrated that while there was only a slight reduction in PpMADS2 protein abundance at 6 h of UV-B irradiation, the protein bands were significantly weakened at 48 h, consistent with the proteome sequencing findings (Fig. 4A). These results indicated that UV-B irradiation significantly reduces PpMADS2 protein levels and is likely causative for reduced linalool levels in fruit.

Plant response is known to UV-B primarily relies on the UVR8-COP1-HY5 signaling pathway. Protein analysis showed that UV-B irradiation significantly induced levels of PpCOP1 and PpHY5 (Fig. 4D). However, no significant changes were observed in the content of PpUVR8 protein, possibly due to its ability for repetitive utilization (Fig. 4D) (Fang et al. 2022).

PpMADS2 interacts with the E3 ubiquitin ligase PpCOP1

The E3 ubiquitin ligase PpCOP1 plays a crucial role in the UV-B signaling pathway. To investigate the underlying mechanism behind the significant reduction in PpMADS2 protein abundance following UV-B treatment, we investigated the interaction between PpCOP1 and TFs (Supplementary Fig. S5). Besides PpMADS2, 2 previously identified peach TFs that regulate linalool biosynthesis, namely PpbHLH1 and PpERF61 (Wei et al. 2021, 2022), were simultaneously subjected to yeast 2-hybrid (Y2H) analysis with PpCOP1 (Supplementary Fig. S5). The Y2H results revealed that out of these 3 TFs, only PpMADS2 interacted with PpCOP1. Furthermore, we observed that neither of the other 2 light signaling elements, namely PpUVR8 and PpHY5, could interact with PpMADS2 (Supplementary Fig. S5). The structure of the PpCOP1 protein consists of 3 domains: RING finger (N-Ring), coiled-coil (N-Coil), and WD40 repeats (WD40) as depicted in Fig. 5A and Supplementary Fig. S6. To identify the specific region responsible for interaction between PpMADS2 and PpCOP1, fragments corresponding to different structural domains (N282, N-Ring, N-Coil, and WD40) of PpCOP1 were cloned into the pGBKT7 (BD) vector for subsequent Y2H analysis with the PpMADS2 protein. As shown in Fig. 5A, the N-Coil and WD40 structural domains of PpCOP1 interacted with PpMADS2, while the N-Ring failed to directly bind to PpMADS2.

Luciferase complemention imaging (LCI) and co-immunoprecipitation (Co-IP) assays were performed in Nicotiana benthamiana leaves to further validate the interaction between PpCOP1 and PpMADS2 in planta. Notably, leaves co-infiltrated with PpMADS2-nLUC (encoding PpMADS2 fused to the N-terminal half of LUC) and PpCOP1-cLUC (encoding the C-terminal half of LUC fused to PpCOP1) exhibited strong luciferase fluorescence intensity. Conversely, no LUC activity was detected in negative control combinations (Fig. 5B). In the Co-IP assay, immunoprecipitation using an anti-MYC antibody revealed a detectable band for PpMADS2 in N. benthamiana leaves co-expressing PpCOP1-MYC and PpMADS2-FLAG (Fig. 5C). However, this band was not observed with “PpCOP1-MYC + FLAG” and “MYC + PpMADS2-FLAG.”

Subsequently, we confirmed the interaction in vitro by a GST pull-down assay (Fig. 5D). In this assay, the elution complex of PpCOP1-GST contained detectable levels of the MBP-tagged form of PpMADS2 protein, while no MBP-tagged protein was detected in the control elution complex. Overall, these data provide evidence for both in vitro and in vivo interactions between PpMADS2 and PpCOP1.

PpMADS2 is targeted by PpCOP1 and degraded by the 26S proteasome

To unequivocally establish PpCOP1 as a functional ortholog of COP1, the interactions between PpCOP1 and both PpUVR8 and PpHY5 were validated using Y2H system. As illustrated in Supplementary Fig. S7, PpCOP1 interacts with PpUVR8 and PpHY5 via the WD40 domain under UV-B irradiation, which is consistent with findings in Arabidopsis (Holm et al. 2001; Lin et al. 2020; Wang et al. 2022; Zhang et al. 2023). Given the demonstrated E3 ubiquitin ligase activity of PpCOP1 (Supplementary Figs. S8 and S9) and its interaction with PpMADS2 (Fig. 5), we conducted an in vitro ubiquitination assay using purified recombinant PpCOP1-MBP along with PpMADS2-GST (Supplementary Fig. S8). Upon co-incubation of PpCOP1-MBP and PpMADS2-GST in the presence of ubiquitin, an E1 ubiquitin-activating enzyme, and an E2 ubiquitin-conjugating enzyme, we observed the appearance of higher molecular weight bands exclusively in the sample containing both proteins, indicating successful ubiquitination of PpMADS2-GST (Lane 4, Fig. 6A). However, no signal indicative of ubiquitination was detected when any one component was missing from the reaction mixture (Fig. 6A).

To validate the ubiquitin-mediated degradation effect of PpCOP1 on PpMADS2 in vivo, we co-expressed PpCOP1-MYC and PpMADS2-FLAG constructs in N. benthamiana leaves, and subjected N. benthamiana to UV-B irradiation to assess the abundance of PpMADS2-FLAG protein. As shown in Fig. 6B, the protein content of PpMADS2 exhibited a significant reduction following UV-B treatment. Conversely, no substantial decrease in PpMADS2 protein content was observed when N. benthamiana proteins were injected with the proteasome inhibitor MG132 compared with the control group. However, these alternations were not evident under the controlled dark conditions (Fig. 6B). These findings demonstrate that UV-B-induced inhibition of PpMADS2 protein is mediated through ubiquitination facilitated by the 26S proteasome.

To confirm the role of PpCOP1 in regulating peach linalool synthesis under UV-B, transient overexpression of PpCOP1 was performed in callus of peach fruit. Our results demonstrate that under UV-B irradiation PpMADS2 protein in PpCOP1-overexpressing peach fruit callus was significantly reduced compared with the control (Fig. 6C). This result was not observed under dark conditions. Consistent with this observation, linalool exhibited a similar trend, with significantly lower levels detected in peach fruit callus overexpressing PpCOP1 under UV-B irradiation compared with the control (Fig. 6, D and E). Notably, these results align with those obtained from peach fruit, where UV-B treatment led to a significant decrease in linalool accumulation relative to dark conditions (Fig. 2A).

Considering the challenges associated with stable genetic transformation (overexpression or knockout) in peach, a perennial woody fruit tree, we conducted an analysis of natural variation in gene expression and volatile contents among 165 peach accessions based on transcriptome and metabolomic data (Fig. 6F; Supplementary Table S15) (Cao et al. 2024). The correlation between transcript level and linalool content was examined using linear regression analysis. Our results revealed a positive correlation between PpMADS2 expression and PpTPS1 expression (R = 0.62, P < 0.001), while a negative correlation was observed with PpCOP1 expression (R = −0.58, P < 0.001). Furthermore, both PpTPS1 (R = 0.60, P < 0.001) and PpMADS2 (R = 0.31, P < 0.001) showed positive correlations with linalool content across the different peach varieties. Conversely, the PpCOP1 transcript exhibited a significant negative correlation with linalool content (R = −0.31, P < 0.001). Collectively, these results suggest that PpMADS2 is a promising TF involved in regulating linalool biosynthesis in peach fruit. However, it appears that an E3 ubiquitin ligase gene known as PpCOP1 plays a suppressive role in linalool accumulation.

Discussion

Through the integration of proteomic, transcriptomic, and metabolomic data, the present study uncovers the regulatory role of PpMADS2 in the production of the monoterpene linalool (C₁₀) in fruit. PpMADS2 directly binds to the promoter of PpTPS1 and activates its expression (Fig. 3). Overexpression of PpMADS2 resulted in a significant increase in PpTPS1 transcript levels in peach fruit. The positive regulation of linalool synthesis by PpMADS2 was further confirmed using a stable transgene overexpression system driven by the fruit-specific E8 promoter in tomatoes (Fig. 3D). Moreover, the expression level of PpMADS2 showed a positive correlation with linalool content across 165 peach accessions (Fig. 6F). Therefore, this study provides insight into the involvement of a MADS TF in regulating monoterpene production in fruit. Although TF PpbHLH1 and PpERF61 have been reported to regulate linalool production in peach fruit (Wei et al. 2021, 2022), their protein contents were under detection levels in this study. In contrast, UV-B irradiation significantly decreased the abundance of PpMADS2 on the basis of proteome data and protein gel blot results (Fig. 4, A to C). Hence, UV-B irradiation inhibits linalool production by reducing the content of the PpMADS2 protein.

MADS TFs have been demonstrated to regulate terpenoid production, including the tetraterpenoid (C₄₀) carotenoids and sesquiterpene (C₁₅) artemisinin (Chen et al. 2022b; Liang and Li 2023). For instance, SlRIN, SlCMB1, and SlTAGL1 in tomato, CsMADS3/5/6 in citrus and AcMADS32 in kiwifruit have been identified as key positive regulators of fruit carotenoid accumulation by modulating the activity of crucial enzymes, including PSY and LCYb (Liang and Li 2023). Meanwhile, the negative regulatory role of MADS TFs on the production of carotenoids has also been reported for tomato SlMADS1, SlFYFL and SlMBP8, apple MdMADS6. AaSEP4, a MADS-box TF, has been found to directly promote artemisinin synthesis and accumulation by binding to the promoter of AaGSW1, an essential WRKY TF (Chen et al. 2022b). However, the role of the MADS-TF family in regulating the synthesis of volatile terpenoids is still poorly understood and primarily focuses on carotenoid-derived volatiles catalyzed by CCD action (Li et al. 2020; Meng et al. 2020; Liang and Li 2023). In RIN-CRISPR tomato fruits, there was a reduction observed in the levels of 6-methyl-5-hepten-2-one derived from carotenoid (Li et al. 2020). Grape VvMADS4 negatively regulates biosynthesis of carotenoid-derived norisoprenoids through direct binding to the promoter of VvCCD4b (Meng et al. 2020). The findings in this study enhances our understanding of the regulatory role of a MADS TF in the biosynthesis of volatile terpenoid by activating TPS expression. Collectively, MADS TFs play pivotal roles in governing terpene formation in plants.

This study and previous studies have reported a decrease in the flavor-related linalool following exposure to UV-B in fruit (Liu et al. 2017; Wei et al. 2021). However, the mechanism by which the light signaling pathway connects with the terpenoid synthesis pathway remains unclear. Our findings indicate that UV-B regulates the synthesis of a terpenoid through COP1, rather than UVR8 and HY5. We observed an increase in PpCOP1 protein in response to UV-B, leading to increased ubiquitin-mediated degradation of the PpMADS2. This suppression of PpMADS2 protein results in reduced levels of terpene synthase PpTPS1 transcript, ultimately leading to the reduction of linalool in fruit. UV-B-induced 26S-proteasome-mediated degradation of PpMADS2 largely depends on PpCOP1 (Fig. 7).

COP1 functions as an E3 ubiquitin ligase in plants and regulates various aspects of plant growth, development, and metabolism. It polyubiquitinates and facilitates the proteasome-mediated degradation of numerous substrates, including TFs such as MYB, bHLH, and bZIP (Han et al. 2020; Ponnu and Hoecker 2021). In darkness, COP1 mainly achieves dark morphogenesis by targeting core photomorphogenic proteins like HY5 within the nucleus and promoting their ubiquitin-mediated degradation (Han et al. 2020). Under dark conditions, the degradation of PpMYB10 and PpbHLH64 by COP1 in pear and MdMYB1 by COP1 in apple results in a blockage of anthocyanin synthesis and prevents red coloration of the peel (Li et al. 2012; Tao et al. 2020). AaMYB108 interacts with AcCOP1 to regulate artemisinin synthesis following light exposure in Artemisia annua (Liu et al. 2023a). Currently, MADS TFs have not been reported as substrates for ubiquitination or show interactions with other light signaling elements mediated by COP1. In this study, we discovered that a linalool regulator called PpMADS2 interacts with PpCOP1 but not previously reported PpbHLH1 or PpERF61 (Supplementary Fig. S5). These findings suggest that PpMADS2 serves as a crucial link between light signals and downstream metabolites, providing preliminary insights into why UV-B exposure leads to loss of fruit flavor compounds.

In this study, we observed that the ubiquitination-mediated degradation of PpMADS2 by PpCOP1, an E3 ubiquitin ligase induced by UV-B irradiation, was a significant factor contributing to the reduction of PpMADS2 protein and decrease in linalool content (Fig. 7). Transcriptome data and RT-qPCR results revealed that PpMADS2 transcript level was significantly reduced by UV-B (Supplementary Fig. S10). This indicates that the decrease in PpMADS2 protein may result from both reduced transcript levels and PpCOP1-mediated ubiquitination degradation. Further investigation into the regulatory mechanism of PpMADS2 transcript reductions is interesting. It is important to note that the present study focused on postharvest UV-B treatment of detached peach fruits, which differs from the conventional UV-B signaling mechanisms in living plants. In Arabidopsis, UV-B-induced photoactivation of UVR8 leads to its high-affinity binding with COP1, thereby inhibiting COP1 interaction with its substrates (Lau et al. 2019). Active UVR8 directly competes for the COP1 substrate-binding site, effectively blocking COP1 activity (Podolec et al. 2021). Conversely, our findings indicate that UV-B activates PpCOP1, promoting the degradation of PpMADS2, and consequently reducing volatiles in detached fruits.

Considering the conserved nature of COP1 as a core element in the light signaling pathway (Hoecker 2017; Podolec and Ulm 2018; Yadav et al. 2020), it is speculated that COP1 may serve as a potential regulator for alterations in MEP pathway metabolites induced by light wavelengths other than UV-B. For example, blue light triggers AmMYB24 binding to the AmOCS promoter, leading to increased emission of the monoterpene ocimene in snapdragon flower. However, the specific role of COP1 remains unclear (Han et al. 2022). Furthermore, monoterpenes play a crucial role in determining the flavor of Muscat grapes. Postharvest irradiation with blue and red-blue light has been shown to increase the content of various monoterpenes such as linalool, limonene, α-terpineol, dihydrolinalool, terpinolene, geraniol, p-cymene, and E-α-ocimene (Li et al. 2023). These results suggest that different visible light regimes exert distinct regulatory effects on the MEP pathway, and highlight the need for further exploration into the involvement of COP1.

In summary, we have identified significant alterations in content of multiple metabolites following UV-B exposure in fruit. The reduction of flavor-associated linalool is regulated by PpMADS2, which interacts with light signaling to modulate the MEP pathway. PpCOP1 facilitates the ubiquitination of PpMADS2, leading to a decrease in linalool synthesis. The degradation of PpMADS2 induced by UV-B largely depends on 26S-proteasome-mediated process and requires the involvement of PpCOP1. Thus, PpMADS2 integrates UV-B signaling by interacting with PpCOP1 to regulate linalool biosynthesis in peach. This study expands our understanding of the intricate interplay between light signaling and flavor quality. Multiomics approaches including metabolomics, proteomics, and transcriptomics provide valuable resources for investigating the mechanism underlying how light affects metabolites in fruit crops.

Materials and methods

Plant materials and treatment

Peach (Prunus persica L. Batsch cv. Hujingmilu) fruits were obtained from the Melting Peach Research Institute in Fenghua, Zhejiang Province, China. Under controlled experimental conditions, UV-B irradiation treatment was conducted as described in our previous study (Liu et al. 2017). Peach fruits at commercial harvest maturity (108 d after bloom) were picked from the tree on the same day, and samples that were more consistent in appearance and free of mechanical injury or pests and diseases were selected for subsequent experimental treatments. After harvest, the detached ripe peach fruits were divided into 2 groups. One group was exposed to UV-B (280 to 315 nm) radiation at 150 μW cm⁻² provided by UV-B lamp tubes (Luzchem Research Inc., Ontario, Canada) for 6 and 48 h. The control group was covered with aluminum foil and placed next to fruits exposed to UV-B radiation in climate chambers with a constant temperature of 20 °C, relative humidity (RH) of 90% to 96%, without any natural light. Each sampling time has 3 biological replicates for the control and treatment groups, with 5 fruits in each replicate. Slices of peel tissue (∼1 mm thick) were separated and sampled for subsequent proteomic and metabolomic analyses. Peach leaves, flowers, and fruit at different developmental stages were also collected as previously described previously, with 3 biological replicates of 20 leaves, 20 flowers, and 5 fruits per replicate, respectively (Wu et al. 2017). One hundred sixty-five accessions of peach fruit were collected as previously described with 3 biological replicates and 5 fruits in each replicate (Cao et al. 2024). Peach samples were then promptly frozen in liquid nitrogen and stored at −80 °C for further analysis.

Quantitative proteomics analysis of fruits

Protein extraction from fruit tissues and tryptic digestion were conducted according to methods described previously (Li et al. 2024) unless otherwise specified. Briefly, proteins were extracted from both control and UV-B group samples, each with 3 biological replicates, using the modified phenol-methanol method (Deng et al. 2007). The extracted proteins were digested with trypsin and labeled using TMT10plex reagents (Thermo Fisher, Waltham, USA). Two sets of TMT experiments were performed: one for the 6 h untreated and treated samples and one for the 48 h untreated and treated samples. In these experiments, control samples were labeled with the TMT10plex reagents 126, 127N, and 128C, while UV-B treated samples were labeled with the reagents 129N, 130N, and 13 °C. Additionally, 2 internal references were pooled from all 12 samples and labeled with the reagents 127C and 128N. The multiplexed TMT-labeled samples were combined and separated using a Waters Acquity BEH C18 1.7 μm, 2.1 × 100 mm column on an H-class UPLC system (Waters, Milford, MA, USA) at a flow rate of 300 μL min⁻¹. A total of 24 fractions were collected, which were then combined into 12 fractions and vacuum-dried for LC-MS/MS analysis.

Peptide samples were analyzed on an Ultimate 3000 nano UHPLC system (Thermo Scientific, USA) coupled online to a Q Exactive HF mass spectrometer (Thermo Scientific). A total of 1.0 μg of the peptide sample was separated using a binary solvent system consisting of 0.1% formic acid (solvent A) and 80% acetonitrile with 0.1% formic acid (solvent B). Peptides were eluted into the mass spectrometer at a flow rate of 300 nL min⁻¹, following a gradient starting from 5% to 7% solvent B within 2 min, increasing to 20% solvent B within 66 min, to 40% solvent B within 33 min, then from 40% to 90% solvent B within 4 min. The oven column temperature was set at 60 °C. The full scan was performed between 350 and 1,650 m z⁻¹, with a resolution of 120,000, and the automatic gain control target at 3 × 10⁶. The MS/MS scan was operated with HCD in top 12 mode using the following settings: resolution 45,000; automatic gain control target at 1 × 10⁵; normalized collision energy at 32%; isolation window of 1.2 m z⁻¹; dynamic exclusion 30 s.

The data-dependent acquisition raw instrument files were processed using the Proteome Discoverer (version 2.4.0.305) software package (Thermo Scientific) using the SequestHT node against the proteome of Prunus persica Whole Genome Assembly v2.0 & Annotation v2.1 (https://www.rosaceae.org/species/prunus_persica/genome_v2.0.a1, total 47,089 entries) with a FDR < 0.01 at the level of PSMs, peptides and proteins. The peptide precursor tolerance was set to 10 ppm, and fragment ion tolerance was set to 0.02 Da. Carbamidomethyl of cysteines and TMT6plex (+ 229.163 Da) on lysine residues and peptide N-termini were set as static modifications, oxidation of methionine residues, and acetylation of protein N-termini were set as dynamic modifications, and 2 missed cleavage trypsin sites were allowed. Normalization was applied to the grand total reporter ion intensity for each channel of the 8-plex experiment. An internal reference scaling (IRS) methodology, as described by Plubell et al. (2017), was used to correct for random MS2 sampling variations between the 2 TMT experiments using the internal references in each experiment. Further downstream analysis was performed in the R scripting and statistical environment, using the limma package from Bioconductor (http://www.bioconductor.org/). The basic statistical method used for significance analysis was the moderated t-statistic.

RNA library construction, sequencing, and bioinformatics analysis

Transcriptome analysis of UV-B treated peach fruits was performed with 3 biological replicates. RNA library construction and RNA-sequencing were performed by LC-Bio Technology Co., Ltd. (Hangzhou, China). The libraries were sequenced using the Illumina HiSeq2000 sequence platform. Transcriptome data were analyzed according to our previous studies (Cao et al. 2024). The raw reads of the transcriptome were quality controlled using fastp (1.21) and mapped to the Prunus persica (“Lovell,” https://phytozome-next.jgi.doe.gov/info/Ppersica_v2_1) genome with Hisat2 (2.0.5). The expression levels were quantified using StringTie (1.3.3b), which normalizes the expression levels of genes to FPKM (fragments per kilobase of transcript per million fragments mapped). The reads count of per gene was calculated using featureCounts (1.5.0-p3).

RNA extraction and gene expression analysis

Total RNA was extracted from frozen peach samples according to a previous protocol (Zhang et al. 2016). For reverse transcription quantitative PCR (RT-qPCR) analysis, PrimeScript RT reagent kit with gDNA Eraser (Takara, Beijing, China) was used to synthesize the first-strand cDNA, and then experiments were performed using a CFX96 instrument (Bio-Rad, Hercules, CA, USA) and Ssofast EvaGreen Supermix (Bio-Rad). Each RT-qPCR analysis contains 3 replicates. Oligonucleotide primers used for RT-qPCR analysis are listed in Supplementary Table S16.

Protein gel blotting

One gram of peach peel or peach callus was placed into BPP lysis solution (200 mm Tris-Base, 200 mm NaCl, 50 mm sodium ascorbate, 1% deoxycholate, 1% NP-40, 1% Triton X-100, 1% Tween-20, 5 mm EDTA, 30% sucrose, 50 mm tetraborate, 1% PVPP, 1× PMSF, 1× Protease inhibitor) and fully lysed for protein extraction. Then, 2 separate centrifugations were performed and equal volumes of Tris-saturated phenol and BPP lysate were mixed thoroughly with the supernatant, respectively, and the final supernatant crude protein extract was obtained. Approximately 5 volumes of saturated ammonium sulfate methanol solution were added and left to precipitate overnight at −20 °C. The next day, the precipitates were centrifuged twice and resuspended in methanol and acetone, respectively, and finally, the protein precipitates were dissolved in 0.5% SDS solution for further WB experiments. Modified BCA Protein Assay Kit (Sangon Biotech, Shanghai, China) was used for protein quantification.

Thirty μg of denatured proteins were electrophoresed by SDS-PAGE (200 V, 35 min), the proteins were transferred from the gel to a PVDF membrane (Immun-Blot PVDF Membrane, Roll, Bio-Rad) using an eBlotTM L1 rapid wet transfer instrument (Genscript, NJ, USA). Then, the membrane was blocked with 5% skim milk for 2 h and incubated overnight at 4 °C with the primary antibody. The next day, the membrane was washed with 1× TBST and incubated with a secondary antibody for 1 h at room temperature. The membrane was washed again with 1× TBST and detected by the FDbio-Femto ECL Kit (Fude Biological Technology Co., Ltd., Hangzhou, China) according to the manufacturer's instructions. The primary antibodies used in this assay were a plant universal internal reference Actin Mouse mAb for PLANTs (1:2,000 dilutions, Abmart, Shanghai, China) antibody, and the affinity-purified rabbit polyclonal antibody to PpMADS2 protein (1:1,000 dilutions). The specific polyclonal antibody against PpMADS2 was prepared with peptide antigen, and the peptide sequence is c- NKALRRKLEETSGQAPPLLAWEAAGHGNNNVQHTGLPHHPHSQGFFHPLGNNSTSQIGYTPLGSDHHEQMNVGNHGQ, Abmart. The secondary antibody used was Goat Anti-Rabbit&Mouse IgG-HRP (1:10,000, M21003, Abmart).

Volatile analysis by gas chromatography–mass spectrometry

Volatile analysis followed methodologies detailed in our previous studies (Zhang et al. 2016; Wei et al. 2021). Peach pericarp (∼1 mm thick) treated with UV-B for 6 and 48 h was sampled for volatile analysis, with the control and treatment groups containing 3 biological replicates of 5 fruits each, respectively. For 165 accessions of peach fruit, the pulp of the middle part of the fruit near the pericarp was sampled for volatile analysis after peeling. Three biological replicates each contained 3 fruits. Three biological replicates of transient overexpression and control callus of peach fruit were directly sampled for volatile analysis. For tomato fruit, T₁ generation of transgenic and WT tomato fruits at Br + 7 d (red ripe stage) were sampled for volatile analysis. The 3 tomato plants of each line were 3 biological replicates, each containing 5 fruits. Frozen peach peel (1 g), peach fruit (5 g), callus of peach fruit (5 g), and tomato fruit (5 g) were ground into powder under liquid nitrogen. The resulting powders were transferred to vials containing a solution of 200 mm ethylenediaminetetraacetic acid (EDTA) and 20% CaCl₂. Prior to sealing, 30 μL of 2-octanol (0.8 mg mL⁻¹) was added as an internal standard. Volatile compounds were collected using a fiber coated with 65 μm polydimethylsiloxane and divinylbenzene (PDMS-DVB) (Supelco Co., Bellefonte, PA). The vials were loaded into a solid-phase microextraction (SPME) autosampler (Combi PAL, CTC Analytics, Agilent Technologies), connected to an Agilent 7890N gas chromatograph and an Agilent 5975C mass spectrometer. Volatiles were separated on a DB-WAX column (30 m × 0.25 mm i.d. × 0.25 μm film thickness; J & W Scientific, Folsom, CA) with helium as the carrier gas at 1.0 mL min⁻¹. The temperature program began at 40 °C, ramped at 3 °C min⁻¹ to 100 °C, and then increased to 245 °C at 5 °C min⁻¹. Electron ionization at 70 eV ionized the column effluent, with a transfer temperature of 250 °C and a source temperature of 230 °C. Identification of volatile compounds relied on comparing their electron ionization mass spectra with those in the NIST Mass Spectral Library (NIST-08) and authentic standards based on retention times. Quantification involved referencing the peak area of the internal standard in the total ion chromatogram.

Fatty acid extraction and analysis

Fatty acids were extracted and analyzed by GC as previously described (Jin et al. 2022). One gram of the peach peel treated with UV-B for 6 and 48 h was taken and fully ground in liquid nitrogen. Frozen fruit tissue powder was added into a solution containing 15 mL of n-hexane:isopropanol (3:2, v/v) and 7.5 mL of 6.7% Na₂SO₄, then vortexed for 2 min and mixed, and centrifuged at 10,000 ×g for 10 min at 4 °C. After taking the supernatant, the lower solution was re-extracted by adding n-hexane: isopropanol solution and the 2 supernatants were combined. After vacuum rotary evaporation until about 1 mL of solution remained, 100 μL of exogenous heptadecanoic acid (C17:0, 8 μg μL⁻¹) was added as an internal standard. After drying with nitrogen, the supernatant was converted into fatty acid methyl esters (FAME) after adding 3 mL methanol:toluene:H₂SO₄ (88:10:2, v/v/v). After vortexing and mixing the sample, it was transferred to a 4 mL glass vial and methyl esterified in a water bath at 80 °C for 1 h. Then, 1 mL of heptane solution was added, vortexed briefly to mix, and allowed to stand for sufficient time to stratify. The supernatant was aspirated and added to ∼0.5 g of anhydrous Na₂SO₄, and 100 μL of the supernatant was taken as the sample to be measured by GC after removing the excess water. An Agilent 6890N gas chromatograph equipped with a flame ionization detector and a DB-Wax column (0.25 mm, 30 m, 0.25 μm, J & W Scientific) was used for fatty acid identification. The temperature of both the injector and detector was 230 °C. The oven temperature was raised from 50 °C to 200 °C at 25 °C min⁻¹ and then to 230 °C at 3 °C min⁻¹. The carrier gas was nitrogen with 1 mL min⁻¹. Quantitative measurement of fatty acids was calculated based on peak area of internal standard C17:0 with known content.

Heterologous expression in Escherichia coli and recombinant protein purification

Recombinant protein purification was performed following the method described previously (Wei et al. 2021). The full-length cDNAs of the PpMADS2 and PpCOP1 genes were inserted into the pGEX-4T-1 (GST-tag) and pMAL-c2x (MBP-tag) expression vectors using the primers in Supplementary Table S16. After sequence validation, recombinant vectors were transformed into E. coli BL21 (DE3) pLysS (Promega, Madison, WI, USA). The recombinant proteins were induced with 0.5 mm isopropyl-β-D-thiogalactopyranoside (IPTG) at 16 °C for 20 h and purified according to GST-tag Protein Purification Kit (Beyotime, Shanghai, China) manufacturer's instructions. MBPSep Dextrin Agarose Resin 6FF (Yeasen, Shanghai, China) was used to purify the MBP-tag proteins. SDS-PAGE was performed and the protein was visualized by Coomassie brilliant blue. pGEX-4T-1 and pMAL-c2x empty vectors as negative control.

Dual-luciferase assays

Following the protocol outlined by Wei et al. (2021), full-length cDNAs of TFs were cloned into the pGreen II 0029 62-SK vector, while the promoter of PpTPS1 was cloned into the pGreen II 0800-LUC vector using primers listed in Supplementary Table S16. The Renilla luciferase gene (REN), driven by a 35S promoter in the LUC vector, served as a positive control for estimating transient expression levels. These constructs were transformed into Agrobacterium tumefaciens GV3101::pSoup via the Gene Pulser Xcell Electroporation System (Bio-Rad). Agrobacterium cultures were prepared using an infiltration buffer (10 mm MES, 10 mm MgCl₂, 150 mm acetosyringone, pH 5.6) and adjusted to an OD₆₀₀ of 0.75. To assess the activity of specific TFs on the PpTPS1 promoter, Agrobacterium culture mixtures consisting of 1 mL of TFs and 100 μL of the PpTPS1 promoter construct were used for transient expression in N. benthamiana leaves. N. benthamiana plants were grown in a climate chamber (16 h:8 h, 25 °C, light:dark) and 4-wk-old seedlings were used for experiments. The ratio of firefly luciferase (LUC) to Renilla luciferase (REN) activities was measured 3 d post-infiltration using a Modulus Luminometer (Promega, Madison, WI, USA). Enzyme activities of LUC and REN were assayed using dual-luciferase assay reagents (Promega). The LUC/REN value for the empty vector SK on the promoter was set to 1 as a calibration. For each TF-promoter interaction, at least 3 independent experiments were performed. Six injected areas of 3 leaves on N. benthamiana were tested in each experiment for both the control and experimental groups.

Electrophoretic mobility shift assay

EMSA was conducted using the LightShift Chemiluminescent EMSA kit (ThermoFisher Scientific) according to the manufacturer's protocol. The details of the EMSA are provided in Wei et al. (2021). Double-stranded probes containing the CArG-binding site were made by annealing separately synthesized strands, with 3′biotin labeling. Unlabeled DNA fragment was used as a competitor, while biotin-labeled probes targeting mutation-specific CArG-binding sites were used in the assay. The probes used for EMSA are listed in Supplementary Table S16. Reaction products were subjected to SDS-PAGE electrophoresis in 0.5× Tris-borate/EDTA (0.5× TBE) buffer at 100 V. Subsequently, DNA fragments were transferred from the gel to a positively charged nylon membrane (Immun-Blot PVDF Membrane, Roll, Bio-Rad) with 0.5× TBE at 380 mA for 30 min at 4 °C. After half an hour of UV cross-linking, the membrane was tested with the chemiluminescent nucleic acid detection module kit (Thermo) according to the manufacturer's instructions.

Gene transient overexpression in peach fruit

Transient overexpression in “Hujingmilu” peach fruit was conducted as described by Liu et al. (2017) to verify gene function in vivo. The empty SK and PpMADS2-SK constructs were electroporated into Agrobacterium GV3101 and cultured at 28 °C until the OD₆₀₀ reached 0.8 to 1.0. The infiltration buffer used was the same as in the dual-luciferase assay method described above. After disinfecting the surface of the fruit, 2 flesh cubes (1 cm thick) were cut from opposite sides of each fruit and halved. Agrobacterium cultures carrying the PpMADS2 construct or the empty construct were infiltrated into the respective halves under a −70 kPa vacuum. The vacuum was gradually released to facilitate bacterial penetration into the flesh tissue. Post-infiltration, the flesh cubes were rinsed 3 times with sterile water and cultured on MS medium in a growth chamber at 20 °C with 85% RH for 3 d. The flesh cubes were then collected for GC−MS analysis. The transient expression treatments were repeated 3 times, with each experiment involving 5 fruits.

Gene transient overexpression in callus of peach fruit

To create PpCOP1-FLAG construct, the CDS sequence of PpCOP1 with the stop codon removed was cloned into pCAMBIA1300-221-UBQ-3flag vector. Then, FLAG and PpCOP1-FLAG constructs were electroporated into Agrobacterium strain GV3101 and cultured at 28 °C until the OD₆₀₀ reached 0.9 to 1.1. After centrifugation, Agrobacterium cultures were resuspended with SH liquid medium (Schenk & Hildebrandt Medium, PVP-K30, sucrose, agar, 0.5 mg L⁻¹ TDZ, 0.5 mg L⁻¹ 2,4-D, pH = 5.7 to 5.9) to OD₆₀₀ = 0.85, and incubated at 28 ℃ for 1 h on a shaker. Callus of “Hujingmilu” peach fruit were immersed in the infestation solution for 10 min, and were shaken upside down every 3 min. The bacterial liquid on the surface was blotted out with filter paper and transferred to SH medium lined with filter paper. After 3 d of cultivation, they were transferred to SH medium containing antibiotics for further incubation for 3 d. Then, control and PpCOP1 overexpressing callus were divided into 2 groups. One group was exposed to UV-B radiation of 40 μw cm⁻² for 24 h, and the other group was left in a dark environment for 24 h. It is processed in climate chambers with a constant temperature of 20 °C, RH of 90% to 96%. Peach callus were sampled and ground in liquid nitrogen. After proteins extraction and quantification by the same method described in Protein gel blotting, the PpCOP1 protein was detected by immunoblotting with anti-FLAG antibody by protein gel blot. The PpMADS2 protein was detected by anti-PpMADS2 polyclonal rabbit antibody. Actin Mouse mAb for PLANTs antibody was used as the loading control.

Stable heterologous overexpression in tomato fruit

Using the primers listed in Supplementary Table S16, the full-length cDNA of PpMADS2 was first inserted into the pDONR207 vector via the Gateway ligation system (11789-020, 11791-020, Invitrogen, Thermo) and then cloned into the pBIN19-E8 vector driven by the E8 promoter. Agrobacterium-mediated tomato (cv Ailsa Craig) transformation was performed following the method described previously (Wang et al. 2005) with modifications. Transformed lines were screened on kanamycin (50 mg L⁻¹). The identified T₁ generation of transgenic and WT tomato plants were grown in a greenhouse (25 °C, 16 h light/8 h darkness). Tomato fruits at Br + 7 d (red ripe stage) were frozen in liquid nitrogen and stored at −80 °C for analysis. Three tomato plants from each line were selected as 3 biological replicates, each containing 5 fruits. Gene expression in transgenic tomato was analyzed by RT-qPCR with the primers listed in Supplementary Table S16.

Y2H assay

The Y2H assay was conducted based on a Matchmaker Gold Yeast Two-Hybrid System (Takara, Beijing, China). The full-length CDS of PpMADS2, PpbHLH1, PpERF61, PpHY5, PpUVR8 were cloned into the AD (pGADT7) vector. The full-length CDS of PpUVR8, PpCOP1, and domain fragments (N282, N-Ring, N-Coil, and WD40) were cloned into the pGBKT7 (BD) vector. After an auto-activation analysis, the gene-AD and gene-BD recombinant plasmids were inserted into Y2HGold cells, which were then added to synthetic dextrose medium (SD) medium lacking leucine (Leu) and tryptophan (Trp) (DDO; SD/-Leu/-Trp). The potential physical interactions between proteins were evaluated by screening the yeast transformants on QDO (Sd medium lacking Leu, Trp, adenine [Ade] and histidine [His]; Sd/-Leu/-Trp/-Ade/-His) and QDO/X/A (Sd/-Leu/-Trp/-Ade/-His supplemented with 40 μg mL⁻¹ X-α-gal and 200 ng mL⁻¹ aureobasidin A [AbA]). The plates were photographed after incubation at 30 °C for 5 d. All transformations and screenings were performed at least 3 times. Primers used in this assay are listed in Supplementary Table S16.

Firefly LCI assay

Agrobacterium strain GV3101 carrying the constructs of the PpMADS2-pCambia1300-nLUC and PpCOP1-pCambia1300-cLUC were mixed and infiltrated into N. benthamiana leaves. The bacteria were resuspended in the infiltration buffer to a final OD₆₀₀ of 0.5. The LUC activity was determined 2 d after infiltration. Before imaging, 0.2 mm luciferin (Yeasen, Shanghai, China) was injected to the position infiltrated by A. tumefaciens and held for 15 min in the dark. The LUC imaging was performed using the NightSHADE LB 985 system. Empty vectors expressing cLUC or nLUC were co-transformed as the negative controls. The primers used for vector constructions are listed in Supplementary Table S16.

Co-immunoprecipitation

To create PpCOP1-MYC and PpMADS2-FLAG constructs, the CDS sequences of PpCOP1 and PpMADS2 with the stop codon removed were cloned into pCAMBIA1300-221-UBQ-4myc and pCAMBIA1300-221-UBQ-3flag vectors, respectively. Then, the resulting constructs were introduced into Agrobacterium strain GV3101 and infiltrated into the abaxial side of N. benthamiana leaves. For proteasome inhibition, leaves were infiltrated with 50 μM MG132 (Yeasen, Shanghai, China) solution for 12 h before sample collection. 48 h after infiltration, N. benthamiana leaves were ground in liquid nitrogen.

Proteins were extracted using Plant Protein Extraction Kit including protease inhibitor cocktail (Fude Biological Technology Co., Ltd.) according to the manufacturer's instructions. One portion of the protein obtained was used as input, and the other portion was incubated with Anti-MYC Tag mAb Magnetic conjugated Beads (Abmart) at 4 °C overnight to capture the MYC-tagged protein. The second day, the beads were separated with a magnetic separation rack and washed with dilution buffer. The beads were then resuspended with 1× Protein Loading Buffer and boiled at 95 °C for 10 min to separate the immunocomplexes from the magnetic beads, and the supernatant obtained was subjected to immunoblotting analysis using anti-MYC antibody (Abmart) and anti-FLAG antibody (Abmart) for Co-IP assay. The primers used for vector constructions are listed in Supplementary Table S16.

Pull-down assay

The purified PpCOP1-GST and PpMADS2-MBP protein were subjected to pull-down assay using Pierce GST Protein Interaction Pull-Down Kit (Thermo Scientific) according to the manufacturer's instructions with minor modifications. Briefly, GST and PpCOP1-GST proteins were incubated with glutathione agarose for 2 to 3 h at 4 °C, respectively. After incubation, the agarose was washed and PpMADS2-MBP protein was added for further incubation. “GST + PpMADS2-MBP” and “PpCOP1-GST + PpMADS2-MBP” proteins were incubated overnight at 4 °C to fully bind. “GST + PpMADS2-MBP” group was used as a control. The second day, the agarose was washed with wash solution and Bait-Prey protein complex was eluted with 10 mm glutathione elution buffer. The samples were subjected to a western blot involving anti-GST antibody (Abmart) and anti-MBP antibody (Abmart). The primers used for vector constructions are listed in Supplementary Table S16.

In vitro ubiquitination assay

Recombinant PpCOP1-MBP and PpMADS2-GST proteins were obtained as described in “Heterologous expression in E. coli and purification” section. In vitro ubiquitination assay was performed as described by previously (Wei et al. 2023). In brief, reactions (30 mL) containing 50 mm Tris-HCl (pH 7.5), 10 mm MgCl₂, 0.05 mm ZnCl₂, 1 mm ATP (Sigma-Aldrich), 0.2 mm dithiothreitol, 10 mm phosphocreatine, 0.1 unit of creatine kinase (Sigma-Aldrich), 50 ng of human E1 (Boston Biochem), 250 ng of human E2 (Boston Biochem), 2 mg of ubiquitin (Boston Biochem), 500 ng of PpCOP1-MBP, and 500 ng of PpMADS2-GST were incubated at 30 °C for 2 h. Reactions were stopped by adding sample buffer and analyzed by SDS-PAGE followed by immunoblot analysis using anti-GST antibody (Sigma-Aldrich).

In vivo protein degradation assay

These 2 combinations of A. tumefaciens strain GV3101 (“MYC + PpMADS2-FLAG” and “PpCOP1-MYC + PpMADS2-FLAG”) were mixed in a 1:1 ratio, respectively, and were injected into the abaxial side of N. benthamiana leaves. One day later, these N. benthamiana were divided into 2 groups, one group was infiltrated with 50 μM of MG132 to inhibit the proteasome and the other group was infiltrated with a mixture of DMSO and water as a control. Then, half the amount of N. benthamiana was exposed to UV-B radiation of 40 μw cm⁻² for 24 h, and the other half was left in a dark environment for 24 h. It is processed in climate chambers with a constant temperature of 20 °C, RH of 90% to 96%. N. benthamiana leaves were sampled and ground in liquid nitrogen. After proteins extraction and quantification by the same method described in the Co-IP assay procedure, the content of PpMADS2 protein in the 4 groups of samples was detected by WB. Actin Mouse mAb for PLANTs (Abmart) antibody was used to detect the content of the internal reference protein and the anti-FLAG antibody (Abmart) was used to detect the content of the PpMADS2 protein.

Statistics

Figures were produced using GraphPad Prism 9.0 (GraphPad Software, San Diego, CA, USA). The 2-sample significance test was calculated using unpaired Student's t-test (*, P < 0.05, **, P < 0.01, and ***, P < 0.001) (SPSS 19.0, SPSS Inc., Chicago, IL, USA).

Data availability

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (https://proteomecentral.proteomexchange.org) via the iProX partner repository (Ma et al. 2019; Chen et al. 2022a) with the dataset identifier PXD056349. The raw RNA-seq data of peach under UV-B treatment have been deposited in the Genome Sequence Archive (Chen et al. 2021) in National Genomics Data Center (CNCB-NGDC Members and Partners 2024), China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (GSA: CRA016707) that are publicly accessible at https://ngdc.cncb.ac.cn/gsa. The RNA-seq raw data of different tissues and developmental stages of peach fruits can be found in the National Center for Biotechnology Information (NCBI) Short Read Archive database with accession no. PRJNA576753. The raw RNA-seq data of 165 accessions of peach fruit are publicly available in the Genome Sequence Archive (GSA: CRA015040) and GC-MS data for 165 accessions of peach fruit are publicly available in the OMIX, China National Center for Bioinformation/Beijing Institute of Genomics, Chinese Academy of Sciences (https://ngdc.cncb.ac.cn/omix: accession no. OMIX006918).

Accession numbers

Sequence data from this article can be found in the Phytozome 13.0 database (https://phytozome-next.jgi.doe.gov/info/Ppersica_v2_1) under accession numbers: PpMADS2 (Prupe.5G208400), PpCOP1 (Prupe.5G031300), PpTPS1 (Prupe.4G030400), PpUVR8 (Prupe.4G277200), PpHY5 (Prupe.1G478400), PpbHLH1 (Prupe.8G157500), and PpERF61 (Prupe.5G117800).

Acknowledgments

We thank Changqing Zhu for help with technical assistance for GC-MS, Rong Jin for help with plant care, and Harry Klee for language improvement.

Author contributions

B.Z. and C.W. conceived the research plans; C.W. and H.Y. performed most of the experiments and analyses with helps from B.Y., W.W., W.S., and J.C.; Z.D. conducted proteomic sequencing in this study; C.W. and B.Z. wrote the article; Z.D., B.Z., X.L., and K.C. reviewed and edited the manuscript. All authors have read and approved to the published version of the manuscript.

Supplementary data

The following materials are available in the online version of this article.

Supplementary Figure S1. The fatty acid content and abundance of related metabolic enzymes under UV-B treatment.

Supplementary Figure S2. The MVA pathway for farnesene biosynthesis and changes in the abundance of several enzymes of this pathway in response to UV-B.

Supplementary Figure S3. The levels of β-ionone and 7,10-dihydro-β-ionone and changes in PpCCD4 abundance under UV-B treatment.

Supplementary Figure S4. Heatmap showing PpMADS2 and PpTPS1 expression levels during peach fruit developmental and in different tissues.

Supplementary Figure S5. Y2H analyses of the protein–protein interactions between PpCOP1 and PpMADS2, PpbHLH1 or PpERF61, as well as between PpUVR8, PpHY5, and PpMADS2.

Supplementary Figure S6. Schematic representation of the PpCOP1 structural domains.

Supplementary Figure S7. Y2H analyses of the protein–protein interactions between PpCOP1 and PpUVR8 or PpHY5.

Supplementary Figure S8. SDS-PAGE analysis of the recombinant PpCOP1 and PpMADS2 proteins expressed in E. coli strain BL21.

Supplementary Figure S9. Self-ubiquitination of PpCOP1.

Supplementary Figure S10.  PpMADS2 transcript levels are significantly reduced by UV-B irradiation.

Supplementary Table S1. RAW intensity data of the proteomics analysis in peach under UV-B treatment.

Supplementary Table S2. Normalized intensity data of the proteomics analysis in peach under UV-B treatment.

Supplementary Table S3. DEPs in peach irradiated with UV-B for 6 h.

Supplementary Table S4. DEPs in peach irradiated with UV-B for 48 h.

Supplementary Table S5. GO analysis of the DEPs in peach irradiated with UV-B for 6 h.

Supplementary Table S6. GO analysis of the DEPs in peach irradiated with UV-B for 48 h.

Supplementary Table S7. KEGG analysis of the DEPs in peach irradiated with UV-B for 6 h.

Supplementary Table S8. KEGG analysis of the DEPs in peach irradiated with UV-B for 48 h.

Supplementary Table S9. Changes in the protein abundance of enzymes from several metabolic pathways in response to UV-B treatment.

Supplementary Table S10. Quality control data of the RNA-seq reads in peach under UV-B treatment.

Supplementary Table S11. Mapped data of the RNA-seq reads in peach under UV-B treatment.

Supplementary Table S12. FPKM of all genes in UV-B-treated peach fruit.

Supplementary Table S13. Transcript levels of MEP pathway genes in response to UV-B treatment presented in Fig. 2B.

Supplementary Table S14. List of the TFs in the 6 and 48 h DEP datasets.

Supplementary Table S15. Transcript levels of PpCOP1, PpMADS2, and PpTPS1 and linalool content in 165 peach fruit accessions.

Supplementary Table S16. Primers used in the present study.

Funding

This research was supported by Zhejiang Provincial Natural Science Foundation (LD22C150001), the National Natural Science Foundation of China (31972379 and 32302492), and the 111 Project (B17039).

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• SDS Gramene: AT1G14750

• SDS Araport: AT1G14750

• protein CHEBI: CHEBI:36080

• PSY Gramene: ostta05g03530

• PSY Araport: ostta05g03530

• COP1 Gramene: AT2G32950

• COP1 Araport: AT2G32950

• UVR8 Gramene: AT5G63860

• UVR8 Araport: AT5G63860

• HMBPP CHEBI: CHEBI:128753

References

Author notes