图1 冷冻温度胁迫下山桃表型（A）及生理指标分析（B、C）数据为3次独立试验的平均值±标准差。不同字母表示经单因素方差分析（ANOVA）及Duncan多重比较检验确定的组间差异显著性（P<0.05）。CK：5 ℃；T5：-5 ℃；T15：-15 ℃；T25：-25 ℃。下同

Fig. 1 Analysis of phenotypic (A) and physiological indices (B, C) of P. davidiana under freezing temperature stressData are means ± standard deviation of three independent experiments. Different letters indicate the significance of differences between groups as determined by one-way analysis of variance (ANOVA) and Duncan’s multiple comparison test (P<0.05). The same below

图2 山桃在不同冷冻胁迫下的差异积累代谢物分析A：不同温度下鉴定代谢物的主成分分析（PCA）；B：4种不同温度处理下所有差异代谢物（DEMs）的热图，颜色表示各差异代谢物的相对含量水平，红色表示高含量，绿色表示低含量；C：CK vs T5、CK vs T15和CK vs T25差异代谢物的火山图；D：所有差异代谢物根据变化趋势进行的聚类分析，详细聚类结果见图S2；E：CK vs T5、CK vs T15和CK vs T25差异代谢物的维恩图；F：CK vs T5、CK vs T15和CK vs T25 3组共有109个差异代谢物的热图，详细差异代谢物结果见表S5

Fig. 2 Analysis of differentially accumulated metabolites in P. davidiana under different freezing stressesA: Principal component analysis (PCA) of the identified metabolites at different temperatures. B: Heat maps of all differential metabolites (DEMs) at four different temperature treatments. Colours indicate the relative content level of each differential metabolite, with red indicating high content and green indicating low content. C: Volcano plots of CK vs T5, CK vs T15 and CK vs T25 differential metabolites. D: Cluster analysis of all differential metabolites according to the trend of changes, and the detailed clustering results are shown in Figure S2. E: Venn diagrams of CK vs T5, CK vs T15 and CK vs T25 differential metabolites. F: Heat maps of a total of 109 differential metabolites in the three groups of CK vs T5, CK vs T15 and CK vs T25, and the detailed results of differential metabolites are shown in Table S5

图3 差异表达代谢物（DEMs）的KEGG通路富集分析A-C：CK vs T5、CK vs T15和CK vs T25比较分析的KEGG 通路富集分析图；D-F：CK vs T5、CK vs T15和CK vs T25比较分析前10种差异表达代谢物

Fig. 3 KEGG pathway enrichment analysis of differentially expressed metabolites (DEMs)A-C: KEGG pathway enrichment analysis of the comparisons CK vs T5, CK vs T15, and CK vs T25. D-F: Top 10 differentially expressed metabolites in the comparing CK vs T5, CK vs T15, and CK vs T25

图4 差异表达基因（DEGs）表达模式及共表达基因的GO与KEGG富集分析A：DEGs上调和下调情况；B：不同处理组间DEGs表达Venn图；C：所有比较组中DEGs的共调控关系；D：510个共表达基因的前20位KEGG和GO富集分析结果；E：510个共表达DEGs被划分为9个聚类，每个聚类的基因数量显示在顶部，绿色线条表示各子聚类相对表达水平的平均值，橘色线条表示各子聚类中单个基因的相对表达水平；F：各聚类中显著富集的KEGG通路（P<0.05）

Fig. 4 Expression patterns of differentially expressed genes (DEGs) and GO and KEGG enrichment analyses of co-expressed genesA: Up- and down-regulation of DEGs. B: Venn map of DEGs expression among different treatment groups. C: Co-regulatory relationships of DEGs in all comparison groups. D: Results of KEGG and GO enrichment analyses of top 20 KEGGs and GOs for the 510 co-expressed genes. E: The 510 co-expressed DEGs were classified into 9 clusters. The number of genes in each cluster is shown at the top, the green line indicates the average of the relative expression in each subcluster, and the pink line indicates the relative expressions of individual genes in each subcluster. F: KEGG pathways significantly enriched in each cluster (P<0.05)

图5 RNA-Seq数据的RT-qPCR验证数据表示为3个独立生物学重复的平均值±标准误

Fig. 5 RT-qPCR validation of RNA-Seq dataData showed the means ± SE of three independent biological replicates

图6 基因WGCNA分析与生理性状的关联性A：基于WGCNA的层次聚类树分析，主分支划分为14个不同颜色模块；B：模块‒性状相关性热图，每行对应1个模块（颜色与A图一致），行列交叉处的颜色反映模块与生理性状的相关性，标有五角星的黑、绿、鲑红、红、棕褐及黄色模块与电解质渗漏率（EL）、丙二醛（MDA）及过氧化物酶活性（POD）显著相关（P<0.05）；C：上述6个模块中DEGs的温度响应趋势及KEGG通路富集分析（通过R语言pHYPER函数实现，P<0.05）

Fig. 6 Association of gene WGCNA analysis with physiological traitsA: Hierarchical clustering tree analysis based on WGCNA, with the main branch divided into 14 modules of different colours. B: Heatmap of module-trait correlation, each row corresponds to one module (colours are consistent with Figure A), and the colours at the intersections of the rows and columns reflect the correlation between the modules and the physiological traits, the black, green, salmon red, red, brownish-brown and yellow modules labelled with an asterisk are significantly (P<0.05) correlated with the electrolyte leakage rate (EL), malondialdehyde (MDA) and peroxidase activity (POD) are significantly correlated (P<0.05). C: Temperature response trend and KEGG pathway enrichment analysis of DEGs in the above six modules (achieved by R language pHYPER function, P<0.05)

图7 山桃响应不同强度冷冻胁迫的代谢组与转录组关联分析A：九象限图展示代谢物与基因的关联性，红色代表基因和代谢物均上调，绿色代表基因和代谢物均下调，黑色代表表示统计不显著的关联，蓝色表示基因与代谢物表达趋势相反；B：处理组（CK vs T5/T15/T25）与对照组间DAMs的相关系数聚类热图（PCCs>0.8）；C：CK vs T5/T15/T25 3组比较中DEMs前20位KEGG富集通路汇总分析

Fig. 7 Metabolome-transcriptome association analysis of P. davidiana in response to different intensities of freezing stressA: Nine-quadrant plot demonstrates metabolite-gene associations, red indicates that both genes and metabolites are upregulated, green indicates that both genes and metabolites are downregulated, black indicates a statistically insignificant association, and blue indicates that the expression trends of genes and metabolites are opposite. B: Clustered heatmap of correlation coefficients of DAMs (PCCs>0.8) between treatment (CK vs T5/T15/T25) and control groups. C: Top 20 DEMs in the comparison of the three groups CK vs T5/T15/T25 KEGG-enriched pathway summary analysis

图8 冷冻胁迫下山桃差异表达代谢物（DEMs）与差异表达基因（DEGs）参与的代谢通路A：淀粉与糖代谢；B：有机酸代谢；C；类黄酮合成。圆形表示受调控代谢物（绿色下调，红色上调），矩形表示受调控基因（绿色下调，红色上调）

Fig. 8 Metabolic pathways involved in differentially expressed metabolites (DEMs) and differentially expressed genes (DEGs) in P. davidiana under freezing stressA: Starch and sugar metabolism. B: Organic acid metabolism. C: Flavonoid synthesis. Circles indicate regulated metabolites (green down-regulation, red up-regulation), rectangles indicate regulated genes (green down-regulation, red up-regulation)