Influences of chemotype and parental genotype on metabolic fingerprints of tansy plants uncovered by predictive metabolomics

Abstract

Intraspecific plant chemodiversity shapes plant-environment interactions. Within species, chemotypes can be defined according to variation in dominant specialised metabolites belonging to certain classes. Different ecological functions could be assigned to these distinct chemotypes. However, the roles of other metabolic variation and the parental origin (or genotype) of the chemotypes remain poorly explored. Here, we first compared the capacity of terpenoid profiles and metabolic fingerprints to distinguish five chemotypes of common tansy (Tanacetum vulgare) and depict metabolic differences. Metabolic fingerprints captured higher variation in metabolites while preserving the ability to define chemotypes. These differences might influence plant performance and interactions with the environment. Next, to characterise the influence of the maternal origin on chemodiversity, we performed variation partitioning and generalised linear modelling. Our findings revealed that maternal origin was a higher source of chemical variation than chemotype. Predictive metabolomics unveiled 184 markers predicting maternal origin with 89% accuracy. These markers included, among others, phenolics, whose functions in plant-environment interactions are well established. Hence, these findings place parental genotype at the forefront of intraspecific chemodiversity. We recommend considering this factor when comparing the ecology of various chemotypes. Additionally, the combined inclusion of inherited variation in main terpenoids and other metabolites in computational models may help connect chemodiversity and evolutionary principles.

Figure 1

Simplified scheme of the analytical workflow.

Figure 2

Identification of tansy chemotypes via GC–MS and LC–MS analyses. (a) Principal component analysis biplot illustrating the discrimination of tansy chemotypes using GC–MS data (52 features). (b) Representation of the levels of the 39 significant GC–MS features (Tukey’s test, P < 0.05, FDR correction) differentiating chemotypes via clustering analysis (Pearson correlation, Ward algorithm). (c) Clustering analysis (Pearson correlation, Ward algorithm) of tansy chemotypes using 809 significant LCMS features (Tukey’s test, P < 0.05 after FDR correction). Keto: artemisia ketone chemotype, Bthu: β-thujone chemotype, ABThu: α-/β- thujone chemotype, Aacet: artemisyl acetate/artemisia ketone/artemisia alcohol chemotype, Myrox: (Z)-myroxide/santolina triene/artemisyl acetate chemotype.

Figure 3

Predictive metabolomics on tansy plants. (a) R² scores of the 500 generalised linear models using GC–MS or LC–MS features (Tukey’s test, P < 0.01). For each condition, 500 permuted datasets were created by randomly swapping chemotype classes. Feat: features, Mar: markers, Perm: permuted, Sig: significant. “GC–MS Sig feat” included 39 significant features, “GC–MS Best mar” contained 7 markers, “LC–MS Sig feat” contained 809 significant features, “LC–MS Best mar” included 39 markers. (b) Upset plot of the top 50 markers for each chemotype from LC–MS modelling. The absence of a dot means that the corresponding markers were not present among the best 50 markers in the corresponding chemotypes. (c) Putative annotation of the best LC–MS predictors. Keto: artemisia ketone chemotype, BThu: β-thujone chemotype, ABThu: α-/β-thujone chemotype, Aacet: artemisyl acetate/artemisia ketone/artemisia alcohol chemotype, Myrox: (Z)-myroxide/santolina triene/artemisyl acetate chemotype.

Figure 4

Effect of maternal genotype on intraspecific chemodiversity. (a) Variation partitioning on 52 GC–MS features (ANOVA test, P < 0.05). (b) Variation partitioning on 5,066 LC–MS features (ANOVA test, P < 0.05). (c) R² scores of the 500 generalised linear models using GC–MS or LC–MS features (Tukey’s test, P < 0.01). For each condition, 500 permuted datasets were developed. Feat: features, Perm: permuted. Sig: significant. “GC–MS Sig feat” included 41 significant features, “LC–MS Sig feat” contained 3,688 significant features and “LC–MS Top 5%” included the best 5% markers (184). (d) Putative annotation of the best LC–MS markers.