The diel course allocation toward primary and secondary metabolisms is distinct between ferns and angiosperms
Ferns and angiosperms have different evolutionary histories (Sussmilchet al. 2019), which may reflect in their metabolome. Furthermore, our previous results suggest that ferns and angiosperms have distinct allocation of the daily CO2 assimilated toward the synthesis of primary/secondary metabolites (Lima et al. 2019), which may have consequences for both A -g sand growth-stress tolerance trade-offs. We then next carried out a LC-MS-based metabolic fingerprinting analysis in leaf samples harvested throughout the diel course to better understand the metabolic differences between these plant groups. Metabolic fingerprinting is an untargeted metabolomics approach suitable to discriminate biological samples without metabolite identification, i.e. based in the intensity of the features (peaks found in the chromatograms) detected in the samples (Scholz et al. 2004; Kruger et al. 2008; Kosmideset al. 2013; Silveira-Sotelo et al. 2015; Perez de Souzaet al. 2019). Given that the LC-MS platform used here mostly detects secondary metabolites (Tohge & Fernie 2010; Perez de Souzaet al. 2021), metabolic fingerprinting analysis was next used to investigate how ferns and angiosperms differ with regard to their contents of secondary metabolites. Analysis of the chromatograms revealed an incredible metabolic diversity in ferns, as evidenced by the higher number of peaks solely detected in ferns. However a high intensity of peaks for a given compounds was found in both plant groups (Figures 4a-d). Similarly, several mass-to-charge ratio (m/z ) features obtained by MS analysis were solely found in ferns, especially at 5:00 h and 14:00 h (Figures 5a-d). A total of 19340, 7741, 9668 and 19598 features were statistically different (P < 0.05) between ferns and angiosperms at 5:00 h, 8:00 h, 14:00 h and 17:00 h, respectively. Several of these features had higher intensity in ferns, compared to angiosperms (Figures 6a-d). Hierarchical clustering analysis (HCA) (Figures 6a-d), orthogonal partial least squares discriminant analysis (orthoPLS-DA) (Figures 7a-d) or principal component analysis (PCA) (Figures S6a-d) of these features indicate that ferns and angiosperms differ substantially at the secondary metabolic level. Similarly, PCA using previously published GC-MS metabolite profiling data (Lima et al. 2019), which mostly refers to primary metabolites, also discriminate ferns and angiosperms (Figures S7a-d).
The results obtained here coupled to previous studies provide compelling evidence highlighting that fern stomata can respond to the circadian rhythm and close in response to sucrose, mannitol, dark, high VPD and high CO2 ­concentration (Franks & Britton-Harper 2016; Hõrak et al. 2017; Lima et al. 2019; Cardoso et al.2019; Gong et al. 2021; Plackett et al. 2021). However, no ABA response was observed in the ferns used here (Figure 8a). These results indicate that the slower ferns stomatal closure is associated to their limited capacity to respond to ABA and to mesophyll-derived metabolites, especially sucrose, when compared to angiosperms. This idea is supported by the fact that the faster high CO2-induced stomatal closure of angiosperms was associated to their higher capacity to produce sucrose (Figure 8b) (Limaet al. 2019), when compared to ferns. Furthermore, our metabolomics analyses suggest that ferns and angiosperms exhibit distinct patterns of allocation toward primary and secondary metabolisms throughout the diel course, which may affect the level of ROS in the guard cells of these species (Figure 8c) and ultimately have implications for the regulation of bothA -g s and growth-stress tolerance trade-offs (Figure 9).