Molecular genetic and biochemical evidence for adaptive evolution of leaf abaxial epicuticular wax crystals in the genus Lithocarpus (Fagaceae)

Abstract

Background: Leaf epicuticular wax is an important functional trait for physiological regulation and pathogen defense. This study tests how selective pressure may have forced the trait of leaf abaxial epicuticular wax crystals (LAEWC) and whether the presence/absence of LAEWC is associated with other ecophysiological traits. Scanning Electron Microscopy was conducted to check for LAEWC in different Lithocarpus species. Four wax biosynthesis related genes, including two wax backbone genes ECERIFERUM 1 (CER1) and CER3, one regulatory gene CER7 and one transport gene CER5, were cloned and sequenced. Ecophysiological measurements of secondary metabolites, photosynthesis, water usage efficiency, and nutrition indices were also determined. Evolutionary hypotheses of leaf wax character transition associated with the evolution of those ecophysiological traits as well as species evolution were tested by maximum likelihood.

Results: Eight of 14 studied Lithocarpus species have obvious LAEWC appearing with various types of trichomes. Measurements of ecophysiological traits show no direct correlations with the presence/absence of LAEWC. However, the content of phenolic acids is significantly associated with the gene evolution of the wax biosynthetic backbone gene CER1, which was detected to be positively selected when LAEWC was gained during the late-Miocene-to-Pliocene period.

Conclusions: Changes of landmass and vegetation type accelerated the diversification of tropical and subtropical forest trees and certain herbivores during the late Miocene. As phenolic acids were long thought to be associated with defense against herbivories, co-occurrence of LAEWC and phenolic acids may suggest that LAEWC might be an adaptive defensive mechanism in Lithocarpus.

Keywords: Adaptive evolution; Chemical defenses; Leaf epicuticular wax; Lithocarpus; Phylogenetic signal; Physical defenses.

Fig. 1

Species tree and hypotheses of positive selection on LAEWC trait-shift events. Three positive selection hypotheses are: (a) positive selection independently acted on lineages that gained LAEWC, (b) positive selection independently acted on lineages of lost LAEWC, and (c) positive selection acted on the ancestor of all LAEWC lineages and secondarily acted on the branch of loss LAEWC. The gray area indicates the species with LAEWC. The solid bars and hollow bars indicate the gain events and loss events of LAEWC, respectively. (d) The species tree of the studied Lithocarpus species reconstructed using six LAEWC unrelated genes. Bold branches indicate > 95% posterior probability supporting values for grouping. Species with LAEWC were marked in bold, which revealed non-monophyletic relationship of either LAEWC species or non-LAEWC species. Values near the nodes are the estimated splitting time (Mya) with 95% highest posterior density (gray bars). P and Q at the geological time scale axis are Pliocene and Quaternary, respectively

Fig. 2

Details of the abaxial layer of leaf epidermis of species of the genus Lithocarpus using Scanning Electron Microscope (SEM). a L. amygdalifolius; (b) L. brevicaudatus; (c) L. cornea; (d) L. dodonaeifolius; (e) L. formosanus; (f) L. glaber; (g) L. hanceii; (h) L. harlandii; (i) L. kawakamii; (j) L. konishii; (k) L. lepidocarpus; (l) L. nantoensis; (m) L. shinsuiensis; (n) L. taitoensis. The scale bar represents 200 μm

Fig. 3

Gene trees of LAEWC related genes (a) CER1, (b) CER3, (c) CER5, and (d) CER7. Tree topologies shown here are based on the neighbor-joining method and the branches with bold indicate bootstrap values > 50% for supporting the deriving groups. Values of the nodes indicate the posterior probabilities of the supporting values inferred by the maximum likelihood (ML) and Bayesian inference (BI) methods (ML/BI). Dashes indicate the posterior probability < 50%. The operational taxonomic units labeled in bold are the species with LAEWC. Codes after the species name are the haplotypes cloned in this study

Fig. 4

Phylogenetic principal component analysis (pPCA) conducted with the reference tree. Black wording indicates abbreviation of each species (see Fig. 1), while the red wording indicates ecophysiological traits. (C:carbon content; N: nitrogen content; PA: phenolic acid; C/N: ratio of carbon and nitrogen content; d13C: δ¹³C; d15N: δ¹⁵N; waxy: LAEWC state; Alt: altitude; YII: phytochemical yield of photosystem II)