Plastome evolution in Santalales involves relaxed selection prior to loss of ndh genes and major boundary shifts of the inverted repeat

Abstract

Background and aims: Biological aspects of haustorial parasitism have significant effects on the configuration of the plastid genome. Approximately half the diversity of haustorial parasites belongs to the order Santalales, where a clearer picture of plastome evolution in relation to parasitism is starting to emerge. However, in previous studies of plastome evolution there is still a notable under-representation of members from non-parasitic and deep-branching hemiparasitic lineages, limiting evolutionary inference around the time of transition to a parasitic lifestyle. To expand taxon sampling relevant to this transition we therefore targeted three families of non-parasites (Erythropalaceae, Strombosiaceae and Coulaceae), two families of root-feeding hemiparasites (Ximeniaceae and Olacaceae) and two families of uncertain parasitic status (Aptandraceae and Octoknemaceae). With data from these lineages we aimed to explore plastome evolution in relation to the evolution of parasitism.

Methods: From 29 new samples we sequenced and annotated plastomes and the nuclear ribosomal cistron. We examined phylogenetic patterns, plastome evolution, and patterns of relaxed or intensified selection in plastid genes. Available transcriptome data were analysed to investigate potential transfer of infA to the nuclear genome.

Results: Phylogenetic relationships indicate a single functional loss of all plastid ndh genes (ndhA-K) in a clade formed by confirmed parasites and Aptandraceae, and the loss coincides with major size and boundary shifts of the inverted repeat (IR) region. Depending on an autotrophic or heterotrophic lifestyle in Aptandraceae, plastome changes are either correlated with or pre-date the evolution of parasitism. Phylogenetic patterns also indicate repeated loss of infA from the plastome, and based on the presence of transcribed sequences with presequences corresponding to thylakoid luminal transit peptides, we infer that the genes were transferred to the nuclear genome.

Conclusions: Except for the loss of the ndh complex, relatively few genes have been lost from the plastome in deep-branching root parasites in Santalales. Prior to loss of the ndh genes, they show signs of relaxed selection indicative of their dispensability. To firmly establish a potential correlation between ndh gene loss, plastome instability and evolution of parasitism, it is pertinent to refute or confirm a parasitic lifestyle in all Santalales clades.

Keywords: Gene loss; Santalales; inverted repeat; parasitism; phylogeny; plastome evolution.

Fig. 1.

Four alternative topologies from previous and present phylogenetic analyses of Santalales: (A) Su et al. (2015), 3 plastid + 1 mitochondrial + 3 nuclear genes; (B) Nickrent et al. (2019), 3 plastid + 2 nuclear genes; (C) Nickrent (2020), 3 plastid + 1 mitochondrial + 3 nuclear genes; (D) this study, 76 plastid + 3 nuclear genes, summary tree of maximum likelihood and Bayesian inference analyses. Green terminals are non-parasitic; brown terminals are parasitic; blue terminals have uncertain parasitic status. Other Santalales comprise 13 families including root and stem hemiparasites as well as holoparasites.

Fig. 2.

Inverted repeat (IR) region in representatives of Santalales. Red bars showing relative length of the IR. Coloured squares indicate a core IR region shared by outgroup representatives and ancestral Santalales. Functional genes in the core IR region are in green, pseudogenes in yellow. Genes of species with expanded IRs are listed upstream or downstream of the core region. Pa = genes partially included in the IR; P = pseudogenes in the non-core IR region.

Fig. 3.

Phylogeny of Santalales based on maximum likelihood analysis of combined plastome gene and nrDNA sequences. Node values are ultrafast bootstrap support. Green branches correspond to non-parasitic lineages, brown branches to parasitic lineages. Blue branches are lineages with uncertain parasitic status. Asterisks (*) after species names indicate stem parasites. Functional gene losses are marked in red. The scale bar indicates inferred substitutions per site.

Fig. 4.

Comparison between plastid (P) and putative nuclear (N) infA amino acid sequences. Predicted cleavage sites are in bold/red.