Origin, loss, and regain of self-incompatibility in angiosperms

Abstract

The self-incompatibility (SI) system with the broadest taxonomic distribution in angiosperms is based on multiple S-locus F-box genes (SLFs) tightly linked to an S-RNase termed type-1. Multiple SLFs collaborate to detoxify nonself S-RNases while being unable to detoxify self S-RNases. However, it is unclear how such a system evolved, because in an ancestral system with a single SLF, many nonself S-RNases would not be detoxified, giving low cross-fertilization rates. In addition, how the system has been maintained in the face of whole-genome duplications (WGDs) or lost in other lineages remains unclear. Here we show that SLFs from a broad range of species can detoxify S-RNases from Petunia with a high detoxification probability, suggestive of an ancestral feature enabling cross-fertilization and subsequently modified as additional SLFs evolved. We further show, based on its genomic signatures, that type-1 was likely maintained in many lineages, despite WGD, through deletion of duplicate S-loci. In other lineages, SI was lost either through S-locus deletions or by retaining duplications. Two deletion lineages regained SI through type-2 (Brassicaceae) or type-4 (Primulaceae), and one duplication lineage through type-3 (Papaveraceae) mechanisms. Thus, our results reveal a highly dynamic process behind the origin, maintenance, loss, and regain of SI.

The Origin and Evolution of RNase T2 Family and Gametophytic Self-incompatibility System in Plants

“…WGDs or segmental duplications could generate a copy of the S -locus, resulting in competitive interaction between the two copies and the loss of SI through Route I ( fig. 5 A ) ( Zhao et al 2021b ). However, the majority of extant species exhibited SI, which led us to query how SI was regained in these species.…”

“…However, the majority of extant species exhibited SI, which led us to query how SI was regained in these species. Reduced expression or inactivation of S-RNases and SLFs in the duplicated copy of the S -locus have been suggested to lead to the regain of SI in Antirrhinum hispanicum ( Fujii et al 2016 ; Zhao et al 2021b ). Based on this observation, we inferred that some species in Rosaceae ( Prunus armeniaca , Pr.…”

“…Previous studies have shown that Class I RNase T2 genes are contained in “early-diverging” groups (marchantiophyta, bryophytes, lycophytes, and ferns), some Class II RNase T2 genes are constitutively expressed (e.g., RNS2 in A. thaliana and RNase LER in S. lycopersicum ) ( Hillwig et al 2011 ; Kothke and Kock 2011 ), and Class III genes may have originated from those in Class I or may share a common ancestor ( MacIntosh et al 2010 ). S-RNases , in Class III, have been the focus of much attention due to their important role in gametophytic self-incompatibility (GSI) ( Ashkani and Rees 2016 ; Fujii et al 2016 ; Zhao et al 2021b ). The molecular function of S-RNase genes has been studied in Rosaceae species such as Prunus mume , Malus × domestica , and Pyrus bretschneideri .…”

“…GSI is controlled by the S -locus, which contains the S-RNase and F-box genes; the pistil determinant of GSI is the S-RNase gene and the male determinants are SLF genes, which are usually linked to the S-RNase and are also known as S-haplotype specific F-box ( SFB ) genes ( Sijacic et al 2004 ). Recently, SI was divided into four specific types ( Zhao et al 2021b ). Type-1 SI, found in most GSI species, is controlled by multiple pollen-specific SLF s from one haplotype and pistil-specific S-RNases from other haplotypes.…”

“…S-RNase -based GSI has been found in the common ancestor of Asteridae and Rosidae ( Vieira et al 2008 ), and S-RNase and SLF genes show co-evolution patterns in this system ( Kubo et al 2010 ). To date, this GSI system has been identified in five families (Solanaceae, Plantaginaceae, Rosaceae, Rutaceae, and Rubiaceae) ( Franklin-Tong and Franklin 2003 ; McClure and Franklin-Tong 2006 ; Vieira et al 2008 , 2021 ; Zhao et al 2021b ). GSI occurs in two divergent patterns (in Maleae and Amygdaleae) in Rosaceae, in which different numbers of pollen-specific SLF s genes imply that there are different evolutionary mechanisms of self-pollen recognition in Maleae and Amygdaleae ( Hua et al 2008 ; Kubo et al 2010 ; Fujii et al 2016 ).…”