The liverwort Marchantia polymorpha has been utilized as a model for biological studies since the 18th century. In the past few decades, there has been a Renaissance in its utilization in genomic and genetic approaches to investigating physiological, developmental, and evolutionary aspects of land plant biology. The reasons for its adoption are similar to those of other genetic models, e.g. simple cultivation, ready access via its worldwide distribution, ease of crossing, facile genetics, and more recently, efficient transformation, genome editing, and genomic resources. The haploid gametophyte dominant life cycle of M. polymorpha is conducive to forward genetic approaches. The lack of ancient whole-genome duplications within liverworts facilitates reverse genetic approaches, and possibly related to this genomic stability, liverworts possess sex chromosomes that evolved in the ancestral liverwort. As a representative of one of the three bryophyte lineages, its phylogenetic position allows comparative approaches to provide insights into ancestral land plants. Given the karyotype and genome stability within liverworts, the resources developed for M. polymorpha have facilitated the development of related species as models for biological processes lacking in M. polymorpha.

Glutaredoxins (GRXs) are ubiquitous oxidoreductases that play a crucial role in response to oxidative stress by reducing disulfides in various organisms. In planta, three different GRX classes have been identified according to their active site motifs. CPYC and CGFS classes are found in all organisms, whereas the CC-type class is specific for higher land plants. Recently, two Arabidopsis CC-type GRXs, ROXY1 and ROXY2, were shown to exert crucial functions in petal and anther initiation and differentiation. To analyze the function of CC-type GRXs in the distantly related monocots, we isolated and characterized OsROXY1 and OsROXY2—two rice homologs of ROXY1. Both genes are expressed in vegetative and reproductive stages. Although rice flower morphology is distinct from eudicots, OsROXY1/2 floral expression patterns are similar to their Arabidopsis counterparts ROXY1/2. Complementation experiments demonstrate that OsROXY1 and OsROXY2 can fully rescue the roxy1 floral mutant phenotype. Overexpression of OsROXY1, OsROXY2, and ROXY1 in Arabidopsis causes similar vegetative and reproductive plant developmental defects. ROXY1 and its rice homologs thus exert a conserved function during eudicot and monocot flower development. Strikingly, overexpression of these CC-type GRXs also leads to an increased accumulation of hydrogen peroxide levels and hyper-susceptibility to infection from the necrotrophic pathogen Botrytis cinerea, revealing the importance of balanced redox processes in flower organ development and pathogen defence.

Evolution of floral monosymmetry is thought to be a major driving force of angiosperm radiation, making angiosperms the most successful land plant group in terms of species richness. Monosymmetry evolved from a polysymmetric ancestor repeatedly in different angiosperm lineages, where it likely facilitated diversification through the interaction with insects. Most monosymmetric taxa are thus dominated by monosymmetric members. However, in the Brassicaceae, only few members develop a monosymmetric corolla with two petal pairs of unequal size, making them an ideal system to study the evolution of molecular mechanisms enhancing flower complexity. Monosymmetry is controlled by the TCP transcription factors that belong to the CYC2 clade in distantly related taxa. In Iberis amara, the first crucifer analyzed in terms of monosymmetry development, unequal corolla formation is due to a stronger CYC2 clade gene expression in the smaller adaxial petals compared with the larger abaxial ones. Phylogenetic reconstruction of the crucifer family reveals that the monosymmetric genera Iberis, Calepina, and Teesdalia belong to one major crucifer lineage. Monosymmetry is most pronounced in Iberis and less so in Calepina and Teesdalia, with a positive dosage-dependent correlation between the strength of a CYC2 expression difference and the extent of monosymmetry formation. An early adaxial CYC2 expression in floral meristems, observed in many distantly related taxa, might have facilitated the repeated evolution of CYC2-controlled monosymmetry. Comparison of early and late CYC2 expression in monosymmetric and polysymmetric crucifers representative for the four major crucifer lineages reveals that an adaxial CYC2 expression in floral meristems is likely ancestral for the Brassicaceae. However, it got lost in all analyzed monosymmetric members and is, as such, not a prerequisite for the establishment of corolla monosymmetry in crucifers. Here, monosymmetry evolved via a heterochronic CYC2 expression shift from an ancestral early adaxial expression in floral meristems to an adaxial CYC2 transcript accumulation later in petal development. This study emphasizes the potential of regulatory changes in the evolution of morphological novelties, like corolla monosymmetry in the Brassicaceae. In combination with a corymboid inflorescence, monosymmetry might have served as a key invention driving diversification in the genus Iberis comprising more than 20 monosymmetric species.

Summary: Flower monosymmetry contributes to specialized interactions between plants and their insect pollinators. In the magnoliids, flower monosymmetry is exhibited only in the Aristolochiaceae (Piperales). Aristolochia flowers develop a calyx‐derived monosymmetric perianth that enhances pollination success by a flytrap mechanism. Aristolochia arborea forms additionally a special perianth outgrowth that mimics a mushroom to attract flies, the mushroom mimicry structure (MMS). In core eudicots, members of the CYC2 clade of TCP transcription factors are key regulators of corolla monosymmetry establishment. The CYC2 clade arose via core eudicot‐specific duplications from ancestral CYC/TB1 genes. CYC/TB1 genes are also thought to affect monosymmetry formation in early diverging eudicot and monocot species. Here, we demonstrate that CYC/TB1 genes, named CYC‐like genes (CYCL) are present in basal angiosperms and magnoliids. Expression analyses in A. arborea indicate that CYCL genes participate in perianth and MMS differentiation processes and do not support a CYCL gene function in initial flower monosymmetry formation. Heterologous CYCL and CYC2 gene overexpression studies in Arabidopsis show that Aristolochia CYCL proteins only perform a CYC2‐like function when the CYCL TCP domain is replaced by a CYC2 domain. Comparative TCP domain analyses revealed that an LxxLL motif, known to mediate protein–protein interactions, evolved in the second helix of the TCP domain in the CYC2 lineage and contributes to CYC2‐related functions. Our data imply that divergent evolution of the CYC/TB1 lineages caused significant changes in their coding regions, which together with cis‐regulatory changes established the key CYC2 function in regulating eudicot flower monosymmetry.