Petals can be simple or elaborate, depending on whether they have lobes, teeth, fringes, or appendages along their margins, or possess spurs, scales, or other types of modifications on their adaxial/abaxial side, or both. Elaborate petals have been recorded in 23 orders of angiosperms and are generally believed to have played key roles in the adaptive evolution of corresponding lineages. The mechanisms underlying the formation of elaborate petals, however, are largely unclear. Here, by performing extensive transcriptomic and functional studies on Nigella damascena (Ranunculaceae), we explore the mechanisms underlying elaborate petal development and specialized character formation. In addition to the identification of genes and programs that are specifically/preferentially expressed in petals, we found genes and programs that are required for elaborate rather than simple petal development. By correlating the changes in gene expression with those in petal development, we identified 30 genes that are responsible for the marginal/ventral elaboration of petals and the initiation of several highly specialized morphological characters (e.g., pseudonectaries, long hairs, and short trichomes). Expression and functional analyses further confirmed that a class I homeodomain-leucine zipper family transcription factor gene, Nigella damascena LATE MERISTEM IDENTITY1 (NidaLMI1), plays important roles in the development of short trichomes and bifurcation of the lower lip. Our results not only provide the first portrait of elaborate petal development but also pave the way to understanding the mechanisms underlying lateral organ diversification in plants.

AP1/FUL, SEP, AGL6, and FLC subfamily genes play important roles in flower development. The phylogenetic relationships among them, however, have been controversial, which impedes our understanding of the origin and functional divergence of these genes. One possible reason for the controversy may be the problems caused by changes in the exon-intron structure of genes, which, according to recent studies, may generate non-homologous sites and hamper the homology-based sequence alignment. In this study, we first performed exon-by-exon alignments of these and three outgroup subfamilies (SOC1, AG, and STK). Phylogenetic trees reconstructed based on these matrices show improved resolution and better congruence with species phylogeny. In the context of these phylogenies, we traced evolutionary changes of exon-intron structures in each subfamily. We found that structural changes have occurred frequently following gene duplication and speciation events. Notably, exons 7 and 8 (if present) suffered more structural changes than others. With the knowledge of exon-intron structural changes, we generated more reasonable alignments containing all the focal subfamilies. The resulting trees showed that the SEP subfamily is sister to the monophyletic group formed by AP1/FUL and FLC subfamily genes and that the AGL6 subfamily forms a sister group to the three abovementioned subfamilies. Based on this topology, we inferred the evolutionary history of exon-intron structural changes among different subfamilies. Particularly, we found that the eighth exon originated before the divergence of AP1/FUL, FLC, SEP, and AGL6 subfamilies and degenerated in the ancestral FLC-like gene. These results provide new insights into the origin and evolution of the AP1/FUL, FLC, SEP, and AGL6 subfamilies.

Pseudonectaries, or false nectaries, the glistening structures that resemble nectaries or nectar droplets but do not secrete nectar, show considerable diversity and play important roles in plant-animal interactions. The morphological nature, optical features, molecular underpinnings and ecological functions of pseudonectaries, however, remain largely unclear. Here, we show that pseudonectaries of Nigella damascena (Ranunculaceae) are tiny, regional protrusions covered by tightly arranged, non-secretory polygonal epidermal cells with flat, smooth and reflective surface, and are clearly visible even under ultraviolet light and bee vision. We also show that genes associated with cell division, chloroplast development and wax formation are preferably expressed in pseudonectaries. Specifically, NidaYABBY5, an abaxial gene with ectopic expression in pseudonectaries, is indispensable for pseudonectary development: knockdown of it led to complete losses of pseudonectaries. Notably, when flowers without pseudonectaries were arrayed beside those with pseudonectaries, clear differences were observed in the visiting frequency, probing time and visiting behavior of pollinators (i.e., honey bees), suggesting that pseudonectaries serve as both visual attractants and nectar guides.

Species of the tribe Delphinieae (Ranunculaceae) have long been the focus of morphological, ecological and evolutionary studies due to their highly specialized, nearly zygomorphic (bilaterally symmetrical) spiral flowers with nested petal and sepal spurs and reduced petals. The mechanisms underlying the development and evolution of Delphinieae flowers, however, remain unclear. Here, by conducting extensive phylogenetic, comparative transcriptomic, expression, and functional studies, we clarified the evolutionary histories, expression patterns, and functions of floral organ identity and symmetry genes in Delphinieae. We found that duplication and/or diversification of APETALA3-3 (AP3-3), AGAMOUS-LIKE6 (AGL6), CYCLOIDEA (CYC) and DIVARICATA (DIV) lineage genes was tightly associated with the origination of Delphinieae flowers. Specifically, an AGL6-lineage member (such as the Delphinium ajacis AGL6-1a) represses sepal spur formation and petal development in the lateral and ventral parts of the flower while determining petal identity redundantly with AGL6-1b. By contrast, two CYC2-like genes, CYC2b and CYC2a, define the dorsal and lateral-ventral identities of the flower, respectively, and form complex regulatory links with AP3-3, AGL6-1a and DIV1. Therefore, duplication and diversification of floral symmetry genes, as well as co-option of the duplicated copies into the pre-existing floral regulatory network, have been key for the origin of Delphinieae flowers.

Petals can be simple or elaborate, depending on whether they have complex basic structures and/or highly specialized epidermal modifications. It has been proposed that the independent origin and diversification of elaborate petals have promoted plant-animal interactions and, therefore, the evolutionary radiation of corresponding plant groups. Recent advances in floral development and evolution have greatly improved our understanding of the processes, patterns and mechanisms underlying petal elaboration. In this review, we compare the developmental processes of simple and elaborate petals, concluding that elaborate petals can be achieved through four main ways of modifications (i.e., marginal elaboration, ventral elaboration, dorsal elaboration and surface elaboration). Meanwhile, although different types of elaborate petals were made through different ways of modifications, they are all results of changes in the expression patterns of genes involved in organ polarity establishment and/or the proliferation, expansion and differentiation of cells. Deployment of existing genetic materials to a new function was also shown to be a key to the making of elaborate petals during evolution.

Summary Complex color patterns on petals are widespread in flowering plants, yet the mechanisms underlying their formation remain largely unclear. Here, by conducting detailed morphological, anatomical, biochemical, optical, transcriptomic, and functional studies, we investigated the cellular bases, chromogenic substances, reflectance spectra, developmental processes, and underlying mechanisms of complex color pattern formation on Nigella orientalis petals. We found that the complexity of the N. orientalis petals in color pattern is reflected at multiple levels, with the amount and arrangement of different pigmented cells being the key. We also found that biosynthesis of the chromogenic substances of different colors is sequential, so that one color/pattern is superimposed on another. Expression and functional studies further revealed that a pair of R2R3‐MYB genes function cooperatively to specify the formation of the eyebrow‐like horizontal stripe and the Mohawk haircut‐like splatters. Specifically, while NiorMYB113‐1 functions to draw a large splatter region, NiorMYB113‐2 functions to suppress the production of anthocyanins from the region where a gap will form, thereby forming the highly specialized pattern. Our results provide a detailed portrait for the spatiotemporal dynamics of the coloration of N. orientalis petals and help better understand the mechanisms underlying complex color pattern formation in plants.

Peltate organs, such as the prey-capturing traps of carnivorous plants and nectary-bearing petals of ranunculaceous species, are widespread in nature and have intrigued and perplexed scientists for centuries. Shifts in the expression domains of adaxial/abaxial genes have been shown to control leaf peltation in some carnivorous plants, yet the mechanisms underlying the generation of other peltate organs remain unclear. Here, we show that formation of various peltate ranunculaceous petals was also caused by shifts in the expression domains of adaxial/abaxial genes, followed by differentiated regional growth sculpting the margins and/or other parts of the organs. By inducing parameters to specify the time, position, and degree of the shifts and growth, we further propose a generalized modeling system, through which various unifacial, bifacial, and peltate organs can be simulated. These results demonstrate the existence of a hierarchical morphospace system and pave the way to understand the mechanisms underlying plant organ diversification.

In the original Abstract, the gene 'NidaYABB5' was missing a Y and should have been NidaYABBY5. In the first paragraph of the Discussion, the sentence 'Specifically, being protrusive, colorful, and reflective make pseudonectaries…' should have read: 'Specifically, being protrusive, colorful, and reflective makes pseudonectaries…'. In the fourth paragraph of the Discussion, the sentence 'Genes involved in nectar development, such as orthologs STY1/2 and LRP…' should have read: 'Genes involved in nectary development, such as orthologs of STY1/2 and LRP…'. In the fourth paragraph of the Discussion, the sentence '…ectopic expression of NidaYAB5 can explain the reason why pseudonectaries was formed in the adaxial side of the lower lip of the petal' should have read: '…ectopic expression of NidaYAB5 can explain the reason why pseudonectaries were formed in the adaxial side of the lower lip of the petal'.