Project Topics
Get Complete Project Material File(s) Now! »
The molecular side of the flower
Despite the great morphological diversity of flowers across angiosperms, the conservation of a basic floral architectural plan in higher eudicots is at the center of the formulation of the ABC model of floral development and organ identity. This model for floral organ identity specification was developed in parallel in the core eudicot model species Arabidopsis thaliana and Antirrhinum majus, upon observation of a curious group of homeotic mutants exhibiting alternative floral architectures. Three classes of mutants could be discerned among them, A, B and C. Class A mutants have carpels in the place of sepals, and stamens in the place of petals, class B mutants have sepals in both perianth whorls and carpels in both inner whorls, and class C mutants have sepals and petals at the place of stamens and carpels respectively (Coen & Meyerowitz, 1991; Bowman et al., 1991). Additionally, class C mutants also show a disruption of floral meristem determinacy resulting in an increase in the number of floral pieces, a phenomenon commonly called ‘double flower’ (Dubois et al., 2010). The careful interpretation of these mutants led to the formulation of a simple developmental model, in which the combinatorial action of three functional classes explains the specification of the four different floral organs (Figure 1). If in class A mutants normal perianth specification is disrupted, then A function must be related to the identity of sepals and petals. Similarly, if class B mutants fail to produce petals and stamens, then the B function pertains to the identity of petals and stamens. Finally, class C mutants suggest a function of this gene class in stamen and carpel identity, as well as meristem determinacy. It then follows that in order to produce sepals the A function alone suffices, whereas to produce petals the concerted action of A and B function is required. Likewise, to specify stamen identity both B and C functions are needed, and to specify carpels the C function alone is required (Coen & Meyerowitz, 1991).
Molecular analysis studies have revealed the genetic basis for these functional classes in Arabidopsis thaliana. In this model species, two A function genes have been identified: APETALA2 (AP2) and APETALA1 (AP1). The AP2 gene is initially expressed in all four whorls of developing flowers but is later believed to be repressed in whorls 3 and 4, becoming restricted to the first and second whorls were it is required to specify sepal and petal identities respectively (Kunst et al., 1989; Jofuku et al., 1994; Chen, 2004). In addition to its role in sepal and petal development, the AP1 gene has an additional and crucial role in the specification of floral meristem identity. Accordingly, AP1 is initially expressed at an early stage of floral development in a broad fashion across the meristem and only later restricted to whorls 1 and 2 (Mandel et al., 1992; Gustafson-Brown et al., 1994). The B function has also been shown to be performed by two genes in A. thaliana, the APETALA3 (AP3) and PISTILLATA (PI), two related genes issued from an ancient gene duplication event predating the angiosperm lineage (Krizek & Meyerowitz, 1996; Kramer et al., 1998). AP3 and PI are both expressed in the petal and stamen primordia and their products form obligate heterodimers in order to regulate expression of downstream genes implicated in petal and stamen development, as well as each other’s (Goto & Meyerowitz, 1994). Finally, only one C function gene is known in Arabidopsis, the AGAMOUS (AG) gene (Yanofsky et al., 1990). Similarly to the AP1 gene, AG has a double function specifying not only stamen and carpel identity but also floral meristem determinacy (Bowman et al., 1989). Parallel studies of ABC functions in Antirrhinum revealed the identity of both B and C function genes. Antirrhinum B function is performed by AP3 and PI homologs DEFICIENS (DEF) and GLOBOSA (GLO) respectively, while C function is performed by PLENA (PLE), an homolog of AG (reviewed in Theißen et al., 2000). The fact that organ identity specification is performed by homologous genes in a similar fashion between the two distinct species, led to the initial assumption of broad conservation of the ABC program of floral development across flowering species (Coen & Meyerowitz, 1991).
Cloning of these ABC function genes revealed that all encode putative transcription factors and, except for AP2, all belong to the MADS-box family of genes (Ma & DePamphilis, 2000). The members of this family share a highly conserved DNA sequence, called the MADS-box, which encodes a DNA-binding motif. The MADS-box genes involved in floral development belong to a special class of transcription factors called MIKC-type proteins for the presence of four different domains: the MADS (M), intervening (I), keratin-like (K) and C-terminal (C) domains. The MADS-domain, the most conserved region, is required for DNA-binding and protein dimerization. The I-domain is also required for DNA binding of dimer forming proteins and it is believed to influence the specificity of the DNA-binding dimer formation. The K-domain is involved in the mediation of interaction between MIKC-type proteins. The C-terminal domain, the least conserved region, acts as a stabilizing/enhancing factor in K-domain mediated protein interactions (MIKC domain functions were reviewd in Kaufmann et al., 2005).
Conservation of developmental programs
The ABCE model provides a solid working base for the study of floral developmental genetics, with many powerful predictions and testable hypotheses. However, because it has been designed based on observations made in highly derived species, one has to be cautious when making inferences in lower groups of angiosperms. Consequently, what had initially been viewed as a considerable degree of conservation, has been subsequently challenged in recurrent studies testing the applicability of the ABC model to groups outside the core eudicots. Among the functional gene classes of the ABCE model the B, C and E classes seem to be the most conserved with homologs of these genes being isolated from a range of species, while A class homologs with functional roles on flower development are yet to be found (Kramer & Hall, 2005; Soltis et al., 2006; Litt & Kramer, 2010). Therefore, developmental molecular clues into the evolution of petals and perianth differentiation must be based on the study of B function conservation.
Contrary observations punctuate the paradigm of B function in petal development across angiosperms. In model core eudicots AP3 and PI are expressed in a specific manner in the second whorl of the perianth (as well as in the third whorl), where their activity is continuously required for petal specification identity and the proper development of mature petal morphological features. In addition, the heterotopic expression of B genes in the first whorl is capable of inducing ectopic petal formation (Krizek & Meyerowitz, 1996). Despite the likely homoplasy of petals in core eudicots and other outside groups, homologs of the AP3 and PI B genes have been recurrently shown to be expressed in association with petalous perianths across angiosperms, apparently supporting a conservation of B function (Rasmussen et al., 2009; Litt & Kramer, 2010).
On the other hand, studies of B gene expression in basal angiosperms have revealed a previously unknown complexity of patterns and dynamics. Most notably, unlike in the core eudicots, activity of AP3 and PI homologs in basal angiosperms is not restricted to certain whorls but is found in broad domains across floral meristems and their continued expression in later stages of primordia differentiation is not required for the proper petal development (Soltis et al., 2006). Additionally, although B gene expression can be found in first whorl petaloid organs in some species (Bowman, 1997; Kramer et al., 2003), other studies have shown a lack of association of B gene expression with the occurrence of petaloid organs outside the second whorl, or an expression of B genes in non-petaloid second whorl organs (Jaramillo & Kramer, 2004; Geuten et al., 2006; Landis et al., 2012). These observations have lead authors to question the role of B genes in specifying petals and petaloidy, and to advance an hypothesis on the decoupling of petaloidy and B gene expression in less derived angiosperms (Ronse De Craene, 2007). The idea that B genes expression does not imply the production of petals or petaloidy was elegantly incorporated in the ‘regional specification’ model (Irish, 2009). This model reiterates the idea that specification of distinct inner perianth organs in association with the expression of B gene orthologs in that domain does not necessarily lead to the production of petals. Therefore, the ancestral role of B gene homologs may not be the specification of organ identity but of a region within the perianth (Figure 2, p. 20) (Drea et al., 2007; Irish, 2009).
Molecular developmental theories of perianth architecture evolution and diversity
The remarkable diversity of perianth architectures across angiosperms has long suggested a complex evolutionary history of organ loss and gain, eliciting several transitions between the undifferentiated and differentiated states. In addition to the diversity of perianth compositions, petals also display a remarkable diversity of form, color and size across angiosperms, which have led to a long standing belief that these structures have evolved several times across angiosperm lineages. Phylogenetic and morphological analyses both seem to support independent origins for the differentiated perianth in the monocots, the Ranunculaceae and in core eudicots (Endress & Doyle, 2009; Ronse De Craene & Brockington, 2013). Bowman (1997) pointed out that evolution of a bipartite perianth from an undifferentiated state, could occur by differentiation of the existing perianth into two distinct organ types, or alternatively, by conversion of stamens to petaloid organs. Based on a number of complex morphological features such as development rate, epidermal cell identity and venation pattern, petals are believed to have arisen as modifications of stamen-like structures in a process of andropetaloidy, or to be derived from bract- or sepal-like organs, bracteopetaloidy, depending on the angiosperm lineage (Kosuge, 1994; Ronse De Craene, 2007; Ronse De Craene & Brockington, 2013).
Understanding floral architecture diversity
In order to understand the evolution of flower diversity beyond the how and the when provided by molecular, morphological and phylogenetic studies, it is indispensible to consider the functional aspects of the flower and the functional roles of the different traits. In fact, long before the development of such disciplines as molecular biology, the evolution of flower diversity already puzzled biologists. We have already seen that this was a major concern of Darwin as it directly challenged his views on gradual evolution by natural selection. When confronted with the evidence for the rapid diversification and biogeographical spread of higher plants, Darwin admitted that it could only be explained by an extremely rapid pace of evolutionary diversification or a strikingly long and missing fossil record. Unsurprisingly he dismissed the absence of fossil evidence for a long and gradual history of evolutionary transformation, which should be an indication of rapid evolutionary change, and instead chose to propose an explanation for that gap in the fossil record. He came, however, to accept an early theory that aimed to explain the seemingly abrupt and highly accelerated diversification of floral morphology, through a strong co-evolutionary interdependence mechanism between insects and flowering plants (reviewed in Friedman, 2009). Darwin’s efforts to understand flowering plants diversity in light of his adaptation by natural selection theory were henceforth marked by his realization of the importance of outcrossing, and the role of insect pollination in such mating systems. The greatest consequence of this work was the promotion of the new discipline of pollination biology, which laid the basis for our current understanding of floral function and provided the most accepted hypothesis for the rapid radiation of flowering plants, that of a co-evolution with insects (Friedman, 2009; Harder & Johnson, 2009). The prevailing view, even today, is that floral diversity is the result of gradual and continuous selection accumulation in a process of adaptation of floral traits to different modes of plant reproduction, of which biotic mediated outcrossing (animal pollination) is the most effective and widespread (Fenster et al., 2004; Kay & Sargent, 2009). As Stebbins (1970) put it: “The diverse floral structures and pollination mechanisms found in angiosperms represent a series of adaptive radiations to different pollen vectors and different ways of becoming adapted to the same vector.”
Pollination biology and flower diversity
The core of pollination biology theories lies in the realization of the extraordinary relationship of dependence between plants and the surrounding environment for their successful mating, i.e. production of high quality offspring via cross-pollination (Barrett, 2010; Schiestl & Johnson, 2013). Indeed, it is frequently observed that differences between major groups of angiosperms are found in association with traits implicated with reproduction efficiency and successful establishment of seedlings, whereas the vegetative aspects of plant diversity largely reflect adaptations of a same body plan to different environments (Stebbins, 1970). Despite not being exclusively necessary (self-pollinating species do exist, although not without constraints on their evolutionary potential due to their particular genetic structure), cross-pollination, that is, mating with another plant, is the norm among the majority of flowering plants reproduction systems (Stebbins, 1970). Due to their immobility plants depend either on biotic or abiotic elements for the successful transport of pollen from one plant to another. Among flowering plants the most frequently used pollen vectors are animal pollinators with which they may establish strong relations of interdependence benefitting both the plant (from a reproductive point of view) and the animal (from a foraging and nourishing point of view) (Mitchell et al., 2009b). Under a comparative evolutionary context, such interactions between flowers and pollinators have given rise to the hypothesis that floral traits have adapted for pollination by different animal groups which in turn has led to convergent evolution of floral traits into common character states or syndromes. That is, that flowers exhibit a certain level of specialization for the attraction and use of specific group of animals as pollinators (Johnson & Steiner, 2000; Fenster et al., 2004).
Table of contents :
Chapter 1. Developmental and molecular origin of the Nigella damascena floral dimorphism
Résumé en français du premier chapitre
Preamble
Article – An APETALA3 homolog controls both petal identity and floral meristem patterning in Nigella damascena L. (Ranunculaceae)
Concluding remarks
Chapter 2. Ecological and evolutionary significance of the Nigella damascena floral dimorphism
Résumé en français du deuxième chapitre 1
Preamble
Introduction
Material and methods
Results
Discussion
References
Supporting information
Conclusion
Brève résumé en français de la conclusion
Loose ends
An essay on finding the meaning of fundamental biology research
Introduction
Version abrégé en français de l’introduction
Le contexte évolutif
La fleur – contexte morphologique, moléculaire et développemental
Méchanismes d’évolution et diversification des traits floraux
Le dimorphisme floral chez Nigella damascena
The evolutionary context
Wonderful mysteries
Darwin’s legacy
“Evolutionary concepts evolve” (Dobzhansky, 1963b)
Hopeful homeotic monsters
The flower – morphological, molecular and historical context
Why flowers?
What’s in a name?
The molecular side of the flower
Conservation of developmental programs
Molecular developmental theories of perianth architecture evolution and diversity
Homeosis in flowers
Mechanisms of floral traits diversification and evolution
Understanding floral architecture diversity
Pollination biology and flower diversity
The Nigella damascena floral dimorphism
Species context
Species description
Question 1. Molecular origin of the Nigella damascena floral dimorphism
Question 2. Evolutionary significance of perianth form variation
References
Chapter 1. Developmental and molecular origin of the Nigella damascena floral dimorphism
Résumé en français du premier chapitre
Preamble
Article – An APETALA3 homolog controls both petal identity and floral meristem patterning in
Nigella damascena L. (Ranunculaceae)
Conclusive remarks
References
Chapter 2. Ecological and evolutionary significance of the Nigella damascena floral dimorphism
Preamble
Introduction
Material and methods
Plant material
Production of a polymorphic population
The experimental set-up
Morphological measurements and notations during flowering – G0
Morphological measurements and notations on capsules and seeds – G0
Insect observations – G0
Morphological measurements and notations on progeny
Morphological measurements and notations during flowering – G1
Morph-ratio determination
Data analysis
Results
Floral traits and morphology associated with reproduction mode
Pollinator identification and behavior
Differences between population replicates
Differences between [P] and [T] – Floral traits
Differences between [P] and [T] – Capsule and seed production
Differences between [P] and [T] – Progeny development and vigor
Estimating fitness
Detection of selection on floral traits
Morph-ratio
Discussion
Morph effect on pollinator behavior
Reproduction strategies
Reproductive success
Selection on floral traits
Major conclusions and perspectives
References
