SEARCH

SEARCH BY CITATION

Keywords:

  • perianth architecture;
  • B–function genes;
  • petal identity;
  • floral meristem patterning;
  • homeotic transformation;
  • Ranunculaceae;
  • Nigella damascena

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Flower architecture mutants provide a unique opportunity to address the genetic origin of flower diversity. Here we study a naturally occurring floral dimorphism in Nigella damascena (Ranunculaceae), involving replacement of the petals by numerous sepal-like and chimeric sepal/stamen organs. We performed a comparative study of floral morphology and floral development, and characterized the expression of APETALA3 and PISTILLATA homologs in both morphs. Segregation analyses and gene silencing were used to determine the involvement of an APETALA3 paralog (NdAP3–3) in the floral dimorphism. We demonstrate that the complex floral dimorphism is controlled by a single locus, which perfectly co-segregates with the NdAP3–3 gene. This gene is not expressed in the apetalous morph and exhibits a particular expression dynamic during early floral development in the petalous morph. NdAP3–3 silencing in petalous plants perfectly phenocopies the apetalous morph. Our results show that NdAP3–3 is fully responsible for the complex N. damascena floral dimorphism, suggesting that it plays a role not only in petal identity but also in meristem patterning, possibly through regulation of perianth organ number and the perianth/stamen boundary.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The flower is very likely the structure that most contributed to the evolutionary success of angiosperms. Flowers typically consist of two inner whorls containing the male and female reproductive organs, surrounded by two outer whorls of sterile organs, the sepals and petals, which collectively form the perianth and often serve as attractive structures for insect-pollinated species. Variations in the composition and shape of the perianth contribute greatly to the remarkable diversity of flower architectures observed across the angiosperms, the origin of which has intrigued developmental and evolutionary biologists for almost two centuries (Friedman and Diggle, 2011). Detailed studies of floral homeotic mutants in the core eudicot species Arabidopsis thaliana and Antirrhinum majus led to the proposal of a genetic model for floral organ identity specification (Bowman et al., 1991; Coen and Meyerowitz, 1991). According to this model, three classes of genes (A, B and C) act in concentric overlapping fields within the floral meristem to specify the identity of the four floral organs: A alone specifies sepals, A + B specify petals, B + C specify stamens, and C alone specifies carpels. Since its description more than 20 years ago, this model has been expanded as new gene classes have been discovered (Angenent et al., 1995; Pelaz et al., 2000), and reappraised, as new studies outside the core eudicots reveal the extent of its conservation across the angiosperms. Notably, expression studies of the various classes of organ identity genes in basal eudicot and basal angiosperm flowers have revealed broad consistency with the predictions of the ABC model, particularly for the B- and C–class genes (Soltis et al., 2007; Theißen and Melzer, 2007; Litt and Kramer, 2010).

In Arabidopsis, the B function is fulfilled by the MADS box genes APETALA3 (AP3) and PISTILLATA (PI) (Bowman et al., 1989; Krizek and Meyerowitz, 1996). AP3 and PI belong to two related gene lineages issuing from an ancient gene duplication that probably occurred before angiosperm diversification. Following this ancient duplication, the two gene lineages were subjected to different evolutionary histories across the angiosperms (Kramer et al., 1998). In the basal eudicot order Ranunculales, two additional gene duplications took place in the AP3 lineage, the last one after divergence of the Papaveraceae, resulting in production of three AP3 paralogs (AP3–1, AP3–2 and AP3–3) (Kramer et al., 2003; Rasmussen et al., 2009; Hu et al., 2012). The retention of these AP3 paralogs has been attributed to gene sub-functionalization and/or neo-functionalization, which may account for the divergent expression patterns observed among them and the possible evolution of novel regulatory functions in organ identity specification (Kramer et al., 2003; Stellari et al., 2004; Rasmussen et al., 2009). Indeed, while the AP3–1 and AP3–2 paralogs have broad and temporally variable expression domains, the AP3–3 paralog exhibits petal-specific expression and an even more remarkable absence of expression in apetalous species, which makes it a likely candidate for the petal identity specification function (Rasmussen et al., 2009; Zhang et al., 2013). Despite this general correlation, the involvement of AP3–3 in the petal identity program has only been functionally validated in the petalous Ranunculaceae species Aquilegia coerulea, in which AP3–3 gene silencing led to conversion of petals into sepals (Sharma et al., 2011). Additional functional studies are required in order to determine whether the AP3–3 role in petal identity specification and perianth architecture is conserved on a wider scale.

The Ranunculaceae family, the richest in species among Ranunculales, exhibits a remarkable diversity of perianth architecture, displaying numerous transitions between a bipartite perianth, possessing morphologically differentiated sepals and petals, and a unipartite perianth, consisting of undifferentiated organs that are either entirely sepaloid or entirely petaloid. Nigella damascena L. (love in a mist) presents a rare case of a bipartite to unipartite perianth transition at the species level, resulting in a perianth architecture dimorphism. This unipartite morph has been previously described by Toxopéus (1927) as a double-flower mutant, lacking petals and instead having a series of sepal-like organs. Studies on the organ identity shift have revealed its monogenic control by a bi-allelic locus, with the petalous form being dominant (Toxopéus, 1927), as well as a tight association of the NdAP3–3 paralog with the petalous form, both at the expression pattern and genomic sequence levels (Zhang et al., 2013).

Here we describe and investigate the genetic and molecular origin of a previously disregarded aspect of the Nigella damascena perianth dimorphism, involving an increase in total perianth organ number and the production of a gradation of organ morphologies from entirely sepal-like to mixed sepal and stamen-like, suggestive of a disruption of the boundary between perianth and reproductive organs. We characterized in detail the floral morphology and floral development of both morphs, and performed a comparative analysis of the gene expression of the APETALA3 and PISTILLATA homologs. Segregation analysis and virus-induced gene silencing (VIGS) studies provide compelling evidence for involvement of the NdAP3–3 paralog in all the aspects of the floral dimorphism. We show that NdAP3–3 plays a dual role in flower development, determining not only the identity of the petals, but also controlling floral meristem patterning through regulation of perianth organ number and proper establishment of a perianth/stamen boundary. We discuss this dual role in an evolutionary context within the Ranunculaceae.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The Nigella damascena floral dimorphism encompasses not only a shift in petal identity but also in perianth organ number and the perianth/stamen boundary

Two Nigella damascena L. floral morphs were observed among the plants sown in greenhouse from a sample of seeds collected in the natural population of Mornas (Vaucluse, France). The most commonly observed form, hereafter referred to as the [P] morph, has four organ types inserted in a spiral arrangement: approximately five sepals and eight petals, a variable number of stamens arranged in eight parastichies, and a gynecium of five proximally connate carpels (Figure 1a). Mature sepals have an ovate simple blade with a lanceolate apex and a range of petaloid characteristics such as bright blue coloration and the presence of papillated striated conical cells on the adaxial surface (Figure 1c,i). On the abaxial side, the cells are irregularly shaped, pavement-like and interspersed with stomata (Figure 1j). The petals are small dark-blue organs with a narrow stalk-like base, two apical lobes bearing two round glistening pseudonectaries, and a nectariferous pouch covered by a flat scale (Figure 1d). The petal cellular epidermis is composed of papillated conical cells with ornamentations, and interspersed with trichomes on the adaxial side of the apical lobes (Figure 1k) and regular elongated ornamented cells on the entire abaxial side as well as on the stalk and the adaxial surface of the nectary operculum (Figure 1l). The flower size and total perianth organ number may vary within an inflorescence according to flower position.

image

Figure 1. Nigella damascena flower dimorphism and perianth organ morphologies at anthesis. (a, b) Open flowers of the petalous [P] morph (a) and the apetalous [T] morph. (c, d) [P] morph sepal (c) and petal (d). (e–h) [T] morph sepal (e), sepal-like organ (f), intermediate sepal-like trifid organ (g), and inner hybrid organ showing stamenoid and sepaloid characteristics (h). (i, j) Scanning electron microscopy of the adaxial (i) and abaxial (j) surfaces of the sepal and sepal-like organs shown in (c) and (e–h). (k, l) Scanning electron microscopy of the petal adaxial (k) and abaxial (l) sides. se, sepals; arrowhead, petal; asterisk, sepal-like organs. Scale bars = 25 μm (i,k) and 50 μm (j,l).

Download figure to PowerPoint

The less commonly observed floral morph in the Mornas population, hereafter referred to as the [T] morph, resembles the double-flowered variant described by Toxopéus (1927). While the five sepals in [T] morph are equivalent in structure and cellular composition to those found in the [P] morph (Figure 1e), the nectariferous petals are absent and are replaced by a variable number of sepal-like organs (Figure 1b). These include a range of shapes from outer-most lanceolate and entirely sepaloid (Figure 1f), to intermediate organs with bifid or trifid apices (Figure 1g), to inner-most hybrid organs (Figure 1h) that have mixed sepal and stamen characteristics, such as a thin filament-like stalk and a half-sporangiate half-sepaloid apical blade. The cellular composition in the inner perianth organs resembles that of the sepal epidermis, with conical cells on the adaxial side and pavement cells interspersed with stomata on the abaxial surface. The inner stamens as well as the carpels of [T] morph flowers are similar to those described above for the [P] morph. In addition to the production of sepal/stamen hybrid organs in the [T] morph, there is a significant increase in the total perianth organ number. Indeed, the number of perianth organs in the [T] morph (sepals + sepal-like) is significantly greater than that of the [P] morph (sepals + petals) ([P] 12.88 ± 1.02, [T] 23.42 ± 5.06, mean ± SD, < 0.001 × 10−16). Thus, although the loss of petals and production of sepal-like organs in their place may resemble a classical homeotic transformation, the N. damascena perianth polymorphism case is more complex, implying a shift in perianth organ number and perianth/stamen boundary disruption.

Perianth organ identity, perianth organ number and boundary shift are determined by a single locus

The N. damascena flower dimorphism studied by Toxopéus (1927) is determined by a single locus, with the petalous form being dominant over the apetalous one. We named this the P locus, and named the dominant allele responsible for the presence of petals the P allele; the homozygous pp combination produces no petals. In our study, organ identity transformation and the shift in organ number were always associated in the [T] morph. In order to confirm the monogenic determinism not only of the organ identity shift but of the whole complex-flower phenotype, we analyzed three F2 populations segregating for floral morph. We confirmed that the [P] and [T] proportions do not differ significantly from the 3:1 ratio expected in a dominance scenario (97 [P] to 41 [T], = 0.1941). Furthermore, the shift in petal identity, the presence of supplementary sepal-like organs and disruption of the perianth/stamen boundary were always associated and observed only in the [T] morph. A test cross of F2 [P] plants indicated approximately a 2:1 ratio of heterozygous to homozygous plants (76 Pp to 21 PP,= 0.0146), confirming the monogenic dominant nature of the flower dimorphism determinism, which encompasses not only perianth organ identity but also perianth organ number and perianth/stamen boundary regulation.

Comparison of floral development in the [P] and [T] morphs

During floral organogenesis in the [P] morph, four types of primordia are initiated in a centripetal way: crescent-shaped and fast growing, hemispherical with a break in development, hemispherical without a break in development, and upper-most horseshoe-shaped (Figure 2a–d) (Jabbour et al., 2009; Zhao et al., 2011). These four types of primordia become sepals, petals, stamens and carpels, respectively. In the present study, we broke down the early development of [P] morph floral buds into four stages. The first stage (S–I) corresponds to calyx initiation (Figure 2a). During the second stage (S–II), the primordia of petals and lower-most stamens initiate and are indistinguishable in shape and size (Figure 2b). At the third stage (S–III), petal primordia display a particular flattened shape and their development is delayed, while the remaining stamens are initiated and their development continues (Figure 2c). The fourth stage (S–IV) is characterized by carpel initiation (Figure 2d). By this time, petal and stamen primordia are easily distinguishable by their shape and differentiation state.

image

Figure 2. Scanning electron micrographs of the developmental sequence of Nigella damascena flowers with ([P]) and without ([T]) petals. (a) [P] Floral meristem at calyx initiation – stage S-I. (b) [P] floral but at initiation of petal and stamen primordia – stage S-II. (c) [P] floral bud showing petal developmental delay and continued initiation of stamen primordia – stage S-III. (d) [P] floral bud at carpel initiation, androecium differentiation and corolla development – stage S-IV. (e) [T] Floral meristem with initiated calyx – stage S-I. (f) [T] floral bud at initiation of sepal-like and stamen primordia – stage S-II. (g) [T] floral bud showing continued sepal-like organ and stamen primordia initiation and beginning of differentiation of the outer sepal-like organ primordia – stage S-III. (h) [T] floral but at carpel initiation, sepal-like organ and stamen differentiation – stage S-IV. br: bracts, se: sepals, arrowhead: petal primordium, asterisk: sepal-like organ primordium, c: carpels. Bars: 100 μm.

Download figure to PowerPoint

In contrast to the [P] morph floral development, only three visibly different types of primordia are initiated during [T] morph organogenesis in a centripetal way, i.e. crescent-shaped and fast growing, crescent-shaped to hemispherical, and upper-most horseshoe-shaped (Figure 2e–h). These three types of primordia become sepals, sepal-like organs or stamens, and carpels, respectively. The pre-anthetic development of the [T] morph was similarly broken down into four stages. The first stage (S–I) corresponds to calyx initiation (Figure 2e). During the second stage (S–II), undifferentiated future sepal-like organ and stamen primordia are initiated (Figure 2f). During the third stage (S–III), the outer-most of these primordia become crescent-shaped and develop in a sepal-like way (outer-most in Figure 2g), while the youngest inner-most primordia remain indistinguishable. The fourth stage (S–IV) is marked by the initiation of carpel primordia (Figure 2h). By this stage, a fair proportion of earlier initiated primordia may be identified as sepal-like organs, but, as the inner-most perianth organs share stamen-like characteristics, the identity of inner primordia and the point beyond which true stamen primordia begin cannot be fully discerned yet.

Comparison of expression patterns of APETALA3 and PISTILLATA homologs in [P] and [T] morphs

We investigated the Nigella damascena APETALA3 and PISTILLATA homolog expression patterns in developing floral organs of [P] and [T] from floral buds at stages > IV using RT–PCR. Our results confirm the previously observed petal-specific pattern of expression of the NdAP3–3 paralog during [P] morph floral development, and its complete absence from developing [T] perianth organs (Figure 3) (Zhang et al., 2013). Although a qualitative difference in expression was detected for the NdAP3–3 paralog, no significant differences in the NdAP3–1, NdAP3–2 and NdPI gene expression patterns were observed (Figure 3). The expression patterns of all four genes were further investigated during early flower development using in situ hybridization.

image

Figure 3. NdAP3–1, NdAP3–2, NdAP3–3 and NdPI locus-specific RT–PCR on RNA from dissected floral organs of N. damascena petalous ([P]) and apetalous ([T]) flower buds (5–6 mm diameter, stage > IV). Sep, sepals; Pet, petals; Spl, sepal-like organs; Sta, stamens; Car, carpels. Expression levels were normalized using the ACTIN gene.

Download figure to PowerPoint

In the [P] morph, NdAP3–1 expression is broadly detected at stage S–I (Figure 4a). Later, this expression becomes restricted to stamen primordia (Figure 4b–d). NdAP3–2 expression is not detected at stage S–I, but only later upon petal and stamen primordia initiation at stage S–II. It first appears in the primordia of these organs where it persists throughout floral development (Figure 4e–h). Expression of NdAP3–3 is first detected in a region of the undifferentiated floral meristem that encompasses the sites of the future petal and stamen primordia (Figure 4i). This expression persists as development proceeds, but becomes restricted to petal primordia at stage S–IV (Figure 4j–l). Transcripts of NdPI were strongly detected in the early floral meristem and in sepal, petal and stamen developing primordia, but not in the future carpel primordia region (Figure 4m–p).

image

Figure 4. In situ hybridization of Nigella damascena APETALA3 and PI homologs in petalous, [P] morph, floral meristems. (a–d) NdAP3–1, (e–h) NdAP3–2, (i–l) NdAP3–3, and (m–p) NdPI. (a, e, i) Early stage S–I floral meristems with developing sepal primordia. (f, m) Stage S–II meristems with undifferentiated petal or stamen primordia initiation. (b, g, i, k, n) Stage S–III meristems with delayed petal primordia. (c, d, h, l, o, p) Stage S–IV floral meristems with different developing organs up to initiation of the future carpels from the flat meristem top. se, developing sepals; arrowheads, petal primordia. Scale bars = 100 μm.

Download figure to PowerPoint

In accordance with the RT–PCR results, in situ hybridization for NdAP3–1, NdAP3–2 and NdPI in developing [T] floral meristems revealed mostly comparable expression patterns to those described for the [P] morph. While the NdAP3–1 paralog is absent from both developing petals and sepal-like organs (Figure 5b–d), and NdPI is similarly expressed in inner perianth organ primordia and stamens of both morphs (Figure 5m–p), NdAP3–2 is expressed in [P] developing petals but is absent from the apetalous developing perianth of [T] buds (Figure 5f–h). Most remarkably, NdAP3–3 transcripts were not detected at any stage in [T] floral buds, indicating that this paralog is not expressed at all in this morph (Figure 5i–l). The striking qualitative difference in NdAP3–3 expression patterns between the [P] and [T] morphs is highly suggestive of a specific role for this gene in the observed N. damascena perianth morphologies.

image

Figure 5. In situ hybridization of Nigella damascena APETALA3 and PI homologs in apetalous [T] morph floral meristems. (a–d) NdAP3–1, (e–h) NdAP3–2, (i–l) NdAP3–3, and (m–p) NdPI. (a,e) Early stage S–I floral meristems with sepal primordia. (i, m) Stage S–II meristems with few sepal-like organ and stamen undifferentiated primordia. (b-d, f, g, j, k, n, o) Stage S–III floral meristems with the earliest initiated sepal-like organ primordia starting differentiation (indicated by asterisks). (h,l,p) Stage S–IV floral buds at carpel initiation. se, sepals; asterisks, sepal-like organs. Scale bars = 100 μm.

Download figure to PowerPoint

A MITE insertion in the NdAP3–3 locus perfectly co-segregates with the P locus

Comparison of the genomic structure and sequence of the NdAP3–3 locus from one [T] and [P] plant from each available accession revealed no polymorphisms except for a 250 bp insertion in the second intron, which was homozygous in all [T] plants and either heterozygous or absent in [P] plants. This insertion bears all the characteristics of a type II non-autonomous transposable element, i.e. short size, terminal inverted repeats and target-site duplication. Based on these structural features, this transposable element is most likely a MITE (miniature inverted repeat transposable element). We investigated the segregation pattern of this insertion in the three segregating F2 populations for the flower morph. In all 136 plants, the insertion perfectly co-segregated with floral morph and genotype at the P locus (Table S1). These results confirm NdAP3–3 as a good candidate for the P locus, or suggest that it is in very close proximity to the responsible gene.

We investigated the possibility of an altered splicing pattern in the [T] morph induced by the presence of the MITE. Using various primer pairs and enhanced PCR, we were able to detect (at very low levels) two alternative transcripts in floral cDNA of the [T] morph, containing either a 3′ fragment of the second intron or the whole second intron including the MITE (Figure S1). No wild-type transcripts were found in this morph.

Functional validation of NdAP3–3

In order to demonstrate the role of NdAP3–3 in perianth organ identity and floral meristem patterning, we used a Tobacco rattle virus (TRV) based VIGS method to reduce its expression in [P] plants. Similarly to previous studies, we used a TRV2 construct containing a fragment of the ANTHOCYANIDIN SYNTHASE (ANS) gene from closely relates species Aquilegia vulgaris, which allows easy identification of effective silencing. Treatment with TRV2-ANS alone did not affect flower development or organ identity (Figure 6a), but generated an array of ANS silencing phenotypes that varied according to the timing of infection. In late and intermediate inoculations, treatment with TRV2-ANS led to completely white sepals and green petals (Figure 6a) or large white sectors in sepals and green sectors in petals. In contrast, flowers from early inoculations showed only small white or green sectors in sepals and petals,respectively, or a light dotted-like discoloration (as seen in Figure 6b,c), indicating that the efficiency of ANS silencing increased with the time of inoculation. In contrast, the effect of NdAP3–3 specific silencing decreased with the time of infection. For this reason, we did not use to ANS silencing phenotypes as a strict guide when detecting NdAP3–3 silenced flowers, but rather considered their organ morphology and organ number, and a priori knowledge of the [T] morph perianth architecture. Expression analysis of presumed transformed organs was used to confirm complete silencing. In addition to the customary VIGS controls, we also performed parallel inoculations in [T] morph plants. The ANS silencing phenotypes in [T] perianth organs were comparable to those observed in [P] plants. NdAP3–3 silencing had no effect on [T] perianth morphology and composition, as expected as no NdAP3–3 transcripts were found in this morph, suggesting that silencing is specific to the AP3–3 paralog.

image

Figure 6. Effect of NdAP3–3 virus-induced gene silencing on N. damascena petalous [P] plants. (a) Flower morphology under ANS silencing. (b–i) ANS-NdAP3–3 silencing phenotypes. (b, c) Partially (b) and completely (c) transformed flowers. (d–g) Various degrees of petal transformation. (h,j) Adaxial (h) and abaxial (j) epidermal cellular morphology observed in the organs shown in (e–g). Arrow, semi-transformed organs. Scale bars = 5 mm (d–g) and 50 μm (h,i).

Download figure to PowerPoint

Effect of NdAP3–3 silencing on perianth organ identity

Upon NdAP3–3 silencing, sepals were unaffected and petals showed a range of organ identity transformation phenotypes into sepal-like organs (Figure 6d–g). Petals on late and intermediate inoculated plants were partially transformed, keeping the same overall shape with a reduced operculum and semi-fused apical lobes in some flowers (Figure 6b,e), or having an elongated blade with a residual nectary, almost or completely fused lobes and no pseudonectaries in others (Figure 6f). All these semi-transformed petals already showed a transformation of cellular identity, having typical sepal cells on both the adaxial and abaxial sides (Figure 6h,i). Additionally, these organs do not express NdAP3–3 but express NdAP3–2, which may be used as a proxy for the switch from petal to sepal identity as this pattern is specifically detected in mature sepal and sepal-like organs (Figure 7). Finally, early NdAP3–3 silencing led to complete morphological transformation of petals into sepal-like organs with an elongated lanceolate single blade, and no sign of fused lobes, nectary crease or operculum (Figure 6g). Like the semi-transformed organs, completely transformed petals have a complete cellular and molecular sepal identity (Figure 7).

image

Figure 7. NdAP3–1, NdAP3–2, NdAP3–3 and NdPI locus-specific RT–PCR on RNA prepared from mature perianth organs from VIGS-treated N. damascena plants with ([P]) and without ([T]) petals. Sepals (Sep), sepal-like organs (Spl) and petals (Pet) from untreated plants (NO), ANS control plants (ANS) and TRV2-ANS-AP3–3 treated plants (AP33) showing mild silencing effects (st), strong silencing effects (ct) or no silencing effects (nt). Expression levels were normalized using the ACTIN gene.

Download figure to PowerPoint

Effect of AP3–3 silencing on perianth organ number

In early silenced plants, in which all petals showed complete morphological transformation, some flowers also showed supplementary perianth organs. This increased perianth organ number to values higher than for control [P] plants, overlapping the normal [T] morph values (Figures 6c and 8). The increase in perianth organ number was accompanied by production of hybrid organs with mixed sepal and stamen characteristics, indicating disruption of the perianth/stamen boundary (Figure 6c). These results confirm that the NdAP3–3 locus alone is responsible for the complex dimorphism, being implicated not only in the petal identity program but also in the control of perianth organ number and the perianth boundary with reproductive organs.

image

Figure 8. Frequency distribution of perianth organ number in N. damascena [P] (petalous) and [T] (apetalous) plants from various VIGS treatments. Control: pooled plants from TRV2 and ANS treatments. AP33: plants treated with the TRV2-ANS-AP3–3 construction. Early and late virus inoculation treatments are shown separately for [P] plants and pooled for [T] plants. The number of observations was 263 for [P] control, 283 for [T] control, 241 for [P] AP33 late, 282 [P] for AP33 early, and 195 for [T] AP33.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Despite being seldom observed in nature, polymorphisms affecting flower architecture may have significant evolutionary potential in the generation of species diversity, provided their effect on reproductive fitness does not significantly decrease the selective value (Hintz et al., 2006; Theißen, 2006). Such polymorphisms also provide valuable insight into the genetic origin of flower architecture. The naturally occurring Nigella damascena mutant described here falls in this category. It exhibits a complex phenotype, encompassing not only a change in the identity of inner perianth organs but also an increase in total perianth organ number and production of novel organ morphologies that suggest disruption of the perianth/stamen boundary. Using segregation analysis and gene silencing, we provide compelling evidence that a single gene, the NdAP3–3 paralog, is responsible for all aspects of this phenotype.

Monogenic control of the dimorphism and co-segregation with a MITE insertion at the NdAP3–3 locus

The complex N. damascena perianth phenotype of the apetalous morph is remarkably different from the classical B–function homeotic mutants described in core eudicot model species, in which petals and stamens are replaced by sepals and carpels, respectively (Sommer et al., 1990; Jack et al., 1992), and other basal eudicot models, where petals are replaced by sepaloid organs upon B–class gene inactivation but organ number is unaffected (Drea et al., 2007; Kramer et al., 2007; Sharma et al., 2011). Using segregation analysis, we confirmed that petal identity and perianth organ number are controlled by the same locus. A MITE inserted in the NdAP3–3 paralog was found to co-segregate completely with petal identity (this study and Zhang et al., 2013), perianth organ number and perianth/stamen boundary regulation (this study). Interestingly, the same NdAP3–3 mutant allele was found in all genetic origins that we studied, as well as those studied by Zhang et al. (2013), suggesting that a single mutation event is at the origin of the N. damascena dimorphism. The MITE insertion has been suggested to be the cause of the absence of expression (Zhang et al., 2013); however, although these transposable elements are preferentially found in the vicinity of plant genes, there is little compelling evidence for a role in gene regulation (Wessler et al., 1995; Casacuberta and Santiago, 2003; Feschotte, 2008). An alternative hypothesis is that the MITE insertion is simply a consequence of a pseudogenization process after an initial mutation, probably in a regulatory sequence.

Comparative floral development

Organ initiation, phyllotaxy and growth rate were previously described for the petalous N. damascena morph by Jabbour et al. (2009) and Zhao et al. (2011). Therefore, we chose to focus our developmental study on the short time window when petals/stamens and sepal-like organs/stamens are initiated and develop in the [P] and [T] morphs, respectively. While petal and stamen primordia in [P] morph floral buds appear morphologically identical at initiation, there is soon a delay in the developmental rate of petals, characterized by flattening of the primordia. This petal-specific developmental delay makes it possible to identify them with certainty while sepals and stamens progress in their development. Both the early morphological similarity between developing petal and stamen primordia, and the subsequent delay in petal development have been previously identified and described during floral development in several species of Ranunculaceae (Kosuge, 1994; Jabbour et al., 2009; Zhao et al., 2011).

Additionally, we present a detailed developmental study of the [T] floral morph. At early stages, all developing inner perianth organ primordia are morphologically identical and present on the same ontogenic spiral. Morphological differences gradually arise as primordia develop. However, because of the similarities between the inner-most sepal-like organ primordia and the outer-most true stamen primordia, it is impossible to determine the physical limit in the ontogenic sequence between perianth and stamen primordia. This situation is reflected in the corresponding adult flower, in which the boundary between perianth and stamens is blurred, as revealed by the occurrence of hybrid organs.

NdAP3–3 is responsible for the complex floral dimorphism through a dual role in petal identity and floral meristem patterning

In order to functionally validate the NdAP3–3 candidate as the locus responsible for the perianth dimorphism, we performed a VIGS experiment at three different time points relative to floral transition. Silencing specificity was confirmed by the absence of an effect on [T] morph flower development and morphology, or on the expression levels of other AP3 paralogs. Depending on the timing of gene silencing, various degrees of floral transformation were observed. While intermediate to late silencing only affected petal identity, early NdAP3–3 silencing resulted in transformation of petals into a range of sepal-like organs and an increase in the total perianth organ number, revealing a dual role for NdAP3–3 in petal identity specification and floral meristem patterning. This dual role parallels the particular expression pattern dynamics of NdAP3–3 observed during early [P] flower development.

At early stages of floral development, our in situ hybridization study revealed a broad expression domain of NdAP3–3, encompassing not only the region of future petal primordia but also adjacent upper regions that give rise to the stamen primordia. We hypothesize that this early broad expression pattern of NdAP3–3 reflects a role in meristem patterning, defining a ‘non-sepal’ morphogenetic domain and controlling perianth organ number, possibly via a cell proliferation repression function. In contrast, in Arabidopsis, AP3 is involved in cell proliferation promotion (Krizek and Meyerowitz, 1996), and a direct role for an AP3 homolog in cell proliferation repression has not been reported.

The early NdAP3–3 broad expression domain and its potential role in domain specification and regulation of organ number fits within the hypothesis of a regional specification function proposed for the ancestral AP3/PI genes (Drea et al., 2007). This ancestral function was suggested to account for the broad and dynamic expression domains of AP3/PI and their ability to specify both petaloid and non-petaloid inner perianth organs in basal angiosperms (Drea et al., 2007; Irish, 2009). Under this hypothesis, absence of NdAP3–3 expression leads to disruption of the perianth/stamen boundary simply by releasing the regional specification program constraint and allowing sepal and stamen identity programs to pervade the unregulated middle ground. The overlapping of sepal and stamen identity programs may then lead to production of transitional hybrid forms with mixed organ identities, much like the mechanism proposed in the fading borders model (Buzgo et al., 2004, 2005; Kim et al., 2005). This model is based on the observation of broad and overlapping expression domains of floral organ identity genes in basal angiosperms, which coincide with the presence of gradual transitions between organ forms, from sepals to petals and petals to stamens in flowers with spiral phyllotaxy (Kim et al., 2005). Alternatively, the boundary disruption in the absence of NdAP3–3 expression may be the result of a more direct role of this paralog in defining boundaries between floral organs by restricting C gene expression to the central region of the floral meristem. While in core eudicot models, this antagonistic role is performed by A–function genes, no A–class homologs have yet been found in Ranunculaceae (Litt and Irish, 2003). NdAP3–3 may perform this function by a mechanism similar to that described in Arabidopsis, whereby AP3 activates the cadastral gene SUPERMAN at the boundary between stamens and carpels (Bowman et al., 1992), but here at the petal/stamen boundary.

The complete transformation of petals into sepal-like organs, both in terms of morphology and cellular composition, is consistent with a role for NdAP3–3 in petal identity specification and morphogenesis. A petal identity role for the AP3–3 orthologs within the Ranunculaceae family has been strongly suspected based on expression pattern analysis in other petalous species, but has only been validated in Aquilegia so far (Rasmussen et al., 2009; Sharma et al., 2011). The restriction of the NdAP3–3 expression domain to petal primordia at the time of their developmental delay may reflect a cell proliferation inhibition function at the onset of the petal identity program, analogous to that suggested for the regional specification role. The sustained NdAP3–3 petal-specific expression throughout petal development supports an additional role in petal morphogenesis and cellular identity specification.

A particular dynamics process is evidenced between the petal identity and meristem patterning roles of NdAP3–3 by the early and late silencing effects. We observed [P] plants that showed complete silencing phenotypes until untransformed flowers reappeared, indicating later reactivation of NdAP3–3 due to the transient nature of VIGS. During this time frame, flowers with organ number transformation but no organ identity transformation (i.e. supplementary petals) were never observed. The absence of such phenotypes indicates an irreversibility of early NdAP3–3 silencing, and a dependency of its function in petal identity on its own early expression. This may either be because the petal identity program is dependent upon the floral meristem patterning role, or because the continued NdAP3–3 expression is dependent on its own downstream targets in a feedback loop mechanism.

Evolutionary perspectives regarding the NdAP3–3 dual role

An interesting issue to be considered is whether this dual role in floral meristem patterning and petal identity specification is specifically derived in Nigella damascena or is perhaps more largely distributed among Ranunculaceae. While a number of studies support an ancestral role for AP3–3 genes in petal identity specification (Rasmussen et al., 2009; Zhang et al., 2013), the lack of a previous report on floral patterning function leaves its evolutionary origin an open question. In Aquilegia vulgaris, the only other Ranunculaceae species to have been studied in detail, AP3–3 paralog expression appears to be restricted to petal primordia throughout floral development (Kramer et al., 2007). Functional studies using AP3–3-specific silencing in the petalous plant Aquilegia coerulea led to transformation of petals into sepals, without affecting the development of other floral organs (Sharma et al., 2011). Interestingly, whereas all floral organs in Nigella are inserted in a spiral, only sepals are spirally inserted in Aquilegia, while petals and stamens are whorled (Tucker and Hodges, 2005). The different consequences of AP3–3 absence of expression in the two species may reflect different developmental constraints imposed by the spiral versus whorled phyllotaxy. Another possibility is that AP3–3 plays no role in floral meristem patterning in Aquilegia. Because functional studies of the AP3–3 paralog have only been performed in two species, it is presently impossible to distinguish between a scenario in which the floral meristem patterning function was lost in Aquilegia or specifically acquired in Nigella. Unipartite perianths evolved several times independently of bipartite ancestors within the Ranunculaceae family (Rasmussen et al., 2009). Such events have been shown to occur in association with an absence of AP3–3 expression (Zhang et al., 2013). However, few cases of transitional organs have been described (Ronse de Craene and Brockington, 2013), questioning the involvement of this gene in regulation of the perianth/stamen boundary. Resolving such issues requires studying additional Ranunculaceae species for the expression and function of AP3–3 in a phylogenetic framework, in parallel with phylogenetic mapping of evolutionary transitions between spiral and whorled phyllotaxy, and between bipartite and unipartite perianths in relation to the presence or absence of transitional organ forms.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Plant material

Plant origin

Nigella damascena L. plants possessing flowers with and without petals ([P] and [T] floral morphs, respectively) originated from a natural population in Mornas (Vaucluse, South of France, accession 04–98 of the French National Museum of Natural History collection) and three cultivated commercial seed lots (Royalfleur (http://www.royalfleur.com/), Truffaut (http://www.truffaut.com/Réf. 46507) and Vilmorin (http://www.vilmorin.com/).

Production of a segregating population

Three F2 segregating populations were produced by selfing heterozygous [P] plants grown from commercial seed lots. Based on the monogenic dominant morph determination (Toxopéus, 1927), [T] F2 plants were considered to have the pp genotype. One flower from each [P] F2 plant was test-crossed using [T] pollen, and the flower morph of eight offspring per test cross was used to determine the original F2 [P] plant genotype (either PP or Pp). Fresh F2 plant material was sampled for DNA extraction and genotyping as described below.

Culture conditions

Plants were grown in greenhouse under long-day conditions (16 h day/8 h night) at 21°C during the day, 17°C during the night, and 60% relative humidity.

Floral morphology and scanning electron microscopy floral development study

Flowers at anthesis were observed in the greenhouse and photographed using an Olympus E410 camera (http://www.olympus-europa.com/). Individual organs were dissected from mature flowers and photographed using the Axio ZoomV16 stereomicroscope system (Zeiss, http://microscopy.zeiss.com/microscopy/en_de/products.html). Floral buds from successive developmental stages were sampled, fixed in 85 ml 55% ethanol, 5 ml glacial acetic acid, 10 ml formaldehyde, and stored in 70% ethanol. Buds were dissected under a MZ6 stereomicroscope (Leica Microsystems, http://www.leica-microsystems.com/products/), dehydrated in an ethanol series, and dried using an Emitech K850 critical-point dryer (Quorum Technologies, http://www.quorumtechnologies.com/products.html). Dried floral structures were mounted on aluminum stubs with colloidal graphite, sputter-coated with gold using a JFC–1200 fine coater (JEOL, http://www.jeol.co.jp/en/products/), and observed using a JSM–840A scanning electron microscope (JEOL) or a SU3500 scanning electron microscope (Hitachi, http://www.hitachi-hitec.com/global/em/index.html).

Candidate gene characterization

During the course of our study, partial coding sequences for the three Nigella damascena AP3 paralogs, as well as the genomic sequence for the NdAP3–3 locus, were published (Zhang et al., 2013). We obtained similar sequences for the three AP3 paralogs and the PI homolog using classical degenerate primer PCR and 5′- and 3′-RACE PCR methods as described in Methods S1.

Expression analysis

RT–PCR

Floral buds ranging from 5–6 mm diameter were dissected into sepals, petals, stamens and carpels for [P] buds, and sepals, sepal-like organs, stamens and carpels for [T] buds. Two biological repeats for each morph were prepared, each consisting of approximately ten dissected buds for each morph. Total RNA was extracted using a Qiagen (http://www.qiagen.com/products/) RNeasy kit according to the manufacturer's instructions. DNase treatment was performed using Ambion DNase I (Invitrogen, http://www.invitrogen.com/), and single-stranded cDNA was produced using SuperScript II reverse transcriptase (Invitrogen) and random hexamers (d(N)6) (Damerval et al., 2007). Possible DNA contamination was excluded by performing negative controls lacking reverse transcriptase using ACTIN-specific primers. Each gene was amplified using specific primers (Table S2). ACTIN was also used as a reference for cDNA quantity calibration among samples.

In situ hybridization

[P] and [T] flower buds were sampled at a range of developmental stages and fixed under vacuum in freshly prepared 4% paraformaldehyde. Dehydrated tissues were embedded in Paraplast Plus (McCormick Scientific, http://www.leicabiosystems.com/specimen-preparation/consumables/), and sectioned to 8 μm thickness (Damerval et al., 2007). Digoxigenin-labeled RNA antisense probes were synthesized from cDNA using T7 RNA polymerase (Riboprobe System T7, Promega, www.promega.com/products) and the primers listed in Table S2. Probes longer than 250 bp were hydrolyzed. Slide pre-treatment, pre-hybridization and hybridization were performed as described by Damerval et al. (2007). Hybridized sections were digitally photographed using a ProGres C10 camera (Jenoptik, http://www.jenoptik.com/en-digital-cameras-for-microscopy) mounted on a Microphot–FXA microscope (Nikon, http://www.nikon.com/products/instruments/index.htm). The contrast and brightness of photomicrographs was adjusted using ImageJ (ImageJ, U. S. National Institutes of Health, http://imagej.nih.gov/ij/) and Gimp software (Gimp, http://www.gimp.org/).

Virus-induced gene silencing

The TRV1 and TRV2 vectors containing the Aquilegia vulgaris ANTHOCYANIDIN SYNTHASE sequence (AqvANS) were kindly provided by E. Kramer, Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA 02138 (Kramer et al., 2007). A 305 bp fragment specific to the NdAP3–3 locus was amplified from the 3′ cDNA and UTR using primers designed to add BamHI and SacI restriction sites at the 5′ and 3′ ends, respectively (5′-GGATCCTGGACATTACAATTTACGACTGG-3′ and 5′-AGCTCTCCCAAACAAGGTCTACTTAATCCC-3′) (restriction sites underlined). This PCR product was purified and cloned using the pGEM–T Easy Vector system (Promega). The fragment was excised by double digestion with BamHI/SacI, and purified before ligation into a similarly digested TRV2-AqvANS construct. The TRV2-AqvANS-NdAP3–3 construct was transformed into Escherichia coli, positive clones were verified by PCR amplification, and plasmid was extracted for transformation into Agrobacterium tumefaciens strain GV3101. Separate liquid cultures for each construct (TRV1, TRV2, TRV2-AqvANS and TRV2-AqvANS-NdAP3–3) were grown overnight, and cells were collected by centrifugation (14 000 rpm for 10 min at 4°C) before resuspension in infiltration buffer (10 mm MES, 10 mm MgCl2, 100 mm acetosyringone) at a final absorbance of 2.0. The TRV1 solution was mixed with each of the other three solutions in equal volumes, and incubated for 3 h on ice prior to infiltration.

Three series of approximately 42 [P] and 28 [T] plants obtained from the selfing progeny of homozygous plants in the F2 segregating populations were used in three assays (early, intermediate and late). In each assay, a group of plants remained untreated (between one and two plants for each morph), a second group was inoculated with TRV1 and empty TRV2 (between four and six plants for each morph), a third group was inoculated with TRV1 and TRV2-AqvANS (between 12 and 14 [P] plants and six and 8 [T] plants), and finally a group of plants was inoculated with TRV1 and TRV2-AqvANS-NdAP3–3 (between 20 and 24 [P] plants and 14 [T] plants). Plants were treated by injection of 1 ml of solution with a needle syringe at the base of the stem, either at 6, 7 and 8 weeks after germination (late assay); 5, 6, 7, 9 and 10 weeks after germination (intermediate assay) or 4, 5, 7 and 9 weeks after germination (early assay).

Inflorescences were regularly inspected for signs of ANS or ANS/AP3–3 silencing. For each plant, 5–50 flowers were observed, and several phenotypic traits were recorded for each flower: ANS silencing phenotype in perianth organs, and perianth organ number including the number of wild-type-like, semi-transformed and transformed organs. Interesting phenotypes were photographed using an E410 Olympus camera or using a ProGres C10 camera (Jenoptik) mounted on a SMZ1500 stereomicroscope (Nikon). Freshly sampled floral organs were also documented using a SH–1500 scanning electron microscope (Hirox, http://www.hirox-europe.com/products/microscope/indexsem.html). Individual organs from flowers showing various degrees of silencing as well as unsilenced controls were sampled for RNA extraction and gene expression analysis as described above.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

We wish to thank Laurent Bonjean (UMR Génétique Végétale, Gif sur Yvette, France), Gilles Santé (Institut de Biologie des Plantes, Orsay, France) and Hervé Ferry (Institut Jean-Pierre Bourgin, Versailles, France) for plant care, Martine Le Guilloux (UMR Génétique Végétale, Gif sur Yvette, France) and Bernard Adroher (Institut Jean-Pierre Bourgin, Versailles, France) for technical assistance, Soizic Hupet (Institut Jean-Pierre Bourgin, Versailles, France) for VIGS observations, and the Plateforme de Microscopies et d'Imagerie (Muséum National d'Histoire Naturelle, Paris, France) where the scanning electron microscopy studies were performed. This work was supported by Fondation pour la Recherche sur la Biodiversité grant number AAS–IN–2009–040, and B.G. was financed by a PhD fellowship from the Ministere de l'Education Nationale de la Recherche et de Technologie.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Angenent, G.C., Franken, J., Busscher, M., van Dijken, A., van Went, J.L., Dons, H.J.M. and van Tunen, A.J. (1995) A novel class of MADS box genes is involved in ovule development in Petunia. Plant Cell, 7, 15691582.
  • Bowman, J.L., Smyth, D.R. and Meyerowitz, E.M. (1989) Genes directing flower development in Arabidopsis. Plant Cell, 1, 3752.
  • Bowman, J.L., Smyth, D.R. and Meyerowitz, E.M. (1991) Genetic interactions among floral homeotic genes of Arabidopsis. Development, 112, 120.
  • Bowman, J.L., Sakai, H., Jack, T., Weigel, D., Mayer, U. and Meyerowitz, E.M. (1992) SUPERMAN, a regulator of floral homeotic genes in Arabidopsis. Development, 114, 599615.
  • Buzgo, M., Soltis, P.S. and Soltis, D.E. (2004) Floral developmental morphology of Amborella trichopoda (Amborellaceae). Int. J. Plant Sci. 165, 925947.
  • Buzgo, M., Soltis, P.S., Kim, S. and Soltis, D.E. (2005) The making of the flower. Biologist, 52, 149154.
  • Casacuberta, J.M. and Santiago, N. (2003) Plant LTR-retrotransposons and MITEs: control of transposition and impact on the evolution of plant genes and genomes. Gene, 311, 111.
  • Coen, E. and Meyerowitz, E.M. (1991) The war of the whorls: genetic interactions controlling flower development. Nature, 353, 3137.
  • Damerval, C., Guilloux, M.Le., Jager, M. and Charon, C. (2007) Diversity and evolution of CYCLOIDEA-like TCP genes in relation to flower development in Papaveraceae. Plant Physiol. 143, 759772.
  • Drea, S., Hileman, L.C., de Martino, G. and Irish, V.F. (2007) Functional analyses of genetic pathways controlling petal specification in poppy. Development, 134, 41574166.
  • Feschotte, C. (2008) The contribution of transposable elements to the evolution of regulatory networks. Nat. Rev. Genet. 9, 397405.
  • Friedman, W.E. and Diggle, P.K. (2011) Charles Darwin and the origins of plant evolutionary developmental biology. Plant Cell, 23, 11941207.
  • Hintz, M., Bartholmes, C., Nutt, P., Ziermann, J., Hameister, S., Neuffer, B. and Theißen, G. (2006) Catching a ‘hopeful monster’: shepherd's purse (Capsella bursa-pastoris) as a model system to study the evolution of flower development. J. Exp. Bot. 57, 35313542.
  • Hu, J., Zhang, J., Shan, H. and Chen, Z. (2012) Expression of floral MADS-box genes in Sinofranchetia chinensis (Lardizabalaceae): implications for the nature of the nectar leaves. Ann. Bot. 110, 5769.
  • Irish, V.F. (2009) Evolution of petal identity. J. Exp. Bot. 60, 25172527.
  • Jabbour, F., Ronse de Craene, L.P., Nadot, S. and Damerval, C. (2009) Establishment of zygomorphy on an ontogenic spiral and evolution of perianth in the tribe Delphinieae (Ranunculaceae). Ann. Bot. 104, 809822.
  • Jack, T., Brockman, L.L. and Meyerowitz, E.M. (1992) The homeotic gene APETALA3 of Arabidopsis thaliana encodes a MADS box and is expressed in petals and stamens. Cell, 68, 683697.
  • Kim, S., Koh, J., Yoo, M.-J., Kong, H., Hu, Y., Ma, H., Soltis, P.S. and Soltis, D.E. (2005) Expression of floral MADS-box genes in basal angiosperms: implications for the evolution of floral regulators. Plant J. 43, 724744.
  • Kosuge, K. (1994) Petal evolution in Ranunculaceae. Plant Syst. Evol. 8(Suppl. 8), 185191.
  • Kramer, E.M., Dorit, R.L. and Irish, V.F. (1998) Molecular evolution of genes controlling petal and stamen development: duplication and divergence within the APETALA3 and PISTILLATA MADS-box gene lineages. Genetics, 149, 765783.
  • Kramer, E.M., Di Stilio, V.S. and Schluter, P.M. (2003) Complex patterns of gene duplication in the APETALA3 and PISTILLATA lineages of the Ranunculaceae. Int. J. Plant Sci. 164, 111.
  • Kramer, E.M., Holappa, L., Gould, B., Jaramillo, M.A., Setnikov, D. and Santiago, P.M. (2007) Elaboration of B gene function to include the identity of novel floral organs in the lower eudicot Aquilegia. Plant Cell, 19, 750766.
  • Krizek, B.A. and Meyerowitz, E.M. (1996) The Arabidopsis homeotic genes APETALA3 and PISTILLATA are sufficient to provide the B class organ identity function. Development, 122, 1122.
  • Litt, A. and Irish, V.F. (2003) Duplication and diversification in the APETALA1/FRUITFULL floral homeotic gene lineage: implications for the evolution of floral development. Genetics, 165, 821833.
  • Litt, A. and Kramer, E.M. (2010) The ABC model and the diversification of floral organ identity. Semin. Cell Dev. Biol. 21, 129137.
  • Pelaz, S., Ditta, G.S. and Yanofsky, M.F. (2000) B and C foral organ identity functions require SEPALLATA MADS-box genes. Nature, 405, 200203.
  • Rasmussen, D.A., Kramer, E.M. and Zimmer, E.A. (2009) One size fits all? Molecular evidence for a commonly inherited petal identity program in Ranunculales. Am. J. Bot. 96, 96109.
  • Ronse de Craene, L.P. and Brockington, S.F. (2013) Origin and evolution of petals in angiosperms. Plant Ecol. Evol. 146, 525.
  • Sharma, B., Guo, C., Kong, H. and Kramer, E.M. (2011) Petal-specific subfunctionalization of an APETALA3 paralog in the Ranunculales and its implications for petal evolution. New Phytol. 191, 870883.
  • Soltis, D.E., Ma, H., Frohlich, M.W., Soltis, P.S., Albert, V.A., Oppenheimer, D.G., Altman, N.S., DePamphilis, C.W. and Leebens-Mack, J. (2007) The floral genome: an evolutionary history of gene duplication and shifting patterns of gene expression. Trends Plant Sci. 12, 358367.
  • Sommer, H., Beltran, J.-P., Huijser, P., Pape, H. and Lonnig, W.–E., Saedler, H. and Schwarz-Sommer, Z., (1990) DEFICIENS, a homeotic gene involved in the control of flower morphogenesis in Antirrhinum majus: the protein shows homology to transcription factors. EMBO J. 9, 605613.
  • Stellari, G.M., Jaramillo, M.A. and Kramer, E.M. (2004) Evolution of the APETALA3 and PISTILLATA lineages of MADS-box-containing genes in the basal angiosperms. Mol. Biol. Evol. 21, 506519.
  • Theißen, G. (2006) The proper place of hopeful monsters in evolutionary biology. Theory Biosci. 124, 349369.
  • Theißen, G. and Melzer, R. (2007) Molecular mechanisms underlying origin and diversification of the angiosperm flower. Ann. Bot. 100, 603619.
  • Toxopéus, H.J. (1927) Erblichkeitsuntersuchungen an Nigella damascena L. Genetica, 9, 341440.
  • Tucker, S.C. and Hodges, S.A. (2005) Floral ontogeny of Aquilegia, Semiaquilegia, and Enemion (Ranunculaceae). Int. J. Plant Sci. 166, 557574.
  • Wessler, S.R., Bureau, T.E. and White, S.E. (1995) LTR-retrotransposons and MITEs: important players in the evolution of plant genomes. Curr. Opin. Genet. Dev. 5, 814821.
  • Zhang, R., Guo, C., Zhang, W. et al. (2013) Disruption of the petal identity gene APETALA3–3 is highly correlated with loss of petals within the buttercup family (Ranunculaceae). Proc. Natl Acad. Sci. USA 110, 50745079.
  • Zhao, L., Liu, P., Che, X.-F., Wang, W. and Ren, Y. (2011) Floral organogenesis of Helleborus thibetanus and Nigella damascena (Ranunculaceae) and its systematic significance. Bot. J. Linn. Soc. 166, 431443.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
tpj12284-sup-0001-FigS1.tifimage/tif3677KFigure S1. Alternative splicing of NdAP3–3 transcripts.
tpj12284-sup-0002-MethodS1.docxWord document15KTable S1. Segregation analysis of floral morph, P locus genotype and NdAP3–3 genotype.
tpj12284-sup-0003-TableS1.docxWord document11KTable S2. List of primers used in this study.
tpj12284-sup-0004-TableS2.docxWord document18KMethods S1. Characterization of candidate genes.
tpj12284-sup-0005-Legends.docxWord document13K 

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.