• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Jul 29, 2008; 105(30): 10414–10419.
Published online Jul 23, 2008. doi:  10.1073/pnas.0803291105
PMCID: PMC2492476
Developmental Biology

Genetic control of floral zygomorphy in pea (Pisum sativum L.)


Floral zygomorphy (flowers with bilateral symmetry) has multiple origins and typically manifests two kinds of asymmetries, dorsoventral (DV) and organ internal (IN) asymmetries in floral and organ planes, respectively, revealing the underlying key regulators in plant genomes that generate and superimpose various mechanisms to build up complexity and different floral forms during plant development. In this study, we investigate the loci affecting these asymmetries during the development of floral zygomorphy in pea (Pisum sativum L.). Two genes, LOBED STANDARD 1 (LST1) and KEELED WINGS (K), were cloned that encode TCP transcription factors and have divergent functions to constitute the DV asymmetry. A previously undescribed regulator, SYMMETRIC PETALS 1 (SYP1), has been isolated as controlling IN asymmetry. Genetic analysis demonstrates that DV and IN asymmetries could be controlled independently by the two kinds of regulators in pea, and their interactions help to specify the type of zygomorphy. Based on the genetic analysis in pea, we suggest that variation in both the functions and interactions of these regulators could give rise to the wide spectrum of floral symmetries among legume species and other flowering plants.

Keywords: symmetric petal, dorsoventral asymmetry, organ internal asymmetry, KEELED WINGS, LOBED STANDARD 1

Flower development in higher plants gives rise to an enormous variation of flower morphology and immense aesthetic diversification in nature. An important aspect for divergent floral developments is the establishment of floral symmetries, where a few distinct basic forms could be distinguished (1, 2): the monosymmetry (zygomorphy, with one symmetric plane), polysymmetry (actinomorphy, with several symmetric planes), and left–right asymmetry (with no symmetric plane). Among these, zygomorphy is considered the more specialized form and has been the most under investigation for its origin and underlying mechanisms.

Fabaceae (legumes) is one of the largest families in angiosperm, with a range of floral symmetric forms, and its success is thought to be coupled with its predominant zygomorphic flowers (3, 4). Most zygomorphic flowers are found in the subfamily Papilionoideae (5, 6), which attracted the attention of researchers since the end of the 18th century (7). Darwin (8) demonstrated the role of this type of zygomorphy in pollination biology, and the special floral shape of papilionoid legumes was an important factor in Mendel's groundbreaking work on the laws of genetic inheritance in the 1850s. Pea flowers, like most zygomorphic flowers, possess prominent corolla with three petal types, which are arranged along a dorsoventral (DV) axis, and manifest two types of asymmetries: DV asymmetry in the floral plane and organ internal (IN) asymmetry in the floral organ plane (Fig. 1a). It is well documented that DV asymmetry in papilionoid legumes commences in the floral meristem when the asymmetric development of floral organ primordia occurs (Fig. 2a) (5, 6). However, IN asymmetry is variable among petals: one dorsal petal (the standard) is IN symmetric, and two lateral (the wing) and two ventral petals are IN asymmetric (the two ventral petals are united on the lower edge and form a keel). This raises the question of how key regulators generate DV and IN asymmetries and superimpose them during zygomorphic development. The conspicuous zygomorphic flower of pea, for which there is a large collection of mutants, makes it a good model for exploring the key regulators to determine floral symmetry. For example, two loci, KEELED WINGS (K) and LOBED STANDARD 1 (LST1), were identified, respectively, in 1919 and 1985 (9, 10). The mutants at the LST1 locus give rise to the abnormal shape in the dorsal petals (Fig. 1c) (10, 11), whereas the k-1 mutant has been used as a morphological marker in genetic analysis for its homeotic transformation phenotype, i.e., the lateral petals mimic the ventral in size, shape, and color (Fig. 1b). K in pea could be an ortholog of KEELED WINGS in Lotus 1 (KEW1), because their mutants share a similar phenotype by bearing ventralized lateral petals (Fig. 1 b and f), and both are located in the syntenic regions in the pea and Lotus genomes, respectively (12). However, neither LST1 nor K has been cloned, partly because of the difficulty caused by the large genome size of pea.

Fig. 1.
Mutants affecting floral dorsoventral asymmetry in papilionoid legumes. Front and side views of flowers and flattened petals from wild type and mutants are shown. (a) Wild-type pea flower. Yellow arrow: two ventral petals form a keel. Red line with arrow: ...
Fig. 2.
Petal development in wild type and mutants. (a) Floral organ development stages 2, 4 and 5 (21, 22). At stage 4, two instead of one ventral sepals are initiated in syp1-1 floral meristem (yellow star). At stage 5, one dorsal petal and two stamens are ...

It has been shown that zygomorphy has multiple origins and plays a key role in the speciation and diversification in flowering plants (3, 13). Concomitant with their independent origins, variations in petal arrangement manifest in various superimpositions of the DV and IN asymmetries in distinct species. The flower in snapdragon (Antirrhinum majus) of the Asterid family represents another type of zygomorphy: an IN symmetric petal is positioned in the ventral, and two dorsal and two lateral petals with IN asymmetry are in pairs along the DV axis, in contrast to the arrangement in pea flower (12). Thus, the constitution of the two asymmetries may establish the developmental framework for different floral zygomorphies and give a clue as to the divergent actions of the underlying regulators among plant genomes. It has been shown that, in Antirrhinum, two closely related genes, CYCLOIDEA (CYC) and DICHOTOMA (DICH) (14, 15), are the key regulators that establish DV asymmetry, encode TCP transcription factors, and are expressed in the dorsal region of the floral meristem (16). These two regulators are also responsible for the elaboration of organ IN asymmetry and regulate and/or interact with two distinct MYB proteins, DIVARICATA (DIV) and RADIALIS (RAD), respectively, to determine lateral and ventral identities (1719). Despite the prominent difference in the zygomorphies between pea and Antirrhinum, recent studies in papilionoid legumes show that CYC orthologs play a key role in determining dorsal identity in two legume species, Lotus japonicus and Cadia (12, 20). Nevertheless, the CYC-like TCP genes, like DV regulators in distinct species, could have divergent functions apart from their common one in the control of DV identity: abolishing the activity of DV regulators in Antirrhinum gives rise to a default ventralized floral form, with all petals manifesting bilateral symmetry (the cyc dich flower, ref. 14), demonstrating that both CYC and DICH work together to determine IN asymmetry (15). However, a default ventralized form in papilionoid legumes should have all petals, which mimic the asymmetric shape of the ventral petal in wild type (12), suggesting that the CYC-like TCP genes might not be a prerequisite in the control of IN asymmetry in legumes. Therefore, different types of zygomorphies could be generated either by the divergent functions of the common key regulators and/or by the evolution of other distinct regulators.

In this study, we cloned K and LST1 loci and demonstrated that they are DV regulators in the control of lateral and dorsal identities, respectively, and that they originate from the duplication of an ancestral TCP gene during the speciation of papilionoid legumes. A locus, SYP1 (SYMMETRIC PETALS 1, named for its mutant flowers bearing all symmetric petals without normal IN asymmetry), was isolated, whose function is to establish IN asymmetry of petals. In the genetic analysis, three default floral forms with absent DV or/and IN asymmetries were identified in pea, demonstrating that DV and IN asymmetries can be independently controlled by distinct regulators, and that their interaction is important for the zygomorphic development. Taking these together, we propose that variation in both the entities and interactions of these regulators in the control of DV and IN asymmetries could give rise to the wide spectrum of floral symmetries among flowering plants.


k and lst1 Display Deficiency in the Development of DV Asymmetry.

The floral morphogenesis of pea mutants at two loci, K and LST1, was characterized by using SEM in the eight developmental stages according to previous studies (20, 21). k-1, k-2, and k-3 display the same phenotype, and no other detectable phenotype was observed apart from the ventralized petals at the lateral position (Fig. 1b). In lst1-1, lst1-2, and lst1-3, flowers bear abnormal standard petals with lobes whose sizes could be sensitive to growth conditions (Fig. 1c). The visible differences between wild type and k or lst1 commence at the same stage (stage 6), when the vascular tissues are developed in petals, and the asymmetric shape of both lateral and ventral petals becomes obvious (Fig. 2d). At this stage, the lateral petals in k mimic the ventral in shape, whereas the dorsal petal in lst1 is retarded in comparison with their wild-type counterparts (Fig. 2d). In mature flowers, alteration of lateral and dorsal identities in both k and lst1 is evident, as judged by the abnormal appearance of the lateralized and ventralized epidermal cell types of the lobed dorsal petals of lst1 and the lateral petals of k, respectively (Fig. 2c).

When lst1-1 was introduced into k-1 backgrounds, a high percentage of flowers displayed variable organ numbers in the dorsal region (Fig. 2b), with absent dorsal petal or more sepals being the most common (data not shown). In a comparison, ≈5% lst1-1 flowers displayed a similar phenotype. SEM analysis showed that primordium initiation was affected in the dorsal region (Fig. 2a), and the ventralized epidermal cell could be found in all petals (Fig. 2c). However, the asymmetric development of the early floral meristem was not altered in the double mutant (Fig. 2c). In the double mutants, not only were the numbers of stamen and sepal at the dorsal region affected, but also the length of the dorsal stamen could be longer than that of wild type (data not shown). Although most lateral petals displayed perfect ventralized forms, occasionally some lateral petals with conspicuous shape could be found (Fig. 1d). In Lotus, ≈3–5% flowers (often higher under stressed conditions) bear repetitive morphs of the lateral petal, rather than the most mirror-image forms in kew1, the putative ortholog of k (Fig. 1g), which manifest an abnormal left–right asymmetry. However, the majority of dorsal petals of k lst1 in pea failed to fully expand and therefore were smaller in size, displaying IN asymmetry to varying degrees [Fig. 1d; supporting information (SI) Fig. S1]. Nevertheless, a small portion of flowers bore the dorsal petals, which mimicked the shape of the ventral petal (Fig. 1d), representing a default state without DV asymmetry. Thus, during zygomorphic development, K and LST1 play a key role in establishing lateral and dorsal identities, respectively, and their interaction for the development of dorsal organs is revealed in the double mutant.

Both K and LST1 Encode CYC-Like TCP Proteins.

The comparative genomics approach was conducted to clone K and LST1 in pea. Previous genetic analysis has mapped the K locus in linkage group II (23). In a parallel mapping experiment in Lotus, kew1 was located into a 295-kb region, where a complete contig and DNA sequence were subsequently obtained (Fig. 3a). However, no recombination haplotype was found in the large mapping population within the 200-kb region containing KEW1, and sequence analysis revealed a candidate gene within the contig, which is predicted to encode a CYC-like TCP protein with 370 aa and has been reported as LjCYC3 (12). However, there is no sequence alteration in a 4-kb region of kew1 containing LjCYC3, but whose expression level was found to be decreased (data not shown). Thus, three CYC homologs, designated PsCYC1, PsCYC2, and PsCYC3, were cloned in pea (GenBank accession nos. EU574913, EU574914, and EU574915). Phylogenetic analysis indicated PsCYC3 is the ortholog of LjCYC3 (Fig. 3b). When the gene structure of PsCYC3 was analyzed, deletions were found in both k-1 and k-3 (Fig. 3c), and a single base deletion was detected at position 350 of PsCYC3 ORF in k-2, which should cause a frame-shift and disrupt the conserved TCP domain (Fig. S2a). However, LST1 was mapped and located in the linkage group VI of pea between two SSR markers, AD51 and AA200 (24). We noticed that lst1 exhibits a similar phenotype to the LjCYC2 mutant, squared standard 1 (squ1) in Lotus (12): both mutants have the dorsal petal phenotype and can interact with k or kew1, respectively, by giving rise to the ventralized character in all petals. Thus, the ortholog of LjCYC2, PsCYC2, was subject to detailed analysis, and an STS marker for PsCYC2 was found to cosegregate with lst1 in our mapping population (data not shown). When the sequence of PsCYC2 was analyzed, a single base substitution (C-437-T) was found in lst1-1, which would cause an amino acid substitution (S-146-L) in the conserved TCP domain (Fig. S2a), and deletion of PsCYC2 was found in lst1-3 (data not shown). Thus, the conserved site of 146S in the TCP domain could be important for protein function. To verify PsCYCs function, a virus-inducible gene silencing assay, based on Pea early browning (PEVB-VIGS), was conducted (25). VIGS-PsCYCs silencing constructs containing different fragments of PsCYC genes (Fig. S2a) were applied to wild-type and different mutant plants, respectively, and semiquantitative RT-PCR was conducted to confirm the reduced expression of target genes (data not shown). In VIGS-PsCYC3 silenced wild-type plants, ≈36.5% and 57.1% of flowers displayed partial and complete conversion of wings into the shape of keels, respectively (Fig. 3 h and i). In VIGS-PsCYC2 silenced plants, ≈30% of flowers displayed lobed or retarded standards (Fig. 3k). VIGS-PsCYC3 silenced lst1 and VIGS-PsCYC2 silenced k plants phenocopied the k lst1 double mutant to various extents (data not shown). When VIGS-PsCYC1 was tested, no phenotype was observed in VIGS-PsCYC1 silenced wild-type plants. No additive or novel phenotypes were found in the VIGS-PsCYC1 silenced lst1 or k lst1 plants. However, in VIGS-PsCYC1 silenced k plants, a portion of flowers with retarded standards were found (Fig. 3l), suggesting PsCYC1 could have a redundant function with LST1 in the interaction with K to control dorsal petal development. Thus, we concluded that both K and LST1 are cloned and encode CYC-like TCP proteins.

Fig. 3.
Cloning of K and LST1 genes. (a) Physical map of the 295-kb region in Chr. 5 of Lotus, where KEW1 was located (upper lane), and the gene structure of LjCYC3 is shown in the lower lane. (b) Phylogenetic tree of TCP family members. CYC ( ...

The expression pattern of these DV regulators was analyzed, and all were found to express in the apex of wild type during floral development, sharing an asymmetric expression pattern in the dorsal organs of the flower. Both PsCYC1 and LST1 are expressed in standards, whereas the transcripts of K can be found in both dorsal and lateral petals (Fig. 3d). As expected, no transcript of K and LST1 was detected in the deletion mutants of k or lst1, respectively; however, the transcription levels of K and LST1 bearing the point mutation were not down-regulated (Fig. 3 e and f). In RNA in situ hybridization experiments, it was found that both PsCYC1 and LST1 are expressed first in the dorsal region of the floral meristem before floral organ initiation and then inside dorsal petals (Fig. S2 b–i). We failed to detect the transcripts of K by RNA in situ hybridization, presumably because of its low expression level. CYC homologs in legumes have been characterized into two major groups, LEGCYC groups I and II (26, 27), and phylogenetic analysis placed PsCYC1 and LST1 into LEGCYC group I and K into group II (Fig. 3b). Thus, the overlapped expression domains and different expression levels are consistent with the functional divergence and phylogenetic relations of these DV regulators.

Mutant syp1 Gives Rise to a Defect in IN Asymmetry in Petals.

The default state without DV asymmetry in the k lst1, in which all petals are ventralized and manifest IN asymmetry (Fig. 1d), suggests other regulators exist, which should control IN asymmetry independent of K/LST1 function in pea. To target the hypothetical IN regulators, we conducted the screen for mutants bearing petals with altered IN asymmetry. In two large-scale mutagenesis experiments, we identified the SYP1 locus with two mutated alleles (Fig. 4 b and c). In syp1-1, nearly all petals are bilaterally symmetrical but maintain their DV identities (Fig. 4b), and the most conspicuous characteristic is that the symmetric ventral petals possess a keel structure, in contrast to the wild-type flower, whose two ventral petals form a keel (Fig. 4 a and b). However, syp1-2, presumably a weaker allele, has a highly variable effect on IN asymmetry of the lateral but not the ventral petals (Fig. 4c). The SYP1 locus has been located in linkage group II of pea, and its syntenic region is anchored in both Lotus and Medicago genomes, to conduct the cloning work (data not shown).

Fig. 4.
Interaction between PsCYCs and SYP1, floral diagrams, and the establishment of floral zygomorphy. (a–f) syp1 mutants and syp1-1 in different genetic backgrounds. Front and side views of flowers and flattened petals from wild type and mutants are ...

In spy1-1, approximately one-third of the flowers have increased organ numbers in the ventral region (Fig. 2b), where abnormal primordium initiation was found during early development of the floral meristem (Fig. 2a). In syp1-1, the abnormal bilaterally symmetric shapes of the lateral and ventral petals can be observed at petal developmental stage 6, when the malfunction of petal development also occurs in k and lst1 (Fig. 2d). However, the whole floral meristem became more symmetrical than in wild type from the beginning of floral organ primordium development (Fig. 2a). When the epidermal cells of the syp-1 petals were analyzed, both cell size and type were found to be the same as their wild-type counterparts (Fig. 2c). Therefore, the syp1-1 flower represents a default state without IN asymmetry in all petals, in contrast to the one without DV asymmetry in the k lst1.

SYP1 and K/LST1 Are Antagonistic at the Early Stage of Floral Development.

To investigate how DV and IN regulators interact with each other during zygomorphic development, syp1-1 was introduced into k, lst1, and k lst1 genetic backgrounds, respectively. The lst1 syp1-1 double mutant displayed an additive phenotype: lobed standards and petals in the lateral and ventral positions without IN asymmetry (Fig. 4d). However, k syp1 flowers possess abnormal standards with an altered shape apart from the expected keels in both lateral and ventral regions (Fig. 4e), revealing a hidden function of SYP1 during the development of the dorsal petal. In the k lst1 syp1-1 triple mutant, the flowers display a radial symmetry (Fig. 4f), and all petals possess ventralized identity with a keel structure. Thus, a default form without DV and IN asymmetries was identified.

In the triple mutant, no other detectable phenotype was found apart from floral symmetry. However, there was little variation in floral organ numbers, in contrast to a high portion of k lst1 flowers with variable organ numbers, in the dorsal region, whereas that in the ventral region is notably reduced in comparison with the one in syp1-1, indicating that the malfunction of organ primordium initiation in k/lst1 and syp1-1 is completely or partially suppressed in the triple mutant. The expression patterns of K, LST1, and PsCYC1 were analyzed in syp1-1 by RT-PCR, and no detectable alteration was found (data not shown). Thus, the antagonistic interaction of DV and IN regulators is found at the early stage during floral development when floral organ primordia initiate.


Zygomorphic development, like other pattern formations in plants and animals, involves the establishment of different body planes with distinct developmental axes, where various asymmetries are generated and superimposed under the control of different regulators. The pea possesses a conspicuous zygomorphic flower with an abundant mutant collection and provides an ideal experimental system to analyze the key regulators in control of floral symmetry. However, its complex large genome (5,000 Mb, 10× the size of either Lotus or Medicago, with >90% repetitive sequence) had obscured progress in molecular analysis (28). By adopting a comparative genomics approach and other molecular genetic tools, we successfully cloned two genes in pea, K and LST1, demonstrating that the limited sequence information, large size, and complex genome are no longer impassable barriers for the molecular study of pea.

In this work, the functions of K and LST1 and PsCYC1 are characterized, which comprise a small CYC-like TCP gene cluster and originate from two duplication events of an ancient TCP gene (26, 27). Because the deletion alleles at both K and LST1 loci show no other detectable malfunction apart from the floral symmetry, it is evident that these TCP genes have been destined for the control of zygomorphic development. Their expression patterns were found to be overlapped but differ spatially and quantitatively, which is consistent with the divergent functions of LST1 and K to determine the dorsal and lateral identities of petals, respectively. It is also consistent with previous reports that alteration of expression pattern and level of TB1, a CYC-like TCP gene in maize, could lead to novel morphogenesis (29, 30). In a comparison, CYC and its close duplicate DICH in Antirrhinum determine DV asymmetry and regulate two distinct small MYB proteins, which are involved in the control of lateral and ventral identities (14, 15, 1719). These indicate that increasing the copy number of TCP genes should be a key step for elaborating their roles in the control of floral DV asymmetry, and then different copies acquire their divergent functions as different DV regulators during zygomorphic evolution. Thus, the independent duplication events giving rise to divergent DV regulators in different species could account for the molecular basis on the development of different types of floral zygomorphies.

The conspicuously different types of zygomorphic flowers in Antirrhinum and pea demonstrate how different origins of regulators could have resulted in distinct flower forms. When DV regulator function is abolished, the default state without DV asymmetry in both species appears to be a ventralized form, and all petals acquire ventral identity (Fig. 4g). However, ventral identities in the two types of zygomorphic flowers have distinct properties: the ventral petal in Antirrhinum displays a bilaterally symmetric shape, whereas that in pea possesses IN asymmetry (Fig. 4g). An IN regulator, SYP1, should be responsible for IN asymmetry of petals when DV regulators are mutated in pea. In syp1-1, where IN asymmetry is abolished, all petals are bilaterally symmetric, but their DV identities are maintained (Fig. 4b), revealing another default state without IN asymmetry. Therefore, DV and IN regulators can separately generate different asymmetries in pea (Fig. 4g). In contrast to Antirrhinum, two DV regulators, CYC and DICH, interplay to modify IN asymmetry of the dorsal petal. Thus, our findings offer evidence to support the notion that organ IN asymmetry also has multiple origins. It is likely that the independent genetic control of DV and IN asymmetry may not be unique in pea, and two relevant pathways could have evolved independently in other species as well. Therefore, recruiting only one or both two kinds of regulators in the control of DV and IN, respectively, could give rise to various floral forms (Fig. 4g). Furthermore, variation or modification of the interaction between DV and IN regulators could generate floral forms with different symmetries, such as the left-right asymmetric flowers in k lst1 in pea or kew1 in Lotus (Fig. 1 d and g) and the radial symmetrical flower in Cadia, where the expanding expression pattern of LegCYC was found (20).

The existence of the two reciprocal default forms in pea with only one DV or IN asymmetry raises the question of how the two asymmetries are superimposed and coordinated during zygomorphic development. In this study, the dorsal petal morphogenesis of pea provides a unique example to examine the interaction between IN and DV factors, because it possesses bilateral symmetry. It has been shown that ectopic expression or the altered expression domain of CYC-like TCP genes in papilionoid legumes can suppress the manifestation of IN asymmetry in lateral and ventral petals by altering their identity (12, 20). Consistently, the bilaterally symmetric shape of dorsal petal in pea can become asymmetric when both K and LST1 are mutated (Fig. 1d), indicating that SYP1 action is suppressed in the dorsal petal, where DV regulators are expressed. Other data also support this interaction: The abnormality of dorsal primordium initiation in the k lst1 is suppressed by introducing syp1-1 (Fig. 2a), and the morphology of the dorsal petal is altered in the k syp1, whereas its symmetric shape is otherwise maintained (Fig. 4e). The interaction is not limited to the dorsal petal. For example, in syp1-1, floral organ initiation in the ventral region is affected and can be partially suppressed in the k lst1 syp1 triple mutants, suggesting that both DV and IN regulators participate in organ initiation in the two polar regions of the floral meristem. Previous studies show that a TCP protein in Arabidopsis should be involved in the regulation of cell division (31). It is most likely that both DV and IN regulators could participate in similar biological function to regulate cell division and differentiation in floral meristem and floral organ primordia. Thus, both the divergent functions and the delicate interaction between the two kinds of regulators are essential to regulate zygomorphic development in pea (Fig. 4h).

Using examples from a range of pea floral mutants (Fig. 4g), we suggest that both DV and IN regulators could have arisen through independent evolutionary routes in Papilionoideae and even in other species. Consequently, possession of only one or two types of regulators involved in the control of floral symmetry and the variation or modification of the interaction between them have the potential to generate the wide spectrum of floral forms with varied symmetries found in Fabaceae and among other flowering plants as well.

Materials and Methods

Plant Material and Growth Conditions.

All pea lines used in this study were obtained from the John Innes Pisum Germplasm Collection, except syp1-1. k-2 arose by x-ray mutagenesis, and k-3 was derived from fast neutron mutagenesis in JI116 and JI2296, respectively. lst1-2, lst1-3, and syp1-2 were obtained by fast neutron mutagenesis in JI2822. syp1-1 was identified from fast neutron mutagenesis in Térèse. Allelisms were confirmed by crosses between k-1 and k-2, k-1 and k-3, lst1-1 and lst1-2, lst1-1 and lst1-3, and syp1-1 and syp1-2, respectively. All plants were grown at 18–20°C with a 16-h light/8-h dark photoperiod at 150 mE·m−2·s−1.


Nonradioactive in situ hybridization was performed essentially as described (32). SEM for mature petals was performed on plastic replicas as described (33). SEM for floral buds was prepared as described (34). Samples were examined in JEOL JSM-6360LV (JEOL).

Gene Cloning and RT-PCR.

Primer sets used for amplifying PsCYC1 (SL0842/G3848), LST1 (SL0773/SL0970), and K (SL1254/SL1103) genes were designed by sequence analysis of ESTs from Medicago truncatula and CYC homologs in Lotus. Homologous alignments were performed by using the ClustalX program (version 1.83), and phylogenetic trees were computed by using the Phylip program (version 3.6). Primer pairs SL1098/SL1099, SL1100/SL1101, and SL1102/SL1103 were used for amplification of the PsCYC1, LST1, and K transcripts, respectively. Histone H4 (GenBank accession no. U10042) was amplified with primers SL1815/SL0363 as an internal control. For in situ probes, PsCYC1 and LST1 transcripts were generated with primer sets SL0932/SL0933 and SL0868/SL0970, respectively, from cDNA fragments. Semiquantitative RT-PCR was performed as described (12). Genomic DNA was digested with HindIII before the genomic Southern blot analysis.

Mapping and Cosegregation Analysis.

kew1 was mapped as described (12). Two SSR markers, AD51 and AA200 (24), were used to locate LST1 in linkage group VI of pea. A STS marker (LjSSR1084 and LjSSR1085) for PsCYC2 was used to conduct the cosegregation test for lst1 in a lst1-1 × JI992 population (n = 280). The cosegregation of PsCYC3 with K was performed in a 140 k-1 × JI992 population by using primers SL1102/SL1103.

Virus-Induced Gene Silencing (VIGS) Assay.

The VIGS assay in pea was carried out as described (25). The fragments being used for the constructs, VIGS-PsCYC1, -2, and -3, are marked in Fig. S2a. Fifteen 2-week-old plants for each construct were agroinoculated and repeated three times independently.

Supplementary Material

Supporting Information:


We thank Dr. Ida Elisabeth Johansen (Danish Institute of Agricultural Sciences, Frederiksberg, Denmark) for providing the pCAPE1, pCAPE2-PDS, and pCAPE2-GFP constructs for VIGS. We acknowledge H. Xia, H. Yin, S. Hao, Z. Xu, J. Yan, and C. Li for critical comments on the manuscript. We also thank Z. Xu, J. Li, Y. Xue, G. Wang, Y. Tian, H. Lin, L. Zhuang, and Y. Liu for encouragement and support for this experiment. This work was supported by the National High Technology Research and Development Program of China (Grant nos. 2006AA10A110 and 2007AA10Z113), the National Nature Science Foundation of China (Grant nos. 30430330 and 30528016), and FP6-2002-FOOD-1-506223 (Grain Legumes) from the European Commission.


The authors declare no conflict of interest.

This article is a PNAS Direct Submission.


1. Endress PK. Symmetry in flowers: Diversity and evolution. Int J Plant Sci. 1999;160:S3–S23. [PubMed]
2. Endress PK. Evolution of floral symmetry. Curr Opin Plant Biol. 2001;4:86–91. [PubMed]
3. Cubas P. Floral zygomorphy, the recurring evolution of a successful trait. BioEssays. 2004;26:1175–1184. [PubMed]
4. Doyle JJ, Luckow MA. The rest of the iceberg. Legume diversity and evolution in a phylogenetic context. Plant Physiol. 2003;131:900–910. [PMC free article] [PubMed]
5. Tucker SC. In: Contemporary Problems in Plant Anatomy. White RA, Dickison WC, editors. New York: Academic; 1984. pp. 354–395.
6. Tucker SC. Floral development in legumes. Plant Physiol. 2003;131:911–926. [PMC free article] [PubMed]
7. Sprengel CK. Berlin: Vieweg; 1793. Das Entdeckte Geheimniss der Natur im Bau und in der Befruchtung der Blumen.
8. Darwin CR. Bees and fertilisation of kidney beans. Gard Chron Agric Gaz. 1857;43:725.
9. Pellew C, Sverdrup A. New observations on the genetics of peas. J Genet. 1923;13:125–131.
10. Winfield P. A new gene (lst) which influences standard shape in Pisum sativum. Pisum Newslett. 1987;19:84–85.
11. Ambrose MJ. Further observations on the expression of lobed standard (lst) and keeled wings (k) and their involvement in petal identity. Pisum Genet. 2003;35:1–2.
12. Feng X, et al. Control of petal shape and floral zygomorphy in Lotus japonicus. Proc Natl Acad Sci USA. 2006;103:4970–4975. [PMC free article] [PubMed]
13. Sargent RD. Floral symmetry affects speciation rates in angiosperms. Proc Biol Sci. 2004;271:603–608. [PMC free article] [PubMed]
14. Luo D, Carpenter R, Vincent C, Copsey L, Coen E. Origin of floral asymmetry in Antirrhinum. Nature. 1996;383:794–799. [PubMed]
15. Luo D, et al. Control of organ asymmetry in flowers of Antirrhinum. Cell. 1999;99:367–376. [PubMed]
16. Cubas P, Lauter N, Doebley J, Coen E. The TCP domain: A motif found in proteins regulating plant growth and development. Plant J. 1999;18:215–222. [PubMed]
17. Almeida J, Rocheta M, Galego L. Genetic control of flower shape in Antirrhinum majus. Development. 1997;124:1387–1392. [PubMed]
18. Galego L, Almeida J. Role of DIVARICATA in the control of dorsoventral asymmetry in Antirrhinum flowers. Genes Dev. 2002;16:880–891. [PMC free article] [PubMed]
19. Corley SB, Carpenter R, Copsey L, Coen E. Floral asymmetry involves an interplay between TCP and MYB transcription factors in Antirrhinum. Proc Natl Acad Sci USA. 2005;102:5068–5073. [PMC free article] [PubMed]
20. Citerne HL, Pennington RT, Cronk QC. An apparent reversal in floral symmetry in the legume Cadia is a homeotic transformation. Proc Natl Acad Sci USA. 2006;32:1217–1220. [PMC free article] [PubMed]
21. Tucker SC. Overlapping organ initiation and common primordia in flowers of Pisum sativum. Am J Bot. 1989;76:714–729.
22. Ferrandiz C, Navarro C, Gomez MD, Canas LA, Beltran JP. Flower development in pisum sativum: from the war of the whorls to the battle of the common primordia. Dev Genet. 1999;25:280–290. [PubMed]
23. Ellis THN, Poyser SJ. An integrated and comparative view of pea genetic and cytogenetic maps. New Phytol. 2002;153:17–25.
24. Loridon K, et al. Microsatellite marker polymorphism and mapping in pea (Pisum sativum L) Theor Appl Genet. 2005;111:1022–1031. [PubMed]
25. Constantin GD, et al. Virus-induced gene silencing as a tool for functional genomics in a legume species. Plant J. 2004;40:622–631. [PubMed]
26. Citerne HL, Luo D, Pennington RT, Coen E, Cronk QC. A phylogenomic investigation of CYCLOIDEA-like TCP genes in the Leguminosae. Plant Physiol. 2003;131:1042–1053. [PMC free article] [PubMed]
27. Fukuda T, Yokoyama J, Maki M. Molecular evolution of cycloidea-like genes in Fabaceae. J Mol Evol. 2003;57:588–597. [PubMed]
28. Choi HK, et al. Estimating genome conservation between crop and model legume species. Proc Natl Acad Sci USA. 2004;101:15289–15294. [PMC free article] [PubMed]
29. Doebley J, Stec A, Hubbard L. The evolution of apical dominance in maize. Nature. 1997;386:485–488. [PubMed]
30. Wang RL, Stec A, Hey J, Lukens L, Doebley J. The limits of selection during maize domestication. Nature. 1999;398:236–239. [PubMed]
31. Li C, Potuschak T, Colon-Carmona A, Gutierrez RA, Doerner P. Arabidopsis TCP20 links regulation of growth and cell division control pathways. Proc Natl Acad Sci USA. 2005;102:12978–12983. [PMC free article] [PubMed]
32. Coen ES, et al. floricaula: A homeotic gene required for flower development in Antirrhinum majus. Cell. 1990;63:1311–1322. [PubMed]
33. Green PB, Linstead P. A procedure for SEM of complex shoot structures applied to the inflorescence of snapdragon (Antirrhinum) Protoplasma. 1990;158:33–38.
34. Chen C, Wang S, Huang H. LEUNIG has multiple functions in gynoecium development in Arabidopsis. Genesis. 2000;26:42–54. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...