|Home | About | Journals | Submit | Contact Us | Français|
The inflorescence of grass species such as wheat, rice and maize consists of a unique reproductive structure called the spikelet, which is comprised of one, a few, or several florets (individual flowers). When reproductive growth is initiated, the inflorescence meristem differentiates a spikelet meristem as a lateral branch; the spikelet meristem then produces a floret meristem as a lateral branch. Interestingly, in wheat, the number of fertile florets per spikelet is associated with ploidy level: one or two florets in diploid, two or three in tetraploid, and more than three in hexaploid wheats. The objective of this study was to identify the mechanisms that regulate the architecture of the inflorescence in wheat and its relationship to ploidy level.
The floral anatomy of diploid (Triticum monococcum), tetraploid (T. turgidum ssp. durum) and hexaploid (T. aestivum) wheat species were investigated by light and scanning electron microscopy to describe floret development and to clarify the timing of the initiation of the floret primordia. In situ hybridization analysis using Wknox1, a wheat knotted1 orthologue, was performed to determine the patterning of meristem formation in the inflorescence.
The recessive natural mutation of tetraploid (T. turgidum ssp. turgidum) wheat, branching head (bh), which produces branched inflorescences, was used to demonstrate the utility of Wknox1 as a molecular marker for meristematic tissue. Then an analysis of Wknox1 expression was performed in diploid, tetraploid and hexaploid wheats and heterochronic development of the floret meristems was found among these wheat species.
It is shown that the difference in the number of floret primordia in diploid, tetraploid and hexaploid wheats is caused by the heterochronic initiation of floret meristem development from the spikelet meristem.
Organogenesis in the plant shoot is controlled by a group of stem cells (called the meristem) at the shoot apex. In grass species such as rice (Oryza sativa), wheat (Triticum aestivum) and maize (Zea mays), when the growth stage switches from the vegetative phase to the reproductive phase, the shoot apical meristem stops differentiating leaves and gradually elongates and changes into an inflorescence meristem (IM). The architecture of the inflorescence is strongly influenced by the timing of meristem initiation and the hierarchical structures of different types of meristem. The spikelet is a reproductive unit unique to the grass inflorescence (Schmidt and Ambrose, 1998; Bommert et al., 2005). The spikelet is comprised of florets and is encompassed by two small bract leaves (called glumes in wheat) (Murai et al., 2002). During inflorescence development, the inflorescence meristem produces spikelet meristem (SM), and the SM produces the floret meristem (FM). Finally, the FM produces the floral organs.
Cereal crops are the major source of carbohydrates for humans, and the yield of grain from these crops is largely dependent on inflorescence architecture (Ashikari et al., 2005). Studies on the genetic regulation of inflorescence formation have provided valuable information for crop breeding. Molecular genetic studies have identified a number of genes involved in controlling inflorescence branching in maize and rice (Bommert et al., 2005). The inflorescence of maize and rice is more complicated than that of wheat because of the presence of additional axillary branch meristems: the tassel branch and spikelet pair meristem in maize, and the panicle branch meristem in rice (Schmidt and Ambrose, 1998; Shitsukawa et al., 2006). In wheat, branching of the inflorescence is regulated by two types of meristem, SM and FM (Murai et al., 2002; Shitsukawa et al., 2006). A comparative analysis of the genetic networks required for axillary meristem formation in wheat, maize and rice should lead to a more complete understanding of inflorescence architecture in these grass species.
The genes that regulate inflorescence formation are largely conserved in rice and maize. Thus, for example, BARREN STALK (BA1) in maize is the orthologue of LAX PANICLE (LAX) in rice (Komatsu et al., 2001, 2003; Gallavotti et al., 2004). BA1 and LAX encode a helix–loop–helix transcription factor that regulates axillary meristem initiation. The gene that functions to initiate axillary meristem and lateral organ development in the maize inflorescence was identified in the barren inflorescence2 (bif2) mutant (McSteen and Hake, 2001; McSteen et al., 2007). BIF2, and its rice orthologue, OsBIF2 (OsPID), is a PINOID serine/threonine kinase that is expressed in all axillary meristems and lateral primordia during axillary meristem initiation (McSteen et al., 2007; Morita and Kyozuka, 2007). A recent study showed that the BIF2 protein is associated with auxin transport, and functions in axillary meristem initiation by interaction with BA1 in the nucleus (Skirpan et al., 2008).
SM identity is regulated by the action of the orthologous genes, BRANCHED SILKLESS 1 (BD1) in maize and FRIZZY PANICLE (FZP1) in rice (Chuck et al., 2002; Komatsu et al., 2003). BD1 and FZP1 belong to the APETALA2 (AP2) transcription factor family, and knockout mutants of these genes exhibit defects in the FM and branch structures at the floret position. Other orthologous AP2-like genes, INDETERMINATE SPIKELET 1 (IDS1) and SISTER OF INDETERMINATE SPIKELET 1 (SID1) of maize and SUPERNUMERARY BRACT (SNB) of rice, play a major role in the determination of the SM (Chuck et al., 1998, 2008; Lee et al., 2007). In ids1 and sid1 mutants, the spikelet loses determinacy and multiple florets are formed by one spikelet. In rice, SNB regulates the transition from SM to FM, and its loss-of-function mutant produces extra bract-like structures at the base of the spikelets and shows abnormal development of the floral organs. AP2-like genes are also associated with inflorescence formation in wheat. The major domestication Q gene has been shown to govern some aspects of inflorescence architecture such as free-threshing character, square spike phenotype and rachis fragility (Faris et al., 2003; Simons et al., 2006). The Q gene encodes an AP2 family transcription factor. Sequencing and Southern analysis of this locus indicated that the dominant Q allele most likely arose through a gain-of-function mutation of q.
Bread wheat (Triticum aestivum) is a hexaploid species with the genome constitution AABBDD that originated from three diploid ancestral species: the A genome came from T. urartu, the B genome from Aegilops speltoides or another species classified in the Sitopsis section, and the D genome from Ae. tauschii (Feldman, 2001). It is known that hexaploid wheats originated on multiple occasions from crosses of the domesticated tetraploid wheats (T. turgidum ssp. dicoccum), and the diploid D genome donor, Aegilops tauschii. During the polyploidization step, polyploid wheats acquired environmental adaptability and high yield potential. Polyploidy has a distinctive effect on inflorescence morphology through influencing floret number per spikelet. As the ploidy level increases, so does the floret number per spikelet: one or two florets per spikelet in diploid, two or three in tetraploid, and more then three in hexaploid wheats. It is estimated that 30–80 % of the angiosperms are polyploids (Leitch and Leitch, 2008). Generally, polyploids exhibit growth vigour and wide adaptation to environment compared with diploid species, indicating that polyploidy is a driving force in plant evolution and speciation (Otto and Whitton, 2000). Diploid, tetraploid and hexaploid wheats should be good material for addressing the question of how mechanisms contribute to the growth vigour and adaptive traits in polyploids.
In this study, the mechanisms that regulate the number of florets per spikelet by a comparative morphological analysis of diploid (T. monococcum), tetraploid (T. turgidum ssp. durum) and hexaploid (Triticum aestivum) wheats were investigated. It was found that the number of floret primordia differed between these species. Furthermore, a comparison of the meristematic tissue distribution showed that the difference in the number of floret primordia was due to the heterochronic development of FM initiation from the SM. Heterochrony is defined as a change in the timing of developmental events. In general, it leads to variation in size and shape, which are important traits for speciation.
The diploid wheat (Triticum monococcum, genome constitution AA) strain KU104-2, tetraploid wheat (T. turgidum ssp. durum, AABB) ‘Langdon’ and hexaploid wheat (T. aestivum, AABBDD) ‘Chinese Spring’ were grown in an experimental field, and were used for light microscope, scanning electron microscope (SEM) and in situ hybridization analyses. A tetraploid mutant, which was designated as branching head (bh) (T. turgidum ssp. turgidum strain KT10-2), is a classic natural mutant of emmer wheat; genetic and mapping analyses indicated it is controlled by single gene locus located on the short arm of chromosome 2A (Klindworth et al., 1990a, b, 1997). The bh mutant has a highly branched spike (inflorescence).
Spikes at the heading stage and at 3 d after flowering from diploid, tetraploid and hexaploid wheats were sampled from the field, and all the florets of a spikelet were analysed by light microscopy. The spikes at the two different stages were sampled from the same spike of the same plant. The analysis was performed using an OLYMPUS SZX12 light microscope.
A low-vacuum SEM (S-3000N, Hitachi Co., Ltd, Japan) was used to observe the morphological features of inflorescences. Young spikes were chilled at −15 °C on a cool stage and examined at under low vacuum conditions (30 Pa) and an accelerating voltage of 15 kV or 20 kV.
In situ hybridization was performed as described previously (Shitsukawa et al., 2006). Spikes at different developmental stages (from pre-double ridge to floral organ differentiation stages) were sampled from CS and bh mutant plants, and fixed with FAA solution (3·7 % paraformaldehyde, 5 % acetic acid, 50 % ethanol) at 4 °C overnight. The fixed tissues were dehydrated and embedded in Paraplast Plus (Oxford Labware). The tissues were cut into 20-μm sections and dried overnight. Hybridization was performed overnight at 55 °C. A DIG-labelled probe was made by in vitro transcription of a plasmid containing the full-length Wknox1 (Takumi et al., 2000) using a DIG RNA labeling kit (Roche Diagnostics). After hybridization, the sections were washed twice with 0·5 × SSC at 52 °C. Immunological detection of the hybridized probe was performed as described by Hama et al. (2004). As controls for specificity, sections were hybridized with sense and antisense probes of the same region of the Wknox1 gene.
The early stages of inflorescence development and the inflorescence structures of hexaploid wheat were described in previous studies (Murai et al., 2002; Shitsukawa et al., 2006). The wheat inflorescence (spike, ear or head) develops at the tip of a stem and is composed of spikelets. The spikelets are arranged as two opposite rows on the main axis, the rachis. The number of spikelets per spike is determined by the timing of terminal spikelet initiation, which depends upon genotype and environmental conditions. In this respect, the rachis meristem in wheat is determinate. The spikelet is composed of florets joined at the axis (rachilla) alternately on opposite sides, and encompassed by two small bract leaves called glumes (Fig. 1A). There are multiple florets in each spikelet: hexaploid wheat usually has four to six fertile florets. In contrast to barley, rice and maize, the wheat rachilla meristem is classified as indeterminate. In each floret, the reproductive organs are enveloped by two leaf-like structures, a lemma and a palea. The lemma and palea have one and two midribs, respectively (Fig. 1B). The lemma and palea correspond to the bract, but their origin is still unclear. An individual wheat flower contains one pistil, three stamens and two lodicules (Fig. 1B, C). The lodicules are formed on the lemma side and are considered as an analogous organ to the petal of other angiosperms.
The anatomy of florets of diploid, tetraploid and hexaploid wheats were analysed by light microscopy to identify morphological differences in their development at different ploidy levels (Fig. 2). The number of florets per spikelet increased with ploidy level: two florets in diploid, four in tetraploid and five in hexaploid wheats. In the heading stage spikelets analysed here, one fertile floret was found in diploid (Fig. 2A), three in tetraploid (Fig. 2D–F) and four in hexaploid wheats (Fig. 2I–L). Hypoplasia, underdevelopment or incomplete development, was observed in the floret primordia at the tips of spikelets of diploid (Fig 2C), tetraploid (Fig. 2H) and hexaploid wheats (Fig. 2N). Thus, SM was similarly indeterminate at all ploidy levels.
In spikes at 3 d after flowering, fertile florets usually contain developing grains. Examples of spikes at this growth stage from diploid, tetraploid and hexaploid plants are shown in Fig. 3A, E and J, respectively. One developing grain was found in diploid (Fig. 3B), three in tetraploid (Fig. 3F–H) and four in hexaploid wheats (Fig. 3K–N). In contrast to fertile florets with developing grains, sterile florets regressed in size (Fig. 3C, D, I, O, P). These observations indicate that the number of fertile florets per spikelet is determined by hypoplasia and abortion of florets in the distal segment of the spikelet.
Morphological differences in inflorescence development in diploid, tetraploid and hexaploid wheats were analysed further using SEM. The early steps of inflorescence development were similar at all ploidy levels (Fig. 4A, E, I). The inflorescence meristem forms the spikelet primordia in two opposite rows, and the glume is developed as the first organ of the spikelet (Fig. 4B, F, J). At the floret differentiation stage, the SM produces the FM at a lateral branch (Fig. 4C, G, K). There were clear differences among the diploid, tetraploid and hexaploid wheats at the floret differentiation to floral organ development stages. In diploid wheat, four floret primordia were initiated (Fig. 4D), whereas five and six floret primordia were initiated in tetraploid (Fig. 4H) and hexaploid wheats (Fig. 4L), respectively. These results are in agreement with those from the light microscope analysis and indicate that the number of floret primordia is correlated with ploidy level in wheat.
The wheat KNOX1 gene, Wknox1, belongs to the knotted1(kn1)-type class I homeobox (KNOX) genes, such as maize kn1 and rice OSH1, and is strongly expressed in the meristem (Takumi et al., 2000; Morimoto et al., 2005). OSH1 has been reported to be a useful molecular marker for meristematic tissues (Kerstetter et al., 1994; Sentoku et al., 1999). To determine whether Wknox1 could likewise prove to be a robust molecular marker for meristematic tissues in wheat, its expression pattern was examined in the spikes of mutant branching head (bh) wheat. In normal wheat, the inflorescence is composed of spikelets and florets, and does not produce any lateral branches from the inflorescence axis (Fig. 1A). In bh wheat, a well-known natural variant of tetraploid wheat (T. turgidum ssp. turgidum), long lateral branches form in the basal segment of inflorescences; these ectopic lateral branches bear spikelets with fertile florets (Fig. 5A, B). Although the BH gene has not yet been cloned, bh remains an important mutant for understanding inflorescence architecture in wheat because the non-branched inflorescence axis is the one of the most remarkable characteristics of wheat compared with rice and maize.
The inflorescence of bh wheat appears normal at the early stages of development (Fig. 5C). The inflorescence meristem produces an SM on the axis alternately on opposite sides (Fig. 5D). Abnormal morphological development compared with normal spikelets was only observed when the SM elongated and formed an ectopic branch after the initiation of glume primordia. Meristem arose from ectopic branches of the axis alternately on opposite sides (Fig. 5D). The meristem was not FM but produced a spikelet-like organ (ectopic spikelet; Fig. 5E) that consisted of a floret-like organ (ectopic floret; Fig. 5F). The ectopic florets had a normal morphology with three fertile stamens and a pistil. The ectopic branch derived from the SM had a pair of glumes (Fig. 5B, D–F) suggesting that it was analogous to a spikelet. These observations indicate that the bh phenotype is caused by the transformation of the FM into an SM-like meristem.
To confirm that Wknox1 is a functional orthologue of other KNOX homeobox genes, an in situ analysis was performed. Wknox1 expression was detected in longitudinal sections of the ectopic branch at the primordia (Fig. 5G) and elongating stages (Fig. 5H) in the bh inflorescence. In horizontal sections of the ectopic branch, the Wknox1 signal showed an expanded distribution across a large part of the ectopic spikelet and also into ectopic floret primordia (Fig. 5I). The pattern of Wknox1 localization corresponded to meristem patterning, suggesting that Wknox1 has functional orthology to the KNOX homeobox genes and that its meristem-specific expression pattern is a valuable tool for studying the distribution of meristematic tissue in wheat inflorescences.
To investigate the meristem patterning of the wheat inflorescence, an in situ hybridization analysis of Wknox1 expression was performed in diploid, tetraploid and hexaploid wheats. At the spikelet differentiation stage, the Wknox1 signal showed a similar localization pattern in the SM of diploid (Fig. 6A), tetraploid (Fig. 6C) and hexaploid wheats (Fig. 6E). As shown in the figure, it is notable that the SM appears to be similar in size at the different ploidy levels. The SM elongates gradually and produces FM as a lateral branch. At the floret differentiation stage, we observed three signals corresponding to the FM in diploid wheat (Fig. 6B), four in tetraploid wheat (Fig. 6D) and six in hexaploid wheat (Fig. 6F). Notably, although the Wknox1 signal was similarly persistent at the tip of the SM in diploid, tetraploid and hexaploid wheats, the numbers of FM per spikelet were correlated with ploidy level. This observation is in agreement with the results of the morphological studies (Figs 2 and and3).3). The length of the floret differentiation stage did not differ in the diploid, tetraploid and hexaploid wheats, i.e. from the middle of April to the beginning of May in the experimental field. This suggests that the timing of FM initiation varied with ploidy level: the FM initiated slowly from the SM in diploid wheat but rapidly in hexaploid wheat. Based on the results of this study, it is concluded that the heterochronic development of the FM at different ploidy levels causes the difference in the number of florets in diploid, tetraploid and hexaploid wheats.
The inflorescence of grass species is composed of a unique unit called the spikelet. The spikelet is encompassed by two small bract leaves called glumes, and composed of florets joined at the rachilla. In barley, rice and maize, the number of florets per spikelet is determinate as one, one (originally three) and two, respectively (Schmidt and Ambrose, 1998). In contrast, wheat has an indeterminate SM, and there are multiple florets in each spikelet: usually four to six fertile florets in hexaploid wheat (Murai et al., 2002). In this study, hypoplasia was observed in the florets in the distal portion of the spikelet (Figs 2 and and3).3). The florets shown in Figs 2 and and33 were sampled from the same spike; therefore, floret positions in the spikelet in Fig. 2 correspond to those in Fig. 3. The present results indicate that there are two types of floret hypoplasia. In the first type, floral organs such as stamens and pistil are differentiated but they are sterile and aborted (e.g. in diploid wheat; compare Figs 2B and and3C).3C). In the second type, the FM initiates but no floral organs develop (e.g. in diploid wheat; compare Figs 2C and and3D).3D). The number of florets with hypoplasia might be affected by genotype and growth conditions, such as use of fertilizer. A better understanding of the mechanism of floret hypoplasia would provide useful information for increasing the grain number per spikelet in crop species.
In this study, it was found that the number of floret primordia increased with the ploidy level. Furthermore, the comparison of floret hypoplasia in diploid, tetraploid and hexaploid wheats showed that hypoplasia occurred in florets in the distal portion of spikelets at all ploidy levels (Figs 2 and and3).3). Therefore, the present study indicates that the difference in number of fertile florets among diploid, tetraploid and hexaploid wheats is associated with the number of floret primordia initiated in spikelets. Why do the numbers of florets differ in diploid, tetraploid and hexaploid wheats? The SEM and in situ analyses indicated that heterochronic development of the FM at different ploidy levels might cause the difference in the number of florets (Figs 4 and and6).6). This conclusion then raises the question of what causes the heterochronic development of the FM in diploid, tetraploid and hexaploid wheats. This may be related to a heterosis effect in the polyploid species; however, the mechanism is unclear at present.
In this study, it was confirmed that the wheat SM is indeterminate (Fig. 6) and it was found that the number of fertile florets per spikelet was determined by the numbers of floret primordia and of florets with hypoplasia (Figs 2 and and3).3). In maize, the IDS1 gene is a major candidate for mediating the number of florets per spikelet (Chuck et al., 1998). In both male and female inflorescences of maize ids1 mutants, the SM becomes indeterminate and produces additional florets, suggesting that IDS1 is required either to suppress FM differentiation as lateral branches of SM or to promote the conversion of the SM to FM. IDS1 encodes an APETALA2 (AP2)-like transcription factor. Another AP2-like gene, SID1, also acts to promote determinacy and produce sex organs in maize (Chuck et al., 2008). A different class of mutant, tasselseed4 (ts4), produces highly branched meristems. TS4 encodes a mir172 microRNA that targets both IDS1 and SID1 in maize (Chuck et al., 2007, 2008). It is possible that the loss-of-function mutations of these genes are involved in the indeterminacy of the SM. To date, wheat orthologues of IDS1 and SID1 have not been identified. In wheat, the AP2-like gene Q is a major regulator of spike formation, for example, of the square-headed phenotype and free-threshing character. Notably, hexaploid wheat mutants and natural variants that possess the q allele have a speltoid-type spike in which the number of florets per spikelet is reduced, especially in the spikelet in the distal portion of the spike. Diploid wheat (T. monococcum) carries the q allele and has a speltoid-type spike. In contrast, cultivated tetraploid (T. turgidum ssp. durum) and hexaploid (T. aestivum) wheats have the Q allele. This may be one of the reasons why the number of florets per spikelet in the diploid is lower than that in polyploid wheats. Furthermore, it is known that the Q gene shows dosage effects, suggesting that the effect of Q is stronger in hexaploid than tetraploid wheat. Although it is possible that Q influences the number of florets per spikelet, it is not clear whether Q might control determinacy of the SM or heterochronic development of the FM.
In rice, the SUPERNUMERARY BRACT (SNB) gene regulates the timing of meristem development and the pattern of formation of floral organs (Lee et al., 2007). The SNB gene encodes a protein that belongs to the IDS1 sub-group of AP2-like genes. SNB knockout plants have extra bract-like structures in the basal portion of the spikelet and show pleiotropic effects on floral organ formation. In addition, snb plants produce extra florets in the tip of the rachilla due to the heterochronic development of the meristem. The phenotype of snb rice plants suggests the possibility that heterochronic development of the FM in the wheat spikelet is associated with an SNB-like gene.
The in situ analysis showed that Wknox1 transcripts were present in the tips of the SMs in diploid, tetraploid and hexaploid wheats (Fig. 6), indicating that the indeterminacy of the SM is maintained in polyploid wheats, i.e. indeterminacy of the SM is not affected by polyploidization. Therefore, it is concluded that differences in the numbers of grains per spikelet in diploid, tetraploid and hexaploid wheats are due to the heterochronic development of FM initiation from the SM. What is the relationship between polyploidization and heterochrony of FM development? Which genes might regulate heterochrony? Polyploidization leads to the generation of duplicated homoeologous genes (homoeologues), as opposed to paralogous genes (paralogues). There are three possible evolutionary fates for homoeologous genes in polyploids: functional diversification, gene silencing, and retention of original or similar function (Wendel, 2000). Functional diversification of homoeologues is one of the important factors in the evolutionary success of polyploid species. Furthermore, the evolutionary success of allopolyploids is due to the retention of the function of all homoeologues at many loci. This facilitates positive inter-genomic interactions that are maintained in a self-pollinating plant such as wheat to produce permanent heterosis. The tetraploid and hexaploid wheat genomes contain duplicated and triplicated homoeologues derived from their respective ancestral diploid species. Dosage effects and/or functional divergence of the genes for heterochrony, e.g. an SNB-like gene (Lee et al., 2007), might regulate FM development in polyploids. This may be one factor causing heterosis in polyploids.
Polyploidy often occurs throughout the evolution of flowering plants, including many important crops (Leitch and Bennett, 1997). Studies of natural and synthetic polyploids revealed that dynamic and stochastic changes occurred in genomic structure and gene expression in polyploids, which maybe provided growth vigour and adaptive traits to polyploids (Chen and Ni, 2006; Chen, 2007). In this study, it was shown that heterochronic development of the FM is associated with the difference in floret number per spikelet among diploid, tetraploid and hexaploid wheats. Of course, it has to be noted that there are wide variations in floret number per spikelet within diploid, tetraploid or hexaploid species. However, heterochronic development of the FM should be one of the important mechanisms to determine the difference in floret number among diploid, tetraploid and hexaploid wheats. In developmental biology, heterochrony is defined as a change in the timing of developmental events, leading to variation in size and shape, which are important traits for speciation. The identification of the genes for the heterochronic development of the FM and a functional analysis of the homoeologous genes is a future goal for research. Increase in the number of florets is one of the important characters that have contributed to the success of polyploids in nature and in agriculture, since it leads directly to an increase in seed production. This topic is therefore of interest both to applied and basic sciences.
We thank Dr Shoji Ohta and Dr Yoshihiro Matsuoka for valuable suggestions, and Ms Ai Takagishi for technical assistance. We are also grateful to the National Bioresource Project – Wheat (NBRP-KOMUGI) for providing wheat materials. This work was supported by KAKENHI (Grant-in-Aid for Scientific Research) on Priority Areas ‘Comparative genomics’ from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to K.M.).