SEARCH

SEARCH BY CITATION

Keywords:

  • TCP4;
  • organ maturation;
  • plastochron;
  • flowering time;
  • reproduction;
  • life cycle

Summary

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

Plant organs are initiated as primordial outgrowths, and require controlled cell division and differentiation to achieve their final size and shape. Superimposed on this is another developmental program that orchestrates the switch from vegetative to reproductive to senescence stages in the life cycle. These require sequential function of heterochronic regulators. Little is known regarding the coordination between organ and organismal growth in plants. The TCP gene family encodes transcription factors that control diverse developmental traits, and a subgroup of class II TCP genes regulate leaf morphogenesis. Absence of these genes results in large, crinkly leaves due to excess division, mainly at margins. It has been suggested that these class II TCPs modulate the spatio-temporal control of differentiation in a growing leaf, rather than regulating cell proliferation per se. However, the link between class II TCP action and cell growth has not been established. As loss-of-function mutants of individual TCP genes in Arabidopsis are not very informative due to gene redundancy, we generated a transgenic line that expressed a hyper-activated form of TCP4 in its endogenous expression domain. This resulted in premature onset of maturation and decreased cell proliferation, leading to much smaller leaves, with cup-shaped lamina in extreme cases. Further, the transgenic line initiated leaves faster than wild-type and underwent precocious reproductive maturation due to a shortened adult vegetative phase. Early senescence and severe fertility defects were also observed. Thus, hyper-activation of TCP4 revealed its role in determining the timing of crucial developmental events, both at the organ and organism level.


Introduction

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

In plants, organs develop throughout life from a mass of pluripotent cells at the shoot and root apices. During growth, the shoot apex progresses through distinct developmental phases such as the juvenile vegetative phase, the adult vegetative phase (during both of which leaves are produced) and the reproductive phase, during which flowers form. In Arabidopsis, the juvenile to adult transition is marked by the appearance of trichomes on the abaxial surface of leaves, increased leaf serration, an increased length/width ratio of leaves, and the ability to respond to extrinsic and intrinsic cues for flowering (Telfer et al., 1997). In annual species, the reproductive shoot apex eventually enters senescence and dies. The precise timing of these individual phases is crucial for normal development, and is regulated by well-studied genetic pathways (Poethig, 2010).

Temporal regulation of coordinated processes is a pre-requisite for growth of lateral organs such as leaves. The basic form of the leaf is established by cell proliferation during the early part of growth, while directed cell expansion increases the leaf surface area by several thousand-fold during the later part of growth (Dale, 1988; Dengler and Tsukaya, 2001). During proliferation, dividing cells are initially found throughout the leaf and are progressively restricted towards basal parts of the lamina (Donnelly et al., 1999; Nath et al., 2003; Efroni et al., 2008). These studies demonstrate that the process of growth involves spatial and temporal regulation of cell division and expansion.

Class II TCP genes, including CINCINNATA (CIN) in Antirrhinum and TCP2, 3, 4, 10 and 24 in Arabidopsis, encode transcription factors that coordinate growth processes during leaf development (Martín-Trillo and Cubas, 2010). They limit cell proliferation in the lamina, particularly at margins. Down-regulation of TCP genes leads to a bigger, crinkly leaf with undifferentiated cells, and their constitutive or mis-expression leads to rapid onset of differentiation (Nath et al., 2003; Palatnik et al., 2003; Koyama et al., 2007; Ori et al., 2007; Efroni et al., 2008; Schommer et al., 2008). This suggests that TCPs restrict the duration of the proliferative phase by promoting the onset of differentiation. Likewise, other genetic factors that control organ morphogenesis affect either cell proliferation or the expansion phase. AINTEGUMENTA, which encodes an AP2 domain transcription factor, controls organ size by determining the competence of cells to divide (Mizukami and Fischer, 2000). JAGGED and NUBBIN are other putative transcription factors that promote cell division in lateral organs, as does KLUH, a cytochrome P450 oxidase (Dinneny et al., 2004; Ohno et al., 2004; Anastasiou et al., 2007). Growth by division along the medio-lateral axis is regulated by ANGUSTIFOLIA3, a transcriptional co-activator (Kim and Kende, 2004; Horiguchi et al., 2005), while cell division along the proximo-distal axis is independently regulated by ROTUNDIFOLIA4, a plant-specific peptide (Narita et al., 2004). In contrast, transcription factors such as PEAPOD (PPD1 and 2), ARF2 and the E3 ubiquitin ligase BIG BROTHER repress cell division (Disch et al., 2006; Schruff et al., 2006; White, 2006). Cell expansion is regulated by the ARGOS-LIKE1 and ROTUNDIFOLIA3 genes that are involved in brassinosteroid-mediated growth control (Kim et al., 2005; Hu et al., 2006). Mutation in genes regulating endo-reduplication, an event that is responsible for the considerable increase in size of leaf epidermal cells, also reduces organ size and slows down growth (Churchman et al., 2006).

How organ growth is coordinated with the overall maturation of an organism is not well understood on the basis of most studies on leaf mutants. Apart from organ size, other traits such as rate of vegetative growth, number of leaves, timing of reproductive maturity and senescence, number of progeny produced and lifespan, directly determine the fitness of an individual. These traits are genetically controlled, and are often linked by common and cross-talking pathways. Empirical observations in many animal and plant species have shown that maturation time is a function of body size (Roff, 1992; Stearns, 1992). The advantages conferred by a short generation time with early reproductive maturation are expected to be counter-balanced by a trade-off with other fitness components, such as decreased seed set. Correlated genetic traits place a constraint on the extent to which an organism can evolve adaptations in response to changes in environment (Blows and Hoffmann, 2005). Thus, studies on mutations that simultaneously affect the growth of organs and of the whole organism provide valuable information on coordination between the two processes.

Although the role of TCP genes in controlling organ growth has been studied in detail, it is not clear whether their activity regulates plant growth and life cycle in general. We wished to investigate the relationship between the level of TCP activity and growth at the cellular, organ and organismal level. Individual TCP mutants display a mild phenotype due to functional redundancy. While their simultaneous down-regulation by constitutive expression of miR319, a microRNA that targets CIN-like TCP genes, strongly alters leaf phenotype in Arabidopsis, it also results in down-regulation of miR159 targets, due to cross-regulation (Palatnik et al., 2003, 2007). Moreover, constitutive expression of miR319-resistant versions of TCPs leads to embryonic lethality, and does not provide accurate information about endogenous TCP function due to ectopic expression of the transgene. To circumvent these problems, we have hyper-activated the function of a single TCP protein, TCP4, without affecting its transcriptional and post-transcriptional regulation, and studied the effect on plant development. Our findings show that the level of TCP4 activity is an important determinant not only of leaf size, but also the timing of other traits, such as the vegetative to reproductive transition, the rate of organ initiation and onset of senescence, thus revealing the novel function of this transcription factor in Arabidopsis.

Results

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

Generation of a stable transgenic line expressing a hyper-active form of TCP4

To examine the effect of heightened TCP4 activity on plant growth, the TCP4 ORF was fused in-frame to the sequence encoding the C-terminal activation domain of VP16, a viral protein, which acts as a potent activator of transcription when fused with known transcription factors in heterologous systems (Sadowski et al., 1988). Three fusion constructs were generated that differed with respect to the site of VP16 insertion (Figure 1a). All fusion proteins were tested for their transcription-activation potential by yeast one-hybrid assay. The TCP4 protein itself shows transactivation activity in yeast (Aggarwal et al., 2010). Although all three fusion proteins showed enhanced transactivation, the C-terminal fusion protein TCP4:VP16-C showed the highest activity, approximately 16-fold higher than TCP4 alone (Figure 1b). Thus, the activation domain of VP16 imparted hyper transcriptional-activation function to TCP4. The TCP4::TCP4:VP16-C construct was introduced into tcp4-1, a null mutant of TCP4 that shows cotyledon epinasty, deeper leaf serrations and mild delay in flowering (Figures 1d and 2a) (Schommer et al., 2008). Ten independent transgenic lines were obtained, which were classified as weak, strong or severe, based on the extent of the reduction in rosette size (Figure 1c). The presence of the transgene was confirmed by genomic PCR (data not shown). All transgenic lines with a severe phenotype were infertile, but one of the lines with a strong phenotype produced viable seeds. This stable line expressed the transgene in the endogenous expression domain of TCP4, as analyzed by RT-PCR and RNA in situ hybridization (Figure S1, see Appendix S1 for experimental procedure). This line was used for further analysis. The TCP4:VP16-C seedlings not only showed rescue of cotyledon epinasty of tcp4-1, but also produced hyponastic cotyledons and longer hypocotyls (Figure 1d,e). Plants expressing miR319-resistant TCP4 showed similar phenotypic effects (Schommer et al., 2008), indicating that the phenotypic changes observed in TCP4:VP16-C plants were due to enhanced TCP4 function. When the TCP4:VP16-C line was crossed with jaw-D, an miR319a over-expressing line, the phenotype of the trans-heterozygote F1 plants resembled that of the wild-type. Thus, hyper-activity of TCP4 fully rescued the morphological defects caused by down-regulation of multiple TCP genes (Figure 1f). The fact that the rescue was not a consequence of silencing the jaw-D allele was confirmed by checking the level of miR319a precursor in these F1 plants (Figure 1g). Compared to the jaw-D homozygous line, the precursor level of miR319 was reduced in the F1 plants, most likely due to an allele-dosage effect, as F1 progeny of a cross between jaw-D and Col-0 showed a similar decrease in the pre-miR319 level.

image

Figure 1.  Generation of transgenic Arabidopsis plants expressing hyper-active TCP4. (a) Schematic of the TCP4 ORF, alone or fused with the activation domain of VP16 (gray box). The asterisks indicate miR319 target sites. TCP indicates the TCP domain. (b) β-galactosidase activity in yeast expressing TCP4 alone (TCP4) or TCP4:VP16 fusions (M1, M2 and C). V indicates the vector control. Values are means ± SEM (five independent experiments). (c) Rosettes of 24-day-old plants. Scale bar = 1 cm. (d, e) Seedling phenotype of a strong TCP4:VP16-C line. The mild and strong cotyledon epinasty in Col-0 and tcp4-1, respectively, is completely rescued and converted to hyponasty in TCP4:VP16-C (d). Scale bar = 5 mm. The length of TCP4:VP16-C hypocotyls is also increased (d, e). Values in (e) are mean lengths ± SD ( 25). Asterisks indicate significant differences (*< 0.05, ***< 0.001). (f) Leaf phenotype in 24-day-old rosettes of the indicated genotypes. Scale bar = 1 cm. (g) Relative expression of pre-miR319a in jaw-D (black bar), jaw-D x TCP4:VP16-C F1 progeny (white bar) and jaw-D x Col-0 F1 progeny (gray bar). Transcript levels were first normalized to those for the ACTIN transcript, and then expressed relative to Col-0. Values are the mean of two independent experiments.

Download figure to PowerPoint

image

Figure 2.  Altered leaf morphology in TCP4:VP16-C. (a) Silhouettes of rosette leaves in 24-day-old plants grown under long-day conditions. Scale bar = 5 mm. (b) Comparison of the area of the 7th rosette leaf in tcp4-1 plants (black circles) and TCP4:VP16-C plants (white circles) grown under long-day conditions. The error bar indicates SD. (c, d) Comparison of the size of 7th rosette leaves in plants of the indicated genotypes grown under short-day conditions. Note that TCP4:VP16-C leaves are smaller with strong cup-shaped morphology (c). Scale bar = 1 cm. Error bars in (d) indicate SD (= 6). (e) Comparison of the leaf index of 7th rosette leave in tcp4-1 plants (black circles) and TCP4:VP16-C plants (white circles) grown under long-day conditions.

Download figure to PowerPoint

TCP4 hyper-activity reduces leaf size

The TCP4:VP16-C plants produced considerably smaller rosette leaves compared to tcp4-1 and Col-0 (Figure 2a). The first two leaves showed a dramatic reduction of the lamina. Detailed analysis of leaf growth was performed for the 7th rosette leaf, which is representative of the adult vegetative phase (Figure 2b). In tcp4-1, leaf growth continued till 15 days after emergence, as indicated by the increase in leaf length and width. However, leaf growth in TCP4:VP16-C plants proceeded at a slower rate and stopped 12 days after emergence. The shorter duration of growth coupled with a slower growth rate led to a threefold reduction in leaf size in TCP4:VP16-C plants (Figure 2b–d).

The TCP4:VP16-C leaves were rounder than usual. The shape of a leaf is determined by relative growth in length and width axes, which is expressed as the leaf index (Tsukaya, 2002a). Although leaf growth was affected in both axes in TCP4:VP16-C plants, the extent of the reduction was proportionate, such that the leaf index was unchanged relative to tcp4-1 (Figure 2e and Figure S2a,b). This was true for other rosette leaves as well (Figure S2c,d). The TCP4:VP16-C leaves also showed cup-shaped morphology, with the central portion of the lamina bulging out, making the leaf un-flat. The loss of flatness was exacerbated under short-day conditions (Figure 2c). These leaves could not be flattened without introducing cuts at the margin (Figure S2e), indicating that there is loss of area at the margin. The reduction in area was approximately 9.2 ± 2.8% compared with Col-0. In addition, the later-formed leaves in TCP4:VP16-C lacked serrations. Leaves of Col-0 plants show serrations at the margin, starting from the 7th leaf (Figure 2a). tcp4-1 leaves display deeper serrations, consistent with the role of TCPs in shaping the margins (Koyama et al., 2007). These serrations were completely abolished in the TCP4:VP16-C line, even in adult leaves, leading to smooth margins.

Although TCP4 is expressed in growing flower buds and controls petal development (Nag et al., 2009), petal growth was unaffected in the TCP4:VP16-C plants. Growth of other floral organs was also normal (Figure S3). Organs with indeterminate fate, such as root and stem, also grew normally (Figure S4), although we observed a reduction in the length of secondary lateral roots and a decrease in the number of tertiary lateral roots in TCP4:VP16-C. Thus, TCP4 activity primarily controls leaf growth.

Premature arrest of cell division and accelerated differentiation in TCP4:VP16-C plants

Reduction in leaf size may be an outcome of reduced cell number or cell size. To distinguish between the two possibilities, cell size in mature leaf epidermis was compared between TCP4:VP16-C and tcp4-1 plants (Figure 3a). The mean cell size in TCP4:VP16-C plants was not significantly different from that in tcp4-1 plants (> 0.01), indicating that the reduction in leaf area was solely due to reduced cell number. Cell number is determined by the duration of the proliferative phase early in development, after which cells stop dividing and undergo differentiation. To determine whether cells were undergoing premature cessation of division and entering differentiation early, we monitored the increase in epidermal cell area at progressive stages of leaf development.

image

Figure 3.  Premature onset of differentiation in TCP4:VP16-C leaves. (a) Mean cell size in the adaxial epidermis of tcp4-1 and TCP4:VP16-C leaves. Error bar indicates SD. (b) Scanning electron micrographs of epidermal cells at the tip, middle and base of the 7th rosette leaf (as indicated in the diagram on the left) at maturity. Scale bar = 100 μm. (c–e) Epidermal cell size at the tip (c), middle (d) and base (e) of the 7th rosette leaf in tcp4-1 plants (black circles) and TCP4:VP16-C plants (white circles). Error bars indicate SEM. (f) Distribution of cell size at the base of tcp4-1 and TCP4:VP16-C leaves (7th position).

Download figure to PowerPoint

The pattern of cell size increase for tcp4-1 and TCP4:VP16-C at the tip, middle and base of developing lamina is shown in Figure 3(c–e). When the leaf was about one-fifth of its final length in tcp4-1 plants, the cells throughout the lamina were small and polygonal. As growth proceeded, the mean cell area increased, with the rate of increase being highest at the tip and lowest at the base. In contrast, TCP4:VP16-C cells were larger throughout the developing lamina, indicating that the cells started differentiating earlier. The TCP4:VP16-C cells underwent a precocious increase in size, reaching mature size when the leaf was half as long as in tcp4-1. The tip-to-base gradient in cell size in young leaves was almost abolished in the TCP4:VP16-C line, as the middle and the basal cells grew to the same size as the tip cells at an earlier stage. Moreover, cells at the base of mature TCP4:VP16-C leaves were larger than cells at the middle or the tip due to excess expansion (Figure 3b,f); over 10% of basal cells in TCP4:VP16-C plants had a cell area >6000 μm2.

Thus, cells in TCP4:VP16-C leaves undergo division arrest and enter the differentiation phase prematurely. In order to directly assess the cell division potential in TCP4:VP16-C leaves, we crossed the TCP4:VP16-C line with a cyc1At::GUS reporter line that expresses GUS only in mitotic cells (Donnelly et al., 1999). Leaves from the reporter line in both Col-0 and TCP4:VP16-C backgrounds were harvested at various developmental stages and analyzed by GUS assay (Figure S5). Wild-type leaves showed cell division activity throughout the lamina at very young stages, but this was progressively restricted towards the base as the leaf matured (Figure S5a–d). In the TCP4:VP16-C leaves, GUS activity decreased at very early stage of leaf development. Even in the 0.2 mm long leaves, little GUS activity could be detected, indicating reduced cell division, and this had completely ceased by the time the leaf was 1 mm long (Figure S5a–e). Thus, hyper-active TCP4 directs cells to exit the proliferation phase earlier and enter differentiation prematurely, leading to a smaller leaf.

Effect of TCP4 hyper-activity on cell size

An effect of TCP4 activity on cell size was also observed in the hypocotyl, which grows only by cell elongation under etiolated conditions (Gendreau et al., 1997). Col-0, tcp4-1 and TCP4:VP16-C seedlings were grown in darkness, and the length and cell size of the hypocotyls were compared. Etiolated TCP4:VP16-C hypocotyls were approximately 1.5 times longer than those of Col-0 and tcp4-1 (Figure 4a,b). This increase in length was solely due to cell expansion, as the TCP4:VP16-C epidermal cells were approximately 1.5 times larger compared to the control plants (Figure 4c). Thus, enhanced TCP4 activity promotes cell expansion in leaves and the hypocotyl. As gibberellic acid (GA) is known to regulate cell expansion (Olszewski et al., 2002), we tested the effect of GA3 on the growth of cotyledons, which grow mostly by cell expansion post-germination (Tsukaya et al., 1994). As there was no difference in the cotyledon growth between Col-0 and tcp4-1 under mock and GA3-treated conditions (Figure S6), we used tcp4-1 as the control for comparison with TCP4:VP16-C line. The tcp4-1 cotyledons grew approximately 1.5 times larger in 10 μm GA3 compared to mock-treated plants. By contrast, the TCP4:VP16-C cotyledons grew approximately three times larger in the presence of 10 μm GA3 compared to mock-treated plants (Figure 4d,e). This increase in organ area was fully accounted for by cell enlargement (Figure 4f,g). Thus, hyper-activation of TCP4 not only increased cell size, it also made the epidermal cells more sensitive to GA-dependent cell expansion.

image

Figure 4.  TCP4 activity regulates cell size. (a, b) Comparison of hypocotyl length of 5-day-old etiolated seedlings of indicated genotypes. Scale bar in (a) =5 mm. Error bar in (b) indicates SD ( 15). (c) Epidermal cells in etiolated hypocotyls. Scale bar = 200 μm. (d–g) Effect of GA3 on cotyledon area (d, e) and epidermal cell size (f, g) in plants of indicated genotypes grown under long-day conditions, at 21 days after sowing. Scale bars = 2 mm (d) and 200 μm (f). Error bars in (e) and (g) indicate SD and SEM, respectively. ***< 0.001.

Download figure to PowerPoint

Hyper-active TCP4 promotes onset of senescence

TCP proteins positively regulate jasmonic acid (JA) biosynthesis by promoting expression of LOX2, a gene that encodes a key enzyme in the JA biosynthetic pathway in Arabidopsis (Schommer et al., 2008). JA is known to control the onset of leaf senescence, which is delayed in the jaw-D mutant that shows reduced LOX2 expression. We observed precocious onset of senescence in TCP4:VP16-C plants (Figure 5a). To estimate the effect of hyper-activity of TCP4 on senescence, an induced senescence assay was performed. Col-0 leaves showed an approximately 20% decrease in chlorophyll content after induction of senescence, but TCP4:VP16-C leaves showed an approximately 80% reduction (Figure 5b). Application of methyl jasmonate (MeJA) elicited dose-dependent senescence response in Col-0, which was moderate at 60 μm and more pronounced at 300 μm. On the other hand, TCP4:VP16-C leaves showed an enhanced response to the lower dose of MeJA within 4 days (Figure 5c). Thus, TCP4 hyper-activity propels leaves to a premature terminal stage, i.e. senescence, possibly as a result of accelerated differentiation. To test whether this premature onset of senescence is due to an enhancement in endogenous LOX2 transcription, we determined the LOX2 transcript level by quantitative RT-PCR using RNA isolated from inflorescences. Although the expression level was significantly reduced in tcp4-1 plants, it was not significantly up-regulated in TCP4:VP16-C plants compared to Col-0 (Figure 5d). This raises the possibility that TCP4 mediates senescence by a LOX2-independent pathway.

image

Figure 5.  Precocious onset of senescence in TCP4:VP16-C plants. (a) Mature rosettes of indicated genotypes at 35 days after sowing. (b) Relative total chlorophyll content in rosette leaves upon induction of senescence. Error bars indicate SD ( 5). (c) Effect of MeJA on progression of senescence in detached leaves. D0 and D4 indicate 0 and 4 days post-detachment, respectively. (d) Relative expression level of LOX2 in inflorescences. Values are means ± SD (three biological samples).

Download figure to PowerPoint

Precocious initiation of leaves in the TCP4:VP16-C line

We observed that TCP4:VP16-C plants produced more leaves than tcp4-1 and Col-0 at any given time, indicating that they initiate leaves at a faster rate (Figure 6a). The time interval between initiations of successive leaf primordia is termed the plastochron index (PI), a trait that is genetically determined. Many Arabidopsis mutants with altered PI have been isolated (Conway and Poethig, 1997; Prigge and Wagner, 2001; Wang et al., 2008; Lohmann et al., 2010). To test whether TCP4 activity affects PI, we determined the leaf initiation rates by counting the number of leaves formed as a function of time. The PI values for Col-0, tcp4-1 and TCP4:VP16-C were 1.89, 1.86 and 1.45, respectively, as calculated from the inverse of the slope of the graph in Figure 6(b). Initiation of flowers was also faster in the TCP4:VP16-C line (data not shown). Thus, enhancement of TCP4 activity accelerates the rate of organ initiation.

image

Figure 6.  TCP4 activity controls the plastochron index and flowering time. (a) Rosettes at 12 days after sowing. There were more leaves in TCP4:VP16-C plants compared to Col-0 and tcp4-1. (b) Appearance of visible leaves as a function of time in Col-0 (gray circles), tcp4 (black circles) and TCP4:VP16-C (white circles) grown under long-day conditions. Error bars indicate SD (= 17). (c) Plants of indicated genotypes at 24 days after sowing. Note the early flowering phenotype of TCP4:VP16-C plants. (d) Total number of rosette leaves, with (gray bar) and without (black bar) abaxial trichomes, formed by plants grown under long-day conditions. Error bars indicate SD ( 15).

Download figure to PowerPoint

Early onset of flowering in TCP4:VP16-C plants

Although TCP4:VP16-C plants showed faster leaf initiation, they produced fewer rosette leaves compared to Col-0 and tcp4-1 and flowered earlier (Figure 6b,c). This result is consistent with the modest delay in flowering shown by a TCP4 loss-of-function mutant (Schommer et al., 2008). Under long-day conditions, Col-0 flowered after producing 12.6 ± 1.2 rosette leaves, tcp4-1 produced 13.4 ± 1.5 leaves, before flowering, and TCP4:VP16-C produced10.0 ± 0.8 leaves (Figure 6d). Early flowering in TCP4:VP16-C was also evident from the shorter FT50 (defined as the number of days taken to flower by 50% of a plant population under study) compared to control plants (Table S1).

As the vegetative to reproductive transition was advanced in TCP4:VP16-C, we determined whether juvenile to adult vegetative transition was similarly affected. This was achieved by scoring the number of leaves produced during these phases, taking the presence of abaxial trichomes as a marker for the adult phase. While TCP4:VP16-C plants did not differ significantly from Col-0 and tcp4-1 in terms of the number of juvenile leaves, they produced only approximately 1.5 leaves in the adult phase, in contrast to approximately 5 leaves produced by Col-0 or tcp4-1 plants (Figure 6d). This suggested that TCP4 activity advances reproductive maturity by limiting only the duration of the adult vegetative phase.

Reduced fertility in TCP4:VP16-C plants

As TCP4::TCP4:VP16-C transgenic lines with a severe phenotype failed to yield seeds, we suspected that hyper-activity of TCP4 reduced fertility. The TCP4:VP16-C line used in this study formed fewer siliques than Col-0 and tcp4-1 (Figure 7a), but produced more flowers. Fertility was dramatically reduced, as almost 80% of the flowers of this TCP4:VP16-C line were aborted (Figure 7b). We examined the TCP4 expression pattern in inflorescences and floral organs by GUS reporter assay. The TCP4 promoter is persistently active in the petals in all buds (Figure 7c). GUS activity was detected in pollen of unopened flowers when the stamen was approximately 0.5 mm long and the filament started elongating. At this stage, the TCP4 promoter is also active in carpel wall tissue (Figure 7d,e). No expression was detected when the flower opened (data not shown). Thus, TCP4 is transiently expressed in pollen and carpel tissues.

image

Figure 7.  Reduced fertility in TCP4:VP16 plants. (a) Mature inflorescence stem bearing siliques in indicated genotypes. The arrowhead indicates aborted siliques in TCP4:VP16-C plants. (b) Flowers that form siliques expressed as a percentage of total flowers produced. Values are means ± SD (at least 20 plants). (c–e) TCP4 promoter activity in the inflorescence, (c) carpel (d) and anther (e) of unopened flowers [inset in (d)] as determined by GUS activity. Expression in the carpel wall is shown in the enlargement. Scale bars = 1 mm (c) and 0.5 mm (d, e). (f–h) DAPI-stained pollen from Col-0 (f), tcp4-1 (g) and TCP4:VP16-C (h) anthers from open flowers. Arrowheads indicate two sperm nuclei and one vegetative nucleus, shown as an inset in (f). Scale bar = 100 μm. (i–k) In vitro germination of pollen from dehisced anthers of Col-0, (i) tcp4-1 (j) and TCP4:VP16-C (k). Scale bar = 1 mm. (l–n) Aniline blue-stained carpels from self-pollinated Col-0 (l), tcp4-1 (m) and TCP4:VP16-C (n). Scale bar = 200 μm.

Download figure to PowerPoint

As the growth of stamens and carpels in TCP4:VP16-C flowers was apparently normal (Figure S3d), we examined pollen function and differentiation. Pollen from mature anthers was stained with DAPI to observe the nuclei. The mature pollen grains in Arabidopsis are tri-cellular, with two small sperm cells enclosed within a large vegetative cell that grows into the pollen tube upon germination. Pollen from Col-0 showed the normal pattern of two small and bright sperm cell nuclei with a diffuse ring-like vegetative nucleus (Figure 7f). The tcp4-1 and TCP4:VP16-C pollen showed similar patterns, indicative of normal meiosis (Figure 7g,h). However, the pollen grains in Col-0 were dispersed, but those in tcp4-1 were clumped, indicating a defect in separation of microspores after meiosis, giving rise to the tetrad phenotype. The extent of clumping was enhanced in TCP4:VP16-C. However, the clumping of microspores did not compromise germination of the mature pollen in vitro and in vivo. Pollen grains in anthers of Col-0, tcp4-1 and TCP4:VP16-C, incubated on germination medium, grew pollen tubes after 6 h (Figure 7i–k). Pollen tube formation also occurred normally for all three genotypes in situ, as determined by aniline blue staining of carpels that were self-pollinated manually (Figure 7l–n). This suggests that pollen maturation was unaffected in the TCP4:VP16-C line.

To test carpel function, TCP4:VP16 carpels were pollinated with pollen from wild-type and tcp4-1 flowers. In both cases, crosses failed to increase the success rate above that of a self-cross, indicating that carpel function was also compromised in the transgenic flowers. Surprisingly, TCP4:VP16 pollen failed to produce viable seeds on Col-0 and tcp4-1 carpels (Table S2). Although the pollen appeared normal, it could not fertilize Col-0 and tcp4-1 ovules. These observations show that TCP4 hyper-activity leads to a general reproductive dysfunction, suggesting a role for TCP4 in regulation of plant reproduction.

Discussion

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

Role of TCP4 activity in determining cell and leaf size

Leaf growth is a consequence of regulated cell proliferation and expansion. As a differentiated cell can no longer contribute to increasing the cell number for further growth, the timing of its division arrest is crucial in the context of overall leaf development. TCP4 and its close homologs are involved in the onset of differentiation and have a dose-dependent effect on leaf development (Efroni et al., 2008). The severity of the phenotype associated with TCP loss of function increases as more TCP4-like genes are knocked-down (Koyama et al., 2007; Schommer et al., 2008). Leaves of strong tcp mutants become excessively crinkly with serrated margins, masking the effect on overall organ size. To obtain an insight into the action of TCPs on organ size, we generated a transgenic line that expressed hyper-active TCP4 in its endogenous domain. The advantage of such an approach is that hyper-morphic variants can reveal novel functions of biological significance that are otherwise masked by genetic redundancy. Moreover, unlike in the case of over-expression of a microRNA-resistant version of TCPs, the endogenous spatial and temporal expression pattern is preserved. However, we are aware that the phenotypic changes in such hyper-active lines could, in part, arise from altered post-translational stability/mis-regulation of the protein due to fusion with a heterologous transcription activator.

TCP4 hyper-activity reduced leaf size by limiting the duration of growth and also slowing down the rate of growth. The premature termination of growth may be due to a decrease in the number of cells undergoing division and/or an increase in the number of cells entering the differentiation phase. Previous studies have implied that TCP4 primarily regulates differentiation. For instance, the mature leaves of 35S::miR319a plants have more small undifferentiated cells due to delayed entry into the differentiation phase (Efroni et al., 2008). Our findings are in agreement with this model, as hyper-activity of TCP4 resulted in precocious onset of differentiation. Further, cells in mature leaves attained a larger size, particularly at the base. As the leaf matures, TCP4 expression is restricted to the base, and hence cells at the base are exposed to TCP4 activity for longer than cells at the distal region. The fact that TCP4 controls cell expansion was also demonstrated by enhanced elongation of TCP4:VP16-C hypocotyls. The effect of TCP4 activity on cell expansion is downstream of GA, as epidermal cells of the TCP4:VP16-C line are hyper-sensitive to GA. Thus, hyper-activity of TCP4 promotes cell expansion in leaves, hypocotyls and cotyledons. This effect is unlikely to be due to a compensatory mechanism that is frequently encountered in cell-division mutants (Tsukaya, 2002b), whereby cell size increases in response to reduced cell division. However, the mechanistic basis of TCP4-mediated regulation of differentiation remains unclear.

Although the shape of the TCP4:VP16-C leaves remained unchanged, the curvature was affected. The leaves were cup-shaped in extreme cases, and could not be flattened without cutting the margins. Such a phenotype may result from reduced growth at the margin relative to the centre. This observation is at the other extreme of the spectrum of phenotypes obtained by modulating TCP activity. On one hand, inactivation of CIN gene results in crinkly leaf, due to excess growth in the margins that creates folds in lamina (Nath et al., 2003). On the other hand, here, hyper-activation of TCP4 makes the leaf cup-shaped, possibly due to reduced growth at the margins. Thus, an optimal level of TCP activity is crucial for the flatness of the lamina.

TCP4 regulates plant maturation programs

Enhancement of TCP4 activity affected not only leaf growth but also the rate of initiation of leaves. TCP4:VP16-C leaves were initiated faster than in wild-type. However, the total number of leaves produced during the vegetative phase was less than in Col-0, as the TCP4:VP16-C plants flowered earlier. This feature is shared by other genes that affect the PI, although most of them affect flowering time only slightly, with the exception of the amp1 mutant, which undergoes precocious flowering transition (Chaudhury et al., 1993). The effect on flowering time may not always correlate with the nature of the change in the PI. Our study shows that the faster rate of leaf initiation and early onset of flowering are coupled by TCP activity. Further, TCP genes may have a dose-dependent effect on the PI, akin to leaf size. This is supported by the fact that the leaf initiation rate in the jaw-D mutant is slower than in tcp4-1 (data not shown). How TCP genes control the PI is not clear. TCP genes are not expressed in the shoot apical meristem, although a non-autonomous effect cannot be ruled out. The leaf initiation rate is correlated with organ size; mutants with small organs often show a faster rate of initiation. This correlation is also seen in TCP4:VP16-C plants. It has been suggested that the two traits are associated in order to compensate for any change in total biomass during vegetative growth (Wang et al., 2008). It would be interesting to study the dose-dependent effect of TCPs on organ size in concert with the rate of initiation.

TCP4 controls flowering time, as evident from the delayed flowering phenotype of its loss-of-function mutant tcp4-1 (Schommer et al., 2008). Interestingly, flowering in the jaw-D mutant is delayed to the same extent as in the tcp4 single mutant, suggesting that TCP4 has a non-redundant role in determination of flowering time. This conclusion is reinforced by the observation that hyper-activity of TCP4 leads to early onset of flowering.

It is not known how TCP4 regulates flowering time. The timing of the vegetative to reproductive switch is governed by several pathways that involve a host of transcription factors that combine the endogenous signals with environmental cues to turn on the floral meristem identity genes LEAFY and APETALA1 (Komeda, 2004). It is possible that TCP4 interacts with one or more of these pathways. There are indications that the SQUAMOSA PROMOTER BINDING PROTEIN-LIKE (SPL) group of genes are involved in the TCP-mediated effect on floral transition. These genes regulate shoot development in Arabidopsis, and three of the members, namely SPL3, 4 and 5, redundantly control the onset of flowering and are targets of miR156 (Cardon et al., 1997; Wu and Poethig, 2006). Gain of function of these genes promotes early flowering. Further, their upstream region has conserved TCP4 binding sites and they are down-regulated in the jaw-D mutant (Schommer et al., 2008). Similarly, we have observed that the transcript level of SPL3 was consistently down-regulated in the tcp4-1 mutant, and that a recombinant TCP4 protein is capable of binding to the consensus motifs in SPL3 regulatory sequence in vitro (data not shown). Thus, it is possible that TCP4 directly controls transcription of these genes. However, TCP4:VP16-C plants did not show any up-regulation of SPL3 transcript level (data not shown). The juvenile to adult vegetative phase change occurred normally in these plants, whereas SPL3 over-expression is known to advance this transition (Wu and Poethig, 2006). It is possible that, during the juvenile phase, TCP-mediated transcription of SPL3 is kept in check by post-transcriptional degradation by miR156. As the transition to the adult phase takes place, miR156 production decreases (Wu and Poethig, 2006). The de-repression of SPL transcript accumulation, coupled with hyper-transcription of these genes mediated by TCP4:VP16-C, may, in principle, be responsible for the precocious adult to reproductive transition.

TCP4 activity controls plant reproduction

TCP4 is expressed in the pollen and carpel. Loss of TCP4 expression does not result in any obvious defect in fertility, apart from the failure of tetrad microspores to separate. However, over-expression of miR319 leads to male sterility (Palatnik et al., 2007). This sterility phenotype is not due to down-regulation of TCP genes, but is rather due to cross-regulation of miR159-targeted GAMYB genes by excess miR319. Our study has uncovered a role for TCP4 in reproductive development. TCP4 hyper-activity drastically reduced the fertility of the plant. Although TCP4:VP16-C plants produced viable pollen that germinated normally, a defect in one of the post-germination steps cannot be ruled out. Carpel function was also aberrant in the transgenic line, as viability was not restored upon crossing with wild-type pollen. Reproduction in plants is modulated by GA and JA (Koornneef and Veen, 1980; Stintzi and Browse, 2000), and defects in their biosynthesis or signaling often lead to sterility. Application of neither GA3 nor MeJA rescued the sterility in the transgenic line (data not shown). It is possible that TCP4 functions downstream of GA3, as observed in cotyledon growth, to control fertility as well.

TCP4 as an integrator of key developmental events

We have shown that an enhanced level of TCP4 activity modulates organ development by promoting its initiation, maturation and senescence. TCP4 activity hastens the maturation of the shoot apex to the reproductive phase and also determines fertility. In general, TCP4 activity controls the onset of maturation programs in the plant life cycle. Traits such as initiation rate, organ size, flowering time and seed yield contribute to the fitness of the plant. A faster rate of initiation, bigger organ size, early onset of flowering and higher seed yield are obvious desirable traits. However, they rarely occur simultaneously in a mutant or a natural variant, suggesting that there is a trade-off among traits. Studies have shown that such traits are linked and are controlled by multiple loci that contribute quantitatively to the phenotype. A change that benefits one trait may adversely affect another (Kozlowski, 1992; Mendez-Vigo et al., 2010; Colautti et al., 2011). Our study shows that TCP4 activity can potentially coordinate these inter-connected traits. We observe that the TCP4:VP16-C plants reach reproductive phase faster, but produce fewer seeds. Although these plants have faster rate of organ initiation, the final size is reduced and senescence is advanced. Such genetic constraint on traits limits the phenotypic variation that can be produced in plants and hence adaptation to the environment. Our study suggests that TCP4 links organ growth with that of the whole organism. It acts as a heterochronic regulator that possibly affects the timing of multiple maturation programs. Any perturbation in TCP activity may have far-reaching effects on plant growth, and thus and optimal level of TCP activity is crucial for plant homeostasis. Further studies are required to elucidate the molecular mechanisms involved in TCP-mediated integration of developmental processes.

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 materials and growth conditions

All mutants and transgenic lines are in the Col-0 background, except TCP4 expression reporter line, which is in the Ler background (TIGR annotation ET5977) and was obtained from the Cold Spring Harbor Laboratory (http://genetrap.cshl.edu/). Plants were grown at 22°C and under 16 h day/8 h night (long day) or 8 h day/16 h night (short day) photoperiods.

Construct generation and plant transformation

The TCP4 ORF (1263 bp) was fused translationally to the C-terminal trans-activation domain of VP16 (240 bp) (Sadowski et al., 1988), which was amplified from the pER8 plasmid (Zuo et al., 2000) using domain-specific primers 5′-GATAGGATCCAGGTGAACCGTCGAGCGCC-3′ (forward) and 5′-CATGCCATGGTCACCCACCGTACTCGTCAATTCC-3′ (reverse). The fusion sequence was cloned downstream of a 2.0 kb upstream region of TCP4 in the pCAMBIA 1390 binary vector (CAMBIA, http://www.cambia.org). tcp4-1 plants were used for transformation (Clough and Bent, 1998). Transformants were selected on half-strength MS medium (Himedia, http://www.himedialabs.com) containing 100 μg ml−1 cefotaxime sodium salt (Himedia) and 20 μg ml−1 hygromycin (Sigma-Aldrich, http://www.sigmaaldrich.com). The presence of the transgene was confirmed by using the TCP4-specific forward primer 5′-ATGTCTGACGACCAATTCCATC-3′ and the VP16-specific reverse primer (see above).

Yeast one-hybrid assay

TCP4 and TCP4:VP16 fusion sequences were tested for trans-activation potential by yeast-one hybrid assay. ORFs were fused in-frame to the GAL4 DNA-binding domain in the pGBT9 prey vector, and introduced into yeast strain Y187 by the LiOAc/ssDNA/PEG method (Gietz et al., 1992; Bartel et al., 1993; Harper et al., 1993) Transformants were selected on synthetic drop-out medium lacking tryptophan, and β-galactosidase activity was measured using o-nitrophenyl galactoside (Sigma-Aldrich) as substrate, as described in the Yeast Protocols Handbook (Clontech, http://www.clontech.com).

Leaf and cell size measurements

The leaf lamina was taped onto a glass slide and photographed under a dissecting microscope. ImageJ software (http://rsbweb.nih.gov/ij/) was used to calculate length, width and area. For scanning electron microscopy (SEM), leaf epidermal impressions were made using dental wax (Coltene, http://www.coltene.com), and these were then used to make araldite casts. These were coated with gold and imaged using SEM (FEI Quanta 200, http://www.fei.com). To estimate the mean cell area from the tip, middle and base of leaf, images were analyzed using Quanta software. The area for at least 100 cells in a given region was estimated to obtain the mean value. Means obtained from 3–5 leaves were averaged, and the standard error of the means was calculated.

Induced senescence assay

Mature leaves were detached and incubated in the dark on water-soaked filter papers for 4 or 5 days, as described previously (Oh et al., 1996). Photographs of the leaves were taken before and after incubation using a SONY Cyber-shot digital camera (http://www.sony.co.in/productcategory/cybershot-digital-camera). For chlorophyll estimation, the middle portion of the detached leaf was weighed, cut into pieces and incubated in 1 ml of 80% acetone overnight at 4°C. The absorbance of acetone-extracted pigments was measured using an ELISA reader (Beckman, https://www.beckmancoulter.com) at 645 and 663 nm against an 80% acetone blank. Measurements were performed in triplicate. The amount of total chlorophyll was calculated as described previously (Arnon, 1949).

GUS assay

Plant tissue was collected in a microfuge tube and placed in 90% acetone on ice until all samples were harvested. Samples were washed several times in staining buffer (50 mm phosphate buffer pH 7.2, 0.5 mm potassium ferricyanide, 0.5 mm potassium ferrocyanide and 10 mm EDTA) on ice (three times, 30 s each), followed by overnight incubation with 2 mm X-Gluc [5-bromo-4-chloro-3-indolyl β-d-glucuronide (cyclohexamine salt), prepared in staining buffer] (Sigma-Aldrich) at 37°C. After washing the samples with 70% ethanol at room temperature, the cleared tissue was analyzed under a dissecting microscope (WILD Heerbrugg, http://www.wild-heerbrugg.com/) and photographed using Nikon Coolpix digital camera (http://www.nikon.com/).

Assays for pollen viability

Mature anthers from open flowers were collected in a solution containing 0.1 m sodium phosphate buffer, pH 7.0, 1 mm EDTA, 0.1% Triton X-100 and 0.4 mg ml−1 DAPI (4′,6-Diamidino-2-phenylindole, Sigma-Aldrich), vortexed to release pollen, and incubated overnight at 4°C. Pollen grains were analyzed under a fluorescence microscope (Olympus, http://www.olympus-global.com/). For in vitro germination, mature anthers were dehydrated at room temperature for 90 min and incubated on germination medium (15% w/v sucrose, 0.4 mm Ca(NO3)2 and 0.4 mm H3BO3, solidified with 1% w/v agarose) for 8 h at room temperature under moist conditions, and observed using a differential interference contrast microscope (Olympus) (Carpenter et al., 1992). For the in vivo germination experiment, carpels were collected 24 h after pollination, fixed in ethanol/acetic acid (3:1) for 1 h at room temperature, and left overnight in 8 m NaOH for softening. After several washes in distilled water, they were incubated in 0.1% aniline blue in 0.1 m K2HPO4/KOH buffer, pH 11, for 3 h, and analyzed using a fluorescence microscope (Olympus) (Muschietti et al., 1994).

Molecular biology

Cloning and PCR were performed according to standard protocols (Sambrook and Russell, 2001). A Qiagen kit (http://www.qiagen.com/) was used for RNA isolation. Reverse transcription was performed on 1 μg of DNase-treated RNA using Revert-aid reverse transcriptase (Fermentas, http://www.fermentas.com) according to the manufacturer’s protocol. Real-time PCR was performed using a SYBR Green kit (Finnzymes, http://www.finnzymes.com), according the manufacturer’s protocol. Data analysis was performed using ABI Prism 7900HT SDS software (Applied Biosystems, http://www.appliedbiosystems.com).

Statistical analysis

All analyses were performed using Student’s t-test (SigmaPlot, http://www.sigmaplot.com), except for the plastochron index, where linear regression analysis was used (GraphPad Prism, http://www.graphpad.com).

Acknowledgements

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

We thank Pilar Cubas (Departamento de Genética Molecular de Plantas, Centro Nacional de Biotecnología/CSIC, Campus Universidad Autónoma de Madrid, Madrid, Spain), Nancy Dengler (Department of Ecology & Evolutionary Biology, University of Toronto, Toronto, Canada) and Detlef Weigel (Department 6 Molecular Biology, Max Planck Institute for Developmental Biology, Tuebingen, Germany) for the tcp4-1, cyc1At::GUS and jaw-D lines, respectively, Nam Hai Chua (Laboratory of Plant Molecular Biology, The Rockefeller University, New York, USA) for the pER8 plasmid, Dr Parag Sadhale for the yeast strain and vectors and Mainak Das Gupta for help with RNA in situ hybridization. K.S. was supported by a fellowship from Indian Institute of Science. U.N. was supported by a grant from Department of Biotechnology, Government of India (BT/PR/5096/AGR/16/431/2004).

References

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

Supporting Information

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

Figure S1. Expression pattern of TCP4:VP16-C transgene. (a). RT-PCR with TCP4:VP16-C-specific primers (upper panel) and UBQ10-specific primers (lower panel) on cDNA prepared from whole seedlings of indicated genotypes. (b). In situ hybridization showing expression pattern of the fusion transcript in a 1.5 mm long TCP4:VP16-C rosette leaf. Transverse leaf sections are arranged from tip (top) to base (bottom). Sense control did not show any expression of the transgene (data not shown). Scale bar = 200 μm.

Figure S2. Leaf growth in TCP4:VP16-C. (a, b). Average length (a) and width (b) of lamina plotted as a function of days after emergence of the 7th rosette leaf for tcp4-1 (●) and TCP4:VP16-C (○). Error bar indicates SD, n = 3. (c, d). Leaf length plotted against leaf width for the 8th (c) and the 9th (d) rosette leaf in tcp4-1 (●) and TCP4:VP16-C (○). (e). A rosette leaf from a short-day grown TCP4:VP16-C plant flattened on a glass plate by introducing cuts at the margins. Triangle marks a representative gap generated in the leaf margin due to flattening. Scale bar = 1 cm.

Figure S3. Floral organ size unaffected in TCP4:VP16-C. (a). Open flowers from indicated genotypes. Scale bar = 1 mm. (b, c). Mature petals are flattened (b) and their average area measured (c). Scale bar = 1 mm. Error bars in (c) indicate SD, n = 5. (d). Average length of stamen and carpel at maturity plotted for different genotypes. Error bar indicates SD, n = 5.

Figure S4. Growth of indeterminate organs in TCP4:VP16-C. (a, b). Comparison of root growth (a) and average lengths of primary (open bar), secondary (black bar) and tertiary (grey bar) roots (b) in indicated genotypes. Scale bar in (a) =1 cm. (c). Comparison of number of lateral roots. Error bar indicates SD, n ≥ 5. (d). Rate of increase in the length of proliferation zone in roots of tcp4-1 (●) and TCP4:VP16-C (○) plotted against days after sowing. Error bars indicate SD, n ≥ 6. (e). Comparison of stem length at maturity in indicated genotypes. Error bars indicate SD, n ≥ 15.

Figure S5. Premature arrest of cell division in TCP4:VP16-C. (a–e). Expression of mitosis-specific cyc1At::GUS reporter in the 7th rosette leaf of Col-0 and TCP4:VP16-C at various stages of leaf growth (leaf length indicated at the bottom). Scale bars = 0.2 mm in (a, b), 0.5 mm in (c, d), 1 mm in (e).

Figure S6. Comparison of cotyledon growth in Col-0 and tcp4-1 in presence of GA3. Average area of mature cotyledons from long-day grown seedlings at 17 DAS plotted against GA3 concentration for Col-0 (grey bar) and tcp4-1 (black bar). Error bars indicate SD, n = 10.

Table S1. Comparison of onset of flowering in tcp4-1 and TCP4:VP16-C under long days. Flowering time for tcp4-1 and TCP4:VP16 was measured as FT50 (number of days taken for 50% of population to flower) and as number of rosette leaves (mean ± SD). The numbers in parentheses refer to number of plants taken for study.

Table S2. Result of crosses performed to test pollen and carpel function in TCP4:VP16-C. “Percentage success” indicates the percentage of crossed flowers that formed elongated siliques with at least one bold seed.

Appendix S1. Experimental procedures for in situ hybridization and root proliferation zone analysis.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
TPJ_4616_sm_FigS1.tif5717KSupporting info item
TPJ_4616_sm_FigS2.tif4084KSupporting info item
TPJ_4616_sm_FigS3.tif7552KSupporting info item
TPJ_4616_sm_FigS4.tif2246KSupporting info item
TPJ_4616_sm_FigS5.tif8743KSupporting info item
TPJ_4616_sm_FigS6.tif3651KSupporting info item
TPJ_4616_sm_FullSupplMaterial.doc40KSupporting info item
TPJ_4616_sm_TableS1.tif5170KSupporting info item
TPJ_4616_sm_TableS2.tif5434KSupporting info item

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.