Brassinosteroid (BR) hormones control plant growth through acting on both cell expansion and division. Here, we examined the role of BRs in leaf growth using the Arabidopsis BR-deficient mutant constitutive photomorphogenesis and dwarfism (cpd).
We show that the reduced size of cpd leaf blades is a result of a decrease in cell size and number, as well as in venation length and complexity. Kinematic growth analysis and tissue-specific marker gene expression revealed that the leaf phenotype of cpd is associated with a prolonged cell division phase and delayed differentiation.
cpd-leaf-rescue experiments and leaf growth analysis of BR biosynthesis and signaling gain-of-function mutants showed that BR production and BR receptor-dependent signaling differentially control the balance between cell division and expansion in the leaf.
Investigation of cell cycle markers in leaves of cpd revealed the accumulation of mitotic proteins independent of transcription. This correlated with an increase in cyclin-dependent kinase activity, suggesting a role for BRs in control of mitosis.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
In Arabidopsis, leaves are initiated at the flanks of the shoot apical meristem by cell division, followed by gradual (tip to the base) maturation marked by cell cycle exit and cell expansion (Donnelly et al., 1999; Efroni et al., 2010; Andriankaja et al., 2012). Genetic, hormonal and environmental factors control leaf growth and development (Tsukaya, 2002; Bögre et al., 2008; Skirycz & Inzé, 2010). Brassinosteroid (BR) hormones are known to be essential for leaf growth because BR loss-of-function mutants display severe leaf phenotypes including small, round-shaped leaves and short petioles (Clouse et al., 1996; Szekeres et al., 1996; Li & Chory, 1997; Choe et al., 1998; Friedrichsen et al., 2000; Li et al., 2001). Conversely, BR gain-of-function mutants show larger and elongated leaf blades and longer petioles (Choe et al., 2001; Wang et al., 2001; Yin et al., 2002; Gonzalez et al., 2010; Oh et al., 2011). The dwarf leaf phenotype of the BR loss-of-function mutants is attributed mainly to impaired cell expansion because of the smaller cell size of these mutants (Chory et al., 1991; Clouse et al., 1996; Kauschmann et al., 1996; Szekeres et al., 1996; Fujioka et al., 1997; Azpiroz et al., 1998; Choe et al., 1999). However, exogenous BRs rescue the leaf size of the de-etiolated2 (det2) mutant, defective in the biosynthesis of BRs, by increasing both mesophyll cell size and number, implying that BRs promote leaf growth by positively regulating both cell expansion and cell division (Nakaya et al., 2002). In agreement with this conclusion, the enlarged leaf size of Arabidopsis plants overexpressing the BR receptor BR INSENSITIVE 1 (BRI1) and the gain-of-function phosphorylation mutant BRI1 (Y831F) is a result of an increase in the epidermal cell number (Gonzalez et al., 2010; Oh et al., 2011), suggesting that BRs promote cell division during leaf growth. A short-term promotion of cell division by BRs is also observed in tobacco (Nicotiana tabacum) protoplasts and Bright Yellow-2 (BY-2) cells (Nakajima et al., 1996; Oh & Clouse, 1998; Miyazawa et al., 2003), whereas impaired BR biosynthesis in callus cells of Arabidopsis results in slower cell division rates (Cheon et al., 2010). Also, the short root phenotype of the null BR receptor mutant bri1-116 is caused by defects in both cell expansion and in the cell cycle progression in the root meristem (González-García et al., 2011; Hacham et al., 2011).
To date, the mechanism of BR-mediated control of cell division remains elusive. Several reports have shown that BRs enhance the transcription of core cell cycle genes such as CYCLIN D3 (CYCD3), CYCLIN B1;1 (CYCB1;1), HISTONE 4 (H4) and the B-type CYCLIN-DEPENDENT KINASE B1;1 (CDKB1;1) in Arabidopsis plants, cultures, and BY-2 cells (Yoshizumi et al., 1999; Hu et al., 2000; Miyazawa et al., 2003). Conversely, BR loss-of-function mutants display a reduced expression of CYCB1;1 and an increased expression of the cell cycle inhibitor KIP-RELATED PROTEIN 2 (KRP2) (González-García et al., 2011; Hacham et al., 2011). Identification of several cell cycle genes as direct targets of both the BRASSINAZOLE RESISTANT 1 (BRZ1) and the BRI1 EMS SUPPRESSOR 1 (BES1)/BZR2 transcription factors might imply a transcriptional control of the cell cycle by BRs (Sun et al., 2010; Gudesblat & Russinova, 2011; Yu et al., 2011).
In addition to cell cycle progression, BRs are required for differentiation of the cells in the leaf margin (Reinhardt et al., 2007) and for defining the timing of differentiation in the root (González-García et al., 2011). BRs also control the differentiation of vascular tissues in the shoot, as Arabidopsis and rice (Oryza sativa) BR loss-of-function mutants have less xylem (Caño-Delgado et al., 2004; Nakamura et al., 2006; Ibañes et al., 2009).
To understand how BRs promote leaf growth, we analyzed cell division, expansion, and differentiation in Arabidopsis leaves of the BR-deficient mutant constitutive photomorphogenesis and dwarfism (cpd). We provide evidence that BRs are essential for cell cycle progression during leaf growth. cpd-rescue experiments and investigation of BR biosynthesis and signaling gain-of-function mutants show that BR production and BR receptor-dependent signaling have distinct effects on cell division and expansion in the leaf. Several mitotic proteins accumulated in cpd mutants independent of transcription. This correlated with an increase in cyclin-dependent kinase activity, suggesting the hypothesis that BRs are required for the degradation of mitotic regulators.
Materials and Methods
Plant material, marker lines and growth conditions
Arabidopsis thaliana (L.) Heynh. ecotype Columbia-0 (Col-0) plants and mutants in the same genetic background were used except for the Pro35S:CYCD3;1 (CYCD3;1OE) transgenic line (Dewitte et al., 2003), which was in the Landsberg erecta ecotype. cpd (Szekeres et al., 1996), bri1-116, and ProBRI1:BRI1-GFP (BRI1OE) (Friedrichsen et al., 2000) were crossed with the following transcriptional or translational reporters: ProCYCA2;1:GUS (Vanneste et al., 2011), ProCYCB1;1:GUS (De Smet et al., 2008), ProCYCB1;1:DB-GUS (Colón-Carmona et al., 1999), ProCYCB1;2:DB-GUS (Schnittger et al., 2002), Pro35S:GFP-TUB6 (Nakamura et al., 2004), ProATHB15:nlsYFP (nls-nuclear localization signal), ProPIN1:GUS (Friml et al., 2003), and ProDR5:GUS (Ulmasov et al., 1997). The ProATHB15:nlsYFP construct contained the 2-kb promoter fragment of the ATHB15 gene cloned in the pDONR™-221 vector and later in the pBGYN binary vector using Gateway® cloning technology (Invitrogen, Carlsbad, CA, USA; Supporting Information Table S1). These constructs were introduced into Agrobacterium tumefaciens C58pMP90 and transformed into Col-0 plants by floral dipping (Clough & Bent, 1998). The mutant Pro35S-DWF4 (DWF4OE) (Wang et al., 2001) was also used. Transgenic lines and mutants were genotyped using primers listed in Table S1.
Before transfer to the light, seeds were incubated in the dark at 4°C for 48 h. Plants were grown under standard growth conditions (21°C; 16 h : 8 h, light : dark photoperiod) on vertical plates containing half-strength Murashige and Skoog medium supplemented with 0.8% plant tissue culture agar (LAB M Ltd, Heywood, UK) and 1% sucrose. The bri1-116 and BRI1OE mutants were grown in growth chambers (21°C; 16 h : 8 h, light : dark photoperiod; light intensity 100 μmol m−2 s−1 photosynthetically active radiation; 55% humidity), in peat pellets (Jiffy International As, Stange, Norway). The cpd-rescue experiment was performed at 8 d after sowing (DAS) in vitro on medium supplemented with dimethylsulfoxide (DMSO) for the mock treatments, or with 100 nM BL (Wako Pure Chemical Industries, Osaka, Japan). The vertical-grown plants at 8 DAS were transferred to DMSO- or BL-containing medium in sterile conditions. To avoid contact with the agar, the rosettes were placed on sterile parafilm strips, whereas the roots were kept in contact with the medium.
Leaves were harvested and cleared in the following series of solutions: absolute ethanol (1–2 d), 1.25 M NaOH : EtOH (1 : 1 v/v) solution for 2 h at 60°C, lactic acid saturated in chloral hydrate, and finally lactic acid. The leaves were mounted on microscopic slides with the abaxial side upwards. Whole leaves were photographed using a Nikon camera connected to a binocular MZ16 Leica microscope for subsequent measuring of the leaf blade area with ImageJ software (Abràmoff et al., 2004). Cells at the abaxial epidermis and the palisade mesophyll were imaged using differential interference contrast (DIC) BX51 microscopes (Olympus Europa, Munich, Germany) at 25% (leaf base) and 75% (leaf base) of the leaf blade length. The number of cells (50 cells on average) per drawn image area was counted in ImageJ and extrapolated to the whole leaf area. The average cell size at the leaf base and tip was calculated from the ratio of image area/cell number. In total eight leaves per genotype or treatment were analyzed and at least two independent experiments were performed. The kinematic analysis was performed according to De Veylder et al. (2001) and Achard et al. (2009). A detailed description of the venation pattern analysis methods is given in Dhondt et al. (2012). The following parameters characterizing the venation pattern were measured: number of areolas (i.e. the area closed by vasculature), the number of branching points (where a vein is crossed by another vein), the number of end points (where a vein stops), and venation pattern complexity (i.e. the sum of the number of all branching and endpoints, and the connections between the veins).
Pavement cell circularity was extracted automatically from the epidermal microscopic drawings using an ImageJ macro. Circularity was calculated according to the following formula: 4πA/P2, where A is the area and P is the perimeter of the cell. A circularity value of 1 represents a perfect circular morphology, and a value of 0 indicates an elongated polygon.
The percentage of mitotic events visualized using TUBULIN6 (TUB6)- and CYCB1;2-positive cells (M cells) per total mesophyll cell number was calculated according to Donnelly et al. (1999). To measure the frequency of mitotic events and M cells at the tip and the base of the leaf, 100 cells per region were taken and the respective ratios (mitotic events per 100 cells, and M cells per 100 cells) were calculated.
For flow cytometry, frozen leaves were chopped with a razor blade in 200 μl of Cystain UV Precise P Nuclei extraction buffer followed by the addition of 800 μl of staining buffer (buffers from Partec, Münster, Germany), and the mix was filtered through a 50-μm mesh. The distribution of the nuclear DNA content was analyzed using a CytoFlow ML flowcytometer and FloMax software (Partec, Münster, Germany).
Transgenic Arabidopsis leaves were assayed for GFP and YFP fluorescence either with a 100 M Zeiss confocal microscope with software package lsm 510 version 3.2 or with an Olympus FluoView FV1000 microscope equipped with a ×63 water-corrected objective (numerical aperture of 1.2). The GFP fluorescence was excited with a 488-nm laser. Emission fluorescence was captured in the frame-scanning mode alternating GFP fluorescence via a 500–550-nm band-pass filter.
For analysis of GUS activity, plants were submerged in 90% acetone for 30 min at 4°C, washed once with phosphate buffer (pH 7) at room temperature and incubated in GUS solution (1 mM X-Glc; 0.5% dimethylformamide, 0.5% Triton X-100, 1 mM EDTA, 0.5 mM K3Fe(CN)6 and 0.5 mM K4Fe(CN)6 in phosphate buffer) at 37°C. After washing in phosphate buffer at room temperature, the samples were kept in 70% ethanol for chlorophyll destaining. The leaves were cleared as described in the ‘Growth analysis’ section and mounted in lactic acid. Images of GUS-stained plants were taken with the binocular microscope (MZ16; Leica) and Nikon camera.
RNA and quantitative RT-PCR
Total RNA was extracted from the first leaf pair with the RNeasy kit (Qiagen). Poly(dT) cDNA was prepared from 1 μg of total RNA with Superscript III reverse transcriptase (Invitrogen, Carlsbad, CA, USA) and analyzed on an LightCycler 480 apparatus (Roche Diagnostics, Basel, Switzerland) with the SYBR Green I Master kit (Roche Diagnostics, Basel, Switzerland) according to the manufacturer's instructions. Gene expression was quantified with a specific primer set (Table S1) designed with Beacon Designer 4.0 (Premier Biosoft International, Palo Alto, CA, USA). All individual reactions were performed in triplicate. Data were analyzed with qBase (Hellemans et al., 2007). Expression levels were normalized to those of 26S PROTEASOME REGULATORY SUBUNIT S2 1A (PRS2)/REGULATORY PARTICLE NON-ATPASE1 (RPN1).
Immunoblot and protein quantification analysis
To extract proteins, the first leaf pair of plants at 8 or 12 DAS was frozen in liquid nitrogen, ground and homogenized in ice-cold extraction buffer (25 mM Tris–HCl, pH 7.6, 15 mM MgCl2, 5 mM EGTA, 150 mM NaCl, 15 mM p-nitrophenyl phosphate, 60 mM β-glycerophosphate, 0.1% (v/v) Nonidet P-40, 0.1 mM sodium vanadate, 1 mM NaF, 1 mM dithiothreitol, 1 mM phenylmethylsulfonylfluoride, and 1 tablet per 10 ml of protease inhibitor cocktail (Roche, Basel, Switzerland)). The homogenate was centrifuged twice for 20 min at 20 800 g and 4°C and the protein concentration determined with QuickStart™ Bradford1X dye reagent (Bio-Rad Laboratories, Hercules, CA, USA). Approximately 45 μg of total proteins were mixed with 5X SDS sample buffer (250 mM Tris–HCl, pH 6.8, 50% glycerol, 5% SDS, 0,02% bromophenol blue and 0.5 M DTT), boiled for 10 min at 95°C, separated on a 15% (w/v) SDS-PAGE gel and further transferred to nitrocellulose membranes (Hybond-C super; GE-Healthcare, Little Chalfont, UK). Membranes were then stained with 0.1% (w/v) Ponceau solution and blocked overnight at 4°C. For immunodetection, monoclonal mouse anti-GFP antibody (Living Colors®; Clontech, Mountain View, CA, USA), polyclonal rabbit anti-KNOLLE antibody (gift from Gerd Jürgens) at 1 : 4000 dilution, and polyclonal rabbit anti-CDKA;1 PSTAIRE (Santa Cruz Biotechnology, Santa Cruz, CA, USA) at 1 : 2000 dilution were used as primary antibodies. A secondary anti-rabbit antibody (Amersham/GE Healthcare, Little Chalfont, UK) was used at 1 : 10 000 dilution. The proteins were detected by chemiluminescence reaction (Perkin-Elmer, Waltham, MA, USA). The images obtained were quantified using ImageJ software (Abràmoff et al., 2004) by inverting the image and measuring the band of interest as the mean value of pixels per unit area. The set of bands were compared with corresponding nonspecific bands in order to subtract the background. Finally, the relative values of the signal were calculated as compared with the control sample. The number of biological repetitions for each experiment is indicated in each legend and the average value of the significant fold-change is presented below the representative blot.
P10CKS1At was purified from an overproducing Escherichia coli strain and linked to CNBr-Sepharose 4B (GE-Healthcare, Little Chalfont, UK) according to Azzi et al. (1994). The beads were washed three times with homogenization buffer (HB) containing 25 mM Tris–HCl (pH 7.6), 60 mM β-glycerophosphate, 15 mM nitrophenyl phosphate, 15 mM EGTA (pH 8), 15 mM MgCl2, 85 mM NaCl, 1 mM dithiothreitol, 0.1 mM vanadate, 1 mM NaF, 0.1 mM benzamidine, 1 mM phenylmethylsulfonylfluoride, 0.1% NP-40, and 1 tablet per 10 ml protease inhibitor cocktail (Roche, Basel, Switzerland). Protein extracts were prepared from the first leaf pair of the Col-0 wild type and cpd mutant in HB. In a total volume of 200 μl of HB, 350 μg of protein extract was loaded on 50 μl of 50% (v/v) p10CKS1At-Sepharose beads and incubated on a rotating wheel for 2 h at 4°C. After brief centrifugation at 110 g and removal of the supernatant, the beads were carefully washed three times with bead buffer containing 50 mM Tris–HCl (pH 7.5), 5 mM NaF, 250 mM NaCl, 5 mM EGTA, 5 mM EDTA, 0.1% NP-40, 0.1 mM benzamidine, 0.1 mM vanadate and 1 tablet per 10 ml protease inhibitor cocktail (Roche, Basel, Switzerland). The beads were washed once with kinase buffer (50 mM Tris–HCl (pH 7.8), 15 mM MgCl2, 5 mM EGTA and 2 mM dithiothreitol), and the supernatant was removed carefully. The histone H1 kinase reactions were initiated by resuspending the pellets of p10CKS1At-Sepharose beads in 35 μl of the reaction mixture containing 5 μCi [γ-33P] ATP (3000 Ci mmol−1), 0.5 mg ml−1 histone H1, 50 mM Tris–HCl (pH 7.8), 15 mM MgCl2, 5 mM EGTA, 1 mM dithiothreitol, 60 μg ml−1 cAmp-dependent kinase inhibitor and 10 μM ATP. After 20 min of incubation at 30°C, the kinase reactions were stopped by the addition of 10X SDS/PAGE loading buffer. Aliquots were boiled, loaded on a 12% (w/v) acrylamide gel, and stained using Coomassie blue. The gel was dried overnight and incorporation of [γ-33P] ATP into histone H1 was quantified with a PhosphorImager (Molecular Dynamics, Sunnyvale, CA, USA).
P values were calculated using a two-tailed Student's t -test and Excel software.
BRs regulate leaf growth by promoting cell expansion, proliferation and differentiation
To understand how BRs regulate leaf growth, we compared leaves of wild-type plants with those of the BR loss-of-function mutant cpd which is deficient in a gene encoding a key enzyme, cytochrome P450 (CYP90), in the BR biosynthesis pathway (Szekeres et al., 1996). The decreased BR levels in the cpd mutant resulted in fewer and significantly smaller, round-shaped leaves, as seen in leaf series of in vitro grown plants at 21 DAS (Fig. 1a). To gain insight into the cellular and developmental basis of this phenotype, the abaxial epidermis and palisade mesophyll of the first true leaf pair of the cpd mutant were analyzed. As under in vitro conditions the growth of those leaves is subdivided into three developmental phases, samples were collected at 8 DAS, corresponding to active cell proliferation, 12 and 16 DAS, corresponding to cell expansion, and 21 DAS, when both cell division and cell expansion have ceased (De Veylder et al., 2001; Beemster et al., 2005). At 21 DAS, the area of the cpd leaf blade showed a 70% reduction when compared with the wild type, which correlated with a significant decrease in both cell size and cell number in the epidermis and mesophyll (Fig. 1a–d). Interestingly, while the cell size was more reduced in the mesophyll, the cell number was more affected in the epidermis. The observed defects in epidermal cell expansion in the cpd mutant correlated with a less complex shape of the pavement cells, demonstrated by the higher circularity value (Andriankaja et al., 2012) than that of the wild type (Fig. S1a,b).
To study the effect of BRs on leaf differentiation, we analyzed the leaf vasculature in the cpd mutant using parameters that quantitatively describe the leaf venation pattern (Dhondt et al., 2012). When compared with the wild type, the leaves of the cpd mutant displayed a reduction in venation pattern length, number of areolas and branching end points, which resulted in an overall reduced venation pattern complexity (Figs 1e, S1c,d). Taken together, the described phenotypes highlight a role for BRs in leaf growth through promoting cell expansion, proliferation and differentiation, thus affecting the final size of the leaf.
BRs balance cell proliferation and expansion in the leaf
We next aimed to clarify whether the observed leaf phenotypes in the cpd mutant are attributable to disturbed coordination between cell division and expansion during leaf growth. Therefore, the cell division rate (De Veylder et al., 2001; Achard et al., 2009) was determined in the abaxial leaf epidermis of the cpd mutant and the wild type (Fig. S2a). Because of the dwarf nature of cpd, the earliest time-point examined was 8 DAS. The kinematic analysis showed that at this stage the cell division rate was lower in cpd than in the wild type, but between 10 and 14 DAS, when the transition from proliferation to expansion occurred, it declined more slowly in the mutant, indicating a delayed exit from cell proliferation. The average cell cycle duration (Fiorani & Beemster, 2006) in the whole leaf at 8 DAS was higher in cpd than in the wild type, being 31 and 25 h, respectively. Ploidy levels were measured at time intervals corresponding to the kinematic analysis and no significant changes were observed when compared with the wild type (Fig. S2b). We conclude that the reduced cell number in the epidermis of the cpd mutant is caused by a slower progression through the cell cycle and a delayed exit from cell division.
We further explored how BR deficiency affects the spatial coordination between cell division and cell expansion in the leaf at early developmental stages, namely 7 and 8 DAS. The cell division activity in the leaf abaxial epidermis was fine-mapped using the tubulin marker Pro35S:GFP-TUB6 (Snustad et al., 1992; Nakamura et al., 2004), which visualizes the mitotic spindles denoting mitotic events (Fig. 2a). The epidermal cell size and the frequency of mitotic events were measured at the leaf base (at 25% of the leaf blade length) and at the leaf tip (at 75% of the leaf blade length) at 7 DAS. When compared with the wild type, the fraction of mitotic cells at the leaf tip of the cpd mutant was higher (Fig. 2b) and correlated with a reduction in epidermal cell size (Fig. 2c) and number (Fig. 1d). Also, the percentage of mitotic events per total epidermal cell number was slightly, but significantly, increased in cpd leaves at 7 DAS (Fig. S3a). These data indicate that in cpd epidermal cells reside in the M phase for longer.
We next investigated whether leaf differentiation is delayed in cpd by analyzing the expression of the transcription factor ATHB15 which negatively regulates vascular cell differentiation (Ohashi-Ito & Fukuda, 2003). At 7 DAS, the ProATHB15:nlsYFP fluorescence was observed in the differentiating vascular bundles of the wild-type leaf base, while in cpd leaves the expression was observed in a broader area along the leaf length (Fig. 2d), suggesting a delay in differentiation. The prolonged ProATHB15:nlsYFP activity in cpd leaves correlated with more than a fourfold increase in the accumulation of YFP protein in whole cpd leaf extracts (Fig. 2e). Thus, our results support a role for BRs in controlling the exit from mitosis in the leaf.
BRs control the exit from mitosis
To further unravel the nature of the M-phase defects in cpd leaves, we analyzed the expression of the B-type cyclin marker ProCYCB1;2:DB-GUS, which accumulates during mitosis and is degraded at anaphase as a result of the presence of a destruction box (DB; Colón-Carmona et al., 1999; Schnittger et al., 2002), in the leaf mesophyll of cpd and wild-type plants at 8 DAS (Fig. 2f). cpd displayed an increase in the number of cells labeled with CYCB1;2:DB-GUS, both at the base and at the tip of the leaf, whereas in the wild type, CYCB1;2:DB-GUS-labeled cells were confined to the leaf base (Figs 2f,g, S3b). In accordance with this finding, the transition toward cell expansion was delayed in the mesophyll cells of cpd (Fig. 2h). To understand why ProCYCB1;2:DB-GUS accumulates in cpd, we measured the CYCB1;2 and GUS transcript levels in leaves of the same lines (Figs 3a, S3c). Although a slight increase in GUS transcript levels was observed in cpd (Fig. S3c), the endogenous B-type cyclin transcripts were slightly reduced in this mutant cpd (Fig. 3a). Similarly, the transcription of the M phase-expressed syntaxin KNOLLE (Völker et al., 2001) was not increased (Fig. 3a), whereas a significant accumulation of its protein was detected in cpd at 8 DAS by immunoblot analysis (Fig. 3b). Because CYCLIN-DEPENDENT KINASE (CDK) activity is known to peak during mitosis, we tested the total CDK activity in leaves. Notably, 20% higher CDK activity was observed in cpd compared with wild-type leaves (Fig. 3c). The CDKA;1 transcription and CDKA;1 protein levels were not changed (Fig. 3a,b).
To investigate whether an increase in mitotic activity is sufficient to reverse the BR-deficient leaf phenotype, a transgenic line overexpressing CYCD3;1 (CYCD3;1OE) (Dewitte et al., 2003) was introduced into the cpd mutant (Fig. 4a,b) and analyzed at 12 DAS. Overexpression of CYCD3;1 restored the mitotic activity in cpd, as leaves of cpd;CYCD3;1OE plants contained numerous small cells both in the epidermis and in the mesophyll, similarly to the CYCD3;1OE line (Fig. 4c,d). As expected for dividing tissues, the protein level of KNOLLE was increased (Fig. S4). Although the high CYCD3;1 expression in cpd was able to restore the cell number deficit, the cells remained small at both the base and the tip of the leaf, demonstrating a defect in the transition toward expansion along the leaf length (Fig. 4e,f). These results show that a complete restoration of the BR-deficient leaf phenotypes requires BR-mediated cell expansion and differentiation.
Exogenous application of BRs triggers differentiation in the leaf
It was previously shown that the dwarf leaf phenotype of the cpd mutant is restored to the wild-type size by exogenous application of the most active BR, brassinoslide (BL; Szekeres et al., 1996). We used this experimental system as a tool to investigate whether the cell expansion and division defects in the cpd mutant are completely restored by the BL treatment. For this, cpd plants at 8 DAS were transferred to medium supplemented with 100 nM BL and analyzed after 24–96 h (Fig. 5a). Forty-eight hours of BL treatment increased the size of the cpd leaf blade by 60%, whereas a longer application of BL had a mostly inhibitory effect on leaf growth. Interestingly, the enlargement of the leaf blade at 48 h was a result of induced cell expansion in both the epidermis and the mesophyll (Fig. 5b). Notably, a slight, but significant, increase in cell number was found only in the epidermis (Fig. 5c). BL also promoted cell expansion along the leaf in the epidermis and mesophyll (Fig. 5d), which correlated with a reduction of the percentage of mitotic cells labeled with CYCB1;2:DB in the mesophyll (Fig. 5e), as well as a marked decrease in the abundance of total KNOLLE and CDKA;1 proteins (Fig. 5f). Our data show that the exogenously applied BL triggers mitotic cell cycle exit and differentiation in the leaf. This observation was further supported by the cellular analysis of leaves of plants overexpressing a key BR biosynthetic enzyme, DWARF4 (DWF4OE), and thus presumably accumulating BRs (Choe et al., 2001; Wang et al., 2001). At 12 DAS, the size of both epidermal and mesophyll cells was increased compared with the wild type, while the total cell number was unaffected in the epidermis and slightly, but significantly, reduced in the mesophyll (Fig. S5a–c).
Application of BL also enhanced the vascular differentiation of the leaf in cpd after treatment with BL for 48 h, as seen from the expression of three vascular markers, that is, the A2-type cyclin ProCYCA2;1:GUS (Fig. 5g; Vanneste et al., 2011), the auxin efflux carrier ProPIN1:GUS (Friml et al., 2003), and the auxin reporter ProDR5:GUS (Ulmasov et al., 1997; Fig. S5e). During early Arabidopsis leaf development, the CYCA2;1 and PIN-FORMED1 (PIN1) markers are expressed in relatively broad regions containing proliferating cells. Later their expression becomes restricted to the provascular strands, which is indicative of leaf differentiation (Scarpella et al., 2006; Wenzel et al., 2007). In cpd leaves at 9 DAS, the diffused GUS expression pattern persisted longer than in the wild type (Fig. S5d), whereas the application of BL accelerated the transition to bundle-specific expression (Fig. S5e). The delayed vascular differentiation caused by BR deficiency correlated with a decrease in auxin responses, as demonstrated by the reduced DR5:GUS expression in cpd leaves, while BL application had the opposite effect (Fig. S5d,e).
The increase in epidermal and mesophyll cell size, the reduced protein level of cell division proteins, and the enhanced vascular differentiation in cpd leaves treated with BL all suggest that addition of BRs stimulates the onset of differentiation. As the exit from the cell cycle is followed by an endoreduplication cycle (De Veylder et al., 2011), we measured the DNA content in leaf samples treated with BL. As predicted, application of BL for 72 h resulted in an accelerated mitotic exit, as demonstrated by the increased fraction of 8C cells, marking endoreduplication (Fig. 5h).
BRI1-dependent signaling fine-tunes leaf growth in Arabidopsis
To explore the role of BR signaling in leaf development, we analyzed the respective gain- and loss-of-function Arabidopsis BR signaling mutants BRI1OE and bri1-116 (Friedrichsen et al., 2000). Plants were grown in soil and the leaf cell size and cell number were analyzed at 8, 12, 16, and 21 DAS (Fig. 6a–c). When compared with the wild type, bri1-116 leaves at 21 DAS had smaller and fewer cells in both the epidermis and the mesophyll (Fig. 6b,c), which correlated with the 80% reduction in the leaf blade area (Fig. 6a). By contrast, the leaves of the BRI1OE line contained a higher cell number and slightly larger cells in both the epidermis and the mesophyll (Fig. 6b,c), resulting in an 80% increase in the leaf blade area when compared with the wild type at 21 DAS (Fig. 6a). Similarly to cpd, the mitotic ProCYCB1;2:GUS accumulated in bri1-116 when the DB motif was present (Fig. 6d). Consistent with the observed increase in cell number in the BRI1OE leaves, the expression of ProCYCB1;2:DB-GUS and ProCYCB1;1:DB-GUS was also increased (Figs 6d, S6a). Correspondingly, the percentage of mitotic cells in both bri1-116 and BRI1OE leaves was higher than in wild-type leaves (Fig. 6e). As in cpd, the bri1-116 leaves displayed a delay in both cell elongation and vascular development along the leaf (Fig. 6f,g). By contrast, compared with the wild type, at 12 DAS the BRI1OE leaves were characterized by accelerated cell expansion along the leaf blade (Fig. 6f), as well as vasculature development defined by the faster decline in the expression of the provascular markers CYCA2;1:GUS and PIN1:GUS at the leaf tip (Figs 6g, S6b). Notably, the expression of DR5:GUS in the BRI1OE leaves did not increase (Fig. S6c), which differed from wild-type leaves treated with BL (Bao et al., 2004). We conclude that the enhanced BR signaling in BRI1OE plants favors optimal BR responses for stimulating epidermal and mesophyll cell division and cell expansion, as well as vasculature development in the leaf.
BR synthesis and BRI1-mediated signaling differentially affect cell division and expansion in the leaf
The final size and shape of the leaf are defined by the genetic program and environmental stimuli (Skirycz & Inzé, 2010; Gonzalez et al., 2012). Our study revealed that the plant hormones BRs promote leaf growth by affecting cell proliferation and expansion and this effect is strongly dependent on BR levels and BRI1-mediated signaling. As a general tendency, a deficiency in either BR biosynthesis or signaling caused smaller leaves with a decreased number and size of epidermal and mesophyll cells. Conversely, plants displaying constitutive BR responses including BR biosynthesis and signaling gain-of-function mutants had larger leaves. Whereas the larger leaf size in plants overproducing the BRI1 receptor (BRI1OE) in its native domain was attributable to an increase mainly in cell number, the larger leaves of plants with enhanced BR biosynthesis (DWF4OE) were solely caused by an increase in cell size. Similarly, the rescue of the small cpd leaves by exogenous application of BRs was a result more of cell expansion than division. Previous reports also showed that the growth promotion in leaves associated with BR signaling, in plants overexpressing either the BRI1 receptor or its phosphorylation variant BRI1 (Y831F), was coupled with an increase in cell number (Gonzalez et al., 2010; Oh et al., 2011). By contrast, the gain-of-function mutant of the BREVIS RADIX gene encoding a putative transcription factor with an impact on BR biosynthesis, resulting in increased BR levels, showed a promoting effect on cell expansion in the same organ (Beuchat et al., 2010). Based on these observations, we conclude that manipulation of BR signaling affects cell division, whereas manipulation of BR levels affects cell expansion. We speculate that the different cellular phenotypes observed in BR signaling and biosynthesis mutants are related to differences in the ratio of receptor (BRI1)/ligand (BL) concentrations, which consequently trigger downstream BR responses with distinctive outputs. Similarly, in Arabidopsis roots optimal BR signaling is required to promote meristematic divisions, whereas enhanced BR signaling caused by a gain-of-function mutation in BES1 (bes1-D) induced early differentiation (González-García et al., 2011). It was recently postulated that the density of the BRI1 receptor in different cell types of the Arabidopsis root is an important element in BRI1-mediated signaling, as it determines the cell sensitivity to BRs in different tissues (González-García et al., 2011; Hacham et al., 2011; Van Esse et al., 2011). It will be of interest to further investigate whether a similar scenario is applicable to the leaf.
The mature vegetative leaf of Arabidopsis consists of an outer epidermis, internal mesophyll and vasculature (Tsukaya, 2005). Earlier reports determined that BRI1 activity in the epidermis only is required for the growth of Arabidopsis leaves, although BRI1 expression was detected in almost all cells (Savaldi-Goldstein et al., 2007). Notably, we observed differences in the BR effect between the epidermis and the mesophyll in the leaves of BR loss- and gain-of-function mutants. Whereas cell division was affected more in the epidermis, the effect on cell size was more pronounced in the mesophyll. The molecular basis of these different responses remains unknown. It has been shown that the epidermis and palisade mesophyll in Arabidopsis leaves display a similar gradient in division pattern and they have a similar frequency of cell division (Andriankaja et al., 2012). Thus, while the basic mechanism of cell division is presumably common to both tissue layers, tissue layer-specific upstream signaling regulation, possibly via BR biosynthesis, catabolism, and signaling, might balance the growth behavior of each tissue within the leaf. In addition to cell size and cell number, the reduced leaf size of the cpd mutant was associated with a shorter venation pattern length and an overall reduced complexity, whereas increased BR signaling or application of BRs accelerated vascular differentiation in the leaf.
BRs regulate the boundary between cell division and expansion in the leaf
Mutants defective in BR biosynthesis or perception have characteristically small, round leaves (Kauschmann et al., 1996; Szekeres et al., 1996; Azpiroz et al., 1998), suggesting that they stopped growing along the leaf length. In agreement with this conclusion, exogenous application of BRs recovered the size and shape of the det2 mutant leaves by a stimulatory effect on cell division and cell elongation, predominately in the longitudinal direction (Nakaya et al., 2002). It was previously shown that leaf growth in Arabidopsis involves the establishment of a basiplastic (apex to base) cell division gradient along the blade. Thus, proliferative cell divisions initially are observed throughout the leaf, but gradually become restricted to more basal portions of the leaf blade (Donnelly et al., 1999). Recently, it was revealed that the boundary between cell division and expansion, called an arrest front, remains at a constant distance from the leaf blade base during early stages of leaf development (4–7 DAS) and later (7–8 DAS) moves to the leaf base and disappears (Kazama et al., 2010; Andriankaja et al., 2012). Indeed, in the cpd mutant this arrest front is retained longer than in the wild type, although our leaf staging might differ from that of Kazama et al. (2010) as a consequence of different growth conditions. However, because fewer cells were found in cpd leaves, we conclude that the longer maintenance of the arrest front in cpd is not a compensatory mechanism for cell expansion defects (Tsukaya, 2006; Bögre et al., 2008) but is probably a result of impaired cell production and differentiation. Further analysis will be essential to clarify the role of BRs in this process.
BRs control the exit from mitosis
Our study shows that BRs control the exit from mitosis. Even though the transcript levels of the mitotic marker genes CYCB1;1, CYCB1;2 and KNOLLE were slightly reduced in the cpd leaves, supporting recent observations made in seedlings of the BR-insensitive mutant bri1-116 (Hacham et al., 2011), an accumulation of proteins of the same markers was detected. Thus, it appears that M-phase-specific cell cycle proteins fail to degrade in the loss-of-function BR mutants. The increased CDK activity in cpd leaves also supports mitotic defects related to anaphase promoting complex/cyclosome (APC/C)-mediated regulation. APC/C has been implicated in leaf development, as the constitutive overexpression of the APC/C subunit APC10 in Arabidopsis enhanced leaf size as a result of enhanced rates of cell division and increased proteolysis of CYCB1;1 (Eloy et al., 2011). Remarkably, post-mitotic cells might also require APC/C activity for differentiation (Marrocco et al., 2009). Therefore, BRs might regulate division and/or differentiation through modulation of APC/C activity via an as yet unknown mechanism. A previous work linked the increase in CYCB1 protein levels with an aberrant function of the Arabidopsis separase, which is required at anaphase onset to separate the sister chromatids (Wu et al., 2010). However, a BR function in separase activities including chromatin maintenance or DNA repair is not known. The CYCB1 accumulation as a consequence of stabilization in different backgrounds results in disturbed cortical microtubule organization (Weingartner et al., 2004; Serralbo et al., 2006; Pérez-Pérez et al., 2008). Although the CYCB1 accumulation and the higher CDK activity in cpd leaves correlated with irregular epidermal cell shape reflecting cytoskeleton organization defects, it remains to be determined whether this is the case in BR loss-of-function mutants.
In conclusion, we demonstrate that BRs are essential for cell division, expansion and differentiation in the leaf. The balance between proliferation and differentiation in a temporal and spatial manner depends on BR levels and BRI1-mediated signaling. However, further analysis will be essential to clarify the exact molecular mechanisms.
We thank M. Szekeres, T. Hashimoto, T. Beeckman, G. Jürgens and N. Raikhel for providing materials; M-C. Baroso for technical assistance; G. Gudesblat, N. Irani, D. Van Damme, V. Vassileva, S. Vanneste, J. Kleine-Vehn, S. Naramoto and P. Marhavy for useful discussions and critical reading of the manuscript; and M. De Cock and A. Bleys for help in preparing the manuscript. This work was supported by the FP7 Marie-Curie Initial Training Network ‘BRAVISSIMO’ (PITN-GA-2008-215118; E.R. and A.C-D), the Belgian Science Policy (BELSPO) (M.K.Z.) and the Agency for Innovation by Science and Technology (IWT) (who awarded a postdoctoral fellowship to M.K.Z., a pre-doctoral fellowship to S.D. and a ‘Strategisch Basisonderzoek’ grant (no. 60839) to C.B.). A.C-D. receives funding through grants from the Spanish Ministry of Science and Innovation (BIO2008/00505).