Plant Mol Biol DOI 10.1007/s11103-014-0180-2
Transcriptional coordination between leaf cell differentiation and chloroplast development established by TCP20 and the subgroup Ib bHLH transcription factors Megan E. Andriankaja · Selahattin Danisman · Lorin F. Mignolet‑Spruyt · Hannes Claeys · Irina Kochanke · Mattias Vermeersch · Liesbeth De Milde · Stefanie De Bodt · Veronique Storme · Aleksandra Skirycz · Felix Maurer · Petra Bauer · Per Mühlenbock · Frank Van Breusegem · Gerco C. Angenent · Richard G. H. Immink · Dirk Inzé Received: 2 September 2013 / Accepted: 4 February 2014 © Springer Science+Business Media Dordrecht 2014
Abstract The establishment of the photosynthetic apparatus during chloroplast development creates a high demand for iron as a redox metal. However, iron in too high quantities becomes toxic to the plant, thus plants have evolved a complex network of iron uptake and regulation mechanisms. Here, we examined whether four of the subgroup Ib basic helix-loop-helix transcription factors (bHLH38, bHLH39, bHLH100, bHLH101), previously implicated in iron homeostasis in roots, also play a role in regulating iron metabolism in developing leaves. These transcription factor genes were strongly up-regulated during the transition from cell proliferation to expansion, and thus sink-source transition, in young developing leaves of Arabidopsis thaliana. The four subgroup Ib bHLH genes also showed reduced expression levels in developing leaves of plants treated with norflurazon, indicating their expression was tightly linked to the onset of photosynthetic Electronic supplementary material The online version of this article (doi:10.1007/s11103-014-0180-2) contains supplementary material, which is available to authorized users. M. E. Andriankaja · L. F. Mignolet‑Spruyt · H. Claeys · M. Vermeersch · L. De Milde · S. De Bodt · V. Storme · A. Skirycz · P. Mühlenbock · F. Van Breusegem · D. Inzé Department of Plant Systems Biology, VIB, Technologiepark 927, 9052 Ghent, Belgium M. E. Andriankaja · L. F. Mignolet‑Spruyt · H. Claeys · M. Vermeersch · L. De Milde · S. De Bodt · V. Storme · A. Skirycz · P. Mühlenbock · F. Van Breusegem · D. Inzé (*) Department of Plant Biotechnology and Bioinformatics, Ghent University, Technologiepark 927, 9052 Ghent, Belgium e-mail:
[email protected]‑ugent.be S. Danisman · G. C. Angenent Laboratory of Molecular Biology, Wageningen University, Droevendaalsesteeg 1, 6708 PB Wageningen, The Netherlands
activity in young leaves. In addition, we provide evidence for a mechanism whereby the transcriptional regulators SAC51 and TCP20 antagonistically regulate the expression of these four subgroup Ib bHLH genes. A loss-offunction mutant analysis also revealed that single mutants of bHLH38, bHLH39, bHLH100, and bHLH101 developed smaller rosettes than wild-type plants in soil. When grown in agar plates with reduced iron concentration, triple bhlh39 bhlh100 bhlh101 mutant plants were smaller than wildtype plants. However, measurements of the iron content in single and multiple subgroup Ib bHLH genes, as well as transcript profiling of iron response genes during early leaf development, do not support a role for bHLH38, bHLH39, bHLH100, and bHLH101 in iron homeostasis during early leaf development. Keywords Leaf development · Arabidopsis thaliana · Iron metabolism · Gene regulation · Basic helix-loop-helix transcription factors
S. Danisman · I. Kochanke Department of Molecular Cell Physiology, Bielefeld University, Universitätsstrasse 25, 33615 Bielefeld, Germany F. Maurer · P. Bauer Department of Biosciences, Plant Biology, Saarland University, Campus A2.4, 66123 Saarbrücken, Germany P. Mühlenbock Department of Plant Protection Biology, Swedish University of Agricultural Sciences, Box 52, 230 53 Alnarp, Sweden G. C. Angenent · R. G. H. Immink Bioscience, Plant Research International, Droevendaalsesteeg 1, 6708 PB Wageningen, The Netherlands
13
Introduction Leaf development can be divided into three phases: leaf initiation, primary morphogenesis, and secondary morphogenesis (Donnelly et al. 1999; Breuninger and Lenhard 2010). During primary morphogenesis the leaf grows largely through cell proliferation, while during secondary morphogenesis the leaf cells stop dividing and grow by post-mitotic expansion (Donnelly et al. 1999; for a review, see Gonzalez et al. 2012). Secondary morphogenesis is a highly dynamic process that does not take place in all leaf cells simultaneously (Donnelly et al. 1999; Efroni et al. 2008; Andriankaja et al. 2012), rather it begins in the tip of the leaf and progresses toward the leaf base. In addition to the onset of cell expansion, secondary morphogenesis also includes the initiation of chloroplast differentiation and the establishment of the photosynthetic machinery (Andriankaja et al. 2012). In fact, in leaves blocked in chloroplast differentiation, cell expansion was inhibited, suggesting that functioning chloroplasts are needed for proper onset of cell expansion (Andriankaja et al. 2012). It is still largely unknown how cell proliferation and cell expansion are coordinated with chloroplast development at the molecular level (Hou et al. 1993; Reiter et al. 1994; Osteryoung and Nunnari 2003; Raynaud et al. 2005; Kobayashi et al. 2009). Teosinte-Like, cycloidea and PCF1 (TCP) transcription factors are plant-specific transcription factors that are involved in controlling both proliferation and differentiation in leaves (Cubas et al. 1999; Palatnik et al. 2003; Crawford et al. 2004; Aguilar-Martínez et al. 2007; Nag et al. 2009; Martín-Trillo and Cubas 2010). Based on sequence homology and domain architecture, the TCP transcription factor family can be divided into two classes: class I and class II TCP genes. In Arabidopsis, class I TCPs are postulated to promote cell division (Li et al. 2005; reviewed in Martín-Trillo and Cubas 2010), and mutant analysis revealed a role for the class I TCP20 protein in the stimulation of cell proliferation (Danisman et al. 2012). Although the class I TCP20 protein was shown to induce expression of CYCB1;1 (Li et al. 2005), it may also promote cell differentiation and expansion by directly inducing genes involved in cell wall biogenesis (Hervé et al. 2009). Class II TCPs were shown to promote cell differentiation in leaves based on mutant analysis (Efroni et al. 2008). Overexpression of the microRNA miR319a in the JAGGED AND WAVY (jaw-D) mutant results in the knock-down of five class II TCPs, TCP2, 3, 4, 10, and 24. As a consequence, cells stop proliferating later than in wild-type leaves, leading to larger leaves, a higher number of smaller leaf cells, and eventually to heavily serrated and wavy leaf margins (Palatnik et al. 2003; Efroni et al. 2008).
13
Plant Mol Biol
Chloroplast development demands iron because of its crucial role in the photosynthetic electron transport chain (Nishio and Terry 1983; Larbi et al. 2006; Jeong and Guerinot 2009). Plants exposed to iron deficiency have symptoms of leaf chlorosis due to reduced photosynthetic efficiency, and thus less carbon fixation (Larbi et al. 2006). However, due to the fact that iron is a catalyst for the Fenton reaction, which produces destructive hydroxyl radicals, iron in excess can be toxic and can seriously alter plant growth and survival (Kampfenkel et al. 1995; Apel and Hirt 2004). This means that iron homeostasis is a tightly regulated process in plants involving a plethora of components. Two known players are the iron-regulated transporter 1 and ferric reductase oxidase2 (IRT1 and FRO2), which constitute the major iron uptake system in the roots of dicotyledonous plants (Eide et al. 1996; Robinson et al. 1999; Henriques et al. 2002; Varotto et al. 2002; Vert et al. 2002), and their regulator fer-like iron deficiency-induced transcription factor (FIT) (Colangelo and Guerinot 2004; Jakoby et al. 2004; Yuan et al. 2005; Bauer et al. 2007). FIT was found to be induced during iron deficiency and essential for iron transport (Colangelo and Guerinot 2004; Jakoby et al. 2004; Yuan et al. 2005). Another group of transcription factors found to be transcriptionally induced by iron deficiency consists of the subgroup Ib (2) of basic helix-loop-helix transcription factors (Pires and Dolan 2010), more specifically bHLH38, bHLH39, bHLH100, and bHLH101 (Wang et al. 2007). bHLH38, bHLH39, bHLH100, and bHLH101 were found to interact with FIT in regulating expression of IRT1 and FRO2 (Yuan et al. 2008; Wang et al. 2013). Another important regulator in maintaining iron homeostasis is ferritin. Ferritin is a large iron trap, which stores free labile iron in the cell, and thus prevents it from inducing ROS formation via the Fenton reaction. This was confirmed by ferritin mutants (fer1fer3fer4), which grew the same as wild-type plants in standard conditions, but were less able to maintain growth in iron accumulating conditions due to iron-induced ROS production (Ravet et al. 2009). To date, iron homeostasis has mainly been investigated with the aim of understanding iron uptake and iron toxicity, particularly in roots (Palmer and Guerinot 2009). Thus, the genes involved in iron homeostasis in leaves, and how these affect photosynthesis and leaf growth, are less well characterized and were explored further in this study. Microarray data from our previous study (Andriankaja et al. 2012) showed that the iron deficiency-responsive bHLH transcription factor genes, bHLH38, bHLH39, bHLH100, and bHLH101, were transcriptionally induced at the transition from primary to secondary leaf morphogenesis, making them good candidates for being markers for the onset of the differentiation of photosynthetically active cells (sink-source transition) and/or the transition from cell proliferation to expansion. These four bHLH
Plant Mol Biol
genes were also suggested to be repressed upon inducible overexpression of the class I TCP gene, TCP20 (Danisman et al. 2012). In this study we provide further evidence for this repression and binding of TCP20 to the bHLH39 promoter by chromatin immunoprecipitation (ChIP) and transactivation experiments. Furthermore, we identified the bHLH transcription factor SAC51 as a potential activator of bHLH39 that may antagonize TCP20 regulation of the target. Growth phenotyping of single bhlh38, bhlh39, bhlh100, and bhlh101 mutants revealed that bHLH38, bHLH39, bHLH100, and bHLH101 strongly repressed rosette growth of plants grown in normal long-day conditions in soil. When grown in agar plates with reduced iron concentration, triple bhlh39 bhlh100 bhlh101 mutant plants were smaller than wild-type plants. However, measurements of the iron content in single and multiple subgroup Ib bHLH genes, as well as transcript profiling of iron response genes during early leaf development, do not further support a role for bHLH38, bHLH39, bHLH100, and bHLH101 in iron homeostasis during early leaf development.
were sown at a density of 20 seeds per Greiner Bio One (90 mm Ø x 20 mm height) petri dishes. They were grown on full-strength MS medium (Murashige and Skoog 1962) (Duchefa Biochemistry, Haarlem, Netherlands) with 1.5 % sucrose. Prior to sowing, the seeds were imbibed with water in 2 mL tubes and stratified at 4 °C for 4 days. After sowing, the plants were grown in long day conditions (growth rooms kept at 22 °C with 16-h day/8-h night cycles, temperature inside the petri dish was 20.8 °C and light intensity was 80 μM) for 14 days. Subsequently, petri dishes were transferred to continuous high light conditions (growth rooms kept at 22 °C with 24-h day/0-h night cycles, temperature inside the petri dish of 27.7 °C and a light intensity of 300– 400 μM), continuous light conditions (growth rooms kept at 22 °C with 24-h day/0-h night cycles, temperature inside the petri dish of 21.5 °C and a light intensity of 100 μM) or stayed at long day conditions for another 7 days. Daily randomization decreased the chance of unwanted positional bias. The experiments were conducted in triplicate.
Materials and methods
Iron deficiency and sufficiency growth conditions
Plant material
Iron deficiency and sufficiency conditions were realized by changing the amount of iron that was supplied to the full-strenght MS agar growth medium. The standard (i.e., control) MS agar growth condition of 100 μM Ethylenediaminetetraacetic Acid Ferric Sodium (FeNa EDTA) is generated by adding 4.3 g/L of unaltered full-strength MS salt mix (Duchefa Biochemistry, Haarlem, The Netherlands). Full-strength MS medium with a concentration of 50 or 300 μM iron was obtained by adding respectively 18.35 or 110.10 mg/L FeNa EDTA (C10H12N2O8FeNa; molecular mass = 367.0), together with 4.3 g/L of specially ordered MS salt mix without iron (Duchefa Biochemistry, Haarlem, Netherlands). In addition, a detailed analysis on the concentrations and activities of the soluble aqueous inorganic species derived from FeNa EDTA was performed by using the chemical equilibrium modelling software Visual MINTEQ 3.0 (Gustafsson J and 2010) (Table S1). In case of the short-term iron deficiency experiment, wild-type Arabidopsis was sown out on medium containing 50 μM iron with agar (6 g/L), using sterilized nylon meshes (mesh pore size: 200 μm) to ease transfer to experimental media (Passarinho et al. 2008; Hong et al. 2013; Salomé et al. 2013). Plants were grown in a 16-h day/8-h night regime. Iron depletion was conducted 8 days after sowing out. Plants were transferred to liquid medium either containing 50 or 0 μM iron. They were harvested 6 and 24 h after and immediately before treatment. The experiment was conducted in triplicate.
The TCP20-GR and TCP20-GFP lines were described in Danisman et al. (2012). The bhlh38-1, bhlh39-1, bhlh1001, bhlh101-1 mutant lines as well as the pbHLH100::GUS line were described in Wang et al. (2007), and the triple bhlh39 bhlh100 bhlh101 mutant was generated by crossing a bhlh39-1 homozygous plant with a bhlh100-1 bhlh101-1 homozygous plant and selecting for the triple mutants by PCR genotyping in the resulting F2. It was then further multiplied by selfing. Plant growth conditions in soil Arabidopsis thaliana (L.) Heyhn. ecotype Col-0 seeds, as well as the bhlh mutants were sown on soil in 6 cm wide square pots with a density of four seeds per pot. Prior to sowing, the seeds were imbibed with water in 1.5 mL tubes and stratified at 4 °C for 2 days. After sowing, the plants were moved to standard conditions (growth rooms kept at 22 °C with 16-h day/8-h night cycles and light intensity of 110 μM). After 7 days in the growth room, the four seedlings were screened, removing all but one seedling per pot, which most closely resembled the genotype average. Plant growth conditions on agar plates Arabidopsis thaliana (L.) Heyhn. ecotype Columbia-0 (Col-0) seeds, as well as those of the bHLH mutants
13
Iron measurement Arabidopsis wild-type and mutant plants were grown under long-day conditions for 4 weeks and then whole rosettes were harvested for iron measurements. Plant iron measurements were conducted as previously described (Li et al. 2011). Iron concentration was determined using 0.05 % 1,10-phenanthroline and measuring the absorbance at 510 nm against an iron standard curve that was established with defined amounts of FeSO4·7H2O. The analysis was conducted in triplicate, per sample ten rosettes were pooled.
Plant Mol Biol
plate. Plants were grown in a 16-h day/8-h night regime. Glucocorticoid induction experiments were conducted 14 days after germination of the plants. Growing the plants on nylon meshes allowed us to transfer the plants onto induction media swiftly and without severely damaging the roots. The induction medium consisted of 2.3 g/L MS, 1 % (w/v) sugar, 10 μM dexamethasone, and 10 μM cycloheximide (Passarinho et al. 2008). Control plants were transferred to the same media except without the dexamethasone. Whole rosettes were harvested at different time points after start of the treatment. RNA isolation and qRT‑PCR
Rosette growth phenotyping Either projected rosette growth or total rosette area was determined for each mutant in the growth conditions mentioned above. For projected rosette measurements, plants were imaged from above in a fixed imaging platform (MIRGIS), which images plants at the same time every day via fixed cameras located directly above the plants. These images were then converted to binary images in ImageJ, which converted the green area in each pot to black. This black area was then measured by the measure particles function in ImageJ using a threshold particle size of 50 pixels. Total rosette area was analyzed by doing leaf series as previously described by Gonzalez et al. (2010). To determine whether cell size or cell number was affected, the third leaf of the bhlh39 mutant was harvested 22 DAS and cells were drawn and analyzed also as previously described in Gonzalez et al. (2010).
RNA was isolated using the QIAGEN RNeasy RNA isolation kit according to the manufacturer’s protocol. DNase treatment took place on-column, following the protocols from the manufacturer. M-MuLV reverse transcriptase (RT) from Promega was used for cDNA synthesis. First, a mix of poly-dT primer and dNTPs was added to 500 ng DNA-free RNA in a volume of 12 μL. This solution was kept in ambient temperature for 2 min before 1 μL RT was added. After addition of the RT, samples were incubated at 25 °C for 15 min and transferred to 42 °C for 50 min. Here, reverse transcription took place and was stopped by heat treatment at 70 °C for 15 min. The cDNA made this way was used for quantitative Realtime PCR (qRT-PCR) using the SYBR green mix from BioRad. The reference genes used for all analyses were the SAND family gene AT2G28390 and the TIP41-like gene AT4G34270, both determined superior reference genes (Czechowski et al. 2005).
Norflurazon treatment
Protoplast assay
Plants were grown on half strength MS on meshes with 20 μm diameter pore sizes. On 8 DAS (days after stratification) the meshes were transferred to either fresh MS plates or MS containing 5 μM norflurazon (Koussevitzky et al. 2007). Plants were then harvested at day 10 (48 h after transfer). Third leaves from the plants were microdissected as described in Skirycz et al. (2011). RNA was extracted using the Qiagen RNeasy plant mini kit using optional oncolumn DNase digestion, and RNA quantification was then performed using qRT-PCR.
Tobacco BY-2 protoplasts were prepared from cell cultures as published in De Sutter et al. (2005). Protoplasts were co-transfected in 48-well multiwell plates with 2 μg each of reporter, effector, and normalization plasmids using the automated protoplast transfection system established by De Sutter et al. (2005). Lastly, renilla and firefly luciferase activity was measured using the Dual Luciferase Reporter Assay 1,000 (Promega) as outlined in the users manual. The dual luciferase activity was measured on a LUMIstar Galaxy luminometer with double injectors (BMG Labtechnologies, Offenburg, Germany) as outlined in De Sutter et al. (2005). Eight transfections were made for each combination per run, and 3 independent runs were performed (n = 24).
GR induction assays for the TCP20‑GR line For dexamethasone induction experiments of the TCP20GR line, 50 mL of ½ MS medium (2.3 g/L) with Agar (6 g/L) was poured per plate and, after polymerization, a sterilized nylon mesh (mesh pore size: 200 μm) was placed on the medium (Passarinho et al. 2008). 30–50 seeds of both the GR-line as well as wild type Col-0 were sown per
13
Chromatin immunoprecipitation Chromatin immunoprecipitation (ChIP) experiments mainly followed protocols described previously (de Folter
Plant Mol Biol
et al. 2007; Kaufmann et al. 2010). TCP20-GFP seedlings were grown in liquid ½ MS medium on a horizontal shaker (30 rpm) for 4 days in a 16-h day/8-h night regime. After this time, seedlings were harvested by pouring the liquid MS medium through a sieve and then 3 g of plant material was fixated with 37 % formaldehyde. Immunoprecipitation was conducted using a GFP antibody coupled to magnetic beads. The magnetic beads were used to precipitate the antibody-protein-GFP complexes. The enrichment of TCP20 binding regions was compared between the immunoprecipitate and 1:1,000 diluted input material. Realtime PCR analysis was then used to check for enrichment of targets. Promoter elements, pSAND, not expected to be bound by TCP20 were used as negative control. The pSAND gene is often used as a qPCR housekeeping gene due to its stable expression profile, and its promoter does not contain any potential TCP binding-sites, making it a good negative control for testing TCP binding specificity. The promoter elements tested for the subgroup Ib bHLH genes were as follows: bHLH038 pro 1: −388 to −295, bHLH038 pro 2: −221 to −164, bHLH039: −334 to −264, bHLH100: −1165 to −943, bHLH101 pro 1: −2,280 to −2,135, bHLH101 pro 2: −2,040 to −1,890. The numbers are nucleotides upstream of the annotated transcriptional start sites for the respective genes. Yeast one‑hybrid assays All previously identified TCP20 heterodimers (Danisman et al. 2012) have been screened against a BHLH39 promoter fragment in a yeast one-hybrid assay based on the Matchmaker Gold Yeast Two-Hybrid System (http:// www.clontech.com). The TCP dimers were expressed in yeast strain PJ69-4A and the pbHLH39 reporter construct (CZN2069) was transformed into PJ69-4α. For the reporter plasmid a 762 bp bHLH39 promoter fragment was recombined into the Gateway compatible pAbAi vector CZN1018 (Danisman et al. 2012). Subsequently background expression of the reporter gene was analyzed in an autoactivation test and growth of yeast up-to 100 ng/mL aurobasidin was observed. Therefore, the final analysis for binding of TCP20 dimers to the bHLH39 promoter was performed at the concentrations 150, 175 and 200 ng/mL aurobasidin. The mating-based yeast one-hybrid assay was performed as described previously (Danisman et al. 2012). Growth, and hence binding of the proteins to DNA, was scored after 5 days incubation of yeast on the selective media at 20 °C. As negative control, two TCP20 protein combinations have been tested for which no dimerization was detected in the previous yeast two-hybrid assay (marked with an asterisk in Fig. 4c). For the identification of other transcription factors able to bind the pbHL39 promoter fragment the REGIA transcription factor collection (Castrillo et al. 2011) was
screened using the yeast-one-hybrid assay described above. All putative positives found in the first screening at 150 ng/mL aurobasidin were re-screened at 150, 175 and 200 ng/mL aurobasidin to confirm binding. PAM fluorometry By using the PAM chlorophyll fluorometer (Pulse-Amplitude-Modulation; Heinz Walz GmbH, Effeltrich/Duitsland), chlorophyll fluorescence parameters (i.e., F0 and Fm) were measured, which were used to obtain the ratio between the variable fluorescence and the maximal fluorescence (Fv/Fm). Fv′/Fm′ (ф PSII) represents the maximum (quantum) efficiency of photosystem II (PSII) for non-darkadapted plants (Baker 2008). This was measured for individual leaves throughout rosette development for in soil grown plants or for plants grown on agar plates by encircling the center of a plant’s rosette via the area of interest (AOI) option. For the statistical analysis, a three way ANOVA model was fit to the data with all higher order terms included in the model. The petri-dishes on which the plants were grown were added as a random effect, nested under the three factors to account for dependencies between plants grown on the same tray. The analysis was done with the mixed procedure from SAS (SAS/STAT analytical product 12.1, SAS Institute Inc., 2012, Cary, NC). Robust standard errors were calculated. User-defined contrasts were calculated and adjusted for multiple testing with FDR (Benjamini and Hochberg 1995) using the multtest procedure. FDR adjusted p values
