|Home | About | Journals | Submit | Contact Us | Français|
Strigolactones are newly identified plant hormones, shown to participate in the regulation of lateral shoot branching and root development. However, little is known about their effects on biological processes, genes, and proteins. Transcription profiling of roots treated with GR24, a synthetic strigolactone with proven biological activity, and/or indole acetic acid (IAA) was combined with physiological and transcriptional analysis of a tomato mutant (Sl-ORT1) deficient in strigolactone production. GR24 treatment led to markedly induced expression of genes putatively involved in light harvesting. This was apparent in both the presence and absence of exogenously applied IAA, but not with IAA treatment alone. Following validation of the microarray results, transcriptional induction by light of the GR24-induced genes was demonstrated in leaves exposed to high or low light intensities. Sl-ORT1 contained less chlorophyll and showed reduced expression of light harvesting-associated genes than the wild type (WT). Moreover, perfusion of GR24 into WT and Sl-ORT1 leaves led to induction of most of the examined light harvesting-associated genes. Results suggest that GR24 treatment interferes with the root's response to IAA treatment and that strigolactones are potentially positive regulators of light harvesting in plants.
Strigolactones have been recently defined as a new group of plant hormones or their biosynthetic precursors (Gomez-Roldan et al., 2008; Umehara et al., 2008). Strigolactone production has been demonstrated in many plant species (e.g. Sato et al., 2005; Gomez-Roldan et al., 2008; Umehara et al., 2008; Koltai et al., 2010b), synthesized mainly in the roots and lower part of the stem (Foo et al., 2001; reviewed by Dun et al., 2009) and then moving towards the shoot apex (Foo et al., 2001; Brewer et al. 2009; Ferguson and Beveridge 2009). Strigolactones have been shown to play a role in inhibition of shoot branching (Gomez-Roldan et al., 2008; Umehara et al., 2008; Brewer et al., 2009; Ferguson and Beveridge, 2009) and thus to affect shoot architecture; they have also been shown to affect root growth (Koltai et al., 2010a) and root system architecture (HK and YK, unpublished results). Strigolactones have been suggested to be derived from the carotenoid pathway (Matusova et al., 2005) via the activity of various oxygenases (Gomez-Roldan et al., 2008; Umehara et al., 2008, and references therein).
Some studies have suggested key hypotheses regarding the role strigolactones may play in restraining lateral shoot bud outgrowth, thereby determining shoot architecture. It was suggested that strigolactones inhibit polar auxin transport from the buds by reducing the capacity for polar auxin transport from the apical meristem, resulting in restrained bud outgrowth (e.g. Bennett et al., 2006; Mouchel and Leyser, 2007; Ongaro and Leyser, 2008; Leyser, 2009). Another possibility is that strigolactones serve as auxin-promoted secondary messengers that move up into the buds to repress their outgrowth (Brewer et al., 2009; Ferguson and Beveridge, 2009; reviewed by Dun et al., 2009), or that both auxin and strigolactones have the ability to change each other's levels and distribution in a dynamic feedback loop, which is required for the coordinated control of axillary branching (Hayward et al., 2009).
Using a synthetic strigolactone (GR24) previously shown to have strigolactone-like biological activity (Gomez-Roldan et al., 2008; Umehara et al., 2008), an effect of strigolactones on tomato root development was recently found. GR24 was shown to interfere with the inhibitory effect of exogenously applied auxin on root elongation, conveyed via an increase in root cell length. Auxin efflux carriers were involved in this effect of strigolactone on root growth and root hair elongation (Koltai et al., 2010a).
However, only little is known about the biological processes, genes, and proteins affected by strigolactones. For a better understanding of the effects of strigolactones on plant development and their cross-talk with auxin, gene expression was profiled following tomato root exposure to GR24 and indole acetic acid (IAA). Of the GR24-induced genes, many were putatively involved in light harvesting. Following validation of the microarray results, light-induced transcription of the GR24-induced genes was demonstrated in leaves exposed to high or low light intensities. Expression of these genes was reduced in a strigolactone-deficient tomato mutant (Sl-ORT1; Koltai et al., 2010b), which was also found to possess reduced levels of chlorophyll. Moreover, exposure of both wild-type (WT) and Sl-ORT1 leaves to GR24 led to induction of the expression of these genes. Together, the results suggest that strigolactones are potentially positive regulators of plant light-harvesting components.
Tomato (Solanum lycopersicum) cv. M82 (WT; Eshed et al., 1992) seedlings were surface-sterilized, immersed in sterile distilled water for 1h, and placed on half-strength Murashige and Skoog (MS) agar medium supplemented with 1.5% (w/v) sucrose, in Petri dishes (50 seeds per Petri dish). Dishes were placed horizontally in the dark at 25°C for 72h to induce germination. Germinated seeds were gently transferred to half-strength MS agar medium supplemented with 1.5% (w/v) sucrose in Petri dishes. Petri dishes included the following hormone concentrations (designated as treatments): IAA at a concentration of 10−8M; IAA at a concentration of 10−8M supplemented with the synthetic strigolactone GR24 (Johnson et al., 1981) at a concentration of 13.5μM; GR24 at a concentration of 27μM; and non-supplemented control. Hormone concentrations were determined, in a previous study, to have significant effects on root growth under the same growth conditions (Koltai et al., 2010a). The plates were placed vertically to allow gravitropic root growth along the surface of the agar under 16/8h (day/night) fluorescent lighting (100μmol m−2 s−1) at 25°C. Roots were marked after 24h of incubation; following an additional 24h, roots were sectioned from the 24h mark to the root tip and were frozen in liquid nitrogen.
WT and Sl-ORT1 seeds were surface-sterilized and allowed to germinate and grow in styrofoam seedling trays in soil:vermiculite (1:1, v/v). Four-week-old seedlings were transferred to 3.0l pots (one plant per pot) with a 1:1 mixture of soil and vermiculite. For each WT or Sl-ORT1 plant, 12 pots with plants were grown under full light intensity (August–September 2009) and another 12 pots under shaded light, the latter provided with black 50% shading nets. Plants were grown in the greenhouse under natural light conditions supplemented with artificial light (100μmol m−2 s−1) to maintain a 16/8h (day/night) photoperiod at 28°C/24°C (day/night). Plants were fertilized with 7:1:7 (NPK) solutions. The dry weight (DW) of shoots was determined for each plant after 45d; 12 plants were examined for each WT and Sl-ORT1 strain, for each shading treatment. DW was determined following 72 h incubation in an air forced oven at 70°C. The experiment was repeated twice. Means ±SE were determined for all replicates; means of replicates were subjected to statistical analysis by multiple-range test (P ≤0.05), using the JMP statistical package (SAS, Cary, NC, USA).
The first fully expanded leaves of M82 or Sl-ORT plants grown under greenhouse conditions as described above were injected with 1ml of either GR24 (2.7μM) or water as a control. Five spots on the leaf were injected with 1ml of fluid into the lower epidermis, such that the injected fluid spread throughout the leaf tissue; injection sites were marked. Marked areas were collected 48h after injection, and were frozen in liquid nitrogen.
RNA extraction was performed, for each biological replicate, from 150mg of roots or leaves of plants grown under in vitro and greenhouse conditions, respectively, as described above. Total RNA was extracted using TRI reagent (MRC, Cincinnati, OH, USA) and treated with Turbo DNase enzyme (Ambion) as per the manufacturer's instructions. For the microarray experiments, no DNase treatment was performed. Rather, RNA was purified and concentrated using an RNeasy MiniElute Cleanup Kit (Qiagen, Valencia, CA, USA) as per the manufacturer's instructions, and was taken for hybridization.
For cDNA synthesis, 2.5μg of total RNA and 0.1μM of random hexamer primers were heated for 5min at 65°C and snap-chilled on ice. The following components were added to the reaction mixture: 0.2mM dNTP mixture (Fermentas, Glen Burnie, MD, USA), M-MuLV-reverse transcriptase (RT) buffer (1× final concentration), 40U of RNase M-MuLV inhibitor (Fermentas), 200U of M-MuLV-RT enzyme (Fermentas), and diethylpyrocarbonate (DEPC)-treated water to a reaction volume of 21μl. The reaction was incubated at 42°C for 60min following an incubation at 70°C for 10min.
The Affymetrics GeneChip® Tomato Genome Array that was used consists of 10227 S. lycopersicum probe sets for examination of ~9200 S. lycopersicum transcripts. Sequence information for this array was selected from public data sources including S. lycopersicum UniGene Build #20 and GenBank mRNAs (http://www.affymetrix.com/products/arrays/specific/tomato.affx); the total number of genes in the tomato euchromatin is ~40000 (Mueller et al., 2009).
RNA, purified as described above, was subjected to Affymetrics microarray hybridization, as instructed by the manufacturer using a GeneChip® 3′ IVT Express Kit (Affymetrix).
GeneChips were scanned using the Affymetrix GeneChip Scanner 3000 7G. Data were quantified with the Affymetrix Expression Console using the MAS 5.0 algorithm with Affymetrix default analysis settings and global scaling as the normalization method. Filtration on confidence was performed based on one-sample t-test (P ≤0.05), and the resulting gene lists were filtered using GeneSpring GX (Agilent) for differentially expressed genes (>2-fold and <0.5-fold change between treatment versus control). Three biological replicates were performed for each examined treatment (i.e. IAA, IAA+13.5μM GR24, and 27μM GR24).
Annotation of part of the array-represented gene sequences is based on Gene Ontology nomenclature (GO; http://www.geneontology.org/) and was supplied by Affymetrix Inc. Additional gene annotations were found by BLAST function for comparisons between array-represented gene sequences and the NCBI non-redundant protein and TAIR databases. Functional classifications were performed by MapMan software (Thimm et al., 2004).
Quantitative real-time (qPCR) was performed on RNA extracted from either roots or leaves, as described above. For microarray validation, three additional (to those of the microarray) and independent experiments were examined. The qPCR was performed using components supplied in the KAPA SYBR FAST qPCR kits (Kapa Biosystems, Woburn, MA, USA) and gene-specific primers (Table 1). The reaction mixture consisted of the following components: 2× Master Mix with integrated antibody-mediated hot start, SYBR® Green I fluorescent dye, MgCl2, dNTPs, stabilizers, 2μl of the template, and PCR-grade water to a final volume of 10μl. The qPCR analysis was carried out on a Rotor gene 6000 instrument (Corbett-Qiagen, Valencia, CA, USA) according to the following program: 3min at 95°C, followed by 49 cycles of 95°C for 3s, 60°C for 20s, and 72°C for 1s. Primers used for qPCR are presented in Table 1. The threshold cycle (Ct) was calculated by the Rotor gene 6000 instrument software. The level of expression of the target genes was calculated relative to that of the reference mRNA; tomato 18S rRNA (accession no. AY552528) served as the reference gene for the amount of RNA, and was amplified using the forward primer 5′TTGATTACGTCCCTGCCCTTTGTACAC3′ and the reverse primer 5′AGGTTCACACCTACGGAAACCTTGTTAC3′. Values of the steady-state level of gene transcripts were determined as a ratio between two conditions (e.g. GR24 treatment versus control, or Sl-ORT1 versus WT) using the 2Δ-ΔCt method (Arocho et al., 2006). A value above or below 1 represents an increase or decrease, respectively, in the steady-state level of gene transcripts for the examined conditions (i.e. that of the nominator versus that of the denominator). Means ±SE were calculated for three biological replicates for each examined treatment. Means of replicates were subjected to statistical analysis by multiple-range test (P ≤0.05), using the JMP statistical package.
Chlorophyll was measured using a hand-held chlorophyll meter (SPAD-502 chlorophyll meter, Minolta Camera Co., Ltd, Osaka, Japan), according to the manufacturer's instructions. The measurements were performed on eight mature (fully exposed) leaves from each of four (45d old) plants per variant in the morning (10:00 to 11:00h) under natural sunlight (1200μmol m−2 s−1, ~12h of daylight) in the greenhouse.
Roots grown on plates supplemented with different hormones: IAA (10−8M), IAA (10−8M)+GR24 (13.5μM), and GR24 (27 μM), and non-treated controls were profiled for gene expression using tomato Affymetrix microarrays (Supplementary Table S1 available at JXB online). Hierarchical clustering of the expression results suggested that the two treatments with GR24, namely IAA+GR24 and GR24, were more similar to each other in terms of gene expression than the IAA+GR24 and IAA treatments (Supplementary Fig. S1).
Statistical and 2-fold cut-off were used to identify genes that were significantly and differentially expressed for each of the treatments (Supplementary Table S2 at JXB online): most differentially expressed genes were found in the GR24 treatment, less in the IAA+GR24 treatment, and even less in the IAA treatment. Moreover, intersection by a Venn diagram of each of the treatments versus non-treated controls suggested that more genes were shared between the GR24 treatments (IAA+GR24 and GR24) than between the IAA treatments (IAA+GR24 and IAA), or between IAA and the GR24 treatments (Fig. 1).
Expression of eight significantly and differentially regulated genes was validated by qPCR. All eight genes showed expression patterns with the same tendency as those detected in the microarray experiments, confirming the accuracy of the microarray data (Table 2).
These results suggested that under the examined conditions, treatment of roots with GR24 interferes with the effects on roots of simultaneous treatment with IAA.
The significantly and differentially expressed genes (Supplementary Table S2 at JXB online) in the three root treatments were divided according to their putative association with biological processes. Significantly, in the GR24 treatments (i.e. IAA+GR24 and GR24), many of the up-regulated genes were involved in light harvesting (Supplementary Fig. S2). Notable among these were genes of photosystem I (PSI) and PSII (Fig. 2). However, in the IAA-only treatment, the light harvesting-associated genes were not induced (Supplementary Fig. S2, Fig. 2).
To examine the association between the GR24-induced, putative light harvesting-associated genes and plant light reactions, the light regulation of the transcription of these genes was examined in leaves of WT tomato plants grown under full and reduced light intensities.
Under 50% shading, total primary shoot DW was reduced by ~66% (8.9±0.4g for full light intensity versus 3.0±0.5g for shaded plants) and chlorophyll levels were reduced by ~20% (47.9±0.7 SPAD units for full light intensity versus 39.2±0.5 SPAD units for shaded plants).
Under these shaded, reduced-chlorophyll conditions, transcription levels of the GR24-induced putative light harvesting-associated genes were reduced in comparison with their transcription levels under full-light conditions (Table 2). Hence, these genes are regulated transcriptionally and are induced by high-light conditions, these latter conditions being associated with increased content of chlorophyll.
The above results suggested that exposure to strigolactones induces light harvesting-associated genes. To validate this notion, the levels of chlorophyll were examined in a mutant, Sl-ORT1, which is deficient in strigolactone production (Dor et al., 2010; Koltai et al., 2010b) in comparison with WT plants.
Under full-light conditions, Sl-ORT1 had a significantly reduced level of chlorophyll relative to the WT (41.5±0.8 SPAD units and 47.9±0.7 SPAD units, respectively).
The level of transcription of the GR24-induced light-responsive genes was determined in WT and Sl-ORT1 leaves grown under full-light conditions. All examined light harvesting-associated genes showed reduced levels of transcription in Sl-ORT1 in comparison with the WT (Table 2).
Together, these and the above results of the chlorophyll level in Sl-ORT1 leaves suggest that lack of strigolactones is associated with a reduced level of light harvesting in plants, reflected as both reduced transcription of light harvesting-associated genes and a decrease in chlorophyll level.
To demonstrate a direct connection between plant exposure to strigolactones (in the form of GR24) and transcription of the light harvesting-associated genes, light harvesting-associated gene transcription was profiled in Sl-ORT1 and WT leaves 48h after injecting leaves with GR24. All but one of the examined light-harvesting genes appeared to be induced upon GR24 injection, in both Sl-ORT1 and WT leaves (Table 2).
In this study, the understanding of plant responses to strigolactones is promoted by a demonstration that strigolactones have a positive effect on the plant's light harvesting.
Gene expression profiling suggested that GR24 treatments are similar in their effect on gene expression in treated roots in the presence and absence of IAA. Gene expression profiling allowed the identification of GR24-induced genes, activated upon GR24 treatment regardless of other conditions (i.e. treatment with IAA or different GR24 concentrations); these genes, many of which are putatively associated with light harvesting, are discussed further on. Gene expression profiling also suggested that treatment of roots with GR24 interferes with the effect of IAA treatment on roots.
Accordingly, unlike in the GR24 treatments, in the IAA-only treatment the light harvesting-associated genes were not induced. This lack of IAA induction of light harvesting-associated genes is in agreement with the results of Volfová et al. (1978), suggesting that IAA treatment reduces chlorophyll content in wheat. This lack of IAA induction of light harvesting-associated genes is also in agreement with the results of Zhong and Ye (2001), suggesting that inhibition of auxin transport in Arabidopsis, leading to reduced auxin polar flow along the inflorescence stems and in the hypocotyls, is associated with increased chlorophyll content.
This interference of strigolactones with auxin responses detected in the present system is reflected at the gene expression level; however, it may also be reflected at the root phenotypic level: exposure of roots to GR24 interfered with the inhibitory effect of exogenously applied auxin on root cell elongation in tomato plants (Koltai et al., 2010a). Although the present results should be further validated, they support the suggestion that strigolactones may interfere with the capacity for polar auxin transport from the apical meristem (e.g. Bennett et al., 2006; Mouchel and Leyser, 2007; Ongaro and Leyser, 2008; Leyser, 2009): exogenous application of strigolactones may interfere with polar auxin transport thereby leading to increased chlorophyll content (see also Zhong and Ye, 2001).
Several lines of evidence in the present study suggest that strigolactones are positive regulators of plant light harvesting. The first is the list of genes induced by GR24, which is enriched in genes putatively associated with light harvesting. These include components of PSI and PSII, which are multisubunit membrane–protein complexes that constitute, as the principal converter of sunlight into chemical energy (reviewed by Nelson and Yocum, 2006), precursors of chlorophyll a/b-binding proteins, which serve as light-harvesting antennae for the capture of light energy and its transfer to the photosynthetic reaction centers (reviewed by Koziol et al., 2007), and ribulose-1,5-bisphosphate carboxylase (Rubisco), the primary CO2-fixing enzyme (reviewed by Portis and Parry, 2007).
Light regulation at the transcription level of these GR24-induced genes was demonstrated in leaves of plants exposed to different light regimes, and is in agreement with multiple studies showing light induction at the transcription level of light harvesting-associated genes (e.g. Guo et al., 2008). Moreover, transcription of some of these genes was induced upon direct GR24 application to leaves. Hence, it is suggested that plant exposure to strigolactones activates light harvesting-related genes.
The second piece of evidence for the association between strigolactones and light harvesting came from analysis of Sl-ORT1, a tomato mutant deficient in strigolactone biosynthesis (Dor et al., 2010; Koltai et al., 2010b). Notably, in Sl-ORT1, all examined GR24- and light-induced genes were transcriptionally down-regulated in comparison with the WT. Accordingly, in Sl-ORT1, a reduced level of chlorophyll was detected in leaves relative to the WT. Together, these results further support the hypothesis that strigolactones are inducers of light harvesting; this induction is associated, at least partially, with effects on chlorophyll levels.
Two other studies have indicated a connection between strigolactones and light reactions. One (Mashiguchi et al., 2009) found that light signalling-related genes are induced in Arabidopsis seedlings upon exposure to GR24. The second (Shen et al., 2007) examined light responses of max2 mutant seedlings (pps). max2 plants are mutated in an F-box protein suggested to be a strigolactone receptor and display an increased level of shoot branching (Stirnberg et al., 2002; Bennett et al., 2006; Umehara et al., 2008). max2 seedlings were shown to have longer hypocotyls and slightly smaller cotyledons under continuous red, far-red, and blue light compared with those of the WT (Shen et al., 2007). Moreover, several genes, including Rubisco small subunit and chlorophyll a/b-binding protein precursors, were found to exhibit a slower rate of induction upon red light exposure in max2 mutants relative to the WT (Shen et al., 2007). However, it was not determined whether MAX2 was involved in light signalling separately or as part of the shoot branching-regulatory pathway.
In the present study, Rubisco and chlorophyll a/b-binding protein precursor were among the GR24-induced genes whose expression was reduced in Sl-ORT1 in comparison with the WT. These results, and the reduced chlorophyll level detected in Sl-ORT1, suggest that mutants flawed in strigolactone perception or synthesis (i.e. max2 and Sl-ORT1, respectively) are also flawed in their light harvesting, in turn suggesting a direct connection between strigolactones and light harvesting, and co-regulation of strigolactones and light on light-harvesting components.
Other plant hormones are also known to be connected with light-harvesting pathways. Brassinosteroids were demonstrated to increase chlorophyll levels in plants (e.g. Krizek and Mandava, 1983); cytokinin co-regulates, along with light, many plant processes (reviewed by Werner and Schmulling, 2009), whereas specifically regarding light harvesting it delays chlorophyll destruction (Riefler et al., 2006; Sergiev et al., 2007). Cytokinin was also shown to affect chloroplast organization during photosynthetic acclimation to canopy density (Boonman et al., 2009). Hence, possible interaction of strigolactones with other hormones such as cytokinin (Ferguson and Beveridge, 2009) or brassinosteroids may be a mediator of the detected connection between strigolactones and plant light harvesting.
To conclude, there are several lines of evidence suggesting a role for strigolactones as positive regulators of light harvesting. Further studies are needed to determine the junction points of the co-regulation of strigolactones and light on light-harvesting components. Moreover, since carotenoid biosynthesis has been shown to be light dependent (e.g. Cazzonelli et al., 2009), and since strigolactones are thought to be derived from this pathway (Matusova et al., 2005), it might be that the cross-talk between strigolactones and light-associated pathways follows a feedback loop which is required for the plant's coordinated growth and development.
Supplementary data are available at JXB online.
Table S1. Raw microarray data of signal intensity for each of the probes on the microarray. Affymetrix probe set ID and signal intensity for each of the probes in each treatment [GR24 (27μM), IAA (10−8M)+GR24 (13.5μM), and IAA (10−8M)], for each of the biological replicates (a–c) is shown.
Table S2. Lists of genes that are significantly and differentially expressed. Genes were selected based on their expression ratio in roots exposed to GR24 and IAA treatments [IAA (10−8M), IAA (10−8M)+GR24 (13.5μM), and GR24 (27μM)] versus non-treated controls. Annotation is based on Gene Ontology (GO) nomenclature and derived from Affymetrix data and BLAST searches. Genes are divided into up- and down-regulated; within each category; lists are divided into numbers of genes differentially regulated, based upon the intersection between treatments of significantly and differentially regulated gene lists (Fig. 2). Probe set ID, gene symbol, gene title, GO biological process term, GO molecular function term, GO cellular component term, and gene accession numbers are presented. Red letters of Probe ID represent probes with no known annotation.
Figure S1. Hierarchical clustering of microarray hybridization gene expression data of roots exposed to GR24 and IAA treatments: IAA (10−8M), IAA (10−8M)+GR24 (13.5μM), GR24 (27μM), and non-treated controls. The colour scale represents fold change in gene expression in treatments versus controls.
Figure S2. Illustration of the general metabolism biological pathways in which differentially regulated genes putatively participate. Differentially regulated genes were identified from hybridization data of roots exposed to GR24 and IAA treatments [IAA (10−8M), IAA (10−8M)+GR24 (13.5μM), and GR24 (27μM)] versus non-treated controls. Blue or red squares represent individual genes. The colour within the squares represents fold change in gene expression in treatments versus controls; values of fold change are as indicated in the colour scale. The figure was adapted from MapMan software (Thimm et al., 2004).
We would like to thank Bruria Ben-Dor, Niv Faig, Einat Eliyahu, Dr Mor Grinberg-Yaari, Dr Yossefa Shahak, and Lior Avraham for technical help. We are grateful to Dr Koichi Yoneyama for the supplement of GR24.