Introduction

Plants are sessile and must endure environmental challenges, which include changes in water and nutrient availability and their heterogenous distribution in the soil. To acquire water and nutrients efficiently, plants dynamically modulate their root systems by regulating primary root growth, lateral root formation, and root hair number (de Dorlodot et al., 2007; Osmont et al., 2007; Nibau et al., 2008). Although progress has been made in dissecting the hormonal, genetic, and nutritional control of lateral root development (de Dorlodot et al., 2007; Osmont et al., 2007; Nibau et al., 2008; Peret et al., 2009), the mechanisms underlying the initiation and emergence of lateral roots are unclear. The formation of lateral roots is dependent on the coordinated action of many factors. In Arabidopsis thaliana, more than seventy genes have been shown to influence lateral root development (De Smet et al., 2006); however, the plant hormone auxin and its polar transport and redistribution have emerged as key factors (Fukaki et al., 2002; Nibau et al., 2008; Peret et al., 2009; Vanneste and Friml, 2009). Intercellular auxin movement is controlled by several plant-specific pin-formed (PIN) proteins that facilitate auxin efflux (Vanneste and Friml, 2009).

Coumarins (1, 2-benzopyrones) are secondary metabolites often found in roots of higher plants where they originate from the general phenylpropanoid pathway (Bourgaud et al., 2006). While their exact biological function in plants remains unclear, historically they are known to be involved in plant defense (Baumert et al., 2001; Carpinella et al., 2005; Brooker et al., 2007) along with flavonoids, phenolics, and alkaloids. Unlike most other secondary metabolites, coumarins appear to specifically inhibit seed germination and root growth (Svensson, 1971; Abenavoli et al., 2001, 2004, 2006, 2008; DeBolt et al., 2007), and this is consistent with previous observations that coumarins accumulate mainly in roots. They also influence membrane transport in the root tip (Lupini et al., 2010) and nitrogen uptake and metabolism (Abenavoli et al., 2001, 2003). Some natural coumarins also have received attention for their pharmacological properties, such as inhibition of HIV-1 and in treating thromboembolic diseases (Kostova et al., 2006).

Chemical genetics has been used extensively in plant biology for the identification of metabolites and their candidate genes involved in specific biochemical pathways (Blackwell and Zhao, 2003). As part of our efforts to characterize natural products that affect seed germination, root growth, and root architecture, we used forward chemical genetics to screen a small-molecule chemical library. Here, we focused on a coumarin derivative, 4-methylumbelliferone (4-MU), which inhibited germination and primary root growth, but promoted root branching after seedling establishment. We investigated the role of this compound on these parameters in more detail and demonstrated that 4-MU specifically accumulates in roots, where it is transformed into 4-methylumbelliferyl-β-D-glucoside (4-MU-Glc) by the action of UDP-glycosyltransferases (UGTs). We propose that UGTs play an integral role in the biochemical detoxification of 4-MU in Arabidopsis during seed germination and seedling development.

Methods and Materials

Plants, Chemicals and Forward Chemical Genetic Screens

All chemical standards were purchased from Sigma-Aldrich (St. Louis, MO, USA). Solvents used were high-pressure liquid chromatography (HPLC) grade. Arabidopsis thaliana ecotype Columbia (Col-0) was obtained from the Arabidopsis Biological Resource Center (Columbus OH, USA). Seeds were cold-treated at 4°C for 2 d before germination. A collection of 128 natural compounds was screened by germinating and growing Arabidopsis on sterile Murashige and Skoog (MS) medium (Sigma) supplemented with 1.2% agarose and 0, 25, 50, or 125 μM compound. Germination and seedling growth were conducted in a growth cabinet with 16 hr of light (22°C), 8 hr dark (18°C) and 60% relative humidity. Plants were grown for different lengths of time in 4-MU MS media based on the objectives of the experiments. For observation of 4-MU effects on seed germination and primary root growth, plants were treated for 6 d. For auxin transport analysis and MUG assays in root systems, plants were harvested after treatment for 14 d when lateral root branching has been initiated and DR5::GUS was actively expressed (Supplementary Data Fig. 1). To measure the effect of 4-MU on seedling establishment and root architecture, plants were treated for up to 22 d when seedlings and root branching were well established. Fresh weight and root length were measured immediately after harvest. The 35S::ABD-GFP (Motes et al., 2005) construct was provided by Dr. Elison B. Blancaflor, and the DR5::GUS transgenic line (Ulmasov et al., 1997) by Dr. Tom Guilfoyle (University of Missouri).

GUS Expression Analysis

Quantitative analysis of GUS activity was conducted using 4-methylumbelliferyl β-D-glucuronide hydrate (MUG) fluorometric assays (Jefferson, 1987). Seedlings (200 mg) were ground with 1 ml of GUS extraction buffer containing 0.1 M Na2HPO4 (pH 7.5), 10 mM EDTA, 0.1% Triton X-100, and 10 mM β-mercaptoethanol at 4°C. After centrifugation, 100 μl of the extract were incubated with 2 mM MUG in extraction buffer in a final volume of 1 ml for 3 hr at 37°C. Three independent replicates were conducted.

Fluorescent Confocal Microscopy

Laser scanning confocal and differential interference contrast microscopy was performed with a Carl Zeiss LSM150 microscope (Department of Veterinary Biomedical Sciences, University of Saskatchewan). GFP fluorescence was detected using excitation and emission wavelengths of 488 and 505 nm, respectively. The data shown are representative of three independent experiments. The images were analyzed using NIH Image J 1.37 software (http://rsb.info.nih.gov/ij/).

Identification and Purification of 4-MU-O-β-D-Glucopyranoside

Arabidopsis roots were harvested and frozen at −80°C from plants grown 6 d either in the presence or absence of 4-MU. A 100 mg sample of frozen roots was ground in liquid nitrogen in a 20 ml Potter-Elvehjem, followed by grinding in 10 ml methanol/water (80:20; v/v) for 10 min. After filtration, the pellet was extracted again with methanol/water overnight at 4°C in the dark. The two extracts were combined and dried at 35°C under vacuum, and the dried extract was resuspended at 1 mg ml−1 with methanol/water (50:50; v/v). HPLC/UV analysis was conducted using a Hewlett Packard Agilent 1100 chromatograph, a G-7120 diode array detector, HP Chemstation ver. 08.01 software and a Zorbax C18 column (150 × 4.6 mm, 5 μm, Mississauga, ON, Canada) with a 60 min linear gradient of 5–75% methanol v/v, a flow rate of 1.0 ml min−1, and an injection volume of 2 μl. LC-MS profiles were obtained on an Agilent 1100 HPLC and an API QStar XL pulsar hybrid system (Applied Biosystems, Foster City, CA, USA) (Li et al., 2011).

Identical chromatography conditions were used for semi-preparative purification of 4-MU-β-D-glucoside (4-MU-Glc) from 5 g of 4-MU-treated frozen root tissue, except that the flow rate was 3.0 ml min−1 and a Zorbax C18 column (150 mm x 2.5 cm, Mississauga, ON, Canada) was used. The major peak was isolated, dried in a rotary evaporatory, and then the compound was dissolved in 510 μl of fresh methanol (Sigma). 1H NMR, 1H-1H COSY, and HMBC spectra were measured on the isolated compound with a Bruker Advance 500 NMR spectrometer equipped with a Bruker 5 mm inverse tripe resonance TXI probe (Bruker Biospin, Rheinstetten, Germany) at the Saskatchewan Structural Science Centre, University of Saskatchewan. Chemical shifts (Table 1) (δ) were expressed in parts per million (ppm) and coupling constants (J) in Hertz (Hz).

Table 1 1H-NMR data of 4-MU and 4-MU-Glc (500 MHZ) a, b
Full size table

To determine the configuration of the sugar moiety, the isolated compound was refluxed with 6 N HCl (5 ml) at 100°C for 2 h (Li et al., 2009b). The mixture was extracted with CHCl3 to remove the aglycone, then the aqueous layer was neutralized with Na2CO3, filtered, and dried under vacuum. A portion of the aqueous residue was dissolved in H2O for sugar analysis by thin layer chromatography (TLC) (Macherey-Nagel, SilG/UV254, 0.20 mm) with n-BuOH:HOAc:H2O (6:1:2) as the solvent. TLC sample spots were detected by spraying aniline hydrogen phthalate reagent (100 ml n-butanol saturated by H2O, 0.96 g aniline and 1.66 g phthalic acid) and heating. The absolute configuration of glucose was determined by chiral GC-FID analysis using an Agilent GC and a 0.25 mm × 25 mm Hydrodexb-6-TBDM chiral capillary column (Macherey-Nagel, Germany). β-D-glucose (Sigma) was used as an authentic GC standard. A portion of the aqueous layer residue also was resuspended in 1 ml dichloromethane followed by the addition of 50 μl trifluoroacetic anhydride. The mixture was allowed to react at room temperature for 16 h and dried under a stream of nitrogen at room temperature. The sugar derivative was separated on the GC using the following temperature program: inlet temperature at 240°C; hydrogen carrier gas and a 1/20 split using nitrogen as the makeup gas; column temperatures starting at 120°C, ramping to 220°C at 50°C min−1 and maintenance for 12 min.

Microarray

Total RNA was extracted from Arabidopsis seedlings grown on agar plates for 6 d with or without 4-MU supplementation using the RNeasy plant mini kit (Qiagen, Valencia, CA, USA). cDNA was prepared using Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Three biological replicates were collected for each sample. An Arabidopsis 26 K cDNA array (Qiagen, Valencia, CA, USA) was hybridized with Cy5- and Cy3-labelled probe pairs from the 4-MU treated and untreated samples. Probe labeling, hybridization, and washing were conducted using a CyScribe Post-Labeling kit according to the manufacture’s instruction (Amersham Biosciences, Pittsburgh, PA, USA). Hybridized slides were scanned on a GenePix 4000B scanner (Axon Instruments Inc., Union City, CA, USA). The intensity (I) of each spot at λ546nm and λ647nm was transformed into a I546:I647ratio with Array-Pro Analyzer 4.0 (Media Cybernetics Inc., Bethesda, MD, USA). Microarray data initially were imported into BASE 1.2.16. The data then were filtered, normalized, and analyzed with Gene-Spring 6.1 (Silicon Genetics, Redwood City, CA, USA) as previously described (Zhao et al., 2009), and imported into MapMan for category analysis.

Real Time RT-PCR

mRNA from the microarray experiment was converted to cDNA with SuperScript III First-Strand Synthesis SuperMix (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. Quantitative real time PCR analysis (qRT-PCR) was carried out using a Platinum SYBR Green qPCR SuperMix-UDG kit (Invitrogen, Carlsbad, CA, USA) on an ABI PRISM StepOnePlus Real-Time PCR System (Applied Biosystems) as previously described (Gao et al., 2009; Li et al., 2011). The Arabidopsis Ef-1α gene was chosen as an endogenous reference gene. Primer sequences for qRT-PCR were designed using Oligo Perfect Primer Design software provided by Invitrogen (http://tools.invitrogen.com/) (Table 2). For each pair of primers, gel electrophoresis and melting curve analyses were performed to ensure that only a single PCR amplicon of the expected length and melting temperature was generated. Each sample was assayed in triplicate, and data were analyzed using Step-One v2.0 (Applied Biosystems). The level of each mRNA was calculated using the mean threshold cycle (Ct) value and normalized to that of the reference gene Ef-1α. All results are shown as means of at least three biological replicates with corresponding standard deviations (SD).

Table 2 Primers used in this study
Full size table

Glucosyltransferase Activity Assay

To generate constructs for GST-UGT expression, the full length open reading frames (ORFs) of UGT73B4 and UGT75B2 were amplified by PCR using primers 73B4-PET-F and 73B4-PET-R for UGT73B4 and primers 75B2-PET-F and 75B2-PET-R for UGT75B2 (Table 2), and cloned into the pET102/D/TOPO vector using the directional TOPO cloning system (Invitrogen, Carlsbad, CA, USA). All constructs were confirmed by DNA sequencing. The recombinant plasmids were transformed into Escherichia coli BL21 Star (DE3) cells (Invitrogen, Carlsbad, CA, USA), and the protein was purified as a GST fusion according to the manufacturer’s instructions (Qiagen, Valencia, CA, USA). Each glucosyltransferase activity assay mix (2 ml) contained 100 μg of recombinant protein, 50 mM Tris-HCl, pH 7.0, 10 mM β-mercaptoethanol, 10 mM UDP-glucose, and 1 mM 4-MU as substrate. The reaction mix was incubated at 28°C for 1 h, stopped with the addition of 200 μM trichloroacetic acid, and subjected to reverse-phase HPLC analysis as described above. Absorbance at λ320 was used to monitor chromatogram profiles, and the identity of peaks (compounds) was confirmed by comparing retention times and photo-diode array profiles with authentic standards.

Statistical Analysis

All experiments were conducted in triplicate and mean data (±SD) analyzed with ANOVA for significance (LSD at P < 0.05) using SAS 9.0 (SAS Institute, Inc. Cary, NC, USA), unless otherwise indicated. For microarray analysis, log ratios of intensity after background correction were scaled to have similar distribution across and consistency among arrays. The ANOVA model “log (y ijkgr ) = μ + A i  + D j  + T k  + G g  + AG ig  + DG jg  + TG kg  + ε ijkr ” was employed to detect differentially expressed genes using the normalized data, where log (y ijkgr ) represents the background-corrected and normalized natural logarithm of the intensity of the r th replicate of gene g on array i, with dye j, and treatment/condition k; μ represents the average natural logarithm of gene intensity over all the genes, arrays and dyes; A, D, T, and G represent the array, dye, treatment and main gene effect, respectively; and ε ijkr represents normal distribution with mean zero (Li et al., 2009a).

Results

Identification of 4-MU as a Root Growth Effector

A set of 128 natural compounds was screened for their impact on Arabidopsis seed germination, primary root growth, root architecture, and seedling establishment. Out of this screening, seven coumarins and five phenolic acids delayed Arabidopsis seed germination and inhibited root growth. One coumarin and two ginsenosides inhibited root growth alone, and one phenolic acid and one flavonoid inhibited seed germination alone (Supplementary Table 1). Of all compounds tested, 4-MU affected all three parameters, including root architecture. Since 4-MU was the only compound to affect root architecture, it became the focus of the remainder of this study

Seed germination was inhibited by a range of 4-MU treatments. Germination rates were 90% and 75% with 125 μM and 250 μM 4-MU, respectively. Seeds were unable to germinate when the concentration of 4-MU was increased to 500 μM (data not shown). Primary root growth was inhibited in 6-d-old seedlings grown on medium supplemented with 125 μM or 250 μM 4-MU (Fig. 1a,b,c). Primary root length was reduced by 25% and 75%, respectively, compared to untreated control seedlings over a 22 d time period (Fig. 2a), but growth of cotyledons, hypocotyls and rosette leaves was not affected at these concentrations (Figs. 1a,b, c and 2b). Accompanying the negative effect on primary root growth, the development of root hairs on primary roots also was affected by 4-MU. Approximately 20% of root hairs on primary roots were bulbous at 125 μM 4-MU (Fig. 1h), while very few root hairs were fully emerged at 250 μM 4-MU (Fig. 1i). After exposure to 125 μM 4-MU, lateral roots began to form as buds by day 14, and abundant lateral roots had formed by 22 days, even though primary root growth was hindered (Fig. 2c). Since accumulation of coumarins can be detected in plant tissues under UV light (Chong et al., 2002; Kai et al., 2008), the 6-d-old Arabidopsis seedlings grown on media supplemented with 4-MU were compared with untreated roots under UV light. Consistent with the inhibition of Arabidopsis root growth by 4-MU, blue fluorescence increased in roots of seedlings treated with this compound (shown for 250 μM 4-MU in Fig. 1e and f).

Fig. 1
figure 1

The effect of exogenous 4-MU on growth of primary roots. Arabidopsis seedlings were grown vertically on media supplemented with different concentrations of 4-MU for 6 d. a An untreated seedling. b A seedling treated with 125 μM 4-MU. c A seedling treated with 250 μM 4-MU. (d, e and f) UV (325 nm) channel of (a), (b) and (c) showing the accumulation of fluorescent 4-MU in primary roots. (g, h and i) Root hairs of (a), (b) and (c) showing bulbous root hairs on the primary root after treatment with 125 μM 4-MU, and inhibition of root hair formation after treatment with 250 μM 4-MU. Scale bars in a–f, 1 cm; in g–i, 2 mm

Full size image
Fig. 2
figure 2

The effect of 4-MU on Arabidopsis growth. a Time course of the effect of 4-MU on primary root elongation over 22 d. Values represent the means ± SD of three replicates. b Fresh weight of aerial parts of plants (mg FW per 10 seedlings without roots) and root length (cm) of 6-d-old plants in the presence or absence of 4-MU. Data represent the mean ± SD from 6 independent experiments. Asterisks indicate significant difference between treated and untreated plants based on a student test (P < 0.05). c Root system architecture and aerial vegetation of 22-d-old plants showing root branching in the presence (right) of 125 μM 4-MU, very little branching without 4-MU (left), and normal rosette growth under both conditions

Full size image

Accumulation and Glucosylation of 4-MU in Arabidopsis Roots

To confirm the presence of 4-MU in Arabidopsis roots, we analyzed the phytochemical composition of roots of 6-d-old seedlings by HPLC-UV/MS. In MeOH extracts from roots exposed to 125 μM 4-MU, 4-MU and a large additional peak were detected at Rt 36.5 min and Rt 23.0 min, respectively. Neither of these peaks was detected in the roots of untreated seedlings (Fig. 3a). Comparison of the 1H NMR spectrum of the new reaction product with that of 4-MU (Supplementary Data Figs. 2 and 3) showed no significant chemical shifts in the aromatic proton signals, but an additional set of sugar moiety signals appeared at δ 4.95 (1H, d, J = 7.6 Hz), 3.82 (1H, d, J = 11.8 Hz), 3.62 (1H, dd, J = 5.6, 11.8 Hz), and 3.32-3.45 (3H, m) (Table 1). Combinational 1H-1H COSY and HMBC NMR spectra (Supplementary Data Figs. 4 and 5) were used to elucidate the structure. The sugar moiety was identified as glucose on the basis of an acid hydrolysis reaction, which gave a reaction product with the same R f on a TLC plate as an authentic glucose standard compound (data not shown). The coupling constant J H1′, H2′ (7.6 Hz) of the new enzyme reaction product, followed by chiral GC analysis, indicated an absolute configuration of β-D-glucose. The new reaction product was defined as C16H18O8 [4-MU-β-D-glucopyranoside (4-MU-Glc)] based on its HRESIMS ([M + Na]+, m/z 361.0906, calc. 361.0899) (Fig. 3b). This indicates that 4-MU is partially transformed into 4-MU-Glc by glucosylation after uptake in Arabidopsis roots.

Fig. 3
figure 3

Glucosylation of 4-MU in Arabidopsis roots. a HPLC-UV profiles (λ325) of 80% methanol extracts from roots of 6-d-old seedlings treated with (upper panel) or without (bottom panel) 125 μM 4-MU. A new product was identified as 4-MU-β-D-glucoside (4-MU-Glc) in the 4-MU treated root extracts. b Schematic of the UGT-catalyzed biotransformation of 4-MU to 4-MU-Glc in Arabidopsis roots

Full size image

To gain additional insight into how Arabidopsis plants react to 4-MU by glucosylation after seed germination, we conducted a microarray analysis of 6-d-old seedlings grown on medium containing 125 μm 4-MU (Supplementary Data Tables 2 and 3). A total of 316 genes showed statistically significant up-regulation, while 243 showed statistically significant down-regulation. Genes determined to be up-regulated or down-regulated ≥ 2-fold in the treated relative to the untreated sample were catalogued and placed into twelve categories based on gene function according to Arabidopsis Gene Ontology (Supplementary Data Fig. 6). Although differences between treated and untreated samples were noted for individual genes, the number of up-regulated and down-regulated genes were similar in the categories of response to stress (51 vs. 48), response to abiotic and biotic stresses (49 vs. 44), and transportation (32 vs. 25), respectively. It was notable that most of the genes with altered expression did not have known functions (99 vs. 145, respectively). Four genes encoding uridine diphosphate (UDP) glycosyltransferases (UGTs) were up-regulated more than two-fold in the 4-MU treated seedlings, these included UGT73B4, UGT73C1, UGT75B2, and GAUT2 (At2g46480). Consistent with the microarray data, qRT-PCR analysis revealed significant up-regulation of these four UGT genes in roots of 6-d-old seedlings exposed to 125 μm 4-MU (Fig. 4). Since UGT73B4, UGT73C1, and UGT75B2 are active on hydroxycoumarins, scopoletin, and esculetin (Lim et al., 2003), we tested whether 4-MU could be glucosylated by these UGTs using purified recombinant UGT73B4 and UGT75B2 enzymes. Reverse-phase HPLC and the identification of compounds in the post-reaction mixture showed a product with a retention time (22.1 min) and a photo-diode array profile identical to the authentic standard 4-MU-Glc in the reaction with UGT73B4, indicating that 4-MU is a substrate of this UGT (Fig. 5). UGT75B2 enzyme activity was not detected with 4-MU (data not shown).

Fig. 4
figure 4

Induction of glycosyltransferase genes by 4-MU in Arabidopsis roots. Quantitative RT-PCR analysis was used to analyze expression of genes encoding glycosyltransferases in the roots of 6-day-old seedlings treated with 125 μM 4-MU relative to levels from untreated seedlings. Ef-1α was used as a reference gene

Full size image
Fig. 5
figure 5

HPLC-UV profiles of the reaction products of UGT73B4 recombinant protein incubated with 4-MU. Reaction products of GST protein (Control) or GST-UGT73B4 fusion (UGT73B4) were subjected to HPLC analysis and the products were detected at λ320, (RT, 22.1 min) and confirmed by NMR, showing that UGT73B4 catalyzes the conversion of 4-MU to a 4-MU-Glc

Full size image

Involvement of Auxin Efflux in 4-MU-Induced Lateral Root Formation

Auxin plays a key role in controlling lateral root formation (Fukaki and Tasaka, 2009). To test whether lateral root formation induced by 4-MU treatment was accompanied by an alteration in auxin signaling, we performed qRT-PCR analysis on genes involved in either auxin signal transduction, flux, or biosynthesis using total RNA extracted from 4-MU-treated roots of 14-d-old seedlings when lateral root branching has been initiated and DR5::GUS was actively expressed (Supplementary Data Fig. 1). For auxin flux facilitator genes, we chose PIN2 and PIN3 in preference to the other five PIN genes, which have additional or other functions. For example, PIN1 is involved in vascular development, phyllotaxis, vein formation, embryogenesis, and lateral organ development (Petrásek and Friml 2009). PIN2 and PIN3 were up-regulated by 4-MU, whereas the expression of an auxin receptor TIR1, a rate-limiting for the auxin biosynthetic gene YUCCA1 (Zhao et al., 2001), and an auxin-responsive GH3 family member At2g14960 (Mishra et al., 2009) was unchanged relative to untreated seedlings (Fig. 6a). Additionally, the transcript level of IAA14/SLR1, one of the most important auxin responsive AUX/IAA genes involved in lateral root initiation (Fukaki et al., 2002; Fukaki and Tasaka, 2009), was elevated by 4-MU. This analysis suggested that 4-MU affected auxin redistribution rather than biosynthesis. To eliminate the possibility that 4-MU treatment affects endogenous auxin biosynthesis, β-glucuronidase (GUS) activity under the control of the synthetic auxin-responsive promoter DR5 (DR5::GUS) was determined in roots of 14-d-old seedlings grown on medium supplemented with 125 μM 4-MU. MUG activity (as measured by a fluorometric assay) in DR5::GUS—transformed Arabidopsis lines grown in the presence of 4-MU showed no difference compared to lines without 4-MU treatment (Fig. 6b), indicating that 4-MU had no effect on the regulation of auxin biosynthetic genes.

Fig. 6
figure 6

(A) Transcript abundance of auxin biosynthesis, efflux, signalling, and responsive genes after 4-MU treatment. Expression was tested for the auxin efflux facilitators PIN2 and PIN3, the auxin-responsive genes IAA14 and GH3 (At2g14960), the major auxin biosynthetic gene YUCCA1 and the auxin receptor TIR1. Real time PCR was used to examine transcript levels of the six genes are relative to that in untreated Arabidopsis seedlings. Ef-1α was used as a reference gene. (B) Auxin biosynthesis as detected by GUS activity levels in DR5::GUS transgenic seedlings untreated or treated 125 μM 4-MU. Values are the means ± SD of triplicate samples

Full size image

Alteration of Actin Filament Organization by 4-MU Treatment

The actin cytoskeleton is important for root growth and is also thought to play a role in the polarization of auxin fluxes (Dhonukshe et al., 2008; Nick et al., 2009). We, therefore, examined the organization of the actin cytoskeleton system in roots in response to 4-MU. Arabidopsis plants were transformed with the actin-binding domain (ABD)-GFP protein fusion construct, which specifically binds filamentous (F)-actin (Motes et al., 2005) and allows visualization of the F-actin cytoskeleton. ABD-GFP florescence strongly accumulated in the root caps of untreated 6-d-old seedlings (Fig. 7b), whereas GFP fluorescence was observed in the meristematic tissues of seedlings treated with 125 μM 4-MU (Fig. 7e). Alteration of the root F-actin organization in plants treated with 125 μM 4-MU was accompanied by an abnormal root tip phenotype in which root caps appeared to be physically detached with irregularly shaped cells (Fig. 7d,f).

Fig. 7
figure 7

Alteration of actin cytoskeleton organization in Arabidopsis root tips in response to 4-MU. Primary root tips of 6-d-old 35S::ABD-GFP transformed seedlings were examined by fluorescence microscopy. (a, b and c) A root tip in the absence of 4-MU. (d, e and f) A root tip treated with 125 μM 4-MU. Brightfield (a and d), GFP channel (b and e), and merged (c and f) images are shown. Scale bar, 100 μm

Full size image

Discussion

We have characterized the response of the Arabidopsis root system to treatment with 4-MU. In the presence of this compound, seed germination was inhibited, primary root growth and root hairs were reduced, and lateral root emergence was stimulated, while cotyledons and seedling rosette leaves were unaffected (at least up to 22 days of evaluation). At high concentrations (500 μM), germination was completely inhibited. Transcripts of three UGTs (UGT73B4, UGT73C1, UGT75B2) also were up-regulated after treatment with 4-MU, and a recombinant UGT73B4 enzyme could catalyze the glucosylation of 4-MU in vitro into the same end product (4-MU-Glc) that was accumulated with 4-MU in vivo in roots of the treated plants.

Plants have evolved complex pathways for responding to chemicals and toxins, such as glutathione sulphotransferase (GST)-catalyzed conjugation of glutathione and UGT-catalyzed glucosylation of diverse acceptors (Edwards et al., 2000; Gandia-Herrero et al., 2008). Glucosylation alters the properties of acceptor chemicals, such as their bioactivity, stability, hydrophilicity, and subcellular localization (Bowles et al., 2005). UGTs are involved in the synthesis and modification of glucoconjugates via transfer of sugars to a wide range of acceptor molecules that include plant secondary metabolites, phytotoxins, and xenobiotics (Bowles et al., 2005). UGT73B4 and several other UGTs have been shown to glucosylate coumarins, scopoletin, and esculetin, the flavonoid quercetin and 2,4,6-trinitrotoluene (TNT) (Lim et al., 2003; Gandia-Herrero et al., 2008). Hence, UGT-catalyzed glucosylation of 4-MU may represent a mechanism to inactivate this compound in Arabidopsis roots.

Auxin and its polar transport and redistribution are key factors controlling lateral root initiation, formation, and development (Nibau et al., 2008; Fukaki and Tasaka, 2009; Peret et al., 2009; Vanneste and Friml, 2009). Lateral root emergence was correlated with changes in auxin responsive gene expression in Arabidopsis challenged with 4-MU. Genes that encode the auxin efflux facilitators PIN2 and PIN3 were strongly up-regulated by 4-MU, and expression of the auxin receptor TIR1 or the key auxin biosynthetic enzyme YUCCA1 was unchanged in the presence of 4-MU. The F-actin fibrils in the primary root also were disrupted, and the root cap was detached under 4-MU exposure. These observations suggest that 4-MU induces lateral root emergence and root branching through auxin redistribution rather than auxin biosynthesis and does so in concert with (and as a consequence of) the inhibition of primary root growth.

Small compounds with impacts on cellular activity are useful, potent chemical tools for defining protein function (Walsh and Chang, 2006) and have provided insight into regulatory mechanisms, e.g., toyocamycin inhibition of auxin signaling (Hayashi et al., 2009). Our identification and characterization of 4-MU as a positive regulator of Arabidopsis root branching as a consequence of primary root inhibition is a starting point for continued chemical genetics investigations aimed at identifying the cellular target of 4-MU. At this point, we also anticipate that 4-MU will prove to be a useful chemical tool for studying root biology and development in crop plants, particularly the Brassicas which are genetically close to Arabidopsis.