Strigolactone analog GR24 triggers changes in PIN2 polarity, vesicle trafficking and actin filament architecture

Authors


Summary

  • Strigolactones (SLs) are plant hormones that regulate shoot and root development in a MAX2-dependent manner. The mechanism underlying SLs' effects on roots is unclear.
  • We used root hair elongation to measure root response to SLs. We examined the effects of GR24 (a synthetic, biologically active SL analog) on localization of the auxin efflux transporter PIN2, endosomal trafficking, and F-actin architecture and dynamics in the plasma membrane (PM) of epidermal cells of the primary root elongation zone in wildtype (WT) Arabidopsis and the SL-insensitive mutant max2. We also recorded the response to GR24 of trafficking (tir3), actin (der1) and PIN2 (eir1) mutants.
  • GR24 increased polar localization of PIN2 in the PM of epidermal cells and accumulation of PIN2-containing brefeldin A (BFA) bodies, increased ARA7-labeled endosomal trafficking, reduced F-actin bundling and enhanced actin dynamics, all in a MAX2-dependent manner. Most of the der1 and tir3 mutant lines also displayed reduced sensitivity to GR24 with respect to root hair elongation.
  • We suggest that SLs increase PIN2 polar localization, PIN2 endocytosis, endosomal trafficking, actin debundling and actin dynamics in a MAX2-dependent fashion. This enhancement might underlie the WT root's response to SLs, and suggests noncell autonomous activity of SLs in roots.

Introduction

The plant hormone strigolactone (SL) has been found in many plant species (Yoneyama et al., 2013). SLs affect different developmental processes in plants. For instance, they suppress the outgrowth of preformed axillary buds in the shoot (e.g. Gomez-Roldan et al., 2008; Umehara et al., 2008), are negative regulators of adventitious root formation (Rasmussen et al., 2012) and are positive regulators of shoot secondary growth (Agusti et al., 2011). In roots, SLs suppress lateral root formation, induce root hair elongation (Kapulnik et al., 2011a) and promote meristem cell number in the primary root meristem (Ruyter-Spira et al., 2011; Koren et al., 2013). SL activities in plants are carried out through a signaling pathway that includes the F-box protein MORE AXILLARY GROWTH 2 (MAX2; Stirnberg et al., 2007; Umehara et al., 2008). Another component of SL signaling is DWARF14 (D14), an α/β-hydrolase (Arite et al., 2009). It has been suggested that in the presence of GR24, D14 interacts with MAX2 to confer SL perception and response (Hamiaux et al., 2012).

Strigolactones are probably regulators of auxin flux. They have been suggested to dampen auxin transport in the shoot (e.g. Crawford et al., 2010; Domagalska & Leyser, 2011). In the stem, SL signaling has been found to trigger depletion of the auxin transporter PIN-FORMED1 (PIN1) from the plasma membrane (PM) of xylem parenchyma cells, supporting the hypothesis that SLs regulate shoot branching by modulating auxin transport (Shinohara et al., 2013). In the root, SLs have been suggested to affect auxin efflux (Koltai et al., 2010) and PIN1 (Ruyter-Spira et al., 2011). Moreover, endodermal SL signaling has been suggested to affect auxin flux in the root tip and thereby lateral-root formation, primary root meristem size and root hair elongation (Koltai & Kapulnik, 2013).

The polar position of PINs in the PM is an important determinant of the direction of auxin flux (Wisniewska et al., 2006). F-actin has been shown to play a role in basal and apical PIN targeting to the PM, and in cellular PIN trafficking (Geldner et al., 2001; Nagawa et al., 2012). Specifically, stabilization of actin filaments has been suggested to slow their dynamics and block PIN subcellular dynamics (Dhonukshe et al., 2008a). In addition, dynamic vesicle trafficking has been shown to determine, to a large extent, the PINs' PM localization (e.g. Geldner et al., 2001; Dhonukshe et al., 2007).

Elongation of the root hair tip is affected by auxin transport in the epidermal cell layer containing the hair cells and flanking nonhair cells, including in the root elongation zone (Jones et al., 2009). To better understand the mechanism underlying SLs' effects on root cells and to examine the suggested effect of SLs on auxin transport in roots, we therefore studied their association with changes in PIN2 expression and localization in the PM of epidermal cells of the primary root elongation zone. We also studied whether the changes in PIN2 are derived from SL activity on vesicle trafficking and F-actin architecture and dynamics.

Root hair number and length are suggested to be directly associated with the plant's ability to absorb nutrients from the soil (Sánchez-Calderón et al., 2005; reviewed by Gilroy & Jones, 2000). In the present study, root hair elongation was used as a measure of the root's response to SLs. We found that, under the experimental conditions, SLs increase PIN2 polarity in the epidermal cell membrane of the root in wildtype (WT) Arabidopsis but not in the SL-insensitive mutant max2-1. SL treatment also increased the accumulation of PIN2-containing brefeldin A (BFA) bodies and induced trafficking of endosomes in the WT, but not in max2-1. Moreover, SLs reduced F-actin filament bundling while inducing its dynamics in a MAX2-dependent manner. Finally, mutants flawed in actin and vesicle trafficking, but not in PIN2, showed a reduced response to SLs. Our results suggest that SLs affect actin filament organization and dynamics, endosome trafficking and PIN2 localization in the PM of the epidermal cell layer.

Materials and Methods

Arabidopsis strains, growth conditions and treatments

Seeds of Arabidopsis thaliana (L.) Heynh. used in this study included WT (Col-0, Ler and C24), Col-0 homozygous max2-1 lines, auxin transport-deficient and ethylene-insensitive eir1-1 and eir1-4, TRANSPORT INHIBITOR RESISTANT3 (TIR3/BIG) tir3-103, tir3-104, tir3-105 (obtained from the ABRC stock center http://abrc.osu.edu/), PIN2::PIN2-GFP (Vieten et al., 2005; seeds kindly provided by the Department of Molecular Biology & Ecology of Plants, Tel Aviv University, Israel), AUX1::AUX1-YFP (Jones et al., 2009; seeds kindly provided by the Department of Molecular Biology & Ecology of Plants, Tel Aviv University, Israel), 35S::TALIN-GFP (Kost et al., 1998; construct kindly provided by N. H. Chua, the Rockefeller University, New York, NY, USA), 35S::ARA7-GFP (Ueda et al., 2004; construct kindly provided by T. Ueda, University of Tokyo, Japan), fABD2-GFP (Sheahan et al., 2004; seeds kindly provided by C. Staiger, Purdue University, West Lafayette, IN, USA), Ler homozygous line max2-8 and C24 homozygous lines DEFORMED ROOT HAIR (der)1-1, der1-2, der1-3 (obtained from the ABRC stock center). max2-1 PIN2::PIN2-GFP-expressing lines (designated max2-1-22), max2-1 35S::ARA7-GFP-expressing lines (designated max2-1-73) and max2-1 35S::TALIN-GFP-expressing lines (designated max2-1-28), were identified from F2 progeny of a cross between max2-1 and Col-0 PIN2::PIN2-GFP, Col-0 35S::ARA7-GFP and Col-0 35S::TALIN-GFP, respectively, based on F2 and F3 shoot branching and lack of a root hair elongation response (determined as described in Kapulnik et al., 2011a) to GR24. Selected seeds of homozygous lines were used in the present study.

Seeds were surface-sterilized and germinated on 1/2 Murashige and Skoog (MS) solidified with 1% (w/v) agar in 90 mm round plates. Plates contained various concentrations of hormones: the SL analog GR24 as a mixture of four diastereomers ((±)-GR24 and (±)-2′-epi-GR24) (Johnson et al., 1981). GR24 treatments were conducted at a concentration of 3 × 10−6 M, determined to be the optimal concentration for inducing root hair elongation under the examined conditions (Kapulnik et al., 2011a). GR24 was initially dissolved in acetone (5 mg GR24 in 3 ml acetone) to give a 5.6 mM solution. This solution was then further diluted with double-distilled sterile water. Hence, experimental controls for the GR24 treatments included roots treated with acetone at the concentration used in the respective GR24 treatment. IAA treatments were conducted at a concentration of 5 × 10−8 M diluted in double-distilled sterile water from a stock solution of 1 × 10−4 M. Hence, experimental controls for the IAA treatments consisted of roots treated with double-distilled sterile water. BFA treatments were given at a concentration of 100 μM for 1 h.

In each of the experiments, hormone-treated roots were compared with the respective controls. Plates were incubated vertically in the dark at 4°C for 2 d to synchronize germination. Plates were left unsealed to prevent accumulation of gases (e.g. ethylene); they were positioned in an upright 90° position, and incubated at 22°C with a light intensity of 50–60 mol photons m−2 s−1 provided by white fluorescent tubes under a photoperiod of 16 : 8 h, light : dark. Germination occurred after 48 h, and plates were incubated for an additional 72–96 h before observation.

Live-cell imaging and quantification

Fluorescence intensity of green fluorescent protein (GFP), yellow fluorescent protein (YFP) and the endocytic tracer fluorescent styryl dye FM4-64 in live cell imaging was examined and photographed with an Olympus IX 81 inverted microscope linked to a laser scanning device (FLUOVIEW 500), equipped with an argon-ion laser and 60X 1.0 NA PlanApo water objective at a wavelength of 488 nm for GFP and FM4-64, and 515 nm for YFP. Emissions were collected through a BA 505–525 filter for GFP, a BA 535–560 filter for YFP and a BA 610 IF filter for FM4-64. The same image acquisition parameters were used for all signal measurements. For quantification of the fluorescent signal in epidermal cell membranes, each cell membrane was carefully encircled with a drawn line, and the mean pixel intensity value of the encircled area was measured using the color histogram (green fluorescence) of ImageJ software (http://rsbweb.nih.gov/ij/). To determine signal intensity and polarity, the mean pixel intensities were obtained from the polar and lateral sides of the individual cells. The polarity index was determined as the ratio of intensity on the polar side to that on the lateral side, divided by two (Dhonukshe et al., 2008b). A minimum 60 cells were analyzed for each treatment from at least 30 seedlings.

To determine F-actin filament architecture, Z optical sections in steps of 0.68 μm from the epidermal layer of the roots (four sections per root treated with GR24, IAA or controls; two sections per root treated with latrunculin B (LatB), owing to the high background) were subjected to processing as described in Higaki et al. (2010) for determination of skewness, using ImageJ. For time lapse analysis of actin filament or vesicle dynamics, single optical sections were taken every 3.2 s. Actin movement was subjected to processing as described by Higaki et al. (2010) for the determination of flow (median and average velocity), using ImageJ. Endosomal movement was subjected to processing as described by Higaki et al. (2008) for the determination of average velocity, using ImageJ. At least 60 cells were analyzed for each treatment and for each measured parameter, from at least 30 seedlings. Means of replicates were subjected to statistical analysis by ANOVA pairwise Student's t-test ( 0.05) using the JMP statistical package.

Determination of root hair length

Root hair length of roots grown on GR24 or IAA and control plates was examined as described by Kapulnik et al. (2011a). Briefly, at 72 h postgermination, roots were examined on the plates using a stereomicroscope (Leica MZFLIII, Leica Microsystems GmbH, Wetzler, Germany) and pictures were taken of the root segments that grew on the plates for 48 h with a Nikon DS-Fi1 camera (Amsterdam, the Netherlands). Experiments were repeated four times; each treatment within each experiment included three replicates, 10 seedlings per replicate. Measurements of root hair length were taken from 10 pictures (10 plants) per treatment, per replicate; 15–30 different root hairs were measured per picture using ImageJ (http://rsbweb.nih.gov/ij/) (= 300). Means of replicates were subjected to statistical analysis by one-way ANOVA pairwise Student's t-test ( 0.05) using the JMP statistical package (SAS Institute, Cary, NC, USA).

RNA extraction and cDNA preparation

For RNA extraction from roots, 2–5 mm of roots from seedlings grown on 1/2 MS plates for 96 h postgermination were sectioned with a sharp blade and snap-frozen in liquid nitrogen. Each replicate consisted of 180 plants. Total RNA was extracted using TRI reagent (Sigma) and digested with Turbo DNase enzyme (Ambion) as per the manufacturer's instructions. Reverse transcription of cDNA was performed as described by Mayzlish-Gati et al. (2010).

Quantitative real-time PCR

Quantitative real-time PCR (qPCR) was performed on cDNA reverse-transcribed from the RNA extracted from the roots. Arabidopsis 15S ribosomal RNA (GenBank accession no. AT1G04270.1) served as the reference gene for RNA amount, and was amplified using specific primers (forward) 5′-CAA AGGAGTTGATCTCGATGCTCTT-3′ and (reverse) 5′-GCCTCCCTTTTCGCTTTCC-3′. To determine PIN2 (GenBank accession no. AF086906.1) gene expression, the following primers were used: (forward) 5′-CTGGTCCAGGTGGAGATGTT-3′ and (reverse) 5′-GCCTCCTCTTCCTGCTTTCT-3′. qPCR amplification was performed as described by Mayzlish-Gati et al. (2010), using the 2−ΔΔCT method (Livak & Schmittgen, 2001). The experiment was performed in three to four biological replicates, each with four technical repeats. Means and standard error were calculated for all biological replicates. Values above or below 1 represented an increase or decrease, respectively, in the steady-state level of gene transcripts for the examined conditions (i.e. that of the numerator vs that of the denominator). Means and standard deviation were calculated for at least three biological replicates in each examined treatment. Means of replicates were subjected to statistical analysis by one-way ANOVA pairwise Student's t-test ( 0.05) using the JMP statistical package.

Results

SL treatment increases PIN2 levels and polarity in the PM of epidermal cells in the WT root elongation zone

We previously showed that SLs lead to an increase in root hair elongation in WT Arabidopsis (Kapulnik et al., 2011a; Supporting Information, Fig. S1). PIN2 auxin transporters play a role in this process, in the epidermal cells of the root, probably maintaining the cell polarity and auxin balance necessary for the developing hair cells (Jones et al., 2009; reviewed by Lee & Cho, 2013). We used root hair elongation as a bioassay for root response to GR24, and examined whether GR24 treatment of seedlings that had grown on plates under conditions that lead to increased root hair elongation in the WT (72–96 h postgermination, 3 μM GR24; Fig. S1) also affects PIN2's presence in the PM of epidermal cells in the primary root elongation zone. Under these conditions, GR24 led to a significant increase in PIN2-GFP signal in the apical PM of the epidermal cells (Fig. 1a,b,e). GR24 treatment also led to an increase in the PIN2 polarization index (i.e. the ratio of intensity at the apical vs lateral PM) in these cells (Fig. 1a,b,f). Interestingly, GR24 did not lead to a significant increase in AUX1-YFP signal in the apical PM of the epidermal cells (Fig. S2a–c), or to an increase in AUX1 polarization index in these cells (Fig. S2a,b,d). These results suggest that, under the examined conditions, SLs affect PIN2 but not AUX1 polarity in the PM of epidermal cells in the root elongation zone.

Figure 1.

Effect of GR24 (3 × 10−6 M) treatment on PIN2 plasma membrane localization and polarity in the epidermal cells of the primary root elongation zone in PIN2::PIN2-GFP Arabidopsis thaliana seedlings grown on acetone or GR24 plates (72–96 h postgermination). (a–d) PIN2-GFP signal in Col-0 (a, b) and max2-1 (max2-1-22; c, d) roots grown on acetone (ac; a, c) or GR24 (b, d) plates (bars, 50 μm). Arrows indicate rootwards. (e) Intensity of the PIN2-GFP signal (arbitrary units) in the apical plasma membrane of Col-0 and max2-1-22 roots grown on acetone (open columns) or GR24 (closed columns) plates. (f) Polarity of the PIN2-GFP signal in the plasma membrane of Col-0 and max2-1-22 roots grown on acetone (open columns) or GR24 (closed columns) plates. Polarity index was determined as the ratio of intensity on the polar side to that on the lateral side, divided by two (Dhonukshe et al., 2008b). A total of 50–60 cells from 10 plants were examined per replicate; three replicates were included. The same lowercase letters above the bars indicate statistically nonsignificant differences between means by one-way ANOVA pairwise Student's t-test ( 0.05). Means are significantly different between hormone treatment (GR24) and control (ac) as determined by ANOVA pairwise Student's t-test (*,  0.05); error bars ± 1 SE.

PIN2 level and polarity in max2 do not respond to SL treatment

To examine whether the effect of GR24 on PIN2 expression and localization is MAX2-dependent, we generated max2-1 SL-response mutants expressing PIN2-GFP under the PIN2 promoter and determined PIN2 expression and distribution in the PM of epidermal cells in the primary root elongation zone. The intensity in the apical PM and polarity of the PIN2-GFP signal in the max2-1 epidermal cell PM were similar to those in the WT (Fig. 1a,c,e,f). Moreover, in agreement with max2-1 roots' nonresponsiveness to SLs (Kapulnik et al., 2011a), neither the intensity of the signal in the apical PM nor its polarity were affected in max2-1 by GR24 treatment (Fig. 1c–f). These results further suggest that SLs affect the level and distribution of PIN2 proteins in the PM of epidermal cells in the primary root elongation zone, and that this effect is dependent on MAX2.

SLs stimulate PIN2 expression

Our results suggested that PIN2-GFP signal in the apical PM increases upon GR24 treatment in the WT but not in max2-1. We therefore examined whether the increase in GFP signal involves an increase in PIN2 gene expression in the root tips following GR24 treatment in the WT but not in max2. First, we compared the levels of PIN2 expression in the root tips of max2 and the WT, in two genotypic backgrounds: Col-0 vs max2-1 and Ler vs max2-8. PIN2 expression was reduced 0.50 ± 0.34-fold in max2-1 vs Col-0 and 0.60 ± 0.23-fold in max2-8 vs Ler. Thus, max2 showed only a minor (and insignificant) reduction in PIN2 expression compared with the WT in the two genotypic backgrounds.

Next, we examined the level of PIN2 expression in the WT (Col-0) treated with GR24 (3 μM) and the acetone control. PIN2 expression in the treated WT was 3.56 ± 0.39-fold that in the control. By contrast, GR24 treatment did not lead to any change in PIN2 expression in max2-1, being 0.96 ± 0.19-fold that in the control. Together, the results suggest that SL signaling leads to induced levels of PIN2 gene expression in correlation with the increase in PIN2-GFP signal in the apical PM, and that this induction is dependent on MAX2.

SLs increase accumulation of PIN2-containing BFA bodies in the WT but not in max2-1

PIN proteins undergo constitutive cycling between the PM and the endosomes; this dynamic vesicle trafficking determines, to a large extent, PINs' PM localization (e.g. Geldner et al., 2001; Dhonukshe et al., 2007). Part of this pathway is sensitive to BFA; BFA preferentially interferes with basal PIN recycling without affecting endocytosis, thereby inducing the accumulation of PIN proteins in so-called BFA bodies (Geldner et al., 2001). We examined whether the effect of SLs on PIN2 localization in WT roots is associated with changes in PIN2 endocytosis. In the WT, GR24 treatment led to increases in the number of PIN2-containing BFA bodies per cell (Fig. 2a,b,e,f,i), the percentage of cells with these BFA bodies (Fig. 2a,b,e,f,j) and their area (Fig. 2a,b,e,f,k) compared with acetone control seedlings. Moreover, in agreement with max2-1 roots' nonresponsiveness to SLs (Kapulnik et al., 2011a), neither the number of PIN2-containing BFA bodies per cell (Fig. 2c,d,g–i), nor the percentage of cells with these BFA bodies (Fig. 2c,d,g,h,j) or their area (Fig. 2c,d,g,h,k) were altered in max2-1 by GR24 treatment in comparison to controls. In max2-1 acetone controls, the number of PIN2-containing BFA bodies, area and percentage of cells containing BFA bodies were similar to the WT acetone control (Fig. 2a,c,e,g,i–k). Together, these results suggest that GR24 treatment induces PIN2 endocytosis in a MAX2-dependent manner.

Figure 2.

Effect of GR24 (3 × 10−6 M) treatments on PIN2-containing brefeldin A (BFA) bodies in the epidermal cells of the primary root elongation zone in PIN2::PIN2-GFP Arabidopsis thaliana seedlings grown on GR24 or acetone plates (72–96 h postgermination). (a–d) PIN2 signal (green) in Col-0 (a, b) and max 2-1 (max2-1-22; c, d) roots grown on acetone (ac; a, c) or GR24 (b, d) plates and treated with BFA (100 μM, 1 h) (bars, 50 μM). (e–h) PIN2-containing BFA bodies (green, PIN-GFP signal; red, FM4-64) in Col-0 (e, f) and max2-1-22 (g, h) roots grown on acetone (ac; e, g) or GR24 (f, h) plates and treated with BFA (100 μM, 1 h). Bars, 50 μm; arrows indicate PIN2-containing BFA bodies. (i) Number of PIN2-containing BFA bodies per cell in the apical plasma membrane of Col-0 and max2-1-22 roots grown on acetone (open columns) or GR24 (closed columns) plates. (j) Percentage of cells with PIN2-containing BFA bodies in the apical plasma membrane of Col-0 and max2-1-22 roots grown on acetone (open columns) or GR24 (closed columns) plates. (k) Area per cell of PIN2-containing BFA bodies in the apical plasma membrane of Col-0 and max2-1-22 roots grown on acetone (open columns) or GR24 (closed columns) plates. A total of 50–60 cells from 10 plants were examined per replicate; three replicates were included. The same lowercase letters above the bars indicate statistically nonsignificant differences between means by one-way ANOVA pairwise Student's t-test ( 0.05). Means are significantly different between hormone treatment (GR24) and control (ac) as determined by ANOVA pairwise Student's t-test (*,  0.05); error bars ± 1 SE.

SLs induce endosomal movement in the WT but not in max2-1

Next, we examined the effect of GR24 on endosomal movement velocity in the PM of epidermal cells in the primary root elongation zone by analyzing a WT line that expresses GFP fusions to the endosomal marker ARA7 (Ueda et al., 2004). Endosomal movement was observed as superpositions of a single optical section taken every 3.2 s from the epidermal layer of the roots. Velocity of endosome movement was measured according to Higaki et al. (2008).

The velocity of endosomal movement in seedlings grown on acetone and GR24 plates (72–96 h post germination, 3 μM GR24; Fig. S1) was examined. Under these conditions, endosomal movement was faster in GR24-treated roots than in the acetone control (Fig. 3a,b,e; see Videos S1, S2). This was evident in the overlay of frame 1 (green) on frame 5 (+16 s, red), and on frame 9 (+29 s, blue) of ARA7- labeled endosomes following GR24 treatment as an increased proportion of green, red or blue dots in the GR24-treated roots (Fig. 3a,b). Suggested ‘trails’ of endosomal movement are shown in Fig. 3(a–d). Together, these results suggest that GR24 treatment enhances the velocity of endosomal movement.

Figure 3.

Effect of GR24 (3 × 10−6 M) treatments on the velocity of ARA7-labeled endosomes in the epidermal cells of the primary root elongation zone in 35S::ARA7-GFP Arabidopsis thaliana seedlings grown on acetone or GR24 plates (72–96 h postgermination). (a–d) Overlay of frame 1 (green) on frame 5 (+16 s, red), and on frame 9 (+29 s, blue) of ARA7 in Col-0 ARA7 (a, b), and max2-1 ARA7 (max2-1-73; c, d) seedlings grown on acetone (ac) or GR24 plates (bars, 10 μm). Suggested ‘trails’ of endosomal movement are shown with three colored arrows in the inserts: frame 1, green; frame 5, red; and frame 9, blue. (e) Average velocity of ARA7-labeled endosomes in Col-0 and max2-1-73 seedlings grown on acetone (open columns) or GR24 (closed columns) plates. A total of 20–30 cells from 10 plants were examined per replicate; three replicates were included. The same lowercase letters above the bars indicate statistically nonsignificant differences between means by one-way ANOVA pairwise Student's t-test ( 0.05). Means are significantly different between hormone treatment (GR24) and control (ac) as determined by ANOVA pairwise Student's t-test (*,  0.05); error bars ± 1 SE.

To examine whether GR24's effect on endosomal movement is MAX2-dependent, we generated a max2-1 mutant expressing ARA7-GFP under a 35S promoter and determined endosomal movement in epidermal cells in the primary root elongation zone in this line. The endosomal velocity in the max2-1 epidermal cell was similar to that in the WT (Fig. 3a,c,e). However, unlike the WT, application of GR24 to this line did not result in induction of endosomal movement relative to the acetone control (Fig. 3c–e). Together, the results suggest that GR24 induction of endosomal movement in the epidermal cells of the primary root elongation zone is dependent on MAX2.

SLs affect actin filament architecture in the WT but not in max2-1

F-actin plays a major role in vesicle trafficking in the cell, including vesicles that are involved in PIN2 recycling in the root epidermal and cortical cells (e.g. Geldner et al., 2001; Kleine-Vehn et al., 2008; Lanza et al., 2012; Nagawa et al., 2012). We therefore examined whether the effect of SLs on PIN2 localization in the PM of epidermal cells in the primary root elongation zone is associated with changes in actin architecture in these cells. Actin filaments were labeled with TALIN-GFP (Kost et al., 1998) and observed under a confocal microscope. Actin bundling was observed as a superposition of Z optical sections from the epidermal layer of the roots. Skewness, a statistical parameter of the extent of actin filament bundling in a population of cells (Higaki et al., 2010), was used to quantify the GR24-induced changes in actin architecture.

In WT seedlings grown on acetone and GR24 plates, a reduction was detected in actin filament bundling in the epidermal cells of the root tip elongation zone in comparison to the acetone control, as evidenced by an increased proportion of thin fibrils and a decreased proportion of thick fibrils in the GR24-treated roots, and measured as a significant reduction in skewness following GR24 treatment compared with the acetone control (Fig. 4a,b,e). As TALIN-GFP has been previously shown to cause defects in actin organization (Ketelaar et al., 2004), we examined GR24's effect on actin architecture in a fABD2-GFP line, which has no detectable adverse effects on actin dynamics or plant morphology (Sheahan et al., 2004). Similar results of reduced skewness upon GR24 treatment in comparison to controls were obtained (Fig. S3).

Figure 4.

Effect of GR24 (3 × 10−6 M) treatments on the skewness of TALIN-GFP F-actin in the epidermal cells of the primary root elongation zone in 35S::TALIN-GFP Arabidopsis thaliana seedlings grown on acetone or GR24 plates (72–96 h postgermination). (a–d) TALIN-GFP signal in Col-0 (a, c) or max2-1 mutant (max2-1-28; b, d) grown on acetone (ac) or GR24 plates (bars, 20 μm). (e) Skewness of F-actin in Col-0 or max2-1-28 mutant seedlings grown on acetone (open columns) or GR24 (closed columns) plates. A total of 20–30 cells from 10 plants were examined per replicate; three replicates were included. The same lowercase letters above the bars indicate statistically nonsignificant differences between means by one-way ANOVA pairwise Student's t-test ( 0.05). Means are significantly different between hormone treatment (GR24) and control (ac) as determined by ANOVA pairwise Student's t-test (*,  0.05); error bars represent ± 1 SE.

To examine whether GR24's effect on actin bundling is MAX2-dependent, we generated a max2-1 mutant that expresses TALIN–GFP under the 35S promoter and determined the amount of actin bundling in epidermal cells in the primary root elongation zone. The skewness in max2-1 epidermal cells was similar to that in the WT (Fig. 4a,c,e). Unlike the WT, however, application of GR24 to this line did not reduce actin filament bundling in comparison to the acetone control (Fig. 4c–e). Together, the results suggest that bundling of actin filaments in the epidermal cells of the primary root elongation zone is reduced upon GR24 treatment in a MAX2-dependent fashion.

To confirm that the effect of SLs on actin is related to debundling, we examined F-actin stability, as actin debundling has been shown to be associated with its destabilization (Staiger & Blanchoin, 2006). We tested the effect of LatB, which leads to depolymerization of actin filaments (Spector et al., 1983), on actin architecture in the epidermal cells of the primary root elongation zone. LatB treatment of seedlings grown on GR24 plates resulted in increased depolymerization of actin in comparison to the acetone control. This was evident as a reduced number of filaments in the GR24-grown seedlings following LatB treatment in comparison to the LatB-treated acetone control seedlings (Fig. 5). Thus, GR24 treatment seems to reduce F-actin bundling, which in turn would enhance actin destabilization and thus actin susceptibility to depolymerization by LatB.

Figure 5.

Sensitivity of TALIN-GFP F-actin in the epidermal cells of the primary root elongation zone to latrunculin B (LatB) treatment in 35S::TALIN-GFP Arabidopsis thaliana seedlings grown on acetone or GR24 plates (72–96 h postgermination). (a–d) TALIN-GFP signal in roots of Col-0 seedlings grown on acetone (ac; a, b) or GR24 (c, d) plates (bars, 20 μm) and treated (b, d) or not treated (a, c) with LatB.

SLs affect actin dynamics in the WT but not in max2-1

As our results showed that SLs affect actin filament bundling, and because reduced actin filament bundling is associated with increased actin filament dynamics (Staiger & Blanchoin, 2006), we examined the effect of GR24 on the dynamics of actin filaments in the root epidermal cells. Actin dynamics was observed as superpositions of a single optical section taken every 3.2 s from the epidermal layer of the roots. Flow, a statistical parameter describing the extent of actin filament movement in a population of cells (Higaki et al., 2010), was used to quantify the GR24-induced changes in actin dynamics.

In seedlings grown on acetone and GR24 plates (72–96 h postgermination, 3 μM GR24; Fig. S1), induction of actin filament flow, quantified as an increase in F-actin average and median velocity, was detected in the epidermal cells of the root tip elongation zone in comparison to the acetone control (Fig. 6a,b,e,f, Videos S3, S4). This was evident in the overlay of frame 1 (green) on frame 5 (+16 s, red), and on frame 9 (+29 s, blue) of TALIN-GFP signal in the roots following treatment, as an increased proportion of green, red or blue filaments in the GR24-treated roots (Fig. 6a,b). As increased actin dynamics has been shown to be a result of reduced actin bundling (Staiger & Blanchoin, 2006), the increased actin dynamics conferred by GR24 is in agreement with the reduced skewness observed under these conditions.

Figure 6.

Effect of GR24 (3 × 10−6 M) treatments on the dynamics of TALIN-GFP F-actin in the epidermal cells of the primary root elongation zone in 35S::TALIN-GFP Arabidopsis thaliana seedlings grown on acetone or GR24 plates (72–96 h postgermination). (a–d) Overlay of frame 1 (green) on frame 5 (+16 s, red), and on frame 9 (+29 s, blue) of TALIN-GFP signal in Col-0 (a, b) or max2-1 mutant (max2-1-28; c, d) grown on acetone (ac) or GR24 plates (bars, 20 μm). (e) Average velocity value (excluding zero values, μm s−1) of F-actin in Col-0 or max2-1-28 seedlings grown on acetone (open columns) or GR24 (closed columns) plates. (f) Median velocity value (excluding zero values, μm s−1) of F-actin in Col-0 or max2-1-28 seedlings grown on acetone (open columns) or GR24 (closed columns) plates. A total of 20–30 cells from 10 plants were examined per replicate; three replicates were included. The same lowercase letters above the bars indicate statistically nonsignificant differences between means by one-way ANOVA pairwise Student's t-test ( 0.05). Means are significantly different between hormone treatment (GR24) and control (ac) as determined by ANOVA pairwise Student's t-test (*,  0.05); error bars ± 1 SE.

To examine whether the GR24 effect on actin dynamics is MAX2-dependent, we determined the effect of GR24 on actin filament dynamics in epidermal cells in the primary root elongation zone of the max2-1-28 35S:GFP-TALIN line. The velocity in max2-1 epidermal cells was similar to that in the WT (Fig. 6a,c,e,f), but unlike the WT, application of GR24 did not increase actin filament dynamics in comparison to the acetone control (Fig. 6c–f). Together, the results suggest that the increase in actin filament dynamics in the epidermal cells of the primary root elongation zone upon GR24 treatment is dependent on MAX2.

The effects of GR24 on actin architecture and dynamics were found to be similar to those induced by auxin (5 × 10−8 M IAA for 72–96 h postgermination; Fig. S4), in agreement with others (Waller et al., 2002; Holweg et al., 2004; Maisch & Nick, 2007; Lanza et al., 2012). Together, these results suggest that actin bundling is reduced and actin dynamics increased as a result of either GR24 or IAA application.

Mutants of components involved in trafficking and actin display an aberrant root hair elongation response to GR24 treatment in comparison to the WT

To further examine whether the effects of SLs on PIN2, vesicle trafficking and actin dynamics are linked to their physiological effect on root hair cells, we examined root hair elongation responses to GR24 in mutants of ACTIN2 (ACT2) (der1; Ringli et al., 2002), PIN2 (eir1; e.g. Luschnig et al., 1998; Abas et al., 2006) and the PIN-trafficking-associated protein TRANSPORT INHIBITOR RESISTANT3 (tir3; Desgagné-Penix et al., 2005). In agreement with Kapulnik et al. (2011a), the WT (Col-0 and C24) responded to GR24 treatment (3 μM) with an increase in root hair elongation, whereas max2-1 did not (Fig. 7). Root hair elongation in the eir1 mutants was sensitive to GR24 (Fig. 7). However, tir3 and most der1 mutant lines (except for der1-1) showed reduced sensitivity to GR24 relative to the WT (Fig. 7). On the other hand, root hair elongation in all of these mutants was sensitive to exogenous application of IAA, another positive regulator of root hair elongation (Pitts et al., 1998) (Fig. 7). The lack of a significant response of C24 (the genetic background of the der mutant line) to IAA might result from its long root hairs under control (water) conditions. The reduced sensitivity of these mutants to GR24 supports the hypothesis that GR24 leads to root hair elongation via regulation of vesicle trafficking associated with actin filament bundling and that, unlike IAA, the effect of GR24 on root hair elongation is at least partially dependent on these processes. However, the sensitivity of eir1 (PIN2) to GR24 suggests that its effect on root hair elongation is not solely dependent on PIN2.

Figure 7.

Lengths of root hairs and the effect of GR24 (3 × 10−6 M) (a) and IAA (5 × 10−8 M) (b) treatments on root hair length in Arabidopsis thaliana wildtype (WT; Col-0, C24), max2-1, eir1-1, der1-1, der1-2, der1-3, tir3-103, tir3-104 and tir3-105. Root hair length was measured at 48 h postgermination in seedlings grown on hormone (GR24 or IAA; closed columns) or control (acetone (ac) or H2O; open columns) plates (= 300). Means are significantly different between hormone treatments (GR24 or IAA) and controls (ac or H2O, respectively) in each examined line as determined by ANOVA pairwise Student's t-test (*,  0.05); error bars ± 1 SE.

Discussion

Herein we report on the cellular events that might underlie the root's response to SL, under GR24 treatment conditions that lead to root hair elongation. We show that this effect is associated with increased PIN2 polarization, increased PIN2 endocytosis, increased endosomal movement, reduced actin filament bundling and increased actin dynamics in the root epidermis.

We found that SLs increase polar localization of PIN2 in the PM of epidermal cells in the elongation zone of the root of WT, but not max2-1 plants. On the other hand, SLs did not affect PM localization of AUX1. Polar trafficking of AUX1 has been suggested to be distinct from that of PINs (Kleine-Vehn et al., 2006). Hence, SLs may affect the trafficking pathway associated with PIN polar localization, but not that associated with AUX1 polar localization.

Similar results suggesting the involvement of SLs in PIN1's presence in the PM were obtained in a study on SL-regulated shoot branching (Crawford et al., 2010; Shinohara et al., 2013). However, our observations of GR24 enhancement of PIN2 polar localization are in contrast to reports of the SLs' capacity to reduce auxin transport in Arabidopsis stems (Crawford et al., 2010), the higher PIN1 expression reported in max1 and max2 mutant plants when compared with their WT counterparts (Crawford et al., 2010), and the effect of GR24 on PIN1 depletion from the PM of xylem parenchyma cells in the stem (Crawford et al., 2010; Shinohara et al., 2013). These differences might be inherent to tissue-specific responses, that is, SLs may regulate PIN protein localization in the PM, but the trend of that regulation (i.e. up or down) may depend on the site of SL activity (e.g. xylem parenchyma cells in the stem or epidermal cells in the root).

On the other hand, the effect of GR24 described in the current study might not be a direct consequence of SL action. An indirect effect of GR24 as a consequence of reduced auxin concentrations has been suggested for PIN proteins in provascular tissue of the primary root (Ruyter-Spira et al., 2011). Most of the analyses in the current study were performed only after 72–96 h of continuous incubation with GR24. As a result, auxin fluxes could, during this time interval, change accordingly as a result of GR24 action elsewhere in the plant, and indirectly affect PIN protein dynamics in the cells under investigation in a MAX2-dependent manner. For example, a putative SL inhibitory effect on PIN lateralization in the root endodermis might lead to auxin accumulation in epidermal cells of the transition zone, the specific cells under investigation in the current study.

The increase in PIN2 signal in the apical PM of epidermal cells upon GR24 treatment of WT roots may result from either increased PIN2 gene expression or increased PIN2 trafficking to the apical PM, or both. Indeed, a threefold induction in PIN2 expression was detected in the WT but not in max2-1 upon GR24 treatment, suggesting that the increase in PIN2 signal under these conditions is at least partly a result of PIN2 expression induced by the GR24 treatment in a MAX2-dependent fashion.

The constitutive trafficking of PIN vesicles between the PM and the endosomes determines, to a large extent, the PINs' PM localization (e.g. Geldner et al., 2001; Dhonukshe et al., 2007, 2008b). GR24 treatment induced the accumulation of PIN2-containing BFA bodies in the WT but not max2-1, in line with the increased PIN2 expression in the WT but not in max2-1 under these conditions. In addition, in the WT, but not in max2-1, GR24 treatment led to an increase in the movement velocity of ARA7-labeled endosomes (Ueda et al., 2001, 2004), suggesting that SLs increase PIN2 endocytosis and movement of endosomes in a MAX2-dependent fashion. The increase in PIN2 endocytosis may have resulted from its increased amount in WT but not in max2 following GR24 treatment.

We also found SLs to reduce F-actin architecture and dynamics. Upon GR24 treatment, a marked reduction in actin bundling was evident in comparison to the control in the WT but not in max2-1. Notably, TALIN-bound actin has been shown to lead to bundling of actin filaments (Ketelaar et al., 2004; Nick et al., 2009). Indeed, our observations revealed bundling of F-actin under control conditions in the TALIN-GFP lines (in agreement with Ketelaar et al., 2004; although no defects were observed in root hair elongation under our experimental conditions). However, reduction in actin filament bundling by GR24 treatment was also present in the fABD2-GFP line; untreated fABD2-GFP lines show no detectable adverse effects on actin dynamics or plant morphology (Sheahan et al., 2004). In addition, the GR24-mediated reduction in F-actin bundling was associated with increased sensitivity to LatB. Taken together, it is suggested that GR24 leads to reduction of actin-fibril bundling and increased actin-fibril dynamics in a MAX2-dependent fashion.

Only a minor and insignificant reduction in PIN2 expression was evident in max2 compared with the WT. Moreover, max2 mutants showed WT-like levels and polarity of PIN2 and similar amounts of endocytosis, endosomal movement and actin architecture under control conditions. This suggests that the SL response is not obligatory for proper expression and localization of PIN2, and is in agreement with observations that show similar root responses (e.g. gravitropic response; Shinohara et al., 2013) for the WT and max2.

To further validate the involvement of PIN2, actin and PIN-associated trafficking in the root's response to GR24, we examined the SL responses in several mutants. These included der1 mutants that are defective in the ACT2 gene and display impaired root hair development (Ringli et al., 2002), and alleles of eir1 – PIN2 loss-of-function mutants that are agravitropic and show reduced sensitivity to ethylene and to internally generated auxin (e.g. Luschnig et al., 1998; Abas et al., 2006). Alleles of tir3 were also examined (Desgagné-Penix et al., 2005): BIG/TIR3 functions in polar auxin transport and endocytic trafficking and in the stabilization of PIN1 on the PM (Gil et al., 2001; Titapiwatanakun et al., 2009).

Most of the der1 and tir3 mutant lines displayed reduced sensitivity to GR24 with respect to root hair elongation. The reduced sensitivity of the mutants associated with endocytic trafficking and actin to GR24 supports the suggestion that GR24 treatment leads to a root response via regulation of vesicle trafficking associated with actin filament bundling. Moreover, our results with tir3 mutants are in agreement with Shinohara et al.'s (2013) findings demonstrating increased sensitivity to GR24 in tir3 with respect to shoot branching and root growth. By contrast, the sensitivity of eir1 (PIN2) to GR24 suggests that the effect of GR24 on root hair elongation is not solely dependent on PIN2. Thus, other mechanisms of root hair elongation may be employed by SLs, such as activation of additional hormonal pathways. These might include the ethylene pathway, as ethylene has been suggested to form a crosstalk junction between the SL and auxin pathways (Kapulnik et al., 2011b; Koltai, 2011). Another possibility is the GA pathway, as both GA-DELLA (Jiang et al., 2007) and SL (Mayzlish-Gati et al., 2010) pathways have been shown to regulate root hair response under conditions of phosphate starvation.

On the other hand, all of the examined mutants associated with endocytic trafficking and actin displayed WT-like sensitivity to auxin, suggesting that the effect of auxin on root hair elongation does not require full functionality of these components. However, sensitivity of one of the der lines (der1-1) to GR24 might indicate the involvement of other actin proteins in this process. Indeed, a point mutation in the gene for ACT2 resulted in remodeling of the actin cytoskeleton and only partially depolarized PIN2 localization (Lanza et al., 2012), and overexpression of ACT7 and ectopic expression of ACT1 rescued the root hair elongation defects of the act2-1 mutant (Gilliland et al., 2002), suggesting some degree of actin protein functional redundancy.

To summarize, we found that the root response to GR24, measured as root hair elongation, is associated with several cellular events that are dependent on MAX2 SL signaling: increased PIN2 polarization in the PM; increased level of PIN2 expression and internalization; increased endosomal movement; and reduced bundling and increased dynamics of F-actin. The effect of GR24 on endosomal trafficking and F-actin architecture and dynamics may be correlated to the changes in PIN2 trafficking and its PM localization. In agreement, other studies showed that endocytosis, endosomes and F-actin are required for acquisition of PIN proteins' polar localization (Geldner et al., 2001; Dhonukshe et al., 2008a,b; Kleine-Vehn et al., 2008; Nick et al., 2009; Nick 2010; Lin et al., 2012; Nagawa et al., 2012).

The suggested mechanism of SL regulation of PIN PM localization seems to be common to other plant hormone responses. IAA affects actin bundling and dynamics in association with PIN proteins' polar localization (e.g. Holweg et al., 2004; Maisch & Nick, 2007; Dhonukshe et al., 2008b; Nick et al., 2009; Lanza et al., 2012; Lin et al., 2012). Similarly, brassinosteroids stimulate plant tropism through modulation of polar auxin transport, associated with differential transcription of PIN genes and enhancement of the polar accumulation of PIN proteins, also involving induction of actin debundling (e.g. Li et al., 2005; Vieten et al., 2005; Nick et al., 2009; Lanza et al., 2012).

However, unlike IAA, the GR24 effect on root hair elongation is at least partially dependent on these processes, and may require actin debundling and enhanced endocytic trafficking for execution of root hair elongation. This dependency of SL-induced root responses on cellular trafficking and intracellular auxin transport suggests a noncell autonomous effect of SLs on roots, in line with Koren et al.'s (2013) findings, which showed that endodermal expression of SL-MAX2 signaling (in the max2-1 mutant) is sufficient to regulate the root response to SLs.

Acknowledgements

We are grateful to Dr Einat Sadot (ARO, Volcani Center, Israel) for consultation during the experiments. We are grateful to Prof. Binne Zwanenburg (Radboud University Nijmegen, the Netherlands) and Prof. Koichi Yoneyama (Utsunomiya University, Japan) for providing us with the SL analog GR24. We thank Dana Laufer and Dikla Koren for technical help.

Ancillary