Indole-3-butyric acid induces lateral root formation via peroxisome-derived indole-3-acetic acid and nitric oxide



  • Controlled plant growth requires regulation through a variety of signaling molecules, including steroids, peptides, radicals of oxygen and nitrogen, as well as the ‘classical’ phytohormone groups. Auxin is critical for the control of plant growth and also orchestrates many developmental processes, such as the formation of new roots. It modulates root architecture both slowly, through actions at the transcriptional level and, more rapidly, by mechanisms targeting primarily plasma membrane sensory systems and intracellular signaling pathways. The latter reactions use several second messengers, including Ca2+, nitric oxide (NO) and reactive oxygen species (ROS).
  • Here, we investigated the different roles of two auxins, the major auxin indole-3-acetic acid (IAA) and another endogenous auxin indole-3-butyric acid (IBA), in the lateral root formation process of Arabidopsis and maize. This was mainly analyzed by different types of fluorescence microscopy and inhibitors of NO production.
  • This study revealed that peroxisomal IBA to IAA conversion is followed by peroxisomal NO, which is important for IBA-induced lateral root formation.
  • We conclude that peroxisomal NO emerges as a new player in auxin-induced root organogenesis. In particular, the spatially and temporally coordinated release of NO and IAA from peroxisomes is behind the strong promotion of lateral root formation via IBA.


Auxin is critical for plant growth and orchestrates almost all developmental processes (Woodward & Bartel, 2005a; Aloni et al., 2006; Teale et al., 2006). Auxin is a collective name for several molecules with a wide spectrum of growth-promoting activities. They are synthesized not only in plants (Woodward & Bartel, 2005a), but also in fungi (Ulrich, 1960; Splivallo et al., 2009; Rao et al., 2010) and bacteria (Bianco et al., 2006a,b; Spaepen et al., 2007; Bai et al., 2012), and can also be produced synthetically (e.g. 1-napthylacetic acid, NAA). In a narrow sense, the term auxin is often used for the most potent native compound, namely indole-3-acetic acid (IAA).

Nearly every developmental stage (embryonic and postembryonic) and every growth process (formation of lateral organs, growth of leaves) of a plant is affected by IAA directly (Heisler et al., 2005; Teale et al., 2005, 2006) or indirectly by secondary induced signaling molecules (Pagnussat et al., 2004). Furthermore, phototropism and gravitropism are closely connected to auxin transport and action (Friml et al., 2002; Li et al., 2005; Wan et al., 2012).

A central point in the understanding of all the facets of auxin activity is the identification of specific receptors and interaction partners at the cellular level. Around 20 yr ago, the first auxin binding protein (ABP1) was found (Hesse et al., 1989). Located at the cellular periphery and being responsible for fast reaction to exogenous auxin (Dahlke et al., 2010; Tromas et al., 2010), ABP1 participates in extracellular auxin-mediated cell elongation by regulating endocytosis (Robert et al., 2010) and activating ROP signaling (Xu et al., 2010), as well as activating ion channels and H+-ATPases at the plasma membrane (Felle et al., 1991; Rück et al., 1993; Maurel et al., 1994; Christian et al., 2006). The proton efflux acidifies and loosens the cell walls, thus enabling turgor-driven cell expansion (e.g. Christian et al., 2006). In addition, an intracellular auxin receptor has been found. This receptor controls the degradation of a subset of transcription factors and thereby modulates transcriptional auxin responses (Dharmasiri et al., 2005; Kepinski & Leyser, 2005). However, many aspects of auxin action are not understood. For instance, the additional dependence of auxin on downstream signals, such as reactive oxygen species (ROS) and nitric oxide (NO), in the regulation of root organogenesis and architecture is still puzzling (Joo et al., 2001, 2005; Pagnussat et al., 2003, 2004; Hu et al., 2005; Yadav et al., 2013).

NO is a well-known stress signaling molecule that plays a crucial role during plant defense against pathogens (e.g. Bolwell et al., 1999). Recently, a more fundamental role in basic growth processes has been discussed. Rather surprisingly, NO has been reported to function as a downstream signaling molecule of auxin-induced lateral and adventitious root formation (Pagnussat et al., 2003, 2004). Moreover, gravistimulation of roots not only induces auxin accumulation at the lower root flank, but also of NO (Joo et al., 2001, 2005; Hu et al., 2005), and a reduction in the NO level inhibits gravitropic bending of gravistimulated root apices (Joo et al., 2005).

An interesting avenue that may help to understand NO-dependent auxin signaling is to take a closer look at the effects of indole-3-butyric acid (IBA), which is a naturally occurring auxin known from several plant species (Epstein & Ludwig-Müller, 1993). Unlike IAA, it is believed that IBA has nearly no auxin-typical transcriptional capacity on its own (e.g. Oono et al., 1998). IBA is considered to serve mainly as a transport and storage form of IAA (Bartel, 1997; Zolman et al., 2000). IAA can be converted to IBA (Ludwig-Müller & Epstein, 1991) and IBA is converted back to IAA in a peroxisomal β-oxidation-like process (Zolman et al., 2000). In plants, β-oxidation is localized in the peroxisomes (Gerhardt, 1992; Kindl, 1993). Mutants with defects of peroxisomal biosynthesis or β-oxidation are resistant to externally applied IBA (Zolman et al., 2000, 2001a,b; Zolman & Bartel, 2004). Several mutants of peroxisomal enzymes appear to be solely linked to β-oxidation-like processing of IBA to IAA (Zolman et al., 2008; Strader et al., 2011). Moreover, mutants defective in β-oxidation are impaired in IBA to IAA conversion (Strader et al., 2010). These genetic approaches indicate that IBA to IAA conversion could be the explanation for the unique biological activity of IBA.

In auxin bioassays, IBA often shows only weak activity (Woodward et al., 2005a), but there are exceptions. IBA, but not IAA, can efficiently induce adventitious roots in Arabidopsis (Ludwig-Müller et al., 2005). Moreover, it has emerged that IBA can promote lateral root formation (LRF) independently of IAA (Ludwig-Müller, 2000; Chhun et al., 2003; Strader et al., 2010, 2011). Altogether, several pieces of evidence indicate that the straightforward scenario of IBA being only a precursor of the active auxin IAA is not sufficient to explain its activity features. It has been shown that IBA, much like IAA, uses NO as a downstream signal to induce lateral roots (Kolbert & Erdei, 2008; Kolbert et al., 2008; Yadav et al., 2013). NO signal transduction cascades relevant for root primordial formation include cyclic GMP, phospholipase D, phosphatidic acid, ROS and PIN transporters of auxin (Pagnussat et al., 2003; Lanteri et al., 2008; Li et al., 2011; Bai et al., 2012; Li & Jia, 2013).

Here, we show the effects of IAA and IBA in different root development model systems using wild-type and mutant plant lines of maize and Arabidopsis. Particular emphasis is paid to NO-mediated IAA and IBA effects on root growth and development of root systems. We use the protein degradation inhibitor Terfestatin A as an additional experimental tool. This inhibitor prevents the auxin-specific changes of gene repression (Yamazoe et al., 2005) and enables discrimination between transcriptional and redox-based activities of auxin.

IBA shows paradoxical behavior relative to other auxins, such as IAA, NAA or 2,4-dichlorophenoxyacetic acid (2,4-D). It displays only weak auxin activity in root growth inhibition assays (Woodward & Bartel, 2005a), but its ability to induce lateral and adventitious roots is comparable with that of other auxins (Nordström et al., 1991; Zolman et al., 2000; Ludwig-Müller et al., 2005). The lateral rootless 1 (lrt1) Oryza sativa (rice) mutant is less sensitive to auxin (IAA, IBA and 2,4-D) with respect to root elongation. In addition, the only auxin that can restore lateral root initiation in the mutant is IBA (Chhun et al., 2003). Overall, this suggests that, at least in lateral root initiation, IBA performs an essential role, which cannot be replaced by IAA. Our present data show that the spatially and temporally coordinated release of IAA and NO via peroxisomes is the critical process, which makes IBA a unique player in lateral root initiation.

Materials and Methods

Plant material and inhibitor treatments

Maize grains (Zea mays L.) of wild-type and the lrt1 mutant were soaked for 6 h and germinated on well-moistened rolls of filter paper for 4 d in the dark at room temperature. Seedlings with straight primary roots, 50–70 mm in length (wild-type, lrt1), were selected for auxin treatments. For pharmacological experiments, root apices were submerged in appropriate solutions in the dark at room temperature. For the treatments of roots with auxin, an effective working concentration of 10 μM IAA or IBA was prepared immediately before submerging root apices for 2 h.

Arabidopsis seeds were surface sterilized and placed on half-strength Murashige & Skoog (1962) (MS) culture medium without vitamins and containing 1% sucrose (1.5% for β-oxidation-defective mutants and associated wild-type controls) which was solidified with 0.8% phytagel. Plates with seeds were stored at 4°C for 48 h to break dormancy and then vertically mounted under continuous yellow light for 3–4 d.

Arabidopsis mutants pex5-1 pex7-1, pxa1-1 and pex6 (originally designated B11 by Zolman et al., 2000), with defects in β-oxidation and showing IBA insensitivity (Zolman et al., 2001a; Zolman & Bartel, 2004; Woodward & Bartel, 2005b), and the mutant noa1 with impaired NO production (originally designated Atnos1 by Guo et al., 2003), were used for our experiments.

For microscopy, 3–4-d-old seedlings were transferred to microscopic slides, which were placed in thin chambers made of coverslips. The chambers were filled with half-strength MS medium, but without phytagel, and placed in sterile glass cuvettes containing the medium at a level that reached the open lower end of the chambers. This allowed free exchange of medium to take place between the chambers and the cuvette. Seedlings were grown in a vertical position under continuous yellow light for up to 24 h. During this period, the seedlings stabilized their root growth and generated new root hairs in the liquid medium. In vivo monitoring of peroxisomes was performed with peroxisome targeting signal 1-green fluorescent protein (PTS1-GFP) (see Woodward & Bartel, 2005b).

Inhibitors and chemicals for treatments (Terfestatin A, IAA, IBA, cPTIO (2-(4-carboxy-2-phenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide) and SNAP (S-nitroso-N-acetylpenicillamine)) were added to half-strength MS culture medium. Terfestatin A (described in Yamazoe et al., 2005) was a gift from H. Nozaki (Okayama University of Science, Okayama City). Unless stated otherwise, all chemicals were obtained from Boehringer-Mannheim (Mannheim, Germany), MBI Fermentas (Darmstadt, Germany), Merck (Darmstadt, Germany), Roth (Darmstadt, Germany) or Sigma (Munich, Germany).

NO labeling and measurements

The detection of NO was achieved by the specific fluorescent probes diamino-rhodamine-4M AM (DAR-4M AM; Calbiochem, Billerica, MA, USA), 4,5-diamino-fluorescein diacetate (DAF-2 DA; Calbiochem) and 4-methoxy-2-(1H-naphtho(2,3-d)imidazol-2-yl)phenol (MNIP) with Cu(II) (MNIP-Cu; Yadav et al., 2013).

DAR-4M AM is a cell-permeable, NO-specific dye. Intracellular esterases release DAR-4M, which reacts with NO in the presence of O2, resulting in a strong fluorescent triazolo-rhodamine analog (DAR-4M T). The DAR-4M T fluorescence intensity is pH independent.

Roots were incubated in the dark for 30 min at 25°C in 10 mM Tris/HCl (pH 6.5) containing 10 μM dye added from a 10 mM stock in dimethylsulfoxide (DMSO) (Sigma). The roots were then washed three times in fresh buffer to remove excess fluorophore, mounted in buffer on microscopic slides and then examined under a confocal laser scanning microscope (see later).

For semiquantitative measurement of NO in extracts, roots were incubated in the dark for 30 min at 25°C in 10 mM Tris/HCl (pH 6.5) containing 10 μM of the specific fluorescent probe DAF-2 DA or with 50 μM MNIP-Cu. Similar to DAR-4M AM, DAF-2 DA is a cell-permeable dye which, after the action of intracellular esterases, reacts with NO in the presence of O2, resulting in a strong fluorescent DAF-2 triazole (DAF-2T). The samples were frozen in liquid nitrogen, homogenized in 1 ml of buffer, incubated for 15 min and spun down. The resulting supernatant was collected in a 1-ml cuvette and measured with a TriStar LB 941 plate-reader (Berthold Technologies, Bad Wildbad, Germany) at 488 nm excitation and 515 nm emission. For semiquantitative measurement of NO with MNIP-Cu, crystals of MNIP were dissolved in DMSO to obtain a 10 mM stock, which was stored at −20°C. MNIP-Cu was always prepared fresh from MNIP just before use. MNIP stock (10 mM) was diluted to 1 mM with DMSO and 20 μl of 50 mM of aqueous copper sulfate was added to 1 ml of MNIP solution. The mixture was stirred for 5 min at room temperature, resulting in the formation of a stable yellow-colored solution of MNIP-Cu.

Seedlings were treated with 50 μM MNIP-Cu. For the measurement of NO fluorescence from the MNIP–NO complex, samples were frozen in liquid nitrogen, homogenized in 1 ml of buffer and spun down. The resulting supernatant was collected in a 1-ml cuvette and measured with a TriStar LB 941 plate-reader at 385 nm excitation and 420 nm emission. For each of the treatments, incubation with 1 mM cPTIO (NO scavenger) for 30 min, followed by co-incubation with MNIP-Cu for 10 min, was also undertaken.

Measurements of the IBA to IAA conversion

The conversion of IBA to IAA was measured by gas chromatography-mass spectrometry (GC-MS) using heavy labeled isotopes. The plant material for studying the conversion of IBA to IAA was grown aseptically on plates containing half-strength MS medium with 1% agar and 1.5% saccharose for better germination of the mutants. To produce sufficient material for the β-oxidation-defective mutants, six plants per plate were grown in a temperature-controlled chamber for 5 wk vertically under a light : dark cycle (16 h : 8 h) and 23°C. For the IBA to IAA conversion experiment, the plants were carefully removed from the agar surface and incubated in 100 mM MES buffer, pH 6, containing [indole-13C8,15N1]-IBA (a generous gift from Dr Jerry D. Cohen, University of Minnesota, Minneapolis, MN, USA; Barkawi et al., 2008) for 16 h in the dark under continuous slow shaking (50 rpm). The plants were removed from the buffer, washed several times with distilled water and frozen in liquid nitrogen before extraction. Duplicate measurements were performed on wild-type, pxa1 and noa1 plants. The concentration of labeled IBA was calculated in the tissue as nanograms per milligram fresh weight. As a second heavy isotope-labeled standard for IBA was not available, this might have caused slight underestimation in the calculation of IBA, and thus may have resulted in slightly overestimated IAA formation rates. For IAA determination, [indole-13C6]-IAA (Cambridge Isotope Laboratories, Andover, MA, USA) was added as an internal standard. IAA and IBA were analyzed as described in Jentschel et al. (2007). The samples were methylated with diazomethane (Cohen, 1984) and analyzed on a Varian Saturn GC-MS system (Varian Inc., Darmstadt, Germany) as described in Campanella et al. (2003).

Microscopy, image processing and cytofluorimetric measurements

Confocal microscopy was carried out with either a Leica TCS 4D or a Zeiss LSM 510 Meta (Wetzlar, Germany). Both were equipped with an argon–mixed gas laser and excitation/emission filter combinations for fluorescein isothiocyanate (FITC)/GFP and rhodamine/tetramethyl rhodamine isothiocyanate (TRITC)/FM dye detection. Samples were examined using × 40 oil immersion and × 63 water immersion objectives. The fluorochrome DAR-4M T was excited by the 488 nm laser line and emission was filtered between 620 and 710 nm. Serial confocal optical sections were taken at different step sizes, ranging from 0.5 to 2 μm. Projections of serial confocal sections and final image processing were performed with open source software Image-J (

For growth and curvature measurements, seedlings were observed directly on the Petri dishes with a binocular (ICS Leica, Wetzlar, Germany) using Discus image software (Carl H. Hilgers, Königswinter, Germany), or the Petri dishes were placed on a standard PC scanner. Analysis and measurements were made with Image-J.


IBA promotes NO pathways for LRF in Arabidopsis and Zea mays

It has been proposed previously that IBA induces LRF by the promotion of NO production (Kolbert et al., 2008). This observation and the behavior of the maize lrt1 mutant to IBA prompted us to perform a lateral root growth assay with wild-type and mutant Arabidopsis thaliana under different conditions (Fig. 1). Both IAA and IBA promoted strong LRF in Arabidopsis wild-type roots at a concentration of 1 μM. The root-specific auxin signaling inhibitor Terfestatin A reduced LRF of wild-type plants in the presence of either IAA or IBA (see also Yamazoe et al., 2005), although the effect on IBA-induced LRF was much less than that on IAA-induced LRF (Fig. 1a).

Figure 1.

Auxin-induced lateral root formation in Arabidopsis. The mean average of emerging lateral roots from 25 plants (7 d after germination (DAG)) per sample was measured and the number of lateral roots per centimeter of primary root was calculated. The nitric oxide (NO) donor SNAP (S-nitroso-N-acetylpenicillamine), NO scavenger cPTIO (2-(4-carboxy-2-phenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide) and auxin signaling inhibitor Terfestatin A (Trf A) were used at a concentration of 10 μM and the auxins were used at a concentration of 1 μM. Wild-type (WT) measurements are shown in (a). The measurements for the peroxisomal/developmental mutants pex5-1 pex7-1 and pxa1-1 and the NO-deficient mutant noa1 are displayed in (b). Error bars indicate ± SD.

β-Oxidation-defective mutants are insensitive to IBA (see, for example, Zolman et al., 2000) and have a very low rate of LRF. Consequently, LRF could be rescued by the addition of IAA, but not IBA (Fig. 1b). To test whether both auxins are dependent on NO for their ability to induce lateral roots, the mutant noa1 and NO level-altering chemicals were used. The noa1 mutant has a strongly diminished ability to generate endogenous NO after exposure to different environmental changes (Guo et al., 2003; Guo & Crawford, 2005). Both auxins were less effective in promoting LRF in the noa1 mutant. Simultaneous treatment of both auxins with the NO scavenger cPTIO led to a similar strong decreased rate of LRF. By contrast, a high NO level induced by treatment with the NO donor SNAP increased the number of lateral roots in all tested plant lines (Fig. 1b).

The maize mutant lrt1 showed a similar behavior to Arabidopsis with regard to LRF. It was insensitive to IAA, whereas treatment with 10 μM IBA resulted in a qualitative rescue of the lateral root phenotype (Table 1). The IBA-induced lateral roots were stunted and grouped closely together, giving rise to small clusters with several lateral roots (Supporting Information Fig. S1). Interestingly, a co-treatment with the NO scavenger cPTIO inhibited the rescuing effects of IBA. Moreover a co-treatment with the NO donor SNAP increased the quality of the rescue to a near wild-type-like phenotype (Fig. S1). Treatment with SNAP alone resulted in only a partial rescue, with c. 60% of all individual mutants showing stunted lateral roots, delayed by 1 or 2 d.

Table 1. Root phenotype of the maize mutant lateral rootless 1 (lrt1)
TreatmentRoot length (cm)Roots with lateral roots
  1. Maize mutant lrt1 (9 d after germination (DAG)) treated for 48 h with auxin (1 μM) and/or SNAP (S-nitroso-N-acetylpenicillamine) (10 μM). Primary root lengths are shown as averages ± SD, = 25 plants per treatment. IAA, indole-3-acetic acid; IBA, indole-3-butyric acid.

Mock (control)10.1 ± 0.3None
IAA (1 μM)9.8 ± 0.8None
IBA (1 μM)6.5 ± 0.522 from 25 (88%)
SNAP (10 μM)11.5 ± 315 from 25 (60%)
IBA (1 μM) + SNAP (10 μM)10.5 ± 1.425 from 25 (100%)

Transcriptional auxin activity

To better understand whether the activity of IBA on gene expression is dependent on β-oxidation, semi-quantitative reverse transcription-polymerase chain reactions (RT-PCRs) were performed to determine the transcript levels of IAA1 and IAA19 in the wild-type and mutants (Fig. 2). Both transcripts were strongly induced by treatment with 1 μM IAA or IBA. Three Arabidopsis mutants pex5-1 pex7-1, pxa1-1 and pex6 with defects in β-oxidation and IBA insensitivity were also tested. All mutants showed the induction of IAA1 and IAA19 transcripts on IAA treatment. However, no transcriptional activation was detectable after IBA treatment. This result indicates that a β-oxidation-like conversion from IBA to IAA is needed for an IBA-induced transcriptional activity of these chosen Aux/IAA genes.

Figure 2.

Testing of indole-3-butyric acid (IBA) for typical transcriptional auxin activity in Arabidopsis. Reverse transcription-polymerase chain reaction (RT-PCR) of auxin-inducible genes (IAA1 and IAA19) from cDNA of roots. Both transcripts are expressed in root apices after indole-3-acetic acid (IAA) treatment. Transcripts are detectable after treatment for 1 h with 1 μM auxin. System was adjusted with ubiquitin expression for semiquantitative comparison. IBA is able to induce transcripts, starting with a 1-h treatment time and a concentration of 1 μM. Comparison of transcriptional activity of IAA1 and IAA19 after 2 h of treatment with 1 μM of IAA or IBA in wild-type and pxa1-1, pex6 and pex 5-1 pex7-1 mutants. IBA fails to induce IAA1 and IAA19 transcripts in the IBA-resistant mutants.

IBA to IAA conversion promotes peroxisomal NO formation

The scenario of a conversion of inactive IBA to active IAA could explain the slower response of transcriptional activation and the weak inhibition of primary root growth on IBA treatment. Nevertheless, this scenario cannot explain the fact that IBA is effective in boosting LRF. Several lines of evidence indicate the ability of IAA to induce these roots by using NO as a downstream signal (Correa-Aragunde et al., 2004). We used the NO-specific fluorescent markers DAF-2 DA, DAR-4M AM and MNIP-Cu to test whether IBA induces NO in a manner comparable with IAA.

Both DAF and MNIP–NO fluorescence measurements of IAA- and IBA-treated Arabidopsis roots showed that IBA promoted NO production in a similar manner to IAA in wild-type seedlings (Fig. 3). Furthermore, IAA, but not IBA, was able to induce NO in the mutants pex5-1 pex7-1, pex6 and pxa1-1. Both auxins induced NO production in nia1nia2 and noa1, although at a lower level than in wild-type plants. Furthermore, treatments with the NO scavenger cPTIO showed that both auxins indeed induced NO itself. In vivo labeling with DAR-4M AM allowed the monitoring of the subcellular distribution of NO sources. In addition to a strong diffusible cytoplasmic signal near the plasma membrane, several strongly fluorescent moving spots were detectable (Fig. 4a). The peroxisomal marker PTS1-GFP showed nearly no co-localization with DAR-4M in untreated root cells. However, peroxisomes appeared as NO-accumulating organelles after 1 μM IAA treatment (Fig. 4b), and even as major NO-accumulating organelles after 1 μM IBA treatment (Fig. 4c,d).

Figure 3.

Auxin-induced nitric oxide (NO) formation in Arabidopsis. The relative fluorescence intensities (RFUs) of 4,5-diamino-fluorescein-2 triazole (DAF-2T and 4-methoxy-2-(1H-naphtho(2,3-d)imidazol-2-yl)phenol (MNIP)–NO reflect the level of NO formation. Bars represent the averages of extracts from 10 plants. Error bars indicate ± SD. Both methods, DAF (a) and MNIP (b), show that a treatment with indole-3-acetic acid (IAA) (1 μM) induced an increase in NO formation (compared with untreated control) in roots of all lines tested. Indole-3-butyric acid (IBA) (1 μM) induced increased NO formation only in wild-type, nia1nia2 and noa1, but not in the β-oxidation-defective mutants pex5-1 pex7-1, pex6 and pxa1-1. Plants co-incubated with one of the auxins and the NO scavenger cPTIO (2-(4-carboxy-2-phenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide) show no increase in detectable NO formation with MNIP–NO (b).

Figure 4.

Co-localization of nitric oxide (NO) with peroxisomes after indole-3-butyric acid (IBA) treatment in Arabidopsis. Roots from 2-wk-old seedlings of a 35s::PTS1-GFP line were labeled with 10 μM diamino-rhodamine-4M (DAR-4M) dye. Green color shows peroxisome targeting signal 1-green fluorescent protein (PTS1-GFP)-positive peroxisomes and red color shows the fluorescent DAR-4M T originating from the reaction of DAR-4M with NO. (a) Mock-treated root (dimethylsulfoxide (DMSO) control). DAR-4M T labeling pattern in root cells consists of a diffuse cytoplasmic signal often near the plasma membrane and a few moving intracellular spots. (b) Root treated for 2 h with indole-3-acetic acid (IAA) (1 μM). The pool of PTS1-GFP-marked peroxisomes shows a partial co-localization with DAR-4MT. (c) Root treated for 2 h with IBA (1 μM). An extremely strong co-localization of PTS1-GFP and DAF-2T is monitored after IBA treatment. Bars, 10 μm.

Measurements of the IBA to IAA conversion

We also measured the IBA to IAA conversion. Five-wk-old Arabidopsis plants were incubated with [indole-13C8,15N1]-IBA for 16 h and harvested. To determine the conversion rate by GC-MS, a labeled standard [indole-13C6]-IAA was added. The 13C8,15N1-label in the IAA peak was calculated on the basis of the standard. Then, based on this value, the conversion of IBA to IAA in each sample was measured (Table 2). Wild-type plants showed an average conversion rate of 55.8% of 13C8,15N1-IBA in this system; the IBA-insensitive mutant pxa1-1 had a rate of 8%. The noa1 mutant showed a comparable conversion rate (55.2%) to the wild-type.

Table 2. Indole-3-butyric acid (IBA) to indole-3-acetic acid (IAA) conversion in Arabidopsis
 Fresh weight (mg)Labeled IBA (ng mg−1 fresh weight)Labeled IAA (ng mg−1 fresh weight)Conversion IBA → IAA (%)
  1. Five-week-old Arabidopsis plants were incubated with [indole-13C8,15N1]-IBA for 16 h and harvested. The conversion rate was determined by GC-MS; a labeled standard [indole-13C6]-IAA was added. The 13C8,15N1-label in the IAA peak was calculated on the basis of the standard. Then, based on this value, the conversion of IBA to IAA in each sample was measured.

noa1 46.70.720.4155.2


IAA influences virtually every aspect of plant growth and development (Woodward & Bartel, 2005a; Aloni et al., 2006; Teale et al., 2006). In addition to the well-documented effect on the transcriptional level (Abel, 2007), other effects occur via signaling at the plasma membrane and endomembranes, via diverse downstream signals including NO (Mockaitis & Howell, 2000; Pagnussat et al., 2003, 2004; Lanteri et al., 2006, 2008; Bai et al., 2012; Chen et al., 2012; Li & Jia, 2013; Wang et al., 2013). Genetic analysis has uncovered many aspects of auxin biosynthesis and metabolism (reviewed in Woodward & Bartel, 2005a), and details of transcriptional regulation pathways of auxin-inducible genes (Sawa et al., 2002; Pufky et al., 2003; Cluis et al., 2004; Himanen et al., 2004). Studies to compare different auxins have shown fundamental differences in their function on plant growth, enabling a classification of active and inactive auxins. However, IBA shows conflicting results in different growth assays, making it difficult to classify it into one or the other group. In this work, we have conducted a set of experiments to examine the role of IBA, IAA and NO, and their connection within a possible signaling pathway in LRF.

IBA occurs naturally in different plant species and tissues (Epstein & Ludwig-Müller, 1993). It can be converted into IAA and vice versa, indicating a strong link between the metabolism of the two auxins (reviewed in Woodward & Bartel, 2005a). IBA shows the unique auxin ability of a polar cell-to-cell transport, but possibly by a different mechanism from IAA (Rashotte et al., 2003; Poupart et al., 2005). IBA and NO are known to promote the formation of lateral roots (Zolman et al., 2000; Correa-Aragunde et al., 2004) and adventitious roots (Nordström et al., 1991; Ludwig-Müller et al., 2005; Yadav et al., 2013). These are all typical auxin traits but, as we have shown here using β-oxidation-defective mutants of Arabidopsis, IBA does not show the transcriptional activity expected of an active auxin (Fig. 2).

The root growth of mutants with specific β-oxidation defects is insensitive to IBA, and our experiments show that IBA is unable to activate the transcription of Aux/IAA genes in such mutants (Fig. 1). Moreover, these mutants show very low IBA to IAA conversion rates (Table 2; Strader et al., 2010, 2011). All this suggests that IBA is converted to IAA in a process paralleling fatty acid β-oxidation (Zolman et al., 2000). β-Oxidation of fatty acids in plants resides solely in the peroxisomes (Gerhardt, 1992; Kindl, 1993). This supports the assumption that the conversion of IBA to IAA is indeed peroxisomal.

Terfestatin A is a root-specific auxin signaling inhibitor which disturbs the SCFTIR proteasome complex and disables the transcription of Aux/IAA response proteins and of the synthetic DR5-reporter (Yamazoe et al., 2005). The blocking of auxin-inducible gene transcription by Terfestatin A prevents the formation of lateral roots and neutralizes the positive LRF-promoting effect of IAA. However, the inhibition of SCFTIR-controlled gene expression has a weaker detrimental effect on IBA activity, pointing to the existence of a second LRF-promoting pathway in response to IBA (Fig. 2).

The fluorescent dye DAF-2 DA is often used as a marker of NO. However, its specificity is sometimes questioned because of the possibility that it detects not NO itself, but its oxidized products, such as N2O3 (Kojima et al., 1998). Indeed, the DAF–NO detection system seems to require oxygen to form fluorescent molecules (Arita et al., 2006). This may constrict the usage of the dye (not suitable under anoxia), but the generated fluorescence signal still reflects the level of NO produced after a stimulus. The fluorescence signal of DAF–NO reveals that IBA treatment generates a considerable amount of NO (Fig. 3). Importantly, this response was not found in the IBA-insensitive mutants used in this study, indicating that intact peroxisomal β-oxidation is necessary for this NO generation. This is consistent with the important finding of our present study that peroxisomes are a major NO source after IBA treatment (Fig. 4). We propose the scenario that the IBA to IAA conversion is responsible for the peroxisomal NO signal of IBA-exposed roots, which drives LRF. Consistent with this hypothesis, the impact of IBA on LRF is more susceptible to changes in the NO level than the impact elicited by IAA (Fig. 1). IBA is nearly inactive on LRF after inhibition of NO production by cPTIO, as well as in the Arabidopsis mutant noa1 with its diminished ability to produce NO (Guo et al., 2003; Guo & Crawford, 2005). Importantly, the IBA to IAA conversion is not significantly affected in noa1 (Table 2).

The gene NOA1 has been suggested to encode a plant NO-synthase (Guo et al., 2003; Guo & Crawford, 2005), even though it shares no sequence homologies to animal NO-synthases, making a role as an NO-synthase in plants implausible. However, the noa1 protein participates in NO production from l-arginine, which is a typical NO-synthase feature. Published results indicate that NOA1 is a plastidic GTPase, and the decreased NO production in the noa1 mutant is an indirect consequence of the mutation (Moreau et al., 2008; Gas et al., 2009). Nevertheless, the mutant shows disturbed NO formation and is, as such, a valid candidate for our purposes. A similar approach using nitrate reductase (NR) mutants (nia1nia2) has suggested that NO formation, which is required for IBA action, is produced by NR (Kolbert & Erdei, 2008; Kolbert et al., 2008). However, these studies failed to provide convincing results to link NR activity with IBA activity, apart from an IBA insensitivity of the NR mutant. Strikingly, the inhibition of NR prevents not only IBA-induced, but also IAA-induced, NO formation (Fig. 3; see also Hu et al., 2005). Both noa1 and nia1nia2 showed only weak NO formation after IBA treatment, but both mutants have, in addition to their reduced capability for constitutive NO formation, pleiotropic phenotypes in metabolism, growth, development and stress responses (Modolo et al., 2005; Moreau et al., 2008; Gas et al., 2009). This makes it difficult, in general, to distinguish between effects caused by impaired NO biosynthesis from those caused by metabolic alterations.

We have shown that peroxisomes are a major cellular site of NO production during the IBA to IAA conversion. Interestingly, other reports have also shown that peroxisomes act as potential NO sources in plants. For example, NO has been detected in peroxisomes of pollen tubes, in which the NO level at the pollen tube tip determines the rate and orientation of pollen tube growth (Prado et al., 2004). This suggests that one mode of the peroxisome signal transduction mechanism is via the ability to act as an NO source. In vitro and in vivo studies have proposed NO-synthase-like activities in peroxisomes from pea leaf (Barroso et al., 1999; del Rio et al., 2004) and Arabidopsis roots (Corpas et al., 2009). An additional nitrite-dependent pathway for NO production, involving xanthine oxidase, has been identified (Zhang et al., 1998). Under anaerobic conditions, when NO is important for the hypoxic acclimation of roots (Mugnai et al., 2012), nitrite can be reduced to NO by the purified peroxisomal enzyme xanthine oxidoreductase (XOR) (Godber et al., 2000). Moreover, a role for XOR in the production of NO was proposed after expression and inhibition studies on phosphate deficiency in cluster roots of white lupin (Wang et al., 2010).

These examples highlight that our knowledge about NO signal transduction pathways is still fragmentary. All the conclusions and most of our knowledge about NO signaling and the genetic modulation of endogenous NO levels in plants are of a correlative nature. Despite the as yet unresolved questions concerning the NO-producing enzyme systems, an additional line of evidence shows the involvement of NO in IBA-induced LRF. The lrt1 maize mutant is insensitive to the auxins IAA, NAA and 2,4-D in terms of lateral root initiation (Hochholdinger & Feix, 1998), but IBA, as well as NO, can induce lateral roots in this mutant, resulting in a qualitative rescue of the phenotype. The primary root shows local induction of stunted lateral roots. Both treatments are amplified if applied together, resulting in a wild-type-like phenotype. An inhibition of NO production by cPTIO abolishes the ability of IBA to induce lateral roots completely, indicating the NO dependence of IBA activity in maize, as in Arabidopsis.

A question which still remains to be answered is the role of endogenous IBA. Recent studies have indicated that the IBA to IAA conversion is relevant for undisturbed seedling development by feeding into internal active auxin pools (Strader et al., 2011). One phenotypic aspect of many mutants that cannot utilize IBA is a shortage of lateral roots (e.g. Zolman et al., 2001a,b; Wiszniewski et al., 2008; Strader et al., 2010, 2011). This could be an indicator that IBA to IAA conversion is necessary for LRF. Our present data show that the spatially and temporally coordinated release of NO and IAA is needed for IBA bioactivity. IBA to IAA peroxisomal conversion not only results in tightly regulated IAA synthesis, but also promotes concomitant NO formation. It can be expected that other processes also requiring IBA, and also involving NO, will be shown to rely on peroxisomal IAA–NO generation. Here, we can mention polarized tip growth of root hairs and root adaptation to salt and water stresses (Lombardo et al., 2006; Strader et al., 2010, 2011; Tognetti et al., 2010; Strader & Bartel, 2011). Future work is necessary to elaborate what kind of NO-producing enzyme systems are responsible for the NO formation induced by IBA, and whether NO-mediated signaling is also utilized in other processes in which IBA is active or essential.


The gifts of labeled IBA (Dr Jerry D. Cohen, University of Minnesota, Minneapolis, MN, USA) and of MNIP-Cu (Professor Bhatla, University of Delhi, Delhi, India) are gratefully acknowledged. We would like to thank Silvia Heinze (Technische Universität Dresden, Dresden, Germany) for technical assistance.