Medicago truncatula Gaertn. cv. Jemalong genotype A17 was used. Seeds were scarified with sandpaper and germinated on germination soil (S-jord, Hasselfors Garden AB, Sweden). Plants were grown in the greenhouse under a 15-h photoperiod and day/night temperatures of 22°C/19°C, with a light intensity of 70–300 μE m⁻² s⁻¹. Plants were watered once per week with water and once per week with Fåhraeus medium (Fåhraeus, 1957). For nodulation, plants were inoculated with Sinorhizobium meliloti 1021 at OD₆₀₀ = 0.1. For auxin quantifications, M. truncatula seeds were scarified with sandpaper, surface sterilized in 6% (w/v) sodium hypochlorite for 30 min, rinsed five times in sterile water, and grown on 15-cm diameter nutrient agar containing nitrogen-free Fåhraeus medium [0.9 mM CaCl₂, 0.5 mM MgSO₄, 0.7 mM KH₂PO₄, 0.8 mM Na₂HPO₄, 20 μM iron (II) citrate, supplemented with 0.1 mg/ml of each of the following microelements: MnCl₂, CuSO₄, ZnCl₂, HBO₃, and Na₂MnO₄]. Plants were grown in a growth room at 22°C with a 16-h day length and ∼150 μE m⁻² s⁻¹ light intensity, and nodules were harvested at 6 weeks postinoculation with S. meliloti strain 1021.

Datisca glomerata (C. Presl.) Baill. seeds were collected from greenhouse plants originating from plants growing in Vaca Hills (California, USA). Seeds were germinated on sand wetted with water; plantlets were transferred to pot soil (S-jord, Hasselfors Garden AB, Hasselfors, Sweden). Plants were grown in a growth chamber under a 15-h photoperiod and day/night temperatures of 25°C/19°C, with a light intensity of 60–100 μE m⁻² s⁻¹ and 65% relative humidity. When the plants had reached a height of ∼20 cm, they were transferred to bigger pots containing soil (S-jord, Hasselfors Garden AB) and crushed nodules harvested from older D. glomerata plants. Plants were watered twice per week, once with water and once with 1/4 strength Hoagland’s medium (Hoagland and Arnon, 1938) without a nitrogen source prepared as follows: 2 mM K₂SO₄, 2 mM KH₂PO₄, 2.5 mM CaSO₄, 2 mM MgSO₄, microelements (2.86 mg/l H₃BO₃, 1.81 mg/l MnCl₂ 4 H₂O, 0.22 mg/l ZnSO₄ 7 H₂O, 0.08 mg/l CuSO₄ 5 H₂O, 0.025 mg/l Na₂MnO₄ 2 H₂O, 0.025 mg/l CoCl₂ 6 H₂O), 100 ml/lchelated iron stock (20 mM FeSO₄ and 20 mM Na₂EDTA), pH 5.8. Nodules were harvested 6–8 weeks after infection.

Coriaria myrtifolia L. seeds collected in Jijel (Algeria) were kindly provided by Amir Ktari (University of Carthage, Tunis, Tunisia). They were vernalized on wetted sand at 7°C for 1 week before transfer to the greenhouse to germination soil. Growth conditions were the same as for D. glomerata.

Casuarina glauca Sieber ex Spreng. seeds were obtained from New Zealand Tree Seeds (nzseeds.co.nz), germinated and grown on germination soil for 3 months before being transferred to pots containing perlite/vermiculite (50/50, v/v) wetted with 1/4 strength Hoagland’s medium containing 10 mM KNO₃ (Hoagland and Arnon, 1938). Growth conditions were as described for D. glomerata. Roots were harvested when the plants were ∼20 cm high.

Glycine max (L.) Merr. cv. “Bragg” (soybean) seeds were obtained from Brett Ferguson (University of Queensland, Australia) and inoculated with Bradyrhizobium japonicum USDA110 (OD₆₀₀ = 0.1).

Lotus japonicus L. cv. Gifu seeds were purchased from the University of Miyazaki (Japan) and inoculated with Mesorhizobium loti MAFF303099 at OD₆₀₀ = 0.1. Soybean plants were grown in 2 l pots in sand, while L. japonicus was grown in Petri dishes (15 cm diameter) on 1/4 strength B&D medium (Broughton and Dilworth, 1971). Both species took 2 days to germinate and were inoculated 1 week postgermination. Nodules were collected 3 weeks postinoculation in both species. Temperature in the greenhouse was maintained at 25°C during the day and 20°C at night, with a natural day–light cycle.

Cucumis sativus L. (cucumber) seeds were purchased from “Seeds2freedom” (Australia) and grown in sand in rectangular 500-ml pots for 3 weeks prior to tissue harvest.

Begonia bowerae Ziesenh. cv. Cleopatra was a gift from the glasshouse facility manager at the Research School of Biology (Australian National University, Canberra, Australia). The adult begonia plant was grown in a 2-L pot in a generic soil mixture. Temperature in the greenhouse was maintained at 25°C during the day and 20°C at night, with a natural day light cycle.

Cicer arietinum L. (chickpea) seeds (kindly provided by Dr Angela Pattison, University of Sydney) were sterilized in 0.5% bleach for 30 min, coated with P-Pickel T fungicide (Nufarm Australia), and germinated on 1% agar plates over 2 days. Germinated C. arietinum seeds were transferred into autoclaved sand in individual 300-ml pots. After 1 week, each seedling was inoculated with 1 ml of Mesorhizobium ciceri strain CC1192 (OD₆₀₀ = 0.1). Nodules were harvested 6 weeks postinoculation, snap frozen, and stored until processing.

All rhizobial strains were maintained on TY medium. Escherichia coli strain TOP10 (Invitrogen, Stockholm, Sweden) was used for plasmid propagation and manipulation. Strains TOP10 and GJ23 helper (Van Haute et al., 1983) were used for triparental mating. E. coli bacteria were grown in Luria–Bertani medium (LB; Miller, 1972) at 37°C. Agrobacterium rhizogenes strains R1000 (Moore et al., 1979) and Arqua1 (Quandt et al., 1993) were used for hairy root transformation of M. truncatula, and LBA1334 (Offringa et al., 1986) and AR1193 (Stougaard et al., 1987) for hairy root transformation of D. glomerata. Agrobacteria were grown in YEB medium (Van Larebeke et al., 1977) at 28°C. M. truncatula plants were inoculated with the strain S. meliloti Sm1021 (Meade and Signer, 1977). D. glomerata plants were inoculated with the strain Candidatus Frankia datiscae Dg1, originating from nodules of Coriaria nepalensis from Pakistan (Mirza et al., 1994; Persson et al., 2011; Persson et al., 2015). Crushed D. glomerata nodules were used for the inoculation of root systems.

M. truncatula seeds were scarified with sandpaper, surface sterilized in 6% (w/v) sodium hypochlorite for 30 min, rinsed five times in sterile water, placed on 1/2 Hoagland’s 1% agar plates (Hoagland and Arnon, 1938), and incubated at 4°C overnight. Hoagland’s medium with a nitrogen source was prepared according to the following protocol: 1 mM KH₂PO₄, 2 mM MgSO₄, 5 mM Ca(NO₃)₂, 5 mM KNO₃, microelements (as above), 100 ml/l chelated iron stock (as above), pH 5.8. D. glomerata seeds were surface sterilized by gentle shaking in 70% (v/v) EtOH/0.05% (w/v) sodium dodecyl sulfate (SDS) for 5 min, followed by 20 min incubation in 2.5% (w/v) sodium hypochlorite containing 0.1% (w/v) SDS. Seeds were rinsed five times in sterile water, placed in an Eppendorf tube covered with sterile water, and incubated at 4°C for 1 week. Afterwards, seeds were placed on 1/2 Hoagland’s medium plates with 1% agar. When seedlings had grown to a length of ∼2 cm, they were transferred to 1/2 Hoagland’s 1% agar plates containing various concentrations of PAA or NAA for 28 days. For each auxin concentration, 9−12 plants were analyzed. Plants were grown in a growth room at 22°C with a 16-h day length and ∼150 μEm⁻² s⁻¹ light intensity. At the time of transfer to the hormone-containing plates, the root tip was marked on the plate. Root length was measured relative to the position of the root tip. Emerged and unemerged lateral roots were counted along the whole length of the root, including the 2 cm of the root that had formed at the time of transfer to hormone plates. For the counting of unemerged lateral root primordia under a light microscope, roots were bleached in 6% (w/v) sodium hypochlorite for 3–10 min, followed with rinsing in water and staining in 0.05% (w/v) methylene blue for 5 min. Roots were rinsed in water again before mounting on slides in water.

Destination vectors used were the pKGW-RR-MGW binary vector (Op den Camp et al., 2011) and an integration vector derived from pIV10 (Radutoiu et al., 2005). An A. rhizogenes strain R1000 carrying DR5-KGW-RR-MGW was available (Ilina et al., 2012). The vector pKGW-RR-MGW showed no background expression in M. truncatula hairy roots (E. Limpens, personal communication). However, when tested in D. glomerata in the present study, pKGW-GGRR showed extensive background GUS expression in hairy roots (data not shown). Therefore, the pIV10 vector (Radutoiu et al., 2005) was chosen for the preparation of a new DR5:GUS fusion construct. The promoter DR5 was PCR amplified from pBI101.3 (Jefferson et al., 1987) with the primers 5’-ACGGATCCGGTATCG-CAGCCCCCTTTTGTCTC-3’ and 5’-ACCTCGAGGGT-CTTG-CGGGGCTGCA-GG-3’. The BamHI/XhoI-digested proDR5 promoter fragment was subcloned in the BamHI/XhoI-digested entry vector pUC18-entry8 (Hornung et al., 2005). Recombination was confirmed with PCR with the forward gene-specific primer 5’-ACGGATCCGGT-ATCGCAGCCCCCTTTTGTCTC-3’ and the M13 reverse primer 5’-CAGGAAA-CAGCTGAC-3’. Then, DR5 was transferred from pUC18-entry8 into the integration vector pIV10 upstream of the GUS reporter gene ORF to yield the DR5:GUS fusion, using the Gateway cloning technology (Clontech, Mountain View, CA, USA). Recombination was confirmed with PCR with the forward gene-specific primer and the EcGUS reverse primer 5’-CCGGCTTTCTTGTAACGC-3’. The positive recombinants were sequenced (Eurofins, Ebersberg, Germany) using the EcGUS reverse primer. The pIV10 vector with the DR5:GUS construct was integrated into A. rhizogenes AR1193 T_L-DNA segment via triparental mating (Radutoiu et al., 2005). Selection of integration events was carried out on YEB agar medium containing 100 µg/ml ampicillin, 100 µg/ml spectinomycin, and 100 µg/ml rifampicin. Selected transformants were confirmed by colony PCR or liquid culture PCR with the forward gene-specific primer and the EcGUS reverse primer.

For DR5:GUS expression studies, axenically grown M. truncatula plants were transformed with A. rhizogenes R1000 as described by Boisson-Dernier et al. (2001). One week prior to inoculation with S. meliloti, plants were transferred to slanted 1% agar containing nitrogen-free Fåhraeus medium. Nodules were harvested at different stages of development.

D. glomerata transformation was performed using a protocol based on Markmann et al. (2008) with several modifications. Surface-sterilized D. glomerata seeds were germinated on Petri dishes with 1/4 strength Hoagland’s medium pH 5.8 containing 5 mM KNO₃, 10 g/l sucrose, and 0.8% micro agar (Duchefa, Haarlem, The Netherlands) and incubated in the dark for 7 days at 25°C. Afterwards, the plates were placed horizontally for seed germination under 16-h/8-h day/night for 20 days at 24°C. Twenty-day-old seedlings were transplanted onto new plates with full strength Murashige and Skoog (MS; Murashige and Skoog, 1962) medium pH 5.8 (Duchefa) containing 30 g/l sucrose, 0.5 μM NAA (Sigma-Aldrich), 5 μM N6-Benzylaminopurine (Duchefa), and 0.8% agar. The transplanted seedlings were grown under a growth cabinet with 16-h/8-h day/night at 25°C until they had thick hypocotyls, which took ∼3 months. A. rhizogenes transformation was performed only on seedlings with thick hypocotyls.

To prepare seedlings for transformation, they were incubated at 4°C in the dark for 4 days, before being pierced at the root–hypocotyl junction with a 0.4 mm × 20 mm syringe needle. Bacterial paste [A. rhizogenes strain AR1193 with pIV10 (empty vector) or with DR5:GUS grown on solid YEB medium (Van Larebeke et al., 1977) with antibiotics for 48 h] was directly applied onto the wound using a spatula. The treated seedlings were transferred to Petri dishes with 1/2 strength MS medium containing 100 µM acetosyringone, 19 g/l sucrose, and 0.8% micro agar (Duchefa). The plates were kept in the dark for 4 days for 25°C before being transferred to a growth cabinet with 16-h/8-h day/night at 25°C. After 2 weeks, transgenic roots appeared at the wound sites, and the plantlets were transferred to small pots containing a mixture of soil (Plugg och Såjord, Weibulls Trädgard AB, Hammenhog, Sweden), sand (Rådasand, Lidköping, Sweden; 0.4–0.6 mm), and autoclaved vermiculite. They were kept in a growth chamber with a 15-h photoperiod and day/night temperatures of 24°C/19°C, with a light intensity of 60–100 μE m⁻² s⁻¹ and 65% relative humidity under a lid for another 2 weeks before either harvesting of roots for GUS activity staining (36 root systems), or transfer to new pots containing the same soil mixture, but with D. glomerata nodules containing Candidatus Frankia datiscae Dg1 (Persson et al., 2015), which had been surface sterilized by incubation in 70% EtOH with 0.5% SDS for 5 min, followed by incubation in 1% sodium hypochlorite, 0.1% SDS for 7 min, followed by six washes with sterile ddH₂O, before being ground in ddH₂O and mixed with the soil. After infection, plants were watered once per week with 1/4 strength Hoagland’s medium pH 6.2 with doubled phosphate concentration containing 1 mM KNO₃, otherwise with deionized water. Infection with Frankia was repeated after 2 weeks by spreading a suspension of surface-sterilized ground nodules over the soil. Nodules were harvested after the leaves of the infected plants had turned dark green, which took 2–3 months. Altogether, 10 nodulated root systems were stained for GUS activity.

M. truncatula hairy roots were harvested before inoculation and at 3, 9, and 21 days postinoculation (dpi) with S. meliloti. M. truncatula hairy roots containing the fluorescent marker protein DsRed were selected with a SteREO Lumar.V12 stereomicroscope (Carl Zeiss, Jena, Germany) equipped with a 43 HE filter set (EX BP 550/25, EM BP 605/70). Afterwards, samples were fixed in 0.5% (w/v) paraformaldehyde in 100 mM potassium phosphate buffer (pH 7) on ice for 45 min under vacuum. Then, the material was rinsed with 100 mM potassium phosphate buffer and stained for GUS activity in GUS reaction buffer {100 mM potassium phosphate buffer (pH 7) containing 0.5 mM EDTA, 0.1% (v/v) Triton X-100, 0.5 mM K₃[Fe(CN)₆], 0.5 mM K₄[Fe(CN)₆], and 1 mM X-Gluc (5-bromo-4-chloro-3-indolyl β-D-glucuronide)} at 37°C overnight in the dark.

D. glomerata hairy roots were harvested and washed in GUS reaction buffer {1/4 strength SB buffer (12.5 mM PIPES, 1.25 mM MgSO₄, 1.25 mM ethylene glycol tetraacetic acid, pH 6.9) containing 1 mM EDTA, 0.1% (v/v) Triton X-100 with 0.25 mM K₃[Fe(CN)₆]}. Then, the samples were transferred to GUS reaction buffer containing 1 mM X-Gluc, vacuum-infiltrated three times, each time for 5 min and incubated for 1–24 hat 37°C in the dark. Then, the stained samples were transferred to fixative solution (0.1 M sodium phosphate buffer pH 6.8, 3% paraformaldehyde, 0.1% Tween 20, 0.1% Triton-X-100) and incubated at 4°C overnight. Later, the root samples were rinsed twice with 0.1 M sodium phosphate buffer (pH 6.8) and embedded in 3% agarose.

M. truncatula roots were sectioned on a vibratome 1000 Plus (Intracel, Royston, UK) at 100 µm thickness. D. glomerata roots were sectioned on a Leica VT1000E vibratome at 50–100 μm thickness (Leica Biosystems, Wetzlar, Germany).

GUS-stained M. truncatula and D. glomerata roots were examined under an SZX9 stereomicroscope (Olympus, Tokyo, Japan) with a DMC-FZ5GN digital camera (Panasonic, Kadoma, Osaka, Japan) or a SteREO Lumar.V12 stereomicroscope (Carl Zeiss) with an Axiocam MRc5 digital microscope camera (Carl Zeiss). M. truncatula root sections were viewed with a Leica DMBL microscope (Leica Microsystems, Wetzlar, Germany) and documented with a SPOT RT slider CCD camera (Diagnostic Instruments, Sterling Heights, MI, US). D. glomerata sections were observed under an Axiovert 200 M microscope (Carl Zeiss) using bright field microscopy; results were documented using an Axiocam 506 color camera (Carl Zeiss).

M. truncatula wild-type roots (∼5 cm from the root tip) and mature nodules were harvested in November–February, whereas D. glomerata wild-type roots (∼5 cm from the root tip), hairy roots, and mature nodules were harvested in June–October. Frozen plant material (50–200 mg of roots and nodules, respectively) was ground in liquid nitrogen. At least three biological and two technical replicates were analyzed; the exceptions were auxin in D. glomerata nodules and chickpea roots (two biological replicates).

For detailed analysis of cytokinins by liquid chromatography–mass spectrometry (LC-MS), trans-zeatin (tZ), trans-zeatin-7-glucoside (tZ7G), trans-zeatin-O-glucoside (tZOG), trans-zeatin-9-glucoside (tZ9G), trans-zeatin riboside O-glucoside (tZROG), cis-zeatin (cZ), trans-dehydrozeatin (tDHZ), trans-dehydrozeatin-riboside (tDZR), trans-zeatin riboside (tZR), and trans-N ⁶ -(Δ²-isopentenyl)adenine (tiP), and IAA were to be extracted and purified. Extraction was conducted with 1.25 ml of 80% (v/v) methanol. Samples were vortexed for 10 s; 4 µl of 5 ppm internal standard (deuterated mix) in 20% (v/v) methanol was added; then, the mixture was vortexed again for 10 s. The deuterium-labeled phytohormones used were [²H₃]tDZR, [²H₆]tiP, [²H₅]tZ, [²H₅]tZOG, [²H₅]tZR, [²H₅]tZROG, [²H₆]ABA, and [²H₅]IAA (Olchemim Ltd, Olomouc, Czech Republic). The suspension was incubated at 4°C for 30 min, vortexed and centrifuged at 20,000 g at 4°C for 15 min. Afterwards, the supernatant was filtered through Chromafix C18 columns (Macherey-Nagel, Düren, Germany) to remove particles of a size above 45 µm. As described above, the pellet was resuspended in the same volume of extraction solvent, then the suspension was vortexed and centrifuged, and the supernatant was filtered. Extracts from two extractions were pooled together and concentrated using a speedvac centrifuge. One milliliter of 80% (v/v) methanol was added to dried samples, and then the samples were incubated in an ultrasonic waterbath for 8 min or until the pellet was dissolved. After that, the suspension was vortexed and filtered through syringe filters Chromafil PET-20/15 MS (Macherey-Nagel, Düren, Germany) and stored at −80°C prior to LC-MS analysis. Analyses were carried out on a high-performance liquid chromatography (HPLC)/MS system consisting of an Agilent 1100 Series HPLC (Agilent Technologies, Santa Clara, CA, USA) equipped with a microwell plate autosampler and a capillary pump and connected to an Agilent Ion Trap XCT Plus mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) using an electrospray interface (ESI) as described in Großkinsky et al. (2014).

For detailed analysis of auxins by LC-MS/MS, the auxins IAA, IAA-alanine (IAA-Ala), IAA-aspartate (IAA-Asp), IAA-leucine (IAA-Leu)/IAA-isoleucine (IAA-Ile), IAA-phenylalanine (IAA-Phe), 4-chloro-IAA (4-Cl-IAA), IAA-tryptophan (IAA-Trp), IAA-valine (IAA-Val), indole-3-butyric acid (IBA), and PAA were extracted and purified. Sample preparation was carried out as described in Müller and Munné-Bosch (2011). The extracts were dried in a speedvac centrifuge with a vacuum cryopumping at 45°C overnight, filled with nitrogen gas, and stored at −80°C. Twenty nanograms of [²H₅]IAA (Cambridge Isotopes Laboratory, MA, USA) was used as an internal standard in individual samples. Prior to injection, dried samples were resuspended in 80% (v/v) methanol and filtered through a 0.45-µm Nanosep MF centrifugal filtration column (Pall Life Sciences, NY, USA). LC-MS/MS was performed using an Agilent 6530 Accurate Mass LC-MS Q-TOF (Santa Clara, CA, USA) or a Thermo UPLC Q Exactive Plus Orbitrap LC-MS/MS system (Thermo Fisher Scientific, Waltham, MA, USA). Samples were subjected to ESI in the Jetstream interface in both positive and negative modes. Conditions in the positive modes were as follows: gas temperature, 250°C; drying gas, 5 L min⁻¹; nebulizer, 30 psig; sheath gas temperature, 350°C and flowrate of 11 L min⁻¹; capillary voltage, 2,500 V; nozzle voltage, 500 V; and fragmentor voltage, 138 V. Conditions in the negative mode were as follows: gas temperature, 300°C; drying gas, 9 L min⁻¹; nebulizer, 25 psig; sheath gas temperature, 350°C and flowrate of 11 L min⁻¹; capillary voltage, 3,000 V; nozzle voltage, 500 V; and fragmentor voltage, 140 V. Samples were injected (7 µl) onto an Agilent Zorbax Eclipse 1.8 µm XDB-C18 2.1 × 50 mm column. Solvent A consists of 0.1% aqueous formic acid, and solvent B consists of 90% methanol with 0.1% formic acid. Free auxins and conjugates were eluted with a linear gradient from 10 to 50% solvent B from 0 to 8 min, 50 to 70% solvent B from 8 to 12 min (then hold from 12 to 20 min), and 70 to 10% solvent B from 20 to 21 min (then hold from 21 to 30 min) at a flowrate of 200 µl min⁻¹. The instrument was run in extended dynamic mode over a range of m/z 50–950 using targeted collision induced dissociation (CID; N₂ collision gas supplied at 18 psi) MS/MS (3 spectra s⁻¹). Analysis of data was performed using Agilent MassHunter software v 5.0. Authentic auxin standards (IAA-Phe, IAA-Leu, IAA-Val, IAA-Trp, 4-Cl-IAA; Olchemim Ltd, Olomouc, Czech Republic; IAA-Asp, IAA-Ala, IAA-Ile, IAA, IBA, PAA; Sigma, St. Louis, MO, USA) were used to determine elution times, collision energies, detection/quantification limits, and for absolute quantification.

A detailed analysis of auxin composition and concentrations was performed for roots and nodules of D. glomerata and M. truncatula, using the extraction method described by Müller and Munné-Bosch (2011) followed by LC-MS/MS. Roots and nodules of a second legume, C. arietinum (chickpea), like M. truncatula a member of the inverted repeat-lacking clade, were included to assess the diversity of auxin patterns in different legumes. The results are depicted in Figure 1 .

This method allowed the detection of the auxins IAA, IBA, 4-chloro-indole-3-acetic acid (4-Cl-IAA), and PAA and of the conjugated auxins IAA-Ala, IAA-Asp, IAA-Leu/IAA-Ile, IAA-Phe, IAA-Val, and IAA-Trp. Generally, IAA, IBA, 4-Cl-IAA, and PAA are considered as active auxins, while conjugated auxins are considered as inactive storage compounds (Korasick et al., 2013); it should be pointed out that conjugated forms of PAA were not analyzed as no standards were available. Concentrations of the auxins IBA and 4-Cl-IAA were below the limit of detection for the three plant species analyzed.

PAA, being present at levels much higher than those of IAA, turned out to be the dominant auxin in roots and nodules of both M. truncatula and C. arietinum as well as in roots and nodules of D. glomerata ( Figures 1A, B ). No significant differences were found between roots and nodules in the contents of PAA or IAA in both M. truncatula and D. glomerata ( Figures 1A, B ). However, this may not be representative of other legumes, as in chickpea we detected a significantly lower content of PAA and a significantly higher content of IAA in nodules than those in roots ( Figures 1A, B ). In M. truncatula and D. glomerata, the only conjugated auxins whose content was significantly higher in nodules than in roots was IAA-Ala , while IAA-Asp levels were only higher in nodules of D. glomerata ( Figures 1C, D ). Other auxin conjugates did not show any significant differences in their contents between roots and nodules in both M. truncatula and D. glomerata ( Figures 1E–H ). The results for chickpea were different: the contents of two conjugated auxins, IAA-Asp and IAA-Trp, were significantly higher in nodules than in roots. Absolute levels of active auxins differed between the three species examined.

In order to find out whether high PAA contents are typical for subterranean organs of actinorhizal plants or members of the Cucurbitales, roots from four other plant species were included in the analysis: Casuarina glauca (Casuarinaceae, Fagales) and Coriaria myrtifolia (Coriariaceae, Cucurbitales) as representatives of actinorhizal plants and B. bowerae and Cucumis sativus (cucumber) (Begoniaceae and Cucurbitaceae, both Cucurbitales) as non-symbiotic representatives of the Cucurbitales. The results are depicted in Table 1 . PAA was the dominant auxin in roots of cucumber and C. glauca; however, no PAA could be detected in roots of B. bowerae and C. myrtifolia. Thus, high PAA levels in roots are neither a general feature of actinorhizal plants nor that of Cucurbitales. To see whether higher levels of PAA were associated with nodulation of determinate nodule-forming legumes, we analyzed the auxin profile of nodules of two more legumes, namely, G. max (soybean) and the model legume L. japonicus ( Table 1 ). The results showed that PAA was indeed present in nodules of both species; however, due to high variability in PAA levels between biological replicates, no conclusion could be drawn regarding a possible role of PAA in nodulation of the determinate nodule-forming legumes.

As auxins typically act in concert with or antagonistically to cytokinins, the concentrations of cytokinins in roots and nodules were also quantified for both M. truncatula and D. glomerata. Endogenous concentrations of different cytokinins [trans-zeatin (tZ), cis-zeatin (cZ), trans-dehydrozeatin (tDHZ), trans-dehydrozeatin-riboside (tDZR), trans-zeatin riboside (tZR), and trans-N⁶-(Δ²-isopentenyl)adenine (tiP)] and their glycosylated forms [trans-zeatin-7-glucoside (tZ7G), trans-zeatin-9-glucoside (tZ9G), trans-zeatin-O-glucoside (tZOG), and trans-zeatin riboside O-glucoside (tZROG)] were determined for both M. truncatula and D. glomerata according to Großkinsky et al. (2014) ( Figure 2 ). Generally, the glycosides are considered stable biologically inactive storage forms, while non-glycosylated forms are considered biologically active (Van der Krieken et al., 1990).

In M. truncatula, concentrations of cytokinins were close to the detection limit, and no significant differences could be detected between roots and nodules in the concentrations of individual cytokinins ( Figures 2 A–J ).

In D. glomerata, the total concentration of glycosylated cytokinins was much higher in nodules than in roots (Mann–Whitney U test, p ≤ 0.05); in particular, the concentrations of tZOG and tZROG were increased ( Figures 2G, H ). Thus, while in roots, the combined concentrations of non-glycosylated cytokinins (tZ, tDHZ, tDZR, cZ, and tiP) were higher than those of glycosylated cytokinins; in nodules, the combined concentrations of glycosylated cytokinins were significantly higher than those of non-glycosylated cytokinins. With regard to non-glycosylated cytokinins, the only detectable difference between roots and nodules was that concentrations of cZ were significantly elevated in nodules compared to those in roots ( Figure 2B ), although cZ is thought to play a lesser role in plant development than tZ and tiP do (Schäfer et al., 2015).

To conclude, M. truncatula roots and nodules did not differ significantly in the contents of active auxins. Levels of the inactive conjugated auxin IAA-Ala were significantly higher in nodules than in roots. Furthermore, M. truncatula roots and nodules did not differ significantly in the contents of non-glycosylated and glycosylated cytokinins, respectively. In D. glomerata, roots and nodules did not differ significantly in the contents of active auxins, whereas the contents of conjugated auxins in roots were slightly higher than those in nodules. As for cytokinins in D. glomerata, the contents of the glycosylated ZROG and tZOG and the non-glycosylated cytokinin cZ were significantly higher in nodules than in roots.

Altogether, auxin and cytokinins profiles of M. truncatula and D. glomerata showed some similarities. Roots and nodules in both species did not differ significantly in the contents of active auxins and non-glycosylated cytokinins, respectively. Based on concentrations of detected active auxins (IAA and PAA) and indisputably active cytokinins (tZ and tiP), auxin/cytokinin ratios were calculated for roots and nodules of both M. truncatula and D. glomerata. In M. truncatula, auxin/cytokinin ratios were 124 and 97 for roots and nodules, respectively, whereas in D. glomerata the ratios were 101 and 96, respectively.

The synthetic auxin-responsive promoter DR5 (Ulmasov et al., 1997) was used to monitor the auxin response in transgenic roots of composite M. truncatula plants. The activity of this promoter had been shown to correlate with the sites of auxin accumulation in A. thaliana roots (Casimiro et al., 2001). In non-inoculated transgenic roots, strong DR5 promoter activity as signified by GUS staining was detected in root tips and lateral root primordia, while lower activities were found in the vascular system. Activities in the root cortex were very low ( Figures 3A–C ). After inoculation with S. meliloti, GUS activity was found in nodule primordia at 4 days postinoculation (dpi; Figure 3D ) and in the tip of differentiating nodules at 9 dpi ( Figure 3E ). In mature, 3-week-old nodules, GUS staining had more or less disappeared ( Figure 3F ).

Transgenic roots of composite D. glomerata plants were obtained by transformation with A. rhizogenes carrying the integration vector pIV10 with a DR5:GUS fusion construct. pIV10 without insert showed no background activity in transgenic roots of D. glomerata (data not shown). Strong DR5 promoter activity was detected in the root tip ( Figure 3G ) and in lateral root primordia, while lower activities were found in the vascular system ( Figure 3H ). No DR5 promoter activity was detected in the root cortex. The root tip staining was due to strong promoter activity in the root cap ( Figure 3G ). GUS activity was also detected in nodule primordia ( Figure 3I ). In nodules, GUS activity was confined to a single-cell layer distal to the nodule lobe meristem, i.e., to the nodule phellogen covering the nodule apex ( Figure 3J shows the whole nodule, Figure 3K shows a nodule section).

The cell response to auxin in M. truncatula roots was studied by application of the synthetic auxin NAA and the natural auxin PAA, respectively, to seedlings in axenic culture at the following concentrations: 10⁻⁹, 10⁻⁸, 10⁻⁷, 10⁻⁶, and 10⁻⁵ M. Some pictures of plant growth under the experimental conditions are shown in Supplementary Figure S1 . These experiments were performed three times for main root length growth and two times for root branching. Both auxins had an inhibitory effect on the elongation of the primary root, with different sensitivity ( Figure 4 ). In this experiment, sensitivity to PAA (10⁻⁸ M) was higher than sensitivity to NAA (10⁻⁷ M); however, in two repetitions of the experiment, sensitivity to PAA was lower than sensitivity to NAA.

Similarly, both auxins enhanced root branching, as evaluated by measuring the numbers of emerged and unemerged lateral roots (lateral root primordia) per main root ( Figure 5 ). In this series, sensitivity to NAA was equal to the sensitivity to PAA (10⁻⁶ M). The optimum curve of the auxin effect on lateral root branching was visible for NAA in that root branching at 10⁻⁵ M ( Figure 5B ) was lower than at control levels ( Figure 5A ), but not for PAA: root branching at 10⁻⁵ M was the same as at 10⁻⁶ M ( Figure 5B ). Concentrations higher than 10⁻⁵ M were not used because, already at 10⁻⁵ M, the effect on root length growth was so strong ( Figure 4 ) that counting of primordia was very difficult. Similar results were obtained by measuring the number of emerged lateral roots only ( Supplementary Figure S2 ). As for M. truncatula, the auxin response was studied in D. glomerata seedlings using NAA and PAA, respectively, at the concentrations 10⁻⁹, 10⁻⁸, 10⁻⁷, 10⁻⁶, and 10⁻⁵ M. The effect on the growth of the main root was similar to that in M. truncatula in that an inhibitory effect was observed at 10⁻⁸ M for NAA and at 10⁻⁷ for PAA, respectively ( Figure 6 ). However, in contrast to M. truncatula, neither auxin promoted root branching at any concentration examined. On the contrary, NAA inhibited root branching (expressed as lateral roots and lateral root primordia per main root) starting at 10⁻⁸ M and PAA at 10⁻⁷ M ( Figure 7 ). With NAA concentrations above 10⁻⁸ M, the inhibitory effect increased, while with PAA concentrations above 10⁻⁷ M, the inhibitory effect remained the same. Similar results were obtained by measuring the number of emerged lateral roots per main root only ( Supplementary Figure S2 ).

Adventitious roots are formed from stem or leaf-derived cells. Development of adventitious roots is a complex process that is affected by multiple factors including phytohormones, e.g., auxin (Steffens and Rasmussen, 2016), light, nutritional status, and stress responses such as wounding (Geiss et al., 2009). The effect of auxin on adventitious root formation requires polar transport (Vidoz et al., 2010; Sun et al., 2015). In the growth system used in this study, D. glomerata forms adventitious roots ( Supplementary Figure S3 ), while M. truncatula does not. Adventitious root formation was not restricted to the hypocotyl or a definable part of the stem. The effect of exogenously applied auxin on adventitious root formation was evaluated for D. glomerata ( Figure 8 ). Application of NAA or PAA, respectively, increased adventitious root formation following an optimum curve, and the induction of the morphogenetic program was more sensitive to NAA than to PAA. Superoptimal NAA concentrations (10⁻⁶–10⁻⁵ M) reduced adventitious root formation to values below those in the control, while superoptimal PAA concentrations reduced it to control values.