Cooperation between a root fungal endophyte and host‐derived coumarin scopoletin mediates Arabidopsis iron nutrition

Summary

• Iron acquisition is a critical challenge for plants, especially in iron‐deficient soils. Recent research underscores the importance of root‐exuded coumarins in modulating the root microbiome community structure and facilitating iron uptake. However, interactions between root‐associated fungi and coumarins in plant iron nutrition remain unknown. We investigated the mechanism by which a fungal endophyte, Macrophomina phaseolina (F80), enhances Arabidopsis iron nutrition.

• Fungal–coumarin interactions were assessed by profiling metabolites and measuring iron mobilisation in F80 cultures supplemented with specific coumarins, alongside quantifying growth performance and iron content in Arabidopsis coumarin‐biosynthesis mutants inoculated with F80.

• Our findings reveal that an interaction between the coumarin scopoletin and F80 in the rhizosphere rescues plant growth under iron‐limiting conditions by resolving the iron mobility bottleneck. F80 exhibits a capacity to modify scopoletin into iron‐chelating catechol coumarin esculetin, thereby releasing available iron.

• We conclude that Arabidopsis‐produced scopoletin functions as a precursor for fungal conversion into iron‐chelating coumarins. By extending the role of coumarins from bacterial to fungal members of the root microbiota, this study places coumarins at the centre of commensal‐mediated enhancement of plant iron nutrition across microbial kingdoms.

Introduction

Plants have evolved sophisticated mechanisms to shape their root microbiomes. Microbial community structure in the rhizosphere is influenced by various root‐exuded compounds, including a diverse class of secondary metabolites known as coumarins (Stassen et al., ²⁰²¹). Coumarins are derived from the phenylpropanoid pathway and are produced by a wide range of dicotyledonous plants (Rajniak et al., ²⁰¹⁸). Their broad biological activities, which include antimicrobial, antioxidant, and anti‐inflammatory properties, have attracted attention in plant and medical sciences (Borges et al., ²⁰⁰⁵). Coumarins can be found in above‐ and below‐ground plant tissues (Stringlis et al., ²⁰¹⁹; Robe et al., ²⁰²¹). Over 50 yr ago, foliar coumarin accumulation was linked to pathogen defence (Sequeira & Kelman, 1962; Reuveni & Cohen, ¹⁹⁷⁸; Zaynab et al., ²⁰²⁴), while more recent studies have highlighted their central role in plant iron acquisition at the root surface (Schmid et al., ²⁰¹⁴; Sisó‐Terraza et al., ²⁰¹⁶; Rosenkranz et al., ²⁰²¹).

Iron is an essential micronutrient for almost all life forms, including plants and their associated microbes (Aznar et al., ²⁰¹⁵). However, despite its high abundance in soil, iron is often bio‐unavailable to soil‐inhabiting organisms due to poor solubility as ferric oxide (Fe³⁺), particularly in highly oxygenated, neutral, or alkaline soils (Hindt & Guerinot, 2022). This limited bioavailability often leads to iron deficiency in plants, and in natural and agricultural environments iron fertilisation is often necessary (Zuo & Zhang, 2011). To counter this challenge, plants have evolved specialised iron acquisition strategies. Dicots and nongrass monocots employ ‘strategy I’, which relies on the reduction of iron before its import (Römheld, 1987; Eide et al., ¹⁹⁹⁶). In the model strategy I plant Arabidopsis thaliana (Arabidopsis), this is conferred by the tripartite reductive import module consisting of H⁺‐ATPase AHA2, FERRIC REDUCTION OXIDASE 2 (FRO2) and IRON‐REGULATED TRANSPORTER 1 (IRT1) (Tsai & Schmidt, 2017). AHA2‐mediated proton extrusion into the rhizosphere is thought to contribute to ferric iron (Fe³⁺) mobility through acidification of the rhizosphere (Santi & Schmidt, 2009). Acidification also lowers the redox potential to support Fe³⁺ reduction to ferrous iron (Fe²⁺) by FRO2, and consequently, Fe²⁺ is taken up via IRT1 (Varotto et al., ²⁰⁰²; Vert et al., ²⁰⁰²). Graminaceous plants, on the other hand, utilise ‘strategy II,’ marked by the release of mugineic acid‐type phytosiderophores that chelate Fe³⁺ and form complexes that can be efficiently absorbed by specific root transporters (Takagi et al., ¹⁹⁸⁴; Römheld & Marschner, ¹⁹⁸⁶; Ishimaru et al., ²⁰⁰⁶; Murata et al., ²⁰⁰⁶; Nozoye et al., ²⁰¹¹).

In addition to acidification, several strategy I plants also exude specific iron‐chelating coumarins to assist in iron solubilisation. Iron chelation is facilitated by catechol coumarins such as esculetin, fraxetin, and sideretin, with two hydroxyl groups in ortho positions (Hider et al., ²⁰⁰¹; Mladěnka et al., ²⁰¹⁰). The effectiveness of these coumarin metabolites is pH‐dependent, and expression of their biosynthesis genes responds to pH fluctuations (Sisó‐Terraza et al., ²⁰¹⁶; Tsai & Schmidt, ^{2017, 2020}; Gautam et al., ²⁰²¹; Paffrath et al., ²⁰²³). Plants also exude the noncatechol‐containing coumarin scopoletin, which is not effective in mobilising iron (Schmid et al., ²⁰¹⁴; Sisó‐Terraza et al., ²⁰¹⁶; Rajniak et al., ²⁰¹⁸). The complexity of coumarin profiles and their biological activity is further increased by their ability to influence the composition and dynamics of the plant root microbiota (Zamioudis et al., ²⁰¹⁴; Stringlis et al., ²⁰¹⁸; Voges et al., ²⁰¹⁹; Harbort et al., ²⁰²⁰). Several coumarins, especially scopoletin, were reported to possess selective antimicrobial activities (Stringlis et al., ²⁰¹⁹) and inhibit the growth of soil‐borne fungi (Carpinella et al., ²⁰⁰⁵; Kai et al., ²⁰⁰⁶; Ba et al., ²⁰¹⁷; Stringlis et al., ²⁰¹⁸). Moreover, Harbort et al. (2020) demonstrated that root commensal bacteria can alleviate iron starvation in Arabidopsis through interactions with root‐secreted fraxetin in alkaline pH conditions. Arabidopsis mutants lacking coumarins harboured altered bacterial communities and had heightened iron stress sensitivity (Harbort et al., ²⁰²⁰).

The first step in the Arabidopsis coumarin biosynthesis pathway involves the 2‐oxoglutarate‐dependent dioxygenase FERULOYL‐CoA 6′‐HYDROXYLASE (F6′H1) (Kai et al., ²⁰⁰⁸; Schmid et al., ²⁰¹⁴; Schmidt et al., ²⁰¹⁴). F6′H1 converts feruloyl‐CoA to 6‐hydroxyferuloyl‐CoA, which is then converted to scopoletin spontaneously or by the BAHD acyltransferase‐COUMARIN SYNTHASE (COSY) in organs shielded from light (Kai et al., ²⁰⁰⁸; Vanholme et al., ²⁰¹⁹). COSY was also found to produce esculetin in vitro from its hydroxycinnamoyl‐CoA thioester (Vanholme et al., ²⁰¹⁹). While the mode of esculetin biosynthesis in the plant remains unclear, Rajniak et al. (2018) suggested that F6'H1 is required, although Arabidopsis Col‐0 exudates contain relatively low levels of esculetin compared to scopoletin (Schmid et al., ²⁰¹⁴; Paffrath et al., ²⁰²³). Fraxetin is synthesised from scopoletin by SCOPOLETIN‐8‐HYDROXYLASE (S8H) and can be further hydroxylated into sideretin by the cytochrome P450 82C4 enzyme (CYP82C4) (Rajniak et al., ²⁰¹⁸; Siwinska et al., ²⁰¹⁸; Tsai et al., ²⁰¹⁸). These last steps in the coumarin biosynthesis pathway (S8H and CYP82C4) are highly induced in response to iron deficiency and lead to higher coumarin exudation rates (Rodríguez‐Celma et al., ²⁰¹³; Fourcroy et al., ²⁰¹⁴; Schmid et al., ²⁰¹⁴; Sisó‐Terraza et al., ²⁰¹⁶; Tsai et al., ²⁰¹⁸; Paffrath et al., ²⁰²³). Coumarins accumulate inside root tissues mainly as glycosylated forms, and their exudation is facilitated by the β‐glycosidase BGLU42 (Zamioudis et al., ²⁰¹⁴; Stringlis et al., ²⁰¹⁸), which deglycosylates scopolin into scopoletin (Ahn et al., ²⁰¹⁰; Zamioudis et al., ²⁰¹⁴). Coumarins are secreted into the rhizosphere via ATP‐binding cassette transporters like PLEIOTROPIC DRUG RESISTANCE 9 (PDR9) (Fourcroy et al., ²⁰¹⁴; Ziegler et al., ²⁰¹⁷), where they can directly contribute to iron uptake by chelating or reducing Fe³⁺ (Hider et al., ²⁰⁰¹; Mladěnka et al., ²⁰¹⁰; Schmidt et al., ²⁰¹⁴; Paffrath et al., ²⁰²³).

While a critical role of coumarin interactions with root‐associated bacterial microbiota for plant iron nutrition has been established (Harbort et al., ²⁰²⁰), contributions of fungal root endophytes and their interplay with root‐secreted coumarins for plant iron nutrition remain unreported. Here we used a fungal endophyte, Macrophomina phaseolina MPI‐SDFR‐AT‐0080 (F80), which is a member of the Arabidopsis mycobiota isolated from the roots of wild Arabidopsis populations and part of a microbial culture collection (Durán et al., ²⁰¹⁸). We demonstrate that M. phaseolina F80 alleviates host iron starvation in a coumarin‐dependent manner. We uncover a previously undescribed mechanism by which a root endophytic fungus promotes plant iron nutrition through cooperation with the plant‐derived coumarin scopoletin. Ex planta assays suggest that F80 mediates the conversion of scopoletin to the iron‐mobilising catechol coumarin esculetin, which then relieves plant iron deficiency by providing soluble Fe³⁺ to FRO2 for reduction and import via IRT1. Our data highlight a role of coumarin–microbiota interactions for plant iron nutrition across microbial kingdoms and further suggest the physiological importance of the highly abundant coumarin scopoletin for plant iron nutrition in nature.

Materials and Methods

Growth conditions

The following Arabidopsis thaliana (L.) Heynh. genotypes were used: Columbia‐0 (Col‐0) wild‐type, f6′h1‐1 AT3G13610; SALK_132418C (f6′h1), s8h‐1 AT3G12900; SM_3_27151 (s8h), cyp82c4‐1 AT4G31940; SALK_001585 (cyp82c4), frd1‐1 AT1G01580; N3777 (fro2); irt1‐1 AT4G19690; and SALK_024525 (irt1). Arabidopsis seeds were surface sterilised for 15 min with 70% ethanol + 0.1% Tween‐20, subsequently for 2 min in 95% ethanol and after drying, seeds were kept in sterile distilled water for 2–3 d in the dark at 4°C for stratification. For germination, seeds were placed onto square Petri plates containing ½‐strength Murashige & Skoog (½MS) medium (vitamins, 0.5 g l⁻¹ MES, pH 5.7, 0.5% sucrose, 1% BD DIFCO™ Agar). The plates were sealed with micropore tape and placed vertically into a growth chamber (10 h : 14 h, 21°C : 19°C, light : dark) for 7 d.

Fungal growth and hyphal harvesting

The fungus Macrophomina phaseolina (Tassi) Goid. (MPI‐SDFR‐AT‐0080 v.1.0) (F80) (Mesny et al., ²⁰²¹) was grown on potato dextrose agar at 22°C for c. 1–2 wk. Hyphae were harvested with a sterilised scalpel, transferred to grinding tubes with metal beads, diluted to 100 mg hyphae ml⁻¹ with 10 mM MgCl₂, and ground for 5 min. To obtain heat‐killed (HK) fungus, samples were incubated at 99°C for 1 h.

Gnotobiotic agar‐plate system

The gnotobiotic agar‐based system was adapted from Harbort et al. (2020) – STAR protocols. Adjusted ½ MS media without iron (750 μM MgSO₄, 625 μM KH₂PO₄, 10.3 mM NH₄NO₃, 9.4 mM KNO₃, 1.5 mM CaCl₂, 55 nM CoCl₂, 53 nM CuCl₂, 50 μM H₃BO₃, 2.5 μM KI, 50 μM MnCl₂, 520 nM Na₂MoO₄, 15 μM ZnCl₂, and 9.4 mM KCl, with 1% BD DIFCO™ Agar, Bacteriological) was prepared from stock solutions and supplied with 10 mM MES, 50 μM iron (FeEDTA or FeCl₃), and fungal hyphae (1 : 2000 dilution, 50 μg ml⁻¹ final concentration) after autoclaving. For coumarin supplementation experiments, coumarins were added from 100 mM DMSO stocks to a final concentration of 10 μM or an equal amount of DMSO as control. After solidifying, the top c. 2.5 cm agar was removed from 12 × 12 square Petri plates filled with 50 ml of medium, and seven germinated seedlings were transferred per plate (four replicate plates per experiment). Plates were sealed with micropore tape and vertically grown for 2 wk in a growth chamber (10 h : 14 h, 21°C : 19°C, light : dark) with roots shielded from light. After 2 wk, shoot samples were collected for Chl measurements, and whole shoots and roots were harvested to record shoot fresh weight (SFW) and assess fungal colonisation, respectively.

Shoot Chl analysis

The protocol for Chl extraction and quantification was based on previous studies (Hiscox & Israelstam, 2011; Harbort et al., ²⁰²⁰). About five to seven shoots from each plate were pooled, weighed, and kept at −80°C until processing. To each sample, 1 ml DMSO per 30 mg of leaf tissue was added, and they were incubated for 45 min at 65°C with 300 rpm agitation. Chl absorbance was measured at 652 nm on a Nanodrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA).

Shoot mineral element analysis

Mineral element analysis of whole shoots was performed as described previously (Paffrath et al., ²⁰²³). Shoot samples were dried to constant weight at 65°C and weighed into polytetrafluoroethylene tubes. Dried plant material was digested using concentrated HNO₃ (67–69% v/v; Bernd Kraft, Munich, Germany) and subjected to pressurised digestion in a high‐performance microwave reactor (UltraCLAVE IV, MLS GmbH). After digestion, the samples were diluted with deionised water (Milli‐Q Reference A+, Merck Millipore, Darmstadt, Germany) and analysed using high‐resolution inductively coupled plasma mass spectrometry (HR‐ICP‐MS) (ELEMENT 2; Thermo Fisher Scientific). Element standards were prepared from certified reference single standards from CPI‐International (USA).

Fungal colonisation assays

To assess fungal colonisation of Arabidopsis roots at 14 d post‐inoculation (dpi), roots were collected into Lysing Matrix E 2 ml tubes (MP Biomedicals, Irvine, CA, USA). The tube content was crushed for 30 s at 6000  g . 980 μl sodium phosphate buffer (MP Biomedicals) and 122.5 μl MT‐buffer (MP Biomedicals) were added, followed by two additional 30‐s rounds of crushing at 6000  g . The samples were then centrifuged for 15 min at 20 000 g . A volume of 50 μl binding matrix (MP Biomedicals) per sample was added to a 96‐well plate. A filter plate (Acroprep Advance, 0.2 μm Supor filter, Pall) was placed on top of the 96‐well plate, and 150 μl supernatant from each sample was transferred onto the filter plate. The plate was centrifuged for 20 min at 300 g ; thereafter, the flow‐through was mixed with the binding matrix by shaking for 3 min. 190 μl of each sample was transferred to a second filter plate and washed twice with 150 μl SEWS‐M (MP Biomedicals) by centrifuging for 5 min at 300 g , and the centrifuging step was repeated to dry the samples. DNA was eluted with 30 μl of ddH₂O by centrifuging at 300 g for 5 min and stored at −20°C until qPCR analysis. Fungal colonisation was then measured by qPCR as described (Mesny et al., ²⁰²¹). Briefly, ITS1F/ITS2 primers for fungal ITS1 and UBQ10F/UBQ10R primers for Arabidopsis Ubiquitin10 were used. Reactions included iQ™ SYBR® Green Supermix, primers and 1 μl DNA template. qPCR was run on a BioRad CFX Connect with 95°C denaturation (3 min), followed by 39 cycles of 95°C (15 s), 60°C (30 s) and 72°C (30 s). Colonisation index was calculated as 2^{−Cq(ITS1)/Cq(UBQ10)}.

96‐well fungal culture assay

Culture medium containing ½MS medium without iron, with ARE, and vitamins (Table S1) was prepared; the pH was adjusted to pH 5.7 (buffered with 10 mM MES) with KOH and sterile‐filtered (0.22 μm) using vacuum filtration. The medium was supplemented with scopoletin, fraxetin, or esculetin (0 or 2 mM) and 190 μl was added per well in a 96‐well plate. The wells were inoculated with 10 μl of fungal hyphal solution (stock solution of 100 mg mycelium ml⁻¹ in 10 mM MgCl₂; final concentration 5 mg ml⁻¹) and 10 μl of 10 mM MgCl₂ was used as a blank control. The plates were sealed and incubated for 12 d at 22°C with shaking (140 rpm); OD₆₀₀ and scopoletin fluorescence (excitation 385 nm, emission 470 nm) measurements were taken daily around the same time using a TECAN spectrophotometer (Infinite® M Plex; Tecan Trading AG, Zurich, Switzerland). To ensure consistency across experiments, detected fluorescence was normalised to OD₆₀₀ and the initial fluorescence at day 0. To evaluate potential fungal growth inhibition by scopoletin, daily OD₆₀₀ values were normalised to growth in the absence of scopoletin.

Metabolite extraction

For F80 liquid culture time‐course assays, pure ethyl acetate (Honeywell Riedel‐de HaënTM, LC‐MS grade) was used to extract metabolites from fungal culture pellet and supernatant fractions. Extraction was performed by adding 500 μl of ethyl acetate to 1.5 ml fungal culture fractions, followed by 30 s vortexing and 10 min sonic bath. The sample was centrifuged for 10 min (20 000 g , 4°C) and 450 μl of the upper phase was collected in a new 1.5‐ml test tube. The extraction was repeated twice. The combined solvent extract was dried in a TurboVap® Classic LV (Biotage, Uppsala, Sweden) under nitrogen flow, and samples were stored at −80°C.

For exudate profiling in the plant medium, 5 ml pure ethyl acetate (Honeywell Riedel‐de HaënTM, LC‐MS grade) was added to 50 ml plant agar medium (harvested 1 wk after transfer of 9‐d‐old seedlings), and samples were vortexed for 30 s and centrifuged for 10 min (4°C, 1800 g ). The supernatant was collected, and two more rounds of extraction were performed. Combined extracts were dried as described above, and samples were stored at −80°C.

For coumarin profiling of plant roots (harvested 1 wk after transfer of 9‐d‐old seedlings on ½MS medium), 200 mg (FW) ground root samples were extracted using a modified version of Blight & Dyer extraction. First, 900 μl dichloromethane : ethanol, 2 : 1 and 100 μl HCl (pH 1.4) were consecutively added to each sample, followed by 30 s vortexing and 5 min centrifugation (4°C, 21 000 g ). The upper phase was discarded, and the lower phase (c. 700 μl) was collected into 2‐ml test tubes. After adding 500 μl tetrahydrofuran to the remaining pellets, samples were vortexed for 30 s and centrifuged for 5 min (4°C, 21 000 g ). The supernatant was collected and combined with the lower phase fraction. The solvent mixture was dried as described above, and the samples were stored at −80°C.

Metabolite profiling with mass spectrometry

Before measurement, each sample was resuspended in 100 μl of pure methanol (Honeywell Riedel‐de HaënTM, LC‐MS grade) and transferred to a liquid chromatography vial with insert. Separation of metabolites by ultrahigh‐performance liquid chromatography (UHPLC) was performed on a Nucleoshell RP18 (2.1 × 150 mm, particle size 2.1 μm, Macherey & Nagel, GmbH, Düren, Germany) using an ACQUITY UPLC System, equipped with a Binary Solvent Manager and Sample Manager (20 μl sample loop, partial loop injection mode, 2 μl injection volume, Waters GmbH Eschborn, Germany). Eluents A and B were aqueous 0.3 mmol l⁻¹ NH₄HCOO (adjusted to pH 3.5 with formic acid) and acetonitrile, respectively. Elution was performed isocratically for 2 min at 5% eluent B, from 2 to 7 min with a linear gradient to 30% B, from 7 to 99 min at 95% B, then 1 min at 95% B, from 10 to 11 min at 5% B, and then another minute at 5% B. The flow rate was set to 400 μl min⁻¹, and the column temperature was maintained at 40°C. Metabolites were detected by negative electrospray ionisation and mass spectrometry (ESI‐QTOF‐MS/MS). Serial dilutions of standards of scopoletin, esculetin, fraxetin, and 3‐(2,4‐dihydroxy‐5‐methoxyphenyl) propanoic acid were injected every 24 samples for identification and quantification purposes. Mass spectrometric analysis was performed by MS‐QTOF‐IDA‐MS/MS (ZenoTOF 7600, AB Sciex GmbH, Darmstadt, Germany) and controlled by SCIEX OS software (AB Sciex GmbH). The source operation parameters were as follows: ion spray voltage, −4500 V; ion source gas 1, 60 psi; source temperature 600°C; ion source gas 2, 70 psi; curtain gas, 35 psi. Instrument tuning and internal mass calibration were performed every five samples with the calibrant delivery system applying X500 ESI Calibrant Solution in negative mode tuning (AB Sciex GmbH). Data analysis was performed in sciex os software and ms‐dial software (Tsugawa et al., ²⁰¹⁵).

Ferrozine iron mobilisation assay

To remove medium residues from the hyphal solution, fungal hyphae (prepared as described above) were washed three times with MgCl₂ by centrifugation (5 min at 11 000 g ) and resuspended in fresh MgCl₂. Coumarins were supplemented to the culture medium (½MS medium without iron, with ARE and vitamins (Table S1), 10 mM MES, pH 5.7) to a final concentration of 400 μM. Cultures were inoculated with washed hyphae and grown as described for the 96‐well fungal culture assay (22°C, 140 rpm, dark). Samples were collected at 0, 6, and 9 dpi, centrifuged at 4000 g for 20 min; the supernatant was transferred into a new 96‐well plate and stored at −80°C until processing.

To assess iron mobilisation, a 120‐μl sample was transferred into a 96‐well plate and supplemented with a final concentration of 0.1 mM FeCl₃ and 0.4 mM ferrozine. Absorbance at 562 nm was measured using a TECAN spectrophotometer (Infinite® M Plex; Tecan Trading AG) before and after ferrozine addition. Controls included wells with water, 0.1 mM FeCl₃ (negative control), and 0.1 mM FeEDTA (positive control). To quantify total mobilised iron, immobile iron was removed by centrifugation (1700 g , 15 min), and the supernatant was transferred to a new 96‐well plate. Samples were then incubated with 10 mM HCl and 10 mM ascorbic acid for 20 min in darkness to reduce the mobile iron fraction, followed by absorbance measurement at 562 nm. All values were normalised to FeEDTA samples (before reduction), and FeSO₄ calibration curves were used for quantification of mobile iron.

Results

Macrophomina phaseolina (F80) enhances iron nutrition in Arabidopsis under iron‐limiting conditions

To test whether fungal endophytes can improve plant iron nutrition, we modified a previously reported gnotobiotic agar‐based plant growth assay that allows precise modulation of iron bioavailability and the co‐cultivation of Arabidopsis plants with root microbiota members (Harbort et al., ²⁰²⁰). The system relies on ½MS medium buffered at pH 5.7 with MES and supplemented with either unchelated iron (FeCl₃) or chelated iron (FeEDTA), which reflect bio‐unavailable iron (unavFe) and bioavailable iron (avFe) conditions, respectively. Using this system, we assessed the ability of the fungal root endophyte F80 to enhance the performance of Arabidopsis Col‐0 wild‐type plants under unavFe conditions. F80 was identified previously as part of a culture collection of 41 fungal isolates representative of the Arabidopsis root mycobiota (Mesny et al., ²⁰²¹). Plant performance was evaluated after 2 wk of co‐cultivation with fungus or mock treatments using SFW and shoot total Chl concentration as a proxy for iron status. Three‐week‐old Arabidopsis seedlings exhibited enhanced growth and reduced leaf chlorosis when inoculated with F80 under iron‐limiting conditions, closely resembling the avFe phenotype (Fig. 1a). Plants grown at unavFe and inoculated with F80 exhibited a significant increase in SFW and shoot Chl concentration compared to mock‐treated controls (Fig. 1b,c). By contrast, we did not observe F80‐mediated plant growth promotion or increased Chl under avFe conditions (Fig. 1b,c). Plants inoculated with heat‐killed F80 behaved similarly to mock‐treated plants (Fig. S1), highlighting the importance of live fungus and indicating that putative inoculum‐derived iron contamination is negligible. To ascertain whether the observed phenotypes reflect the iron status of the plant, the shoot elemental content was measured by HR‐ICP‐MS (Fig. 1d). This analysis showed that F80‐inoculated plants accumulated significantly more iron in their shoots under unavFe compared to mock conditions, almost reaching the iron content of plants grown at avFe (Fig. 1d). In line with SFW and shoot Chl concentration, no F80‐mediated increase in shoot iron content was observed in avFe conditions (Fig. 1d). Taken together, these results show that F80 mediates Arabidopsis iron nutrition specifically in conditions of limited iron bioavailability. This implies that F80 improves plant access to immobile iron pools in the growth medium.

Fig. 1.

Fungal endophyte Macrophomina phaseolina (F80) improves Arabidopsis iron nutrition under iron limitation. (a) Representative phenotypes, (b) shoot fresh weight (SFW), (c) shoot Chl concentration, and (d) shoot Fe content of Col‐0 Arabidopsis seedlings. Seven‐day‐old seedlings were transferred to ½MS medium and grown for 2 wk on mock or F80‐inoculated plates in unavFe (50 μM FeCl₃) or avFe (50 μM FeEDTA) conditions. Medians are shown as center lines, boxes indicate 25^th–75^th percentiles, and whiskers extend 1.5× the interquartile range. Letters indicate significant pairwise differences between groups (P _adj ≤ 0.05) by Dunn pairwise comparison test with Benjamini–Hochberg correction (b, c) or a Tukey's HSD corrected for multiple comparisons (d) Bars, 1 cm (a). Data are from two (b, c) or three (d) full factorial replicates (represented by different shapes).

Root‐derived scopoletin and host iron‐reductive import machinery are essential for F80‐mediated plant iron nutrition

Given that F80‐mediated plant iron nutrition is specific to iron‐limiting conditions and the established link between iron limitation‐induced coumarin production and bacterial‐mediated improvements in plant iron nutrition (Harbort et al., ²⁰²⁰), we investigated whether F80‐mediated improvement in the iron nutrition of Arabidopsis also requires plant‐derived coumarins. To this end, we grew the coumarin‐deficient Arabidopsis mutant f6′h1 lacking the first step of the coumarin biosynthesis pathway (Kai et al., ²⁰⁰⁸; Schmid et al., ²⁰¹⁴; Schmidt et al., ²⁰¹⁴) in unavFe conditions. Compared to Col‐0, f6′h1 plants exhibited reduced ability to tolerate iron deficiency, indicated by their lower SFW and shoot Chl concentration under mock conditions (Fig. 2a,b). While F80‐mediated rescue of plant iron deficiency was observed in Col‐0 plants, F80 failed to improve the performance of f6′h1 plants, with both SFW and Chl levels remaining below the mean values observed for Col‐0 (Fig. 2a,b). Consistently, shoot iron content in f6′h1 plants did not significantly increase upon F80 inoculation and remained comparable to mock‐treated controls (Fig. 2c). Notably, F80 inoculation caused a slight increase in SFW and Chl concentration of f6′h1 plants (Fig. 2a,b), suggesting a small coumarin‐independent F80 growth promotion effect.

Fig. 2.

Rescue of iron‐deficient Arabidopsis plants by the fungal root endophyte F80 requires the coumarin scopoletin and the plant's iron reductive import system. (a) Shoot fresh weight (SFW), (b) shoot Chl concentration, and (c) shoot Fe content measured by ICP‐MS at 2 wk of growth after the transfer of the indicated mutants in the coumarin biosynthesis pathway and the iron reductive import mechanism. Seven‐day‐old seedlings were transferred to ½MS medium with unavFe (50 μM FeCl₃) at pH 5.7, mock or inoculated with F80. The dashed line indicates the mean of Col‐0. Letters indicate significant pairwise differences between groups (P _adj ≤ 0.05) by a Dunn pairwise comparison test with Benjamini–Hochberg correction. Statistical significance within genotypes was determined by a Kruskal–Wallis test. Data are from four (a, b) or three (c) full factorial replicates (represented by different shapes).

To examine whether F80 interacts with specific coumarins to facilitate rescue of plant iron deficiency, we used the s8h and cyp82c4 mutant lines. In s8h, scopoletin accumulates, but it fails to produce fraxetin and sideretin. In cyp82c4, both scopoletin and higher amounts of fraxetin are produced, but not sideretin (Rajniak et al., ²⁰¹⁸; Tsai et al., ²⁰¹⁸). Inoculation of s8h plants with F80 led to increased SFW and shoot Chl concentration, reaching levels comparable to those of F80‐inoculated Col‐0 plants. This trend was also observed for shoot iron content, as s8h plants treated with F80 accumulated similar levels of iron as Col‐0 (Fig. 2c). Similarly, F80 inoculation improved the performance of cyp82c4 plants, reflected in increased SFW, Chl concentration, and shoot iron content compared to mock‐treated controls, although levels in cyp82c4 remained slightly lower than those of Col‐0 and s8h plants (Fig. 2a–c). These data show that the fungal root endophyte alleviates plant iron starvation in a coumarin‐dependent manner and that the coumarin scopoletin interacts with F80 to rescue host iron deficiency.

Available Arabidopsis mutants of FRO2 and IRT1, which have disabled Fe³⁺ reduction and Fe²⁺ import, respectively, allowed us to assess whether these functions are required for F80‐mediated plant iron rescue. As previously reported, the fro2 and irt1 mutants were acutely sensitive to iron limitation, resulting in small and chlorotic plants under unavFe conditions (Robinson et al., ¹⁹⁹⁹; Varotto et al., ²⁰⁰²; Vert et al., ²⁰⁰²; Paffrath et al., ²⁰²³). F80 failed to effectively increase fro2 and irt1 SFW, shoot Chl concentration, and shoot iron content (Fig. 2a–c). This suggests that the F80‐mediated rescue involves ferric iron mobilisation and requires downstream Fe³⁺ reduction by FRO2 as well as ferrous iron transport by IRT1. The concentrations of other elements in the shoots were also evaluated for all genotypes by HR‐ICP‐MS; however, no similar trends were observed (Fig. S2). The observed differences in F80‐mediated plant iron nutrition were not attributed to variation in fungal colonisation, as F80 colonisation levels were similar in all plant genotypes (Fig. S3c). Analysis of all plant genotypes at avFe confirmed that F80‐mediated rescue is specific to iron deficiency, as there were no differences observed in shoot Chl of mock vs F80‐inoculated plants (Fig. S3a,b). Additionally, the presence of FeEDTA completely restored plant growth of all tested coumarin mutants, indicating that coumarins are not required for plant growth at avFe. As anticipated, FeEDTA did not fully restore Chl levels in fro2 and irt1, as iron reduction and uptake are still required under this condition (Fig. S3a,b). For all genotypes except s8h, an increase in SFW was observed in avFe conditions in the presence of F80, which might reflect a general beneficial effect of the fungus on plant growth that is independent of iron. The results suggest that F80‐mediated Arabidopsis iron acquisition depends on plant coumarin synthesis, especially scopoletin, and requires a functional plant reductive iron transport machinery to facilitate iron uptake.

F80 converts scopoletin into iron‐chelating esculetin

Scopoletin is recognised for its selective antimicrobial activity (Sun et al., ²⁰¹⁴; Zamioudis et al., ²⁰¹⁴; Stringlis et al., ²⁰¹⁸; Harbort et al., ²⁰²⁰; Lemos et al., ²⁰²⁰). To evaluate the sensitivity of F80 to scopoletin, an in vitro time‐course assay was performed to monitor fungal growth in the presence of 2 mM scopoletin in a liquid culture. To additionally assess potential F80‐mediated alterations in scopoletin, we tracked scopoletin‐derived fluorescence in the cultures over time under long‐wave UV (Goodwin & Kavanagh, 1949). Under mock conditions, the scopoletin‐derived fluorescence remained largely constant, confirming the stability of scopoletin in the absence of the fungus. This is consistent with a previous study using mass spectrometry‐based quantification of coumarins (Rajniak et al., ²⁰¹⁸). In the presence of F80, scopoletin‐derived fluorescence decreased rapidly between 3 and 6 d dpi (Fig. 3a). A scopoletin alteration effect was also indicated by the fungal growth pattern: relative growth of F80 (growth in 2 mM scopoletin/growth in control) decreased from 2 to 5 dpi but recovered from 6 dpi onwards, coinciding with a near‐complete disappearance of scopoletin fluorescence (Fig. S4b). These data suggest that F80 can evade scopoletin antimicrobial activity by a possible detoxification mechanism that alters scopoletin.

Fig. 3.

The fungal endophyte F80 converts scopoletin into esculetin and further degradation compounds. (a) Time course of 0–12 d post‐inoculation (dpi) displaying specific scopoletin fluorescence measured in in vitro mock and F80 cultures supplemented with 2 mM scopoletin (excitation at 385 nm and emission detected at 470 nm). Data combined from three biological replicates. Curve‐shade indicates SE. The area under the curve of graph (a) is shown in Fig. S4(a). (b) MS‐QTOF‐IDA‐MS/MS peak area of scopoletin, esculetin, and 3‐(2,4‐dihydroxy‐5‐methoxyphenyl) propanoic acid (C10H12O5) in F80 culture supernatant (dashed line) and the fungal pellet (full line) grown for 0–12 dpi supplemented with 2 mM scopoletin. Curve‐shading indicates mean ± SD. Absolute concentration values are presented in Fig. S4(d). (c) MS‐QTOF‐IDA‐MS/MS peak area of esculetin in F80 culture supernatant (dashed line) and the fungal pellet (full line) grown for 0–12 dpi supplemented with 2 mM esculetin. Curve‐shade indicates mean ± SD. Absolute concentration values are presented in Fig. S4(e). (d) MS‐QTOF‐IDA‐MS/MS peak area of fraxetin in F80 culture supernatant (dashed line) and the fungal pellet (full line) grown for 0–12 dpi supplemented with 2 mM fraxetin. Curve‐shade indicates mean ± SD. Absolute concentration values are presented in Fig. S4(f). (e) Total amount of iron mobilised from poorly soluble FeCl₃ (0.1 mM) in the F80 culture supernatant or axenic medium (mock) supplemented with scopoletin, fraxetin, or esculetin (final concentration of 0.4 mM) at 0, 6, or 9 dpi. Total mobilised iron was assessed spectrophotometrically based on a Fe(II)‐ferrozine assay after reduction of soluble ferric pools with ascorbic acid. Control values for treatments without fungus (red) or without coumarins (green) are included as dashed lines. Error bars represent mean ± SD; statistical differences between treatments were assessed using t‐tests for normally distributed residuals or, otherwise, Wilcoxon rank‐sum tests. Significance levels are indicated by asterisks with *, P < 0.05 and ***, P < 0.001. All panels represent data from three biological replicates (represented by different shapes).

To validate whether F80 alters scopoletin, liquid cultures of the fungus were separated into supernatant and mycelia, and metabolites were profiled by UHPLC‐MS/MS. The MS/MS data corroborated the fluorescence‐based results in showing a steep decrease in scopoletin levels from 3 to 6 dpi with F80 (Figs 3b, S4d). Notably, a different coumarin molecule, esculetin, was detected in the samples supplemented with scopoletin and displayed increased abundance from 3 dpi onwards. At the same time, additional degradation products containing hydrolysed lactone rings were observed. Untargeted UHPLC–MS/MS profiling of the samples revealed that the most abundant compound detected, besides scopoletin, was 3‐(2,4‐dihydroxy‐5‐methoxyphenyl) propanoic acid (C₁₀H₁₂O₅) (Figs 3b, S4d). At 12 dpi, all coumarin compounds, including scopoletin, esculetin, and 3‐(2,4‐dihydroxy‐5‐methoxyphenyl) propanoic acid, were nearly undetectable. This suggests that further modification or complete degradation through F80, or compound instability, occurs in this assay. Both esculetin and the open‐lactone structure were detected exclusively in scopoletin‐supplemented samples in the presence of the fungus but not in fungal cultures alone (Figs 3b, S4d).

Given that esculetin, an iron‐mobilising coumarin (Schmid et al., ²⁰¹⁴; Paffrath et al., ²⁰²³), was identified as a scopoletin‐conversion product, we next tested the interaction of esculetin with F80. First, esculetin stability in F80 cultures supplemented with 2 mM esculetin was assessed by measuring its levels in the supernatant and fungal pellet over time. Esculetin showed only a slow decline and was still detectable at 12 dpi (Figs 3c, S4e). In parallel, since fraxetin accumulates in the roots of cyp82c4 plants, which support F80‐mediated rescue, the stability of fraxetin was evaluated in the presence of F80 in culture (Fig. 2a–c). Compared to esculetin, fraxetin was more rapidly depleted by the fungus, especially in the supernatant (Figs 3d, S4f). To further investigate the interaction of F80 with these coumarins, the growth of F80 in the presence of scopoletin, esculetin, or fraxetin was evaluated over time. In comparison with F80 cultures supplemented with scopoletin, F80 cultures supplemented with esculetin or fraxetin did not show a decrease in growth, similar to F80 in control conditions (Fig. S4c). The observed F80 conversion of scopoletin to esculetin, a compound that does not affect F80 growth, further supports a fungal detoxification mechanism.

Because our in planta data imply a role for F80 in improving plant access to immobile ferric iron through interactions with scopoletin – a coumarin which lacks a catechol moiety and in vitro iron mobilising capacity (Fig. 2a–c) (Schmidt et al., ²⁰¹⁴; Rajniak et al., ²⁰¹⁸; Paffrath et al., ²⁰²³) – we hypothesised that F80‐mediated alteration in scopoletin is the cause of iron mobilisation to the plant. To test this, we made use of a spectrophotometric assay which quantifies mobile iron based on the ferrozine‐Fe(II) complex (see Materials and Methods section). Fraxetin and esculetin were included in the analysis because esculetin was detected in the F80‐scopoletin cultures (Figs 3b, S3d) and fraxetin produced in F80‐rescued cyp82c4 plants might also interact with the fungus. As expected, the catecholic coumarins fraxetin and esculetin, but not scopoletin, mobilised iron under mock conditions (Fig. 3e). However, when F80 was cultured with these coumarins, differential effects on total iron mobilisation were observed. In F80 cultures supplemented with scopoletin, there was an increase in mobilised iron at 6 and 9 dpi (Fig. 3e), perhaps due to the conversion of scopoletin to esculetin by F80 (Fig. 3b). Conversely, total iron mobilisation in cultures supplemented with fraxetin or esculetin in the presence of the fungus decreased significantly at 6 and especially 9 dpi (Fig. 3e). The results suggest that F80 affects iron mobilisation differently depending on the coumarin substrate it encounters. While scopoletin likely serves as a precursor to iron‐chelating compounds such as esculetin, fraxetin and esculetin participate directly in iron‐binding dynamics.

The above findings show that an interaction between F80 and scopoletin yields synergistic amounts of mobilised iron. Put together with F80's capability to convert scopoletin to the catechol coumarin esculetin, we reasoned that the observed F80‐mediated Arabidopsis iron nutrition could be due to the presence of esculetin.

F80 and scopoletin cooperate to mobilise ferric iron in the rhizosphere

To assess whether F80 interaction with coumarins observed in culture translates into iron mobilisation effects in planta, we conducted coumarin supplementation assays using the coumarin‐deficient f6′h1 mutant. This approach provided a coumarin‐free plant background, enabling evaluation of the importance and sufficiency of specific coumarins for plant physiological fitness and growth under iron limitation. Based on the positive interaction of F80 with scopoletin in enhancing iron mobilisation and growth recovery of f6′h1 plants at avFe (Fig. S3a,b), we hypothesised that combinatorial treatment of the plant growth medium with F80 and scopoletin would result in increased mobile ferric iron and, in turn, restore iron‐limiting plant growth of f6′h1. Addition of 10 μM scopoletin to the plant growth medium under mock (non‐F80) conditions did not significantly increase SFW or Chl concentration in f6′h1 to levels observed in Col‐0 inoculated with F80 (Fig. 4a,b). When combined with F80 inoculation, scopoletin supplementation restored F80‐mediated rescue, with both SFW, Chl, and shoot iron concentrations being similar to Col‐0 inoculated with F80 (Figs 4a,b, S2). Strikingly, the cooperative effect of F80 and scopoletin resembled the mock phenotype of 10 μM esculetin supplementation, which increased SFW and Chl concentration to levels comparable to Col‐0 with F80 (Fig. 4a,b). By contrast, supplementation of fraxetin or esculetin in the presence of F80 resulted in slightly reduced SFW and Chl levels compared to the corresponding mock conditions (Fig. 4a,b). This result is in line with the fungal culture supernatant iron mobilisation data (Fig. 3e) and is likely explained by the degradation of these coumarins observed in MS/MS analysis of F80 cultured with esculetin and fraxetin (Figs 3c,d, S4e,f).

Fig. 4.

Interaction between the fungal endophyte F80 and scopoletin rescues the coumarin‐deficient mutant f6′h1. (a) Shoot fresh weight (SFW) and (b) shoot Chl content of Arabidopsis Col‐0 and f6′h1 plants 2 wk past inoculation. Seven‐day‐old seedlings were transferred to ½MS medium with unavailable iron (50 μM FeCl₃) at pH 5.7, supplemented with 10 μM of the indicated coumarins or an equal amount of DMSO for the control, and mock or inoculated with F80. The dashed line indicates the mean of Col‐0. Letters indicate significant pairwise differences between groups (P _adj ≤ 0.05) by a Dunn pairwise comparison test with Benjamini–Hochberg correction. Data are from three full factorial replicates (represented by different shapes). (c) The concentration of scopoletin exuded after 1 wk in the medium of 9‐d‐old seedlings transferred to ½ MS medium with unavailable (50 μM FeCl₃) or available (50 μM FeEDTA) iron at pH 5.7, mock or inoculated with F80. Metabolites were extracted from c. 50 ml of the medium and analysed with MS‐QTOF‐IDA‐MS/MS. Bars represent the mean from four biological replicates (n = 2–3) with SE bars. Letters indicate significant pairwise differences between groups (P _adj ≤ 0.05) by a Tukey's HSD corrected for multiple comparisons.

To further understand the role of F80 in the iron import pathway of Arabidopsis plants, scopoletin, fraxetin, or esculetin were added to the growth medium of ferric chelate reductase‐deficient fro2 mutant plants. Importantly, none of these compounds enhanced SFW or Chl concentration in fro2 to levels of Col‐0, either alone or in combination with F80 (Fig. S5a,b). However, supplementing with these coumarins slightly increased Chl concentration in fro2, and esculetin together with F80 improved both SFW and Chl concentration compared to control and esculetin‐mock treatments (Fig. S5a,b). These results suggest that partial iron reduction might occur independently of FRO2, while FRO2 is required to fully rescue plant growth. These data underscore the necessity of a functional FRO2 for F80‐mediated alleviation of plant iron deficiency. They also support the notion that F80 cooperatively interacts with scopoletin to mediate ferric iron mobilisation in the rhizosphere. To critically test the impact of chemical iron mobilisation on f6′h1 and fro2 mutants, 10 μM NaEDTA (a synthetic iron chelator) was added to the medium. In f6′h1 plants, NaEDTA supplementation almost fully rescued the f6′h1 iron‐deficiency phenotype under mock and F80‐inoculated conditions (Fig. S5c,d). This aligns with a previous study showing that iron chelators like NaEDTA can bypass the requirement for microbial interactions by artificially increasing iron solubility (Jin et al., ²⁰⁰⁶). By contrast, no NaEDTA‐mediated rescue was observed in fro2 mutants (Fig. S5c,d). This result reinforces the conclusion that FRO2 is essential for the iron uptake pathway when bioavailable ferric iron is present in the form of Fe(III)‐chelates (Figs S5c, S4d). F80 colonisation of roots remained consistent across all experimental conditions, indicating that differences in rescue phenotypes were not due to variation in fungal colonisation (Fig. S5e).

To test whether F80 can also use plant‐synthesised scopoletin, we quantified differences in scopoletin concentration in root exudates of Col‐0 plants grown over 7 d in the presence or absence of F80. HRLC‐MS/MS analysis revealed that the presence of F80 strongly decreased scopoletin levels in the agar medium in both unavFe and avFe conditions (Fig. 4c). Levels of scopolin and scopoletin quantified inside root tissues from this experiment revealed no significant differences between treatments (Fig. S5f). Altogether, the data suggest that F80 and scopoletin interact cooperatively in the rhizosphere to mediate ferric iron mobilisation upstream of the FRO2‐IRT1 reductive import module.

Discussion

Coumarins are crucial for plant iron uptake in conditions of low iron availability. Scopoletin is one of the most abundant coumarins in root exudates of Arabidopsis plants grown under iron‐limiting conditions (Sisó‐Terraza et al., ²⁰¹⁶; Rosenkranz et al., ²⁰²¹), despite its inability to chelate and mobilise iron. Scopoletin is mainly described for its selective antimicrobial activity against fungi and commensal bacteria (Ba et al., ²⁰¹⁷; Harbort et al., ²⁰²⁰; Yu et al., ²⁰²¹), but its potential role in plant iron nutrition remains undefined. Here, we show that scopoletin and the fungal root endophyte Macrophomina phaseolina F80 cooperate to provide mobile Fe³⁺ for subsequent import via the plant reductive import module consisting of the ferric chelate reductase FRO2 and the ferrous iron importer IRT1 (Fig. 5). Our results support a pivotal role of scopoletin in the intricate interplay between F80 and its host plant, whereby scopoletin production and interaction with F80 in the rhizosphere enable the rescue of iron‐deficient Arabidopsis.

Fig. 5.

Model for the interaction of the fungal root endophyte Macrophomina phaseolina F80 with root‐derived scopoletin to mediate iron nutrition in Arabidopsis. Iron deficiency promotes coumarin biosynthesis and exudation in Arabidopsis roots. The first steps in coumarin biosynthesis are catalysed by F6′H1 and COSY, leading to the production of scopoletin. Subsequent conversion to fraxetin and sideretin occurs through the activity of S8H and CYP82C4, respectively. Coumarins are secreted into the rhizosphere by transporters from the PDR family. Once secreted to the rhizosphere, coumarins assist plants in iron acquisition from the soil environment. The catechol coumarins fraxetin, sideretin, and esculetin can directly mobilise iron. Our model proposes a cooperative interaction between the fungal root endophyte F80 and the coumarin scopoletin, which enhances ferric iron mobilisation in the rhizosphere, upstream of the plant's FRO2‐IRT1 reductive iron uptake system. Ex planta data suggest that F80 mediates the conversion of scopoletin into the catechol coumarin esculetin, facilitating iron solubility and rescuing iron‐deficient plant growth (created in BioRender. Van dijck (2025) https://BioRender.com/b4x4h7s).

Our data suggest that ferric iron mobilisation is conferred through fungal conversion of the F80‐inhibitory noncatechol coumarin scopoletin to the noninhibitory catechol coumarin esculetin, which is a potent Fe³⁺ chelator (Figs 3, S4, S5) (Schmid et al., ²⁰¹⁴; Schmidt et al., ²⁰¹⁴; Paffrath et al., ²⁰²³). Because fungal‐mediated scopoletin conversion also occurs in the rhizosphere during plant colonisation (Fig. 4c), it is likely to represent a key mechanism by which the fungus enhances iron availability. The scopoletin‐to‐esculetin transition not only enhances iron solubility but might also represent a ‘win–win’ scenario benefiting both the fungal endophyte and its host: the fungus mitigates scopoletin's antimicrobial inhibitory effects while the plant gains access to more bioavailable iron. This scenario is supported by (1) ex planta time‐course experiments that showed an initial delay in F80 growth in the presence of scopoletin, which reversed as scopoletin is converted into esculetin by F80; and (2) cooperative iron mobilisation by scopoletin and F80 yielding similar levels of mobilised iron as those provided by esculetin alone (Figs 3, S4). Other catechol coumarins have been shown to positively affect the growth of rhizosphere pseudomonads (McRose et al., ²⁰²³). Host rescue from limiting iron was entirely dependent on coumarin production, as no rescue was observed in the coumarin‐deficient f6′h1 mutant (Fig. 2). By contrast, s8h mutants, which produce scopoletin but lack fraxetin and sideretin, supported fungal‐mediated rescue of iron deficiency (Fig. 2). Thus, scopoletin appears to be necessary and sufficient for plant iron nutrition in the presence of this fungal root endophyte. A role of scopoletin in iron mobilisation was further confirmed by chemical complementation experiments. Scopoletin supplementation restored F80‐mediated rescue of f6′h1 plants to similar levels as the catechol coumarins or the synthetic iron chelator EDTA in mock conditions (Fig. 4). Collectively, these findings highlight a new type of plant–microbe interaction in which a fungal endophyte repurposes a plant‐derived coumarin precursor to generate bioavailable iron‐chelating compounds.

Coumarin production and exudation are highly induced upon iron starvation (Fourcroy et al., ²⁰¹⁴; Schmid et al., ²⁰¹⁴; Sisó‐Terraza et al., ²⁰¹⁶; Tsai et al., ²⁰¹⁸; Paffrath et al., ²⁰²³). Current literature suggests that fraxetin and sideretin are the primary coumarins responsible for iron acquisition (Rajniak et al., ²⁰¹⁸; Tsai et al., ²⁰¹⁸; Paffrath et al., ²⁰²³; Robe et al., ²⁰²⁵). Based on their exudation profiles and ability to chelate and reduce iron in varying pH conditions, fraxetin appears to be more important at alkaline pH whereas sideretin primarily functions at more acidic pH (Paffrath et al., ²⁰²³). Scopoletin, although not an efficient iron mobiliser, is, together with the redox‐active sideretin, the most abundant coumarin detected in Arabidopsis root exudates under acidic conditions (Sisó‐Terraza et al., ²⁰¹⁶; Rosenkranz et al., ²⁰²¹; Paffrath et al., ²⁰²³). Thus, a chief role of scopoletin in shaping the rhizosphere microbiome was thought to be through its inhibitory properties (Stringlis et al., ²⁰¹⁸). Our results demonstrate that scopoletin can also contribute indirectly to iron nutrition as a precursor to iron‐chelating coumarins generated by microbial conversion.

F80 was isolated from apparently healthy Arabidopsis plants in natural populations sampled in Europe (Durán et al., ²⁰¹⁸; Thiergart et al., ²⁰²⁰). Previous studies demonstrated that this fungal strain alleviates phosphate starvation in agar‐plate assays (Mesny et al., ²⁰²¹), suggesting that F80 may also be relevant in mitigating other forms of nutrient stress. These and our data reinforce the notion that F80 can exist as a neutral or beneficial endophyte associated with Arabidopsis roots. Conversely, M. phaseolina strains are well‐documented plant pathogens responsible for charcoal rot, a disease affecting over 500 plant species including crops such as maize, soybean, and canola (Su et al., ²⁰⁰¹; Bandara et al., ²⁰¹⁸; Sinha et al., ²⁰²²). The ability of M. phaseolina F80 to degrade scopoletin might have ecological implications, for example, in promoting plant iron nutrition in certain environments and/or conferring microbial competition for niche establishment. Esculetin, for example, was found to inhibit certain fungal pathogens (Beesley et al., ²⁰²³). Notably, scopoletin has been implicated in plant defence as its increased accumulation was associated with enhanced resistance against pathogens (Zamioudis et al., ²⁰¹⁴). Therefore, fungal‐mediated degradation of scopoletin could benefit pathogens by mitigating plant defence (Sun et al., ²⁰¹⁴; Gao et al., ²⁰²⁴). At this stage, it remains unclear whether scopoletin conversion by F80 is driven by a specific enzymatic activity. Alternatively, the fungus might generate a physicochemical environment that facilitates nonenzymatic demethylation to esculetin, as previously observed during coumarin‐mediated iron oxide dissolution at the surface of iron minerals (Baune et al., ²⁰²⁰). Scopoletin was found to be degraded in four different unsterilised soils, indicating that this trait might be ecologically relevant and prevalent (Galán‐Pérez et al., ²⁰²¹).

Microbe‐assisted nutrient acquisition is an emerging concept in plant nutrition, with many studies focusing on bacterial siderophores as key contributors to plant iron uptake (Kramer et al., ²⁰¹⁹). In this context, the role of fungal endophytes remains largely unexplored. To our knowledge, this study provides the first mechanistic basis for an endophytic fungus directly enhancing plant iron nutrition through its interaction with a plant‐derived coumarin. It therefore represents a significant advance in our understanding of how commensal fungal associations can strengthen plant nutritional resilience and productivity. Given the widespread occurrence of endophytes in natural and agricultural ecosystems and scopoletin‐secreting dicot plants, future research might identify similar coumarin transformations in other plant–microbe associations.

In conclusion, we reveal here a fungal endophyte contribution to plant iron nutrition. By demonstrating that M. phaseolina (F80) converts scopoletin into bioactive iron‐chelating coumarins, this study expands our understanding of how plant–microbe interactions influence nutrient acquisition. Fungal endophytes such as F80 could be valuable allies in enhancing plant resilience to nutrient stress, particularly in iron‐deficient soils. Alongside the role of bacteria in iron nutrition (Harbort et al., ²⁰²⁰), our findings highlight the cross‐kingdom significance of coumarins in microbiota‐mediated plant iron acquisition. The insights gained might have implications for developing sustainable agricultural practices that leverage beneficial microbial partnerships to enhance plant nutrition and stress resilience.

Competing interests

None declared.

Author contributions

LVD conceived the study with AP and JEP. LVD designed and performed the experiments and analysed the data. DE performed and analysed metabolite profiling. CH performed and analysed the ferrozine iron mobilisation assay. MM provided materials, analytical tools, and experimental advice. RFHG supervised the ICP‐MS analyses. GUB and AT supervised metabolite profiling. LVD and JEP wrote the paper with inputs from all authors.

Disclaimer

The New Phytologist Foundation remains neutral with regard to jurisdictional claims in maps and in any institutional affiliations.

Supporting information

Dataset S1 Raw data excel file.

Fig. S1 Heat‐killed fungal endophyte F80 does not improve Arabidopsis growth under iron‐limited conditions.

Fig. S2 The concentration of mineral elements in shoots measured by ICP‐MS.

Fig. S3 Phenotypes of mutants disrupted in coumarin biosynthesis or reductive iron uptake under available iron conditions in the presence or not of the fungal endophyte F80.

Fig. S4 Changes in metabolite profile of M. phaseolina F80 cultures when supplemented with scopoletin.

Fig. S5 Coumarin supplementation does not rescue fro2, while Na₂EDTA supplementation rescues f6'h1 but not fro2.

Table S1 Medium composition of ARE and vitamin solutions used in 96‐well fungal culture assay.

Please note: Wiley is not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.

Data availability

Data supporting the findings of this study are available in Figs S1–S5, Table S1, and Dataset S1.