Strigolactones modulate iron uptake in plants
Model for strigolactone regulation of basal iron uptake in Arabidopsis and maize.
Model for strigolactone regulation of basal iron uptake in Arabidopsis and maize.

Iron is an essential micronutrient for plants and animals alike. Insufficient iron intake leads to iron deficiency anemia that affects billions of people world-wide (Turawa et al., 2021). Iron is also required by plants and is critical for Chl biosynthesis, photosynthesis, respiration and many other processes. Although iron is abundant in soil, its low solubility makes it poorly bioavailable; thus, plants have evolved mechanisms that increase availability (Marschner & Römheld, 1994). Most plants use an acidification-reduction strategy to acquire iron from the soil (Strategy I), while plants belonging to the grass family (Poaceae) have evolved a distinct strategy (Strategy II) based on the secretion of iron-chelating molecules called phytosiderophores (PS). Grasses include most of the world's staple crops such as maize and rice. Maize mutants have been essential in identifying the core components of Strategy II iron uptake (von Wiren et al., 1994; Curie et al., 2001; Nozoye et al., 2013; Chan-Rodriguez & Walker, 2018).
To identify new genes involved in iron homeostasis in maize, we screened publicly available maize mutants for iron-deficiency (yellow-striped) phenotypes (Chan-Rodriguez & Walker, 2018). During this process, we identified a yellow-striped mutant that we named interveinal yellow (ivy) (Chan-Rodriguez & Walker, 2018). Leaves of ivy plants grown in soil in the glasshouse contained significantly low iron levels. Complementation was observed following crossing to both ys1 and ys3 mutants, indicating that IVY is a novel maize gene involved in iron homeostasis in maize (Chan-Rodriguez & Walker, 2018).
Strigolactones (SLs) have important roles in the rhizosphere (Waters et al., 2017), and are also integral to determination of plant form (Gomez-Roldan et al., 2008; Umehara et al., 2008). SL hormones derive from the carotenoid pathway (Matusova et al., 2005), and signal transduction has been well described (Seto, 2023; Hu et al., 2024; Zaidi et al., 2024). When SLs bind to their receptor (D14), an F-box protein (D3/MAX2) is recruited which targets D53/SMXL repressors for degradation (Jiang et al., 2013; Zhou et al., 2013; Yao et al., 2018; Li et al., 2022). D53/SMXLs are transcriptional repressors (Jiang et al., 2013) that restrict expression of SL-responsive genes (Waters et al., 2017; Liu et al., 2021; Guan et al., 2023). SLs are well known as one component of plant adaptation to nitrogen and phosphate limitation. Restricted N or P induces SL biosynthesis, which in turn promotes altered patterns of root development (Mayzlish-Gati et al., 2012; Sun et al., 2014; Wang et al., 2021). Recent work in flooded rice demonstrated that SLs regulate adaptive aerenchyma formation, whereas abscisic acid controls iron plaque formation, highlighting that the roles of SLs are context-dependent and may differ between wetland and upland species (Shrestha et al., 2026).
Here, we report mapping of the ivy mutation and identification of its causal gene, ZmCCD8, encoding an enzyme critical to production of SL hormones. A connection between SLs and iron homeostasis has not been previously reported, although symptoms of iron deficiency have been noted for the SL-deficient ccd8 mutant (Guan et al., 2023; Quaggiotti et al., 2024). The work presented here shows that loss of SL biosynthesis or signaling results in defects in basal iron homeostasis that are not linked to root architecture but instead result from downregulation of genes involved directly in iron uptake. Results show that SLs positively regulate iron homeostasis essential for optimal plant growth and development. This previously unappreciated role of SLs may have implications for the use of strigolactones in agricultural contexts.
The Zea mays L. ivy mutant seed stock was obtained from the Maize Genetics Cooperation Stock Center (MGCSC #3812O; ys*-N2398). The ccd8::Ds mutant allele was originally developed in a W22 inbred (Vollbrecht et al., 2010; Guan et al., 2012) and was subsequently introgressed into B73 inbred background by six generations of back-crossing. The ccd7 mutant was obtained from the Mu-Illumina project (mu-illumina_578516.6, Zm00001eb074640) (Williams-Carrier et al., 2010) and yellow-striped individuals were confirmed to be homozygous for a Mutator (Mu) insertion in the first exon (Williams-Carrier et al., 2010). The max4-6 mutant stock was previously described (Auldridge et al., 2006) and maintained in homozygosity. The wild-type Arabidopsis thaliana (L.) Heynh. accession Columbia-0 (Col-0) was used as control. The Mutator (Mu) insertion lines (W22 background) of maize D53 (Zm00001eb404750), D14a (Zm00001eb007670) and D14b (Zm00001eb400720) were initially characterized by Guan et al. (2023). The d53-2 mutant was identified in the UniformMu population (Settles et al., 2007) and has a Mu insertion (mu1094413) in the first exon, 11-bp downstream of the start codon. The d14a-4 mutant was identified from Mu-illumnina seed stocks distributed by the Maize Genetics Cooperation Stock Center through the Maize Genetics and Genomics Database (MaizeGDB). The mutant has a Mu insertion (Mu-illumina_232551.5) in the coding sequence of the first exon. The d14b-3 mutant was identified in the UniformMu population and has a Mu insertion (mu1078949) in the coding sequence of the second exon. To minimize effects of hybrid vigor or additional Mu insertions in the mutant lines, all experimental material was back-crossed to the W22 inbred for at least 2 (UniformMu) or 5 (Mu-illumina) generations. Chlorophyll content was quantified using glasshouse-grown material. The d14a, d14b double mutants were grown at the University of Florida research farm (PSREU, Plant Science, Research and Education Unit) located at Citra, Florida. The d53 and ccd8, d53 mutants were grown in a glasshouse with supplemental lighting (14 h : 8 h, light : dark) at 18°C to 32°C. Growth media (Happy Frog potting soil) was obtained from BWI in Apoka, FL.
For growth in the soil, maize or Arabidopsis seeds were planted in plastic pots with a 4 : 1 mixture of soil (Sun Gro® Professional Growing Mix) and Turface® MVP®, pretreated with Gnatrol (Valent BioSciences, Libertyville, IL, USA). Rice plants were planted in plastic pots with an equal mixture of soil and Profile® Greens Grade™. Arabidopsis plants were grown at room temperature (20–25°C) with 16 h : 8 h, light : dark. Maize and rice seeds were grown in the glasshouse with 16 h : 8 h, light : dark. For growth on plates, Arabidopsis seeds were sterilized (70% ethanol, 0.05% Triton X-100) and grown on sterile culture plates as described previously (Bakirbas & Walker, 2022). For growth in hydroponics, maize seedlings were grown as described previously (Fili & Walker, 2023). Throughout this article, the term ‘hydroponic culture’ refers to plants growing in solution with normal iron concentration and the term ‘iron-deficient hydroponic culture’ refers to plants growing in solution with no added iron. For supplementation of synthetic SL in the hydroponic culture (+/−)-GR24 (rac-GR24) was used (G3324; PhytoTech Labs, Lenexa, KS, USA) at a 1 μM final concentration (Guan & Koch, 2020). The hydroponic solution was renewed every 3 d.
Morphological measurements in ivy mutants were performed in siblings from F2 populations generated by crossing ivy to W22 or B73. The W22 population consisted of 107 individuals (25 mutants, 82 wild-type) and the B73 population consisted of 139 individuals (39 mutant, 100 wild-type). Plants were grown in the field and all measurements were performed when the plants reached maturity.
For quantification of crown root emergence in the ivy mutants, seedlings were grown in hydroponic culture. One set of ivy plants was supplemented with rac-GR24 every 3 d in the hydroponic solution (ivy + rac-GR24), while the other set of ivy plants received mock treatment (ivy). The length of each crown root was measured on day 9. Six biological replicates were used per genotype.
To obtain DNA for genotyping of ivy mutants and Sanger sequencing a modified version of the Dellaporta miniprep was used (Dellaporta et al., 1983). For genotyping of ccd7 mutants using TaqMan assays, a modified quick DNA extraction protocol was used (Wang et al., 1993).
To obtain RNA for gene expression measurements with qRT-PCR, total RNA was extracted using the RNeasy Plant Mini Kit (Qiagen) with on-column DNAse treatment. At least three biological replicates were used per genotype or condition for all gene expression measurements.
cDNA was synthesized from at least 500 ng of total RNA using SuperScript IV VILO Master Mix (Invitrogen™, Thermo Fisher Scientific, Waltham, MA, USA). qRT-PCR reactions were set up using the PowerUp™ SYBR™ Green Master Mix (Applied Biosystems™; Thermo Fisher Scientific) in MicroAmp™ Fast Optical 96-Well Reaction Plates (Applied Biosystems™; Thermo Fisher Scientific). The reactions were run on a StepOnePlus™ Real-Time PCR System (Applied Biosystems™; Thermo Fisher Scientific). Relative expression was determined using the ΔΔCT method with CT values of target genes normalized to ZmGAPDH.
For additional analysis of key transcripts in RNA-seq profiles of ccd8, 5.5 μg of the same total RNA used for the RNA-seq libraries were first treated with RQ1 RNase-free DNase (Promega) to confirm (via negative controls) that there was no carryover of genomic DNA. For quantitative PCR, a Power SYBR green RNA-to-CT, 1-step kit (Applied Biosystems) was used with an iCycler iQ real-time PCR detection system (Bio-Rad). The 18S rRNA was used as an internal reference. Primer sequences were adapted from (Nozoye et al., 2013). Statistical significance of differential expression was tested by analysis of variance (ANOVA).
The original ivy stock was crossed to the W22 inbred. After self-crossing the F1 generation, a segregating F2 population was generated. The population was grown in the field and scored for the yellow-stripe phenotype. Leaf collection was performed from a total of 120 yellow-stripe individuals and 120 wild-type individuals. To obtain high-quality genomic DNA for BSA-seq, the CTAB method for DNA extraction was used as described previously (Healey et al., 2014). Library preparation and sequencing were performed by Novogene (Hi-seq 4000; Illumina, San Diego, CA, USA). The quality of the raw reads was assessed using FastQC (Andrews, 2010). Adapter sequences and low-quality reads were removed using Trimmomatic (Bolger et al., 2014) with an average quality cutoff of 24. Trimmed reads were aligned to the maize W22v2 genome (assembly version: Zm-W22-REFERENCE-NRGENE-2.0) using the Bowtie2 aligner (Langmead & Salzberg, 2012). Variant calling was performed using SAMtools mpileup and variants were filtered using a base quality above 20 and a base coverage above 8 (Li et al., 2009). Allele frequency was calculated and plotted using custom R scripts. Variants were also identified using Varscan (Koboldt et al., 2012) for downstream SNP analysis. All SNPs within the mapping region were extracted and filtered only for homozygous in the mutant pool. SNPs that were also homozygous in the wild-type pool were filtered out. Naturally occurring variants were further filtered out using the available genomes of the following maize inbreds: A632, B73, B104, EP1, F7, Mo17, Oh43 and PH207 (Hirsch et al., 2016; Jiao et al., 2017; Springer et al., 2018; Sun et al., 2018; Haberer et al., 2020; Wang et al., 2023). The region corresponding to the mapping interval in W22 was identified in each inbred through synteny and multiple alignment was performed using Mauve (Darling et al., 2004) and setting the W22 region as reference. The effect of each of the remaining SNPs on the protein product was analyzed using SNPeff (Cingolani et al., 2012), identifying moderate and high-effect SNPs. For moderate-effect SNPs their impact on the protein function was further analyzed using PROVEAN (Choi & Chan, 2015). All genome assembly and annotation files were retrieved directly from MaizeGDB (maizegdb.org).
For transcriptomic analysis of iron deficiency and resupply, wild-type plants (W22) were grown in hydroponic culture for 12 d and then switched to +Fe or −Fe for 3 d to induce an iron deficiency environment. After 3 d, one set of −Fe plants was resupplied with hydroponic solution containing 0.1 mM FeSO4-EDTA (±Fe plants), while the +Fe and −Fe plants received new solution. After 24 h, the roots and shoots were sampled separately and immediately frozen in liquid nitrogen. Three biological replicates were used for each condition. Total RNA from roots and shoots was isolated using the Direct-zolTM RNA MiniPrep Kit (Zymo Research, Irvine, CA, USA) with DNase I treatment. The extracted RNA was quantified and assessed for quality using an Agilent 2100 Bioanalyzer. Library preparation and sequencing were performed by Novogene (Hi-seq 4000; Illumina). The adapter sequences and low-quality reads were removed using Trim Galore (Andrews, 2010; Martin, 2011; Krueger et al., 2023). Trimmed reads were aligned to the maize W22v2 genome (assembly version: Zm-W22-REFERENCE-NRGENE-2.0) using the HISAT2 aligner (Kim et al., 2015). Transcript assembly and quantification were performed using StringTie (Pertea et al., 2015). Differential gene expression analysis was performed using DESeq2 (Love et al., 2014). The cutoff criteria that were used were log2(|FC|) ≥ 1 and Benjamini–Hochberg (FDR) adjusted P-value ≤ 0.01 (Benjamini & Hochberg, 1995). Genome assembly and annotation files of the W22v2 genome were retrieved directly from MaizeGDB (maizegdb.org).
For analysis of expression profiles of ccd8 mutants, both wild-type B73 and ccd8 mutant seedlings were grown in plastic containers (diameter = 11.2 cm & height = 9.4 cm) with propagation mix for growth media (Sun Gro®). The growth chamber environment was set at 25°C with a diurnal cycle of 14 h : 10 h, light : dark (150 μmol m−2 s−1). Root and shoot tissues were sampled from seedlings at 14 d after planting, immediately frozen in liquid nitrogen and stored at −80°C until use. Tissues were ground in liquid nitrogen before extraction and purification of total RNA using plant RNeasy (Qiagen). The resulting RNA was quantified by NanoDrop 1000 (Thermo Fisher Scientific). Libraries for RNA-seq were prepared with a TruSeq sample preparation kit (Illumina) and sequenced by Hi-seq 2000 (ICBR, University of Florida). The data were analyzed using the Tuxedo Suite with statistical significance of differentially expressed genes set at FDR ≤ 0.01 (geometric normalization).
For measurements in Arabidopsis, seeds were germinated and grown on ½ × MS agar (no sucrose), containing low iron (10 μM) for 12 d at 22°C, with 16 h : 8 h, light : dark. Rosettes were then collected, weighed and placed into glass tubes with screw caps containing 1 ml of N,N-dimethylformamide (DMF) overnight. Six biological replicates were used per genotype. The total Chl content was calculated as Chl = 17.9A647 + 8.08A664.5 (Inskeep & Bloom, 1985).
For maize seedlings grown in hydroponic culture, a SPAD meter was used (SPAD 502 Plus Chlorophyll Meter 2900PDL; Konica Minolta, Ramsey, NJ, USA). Sampling was performed as previously described, by averaging all measurements within a leaf (Süß et al., 2019). Six biological replicates were used per genotype.
Leaves of at least 10 individual plants growing in hydroponic culture at the V4 stage were collected. The tissues were rinsed multiple times with double-distilled water (ddH2O). The samples were dried at 65°C for 72 h. Metal concentrations were determined by inductively coupled plasma mass spectrometry (ICP-MS) at the Donald Danforth Plant Research Institute. The translocation factor was calculated by determining the total Fe content in each tissue (whole shoot and whole root). To do this, the entire shoot or root system was weighed, then frozen in liquid nitrogen and ground to a fine powder. A portion of that powder was used in ICP-MS, and the total Fe in the shoot and root was then calculated. The translocation factor was then expressed as the ratio of total shoot Fe to total root Fe.
All statistical analyses were performed using custom R scripts in RStudio. When comparing the means between two groups, either the Student's t-test (unpaired and unequal variance) was used or the Mann–Whitney test. When comparing the means between three groups, ANOVA was used with post hoc Tukey Honestly Significant Difference (HSD). The rstatix package was utilized for statistical analysis (Kassambara, 2020).
The ivy mutant seed stock was obtained from the Maize Genetics Cooperation Stock Center (MGCSC #3812O; ys*-N2398), which listed it as a yellow-stripe mutant with a recessive pattern of inheritance, generated by EMS mutagenesis (Chan-Rodriguez & Walker, 2018). Individuals from the ivy Maize COOP stock were self-pollinated, and their progeny were crossed to inbred lines W22 and B73 to produce segregating populations. Segregating F2 populations showed 1 : 3 phenotypic ratios. The yellow-striped phenotype of the ivy mutants is similar to that of the canonical mutant ys1, which has a mutation in the gene encoding the transporter of iron phytosiderophore complexes (Fig. 1a). The low iron phenotype described previously for soil-grown plants (Chan-Rodriguez & Walker, 2018) was confirmed here using F2 plants grown hydroponically. The leaves of ivy mutant plants had low iron (Supporting Information Fig. S1a), while no significant difference in the iron concentration in the roots of mutant and WT siblings was observed (Fig. S1b). We calculated the translocation factor, which captures the outcome of differential rates of translocation of iron from root to shoot, but we observed no difference between the two genotypes, indicating that translocation of iron from root to shoot is not impaired in ivy mutants (Fig. S1c).

The ivy mutant. (a) Phenotypes of ivy, ccd8 and ys1 yellow-striped mutants of maize. (b) Plot of homozygous variants on chromosome 3. (c) CCD8 gene in maize and relative location of the stop codon introduced in ivy. (d) Exact location of the nonsense mutation in the ivy mutant.
The same F2 population was used for bulked-segregant analysis coupled with sequencing (BSA-seq) to identify the mutation. Reads from individually sequenced mutant and wild-type pools were aligned to the W22 reference genome (version: Zm-W22-REFERENCE-NRGENE-2.0) to reveal homozygous variants among mutant but not wild-type siblings (Fig. S2). Multiple homozygous regions were present on several chromosomes. Based on our prior experience analyzing BSA-seq datasets, we recognized that many of these regions represented background homozygosity commonly observed in W22 mapping populations. Indeed, all but one of these regions were also homozygous in at least two independent BSA-seq analyses performed previously in our laboratory. We therefore focused on the single region uniquely associated with the ivy mutation, located on the long arm of chromosome 3 (Figs 1b, S2). We used SNP filtering to exclude non-EMS type mutations and rejected variants evident in any of the eight sequenced maize lines (Hirsch et al., 2016; Jiao et al., 2017; Springer et al., 2018; Sun et al., 2018; Haberer et al., 2020; Wang et al., 2023) and excluded those, as they are unlikely to be responsible for the ivy phenotype when homozygous. Finally, we analyzed effects of remaining SNPs at the protein level and selected candidate genes with mutations predicted to have moderate to high impacts on the function of the encoded protein.
Out of 32 candidates meeting all the above criteria, only 2 mutations were predicted to strongly affect the encoded protein. One of the mutant genes (Zm00004b018959) in the W22 annotation; Zm00001eb153000 in (B73v5) encodes the carotenoid cleavage dioxygenase, CCD8, a key enzyme in the biosynthesis of SLs (Sorefan et al., 2003; Gomez-Roldan et al., 2008; Umehara et al., 2008) (Fig. 1c); the other was in the gene Zm00004b018907, annotated as a hypothetical ATPase, which is present only in the W22 annotation. Sequence comparison between mutant and wild-type DNA revealed a single nucleotide C to T substitution in codon 408 of the CCD8 CDS in ivy mutants (Fig. 1d). This mutation causes a premature stop codon (Gln408*/572), resulting in a truncated protein that is predicted to be 164 amino acids shorter than wild-type.
The iron-deficient, yellow-striped phenotype of ivy plants was visually identical to that previously characterized for a ccd8 mutant (Guan et al., 2012). To independently confirm that the ivy phenotype was caused by mutation of CCD8, we crossed ivy mutant individuals with an independent ccd8 mutant allele and performed an allelic complementation test. Four independent crosses between ivy and ccd8 mutants were used to confirm complementation. All progeny shared the yellow-striped phenotype (Fig. 2a). Both ivy (Fig. 2b–d) and ccd8 (Fig. S3) mutants showed growth defects typical of SL deficiency (Guan et al., 2012), including more tillers, short stature and a higher tassel-to-plant length ratio.

Phenotype and non-complementation of ivy and ccd8 mutants. (a) Results of complementation crosses between ivy and zmccd8 mutants of maize. Leaves from F1 progeny (CC1, CC2, CC3 and CC4) compared to ys1, ivy and zmccd8 homozygotes. Soil-grown plants 19 d after sowing. (b–d) Morphological measurements in ivy mutants and wild-type siblings. Significance was determined using Student's t-test (unpaired, unequal variance). Boxes represent the interquartile range (25th–75th percentiles), with horizontal lines indicating the median. Whiskers indicate the range of the data. Asterisks (*) indicate significance (***, P < 0.001; ****, P < 0.0001).
Mutant ivy/ccd8 plants have fewer crown roots and many genes involved in iron uptake are strongly expressed in this root type (Stelpflug et al., 2016; Fig. 4a). To test whether ivy/ccd8 plants are anatomically impaired in iron uptake, we examined other mutants with defective crown root development: rootless concerning crown and seminal roots (rtcs) and rootless1 (rt1; Hetz et al., 1996; Hochholdinger et al., 2004; Taramino et al., 2007). Neither mutant had yellow stripes (Fig. S4b,c). The CCD8 enzyme converts β-apo-10′-carotenal to carlactone (Fig. 3a). However, because CCD8 is an enzyme, we considered the possibility that an additional biochemical activity of CCD8, independent of strigolactone biosynthesis, could contribute to the iron-deficient phenotype. To test this possibility, we used the synthetic SL analog, rac-GR24, to determine if this SL could reverse chlorosis of ccd8 mutants. We grew ivy and wild-type W22 plants in hydroponic culture with normal iron nutrition for 10 d. Half the ivy plants received rac-GR24 in the medium. Efficacy of the rac-GR24 treatment was confirmed by quantifying emergence of crown roots in treated and untreated plants (Fig. S5). Chlorosis was assessed visually and as relative Chl content (Gianquinto et al., 2004; Hawkins et al., 2007). Chlorophyll levels of ivy plants supplemented with rac-GR24 were higher than untreated counterparts, which had significantly less than wild-type controls (Fig. 3b,c). The rescue of chlorosis by exogenous strigolactone supports the conclusion that the phenotype of ivy/ccd8 results from strigolactone deficiency rather than from an unexpected additional enzymatic function of CCD8 outside of strigolactone biosynthesis. If the phenotype were caused by the absence of another metabolite produced by CCD8, application of rac-GR24 would not be expected to restore the wild-type phenotype.

Corroborating evidence that CCD8 is the causal gene for the ivy mutant phenotype. (a) The strigolactone biosynthetic pathway. The gene product is given in green; gene names are given in the rice (blue) and Arabidopsis (purple) nomenclature (Booker et al., 2005; Zou et al., 2006; Arite et al., 2007, 2009; Nelson et al., 2011; Alder et al., 2012; Zhang et al., 2014; Abuauf et al., 2018; Wang et al., 2020). (b, c) Hydroponically grown maize ivy mutant plants were treated with rac-GR24 (orange bars) or were mock treated (pink bars). Chlorosis was assessed in a 3-d window for the second (b) and third (c) leaves. Mock-treated W22 plants (green bars) served as controls. Statistical analysis was performed using one-way ANOVA with post hoc Tukey HSD (n = 6). Boxes represent the interquartile range (25th–75th percentiles), with horizontal lines indicating the median. Whiskers indicate the range of the data. Asterisks (*) indicate significance (*, P < 0.005, **, P < 0.01). (d) Comparison of maize ccd7 and ccd8 (Williams-Carrier et al., 2010) mutant phenotypes in flag leaves. (e) Structure of the maize CCD7 gene and location of the Mutator insertion present in the ccd7 mutant (mu-illumina_578516.6).
We next examined whether other SL-related mutants in maize exhibit iron-deficiency phenotypes. First, we looked at a second gene from the SL pathway: CCD7. The CCD7 enzyme is a carotenoid cleavage dioxygenase like CCD8 but acts further upstream in SL biosynthesis (Booker et al., 2004; Pan et al., 2016) (Fig. 3a). The ccd7 phenotype was identical to that of ivy and ccd8 mutants when grown alongside one another in the field (Fig. 3d). These results further support a causal role for SL deficiency in the yellow-stripe phenotype of ivy/ccd8 plants.
A key question was whether mutation of the SL signaling system alone could also confer the iron-deficient phenotype. To test this, we first used a double mutant in which both maize genes for the SL receptor (D14) were dysfunctional. This receptor-deficient mutant (double mutant for d14a and d14b) had a clear, yellow-striped phenotype (Fig. 4a). We next examined a d53 mutant. The mechanism of SL signaling depends on degradation of the D53 repressor, a process hypothesized to release iron-related transcription regulators. If so, an absence of SL signals would leave the D53 repressor bound to iron-regulated genes and prevent their induction by iron deficiency. By contrast, an absence of D53 (as in the d53 mutant) would allow expression of iron-induced genes regardless of whether SL was present or not. In this way, the iron-deficient yellow-stripe phenotype of ivy/ccd8 could be rescued. As predicted, both d53 single and the ccd8 d53 double mutant showed no chlorosis (Fig. 4b,c). Collective genetic evidence thus indicates that the iron-deficient, yellow-striped phenotype arises specifically from a lack of SLs and from aberrant SL signals.

Genetic evidence that strigolactone signaling affects iron deficiency chlorosis. (a) Maize strigolactone (SL) receptor d14a d14b double mutants have a yellow-striped phenotype. (b) Genetic epistasis in maize d53 ccd8 double mutants. (c) Chlorophyll levels in maize ccd8, d53 and ccd8 d53 double mutants. (c) Chlorophyll levels in ccd8, d53 and ccd8 d53 double mutants. Statistical analysis was performed using one-way ANOVA with posthoc Tukey Honestly Significant Difference (HSD). Bars with the same letter above are not significantly different from one another. Asterisk (*) indicates significance (P < 0.05). The error bars are ± standard error of the mean (SEM; n = 10 for each genotype.)
To understand the mechanism through which SLs affect iron homeostasis, we first identified the set of maize genes that respond to iron deficiency. W22 plants were grown in three conditions: +Fe (iron supplied), −Fe (iron starved) and ±Fe (iron resupplied after a period of starvation). The resupplied iron condition was used to distinguish those genes that respond quickly to changes in iron availability in the environment.
A total of 216 genes in the roots and 71 genes in the shoots were differentially expressed between −Fe and +Fe plants (|FC| ≥ 2, P-adj (FDR) ≤ 0.01). In roots, 188 transcripts increased in abundance, and 28 transcripts decreased in abundance, while in shoots, 15 transcripts increased in abundance, and 56 transcripts decreased in abundance during iron deficiency. These define a set of iron-responsive transcripts in maize (Dataset S1).
Notably, we found that genes encoding SL biosynthetic enzymes or those involved in SL signaling (CCD8/Zm00001eb153000; D53/Zm00001eb404750; D14A/Zm00001eb007670; D14B/Zm00001eb400720; CCD7/Zm00001eb074640; ZmMax1b/Zm00001eb123090; ZmCLAMT1/Zm00001eb123070; and ZmCYP706C37/Zm00001eb123110) were not among the iron-regulated genes that we identified, highlighting the idea that iron levels do not modulate the levels of SL-associated genes.
During iron resupply, the majority (178 in roots and 63 in shoots) of transcripts returned to basal levels of expression after 24 h of iron resupply. We consider these fast-responding transcripts as likely to be directly involved in iron homeostasis during iron deficiency, and indeed, many of these transcripts encode proteins with familiar roles in iron uptake or homeostasis.
To better understand the influence of SLs on iron homeostasis, we explored expression of iron-regulated genes in the ccd8 mutant (in a B73 background) by comparing transcriptomes of wild-type (B73) seedlings and SL-deficient ccd8 mutant seedlings. Overall, in ccd8 plants, 99 genes in shoots and 151 genes in roots were significantly upregulated twofold or more, while 62 genes in shoots and 915 in roots were downregulated (Dataset S2).
Using the previously generated list of iron-responsive genes in maize (Dataset S2), we identified all mRNAs that were both responsive to iron deficiency and misregulated in ccd8 transcriptomes. In addition, we considered genes with previously identified roles in iron homeostasis, even if they were not formally included in the set of iron-responding transcripts. Transcripts for iron homeostasis were among those significantly downregulated by SL deficiency and included most genes for phytosiderophore biosynthesis (those for generating precursors from the S-adenosyl-L-methionine cycle and 7 out of 10 nicotianamine (NA) synthases (ZmNAS1;1~6;2; Fig. 5a)). The downregulation of many genes involved in Strategy II uptake – specifically in the biosynthesis of phytosiderophores – is consistent with the idea that a major effect of SL is in the uptake of iron from soil. This is in agreement with our finding that translocation of iron from root to shoot is not strongly affected in the mutant (Fig. S1c). In ccd8 roots, 30 iron-responsive genes were misregulated, while in shoots there was only one (Fig. 5b; Table S1). The observation that c. 3% of genes misregulated in ccd8 mutants are iron-responsive (30/1066), compared to only c. 0.7% iron-regulated genes in the total genome (216/32000), suggests that iron-regulated mRNAs are over-represented among the ccd8 misregulated genes in roots (hypergeometric P-value = 4.23−11). Importantly, 22 of the 30 genes downregulated in ccd8 mutant roots are normally upregulated in iron-deficient roots (Fig. 5c; Table S1). Thus, in the strigolactone-deficient ccd8 mutant, genes associated with iron uptake are expressed at abnormally low levels.

Iron-responsive gene expression in maize ivy/ccd8. (a) The pathway of phytosiderophore biosynthesis in maize. Enzymes and genes downregulated by twofold or more in roots are shown in blue. (b) Venn diagrams showing the comparisons between the two RNA-seq datasets used in this study. (c) Differentially expressed genes in root tissues overlapping the two datasets. (d) RT-qPCR analysis of selected genes involved in iron uptake. Levels of ZmIRO2;2 and ZmYS1 mRNAs were evaluated in both shoot and root tissues. Values are means ± SD (n = 3). Asterisks indicate statistically significant differences between wild-type and the zmccd8 mutant (P < 0.01, t-test). (e) Expression of IRO2;1 in wild-type and mutant plants during iron deficiency. Significance was determined using Student's t-test (unpaired, unequal variance). Asterisks (*) indicate significance (*, P < 0.05; **, P < 0.01).
Notably, transcripts encoding orthologs of a key rice iron-regulatory transcription factor, OsIRO2 (Ogo et al., 2007, 2011) (ZmIRO2;1 and ZmIRO2;2) are also downregulated. In our RNA-seq dataset, ZmIRO2;1 is an iron-responding transcript, while ZmIRO2;2 was only mildly upregulated and did not pass strict criteria as an iron-responsive transcript. A previous transcriptomic study using microarrays showed a twofold change for ZmIRO2;2 in –Fe roots (Zanin et al., 2017). To confirm and extend the RNA-seq data, the expression change of ZmIRO2;2, ZmYS1, ZmNAS6 and ZmDMAS1 in ccd8 mutants was determined using quantitative real-time RT-PCR (Fig. 5d).
Next, we asked whether ivy/ccd8 mutants appropriately alter gene expression in response to limited iron supply. Plants from an ivy-segregating population were grown in hydroponic culture, and iron was withheld from half the plants. In both ivy mutant and WT roots, growth in -Fe medium markedly increased the level of ZmIRO2;1 mRNA (Fig. 5e). Likewise, ZmIRO2;1 mRNA increased after growth in -Fe medium in both ivy mutant and WT shoots. This result indicated that ivy mutants can respond vigorously to exogenously imposed iron deficiency, that is, withdrawal of iron from the growth medium, even though basal levels of iron-responsive genes are abnormally low in ivy plants during growth in normal, iron-sufficient medium or soil.
We tested whether the external provision of synthetic SL to ivy mutant plants could raise the mRNA expression levels of iron-regulated genes to wild-type levels. We grew ivy and wild-type W22 plants in hydroponic culture. One set of ivy plants was provided with rac-GR24 in the hydroponic media, while this synthetic SL was withheld from control ivy and W22 wild-type plants. We then measured expression of ZmIRO2;1, ZmNAS2, ZmNAS6, ZmYS1 and ZmTOM1 in the roots. As expected, ZmIRO2;1 expression in ivy roots was more than fourfold lower than in wild-type. Notably, ivy plants treated with rac-GR24 expressed ZmIRO2;1 at levels indistinguishable from wild-type (Fig. 6a). Results were similar for ZmNAS2, ZmNAS6 and ZmYS1 (Fig. 6b–d). For ZmTOM1, the expression rose significantly above wild-type in the presence of rac-GR24 (Fig. 6e). These findings suggest that SLs can alter expression of these iron-regulated genes.

The effect of SL application on iron-regulated gene expression in ivy mutant plants. One set of ivy plants was supplemented with rac-GR24 in the hydroponic solution (ivy + rac-GR24), while the other set of ivy plants received mock treatment (ivy). Gene expression was measured using RT-qPCR. Statistical analysis was performed using one-way ANOVA with post hoc Tukey HSD. Three independent biological replicates were included for each control or experimental condition. Error bars for RT-PCR data were calculated asymmetrically, with the lower error representing the difference between the mean fold change and the minimum fold change of the control, and the upper error representing the difference between the maximum fold change of the control and the mean fold change. Asterisks (*) indicate significance (*, P < 0.05, **, P < 0.01, ***, P < 0.001). (a) RT-PCR of ZmIRO2;1. (b) RT-PCR of ZmNAS2. (c) RT-PCR of ZmNAS6. (d) RT-PCR of ZmYS1. (e) RT-PCR of ZmTOM1 (YS3).
The unexpected relationship between iron-deficiency phenotypes and a lack of SLs was initially puzzling due to an absence of chlorosis noted in previous studies of ccd8 mutants in either rice or Arabidopsis (Auldridge et al., 2006; Arite et al., 2007). We thus re-examined ccd8/d10 and d14 mutants of rice and ccd8/max4 mutants of Arabidopsis. When the rice mutants were grown in normal culture, an obvious chlorotic phenotype was observed (Fig. 7a). Although soil-grown Arabidopsis mutants lacked visible chlorosis, a pronounced chlorotic phenotype was observed in plants grown on plates with 10 μM instead of the standard 50 μM iron (Fig. 7b–d). The wild-type control plants grown in the same 10 μM iron condition were not chlorotic (Fig. 7b,c). These results demonstrate that the role of SL hormones in iron homeostasis is not unique to maize and extends across other monocot and dicot species.

Strigolactone (SL) mutants in rice and Arabidopsis. Asterisks (*) indicate significant difference between wild-type and mutant (*, P < 0.05). (a) Leaves of rice d10 (ccd8) and d14 mutants compared to wild-type (background ‘Shiokari’). (b) Arabidopsis Col-0 growing in 10 μM iron. (c) Arabidopsis max4 (ccd8) mutants growing in 10 μM iron. (d) Chlorophyll concentration of shoots from Arabidopsis plants grown in 10 μM iron. Significance was determined using pairwise Student's t-test (unpaired, unequal variance). (e) RT-PCR comparing AtFIT expression in WT (Col0) and smxl6 smxl7smxl8 triple mutant plants. (f) RT-PCR comparing ZmIRO2;1 expression in WT (B73) and d53 mutant plants. Three independent biological replicates were included for each control or experimental condition. Error bars for RT-PCR data were calculated asymmetrically, with the lower error representing the difference between the mean fold change and the minimum fold change of the control, and the upper error representing the difference between the maximum fold change of the control and the mean fold change.
In Arabidopsis, gene expression after treatment with the SL analog rac-GR24 has been examined (Wang et al., 2020). Further, the sites of binding of a key negative regulator involved in SL signaling, SMXL6, have been examined using ChIP-seq (Wang et al., 2020). Interestingly, the Arabidopsis FIT gene, which is well characterized as an essential component in iron deficiency gene regulation (Bauer et al., 2007; Yuan et al., 2008; Fig. 6a), is not only upregulated following rac-GR24 treatment, but also SMXL6 binds to the upstream portion of AtFIT (Wang et al., 2020). Moreover, AtFIT mRNA expression is downregulated in the SMXL7-D mutant, a stable gain-of-function mutant in which SMXL7 is not degraded normally after SL stimulation (Hu et al., 2020). These observations support the model for SL regulation of iron homeostasis illustrated in Fig. 8. In the absence of SL, no complex is formed between the SL receptor D14, the ubiquitin ligase SCFmax2 and the SMXL6 negative transcription regulator. This allows SMXL6 to interact with the AtFIT gene, suppressing its transcription, which in turn is expected to decrease basal iron uptake. In ccd8 mutants, the lack of SL stabilizes SMXLs, permanently lowering iron uptake, leading to chlorosis. In the presence of SL hormones, the hormone receptor complex is formed, SMXL6 is degraded, and AtFIT is released from repression. A direct implication of this model is that in smxl6/7/8 mutants, FIT should be released from repression and expressed at higher levels. SMXL6 is functionally redundant with two other closely related Arabidopsis genes, SMXL7 and SMXL8 (Soundappan et al., 2015). We tested this using RT-PCR, and observed that, as expected, AtFIT expression is significantly increased in smxl6/7/8 triple mutants (Fig. 7e).

Model for strigolactone (SL) regulation of basal iron uptake in Arabidopsis and maize. Upon SL binding to its receptor D14, the F-box protein MAX2 (Arabidopsis; D3 in rice and maize) promotes assembly of an SCF E3 ubiquitin ligase complex that targets the transcriptional repressors SMXL6/7/8 (Arabidopsis) or the orthologous repressor D53 (maize) for degradation. In Arabidopsis, SMXL6 binds the FIT promoter and represses its transcription (Wang et al., 2020), thereby reducing expression of iron uptake genes. In maize, FIT does not appear to be regulated by SL. Instead, expression of IRO2, the bHLH transcription factor that partners with FIT, is SL-dependent. We propose that D53 represses IRO2 expression, such that SL-induced degradation of D53 relieves this repression and promotes expression of genes required for iron uptake.
Our transcriptomic data from the maize ivy/ccd8 mutant suggested that the gene encoding the binding partner of ZmFITs, ZmIRO2;1 (Fig. 6c) is the most clearly downregulated transcription factor when SLs are lacking. Neither of the two maize FIT genes were significantly downregulated in ccd8 plants. We hypothesize that the target of SL regulation in maize is not FIT, as in Arabidopsis, but instead is its binding partner, IRO2 (Fig. 8). We examined ZmIRO2;1 expression in a maize d53 mutant. In maize, a single gene, D53, corresponds to the three closely related Arabidopsis genes AtSMXL6, AtSMXL7 and AtSMXL8. As predicted, the level of ZmIRO2;1 mRNA increases in the d53 mutant (Fig. 7f).
Strigolactones (SLs) are plant hormones known for regulating development, but their roles in nutrient homeostasis are less clear. This study reveals that SLs are essential for maintaining iron balance in maize, as loss of SL biosynthesis leads to chlorosis and disrupted iron gene expression. Similar effects in Arabidopsis and rice suggest a conserved role across species. These findings uncover a previously unknown connection between SL hormone signaling and iron regulation, offering new insight into the mechanisms that plants use to regulate iron uptake.
Using BSA-seq, we tentatively concluded that the mutated locus responsible for a novel iron-deficient maize mutant, interveinal yellow (ivy), encodes CCD8, a key enzyme in SL biosynthesis (Fig. 1). Genetic complementation with a previously characterized ccd8 mutant (Guan et al., 2012) confirmed this finding (Fig. 2). This reveals a previously unknown connection between SL hormones and iron homeostasis. The yellow-striped phenotype of a maize ccd7 mutant identified in this study further confirms that genetically impaired biosynthesis of SL leads to iron deficiency symptoms. Moreover, control of this phenotype by SL signals alone is demonstrated by blocking transduction (SL receptor double mutant) and by initiating signals in the absence of SL (ccd8 with a mutated D53 transcriptional repressor). Our findings in maize extend to both rice and Arabidopsis, where chlorosis emerges when SL-deficient mutants are grown in conditions where the WT is not chlorotic.
Our results suggest that strigolactones regulate the Strategy II iron acquisition pathway in maize. Expression of ZmIRO2 and its downstream targets, including ZmNAS2/6, ZmDMAS1, ZmTOM1 and ZmYS1, was reduced in the ccd8 mutant and restored by rac-GR24 treatment (Figs 5, 6), consistent with strigolactones acting upstream of the transcriptional network controlling phytosiderophore-mediated iron uptake. Although we did not directly measure 2′-deoxymugineic acid (DMA) production or secretion, these findings suggest that SL deficiency may impair Strategy II iron acquisition through reduced expression of genes required for phytosiderophore biosynthesis, transport and uptake.
Strigolactones (SLs) have traditionally been viewed as regulators of plant growth and development in response to environmental conditions. For example, phosphate and nitrate deficiency stimulate SL exudation from roots, promoting arbuscular mycorrhizal associations (Yoneyama et al., 2007a,b) and reshaping root architecture (Arite et al., 2012; Sun et al., 2014, 2021). In addition, the abundance of certain intermediary metabolites influences SL levels and signaling to better align metabolism and growth (Bertheloot et al., 2020; Tal et al., 2022). Together, these findings support a model in which SLs help tune plant growth to prevailing abiotic conditions (recently reviewed in Barbier et al., 2023). Iron appears to differ from other nutrients in its relationship to SL. For phosphorus and nitrogen, nutrient deficiency stimulates SL production, and SL in turn influences nutrient homeostasis (and likely patterns of growth as well). By contrast, iron deficiency does not appear to induce SL biosynthesis. Instead, our results suggest that SLs provide a constitutive regulatory input that maintains the basal expression of genes required for iron acquisition. In this model, SLs are not an inducible signal of iron deficiency but rather are required to sustain the IRO2-dependent transcriptional network that enables normal iron uptake. Loss of SL biosynthesis therefore compromises this basal regulatory program, resulting in insufficient iron acquisition and chlorosis even in the absence of severe iron stress.
To explore how a mutant CCD8 enzyme can cause an iron-deficient phenotype in maize, we considered two distinct hypotheses. First, we posited that CCD8 could have a second, poorly understood enzymatic activity outside of the SL biosynthetic pathway. The observation that supplying synthetic SL to mutants alleviated chlorosis indicated that the iron-deficient phenotype is caused specifically by the lack of SLs (Fig. 3). The chlorotic phenotype of a second SL biosynthetic mutant, ccd7, also corroborates this idea. The carotenoid pathway is also important for abscisic acid (ABA) and glycoside synthesis, raising the concern that disruption of SL synthesis may lead to imbalanced ABA levels. However, the yellow-striped, iron-deficient phenotype of d14 (SL receptor) mutants (Figs 4, 7) indicates a direct link with SL hormone signaling. The second hypothesis that we considered was that the lack of crown roots caused altered iron uptake. To test this, we examined other mutants that lack crown roots, but found that these do not exhibit iron deficiency phenotypes (Fig. 4b,c).
Instead, our data indicate that SL modulates expression of iron-responsive genes, particularly those upregulated by iron deficiency. The expression of such genes is low in ccd8 mutants (Fig. 5). Application of the synthetic SL hormone rac-GR24, increases expression of these genes (Fig. 6), and, importantly, application of rac-GR24 to ccd8 mutants leads to greening of their otherwise chlorotic leaves (Fig. 3). It is important to note that rac-GR24 can activate both the strigolactone (D14) and karrikin (KAI2) signaling pathways, introducing some ambiguity into the interpretation of these experiments. However, the ccd8 mutants used in this study are specifically defective in SL biosynthesis, and the rice and maize d14 mutants used here are specifically impaired in SL perception while retaining an intact karrikin signaling pathway. Taken together, these complementary genetic and physiological data support the conclusion that strigolactones directly or indirectly modulate the expression of genes involved in iron uptake.
Mutants incapable of producing SL have been studied in both rice and Arabidopsis. We re-examined the rice d10 (ccd8) and d14 (SL receptor) mutants and observed chlorosis in normal culture conditions, indicating that in rice, too, the inability to make or sense SL hormones leads to iron deficiency. Although Arabidopsis ccd8 (max4) mutants are not chlorotic on soil, when iron levels were restricted (to 10 μM), chlorosis was evident. WT plants under the same conditions were green with abundant Chl. Moreover, the absence of the key transcriptional repressors associated with SL signaling, in both Arabidopsis and rice, caused increased expression of key iron-deficiency regulators in each species. The impact of SL hormones on iron thus extends to plant species including both monocots and dicots.
A potential mechanism linking SL signaling to basal iron uptake is suggested by Arabidopsis studies showing that AtFIT is induced by the synthetic SL analog rac-GR24 and that the SL-associated repressor SMXL6 binds the AtFIT promoter, implicating SL signaling in the transcriptional regulation of iron uptake genes (Wang et al., 2020) (Figs 8, S6a). We propose that SL signaling similarly promotes iron acquisition in maize (Fig. 8), but through regulation of IRO2 rather than FIT, as FIT transcript levels are unaffected in maize SL mutants whereas IRO2 expression is markedly reduced (Figs 5, 6). These findings identify IRO2 as a candidate downstream target of SL signaling in the regulation of basal iron uptake in maize.
We thank Chris Phillips and Dan Jones at the UMass Morrill Greenhouses for assistance with plant husbandry. This work was supported by the US Department of Agriculture, National Institute of Food and Agriculture, 2020-67013-30910, and the Department of Biology at the University of Massachusetts Amherst, under project no. MAES00595 to ELW and by the NSF, NSF-IOS-PGRP-1116561, NSF-IOS-PGRP-1521100, NSF-IOS-PGRP-1748105 and NSF-IOS-PGRP-2435687 to KEK.
None declared.
ELW, SF, J-CG and KEK conceived and designed the research. SF, J-CG and RK performed the research and data collection. ELW, SF, J-CG and KEK contributed to data analysis and interpretation and to writing and editing of the manuscript.
The data that support the findings of this study are available in Figs S1–S6 and Datasets S1 and S2. RNA-seq data for iron-regulated gene expression are available at the NCBI GEO database under accession no. GSE280702. The RNA-seq data for ccd8 and B73 root tissues are available in the NCBI database (BioProject PRJNA757767) under accession nos. SRR15613590, SRR15613591, SRR15613599, SRR15613593, SRR15613594 and SRR15613595.
References
The New Phytologist Foundation remains neutral with regard to jurisdictional claims in maps and in any institutional affiliations.
Texto de New Phytologist tal como lo publicó, leído de su página y mostrado aquí con el formato de El Autómata. El original, en nph.onlinelibrary.wiley.com · Volver a la noticia de El Autómata