Sulfur in dialogue with phosphorus, nitrogen, and iron: regulatory networks in plant nutrient homeostasis Authors: Afm Mohabubul Haque, Hideki Takahashi, Hatem Rouached Published: 2026 ABSTRACT AbstractSulfur (S) is an essential macronutrient for plant growth and resilience. The S-amino acids cysteine (Cys) and methionine (Met) are indispensable for protein synthesis and structural integrity, as well as redox homeostasis and cofactor assembly. Over the past several decades, biochemical and molecular genetic studies demonstrated the core steps in sulfate uptake and assimilation pathways, while it has become increasingly evident that S homeostasis in plants cannot be understood in isolation. Robust and reciprocal regulatory interactions link S with phosphorus (P), nitrogen (N), and iron (Fe). Plants remodel membrane lipid compositions, replacing the phospholipids with sulfolipids under P deficiency. Cys/Met biosynthesis is coordinated with N metabolism. The Fe-S cluster assembly requires a balanced supply of Fe and S. These interactions are orchestrated through shared regulatory circuits and specific hub-regulatory transcription factors, including SULFUR LIMITATION 1, PHOSPHATE STARVATION RESPONSE 1, NIN-LIKE PROTEIN 7, and FER-LIKE IRON DEFICIENCY-INDUCED FACTOR. Comparative studies reveal both species-specific and evolutionarily conserved regulatory networks. This review focuses on mechanistic insights into the regulatory circuits revealed from studies with the model plant Arabidopsis thaliana, where the genetic and molecular resolution enabled detailed dissection of the signaling and regulatory networks. This review also highlights unresolved mechanistic gaps and provides insights into systems-level understanding and potential translational approaches that can be implemented to improve crop nutrient use efficiency and stress resilience. Introduction Sulfur (S) is an essential macronutrient that underpins multiple facets of plant metabolism and physiology (Zenda et al., 2021). Plants use photosynthetically captured energy to assimilate inorganic sulfate (SO42−) from the soil into a broad spectrum of organic compounds, including the amino acids cysteine (Cys) and methionine (Met) which are essential for protein synthesis and protein structural conformation (Ma et al., 2024). The thiol group of Cys enables the formation of protein disulfide bonds and mediates redox regulation. In addition, Cys serves as the precursor of glutathione (GSH; γ-glutamyl-cysteinyl-glycine), a tripeptide that functions as a central hub for redox cycling and xenobiotic detoxification. This versatile redox interchange provides plants, like other aerobic organisms, with a flexible mechanism to alleviate oxidative stress (Saito, 2004). Beyond redox homeostasis, S-containing metabolites have important roles in abiotic stress adaptation and defense. For example, phytochelatins derived from GSH chelate and sequester toxic heavy metals not only in vascular plants, but also in certain fungi, algae, diatoms, and invertebrates (Rea, 2012). S metabolism also supports essential cellular processes through the production of S-adenosylmethionine, which donates methyl groups for nucleic acid and protein methylation (Lee et al., 2023). Moreover, S is incorporated into key cofactors such as lipoic acid, biotin, and iron–sulfur (Fe–S) clusters, which are indispensable for enzyme catalysis and energy metabolism, as well as sulfolipids, polysaccharides, flavonoids, and specialized metabolites such as glucosinolates and camalexin contributing to chemical defense against pathogens and herbivores, further underscoring its central role in plant physiology (Nakai and Maruyama-Nakashita, 2020; Kopriva et al., 2024). S deficiency imposes clear physiological and developmental constraints (Bouranis and Chorianopoulou, 2023). Above ground, plants exhibit reduced biomass, chlorosis, and impaired photosynthesis upon S deficiency. Below ground, the root system architecture can be altered: while the primary root extends, the lateral roots proliferate near the root tip in response to shortage of S supply. These S-deficiency phenotypes are characteristic of demand-driven adaptive strategies plants implement to enhance S acquisition from the soil environment (Koprivova and Kopriva, 2016; Aarabi et al., 2020; Narayan et al., 2023). For vascular plants, the inorganic sulfate (SO42−) is the primary S source available in the environment (Wawrzyńska and Sirko, 2024). The canonical machinery of sulfate uptake and transport has been well characterized in the model plant Arabidopsis thaliana (hereafter Arabidopsis), as represented by function of distinct members of the SULFATE TRANSPORTER (SULTR) gene family: The high-affinity sulfate transporters SULTR1;1 and SULTR1;2 mediate sulfate uptake across the plasma membrane of the root epidermis and the outermost cell layers in contact with soil mineral nutrients known as the cortex (Takahashi, 2019). The low-affinity transporter SULTR2;1 mediates long-distance sulfate transport from root to shoot, and is co-expressed with SULTR3;5 in the vasculature, while it also partially contributes to sulfate uptake, along with SULTR1;1 and SULTR1;2, in the root epidermis and cortex under S deficiency (Kataoka et al., 2004a; Maruyama-Nakashita et al., 2015; Takahashi, 2019). The tonoplast-localizing SULTR4;1 and SULTR4;2 control internal sulfate storage and distribution, promoting sulfate transport from root to shoot by releasing sulfate from the vacuole (Kataoka et al., 2004b). These essential steps in sulfate uptake and internal S distribution are orchestrated by the transcription factor SULFUR LIMITATION 1 (SLIM1)/ETHYLENE-INSENSITIVE 3-LIKE 3 (EIL3) (Maruyama-Nakashita et al., 2006), and, in part, its downstream effector microRNA395 (miR395). miR395 can specifically target and fine-tune spatiotemporal expression of SULTR2;1 and chloroplast-localizing ATP sulfurylase (ATPS1, ATPS3 and ATPS4), the first committed enzyme of the S assimilation pathway in the vasculature under S deficiency (Kawashima et al., 2009; 2011; Liang et al., 2010) (Fig. 1). Our current model suggests that, under S deficiency, SULTR2;1 is transcriptionally activated in roots and promotes sulfate uptake and root-to-shoot sulfate transport (Maruyama-Nakashita et al., 2015). miR395 expressed in the phloem post-transcriptionally silences the target gene transcripts, i.e. SULTR2;1 and ATPS1/3/4, in nearby phloem parenchyma cells and restricts S translocation to sink organs (e.g. root tips) (Kawashima et al., 2009; 2011; Liang et al., 2010). These transcriptional and post-transcriptional mechanisms are operated in parallel to control S distribution. Beyond these established transport functions, a possible ‘transceptor’-like role, defined as the ability to mediate nutrient signaling independently of transport activity, has been proposed for SULTR1;2 based on genetic and biochemical observations; however, direct experimental uncoupling of transport and signaling functions has not yet been demonstrated (Shibagaki and Grossman, 2010; Zhang et al., 2014; Zheng et al., 2014). Collectively, these findings from molecular genetics, functional genomics, and biochemical investigations reveal a sophisticated system for sulfate acquisition and its homeostatic regulation pivotal for plant acclimation to varying S conditions in the environment. Integration of sulfur (S), phosphorus (P), nitrogen (N), and iron (Fe) signaling pathways into coordinated regulatory networks underlying nutrient-dependent phenotypes. The transcription factor PHR1/PHLs (PHOSPHATE STARVATION RESPONSE 1/PHR1-LIKE proteins) functions as a central regulator coordinating phosphorus (P) homeostasis with nitrogen (N), iron (Fe), and sulfur (S) signaling pathways. Under P deficiency, PHR1/PHLs induce the expression of SPX domain-containing proteins (SYG1/Pho81/XPR1 domain proteins), microRNAs (miRs) and nutrient transporters to maintain nutrient balance. In the P signaling module, PHR1/PHLs positively regulate miR399 and miR827. These microRNAs repress PHO2 (PHOSPHATE2, an E2 ubiquitin-conjugating enzyme) and NLA (NITROGEN LIMITATION ADAPTATION, an E3 ubiquitin ligase), respectively, thereby modulating P uptake and translocation via PHO1 (PHOSPHATE1, a phosphate exporter) and PHT1 phosphate transporter. PHR1/PHLs also regulate expression of Pi transporter SPDT (SULTR-like P Distribution Transporter) responsible for xylem-to-phloem P distribution. In the N signaling pathway, PHR1/PHLs interact with N-regulated transcription factors HRS1 (GARP family transcription factor) and HHO1s (HRS1 HOMOLOG 1). These factors regulate gene expression of nitrate transporters NRT2.1, NRT2.4 and NRT2.5 responsible for nitrate uptake. N signaling components further include NF-YA (NUCLEAR FACTOR Y subunit A), which is controlled by miR169, and CBL7 (CALCINEURIN B-LIKE PROTEIN 7). The transcription factor NLP7 (NIN LIKE PROTEIN 7) is one of the master regulators involved in nitrate response and plant developmental reprograming, and it is required for induction of gene expression of nitrate transporters (e.g. NRT1.1 and NRT2.1). The Fe regulatory module involves transcription factors bHLH38/39/100/101 (basic HELIX-LOOP-HELIX transcription factors 38, 39, 100 and 101) that interact with FIT (FER-LIKE IRON DEFICIENCY-INDUCED TRANSCRIPTION FACTOR) to regulate genes expression of Fe transporter IRT1 (IRON-REGULATED TRANSPORTER 1) and ferric-chelate reductase FRO2 (FERRIC REDUCTION OXIDASE 2) for Fe uptake and Fe storage protein FER1 (FERRITIN 1). PHR1/PHLs regulates FER1 expression (Bournier et al., 2013). Fe homeostasis is also regulated by ubiquitin ligases BTS/BTSL (BRUTUS/BRUTUS-LIKE proteins). In the S signaling pathway, SLIM1 (SULFUR LIMITATION 1) acts as a master regulator. The function of SLIM1 is essential for induction of gene expression of sulfate transporters SULTR1;1/SULTR1;2 and SULTR4;1/SULTR4;2 that are responsible for sulfate uptake in roots and unloading of sulfate pools in the vacuole, respectively. SLIM1 also positively regulates expression of miR395. This microRNA is expressed in the phloem and restricts S distribution to sink organs by post-transcriptionally targeting and silencing transcript expression of sulfate transporter SULTR2;1) and ATP sulfurylases ATPS1, ATPS3 and ATPS4. In parallel, SULTR2;1 controls root-to-shoot transport of sulfate together with SULTR3;5 coexpressed in the vasculature. Crosstalk between S and P signaling occurs through PHR1/PHLs-mediated regulation of SULTR2;1 and SULTR1;3. Solid arrows indicate positive regulation or activation, blunt-ended lines indicate repression, and dashed lines represent indirect or proposed regulatory interactions. The colored panels represent nutrient-specific signaling modules and their integration through PHR1/PHLs. The S assimilation process does not operate in isolation. Plants rarely experience strictly isolated nutrient limitations, and overlapping nutrient constraints are common in natural and agricultural environments necessitating coordinated regulatory responses (Bouain et al., 2019; Nam et al., 2021; Cho et al., 2026).Interactions of S with phosphorus (P), nitrogen (N), and iron (Fe) are particularly relevant given the metabolic fates of these nutrients. These interactions are likely driven by metabolic demands and shared transcriptional and post-transcriptional regulatory mechanisms integrated in part through signaling crosstalk. In this review, we place S at the center of plant nutrient response networks, emphasizing the molecular mechanisms and regulatory processes that coordinate multifactorial interactions among S, P, N, and Fe-dependent signaling pathways. By situating S-dependent pathways within the broader web of gene regulatory networks, we propose an integrated framework of S/P/N/Fe nutrient signaling interactions (Fig. 1). It is noteworthy that recent studies in Arabidopsis conceptualize such regulatory interactions but predominantly with a focus on P and N signaling pathways, defining the core regulatory principles and the network architecture (Krouk and Kiba, 2020; DeLoose et al., 2024; Ruffel et al., 2025). N limitation (−N) triggers a pronounced repression of microRNA169 (miR169), accompanied by the strong up-regulation of its targets, NUCLEAR FACTOR Y subunit A (NF-YA) transcription factors, highlighting the central role of the miR169-NF-YA regulatory module in N stress adaptation (Zhao et al., 2011). N signaling further integrates into broader regulatory networks. The PHOSPHATE STARVATION RESPONSE 1 (PHR1)/PHR1-LIKE proteins interact with N-responsive transcription factors, including a GARP (GOLDEN2, Arabidopsis RESPONSE REGULATOR-B (ARR-B), PHOSPHATE STARVATION RESPONSE 1 (PSR1)) family transcription factor HRS1 (HYPERSENSITIVE TO LOW Pi-ELICITED PRIMARY ROOT SHORTENING1) and HRS1 HOMOLOG 1 (HHO1), which collectively regulate the expression of high-affinity nitrate transporters such as NITRATE TRANSPORTER 2.1, 2.4 and 2.5 (NRT2.1, NRT2.4 and NRT2.5). Additional N signaling components, such as NIN LIKE PROTEIN 7 (NLP7) and CALCINEURIN B-LIKE PROTEIN 7 (CBL7), contribute to nitrate-dependent regulatory processes (Fig. 1) (Ruffel et al., 2025). This review sheds light on S-centered regulatory integrations with P, N and Fe signaling pathways in Arabidopsis, with emphasis on the genetic, molecular, and physiological mechanisms. Particularly, we focus on the core regulatory principles and model network architecture defined with pathway components identified from Arabidopsis. The conceptual model highlighted in this review aims to advance a systems-level understanding of mechanisms underlying S utilization and plant nutrition, in general. Sulfur-phosphorus interactions S and P nutrition form an integrated regulatory network that couples metabolic remodeling, long-distance transport, and plant hormone actions to stabilize plant performance when supply of these nutrients fluctuates. The MYB transcription factor PHR1 is the master regulator for the phosphate starvation response (Rubio et al., 2001; Bustos et al., 2010) (Fig. 1). PHR1, SPX1 (SYG1 (suppressor of G-protein alpha 1)/Pho81 (cyclin-dependent kinase inhibitor)/XPR1 (xenotropic and polytropic retrovirus receptor 1)), miR399, PHOSPHATE 2 (PHO2; an E2 ubiquitin-conjugating (UBC) enzyme), and PHOSPHATE TRANSPORTER (PHTs; high-affinity Pi transporters) define a phosphate-signaling pathway in plants (Bari et al., 2006) (Fig. 1). miR399, a phosphate (Pi) starvation-induced long-distance signal, is first produced in shoots and moves to roots to suppress PHO2 mRNA and consequently the E2 enzyme, leading to enhanced Pi uptake and root-to-shoot translocation (Chiang et al., 2023; Garcia et al., 2023). PHR1 also positively regulates miR827 expression (Guo et al., 2015). miR827 targets and silences NLA (NITROGEN LIMITATION ADAPTATION, an E3 ubiquitin ligase) mRNA post-transcriptionally, thereby modulating Pi transport (Kant et al., 2011). Besides its function in controlling the core phosphate-signaling pathway, PHR1 diminishes expression of a SULTR family member SULTR3;4 in shoots under P limitation (−P) (Rouached et al., 2011). SULTR3;4 was found to encode a Pi transporter SPDT (SULTR-like P Distribution Transporter) transferring P from xylem to phloem in Oryza sativa (rice) and Arabidopsis (Yamaji et al., 2017; Ding et al., 2020). PHR1 further contributes to reprograming of membrane lipid compositions, as it activates gene expression of UDP-sulfoquinovose synthase (SQD1) and UDP-sulfoquinovose:disacylglycerol sulfoquinovosyltransferase (SQD2) (Rouached et al., 2011; Pant et al., 2015), promoting biosynthesis of sulfoquinovosyldiacylglycerol (SQDG), replacing the phospholipids recycled under −P without compromising the membrane lipid bilayer function or exhausting the inorganic phosphate (Pi) pools (Essigmann et al., 1998; Yu et al., 2002; Pant et al., 2015). In contrast, under S limitation (−S), the S assimilatory fluxes and S transport activities are reweighted to sustain growth despite curtailed S supply. Importantly, the sulfolipid-to-phospholipid substitution is species- and context-dependent (Sato et al., 2017): i.e. a full replacement of sulfolipids by phospholipids under −S is limited in Arabidopsis, indicating the membrane remodeling capacity does not mirror the extensive sulfolipid substitution observed under −P (Essigmann et al., 1998; Yu et al., 2002; Pant et al., 2015). The S and P regulatory interactions can be seen in mechanisms implicated in the control of long-distance P transport. During −S, Arabidopsis rapidly elevates Pi acquisition and reallocates Pi to shoots, indicating that networks under control of plant S status can actively shape phosphate homeostasis (Allahham et al., 2020). The impact of −S on phosphate uptake and distribution in Arabidopsis manifests within 1–2 days and is largely transport-centric (Allahham et al., 2020). Short-term phosphate uptake assays show that plant exposure to −S results in increasing the radioactive phosphate (32P) influx into the roots by ∼3-folds relative to +S, accompanied by a measurable rise in the xylem-sap Pi and net Pi accumulation in the leaves even as the shoot biomass declines (Allahham et al., 2020). Genetic evidence pinpoints transport nodes: the −S-dependent surge of Pi in xylem requires function of two phosphate transporters, the PHOSPHATE1 (PHO1) and the PHOSPHATE TRANSPORTER 1;9 (PHT1;9), because the Pi surge is lost in the pho1 and pht1;9 mutants. However, the total shoot Pi still increases in these backgrounds, suggesting that additional, as yet unresolved, routes may contribute to Pi loading and redistribution (Rouached, 2011; Allahham et al., 2020). Three non-mutually exclusive possibilities have been proposed or inferred: (i) dampened shoot-to-root Pi remobilization (e.g. reduction in phloem Pi export); (ii) shifts in phloem Pi partitioning via transporters such as PHT1;5 and PHO1 homologs (PHO1;H); and (iii) enhanced vacuolar–cytosolic Pi cycling through the function of tonoplast carriers such as PHT5/VACUOLAR PHOSPHATE TRANSPORTER 1 (VPT1) that may increase cytosolic Pi availability without additional Pi uptake (Liu et al., 2016). The kinetics are reversible: resupplying sulfate restores xylem and shoot Pi toward the +S baseline values, indicating rapid, status-dependent regulation of Pi transport and allocation rather than slow developmental reprograming (Rouached, 2011; Allahham et al., 2020; Bouranis et al., 2020). Notably, several −S effects proceed largely outside the influence of canonical master regulators, SLIM1/EIL3 for S and PHR1/PHL1 for P utilization, highlighting alternative control points. Post-translational mechanisms (including ubiquitin-mediated endocytosis) control protein trafficking and turnover of the PHT1 phosphate transporter (Wang et al., 2021). SPX-inositol-pyrophosphate sensing gates PHR1 activity as a function of cellular energy and Pi charge (Wild et al., 2016; Ried et al., 2021). Thus, metabolic cues associated with altered cellular redox status and carbohydrate availability under −S may modulate Pi transport activity through post-translational and sensing mechanisms, providing alternatives to transcriptional induction of PHT1 transporters (Allahham et al., 2020). Together, these observations argue that plants maintain Pi homeostasis under −S not simply by ‘turning on’ more transcription, but by dynamically redistributing available Pi pools and modulating fluxes through existing Pi transport machineries, as previously demonstrated for phosphate transporter regulation under arsenate stress (Castrillo et al., 2013). The S and P interplay also impacts long-distance sulfate transport and S distribution (Sacchi and Nocito, 2019). The S regulatory network fine-tunes sulfate distribution through spatially organized transcriptional and post-transcriptional regulation via SLIM1, miR395, and PHR1 (Fig. 1). Under −S, induction of SULTR2;1 expression in roots promotes root-to-shoot sulfate transport (Kataoka et al., 2004a; Maruyama-Nakashita et al., 2015). In parallel, miR395 expressed and accumulated under control of SLIM1 silences transcript accumulation of SULTR2;1 and ATP sulfurylases (i.e. ATPS1, ATPS3 and ATPS4) in cells adjacent to the phloem, preventing distribution of sulfate and S metabolites from source to sink organs under −S (Kawashima et al., 2009, 2011; Liang et al., 2010). By contrast, under −P, SULTR2;1 is negatively regulated by PHR1 in shoots (Rouached et al., 2011). Since miR395 does not accumulate under −P, SULTR2;1 and ATPS1/3/4 transcripts could remain intact and may contribute to sustain source-to-sink S distribution through the phloem. Nevertheless, SULTR2;1 appears to have little impact on both root-to-shoot and shoot-to-root sulfate transport under −P (Rouached et al., 2011). Instead, SULTR1;3 impacts S distribution from shoot to root when it is expressed under control of PHR1 under −P (Rouached et al., 2011), corroborating an established role of this sulfate transporter in the phloem companion cells (Yoshimoto et al., 2003). Following sulfate transport to aboveground tissues, S needs to be internally distributed to meet demand for sulfolipid biosynthesis as the phospholipid content declines under −P. To fulfill sulfolipid biosynthesis that substitutes phospholipids under −P (Essigmann et al., 1998; Yu et al., 2002; Pant et al., 2015), the substrate sulfite from the reductive S assimilation pathway is expected to be exported from bundle sheath cells and redistributed throughout the leaf tissue where both SQD1 and SQD2 are present (Aubry et al., 2014). However, the molecular mechanism mediating sulfite transport and redistribution remains unresolved and is currently hypothetical. Functionally, the −S-triggered preference to load more Pi into shoots can be interpreted as a stabilizer for adenylate charge, nucleic acid synthesis, and ionic balance when sulfate and S-containing antioxidant GSH contents decline. By loading more Pi and promoting Pi-dependent metabolism in leaves, the plants preserve photosynthetic and photorespiratory performances, preventing energy crises that would otherwise amplify the S stress (Xiao et al., 2024; Zenzen et al., 2024). Conversely, under −P, allocating sulfate to shoots and remodeling the membrane lipid composition toward recycling and replacement of the phospholipids with SQDG safeguards the lipid bilayer function and redox capacity while freeing Pi for core metabolism. In both scenarios, the same architecture of the regulatory scheme is visible: the set-points are PHR-SPX-inositol-pyrophosphates and SLIM1-miR395 modules, hormonal gating prioritizing plant development, and a rapidly adjustable transport layer centered on PHO1, PHT1;9, and their phloem and vacuolar-localizing homologs of Pi transporters. What distinguishes the −S-to-Pi directed scenario is the prominence of post-translational mechanisms and Pi partitioning that can operate within hours, offering a plausible explanation for why shoot Pi rises even when xylem Pi does not in certain mutants (e.g. pho1). The system has parallel levers, and it will use whichever are available to keep the shoot supplied (Lapis-Gaza et al., 2014; Wild et al., 2016; Puga et al., 2024; Landaeta-Sepúlveda et al., 2025). These reciprocal adjustments through the function of PHR1 under −P and SLIM1/miR395 under −S illustrate how nutrient-specific transcriptional programs can be repurposed to support the opposing pathway while still allowing rapid metabolic or transporter-level regulation that operates independently of transcriptional latency. Metabolic signaling adds a further layer of integration. Inositol phosphates and inositol pyrophosphates couple adenylate status, Pi charge, and transcriptional control through the SPX-domain sensors that modulate PHR1 activity (Wild et al., 2016; Ried et al., 2021). Perturbations that lower phytic acid (inositol hexakisphosphate) commonly mis-accumulate both sulfate and phosphate transporters, suggesting that a shared integrator may be present to control both the S and P fluxes to be aligned with the cellular demand (Cominelli et al., 2020). Studies on plant hormone cytokinin signaling demonstrate its co-regulation of S and P uptake. Cytokinin mechanistically functions as a repressive signal, dampening expression of SULTR1;1/SULTR1;2 sulfate transporters and Pi transporters, as well as other PHR1 target genes (Maruyama-Nakashita et al., 2004a; Franco-Zorrilla et al., 2005; Wang et al., 2006; Bustos et al., 2010). The necessity of cytokinin receptors in these observations points to a mechanism coordinately regulating −S and −P-responsive genes; however, an alternative nutrient-specific mechanism driven by +S or +P appears to be responsible for systemic repression (Maruyama-Nakashita et al., 2004a; Franco-Zorrilla et al., 2005; Bustos et al., 2010). Through systemic repression, plants can avoid P over-acquisition when Pi supply is adequate, and may preserve S-P balance by preventing Pi-driven dilution or misallocation of sulfate. Cytokinin may also be acting as a repressive signal dampening P and S uptake and coordinating the N-S-P balance as it functions as a long-distance N-demand signal (Ruffel et al., 2011). We clearly find ethylene’s footprint on the core −S signaling network, i.e. the ETHYLENE-INSENSITIVE 3 (EIN3)-like identity of SLIM1 (Maruyama-Nakashita et al., 2006). Additionally, previous work demonstrated that EIN3 can inhibit SLIM1 activity through protein-protein interaction (Wawrzyńska and Sirko, 2016). Later investigations on sugar signaling suggest the hexokinase-dependent pathway is involved in glucose signaling, promoting −S-responsive gene expression (Wawrzyńska et al., 2022). These observations possibly implicate an outcome associated with EIN3 protein destabilization (Yanagisawa et al., 2003), a mechanism that would have mitigated negative impact of EIN3 on the interacting partner SLIM1. Sugars are known to positively control expression of both −S and −P-responsive genes (Maruyama-akashita et al., 2004b; Franco-Zorrilla et al., 2005; Karthikeyan et al., 2007; Bustos et al., 2010). Together, current knowledge illustrates P and S cues intersect and control Pi and sulfate uptake capacity. While evidence to support the causes of co-regulation remains fragmentary, possible links to cytokinin and ethylene signaling modules provide a framework to explore yet-to-be-determined network connections. Other plant hormones may also impose organ-level priorities atop these core −P and −S signaling mechanisms, linking the nutrient status to plant architecture and resource allocation strategies. Auxin and ethylene potentiate the plant response to phosphate starvation and promote development of lateral roots and root hairs, enlarging the absorptive interface for nutrient uptake (Růzicka et al., 2007; Neumann, 2016; Bhosale et al., 2018; Qin and Huang, 2018). Strigolactones are typically elevated under −P, and reallocate carbon to below-ground foraging while also suppressing shoot branching. In hosts that form arbuscular mycorrhiza (AM) symbioses, strigolactones signal to enhance symbiotic Pi capture, which may in turn shift S demand and long-distance S transport (Gomez-Roldan et al., 2008; Yoneyama et al., 2012; Kodama et al., 2022). As mentioned previously, the integrated frameworks generate concrete, testable predictions, including how perturbations of hormonal signaling, SPX-mediated Pi sensing, or vacuolar Pi exchangers should quantitatively alter shoot Pi accumulation and transport dynamics under −S, while several regulatory links remain unresolved. Perturbation of ethylene or auxin metabolism and signals may change the magnitude, but not the direction, of increases in shoot Pi accumulation under −S (Allahham et al., 2020), as these hormones modulate root-hair density and lateral-root formation (Růzicka et al., 2007; Neumann, 2016; Bhosale et al., 2018; Qin and Huang, 2018). Cytokinin elevation should constrain excessive Pi loading under −S by restricting PHR1-dependent regulatory programs to promote expression of PHT1 phosphate transporters, helping plants to maintain ionic balance while sulfate is unavailable. Manipulating the PHR-SPX-inositol-pyrophosphate sensing module is expected to decouple the Pi and −S signals integrated for Pi-transport plasticity, revealing whether shoot Pi loading under −S is limited by perceived Pi sufficiency (Satheesh et al., 2022; Yang et al., 2024). Finally, targeted disruption of vacuolar Pi exchangers (e.g. PHT5/VPT1) or phloem loaders (e.g. PHT1;5/PHO1;H) should define the relative contributions of vacuolar Pi storage and long-distance Pi redistribution to the observed rise in the shoot Pi content (Sun et al., 2023). Framed this way, as a reversible program that coordinates membrane lipid composition, metabolic demands, and long-distance transport, the S-P crosstalk reads not as two parallel stress responses, but as a coordinated homeostatic response tuned to the most limiting element and capable of rebalancing within hours to preserve whole-plant function. Sulfur-nitrogen interactions Biosynthesis of the S-amino acids Cys and Met occurs at the metabolic ‘gates’ where the S and N assimilation pathways converge (Watanabe et al., 2021; Kopriva et al., 2024). Cys is synthesized through condensation of sulfide from the reductive sulfate assimilation pathway and O-acetylserine derived from the Ser biosynthesis pathway. Ser acetyltransferase and O-acetylserine(thiol)lyase, the two enzymes responsible for Cys biosynthesis, assemble into a heteromeric Cys synthase complex (CSC), demonstrating an intriguing metabolic gating feature associated with actions of sulfide and O-acetylserine: CSC serves as the catalyst for O-acetylserine biosynthesis when sulfide accumulates and stabilizes the complex. CSC dissociates when O-acetylserine accumulates, and subsequently, the free O-acetylserine(thiol)lyase serves as the enzyme for Cys biosynthesis (Hell and Wirtz, 2011; Wirtz et al., 2012). Met is synthesized from Cys through three consecutive enzyme reactions, which involve cystathionine γ-synthase, cystathionine β-lyase, and Met synthase (Hesse et al., 2004; Hell and Wirtz, 2011). The first key enzyme reaction with cystathionine γ-synthase that generates cystathionine involves γ-replacement of the Asp-derived substrate O-phosphohomoserine while it combines with Cys. This draws attention to the fact that Met biosynthesis also occurs at a convergent point of S and N metabolism. When S is limiting, restricted metabolic flux toward Cys biosynthesis promotes incorporation of nitrate or ammonium, shifting the metabolic network to allocate N to free non-S amino acids and other metabolites (Nikiforova et al., 2005). Conversely, with sufficient S supply, N metabolic fluxes increase through glutamine synthetase/glutamate synthase and toward downstream pathways to synthesize amino acids, improving N use efficiency (Yu et al., 2021). The S and N metabolic pathways are therefore jointly regulated. This biochemical coupling is likely enforced by interaction of hub-regulatory circuits. On the N side, primary nitrate signaling via NLP7 and the dual-affinity sensor/transporter NPF6.3/NRT1.1, under control of the Calcineurin B-like protein-CBL-interacting protein kinase (CBL-CIPK) phosphorylation networks, activate expression of high-affinity transporters for nitrate uptake (NRT2.1/NRT2.4) (Fig. 1) and nitrate assimilation enzymes (nitrate reductase and nitrite reductase), with the outcome gated by carbon and redox status (Konishi and Yanagisawa, 2013; Wang et al., 2018; Deloose et al., 2024; Ruffel et al., 2025). The S status feeds into this module via SLIM1, which shapes nitrate allocation: the loss of SLIM 1 depresses NRT2.5 and elevates phloem nitrate loaders NRT1.7/NRT1.9, shifting the source-sink nitrate flow. SLIM1 also restrains expression of chloroplastic glutamine synthetase (GLN2), tying S signaling to the glutamine synthetase/glutamine oxoglutarate aminotransferase (GS/GOGAT) cycle (Ristova and Kopriva, 2022). At the same time, S-derived redox buffers set the limit for nitrate reduction throughput by maintaining ferredoxin supply in photosynthetic tissues (Hell and Wirtz, 2011; Takahashi et al., 2011). The overall outcome is a coordinated throttle: under S deficiency, plants down-shift nitrate uptake and reduction; when S is adequate, the ‘cysteine gate’ opens and the N assimilation cascade stabilizes (Kopriva and Rennenberg, 2004). This logic generalizes beyond Arabidopsis. In Triticum aestivum (wheat), coordinated S–N supply preserves grain protein composition: balanced S inputs restore GSH and Met pools and raise N recovery efficiency (Yu et al., 2021). In Zea mays (maize), S co-fertilization improves agronomic N use efficiency under nitrate-based regimes by sustaining chloroplastic assimilation capacity and preventing late-season leaf senescence (Li et al., 2019; Ma et al., 2021). In rice, allelic variation of the nitrate transporter OsNRT1.1B, which enhances nitrate responsiveness, amplifies yield benefits only when S is sufficient, consistent with the notion of S-gated ceiling on nitrate assimilation (Hu et al., 2015). These observations with crops underscore that the N response curve is S-limited in practice. The agronomic N gains saturate unless S supply keeps pace. Seed protein compositions are also tightly coordinated to meet this joint demand. The S-N imbalance caused by S deficiency therefore poses a significant impact on seed protein qualities (Bonnot et al., 2020; 2023). In wheat, S deficiency at a fixed rate of N fertilizer application elevates the grain N:S ratio, decreases synthesis of S-rich seed storage proteins, and degrades the baking quality (Yu et al., 2021). Earlier investigations with legumes, cereals, and other species demonstrate that S-rich seed storage proteins are dispensable, as S-poor alternates accumulate in S-deficient plants, attributable to gene expression of specific storage proteins that differ in S content (Chandler et al., 1983; Holowach et al., 1984; Wrigley et al., 1984; Gayler and Sykes, 1985; Fujiwara et al., 1992; Hirai et al., 1995). Notably, the tonoplast-localizing SULTR4 sulfate transporter facilitates the release of sulfate from the vacuole and contributes to the quality of Pisum sativum (pea) seeds under +S and the yield under –S (Bachelet et al., 2024). While the underlying regulatory mechanism for S sparing is not fully understood, physiological and molecular features of SULFUR DEFICIENCY INDUCED 1 (SDI1) and SDI2 address their pivotal function in transcriptional regulation (Aarabi et al., 2016; 2021). SDI1 and SDI2 are homologous transcriptional repressor proteins expressed under control of SLIM1 (Maruyama-Nakashita et al., 2006; Aarabi et al., 2016). They interact with MYB28 and MYC2 transcription factors, acting as repressors to inhibit aliphatic glucosinolate biosynthesis and expression of S-rich seed storage proteins. These findings imply a homeostatic control mechanism designed to economize S incorporation into S-rich compounds. Plant hormones provide the architecture for this integration. Auxin and ethylene promote lateral root and root-hair proliferation, increasing absorptive surface for nutrient uptake (Růzicka et al., 2007; Neumann, 2016; Bhosale et al., 2018; Qin and Huang, 2018). It is worthy of note that the NPF6.3/NRT1.1 nitrate transporter facilitates auxin uptake and may take part in the hormonal crosstalk that modulates lateral root development by sensing nitrate (Krouk et al., 2010). Plants respond to nutrients with reciprocal transcriptional programs that couple uptake and assimilation of these nutrients and long-distance signals translating the nutrient status (Ruffel, 2018). Cytokinin acts as a long-distance N-demand signal dampening nitrate and sulfate uptake, preserving S-N balance by preventing nitrate-driven dilution of S metabolites (Ruffel et al., 2011; Kiba et al., 2010). Functionally, three features define the S-N axis across species: (i) metabolic gating at Cys and GSH sets the admissible flux of nitrate reduction; (ii) allocation control via SLIM1 and nitrate transporters redistributes N between leaves, phloem, and sinks as S status changes; and (iii) hormonal priority rules determine where new biomass and transport capacity can be installed. In cropping systems, this means N recommendations should be paired with S targets to maintain the tissue N:S ratio within an optimal range; otherwise, excess N application raises nitrate pools rather than proteins, shortening the photosynthetic lifetime of leaves and reducing the grain quality. Sulfur-iron interactions Fe and S metabolism converge in Fe-S cluster assembly, the prosthetic core for electron transfer in photosynthesis and respiration. Because cluster biogenesis draws simultaneously Fe and S, limitation of either element depresses energy metabolism and heightens oxidative stress (Couturier et al., 2013; Balk and Schaedler, 2014). In strategy I dicots such as Arabidopsis and Solanum lycopersicum (tomato) Fe deficiency (−Fe) normally triggers the FER-LIKE FE-DEFICIENCY INDUCED TRANSCRIPTION FACTOR (FIT)-basic helix-loop-helix (bHLH) cascade (e.g. clade IVc bHLHs, clade Ib bHLHs, and BRUTUS E3 ligase), inducing expression of ferric-chelate reductase (FERRIC REDUCTION OXIDASE 2; FRO2) and high-affinity ferrous iron uptake transporter (IRON-REGULATED TRANSPORTER 1; IRT1) (Vert et al., 2002; Brumbarova et al., 2015) (Fig. 1). Strategy I Fe uptake functionally couples with PLEIOTROPIC DRUG RESISTANCE 9 (PDR9/ABCG37), supporting exudation of phenolics, and FERRIC REDUCTASE DEFECTIVE 3 (FER3) and NICOTIANAMINE SYNTHASE 4 (NAS4), mediating xylem-phloem partitioning of Fe (Robinson et al., 1999; Gao and Dubos, 2020; Trofimov et al., 2024). Plant S status rewires the output of this cascade. Under dual Fe and S deficiency (−Fe−S), IRT1, FRO2, PDR9/ABCG37, and NAS4 gene expression is dampened, while early activators for the −Fe response (i.e. FIT, clade IVc bHLHs) remain comparatively stable (Zuchi et al., 2009; Robe et al., 2020). This selective dampening of −Fe-responsive gene expression curtails an unintended influx of non-Fe divalent cations through broad-specificity transporters, such as NATURAL RESISTANCE-ASSOCIATED MACROPHAGE PROTEIN 1 (NRAMP1) for Mn (manganese) (Cailliatte et al., 2010) and IRT1 for Zn (zinc) and Mn (Korshunova et al., 1999; Vert et al., 2002; Dubeaux et al., 2018; Long et al., 2018), attenuating chlorosis (Fig. 2) and improving quantum yield relative to −Fe alone (Robe et al., 2020). These observations are consistent with the hypothesis that an S-responsive checkpoint may be embedded within the −Fe regulatory network rather than within a global plastid stress signaling pathway; however, direct molecular evidence for this checkpoint is still lacking (Estavillo et al., 2011; Tee et al., 2025). This model aligns with recent findings showing that, although Fe deficiency impairs photosynthesis and induces chlorosis, the severity of this chlorosis is strongly conditioned by P availability (Nam et al., 2021) (Fig. 2). In tomato, −S similarly depresses ferric-chelate reductase activity and radiolabeled Fe uptake, reducing leaf Fe accumulation (Zuchi et al., 2009; Coppa et al., 2018), reinforcing the notion that plant S status sets the safe operating range for Fe acquisition mediated by IRT1 (Korshunova et al., 1999; Vert et al., 2002; Dubeaux et al., 2018). A representative composite color illustration of Arabidopsis rosette leaf phenotypes under single and combined nutrient deficiencies. Plants grown under complete nutrition (+Fe+S+P) display uniform green rosettes, whereas those under Fe deficiency (−Fe+S+P) result in strong interveinal chlorosis. Combined Fe and S deficiency (−Fe−S+P) produces a chlorosis phenotype that is less pronounced in young leaves (Robe et al., 2020). Combined Fe and P deficiency (−Fe+S−P) causes a mild and uniform chlorosis phenotype affecting both young and mature leaves (Nam et al., 2021). These non-additive phenotypic outcomes indicate that nutrient interactions alter adaptive responses beyond the extent and spatial patterns expected from each single nutrient limitation. AI (ChatGPT) was used to create the Fig. 2. In strategy II grasses, S supply directly conditions biosynthesis of mugineic acids, the Fe(III)-chelating phytosiderophores synthesized from nicotianamine (Mori and Nishizawa, 1987; Kobayashi and Nishizawa, 2012). Since nicotianamine derives from S-adenosylmethionine, −S constrains both the biosynthesis and efflux of mugineic acids via the transporter of mugineic acid family phytosiderophores 1 (TOM1), thereby diminishing the uptake of Fe(III)-phytosiderophore complexes through YELLOW STRIPE1 (YS1) and its orthologs and homologs (Curie et al., 2001; Nozoye et al., 2011; Masuda et al., 2012; Yoneyama, 2021). In Hordeum vulgare (barley), restoring S supply rescues phytosiderophore release and the Fe-acquisition phenotype under −Fe stress; in durum wheat, −Fe co-induces activities of enzymes for the reductive sulfate assimilation and Cys biosynthesis pathways, such as ATP sulfurylase, adenosine-5′-phosphosulfate reductase, sulfite reductase, and Ser acetyltransferase, indicating an anticipatory ramp-up of S assimilation to support impending Fe-S cluster assembly once Fe becomes available (Astolfi et al., 2010, 2018; Bocchini et al., 2015; Celletti et al., 2016; Quagliata et al., 2025). Together, these findings show that grasses integrate S metabolism not only as a way to passively support cofactor supply, but as a rate-limiting determinant of Fe mobilization capacity. The function of sulfate transporters and the S allocation layers echo this S-Fe integration across clades. In dicots, sulfate influx (SULTR1;1/SULTR1;2), long-distance transfer (SULTR2;1/SULTR3;5), and vacuolar release (SULTR4;1/SULTR4;2) adjust S uptake and distribution under conditions with Fe perturbations, though the direction can be species-specific, e.g. SULTR1;1 suppression in Arabidopsis vs. induction of the orthologs observed in tomato (Paolacci et al., 2014; Khan et al., 2022). In grasses, S allocation to leaves under −Fe supports the S metabolic flux through ATP sulfurylase, adenosine-5′-phosphosulfate reductase, and sulfite reductase, which is followed by the pathway for Met biosynthesis and nicotianamine/phytosiderophore production, thereby coupling the shoot demand to root acquisition (Astolfi et al., 2003, 2021; Vigani and Briat, 2016; Khan et al., 2022). Furthermore, by maintaining GSH pools, plants can prevent chlorophyll degradation and secondary photochemical damage that otherwise amplifies −Fe symptoms (Ramírez et al., 2013). Across plants species, three mechanistic rules emerge. First, a S-responsive checkpoint modulates the Fe-uptake cascade downstream of early bHLH activators (Gao et al., 2020a; 2020b), prioritizing protection from non-Fe divalent cation overload during −Fe (Robe et al., 2020). Second, S assimilation is Fe-responsive, rising under Fe deficiency in both dicots and grasses to pre-stage Fe-S cluster capacity (Briat et al., 2015; Forieri et al., 2017). Third, strategy-dependent wiring matters: in grasses, the Met-nicotianamine-phytosiderophore axis (Kobayashi and Nishizawa, 2012) makes the S availability a direct determinant of Fe mobilization; in dicots, S primarily limits and safeguards intensity of reduction-based Fe uptake and long-distance Fe partitioning (Robe et al., 2020). These principles argue for awareness of crosstalk in Fe nutrition and developing application practice. In calcareous soils where Fe deficiency is chronic, maintaining S sufficiency preserves phytosiderophore release in cereals (Astolfi et al., 2010) and protects dicot leaves from Mn-driven photoinhibition (Robe et al., 2020). Conversely, under S scarcity, dialing back aggressive Fe-uptake programs, for example, through cultivar choice with tampered IRT1 induction or function, may prevent Mn/Zn toxicity while sustaining photosynthesis until S supply recovers (Korshunova et al., 1999; Vert et al., 2002; Dubeaux et al., 2018; Long et al., 2018). Conclusions and perspectives The dialogue between S, P, N, and Fe reveals a unifying principle: plant nutrition is governed not by parallel pipelines but by an interconnected control system that aligns membrane composition, redox balance, and cofactor assembly with resource availability. Across this network, Cys/Met biosynthesis, SQDG-phospholipid remodeling, and Fe-S cluster biogenesis act as load-bearing nodes. These processes are wired to shared regulatory hubs, SLIM1/EIL3, PHR1, NLP7, and FIT, and to nutrient transport and storage modules (e.g. SULTR families, PHO1/PHT1, and transporters for vacuolar nutrient mobilization) that redistribute the nutrient flux within hours. As a result, local changes in S, P, N source, or Fe status propagate into system-level reallocations of carbon source, photosynthesis-derived energy, and non-Fe metals, stabilizing growth and defense when any single nutrient becomes limiting. Three general principles emerge. First, nutrient ratios matter. The balance between N and S determines whether nitrate reduction translates into proteins or stalls at intermediate metabolite pools; the balance between P and S determines whether membranes retain function by substituting SQDG for phospholipids; the balance between Fe and S sets the capacity to assemble Fe-S clusters that power the photosynthesis and respiration. Second, cues and commands for control are layered. Transcriptional programs with the hub-regulatory transcription factors (i.e. SLIM1/EIL3, PHR1, NLP7, and FIT) establish the set-points to investigate potential integrators connecting signals under combined nutrient stresses (Fig. 3), small RNAs (e.g. miR395 and miR399) refine spatial deployment of signals for fine tuning, and post-translational switches tune transporter trafficking and enzyme activity to match the short-term demand for nutrients. Third, system plasticity is purposeful. Rapid reversibility, such as the recovery of xylem/shoot Pi upon sulfate resupply, signals a system built to sense, adjust and reset. A conceptual model illustrating the convergence of nutrient signaling pathways at the level of transcriptional regulation. The transcription factors SLIM1 (SULFUR LIMITATION 1), PHR1 (PHOSPHATE STARVATION RESPONSE 1), NLP7 (NIN LIKE PROTEIN 7), and clade Ib bHLH factors including FIT (FER-LIKE IRON DEFICIENCY-INDUCED TRANSCRIPTION FACTOR) each regulate distinct nutrient-responsive gene networks while also engaging in cross-regulatory interactions. The overlapping regions highlight shared downstream targets associated with nutrient transport, assimilation, and developmental adjustment. Although these regulatory modules are increasingly well described, the mechanisms by which multiple nutrient signals are integrated at the network level remain unresolved. The ‘Integrators?’ element denotes a conceptual placeholder for distributed and potentially redundant processes rather than a single central regulator. The schematic only focuses on core representative transcriptional regulatory modules, thus simplifying the full complexity of regulatory interactions with post-transcriptional processes and hormonal regulation additionally discussed in this review. Created in BioRender. Rouached, H. (2026) https://BioRender.com/8kcbcw3. Crop management can move from single-nutrient prescriptions to crosstalk-aware regimes in order to better co-optimize N:S:P:Fe supply as a vector rather than as independent scalars. In cereals, aligning N with S can preserve grain protein quality; in P-limited soils, maintaining S supply supports SQDG-mediated membrane resilience while in Fe-limited calcareous soils, ensuring S sufficiency prevents metal toxicity and safeguards phytosiderophore synthesis and Fe–S assembly. Breeding and gene editing can target conserved hubs (i.e. SLIM1/EIL3, PHR1, NLP7, and FIT) and their downstream transporters to widen the operating window under fluctuating inputs of nutrients, while safeguarding plant yield and micronutrient content. Molecular diagnostics tools that couple soil tests with plant readouts, representing transcript, protein, or metabolite signatures, could guide in-season adjustments before physiological penalties accrue. Key gaps define priorities for future research. We still lack the definitive identity of sensors that will allow us to compare combined nutrient status in vivo and resolve them into actionable signals. Candidate integrators for S and P sensing, such as Cys, O-acetylserine, sulfide, GSH, inositol pyrophosphates, and Fe-S cluster load, remain speculative and require causal testing that link their dynamics to SLIM1/EIL3 and PHR1 activity, transporter trafficking, and long-distance S/P allocation. Conservation versus diversification of these circuits across lineages (e.g. Arabidopsis, rice, maize, and legumes) remains uncertain, as do the network’s behaviors under combinations of abiotic stresses such as drought, salinity, and elevated CO2. Addressing these questions will require systems-level understanding of plant nutrient responses characterized with large-scale lab experiments and field tests that are factorial by design (i.e. N × S × P × Fe and water), resolved in time and space (e.g. cell types along the root–shoot axes), and integrated across modalities (e.g. transcriptome, metabolome, lipidome, metallome, and flux tracing). Coupling such datasets with mechanistic models capable of addressing emergent states will convert descriptive crosstalk into predictive control. In summary, S sits at the intersection of macronutrient (P and N) signaling networks and Fe micronutrient homeostasis. Articulating the principles that coordinate S with P, N, and Fe, from membrane remodeling and amino-acid synthesis to Fe-S assembly and whole-plant allocation, may give deeper mechanistic insights and practical leverage points. Embedding these principles and models into breeding targets, nutrient input strategies, and decision tools offers a path to crops that use nutrients more efficiently, sustain yield and quality, and remain resilient as climates and soils change.