Sulfur metabolism-dependent retrograde signaling for oxidative stress acclimation Authors: Riley Furbank, Zuzana Plskova, Barry Pogson, Kai Xun Chan, Hideki Takahashi Published: 2026 ABSTRACT AbstractPlant sulfur metabolism is crucial to the plant acclimation response to abiotic stresses, providing the redox-active compounds cysteine and glutathione for redox buffering, as well as the chloroplast-to-nucleus retrograde signal 3′-phosphoadenosine 5′-phosphate (PAP) for activation of gene expression changes. Whilst these processes have been conventionally considered separately, here we review chloroplast retrograde signaling in the context of plant sulfur metabolism, with a focus on the biosynthesis and degradation of PAP in secondary sulfur metabolism. We outline mechanisms by which primary sulfur metabolism via cysteine and glutathione contribute to the modulation of chloroplast retrograde signaling. We examine emerging questions in how plant sulfur metabolism is coordinated in different cell types of a plant leaf for synthesis and accumulation of PAP. Finally, while the majority of chloroplast retrograde signaling research has focused on the model Brassicaceae plant species Arabidopsis thaliana, here we outline the opportunities for novel insights from non-Brassicaceae plants to enable an integrated understanding of the intersection of sulfur metabolism and retrograde signaling.Keywords: 3′-phosphoadenosine 5′-phosphate, cysteine, glutathione, oxidative stress, retrograde signaling, SAL1, sulfur Introduction Environmental extremes such as heat, drought, and excessive light stresses cause biochemical perturbations in plant chloroplasts by promoting the over-production of reactive oxygen species (ROS). ROS over-accumulation can result in lipid peroxidation, DNA damage, and protein modifications through cysteine and methionine oxidation (Zheng et al., 2025), and ultimately cell death, if not managed appropriately. Consequently, plant cells have evolved a suite of cellular and biochemical mechanisms to sense and respond to abiotic stresses by targeting ROS and redox state of chloroplasts. A crucial cellular mechanism for sensing chloroplastic ROS is chloroplast-to-nucleus signaling, also known as chloroplast retrograde signaling (Chan et al., 2016b), which activates cellular responses such as transcription and translation of proteins important for stress acclimation (Zandalinas et al., 2018; Molina-Moya et al., 2025). Concurrently, a key biochemical strategy is the management of levels and/or impacts of the ROS molecule with the longest half-life, which is hydrogen peroxide (H2O2). Various antioxidant and redox buffering compounds such as glutathione (GSH) act as substrates for (i) dedicated redox homeostasis pathways and enzymes that scavenge H2O2 such as the ascorbate–glutathione cycle and glutathione transferases (GSTs), respectively (Dixon et al., 2009; Mano et al., 2019); and (ii) cysteine-containing redox enzymes such as glutaredoxins to reverse H2O2-induced oxidation of cysteine residues on proteins (Wu et al., 2017). H2O2 can also be converted to water by glutathione peroxidases (GPXs), although unlike mammalian GPXs, the plant GPXs preferentially utilize thioredoxins rather than GSH as the reductant (Bela et al., 2022). Plant sulfur metabolism is tightly linked to plant abiotic stress responses Plant sulfur metabolism provides the critical link between the cellular signaling mechanisms of H2O2 sensing and the biochemical mechanisms of H2O2 scavenging and redox state maintenance. The primary branch of sulfur metabolism in plants provides the cell with the amino acids methionine and cysteine (Takahashi et al., 2011; Romero et al., 2014; Jez, 2019), the latter being crucial for H2O2-modulated post-translational regulation of chloroplast retrograde signaling proteins, as discussed later in this review. Cysteine is additionally a substrate for the biosynthesis of GSH for redox buffering and H2O2 scavenging via redox enzymes, as described above (reviewed in Hasanuzzaman et al., 2017; Jez, 2019; Noctor et al., 2024). The secondary branch of sulfur metabolism in plants enables the regulation of stress-responsive plant hormones and metabolites such as jasmonic acid, brassinosteroids, salicylic acid and glucosinolates via sulfation (Hirschmann et al., 2014). The by-product of these reactions is 3′-phosphoadenosine 5′-phosphate (PAP), which then functions as a key chloroplast-to-nucleus retrograde signal during drought, high light and heat stress (Estavillo et al., 2011; Fang et al., 2019). Plant sulfur assimilation begins with uptake of the sulfate (SO42−) or thiosulfate (S2O32−) anions via roots (Nakajima et al., 2019). The transport of sulfate in the roots is facilitated by sulfate transporters (SULTRs), which co-transport sulfate and H+ ions (for review, see Takahashi, 2019). The Arabidopsis SULTRs are distributed across different tissues for consistent uptake of sulfate into the plant, and for efficient source-to-sink transport. Different SULTR groups have different affinities to sulfate corresponding to their function and localization within the plant. For example, Group 2 SULTRs (SULTR2;1, SULTR2;2) are low-affinity transporters which are sufficient to transfer the highly abundant sulfate in xylem sap between roots and shoots under sulfur-replete conditions, while the high affinity for sulfate in group 1 SULTRs (SULTR1;1, SULTR1;2 ,SULTR1;3) is consistent with their role in loading sulfate from soil into root cells (Takahashi, 2019). SULTR1;1 and 1;2, are localized in the roots (Takahashi et al., 2000; Yoshimoto et al., 2002), SULTR2;1 is localized in the root and leaf xylem and phloem (Takahashi et al., 2000), SULTR2;2 is localized in the root phloem and leaf bundle sheath (BS) cells (Takahashi et al., 2000), and SULTR1;3 is localized in the phloem companion cells (Yoshimoto et al., 2003). Within the cell, the Group 3 SULTRs, such as SULTR3;1, transport sulfate into the chloroplast, providing material for cysteine biosynthesis (Cao et al., 2013). The Group 1 SULTRs may also perform the role of low-affinity SULTRs during sulfur deficiency. For example, SULTR1;3 involved in loading sulfate to the phloem is transcriptionally up-regulated under low sulfur conditions (Yoshimoto et al., 2003), which may have an opposing effect against the miR395-mediated suppression of SULTR2;1 in the phloem companion cells under sulfur limitation (Kawashima et al., 2009). SULTR expression is responsive to abiotic stress to enable stress acclimation. For instance, expression of both Arabidopsis and Medicago SULTR2;1 is down-regulated in shoots under drought stress, whereas SULTR3;1 is up-regulated (Gallardo et al., 2014). This can potentially enable a high root-to-shoot sulfate gradient, thereby facilitating efficient transport of sulfate into guard cells, where sulfate can contribute to stomatal closure via abscisic acid (ABA) signaling (Batool et al., 2018; Sun et al., 2025). In this context, bioinformatic analysis has identified cis-regulatory elements in SULTR promoters that may be responsive to transcription factors regulated by ABA and methyl jasmonate, providing a hypothetical mechanism for these stress-responsive hormones to regulate the expression of SULTRs (Puresmaeli et al., 2023). After assimilation, sulfate is activated to adenosine 5ʹ-phosphosulfate (APS) by ATP-sulfurylase. Most of the APS in plant cells enters the reductive primary sulfur assimilation pathway where it is reduced to sulfite by APS reductase (APR) (Fig. 1). The sulfite is reduced by sulfite reductase to form sulfide, which is used to form cysteine by cysteine synthase complex (CSC). The complex is composed of serine acetyltransferase (SAT) hexamers and O-acetylserine(thiol)lyase (OAS-TL) dimers. Under sulfur-sufficient conditions, the complex remains intact, and OAS is synthesized by SAT. Cysteine is synthesized by free OAS-TL dimers, while the OAS-TLs in the complex are inactive. As the sulfur pool reduces, OAS accumulates and promotes dissociation of the complex. SAT freed from the complex is inactive due to feedback inhibition by cysteine (Wirtz et al., 2010; Hell and Wirtz, 2011; Takahashi et al., 2011; Wirtz et al., 2023). OAS can then act as a signal on its own or spontaneously convert to N-acetylserine at neutral pH (Flavin and Slaughter, 1965; Hopkins et al., 2005). Cysteine biosynthesis in chloroplasts is required for short-term response to high light stress, but is dispensable in long-term growth under elevated light, presumably due to compensation by other CSC isoforms such as those in the mitochondria (Speiser et al., 2015). Retrograde signaling pathways in context of sulfur metabolism. Assimilated sulfate is activated by ATPS to APS, most of which enters the reductive pathway and gets reduced to sulfite. From the cytosol, sulfite can enter the peroxisome and become oxidized by SOX, forming H2O2 as a by-product. In the chloroplast, sulfite is reduced by SIR to sulfide, which is then used by CSC to form cysteine, serving as a precursor for Met, GSH, and other S-containing compounds. Non-reduced APS can enter the oxidative pathway and provide sulfur to form PAPS, catalyzed by APK. PAPS can relocate between the cytosol and the chloroplast through PAPST, providing a substrate for SOT to form PAP, which serves as a substrate for SAL1 in the chloroplast and the mitochondria, and can also be transported by PAPST. Under stress conditions, ETC-produced ROS can inhibit SAL1, leading to the accumulation of PAP and its migration to the nucleus. PAP then inhibits XRNs, causing accumulation of transcripts involved in stress responses. Chloroplastic ROS can also damage the 4Fe-4S cofactor of HDS, leading to accumulation of another retrograde signal, MEcPP. Similar to PAP, MEcPP migrates to the nucleus, inducing the expression of stress-related genes. While some chloroplast-produced ROS may be quenched by antioxidants such as GSH, others can migrate to the nucleus through stromules. In the nucleus, ROS can oxidize cysteines of transcription factors leading to disulfide bond formation, which can be reversed by nuclear TRXs/GRXs. Legend: Light blue boxes indicate enzymes, dark blue boxes indicate transporters, yellow ovals indicate metabolites. Black arrows indicate enzymatic reactions, blue arrows indicate movement, red inhibitory lines indicate enzymatic inhibition. Dashed blue lines indicate movement of ROS and metabolites yet to be proven, dashed red line suggests ROS implications of the mitochondrial SAL1–PAP reaction. Red lightning symbol indicates environmental stress. Abbreviations: 4Fe-4S, ferredoxin-type Fe-S cluster; AMP, adenosine monophosphate; APK, APS kinase; APR, APS reductase; APS, adenosine 5′-phosphosulfate; ATPS, ATP-sulfurylase; CPKs, calcium-dependent protein kinases; CSC, cysteine synthase complex; Cys, cysteine; ETC, electron transport chain; GRXs, glutaredoxins; GSH, glutathione; H2O2, hydrogen peroxide; HDS, hydroxymethylbutenyl diphosphate synthase; HMBPP, 4-hydroxy-3-methylbut-2-enyl diphosphate; HPL, hydroperoxide lyase; MEcPP, 2 -C-methyl-derythritol2,4-cyclodiphosphate; Met, methionine; PAP, 3′-phosphoadenosine 5′-phosphate; PAPS, 3′-phosphoadenosine 5′-phosphosulfate; PAPST, 3′-phosphoadenosine 5′-phosphosulfate transporter; PRANGs, plastid redox-associated nuclear genes; SAL1, SAL1 phosphatase; SIR, sulfite reductase; SOT, sulfotransferase; SOX, sulfite oxidase; SULTR, sulfate transporter; TFs, transcription factors; TRXs, thioredoxins; UPR, unfolded protein response; XRNs, 5′-3′ exoribonucleases. Alternatively, APS can be utilized in the secondary, oxidative sulfur assimilation pathway (Fig. 1). Here, APS functions as a sulfur donor for the formation of 3′-phosphoadenosine 5′-phosphosulfate (PAPS) catalyzed by APS kinase (APK). PAPS can be transported between its sites of synthesis in the cytosol and chloroplasts to mitochondria—and presumably the Golgi—via PAPS/PAP transporters (PAPSTs) (Gigolashvili et al., 2012). PAPS is an active form of APS that supplies sulfur to recipient substrates in a reaction catalyzed by sulfotransferases (SOTs), such as in the biosynthesis of glucosinolates and sulfated signaling peptides (Kaufmann and Sauter, 2019). SOT activity is crucial for abiotic stress responses; for example, the sulfotransferase ST2A/SOT15 is required for regulation of jasmonic acid activity for heat acclimation (Zhu et al., 2021). The SOT-catalyzed reactions produce a by-product, PAP (Hirschmann et al., 2017), which is then recycled by the SAL1 phosphatase to adenosine monophosphate (AMP) (Quintero et al., 1996). Sulfur metabolism as a source and regulator of retrograde signals The SAL–PAP chloroplast retrograde signaling pathway To date, one of the best characterized chloroplast-to-nucleus signaling pathways is the aforementioned SAL1 phosphatase and its catabolic substrate PAP from secondary sulfur metabolism (Fig. 1). As PAP is inhibitory to SOT activity (Hirschmann et al., 2017), PAP levels are tightly regulated in the cell primarily via the catabolic activity of SAL1 (Ashykhmina et al., 2021). SAL1 is a bifunctional enzyme that primarily dephosphorylates PAP to AMP, and PAPS to APS (Quintero et al., 1996; Zhao et al., 2019). Inactivation of SAL1 due to cysteine oxidation by H2O2 during stress (Chan et al., 2016a) enables PAP accumulation in wild type plants under various abiotic stresses, including drought, high light, and heat (Estavillo et al., 2011; Fang et al., 2019). The accumulated PAP translocates to the nucleus, where it binds and inhibits 5′-3′ exoribonucleases (XRNs) (Gy et al., 2007; Rodriguez et al., 2010). This leads to increased polymerase II read-through from highly expressed genes to downstream loci, which are enriched in transcripts promoting stress tolerance (Crisp et al., 2018). These include plastid redox-associated nuclear genes (PRANGs), such as Ascorbate Peroxidase 2 (APX2), involved in detoxification of H2O2 (Rossel et al., 2006; Estavillo et al., 2011). The role of PAP as a stress-responsive retrograde signal mediating plant environmental acclimation is demonstrated by Arabidopsis sal1 mutants with constitutively accumulated PAP, which show decreased cellular damage and enhanced survivorship under drought and excess light (Crisp et al., 2018) and cadmium toxicity (Xi et al., 2016). The role of PAP extends beyond H2O2 detoxification, as sal1 mutants also have enhanced stomatal closure, which is recapitulated in wild type guard cells treated with exogenous PAP (Pornsiriwong et al., 2017; Tee et al., 2025). Interestingly, PAP induces stomatal closure by up-regulating calcium-dependent protein kinases (CPKs), which activate a ROS burst, most likely both H2O2 and superoxide, in both chloroplasts and the apoplast (Tee et al., 2025). Whilst the SAL–PAP pathway has been primarily studied in the context of chloroplast signaling, SAL1 is localized to both chloroplasts and mitochondria in Arabidopsis (Estavillo et al., 2011). This localization is consistent with the presence of the PAPS/PAP transporters, PAPST1 and PAPST2, on the chloroplast and mitochondrial membranes (Gigolashvili et al., 2012; Ashykhmina et al., 2019). SAL1 is functional in mitochondria as the mutant phenotypes of sal1 in growth, morphology, and sulfur metabolism can be complemented with mitochondrial-targeted SAL1 (Ashykhmina et al., 2019; Ashykhmina et al., 2021). Significant overlaps in the transcriptome and stress phenotypes of sal1 and the radical cell death induced 1 (rcd1) mutant further suggest an interaction between SAL1–PAP signaling and RCD1-regulated mitochondrial retrograde signaling (Van Aken and Pogson, 2017; Shapiguzov et al., 2019). Given that mitochondria are also oxidized under conditions where PAP signaling is active, such as chloroplast photooxidative stress (Ugalde et al., 2021) and stomatal closure (Postiglione and Muday, 2023), it is possible that SAL1 is also regulated by the mitochondrial redox state for PAP accumulation. SAL1 is present in sufficiently high abundance in both purified chloroplast and mitochondrial fractions in wild type leaves for detection by western blot (Estavillo et al., 2011). In contrast, genetic complementation of sal1 with SAL1 targeted exclusively to either chloroplasts or mitochondria resulted in transgenic lines with unequal abundance of SAL1 (Ashykhmina et al., 2021). Consequently, delineating the role(s) of SAL1 in chloroplasts versus mitochondria will require careful analysis. Regardless of the subcellular localization of SAL1, PAP levels in the cell need to be tightly regulated. Whilst small increases in cytosolic PAP have been linked to growth promotion (Ashykhmina et al., 2019), plants that constitutively accumulate 5- to 10-fold higher PAP than wild type show a range of developmental defects, including hindered growth, altered leaf morphology and delayed flowering. These phenotypes are linked to perturbations in hormonal homeostasis (Rodriguez et al., 2010; Phua et al., 2018) and light signaling (Kim and von Arnim, 2009). Constitutive PAP accumulation in sal1 mutants also causes extensive disruption to sulfur metabolism, including SOT inhibition leading to decreased levels of key sulfated metabolites such as glucosinolates and decreased glutathione (Lee et al., 2012). These impacts are consistent with the known PAP inhibition of SOTs (Hirschmann et al., 2017) and the extensive crosstalk between primary and secondary sulfur metabolism (Li et al., 2020; Kopriva et al., 2024). However, the effects of PAP on sulfur metabolism have been observed from genetic mutants accumulating very high PAP levels throughout their entire life cycle. Furthermore, recombinant SAL1 shows enzymatic activity against additional substrates including PAPS and inositol phosphates (Quintero et al., 1996). The alx8 mutant allele of SAL1 accumulates 20-fold higher PAP and a relatively small, but significant, 2-fold increase in PAPS (Estavillo et al., 2011). The latter is likely a compensatory feedback response to inhibition of SOT activity by PAP as discussed above, although PAPS accumulation directly due to loss of SAL1 catabolic activity against PAPS cannot be ruled out. The alx8, fry1-6 and fry1-1 alleles of SAL1 did not accumulate inositol phosphates (Estavillo et al., 2011), although fry1-1 did show a 1.5- to 2-fold increase in inositol 1, 4, 5-trisphosphate (IP3) in a separate study (Zhang et al., 2011) and another mutant allele of SAL1 with altered activity against IP3 has been described (Shirzadian-Khorramabad et al., 2022). Consequently, pleiotropic effects from either PAP, PAPS and/or IP3 over-accumulation on sulfur metabolism may need to be considered. How these metabolites intersect with sulfur metabolism in wild type plants during short-term exposure to oxidative stress also remains to be elucidated. Redox biochemistry of cysteine is critical to chloroplast signaling Disulfide bond formation in SAL1 The presence of hydrogen-bond donors and an electropositive micro-environment surrounding a cysteine residue on a protein can decrease its pKa from the typical value of ∼8.5. This increases the propensity of the thiol group in cysteine to deprotonation, producing a negatively charged thiolate anion which can react directly with H2O2 spontaneously or through various peroxidases (Poole, 2015). These cysteine oxidation events are essential to sensing changes in the redox environment of the cell (Mittler and Jones, 2024). Known protein targets of cysteine oxidation include transcription factors, peroxidases, and metabolic enzymes, with upwards of 200 000 cysteines estimated to be implicated in redox networks (Jones and Sies, 2015; Foyer et al., 2018; Zhou et al., 2023). Cysteine oxidation by H2O2 leading to disulfide bond formation can dramatically impact protein structure and function, often playing key regulatory and signaling roles, including in retrograde signaling (Chan et al., 2016a; Young et al., 2019). This is exemplified by the redox-sensitive cysteines in SAL1, where H2O2-responsive disulfide bonding significantly decreases SAL1 phosphatase activity (Chan et al., 2016a) (Fig. 1). Under oxidizing conditions, the Cys119 residue at the dimerization interface of the Arabidopsis SAL1 protein can facilitate an inter-molecular disulfide bond between two SAL1 monomers. Furthermore, the oxidized SAL1 protein also contains a Cys167–Cys190 intramolecular disulfide pair, which appears to be highly conserved across the plant kingdom (Chan et al., 2016a). Molecular dynamics simulations of oxidized SAL1 possessing both the inter- and intra-molecular disulfide bonds indicate significantly decreased mobility of a flexible loop mediating substrate entry and/or product release at the active site, providing a mechanistic explanation for its decreased enzymatic activity when oxidized (Chan et al., 2016a). The redox-sensitive cysteines in SAL1 can also be regulated via glutathionylation (Chan et al., 2016a) or S-nitrosylation (Hu et al., 2015). Glutathionylation of cysteine residues can simultaneously regulate protein activity and protect the oxidized sulfenic acid of cysteine residues from irreversible oxidation to sulfonic acid, which can render proteins non-functional (Huang et al., 2018). Thus, it is possible that reversible S-glutathionylation of cysteine residues on SAL1 may enable more rapid regulation of PAP levels by SAL1 upon recovery from stress conditions. Cysteine oxidation and reduction There is also increasing attention on H2O2 itself acting as a retrograde signal. Direct H2O2 transfer from chloroplast to the nucleus can occur via chloroplast projections termed stromules during viral infection (Caplan et al., 2015), or via direct chloroplast–nuclear associations during high light stress (Exposito-Rodriguez et al., 2017) and stomatal closure (Tee et al., 2025). Consistent with these observations, several nuclear proteins with redox-sensitive cysteines have been identified by proteomics (Dard et al., 2024), although it is important to note that not all of these proteins have been experimentally validated to have redox-sensitive activity. Nevertheless, various components of the nuclear gene expression machinery can be regulated via cysteine oxidation, including proteins involved in the mediator complex and in transcriptional activation (Shaikhali et al., 2008, 2015; De Smet et al., 2025; Huang et al., 2025). For example, GENERAL CONTROL NON-DEPRESSIBLE 5 (GCN5), which is part of the Spt–Ada–Gcn5 acetyltransferase (SAGA) complex (Vlachonasios et al., 2021) that serves as a transcriptional coactivator, contains several cysteines which can be S-sulfenylated, thereby negatively impacting its activity in vitro. Furthermore, plants with Cys293-Ser mutation displayed phenotypes similar to gcn5 mutant plants under oxidative stress induced by paraquat, suggesting the importance of Cys293 in redox responses of GCN5. The aforementioned RCD1 protein also contains seven interdomain cysteine residues, which can affect the oligomerization state and activity of RCD1 in response to chloroplastic ROS, thereby influencing expression of genes responsive to mitochondrial dysfunction (Shapiguzov et al., 2019). Overall, these results support a link between redox signaling via sulfenylation and the regulation of gene expression. Another role of cysteine residues is in the coordination of essential enzymatic cofactors such as 2Fe–2S and 4Fe–4S ferredoxin-type iron-sulfur clusters, both of which are acutely sensitive to oxidation by ROS such as superoxide and H2O2 (Balk and Pilon, 2011; Couturier et al., 2013). In plants, the hydroxymethylbutenyl diphosphate synthase (HDS) enzyme contains a ferredoxin-type cluster, and is responsible for converting 2-C-methyl-derythritol 2,4-cyclodiphosphate (MEcPP) into (E)-4-hydroxy-3-methylbut-2-en-1-yl diphosphate (HMBPP) during isoprenoid biosynthesis in chloroplasts (Fig. 1) (Seemann et al., 2005). Under oxidative stress conditions such as high light, ROS produced at photosystem I damage the HDS 4Fe-4S cofactor, inhibiting its activity and leading to the buildup of its native substrate MEcPP, which functions as a chloroplast-to-nucleus retrograde signal (Xiao et al., 2012; Benn et al., 2016) (Fig. 1). MEcPP induces stress-responsive genes such as hydroperoxide lyase (HPL) and many genes associated with the unfolded protein response in the endoplasmic reticulum (Walley et al., 2015; Bjornson et al., 2017). Similarly to SAL1–PAP, the MEcPP retrograde signaling pathway arose early in plant evolution, and interacts with hormonal signaling (Jiang et al., 2018). Whilst cysteine oxidation enables activation of chloroplast retrograde signaling, how the reverse process impacts chloroplast communication is an equally important question. Thioredoxins (TRXs) and glutaredoxins (GRXs) are small proteins that catalyze the reversible oxidation-reduction of cysteine residues in specific substrate proteins to regulate their activity. As the roles of TRXs and GRXs in retrograde signaling have been recently reviewed in detail (Sevilla et al., 2023), we will only highlight selected points of interest. Several chloroplast TRXs including Trx f, m, z and NADPH-dependent thioredoxin reductase C (NTRC) have been implicated in the redox regulation of chloroplast gene expression proteins and signaling pathways in response to abiotic stresses, as well as during chloroplast biogenesis (Chan et al., 2016a; Yokochi et al., 2019). In the nucleus, several TRXs and GRXs in different plant species have been identified to interact with and/or revert the oxidation of transcription factors (reviewed in Sevilla et al., 2023), presenting a potential mechanism for rapid and reversible modulation of protein activity in the nucleus in response to H2O2 from chloroplasts. Thus, collectively, the redox biochemistry of cysteine provides several mechanisms for the modulation of organelle retrograde signaling, including at the organellar sites of H2O2 generation and in the nucleus (Fig. 1). Does cysteine biosynthesis regulate chloroplast signaling? The chloroplast redox state is influenced by several processes, including the Mehler reaction, which reduces O2 at photosystem I to form superoxide and ultimately H2O2, and the removal of H2O2 via the ascorbate-glutathione cycle, both of which have been comprehensively studied (Foyer and Noctor, 2011, 2020). How these well-established redox homeostasis pathways can be supplemented by cysteine biosynthesis for maintenance of chloroplast redox balance, and if cysteine biosynthesis itself is also an origin of retrograde signals, have long been topics of interest (López-Martín et al., 2008; Speiser et al., 2015). Mutation of the cytosolic O-acetylserine(thiol)lyase OAS-A1 and decreased cysteine content compromises the cellular redox balance, leading to overaccumulation of H2O2 even under non-stress conditions (López-Martín et al., 2008; Romero et al., 2014). Inversely, the des1 mutant which lacks cysteine desulfuration activity has increased foliar cysteine content, promoting oxidative stress tolerance by limiting H2O2 accumulation (Heeg et al., 2008; Alvarez et al., 2010; Zhou et al., 2021). Interestingly, des1 plants also exhibited an up-regulation in APX activity, potentially suggesting a connection between cysteine biosynthesis and expression of H2O2 homeostasis enzymes, in addition to the direct role of cysteine as a substrate for GSH synthesis and redox buffering. Indeed, overexpression of the plastid-localized CSC composed of OAS-B and serine acetyltransferase in tobacco drives cysteine accumulation in chloroplasts, but does so regardless of serine acetyltransferase activity (Wirtz et al., 2023). Plants overexpressing either the enzymatically active or inactive plastid CSC both exhibit similar chlorotic phenotypes, and have impaired grana formation, but show little change in glutathione redox homeostasis. This suggests that CSC formation, rather than activity, is responsible for these phenotypes, hinting towards cellular sensing and signaling mechanisms that connect cysteine biosynthesis to nuclear gene expression (Fig. 2). Putative sensing and signaling pathways linked to the plastid CSC in guard cells and foliar tissue. In guard cells, the CSC is a key regulator that integrates signals from sulfur metabolism, 12-oxo-phytodienoic acid (OPDA), and hormones to control stomatal dynamics. OPDA (derived from alpha-linoleic acid metabolism) accumulates in response to oxidative stress and drought, and via cyclophilin 20-3 (CYP20-3) stabilizes the CSC, leading to increased production of cysteine, glutathione, and ABA for stomatal closure. In contrast, in foliar tissue, association of the CSC, independent of its activity, functions to regulate the expression of genes implicated in sulfur metabolism, and regulates photosynthesis and chloroplast structural dynamics. Note that this has only been demonstrated in tobacco leaves; the downstream components in this signaling cascade and the involvement of other interactors such as CYP20-3 are currently unknown. In Arabidopsis, SCS in the thylakoid lumen catalyzes the formation of S-sulfocysteine from thiosulfate, where mutation of SCS leads to sensitivity to oxidative stress. How the SCS enzyme, S-sulfocysteine itself, or other interacting partners help modulate stress-responsive gene expression and protect the chloroplast redox state alongside the CSC is unclear. Abbreviations: ABA, abscisic acid; CSC, cysteine synthase complex; CYP20-3, cyclophilin 20-3; OAS-B, O-acetylserine(thiol)lyase B; OPDA, 12-oxo-phytodienoic acid; SAT, serine acetyltransferase; SCS, S-sulfocysteine synthase; SULTR, sulfate transporter. Analysis of serine acetyltransferase single mutants revealed that the mitochondrial isoform is mainly responsible for OAS formation and downstream cysteine biosynthesis in leaves, while the plastidial isoform alone is insufficient to support wild type levels of cysteine biosynthesis and growth (Watanabe et al., 2008; Wirtz et al., 2012). In contrast, in guard cells specifically, the chloroplast localized CSC is essential in driving ABA biosynthesis and stomatal closure (Bauer et al., 2013; Sun et al., 2025). Plants lacking only the chloroplast-localized CSC still close stomata in response to ABA, but are insensitive to exogenous sulfate treatment, in contrast to the wild type, cytosolic, and mitochondrial csc mutants (Sun et al., 2025). For the plastid csc mutant, this was associated with an impaired induction of ABA-responsive gene expression only in guard cells of epidermal peels. Interestingly, 12-oxo-phytodienoic acid (OPDA), a precursor in jasmonic acid biosynthesis, promotes CSC formation in plastids by interacting with the jasmonate family binding protein cyclophilin 20-3 (CYP20-3), stabilizing the OASTL-B-serine acetyltransferase interaction (Park et al., 2013; Müller et al., 2017). The plastid CSC in guard cells thus acts as a hub that integrates signals such as sulfate, OPDA, and ABA to drive light-dependent stomatal closure (Fig. 2). Notably, the plastid CSC seems especially important in supplying cysteine for autonomous ABA biosynthesis in guard cells (Sun et al., 2025). Thus, while potential retrograde signaling roles of the CSC can be decoupled from its enzymatic activity in some cases, its role in driving OPDA-dependent stomatal closure in guard cells seems activity-dependent, highlighting potential cell-specific functions of CSC signaling. It is noteworthy that OPDA itself has been proposed to function as a signal from chloroplasts as it accumulates during high light stress (Kılıç et al., 2025) and can alter nuclear gene expression independently of JA (Taki et al., 2005), most likely after conversion to downstream metabolites in the peroxisomes (Ueda et al., 2025). While CSC activity accounts for most of the OAS production in plastids and is sufficient for plant survival (Watanabe et al., 2008), and seems especially important in regulating guard cell dynamics (Sun et al., 2025), an intriguing question is whether secondary cysteine biosynthetic pathways also function for maintenance and growth, or instead may contribute to the regulation of chloroplast signals and stress acclimation. In Arabidopsis, chloroplast cysteine synthesis is catalyzed by both the CSC and S-sulfocysteine synthase (SCS), the latter incorporating thiosulfate into OAS to form S-sulfocysteine within the thylakoid lumen (Bermúdez et al., 2010). This contrasts with the function of OAS-B in primary sulfur assimilation and the incorporation of sulfide to form cysteine directly in the chloroplast stroma. Intriguingly, OAS-B does not appear to be the main contributor to total OASTL activity and thus cysteine synthesis in foliar tissue, as oas-b null mutants exhibit no severe phenotypes (Heeg et al., 2008; Watanabe et al., 2008). However, OAS-B and the plastid CSC alone are sufficient to support plant growth, suggestive of plasticity in cysteine synthase activity and exchange between different sub-cellular cysteine pools (Watanabe et al., 2008). SCS, on the other hand, seems to have a unique function, as loss of this enzyme results in unique light regime-dependent phenotypic differences (Bermúdez et al., 2010; Bermudez et al., 2012). Under long days scs plants are dwarfed, have decreased chlorophyll content and photosynthetic activity, and over accumulate both H2O2 and superoxide despite accumulating significantly more glutathione. These phenotypes were specific to long day conditions and attributed to damage of the photosynthetic machinery arising from perturbations in chloroplast redox balance during periods of prolonged light exposure, conditions mimicking high light stress in a wild type plant. Together with the unique localization of SCS to the thylakoid lumen, these results could suggest a distinct, specialized role of S-sulfocysteine synthesis in regulating high light-dependent chloroplast signals (Fig. 2). However, whether the scs phenotypes are driven by a direct change in the chloroplast cysteine pool or by a unique signaling function of SCS remains unclear. While the scs mutant does not show decreased cysteine content at the whole-leaf level, local changes in the chloroplast cysteine pool may be masked when cysteine content is measured at the whole cell level. Despite their functional differences, it is important to note that SCS and the CSC both depend on OAS availability for their activity. OAS has a dual function as substrate for the CSC, but also promotes its dissociation, effectively acting as a product inhibitor of CSC signaling (Droux et al., 1998; Wirtz et al., 2001; Sun et al., 2021). Considering that the OAS-B of the CSC is significantly more abundant than SCS and accounts for the majority of plastid OAS production, CSC activity in the stroma likely limits supply of OAS to SCS in the lumen. Whether these different plastid cysteine synthesis signaling pathways are functionally distinct, interact, or are co-regulated to maintain the chloroplast redox balance remains unexplored. Cellular compartmentalization of sulfur metabolism and retrograde signaling Intriguingly, there is an emerging hypothesis that chloroplast retrograde signals may function in a cell-specific manner (Mackenzie and Mullineaux, 2022), as the chloroplast proteomes of mesophyll cells vary significantly from the plastid proteomes of epidermal and vascular parenchyma cells (Beltrán et al., 2018). Indeed, PAP function is not uniform across all cell types (Fig. 3). In foliar tissue of the Arabidopsis sal1 mutant, PAP-induced up-regulation of APX2 leads to decreased ROS accumulation in vascular bundles (Estavillo et al., 2011). Whilst PAP has only been measured at the whole-leaf level, the APX2 and ROS phenotypes in sal1 are consistent with high SAL1 promoter-driven GFP abundance in vascular bundles (Estavillo et al., 2011) and enriched SAL1 mRNA on ribosomes of leaf BS cells (Aubry et al., 2014). In contrast, exogenous PAP treatment on wild-type leaves, and constitutive PAP accumulation in the sal1 mutant induce ROS production in both the chloroplasts and nuclei of guard cells (Pornsiriwong et al., 2017; Tee et al., 2025). In this case, PAP induces chloroplast ROS accumulation via photosynthetic electron transport. PAP also directly induces expression of calcium-dependent protein kinases that activate Respiratory Burst Oxidase Homolog D (RBOHD) at the plasma membrane, resulting in an apoplastic ROS burst (Tee et al., 2025). Could the differential impacts of PAP in BS and guard cells be explained, at least partially, by cell-specific differences in sulfur metabolism? Cellular compartmentation of SAL–PAP signaling in Arabidopsis. (A) Summary of major cell types and their positions within Arabidopsis leaves. (B) Specialization of PAP-dependent chloroplast signaling in vascular bundles (vasculature + bundle sheath cells) and guard cells. While both cell types may have the capacity for autonomous sulfur assimilation and PAP biosynthesis via SOTs, the SAL–PAP pathway has unique outcomes depending on the cell type. PAP-induced gene expression in vascular bundles suppresses ROS and remodels photosynthesis, while in guard cells it drives ROS accumulation in both chloroplasts and nuclei, interacting with ABA to induce stomatal closure. In (A), cells with high relatively high photosynthetic activity and/or number of chloroplasts are shown in green. In (B), chloroplasts are indicated by green rectangles. Abbreviations: ABA, abscisic acid; AMP, adenosine monophosphate; APXs, Ascorbate Peroxidases; GSHS, Glutathione Synthase; PAP, 3’-phosphoadenosine 5’-phosphate; PAPS, 3’-phosphoadenosine 5’-phosphosulfate; PAPST; PAPS/PAP Transporter; RBOHD, Respiratory Burst Oxidase Homolog D; ROS, reactive oxygen species; SOT, sulfotransferase; TGG1, Thioglucoside Glucohydrolase 1. A growing body of work utilizing single-cell RNA sequencing (scRNA–seq), spatial metabolomics, and multi-omics has enabled quantification of gene expression and metabolites with unparalleled cellular resolution (Angelovici and Kliebenstein, 2022; Delannoy et al., 2023; Ntelkis et al., 2025). Cell-specific profiling of the Arabidopsis translatome highlighted dramatic enrichment of sulfur assimilatory and metabolic pathways (Aubry et al., 2014). This included the BS-preferential ribosomal association of 11 transcripts encoding sulfate transporters, two isoforms of ATP sulfurylase, three APS reductases, and four serine O-acetlytransferases. This enrichment also extended to the secondary oxidative pathway, including APS kinase and several sulfotransferases, along with transcripts implicated in both aliphatic and indolic glucosinolate biosynthesis. Interestingly, the BS of rice is similarly enriched for these transcripts, suggesting that the complete sulfur assimilation pathway is preferentially abundant in the BS of C3 plants, possibly conditioning this cell type for both glutathione and secondary product biosynthesis (Hua et al., 2021). Similarly, transcripts for APR2 and 4 of the five SULTR genes are highly abundant in guard cells (Bauer et al., 2013; Ren et al., 2022). Together with the fact that mature guard cells are symplastically isolated from the mesophyll and epidermal pavement cells, collectively these observations—albeit at the transcript level—suggest a possibility for autonomous sulfur metabolism (Bauer et al., 2013; Wilson et al., 2025) in driving cell-specific PAP signaling in Arabidopsis. In the following sections we will examine correlations between PAP accumulation and aspects of sulfur metabolism which may be differentially regulated in specific cell types. Glutathione and the chloroplast redox balance The redox state of the cell is ultimately determined by its metabolic activity, where the balance of antioxidants, reducing equivalents, and energy production are closely tied to levels of ROS such as H2O2 (Fukushima et al., 2017; Mittler and Jones, 2024). While the direct reactivity of glutathione with H2O2 is likely too low to act as an efficient scavenger under most cellular conditions, it fulfills a plethora of roles critical in buffering the chloroplast redox state against environmental changes (Noctor et al., 2018). Glutathione is essential for regenerating ascorbate from its oxidized form (dehydroascorbate) after reaction with H2O2 in APXs, directly linking H2O2 scavenging to glutathione (Schnaubelt et al., 2015; Noctor, 2025). Glutathione is also linked to H2O2 reduction by ascorbate-independent routes via the action of GSTs. Several plant GSTs exhibit direct glutathione-dependent peroxidase activity against either H2O2 or other organic peroxides such as lipid peroxides or reactive carbonyl species, further helping to buffer rapid changes in H2O2 and redox (Dixon et al., 2009; Mano et al., 2019). Reduced glutathione consumed via these reactions can then be replenished by glutathione reductases and NADPH. Alterations in the chloroplast redox state driven by glutathione can influence H2O2 homeostasis and H2O2-responsive gene expression. In wild-type Arabidopsis expressing the glutathione redox potential sensor roGFP2, a less negative glutathione redox potential in chloroplasts is correlated with nuclear gene expression changes (Bratt et al., 2016; Haber et al., 2021) (Fig. 2). This correlation was observed under various conditions, including specific high light stress or photosynthetic inhibitor treatments that alter the glutathione redox potential in chloroplasts but not in mitochondria, cytosol, and peroxisomes (Bratt et al., 2016). The Arabidopsis cad2 mutation in glutamate-cysteine ligase (GCL) was originally discovered due to its sensitivity to cadmium treatment, which disrupts the redox balance and induces H2O2 production (Howden and Cobbett, 1992). This mutation diminishes glutathione biosynthesis to ∼30% of that seen in wild type, resulting in sensitivity to oxidative stress resulting from salinity stress and over-accumulation of ROS (Cobbett et al., 1998; Xiang et al., 2001; Queval et al., 2007; Han et al., 2013). The pad2 mutant of GCL also has a less negative glutathione redox potential, resulting in greater up-regulation of stress-responsive markers such as Glutathione Reductase and RbohD in response to pathogen challenge (Dubreuil-Maurizi et al., 2011). Similarly, in other GCL mutants such as rax1, decreased foliar glutathione drives constitutive up-regulation of stress-responsive genes like APX2, despite no clear differences in H2O2 accumulation (Ball et al., 2004). Interestingly, live-cell imaging of chloroplast redox dynamics and H2O2 using the fluorescent reporter Hyper7 have demonstrated heterogeneity in H2O2 accumulation between cell types under both basal and high light conditions (Dopp et al., 2023). In these experiments, epidermal pavement plastids showed the greatest degree of H2O2 accumulation, while guard cell chloroplasts exhibited an attenuated induction of H2O2 production. This coincided with a more reduced glutathione pool in guard cells, suggesting that heterogeneity in glutathione redox buffering between cell types influences capacity for H2O2 accumulation. How the cell-specific differences in chloroplast glutathione redox state may go on to influence the dynamics of H2O2-dependent retrograde signaling through pathways such as SAL–PAP is unclear. A more reduced chloroplast glutathione pool and lesser H2O2 accumulation may present lower capacity for SAL1 inactivation via cysteine oxidation (Chan et al., 2016a), or conversely be the consequence of increased PAP-dependent retrograde signaling, which up-regulates photoprotective genes and genes encoding H2O2 scavengers such as APX2 (Estavillo et al., 2011; Pornsiriwong et al., 2017). Notably, SAL1 enzymatic activity against PAP is decreased not only in leaf extracts of rax1-1 that has lower levels of GSH, but also in mutants less able to maintain high NADPH:NADP+ ratio and in regenerating 2-cys peroxiredoxins that modulate disulfide-bonding in their target proteins (Chan et al., 2016a). Consequently, SAL1 activity in vivo is likely to be regulated by additional factors beyond GSH abundance. Nevertheless, whether differential capacity for GSH synthesis and/or GSH:GSSG redox state impacts SAL1 redox control in guard cells, and if this extends to cell types such as the BS, merits further investigation involving concurrent measurements of plastid H2O2 and glutathione redox dynamics across multiple cell types. A related consideration is that perturbations of plastidial redox state under abiotic stress also leads to subsequent changes in the redox state of other subcellular compartments such as the cytosol and mitochondria (Ugalde et al., 2021), raising questions as to how redox control of mitochondrial SAL and other PAP phosphatases such as cytosol-localized AHL may collectively regulate PAP-mediated retrograde signaling. A correlation between glucosinolate metabolism and PAP in Brassicaceae Glucosinolate metabolism is almost exclusive to Brassicaceae, which includes Arabidopsis where most of the research on SAL1–PAP retrograde signaling research has been performed. Several aspects of glucosinolate metabolism are linked or correlated to PAP. The most direct link is that the SOT16-18 enzymes catalyze the final step of glucosinolate biosynthesis by transferring the sulfate group from PAPS to desulfo-glucosinolates, thereby producing PAP as the by-product (Hirschmann et al., 2014; Iwamoto et al., 2023). Similar to PAP accumulation under high light, heat and drought stresses (Estavillo et al., 2011; Fang et al., 2019), foliar glucosinolate accumulation is stimulated by abiotic stress in multiple Brassica species (Xu et al., 2019; Huang et al., 2024). Indeed, glucosinolate biosynthesis genes are generally strongly induced by oxidative stress, and glucosinolate-deficient mutants show increased sensitivity to abiotic stress (Ludwig-Müller et al., 2000; Huang et al., 2019; Rao et al., 2021). These observations are consistent with the role of several glucosinolate breakdown products in activating oxidative stress responses (Xu et al., 2019; Huang et al., 2024). Glucosinolate metabolism and its downstream functions in abiotic stress responses also appear correlated to PAP in the vascular bundle and guard cells. Glucosinolate biosynthetic enzymes such as CYP83s and SOT17 are enriched in accessory cells such as phloem parenchyma in the vascular bundle, consistent with imaging mass spectroscopy showing glucosinolates being concentrated near the midvein (Shroff et al., 2008; Koroleva and Cramer, 2011; Sarsby et al., 2012; Nintemann et al., 2018; Maeda et al., 2023). The myrosinase Thioglucoside Glucohydrolase 1 (TGG1) responsible for degrading glucosinolates into isothiocyanates is highly abundant in both guard cells and the vasculature, becoming significantly vasculature-enriched at the transcript level under ultraviolet stress (Koroleva and Cramer, 2011; Maeda et al., 2023). Exogenous isothiocyanate application induces broad transcriptomic remodeling, down-regulating expression of SOTs and APK but up-regulating a significant number of transcripts implicated in singlet oxygen (1O2) and H2O2 responses (Khokon et al., 2011; Kissen et al., 2016; Huang et al., 2024). This includes RBOHC and RBOHD, whose expression is also enriched in vasculature cells to drive systemic ROS signaling. In guard cells, direct breakdown products of glucosinolates such as allyl isothiocyanate, 3-butenenitrile, and ethyl thiocyanate act as stress signals that induce stomatal closure (Zhao et al., 2008; Hossain et al., 2013), doing so through similar mechanisms to PAP such as activating ROS production and elevating cytosolic Ca2+ (Khokon et al., 2011). Together with the fact that glucosinolates represent a significant sink of sulfur in Arabidopsis leaves, often accumulating to total concentrations similar to or greater than free cysteine and glutathione (in the nanomole per mg of fresh weight range) (Brown et al., 2003; Zhang et al., 2020), it is tempting to speculate that glucosinolate biosynthesis is directly linked to cell-specific PAP accumulation and signaling during abiotic stresses. That is, BS and other cell types with high SOT16-18 activity for glucosinolate biosynthesis and consequently high rates of PAP biosynthesis could be primed for retrograde signaling via the SAL–PAP pathway. In turn, the enrichment of SAL transcripts on BS ribosomes and its localization to the Arabidopsis vasculature may offset this increased capacity for PAP production by helping to catabolically balance PAP levels under unstressed conditions (Estavillo et al., 2011; Aubry et al., 2014). Similarly, in guard cells, glucosinolates and PAP could both concurrently regulate stomatal closure. Several Arabidopsis mutants deficient in glucosinolate biosynthesis show impaired drought responses and defects in stomatal regulation, whereas overexpression of key glucosinolate biosynthesis regulators improves tolerance (Salehin et al., 2019; Abdelgawad et al., 2023; Kliebenstein, 2023). However, direct experimental evidence linking glucosinolate metabolism to PAP accumulation and signaling is still lacking; for example, it remains unclear if there is altered PAP accumulation and/or signaling in these glucosinolate biosynthesis mutants and overexpressors. A key argument against glucosinolate metabolism being the sole explanation for cell-specific PAP signaling is that PAP-induced stomatal closure is evolutionarily conserved in land plants beyond Brassicaceae, having been observed in multiple species including Arabidopsis, ferns, mosses, barley, and wheat (Zhao et al., 2019; Jiang et al., 2024; Bowerman et al., 2026). PAP accumulation has also been observed in drought-stressed wheat (Bowerman et al., 2026) and in dehydrated Medicago truncatula seeds (Pagano et al., 2022). Except for Arabidopsis, all the other plant species listed above lack capacity for glucosinolate biosynthesis. Consequently, it is possible that the relationship between PAP and sulfur assimilation could be family-specific, and that Brassicaceae may be the exception, rather than the rule, for how plant cells regulate PAP biosynthesis during abiotic stress. Extensive study of PAP metabolism in non-Brassicaceae, at whole-leaf and cell type-specific levels, will be required to understand how PAP biosynthesis for retrograde signaling may occur in non-Brassicaceae plants. Beyond Arabidopsis: sulfur metabolism and retrograde signaling in non-Brassicaceae Sulfur-dependent chloroplast retrograde signaling beyond Brassicaceae remains largely enigmatic. Cysteine feeding experiments under low sulfur conditions by Rahimzadeh Karvansara et al. (2023) revealed significant accumulation of sulfate in the shoots of both rice and Setaria, but not in Arabidopsis, hinting towards monocot-specific pathways for remobilization and utilization of sulfur compounds. However, even within monocots, C3 rice and C4 Setaria viridis differ drastically in their root-shoot thiol allocation and responses to sulfur deprivation (Rahimzadeh Karvansara et al., 2023; Zenzen et al., 2024). Whether and how family- and/or species- specific differences in sulfur metabolism may impact on PAP homeostasis remains unclear. However, recent advances have begun to reveal the function of SAL-PAP signaling in monocots. While Arabidopsis has only one copy of SAL1, several monocot species contain multiple copies of SAL1 in their genomes due to genome and/or gene duplication events. Wheat, one of the world’s most important crops, has between five and seven homeologues of the SAL1 gene depending on the cultivar (Manmathan et al., 2013; Mohr et al., 2022; Bowerman et al., 2026). Targeting all six SAL1 homeologues in wheat ‘Bobwhite’ using CRISPR-mediated knockout technology did not significantly increase yield rescue under drought, and the sal1 knockout lines had decreased growth (Mohr et al., 2022). The latter is consistent with Arabidopsis sal1 growth deficiency, which may be explained by the negative effect of constitutively accumulated PAP on growth hormone signaling, such as down-regulation of brassinosteroid target genes and increased levels of jasmonic acid (Phua et al., 2018). In contrast, decreasing wheat SAL transcript levels by RNAi-mediated knockdown partially improved survivorship under drought stress (Manmathan et al., 2013). This opens the question of whether titrating the amount of PAP by partially decreasing SAL function in crops with multiple SAL homeologues can increase drought tolerance without hindering growth. Indeed, recent analysis of wheat ‘Chara’ mutants lacking only one or two SAL isoforms showed that this is likely the case (Bowerman et al., 2026). Unlike the Arabidopsis sal1 mutant, wheat ‘Chara’ heavy ion bombardment (HIB) deletion lines deficient in either one or two SAL genes showed no constitutive PAP accumulation and no growth or morphological phenotypes. However, under glasshouse drought conditions, several of these lines exhibited lower leaf SAL activity and higher PAP accumulation, correlating with enhanced leaf water retention and photosynthesis (Bowerman et al., 2026). Therefore, in vivo fine-tuning, rather than complete inactivation, of SAL phosphatase activity against PAP may avoid the yield versus resilience trade-off often encountered in crop breeding (Dietz et al., 2021). Intriguingly, the wheat SAL deletion analysis also suggests the possibility that certain plant species have evolved specialization of specific SAL1 homeologues, such that they have different roles in metabolism and drought resilience (Bowerman et al., 2026). The genome of wheat ‘Chara’ contains six SAL1 homeologues located on chromosomes 4A, 5A, 5B, 5D, 7A, and 7D, and each of these has a different gene expression profile across developmental stages (Bowerman et al., 2026). Based on phylogenetic analysis, these SALs were assigned to two groups, named TaSAL1 and TaSAL2. Co-expression data connected TaSAL1 to aminoacyl-tRNA pathway, DNA repair, and sulfur metabolism, while TaSAL2 was connected to carbon metabolism and amino acid biosynthesis. When exposed to fluctuating growth conditions in field trials, a TaSAL1-5D null line exhibited higher chlorophyll and nitrogen content as well as greater drought resilience, but had lower yield than wild type. In contrast, the TaSAL2-4A null line exhibited up to 8% higher yield, increased water productivity, and yield stability (Bowerman et al., 2026). Whether this homeologue specialization is purely controlled by the timing of gene expression across plant development, or also involves differences in cell type expression, subcellular localization and/or biochemical activity, remains unclear. The interaction between PAP signaling and other components of sulfur metabolism in non-Brassica species such as wheat could be an interesting future research area. Wheat ‘Chara’ SAL deletion lines exhibited diurnally-regulated enhanced leaf temperatures in the field under water-limiting conditions. However, the leaf temperatures were the same as the parental lines under well-watered conditions and at dawn. This likely reflects more dynamically regulated stomata in the presence of PAP (Bowerman et al., 2026). Interestingly, the SAL mutants had decreased ABA levels under drought, suggesting down-regulation or by-pass of the ABA pathway due to enhanced PAP signaling (Pornsiriwong et al., 2017; Bowerman et al., 2026). Given that cysteine biosynthesis in guard cells enables ABA biosynthesis and stomatal closure (Sun et al., 2025), the cross-talk between PAP signaling and cysteine biosynthesis in guard cells of monocot species may reveal crucial insights into the coordination of metabolism and stress responses. Despite these insights into SAL1–PAP signaling from wheat, however, C3 monocots may not be ideal for exploring cell-specific sulfur metabolism and retrograde signaling. A key limitation to cell type-resolved insights into SAL1–PAP signaling is the requirement for cell-separation and analysis methods that are both high-resolution and of sufficient high-sensitivity to detect low-abundance secondary metabolites such as PAP. However, plants performing C4 photosynthesis represent a unique system in which the three major cell layers of the leaf [mesophyll (M), bundle sheath (BS)/vasculature, and epidermis] can be mechanically separated with relative ease and high purity, allowing for measurements of protein, metabolites, and transcripts in these cell type-enriched fractions (Döring et al., 2016; Rao et al., 2016; Denton et al., 2017). Interestingly, despite the general specializations of M and BS cells being more extensive in C4 plants, C4 monocots exhibit similar compartmentation of sulfur metabolism to Arabidopsis. APR activity is similarly restricted to the BS of maize, along with transcripts encoding other components of both primary and sulfur assimilation such as ATP Sulfurylase (ATPS), Sulfite Reductase (SIR), and APK (Burgener et al., 1998; Kopriva and Koprivova, 2005; Weckopp and Kopriva, 2015). A comparison of published mesophyll and BS transcriptomes across C3 Arabidopsis and rice as well as several C4 monocots (John et al., 2014; Döring et al., 2016; Rao et al., 2016; Berkowitz et al., 2021; Hua et al., 2021) shows broad conservation for the BS-enrichment of SULTR, ATPS, APR and SIR transcripts, whereas the rest of the primary and secondary sulfur metabolism transcripts show species-specific trends of BS or mesophyll enrichment (Fig. 4). However, this cell type-compartmentation seems to be unique to monocot C4 plants specifically, as transcript abundances of ATPS and APR are comparable between M and BS cells in the C4 eudicot species of the Flaveria genus (Koprivova et al., 2001; Gerlich et al., 2018). While the evolutionary and physiological relevance of this compartmentation for facilitating C4 photosynthesis is still unclear, C4 monocots such as maize and the model plant Setaria viridis may be useful tools for exploring cellular compartmentalizations in sulfur metabolism, and whether these drive concurrent specializations in retrograde signaling. Bundle sheath versus mesophyll compartmentation of sulfur metabolism across Arabidopsis and various monocots. Values represent the bundle sheath/mesophyll log2 ratio as calculated from existing transcriptomic data of laser-dissected A. thaliana (Berkowitz et al., 2021) and Oryza sativa (Hua et al., 2021), and mechanically separated Panicum virgatum (Rao et al., 2016), Zea mays and Setaria viridis (John et al., 2014), and Sorghum bicolor (Döring et al., 2016). Notation of −A/B next to the gene name indicates instances where a single A. thaliana gene has multiple orthologues in the monocots, e.g. GSH1-A and GSH1-B. N/A values are indicated in gray for genes without direct mapping to Arabidopsis orthologues or those that were not detected in respective datasets. Cells with shades of red indicate a positive bundle sheath/mesophyll log2 ratio, i.e. higher expression of the transcript in bundle sheath cells compared to mesophyll. Cells with shades of blue indicate a negative bundle sheath/mesophyll log2 ratio, i.e. lower expression of the transcript in bundle sheath cells compared to mesophyll. Cells coloured in white indicate no difference in expression of the transcript between bundle sheath and mesophyll cells. Note that the SOT17 and SOT18 orthologues detected in O. sativa, P. virgatum, S. viridis, and S. bicolor are likely to act on different substrates compared with A. thaliana, as these monocots lack glucosinolate metabolism. Abbreviations: APK, APS kinase; APR, APS reductase; APS, adenosine 5′-phosphosulfate; ATPS, ATP-sulfurylase; GSH1, glutamate-cysteine ligase; GSH2, Glutathione Synthetase; OAS-A/B/C, O-acetylserine(thiol)lyase A/B/C; SERAT, serine acetyltransferase; SIR, sulfite reductase; SOT, sulfotransferase; SULTR, sulfate transporter. Conclusion Plant sulfur metabolism plays a unique role in mediating plant cell responses to challenging environmental conditions. Through the redox biochemistry of the sulfhydryl group in cysteine residues (both within proteins and in the redox buffer GSH), primary sulfur metabolism directly influences the structure, function, activity, and regulation of proteins across the cell. This allows proteins to sense, transduce, and respond to H2O2 over-accumulation as part of retrograde signaling. Secondary sulfur metabolism is a source of at least one chloroplast/mitochondrial retrograde signal, and is regulated via redox regulation of cysteines in the phosphatase SAL1, demonstrating the influence of the two branches of sulfur metabolism on each other during retrograde signaling. These interactions may have implications for crop bio-engineering, for instance how enhanced PAP retrograde signaling in the wheat single SAL deletion lines (Bowerman et al., 2026) might impact on cysteine biosynthesis, and conversely how PAP retrograde signaling may be altered when cysteine biosynthesis is enhanced (Xu et al., 2024). Whilst most retrograde signaling studies to date have been conducted in Arabidopsis, emerging genomic resources and cell type-resolved technologies present unprecedented opportunities to interrogate sulfur metabolism and retrograde signaling in the context of cell type-specialization across land plant evolution. It may also be timely to expand our focus beyond photosynthetic leaves to under-studied tissues such as seeds where retrograde signals like PAP also function (Pornsiriwong et al., 2017; Pagano et al., 2022), and where sulfur nutrition plays an important role. Supplementary Material Contributor Information Riley Furbank, Research School of Biology, The Australian National University, Canberra, Australian Capital Territory 2601, Australia. Zuzana Plskova, Research School of Biology, The Australian National University, Canberra, Australian Capital Territory 2601, Australia; ARC Training Centre for Future Crops Development, Research School of Biology, The Australian National University, Canberra, Australian Capital Territory 2601, Australia. Barry Pogson, Research School of Biology, The Australian National University, Canberra, Australian Capital Territory 2601, Australia; ARC Training Centre for Future Crops Development, Research School of Biology, The Australian National University, Canberra, Australian Capital Territory 2601, Australia. Kai Xun Chan, Research School of Biology, The Australian National University, Canberra, Australian Capital Territory 2601, Australia; ARC Training Centre for Future Crops Development, Research School of Biology, The Australian National University, Canberra, Australian Capital Territory 2601, Australia. Hideki Takahashi, Michigan State University, USA. Supplementary data The following supplementary data are available at JXB online. Table S1. Relative expression of sulfur metabolism genes in bundle sheath versus mesophyll cells in various plant species. Funding This work was supported by the Australian Research Council to BJP [Laureate Fellowship (FL190100056), Training Centre for Accelerated Future Crops Development (IC210100047)] ; KXC [Discovery Early Career Researcher Award (DE210100466), Discovery Project (DP250100418)], and ZP [Training Centre for Accelerated Future Crops Development (IC210100047) PhD Scholarship]. ZP is also supported by the Grains Research and Development Corporation (GRDC) [PhD top-up scholarship (ANU2501-003RSX)]. RTF is supported by the Department of Education, Australian Government through a Research Training Program (RTP) PhD scholarship. Data availability This review contains no new experimental data. Supporting gene expression data that were analyzed, but not generated, as part of this review are from previously published studies and datasets [laser-dissected A. thaliana (Berkowitz et al., 2021) and Oryza sativa (Hua et al., 2021), and mechanically separated Panicum virgatum (Rao et al., 2016), Zea mays and Setaria viridis (John et al., 2014), and Sorghum bicolor (Döring et al., 2016)]. These datasets are cited within the text at relevant places. The compiled gene expression data used to generate Fig. 4 is available in Supplementary Table S1. References Associated Data Supplementary Materials Data Availability Statement This review contains no new experimental data. Supporting gene expression data that were analyzed, but not generated, as part of this review are from previously published studies and datasets [laser-dissected A. thaliana (Berkowitz et al., 2021) and Oryza sativa (Hua et al., 2021), and mechanically separated Panicum virgatum (Rao et al., 2016), Zea mays and Setaria viridis (John et al., 2014), and Sorghum bicolor (Döring et al., 2016)]. These datasets are cited within the text at relevant places. The compiled gene expression data used to generate Fig. 4 is available in Supplementary Table S1.