{"paper_id":"0966a36d-d17e-44a3-bfb8-a6946065fa23","body_text":"Metabolic enzymes are increasingly recognized as multifunctional proteins that extend beyond their classical catalytic roles in cellular metabolism. In addition to regulating metabolic flux, many exhibit non-canonical or “moonlighting” functions, mediated by changes in subcellular localization, post-translational modifications, or protein interactions. Through these mechanisms, metabolic enzymes can function as scaffolds, RNA-binding proteins (RBPs), and regulators of transcriptional or translational processes [ 1 , 2 ]. Importantly, such non-canonical functions are often highly context-dependent, emerging under specific physiological or stress conditions that alter protein localization, interaction networks, or structural states [ 1 ].\nProtein–protein interactions (PPIs) represent a central mechanism through which proteins acquire such multifunctionality. Through dynamic and context-dependent interactions, proteins can be repurposed to support regulatory functions that are not predicted by their canonical activity [ 3 ]. For example, the transcription factor NF-κB recruits ribosomal protein S3 (RPS3) as a non-canonical subunit to enhance DNA binding at specific target genes [ 4 ], illustrating how protein–protein interactions can repurpose proteins for regulatory functions.\nBeyond PPIs, the identification of RNA-binding capabilities in numerous metabolic enzymes—including GAPDH, PKM2, ASS1, and ADK—has further expanded the concept of protein multifunctionality, revealing that metabolic enzymes can directly participate in post-transcriptional regulation and feedback control of gene expression, thereby adding another layer of complexity to cellular and disease processes, including cancer, reviewed in [ 5 ].\nThe transsulfuration (TSS) pathway represents a metabolic system in which such multifunctionality may be particularly relevant. While traditionally associated with sulfur metabolism and redox homeostasis [ 6 ], emerging evidence suggests that enzymes within this pathway may participate in regulatory processes beyond their canonical biochemical roles, including functions related to post-transcriptional gene regulation [ 7 ]. However, the extent to which TSS enzymes contribute to these processes remains poorly defined.\nIn this review, we focus on cystathionine γ-lyase (CTH), a key enzyme of the TSS pathway, and examine evidence suggesting functions beyond its established roles in cysteine metabolism and hydrogen sulfide (H 2 S) production. Particular emphasis is placed on emerging links between CTH and translational regulation, including its potential role in regulating HIF-1α expression in clear cell ovarian carcinoma (CCOC) [ 7 ].\n\nThe TSS pathway is a central component of sulfur metabolism that links methionine metabolism to cysteine biosynthesis while generating key bioactive molecules. It produces cysteine, which serves as a precursor for glutathione (GSH) and supports cellular redox homeostasis, as well as H 2 S, a gaseous signaling molecule with diverse roles in physiological and pathological processes [ 8 , 9 , 10 ]. This pathway is primarily mediated by the pyridoxal-5′-phosphate-dependent enzymes cystathionine β-synthase (CBS) and cystathionine γ-lyase (CTH, also known as CSE or CGL), which catalyze the conversion of homocysteine to cysteine. Canonically, CBS condenses homocysteine with serine to form cystathionine, which is subsequently cleaved by CTH to yield cysteine. Beyond these canonical reactions, both CBS and CTH can also utilize cysteine and/or homocysteine in β- or γ-elimination reactions that release H 2 S, making sulfide production an intrinsic output of the TSS pathway [ 10 , 11 ]. In addition, 3-mercaptopyruvate sulfurtransferase (MPST) contributes to H 2 S production within the broader sulfur metabolic network [ 10 , 12 ] ( Figure 1 ).\nA considerable body of evidence indicates that H 2 S and its primary producing enzymes—CBS, CTH, and MPST— play important roles in cancer progression by promoting tumor cell proliferation, supporting cellular bioenergetics, and stimulating angiogenesis [ 13 , 14 , 15 , 16 ]. While the catalytic activities of TSS enzymes and regulation by metabolites such as homocysteine and cysteine are well characterized, increasing evidence suggests additional non-catalytic and regulatory functions. These activities extend beyond canonical roles in cysteine biosynthesis and H 2 S generation. For instance, CBS contains a unique N-terminal heme-binding domain in which the heme is not required for catalytic activity but instead functions as a redox sensor that regulates both CBS enzymatic activity and broader cellular processes [ 17 , 18 ]. Moreover, CBS interacts with the core circadian protein cryptochrome 1 (CRY1) in U2-OS and NIH3T3 cells, implicating it in circadian regulation [ 19 ]. In addition, MPST has been shown to regulate inflammatory and epithelial homeostasis in an H 2 S-independent manner through direct interaction with AKT and modulation of its phosphorylation status [ 20 ]. Despite these findings, systematic characterization of the non-canonical functions of TSS enzymes remains limited. Notably, CTH is of particular interest because it occupies a central position in sulfur metabolism as the primary mammalian enzyme responsible for de novo cysteine synthesis from cystathionine, thereby supporting glutathione (GSH) production, especially under conditions of limited exogenous cysteine availability [ 6 , 15 ], and as a major source of H 2 S in peripheral tissues [ 21 ], with emerging evidence also supporting important roles in the central nervous system [ 22 ].\nCTH is a tetrameric enzyme, with each monomer covalently bound to a pyridoxal 5′-phosphate (PLP) cofactor, the active form of vitamin B6. Consistent with its central metabolic and signaling roles, CTH is tightly regulated, and its expression and activity are highly context-dependent. Epigenetic, transcriptional, and post-transcriptional mechanisms collectively modulate CTH expression [ 10 , 15 , 23 , 24 ] ( Figure 2 ).\nLoss of CTH function causes the rare metabolic disorder cystathioninemia (or cystathioninuria) [ 25 ]. While vitamin B6-responsive and nonresponsive forms have been described [ 26 , 27 ], the condition is generally considered clinically mild, and individuals lacking functional enzyme often exhibit favorable outcomes, suggesting that CTH deficiency is well tolerated in humans and is not associated with severe systemic dysfunction [ 28 , 29 , 30 ].\nMultilevel regulation of CTH. CTH expression and activity are regulated at epigenetic, transcriptional, and post-transcriptional levels. (Top panel: Epigenetic and transcriptional regulation)  CTH  transcription is modulated by diverse cellular stress signals, including hypoxia, oxidative stress, endoplasmic reticulum (ER) stress, Golgi stress, and mitochondrial stress, as well as inflammation, nutrient deprivation, and hormonal cues (e.g., growth hormone (GH) and thyroid hormone (TH)). Epigenetic regulation via CpG island methylation within the  CTH  promoter represses gene expression. Multiple transcription factors directly bind the  CTH  promoter to regulate transcription: specificity protein 1 (SP1) maintains basal expression; nuclear factor erythroid 2-related factor 2 (Nrf2) mediates induction under oxidative stress; activating transcription factor 4 (ATF4) regulates expression during amino acid starvation; ETS Like-1 protein (Elk1) contributes to promoter activation; and cAMP response element-binding protein (CREB) enhances transcription. In addition, farnesoid X receptor (FXR) directly binds a specific response element (AGTTCAgTGTACCT) within the  CTH  promoter to regulate its expression. (Bottom panel: Post-transcriptional regulation) CTH expression is further controlled by microRNAs (e.g., miR-21 and miR-30s family), which modulate mRNA stability and translation. Protein abundance and enzymatic activity are influenced by substrate availability and protein degradation pathways. Post-translational modifications (PTMs), including phosphorylation, S-sulfhydration, SUMOylation, ubiquitination, and acetylation, as reported in the literature [ 31 , 32 ] and curated in PhosphoSitePlus ( www.phosphosite.org , accessed on 13 February 2026), dynamically regulate CTH stability and function. Together, these multilayered regulatory mechanisms coordinate CTH-dependent hydrogen sulfide (H 2 S) production and transsulfuration pathway activity under physiological and pathological conditions. Some elements of the schematic were generated with assistance from ChatGPT (OpenAI, GPT-5).\nStudies using  Cth − / −  mouse models have yielded variable results. While some reports suggest a role for CTH in neuroprotection, its deficiency has been associated with cognitive impairment [ 22 ], and others describe largely normal phenotypes aside from hypertension [ 33 ]. Moreover, additional studies have demonstrated that  Cth − / −  mice, despite developing normally, exhibit increased susceptibility to oxidative stress under cysteine-restricted conditions [ 34 ]. Collectively, these discrepancies likely reflect differences in genetic background, experimental conditions, and species-specific compensatory mechanisms between mice and humans.\nIn addition to these context-dependent effects observed in loss-of-function models, emerging evidence suggests context-specific non-canonical functions of CTH in distinct cellular settings. Notably, CTH overexpression in CD8 +  T cells enhances antitumor activity by reshaping the tumor microenvironment through depletion of extracellular glycine, serine, and proline, rather than by promoting T cell proliferation. Although the TSS pathway is present in T cells, CTH overexpression does not confer cysteine independence or rescue proliferation under cysteine/cystine-deprived conditions [ 35 ]. Additionally, evidence for non-canonical CTH functions has emerged in neurological disease context. In a Parkinson’s disease-related astrocyte model, increased CTH/CSE expression was associated with astrocyte state transitions towards a neurotoxic phenotype, alongside an overall reduction in H 2 S levels caused by decreased CBS and MPST expression. This effect on astrocytes was linked to FOXD3-mediated transcriptional reprogramming [ 36 ], supporting context-dependent regulatory interactions of CTH beyond its canonical role in H 2 S production.\nIn cancer, increased CTH expression has been associated with cancer progression and metastasis in multiple cancers, including breast cancer [ 37 ], prostate cancer [ 38 ], glioblastoma [ 39 ], nasopharyngeal carcinoma [ 40 ], and other malignancies, where the pro-tumorigenic effects of CTH have largely been attributed to activation of the CTH/H 2 S axis; however, direct phenotypic rescue by exogenous H 2 S at physiologically relevant concentrations has not consistently been demonstrated, leaving the specific contribution of H 2 S to these effects incompletely resolved.\nFunctional studies in specific tumor contexts further challenge an H 2 S-centric model. In Ewing sarcoma, CTH supports survival under cystine limitation by maintaining glutathione-dependent redox homeostasis, and viability can be restored in  CTH  knockdown cells by antioxidant supplementation but not by H 2 S donors [ 41 ]. Together, these findings point to context-dependent functions of CTH that are not universally explained by H 2 S production and may extend beyond canonical redox metabolism.\n\nH 2 S is now recognized as an important gasotransmitter generated through enzymatic and non-enzymatic pathways, as well as from dietary sources and the gut microbiota [ 8 , 42 ]. It interacts with diverse biomolecules and regulates processes including metabolism, vascular tone, immune responses, and neuronal signaling with potential implications for both health and disease [ 8 , 43 ].\nEarly investigations into H 2 S biology were shaped by its established toxicity as an environmental hazard [ 44 , 45 ], driven by its potent inhibition of cellular respiration, later attributed to the suppression of cytochrome c oxidase (complex IV) in the mitochondrial electron transport chain [ 46 , 47 , 48 ]. This toxicological view was further reinforced by early experimental studies employing supraphysiological (high micromolar to millimolar) concentrations of fast-releasing inorganic donors (e.g., Na 2 S or NaHS) [ 49 , 50 ]. While earlier studies were instrumental in defining the chemical reactivity of H 2 S, the concentrations used often exceed those found in vivo and may not fully reflect endogenous signaling conditions. Consequently, these approaches may have disproportionately emphasized inhibitory effects that differ from those occurring under physiological conditions. High donor concentrations likely overwhelmed redox-sensitive targets, eliminated signaling specificity, and induced acute mitochondrial inhibition that dominated cellular phenotypes. Such conditions may also have obscured dose-dependent signaling effects, reinforcing the misconception that H 2 S functions primarily as a toxic or metabolic inhibitor (see reviews [ 51 , 52 , 53 , 54 ]).\nMore recent quantitative analyses indicate that free H 2 S levels in tissues are typically maintained within the low nanomolar to low micromolar range, with rapid turnover and tightly regulated spatial dynamics [ 55 , 56 ]. Within this physiological window, H 2 S exerts context- and compartment-dependent effects on mitochondrial function and cellular bioenergetics and engages signaling mechanisms distinct from those observed at higher concentrations, potentially including protein persulfidation of reactive cysteine residues or signaling via reactive sulfur species such as polysulfides [ 52 , 53 , 54 , 57 ]. Collectively, these findings support a model in which H 2 S functions as a regulated signaling molecule whose biological effects are highly dependent on concentration, cellular context, and subcellular localization.\nIn cancer, H 2 S can exert context-dependent and sometimes opposing effects depending on tumor type, metabolic state, and microenvironment, displaying both tumor-promoting activities, as reported in NSCLC [ 58 ], esophageal cancer [ 59 ], colorectal cancer [ 60 ], among others, and tumor-suppressive effects, as observed in the ID8 murine epithelial ovarian cancer cell line [ 61 ] and in human breast adenocarcinoma as well as hepatocellular carcinoma cell lines [ 62 ] (see also the review in [ 63 ]).\nExperimental studies using exogenous H 2 S donors further support the idea that sulfide signaling produces distinct effects across cellular systems. Lee et al. [ 64 ] showed that H 2 S donors (GYY4137 and NaHS) induce G2/M cell cycle arrest and apoptosis across multiple cancer cell lines, with limited effects in normal fibroblasts. Similarly, Xiao et al. [ 65 ] reported that Na 2 S increases reactive oxygen species (ROS) levels in glioblastoma cells, promoting apoptosis and enhancing radiosensitivity, effects not observed in normal human brain microvascular endothelial cells. In line with these findings, Zhao et al. [ 66 ] demonstrated that NaHS induces apoptosis in glioma cells through activation of p38/MAPK and p53 signaling pathways. Collectively, these studies highlight the context-dependent nature of H 2 S signaling across cellular systems.\nCTH has historically been considered a major enzymatic source of H 2 S, supported by its tissue expression profile and murine knockout studies showing reduced H 2 S levels upon genetic deletion [ 33 , 67 , 68 ]. However, this model is increasingly challenged by human genetic and metabolic data. Kozich et al. demonstrated that in patients with rare inborn errors affecting CBS or CTH, circulating bioavailable sulfide is not decreased and may even increase in CBS deficiency, indicating strong compensatory mechanisms that preserve systemic sulfur homeostasis [ 69 ]. These findings are particularly relevant for cancer biology, where enzymatic dependencies inferred from preclinical models may not directly reflect human metabolic adaptation. Instead, they suggest that disruption of canonical TSS enzymes may be buffered by alternative enzymatic, nutritional, or microbiota-derived sulfur sources.\nFurther supporting this concept, recent studies in mammalian systems, including mouse models, show that sulfide and persulfide production can persist even when canonical TSS enzymes are disrupted, owing to the activity of cysteinyl-tRNA synthetase (CARS), which functions as a cysteine persulfide synthase and contributes substantially to intracellular sulfide and persulfide production in mammals [ 70 ]; these species are also proposed to serve as a reservoir for H 2 S [ 71 ]. Of interest, earlier work suggested that CBS- and CTH-mediated reactions predominantly generate persulfides rather than free H 2 S, with H 2 S largely arising as a downstream product of persulfide degradation [ 72 ]. This view has since evolved with evidence that persulfides function as distinct bioactive sulfur species rather than merely intermediates in H 2 S release [ 73 ]. Recent work further indicates that their signaling properties are chemically diverse and may involve electrophilic redox signaling and direct sulfur transfer through transpersulfidation reactions independent of free H 2 S generation [ 74 ]. These findings further highlight that persulfide biology cannot be reduced to H 2 S release alone, but involves direct and chemically diverse modes of action. Consistent with this buffered and distributed sulfur network, endothelial-specific deletion of CTH in mice does not significantly alter tissue H 2 S levels but instead affects sulfane sulfur pools [ 75 ], indicating that H 2 S output is maintained through metabolic redundancy and that sulfur regulation extends beyond single-enzyme control.\nBeyond sulfur buffering, several studies also suggest that CTH-associated phenotypes may diverge from canonical H 2 S output. Geng et al. reported that both pharmacological elevation of H 2 S levels using a donor and inhibition of endogenous H 2 S production through CTH blockade improved insulin sensitivity in obese mice, despite exerting opposite effects on lipolysis [ 76 ]. These findings suggest that CTH-associated metabolic effects are not strictly concordant with H 2 S signaling and may reflect partially decoupled roles in metabolic regulation.\nFurthermore, Bibli et al. showed that inflammatory signaling induces phosphorylation of CTH at Ser377, resulting in enzymatic inactivation and reduced H 2 S production despite maintained or increased CTH protein expression [ 77 ]. This indicates that post-translational modification can uncouple CTH abundance from H 2 S production, suggesting that CTH activity—and its downstream functions—may be regulated independently of canonical sulfide output. Importantly, enzyme-selective regulation of H 2 S production has also been reported, as Qi et al. demonstrated that estrogen-dependent H 2 S production in endometrial stromal cells is mediated specifically through CBS rather than CTH, supporting enzyme-selective regulation of H 2 S signaling in a physiological context [ 78 ].\nCollectively, these observations challenge the assumption that CTH-derived H 2 S, and potentially H 2 S produced by other enzymes, represents a stable or direct functional output across biological contexts. Instead, they suggest that H 2 S may not fully account for CTH-associated biological activity, and support a model in which tightly regulated sulfide levels operate as part of a broader network of reactive sulfur species and metabolic fluxes.\nCysteine metabolism highlights a parallel limitation of the canonical model. Cysteine is central to redox homeostasis and anabolic metabolism, serving as a precursor for glutathione (GSH), iron–sulfur clusters, coenzyme A, and other sulfur-containing metabolites, while also contributing directly to protein synthesis [ 6 , 79 ]. Although TSS-derived cysteine production has been proposed to support survival under cysteine-limiting conditions in certain cancers such as neuroblastoma [ 13 , 80 ] and Ewing sarcoma [ 41 ], its importance is highly context-dependent and not universally rate-limiting across cancers. In line with this, Zhang et al. found that CBS and CTH expression levels do not correlate with cysteine starvation sensitivity across cancers, which instead depends on broader metabolic context, including polyamine-associated metabolic rewiring that increases oxidative stress under cysteine limitation [ 81 ].\nFunctional studies further support the limited buffering potential of TSS under nutrient stress. Kang et al. showed that in a broad panel of non-small cell lung cancer (NSCLC) cell lines, de novo cysteine synthesis is insufficient to sustain intracellular cysteine pools during cystine starvation, leading to impaired GSH synthesis and ferroptosis, an oxidative stress-driven form of cell death. Although partial compensation occurs through glutamate–cysteine ligase catalytic subunit (GCLC)-mediated γ-glutamyl peptide production, this mechanism primarily limits glutamate accumulation rather than restoring canonical antioxidant capacity [ 82 ]. Extending these observations to hematological malignancies, studies in acute myeloid leukemia (AML) showed that CRISPR-mediated CTH loss did not substantially sensitize cells to cystine deprivation, although differences in CBS and CTH expression influenced the ability of AML cells to adapt to cysteine stress [ 83 ]. In vivo metabolic tracing studies corroborate these findings, demonstrating that TSS contributes minimally to cysteine pools in most non-hepatic tissues, where extracellular cyst(e)ine remains the dominant source. Tumor cysteine metabolism therefore largely reflects tissue-of-origin constraints rather than intrinsic TSS capacity [ 84 ].\nCCOC may represent a useful model system for investigating the TSS, as it exhibits a functional TSS pathway, relies on cysteine metabolism for growth, and displays heterogeneous sensitivity to cysteine deprivation across cell lines. Cysteine deprivation is further associated with context-dependent oxidative stress responses, including necrosis, ferroptosis, and apoptosis depending on metabolic state. Notably, intracellular cysteine levels vary across cell lines, whereas glutathione levels remain comparable despite differences in basal cysteine abundance [ 85 ], suggesting that cysteine availability may not be directly reflected in steady-state glutathione pools. Furthermore, ferroptosis sensitivity is influenced not only by sulfur metabolism but also by oncogenic signaling. Hyperactivation of the PI3K/AKT/mTORC1 pathway promotes ferroptosis resistance through lipid remodeling programs involving SREBP1 and the accumulation of monounsaturated fatty acids, potentially mediated by SCD1 [ 86 , 87 ]. This is particularly relevant for CCOC, which frequently harbors PI3K/AKT pathway activation, suggesting that ferroptosis resistance may arise from the integration of metabolic, lipid, and signaling networks rather than isolated metabolic inputs or redox pathways alone.\nCCOC also provides a compelling model for examining potential non-canonical functions of CTH. We previously demonstrated that CTH, but not CBS or MPST, is required for hypoxia-induced HIF-1α protein accumulation in CCOC cells [ 7 ], indicating a CTH-specific effect. At the transcriptional level,  HIF1A  mRNA levels in  CTH  knockout (KO) cells were comparable to control cells, and actinomycin D-based mRNA decay assays showed no difference in  HIF1A  transcript stability, arguing against transcriptional regulation. At the protein level, neither the proteasome inhibitor MG132 nor the prolyl hydroxylase inhibitor DMOG fully restored HIF-1α in  CTH  KO cells to levels comparable to those in control cells, and cycloheximide chase experiments showed similar HIF-1α degradation rates between conditions. Together, these data exclude altered degradation or protein stability as mechanisms [ 7 ].\nGiven known links between CTH and H 2 S, our previous study observed increased H 2 S levels in  CTH  KO cells, likely due to compensatory CBS upregulation and enhanced CBS activity, both of which were experimentally demonstrated [ 7 ]. This was particularly relevant in light of prior studies demonstrating that H 2 S can regulate HIF-1α in a context-dependent manner, with reports of both inhibitory [ 88 ] and stimulatory [ 89 ] effects depending on cellular system and experimental conditions, and that CBS-derived H 2 S can regulate HIF-1α stability in non-ovarian models through persulfidation and activation of prolyl hydroxylase 2 (PHD2), thereby promoting HIF-1α hydroxylation and degradation [ 90 ]. However, several observations argued against H 2 S-mediated regulation in the CCOC model. Pharmacologic supplementation with either the slow-releasing H 2 S donor GYY4137 or the fast-releasing donor Na 2 S failed to alter HIF-1α protein expression in control cells. Furthermore, a catalytically impaired CTH mutant carrying an arginine-to-alanine substitution at residue 62 (R62A), a residue reported to be essential for CTH enzymatic activity [ 91 ], restored HIF-1α expression in  CTH  KO cells, supporting a potential non-enzymatic role for CTH. In addition, CTH-driven phenotypes, including reduced viability and motility, are not fully explained by redox metabolism, as antioxidant supplementation or glutathione monoethyl ester (GSH-MEE) fails to fully rescue the effects of CTH loss under cysteine-replete conditions [ 7 ]. Together, these findings indicate that CTH-dependent regulation of HIF-1α cannot be fully explained by established roles in sulfur metabolism, redox control, or H 2 S production.\nCollectively, these observations across  Section 3.1 ,  Section 3.2  and  Section 3.3  support a conceptual shift in which sulfur metabolism in mammalian systems is better described as a buffered, redundant, context-dependent network of reactive sulfur species and metabolic fluxes rather than linear outputs of H 2 S or cysteine production. Within this framework, CTH may function as a node in this distributed regulatory system contributing to redox balance, metabolic plasticity, and stress adaptation. In cancer, the frequent upregulation of CTH may reflect integration into broader metabolic and stress-response programs, alongside additional non-canonical functions beyond its enzymatic roles that remain to be defined. Structural and spatial mechanisms may further contribute to these non-canonical functions.\n\nCTH is classically reported as a predominantly cytosolic enzyme [ 92 ]; however, multiple studies suggest that it may localize to diverse intracellular and extracellular compartments, including the nucleus [ 93 ], mitochondria [ 94 ], plasma membrane [ 95 ], endoplasmic reticulum [ 95 ], plasma [ 96 ], and urine [ 97 ], where it may exert compartment-specific functions. In addition, large-scale proteomic analyses [ 98 ], sequence-based predictions, and corresponding annotations in curated databases such as GeneCards [ 99 ] support a multi-compartment distribution of CTH across intracellular and extracellular environments ( Table 1 ), warranting further investigation of its context-specific functions.\nPost-translational modifications (PTMs) may provide a mechanistic basis for such dynamic subcellular distribution. Agrawal et al. demonstrated that both CBS and CTH can undergo SUMOylation in vitro, a modification proposed to regulate nuclear translocation and potentially enable compartment-specific functions [ 31 ]. Consistent with this, Drekolia et al. [ 93 ] recently demonstrated that CTH can translocate to the nucleus in a cystine-dependent manner despite lacking a classical nuclear localization signal, where it contributes to cystine oxidation and the generation of acetyl units that promote histone H3 acetylation and chromatin remodeling, thereby linking metabolism to transcriptional regulation.\nIn addition, both CTH and CBS are secreted by endothelial cells and hepatocytes, circulate in plasma, and remain catalytically active in blood, generating H 2 S from homocysteine [ 96 ]. This extracellular activity further indicates that TSS enzymes are not confined to intracellular metabolism. Collectively, these observations support a broader model in which CTH operates across intracellular and extracellular environments rather than as a solely cytosolic enzyme in cysteine metabolism and H 2 S generation. This spatial distribution may enable context-dependent functions through both catalytic and interaction-driven mechanisms.\nBeyond spatial compartmentalization, intrinsic structural properties of CTH may further contribute to its functional plasticity. Structural studies of human CTH reveal that the apo enzyme adopts an open conformation in the absence of PLP, transitioning to a more closed state upon cofactor binding. This ligand-dependent conformational shift highlights the intrinsic flexibility of the protein [ 100 ], suggesting that CTH can sample multiple structural states. Such conformational plasticity may enable interactions beyond its canonical catalytic role, including context-dependent non-canonical functions.\nIn line with this conformational plasticity, additional structural analyses indicate that CTH is not rigidly folded but exhibits localized flexibility and partial disorder in specific regions [ 91 ]. Notably, two loop regions (Met110-Asn118 and Thr210-Met216), which flank the PLP-binding cleft, undergo conformational rearrangement upon cofactor binding, folding back over the active site to stabilize ligand interactions ( Figure 3 ). Although these observations do not indicate PLP-independent catalytic activity, they suggest that CTH structure is sensitive to its molecular environment, a property that may facilitate regulated interactions or non-canonical functions in specific cellular contexts. Such structural plasticity is a recognized feature of proteins capable of transient and multivalent interactions.\nPartial disorder and conformational flexibility are increasingly recognized as enabling features for nucleic acid interactions. Although classical RBPs typically contain defined RNA recognition motifs (RRMs) or KH domains, many non-canonical RBPs lack these canonical features yet still associate with RNA [ 101 , 102 , 103 ]. Intrinsically disordered regions (IDRs) play a key role in mediating such interactions by enabling conformational adaptability and dynamic binding, and are increasingly recognized as hallmark features of proteins that function as interaction hubs within signaling networks [ 104 , 105 , 106 , 107 ]. In this context, CTH exhibits features consistent with RNA-binding potential, which may contribute to its capacity to engage diverse interaction partners. For example, its N-terminal region includes an unstructured segment, and specific regions flanking the PLP-binding cleft can show flexibility and can become partially disordered under certain conditions [ 91 ]. Computational analysis using the Database of Disordered Protein Predictions (D2P2) [ 108 ] ( https://d2p2.pro/search , accessed on 24 March 2026), by querying human CTH UniProt ID  P32929 , predicts multiple intrinsically disordered regions in human CTH, some of which overlap with potential post-translational modification sites ( Figure 4 ), suggesting regulatory complexity. Consistent with this structural disorder,  Pseudomonas aeruginosa  CTH/CGL (PaCGL) shows similar features, including an unstructured N-terminal segment (first 13 residues), a disordered loop region (residues 46–57) [ 109 ], and localized conformational variability in two long loops (L23–60 and L347–370), which can adopt partially disordered or alternative conformations, including an extended non-helical state in one subunit [ 109 ]. Together, these structural and spatial properties may provide a mechanistic basis for the involvement of CTH in diverse molecular interaction networks.\n\nBuilding on its multi-compartment distribution, CTH engages in protein interaction networks that may underlie its context-dependent functions. Emerging evidence suggests that these interactions extend beyond canonical metabolic roles, implicating CTH in signaling processes, stress responses, and transcriptional regulation. Hu et al. showed that CTH counteracts endothelial cell senescence by binding to p53, promoting its cytoplasmic retention and preventing its acetylation and transcriptional activation, independently of p53 S-sulfhydration by H 2 S [ 110 ]. Furthermore, Zhu et al. demonstrated that CTH forms a complex with YAP in a Parkinson’s disease-related astrocyte model, facilitating FOXD3-mediated transcriptional regulation while modulating YAP nuclear translocation [ 36 ]. In line with this emerging view of CTH as a multifunctional regulatory protein, Drekolia et al. [ 93 ] reported that CTH/CSE associates with pyruvate dehydrogenase E1 subunit alpha 1 (PDHA1) in both cytosolic and nuclear compartments of pre-proliferative endothelial cells in a cystine-dependent manner, with coordinated nuclear localization suggesting compartment-specific metabolic–nuclear coupling. Reciprocal immunoprecipitation further revealed that PDHA1 complexes contain histone acetyltransferases, including GCN5 and HAT1, suggesting the assembly of a nutrient-sensitive multi-protein complex linking metabolic enzymes to chromatin-modifying machinery. Notably, this interaction network is associated with selective regulation of histone H3-, H4-, and H2A-associated acetylation marks, particularly H3K9ac, H3K23ac, and H2AK5ac, accompanied by downstream changes in chromatin accessibility and transcriptional programs. Importantly, Drekolia et al. [ 93 ] further showed that perturbation of sulfide flux using a fast-releasing H 2 S donor does not significantly impact vascular growth in vivo, suggesting a limited contribution of H 2 S signaling in this context. Together, these findings support a model in which CTH participates in context-dependent protein interaction networks that integrate metabolic status with signaling and transcriptional regulation.\nAt a systems level, curated interaction databases (BioGRID [ 111 , 112 ], IntAct [ 113 , 114 ], HitPredict [ 115 ]) and large-scale affinity purification–mass spectrometry (AP-MS)-based proteomic interactome datasets [ 98 , 116 , 117 , 118 , 119 ] consistently place CTH within extensive protein association networks across eukaryotic systems. These resources collectively indicate that CTH participates in broad functional modules spanning immune, metabolic, transcriptional, and structural pathways. To provide a representative overview,  Figure 5  summarizes a curated subset of CTH-associated interactions derived from the BioGRID database (version 5.0.257), highlighting representative proteins involved in translational regulation, signaling, and cellular homeostasis. This includes YTHDF1 [ 120 ], PARK2 [ 121 ], SDC1 [ 98 ], CUL4A [ 122 ], along with additional network-associated proteins such as RECK, WDYHV1, PTCRA, SMOC1, SLC25A32, HOXD3, KLHL20, ARHGEF39, and SCG3.\nImportantly, additional CTH-associated proteins identified through independent interaction studies and curated datasets, including YWHAZ (14-3-3ζ/δ) [ 123 ] and the RBPs ELAVL1/HuR and PTBP1 [ 77 ], further expand this interaction landscape beyond the BioGRID-derived subset shown in  Figure 5 . Collectively, these observations position CTH within broader signaling and post-transcriptional regulatory networks potentially linked to processes such as mRNA translation and cellular stress responses. The functional and subcellular diversity of these interacting partners—spanning cytoplasmic, mitochondrial, nuclear, and secretory compartments—is consistent with the reported multi-compartment localization of CTH [ 98 ] and supports its activity across diverse cellular contexts. However, functional validation of many of these interactions remains unresolved. Together, these findings support the emerging view that CTH interfaces with multiple signaling and RNA-regulatory networks.\n\nEmerging evidence places CTH at the interface between cellular signaling pathways and the molecular machinery governing protein synthesis, suggesting it may influence translation through both indirect signaling effects and proximity-based interactions with translational regulators.\nCTH has been linked to modulation of the PI3K/AKT/mTOR pathway, a central regulator of mRNA translation. In breast cancer models, CTH has been shown to positively regulate the PI3K/AKT pathway, with its upregulation increasing PI3K, Akt, and phospho-Akt levels and its knockdown producing the opposite effect [ 37 ]. This is consistent with findings in triple-negative breast cancer (TNBC) cells, where pharmacological inhibition of CTH reduces PI3K/AKT signaling, as evidenced by decreased PI3K and AKT levels as well as AKT phosphorylation in MDA-MB-231 and MDA-MB-468 cells [ 124 ]. Similarly, in nasopharyngeal carcinoma models, CTH overexpression increases phospho-PI3K, phospho-AKT, and phospho-mTOR levels, whereas knockdown reduces their expression [ 40 ]. Through this signaling axis, CTH may influence downstream translational regulators such as 4E-BPs and S6K, thereby affecting mRNA translation through mechanisms that include modulation of eIF4F complex assembly; however, direct evidence for regulation of these translational effectors by CTH remains lacking. These effects are likely context-dependent and shaped by cellular redox state and metabolic flux.\nBeyond upstream signaling, large-scale interactome studies suggest that CTH is linked to components of the translational machinery identified across species. CTH has been shown to associate with ribosomal protein S2 (RpS2) in  Drosophila melanogaster  and  Saccharomyces cerevisiae  [ 125 ], as well as with eIF3j, a subunit of the eukaryotic translation initiation factor 3 complex, in  Drosophila melanogaster  [ 126 ]. Given the evolutionary conservation of both CTH and eIF3j, these findings suggest a potential conserved association with translation initiation machinery.\nBroader interactome datasets extend this network to additional translational and signaling-associated proteins. CTH-associated proteins include YTHDF1 [ 120 ] and YWHAZ (14-3-3ζ/δ) [ 123 ]. YTHDF1, a canonical m 6 A reader, promotes translation efficiency of target mRNAs through recruitment of eukaryotic initiation factor 3 (eIF3) [ 127 , 128 ], while YWHAZ has been implicated in selective mRNA translation [ 123 , 129 ], and in PI3K/AKT/mTOR pathway activation [ 130 ]. Additional network-level evidence links CTH to PARK2 [ 121 ], which has been reported to interact with translation initiation factors including eIF4B in mammalian cells [ 131 , 132 ] and eIF4E in  Drosophila melanogaster  [ 133 ]. Furthermore, CTH-associated AP-MS datasets identify interacting partners including SDC1 [ 98 ] and CUL4A [ 122 ], both of which function as modulators of PI3K/AKT signaling [ 134 , 135 ], thereby providing additional indirect links to mRNA translational control.\nAdditional interactome analyses in murine endothelial cells link CTH to RBPs, including ELAVL1 and PTBP1 [ 77 ]. Both proteins are established regulators of  HIF1A  mRNA translation [ 136 ], suggesting a potential link between CTH-associated complexes and transcript-selective translational control.\nCollectively, these findings support a model in which CTH is positioned within translational regulatory networks through a combination of signaling-mediated effects and protein interaction-dependent associations. This framework suggests involvement of CTH in both global translational control and more selective regulatory processes, although direct mechanistic evidence remains limited and further experimental validation is required.\n\nMotivated by findings presented in preceding sections, CCOC provides a biologically and clinically relevant context to further investigate the non-canonical functions of CTH. Although current mechanistic insights in this context are still emerging and are informed by prior studies, including our own [ 7 ], CCOC provides a valuable framework for exploring alternative paradigms of CTH biology. CCOC, which often arises from ovarian endometriotic cysts (endometriomas) [ 137 ], is an aggressive subtype of epithelial ovarian cancer, accounting for approximately 12% of cases and characterized by limited treatment options and poor clinical outcomes [ 138 , 139 ]. This malignancy is defined by a distinct molecular and metabolic landscape, including frequent loss of ARID1A, hyperactivation of the PI3K/AKT/mTOR pathway, and a pronounced hypoxic signature [ 138 , 140 , 141 , 142 ] ( Figure 6 ). In addition, CCOC exhibits marked resistance to oxidative and metabolic stress through adaptive antioxidant mechanisms that extend beyond glutathione-dependent redox homeostasis [ 143 ]. Collectively, these features create a cellular environment in which translational control plays a critical role in tumor adaptation and progression.\nNotably, CTH expression is elevated in CCOC relative to other epithelial ovarian cancer subtypes [ 144 , 145 ]. In this context, and as discussed in  Section 3.3 , CTH-dependent regulation of HIF-1α in CCOC is not explained by canonical enzymatic functions or by transcriptional and protein stability mechanisms [ 7 ]. Given the central role of translational regulation in controlling HIF-1α protein expression [ 146 , 147 , 148 ], these observations raise the possibility that CTH may contribute to HIF-1α regulation and stress adaptation through mechanisms potentially linked to translational regulation. Of note, CTH expression is further upregulated in CCOC cells under hypoxia and following treatment with endometriotic cyst contents, consistent with a potential role in adaptive responses to the endometriosis-associated tumor microenvironment [ 7 ]. Within this framework, activation of the PI3K/AKT/mTOR pathway in CCOC provides a strong driver of mRNA translation, while hypoxic and oxidative stress conditions favor selective translation of stress-responsive transcripts [ 149 , 150 ]. CTH may therefore function as a modulatory factor integrating metabolic signaling with mRNA translation.\n\nBased on converging structural, interactome, and functional evidence, we propose a model in which CTH may function as a context-dependent regulator of mRNA translation, including potentially  HIF1A  and other adaptive transcripts, thereby linking cellular metabolic state to adaptive gene expression programs in CCOC.  Figure 7  summarizes four regulatory axes through which CTH may influence CCOC progression and adaptation to its microenvironment: (1) modulation of PI3K/AKT signaling, a pathway frequently reported to be altered upon CTH modulation; (2) association with components of the translational initiation machinery; (3) interaction with  HIF1A  mRNA, either directly or indirectly through RBP-mediated mechanisms; and (4) maintenance of glutathione-dependent redox homeostasis and ferroptosis resistance. Collectively, these pathways may promote cellular adaptation to hypoxic and oxidative stress conditions characteristic of the CCOC microenvironment. The first three axes form the basis of the translational regulatory mechanisms discussed below, whereas redox homeostasis represents a parallel adaptive pathway that may cooperate with translational control.\nCTH function in translational regulation can be mechanistically conceptualized across three complementary and non-mutually exclusive models.\nModel 1: CTH may modulate global translational control through the PI3K/AKT/mTOR signaling axis, thereby modulating overall translation capacity. This mechanism is supported most directly by evidence linking CTH perturbation to changes in this pathway across tumor contexts.\nModel 2: CTH may contribute to transcript-selective translation through associations with components of the translational machinery in both cross-species systems (e.g., eIF3j and RpS2) and mammalian systems (e.g., YTHDF1, which promotes translation via recruitment of eIF3), as well as RBPs such as the  HIF1A  mRNA translational regulators ELAVL1 and PTBP1. Experimental validation of ELAVL1 [ 77 ] supports the RNA-binding protein-associated interactions, although direct evidence linking CTH to selective translational control remains limited.\nModel 3: CTH could potentially interact with  HIF1A  mRNA. This possibility is currently hypothesis-generating and supported primarily by computational prediction. Sequence-based prediction using RPISeq [ 151 , 152 ], which applies Support Vector Machine (SVM) and Random Forest (RF) classifiers trained on PRIDB datasets [ 153 ], suggests a potential interaction between CTH and  HIF1A  mRNA. As shown in  Figure 8 A, CTH exhibits higher predicted interaction probabilities with  HIF1A  (RF: 0.75; SVM: 0.99) than with a comparison transcript ( PTBP1 ; RF: 0.60; SVM: 0.68), indicating potential transcript selectivity rather than nonspecific RNA binding; however, these differences should be interpreted only as suggestive of transcript selectivity and not as evidence of binding specificity. Region-resolved analysis across the HIF1A transcript further reveals heterogeneous interaction probabilities ( Figure 8 B), with stronger predicted signals in the coding sequence (RF: 0.70; SVM: 0.99) and 3′UTR (RF: 0.65; SVM: 0.99), and weaker signals in the 5′UTR (SVM: 0.38). Based on these predictions, we present a schematic hypothesis illustrating preferential interaction within the coding sequence ( Figure 8 C). This is intended as a testable Model 3 framework rather than evidence of validated binding. Collectively, these findings support a potential non-random, transcript- and region-specific interaction profile between CTH and  HIF1A  mRNA, warranting further experimental validation. The converging lines of evidence supporting Models 1–3, together with additional structural and non-canonical functional support, are summarized in  Table 2 .\nTogether, these models provide mechanistic hypotheses for how CTH may integrate signaling and translational control pathways to regulate gene expression programs involved in tumor progression and adaptation.\nEvidence types are classified as follows: Direct experimental refers to functional studies demonstrating changes in HIF-1α protein or translation-related outputs following CTH perturbation. Experimental (indirect mechanistic) refers to pathway-level or interactome-based evidence suggesting mechanistic involvement without direct measurement of translational regulation. Computational refers to in silico analyses without experimental validation in this framework. Conceptual synthesis denotes integrative model-based interpretation derived from multiple evidence layers. “Hypothesis-generating” and “speculative” denote mechanistic interpretations not directly supported by functional or biochemical validation, and are included to reflect emerging or conceptual extensions of the model. “Limitations” indicate unresolved mechanistic gaps or lack of functional validation in CCOC. Abbreviations: CCOC, clear cell ovarian carcinoma; RBPs, RNA-binding proteins; IP-MS, immunoprecipitation coupled with mass spectrometry; RPIseq, RNA-protein interaction prediction using sequence information; CDS, coding sequence; UTR, untranslated region; IDRs, intrinsically disordered regions; WB, Western blotting.\n\nCollectively, the evidence presented here supports a conceptual shift in the understanding of cystathionine γ-lyase (CTH). While traditionally viewed as a metabolic enzyme involved in cysteine biosynthesis and hydrogen sulfide (H 2 S) production, CTH may also exhibit broader, context-dependent functions extending to the regulation of mRNA translation.\nThis emerging perspective is supported by multiple converging features, including structural flexibility, multi-compartment localization, interaction with translation-associated proteins and RBPs, and modulation of translational signaling pathways such as PI3K/AKT/mTOR. Collectively, these observations position CTH within regulatory networks that coordinate cellular responses to metabolic and environmental stress. However, the current evidence remains largely indirect, and whether CTH directly regulates mRNA translation remains unresolved. The hypothesis that CTH contributes to translational control—particularly of transcripts such as  HIF1A —should therefore be viewed as a testable framework rather than an established mechanism.\nFuture studies in CCOC cell lines, organoid systems, and other patient-derived models should therefore integrate loss- and gain-of-function genetic models of CTH (knockdown/knockout and overexpression systems) with approaches that directly interrogate the relationship between CTH and the translational machinery. Addressing these hypotheses will require systematic experimental validation using complementary approaches, including: RNA association assays, including CLIP-seq to assess whether CTH is associated with RNA, and RIP-qPCR to evaluate its presence in ribonucleoprotein complexes. Protein interactome profiling (e.g., immunoprecipitation coupled with mass spectrometry, IP-MS) to identify CTH-interacting partners, particularly components of RNA-binding and translational regulatory complexes. Polysome profiling to determine whether CTH alters global translational state and ribosome distribution across mRNA populations, with fraction-specific RNA analysis (qPCR or sequencing) to identify transcripts enriched in translating fractions. Ribosome profiling to evaluate whether CTH modulates genome-wide and transcript-specific translational efficiency at nucleotide resolution. Reporter assays using  HIF1A  5′ UTR/CDS/3′ UTR or full-length constructs in the presence of recombinant or cellular CTH modulation to determine whether CTH directly regulates  HIF1A  mRNA translation, particularly given prior evidence that CTH modulates HIF-1α abundance [ 7 ], and affects the expression of HIF-1α downstream targets such as VEGF [ 7 , 37 ]. Functional rescue experiments using catalytic-dead CTH mutants to distinguish enzymatic from non-enzymatic roles.\nRNA association assays, including CLIP-seq to assess whether CTH is associated with RNA, and RIP-qPCR to evaluate its presence in ribonucleoprotein complexes.\nProtein interactome profiling (e.g., immunoprecipitation coupled with mass spectrometry, IP-MS) to identify CTH-interacting partners, particularly components of RNA-binding and translational regulatory complexes.\nPolysome profiling to determine whether CTH alters global translational state and ribosome distribution across mRNA populations, with fraction-specific RNA analysis (qPCR or sequencing) to identify transcripts enriched in translating fractions.\nRibosome profiling to evaluate whether CTH modulates genome-wide and transcript-specific translational efficiency at nucleotide resolution.\nReporter assays using  HIF1A  5′ UTR/CDS/3′ UTR or full-length constructs in the presence of recombinant or cellular CTH modulation to determine whether CTH directly regulates  HIF1A  mRNA translation, particularly given prior evidence that CTH modulates HIF-1α abundance [ 7 ], and affects the expression of HIF-1α downstream targets such as VEGF [ 7 , 37 ].\nFunctional rescue experiments using catalytic-dead CTH mutants to distinguish enzymatic from non-enzymatic roles.\nThese approaches are intended to both validate and potentially falsify key components of the proposed model, enabling rigorous testing of the framework. They will also be essential to establish whether CTH functions as a direct or indirect regulator of mRNA translation and to define the molecular mechanisms involved. From a therapeutic perspective, these considerations suggest that current strategies focused solely on inhibiting CTH enzymatic activity, such as the use of competitive inhibitors like propargylglycine (PAG) [ 154 ], may not fully capture its functional roles, particularly if non-canonical interaction-mediated mechanisms are involved. In this context, our recent findings in CCOC models showed that both genetic and pharmacological targeting of CTH enhanced sensitivity to the mTOR inhibitor everolimus, with a noticeably more pronounced effect observed following genetic modulation of CTH [ 155 ]. Nevertheless, given the limited direct mechanistic evidence, such implications should be regarded as future possibilities rather than immediate therapeutic opportunities.\nAlthough this review focuses on CCOC as a biologically informative model system, the principles outlined here may extend to other cancer types and physiological contexts in which CTH is expressed and dynamically regulated. More broadly, this work highlights a growing paradigm in which metabolic enzymes contribute to gene regulation beyond their canonical biochemical functions, revealing additional layers of post-transcriptional control linking metabolic state to selective protein synthesis.\n\nWhile our findings and the available evidence support a potential role for CTH in post-transcriptional regulation, whether this extends to direct regulation of mRNA translation remains unresolved and requires experimental validation. Should such a role exist, its effects may not be restricted to translation initiation but could extend to other stages of the translational cycle. Translation is a multi-step, highly coordinated process [ 156 ] encompassing initiation, elongation, termination, and ribosome recycling. For instance, CTH activity may influence elongation rates or ribosome processivity under specific cellular contexts, including hypoxia and oxidative stress. Notably, emerging evidence links cellular redox status to ribosome pausing and codon-specific translation dynamics [ 157 , 158 ], raising the possibility that CTH-dependent sulfur metabolism could intersect with these processes.\nIn addition, CTH may influence mRNA partitioning into stress granules or processing bodies—dynamic non-membranous cytoplasmic assemblies that form under stress and function as transient storage sites for mRNAs [ 159 ]—thereby indirectly shaping translation efficiency. This possibility is supported by the close coupling of stress granule assembly to translational arrest and redox signaling [ 160 ]. These processes are particularly relevant in hypoxic and metabolic stress contexts, where both CTH expression and translational control pathways are dynamically regulated. Notably, exogenous H 2 S has been shown to promote stress granule formation and induce translational repression [ 161 ], suggesting that CTH may participate in these regulatory networks.\nFinally, we cannot exclude potential crosstalk with non-canonical translation pathways, including internal ribosome entry site (IRES)-mediated initiation. These mechanisms may become especially important when canonical translation initiation is altered or restricted. Together, these considerations highlight that although our data emphasize translation as a key node of CTH function, additional mechanisms may contribute to the observed phenotypes. Dissecting these possibilities will be important for developing a more comprehensive understanding of how CTH integrates metabolic and translational control.","source_license":"CC-BY-4.0","license_restricted":false}