Chemogenetic Manipulation of H₂S with Spatiotemporal Precision

Abstract

Hydrogen sulfide (H₂S) is a signaling molecule with a plethora of biological functions, yet precision tools for modulating its intracellular flux remain scarce. Conventional small-molecule donors and enzymatic systems often suffer from off-target reactivity, uncontrolled release kinetics, and redox crosstalk, confounding mechanistic studies. Here, we establish a Salmonella typhimurium d-cysteine desulfhydrase (stDCyD)-derived chemogenetic tool for controlled H₂S manipulation in living cells. stDCyD catalyzes the α,β-elimination of d-cysteine to selectively yield bioavailable H₂S. We term this tool H₂SWITCH. Our approach exhibits pronounced enantioselectivity for d-cysteine, robust catalytic efficiency at physiological temperatures, and temporal tunability through substrate dosing. This chemogenetic tool provides a chemically defined and interference-free method to unravel the physiological and pathological roles of H₂S with unprecedented precision in complex biological systems.

Hydrogen sulfide (H₂S) has emerged as a critical gasotransmitter alongside nitric oxide (NO) and carbon monoxide (CO), with growing recognition of its diverse roles in physiological and pathological processes. , Originally regarded as merely a toxic gas, H₂S is now known to be endogenously produced in mammalian cells by enzymes such as cystathionine β-synthase (CBS), cystathionine γ-lyase (CSE), and 3-mercaptopyruvate sulfurtransferase (MPST). − H₂S regulates key biological functions including vascular tone, mitochondrial bioenergetics , and neuromodulation. Its homeostatic imbalance has been implicated in a range of disorders. −

Traditional approaches to modulate intracellular H₂S levels largely rely on genetic manipulations, such as the overexpression or knockout of H₂S-producing enzymes, or pharmacological interventions, including the administration of sulfide-releasing salts like sodium hydrosulfide (NaHS) or slow-releasing H₂S donors. However, these methods suffer from major drawbacks, including poor spatiotemporal resolution, a lack of reversibility, and off-target effects that complicate the interpretation of results. Particularly in single-cell studies or in complex tissue environments, the inability to achieve precise, real-time control of H₂S concentrations limits the scope of mechanistic investigations. In recent years, several chemical probes, such as Reso-N3, resorufin-functionalized coumarin probe (RC), and resorufin-based probe RH for detection of H₂S and thiols have been developed. Besides, genetically encoded sensors for real-time detection of H₂S have been established (hsGFP and hsFRET biosensors). These advances have provided valuable insights into the spatial and temporal dynamics of H₂S signaling. However, detection alone is insufficient for functional studies; tools are needed that not only report on H₂S levels but also enable active manipulation with high precision.

In our previous work, we demonstrated that modified D-amino acid oxidase (mDAAO) can catalyze H₂S production in the presence of d-cysteine. In a recent study, we introduced a chemogenetic approach for manipulation of intracellular pH, named pH-Control, using the Salmonella typhimurium d-cysteine desulfhydrase (stDCyD) enzyme. In the presence of its substrate β-chloro-d-alanine (βCDA), stDCyD generates significant levels of hydrochloric acid (HCl) along with the byproducts, pyruvate, and ammonia in negligible amounts. In this study, we establish and characterize stDCyD as a chemogenetic tool that, in the presence of d-cysteine, can generate H₂S in living mammalian cells. By introducing a robust, targeted, and highly specific strategy to control intracellular H₂S concentrations, we provide a powerful tool to unravel the complex biology of this important gaseous signaling molecule.

As shown in Figure , stDCyD directly converts d-cysteine into H₂S, pyruvate and ammonium (NH₄ ⁺), utilizing pyridoxal 5′-phosphate (PLP) as a cofactor. In contrast, mDAAO first oxidizes d-cysteine into its corresponding α-keto acid, mercaptopyruvate, which is subsequently metabolized to H₂S by endogenous MPST. Both MPST and stDCyD produce pyruvate as a byproduct. Therefore, to indirectly compare the efficiencies of these two enzymatic pathways (as well as the endogenous pathway (MPST)), we used a genetically encoded pyruvate biosensor to monitor this metabolite’s production in real time (Figure S1). Our results showed that significant pyruvate generation occurred only in cells expressing stDCyD, but not in wild-type (WT) cells or those expressing mDAAO (Figure S1). These findings suggest that stDCyD exhibits superior catalytic efficiency for d-cysteine metabolism under the tested conditions.

1.

Mechanism of H₂S production by stDCyD and mDAAO. The stDCyD enzyme catalyzes the conversion of d-cysteine to H₂S, with the coenzyme pyridoxal 5′-phosphate (PLP) forming an aminoacrylate intermediate. This intermediate is subsequently converted to pyruvate and ammonium. In contrast, mDAAO oxidizes d-cysteine to 3-mercaptopyruvate using flavin adenine dinucleotide (FAD) as a cofactor, producing H₂O₂ as a byproduct. The resulting 3-mercaptopyruvate is then converted to H₂S by the endogenous enzyme 3-mercaptopyruvate sulfurtransferase, generating pyruvate as a byproduct.

Further characterization of the enzymatic activities focused on H₂O₂ generation. It is well established that mDAAO, while catalyzing the oxidation of D-amino acids, produces H₂O₂ as a byproduct. Live-cell H₂O₂ imaging revealed that, in mDAAO-expressing cells, 5 mM d-cysteine produced intracellular H₂O₂ levels comparable to those induced by 50 μM exogenous H₂O₂ (Figure S2). In contrast, cells expressing stDCyD showed no detectable H₂O₂ production under the same conditions (Figure S2). This distinction further highlighted the advantage of stDCyD for specific H₂S manipulation without the confounding effects of the oxidative stress induced by H₂O₂. Based on these observations, we selected stDCyD as the preferred chemogenetic tool for further characterization.

In a cell-free (“in chemico”) assay, we quantified H₂S production by purified stDCyD using RH probe. (See Supplementary Note 1 for details.) As shown in Figure B, purified stDCyD was incubated with d-cysteine and H₂S formation was monitored in real time via RH fluorescence. In control reactions lacking stDCyD, increasing d-cysteine produced only a slight fluorescence rise, consistent with the probe’s low background reactivity toward the thiol of d-cysteine (Figure C). In the presence of stDCyD, RH fluorescence increased in a d-cysteine concentration-dependent manner and exceeded no-enzyme controls, especially at higher d-cysteine concentration (Figure C). To assess substrate specificity, we compared RH responses to d-cysteine, l-cysteine, and d-methionine with or without stDCyD (Figure D). d-cysteine plus stDCyD elicited a strong signal, consistent with efficient H₂S formation. l-cysteine decreased fluorescence relative to the enzyme-free control, consistent with competitive binding to stDCyD without turnover and suppression of the residual probe background (Figure D). d-Methionine had no measurable effect on the RH response (Figure D). Fluorescence was higher at 37 °C than at 25 °C (Figure E), indicating greater catalytic activity at physiological temperature. In addition, we have investigated the functionality of stDCyD from acidic pH to physiological range (Figure S3) against one of its known substrates, βCDA, and showed that, at neutral pH, stDCyD shows a better catalytic activity (Figure S3). These results validate stDCyD as a selective, physiologically compatible, and dose-tunable chemogenetic approach for on-demand generation of H₂S from d-cysteine.

2.

Characterization of H₂SWITCH using the RH probe. A) Schematic representation of the RH probe and its activation by H₂S. B) Schematic illustration of the experimental protocol for H₂S measurement based on the enzymatic activity of stDCyD using the RH fluorescent probe. Desired concentrations of d-cysteine were mixed with purified stDCyD in PBS and incubated for 10 min at 37 °C. Subsequently, 2 μL of the reaction mixture was added to the RH probe solution, followed by PBS to reach the final assay volume for fluorescence measurements using a plate reader. C) Dose-dependent generation of H₂S by stDCyD in the presence of indicated concentrations of d-cysteine (orange curve). The blue curve represents the response of the RH probe to d-cysteine alone. D) RH fluorescence changes in response to 10 mM d-cysteine, 10 mM l-cysteine, and 10 mM d-methionine in the presence or absence of stDCyD. E) Quantification of H₂S generation by stDCyD at room temperature (25 °C) and physiological temperature (37 °C). Statistical analysis for C, D, E: n = 3 for all experiments, and two-tailed Student’s t test was applied to compare two independent groups; *** P < 0.001. d-Cys: d-cysteine. l-Cys: l-cysteine. d-Met: d-methionine.

To evaluate the applicability of H₂SWITCH in mammalian cells, we first assessed the potential toxicity of d-cysteine in wild-type HEK 293T cells and in cells stably expressing DsRed–stDCyD–NES (Figure S4). No cytotoxic effects were observed in either cell line at concentrations of up to 30 mM d-cysteine. We then proceeded to generate and directly visualize H₂S in cellulo. Wild type HEK 293T cells and cells expressing stDCyD were transiently transfected with the hsGFP sensor (Figure A). Administration of 10 mM d-cysteine increased H₂S in both WT and stDCyD-expressing cells (Figure B), but the maximal response was significantly greater in the latter (Figure C).In our previous study, we demonstrated that the DsRed-stDCyD construct can be selectively targeted to distinct subcellular compartments, where it retained full enzymatic functionality. To further verify compartment-specific H₂S generation, we employed a mitochondria targeted DsRed-stDCyD together with an untargeted hsGFP biosensor (Figure S5). Our data show that stDCyD efficiently metabolizes d-cysteine to produce H₂S and validate its utility as a robust tool for intracellular H₂S generation. The modest H₂S signal in WT cells upon d-cysteine treatment can be explained by (i) oxidation of d-cysteine to mercaptopyruvate by endogenous d-amino-acid oxidase followed by MPST-mediated H₂S formation, (ii) racemization of d-cysteine to l-cysteine (e.g., via serine racemase), enabling metabolism by canonical H₂S-producing enzymes (CBS, CSE), and (iii) the basal reactivity of the probe with d-cysteine. Regarding the latter, we showed that pAzF-treated HEK 293T cells respond to d-cysteine addition, evidenced by an increase in the GFP-channel fluorescence intensity (Figure S6A). Moreover, a highly significant increase in fluorescence intensity was observed when pAzF and d-cysteine were incubated together in a cell-free buffer (Figure S6B). Together, these data underscore the urgent need for more specific probes for H₂S visualization.

3.

H₂S generation in mammalian cells upon the enzymatic activity of stDCyD. A) Wide-field images show the change in the hsGFP fluorescence upon administration of d-cysteine in cells coexpressing hsGFP and stDCyD enzyme. Scale bar represents 10 μm. B) Panel shows representative real-time traces of produced H₂S in response to 10 mM d-cysteine in HEK 293T cells coexpressing untargeted hsGFP and stDCyD (orange curve, 3/41) or only hsGFP (blue curve, 3/54). C) The bar graph shows statistical analysis of the maximum response of hsGFP in response to 10 mM d-cysteine. Statistical analysis was performed using Student’s t test. *** refers to P < 0.001.

In conclusion, we present H₂SWITCH, a substrate-gated chemogenetic strategy that enables dose-tunable, enantioselective, and redox-orthogonal modulation of intracellular H₂S, which establish H₂SWITCH as a precise, interference-free tool for manipulating H₂S flux. Mechanistically, stDCyD offers several advantages: (i) orthogonalitythe d-cysteine/stDCyD pair minimizes engagement of native pathways; (ii) spatiotemporal control, substrate dosing affords rapid on-/off-like behavior without slow donor kinetics; (iii) compatibility with sensors, H₂S formation can be read out in real time with genetically encoded or chemical reporters, enabling closed-loop experiments. The genetically encodable nature of H₂SWITCH allows precise spatial and temporal control of H₂S production, a feature that is unattainable with small-molecule donors. In contrast to donor-based systems that rely on passive diffusion and trigger-dependent chemical release, , stDCyD affords on-demand manipulation of H₂S flux within specific cellular microdomains. Such modularity parallels our previous demonstration of the same enzyme scaffold’s functional robustness and targeting flexibility in a chemogenetic pH-control system, where compartment-specific localization did not compromise catalytic efficiency. While recent developments in stimuli-responsive H₂S donors have enhanced the temporal control of chemical donors, these molecules still exhibit intrinsic limitations. Specifically, they display poor cell-type selectivity, unpredictable and burst-type release kinetics, limited intracellular stability, and potential off-target oxidation or cytotoxic byproducts that obscure mechanistic readouts. , Conversely, the H₂SWITCH system provides a genetically precise, catalytically orthogonal, and imaging-compatible means to modulate H₂S signaling, enabling kinetic and spatial dissection of this gasotransmitter’s roles in mitochondria, cytosol, or other subcellular niches.

At the same time, important caveats warrant consideration. d-Cysteine can be racemized or enter endogenous routes , (e.g., DAAO→MPST), which likely explains the modest WT signals (Figure B); these effects can be bounded with catalytically dead stDCyD controls, pathway inhibitors and knockdowns (i.e., MPST or serine racemase K.O. cell lines). Second, pyruvate is a stoichiometric byproduct and may influence metabolism; concurrent pyruvate sensing, short pulses, or compartment targeting help deconvolute these effects. Third, effective use depends on intracellular PLP availability; substrate uptake through transporters and media composition should be validated per model.

Our H₂SWITCH approach opens experimental spaces that were previously hard to access. Compartment-resolved H₂S biology becomes tractable by directing stDCyD to mitochondria, ER, or endothelium-specific microdomains to test local persulfidation events, K_ATP channel regulation, or crosstalk with NO/CO signaling. Future studies emphasizing subcellular targeting and local H₂S generation (e.g., mitochondrial matrix, cytosol, ER) will elucidate the biological roles of H₂S microdomains in specific pathways.

Kinetic causalityhow fast and how much H₂S is required to trigger specific pathwayscan be quantified by pairing H₂SWITCH with real-time reporters (H₂S, redox, Ca²⁺, ATP) ,− and time-stamped perturbations.

Looking ahead, in vivo translation is feasible with AAV-mediated or transgenic delivery and cell-type–specific promoters or Cre lines, combined with localized d-cysteine administration (e.g., cerebral open fluid microperfusion (cOFM). Coupling H₂SWITCH with omics readouts and proteomic persulfidation mapping will enable systems-level causal links between the H₂S flux and phenotype.

In summary, H₂SWITCH provides a precision chemogenetic route to interrogate H₂S signaling with temporal, spatial, and mechanistic clarity. By minimizing redox crosstalk and enabling compartment-specific, dose-controlled generation, it sets the stage for definitive causal tests of H₂S function across vascular, cancer, and neurobiological contextsexperiments that have remained largely out of reach with existing donor or genetic strategies.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/prechem.5c00189.

• Detailed materials and methods and supplementary Note 1; Figures S1–S13 of cellulo experiments for the live-cell imaging of pyruvate and hydrogen peroxide dynamics, as well as in chemico characterization of the RH probe (PDF)

A.G.Z. and E.E. conceived the research idea and designed the experiments. A.G.Z., R.Z.M., and O.A. performed the in cellulo experiments. H.I. and S.M.M. carried out the in chemico experiments. Ş.Ç. and A.A. designed and performed Seahorse experiments. T.A. synthesized and provided the RH probe. A.G.Z., H.I., S.M.M., J.A., A.A., S.Y., E.V., T.A.C., E.N.Y., M.Ş.A., M.I.A., T.A., S.V., and E.E. contributed intellectually to data analysis, discussion, and manuscript preparation.

EMBO Installation Grant (EMBO IG J4113) to E.E.

The authors declare the following competing financial interest(s): A.G.Z., M.M., and E.E. have filed a patent application (patent application number 2023/000206) describing parts of this manuscript's research, which does not alter the authors' adherence to the policies on sharing data and materials presented in this study. The remaining authors declare no competing financial interests.