Bioelectronic synthesis of hydrogen sulfide enables spatiotemporal regulation of protein modification and cellular redox

Abstract

Reactive signaling molecules such as hydrogen sulfide (H₂S) regulate protein function and cellular redox balance, yet their instability makes precise delivery in biological systems challenging. Existing bioelectronic platforms primarily target stable molecules and often lack the ability to control transient molecules with spatiotemporal precision. We develop a bioelectronic platform that uses electrochemical reactions to directly generate and deliver H₂S from biocompatible thiosulfate precursors near living cells. Through electrocatalyst screening, theoretical modeling, and product analysis, we demonstrate that biocompatible metal cathodes with low metal-hydrogen binding energy catalyze H₂S production while suppressing side reactions. Programmable electronic inputs, including electrolysis time and applied voltage, quantitatively control distance- and time-dependent H₂S release at the bioelectronic interface while maintaining physiological compatibility. This spatiotemporally modulated H₂S synthesis enables on-demand activation of ion channels through protein sulfhydration and restoration of intracellular redox balance under oxidative stress. Our platform broadens the functional scope of bioelectronics and establishes electrosynthesis as a modality for dynamic communication between electronics and biology.

Fig. 1.. A schematic of the bioelectronic platform for spatiotemporally controlled synthesis and delivery of H₂S.

(A) Overview of platform development, beginning with the design of an electrochemical reaction to synthesize H₂S from S₂O₃²⁻ ions at the cathode. Biocompatible metal cathodes were screened to identify catalysts for selective H₂S production, followed by mechanistic investigations of H₂S electrosynthesis. (B) Application of the bioelectronic platform for spatiotemporally controlled H₂S delivery through tunable electronic inputs. This system enabled on-demand modulation of protein function and restoration of cellular redox balance under both physiological and pathological conditions.

Fig. 2.. Biocompatible metal cathodes for the electrochemical reduction of S₂O₃²⁻ to H₂S.

(A) CV curves of various biocompatible metal cathodes in the presence and absence of 0.1 M S₂O₃²⁻ at pH 7.4. Scan rate: 20 mV/s. (B) A schematic illustrating the modified methylene blue method for quantifying electrochemically produced H₂S without interference from S₂O₃²⁻. Here, MB refers to methylene blue. (C) Calibration curve for H₂S in the presence of 0.1 M S₂O₃²⁻ obtained by the modified methylene blue method. H₂S-donor sodium sulfide (Na₂S) was used in these experiments, and absorbance at 670 nm was used for the analysis. (D) FE_H2S (mean ± SEM) for various biocompatible metal cathodes in the presence of 0.1 M S₂O₃²⁻ at pH 7.4 (n ≥ 3 independent experiments for each condition). FE_H2S was calculated from the number of moles of H₂S produced (N_H2S) and the total charge passed (Q) during electrolysis. The equation used for calculating FE_H2S is shown in the inset, where F is the Faraday constant, and the factor of two accounts for the two-electron process of H₂S electrosynthesis. No measurable H₂S was detected (*). (E) An illustration of the H₂ measurement setup comprised of a gas-tight electrochemical cell and H₂ sensor. (F) Output signal profiles of the H₂ sensor as a function of time, showing the response before, during, and after applying voltages to Ag and Mo cathodes in the presence of 0.1 M S₂O₃²⁻ at pH 7.4. (G) FE_H2 (mean ± SEM) for Ag and Mo cathodes, calculated from the H₂ sensor output profiles (n = 3 independent experiments for each condition). (H) Voltage- and electrolysis duration–dependent H₂S generation (mean ± SEM) using an Ag cathode in the presence of 0.1 M S₂O₃²⁻ at pH 7.4 (n ≥ 3 independent experiments for each condition).

Fig. 3.. Computational studies on the mechanism of electrochemical H₂S synthesis.

(A) GIXRD patterns of Ag and Mo cathodes before and after electrolysis at −1.5 V versus Ag/AgCl in the presence of 0.1 M S₂O₃²⁻ for 30 min, showing structural stability under reductive potentials. (B) Cross-sectional TEM image of the Ag cathode surface after electrochemical S₂O₃²⁻ reduction at −1.5 V versus Ag/AgCl for 30 min. Scale bar, 5 nm. (C and D) Schematic representations of the (111) and (110) crystallographic surfaces of (C) Ag and (D) Mo, respectively, used in the DFT simulation. (E and F) Gibbs free-energy diagrams for the *S to H₂S reduction pathway catalyzed by (E) Ag and (F) Mo cathodes at U = 0 (black) and their respective activation potentials [U = 0.44 V for Ag and 1.22 V for Mo (blue)]. (G) Calculated hydrogen binding energies on bare metal sites and *S sites for Ag and Mo surfaces, showing notable differences in hydrogen adsorption behavior. (H) Correlation between calculated metal-hydrogen binding energies and FE_H2S across various biocompatible metal cathodes analyzed in this study. Here, FE_H2S values were obtained from Fig. 2D.

Fig. 4.. TRPA1 activation through bioelectronic H₂S synthesis.

(A) Representative fluorescent images of HEK 293T cells co-expressing TRPA1 and GCaMP6s. Scale bar, 100 μm. (B) Averaged GCaMP6s fluorescent traces of TRPA1⁺ and TRPA1⁻ cells (n = 300 cells for each trace) following administration of 1 or 5 mM H₂S-donor Na₂S at 30 s (dashed lines). The solid lines represent the mean, and the shaded areas indicate the SEM. (C) Averaged GCaMP6s fluorescence traces of TRPA1⁺ cells (n = 300 cells for each trace) stimulated with 1 mM Na₂S at 30 s and subsequently treated with 0 or 1 mM DTT at 400 s (dashed lines). (D) A schematic of the experimental setup to monitor TRPA1 activation at the bioelectronic interface. (E) An illustration of the Ca²⁺ influx through TRPA1 in response to electrochemically generated H₂S. (F and G) Individual GCaMP6s fluorescence traces of (F) TRPA1⁺ and (G) TRPA1⁻ cells during electrochemical synthesis of H₂S at the Ag cathode. Voltages of −1.25 V versus Ag/AgCl were applied from 30 s (dashed lines) until the end of the recording (n = 300 cells for each condition). (H) Averaged GCaMP6s fluorescence traces of TRPA1⁺ cells (n = 300 cells for each trace) prestimulated with electrochemically produced H₂S at −1.25 V versus Ag/AgCl, followed by treatment with 0 or 5 mM DTT. The dashed line indicates the time of DTT administration. (I) Individual GCaMP6s fluorescence traces of TRPA1⁺ cells pre-incubated with 100 μM HC-030031 (TRPA1 antagonist) during electrochemical synthesis of H₂S at the Ag cathode. Voltages of −1.25 V versus Ag/AgCl were turned on at 30 s (dashed lines) and applied until the end of the recordings.

Fig. 5.. Spatiotemporally controlled TRPA1 activation with H₂S electrosynthesis.

(A) Normalized GCaMP6s fluorescence (ΔF/F₀) during electrochemical H₂S generation at the Ag cathode (n = 300 cells for each trace). Solid traces represent mean fluorescence, and shaded regions indicate SEM. Dashed lines represent the fitted slopes between two time points at which the normalized GCaMP6s fluorescence reached 20% and 80% of its maximum value used to calculate temporal regulation factors. (B) Representative images of TRPA1⁺ cells positioned in the vicinity of the Ag cathode after 800 and 1200 s at −1.25 and −1.5 V versus Ag/AgCl. Scale bars, 400 μm. (C and D) Scatter plots showing the activation time of individual TRPA1⁺ cells as a function of their distance from the imaging-field center during H₂S electrosynthesis at (C) −1.25 and (D) −1.5 V versus Ag/AgCl. Negative x-axis values indicate cells positioned closer to the Ag cathode. Dashed lines represent linear fits used to derive spatial regulation factors. (E and F) Representative time-lapse images of TRPA1⁺ cells showing spatially resolved TRPA1 activation mediated by electrosynthesized H₂S at (E) −1.25 and (F) −1.5 V versus Ag/AgCl. Scale bars, 400 μm. (G) Colormap showing ΔF/F₀ at various distances from the imaging center and times after the first detectable TRPA1⁺ cell activation at −1.25 or −1.5 V versus Ag/AgCl (n = 3 independent experiments for each voltage condition). (H and I) Quantitative analysis of (H) temporal and (I) spatial regulation factors (mean ± SEM) under −1.25 and −1.5 V versus Ag/AgCl. The representative procedures used to derive these factors are illustrated in (A) and (C) and (D), respectively. Statistical differences between the two voltage conditions were determined by one-tailed Student’s t tests (n = 3 independent experiments; **P = 0.0068 and *P = 0.0247).

Fig. 6.. Bioelectronic restoration of cellular redox balance via H₂S.

(A) A schematic illustrating H₂S-mediated recovery of GSH levels in HEK 293 T cells under oxidative stress. (B) Changes in intracellular ROS and GSH levels following rotenone treatment (0.1 μM, 24 hours), expressed as fold changes relative to 0 μM rotenone control. Statistical differences were determined by one-tailed Student’s t tests (ROS: n = 3 per group, *P = 0.0104; GSH: n = 5 per group, ****P < 0.0001). (C) Na₂S concentration–dependent recovery of GSH levels, expressed as fold changes relative to 0 μM rotenone control. Rotenone pretreated cells (0.1 μM, 24 hours) were incubated with Na₂S (1 to 10 mM, 6 hours), which restored GSH levels, as determined by one-way ANOVA tests (n = 5 per group; P values provided in table S3A). (D) A schematic of the experimental setup with a salt bridge to investigate GSH restoration upon electrochemical H₂S delivery to rotenone pretreated cells (0.1 μM, 24 hours). (E) A workflow of GSH quantification after electrochemical H₂S delivery. (F to I) GSH levels in rotenone pretreated cells at various positions (positions 1, 2, and 3) following electrochemical H₂S delivery. H₂S electrosynthesis was performed at either (F and G) −1.25 or (H and I) −1.5 V versus Ag/AgCl for 15 or 30 min, respectively. Cells without rotenone pretreatment and H₂S delivery, as well as cells with rotenone pretreatment but no H₂S delivery, were used as controls. One-way ANOVA tests were used to assess statistical differences between groups (n ≥ 3 independent experiments; P values provided in table S3, B to E). (G and I) Colormaps depicting GSH levels in rotenone pretreated cells after electrochemical H₂S delivery as a function of distance from the cathode, electrolysis time, and applied voltage. GSH levels in (F) to (I) are normalized to control groups without rotenone treatment and H₂S delivery. Data are presented as mean ± SEM. (*P < 0.05, **P < 0.01, and ****P < 0.0001).