Design, Validation, and Application of an Enzyme-Coupled Hydrogen Sulfide Detection Assay

Abstract

Hydrogen sulfide (H₂S) is a key metabolite in biosynthesis and is increasingly being recognized as an essential gasotransmitter. Owing to its diffusible and reactive nature, H₂S can be difficult to quantify, particularly in situ. Although several detection schemes are available, they have drawbacks. In efforts to quantify sulfide release in the cross-linking reaction of the flagellar protein FlgE, we developed an enzyme-coupled sulfide detection assay using the Escherichia coli O-acetylserine sulfhydrylase enzyme CysM. Conversion of HS^- to l-cysteine via CysM followed by derivatization with the thiol-specific fluorescent dye 7-diethylamino-3-(4-maleimidophenyl)-4-methylcoumarin enables for facile detection and quantification of H₂S by fluorescent HPLC. The assay was validated by comparison to the well-established methylene blue sulfide detection assay and the robustness demonstrated by interference assays in the presence of common thiols such as glutathione, 2-mercaptoethanol, dithiothreitol, and l-methionine, as well as a range of anions. We then applied the assay to the aforementioned lysinoalanine cross-linking by the Treponema denticola flagellar hook protein FlgE. Overall, unlike previously reported H₂S detection methods, the assay provides a biologically compatible platform to accurately and specifically measure hydrogen sulfide in situ, even when it is produced on long time scales.

Figure 1:. Overview of (A) CysM enzyme-coupled hydrogen sulfide detection assay and (B) CysM catalyzed L-cysteine production.

A) Step one involves incubation of the sulfide-containing samples with OAS and CysM to form L-cysteine and acetate. L-cysteine is then reduced with TCEP prior to derivatized with thiol-reactive dye DMM prior to HPLC separation and fluorescence detection. B) Abbreviated reaction cycle of L-cysteine production by CysM.^, (1. Internal Schiff Base [ISB], 2. Geminal diamine [GDA], 3. External Schiff Base [ESB], and 4. α-Aminoacrylate Intermediate [αAI]).

Figure 2:. Purification and characterization of E. coli CysM.

(A) Size exclusion chromatograph of overexpressed CysM reveals one major peak with an elution volume corresponding to a CysM homodimer. The final CysM stock was bright yellow in color, with a purity of > 95% and molecular weight of 33 kDa as estimated by SDS-PAGE. (B) Qualitative assessment of CysM activity; 1.0 mL of 60 μM CysM in 1X XLB pH 7.5 was stirred and absorbance measured in 5 second intervals at 412nm and 470nm. To test for sulfide specificity, 10 μL 10 mM OAS was added (indicated with *) and then 10 μL 10 mM reagent added (indicated with * and labeled above) once absorbance at 470nm reached a maximum.

Figure 3:. Resolution of L-cysteine and TCEP after DMM derivatization.

For each sample, 50 μM DMM in 50 mM Tris pH 7.5 was treated with excess L-cysteine (magenta) or TCEP (blue), reacted for 20 minutes and 10 μL injected onto a RP-HPLC equilibrated with 65:35 10 mM TCA pH 2.5:acetonitrile at a flow rate of 1.0 mL/min. Comparing both runs reveal two well resolved peaks corresponding to either the L-cysteine or TCEP derivatized DMM products.

Figure 4:. L-cysteine and Na₂S HPLC chromatograms and calibration curves.

A) Representative elution pattern of L-cysteine standards. B) Elution pattern of 0 – 25 μM Na2S standards incubated with 60 μM CysM and 1 mM OAS. Standards were run in triplicate and aligned to the largest peak marked with *. C) Comparison of L-cysteine and CysM-derived L-cysteine calibration curves plotting peak area versus sulfide/L-cysteine concentration. D) Low sulfide calibration curves (0–2 μM), Error bars reported as +/− the standard deviation of the three replicate samples. See Figure S2 for 0–2 μM chromatographs and unaligned chromatographs.

Figure 5:. Comparing CysM and MB sulfide detection assays.

For each set of standards, the total amount of sulfide present in each sample was calculated and plotted versus the normalized signal intensity. A) CysM (blue spheres), or B) MB (purple spheres) assay data points were fit to a line and the slopes calculated and compared. Standards were prepared in triplicate with +/− the standard deviation represented by error bars. The CysM standards were taken from Figure 4B and the MB assay standards reported in Figure S5.

Figure 6:. Measuring sulfide release by T. denticola FlgE lysinoalanine cross-linking.

A) Minimal mechanism of lysinoalanine (LA) cross-link formation, (cysteine-178 [C178], dehydroalanine [DHA] and lysinoalanine [LA]). B. H₂S calibration curve using 5 μM CysM and 1 mM OAS in 1X XLB pH 7.5. C) HPLC chromatograms of wild type (WT, red) and C178A (blue) full-length TdFlgE crosslinking samples where the inset shows the SDS-PAGE gel of WT TdFlgE with a quantification of the HMWC relative to the monomeric TdFlgE. Band intensity measurements were carried out using FIJT D) Peak area measurements of (C) result in an 8-fold increase in the peak area for WT compared to C178A TdFlgE. E) Buffer-subtracted H₂S quantities of WT TdFlgE (9.3 +/− 2.0 ng corresponding to a concentration of 5.6 +/− 1.1 μM) and C178A TdFlgE (2.2 +/− 1.8 ng corresponding to a concentration of 1.3 +/− 0.7 μM). For WT TdFlgE, 5.5 +/− 1.1 μM HS⁻ corresponds to approximately 18.6 +/− 3.7% of the total TdFlgE concentration, in agreement with the 21.8% of cross-linked species observed via SDS-PAGE. Statistical significance was confirmed via a two-tailed unpaired students t-test (** = p < 0.01, * = p < 0.05).

Figure 6:. Measuring sulfide release by T. denticola FlgE lysinoalanine cross-linking.

A) Minimal mechanism of lysinoalanine (LA) cross-link formation, (cysteine-178 [C178], dehydroalanine [DHA] and lysinoalanine [LA]). B. H₂S calibration curve using 5 μM CysM and 1 mM OAS in 1X XLB pH 7.5. C) HPLC chromatograms of wild type (WT, red) and C178A (blue) full-length TdFlgE crosslinking samples where the inset shows the SDS-PAGE gel of WT TdFlgE with a quantification of the HMWC relative to the monomeric TdFlgE. Band intensity measurements were carried out using FIJT D) Peak area measurements of (C) result in an 8-fold increase in the peak area for WT compared to C178A TdFlgE. E) Buffer-subtracted H₂S quantities of WT TdFlgE (9.3 +/− 2.0 ng corresponding to a concentration of 5.6 +/− 1.1 μM) and C178A TdFlgE (2.2 +/− 1.8 ng corresponding to a concentration of 1.3 +/− 0.7 μM). For WT TdFlgE, 5.5 +/− 1.1 μM HS⁻ corresponds to approximately 18.6 +/− 3.7% of the total TdFlgE concentration, in agreement with the 21.8% of cross-linked species observed via SDS-PAGE. Statistical significance was confirmed via a two-tailed unpaired students t-test (** = p < 0.01, * = p < 0.05).