Introduction

Superficial fungal infections are very common, with a global prevalence of 20–25%[1](https://www.nature.com/articles/s41598-025-22062-7#ref-CR1 “Havlickova, B., Czaika, V. A. & Friedrich, M. Epidemiological trends in skin mycoses worldwide. Mycoses 51(Suppl 4), 2–15. https://doi.org/10.1111/j.1439-0507.2008.01606.x

(2008).“). This disease, dermatophytosis, can be uncomfortable and painful, reducing the quality of life[2](https://www.nature.com/articles/s41598-025-22062-7#ref-CR2 “Murdan, S. Drug delivery to the nail following topical application. Int. J. Pharm. 236, 1–26. https://doi.org/10.1016/s0378-5173(01)00989-9

(2002).“). In high-risk groups such as the elderly and diabetics, the consequences can be severe. For example, diabetics are 2.7 fold more likely to develop fungal nail infections (onychomycosis) than non-diabetics[3](https://www.nature.com/articles/s41598-025-22062-7#ref-CR3 “Gupta, A. K., Daigle, D. & Foley, K. A. The prevalence of culture-confirmed toenail onychomycosis in at-risk patient populations. J. Eur. Acad. Dermatol. Venereol. 29, 1039–1044. https://doi.org/10.1111/jdv.12873

(2015).“). In these patients, fungal infections may precede foot ulcers, potentially leading to lower-limb amputation[4](https://www.nature.com/articles/s41598-025-22062-7#ref-CR4 “Boyko, E. J., Ahroni, J. H., Cohen, V., Nelson, K. M. & Heagerty, P. J. Prediction of diabetic foot ulcer occurrence using commonly available clinical information: The seattle diabetic foot study. Diabetes Care 29, 1202–1207. https://doi.org/10.2337/dc05-2031

(2006).“). With rising obesity and type 2 diabetes, more serious cases of dermatophytosis are expected to increase.

The most common cause of dermatophytosis is Trichophyton rubrum, though other dermatophytes (Trichophyton, Microsporum and Epidermaphyton) and non-dermatophytes like Candida, Aspergillus, Fusarium, and (Neo)scytalidium spp., can also be responsible[5](https://www.nature.com/articles/s41598-025-22062-7#ref-CR5 “Moreno, G. & Arenas, R. Other fungi causing onychomycosis. Clin. Dermatol. 28, 160–163. https://doi.org/10.1016/j.clindermatol.2009.12.009

(2010).“),[6](https://www.nature.com/articles/s41598-025-22062-7#ref-CR6 “Machouart, M., Menir, P., Helenon, R., Quist, D. & Desbois, N. Scytalidium and scytalidiosis: What’s new in 2012?. J. Mycol. Med. 23, 40–46. https://doi.org/10.1016/j.mycmed.2013.01.002

(2013).“). Finally, bacteria such as Pseudomonas aeruginosa or Staphylococcus aureus can also infect nails.

Treatment options are limited, mainly involving azoles or allyamines. First-line oral treatments such as itraconazole and terbinafine are reasonably effective, but this takes several weeks or months and may cause side effects and drug interactions. Topical treatments, which are suitable for mild or moderate cases, offer more variety of drugs[7](https://www.nature.com/articles/s41598-025-22062-7#ref-CR7 “Gupta, A. K., Foley, K. A. & Versteeg, S. G. New antifungal agents and new formulations against dermatophytes. Mycopathologia 182, 127–141. https://doi.org/10.1007/s11046-016-0045-0

(2017).“) and are often preferred by patients. However, they act very slowly and have low cure rates. Nail infections are particularly challenging, often requiring treatment of a year or longer for toenails, with high rates of relapse8. Even one of the most effective topical agents, efinaconazole, achieves only a complete cure rate (resolution of clinical signs and symptoms) of 15–18%, and a mycological cure rate (negative for culture or microscopy) of 53–55%[9](https://www.nature.com/articles/s41598-025-22062-7#ref-CR9 “Elewski, B. E. et al. Efinaconazole 10% solution in the treatment of toenail onychomycosis: Two phase III multicenter, randomized, double-blind studies. J. Am. Acad. Dermatol. 68, 600–608. https://doi.org/10.1016/j.jaad.2012.10.013

(2013).“).

A key issue is slow penetration of antifungals into the nail plate[10](https://www.nature.com/articles/s41598-025-22062-7#ref-CR10 “Kubota-Ishida, N. et al. Human onychopharmacokinetic and pharmacodynamic analyses of ME1111, a new topical agent for onychomycosis. Antimicrob. Agents Chemother. 62, 10. https://doi.org/10.1128/AAC.00779-17

(2018).“), influenced by factors such as molecular weight and polarity[10](#ref-CR10 “Kubota-Ishida, N. et al. Human onychopharmacokinetic and pharmacodynamic analyses of ME1111, a new topical agent for onychomycosis. Antimicrob. Agents Chemother. 62, 10. https://doi.org/10.1128/AAC.00779-17

(2018).“),[11](#ref-CR11 “Hui, X. et al. In vitro penetration of a novel oxaborole antifungal (AN2690) into the human nail plate. J. Pharm. Sci. 96, 2622–2631. https://doi.org/10.1002/jps.20901

(2007).“),[12](https://www.nature.com/articles/s41598-025-22062-7#ref-CR12 “McAuley, W. et al. An investigation of how fungal infection influences drug penetration through onychomycosis patient’s nail plates. Eur. J. Pharm. Biopharm. 102, 178–184. https://doi.org/10.1016/j.ejpb.2016.03.008

(2016).“). An “ideal” compound for the treatment should be small, polar, and have good antimicrobial activity. Recently, we identified H2S as a promising candidate due to its good nail penetration[13](https://www.nature.com/articles/s41598-025-22062-7#ref-CR13 “Malallah, O. et al. Systematic review and QSPR analysis of chemical penetration through the nail to inform onychomycosis candidate selection. Drug Discov. Today 29, 103844. https://doi.org/10.1016/j.drudis.2023.103844

(2024).“) and known antimicrobial properties[14](https://www.nature.com/articles/s41598-025-22062-7#ref-CR14 “Fu, L. H. et al. An antifungal role of hydrogen sulfide on the postharvest pathogens Aspergillus niger and Penicillium italicum. PLoS ONE 9, e104206. https://doi.org/10.1371/journal.pone.0104206

(2014).“),[15](https://www.nature.com/articles/s41598-025-22062-7#ref-CR15 “Fu, L. H. et al. Hydrogen sulfide inhibits the growth of Escherichia coli through oxidative damage. J. Microbiol. 56, 238–245. https://doi.org/10.1007/s12275-018-7537-1

(2018).“). Here we aimed to investigate its activity against microbes causing nail infections and to understand the mechanism of action of this gaseous molecule.

Materials and methods

Release of gaseous H2S from NaHS

The release of H2S gas from 20 mL solutions of 0.1 mM, 1 mM or 10 mM sodium hydrogen sulphide (NaHS) in water was measured in airtight boxes (2.1L; FoodSaver® Quick Marinator). At designated time points, a T101 H2S gas analyser (Teledyne) was attached to the box, and the entire box volume was extracted in 4 min at a rate of 650 mL/min for measurement. A separate box was used for each time point/concentration combination. The analyser had a limit of detection (LOD) of 0.4 ppb using a 10 V analogue output range and was calibrated using standard H2S gas dilutions provided by the manufacturer.

Strains and culture conditions

A list of strains is presented in Table 1. Where indicated, isolates were obtained from the American Type Culture Collection (ATCC), the National Collection of Type Cultures (NCTC), the National Collection of Pathogenic Fungi (NCPF), and the Belgian Coordinated Collection of Microorganisms (BCCM). Filamentous fungi were maintained on potato dextrose agar plates that were incubated at 30 °C. Conidia were isolated as described[16](https://www.nature.com/articles/s41598-025-22062-7#ref-CR16 “Ho, F. K., Delgado-Charro, M. B. & Bolhuis, A. Evaluation of an explanted porcine skin model to investigate infection with the dermatophyte Trichophyton rubrum. Mycopathologia 185, 233–243. https://doi.org/10.1007/s11046-020-00438-9

(2020).“). Candida albicans and bacteria were routinely cultured on Brain Heart Infusion (BHI) medium at 37 °C.

Antimicrobial activity of H2S

The antifungal activity of the H2S donor NaHS was tested in liquid using RPMI-1640 medium with glucose and MOPS (RPMI-GM)[22](https://www.nature.com/articles/s41598-025-22062-7#ref-CR22 “Rodriguez-Tudela, J. L. & Martinez-Suarez, J. V. Defining conditions for microbroth antifungal susceptibility tests: Influence of RPMI and RPMI-2% glucose on the selection of endpoint criteria. J. Antimicrob. Chemother. 35, 739–749. https://doi.org/10.1093/jac/35.6.739

(1995).“). Where indicated, the pH was adjusted to pH 5, 7, or 8. The medium was inoculated with ~ 105 conidia/mL of T. rubrum ATCC 28,188 and incubated at 30 °C. To limit outgassing of H2S, the headspace was kept to ~ 20% of the total volume. After 4 days, cultures were inspected and the minimal inhibitory concentration in aqueous conditions (MICaq) was defined as the lowest concentration with no visible growth.

To test the activity of gaseous H2S, conidia of fungi (2 µl of a suspension containing ~ 500 conidia) were inoculated on Sabouraud dextrose agar (SDA) plates, and then the plates (triple vented) were placed in a 2.1 L airtight box containing an open dish with 20 mL of freshly prepared NaHS in water (Fig. 1). Boxes were incubated for 4 days at 30 °C, after which fungal growth was assessed. The MICg_max was defined as the lowest concentration (assuming that 100% of sulphide is released as gaseous H2S) with no visible fungal growth. Plates were then incubated for an additional 3 days at 30 °C without H2S: the lowest concentration showing no growth was defined as the MFCg_max.

Fig. 1

Release of gaseous H2S. (A) Experimental setup for exposure of microbes to gaseous H2S. The gas is released from the donor NaHS, dissolved in 20 mL H2O, that is present in an open petri dish that was placed inside a 2.1 L airtight box. (B) and (C) show the concentration and percentage, respectively, of sulphide released at specific time points in the box. All experiments were repeated independently three times, and the error bars represent the standard deviation.

To assess the activity of gaseous H2S against bacteria and C. albicans, overnight cultures were serially diluted in phosphate-buffered saline (PBS), and 10 µL spots were deposited on BHI agar. These plates were placed in boxes with NaHS solution, and incubated at 37 °C. After 24 h, colony-forming units (CFU) were counted. The MIC was defined as the lowest concentration of H2S that fully inhibited growth. Plates were incubated for another 24 h in the absence of H2S to determine the MFC or minimal bactericidal concentration (MBC).

Confocal laser scanning microscopy (CLSM)

To assess viability of T. rubrum ATCC 28,188 when exposed to H2S, conidia (2.5 × 105 CFU/mL) were grown in RPMI-GM (pH 5) for 24 h, in the presence or absence of a sublethal concentration of NaHS. This was followed by staining with the BacLight stain (Invitrogen) according to manufacturer’s instructions. Images were captured using a Zeiss LSM880 laser scanning confocal microscope equipped with Airyscan technology and Zen 3.5 software.

Oxidative stress induction in T. rubrum by NaHS exposure was assessed using 2’,7’-dichlorodihydrofluorescein diacetate (DCHF-DA) staining, following a protocol described previously[23](https://www.nature.com/articles/s41598-025-22062-7#ref-CR23 “Kim, H. & Xue, X. Detection of total reactive oxygen species in adherent cells by 2’,7’-dichlorodihydrofluorescein diacetate staining. J. Vis. Exp. https://doi.org/10.3791/60682

(2020).“). T. rubrum conidia (2.5 × 105 CFU/mL) were cultured at 30 °C for 12 h in Sabouraud dextrose broth (SDB) to induce germination. Next, the germlings were resuspended in RPMI-GM (pH 5) and treated for 2 h with NaHS. Following treatment, the samples were centrifuged at 11,600 g and washed once with RPMI-GM to remove NaHS, resuspended in RPMI-GM containing 10 µM DCHF-DA and 25 µM Calcofluor white, and incubated for 30 min at 30 °C in the dark. After incubation, the samples were washed once with medium and twice with PBS. Images were captured using a Zeiss LSM880 laser scanning confocal microscope as described above.

Flow cytometry

Fungal germination of T. rubrum under exposure to NaHS was examined using flow cytometry. A suspension of 2.5 × 105 CFU/mL of conidia was grown in RPMI-GM (pH 5) with NaHS for 0, 6, 12 and 18 h. Formaldehyde (final concentration at 10%) was added, followed by staining with 20 μg/mL calcofluor white. The cells were left at 4 °C overnight and washed twice with PBS. Blue fluorescence emitted by 20,000 cells was quantified using a BD FACSAria™ III (Becton Dickinson, San Jose, CA, USA) with an 85 μm nozzle tip. Data were collected using BD FACSDiva software and analysed with FCS express.

Cytochrome C oxidase (COX) assay

T. rubrum germlings were prepared by culturing conidia in SDB in a shaking incubator at 30 °C for 19 h, and protoplasts were generated as described[24](https://www.nature.com/articles/s41598-025-22062-7#ref-CR24 “Kaufman, G., Horwitz, B. A., Hadar, R., Ullmann, Y. & Berdicevsky, I. Green fluorescent protein (GFP) as a vital marker for pathogenic development of the dermatophyte Trichophyton mentagrophytes. Microbiology (Reading) 150, 2785–2790. https://doi.org/10.1099/mic.0.27094-0

(2004).“). Protoplasts were separated from cell debris by 1 min of vortexing, filtration through a 40 µm strainer, and centrifugation at 700 g for 10 min at 4 °C. The supernatant was carefully collected and centrifuged at 3000 g for 15 min at 4 °C. The pellet was washed with 0.5 mL PBS and centrifuged at 10,000 g for 30 min at 4 °C, and resuspended in 60 µL of PBS. COX activity, with or without NaHS, was measured using a cytochrome C oxidase activity kit (Thermo Fisher). Statistical significance was assessed using a One-way ANOVA with a Dunnett’s post-hoc test using GraphPad Prism 10.4, with p < 0.05 considered significant.

Thiol modification

5’-IAF labelling was used to assess the ability of H2S to interact with thiol groups on proteins, using a previously published method[25](https://www.nature.com/articles/s41598-025-22062-7#ref-CR25 “Williams, E. et al. Detection of thiol modifications by hydrogen sulfide. Methods Enzymol. 555, 233–251. https://doi.org/10.1016/bs.mie.2014.11.026

(2015).“). T. rubrum conidia were treated with NaHS in RPMI-GM (pH 5.0) for 3 h, washed with PBS, and ground in liquid nitrogen. 2 mL of lysis buffer was added, the samples were vortexed for 10 min at 4 °C, kept on ice for 30 min, and centrifuged at 13,300 g for 20 min at 4 °C. For fluorescent labelling, 2 µL of 20 mM 5’-IAF was added to a 200 µL sample for 10 min, at room temperature in the dark. Proteins were precipitated by adding 1 mL ice-cold acetone and incubation at −20 °C for 10 min. After centrifugation (13,300 g) for 20 min, pellets were air-dried in the dark for 10 min, resuspended in 20 µL of SDS sample loading buffer, and mixed with 2 µL of 10 × DTT. Samples were denatured at 70 °C for 10 min and separated with SDS-PAGE. 5-IAF-labelled proteins were visualised using a UV transilluminator, using a Biorad GelDoc XR + imaging system, followed by Coomassie Brilliant Blue staining to show total protein. Total protein in each lane of the SDS-PAGE gels was quantified using ImageJ/Fiji software (version 2.14.0/1.54f.).

Transcriptome analysis

Culture preparation and RNA extraction were performed using a previously described method[26](https://www.nature.com/articles/s41598-025-22062-7#ref-CR26 “Mendes, N. S. et al. Transcriptome-wide survey of gene expression changes and alternative splicing in Trichophyton rubrum in response to undecanoic acid. Sci. Rep. 8, 2520. https://doi.org/10.1038/s41598-018-20738-x

(2018).“). T. rubrum ACTC 28,188 (1 × 106 conidia/mL) was inoculated in 100 mL RPMI-GM (pH 5.0) and incubated at 30 °C for 96 h with agitation. Mycelium was harvested by filtration and transferred to RPMI-GM (pH 5.0) with or without 0.1 mM NaHS, followed by incubation at 30 °C for 3 h. Samples were stored at –80 °C until RNA extraction. Total RNA was extracted from ~ 100 mg frozen mycelia using the Illustra RNAspin Mini RNA Isolation Kit (GE Healthcare) per the manufacturer’s instructions. RNA quantity and integrity were assessed using Qubit RNA BR and IQ assay kits with a Qubit 4 Fluorometer (Thermo Fisher).

Paired-end 100 bp sequencing was performed by BGI Tech Solutions on a DBN sequencer following the construction of an rRNA depletion library. For each condition (with/without 0.1 mM NaHS), 4 biological replicates were sequenced. FastQ sequencing files were quality checked with FastQC, yielding scores of > 36[27](https://www.nature.com/articles/s41598-025-22062-7#ref-CR27 “Andrews, S. C. FastQC: A quality control tool for high throughput sequence data. http://www.bioinformatics.babraham.ac.uk/projects/fastqc/

(2010).“). Reads were splice-aligned to the T. rubrum CBS118892 reference genome (ASM151142v1 in NCBI, GCF_000955945.1 Refseq) using Hisat2[28](https://www.nature.com/articles/s41598-025-22062-7#ref-CR28 “Pertea, M., Kim, D., Pertea, G. M., Leek, J. T. & Salzberg, S. L. Transcript-level expression analysis of RNA-seq experiments with HISAT, StringTie and ballgown. Nat. Protoc. 11, 1650–1667. https://doi.org/10.1038/nprot.2016.095

(2016).“), with a mean alignment rate of 54.4%. Differential expression analysis was analysed using DESeq2[29](https://www.nature.com/articles/s41598-025-22062-7#ref-CR29 “Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 550. https://doi.org/10.1186/s13059-014-0550-8

(2014).“), using the NCBI ASM151142v1 reference database. P values were adjusted for transcriptome-wide false discovery rate (FDR), with an adjusted significance threshold of q < 0.05. Genes were classified based on Uniprot (www.uniprot.org) annotation, and GO enrichment analysis was performed using FungiDB[30](https://www.nature.com/articles/s41598-025-22062-7#ref-CR30 “Basenko, E. Y. et al. FungiDB: An integrated bioinformatic resource for fungi and oomycetes. J. Fungi (Basel) 4, 39. https://doi.org/10.3390/jof4010039

(2018).“), using a Benjamini-adjusted p value of < 0.05 as significant). The RNAseq data are available in NCBI BioProject PRJNA1242198.

Results

Release of H2S gas from the donor NaHS

The setup of exposure of microbes to H2S is shown in Fig. 1A, with microbes grown on agar plates placed inside an airtight 2.1 L box, with 20 mL NaHS solution in an open dish. NaHS dissociates in water into Na+ and HS−, the latter converting to gaseous H2S. To quantify H2S release, boxes were connected to a gas analyser, and the release of H2S was measured after 1, 3, 6, and 24 h, using a separate box for each measurement. The experiments showed that at all concentrations of the donor, H2S release peaked within 3 h (Fig. 1B). At 0.1 or 1 mM NaHS, 16–18% of the total available sulphide was released as H2S, compared to only 2% with 10 mM NaHS (Fig. 1C). H2S levels declined at 6 and 24 h, probably due to adsorption by the box’s plastic[31](https://www.nature.com/articles/s41598-025-22062-7#ref-CR31 “Nielsen, A. H., Vollertsen, J., Jensen, H. S., Wium-Andersen, T. & Hvitved-Jacobsen, T. Influence of pipe material and surfaces on sulfide related odor and corrosion in sewers. Water Res. 42, 4206–4214. https://doi.org/10.1016/j.watres.2008.07.013

(2008).“).

Antimicrobial activity of H2S

To assess the antimicrobial activity of H2S, we tested aqueous sulphide against T. rubrum using a standard macrobroth dilution method in which NaHS was dissolved in RPMI-GM. In water, the sulphide exists as H2S, HS− or S2−, with their proportions pH-dependent (pKa1 = 7.04, pKa2 = 11.96)32. The activity of NaHS was determined at pH 5, 7, and 8. While not affecting the growth of T. rubrum significantly, they altered the ratio of [H2S]:[HS−] to ~ 100:1, 1:1, and 1:10, respectively, with negligible amounts of S2−. Notably, T. rubrum was more sensitive at pH 5, showing a 20-fold lower MICaq than at pH 7 or 8 (Table 2). Considering the H2S:HS− ratios, this suggests that H2S has higher antifungal activity than HS−. To note, concentrations of antimicrobials are traditionally expressed in mass per volume, and MIC values of sulphide (i.e. H2S plus HS−) are therefore also expressed as equivalent concentrations of H2S in µg/mL (Table 2).

To measure gaseous H2S activity, SDA plates were inoculated with T. rubrum conidia. Similar to the macrobroth tests, SDA plates were also adjusted to pH 5, 7, or 8. Plates inside a box were incubated for 4 days. Supplementary Fig. S1 (top row) shows an example, with the lowest concentration preventing growth defined as the gaseous MIC. Exact H2S levels are hard to work out, as they depend on donor concentration and exposure time. Therefore, MIC values are expressed as the MICg_max: the theoretical MIC if all available sulphide is released as H2S. Independent of the pH of the agar plates, the MICg_max was reached with 0.25–0.5 mM NaHS (in 20 mL in the 2.1 L box) that, if all sulphide is released as gas, corresponds to 2.4–4.8 µM (0.082–0.16 µg/mL) of H2S (Table 2). Interestingly, H2Sg was far more potent than H2Saq: at pH 5, the MICg_max was 50-fold lower than the MICaq, while at pH 7 or 8 this was 1000–2000 fold lower. It should be noted that the MICg_max was twofold lower at pH 8 when compared to pH 5 or 7. However, this represents a difference of one dilution step, which is generally considered not significant in MIC testing.

To evaluate the spectrum of activity of H2S, several fungal isolates causing superficial infections were tested. These included T. rubrum isolates with different morphotypes33 (e.g. granular NCPF 0420 and the others with a downy morphology), a terbinafine-resistant isolate of T. indotinea, and various non-dermatophytes. The minimal fungicidal concentration (MFCg_max) was also determined by incubating SDA plates that were exposed for 4 days for an additional 3 days in the absence of H2S. Supplementary Fig. S1 shows an example, and results are summarised in Table 3. From this, dermatophytes (Trichophyton and Microsporum spp.) appeared more sensitive to H2S than most non-dermatophytes, with MICg_max values of 0.082–0.16 µg/mL vs. 0.32–1.6 µg/mL, respectively. The MFCg_max values were generally the same or double the MICg_max, indicating that H2S is fungicidal.

Next, activity of gaseous H2S was tested on bacteria that can cause nail infections[34](https://www.nature.com/articles/s41598-025-22062-7#ref-CR34 “Iorizzo, M. & Pasch, M. C. Bacterial and viral infections of the nail unit: Tips for diagnosis and management. Hand Surg. Rehabil. https://doi.org/10.1016/j.hansur.2022.11.006

(2022).“). The data (Table 4) illustrated that two S. aureus strains, including multidrug-resistant MRSA252, were sensitive to H2S, while P. aeruginosa was not inhibited at the highest concentration tested. To test resistance in another Gram-negative species, two laboratory strains of E. coli were tested. E. coli BW25113 was highly resistant (MICg_max value > 8.2 µg/mL), but the recA- strain DH5α showed partial sensitivity with an MICg_max of 3.2 µg/mL.

To determine the minimal H2S exposure time, SDA agar plates were inoculated with T. rubrum and incubated with H2S for 1, 3, 6, or 24 h in an airtight box. Plates were then incubated for a further 7 days without H2S. This showed that between 3 and 6 h of exposure is sufficient reach the MICg_max value (Fig. 2).

Fig. 2

Effect of H2S exposure time and concentration against T. rubrum. Agar plates were inoculated with 3 spots of T. rubrum ATCC 28,188, incubated in airtight boxes for 1, 3, 6, and 24 h, in the presence of different concentrations of H2S, followed by a further incubation without H2S for a further 7 days. The concentrations H2S indicated are those which would be achieved if all available sulphide is released as gaseous H2S.

Germination of T. rubrum spores takes several hours, with conidial swelling after 3–4 h and germ tube formation by 9–10 h[16](https://www.nature.com/articles/s41598-025-22062-7#ref-CR16 “Ho, F. K., Delgado-Charro, M. B. & Bolhuis, A. Evaluation of an explanted porcine skin model to investigate infection with the dermatophyte Trichophyton rubrum. Mycopathologia 185, 233–243. https://doi.org/10.1007/s11046-020-00438-9

(2020).“),[35](https://www.nature.com/articles/s41598-025-22062-7#ref-CR35 “Liu, T. et al. The use of global transcriptional analysis to reveal the biological and cellular events involved in distinct development phases of Trichophyton rubrum conidial germination. BMC Genomics 8, 100. https://doi.org/10.1186/1471-2164-8-100

(2007).“). To test post-germination activity, conidia were pre-incubated to allow germination before H2S exposure. The MICg_max of H2S increased after 9–12 h preincubation, suggesting that once germ tubes form, T. rubrum becomes more tolerant to H2S: the MICg_max increased from 0.082 to 0.32 µg/mL (Fig. 3A) and the MFCg_max increased from 0.082 to 1.6 µg/mL (Fig. 3B).

Fig. 3

The effect of pre-incubation of T. rubrum conidia on the MIC of gaseous H2S. T. rubrum ATCC 28,188 conidia were deposited on SDA plates and then incubated at 30 °C for 0–12 h, followed by incubation for 4 days in an airtight box containing different concentrations of H2S, to determine the MICg_max (A). The plates were removed from the box and incubated for a further 3 days to determine the MFCg_max (B).

To visualise the effects of H2S during germination, T. rubrum conidia were exposed to H2S in RPMI-GM, stained with Live/Dead Baclight, and examined by confocal microscopy. As shown in Fig. 4A, a sublethal concentration of H2S (40% of the MICaq) inhibited mycelium formation (compare top and bottom panels) and resulted conidial clumping, with some conidia staining red, indicating membrane damage. Flow cytometry showed similar results using the fluorophore calcofluor white, which binds fungal cell walls. Without H2S, fluorescence increased over time, indicating growth, but with sublethal (3.4 µg/mL) or lethal (8.5 µg/mL) concentrations of H2S, fluorescence remained low (compare light blue to teal and dark blue), indicating no conidial outgrowth (Fig. 4B).

Fig. 4

H2S inhibits the germination of T. rubrum conidia. (A) T. rubrum ATCC 28,188 without (top panel) or exposed to NaHS in RPMI-GM medium (pH 5) for 24 h, at a sulphide concentration that is equivalent to 3.4 µg/mL of H2S. This was followed by staining with the LIVE/DEAD BacLight stain and fluorescence microscopy. (B) Flow cytometry analysis of T. rubrum ATCC 28,188 conidia exposed to different concentrations of H2S in RPMI-GM (pH 5) for 0–18 h. Conidia were identified and delimited on the scatterplot (granularity [SSC-A] vs. blue fluorescence [Calcofluor white; DAPI-A filter]) at the top right corner. Histogram plots show only events from the conidia gate in the scatterplot.

Mechanisms of action of H2S

Previous studies reported that H2S increases reactive oxygen species (ROS) in fungal plant pathogens[14](https://www.nature.com/articles/s41598-025-22062-7#ref-CR14 “Fu, L. H. et al. An antifungal role of hydrogen sulfide on the postharvest pathogens Aspergillus niger and Penicillium italicum. PLoS ONE 9, e104206. https://doi.org/10.1371/journal.pone.0104206

(2014).“). Using the redox-sensitive probe DCHF-DA, we investigated ROS levels in T. rubrum after 12 h pre-incubation followed by 2 h of treatment with 3.4 µg/mL H2S (= 40% MICaq). ROS levels increased compared with the control (Fig. 5A). To confirm this, MICaq values were determined in the presence of antioxidants. Adding N-acetyl cysteine (NAC) or glutathione indeed significantly increased the tolerance to H2S, increasing the MICaq eightfold (Fig. 5B).

Fig. 5

Mechanism of action of H2S. (A) CLSM images of 12 h pre-incubated T. rubrum ATCC 28,188 exposed to NaHS in RPMI-GM medium (pH 5) for 2 h, at a sulphide concentration that is equivalent to 3.4 µg/mL of H2S. The samples were stained with Calcofluor white and DCHF-DA. (B) 60 µg/mL of the antioxidants N-acetylcysteine (NAC) and reduced glutathione (GSSH) counteract the effect of H2S on T. rubrum. (C) Effect of H2S on cytochrome c oxidase activity in isolated mitochondria from T. rubrum. Statistical analysis was done by a One-way ANOVA followed by Dunnett’s multiple comparison test (**p < 0.01; ***p < 0.001, n = 3). (D) 5-IAF labelling of proteins from T. rubrum conidia, separated by SDS-PAGE. (Left) Coomassie brilliant blue stain of the gel showing molecular weight marker bands (in kDa) and protein extracts from samples that were untreated ( −) or treated with NaHS ( +) for 3 h. (Right) 5-IAF labelled proteins.

Cytochrome C oxidase (COX), the terminal oxidase in the respiratory chain, is a known cellular target of H2S[36](https://www.nature.com/articles/s41598-025-22062-7#ref-CR36 “Cooper, C. E. & Brown, G. C. The inhibition of mitochondrial cytochrome oxidase by the gases carbon monoxide, nitric oxide, hydrogen cyanide and hydrogen sulfide: Chemical mechanism and physiological significance. J. Bioenerg. Biomembr. 40, 533–539. https://doi.org/10.1007/s10863-008-9166-6

(2008).“). We tested effects of H2S on COX activity in crude mitochondrial extracts of T. rubrum. As shown in Fig. 5C, H2S inhibited COX activity, with 42% of the activity remaining at 0.4 times the MICaq (3.4 µg/mL) and 15% at twice the MICaq (17 µg/mL H2S).

H2S is also known to induce S-sulfhydration, modifying thiol groups in proteins into persulphides (R–S–S–H) [37](https://www.nature.com/articles/s41598-025-22062-7#ref-CR37 “Paul, B. D. & Snyder, S. H. H2S: A novel gasotransmitter that signals by sulfhydration. Trends Biochem. Sci. 40, 687–700. https://doi.org/10.1016/j.tibs.2015.08.007

(2015).“). To assess this, T. rubrum was treated with H2S and proteins were extracted and labelled with the fluorescent probe 5’-iodoacetamide fluorescein (5-IAF). S-sulfhydrated thiols cannot react with 5-IAF, and thus are not labelled. As is shown in Fig. 5D, total protein levels (Coomassie staining, left) were similar between treated and untreated cells. In contrast, 5-IAF labelling revealed many fluorescent proteins in untreated cells, but much less in treated cells. To quantify this, total 5-IAF labelled protein with or without H2S was measured by analysing pixel intensity of the SDS_PAGE with ImageJ. To account for differences in protein levels between treated and untreated cells, 5-IAF signals were normalised against Coomassie-stained protein. This showed a 69.5% (+ /−9.7%) reduction in 5-IAF labelled protein when comparing treated against untreated sample (n = 3, p = 0.012, paired t test). This indicates that H2S results in extensive thiol modification.

Transcriptomic analysis of H2S-treated T. rubrum

Transcriptomic analysis was performed to obtain a more global view of the impact of H2S. T. rubrum was cultured in RPMI-GM (pH 5) and then treated for 3 h in the absence or presence of a sublethal concentration of NaHS (equivalent to 3.4 µg/mL H2S), followed by RNA isolation. After sequencing and data filtering, an average of 24 million high-quality reads of 100-bp paired-end sequences were obtained per sample. The average Q20 and Q30 scores were 97.1% and 91.1%, respectively, confirming high sequence quality. The sequencing files were aligned to the T. rubrum CBS118892 reference genome, and differential expression analysis was then performed using DESeq. This revealed that H2S resulted in the upregulation of 96 genes and the downregulation of 117 genes. The distribution of the log fold-change and adjusted p values (p < 0.05) are shown in Fig. 6A. Functional classification of these genes is shown in Fig. 6B and C. A complete list of differentially expressed genes (log2-fold change ≥ 1) and their classification is shown in Supplementary Table S1.

Fig. 6

Transcriptome analysis of T. rubrum ATCC28188 treated with H2S. (A) Volcano plot showing differentially expressed genes (absolute log2 fold change > 1, adjusted p value < 0.05). Red dots represent upregulated genes, blue dots represent downregulated genes, and grey dots indicate genes with no statistically significant differences in their expression. (B) and (C) show the classification of genes that are upregulated or downregulated, respectively.

Gene Ontology (GO) enrichment analysis was performed using the FungiDB platform [30](https://www.nature.com/articles/s41598-025-22062-7#ref-CR30 “Basenko, E. Y. et al. FungiDB: An integrated bioinformatic resource for fungi and oomycetes. J. Fungi (Basel) 4, 39. https://doi.org/10.3390/jof4010039

(2018).“). As shown in Fig. 7, genes encoding membrane proteins—particularly transporters—were enriched among those that are upregulated. Among downregulated genes, there was enrichment for those encoding ribosomal components or factors involved in ribosome biogenesis.

Fig. 7

Enrichment analysis of genes that are up- or down-regulated in response to H2S. The panels show only GO terms with an adjusted p value (Benjamini) of p < 0.05. Biological process is shown in green, molecular function in blue, and cellular component in red.

In the study by Fu et al.[14](https://www.nature.com/articles/s41598-025-22062-7#ref-CR14 “Fu, L. H. et al. An antifungal role of hydrogen sulfide on the postharvest pathogens Aspergillus niger and Penicillium italicum. PLoS ONE 9, e104206. https://doi.org/10.1371/journal.pone.0104206

(2014).“), reduced expression of catalase (CAT) and superoxide dismutase (SOD) genes was suggested as a reason for H2S-dependent ROS development. The genome of T. rubrum contains three putative CAT genes (gene IDs: TERG_01252, TERG_06053, and TERG_02005), and four SOD genes (TERG_07262, TERG_04335, TERG_04819, and TERG_08969). None were downregulated in response to H2S in T. rubrum, and all but one showed no significant change in expression. The exception was TERG_08969, a gene encoding a putative Cu/Zn SOD, which was upregulated nearly threefold in the presence of H2S (Supplementary Table S1).

Due to ROS production in the presence of H2S, it was expected that several genes encoding oxidative stress regulators would be differentially expressed. A recent study in T. rubrum identified that the transcription factor StuA is involved in regulation of essential antioxidant genes[38](https://www.nature.com/articles/s41598-025-22062-7#ref-CR38 “Petrucelli, M. et al. The transcription factor StuA regulates oxidative stress-responsive genes in Trichophyton rubrum. Int. J. Mol. Sci. 25, 12959. https://doi.org/10.3390/ijms252312959

(2024).“). Several of the genes that were identified in that study showed statistically significant increased expression in response to H2S (Table 5), but the levels of overexpression were modest (1.3–1.8-fold).

Discussion

We examined whether exogenous H2S has activity against microbes causing nail infections, which are difficult to treat due to poor penetration of antifungals into the nail plate[10](#ref-CR10 “Kubota-Ishida, N. et al. Human onychopharmacokinetic and pharmacodynamic analyses of ME1111, a new topical agent for onychomycosis. Antimicrob. Agents Chemother. 62, 10. https://doi.org/10.1128/AAC.00779-17

(2018).“),[11](#ref-CR11 “Hui, X. et al. In vitro penetration of a novel oxaborole antifungal (AN2690) into the human nail plate. J. Pharm. Sci. 96, 2622–2631. https://doi.org/10.1002/jps.20901

(2007).“),[12](https://www.nature.com/articles/s41598-025-22062-7#ref-CR12 “McAuley, W. et al. An investigation of how fungal infection influences drug penetration through onychomycosis patient’s nail plates. Eur. J. Pharm. Biopharm. 102, 178–184. https://doi.org/10.1016/j.ejpb.2016.03.008

(2016).“). A recent study showed that polar molecules with a low molecular weight (< 120 g/mol) readily penetrate the nail plate. H2S, being weakly polar with a molecular weight of 34 g/mol, penetrates nails much more efficiently than topical antifungals such as amorolfine or ciclopirox[13](https://www.nature.com/articles/s41598-025-22062-7#ref-CR13 “Malallah, O. et al. Systematic review and QSPR analysis of chemical penetration through the nail to inform onychomycosis candidate selection. Drug Discov. Today 29, 103844. https://doi.org/10.1016/j.drudis.2023.103844

(2024).“). Because of its antimicrobial activity, H2S was proposed as a promising candidate for onychomycosis treatment, capable of reaching pathogens deep within the nail.

We demonstrated that H2S is effective against pathogens causing nail infections, including a terbinafine-resistant isolate of T. indotineae. The effects of sulphide differ between gaseous and aqueous forms and, in aqueous conditions, depend on the pH. In solution, there is an equilibrium of H2S ⇌ HS− ⇌ S2−, with pKa values of 7.04 and 11.96 for the first and second step, respectively. At pH 5, H2S predominates, at pH 7 the ratio of H2S to HS− is ~ 1:1, and at pH 8, HS− is the major form. The highest activity was observed at pH 5 (a 20-fold lower MICaq value), suggesting that H2S is more active than HS−. However, the cytoplasmic pH remains near neutral[39](https://www.nature.com/articles/s41598-025-22062-7#ref-CR39 “Kane, P. M. Proton transport and pH control in fungi. Adv. Exp. Med. Biol. 892, 33–68. https://doi.org/10.1007/978-3-319-25304-6_3

(2016).“), and the intracellular ratio of H2S:HS− is likely constant. This suggests the differences in activity are primarily related to uptake: H2S is less polar than water and can diffuse freely across membranes[40](https://www.nature.com/articles/s41598-025-22062-7#ref-CR40 “Riahi, S. & Rowley, C. N. Why can hydrogen sulfide permeate cell membranes?. J. Am. Chem. Soc. 136, 15111–15113. https://doi.org/10.1021/ja508063s

(2014).“), whereas the charged HS− ion is likely transported more slowly via proteins. However, membrane permeability is pH-dependent[41](https://www.nature.com/articles/s41598-025-22062-7#ref-CR41 “Miedema, H., Staal, H. & Prins, H. B. A. pH-induced proton permeability changes of plasma membrane vesicles. J. Membr. Biol. 152, 159–167. https://doi.org/10.1007/s002329900094

(1996).“), and because the relative abundance of H2S and HS− also varies with pH, it becomes difficult to predict precisely how much of each species crosses the membrane under physiological conditions. For instance, if the external acidic environment was to facilitate the uptake of HS−, this could shift the equilibrium towards greater internalisation of the ionised form. Nevertheless, the difference in species activity and the known properties of H2S diffusion are the most likely explanations for the observed results.

It should be noted that the pH in the cytoplasm is close to neutral, thus once H2S internalises, approximately half of that converts to the anion and the cytoplasmic concentrations of H2S and HS− will be similar. This equilibrium makes it difficult to determine which species is primarily responsible for the observed biological effects, and the interconversion between the two further complicates such distinctions. Consequently, in discussions on the antifungal activity of H2S, the term also implicitly includes HS−.

In the gaseous state, the pH of the growth medium has no effect on H2S activity, as in air all sulphide will be in the form of H2S which, as mentioned before, freely diffuses into cells. Interestingly, gaseous H2S is more active than aqueous H2S (at pH 5, the MICg_max is 50-fold lower than the MICaq). We cannot exclude different modes of action of gaseous and aqueous forms, for instance if some targets are extracellular. However, once H2S diffuses into the cytoplasm, the cellular targets are likely to be the same, irrespective of whether H2S entered in the gaseous or aqueous form. Therefore, we think that the difference in activity between gaseous and aqueous forms is, similar to what we observed at different pH values, more likely related to the uptake of H2S. In the gaseous form, H2S can diffuse rapidly across membranes, whereas the aqueous form will interact with water or other solutes through hydrogen bonding42, thereby slowing down the rate of diffusion into cells.

It should be noted that the true gaseous MIC values are hard to define in a closed system, as H2S release varies with donor concentration and time. For dermatophytes, we found MIC values with 0.25–0.5 mM NaHS in 20 mL, equating to MICg_max levels of 0.082–0.16 µg/mL. At those concentrations, approximately 16% of sulphide is released as H2S after 6 h—time sufficient to kill the fungi—suggesting true gaseous MICs of 0.013–0.026 µg/mL. That compares favourably with topical antifungals such as ciclopirox (0.25 µg/mL) and amorolfine (0.06 µg/mL)[43](https://www.nature.com/articles/s41598-025-22062-7#ref-CR43 “Ghannoum, M., Isham, N. & Long, L. In vitro antifungal activity of ME1111, a new topical agent for onychomycosis, against clinical isolates of dermatophytes. Antimicrob. Agents Chemother. 59, 5154–5158. https://doi.org/10.1128/AAC.00992-15

(2015).“). However, these values were determined using different methodology, making a direct comparison difficult. This comparison is also relevant when comparing the MICaq and MICg_max. Both metrics relate to the total amount of sulphide present. However, in the case of H2S gas, the fungi are not exposed to all available sulphide. This indicates that the difference between the aqueous MIC and (true) gaseous MIC is even greater than indicated above. Nevertheless, as the gaseous H2S concentration is not constant, a direct comparison between the aqueous and true gaseous MIC is likely unfeasible.

In T. rubrum, early stages of growth such as spore germination were more sensitive to H2S, as the MICg_max increased when applied during hyphal formation, which was also observed in plant pathogens44. Germination involves rapid metabolic and respiratory changes[45](#ref-CR45 “Stade, S. & Brambl, R. Mitochondrial biogenesis during fungal spore germination: respiration and cytochrome c oxidase in Neurospora crassa. J. Bacteriol. 147, 757–767. https://doi.org/10.1128/jb.147.3.757-767.1981

(1981).“),[46](#ref-CR46 “Novodvorska, M. et al. Metabolic activity in dormant conidia of Aspergillus niger and developmental changes during conidial outgrowth. Fungal Genet. Biol. 94, 23–31. https://doi.org/10.1016/j.fgb.2016.07.002

(2016).“),47. In Aspergillus fumigatus, increased oxygen uptake is observed 3.5 h after germination[46](https://www.nature.com/articles/s41598-025-22062-7#ref-CR46 “Novodvorska, M. et al. Metabolic activity in dormant conidia of Aspergillus niger and developmental changes during conidial outgrowth. Fungal Genet. Biol. 94, 23–31. https://doi.org/10.1016/j.fgb.2016.07.002

(2016).“), while in Botryodiplodia theobromae, the activity of COX peaks at 3.5–4 h47. H2S is known to inhibit COX through binding to the haem moiety, competing with oxygen[48](https://www.nature.com/articles/s41598-025-22062-7#ref-CR48 “Nicholls, P., Marshall, D. C., Cooper, C. E. & Wilson, M. T. Sulfide inhibition of and metabolism by cytochrome c oxidase. Biochem. Soc. Trans. 41, 1312–1316. https://doi.org/10.1042/BST20130070

(2013).“). Our observation that 3–6 h of H2S exposure is needed to inhibit T. rubrum conidia aligns with this, and we also identified COX as one of the targets of H2S in T. rubrum. However, germination is a complex process and other cellular processes may also be affected, since H2S can also inhibit other proteins with metal centres[49](https://www.nature.com/articles/s41598-025-22062-7#ref-CR49 “Mantle, D. & Yang, G. Hydrogen sulfide and metal interaction: The pathophysiological implications. Mol. Cell. Biochem. 477, 2235–2248. https://doi.org/10.1007/s11010-022-04443-y

(2022).“).

Similar to previous studies[14](https://www.nature.com/articles/s41598-025-22062-7#ref-CR14 “Fu, L. H. et al. An antifungal role of hydrogen sulfide on the postharvest pathogens Aspergillus niger and Penicillium italicum. PLoS ONE 9, e104206. https://doi.org/10.1371/journal.pone.0104206

(2014).“),[15](https://www.nature.com/articles/s41598-025-22062-7#ref-CR15 “Fu, L. H. et al. Hydrogen sulfide inhibits the growth of Escherichia coli through oxidative damage. J. Microbiol. 56, 238–245. https://doi.org/10.1007/s12275-018-7537-1

(2018).“), we found that H2S increases ROS, based on the response of the redox-sensitive probe DCHF-DA and antioxidant-mediated increases in MICaq values of H2S. Notably, this effect is concentration-dependent; at low endogenous concentrations, H2S can function as an antioxidant and a signalling molecule[36](https://www.nature.com/articles/s41598-025-22062-7#ref-CR36 “Cooper, C. E. & Brown, G. C. The inhibition of mitochondrial cytochrome oxidase by the gases carbon monoxide, nitric oxide, hydrogen cyanide and hydrogen sulfide: Chemical mechanism and physiological significance. J. Bioenerg. Biomembr. 40, 533–539. https://doi.org/10.1007/s10863-008-9166-6

(2008).“). However, higher concentrations of exogenous H2S lead to ROS that potentially damages DNA, lipids and proteins[50](https://www.nature.com/articles/s41598-025-22062-7#ref-CR50 “Juan, C. A., Perez de la Lastra, J. M., Plou, F. J. & Perez-Lebena, E. The chemistry of reactive oxygen species (ROS) revisited: Outlining their role in biological macromolecules (DNA, lipids and proteins) and induced pathologies. Int. J. Mol. Sci. 22, 642. https://doi.org/10.3390/ijms22094642

(2021).“). In this respect, the observation that E. coli BW25113 was more resistant than E. coli DH5α is interesting: the latter is recA*-*, a gene involved in DNA repair[51](https://www.nature.com/articles/s41598-025-22062-7#ref-CR51 “Del Val, E., Nasser, W., Abaibou, H. & Reverchon, S. RecA and DNA recombination: A review of molec