Jun Lei Gao, Bai Hao Li, Zhi Bin Ban, Li Jia Li, Hao Liang, Xiao Gang Yan, Cheng Xiao, Fang Yu Zhang

Jilin Academy of Agricultural Sciences ( Northeast Agricultural Research Center of China ), Gongzhuling, Jilin, 136100, China. email:gaojunlei199405@163.com

Hydrogen sulfide (H₂S) is a common industrial hazard that originates from the petroleum and gas industries, intensive animal farming, sewage, and waste treatment plants^1–2. H₂S is a highly toxic and malodorous gas, but H2S endogenous to organisms affects various biological activities and signal transduction functions³. The central nervous, respiratory, and cardiovascular systems are the tissues most sensitive to H₂S, and the clinical manifestations of their exposure depend on the concentration of H₂S and the exposure time⁴. In this regard, exposure to between 10 ppm and 500 ppm induces respiratory symptoms that range from rhinitis to acute respiratory failure, whereas exposure to between 500 ppm and 1000 ppm may cause immediate loss of consciousness and death⁵.

At concentrations between150 ppm and 250 ppm, H₂S can induces neurotoxic damage leading to olfactory nerve paralysis after only a few breaths of contaminated air are inhaled, according to UK Health and Safety Executive Guidance. Worldwide occupational exposure limits for H₂S are in the range of 1 ppm to 10 ppm ⁶. However, maintaining slurry systems on farms with powdered gypsum as bedding⁷, power washing in swine operations, and removing mushroom compost can cause increases in environmental H₂S levels ⁸. The H₂S concentrations in floor-based poultry houses and swine barns range from 0 to 97 ppm⁸. In this regard, the effects on health caused by exposure to H₂S at concentrations less than or equal to 100 ppm have not yet been thoroughly investigated. On the other hand, one study showed that inhalation of 80 ppm H₂S for 6 h reduces oxidative stress and protects the lungs from ventilator damage, intranasal lipopolysaccharide exposure, and inhalation of cotton smoke⁹. Furthermore, inhalation of 40 ppm H₂S for 8 h per day for 7 days induced the degeneration of tyrosine hydroxylase-containing neurons and caused gliosis in the nigrostriatal region of the brains of mice induced with neurotoxin 1-methyl-4-1,2,3,6-tetrahydropridine, possibly because of upregulation of antioxidant defense mechanisms¹⁰. Rats were administered with hydrogen sulfide (H₂S) by inhalation (80 ppm at 0, 1.5, and 3 hours after reperfusion) or intravenous injection of the slow-releasing H₂S donor GYY 4137. The results showed that both inhaled H₂S and intravenous GYY 4137 administration could improve neuronal cell survival¹¹. The findings of these studies changed the prevailing impression of the toxicity of low-level H2S inhalation, necessitating a systematic evaluation of how relatively low concentrations (below 100 ppm) of H2S affect normal physiological status, especially with respect to the molecular mechanism of this effect.

Therefore, the aim of the present study was to investigate the effects of between 60 ppm and 100 ppm H2S on immunological parameters in mouse blood, liver and kidney enzymatic activity levels, and alterations in lung gene expression by utilizing RNA sequencing (RNA-seq) and RT-qPCR technologies. These results might be important to individuals who work in industrial environments contaminated with H₂S and exposed to the gas for several hours and are expected to contribute to the occupational care and safety management needed for mitigating H₂S exposure.

The in vivo research was conducted at Jilin Academy of Agricultural Sciences, Gongzhuling, Jilin, China. All methods were performed in accordance with the relevant guidelines, all methods in this study following the Principles of Regulations on Administration of Experimental Animals of the Chinese government and relevant animal ethics standards of the Jilin Academy of Agricultural Sciences. Sixty adult female Kunming mice (22–25 g) were obtained simultaneously with same age and randomly divided into four groups in triplicate, and five mice selected randomly from each group were designed as one replicate and placed in one chamber, each mouse was exposed to H₂S only once. The mice were allocated into treatment groups according to their exposure to H₂S at the following concentrations: 0 ppm (group A), 60 ppm (group B), 80 ppm (group C), and 100 ppm (group D). Animals were consecutively enclosed in 4 sealed glass chambers (30.5 cm in length, 35 cm in width, 37 cm in height, and 0.0395 m3 in volume) with stainless steel frames and covers and a polyvinyl chloride back side. A miniature electronic fan was placed inside the chamber, and a gas injection orifice with a polycarbonate tube and switch were installed on the back side. Chamber connecting points were sealed with structural adhesive, whereas junctions between the chamber and the cover were sealed with rubber sealing strips. The seals on the chamber were tested by the white smoke method with incense, not gas, to prevent gas leaks, and the maximum barometric air pressure in the chamber was 0.95- to 1.05-fold the pressure of the external ambient atmosphere, as tested through air extraction and injection. The mice were maintained in a pathogen- and stress-free environment from 22°C to 25°C and from 50–60% humidity under a light-dark cycle (light phase, 07:00–19:00 h) and given water and food ad libitum. Before the experiment, the mice were acclimated to this environment for one week.

H₂S gas (10,084 ppm) was obtained from Jining Xieli Special Gas Co., Ltd., China. At 06:00 h on the 8th day, food and water were removed, and the chambers were carefully cleaned to avoid stressing the mice. At 07:00 h, the chambers housing mouse groups A, B, C, and D were injected with 0.235 L air (placebo), 0.235 L (60 ppm), 0.313 L (80 ppm), and 0.392 L (100 ppm) pure H₂S, respectively, as well as 2.0 L oxygen in all chambers (99.99%, Jining Xieli Special Gas Co., Ltd., China) with medical injectors that had been installed and adapted with small gas valves before H₂S injection to prevent gas leaks. After gas injection, the miniature fan in each chamber was turned on to mix the gas, and the animals were exposed for 6 h. The O₂ percentage remaining and CO₂ percentage accumulating in the chamber after 6 h were calculated according to the amount of O₂ consumption and the respiratory quotient, avoiding O₂ and CO₂ concentrations exceeding the limit to maintain normal life functions.

After H₂S treatment, two mice from each identical batch of the corresponding chamber were randomly selected and immediately sacrificed by intraperitoneal injection of pentobarbital sodium(Injected at a dose of 45 mg/kg). The mice were bled from the retroorbital plexus using 0.2% heparin sodium, and the blood was centrifuged at 3000 rpm for 10 min and stored at -20°C for immune index testing. Liver, kidney, and lung tissue samples were collected, cut into 1 cm3 pieces, and stored in liquid nitrogen until use. Serum interleukin-8 (IL-8, Xinbosheng Biotech. Co. Ltd., Shenzhen, China), tumor necrosis factor- (TNF-α), interleukin-1 (IL-1), and interleukin-10 (IL-10) (Jianglai Biotech. Co. Ltd., Shanghai) levels were measured by double-antibody sandwich ELISA using an avidin-biotin-peroxidase complex.

In addition, mouse liver and kidney tissues were mixed with cold 0.90% saline for the homogenate and single-cell suspension preparation (Medimachine II, Syntec Co. Ltd., Ireland). Homogenates were centrifuged at 3000 rpm for 10 min using refrigerated equipment (Biofuge Stratos, Thermo Co. Ltd., Germany), and supernatant fluids were obtained for malonaldehyde (MDA; thiobarbituric acid colorimetry), superoxide dismutase (SOD; xanthine oxidase colorimetry), catalase (CAT; ammonium molybdate colorimetry), glutathione peroxidase (GSH-Px; dithiodinitrobenzoic acid colorimetry), glutathione S-transferase (GST; 1-chloro-2,4-dinitrobenzene chromogenic assay), and glutathione reductase activation coefficient (GRAC; nicotinamide adenine dinucleotide phosphate colorimetry) analyses using an ultraviolet and visible spectrophotometer (UV-1780, Shimadzu Co. Ltd., Japan) and corresponding analytic kits (Ruixin Biotech. Co. Ltd., Quanzhou, China). Single-cell suspensions were cultured with a fluorescence probe (2,7-dichlorofluorescin diacetate) at 37°C for 20 min, and the cells were subsequently collected to measure reactive oxygen species (ROS) activity with a fluorospectrophotometer (RF-540, Shimadzu Co. Ltd., Japan) at a 488-nm excitation wavelength and 525-nm emission wavelength based on the analytic kit protocols (Ruixin Biotech. Co. Ltd., Quanzhou, China).

Total RNA of mouse lung tissue was extracted with an RNeasy Micro Kit according to the manufacturer’s instructions (Qiagen, Maryland, Cat no. 74004). Total RNA was detected with 1% agarose gel electrophoresis. The results showed that the RNA bands were clear, and extra bands were not observed. The purity of the RNA detected by the spectrophotometer was between 1.8 and 2.0. The RNA was verified to be of good integrity and at a high concentration by using an Agilent 2100 RNA Nano 6000 assay kit (Agilent Technologies Inc., Palo Alto, CA, USA), which was also used for subsequent detection. Oligoattached magnetic beads were utilized to purify the mRNA with a poly-A structure from the pool of total RNA. Next, 300-bp fragments were obtained through divalent cation fragmentation in Illumina proprietary fragmentation buffer. First-strand cDNA was synthesized by using random oligonucleotides and Super Script II reverse transcriptase (Thermo Fisher Scientific, Waltham, MA, USA); second-strand cDNA was subsequently synthesized from the first cDNA template using DNA polymerase I and RNase H (Illumina Inc., San Diego, CA, USA). Illumina PCR Primer Cocktail (Illumina Inc.) in a 15-cycle PCR (Applied Biosystems, Thermo Fisher Scientific Inc., Shanghai, China) was adopted for library fragment enrichment after library construction. The 450-bp library was selected according to fragment length. The products were subsequently purified (AMPure XP system) and quantified by a high-sensitivity DNA assay in an Agilent 2100 Bioanalyzer (Agilent Technologies Inc.). The HiSeq platform (Illumina Inc.) from Personalbio Co. Ltd. (Shanghai, China) was used for library sequencing with high-throughput sequencing technology.

The samples were converted into image files by sequencing, and the files were subsequently converted to the FASTQ format called Raw Data. We evaluated the raw data of each sample, including the sample name, percentage of fuzzy bases, Q20 (%) and Q30 (%). Sequencing data contain some low-quality reads, which interfere with the subsequent information analysis; therefore, it is necessary to filter the sequencing data. We obtained high-quality data by removing low-quality reads, empty reads, adapter sequences, and reads with more than 10% N sequences. Next, the high-quality reads were subjected to base quality and content distribution tests. High-quality clean data were mapped to the Mus_musculus.GRCm38.dna.fa genome using HISAT software (http://ccb.jhu.edu/software/hisat2/index.shtml). The distribution of the reads mapped to the genome was statistically analyzed, including CDSs (coding sequences), introns (introns), intergenic regions and UTRs (5’ and 3’ untranslated regions). The reads distributed in intergenic regions were considered new genes or new noncoding RNAs.

Fragments per kilobase per million mapped reads (FPKM) and HTSeq (0.9.1) statistics (https://htseq.readthedocs.io/en/release_0.9.1) were utilized to qualify and analyze the gene expression levels. A log2-fold change| >1 was considered significant (p < 0.05) when evaluating the A vs. B, A vs. C, A vs. D, B vs. C, B vs. D, and C vs. D groups. The R language Pheatmap (1.0.8) software package (https://cran.r-project.org/web/packages/pheatmap/index.html) was also used to perform bidirectional clustering analysis for all different genes.

DEGs of each group were subjected to GO and KEGG pathway (https://www.genome.jp/kegg/) analyses. The top enriched GO terms and KEGG pathways corresponding to physiological processes affected by H₂S were identified by the biological activity, cellular component, and molecular function categories.

The 10 differentially expressed genes were obtained from RNA-seq to verify the accuracy of the sequencing data obtained by qRT-PCR. Total RNA was extracted from lung tissue with TRIzol (Thermo Fisher Scientific), and the RNA concentration was measured using a NanoDrop 5000 spectrophotometer (Thermo Fisher Scientific) following the manufacturer’s instructions. In addition, qRT-PCR was performed by the Reverse Transcription System (TaKaRa) in a C1000 Thermal Cycler (Bio-Rad Inc., Hercules, CA, USA) using the glyceraldehyde-3-phosphate dehydrogenase gene (GAPDH) as an internal control. qRT-PCR analysis was developed on a LightCycler 480 system (Roche Applied Science, Penzberg, Germany) with SuperMix Real PreMix Plus (SYBR green) (Roche Applied Science). PCR mixtures were initially heated to 95°C for 15 min, followed by 40 cycles of 95°C for 10 s, 60°C for 20 s, and 72°C for 32 s. The relative expression levels of genes were calculated using the 2- ΔΔCT method. Primer sequences are shown in Table 1.

Independent experiments in three replications were performed for all analyses. Five biological replicate experiments were performed for the sequencing analyses. Experimental data were assessed with one-way ANOVA, and Tukey,s multiple range test was used to identify differences due to treatment with SPSS software (Version 21.0 SPSS, Inc., Chicago, IL, USA). A p-value < 0.05 was considered to indicate a significant difference, and a p-value < 0.01 indicated a very significant difference.

After H₂S exposure, all the mice were removed from the chambers and the behavior and physiological statuses of the mice were normal; they were quiet, showing no agitation, vomiting, ataxia or unconsciousness. For serum immune indexes, H₂S treatments significantly increased TNF-α levels compared with the 0-ppm treatment (p < 0.05), and the TNF-α level after 80 ppm treatment was higher than that after other treatments. In addition, exposure to 80 ppm and 100 ppm H₂S increased the IL-1 levels compared with exposure to 60 ppm and 0 ppm H₂S (p < 0.05), but IL-8 levels were not significantly altered in any treatment group. Furthermore, exposure to 100 ppm H₂S lead to a significant increase in IL-10 levels compared with the other treatments (p < 0.01), and exposure to 80 ppm H₂S induced significantly higher levels of IL-10 than the levels induced by 0 ppm H₂S (p < 0.01) (Table 2).

a,b,c Means within a row with different superscripts differ (p < 0.05).

Table 2. Effects of H₂S on blood immune indexes and enzymatic activity levels of liver and kidney tissues.

In regard to redox-reaction enzymatic activities of liver and kidney tissues, the 80 ppm and 100 ppm H2S treatments significantly decreased the MDA content (p < 0.05), compared with the 0 and 60 ppm H2S treatments, and the MDA content after the 80 ppm treatment was significantly lower than that after the 60 ppm treatment (p < 0.05). In contrast, the 80 ppm and 100 ppm H₂S treatments significantly decreased SOD activity levels (p < 0.05) compared with the 0 and 60 ppm H₂S treatments. Furthermore, H₂S significantly increased CAT (p < 0.01) and GSP-Px (p < 0.05) activity levels compared with those in the untreated control, but there was no difference among H₂S treatments. The GST activity of the mice exposed to 100 ppm H₂S was significantly increased compared with that of the mice exposed to other treatments (p < 0.01), whereas 80 and 100 ppm H₂S induced significantly higher GRAC than did 60 ppm (p < 0.01), which was higher than induced by the 0 ppm treatment (p < 0.01). The ROS observed in mice exposed to the 100 ppm H₂S treatment were significantly higher than those observed in the mice exposed to the 0 and 60 ppm treatments (p < 0.01), and the ROS levels in the mice exposed to the 60 and 80 ppm treatments were significantly higher than those in the mice exposed to the 0 ppm treatment (p < 0.01) (Table 2).

A total of 20 cDNA libraries were constructed, and we obtained more than 6 GB of clean bases for each sample. As shown in Table 3, every sample generated numerous raw reads, whose Q30 rate was more than 94%, suggesting that sequencing data were of high quality and reliable. After poly(N)-containing, low-quality, and adapter-containing reads were removed from the raw data, the clean read rate was more than 93%. When the filtered reads were compared with a reference genome by using HISAT2 software, the mapping rate was more than 96%, and the CDS region of read content showed the highest rate of clean reads.

The differentially expressed genes were analyzed based on comparisons among six groups: A vs. B, A vs. C, A vs. D, B vs. C, B vs. D, and C vs. D (Table 3). We clustered the analyzed differentially expressed genes and generated six heat maps using R language software, as shown in Fig. 1. We identified 70 significantly dysregulated genes (29 and 41 genes were upregulated and downregulated) between groups A and B, whereas 49 genes were differentially expressed (25 mRNAs increased and 24 mRNAs decreased) between groups A and C. Furthermore, 333 genes were differentially expressed (54 upregulated and 279 downregulated) between groups A and D; 141 genes were differentially expressed (98 upregulated and 43 downregulated) between groups B and C; 188 genes were differentially expressed (48 upregulated and 140 downregulated) between groups B and D; and 182 genes were differentially expressed (31 upregulated and 151 downregulated) between groups C and D. The detailed information was showed in TableS1-S6. We uploaded the raw sequencing data on the SRA (Sequence Read Archive) accession number SUB8720470.

(A–F) Expression profiles of mRNAs. In the volcano plots, blue, red, and gray points represent mRNAs that were downregulated, upregulated, and not significantly different in the different groups, respectively. x axis: log2 ratio of mRNA expression levels between different groups (A_B, A_C, A_D, B_C, B_D, C_D). y axis: false discovery rate values (-log10 transformed) of the mRNAs. (G) Cluster analysis of expression of mRNAs. Red and green: increased and decreased expression, respectively.

The GO analysis was performed according to the DEGs. The most highly enriched GO terms in the molecular function (MF), cellular component (CC) and biological process (BP) categories are presented in Fig. 2. The top three terms between groups A and B were protein antigen binding (GO: 1990405), MHC protein complex (GO: 0042611) and immune response (GO: 0006955), whereas the top three terms between groups A and C were L-tyrosine:2-oxoglutarate aminotransferase activity (GO: 0004838), perineuronal net (GO: 0072534) and DNA demethylation (GO: 0080111); the top three terms between groups A and D were nucleoside-triphosphatase activity (GO: 0017111, chromosome, centromeric region (GO: 0000775) and mitotic cell cycle process (GO:1903047). The top three terms between groups B and C were actin binding (GO: 0003779), contractile fiber (GO: 0043292) and muscle system process (GO: 0003012); the top three terms between groups B and D were protein binding (GO: 0005515), chromosomal region (GO: 0098687) and mitotic cell cycle process (GO: 1903047). The top three terms between groups C and D were purine ribonucleoside triphosphate binding (GO: 0035639), chromosome, centromeric region (GO: 0000775), and mitotic cell cycle process (GO: 1903047).

KEGG pathway enrichment analysis indicated that signaling pathways were significantly different among the experimental groups. The top three pathways of DEG enrichment between groups A and B were graft-versus-host disease, type-I diabetes mellitus, and allograft rejection; those between groups A and C were cytokine-cytokine receptor interaction, intestinal immune network for IgA production, and acute myeloid leukemia; those between groups A and D were cell cycle, malaria, and cytokine-cytokine receptor interaction; those between groups B and C were cardiac muscle contraction, dilated cardiomyopathy, and hypertrophic cardiomyopathy; those between groups B and D were cell cycle, p53 signaling pathway, and cytokine-cytokine receptor interaction; and those between C and D were oocyte meiosis, progesterone-mediated oocyte maturation, and cell cycle, respectively. The detailed information was showed in TableS7-S18.

(A, C, E, G, I, K) GO analysis results showing differentially expressed genes. In the histogram, red, green, and blue represent genes enriched in the cellular component, molecular function, and biological process categories, respectively. x axis: GO terms. y axis: false discovery rate values (-log10 transformed).

(B, D, F, H, J, L) Significantly enriched KEGG pathways with differences meeting the cut-off of a p value < 0.05. Each line represents a pathway, and the red dotes indicate the genes significantly enriched.

To confirm the accuracy of RNA-seq in this experiment, 10 significantly differentially expressed genes in the six groups were randomly determined by qPCR. These genes included fibroblast growth factor binding protein 1, FGFBP1: GATA-binding protein 3, GATA3: killer cell lectin-like receptor G1, KLRG1; lysyl oxidase, LOX; microfibril-associated protein 4, MFAP4; NLR family, pyrin domain containing 1A, NLRP1A; NUF2 component of the NDC80 kinetochore complex, NUF2; proliferating cell nuclear antigen pseudogene 2, PCNA-PS2; solute carrier family 7 member 11, SLC7A11; and T-box transcription factor 21, TBX21. As shown in Fig. 3, the expression levels of these genes exhibited significant differences in the four groups. In addition, the qPCR results were highly consistent with the RNA-seq data, which suggested that the RNA-seq data were reliable and accurate and that the genes may be identified as candidate targets.

mRNA expression was quantified relative to the Gapdh expression level by using the comparative cycle threshold (△CT) method.

In livestock and poultry breeding, H₂S primarily originates from animal manure and microbial decomposition processes¹². High concentrations of H2S cause respiratory diseases and reduce immune function and production performance⁴. H₂S toxicity primarily affects the nervous system, but some studies have reported that H2S has anti-inflammatory, antitumor, cardioprotective and regulates redox^13–16. The effects of H2S on mouse immune indexes in serum were evaluated first. TNF-α is a tumor necrosis factor primarily secreted by monocyte/macrophages and can kill tumor cells and promote inflammatory reactions. The increase in serum TNF-α levels induced by H₂S indicated that H₂S can cause certain inflammatory reactions in the body. Interleukins are lymphokines that interact with leukocytes or immune cells and can transmit signals, activate and regulate immune cells and T and B cells and play important roles in the inflammatory response¹⁷. The increase in serum IL-1 and IL-10 levels indicated that H2S concentrations of 80 ppm and 100 ppm can elicit a systemic immune response and promote B cell proliferation and antibody secretion. The TNF-α and IL-1 in broiler spleens were also increased when broilers were exposed to 4 or 20 ppm H2S for six weeks¹⁸ .IL-8 was the first chemokine to be discovered, exhibiting pro-inflammatory activity¹⁹. Although there was no significant variation in the effects on IL-8 between different H₂S inhalation treatments, the IL-8 level after 100 ppm exposure was increased by 29.13% compared with that after 0 ppm exposure. These results of the serum immunological indicators analysis suggest that the inflammatory reaction of the mice was activated when the H2S inhalation concentration was increased, especially when the H2S concentration exceeded 80 ppm. When the anesthetized mice continuously inhaled 60 ppm H2S, the potential anti-inflammatory effects of the inhaled H2S were outweighed by the deleterious effects²⁰ .

MDA and ROS are oxidative agents that induce oxidative injury in organisms, whereas SOD, CAT, GSH-Px, GST and GRAC are antioxidative substances that eliminate excess free radicals and peroxides ^12,21. MDA is an indicator of lipid peroxidation and the peroxidation of biomembranes and lipoproteins²², but ROS are metabolites of oxygen and primarily include superoxide anions (O₂-), hydroxyl radicals (OH-) and hydrogen peroxide (H₂O₂)²³. In our study, differences in the mechanism modulated by MDA and the redox balance mediated by ROS led to various results. The reduction in MDA levels in the 80 and 100 ppm treatment groups might be attributed to the inhibiting effects of H₂S on nuclear factor kappa-B (NF-κB) activation, and nitric oxide synthase (iNOS), carbon monoxide (NO)²⁴. In addition, the increase in ROS accumulation might be related to oxygen utilization efficiency. The toxicity induced by H₂S exposure is involved in the inhibition of cytochrome coxidase, the terminal enzyme complex in the electron transport chain, which, in turn, blocks oxidative phosphorylation and limits the oxygen available for oxidative metabolism²⁵. The changes in these two oxidative agents indicate that H₂S inhalation may relieve lipid peroxidation to a certain degree but intensifies the accumulation of ROS; in particular, the ROS levels in 100 ppm H₂S treatment group was increased by nearly twice that in 0 ppm H₂S treatment group.

Antioxidative enzymes constitute the organism defenses by scavenging oxidants and repairing the damage caused by ROS; in particular, SOD is critical for catalyzing the dismutation of O₂- into H₂O₂, and the generated H₂O₂ and other hydroperoxides are catalyzed by CAT or GPS-Px into H₂O and O₂ by the oxidation of glutathione (GSH) to generate glutathione disulfide (GSSG)²⁶. Glutathione reductase (GR) subsequently reduces GSSG to GSH, and GST catalyzes the reaction of GSH conjugation to a wide range of electrophilic metabolites of xenobiotics²⁷. In contrast to that of SOD, the activity level of the four other antioxidant enzymes we evaluated in our study were all upregulated with increasing H₂S inhalation. Using a mouse model of Parkinson’s disease, a previous study found that inhalation of 40 ppm can lead to an increase in the gene expression of GST Mu1 and GST A4, but not GSH¹⁰. The different effects of H₂S exposure on the antioxidative enzymes in the present study and other studies might be related to differences in animal species, pathological or physiological states, inhaled H₂S concentrations or respiratory times, which require accurate evaluation. In addition, the opposite trend exhibited by SOD compared to the other antioxidant enzymes might be attributed to its relatively extensive set of regulatory substrates, which includes not only ROS but also lipid peroxides. For these reasons, the variation in the levels of ROS and the antioxidant enzymes, except SOD, might be closely related to the negative feedbacks induced by the accumulation of MDA or ROS. Organic redox balance is a complex process that involves various metabolic pathways; therefore, the regulatory mechanism governing the effect of H₂S inhalation on the redox system warrants in-depth investigation. The enzymatic activities of liver and kidney tissues suggest that H₂S inhalation induced oxidative stress, reactive oxygen species production and activation of the antioxidation pathways in mice.

Many DEGs in the groups exposed to different H₂S concentrations were identified in the present study. These DEGs included genes related to oxidative stress, inflammation and apoptosis. IL1F9 is a member of the interleukin 1 (IL-1) family and it promotes inflammation and is an agonist of IL-1²⁸. IL-1 can bind with IL-1 receptor-related protein 2 (IL-1 Rrp2) to activate the NF-κB signaling pathway²⁹. Deficiency in the mouse immune gene IL1F9 was impaired and allowed an inflammatory reaction to be induced by a virus³⁰. In this study, IL1F9 expression was increased in the 100 ppm H₂S treatment group, which suggests that H₂S caused cellular injury, and IL1F9 subsequently induced inflammation to eliminate the harmful factors. We selected 10 DEGs for identification by qPCR, and some of these can be considered candidate genes as possible mitigation targets. Fibroblast growth factor-binding protein 1 (FGFBP1) can bind with secreted fibroblast growth factor (FGF) extracellularly and enhance the biology of FGF to improve cell proliferation^[31]. FGFBP1 expression in this study was significantly changed, which suggested that H₂S can induce cell proliferation. GATA-binding protein 3 (GATA3) is a member of the GATA family and is the key regulator of innate immunity and adaptive immunity. GATA3 is highly expressed in T cells, B cells and other immune cells and plays important roles in controlling the development and peripheral maturation of T cells³². Several studies have shown that GATA3 binds to the promoter/enhancer region of the Il7r gene in CD8 + T cells to maintain immune function^33–34. In our study, GATA3 was significantly changed upon H₂S exposure, and this change was confirmed by qPCR. Therefore, GATA3 may be a candidate gene involved in the mechanism of mouse organism impairment. Like GATA3, killer cell lectin-like receptor G1 (KLRG1) can be expressed in natural killer (NK) cells and T cells. A special domain in KLRG1 plays an important role in the immune system^35–36. Therefore, we considered KLRG1 to be a candidate gene. T-box transcription factor 21 (TBX21) is an immune cell-specific member of the T-box family of transcription factors, and it is highly expressed in innate immune cells and can regulate the development and terminal maturation of NK cells³⁷. In the absence of TBX21, the function of NK cells was impaired, and the survival rate was decreased³⁸. TBX21 can bind to the GATA3 gene, form a complex and promote the inhibitory chromatin modification of the Gata3 site, thereby inhibiting the expression of Gata3 in TH1 cells³⁹. TBX21 plays an important role in the regulation of the immune response by promoting the development of the innate immune system and affecting the transport of innate immune cells and adaptive immune cells. In this study, FGFBP1, GATA3, KLRG1 and TBX21 were identified as candidate genes because these genes were not merely involved in the immune system but also exhibited significant changes in expression, as indicated by RNA-seq and confirmed by qPCR.

H₂S inhalation at 60 ppm to 100 ppm for 6 h was able to activate inflammatory reactions, induce oxidative stress through an increase in reactive oxygen species, and accordingly trigger improved antioxidation capability of the mice, especially when the H₂S concentration was close to 100 ppm. Furthermore, several genes related to oxidative stress, inflammation, and apoptosis, namely, FGFBP1, GATA3, KLRG1, and TBX21, were identified as candidate genes and played crucial roles in the response to H2S exposure. Moreover, these results suggest that inhalation of H₂S at concentrations lower than 100 ppm can also harm organisms; therefore, H₂S inhalation for therapeutic purposes should be practiced with caution.