Hydrogen sulfide inhalation induces changes in body physiology and lung gene expression in mice Authors: Jun Lei Gao, Bai Hao Li, Zhi Bin Ban, Li Jia Li, Hao Liang, Xiao Gang Yan, Cheng Xiao, Fang Yu Zhang Published: 2026 ABSTRACT AbstractHydrogen sulfide (H2S) inhalation injures the respiratory, immune, and nervous systems, altering homeostasis and health, but the molecular mechanisms have been still unclear. In this study, we tested the effects of different concentrations of H2S on immune index and enzyme activity of mouse tissues and measured differentially expressed genes (DEGs) in mouse lung tissues by transcriptome sequencing. The results showed that H2S significantly increased tumor necrosis factor-alpha (TNF-α) (p < 0.05), catalase (CAT) (p < 0.01) and glutathione peroxidase (GSH-Px) (p < 0.05) activity levels; 80 and 100 ppm H2S treatment significantly increased interleukin-1β(IL-1β), interleukin-10 (IL-10) (p < 0.01) and glutathione reductase activation coefficient (GRAC) (p < 0.01) but decreased malonaldehyde (MDA) and superoxide dismutase (SOD) levels (p < 0.05); 100 ppm H2S treatment significantly increased glutathione S-transferase (GST) and reactive oxygen species (ROS) levels (p < 0.01). In addition, 963 DEGs were identified in different group. These genes were notably involved in immune responses, DNA modification, enzymatic activity, and cell cycle processes. FGFBP1, GATA3, KLRG1 and TBX21 may be crucial roles in the response to H2S exposure. This study provides a reference and many differential genes for the study of hydrogen sulfide damage mechanism.Keywords: Hydrogen sulfide, Lung gene expression, MiceSubject terms: Biochemistry, Cell biology, Diseases, Immunology, Molecular biology Introduction Hydrogen sulfide (H2S) is a common industrial hazard originating from petroleum and gas industries, intensive animal farming, sewage, and waste treatment plants1,2. While endogenous H2S modulates various biological activities and signal transduction functions3, exogenous exposure primarily affects the central nervous, respiratory, and cardiovascular systems in a concentration- and duration-dependent manner4. Occupational exposure limits for H2S are typically set between 1 and 10 ppm5, yet concentrations in agricultural settings such as poultry houses and swine barns can range from 0 to 97 ppm6,7. Notably, exposure to 80–100 ppm H2S has been reported to exert both protective and deleterious effects depending on context8–10, suggesting that the biological impact of low-to-moderate H2S exposure is complex and remains incompletely understood. Despite these observations, existing studies have largely focused on individual endpoints, such as isolated immune markers or oxidative stress parameters, and have rarely integrated multiple physiological layers to elucidate the underlying molecular mechanisms. In particular, how H2S concentrations between 60 and 100 ppm simultaneously influence systemic immune responses, tissue redox balance, and pulmonary gene expression has not been systematically characterized. Moreover, while transcriptomic alterations following H2S exposure have been reported, the functional implications of these gene expression changes for lung tissue phenotypes remain largely unexplored. To address these gaps, we formulated the following hypothesis: inhalation of H2S at concentrations of 60–100 ppm induces concentration-dependent alterations in lung homeostasis through coordinated changes in inflammatory and redox pathways, with specific transcriptionally regulated genes serving as key mediators. 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. Materials and methods Animals and chambers 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) of the same age were obtained and randomly assigned to four treatment groups, with three biological replicates per group. Each biological replicate consisted of five mice housed together in one chamber, and 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% to 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 exposure 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. During the 6-hour exposure period, the H2S concentration inside each chamber was continuously monitored using an in-chamber gas detector (Model BH-90 A; detection range: 0–200 ppm; resolution: 1 ppm; ) calibrated prior to each experimental run. The detector probe was inserted through a dedicated airtight port on the chamber wall. Throughout the entire exposure period, the measured concentrations deviated by no more than ± 2% from the preset target values (60, 80, and 100 ppm), thus verifying the reliability of our dosing procedure. Immune indexes and redox-reaction enzymatic activity assays 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). All biochemical assays were performed in triplicate technical replicates, and the mean values were used for statistical analysis.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). RNA isolation and RNA-seq analysis 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. Oligo-attached 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. Raw data processing and alignment analysis 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. DEG analysis 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. Quantitative real-time Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses 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. Quantitative real-time quantitative PCR (qRT-PCR) validation of the RNA-seq results 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. Table 1Primer sequences used for quantitative real-time PCR.mRNAsOligoPrimer sequence (5’to3’)GapdhForward primerReverse primerTGTTTCCTCGTCCCGTAGAATCTCCACTTTGCCACTGCFgfbp1Forward primerReverse primerAGCCCAGGAAGGAGAAAGTGGTTGAGCACATCTGGAGata3Forward primerReverse primerCATGGGTTAGAGAGGCAGATAGAAGGGGTCGGAGGAKlrg1Forward primerReverse primerGGCTCTCTCATCCCTTCCGGTGTTTGCGTCTTTCTGTLoxForward primerReverse primerATAGGGCGGATGTCAGAGTGCTGTGGTAATGTTGGTGMfap4Forward primerReverse primerGGCCTTTCTGTGGCTGACTGAGGTGGGGTGCAAANlrp1aForward primerReverse primerGCTGGGCAGTTTACTTGGCACTTGACATGGCTGCTCNuf2Forward primerReverse primerCGATGTGCTTCCTGTGTGGTTGGGCCTGGGTAGAA Primer sequences used for quantitative real-time PCR. Forward primer Reverse primer TGTTTCCTCGTCCCGTAGA ATCTCCACTTTGCCACTGC Forward primer Reverse primer AGCCCAGGAAGGAGAAAG TGGTTGAGCACATCTGGA Forward primer Reverse primer CATGGGTTAGAGAGGCAGA TAGAAGGGGTCGGAGGA Forward primer Reverse primer GGCTCTCTCATCCCTTCC GGTGTTTGCGTCTTTCTGT Forward primer Reverse primer ATAGGGCGGATGTCAGAG TGCTGTGGTAATGTTGGTG Forward primer Reverse primer GGCCTTTCTGTGGCTGA CTGAGGTGGGGTGCAAA Forward primer Reverse primer GCTGGGCAGTTTACTTGG CACTTGACATGGCTGCTC Forward primer Reverse primer CGATGTGCTTCCTGTGTG GTTGGGCCTGGGTAGAA Statistical analysis For biochemical assays, three biological replicates were performed per treatment group, and each sample was measured in technical triplicates. The mean of the technical replicates was used as a single data point for statistical analysis. For RNA-seq, five biological replicates were analyzed per treatment group.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. Results Immune indexes 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). Table 2Effects of H2S on blood immune indexes and enzymatic activity levels of liver and kidney tissues.ItemsH2S inhalationp valueGroup A(0 ppm)Group B(60 ppm)Group C(80 ppm)Group D(100 ppm) Serum immune indexes TNF-α pg/ml333.79 ± 58.33c372.59 ± 43.06b503.79 ± 85.33a490.56 ± 93.44ab0.019IL-1β pg/ml176.61 ± 14.94b173.67 ± 9.24b200.23 ± 17.40a200.41 ± 14.82a0.011IL-8 pg/ml122.93 ± 28.3787.90 ± 27.38101.68 ± 27.02158.74 ± 45.080.211IL-10 pg/ml446.14 ± 39.51c461.76 ± 20.91bc514.91 ± 56.09b599.69 ± 53.78a< 0.001 Enzymatic activity levels of liver and kidney tissues MDA nmol/ml24.40 ± 3.77a23.01 ± 2.33ab18.69 ± 3.52c19.17 ± 2.47bc0.023ROS U/mg protein137.04 ± 18.61c218.83 ± 34.02b255.66 ± 27.48ab270.88 ± 53.06a< 0.001SOD U/mg protein283.91 ± 22.88a284.46 ± 27.63a228.24 ± 32.89b210.03 ± 56.86b0.010CAT U/mg protein14.52 ± 2.02b21.28 ± 2.66a21.22 ± 4.28a22.50 ± 2.68a0.003GSH-Px U/mg protein60.32 ± 14.75b97.66 ± 19.94a94.87 ± 9.28a101.10 ± 26.12a0.026GST U/mg protein326.40 ± 87.68b434.14 ± 98.48b419.83 ± 37.39b615.14 ± 107.92a0.009GRAC0.427 ± 0.076c0.564 ± 0.065b0.673 ± 0.115a0.699 ± 0.055a< 0.001a, b,c Means within a row with different superscripts differ (p < 0.05). Effects of H2S on blood immune indexes and enzymatic activity levels of liver and kidney tissues. a, b,c Means within a row with different superscripts differ (p < 0.05). Redox-reaction enzymatic activity levels 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). RNA-seq data analysis A total of 20 cDNA libraries were constructed, representing five biological replicates per treatment group. Each library was sequenced individually, and more than 6 GB of clean bases were obtained per 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. Table 3Alignment and quantification statistics in each sample in the RNA-seq library.SampleRaw readsClean readsQ30 (%)Clean reads (%)Mapping rate(%)A1429398144007149894.9193.3297.03A2433729444058966095.0193.5896.79A3435123704061525895.2293.3496.98A4473990984420836694.7793.2696.91A5514518504810735294.9193.4996.84B1537875145039135294.9193.6896.94B2500503184688556895.4193.6796.46B3444249504165107095.1493.7596.89B4502312104708629694.4793.7396.81B5490535944595628495.2593.6896.89C1541370185067729095.1093.6096.86C2510882784774217295.3693.4596.83C3491731184578348295.2593.1097.05C4552593405176750295.1693.6896.84C5557842805226939095.0093.6996.88D1486483984554809894.8393.6296.83D2434666044068897895.0893.696.65D3468650044397919495.1193.8496.72D4498592324668437095.0793.6396.97D5462469064330202494.8893.6396.85 Alignment and quantification statistics in each sample in the RNA-seq library. Differential expression analysis of the RNA-seq data 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 Tables S1-S6. We uploaded the raw sequencing data on the SRA (Sequence Read Archive) accession number SUB16200134. Expression profiles of distinct RNAs. (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. Gene Ontology and Kyoto Encyclopedia of genes and genomes pathway analyses 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). Significantly enriched GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) genes. (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. 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 Table S7-S18. Validation of the DEGs by qPCR 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 1 A, 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. Validation of mRNA expression from the A to D groups by RT-qPCR. mRNA expression was quantified relative to the Gapdh expression level by using the comparative cycle threshold (△CT) method. Discussion Immune indexes In livestock and poultry breeding, H2S primarily originates from animal manure and microbial decomposition processes11. 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 redox12–15. 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 response16. 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 weeks17. IL-8 was the first chemokine to be discovered, exhibiting pro-inflammatory activity18. 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 effects19. Redox-reaction enzymatic activities 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 peroxides11,20. MDA is an indicator of lipid peroxidation and the peroxidation of biomembranes and lipoproteins21, but ROS are metabolites of oxygen and primarily include superoxide anions (O2-), hydroxyl radicals (OH-) and hydrogen peroxide (H2O2)22. 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)23. 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 metabolism24. 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)25. Glutathione reductase (GR) subsequently reduces GSSG to GSH, and GST catalyzes the reaction of GSH conjugation to a wide range of electrophilic metabolites of xenobiotics26. 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 GSH9. 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. Gene expression in lung tissue 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-127. IL-1 can bind with IL-1 receptor-related protein 2 (IL-1 Rrp2) to activate the NF-κB signaling pathway28. Deficiency in the mouse immune gene IL1F9 was impaired and allowed an inflammatory reaction to be induced by a virus29. 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 proliferation30. 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 cells31. Several studies have shown that GATA3 binds to the promoter/enhancer region of the Il7r gene in CD8 + T cells to maintain immune function32,33. 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 system34,35. 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 cells36. In the absence of TBX21, the function of NK cells was impaired, and the survival rate was decreased37. 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 cells38. 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. Nrf2 (nuclear factor erythroid 2-related factor 2) and its negative regulator Keap1 constitute a core antioxidant defense pathway. In models of cadmium-induced hepatotoxicity, chromium-induced nephrotoxicity, and PFOS-induced cardiotoxicity, Nrf2 expression was significantly downregulated while Keap1 expression was upregulated, leading to decreased activities of downstream antioxidant enzymes such as HO-1, GST, GSR, and GPx39. In the present study, similar changes in antioxidant enzyme activities (e.g., decreased SOD, increased CAT and GPx) were observed after H2S exposure, suggesting that the Nrf2/Keap1 pathway may be involved in the H2S-induced oxidative stress response, warranting further investigation. Furthermore, the TLR4/MyD88 signaling pathway was significantly activated in chromium-induced nephrotoxicity, accompanied by upregulation of NF-κB, TNF-α, IL-1β, IL-6, and COX-2, indicating its central role in inflammatory responses40. In our study, significant changes in TNF-α, IL-1β, and IL-10 levels were also observed after H2S exposure, implying that the TLR4/MyD88/NF-κB axis may also play a role in H2S-induced pulmonary inflammation.Regarding apoptosis, Bax, Bcl-2, and Caspase-3 are key regulatory genes in multiple toxicological models. Exposure to cadmium, chromium, PFOS, and trans fatty acids (e.g., elaidic acid) all led to upregulation of Bax and Caspase-3 and downregulation of Bcl-2, thereby promoting apoptosis41,42. Although we did not directly measure the expression of these apoptotic genes in the present study, we observed GO enrichment related to DNA modification and cell cycle processes in lung tissues after H2S exposure, suggesting that H2S may mediate lung injury by affecting cell survival and apoptosis balance. Finally, the HMGB1/RAGE signaling axis was significantly activated in chromium-induced renal injury, promoting the release of inflammatory cytokines40. HMGB1, as a damage-associated molecular pattern (DAMP) molecule, may also play a similar role in H2S-induced lung injury, a possibility worth exploring in future studies. Linking gene expression changes to potential lung phenotypes Beyond identifying differentially expressed genes, we further explored their potential functional consequences in lung tissue. For instance, FGFBP1, which was significantly upregulated, enhances fibroblast growth factor signaling and has been implicated in epithelial proliferation and repair43. Its sustained elevation following H2S exposure may contribute to aberrant airway epithelial remodeling or fibrotic changes, although this requires direct histological validation. Similarly, the coordinated changes in GATA3 and TBX21 master regulators of T-cell differentiation suggest a possible shift in the Th1/Th2 balance within the pulmonary microenvironment, which could influence the nature and chronicity of the inflammatory response. KLRG1, a marker of terminally differentiated NK and T cells, was also altered, raising the possibility of impaired immune surveillance in the lung following H2S exposure. In line with these observations, our KEGG pathway enrichment analyses identified immune-related pathways and cell cycle regulation as top enriched terms, further supporting the notion that H2S may disrupt lung immune homeostasis and cellular turnover. Although direct phenotypic assessments such as lung histopathology or functional assays were not performed in this study, the gene expression patterns we observed provide testable hypotheses for future investigations. Conclusion 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 indicate that inhalation of hydrogen sulfide (H2S) at concentrations below 100 ppm can still cause damage to organisms and trigger molecular alterations in the body, which may in turn exert potential impacts on lung structure and immune function. Together, the inhalation of H2S for therapeutic purposes should be conducted with extreme caution. Author contributions Jun Lei Gao: Writing – review & editing, Writing – original draft, Visualisation, Validation, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Con- ceptualisation. Bai Hao Li: Writing – review & editing, Writing – original draft, Visualisation, Validation, Resources, Pro- ject administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualisation. Fang Yu Zhang: Writing – review & editing, Writing – original draft, Visualisation, Validation, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Con- ceptualisation. Cheng Xiao: Writing – review & editing, Writing – original draft, Visualisation, Validation, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Con- ceptualisation. Zhi Bin Ban: Writing – review & editing, Methodology, Investigation, Formal analysis, Data cura- tion, Conceptualisation. Li Jia Li: Writing – review & editing, Supervision, Methodology, Investigation, Conceptualisation. Hao Liang: Writing – review & editing, Supervision, Methodology, Investigation, Conceptualisation. Xiao Gang Yan: Writing. Funding This study was supported by the National Modern Industrial Technology System for Wool Sheep (project no.CARS-38-18),Research and Industrialization Demonstration Application of Key Technologies for Feed Utilization of Functional Traditional Chinese Medicine Residues (20250202009NC), Research and Application of High-efficiency Utilization Technology for Corn Processing By-product Feed (2025ZDYFNS04). Data availability We uploaded the raw sequencing data to the SRA (Sequence Read Archive) under the accession number SUB16200134. Declarations Competing interests The authors declare no competing interests. Institutional review board statement All operations in this study followed the rules concerning animal experiments and biodiversity rights as stated in the Principles of Regulations on Administration of Experimental Animals of the Chinese government (Revised Edition, 2017), and the study protocol was approved by the Animal Ethics Committee of the Jilin Academy of Agricultural Sciences (AWEC 2019A05, May 16th, 2019). Statement on ARRIVE guidelines The study is reported in accordance with ARRIVE guidelines 2.0. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Jun Lei Gao and Bai Hao Li contributed equally to this work. Contributor Information Cheng Xiao, Email: 1174114261@qq.com. Fang Yu Zhang, Email: zhangfangyu610@163.com. References Associated Data Data Availability Statement We uploaded the raw sequencing data to the SRA (Sequence Read Archive) under the accession number SUB16200134.