Objective:To explore the effect of hydrogen sulfide (H 2S) on modulating the subunit Kir6.2 of adenosine triphosphate sensitive potassium channels via the cyclic guanosine monophosphate-dependent protein kinase (cGMP/PKG) signaling pathway in epileptic rat models. Methods:Sixty adult male SD rats were randomly divided into the following six groups (10 rats in each group) by random number table method: control, epileptic, H 2S donor, H 2S donor+epileptic, KT5823 (one inhibitor of the cyclic guanosine monophosphate-dependent protein kinase)+H 2S donor+epileptic, and glibenclamide (one inhibitor of the adenosine triphosphate sensitive potassium channels)+H 2S donor+epileptic groups. Except the control group, SD rats were intraperitoneally injected with plentylenetetrazole to make the kindling models and their behaviours were recorded including the latency period, the grade, and the duration of the first epileptic seizure according to the Racine′s standard. The waveforms of electroencephalogram (EEG) in hippocampus were also recorded during the seizure. The mRNA and protein levels of PKG and Kir6.2 in hippocampus were evaluated by Western blotting and quantitative real-time polymerase chain reaction, and the hippocampal concentrations of cGMP and phosphorylation of cyclic guanosine monophosphate-dependent protein kinase (p-PKG) were detected by enzyme linked immunosorbent assay. Results:Rats in the epileptic group showed Ⅳ-Ⅴ grade of epileptic seizure [4.500 (4.000, 4.875)], short latency period [(10.37±8.21) min] but long duration [(69.50±24.37) s] of seizure. Compared to the epileptic group, rats in the H 2S donor group showed Ⅱ-Ⅲ grade of epileptic seizure ( P=0.004), significantly longer latency period ( P<0.001), and shorter duration of seizure ( P<0.001). Compared to the H 2S donor+epileptic group, rats in the KT5823+H 2S donor+epileptic group showed Ⅲ-Ⅳ grade of epileptic seizures, significantly shorter latency period ( P<0.001), and longer duration of seizure ( P<0.001). The results of EEG showed that the wave patterns in the epileptic group were spike or sharp waves and the amplitudes were largest [(190.570±23.590) μV]. Compared with the epileptic group, amplitudes were reduced ( P<0.001) in the H 2S donor+epileptic group. PKG mRNA and PKG protein were expressed differently among all groups (PKG mRNA: n=5, H=26.714, P<0.001; PKG protein: n=5, F=30.597, P<0.001). Compared with the control group, the expression of both PKG mRNA and PKG protein was decreased (PKG mRNA: 1.000±0.001 vs 0.782±0.064, P=0.023; PKG protein: 0.550±0.037 vs 0.145±0.020, P=0.042) in the epileptic group. Besides, Kir6.2 mRNA and Kir6.2 protein were expressed differently among all groups (Kir6.2 mRNA: n=5, H=27.761, P<0.001; Kir6.2 protein: n=5, F=60.659, P<0.001). Compared with the control group, the expression of both Kir6.2 mRNA and Kir6.2 protein was decreased (Kir6.2 mRNA: 1.000±0.001 vs 0.897±0.033, P=0.004; Kir6.2 protein: 0.384±0.035 vs 0.215±0.016, P=0.024) in the epileptic group. And the concentrations of cGMP and p-PKG were decreased (cGMP: P<0.001; p-PKG: P<0.001) in the epileptic group. The results in the H 2S donor+epileptic group were up-regulated (PKG mRNA: P=0.047; PKG protein: P<0.001; Kir6.2 mRNA: P=0.011; Kir6.2 protein: P<0.001; cGMP: P<0.001; p-PKG: P<0.001) compared with the epileptic group. However, the results in the KT5823+H 2S donor+epileptic group were down-regulated (PKG mRNA: P=0.015; PKG protein: P=0.027; Kir6.2 mRNA: P=0.013; Kir6.2 protein: P=0.017; cGMP: P=0.005; p-PKG: P<0.001) compared with the H 2S donor+epileptic group. Conclusion:A possible mechanism is that H 2S prevents the epileptic seizure from modulating the subunit Kir6.2 of ATP sensitive potassium channels via the cGMP/PKG signaling pathway.