Reduced Levels of H2S in Diabetes-Associated Osteoarthritis Are Linked to Hyperglycaemia, Nrf-2/HO-1 Signalling Downregulation and Chondrocyte Dysfunction

Abstract

Different findings indicate that type 2 diabetes is an independent risk factor for osteoarthritis (OA). However, the mechanisms underlying the connection between both diseases remain unclear. Changes in the balance of hydrogen sulphide (H₂S) are thought to play an important role in the pathogenesis of diabetes and its complications, although its role is still controversial. In this study, we examined the modulation of H₂S levels in serum and chondrocytes from OA diabetic (DB) and non-diabetic (non-DB) patients and in cells under glucose stress, in order to elucidate whether impairment in H₂S-mediated signalling could participate in the onset of DB-related OA. Here, we identified a reduction in H₂S synthesis in the cartilage from OA-DB patients and in cells under glucose stress, which is associated with hyperglycaemia-mediated dysregulation of chondrocyte metabolism. In addition, our results indicate that H₂S is an inductor of the Nrf-2/HO-1 signalling pathway in cartilage, but is also a downstream target of Nrf-2 transcriptional activity. Thereby, impairment of the H₂S/Nrf-2 axis under glucose stress or DB triggers chondrocyte catabolic responses, favouring the disruption of cartilage homeostasis that characterizes OA pathology. Finally, our findings highlight the benefits of the use of exogeneous sources of H₂S in the treatment of DB-OA patients, and warrant future clinical studies.

Keywords: chondrocytes; heme oxygenase-1; hydrogen sulphide; inflammation; nuclear factor-erythroid 2-related factor-2; osteoarthritis; type 2 diabetes.

Figure 1

Serum H₂S levels in osteoarthritic patients with or without diabetes. Using an ion-selective microelectrode, (a) the levels of H₂S were measured in the sera from osteoarthritis patients without diabetes (non DB-OA, blue dots; n = 34) or with diabetes (DB-OA, green dots; n = 21). Additionally, the study population was divided into patients with normoglycaemia (n = 29) or hyperglycaemia (n = 14) (b) or individuals with normolipidaemia (n = 20) or dyslipidaemia (n = 35) background (c) and H₂S levels were analysed in these subsets of patients. Values are mean ± SEM. * p ≤ 0.05 analysed by the Mann–Whitney test.

Figure 2

Levels of H₂S synthesising enzymes in cartilage from osteoarthritic patients with or without diabetes. The protein expression of enzymes involved in H₂S biosynthesis, (a) cystathionine β-synthase (CBS), (b) cystathionine γ-lyase (CSE), and (c) 3-mercaptopyruvate sulphurtransferase (MPST), was evaluated by western blot in freshly isolated chondrocytes from OA patients without DB (non DB-OA, blue dots; n = 14) or with DB (DB-OA, green dots; n = 10). Additionally, the study population was divided into patients with normoglycaemia or hyperglycaemia and (d–f) levels of these enzymes were analysed in these different subsets of patients. Values were mean ± SEM. * p ≤ 0.05 analysed by the Mann–Whitney test.

Figure 3

HO-1 expression in cartilage from osteoarthritic patients with or without diabetes and its association with H₂S synthesising enzymes. (a) The protein expression HO-1 was evaluated by western blot in freshly isolated chondrocytes from OA patients without DB (non DB-OA, blue dots; n = 14) or with DB (DB-OA, green dots; n = 10). Additionally, samples were grouped into patients with normoglycaemia or hyperglycaemia (b) and levels of HO-1 were analysed in these different subsets of individuals. (c) Association between glucose levels, HO-1 and H₂S synthesising enzymes (p ≤ 0.05 shown in bold). (d) Graph representing Spearman correlation between HO-1 and CBS. Additionally, linear regression is shown. (e) Hierarchical clustering heatmap analysis was carried out to identify different OA phenotypes based on a specific profile of variables. Values are mean ± SEM. * p ≤ 0.05 analysed by the Mann–Whitney test. CBS, cystathionine β-synthase; CSE, cystathionine γ-lyase; HO-1, hemeoxygenase-1; MPST, 3-mercaptopyruvate sulphurtransferase.

Figure 4

IL-1β-stimulated expression of H₂S synthesising enzymes under high glucose environment. OA human chondrocytes were incubated in growth media with normal (NG) or high (HG) levels of glucose in the presence or absence of IL-1β. Thereafter, the gene expression of (a) CBS and (b) CSE was analysed. Values are mean ± SEM (n = 5). * p ≤ 0.05 analysed by the Wilcoxon test. CBS, cystathionine β-synthase; CSE, cystathionine γ-lyase.

Figure 5

Effect of exogenous administration of H₂S on pro-inflammatory and pro-oxidant responses to IL-1β and HO-1 expression recovery in chondrocytes exposed to a HG environment. OA human chondrocytes were incubated in growth media with normal (NG) or high (HG) levels of glucose in the presence or absence of IL-1β and co-treated with a fast or slow-releasing H₂S donor, NaSH 500 µM and GYY-4137 500 µM, respectively. Then, (a) protein levels of COX-2 were measured by western blot. Below panel shows representative images of COX-2 expression from one experiment. Expression levels of β-tubulin were employed as the loading control. (b) IL-6 released in the cell culture supernatant was measured by ELISA. (c) ROS production was monitored by flow cytometry using the fluorogenic dye 2′, dichlorofluorescein (DCF). Antimycin A (AA) was employed as a positive control of ROS production. (d) Protein HO-1 expression was also evaluated. Below panel shows representative images of HO-1 expression from one experiment. Values are mean ± SEM. * p ≤ 0.05 analysed by the Wilcoxon test. COX-2, cyclooxygenase 2; HO-1, hemeoxygenase-1; IL-6, interleukin 6.

Figure 6

siRNA-induced reduction in Nrf-2 expression modulates HO-1 and CBS levels and attenuates anti-inflammatory effects of exogenous administration of H₂S on IL-1β-stimulated chondrocytes under a HG environment. TC28a2 cells were transfected with the control or NRF-2 siRNA and then incubated in growth media with high levels of glucose in the presence or absence of IL-1β. Then, the gene expression of (a) Nrf-2, (b) HO-1, (c) CBS, and (d) CSE were analysed. (e) Representative images of one experiment evaluating protein expression of Nrf-2, HO-1, CBS, and CSE by western blot. Protein levels of tubulin were used as the loading control. (f) IL-6 production was analysed by ELISA. Basal condition in cells transfected with siRNA = 1. Values are mean ± SEM. * p ≤ 0.05 vs. respective condition in cells transfected with siRNA control, analysed by the Wilcoxon test. ^# p ≤ 0.05 vs. cells treated with IL-1β alone (n = 3). CBS, cystathionine β-synthase; CSE, cystathionine γ-lyase; HO-1, hemeoxygenase-1; IL-6, interleukin 6; Nrf-2, nuclear factor erythroid 2-related factor 2.

Figure 7

Expression of Nrf-2 and H₂S synthesising enzymes is diminished in the cartilage from diabetic mice. Histological slides of joints from lean or diabetic (db/db) mice were employed to analyse cartilage expression of (a) Nrf-2 and (b) H₂S synthesising enzymes CBS and CSE by immunohistochemistry. Images on the left show representative micrographs from one experiment (n = 7). Values are mean ± SEM. * p ≤ 0.05 CBS, cystathionine β-synthase; CSE, cystathionine γ-lyase; Nrf-2, nuclear factor erythroid 2-related factor 2.