Isoflurane postconditioning (IsoPostC) attenuates myocardial ischaemia/reperfusion injury (IRI). Signal transducer and activator of transcription-3 (STAT3) is critical in ischaemic postconditioning cardioprotection, which can be regulated by the Brahma-related gene (Brg1) and nuclear factor-erythroid 2-related factor 2 (Nrf2), although they are both reduced in diabetic hearts. We hypothesized that reduced Brg1/Nrf2 and STAT3 activation may jeopardize IsoPostC-mediated cardioprotection in diabetic hearts. In the present study, Langendorff-perfused, non-diabetic (control) and 8-week-old streptozotocin-induced Type 1 diabetic rat hearts were subjected to 30 min of global ischaemia and 120 min of reperfusion without or with IsoPostC, which was achieved by administering emulsified isoflurane (2.0%, v/v) in Krebs–Henseleit (KH) solution immediately at the onset of reperfusion for 10 min and switching to KH solution perfusion alone thereafter. Cultured H9C2 cells were exposed to normal glucose (NG, 5.5 mM) or high glucose (HG, 30 mM) and subjected to hypoxia/reoxygenation (HR) in the presence or absence of IsoPostC. Diabetic rats displayed larger post-ischaemic myocardial infarction and more severe haemodynamic dysfunction, associated with increased myocardial oxidative stress and reduced cardiac Brg1, Nrf2 and STAT3 phosphorylation/activation (p-STAT3), compared with controls. These changes were reversed/prevented by IsoPostC in control but not in diabetic rats. In H9C2 cells exposed to NG but not HG, IsoPostC significantly attenuated HR-induced cellular injury and superoxide anion production with increased Brg1, Nrf2 and p-STAT3. These beneficial effects of IsoPostC were abolished by Brg1, Nrf2 or STAT3 gene knockdown. Brg1 or Nrf2 gene knockdown abolished IsoPostC-induced STAT3 activation. N-acetylcysteine restored Brg1, Nrf2 and p-STAT3, and IsoPostC-induced protection in H9C2 cells exposed to HG and HR. In conclusion, IsoPostC confers cardioprotection through Brg1/Nrf2/STAT3 signalling, and impairment of this pathway may be responsible for the loss of IsoPostC cardioprotection in diabetes.
The volatile anaesthetic isoflurane, applied immediately at the onset of reperfusion (IsoPostC), confers cardioprotective effects in non-diabetes but loses its effectiveness in diabetes, and the reason for this is unclear. The present study was designed to investigate the mechanism underlying the reduced cardioprotective effect of IsoPostC in diabetes.
STAT3 activation plays a critical role in IsoPostC cardioprotection. IsoPostC, by increasing Brg1 and Nrf2, activates STAT3, which subsequently reduces myocardial IRI-induced oxidative stress and eventually improves post-ischaemic cardiac functional recovery.
In diabetes, STAT3 is reduced in association with decreased Brg1 and Nrf2. This impairment of Brg1/Nrf2/STAT3 signalling is responsible for the loss of cardioprotection of IsoPostC in diabetes. Thus, effective means such as the antioxidant NAC which can increase Brg1 in diabetes may facilitate IsoPostC cardioprotection by activating STAT3 in diabetic hearts.
Ischaemic heart disease is the leading cause of death in patients with diabetes . Reperfusion restores coronary flow, but reperfusion itself may cause lethal tissue injury called ischaemia/reperfusion injury (IRI). The volatile anaesthetics such as isoflurane, when applied immediately at the onset of reperfusion, can reduce myocardial infarct size (IS)  and this phenomenon is termed ‘isoflurane postconditioning’ (IsoPostC) . However, the cardioprotective effects of volatile anaesthetic postconditioning are attenuated or abolished in rabbits  and mice  with hyperglycaemia or in individuals with diabetes , and the mechanism is unclear.
IsoPostC confers cardioprotection against myocardial IRI by inhibiting apoptosis via modulation of pro- and anti-apoptotic proteins  and preservation of mitochondrial function [8,9], which requires the activation of signal transducer and activator of transcription-3 (STAT3) . However, in diabetes, myocardial STAT3 activation is reduced , suggesting that reduced STAT3 may render diabetic hearts more sensitive to IRI, and less responsive or unresponsive to IsoPostC.
The activation of STAT3 is dependent on the Brahma-related gene 1 (Brg1) [11,12], a chromatin-remodelling enzyme that regulates the chromatin accessibility at STAT3-binding targets [13,14]. Brg1 also facilitates nuclear factor-erythroid 2-related factor 2 (Nrf2) to induce haem oxygenase-1 (HO-1) in response to oxidative stress . These, together with our finding that STAT3 functions as the downstream target of HO-1 in combating hyperglycaemia-induced myocardial dysfunction , suggest that Brg1/Nrf2 may play an important role in IsoPostC-mediated STAT3 activation and IsoPostC-mediated cardioprotection. The fact that myocardial Brg1 and Nrf2 are reduced in diabetes [17,18] prompts us to hypothesize that reduction in myocardial Brg1 may undermine STAT3 activation in diabetes and render diabetic hearts less responsive or unresponsive to IsoPostC cardioprotection.
Therefore, in the present study, we integrate an ex vivo ischaemia/reperfusion (IR) model of Langendorff-perfused heart and an in vitro cell hypoxic reoxygenation model to explore the role of Brg1 and Nrf2 in facilitating IsoPostC-mediated STAT3 activation and cardioprotection under normal and diabetic conditions, and to determine whether up-regulation of Brg1/Nrf2 may restore IsoPostC cardioprotection in diabetes.
MATERIALS AND METHODS
Induction of diabetes
Male Sprague–Dawley rats (250±10 g, aged approximately 6 weeks), obtained from the Laboratory Animal Service Centre, University of Hong Kong, were housed and given free access to standard food and water, and the experiments were approved by the Committee on Use of Live Animals in Teaching and Research (CULATR no. 2972-13). Diabetes was induced by a single tail-vein injection of streptozotocin (STZ) at a dose of 60 mg/kg in 0.1 M citrate buffer (pH 4.5) or citrate buffer alone as control. Blood glucose levels were monitored 72 h after STZ injection by using a One Touch Ultra Glucose meter (Life Scan Inc.). Rats with blood glucose levels >15 mM were considered to be diabetic. Water intake and food consumption were recorded daily, whereas plasma glucose (detected after 6 h of fasting) and body weight were monitored weekly. At termination (8 weeks after diabetic induction), rats were weighed and subjected to IR as described below. Before IR, plasma was extracted from blood samples and stored at −80°C until assay.
Measurement of cardiac and plasma level of free 15-F2t-isoprostane
Measurement of cardiac level of HO-1 activity
HO-1 activity was measured by a commercial assay kit (Enzo Life Sciences) as we have described in . Homogenized heart tissues were purified using an Affinity Sorbent and Affinity Column (Cayman Chemical) and then processed for analysis. HO-1 activity was expressed in ng per mg of protein.
Isolated rat hearts were Langendorff-perfused as previously described . Briefly, after being anaesthetized by the injection of intraperitoneal pentobarbital sodium (65 mg/kg), thoracotomy was performed for quick excision of the hearts and immersion in ice-cold Krebs–Henseleit (KH) solution. The perfusate was equilibrated with a mixture of 95% O2 and 5% CO2, and maintained at 37°C using a thermostatically controlled water-circulating system. The hearts were kept warm in a chamber with circulating water controlled at 37°C. A water-filled balloon was then inserted into the left ventricle (LV) through the mitral valve and connected to a pressure transducer for continuous measurement of LV pressure. After 10 min of equilibration, the hearts were subjected to 30 min of global ischaemia by turning off the perfusate, followed by 120 min of reperfusion with KH solution at a constant flow rate of 10 ml/min. The LV pressure signal, left ventricular end-diastolic pressure (LVEDP), the peak rate of pressure increase (dP/dtmax) and the peak rate of pressure decrease (dP/dtmin) were recorded using a real-time data acquisition and analysis system (PowerLab data acquisition system, Chart 5 Recorder Software, ADInstruments Inc.). At the end of reperfusion, hearts were removed and immediately frozen in liquid nitrogen and stored at −80°C until analysed or processed for myocardial IS measurement as described below.
IsoPostC model in rats
The experimental design is illustrated in Supplementary Figure S1(A). Non-diabetic (control or C) and diabetic (D) rats (n=6 per group) were randomly allocated to the sham-operated group (sham), the IR-treated group (IR) and the isoflurane postconditioning group (IsoPostC+IR). IsoPostC was induced by administering emulsified isoflurane in KH solution immediately at the onset of reperfusion for 10 min at 2.0% (v/v) concentration (this isoflurane concentration was chosen based on a previous study  and our preliminary study), and switched to KH solution perfusion only thereafter. For 118 successfully completed studies 130 rats were used. Two rats were excluded due to the failure in surgical procedures and ten because intractable ventricular fibrillation occurred during coronary artery occlusion.
Determination of myocardial IS
Measurement of creatinine kinase-MB
Assessment of myocardial apoptosis
Myocardial apoptosis was assessed by measuring cleaved caspase 3/total caspase 3 ratio and cytosolic cytochrome c release using Western blotting.
Adult rat cardiomyocyte isolation and hypoxia/reoxygenation (HR)
An additional series of experiments in vitro (see Figure 2) were designed for further confirmation of the findings from the ex vivo study. Isolated cardiomyocytes from adult C and D rat ventricles were prepared as described in our previous study , which showed that the ROS level and oxidative stress were significantly increased in cultured myocytes isolated from D rats compared with those from C rats. Confluent beating cardiomyocytes were subjected to hypoxia for 30 min and reoxygenated for 120 min. Hypoxia was obtained by equilibrating a humidified glass chamber containing myocytes with 95% N2 and 5% CO2 via a deoxygenator. Reoxygenation was achieved by exposing cells to room air (CO2 incubator). IsoPostC was induced by giving 0.2% emulsified isoflurane for 30 min at the onset of reoxygenation (this concentration of isoflurane was chosen based on a previous study  and our preliminary study). Each experiment was performed independently six times.
Impact of Brg1, Nrf2 or STAT3 on post-hypoxic cellular injury and IsoPostC protection in H9C2 cells
To confirm the role of Brg1/Nrf2/STAT3 in IsoPostC cardioprotection, gene silencing was performed using Brg1 siRNA, Nrf2 siRNA, STAT3 siRNA, control siRNA and transfection reagent (Santa Cruz Biotechnology) as described in  in H9C2 cells (purchased from the Chinese Academy of Science Cell Band) under normal glucose (NG, 5 mM). Efficiency of gene knockdown has been confirmed using Western blotting (see Supplementary Figures S2A–S2C). At 2 days after transfection, H9C2 cells were challenged with 30 min of hypoxia and 120 min of reoxygenation. N-acetylcysteine (NAC) treatment was started 24 h before transfection (i.e. total NAC treatment time was 72 h before hypoxia), in order to confirm the role of high glucose (HG)-induced oxidative stress in the loss of IsoPostC protection. Glucose concentration was chosen based on our previous studies [25,26] showing that it is sufficient to induce oxidative stress and cell apoptosis, the main pathogenesis of diabetes-induced cell injury in vivo. Similar to what was reported in our previous study , our preliminary data showed that osmolarity at the level equivalent to that induced by the HG we used had no significant impact on cell death and apoptosis (results not shown). Cells were incubated in HG and treated with NAC for a total of 72 h before being subjected to hypoxia and IsoPostC. The cells were incubated with HG and NAC for 24 h before being transfected with the respective siRNA, and last kept in the presence of HG and NAC for 48 h; then the cells were subjected to hypoxia and IsoPostC. Each experiment was performed independently six times.
Superoxide anion production assessment by DHE
Dihydroethidium (DHE) binds to DNA and produces fluorescent ethidium bromide when oxidized by the superoxide anion. At the end of reoxygenation, cells were incubated with 7.5 μM DHE at 37°C for 30 min and a fluorescence microscope (BX41 System microscope, Olympus) was used to detect positive-staining nuclei captured using a DP72 digital camera. The fluorescence images (magnification ×400) were calculated in each of five randomly selected fields and expressed as a percentage of the DHE-stained positive nuclei compared with control by a quantitative morphometric method.
Determination of post-hypoxic cell injury by measuring the release of LDH and cytochrome
c, and cleaved caspase 3 protein expression
Cell injury was assessed by quantification of lactate dehydrogenase (LDH) in the medium using a cytotoxicity assay kit (Roche Diagnostics) as described in [19,20]. In addition cell apoptosis was assessed by measuring the cleaved caspase 3/total caspase 3 ratio and cytosolic cytochrome c release, using Western blotting as described in our previous studies using the same experimental model, which showed that an increase in the cleaved caspase 3/total caspase 3 ratio corresponded to the increase in apoptotic cell death detected using a terminal deoxynucleotidyl-transferase dUTP nick-end labelling (TUNEL) assay [16,18].
Protein extraction and immunoblotting
Equal protein amounts from rat LV tissues, isolated cardiomyocytes and H9C2 cell homogenates were resolved using SDS/7.5–12.5% PAGE, subsequently transferred to polyvinylidene fluoride (PVDF) membranes and then Western blotting as described previously [28,29]. The protein expression of p-STAT3 (Tyr705), STAT3, Nrf2, Brg1, cleaved caspase 3, total caspase 3, cytosolic cytochrome c and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; Cell Signaling Technology) was analysed by immunoblotting.
Values are presented as means±S.E.M. Intergroup comparisons of dependent variables were performed using ANOVA with Tukey's test to analyse multiple comparisons of group means. Differences were considered to be statistically significant when P<0.05.
General characteristics and the plasma 15-F2t-isoprostane level in STZ-induced diabetic rats
As shown in Supplementary Table S1, 8 weeks after diabetic induction, daily water and food intake were significantly increased in D rats compared with C rats (both P<0.05). Body weight gain and heart weight in D rats were reduced, whereas the heart weight/body weight ratio was significantly increased in D rats compared with that in C rats. Plasma glucose levels were significantly increased in D rats compared with those in C rats. The levels of plasma free 15-F2t-isoprostane in D rats were higher than those in C rats (P<0.05).
IsoPostC attenuated post-ischaemic myocardial injury and oxidative stress that were associated with enhanced cardiac Brg1, Nrf2 and p-STAT3 in normal but not in diabetic rats
As shown in Figure 1, in C rats, post-ischaemic myocardial injury was increased and manifested as increased myocardial IS, effluent CK-MB, cytosolic cytochrome c release and cleaved caspase 3 protein expression, which was associated with elevated myocardial oxidative stress, evidenced by increased 15-F2t-isoprostane in the IR compared with the sham group (Figures 1A–1E). All of these changes were attenuated by IsoPostC which was associated with increased cardiac Brg1, Nrf2 and p-STAT3 (Tyr705) protein expression (P<0.05 vs C+IR, Figures 1G–1I). HO-1 activity was significantly increased after myocardial IR and was further enhanced by IsoPostC in C rats (Figure 1F). Myocardial IS, CK-MB, cytosolic cytochrome c release and cleaved caspase 3 protein expression were severe in D rats subjected to IR, which was associated with increased myocardial oxidative stress evidenced by increased 15-F2t-isoprostane and reduced HO-1 activity as well as lower Brg1, Nrf2 and p-STAT3 (Tyr705) protein expression (P<0.05 vs C+IR and C+IsoPostC+IR). Furthermore, diabetes cancelled out IsoPostC-mediated myocardial protection and enhancement of myocardial Brg1, Nrf2 and p-STAT3 (Tyr705) protein expression (P>0.05 vs D+IR).
Effects of IsoPostC on myocardial injury and myocardial oxidative stress in non-diabetic (C) and STZ-induced diabetic (D) rats subjected to IR
IsoPostC attenuated myocardial dysfunction in normal but not in diabetic rats
As shown in Supplementary Figure S3, dP/dtmax decreased progressively in all groups, accompanied by concomitant and continuous increases in dP/dtmin and LVEDP. IsoPostC attenuated all of these changes in C but not in D rats. Overall, the value of the area under the curve (AUC) for dP/dtmax in C rats was higher in the IsoPostC+IR than in the IR group, whereas the AUC for dP/dtmax in D rats was markedly lower than that in the C+IR and C+IsoPostC+IR groups (P<0.05, see Supplementary Figure S3B). Similarly, the AUC for dP/dtmin in the IsoPostC+IR group was higher than in the IR group in C rats and diabetes abolished IsoPostC-mediated enhancement of dP/dtmin (see Supplementary Figure S3D). The AUC value of LVEDP in the IsoPostC+IR group was lower than that in the IR group in C rats, and the values for LVEDP in D rats with and without IsoPostC were similarly dramatically increased compared with those in C rats (see Supplementary Figure S3F).
IsoPostC reduced post-hypoxic cell death and increased Brg1, Nrf2 and p-STAT3 (Tyr705) protein expression in cardiomyocytes isolated from normal but not diabetic rats
As shown in Figure 2, in primary cardiomyocytes isolated from C rats, HR significantly increased cardiomyocyte cellular injury assessed as elevation in LDH release, cleaved caspase 3/total caspase 3 ratio and cytosolic cytochrome c release (P<0.05 vs normoxia control group, Figures 2A–2C), which were remarkably reduced by IsoPostC that was associated with enhanced Brg1, Nrf2 and p-STAT3 (Tyr705) protein expression (P<0.05 vs C+HR, Figures 2D–2F). Compared with cardiomyocytes isolated from C rats, those isolated from D rats showed decreases in Brg1, Nrf2 and p-STAT3 (Tyr705), accompanied by increased cellular injury, which manifested as increases in LDH, cleaved caspase 3/total caspase 3 ratio and cytosolic cytochrome c release in both D+HR and D+IsoPostC+HR groups (P<0.05 vs C+HR and C+IsoPostC+HR). There was no significant difference between HR and IsoPostC+HR groups in cardiomyocytes isolated from D rats (P>0.05 vs D+HR).
Effects of IsoPostC in cardiomyocytes isolated from C and D rats subjected to HR
IsoPostC attenuated post-hypoxic cardiomyocyte apoptosis and decreased superoxide anion production in H9C2 cells exposed to NG which were abolished by
Brg1, Nrf2 or STAT3 gene knockdown
As shown in Figures 3–5, in H9C2 cells under NG conditions, HR significantly increased cell death and apoptosis as evidenced by increased LDH release (Figures 3, 4 and 5A) and cytosolic cytochrome c release (Figures 3, 4 and 5B), compared with untreated controls. This was concomitant with increased superoxide anion production (Figures 3, 4 and 5C), and these changes were prevented by IsoPostC. Brg1, Nrf2 and STAT3 gene knockdown by siRNAs respectively abolished IsoPostC-mediated reductions in LDH and cytosolic cytochrome c release, and reversed IsoPostC-mediated reduction in DHE staining. Brg1 or Nrf2 gene knockdown in the presence of IsoPostC significantly reduced p-STAT3 (Tyr705) protein expression. However, in the presence of IsoPostC, STAT3 gene knockdown had no effect on the protein expression of Brg1 and Nrf2 (P>0.05 vs IsoPostC+HR, Figures 5D and 5E).
Brg1 gene knockdown on IsoPostC protection against HR under NG conditions
Nrf2 gene knockdown on IsoPostC protection against HR under NG conditions
STAT3 gene knockdown on IsoPostC protection against HR under NG conditions
IsoPostC cellular protection was lost in H9C2 cells exposed to HG and can be restored by the antioxidant NAC
Cell death that manifested as LDH release was significantly increased and associated with reduced Brg1, Nrf2 and STAT3 activation in H9C2 cells incubated with HG compared with cells incubated with NG (see Supplementary Figure S4). In cells exposed to HG, no significant difference was observed in LDH release and cytosolic cytochrome c release, as well as in superoxide anion generation between HR and IsoPostC+HR groups (P>0.05; Figures 6A–6C). NAC significantly attenuated HR-induced cell injury, reflected as reductions in post-hypoxic LDH, cytosolic cytochrome c release and superoxide anion generation (P<0.05 vs HG+HR, Figures 6A–6C), with concomitant enhancement of Brg1, Nrf2 and p-STAT3 (Tyr705) protein expression (P<0.05 vs HG+HR, Figures 6D–6F), and all of these changes were abolished by Brg1 gene knockdown. NAC partially restored the sensitivity of cardiomyocytes to IsoPostC cellular protection under HG accompanied by increased Brg1, Nrf2 and p-STAT3 (Tyr705) protein expression (P<0.05 vs HG+IsoPostC+HR). However, Brg1 gene knockdown cancelled NAC-mediated restoration of IsoPostC cellular protection in cells exposed to HG (P<0.05 vs HG+NAC+IsoPostC+HR).
Effect of NAC treatment on restoring IsoPostC cellular protection under HG conditions
In the present study, we have demonstrated that, in normal non-diabetic rats, IsoPostC can reduce post-ischaemic myocardial injury manifested as reduced IS and CK-MB, and ameliorate cardiac dysfunction. Consistently, in H9C2 cells incubated with NG, IsoPostC can decrease post-hypoxic LDH release, attenuate superoxide anion generation and reduce cytosolic cytochrome c release. Moreover, we found that p-STAT3 was significantly increased, accompanied by enhanced myocardial Brg1 and Nrf2 subsequent to IsoPostC, whereas gene knockdown of Brg1 or Nrf2 not only diminished IsoPostC-induced STAT3 activation but also cancelled the above-mentioned cardioprotective effects of IsoPostC. This indicates that Brg1/Nrf2 signalling plays a critical role in IsoPostC-mediated STAT3 activation and the consequent cardioprotection. However, in D rats, the cardioprotection of IsoPostC was diminished and this was associated with reduced Brg1, Nrf2 and p-STAT3. Antioxidant treatment with NAC restored IsoPostC cardioprotection in diabetes which was associated with increased Brg1. These beneficial effects of NAC were cancelled by Brg1 gene knockdown, suggesting that the fundamental mechanism responsible for the loss of cardioprotection of IsoPostC in hearts from D rats might be inactivation of STAT3 due to impaired Brg1/Nrf2 (see Supplementary Figure S5). Taken together, the present study, for the first time, points out the importance of Brg1/Nrf2 in facilitating IsoPostC-mediated STAT3 activation and cardioprotection, and that the impairment of Brg1/Nrf2 signalling in diabetes may represent a mechanism governing the loss of IsoPostC myocardial protection in diabetes.
Increasing evidence demonstrates that both the cardioprotective reperfusion injury salvage kinase (RISK) signalling pathway [i.e. phosphoinositide 3-kinase (PI3K)/protein kinase B (Akt)], and especially the more recently identified survivor activating factor enhancement (SAFE) pathway [i.e. Janus kinase (JAK)/STAT3] play critical roles in the cardioprotection afforded by IsoPostC [30,31]. It is of interest that Goodman et al.  reported, during ischaemic postconditioning, that JAK/STAT3 worked upstream of PI3K/Akt, and JAK/STAT3 activation alone was insufficient to provide cardioprotection in ischaemic postconditioning without subsequent PI3K/Akt activation. Similarly, Tian et al.  demonstrated that, during ischaemic postconditioning, JAK/STAT3 signalling regulated the activation of the PI3K/Akt pathway and that blockage of PI3K/Akt by wortmannin partially attenuated but did not abolish ischaemic postconditioning cardioprotection. These results not only indicate that cross-talk exists between the RISK (i.e. PI3K/Akt) and the SAFE (i.e. JAK/STAT3) pathways, but also provide evidence that JAK/STAT3 may play an essential role in ischaemic postconditioning cardioprotection. In other words, the SAFE pathway can function independently of the RISK pathway to confer cardioprotection during postconditioning. Findings of the present study that STAT3 gene knockdown abolished IsoPostC cardioprotection highlight the importance of STAT3 activation in IsoPostC cardioprotection.
STAT3 is activated via the tyrosine phosphorylation cascade and then translocates into the nucleus and binds to specific target genes. Brg1, a major catalytic subunit of several chromatin-remodelling complexes, is essential for the promotion of STAT3 transcription . However, very little is known about the role of Brg1 in myocardial IRI and, in particular, its role in heart responsiveness to IsoPostC-induced cardioprotection. To gain insight into the relationship between Brg1 and STAT3 in IsoPostC cardioprotection, we performed siRNA transfection to knock down Brg1 and STAT3 genes in cultured cardiomyocytes with normal concentrations of glucose, and found that knockdown of these genes respectively abolished IsoPostC cardioprotection. More importantly, STAT3 gene knockdown in the presence of IsoPostC did not have a significant impact on Brg1 expression, whereas Brg1 gene knockdown adversely reduced p-STAT3 expression, indicating that Brg1 might regulate STAT3 activation in the context of IsoPostC cardioprotection. Brg1-mediated modification of chromatin architecture requires transcriptional regulation through the recruitment of its essential transcription factors . Multiple studies have provided evidence that Brg1 can regulate gene expression through interaction with specific transcription factors in a given signalling pathway, which is dependent on chromatin remodelling. In the presence of oxidative stress, Brg1 gene knockdown can selectively reduce Nrf2-targeted gene expression of HO-1 , a cytoprotective defence enzyme and an upstream effector of STAT3 . It is interesting that, in the present study, gene knockdown of Brg1 reduced the IsoPostC-induced increase of Nrf2 protein expression under NG conditions and cancelled IsoPostC cardioprotection. Similarly, under HG conditions, NAC in combination with IsoPostC increased Brg1 which was associated with increased Nrf2 protein expression, but this increase was diminished by Brg1 gene knockdown, suggesting that, in the context of IsoPostC, Brg1 not only facilitates Nrf2-induced induction of STAT3, but also positively regulates Nrf2. Unexpectedly, in the present study, Nrf2 gene knockdown abolished IsoPostC-induced up-regulation of Brg1; one possible exploration for this interesting result may be that knockdown of Nrf2 reduced the IsoPostC-induced increase of cell antioxidant capacity, which increased oxidative stress, and this subsequently down-regulated Brg1; this is similar to our previous finding that diabetes-induced oxidative stress down-regulated Brg1 and could be reversed by antioxidant treatment with NAC . However, whether Nrf2 has a direct regulatory effect on Brg1 needs further investigation. Collectively, findings from the present study and others provide evidence that Brg1 may functionally interact with Nrf2 on oxidant-induced stress and subsequently participate in Nrf2-mediated activation of the STAT3 gene. This then transfers cardioprotective signals to targeted genes, which may serve as an important regulatory checkpoint in the process of IsoPostC-mediated cardioprotection. However, more evidence is needed to delineate how Brg1, Nrf2 and STAT3 interact with each other in IsoPostC cardioprotection, in order to clarify the mechanisms responsible for the IsoPostC-mediated effects in normal rodents.
Some volatile anaesthetics such as desflurane have been shown to be cardioprotective when applied at the onset of reperfusion in the myocardium for patients with diabetes . However, in most cases, the cardioprotective effects of anaesthetics including isoflurane have been shown to be diminished or even abolished in hearts from individuals with diabetes [2,6]. Hyperglycaemia-induced oxidative and nitrosative stress, increased quantities of circulating inflammatory cytokines and dysfunction of mitochondrial KATP channels are all plausible factors to explain the susceptibility of diabetic hearts to IRI [36–39]. Acute hyperglycaemia, a common perioperative finding in cardiac surgery, is associated with high morbidity in diabetic and non-diabetic patients. In fact, there is a high proportion (approximately 20%) of patients who have diabetes  undergoing cardiac surgery (e.g. coronary artery bypass grafting) and they have more severe post-operative complications and increased morbidity and mortality during or after surgery . Thus, experimental models using animals with a relatively long duration of diabetes for the study of diabetes tolerance to myocardial IR would be more clinically relevant. So far, the effect of IsoPostC on myocardial IR in diabetes is conducted in animals with acute hyperglycaemia (induced by dextrose) [4,5].
Our most recent study showed that the postconditioning cardioprotection of the volatile anaesthetic sevoflurane was lost in D rats with a relatively long duration of disease (8 weeks of diabetes), which was associated with reduced cardiac STAT3 activation, and that treatment with the antioxidant NAC could restore cardiac STAT3 activation and preserve sevoflurane postconditioning cardioprotection . However, the mechanism by which hyperglycaemia reduces cardiac STAT3, and how the antioxidant NAC restores diabetic heart STAT3 activation and preserves volatile anaesthetic postconditioning cardioprotection in diabetes has not been explored. We found that the loss of sensitivity to IsoPostC in D rats was associated with decreased myocardial Brg1 expression, which was consistent with our previous study showing that Brg1 was reduced in diabetes [17,18]. Moreover, the decreases in Nrf2 and p-STAT3 were also detected in diabetes, which suggests that impairment in Brg1/Nrf2/STAT3 signalling is probably responsible for compromised IsoPostC cardioprotection in diabetes. Intriguingly, in H9C2 cells incubated with HG, NAC treatment restored the IsoPostC protective effect that was associated with enhanced Brg1 expression, whereas Brg1 knockdown cancelled the NAC-mediated restoration of IsoPostC protection. This supports the notion that Brg1 plays a critical role in IsoPostC cardioprotection. The present study provides evidence to suggest that reduced STAT3 in diabetes may be a consequence of reduction in cardiac Brg1 and Nrf2 expression, which is a critical step towards understanding the pathophysiology of heart disease in diabetes.
Limitations of the study
In the present study, isolated perfused heart models and cultured cell models were used. Although these two models are good for mechanistic studies because all of the conditions in these models can be precisely controlled, there are some limitations to the methods including the absence of normal humoral influences and neuronal regulation. Therefore, the extension of the related study to the in vivo model of myocardial IR, which is more relevant to clinical settings, should offer better understanding of the cardioprotective effects and mechanisms of isoflurane in hearts from individuals with diabetes. In addition, there is cross-talk between the SAFE and RISK pathways, and these two important pro-survival pathways may converge at STAT3, the main target that we studied in the present study. Therefore, we could not completely exclude the possibility that IsoPostC may confer cardioprotective effects in part through the RISK pathway, which needs further investigation.
Yan Wang and Haobo Li performed the experiments, analysed the data, interpreted the experimental results and drafted the paper. Zhengyuan Xia and Michael G. Irwin designed the study. Haobo Li, Sheng Wang and Stanley Sau-ching Wong analysed the data, interpreted the experimental results and revised the paper. Huansen Huang, Shiming Liu and Xiaowen Mao interpreted the experimental results and analysed the data. Zhengyuan Xia and Michael G. Irwin edited, revised and approved the final version of the paper before submission.
We thank Professor Jin Liu (Department of Anaesthesiology, West China Hospital, Sichuan University) for providing the emulsified isoflurane.
This work was supported by a General Research Fund grant [grant numbers 7123915, 784011M, 17124614 to Z. Xia, 768211M and 17121315 to M.G.I.] from the Research Grants Council of Hong Kong, and in part by grants from the National Natural Science Foundation of China (NSFC) [grant number 81270899] and the University of Hong Kong Seeding grant for basic sciences.
area under the curve
Brahma-related gene 1
- C rat
- D rat
- KH solution
left ventricular end-diastolic pressure
nuclear factor-erythroid 2-related factor 2
- RISK pathway
reperfusion injury salvage kinase pathway
reactive oxygen species
- SAFE pathway
survivor activating factor enhancement pathway
signal transducer and activator of transcription-3
These authors contributed equally to this work.