N‐n‐butyl haloperidol iodide mediates cardioprotection via regulating AMPK/FoxO1 signalling

Abstract Derangement of redox condition largely contributes to cardiac ischemia/reperfusion (I/R) injury. FoxO1 is a transcription factor which transcripts a series of antioxidants to antagonize I/R‐induced oxidative myocardial damage. N‐n‐butyl haloperidol iodide (F2) is a derivative derived from haloperidol structural modification with potent capacity of inhibiting oxidative stress. This investigation intends to validate whether cardio‐protection of F2 is dependent on FoxO1 using an in vivo mouse I/R model and if so, to further elucidate the molecular regulating mechanism. This study initially revealed that F2 preconditioning led to a profound reduction in I/R injury, which was accompanied by attenuated oxidative stress and upregulation of antioxidants (SOD2 and catalase), nuclear FoxO1 and phosphorylation of AMPK. Furthermore, inactivation of FoxO1 with AS1842856 abolished the cardio‐protective effect of F2. Importantly, we identified F2‐mediated nuclear accumulation of FoxO1 is dependent on AMPK, as blockage of AMPK with compound C induced nuclear exit of FoxO1. Collectively, our data uncover that F2 pretreatment exerts significant protection against post ischemic myocardial injury by its regulation of AMPK/FoxO1 pathway, which may provide a new avenue for treating ischemic disease.

8][9] Especially, FoxO1 elevates the transcription level of a series of targeted genes encoding antioxidant proteins including SOD2 and catalase to mediates its protective effects against various stimuli-induced oxidant stress. 10,11Cardiac-specific knock-out of FoxO1 aggravates myocardial function and injury elicited by IR. 12 However, the molecular signalling that modulates FoxO1 activity is still largely unknown.4][15] AMPK has been reported to phosphorylate FoxO1 and to enhance FoxO1 nuclear translocation. 16,17Therefore, considering the significance of preservation of redox balance during heart I/R, it is important to elucidate the regulating mechanism of FoxO1 nuclear accumulation.
N-n-butyl haloperidol iodide (F 2 ) is a compound derived from haloperidol (a classical and widely used antipsychotic agent), which confers maintained relaxing effect of coronary artery on F 2 with no extrapyramidal reactions for its increased hydrophilicity.
9][20][21] In particular, an investigation from our group has indicated that F 2 activates AMPK and subsequently suppresses ROS generation to mitigate hypoxia/ reoxygenation (H/R)-provoked oxidative damage in primary cultured heart endothelial cells. 22However, the mechanism by which AMPK exerts its antioxidative effects yet to be fully elucidated.
Based on above-mentioned evidence, this research is designed to decipher whether FoxO1 is vital in F 2 -induced cardioprotection, and to further investigate the underlying regulating mechanism of FoxO1 signalling.F 2 was dissolved in DMSO as a stock solution with a concentration of 50 mg/mL.After 1:100 dilution in normal saline, F 2 (5 mg/ kg) was intraperitoneally administered to test animals 24 h before myocardial I/R.DMSO administered to test animals was limited at 1% of all injections.

| Animals
Adult C57BL/6 mice (male, 2-3 months) and neonatal mice (1-2 day) were used (Vital River Laboratory, Beijing).All experiments conducted to mouse were approved by the Institutional Animal Care and Use Committee of Shantou University Medical College and consistent with the Guidelines for the Care and Use of Laboratory Animals published by the National Institutes of Health.Animal suffering was minimized at the best of the current research.

| Myocardial I/R protocol
Global I/R surgery was prepared as previously described. 23llowing 1-week adaption, one single dose of sodium pentobarbital (50 mg/kg) was intraperitoneally injected to mice to induce general anaesthesia.After complete anaesthesia, mice were intubated and connected to a rodent ventilator to ensure effective ventilation.Next, a midline sternotomy was applied to mice to expose heart.Subsequently, a slipknot was tied to the left anterior descending (LAD) coronary artery for 45 min to induce myocardial ischemia, followed by loosening the slipknot to induce reperfusion for 24 h.Successful preparation of this model was confirmed by the observation of cyanosis at apex of heart and elevated STsegment (Figure S1).Sham-operated mice underwent the same procedure without LAD occlusion.
A preconditioning regimen was used to study the effect of F 2 on cardiac I/R (Figure S2).The mice were divided into seven groups: (1)   Sham; (2) I/R; (3) I/R + F 2 (5 mg/kg); (4) I/R + AS (10 mg/kg); (5) I/R + AS (10 mg/kg) + F 2 (5 mg/kg); (6) I/R + CC (10 mg/kg); (7) I/R + CC (10 mg/ kg) + F 2 (5 mg/kg).We previously demonstrated the safety of F 2 on healthy mice. 24Thus, a F 2 alone group was not included.In each experiment, both groups comprised six mice.anaesthesia procedure (see Section 2.3), followed by rapid excision of the hearts.Following re-ligation of LAD coronary artery at the same occlusion place, the aorta of excised heart was cannulated with a catheter to perfuse 1% EB dye retrogradely to characterize the ischemic zone.Then, the hearts were frozen for 30 min at −80°C and cut into slices.These heart slices were incubated in 2% TTC for 30 min at 37°C to delineate the infarct zone.After fixation with 4% paraformaldehyde overnight, the slices were scanned and an ImageJ software (Media Cybernetics, Bethesda, MD, USA) was used to determine the infarction, risk and left ventricular (LV) areas.The infarct size was presented as infarct size (INF)/area at risk (AAR).

| Evaluation of cardiac function
Echocardiography (ECG) was carried out to estimate the heart function by the Vevo 2100 imaging system (Visual Sonics Inc., Toronto, Ontario, Canada) as previously described. 23After mild anaesthesia using isoflurane, different groups of mice underwent transthoracic ECG to acquire high-resolution echocardiographic images using 30-MHz probe.The images were applied to measure a battery of indices of left ventricular function.These parameters were then used to calculate ejection fraction (EF) and fractional shortening (FS).

| TUNEL assay
After anaesthesia, a separate group of mouse hearts underwent rapid excision and fixation by 4% paraformaldehyde solution for 12 h at 4°C.Following dehydration by a vacuum tissue processor (Thermo Fischer Scientific), hearts from different groups were embedded by paraffin and cut into 3-4 sections.To detect apoptosis, a commercial TUNEL assay kit (Cat No. G3250) from Promega (Madison, WI, USA) was employed following the instructions of the manufacturer.
After the TUNEL and immunofluorescence staining for the myocardium marker cTnT, a LSM810 confocal microscope (Carl Zeiss AG, Oberkochen, Baden voorburg, Germany) was used to capture corresponding pictures.For quantification, myocardium apoptosis was presented as total nuclei (blue) count devided by TUNEL-positive nuclei (green) count in every sharp field of each section in AAR region.

| Measurement of myocardial enzyme release
The qualified serum samples were obtained by the centrifugation (3000 rpm) of collected blood specimens for 10 min.For the assessment of myocardial enzyme leakage, activity of serum creatine kinase (CK), lactate dehydrogenase (LDH) and MB isoenzyme of CK (CK-MB) were assessed by commercial Bio-sinew kits using Automatic Chemistry Analyzer (Toshiba Medical Systems Corporation, Otawara, Tochigi, Japan).

| Cardiomyocytes isolation
Neonatal mice cardiomyocytes were isolated by enzymatic digestion from mice (1-2 day).Briefly, ventricle part of mice hearts was carefully collected and cut into pieces.After washing with PBS three times, the ventricular tissue was digested by collagenase type II (0.8 mg/mL) for 5 min at 37°C.The digested cell suspension was filtered and centrifugated for 6 min at 800 rpm.After removing supernatant, cell pellets were re-suspended in M199 supplemented with 10% NBS and transferred to cell culture incubator for 30 min to discard adherent fibroblasts.Then, non-adherent cardiomyocytes kept culturing for another 48 h.

| Immunofluorescence
Detection of immunofluorescence was carried out as previously described. 22Cardiomyocytes cultured on coverslips were fixed and permeabilized with 0.2% Trition X-100 for 30 min.Next, blocking buffer was employed to reduce background for 1 h at room temperature.Primary antibodies (AMPK and FoxO1) with 1:200 dilution were used to probe respective proteins at 4°C overnight.Cells were washed with PBS and incubated with secondary fluorescent FITC (green)-and CY3 (red)-conjugated antibodies.DAPI was used to stain the nuclei.A confocal LSM 810 (Carl Zeiss AG) microscopy was exploited to capture the images of distinct groups.

| Western blot
Western blot was conducted as previously described. 24Equal heart protein samples (about 50 μg) were loaded to 10% SDS-PAGE gels.
The gels were then transferred to nitrocellulose membrane.The membranes were then blocked and incubated with primary antibodies (described in Section 2.1) at 4°C overnight.Following incubation with HRP-conjugated secondary antibodies, the immunoblots were probed by super ECL reagent.The signals of corresponding bands were detected using a Bio-Rad imaging system and quantified using a Gel-Pro software (Media Cybernetics, USA).GAPDH was defined as the loading control to compare band intensity of different groups.

| Co-immunoprecipitation
Co-immunoprecipitation analysis was conducted to detect AMPK-FoxO1 interaction.In brief, heart tissue lysates were acquired using NP-40 lysis buffer (Cat No. P0013F, Beyotime).Following protein concentration determination, equal amount of samples (about 500 μg) were incubated with 1.5 μg of primary antibodies at 4°C overnight on a rocker.Protein A/G magnetic beads (Cat No. HY-K0202, MedChemExpress) were then used to precipitate antibody-bound proteins.After 4 h precipitation at room temperature, the beads were separated by a magnetic stand (Thermo Fischer Scientific) and washed with ice-cold PBST solution (0.1% Tween-20 in PBS).After elution in SDS loading buffer in boiling water for 5 min, eluted proteins were further analysed by western blot.

| Real-time PCR (RT-PCR)
RT-PCR was performed as previously described. 24Total mouse heart RNA was isolated using Trizol (Cat No. 9108, Takara, Kyoto, Japan) and reversely transcribed to cDNA by employing a reverse transcription system (Cat No. RR037A, Takara).SYBR Green master mix (Cat No. Q121-1, Vazyme Biotech, Nanjing, China) was exploited to prepare the reaction system to perform RT-PCR experiment using a LightCycler480 system (Roche Inc., Basel, Switzerland).Relative expression of targeted genes was normalized over 18S RNA.All primer pairs were derived from a previously published document. 25

| Nuclear and cytosolic protein isolation
Isolation of nuclear and cytosolic fraction was developed based on a method previously reported by Calvert et al. 26 Briefly, mouse heart samples were homogenized in buffer A with protease inhibitor cocktail, EDTA, Tris-HCl, sucrose and then centrifugated at 1000 g for 10 min.Cell pellets were washed and underwent resuspension in RIPA lysis buffer.Following centrifugation, supernatants were collected as the nuclear fractions.In contrast, followed by the first round of centrifugation, the supernatants were subjected to a second round of 100,000 g centrifugation for 1 h to obtain the cytosolic fractions.

| Statistics
All data were presented as mean ± SEM.Comparisons between two groups were assessed by two-sided unpaired Student's t-test in Prism 9 (GraphPad Software, La Jolla, CA, USA).Data groups were compared using one-way anova with Dunnett post hoc analysis.p < 0.05 was considered significant.S1).Therefore, these data demonstrate that F 2 is a potent cardioprotective compound.

| Preconditioning of F 2 alleviates the extent of apoptosis and inflammation
Because apoptosis substantially contributes to the cardiomyocytes death in ischemic heart disorder, [27][28][29] we evaluated whether F 2 preconditioning could reduce this form of cell death.Western blot analysis showed that myocardial apoptosis was evidently restricted by F 2 preconditioning.This was as evidenced by the prevention of increase in Bax, cleaved caspase-3 expression, and reduction in Bcl-2 expression (Figure 2A-C), which was also supported by the decreased TUNEL positive myocyte counts (Figure 2D,E).Given that inflammation exerts a key role in postischemic injury, 30,31 we next estimated the potential effects of F 2 on inflammation responses.The transcription of pro-inflammatory cytokines such as IL-6, IL-1β and TNF-α was markedly amplified during I/R but not in F 2 -preconditioned hearts (Figure 2F-H).
These results suggest that repression of inflammation and apoptosis may be responsible, at least partially, for cardioprotection conferred by F 2 .

| F 2 stimulates AMPK phosphorylation and increases nuclear FoxO1 import and the expression of SOD2 and catalase
Next, we tested whether F 2 might maintain redox homeostasis through AMPK and FoxO1 signals.As expected, I/R induced an adaptive increase in phosphorylation of AMPK and nuclear import of FoxO1, and this induction was augmented in F 2 -preconditioned hearts (Figure 3A,B,D).Furthermore, I/R resulted in downregulation of SOD2 and catalase expression, which was corrected by F 2 (Figure 3E,F).Intriguingly, I/R caused a moderate but not significant decrease of total FoxO1 expression, whereas F 2 could mildly upregulate FoxO1 expression (Figure 3C).F 2 also relieved I/R-stimulated oxidant stress, as detected by DHE staining (Figure 3G,H).These results indicate that F2 might attenuate oxidative stress by stimulating AMPK/FoxO1 pathway.

| FoxO1 is involved in the cardioprotection of F 2
To determine whether FoxO1 is essential for the repressive effects of F 2 on cardiac I/R lesion, FoxO1-specific inhibitor AS was used.
For these experiments, AS (10 mg/kg) was intraperitoneally administered to mice at 48 h and 24 h before ischemia.AS blunted the

| AMPK mediates F 2 stimulation of FoxO1 signalling
We subsequently aimed at determining whether F2-induced activation of FoxO1 signalling was associated with AMPK.In this context, AMPK inhibitor CC (10 mg/kg) was also intraperitoneally

| AMPK interacts with FoxO1
Given our findings and the close relationship between AMPK and FoxO1 signalling, AMPK-FoxO1 interaction was confirmed by co-immunoprecipitation and immunofluorescence in cardiomyocytes.As shown in Figure 6A,B, FoxO1 was detected in AMPK immunoprecipitates and AMPK also co-immunoprecipitated with FoxO1, suggesting a physical interaction between these two proteins.Moreover, consistent with the results of co-immunoprecipitation, substantial colocalization of AMPK and FoxO1 were observed in cardiomyocytes using immunofluorescence staining (Figure 6C).Under physiological conditions, a small proportion of ROS generated is rapidly scavenged by internal antioxidative systems. 32wever, these systems are severely disrupted during I/R, which contributes to excessive ROS production and leads to the initial myocardial reperfusion injury. 33Thus, enhancing the activity of antioxidative enzymes may be a good strategy to decrease the oxidative stress-elicited reperfusion damage. 34,35Our previous work showed that F 2 protected against H/R model-evoked oxidative injury in CMECs and myocytes. 21,22In agreement with these results, we found that the in vivo I/R-induced oxidative injury was also significantly alleviated by F 2 pretreatment.This alleviation was also associated with the upregulation of nuclear FoxO1 translocation and its downstream antioxidants SOD2 and catalase.Moreover, FoxO1 blockage restricted the inhibitory effects of F 2 on myocardial injury.Thus, myocardial protection of F 2 is mediated by its activation of FoxO1-dependent antioxidative signalling.As a major energy metabolic regulator, AMPK has been extensively documented to confer benefits on myocardium during I/R. 14,15 previously reported that F 2 promoted the phosphorylation of AMPK in endothelial cells, 22 whereas this is the first study reporting AMPK activation by F 2 in vivo.Importantly, we observed that the cardio-protection of F 2 was blocked by the administration of AMPK selective inhibitor, suggesting that activating AMPK is indispensable for its resistant effects against I/R injury.
We also found that AMPK was implicated in F 2 -induced FoxO1 activation, which was evidenced by increased nuclear exit of FoxO1 in case of the F 2 pretreatment group by AMPK inhibition.This observation is in parallel with the results of several studies. 16,36rthermore, suppression of AMPK likewise reversed the subsequent upregulation of antioxidant proteins.Importantly, the current study firstly reports that AMPK interacts with FoxO1 in cardiomyocytes.Collectively, these data indicate that F 2 counteracts myocardial I/R damage via AMPK-FoxO1 signalling.Notably, an important unanswered question exists with regard to the precise mechanisms by which AMPK activation activates FoxO1, which may relate to the phosphorylation of Thr 649 residue and will be investigated in vitro cardiomyocytes H/R model in our further studies using genetical methods such as small interfering RNA other than pharmacological inhibition.
Another question worthy of explanation is that there are some discrepancies existing in our work.Here, we showed that, during I/R, phosphorylated AMPK and nuclear FoxO1 accumulation were enhanced, whereas the expression of downstream antioxidants including SOD2 and catalase was decreased.These seemingly contradictory data may be deciphered by a documented landmark investigation which showed that myocardial I/R activates Hippo pathway and induced the reduced accumulated Yes-associated protein (YAP, a nuclear co-factor of FoxO1) in the nucleus, resulting in the downregulation of FoxO1-targeting gene transcription, despite elevated nuclear translocation of FoxO1. 37Whether F 2 regulates the Hippo-YAP signalling axis needs investigation in future studies.
To sum up, we substantiated that F 2 negatively affects I/Rinduced heart oxidative lesion by activating AMPK-FoxO1 signalling pathway (Figure 7).Moreover, targeting this antioxidant axis may represent a novel strategy to prevent any other pathologic stimuli-triggered oxidative injury apart from ischemic heart.

| 3 of 12 LU
Measurement of myocardial infarct size was carried out by TTC-EB double staining.Briefly, after I/R, mice underwent the same et al.
Initial experiments were performed to test whether F 2 pretreatment could ameliorate cardiac injury following I/R.F 2 were intraperitoneally administered to mice 24 h before I/R.F 2 significantly reduced the INF:AAR ratio (Figure 1E,F), although AAR per LV were comparable (data not shown).Of note, F 2 -treated hearts were tolerant against myocardial I/R-induced necrosis, as indicated by deceased leakage of myocardial enzymes including LDH, CK and CK-MB (Figure 1A-D).Furthermore, cardiac function almost restored to baseline in F 2 -treated mice, as evidenced by the apparent preservation in EF and FS (Figure 1G-I; Table

F 2 -
mediated increment of SOD2 and catalase expression in ischemic myocardium (Figure4A-C).On the contrary, AS had no significant influence on the INF:AAR ratio in I/R-challenged hearts and nearly abrogated the depressive effects of F 2 on myocardial infarct size (Figure4D,E).Additionally, F 2 -induced reduction of TUNEL-positive nuclei and ROS production was also abolished by AS (Figure4F-I).

F
I G U R E 1 F 2 protects mouse hearts against myocardial ischemic damage.(A-D) Effects of F 2 on serum LDH, CK, CK-MB and TnT concentration (n = 6).(E, F) Representative images and quantification analysis for infarction in hearts, scale bar = 1 mm (n = 5).(G-I) Representative echocardiograms and quantification analysis for EF and FS (n = 5).Data represent mean ± SEM. # p < 0.05 versus the Sham group; *p < 0.05 versus the I/R group.Thus, F 2 -mediated cardioprotection is correlated with FoxO1evoked antioxidant signalling pathway.
administered to mice at 48 h and 24 h prior to I/R.AMPK Inhibition with CC abrogated the F 2 -mediated nuclear FoxO1 accumulation and upregulation of SOD2 and catalase in I/R-exposed hearts (Figure5A-E).Although CC did not impact on myocardial infarct size, TUNEL-positive cardiac cells and ROS generation in vehicletreated mice, CC almost abolished the suppressive effects of F 2 on cardiac infarct area, apoptosis and oxidative stress after I/R (Figure5F-K).These findings suggest that AMPK is an upstream mediator for F 2 -induced activation of FoxO1 antioxidative signalling cascade.

F I G U R E 2 F
2 ameliorates I/R-triggered apoptosis and inflammation.(A-C) Western blot was employed to determine the expression of cleaved caspase 3, caspase3, Bax and Bcl-2 (n = 6); (D, E) Apoptotic cardiac myocytes were assessed using the TUNEL assay, scale bar = 20 μm (n = 6); (F-H) RT-PCR was exploited to detect the extent of inflammation (n = 6).Data represent mean ± SEM. # p < 0.05 versus the Sham group; *p < 0.05 versus the I/R group.

F
I G U R E 3 F 2 enhances AMPK phosphorylation and nuclear FoxO1 accumulation in the I/R-challenged heart.(A-F) Expression of nuclear FoxO1, FoxO1, AMPK, phospho-AMPK, Catalase and SOD2 was examined by western blot (n = 6).(G, H) ROS generation was detected by DHE staining, scale bar = 20 μm (n = 6).Data represent mean ± SEM. # p < 0.05 versus the Sham group; *p < 0.05 versus the I/R group.H3 indicates the histone H3, a nuclear marker.that F 2 -induced protection against myocardial I/R insult was FoxO1dependent.Furthermore, AMPK inhibition with CC abolished F 2 -induced nuclear FoxO1 accumulation and the subsequent upregulation of antioxidant factors including SOD2 and catalase, resulting in the loss of F 2 cardioprotective effects.Therefore, we conclude that pharmacological F 2 pretreatment attenuates cardiac I/R injury via activating AMPK-FoxO1 axis.

F I G U R E 5
Blockade of AMPK relieves effects of F 2 on FoxO1 activation in the ischemic heart.(A-E) Western blot was employed to examine expression of nuclear FoxO1, FoxO1, AMPK, phospho-AMPK, Catalase and SOD2 (n = 6); (F, G) Representative pictures and quantification analysis for infarction, scale bar = 1 mm (n = 5); (H, I) TUNEL staining was employed to evaluate apoptotic cardiac cells, scale bar = 20 μm (n = 6); (J, K) DHE staining was utilized to estimate ROS production, scale bar = 20 μm (n = 6).Data represent mean ± SEM. *p < 0.05 versus the I/R + F 2 group.F I G U R E 6 The interaction between AMPK and FoxO1.(A) After immunoprecipitation with anti-AMPK antibody or rabbit IgG, western blot was employed to examine the expression of AMPK and FoxO1; (B) After immunoprecipitation with anti-FoxO1 antibody or rabbit IgG, western blot was employed to examine expression of AMPK and FoxO1; (C) Representative immunofluorescent images showing the colocalization of AMPK (green) and FoxO1 (red) in cardiomyocytes, scale bar = 20 μm.HC indicates the non-specific IgG heavy chain bands.F I G U R E 7 Scheme illustrating the current investigation.We propose that F 2 increases FoxO1 nuclear translocation through activating AMPK, thereby promoting the transcription of antioxidant proteins, including SOD2 and Catalase.These increased levels of antioxidants negatively regulate ROS production, ultimately mitigating I/R-caused heart injury.