Intermittent insulin treatment mimics ischemic postconditioning via MitoKATP channels, ROS, and RISK

Abstract

Objectives. It has previously been demonstrated that 15-min continuous insulin infusion at immediate reperfusion affords cardioprotection. This study sought to reduce the treatment time of insulin and test if intermittent insulin infusions can mimic ischemic postconditioning. Design. In a Langendorff perfused rat heart model of regional ischemia, hearts were at the onset of reperfusion subjected to either 5- or 1-min continuous insulin infusion or 3 × 30 s intermittent insulin infusions (InsPost); with or without inhibitors of Akt (SH-6), p70s6-kinase (rapamycin), mitochondrial ATP-sensitive potassium channels (5-hydroxydecanoic acid [5-HD]), or a scavenger of reactive oxygen species (ROS; 2-mercaptopropionyl glycine [MPG]). Infarct size is expressed as percent of area at risk and presented as mean ± standard error of the mean or s.e.m. Results. Only InsPost was able to reduce infarct size compared with controls (InsPost 33 ± 6% vs. Ctr 52 ± 4%, p < 0.05.). This cardioprotection was abrogated by co-administering SH-6, rapamycin, 5-HD, or MPG. (InsPost + SH-6 56 ± 9%, InsPost + Rapa 55 ± 8%, InsPost + 5-HD 56 ± 7%, InsPost + MPG 60 ± 3% vs. InsPost 33 ± 6% p < 0.05). These results were corroborated by a significant increase in phosphorylated Akt and p70s6k in the InsPost group compared with controls. Conclusion. Short intermittent insulin infusions can mimic ischemic postconditioning and reduce myocardial infarct size via Akt/p70s6k and mK_ATP channels/ROS-dependent signaling.

Key words::

• cardioprotection
• insulin
• ischemia
• postconditioning
• reperfusion injury
• reperfusion therapy
• ROS
• RISK

Introduction

It has previously been demonstrated that insulin can afford cardioprotection from ischemia–reperfusion-induced injury when administered at the immediate onset of reperfusion (1–3). Recent evidence also demonstrates that insulin can alleviate chronic adverse changes in postischemic cardiac structure and function (4). The binding of insulin to its receptor activates a series of downstream prosurvival kinases, collectively called the Reperfusion Injury Salvaging Kinases (RISK) pathway, which plays a crucial role in both ischemic preconditioning (IPC) and ischemic postconditioning (IPost) (5,6), whereby short alternating cycles of ischemia and reperfusion are applied immediately before or after a prolonged ischemic event, respectively. Activation of phosphoinositide 3-kinase (PI3K) and Akt is central in RISK signaling, which when in a phosphorylated state activates several downstream targets such as p70 ribosomal S6 protein kinase (p70s6k) (3), endothelial NO synthase (eNOS) (7), and glycogen synthase kinase 3 β or GSK3β (8). The end result of RISK activation is preventing or delaying opening of mitochondrial permeability transition pores or mPTPs during early reperfusion, and thus, preventing the associated cell death (9).

Results also indicate that insulin only needs to be administered for the first 15 min of reperfusion to achieve cardioprotection, and if the treatment is delayed 15 min the protection is lost. The same study also indicates that Akt is significantly phosphorylated after only 2 min of insulin administration at reperfusion (2). As Akt phosphorylation plays a critical role in conveying insulin cardioprotection, it should, theoretically, be possible to reduce the treatment time from 15 min. This study therefore explores the cardioprotective potential of shorter insulin treatment protocols. Secondly, we explored a strategy of intermittent insulin infusions in an attempt to shorten treatment time and lower treatment dose by mimicking the treatment protocol of ischemic postconditioning. Both we and others have previously been successful with this conditioning strategy (10,11).

Our final objective was to assess the role of mitochondrial ATP-sensitive potassium (mK_ATP) channels in insulin treatment at reperfusion. These channels are thought to open at reperfusion in response to IPost (12,13), IPC (14,15), and several other pharmacological treatments (16,17), creating a reactive oxygen species (ROS) burst which phosphorylates and translocates protein kinase Cϵ to the mitochondria, leading to cardioprotection. Previous results from others where insulin was used to precondition hearts or myocytes, however, failed to show any involvement of mK_ATP channels or the associated mitochondrial ROS burst (18,19), but this has as of yet not been tested at reperfusion.

Methods

Langendorff perfusion procedure

All experiments were approved by the Norwegian State Commission for Laboratory Animals, and carried out in accordance with the European Communities Council Directive of 1986 (2010/63/EEC). Male Wistar rats (300–350 g) fed a standard diet ad libitum were heparinized (200 IU) and anesthetized with sodium pentobarbital (50 mg/kg) i.p. The hearts were excised, placed in ice-cold Krebs–Henseleit buffer (KHB), and rapidly mounted onto a constant-pressure Langendorff perfusion system (80 mmHg). The KHB (pH: 7.4, oxygenated with 95% O₂/5% CO₂) contained in mM: 118 NaCl, 25.2 NaHCO₃, 4.7 KCl, 1.84 CaCl₂, 1.21 KH₂PO₄, 1.22 MgSO₄, and 11 glucose. A water-filled latex balloon, connected to a hydrostatic pressure transducer (TrueWave, Edwards Lifesciences, USA) and coupled to a data acquisition system (PowerLab 8/30, Chart Pro software-MLS250), was inserted into the left ventricle (LV) through an incision in the left auricle, and inflated to set a left ventricular end diastolic pressure of 5–10 mmHg during stabilization. The ensuing experiments were performed under isovolumetric conditions, and recordings of LV developed pressure (LVDP) and heart rate were made and used for calculation of rate pressure product (RPP). Coronary flow (CF) was measured by timed collection of effluent over 30 s at each sampling point. Myocardial temperature (37°C) was monitored by inserting a thermo-probe into the pulmonary artery. A 3–0 silk suture was passed around the main branch of the left coronary artery, and the ends were threaded through a small vinyl tube to form a snare. Regional ischemia (RI) was achieved by pulling the snare and confirmed by a substantial fall in both LVDP and CF. Reperfusion was achieved by releasing the snare.

Measurement of ischemic risk zone and infarct size

Infarct size (IS) and area at risk (AAR) measurements were performed according to the gold standard (20). At the end of each experiment, the silk suture was securely tightened and a 0.2% (w/v) suspension of Evans Blue infused to demarcate the risk zone (Duke Scientific Corp., Palo Alto, CA, USA). Thereafter the hearts were frozen (− 20°C) overnight and cut into 2-mm-thick slices from apex to the atrioventricular groove, thawed and stained with 1% triphenyltetrazolium chloride (TTC) in phosphate buffer (pH: 7.4) at 37°C for 15 min, before fixation in 4% formalin to enhance the contrast of the stain. The area of the LV, the infarcted area (IS, TTC negative), and the risk zone (AAR, blue) were determined using a computerized planimetry program (Planimetry⁺ v2.0; Erik Traasdahl, ENK, Norway) and multiplied by slice thickness (2 mm) to give calculated volumes. IS is expressed as the volume of IS/AAR(%) and risk zone is expressed as the volume of AAR/LV(%) (Table I).

Table I. Ratio of volume of AAR and LV by Evans Blue staining.

Group	AAR/LV (%)
Ctr	43.6 ± 3.7
InsR + 1	46.2 ± 3.2
InsR + 5	42.4 ± 2.7
InsR + 15	46.0 ± 3.3
InsPost	42.6 ± 2.9
IPost	50.6 ± 3.3
SH-6	41.0 ± 4.2
InsPost + SH-6	47.4 ± 2.1
Rapa	43.2 ± 1.8
InsPost + Rapa	47.4 ± 1.8
5-HD	50.3 ± 2.3
InsPost + 5-HD	33.6 ± 2.5*¤
MPG	48.6 ± 3.6
InsPost + MPG	40.9 ± 2.2

Values represent means ± s.e.m.

Ctr = Ischemia reperfusion control; IPost = ischemic postconditioning, 3 × 30 s; InsPost = insulin postconditioning, 3 × 30 s; Ins_{R + 1}/Ins_{R + 5}/Ins_{R + 15} = 1, 5 or 15 min insulin administration at immediate reperfusion; SH-6 = Akt inhibitor at reperfusion; Rapa = p70s6k inhibitor at reperfusion; MPG = ROS scavenging at reperfusion; 5-HD = mK_ATP channel inhibition at reperfusion.

*P < 0.05 versus Ctr, ¤P < 0.05 versus InsPost.

Experimental protocol

All hearts were stabilized for 20 min before being subjected to 30-min RI and 120-min reperfusion (Figure 1). Baseline values for functional parameters were obtained after 18 min of stabilization. Initially, the temporal effects of insulin administration [0.3 mU ml-1 (2)] (Novo Nordisk A/S, Bagsværd, Denmark) were studied by administering insulin for 15 min (Ins_{R + 15}), 5 min (Ins_{R + 5}), or 1 min (Ins_{R + 1}) at the start of reperfusion. Subsequently, the ability of insulin to mimic ischemic postconditioning was studied by administering insulin [0.3 mU ml-1] for 3 × 30 s at immediate reperfusion (Figure 1). Furthermore, to explore if the pro-survival RISK signaling pathway was underlying the cardioprotective effect of insulin treatment, the following inhibitors ± insulin treatment were administered for 15 min at reperfusion (Figure 1): Akt inhibitor [SH-6; 10 μM (11)] and the mTOR kinase inhibitor rapamycin [Rapa; 1 nM (2)] (both from Calbiochem). In order to investigate the involvement of mK_ATP channels and ROS in insulin-induced signaling at reperfusion, the putatively selective inhibitor of mK_ATP channels, 5-hydroxydecanoic acid [5-HD; 100 μM (14)] and the free radical scavenger, 2-mercaptopropionyl glycine [MPG; 1 mM (14)] was applied for 15 min at reperfusion with or without InsPost (both from Sigma-Aldrich).

Figure 1. Experimental protocol. Stab = stabilization; RI = regional ischemia; Open bars = KHB buffer perfusion; Ins_{R + 1} = insulin [0.3 mU ml^-1] for the first 1 min of reperfusion; Ins_{R + 5} = insulin for the first 5 min of reperfusion; Ins_{R + 15} = insulin for the first 15 min of reperfusion; InsPost = insulin postconditioning, alternating cycles of 3 × 30 s with insulin and KHB infusion; IPost = ischemic postconditioning, 3 × 30 s of global ischemia; The inhibitors (SH-6, Rapa, MPG, and 5-HD) were co-administered with or without InsPost.

Immunoblot analysis

Myocardial Akt (Phospho-Akt, Ser⁴⁷³, and total-Akt) and p70s6 kinase phosphorylation (Phospho-p70s6k, Thr⁴²¹/Ser⁴²⁴ and total p70s6k) in the area at risk was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (all antibodies from Cell Signaling Technology, USA). A separate set of hearts underwent the protocol as previously described (Figure 1) where tissue was collected at 15 min of reperfusion. Hearts perfused with KHB for 20 min served as baseline controls (Ctr_B), while ischemia–reperfusion controls are Ctr_Rep. Cardiac ventricular tissue was homogenized in a lysis buffer (10 mM Tris at pH of 7.5, 2 mM ethylenediaminetetraacetic acid or EDTA, 330 mM sucrose, and a complete protease inhibitor cocktail from Roche)(Roche Applied Science, Penzberg, Germany) using a Precellys homogenizer (Montigny-le-Bretonneux, France), and tissue debris was removed by centrifugation at 16000 rcf (10 min). Protein quantification, sample preparation (40 μg/lane), and electrophoresis were performed as previously described (2). Ponceau S staining (Sigma, St. Louis, USA) confirmed successful transfer. Densitometric analysis was performed using Quantity One software (Bio-Rad, CA, USA), and phosphorylation of Akt and p70s6k was expressed as the ratio between phosphorylated and total protein levels. GAPDH expression was determined to ensure equal loading.

Statistical analysis

Values are presented as mean ± standard error of the mean (s.e.m). IS, AAR, and SDS-PAGE results were tested for group differences by one-way analysis of variance (ANOVA) combined with Fisher's post hoc test. Comparisons of LVDP, CF, and RPP both within and between groups were tested with mixed ANOVA combined with Tukey's post hoc test for significant differences. We tested stabilization versus RI to confirm equal degree of ischemia between groups, and tested % recovery of 18-min stabilization at 5, 30, and 120 min of reperfusion to determine any differences after treatment. All statistics were performed in IBM SPSS (version 20.0.0). A value of p < 0.05 was considered statistically significant.

Results

Cardioprotection from insulin reperfusion therapy is temporally dependent

Continuous insulin treatment for the first 5 or 1 min did not reduce IS compared with controls (Ins_{R + 1} 49 ± 5%, Ins_{R + 5} 48 ± 4% vs. Ctr 52 ± 4%, ns) (Figure 2A), while 15-min treatment did (Ins_{R + 15} 36 ± 6% vs. Ctr 52 ± 4%, p < 0.05) (Figure 2A). These findings were corroborated by results from Akt immunoblots where Ins_{R + 15} showed significantly increased levels of phosphorylated Akt compared with Ins_{R + 5} and Ins_{R + 1} (p < 0.05, Figure 2B and C), and Ins_{R + 5} showed twofold higher Akt phosphorylation compared with Ins_{R + 1} (p < 0.05, Figure 2B and C).

Figure 2. The temporal effect of insulin administration for 1, 5, or 15 min at reperfusion. (A) Only Ins_{R + 15} was able to significantly reduce IS versus Ctr. IS is expressed as % of AAR. Bars represent mean ± s.e.m. (n ≥ 6) *p < 0.05 versus Ctr. (B and C) In a parallel set of hearts, Akt phosphorylation was determined at 15-min reperfusion. The longer the duration of insulin therapy, the higher the levels of phosphorylated Akt. Akt phosphorylation at (Ser⁴⁷³) is expressed in arbitrary units (AU) as a ratio relative to total Akt. GAPDH verifies equal loading. Bars represent means ± s.e.m (n ≥ 3).* p < 0.05 versus Ins_{R + 1}. #p < 0.05 versus Ins_{R + 5}.

Insulin postconditioning is equally cardioprotective as ischemic postconditioning or Ins_{R + 15}

Administration of insulin in bursts of 3 × 30 s at the onset of reperfusion, significantly reduced IS as compared with control hearts (InsPost 33 ± 6% vs. Ctr 52 ± 4%, p < 0.05), but was not statistically different from ischemic postconditioning or Ins_{R + 15} (IPost 36 ± 3%, Ins_{R + 15} 36 ± 6% vs. InsPost 33 ± 6%, ns) (Figures 2 and 3). To rule out any possible effects of the conditioning protocol itself, an additional group was administered KHB for 3 × 30 s at reperfusion from a separate reservoir to serve as sham control, and it was not different from Ctr (data not shown).

Figure 3. Insulin postconditioning is equally cardioprotective as ischemic postconditioning and the protection is abrogated by inhibitors of Akt, p70s6k, or mK_ATP channels, or a ROS scavenger. InsPost showed similar degree of cardioprotection as ischemic postconditioning (IPost). Cardioprotection from InsPost was abrogated by inhibiting Akt (SH-6), mTOR/p70s6k (Rapa), or mK_ATP channels (5-HD), or by scavenging ROS (MPG). Neither of the inhibitors affected IS alone. IS is expressed as % of AAR. Bars represent mean ± s.e.m. (n ≥ 6) *p < 0.05 versus Ctr.

The infarct sparing effect of insulin postconditioning is abrogated by inhibitors of Akt, p70s6k, or mK_ATP channels, or ROS scavenging

When inhibitors of Akt, p70s6k, or mK_ATP channels or a ROS scavenger were co-administered with insulin postconditioning, the cardioprotection was lost (InsPost + SH-6 56 ± 9%, InsPost + Rapa 55 ± 8%, InsPost + 5-HD 64 ± 4%, InsPost + MPG 56 ± 7% vs. InsPost 33 ± 6%, p < 0.05) (Figure 3). The inhibitors/scavenger had no effect on IS themselves (SH-6 55 ± 4%, Rapa 48 ± 3%, 5-HD 57 ± 4%, MPG 60 ± 3% vs. Ctr 52 ± 4%, ns) (Figure 3).

Functional parameters are shown in Table II. There were no major group differences at baseline or RI, and all groups had a significant fall in LVDP, CF, and RPP at RI, thus confirming that all groups received sufficient and similar degree of ischemia. The effects of treatment on IS were only marginally reflected in functional data (CF significantly higher in Ins_{R + 15} vs. Ctr at 120 min rep). This is often the case in models of RI, where persistent stunning is thought to obscure the functional effects of more viable tissue in the cardioprotected groups (21).

Table II. Functional parameters recorded during the experimental protocol.

	Group	18’ stab	5’ RI	5’ Rep	30’ Rep	120’ Rep
				% recovery from 18’ stab
LVDP	Ctr	155 ± 14	75 ± 10^#	66 ± 9	59 ± 9	39 ± 7
mmHg	InsR + 1	152 ± 19	90 ± 14^#	79 ± 11	62 ± 12	56 ± 10
	InsR + 5	132 ± 8	70 ± 8^#	65 ± 10	61 ± 7	39 ± 5
	InsR + 15	153 ± 13	81 ± 8^#	86 ± 8	85 ± 4¤	61 ± 4¤
	InsPost	132 ± 13	68 ± 7^#	66 ± 8	55 ± 7	46 ± 7
	IPost	122 ± 11	79 ± 13^#	72 ± 7	71 ± 9	50 ± 9
	SH-6	170 ± 10	74 ± 11^#	54 ± 13	36 ± 8	25 ± 9
	InsPost + SH-6	137 ± 21	46 ± 13^#	60 ± 13	70 ± 10	54 ± 6
	Rapa	149 ± 20	59 ± 11^#	91 ± 7	78 ± 8¤	58 ± 9
	InsPost + Rapa	135 ± 11	68 ± 14^#	76 ± 16	68 ± 14	53 ± 8
	5-HD	146 ± 9	92 ± 9^#	57 ± 7	60 ± 5	43 ± 5
	InsPost + 5-HD	153 ± 9	112 ± 10^#	74 ± 5	60 ± 9	47 ± 6
	MPG	158 ± 13	71 ± 7^#	84 ± 8	85 ± 5¤	61 ± 6¤
	InsPost + MPG	166 ± 16	67 ± 14^#	73 ± 7	71 ± 4	54 ± 4
CF	Ctr	14.2 ± 1.0	8.4 ± 0.9^#	75 ± 7	66 ± 8	40 ± 4
ml min^− 1	InsR + 1	10.8 ± 0.7	6.9 ± 0.5^#	89 ± 4	80 ± 9	60 ± 8
	InsR + 5	11.2 ± 0.7	7.2 ± 0.7^#	92 ± 9	74 ± 7	55 ± 5
	InsR + 15	11.6 ± 1.5	7.4 ± 0.7^#	112 ± 13	98 ± 10¤	78 ± 7*¤
	InsPost	9.9 ± 1.5	6.4 ± 0.8^#	92 ± 9	91 ± 10	64 ± 8
	IPost	12.1 ± 1.3	8.9 ± 1.1^#	94 ± 10	80 ± 9	62 ± 9
	SH-6	12.4 ± 0.8	7.9 ± 0.9^#	64 ± 24	30 ± 10	30 ± 8
	InsPost + SH-6	10.8 ± 0.8	6.0 ± 0.3^#	126 ± 21	77 ± 15¤	59 ± 8¤
	Rapa	10.9 ± 1.6	6.4 ± 0.7^#	95 ± 4	85 ± 3	67 ± 3
	InsPost + Rapa	9.5 ± 0.5	5.5 ± 0.7^#	107 ± 11	92 ± 14	65 ± 5
	5-HD	11.5 ± 0.6	7.1 ± 0.3^#	87 ± 4	70 ± 4	52 ± 4
	InsPost + 5-HD	12.3 ± 1.2	9.2 ± 0.6^#	100 ± 11	76 ± 9	57 ± 6
	MPG	13.5 ± 0.8	7.2 ± 0.3^#	106 ± 7	98 ± 7¤	71 ± 7¤
	InsPost + MPG	15.2 ± 0.8	9.1 ± 1.4^#	98 ± 8	91 ± 4¤	65 ± 6¤
RPP	Ctr	45655 ± 3735	21795 ± 3053^#	54 ± 8	52 ± 9	33 ± 5
rate*mmHg	InsR + 1	41279 ± 6539	23915 ± 4737^#	66 ± 10	58 ± 13	53 ± 11
	InsR + 5	37140 ± 3699	18046 ± 3036^#	51 ± 8	50 ± 6	35 ± 5
	InsR + 15	42351 ± 4563	22149 ± 2006^#	87 ± 17	83 ± 9¤	58 ± 6
	InsPost	33569 ± 4976	16900 ± 2035^#	66 ± 7	61 ± 6	50 ± 9
	IPost	37436 ± 3840	23237 ± 4294^#	66 ± 8	66 ± 9	47 ± 11
	SH-6	47550 ± 3254	18960 ± 3175^#	40 ± 17	22 ± 12	24 ± 11
	InsPost + SH-6	39680 ± 6016	13295 ± 3961^#	50 ± 13	55 ± 10	48 ± 6
	Rapa	40987 ± 6551	16777 ± 3551^#	75 ± 8	79 ± 6¤	56 ± 8
	InsPost + Rapa	36040 ± 7762	18165 ± 5037^#	76 ± 15	85 ± 18¤	54 ± 7
	5-HD	41611 ± 2256	27417 ± 2655^#	53 ± 8	60 ± 4	43 ± 6
	InsPost + 5-HD	37813 ± 5064	30969 ± 3134	84 ± 17	76 ± 17¤	53 ± 12
	MPG	51060 ± 5290	21235 ± 1771^#	66 ± 6	71 ± 7¤	56 ± 7
	InsPost + MPG	51989 ± 4202	20253 ± 5172^#	63 ± 7	70 ± 4¤	49 ± 4

Values represent means ± s.e.m.

Ctr = Ischemia–reperfusion control; IPost = ischemic postconditioning, 3 × 30 s; InsPost = insulin postconditioning, 3 × 30 s; Ins_{R + 1}/Ins_{R + 5}/Ins_{R + 15} = 1-, 5-, or 15-min insulin administration at immediate reperfusion; SH-6 = Akt inhibitor at reperfusion; Rapa = p70s6k inhibitor at reperfusion; MPG = ROS scavenging at reperfusion; 5-HD = mK_ATP channel inhibition at reperfusion.

#P < 0.05 versus corresponding Stab. ¤P < 0.05 versus SH-6, *P < 0.05 versus Ctr.

Insulin postconditioning increases Akt and p70s6 phosphorylation during early reperfusion

Insulin is known to activate the pro-survival kinase Akt, and we have previously shown that reperfusion therapy with insulin maintains Akt phosphorylation at Ser⁴⁷³ during early reperfusion (2). Insulin postconditioning also maintained Akt in a phosphorylated state as compared with vehicle-treated control hearts at 15-min reperfusion (Figure 4A and B). The degree of Akt phosphorylation at Ser⁴⁷³ by InsPost was comparable to Akt phosphorylation induced by IPost and Ins_{R + 15} (Figure 4A and B). The SH-6 inhibitor is a phosphatidylinositol (PI) analog that inhibits Akt without affecting the activity of the upstream kinase PDK-1 (22) and also prevents phosphorylation of Akt at Ser⁴⁷³ (9,23). Administration of SH-6 significantly reduced Akt phosphorylation in hearts exposed to insulin therapy at immediate reperfusion (Figure 4C), validating SH-6 as an Akt inhibitor.

Figure 4. Phosphorylation status of myocardial Akt in hearts exposed to insulin postconditioning. (A) Representative immunoblots of Akt phosphorylation (Ser⁴⁷³) and total Akt, demonstrating the effects of treatment with InsPost compared with ischemic postconditioning (IPost) and 15 min of insulin administration at reperfusion (Ins_{R + 15}). Ctr_B = baseline KHB perfusion for 20 min. Ctr_Rep = ischemia–reperfusion control. GAPDH verifies equal loading. (B) Densitometric analysis of total and phosphorylated Akt immunoblots expressed in AU where p-Akt was expressed as a ratio of total Akt with Ctr_B = 1. The cardioprotective effect of Ins_{R + 15}, IPost, and InsPost was associated with sustained phosphorylation of Akt for the first 15 min of ischemic reperfusion. (C) Blots from insulin-treated hearts confirm that SH-6 indeed abrogates Akt phosphorylation. Bars represent means ± s.e.m (n ≥ 3). *p < 0.05 versus Ctr_Rep and Ctr_B.

Multiple and divergent pathways activated by Akt are postulated to promote cell survival (24). The mTOR/p70s6k signaling pathway is thought to regulate translational protein synthesis and is central in mammalian cellular growth (25). This signaling pathway also conveys a pro-survival effect (25,26) and is activated by insulin administrated at the time of reperfusion (2,3). In concordance with the elevated levels of phosphorylated Akt, InsPost hearts showed an approximately four-fold increase in p70s6k phosphorylation at Ser⁴²¹/Thr⁴²⁴ as compared with control hearts, as did IPost and Ins_{R + 15} (p < 0.05, Figure 5A and B).

Figure 5. Phosphorylation status of myocardial p70s6k in hearts exposed to insulin postconditioning. (A) Representative immunoblot of p70s6k phosphorylation (Thr⁴²¹/Ser⁴²⁴) and total p70s6k demonstrating the effects of InsPost compared with IPost and 15 min of insulin administration at reperfusion (Ins_{R + 15}). Ctr_B = baseline KHB perfusion for 20 min. Ctr_Rep = ischemia–reperfusion control. GAPDH verifies equal loading. (B) Densitometric analysis of total and phosphorylated p70s6k immunoblots expressed in AU were p-p70s6k was expressed as a ratio of total p70s6k with Ctr_B = 1. The cardioprotective effect of Ins_{R + 15}, IPost, and InsPost was associated with sustained phosphorylation of p70s6k for the first 15 min of ischemic reperfusion. Bars represent means ± s.e.m (n ≥ 3). *p < 0.05 versus Ctr_Rep and Ctr_B.

Discussion

In summary, our data demonstrate that short, continuous insulin treatment for 5 or 1 min at reperfusion fails to offer cardioprotection, while intermittent insulin infusions, given as a mimetic of IPost, afforded cardioprotection to the same extent as Ins_{R + 15} or IPost. Moreover, our data suggest that this cardioprotection is mediated via Akt/p70s6k and mK_ATP/ROS-dependent RISK survival signaling at reperfusion.

Previously, it has been shown that continuous insulin administration at reperfusion substantially phosphorylated myocardial Akt when the tissue was isolated immediately after 2, 5, or 15 min (2). As Akt seems to be a critical kinase for cardioprotection (reviewed in (27,28)), we therefore expected insulin therapy to be effective in shorter treatment protocols. In the present study, however, all tissues were harvested 15 min into the reperfusion period, and tissue from the shorter insulin treatments showed reduced myocardial Akt phosphorylation (Figure 2B). The blunted Akt phosphorylation could explain the loss of cardioprotection observed in these groups, and taken together with previous reported results, it seems that shorter insulin treatment gives a non-sustained Akt phosphorylation where Akt is phosphorylated acutely, but dephosphorylated upon withdrawal of insulin treatment. How this occurs, however, is yet to be determined.

Recent evidence suggests that co-activation of Akt and AMPK is essential for the cardioprotection afforded by IPC via enhanced myocardial glucose uptake at postischemic reperfusion (29). Ji et al. further show that the reduced ability of IPC to mediate cardioprotection in the streptozotocin-diabetic rat heart could be reversed by insulin, indicating an intrinsic metabolic capacity that could be stimulated by insulin stimulation. Whether pharmacological postconditioning with insulin is dependent on glucose uptake at early reperfusion needs further elucidation. However, previous results have indicated that insulin-mediated Akt-dependent cardioprotection is independent of the presence of glucose at reperfusion, and further, that sarcolemmal glucose transporter type 4 or GLUT4 levels were unchanged in the presence or absence of insulin administration at early reperfusion (2).

In contrast, intermittent infusion of insulin (InsPost), administered as a mimetic of IPost 3 × 30 s, offers equally strong Akt and p70s6k phosphorylation and a similar reduction in IS as Ins_{R + 15}, despite a total treatment time of only 1.5 min. It would therefore seem that this cyclic administration of insulin somehow maintains Akt in a phosphorylated state. These findings are in concordance with the study from Penna et al. (10), where a 3-min continuous bradykinin infusion at reperfusion failed to offer cardioprotection, but a 5 × 10 s intermittent infusion succeeded. It is also in concordance with our own findings, where effluents from IPC hearts were cardioprotective when administered for 3 × 30 s at reperfusion (11). In Penna et al.'s study the same pattern could also be mimicked by specific activation of mK_ATP channels by diazoxide for 5 × 10 s, but not by continuous administration for 3 min.

Controversy exists on whether insulin cardioprotection is mediated via mK_ATP channels and ROS. LaDisa et al. reported that glucose-insulin-potassium or GIK cardioprotection could be abolished by non-specific inhibition of K_ATP channels (30), while Baines et al. found no inhibitory effect of either mitochondrial K_ATP channels or PKCε on insulin preconditioning (18). Juhazhova et al., in an extensive set of experiments, categorized insulin as a “non-sweller,” that is, its protection was independent of mK_ATP channels, ROS and PKCε translocation, and the associated mitochondrial swelling this would lead to (19). Our results contrast these and indicate that, at least in the setting of reperfusion, intermittent insulin infusions convey cardioprotection via mK_ATP channels. An important limitation of this interpretation is a series of reports from Hanley et al. showing that 5-HD can be beta-oxidized, and thus create a bottleneck for fatty acid oxidation (31–33), suggesting an alternative explanation for the inhibitory effects of 5-HD. Our preparation, however, used glucose as the only metabolic substrate, so it is uncertain whether this effect pertains to the present study.

The role of ROS in myocardial ischemia–reperfusion damage is well established, and IPost and other cardioprotective treatments have been demonstrated to inhibit ROS generation at reperfusion (13,34,35). However, several studies have reported that pharmacological or mechanical cardioprotective treatments are dependent on ROS signaling (10,12,36,37), creating what Hausenloy et al. coined the free radical signaling paradox (28). Insulin has previously been shown to be cardioprotective by reducing oxidative/nitrative stress (38), and although studies from adipocytes indicate that insulin elicits a burst of hydrogen peroxide that enhances insulin receptor tyrosine phosphorylation, and thus, insulin signaling (39,40), no studies have to date investigated if cardioprotection from insulin treatment at reperfusion is dependent on ROS signaling. In the present study, addition of MPG at reperfusion abrogated the cardioprotection afforded by InsPost, while having no effect on IS alone. This supports a role for ROS signaling in insulin cardioprotection as well, albeit with insufficient evidence for what kind of subspecies, at which concentration and where in the signaling cascade this takes place. However, MPG does not appear to scavenge hydrogen peroxide (41), so this is unlikely to be an effect of hydrogen peroxide on the insulin receptor. A few have reported reduced IS when treating with MPG alone (42–44). The present study, in concordance with several others (10,14,36,45), found no such effect, and this could be related to differences in concentration of MPG or the administration protocol.

There are some important limitations to consider: The Langendorff heart perfusion is isolated from blood, immune system, and neuro-hormonal influence, thus our results could be different in the in vivo setting. We found no correlation between IS and functional recovery at reperfusion. This might be surprising, especially considering the inotropic effect of insulin. However, it is in concordance with others, who have used even higher concentrations of insulin (46), and in general, the lack of correlation between heart function and IS has been observed by several others and is believed to be due to persistent stunning (21,47).

In conclusion, intermittent insulin treatment of 3 × 30 s offers strong cardioprotection in the ex vivo rat heart by activating Akt/p70s6k and mK_ATP/ROS-dependent RISK survival signaling at reperfusion. In contrast, longer but continuous insulin reperfusion therapy for 5 min offers no cardioprotection.

Acknowledgements

Ingrid Strand is thanked for technical assistance. AKJ is supported by the Bergen Medical Research Foundation, L. Meltzer Foundation, Norwegian Council on Cardiovascular Diseases, and University of Bergen Heart Foundation. LB is supported by the Norwegian Council on Cardiovascular Diseases. EH is supported by Norwegian Research Council.

Declaration of interest: The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper.