A new calpain inhibitor protects left ventricular dysfunction induced by mild ischemia-reperfusion in in situ rat hearts

Abstract

We have previously indicated that a new soluble calpain inhibitor, SNJ-1945 (SNJ), attenuates cardiac dysfunction after cardioplegia arrest-reperfusion by inhibiting the proteolysis of α-fodrin in in vitro study. Nevertheless, the in vivo study design is indispensable to explore realistic therapeutic approaches for clinical use. The aim of the present in situ study was to investigate whether SNJ attenuated left ventricular (LV) dysfunction (stunning) after mild ischemic-reperfusion (mI-R) in rat hearts. SNJ (60 μmol/l, 5 ml i.p.) was injected 30 min before gradual and partial coronary occlusion at proximal left anterior descending artery. To investigate LV function, we obtained curvilinear end-systolic pressure–volume relationship by increasing afterload 60 min after reperfusion. In the mI-R group, specific LV functional indices at midrange LV volume (mLVV), end-systolic pressure (ESP_mLVV), and pressure–volume area (PVA_mLVV: a total mechanical energy per beat, linearly related to oxygen consumption) significantly decreased, but SNJ reversed these decreases to time control level. Furthermore, SNJ prevented the α-fodrin degradation and attenuated degradation of Ca²⁺ handling proteins after mI-R. Our results indicate that improvements in LV function following mI-R injury are associated with inhibition of the proteolysis of α-fodrin in in situ rat hearts. In conclusion, SNJ should be a promising tool to protect the heart from the stunning.

Introduction

It is well known that one of underlying mechanisms for ischemic-reperfusion (I-R) injury is Ca²⁺ overload resulting from increased Ca²⁺ influx mediated via reverse-mode Na⁺–Ca²⁺ exchanger (NCX) [1–4]. We have previously reported that reperfusion injury after KCl cardioplegic cardiac arrest leads to Ca²⁺ overload and the resultant left ventricular (LV) dysfunction similar to I-R injury [5]. The mechanisms of Ca²⁺ overload in this model are likely accumulation of intracellular Na⁺ and subsequent activation of reverse-mode NCX activity [6]. However, previous studies in our laboratory suggest that proteolysis of the cytoskeletal protein α-fodrin by calpains may also play a role in LV dysfunction [5, 7, 8].

Calpain inhibition was found to prevent the proteolysis of α-fodrin due to reperfusion injury after global ischemia [8] and after KCl cardioplegic cardiac arrest [9]. It has been proposed that α-fodrin maintains the integrity of the plasma membranes as a constituent of the membrane skeleton [10, 11]. Therefore, it seems likely that the degradation of α-fodrin in membrane fractions would alter the properties of ion channels [12]. Indeed, the possibility that disruption of cytoskeletal proteins inactivates L-type Ca²⁺ channels has been reported [13].

Recently, we have reported that a novel calpain inhibitor, {(1S)-1[(((1S)-1-benzyl-3-cyclopropyl-amino-2,3-di-oxopropyl) amino) carbonyl]-3-methylbutyl} carbamic acid 5-methoxy-3-oxapentyl ester (SNJ-1945; SNJ) attenuates left ventricular (LV) dysfunction induced by reperfusion-injury after cardioplegic cardiac arrest, by inhibiting the proteolysis of α-fodrin, without any effects on protein levels of sarcoplasmic reticulum Ca²⁺ ATPase (SERCA2a) and L-type Ca²⁺ channel (LTCC) [9]. This novel calpain inhibitor (SNJ) has been shown to have good aqueous solubility, good plasma exposure, and good tissue penetration in rats and monkeys [14, 15]. Intraperitoneal administration of SNJ (160 mg/kg) for 14 days produced no obvious toxicity or abnormalities in rats [14]. Furthermore, it has been reported that SNJ was effective against cerebral ischemia-induced damage [16], but there have been no reports on its efficacy against cardiac I-R injury in in situ hearts. Only our recent study has reported that SNJ was effective against reperfusion injury after cardioplegic cardiac arrest, though in in vitro hearts [9].

Accordingly, the in vivo study design is indispensable to explore realistic therapeutic approaches for clinical use, since in vivo cardiac hemodynamics is regulated by the autonomic nervous system and endocrine gland. These regulations are lacking in in vitro hearts. The aim of the present study was to investigate the cardioprotective effects of SNJ against mild I-R (mI-R)-induced injury after gradual and partial coronary occlusion at the proximal left anterior descending artery in rat in situ hearts using analysis of LV mechanical work.

Methods

The investigation conformed with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996), and reviewed and approved by the animal care and use committee of Nara Medical University.

Surgical preparation

The trachea was intubated, and the rat was ventilated with room air under pentobarbital (50 mg/kg, i.p.) anesthesia. Body temperature was maintained normal using the warming plate. The chest was opened, and a conductance catheter (1.5 Fr) [17] was introduced into the LV through an apical stab to obtain reliable LV volume (LVV) signal. A 1.5-Fr pressure catheter was also inserted through the apex into the LV to obtain reliable LV pressure (LVP) signal. Anesthetic level was sustained with pentobarbital intravenous infusion at 0.5 mg kg⁻¹ h⁻¹ throughout the experiment.

Mild ischemic-reperfusion heart preparation

Coronary gradual and partial occlusion was performed in 10 s by ligation of proximal left anterior descending artery and attached 1-mm-diameter soft tube using a bulldog clamp at the knot of suture (Fig. 1a) under monitoring LV pressure–volume (P–V) loop. During this occlusion, the P–V loop moderately shifted rightward and each stroke volume gradually decreased accompanied with increases in heart rate, but end-systolic pressure (ESP) hardly changed (Fig. 1b). The rats did not cause lethal arrhythmia because of adequate perfusion pressure. The total mean survival rate of the protocol was 80.0 ± 7.1 %; 76.9 % in mI-R and 88.2 % in SNJ + mI-R group in the present mI-R injury model. No differences in the occurrence of arrhythmia in the SNJ + mI-R group were observed compared with that in the mI-R group. This preparation is a stunned heart model [18, 19] and appropriate for our LV functional analysis because no myocardial infarctions were observed 20 min after reperfusion (Fig. 1d) as previously reported [19], although the mild ischemia period persisted for 30 min in the present protocol.

Fig. 1

Mild ischemic-reperfusion (mI-R) heart preparation. a Coronary gradual and partial occlusion was performed in 10 s by ligation of proximal left anterior descending artery and attached 1-mm-diameter soft tube (as buffer) using a bulldog clamp at the knot of 4-0 polypropylene suture under monitoring LV pressure–volume (P–V) loop. b Typical left ventricular (LV) pressure–volume loops during gradual and partial coronary occlusion to make mild ischemic-reperfusion heart preparations. c A typical macrograph of the preparation of time control. d A typical LV macrograph of the mI-R preparation obtained 20 min after mI-R. No myocardial infarction was observed. e A typical macrograph of the preparation of acute myocardial infarction (MI) obtained 20 min after severe I-R. The infarct tissue was indicated by the white dotted boundary. It appeared white even by naked eyes. The non-injured tissue was indicated by an arrow outside the infarct tissue

Measurements of LV stroke volume and pressure volume area (PVA)

Forty-one male Wistar rats (7–12 weeks) were used in the present experiments. The experimental protocol is shown in Fig. 2. LV functions were analyzed in 4 randomly divided groups, i.e., time control, SNJ, mI-R, and SNJ + mI-R. The sample number of SNJ + mI-R was about three times higher than the other groups because it was the most important group to evaluate the effect of SNJ on mI-R.

Fig. 2

Experimental protocols of the four groups. 1 Time control group (n = 5); 2 SNJ group (n = 5); 3 mI-R group (n = 6); 4 SNJ + mI-R group (n = 15)

LV pressure–volume measurements at “Functional analysis” in Fig. 2 were performed 60 min after coronary reperfusion. The detailed methods using conductance and pressure catheters have been described in previous reports [20–23]; LV stroke volume (SV) [=LV end-diastolic volume (EDV) − LV end-systolic volume (ESV)], end-systolic pressure at ESV (ESP_ESV) and Ea (arterial effective elastance; ESP_ESV SV⁻¹) calculated from steady-state P–V loops were evaluated (Fig. 3a). At midrange LVV (mLVV), end-systolic pressure (ESP_mLVV) and systolic pressure–volume area (PVA_mLVV, an appropriate index for evaluating cardiac total mechanical energy per beat) calculated from end-systolic pressure–volume relationship (ESPVR) curve, were also evaluated (Fig. 3b).

Fig. 3

Evaluation of left ventricular (LV) function by end-systolic pressure–volume relation (ESPVR) using pressure–volume (P–V) measurements. Enveloping curve of end-systolic pressure at end-systolic volume (ESP _ESV ) data of several P–V loops during aortic gradual occlusion (gradually increasing afterload) corresponds to ESPVR. SV stroke volume, ESV end-systolic volume, EDV end-diastolic volume, Ea effective arterial elastance. mLVV midrange LV volume, V ₀ LV volume intercept, ESP _mLVV end-systolic pressure at mLVV, PVA _mLVV systolic pressure–volume area at mLVV

The LV end-systolic P–V data on the upper left shoulder of multiple P–V loops during increasing afterload by aortic gradual occlusion were plotted and fitted by the method of least squares using the following equation: LVP = A{1 − exp[−B(LVV − V ₀)]}, where A and B are fitted parameters and V ₀ is systolic unstressed volume [20, 23–25]. Aortic gradual occlusion was performed to tighten a string occluder placed loosely around the ascending aorta until EDV increased slightly as previously reported [20, 21].

The PVA as a function of LVV was obtained by integrating the above exponential function from the extrapolated V ₀ along the volume axis: PVA = A(LVV − V ₀) − A{1 − exp[−B (LVV − V ₀)]}/B [20, 23–25]. PVA is linearly related to myocardial oxygen consumption per beat [22]. Therefore, PVA_mLVV is a better cardiac functional index in terms of cardiac mechanoenergetics. In the present study, we calculated mLVV, which was the value of [V ₀ + (maximum ESV − minimum ESV) on the ESPVR × 1/2] from all P–V loops in each group [20, 24, 25].

Drugs

A synthetic, water-soluble calpain inhibitor, SNJ-1945, (SNJ: provided from Senju Pharmaceutical, Kobe, Japan) was dissolved in Lactate Ringer solution at 60 μmol/l. Other lipophilic calpain inhibitors, such as MDL28170, calpain inhibitor-3 [26], and calpain inhibitor-1 [8] had been previously dissolved in DMSO, which has antioxidant properties. In the present study, we did not use DMSO to dissolve SNJ and thus our study did not include a vehicle control group. We administered SNJ solution at a dose of 0.387 mg/kg on average (60 μmol/l, 5 ml, i.p.); 60 μmol/l is the limit of its solubility [9]. SNJ has a β₁ receptor stimulating action [27], but in the present study this action was not detected (Tables 2, 3).

Polyacrylamide gel electrophoresis and immunoblottings of 150- and 145-kD fragments of α-fodrin (250-kD), L-type Ca²⁺ channel (LTCC) and sarcoplasmic reticulum Ca²⁺ ATPase (SERCA2a)

Membrane proteins were isolated from the LV wall of each frozen heart stored at −80 °C after the mechanoenergetic studies. The frozen hearts (0.1 g) were homogenized in 1 ml of the STE buffer containing 0.32 mol/l sucrose, 10 mmol/l Tris–HCl, pH 7.4, 1 mmol/l EGTA, 5 mmol/l NaN₃, 10 mmol/l β-mercaptoethanol, 20 mmol/l leupeptin, 0.15 mmol/l pepstatin A, 0.2 mmol/l phenylmethanesulfonyl fluoride, and 50 mmol/l NaF with a Polytron homogenizer (NS-310E; Micotec) and centrifuged at 1,000g for 10 min. The supernatants were centrifuged at 100,000g for 60 min at 4 °C. The 100,000g pellets were cellular membrane fractions. Membrane proteins (50 μg/lane) were subjected to SDS-poly-acrylamide gel electrophoresis, followed by immunoblotting of 150- and 145-kD fragments of α-fodrin (240-kD)[8, 28, 29], LTCC, and SERCA2a [9]. The membranes were blocked (4 % Block Ace; Dainippon Pharmaceutical, Osaka, Japan) and then incubated with 2,000-fold diluted primary antibody against anti-α-fodrin (1:2,000 dilution, Biohit; Genex), anti-LTCC antibody (1:300 dilution; Alomone Labs, Israel) and anti-SERCA2a antibody (1:1,000 dilution; Affinity Bio Reagents). The antigens were detected by the luminescence method (ECL Western blotting detection kit, Amersham) with peroxidase-linked anti-mouse IgG (1:2,000 dilution) or peroxidase-linked anti-rabbit IgG (1:2,000). The amounts of membrane proteins were determined to obtain the linear response of ECL-immunoblot. After immunoblotting, the film was scanned with a scanner, and the intensity of the bands was calculated by NIH image analysis. The intensity ratio of the 145- and 150-kD bands versus the 240-kD band (α-fodrin) was expressed in an arbitrary unit and compared with that in the time control (average = 1.0).

Statistics

Comparison of paired and unpaired individual values was performed by paired and unpaired t test, respectively. Multiple comparisons were performed by one-way analysis of variance (ANOVA) with post hoc Bonferroni’s test or Tukey HSD test. A value of P < 0.05 was considered statistically significant. All data are expressed as the mean ± SD.

Results

All cardiac weights data comparing among different 4 groups are shown in Table 1. The smaller BWs in the mI-R (7–11 weeks) and SNJ + mI-R (7–11 weeks) groups were due to younger age than that in the time control (10–12 weeks). All mean pre- and post-data comparing among the different 4 groups are shown in Tables 2 and 3. There were no significant differences in any of the hemodynamic indices of ESV, EDV, Ea, and ESP_ESV except for post-ESP_ESV in mI-R among the time control, mI-R, SNJ + mI-R, and SNJ groups (Table 2).

Table 1 Comparison of cardiac weights among time control, mI-R, SNJ + mI-R, and SNJ groups

Table 2 Hemodynamics in time control, mI-R, SNJ + mI-R, and SNJ groups

Table 3 Variables of left ventricular mechanics in time control, mI-R, SNJ + mI-R, and SNJ groups

Effects of SNJ on LV functions

In SNJ group, each mean SV, ESP_ESV, Ea, ESP_mLVV, and PVA_mLVV was not significantly different from each baseline data for 120 min (Tables 2, 3), indicating no β₁ receptor stimulating action was detected under the present experimental conditions.

Effects of SNJ on LV mechanoenergetics after mI-R

A representative set of P–V loops and ESPVRs during aortic occlusion in each group is shown in Fig. 4. In mI-R group, ESPVR markedly shifted downward 60 min after mI-R (Fig. 4b). In contrast, the P–V loops and ESPVR in SNJ + mI-R (Fig. 4c) 30 min before and 60 min after mI-R were similar to those in the time control (Fig. 4a).

Fig. 4

A representative set of P–V loops and ESPVRs during aortic occlusion in each group. Each onset and offset P–V loop and ESPVRs at baseline and 120 min after baseline are indicated by gray solid and black dotted lines in the time control group (a), respectively. Each onset and offset P–V loop and ESPVRs 30 min before and 60 min after mI-R are indicated by gray solid and black dotted lines in the mI-R (b) and SNJ + mI-R groups (c), respectively. Solid arrow in b indicates downward shift of ESPVR. V ₀ LV volume intercept

Mean fitting parameters, A and B of ESPVR, in each group are shown in Table 3. In the mI-R group, the post-fitting parameter A significantly (P < 0.05) decreased compared with the pre-one. This was consistent with a marked downward-shift of ESPVR after mI-R (Fig. 4b). A representative set of PVA–LVV relationship curves calculated using A and B in each group is shown in Fig. 5. Post-mI-R PVA is decreased at any LVV in the mI-R group, although the larger the LVV, the larger the decrease in PVA that is observed (Fig. 5b).

Fig. 5

PVA–LVV relationship curves in the time control (a), mI-R (b) and SNJ + mI-R groups (c), respectively. PVA–LVV relationship curves 30 min before and 60 min after mI-R are indicated by gray solid and black dotted lines, respectively. Solid arrow in b indicates downward shift of PVA–LVV relation curve. V ₀ LV volume intercept

The mean absolute values of ESP_mLVV, PVA_mLVV and ESP_ESV in each group are also shown in Tables 2 and 3. Post-mI-R mean values of ESP_mLVV and PVA_mLVV in the mI-R group were significantly (P < 0.05) smaller than those in the time control group whereas those in the SNJ + mI-R group were significantly (P < 0.05) larger than those in the mI-R group. Post-mI-R mean values of ESP_mLVV, PVA_mLVV, and ESP_ESV in the mI-R group were significantly (P < 0.05) smaller than the pre-mI-R mean values whereas those in the SNJ + mI-R group were not significantly different from the pre-mI-R mean values.

The mean % of baseline data in ESP_mLVV, PVA_mLVV, and ESP_ESV significantly (P < 0.05) decreased compared to the time control group (=100 %) and that in SV moderately decreased but not significantly in the mI-R group (Fig. 6b). In the SNJ + mI-R group, the mean % of baseline data in ESP_mLVV, PVA_mLVV, and ESP_ESV significantly (P < 0.05) increased from those in the mI-R group to time control level (Fig. 6a, c, d).

Fig. 6

Mean percent of baseline data in ESP_mLVV (a), SV (b), PVA_mLVV (c). and ESP_ESV (d) 60 min after mI-R in each group. Data show the mean with SD. *P < 0.05 versus time control, ^# P < 0.05 versus mI-R

Immunoblotting of 150- and 145-kD fragments of α-fodrin (240-kD)

Figure 7 shows immunoblottings of 240-kD α-fodrin and 145- and 150-kD α-fodrin proteolytic fragments in the time control, mI-R, and SNJ + mI-R groups. The mean amounts of the 145- and 150-kD fragments in the mI-R group were significantly (P < 0.05) larger than those in the time control group. The mean amounts of the 145- and 150-kD fragments in the SNJ + mI-R group were significantly (P < 0.05) smaller than those in the mI-R group. α-fodrin mean degradation 120 min after SNJ did not increase (1.16 ± 0.34 fold; n = 5) compared with the time control group (n = 3). The results indicated that α-fodrin degradation after mI-R was completely prevented by SNJ treatment.

Fig. 7

Ratio of 150- and 145-kD proteolytic fragments by calpain to 240-kD intact α-fodrin in the time control, mI-R, and SNJ + mI-R groups. Upper panel each representative band in the time control, mI-R, and SNJ + mI-R groups. All bands were obtained from the same SDS–polyacrylamide gel. Lower panel mean fold change of 150- and 145-kD fragments to 240-kD α-fodrin compared to mean value (=1.0) in the time control (n = 9) group among the mI-R (n = 9) and SNJ + mI-R (n = 5) groups. Data show the mean with SD. *P < 0.05 versus time control, ^# P < 0.05 versus mI-R. Immunoblotting of 150- and 145-kD fragments of α-fodrin (240-kD)

Figure 8a shows immunoblottings of LTCC in the time control, mI-R, and SNJ + mI-R groups. The mean amount of LTCC protein in mI-R was significantly (P < 0.05) smaller than that in the time control group. The mean amounts of LTCC protein in the SNJ + mI-R group were not significantly different from those in the time control and mI-R groups. The results indicated that moderate prevention of LTCC degradation after mI-R was attained by SNJ treatment.

Fig. 8

Comparison of LTCC protein levels (a) and SERCA2a (b) among the time control, mI-R, and SNJ + mI-R groups. Upper panels each representative band in the time control, mI-R, and SNJ + mI-R groups. All bands of LTCC (a) and SERCA2a (b) were obtained from the same SDS–polyacrylamide gels, respectively. Lower panels mean fold changes of LTCC (a) and SERCA2a (b) compared to the mean value (=1.0) in the time control (n = 6) group among the mI-R (n = 6) and SNJ + mI-R (n = 6) groups. Data show the mean with SD. *P < 0.05 versus time control

Figure 8b shows immunoblottings of SERCA2a in the time control, mI-R, and SNJ + mI-R groups. The mean amount of SERCA2a protein in the mI-R group was significantly (P < 0.05) smaller than that in the time control group. The mean amount of SERCA2a protein in the SNJ + mI-R group increased but not significantly (P > 0.05) and remained significantly (P < 0.05) smaller than that in the time control group, indicating partial prevention of SERCA2a degradation after mI-R was attained by SNJ treatment.

Discussion

We have previously demonstrated that, in in vitro rat hearts, the lipophilic calpain inhibitor, calpain inhibitor-1, attenuates α-fodrin proteolysis and cardiac dysfunction due to I-R injury [8] and acute Ca²⁺ overload [7]. Recently, we have also demonstrated that a new water-soluble calpain inhibitor, SNJ, exerts similar cardioprotective actions in in vitro rat hearts after KCl (30 mEq) cardioplegia arrest-reperfusion [9]. In the present study, we found that, in in situ rat hearts, pretreatment of SNJ also completely prevented cardiac dysfunction and α-fodrin degradation induced by mI-R.

We have previously reported that proteolysis of a cytoskeleton protein, α-fodrin, is found without proteolysis of ankyrin, connexin 43, and troponin I in high Ca²⁺-infusion-induced Ca²⁺ overloaded contractile failing hearts associated with the impairment of the total Ca²⁺ handling in the excitation–contraction (E–C) coupling [7]. It seems likely that α-fodrin is the most sensitive membrane protein to Ca²⁺ overload. To investigate the cardiac protective effect of SNJ mediated via calpain inhibition, we focused on proteolysis of α-fodrin in the present mI-R injury model, because there is a close correlation between the membrane α-fodrin proteolysis and the impairment of the total Ca²⁺ handling [7–9].

In addition, recent studies have revealed that calpain activation induced by I-R injury causes degradation of Ca²⁺-handling proteins such as LTCC [30] and SERCA2a [26, 30]. Thus, we also examined the degradation of Ca²⁺-handling proteins such as LTCC and SERCA2a in the present mI-R injury model.

It has been proposed that fodrin maintains the integrity of the plasma membranes as a constituent of the membrane skeleton [10]. Therefore, it seems likely that the degradation of fodrin in membrane fractions would alter the properties of ion channels [12]. From the possibility that disruption of cytoskeletal proteins inactivates LTCC [13, 31], we speculate that the linkage of the LTCC to the membrane fodrin tethers the channel in place, which somehow modulates the basal activity of the channel, and a loss of the linkage may impair its regulation. Therefore, the calpain inhibitor may have protected against LV dysfunction by preserving the structural integrity of the LTCC in the cell membrane [13, 31], as in in vitro rat hearts after cardioplegia arrest-reperfusion [9].

Although the degradation of LTCC and SERCA2a was not identified in reperfusion injury after cardioplegia cardiac arrest in in vitro rat hearts [9], in the present mild I-R injury in in situ rat hearts, the degradation of LTCC and SERCA2a was identified. SNJ attenuated the degradation of LTCC and SERCA2a, though not completely. Therefore, SNJ attenuated cardiac dysfunction, possibly and partially by preventing the dysfunction of Ca²⁺ handling proteins, such as LTCC and possibly SERCA2a. Although the protein level of SERCA2a remained low, large amounts of SERCA2a protein are originally expressed in the rat myocardium, and thus this level may be sufficient if some functional compensation such as activation of phosphorylation of phospholamban occurs mediated via the autonomic nervous system and/or hormonal control system.

We have reported a marked β₁ action of SNJ in the blood-perfused excised rat hearts [27]. However, in the present in situ heart study, no marked β₁ action of SNJ was observed (see Tables 2, 3). The differences between in vitro and in situ studies might have been caused by the different application methods of SNJ. In in vitro hearts, the direct infusion of SNJ into the coronary artery was performed, whereas in in situ hearts, the indirect intraperitoneal injection of SNJ was performed.

In conclusion, a new water-soluble (thus more beneficial for clinical use) calpain inhibitor, SNJ, should be a promising tool for pharmacotherapy to exert cardioprotective actions in clinical setting. Present in situ mild I-R injury model hearts correspond to stunned myocardium around the necrotic focus in acute myocardial infarction in clinical setting. Upon direct percutaneous transluminal coronary angioplasty (PTCA) for acute myocardial infarction, the stunned myocardium generates arrhythmia and myocardial contractile dysfunction. Pretreatment with SNJ upon PTCA would prevent in vivo heart from generating arrhythmia and myocardial dysfunction.

Limitation of the study

The measurement of the expression levels of other Ca²⁺ handling proteins, such as Na⁺–Ca²⁺-exchanger, phospholamban and ryanodine receptor would be expected for better understanding of the function of Ca²⁺ handling in E–C coupling in the future work.