Transgenic overexpression of the SUR2A-55 splice variant in mouse heart reduces infract size and promotes protective mitochondrial function.

ATP-sensitive potassium channels found in both the sarcolemma (sarcKATP) and mitochondria (mitoKATP) of cardiomyocytes are important mediators of cardioprotection during ischemic heart disease. Sulfonylurea receptor isoforms (SUR2), encoded by Abcc9, an ATP-binding cassette family member, form regulatory subunits of the sarcKATP channel and are also thought to regulate mitoKATP channel activity. A short-form splice variant of SUR2 (SUR2A-55) was previously shown to target mitochondria and display diaxoxide and ATP insensitive KATP activity when co-expressed with the inward rectifier channels Kir6.2 and Kir6.1. We hypothesized that mice with cardiac specific overexpression of SUR2A-55 would mediate cardioprotection from ischemia by altering mitoKATP properties. Mice overexpressing SUR2A-55 (TGSUR2A-55) in cardiomyocytes were generated and showed no significant difference in echocardiographic measured chamber dimension, percent fractional shortening, heart to body weight ratio, or gross histologic features compared to normal mice at 11-14 weeks of age. TGSUR2A-55 had improved hemodynamic functional recovery and smaller infarct size after ischemia reperfusion injury compared to WT mice in an isolated hanging heart model. The mitochondrial membrane potential of TGSUR2A-55 mice was less sensitive to ATP, diazoxide, and Ca2+ loading. These data suggest that the SUR2A-55 splice variant favorably affects mitochondrial function leading to cardioprotection. These data support a role for the regulation of mitoKATP activity by SUR2A-55.

Introduction

Myocardial ATP-sensitive potassium channels (KATP) were first described more than 30 years ago [1] and are major contributors to pre-conditioning and cardioprotection [2, 3, 4]. They were first identified in the sarcolemma (sarcKATP) but soon after KATP channels were also reported in the inner membrane of mitochondria (mitoKATP) [5]. These channels are closed under normal physiologic conditions but open in response to acute stress leading to changes in cytoplasmic and mitochondrial function that promote cardioprotection [6, 7, 8].

The molecular composition of sarcKATP has been identified as a hetero-octomer of four Kir6.x (Kir6.1 or Kir 6.2) pore forming subunits and four sulfonylurea receptor (SUR1 or SUR2) regulatory subunits from the ABC binding cassette family of membrane proteins [9]. In addition, both splice variants of the SUR2 gene (SUR2A and SUR2B) form functional sarcKATP [10]. Different tissues display different subunit expression, giving rise to functionally distinct sarcKATP [11, 12, 13]. The cardiac ventricular sarcKATP is composed of SUR2A and Kir6.2 [14]. While there is a large body of literature studying the constituents of ventricular sarcKATP, the molecular makeup of mitoKATP in the heart has been more elusive. The most recent candidate for the pore forming subunit of mitoKATP is the renal outer medullary potassium channel (ROMK or Kir1.1) which was first supported using a proteomic analysis of the mitochondrial inner membrane [15, 16]. In addition, evidence exists that SUR2 gene products exist in mitochondria and play a role in mitoKATP channel activity [17, 18, 19].

We described a 55 kDa splice variant (SUR2A-55) formed by intra-exonic splicing that is located predominantly in mitochondria [20] which other labs have also identified [21]. KATP activity was recorded when SUR2A-55 was targeted to the sarcolemma by deleting a mitochondrial targeting sequence and when co-expressed with Kir6.2 and Kir6.1 in heterologous cells [20, 22]. We also found that mice engineered with targeted disruption of exon 14–18 of SUR2 (SUR2−/−) lacked full-length SUR2 but unexpectedly exhibited cardioprotection from ischemia reperfusion (IR) injury despite the absence of pinacidil, diazoxide, or glybenclamide sensitive sarcKATP currents [23, 24, 25]. The gene product that encodes the SUR2A-55 splice variant was not perturbed by the deletion used to generate SUR2−/− mice, and correspondingly mitochondria from these mice had favorable mitochondrial bioenergetics characterized by slight mitochondrial membrane depolarization at rest, insensitivity to ATP and diazoxide, increased reactive oxygen species (ROS), and tolerance to Ca2+ loading [26]. Furthermore, KATP currents in heterologous systems produced by co-expression of SUR2A-55 and Kir were less sensitive to block by ATP, more resistant to intracellular Ca2+, and less sensitive to KATP drugs [22] compared to full-length SUR2A. Recently, a second SUR2 knockout mouse was created by removing exon 5, an exon found in both SUR2 and SUR2A-55. This mutation produced early lethality and failure to transition from fetal to adult myocardial metabolism [27]. These mice, unlike the prior SUR2−/− mice, lacked both full-length SUR2 and the SUR2A-55 splice variant. From these mouse studies, we hypothesized that the SUR2A-55 splice variant plays an important role in mitochondrial function and cardioprotection.

To further understand the physiologic implication of SUR2A-55 we present here an investigation of novel mice overexpressing the SUR2A-55 splice variant in cardiac tissue, which we refer to as TGSUR2A-55.

Materials and methods

Construction of a cardiomyocyte-specific SUR2A-55 overexpressing mouse (TGSUR2A-55)

Transgenic mice expressing SUR2A-55 were generated in the University of Wisconsin Transgenic Animal facility. The 1.5 kb cDNA for SUR2A-55 bracketed by Xho I sites was cut with and ligated into the Sal I cloning site downstream of the 5.5 kb α-myosin heavy chain (α-MHC) cardiac-specific promoter and upstream from the 0.6 kb human growth hormone poly A [28]. The α-MHC-SUR2A-55 construct was linearized with Sac II and gel purified and injected into the male pronucleus of FVB single-cell mouse embryos and implanted into psuedopregnant females. Offspring were screened by Southern blotting of genomic DNA using a 0.7 kb probe corresponding to the first 0.7 kb of the SUR2A-55 cDNA. Chimeric mice were bred with FVB mice and colonies were maintained by breeding hemizygous transgenic mice with WT FVB mice in accordance with the standards set by the Animal Care and Use Committees at the University of Wisconsin. All animal experiments were performed in compliance with relevant laws and institutional guidelines and approved by the local animal ethics committee of the University of Wisconsin. Male mice were selected for study owing to evidence of sex differences in myocardial protection [29, 30]. Sex differences are likely to be important and will be the subject of additional study.

RNA analysis

Total RNA was extracted from mouse ventricle using TRIzol reagent (Invitrogen ThermoFisher, Waltham, MA) according to the manufacturer's instructions. First-strand cDNA was synthesized from 3 μg of total RNA using the SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen ThermoFisher, Waltham, MA). The quantitative real-time PCR (qPCR) reaction was performed using a 1:2 dilution of cDNA in a total of 20 μL with 10 μL Gene Expression Master Mix (Integrated DNA Technologies, Coralville, IA) and a final concentration of 500 nM forward primer, 500 nM reverse primer, and 250 nM probe. Each cDNA sample analyzed for gene expression by qPCR using the BioRad CFX 96 Real-Time PCR Detection System. Amplification data were analyzed using the delta-delta Cq method with the relative standard curve method to obtain reaction efficiencies of 90–105%. The conditions of qPCR for all primer/probe sets were as follows: an initial denaturation at 95 °C for 3 min followed by 40 cycles at 95 °C for 15 s and 52 °C for 1 min. Oligonucleotide primers were designed using published sequence data from GenBank with the aid of PrimeQuest Design Tool (Integrated DNA Technologies, Coralville, IA). The primer and hydrolysis probe sequences used are shown in Table 1. The mRNA expression was normalized to the levels of Actb and Gapdh mRNA as the reference genes.

Table 1

Primers used to detect KATP channel subunits (SUR2-55 and SUR2) and reference genes in qPCR studies.

Encoding protein name	Gene name	Reference sequence IDs	Primer and probe sequences (5′-3′)
SUR2-55	Abcc9	DQ991969DQ991970	F-TGTACTGGGTAATGGCCTTTATTAR-TGGAGGCTGTCAGAACAATGPRB-/56-FAM/AGATAGTCC/ZEN/GTCCTGACCTCCCAC/3IABkFQ/
SUR2	Abcc9	D86037.2D86038.2	F-CCACCACAACCTCCTCAATAAR-TGGTCGATGATGTTCGTATCAGPRB-/5HEX/AGGTTCTTT/ZEN/GATACCACACCGCTGG/3IABkFQ/
ACTB	Actb	NM_007393.5	F-GATTACTGCTCTGGCTCCTAGR-GACTCATCGTACTCCTGCTTGPRB-/56-FAM/CTGGCCTCA/ZEN/CTGTCCACCTTCC/3IABkFQ/
GAPDH	Gapdh	NM_008084.3	F-AATGGTGAAGGTCGGTGTGR-GTGGAGTCATACTGGAACATGTAGPRB-/56-FAM/TGCAAATGG/ZEN/CAGCCCTGGTG/3IABkFQ/

Western blot analysis

Total protein was isolated from mouse ventricle. Ventricle tissue was solubilized in RIPA buffer (150 mM NaCl, 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, Tris, pH 7.5; Pierce, Rockford, IL) with added protease inhibitors and phenylmethylsulfonyl fluoride. All samples were sonicated 3 times, 10 seconds each at an amplitude of 60% and incubated at 4 °C for 2 hours. Protein concentrations were determined using the BCA method. Tissue and whole cell lysates were fractionated on 4–12% Bis-Tris SDS-PAGE gradient gel (Invitrogen ThermoFisher, Waltham, MA) and transferred to a polyvinylidene difluoride membrane using a transfer apparatus according to the manufacturer's protocols (Invitrogen ThermoFisher, Waltham, MA). After blocking by incubation with 5% nonfat milk in PBST (137 mM NaCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, 2.7 mM KCl, 0.1% (v/v) Tween 20, pH 7.4) for 60 min, the membrane was washed with PBST and incubated with antibodies at 4 °C overnight on a rocker. Immunoblots were performed with SUR2 specific antibodies that were generated as previously described [31] (1:500 dilution). Membranes were washed five times PBST for 5 min each and incubated with a 1:10,000 dilution of horseradish peroxidase-conjugated anti-rabbit antibodies for 1 h at room temperature on a rocker. Blots were washed with PBST five times and developed with the ECL system (Pierce, Rockford, IL) according to the manufacturer's protocols.

Echocardiography

Transthoracic echocardiography was performed on 11–14 week old mice after sedation with 3% isoflurane using a VisualSonics Vevo 770 machine with a 30-MHz transducer [18]. Left ventricular end diastolic (LVEDD), left ventricular end systolic (LVESD), anterior wall thickness in diastole (LVAW;d) and posterior wall thickness in diastole (LVPW;d) were obtained from M-mode tracings. LV fractional shortening was calculated as (LVEDD−LVESD)/LVEDD×100%.

Histopathology

Twelve week old TGSUR2A-55 and WT mice hearts were sectioned and stained with hematoxylin and eosin to assess myofiber architecture and Masson's trichrome to assess interstitial fibrosis.

Langendorff ischemia reperfusion experiments

Mice were anesthetized with inhaled 3% isoflurane and then euthanized by cervical dislocation. Hearts were rapidly excised and placed in chilled heparinized modified Krebs-Henseleit buffer (KHB) (118 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO4, 1.2 mM KH2PO4, 25 mM NaHCO3, 2.5 mM CaCl2, 0.5 mM EDTA, 5 mM glucose, and 5 mM sodium pyruvate). Extra-cardiac tissue was dissected and discarded while the aorta was located. The aorta was then cannulated with the use of a 22 gauge cannula. The cannula was secured in place with a 6.0 silk suture. Hearts were then perfused at a constant pressure of 80 mmHg on a custom made Langendorff apparatus with the modified KHB and equilibrated with 95% O2 and 5% CO2 at 37 °C. The left auricle was then excised and a fluid filled balloon catheter constructed from a commercially available kit (Harvard Apparatus, Holliston, MA) was placed in the left ventricle. The balloon catheter was attached to a APT300 pressure transducer (Harvard Apparatus, Holliston, MA) and baseline left ventricular pressure was set between 5-10 mm Hg. Baseline cardiac function was recorded for 30 minutes followed by 45 minutes of ischemia and then 60 minutes of reperfusion. Hearts were paced at 360 bpm via epicardial pacing leads with a Grass SD9 stimulator. Left ventricular pressure was recorded throughout the experimental protocol and analyzed using LabChart Pro (ADInstruments, Colorado Springs, CO). Hemodynamic variables are presented as % recovery to baseline values. Upon completion of reperfusion, hearts were rapidly removed from the Langendorff apparatus and perfused with 30 mM of KCl solution to arrest the hearts in diastole. Then the hearts were stained with 1% tetrazolium chloride solution for 10 minutes. The hearts were then sectioned into 7–8 slices. The sections were placed in 10% formalin and photographed the following day for the quantification of infarct size.

Isolation of mitochondria

Mitochondria from murine ventricular cardiac tissue were isolated by homogenization and differential centrifugation as previously described [26, 32]. Briefly, hearts were quickly excised after thoracotomy and placed in ice-cold isolation buffer (50 mM sucrose, 200 mM mannitol, 5 mM KH2PO4, 1 mM EGTA, and 5 mM 3-(N-morpholino)propanesulfonic acid). Hearts were manually minced, digested for 10 minutes with 0.1 mg/ml of trypsin and mitochondria were isolated with differential centrifugation. The final mitochondrial pellet was resuspended in ice cold isolation buffer with 0.1% bovine serum albumin. Protein concentrations were determined and 0.5 mg of mitochondria were used to perform each mitochondrial experimental run. Total isolation time was <2 hours and mitochondria were used within 1–2 hours of isolation.

Mitochondrial membrane potential and tolerance to Ca2+

Mitochondrial membrane potential (Δψm) was monitored spectrophotometrically as previously described [26]. Briefly, with the use of rhodamine 123 (5 nM) and excitation λex of 503-nm and emission λem of 527-nm, 0.5 mg of mitochondria were added to a cuvette containing malate (5 mM), pyruvate (5 nM) and oligomycin (5 μM). Δψm was assessed in the presence of high (500 μM) and low (10 μM) ATP concentrations as well as the addition of diazoxide (0.1 mM) where indicated. The ionophore carbonylcyanide-4-trifluoromethoxyphenylhydrazone (FCCP; 1 μM) was used to depolarize the mitochondrial membrane at the end of each experiment. The Δψm is expressed as the difference in measured rhodamine 123 fluorescence between the FCCP induced maximum and a stable baseline after mitochondrial addition represented as rhodamine 123 arbitrary florescense units (afu). The Δψm response to diazoixde is represented as the change in rhodamine 123 florescense after a stable baseline is achieved from the addition of mitochondria. Tolerance to Ca2+ of the Δψm was assessed as previously described [26]. Briefly, in the presence of 200 μM of ATP and EGTA-free buffer, 20-μM increments of Ca2+ were added and Δψm was monitored until total Ca2+ concentration reached 140 μM. The membrane potential uncoupler carbonylcyanide-4-trifluoromethoxyphenylhydrazone (FCCP; 1 μM) was used at the end of each experimental run and Δψm was analyzed relative to the fluorescence after the addition of FCCP. The cumulative amount of Ca2+ needed to depolarize the Δψm by 20% was determined.

Statistical analysis

Data are reported as mean ± standard error. Statistical analysis were performed using the R program 3.4.2 (R Foundation for Statistical Computing) and Microsoft Excel (Microsoft, Redmond, WA) with a P value < 0.05 considered significant. Comparison between 2 groups were made using the two-tailed Student's t-test. Ischemia-reperfusion experiments of Langendorff-perfused hearts were subjected to two-way ANOVA for comparison between groups.

Results

Overexpression of SUR2A-55 in TGSUR2A-55 mice

Hearts from 10-14 week old mice were used for mRNA and protein analysis. Hearts from TGSUR2A-55 mice showed similar expression of full-length SUR2A mRNA but significantly increased SUR2A-55 mRNA compared to littermate WT mice (Fig. 1B). Qualitatively, the protein expression of SUR2A in the transgenic mice was not changed but there was increased protein expression of SUR2A-55 in TGSUR2A-55 mice compared to WT mice (Fig. 1C). While previous reports have detected SUR2A-55 in WT mouse hearts [20] and cardiomyoctyes [21], in our current FVB mouse line the level of SUR2A-55 was below the level of detection of western blotting and qPCR.

Fig. 1

Construction of a cardiomyocyte specific SUR2A-55 overexpressing mouse. (A) Organization of the α-MHC-SUR2A-55 construct. (B) Relative total cardiac SUR2A and SUR2A-55 mRNA expression determined by qPCR and (C) protein expression. N = 3/group for qPCR, *P < 0.001. Full images of western blots are presented in supplementary Fig. 1.

Cardiac structure and function in TGSUR2A-55 mice

Wall thickness and chamber dimensions were not significantly different between WT and TGSUR2A-55 mice during baseline echocardiography (Fig. 2). Heart rates were not significantly different between WT and TGSUR2A-55 mice (470 bpm ± 66 vs. 461 bpm ± 61). Left ventricular mass to body weight ratio was not significantly different between WT and TGSUR2A-55 mice (3.58 ± 0.25 vs. 3.37 ± 0.12, respectively). Baseline ex vivo hemodynamic characteristics showed a reduction in +dp/dt and −dp/dt in TGSUR2A-55 mice compared to WT mice (Fig. 3A) but in vivo assessment of % fractional shortening (35 ± 2 vs. 40 ± 3, respectively) did not show any significant difference (Fig. 2). There were no gross histologic differences seen between WT and TGSUR2A-55 mice with hematoxylin and eosin staining (Fig. 2). Trichrome staining did not show any gross differences in cardiac fibrosis.

Fig. 2

Echocardiography and histology for WT and TGSUR2A-55 mice. There were no significant difference in echocardiographic chamber dimensions and % fractional shortening between WT and TGSUR2A-55 mice (A, B). The heart to body weight ratio was also not significantly different between WT and TGSUR2A-55 mice (C). There were no gross changes seen in histology between WT and TGSUR2A-55 mice (D). N = 12 for WT mice and N = 13 for TGSUR2A-55 mice in A,B and C. LVAW;d, left ventricular anterior wall in diastole; LVPW;d, left ventricular posterior wall in diastole; LVEDD, left ventricular end diastolic dimension; LVESD, left ventricular end systolic dimension; H&E, Hematoxylin Eosin; MT, Mason trichrome. (D) Bar = 200 μm.

Fig. 3

The effect of global ischemia on cardiac hemodynamic function in WT vs. TGSUR2A-55 mice. After IR injury, TGSUR2A-55 mice have improved functional recovery compared to WT mice in respect to the % recovery of baseline LV developed pressure (A), +dp/dt (B) and –dp/dt (C). N = 7/group. *P < 0.05 for two-way ANOVA.

TGSUR2A-55 mice are protected against ischemia-reperfusion injury

Isolated hearts from TGSUR2A-55 and WT mice were subjected to 45 minutes of no flow ischemia and then 60 minutes of reperfusion. Baseline hemodynamic characteristics are shown in Table 2 and subsequent hemodynamic variables are presented as % recovery to baseline values (Fig. A–C). TGSUR2A-55 mice had improved left ventricular developed pressure (LVDP) compared to WT mice during reperfusion (P < 0.0001 by two-way ANOVA, Fig. 3A). TGSUR2A-55 mice also had improved +dp/dt and −dp/dt compared to WT mice during reperfusion (P < 0.0001 by two-way ANOVA, Fig. 3B and C). Representative heart sections stained with 1% TTC from TGSUR2A-55 and WT mice after IR injury (Fig. 4A) showed significantly reduced myocardial infarction in TGSUR2A-55 compared to WT mice (Fig. 4B).

Table 2

Baseline hemodynamic characteristics of isolated perfused hearts.

	WT	TGSUR2A-55	P value
LVDP (mmHg)	89.2 ± 15.1	77 ± 10.9	0.067
EDP (mmHg)	8.4 ± 3.7	8.9 ± 4.3	0.41
+dp/dt (mmHg/s)	2958 ± 378	2471 ± 488	0.03
−dp/dt (mmHg/s)	−1836 ± 234	−1435 ± 232	0.003

Fig. 4

The effect of global ischemia on infarct size in WT vs. TGSUR2A-55 mice. (A) Representative sections after IR injury from WT and TGSUR2A-55 mice. (B) TGSUR2A-55 mice have significantly less infarct size compared to WT littermate controls after IR injury. Circles represent individual data points. Squares represent mean values. N = 7/group. *P < 0.01.

The Δψm of TGSUR2A-55 mice are less sensitive to ATP and diazoxide

The mitochondrial membrane potential (Δψm) was measured at low (10 μM) and high (500 μM) ATP concentrations in isolated mitochondria from TGSUR2A-55 and WT mice with the substrates malate and pyruvate. Representative traces show that the sensitivity of Δψm to ATP was diminished in TGSUR2A-55 mitochondria compared to WT mitochondria (Fig. 5A and B). There was a significant 14% difference in Δψm for WT mitochondria measured at low ATP vs. high ATP concentrations (90.1 ± 6.7 afu vs. 104.7 ± 1.6 afu, P < 0.05), while there was only a 4% difference in Δψm for TGSUR2A-55 mitochondria measured at low vs. high ATP concentrations (97 ± 8.9 afu vs. 101 ± 8 afu, P = 0.22). Previous studies of SUR2−/− mice mitochondria found a significantly more polarized Δψm at high ATP concentrations compared to WT mice mitochondria [26]. However our current study did not find such a difference. In the presence of ATP, diazoxide decreased the Δψm in WT mitochondria more than the Δψm of TGSUR2A-55 mitochondria (Fig. 6A). Cumulative experiments showed that WT mitochondria had a significantly greater response to diazoxide compared to TGSUR2A-55 mice (18% ± 0.02 vs. 8% ± 0.02, P < 0.05, Fig. 6B).

Fig. 5

ATP sensitivity of mitochondrial membrane potential (ψm) from WT and TGSUR2A-55 mice mitochondria. (A) and (B) show measurements of ψm using rhodamine 123 fluorescence in isolated mitochondria from WT and TGSUR2A-55 mouse hearts at 10 μM [ATP] and 500 μM [ATP]. (C) Summary data of Δψm (FCCP induced maxium – resting mitochondrial fluorescence) using high and low concentrations of ATP in WT and TGSUR2A-55 mice. N = 6/group for experiments performed with 10 μM [ATP]; N = 4/group for experiments performed with 500 μM [ATP]. *P < 0.05.

Fig. 6

Sensitivity of ψm to diazoxide in WT and TGSUR2A-55 mice mitochondria. Representative measurements (A) of ψm using rhodamine 123 fluorescence showing a reduction in TGSUR2A-55 mitochondria's response to additions of diazoxide (indicted by broken arrows) relative to WT mitochondria. (B) Summary data showing a greater change in ψm in WT mitochondria compared to TGSUR2A-55 mitochondria to diazoxide. N = 4/group; *P < 0.05.

The Δψm of TGSUR2A-55 mitochondria are more tolerant to Ca2+

To test Δψm resistance to Ca2+, TGSUR2A-55 and WT mitochondria were challenged with 20 μM pulses of Ca2+ while Δψm was measured. Representative plots (Fig. 7A) show that WT mitochondria start to depolarize sooner and less Ca2+ is needed to depolarize mitochondria from WT compared to TGSUR2A-55 hearts. (P < 0.05, Fig. 7B).

Fig. 7

Mitochondrial tolerance to Ca2+ overload. After addition of mitochondria and stabilization of florescence, 20 μM additions of Ca2+ (indicted by broken arrows) were given to the mitochondrial preparation as shown in A. The cumulative amount of calcium needed to depolarize the ψm at least 20% was quatified and shown in B among WT and TGSUR2A-55 heart mitochondria. N = 4/group; *P < 0.05.

Discussion

In this investigation we studied whether the cardiomyocyte specific overexpression of the SUR2A-55 splice variant affects the response to myocardial IR injury and alters mitoKATP pharmacology. We found that at baseline there was no significant difference in echocardiographic parameters and gross histology between TGSUR2A-55 and littermate WT control mice. TGSUR2A-55 mice had improved infarct size and hemodynamic response when subjected to IR injury compared to WT mice. We further assessed isolated mitochondria from TGSUR2A-55 mice to see if they have a bioenergentic profile consistent with a protected phenotype [26, 33]. The Δψm in TGSUR2A-55 mitochondria were less responsive to ATP and diazoxide and exhibited an increased tolerance to Ca2+ loading compared to the Δψm in mitochondria from WT mice. KATP channels composed of SUR2A-55 subunits have reduced sensitivity to ATP and diazoxide [22], which is predicted since alternative mRNA splicing removes putative ATP binding sites. These results show that overexpression of SUR2A-55 leads to preserved resting in vivo cardiac structure and function while promoting cardioprotection from myocardial stress associated with a favorable mitochondrial phenotype.

SUR2A-55 overexpression leads to a viable phenotype

While we were unable to identity significant amounts of SUR2A-55 in resting WT hearts, prior studies have shown the presence of SUR2A-55 in mouse heart and localization to the mitochondria [20, 27]. In our study we found that baseline echocardiography-derived cardiac chamber size and function as well as heart to body weight ratio of TGSUR2A-55 mice compared to WT mice were not significantly different. We did, however, notice that in the Langendorff experiments, baseline +dp/dt and −dp/dt values were significantly less in the TGSUR2A-55 compared to WT mice. The mechanism of the isolated heart baseline differences are unclear. It may be that TGSUR2A-55 mice are more susceptible to cardiac depression when ex vivo or that the mitochondrial changes observed translate to subtle changes in cardiac performance when measured ex vivo compared to in vivo. In addition, TGSUR2A-55 may be more sensitive to cardiac depression with the dose of anesthesia used for cervical dislocation. While the current overexpression model of SUR2A-55 leads to a viable mouse phenotype, recent data has suggested that deletion of all SUR2 products including the long-form and SUR2A-55 splice variant lead to lethality with a failure to transition from neonatal to mature adult metabolism [27]. Therefore, it is likely that the SUR2A-55 splice variant carries not only an important role in cardioprotection in adulthood but also in development during the neonatal period. Transgenic mice that conserve the SUR2A-55 variant but do not have the long-form SUR2 due to deletion of exons 14–18 are viable and live to adulthood [34]. While studying substrate utilization and cardiac metabolism were outside the scope of this investigation, future studies testing how SUR2A-55 affects myocardial metabolism may lead to greater understanding of how mitoKATP affects both mitochondrial and cytoplasmic energy production.

Cardiomyocytes with increased expression of SUR2A-55 leads to cardioprotection

The location and significance of the SUR2-55 splice variant was originally presented by Ye et. al. [20] who described a SUR2-55 short-form based ATP-sensitive potassium channel that was insensitive to diazoxide and targeted to the mitochondria. SUR2−/− mice, which removed full length SUR2 but not SUR2A-55, displayed episodic vasospasm, sudden cardiac death, and enhanced insulin dependent glucose uptake [20, 23, 24]. The splice variant was formed through a nonconventional intraexonic splicing event within the 5th and 30th exons of the SUR2 mRNA. When subjected to IR injury, this SUR2−/− mouse was found to have reduced infarction compared to controls [20, 34]. In this current investigation we found that mice overexpressing the SUR2A-55 splice variant have improved tolerance to IR injury compared to controls. Unlike SUR2−/− mice, the TGSUR2A-55 mice express the long form of SUR2 which is also found in mitochondria [20] and thus cardioprotection can occur with the expression of both splice variants. In addition, periodic vasospasm which is reported in the SUR2−/− mouse, is not expected in the TGSUR2A-55 mouse and therefore does not explain the protected phenotype. We speculate from these studies that a switch from mitoKATP formed predominantly with the long-form SUR2 splice variant to the short-form splice variant channel can cause a cardioprotective phenotype. Whether or not a switch to SUR2A-55 dominated mitoKATP underlies cardioprotection in ischemic preconditioning requires further study. Therapies aimed at triggering a switch in SUR2 splice variant composition of mitoKATP may be an option for cardioprotective drugs in the future.

Overexpression of SUR2A-55 changes Δψm response to pharmacology and has implication for the regulation of mitoKATP

Mitochondria play a paramount role in cell death. During myocardial ischemia reperfusion injury, Ca2+ entry into the mitochondria ultimately leads to the mitochondrial transition pore (MPTP) opening and cell death [35, 36]. In our investigation, we found that mitochondria from TGSUR2A-55 are more resistant to Ca2+ loading. In addition, we observed that the Δψm from TGSUR2A-55 mitochondria are less sensitive to ATP concentration and have less responsiveness to diazoxide. Interestingly, the percent difference in Δψm at low and high ATP concentrations in this report is similar to prior studies on the SUR2−/− mice [26] in which the SUR2A-55 isoform was preserved. While the electron transport chain and a number of channels in the mitochondria may be the reason for changes in Δψm [37], the change in sensitivity to ATP and diazoxide, as well as prior studies on SUR2A-55 make mitoKATP the most likely target responsible for our observed data. However, direct measurements of mitoKATP currents were not assessed. These data support prior work that SUR2A-55 promotes a mitoKATP channel that is insentive to diaxoxide and ATP. The lack of ATP or diazoxide to change Δψm suggest a persistent potassium leak and constituently active mitoKATP channel [22, 26].The mitochondrial membrane depolarization that ensues even in times of increased ATP, such as reperfusion, can result in lower mitochondrial calcium entry and reduced MPTP opening. Prior reports have suggested this as a mechanisms for protection from mitoKATP activation and our data would continue to support this theory [38, 39, 40].

Further study in this area is challenging given the elusive nature of the mitoKATP channel. However, recent data suggested that ROMK2 may be the pore forming unit of mitoKATP [15]. Molecular reconstitution studies and gene deletion studies show that ROMK interacts with an ABC binding cassette to manifest native channel properties [41]. Additional studies have shown that ROMK2 binds to an another SUR2 splice variant, SUR2B, to form sulfonylurea sensitive K+ channels that were not observed with ROMK1 or ROMK3. It is unknown if other splice variants of SUR can create native functional channels with ROMK. Further investigations studying interactions of ROMK2 and SUR2A splice variants may shed light on the molecular identity of mitoKATP. Our research suggests considerable plasticity of the mitoKATP channel with different subunits affecting the channels response to pharmacology. This may also explain the elusive nature of the mitoKATP channel and conjecture surrounding its existence.

While our current investigations describe the response of SUR2A-55 overexpression in mice, certain limitations exist. For practical reasons of scope, we restricted study to overexpression of the A isoform of SUR2. Prior studies have shown that both SUR2A-55 and SUR2B-55 isoforms exist [20, 21]. Therefore the SUR2B isoforms may also have physiologic significance and play a role in mitochondrial function. We concentrated on the SUR2A-55 isoform as our prior work showed qualitiatvely increased expression of SUR2A-55 over SUR2B-55 in mouse hearts [20]. In addition, due to the scope of our research, we restricted our study to male mice. While previous studies have shown targeting of the SUR2A-55 protein to mitochondria, we did not assess its location in TGSUR2A-55 and did not measure potassium flux in mitochondria. However, our mitochondrial biophysical studies suggest a direct effect for SUR2A-55 on mitochondria.We also used a global ischemia ex vivo model for IR injury which has different physiologic stress than a coronary ligation model and may not be as translational to human disease. In addition, while we report cardioprotection to IR injury and changes in mitonchondrial biophysical properties in TGSUR2A-55 mice, there may be other cellular mechanisms including changes in sarcKATP activity that could be responsible for the cardioprotective response. Further studies on the role of SUR2A-55 in sarcKATP activity are needed.

Conclusion

In conclusion, elucidating pathways in cardioprotection may lead to improved clinical treatment of heart disease. While great advances have been made in the treatment of ischemic heart disease, it remains the leading cause for the development of heart failure [42, 43] and imposes a large burden on society [44]. In this research investigation, we showed that SUR2 splice variants offer an interesting area of further research to clarify mitochondrial function in ischemic heart disease. By altering the expression of SUR2 splice variants, different cardiac phenotypes arise. The overexpression of the SUR2A-55 splice variant leads to cardioprotection likely through a constitutively active mitoKATP channel. Further research to identify the components of the mitoKATP channel that enhance the cardioprotective phenotype should be considered using the TGSUR2A-55 mouse.

Declarations

Author contribution statement

Mohun Ramratnam: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Barrett Kenny: Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data.

John W. Kyle: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data.

Brandi Wiedmeyer: Performed the experiments; Analyzed and interpreted the data; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Timothy A. Hacker: Contributed reagents, materials, analysis tools or data.

David Y. Barefield: Analyzed and interpreted the data.

Elizabeth M. McNally, Jonathan C. Makielski: Conceived and designed the experiments; Wrote the paper.

Funding statement

This work was supported by National Institutes of Health grant R01HL122109 (to EMM and JCM).

Competing interest statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.