Background
In addition to skeletal muscle, liver and adipose tissue, the heart with 10,000–100,000 expressed insulin receptors (IRs) per cardiomyocyte [
1] must be considered as an additional major organ affected by insulin (Ins). In cardiomyocytes, Ins modulates glucose transport, metabolism, protein synthesis, hypertrophy, contractility, beating-rate, and apoptosis [
1‐
3].
Long-acting insulin analogues are designed to deliver a constant basal Ins supply throughout the day via the subcutaneous route resulting in improved fasting blood glucose and overall glycaemic control while reducing the risk of hypoglycaemia [
4‐
6]. Insulin glargine (IGla) has one amino acid exchange in the A-chain and two additional arginine residues at the B-chain, resulting in an altered isoelectric point that leads to local precipitation after subcutaneous injection [
7,
8]. From this depot, IGla is slowly dissolved followed by immediate biotransformation into the active metabolite M1 (IGlaM1) [
9,
10]. For insulin degludec (IDeg) it has been proposed that its protracted action results from slow dissolution of subcutaneous multi-hexamer assemblies [
11]. However the structure of the side chain attached to this molecule (n-16 fatty acid) [
11] suggests that the protraction mode may be similar to that of insulin detemir, which binds to human serum albumin (HSA) via its fatty acid side chain, thereby protracting its duration of action by providing a ‘floating depot’ with the consequence of a reduced biological availability [
12‐
14].
Besides the benefits of Ins analogue modification, modifying the Ins molecule may lead to an altered activation profile such as receptor signalling or pharmacodynamics and pharmacokinetics. Recently the effects of Ins analogues on the cardiovascular system gained considerable interest. So far, finalised clinical data is only available for IGla, whereas for IDeg the investigation is still ongoing. Results from the Outcome Reduction With Initial Glargine Intervention (ORIGIN) trial proved non-inferiority of IGla compared to standard care treatment in regard to cardiovascular effects (Trial number NCT00069784) [
15]. Just recently the interim analysis of the dedicated cardiovascular outcome trial for IDeg (DEVOTE) (Trial number NCT01959529), requested in 2013 by the American Food and Drug Administration (FDA) [
16] suggested non-inferiority to IGla, leading to approval of IDeg for the US market with a post marketing commitment to provide further non-inferiority data in major adverse cardiovascular events [
17]. These clinical trials were designed to compare systemic metabolic effects of the analogues in patients. However, strong evidence for tissue-specific action of Ins exists [
18] and to the best of our knowledge, there are no in vitro studies published which elucidate and compare the effect of different long-acting Ins analogues on cardiomyocyte cell models. Thus, data regarding the signalling and function of Ins analogues in cardiomyocyte cell models might shed light on potential differences of these drugs in relation to their cardiac action. Therefore, we compared the impact of IGla, IGlaM1, and IDeg to Ins in regard to signalling, contractility and anti-apoptotic potency in HL–1 cardiomyocytes, ARVM, H9c2-E2, and Cor.4U
® cells. Our data show a very similar cardiomyocyte action profile for both IGla and IDeg, at least under steady-state conditions.
Methods
Cell culture
The cardiac mouse cell line HL-1, a cell line derived from the AT-1 mouse atrial cardiomyocyte tumour lineage [
19], was kindly provided by Dr. W.C. Claycomb (Louisiana State University, New Orleans, LA, USA). HL-1 cells were cultivated in Claycomb medium containing 10 % fetal calf serum (FCS), 100 µM norepinephrine and 4 mM
l-glutamine (all from Sigma-Aldrich, Munich, Germany) on gelatine/fibronectin coated plates. H9c2 cells (ATCC CRL-1446), stably transfected with the human IR in our laboratory [
20] (H9c2-E2) were cultivated in DMEM, low glucose containing 10 % FCS, 1 % non–essential amino acids and 600 µg/ml G418 (all from Invitrogen, Carlsbad, CA, USA). Commercially available iPS-derived human cardiomyocytes (Cor.4U
®) (Axiogenesis, Cologne, Germany) were cultured in Cor.4U
® Complete Medium containing 10 % FCS on fibronectin–coated 96-well E-Plate (Acea Biosciences, San Diego, CA, USA). The medium was changed twice daily. All cells were incubated at 37 °C with 5 % CO
2 in a humidified incubator.
Competition binding experiments on membrane embedded and solubilised insulin receptor preparations
Isolation of insulin receptor embedded plasma membranes (M-IR) and competition binding experiments were performed as previously described [
21]. Briefly, CHO-cells overexpressing the IR were collected and re-suspended in ice-cold 2.25 STM buffer (2.25 M sucrose, 5 mM Tris–HCl pH 7.4, 5 mM MgCl
2, complete protease inhibitor) and disrupted using a Dounce homogenizer followed by sonication. The homogenate was overlaid with 0.8 STM buffer (0.8 M sucrose, 5 mM Tris–HCl pH 7.4, 5 mM MgCl
2, complete protease inhibitor) and ultra-centrifuged for 90 min at 100,000
g. Plasma membranes at the interface were collected and washed twice with phosphate buffered saline (PBS). The final pellet was re-suspended in dilution buffer (50 mM Tric-HCl pH 7.4, 5 mM MgCl
2, complete protease inhibitor) and again homogenised with a Dounce homogenizer. Competition binding experiments were performed in a binding buffer (50 mM Tris–HCl, 150 mM NaCl, 0.1 % BSA, complete protease inhibitor, adjusted to pH 7.8) in 96-well microplates. In each well 2 µg isolated membrane were incubated with 0.25 mg wheat germ agglutinin polyvinyltoluene polyethylenimine scintillation proximity assay (SPA) beads. Constant concentrations of [
125I]-labelled human Ins (100 pM) and various concentrations of respective unlabelled Ins (0.001–1000 nM) were added for 12 h at room temperature (23 °C). The radioactivity was measured at equilibrium in a microplate scintillation counter (Wallac Microbeta, Freiburg, Germany).
Binding on a freshly solubilised IR preparation (S-IR) was performed as previously described [
22] with some modifications. Aliquots of membranes were incubated at 4 °C for 30 min in a solubilisation buffer (20 mM HEPES–NaOH, 100 mM NaCl, 10 mM MgSO4, 1 % (w/v) n-Dodecyl-ß-
d-maltoside (Sigma-Aldrich, Munich, Germany), adjusted to pH 7.8 and Complete TM Protease Inhibitor cocktail). Thereafter, ultra-centrifugation was performed at 100,000
g for 30 min and 4 °C to remove non–solubilised debris. Protein concentration in the supernatant was adjusted to 0.15 mg/ml with binding buffer (100 mM HEPES–NaOH, 100 mM NaCl, 10 mM MgSO
4, 0.025 % (v/v) Tween-20, adjusted to pH 7.8 and complete TM protease inhibitor cocktail). To streptavidin SPA beads (5 mg in 1000 ml binding buffer), 50 µl of an anti-IR alpha-antibody 83-7 (Abcam, Cambridge, UK) was added. After incubation for 30 min, SPA beads were once washed and finally re-suspended in 500 µl binding buffer. A solution of solubilised receptor (1 ml, 0.15 mg/ml) was added and incubated for further 60 min, before washing and resuspension in 1.5 ml. Subsequently, 100 µl re-suspended IR-Antibody-SPA beads (containing 10 µg total protein) were mixed with 50 µl [
125I]-labelled insulin tracer (100 pM) and 50 µl non-radioactive Ins (0.001 – 1000 nM), incubated for 12 h at room temperature (23 °C) under shaking, centrifuged for 2 min and measured in the scintillation counter (Wallac Microbeta, Freiburg, Germany).
Effect of insulin and insulin analogues on contractility of primary adult rat ventricular cardiomyocytes
Adult rat ventricular cardiomyocytes (ARVM) were isolated from wild-type Lewis rats (Lew/Crl) as previously described [
23]. ARVM were cultivated 3 h in Medium 199 with Hanks’ balanced salts containing 5 mM creatin, 2 mM carnitine and 5 mM taurine supplemented with 10 % FCS and 1 % insulin/transferrin/selene on laminin–coated dishes (ibidi GmbH, Martinsried, Germany). Subsequently, ARVM were cultivated over-night in DMEM/F12 containing 33 µM biotin and 17 µM pantothenate (Invitrogen, Carlsbad, CA, USA). Prior to measurement, ARVM were pre-incubated for 5 min with 100 nM of Ins (porcine Ins, Cat. No.: I5523, Sigma-Aldrich, Munich, Germany), IGla, IGlaM1 or IDeg (provided by Sanofi-Aventis, Frankfurt a.M., Germany) in modified Tyrodes solution: 125 mM NaCl; 1.2 mM KH
2PO
4; 2.6 mM KCl; 1.2 mM MgSO
4*7H
2O; 1 mM CaCl
2*2H
2O; 10 mM Glucose; 10 mM HEPES; adjusted to pH = 7.4 prior to measurement. Furthermore, untreated ARVM or ARVM treated with 10 nM isoproterenol (Sigma-Aldrich, Munich, Germany) were immediately measured. ARVM were paced with bipolar pulses in a contractility and fluorescence system (IonOptix, Milton, MA, USA) at 15 V, 1 Hz, 0.5 ms, at 37 °C for up to 10 min and 10–14 contractions of at least 10 rod–shaped ARVM per condition were recorded. Sarcomeric shortening, shortening rate and re-lengthening rate were calculated using the IonWizard software (IonOptix, Milton, MA, USA). To determine the role of Akt for the positive inotropic effect of Ins and the analogues, ARVM were pre-treated with 10 µM of the specific Akt–inhibitor triciribine (Sigma-Aldrich, Munich, Germany) for 30 min in contraction buffer. Afterwards, ARVM were treated as described above. After 30 min treatment with 10 µM triciribine, ARVM viability was assessed by incubating the cells with 0.1 % trypan blue in PBS for 5 min. Microscopic pictures were taken randomly with at least 10 pictures per condition. As a positive control 200 µM H
2O
2 was utilised. For each condition at least 400 cells were counted per experiment.
Immunoblotting
ARVM and HL-1 cells were treated as indicated and lysed in buffer containing 50 mM HEPES (pH 7.4) (PromoCell, Heidelberg, Germany), 1 % Triton X-100 (Sigma-Aldrich, Munich, Germany), PhosSTOP and CompleteTM protease inhibitor cocktail (Roche, Basel, Switzerland). After incubation for 2 h at 4 °C, the suspension was centrifuged at 10,000g for 15 min. 5 μg protein sample of the total cell lysate was separated by SDS/PAGE (10 % gel) and transferred to a polyvinylidene fluoride (PVDF) membrane. Membranes were blocked in tris-buffered saline (TBS) containing 0.1 % tween 20 and 5 % (w/v) non-fat dried skimmed milk powder and incubated overnight with anti-phospho Akt(Ser473) antibody, anti-phospho Akt(Thr308), anti-GAPDH antibody (all Cell Signalling Technology, Danvers, MA, USA) or anti-tubulin antibody (Abcam, Cambridge, UK). After washing, membranes were incubated with appropriate horseradish peroxidase-coupled secondary antibody and processed for enhanced chemiluminescence (ECL) detection using Immobilion horse radish peroxidase (HRP) substrate (Millipore, Darmstadt, Germany). Signals were visualised and evaluated on a VersaDoc 4000 MP Bio-Rad Laboratories work station and analysed by Quantity One analysis software (version 4.6.7) (both Bio-Rad Laboratories, Hercules, CA, USA).
Impedance measurement in human Cor.4U® cells
Cor.4U
® cardiomyocytes were seeded at a density of 30,000 cells per well. From day 3 on the cells were cultured in Iscove’s Basal Medium containing 1 % GlutaMAX supplement (both Life Technologies, Carlsbad, CA, USA) and 2 µg/ml Ciprobay (Bayer, Leverkusen, Germany). Treatment with 500 nM of the designated Ins or analogue or 100 nM isoproterenol was started at day 4 after seeding. Impedance of each well was measured with the RTCA Cardio xCELLigence Analyser [
24] (Acea Biosciences, San Diego, CA, USA) during the whole experiment. Beating-rate and cell index were analysed by RTCA cardio software (Acea Biosciences, San Diego, CA, USA).
Glucose uptake in HL-1 cardiomyocytes
HL-1 cells were seeded at a density of 400,000 cells per well in a 12-well plate. Glucose uptake was measured in serum-starved HL-1 cells, either kept untreated or exposed for 60 min to 200 nM Ins and the analogues, respectively. Subsequently 0.12 mM deoxy–d–glucose (Sigma-Aldrich, Munich, Germany) with 0.055 mCi 2–deoxy–D–[14C]glucose (PerkinElmer, Waltham, MA, USA) was added to the cells. After 10 min incubation the uptake was terminated by repeated washing with ice cold PBS. Afterwards, the cardiomyocytes were lysed with lysis buffer containing 1 % SDS and 200 mM NaOH. Incorporated glucose was measured by scintillation counting of the cell lysates in a liquid scintillation counter (Beckman Coulter, Pasadena, CA, USA). Values were corrected for non-specific uptake as measured by incubation with L-[14C]glucose (PerkinElmer, Waltham, MA, USA).
Caspase 3/7 activity assay
To assess the anti-apoptotic effect of Ins and its analogues, H9c2-E2 cells were seeded in a density of 5000 cells per well of a 96-well plate. The next morning cells were treated with 100 nM of the respective Ins either in the presence or absence of 800 µM of H2O2 for 2 h. Caspase 3/7 activity was then measured by the caspase-Glo® 3/7 assay system (Promega, Madison, Wisconsin, USA) as described in the manual. After 2 h incubation period caspase 3/7 activity was analysed by measuring the luminescence in an Infinite 200 plate reader (Tecan, Männedorf, Switzerland).
Statistical analysis
Results are expressed as mean values ± SEM of at least three independent experiments. For statistical analysis Graphpad Prism v5.00 (Graphpad Software, San Diego, CA, USA) was used. One-way ANOVA was performed to determine significance between conditions, with level of significance chosen at p < 0.05. In case of IC50 determinations for binding, a non-parametric Kruskal–Wallis testing was performed, again with level of significance chosen at p < 0.05.
Discussion
Treatment of diabetic patients with insulin analogues has been shown to provide a more efficient, reproducible, and convenient therapy than regular insulin. The analogues may vary from insulin with respect to metabolic potency, stability or onset, and duration of action that is achieved by either sequence or secondary structural modifications. These changes may lead to an altered functional profile, emphasizing the importance of examining all steps in the action of an insulin analog in vitro and in vivo. Regarding the effects of insulin analogues in cardiomyocyte cell models no published data is available. Therefore, we compared the long–acting insulin analogues IGla and IDeg in cardiac cell models. Using this in vitro setting, we could show the absence of any difference in functional cardiac endpoint measurements and insulin signalling between the long-acting insulin analogues IGla and IDeg under steady state conditions.
Since Akt is a major element of the insulin signalling pathway, we first analysed the activation of Akt in ARVM and HL-1 cells after treatment with the different insulin analogues. Although our results show a slower onset of Akt activation in HL-1 cells treated with IDeg for 5 and 10 min compared to the other insulins (Fig.
1b, c), we did not observe a difference in Akt activation in HL-1 cells treated for 60 min and acutely treated ARVM (Fig.
1d and
2a). We speculate that a possible explanation for the slower onset of Akt phosphorylation in HL-1 cells might be the low binding affinity of IDeg towards the IR. In receptor binding studies using S-IR of both isoforms, the binding affinity for IDeg was found to be 13–15 % relative to human insulin [
25]. The results from our indirect binding assays with the S-IR show a similar binding affinity of IDeg with ~18.6 % relative to human insulin. However, it should be noted that S-IR displays a less complex construct compared to M-IR. In these preparations the IR is surrounded by lipids and protein complexes. In M-IR preparations IDeg showed a binding affinity of ~4 % compared to Ins and ~13 % compared to IGlaM1. It could be possible that the fatty acid residue attached to IDeg interacts with other components of the membrane, and therefore leading to a reduced binding affinity towards the IR. The very low binding affinity of IDeg in M-IR (Fig.
1a) could be an explanation for the slower onset of action observed in HL-1 and Cor.4U
® cells. The results obtained in M-IR preparations with Ins and IGlaM1 and IDeg in S-IR are comparable to previously published data [
21,
25].
Insulin-mediated Akt activation in the myocardium triggers a variety of processes, like glucose uptake, modification of calcium signalling and anti-apoptotic effects [
26]. Even though under physiological conditions the main energy source for the heart is fatty acids, about one third is derived from glucose [
27] and with increasing blood glucose and insulin level, glucose becomes the favoured substrate in the heart [
2]. Therefore, we next analysed the ability of IGla and IDeg to stimulate glucose uptake under steady-state conditions in HL-1 cells, since we observed a full Akt activation with IDeg after 60 min. Under these conditions we observed no difference between IGla and IDeg in the insulin stimulated glucose uptake compared to Ins. We therefore conclude that IGla and IDeg are equipotent in regulating cardiac glucose consumption, at least under steady-state conditions.
Another important aspect of insulin function in cardiomyocytes is the positive inotropic effect which was already described in the 1920s by Visscher and Müller [
28] and is due to an increased excitation–contraction coupling, which in turn is controlled by entry and release of Ca
2+ from the sacroplasmatic reticulum (SR). Using insulin and the insulin analogues we observed a similar positive inotropic effect, as shown by comparable increased sarcomeric shortening, departure- and return-velocity in ARVM. Furthermore, we could show that the positive inotropic effect in ARVM is completely Akt dependent. The Akt dependency of cardiomyocyte contraction was shown previously by Graves et al. in HL-1 cells, where inhibition of Akt activation leads to decreased total [Ca
2+]
i, intracellular Ca
2+ transients and membrane I
Ca [
29]. Furthermore, Reinartz et al. [
30] recently showed that Akt1 and Akt2 knockdown affected phosphorylation of proteins involved in regulation of heart contraction as well as relaxation and regulation of heart rate. Additionally, proteins involved in Ca
2+ release and re-entry into the SR are affected (e.g. CaMKII or phospholamban, a direct target of Akt) [
30,
31], which could be a possible explanation for the complete abrogation of the positive inotropic effect after Akt-inhibition. Insulin increases the beating–rate of cardiac muscle, but the underlying mechanism is controversially discussed in the literature. While some groups found evidence for insulin to directly increase the beating-rate in vivo [
32,
33], others claimed that insulin acts on the nervous system and thereby leads to beta-adrenergic stimulation of the heart [
34,
35]. In our in vitro experiments with Cor.4U
® cells we observed a slight but significant increase in beating-rate of Cor.4U
® cardiomyocytes for up to 6 h. As observed in HL–1 signalling, we measured a slower onset of action with IDeg. However under steady state conditions no differences between the different analogues could be detected. Together with the results from the ARVM contraction experiments we conclude that the increase in beating–rate of cardiomyocytes is at least partly independent of nervous system activity and directly affected by insulin itself.
Furthermore, insulin is known to reduce the damage of ischemia/reperfusion injury (IRI) in vivo as well as in vitro [
36‐
39]. This damage is induced by massive production of reactive oxygen species (ROS) [
40]. While diabetes per se is a risk factor for ischemic heart disease, the damage inflicted by IRI is even worse in diabetic patients [
41]. Therefore, we aimed to mimic IRI in H9c2–E2 cardiomyocytes by challenging the cells with H
2O
2 with subsequent measurement of caspase 3/7 activity. To elucidate the potency of IGla and IDeg in prevention of caspase 3/7 activation during ROS treatment, we treated part of the cells with a combination of H
2O
2 and the respective insulin. With our results we were able to reproduce the protective effect of Ins during IRI and furthermore we were able to show that both, IGla and IDeg, have the same potency to prevent caspase 3/7 activation as Ins.
Conclusion
In conclusion, the long-acting insulin analogues IGla and IDeg show no major differences in several cardiomyocyte in vitro models regarding insulin signalling, contractility parameters, beating–rate, glucose uptake, and protection from oxidative stress–induced caspase 3/7 activation under steady-state conditions. However, for IDeg we observed a slower onset of action in Akt phosphorylation in HL-1 cells as well as slower response to IDeg in human Cor.4U® cardiomyocytes. Additionally, we observed very low binding affinities of IDeg in M-IR preparations. Whether these effects translate to the complex in vivo situation needs further evaluation.
Authors’ contributions
TH: data collection and analysis, study design, drafted the manuscript. SO: data collection, critical revision of the manuscript. MO: critical revision of the manuscript. SR: study design, critical revision of the manuscript. PW: data collection and analysis, study design, critical revision of the manuscript. NT: study design, critical revision of the manuscript. NW: data analysis, study design, critical revision of the manuscript. JE: study design, critical revision of the manuscript. All authors read and approved the final manuscript.