Animals & Tissue Collection
Twelve and 20-week-old male type 2 diabetic Zucker Diabetic Fatty (DM) (ZDF fa/fa, N = 10 per age group) rats and their lean non-diabetic (nDM) littermates (ZDF +/+, N = 10 per age group) were used in this study. All experiments were approved and conducted within the guidelines of the Animal Ethics Committee of the University of Otago, New Zealand, and adhered to the NZ Animal Welfare Act of 1991, which complies with the US National Institutes of Health Guide for the Care and use of Laboratory Animals. Rats were euthanized with a lethal dose (60 mg/kg) of pentobarbital (Provit, Canada). Upon loss of orbital and pedal reflexes, the chest cavity was opened and the heart was quickly excised and placed into a bath of cold relaxing buffer (100 mM potassium chloride, 1.75 mM ethylene glycol tetraacetic acid (EGTA), 10 mM Imidazole, 4 mM adenosine triphosphate magnesium, 5 mM magnesium chloride, pH 7.0 with potassium hydroxide). Heart and left ventricle (LV) weight were recorded and each chamber was dissected, quickly frozen in liquid nitrogen and stored at -80°C.
Echocardiography
Echocardiography was carried out at 12 and 20 weeks of age on nDM and DM rats using an echocardiography ultrasound machine (Vivid Q, General Electric Healthcare) to determine the effects of the progression of diabetes on systolic and diastolic left ventricular (LV) morphology and function. Rats were initially anaesthetized with 4% isoflurane in a gas chamber, and then transferred onto a heating pad. Anaesthesia was maintained at 2–3% isoflurane by nose cone during echocardiography. All exams were performed with the animal on its back and a shaven chest. Two-dimensional images were obtained from a parasternal long axis view. LV structure and systolic function was determined by measuring in M-mode. LV diastolic function was measured by pulse-wave Doppler images. To minimize the effects of isoflurane on the heart, echocardiography was performed two to three days prior to euthanasia to ensure washout.
Isolation of Rat Cardiomyocytes
Previously frozen LV samples were partially homogenized in relaxing buffer with HALT Protease and Phosphatase Inhibitor Cocktail 100X (Thermo Scientific, 78430). The myocyte suspension was centrifuged at 10000 rpm, re-suspended and permeabilized with 1.1% Triton X-100 detergent (Thermo Scientific, 28314) for 8 minutes. The resulting pellet of skinned myocytes was washed with relaxing buffer 3 times. Myocytes were stored on ice and used for up to two days, as previous studies found cells were equally viable for up to 48 hours (26, 35).
Cardiomyocyte Force Measurements
A single cardiomyocyte or small bundle of cells (approximately two to five) were mounted on an inverted phase contrast microscope (Nikon) and fixed between a piezoelectric motor and a force transducer (Aurora Scientific, 405 model) with the use of silicone glue (Marineland Aquarium Sealant) in a relaxing buffer. Stage temperature was set to 15°C and glue was left to cure for 30 minutes. Once cured, the cell could be moved into the well containing pCa 9.0 (lowest concentration of calcium) and stretched until the sarcomere length reached 2.3 um. Sarcomere length, cell length and cell height were visualized and measured using an IDS camera and VSL 900B software (Aurora Scientific).
The pCa50, or the [calcium (Ca2+)] at which half-maximal tension was produced, is a measure of the calcium sensitivity of cardiomyocytes. Myofilament tension was developed by placing each cell into baths containing different concentrations of calcium (-log[Ca2+] or pCa). Force was measured by slacking the myocyte by 20% of its original length (slack test). The force at a given pCa was measured as the change in force from peak length to 80% of initial length (0.8 Lo). Initially, force was measured in pCa 4.5 and then randomly selected submaximal pCa solutions, with every fourth or fifth activation made in 4.5 to assess any decline in myocyte performance. Submaximal pCa solutions were created by mixing different ratios of pCa 9.0 (EGTA 7mM, Imidazole 20 mM, magnesium chloride 5.42 mM, potassium chloride 79.16 mM, calcium chloride 16.33 uM, creatine phosphate 14.5 mM, magnesium adenosine triphosphate 4.74 mM) and 4.5 (EGTA 7mM, Imidazole 20 mM, magnesium chloride 5.26 mM, potassium chloride 64 mM, calcium chloride 7 mM, creatine phosphate 14.5 mM, magnesium adenosine triphosphate 4.81 mM), determined from the computer program Fabiato (36). The experimenter was blinded regarding whether a given cell was from a nDM or DM LV sample. Myocytes were excluded from data analysis if 80% of maximal force was not maintained by the end of the experimental protocol.
Maximal force, or the force at pCa 4.5 subtracted by the passive force measured at pCa 9.0, was normalized to the cross-sectional area (CSA) of the cell (CSA = 3.14*(cell width/2)2). Active tension was calculated by subtracting passive tension at pCa 9.0 from the total tension measured by the slack test. Tension at each pCa was expressed as a fraction of the maximum tension (pCa 4.5) obtained for that cell under the same conditions. As described by Hofmann et al., data were analysed by least-squares regression using the Hill equation,
Log [(Prel)/(1 – Prel)] = n(log [Ca2+] + k)
where Prel is tension expressed as a fraction of maximal tension, n is the Hill coefficient, and k is the intercept of the fitted line with the x-axis, which corresponds to the [Ca2+] at half-maximal (50%) tension (pCa50) (37). With the use of constants derived from the Hill equation, tension data were fit using Prism software with the following Eq. (35).
Prel = [Ca2+]n/(kn + [Ca2+]n)
ProQ Diamond Phospho-Fluorescent Gel Staining
ProQ Diamond Phosphoprotein gel stain (Invitrogen, P33300) was used to determine myofilament protein phosphorylation from nDM and DM LV samples. Protein lysates were prepared by homogenizing the same LV samples used previously in RIPA buffer (sodium chloride 50 mM pH 7.4 with hydrogen chloride, sodium dodecyl sulfate (SDS) 1%, Triton X-100 1%, edetate disodium 1 mM). Protein samples (20 ug) were prepared for gel electrophoresis by adding 1X sample buffer, water and protein lysate of interest. Protein was denatured by heating at 95°C for 5 minutes and loaded onto a 4–20% gradient gel (BioRad, #4561096). A PeppermintStick Phosphoprotein Standard (Invitrogen, P27167) was used to identify myofilament proteins of interest and also as positive and negative controls for the phospho-staining technique. After gel electrophoresis was complete, gels were fixed overnight in 50% methanol and 10% acetic acid to remove remaining SDS. Gels were washed 3 times for a total of 30 minutes with ultrapure water to remove fix solution and subsequently stained with ProQ Diamond gel stain. Gels were protected from light, placed on shaker at 50 rpm and stained for 75 minutes. Next, gels were destained using ProQ Diamond Phosphoprotein Gel Destaining Solution (Invitrogen, P33310) for a total of 90 minutes. Each gel was then washed with ultrapure water before imaging on the VersaDoc Imager using a green light filter at 605 nm and exposed for 20 seconds.
Gels were then immersed in Sypro Ruby Protein Gel Stain (Invitrogen, S12000) overnight in the dark at 50 rpm to allow for the measurement of total protein and ensure equal loading across lanes. Before imaging, gels were washed in a Sypro Wash solution (10% methanol, 7% acetic acid) for 30 minutes and washed with ultrapure water. Sypro Ruby stained gels were imaged with the VersaDoc Imager using the UV Transilluminator filter at 605 nm and exposed for 1 second.
Gel band intensities were quantified using ImageStudio software (LiCor Tech). The amount of phosphorylation was normalized to total protein using a ratio of the signal intensity of the phospho-stained band divided by the signal intensity of the total protein at each respective protein of interest: cMyBP-C, cTnT, cTnI and myosin light chain II (MLCII). Due to high quantities of protein bands present near cTnT and cTnI bands, native cTnT and cTnI protein were run in neighboring lanes to confirm band identifies (Abcam, ab9937 & ab9936 respectively).
Western Blot Analysis of Phospho-Specific Proteins and O-GlcNAc
To measure the phosphorylation occurring at specific sites on cTnI, LV tissue was homogenized in RIPA buffer and protein lysates were prepared similarly to above. Proteins were separated on a 15% SDS PAGE gel and a molecular weight marker was used for protein identification (BioRad, 1610374). Each gel was run for approximately one hour at 120 V using a BioRad PowerPac blot system. Gels were then transferred onto a nitrocellulose membrane, 100 V for 90 minutes. After protein transfer, membranes were incubated in 5% milk powder TBST solution (Tris 50 mM, sodium chloride 150 mM, Tween-20 0.5%). Primary antibodies for total O-GlcNAc, cTnI and β-actin (loading control) were used (Abcam ab2739, ab4002, ab49900 respectively). All primary antibody incubations were performed overnight at 4°C. For secondary anti-rabbit and anti-mouse antibodies, incubation was for 1.5 hours at room temperature (Abcam ab97051 and ab97023 respectively).
After obtaining the total cTnI protein expression signal, the same membrane underwent a stripping protocol. To remove the cTnI antibody, a stripping buffer (β-mercaptoethanol 100mM, SDS 2%, Tris 62.8 mM pH 6.8) was used at 50°C for 30 minutes. Imaging was completed to confirm stripping of the membrane. Lastly, site-specific phosphorylation expression could be determined using one of the following primary antibodies: cTnI Ser 23/24, Thr 144 and Ser 150 (Cell Signaling 4004, Abcam ab58546 and ab169867 respectively). After an overnight incubation, membranes were washed with TBST and incubated with an anti-rabbit secondary antibody. To detect differences in cTnI phospho-specific levels we determined, for each sample, the signal intensity of phospho-cTnI site of interest normalized to total cTnI. To ensure equal loading, protein quantity was controlled by probing for and measuring β-actin signal intensity (Abcam, ab49900). To determine the amount of O-GlcNAc modified cTnI, total O-GlcNAc at the cTnI band was normalized to total cTnI expressed. All blots were imaged using the Syngene Imager after incubation with West Pico solution (ThermoFisher, 34580) to detect protein bands. Protein bands were analyzed using ImageJ Software or Image Studio software for total O-GlcNAc gel lanes.
Statistical Analysis
The four groups (12 nDM, 12 DM, 20 nDM and 20 DM) were compared using a two-way ANOVA and a Levene’s test of normality was used to check for a normal distribution. An exception was made for site-specific western blotting, as experiments were completed at different time points (months apart) and made a two-way ANOVA comparison unfeasible. In this case, independent t-tests were run to compare 12-week-old nDM and DM rats and also 20-week-old nDM and DM rats and corrected by a Bonferroni’s correction. Both SPSS (IBM) and Prism-8 (GraphPad) software were utilized to analyze and create figures, respectively. Statistical significance was set at p ≤ 0.05.