Sex differences in ventricular muscle energetics in a type 2 diabetic rat model

Abstract

Background

Sex differences in cardiac function under diabetic conditions have been extensively studied. However, the impacts of type 2 diabetes on cardiac energetics between sexes remain poorly defined. Likewise, whether sex-specific differences in cardiac efficiency reported at the whole heart level manifest at the muscle level is uncertain. This study is the first to assess sex-specific cardiac energetics in type 2 diabetes by directly measuring heat production in isolated rat left ventricular trabeculae.

Methods

Induction of diabetes in Wistar rats by a high-fat diet (23.5% kcal vs. 5% kcal from fat) and low-dose streptozotocin (30 mg/kg) was assessed through measurements of fasting blood glucose and glucose tolerance, and plasma biomarkers. Mechanoenergetics of isolated trabeculae were characterised using our work-loop calorimeter at body temperature. Experiments were conducted under loading conditions of varying muscle lengths, contraction modes, and afterloads. Force, muscle length, and heat output were simultaneously recorded. Metrics including twitch kinetics, shortening, mechanical work, activation heat, and cross-bridge heat were extracted, and mechanical efficiency was estimated and evaluated over a wide range of loading conditions.

Results

At the organism level, diabetic rats exhibited increased blood glucose levels and impaired glucose handling. While diabetes caused sex-specific effects on body mass, corticosterone, plasma insulin, and biomarkers, our results at the muscle level showed no effects of diabetes on any measured indices of cardiac mechanoenergetics and, hence, mechanical efficiency, in either sex.

Conclusions

Effects of type 2 diabetes do not manifest in the mechanoenergetic functional performance of isolated cardiac muscles, even when challenged to a wide range of loading conditions. This conclusion is upheld in both male and female rats. These findings underscore the need for assessing cardiac function beyond the muscle level, as muscle-specific thermodynamics in diabetes may not be captured from systemic whole-organism measurements.

Graphical abstract

Background

Sex differences in cardiovascular complications are increasingly recognised in type 2 diabetes. Diabetic women often experience more severe outcomes than diabetic men [1–4]. This disparity is especially concerning given that type 2 diabetes is one of the most common and fastest-growing long-term diseases worldwide [5].

The Framingham Study reported a three-fold increase in cardiovascular mortality among individuals with diabetes, and a 5.1-fold higher risk of heart failure in diabetic women compared to a 2.4-fold rise in diabetic men [6]. Studies show that women with type 2 diabetes face a higher relative risk for coronary heart disease [7], stroke [8], left ventricular (LV) hypertrophy [9, 10], heart failure progression [11–13] and overall mortality [3, 14, 15] compared to diabetic men. Diabetic women also experience more severe forms of diabetic cardiomyopathy, with increased myocardial fibrosis and diastolic dysfunction [4, 9, 10, 16]. On the contrary, one study observed no discernible sex differences in diabetes associated abnormalities of left ventricular structure and function, and of arterial stiffness [17].

Animal studies reported an earlier and more pronounced diastolic dysfunction and remodelling in diabetic females [18], despite less pronounced hyperglycaemia [19]. It is evident that decreased developed LV pressure and increased LV end-diastolic pressure were found in diabetic female rats, while these parameters remained unaffected in diabetic males, indicating more severe dysfunction in females [20].

Several studies have presented contrasting findings, showing less severe dysfunction in female rodents compared to males [21]. One study demonstrated that diabetes caused a significant reduction in developed LV pressure in both sexes, but the reduction was less severe in females, which was associated with less reduction in the rates of pressure development and decay in female diabetic hearts [22]. This aligns with the observations of less severe dysfunction in contractile properties, including peak shortening, velocity of shortening, and sarcoplasmic reticular calcium ion (Ca²⁺) release in female diabetic mouse myocytes [23], and less impairment in the maximal velocity of tension development in diabetic female rats compared to their male counterparts [24, 25]. Diabetes also alters signalling pathways governing cardiac function differently in males and females by impairing ß-adrenergic receptor responsiveness in male rat hearts, whereas female hearts maintain more robust contractile responses [26]. Other studies show no significant sex differences in the myocardial response to Ca²⁺ changes under diabetic conditions [24, 25, 27]. Diabetes was associated with prolonged contraction and relaxation durations, which were more pronounced in male diabetic rats than in female diabetic rats [24, 27].

Moreover, type 2 diabetic cardiomyopathy involves dysregulated Ca²⁺ handling [28–31]. Diabetes did not significantly affect Ca²⁺ handling in female mouse myocytes, while it notably impaired these changes in male mouse myocytes [23]. However, female mouse myocytes had smaller increases in intracellular Ca²⁺ concentration and slower Ca²⁺ transient decay compared to males [23]. Diabetic female rat myocytes had more pronounced reduction of Ca²⁺ transient amplitudes, higher ryanodine receptor type 2 (RyR2) levels and less significant hyperphosphorylation [22].

In terms of cardiac energetics, while our earlier research showed no sex-specific disparities in cardiac energetics of healthy rat hearts [32], growing evidence suggests that diabetes may exemplify sex differences in cardiac energetics and efficiency. However, there have been only a few studies in this area. We are aware of only two studies reporting diabetic women experience reduced myocardial mechanoenergetic efficiency [33] and left ventricular efficiency [13], attributed to a shift from glucose to fatty acid oxidation, leading to higher oxygen consumption compared to diabetic men [11, 34, 35]. Despite the growing evidence for sex-specific differences in cardiac function and energetics under diabetic conditions, whether the diabetes induced sex difference in cardiac efficiency at the whole heart level manifests at the muscle level is uncertain.

This study aims to examine the effects of type 2 diabetes on cardiac energetics and efficiency in female versus male rat ventricular muscles. Using a high-fat diet (HFD) and low-dose streptozotocin (STZ) to induce diabetes, we measured the heat output and mechanical work in isolated left ventricular trabeculae using a work-loop calorimeter. By comparing control and diabetic female and male groups, we aim to determine whether type 2 diabetes imposes a greater energetic burden on one sex, using experimental protocols that measure activation heat, cross-bridge heat and mechanical efficiency under conditions of varying preloads and afterloads.

Methods

Female and male Wistar rats (6 weeks of age) were obtained from the University’s animal facility. Protocols for animal handling and euthanasia were approved by the Animal Ethics Committee of The University of Auckland (Ethics No. R002265).

Induction protocol for type 2 diabetes

We chose to start dietary intervention when the rats were 6 weeks old for two reasons. First, at this age, rats are considered adults – their vital systems are fully developed, and they have reached sexual maturity [36]. Second, 6 weeks is a commonly used age for inducing type 2 diabetes using the HFD/STZ model [37, 38], allowing our findings to be compared with existing studies. The rats were randomly assigned to the Control group or the Diabetes group. This resulted in four groups of 12 rats each: Control Female, Control Male, Diabetes Female, and Diabetes Male. On Day 1, control rats were fed a standard diet containing 5% fat (providing 13% of energy from fat), while diabetic rats were given a high-fat diet with 23.5% fat (diet code SF04-001 from Specialty Feeds, providing 45% of energy from fat).

At Week 3 post-diet, all diabetes groups received an intraperitoneal injection of low-dose STZ (30 mg/kg in citrate buffer, pH 4), while control groups were given an injection of citrate buffer. Blood glucose levels were measured on the 3rd, 5th, and 7th days post-STZ injection. If a rat’s glucose level was below 8 mmol/L on day 7 after STZ injection, a second STZ injection was carried out in Week 4 post-diet. Blood glucose levels and body weight were monitored weekly until Week 16. Experiments studying isolated ventricular trabeculae were conducted within Week 16–17 post-diet.

Rats that may have developed type 1 diabetes, following blood glucose of over 25 mmol/L consistently for eight weeks post-injection of STZ, were excluded from the study. In total, there remained n = 12 Control Female rats, n = 12 Control Male rats, n = 9 Diabetic Female rats, and n = 11 Diabetic Male rats.

Glucose tolerance monitoring

At the start of Week 16, a glucose tolerance test (GTT) was conducted on each rat after a 6-hour fasting period. At Time 0, blood glucose was measured with a glucometer and a sample of blood was collected and kept frozen. Each rat then received an intraperitoneal injection of glucose (1.5 g/kg). Blood glucose levels were measured at specific time points post-injection ((0, 15, 35, 65, 95, and 125) minutes), while blood samples were collected at (0, 35, and 125) minutes. Blood was collected via tail-tip puncture using a hypodermic 26G needle, with one needle used per rat. For each sampling session, approximately 150 µL of blood was collected into a Microvette CB300 Lithium heparin tube and kept on ice for later plasma insulin and corticosterone measurements. Blood samples were then centrifuged at 1000 relative centrifugal force (rcf) for 10 min at 4 °C to obtain plasma. The plasma samples were analysed for insulin and corticosterone using Enzyme-Linked Immunosorbent Assay (ELISA) kits; corticosterone (Invitrogen Cat# EIACORT) and Insulin (Crystal Chem Cat# 90060).

Heart isolation

On the day of the experiment within Week 16–17 post-diet, each rat was anaesthetised with isoflurane, given a subcutaneous heparin injection (1,000 IU/kg), and body mass measured prior to euthanasia by cervical dislocation. A standard protocol of brief induction with isoflurane (< 5% in O₂ (Oxygen) for 2–3 min), followed by maintenance at (1-1.5) % for approximately 10 min post-injection of heparin [39], was used. All animals treated identically. These short exposures at these doses have been shown to have negligible effects on cardiac function [40]. The heart was quickly excised and immersed in cold Tyrode's solution. Within 30 s, the heart was Langendorff perfused retrogradely at room temperature with oxygenated Tyrode's solution containing (in mmol/L): 130 Sodium Chloride (NaCl), 6 Potassium Chloride (KCl), 1 Magnesium Chloride (MgCl₂), 0.5 Sodium dihydrogen phosphate (NaH₂PO₄), 0.3 Calcium Chloride (CaCl₂), 10 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 10 glucose, 20 2,3-butanedione monoxime (to inhibit muscle contraction), and Tris (Tris(hydroxymethyl)aminomethane, to adjust pH to 7.4). Hearts were perfused to flush out blood and maintain tissue viability during dissection. The perfusion was performed by cannulating the aorta and delivering oxygenated Tyrode’s solution at a pressure of approximately 5.3 kPa (40 mmHg) via gravity, maintaining the aortic valve in a closed position to direct flow through the coronary circulation via the aortic ostia. It was performed under constant pressure mode and maintained for 30 min, during which time trabeculae from the heart was dissected. Dissected trabeculae were kept in the dissection bath receiving the perfusate until completion of the calorimetry experiments on the same day.

Blood collection at sacrifice

Blood samples were also collected immediately after cervical dislocation and heart excision by drawing blood from the thoracic cavity using a sterile syringe and transferring it into a 5 mL Ethylenediaminetetraacetic acid (EDTA) K2 vacuum tube for storage at -80 °C. Whole blood samples were preserved for HbA1c analysis, while plasma was separated by centrifugation (1000 rcf, 10 min, 4 °C) and stored at -80 °C for subsequent biomarker assessments. HbA1c was measured on a c311 autoanalyser by turbidimetric inhibition immunoassay for hemolysed whole blood utilising off-board dilution (Roche Diagnostics). Metabolite concentrations were measured on a Hitachi c311 autoanalyser (Hitachi High Technologies Corporation, Tokyo, Japan). Plasma concentrations of glucose (Glu), urea, alanine aminotransferase (ALT), Albumin (Alb), lactate dehydrogenase (LDH), aspartate aminotransferase (AST), creatinine (Creat), uric acid (UA), and lipid profile markers (HDL, LDL, and Trig) were measured using enzymatic UV or colorimetric assays. C-reactive protein (CRP) was analysed by particle enhanced immunoturbidimetric assay (Roche Diagnostics, Mannheim, Germany).

Trabeculae preparation

Trabeculae were isolated from the left ventricle under a dissecting microscope. A single trabecula was transferred and mounted between platinum hooks in a work-loop calorimeter [41]. The trabecula was continuously superfused with oxygenated experimental Tyrode's solution, identical in composition to the Tyrode's solution detailed above but with 1.5 mmol/L CaCl₂, without 2,3-butanedione monoxime, and with the pH adjusted with Tris to 7.4 at 37 °C. The superfusate was gravity-fed into the calorimeter chamber at a controlled rate of 0.6 µL/s, flowing from upstream to downstream along the muscle to ensure optimal heat sensitivity [42] and maintain muscle viability [43]. The upstream hook was connected to a linear length motor for muscle length control, while the downstream hook was attached to a force transducer to measure muscle force. Thermopile arrays beneath the calorimeter chamber measured the superfusate temperature upstream and downstream of the muscle. The rate of heat production of the trabecula was determined from the voltage difference between the upstream and downstream thermopiles.

After being mounted in the calorimeter, the trabecula was given an acclimatisation period of approximately 1 h. During this time, the trabecula was electrically stimulated at 2 Hz using a pair of platinum electrodes within the calorimeter. The device was contained within a light-proof and thermally insulated lid for maintaining an internal temperature of 37 °C. Experiments began once the contracting trabecula reached its mechanical performance acclimatisation and the calorimeter reached thermal equilibrium. The stimulation frequency was then increased to 4 Hz, and the trabecula was gradually stretched using the linear length motor until maximum active twitch force was achieved, identifying the optimal muscle length (L_o). Two experimental protocols (length-change and work-loop) were performed, with simultaneous measurements of muscle force, muscle length, and heat production rate throughout the experiment.

Length-change protocol

The trabecula was subjected to undergo a series of isometric (Isom) contractions at six different muscle lengths, ranging from the optimal length (L_o) to approximately 0.75 L_o, where muscle force was minimal. At each length, muscle isometric force and heat production rate reached steady states within ( [2, 3]) minutes. Electrical stimulation was halted between each length step to measure baseline heat when the trabecula was rendered quiescent.

Work-loop protocol

The trabecula was subjected to perform work-loop (WL) contractions at initial muscle length of L_o at six different afterloads, approximately set to 0.8, 0.7, 0.6, 0.5, 0.3, and 0.1 of the isometric active stress. The lowest afterload was close to the muscle passive force at L_o. Before each afterloaded work-loop contraction step, isometric contraction mode was performed to measure isometric force and heat production rate. Steady-state force, length, and heat rate during work-loop contractions were achieved within 2 min.

Post-experimental measurement and quantification

At the completion of the two contraction protocols, heat measurements from the work-loop calorimeter were corrected by subtracting two sources of artefacts. First, with the muscle quiescent by halting the stimulation, the length motor was oscillated at 4 Hz between muscle optimal length and the peak work-loop shortening length to measure the heat artefact generated by the motor movement. Second, to measure the stimulator heat, the muscle was transferred downstream of the measurement chamber into the mounting chamber. In the absence of the muscle in the measurement chamber, the stimulator was turned on and stimulus heat was measured.

On the day of the experiment, the rat lower left hindlimb was dissected to measure tibial length by placing a ruler along the bone. In the dissection bath, the wall thicknesses of both ventricles were measured under a microscopic graticule. The ventricles and atria were then weighed to determine their individual masses. Heart mass was measured by summing the wet mass of the ventricles and atria.

In data analysis, force was converted to stress (kPa) by normalising it to the muscle cross-sectional area, which was determined from measurements of muscle diameters at L_o in the calorimeter chamber taken in two perpendicular views (top and side) using a microscopic eye-piece graticule. Muscle volume was then calculated by using these diameters and muscle length. The calorimeter has a side mirror (45˚) that allows measurement of side diameter of a mounted trabecula. Twitch duration was measured at 5% of peak twitch force, separated into rise and fall. Twitch heat (kJ·m⁻³) was calculated by dividing the steady-state heat production rate by stimulus frequency and normalising it to muscle volume. Activation heat (kJ·m⁻³) was determined from the y-intercept of the relation describing steady-state isometric heat versus active stress data, while the inverse of the slope of the relation represented muscle economy. Afterload was defined as the stress at which the muscle transitioned from isometric to isotonic shortening during a work-loop, and relative afterload was calculated as the ratio of afterload to peak isometric total stress (active + passive). The extent of shortening was the difference between end-systolic and end-diastolic lengths during a work-loop. Muscle shortening velocity was measured from the length–time trace during the initial 10 ms to 20 ms of isotonic shortening, where velocity was maximal and nearly constant. This value was normalised to muscle length and expressed in units of s⁻¹. Work output (kJ·m⁻³) was determined by integrating stress as a function of relative muscle length over the twitch duration. The area within each work-loop quantified the external work output. Change of enthalpy (ΔH) was taken as the sum of work and active heat (kJ·m⁻³). Mechanical efficiency was expressed as the ratio of work output to ΔH (%).

Statistical analyses

Statistical analyses were performed using SAS software (SAS Institute Inc., Cary, NC, USA). A Two-way ANOVA was applied using the PROC MIXED procedure to assess the fixed effects of sex, diabetes, and their interaction, with animal as a random effect. Tukey-Kramer post-hoc tests were applied where appropriate. For analyses involving repeated measurements over time, repeated measures ANOVA was performed using PROC MIXED procedure, with time as the repeated factor and diabetes status and sex as the independent variables. To estimate variables of interest, i.e., peak values, polynomial regressions (up to the 3rd order) were fitted to the data, with regression lines (each derived from a single trabecula) averaged within groups using the random coefficient model in the PROC MIXED procedure. Data are reported as means ± standard errors, unless otherwise stated. Statistical significance was set at p < 0.05.

Results

Measurement of muscle dimensions

Each trabecula was assumed to have an elliptical cross-section. The study included a total of 46 female trabeculae (21 from the Control group and 25 from the Diabetes group) and 45 male trabeculae (18 Control and 27 Diabetes). These were isolated from the hearts of 21 female rats (12 Control, 9 Diabetes) and 23 male rats (12 Control, 11 Diabetes). Student’s t-test showed no significant differences in cross-sectional areas and muscle volumes between the Control and the Diabetes groups within each sex group. For female trabeculae, the Control group had average values (mean ± standard error ) of (0.07 ± 0.01) mm² and (0.29 ± 0.03) mm³, while the Diabetes group had (0.08 ± 0.01) mm² and (0.29 ± 0.04) mm³, respectively. For male trabeculae, the Control group measured (0.08 ± 0.01) mm² and (0.32 ± 0.04) mm³, whereas the Diabetes group had (0.08 ± 0.01) mm² and (0.31 ± 0.04) mm³, respectively.

Body mass and blood glucose weekly monitoring

Figure 1 presents weekly body mass and blood glucose measurements of rats (Female, Control: n = 12, Diabetes: n = 9; Male, Control: n = 12, Diabetes: n = 11) for the entire period (16 weeks of high-fat diet) of the study. There was an effect of sex (^#p < 0.001) and effect of diabetes (^*p < 0.001) in body mass, as tested by Two-way ANOVA. In females (Fig. 1A, left), diabetic rats exhibited a significant increase in body mass compared to controls (^*p = 0.0200), with a notable sex difference (^#p < 0.0001) compared to diabetic males. In contrast, male diabetic rats (Fig. 1A, right) showed no significant deviation in body mass from their respective controls. Sex difference was also evident between control groups of female and male rats (^&p < 0.0001). Blood glucose levels (Fig. 1B) were significantly elevated following injection of streptozotocin in diabetic rats compared to their controls tested by Two-way ANOVA (no effect of sex: p = 0.9056, effect of diabetes: ^*p = 0.002).

Fig. 1.

Weekly monitoring of type 2 diabetes progression over 16 weeks. Diet commenced at Week 0. At Week 3, and if applicable at Week 4, the diabetes groups received streptozotocin (STZ) injection, whereas the control groups received a buffer injection only at Week 3. (A) Body mass of female (Control: n = 12, Diabetes: n = 9) and male (Control: n = 12, Diabetes: n = 11) rats. Two-way ANOVA detects effects of sex (*p < 0.001) and diabetes on body mass (*p < 0.001). Diabetic female rats showed a significant increase in body mass compared to female control rats (^*p = 0.0200). A significant sex difference in body mass was evident (^#p < 0.0001) between male and female diabetes groups as well as between male and female control groups (^&p < 0.0001). (B) Blood glucose levels of the same female and male rats as in panel A. Two-way ANOVA detects no effect of sex (p = 0.9056) but effect of diabetes (^*p = 0.002) on blood glucose levels. ^*p < 0.05, Control vs. Diabetes; ^#p < 0.05 Diabetes Female vs. Diabetes Male; ^&p < 0.05 Control Female vs. Control Male. Values are means ± standard errors

Glucose tolerance test

Figure 2 illustrates blood glucose, plasma insulin, and corticosterone levels in rats during the glucose tolerance test at the start of Week 16 post-diet. There was no effect of sex (p = 0.4626) and effect of diabetes (^*p < 0.001) in blood glucose during glucose tolerance test, as tested by Two-way ANOVA. Glucose tolerance test revealed that diabetic rats exhibited significantly higher blood glucose levels than controls in both sexes (Female: ^*p = 0.0122, Male: ^*p = 0.0002; Fig. 2A). No significant differences in blood glucose levels were observed between diabetes male and diabetes female groups during the test. Plasma insulin levels (Fig. 2B) were not significantly different between diabetic rats compared to their controls in both sexes, as Two-way ANOVA detected effect of sex (^#p < 0.001) but not diabetes (p = 0.7727). Insulin levels in diabetic male rats were significantly higher compared to diabetic female rats (^#p = 0.0301). There was also a significant difference in plasma insulin between control male and control female (^&p < 0.0001). Corticosterone levels (Fig. 2C) in diabetic rats did not show a significant difference compared to their controls as detected by Two-way ANOVA (effect of sex: (^#p < 0.001, no effect of diabetes: p = 0.2799). Diabetic female rats had higher levels of corticosterone compared to diabetic male rats (^#p = 0.0423). Control female rats had significantly higher corticosterone levels than control males (^&p < 0.0001).

Fig. 2.

Glucose tolerance test of the rats. (A) In the glucose tolerance test, 1.5 g/kg of glucose bolus was injected intraperitoneally. The blood glucose level was measured before injection (0 min) and at 15, 35, 65, 95, and 125 min after injection. Two-way ANOVA detects no effects of sex (p = 0.4626), but effect of diabetes (^*p < 0.001) on blood glucose. Blood glucose in diabetic rats was significantly higher than their controls (Female: ^*p = 0.0122, Male: ^*p = 0.0002). No significant differences were observed between diabetes male and diabetes female groups. (B) Plasma insulin levels were not significantly different between diabetic rats compared to their control in both sexes as detected by Two-way ANOVA (effect of sex: ^#p < 0.001, but not diabetes: p = 0.7727). Insulin levels were lower in diabetic female rats compared to diabetic male rats (^#p = 0.0301). It was also were lower in control female rats compared to control male rats (^&p < 0.0001). (C) Corticosterone levels in diabetic rats were not significantly different from their controls as detected by Two-way ANOVA (effect of sex: (^#p < 0.001, but no effect of diabetes: p = 0.2799). Diabetes female group had higher levels of corticosterone compared to diabetes male group (^#p = 0.0423). Meanwhile, control female rats had significantly higher corticosterone levels than control males (^&p < 0.0001). ^*p < 0.05, Control vs. Diabetes; ^#p < 0.05 Diabetes Female vs. Diabetes Male; ^&p < 0.05 Control Female vs. Control Male. Values are means ± standard errors

Plasma biomarkers

Figure 3 indicates significant alterations in several plasma biomarkers between diabetic and control rats, with notable sex-specific differences. In female diabetic rats, high-density lipoprotein (HDL) cholesterol was significantly lower than their control (^*p = 0.0256), but in diabetes male group HDL approached significant difference compared to their control (^*p = 0.0583). Diabetic male rats had higher levels of low-density lipoprotein (LDL) cholesterol (^*p < 0.0001) and blood glucose (^*p = 0.0201) compared to their controls. Albumin (Alb) was higher in control female (^&p < 0.0001), but LDL was lower compared to control male (^&p = 0.0306). Diabetic values of Alb (^#p < 0.0001), HDL (^#p = 0.0012), LDL (^#p < 0.0001), CRP (^#p = 0.0475), and uric acid (UA) (^#p = 0.0167) were sex different. In Fig. 3K and L, triglycerides and glycated haemoglobin (HbA1c) were not significantly different between the control group and the diabetes group in both sexes.

Fig. 3.

Plasma biomarkers of the rats at the time of sacrifice. (A) Lactate dehydrogenase (LDH), (B) C-reactive protein (CRP), (C) Albumin (Alb), (D) High-density lipoprotein (HDL), (E) Low-density lipoprotein (LDL), (F) Alanine aminotransferase (ALT), (G) Aspartate aminotransferase (AST), (H) Creatinine (Creat), (I) Uric acid (UA), (J) plasma glucose (Glu), (K) Triglycerides (Trig), and (L) Glycated haemoglobin (HbA1c). Data are presented for female (Control: n = 12, Diabetes: n = 9) and male (Control: n = 12, Diabetes: n = 11) rats. ^*p < 0.05, Control vs. Diabetes; ^#p < 0.05 Diabetes Female vs. Diabetes Male; ^&p < 0.05 Control Female vs. Control Male. Values are means ± standard errors

Morphometrics of rats

Table 1 summarises the morphometric characteristics of the rats at sacrifice (at week 17 post-diet). Diabetes had no significant effects on the male rats. Diabetic female rats were heavier than control females (^*p = 0.0421); their normalised body mass (body mass normalised to tibial length) was also higher (^*p = 0.0183). These diabetic values were sex different: body mass (^#p = 0.0001), tibial length (^#p = 0.0001), normalised body mass (^#p = 0.0001), heart mass (^#p = 0.0001), normalised heart mass (^#p = 0.0006), heart mass to body mass (^#p = 0.0323), LV mass (^#p = 0.0001), normalised LV mass (^#p = 0.003), LV wall thickness normalised to heart mass (^#p = 0.0001), RV mass (^#p = 0.0019), normalised RV mass (^#p = 0.045), and RV wall thickness (^#p = 0.0186). There was evidence of sex difference between control male and control female in almost all variables.

Table 1.

Morphometric characteristics of the rats at sacrifice. Values are means ± standard errors. ^*p < 0.05, control vs. diabetes; ^#p < 0.05 diabetes female vs. diabetes male; ^&p < 0.05 control female vs. control male

	Control Female (n = 12)		Diabetes Female (n = 9)	Control Male (n = 12)	Diabetes Male (n = 11)
Age (Days)		165 ± 1&	161 ± 1*#	162 ± 1	163 ± 0
Body mass (g)		303.1 ± 4.5&	363.1 ± 8.4*#	563.6 ± 8.2	569.8 ± 27.1
Tibial length (mm)		40.5 ± 0.5&	40.6 ± 0.5#	46.0 ± 0.5	44.5 ± 0.5
Body mass/tibial length (%)		7.51 ± 0.17&	8.95 ± 0.17*#	12.26 ± 0.17	12.80 ± 0.57
Heart
Mass (g)		1.08 ± 0.03&	1.19 ± 0.03#	1.68 ± 0.06	1.62 ± 0.06
Mass/tibial length (g/mm)		0.027 ± 0.001&	0.029 ± 0.001#	0.037 ± 0.001	0.037 ± 0.001
Mass/body mass (%)		0.36 ± 0.00&	0.33 ± 0.01#	0.30 ± 0.01	0.29 ± 0.01
Left ventricle
Mass (g)		0.79 ± 0.02&	0.87 ± 0.02#	1.19 ± 0.06	1.18 ± 0.05
Mass/tibial length (g/mm)		0.019 ± 0.000&	0.021 ± 0.000#	0.026 ± 0.001	0.026 ± 0.001
Mass/body mass (%)		0.260 ± 0.004&	0.240 ± 0.007	0.211 ± 0.010	0.209 ± 0.008
Wall thickness (mm)		3.33 ± 0.10	3.55 ± 0.10	3.58 ± 0.06	3.56 ± 0.10
Wall thickness/tibial length (%)		8.25 ± 0.27	8.76 ± 0.27	7.79 ± 0.15	8.02 ± 0.28
Wall thickness/heart mass (mm/g)		3.09 ± 0.08&	2.99 ± 0.10#	2.16 ± 0.07	2.22 ± 0.09
Right ventricle
Mass (g)		0.18 ± 0.01&	0.21 ± 0.02#	0.32 ± 0.01	0.28 ± 0.01
Mass/tibial length (g/mm)		0.0045 ± 0.0000&	0.0053 ± 0.0000#	0.0069 ± 0.0000	0.0064 ± 0.0000
Mass/body mass (%)		0.060 ± 0.002	0.060 ± 0.005	0.056 ± 0.002	0.051 ± 0.002
Wall thickness (mm)		1.40 ± 0.05&	1.39 ± 0.04#	1.71 ± 0.08	1.74 ± 0.11
Wall thickness/tibial length (%)		3.46 ± 0.14	3.42 ± 0.10	3.72 ± 0.18	3.92 ± 0.24
Wall thickness/heart mass (mm/g)		1.31 ± 0.07&	1.17 ± 0.05	1.04 ± 0.07	1.10 ± 0.09

Mechanoenergetics of contracting trabeculae

Figure 4 displays representative data obtained from an isolated trabecula when subjected to experimental protocols. Figure 4A (left-hand panel) shows simultaneous recordings of twitch force and rate of heat output from a representative trabecula undergoing the isometric length-change protocol. Panels (i)–(iii) illustrate representative twitches and corresponding heat rate recordings at progressively shorter muscle lengths, ranging from the L_o to approximately 0.75 L_o. Both twitch force and rate of heat production progressively decreased as muscle length was decreased. On the right-hand panel, steady-state single twitch profiles at various muscle lengths are presented. Figure 4B presents representative results from the work-loop protocol applied to the same representative trabecula. The left-hand panel shows raw experimental data for stress and rate of heat production during work-loop contractions at various afterloads. Muscle rate of heat production was highest during isometric contractions and gradually decreased in work-loop contractions at lower afterloads. On the right-hand panel, steady-state force twitches at different afterloads are shown, along with associated length changes throughout each twitch. Parametric plots of twitch stress and length resulted plots of stress-length work-loops, where the width of each loop reflects the extent of muscle shortening, which increased as afterloads decreased.

Fig. 4.

Representative records of experimental protocols imposed on an isolated trabecula. (A) Representative records of length-change isometric twitches and heat production rate from a trabecula at progressively reduced muscle lengths ((“i” – “iii”) are L_o, 0.92 L_o and 0.82 L_o, respectively). Twitches were recorded upon electrical stimulation until steady states of force and heat rate were achieved. The right-hand panel shows superimposed twitch profiles at steady state across different muscle lengths, corresponding to “i” – “iii” in the left-hand panel. (B) Representative records of the work-loop protocol imposed on the same isolated trabecula. After achieving a steady state under isometric contractions (Isom), the trabecula underwent six afterloaded work-loop contractions (“a” – “f”). The right-hand panel shows superimposed steady-state profiles of relative muscle length versus time (top) and isotonic twitches at various afterloads (bottom), and the corresponding stress–length work-loops. The loop width indicates the extent of muscle shortening, while the area within each loop represents the muscle work output

Figure 5 (top) presents data from a single trabecula, illustrating increases in both total stress and passive stress with increasing muscle length. The work-loop (WL) end-systolic stress-length relation was lower than the isometric (Isom) total stress-length relation. The average of each of these stress-length relations is plotted in the middle panels. Diabetes and sex had no effect on these average relations, and had no effect on the peak value of stresses (obtained at L = L_o; i.e., isometric contraction) depicted in the bottom panels. We use the same approach for labelling and lettering in the subsequent figures, i.e. top rows contain representative data, middle rows contain average data, and bottom rows contain peak values.

Fig. 5.

Steady-state twitch stresses. Top row: data points from representative trabeculae are plotted against relative muscle length for the total stress consisting of the steady-state isometric (Isom) stress (derived from Fig. 4A) and the work-loop (WL) end-systolic stress (derived from Fig. 4B), and the passive stress (derived from Fig. 4). Stress length data are fitted with third-order polynomials. Middle row: Group-averaged stress-length relationships for female (Control: n = 21; Diabetes: n = 25) and male (Control: n = 18; Diabetes: n = 27) trabeculae. Bottom row: Peak values of passive, active, and total stresses; data obtained at L_o. Active stress is the difference between the total stress and the passive stress. There were no significant differences between groups. Two-way ANOVA detects no difference in main effects (sex, diabetes) and interaction term. Values are means ± standard errors

Figure 6 illustrates the steady-state isometric twitch kinetics obtained under the length-change protocol, with Panel A showing the twitch duration and Panel B showing the rates of twitch stress development. Twitch duration rise and fall increased with active stress (representative trabeculae shown in the top row; group-averaged regression lines shown in the middle row. Peak values of twitch duration (obtained at L_o) are shown in the bottom row, with no significant differences between the diabetes group and the control group within each sex group. In Fig. 6B, twitch stress development (rise; +dS/dt) and relaxation (fall; −dS/dt) as functions of active stress were not different between the diabetes group and the control group within each sex group. Peak values of ± dS/dt (bottom row) showed no significant differences between control and diabetic groups for both female and male trabeculae.

Fig. 6.

Steady-state isometric twitch kinetics measured under the length-change protocol. (A) Top row: Twitch duration as functions of active stress of representative female and male trabeculae. Data fitted using linear regression. Middle row: Group-averaged relations for twitch duration both fall and rise as functions of active stress in trabeculae of female (Control: n = 21; Diabetes: n = 25) and male (Control: n = 18; Diabetes: n = 27) rats. Bottom Row: Peak twitch durations (fall and rise), obtained at L_o, for female and male groups. (B) Top row: Rate of twitch stress development (rise; +dS/dt) and relaxation (fall; -dS/dt) as functions of active stress of representative female and male trabeculae. Data fitted using linear regression. Middle row: Group-averaged ± dS/dtstress relations of female (Control: n = 21; Diabetes: n = 25) and male (Control: n = 18; Diabetes: n = 27) trabeculae. Bottom row: Peak ± dS/dt, obtained at L_o , for female and male groups. No statistical differences were detected between Control and Diabetes groups in each sex group. Two-way ANOVA detects no difference in main effects (sex, diabetes) and interaction term. Values are means ± standard errors

Muscle shortening during work-loop contractions was calculated from the length trace (Fig. 4B) and expressed as a percentage of L_o. There were no effects of diabetes or sex in the average relationship between shortening extent and relative active afterload (Fig. 7A, middle) nor in the peak shortening extent (estimated by extrapolation to zero relative active afterload) (Fig. 7A, Bottom). Shortening velocity was determined as the maximum slope of the length-time trace (Fig. 4B) and normalised to L_o. Similarly, sex had no effect on the average relationship between shortening velocity and relative active afterload (Fig. 7B, middle) or its peak value (Fig. 7B, bottom).

Fig. 7.

Shortening kinetics of trabeculae subjected to the work-loop protocol. (A) Top rows: The extent of muscle shortening (calculated from the length trace in Fig. 4B) as a function of the relative active afterload in representative female and male trabeculae. Data were fitted using a 2nd order polynomial. Middle rows: Average relation of the extent of shortening and relative active afterload from trabeculae of female (Control: n = 21; Diabetes: n = 25) and male (Control: n = 18; Diabetes: n = 27) rats. Bottom rows: Peak extent of shortening for the muscle groups, calculated from the y-intercepts of the relations in the middle row above. (B) Top rows: Velocity of shortening as a function of relative active afterload (calculated from the length trace as in Fig. 4B) in representative female and male trabeculae. Data were fitted using a 3rd order polynomial. Middle rows: Average relation of velocity of shortening and relative active afterload from trabeculae of female (Control: n = 21; Diabetes: n = 25) and male (Control: n = 18; Diabetes: n = 27) rats. Bottom rows: Peak velocity of shortening for the muscle groups, calculated from the y-intercepts of the relations in the middle row above. Two-way ANOVA detects no difference in main effects (sex, diabetes) and interaction term. Values are means ± standard errors

Figure 8 plots heat-stress relations obtained from both the length-change and the work-loop protocols. There were no effects of diabetes and sex in the average heat-stress relationships (Fig. 8, middle). Similarly, diabetes and sex did not influence the cross-bridge heat (Q_xb) and the activation heat (Q_A), as shown in Fig. 8, bottom. Peak shortening heat (Q_s), calculated from the difference between the y-intercepts of the work-loop and isometric heat-stress relationships (Fig. 8, middle), was not affected by diabetes and sex (Fig. 8, bottom).

Fig. 8.

Twitch heat as a function of active stress. Top Rows: Isometric (Isom) heat (derived from Fig. 4A) and work-loop (WL) heat (derived from Fig. 4B) data from representative female and male trabeculae. Data were fitted using linear regression. Estimations of activation heat (Q_A) and peak cross-bridge heat (Peak Q_xb) were depicted. Cross-bridge heat (Q_xb) was quantified by subtracting Q_A from the heat measured under work-loop contractions – it consists of cross-bridge isometric heat and cross-bridge shortening heat (Q_s). Peak shortening heat (Q_s) was quantified by subtracting Q_A from the y-intercept of the work-loop heat-stress relation, i.e., the difference between the two y-intercepts. Middle rows: The average relations of heat and active stress from trabeculae of female (Control: n = 21; Diabetes: n = 25) and male (Control: n = 18; Diabetes: n = 27) rats. Bottom rows: Activation heat (Q_A), peak cross-bridge heat (Peak Q_xb), and peak shortening heat (Q_s) were not different between Control and Diabetes in both sex groups. Two-way ANOVA detects no difference in main effects (sex, diabetes) and interaction term. Values are means ± standard errors

Figure 9 shows external work output and the change of enthalpy (work plus heat) as functions of relative afterload. The middle rows of Fig. 9 indicate that the relationships were not different between the diabetes and the control groups, and between sexes. Peak values were calculated from these relations. As shown in the bottom rows, there were no significant effects of diabetes and sex in peak work (Female, Control: (0.66 ± 0.13) kJ/m³; Diabetes: (0.60 ± 0.13) kJ/m³; Male, Control: (0.44 ± 0.05) kJ/m³; Diabetes: (0.43 ± 0.07) kJ/m³) and peak change of enthalpy (Female, Control: (9.20 ± 1.36) kJ/m³; Diabetes: (8.37 ± 1.17) kJ/m³; Male, Control: (6.20 ± 0.55) kJ/m³; Diabetes: (6.63 ± 0.93) kJ/m³).

Fig. 9.

Mechanoenergetics of steady-state work-loop contractions. (A) Top rows: Mechanical work (the area of each afterloaded work-loop) as a function of relative afterload from representative female and male trabeculae. Data were fitted using 3rd order polynomial, and the peak value of work was quantified for each trabecula. Middle rows: Average relation of work and relative afterload from trabeculae of female (Control: n = 21; Diabetes: n = 25) and male (Control: n = 18; Diabetes: n = 27) rats. Bottom rows: Peak work was not different between Control and Diabetes in each sex group. (B) Top rows: Change of enthalpy as a function of relative afterload from representative female and male trabeculae. Data were fitted using 2nd order polynomial. Middle rows: Average relation of change of enthalpy and relative afterload from trabeculae of female (Control: n = 21; Diabetes: n = 25) and male (Control: n = 18; Diabetes: n = 27) rats. Bottom rows: The peak of change of enthalpy was not different between Control and Diabetes in each sex group. Two-way ANOVA detected no difference in main effects (sex, diabetes) and interaction term. Values are means ± standard errors

Figure 10A illustrates mechanical efficiency (the ratio of work to enthalpy of change) as a function of relative afterload. No significant difference in peak mechanical efficiency was found between control and diabetic males (Control: (10.34 ± 1.11)%; Diabetes: (7.66 ± 0.57) %), nor between control and diabetic females (Control: (10.02 ± 0.79)%; Diabetes: (9.41 ± 1.13) %).

Fig. 10.

The efficiencies of trabeculae under steady-state work-loop contractions. Top rows: Mechanical efficiency as a function of relative afterload from representative female and male trabeculae. Mechanical efficiency is the ratio of work to change of enthalpy (Fig. 9), where change of enthalpy is the sum of work and heat. The heat component consists of both cross-bridge heat and activation heat. Data were fitted using third-order polynomials. Middle rows: Average relation of mechanical efficiency and relative afterload from trabeculae of female (Control: n = 21; Diabetes: n = 25) and male (Control: n = 18; Diabetes: n = 27) rats. Bottom rows: There was no significant difference in peak mechanical efficiency between the diabetic male group and its respective control, nor between the diabetic female group and its respective control. Two-way ANOVA detects no difference in main effects (sex, diabetes) and interaction term. Values are means ± standard errors

Discussion

This study is the first to assess the impact of type 2 diabetes on cardiac energetics in a sex-specific manner. Type 2 diabetes was induced in rats by a combination of a high-fat diet and low-dose streptozotocin (STZ) injections. Rats were assessed for type 2 diabetes in terms of body mass, blood glucose (Fig. 1), and a glucose tolerance test (Fig. 2). Blood and plasma biomarkers were also analysed (Fig. 3). Mechanoenergetics of isolated left ventricular trabeculae were measured under various conditions, including different muscle lengths (Fig. 4A) and afterloads (Fig. 4B). We evaluated mechanoenergetics through protocols that allowed the muscle to contract either isometrically or shorten to generate mechanical work. These protocols enabled us to examine twitch kinetics (Fig. 6), muscle shortening (Fig. 7), work output (Fig. 9), and the energy involved in cross-bridge cycling (actomyosin adenosine triphosphatase) and intracellular Ca²⁺ cycling (activation heat) (Fig. 8). As such, we were able to estimate mechanical efficiency (Fig. 10) as a function of afterload as well as the peak value of efficiency. At the rat level, we induced type 2 diabetes, monitored body mass and blood glucose, and confirmed through blood glucose measurements, tolerance to glucose and blood biomarkers results. At the trabecula level, diabetes had no effect on muscle shortening, twitch force, mechanical work, heat production and, hence, mechanical efficiency within sexes.

We used a high-fat diet (23.5% kcal from fat) combined with low-dose STZ injections (30 mg/kg) to induce type 2 diabetes in our rats. This is a well-established model [44–47] for mimicking type 2 diabetes progression in humans [48]. This approach induces partial insulin deficiency [49] and increases insulin resistance [50, 51], effectively replicating disease development [52–54].

Different protocols, STZ doses and diet durations have been employed for the induction of type 2 diabetes in rats. Most studies used low-doses of STZ of 30–35 mg/kg [55–57], though higher doses (up to 50 mg/kg) were also tested [44, 48, 58] but more resembles type 1 diabetes [57]. Total feeding durations across studies ranged from 4 to 12 weeks, with shorter durations typically used in conjunction with higher STZ doses of 40–50 mg/kg [44]. In line with these, we used STZ at the lower range of concentration (30 mg/kg) with a prolonged diet duration (16–17 weeks).

Most previous studies using high-fat diet and low-dose STZ models have studied, independently, either male [45, 47] or female [56, 59] rodents, and, thus, there is limited research specifically examining sex differences. Studies focusing solely on high-fat diet interventions, without STZ injection, suggest females may require longer exposure or higher fat percentage diet to achieve similar diabetes severity as males [60–62]. This sex-specific response to a high-fat diet has not been investigated in the diabetes model with low-dose STZ. We note that a recent study employed a higher fat diet (60% of energy from fat) and a different STZ dose (30 mg/kg/day for 3 days of consecutive injections) reported successful diabetes induction only in male mice [63]. Another study using a slightly altered protocol (58% of energy from fat diet, single 35 mg/kg STZ dose) successfully induced diabetes in female Sprague-Dawley rats [56]. In our study, we applied the same diabetes induction protocol to both sexes in order to directly examine biological sex differences in cardiac response to diabetes. We did not apply a more severe diabetes induction protocol to female rats (i.e., >30 mg/kg STZ and/or > 17 weeks of high-fat diet) as two of the diabetic female rats developed type 1 diabetes, whereas only a diabetic male rat developed type 1 diabetes; these animals were excluded from the study.

Our diabetic rats exhibited glucose intolerance (Fig. 2A), and sex-specific plasma insulin response during the glucose tolerance test (Fig. 2B). Our findings of lower levels of plasma insulin in diabetic females than diabetic males align with findings of other studies [61, 64]. We also found blunted plasma insulin response to diabetes in both sexes, which does not appear to support the notion that females may have greater resilience against β-cell destruction but develop insulin resistance in diabetes [65–68]. We also observed no change in plasma corticosterone levels in diabetic rats (Fig. 2C), consistent with previous findings studying STZ-induced rats [69, 70]. Our observation of higher corticosterone levels in diabetic females compared to diabetic males may stem from the inherently higher basal corticosterone levels we saw in control females compared to control males (Fig. 2C), which aligns with findings from other studies [71, 72].

Our findings reveal sex-specific differences in blood and plasma biomarkers among diabetic rats (Fig. 3). In line with previous reports, LDL levels were elevated in diabetic males compared to controls [55, 73]. However, unlike those male-focused studies, we observed reduced HDL levels in diabetic females - a pattern not previously reported in this diabetic rat model. In contrast to human studies reporting higher LDL and HDL levels in diabetic females compared to diabetic males [74], we observed lower levels of both lipoproteins in diabetic female rats. Aligning with this trend, control females displayed lower LDL levels than control males. Unlike humans, rats lack cholesteryl ester transfer protein leading to fundamentally different lipoprotein profiles, typically characterized by lower LDL and higher HDL levels compared with humans [75]. The higher albumin (Alb) levels observed in females, both in control and diabetic groups, may reflect enhanced hepatic protein synthesis potentially mediated by estrogen or less diabetes-induced organ damage in females [76, 77]. Human studies indicate that elevated serum uric acid (UA) levels are associated with an increased risk of developing type 2 diabetes in both sexes, with some research suggesting a stronger association in women [78, 79]. However, in our study, UA levels were not significantly altered by diabetes in either sex, aligning with previous findings [57]; although we observed lower UA levels in diabetic females compared to diabetic males align with another paper on uricase-deficient rats, as hyperuricemia is a risk factor for diabetes [80].

Our observations of increased body mass in diabetic females compared to their controls (Table 1) are consistent with previous findings linking diabetes to greater adiposity in female rats [51, 81]. However, there are few studies reported lower body mass using same STZ injection dose but shorter diabetes induction period (10–12 weeks) [44, 82]. In particular, one study [82] performed cardiac echocardiography and showed lower ejection fraction, fractional shortening, diastolic dysfunction and increased left ventricular diastolic dimension in diabetic rats compared to their controls. That study also reported ventricular hypertropy, which we did not observe in our rats (Table 1). Interestingly, in Type 1 diabetes rats lower body mass [83] and hypertrophy [84, 85] has been reported, and the effects of diabetes on muscle mechanoenergetics were not observed [86].

Our findings of sex differences between control groups are consistent with previous studies highlighting morphological differences [87–89]. LV wall thickness was not significantly different between sexes and was not affected by diabetes; however, when normalised, it was higher in diabetic females. This is consistent with findings from both animal and clinical studies showing lower absolute LV mass in diabetic females ​ [13, 19], yet a greater increase in LV mass index in diabetic women compared to men [33]. However, RV wall thickness was found to be smaller in control female rats than in males and remained lower in diabetic females, potentially reflecting inherent anatomical sex differences, although studies specifically focusing on sex differences in RV wall thickness in diabetic rats are few.

Despite diabetes induced and sex-specific differences manifesting at the rat level, our results show no difference among groups in mechanoenergetics at the muscle level. We obtained null effects of diabetes on muscle force production (Fig. 5), shortening kinetics (Fig. 7), work output (Fig. 9) and twitch force kinetics (Fig. 6) in both sexes. We discuss our findings of muscle mechanics by comparing them to results from inherited and genetic diabetic rat models. In males, baseline left ventricular pressure magnitude and kinetics are preserved in Zucker type 2 diabetic and obese rats compared to their non-diabetic and lean counterparts [90, 91]. Similarly, maximal developed stress, twitch kinetics, and cardiomyocyte force remain unaffected in high-fat diet-fed obese rats and Zucker diabetic rats [92, 93]. However, impaired maximal rate of change of isometric twitch in papillary muscle [24, 25] and greater impairment in myocyte shortening [23] has been observed predominantly in male diabetic rodents. Another study also reported prolonged contraction and relaxation durations in diabetic rodent ventricular myocytes across both sexes [22]. Unlike these studies, our findings indicate no sex-specific diabetic effects on trabeculae, suggesting variability in the model of diabetes, the influence of sex on myocardial responses to diabetes across different choices of muscle preparation, experimental models and conditions.

In terms of muscle energetics, we found no effect of diabetes on activation heat (Fig. 8), which represents the energy expenditure arising from cellular Ca²⁺ cycling [94], primarily due to ATP hydrolysis by sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA). Reports on Ca²⁺ transient amplitude in diabetes are inconsistent, with greater reductions typically observed in males. While Yaras, Tuncay [22] noted smaller reductions of Ca²⁺ transient amplitude in diabetic females, Ceylan-Isik, LaCour and Ren [23] found preserved Ca²⁺ transient in female mice. These discrepancies may be due to sex differences in SERCA activity or Ca²⁺ uptake [95], as well as the varying effects of diabetes on these processes. Although SERCA protein expression and Ca²⁺ uptake appear unaffected by sex in non-diabetic models [89, 96, 97], diabetes may alter this balance. Mild type 2 diabetes may involve compensatory upregulation of SERCA2a [98], whereas advanced stage of type 2 diabetes or type 1 diabetes is often associated with SERCA2a downregulation [99]. Conversely, in the Zucker Diabetic Fatty rat heart, SERCA expression and its regulatory phosphorylation are reduced, yet Ca²⁺ uptake activity remains unchanged [100]. These suggest that inconsistencies in Ca²⁺ transient data may reflect differences in SERCA activity under varying diabetic conditions [98, 99]. Our findings of a null effect of diabetes on activation heat show that the same energy expenditure from intracellular Ca²⁺ cycling in control and diabetic muscles in both sexes.

We found a null effect of diabetes on cross-bridge heat, showing that cycling cross-bridges liberate the same amount of heat in control and diabetic muscles (Fig. 8). Diabetes is known to induce a shift toward β-myosin heavy chain expression, which is associated with lower actomyosin ATPase activity and slower shortening velocity [101]. Diabetes is also known to reduce actomyosin ATPase and myosin Ca²⁺-ATPase activity in rat hearts [102]. Our results of comparable cross-bridge heat, over a wide range of active force (Fig. 8), are not consistent with these reported effects of diabetes on actomyosin activity. A way to explain this discrepancy could be the difference in experimental preparations. While we used intact muscle, most other studies used skinned fibres with the preparation maximally activated and the investigators examined the contractile proteins in isolation and did not include the excitation functions [103]. In our experiments, muscle force was varied by using two different modes of contraction (isometric and work-loop and by varying muscle length and afterload). Given that we found muscle force, under these various loading conditions, is not affected by diabetes, it is not surprising that cross-bridge heat – energetic output arising from actomyosin activity, is also not affected by diabetes.

At any given muscle length (Fig. 5), end-systolic stress is higher during isometric contractions compared to work-loop contractions, which involve muscle shortening. This demonstrates the dependence of the cardiac force-length relationship on the mode of contraction. A similar dependency is observed in the heat-force relationship (Fig. 8), that is, for a given force, active heat production is greater during work-loop contractions than during isometric contractions, reflecting the heat associated with muscle shortening [104]. We found no effect of diabetes in the work-loop heat stress relation, the isometric heat-stress relation, and shortening heat. These results are consistent with our findings of no effects of diabetes on active force (Fig. 5) extent of shortening (Fig. 7), and active heat (Fig. 8). Thus, diabetes has no effect on cardiac contraction mode dependency, on both the force length plane (mechanics) and on the heat-force plane (energetics).

Diabetes had no impact on mechanical work or active heat, resulting in no impact on the change of enthalpy (Fig. 9) and, thus, no impact on mechanical efficiency within sexes (Fig. 10). In this study, we directly quantified energy expenditure by measuring heat output and calculated mechanical efficiency as the ratio of work output to the corresponding enthalpy change (the sum of work and heat). By contrast, previous assessments of cardiac efficiency in studying diabetes have relied on positron emission tomography and echocardiographic indices in humans [13, 33, 35], and mitochondrial respiration assays in isolated rat hearts [105]. These previous studies, which measured myocardial oxygen consumption using different methodologies, consistently reported that cardiac efficiency is impaired in diabetes, and more severely in diabetic females than in diabetic males. The divergence from our findings may reflect methodological differences, particularly the ability of direct calorimetry to assess the heat output arising from contraction, specifically the energy expenditure specific to intracellular Ca²⁺ cycling and to cross-bridge cycling. Measurement of oxygen consumption is not specific to these cycling processes.

Limitations

First, although our protocol allowed for a second STZ injection when initial hyperglycemia was not achieved, only a few animals required it, limiting our ability to assess the effects of double-injection versus single-injection. An echocardiography cardiac study [82] found that one to three STZ injections induced similar levels of cardiac dysfunction compared to controls, suggesting repeated dosing increases diabetes induction success without worsening pathology. However, subtle effects on energetics or sex-specific responses may still exist. Second, although we used brief isoflurane anesthesia which is unlikely to affect cardiac function of isolated trabeculae, longer exposures [39, 40] such as those used in the echocardiography study [82] may contribute to the cardiac dysfunction they observed in diabetic hearts. Finally, in our mechanoenergetic experiments we used a glucose-based Tyrode's solution without fatty acids, which may have masked diabetes-induced metabolic shifts toward fatty acid oxidation, particularly in females; this shift has been linked to reduced myocardial efficiency [11, 35].

Conclusions

In conclusion, this study provides the first direct assessment of sex-specific cardiac mechanoenergetics in type 2 diabetes. Although type 2 diabetes elicited sex-specific effects at the organism level, i.e., on rat’s body mass, corticosterone levels, plasma insulin, and other biomarkers, it does not affect any measured indices of cardiac energetics or mechanical efficiency in either sex. These results indicate that ventricular muscle mechanoenergetics functional performance in both males and females is preserved in the context of diabetes, revealing that sex-specific energetic differences are not manifested at the myocardial level in this rat model of high-fat diet and low-dose streptozotocin-induced diabetes.

Abbreviations

Alb

Albumin

ALT

Alanine Aminotransferase

ANOVA

Analysis of Variance

AST

Aspartate Aminotransferase

ATPase

Adenosine Triphosphatase

CaCl₂

Calcium Chloride

Creat

Creatinine

CRP

C-reactive protein

dS/dt

Rate of change of twitch stress

EDTA

Ethylenediaminetetraacetic acid

ELISA

Enzyme-Linked Immunosorbent Assay

Glu

Glucose

GTT

Glucose Tolerance Test

H

Change of Enthalpy

HbA1c

Glycated Hemoglobin

HDL

High-Density Lipoprotein

HEPES

4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

Isom

Isometric

KCl

Potassium Chloride

LDH

Lactate Dehydrogenase

LDL

Low-Density Lipoprotein

Lo

Optimal Muscle Length

MgCl₂

Magnesium Chloride

NaCl

Sodium Chloride

NaH₂PO₄

Sodium dihydrogen phosphate

O2

Oxygen

PROC

Procedure

QA

Activation Heat

Qs

Shortening Heat

Qxb

Cross-bridge Heat

rcf

Relative Centrifugal Force

RyR2

Ryanodine Receptor 2

SAS

Statistical Analysis System

SERCA

Sarcoplasmic/endoplasmic reticulum calcium ATPase

STZ

Streptozotocin

Trig

Triglycerides

Tris

Tris(hydroxymethyl)aminomethane

UA

Uric Acid

WL

Work-loop

Author contributions

M.R. and J. C.H. conceived and designed the study. M.R. performed experiments, conducted statistical analysis, prepared figures, and drafted the manuscript. L.N. was involved in the experimental induction of diabetes in rats and data collection. T.P., D.C., K.T., and A.T. contributed to the conception of the research and interpretation of the data. All authors edited and revised the manuscript. All authors approved the final version of the manuscript.

Funding

This study was funded by the Heart Foundation of New Zealand through Postgraduate Scholarship (1875, awarded to M.R.), Project Grant (1929, awarded to J. C.H.) Research Fellowship Grant (1896, awarded to T.P.), and Senior Research Fellowship Grant (awarded to D.C), the Health Research Council of New Zealand through Sir Charles Hercus Health Research Fellowship Grants (20/011 and 21/116; awarded to J. C.H., K.T., respectively), Explorer Grant (21/758, awarded to J. C.H.) and Emerging Researcher First Grant (21/653, awarded to T.P.), and the Royal Society of New Zealand through Marsden Project Grant (MFP UOA2206, awarded to J. C.H.) and James Cook Research Fellowship (awarded to A.T.).

Data availability

The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

All experimental procedures were approved by the Animal Ethics Committee of The University of Auckland (Ethics No. R002265).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.