Proteomic Analysis Reveals Sex-Specific Changes in Chronically Ischemic Swine Myocardium Treated with Sodium-Glucose Cotransporter-2 Inhibitor Canagliflozin

Abstract

Introduction:

Although sodium-glucose cotransporter-2 (SGLT-2) inhibitors have been shown to improve cardiovascular outcomes in general, little is presently known about any sex-specific changes that may result from this therapy. We sought to investigate and quantify potential sex-specific changes seen with the SGLT-2 inhibitor canagliflozin (CAN) in a swine model of chronic myocardial ischemia.

Methods:

Eighteen Yorkshire swine underwent left thoracotomy with placement of an ameroid constrictor. Two weeks post-op, swine were assigned to receive either control (F=5, M=5), or CAN 300 mg daily (F=4, M=4). Following five weeks of therapy, swine underwent myocardial functional measurements and myocardial tissue was sent for proteomic analysis.

Results:

Functional measurements showed increased cardiac output, stroke volume, ejection fraction, and ischemic myocardial flow at rest in CAN males compared to control males (all p<0.05). The CAN females had no change in cardiac function when compared to control. Proteomic analysis demonstrated six total up-regulated and 97 down-regulated proteins in the CAN female group compared to the female control. Pathway analysis showed decreases in proteins in the tricarboxylic acidic cycle. The CAN male group had 639 up-regulated and 172 down-regulated proteins compared to male control. Pathway analysis showed increases in pathways related to cellular metabolism and decreases in pathways relevant to the development of cardiomyopathy and to oxidative phosphorylation.

Conclusions:

Males treated with CAN had significant improvements in cardiac function that were not observed in females. Moreover, CAN treatment in males was associated with significantly more changes in protein expression than in females. The increased proteomic changes seen in the male CAN group likely contributed to the more robust changes in cardiac function seen in males treated with CAN.

Graphical Abstract:

Visual Abstract: Proteomic Analysis Reveals Sex-Specific Changes in Chronically Ischemic Swine Myocardium Treated with Sodium-Glucose Cotransporter-2 Inhibitor Canagliflozin (CAN). Eighteen Yorkshire swine underwent placement of an ameroid constrictor. Two weeks post-op, swine were assigned to receive either no drug (CON, F=5, M=5), or CAN at a dose of 300 mg daily (F=4, M=4). Following five weeks of therapy, swine underwent terminal harvest. Males treated with CAN exhibited significant improvements in cardiac function that were not observed in females. Moreover, CAN treatment in males was associated with significantly more changes in protein expression than in females. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Introduction:

Cardiovascular disease (CVD) is the leading cause of morbidity and mortality worldwide, with an estimated 19 million deaths attributed to CVD globally every year^1,2. It has accordingly been the subject of extensive research aimed at its prevention and amelioration, both across demographic lines and as it pertains to specific groups. Many studies have sought, for instance, to investigate sex differences in the clinical presentation, management, and outcomes of patients with chronic myocardial ischemia. This research has produced several notable findings, both on the individual and on the population levels^3–7. It has been shown, for instance, that the annual CVD mortality rate has been higher in women than in men since 1984. Perhaps related to this, women with acute coronary syndromes are less likely to undergo cardiac catheterization, less likely to receive timely reperfusion, and less likely to be treated with optimized medical therapies^3–7. Clinically, women are less likely to present with typical anginal chest pain, and instead often exhibit other symptoms such as dyspnea, arm/back/jaw pain, lightheadedness, or loss of appetite^3–7. Given these disparate clinical manifestations and outcomes, it is essential that basic science be conducted in a manner that renders it sensitive to sex differences. Such methodology will help to reveal whether there are true biological underpinnings to these phenomena, or whether they are instead attributable to sociologic distortion.

In the nearly one-third of all patients with coronary disease who are poor candidates for surgical or percutaneous intervention, these considerations are compounded. if diverse therapeutic options for patients with insufficient revascularization options are limited in general, they are all the more circumscribed in the event that there are important sex-specific differences neglected during the cultivation of novel therapies^8,9. Moreover, the stakes are high. Chronic untreated myocardial ischemia ultimately precipitates adverse myocardial remodeling, necrosis, and subsequent heart failure¹⁰. Thus, investigations into potential medical therapies to attenuate the complications caused by chronic myocardial ischemia are critical to quality of life and longevity within this population.

Sodium-glucose cotransporter-2 (SGLT-2) inhibitors are a group of medications primarily used as antihyperglycemic agents in patients with diabetes mellitus¹¹. Recent studies, however, have shown them to possess cardioprotective effects that appear to arise independently from the effect of these agents on glycemic regulation¹². Human outcomes studies of these effects have been very promising, showing decreased cardiovascular mortality and decreased heart failure readmissions in patients treated with SGLT-2 inhibitors and earing these medications a 1A recommendation within the heart failure treatment algorithm^13–16.

Despite their clinical promise, the precise mechanism whereby SGLT-2 inhibitors produce cardioprotective effects is poorly understood. Non-diabetic animal models have produced several hypotheses in this regard, however: in animals, SGLT-2 inhibitor therapy has decreased infarct size, augmented endothelium-dependent vasorelaxation, mitigated oxidative stress, and improved myocardial perfusion and fibrosis^17–20. Our group has recently demonstrated, in particular, that treatment with the SGLT-2 inhibitor canagliflozin (CAN) results in improved cardiac function and perfusion in a swine model of chronic myocardial ischemia²¹. Given the paucity of current knowledge regarding sex-specific responses to CAN, we utilized our swine model in an effort to characterize the same, drawing for this purpose on a combined myocardial functional and proteomic analysis.

Methods:

Humane Animal Care:

The Principles of Laboratory Animal Care and the Guide for the Care and Use of Laboratory Animals were followed for the design of all animal experiments²¹. The Rhode Island Hospital Institutional Animal Care and Use Committee approved and monitored the conduction of our protocol (Protocol #505821).

Model:

Chronic myocardial ischemia was modeled in 18 Yorkshire swine (Cummings School of Veterinary Medicine of Tufts University Farm, North Grafton, MA, USA) by placing an ameroid constrictor (Research Instruments SW, Escondido, CA, USA) onto the left circumflex coronary artery (LCx). Swine then recovered for two weeks before receiving either control (CON, n=10, F=5, M=5) or CAN 300 mg daily (n=8, F=4, M=4). After 5 weeks of therapy, all swine were sedated for functional measurements and subsequently underwent tissue harvest for further analysis²¹.

Ameroid Constrictor:

Anesthesia and preoperative care were performed as previously described²¹. The swine were placed in a modified right lateral decubitus position and prepped with betadine. The LCx was exposed using a left thoracotomy as detailed in our previous work, isolated back to the left main coronary artery, encircled with a vessel loop, and finally occluded for two minutes. A 5 ml solution of gold microspheres (BioPal, Worcester, MA, USA) was concomitantly injected into the right atrium to delineate the area at risk for ischemia. After the two minutes elapsed, the ameroid constrictor was placed at the base of the LCx as proximal to the left main coronary artery as possible to create consistent infarcts across subjects. The pericardium and chest were closed with absorbable suture as previously described once adequate ameroid positioning was visually confirmed. Post-operative management then proceeded as previously described²¹.

Harvest and Functional Measurements:

Anesthesia and preoperative care were performed analogously to our methods as previously reported²¹. The swine were placed supine and prepped with betadine. A median sternotomy was performed, and the heart was dissected free from pericardial and sternal adhesions using a combination of blunt and sharp dissection. The left and right atrium were connected to a pacemaker (OSCOR, Palm Harbor, FL, USA). The femoral artery was accessed via inguinal cutdown. The femoral artery was cannulated with a 7 French vascular sheath. Perfusion was measured by injecting microspheres (BioPal, Worcester, MA, USA) into the left atrium while withdrawing a constant volume of fluid from the femoral artery. This was performed both at rest and while pacing the heart at 150 beats per minute. A PV loop catheter (Transonic, Ithica, NY) was placed into the left ventricle using Seldinger technique to facilitate acquisition of functional data. After these were gathered, the swine were euthanized, their organs removed, and hearts sectioned and snap-frozen in liquid nitrogen. Gold microsphere analysis was employed to reveal the most ischemic tissue for analysis²¹. An expedited harvest was conducted on one of the female controls (CONF) due to interoperative instability. Accordingly, we do not have perfusion or functional data for this animal. Additionally, the rest flow failed to map in one control female animal. These circumstances resulted in an n of four for most female animal functional data and n of three for female animal rest perfusion.

Proteomics:

Tissue samples obtained from the most ischemic aspects of the sectioned heart were sent to the proteomics core at The University of Massachusetts-Boston (Boston, MA, USA) for analysis. Samples were extracted using T-PER tissue lysis buffer (Thermo Fisher Scientific, Waltham, MA, USA). Overnight acetone precipitation was used to buffer-exchange samples into digestion buffer/TMTpro labeling buffer (100 mM TEAB) (Thermo Fisher Scientific, Waltham, MA, USA). Samples were reduced with DTT, alkylated with IAA, digested overnight with Trypsin/LysC, labeled with TMTpro 16 plex reagent (Thermo Fisher Scientific, Waltham, MA, USA), and multiplexed. Multiplexed samples were then fractionated into 10 constituents via RP-HPLC (Thermo UltiMate 3000), and fractions were analyzed via nLC-MS on an EASY-1200 nLC (Thermo Fisher Scientific, Waltham, MA, USA) coupled online to an Orbitrap Fusion Lumos (Thermo Fisher Scientific, Waltham, MA, USA). Raw data were processed using Thermo Fisher’s Proteome Discoverer (Thermo Fisher Scientific, Waltham, MA, USA) with an FDR of 0.1%. Proteins with unadjusted p-values less than 0.05 and fold changes of greater than two or less than one-half were considered significant. Pathway analysis was performed using ShinyGO 0.77 (South Dakota State University, Brookings, South Dakota, USA).

Immunoblotting:

Immunoblotting was used for protein quantification/proteomics validation. Lysates were made from tissue regions analogously ischemic to those sent for proteomics using RIPA Lysis Buffer and the Halt Protease Inhibitor Cocktail (ThermoFisher Scientific, Waltham, MA, USA) as previously described. The lysate (40μg) was run on a 4–12% Bis-Tris gel (ThermoFisher Scientific, Waltham, MA, USA) and transferred overnight²². Primary antibody dilutions were prepared according to manufacturer recommendations and incubated overnight in a cold room. Horseradish peroxidase-linked secondary antibodies to mouse or rabbit (Cell Signaling, Danvers, MA. USA) were incubated for one hour at room temperature (Supplemental Table 1). Membranes were developed and imaged on a ChemiDoc Imaging System (Bio-Rad, Hercules, CA, USA) in the manner used and delineated previously. Restore PLUS Western Blot Stripping Buffer (ThermoFisher Scientific, Waltham, MA, USA) was used to strip the membrane to allow for repeat probing. Immunoblot intensity was measured using NIH Image J software²².

Antibodies:

Primary antibodies to isocitrate dehydrogenase 2 and mitochondrial pyruvate carrier 1 were obtained from Cell Signaling (Danvers, Massachusetts, USA). Primary antibody to total oxidative phosphorylation was obtained from Abcam (Cambridge, England, UK) (Supplemental Table 1).

Immunofluorescence:

Immunofluorescence staining was performed on frozen sections of the ischemic myocardium as previously detailed in our work²¹. Slides were incubated with primary antibody to α smooth muscle actin (Abcam, Cambridge, UK), isolectin B4 conjugated to Alexa Fluor 647 (Thermo Fisher Scientific, Waltham, MA. USA), and DAPI (Cell Signaling, Danvers, Massachusetts, USA) (Supplemental Table 1). Images were acquired at 20X magnification with an Olympus VS200 Slide Scanner (Olympus Corporation, Tokyo, Japan), and analyzed with QuPath software²¹. Because tissue staining was conducted before completion of all harvests, capillary and arteriolar density were measured only for three of the five CON F animals.

Statistical Analysis:

All statistical analysis other than that pertaining to proteomics data was performed using Prism 9 (GraphPad Software, San Diego, CA, USA). Data was tested for normality with the Shapiro-Wilk test. Parametric data was analyzed using Student’s t-test or one-way analysis of variance (ANOVA). Non-parametric data was analyzed with the Mann–Whitney U test or the Kruskal–Wallis one-way ANOVA. Immunoblot data is shown as mean fold change of the band intensity normalized to the average control. Data is presented as mean and standard deviation. Outliers greater than two standard deviations from the mean were excluded.

Results:

Cardiac Function:

The CON male group (CON M) exhibited lower stroke work, stroke volume, cardiac output, and ejection fraction than the CON F group (all p<0.05, Figure 1). There was a significant increase in stroke work, stroke volume, cardiac output, ejection fraction, and the preload recruitable stroke work (PRSW) slope in the CAN male group (CAN M) compared to the CON M (all p<0.05) (Figure 1). There was no change in stroke work, stroke volume, cardiac output, ejection fraction, myocardial perfusion at rest, or in the PRSW slope between the CAN female group (CAN F) and the CON F (all p>0.05, Figure 1). There was no change in the end-systolic pressure-volume relationship (ESPVR), systolic blood pressure, diastolic blood pressure, mean arterial blood pressure, or heart rate between any of the groups (all p>0.05,Figure 1 and Supplemental Figure 1).

Figure 1: Cardiac Function.

The CON male group (CON M, n=5) exhibited lower stroke work, stroke volume, cardiac output, and ejection fraction than the CON female group (CON F, n=4). There was a significant increase in stroke work, stroke volume, cardiac output, ejection fraction, and the preload recruitable stroke work (PRSW) slope in the canagliflozin male group (CAN M, n=4) group compared to the CON M group. There was no change in stroke work, stroke volume, cardiac output, ejection fraction, myocardial perfusion at rest, or the PRSW slope between the canagliflozin female group (CAN F, n=4) and the CON F group. There was no change in the end-systolic pressure-volume relationship (ESPVR) between any of the groups. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Myocardial Perfusion and Vascular Density:

There was a significant increase in myocardial perfusion at rest in the CAN M group compared to the CON M group (p=0.02). There was no change in ischemic myocardial flow with pacing between the CAN M and CON M groups (p=0.23). There was no change in ischemic myocardial flow at rest or with pacing between the CAN F and CON F groups (all p>0.05, Figure 2A). There was no significant difference in ischemic myocardial flow at rest or with pacing between the CON M and CON F groups (Figure 2A). There was no difference in capillary density using isolectin B4 staining or arteriolar count using α smooth muscle actin between any of the groups (all p>0.05, Figure 2B and Figure 2C).

Figure 2: Ischemic Myocardial Perfusion and Vascular Density.

Figure 2 A: There was a significant increase in myocardial perfusion at rest in the canagliflozin male group (CAN M, n=4) compared to control males (CON M, n=5). There was no change in ischemic myocardial flow with pacing between the CAN M and CON M groups. There was no change in ischemic myocardial flow at rest between the canagliflozin female group (CAN F, n=4) and control females (CON F, n=3). There was no change in ischemic myocardial flow with pacing between the CAN F group (n=4) and control females CON F (n=4). There was no significant difference in ischemic myocardial flow at rest or with pacing between the CON M and CON F group. Figure 2B: There was no difference in capillary or arteriolar count between any of the groups (CON F n=3, CON M n=5, CAN F n=4, CAN M n=4). Data is represented as mean fold change normalized to CON F. Figure 2C: immunofluorescent stain for arterioles using alpha-SMA (yellow), capillaries using isolectin B4 (magenta), and cell nuclei with DAPI (blue). There was no difference in capillary density or arteriolar count between any of the groups. *p<0.05.

Overall Proteomics Results:

Proteomic analysis of the ischemic myocardium quantified 3,256 proteins. There were 97 significantly down-regulated proteins and six significantly up-regulated proteins in the female CAN group compared to the female CON group (Figure 3A). Pathway analysis showed decreases in proteins involved in the tricarboxylic acid cycle (TCA, Figure 3C). The male CAN group had 172 significantly down-regulated and 639 significantly up-regulated proteins when compared to male controls (Figure 3B). Pathway analysis showed increases in pathways contributing to glycolysis and complex carbohydrate metabolism (Figure 3D). The male CAN group had decreases in pathways contributing to oxidative phosphorylation and myocardial contraction (Figure 3E).

Figure 3: Volcano Plots and Enrichment Analysis.

Figure 3A: Volcano plot showing all proteins in the female myocardium. There were 97 significantly down-regulated proteins (green dots in the green box) and six significantly up-regulated proteins (red dots in the red box) in the canagliflozin female group when compared to the control female group. Figure 3B: Volcano plot showing all proteins in the male myocardium. The male canagliflozin group had 172 significantly down-regulated (green dots in the green box) and 639 significantly up-regulated proteins (red dots in the red box) when compared to the male control group. Figure 3C: Pathway analysis showed decreases in proteins involved in the tricarboxylic acid cycle in the canagliflozin female group when compared to the control female group. Figure 3D: Pathway analysis showed increases in pathways contributing to glycolysis and complex carbohydrate metabolism in the male canagliflozin group when compared to controls. Figure 3E: The canagliflozin male group demonstrated decreases in pathways contributing to oxidative phosphorylation and myocardial contraction.

TCA and Glycolysis:

Proteomics enrichment analysis yielded significant decreases in TCA cycle proteins in the CAN F group when compared to CON F counterparts, including malate dehydrogenase and succinate ligase (all p<0.05, Figure 4A). Interestingly, enrichment also demonstrated a significant increase in TCA and glycolysis proteins in the CAN M group when compared to CON M counterparts, including glucose 6 phosphate isomerase, hexokinase 1, hexokinase 2, malate dehydrogenase, pyruvate kinase, and fumarate hydratase (all p<0.05, Figure 4A). Immunoblotting showed further increases in the TCA-related proteins isocitrate dehydrogenase and mitochondrial pyruvate carrier 1 in the CAN M group compared to the CON M group (p=0.002 and p=0.03, Figure 4B and Figure 4C). Mitochondrial pyruvate carrier 1 was significantly decreased in the CAN F compared to CON F (p=0.03, Figure 4B and Figure 4C).

Figure 4: The Tricarboxylic Acid Cycle and Glycolysis.

Figure 4A: Proteomic analysis showed a significant decrease in tricarboxylic acid cycle (TCA) cycle proteins in canagliflozin-treated females (CAN F, red) compared to control females (CON F), including malate dehydrogenase and succinate ligase. Proteomics also showed a significant increase in TCA and glycolysis proteins in canagliflozin males (CAN M, blue) compared to control males (CON M), including glucose 6 phosphate isomerase, hexokinase 1, hexokinase 2, malate dehydrogenase, pyruvate kinase, and fumarate hydratase. Female changes are rendered in red and male in blue. Data is represented as mean fold change. The fold change for females is calculated as average CAN F divided by average CON F. The fold change for males is calculated as average CAN male divided by average CON male. Figure 4B: Immunoblotting showed increases in TCA-related proteins isocitrate dehydrogenase 2 (IDH2) and mitochondrial pyruvate carrier 1 (MCP-1) in the CAN M group compared to CON M. MCP-1 was significantly decreased in the CAN F compared to CON F. Data is represented as mean fold change normalized to CON F. Figure 4C: immunoblotting bands for IDH 2 and MCP-1. *p<0.05, **p<0.01.

Oxidative Phosphorylation:

Enrichment also showed a significant decrease in mitochondrial complexes in the CAN M group when compared to the CON M group, including in Complexes III, IV, and V (all p<0.05, Figure 5A). This was validated by immunoblotting for total oxidative phosphorylation that analogously showed significant decreases in Complexes I, II, and V (all p <0.05), and a trend towards decreased Complex IV (p=0.06, Figure 5B and Figure 5C).

Figure 5: Oxidative Phosphorylation.

Figure 5A: Proteomic analysis showed a significant decrease in mitochondrial complexes in the canagliflozin male (CAN M) group compared to control males (CON M), including in Complexes III, IV, and V. Data is represented as mean fold change. The fold change is calculated as average CAN male divided by average CON male. Figure 5B: Mitochondrial complex changes were validated by immunoblotting for total oxidative phosphorylation, which yielded significant decreases in Complexes I, II, and V, as well as a trend towards decreased Complex IV (p=0.06). Data is represented as mean fold change normalized to control females (CON F). Figure 5C: immunoblotting bands for total oxidative phosphorylation. *p<0.05.

Hypertrophic Cardiomyopathy:

Proteomic analysis of the CAN M myocardium additionally revealed a decrease in proteins upregulated in cardiac hypertrophy and hypertrophic cardiomyopathy, including myosin-4, myosin heavy chain 7, desmin, collagen type 1, alpha actinin, and tropomyosin alpha-1 compared to the CON M group (all p<0.05, Figure 6).

Figure 6: Hypertrophic Cardiomyopathy.

Proteomic analysis of the canagliflozin male myocardium demonstrated a decrease in proteins upregulated in cardiac hypertrophy and hypertrophic cardiomyopathy, including myosin-4, myosin heavy chain 7, alpha actinin, desmin, collagen type 1, and tropomyosin alpha-1 compared to control males. Data is represented as mean fold change. The fold change is calculated as average canagliflozin male divided by average control male.

Discussion:

Despite the established status of CAD as a sex-specific illness, unfortunately little is known about sex-specific responses to CAD-targeted therapies. Our study attempted to make headway in this regard, and indeed showed several sex-specific responses to the use of canagliflozin in a swine model of chronic myocardial ischemia. We showed a significant increase in several key markers of cardiac function in males treated with CAN that were not seen in females, including stroke work, stroke volume, cardiac output, ejection fraction, and the PRSW slope. These differences carry the important implication that CAN therapy appears to produce more significantly augmented cardiac function in males than in female subjects. It is further important to note, however, that the CON M had a lower stroke work, stroke volume, cardiac output, and ejection fraction than the CON female group. This could suggest that some of the more robust improvements in cardiac functions seen in males is due to lower baseline function, potentially rendering this study more sensitive to improvements in male function compared with that in female counterparts.

Beyond these functional differences, male swine treated with CAN also demonstrated a significant increase in myocardial perfusion that was not seen in the female cohort. Interestingly, this increase in myocardial perfusion was not accompanied by a change in vascular density, and the proteomics enrichment analysis did not highlight any pathways involving angiogenesis. These data, when taken together with the changes in microvascular reactivity we have demonstrated with CAN in a rodent model, suggest that CAN treatment may selectively modulate vascular reactivity in the tissue of male subjects, but microvascular studies in swine are needed to confirm this hypothesis²³. Additionally, it is important to note that conclusions regarding perfusion at this time are limited significantly by sample size, as the female group only has three data points for perfusion at rest.

Healthy myocardial tissue derives much of its energy from fatty acid oxidation, while myocardial ischemia precipitates dysregulation in myocardial metabolism^24,25. Data from our study showed a significant decrease in several enzymatic steps within the TCA cycle, including malate dehydrogenase and succinate ligase, in females treated with CAN. Males treated with CAN, on the other hand, showed the inverse: an increase in several TCA-related and glycolytic markers, including glucose 6 phosphate isomerase, hexokinase 1, hexokinase 2, malate dehydrogenase, pyruvate kinase, and fumarate hydratase. These increases seen in the male ischemic myocardium are mechanistically as well as clinically significant. Mechanistically, increased glycolysis and TCA cycle activity likely represent a shift towards increased myocardial energy production. Clinically, modulation of TCA cycle metabolites has been linked to disease amelioration, with increasing fumarate and decreasing succinate accumulation having both been linked to decreased infarct size and augmented cardioprotective efficacy^26,27.

Proteomic analysis also demonstrated differences related to oxidative phosphorylation. Specifically, CAN treatment was correlated with decreased mitochondrial complex expression in the male myocardium. Decreased oxidative phosphorylation is a key pathophysiologic component of myocardial ischemia and often represents a shift to increased glycolytic metabolism. The increased glycolytic proteins seen in the CAN M group are likely related to decreased oxidative phosphorylation; the decrease in oxidative phosphorylation seen in the CAN M group may point towards an underlying protective metabolic state characterized by decreased oxidative stress and, consequently, reduced myocardial damage²⁸.

Another modality of difference seen with CAN treatment related to decreases in several myocardial proteins of relevance to the pathogenesis of hypertrophic cardiomyopathy and heart failure. Tropomyosin and myosin, contractile proteins linked to cardiac hypertrophy, were decreased in the CAN M ischemic myocardium^29,30. Collagen type I and alpha actin, which are structural proteins also linked to cardiac hypertrophy, were also decreased in the CAN M ischemic myocardium^31,32. Finally, desmin, an intermediate filament protein linked to hypertrophic cardiomyopathy, was decreased in the CAN M group as well when compared to the CON M ischemic myocardium³³.

Synthesizing these differences, our work revealed that CAN treatment was associated with significant improvements in myocardial function and perfusion in male swine. It was also associated with significantly more changes in protein expression in the ischemic male swine than were seen in female swine counterparts. The protein expressive changes in CAN males included increases in glycolytic and TCA cycle-related intermediates and decreases in pathways contributing to oxidative phosphorylation and hypertrophic cardiomyopathy. The increased proteomic changes seen with male CAN likely contribute to the more robust changes in cardiac function also seen in males treated with CAN. Likewise, the lack of significant functional changes in CAN F is likely related to the relatively small number of proteomic changes in CAN F from CON F. These sex-specific results may aid in the ongoing development of precision medicine for ischemic heart disease.

While this study greatly expands our understanding of the sex-specific changes seen with CAN treatment, it is not without limitations in this regard. Our low sample size, while selected in accordance with humane principles to minimize the use of experimental animals, may secondarily imply a study that is underpowered to detect small changes. Additionally, the female pigs in this study are young with intact ovaries: although we collected blood samples from a small set of the harvest pigs confirming low estrogen levels in females, it is important to remember that most females with CAD are postmenopausal³⁴. Methodologically, our selection of proteomics as a means of assessing differences in treatment effects constitutes an inherent limitation, as proteomic analyses are heterogeneous, raising the possibility that another proteomic modality might discover different numbers and forms of proteins not captured by the results of our study. Finally, our decision to study CAN confers the limit that the results of this study could be different with other SGLT-2 inhibitors. Follow-up studies would ideally compare effects across multiple agents in this class with a similar analytic strategy.

Conclusion:

Males treated with CAN demonstrated significant improvements across multiple cardiac functional parameters, including stroke work, stroke volume, cardiac output, ejection fraction, and the PRSW slope. These improvements were not replicated in females treated with CAN. CAN treatment was associated with increased blood flow to ischemic myocardial tissue independently of vascular density. Proteomic analysis of CAN-treated tissue yielded significantly more changes in protein expression among ischemic male swine than were seen in female swine, including decreased protein expression in pathways contributing to hypertrophic cardiomyopathy. The increased proteomic changes seen with the male CAN group likely contributed to the more robust changes in cardiac function delineated previously. These results may augment the ongoing effort to develop novel therapies for ischemic heart disease in a manner that remains faithful to the unique molecular environments seen across demographic divisions.

Supplementary Material

SDC 1-2

Supplemental Figure 1: Heart Rate and Blood Pressure. There was no change mean arterial pressure, systolic blood pressure, diastolic blood pressure, and heart rate between any of the groups. Groups are as follows: canagliflozin female group (CAN F, n=4), canagliflozin male group (CAN M, n=4), control female group (CON F, n=4), control male group (CON M, n=5).

Supplemental Table 1: Antibody Catalog Number.

Abbreviations:

CAD

coronary artery disease

CAN

canagliflozin

CAN F

canagliflozin female

CAN M

canagliflozin Male

CON

control

CON F

control female

CON M

control male

CVD

cardiovascular disease

ESPVR

end-systolic pressure-volume relationship

LCx

left coronary circumflex

PRSW

preload recruitable stroke work

SGLT-2

Sodium-glucose Cotransporter-2

TCA

Tricarboxylic Acid Cycle