Cardioprotective O-GlcNAc signaling is elevated in murine female hearts via enhanced O-GlcNAc transferase activity

Abstract

The post-translational modification of intracellular proteins by O-linked β-GlcNAc (O-GlcNAc) has emerged as a critical regulator of cardiac function. Enhanced O-GlcNAcylation activates cytoprotective pathways in cardiac models of ischemia-reperfusion (I/R) injury; however, the mechanisms underpinning O-GlcNAc cycling in response to I/R injury have not been comprehensively assessed. The cycling of O-GlcNAc is regulated by the collective efforts of two enzymes: O-GlcNAc transferase (OGT) and O-GlcNAcase (OGA), which catalyze the addition and hydrolysis of O-GlcNAc, respectively. It has previously been shown that baseline heart physiology and pathophysiology are impacted by sex. Here, we hypothesized that sex differences in molecular signaling may target protein O-GlcNAcylation both basally and in ischemic hearts. To address this question, we subjected male and female WT murine hearts to ex vivo ischemia or I/R injury. We assessed hearts for protein O-GlcNAcylation, abundance of OGT, OGA, and glutamine:fructose-6-phosphate aminotransferase (GFAT2), activity of OGT and OGA, and UDP-GlcNAc levels. Our data demonstrate elevated O-GlcNAcylation in female hearts both basally and during ischemia. We show that OGT activity was enhanced in female hearts in all treatments, suggesting a mechanism for these observations. Furthermore, we found that ischemia led to reduced O-GlcNAcylation and OGT-specific activity. Our findings provide a foundation for understanding molecular mechanisms that regulate O-GlcNAcylation in the heart and highlight the importance of sex as a significant factor when assessing key regulatory events that control O-GlcNAc cycling. These data suggest the intriguing possibility that elevated O-GlcNAcylation in females contributes to reduced ischemic susceptibility.

Keywords: O-GlcNAc; O-GlcNAc transferase; O-GlcNAcase; cardiac; glycosylation; hexosamine biosynthetic pathway; sex differences.

Figure 1

O-GlcNAc levels are enhanced in female hearts compared with age-matched male hearts.A, total tissue lysates of hearts from C57B6/J WT mice (14–18 weeks, N = 5 per sex) were generated using 9 M urea. Proteins were separated by SDS-PAGE (15 or 22.5 μg), electroblotted to nitrocellulose, and the following were detected and quantified: O-GlcNAc (RL2), O-GlcNAc specificity control (RL2 with 500 mM GlcNAc), O-GlcNAc transferase (OGT), O-GlcNAcase (OGA), and glutamine fructose-6-phosphate amidotransferase 2 (GFAT2). Protein load was assessed by total protein membrane stain (SYPRO Ruby). The following signals were quantified and normalized to total protein (B) O-GlcNAc (RL2); (C) OGT; (D) OGA; and (E) GFAT2. Statistical test: unpaired Student’s t test. ∗p < 0.05. B–E, data are represented as mean ± SD. ns, nonspecific band; O-GlcNAc, O-linked β-GlcNAc.

Figure 2

Female hearts begin ischemic contracture and reach hypercontracture earlier than age-matched male hearts. Hearts from C57B6/J WT mice (15–21 weeks, N = 6–12 per sex per treatment) were treated ex vivo on the Langendorff apparatus. A, perfusion protocol: control—40 min perfusion; ischemia—20 min perfusion, 20 min—no-flow ischemia; ischemia–reperfusion (I/R) injury—20 min perfusion, 20 min ischemia, 120 min reperfusion. B, representative I/R left ventricular developed pressure (LVDP) trace obtained during perfusion. Critical time points are highlighted. C, summary of preischemic and postischemic contractile parameters. Heart rate (HR), LVDP, and rate pressure product (RPP) were recorded for all replicates and represented as mean ± SD. D, comparison of time in contracture (•), time in hypercontracture (), and the duration of contracture () between males and females. Statistical test: two-way ANOVA. ∗∗∗p < 0.001. Data are represented as mean ± SD.

Figure 3

O-GlcNAc levelscycle in response to I/R injury with little impact on OGT and OGA abundance.A, hearts from C57B6/J WT mice (15–21 weeks, N = 6–12 per sex per treatment) were perfused ex vivo on the Langendorff apparatus (control [•]; ischemia 20 min []; I/R 20 min ischemia/120 min reperfusion []). Hearts were extracted in 1% N-P40 buffer, and proteins were separated by SDS-PAGE (15 or 22.5 μg), electroblotted to nitrocellulose, and the following were detected: OGT, OGA, GFAT2, heat shock cognate 70 (Hsc70), O-GlcNAc (RL2), and O-GlcNAc specificity control (RL2 with 500 mM GlcNAc, not shown). Hsc70 was used as a representative immunoblot for protein load. Protein load was assessed by total protein membrane stain (SYPRO Ruby or Direct Blue-71, Fig. S2). The following were quantified and normalized to total protein. (B) O-GlcNAc (RL2); (C) OGT; (D) OGA; and (E) GFAT2. B–E, pooled male and female data are reported on the left. Statistical test: one-way ANOVA. ∗p < 0.05. Data parsed by sex are reported on the right. Statistical test: two-way ANOVA. ∗∗p < 0.01; ∗∗∗∗p < 0.0001. Data are represented as mean ± SD. GFAT2, glutamine-fructose 6 phosphate amidotransferase 2; IR, I/R, ischemia–reperfusion; OGA, O-GlcNAcase; O-GlcNAc, O-linked β-GlcNAc; OGT, O-GlcNAc transferase.

Figure 4

OGT activity is elevated in female hearts. Hearts from C57B6/J WT mice (15–21 weeks, N = 6–12 per sex per treatment) were perfused ex vivo on the Langendorff apparatus (control [•]; ischemia 20 min []; I/R 20 min ischemia/120 min reperfusion []). Hearts were extracted in 1% NP-40, and equal protein was desalted using a Zebaspin column. A, OGT activity was assessed by measuring transfer of ³H-GlcNAc to the casein kinase II (CK2) acceptor peptide. B, OGA activity was detected using the pseudosubstrate 4-methylumbelliferyl-β-GlcNAc. A and B, pooled male and female data are reported on the left. Statistical test: one-way ANOVA. ∗∗p < 0.01; ∗∗∗∗p < 0.0001. Data parsed by sex are reported on the right. Statistical test: two-way ANOVA. ∗p < 0.05; ∗∗p < 0.01. Data are represented as mean ± SD. C, correlation of OGT activity and O-GlcNAc (RL2). D, correlation of OGA activity and O-GlcNAc (RL2). C and D, r = Pearson correlation coefficient. I/R, ischemia–reperfusion; OGA, O-GlcNAcase; OGT, O-GlcNAc transferase.

Figure 5

UDP-GlcNAc levels do not change during injury in both male and female hearts. Hearts from C57B6/J WT mice (15–21 weeks, N = 5–6 per sex per treatment) were perfused ex vivo on the Langendorff apparatus (control [•]; ischemia 20 min []; I/R 20 min ischemia/120 min reperfusion []). Heart nucleotides and nucleotide sugars were extracted in methanol:chloroform, desalted using solid-phase extraction, and analyzed using high-performance anion-exchange chromatography. Nucleotide/nucleotide sugar levels were normalized to total protein: (A) UDP-GlcNAc; (B) ATP; (C) ADP; (D) AMP. A–D, pooled male and female data are reported on the left. Statistical test: one-way ANOVA. ∗p < 0.05; ∗∗p < 0.01; and ∗∗∗∗p < 0.0001. Data parsed by sex are reported on the right. Statistical test: two-way ANOVA. ∗p < 0.05; ∗∗p < 0.01. Data are represented as mean ± SD. E, representative separation of nucleotide sugars. Peaks are numbered as follows: (1) CMP, CMP-neuraminic acid; (2) AMP; (3) UMP; (4) UDP-GlcNAc; (5) UDP-GalNAc; (6) UDP-glucose, UDP-galactose, UDP-xylose, CDP, GMP; (7) ADP; (8) UDP, GDP-mannose, GDP-fucose, CTP, UTP; (9) ATP, GDP; and (10) UDP-glucuronic acid. The sodium acetate gradient used to achieve nucleotide separation is depicted on the right Y-axis. F, correlation of UDP-GlcNAc and O-GlcNAc (RL2). G, correlation of AMP:ATP ratio and OGT activity. F and G, r = Pearson's correlation coefficient. I/R, ischemia–reperfusion.