Exercise-Induced SOD3/H₂O₂ Sulfenylates PKG1α to Restore Vasodilation in Diabetes

Endothelial dysfunction, a hallmark of type 2 diabetes (T2D), is caused by reduced nitric oxide (NO) bioavailability. Exercise restores endothelial function in T2D by enhancing resilience¹. Under physiological conditions, exercise activates endothelial nitric oxide synthase (eNOS), promoting NO-dependent vasodilation. However, in T2D, where eNOS uncoupling and excess superoxide (O₂^•−) contribute to oxidative stress, exercise may exert vascular benefits through alternative pathways. One such mechanism involves the exercise-induced upregulation of antioxidant defenses, including superoxide dismutases (SODs), which convert O₂^•− to hydrogen peroxide (H₂O₂). This process may enable H₂O₂-mediated vasodilation, either through NO-dependent or NO-independent pathways. These adaptations enhance endothelial resilience, restore vascular homeostasis and mitigate oxidative stress-induced endothelial dysfunction. However, the precise mechanisms underlying exercise-induced H₂O₂-mediated vasodilation in T2D remain poorly understood.

Extracellular SOD (ecSOD, SOD3) is a major secretory copper (Cu)-containing antioxidant enzyme in the vasculature, whose activity requires the Cu transporter ATP7A. SOD3 is secreted by vascular smooth muscle cells and anchors at the endothelial surface². In T2D arteries, ATP7A expression is downregulated, leading to reduced SOD3 function, increased O₂^•−levels, and impaired NO-depending vasodilation³. Evidence suggests that H₂O₂, rather than NO, can directly activate cGMP-dependent protein kinase (PKG) through Cys42 oxidation⁴. Here we show that exercise in T2D promotes PKG1α-CysOH formation at Cys42, thereby enhancing vasodilation. Mechanistically, exercise restores the ATP7A-SOD3 axis, facilitating SOD3-mediated out-side in H₂O₂ signaling to improve vascular function (Figure F).

Figure. Exercise-induced SOD3-Derived H₂O₂ Restores Endothelial Function in Type2 Diabetes via Sulfenylation of PKG1α at Cys42.

A. Schematic of the experimental protocol. B. Exercise restores endothelial function via SOD3-derived H₂O₂, but not NO-dependent manner in T2D mice. Mesenteric resistance arteries were pretreated (30 min) with vehicle, catalase (1000 U/ml) or L-NAME (100 μM) and evaluated acetylcholine (ACh)-induced vasodilation in WT (B1), T2D (B2), SOD3KO/T2D mice (B3) with exercise (Exe) or sedentary (Sed) using wire myograph. H₂O₂ (B4, Control (n=5), T2D (n=6), T2D/SOD3 KO (n=4)) and (B5) cGMP levels (n=4) in mesenteric arteries. C. Exercise restores decreased SOD3 activity in T2D vessels via increasing expression of ATP7A, a Cu transporter for SOD3. SOD (C1) and ATP7A (C2) protein expression, and activity of SOD3 (C3) and SOD1 (C4) was measured in mesenteric arteries from the control, T2D, or Cu transport-dysfunctional ATP7A mutant (ATP7A^mut)/T2D mice with Exe or Sed (n=3). D. Exercise induces PKG1α-CysOH formation in T2D vessels via SOD3-dependent manner, which restores endothelial function. DCP-Bio1 assay to detect PKGIα-CysOH (D1), DCP-Bio1-labeled lysates from vessels of WT, SOD3 KO (D3, n=3), PKG1α^C42SKI (D5, n=4) mice with or without T2D and Exe were used to measure PKGIα-CysOH formation (D2–D5). PKG activity was assessed by pVASP/VASP (D2, n=4). ACh-induced vasodilation in WT, T2D, PKG1α^C42SKI/T2D mice with Exe or Sed (D6). E. ATP7A is required for exercise-induced PKG1α Cys oxidation and restoring endothelial function in T2D. PKG1α-CysOH formation was measured in T2D or T2D/ATP7A^mut (E2, n=4) or T2D/ATP7A-Tg or T2D/ATP7A-Tg/PKG1α^C42SKI (E3, n=3) mice with Exe or Sed (E1–E3). ACh-induced vasorelaxation in T2D or T2D/ATP7A-Tg or T2D/ATP7A-Tg/PKG1α^C42SKI mice with Sed (E4). F. Proposed model: Exercise-induced ATP7A-mediated SOD3-derived out-side in H₂O₂ signaling restore endothelial function in T2D mice via PKG1α sulfenylation. All data were expressed as mean ± SEM. Normality was assessed using the Shapiro-Wilk test. Statistical significance was determined two-way ANOVA (Figures B1–B3, D6, E4) or one-way ANOVA (all other figures) followed by post-hoc Turkey test using Graphpad Prism 10. NS: no significant.

To assess the beneficial effects of exercise on impaired vasodilation in T2D, we used male 6 weeks old male C57BL/6 background mice because female is resistance to induce diabetes. T2D was induced as reported³ with voluntary wheel exercise training (Figure A). After 14 days, the running distance was similar between control and T2D mice. This exercise intervention did not affect metabolic profile in control or T2D mice. Wire myograph showed that exercise restored endothelium-dependent relaxation (EDR) in mesenteric resistance arteries (≈230 μm) of T2D mice (Figure B2). In control mice, EDR in arteries from exercised mice was abolished by the eNOS inhibitor L-NAME but only partially inhibited by catalase (Figure B1), consistent with increased cGMP levels (Figure B5), indicating a predominantly NO-dependent mechanism. In T2D mice, exercise-enhanced EDR was abolished by catalase but unaffected by L-NAME (Figure B2), correlating with increased H₂O₂ levels without altering cGMP (Figure B4–B5). Endothelium-independent vasorelaxation induced by sodium nitroprusside remained unchanged (data not shown). These findings suggest that in T2D, exercise restores EDR through an H₂O₂-dependent mechanism, while in control mice, EDR is primarily mediated by a NO dependent pathway (Figure F).

To identify the source of exercise-induced H₂O₂ production, we examined the role of SOD in vascular adaptations to exercise in T2D mice. Under sedentary conditions, SOD3 activity, but not protein levels, was significantly reduced in T2D mesenteric arteries, whereas SOD1 activity and protein levels remained unchanged (Figures C1–C3–C4). Exercise restored the SOD3 activity without affecting SOD1 (Figure C3–C4), correlating with increased extracellular H₂O₂ (Figure B4) and improved EDR in T2D mesenteric arteries (Figure B2). However, these effects were abolished in SOD3^−/−/T2D mice (Figures B3). Notably, exercise restored ATP7A levels in T2D mice (Figure C2), while Cu transport dysfunctional ATP7A mutant (ATP7A^mut) mice showed no exercise-induced SOD3 activation in diabetic arteries without affecting SOD1 activation (Figures C3–C4). Thus, exercise restores SOD3 activity by restoring ATP7A, enhancing EDR through SOD3-derived H₂O₂ in T2D. Importantly, Nox4 knockout mice exhibited no effect on exercise-induced restoration of EDR in T2D, indicating that Nox4 is not involved in this process (data not shown).

To elucidate the mechanism underlying exercise-induced EDR in T2D mice, we measured cGMP levels in mesenteric arteries. While exercise increased cGMP levels in control mice, no such increase was observed in T2D mice (Figure B5). To further assess the NO-cGMP-PKG1α pathway, we examined the phosphorylation of vasodilator-stimulated phosphoprotein (p-VASP) at Ser239. Despite the lack of cGMP induction, exercise increased p-VASP levels in T2D arteries (Figure D2), suggesting that PKG1α activation occurs through an alternative mechanism. To explore this, we assessed PKG1α cysteine oxidation using the CysOH trapping probe, DCP-Bio1⁵ (Figure D1). Exercise enhanced PKG1α-CysOH formation in T2D arteries, but this effect was abolished in SOD3^−/− mice (Figure D2), indicating that SOD3-derived H₂O₂ mediates exercise-induced EDR via PKG1α-CysOH formation. To confirm the functional relevance of PKG1α-CysOH in EDR, we used “redox-dead” Cys42Ser PKG1α knock-in (KI) mice, which cannot form CysOH in response to H₂O₂⁴. In diabetic PKG1α KI mice, both exercise-induced PKG1α-CysOH formation (Figures D4–D5) and EDR (Figure D6) were significantly reduced, supporting the conclusion that aerobic exercise promotes PKG1α-CysOH formation via a SOD3-H₂O₂-dependent mechanism. Notably, while resistance training has been shown to improve endothelial dysfunction in T2D, it remains unclear whether this benefit involves H₂O₂-mediated PKGIα oxidation. This point warrants further investigation in future studies.

To determine the role of the ATP7A-SOD3 axis in exercise-induced EDR in T2D, we examined PKG1α oxidation in ATP7A^mut mice. These mice exhibited impaired exercise-induced PKG1α-CysOH formation (Figures E1–E2) in T2D arteries, indicating that exercise restores ATP7A expression, which enhances the SOD3-PKG1α CysOH-EDR axis. To further explore the functional link between ATP7A and PKG1α, we utilized ATP7A-overexpressing transgenic (ATP7A-Tg) mice crossed with PKG1α KI mice. In diabetic ATP7A-Tg mice, both PKG1α-CysOH formation and EDR were significantly improved compared to diabetic controls; however, these beneficial effects were absent in ATP7A-Tg/PKG1α KI mice (Figures E3–E4). These results indicate that ATP7A upregulation induced by exercise in T2D is essential for enhancing the SOD3-H₂O₂-PKG1α CysOH axis that drives EDR. Future studies will aim to determine: (1) whether this pathway is functional in exercise-trained patients with T2D, (2) whether sex differences influence this mechanism, (3) the reversibility and time course of exercise-induced PKG1α oxidation, and (4) the molecular mechanisms underlying exercise-induced ATP7A protein expression.

In summary, we demonstrate that exercise restores the ATP7A-SOD3 pathway, facilitating outside-in H₂O₂ signaling that promotes NO-independent EDR in T2D mice via PKG1α oxidation (Figure F). These findings uncover a mechanism by which exercise-mediated redox regulation of PKG1α through the ATP7A-SOD3-H₂O₂ axis enhances endothelial resilience and improves vascular function in diabetes.

Nonstandard Abbreviations and Acronyms

T2D

Type 2 diabetes

NO

Nitric oxide

O2^•−

Superoxide

H₂O₂

Hydrogen peroxide

eNOS

Endothelial nitric oxide synthase

SOD

Superoxide dismutase

PKG

cGMP-dependent protein kinase

EDR

Endothelium dependent relaxation

L-NAME

L-NG-Nitro arginine methyl ester

Nox4

NADPH oxidase 4

VASP

Vasodilator-stimulated phosphoprotein

Data Availability

Data and materials supporting the results or analyses presented in this publication are available on reasonable request to the corresponding authors.