Ceramide changes the mediator of flow-induced vasodilation from nitric oxide to hydrogen peroxide in the human microcirculation

Abstract

Rationale: Mitochondrial-derived hydrogen peroxide (H2O2) regulates flow-induced dilation (FID) in microvessels from patients with coronary artery disease. The relationship between ceramide, an independent risk factor for coronary artery disease and a known inducer of mitochondrial reactive oxygen species, and FID is unknown.

Objective: We examined the hypothesis that exogenous ceramide induces a switch in the mediator of FID from nitric oxide to H2O2.

Methods and results: Internal diameter changes of resistance arterioles from human adipose and atrial tissue were measured by video microscopy. Mitochondrial H2O2 production was assayed in arterioles using mito peroxy yellow 1. Polyethylene glycol-catalase, rotenone, and Mito-TEMPO impaired FID in healthy adipose arterioles pretreated with ceramide, whereas N(ω)-nitro-l-arginine methyl ester had no effect. Mitochondrial H2O2 production was induced in response to flow in healthy adipose vessels pretreated with ceramide, and this was abolished in the presence of polyethylene glycol-catalase. Immunohistochemistry demonstrated ceramide accumulation in arterioles from both healthy patients and patients with coronary artery disease. N(ω)-nitro-l-arginine methyl ester reduced vasodilation to flow in adipose as well as atrial vessels from patients with coronary artery disease incubated with GW4869, a neutral sphingomyelinase inhibitor, whereas polyethylene glycol-catalase had no effect.

Conclusions: Our data indicate that ceramide has an integral role in the transition of the mediator of FID from nitric oxide to mitochondrial-derived H2O2 and that inhibition of ceramide production can revert the mechanism of dilation back to nitric oxide. Ceramide may be an important target for preventing and treating vascular dysfunction associated with atherosclerosis.

Keywords: ceramides; mitochondria; nitric oxide; reactive oxygen species; sphingomyelin phosphodiesterase.

Figure 1. Effect of exogenous ceramide on FID

A, The magnitude of dilation is not affected with overnight incubation of ceramide alone compared to vehicle-treated control (n= 10 and 10, respectively). B, FID in healthy adipose arterioles is reduced in the presence of L-NAME however is maintained if first pre-incubated with ceramide (n= 7 and 5, respectively). C, PEG-catalase has minimal effect on FID in healthy arterioles, however impairs FID in ceramide-treated arterioles (n= 5 and 5, respectively). D, SNP-induced dilation is not reduced by ceramide compared to vehicle-treated control (n=4 and 6, respectively). *p<0.05 vs. vehicle at specific pressure gradients. n indicates number of patients.

Figure 2. Effect of ceramide on production of mitochondrial H₂O₂ in response to flow

A, Representative images of cannulated vessels incubated in vehicle or ceramide overnight +/-PEG-catalase, following intraluminal incubation with MitoPY1 at baseline, 1 min, and 5 min of max intraluminal flow. B, Quantification of mitochondrial H₂O₂ expressed as percent change in MitoPY1 fluorescence from baseline at 1 min and 5 min of max flow. * p<0.05 vs. vehicle, † p<0.05 vs. ceramide+catalase.

Figure 2. Effect of ceramide on production of mitochondrial H₂O₂ in response to flow

A, Representative images of cannulated vessels incubated in vehicle or ceramide overnight +/-PEG-catalase, following intraluminal incubation with MitoPY1 at baseline, 1 min, and 5 min of max intraluminal flow. B, Quantification of mitochondrial H₂O₂ expressed as percent change in MitoPY1 fluorescence from baseline at 1 min and 5 min of max flow. * p<0.05 vs. vehicle, † p<0.05 vs. ceramide+catalase.

Figure 3. Role of mitochondria in FID in ceramide exposed vessels

A, Vasodilation in response to flow is impaired in arterioles pre-treated with ceramide and rotenone compared to ceramide alone. Rotenone alone had no effect on FID (n=5, 10, and 5, respectively). B, Mito-TEMPO significantly decreased the response to flow in ceramide-treated vessels compared to ceramide alone. Mito-TEMPO alone had no effect (n=5, 10, and 5, respectively). † p<0.05 vs. ceramide curve. *p<0.05 vs. ceramide at specific pressure gradients.

Figure 4. Ceramide accumulation in arterioles from healthy versus CAD patients

Representative images from 3 patients (3 healthy, 3 CAD). The total area of staining is decreased in arterioles from healthy patients (A) versus patients with CAD (B), however area stained/total area did not differ between groups. Specificity of the antibody was examined by removal of the primary antibody (C). Bar=40 μm.

Figure 5. Inhibition of NSmase reverts the mediator of FID back to NO in vessels from adipose tissue from patients with CAD

A, Incubation with the specific NSmase inhibitor GW4869 (4μM, 16-20hrs) did not affect the overall magnitude of dilation to flow compared to vehicle-treated control (n=9 and 8, respectively). B, The response to flow was inhibited in CAD vessels first incubated with GW4869 in the presence of L-NAME (100μM) compared to GW4869 alone, whereas PEG-catalase (500U) had no effect (n=5, 9, and 5, respectively). *p<0.05 vs. GW4869 at specific pressure gradients.

Figure 6. Inhibition of NSmase reverts the mediator of FID back to NO in atrial vessels from patients with CAD

Incubation with the specific NSmase inhibitor GW4869 (4μM, 16-20hrs) had no effect on FID compared to vehicle (n=4). B, The response to flow was inhibited in CAD vessels first incubated with GW4869 in the presence of L-NAME (100μM), whereas PEG-catalase (500U) had no effect (n=4). *p<0.05 vs. GW4869 at specific pressure gradients.

Figure 7. A schematic diagram illustrating the involved pathways

Under normal conditions in healthy adults, exposure of the endothelial layer to shear stress activates endothelial nitric oxide synthase (eNOS) causing elevation of nitric oxide (NO) which primarily serves as the mediator of smooth muscle dilation. In the disease state, increased levels of ceramide formed via neutral sphingomyelinase (NSmase) trigger mitochondrial reactive oxygen species (ROS) formation which both decrease the bioavailability of NO and ultimately change the mediator of vasodilation to hydrogen peroxide (H₂O₂). Inhibition of NSmase with the specific non-competitive inhibitor GW4869 can revert the mechanism of dilation back to NO.