NADPH Oxidase 4: Crucial for Endothelial Function under Hypoxia-Complementing Prostacyclin

Abstract

Aim: The primary endothelial NADPH oxidase isoform 4 (NOX4) is notably induced during hypoxia, with emerging evidence suggesting its vasoprotective role through H₂O₂ production. Therefore, we aimed to elucidate NOX4's significance in endothelial function under hypoxia. Methods: Human vessels, in addition to murine vessels from Nox4^-/- mice, were explored. On a functional level, Mulvany myograph experiments were performed. To obtain mechanistical insights, human endothelial cells were cultured under hypoxia with inhibitors of hypoxia-inducible factors. Additionally, endothelial cells were cultured under combined hypoxia and laminar shear stress conditions. Results: In human occluded vessels, NOX4 expression strongly correlated with prostaglandin I2 synthase (PTGIS). Hypoxia significantly elevated NOX4 and PTGIS expression and activity in human endothelial cells. Inhibition of prolyl hydroxylase domain (PHD) enzymes, which stabilize hypoxia-inducible factors (HIFs), increased NOX4 and PTGIS expression even under normoxic conditions. NOX4 mRNA expression was reduced by HIF1a inhibition, while PTGIS mRNA expression was only affected by the inhibition of HIF2a under hypoxia. Endothelial function assessments revealed hypoxia-induced endothelial dysfunction in mesenteric arteries from wild-type mice. Mesenteric arteries from Nox4^-/- mice exhibited an altered endothelial function under hypoxia, most prominent in the presence of cyclooxygenase inhibitor diclofenac to exclude the impact of prostacyclin. Restored protective laminar shear stress, as it might occur after thrombolysis, angioplasty, or stenting, attenuated the hypoxic response in endothelial cells, reducing HIF1a expression and its target NOX4 while enhancing eNOS expression. Conclusions: Hypoxia strongly induces NOX4 and PTGIS, with a close correlation between both factors in occluded, hypoxic human vessels. This relationship ensured endothelium-dependent vasodilation under hypoxic conditions. Protective laminar blood flow restores eNOS expression and mitigates the hypoxic response on NOX4 and PTGIS.

Keywords: NADPH oxidase 4; PTGIS; endothelial function; human endothelial cells; hypoxia; laminar flow.

Figure 1

NOX4 highly correlates with PTGIS in human occluded vessels. (A) Correlation of NOX4 mRNA expression and prostaglandin I2 synthase (PTGIS) in femoral artery walls from patients with peripheral arterial disease (n = 11). POLRIIa mRNA expression was used as reference. (B) Correlation of NOX4 mRNA expression and prostaglandin I2 synthase (PTGIS) in human non-occluded left internal mammary arteries (n = 30). POLRIIa mRNA expression was used as reference. Statistics: Spearman’s correlation coefficient (r_S) between PTGIS and NOX4 expression. Abbreviations: NOX4, NADPH oxidase 4; PTGIS, prostaglandin I2 synthase; POLRIIa, RNA polymerase II subunit A.

Figure 2

Increased NOX4 and PTGIS expression after hypoxia in human endothelial cells and murine aorta. (A) NOX4 mRNA expression after 8 h, 16 h, and 24 h normoxia (21% oxygen) or hypoxia (1% oxygen) (n = 3). TBP mRNA expression was used as reference. (B) Hydrogen peroxide production in HUVECs subjected to transduction with lentiviral particles containing shNOX4 and exposed to 24 h hypoxia (1% oxygen) (n ≥ 5). Untransduced cells (untr) and cells transduced with scrambled shRNA (shC) were used as controls. (C) Relative Nox4 mRNA expression in aorta from WT and Nox4^−/− mice after 24 h hypoxia (1% oxygen) (n = 6). B2m mRNA expression was used as reference. (D) PTGIS mRNA expression after 8 h, 16 h, and 24 h normoxia (21% oxygen) or hypoxia (1% oxygen) (n = 4). TBP mRNA expression was used as reference. (E) 6-keto-PGF1 alpha concentration in supernatant of HUVECs exposed to 24 h hypoxia (1% oxygen) (n ≥ 5). (F) Relative Ptgis mRNA expression in aorta from WT and Nox4^−/− mice after 24 h hypoxia (1% oxygen) (n = 6). B2m mRNA expression was used as reference. (G) Relative NOX4 mRNA expression in HUVECs treated subsequently with actinomycin D (ActD) for 8 h under normoxia and hypoxia (1% oxygen) (n = 5). POLRIIa mRNA expression was used as reference. Mean relative NOX4 mRNA expression in HUVECs treated subsequently with actinomycin D (ActD) for 8 h under normoxia and hypoxia (1% oxygen) (n = 5). POLRIIa mRNA expression was used as reference. (H) Relative PTGIS mRNA expression in HUVECs treated subsequently with actinomycin D (ActD) for 8 h under normoxia and hypoxia (1% oxygen) (n = 5). POLRIIa mRNA expression was used as reference. Mean relative PTGIS mRNA expression in HUVECs treated subsequently with actinomycin D (ActD) for 8 h under normoxia and hypoxia (1% oxygen) (n = 5). POLRIIa mRNA expression was used as reference. Statistics: Normal (Gaussian) distribution was tested by Shapiro–Wilk normality tests. All data had Gaussian distributions. Data were then analyzed with t-tests or one-way ANOVAs followed by Tukey’s multiple comparisons tests or two-way ANOVA and Sidak’s multiple comparisons tests, respectively. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. Abbreviations: ActD, actinomycin D; B2m, beta-2 microglobulin; crtl, control; HUVECs, human umbilical vein endothelial cells; NOX4, NADPH oxidase 4; PTGIS, prostaglandin I2 synthase; POLRIIa, RNA polymerase II subunit A; shRNA, short hairpin RNA; TBP, TATA-box binding protein; untr, untreated; WT, wild-type.

Figure 3

NOX4 and PTGIS were predominantly regulated by different HIFs. (A) Relative NOX4 mRNA expression in HUVECs after 24 h 10 mM DMOG treatment under normoxic conditions (n ≥ 7). TBP mRNA expression was used as reference. (B) Relative NOX4 mRNA expression in HUVECs treated subsequently with HIF1a inhibitor for 16 h under normoxia and hypoxia (1% oxygen) (n = 7). POLRIIa mRNA expression was used as reference. (C) Relative NOX4 mRNA expression in HUVECs treated subsequently with HIF2a inhibitor for 16 h under normoxia and hypoxia (1% oxygen) (n = 5). POLRIIa mRNA expression was used as reference. (D) Relative PTGIS mRNA expression in HUVECs after 24 h 10 mM DMOG treatment under normoxic conditions (n ≥ 7). TBP mRNA expression was used as reference. (E) Relative PTGIS mRNA expression in HUVECs treated subsequently with HIF1a inhibitor for 16 h under normoxia and hypoxia (1% oxygen) (n = 8). POLRIIa mRNA expression was used as reference. (F) Relative PTGIS mRNA expression in HUVECs treated subsequently with HIF2a inhibitor for 16 h under normoxia and hypoxia (1% oxygen) (n = 6). POLRIIa mRNA expression was used as reference. Statistics: Normal (Gaussian) distribution was tested by Shapiro–Wilk normality tests. Non-Gaussian distributed data were analyzed by Kruskal–Wallis and Dunn’s multiple comparison tests (D,E). Gaussian distributed data (A,B,C,F) were analyzed with one-way ANOVA followed by Tukey’s multiple comparisons tests or two-way ANOVA and Sidak’s multiple comparisons tests, respectively. * p < 0.05; ** p < 0.01; **** p < 0.0001. Abbreviations: crtl, control; DMOG, dimethyloxallyl glycine; HIF, hypoxia-inducible factor; HUVECs, human umbilical vein endothelial cells; NOX4, NADPH oxidase 4; PTGIS, prostaglandin I2 synthase; POLRIIa, RNA polymerase II subunit A; TBP, TATA-box binding protein.

Figure 4

H₂O₂ derived from Nox4 maintained vascular response of mesenteric arteries under hypoxia. (A) Endothelium-dependent vasorelaxation of mesenteric arteries from WT mice under normoxic and hypoxic conditions ex vivo (n ≥ 11). (B) Hydrogen peroxide release in WT and Nox4^−/− mice mesenteric arteries ex vivo after hypoxia (n = 6). (C) Endothelium-dependent vasorelaxation of mesenteric arteries from WT mice and catalase treatment under hypoxic conditions ex vivo (n ≥ 6). (D) Endothelium-dependent vasorelaxation of mesenteric arteries from WT mice and Nox4^−/− mice under hypoxic conditions ex vivo (n ≥ 11). (E) Endothelium-dependent vasorelaxation of mesenteric arteries from WT mice and diclofenac treatment under hypoxic conditions ex vivo (n ≥ 6). (F) Endothelium-dependent vasorelaxation of mesenteric arteries from WT mice with diclofenac and catalase treatment under hypoxic conditions ex vivo (n ≥ 5). (G) Endothelium-dependent vasorelaxation of mesenteric arteries from WT mice and Nox4^−/− mice under hypoxic conditions and diclofenac treatment ex vivo (n ≥ 6). (H) Endothelium-dependent vasorelaxation of mesenteric arteries from WT mice and Nox4^−/− mice under hypoxic conditions and L-NAME treatment ex vivo (n ≥ 8). (I) Relative eNOS mRNA expression in HUVEC treated with hypoxia and L-NAME or L-NAME and catalase combination (n ≥ 5). POLRIIa mRNA expression was used as reference. Statistics: Normal (Gaussian) distribution was tested by Shapiro–Wilk normality tests. All data has Gaussian distribution. Data was then analyzed with t-tests or one-way ANOVAs, followed by Tukey’s multiple comparisons tests or two-way ANOVA and Sidak’s multiple comparisons tests, respectively. * p < 0.05; ** p < 0.01; *** p < 0.001. Abbreviations: Ach, acetylcholine; eNOS, endothelial nitric oxide synthase; L-NAME, N^G-Nitro-L-arginine methyl ester; NOX4, NADPH oxidase 4; PE, phenylephrine; WT, wild-type.

Figure 5

Laminar flow of 30 dyn/cm² on human endothelial cells prevented upregulation of NOX4 and PTGIS under hypoxia. (A) Relative NOX4 mRNA expression in endothelial cells under laminar 30 dyn/cm² flow after 24 h and (B) 48 h normoxic or hypoxic conditions (n = 8). TBP mRNA expression was used as reference. (C) Relative PTGIS mRNA expression in endothelial cells under laminar 30 dyn/cm² flow after 24 h and (D) 48 h normoxic or hypoxic conditions (n = 8). TBP mRNA expression was used as reference. (E) Relative eNOS mRNA expression in endothelial cells under laminar 30 dyn/cm² flow after 24 h and (F) 48 h normoxic or hypoxic conditions (n = 8). TBP mRNA expression was used as reference. (G) Nitrite concentration in supernatant of endothelial cells under laminar 30 dyn/cm² flow after 24 h normoxic or hypoxic conditions (n = 8). (H) Relative eNOS protein expression of endothelial cells under static and laminar 30 dyn/cm² flow after 24 h normoxic or hypoxic conditions (n = 6). (I) Relative HIF1a protein expression in endothelial cells under static and laminar 30 dyn/cm² flow after 24 h hypoxic conditions (n = 5). Statistics: Normal (Gaussian) distribution was tested by Shapiro–Wilk normality tests. All data had Gaussian distributions. Data were then analyzed with two-way ANOVAs and Sidak’s multiple comparisons tests, respectively. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. Abbreviations: eNOS, endothelial nitric oxide synthase; HIF1a, hypoxia-inducible factor 1a; NOX4, NADPH oxidase 4; PTGIS, prostaglandin I2 synthase; TBP, TATA-box binding protein.