NO-ferroheme is a signaling entity in the vasculature

Abstract

Despite wide appreciation of the biological role of nitric oxide (NO) synthase (NOS) signaling, questions remain about the chemical nature of NOS-derived bioactivity. Here we show that NO-like bioactivity can be efficiently transduced by mobile NO-ferroheme species, which can transfer between proteins, partition into a hydrophobic phase and directly activate the sGC-cGMP-PKG pathway without intermediacy of free NO. The NO-ferroheme species (with or without a protein carrier) efficiently relax isolated blood vessels and induce hypotension in rodents, which is greatly potentiated after the blockade of NOS activity. While free NO-induced relaxations are abolished by an NO scavenger and in the presence of red blood cells or blood plasma, a model compound, NO-ferroheme-myoglobin preserves its vasoactivity suggesting the physiological relevance of NO-ferroheme species. We conclude that NO-ferroheme behaves as a signaling entity in the vasculature.

Fig. 1. Different NO-ferroheme preparations potently relax blood vessels.

a–d, EPR spectra (X band; 77K) of different NO-ferroheme preparations together with ex vivo vasorelaxation responses of aortic rings to increasing concentrations of (NO-ferroheme)4-Hb (n = 14) (a), NO-ferroheme-Mb (n = 31) (b), NO-ferroheme-BSA (n = 4) (c) and NO-ferroheme-l-Cys (n = 27) (d), using the myograph system in the presence of 0.1 mM L-NAME. n represents the number of mouse aortic rings used in different myograph chambers. Vasorelaxation responses are shown as percent of PE-induced plateau, and the data are presented as mean ± s.e.m. Half maximal EC50 values were calculated using least squares nonlinear regression analysis and are presented as log and absolute values with a 95% CI. CI, confidence interval; EC50, effective concentration; PE, phenylephrine. Source data

Fig. 2. NO-ferroheme vasorelaxation is mediated by the sGC/PKG pathway.

Ex vivo vasorelaxation responses of mouse aortic rings to increasing concentrations of NO-ferroheme-Mb. a,b, The responses to NO-ferroheme-Mb were abolished by simultaneous treatment with ODQ (10 μM), a cell-permeable and potent inhibitor of sGC (placebo, n = 4; ODQ, n = 4) (a), but sensitized by the PDE5 inhibitor, sildenafil (1 μM) placebo, n = 4; sildenafil, n = 4) (b). c, Simultaneous addition of X (0.37 mM) and XO (5 mU ml⁻¹), to induce superoxide production and alter the redox status, attenuated the response to NO-ferroheme-Mb (placebo, n = 3; X/XO, n = 2). d, In isolated rat aortic segments treated with the NOS inhibitor, L-NAME (300 µM), co-incubation with NO-ferroheme-Mb (0.5 µM) significantly increased pVASP, thus indicating the activation of the sGC/PKG pathway. Data in a–c were analyzed by two-way repeated measures ANOVA followed by Šídák’s multiple comparisons test. Vasorelaxation responses are shown as percent of PE-induced plateau, and the data are presented as mean ± s.e.m. n in a–c represents the number of mouse aortic rings used in different myograph chambers. Data in d were analyzed by nonparametric Kruskal–Wallis test followed by Dunn’s multiple comparisons test and presented as mean ± s.d. Dots in d represent different aortic segments from three different rounds of western blot analysis. In a–d, *P ≤ 0.05, **P ≤ 0.01 and ****P ≤ 0.0001. Source data

Fig. 3. The bioactivity of NO-ferroheme is unrelated to the release of NO.

a–f, Ex vivo vasorelaxation responses of mouse aortic rings to increasing concentrations of NO-ferroheme-Mb (a–c) and NO (d–f), in the absence or presence of the NO scavenger cPTIO (100 μM) (a,d), RBC (5 vol%) (b,e) or blood plasma (5 vol%) (c,f). The vasodilatation by NO was effectively blocked or greatly attenuated by cPTIO (d), RBC (e) or plasma (f). In sharp contrast, none of these scavengers had any major effect on NO-ferroheme-Mb-induced vasorelaxations (a–c). Data in a–f were analyzed by two-way repeated measures ANOVA followed by Šídák’s multiple comparisons test. Vasorelaxation responses are shown as percent of PE-induced plateau, and the data are presented as mean ± s.e.m. The number (n) of mouse aortic rings used in different myograph experiments were as follows: a, placebo, n = 5; cPTIO, n = 6; b, placebo, n = 14; RBC, n = 14; c, placebo, n = 6; plasma, n = 6; d, placebo, n = 7; cPTIO, n = 8; e, placebo, n = 27; RBC, n = 16; and f, placebo, n = 6; plasma, n = 6. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001. Source data

Fig. 4. No release of NO from NO-ferroheme-Mb.

a,b, Original tracings (a) and mean NO concentrations (b) in headspace gas of solutions containing mouse aortic rings and 1 µM of an NO donor (DEA-NONOate, n = 7) or NO-ferroheme-Mb (Mb-NO, n = 5). c, EPR spectra of rat aortic rings incubated either with 10 µM NO-ferroheme-Mb (upper trace) or with 10 µM spermine NONOate (lower trace) in the presence of hydrophobic NO spin trap, colloid Fe(DETC)2 (100 µM; 30 min), following washout. Intravascular NO was detected only in NO donor-treated rings; in NO-ferroheme-Mb-treated rings, a six-coordinated NO-ferroheme EPR signal can be observed. Representative EPR spectra of three independent experiments. d, Partition of NO-ferroheme entity as five-coordinated species from NO-ferroheme-Mb water solution to 1-octanol phase. NO-ferroheme-Mb solution (1 mM; 1 ml) was mixed with an equal volume of deoxygenated 1-octanol, vortexed for 5 min and centrifugated (5 min). Samples from the upper (1-octanol) phase and lower (water) phase were analyzed by EPR at 77K. The 1-octanol samples exhibited an EPR signal characteristic of five-coordinated NO-ferroheme species, while the water samples showed a signal of parent six-coordinated NO-ferroheme-Mb. Double integration of the EPR signals indicated that 20–30% of NO-ferroheme groups were partitioned from Mb to 1-octanol phase. Representative spectra of two experiments. Data presented in b were analyzed by unpaired t-test and presented as mean ± s.d. P value for statistically significant difference is indicated in b. Source data

Fig. 5. Effects of NO-ferroheme on mitochondria and apo-sGC activation.

a,b, The effect of NO-ferroheme-Mb and the NO donor DEA-NONOate on mitochondrial (CI + CII) state 3 respiration evaluated by high-resolution respirometry. a, DEA-NONOate clearly inhibited mitochondrial respiration, whereas NO-ferroheme-Mb had no inhibitory effect. b, Original tracings showing mitochondrial respiration in an oxygraph and the effects of DEA-NONOate (1 μM, top) and Mb-NO (1 μM, middle and 10 μM, bottom). DEA-NONOate completely inhibited (CI + CII-dependent) state 3 respiration (red curve, top), whereas NO-ferroheme-Mb had no effect (red curve, middle and bottom). Blue curves represent oxygen concentration in the chamber. c, Apo-sGC was prepared and supplemented with either NO-ferroheme-Mb (10 μM) or DEA-NONOate (10 μM), followed by an analysis of cGMP production using an ELISA kit. Apo-sGC was clearly activated by NO-ferroheme-Mb, but not by NO released from DEA-NONOate. Respirometry data in a were repeated three times and analyzed by paired two-way repeated measures ANOVA followed by Šídák’s multiple comparisons test. Data in c were analyzed by nonparametric Kruskal–Wallis test followed by Dunn’s multiple comparisons test. Dots connected with lines in a represent paired data from three independent experiments. Dots in c represent the number of independent observations, and the data are presented as mean ± s.d. a,c, P values for statistically significant differences are indicated. ****P ≤ 0.0001. Differences not significant are indicated as NS. Source data

Fig. 6. Cardiovascular effects in vivo of NO-ferroheme-Mb and NO.

a, Blood responses to intravenous injection of different doses of NO-ferroheme-Mb and authentic NO were analyzed in anesthetized Wistar rats with regular water or supplemented with L-NAME (1 g l⁻¹). b, Traces of blood pressure in response to increasing bolus doses of NO-ferroheme-Mb and authentic NO in control rats. Administration of NO-ferroheme-Mb, but not authentic NO, showed profound and dose-dependent reductions of blood pressure, which were potentiated in rats with NOS inhibition. c,d, Summarized blood pressure responses and the duration of the responses to different doses of NO-ferroheme-Mb (c) and authentic NO (d) in control and L-NAME-treated rats. Data were analyzed with two-way ANOVA (mixed-effects model) followed by Šídák’s multiple comparisons test. The dots in c and d (circles, controls; squares, L-NAME) represent the responses in different animals, and data are presented in bars as mean ± s.d. In c,d, P values for statistically significant differences are indicated. Blue shading denotes control and orange shading denotes L-NAME. ***P ≤ 0.001 and ****P ≤ 0.0001 between the same dose given in control and L-NAME-treated rats. Source data