Cytochrome b5 Reductase 3 Modulates Soluble Guanylate Cyclase Redox State and cGMP Signaling

Abstract

Rationale: Soluble guanylate cyclase (sGC) heme iron, in its oxidized state (Fe³⁺), is desensitized to NO and limits cGMP production needed for downstream activation of protein kinase G-dependent signaling and blood vessel dilation.

Objective: Although reactive oxygen species are known to oxidize the sGC heme iron, the basic mechanism(s) governing sGC heme iron recycling to its NO-sensitive, reduced state remain poorly understood.

Methods and results: Oxidant challenge studies show that vascular smooth muscle cells have an intrinsic ability to reduce oxidized sGC heme iron and form protein-protein complexes between cytochrome b5 reductase 3, also known as methemoglobin reductase, and oxidized sGC. Genetic knockdown and pharmacological inhibition in vascular smooth muscle cells reveal that cytochrome b5 reductase 3 expression and activity is critical for NO-stimulated cGMP production and vasodilation. Mechanistically, we show that cytochrome b5 reductase 3 directly reduces oxidized sGC required for NO sensitization as assessed by biochemical, cellular, and ex vivo assays.

Conclusions: Together, these findings identify new insights into NO-sGC-cGMP signaling and reveal cytochrome b5 reductase 3 as the first identified physiological sGC heme iron reductase in vascular smooth muscle cells, serving as a critical regulator of cGMP production and protein kinase G-dependent signaling.

Keywords: guanosine; heme; iron; nitric oxide; reactive oxygen species.

Figure 1. Oxidized sGC heme iron can be reduced and Cyb5R3 associates with oxidized sGC β

a, RASMCs were treated with ODQ (10 μM, 15 min), washed 2X with PBS, and allowed to recover for 60 min in the presence of 10 μM sildenafil. Cells were then treated DEA NONOate (1 μM, 30 min), lysed and subjected to cGMP measurements via ELISA (n=3). b, Primary RASMCs were treated with ODQ (10 μM, 10 min) and fixed in paraformaldehyde. Cells were then subjected to a proximity ligation assay to measure sGC-Cyb5R3 macromolecular complexes (red punctates) using confocal microscopy. IgG antibody served as a negative control. Green labeling is smooth muscle actin and blue are nuclei. Images on right show zoomed in areas (in white box) from images on the left. Quantification is from 10 images and complexes were normalized per nuclei as shown in graph. c, RASMCs were treated with and without ODQ (10 μM, 15 min), followed by co-immunoprecipitation of sGC β using a rabbit anti-sGC β antibody. Lysates were blotted for Cyb5R3 using a goat anti-Cyb5R3 followed by a LI-COR anti-goat 700 secondary antibody (red channel), and sGC β using a rabbit anti- sGC β followed by a LI-COR anti-rabbit 800 secondary antibody (green channel). Merge channels are shown in bottom blot. Total cell lysate input or pull-down using IgG or anti sGC β are shown for each condition. Cyb5R3 band density normalized to sGC β is shown in graph on right (n=3). * = p<0.05, ** = p<0.01, *** = p<0.001 and **** = p<0.0001 using a 1-way ANOVA (a) or a Student’s t-test (b, c) and error bars are s.e.m.

Figure 2. Loss of Cyb5R3 expression or activity, but not Cyb5b, causes sGC β protein loss and decreased basal cGMP levels

a, RASMCs were transduced with lentivirus to express non-targeting shRNA (nt shRNA) or Cyb5R3 shRNA. Following seven days of selection, cells were lysed and Western blotted for Cyb5R3, sGC β, Cyb5b and α-tubulin expression. b, shows quantification of Western blot on left (n=4). c, shows accumulation of cGMP levels from nt shRNA and Cyb5R3 shRNA transduced RASMCs treated with Sildenafil (10 μM, 18 hrs) (n=4). d, shows a Western blot of Cyb5R3, sGC β, Cyb5b and α-tubulin expression from RASMCs transduced with nt shRNA or Cyb5b shRNA. e, shows quantification of Western blot on left (n=3) and f, shows basal cGMP levels (n=4). g, Cyb5R3, sGC β, Cyb5b and α-tubulin expression from RASMCs treated with the ZINC39395747 (30 μM, 24hrs), h shows protein quantification from blot on left (n=3) and i, shows basal cGMP levels (n=3). * = p<0.05, *** = p<0.001 and **** = p<0.0001 using a Student’s t-test and error bars are s.e.m.

Figure 3. Cyb5R3 expression or activity modulates cGMP signaling following NO-stimulation in RASMCs

a, nt shRNA or Cyb5R3 shRNA transduced primary RASMCs were pretreated with sildenafil (10 μM, 15 min) followed by stimulation with DEA NONOate (0–25 μM, 30 min) or DETA NONOate (0–25 μM, 30 min). Cells were lysed and subjected to cGMP measurements via ELISA (n=3–4). b, Quantification of superoxide was accomplished by measuring and quantifying the abundance of 2-OH-E⁺; the superoxide-specific oxidation product of hydroethidine HE oxidation via HPLC coupled to an electrochemical detector (n=3). c, shows DETA-NONOate stimulated (10 μM, 30 min) cGMP response from RASMCs treated with the ZINC39395747 (30 μM, 24 hrs) (n=4). d, Measurement of intracellular cGMP via a genetically encoded cGMP biosensor in nt shRNA and Cyb5R3 shRNA transduced RASMCs. Cells were stimulated by micro spritzing with 100 mM DEA NONOate. Graph below shows the change in fluorescence from baseline per cell (n=35). e, Western blot for p-VASP²³⁹, Cyb5R3 and α-tubulin nt shRNA and Cyb5R3 shRNA transduced RASMCs treated with DEA NONOate (1 μM, 30 min) or DETA NONOate (10 μM, 30 min) (n=3). f, Western blot for p-VASP²³⁹, Cyb5R3 and α-tubulin nt shRNA and Cyb5R3 shRNA transduced RASMCs stimulated with 8-bromo cGMP (100 μM, 15 min) (n=3). * = p<0.05, ** = p<0.01, *** = p<0.001 and **** = p<0.0001 using a 1-way ANOVA (a, b, d) or a Student’s t-test (c) and error bars are s.e.m.

Figure 4. Loss of Cyb5R3 causes sGC heme iron oxidation

a, nt shRNA or Cyb5R3 shRNA transduced primary RASMCs were pretreated with sildenafil (10 μM, 15 min) followed by stimulation with BAY 58-2667 (0–10 μM, 1 hr) (n=3). b, RASMCs were treated with the Cyb5R3 inhibitor (30 μM, 24 hrs) followed by incubation with sildenafil (10 μM, 15 min) and stimulation with BAY 58-2667 (1 μM, 1 hr) (n=4). c, Naive RASMCs were simultaneously treated with ODQ (10 μM, 1 hr) and the Cyb5R3 inhibitor (0–50 μM, 1 hr) followed by stimulation with BAY-58-2667 (100 nM, 1 hr) (n=4) d, RASMCs were transduced with GFP or rat Cyb5R3 AV for 24 hrs followed by ODQ treatment for 60 mins. After 60 mins, cell were then treated with sildenafil (10 μM, 15 min), followed by BAY 58-2667 (100 nM, 60 min1 hr) stimulation and subjected to a cGMP ELISA assay (n=3–4). * = p<0.05, ** = p<0.01 and **** = p<0.0001 using a 1-way ANOVA and error bars are s.e.m.

Figure 5. Cyb5R3 directly reduces oxidized sGC

a–b, Recombinant human Cyb5R3 (10 μM), Cyb5b (10 μM), and human oxidized sGC (500 ng) were combined and incubated in 250 mM triethanolamine buffer, 1 mM DTT (positive control), ZINC39395747 (50 μM), +/- 100 μM NADH (required co-factor for Cyb5R3) in the absence of oxygen. Samples were then combined with GTP (100 μM), and Mn²⁺ (1 mM) followed by stimulation with DEA NONOate (1 μM) or BAY 58-2667 (1 μM). After 5 min, reactions were stopped and cGMP was measured. (n=3) c–d. UV-Vis spectra showing time dependent reduction of oxidized sGC by mixing purified Cyb5R3 (5 μM) and sGC (1 μM) in the presence of NADH (10 μM). * = p<0.05 using a 1-way ANOVA and error bars are s.e.m.

Figure 6. Inhibition of Cyb5R3 activity or overexpression of Cyb5R3 modulates BAY 58-2667 induced vasodilation in mouse thoracic aortas

a, Western blot of GFP and Cyb5R3 from pooled mouse thoracic aortas transduced with GFP or rat Cyb5R3 AV (n=3–5). b, Immunofluorescence images show GFP (green), Cyb5R3 (red) and nuclei (blue) in medial layers following GFP and rat Cyb5R3 AV overexpression. c, Cumulative dose response curve to Ach or d, SNP 24 hrs after Cyb5R3 inhibition (n=8–10). e, Isolated thoracic aorta segments were mounted on a two pin myograph and incubated with ZINC39395747 and ODQ. After 60 min, aortas were subjected to cumulative doses of BAY 58-2667 shown in graph. (n=5–7 vessels). f, Isolated thoracic aortic rings were incubated with GFP or Cyb5R3 AV for 24 hrs, mounted on a two pin myograph and subjected to cumulative doses of BAY 58 2667 (n=6 vessels). * = p<0.05, ** = p<0.01, and **** = p<0.0001 using a 2-way ANOVA and error bars are s.e.m.

Figure 7. Schematic of Cyb5R3 regulating sGC heme redox state

Illustration shows Cyb5R3 sensitizes sGC to NO by reducing the sGC heme iron, controlling cGMP production and vessel relaxation.