Cytochrome b₅ reductase is the component from neuronal synaptic plasma membrane vesicles that generates superoxide anion upon stimulation by cytochrome c

Abstract

In this work, we measured the effect of cytochrome c on the NADH-dependent superoxide anion production by synaptic plasma membrane vesicles from rat brain. In these membranes, the cytochrome c stimulated NADH-dependent superoxide anion production was inhibited by antibodies against cytochrome b₅ reductase linking the production to this enzyme. Measurement of the superoxide anion radical generated by purified recombinant soluble and membrane cytochrome b₅ reductase corroborates the production of the radical by different enzyme isoforms. In the presence of cytochrome c, a burst of superoxide anion as well as the reduction of cytochrome c by cytochrome b₅ reductase was measured. Complex formation between both proteins suggests that cytochrome b₅ reductase is one of the major partners of cytochrome c upon its release from mitochondria to the cytosol during apoptosis. Superoxide anion production and cytochrome c reduction are the consequences of the stimulated NADH consumption by cytochrome b₅ reductase upon complex formation with cytochrome c and suggest a major role of this enzyme as an anti-apoptotic protein during cell death.

Highlights

• •Cyt c stimulates the NADH-dependent O₂^·- production by SPMV.
• •Antibodies against Cb₅R inhibit the Cyt c-stimulated O₂^·- production by SPMV.
• •The O₂^·- production by purified Cb₅R was assessed through biochemical methods.
• •Oxidized Cyt c stimulates the O₂^·- production by Cb₅R upon complex formation.
• •Reduced Cyt c and O₂^·- are the products of the NADH-dependent activities of Cb₅R.

1. Introduction

The plasma membrane NADH oxidase activity of cerebellar granule neurons represents a disguisable activity producing superoxide anion (O₂^·-) as a collateral product of NADH consumption [1], [2], [3], [4]. The plasma membrane constituents associated to this activity are not well defined although it is known that cytochrome b₅ reductase (Cb₅R) is one of its major components present at the plasma membrane of rat cerebellar granule neurons in culture and of synaptic plasma membrane vesicles (SPMV) from rat brain [1]. This protein increases its association to lipids rafts in apoptosis [2]. In addition, 1–3 h after apoptosis induction an increment of O₂^·- has been detected at the peripheral neuronal plasma membrane [2]. This event correlates with the observed times for cytochrome c (Cyt c) release from mitochondria to the cytosol, as soon as 1 h after apoptosis induction, although the maximum peak for its release was found at 3 h [2].

In this work, we described the function of Cyt c as activator of the O₂^·- production by Cb₅R, as a component of SPMV, and results were experimentally confirmed with two isoforms of human Cb₅R. Due to the important role of Cyt c redox state in apoptosis and its reduction by Cb₅R, we propose a function of Cb₅R, as one the main defensive components during apoptosis after Cyt c release from mitochondria to the cytosol.

2. Materials and methods

2.1. SPMV preparation

Rat brain SPMV were prepared using a standard procedure as described in [1], [3].

2.2. Human Cb₅R isoforms cloning

Cloning of Cb₅R isoforms was performed as indicated in [5] using commercially available construct for soluble and primers described in Supplementary material.

2.3. Purification of recombinant human Cb₅R isoforms

Clones of Cb₅R isoforms were overexpressed in DE3 competent cells (Rosetta Gammi 2, Novagen) and the recombinant protein purified as indicated in [5].

2.4. NADH oxidase activity

NADH oxidase was measured at 37 °C as in [1], [3], [4], [6], [7].

2.5. O₂ consumption

O₂ consumption was measured using an Oxygraph Plus DW1 (Hansatech instruments) electrode in the same buffer described above, in presence of NADH (50 μM) and purified human recombinant Cb₅R isoforms at 37 °C.

2.6. O₂^·- measurement with NBT

O₂^·- production by Cb₅R was calculated measuring the reduction of NBT in the same buffer described above at pH 7.0, with NBT 200 μM and SOD 1 U/mL at 560 nm at 37 °C using a ε of 27.8 mM⁻¹ cm⁻¹ [8], [9].

2.7. Cyclic voltammetry

Qualitative measurement of the O₂^·- generated by Cb₅R was performed by cyclic voltammetry with a pyrolytic graphite electrode using the thin layer technique (membrane cut off 3.5 kDa) [5]. Cb₅R (0.6 mM) or albumins (0.6 mM) as a control were loaded onto the electrode. The set up was completed with a silver/silver chloride (Ag/AgCl) reference electrode and a platinum counter electrode to complete the three electrodes cell configuration.

2.8. O₂^·- measurement with DHE

O₂^·- formation was measured by fluorescence using dihydroethidium (DHE) [10]. Measurements were performed at 37 °C in buffer (pH 7.0) potassium phosphate 20 mM, DTPA 0.1 mM, and DHE 2 μM and Cyt c at the concentration indicated in each experiment, using a quartz cuvette. Fluorescence of DHE was measured with 470 nm and 605 nm excitation and emission wavelengths, respectively, and slits of 10 nm. Xanthine/Xanthine oxidase (XA/XO) was used to calibrate the signal.

2.9. Cb₅R:Cyt c complex formation

Complex formation was measured at 37 °C as indicated in [5].

3. Results

3.1. O₂^·- production by SPMV NADH oxidase activity is stimulated by Cyt c

We measured the effect of oxidized Cyt c (Fe³⁺) on the NADH-dependent O₂^·- production by SPMV with DHE. Addition of Cyt c (2.5 μM) to the assay produced more than 3-fold increase in the oxidation of DHE, in the presence of SPMV (7.5 μg/mL) and NADH (50 μM) (Fig. 1A, continuous line and B). In addition, SOD added to the assay blocked the Cyt c stimulated DHE oxidation rate by SPMV (Fig. 1A, dotted line and B), pointing out that the increased DHE oxidation rate was due to production of O₂^·-, as expected for a O₂^·- responsive dye [11]. The effect of a specific antibody against Cb₅R (ProteinTech, Cat #4668234) in this assay was also tested (Fig. 1A, dashed line and B). The O₂^·- production by SPMV was almost completely inhibited, i.e. ≥ 90 % inhibition, in the presence of the specific antibody against Cb₅R. We measured the DHE oxidation rate dependence upon Cyt c (Fe³⁺) concentration, in the absence (filled squares) and presence of SOD (1 U/mL) (open squares) (Fig. 1C). Addition of increasing concentrations of Cyt c to the assay produced a Cyt c dependent increase of the DHE oxidation rate. Calibration curves for O₂^·- production vs. DHE oxidation were generated using increasing XO concentrations (Supplementary Fig. S1). Thereafter, we calculated that Cyt c was stimulating the NADH-dependent O₂^·- production by SPMV almost 20-fold, reaching a maximum value of 192 ± 41 nmoles/min/mg protein, in comparison to the activity measured in absence of Cyt c (10 nmoles/min/mg protein) (Fig. 1D). The NADH dependent O₂^·- production dependence upon Cyt c concentration yielded a K_m for Cyt c stimulation of 0.2 ± 0.03 μM.

Fig. 1.

Cytcstimulated NADH-dependent O₂^·-production by SPMV. Panel A : Kinetics of the NADH dependent DHE oxidation by SPMV measured by fluorescence in the absence and presence of SOD (1 U/mL) and anti- Cb₅R antibody (1.5 μg/mL). DHE oxidation was measured at 37 °C in potassium phosphate 20 mM plus DTPA 0.1 mM (pH 7.0), using a Perkin Elmer spectrofluorimeter with 470 nm and 605 nm excitation and emission wavelengths, respectively, and 10 nm excitation and emission slits. Representative traces of DHE oxidation by SPMV (7.5 μg/mL) in the presence of NADH (50 μM), oxidized Cyt c (Fe³⁺) (2.5 μM) and DHE (2 μM),in the presence of 1.5 μg/mL anti-Cb₅R (dashed line) or 1 U/mL SOD (dotted line) are shown. Panel B: Quantification of the inhibition induced by anti-Cb₅R (1.5 μg/mL) and SOD (1 U/mL) on the DHE oxidation rate by SPMV (7.5 μg/mL) in the presence of NADH (50 μM) and oxidized Cyt c (Fe³⁺) (2.5 μM). Panel C: Dependence of the NADH-dependent DHE oxidation rate by SPMV (7.5 μg/mL) upon Cyt c concentration in the absence (filled squares) or in the presence of SOD (1 U/mL) (open squares). Panel D: NADH dependent O₂^·- production by SPMV (7.5 μg/mL) dependence upon Cyt c concentration, measured with DHE. All the results shown in this Figure are the average (± standard errors) of experiments done by triplicate.

3.2. Measurement of the O₂^·- production by recombinant Cb₅R isoforms

3.2.1. O₂^·- production by Cb₅R

The oxidation of NADH by soluble and membrane purified Cb₅R isoforms (Fig. 2 and Supp. Fig. S2A, respectively) was linearly dependent upon protein concentration (Fig. 2D). The calculated NADH oxidase activity of soluble and membrane Cb₅R was 0.27 ± 0.02 and 0.15 ± 0.02 μmoles/min/mg of protein, respectively. Under the same experimental conditions the kinetics of O₂ consumption, in the presence of NADH, by the soluble and membrane isoform of Cb₅R (Fig. 2B and Supp. Fig. S2B, respectively) yielded an O₂ consumption rate of 0.54 ± 0.02 and 0.33 ± 0.04 μmoles/min /mg protein for soluble and membrane Cb₅R respectively, calculated from the linear regression plot obtained with increasing enzyme concentrations (Fig. 2E). O₂^·- was measured from the SOD-inhibited NBT reduction (Fig. 2C and Supp. Fig. S2C), yielding a rate of O₂^·- production by soluble and membrane Cb₅R of 0.49 ± 0.02 and 0.26 ± 0.02 μmoles/min /mg of protein, respectively, from the slope of the linear dependence with Cb₅R concentration (Fig. 2F). Thus, these results yielded a stoichiometry of ≈ 2 molecules of O₂^·- generated per molecule of oxidized NADH and a good coherence for the results obtained with these three methods.

Fig. 2.

Correlation between NADH oxidation, O₂consumption and superoxide anion production by soluble Cb₅R. Panel A : Representative traces of the NADH oxidation by soluble Cb₅R (2.5 and 5 μg/mL) are shown. NADH oxidase activity was measured from absorbance decay at 340 nm at 37 °C, in the following assay medium (pH 7.0): potassium phosphate 20 mM, DTPA 0.1 mM, NADH 100 μM in presence of soluble Cb₅R 2.5 (black line) and 5 μg/mL (grey line). Dotted lines indicate the slopes used to calculate the activity. Panel B: Oxygen consumption kinetics for soluble Cb₅R 1 μg/mL (black line) 5 μg/mL (grey line). The reaction was started by addition of Cb₅R at the time marked by a large drop of trace signals. O₂ consumption was measured in presence of soluble Cb₅R using a Oxygraph Plus DW1 (Hansatech Instruments) electrode, filled with 2 mL of the assay medium indicated in the Panel A. Dotted lines indicate the slopes used to calculate the activity. Panel C: O₂^·- production by soluble Cb₅R was measured with NBT. Representative traces for the kinetics of NBT reduction by soluble Cb₅R isoform, and sensitivity to SOD is shown. NBT reduction was measured at 37 °C at 560 nm with soluble Cb₅R 2.5 μg/mL in absence (black line) or presence of SOD 1 U/mL (dotted line) in the assay medium indicated in the Panel A, supplemented with NBT 200 μM. The dotted line indicates the slope used to calculate the activity. Panels D, E and F: Dependence upon Cb₅R concentration for soluble (filled squares) and membrane (open squares) Cb₅R, respectively, of NADH oxidation, oxygen consumption and superoxide production. All the results shown in this Figure are the average (± standard errors) of triplicate experiments.

Cyclic voltammetry can be used for experimental assessment of O₂^·- production [12], [13], [14], [15], [16]. On these grounds, we have used this technique to further confirm O₂^·- production by Cb₅R. The measurement of this radical was first calibrated using KO₂ as a model compound. Two peaks appeared dependent on two generated components over the control: one at 1.25 V and another one at 0.8 V that were assigned to O₂^·- and O₂, respectively (Fig. 3A). In presence of Cb₅R (panel B), a signal similar to the observed for O₂^·- (using KO₂) appears, at the same potential, and the O₂ signal decreased correlating with O₂ consumption by the enzyme to generate O₂^·-. In presence of SOD, the O₂^·- measured signal generated by Cb₅R was equal to control (panel 3C).

Fig. 3.

Qualitative measurement of the O₂^·-production by soluble Cb₅R using cyclic voltammetry. Representative voltammograms of KO₂ added to the buffer (continuous line) vs its absence (dashed line) measured using thin layer technique with a membrane (cut off 3.5 kDa) with control/albumin is shown in panel A. Electrolyte: 20 mM Phosphate buffer / 0.1 M KCl pH 7.0 in presence of atmospheric O₂. Scan rate: 5 mV/s. Proteins concentration: 0.1 mM. Layer thickness approx. 18 mm. Panel B: Typical voltammograms of Cb₅R (continuous line) vs. the control/albumin (dashed line) using thin layer technique with a membrane (Cutoff 3.5 kDa), recorded under the same conditions of Panel A. Panel C: Effect of SOD addition over the O₂^·- production by soluble Cb₅R. Representative voltammograms in the presence of SOD 0.1 U/mL (dotted line) vs. control/albumin (dashed line) are compared to the ones obtained for soluble Cb₅R (0.1 mM) (black line) vs. soluble Cb₅R (0.1 mM) in presence of SOD (0.1 U/mL) (grey line). Conditions were the same used in previous panels. All the results shown in this Figure are representative of triplicate experiments.

3.2.2. Cyt c stimulated O₂^·- production by Cb₅R

The NADH-dependent DHE oxidation rate by purified Cb₅R was almost completely inhibited by the presence of SOD in the assay medium (Fig. 4A). The Cyt c stimulated O₂^·- production by Cb₅R (1 mg/mL) was also reliably monitored with DHE (Fig. 4B) by the dependence upon Cyt c (Fe³⁺) of the initial DHE oxidation rate. As Cyt c reduction has also been used as an indicator to monitor O₂^·- production [17], [18], we have experimentally assessed whether the SOD inhibited reduction of Cyt c can reliably monitor the NADH-dependent O₂^·- production by purified Cb₅R. The kinetics of Cyt c reduction by Cb₅R in absence (continuous line) and presence of SOD (dashed line) is shown in Fig. 4C. These results showed that SOD (1 U/mL) inhibits by 40-45 % the reduction of Cyt c upon incubation in the assay for 45 min, and about the same reduction of the initial rate of reduction up to 5–10 min. This result is in contrast with the almost complete inhibition by SOD (1 U/ml)of the Cyt c stimulated DHE oxidation by Cb₅R, and pointed out that the reduction of Cyt c was the sum of two different kinetic processes: (1) direct reduction by Cb₅R which can use Cyt c as a final electron acceptor, and (2) reduction of Cyt c by the O₂^·- released by Cb₅R.

Fig. 4.

DHE oxidation by Cb₅R is almost completely inhibited by SOD and Cytcstimulated O₂^·-production by Cb₅R. Panel A: Kinetics of DHE oxidation by soluble Cb₅R (continuous line) vs. blank (dashed line) in absence (black line) and presence of 1 U/mL SOD (grey line). DHE oxidation was measured at 37 °C in the following assay medium (pH 7.0): potassium phosphate 20 mM, DTPA 0.1 mM, NADH 50 μM and 2 μM DHE, with a fixed reductase concentration (1 μg/mL or 27.6 nM). Excitation and emission wavelengths 470 nm and 605 nm, respectively, and 10 nm excitation and emission slits. Panel B: Kinetics of DHE oxidation by soluble Cb₅R sensitive to SOD (1 U/mL), in the presence of increasing Cyt c concentrations. DHE oxidation was measured as indicated above, in the presence of the Cyt c concentrations listed in the figure. The traces shown are averages of experimental triplicates. Panel C: Representative traces for the kinetics of Cyt c reduction by soluble (black line) and membrane Cb₅R (grey line) with Cyt c (3.75 μM), in the absence (continuous lines) and presence (dashed lines) of SOD 1 U/mL. Cyt c reduction was measured at 550 nm at 37 °C in the following assay medium (pH 7.0): potassium phosphate 20 mM, DTPA 0.1 mM, Cyt c 3.75 μM and NADH 100 μM, with soluble or membrane Cb₅R (0.1 μg/mL). Panel D: The O₂^·- production rate by soluble (filled squares) and membrane (open squares) Cb₅R dependence upon Cyt c is shown. The DHE oxidation rate was measured as indicated in the Panel A, with a fixed Cb₅R (1 μg/mL or 27.6 nM)) and DHE concentration (2 μM) in the assay medium. Panel E: The O₂^·- production by soluble (filled squares) and membrane (open squares) Cb₅R isoforms dependence upon DHE concentration is shown. The DHE oxidation rate was measured as indicated in the Panel A, with a fixed Cb₅R concentrations(1 μg/mL or 27.6 nM) and Cyt c (2.5 μM). Panel F: Soluble (filled squares) and membrane Cb₅R (open squares) flavin autofluorescence dependence upon Cyt c concentration. Cyt c elicits a large increase of Cb₅R-flavin autofluorescence that allows measuring Cb₅R: Cyt c complex formation by fluorescence. Y-axis: molar fraction of Cb₅R saturated with Cyt c, which has been calculated as follows: ΔF/ΔF_max = (F-F₀)/(F_max-F₀), where F is the fluorescence intensity at each Cyt c concentration, and F₀ and F_max are the fluorescence intensity in the absence and saturating concentrations of Cyt c, respectively. Fluorescence measurements were performed at 37 °C in the following buffered solution (pH 7.0): 20 mM potassium phosphate, 1 mM EDTA, and 2 μM Cb₅R. Excitation and emission wavelengths 470 nm and 520 nm, respectively, with excitation and emission slits of 10 nm. The averages (± standard errors) of triplicate experiments are shown. Solid and dashed lines are the best non-linear squares fit to the one-binding site equation (F-F₀)/(F_T -F₀) = [Cyt c]/(K_d+[Cyt c]), and yielded values of R² = 0.99 and 0.98, respectively, for soluble (black line) and membrane (grey line) Cb₅R isoforms.

Therefore, for a proper kinetic analysis of O₂^·- production we measured the dependence of the DHE oxidation rate upon Cyt c and DHE concentration, using a fixed Cb₅R concentration and a fixed concentration of one of the twosubstrates for O₂^·- detection, as indicated in the Supp. material. The data were fit to a two substrate Michaelis-Menten kinetic model (Fig. 4 D and E). To calculate the O₂^·- production, we calibrated the oxidation of DHE by XA/XO (Supp. Fig. 1B and C). From titration results with different Cyt c concentrations and fixed DHE (2 μM) and Cb₅R (1 μg/mL) concentration, we calculated a k_cat for O₂^·- production by soluble and membrane Cb₅R of 1.37 ± 0.02 and 1.17 ± 0.02 s⁻¹, with a K_m for Cyt c of 0.29 ± 0.01 and 0.42 ± 0.02 μM, respectively (Fig. 4D). From titration with different DHE concentrations and fixed Cyt c (2.5 μM) and Cb₅R (1 μg/mL) concentrations, we calculated a k_cat for O₂^·- production by soluble and membrane Cb₅R of 1.45 ± 0.11 and 1.49 ± 0.03 s⁻¹ and a K_m for DHE of 0.19 ± 0.01 and 0.25 ± 0.04 μM, respectively (Fig. 4E).

3.3. Measurement of Cb₅R and Cyt c dissociation constant

The results shown above pointed out that Cyt c behaves as a redox partner of Cb₅R, opening the possibility to use flavin autofluorescence of Cb₅R to measure the interaction between these two proteins, see e.g. [5].Fig. 4F shows Cb₅R flavin autofluorescence intensity dependence upon Cyt c concentration, yielding a large increase of the fluorescence intensity, i.e. between 60 % and 300 % for soluble Cb₅R and for membrane Cb₅R, at saturating concentration of Cyt c (5 μM). These results revealed that Cyt c interaction with Cb₅R can be appropriately monitored by Cb₅R flavin autofluorescence. The data can be fit to a hyperbolic curve as indicated in the Material and Methods section, yielding a dissociation constant of the Cyt c/Cb₅R complex of 0.40 ± 0.05 and 0.38 ± 0.02 μM for soluble and membrane Cb₅R, respectively. Scatchard plot analysis (Supp. Fig. S3) is consistent with the binding of one Cyt c molecule per Cb₅R molecule for soluble and membrane isoforms.

4. Discussion

A scheme or the reactions described in this manuscript is shown in Supp. Fig. S4. Our data demonstrate that Cb₅R can use O₂ as an electron acceptor using NADH as substrate. Although the use of DHE formeasurement of O₂^·- has been stated to be useful for qualitative purposes in biological systems [19], we achieved to quantify O₂^·- production by purified Cb₅R with DHE, using proper controls (i.e. calibrating the signal with XA/XO in the presence of a large amount of catalase that avoid E⁺ formation when H₂O₂ is also produced) as shown in other reports [10], [11], [20]. Stoichiometric ratios between NADH and O₂ consumption indicated that Cb₅R uses one NADH molecule to reduce two O₂ molecules. Moreover, the values obtained for O₂^·- production correlated with O₂ consumption, indicating that the O₂ consumption is mainly due to O₂^·- production (Table 1). With the use of an anti-Cb₅R antibody, we confirmed that Cb₅R was responsible of the Cyt c (Fe³⁺) stimulated production by SPMV, since 90 % of the O₂^·- production was blocked by addition of specific antibodies against Cb₅R, added to the assay. As cytochrome P450s also display NAD(P)H-dependent production of O₂^.- [21] and some cytochrome P450 isoforms are associated the plasma membrane [22], it is likely that cytochrome P450s account for most of the Cb₅R-independent O₂^.- production [22], [23], although we cannot discard other O₂^.- sources. Our results also show that Cyt c binds to purified Cb₅R isoforms with dissociation constants similar to the K_m values for the Cyt c stimulated O₂^·- production by Cb₅R isoforms and close to the K_m value obtained from the NADH-dependent production of O₂^·- by SPMV.

Table 1.

O₂^·- production by human Cb₅R.

	Soluble Cb5R	Membrane Cb5R
	(μmoles/min /mg protein)	(μmoles/min /mg protein)
NADH oxidase	0.27 ± 0.02	0.15 ± 0.02
O2 consumption	0.54 ± 0.02	0.33 ± 0.04
O2·- production (SOD-inhibited NBT reductase)	0.49 ± 0.02	0.26 ± 0.02
Cyt c stimulated O2·- production (DHE)	4.1 ± 0.2a	3.5 ± 0.3a
Cyt c stimulated O2·- production (DHE)	4.4 ± 0.1b	3.8 ± 0.4b

^aValues calculated by fitting the data obtained with DHE to one substrate Michaelis-Menten kinetics.

^bValues calculated by fitting the data obtained with DHE to two substrates Michaelis-Menten kinetics.

In the context of apoptosis, the function of Cyt c reduction by Cb₅R can be seen as part of the cellular defense system, because this protein shows a widespread subcellular membrane localization, namely, endoplasmic reticulum, outer mitochondrial membrane and plasma membrane [24]. Cyt c reduction blocks apoptosis since its role in this type of cell death has been mainly attributed to the oxidized form [25], [26], [27]. For this reason, systems with ability to reduce Cyt c have an intrinsic anti-apoptotic function. Noteworthy, the payback for this reduction exerted by Cb₅R is the formation of O₂^·-, a radical also described to be formed in mitochondria upon Cyt c release [28].

Appendix A. Supplementary material

Supplementary material Fig. S1. Calibration curves for O₂^·-production vs DHE oxidation. Panel A: The DHE oxidation rate at 37 °C was calibrated with XA (0.4 mM) with increasing concentrations of XO (18, 36, and 52 mU/mL) in the following assay medium (pH 7.0): potassium phosphate 20 mM, DTPA 0.1 mM, pluscatalase (3 U/mL) and NADH 50 μM. Panel B: XA/XO (same concentrations used above) was used to calibrate the O₂^·- production using the SOD-sensitive Cyt c reduction assay under the same conditions of Panel A, in the absence and presence of 1 U/mL of SOD. Panel C: Calibration curve for DHE oxidation rates in the presence of XA (0.5 mM) at fixed XO concentrations (16, 36 and 52 mU/mL of XO) vs superoxide anion production measured by the reduction of Cyt c inhibited by SOD at the same XA/XO concentrations. All the results shown in this Figure are the average (± standard errors) of triplicate experiments. Dashed lines indicate the best linear regression fits, R² 0.99 in all cases. Fig. S2. Correlation between NADH, oxygen consumption and superoxide anion production by membrane Cb₅R. Panel A: Representative traces for NADH oxidation by membrane Cb₅R measured at 340 nm and 37 °C, in the following assay medium (pH 7.0) potassium phosphate 20 mM, DTPA 0.1 mM and NADH 100 μM, in presence of membrane Cb₅R 2.5 (black line) and 5 μg/mL (grey line). Dashed lines are the slopes used to calculate the NADH oxidase activity. Panel B: O₂ consumption kinetics at 37 °C for membrane Cb₅R 1 μg/mL (black line) 5 μg/mL (grey line) measured with an Oxygraph Plus DW1 (Hansatech Instruments) electrode, filled with 2 mL of the assay medium indicated in Panel A. The large drop of the traces monitors the time at which Cb₅R was added and dashed lines are the slopes used to calculate the O₂ consumption activity. Panel C: O₂^·- production by membrane Cb₅R measured with NBT. Representative traces for the kinetics of NBT reduction (wavelength 560 nm) by 2.5 μg/mL of the membrane Cb₅R isoform at 37 °C in the assay medium, indicated in Panel A, supplemented with NBT 200 μM, in absence (black line) or presence (dotted line) of SOD 1 U/mL. The dashed line indicates the slope used to calculate the initial NBT reduction rate. All the results shown in this Figure are representative of triplicate experiments. Fig. S3. Determination of the stoichiometry of the human Cb₅R: Cb₅complex. The Cb₅R-flavin fluorescence was titrated with Cyt c at different soluble (Panel A) and membrane (Panel B) Cb₅R concentrations at 37 °C in potassium phosphate 20 mM plus DTPA 0.1 mM buffer (pH 7.0). Cb₅R concentrations: 1 μM (square), 2 μM (circle) and 5 μM (up triangle), for soluble (filled symbols) and membrane isoforms (open symbols). Fluorescence intensity was measured with 470 nm and 520 nm excitation and emission wavelengths, respectively, and 10 nm excitation and emission slits. Panel C: The fluorescence intensity increase (F-F₀) of soluble and membrane Cb₅R autofluorescence dependence upon Cyt c concentration (symbols indicate the same Cb₅R concentrations used in Panels A and B). All curves for Cb₅R autofluorescence dependence upon Cyt c fit well to an hyperbolic curve (dashed line), with non-statistically significant differences in the half saturation value, pointing out that the binding affinity was not significantly altered in the Cb₅R concentration range used in these assays. Panel D: Scatchard-like plot of the data obtained in the Cb₅R-flavin fluorescence titration with Cyt c assuming a 1:1 stoichiometric ratio of the complex Cb₅R:Cb₅. Bound and free [Cyt c] have been calculated as follows: [Cyt c]_bound = molar fraction of Cb₅R saturated with Cyt c x [Cyt c]_total, and [Cyt c]_free = [Cyt c]_total -[Cyt c]_bound. The molar fraction of Cb₅R saturated with Cyt c has been calculated as follows ΔF/ΔF_max = (F-F₀)/(F_max-F₀), where F is the fluorescence intensity at each Cyt c concentration, and F₀ and F_max are the fluorescence intensity in the absence and saturating concentrations of Cyt c, respectively. The data lead to a value of 1.0 ± 0.07 for the intercept with the X-axis, meaning 1 Cb₅ binding site per Cb₅R molecule. The average of the plots obtained at different Cb₅R concentrations is shown as a continuous line and SD error is shown as dashed lines. Fig. S4. O₂^·-production by Cb₅R (panel A) and its stimulation by Cytc(panel B) upon complex formation with Cytc. Panel A shows the production of O₂^·- (red) by Cb₅R (yellow backbone) (PDB file: 1UMK, is the surface representation of the crystal model for erythrocyte NADH-Cb₅R) using NADH and O₂ as substrates (in black). Panel B shows the two observed pathways for Cyt c reduction (pink backbone) by Cb₅R upon complex formation with oxidized Cyt c (Fe³⁺) (dark red backbone). Cyt c can be reduced through electron transfer from Cb₅R or through the stimulated superoxide anion production generated upon formation of the complex between Cb₅R and Cyt c. In both cases NADH is used as a substrate. Arrow thickness indicates the stimulation of activities.

Supplementary material