Differences In Oxygen-Dependent Nitric Oxide Metabolism By Cytoglobin And Myoglobin Account For Their Differing Functional Roles

Abstract

The endogenous vasodilator nitric oxide (NO) is metabolized in tissues in an O₂-dependent manner. In skeletal and cardiac muscle, high concentrations of myoglobin (Mb) function as a potent NO scavenger. However, Mb concentration is very low in vascular smooth muscle where low concentrations of cytoglobin (Cygb) may play a major role in metabolizing NO. Questions remain regarding how low concentrations of Cygb and Mb differ in their NO metabolism and the basis for their different cellular roles and functions. In this study, electrode techniques were applied to perform comparative measurements of the kinetics of NO consumption by Mb and Cygb. UV/Vis spectroscopic methods and computer simulations were performed to study the reaction of Mb and Cygb with ascorbate (Asc) and the underlying mechanism. It was observed that the initial rate of Cygb³⁺ reduction by Asc was 415-fold greater than that of Mb³⁺. In the low [O₂] range (0-50 μM), Cygb-mediated NO consumption rate is ~500 times more sensitive to changes in O₂ concentration than that of Mb. The reduction of Cygb³⁺ by Asc follows a reversible kinetic model while its reduction of Mb³⁺ is irreversible. A reaction mechanism for Cygb³⁺ reduction by Asc was proposed and the reaction equilibrium constants determined. Our results suggest that the rapid reduction of Cygb by cellular reductants enables Cygb to efficiently regulate NO metabolism in the vascular wall in an oxygen-dependent manner, while the slow rate of Mb reduction does not provide this oxygen dependence.

Introduction

Recent evidence has shown that cytoglobin (Cygb) is expressed not only in adventitial fibroblasts but also in smooth muscle cells [1, 2], playing a role in vascular nitric oxide (NO) catabolism. NO is an important signaling molecule regulating vascular tone or resistance and controlling blood flow[3, 4]. The rate of NO metabolism in the intervascular tissue [5] and in the vascular wall [6] is sensitive to change in O₂ concentration. Cygb metabolizes NO in an O₂-dependent manner [1, 7, 8]. We have recently characterized the kinetics of the O₂-dependent NO metabolism by Cygb in the presence of cellular reductants and demonstrated that Cygb can sensitively transduce a change in O₂ concentration to a change in the rate of NO consumption [9], allowing Cygb to sense and regulate the NO diffusion distance in the vascular wall in response to changes in O₂ concentration. However, metabolism of NO is not unique to Cygb. Other globins such as hemoglobin (Hb), myoglobin (Mb) and neuroglobin (Ngb) also metabolize NO [7, 8, 10, 11], and the reaction rates of NO with oxyhemoglobin (Hb²⁺O₂), oxymyoglobin (Mb²⁺O₂), oxyneuroglobin (Ngb²⁺O₂), and oxycytoglobin (Cygb²⁺O₂) are very rapid and similar to each other.

Each of the globins has a unique pattern of cellular expression and localization. Hb is mainly located in red blood cells [12]; Mb is mainly located in heart and skeletal muscles [13]; and Ngb is mainly present in neurons [14, 15]. Cygb has been found in fibroblasts and in the vascular wall [1, 2, 16]. The concentration of Hb hemes in blood is ~8 mM and the concentration of Mb in heart and skeletal muscles is several hundred μM or higher. At such high globin concentrations, the rates of NO consumption in blood, heart and skeletal muscles are very rapid so that most NO molecules will be scavenged by Hb or Mb in these locations [17-21]. Unlike Hb and Mb, Cygb concentration in vivo is in the low μM range [22, 23]. At this low concentration, Cygb does not effectively scavenge NO in the vascular wall, but likely plays a role in the regulation of vascular NO concentration. Interestingly Cygb has been reported to have a binding site for ascorbate that may facilitate its role as an effective reducing substrate [8]. Cygb is considered to have a common evolutionary ancestor with Mb [24]. Mb concentrations in the vascular wall are very low if present at all [25]. Thus, the preferential expression of Cygb over Mb in the smooth muscle suggests that there is an important functional benefit or role provided by Cygb but not Mb.

Questions remain regarding whether Cygb and Mb have similar behavior in their reaction rate and kinetics of NO metabolism and regarding the basis for the different roles and functions of these two similar globins. In order to address these questions, in this study we compare Cygb and Mb with regard to their ability to regulate the O₂-dependent NO consumption rate and with regard to the rate of their reduction by the cellular reductant ascorbate. The reaction mechanism of Cygb reduction by ascorbate is also studied and characterized.

Results

Effect of oxygen concentration on the rate of NO consumption by Mb compared to Cygb

It was previously demonstrated that the rate of NO consumption by Cygb (V_Cygb-NO) significantly decreases when oxygen concentrations change from 200 μM to 0 μM [9]. In this study, we measured the rate of NO consumption by 0.3 μM Mb in the presence of 300 μM Asc and 400 units/mL SOD with varying oxygen concentration (Fig. 1A). It can be seen that the rate of NO consumption by Mb (V_Mb-NO) only slightly decreases as the O₂ concentration decreases. In contrast, NO consumption by 0.3 μM Cygb, 300 μM Asc, and 400 units/mL SOD is much faster under room air and the rate of NO consumption significantly decreased when O₂ concentration changes from 200 μM (room air) to near 0 μM (anaerobic conditions) (Fig. 1B).

Fig. 1.

Measurements of oxygen-dependent NO consumption rate. NO (final concentration 0.5 μM) was repeatedly added to the solution in the presence of 0.3 μM Mb (A) or Cygb (B), 300 μM Asc and 400 units/mL SOD. The upwards arrows designate the time that argon gas started to flow over the solution surface to remove oxygen in the solution.

Reduction of Mb³⁺ by ascorbate

A typical change in the UV-Vis spectrum of the reduction of Mb³⁺ by Asc (10 mM) under anaerobic conditions is shown in Fig. 2A. The Soret band of Mb³⁺ (409 nm) shifts to that of the deoxy Mb²⁺ (434 nm) gradually with time, indicating an Asc-dependent reduction of Mb³⁺ to the Mb²⁺. The reduction rate of Mb³⁺ by Asc was reported as second order, that is, first order with respect to both Mb and Asc concentration [26]. Since Asc concentrations in the experiments were much greater than the Mb concentration, the reduction of Mb³⁺ by Asc in our experiments was a pseudo first order reaction. Using the time course of the change in absorbance at 409 nm, we determined the pseudo first order rate constant (k’_Mb) of Mb³⁺ reduction at 4 different Asc concentrations (10, 30, 50 and 100 mM). The experiments at each concentration were performed 5 times (n=5). The second order rate constant of Mb reduction by Asc was then obtained by plotting k’_Mb versus Asc concentration as shown in Fig. 2B. The second order rate constant of Mb reduction by Asc was determined as k_Mb=(8.7 ± 0.3)×10⁻² M⁻¹s⁻¹ (n=5).

Fig. 2.

Reduction of Mb by Asc. (A) An illustration of reduction of 10 μM Mb(Fe³⁺) to Mb(Fe²⁺) by 10 mM ascorbate at 37 °C under anaerobic conditions, which was indicated by a shift of the heme Soret band from 409 nm to 434 nm. (B) The second order rate constant k_Mb at 37 °C was determined as (8.7 ± 0.3) × 10⁻² M⁻¹s⁻¹ (n=5) by the plot of the pseudo first order rate constant k’_Mb versus Asc concentration (10-100 mM).

Computer simulations of NO consumption by Mb compared to Cygb

Using the rate constants reported in the literature and measured in our experiments (Table 1), we performed computer simulations of V_Mb-NO and V_Cygb-NO at varying O₂ concentration (using Eq. (7) for Mb and using the Equations in our previous paper [9] for Cygb). Experimental parameters in the equations, such as Mb or Cygb concentration [E], Asc concentration [A], oxygen concentration [O₂], and NO concentration [NO], were the same as the values used in the experiments. Since some rate constants and equilibrium constants in Eq. (7) at 37 °C were not available in the literature, the values of these parameters at 37 °C had to be estimated from available data at other temperatures. The reported k₁ for Mb is from 3.1 × 10⁷ M⁻¹s⁻¹ to 3.7 × 10⁷ M⁻¹s⁻¹ in the temperature range 10 °C to 25 °C [10, 27, 28]. From this temperature dependence of k₁ on Mb, we used k₁=4.5 × 10⁷ M⁻¹s⁻¹ for Mb at 37 °C. The rate constant k₂ for Mb is measured in our experiments as k_Mb. The reported k₄ for Mb at 20 °C is 1.7 × 10⁷ M⁻¹s⁻¹ and k₋₄ at 20 °C is 1.2 × 10⁻⁴ s⁻¹, so the value of k₋₄/k₄ at 20 °C is 7.1 pM. Considering that when the temperature changes from 20 °C to 37 °C, k₋₃/k₃ of Mb increases by nearly 5 times [29], we estimated that k₋₄/k₄ of Mb at 37 °C is 35 pM assuming that k₋₄/k₄ and k₋₃/k₃ have a similar temperature dependence. Simulated curves (solid lines) of V_Cygb-NO and V_Mb-NO versus [O₂] are shown in Fig. 3. The simulated curves are a good fit with the experimental V_Cygb-NO (▽) and V_Mb-NO (•) versus [O₂]. From the slopes of NO decay curves immediately after each NO concentration peak (Fig. 1), we can obtain an NO decay rate that is the sum of NO consumption (V_total) by Cygb or Mb, by O₂ and physical diffusion out of the solution. The experimental data of V_Mb-NO and V_Cygb-NO in Fig. 3 were obtained by subtracting the rate of NO diffusion out of the solution from V_total for Mb and Cygb, repectively [9]. Final parameters used in the simulated curve are listed (bold fonts) in Table 1. Our simulations show that the slope of the V_Mb-NO-[O₂] plot (Fig. 3) is mainly contributed by the NO autoxidation rate V_au rather than the Mb-mediated NO consumption rate V_NO. In contrast, V_Cygb-NO is mainly contributed by the Cygb-mediated NO consumption rate V_NO rather than the NO autoxidation rate V_au.

Table 1.

Kinetic and equilibrium constants involved in the oxygen-dependent NO consumption by Cygb and Mb.

Parameters	Values for Cygb	References	Values for Mb	References
k 1	2.2×107 M−1s−1 (20 °C) 3 ×107 M−1s−1 (37 °C)	[7] [9]	3.1×107 M−1s−1 (10 °C) 3.4×107 M−1s−1 (20 °C) 3.7×107 M−1s−1 (25 °C) 4.5×107 M−1s−1 (37 °C)	[27] [10] [28] Estimated
k2 (kCygb or kMb)	36 (56) M−1 s−1 (37 °C)	[9]	8.7×10−2 M−1s−1 (37 °C)	This work
k−3/k3	35 nM (25 °C) 37 nM (37 °C)	[22] [9]	3.2 μM (2.4 torr) (37 °C)	[29]
k −3	0.9 s−1 (25 °C) 1.3 s−1 (37 °C)	[38] [9]	15 s−1 (37 °C)	[39]
k 3	2.5×107 M−1s−1 (25 °C) 2.7×107 M−1s−1 (25 °C) 3.5×107 M−1s−1 (37 °C)	[22] [38] [9]	1.7×107 M−1s−1 (37 °C)	[39]
kh/k-h	140(25 °C) 165 (37 °C)	[22] [9]
k 4			1.7×107 M−1s−1 (20 °C)	[40]
k −4			1.2×10−4 s−1 (20 °C)	[40]
k−4/k4	8 pM (37 °C)	[9]	35 pM (37 °C)	Estimated
kau	11×106 M−2s−1 (37 °C)	[37]	11×106 M−2s−1 (37 °C)

Fig. 3.

Simulated and measured V_Cygb-NO-[O₂] and V_Mb-NO-[O₂] curves at varying O₂ concentrations. Experimental data V_Mb-NO (•) and V_Cygb-NO (▽) versus [O₂] were obtained from Figs. 1A and 1B, respectively. The V_Mb-NO-[O₂] and V_Cygb-NO-[O₂] curves (solid lines) were simulated from Eqs. (7)&(8) and equations in our previous paper[9]. In the simulations, [NO] is 500 nM, Asc concentration is 0.3 mM, and all other parameters are listed in Table 1.

The rate equations for Mb and Cygb in this study and our previous paper [9] were derived under steady-state conditions at any given NO concentration. The in vivo NO concentration at a location and at a certain time period can be considered as a steady-state concentration achieved by continuous NO generation and NO consumption. The NO concentration used in this study is 500 nM. Although NO concentrations used in tests of vascular activity can be as high as μM, the physiological NO concentration in the vascular wall for dilating blood vessels may be much lower, which is in the range of nM or below [30]. Very recently, by using ultrasensitive detector cells, it was observed that the average NO concentrations in the cerebellum and the hippocampus are 200 pM or below [31]. Direct experimental measurements of Cygb-mediated NO metabolism at the nM and sub-nM NO concentration range in our experiments are not possible because of the detection limit of NO electrodes. To examine how O₂ regulates V_Cygb-NO under physiological conditions ([NO] in nM to sub-nM range [30] and [Asc] in mM range [32]), we performed computer simulations based on rate equations (7) and (8) in this paper and rate equations in our previous paper [9] to observe changes of V_Cygb-NO and V_Mb-NO versus O₂ concentration at 3 different NO concentrations 10 nM, 1 nM and 0.1 nM (Fig. 4). In these simulations, we defined V_Cygb-NO and V_Mb-NO at PO₂=100 torr (or [O₂]=133 μM at 37 °C in buffer solution) as (V₁₀₀)_Cygb and (V₁₀₀)_Mb, respectively (Note: In experiments in air (21% O2) at 37 °C, PO₂ is ~150 mmHg after correcting for saturated vapor pressure, and the O₂ concentration in buffer solution is ~200 μM. The conversion factor from PO₂ (torr) to O₂ concentration (μM) is 200 μM /150 torr=1.33 μM/torr. When PO₂ is 100 torr or 10 torr, the corresponding O₂ concentration in buffer solution at 37 °C is 133 μM or 13 μM, respectively). The maximal PO₂ in arteries under normoxic conditions is near 100 torr [33, 34], so we consider (V₁₀₀)_Cygb and (V₁₀₀)_Mb to be the maximal rate of NO metabolism that can be achieved at a given NO concentration and other conditions (Table 1) if the reaction occurs in vivo under normoxic conditions. Both (V₁₀₀)_Cygb and (V₁₀₀)_Mb are dependent on NO concentration. To compare the O₂ dependence of V_Cygb-NO and V_Mb-NO at different NO concentrations, we normalized the rate of NO metabolism V_Cygb-NO and V_Mb-NO (dividing V_Cygb-NO and V_Mb-NO by (V₁₀₀)_Cygb and (V₁₀₀)_Mb, respectively). The normalized rate of NO metabolism increases as [O₂] increases. The O₂ concentration range corresponding to the normalized rate of NO metabolism below 0.9 is the major working range for O₂ concentration to regulate the rate of NO metabolism. From the simulated curves in Fig. 4 we can see that this working range of [O₂] for regulating the Cygb-mediated NO metabolism is from 0 to 41 μM (31 torr at 37 °C). In contrast, the [O₂] working range for regulating the Mb-mediated NO metabolism is only from 0 to 0.43 μM (0.32 torr), nearly 100 times lower than that for Cygb-mediated NO metabolism.

Fig. 4.

Computer simulations to examine the O₂ dependence of Cygb- and Mb-mediated NO consumption at different NO concentrations. The normalized rate of Cygb-mediated NO metabolism (A) and Mb-mediated NO metabolism (B) versus [O₂] were simulated at [NO]=10 nM (dashed), 1 nM (dotted) and 0.1 nM (solid) (In A, the [NO] = 1 nM and [NO] = 0.1 nM curves are nearly overlapping, while in B, all curves nearly overlap), and [Asc]=1 mM [32]. Other related parameters for computer simulations are listed in Table 1. (V₁₀₀)_Cygb and (V₁₀₀)_Mb are the rates of NO metabolism by Cygb and Mb at PO₂=100 torr, respectively. [O₂] working range for regulating the Cygb-mediated NO metabolism is from 0 to [O₂] corresponding to the normalized rate at 0.9, which is 41.3±2.4 μM for Cygb and 0.43±0.03 μM for Mb.

Effect of oxygen concentration on different globin species

The concentrations of each globin species in the reaction system (including 0.5 μM NO, 300 μM Asc, 400 units/mL SOD, 0.3 μM Mb or Cygb and varied [O₂] were calculated from Eqs. (9)-(12) or from equations in our previous study [9], respectively. The simulated concentration changes of Cygb³⁺, Cygb²⁺NO, Cygb²⁺O₂, Mb³⁺, Mb²⁺NO, and Mb²⁺O₂ versus [O₂] are shown in Fig. 5. The concentrations of Cygb³⁺, Cygb²⁺NO and Cygb²⁺O₂ were observed to significantly change with [O₂] while the concentrations of Mb³⁺, Mb²⁺NO and Mb²⁺O₂ change very little when [O₂] is lowered from 200 μM down to ~0.2 μM (Figs. 5A & 5B). When [O₂] varies from 0 to 50 μM, Mb²⁺O₂ concentration only increases from 0 to 0.35 pM, while in contrast the Cygb²⁺O₂ concentration increases much more from 0 to 0.17 nM (Fig. 5B). Therefore, the concentration increase of Cygb²⁺O₂ is nearly 500 times greater than the concentration increase of Mb²⁺O₂ in the low oxygen concentration range between 0 and 50 μM.

Fig. 5.

Effect of oxygen concentration on globin species. A: Simulated concentration changes of Cygb(Fe³⁺), Cygb(Fe²⁺NO), Mb(Fe³⁺) and Mb(Fe²⁺NO) when [O₂] varies from 0 to 200 μM. B: Simulated concentration changes of Cygb(Fe²⁺O₂) and Mb(Fe²⁺O₂) when [O₂] varies from 0 to 200 μM.

Kinetic measurements of Cygb³⁺ and Mb³⁺ reduction by ascorbate at different concentrations

Since Cygb³⁺ reduction by Asc is much faster than Mb³⁺, we further examined if the mechanism for Cygb³⁺ reduction is different from that for Mb³⁺ reduction. It was observed that the reduction of Mb³⁺ to Mb²⁺ by either 10 mM or 30 mM Asc went to completion, albeit rather slowly (Fig. 6A). In contrast, Asc can only partially reduce Cygb³⁺ to Cygb²⁺, with the final ratio of Cygb²⁺ concentration to the initial Cygb³⁺ concentration increasing with Asc concentration, reaching an equilibrium between the reactants and products at each given Asc concentration, To simultaneously record Cygb³⁺ reduction (absorbance peak at 416 nm) and Cygb²⁺ formation (absorbance peak at 428 nm), the reduction process of Cygb³⁺ was recorded by repetitively scanning from 350 nm to 700 nm on a Cary 50 UV/VIS spectrophotometer while different concentrations of Asc (0.1, 0.2, 0.7, 2, 7, 20 mM) were injected into the chamber. From these recorded spectra, we can examine the kinetic characteristics of conversion of Cygb³⁺ to Cygb²⁺ by plotting the absorbance changes vs. time at wavelengths 416 nm and 428 nm (Fig. 6B).

Fig. 6.

Kinetic characteristics of reduction of Mb³⁺ and Cygb³⁺ by different concentrations of Asc. A: Reduction of Mb³⁺ by 10 mM and 30 mM Asc. B: Simultaneous measurements of Cygb³⁺ reduction and Cygb²⁺ formation at different Asc concentrations. The reaction process of Cygb³⁺ with Asc was recorded by repetitively scanning from 350 nm and 700 nm on a Cary 50 UV/VIS spectrophotometer while different concentrations of Asc (0.1, 0.2, 0.7, 2, 7, 20 mM) were injected into the chamber at the time indicated by vertical lines. The kinetic curves of Cygb³⁺ reduction and Cygb²⁺ formation can be observed by plotting the absorbance changes vs. time at wavelength 416 nm or 428 nm.

Reaction mechanism of Cygb³⁺ reduction by Asc

The large difference between reduction of Mb³⁺ and Cygb³⁺ by Asc prompts us to think if the reaction kinetics and mechanism for Cygb³⁺ reduction by Asc has some difference from Mb³⁺ reduction by Asc. The formation of different equilibrium concentrations between Cygb²⁺ and Cygb³⁺ at different concentrations of Asc suggests that the reduction of Cygb³⁺ by Asc is a reversible reaction. Furthermore, 10 μM Cygb³⁺ cannot be completely reduced even with the addition of as high as 30 mM Asc. Based on these observations, we proposed the following reaction mechanism for Cygb³⁺ reduction by Asc:

$${\mathit{Cygb}}^{3+}+A\underset{k_{-1}}{\overset{k_1}{\leftrightharpoons }}{\mathit{Cygb}}^{3+}A\underset{k_{-2}}{\overset{k_2}{\leftrightharpoons }}{\mathit{Cygb}}^{2+}{A}^{+}$$ (1)

Assuming that the initial Cygb³⁺ concentration is c₀, the concentration of intermediate Cygb³⁺A is x, the concentration of product Cygb²⁺A⁺ is y, and the Asc concentration [A] is a constant in each measurement because Asc concentration greatly exceeds c₀, we have the following equations when an equilibrium is reached:

$$k_1(c_0-x-y)\left[A\right]=k_{-1}x$$ (2)

$$k_{2}x=k_{-2}y$$ (3)

From Eq. (3) we have:

$$y=\frac{k_2}{k_{-2}}x=K_{2}x$$ (4)

Substitution of Eq. (4) into (2) gives:

$$x=\frac{K_1{c}_0\left[A\right]}{1+K_1(1+K_2)\left[A\right]}$$ (5)

where K₁=k₁/k₋₁. Substitution of Eq. (5) into (4) gives:

$$\frac{y}{c_0}=\frac{K_1{K}_2\left[A\right]}{1+K_1(1+K_2)\left[A\right]}$$ (6)

Eq. (6) indicates that the ratio of the equilibrium concentration of Cygb²⁺ (y) to the initial Cygb³⁺ concentration (c₀) increases with Asc concentration but will not approach 1 unless K₂ approaches infinity. We further experimentally examined the formation of Cygb²⁺ in the presence of different Asc concentrations under anaerobic conditions at 37 °C by recording the change of absorbance at 428 nm with time on a Cary 50 spectrophotometer. From the Cygb²⁺ equilibrium concentration (y) at each Asc concentration ([A]), we can plot the ratio y/c₀ (or Cygb²⁺/[Cygb³⁺]₀) vs. [A] (•) and then use Eq. (6) to fit the experimental data. As shown in Fig. 7, the best fitted curve (solid line) is in excellent agreement with the experimental data. From the best fitting curve, we obtained the value of K₁K₂ (0.0633±0.002 mM⁻¹) and the value of K₁(1+K₂) (0.0751±0.0035 mM⁻¹) (n=7). The values of K₁ and K₂ calculated from the measured K₁K₂ and K₁(1+K₂) are 0.012±0.006 mM⁻¹ and 5.4±1.8, respectively.

Fig. 7.

Plot of Cygb²⁺/[Cygb³⁺]₀ vs. [A]. Eq. (6) is used to fit the experimental data (•). The best fitted curve (solid line) is in excellent agreement with the experimental data (n=7). The values of K₁K₂ (0.0633 ± 0.002 mM⁻¹) and K₁(1+K₂) (0.0751 ± 0.0035 mM⁻¹) were obtained from the best fitting curve. The values of K₁ and K₂ calculated from the measured K₁K₂ and K₁(1+K₂) are 0.012 ± 0.006 mM⁻¹ and 5.4 ± 1.8, respectively.

Discussion

The rate of Mb-mediated NO consumption (Fig. 1A) is much less regulated by oxygen concentration than that of Cygb-mediated NO consumption (Fig. 1B). This large difference implies that the rate of Cygb³⁺ reduction by Asc may be much greater than the rate of Mb³⁺ reduction by Asc. The rate of Mb³⁺ reduction by Asc is relatively slow (Fig. 2A) and the apparent first order rate constant is linear with Asc in the concentration range we examined (Fig. 2B). In contrast, the rate of Cygb reduction by Asc is much faster and the rate is not linear with Asc concentration[9]. Calculation from their rate constants gives that the reduction of Cygb³⁺ by Asc is 415 times faster than the reduction of Mb by Asc. Using Eq. (8), the rate equation for Cygb[9], and the kinetic parameters in Table 1, we calculated V_Mb-NO-[O₂] and V_Cygb-NO-[O₂] curves, which fit the experimental data very well (Fig. 3). Our results show that in the presence of hundred μM Asc and sub μM globin, Mb-mediated NO consumption is much slower than the NO autoxidation rate making autoxidation the main component in the V_Mb-NO term for the reaction with Mb. In contrast, Cygb-mediated NO consumption rather than NO autoxidation is the main component in the V_Cygb-NO term. In Fig. 4, we demonstrate O₂ dependence of the Cygb and Mb mediated NO metabolism at three NO concentrations (10 nM, 1 nM and 0.1 nM) showing that the major working range of PO₂ for the Cygb-mediated NO metabolism in our experiments is within 0-31 torr (41 μM at 37 °C). The curve shifts to the left when [NO] decreases from 10 nM to 1 nM, but the curve shift is very small when [NO] is further decreased from 1 nM to 0.1 nM. As a result of this shift, O₂ can more sharply regulate the Cygb-mediated NO consumption within the physiological range of NO concentration (low nM or below) when PO₂ is lower than 31 torr. It was reported that for wild type rats under normoxic conditions (21% O₂ in air), O₂ partial pressure (PO₂) is ~100 torr in the arteries, and 27 torr in tissues [33]. When the inspired O₂ tension dropped from 21% to 7% (hypoxia), PO₂ in tissues dropped below 10 torr. So PO₂ in these hypoxic tissues is within the major working range of PO₂ for the Cygb-mediated NO metabolism at physiological NO concentrations. Under the same conditions, the [O₂] working range for the Mb-mediated NO metabolism is only 0.32 torr, indicating that the rate of NO metabolism by Mb in the vascular wall does not change with [O₂] until [O₂] is below 0.32 torr. However, even if the Mb-mediated NO consumption can be regulated by O₂ concentration when [O₂] drops below 0.32 torr, this O₂ dependence of Mb-mediated NO metabolism is not important in the vascular wall because it is smaller than the rate of NO autoxidation (Fig. 3).

Computer simulations show that changes in oxygen concentration have little effect on concentrations of Mb³⁺, Mb²⁺NO, and Mb²⁺O₂, under the experimental conditions of this study, and the majority of Mb exists in the form of Mb³⁺ due to the slow rate of Mb³⁺ reduction by Asc (Fig. 5). In contrast, concentrations of Cygb³⁺, Cygb²⁺NO, and Cygb²⁺O₂ are very sensitive to changes in oxygen concentration. The high sensitivity of Cygb to transduce signals from changes in [O₂] into changes in the rate of NO consumption can be attributed to the significant formation rate of Cygb²⁺O₂. Since the concentration increase of Cygb²⁺O₂ is nearly 500 times greater than the concentration increase of Mb²⁺O₂ in the oxygen concentration range between 0 and 50 μM, and the rate of NO dioxygenation by Cygb and Mb is proportional to Cygb²⁺O₂ and Mb²⁺O₂ concentration, respectively, this indicates that Cygb is nearly 500 times more sensitive than Mb for “detecting” changes of oxygen concentration in regulating the rate of NO consumption in the presence of ascorbate as reductant.

It is interesting that the reaction kinetics for the reduction of Cygb by Asc is very different from that of the reduction of Mb by Asc (Fig. 6). The existence of a corresponding equilibrium concentration of Cygb²⁺ product for a given Asc concentration indicates that the reduction of Cygb³⁺ by Asc is reversible. To explain the phenomenon that Cygb³⁺ cannot be completely reduced by Asc even with Asc in gross excess, we suggest as reported previously[8] that Asc binds to Cygb³⁺ prior to the reduction step. Thus, since the reduction is reversible, Cygb³⁺A cannot be completely converted into Cygb²⁺A⁺. Eq. (6) derived from the reaction mechanism shown in Eq. (1) fits the experimental data very well (Fig. 7), indicating that the proposed reaction mechanism can explain the experimental results for the reduction of Cygb³⁺ by Asc. The physiological role of the reversible reduction of Cygb remains to be further understood; however, existence of this reversible reduction can prevent Cygb³⁺ from being fully reduced to Cygb²⁺ in hypoxia or anoxia. This lower concentration of reduced Cygb has a benefit in reducing the production of reactive oxygen species and decreasing the rate of NO consumption during reoxygenation.

If a globin acts as an oxygen sensor for the regulation of free NO and secondary vasodilation in the vascular wall, the rate of NO catabolism by this globin must be appropriate. If the NO decay rate is too high in the vascular wall, the effective NO concentration may be too low to regulate vascular tone. Conversely, if the globin-mediated rate of NO catabolism is too low, this globin cannot be the major catabolic pathway to control NO concentration. Cygb seems to meet these requirements. Cygb concentration in tissue is nearly 4 orders of magnitude lower than Hb concentration in blood and nearly 2-3 orders of magnitude less than Mb concentration in skeletal and cardiac muscle. In this low concentration range, Cygb cannot act as an efficient NO scavenger as Hb and Mb do in RBCs[19, 21], skeletal and cardiac muscle[18, 20], but it is able to efficiently regulate the rate of NO metabolism in the vascular wall because Cygb³⁺ can be rapidly recycled to Cygb²⁺ by cellular reductants to continuously metabolize NO. Since the rate of Mb reduction by Asc is 2 orders of magnitude slower than Cygb, and the vascular wall contains only very low levels of Mb, vascular Mb cannot efficiently use cellular Asc as a reductant to metabolize NO. However, evidence has shown that Mb in myocardium can efficiently generate NO by reducing nitrite as O₂ concentration decreases [35]. This Mb-mediated NO generation from nitrite has protective effects on ischemia-reperfusion injury. It has been demonstrated that some other reductants and cellular reductases can efficiently supply electrons to Cygb³⁺ for Cygb-mediated NO consumption, but these electron donors are not effective in supporting Mb-mediated NO consumption[8]. Cygb is a bis-histidyl hexacoordinate globin. It has been shown that bis-histidyl hexacoordination in globins facilitates heme reduction kinetics by small molecules[36]. These lines of evidence indicate that Cygb would be much more efficient than Mb in regulating NO metabolism in the vascular wall, where globin concentrations are in the range of low μM or less, by using cellular reductants as electron donors. In comparison, Mb is mainly present in skeletal or cardiac muscle where its high concentration enables O₂ storage, in turn facilitating O₂ diffusion in cardiac and skeletal muscles. The high concentration of Mb in these tissues along with its high affinity for O₂ enables a role in providing oxygen to the tissue under severe hypoxia. Its high rate of NO diooxygenation imparts its additional role in the oxidative degradation of NO in these muscles.

Materials and methods

Expression and Purification of recombinant Cygb

The expression plasmid for Cygb (human Cygb cDNA in pET3a) was obtained from Thorsten Burmester (Mainz, Germany) and transformed into Escherichia coli strain C41(DE3)pLysS. Cells were grown, harvested, and lysed as detailed in reference [9]. A 35 % ammonium sulfate cut was done on the supernatant, the pellet was discarded, and the supernatant was dialysed against 2 L of 50 mM TRIS-HCl, 1 mM dithiothreitol and 0.1 mM EDTA, pH 7.5 with a total of 3 buffer exchanges. After dialysis, insoluble material was removed by centrifugation (45,000 × g for 1 hour) and the human Cygb was concentrated to 50 mL using Amicon Ultra-15 centrifugal filters (Millipore Corporation) with a 10,000 molecular weight cutoff. Final purification was done on a GE Healthcare AKTA Purifier system using a 50 mL SuperLoop for sample loading. A HiPrep 16/10 DEAE FF anion-exchange column was run with a sodium chloride gradient elution followed by a HiPrep 26/60 Sephacryl S-300 High Resolution size-exclusion column eluted with 50 mM Tris-HCl, pH 7.5, 100 mM NaCl, and 0.1 mM EDTA. The protein was concentrated and stored in 50 μL aliquots in a −80° C freezer.

Reduction of Cygb³⁺ or Mb³⁺ by ascorbate

The reduction of Cygb³⁺ and Mb³⁺ by Asc in a chamber (cuvette) was performed in a similar manner to that described in [9]. Briefly, the chamber was covered by Parafilm membrane and the buffer solution (pH 7.0, 37 °C) in the chamber was deaerated by bubbling argon gas into the solution for at least 15 minutes through a tube. After adding 10 μM deaerated Cygb³⁺ or Mb³⁺ into the chamber, the tube was moved to the gas phase above the solution surface and the argon gas flow was maintained in the chamber during the experiments. After 5-10 minutes, varying concentrations of deaerated Asc were then added into the chamber using a Hamilton gastight syringe to start the reduction. The reduction of Cygb³⁺ or Mb³⁺ in the solution was monitored using a Cary 50 UV/Vis spectrophotometer scanning from 350 nm to 700 nm or fixing the wavelength at the absorbance peak of the corresponding Soret band.

Simultaneous Measurements of [NO] and [O₂]

The measurements were performed in a 4-port water-jacketed electrochemical chamber (NOCHM-4 from WPI, Sarasota FL) containing 1 or 2 mL HyClone Dulbecco’s Phosphate Buffered Saline (DPBS, pH 7.0, Thermo Scientific). A NO electrode and an O₂ electrode were connected to an Apollo 4000 electrochemical instrument (WPI, Sarasota FL). NO solution was prepared as described[5, 37]. To measure the rate of NO consumption by Cygb at different oxygen concentrations and to compare the results with that of Mb, NO (final concentration 0.5 μM) was added to the solution in the presence of 0.3 μM Cygb (or Mb), 300 μM Asc and 400 units/mL superoxide dismutase (SOD) to measure the rate of NO consumption in room air. Argon gas was then introduced in the chamber headspace to remove O₂ from the solution. While the oxygen concentration was gradually decreased by the flow of argon gas, 0.5 μM NO was repeatedly injected into the solution. From the recorded [NO]-t and [O₂]-t curves, the oxygen concentration and the NO decay rate (V_NO) at each NO peak were measured and used for plotting the V_NO-[O₂] curve.

Computer simulations of NO consumption by Cygb and Mb

Computer simulations of the rate of NO consumption by Cygb and Mb were performed on a PC using MATLAB software. The equations used in the simulations for Cygb are given in our previous paper[9]; and the equations for Mb are similar to those for Cygb except that the term k_h/k_−h is not involved:

$$V_{\mathit{NO}}=-\frac{k_1\left[E\right]\left[\mathit{NO}\right]\left[O_2\right]}{\left[O_2\right](1+\frac{k_1\left[\mathit{NO}\right]}{k_2\left[R\right]})+(\frac{k_{-3}}{k_3}+\frac{k_1}{k_3}[\mathit{NO}\left]\right)(1+\frac{k_4}{k_{-4}}[\mathit{NO}\left]\right)}$$ (7)

$$V_{\mathit{Mb}-\mathit{NO}}=V_{\mathit{NO}}+k_{\mathit{au}}\left[O_2\right]{\left[\mathit{NO}\right]}^2$$ (8)

$$\left[{\mathit{Mb}}^{2+}\right]=\frac{\left[E\right]}{1+\frac{k_4\left[\mathit{NO}\right]}{k_{-4}}+(1+\frac{k_1\left[\mathit{NO}\right]}{k_2\left[R\right]})\frac{k_3\left[O_2\right]}{(k_{-3}+k_1[\mathit{NO}\left]\right)}}$$ (9)

$$\left[{\mathit{Mb}}^{2+}{O}_2\right]=\frac{k_3\left[{\mathit{Mb}}^{2+}\right]\left[O_2\right]}{k_{-3}+k_1\left[\mathit{NO}\right]}$$ (10)

$$\left[{\mathit{Mb}}^{3+}\right]=\frac{k_1{k}_3\left[{\mathit{Mb}}^{2+}\right]\left[O_2\right]\left[\mathit{NO}\right]}{k_2\left[R\right](k_{-3}+k_1[\mathit{NO}\left]\right)}$$ (11)

$$\left[{\mathit{Mb}}^{2+}\mathit{NO}\right)]=\frac{k_4\left[{\mathit{Mb}}^{2+}\right]\left[\mathit{NO}\right]}{k_{-4}}$$ (12)

In the above equations, V_NO is the rate of NO consumption by Mb, [E] is the total Mb concentration, [R] is the Asc concentration, V_Mb-NO is the sum of V_NO and the rate of NO autoxidation in solution as defined in Eq. (8).

Conclusion

The rate of metCygb reduction by Asc is hundreds of times greater than the rate of metMb reduction by Asc. As a result, Cygb regulates the rate of NO catabolism in response to a change in oxygen concentration nearly 500 times more efficiently than its family member Mb. In addition to the large difference in the rate constants between Cygb reduction and Mb reduction by Asc, the reduction of Cygb by Asc is reversible so that Cygb³⁺ cannot be fully reduced to Cygb²⁺ in the presence of mM or even tens of mM Asc. The rapid reduction of Cygb by cellular reductants enables Cygb to efficiently regulate the oxygen-dependent NO metabolism in the vascular wall.

Abbreviations

Cygb

cytoglobin

Mb

myoglobin

Ngb

neuroglobin

Hb²⁺O₂

oxyhemoglobin

Mb²⁺O₂

oxymyoglobin

Ngb²⁺O₂

oxyneuroglobin

Cygb²⁺O₂

oxycytoglobin

Asc

ascorbate

NO

nitric oxide

O₂

oxygen

PO₂

O₂ partial pressure