Mitochondrial respiration in the low oxygen environment of the cell effect of ADP on oxygen kinetics

B1OCHIMICA ET BIOPHYS1CA ACTA

ELSEVIER

Biochimica et Biophysica Acta, 1365 (1998) 249-254

Mitochondrial respiration in the low oxygen environment of the cell Effect of ADP on oxygen kinetics Erich Gnaiger*, Barbara Lassnig, Andrey V. Kuznetsov, Raimund Margreiter Department of Transplant Surgery, D. Swarovski Research Laboratory, University Hospital Innsbruck, A-6020 Innsbruck, Austria Received 10 February 1998; received in revised form 12 March 1998; accepted 12 March 1998

Abstract Oxygen levels in the intracellular microenvironment of tissues such as heart are extremely low, at 1-2% of standard atmospheric oxygen pressure. Kinetic studies with isolated mitochondria suggest a regulatory role of oxygen under these conditions, particularly in active states at high ADP concentration, when oxygen affinity was lower than in the resting state at ADP limitation. The oxygen pressure at 50% of maximum flux, Ps0, was 0.035 and 0.057 kPa in heart and liver mitochondria, respiring in state 3 on substrates for complex I or II and II, respectively. Ps0 in the resting state 4 was 0.02 kPa. The apparent kinetic efficiency, Jmax/Pso, increased from the resting to the active state, despite the decrease of oxygen affinity, 1/Pso. Consequently, the relative increase of respiratory flux by ADP activation, expressed as the adenylate control ratio, declined under hypoxia, but not to the extreme of a complete loss of the scope for activation, which would occur at constant Jmax/Pso" High oxygen affinity is achieved by an excess capacity of cytochrome c oxidase relative to the respiratory chain and a correspondingly low turnover rate of this enzyme, consistent with the concept of kinetic trapping of oxygen [1]. © 1998 Elsevier Science B.V. Keywords: Oxygen affinity; Catalytic efficiency; Respiratory control; Hypoxia; Mitochondrion; Heart; Liver

1. Introduction: bioenergetics and intracellular oxygen levels Oxygen pressure in the intracellular microenvironment within tissues is very low, setting kinetic constraints on mitochondrial respiration under normoxic conditions and hypoxic stress [2]. Myoglobin saturation levels in heart and skeletal muscle indicate *Corresponding author. Fax: +43-512-5044625; E-mail: [email protected] Abbreviations: Jo2, oxygen flux through the respiratory chain, per mg mitochondrial protein; Jmax, maximum flux at oxygen saturation; Jo2.v, volume-specific oxygen flux in the Oxygraph chamber; Po2, partial oxygen pressure; Pso, oxygen pressure at 50% of Jmax; Jmax/Pso, apparent catalytic efficiency.

average intracellular oxygen pressures of 0.3 kPa (2 mmHg; 3 IxM O2), c. 2% of air saturation [3,4]. Relatively few studies address the role of oxygen in mitochondrial respiratory control [5,6], whereas a substantial literature exists on the regulation of oxidative phosphorylation by ADP, Pi, creatine, Ca 2÷ and substrates. In the latter bioenergetic studies with isolated mitochondria, saturating oxygen levels are maintained which are at least one order of magnitude above in vivo concentrations. On the other hand, several investigations of oxygen kinetics use isolated mitochondria only in the resting state, without addition of ADP, or after uncoupling [7,8]. For a physiological interpretation of oxygen kinetics, however, information is required on passive and active oxida-

0005-2728/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII: S0005-2728(98)00076-0

250

E. Gnaiger et al. / Biochimica et Biophysica Acta 1365 (1998) 249-254

five phosphorylation. This is particularly evident for the heart, where the passive mitochondrial state is clearly not a standard condition. In the present study on oxygen kinetics of isolated mitochondria, we address the following questions: (1) Are the current controversies on mitochondrial oxygen kinetics a result of physiological differences or related to methodological artefacts? (2) How is oxygen affinity regulated by metabolic state and rate? (3) Does the control by oxygen of mitochondrial oxidative phosphorylation follow a general pattern, or are tissue-specific differences important? ( 4 ) W e discuss the physiological role of mitochondrial oxygen affinity under the complex conditions of multiple respiratory control by oxygen and adenylate levels.

2. Measurement of mitochondrial oxygen kinetics

The measurement of mitochondrial oxygen kinetics presents an experimental challenge, and precautions are required for the appropriate instrumental basis and methodological controls at extremely low oxygen. In our view, any study of cellular or mitochondrial oxygen kinetics has to observe the following practical considerations [2,9]: (1) The respirometric measuring chamber must be made of inert materials which do not store and release dissolved oxygen. For example, PTFE-coated magnetic stirrers present a serious problem for back-diffusion of oxygen. (2) Gas-aqueous phase boundaries must be avoided to exclude uncontrolled oxygen gradients within the measuring system. This can be achieved in a well stirred closed system. (3) Measurement of oxygen back-diffusion and mathematical correction for background are required for high accuracy of flux at low oxygen. (4) Digital data recording at moderately high frequency (one data point per second) is a basis for deconvolution of the oxygen signal with the calibrated time constant of the oxygen sensor, and for calculating oxygen flux, Jo2.v [pmol's l'cm-3], as the time derivative of oxygen concentration. (5) High sensitivity must be ensured by accurate calibration of the oxygen zero signal and correction for signal drift at near-zero oxygen levels. (6) The enzyme concentration must be chosen in a tested range compat-

ible with the instrumental sensitivity and time resolution. In this range, protein-specific flux, Jo2 [nmol's-~-mg-~], and oxygen affinity must be independent of enzyme concentration. With decreasing enzyme concentration, the volume-specific flux declines toward the instrumental limit of detection. Then oxygen depletion becomes too slow for accurate resolution of oxygen kinetics in a closed chamber. At increasing enzyme concentration and high activity, on the other hand, volume-specific flux in the respirometric chamber causes a rapid depletion of oxygen over time. This sets increasing demands on the instrumental time resolution. Moreover, an absolute limit for the analysis of oxygen kinetics in a closed chamber is reached when the number of data points reduces the degrees of freedom in the oxygensensitive range below a limit set by statistics. (7) The oxygen range for kinetic analysis must be carefully chosen. While the full oxygen-dependent range must be included, it is important to exclude effects at nonphysiological high oxygen pressure which are related to higher-order kinetic terms or unrelated to oxygen, e.g., time of exposure. Data points, which are recorded after zero (or minimum) oxygen is reached, are omitted for statistical reasons. (8) Nonlinear hyperbolic fits of the flux-oxygen pressure relation must be tested by observation of residuals rather than linearity of double-reciprocal plots [10], as a basis for expressing results in terms of the parameters Jmax (maximum flux at oxygen saturation) and Ps0 (oxygen pressure at half-maximum flux), Jmax " Po 2

J ° 2 - Pso + P%

(1)

Importantly, Jmax is constant within, but variable between experiments, i.e., under the control of ADP in our study. We studied oxygen kinetics of mitochondria isolated from rat heart and liver with the Oroboros Oxygraph and DatLab Software (Oroboros, Innsbruck, Austria) [2,9]. The incubation medium contained 200 mM sucrose, 20 mM Hepes, 0.5 mM EGTA, 1 g/1 bovine serum albumin (essentially fatty acid free), 3 mM MgC12, 20 mM taurine and 10 mM KH2PO4, osmolarity - 3 0 0 mosM, pH 7.1 at 30°C. By selecting oxygen pressures <1.1 kPa for hyperbolic analysis [2], oxygen levels > 10 times the

E. Gnaiger et al. / Biochimica et Biophysica Acta 1365 (1998) 249-254

Ps0 were covered, following a well-established rule in enzyme kinetics. Mitochondrial [2] and cellular respiration [9] could be described by the hyperbolic relationship (Eq. (1); Fig. 1). Contrary results with rat liver mitochondria [ 11] are based on a too narrow

A

,,Tu~150 "~100 &

5o

..,Oo 0,0

0,5

1,0

PO2 [kPa] B • 300

"~150 Q,.

P50 = 0.029 kPa 0.094 mg.cm -3

& o

0,0

I

I

0,5

1,0

PO2 [kPa]

,_ C

03

! 1000

-~500

~= & ~O

P50 = 0.109 kPa 0.240 mg.cm -3 .. , ,

0 0,0

0,5

1,0

PO2 [kPa] Fig. l. Oxygen flux per volume of the Oxygraph chamber, Joz.v (pmol1 -3 s -cm ) , as a function of oxygen pressure, Poz (kPa), in active heart mitochondria (state 3, 1 m M A D E 1 mM ATE 2 mM pyruvate, 5 mM malate), at different mitochondrial protein concentrations (mg-cm-3). The low-oxygen range < 1.1 kPa is shown as relevant for mitochondrial oxygen kinetics. Protein-specific maximum flux and Pso were independent of protein concentration (A and B), but hyperbolic fitting yields artefacts at protein concentrations >0.15 mg.cm -3, corresponding to volume-

specific Jmax>500 pmol's 1.cm-3 (C).

251

oxygen range <0.1 kPa ( < 1 ~M 0 2 ) and doublereciprocal plots. Moreover, the dependence of the apparent Pso on enzyme concentration [12] raises doubts on the instrumental and methodological reliability. Volume-specific flux is shown as a function of oxygen pressure at three protein concentrations of heart mitochondria (Fig. 1). Protein-specific flux at -1 kinetic oxygen saturation, Jrnax' was 4.5 nmol.s -1 mg , and Ps0 was 0.035 kPa in the active state, measured in the range of protein concentrations from 0.02 to 0.12 m g . c m 3 (Fig. 1A and B). At higher protein concentrations, volume-specific fluxes > 5 0 0 pmol. s-] .cm-3 resulted in artificially high Pso values (Fig. 1C). Using a steady-state approach with oxygen transfer from the gas to aqueous phase, Sugano et al. [13] found a Ps0 of 0.08 kPa in pigeon heart mitochondria at a protein concentration of 0.2 mg-cm -3 (for a review see Ref. [2]). At 10 m g . c m -3, however, Cole et al. [14] obtained a Pso of 0.005 kPa in rat skeletal muscle mitochondria at state 3. Under these conditions, oxygen flux per volume of the mitochondrial suspension is extremely high, to the extent that oxygen gradients between gas phase and oxygen sensor may explain the erroneously low Pso value, reference to which leads to the conclusion that oxygen does not play any regulatory role under physiological conditions [15]. A Ps0 of 0.04 kPa was obtained in a closed system for rat heart mitochondria respiring in the absence of ADP, without a value reported for the active state 3 [8]. Costa et al. [16] incubated rat heart mitochondria at concentrations of 0.2 to 0.5 mg protein.cm 3 in a closed chamber. At state 3 (25°C), volume-specific fluxes reach up to 900 p m o l ' s - ~ ' c m -3, and the resulting Pso is too high (0.15 kPa [16]) owing to an insufficient number of kinetically relevant data points and a rapid oxygen depletion in excess of the instrumental time resolution (see above). The resulting Pso in active heart mitochondria would correspond to a serious oxygen limitation by 30% under normoxia at an intracellular oxygen pressure of 0.3 kPa (see Eq. (2)). This appears to be incompatible with the close matching of maximum respiration in the intact heart and isolated mitochondria [17]. Both extreme positions on the regulatory role of oxygen must be revised on the basis of our results in heart and liver mitochondria.

E. Gnaiger et al. / Biochimica et Biophysica Acta 1365 (1998) 249-254

252

3. Oxygen and adenylate control of mitochondrial respiration

oxygen affinity, 1/Pso, of mitochondrial respiration is a complex function of metabolic state and rate [9].

The Ps0 of active rat liver mitochondria was 0.057 kPa, nearly twice as high as in heart. In both types of mitochondria, the Ps0 increased with Jmax, from a minimum at rest (state 4) to the active coupled state 3 (Fig. 2). Uncoupling with optimum carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) concentration, however, resulted in a decrease of Ps0 at state 3u, at maximum flux and adenylate concentrations similar to state 3. This pattern is consistent with a thermodynamic contribution to the regulation of Ps0 [2]. A small but significant effect of ATP on Pso was observed at resting states 2 and 4, unrelated to the change in Jmax: whereas flux increased slightly after addition of ATP, which is generally attributed to nonmitochondrial ATPase activity in the presence of 2+ Mg , the corresponding Pso decreased (Fig. 2). The

°

e, . ,

°'°'l to o

I

0,02 0,00

0,0

0,4

0,8

1,2

Jmax [nmol.sl.mg1] Fig. 2. Oxygen pressure at half-maximum flux, Ps0 (kPa) (reciprocal value of oxygen affinity), as a function of maximum flux, Jmax (umol O z . s - l . m g I mitochondrial protein) (at kinetic oxygen saturation), in isolated rat liver mitochondria with 10 mM succinate and 0.5 I~M rotenone. (&) State 2 (with substrate, no adenylates); (©) state 4, resting (1 mM ATP); ([:3) state 3, active (1 mM ADR 1 mM ATP); (V) state 3u, uncoupled (2 ~M FCCP). The linear regression shows the increase of p50 with rate (passive to active state, open symbols; dotted lines: 95% confidence limits of the regression). Each symbol represents the result of a single aerobic-anoxic transition, showing the full range of experimental variation. Po2 was dependent on respiratory rate, but also on metabolic state.

4. Pso and catalytic efficiency in passive and active states What is the physiological significance of the increase of p5 0 in the transition from passive to active states (linear regression in Fig. 2)? As derived from Eq. (1), the Pso determines the limitation of flux relative to Jmax at any given oxygen pressure, which is equivalent to the elasticity, eo2,J of oxygen flux, with respect to oxygen pressure [18], j

1

~o2 - 1 + po2/Pso

(2)

At an intracellular Po2 of 0.3 kPa, therefore, mitochondrial respiration in the active heart is oxygen limited by a factor of 0.10, whereas mitochondrial respiration in active liver would be more oxygen limited with an elasticity of 0.16, at an identical intracellular Po2 (Eq. (2)). If tissue oxygenation does not change between active and passive states, oxygen limitation would amount to only 6% of Jmax at state 4, due to the lower Ps0 of 0.02 kPa (Fig. 2). Nevertheless, active oxygen flux was higher than passive flux at any oxygen level (Fig. 3). Therefore, respiratory control by oxygen cannot be discussed on the basis of po2/Pso ratios alone. The apparent catalytic efficiency, Jmax/Pso, expresses the response of oxygen flux to Po2 at extremely low oxygen levels ( c o m p a r e kcat[K M [19]). A constant ratio of Jmax/Pso is expected on the basis of kinetic models of cytochrome c oxidase [1,5,20]. The increase of Pso with Jmax agrees with this model. However, the apparent catalytic efficiency was not constant but increased by adenylate activation, despite the decline of the affinity. Therefore, a scope of activation by ADP was maintained down to minimum oxygen levels. A maximum adenylate control ratio (Jo2 at state 3/Jo2 at state 4) was obtained at kinetic oxygen saturation, corresponding to the classical respiratory control ratio. The adenylate control ratio decreased under hypoxia (Fig. 4).

E. Gnaiger et al. / Biochimica et Biophysica Acta 1365 (1998) 249-254

C02 [p.M] 0.0

1.0

2.0

253

Co2 [I.IM] 3.0

4.0

0.0

100

9

1.0 '

=o

~_~ 75

2.0 '

3.0 '

iii

~7

4.1

1

i ,,,,i iiii ::::::::..:.::::~

~:::~::~!~J

~

~.

-

95

<

0 0.0

0.1

0.2 P02 [kPa]

0.3

0.4

3

:::$

1 , 0.0

, 0.1

, 0.2

, 0.3

0.4

0o2 [kPa]

Fig. 3. Oxygen flux of heart mitochondria as a function of oxygen pressure, Po2, in passive and active states with pyruvate and malate. The vertical arrow indicates the scope of activation by ADP at an intracellular Po2 of 0.3 kPa (3 txM). The circle and square show the average Pso in state 4 and 3 (error bars are S.D.; N = 9 and 26, respectively). The dotted lines through the Ps0 values illustrate slopes (0.5 Jm,xlPso)proportional to the apparent catalytic efficiency, which increases with activation by ADP (bow arrow).

Fig. 4. Adenylate control ratio, ACR (ratio of flux at state 3 and 4), as a function of oxygen pressure, Po2, or oxygen concentration, Co2, in heart mitochondria, showing the decline in the scope for activation under hypoxia (thick line, from Fig. 3). A C R = 1.0 indicates a zero scope for ADP activation. Upper limit (horizontal dotted line, for the theoretical case of constant Ps0 in state 4 and 3): ACR is constant and independent of Poz, equal to the classical respiratory control ratio measured at kinetic oxygen saturation. Lower limit (for a theoretically constant catalytic efficiency, Jmax/Pso=const.): ACR declines to 1.0 at zero oxygen.

5. Conclusions: hypoxic versus hyperoxic stress

effect of hypoxia, particularly in inflammatory and ischemia/reperfusion pathologies [21]. Kinetics of the isolated cytochrome c oxidase [2224] must be complemented by phenomenological kinetics of the respiratory chain for improving the conceptual link between mitochondrial bioenergetics, cell physiology and mitochondrial pathology. This contributes to a functional interpretation of the apparent excess capacity of cytochrome c oxidase, not merely in terms of a threshold effect providing a safety margin against the accumulation of mitochondrial damage [25], but as the mechanism for the regulation of high oxygen affinity. Whereas the flux control coefficient of cytochrome c oxidase is low at high oxygen concentration [26], it is expected to increase when measured under physiological Po2 [2,27]. Excess capacity of cytochrome c oxidase in heart mitochondria is about twice as high as in liver, which explains the higher oxygen affinity of heart

For evaluation of the physiological role of intracellular oxygen in the control of metabolic flux, the experimental data base must be critically evaluated, and oxygen affinities must be measured under relevant metabolic conditions in the range of mitochondrial states 4 and 3, approaching state 3 at maximum activity. Intracellular Po2/Pso ratios of 5 10 indicate that mitochondrial respiratory capacities are slightly but not excessively higher than the organism's capacity for oxygen delivery to the tissue. Under these conditions, oxygen delivery is a limiting factor for aerobic performance, and the corresponding high fluxes of mitochondrial electron transport and oxidative phosphorylation take place in a low oxygen environment at reduced oxidative stress. Competitive inhibition of cytochrome c oxidase by NO increases the mitochondrial Ps0 and may thus accentuate the

254

E. Gnaiger et al. / Biochimica et Biophysica Acta 1365 (1998) 249-254

mitochondria in terms of a lower turnover rate of this enzyme even at maximum flux through the respiratory chain ( [18]; Lassnig et al., in preparation).

Acknowledgements Supported by a research grant of the University of Innsbruck. A.K. was supported by an Oroboros Research Award.

References [1] M.I. Verkhovsky, J.E. Morgan, A. Puustinen, M. Wikstr6m, Nature 380 (1996) 268-270. [2] E. Gnaiger, R. Steinlechner-Maran, G. M6ndez, T. Eberl, R. Margreiter, J. Bioenerg. Biomembr. 27 (1995) 583-596. [3] B.A. Wittenberg, J.B. Wittenberg, Annu. Rev. Physiol. 51 (1989) 857-878. [4] R.S. Richardson, E.A. Noyszzewski, K.F. Kendrick, J.S. Leigh, P.D. Wagner, J. Clin. Invest. 96 (1995) 1916-1926. [5] B. Chance, J. Gen. Physiol. 49 (1965) 163-188. [6] D.P. Jones, Am. J. Physiol. 250 (1986) C663-C675. [7] D.F. Wilson, W.L. Rumsey, T.J. Green, J. Vanderkooi, J. Biol. Chem. 263 (1988) 2712-2718. [8] W.L. Rumsey, C. Schlosser, E.M. Nuutinen, M. Robiolio, D.F. Wilson, J. Biol. Chem. 265 (1990) 15392-15399. [9] R. Steinlechner-Maran, T. Eberl, M. Kunc, R. Margreiter, E. Gnaiger, Am. J. Physiol. 271 (1996) C2053-C2061. [10] A. Cornish-Bowden, Fundamentals of Enzyme Kinetics, Portland Press, London, 1995.

[11] L.C. Petersen, E Nicholls, H. Degn, Biochem. J. 142 (1974) 247-252. [12] L.C. Petersen, E Nicholls, H. Degn, Biochim. Biophys. Acta 452 (1976) 59-65. [13] T. Sugano, N. Oshino, B. Chance, Biochim. Biophys. Acta 347 (1974) 340-358. [14] R.C. Cole, EC. Sukanek, J.B. Wittenberg, B.A. Wittenberg, J. Appl. Physiol. 53 (1982) 1116-1124. [15] T.E.J. Gayeski, C.R. Honig, Am. J. Physiol. 260 (1991) H522-H531. [16] L.E. Costa, G. M6ndez, A. Boveris, Am. J. Physiol. 273 (1997) C852-C858. [17] V.K. Mootha, A.E. Arai, R.S. Balaban, Am. J. Physiol. 272 (1997) H769-H775. [18] E. Gnaiger, B. Lassnig, A. Kuznetsov, G. Rieger, R. Margreiter, J. Exp. Biol. 201 (1998) 1129-1139. [19] A. Fersht, Enzyme Structure and Mechanism, Freeman, New York, 1985. [20] K. Minnaert, Biochim. Biophys. Acta 50 (1961) 23-34. [21] G.C. Brown, Mol. Cell. Biochem. 174 (1997) 189-192. [22] B.G. Malmstr6m, Acc. Chem. Res. 26 (1993) 332-338. [23] E Nicholls, E Butko, J. Bioenerg. Biomembr. 25 (1993) 137-143. [24] S. Riistama, A. Puustinen, A. Garcia-Horsman, S. Iwata, H. Michel, M. Wikstr6m, Biochim. Biophys. Acta 1275 (1996) 1-4. [25] T. Letellier, M. Malgat, J.-E Mazat, Biochim. Biophys. Acta 1141 (1993) 58-64. [26] A. Kuznetsov, J.F. Clark, K. Winkler, W.S. Kunz, J. Biol. Chem. 271 (1996) 283-288. [27] B. Korzeniewski, J.-E Mazat, Acta Biotheor. 44 (1996) 263-269.