Biochimicu et Biophysics Acta, 456 (1976) 129-148 ‘0 Elsevier Scientific Publishing Company, Amsterdam
Printed in The Netherlands
BBA 86031
UNCOUPLING
OF OXIDATIVE
WALTER
G. HANSTEIN
Depurtmenr
qf Biochemistry,
(Received November
Scripps
PHOSPHORYLATION
Clinic and Research
Foundation,
La Jolla.
Cal&
92037 ( U.S.A)
19th. 1975)
CONTENTS
III.
................................. efficiency ........................... Respiratory control ..............................
IV.
Types of uncoupling
V.
Interactions
of uncouplers with other inhibitors.
VI.
Interactions
of uncouplers
components
.................................
I. II.
VII.
Introduction
129
Phosphorylation
.............................
VIII. The concept of stoichiometry.
132
.................
with the mitochondrial
Affinity labeling of mitochondria by picrate.
130 I31
by uncouplers
membrane
134 and with membrane 136
.................
137
.........................
140
IX.
Uncoupling
X.
On the mechanism of uncoupling of oxidative phosphorylation
............................
................................. ....................................
I41
...........
143
Acknowledgements
I46
References
146
I. INTRODUCTION
Food stuff such as lipids, carbohydrates presence of oxygen. Were it not for the hurdles material would react with oxygen and produce and other inorganic compounds. Living things organic materials and oxygen, and convert it useful than heat. In cells of higher organisms,
and proteins are metastable in the of high activation energies, all organic heat, carbon dioxide, water, nitrogen conserve the redox energy present in into forms of energy which are more this vital function is performed by
Abbreviations: CCCP, Cl-CCP, carbonyl cyanide m-chlorophenylhydrazone; FCCP, carbonyl cyanide p-trifluoromethylphenylhydrazone; HQNO, 2-heptyl-4-hydroxyquinoline N-oxide; S-6, 5-chloro-3-(p-chlorophenyl)-4’-chlorosalicylanilide ; S-13, 5-chloro-3-r-butyl-2’-chloro-4’-nitrosalicylanilide; SF-6847, 3,5-di-r-butyl-4-hydroxybenzylidene malononitrile; TTFB, 4,5,6,7-tetrachloro-2trifluoromethylbenzimidazole.
130 mitochondria,
which have therefore
been called
“the microscopic
powerplants”
of
the cells [I]. Specifically, several complex enzyme assemblies [2,3] located in the inner mitochondrial membrane catalyze oxidative phosphorylation, a network of reactions
in which the oxidation
the phosphorylation source of chemical
of ADP.
of substrates
is coupled
ATP, a high-energy
to the synthesis
compound,
energy for a large variety of metabolic
of ATP by
serves as the general
processes
such as muscular
contraction, transport of ions and small molecules, and syntheses necessary for growth and maintenance of the organism. In tightly coupled mitochondria, very little oxidation occurs in the absence of ATP synthesis. This phenomenon is known as respiratory control. Uncouplers, a large class of relatively simple organic chemicals, have the ability to abolish the coupling of substrate oxidation to ATP synthesis. As a consequence, the latter reaction comes to a halt, because it is deprived of the necessary energy input, whereas the former reaction, uninhibited by respiratory control, proceeds at maximal rate and produces heat instead of ATP [4]. There is much detailed knowledge about the composition and the mode of action of the electron transport system [2,5-71, and the ADP-phosphorylation system [8-IO] in mitochondria. In contrast, the nature of the device which couples these two systems is not known with certainty, nor is there a generally accepted mechanism of uncoupling of oxidative phosphorylation. Several general reviews on this topic covering the literature prior to 1974 are available [I lLl5]. The scope of this article does not permit exhaustive treatment of even the most recent literature. It will only be possible to introduce the problem, to discuss selected aspects which have been most actively investigated in the last several years, and to summarize what recent studies of uncouplers may have taught us about the mechanism of oxidative phosphorylation.
II. PHOSPHORYLATION
EFFICIENCY
Some basic facts, summarized important
to understand
uncoupling
in Fig. 1, are necessary of mitochondrial
to appreciate
oxidative
why it is
phosphorylation.
Electron transport from NADH or succinate to oxygen, via coenzyme Q and cytochrome c, is catalyzed by a number of flavoproteins, cytochromes, nonheme iron and copper proteins (not shown), and accompanied by the release of free energies amounting to 52 or 36 kcal/mol, respectively [4]. The energy in each major step is conserved in a common pool [16], designated by a squiggle sign (-). At present, it is not known (and is a matter of heated debates) whether the form of potential energy produced in the coupling events is best described as a chemical intermediate, as a combination of membrane potential and pH-gradient, or as a conformational state. The energy for the synthesis of just one molecule of ATP is generated by the passage of two electrons through each of the coupling sites 1, II and III. Thus, oxidation of one molecule of NADH supports the formation of no more than three
131 SUCCINATE/FLlhWRATE
Electron transportsystem (Flawproteins, nonheme iron proteins, cytochromes, copper, coenzyme
Q)
SITE
.
aride hvdroxylamine cyanide
lr
amvtal rotenone
SITE
I
SITE
II
III
gua"i%$q+/& Coupling system (OSCP, B-type factors, UBP. DCCD-binding protein)
ION TRANSPORT-
. (161
HEAT
II
oligomycin,DCCD ADP
PHOSPMTE ATP-synthesizingsystem (F,-ATPase)
arsenate
aurovertin it ATP
Fig. I. Coupling of electron-transport and phosphorylation in mitochondria, and the sites of action of inhibitors and activators. CoQ, co-enzyme A: cyt. c, cytochrome c; DCCD, dicyclohexyl carbodiimide; HQNO, 2-heptyl-4-hydroxyquinoline-N-oxide; OSCP, oligomycin-sensitivity conferring protein: UBP. uncoupler binding protein (see Section VII).
molecules of ATP, and the molar ratio of esterified phosphate to consumed oxygen (P/O ratio) is maximally three. As seen in Fig. 1, the theoretical P/O ratio with succinate as substrate is 2. Uncouplers lower the P/O ratio by converting essentially all of the potential energy (-) into heat [4]. Fig. 1 also indicates that all partial reactions involving utilization of high energy are also subject to inhibition by uncouplers. These include, among others, ATP-32Pi exchange (reactions 17 and 18), ATP-driven reverse electron-flow from succinate to NAD (reactions 3, 2, 17, 14), respiration-driven reverse electron-flow from succinate to NAD (reactions 3, 2, 7, 9, 14), and ATP synthesis driven by an ion gradient (reactions 1.5, 18). Indeed, the existence of a common high-energy intermediate X - I (a precursor of the noncommittal -) was postulated to a large degree on the basis of the effects of uncouplers [17].
III. RESPIRATORY
CONTROL
In the presence of oxidizable substrate, oxygen, ADP and phosphate, the rates of respiration and phosphorylation are fast, and mitochondria are in the active state (state 3) [17]. In the controlled state (state 4), i.e., in the presence of substrate and oxygen, but in the absence of either ADP or phosphate, there is little respiration in tightly coupled mitochondria. The respiratory control ratio (RCR), the quotient of the rates in the active state and the controlled state, is a measure of the degree of
132 control imposed on oxidation by phosphorylation. Appropriate concentrations of uncouplers raise the rate of respiration from the state 4 to the state 3 level, and inhibit
the rate of phosphorylation
nearly
completely,
the uncoupled state (state 3~) [I 81. Respiratory control is not necessarily
an integral
rylation.
(light)
Aged mitochondria
[l9],
broken
bringing
the mitochondria
part of oxidative
mitochondria
to
phospho-
[20] and submito-
chondrial particles obtained by sonic disruption of mitochondria [2l] lack respiratory control nearly completely, but retain the full capacity of phosphorylation coupled to oxidation (loose coupling). Nevertheless, the release of respiratory control is a useful test for uncoupling agents. This is in spite of the fact that there are agents which release respiratory control without being uncouplers in the usual sense, (see Section IV), and the existence of membrane-impermeable uncouplers which are unable to affect respiratory control in intact mitochondria (see Section IX). The stimulation of an oligomycin-sensitive ATPase by uncouplers (reactions 17, 19) in intact mitochondria is in many respects similar to the release of respiratory control. This aspect of uncoupling, and the inhibitory effect of high concentrations of uncouplers on electron transport and ATPase activity [22] will not be discussed in this article.
IV. TYPES OF UNCOUPLING
Uncoupling in a broad sense may be brought about by a variety of agents or treatments which have nothing more in common than their effect on respiratory control. on mitochondrial ATPase and on the phosphorylation efficiency. For the study of the mechanism of uncoupling it is therefore necessary to classify uncoupling according to the agents and procedures which elicit this kind of response in mitochondria. 1. Strucfural uncoupfing. All procedures which impair the integrity of the inner mitochondrial membrane also decrease respiratory control, increase the ATPase activity, and often lower the phosphorylation efficiency. Such treatments include mechanical disruptions (freezing and thawing, shearing forces, sonication), aging [ 191, incubation with phospholipases [23], and addition of detergents [24]. 2. Uncoupling by cations and ionophorrs. Cations such as ethidium bromide [26] are known [25] and K +, in the presence of the antibiotic ionophore valinomycin to uncouple oxidative phosphorylation, presumably by the use of potential energy for ion transport in futile cycles (Fig. I, reaction 15). Under certain conditions, the antibiotics Dio-9 [27] and A 20668 [28] act as uncouplers, possibly by a similar mechanism, since they increase the proton permeability in liposomes [29] and mitochondria [28], respectively. The release of respiratory control by arsenate is oligomycin-sensitive [30] (see Fig. I ), in contrast to the effects of cations and ionophores. It appears that uncoupling by arsenate is due to direct, nonhydrolytic interference with ATP synthesis [31].
133 3. Uncoupling
involving
covalent
binding.
Alkylating
agents of the mustard
gas
type [32] and electrophiles
such as isothiocyanates [33] have been reported to have Although not much is known about the meuncoupling effects on mitochondria. chanism, the inhibitory effect of an alkylating uncoupler introduced by Wang [34]
demonstrates that alkylation of the membrane does not per se prompt a release of respiratory control. In mitochondria, uncoupling by CCCP and I, I ,3-tricyano-2-aminopropene is prevented, but not reversed by aminothiols [35]. In isolated bacterial membrane vesicles, however, uncoupling by CCCP is blocked as well as reversed by dithiothreitol, cysteine and other thiols [36]. Furthermore, high concentrations of CCCP diminish the reactivity of such vesicles toward N-ethyl maleimide, a powerful SHreagent. In contrast, none of the uncoupling effects of dinitrophenol is influenced by thiols. These and other data [37] indicate that carbonyl cyanide hydrazones may react in two different ways: as a sulfhydryl reagent, and as an uncoupler similar to dinitrophenol. It is, of course, a very interesting question whether reactions with certain SH-groups alone will result in uncoupling, as suggested by data obtained with 4-bromophenylisothiocyanate [33]. 4. Uncoupling by phenols and other anionic aromatic compounds. Many of the effects of the classical uncoupler 2,4-dinitrophenol were known, in a general way [l 11, long before uncoupling as a phenomenon associated with mitochondrial oxidative phosphorylation had been recognized by Lardy and Elvehjem [38] and demonstrated by Loomis and Lipmann [39]. The overall physiological effects of uncouplers are profound and dramatic [40]: “Dinitrophenol exerts a remarkable stimulating effect on fat metabolism, and the metabolism is sufficient to produce hyperthermia. Dinitrophenol has been tried extensively for clinical reduction of obesity; it is very effective. Unfortunately, its action is not always reliable and toxic manifestations, frequently with fatal results, appear unexpectedly”. The radical, biocidal effects of uncouplers have made them useful as herbicides and fungicides [4l]. Phenols are the largest group (about 40 different compounds have been described in the literature) in a class of uncouplers which includes other groups such as salicylanilides (S-6, S-1 3), carbonyl cyanide hydrazones (FCCP, CCCP), benzimiamong dazoles (TTFB), and similar heterocyclic systems. A common characteristic the groups in this class of uncouplers is a phenolic or anilinic configuration, electronegatively
0 /
\
-
substituted
XH
(X=
at the benzene
ring and on the nitrogen
in position
X:
NH,O,S)
-
Extensive qualitative and quantitative structure-function relationship studies have been published which stress the importance of lipophilicity and acidity for the ability to uncouple effectively [41l481. Nevertheless, these criteria do not fully define the effectiveness of uncouplers: 3,5-dibromo-4-cyanophenol is 6-9 times more
134 effective than 2,4_dinitrophenol, identical [44].
even though
their acidity and lipophilicity
are nearly
There is evidence that all uncouplers in this class of anionic aromates uncouple in the same or a very similar fashion, and the remainder of this article will concentrate on the discussion mechanism
of the properties
and interactions
of these compounds
and on the
by which they act as uncouplers.
V. INTERACTIONS
OF UNCOUPLERS
WITH
OTHER
INHIBITORS
In the controlled state of tightly coupled mitochondria (state 4), electron transport is inhibited, either as a result of direct interactions between components of the oxidative phosphorylation system, or indirectly through an electrochemical proton gradient (see Section X). Other, similar controls of electron transport by the energy conservation system are apparent in the effects of uncouplers on the potencies of respiratory inhibitors. These include (a) amytal and guanidine, (b) HQNO and (c) azide and hydroxylamine, which inhibit electron transport at phosphorylation sites I, II, and III, respectively (see Fig. 1). Chance and Hollunger [I81 showed that, in the presence of uncouplers, amytal and guanidine inhibit electron transport through site I much less efficiently than in the presence of ADP and phosphate (state 3). Similarly, Howland [49] demonstrated that concentrations of HQNO capable of inhibiting more than 90”/,, of state 3 respiration decreased the rate only by 50:;, in the presence of 2,4_dinitrophenol as uncoupler (state 3~). This corresponds to a difference by a factor of about 10 between the Ki values in state 3u and in state 3. As shown by Wilson and Chance [50], azide inhibits respiration in the presence of ADP and phosphate (state 3) or of Ca’+ much more than in their absence (state 4) or in the presence of uncouplers (state 3~). Uncouplers such as 2,4_dinitrophenol, pentachlorophenol, TTFB, and FCCP increase the apparent Ki of azide (0. I mM) by factors ranging from I .3 to about 10. Similar results have been obtained with hydroxylamine [.51,52], with the important by hydroxylamine, but difference that Ca’+ releases effectively the state 3 inhibition not by azide. Azide [53], and to a certain degree hydroxylamine at higher concentrations than those necessary for the inhibition
[52], are uncouplers of electron transport
(see Table I). These data show that the degree of energization in the energy conservation system not only influences electron transport per se (respiratory control) but also the effectiveness of respiratory inhibitors. In addition, it appears that these effects are not independent from the means by which deenergization is achieved, e.g., by ADP + phosphate, uncouplers or calcium uptake. Differential effects of this kind are very difficult to rationalize without invoking direct and multiple interactions between components of the oxidative phosphorylation system.
135 VI. INTERACTIONS OF AND WITH MEMBRANE
As expected philic structural
UNCOUPLERS COMPONENTS
from relatively characteristics,
lipids and biomembranes.
WITH
THE
small molecules uncouplers
interact
Several enzymes
MITOCHONDRIAL
with both hydrophobic
MEMBRANE
and hydro-
with a variety of soluble proteins,
such as kinases,
the soluble
F,-ATPase
and pyridine nucleotide-dependent dehydrogenases, which are all specific for adeninecontaining substrates, are inhibited by uncouplers [55-571. Serum albumin binds uncouplers strongly enough to reverse uncoupling of mitochondria completely [58]. The interactions of uncouplers with phospholipid liposomes and bilayers manifest themselves in increased swelling (decreased light-scattering) [59,60] and increased electric conductance [32,59-651, respectively. Apparently, these effects are brought about by complexes consisting of pairs of uncharged uncoupler acids and anions [64-661 present in the lipophilic phase, and also by uncoupler acids and anions imbedded in the polar-unpolar interphase of the phospholipid assembly [64,65,67,68]. There is often [32], but not always [60,62,63] a moderately good correlation between the ability of uncoupling agents to release respiratory control in mitochondria and the capacity to increase electric conductance in phospholipid bilayers. The correlation between the efficiencies of uncouplers to induce swelling in liposomes and to uncouple mitochondria is better [60] and appears to be more significant in view of similar parallelities between mitochondrial swelling and respiratory stimulation [28,69]. As seen in Fig. 2, the latter correlation accomodates not only un-
TBA
. . DOC.
TPB /
Fig. 2. Correlation between uncoupler induced swelling and respiratory stimulation in mitochondria. DNP, dinitrophenol; BTZ, SH-benzothiazole; SA, thiosalicylic acid; FCCP, carbonyl cyanide-p-trifluoromethoxyphenyl hydrazone; CL-CCP, carbonyl cyanide m-chlorophenyl hydrazone; 1799, 2,2’-bis (hexafluoroacetonyl )acetone: Dicum, dicumarol; 6847, SF-6847; TTFB, 2-trifluoromethyltetrachlorobenzimidazole: S13, 5-chloro-3-t-butyl-2’-chloro-4’-nitro-salicylanilide; PCP, pentachlorophenol; TPB, tetraphenyl boron; TBA, tributylamine; O.A., oleic acid; M.A., myristic acid; DOC, deoxycholate; ARS, arsenate. From Cunarro and Weiner [69].
136 couplers of the anionic aromatic type, but also detergents, amines and arsenate (see Section IV). The type of mitochondrial swelling used in these experiments is assumed to be due to an uncoupler-mediated shown control
in Fig. 2 has therefore
increase
in proton
been taken as evidence
permeability.
The correlation
that the release of respiratory
by uncouplers is the result of enhanced proton permeability. Early studies of uncoupler binding by mitochondria and mitochondrial
proteins
[70] have led Weinbach and Garbus to conclude that uncoupler-induced conformational changes in mitochondrial proteins are the basis of uncoupling [71]. It was already clear from their work that mitochondria can bind lipophilic uncouplers in amounts which are orders of magnitudes higher than the minimum necessary for uncoupling. Later studies by Wang et al. [72] and Bakker et al. [73] confirmed that, in direct binding studies, uncouplers such as CCCP, pentachlorophenol and TTFB appear to bind predominantly in a non-specific, partition-like manner. With the use of a new, largely hydrophilic uncoupler, 2-azido-4-nitrophenol Hanstein and Hatefi have demonstrated the existence of specific uncoupler binding in addition to unspecific partitioning [37] (Fig. 3). The specific, high-affinity un-
Fig.
3. Concentration
dependence
of uncoupler
equilibrium
binding
by mitochondria.
total binding; Curve B: specific binding (derived from Curve A by graphical NPA, 2-azido-4-nitrophenol. From Hanstein and Hatefi [37]
or computational
Curve
A:
means);
coupler binding site in mitochondria has many of the properties which one would expect from a component involved in uncoupling of oxidative phosphorylation : (I) It is specific for uncouplers of the anionic aromatic type. (2) Specific binding is not affected by other types of uncouplers or inhibitors such as ionophores and arsenate, or rotenone, antimycin, cyanide and oligomycin (see Fig. 1). (3) It is independent of the energy state of mitochondria, e.g.. the presence or absence of ATP or substrate,
137 and partial or full deenergization number of binding sites, about
by arsenate 0.6 nmol/mg
or valinomycin-K+ (& picrate). (4) The protein in beef heart mitochondria and
0.3-0.4 nmol/mg protein in rat liver mitochondria, is comparable to the concentration of other components of the oxidative phosphorylation system [74]. (5) The uncoupler binding site is apparently ubiquitous in mitochondria, e.g., beef heart [37], rat liver [54], yeast (Hanstein, Eugha
W. G. and
Griffiths,
D. E., unpublished
observations)
and
gracilis
mitochondria (Hanstein, W. G. and Kahn, J. S., unpublished observations), but not in other membrane systems such as erythrocyte ghosts [75]. Other important characteristics of specific uncoupler binding include the pH-independence of the dissociation constant, a Hill-slope of 1.0 and the almost entirely enthalpic nature of the free energy of binding, and suggest that uncouplers bind as single, anionic molecules in a non-cooperative fashion, without inducing net conformational changes. The experimental difficulties in determining the specific binding parameters of uncouplers which are much more lipophilic than 2-azido-4-nitrophenol and 2,4-dinitrophenol can be overcome in indirect, competitive binding experiments (see Fig. 5 in ref. 37), which allow the calculation of a competitive dissociation constant and of the extent of specific and unspecific binding [54]. This technique has made it possible to determine the dissociation constants of pentachlorophenol and S-13 in addition to azide and, in submitochondrial particles, to picrate. Table I shows dissociation constants of uncouplers obtained by direct or competitive binding studies at 3 “C, together with the ranges of concentrations (up) in which those uncouplers abolish 50 y(‘,of oxidative phosphorylation or respiratory control at 30 ‘C. This table also includes dissociation constants and ‘p,, values in terms of concentrations of uncouplers present in the mitochondrial phase. It is seen that, over a range of more than three orders of magnitude, there is a close correlation (r > 0.99) between the kinetic (y,,) and thermodynamic (I&) values. After application of the appropriate temperature corrections for specific binding ( 1H = -8 kcal and -4.6 kcal [54], for the aqueous
which applies
and the mitochondrial
to p,