Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

that contains heme c

and to which cytochrome c docks

(Figure 22-23). The ISP is likewise anchored by a single

transmembrane helix and extends into the intermembrane

space. Interestingly, the two ISPs of the dimeric complex

are domain swapped (intertwined) such that the [2Fe–2S]

cluster-containing domain of one protomer interacts with

the cytochrome b and cytochrome c

subunits of the other

protomer. The distances between the various metal cen-

ters are all quite large, ranging from 21 to 34 Å. The por-

tion of the complex that occupies the matrix, which ac-

counts for more than half the mass of cytochrome bc

consists largely of the structurally homologous Core 1 and

Core 2 proteins.

The route of electrons through the cytochrome bc

com-

plex is discussed in Section 22-3B, together with the mech-

anism by which the complex preserves the free energy of

electron transfer from CoQH

to cytochrome c for ATP

synthesis.

4. Cytochrome c

Cytochrome c, whose evolution we discussed in Section

7-3B, is a peripheral membrane protein of known crystal

structure (Figs. 8-42 and 9-41c) that is loosely bound to the

outer surface of the inner mitochondrial membrane. It

alternately binds to cytochrome c

(of Complex III) and to

cytochrome c oxidase (Complex IV) and thereby functions

to shuttle electrons between them.

The X-ray structure of yeast cytochrome bc

in complex

with cytochrome c,determined by Carola Hunte, reveals, as

expected, that cytochrome c binds to the cytochrome c

subunit of cytochrome bc

(Fig. 22-23). This association ap-

pears to be particularly tenuous because its interfacial area

(880 Å

) is significantly less than that exhibited by

protein–protein complexes known to have low stability

(typically ⬍1600 Å

). Such a small interface is well suited

for fast binding and release. The closest approach between

the heme groups of the contacting proteins is 4.1 Å

between atoms of their respective thioether-bonded sub-

stituents, and their Fe–Fe distance is 17.4 Å. This accounts

for the 8.3 ⫻ 10

⫺1

rate of electron transfer between these

two redox centers (see below).

a. Protein Structure Influences the Rate

of Electron Transfer

Reduced hemes are highly reactive entities; they can

transfer electrons over distances of 10 to 20 Å at physiolog-

ically significant rates. Hence cytochromes, in a sense, have

the opposite function of enzymes: Instead of persuading

unreactive substrates to react, they must prevent their

hemes from transferring electrons nonspecifically to other

cellular components. This, no doubt, is why these hemes

are almost entirely enveloped by protein. However,

cytochromes must also provide a path for electron transfer

to an appropriate partner.

In proteins, electron transfers occur between redox-

active cofactors such as hemes and/or Fe–S clusters. How-

ever, a survey of proteins of known structure that function

in electron transfer reveals that electrons travel no more

than 14 Å between protein-embedded redox centers and

that transfers over longer distances always involve chains

of redox-active cofactors (e.g., Figs. 22-19b and 22-22b).

Electron transfers occur far more efficiently through bonds

than through space via a quantum mechanical process

known as electron tunneling. Thus, as Harry Gray has ex-

perimentally demonstrated, electron tunneling within

proteins occurs largely through the polypeptide chains be-

tween bound redox-active groups and the rate of electron

transfer varies with the structure of the intervening

polypeptide. Moreover, tunneling across protein–protein

interfaces is largely mediated by van der Waals interactions

and water-bridged hydrogen bonds. Nevertheless, Leslie

Dutton has shown that the experimentally measured elec-

tron transfer rates within proteins vary only with the dis-

tance between the electron donor and the electron accep-

tor and fall off with an ⬃10-fold decrease in rate for each

1.7 Å increase in this distance.

5. Complex IV (Cytochrome c Oxidase)

Cytochrome c oxidase (COX or CcO), the terminal en-

zyme of the electron-transport chain, catalyzes the one-

electron oxidations of four consecutive reduced cytochrome

c molecules and the concomitant four-electron reduction of

one O

molecule to yield 2H

Eukaryotic COX is an ⬃200-kD transmembrane pro-

tein composed of 8 to 13 subunits, whose largest and most

hydrophobic chains, Subunits I, II, and III, are encoded by

mitochondrial genes, with the remaining subunits encoded

by nuclear genes. Eukaryotic COX exists in membranes as

a dimer. Subunits I and II of this complex contain all four

of its redox-active centers: two a-type hemes, a and a

, and

two Cu-containing centers, Cu

and Cu

. The Cu

center

and heme a are of low potential (0.245 and 0.210 V; Table

22-1), whereas Cu

and heme a

are of higher potential

(0.340 and 0.385 V). Spectroscopic studies indicate that

electrons are passed from cytochrome c to the Cu

center,

then to the heme a, and finally to a binuclear complex of

heme a

and Cu

binds to this binuclear complex and is

reduced to H

O in a complex four-electron reaction (see

below).

a. X-Ray Structures of Cytochrome c Oxidase

The X-ray structures of three species of cytochrome c

oxidase have been determined: two relatively simple

forms from the soil bacterium Paracoccus denitrificans

(1106 residues) by Michel and the purple photosynthetic

bacterium Rhodobacter sphaeroides by Iwata, and a

more complex form from bovine heart (1806 residues) by

Shinya Yoshikawa. Each protomer of the 2-fold symmetric

dimeric bovine COX has an ellipsoidal (potatolike)

shape comprised of a 48-Å-thick transmembrane portion

and hydrophilic portions that protrude 32 and 37 Å into

the mitochondrial matrix and the intermembrane space,

4 cytochrome c

3⫹

⫹ 2H

4 cytochrome c

2⫹

⫹ 4H

⫹

⫹ O

Section 22-2. Electron Transport 841

JWCL281_c22_823-870.qxd 3/19/10 11:00 AM Page 841

respectively (Fig. 22-24).These protomers each consist of 13

different subunits that mainly form 28 transmembrane

helices. The protomer surfaces that comprise the dimer in-

terface are concave (Fig. 22-24) and hence their relatively

tenuous contacts enclose a lipid-filled cavity. Moreover,

mitochondrial COX is active as a monomer.Thus a mecha-

nistic role for dimer formation seems unlikely. The P. deni-

trificans and Rb. sphaeroides COXs are monomeric com-

plexes, each of which consists of only 4 subunits that

collectively contain 22 transmembrane helices.

The structures of bovine COX Subunits I (12 transmem-

brane helices) and II (2 transmembrane helices), which

bind all of the complex’s four redox centers, are closely

similar to those of P. denitrificans and Rb. sphaeroides

COXs.These are the subunits that carry out the main func-

tions of the complex: transporting electrons from cy-

tochrome c to O

to yield water, while pumping protons

from inside (as we shall refer to the mitochondrial matrix

and the bacterial cytoplasm; also called the N-side due to

its negative charge) to outside (as we shall refer to the

mitochondrial intermembrane space and the bacterial

periplasmic space; also called the P-side due to its positive

charge). Bovine Subunit III (7 transmembrane helices),

whose structure also resembles that of its P. denitrificans

and Rb. sphaeroides counterparts, does not appear to di-

rectly participate in electron transfer or proton transloca-

tion. Indeed, a complex consisting of only P. denitrificans

Subunits I and II can actively transport electrons and pump

protons. Thus the function of Subunit III is unknown,

although there is some evidence that it facilitates the assem-

bly of Subunits I and II to form the active complex. Note,

however, that Subunit III does not contact Subunit II.

None of the 10 nuclear-encoded subunits of bovine COX

resemble Subunit IV of P.denitrificans and Rb. sphaeroides

COXs (1 transmembrane helix). Seven of these bovine

subunits each have one transmembrane helix, all oriented

with their N-terminal ends on the matrix side of the mem-

brane. These helices are distributed about the periphery of

the dimeric core formed by Subunits I, II, and III. The re-

maining three bovine subunits are globular and associate

entirely with the extramembrane portions of the complex.

The X-ray structure of bovine COX provides little indica-

tion of the function of any of its nuclear-encoded subunits.

Perhaps they have regulatory roles.

Subunit I binds heme a and the heme a

–Cu

binu-

clear center (Fig. 22-25), whose metal ions are all located

⬃13 Å below the membrane surface on its intermem-

brane/periplasmic side (Fig. 22-24).The heme a

Fe has one

axial His ligand, the heme a Fe has two axial His ligands

842 Chapter 22. Electron Transport and Oxidative Phosphorylation

Figure 22-24 X-ray structure of the bovine heart cytochrome c

oxidase homodimer. The view is perpendicular to its 2-fold axis

and parallel to the membrane with the matrix below. The 13

subunits in each protomer, which collectively have 28

transmembrane helices, are drawn in semitransparent ribbon

form and differently colored according to type with Subunit I

pale green, Subunit II pink, and Subunit III pale yellow. The

protein’s bound heme groups and Cu ions are drawn in

space-filling form with C green, N blue, O red, Fe red-brown, and

Cu cyan.The horizontal black lines delineate the inferred

position of the inner mitochondrial matrix. [Based on an X-ray

structure by Shinya Yoshikawa, Himeji Institute of Technology,

Hyogo, Japan. PDBid 1V54.]

See Interactive Exercise 18.

Figure 22-25 The redox centers in the X-ray structure of

bovine heart cytochrome c oxidase. The view is similar to that of

the left protomer in Fig. 22-24. The Fe and Cu ions are

represented by orange and cyan spheres. Their liganding heme

and protein groups (from subunit II for the Cu

center and

subunit I for the others) are drawn in stick form colored

according to atom type with heme C magenta, protein C green,

N blue, O red, and S yellow. The peroxy group that bridges the

and heme a

Fe ions is shown in ball-and-stick form in red.

Coordination bonds are drawn as thin gray lines. Note that the

side chains of His 240 and Tyr 244 are joined by a covalent bond

(lower right). [Based on an X-ray structure by Shinya Yoshikawa,

Himeji Institute of Technology, Hyogo, Japan. PDBid 2OCC.]

JWCL281_c22_823-870.qxd 10/19/10 8:07 AM Page 842

(as in Fig. 22-21b, top), and the Cu

atom has three His

ligands, whose coordinating N atoms are arranged in an

equilateral triangle that is centered on Cu

and is paral-

lel to heme a

. The X-ray structure of fully oxidized

bovine COX reveals a peroxide group (O

2⫺

; Fig. 22-25)

that bridges the heme a

Fe atom and Cu

(which are

separated by 4.9 Å), thereby liganding Cu

via a dis-

torted square-planar arrangement, a stable coordination

geometry for Cu(II). However, in the X-ray structure of

the fully reduced form of bovine COX (in which the

heme a

Fe¬Cu

distance is 5.2 Å), this ligand is absent,

so that Cu

is trigonally liganded, a stable coordination

geometry for Cu(I). The closest approach of the two

heme groups is 4 Å, and the distance between their Fe

atoms is 13.2 Å.

Subunit II, in addition to its two transmembrane helices,

has a globular domain on the outside surface that binds the

center and largely consists of a 10-stranded ␤ barrel

(Fig. 22-24). The Cu

center is located ⬃8 Å above the out-

side membrane surface. Although the Cu

center was, for

many years, widely believed to contain only one Cu atom,

the X-ray structures of COX clearly indicate that Cu

con-

tains two Cu atoms (Fig. 22-25). These are bridged by two

Cys S atom ligands and have two additional protein ligands

each to form an arrangement similar to that of a [2Fe–2S]

cluster (Fig. 22-15b) in which the two Cu atoms are 2.4 Å

apart. Spectroscopic measurements indicate that in the Cu

center’s reduced form, both of its Cu atoms are in their

Cu(I) states, whereas in its fully oxidized form, an electron

appears to be delocalized between the two Cu atoms such

that they assume the [Cu

1.5⫹

...

1.5⫹

] state.

b. Electron and Proton Acquisition

COX’s cytochrome c binding site is postulated to be in a

corner formed by the globular domain of Subunit II and

the outside surface of Subunit I, since this region is close to

the Cu

site and contains 10 acidic side chains that could

interact with a ring of Lys side chains that surrounds cy-

tochrome c’s heme crevice. Indeed, differential labeling of

cytochrome c oxidase’s carboxyl groups in the presence and

absence of cytochrome c demonstrated that cytochrome c

shields the invariant residues Asp 112, Glu 114, and Glu 198

of Subunit II (bovine numbering). Glu 198 is located be-

tween the two Cys residues of Subunit II that ligand Cu

(Fig. 22-25). This observation supports the spectroscopic

evidence that places the cytochrome c binding site on Sub-

unit II in close proximity to Cu

. Cross-linking studies

have additionally shown that the cytochrome c surface op-

posite the electron-transferring site interacts with Subunit

III, suggesting that Subunit III also participates in binding

cytochrome c.

Time-resolved spectroscopic studies indicate that an

electron obtained from cytochrome c is first acquired by

the Cu

center and is then transferred to heme a rather

than heme a

, probably because the shortest Cu

...

heme a

distance of 11.7 Å is less than the shortest Cu

...

heme a

distance of 14.7 Å. The electron is then rapidly transferred

to the heme a

–Cu

binuclear center, where it participates

in reducing the bound O

to H

O. Note that the fifth ligand

of heme a, His 378, is separated from the fifth ligand of

heme a

, His 376, by only one residue, and hence there is a

relatively short through-the-bonds electron transfer path-

way between heme a and a

(Fig. 22-25, lower left). In addi-

tion, the closest approach of the two hemes is 4 Å.

COX must acquire four so-called chemical or scalar

protons from the inside for every molecule of O

it reduces

to H

O.This four-electron process is coupled to the translo-

cation of up to four so-called pumped or vectorial protons

from the inside to the outside, thereby contributing to the

proton gradient that powers ATP synthesis (Section

22-3C). Note that for each turnover of the enzyme,

a total of eight positive charges are transported across the

membrane, thereby contributing to its membrane potential.

c. The Reduction of O

by Cytochrome c Oxidase

Occurs in Multiple Steps

The reduction of O

to 2H

O by cytochrome c oxidase

takes place on the cytochrome a

–Cu

binuclear complex

(Fig. 22-25). Indeed, a synthetic model of this binuclear

complex (Fig. 22-26), synthesized by James Collman, effi-

ciently catalyzes the reduction of O

to H

O when attached

to an electrode.

The COX-mediated reduction of O

requires, as we

shall see, the nearly simultaneous input of four electrons.

However, the fully reduced – binuclear complex

can readily contribute only three electrons to its bound O

in reaching its fully oxidized – state [cytochrome a

transiently assumes its Fe(IV) or ferryl oxidation state

during the reduction of O

; see below]. What is the source

of the fourth electron?

2⫹

4⫹

1⫹

2⫹

4 cyt c

3⫹

⫹ 2H

O ⫹ 4H

⫹

outside

⫹

inside

⫹ 4 cyt c

2⫹

⫹ O

Section 22-2. Electron Transport 843

2⫹

Figure 22-26 Synthetic model of the cytochrome a

–Cu

binuclear complex. This structure efficiently reduces O

to H

when attached to an electrode. The pyridine group that axially

ligands the Fe ion (bottom) can be replaced by an imidazole

group.

JWCL281_c22_823-870.qxd 6/8/10 9:19 AM Page 843

The X-ray structures of both bovine and P. denitrificans

COX clearly indicate that the His 240 ligand of Cu

(bovine numbering) is covalently cross-linked to the side

chain of the conserved Tyr 244 (Fig. 22-25, lower right).This

places Tyr 244’s phenolic ¬OH group in close proximity to

the heme a

-ligated O

such that Tyr 244 can supply the

fourth electron by transiently forming a tyrosyl radical

(TyrOⴢ). In fact, adding peroxide to the resting enzyme

generates a tyrosyl radical, whereas mutating Tyr 244 to

Phe inactivates the enzyme. Moreover, tyrosyl radicals

have been implicated in several enzymatically mediated re-

dox processes, including the generation of O

from H

O in

photosynthesis (in a sense, the reverse of the COX reac-

tion; Section 24-2Cd), and in the ribonucleotide reductase

reaction (which converts NDP to dNDP; Section 28-3Aa).

Tyr 244’s phenolic¬OH group is within hydrogen bonding

distance of the COX-bound O

and hence is a likely H

⫹

donor during O¬O bond cleavage. The formation of the

covalent cross-link is expected to lower both the reduction

potential and the pK of Tyr 244, thereby facilitating both

radical formation and proton donation (the synthetic binu-

clear complex in Fig. 22-26 can function without an associ-

ated tyrosyl radical, presumably because its associated

electrode can supply it with electrons much faster than cy-

tochrome c can supply them to COX).

The COX reaction, elucidated in large part by Mårten

Wikström and Gerald Babcock using a variety of spectro-

scopic techniques, involves four consecutive one-electron

transfers from the Cu

and cytochrome a sites and occurs

as follows (Fig. 22-27):

1 and 2. The oxidized binuclear complex [Fe(III)

¬OH

⫺

Cu(II)

] is reduced to its [Fe(II)

Cu(I)

] state by two con-

secutive one-electron transfers from cytochrome c via cy-

tochrome a and Cu

.A proton from the matrix is concomi-

tantly acquired and an H

O is released in this process. Tyr

244 (Y¬OH) is in its phenolic state.

3. O

binds to the reduced binuclear complex so as to

ligand its Fe(II)

atom. It binds to the heme with much the

same configuration it has in oxymyoglobin (Fig. 10-12).

4. Internal electron redistribution rapidly yields the

oxyferryl complex [Fe(IV) O

2⫺

⫺

¬Cu(II)] in which

Tyr 244 has donated an electron and a proton to the com-

plex and thereby assumed its neutral radical state (Y¬Oⴢ).

This is known as compound P because it was originally

thought that this spectroscopically identified state was a

peroxy complex. However, it has since been shown that a

peroxy compound is not on the reaction pathway. The per-

oxy complex displayed in Fig. 22-25 is a two-electron

reduced “mixed valence” state of the enzyme that cannot

reduce O

past its peroxy form.

5. A third one-electron transfer from cytochrome c to-

gether with the acquisition of two protons reconverts Tyr

244 to its phenolic state, yielding compound F (for ferryl)

and releasing an H

6. A fourth and final electron transfer and proton ac-

quisition yields the oxidized [Fe(III)

¬OH

⫺

Cu(II)

]

complex, thereby completing the catalytic cycle.

COX typically undergoes 100 to 200 turnovers per second

so that one catalytic cycle takes only a few milliseconds.

Note that the COX reaction proceeds without the release

of the destructive partially reduced reactive oxygen

species (ROS) from its active site. The positions in this

proposed catalytic cycle at which protons appear to be

“

844 Chapter 22. Electron Transport and Oxidative Phosphorylation

Figure 22-27 Proposed reaction sequence for the reduction of

by the cytochrome a

–Cu

binuclear complex of cytochrome

c oxidase. The numbered steps are discussed in the text.The

–

, H

⫹

–

, 2H

⫹

Cu(II)

–

OHFe(III)

HOY

oxidized

Cu(II)

⫺

Fe(IV)

–

Cu(II)

⫺

Fe(IV)

•

–

, H

⫹

–

Cu(I)O

Fe(II)

HOY

oxy

Cu(I)Fe(II)

HOY

reduced

entire reaction is extremely fast; it goes to completion in ⬃1 ms

at room temperature. [Modified from Babcock, G.T., Proc. Natl.

Acad. Sci. 96, 12971 (1999).]

JWCL281_c22_823-870.qxd 3/19/10 11:00 AM Page 844

pumped from the matrix (bacterial cytoplasm) to the in-

termembrane (periplasmic) space are discussed in Section

22-3B. Keep in mind, however,that aspects of this cycle are

uncertain and/or disputed and hence it remains under in-

vestigation.

d. Complexes I, III, and IV Form Supercomplexes

For many years, it had been widely assumed that the

complexes of the respiratory chain were laterally mobile

within the inner mitochondrial membrane and hence did

not associate. However, the development of gentle meth-

ods of separating the components of the inner mitochondr-

ial membrane has made it increasingly clear that these al-

ready large protein complexes form supercomplexes. In

fact, a variety of supercomplexes from several organisms

have been characterized, including yeast III

, bovine

and Arabadopsis thaliana (a plant) I

III

, and bovine

III

(in which, for example, IV

represents a protomer

of Complex IV). Electron microscopy–based images such

as Fig. 22-28 indicate that Complex IV monomers associate

with Complex III dimers, which limits the distance that cy-

tochrome c must travel to transport an electron from Com-

plex III to Complex IV to ⬍40 Å. Similarly, the association

of Complex III with Complex I reduces the distance over

which CoQH

must diffuse between these two complexes.

Apparently, these supermolecular complexes function to

increase the efficiency of electron transport via the chan-

neling of their intermediates, a conclusion supported by ki-

netic measurements. Note that none of these supercom-

plexes contain Complex II, also in agreement with kinetic

measurements.

3 OXIDATIVE PHOSPHORYLATION

The endergonic synthesis of ATP from ADP and P

in mito-

chondria, which, as we shall see, is catalyzed by proton-

translocating ATP synthase (Complex V), is driven by the

electron-transport process. Yet, since Complex V is physi-

cally distinct from the proteins mediating electron trans-

port (Complexes I–IV), the free energy released by electron

transport must be conserved in a form that ATP synthase

can utilize. Such energy conservation is referred to as en-

ergy coupling or energy transduction.

The physical characterization of energy coupling proved

to be surprisingly elusive; many sensible and often ingen-

ious ideas have failed to withstand the test of experimental

scrutiny. In this section we first examine some of the hy-

potheses that have been formulated to explain the cou-

pling of electron transport and ATP synthesis. We shall

then explore the coupling mechanism that has garnered

the most experimental support, analyze the mechanism by

which ATP is synthesized by ATP synthase, and, finally, dis-

cuss how electron transport and ATP synthesis can be

uncoupled.

A. Energy Coupling Hypotheses

In the more than 70 years that electron transport and ox-

idative phosphorylation have been studied, numerous

mechanisms have been proposed to explain how these

processes are coupled. In the following paragraphs, we ex-

amine the mechanisms that have received the greatest ex-

perimental attention.

1. The chemical coupling hypothesis. In 1953, Edward

Slater formulated the chemical coupling hypothesis, in

which he proposed that electron transport yielded reactive

intermediates whose subsequent breakdown drove oxida-

tive phosphorylation. We have seen, for example, that such

a mechanism is responsible for ATP synthesis in glycolysis

(Sections 17-2G and 17-2J). Thus, the exergonic oxidation

of glyceraldehyde-3-phosphate by NAD

⫹

yields 1,3-

bisphosphoglycerate, a reactive (“high-energy”) acyl phos-

phate whose phosphoryl group is then transferred to ADP

to form ATP in the phosphoglycerate kinase reaction. The

difficulty with such a mechanism for oxidative phosphoryla-

tion, which has caused it to be abandoned, is that despite in-

tensive efforts in numerous laboratories over many years,

no appropriate reactive intermediates have been identified.

2. The conformational coupling hypothesis. The con-

formational coupling hypothesis, which Paul Boyer formu-

lated in 1964, proposes that electron transport causes

proteins of the inner mitochondrial membrane to assume

“activated” or “energized” conformational states. These

proteins are somehow associated with ATP synthase such

that their relaxation back to the deactivated conformation

drives ATP synthesis. As with the chemical coupling hy-

pothesis, the conformational coupling hypothesis has

found little experimental support. However, conforma-

tional coupling of a different sort appears to be involved in

ATP synthesis (Section 22-3Cd).

Section 22-3. Oxidative Phosphorylation 845

Figure 22-28 Electron microscopy–based image of the bovine

supercomplex I

III

at 32 Å resolution. The supercomplex is

represented by its semitransparent surface (blue) fitted with the

EM-based structure of Complex I (yellow) and the X-ray

structures of the Complex III dimer (red) and the Complex IV

monomer (green).The view is parallel to the plane of the

membrane with the matrix above. [Courtesy of Eva Schäfer,

Birbeck College, London, U.K.]

JWCL281_c22_823-870.qxd 3/19/10 11:00 AM Page 845

3. The chemiosmotic hypothesis. The chemiosmotic hy-

pothesis, proposed in 1961 by Peter Mitchell, has spurred

considerable controversy, as well as much research, and is

now the model most consistent with the experimental evi-

dence. It postulates that the free energy of electron transport

is conserved by pumping H

⫹

from the mitochondrial matrix

to the intermembrane space so as to create an electrochemi-

cal H

⫹

gradient across the inner mitochondrial membrane.

The electrochemical potential of this gradient is harnessed to

synthesize ATP (Fig. 22-29).

Several key observations are explained by the chemios-

motic hypothesis:

(a) Oxidative phosphorylation requires an intact inner

mitochondrial membrane.

(b) The inner mitochondrial membrane is imperme-

able to ions such as H

⫹

,OH

⫺

⫹

, and Cl

⫺

, whose

free diffusion would discharge an electrochemical

gradient.

⫹

out

of intact mitochondria, thereby creating a measura-

ble electrochemical gradient across the inner mito-

chondrial membrane.

(d) Compounds that increase the permeability of the in-

ner mitochondrial membrane to protons, and thereby

dissipate the electrochemical gradient, allow electron

transport (from NADH and succinate oxidation) to

continue but inhibit ATP synthesis; that is, they “un-

couple” electron transport from oxidative phospho-

rylation. Conversely, increasing the acidity outside

the inner mitochondrial membrane stimulates ATP

synthesis.

In the remainder of this section we examine the mecha-

nisms through which electron transport can result in pro-

ton translocation and how an electrochemical gradient can

interact with ATP synthase to drive ATP synthesis.

B. Proton Gradient Generation

Electron transport, as we shall see, causes Complexes I, III,

and IV to transport protons across the inner mitochondrial

membrane from the matrix, a region of low [H

⫹

] and neg-

ative electrical potential, to the intermembrane space

(which is in contact with the cytosol), a region of high [H

⫹

]

and positive electrical potential (Fig. 22-14). The free en-

ergy sequestered by the resulting electrochemical gradi-

ent [which, in analogy to the term electromotive force

(emf), is called proton-motive force (pmf)] powers ATP

synthesis.

a. Proton Pumping Is an Endergonic Process

The free energy change of transporting a proton out of

the mitochondrion against an electrochemical gradient is

expressed by Eq. [20.3], which, in terms of pH, is

[22.1]

where Z is the charge on the proton (including sign), f is

the Faraday constant, and ⌬G is the membrane potential.

The sign convention for ⌬G is that when a positive ion is

transported from negative to positive, ⌬G is positive.

Since pH(out) is less than pH(in), the export of protons

from the mitochondrial matrix (against the proton gradi-

ent) is an endergonic process. In addition, proton trans-

port out of the matrix makes the inner membrane’s internal

surface more negative than its external surface. Outward

transport of a positive ion is consequently associated with

a positive ⌬G and an increase in free energy (endergonic

process), whereas the outward transport of a negative ion

yields the opposite result. Clearly, it is always necessary to

describe membrane polarity when specifying a membrane

potential.

The measured membrane potential across the inner

membrane of a liver mitochondrion, for example,is 0.168 V

¢G ⫽ 2.3 RT[pH(in) ⫺ pH(out)] ⫹ Z f ¢⌿

846 Chapter 22. Electron Transport and Oxidative Phosphorylation

Figure 22-29 Coupling of electron transport (green arrow) and

ATP synthesis. H

⫹

is pumped out of the mitochondrion by

Complexes I, III, and IV of the electron-transport chain (blue

arrows), thereby generating an electrochemical gradient across

the inner mitochondrial membrane. The exergonic return of these

Intermembrane

space

]Low

]High

–

––

–

ATP

ADP

+ P

–

III IV

Outer mitochondrial

membrane

Inner mitochondrial

membrane

protons to the matrix powers the synthesis of ATP (red arrow).

Note that the outer mitochondrial membrane is permeable to

small molecules and ions, including H

⫹

. See the Animated

Figures

JWCL281_c22_823-870.qxd 3/19/10 11:00 AM Page 846

(inside negative; this corresponds to an ⬃210,000 V ⴢ cm

⫺1

electric field across its ⬃80 Å thickness).The pH of the ma-

trix is 0.75 units higher than that of the intermembrane

space. ⌬G for proton transport out of this mitochondrial

matrix is therefore 21.5 kJ ⴢ mol

⫺1

b. The Passage of About Three Protons Is

Required to Synthesize One ATP

An ATP molecule’s estimated physiological free en-

ergy of synthesis, around ⫹40 to ⫹50 kJ ⴢ mol

⫺1

, is too

large to be driven by the passage of a single proton back

into the mitochondrial matrix; at least two protons are re-

quired. This number is difficult to measure precisely, in

part because transported protons tend to leak back across

the mitochondrial membrane. However, most estimates

indicate that around three protons are passed per ATP

synthesized.

c. Two Mechanisms of Proton Transport

Have Been Proposed

Three of the four electron-transport complexes, Com-

plexes I, III, and IV, are involved in proton translocation.

Two mechanisms have been entertained that would couple

the free energy of electron transport with the active trans-

port of protons: the redox loop mechanism and the proton

pump mechanism.

d. The Redox Loop Mechanism

This mechanism, proposed by Mitchell, requires that the

redox centers of the respiratory chain (FMN, CoQ, cy-

tochromes, and iron–sulfur clusters) be so arranged in the

membrane that reduction would involve a redox center si-

multaneously accepting e

⫺

and H

⫹

from the matrix side of

the membrane. Reoxidation of this redox center by the

next center in the chain would involve release of H

⫹

on the

cytosolic side of the membrane together with the transfer

of electrons back to the matrix side (Fig. 22-30). Electron

flow from one center to the next would therefore yield net

translocation of H

⫹

and the creation of an electrochemical

gradient (⌬G and ⌬pH).

The redox loop mechanism requires that the first redox

carrier contain more hydrogen atoms in its reduced state

than in its oxidized state and that the second redox carrier

have no difference in its hydrogen atom content between its

reduced and oxidized states. Are these requirements met in

the electron-transport chain? Some of the redox carriers,

FMN and CoQ, in fact, contain more hydrogen atoms in

their reduced state than in their oxidized state and thus can

qualify as proton carriers as well as electron carriers. If

these centers were spatially alternated with pure electron

carriers (cytochromes and iron–sulfur clusters), such a

mechanism could well be accommodated.

The main difficulty with the redox loop mechanism in-

volves the deficiency of (H

⫹

⫹ e

⫺

) carriers that can alter-

nate with pure e

⫺

carriers. Whereas the electron-transport

chain has as many as 15 pure e

⫺

carriers (up to 8 iron–sulfur

proteins, 5 cytochromes, and 2 Cu centers), it has only 2

⫹

⫹ e

⫺

) carriers. The fact that there are three complexes

with standard reduction potential changes large enough to

provide free energy for ATP synthesis suggests the need

for at least three proton-transport redox carriers. As we

shall see, however, there are, in fact, three proton-transport

sites but only two proton-transport redox carriers: Both the

redox loop mechanism and the proton pump mechanism

(discussed below) are employed.

e. Complex III Pumps Protons via the Q Cycle, a

Type of Redox Loop

See Guided Exploration 20: The Q cycle Mitchell postulated

that Complex III functions in a way that permits one mole-

cule of CoQH

, the two-electron carrier, to sequentially re-

duce two molecules of cytochrome c, a one-electron carrier,

while transporting four protons. This occurs via a modified

redox loop mechanism involving a remarkable bifurcation

of the flow of electrons from CoQH

to cytochrome c

and

to cytochrome b. It is through this so-called Q cycle that

Complex III pumps protons from the matrix to the inter-

membrane space.

The essence of the Q cycle is that CoQH

undergoes a

two-cycle reoxidation in which the semiquinone is a sta-

ble intermediate. This involves two independent binding

sites for coenzyme Q: Q

, which binds CoQH

and is lo-

cated between the ISP and heme b

in proximity to the in-

termembrane space (Fig. 22-23); and Q

, which binds both

and CoQ and is located near heme b

in proximity toQ

⫺

ⴢ

⫺

ⴢ

Section 22-3. Oxidative Phosphorylation 847

Intermembrane

space

Inner

membrane

Matrix

SSH

–

red

Figure 22-30 The redox loop mechanism for electron transport–

linked H

ⴙ

translocation. AH

represents (H

⫹

⫹ e

⫺

) carriers such

as FMNH

and CoQH

, whereas B represents pure e

⫺

carriers

such as iron–sulfur clusters and the cytochromes. These compo-

nents are so arranged as to require that electron transport be

accompanied by H

⫹

translocation.

JWCL281_c22_823-870.qxd 3/19/10 11:00 AM Page 847

the matrix. In the first cycle (Fig. 22-31a), CoQH

, which is

supplied by Complexes I or II on the matrix side of the in-

ner mitochondrial membrane (1), diffuses through the

membrane to its cytoplasmic side, where it binds to the Q

site (2).There it transfers one of its electrons to the ISP (3),

releasing its two protons into the intermembrane space

and yielding . The ISP then reduces cytochrome c

whereas the transfers its remaining electron to heme b

(4), yielding fully oxidized CoQ. Heme b

then reduces

heme b

(6).The CoQ from Step 4 is released from the Q

site and diffuses back through the membrane to rebind at

the Q

site (5), where it picks up the electron from heme

(7), reverting to the semiquinone form, .Thus, the re-

action for this first cycle is

In the second cycle (Fig. 22-31b), another CoQH

re-

peats Steps 1 through 6: One electron reduces the ISP and

then cytochrome c

, and the other electron sequentially re-

duces heme b

and then heme b

. This second electron

then reduces the at the Q

site produced in the first cy-

cle (8), yielding CoQH

. The protons taken up in this last

step originate in the mitochondrial matrix.The reaction for

the second cycle is therefore

For every two CoQH

that enter the Q cycle, one

CoQH

is regenerated. The combination of both cycles, in

which two electrons are transferred from CoQH

to cy-

tochrome c

, results in the overall reaction

X-ray studies of Complex III provide direct evidence

for the independent existence of the Q

and Q

sites. The

antifungal agents myxothiazol and stigmatellin,

CoQ ⫹ 2 cytochrome c

(Fe

2⫹

) ⫹ 4H

⫹

(outside)

CoQH

⫹ 2 cytochrome c

(Fe

3⫹

) ⫹ 2H

⫹

(matrix)

CoQ ⫹ CoQH

⫹ cytochrome c

(Fe

2⫹

) ⫹ 2H

⫹

(outside)

CoQH

⫹ Q

⫺

ⴢ ⫹ cytochrome c

(Fe

3⫹

)⫹ 2H

⫹

(matrix)

⫺

ⴢ

⫺

ⴢ ⫹ cytochrome c

(Fe

2⫹

) ⫹ 2H

⫹

(outside)

CoQH

⫹ cytochrome c

(Fe

3⫹

)

⫺

ⴢ

⫺

ⴢ

⫺

ⴢ

848 Chapter 22. Electron Transport and Oxidative Phosphorylation

Figure 22-31 The Q cycle. The Q cycle is an electron-

transport cycle in Complex III that accounts for H

⫹

translocation

during the transport of electrons from cytochrome b to

cytochrome c: The overall cycle is actually two cycles, the first

(a) requiring reactions 1 through 7 and the second (b) requiring

reactions 1 through 6 and 8. (1) Coenzyme QH

is supplied by

Complex I on the matrix side of the membrane. (2) QH

diffuses

to the outside of the membrane. (3) QH

reduces the Rieske

iron–sulfur protein (ISP) forming semiquinone and releasing

⫹

.The ISP goes on to reduce cytochrome c

. (4) reduces

heme b

to form coenzyme Q. (5) Q diffuses to the matrix side.

(6) Heme b

reduces heme b

.(7, cycle 1 only) Q is reduced to

by heme b

.(8, cycle 2 only) is reduced to QH

by heme

. [After Trumpower, B.L., J. Biol. Chem. 265, 11410 (1990).]

⫺

ⴢQ

⫺

ⴢ

⫺

ⴢ

⫺

ⴢ

ISP

Inter-

membrane

space

Cycle 1

Matrix

–

From

complex I

ISP

–

Cycle 2

–

From

complex I

–

(b)

(a)

OCH

Myxothiazol

Stigmatellin

OCH

OH CH

JWCL281_c22_823-870.qxd 6/8/10 9:19 AM Page 848

which both block electron flow from CoQH

to the ISP and

to heme b

(Steps 3 and 4 of both cycles), bind in a pocket

within cytochrome b between the ISP and heme b

(Fig.

22-23). Evidently, this binding pocket overlaps the Q

site.

Similarly, antimycin (Section 22-2Ba), which blocks elec-

tron flow from heme b

to CoQ and (Step 7 of Cycle 1

and Step 8 of Cycle 2), binds in a pocket near heme b

thereby identifying this pocket as site Q

The circuitous route of electron transfer in Complex III

is tied to the ability of coenzyme Q to diffuse within the

hydrophobic core of the membrane in order to bind to

both the Q

and Q

sites.This process is facilitated by an in-

dentation in the surface of cytochrome b’s transmembrane

region that contains Q

from one protomer and Q

from

the other. When CoQH

is oxidized, two reduced cy-

tochrome c molecules and four protons appear on the outer

side of the membrane. Proton transport by the Q cycle thus

follows the redox loop mechanism of proton transport in

which a redox center itself (CoQ) is the proton carrier.As

we shall see below, however, Complexes I and IV follow a

different mechanism of proton transport, the proton pump

mechanism.

f. The Bifurcation of the Q Cycle’s Electron Flow

Occurs via Domain Movement

Why does Q

-bound exclusively reduce heme b

(Step 4 of the Q cycle) rather than the Rieske [2Fe–2S]

cluster of the ISP, despite the greater reduction potential

difference (⌬e) favoring the latter reaction (Table 22-1)?

⫺

ⴢ

⫺

ⴢ

The remarkable answer to this question provides a fasci-

nating insight into the inner workings of Complex III. Al-

though the binding of stigmatellin and myxothiazol to Q

are mutually exclusive, these inhibitors affect this site dif-

ferently: Stigmatellin perturbs the spectrum and redox

properties of the ISP’s Rieske [2Fe–2S] cluster as well as

prevents it from oxidizing cytochrome c

, whereas myxoth-

iazol does not interact with the ISP but, instead, shifts the

spectrum of heme b

. Evidently, stigmatellin is a CoQH

analog and myxothiazol is a mimic.

X-ray structures of Complex III reveal that its Q

site is

a bifurcated pocket in which stigmatellin binds close to the

ISP docking interface (see below), whereas myxothiazol

binds in the vicinity of heme b

(Fig. 22-32; their binding

to Q

is mutually exclusive because their hydrophobic

tails would overlap). Furthermore, the globular Rieske

[2Fe–2S] cluster–containing domain of the ISP (Fig. 22-23,

top) is conformationally mobile and assumes a conforma-

tional state that is controlled by the ligand-binding state of

: It binds to cytochrome b near the heme b

site when

stigmatellin is bound to the Q

site, but has swung around

by ⬃20 Å (via an ⬃57º hinge motion that leaves intact its

tertiary structure) to bind to the cytochrome c

near its

heme c when myxothiazol is bound to Q

. Apparently the

globular domain of the ISP functions to shuttle an electron

from CoQH

, which is bound to Q

near the ISP docking

interface, to heme c

by mechanically swinging the reduced

Rieske [2Fe–2S] cluster between these sites. The resulting

shifts to the position near heme b

(probably via aQ

⫺

ⴢ

⫺

ⴢ

Section 22-3. Oxidative Phosphorylation 849

Figure 22-32 X-ray structures of the Q

binding site of the

chicken cytochrome bc

complex occupied by inhibitors. The

structures show (a) its complex with stigmatellin and (b) its

complex with myxothiazol.The protein surface (white) has been

cut away to show the Q

pocket. Heme b

(upper right) is shown

in ball-and-stick form with C gray, N blue, O red, and Fe tan. In

Part a, stigmatellin is drawn as a stick model with C yellow and

its volume represented by the dotted surface. The Rieske [2Fe–2S]

cluster (lower left) is represented by gold spheres, and the ISP

domain to which it is bound is drawn as a cyan ribbon. Note that

His 161, which is a ligand of the Rieske [2Fe–2S] cluster, forms a

hydrogen bond to stigmatellin. In Part b, the myxothiazol is

drawn as a stick model with C orange. Note that the semiquinone-

mimicking portion of myxothiazol binds to Q

in proximity to

heme b

, whereas its hydrophobic tail occupies the same position

as that of stigmatellin.Also note that the [2Fe–2S] cluster–

containing ISP domain is not visible in this diagram; it has

rotated into proximity with cytochrome c

. [Courtesy of Antony

Crofts, University of Illinois at Urbana–Champagne, and Edward

Berry, University of California at Berkeley. PDBid 3BCC.]

(a)

(b)

JWCL281_c22_823-870.qxd 3/19/10 11:00 AM Page 849

rotation about a bond connecting the semiquinone ring to its

nonpolar tail), which it then reduces. Thus, is unable to

reduce the ISP (after it has reduced cytochrome c

) be-

cause it is too far away to do so. This novel mechanism is

supported by the observation that mutagenically inserting

a disulfide bond either within the hinge of the ISP or be-

tween it and cytochrome b greatly diminishes the activity

of Complex III but that its activity is restored on exposure

to reducing agents.

g. The Proton Pump Mechanism

Complex IV (COX) transports four protons from the

matrix to the intermembrane space for each O

it reduces

(2H

⫹

per electron pair; Fig. 22-14). It contains no (H

⫹

⫹ e

⫺

)

carriers and hence cannot do so via a redox loop (Q cycle-

like) system. Rather, as we shall see, it does so via the pro-

ton pump mechanism (Fig. 22-33), which does not require

that the redox centers themselves be H

⫹

carriers. In this

model, the transfer of electrons results in conformational

changes to the complex. The unidirectional translocation of

protons occurs as a result of the influence of these confor-

mational changes on the pK’s of amino acid side chains and

their alternate exposure to the internal and external side of

the membrane. We have previously seen that conformation

can influence pK. The Bohr effect in hemoglobin, for ex-

ample, results from conformational changes induced by O

binding, which causes pK changes in protein acid–base

groups (Section 10-2E). If such a protein were located in a

membrane and if, in addition to pK changes, the conforma-

tional changes altered the side of the membrane to which

the affected amino acid side chains were exposed, the re-

sult would be H

⫹

transport and the system would be a

proton pump.

Keep in mind that protons, being atomic nuclei, must al-

ways be associated with molecules or ions. Consequently, a

proton cannot be transported across a membrane in the

same way that, say, a K

⫹

ion is. Rather, protons are translo-

cated by hopping along chains of hydrogen bonded groups

⫺

ⴢ

in the transport protein in much the same way that hydro-

nium ions migrate through an aqueous solution (Fig. 2-10),

that is, they move along a proton wire (Section 21-2Al).

However, unlike a wire in an electrical circuit, all elements

of a proton wire need not be connected at the same time

and, moreover, internal water molecules, which are not al-

ways apparent in protein X-ray structures, are likely to be

integral parts of the proton wire. Hence, elucidating the

precise pathway of proton transport through a protein is a

difficult and uncertain task. Note that proton pumps, like

other active transport systems, must be gated to prevent

protons from leaking back through the pump and thus

short-circuiting the system.

h. Bacteriorhodopsin Is a Light-Driven Proton Pump

The simplest known and best characterized proton

pump is the intrinsic membrane protein bacteriorhodopsin

of Halobacterium halobium. It consists mainly of seven

transmembrane helices, A through G, that form a central

polar channel (Section 12-3Ab). This channel contains a

retinal prosthetic group that is covalently linked, via a pro-

tonated Schiff base, to Lys 216, which extends from helix G

(Fig. 12-24). The protein obtains the free energy required

for unidirectionally pumping protons from the absorption

of a photon by the retinal. This initiates a sequence of

events in which the protein conformationally adjusts

through successive spectroscopically characterized inter-

mediates, designated J, K, L, M, N, and O, as the system de-

cays back to its ground state over a period of ⬃10 ms. The

result of this cycle is the net translocation of one proton

from the cytoplasm to the extracellular medium, thereby

converting light energy to proton-motive force. The as yet

incompletely understood mechanism of this process, which

was elucidated through detailed structural, mutational, and

time-resolved spectroscopic studies carried out in several

laboratories, is outlined in Fig. 22-34:

1. On absorbing a photon, the ground state all-trans-

retinal photoisomerizes to its 13-cis form. This is a multi-

step process that rapidly passes (in ⬃3 ps) through the J

and K states. The free end of the retinal, which is now

twisted around the newly cis double bond, moves relative

to the protein scaffold such that retinal’s C13 methyl group

and its C14 atom shift toward the inside by 1.3 and 1.7 Å,

respectively. This yields the L state.

2. Further conformational adjustments yield the M

state. Here the N atom of the Schiff base has rotated and

shifted from its ground state position, in which it is hydro-

gen bonded to an internal water molecule, to one in which

it points toward the inside face of the protein in the vicinity

of the hydrophobic side chains of Val 49 and Leu 93. This

reduces the pK of the protonated Schiff base. In contrast,

the pK of Asp 85 increases. This is because, in the ground

state,Asp 85 effectively serves as the counterion of the pro-

tonated Schiff base and participates in a hydrogen bonded

network with three internal water molecules, but in the M

state, it is only associated with a single water molecule.

Consequently, the Schiff base protonates Asp 85. This

process is facilitated by a slight movement of helix C that

850 Chapter 22. Electron Transport and Oxidative Phosphorylation

Figure 22-33 Proton pump mechanism of electron

transport–linked proton translocation. At each H

⫹

translocation

site, n protons bind to amino acid side chains on the matrix side

(inside) of the membrane. Reduction causes a conformational

change that decreases the pK’s of these side chains and exposes

them to the cytosolic side (outside) of the membrane, where the

protons dissociate. Reoxidation results in a conformational

change that restores the pump to its original conformation.

–

Intermembrane

space

Matrix

reduction

reoxidation

JWCL281_c22_823-870.qxd 6/8/10 9:19 AM Page 850