19.

2 ATP Synthesis 747

Table 19–4), that transfers electrons from NADH in alternative respiratory pathway through the rote-
the matrix directly to ubiquinone, bypassing Com- none-insensitive NADH dehydrogenase, the external
plex I and its associated proton pumping. And plant NADH dehydrogenase, or succinate dehydrogenase
mitochondria have yet another NADH dehydroge- (Complex II), and pass to O2 via the cyanide-resistant
nase, on the external face of the inner membrane, alternative oxidase, energy is not conserved as ATP
that transfers electrons from NADPH or NADH in but is released as heat. A skunk cabbage can use the
the intermembrane space to ubiquinone, again heat to melt snow, produce a foul stench, or attract
bypassing Complex I. Thus when electrons enter the beetles or flies.

Alternative NAD(P)+ External NAD(P)H

NADH dehydrogenase Cyt c
Intermembrane dehydrogenase
space (P side) NAD(P)H

I Q III IV

Matrix (N side)
NADH NAD+ 1– O
H2O
2 2
Alternative Heat
oxidase

FIGURE 2 Electron carriers of the inner membrane of plant mito- animal mitochondria, or through plant-specific alternative carriers
chondria. Electrons can flow through Complexes I, III, and IV, as in by the paths shown with blue arrows.

which transfers the electrons to cytochrome b, the

first carrier in Complex III. In this complex,
19.2 ATP Synthesis
electrons take two separate paths through two How is a concentration gradient of protons transformed
b-type cytochromes and cytochrome c1 to an Fe-S into ATP? We have seen that electron transfer releases,
center. The Fe-S center passes electrons, one at a and the proton-motive force conserves, more than
time, through cytochrome c and into Complex IV, enough free energy (about 200 kJ) per “mole” of elec-
cytochrome oxidase. This copper-containing tron pairs to drive the formation of a mole of ATP,
enzyme, which also contains cytochromes a and which requires about 50 kJ (p. 519). Mitochondrial oxi-
a3, accumulates electrons, then passes them to O2, dative phosphorylation therefore poses no thermody-
reducing it to H2O. namic problem. But what is the chemical mechanism
 Some electrons enter this chain of carriers through that couples proton flux with phosphorylation?
alternative paths. Succinate is oxidized by The chemiosmotic model, proposed by Peter
succinate dehydrogenase (Complex II), which Mitchell, is the paradigm for this mechanism. Accord-
contains a flavoprotein that passes electrons ing to the model (Fig. 19–19), the electrochemical
through several Fe-S centers to ubiquinone. energy inherent in the difference in proton concentra-
Electrons derived from the oxidation of fatty acids tion and the separation of charge across the inner
pass to ubiquinone via the electron-transferring mitochondrial membrane—the proton-motive force—
flavoprotein. drives the synthesis of ATP as protons flow passively
back into the matrix through a proton pore in ATP
 Potentially harmful reactive oxygen species synthase. To emphasize this crucial role of the pro-
produced in mitochondria are inactivated by a set ton-motive force, the equation for ATP synthesis is
of protective enzymes, including superoxide sometimes written
dismutase and glutathione peroxidase.
ADP 1 Pi 1 nHP1 ¡
 Plants, fungi, and unicellular eukaryotes have, in
ATP 1 H2O 1 nHN
1
(19–10)
addition to the typical cyanide-sensitive path for
electron transfer, an alternative, cyanide-resistant Mitchell used “chemiosmotic” to describe enzymatic reac-
NADH oxidation pathway. tions that involve, simultaneously, a chemical reaction
748 Oxidative Phosphorylation and Photophosphorylation

Cyt c
Intermembrane
space (P side) 4H+ 2H+

QH2 II QH2 III IV Fo

Q

2H+
Matrix
(N side) 2H+ + 21 O2 H2 O ADP
nH+ F1 + Pi
NAD+ Succinate Fumarate ATP
2H+
NADH + H+
Proton-motive force
Chemical potential Electrical potential ATP synthesis driven
∆pH + ∆w by proton-motive force
(inside alkaline) (inside negative)

FIGURE 19–19 Chemiosmotic model. In this simple representation of the and an electrical gradient (Dc) (combined, the proton-motive force). The
chemiosmotic theory applied to mitochondria, electrons from NADH and inner mitochondrial membrane is impermeable to protons; protons can
other oxidizable substrates pass through a chain of carriers arranged asym- reenter the matrix only through proton-specific channels (Fo). The proton-
metrically in the inner membrane. Electron flow is accompanied by proton motive force that drives protons back into the matrix provides the energy for
transfer across the membrane, producing both a chemical gradient (DpH) ATP synthesis, catalyzed by the F1 complex associated with Fo.

and a transport process, and tions, no ATP synthesis can

the overall process is some- occur and electron transfer to
times referred to as “chemios- O2 does not proceed. Henry
motic coupling.” Coupling Lardy, who pioneered the use
refers to the obligate connec- of antibiotics to explore mito-
tion between mitochondrial chondrial function, demon-
ATP synthesis and electron strated coupling of oxidation
flow through the respiratory and phosphorylation by using
chain; neither of the two pro- oligomycin and venturicidin,
cesses can proceed without the toxic antibiotics that bind to
Henry Lardy, 1917–2010
Peter Mitchell,
other. The operational defini- the ATP synthase in mito-
1920–1992 tion of coupling is shown in chondria. These compounds are potent inhibitors of
Figure 19–20. When isolated both ATP synthesis and the transfer of electrons
mitochondria are suspended in a buffer containing through the chain of carriers to O2 (Fig. 19–20b).
ADP, Pi, and an oxidizable substrate such as succinate, Because oligomycin is known to interact not directly
three easily measured processes occur: (1) the sub- with the electron carriers but with ATP synthase, it fol-
strate is oxidized (succinate yields fumarate), (2) O2 is lows that electron transfer and ATP synthesis are obli-
consumed, and (3) ATP is synthesized. Oxygen con- gately coupled: neither reaction occurs without the
sumption and ATP synthesis depend on the presence other.
of an oxidizable substrate (succinate in this case) as Chemiosmotic theory readily explains the depen-
well as ADP and Pi. dence of electron transfer on ATP synthesis in mito-
Because the energy of substrate oxidation drives chondria. When the flow of protons into the matrix
ATP synthesis in mitochondria, we would expect inhib- through the proton channel of ATP synthase is
itors of the passage of electrons to O2 (such as cyanide, blocked (with oligomycin, for example), no path
carbon monoxide, and antimycin A) to block ATP syn- exists for the return of protons to the matrix, and the
thesis (Fig. 19–20a). More surprising is the finding that continued extrusion of protons driven by the activity
the converse is also true: inhibition of ATP synthesis of the respiratory chain generates a large proton gra-
blocks electron transfer in intact mitochondria. This dient. The proton-motive force builds up until the
obligatory coupling can be demonstrated in isolated cost (free energy) of pumping protons out of the
mitochondria by providing O2 and oxidizable sub- matrix against this gradient equals or exceeds the
strates, but not ADP (Fig. 19–20b). Under these condi- energy released by the transfer of electrons from
 19.2 ATP Synthesis 749

Add Add Add

CN venturicidin DNP
or Uncoupled
oligomycin

ATP synthesized

ATP synthesized
O2 consumed

O2 consumed
Add Add
succinate ADP ⫹ Pi
Add Add
ADP ⫹ Pi succinate

(a) Time (b) Time

FIGURE 19–20 Coupling of electron transfer and ATP synthesis in mito- sized. Addition of cyanide (CN), which blocks electron transfer between
chondria. In experiments to demonstrate coupling, mitochondria are sus- cytochrome oxidase (Complex IV) and O2, inhibits both respiration and
pended in a buffered medium and an O2 electrode monitors O2 consump- ATP synthesis. (b) Mitochondria provided with succinate respire and
tion. At intervals, samples are removed and assayed for the presence of synthesize ATP only when ADP and Pi are added. Subsequent addition of
ATP. (a) Addition of ADP and Pi alone results in little or no increase in venturicidin or oligomycin, inhibitors of ATP synthase, blocks both ATP
either respiration (O2 consumption; black) or ATP synthesis (red). When synthesis and respiration. Dinitrophenol (DNP) is an uncoupler, allowing
succinate is added, respiration begins immediately and ATP is synthe- respiration to continue without ATP synthesis.

NADH to O2. At this point electron flow must stop; OH O

the free energy for the overall process of electron NO2 NO2
flow coupled to proton pumping becomes zero, and ⫹ H⫹
the system is at equilibrium.
Certain conditions and reagents, however, can
NO2 NO2
uncouple oxidation from phosphorylation. When intact
2,4-Dinitrophenol
mitochondria are disrupted by treatment with deter- (DNP)
gent or by physical shear, the resulting membrane
fragments can still catalyze electron transfer from suc-
N N N N
cinate or NADH to O2, but no ATP synthesis is coupled C C C C
to this respiration. Certain chemical compounds cause C C
uncoupling without disrupting mitochondrial struc- N N
ture. Chemical uncouplers include 2,4-dinitrophenol
NH N
(DNP) and carbonylcyanide-p-trifluoromethoxy-
phenylhydrazone (FCCP) (Table 19–4; Fig. 19–21),
weak acids with hydrophobic properties that permit ⫹ H⫹
them to diffuse readily across mitochondrial mem-
branes. After entering the matrix in the protonated O O
form, they can release a proton, thus dissipating the F C F F C F
proton gradient. Resonance stabilization delocalizes
F F
the charge on the anionic forms, making them suffi-
Carbonylcyanide-p-
ciently permeant to diffuse back across the membrane, trifluoromethoxyphenylhydrazone
where they can pick up a proton and repeat the pro- (FCCP)
cess. Ionophores such as valinomycin (see Fig. 11–44)
allow inorganic ions to pass easily through membranes. FIGURE 19–21 Two chemical uncouplers of oxidative phosphorylation.
Both DNP and FCCP have a dissociable proton and are very hydrophobic.
Ionophores uncouple electron transfer from oxidative
They carry protons across the inner mitochondrial membrane, dissipat-
phosphorylation by dissipating the electrical contribu-
ing the proton gradient. Both also uncouple photophosphorylation (see
tion to the electrochemical gradient across the mito-
Fig. 19–65).
chondrial membrane.
A prediction of the chemiosmotic theory is that,
because the role of electron transfer in mitochondrial (Fig. 19–22). Mitochondria manipulated so as to
ATP synthesis is simply to pump protons to create impose a difference of proton concentration and a
the electrochemical potential of the proton-motive separation of charge across the inner membrane syn-
force, an artificially created proton gradient should thesize ATP in the absence of an oxidizable sub-
be able to replace electron transfer in driving ATP strate; the proton-motive force alone suffices to drive
synthesis. This has been experimentally confirmed ATP synthesis.
750 Oxidative Phosphorylation and Photophosphorylation

(a) sensitive), which is integral to the membrane. F1, the

Matrix [H⫹]  10⫺9 M first factor recognized as essential for oxidative phos-
[K⫹]  [Cl⫺]  0.1 M phorylation, was identified and purified by Efraim
Racker and his colleagues in the early 1960s.
In the laboratory, small membrane vesicles formed
F o F1 from inner mitochondrial membranes carry out ATP
synthesis coupled to electron transfer. When F1 is gen-
tly extracted, the “stripped” vesicles still contain intact
[H⫹]  10⫺9 M respiratory chains and the Fo portion of ATP synthase.
Intermembrane space
The vesicles can catalyze electron transfer from NADH
to O2 but cannot produce a proton gradient: Fo has a
pH lowered from 9 to 7; proton pore through which protons leak as fast as they
valinomycin added; no K⫹ are pumped by electron transfer, and without a proton
gradient the F1-depleted vesicles cannot make ATP.
(b) Isolated F1 catalyzes ATP hydrolysis (the reversal of
synthesis) and was therefore originally called F1 ATPase.
[H⫹]  10⫺9 M When purified F1 is added back to the depleted vesicles,
[K⫹] ⬍⬍ [Cl⫺]
⫺ it reassociates with Fo, plugging its proton pore and
(b) ⫺
⫹ ⫺ restoring the membrane’s capacity to couple electron
K⫹ ATP
⫺ ⫹⫹
⫺ transfer and ATP synthesis.
⫺ ⫹⫹ ADP⫹ Pi
⫺ ⫺ [H⫹]  ATP Is Stabilized Relative to ADP on the Surface of F1
10⫺7 M
⫹ ⫹ Isotope exchange experiments with purified F1 reveal a
K⫹
remarkable fact about the enzyme’s catalytic mecha-
FIGURE 19–22 Evidence for the role of a proton gradient in ATP synthe- nism: on the enzyme surface, the reaction
sis. An artificially imposed electrochemical gradient can drive ATP syn-
ADP 1 Pi ∆ ATP 1 H2O is readily reversible—the
thesis in the absence of an oxidizable substrate as electron donor. In this free-energy change for ATP synthesis is close to zero!
two-step experiment, (a) isolated mitochondria are first incubated in a When ATP is hydrolyzed by F1 in the presence of
18
pH 9 buffer containing 0.1 M KCl. Slow leakage of buffer and KCl into the O-labeled water, the Pi released contains an 18O atom.
mitochondria eventually brings the matrix into equilibrium with the sur- Careful measurement of the 18O content of Pi formed in
rounding medium. No oxidizable substrates are present. (b) Mitochondria vitro by F1-catalyzed hydrolysis of ATP reveals that the
are now separated from the pH 9 buffer and resuspended in pH 7 buffer Pi has not one, but three or four 18O atoms (Fig. 19–23).
containing valinomycin but no KCl. The change in buffer creates a differ- This indicates that the terminal pyrophosphate bond in
ence of two pH units across the inner mitochondrial membrane. The out- ATP is cleaved and re-formed repeatedly before Pi
ward flow of K⫹, carried (by valinomycin) down its concentration gradient leaves the enzyme surface. With Pi free to tumble in its
without a counter-ion, creates a charge imbalance across the membrane binding site, each hydrolysis inserts 18O randomly at one
(matrix negative). The sum of the chemical potential provided by the pH of the four positions in the molecule. This exchange
difference and the electrical potential provided by the separation of charges reaction occurs in unenergized FoF1 complexes (with no
is a proton-motive force large enough to support ATP synthesis in the proton gradient) and with isolated F1—the exchange
absence of an oxidizable substrate. does not require the input of energy.
Kinetic studies of the initial rates of ATP synthesis
and hydrolysis confirm the conclusion that DG98 for ATP
ATP Synthase Has Two Functional Domains, Fo and F1
synthesis on the enzyme is near zero. From the mea-
Mitochondrial ATP synthase is an F-type ATPase (see Fig. sured rates of hydrolysis (k1 5 10 s21 ) and synthesis
11–39) similar in structure and mechanism to the ATP (k21 5 24 s21 ), the calculated equilibrium constant for
synthases of chloroplasts and the reaction
bacteria. This large enzyme
complex of the inner mitochon- Enz-ATP ∆ Enz-(ADP 1 Pi )
drial membrane catalyzes the is
formation of ATP from ADP k21 24 s21
and Pi, driven by the flow of K¿eq 5 5 5 2.4
k1 10 s21
protons from the P to the N
side of the membrane (Eqn From this K¿eq, the calculated apparent DG98 is close to
19–10). ATP synthase, also zero. This is much different from the K¿eq of about 105
called Complex V, has two (¢G¿8 5 230.5 kJ/mol) for the hydrolysis of ATP free
distinct components: F1, a in solution (not on the enzyme surface).
peripheral membrane protein, What accounts for the huge difference? ATP syn-
Efraim Racker, 1913–1991 and Fo (o denoting oligomycin- thase stabilizes ATP relative to ADP 1 Pi by binding
 19.2 ATP Synthesis 751

(a) FIGURE 19–23 Catalytic mechanism of F1. (a) 18O-exchange experiment. F1 solubilized from
Enzyme
mitochondrial membranes is incubated with ATP in the presence of 18O-labeled water. At
(F1)
intervals, a sample of the solution is withdrawn and analyzed for the incorporation of 18O
into the Pi produced from ATP hydrolysis. In minutes, the Pi contains three or four 18O atoms,
ATP ⫹ H218O
indicating that both ATP hydrolysis and ATP synthesis have occurred several times during
the incubation. (b) The likely transition state complex for ATP hydrolysis and synthesis by
F1 ATP synthase (derived from PDB ID 1BMF). The ␣ subunit is shown in gray, ␤ in purple. The
ADP b a b positively charged residues ␤-Arg182 and ␣-Arg376 coordinate two oxygens of the pentavalent
⫹
18
phosphate intermediate; ␤-Lys155 interacts with a third oxygen, and the Mg2⫹ ion further sta-
O bilizes the intermediate. The blue sphere represents the leaving group (H2O). These interac-
18
O P 18
O tions result in the ready equilibration of ATP and ADP 1 Pi in the active site.
18
O
FO
ATP more tightly, releasing enough energy to counter-
balance the cost of making ATP. Careful measurements
of the binding constants show that FoF1 binds ATP with
(b)
very high affinity (Kd # 10212 M) and ADP with much
b-Glu181 lower affinity (Kd < 1025 M). The difference in Kd cor-
H2O
Mg 2+
responds to a difference of about 40 kJ/mol in binding
b-Arg182
energy, and this binding energy drives the equilibrium
toward formation of the product ATP.

a-Arg376 The Proton Gradient Drives the Release of ATP from

the Enzyme Surface
ADP Although ATP synthase equilibrates ATP with ADP 1
b-Lys155 Pi, in the absence of a proton gradient the newly synthe-
sized ATP does not leave the surface of the enzyme. It
is the proton gradient that causes the enzyme to
release the ATP formed on its surface. The reaction
coordinate diagram of the process (Fig. 19–24) illus-
trates the difference between the mechanism of ATP
synthase and that of many other enzymes that catalyze
endergonic reactions.
For the continued synthesis of ATP, the enzyme
must cycle between a form that binds ATP very
tightly and a form that releases ATP. Chemical and

80
Typical enzyme ATP synthase ATP
(in solution)
60 ‡

P ADP⫹Pi
G (kJ/mol)

40
ES [E ATP]

20 E ADP⫹Pi

0
E⫹S

Reaction coordinate

FIGURE 19–24 Reaction coordinate diagrams for ATP synthase and for from ADP and Pi in aqueous solution is large and positive, but on the en-
a more typical enzyme. In a typical enzyme-catalyzed reaction (left), zyme surface, the very tight binding of ATP provides sufficient binding
reaching the transition state (‡) between substrate and product is the energy to bring the free energy of the enzyme-bound ATP close to that of
major energy barrier to overcome. In the reaction catalyzed by ATP syn- ADP 1 Pi, so the reaction is readily reversible. The equilibrium constant is
thase (right), release of ATP from the enzyme, not formation of ATP, is near 1. The free energy required for the release of ATP is provided by the
the major energy barrier. The free-energy change for the formation of ATP proton-motive force.
752 Oxidative Phosphorylation and Photophosphorylation

crystallographic studies of the ATP synthase have The Fo complex, with its proton pore, is composed
revealed the structural basis for this alternation in of three subunits, a, b, and c, in the proportion ab2cn.,
function. where n ranges from 8 to 15 in various organisms; in
yeast, it is 10. Subunit c is a small (Mr 8,000), very
hydrophobic polypeptide, consisting almost entirely
of two transmembrane helices, with a small loop
Each ␤ Subunit of ATP Synthase Can Assume Three extending from the matrix side of the membrane. The
Different Conformations crystal structure of the yeast FoF1, solved in 1999,
Mitochondrial F1 has nine subunits of five different shows 10 c subunits, each with two transmembrane
types, with the composition ␣3␤3␥␦␧. Each of the helices roughly perpendicular to the plane of the
three ␤ subunits has one catalytic site for ATP synthe- membrane and arranged in two concentric circles.
sis. The crystallographic determination of the F1 The inner circle is made up of the amino-terminal
structure by John E. Walker helices of each c subunit; the outer circle, about 55 Å
and colleagues revealed in diameter, is made up of the carboxyl-terminal heli-
structural details very helpful ces. The c subunits in this c ring rotate together as
in explaining the catalytic a unit around an axis perpendicular to the mem-
mechanism of the enzyme. brane. The  and  subunits of F1 form a leg-and-foot
The knoblike portion of F1 is that projects from the bottom (membrane) side of F1
a flattened sphere, 8 nm high and stands firmly on the ring of c subunits. The a sub-
and 10 nm across, consisting unit consists of several hydrophobic helices that span
of alternating ␣ and ␤ sub- the membrane in close association with one of the
John E. Walker units arranged like the sec- c subunits in the c ring. The schematic drawing in
tions of an orange (Fig. 19–25a, b, c). The polypep- Figure 19–25a combines the structural information
tides that make up the stalk in the F1 crystal structure from studies of bovine F1, yeast FoF1, and the c ring of
are asymmetrically arranged, with one domain of the Ilyobacter tartaricus.
single  subunit making up a central shaft that passes
through F1, and another domain of  associated pri- Rotational Catalysis Is Key to the Binding-Change
marily with one of the three ␤ subunits, designated
␤-empty (Fig. 19–25b). Although the amino acid Mechanism for ATP Synthesis
sequences of the three ␤ subunits are identical, their On the basis of detailed kinetic and binding studies of
conformations differ, in part because of the associa- the reactions catalyzed by FoF1, Paul Boyer proposed
tion of the  subunit with just one of the three. The a rotational catalysis mechanism in which the
structures of the ␦ and ␧ subunits are not revealed in three active sites of F1 take turns catalyzing ATP syn-
these crystallographic studies. thesis (Fig. 19–26). A given ␤ subunit starts in the
The conformational differences among ␤ subunits ␤-ADP conformation, which binds ADP and Pi from
extend to differences in their ATP/ADP-binding sites. the surrounding medium. The subunit now changes
When researchers crystallized the protein in the pres- conformation, assuming the ␤-ATP form that tightly
ence of ADP and App(NH)p, a close structural analog of binds and stabilizes ATP, bringing about the ready
ATP that cannot be hydrolyzed by the ATPase activity of equilibration of ADP 1 Pi with ATP on the enzyme
F1, the binding site of one of the three ␤ subunits was surface. Finally, the subunit changes to the ␤-empty
filled with App(NH)p, the second was filled with ADP, conformation, which has very low affinity for ATP,
and the third was empty. The corresponding ␤ subunit and the newly synthesized ATP leaves the enzyme
conformations are designated ␤-ATP, ␤-ADP, and surface. Another round of catalysis begins when this
␤-empty (Fig. 19–25b). This difference in nucleotide subunit again assumes the ␤-ADP form and binds
binding among the three subunits is critical to the mech- ADP and Pi.
anism of the complex. The conformational changes central to this mecha-
nism are driven by the passage of protons through the
Fo portion of ATP synthase.
NH2 The streaming of protons
through the Fo pore causes the
␥ ␤ ␣
N
O O O N cylinder of c subunits and the
H

O P N P O P O CH2 N N
attached  subunit to rotate
O
about the long axis of , which
O
 
O 
O is perpendicular to the plane
H H of the membrane. The  sub-
Nonhydrolyzable H H
unit passes through the center
␤␤-␥
␥ bond
OH OH of the ␣3␤3 spheroid, which is
App(NH)p (␤,-imidoadenosine 5⬘-triphosphate) Paul Boyer held stationary relative to the
 19.2 ATP Synthesis 753

FIGURE 19–25 Mitochondrial ATP synthase complex. (a) A cartoon view (b) Top view of F1 a
of the FoF1 complex. (b) (PDB ID 1BMF and PDB ID 1JNV) F1 viewed from
above (that is, from the N side of the membrane), showing the three ␤
(shades of purple) and three ␣ (shades of gray) subunits and the central b-ATP
b-ADP
shaft ( subunit, green). Each ␤ subunit, near its interface with the neigh-
boring ␣ subunit, has a nucleotide-binding site critical to the catalytic ac-
ADP
tivity. The single  subunit associates primarily with one of the three ␣␤
pairs, forcing each of the three ␤ subunits into slightly different conforma-
tions, with different nucleotide-binding sites. In the crystalline enzyme, ATP
one subunit (␤-ADP) has ADP (yellow) in its binding site, the next (␤-
ATP) has ATP (red), and the third (␤-empty) has no bound nucleotide.
(c) The entire enzyme viewed from the side (in the plane of the mem-
brane). The F1 portion (PDB IDs 1BMF, 1JNV, and 2A7U) has three ␣ and
a
three ␤ subunits arranged like the segments of an orange around a central
shaft, the  subunit (green). (Two ␣ subunits and one ␤ subunit have been
a
omitted to reveal the  subunit and the binding sites for ATP and ADP on
the ␤ subunits.) The ␦ subunit confers oligomycin sensitivity on the ATP
synthase, and the ␧ subunit may serve to inhibit its ATPase activity under b-empty
some circumstances. The Fo subunit consists of one a subunit and two b
subunits (PDB ID 2CLY and PDB ID 1B9U), which anchor the FoF1 complex
in the membrane and act as a stator, holding the ␣ and ␤ subunits in place.
Fo also includes the c ring, made up of a number (8 to 15, depending on the
species) of identical c subunits, small, hydrophobic proteins. The c ring and (c) Side view of FoF1
the a subunit interact to provide a transmembrane path for protons. Each c d
subunit has an Asp residue near the middle of the membrane, which can
bind or give up a proton. In this structure (PDB ID 1YCE), we have shown
the homologous c11 ring of the Na1-ATPase of Ilyobacter tartaricus, for
which the structure is well established. The Na1-binding sites in it, which
correspond to the proton-binding sites of the FoF1 complex, are shown with
their bound Na⫹ ions (red spheres). (d) A view of Fo perpendicular to the
membrane. Each of the c subunits in Fo has a critical Asp residue near the
middle of the membrane, which undergoes protonation/deprotonation
ATP
during the catalytic cycle of the ATP synthase. As in (c), red spheres repre-
F1
sent the Na⫹- or proton-binding sites in Asp residues.

(a) b
b2
a d a ADP
g
b b
a e
ATP

b2 ADP + Pi
N side
a
e g
Proton-binding
a sites on Asp Fo
residues

c11
P side
c8 to c15
H+ (c10 in yeast mitochondria)

(d) Bottom view of Fo

a
Na+- or
proton-binding
b2 sites on Asp
residues
754 Oxidative Phosphorylation and Photophosphorylation

␤ FIGURE 19–26 Binding-change model for ATP synthase. The F1 complex has three

nonequivalent adenine nucleotide–binding sites, one for each pair of ␣ and ␤ subunits.
ATP
At any given moment, one of these sites is in the ␤-ATP conformation (which binds
␤ ADP ATP tightly), a second is in the ␤-ADP (loose-binding) conformation, and a third is in
+Pi 
the ␤-empty (very-loose-binding) conformation. The proton-motive force causes rota-
tion of the central shaft—the  subunit, shown as a green arrowhead—which comes
into contact with each ␣␤ subunit pair in succession. This produces a cooperative con-
 ␤
3 HP+ 3 HN+
formational change in which the ␤-ATP site is converted to the ␤-empty conforma-
tion, and ATP dissociates; the ␤-ADP site is converted to the ␤-ATP conformation,
ATP which promotes condensation of bound ADP 1 Pi to form ATP; and the ␤-empty site
+
3 HN becomes a ␤-ADP site, which loosely binds ADP 1 Pi entering from the solvent. This
3 HP+ model, based on experimental findings, requires that at least two of the three catalytic
ATP sites alternate in activity; ATP cannot be released from one site unless and until ADP
␤ and Pi are bound at the other.
  ␤
ATP
ADP
+Pi One strong prediction of this binding-change
␤ ATP  ␤  model is that the  subunit should rotate in one direc-
tion when FoF1 is synthesizing ATP and in the opposite
ADP
+Pi ATP direction when the enzyme is hydrolyzing ATP. This
 ␤  ␤ prediction of rotation with ATP hydrolysis was con-
firmed in elegant experiments in the laboratories of
3 HP+ 3 HN+ Masasuke Yoshida and Kazuhiko Kinosita, Jr. The rota-
tion of  in a single F1 molecule was observed micro-
scopically by attaching a long, thin, fluorescent actin
membrane surface by the b2 and ␦ subunits (Fig. 19–25a). polymer to  and watching it move relative to ␣3␤3
With each rotation of 1208, ␥ comes into contact with a immobilized on a microscope slide as ATP was hydro-
different ␤ subunit, and the contact forces that ␤ sub- lyzed. (The expected reversal of the rotation when ATP
unit into the ␤-empty conformation. is being synthesized could not be tested in this experi-
The three ␤ subunits interact in such a way that ment; there is no proton gradient to drive ATP synthe-
when one assumes the ␤-empty conformation, its neigh- sis.) When the entire FoF1 complex (not just F1) was
bor to one side must assume the ␤-ADP form, and the used in a similar experiment, the entire ring of c sub-
other neighbor the ␤-ATP form. Thus one complete rota- units rotated with  (Fig. 19–27). The “shaft” rotated in
tion of the  subunit causes each ␤ subunit to cycle the predicted direction through 3608. The rotation was
through all three of its possible conformations, and for not smooth, but occurred in three discrete steps of 1208.
each rotation, three ATP are synthesized and released As calculated from the known rate of ATP hydrolysis by
from the enzyme surface. one F1 molecule and from the frictional drag on the long

FIGURE 19–27 Experimental demonstration of rotation of Fo and ␥. F1 geneti-

Actin filament
cally engineered to contain a run of His residues adheres tightly to a microscope
slide coated with a Ni complex; biotin is covalently attached to a c subunit of Fo.
The protein avidin, which binds biotin very tightly, is covalently attached to long
Avidin
filaments of actin labeled with a fluorescent probe. Biotin-avidin binding now at-
Biotin
taches the actin filaments to the c subunit. When ATP is provided as substrate
for the ATPase activity of F1, the labeled filament is seen to rotate continuously
Fo in one direction, proving that the Fo cylinder of c subunits rotates. In another ex-
a
c10 periment, a fluorescent actin filament was attached directly to the  subunit. The
series of fluorescence micrographs (read left to right) shows the position of the
e
actin filament at intervals of 133 ms. Note that as the filament rotates, it makes
g a discrete jump about every eleventh frame. Presumably the cylinder and shaft
b2
move as one unit.

F1 b b
a ATP

ADP + Pi

His His
His
Ni complex
 19.2 ATP Synthesis 755

actin polymer, the efficiency of this mechanism in con- arranged in a circle about a central core that is prob-
verting chemical energy into motion is close to 100%. It ably filled with membrane lipids. In each c subunit,
is, in Boyer’s words, “a splendid molecular machine!” there is a critical Asp (or Glu) residue located at
about the middle of the membrane. Protons cross the
How Does Proton Flow through the Fo Complex membrane through a path made up of both a and c
subunits. The a subunit has a proton half-channel
Produce Rotary Motion? leading from the cytosol (P side) to the middle of the
One feasible model to explain how proton flow and membrane, where it ends near the Asp residue of the
rotary motion are coupled in the Fo complex is shown adjoining c subunit. A proton diffuses from the cyto-
in Figure 19–28. The individual subunits in Fo are solic side (where the proton concentration is rela-
tively high) down this half-channel and binds to the
Asp residue, displacing a positively charged Arg resi-
4 Displaced H+ due that had been associated with the Asp. This Arg
exits on N side H+ residue swings aside, forming an interaction with the
5 c10 ring rotates, Asp on the adjacent c subunit in the ring and displac-
Arg210 returns to
P-side half-channel ing the proton from that Asp; this adjacent proton
exits through the second half-channel to the N side,
c10
+H where the proton concentration is relatively low, com-
Arg210 H+
+H
H+
pleting the movement of one proton equivalent from
H+
+
outside to inside the matrix. Now another proton has
a entered the half-channel on the cytosolic side, moved
3 Arg rotates,
210 to the Asp of the next c subunit, and protonated it,
displacing H+ 6 Process repeats again displacing Arg, which in turn displaces a proton
from Asp from the next c subunit, and so forth. The rotary
movement of the c ring is the result of thermal
(Brownian) motion, made unidirectional by the large
Half-channel (N side) c10 difference in proton concentration across the mem-
+H
Arg210 H+ brane. The number of protons that must be trans-
+H
H+ ferred to produce one complete rotation of the c ring
H+ +
is equal to the number of c subunits in the ring. Stud-
a ies of the c ring with atomic force microscopy (Box
Half-channel (P side)
19–2) or x-ray diffraction have shown that the number
2 Proton displaces H+ 1 Proton enters half- of c subunits is different in different organisms (Fig.
Arg210 to adjacent channel on P side
c subunit 19–29). In animal mitochondria this number is 8, in
yeast mitochondria and in E. coli it is 10, and the
number of c subunits can range as high as 15 in the
FIGURE 19–28 A model for proton-driven rotation of the c ring. The cyanobacterium Spirulina platensis. The rate of
a subunit of the Fo complex of the ATP synthase (see Fig. 19–25a) has rotation in intact mitochondria has been estimated at
two hydrophilic half-channels for protons, one leading from the P side to about 6,000 rpm—100 rotations per second!
the middle of the membrane, and another leading from the middle of the
membrane to the N side (matrix). The individual c subunits in Fo (10 in the
yeast enzyme) are arranged in a circle about a central core. Each c subunit Chemiosmotic Coupling Allows Nonintegral
has a critical Asp residue about midway across the membrane, which can Stoichiometries of O2 Consumption and ATP Synthesis
donate or accept a proton (red H⫹). The a subunit has a positively charged
Before the general acceptance of the chemiosmotic
Arg side chain that forms an electrostatic interaction with the negatively
model for oxidative phosphorylation, the assumption
charged carboxylate of Asp on the adjoining c subunit. This c subunit is
was that the overall reaction equation would take the
initially positioned so that a proton that enters the half-channel on the P
following form:
side (where the proton concentration is relatively high) encounters and
protonates the Asp residue, weakening its interaction with Arg. The Arg xADP 1 xPi 1 12O2 1 H1 1 NADH ¡
side chain rotates toward the protonated Asp residue in the next c sub- xATP 1 H2O 1 NAD1 (19–11)
unit, and displaces its carboxyl proton, as a new electrostatic Arg-Asp
interaction forms. The displaced proton moves through the second half- with the value of x—sometimes called the P/O ratio or
channel in subunit a and is released on the N side. The c subunit with its the P/2e ratio—always an integer. When intact mito-
unprotonated Asp residue moves so that its Arg-Asp pair faces the half- chondria are suspended in solution with an oxidizable
channel on the P side, and a second cycle begins: proton entry, protonation substrate such as succinate or NADH and are provided
of Asp, Arg movement, and proton exit. Rotation of the ring occurs by with O2, ATP synthesis is readily measurable, as is the
thermal (Brownian) motion and is effectively ratcheted; the orientation decrease in O2. In principle, these two measurements
of the proton gradient dictates the direction of proton flow and makes the should yield the number of ATP synthesized per ½O2
rotation of the c ring essentially unidirectional. consumed, the P/O ratio. Measurement of P/O, however,
756 Oxidative Phosphorylation and Photophosphorylation

(a) Top view Side view FIGURE 19–29 Different organisms have different numbers of c subunits
in the c ring of the Fo complex. The structure of the c ring from several
species has been determined by x-ray crystallography. Each helix in the
inner ring is half of a hairpin-shaped c subunit; the outer ring of helices is
the other half of the hairpin structure. Views of the c ring perpendicular to
the membrane show the number of c subunits for (a) bovine mitochondria
(8) and (b) yeast mitochondria (10). Atomic force microscopy has been
used to visualize the c rings of (c) the thermophilic bacterium Bacillus
(b) species TA2.A1 (13) and (d) spinach (14). According to the model in
Fig. 19–28, different numbers of c subunits in the c ring should result in
different ratios of ATP formed per pair of electrons passing through the
respiratory chain (different P/O ratios).

is complicated by the fact that intact mitochondria

(c) (d)
consume ATP in many unrelated reactions taking place
in the matrix, and they consume O2 for purposes other
than oxidative phosphorylation. The contribution of
these interfering reactions must be carefully mea-
sured, and the calculation of P/O must be corrected
for their contributions. Most experiments have yield-
ed P/O (ATP to ½O2) ratios of between 2 and 3 when
NADH was the electron donor, and between 1 and 2
when succinate was the donor. Given the assumption
that P/O should have an integral value, most experi-
menters agreed that the P/O ratios must be 3 for
NADH and 2 for succinate, and for years those values
5 nm
appeared in research papers and textbooks.
With introduction of the chemiosmotic paradigm for
coupling ATP synthesis to electron transfer, there was no

BOX 19–2 METHODS Atomic Force Microscopy to Visualize Membrane Proteins

In atomic force microscopy (AFM), the sharp tip of Laser
a microscopic probe attached to a flexible cantilever
Laser light
is drawn across an uneven surface such as a mem- + detector (detects Cantilever
brane (Fig. 1). Electrostatic and van der Waals cantilever deflection)
interactions between the tip and the sample pro- –
duce a force that moves the probe up and down (in
the z dimension) as it encounters hills and valleys in
the sample. A laser beam reflected from the cantile-
ver detects motions of as little as 1 Å. In one type of
Sample
atomic force microscope, the force on the probe is
z
held constant (relative to a standard force, on the
order of piconewtons) by a feedback circuit that
causes the platform holding the sample to rise or fall x y
to keep the force constant. A series of scans in the Platform moves to maintain
x and y dimensions (the plane of the membrane) constant pressure on cantilever
yields a three-dimensional contour map of the sur- tip. Excursions in the z dimension
are plotted as a function of x, y.
face with resolution near the atomic scale—0.1 nm
in the vertical dimension, 0.5 to 1.0 nm in the lat-
eral dimensions.
In favorable cases, AFM can be used to study
single membrane protein molecules, such as the c
subunits of the Fo complex (Fig. 19–29c, d).
FIGURE 1 The principle of atomic force microscopy.
 19.2 ATP Synthesis 757

theoretical requirement for P/O to be integral. The rele- Adenine nucleotide Phosphate
vant questions about stoichiometry became, How many translocase ATP synthase translocase
(antiporter) (symporter)
protons are pumped outward by electron transfer from
Intermembrane ATP4⫺ H⫹
one NADH to O2, and how many protons must flow H2PO4⫺
space ADP3⫺
inward through the FoF1 complex to drive the synthesis ⫹⫹⫹⫹⫹⫹⫹⫹ ⫹⫹⫹
H+
⫹⫹⫹ ⫹
of one ATP? The measurement of proton fluxes is techni-
cally complicated; the investigator must take into account
the buffering capacity of mitochondria, nonproductive ⫺⫺⫺⫺⫺⫺⫺⫺ ⫺⫺⫺ ⫺⫺⫺ ⫺
leakage of protons across the inner membrane, and use Matrix
of the proton gradient for functions other than ATP syn-
thesis, such as driving the transport of substrates across
the inner mitochondrial membrane (described below).
The consensus experimental values for number of pro- ⫹
tons pumped out per pair of electrons are 10 for NADH H2PO4⫺ H
ADP3⫺
and 6 for succinate (which sends electrons into the respi-
ATP4⫺
ratory chain at the level of ubiquinone). The most widely
accepted experimental value for number of protons
H+
required to drive the synthesis of an ATP molecule is 4,
of which 1 is used in transporting Pi, ATP, and ADP FIGURE 19–30 Adenine nucleotide and phosphate translocases. Trans-
across the mitochondrial membrane (see below). If 10 port systems of the inner mitochondrial membrane carry ADP and Pi into
protons are pumped out per NADH and 4 must flow in to the matrix and newly synthesized ATP into the cytosol. The adenine nu-
produce 1 ATP, the proton-based P/O ratio is 2.5 for cleotide translocase is an antiporter; the same protein moves ADP into the
NADH as the electron donor and 1.5 (6/4) for succinate. matrix and ATP out. The effect of replacing ATP4 with ADP3 in the ma-
trix is the net efflux of one negative charge, which is favored by the charge
difference across the inner membrane (outside positive). At pH 7, Pi is
The Proton-Motive Force Energizes Active Transport present as both HPO242 and H2PO 2 4 ; the phosphate translocase is specific

Although the primary role of the proton gradient in for H2PO2 4 . There is no net flow of charge during symport of H2PO2 4 and
H , but the relatively low proton concentration in the matrix favors the in-
⫹
mitochondria is to furnish energy for the synthesis of
ward movement of H⫹. Thus the proton-motive force is responsible both
ATP, the proton-motive force also drives several trans-
for providing the energy for ATP synthesis and for transporting substrates
port processes essential to oxidative phosphorylation.
(ADP and Pi) into and product (ATP) out of the mitochondrial matrix. All
The inner mitochondrial membrane is generally imper-
three of these transport systems can be isolated as a single membrane-
meable to charged species, but two specific systems
bound complex (ATP synthasome).
transport ADP and Pi into the matrix and ATP out to the
cytosol (Fig. 19–30).
The adenine nucleotide translocase, integral ATP synthase and both translocases, the ATP syntha-
to the inner membrane, binds ADP3 in the intermem- some, can be isolated from mitochondria by gentle dis-
brane space and transports it into the matrix in section with detergents, suggesting that the functions of
exchange for an ATP4 molecule simultaneously these three proteins are very tightly integrated.
transported outward (see Fig. 13–11 for the ionic
forms of ATP and ADP). Because this antiporter
moves four negative charges out for every three WORKED EXAMPLE 19–2 Stoichiometry of ATP Production:
moved in, its activity is favored by the transmembrane Effect of c Ring Size
electrochemical gradient, which gives the matrix a net (a) If bovine mitochondria have 8 c subunits per c ring,
negative charge; the proton-motive force drives ATP- what is the predicted ratio of ATP formed per NADH
ADP exchange. Adenine nucleotide translocase is oxidized? (b) What is the predicted value for yeast
specifically inhibited by atractyloside, a toxic glyco- mitochondria, with 10 c subunits? (c) What are the
side formed by a species of thistle. If the transport of comparable values for electrons entering the respira-
ADP into and ATP out of mitochondria is inhibited, tory chain from FADH2?
cytosolic ATP cannot be regenerated from ADP,
explaining the toxicity of atractyloside. Solution: (a) The question asks us to determine how
A second membrane transport system essential to many ATP are produced per NADH. This is another way
oxidative phosphorylation is the phosphate translo- of asking us to calculate the P/O ratio, or x in Equation
case, which promotes symport of one H2PO2 4 and one H
⫹
19–11. If the c ring has 8 c subunits, then one full rota-
into the matrix. This transport process, too, is favored by tion will transfer 8 protons to the matrix and produce 3
the transmembrane proton gradient (Fig. 19–30). Notice ATP molecules. But this synthesis also requires the
that the process requires movement of one proton from transport of 3 Pi into the matrix, at a cost of 1 proton
the P to the N side of the inner membrane, consuming each, adding 3 more protons to the total number
some of the energy of electron transfer. A complex of the required. This brings the total cost to (11 protons)/
758 Oxidative Phosphorylation and Photophosphorylation

(3 ATP) 5 3.7 protons/ATP. The consensus value for the ATP/NADH and 1.5 ATP/FADH2, and we therefore use
number of protons pumped out per pair of electrons these values throughout this book.
transferred from NADH is 10 (see Fig. 19–19). So, oxidiz-
ing 1 NADH produces (10 protons)/(3.7 protons/ATP) 5
2.7 ATP.
Shuttle Systems Indirectly Convey Cytosolic NADH
(b) If the c ring has 10 c subunits, then one full rota-
tion will transfer 10 protons to the matrix and produce 3 into Mitochondria for Oxidation
ATP molecules. Adding in the 3 protons to transport the The NADH dehydrogenase of the inner mitochondrial
3 Pi into the matrix brings the total cost to (13 protons)/ membrane of animal cells can accept electrons only
(3 ATP) 5 4.3 protons/ATP. Oxidizing 1 NADH produc- from NADH in the matrix. Given that the inner mem-
es (10 protons)/(4.3 protons/ATP) 5 2.3 ATP. brane is not permeable to NADH, how can the NADH
(c) When electrons enter the respiratory chain generated by glycolysis in the cytosol be reoxidized to
from FADH2 (at ubiquinone), only 6 protons are NAD⫹ by O2 via the respiratory chain? Special shuttle
available to drive ATP synthesis. This changes the systems carry reducing equivalents from cytosolic
calculation for bovine mitochondria to (6 protons)/ NADH into mitochondria by an indirect route. The most
(3.7 protons/ATP) 5 1.6 ATP per pair of electrons active NADH shuttle, which functions in liver, kidney,
from FADH2. For yeast mitochondria, the calculation is and heart mitochondria, is the malate-aspartate
(6 protons)/(4.3 protons/ATP) 5 1.4 ATP per pair of shuttle (Fig. 19–31). The reducing equivalents of
electrons from FADH2. cytosolic NADH are first transferred to cytosolic oxalo-
These calculated values of x or the P/O ratio define acetate to yield malate, catalyzed by cytosolic malate
a range that includes the experimental values of 2.5 dehydrogenase. The malate thus formed passes through

Intermembrane space Matrix

Malate–
(P side) (N side)
OH a-ketoglutarate OH
 transporter
OOC CH2 C COO
2 
OOC CH2 C COO
NAD+ H H NAD+
Malate Malate

H⫹ ⫹ NADH 1 3
malate malate NADH ⫹ H⫹
O dehydrogenase dehydrogenase O

OOC CH2 C COO 
OOC CH2 C COO
Oxaloacetate NH⫹ NH⫹
3 3 Oxaloacetate

OOC CH2 CH2 C COO 
OOC CH2 CH2 C COO
H H
Glutamate Glutamate
aspartate aspartate
aminotransferase 6 4 aminotransferase
a-Ketoglutarate a-Ketoglutarate
O O

OOC CH2 CH2 C COO 
OOC CH2 CH2 C COO

NH⫹ Aspartate Aspartate NH⫹

3 3

OOC CH2 C COO 
OOC CH2 C COO
H 5 H
Glutamate-aspartate
transporter

Net efect: NADHP NADHN

FIGURE 19–31 Malate-aspartate shuttle. This shuttle for transporting transporter. 3 In the matrix, malate passes two reducing equivalents to
reducing equivalents from cytosolic NADH into the mitochondrial matrix NAD⫹, and the resulting NADH is oxidized by the respiratory chain; the
is used in liver, kidney, and heart. 1 NADH in the cytosol enters the in- oxaloacetate formed from malate cannot pass directly into the cytosol.
termembrane space through openings in the outer membrane (porins), 4 Oxaloacetate is first transaminated to aspartate, and 5 aspartate can
then passes two reducing equivalents to oxaloacetate, producing malate. leave via the glutamate-aspartate transporter. 6 Oxaloacetate is regener-
2 Malate crosses the inner membrane via the malate–␣-ketoglutarate ated in the cytosol, completing the cycle.
 19.3 Regulation of Oxidative Phosphorylation 759

the inner membrane via the malate–␣-ketoglutarate SUMMARY 19.2 ATP Synthesis
transporter. Within the matrix the reducing equivalents  The flow of electrons through Complexes I, III,
are passed to NAD⫹ by the action of matrix malate dehy-
and IV results in pumping of protons across the
drogenase, forming NADH; this NADH can pass electrons
inner mitochondrial membrane, making the matrix
directly to the respiratory chain. About 2.5 molecules of
alkaline relative to the intermembrane space. This
ATP are generated as this pair of electrons passes to O2.
proton gradient provides the energy (in the form
Cytosolic oxaloacetate must be regenerated by transami-
of the proton-motive force) for ATP synthesis
nation reactions and the activity of membrane transport-
from ADP and Pi by ATP synthase (FoF1 complex)
ers to start another cycle of the shuttle.
in the inner membrane.
Skeletal muscle and brain use a different NADH
shuttle, the glycerol 3-phosphate shuttle (Fig.  ATP synthase carries out “rotational catalysis,” in
19–32). It differs from the malate-aspartate shuttle in which the flow of protons through Fo causes each
that it delivers the reducing equivalents from NADH to of three nucleotide-binding sites in F1 to cycle
ubiquinone and thus into Complex III, not Complex I from (ADP 1 Pi)–bound to ATP-bound to empty
(Fig. 19–8), providing only enough energy to synthesize conformations.
1.5 ATP molecules per pair of electrons.  ATP formation on the enzyme requires little
The mitochondria of plants have an externally ori- energy; the role of the proton-motive force is to
ented NADH dehydrogenase that can transfer electrons push ATP from its binding site on the synthase.
directly from cytosolic NADH into the respiratory chain  The ratio of ATP synthesized per ½O2 reduced to
at the level of ubiquinone. Because this pathway
H2O (the P/O ratio) is about 2.5 when electrons
bypasses the NADH dehydrogenase of Complex I and
enter the respiratory chain at Complex I, and 1.5
the associated proton movement, the yield of ATP from
when electrons enter at ubiquinone. This ratio may
cytosolic NADH is less than that from NADH generated
vary somewhat in different organisms based on the
in the matrix (Box 19–1).
number of c subunits in the Fo complex.
 Energy conserved in a proton gradient can drive
solute transport uphill across a membrane.
Intermembrane Glycolysis  The inner mitochondrial membrane is
space (P side) impermeable to NADH and NAD⫹, but NADH
equivalents are moved from the cytosol to the
NAD+ NADH + H+ matrix by either of two shuttles. NADH equivalents
cytosolic
glycerol 3-phosphate moved in by the malate-aspartate shuttle enter the
dehydrogenase respiratory chain at Complex I and yield a P/O
ratio of 2.5; those moved in by the glycerol
CH2OH 3-phosphate shuttle enter at ubiquinone and give a
–

C–
–O P/O ratio of 1.5.
–

Glycerol 3- Dihydroxyacetone
phosphate phosphate CH2 – O – P

CH2OH 19.3 Regulation of Oxidative Phosphorylation

–

CHOH mitochondrial Oxidative phosphorylation produces most of the ATP

glycerol 3-phosphate
–

CH2 – O – P dehydrogenase made in aerobic cells. Complete oxidation of a molecule

FAD
FADH2 of glucose to CO2 yields 30 or 32 ATP (Table 19–5). By
comparison, glycolysis under anaerobic conditions (lac-
QH2 Complex tate fermentation) yields only 2 ATP per glucose.
III Clearly, the evolution of oxidative phosphorylation pro-
vided a tremendous increase in the energy efficiency of
Matrix (N side) catabolism. Complete oxidation to CO2 of the coenzyme
A derivative of palmitate (16:0), which also occurs in
the mitochondrial matrix, yields 108 ATP per palmitoyl-
FIGURE 19–32 Glycerol 3-phosphate shuttle. This alternative means of CoA (see Table 17–1). A similar calculation can be made
moving reducing equivalents from the cytosol to the mitochondrial ma- for the ATP yield from oxidation of each of the amino
trix operates in skeletal muscle and the brain. In the cytosol, dihydroxy- acids (Chapter 18). Aerobic oxidative pathways that
acetone phosphate accepts two reducing equivalents from NADH in a result in electron transfer to O2 accompanied by oxida-
reaction catalyzed by cytosolic glycerol 3-phosphate dehydrogenase. An tive phosphorylation therefore account for the vast
isozyme of glycerol 3-phosphate dehydrogenase bound to the outer face majority of the ATP produced in catabolism, so the
of the inner membrane then transfers two reducing equivalents from glyc- regulation of ATP production by oxidative phosphoryla-
erol 3-phosphate in the intermembrane space to ubiquinone. Note that tion to match the cell’s fluctuating needs for ATP is
this shuttle does not involve membrane transport systems. absolutely essential.