Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

5. Succinyl-CoA synthetase converts succinyl-coenzyme

A to succinate.The free energy of the thioester bond is con-

served in this reaction by the formation of “high-energy”

GTP from GDP ⫹ P

6. The remaining reactions of the cycle serve to oxidize

succinate back to oxaloacetate in preparation for another

round of the cycle. Succinate dehydrogenase catalyzes the

oxidation of succinate’s central single bond to a trans dou-

ble bond, yielding fumarate with the concomitant reduc-

tion of the redox coenzyme FAD to FADH

(the molecular

formulas of FAD and FADH

and the reactions through

which they are interconverted are given in Fig. 16-8).

7. Fumarase then catalyzes the hydration of fumarate’s

double bond to yield malate.

8. Finally, malate dehydrogenase reforms oxaloacetate

by oxidizing malate’s secondary alcohol group to the corre-

sponding ketone with concomitant reduction of a third

NAD

⫹

to NADH. Acetyl groups are thereby completely

oxidized to CO

with the following stoichiometry:

The citric acid cycle functions catalytically as a consequence

of its regeneration of oxaloacetate: An endless number of

acetyl groups can be oxidized through the agency of a single

oxaloacetate molecule.

NADH and FADH

are vital products of the citric acid

cycle.Their reoxidation by O

through the mediation of the

electron-transport chain and oxidative phosphorylation

(Chapter 22) completes the breakdown of metabolic fuel

in a manner that drives the synthesis of ATP. Other func-

tions of the cycle are discussed in Section 21-5.

B. Historical Perspective

The citric acid cycle was proposed in 1937 by Hans Krebs, a

contribution that ranks as one of the most important

achievements of metabolic chemistry.We therefore outline

the intellectual history of this cycle’s discovery.

By the early 1930s, significant progress had been made

in elucidating the glycolytic pathway (Section 17-1A). Yet

the mechanism of glucose oxidation and its relationship to

cellular respiration (oxygen uptake) was still a mystery.

Nevertheless, the involvement of several metabolites in

cellular oxidative processes was recognized. It was well

known, for example, that in addition to lactate and acetate,

the dicarboxylates succinate, malate, and ␣-ketoglutarate,

as well as the tricarboxylate citrate, are rapidly oxidized by

muscle tissue during respiration. It had also been shown

that malonate (Section 21-3F), a potent inhibitor of succi-

nate oxidation to fumarate, also inhibits cellular respira-

tion, thereby suggesting that succinate plays a central role

in oxidative metabolism rather than being just another

metabolic fuel.

In 1935, Albert Szent-Györgyi demonstrated that cellu-

lar respiration is dramatically accelerated by catalytic

amounts of succinate, fumarate, malate, and oxaloacetate;

3NADH ⫹ FADH

⫹ GTP ⫹ CoA ⫹ 2CO

3NAD

⫹

⫹ FAD ⫹ GDP ⫹ P

⫹ acetyl-CoA

that is, the addition of any of these substances to minced

muscle tissue stimulates O

uptake and CO

production far

in excess of that required to oxidize the added dicarboxylic

acid. Szent-Györgyi further showed that these compounds

were interconverted according to the reaction sequence:

Shortly afterward, Carl Martius and Franz Knoop dem-

onstrated that citrate is rearranged, via cis-aconitate, to

isocitrate and then dehydrogenated to ␣-ketoglutarate.

␣-Ketoglutarate was already known to undergo oxidative

decarboxylation to succinate and CO

. This extended the

proposed reaction sequence to

What was necessary to close the circle so as to make the

system catalytic was to establish that oxaloacetate is con-

verted to citrate. In 1936, Martius and Knoop demon-

strated that citrate could be formed nonenzymatically from

oxaloacetate and pyruvate by treatment with hydrogen

peroxide under basic conditions. Krebs used this chemical

model as the point of departure for the biochemical exper-

iments that led to his proposal of the citric acid cycle.

Krebs’ hypothesis was based on his investigations, starting

in 1936, on respiration in minced pigeon breast muscle (which

has a particularly high rate of respiration). The idea of a cat-

alytic cycle was not new to him: In 1932, he and Kurt Hense-

leit had elucidated the outlines of the urea cycle, a process in

which ammonia and CO

are converted to urea (Section

26-2). The most important observations Krebs made in sup-

port of the existence of the citric acid cycle were as follows:

1. Succinate is formed from fumarate, malate, or oxalo-

acetate in the presence of the metabolic inhibitor malonate.

Since malonate inhibits the direct reduction of fumarate to

succinate, the succinate must be formed by an oxidative cycle.

2. Pyruvate and oxaloacetate can form citrate enzymat-

ically. Krebs therefore suggested that the metabolic cycle is

closed with the reaction:

3. The interconversion rates of the cycle’s individual

steps are sufficiently rapid to account for observed respira-

tion rates, so it must be (at least) the major pathway for

pyruvate oxidation in muscle.

Although Krebs had established the existence of the cit-

ric acid cycle, some major gaps still remained in its com-

plete elucidation. The mechanism of citrate formation did

not become clear until Nathan Kaplan and Fritz Lipmann

discovered coenzyme A in 1945 (Section 21-2), and Severo

Ochoa and Feodor Lynen established, in 1951, that acetyl-

CoA is the intermediate that condenses with oxaloacetate

to form citrate. Oxidative decarboxylation of ␣-ketoglutarate

to succinate was also shown to involve coenzyme A with

succinyl-CoA as an intermediate.

Pyruvate ⫹ oxaloacetate

citrate ⫹ CO

malate

oxaloacetate

␣-ketoglutarate

succinate

fumarate

Citrate

cis-aconitate

isocitrate

Succinate

fumarate

malate

oxaloacetate

Section 21-1. Cycle Overview 791

JWCL281_c21_789-822.qxd 3/17/10 11:33 AM Page 791

Acetyl-coenzyme A (acetyl-CoA)

⬃

O OH

P O P

O CH

–

Adenosine-3⬘-

phosphate

Pantothenic

acid residue

Acetyl group

␤-Mercaptoethylamine

residue

The elucidation of the citric acid cycle was a major

achievement and, like all achievements of this magnitude,

required the efforts of numerous investigators. Indeed, bio-

chemists are still working to understand the cycle on a mo-

lecular and enzymatic level. We shall study the eight en-

zymes that catalyze the cycle after first discussing the cycle’s

major fuel, acetyl-CoA, and its formation from pyruvate.

2 METABOLIC SOURCES OF

ACETYL-COENZYME A

Acetyl groups enter the citric acid cycle as acetyl-coenzyme

A (acetyl-SCoA or acetyl-CoA; Fig. 21-2), the common

product of carbohydrate, fatty acid, and amino acid break-

down. Coenzyme A (CoASH or CoA) consists of a ␤-

mercaptoethylamine group bonded through an amide linkage

to the vitamin pantothenic acid, which, in turn, is attached to

a 3¿-phosphoadenosine moiety via a pyrophosphate bridge.

The acetyl group of acetyl-CoA is bonded as a thioester to the

sulfhydryl portion of the ␤-mercaptoethylamine group. CoA

thereby functions as a carrier of acetyl and other acyl groups

(the A in CoA stands for “Acetylation”).

Acetyl-CoA is a “high-energy” compound: The ⌬G°¿ for

the hydrolysis of its thioester bond is ⫺31.5 kJ ⴢ mol

⫺1

which makes this reaction slightly (1 kJ ⴢ mol

⫺1

) more

exergonic than that of ATP hydrolysis (Section 16-4A).

The formation of this thioester bond in a metabolic inter-

mediate therefore conserves a portion of the free energy of

oxidation of a metabolic fuel.

A. Pyruvate Dehydrogenase Multienzyme

Complex (PDC)

The immediate precursor to acetyl-CoA from carbohydrate

sources is the glycolytic product pyruvate. As we saw in Sec-

tion 17-3, under anaerobic conditions the NADH produced

by glycolysis is reoxidized with concomitant reduction of

pyruvate to lactate (in muscle) or ethanol (in yeast). Under

aerobic conditions, however, NADH is reoxidized by the

mitochondrial electron-transport chain (Section 22-2), so

that pyruvate, which enters the mitochondrion via a spe-

cific pyruvate–H

⫹

symport (membrane transport nomen-

clature is discussed in Section 20-3), can undergo further

oxidation. (The formation of acetyl-CoA from fatty acids

and amino acids is discussed in Sections 25-2 and 26-3.)

Acetyl-CoA is synthesized from pyruvate and CoA

through oxidative decarboxylation by a multienzyme com-

plex named pyruvate dehydrogenase. In general, multien-

zyme complexes are groups of noncovalently associated en-

zymes that catalyze two or more sequential steps in a

metabolic pathway. The pyruvate dehydrogenase multien-

zyme complex (PDC) consists of three enzymes: pyruvate

dehydrogenase (E1), dihydrolipoyl transacetylase (E2), and

dihydrolipoyl dehydrogenase (E3). The E. coli pyruvate de-

hydrogenase complex, whose characterization was pio-

neered by Lester Reed, is an ⬃4600-kD polyhedral particle

that is ⬃300 Å in diameter (Fig. 21-3a). Isolated E. coli E2

forms a particle with 24 identical subunits, which electron

micrographs (Figs. 21-3b and 21-4a), together with an X-ray

structure by Wim Hol (Fig. 21-5a), indicate are arranged

with cubic symmetry.The E1 subunits form dimers that asso-

ciate with the E2 cube at the centers of the cube’s 12 edges

(Fig. 21-4b,c), whereas the E3 subunits form dimers that are

located at the centers of this cube’s 6 faces.We discuss the X-

ray structures of the E1, E2, and E3 subunits below.

a. Some PDCs Have a Dodecahedral Form

Although all PDCs catalyze the same reactions by similar

mechanisms, they may have different quaternary structures.

Whereas the PDCs of E. coli and most other gram-negative

bacteria have the foregoing cubic symmetry, those of

792 Chapter 21. Citric Acid Cycle

Figure 21-2 Chemical structure of acetyl-CoA.

The thioester bond is drawn with a squiggle (

⬃

) to

indicate that it is a “high-energy” bond. In CoA, the

acetyl group is replaced by a hydrogen atom.

JWCL281_c21_789-822.qxd 3/17/10 11:33 AM Page 792

(a) (b) (c)

Section 21-2. Metabolic Sources of Acetyl-Coenzyme A 793

Figure 21-3 Electron micrographs of the

E. coli pyruvate dehydrogenase multienzyme

complex. (a) The intact complex. (b) The

dihydrolipoyl transacetylase (E2) “core”

complex. [Courtesy of Lester Reed,

University of Texas.]

Figure 21-4 Structural organization of the E. coli PDC. (a)

The dihydrolipoyl transacetylase (E2) “core.” Its 24 subunits

(green spheres) associate as trimers located at the corners of a

cube to form a particle that has cubic symmetry (O symmetry;

Section 8-5B). (b) The 24 pyruvate dehydrogenase (E1) subunits

(orange spheres) form dimers that associate with the E2 core

(shaded cube) at the centers of each of its 12 edges, whereas the

12 dihydrolipoyl dehydrogenase (E3) subunits (purple spheres)

form dimers that attach to the E2 cube at the centers of each of its

6 faces. (c) Parts a and b combined to form the entire 60-subunit

complex.

Figure 21-5 Comparison of the X-ray structures of the

dihydrolipoyl transacetylase (E2) cores of PDCs. Both complexes

are drawn mainly in space-filling representation and viewed

along their respective 2-fold axes of symmetry. (a) The 125-Å

high cubic (O symmetry; Fig. 8-65c) E2 core from the gram-

negative bacterium Azotobacter vinelandii. It consists of 24

subunits that form 8 trimers, shown here in different colors. The

positions of a 2-fold, a 3-fold, and a 4-fold axis of symmetry are

indicated.The inset (upper right) is a perspective drawing of a

cube. (b) The 237-Å-diameter dodecahedral (I symmetry;

Fig. 8-65c) E2 core from the gram-positive bacterium Bacillus

stearothermophilus. It consists of 60 subunits that form 20 trimers,

shown here in different colors.The positions of a 2-fold, a 3-fold,

and a 5-fold axis of symmetry are indicated.The inset (upper

right) is a perspective drawing of a dodecahedron.The rear

portion of each complex, which is largely eclipsed by the front

portion, is gray. The subunits forming the trimers closest to the

viewer in each drawing are individually colored with the lower

trimer drawn in ribbon form. Note that the subunits in each

trimer are extensively associated and that the trimers in the two

complexes have closely similar structures. In contrast, the

interactions between contacting trimers in both types of

complexes are relatively tenuous.Also note that these contacting

trimers form the 4- and 5-membered rings that comprise the

square and pentagonal faces of the cubic and dodecahedral

complexes, respectively. [Based on X-ray structures by Wim Hol,

University of Washington. PDBids 1EAB and 1B5S.]

(a)

(b)

(a)

(b)

JWCL281_c21_789-822.qxd 6/5/10 9:53 AM Page 793

eukaryotes and some gram-positive bacteria have an analo-

gous dodecahedral form [Fig. 21-5b; a dodecahedron is a

regular polyhedron with I symmetry (Section 8-5B) that has

20 vertices, each lying on a 3-fold axis, and 12 pentagonal

faces having an aggregate of 30 edges]. Thus, the mitochon-

drially located ⬃10,000-kD eukaryotic complex, the largest

known multienzyme complex, consists of a dodecahedral

core of 20 E2 trimers (one centered on every vertex) sur-

rounded by 30 E1 ␣

␤

heterotetramers (one centered on

every edge) and 12 E3 dimers (one centered in every face).

b. Multienzyme Complexes Are Catalytically Efficient

Multienzyme complexes are a step forward in the evolu-

tion of catalytic efficiency.They offer the following mecha-

nistic advantages:

1. Enzymatic reaction rates are limited by the frequency

at which enzymes collide with their substrates (Section 14-

2Ba). If a series of reactions occurs within a multienzyme

complex,the distance that substrates must diffuse between ac-

tive sites is minimized, thereby achieving a rate enhancement.

2. Complex formation provides the means for channel-

ing (passing) metabolic intermediates between successive

enzymes in a metabolic pathway, thereby minimizing side

reactions.

3. The reactions catalyzed by a multienzyme complex

may be coordinately controlled.

c. Acetyl-CoA Formation Occurs in Five Reactions

The PDC catalyzes five sequential reactions (Fig. 21-6)

with the overall stoichiometry:

The coenzymes and prosthetic groups required in this reac-

tion sequence are thiamine pyrophosphate (TPP;Fig.17-26),

flavin adenine dinucleotide (FAD; Fig. 16-8), nicotinamide

adenine dinucleotide (NAD

⫹

; Fig. 13-2), and lipoamide

acetyl-CoA ⫹ CO

⫹ NADH

Pyruvate ⫹ CoA ⫹ NAD

⫹

(Fig. 21-7); their functions are listed in Table 21-1.

Lipoamide consists of lipoic acid joined in amide linkage to

the ε-amino group of a Lys residue. Reduction of its cyclic

disulfide to a dithiol, dihydrolipoamide, and its reoxidation

(Fig. 21-7) are the “business” of this prosthetic group.

The five reactions catalyzed by the PDC are as follows

(Fig. 21-6):

1. Pyruvate dehydrogenase (E1), a TPP-requiring enzyme,

decarboxylates pyruvate, with the intermediate formation of

hydroxyethyl-TPP.This reaction is identical with that catalyzed

by yeast pyruvate decarboxylase (Section 17-3B):

R⬘

SE1

E1TPP

Pyruvate

Hydroxyethyl-

TPP

CCCH

HOC CH

R⬘

SE1

R⬘

SE1

OCCCH

O OH

794 Chapter 21. Citric Acid Cycle

C C

–

pyruvate

dehydrogenase

(E1)

–

TPPCH

TPP

Hydroxyethyl-

TPP

CCH

Lipoamide

Acetyl-dihydrolipoamide

dihydrolipoyl

transacetylase

(E2)

dihydrolipoyl

dehydrogenase

(E3)

CoA

CCH

S CoA

Acetyl-CoA

FAD

NADH

+ H

NAD

Pyruvate

Figure 21-6 The five reactions of the PDC. E1 (pyruvate

dehydrogenase) contains TPP and catalyzes Reactions 1 and 2.

E2 (dihydrolipoyl transacetylase) contains lipoamide and

catalyzes Reaction 3. E3 (dihydrolipoyl dehydrogenase) contains

FAD and a redox-active disulfide and catalyzes Reactions 4

and 5.

JWCL281_c21_789-822.qxd 3/17/10 11:33 AM Page 794

CoA

Acetyl-CoA

Acetyl-

dihydrolipoamide–E2

Dihydrolipoamide–E2

2. Unlike pyruvate decarboxylase, however, pyruvate

dehydrogenase does not convert the hydroxyethyl-TPP in-

termediate into acetaldehyde and TPP. Instead, the hy-

droxyethyl group is transferred to the next enzyme in the

multienzyme sequence, dihydrolipoyl transacetylase (E2).

The reaction occurs by attack of the hydroxyethyl group

carbanion on the lipoamide disulfide, followed by the

elimination of TPP from the intermediate adduct to form

acetyl-dihydrolipoamide and regenerate active E1. The

hydroxyethyl carbanion is thereby oxidized to an acetyl

group by the concomitant reduction of the lipoamide

disulfide bond:

CR⬘

Lipoamide-E2

Reaction 1

S • E1

CR⬘

S • E1

Acetyl-

dihydrolipoamide-E2

CR⬘

S • E1

TPP • E1

3. E2 then catalyzes the trans-

fer of the acetyl group to CoA,

yielding acetyl-CoA and dihy-

drolipoamide–E2:

This is a transesterification in which the sulfhydryl group of

CoA attacks the acetyl group of acetyl-dihydrolipoamide–

E2 to form a tetrahedral intermediate (not shown), which

decomposes to acetyl-CoA and dihydrolipoamide–E2.

Section 21-2. Metabolic Sources of Acetyl-Coenzyme A 795

Figure 21-7 Interconversion of lipoamide and

dihydrolipoamide. Lipoamide is lipoic acid covalently joined to

the ε-amino group of a Lys residue via an amide linkage.

C NH

C O

(CH

)

Lipoamide

Lipoic acid Lys

–

C NH

C O

(CH

)

Dihydrolipoamide

Table 21-1 The Coenzymes and Prosthetic Groups of Pyruvate Dehydrogenase

Cofactor Location Function

Thiamine pyrophosphate (TPP) Bound to E1 Decarboxylates pyruvate, yielding a

hydroxyethyl-TPP carbanion

Lipoic acid Covalently linked to a Lys on Accepts the hydroxyethyl carbanion from

E2 (lipoamide) TPP as an acetyl group

Coenzyme A (CoA) Substrate for E2 Accepts the acetyl group from acetyl-

dihydrolipoamide

Flavin adenine dinucleotide (FAD) Bound to E3 Reduced by dihydrolipoamide

Nicotinamide adenine dinucleotide (NAD

⫹

) Substrate for E3 Reduced by FADH

JWCL281_c21_789-822.qxd 3/17/10 11:33 AM Page 795

~35 residues

COO

–

~80 residues each

Lipoyl

domains

Outer linker Inner linker

Catalytic

domain

Peripheral

subunit-binding

domain

(

~250 residues

FAD

E3 (oxidized)

FAD

FADH

NAD

NADH

Reaction 4

4. Dihydrolipoyl dehydrogenase (E3; also called dihy-

drolipoamide dehydrogenase) reoxidizes dihydrolipo-

amide, thereby completing the catalytic cycle of E2:

Oxidized E3 contains a reactive disulfide group and a

tightly bound FAD.The oxidation of dihydrolipoamide is a

disulfide interchange reaction (Section 9-1A):The lipoamide

disulfide bond forms with concomitant reduction of E3’s

reactive disulfide to two sulfhydryl groups.

5. Reduced E3 is reoxidized by NAD

⫹

The enzyme’s active sulfhydryl groups

are reoxidized by the enzyme-bound

FAD, which is thereby reduced to

FADH2.The FADH2 is then reoxidized

to FAD by NAD1, producing NADH.

d. The Structure of E2

Dihydrolipoyl transacetylase (E2) consists of several do-

mains (Fig. 21-8): one to three N-terminal lipoyl domains

(⬃80 residues) that each covalently bind a lipoyl group; a

peripheral subunit-binding domain (PSBD; ⬃35 residues)

that binds to both E1 and E3 and hence holds the complex

together; and a C-terminal catalytic domain (⬃250

residues) that contains the enzyme’s catalytic center and its

intersubunit binding sites.These domains are linked by 20- to

(oxidized)

FAD

(reduced)

FAD

Reaction 2

40-residue Pro- and Ala-rich segments that are largely ex-

tended and highly flexible and thereby provide the lipoyl

domains with the mobility they require to interact with E1

and E3 and with neighboring E2 subunits (see below).

The hollow cagelike structures formed by the E2 cat-

alytic domains (Fig. 21-5) contain channels large enough to

allow substrates to diffuse in and out. In fact, coenzyme A

and lipoamide bind in extended conformations at opposite

ends of a 30-Å-long channel that is located at the interface

between each pair of the subunits in a trimer (Fig. 21-9).

This arrangement requires CoA to approach its binding

site from inside the cage.

The NMR structures of the lipoyl domains of E2 from

several sources verify the exposed nature of the site of the

lipoyl group.They consist of a ␤ barrel/sandwich containing

a 4-stranded and a 3-stranded antiparallel ␤ sheet, with the

Lys to which the lipoyl group would be attached extending

from an exposed position on a type I ␤ bend linking two of

the ␤ strands in the 4-stranded sheet (Fig. 21-10a).

The NMR structure of the PSBD from B. stearother-

mophilus, determined by Richard Perham, reveals that its

⬃35-residue ordered region consists of two parallel helices

separated by a loop that form a close-packed hydrophobic

796 Chapter 21. Citric Acid Cycle

Figure 21-8 Domain structure of the dihydrolipoyl

transacetylase (E2) subunit of the PDC. The number of lipoyl

domains, n, is species dependent: n ⫽ 3 for E. coli and A.

vinelandii, n ⫽ 2 for mammals and Streptococcus faecalis, and

n ⫽ 1 for B. stearothermophilus and yeast.

Figure 21-9 X-ray structure of a trimer of A. vinelandii

dihydrolipoyl transacetylase (E2) catalytic domains in complex

with CoA and lipoamide. The trimer, a portion of the cubic

complex (Fig. 21-5a), is drawn in ribbon form and viewed along

its 3-fold axis (the cubic complex’s body diagonal) from inside

the cube. The three identical domains have different colors with

the upper domain colored in rainbow order from its N-terminus

(blue) to its C-terminus (red). The CoA and lipoamide bound to

the upper and lower left subunits are drawn in space-filling form

and those bound to the lower right subunit are drawn in stick form,

with CoA C green, lipoamide C cyan, N blue, O red, P orange, and

S yellow. Note how the N-terminal “elbow” of each subunit

extends behind the counterclockwise adjacent subunit; its deletion

greatly destabilizes the complex. [Based on an X-ray structure by

Wim Hol, University of Washington. PDBid 1EAB.]

JWCL281_c21_789-822.qxd 6/5/10 9:53 AM Page 796

core (Fig. 21-10b). It is one of the smallest known polypep-

tides that have a globular structure but lack disulfide bonds

or prosthetic groups.

e. Intermediates Are Transferred between Enzyme

Subunits by Flexible Tethers

How are reaction intermediates transferred between

the component enzymes of the PDC? The group between

the lipoamide disulfide bond and the E2 polypeptide back-

bone, the so-called lipoyllysyl arm, has a fully extended

length of 14 Å:

S S

14 Å

Lipoyllysyl arm

(fully extended)

Evidently, the lipoyllysyl arm, in combination with the

⬎140-Å-long outer linker (Fig. 21-8), acts as a long flexible

tether that swings its attached lipoyl group among the active

sites of E1, E2, and E3. Moreover, there is rapid inter-

change of acetyl groups among the lipoyl groups of the

E2 core (3 ⫻ 24 ⫽ 72 lipoyl groups in an E. coli PDC;

2 ⫻ 60 ⫽ 120 in mammals); the tethered arms also swing

among themselves, exchanging both acetyl groups and

disulfides:

One E1 subunit can therefore acetylate numerous E2 sub-

units and one E3 subunit can reoxidize several dihy-

drolipoamide groups.

SSSSH

E2 E2⬘

SSSHS

E2 E2⬘

HS S SS

E2 E2⬘

SS SSH

E2 E2⬘

Section 21-2. Metabolic Sources of Acetyl-Coenzyme A 797

Figure 21-10 NMR structures of the lipoyl

and peripheral subunit-binding domains of

dihydrolipoyl transacetylase (E2). (a) The

NMR structure of the A. vinelandii

dihydrolipoyl transacetylase (E2) lipoyl

domain.The polypeptide chain is colored in

rainbow order from its N-terminus (blue) to its

C-terminus (red).The Lys side chain to which the

lipoyl group would be bound is shown in stick

form colored according to atom type (C green,

H white, and N blue) and is located in a type I ␤

turn. [Based on an NMR structure by Aart de

Kok, Wageningen Agricultural University,

Wageningen, Netherlands. PDBid 1IYU.]

(b) The NMR structure of the peripheral

subunit-binding domain (PSBD) from B.

stearothermophilus E2.The polypeptide chain is

colored in rainbow order from its N-terminus

(blue) to its C-terminus (red). [Based on an

NMR structure by Richard Perham, Cambridge

University, U.K. PDBid 2PDD.]

(a)

(b)

JWCL281_c21_789-822.qxd 6/5/10 9:53 AM Page 797

f. Mammalian and Yeast PDCs Contain

Additional Subunits

In mammals and yeast, the dodecahedral PDC’s already

complicated structure has a further level of complexity in

that it contains additional subunits: About 12 copies of E3

binding protein (E3BP) facilitate the binding of E3 to the

E2 core of the eukaryotic dodecahedral complex. E3BP

has a lipoyllysine-containing domain similar to E2 and can

accept an acetyl group, but its C-terminal domain has no

catalytic activity, and the removal of its lipoyllysine domain

does not diminish the catalytic activity of the complex.

E3BP’s main role seems to be to aid in the binding of E3,

since limited proteolysis of E3BP decreases E3’s binding

ability.

g. Cryoelectron Microscopy Reveals the

Structure of the Dodecahedral PDC

James Stoops and Reed have determined the organiza-

tion of the bovine kidney dodecahedral complex through

cryoelectron microscopy (Fig. 21-11). As expected, the E2

subunits form a dodecahedral core that is surrounded by a

concentric dodecahedron of E1 subunits (Fig. 21-11a,b).

The E2 subunits bind the E1 subunits via the radially ex-

tending, ⬃50-Å-long inner linkers that precede the PSBD

of E2 (Fig. 21-8; these linkers are not present in Fig. 21-5b

because that X-ray structure only contains E2’s catalytic

domain). Although the bovine kidney PDC used in the

structure determination lacked sufficient E3BP ⴢ E3 to be

visible, its position in the electron microscopy–based struc-

ture of yeast PDC indicates that an E3 dimer occupies and

largely fills each pentagonal opening of the E2 core

(Fig. 21-11c). Hence the E3 subunits are not arranged with

dodecahedral symmetry (similarly, the E3 subunits in the

cubic E. coli PDC are not arranged with cubic symmetry;

798 Chapter 21. Citric Acid Cycle

Figure 21-11 Electron microscopy–based images of the bovine

kidney pyruvate dehydrogenase complex at ⬃35 Å resolution.

(a) The entire particle as viewed along its 3-fold axis of

symmetry. E1 is yellow, the E2 catalytic core is green, and the

inner linkers that connect E2’s catalytic domains to its

E1-binding domains, the PSBDs (Fig. 21-8), are cyan.The particle

is viewed along a 3-fold axis of symmetry with the positions of a

5-fold axis and a 2-fold axis also marked. (b) A cutaway diagram

viewed and colored as in Part a but with the particle’s closest half

removed to reveal the E2 catalytic core and the inner linkers.

Compare the green portion with the X-ray structure of a

dodecahedral E2 catalytic core as viewed along a 2-fold axis

(Fig. 21-5b). (c) A cutaway diagram as in Part b but with E3

dimers (Fig. 21-13a) shown at 20 Å resolution (red) modeled

into the pentagonal openings of the E2 core. The position of a

peripheral subunit-binding site at the end of an E2 inner linker is

marked by an asterisk (*). [From Zhou, Z.H., McCarthy, D.B.,

O’Connor, C.M., Reed, L.J., and Stoops, J.K., Proc. Natl. Acad.

Sci. 98, 14802 (2001).]

(a)

(b)

(c)

JWCL281_c21_789-822.qxd 3/17/10 11:33 AM Page 798

Fig. 21-4b).The PSBD, which is located at the end of the E2

inner linker (* in Fig. 21-11c), is the point from which the

E2 lipoyl domains (Fig. 21-10a) swing. It is ⬃50 Å from the

nearest active sites of E1, E2, and E3.

In addition to E3BP, the mammalian PDC contains one

to three copies each of pyruvate dehydrogenase kinase and

pyruvate dehydrogenase phosphatase. The kinase and

phosphatase function to regulate the catalytic activity of

the complex (Section 21-2Cb).

h. Two Other Multienzyme Complexes Closely

Resemble the PDC

In addition to the PDC, most cells contain two other

closely related multienzyme complexes: the ␣-ketoglu-

tarate dehydrogenase complex [also called 2-oxoglutarate

dehydrogenase (OGDH), which catalyzes Reaction 4 of

the citric acid cycle; Fig. 21-1] and the branched-chain

␣-keto acid dehydrogenase complex (BCKDH; which par-

ticipates in the degradation of isoleucine, leucine, and va-

line; Section 26-3Ee). These multienzyme complexes all

catalyze similar reactions: the NAD

⫹

-linked oxidative de-

carboxylation of an ␣-keto acid with the transfer of the re-

sulting acyl group to CoA.In fact, all three members of this

2-ketoacid dehydrogenase family (alternatively, the 2-oxo

acid dehydrogenase family) of multienzyme complexes

share the same E3 subunit, and their E1 and E2 subunits,

which are specific for their corresponding substrates, are

homologous and use identical cofactors.Thus, to differenti-

ate them, the E1s for PDC,OGDH, and BCKDH are often

referred to as E1p, E1o, and E1b, respectively, and like-

wise, their E3s are called E3p, E3o, and E3b.

i. Arsenic Compounds Are Poisonous because

They Sequester Lipoamide

Arsenic has been known to be a poison since ancient

times.As(III) compounds, such as arsenite (AsO

3⫺

) and or-

ganic arsenicals, are toxic because of their ability to cova-

lently bind sulfhydryl compounds. This is particularly true

of vicinal (adjacent) sulfhydryls such as those of lipoamide

because they can form bidentate adducts:

The resultant inactivation of lipoamide-containing enzymes,

especially pyruvate dehydrogenase and ␣-ketoglutarate

dehydrogenase (Section 21-3D), brings respiration to a

halt.

Organic arsenicals are more toxic to microorganisms

than to humans, apparently because of differences in the

sensitivities of their various enzymes to these compounds.

This differential toxicity is the basis for the early twentieth

century use of organic arsenicals in the treatment of

syphilis (now superseded by penicillin) and trypanosomia-

sis (typanosomes are parasitic protozoa that cause several

diseases including African sleeping sickness and Chagas-

Cruz disease). These compounds were really the first

antibiotics, although, not surprisingly, they had severe side

effects.

j. Arsenic Poisoning, Napoleon, and Darwin

Arsenic is often suspected as a poison in untimely

deaths. In fact, it has long been thought that Napoleon

Bonaparte died from arsenic poisoning while in exile on

St. Helena, an island in the Atlantic Ocean. This suspicion,

and the chemical analyses it sparked, makes a fascinating

chemical anecdote. The finding that a lock of Napoleon’s

hair indeed contains a high level of arsenic strongly sup-

ports the notion that arsenic poisoning at least contributed

to his death. But was it murder or environmental pollu-

tion? A sample of the wallpaper from Napoleon’s drawing

room was found to contain the commonly used (at the

time) green pigment copper arsenate (CuHAsO

). It was

eventually determined that in a damp climate, as occurs on

St. Helena, fungi growing on the wallpaper eliminate the

arsenic by converting it to the volatile and highly toxic

trimethyl arsine [(CH

)

As]. Indeed, Napoleon’s regular

visitors also suffered from symptoms of arsenic poisoning

(e.g., gastrointestinal disturbances), which appeared to

moderate when they spent much of their time outdoors.

Thus, Napoleon’s arsenic poisoning may have been unin-

tentional.

Retrospective detective work also suggests that

Charles Darwin was a victim of chronic arsenic poison-

ing. For most of his life after he returned from his epic

voyage, Darwin complained of numerous ailments, in-

cluding eczema, vertigo, headaches, arthritis, gout, palpi-

tations, and nausea, all symptoms of arsenic poisoning.

Fowler’s solution, a common nineteenth century tonic,

contained 10 mg of arsenite ⴢ mL

⫺1

. Many individuals,

quite possibly Darwin himself, took this “medication” for

years.

k. The Structure of E1

In dodecahedral PDCs, E1 (pyruvate dehydrogenase) is

a 2-fold symmetric ␣

␤

heterotetramer. However, in cubic

PDCs and in ␣-ketoglutarate dehydrogenase, which is also

cubic, the ␣ and ␤ subunits are genetically fused to form a

single polypeptide and hence a homodimeric enzyme.

The X-ray structure of E1 from the (dodecahedral)

B. stearothermophilus PDC in complex with TPP and a

Section 21-2. Metabolic Sources of Acetyl-Coenzyme A 799

–

Arsenite

R R

–

O As

+ 2 H

AsR⬘ +

R R

R⬘ As

+ H

Dihydro-

lipoamide

Organic

arsenical

JWCL281_c21_789-822.qxd 3/17/10 11:33 AM Page 799

PSBD (Fig. 21-12a), determined by Perham and Ben

Luisi, reveals a tightly associated ␣

␤

heterotetramer,

whose 368-residue ␣ subunits and 324-residue ␤ sub-

units each consist of two domains. A single PSBD binds

to the C-terminal domains of the ␤ subunits along the 2-

fold axis relating them in a way that sterically prevents

the binding of a second PSBD. Hence the PSBD interacts

asymmetrically with the two chemically identical do-

mains in a manner reminiscent of the way that human

growth hormone interacts with its homodimeric receptor

(Section 19-1J). E1’s conserved ⬃310-residue core occurs

in other TPP-utilizing enzymes of known structure, in-

cluding pyruvate decarboxylase (Section 17-3Ba). Each

TPP is bound between the N-terminal domains of an ␣

and a ␤ subunit at the end of an ⬃21-Å-deep funnel-

shaped channel with its thiazolium ring (its reactive

group) closest to the channel entrance. Apparently, the

lipoyllysyl arm at the end of an E2 lipoyl domain is in-

serted into this channel in an extended conformation for

the transfer of the TPP’s hydroxyacyl substituent to

lipoamide.

l. E1 Active Sites Are Coordinated via a Proton Wire

When apo-E1 is mixed with TPP, the first TPP to bind

to this apparently 2-fold symmetric protein does so or-

ders of magnitude faster than the second TPP. This indi-

cates that these two binding sites, which are ⬃20 Å apart,

somehow communicate. In fact, the enzyme’s two active

sites in the foregoing X-ray structure have different

structures in that two conserved loops (residues 203–212

and 276–287 of the ␣ subunits) at the entrance to the

channel leading to TPP are ordered in one subunit in a

way that blocks the active site but are disordered in the

other subunit.That this is not an artifact of crystallization

was demonstrated by the observation that limited prote-

olysis cleaves the more C-terminal of these loops on one

subunit, but not the other. Similar asymmetries have

been observed in all TPP-dependent enzymes of known

structure.

The X-ray structure of E1 reveals a solvent-filled tunnel

linking its two active sites that is lined with 10 conserved

acidic residues (Fig. 21-12b). Similar acidic tunnels are

present in all TPP-dependent enzymes of known structure,

all of which are either dimeric or tetrameric.This led Luisi

and Perham to propose that as pyruvate proceeds through

the E1-catalyzed Reactions 1 and 2 discussed in Section 21-

2Ac, the proton required by Reaction 1 in one active site is

supplied by the proton released in Reaction 2 in the other

active site. The protons are presumably transferred be-

tween the two active sites via series of proton jumps

(Section 2-1C) through the acidic tunnel, which thereby

functions as a proton wire. Consequently, as the two active

sites progress through their catalytic cycles, they must

remain out of phase with one another; that is, one site re-

quires a general base when the other site requires a general

acid and vice versa so that they reciprocate their catalytic

needs via the proton wire.This explains the observed nega-

tive cooperativity of substrate binding exhibited by E1.

Structurally, this is rationalized by the alternate closing and

opening of the active sites through the out of phase order-

ing and disordering of the loops at the entrance of each

active site channel.

800 Chapter 21. Citric Acid Cycle

Figure 21-12 X-ray structure of B. stearothermophilus E1 in

complex with TPP and the peripheral subunit-binding domain

(PSBD). (a) The E1 ␣

␤

heterotetramer, viewed with its 2-fold

axis of symmetry vertical and the E2 dodecahedron below, is

drawn with its ␣ subunits in worm form and its ␤ subunits in

ribbon form.The ␣ and ␤ subunits that are closest to the viewer

and predominantly on the left are each colored in rainbow order

from their N-termini (blue) to their C-termini (red); the remaining

␣ and ␤ subunits are pale cyan and pale green. The PSBD (below)

is light blue. The TPP is shown in space-filling form with C green,

N blue, O red, P orange, and S yellow. The entrances to the

enzyme’s active site channels are indicated by black arrows. (b) A

cutaway surface diagram of the solvent-filled tunnel connecting

the enzyme’s two active sites. The hydroxyethyl-TPPs bound at

the enzyme’s active sites and the Asp and Glu side chains lining

the tunnel are drawn in stick form with C gray, N blue, O red, P

green, and S yellow. Water molecules and Mg

2⫹

ions are

represented by red and magenta spheres. [Part a based on an

X-ray structure by and Part b courtesy of Richard Perham and

Ben Luisi, University of Cambridge, U.K. PDBid 1W85.]

(a)

(b)

JWCL281_c21_789-822.qxd 7/2/10 9:05 AM Page 800