Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

mechanism of PLP participation in the phosphorylase re-

action must differ from that in these other enzymes be-

cause, for example, reduction of the Schiff base with

NaBH

(HCN

C NH ) has no effect on

the activity of phosphorylase, whereas it inactivates the

PLP-requiring enzymes of amino acid metabolism. This is

¬¬¬¬“¬

an intriguing example of nature’s opportunism in using the

same cofactor to perform different chemistries.

Extensive studies on phosphorylase using PLP analogs

in which various parts of this molecule are missing or mod-

ified indicate that only its phosphate group participates in

the catalytic process. Indeed, the X-ray structures of phos-

phorylase reveal that only PLP’s phosphate group is near

this enzyme’s active site. This phosphoryl group most prob-

ably functions as an acid–base catalyst.

c. Kinetics and Reaction Mechanism

The phosphorylase reaction results in the cleavage of

the C1 O1 bond from a nonreducing terminal glucosyl

unit of glycogen, yielding G1P. This reaction proceeds with

retention of configuration, which suggests that the phos-

phorolysis occurs via a double displacement mechanism

(two sequential nucleophilic substitutions, each occurring

with inversion of configuration; Fig. 16-6c) involving a

covalent glucosyl–enzyme intermediate. Yet, phosphory-

lase exhibits Rapid Equilibrium Random Bi Bi kinetics

(Section 14-5), not Ping Pong kinetics, as would be

expected for a double displacement mechanism. Further-

more, all attempts to establish the existence of the putative

covalent intermediate have been unsuccessful.

An alternative mechanism (Fig. 18-3), which is compat-

ible with the available kinetic, chemical, and structural

data, commences with the formation of a ternary enzyme ⴢ

ⴢ glycogen complex, followed by the generation of an

intermediate shielded oxonium ion similar to the transi-

tion state in the lysozyme reaction (which also involves

Section 18-1. Glycogen Breakdown 641

Figure 18-3 The reaction mechanism of glycogen

phosphorylase. Here PL is an enzyme-bound pyridoxal group

and BH



is a positively charged amino acid side chain, probably

that of Lys 568, necessary for the maintenance of PLP electrical

neutrality. (1) Formation of an E ⴢ P

ⴢ glycogen ternary complex.

(2) Formation of a shielded oxonium ion intermediate from the

-linked terminal glucosyl residue involving acid catalysis by P

as facilitated by proton transfer from PLP. The oxonium ion has

the half-chair conformation. (3) Reaction of P

with the oxonium

ion with overall retention of configuration about C1 to form

-

D-glucose-1-phosphate. The glycogen, which has one less

residue than before, cycles back to Step 1.

Nonreducing end

Glycogen

(n glucosyl units)

Glycogen

(n – 1 glucosyl units)

Half-chair oxonium ion

intermediate

+ Glycogen

–

α(1 4) linkage

…

–

α linkage

to phosphate

JWCL281_c18_638-670.qxd 2/26/10 2:24 PM Page 641

glycosidic bond cleavage in a polysaccharide; Section 15-2B).

Bond cleavage, with its consequent oxonium ion formation,

is assisted by protonation of the glycosidic oxygen by the P

substrate (acid catalysis). Phosphorylase has no protein

nucleophilic or carboxylate groups in the vicinity of the

scissile glycosidic bond and hence would be unable to

form a covalent intermediate as does lysozyme. However,

since the PLP phosphoryl group is within hydrogen bond-

ing distance of the P

, it appears that bond cleavage is facil-

itated by the simultaneous protonation of the reacting P

the PLP phosphoryl group in a kind of proton relay.The re-

sulting oxonium ion (Fig. 18-3) is stabilized through its for-

mation of an ion pair with the anionic P

(electrostatic

catalysis), which subsequently collapses to yield product,

G1P, in a reaction step that is facilitated by the abstraction

of a proton from P

by the PLP phosphoryl group (base

catalysis).

Support for the oxonium ion mechanism comes from

the observation that 1,5-gluconolactone

is a potent inhibitor of phosphorylase. 1,5-Gluconolactone

has the same half-chair conformation as the proposed oxo-

nium ion, suggesting that it is a transition state analog that

mimics the oxonium ion at the active site of phosphorylase

(Section 15-1F).

B. Phosphoglucomutase

Phosphorylase converts the glucosyl units of glycogen to

G1P, which, in turn, is converted by phosphoglucomutase

to G6P either for entry into glycolysis in muscle or hydrol-

1,5-Gluconolactone

ysis to glucose in liver. The X-ray structure of rabbit mus-

cle phosphoglucomutase indicates that the active site of

this 561-residue monomeric enzyme is largely buried at

the bottom of a particularly deep crevice in the protein.

The phosphoglucomutase reaction is similar to that cat-

alyzed by phosphoglycerate mutase (Section 17-2Ha). A

phosphoryl group is transferred from the active phospho-

enzyme to G6P, forming glucose 1,6-bisphosphate

(G1,6P), which then rephosphorylates the enzyme to yield

G1P (Fig. 18-4). An important difference between this en-

zyme and phosphoglycerate mutase is that the phosphoryl

group in phosphoglucomutase is covalently bound to a Ser

hydroxyl group rather than to a His imidazole nitrogen.

G1,6P occasionally dissociates from phosphoglucomu-

tase, resulting in the inactivation of this enzyme. The pres-

ence of small amounts of G1,6P is therefore necessary to

keep phosphoglucomutase fully active.This intermediate is

provided by phosphoglucokinase, which catalyzes the

phosphorylation of the C6 OH group of G1P by ATP.

C. Glycogen Debranching Enzyme

Glycogen debranching enzyme, an ⬃1540-residue monomer,

acts as an (1 S 4) transglycosylase (glycosyl transferase) by

transferring an (1 S 4)-linked trisaccharide unit from a

“limit branch”of glycogen to the nonreducing end of another

branch (Fig. 18-5). This reaction forms a new (1 S 4) link-

age with three more units available for phosphorylase-

catalyzed phosphorolysis. The (1 S 6) bond linking the

remaining glycosyl residue in the branch to the main chain

is hydrolyzed (not phosphorylized) by the same debranching

enzyme to yield glucose and debranched glycogen. Thus

debranching enzyme has different active sites for the trans-

ferase reaction and the (1 S 6)-glucosidase reaction. The

presence of two independent catalytic activities on the

same enzyme no doubt improves the efficiency of the de-

branching process.

642 Chapter 18. Glycogen Metabolism

Figure 18-4 The mechanism of action of phosphoglucomutase.

(1) The OH group at C1 of G6P attacks the phosphoenzyme to

form a dephosphoenzyme–G1,6P intermediate. (2) The Ser OH¬

–2

POCH

–2

POCH

OH H

HOH

–

Ser

Enzyme

Glucose-6-

phosphate

(G6P)

OH H

HOH

CH OH

Ser

Enzyme

Glucose-1,6-

bisphosphate

(G1,6P)

HOCH

OH H

HOH

–

Ser

Enzyme

Glucose-1-

phosphate

(G1P)

OPO

–

OPO

–

group on the dephosphoenzyme attacks the phosphoryl group

at C6 to regenerate the phosphoenzyme with the formation of

G1P.

JWCL281_c18_638-670.qxd 2/26/10 2:24 PM Page 642

The maximal rate of the glycogen phosphorylase reac-

tion is much greater than that of the glycogen debranching

reaction. Consequently, the outermost branches of gly-

cogen, which comprise nearly half of its residues, are

degraded in muscle in a few seconds under conditions of

high metabolic demand. Glycogen degradation beyond this

point requires debranching and hence occurs more slowly.

This, in part, accounts for the fact that a muscle can sustain

its maximum exertion for only a few seconds.

D. Thermodynamics of Glycogen Metabolism:

The Need for Separate Pathways of Synthesis

and Breakdown

The G°¿ (G under standard biochemical conditions) for

the phosphorylase reaction is 3.1 kJ ⴢ mol

1

, so, as Eq.

[3.15] indicates, this reaction is at equilibrium (G  0) at

25°C when [P

]冫[G1P]  3.5. In the cell, however, this

concentration ratio varies between 30 and 100, which

places G in the range 5 to 8 kJ ⴢ mol

1

; that is, under

physiological conditions, glycogen breakdown is exer-

gonic. The synthesis of glycogen from G1P under physio-

logical conditions is therefore thermodynamically unfa-

vorable without free energy input. Consequently, glyco-

gen biosynthesis and breakdown must occur by separate

pathways. Thus we encounter a recurrent metabolic strat-

egy: Biosynthetic and degradative pathways of metabolism

are almost always different (Section 16-1). There are two

important reasons for this. The first, as we have seen, is

that both pathways may be required under similar in vivo

metabolite concentrations. This situation is thermody-

namically impossible if one pathway is just the reverse of

the other. The second reason is equally important: Reac-

tions catalyzed by different enzymes can be independ-

ently regulated, which permits very fine flux control. We

have seen this principle in operation in the glycolytic con-

version of fructose-6-phosphate (F6P) to fructose-1,6-

bisphosphate (F1,6P) by phosphofructokinase (PFK;

Section 17-4F). The reverse process in that case (hydroly-

sis of F1,6P) is catalyzed by fructose bisphosphatase

(FBPase). Independent control of those two enzymes

provides precise regulation of glycolytic flux.

Glycogen metabolism, like glycolysis, is exquisitely reg-

ulated by the independent control of its synthetic and

Section 18-1. Glycogen Breakdown 643

glycogen debranching

enzyme

Limit branch

Available for

hydrolysis

Available for

further phosphorolysis

Outer glycogen chains

(after phosphorylase action)

...

Figure 18-5 Reactions catalyzed by debranching enzyme. The

enzyme transfers the terminal three (1 S 4)-linked glucose

residues from a “limit branch” of glycogen to the nonreducing

end of another branch.The (1 S 6) bond of the residue

remaining at the branch point is hydrolyzed by further action of

debranching enzyme to yield free glucose. The newly elongated

branch is subject to degradation by glycogen phosphorylase.

JWCL281_c18_638-670.qxd 2/26/10 2:24 PM Page 643

degradative pathways. In the next section we examine the

pathway of glycogen synthesis and, in Section 18-3, we ex-

plore the regulatory process.

2 GLYCOGEN SYNTHESIS

Although the thermodynamic arguments presented in Sec-

tion 18-1D demonstrate that glycogen synthesis and break-

down must occur by separate pathways, it was not thermo-

dynamic arguments that led to the general acceptance of

this idea. Rather, it was the elucidation of the cause of

McArdle’s disease, a rare inherited glycogen storage dis-

ease that results in painful muscle cramps on strenuous

exertion (Section 18-4).The muscle tissue from individuals

with McArdle’s disease exhibits no glycogen phosphory-

lase activity and is therefore incapable of glycogen break-

down.Their muscles, nevertheless, contain moderately high

quantities of normal glycogen. Clearly, there must be sepa-

rate pathways for glycogen synthesis and breakdown.

Since the direct conversion of G1P to glycogen and P

thermodynamically unfavorable (positive G) under all

physiological P

concentrations, glycogen biosynthesis re-

quires an additional exergonic step. This is accomplished,

as Luis Leloir discovered in 1957, by combining G1P with

uridine triphosphate (UTP) to form uridine diphosphate

glucose (UDP–glucose or UDPG):

UDPG’s “high-energy” status permits it to spontaneously

donate glucosyl units to the growing glycogen chain.

The enzymes catalyzing the three steps involved in the

glycogen synthesis pathway are UDP–glucose pyrophos-

phorylase, glycogen synthase, and glycogen branching

enzyme. In this section, we examine the reactions catalyzed

by these enzymes. Discussion of how these enzymes are

controlled is reserved for Section 18-3.

A. UDP–Glucose Pyrophosphorylase

UDP–glucose pyrophosphorylase catalyzes the reaction of

UTP and G1P (Fig. 18-6). In this reaction, the phosphoryl

oxygen of G1P attacks the  phosphorus atom of UTP to

form UDPG and release PP

. The G°¿ of this phosphoan-

hydride exchange is, as expected, nearly zero. However,

the PP

formed is hydrolyzed in an exergonic reaction by

the omnipresent enzyme inorganic pyrophosphatase. The

overall reaction for the formation of UDPG is therefore

also exergonic:

⬃0

Overall:

The cleavage of a nucleoside triphosphate to form PP

is a

common biosynthetic strategy. The free energy of PP

hy-

drolysis can then be utilized together with the free energy of

nucleoside triphosphate hydrolysis to drive an otherwise

endergonic reaction to completion (Section 16-4Ca).

19.2G1P  UTP Δ UDPG  2P

19.2H

O  PP

Δ 2P

G1P  UTP Δ UDPG  PP

¢G°¿ (kJ ⴢ mol

1

)

–

Uridine diphosphate glucose

(UDPG)

644 Chapter 18. Glycogen Metabolism

Figure 18-6 Reaction catalyzed by UDP–glucose

pyrophosphorylase. The reaction is a phosphoanhydride

exchange in which the phosphoryl oxygen of G1P attacks the 

phosphorus atom of UTP to form UDPG and release PP

.The

is rapidly hydrolyzed by inorganic pyrophosphatase.

HO OH

UTP

–



–

UDP–glucose

2 P

inorganic

pyrophos-

phatase

HO OH

–

O OH

G1P

–

JWCL281_c18_638-670.qxd 6/3/10 1:48 PM Page 644

B. Glycogen Synthase

In the next step of glycogen synthesis, the glycogen syn-

thase reaction, the glucosyl unit of UDPG is transferred to

the C4 OH group on one of glycogen’s nonreducing ends

to form an (1 S 4)-glycosidic bond (Fig. 18-7). The glyco-

gen synthase reaction, like those of glycogen phosphory-

lase and lysozyme, is thought to involve a glucosyl oxonium

ion intermediate or transition state since it is also inhibited

by 1,5-gluconolactone, an analog that mimics the oxonium

ion’s half-chair geometry.

The G°¿ for the glycogen synthase reaction is –13.4

kJ ⴢ mol

1

, making the overall reaction spontaneous under

the same conditions that glycogen breakdown by glycogen

phosphorylase is also spontaneous. The rates of both reac-

tions may then be independently controlled.There is, how-

ever, an energetic price for doing so. In this case, for each

molecule of G1P that is converted to glycogen and then re-

generated, one molecule of UTP is hydrolyzed to UDP and

.The cyclic synthesis and breakdown of glycogen is there-

fore not a perpetual motion “machine” but, rather, is an “en-

gine” that is powered by UTP hydrolysis. The UTP is

replenished through a phosphate-transfer reaction medi-

ated by nucleoside diphosphate kinase (Section 28-1Ba):

so that UTP hydrolysis is energetically equivalent to ATP

UDP  ATP Δ UTP  ADP

hydrolysis.This reaction occurs via a Ping Pong mechanism

in which an active site His residue is transiently phosphory-

lated at its N



position much as occurs in the phosphoglyc-

erate mutase reaction of glycolysis (Section 17-2Ha).

Mammals express two ⬃70% identical isoforms of

glycogen synthase, one mainly in muscle, and the other in

liver, much as is the case with glycogen phosphorylase.

Plants and bacteria express starch/glycogen synthases but

these employ ADP–glucose as glucose donors rather than

UDP–glucose and exhibit very little sequence similarity to

animal glycogen synthases.

a. Glycogenin Initiates Glycogen Synthesis

Glycogen synthase cannot simply link together two glu-

cose residues; it can only extend an already existing

(1 S 4)-linked glucan chain. How, then, is glycogen syn-

thesis initiated? The answer is that the first step in glycogen

synthesis is the self-catalyzed attachment of a glucose

residue to the Tyr 194 OH group of a 332-residue homo-

dimeric protein named glycogenin. Glycogenin, which was

discovered by William Whelan, further extends the glucan

chain by up to ⬃9 additional UDP–glucose-supplied

residues, forming a “primer” for the initiation of glycogen

synthesis. Only at this point does glycogen synthase com-

mence glycogen synthesis, which it initiates on the

“primer” while tightly complexed to glycogenin. However,

these proteins dissociate after the growing glycogen

Section 18-2. Glycogen Synthesis 645

Figure 18-7 Reaction catalyzed by glycogen synthase. The reaction involves a glucosyl

oxonium ion intermediate.

UDP

UDP–glucose

Glycogen (n residues)

Oxonium

ion intermediate

Glycogen (n + 1 residues)

HO OH

–

O OH

...

JWCL281_c18_638-670.qxd 2/26/10 2:24 PM Page 645

granule has reached some minimum size.Analysis of glyco-

gen granules for glycogenin and glycogen synthase shows

that they are present in a 1:1 ratio. Evidently, each glycogen

molecule (Fig. 18-1c) is associated with one molecule each

of glycogenin and glycogen synthase.

The X-ray structure of rabbit muscle glycogenin in com-

plex with UDPG (Fig. 18-8), determined by Thomas Hurley

and Peter Roach, reveals that the enzyme’s N-terminal por-

tion consists of a Rossmann-like fold (Section 8-3Bi) that is

common to the nucleotide-binding domains of most

glycosyltransferases.The OH group of its Tyr 194 is 21 and 16

Å distant from C1 of the UDPG’s glucosyl group on the

same and on the opposite subunit. Moreover, the glucosyl-

transferase reaction occurs with retention of configuration

about the glucosyl C1 atom. This suggests that the glucosyl

group is first transferred to an intermediate nucleophile and

then transferred to Tyr 194 or to a glucose group that is

linked to Tyr 194 (a double displacement reaction). Site-

directed mutagenesis has identified Asp 162 as the probable

nucleophile, whereas Asp 159 appears to have a role in

binding and activating the acceptor molecule. Evidently,

glycogenin undergoes large conformational changes during

its catalytic cycle, although the nature of these changes is

unknown. Indeed, it is unclear whether a given subunit

transfers glucosyl groups to itself or to the opposite subunit.

C. Glycogen Branching Enzyme

Glycogen synthase catalyzes only (1 S 4)-linkage forma-

tion to yield -amylose. Branching to form glycogen is

accomplished by an ⬃700-residue monomeric enzyme,

amylo-(1,4 S 1,6)-transglycosylase (glycogen branching

enzyme), which is distinct from glycogen debranching en-

zyme. Branches are created by the transfer of terminal

chain segments consisting of ⬃7 glucosyl residues to the

C6 OH groups of glucose residues on the same or an-

other glycogen chain (Fig. 18-9). Each transferred segment

must come from a chain of at least 11 residues, and the new

branch point must be at least 4 residues away from other

branch points.

Debranching (Section 18-1C) involves breaking and

reforming (1 S 4)-glycosidic bonds and only the hydroly-

sis of (1 S 6)-glycosidic bonds; branching, on the other

hand, involves breaking (1 S 4)-glycosidic bonds and

reforming (1 S 6) linkages. The need to hydrolyze

glycogen’s (1 S 6)-glycosidic bonds rather than to convert

them to (1 S 4) linkages is explained by the energetics of

these reactions. The free energy of hydrolysis of an

(1 S 4)-glycosidic bond is 15.5 kJ ⴢ mol

1

, whereas that

of an (1 S 6)-glycosidic bond is only 7.1 kJ ⴢ mol

1

.Con-

sequently, the hydrolysis of an (1 S 4)-glycosidic bond

drives the synthesis of an (1 S 6)-glycosidic bond, but the

reverse reaction is endergonic.

a. Glycogen Particles Are Fabricated to Optimize

Glucose Mobilization

The biological function of glycogen is to maximize the

density of stored glucose units consistent with the need to

rapidly mobilize it under conditions of high metabolic

demand. To do so, three related parameters must be opti-

mized: the number of tiers of branches in a glycogen mole-

cule, the number of branches per tier, and the average

chain length per tier. For a glycogen molecule with a fixed

646 Chapter 18. Glycogen Metabolism

Figure 18-8 X-ray structure of rabbit muscle glycogenin. The

upper subunit of this homodimeric enzyme is colored in rainbow

order from its N-terminus (blue) to its C-terminus (red) and the

lower subunit is pale green. UDPG is drawn in ball-and-stick

form with C green, N blue, O red, and P orange. The side chains

of Tyr 194, to which the glucose residue at glycogen’s reducing

end is linked, and Asp 159 and Asp 162, are drawn in stick form

with C cyan and O red. Mn

2

ions, which are thought to

electrostatically stabilize the leaving group, UDP, are represented

by lavender spheres. Residues 262 to 332, which are implicated in

binding glycogen synthase, are not visible in the X-ray structure

and hence disordered. [Based on an X-ray structure by Thomas

Hurley and Peter Roach, Indiana University School of Medicine.

PDBid 1LL2.]

JWCL281_c18_638-670.qxd 2/26/10 2:24 PM Page 646

number of residues, the number of outer branches from

which glucose can be mobilized before debranching is

required decreases as the average chain length increases

(recall that debranching is a slower process than phospho-

rolysis). However, molecules with longer chains have a

greater number of glucose residues that can be phospho-

rylized between branch points. Since the density of outer-

most branches is sterically limited, the maximum size of a

glycogen molecule decreases as the average number of

branches per tier increases. Mature glycogen particles from

a variety of animals have ⬃12 tiers of branches, with ⬃2

branches per tier and branch lengths averaging ⬃13

residues. Mathematical analysis suggests that these values

are close to optimal for mobilizing the greatest amount of

glucose in the shortest possible time.

3 CONTROL OF GLYCOGEN

METABOLISM

We have just seen that both glycogen synthesis and break-

down are exergonic under the same physiological condi-

tions. If both pathways operate simultaneously, however,

all that is achieved is wasteful hydrolysis of UTP. This

situation is similar to that of the phosphofructokinase–

fructose bisphosphatase substrate cycle (Section 17-4Ff).

Glycogen phosphorylase and glycogen synthase therefore

must be under stringent control such that glycogen is ei-

ther synthesized or utilized according to cellular needs.

The astonishing mechanism of this control is our next

topic of discussion. It involves not only allosteric control

and substrate cycles but enzyme-catalyzed covalent modi-

fication of both glycogen synthase and glycogen phospho-

rylase.The covalent modification reactions are themselves

ultimately under hormonal control through an enzymatic

cascade.

A. Direct Allosteric Control of Glycogen

Phosphorylase and Glycogen Synthase

As we saw in Section 17-4B, the net flux of reactants, J,

through a step in a metabolic pathway is the difference be-

tween the forward and reverse reaction velocities, v

and v

The fractional variation in the flux through any step in a path-

way with a change in substrate concentration approaches

infinity as that reaction step approaches equilibrium (v

⬇ v

;

Eq. [17.4]). The flux through a near-equilibrium reaction

Section 18-3. Control of Glycogen Metabolism 647

Figure 18-9 The branching of glycogen. Branches are formed by transferring a 7-residue

terminal segment from an ␣(1 S 4)-linked glucan chain to the C6 OH group of a glucose

residue on the same or another chain.

glycogen branching enzyme

α(1 4)-terminal

chains of glycogen

JWCL281_c18_638-670.qxd 6/3/10 1:48 PM Page 647

is therefore all but uncontrollable. As we have seen for the

case of PFK and FBPase, however, precise flux control of a

pathway is possible when an enzyme functioning far from

equilibrium is opposed by a separately controlled enzyme.

Then, v

and v

vary independently. In fact, under these cir-

cumstances, even the flux direction is controllable if v

can be

made larger than v

. Exactly this situation occurs in glyco-

gen metabolism through the opposition of the glycogen

phosphorylase and glycogen synthase reactions. The rates

of both of these reactions are under allosteric control by

effectors that include ATP, G6P, and AMP. In muscle, glyco-

gen phosphorylase is activated by AMP and inhibited by

ATP and G6P (Fig. 18-10, left). Glycogen synthase, on the

other hand, is activated by G6P.When there is high demand

for ATP (low [ATP], low [G6P], and high [AMP]), glycogen

phosphorylase is stimulated and glycogen synthase is inhib-

ited, so flux through this pathway favors glycogen break-

down. When [ATP] and [G6P] are high, the reverse is true

and glycogen synthesis is favored.

The structural differences between the active (R) and

inactive (T) conformations of glycogen phosphorylase

(Figs. 18-11a and 18-11b) can be understood in terms of

the symmetry model of allosterism (Section 10-4B). The

T-state enzyme has a buried active site and hence a low

affinity for its substrates, whereas the R-state enzyme has

an accessible catalytic site and a high-affinity phosphate

binding site.

AMP promotes phosphorylase’s T (inactive) S R (ac-

tive) conformational shift by binding to the R state of the

enzyme at its allosteric effector site (Fig. 18-10, left). In do-

ing so, AMP’s adenine, ribose, and phosphate groups bind

to separate segments of the polypeptide chain so as to link

the active site, the subunit interface, and the N-terminal re-

gion (Fig. 18-11c), the latter having undergone a large con-

formational shift (36 Å for Ser 14) from its position in the

T-state enzyme (Figs. 18-11a and 18-11b).AMP binding also

causes glycogen phosphorylase’s tower helices (Figs. 18-2

and 18-11) to tilt and pull apart so as to achieve a more fa-

vorable packing. These tertiary movements trigger a con-

certed T S R transition, which largely consists of an ⬃10°

relative rotation of the two subunits about an axis at the

subunit interface that is perpendicular to the dimer’s 2-fold

axis of symmetry. The enzyme’s 2-fold symmetry is thereby

preserved in accordance with the symmetry model of al-

losterism (Section 10-4B). The movement of the tower he-

lices also displaces and disorders a loop (the 280s loop,

residues 282–286), which covers the T-state active site so as

to prevent substrate access. It also causes the Arg 569 side

chain, which is located in the active site near the PLP phos-

phoryl group and the P

-binding site, to rotate in a way that

increases the enzyme’s binding affinity for its anionic P

substrate (Figs. 18-11a and 18-11b).

Curiously, ATP also binds to the allosteric effector site,

but in the T state, so that it inhibits rather than promotes

648 Chapter 18. Glycogen Metabolism

Figure 18-10 The control of glycogen phosphorylase activity.

The enzyme may assume the enzymatically inactive T

conformation (above) or the catalytically active R form (below).

The conformation of phosphorylase b is allosterically controlled

by effectors such as AMP,ATP, and G6P and is mostly in the T

state under physiological conditions. In contrast, the modified

form of the enzyme, phosphorylase a, is largely unresponsive to

2ADP2ATP

phosphorylase

kinase

T form

(inactive)

phosphoprotein

phosphatase

AMP

Glucose

ATP

and/or

G6P

Phosphorylase b Phosphorylase a

R form

(active)

these effectors and is mostly in the R state unless there is a high

level of glucose. Under usual physiological conditions, the

enzymatic activity of glycogen phosphorylase is essentially

determined by its rates of modification and demodification. Note

that only the T form enzyme is subject to phosphorylation and

dephosphorylation, so effector binding influences the rates of

these modification/demodification events.

JWCL281_c18_638-670.qxd 6/3/10 1:48 PM Page 648

the T S R conformational shift.This is because, as structural

analysis indicates, the  and  phosphate groups of ATP

bind to the enzyme such that its ribose and  phosphate

groups are displaced relative to those of AMP, thus desta-

bilizing the R state. The inhibitory action of ATP on phos-

phorylase is therefore simply understood: It competes with

Section 18-3. Control of Glycogen Metabolism 649

Figure 18-11 Conformational changes in glycogen

phosphorylase. Ribbon diagrams of one subunit of the dimeric

enzyme glycogen phosphorylase b (a) in the T state in the

absence of allosteric effectors and (b) in the R state with bound

AMP. The view is of the lower (orange) subunit in Fig. 18-2

rotated by

⬃45° about a horizontal axis.The tower helix is blue;

the N-terminal helix is cyan; the N-terminal residues that change

conformation on AMP binding are red; and residues 281 to 288,

the 280s loop, which in the R state are mostly disordered and

hence not seen, are green. Of the groups that are shown in

space-filling representation, the side chain of Ser 14, the

phosphorylation site, and AMP, which binds only to the R state

enzyme, have C green, N blue, O red, and P orange; the side

chain of Arg 569, which reorients in the T S R transition so as to

interact with the substrate phosphate, has C cyan and N blue; and

sulfate ions bound near Ser 14 and at the active site of the R

state enzyme, which mimic the sterically similar phosphate ions,

have O red and S orange. The PLP at the active site is drawn in

stick form with C magenta, N blue, O red, and P orange. [Based

on X-ray structures by Louise Johnson, Oxford University, U.K.

PDBids 8GPB and 7GPB.] (c) The portion of the glycogen

phosphorylase a dimer in the vicinity of the dimer interface

showing the position of the Ser 14 phosphate group, the AMP

bound in the allosteric effector site, and the active site–bound

glucose molecule. The view is along the molecular 2-fold axis and

hence is similar to that in Fig. 18-2. Residues 5 to 80 and 265 to

310 are, respectively, cyan and blue in one subunit and pink and

magenta in the other.The AMP and glucose are shown in

space-filling representation with C green, N dark blue, O red,

and P orange. The Ser 14 phosphate group is also shown in

space-filling representation with O dark red and P yellow. [X-ray

structure coordinates courtesy of Stephen Sprang, University of

Texas Southwest Medical Center.]

See Kinemage Exercises

14-2 and 14-3

(a) (b)

(c)

JWCL281_c18_638-670.qxd 2/26/10 2:24 PM Page 649

AMP for binding to phosphorylase and, in doing so, pre-

vents the relative motions of the three polypeptide seg-

ments required for phosphorylase activation.

The above allosteric interactions are superimposed on

an even more sophisticated control system involving cova-

lent modifications (phosphorylation/dephosphorylation) of

glycogen phosphorylase and glycogen synthase.These mod-

ifications alter the structures of the enzymes so as to change

their responses to allosteric regulators. We shall therefore

discuss the general concept of covalent modification and

how it increases the sensitivity of a metabolic system to ef-

fector concentration changes.We subsequently consider the

functions of such modifications in glycogen metabolism.

Only then will we be in a position to take up the detailed

consideration of allosteric control in glycogen metabolism.

B. Covalent Modification of Enzymes by Cyclic

Cascades: Effector “Signal” Amplification

Glycogen synthase and glycogen phosphorylase can each be

enzymatically interconverted between two forms with differ-

ent kinetic and allosteric properties through a complex

series of reactions known as a cyclic cascade. The intercon-

version of these different enzyme forms involves distinct,

enzyme-catalyzed covalent modification and demodifica-

tion reactions.

Compared with other regulatory enzymes, enzymati-

cally interconvertible enzyme systems:

1. Can respond to a greater number of allosteric stimuli.

2. Exhibit greater flexibility in their control patterns.

3. Possess enormous amplification potential in their re-

sponses to variations in effector concentrations.

This is because the enzymes that modify and demodify a tar-

get enzyme are themselves under allosteric control. It is

therefore possible for a small change in concentration of an

allosteric effector of a modifying enzyme to cause a large

change in the concentration of an active, modified target en-

zyme. Such a cyclic cascade is diagrammed in Fig. 18-12.

a. Description of a General Cyclic Cascade

Figure 18-12a shows a general scheme for a cyclic cascade

where, by convention, the more active target enzyme form

has the subscript a and the less active form has the subscript

b. Here, modification, in this case, phosphorylation, activates

the enzyme. Note that the modifying enzymes, F and R, are

active only when they have bound their respective allosteric

effectors e

and e

.The kinetic mechanisms for the intercon-

version of the unmodified and modified forms of the target

enzyme, E

and E

, are indicated in Fig. 18-12b.

In the steady state, the fraction of E in the active form,

]冫[E]

(where [E]

 [E

]  [E

] is the total enzyme con-

centration),determines the rate of the reaction catalyzed by

E. This fraction is a function of the total concentrations of

the modifying enzymes, [F]

and [R]

, the concentrations of

their allosteric effectors, e

and e

, the dissociation constants

of these effectors, K

and K

, and the substrate dissociation

constants, K

and K

, of the target enzymes, as well as the

rate constants, k

and k

,for the interconversions themselves

(Fig. 18-12). This relationship is obviously quite complex.

Nevertheless, it can be shown that, in a cyclic cascade, a rel-

atively small change in the concentration of e

, the allosteric

effector of the modifying enzyme F, can result in a much

larger change in [E

]冫[E]

, the fraction of E in the active

form. In other words, the cascade functions to amplify the

sensitivity of the system to an allosteric effector.

We have so far considered the covalent modification of

only one enzyme, a monocyclic cascade. Imagine a bicyclic

cascade involving the covalent modification of one of the

modifying enzymes (F), as well as the metabolic target en-

zyme (E) (Fig. 18-13). As you might expect, the amplifica-

tion potential of a “signal,” e

, as well as the control flexibil-

ity of such a system, is enormous.

650 Chapter 18. Glycogen Metabolism

Figure 18-12 A monocyclic enzyme cascade. (a) General

scheme, where F and R are, respectively, the modifying and

demodifying enzymes.These are allosterically converted from

their inactive to their active conformations on binding their

respective effectors, e

and e

.The target enzyme, E, is more

ATP  F • e



• e



• ATP • F • e



(b)

ADP



•







•





(a)

ADP

ATP



R • e

F • e

active

Less

active

Less

active

Less

active

active in the modified form (E

) and less active in the unmodified

form (E

). Dashed arrows symbolize catalysis of the indicated

reactions. (b) Chemical equations for the interconversion of the

target enzyme’s unmodified and modified forms E

and E

JWCL281_c18_638-670.qxd 2/26/10 2:24 PM Page 650