Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

formulate the following catalytic mechanism for E. coli

RNR (Fig. 28-13):

1. RNR’s free radical (X) abstracts an H atom from

C3¿ of the substrate in the reaction’s rate-determining step.

2 and 3. Acid-catalyzed cleavage of the C2¿¬OH bond

releases H

O to yield a radical cation intermediate. The

radical mediates the stabilization of the C2¿ cation by the

3¿¬OH group’s unshared electron pair, thereby account-

ing for the radical’s catalytic role.

4. The radical cation intermediate is reduced by the en-

zyme’s redox-active sulfhydryl pair to yield a 3¿-deoxynu-

cleotide radical and a protein disulfide group.

5. The 3¿ radical reabstracts an H atom from the protein

to yield the product deoxynucleoside diphosphate and re-

store the enzyme to its radical state.A small fraction of the

originally abstracted H atom exchanges with solvent be-

fore it can be replaced, thus accounting for the release of

H on reduction of [3¿-

H]UDP.

The Tyr 122 radical in R2 is too far away (10 Å) from the

enzyme’s catalytic site to abstract an electron directly from

the substrate. Evidently, the protein mediates electron

transfer from this tyrosyl radical to some other group (X in

Fig. 28-13) that is in close proximity to the substrate

C3¿¬H group. Site-directed mutagenesis studies suggest

that Cys 439 of R1, in its thiyl radical form (—S), is the

most plausible candidate for X (which makes RNR the

only enzyme in which a Cys residue is known to reduce a

carbohydrate substrate). Similar studies suggest that Cys

225 and Cys 462 of R1 form the redox-active sulfhydryl

pair that directly reduces substrate. Moreover,the resulting

disulfide bond is subsequently reduced to regenerate ac-

tive enzyme via disulfide interchange with Cys 754 and Cys

759 on R1, which are apparently positioned to accept elec-

trons from external reducing agents (see below).Thus, each

R1 subunit contains at least five Cys residues that chemi-

cally participate in nucleotide reduction.

These observations are confirmed by the X-ray struc-

ture of R1 in complex with R2’s 20-residue C-terminal

Section 28-3. Formation of Deoxyribonucleotides 1121

Figure 28-13 Enzymatic mechanism of ribonucleotide

reductase. The reaction occurs via a free radical–mediated

process in which reducing equivalents are supplied by the

OHOH

Base

POP OO

–

EX•

•

OHOH

Base

POP OO

–

•

OHOH

Base

POP OO

–

•

Base

POP OO

–

•

Base

POP OO

–

•

Base

POP OO

–

•

Base

POP OO

–

3 2

formation of an enzyme disulfide bond. [After Stubbe, J.A., Biol.

Chem. 265, 5330 (1990).]

JWCL281_c28_1107-1142.qxd 4/22/10 9:16 AM Page 1121

polypeptide (R1 does not crystallize satisfactorily in the

absence of this polypeptide), also determined by Eklund

(Fig. 28-12d). The central domain of the three-domain R1

monomer consists of a novel 10-stranded ␣/␤ barrel that is

formed by the antiparallel joining of two topologically sim-

ilar half-barrels, each comprising five parallel ␤ strands

connected by four ␣ helices. As with the similar 8-stranded

␣/␤ barrels that form the active sites of numerous enzymes

(Section 8-3Bh), R1’s active site Cys residues (439, 225,

462) are located in the mouth of the 10-stranded ␣/␤ barrel.

The two R1 Cys residues, 754 and 759, that are impli-

cated in the regeneration of the active enzyme are compo-

nents of R1’s C-terminal segment, which is not visible in

the X-ray structure of R1 and is presumably disordered.

This observation supports the hypothesis that this C-terminal

segment acts to flexibly shuttle reducing equivalents from

the enzyme surface to its active site.

b. Radical Generation in Class I RNR Requires

the Presence of O

One of the most remarkable aspects of Class I RNR is

its ability to stabilize its normally highly reactive TyrOⴢ

radical (its half-life is 4 days in the protein vs milliseconds

in solution). Yet quenching the radical, say, by hydrox-

yurea, inactivates the enzyme.How, then, is the radical gen-

erated in the first place? The radical may be restored in

vitro by simply treating the inactive enzyme with Fe(II)

and a reducing agent in the presence of O

This is a four-electron reduction of O

in which the reduc-

ing agent that supplies the electron represented by e

–

may

be ascorbate or even excess Fe

2⫹

c. The Inability of Oxidized RNR to Bind Substrate

Serves an Important Protective Function

Comparison of the X-ray structures of reduced R1 (in

which the redox-active Cys 225 and Cys 462 residues are in

their SH forms) with that of oxidized R1 (in which Cys 225

and Cys 462 are disulfide linked) reveals that Cys 462 in re-

duced R1 has rotated away from its position in oxidized R1

to become buried in a hydrophobic pocket, whereas Cys 225

moves into the region formerly occupied by Cys 462.The dis-

tance between the formerly disulfide-linked S atoms thereby

increases from 2.0 to 5.7 Å.These movements are accompa-

nied by small shifts of the surrounding polypeptide chain.

Oxidized RNR does not bind substrate because its R1 Cys

Tyr 122-R2

O•

++++2Fe

+ Fe

+ H

OFe

Tyr 122-R2

225 would prevent the binding of substrate through steric in-

terference of its S atom with the substrate NDP’s O2¿ atom.

The inability of oxidized RNR to bind substrate has

functional significance. In the absence of substrate, the en-

zyme’s free radical is stored in the interior of the R2 pro-

tein, close to its dinuclear iron center. When substrate is

bound, the radical is presumably transferred to it via a se-

ries of protein side chains in both R2 and R1. If the sub-

strate is unable to properly react after accepting this free

radical, as would be the case if RNR were in its oxidized

state, this could result in the destruction of the substrate

and/or the enzyme. Indeed, the mutation of the redox-

active Cys 225 to Ser results in an enzyme that permits the

formation of the substrate radical (Fig. 28-13); however,

since the mutant enzyme is incapable of reducing it,the sub-

strate radical instead decomposes followed by the release of

its base and phosphate moieties. More importantly, a tran-

sient peptide radical forms, which cleaves and inactivates

the R1 polypeptide chain while consuming the radical and

thereby inactivating R2. Thus, an important role of the en-

zyme is to control the release of the radical’s powerful oxi-

dizing capability. It does so in part by preventing the bind-

ing of substrate while the enzyme is in its oxidized form.

d. Ribonucleotide Reductase Is Regulated by

Effector-Induced Oligomerization

The synthesis of the four dNTPs in the amounts re-

quired for DNA synthesis is accomplished through feed-

back control. The maintenance of the proper intracellular

ratios of dNTPs is essential for normal growth. Indeed, a

deficiency of any dNTP is lethal, whereas an excess is muta-

genic because the probability that a given dNTP will be

erroneously incorporated into a growing DNA strand

increases with its concentration relative to those of the other

dNTPs.

The activities of both E. coli and mammalian Class I

RNRs are allosterically responsive to the concentrations of

various (d)NTPs. Thus, as Reichard has shown, ATP in-

duces the reduction of CDP and UDP; dTTP induces the

reduction of GDP and inhibits the reduction of CDP and

UDP; dGTP induces the reduction of ADP and, in mam-

mals but not E. coli, inhibits the reduction of CDP and

UDP; and dATP inhibits the reduction of all NDPs.

Barry Cooperman has shown that the catalytic activity

of mouse RNR varies with its state of oligomerization (that

is, RNR is a morpheein; Section 26-4Ac), which in turn is

governed by the binding of nucleotide effectors to three in-

dependent allosteric sites on R1: (1) the specificity site,

which binds ATP, dATP, dGTP, and dTTP; (2) the activity

site, which binds ATP and dATP; and (3) the hexameriza-

tion site, which binds only ATP. On the basis of molecular

mass, ligand binding, and activity studies on mouse RNR,

Cooperman formulated a model that quantitatively ac-

counts for the allosteric regulation of Class I RNR. It has

the following features (Fig. 28-14a):

1. The binding of ATP, dATP, dGTP, or dTTP to the

specificity site induces the catalytically inactive R1

monomers to form a catalytically active dimer, R1

1122 Chapter 28. Nucleotide Metabolism

JWCL281_c28_1107-1142.qxd 8/9/10 9:46 AM Page 1122

2. The binding of dATP or ATP to the activity site

causes the dimers to form catalytically active tetramers,

, that slowly but reversibly change conformation to a

catalytically inactive state, R1

3. The binding of ATP to the hexamerization site in-

duces the tetramers to further aggregate to form catalyti-

cally active hexamers, R1

, RNR’s major active form.

The concentration of ATP in a cell is such that, in vivo,

R1 is almost entirely in its tetrameric or hexameric forms.

As a consequence, ATP couples the overall rate of DNA

synthesis to the cell’s energy state.

The specificity and activity sites have been located in

X-ray structures of E. coli R1 (Fig. 28-12d);the hexameriza-

tion site has not yet been identified.The R1 hexamer had, in

fact, been previously observed in the X-ray structures of R1

(Fig. 28-14b,c), but the interactions between its contacting

dimers are so tenuous that it was assumed that they are

merely artifacts of crystallization with no physiological sig-

nificance.Yet, since the activity site is located at this contact

Section 28-3. Formation of Deoxyribonucleotides 1123

Figure 28-14 Ribonucleotide reductase regulation. (a) A

model for the allosteric regulation of Class I RNR via its

oligomerization. States shown in green have high activity and

those shown in red have little or no activity. R2 has been omitted

for simplicity. [After Kashlan, O.B., Scott, C.P., Lear, J.D., and

Cooperman, B.S., Biochemistry 41, 461 (2002).] (b) The X-ray

structure of the R1 hexamer, which has D

symmetry, in complex

with AMPPNP as viewed along its 3-fold axis. Each of its three

dimers are differently colored (the X-ray structure of a dimer is

ATP

dATP

dGTP

dTTP

(d)NTP

dATP

ATP

R1 R1

d(NTP)

(d)NTP

(d)ATP

Specificity site

Activity site

Hexamerization site

(slow)

shown in Fig. 28-12d). The AMPPNP, which binds to the

enzyme’s activity sites, is drawn in space-filling form with

C green, N blue, O red, and P gold.The black arrows point along

the R1 dimers’ 2-fold axes and indicate the probable docking

sites for the binding of R2 dimers. (c) The R1  AMPPNP

hexamer as viewed along the vertical 2-fold axis in Part b. [Parts

b and c based on an X-ray structure by Hans Eklund, Swedish

University of Agricultural Sciences, Uppsala, Sweden. PDBid

3R1R.]

(a)

(b)

(c)

JWCL281_c28_1107-1142.qxd 4/22/10 9:16 AM Page 1123

site, it now seems likely that its binding of (d)ATP induces

R1 oligomerization through local conformational changes.

The foregoing model has, for simplicity, neglected the

presence of R2 subunits although, of course, R1 and R2

must be present in equimolar amounts in the active en-

zyme. Presumably, the R1 and R2 dimers bind to one an-

other such that their 2-fold axes coincide.The lack of space

on the inside of the R1 hexamer dictates that the R2 dimers

must contact the R1 dimers from outside the hexamer (Fig.

28-14b).

dCTP is not an effector of RNR. This is presumably be-

cause the intracellular balance between dCTP and dTTP is

not controlled by RNR but, rather,is maintained by deoxy-

cytidine deaminase, which converts dCTP to dUMP, the

precursor of dTTP. This enzyme is activated by dCTP and

inhibited by dTTP.

e. Thioredoxin and Glutaredoxin Are Class I

Ribonucleotide Reductase’s Physiological

Reducing Agents

The final step in the RNR catalytic cycle is the reduction of

the enzyme’s newly formed disulfide bond to reform its re-

dox-active sulfhydryl pair. Dithiols such as that of

2-mercaptoethanol (Section 7-1B) can serve as the reducing

agent for this process in vitro through a disulfide interchange

reaction. One of the enzyme’s physiological reducing agents,

however, is thioredoxin (Trx), a ubiquitous monomeric 105-

residue protein that has a pair of closely proximal redox-

active Cys residues, Cys 32 and Cys 35 (we have previously

encountered thioredoxin in our study of the light-induced

activation of the Calvin cycle; Section 24-3B). Thioredoxin

reduces oxidized RNR via disulfide interchange.

The X-ray structure of reduced E. coli Trx (Fig. 28-15)

reveals that the side chain of the redox-active Cys 32 is

exposed on the protein’s surface, where it is available for

oxidation. Oxidized thioredoxin is, in turn, reduced by

Thioredoxin

(reduced)

Ribonucleotide

reductase

(oxidized)

Ribonucleotide

reductase

(reduced)

Thioredoxin

(oxidized)

NADPH in a reaction mediated by the flavoprotein thiore-

doxin reductase. NADPH therefore serves as the terminal

reducing agent in the RNR-mediated reduction of NDPs

to dNDPs (Fig. 28-16).

The existence of a viable E. coli mutant devoid of thiore-

doxin indicates that this protein is not the only substance

capable of reducing oxidized RNR in vivo. This observation

led to the discovery of glutaredoxin, a disulfide-containing,

monomeric, 85-residue protein that can also reduce RNR

(mutants devoid of both thioredoxin and glutaredoxin are

nonviable). Oxidized glutaredoxin is reduced, via disulfide

interchange, by the Cys-containing tripeptide glutathione

which, in turn, is reduced by NADPH as catalyzed by

glutathione reductase (GR; Section 21-2Ba). The relative

1124 Chapter 28. Nucleotide Metabolism

Figure 28-15 X-ray structure of human thioredoxin in its

reduced (sulfhydryl) state. The 105-residue polypeptide chain is

drawn in ribbon form colored in rainbow order from its N-terminus

(blue) to its C-terminus (red).The side chains of the redox-active

residues, Cys 32 and Cys 35, are shown in space-filling form with

C green and S yellow. This structure closely resembles those of

the homologous a and a¿ domains of protein disulfide isomerase

(PDI, Fig. 9-16). [Based on an X-ray structure by William

Montfort, University of Arizona. PDBid 1ERT.]

Figure 28-16 Electron-transfer pathway for nucleoside

diphosphate (NDP) reduction. NADPH provides the reducing

NADPH

NADP

FAD

FADH

SH SH

NDP

dNDP

SH SH

Thioredoxin reductase Thioredoxin Ribonucleotide

reductase

equivalents for this process through the intermediacy of

thioredoxin reductase, thioredoxin, and ribonucleotide reductase.

JWCL281_c28_1107-1142.qxd 4/22/10 9:16 AM Page 1124

importance of thioredoxin and glutaredoxin in the reduc-

tion of RNRs remains to be established.

f. Thioredoxin Reductase Alternates Its

Conformation with Its Redox State

Thioredoxin reductase (TrxR), a homodimer of 316-

residue subunits, is a homolog of GR that catalyzes a simi-

lar reaction: the reduction of a substrate disulfide bond by

NADPH as mediated by an FAD prosthetic group and a

redox active sulfhydryl pair (Cys 135 and Cys 138). How-

ever, the X-ray structure of the C138S mutant of E. coli

TrxR in complex with NADP



(Fig. 28-17a), determined by

Charles Williams and John Kuriyan, reveals that TrxR and

GR differ in their active site arrangements such that their

redox-active sulfhydryl pairs are on opposite sides of the

flavin rings in the two enzymes. Nevertheless,TrxR’s redox-

active sulfhydryl pair appears properly positioned to re-

duce the flavin ring. However, the NADP



’s nicotinamide

ring is 17 Å from the flavin ring and the redox-active

sulfhydryl pair is buried such that it could not react with

the enzyme’s Trx substrate. How then does TrxR manage to

transfer an electron pair from its bound NADPH via its

flavin ring and redox-active sulfhydryl pair to Trx?

This question was answered by Williams and Martha

Ludwig through their X-ray structure determination of the

C135S mutant of TrxR, whose Cys 138 is disulfide-linked to

Cys 32 of the C35S mutant of E. coli Trx (probably the

physiologically relevant disulfide bond) and which is in

complex with the NADP



analog 3-aminopyridine adenine

dinucleotide phosphate (AADP



). In this complex (Fig.

28-17b),TrxR’s NADP



-binding domain has rotated by 67°

Section 28-3. Formation of Deoxyribonucleotides 1125

Figure 28-17 X-ray structures of E. coli

thioredoxin reductase (TrxR). (a) The C138S

mutant TrxR in complex with NADP



.The

protein is shown in ribbon form colored

according to its secondary structure. The

NADP



, the FAD, and the side chains of

Cys 135 and Ser 138 are drawn in ball-and-

stick form with C yellow, N blue, O red, S

green, and P magenta. (b) The C135S

mutant TrxR in complex with AADP



and

covalently linked to the C35S mutant of Trx

via a disulfide bond between TrxR Cys 138

and Trx Cys 32. The TrxR is represented as in

Part a, the Trx ribbon is blue-gray, and its

Cys 32 and Ser 35 side chains are drawn in

ball-and-stick form. Comparison of these

two structures reveals that TrxR’s

NADP



-binding domain (residues 120–243)

undergoes a 67° rotation about the axis

drawn in blue relative to the rest of the

protein, which is shown in the same

orientation in both structures. [Courtesy of

Martha Ludwig, University of Michigan.

PDBids (a) 1TDF and (b) 1F6M.]

(a)

(b)

relative to the rest of the protein compared to its position

in TrxR alone (Fig. 28-17a). This positions the AADP



’s

pyridine ring to react with the flavin ring and positions

TrxR’s redox-active sulfhydryl pair to undergo a disulfide

interchange reaction with that of Trx. Moreover, in this

latter conformation, the NADP



-binding domain appears

to provide the recognition site for the substrate Trx.

Evidently, TrxR alternates its conformation with each suc-

cessive step in the process of transferring an electron pair

from NADPH to the flavin to its redox-active sulfhydryl

pair to its bound Trx substrate. This added mechanistic

complication relative to that of GR, which does not un-

dergo a significant conformational change in reducing

glutathione disulfide (Section 21-2Ba), has apparently

evolved to permit TrxR to reduce its protein substrate: Trx

would be too large for its redox-active sulfhydryl pair to

properly approach the active site sulfhydryl pair in a GR-

like enzyme.

g. The Three Classes of Ribonucleotide

Reductases Are Evolutionarily Related

We have seen that the active forms of Class I RNRs are

,R1

, and R1

oligomers that have mecha-

nistically essential tyrosyl radicals that are stabilized by

oxo-bridged binuclear Fe(III) complexes, have NDPs for

substrates, and obtain their reducing equivalents from

thioredoxin and glutaredoxin. In contrast, Class II RNRs,

which are  monomers or 

dimers, utilize a 5¿-deoxyadeno-

sylcobalamin cofactor (coenzyme B

; Section 25-2Eb) for

radical generation,have NDPs for substrates, and are reduced

by thioredoxin and glutaredoxin; whereas Class III RNRs,

JWCL281_c28_1107-1142.qxd 4/22/10 9:16 AM Page 1125

which are 

dimers that interact with a radical-generating

protein 

that contains a [4Fe–4S] cluster and requires

S-adenosylmethionine (SAM; Section 26-3Ea) and NADPH

for activity, have NTPs for substrates, and their reducing

equivalents are provided by the oxidation of formate to CO

Since all known cellular life synthesizes its deoxy-

ribonucleotides from ribonucleotides, the rise of an RNR

must have preceded the evolutionary transition from the

RNA world (Section 1-5Ca) to DNA-based life-forms. Did

the three classes of RNRs arise independently or are they

evolutionarily related? Despite the seemingly large differ-

ences between these different classes of RNRs, the reac-

tions they catalyze are surprisingly similar. All replace the

2¿ OH group of ribose with H via a free radical mechanism

involving a thiyl radical with the reducing equivalents pro-

vided by a Cys sulfhydryl group (Fig. 28-13; the second Cys

residue of the redox-active sulfhydryl pair in Class I and II

RNRs is replaced by formate in Class III RNRs).They dif-

fer mainly in the way they generate the free radical. [In

Class II RNRs, the radical is generated by the homolytic

cleavage of its 5¿-deoxyadenosylcobalamin cofactor’s

C¬Co(III) bond (Section 25-2Ec). In Class III RNRs, it is

generated by the NADPH-supplied and [4Fe–4S] cluster–

mediated one-electron reductive cleavage of SAM by the



protein to yield methionine and the 5¿-deoxyadenosyl

radical (the same radical generated by the homolytic cleav-

age of 5¿-deoxyadenosylcobalamin), which then abstracts

the H atom from a C



¬H group of a specific Gly on the 

subunit to yield 5¿-deoxyadenosine and a stable but O

sensitive glycyl radical.] Moreover, the X-ray structures of

both Class II and Class III RNRs reveal that their active

sites are formed by 10-stranded / barrels that have the

same connectivity as and are closely superimposable on

that of Class I RNRs. It therefore appears that all three

classes of RNRs are evolutionarily related. Reichard has

proposed that, since life arose under anaerobic conditions

and that formate, one of simplest organic reductants, was

probably widely available on primitive Earth (Section

1-5B), the primordial RNR was a Class III–like enzyme.

The rise of photosynthetic organisms that generated O

then promoted the evolution of Class II RNRs, which can

function under both anaerobic and aerobic conditions.

Class I RNRs, which require the presence of O

for activa-

tion, evolved last, presumably from a Class II RNR.

h. dNTPs Are Produced by Phosphorylation

of dNDPs

In pathways involving Class I and Class II RNRs, the

final step in the production of dNTPs is the phosphorylation

of the corresponding dNDPs:

This reaction is catalyzed by nucleoside diphosphate

kinase, the same enzyme that phosphorylates NDPs (Sec-

tion 28-1Ba).As before, the reaction is written with ATP as

the phosphoryl donor, although any NTP or dNTP can

function in this capacity. In pathways involving Class III

RNRs, the production of NTPs from NDPs precedes the

reduction of NTPs to dNTPs.

dNDP  ATP Δ dNTP  ADP

B. Origin of Thymine

a. dUTP Diphosphohydrolase

The dTMP component of DNA is synthesized,as we dis-

cuss below, by methylation of dUMP. The dUMP is gener-

ated through the hydrolysis of dUTP by dUTP diphospho-

hydrolase (dUTPase; also called dUTP pyrophosphatase):

The reason for this apparently energetically wasteful

process (dTMP, once formed, is rephosphorylated to dTTP)

is that cells must minimize their concentration of dUTP in

order to prevent incorporation of uracil into their DNA.

This is because, as we discuss in Section 30-5Bd, DNA poly-

merase does not discriminate between dUTP and dTTP.

The X-ray structure of human dUTPase, determined by

John Tainer, reveals the basis for this enzyme’s exquisite

specificity for dUTP. This homotrimer of 141-residue subunits

binds dUTP in a snug-fitting cavity that sterically excludes

thymine’s C5 methyl group via the side chains of conserved

residues (Fig. 28-18a). It differentiates uracil from cytosine via

a set of hydrogen bonds from the protein backbone that

mimic adenine’s base-pairing interaction (Fig. 28-18b), and it

differentiates dUTP from UTP by the steric exclusion of

ribose’s 2¿ OH group by the side chain of a conserved Tyr.

b. Thymidylate Synthase

dTMP is synthesized from dUMP by thymidylate

synthase (TS) with N

-methylenetetrahydrofolate

-methylene-THF) as the methyl donor:

(THF cofactors are discussed in Section 26-4D). Note that

the transferred methylene group (in which the carbon has

the oxidation state of formaldehyde) is reduced to a methyl

group (which has the oxidation state of methanol) at the

R =

COO

;

dTMP

dUMP N

-Methylenetetrahydrofolate

Dihydrofolate

()

n =

16

dRib P

dUTP  H

O Δ dUMP  PP

1126 Chapter 28. Nucleotide Metabolism

JWCL281_c28_1107-1142.qxd 7/21/10 7:09 PM Page 1126

expense of the oxidation of the THF cofactor to dihydro-

folate (DHF).

The catalytic mechanism of TS, a highly conserved ho-

modimeric protein (with 264-residue subunits in E. coli),

has been extensively investigated. On incubation of the en-

zyme with N

-methylene-[6-

H]THF and dUMP, the

H is quantitatively transferred to the methyl group of

the product dTMP. When [5-

H]dUMP is the substrate,

however, the

H is released into the aqueous solvent. Such

information, together with the knowledge that uracil C6,

which occupies the ␤ position of an ␣,␤-unsaturated ke-

tone, is susceptible to nucleophilic attack, led Daniel Santi

to propose the following mechanistic scheme for the TS

reaction (Fig. 28-19):

1. An enzyme nucleophile, identified as the thiolate

group of Cys 146, attacks C6 of dUMP to form a covalent

adduct.

2. C5 of the resulting enolate ion attacks the CH

group

of the iminium cation in equilibrium with N

-methylene-

THF to form an enzyme–dUMP–THF ternary covalent

complex.

3. An enzyme base abstracts the acidic proton at the C5

position of the enzyme-bound dUMP, forming an exocyclic

methylene group and eliminating the THF cofactor. The

abstracted proton subsequently exchanges with solvent.

4. The redox change occurs via the migration of the

N6¬H atom of THF as a hydride ion to the exocyclic

methylene group, converting it to a methyl group (thus

accounting for the above described transfer of

H) and

yielding DHF. This reduction promotes displacement of

the Cys thiolate group from the intermediate so as to

release product, dTMP, and reform active enzyme.

c. 5-Fluorodeoxyuridylate Is a Potent

Antitumor Agent

The above mechanism is supported by the observation

that 5-fluorodeoxyuridylate (FdUMP)

is an irreversible inhibitor of TS.This substance, like dUMP,

binds to the enzyme (an F atom has approximately the

same radius as an H atom) and undergoes the first two steps

of the normal enzymatic reaction. In Step 3, however, the

enzyme cannot abstract the F atom as F

⫹

(recall that F is the

most electronegative element), so that the enzyme is all but

permanently immobilized as the enzyme–FdUMP–THF

5-Fluorodeoxyuridylate (FdUMP)

HOH

–

Section 28-3. Formation of Deoxyribonucleotides 1127

Figure 28-18 X-ray structure of human dUTPase. (a) The

molecular surface at the substrate binding site showing how the

enzyme differentiates uracil from thymine. Bound dUTP is

drawn in stick form with its N, O, and P atoms represented by

blue, red, and yellow spheres. Mg

2⫹

ions that were modeled into

the structure are represented by green spheres.The protein’s

molecular surface is colored according to its electrostatic

potential with positive, negative, and near neutral regions blue,

red, and white, respectively. Note how the snug fit of the uracil

ring into its binding site would sterically exclude thymine’s C5

methyl group. (b) The substrate binding site indicating how the

enzyme differentiates uracil from cytosine and 2¿-deoxyribose

from ribose. dUMP bound at the active site is drawn as in Part a.

The protein, mainly the backbone of a ␤ hairpin motif, is

similarly drawn but with thinner gray bonds. Hydrogen bonds are

shown as dotted white lines, and a tightly bound, conserved water

molecule is represented by a pink sphere. The pattern of

hydrogen bonding donors and acceptors on the protein would

prevent cytosine from binding in the active site pocket.The

conserved Tyr side chain sterically excludes ribose’s 2’ OH group.

[Courtesy of John Tainer,The Scripps Research Institute, La

Jolla, California.]

(a)

(b)

JWCL281_c28_1107-1142.qxd 8/9/10 9:46 AM Page 1127

ternary covalent complex analogous to that after Step 2 in

Fig. 28-19. Indeed, X-ray structural analysis by William

Montfort revealed that crystals of E. coli TS that had been

soaked in a solution containing FdUMP and N

methylene-THF contain precisely this complex (Fig. 28-20).

Enzymatic inhibitors such as FdUMP, which inactivate an

enzyme only after undergoing part or all of its normal

catalytic reaction, are called mechanism-based inhibitors

(alternatively, suicide substrates because they cause the en-

zyme to “commit suicide”). Mechanism-based inhibitors, be-

ing targeted for particular enzymes, are among the most pow-

erful, specific, and therefore useful enzyme inactivators.

The strategic position of thymidylate synthase in DNA

biosynthesis has led to the clinical use of FdUMP as an an-

titumor agent. Rapidly proliferating cells, such as cancer

cells, require a steady supply of dTMP in order to survive

and are therefore killed by treatment with FdUMP. In con-

trast, most normal mammalian cells, which grow slowly if at

all, have a lesser requirement for dTMP, so that they are

1128 Chapter 28. Nucleotide Metabolism

Figure 28-20 The X-ray structure of the E. coli thymidylate

synthase in covalent complex with FdUMP and THF. The active

site region of one subunit of this dimeric enzyme is shown in

ribbon form with its polypeptide chain colored in rainbow order

from its N-terminus (blue) to its C-terminus (red).The FdUMP

and N

-methylene-THF are drawn in stick form with FdUMP

C cyan, N

-methylene-THF C green, N blue, O red,

F purple, and P orange. The C5 and C6 atoms of FdUMP form

covalent bonds with the CH

group substituent to N5 of THF

and the S atom of Cys 146, whose side chain is drawn in stick

form with C green and S yellow. [Based on an X-ray structure by

William Montfort, University of Arizona, PDBid 1TSN.]

–

dRib P

, N

-Methylene-THF

dRib P

dTMP

dRib P

dUMP

–

dRib P

–

+ H

DHF

Figure 28-19 Catalytic mechanism of thymidylate synthase. The methyl group is supplied by

-methylene-THF, which is concomitantly oxidized to dihydrofolate.

JWCL281_c28_1107-1142.qxd 6/8/10 10:39 AM Page 1128

relatively insensitive to FdUMP (some exceptions are the

bone marrow cells that comprise the blood-forming tissues

and much of the immune system, the intestinal mucosa, and

hair follicles). 5-Fluorouracil and 5-fluorodeoxyuridine are

also effective antitumor agents since they are converted to

FdUMP through salvage reactions.

d. N

-Methylene-THF Is Regenerated

in Two Reactions

The thymidylate synthase reaction is biochemically unique

in that it oxidizes THF to DHF; no other enzymatic reaction

employing a THF cofactor alters this coenzyme’s net oxida-

tion state. The DHF product of the thymidylate synthase re-

action is recycled to the enzyme’s N

-methylene-THF

cofactor through two sequential reactions (Fig. 28-21):

1. DHF is reduced to THF by NADPH as catalyzed by

dihydrofolate reductase (DHFR; Section 26-4D).Although,

in most organisms, DHFR is a monomeric monofunctional

enzyme, in protozoa and at least some plants, DHFR and TS

occur on the same polypeptide chain to form a bifunctional

enzyme that has been shown to channel DHF from its TS to

its DHFR active sites.

2. Serine hydroxymethyltransferase (Section 26-3Bb)

transfers the hydroxymethyl group of serine to THF, yield-

ing N

-methylene-THF and glycine.

e. Antifolates Are Anticancer Agents

Inhibition of DHFR quickly results in all of a cell’s lim-

ited supply of THF being converted to DHF by the thymidy-

late synthase reaction. Inhibition of DHFR therefore not

only prevents dTMP synthesis (Fig. 28-21), but also blocks

all other THF-dependent biological reactions such as the

synthesis of purines (Section 28-1A),methionine (Section 26-

5Ba), and, indirectly, histidine (Section 26-5Be). DHFR (Fig.

28-22) therefore offers an attractive target for chemotherapy.

Methotrexate (amethopterin), aminopterin, and tri-

methoprim

are DHF analogs that competitively although all but irre-

versibly bind to DHFR with an ⬃1000-fold greater affinity

than does DHF. These antifolates (substances that interfere

with the action of folate cofactors) are effective anticancer

agents, particularly against childhood leukemias. In fact, a

successful chemotherapeutic strategy is to treat a cancer

OCH

COOCH

–

Trimethoprim

COO

–

NCH

R = H

R = CH

Aminopterin

Methotrexate (amethopterin)

OCH

Section 28-3. Formation of Deoxyribonucleotides 1129

Figure 28-21 Regeneration of N

-methylenetetrahydro-

folate. The DHF product of the thymidylate synthase reaction is

converted back to N

-methylene-THF by the sequential

actions of (1) dihydrofolate reductase and (2) serine

hydroxymethyltransferase. Thymidylate synthase is inhibited

by FdUMP, whereas dihydrofolate reductase is inhibited by the

antifolates methotrexate, aminopterin, and trimethoprim.

Figure 28-22 X-ray structure of human dihydrofolate reduc-

tase in complex with folic acid. The polypeptide is colored in

rainbow order from its N-terminus (blue) to its C-terminus (red).

The folic acid is drawn in space-filling form with C green, N blue,

and O red. [Based on an X-ray structure by Joseph Kraut,

University of California at San Diego. PDBid 1DHF.]

See Interactive Exercise 29.

-Methylene-THF

NADPH + H

NADP

Serine

dTMP

serine

hydroxymethyl

transferase

dihydrofolate

reductase

FdUMP

Methotrexate

Aminopterin

Trimethoprim

thymidylate

synthase

DHF

THF

dUMP

CH COO

–

Glycine

COO

–

JWCL281_c28_1107-1142.qxd 10/19/10 9:58 AM Page 1129

Rib

AMP

Adenosine

nucleotidase

IMP

Inosine

nucleotidase

XMP

Uric acid

Xanthosine

nucleotidase

GMP

Guanosine

Hypoxanthine Xanthine Guanine

nucleotidase

O NH

AMP

deaminase

Ribose-1-P

purine

nucleoside

phosphorylase

(PNP)

Ribose-1-P

purine

nucleoside

phosphorylase

(PNP)

Ribose-1-P

purine

nucleoside

phosphorylase

(PNP)

+ H

xanthine

oxidase

adenosine

deaminase

xanthine

oxidase

guanine

deaminase

victim with a lethal dose of methotrexate and some hours

later “rescue” the patient (but hopefully not the cancer) by

administering massive doses of 5-formyl-THF and/or thymi-

dine.A low dose of methotrexate is also effective in the treat-

ment of rheumatoid arthritis, inhibiting immune system activ-

ity and thus decreasing inflammation. Trimethoprim, which

was discovered by George Hitchings and Gertrude Elion,

binds much more tightly to bacterial DHFRs than to those of

mammals and is therefore a clinically useful antibiotic.

4 NUCLEOTIDE DEGRADATION

Most foodstuffs, being of cellular origin, contain nucleic

acids. Dietary nucleic acids survive the acid medium of the

stomach; they are degraded to their component nu-

cleotides, mainly in the duodenum, by pancreatic nucleases

and intestinal phosphodiesterases. These ionic compounds,

which cannot pass through cell membranes, are then hy-

drolyzed to nucleosides by a variety of group-specific nu-

cleotidases and nonspecific phosphatases. Nucleosides may

be directly absorbed by the intestinal mucosa or first

undergo further degradation to free bases and ribose or

ribose-1-phosphate through the action of nucleosidases

and nucleoside phosphorylases:

Radioactive labeling experiments have demonstrated

that only a small fraction of the bases of ingested nucleic

acids are incorporated into tissue nucleic acids. Evidently,

the de novo pathways of nucleotide biosynthesis largely sat-

isfy an organism’s need for nucleotides. Consequently, in-

gested bases, for the most part, are degraded and excreted.

Cellular nucleic acids are also subject to degradation as part

of the continual turnover of nearly all cellular components.

In this section we outline these catabolic pathways and dis-

cuss the consequences of several of their inherited defects.

A. Catabolism of Purines

The major pathways of purine nucleotide and deoxynu-

cleotide catabolism in animals are diagrammed in

Fig. 28-23. Other organisms may have somewhat different

Nucleoside  P

¬¬¬¬

nucleoside

phosphorylase

base  ribose-1-P

Nucleoside  H

¬¬¬¬

nucleosidase

base  ribose

1130 Chapter 28. Nucleotide Metabolism

Figure 28-23 Major pathways of purine catabolism in animals.

The various purine nucleotides and deoxynucleotides are all

degraded to uric acid.

JWCL281_c28_1107-1142.qxd 4/22/10 9:16 AM Page 1130