Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

RNA is eventually replaced by DNA under conditions that

permit accurate base pairing to be achieved.

One might wonder why cells have evolved the complex

system of discontinuous lagging strand synthesis rather

than a DNA polymerase that could simply extend DNA

chains in their 3¿S5¿ direction. Consideration of the chem-

istry of DNA chain extension also leads to the conclusion

that this system promotes high-fidelity replication.The link-

ing of 5¿-deoxynucleotide triphosphates in the 3¿S5¿ direc-

tion would require the retention of the growing chain’s 5¿-

terminal triphosphate group to drive the next coupling step

(Fig. 30-39a). On editing a mispaired 5¿-terminal nucleotide

(Fig. 30-39b), this putative polymerase would—in analogy

with Pol I, for example—excise the offending nucleotide,

leaving either a 5¿-OH or a 5¿-phosphate group. Neither of

these terminal groups is capable of energizing further chain

extension. A proofreading 3¿S5¿ DNA polymerase would

therefore have to be capable of reactivating its edited prod-

uct. The inherent complexity of such a system has presum-

ably selected against its evolution.

4 EUKARYOTIC REPLICATION

There is a remarkable degree of similarity between eukary-

otic and prokaryotic DNA replication mechanisms. Never-

theless, there are important differences between these two

replication systems as a consequence of the vastly greater

complexity of eukaryotes in comparison to prokaryotes.

For example, eukaryotic chromosomes are structurally

complicated and dynamic complexes of DNA and protein

(Section 34-1) with which the replication machinery must

interact in carrying out its function. Consequently, as is true

of most aspects of biochemistry, our knowledge of how

DNA is replicated in eukaryotes has lagged well behind

that for prokaryotes, although in recent years there has

Section 30-4. Eukaryotic Replication 1201

at aberrantly high levels is more likely to be misincorpo-

rated and, conversely, one present at low levels is more

likely to be replaced by the dNTPs present at higher levels.

2. The polymerase reaction itself has extraordinary fi-

delity. This is because, as we have seen (Section 30-2Ae),

the polymerase reaction occurs in two stages: (1) a binding

step in which the incoming dNTP base-pairs with the tem-

plate while the enzyme is in an open conformation that

cannot catalyze the polymerase reaction; and (2) a catalysis

step in which the polymerase forms a closed conformation

about the newly formed base pair,which properly positions

its catalytic residues (induced fit). Since the formation of

the closed conformation requires that the incoming dNTP

form a Watson–Crick-shaped base pair with the template,

the conformation change constitutes a double check for

correct base pairing.

3. The 3¿S5¿ exonuclease functions of Pol I and Pol III

detect and eliminate the occasional errors made by their

polymerase functions. In fact, mutations that increase a

DNA polymerase’s proofreading exonuclease activity de-

crease the rates of mutation of other genes.

4. A remarkable battery of enzyme systems, contained

in all cells, function to repair residual errors in the newly

synthesized DNA as well as any damage that it may incur

after its synthesis through chemical and/or physical insults.

We discuss these DNA repair systems in Section 30-5.

In addition, the inability of a DNA polymerase to initiate

chain elongation without a primer is a feature that increases

DNA replication fidelity. The first few nucleotides of a

chain to be coupled together are those most likely to be

mispaired because of the cooperative nature of base pairing

interactions (Section 29-2). The editing of a short duplex

oligonucleotide is similarly an error-prone process.The use

of RNA primers eliminates this source of error since the

ppp ppp

(a)

…

ppp

…

pp p

ppp

(b)

…

pp pppp

…

pp p

3⬘5⬘

Figure 30-39 Chemical consequences if a DNA polymerase

could synthesize DNA in its 3ⴕ S 5ⴕ direction. (a) The coupling

of each nucleoside triphosphate to the growing chain would be

driven by the hydrolysis of the previously appended nucleoside

triphosphate. (b) The editorial removal of an incorrect 5¿-

terminal nucleoside triphosphate would render the DNA chain

incapable of further extension.

JWCL281_c30_1173-1259.qxd 8/10/10 9:11 PM Page 1201

1202 Chapter 30. DNA Replication, Repair, and Recombination

cer; Sections 19-3B and 34-4C); by the surgical removal of

a tissue, which results in its rapid regeneration; or by pro-

teins known as mitogens, which bind to cell-surface recep-

tors and induce cell division (Section 34-4D).

a. The Cell Cycle Is Controlled by Cyclins and

Cyclin-Dependent Protein Kinases

The progression of a cell through the cell cycle is regu-

lated by proteins known as cyclins and cyclin-dependent

protein kinases (Cdks). Cyclins are so named because they

are synthesized during one phase of the cell cycle and com-

pletely degraded during a succeeding phase (protein degra-

dation is discussed in Section 32-6). A particular cyclin

specifically binds to and thereby activates its corresponding

Cdk(s) to phosphorylate its target proteins, thus activating

these proteins to carry out the processes comprising that

phase of the cell cycle. In order to enter a new phase in the

cell cycle, a cell must satisfy a corresponding checkpoint,

which monitors whether the cell has satisfactorily completed

the preceding phase [e.g., the attachment of all chromo-

somes to the mitotic spindle must precede mitosis (Section

1-4Aa); if this were not the case for even one chromosome,

one daughter cell would lack this chromosome and the other

would have two, both deleterious if not lethal conditions]. If

the cell has not met the criteria of the checkpoint, the cell cy-

cle is slowed or even arrested until it does so.We further dis-

cuss cell cycle control in Section 34-4C.

B. Eukaryotic Replication Mechanisms

Much of what we know about eukaryotic DNA replication

has been learned from studies on budding yeast (Saccha-

romyces cerevisiae) and fission yeast (Schizosaccharomyces

pombe), the simplest eukaryotes, and on simian virus 40

(SV40), which has a 5243-bp circular DNA chromosome

that has only one replication origin. However, studies of

DNA replication in the cells of metazoa (multicellular ani-

mals), particularly Drosophila, Xenopus laevis (an African

clawed toad, whose eggs are easily studied), and humans,

have also led to important advances in our knowledge.

a. Eukaryotic Cells Contain Numerous

DNA Polymerases

The many known DNA polymerases can be classified

into six families based on phylogenetic relationships. Mem-

bers of families A (e.g., E. coli Pol I), B (e.g., E. coli Pol II),

and C (e.g., E. coli Pol III) encompass all replicative poly-

merases as well as some repair polymerases, family D occurs

only in archaea where its functions are poorly understood,

and families X and Y participate in DNA repair. The fingers

and thumb domains have structures that are unique to each

family, whereas the catalytic residue–containing palm do-

mains are similar in families A, B, and Y. Animal cells ex-

press at least four distinct types of DNA polymerases that

are implicated in DNA replication (Table 30-5). They are

designated, in the order of their discovery, DNA poly-

merases (pols) ␣, ␥, ␦, and ε (alternatively, POLA, POLG,

POLD1, and POLE), of which pol ␥ is a member of family A

and the others are members of family B.

been significant progress in our understanding of this es-

sential process. In this section, we outline what is known

about DNA replication in eukaryotes. We also discuss two

DNA polymerases that are peculiar to eukaryotic systems:

reverse transcriptase and telomerase.

A. The Cell Cycle

The cell cycle, the general sequence of events that occur

during the lifetime of a eukaryotic cell, is divided into four

distinct phases (Fig. 30-40):

1. Mitosis and cell division occur during the relatively

brief M phase (for mitosis).

2. This is followed by the G

phase (for gap),which cov-

ers the longest part of the cell cycle.This is the main period

of cell growth.

3. G

gives way to the S phase (for synthesis), which in

contrast to events in prokaryotes, is the only period in the

cell cycle when DNA is synthesized.

4. During the relatively short G

phase, the now

tetraploid cell prepares for mitosis. It then enters M phase

once again and thereby commences a new round of the cell

cycle.

The cell cycle for cells in culture typically occupies a 16-

to 24-h period. In contrast, cell cycle times for the different

types of cells of a multicellular organism may vary from as

little as 8 h to ⬎100 days. Most of this variation occurs in

the G

phase. Moreover, many terminally differentiated

cells, such as neurons or muscle cells, never divide; they as-

sume a quiescent state known as the G

phase.

A cell’s irreversible “decision” to proliferate is made

during G

. Quiescence is maintained if, for example, nutri-

ents are in short supply or the cell is in contact with other

cells (contact inhibition). Conversely, DNA synthesis may

be induced by various agents such as carcinogens or tumor

viruses, which trigger uncontrolled cell proliferation (can-

Figure 30-40 The eukaryotic cell cycle. Cells in G

may enter

a quiescent phase (G

) rather than continuing about the cycle.

Preparation

for mitosis

(2–6 h)

Mitosis

and cell

division

(1 h)

DNA

replication

(6–8 h)

Commitment

to DNA

replication

(~10 h)

Quiescence

(variable)

JWCL281_c30_1173-1259.qxd 8/10/10 9:11 PM Page 1202

3. There is a large N-terminal domain that A-family

polymerases lack.

4. The template strand enters the active site from a cleft

between the N-terminal and exonuclease domains,

whereas in A-family polymerases, it does so from the fin-

gers domain.

5. The newly formed dsDNA nearest the active site has

a B-DNA-like conformation instead of the A-DNA-like

conformation observed in A-family polymerases (Section

30-2Ae).

Section 30-4. Eukaryotic Replication 1203

Pol ␣ occurs only in the cell nucleus where it partici-

pates in the replication of chromosomal DNA. This func-

tion was largely established through the use of its specific

inhibitor aphidicolin

and by the observation that pol ␣ activity varies with the rate

of cellular proliferation. Pol ␣, as do all DNA polymerases,

replicates DNA by extending a primer 5¿S3¿ under the di-

rection of a single-stranded DNA template. This hetero-

tetramer, which lacks exonuclease activity, consists of a 167-

kD polymerase subunit, a 48-kD primase subunit, a 62-kD

subunit that is required for full primase activity, and a 79-kD

subunit that is implicated in the regulation of initiation,all of

which are collectively known as pol ␣/primase.

Pol ␦ is a heterotrimer whose 125-kD catalytic subunit

lacks an associated primase but contains a proofreading

3¿S5¿ exonuclease domain. The X-ray structure of the

yeast pol ␦ catalytic subunit (also called pol ␦), determined

by Aneel Aggarwal, reveals that this enzyme consists of

five domains arranged around a central hole that is near its

polymerase active site (Fig. 30-41). It has the right-hand-

like architecture first seen in A-family DNA polymerases

(Figs. 30-8 and 30-9), and its palm domain has a structurally

similar core that contains the two invariant Asp residues

implicated in the nucleotidyl transfer mechanism (Fig.30-10).

However, there are major differences between A-family

polymerases (e.g., Fig. 30-8) and pol ␦, which is representa-

tive of B-family polymerases. Most notably, in pol ␦:

1. The fingers domain, which consists of only a pair of

antiparallel helices, is rotated by ⬃60° relative to that in A-

family polymerases.

2. The exonuclease domain projects from the top of the

fingers domain rather than from the bottom of the palm

domain, as it does in A-family polymerases.

Aphidicolin

HOH

C CH

Figure 30-41 X-ray structure of yeast DNA polymerase ␦ (pol

␦) in complex with primer–template DNA and dCTP. The

protein is drawn in ribbon form with its five domains differently

colored as indicated.The DNA, whose primer and template

strands consist of 12 and 16 nt, together with the incoming dCTP,

is drawn in stick form with template C green, primer C cyan,

dCTP C magenta, N blue, O red, and P orange and with

successive P atoms in each DNA strand connected by orange

rods.The Ca

2⫹

ions at the polymerase (Pol) active site are

represented by dark green spheres as is the Ca

2⫹

ion at the

exonuclease (Exo) active site. [Based on an X-ray structure by

Aneel Aggarwal, Mount Sinai School of Medicine, New York,

New York. PDBid 3IAY.]

␣␥␦ε

Location Nucleus Mitochondrion Nucleus Nucleus

Subunit masses (kD)

167, 79, 62, 48 144 125, 55, 40 256, 78, 23, 22

(166, 66, 59, 50) (140, 55) (124, 51, 51) (262, 60, 17, 12)

Family B A B B

Table 30-5 Properties of Eukaryotic DNA Polymerases That Participate

in DNA Replication

Yeast S. cerevisiae (human cells).

Source: Mainly Johnson, A. and O’Donnell, M., Annu. Rev. Biochem. 74, 283 (2005).

JWCL281_c30_1173-1259.qxd 8/10/10 9:11 PM Page 1203

1204 Chapter 30. DNA Replication, Repair, and Recombination

near the primer strand, following which pol ␦ binds to the

PCNA and processively extends the DNA strand.

RFC, like the E. coli clamp loader, is a heteropentamer

of AAA⫹ family subunits, but whereas the clamp loader

contains three identical subunits (␥

␦␦¿), RFC consists of

five different subunits, RFC-A through RFC-E (alterna-

tively, RFC1–RFC5).The X-ray structure of yeast RFC in

complex with PCNA (Fig. 30-43), determined by O’Don-

nell and Kuriyan, reveals that RFC’s A, B, and C subunits

bind in conserved hydrophobic grooves on the face of

PCNA on which its C-terminal residues are located (since

PCNA’s three identical subunits are linked head-to-tail,

its two faces are different). In fact, this so-called C side of

PCNA, which faces away from the direction of poly-

merase motion, binds many of the proteins that partici-

pate in replicative processes, including most DNA poly-

merases, and hence PCNA plays a major role in

In contrast to pol ␣, which exhibits only moderate pro-

cessivity (⬃100 nucleotides), that of pol ␦ is essentially un-

limited (replicates the entire length of a template), but only

when it is in complex with a protein named proliferating

cell nuclear antigen (PCNA; so named because it occurs

only in the nuclei of proliferating cells and reacts with anti-

bodies produced by a subset of patients with the autoim-

mune disease systemic lupus erythematosus). The X-ray

structure of PCNA (Fig. 30-42), determined by Kuriyan, re-

veals that it forms a trimeric ring with almost identical

structure (and presumably function) as the E. coli ␤

slid-

ing clamp (Fig. 30-13).Thus, each PCNA subunit consists of

four rather than six of the structurally similar ␤␣␤␤␤ mo-

tifs from which the E. coli ␤ subunit is constructed. Intrigu-

ingly, PCNA and the ␤ subunit exhibit no significant se-

quence identity, even when their structurally similar

portions are aligned.Archaea also have sliding clamps with

pseudohexagonal symmetry.

Pol ␦ in complex with PCNA is required for lagging

strand synthesis. In contrast, pol ␣/primase functions to

synthesize ⬃12-nt RNA primers, which it extends by an ad-

ditional ⬃20 nt of DNA. Then, in a process called poly-

merase switching, the eukaryotic counterpart of the E. coli

clamp loader (Section 30-3Cc), replication factor C (RFC),

displaces the pol ␣ and loads PCNA on the template DNA

Figure 30-42 X-ray structure of human PCNA. Its three

subunits, which form a 3-fold symmetric ring, are drawn in ribbon

form embedded in their semitransparent surface diagram. One of

these subunits is colored in rainbow order from its N-terminus

(blue) to its C-terminus (red), another is pink, and the third is

light green.A space-filling model of B-DNA viewed along its

helix axis has been drawn in the center of the PCNA ring.

Compare this structure with that of the ␤

sliding clamp of the

E. coli Pol III holoenzyme (Fig. 30-13). [Based on an X-structure

by John Kuriyan, University of California at Berkeley. PDBid

1AXC.]

See Interactive Exercise 35

Figure 30-43 X-ray structure of yeast RFC in complex with

PCNA, ADP, and ATP␥S. Both proteins are drawn in tube-and-

arrow form with the RFC subunits colored as their homologs in

Fig. 30-34a (with the RFC-A, B, C, D, and E subunits

corresponding to the E. coli clamp loader’s ␦, ␥3, ␥2, ␥1, and ␦¿

subunits, respectively) and the PCNA colored as in Fig. 30-42.

The view is similar to that of the E. coli clamp loader in Fig. 30-34a.

ADP, which binds to only RFC-E, and ATP␥S, which binds to the

other four subunits, are shown in space-filling form with ATP␥S

C green,ADP C magenta, N blue, O red, P orange, and S yellow.

Note that all five RFC subunits have a nucleotide binding site,

whereas the E. coli ␦ and ␦¿ subunits do not. [Based on an

X-ray structure by Mike O’Donnell,The Rockefeller University,

and John Kuriyan, University of California at Berkeley. PDBid

1SXJ.]

JWCL281_c30_1173-1259.qxd 10/19/10 10:29 AM Page 1204

(replication units; DNA segments that are each served by

a replication origin) are activated simultaneously. New

replicons are activated throughout S phase until the en-

tire chromosome has been replicated. During this

process, replicons that have already been replicated are

distinguished from those that have not; that is, a cell’s

chromosomal DNA is replicated once and only once per

cell cycle.

c. The Assembly of the Eukaryotic Initiation Complex

Occurs in Two Stages

The once-and-only-once replication of eukaryotic DNA

per cell cycle is conferred by a type of binary switch.A pre-

replicative complex (pre-RC) is assembled at each replica-

tion origin during the G

phase of the cell cycle. This is the

only period of the cell cycle during which the pre-RC can

form and hence this process is known as licensing. How-

ever, a licensed pre-RC cannot initiate DNA replication.

Rather, it must be activated to do so, a process that occurs

only during S phase. This temporal separation of pre-RC

assembly and origin activation ensures that a new pre-RC

cannot assemble on an origin that has already “fired” (com-

menced replication) so that an origin can only fire once per

cell cycle. How does this occur?

The licensing process and how the pre-RC is activated

to form an initiation complex are still incompletely under-

stood. Thus, although it appears that most of the proteins

forming these complexes have been identified, their struc-

tures, interactions, and, in many cases, their functions are

largely unknown. Keeping this in mind, let us consider

what is known about these processes.

Replication origins are surprisingly variable among

species, often within the same organism, and even vary with

a given organism’s developmental stage. Thus, whereas S.

cerevisiae origins, which are known as autonomously repli-

cating sequences (ARS), contain a highly conserved 11-bp

AT-rich sequence within a less well defined ⬃125-bp re-

gion, some metazoan origins are dispersed over 10 to 50 kb

“initiation zones” that contain multiple origins and, in

some cases, require no specific DNA sequence at all. De-

spite this disparity, the proteins that participate in eukary-

otic DNA replication are highly conserved from yeast to

humans.

Section 30-4. Eukaryotic Replication 1205

recruiting the components of the various types of replica-

tion forks. The C-terminal domains of each RFC subunit

associate to form a ring-shaped collar, as do the C-termi-

nal domains of the E. coli clamp loader (Fig. 30-34). Like-

wise, RFC’s AAA⫹ domains are arranged in a right-

handed spiral that matches the helical path of the

sugar–phosphate backbone of B-DNA. Presumably,

prokaryotic and eukaryotic clamp loaders interact with

primer–template DNA and their corresponding sliding

clamps in similar ways.

Pol d, a heterotetrameric nuclear enzyme, is the most

enigmatic participant in DNA replication. Pol ε is highly

processive in the absence of PCNA and has a 3¿S5¿ ex-

onuclease activity that degrades single-stranded DNA to

6- or 7-residue oligonucleotides rather than to mononu-

cleotides, as does that of pol ␦. Although pol ε is necessary

for the viability of yeast, its essential function can be car-

ried out by only the noncatalytic C-terminal half of its

256-kD catalytic subunit, which is unique among B-family

DNA polymerases. This suggests that the C-terminal half

of the pol ε catalytic subunit is required for the assembly

of the replication complex. Nevertheless, Thomas Kunkel

has shown that pol ε is probably the leading strand repli-

case, although it may also contribute to lagging strand syn-

thesis. Moreover, pol ␦ may also participate in leading

strand synthesis.

Pol ␥, a monomer, occurs exclusively in the mitochon-

drion, where it presumably replicates the mitochondrial

DNA. Chloroplasts contain a similar enzyme.

Eukaryotic cells contain batteries of DNA polymerases.

These include the DNA polymerases that participate in

chromosomal DNA replication (pols ␣, ␦, and ε) and sev-

eral that take part in DNA repair processes (Section 30-5)

including pols ␤, ␩, ␫, ␬, and ␨ (alternatively, POLB, POLH,

POLI, POLK, and POLZ). Pol ␤, an X-family enzyme, is

remarkable for its small size (a 335-residue monomer in

humans).

b. Eukaryotic Chromosomes Consist of

Numerous Replicons

Eukaryotic and prokaryotic DNA replication systems

differ most obviously in that eukaryotic chromosomes

have multiple replication origins in contrast to the single

replication origin of prokaryotic chromosomes. Eukary-

otic cells replicate DNA at the rate of ⬃50 nt/s (⬃20

times slower than does E. coli) as was determined by

autoradiographically measuring the lengths of pulse-

labeled sections of eukaryotic chromosomes. Since a eu-

karyotic chromosome typically contains 60 times more

DNA than those of prokaryotes, its bidirectional replica-

tion from a single origin would require ⬃1 month to

complete. Electron micrographs such as Fig. 30-44, how-

ever, reveal that eukaryotic chromosomes contain multi-

ple origins, one every 3 to 300 kb depending on both the

species and the tissue, so that S phase usually occupies

only a few hours.

Cytological observations indicate that the various

chromosomal regions are not all replicated simultane-

ously; rather, clusters of 20 to 80 adjacent replicons

Figure 30-44 Electron micrograph of a fragment of replicating

Drosophila DNA. The arrows indicate its multiple replication

eyes. [From Kreigstein, H.J. and Hogness, D.S., Proc. Natl. Acad.

Sci. 71, 136 (1974).]

JWCL281_c30_1173-1259.qxd 8/10/10 9:11 PM Page 1205

1206 Chapter 30. DNA Replication, Repair, and Recombination

The conversion of a licensed pre-RC to an active initia-

tion complex requires the addition of pol ␣/primase, pol ε,

and several accessory proteins, which only occurs at the on-

set of S phase. This process begins with addition of Mcm10

protein (which shares no sequence similarity with any of

the subunits of the MCM complex) to the pre-RC, which

probably displaces Cdt1.This is followed by the addition of

at least two protein kinases, a Cdk and Ddk, the latter be-

ing a heterodimer of the protein kinase Cdc7 with its

activating subunit Dbf4 (Ddk stands for Dbf4-dependent

kinase). Ddk acts to phosphorylate five of the six MCM

subunits (all but Mcm2) so as to activate the MCM com-

plex as a helicase. In contrast, the way in which Cdks acti-

vate the pre-RC is poorly understood, although several

ORC and MCM proteins as well as Cdc6/Cdc18 are phos-

phorylated by Cdks. Ddk together with a Cdk also recruits

Cdc45 to the growing initiation complex. Cdc45, in turn, is

required for the assembly of the initiating synthetic ma-

chinery at the replication fork, including pol ␣/primase, pol

ε, PCNA, and replication protein A (RPA), the het-

erotrimeric eukaryotic counterpart of SSB, thereby form-

ing an active initiation complex.

d. Re-Replication Is Prevented through the Actions

of Cdks and Geminin

Once initiation (priming) has occurred, the initiation

complex is joined by RFC and pol ␦ and, as is described

above, is converted to an active replicative complex by

polymerase switching. DNA replication then proceeds

bidirectionally until each replication fork has collided with

an oppositely traveling replication fork, thereby complet-

ing the replication of the replicon. An active replication

fork will destroy any licensed pre-RCs and unfired initia-

tion complexes in its path, thereby preventing the DNA at

such sites from being replicated twice. Eukaryotes appear

to lack termination sequences and proteins analogous to

the Ter sites and Tus protein in E. coli.

Several redundant mechanisms ensure that a pre-RC

can initiate DNA synthesis only once. Cdks are active

from late G

phase through late M phase. These elevated

Cdk levels, which are required to activate initiation, also

prevent reinitiation. The Cdk-mediated phosphorylation of

Cdc6/Cdc18, which occurs late in G

after the pre-RCs have

formed, causes Cdc6/Cdc18 to be proteolytically degraded

in yeast and exported from the nucleus in mammalian cells.

Evidently, Cdc6/Cdc18 is only required for the assembly of

the pre-RC, not its activation. The helicase activity of the

MCM complex is inhibited by phosphorylation, at least in

vitro. Moreover, MCM proteins are exported from the nu-

cleus in G

and M phases, a process that is interrupted by

Cdk inactivation. However, the function of Cdk-mediated

phosphorylation of ORC proteins is unclear.

Metazoan cells have yet another mechanism to prevent

the assembly of a licensed pre-RC on already replicated

DNA. High levels of a protein named geminin appear in S

phase and continue to accumulate until late M phase, when

geminin is degraded. Geminin associates with Cdt1 (which

together with Cdc6/Cdc18 loads the MCM complex onto

The assembly of the pre-RC (Fig. 30-45) begins late in

M phase or early in G

phase with the binding of the ori-

gin recognition complex (ORC), a hexamer of related

proteins (Orc1 through Orc6), to the origin, where it re-

mains bound during most or all of the cell cycle. ORC, the

functional analog of DnaA protein in E. coli replication

initiation (Section 30-3Ca), then recruits two proteins,

Cdc6 in S. cerevisiae (Cdc18 in S. pombe; Cdc for cell divi-

sion cycle) and Cdt1. These proteins then cooperate with

the ORC to load the MCM complex [named for its

minichromosome (plasmid) maintenance functions], a

hexamer of related subunits (Mcm2 through Mcm7), onto

the DNA to yield the licensed pre-RC. The MCM com-

plex, a ring-shaped ATP-driven helicase, is the analog of

E. coli DnaB helicase, whereas Cdc6/Cdc18 together with

Cdt1 appears to be an analog of E. coli DnaC (which fa-

cilitates DnaB loading).With the exception of Cdt1, all of

these proteins, Orc1 through Orc6, Cdc6/Cdc18, Mcm2

through Mcm7, as well as E. coli DnaA, DnaB, and DnaC,

are AAA⫹ ATPases.

Figure 30-45 Schematic diagram for the assembly of the

eukaryotic pre-replicative complex (pre-RC). The actual

stoichiometries, positions, and interactions of its various

components are largely unknown.The pre-RC only forms during

the G

phase of the cell cycle.

Origin-containing DNA

Licensed pre-RC

ORC

Cdc6/Cdc18 Cdt1

MCM Complex

JWCL281_c30_1173-1259.qxd 8/10/10 9:11 PM Page 1206

C. Reverse Transcriptase

The retroviruses, which are RNA-containing eukaryotic

viruses such as certain tumor viruses and human immuno-

deficiency virus (HIV), contain an RNA-directed DNA

polymerase (reverse transcriptase). This enzyme, which

was independently discovered in 1970 by Howard Temin

and David Baltimore, acts much like Pol I in that it synthe-

sizes DNA in the 5¿S3¿ direction from primed templates.

In the case of reverse transcriptase, however, RNA is the

template. The discovery of reverse transcriptase caused a

mild sensation in the biochemical community because it

was perceived by some as being heretical to the central

dogma of molecular biology (Section 5-4). There is, how-

ever, no thermodynamic prohibition to the reverse tran-

scriptase reaction; in fact, under certain conditions, Pol I

can likewise copy RNA templates.

Reverse transcriptase transcribes the retrovirus’s single-

stranded RNA genome to a double-stranded DNA as fol-

lows (Fig. 30-47):

1. The retroviral RNA acts as a template for the synthe-

sis of its complementary DNA (RNA-directed DNA poly-

merase activity), yielding an RNA–DNA hybrid helix. The

DNA synthesis is primed by a host cell tRNA whose 3¿ end

partially unfolds to base pair with a complementary seg-

ment of the viral RNA.

2. The RNA strand is then nucleolytically degraded

(RNase H activity; H for hybrid).

3. The DNA strand acts as a template for the synthesis

of its complementary DNA (DNA-directed DNA poly-

merase activity), yielding double-stranded DNA.

The DNA is then integrated into a host cell chromosome.

Reverse transcriptase has been a particularly useful tool

in genetic engineering because of its ability to transcribe

mRNAs to complementary strands of DNA (cDNA). In

transcribing eukaryotic mRNAs, which have poly(A) tails

(Section 31-4Ab), the primer can be oligo(dT). cDNAs

have been used, for example, as probes in Southern blot-

ting (Section 5-5D) to identify the genes coding for their

Section 30-4. Eukaryotic Replication 1207

the ORC) so as to inhibit the assembly of the pre-RC.This

inhibition can be reversed by the addition of excess Cdt1. It

therefore seems likely that the presence of geminin pro-

vides protection against DNA re-replication under condi-

tions when Cdks are inhibited by checkpoint activation. In

addition, re-replication is also prevented by the degrada-

tion of Cdt1 after the origin with which it is associated has

fired.The requirement for DNA replication in this process

is indicated by its blockage by the DNA polymerase in-

hibitor aphidicolin.

Finally, cells that have shifted to the G

(quiescent)

phase of the cell cycle (Fig. 30-40)—the majority of cells in

the human body—cease making DNA. Such cells are char-

acterized by the absence of Cdk activity. In proliferating

cells, this would permit the re-replication of DNA. How-

ever,cells in G

also lack the proteins of the MCM complex

and are therefore incapable of assembling licensed pre-

RCs. Since cancerous cells are characterized by being in a

state of rapid proliferation (Section 19-3B), the presence of

MCM complex proteins in what should be quiescent cells is

a promising diagnostic marker for cancer.

e. Primers Are Removed by RNase H1 and Flap

Endonuclease-1

In lagging strand synthesis, when pol ␦ reaches the pre-

viously synthesized Okazaki fragment, it partially displaces

its RNA primer through DNA synthesis, thereby generat-

ing an RNA flap. The primer is then removed through the

actions of two enzymes: RNase H1 removes most of the

RNA leaving only a 5¿ ribonucleotide adjacent to the DNA,

which is then removed by flap endonuclease-1 (FEN1).

However, as we have seen, pol ␣/primase extends the RNA

primers it has made by ⬃20 nt of DNA before it is dis-

placed by pol ␦. Since pol ␣ lacks proofreading ability, this

primer extension is more likely to contain errors than the

DNA synthesized by pol ␦. However, FEN1, which is re-

cruited to the replication fork by its binding to the C side of

PCNA, provides what is, in effect, pol ␣’s proofreading

function: It is also an endonuclease that excises mismatch-

containing oligonucleotides up to 15 nt long from the 5¿

end of an annealed DNA strand. Moreover, FEN1 can

make several such excisions in succession to remove more

distant mismatches. The excised segment is later replaced

by pol ␦ as it synthesizes the succeeding Okazaki fragment.

f. Mitochondrial DNA Is Replicated via RNA

Okazaki Fragments

Mammalian mitochondria contain two to ten copies of

their ⬃16-kb circular chromosome. The pol ␥–mediated

replication of this chromosome occurs unidirectionally

from a single origin. In this process, as Ian Holt showed, the

lagging strand Okazaki fragments are entirely synthesized

as RNA (Fig. 30-46). The RNA is then replaced by DNA,

although the way this occurs is poorly understood. One

possibility is that this RNA is excised in much the same

way as the primers in the nucleus, that is, through the ac-

tions of RNase H1 and FEN1, followed by the synthesis of

lagging strand DNA by pol ␥.

Leading strand DNA

Lagging strand RNA

Okazaki fragments

Motion of

replication fork

Origin

Mitochondrial chromosome

Figure 30-46 Replication of the mammalian mitochondrial

chromosome. This circular chromosome is unidirectionally

replicated from a single origin by a process in which the Okazaki

fragments are entirely RNA.

JWCL281_c30_1173-1259.qxd 8/26/10 8:20 PM Page 1207

1208 Chapter 30. DNA Replication, Repair, and Recombination

NNN H

HOCH

⫺⫹

3'-Azido-3'-deoxythymidine

(AZT; zidovudine)

H H

HOCH

Hypoxanthine

2',3'-Dideoxyinosine,

(ddI; didanosine)

H H

HOCH

2',3'-Dideoxycytidine

(ddC; zalcitabine)

HOCH

2',3'-Didehydro-3'-

deoxythymidine

(stavudine)

are RT inhibitors. Unfortunately, resistant strains of HIV-1

arise quite rapidly because RT lacks a proofreading exonu-

clease function and hence is highly error prone.Thus, as we

have seen (Section 15-4Cd), effective long-term anti-HIV

therapy requires the concurrent administration of at least

one RT inhibitor and an HIV protease inhibitor.

Edward Arnold determined the X-ray structure of RT

complexed to a 21-bp primer–template DNA with a 5-nt

overhang at the 5¿ end of its template strand, a dideoxy-dG

(ddG) residue at the 3¿ end of its primer strand (which pre-

vents its further extension), and in complex with dATP

(Fig. 30-48). The polymerase domains of p66 and p51 each

contain four subdomains, which, because of their collective

resemblance in p66 to DNA polymerases, are named, from

N- to C-terminus, fingers, palm, thumb, and connection. In-

deed, reverse transcriptases form a separate family of DNA

polymerases, family RT. In p66, the RNase H domain fol-

lows the connection.

p51 has undergone a remarkable conformational

change relative to p66: The connection has rotated by 155°

and translated by 17 Å to bring it from a position in p66 in

which it contacts only the RNase H domain (Fig. 30-48a) to

one in p51 in which it contacts all three other polymerase

subdomains (Fig. 30-48b). This permits p66 and p51 to

bring different surfaces of their connections into juxtaposi-

tion to form, in part, RT’s DNA-binding groove. Thus, the

chemically identical polymerase domains of p66 and p51

are not related by 2-fold molecular symmetry (a rare but

not unprecedented phenomenon), but instead, associate in

a sort of head-to-head and tail-to-tail arrangement (Fig. 30-

48c). Consequently, RT has only one polymerase active site

and one RNase H active site.This is an example of viral ge-

netic economy: HIV-1, with its limited genome size, has

corresponding mRNAs. An RNA’s base sequence can be

easily determined by sequencing its cDNA (Section 7-2A).

a. X-Ray Structure of HIV-1 Reverse Transcriptase

HIV-1 reverse transcriptase (RT) is a dimeric protein

whose subunits are synthesized as identical 66-kD

polypeptides, known as p66 (p for protein), that each con-

tain a polymerase domain and an RNase H domain. How-

ever, the RNase H domain of one of the two subunits is

proteolytically excised, thereby yielding a 51-kD polypep-

tide named p51. Thus, RT is dimer of p66 and p51.

The first drugs to be clinically approved to treat AIDS,

3ⴕ-azido-3ⴕ-deoxythymidine (AZT; zidovudine), 2ⴕ,3ⴕ-

dideoxyinosine (ddI; didanosine), 2ⴕ,3ⴕ-dideoxycytidine

(ddC; zalcitabine), and 2ⴕ,3ⴕ-didehydro-3ⴕ-deoxythymidine

(stavudine),

3'5'

5'3'

RNA

DNA hybrid

RNase H

Double-stranded DNA

Single-stranded RNA

RNA-directed DNA polymerasedNTP

NMP

Single-stranded DNA

DNA-directed DNA polymerasedNTP

Figure 30-47 The reactions catalyzed by

reverse transcriptase.

JWCL281_c30_1173-1259.qxd 8/10/10 9:11 PM Page 1208

succeeded in using a single polypeptide for what are essen-

tially two different functions.

The dATP binds at the 3¿ end of the primer strand, near

p66’s three catalytically essential Asp side chains, where it

pairs with a template dT base.The DNA assumes a confor-

mation that, near the polymerase active site, resembles A-

DNA (note the A-DNA-like tilt of the bases with respect

to the helix axis below the dATP in Fig. 30-48c), but else-

where more closely resembles B-DNA (in which the bases

are nearly perpendicular to the helix axis), a phenomenon

that also has been observed in several structures of A-

family DNA polymerases in their complexes with DNA

(Section 30-2Ae). Most of the protein–DNA interactions

involve the DNA’s sugar–phosphate backbone.

The RT active site region contains the few sequence

motifs that are conserved among the various polymerases.

Indeed, this region of p66 has a striking structural resem-

blance to DNA polymerases of known structure (Sections

30-2A and 30-4Ba). This suggests that other DNA poly-

merases are likely to bind DNA in a similar manner.

D. Telomeres and Telomerase

The ends of linear chromosomes cannot be replicated by

any of the mechanisms we have yet considered. This is be-

cause the RNA primer at the 5¿ end of a completed lagging

strand cannot be replaced with DNA; the primer required

to do this would have no place to bind. How, then, are the

DNA sequences at the ends of eukaryotic chromosomes,

the telomeres (Greek: telos, end), replicated?

Telomeric DNA has an unusual sequence: It consists of

up to several thousand tandem repeats of a simple, species-

dependent, G-rich sequence concluding the 3¿-ending

strand of each chromosomal terminus. For example, the cil-

iated protozoan Tetrahymena has the repeating telomeric

sequence TTGGGG, whereas in all vertebrates it is

Section 30-4. Eukaryotic Replication 1209

Figure 30-48 X-ray structure of HIV-1 reverse transcriptase in

complex with primer–template DNA and dATP. (a) A ribbon

diagram of the p66 subunit in which the N-terminal fingers

domain is blue, the palm is red, the thumb is green, the

connection is yellow, and the RNase H domain is magenta.

(b) The p51 subunit with its palm subunit oriented similarly to

that of p66. Note the different orientations of its four domains

relative to p66. (c) A ribbon diagram of the HIV-1 RT p66/p51

heterodimer in complex with DNA and dATP.The domains of

p66 and p51 are colored as in Parts a and b (the labels indicate

subunit and domain; e.g., 51F and 66R denote the p51 finger

domain and the p66 RNase H domain).The DNA is drawn in

ladder form.The complex is oriented with its p66 polymerase

domain toward the top of the figure and viewed toward the

protein’s template–primer binding cleft (whose floor is largely

composed of the connection domains of p66 and p51). [Based on

an X-ray structure by Edward Arnold, Rutgers University.

PDBid 3JYT.]

See Interactive Exercise 36

(a)

(b)

(c)

JWCL281_c30_1173-1259.qxd 10/19/10 10:30 AM Page 1209

1210 Chapter 30. DNA Replication, Repair, and Recombination

TTAGGG. Moreover, this strand ends with an overhang

that varies from ⬃20 nt in yeast to ⬃200 bp in humans.

Elizabeth Blackburn, Carol Greider, and Jack Szostak

have shown that telomeric DNA is synthesized by a novel

mechanism.The enzyme that synthesizes the G-rich strand

of telomeric DNA is named telomerase. Tetrahymena

telomerase, for example, adds tandem repeats of the telo-

meric sequence TTGGGG to the 3¿ end of any G-rich

telomeric oligonucleotide independently of any exoge-

nously added template. A clue as to how this occurs came

from the discovery that telomerases are ribonucleopro-

teins whose RNA components contain a segment that is

complementary to the repeating telomeric sequence. This

sequence apparently acts as a template in a kind of reverse

transcriptase reaction that synthesizes the telomeric se-

quence, translocates to the DNA’s new 3¿ end, and repeats

the process (Fig. 30-49).This hypothesis is confirmed by the

observation that mutationally altering the telomerase

RNA gene segment complementary to telomere DNA re-

sults in telomere DNA with the corresponding altered se-

quence. In fact, telomerase’s highly conserved protein com-

ponent, which is named TERT, is homologous to known

reverse transcriptases (its RNA component is called TER).

The DNA strand complementary to the telomere’s G-rich

strand is apparently synthesized by the normal cellular ma-

chinery for lagging strand synthesis, thereby accounting for

the 3¿ overhang of the G-rich strand.

a. TERT Resembles Other DNA Polymerases

The X-ray structure of the 596-residue TERT from the

red flour beetle Tribolium castaneum, determined by Em-

manuel Skordalakes, reveals that this subunit contains the

familiar fingers–palm–thumb domain organization of

other DNA polymerases (Fig. 30-50) and, in particular, re-

sembles the corresponding domains of the HIV-1 reverse

transcriptase p66 subunit (Fig. 30-48). In addition, TERT

has an N-terminal RNA-binding domain named TRBD

(for telomere repeat binding domain). TRBD closes the

gap between the thumb and fingers, thereby yielding a

ringlike protein with a hole that is ⬃26 Å wide and ⬃21 Å

deep. This is sufficient to accommodate an ⬃8-bp segment

of double-stranded nucleic acid.

b. Telomeres Must Be Capped

Without the action of telomerase, a chromosome

would be shortened at both ends by 50 to 100 nt with

every cycle of DNA replication and cell division. It was

therefore initially assumed that, in the absence of active

telomerase, essential genes located near the ends of chro-

mosomes would eventually be lost, thereby killing the de-

scendents of the originally affected cells. However, it is

now evident that telomeres serve a vital chromosomal

function that is compromised before this can happen.

Free DNA ends, which are subject to nuclease degrada-

tion, trigger DNA damage repair systems that normally

function to rejoin the ends of broken chromosomes (as

well as cell cycle arrest until this has happened).Thus ex-

posed telomeric DNA would result in the end-to-end fu-

sion of chromosomes, a process that leads to chromoso-

mal instability and eventual cell death [fused

chromosomes often break in mitosis (their two cen-

tromeres may cause them to be pulled in opposite direc-

tions), activating DNA damage checkpoints]. However, in

a process known as capping, telomeric DNA is specifi-

cally bound by proteins that sequester the DNA ends.

There is mounting evidence that capping is a dynamic

process in which the probability of a telomere sponta-

neously upcapping increases as telomere length de-

creases. Since most somatic cells in multicellular organ-

isms have very low levels of telomerase activity, this

explains why such cells in culture can only undergo a lim-

ited number of doublings (20–60) before they reach

senescence (a stage in which they cease dividing) and

eventually die (Section 19-3B). Indeed, otherwise immor-

tal Tetrahymena cultures with mutationally impaired

telomerases exhibit characteristics reminiscent of senes-

cent mammalian cells before dying off. Apparently, the

loss of telomerase function in somatic cells is a basis for ag-

ing in multicellular organisms.

c. Telomere Length Correlates with Aging

There is strong experimental evidence in support of this

theory of aging.The analysis of cultured human fibroblasts

Figure 30-49 Proposed mechanism for the synthesis of

telomeric DNA by Tetrahymena telomerase. The telomere’s 5¿-

ending strand is later extended by normal lagging strand

synthesis. [After Greider, C.W. and Blackburn, E.H., Nature 337,

336 (1989).]

5′

3′ 5′

3′

5′3′

T T G

A A C C C C A A C

Telomeric DNA

Telomerase RNA

polymerize

5′

3′ 5′

5′3′

5′

3′ 5′

5′3′

TTGGGG

CCCAAC

translocate

dGTP + dTTP

3′

JWCL281_c30_1173-1259.qxd 8/10/10 9:11 PM Page 1210