9

DNA Replication
Living cells must be able to duplicate their entire set of genetic instructions
Goal To understand the chemistry
of DNA synthesis and how every time they divide. Likewise, multicellular organisms must be able to
DNA is replicated with high pass on complete copies of their genetic information to future generations.
accuracy. This requires that the instructions are stored in a form that is capable of
Objectives being duplicated. As we have seen (and will return to in Chapter 11), genetic
instructions are embedded in the order of nucleobases in the DNA. As we
After this chapter, you should be able to have also seen, the self-complementary nature of the double helix provides
• describe the experiment that proved a simple (in principle) templating mechanism for replicating DNA into two
that DNA replication is semi- identical copies. Only DNA (and in some instances RNA when, as in the
conservative. case of the genomes of RNA viruses, it is used as a repository of genetic
• diagram the reaction for information) are self-complementary and hence capable of being replicated.
phosphodiester bond formation. The other three categories of macromolecules—carbohydrates, lipids, and
• explain the energetics of DNA proteins—are not self-complementary and do not serve as templates for
synthesis. their own production. Instead, the synthesis of these macromolecules is
• explain why the 5’-to-3’ rule creates a ultimately directed by information stored in DNA through the action of
conundrum during replication. enzymes and other proteins. Here we focus on the chemical and enzymatic
• explain how DNA is replicated
mechanisms by which DNA acts as a template for its own duplication and
accurately. how this replication process is carried out accurately and rapidly.
• explain how and why damage to DNA
is repaired. DNA replication is semi-conservative
As we saw in Chapter 8, the self-complementarity of DNA suggested a
mechanism for its replication. Specifically, Watson and Crick imagined
that base pairing would make it possible for each polynucleotide strand
of the double helix to serve as a template for the synthesis of a new strand.
Chapter 9 DNA Replication 2
That is, and as depicted in Figure 10 of Chapter 8, the parental double helix
would separate into two strands, each of which would serve as a template
for the synthesis of a new, complementary strand. In this scenario, known
as semi-conservative replication, each daughter molecule is a hybrid helix
in which one strand is parental (conserved from the parental helix) and
the other newly synthesized. At the same time, an alternative mechanism
known as conservative replication was a formal possibility. In this scenario
the original double helix is left intact (both of its original strands are
conserved) and an entirely new helix is somehow generated that consists
of two newly synthesized strands. Certainly, semi-conservative replication
was more appealing because it was hard to imagine a mechanism by which
conservative replication could take place. But this didn’t mean that DNA is
not replicated by a conservative mechanism. How then was it determined
which of the two models is correct?
In 1958, Matthew Meselson and Franklin Stahl reported an experiment
that elegantly distinguished between semi-conservative and conservative
replication (Figures 1 and 2). Thanks to its simplicity and decisiveness, the
Meselson-Stahl experiment has been called “the most beautiful experiment
in biology.” The key to the Meselson-Stahl experiment was growing cells
of Escherichia coli in medium containing a heavy isotope of nitrogen, 15N,
for many generations. E. coli needs nitrogen as a nutrient and incorporates
the isotope from the growth medium into its proteins and nucleic acids.

Conservative Semi-conservative
H “Heavy” strand, H
H contains 15N H

Grown in 15N

“Light” strand,
contains 14N
H L L H
H L H L

One cycle of division

in 14N

H L L L L L L H
H L L L L H L L

Two cycles of division

in 14N

Figure 1 The Meselson-Stahl experiment showed that DNA replication is semi-conservative

Cells were grown in medium with the heavy isotope of nitrogen (15N), which led to the synthesis of “heavy” DNA strands, shown in blue.
These cells were then grown for either one or two division cycles in medium containing the common isotope of nitrogen (14N). Newly
synthesized “light” DNA strands, shown in red, are produced using the 14N.
Chapter 9 DNA Replication 3

(A) (B)

centrifugal force

Grown in 15N

One cycle of
division in 14N

Two cycles of
division in 14N

light/ heavy/ heavy/

light light heavy

Figure 2 The Meselson-Stahl experiment analyzed the isotopic composition of DNA by measuring its buoyant
density
Shown in (A) are the original experiments from the publication of Meselson and Stahl [Proc. Natl. Acad. Sci. 44: 671 (1958)] and in (B)
a cartoon that depicts how centrifugation separates DNA molecules according to their bouyant density in a cesium gradient. The dark
vertical bands in the left panel of (A) show the positions of the DNA molecules that were visualized by staining. The column of adjacent
plots show quantitative distributions of the DNA. DNA was extracted from cells after the indicated number of generations (rounds of cell
division) after transfer from growth in medium with 15N to medium with 14N (from generation 0 to generation 4.1). The red rectangle
highlights the densities after approxiately two (generation 1.9) rounds of division.

Next, they transferred the bacteria for various periods of time to medium
containing 14N (the common form of nitrogen with an atomic mass of 14)
such that any DNA synthesized after the transfer would not contain the
heavy nitrogen isotope. Finally, they extracted DNA from the cells and
subjected the DNA to high centrifugal forces in tubes containing cesium
chloride (using an instrument called an ultracentrifuge, in which tubes
are spun at high velocity). Cesium forms a density gradient under such
conditions; at equilibrium, DNA molecules in the gradient localize to a
position that matches their own buoyant density. Thus, parental DNA in
which both strands contain the heavy isotope would have a high density
(blue-blue in Figures 1 and 2) and localize near the bottom of the gradient.
DNA in which both strands are newly synthesized and hence light (red-
red) would localize near the top. Finally, hybrid molecules (blue-red) in
which one strand was heavy and one light would localize at an intermediate
position.
What did Meselson and Stahl observe? When E. coli cells were shifted to
14
N-containing medium for enough time for the DNA molecules to undergo
Chapter 9 DNA Replication 4
one round of replication, the DNA was found to exhibit an intermediate
buoyant density consistent with that expected for hybrid DNA instead
of the high density of the parental, heavy-heavy DNA (Figure 2). This fit
with the expectation from the semi-conservative model that each strand
was serving as a template for the other strand. Next, Meselson and Stahl
went a step further: they extracted DNA from cells that had been grown
for two cycles of division in the light, 14N-medium. Half of the DNA from
these cells exhibited the same buoyant density as that seen after a single
round of division (halfway between the densities of 14N- and 15N-labeled
DNA), and the other half exhibited a density corresponding to that of DNA
that only contained 14N (Figure 2). Thus after two rounds of replication,
DNA was generated in which both strands were newly synthesized. In toto,
these results met the expectation from a model in which parental DNA
contributes one of its two strands to each of the double helices generated
during each round of replication but were inconsistent with a model in
which replication generates a double helix in which both strands are newly
synthesized. Ergo, DNA replication is semi-conservative.

DNA is synthesized by the repetitive addition of nucleotides to the 3’

end of the growing polynucleotide chain
DNA is synthesized by an iterative process in which nucleotides are added
(one after another) to the 3’ end of a growing polynucleotide chain. This
growing strand, which we also refer to as the daughter strand, is base
paired in anti-parallel orientation to a second strand, the template strand,
which provides instructions via base pairing for the successive addition
of nucleotides (Figure 3). Each incoming nucleotide must properly pair
with the corresponding base in the template in order for a phosphodiester
bond to form and for the process to proceed on to the addition of the next
nucleotide. Because nucleotides are added at the 3’ end of the growing
strand, DNA synthesis proceeds in a 5’-to-3’ direction.

Figure 3 DNA synthesis proceeds 5’

in a 5’-to-3’ direction
C
During DNA synthesis one strand of
the helix serves as a template to specify Template strand C
incoming nucleotides that are added at the Polymerization
3’ end of the growing daughter strand. T
C G 3’
G C
T A Growing daughter
strand
G C
3’ 5’
Chapter 9 DNA Replication 5

γ β α NH2 O
gamma beta alpha O O O N O O O
O P O P O P O N O P O P O P O NH
O O O O N O O O O N
N
O O O O
dATP dTTP
O P O P O P O OH OH

O O O O Base
O NH2
O O O N O O O
O P O P O P O NH O P O P O P O N
dNTP O O O O N
N
O O O O N
O
OH dGTP
NH2
dCTP
OH OH

Figure 4 2’-deoxynucleoside triphosphates are substrates for DNA replication

The nucleotide substrates for DNA synthesis are 2’-deoxynucleoside

triphosphates
The substrates for DNA synthesis are 2’-deoxynucleoside triphosphates
(Figure 4). As we learned in Chapter 8, the term nucleoside refers to a
base and a sugar, in this case the sugar 2’-deoxyribose. The substrates for
DNA synthesis are 2’-deoxynucleosides that additionally carry a chain of
three phosphates at the 5’ carbon of the nucleoside. Hence these substrates
can be referred to as 2’-deoxynucleoside triphosphates or more simply as
2’-deoxynucleotides (in that they consist of a base, a sugar, and phosphates)
or even more simply as nucleotides, a shorthand that is frequently used. The
three phosphate groups in 2’-deoxynucleoside triphosphates are designated
using Greek letters as alpha (α), beta (β), and gamma (γ), with the phosphate
that is directly attached to the sugar being the α-phosphate, the middle
phosphate being the β-phosphate, and the phosphate most distal to the
sugar being the γ-phosphate. As we will explain, the β- and γ-phosphates
are released from the 2’-deoxynucleoside triphosphate substrates during
the chemical reaction in which a 2’-deoxynucleotide is added to the 3’ end
of a growing chain; that is, only the α-phosphate is retained in the growing
chain.
Henceforth, for simplicity we will use the abbreviations dNMP, dNDP, and
dNTP for 2’-deoxynucleosides bearing one (mono), two (di) and three (tri)
phosphates, respectively.

Polymerization of 2’-deoxynucleoside triphosphates involves the release

of the β- and γ-phosphates as pyrophosphate
The addition of a dNTP (2’-deoxynucleoside triphosphate) to the 3’
end of the growing strand during DNA synthesis involves the release
of pyrophosphate (Figure 5). Pyrophosphate is a diphosphate (P2O74−,
abbreviated PPi), and in DNA synthesis it is derived from the β- and
γ-phosphates. Pyrophosphate is released from the substrate when a covalent
bond is formed between the 3’ oxygen of the growing polynucleotide and
the α-phosphate of the incoming nucleotide. Pyrophosphate is a leaving
group (Chapter 4) for the DNA synthesis reaction.
Chapter 9 DNA Replication 6

Figure 5 Each step of DNA (A)

synthesis consumes a dNTP (dNMP)n + dNTP (dNMP)n+1 + PPi
molecule and releases a
pyrophosphate molecule (B) 5’ 5’
(A) The chemical reaction for DNA C C
synthesis in which the subscript “n” C C
represents the length of the growing
3’
daughter strand (dNMP)n. (B) A new T 3’ T A
nucleotide is added to the 3’ end of a C G + dATP C G + PPi
growing DNA strand (red) during each G C G C
DNA synthesis reaction. The identity of the
T A T A
nucleotide is specified by complementarity
with the template strand (blue). The G C G C
incoming nucleotide is incorporated from 3’ 5’ 3’ 5’
a dNTP (green), and the addition of the
nucleotide to the growing strand releases a
5’
molecule of pyrophosphate (PPi). 5’

3’
3’

3’ 5’ 3’ 5’

Phosphodiester bonds are formed by nucleophilic attack of the 3’ oxygen

on the α-phosphate of the incoming dNTP
The mechanism for the DNA polymerization reaction is shown in Figure
6. The terminal 3’ oxygen of the growing strand acts as a nucleophile and
forms a covalent bond with the phosphorus atom of the α-phosphate group
of the incoming dNTP. Because phosphorus cannot form six bonds, it
must lose one of its bonds, namely the bond to the oxygen that connects
it to the β- and γ-phosphates, resulting as we have seen in the release of
pyrophosphate.
Notice also in Figure 6, and as introduced above, that the incoming
nucleotide, dATP (green), is specified by base pairing with a thymine (blue)
at the corresponding position on the complementary template strand and
that the 3’ oxygen that acts as a nucleophile arises from a dGMP (red) that
had been incorporated in the previous round of phosphodiester bond
formation by pairing with a cytosine in the template strand.

DNA polymerization is coupled to pyrophosphate hydrolysis

DNA synthesis is catalyzed by an enzyme called DNA polymerase. It turns
out that the reaction catalyzed by DNA polymerase—the formation of a
phosphodiester bond and the release of pyrophosphate—is only modestly
favorable (Figure 7, Reaction 1). Yet, the cell needs to replicate DNA
Chapter 9 DNA Replication 7

5’
Figure 6 The 3’ oxygen acts
as a nucleophile during DNA
5’ O O H N
H
A N 3’
polymerization N H
A 3’ O N N N OH
dATP
The arrow-pushing mechanism for DNA O N
polymerization is shown. The nucleophile
T dATP O
O O O
P O O
is the 3’ oxygen at the 3’ end of the growing C G O
O
H
N H
O P
O P O P O
O O
DNA strand. Electrons from this oxygen C G N O O
are used to form a new bond with the O N H N O H
N
phosphorus atom of the α-phosphate G C N
N
O O H N
group of the incoming dNTP. The β- and A T P
O H
O
O
γ-phosphate groups leave together, forming 3’ 5’ O
O O P
a molecule of pyrophosphate. O
3’ O

5’

5’
5’ H
A O O H N
N 3’ Pyrophosphate
A 3’ O N N H
N N OH O O
T A O O N O P O P O
O O O O
C G O
P H O
O N H O
C G O N P
O
O
G C O
N
N H N N
N
A T O O H N O
O
3’ 5’ O
P
O
H O
O P
O
3’ O

5’

efficiently and completely. How then does the cell succeed in replicating its
genetic material if the fundamental step in DNA synthesis is not energetically
highly favorable? The answer is that phosphodiester bond formation is
coupled to a second, subsequent reaction in which the newly released
pyrophosphate is hydrolyzed to give rise to two molecules of inorganic
phosphate (PO43− or Pi) by the enzyme pyrophosphatase (Figure 7, Reaction
2). The hydrolysis of pyrophosphate is also moderately favorable. However,
in sum, the thermodynamics of these two successive reactions is highly
favorable (Figure 7, combined reaction). That is, the sum of the ΔG°rxn for

Reaction 1 (dNMP)n + dNTP (dNMP)n+1 + PPi ∆G°rxn = -3.5 kcal/mol

Reaction 2 PPi + H2O 2 Pi ∆G°rxn = -4.0 kcal/mol

Combined (dNMP)n + dNTP + H2O (dNMP)n+1 + 2 Pi ∆G°rxn = -7.5 kcal/mol

reaction
Figure 7 The favorability of DNA polymerization is enhanced by the hydrolysis of pyrophosphate
Chapter 9 DNA Replication 8
Reactions 1 and 2 equals −7.5 kcal/mol. Using the equation that we learned
in Chapter 3 relating free energy to the equilibrium constant (ΔG°rxn = −RT
ln Keq ), we can calculate that the Keq for the combined reaction is about
300,000, meaning that the products of the reaction are much more stable
than the reactants.
A simple mnemonic for quickly carrying out such calculations is that a
ΔG°rxn of ~−2.7 kcal/mol is roughly equivalent to a Keq of ~102. Ergo, −7.5
kcal/mol would correspond to an equilibrium constant of (102)(7.5/2.7) or ~3
x 105. Because many reactions in living systems have a ΔG°rxn of only a
few kilocalories/mole, this shorthand is a simple and convenient way to
estimate equilibrium constants.
Finally, we note that this coupling of phosphodiester bond formation to
pyrophosphate hydrolysis is a striking example of Le Châtelier’s principle
(analogous to the case of photosynthesis we considered in Box 1 of Chapter
3) operating in a living system. By depleting the cell of pyrophosphate,
pyrophosphatase perturbs the equilibrium for Reaction 1, draining
pyrophosphate and driving the reaction further in the direction of
phosphodiester bond formation.

DNA synthesis takes place at moving replication forks

DNA replication requires that the two strands of the helix are pulled apart
and that both are copied into daughter strands. This strand separation
creates a moving replication fork downstream of which the DNA remains
wound in a double helix and upstream of which the two strands are
separated and are each serving as a template for the synthesis of daughter
strands. It is a moving replication fork (as we will see, it moves very fast!) in
that downstream DNA is continuously being unwound and upstream DNA
is continuously being replicated into hybrid helices of old and new DNA
strands.
But this creates a conundrum! If DNA only grows in a 5’-to-3’ direction and
if the two strands of the double helix have an anti-parallel orientation, then
how can DNA be copied continuously on both template strands without
violating one of the two fundamental rules of nucleic acid chemistry?

DNA synthesis at the replication fork is continuous and discontinuous

The answer to the conundrum is that DNA is copied continuously on
one template strand and discontinuously on the other (Figure 8). One
template strand is oriented such that the 5’-to-3’ direction in which the
daughter strand synthesis is taking place is aligned with the direction of
fork movement; that is, it points toward the base of the fork where the helix
is being unwound and where fresh, single-stranded template DNA is being
generated. This is known as leading strand synthesis, and it take places in
a continuous manner.
DNA copied from the other template strand, on the other hand, is
discontinuous. It occurs in short bursts that point away from the fork. This
is known as lagging strand synthesis. Lagging strand synthesis takes place
in a 5’-to-3’ direction. So the rule that polynucleotide synthesis proceeds
Chapter 9 DNA Replication 9

leading strand 3'

continuous synthesis 5'

Okazaki fragments 3' 5'

5' 3'

3'

5'
lagging strand 3'
discontinuous synthesis
5'

overall replication fork movement

Figure 8 The leading strand is synthesized continuously, but the lagging strand is not

5’ to 3’ is not violated. Instead, short stretches of DNA are synthesized

and then joined to each other by an enzyme known as a DNA ligase to
create long, continuous DNA strands. A ligase is an enzyme that joins two
polynucleotide chains together by creating a phosphodiester bond between
the 5’ end of one chain and the 3’ end of the other.
The short stretches of DNA generated during lagging strand
synthesis are commonly known as Okazaki fragments after their discoverer
Reiji Okazaki. In bacteria, Okazaki fragments are 1,000-2,000 nucleotides
in length, whereas in the cells of higher organisms they are typically only
100-200 nucleotides long.
To sum up, as the replication fork proceeds in the downstream direction,
additional single-stranded templates are exposed for leading strand synthesis
and lagging strand synthesis. Both take place in a 5’-to-3’ direction. But in
the case of lagging strand synthesis, this synthesis is discontinuous, with
each newly synthesized segment of DNA being joined to the previously
synthesized segments by ligase to create an intact daughter strand.

DNA polymerase catalyzes template-directed DNA synthesis

As we learned in Chapter 4, thermodynamically favorable reactions in the
cell often require a protein catalyst in order for the reaction to proceed
rapidly. This is because even highly favorable reactions often must overcome
a kinetic barrier before the reactants can proceed on to products in a time
frame compatible with life. In other words, many biological reactions have
a high activation energy or ΔG‡.
DNA polymerase is the enzyme that accelerates the rate of phosphodiester
bond formation between the 3’ end of a polynucleotide chain and a dNTP
Chapter 9 DNA Replication 10

Figure 9DNA polymerase (A) “fingers”

“thumb”
catalyzes DNA polymerization
(A) Cartoon diagram of DNA polymerase, newly polymerized DNA
polymerization
showing the fingers, thumb, and site
polymerization site. (B) Crystal structure of
DNA polymerase bound to a piece of DNA.
The fingers, thumb, and polymerization
site are indicated. The location of the template
proofreading site or editing pocket, which DNA
we will discuss shortly, is also shown.

(B)

“fingers” “thumb”

template
DNA

newly
polymerized
DNA
polymerization
site proofreading
site

molecule. The substrate for DNA polymerase is a polynucleotide strand

annealed to a longer template strand. The polynucleotide provides the 3’
hydroxyl at which the dNTP is attached. The template strand, which extends
past the 3’ hydroxyl, specifies the specific nucleotide that will be appended
to the growing strand through base pairing (as we have seen).
Figure 9B shows the atomic structure of DNA polymerase as determined
by X-ray crystallography. Roughly speaking, it resembles a hand in which
the polymerization site or catalytic center is in the palm and is surrounded
by thumb-like and finger-like features of the enzyme (Figure 9A). The
3’ end of the growing DNA strand is located near this catalytic center
during replication. The catalytic center is formed from a set of two to three
aspartate residues that bind two magnesium ions and hold them in specific
positions (Figure 10). These magnesium ions accelerate the polymerization
reaction by making the α-phosphate more electrophilic and the 3’ hydroxyl
group more nucleophilic. The magnesium ions bind the triphosphate
portion of the dNTP and thereby remove some of the negative charge that
otherwise shields the phosphorus atom of the α-phosphate from incoming
nucleophiles. In addition, one of the magnesium ions forms an ion-dipole
interaction with the 3’ hydroxyl group at the 3’ end of the growing strand.
By donating some of its electron density to form this interaction, the
oxygen atom of this 3’ hydroxyl group becomes more positive. To counter
this withdrawal of electrons, oxygen, as the more electronegative partner,
attracts the electrons it shares with hydrogen even more strongly than
Chapter 9 DNA Replication 11

Figure 10 The active site of DNA (A)

polymerase accelerates the rate of γ
the polymerization reaction β
(A) Shown is the X-ray crystal structure Mg2+
dCTP
of DNA polymerase showing the
polymerization active site. The template α
strand is shown in cyan, the growing strand
is shown in magenta, and DNA polymerase Asp
is shown in green. (B) Shown is a two-
dimensional representation of the structure Mg 2+

shown in (A).
template
strand

(B)

usual, ultimately breaking the O-H bond and creating an oxygen anion.
The additional negative charge carried by oxygen makes it a more effective
nucleophile than it would have been had it been part of a hydroxyl group.
Proper alignment of the substrate dNTP with the oxygen anion requires that
the incoming nucleotide pair with the corresponding base on the template
strand. Conversely, an incoming nucleotide with a non-complementary base
would be unable to align properly in the active site and hence fail to react
with the oxygen anion. In other words, pairing with the template strand is
intimately involved in the chemical reaction that promotes phosphodiester
bond formation. Therefore, the ability of DNA polymerase to ensure that
the correct, complementary nucleotide is incorporated at the 3’ hydroxyl of
the growing strand is embedded in the architecture of the active site of the
enzyme.
In sum, DNA polymerase lowers the activation energy for phosphodiester
bond formation (reduces ΔG‡) by aligning the substrates for the reaction in
its catalytic center and by promoting nucleophilic attack of the 3’ oxygen on
the α-phosphate of the incoming dNTP.

DNA polymerase is processive

Not only does DNA polymerase catalyze phosphodiester bond formation,
but it also does so repeatedly. Indeed, the same DNA polymerase molecule
Chapter 9 DNA Replication 12

(A) (B)

Sliding clamp DNA polymerase

5’
Template DNA
3’

5’

New DNA

Direction of polymerization

Figure 11 The sliding clamp allows DNA polymerase to be processive

(A) Cartoon representation of the ring-shaped sliding clamp surrounding the DNA helix and anchoring DNA polymerase to the DNA.
(B) X-ray structure of a sliding clamp with a cartoon representation of DNA passing through the hole in the center. The clamp is composed
of two proteins, which are shown in green and cyan.

successively attaches one nucleotide after another to the growing daughter

strand much faster than it dissociates from the DNA. This property of
remaining attached to the DNA through many rounds of nucleotide
addition is referred to as processivity. Processivity maximizes the speed
of DNA synthesis. If the polymerase frequently fell off the DNA and had
to re-bind in order to resume nucleotide incorporation, the rate of DNA
synthesis would be much slower.
What keeps the enzyme anchored to the DNA? The answer is geometrically
elegant and simple: DNA polymerase is tethered to the DNA by a sliding
clamp that fully encircles the DNA helix (Figure 11A). The sliding clamp
is a complex of proteins that forms a complete circle around the DNA. In
effect, it resembles a doughnut, with the DNA passing through the hole in
the doughnut. The sliding clamp slides on the DNA along with the DNA
polymerase to which it is attached and travels with it as DNA synthesis
proceeds. Thus, the sliding clamp tethers the polymerase to the DNA by
means of a topologically closed structure. The beautiful X-ray structure of
the clamp in Figure 11B shows how the hole in the doughnut accommodates
the DNA.

DNA polymerase and the sliding clamp are part of a larger complex of
proteins that constitutes a DNA replication machine
DNA polymerase and the sliding clamp act in concert with a larger cohort
of proteins that drive movement of the replication fork and coordinate
DNA synthesis on the leading and lagging strands. For example, an enzyme
Chapter 9 DNA Replication 13
known as DNA helicase unwinds double-stranded DNA downstream of
the replication fork to separate the two strands that will serve as templates
for daughter strand synthesis. Another protein is involved in priming the
synthesis of the Okazaki fragments on the lagging strand. DNA polymerase
can only extend an existing polynucleotide chain. Bursts of DNA synthesis
on the lagging strand require a special enzyme known as primase that
produces a short stretch of nucleotides that can be extended by the DNA
polymerase. A clamp loader then assembles a sliding clamp on the DNA
behind the DNA polymerase that will produce the Okazaki fragment. And,
as we have seen, a DNA ligase will merge the Okazaki fragments into an
uninterrupted polynucleotide strand.
The entire set of proteins that mediate DNA replication can be thought of as
a molecular machine with multiple working parts that acts in a coordinated
fashion to bring about the replication of the genetic material. And it does
so remarkably rapidly. The replication machine synthesizes DNA at a rate
of 800 nucleotides per second. Consider that the chromosome of E. coli
consists of almost five million base pairs. Because DNA replication in E.
coli takes place simultaneously from two replication forks, the overall rate
of DNA synthesis is 1,600 nucleotides per second. Thus, E. coli is capable of
duplicating its entire chromosome in as little as 40 minutes, and as we will
see below, it does so with considerable accuracy.
The replication machine is only one of many molecular machines that
carry out the workings of the cell. We will encounter others below and
in later chapters. The very basis for life can be thought of as an ensemble
of molecular machines that carry out the chemical transactions of living
systems.

DNA is replicated with high fidelity

An extraordinary feature of DNA synthesis is its accuracy; it achieves an
error rate of only about one mistake for every 1010 nucleotides incorporated.
How is this accuracy achieved? Part of the answer is that DNA polymerase
incorporates correct nucleotides more effectively than incorrect
nucleotides. This selectivity is dependent on DNA polymerase’s ability to
recognize correctly matched base pairs in its active site. DNA polymerase
does not recognize specific base pairs; instead, it recognizes the geometry
(i.e., shape and size) of the base pair (Figure 12). As we have seen, A:T
and G:C base pairs are similar in structure and are approximately the same
width. When a correctly matched base pair is present in the enzyme active
site, the α-phosphate of the incoming nucleotide is optimally positioned
for the nucleophilic attack of the 3’ hydroxyl at the end of the growing
strand. In contrast, a mismatched pair of nucleotides has a different shape,
and as a consequence, the α-phosphate is not positioned properly to
facilitate the nucleophilic attack of the 3’ hydroxyl. Because the reactants
are not optimally positioned, the reaction proceeds slowly for mismatched
nucleotides; this results in a lag time that gives the mismatched nucleotide
time to dissociate from the polymerase.
Nonetheless, DNA synthesis is not error-free, allowing about one
Chapter 9 DNA Replication 14

(A) Base pair match (B) Base pair mismatch

DNA polymerase Improper DNA polymerase
base pairing H H
N
H
O O H N O O H
N N N α-phosphate
T G
O N T N H
N OH O N N H
O
not positioned
A N N O near 3’ OH
N N O
O O O O
O O O O P O O O
O O
O
P H O O O
P H OH O
O P O P O P O O P O P O
N H N H
O N O N O O
O O O
O C N O H O C N O H
N
H N G N
N
H N G N
N N
O O H N O O H N
O O
O H O O H O
P P
O O
O O P O O P
O O
O O

Figure 12 DNA polymerase catalyzes the addition of correct base pairs more effectively than it catalyzes the
addition of incorrect base pairs

mistake for every 105 phosphodiester bonds formed. Therefore, the ability
of DNA polymerase to recognize correctly paired nucleotides in its active
site is insufficient to explain the extraordinary fidelity of DNA synthesis.

DNA polymerase removes misincorporated nucleotides by proofreading

DNA polymerase is able to achieve further accuracy by proofreading its own
work. That is, if an incorrect nucleotide is incorporated, DNA polymerase
is often capable of catching the error and removing the misincorporated
nucleotide. Proofreading takes advantage of the fact that it is difficult for
DNA polymerase to add a new nucleotide onto an improperly incorporated
nucleotide at the 3’ end of the growing chain. In other words, the nucleotide
at the 3’ end must be correctly paired with the template strand in order
to be properly aligned for polymerization to proceed rapidly to the next
nucleotide. If not correctly paired, the 3’ nucleotide in the stalled polymerase
is able to slip out of the catalytic center and into an editing pocket on the
enzyme (Figure 9). The editing pocket contains a catalytic activity that
removes the nucleotide at the 3’ end of the daughter strand (a 3’ nuclease).
The resulting 3’ end of the growing chain with the misincorporated
nucleotide removed is able to slip back into the catalytic center, where a
fresh round of base pairing and phosphodiester bond formation can take
place with a new incoming nucleotide.
Proofreading improves the accuracy of DNA synthesis by a factor of
about 100. But what happens to errors that escape proofreading and are
incorporated into the daughter strand of the double helix? Cells have
a further back-up system that catches and repairs such errors known as
Chapter 9 DNA Replication 15
mismatch repair. Mismatch repair improves the accuracy of DNA synthesis
by another two or three orders of magnitude. The proteins of the mismatch
repair system scan the double helix for distortions (recall that only properly
paired bases are accommodated within a diameter of 20 Å). The repair
system then removes a stretch of DNA containing the mismatch from the
daughter strand and replaces it with newly synthesized DNA using the
parental strand as a template.
Mismatch repair is in a race against time. If the mismatch is not caught and
repaired before the DNA is fully replicated and a new round of replication
has commenced, then the misincorporated nucleotide will now serve as a
template for a new daughter strand. At this point, the DNA will no longer
contain a mismatch, and the misincorporated nucleotide will have been
permanently incorporated into the DNA as a mutation.

DNA damage can occur independently of replication

DNA is susceptible to damage even after it is replicated. Chemicals and
radiation in the environment can damage DNA. If left unrepaired,
environmentally induced DNA damage results in mutations. Cells have
multiple molecular machines for detecting and repairing such damage. For
example, ultraviolet light from the sun damages pyrimidines. Individuals
with a condition known as xeroderma pigmentosum are defective in the
machine for repairing this damage. As a consequence, such individuals are
hypersensitive to sunlight and readily contract skin cancer.
Some mutations arise without environmental influences. Remember that
DNA is bathed in water at a concentration of about 55 M. Water, which
we usually consider to be innocuous, can cause hydrolytic damage to
DNA bases. One example of such damage is the deamination of the base
cytosine (Figure 13). In this reaction, the exocyclic amino group (i.e., the
amino group that is not part of the ring) is replaced with a carbonyl group,
thereby converting cytosine to uracil, a base usually only found in RNA.
Like cytosine, which pairs with guanine, uracil is a pyrimidine but unlike
cytosine, uracil (like thymine) pairs with adenine. Thus, if left unrepaired,
the resulting uracil will pair with adenine during replication, and after a
further round of replication, that adenine will, in turn, pair with thymine,
resulting in the permanent replacement of the original G:C base pair with
an A:T base pair.
Cells have a repair system that detects uracil in DNA and replaces it
with cytosine. This repair mechanism is only possible because uracil is
not normally found in DNA. Indeed, the susceptibility of cytosine to
deamination provides an appealing explanation for why DNA has thymine
as a base rather than uracil. If uracil were normally found in DNA, then
the cell could not distinguish uracil that arose aberrantly from cytosine
by deamination from all the other uracils in the DNA. RNA, on the other
hand, does not generally serve as a genetic information carrier, and so
conversions of cytosine to uracil do not normally have permanent genetic
(mutational) consequences.
Chapter 9 DNA Replication 16

Figure 13 Cytosine deamination (A)

converts cytosine into uracil H H
N O
(A) As shown in the arrow-pushing cytosine
mechanism, water acts as a nucleophile, N deamination
NH
forming a covalent bond to carbon 4 (“C4”) C + H 2O U + NH3
of cytosine. In this same step, the double N O N O
bond between C4 and nitrogen 3 (“N3”)
breaks, and N3 becomes protonated. In the
second step, a second water molecule acts
as a base, abstracting a hydrogen atom from
the hydroxyl group attached to C4. The
electrons from the broken H-O bond form H H H H
N
a new O-C4 double bond, and the C4-NH2 O H O NH2 O
H H H H
bond is broken as a consequence. The -NH2 O N O
leaving group becomes protonated to form H H H H NH
ammonia (NH3). Cytosine is converted to N O N O
uracil by this reaction. (B) The absence of
the methyl group at pyrimidine position 5
allows repair enzymes to distinguish uracil
from thymine in DNA.
(B)
H
O
N H
N
H N T Uracil lacks the methyl
N A N N group found in thymine,
N O allowing it to be identified
by DNA repair enzymes.

H
O
N H
N
H N U
N A N N
N O

Box 1 The Polymerase Chain Reaction is a technique for amplifying specific DNA
sequences exponentially
An extremely powerful technique that is based on the concepts we have been considering in this chapter is
the Polymerase Chain Reaction or PCR. PCR makes it possible to amplify a desired stretch of DNA from
a much larger DNA molecule. The DNA is amplified in an exponential fashion, meaning that each cycle of
PCR, in principle, doubles the number of DNA molecules with a specific sequence present in a test tube. As
a result, minute quantities of DNA, as little as a single molecule, can be amplified into billions of molecules
in just a few hours using relatively simple equipment. PCR was invented by Kary Mullis, who earned the
1993 Nobel Prize in Chemistry for his achievement.
A PCR reaction contains one or more DNA template molecules, many copies of each of two short single-
stranded DNAs that will prime the reaction, each of the four dNTPs, and DNA polymerase. The sequences
of the two DNA primers are chosen to anneal to each of the template strands at locations that flank the
desired region to be amplified, with the 3’ ends of the primers pointing at each other.
Figure 14 illustrates the PCR process for the amplification of a segment of DNA highlighted in gray. Single-
stranded DNAs (purple and light blue) that are complementary to the ends of this region are used to prime
Chapter 9 DNA Replication 17

3’ 5’

Cycle 1
5’ 3’
3’ 5’
template +
DNA 5’ 3’
region to amplify primers
Heat to 95°C

5’ 3’
denatured 3’ 5’
template
+
3’ 5’ 5’ 3’

Cool to ~60°C

5’ 3’ 5’ 3’
3’ 5’ 3’ 5’

Incubate at ~68-72°C

5’ 3’ 5’ 3’
3’ 5’ 3’ 5’
overhang products

Heat to 95°C
Cycle 2

Cool to ~60°C
5’ 3’ 5’ 3’
3’ 5’ 3’ 5’
5’ 3’ 5’ 3’
3’ 5’ 3’ 5’

Incubate at ~68-72°C

5’ 3’ 5’ 3’
3’ 5’ desired 3’ 5’
5’ 3’ products 5’ 3’
3’ 5’ 3’ 5’

Heat to 95°C
Cycle 3 and above

Cool to ~60°C
Incubate at ~68-72°C

5’ 3’ 5’ 3’
3’ 5’ desired 3’ 5’
5’ 3’ products 5’ 3’
3’ 5’ 3’ 5’
5’ 3’ 3’ 5’
3’ 5’ 5’ 3’
desired
products

Figure 14 PCR uses specially designed primers to amplify a desired DNA sequence
Chapter 9 DNA Replication 18

the reaction. At the start of the first PCR cycle, the test tube containing all of these ingredients is heated to
95°C, which is close to the temperature at which water boils. At such a high temperature, the two strands
of the DNA template denature into single strands. The test tube is then cooled, typically to about 60°C
(the temperature of uncomfortably hot, but not scalding, water). At this intermediate temperature, the
DNA primers anneal to the dissociated template strands, forming two template-primer complexes. So far
no DNA synthesis has taken place. After the template-primer complexes form, the temperature is raised
to 68°C-72°C. DNA polymerase begins to extend each primer along its template, polymerizing a strand
complementary to the template in the 5’-to-3’ direction and consuming dNTPs in the test tube. When
both of the polymerization reactions are complete, the end result is that the desired region of the template
molecule has been copied, along with additional DNA past the region of interest. We will call these pieces of
extended DNA overhang products; they are longer than the segment of DNA we seek to amplify.
Next, the test tube is again heated to 95°C so that the double-stranded products from the first round of
polymerization denature into single strands. Now these amplified single strands can serve as the templates
for the next cycle of PCR. The test tube is once again cooled to about 60°C at which temperature the primers
anneal to the templates and DNA polymerization takes place once again. Notice what happens: when an
overhang product is used as a template for the next round of PCR, the primer anneals at the point where the
overhang begins. When polymerization stops at the end of the DNA, the result is only the desired product,
with no overhang. These correctly-sized products, as well as the overhang products, can be used as templates
again in the next cycle.
In each cycle, the undesired overhang products continue to be produced from the original template DNA,
but these increase only linearly with the number of cycles. Because every product can be used as a template
to make the desired product, the desired product is amplified exponentially, doubling in number during
each cycle. The net result of this process is that each cycle (template denaturation, primer annealing, and
DNA polymerization) of PCR can double the number of desired double-stranded DNA molecules in the
tube. After 26 PCR cycles, one single-stranded template molecule can be theoretically copied more than 30
million (226) times. At this point, the desired product is vastly more abundant than the unwanted overhang
products.

Summary
The Meselson-Stahl experiment showed that DNA is replicated by a semi-
conservative mechanism in which each of the two strands of the double
helix separately serve as templates for the synthesis of new daughter
strands. Using the heavy isotope of nitrogen 15N to density-label DNA, the
Meselson-Stahl experiment showed that after one round of replication in
the presence of 14N all the DNA had a hybrid density, as predicted for semi-
conservative replication.
DNA synthesis takes place in a 5’-to-3’ direction using as substrates the
four 2’-deoxynucleoside triphosphates (dNTPs). The 3’ hydroxyl group
of the growing chain acts as a nucleophile that attacks the α-phosphate of
the incoming dNTP, resulting in the formation of a phosphodiester bond
and the release of pyrophosphate as a leaving group. The complementary
template strand to which the growing daughter strand is annealed specifies
the identity of each nucleotide to be added by base pairing between the
incoming nucleotide and the corresponding base on the template strand.
The enzyme for DNA synthesis is DNA polymerase. The reaction that it
catalyzes, phosphodiester bond formation, is thermodynamically favorable
Chapter 9 DNA Replication 19
but only moderately so. Phosphodiester bond formation is, however,
coupled to a second moderately favorable process, the hydrolysis of the
pyrophosphate leaving group. Together, the two coupled reactions are quite
favorable. The coupling of phosphodiester bond formation to pyrophosphate
hydrolysis is an example of Le Châtelier’s principle.
Replication takes place at replication forks, which are moving sites on
DNA where the two strands are being separated so that each can serve as a
template for daughter strand synthesis. Replication forks are asymmetric in
that synthesis takes place continuously in the direction in which the fork is
moving (leading strand synthesis) and discontinuously on the other strand
(lagging strand synthesis). Lagging strand synthesis takes place in short
bursts that point away from the movement of the fork. The resulting short
segments of DNA are called Okazaki fragments and are stitched together by
DNA ligase to generate an uninterrupted daughter strand.
DNA polymerase synthesizes DNA accurately by requiring that each incoming
nucleotide pair with the corresponding base on the template. Incoming
nucleotides that do not properly pair are disfavored for phosphodiester
bond formation. Nucleotides that are nonetheless misincorporated are
removed by a proofreading mechanism in which a 3’ nuclease in the editing
pocket of DNA polymerase releases the misincorporated nucleotide from
the 3’ end of the growing chain, allowing the polymerase to incorporate
a fresh nucleotide. Misincorporated nucleotides that escape proofreading
are, in turn, removed by a mismatch repair system.
Chemicals, radiation, and hydrolytic damage also contribute to introducing
mutations in DNA. Dedicated enzyme systems scan the DNA for damage,
such as damage to pyrimidines caused by ultraviolet light, and repair the
lesions. Hydrolytic damage resulting from deamination of cytosine is
repaired by a system that detects and removes uracil, the product of the
deamination reaction. Cytosine deamination offers an explanation for why
DNA contains thymine instead of uracil, which is normally only found in
RNA; if DNA contained uracil, then the cell could not distinguish uracil
arising from deamination from uracil normally present in DNA.
The PCR reaction is a powerful technique for amplifying specific segments
of DNA exponentially from much larger DNA molecules by the use of short
DNA primers.