DNA Replication, Repair,and Recombination

The ability of a cell to maintain order in a chaotic environment dependson the accurate duplication of the vast quantity of genetic informationcarried in its DNA. This duplication process, called DNA replication,must occur before a cell can produce two genetically identical daughtercells. Maintaining order in a cell also requires the continual surveillanceand repair of its genetic information, as DNA is subject to damage bychemicals and radiation from the environment, and by accidents andreactive molecules that occur inside the cell. As we shall see in thischapter, each cell contains elaborate machinery for accurately copyingits store of genetic information, as well as specialized enzymes forrepairing DNA when it is damaged. These enzymes catalyze some of themost rapid and accurate processes that take place within cells, and theiractions reflect the elegance and efficiency of cellular chemistry.

Despite these systems for protecting the genetic instructions fromcopying errors and accidental damage, permanent changes, or muta-tions, sometimes do occur. Mutations in the DNA often affect the infor-mation it encodes. Occasionally, this can benefit the organism in whicha mutation occurs: for example, mutations can make bacteria resistantto antibiotics that are used to kill them. Indeed, the accumulation ofchanges in DNA over millions of years provides the variety in geneticmaterial that makes one species distinct from another, as we discuss inChapter 9. Mutations also produce the smaller variations that underliethe differences between individuals of the same species that we can eas-ily see in humans and other animals (Figure 6–1).

However, mutations are often detrimental: in humans, mutationsare responsible for thousands of inherited diseases, and mutations thatarise in the cells of the body throughout the lifetime of an individualmay also cause disease, most notably the many types of cancer. Thussurvival of a cell or organism can depend on preventing changes to itsDNA. Without the cellular systems that are continually monitoring andrepairing damage to DNA, it is questionable whether life could exist atall.

In this chapter, we begin by reviewing the cellular mechanisms—DNA replication and repair—that are responsible for keeping mutationsto a minimum. Finally, we consider some of the intriguing ways inwhich cells alter their genetic information, including DNA recombina-tion and the movement of the special DNA sequences in our chromo-somes called transposable elements.

DNA ReplicationBase-Pairing Enables DNA ReplicationDNA Synthesis Begins at Replication OriginsNew DNA Synthesis Occurs at Replication

ForksThe Replication Fork Is AsymmetricalDNA Polymerase Is Self-correctingShort Lengths of RNA Act as Primers for DNA

SynthesisProteins at a Replication Fork Cooperate to

Form a Replication MachineTelomerase Replicates the Ends of

Eucaryotic ChromosomesDNA Replication Is Relatively Well Understood

DNA RepairMutations Can Have Severe Consequences

for an OrganismA DNA Mismatch Repair System Removes

Replication Errors That Escape theReplication Machine

DNA Is Continually Suffering Damage in CellsThe Stability of Genes Depends on DNA

RepairThe High Fidelity of DNA Maintenance Allows

Closely Related Species to Have Proteinswith Very Similar Sequences

DNA Recombination Homologous Recombination Results in an

Exact Exchange of Genetic InformationRecombination Can Also Occur Between

Nonhomologous DNA SequencesMobile Genetic Elements Encode the

Components They Need for Movement A Large Fraction of the Human Genome Is

Composed of Two Families of TransposableSequences

Viruses Are Fully Mobile Genetic ElementsThat Can Escape from Cells

Retroviruses Reverse the Normal Flow ofGenetic Information

PRELIMINARY VERSION© 2003 Garland Science

DNA ReplicationAt each cell division, a cell must copy its genome with extraordinaryaccuracy. In this section, we explore how the cell achieves this precision,while duplicating DNA at rates as high as 1000 nucleotides per second.

Base-Pairing Enables DNA Replication

In the preceding chapter, we saw that each strand of the DNA doublehelix contains a sequence of nucleotides that is exactly complementaryto the nucleotide sequence of its partner strand. Each strand can there-fore act as a template, or mold, for the synthesis of a new complemen-tary strand (Figure 6–2). In other words, if we designate the two DNAstrands as S and S¢, strand S can serve as a template for making a newstrand S ¢, while strand S¢ can serve as a template for making a newstrand S (Figure 6–3). Thus, the genetic information in DNA can beaccurately copied by the beautifully simple process in which strand Sseparates from strand S ¢, and each separated strand then serves as atemplate for the production of a new complementary partner strandthat is identical to its former partner.

The ability of each strand of a DNA molecule to act as a template forproducing a complementary strand enables a cell to copy, or replicate,its genes before passing them on to its descendants. But the task is awe-inspiring, as it can involve copying billions of nucleotide pairs everytime a cell divides. The copying must be carried out with speed andaccuracy: in about 8 hours, a dividing animal cell will copy the equiva-

Chapter 6: DNA Replication, Repair, and Recombination6:2

Figure 6–1 Hereditary information is passed faithfully from onegeneration to the next. Changes in the DNA, however, can producethe variations that underlie the differences between individuals of thesame species—or, over time, the differences between one species andanother. In this family photo, the children resemble one another andtheir parents more closely than they resemble other people becausethey inherit their particular genes from their parents. The cat sharesmany features with humans, but during the millions of years ofevolution that have separated humans and cats, we both haveaccumulated many hereditary changes that now make us quitedifferent species. The chicken is an even more distant relative.

Figure 6–2 A DNA strand can serve as atemplate. Preferential binding occursbetween pairs of nucleotides (A with T, andG with C) that can form base pairs. Thisenables each strand to act as a templatefor forming its complementary strand.

Figure 6–3 DNA acts as a template for its own duplication. Because thenucleotide A will successfully pair only with T, and G with C, each strand of DNAcan serve as a template to specify the sequence of nucleotides in its complementarystrand. In this way, double-helical DNA can be copied precisely. Keep in mind thatalthough they are colored differently here, the template strands (orange) and thenew strands (red) are chemically identical.

3¢

3¢5¢

5¢

parent DNA double helix

template S strand

template S¢ strand

new S¢ strand

new S strand

5¢ 3¢

3¢ 5¢

5¢ 3¢

3¢ 5¢

5¢ 3¢

3¢ 5¢

S strand

S¢ strand

lent of 1000 books like this one and, on average, get no more than a sin-gle letter or two wrong. This feat is performed by a cluster of proteinsthat together form a “replication machine.” DNA replication producestwo complete double helices from the original DNA molecule, each newDNA helix identical (except for rare errors) in nucleotide sequence tothe parental DNA double helix (see Figure 6–3). Because each parentalstrand serves as the template for one new strand, each of the daughterDNA double helices ends up with one of the original (old) strands plusone strand that is completely new; this style of replication is said to besemiconservative (Figure 6–4).

DNA Synthesis Begins at Replication Origins

The DNA double helix is normally very stable: the two DNA strands arelocked together firmly by the large numbers of hydrogen bondsbetween the bases on both strands (see Figure 5–2). As a result, onlytemperatures approaching those of boiling water provide enough ther-mal energy to separate these strands. In order to be used as a template,however, the double helix must first be opened up and the two strandsseparated to expose unpaired bases. How does this occur at the tem-peratures found in living cells?

The process of DNA replication is begun by initiator proteins thatbind to the DNA and pry the two strands apart, breaking the hydrogenbonds between the bases (Figure 6–5). Although the hydrogen bondscollectively make the DNA helix very stable, individually each hydrogenbond is weak (Chapter 2). Separating a short length of DNA does nottherefore require a large energy input and can occur with the assistanceof these proteins at normal temperatures.

The positions at which the DNA is first opened are called replica-tion origins, and they are marked by a particular sequence ofnucleotides. In simple cells like those of bacteria or yeast, replicationorigins span approximately 100 base pairs; they are composed of DNAsequences that attract the initiator proteins, as well as stretches of DNAthat are especially easy to open. We saw in Chapter 5 that an A-T basepair is held together by fewer hydrogen bonds than is a G-C base pair.Therefore, DNA rich in A-T base pairs is relatively easy to pull apart, andA-T–rich stretches of DNA are typically found at replication origins.

A bacterial genome, which is typically contained in a circular DNAmolecule of several million nucleotide pairs, has a single origin of repli-cation. The human genome, which is very much larger, has approxi-mately 10,000 such origins. In humans, beginning DNA replication atmany places at once allows a cell to replicate its DNA relatively quickly.In How We Know, pp. –, we discuss experiments that reveal the loca-tions of the replication origins in an organism’s genome.

Once an initiator protein binds to DNA at the replication origin andlocally opens up the double helix, it attracts a group of proteins thatcarry out DNA replication. This group operates as a protein machine,with each member carrying out a specific function. We will introduceeach of these proteins shortly, after we consider the overall process ofDNA replication.

6:3DNA Replication

Figure 6–4 In each round of replication, each of the two strands ofDNA is used as a template for the formation of a complementaryDNA strand. The original strands, therefore, remain intact throughmany cell generations. DNA replication is “semiconservative” becauseeach daughter DNA double helix is composed of one conserved strandand one newly synthesized strand.

REPLICATION

REPLICATION

REPLICATION

Figure 6–5 A DNA double helix is openedat its replication origin. Replicationinitiator proteins recognize sequences ofDNA at replication origins and locally pryapart the two strands of the double helix.The exposed single strands can then serveas templates for copying the DNA.

replication origindouble-helicalDNA

double helix opened with the aid ofinitiator proteins

single-stranded DNA templatesready for DNA synthesis

5¢3¢

3¢5¢

5¢3¢

3¢5¢

Chapter 6: DNA Replication, Repair, and Recombination6:4

How We Know: Finding Replication Origins

For eucaryotic cells, DNA replication is a monumental task.To transmit its genetic information to its daughters, a cellmust copy its entire genome quickly, carefully, and com-pletely. And it must copy its genetic material once and onlyonce during each round of cell division. Failure to performthese tasks properly—essentially perfectly—can have cata-strophic consequences. If the whole genome is not repli-cated, critical information may be lost; copying some infor-mation more than once can be equally disastrous. Both sit-uations can cause mutations, duplications, rearrangements,even massive chromosomal breakage—defects that in amulticellular organism can cause cancer or other diseases.

But how do cells control this complicated molecular maneu-ver? We know quite a bit about the control of replication inprocaryotes, where the situation is simpler: bacterialgenomes have a single origin of replication, so all a bacter-ial cell has to do is start replication and make sure that it fin-ishes.

Eucaryotic genomes contain many origins of replication—hundreds in simple yeasts, thousands in humans—and notall of these origins are equal. Some are used faithfully ineach round of replication; others are used periodically, occa-sionally, or even rarely. Furthermore, different origins arecalled into service at different times during S phase, thestage in the cell cycle when DNA is replicated. Some “fire”early in S phase, others launch replication later. Their activ-ity depends, at least in part, on where in the genome theyare located.

With so many replication origins all over the genome, firingat different times, it is hard to imagine how a cell can coor-dinate the process with such extraordinary accuracy. Currentstudies aimed at identifying and locating replication origins,and determining when they are used, are beginning toreveal some answers as to how origins are chosen, how thetiming of their activity is determined, and how the wholeprocess is controlled.

Origins at workUnlike the situation in procaryotes, origins of replication ineucaryotes do not contain an easily identifiable DNAsequence. Most replication origins in the yeast genome con-tain variations of an 11-nucleotide sequence, but not allDNA fragments with this sequence can act as origins. Inother eucaryotes, almost any sequence seems capable ofinitiating replication. So how can researchers identify an ori-gin?

The simplest approach to determining whether a piece ofDNA contains an origin is to see whether the fragment candirect the replication of a plasmid. Plasmids are small loops

of DNA that can exist inside a yeast or bacterial cell, sepa-rate from the cell’s own genome. These circular DNA mole-cules have their own replication origins, which allow themto replicate independently of the host cell’s genome; they aredescribed in greater detail in Chapter 10.

Using recombinant DNA techniques (which we also discussin Chapter 10), an investigator can remove a plasmid’s ori-gin of replication and replace it with a DNA fragment thatmight contain a yeast replication origin. The modified plas-mids are then introduced into a yeast cell and the cells areincubated to allow replication to occur: replication of theplasmid can only occur if its inserted DNA fragment con-tains an origin of replication. In this way, yeast origins ofreplication can be identified (Figure 6–6).

Figure 6–6 Segments of DNA can be analyzed for their potentialto serve as a replication origin by inserting them into a plasmid.Random DNA fragments are inserted into a plasmid whose origin ofreplication has been removed. This plasmid, which contains a genethat allows yeast to grow in the absence of histidine, is thenintroduced into yeast cells, which are grown in medium lackinghistidine. To survive and divide in this medium, yeast cells mustcontain a replicating plasmid. Thus any cells that grow intocolonies will contain a plasmid that bears a DNA fragment that iscapable of driving replication. If the fragment does not supportplasmid replication, the cells will not survive.

INTRODUCTION OF PLASMID DNA INTO YEAST CELLS THAT LACK THE HIS GENE AND THEREFORE CANNOT GROW IN THE

ABSENCE OF HISTIDINE

HIS HIS HIS

randomly selected yeast DNA fragments

selective medium without histidine

DNA fragment (green )lacks an originof replication

DNA fragment (red )lacks an originof replication

DNA fragment (blue )contains an origin

of replication

ATAT

CTGAGCCC

AAGG

AAGG

ATAT GCCCAAGG CTGA

AA GG

AA G G

ATAT GCCC CTGA

ATAT

TATA

TATA

CTGA

GACT

GACT

GCCC

CGGG

CGGG

AAGG

AAGG

TTCC

TTCC

allow DNA replicationto begin at origin

break DNA into fragments,separate DNA strands, andfluorescently label

add labeled fragments toDNA microarray

allow complementary base-pairing to occur

examine microarraywith fluorescent scanner

spot containing replicated DNA appears twice as bright as other spots on the microarray

6:5DNA Replication

Arrays to the rescue

The plasmid maintenance method is fine when one startswith a piece of DNA—or a small collection of DNA frag-ments—that potentially contain a replication origin. Butwhat if a researcher has no idea where the origins lie, orwishes to identify and study all of the origins in a eucaryotesuch as yeast? Now some powerful new techniques areallowing investigators to locate the replication origins acrossa whole genome—and determine when they fire—in one fellswoop.

To map the positions and timing of all the replication originsin the yeast genome, researchers have turned to DNAmicroarrays—grids studded with thousands of DNA frag-ments of known sequence. As we will see in Chapter 10,these microarrays are used to determine whether nucleotidefragments of a particular sequence are present in a sample.They are widely used, for example, to monitor which genesare being expressed in a cell. Because the exact sequence—and position—of every DNA fragment on the microarray isknown, the exact sequence of any nucleotide fragment thatbinds to its complementary sequence in the array can beascertained simply by determining to which position on thearray it is bound.

For the replication origin experiment, researchers use amicroarray that contains DNA fragments that cover the entireyeast genome—tens of thousands of probes that representunique sequences found at consecutive intervals of, say,500 nucleotides along the genome. The yeast are allowed tobegin replication and their DNA is collected, broken intofragments, and labeled with a fluorescent marker (Figure6–7). These fragments are washed over the whole-genomemicroarray and allowed to bind to their complementarysequences fixed to the surface of the array. Researchers canthen locate the replication origins by seeing which DNA seg-ments are duplicated first. As shown in Figure 6–7, DNA seg-ments that have been replicated will be present at twice theconcentration of nonreplicated segments, generating a fluo-rescent spot with twice the intensity on the DNA microarray.

Ready, set, replicateThe situation outlined in Figure 6–7 is an overly simple one:it shows replication beginning at a single origin and pro-ceeding for only a short distance. In reality, replication in

Figure 6–7 DNA microarrays can be used to locate DNA replication origins. A culture of yeast cells is allowed to beginreplication, and the partially replicated DNA is collected. This DNA is broken into fragments, separated into single strands, andthen labeled with a fluorescent marker. The labeled fragments are washed over the microarray grid and are allowed to bind tocomplementary fragments on the array. When this microarray is examined under a fluorescence scanner, the positions to whichlabeled DNA has bound will light up. The more DNA bound to a spot, the brighter that spot will appear. Spots representingreplicated DNA will appear twice as bright, as they will bind twice as much DNA as spots representing nonreplicated DNA. Shownhere is a grid spotted with four different DNA fragments only four nucleotides in length. Actual microarrays contain thousands ofDNA fragments, typically tens to hundreds of nucleotides in length. For simplicity, each spot on the sample microarray shows onlytwo DNA fragments bound to it; in reality, each spot on a microarray contains millions of DNA fragments of identical sequence.

Chapter 6: DNA Replication, Repair, and Recombination6:6

yeast cells is initiated at hundreds of origins that fire at dif-ferent times. Fortunately, the same experimental protocolcan be adapted to determine which regions of DNA repli-cate first, which start later, and how quickly replicationforks sweep through different segments of the genome.

To follow replication over time, researchers collect DNAfrom yeast cells at different times after the start of replica-tion (Figure 6–8). They begin by synchronizing the yeastcells so that each cell in the population will begin to repli-cate its DNA at exactly the same time. This can be done byexposing the yeasts to a molecule that will cause them toarrest in G1, the phase of the cell cycle in which the repli-cation machinery assembles on the origins but in whichDNA synthesis has yet to begin. When the drug isremoved, the yeast cells will all enter S phase.

DNA samples are collected, fluorescently labeled, andwashed over the microarray as before. Now, however,researchers can monitor the progression of replicationthroughout the genome. Using such microarray-basedtechniques, researchers have identified some 300 originsof replication in the yeast genome; some of these had pre-viously been identified using the plasmid maintenancemethods described above, but many are new. They alsofound that replication origins are activated throughout Sphase, with most firings occurring near mid-S phase.

Although the experiment was performed with yeast DNA,this method can be adapted to look for replication originsin human DNA. Such studies should enhance our under-standing of how all cells achieve such exquisite controlover a process as critical and complex as DNA replication.

Figure 6–8 Collecting DNA at different times after replication begins allows an investigator to monitor the progress ofreplication through the genome. Cells are synchronized so that they begin replication at the same time. DNA is collected andapplied to the microarray as shown in Figure 6–7. Replication begins at an origin and proceeds, bidirectionally, until the entiregenome has been copied. For simplicity only one origin is shown here. In yeast cells, replication begins at hundreds of originslocated throughout the genome. The spots on these microarrays represent consecutive sequences along the yeast genome. Only81 spots are shown here, but the actual arrays contain tens of thousands of sequences that span the entire yeast genome.Because the sequence of the DNA at each spot on the microarray is known, the location of every replication origin can bedetermined by monitoring which spots light up first.

0 min 5 min 10 min 20 min

allowreplicationto begin

fragment DNA separate strands fluorescently label

NO REPLICATION REPLICATIONBEGINS AT

ORIGIN

REPLICATIONCONTINUES

DNA FULLYREPLICATED

culture of yeast cellsarrested before DNA

replication begins

New DNA Synthesis Occurs at Replication Forks

DNA molecules in the process of being replicated can be observed inthe electron microscope (Figure 6–9), where it is possible to see Y-shaped junctions in the DNA, called replication forks. At these forks,the replication machine is moving along the DNA, opening up the twostrands of the double helix and using each strand as a template to makea new daughter strand. Two replication forks are formed starting fromeach replication origin, and they move away from the origin in bothdirections, unzipping the DNA as they go. DNA replication in bacterialand eucaryotic chromosomes is therefore termed bidirectional. Theforks move very rapidly—at about 1000 nucleotide pairs per second inbacteria and 100 nucleotide pairs per second in humans. The slowerrate of fork movement in humans (indeed, in all eucaryotes) may be dueto the difficulties in replicating through the more complex chromatinstructure found in these higher organisms.

At the heart of the replication machine is an enzyme called DNApolymerase, which synthesizes the new DNA using one of the oldstrands as a template. This enzyme catalyzes the addition of nucleotidesto the 3¢ end of a growing DNA strand by the formation of a phosphodi-ester bond between this end and the 5¢-phosphate group of the incom-ing nucleotide (Figure 6–10). The nucleotides enter the reaction initiallyas energy-rich nucleoside triphosphates, which provide the energy forthe polymerization reaction. The hydrolysis of one phosphoanhydridebond in the nucleoside triphosphate provides the energy for the con-densation reaction that links the nucleotide monomer to the chain andreleases pyrophosphate (PPi). The DNA polymerase couples the releaseof this energy to the polymerization reaction. Pyrophosphate is furtherhydrolyzed to inorganic phosphate (Pi), which makes the polymeriza-tion reaction effectively irreversible (see Figure 3–).

DNA polymerase does not dissociate from the DNA each time itadds a new nucleotide to the growing chain; rather, it stays associatedwith the DNA and moves along it stepwise for many cycles of the poly-merization reaction. We will see, later in this chapter, how a special pro-tein keeps the polymerase attached in this way.

6:7DNA Replication

0.1 �m

origins of replication

direction offork movement

1

2

3

replication forks

Figure 6–9 Replication forks move awayin both directions from multiplereplication origins in a eucaryoticchromosome. The electron micrographshows DNA replicating in the early embryoof a fly. The particles visible along the DNAare nucleosomes, protein complexespresent in eucaryotic chromosomes aroundwhich the DNA is wrapped. (1), (2), and(3) are drawings of the same portion of aDNA molecule as it might appear atsuccessive stages of replication, drawnfrom electron micrographs. (2) is drawnfrom the electron micrograph shown here.The yellow lines represent the parentalDNA strands; the solid red lines representthe newly synthesized DNA. (Electronmicrograph courtesy of Victoria Foe.)

Question 6–1

Look carefully at themicrograph in Figure6–9.

A. Using the scale bar,estimate thelengths of the DNA strandsbetween the replication forks.Numbering the replication forkssequentially from the left, howlong will it take until forks 4 and5, and forks 6 and 7, respectively,collide with each other? Thedistance between the bases inDNA is 0.34 nm, and eucaryoticreplication forks move at about100 nucleotides per second. Forthis question disregard thenucleosomes seen in themicrograph and assume that theDNA is fully extended.

B. The fly genome is about 1.8 ¥ 108

nucleotide pairs in size. Howmuch of the total fly DNA isshown in the micrograph?

The Replication Fork Is Asymmetrical

The 5¢-to-3¢ direction of the DNA polymerization mechanism poses aproblem at the replication fork. We saw in Figure 5– that thesugar–phosphate backbone of each strand of a DNA double helix has aunique chemical direction, or polarity, determined by the way eachsugar residue is linked to the next, and that the two strands in the dou-ble helix run in opposite orientations. As a consequence, at the replica-tion fork, one new DNA strand is being made on a template that runs inone direction (3¢ to 5¢), whereas the other new strand is being made ona template that runs in the opposite direction (5¢ to 3¢) (Figure 6–11).The replication fork is therefore asymmetrical. Both the new DNAstrands appear to be growing in the same direction, that is, the directionin which the replication fork is moving. On the face of it this suggeststhat one strand is being synthesized in the 3¢-to-5¢ direction and one isbeing synthesized in the 5¢-to-3¢ direction.

DNA polymerase, however, can catalyze the growth of the DNAchain in only one direction; it can add new subunits only to the 3¢ endof the chain (see Figure 6–10). As a result, a new DNA chain can be syn-thesized only in a 5¢-to-3¢ direction. This can easily account for the syn-thesis of one of the two strands of DNA at the replication fork, but notthe other. One might have expected a second DNA polymerase to syn-thesize the other DNA strand—one that works by adding subunits to the5¢ end of a DNA chain. However, no such enzyme exists. Instead, the

Chapter 6: DNA Replication, Repair, and Recombination6:8

Figure 6–10 DNA is synthesized in the5¢-to-3¢ direction. Addition of adeoxyribonucleotide to the 3¢-hydroxyl endof a polynucleotide chain is thefundamental reaction by which DNA issynthesized; the new DNA chain istherefore synthesized in the 5¢-to-3¢

direction. Base-pairing between theincoming deoxyribonucleotide and thetemplate strand guides the formation of anew strand of DNA that is complementaryin nucleotide sequence to the templatechain (see Figure 6–2). The enzyme DNApolymerase catalyzes the addition ofnucleotides to the free 3¢ hydroxyl on thegrowing DNA strand. The nucleotides enterthe reaction as nucleoside triphosphates.Breakage of a phosphoanhydride bond(indicated by the asterisk) in the incomingnucleoside triphosphate releases a largeamount of free energy and thus providesthe energy for the polymerization reaction.

Figure 6–11 At a replication fork, the twonewly synthesized DNA strands are ofopposite polarity.

newly synthesizedstrands

5¢5¢3¢

5¢3¢

5¢3¢

3¢

parentalDNA helix

direction of replication-fork movement

O

P O

OOH2C

O

P_O O

OOH2C

OCH2

O

OP

O

OP

O

O

O_

P_O

OH

P_

O

O

OO

O

CH2

P_

O

O

OO

O

CH2

P_

O

O

OO

O

CH2

P_

O

O

OO

O

CH2

OO

O

P_

O

O

O

CH2

C G

A T

C G

A

T

incoming deoxyribonucleoside triphosphate

primerstrand template

strand

O_

O_

5¢ end of strand

5¢ end of strand

3¢ end of strand

3¢ end of strand

OHpyrophosphate

_O

problem is solved by the use of a “backstitching” maneuver. The DNAstrand whose 5¢ end must grow is made discontinuously, in successiveseparate small pieces, with the DNA polymerase working backwardfrom the replication fork in the 5¢-to-3¢ direction for each new piece.These pieces—called Okazaki fragments after the biochemist who dis-covered them—are later “stitched” together to form a continuous newstrand (Figure 6–12). The DNA strand that is synthesized discontinu-ously in this way is called the lagging strand; the strand that is synthe-sized continuously is called the leading strand.

Although they differ in subtle details, the replication forks of allcells, procaryotic and eucaryotic, have leading and lagging strands. Thecommon feature arises from the fact that all of the DNA polymerasesused to replicate DNA polymerize in the 5¢-to-3¢ direction only. We shalllook at events on the lagging strand in more detail later in this chapter;first, we consider another feature of DNA polymerase that is common toall cells.

DNA Polymerase Is Self-correcting

DNA polymerase is so accurate that it makes only about one error inevery 107 nucleotide pairs it copies. This error rate is much lower thancan be accounted for simply by the accuracy of complementary base-pairing. Although A-T and C-G are by far the most stable base pairs,other, less stable base pairs—for example, G-T and C-A—can also beformed. Such incorrect base pairs are formed much less frequently thancorrect ones, but they occur often enough that they would kill the cellthrough an accumulation of mistakes in the DNA if they were allowed toremain. This catastrophe is avoided because DNA polymerase can cor-rect its mistakes. As well as catalyzing the polymerization reaction, DNApolymerase has an error-correcting activity called proofreading. Beforethe enzyme adds a nucleotide to a growing DNA chain, it checkswhether the previous nucleotide added is correctly base-paired to thetemplate strand. If so, the polymerase adds the next nucleotide; if not,the polymerase removes the mispaired nucleotide by cutting the phos-phodiester bond it has just made, releases the nucleotide, and triesagain (Figure 6–13). Thus, DNA polymerase possesses both a 5¢-to-3¢

6:9DNA Replication

Figure 6–12 DNA replication forks areasymmetrical. Because both of the newstrands are synthesized in the 5¢-to-3¢

direction, the lagging strand of DNA mustbe made initially as a series of short DNAstrands that are later joined together. Theupper diagram shows two replication forksmoving in opposite directions; the lowerdiagram shows the same forks a short timelater. To synthesize the lagging strand, DNApolymerase “backstitches”: it mustsynthesize short fragments (called Okazakifragments) in the 5¢-to-3¢ direction, andthen move in the opposite direction alongthe template strand (toward the fork) beforeit synthesizes the next fragment.

5¢

5¢

3¢

3¢

5¢

5¢

3¢

3¢

5¢

3¢5¢

5¢ 5¢3¢ 3¢

3¢

direction of fork movement

leading-strand templateof left-hand fork

lagging-strand templateof right-hand fork

lagging-strand templateof left-hand fork leading-strand template

of right-hand fork

most recentlysynthesized DNA

POLYMERASE ADDS ANINCORRECT NUCLEOTIDE

5¢

5¢3¢

3¢

templateDNA strand

MISPAIRED NUCLEOTIDEREMOVED BY 3¢ 5¢PROOFREADING

5¢

5¢3¢

3¢

CORRECTLY PAIRED 3¢ ENDALLOWS ADDITION OFNEXT NUCLEOTIDE

5¢

5¢3¢

3¢

SYNTHESIS CONTINUES INTHE 5¢ 3¢ DIRECTION

5¢

5¢3¢

3¢

DNA polymerase

Figure 6–13 During DNA synthesis, DNApolymerase proofreads its own work. If anincorrect nucleotide is added to a growingstrand, the DNA polymerase will cleave itfrom the strand and replace it with thecorrect nucleotide before continuing.

polymerization activity and a 3¢-to-5¢ exonuclease (nucleic acid–degrad-ing) activity. These activities are carried out by different domains withinthe polymerase molecule (Figure 6–14).

This proofreading mechanism explains why DNA polymerases syn-thesize DNA only in the 5¢-to-3¢ direction, despite the need this imposesfor a cumbersome backstitching mechanism at the replication fork. Asshown in Figure 6–15, a hypothetical DNA polymerase that synthesizedin the 3¢-to-5¢ direction (and would thereby circumvent the need forbackstitching) would be unable to proofread: if it removed an incor-rectly paired nucleotide, the polymerase would create a chain end thatis chemically dead, in the sense that it would no longer be able to elon-gate. Thus, for a DNA polymerase to function as a self-correctingenzyme that removes its own polymerization errors as it moves alongthe DNA, it must proceed only in the 5¢-to-3¢ direction.

Short Lengths of RNA Act as Primers for DNA Synthesis

We have seen that the accuracy of DNA replication depends on therequirement of the DNA polymerase for a correctly base-paired endbefore it can add more nucleotides. But since the polymerase can join anucleotide only to a base-paired nucleotide in a DNA double helix, itcannot start a completely new DNA strand. A different enzyme isneeded to begin a new DNA strand, an enzyme that can begin a newpolynucleotide chain simply by joining two nucleotides together with-out the need for a base-paired end. This enzyme does not, however, syn-thesize DNA. It makes a short length of a closely related type of nucleicacid—RNA (ribonucleic acid)—using the DNA strand as a template.This short length of RNA, around 10 nucleotides long, is base-paired tothe template strand and provides a base-paired 3¢ end as a starting pointfor DNA polymerase. It thus serves as a primer for DNA synthesis, andthe enzyme that synthesizes the RNA primer is known as primase. Astrand of RNA is very similar chemically to a single strand of DNA exceptthat it is made of ribonucleotide subunits, in which the sugar is ribose,not deoxyribose; RNA also differs from DNA in that it contains the baseuracil (U) instead of thymine (T) (see Panel 2–6, pp. –). However,since U can form a base pair with A, the RNA primer is synthesized onthe DNA strand by complementary base-pairing in exactly the sameway as is DNA.

For the leading strand, an RNA primer is needed only to start repli-cation at a replication origin; once a replication fork has been estab-

Figure 6–14 DNA polymerase containsseparate sites for DNA synthesis andediting. (A) The structure of an E. coli DNApolymerase molecule, as determined by X-ray crystallography. Roughly speaking,the enzyme resembles a right hand inwhich the palm, fingers, and thumb graspthe DNA. (B) A cutaway outline of thestructures of DNA polymerase complexedwith the DNA template in the polymerizingmode (left) and the editing mode (right).The catalytic site for the error-correctingexonuclease activity (E) and thepolymerization activity (P) are indicated. Todetermine these structures by X-raycrystallography, researchers “froze” thepolymerases in these two states, by using amutant polymerase defective in theexonuclease domain (right) or bywithholding the Mg2+ required forpolymerization (left). These drawingsillustrate a DNA polymerase that functionsduring DNA repair, but the enzymes thatreplicate DNA have similar features.(A, adapted from L.S. Beese, V. Derbyshire,and T.A. Steitz, Science 260:352–355,1993.)

Chapter 6: DNA Replication, Repair, and Recombination6:10

incoming deoxyribonucleosidetriphosphate

gap inhelix

“fingers“

“palm“

primerstrand

templatestrand

“thumb“

3¢

5¢

3¢

5¢

(A) (B)

E

P

E

POLYMERIZING EDITING

P

5¢ 5¢

primer strand

5¢3¢5¢

3¢

3¢

3¢

templatestrand

6:11DNA Replication

lished, the DNA polymerase is continuously presented with a base-paired 3¢ end as it tracks along the template strand. But on the laggingstrand, where DNA synthesis is discontinuous, new primers are neededcontinually, as one can see from Figure 6–12. As the movement of thereplication fork exposes a new stretch of unpaired bases, a new RNAprimer is made at intervals along the lagging strand. DNA polymeraseadds a deoxyribonucleotide to the 3¢ end of this primer to start a DNAstrand, and it will continue to elongate this strand until it runs into thenext RNA primer (Figure 6–16).

To produce a continuous new DNA strand from the many separatepieces of RNA and DNA made on the lagging strand, three additionalenzymes are needed. These act quickly to remove the RNA primer,replace it with DNA, and join the DNA fragments together: a nucleasebreaks apart the RNA primer, a DNA polymerase called a repair poly-merase replaces the RNA with DNA (using the adjacent Okazaki frag-

Figure 6–15 Proofreading explains why DNA chains are synthesized only in the 5¢ to 3¢

direction. (A) Proofreading in the hypothetical 3¢-to-5¢ polymerization scheme would allow theremoval of an incorrect nucleotide (dark green), but would block addition of the correct nucleotide(red) and thereby prevent further chain elongation. (B) Growth in the 5¢-to-3¢ direction allows thechain to continue to be elongated when an incorrect nucleotide has been added and then removedby proofreading (see Figure 6–13).

P P P P P P P P

P

P P P P P P

P

PROOFREADING PROOFREADING

5¢ 5¢ 3¢

P P P P P P P

PPP

PP

PPP

PP

PPP

PP

PPP

PP

PP

PPP

5¢ 5¢ 3¢

5¢ 5¢ 3¢

P P P P P P

5¢ 5¢ 3¢

3¢

3¢

3¢

3¢

HYPOTHETICAL3¢-TO-5¢

STRAND GROWTH

ACTUAL5¢-TO-3¢

STRAND GROWTH

5¢ end producedif one nucleotide

is removed byproofreading

3¢ end producedwhen one nucleotide

is removed byproofreading

incoming correctdeoxyribonucleoside

triphosphate

incoming correctdeoxyribonucleoside

triphosphate

REACTION DOES NOT PROCEED,AS NO HIGH-ENERGY BOND

WOULD BE CLEAVED

HIGH-ENERGY BOND ISCLEAVED, PROVIDING THE

ENERGY FOR POLYMERIZATION

(A) (B)

Chapter 6: DNA Replication, Repair, and Recombination6:12

ment as a primer), and the enzyme DNA ligase joins the 5¢-phosphateend of one new DNA fragment to the 3¢-hydroxyl end of the next (seeFigure 6–16). ATP or NADH is required for ligase activity. We will discussthese three enzymes in more detail in the section on DNA repair later inthis chapter.

Primase can begin new polynucleotide chains, but this is possiblebecause it does not proofread its work. As a result, primers contain ahigh frequency of mistakes. But since they are made of RNA instead ofDNA, the primers stand out as “suspect copy” to be automaticallyremoved and replaced by DNA. This DNA is put in by DNA repair poly-merases, which, like the replicative polymerases, proofread as they syn-thesize. In this way, the cell’s replication machinery is able to begin newDNA chains and, at the same time, ensure that all of the DNA is copiedfaithfully.

Proteins at a Replication Fork Cooperate to Form aReplication Machine

As mentioned earlier, DNA replication requires a variety of proteins thatact in concert with DNA polymerase. Here, we will discuss the addi-tional proteins that, together with DNA polymerase and primase, formthe protein machine that powers the replication fork forward and syn-thesizes new DNA behind it. Although it would make good sense for thethree proteins that replace RNA primers with DNA—nuclease, repairpolymerase, and ligase—to also be a part of the replication machine, itis not yet known whether this is the case.

At the head of the replication machine is a helicase, a protein thatuses the energy of ATP hydrolysis to speed along DNA, opening the dou-ble helix as it moves (Figure 6–17). We saw earlier in this chapter that theDNA double helix must be opened to begin DNA replication, and itmust also be opened continuously as the replication fork progresses, inorder to provide exposed templates for the polymerase. Another com-ponent of the replication machine—single-strand binding protein—clings to the single-stranded DNA exposed by the helicase and tran-siently prevents it from re-forming base pairs. Yet another protein,called a sliding clamp, keeps the DNA polymerase firmly attached to theDNA template; on the lagging strand, the sliding clamp releases thepolymerase from the DNA each time an Okazaki fragment is completed.This clamp protein forms a ring around the DNA helix and binds poly-merase, allowing it to slide along a template strand as it synthesizes newDNA (see Figure 6–17).

Figure 6–16 On the lagging strand, DNA is synthesized infragments. In eucaryotes, RNA primers are made at intervals of about200 nucleotides on the lagging strand, and each RNA primer isapproximately 10 nucleotides long. In the bacterium E. coli, theprimers and Okazaki fragments are about 5 and 1000 nucleotideslong, respectively. Primers are erased by nucleases that recognize anRNA strand in an RNA/DNA helix and degrades it; this leaves gapsthat are filled in by a DNA repair polymerase that can proofread as itfills in the gaps. The completed fragments are finally joined togetherby an enzyme called DNA ligase, which catalyzes the formation of aphosphodiester bond between the 3¢-OH end of one fragment and the5¢-P end of the next, thus linking up the sugar–phosphate backbones.

3¢5¢

5¢

RNAprimer

lagging-strandtemplate

3¢ 5¢3¢

3¢5¢

5¢3¢ 5¢3¢

3¢5¢

5¢3¢

3¢5¢

5¢3¢

3¢5¢

5¢3¢

new RNA primersynthesis by DNAprimase

DNA polymerase adds to newRNA primer to start newOkazaki fragment

DNA polymerase finishesDNA fragment

old RNA primer erasedand replaced by DNA

nick sealing by DNA ligasejoins new Okazaki fragmentto the growing chain

Question 6–2

Discuss the followingstatement: “Primase is asloppy enzyme thatmakes many mistakes.Eventually, the RNAprimers it makes are

disposed of and replaced with DNAby a polymerase with higher fidelity.This is wasteful. It would be moreenergy-efficient if a DNA polymerasemade an accurate copy in the firstplace.”

6:13

Most of the proteins involved in DNA replication are thought to beheld together in a large multienzyme complex that moves as a unitalong the DNA, enabling DNA to be synthesized on both strands in acoordinated manner. This complex can be likened to a tiny sewingmachine composed of protein parts and powered by nucleosidetriphosphate hydrolysis. Although the structures of the individual pro-tein components of the replication machine have been determined,how these components fit together is not known in detail. Some ideasabout the general appearance of the complex, however, have been pro-posed (Figure 6–17B).

Telomerase Replicates the Ends of EucaryoticChromosomes

Having discussed how DNA replication begins at origins and how move-ment of the replication fork proceeds, we now turn to the special prob-lem of replicating the very ends of eucaryotic chromosomes. As we dis-cussed previously, the fact that DNA is synthesized only in the 5¢-to-3¢direction means that the lagging strand of the replication fork is synthe-sized in the form of discontinuous DNA fragments, each of which is

DNA Replication

Figure 6–17 A group of proteins acttogether at a replication fork. (A) Twomolecules of DNA polymerase are shown,one on the leading strand and one on thelagging strand. Both are held on to theDNA by a circular protein clamp thatallows the polymerase to slide. DNAhelicase uses the energy of ATP hydrolysisto propel itself forward and therebyseparate the strands of the parental DNAdouble helix ahead of the polymerase.Single-stranded DNA-binding proteinsmaintain these separated strands as single-stranded DNA to provide access for theprimase and polymerase. For simplicity,this figure shows the proteins workingindependently; in the cell they are heldtogether in a large replication machine, asshown in (B). (B) This diagram shows acurrent view of how the replication proteinsare arranged at the replication fork whenthe fork is moving. The structure in (A) hasbeen altered by folding the DNA on thelagging strand to bring the lagging-strandDNA polymerase molecule in contact withthe leading-strand DNA polymerasemolecule. This folding process also bringsthe 3¢ end of each completed Okazakifragment close to the start site for the nextOkazaki fragment. Because the lagging-strand DNA polymerase molecule is held tothe rest of the replication proteins, it canbe reused to synthesize successive Okazakifragments; in this diagram, it is about to letgo of its completed DNA fragment andmove to the RNA primer that will besynthesized nearby, as required to start thenext DNA fragment on the lagging strand.

newly synthesizedstrand

leading-strand template

DNA polymerase onleading strand

sliding clamp

parentalDNA helix

DNA helicase

single-strand DNA-binding proteinlagging-strand template

RNA primer

new Okazaki fragment DNA primase

DNA polymerase on lagging strand (just finishing an Okazaki fragment)

next Okazaki fragmentwill start here

primosome

clamp loader

parentalDNA helix

new Okazakifragment

lagging-strandtemplate

leading-strandtemplate

newly synthesized strand

newly synthesized strand

(A)

(B)

Chapter 6: DNA Replication, Repair, and Recombination6:14

primed with an RNA primer laid down by a separate enzyme (see Figure6–15). When the replication fork approaches the end of a chromosome,however, the replication machinery encounters a serious problem:there is no place to lay down the RNA primer needed to start theOkazaki fragment at the very tip of the linear DNA molecule. Therefore,some DNA could easily be lost from the ends of a DNA molecule eachtime it is replicated.

Bacteria solve this “end-replication” problem by having circularDNA molecules as chromosomes. Eucaryotes solve it by having specialnucleotide sequences at the ends of their chromosomes which areincorporated into telomeres. These repetitive telomeric DNA sequencesattract an enzyme called telomerase to the chromosome. Telomeraseadds multiple copies of the same telomere DNA sequence to the ends ofthe chromosomes, thereby producing a template that allows replicationof the lagging strand to be completed (Figure 6–18).

In addition to allowing replication of chromosome ends, telomeresserve additional functions: for example, the repeated telomere DNAsequences, together with the regions adjoining them, form structuresthat are recognized by the cell as the true ends of chromosomes ratherthan breaks that sometimes occur in the middle of chromosomes andmust be repaired.

DNA Replication Is Relatively Well Understood

Our current understanding of DNA replication is much more completethan that of many other aspects of cell biology, yet many mysteries stillremain. For example, it is not yet understood how the polymerase onthe leading strand is connected with that on the lagging strand in orderto allow replication to proceed synchronously on both strands.Moreover, although we know in some detail how DNA replicationbegins at replication origins in bacteria, our understanding of thisprocess in eucaryotes—including humans—is only just beginning.

Given the demands for accuracy during DNA replication, and thelengths to which cells go to achieve this precision, it is not surprising, aswe shall see shortly, that cells have also evolved elaborate proteinmachines to scan the finished product for mistakes. These proteinmachines then correct any errors made during DNA replication (rare asthey are) and repair any nucleotides that may have been accidentallydamaged by light, by chemicals in the cell, or by other mutation-caus-ing agents.

Figure 6–18 Telomeres allow thecompletion of DNA synthesis at the endsof eucaryotic chromosomes. To synthesizethe lagging strand at the very end of aeucaryotic chromosome, the machinery ofDNA replication requires a length oftemplate DNA extending beyond the DNAthat is to be copied. In a linear DNAmolecule, synthesis of the lagging strandthus stops short just before the end of thetemplate. But the enzyme telomerase addsa series of repeats of a DNA sequence tothe template strand, which allows thelagging strand to be completed by DNApolymerase, as shown. In humans, thenucleotide sequence of the repeat isGGGGTTA. The telomerase enzymecontains within it a short piece of RNA ofcomplementary sequence to the DNArepeat sequence; this RNA acts as thetemplate for the telomerase DNA synthesis.

3¢

5¢

5¢

3¢

5¢

3¢

5¢3¢

5¢

3¢

3¢

5¢

5¢

3¢

3¢

5¢

template strand

incomplete newly synthesizedlagging strand

telomerase addsadditional repeats tothe template strand

DNA synthesis from RNA primercompletes lagging strand

replicated chromosome end

DNA REPAIR

RNAprimer

Question 6–3

A gene encoding one ofthe proteins involved inDNA replication hasbeen inactivated by amutation in a cell. Inthe absence of this pro-

tein the cell attempts to replicate itsDNA for the very last time. WhatDNA products would be generatedin each case if the following proteinwere missing?

A. DNA polymerase

B. DNA ligase

C. Sliding clamp for DNApolymerase

D. Nuclease that removes RNAprimers

E. DNA helicase

F. Primase

6:15

DNA RepairThe diversity of living organisms and their success in colonizing almostevery part of the Earth’s surface depends on genetic changes accumu-lated gradually over millions of years, allowing organisms to adapt tochanging conditions and to colonize new habitats. However, in theshort term, and from the perspective of an individual organism, geneticchange is often detrimental, especially in multicellular organisms,where a genetic change can upset an organism’s extremely complex andfinely tuned development and physiology. To survive and reproduce,individuals must be genetically stable. This stability is achieved not onlythrough the extremely accurate mechanism for replicating DNA that wehave just discussed, but also through mechanisms for correcting therare copying mistakes made by the replication machinery and forrepairing the accidental damage that is continually occurring to theDNA. Most of these changes in DNA are only temporary because theyare immediately corrected by processes collectively called DNA repair.

Mutations Can Have Severe Consequences for anOrganism

Only rarely do the cell’s DNA replication and repair processes fail andallow a permanent change in the DNA. Such a permanent change iscalled a mutation, and it can have profound consequences. A mutationaffecting just a single nucleotide pair can severely compromise anorganism’s fitness if the change occurs in a vital position in the DNAsequence. Because the structure and activity of a protein depend on itsamino acid sequence, a protein with an altered sequence may functionpoorly or not at all. For example, humans use the protein hemoglobin totransport oxygen in the blood; the sequence of nucleotides that encodesthe amino acid sequence of one of the two types of protein chains (theb-globin chain) of the hemoglobin molecule is shown in Figure 5–11. Apermanent change in a single nucleotide in this sequence can causecells to make a b-globin chain with an incorrect sequence of aminoacids. Such a mutation causes the disease sickle-cell anemia (Figure6–19). The sickle-cell hemoglobin is less soluble than normal hemoglo-bin and forms fibrous precipitates, which lead to the characteristic

DNA Repair

G T G C A C C T G A C T C C T G A G G A G ---

G T G C A C C T G A C T C C T G T G G A G ---

single strand of normal �-globin gene

single strand of mutant �-globin gene

single nucleotidechanged (mutation)

(B)

(A)

(C)

5

5 mm 5 mm

Figure 6–19 A single nucleotide change causes the disease sickle-cell anemia. The complete nucleotide sequence of the b-globin geneis given in Figure 5–11. Only a small portion of the sequence nearthe beginning of the gene is shown in (A). The single nucleotidechange (mutation) in the sickle-cell gene produces a b-globin thatdiffers from normal b-globin only by a change from glutamic acid tovaline at the sixth amino acid position. (The b-globin moleculecontains a total of 146 amino acids.) Humans carry two copies ofeach gene (one inherited from each parent); a sickle-cell mutation inone of the two b-globin genes generally causes no harm to theindividual, as it is compensated for by the normal gene. However, anindividual who inherits two copies of the mutant b-globin genedisplays the symptoms of sickle-cell anemia. Normal red blood cellsare shown in (B), and those from an individual suffering from sickle-cell anemia in (C). Although sickle cell anemia can be a life-threatening disease, the mutation responsible can also be beneficial:patients with the disease, or who are heterozygous carriers of themutation, are more resistant to malaria than unaffected individuals.The parasite that causes malaria grows poorly in red blood cells fromhomozygous sickle-cell patients or from heterozygous carriers.

Chapter 6: DNA Replication, Repair, and Recombination6:16

sickle shape of affected red blood cells. Because these cells are morefragile and frequently break in the bloodstream, patients with thispotentially life-threatening disease have a reduced number of red bloodcells (Figure 6–19C), a deficiency that can cause weakness, dizziness,headaches, pain, and total organ failure.

The example of sickle-cell anemia, which is an inherited disease,illustrates the importance of protecting reproductive cells (germ cells)against mutation. A mutation in one of these will be passed on to all thecells in the body of the multicellular organism that develops from it,including the germ cells for production of the next generation.However, the many other cells in a multicellular organism (its somaticcells) must also be protected from genetic change to safeguard thehealth and well-being of the individual. Nucleotide changes that occurin somatic cells can give rise to variant cells, some of which grow in anuncontrolled fashion at the expense of the other cells in the organism.In the extreme case, an uncontrolled cell proliferation known as cancerresults. This disease, which is responsible for about 30% of the deathsthat occur in Europe and North America, is due largely to a gradualaccumulation of changes in the DNA sequences of somatic cells that iscaused by random mutation (Figure 6–20). Increasing the mutation fre-quency even two- or threefold would cause a disastrous increase in theincidence of cancer by accelerating the rate at which somatic cell vari-ants arise.

Thus the high fidelity with which DNA sequences are replicated andmaintained is important both for the reproductive cells, which transmitthe genes to the next generation, and for the somatic cells, which nor-mally function as carefully regulated members of the complex commu-nity of cells in a multicellular organism. We should therefore not be sur-prised to find that all cells have acquired an elegant set of mechanismsto reduce the number of mutations that occur in their DNA.

A DNA Mismatch Repair System Removes ReplicationErrors That Escape the Replication Machine

In the first part of this chapter, we saw that the high fidelity of the cell’sreplication machinery generally prevents copying mistakes. Despitethese safeguards, however, such mistakes do occur. Fortunately, the cellhas a backup system—called DNA mismatch repair—which is dedi-cated to correcting these rare mistakes. The replication machine itselfmakes approximately one error per 107 nucleotides copied; DNA mis-match repair corrects 99% of these errors, increasing the overall accu-racy to one mistake in 109 nucleotides copied. This level of accuracy ismuch higher than that generally encountered in the visible worldaround us (Table 6–1).

Figure 6–20 Cancer incidence increasesdramatically as a function of age. Thenumber of newly diagnosed cases ofcancer of the colon in women in Englandand Wales in one year is plotted as afunction of age at diagnosis. Since cells arecontinually experiencing accidentalchanges to their DNA that accumulate andare passed on to progeny cells, the chancethat a cell will become cancerous increasesgreatly with age. (Data from C. Muir et al.,Cancer Incidence in Five Continents, Vol. V.Lyon: International Agency for Research onCancer, 1987.)

0

20

40

60

80

100

120

140

160

180

10 20 30 40 50 60 70 80

inci

den

ce o

f ca

nce

r p

er 1

00,0

00 w

om

en

age (years)

Table 6–1 Error Rates

US Postal Service on-time delivery of 13 late deliveries per 100 parcelslocal first-class mail

Airline luggage system 1 lost bag per 200

A professional typist typing at 120 words 1 mistake per 250 charactersper minute

Driving a car in the United States 1 death per 104 people per year

DNA replication (without mismatch repair) 1 mistake per 107 nucleotides copied

DNA replication (including mismatch repair) 1 mistake per 109 nucleotides copied

6:17

Whenever the replication machinery makes a copying mistake, itleaves a mispaired nucleotide (commonly called a mismatch) behind. Ifleft uncorrected, the mismatch will result in a permanent mutation inthe next round of DNA replication (Figure 6–21A). A complex of mis-match repair proteins recognizes these DNA mismatches, removes(excises) one of the two strands of DNA involved in the mismatch, andresynthesizes the missing strand (Figure 6–22). To be effective in cor-recting replication mistakes, this mismatch repair system must alwaysexcise only the newly synthesized DNA strand: excising the other strand(the old strand) would preserve the mistake instead of correcting it (seeFigure 6–21).

In eucaryotes, it is not yet known for certain how the mismatchrepair machinery distinguishes the two DNA strands. However, there isevidence that newly replicated DNA strands—both leading and lag-ging—are preferentially nicked; it is these nicks (single-stranded breaks)that appear to provide the signal that directs the mismatch repairmachinery to the appropriate strand (see Figure 6–22).

The importance of mismatch repair in humans was recognizedrecently when it was discovered that an inherited predisposition to cer-tain cancers (especially some types of colon cancer) is caused by amutation in the gene responsible for producing one of the mismatchrepair proteins. Humans inherit two copies of this gene (one from eachparent), and individuals who inherit one damaged mismatch repairgene show no symptoms until the undamaged copy of the gene is acci-dentally mutated in a somatic cell. This gives rise to a clone of somaticcells that, because they are deficient in mismatch repair, accumulatemutations more rapidly than do normal cells. Because most cancers

DNA Repair

A

G

A

T

C

G

mutated

unchanged

next roundof DNAreplication

newlysynthesizedstrand

templatestrand

NO REPAIR

A

G

A

T

A

T

mutated

mutated

next roundof DNAreplication

newlysynthesizedstrand

templatestrand

EXCISION AND REPAIR OF ONLYTHE TEMPLATE (OLD) STRAND

EXCISION AND REPAIR OF ONLYTHE NEWLY SYNTHESIZED STRAND

A

G

C

G

C

G

unchanged

unchanged

next roundof DNAreplication

newlysynthesizedstrand

templatestrand

(A) (B) (C)

Figure 6–21 DNA mismatch repaircorrects errors made during DNAreplication. (A) If uncorrected, themismatch will lead to a permanentmutation in one of the two DNA moleculesproduced by the next round of DNAreplication. (B) If the mismatch is“repaired” using the newly synthesizedDNA strand as the template, both DNAmolecules produced by the next round ofDNA replication will contain a mutation.(C) If the mismatch is corrected using theoriginal template (old) strand as thetemplate, the possibility of a mutation iseliminated. The scheme shown in (C) isused by cells to repair mismatches, asshown in Figure 6–22.

BINDING OF DNAMISMATCH REPAIRPROTEINS

REMOVAL OF NEWLYSYNTHESIZED DNASTRAND

REPAIR OF GAP BYDNA POLYMERASE AND LIGASE

DNA mismatch

DNA mismatchrepair proteins

nick

newly madeDNA strand

old DNA strandFigure 6–22 DNA mismatch repair proteins correct errors thatoccur during DNA replication. A DNA mismatch, formed when anincorrectly matched base is incorporated into a newly synthesizedDNA chain, distorts the geometry of the double helix. This distortion issubsequently recognized by the DNA mismatch repair proteins, whichthen remove the newly synthesized DNA. The gap in the newlysynthesized DNA is replaced by a DNA polymerase that proofreads asit synthesizes and is sealed by DNA ligase. As shown in the figure, anick in the DNA has been proposed as the signal that allows themismatch repair proteins to distinguish the newly synthesized DNA(which contains the mistake) from the old DNA. Such nicks areknown to occur in the lagging strands (see Figure 6–12) and areobserved to also occur, although less frequently, in the leadingstrands. These nicks remain for only a short period after a replicationfork passes (see Figure 6–16 or 6–17), so that mismatch repair mustoccur quickly.

Question 6–4

Discuss the followingstatement: “The DNArepair enzymes thatcorrect defects intro-duced by deaminationand depurination reac-

tions must preferentially recognizesuch defects on newly synthesizedDNA strands.”

Chapter 6: DNA Replication, Repair, and Recombination6:18

arise from cells that have accumulated multiple mutations (see Figure6–20), a cell deficient in mismatch repair has a greatly enhanced chanceof becoming cancerous. Thus, inheriting a damaged mismatch repairgene predisposes an individual to cancer.

DNA Is Continually Suffering Damage in Cells

Rare mistakes in DNA replication, as we have seen, can be corrected bythe mismatch repair mechanism. There are also other ways in which theDNA can be damaged, and these require other mechanisms for theirrepair. Just like any other molecule in the cell, DNA is continuallyundergoing thermal collisions with other molecules. These often resultin major chemical changes in the DNA. For example, during the time ittakes to read this sentence, a total of about a trillion (1012) purine bases(A and G) will be lost from the DNA of your cells by a spontaneous reac-tion called depurination (Figure 6–23). Depurination does not break thephosphodiester backbone but, instead, gives rise to lesions that resem-ble missing teeth. Another major change is the spontaneous loss of anamino group (deamination) from cytosine in DNA to produce the baseuracil (see Figure 6–23). Some chemically reactive by-products ofmetabolism also occasionally react with the bases in DNA, alteringthem in such a way that their base-pairing properties are changed. Theultraviolet radiation in sunlight is also damaging to DNA; it promotescovalent linkage between two adjacent pyrimidine bases, forming, forexample, the thymine dimer shown in Figure 6–24.

These are only a few of many chemical changes that can occur inour DNA. If left unrepaired, many of them would lead either to the sub-stitution of one nucleotide pair by another as a result of incorrect base-pairing during replication or to deletion of one or more nucleotide pairsin the daughter DNA strand after DNA replication (Figure 6–25). Sometypes of DNA damage (thymine dimers, for example) often stall theDNA replication machinery at the site of the damage. All of these typesof damage, if unrepaired, would have disastrous consequences for anorganism.

Figure 6–23 Depurination anddeamination are the most frequentchemical reactions known to createserious DNA damage in cells.(A) Depurination can release guanine aswell as adenine from DNA. (B) The majortype of deamination reaction convertscytosine to an altered DNA base, uracil,but deamination can occur on other basesas well. Both of these reactions take placeon double-helical DNA; for convenience,only one strand is shown.

GUANINE

OO CH2P

O

O

O_

N

N N

NH

N

H

H

HO

GUANINE

OO CH2P

O

O

O_N

N N

NH

N

H

H

HO

DNA strand DNA strand

H

OH

H2O

depurinatedsugar

CYTOSINE URACIL

N

N

N

H H

H

H O

OO CH2P

O

O

O_

O

N

NHH

H O

OO CH2P

O

O

O_

H2O

NH3

DEAMINATION

DEPURINATION (A)

(B)

6:19

The Stability of Genes Depends on DNA Repair

The thousands of random chemical changes that occur every day in theDNA of a human cell, by metabolic accidents or exposure to DNA-dam-aging chemicals, are repaired by a variety of mechanisms, each cat-alyzed by a different set of enzymes. Nearly all these mechanismsdepend on the existence of two copies of the genetic information, onein each strand of the DNA double helix: if the sequence in one strand isaccidentally damaged, information is not lost irretrievably, because abackup version of the altered strand remains in the complementarysequence of nucleotides in the other strand. Most damage creates struc-tures that are never encountered in an undamaged DNA strand; thusthe good strand is easily distinguished from the bad. The basic pathwayfor repairing damage to DNA is illustrated schematically in Figure 6–26.

DNA Repair

Figure 6–24 The ultraviolet radiation insunlight causes DNA damage. Twoadjacent thymine bases have becomecovalently attached to one another to forma thymine dimer. Skin cells that areexposed to sunlight are especiallysusceptible to this type of DNA damage.

Figure 6–25 Chemical modifications of nucleotides, if left unrepaired, produce mutations. (A) Deamination of cytosine, ifuncorrected, results in the substitution of one base for another when the DNA is replicated. As shown in Figure 6–23, deamination ofcytosine produces uracil. Uracil differs from cytosine in its base-pairing properties and preferentially base-pairs with adenine. The DNAreplication machinery therefore inserts an adenine when it encounters a uracil on the template strand. (B) Depurination, ifuncorrected, can lead to the loss of a nucleotide pair. When the replication machinery encounters a missing purine on the templatestrand, it can skip to the next complete nucleotide, thus producing a nucleotide deletion in the newly synthesized strand. In othercases, the replication machinery places an incorrect nucleotide across from the missing base, again resulting in a mutation.

O

P_O O

OOH2C

O

P_O O

OOH2C

O

O

P_O O

OOH2C

O

P_O O

OOH2C

O

H

O

O

CH3

CCN

N

O

O

CH3

CCN

O

O

CH3

CCN

N

CC

H

O

O

CH3

CCN

N

CC

CC

CC

NH

H

H

H

H

Hthymine

thymine

thymine dimer

UV light

UAT T

G AA T

U AT T

A AA T

C AT T

G AA T

a G has beenchanged to an A

DNAREPLICATION

DNAREPLICATION

deaminated C

new strand

new strand

old strand

old strand

(A)

an A-T nucleotidepair has been deleted

mutated

unchanged

CT T

G AA T

CT T

G AA

C AT T

G AA T

depurinated A

new strand

new strand

old strand

old strand

(B)

mutated

unchanged

Chapter 6: DNA Replication, Repair, and Recombination6:20

As indicated, it involves three steps:

1. The damaged DNA is recognized and removed by one of a varietyof different nucleases, which cleave the covalent bonds that jointhe damaged nucleotides to the rest of the DNA molecule, leavinga small gap on one strand of the DNA double helix in this region.

2. A repair DNA polymerase binds to the 3¢-hydroxyl end of the cutDNA strand. It then fills in the gap by making a complementarycopy of the information stored in the undamaged strand.Although a different enzyme from the DNA polymerase that repli-cates DNA, a repair DNA polymerase synthesizes DNA strands inthe same way. For example, it synthesizes chains in the 5¢-to-3¢direction and has the same type of proofreading activity to ensurethat the template strand is accurately copied. In many cells, this isthe same enzyme that fills in the gap left after the RNA primers areremoved in normal DNA replication (see Figure 6–16).

3. When the repair DNA polymerase has filled in the gap, a breakremains in the sugar–phosphate backbone of the repaired strand.This nick in the helix is sealed by DNA ligase, the same enzyme thatjoins the lagging-strand DNA fragments during DNA replication.

Steps 2 and 3 are nearly the same for most types of DNA repair, includ-ing mismatch repair. However, step 1 uses a series of different enzymes,each specialized for removing different types of DNA damage.

The importance of these repair processes is indicated by the largeinvestment that cells make in DNA repair enzymes. Single-celled organ-isms such as yeasts contain more than 50 different proteins that func-tion in DNA repair, and DNA repair pathways are likely to be even morecomplex in humans. The importance of these DNA repair processes isalso evident from the consequences of their malfunction. Humans withthe genetic disease xeroderma pigmentosum, for example, cannot repairthymine dimers (see Figure 6–24) because they have inherited a defec-tive gene for one of the proteins involved in this repair process. Suchindividuals develop severe skin lesions, including skin cancer, because ofthe accumulation of thymine dimers in cells that are exposed to sunlightand the consequent mutations that arise in the cells that contain them.

The High Fidelity of DNA Maintenance Allows CloselyRelated Species to Have Proteins with Very SimilarSequences

We have seen in this chapter that DNA is replicated and maintainedwith remarkable fidelity. As a consequence, changes in the DNA accu-mulate remarkably slowly in the course of evolution. Of course, the rateof evolutionary change in the DNA of a species depends also on theeffects of natural selection: DNA copying errors that have harmful con-sequences for the organism are eliminated from the population throughthe death or reduced fertility of individuals carrying the misreplicatedDNA. But the mechanisms of DNA replication and repair are so accuratethat even where no such selection operates—at the many sites in theDNA where a change of nucleotide has no effect on the fitness of theorganism—the genetic message is faithfully preserved over tens of mil-lions of years. Thus humans and chimpanzees, after about 5 millionyears of divergent evolution, still have DNA sequences that are at least98% identical. Even humans and whales, after 10 or 20 times this period,still have chromosomes that are unmistakably similar in their DNAsequence and many proteins with amino acid sequences that arealmost identical (Figure 6–27). Thus, in our genomes, we and our rela-tives receive a message from the distant past—a message that is longer

Figure 6–26 The basic mechanism ofDNA repair involves three steps: excision,resynthesis, and ligation. In step 1(excision), the damage is cut out by one ofa series of nucleases, each specialized fora type of DNA damage. In step 2(resynthesis), the original DNA sequence isrestored by a repair DNA polymerase,which fills in the gap created by theexcision events. In step 3 (ligation), DNAligase seals the nick left in thesugar–phosphate backbone of the repairedstrand. Nick sealing, which requires energyfrom ATP hydrolysis, remakes the brokenphosphodiester bond between the adjacentnucleotides. Some types of DNA damage(the deamination of cytosine [Figure6–23], for example) involve thereplacement of a single nucleotide, asshown in the figure. For the repair of otherkinds of DNA damage, such as thyminedimers (see Figure 6–24), a longer stretchof 10 to 20 nucleotides is removed fromthe damaged strand.

DAMAGE TOTOP STRAND

EXCISION OFDAMAGED REGION

DNA POLYMERASE MAKES NEW TOP STRAND USING BOTTOM STRAND AS A TEMPLATE

DNA LIGASESEALS NICK

NET RESULT: REPAIRED DNA

step 1

step 2

step 3

6:21

and more detailed than any book. Thanks to the faithfulness of DNAreplication and repair, 100 million years have scarcely changed itsessential content.

DNA Recombination Thus far we have discussed how the DNA sequences in cells can be main-tained from generation to generation with very little change. However, itis also clear that the DNA sequence in chromosomes does change withtime and can, as we discuss in this section, even be rearranged. The par-ticular combination of genes present in any individual genome is oftenaltered by such DNA rearrangements. In a population, this sort of geneticvariation is important to allow organisms to evolve in response to achanging environment. These DNA rearrangements are caused by a classof mechanisms called genetic recombination.

We begin by reviewing several types of recombination mecha-nisms, each of which can introduce change into the genome of an indi-vidual cell. We discuss later in Chapter 9 how such rearrangements con-tribute to the genetic diversity and evolution of entire species.

Homologous Recombination Results in an ExactExchange of Genetic Information

Of all the recombination mechanisms that exist, perhaps the most fun-damental is homologous recombination. The central features that lieat the heart of homologous recombination seem to be the same in allorganisms on Earth. Although the mechanism is not completely under-stood, the following characteristics are probably common to homolo-gous recombination in all cells:

1. Two double-stranded DNA molecules that have regions of verysimilar (homologous) DNA sequence align so that their homolo-gous sequences are in register. They can then “cross over”: in acomplex reaction, both strands of each double helix are brokenand the broken ends are rejoined to the ends of the opposite DNAmolecule to re-form two intact double helices, each made up ofparts of the two different DNA molecules (Figure 6–28).

2. The site of exchange (that is, where a red double helix is joined toa green double helix in Figure 6–28) can occur anywhere in thehomologous nucleotide sequences of the two participating DNAmolecules.

3. No nucleotide sequences are altered at the site of exchange; thecleavage and rejoining events occur so precisely that not a singlenucleotide is lost or gained.

DNA Recombination

Figure 6–27 The sex determination genesfrom humans and whales areunmistakably similar. Although their bodyplans are strikingly different, humans andwhales are built from the same proteins.Despite the length of time since humansand whales diverged, the nucleotidesequences of many of their genes are stillclosely similar. The sequences of a part ofthe gene encoding the protein thatdetermines maleness in humans andwhales are shown one above the other,and positions where the two are identicalare shaded.

whale

human

two homologous DNA double helices

DNA molecules that have crossed over

Xcrossoverpoint

Figure 6–28 Homologous recombinationtakes place between DNA molecules withsimilar nucleotide sequences. Thebreaking and rejoining of two homologousDNA double helices creates two DNAmolecules that have “crossed over.”Although the two original DNA moleculesmust have similar nucleotide sequences inorder to cross over, they do not have to beidentical; thus a crossover can create DNAmolecules of novel nucleotide sequence.

Chapter 6: DNA Replication, Repair, and Recombination6:22

Homologous recombination begins with a bold stroke: a specialenzyme simultaneously cuts both strands of the double helix, creating acomplete break in the DNA molecule (Figure 6–29). The 5¢ ends at thebreak are then chewed back by a DNA-digesting enzyme, creating pro-truding single-stranded 3¢ ends. Each of these single strands thensearches for a homologous, complementary DNA helix with which topair—leading to the formation of a “joint molecule” between the twochromosomes. The nicks in the DNA strands are then sealed so that thetwo DNA molecules are now held together physically by a crossing-overof one of each of their strands. This crucial intermediate in homologousrecombination is known as a cross-strand exchange, or Holliday junc-tion (Figure 6–30).

To regenerate two separate DNA molecules, the two crossingstrands must be cut. But if they are cut while the structure is still in theform shown in Figure 6–30A, the two original DNA molecules wouldseparate from each other almost unaltered (Figure 6–30D). The struc-ture can, however, undergo a series of rotational movements so that thetwo original noncrossing strands become crossing strands and viceversa (Figure 6–30B and C, and Figure 6–31). If the crossing strands arecut after rotation, one section of each original DNA helix is joined to asection of the other DNA helix; in other words, the two DNA moleculeshave crossed over, and two molecules of novel DNA sequence have beenproduced (Figure 6–30E).

As might be expected, cells use a set of specialized proteins to facil-itate homologous recombination; these proteins break the DNA, cat-alyze strand exchange, and cleave Holliday structures. Because theessential features of homologous recombination are highly conserved,the proteins that carry out this process in different organisms are oftenvery similar to one another in amino acid sequence.

Homologous recombination provides many advantages to cells andorganisms. The process allows an organism to repair DNA that is dam-aged on both strands of the double helix, and it can fix other geneticaccidents that occur during nearly every round of DNA replication. It isalso essential for the accurate chromosome segregation that occursduring meiosis in fungi, plants, and animals, as we shall see in Chapter19. The chromosomal “crossing-over” that occurs when homologouschromosomes come together causes bits of genetic information to beexchanged, generating new combinations of DNA sequences in eachchromosome. The benefit of such gene mixing for the progeny organ-isms is apparently so great that the reassortment of genes by homolo-gous recombination is not confined to sexually reproducing organ-isms; it is also widespread in asexually reproducing organisms, such asbacteria.

Recombination Can Also Occur BetweenNonhomologous DNA Sequences

In homologous recombination, DNA rearrangements occur betweenDNA segments that are very similar in sequence. A second, more spe-cialized type of recombination, called site-specific recombination,

Figure 6–29 Homologous recombination begins with a double-strand break in a chromosome. A DNA-digesting enzyme thencreates protruding 3¢ ends, which find the homologous region of asecond chromosome. The joint molecule formed can eventually beresolved by selective strand cuts to produce two chromosomes thathave crossed over, as shown.

DOUBLE-STRANDBREAK

LIMITEDDEGRADATIONFROM 5¢ ENDS

PAIRING OFONE END WITHMATERNAL CHROMOSOME A

3¢

3¢

3¢

3¢

3¢ 5¢

DNA SYNTHESIS

+

5¢

PAIRING OFOTHER END; DNA SYNTHESIS

joint molecule

crossing-overbetween chromosomes

A and B

chromosome A

RESOLUTION BYSTRAND CUTTING

3¢5¢

5¢3¢

3¢5¢

5¢3¢

chromosome B

1

2

3

4

5

6

allows DNA exchanges to occur between DNA double helices that aredissimilar in nucleotide sequence. Although site-specific recombina-tion performs a variety of tasks in the cell, perhaps its most prevalentfunction is to shuffle specialized bits of DNA called mobile genetic ele-ments. These elements, found in the genomes of nearly all organisms,are short sequences of DNA that can move from one position in thegenome to another through site-specific recombination.

Some of these mobile genetic elements are viruses that take advan-tage of site-directed recombination to move their genomes into and outof the chromosomes of their host cell. A virus can package its nucleicacid into viral particles that can move from one cell to another throughthe extracellular environment. However, most mobile elements canmove only within a single cell and its descendants, as they lack anyintrinsic ability to leave the cell in which they reside.

Mobile genetic elements often comprise a sizable fraction of anorganism’s DNA. For example, approximately 45% of the humangenome is made up of mobile genetic elements; most of these elements,however, are fossils that—because they have been accumulating ran-dom mutations throughout the course of human evolution—have lostthe ability to move within the genome.

Because they have a tendency to multiply, mobile DNA elementsare sometimes called parasitic DNA. However, as we discuss in Chapter9, mobile genetic elements also provide some advantages to their hostgenomes by generating the genetic variation upon which evolutiondepends.

Mobile Genetic Elements Encode the Components TheyNeed for Movement

Unlike homologous recombination, site-specific recombination isguided by recombination enzymes that recognize short, specificnucleotide sequences present on one or both of the recombining DNAmolecules; extensive DNA homology is not required. Each type ofmobile genetic element generally encodes the enzyme that mediates itsown movement and contains special sites upon which the enzyme acts

6:23DNA Recombination

Figure 6–30 The rotation of a Hollidayjunction allows recombination to occur. Instep A, a cross-strand exchange hasformed, as shown in Figure 6–29. (A) Without a rotation (also calledisomerization), cutting of the two crossingstrands would terminate the exchange andhomologous recombination would notoccur (D). After rotation (B and C), cuttingthe two crossing strands creates two DNAmolecules that have exchanged pieces ofDNA (E). In the DNA strands shown here,the entire structure is rotated to yield thecross-strand exchange shown in C; to formB, the chromosome strands on the left-hand side of A are held steady whilethe strands on the right-hand side of thestructure are rotated as indicated; to formC, the strands in the upper portion of B are held steady while the lower strandsare rotated as indicated. In reality, however,only the regions immediately surroundingthe site of crossing over need to rotaterelative to one another to achieve the same result.

Figure 6–31 A cross-strand exchange(Holliday junction) can be seen in theelectron micrograph. This view of themolecule corresponds to the open structureillustrated in Figure 6–30B. (Courtesy ofHuntington Potter and David Dressler.)

(A) (B) (C)

(D)

(E)

Chapter 6: DNA Replication, Repair, and Recombination6:24

(Figure 6–32). Many elements also carry other genes. For example,viruses encode coat proteins that enable them to exist outside cells, inaddition to essential viral enzymes. The spread of mobile genetic ele-ments that carry antibiotic resistance genes is a major factor underlyingthe widespread dissemination of antibiotic resistance in bacterial pop-ulations.

In bacteria, the most common genetic elements are called DNA-only transposons; these generally have only modest selectivity for theirtarget sites and can thus insert themselves into many different DNAsequences. These transposons move from place to place within thegenome by means of specialized recombination enzymes, called trans-posases, that are encoded by the transposable elements themselves (seeFigure 6–32). The transposase first disconnects the transposon from theflanking DNA and then inserts it into a new target DNA site. Again, thereis no requirement for homology between the ends of the element andthe insertion site. Some bacterial transposons move to the target siteusing a cut-and-paste mechanism; others replicate before inserting intothe new chromosomal site, leaving the original copy intact at its previ-ous location (Figure 6–33).

A Large Fraction of the Human Genome Is Composed ofTwo Families of Transposable Sequences

As discussed in Chapter 5, a significant fraction of many vertebratechromosomes is made up of repeated DNA sequences. Many of theserepeated sequences are mobile DNA elements that have proliferatedover evolutionary time-scales, although most are no longer actively

Figure 6–32 Bacteria contain many typesof mobile genetic elements three of whichare shown. Each of these DNAtransposons contains a gene that encodesa transposase (blue), an enzyme thatcarries out some of the DNA breaking andjoining reactions needed for the transposonto move. Each transposon also carries DNAsequences (indicated in red) that arerecognized only by the transposaseencoded by that element and are necessaryfor movement of the transposon. Sometransposons carry, in addition, genes thatencode enzymes that inactivate antibioticssuch as ampicillin (ampR) and tetracycline(tetR). Movement of these genes presents agrowing problem in medicine, as manydisease-causing bacteria have becomeresistant to many of the antibioticsdeveloped during the twentieth century.

IS3

Tn3

Tn10

2 kb

transposase gene

transposase geneampR

tetRtransposase gene

Figure 6–33 Bacterial transposons moveby two mechanisms. (A) In cut-and-pastenonreplicative transposition, the transposonis cut out of the donor DNA and insertedinto the target DNA, leaving behind abroken donor DNA molecule. (The donorcan be repaired in a variety of ways, butthis sometimes results in deletions orrearrangements of the donor molecule.) (B) In the course of replicativetransposition, the transposon is copied byDNA replication. The end products are amolecule that appears identical to theoriginal donor and a target molecule thathas a transposon inserted into it. Ingeneral, a particular type of transposonmoves by only one of these mechanisms.However, the two mechanisms have manyenzymatic similarities, and a fewtransposons can move by eithermechanism. The donor and target DNAscan be part of the same DNA molecule orreside on different molecules.

CUT-AND-PASTENONREPLICATIVETRANSPOSITION

donor DNA

target DNA +

transposon

(A)

REPLICATIVETRANSPOSITION

donor DNA

target DNA +

(B)

new DNAsequence

new DNAsequence

6:25

moving due to the accumulation of deleterious mutations. Some verte-brate mobile elements have moved from place to place within their hostchromosomes using the cut-and-paste mechanism discussed above forbacterial transposons (see Figure 6–33A). However, many others havemoved not as DNA but via an RNA intermediate. These are called retro-transposons and are, as far as is known, unique to eucaryotes.

One type of retrotransposon, the L1 element (sometimes referred toas LINE-1), is a highly repeated sequence that constitutes about 15% ofthe total mass of the human genome. Although most copies of the L1element are immobile, a few still retain the ability to move.Translocation of L1 can sometimes result in human disease: a particu-lar form of hemophilia, for example, is caused by insertion of an L1 ele-ment into the gene that encodes Factor VIII, a protein essential forproper blood clotting.

The L1 element transposes by first being transcribed into RNA bycellular RNA polymerases. A DNA copy of this RNA is then made usingthe enzyme reverse transcriptase, an unusual DNA polymerase that canuse RNA as a template. Reverse transcriptase is encoded by the L1 ele-ment itself. The DNA copy can then reintegrate into another site in thegenome (Figure 6–34).

Almost as abundant in the human genome is the Alu sequence,which is unusually short (about 300 nucleotide pairs). Alu is present inabout 1 million copies in the genome and constitutes about 11% ofhuman DNA; thus it appears on average about once every 5000nucleotide pairs. Only some of the Alu sequences in the genome canstill be copied into RNA. Because Alu elements do not encode their ownreverse transcriptase, they depend on enzymes already present in thecell to help them move.

Comparisons of the sequence and locations of the L1 and Alu-likeelements in different mammals suggest that these sequences have mul-tiplied to high copy numbers in primates relatively recently in evolu-tionary time (Figure 6–35). These highly abundant sequences, scattered

DNA Recombination

Figure 6–34 Retrotransposons move viaan RNA intermediate. These transposableelements are first transcribed into an RNAintermediate. A DNA copy of this RNA ismade by reverse transcriptase. The DNAcopy of the transposon is then inserted intothe target location. The target can be onthe same or a different DNA molecule fromthe donor. These transposable elements are called retrotransposons because at onestage in their transposition their geneticinformation flows backward, from RNA to DNA.

RNA polymerase

reverse transcriptase

DONOR DNA

TARGET DNA

RNA

DNAcopy

+INSERTION OF

DNA COPY

human �-globin gene cluster

mouse �-globin gene cluster

� � ��G

� � �major �minor

�A

10,000nucleotide pairs

Figure 6–35 L1 and Alu-like elementshave multiplied to high copy numbersrelatively recently in evolutionary time.The human genome contains a cluster offive globin genes (top). Each gene (shownin orange and designated by a Greek letter)encodes a protein that carries oxygen inthe blood. The comparable region from themouse genome (bottom) contains only fourglobin genes. The positions of the humanAlu sequences are indicated by greencircles, and the human L1 elements by redcircles. The mouse genome containsdifferent transposable elements: thepositions of B1 elements (which are relatedto the human Alu sequences) are indicatedby blue triangles, and the positions of themouse L1 elements (which are related tothe human L1) are indicated by yellowtriangles. Because the DNA sequences ofthe mouse and human transposableelements are distinct and because thepositions of these transposable elements onthe b-globin gene cluster are very differentbetween human and mouse, it is believedthat they accumulated in mammaliangenomes relatively recently in evolutionarytime. (Courtesy of Ross Hardison andWebb Miller.)

Chapter 6: DNA Replication, Repair, and Recombination6:26

throughout our genome, must have had major effects on the expressionof many of our genes. It is perhaps humbling to contemplate how manyof our uniquely human qualities we might owe to these parasitic geneticelements.

Viruses Are Fully Mobile Genetic Elements That CanEscape from Cells

Viruses were first noticed as disease-causing agents that, by virtue oftheir tiny size, passed through ultrafine filters that can hold back eventhe smallest cell. We now know that viruses are essentially genesenclosed by a protective coat. However, these genes must enter a celland utilize the cell’s machinery to express their genes as proteins and toreplicate their chromosomes, so as to package themselves into newlymade protective coats. Virus reproduction per se is often lethal to thecells in which it occurs; in many cases the infected cell breaks open(lyses) and thereby releases the progeny viruses and allows them accessto nearby cells. Many of the medical symptoms of viral infection reflectthe lytic effect of the virus. The cold sores formed by herpes simplexvirus and the blisters caused by the chicken pox virus, for example, bothreflect the localized killing of skin cells. Although the first viruses thatwere discovered attack mammalian cells, it is now recognized thatmany types of viruses exist. Some of these infect plant cells, while oth-ers use bacterial cells as their hosts.

Viral genomes can be made of DNA or RNA and can be single-stranded or double-stranded (Table 6–2 and Figure 6–36). The amountof DNA or RNA that can be packaged inside a protein shell is limited,and is too small to encode the many different enzymes and other pro-teins that are required to replicate even the simplest virus. Viruses aretherefore parasites that can reproduce themselves only inside a living

Table 6–2 Viruses That Cause Human Disease

VIRUS GENOME TYPE DISEASE

Herpes simplex virus double-stranded DNA recurrent cold sores

Epstein-Barr virus (EBV) double-stranded DNA infectious mononucleosis

Varicella-zoster virus double-stranded DNA chicken pox and shingles

Smallpox virus double-stranded DNA smallpox

Hepatitis B virus part single-, part serum hepatitisdouble-stranded DNA

Human immuno- single-stranded RNA acquired immunodeficiency deficiency virus (HIV) syndrome (AIDS)

Influenza virus type A single-stranded RNA respiratory disease (flu)

Poliovirus single-stranded RNA infantile paralysis

Rhinovirus single-stranded RNA common cold

Hepatitis A virus single-stranded RNA infectious hepatitis

Hepatitis C virus single-stranded RNA non-A, non-B type hepatitis

Yellow fever virus single-stranded RNA yellow fever

Rabies virus single-stranded RNA rabies

Mumps virus single-stranded RNA mumps

Measles virus single-stranded RNA measles

6:27

cell, where they are able to hijack the cell’s own biochemical machinery.Viral genomes typically encode the viral coat proteins as well as pro-teins that attract host enzymes to replicate their genome at the viralreplication origin (Figure 6–37).

The simplest viruses consist of a protein coat made up primarily ofmany copies of a single polypeptide chain surrounding a small genomecomposed of as few as three genes. More complex viruses have largergenomes of up to several hundred genes, surrounded by an elaborateshell composed of many different proteins (Figure 6–38).

Even the largest viruses depend heavily on their host cells forbiosynthesis; no known virus makes its own ribosomes or generates theATP needed for nucleic acid replication, for example. Clearly, cells musthave evolved before viruses.

If we accept the proposal outlined in Chapter 7 that the first cellswould have had RNA-based information storage, the first virus mayhave formed from an RNA plasmid that had acquired a gene coding fora protein that could assemble into a protein coat. Protected by this coat,the viral RNA could now leave the cell and seek a new host.

Retroviruses Reverse the Normal Flow of GeneticInformation

Although there are many similarities between bacterial and eucaryoticviruses, one important type of virus—the retrovirus—is found only ineucaryotic cells. In many respects, retroviruses resemble the retrotrans-posons we discussed earlier. A key feature found in both these geneticelements is a step where DNA is synthesized using RNA as a template(the term “retro” refers to this backward flow of the central dogma). Theenzyme that carries out this step is reverse transcriptase; the retroviralgenome (which is single-stranded RNA) encodes this enzyme, and a fewmolecules of the enzyme are packaged along with the RNA genome ineach individual virus.

DNA Recombination

Figure 6–36 Viral genomes differ instructure. The smallest viruses containonly a few genes and can have an RNA ora DNA genome; the largest viruses containhundreds of genes and have a double-stranded DNA genome. Some examples ofthese types of viruses are as follows:single-stranded RNA—tobacco mosaicvirus, bacteriophage R17, poliovirus;double-stranded RNA—reovirus; single-stranded DNA—parvovirus; single-stranded circular DNA—M13 and fX174bacteriophages; double stranded circularDNA—SV40 and polyoma-viruses; double-stranded DNA—T4 bacteriophage,herpesvirus; double-stranded DNA withcovalently linked terminal protein—adenovirus; double-stranded DNA withcovalently sealed ends—poxvirus. Thepeculiar ends (as well as the circularforms) present in some viral genomesovercome the difficulty of replicating thelast few nucleotides at the end of a DNAchain (see Figure 6–18).

single-stranded RNA single-stranded DNAdouble-stranded

circular DNA

double-stranded DNA witheach end covalently sealed

single-strandedcircular DNA

double-stranded RNA double-stranded DNA

double-stranded DNAwith covalently linked

terminal protein

Figure 6–37 The life cycle of a hypothetical virus. The simple virusillustrated consists of a small double-stranded DNA molecule thatencodes just a single type of viral coat protein. No known virus isquite this simple. In order to replicate, the viral genome must enterthe cell. This is followed by replication of the viral DNA to form manycopies. At the same time, the viral genes are expressed into proteinthrough the steps of transcription and translation, which will bedescribed in the next chapter. The viral genomes assemblespontaneously with the coat proteins to form new virus particles.

virusDNA

coat protein

ENTRY OF DNA INTO CELL

DNA

TRANSCRIPTION REPLICATION

DNARNA

TRANSLATION

coat protein

ASSEMBLY OF PROGENYVIRUS PARTICLES AND

CELL LYSIS

cell

The life cycle of a retrovirus is shown in Figure 6–39. When the sin-gle-stranded RNA genome of the retrovirus enters a cell, the reversetranscriptase brought in with it makes a complementary DNA strand toform a DNA/RNA hybrid double helix. The RNA strand is removed, andthe reverse transcriptase (which can use either DNA or RNA as a tem-plate) now synthesizes a complementary strand to produce a DNA dou-ble helix. This DNA is then integrated into a randomly selected site inthe host genome by a virally encoded integrase enzyme. In this state, thevirus is latent: each time the host cell divides, it passes on a copy of theintegrated viral genome to its progeny cells.

The next step in the replication of a retrovirus—which can takeplace long after its integration into the host genome—is the transcrip-tion of the integrated viral DNA by the host cell RNA polymerase, whichcan produce large numbers of single-stranded RNAs identical to theoriginal infecting genome. These RNA molecules are then translated bythe host cell machinery to produce the protein shell, the envelope pro-teins, and reverse transcriptase—all of which are assembled with theRNA genome into new virus particles.

The human immunodeficiency virus (HIV), which is the cause ofAIDS, is a retrovirus. As with other retroviruses, the HIV genome canpersist in the latent state as a DNA provirus embedded in the chromo-somes of an infected cell. This ability of the virus to hide within hostcells complicates any attempt to treat the infection with antiviral drugs.But because the HIV reverse transcriptase is not used by cells for anypurpose of their own, one of the prime targets of drug developmentagainst AIDS is the viral reverse transcriptase.

6:28

Figure 6–38 The coats of viruses. Theseelectron micrographs of virus particles areall shown at the same scale. (A) T4, alarge DNA-containing virus that infects E. coli cells. The DNA is stored in thebacteriophage head and injected into thebacterium through the cylindrical tail. (B) Potato virus X, a plant virus thatcontains an RNA genome. (C) Adenovirus,a DNA-containing virus that can infecthuman cells. (D) Influenza virus, a largeRNA-containing animal virus whose proteincapsid is further enclosed in a lipid-bilayer-based envelope. The spikes protrudingfrom the envelope are viral proteinsembedded in the membrane bilayer (seeFigure –). (A, courtesy of JamesPaulson; B, courtesy of Graham Hills; C, courtesy of Mei Lie Wong; D, courtesyof R.C. Williams and H.W. Fisher.)

(A) (B)

(D)(C)

100 nm

Chapter 6: DNA Replication, Repair, and Recombination

Question 6–5

Reverse transcriptasesdo not proofread asthey synthesize DNAusing an RNA template.What do you think theconsequences of this

are for the treatment of AIDS?

Since some transposable elements that move through the genomevia an RNA intermediate replicate in a manner similar to that of retrovi-ral genomes (see Figure 6–), it is thought that retroviruses are derivedfrom such a retrotransposon that long ago acquired additional genesencoding coat proteins and other proteins required to make a virus par-ticle. The RNA stage of its replicative cycle could then be packaged intoa virus particle and leave the cell.

Essential Concepts• The ability of a cell to maintain order in a chaotic environment

depends on the accurate duplication of the vast quantity of geneticinformation carried in its DNA.

• Each of the two DNA strands can act as a template for the synthesisof the other strand. A DNA double helix thus carries the same infor-mation in each of its strands.

• A DNA molecule is duplicated (replicated) by the polymerization ofnew complementary strands onto each of the old strands of the DNAdouble helix. This process of DNA replication, in which two identicalDNA molecules are formed from the original molecule, enables thegenetic information to be copied and passed on from cell to daugh-ter cell and from parent to offspring.

• As a DNA molecule replicates, its two strands are pulled apart toform one or more Y-shaped replication forks. The enzyme DNA poly-merase, situated in the fork, lays down a new complementary DNA

6:29Essential Concepts

Figure 6–39 The life cycle of a retrovirus. The retrovirus genome consists of an RNA molecule ofabout 8500 nucleotides; two such molecules are packaged into each viral particle. The enzymereverse transcriptase first makes a DNA copy of the viral RNA molecule and then a second DNAstrand, generating a double-stranded DNA copy of the RNA genome. The integration of this DNAdouble helix into the host chromosome is then catalyzed by a virus-encoded integrase enzyme (seeFigure –). This integration is required for the synthesis of new viral RNA molecules by the hostcell RNA polymerase, the enzyme that transcribes DNA into RNA (discussed in Chapter 7).

capsid

envelope

RNA

reversetranscriptase

ENTRY INTOCELL AND LOSSOF ENVELOPE

RNA

RNA

DNA

DNA

DNA

REVERSE TRANSCRIPTASEMAKES DNA/RNA ANDTHEN DNA/DNA DOUBLE HELIX

INTEGRATION OF DNA COPYINTO HOST CHROMOSOME integrated DNA

TRANSCRIPTION

manyRNAcopies

TRANSLATION

capsid protein

+

envelope protein

+

reverse transcriptase

ASSEMBLY OF MANYNEW VIRUS PARTICLES,EACH CONTAININGREVERSE TRANSCRIPTASE,INTO PROTEIN COATS

strand on each parental strand, thereby making two new double-hel-ical molecules.

• DNA polymerase replicates a DNA template with remarkable fidelity,making less than one error in every 107 bases read. This is possiblebecause the enzyme removes its own polymerization errors as itmoves along the DNA (proofreading).

• Since DNA polymerase can synthesize new DNA in only one direc-tion, only one of the strands in the replication fork, the leadingstrand, can be replicated in a continuous fashion. On the laggingstrand DNA is synthesized by the polymerase in a discontinuous“backstitching” process, making short fragments of DNA that arelater joined up by the enzyme DNA ligase to make a single continu-ous DNA strand.

• The proofreading feature of DNA polymerase makes it incapable ofstarting a new DNA chain. DNA synthesis is primed by an RNA poly-merase, called primase, that makes short lengths of RNA, calledprimers, that are subsequently erased and replaced with DNA.

• DNA replication requires the cooperation of many proteins, whichform a multienzyme replication machine, situated at the replicationfork, that catalyzes DNA synthesis.

• Errors in the replication of DNA and chemical reactions that damagethe nucleotides in DNA cause changes in the nucleotide sequence ofDNA. Such mutations, if not efficiently corrected, can be harmful tothe organism. Genetic information can be stored stably in DNAsequences only because a variety of DNA repair enzymes continu-ously scan the DNA and correct replication mistakes and replacedamaged nucleotides. DNA can be repaired easily because onestrand can be corrected using the other strand as a template.

• In eucaryotes, a special enzyme called telomerase replicates theDNA at the ends of the chromosomes.

• The rare copying mistakes that slip through the DNA replicationmachinery are dealt with by the mismatch repair proteins, whichmonitor newly replicated DNA and repair copying mistakes. Theoverall accuracy of DNA replication, including mismatch repair, isone mistake per 109 nucleotides copied.

• DNA damage caused by chemical reactions and ultraviolet irradia-tion is corrected by a variety of enzymes that recognize damagedDNA and excise a short stretch of the DNA strand that contains it.The missing DNA is resynthesized by a repair DNA polymerase thatuses the undamaged strand as a template. DNA ligase reseals theDNA to complete the repair process.

• Homologous recombination is the process by which two double-stranded DNA molecules of similar nucleotide sequence can crossover to create DNA molecules of novel sequence.

• Mobile genetic elements are DNA sequences that can move fromplace to place in the genomes of their hosts. This movement createschange in the host genomes and provides a source of genetic varia-tion.

• More than 50% of the human genome consists of DNA that isrepeated many times in the genome. Approximately two-thirds ofthis repeated DNA (about 34% of the total genome) consists of twoclasses of transposable elements that have multiplied to especiallyhigh copy numbers in the genome.

• Viruses are little more than genes packaged in protective coats. Theyrequire host cells in order to reproduce themselves.

• Some viruses have RNA instead of DNA as their genomes. One groupof RNA viruses—the retroviruses—must copy their RNA genomesinto DNA in order to replicate.

Chapter 6: DNA Replication, Repair, and Recombination6:30

6:31End-of-Chapter Questions

Key TermsDNA polymerase plasmidDNA repair proofreadingDNA replication replication forkgenetic recombination retrotransposonhomologous recombination retroviruslagging strand RNA (ribonucleic acid)leading strand site-specific recombinationmobile genetic element telomerasemutation templateOkazaki fragment transposon

virus

Questions

Question 6–3/6–6

DNA repair enzymes preferentially repair mismatchedbases on the newly synthesized DNA strand, using theold DNA strand as a template. If mismatches were sim-ply repaired without regard for which strand served astemplate, would this reduce replication errors?Explain your answer.

Question 6–7/6–7

Suppose a mutation affects an enzyme that is requiredto repair the damage to DNA caused by the loss ofpurine bases. This mutation causes the accumulationof about 5000 mutations in the DNA of each of yourcells per day. As the average difference in DNAsequence between humans and chimpanzees is about1%, how long will it take for you to turn into a monkey?What is wrong with this argument?

Question 6–8/6–8

Which of the following statements are correct? Explainyour answers.

A. The replication fork is asymmetrical because itcontains two DNA polymerase molecules thatare structurally distinct.

B. Okazaki fragments are removed by an RNAnuclease.

C. The error rate of DNA replication is reducedboth by proofreading of the DNA polymeraseand by DNA repair enzymes.

D. In the absence of DNA repair, genes are unsta-ble.

E. None of the aberrant bases formed by deami-nation occur naturally in DNA.

F. Cancer results from uncorrected mutations insomatic cells.

Question 6–14/6–9

Being a born skeptic, you plan to confirm for yourselfthe results of a classic experiment originally per-formed in the 1960s by Meselson and Stahl from whichthey concluded that each daughter cell inherits oneand only one strand of its mother’s DNA. To do so, you“synchronize” (using established methods that neednot concern us here) a culture of growing cells, so thatvirtually all cells in your flask begin and then completeDNA synthesis at the same time. Your cells are firstgrown in a normal growth medium and then, after oneround of DNA synthesis, grown further in a speciallyconcocted (and very expensive) growth medium thatcontains nutrients highly enriched in heavy isotopesof nitrogen and carbon (15N and 13C in place of thenaturally abundant 14N and 12C). Cells growing on thismedium use the heavy isotopes to build all of theirmacromolecules, including nucleotides and nucleicacids. You then isolate DNA from cells that have grownfor a different number of generations in the heavy-iso-tope medium and analyze the DNA for its densityusing a gradient centrifugation technique (see Panel4–4, pp. –). The more heavy isotopes have beenbuilt into the DNA, the heavier it appears in this analy-sis. Your data, plotting the amount of DNA isolatedover its density, are shown in Figure Q6–9. Are theseresults in agreement with your expectations? Explainthe results.

Question 6–15/6–10

The speed of DNA replication at a replication fork isabout 100 nucleotides per second in human cells.What is the minimum number of origins of replicationthat a human cell must have in order to replicate itsDNA once every 24 hours? Recall that a human cellcontains two copies of the human genome, one inher-ited from the mother, the other from the father, eachconsisting of 3 ¥ 109 nucleotide pairs.

Chapter 6: DNA Replication, Repair, and Recombination6:32

Question 6–16/6–11

Look carefully at the structures of the compoundsshown in Figure Q6–11. One or the other of the twocompounds is added to a DNA replication reaction.

A. What would you expect if compound A wereadded in large excess over the concentration ofthe available deoxycytosine triphosphate(dCTP)?

B. What would happen if it were added at 10% ofthe concentration of the available dCTP?

C. What effects would you expect if compound Bwere added under the same conditions?

Question 6–17/6–12

The genetic material of a hypothetical organism isstructurally indistinguishable from DNA of normalcells. Surprisingly, analyses reveal that the DNA is syn-thesized from nucleoside triphosphates that containfree 5¢-hydroxyl groups and triphosphate groups at the3¢ position. In what way must this organism’s DNApolymerase differ from that of normal cells? Could itstill proofread?

Question 6–18/6–13

Figure Q6–13 shows a snapshot of a replication fork inwhich the RNA primer has just been added to the lag-ging strand. Using this diagram as a guide, sketch thepath of the DNA as the next Okazaki fragment is syn-thesized. Indicate the sliding clamp and the single-strand binding protein as appropriate.

Question 6–19/6–14

Approximately how many high-energy bonds are usedto replicate a bacterial chromosome? How much glu-cose (compared with its own weight of 10–12 g) does abacterium need to consume to provide enough energyto copy its DNA once? The number of base pairs in thebacterial chromosome is 3 ¥ 106. Oxidation of one glu-cose molecule yields about 30 high-energy phosphatebonds. The molecular weight of glucose is 180 g/mole.(Recall that there are 6 ¥ 1023 molecules in a mole; seeChapter 2.)

Question 6–20/6–15

What, if anything, is wrong with the following state-ment: “Both reproductive-cell DNA stability andsomatic-cell DNA stability are essential for the survivalof a species.” Explain your answer.

Question 6–21/6–16

A common type of error in DNA is produced by a spon-taneous reaction termed deamination in which anucleotide base loses an amino group (NH2), which isreplaced by a keto group (C=O) by the general reaction

amo

un

t o

f D

NA 1

starting cells

second generation third generation

first generation grownin heavy isotopes

light heavydensity

1

2

2

2 4

6

Figure Q6–9

O

O–

–O P

O

O–

PO

O

O

O–

PO O

O

O

O–

–O P CH2

NH2

H H

ON

N

OCH2

NH2

H H

ON

N

(A) (B)

dideoxycytosinetriphosphate

dideoxycytosinemonophosphateFigure Q6–11

next primer

Figure Q6–13

6:33End-of-chapter Questions

shown in Figure Q6–16. Write the structures of thebases A, G, C, T, and U and predict the products thatwill be produced by deamination. By looking at theproducts of this reaction—and remembering that, inthe cell, these will need to be recognized andrepaired—can you propose an explanation why DNAcannot contain uracil?

Question 8–9/6–17

A. Explain why telomeres and telomerase areneeded for replication of eucaryotic chromo-somes but not for replication of a circular bac-terial chromosome. Draw a diagram to illus-trate your explanation.

B. Would you still need telomeres and telomeraseto complete eucaryotic chromosome replica-tion if DNA primase always laid down the RNAprimer at the very 3¢ end of the template for thelagging strand?

Question 9–16/6–18

Discuss the following statement: “Viruses exist in thetwilight zone of life: outside cells they are simply deadassemblies of molecules; inside cells, however, theyare alive.”

NC

NH2

NH3

H2O

NC

O

H

Figure Q6–16