Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

A peptide fragment generated by a specific cleavage

process may still be too large to sequence. In that case, af-

ter its purification, it can be subjected to a second round of

fragmentation using a different cleavage process.

E. Separation and Purification of the

Peptide Fragments

Once again we must employ separation techniques, this

time to isolate the peptide fragments of specific cleavage

operations for subsequent sequence determinations. The

nonpolar residues of peptide fragments are not excluded

from the aqueous environment as they are in native pro-

teins (Chapter 8). Consequently, many peptide fragments

aggregate, precipitate, and/or strongly adsorb to chromato-

graphic materials, which can result in unacceptable peptide

losses. Until around 1980, the trial-and-error development

of methods that could satisfactorily separate a mixture of

peptide fragments constituted the major technical chal-

lenge of a protein sequence determination, as well as its

most time-consuming step. Such methods involved the use

of denaturants, such as urea and SDS, to solubilize the pep-

tide fragments, and the selection of chromatographic mate-

rials and conditions that would reduce their adsorptive

losses. The advent of reverse-phase chromatography by

HPLC (Section 6-3Dh), however, has largely reduced the

separation of peptide fragments to a routine procedure.

F. Sequence Determination

Once the manageably sized peptide fragments that were

formed through specific cleavage reactions have been iso-

lated, their amino acid sequences can be determined. This

is done through repeated cycles of the Edman degradation

(Section 7-1Aa). An automated device for doing so was

first developed by Edman and Geoffrey Begg. In modern

sequencers, the peptide sample is adsorbed onto a

polyvinylidene fluoride (PVDF) membrane or dried onto

glass fiber paper which is impregnated with polybrene (a

polymeric quaternary ammonium salt). In either case, the

peptide is immobilized but is readily accessible to Edman

reagents. Accurately measured quantities of reagents, ei-

ther in solution or as vapors in a stream of argon (which

minimizes peptide loss), are then delivered to the reaction

cell at programmed intervals.The thiazolinone-amino acids

are automatically removed, converted to the correspon-

ding PTH-amino acids (Fig. 7-4), and identified via HPLC.

Such instruments are capable of processing up to one

residue per hour.

A peptide’s 40 to 60 N-terminal residues can usually be

identified (100 or more with the most advanced systems) be-

fore the cumulative effects of incomplete reactions, side reac-

tions, and peptide loss make further amino acid identifica-

tion unreliable. As little as 0.1 pmol of a PTH-amino acid

can be reliably identified by the UV detector–equipped re-

verse-phase HPLC systems used in advanced sequencers.

Consequently, a peptide’s N-terminal 5 to 25 residues can,

respectively, be determined with as little as 1 to 10 pmol of

the peptide—invisibly small amounts.

G. Ordering the Peptide Fragments

With the peptide fragments individually sequenced, what

remains is to elucidate the order in which they are con-

nected in the original polypeptide. We do so by comparing

the amino acid sequences of one set of peptide fragments

with those of a second set whose specific cleavage sites over-

lap those of the first set (Fig. 7-6). The overlapping peptide

segments must be of sufficient length to identify each

cleavage site uniquely, but as there are 20 possibilities for

each amino acid residue, an overlap of only a few residues

is usually enough.

Section 7-1. Primary Structure Determination of Proteins 171

Figure 7-6 The amino acid sequence of a polypeptide chain is

determined by comparing the sequences of two sets of mutually

overlapping peptide fragments. In this example, the two sets of

peptide fragments are generated by cleaving the polypeptide

after all its Arg and Lys residues with trypsin and, in a separate

reaction, after all its Met residues by treatment with cyanogen

bromide. The order of the first two tryptic peptides is established,

CNBr

fragments

CNBr

Trp Met Gly Ala Lys Leu Pro Met Asp Gly Arg Cys Ala GlnPhe

CNBr

Trypsin

fragments

trypsin

for example, by the observation that the Gly-Ala-Lys-Leu-Pro-Met

cyanogen bromide peptide has its N- and C-terminal sequences

in common with the C- and N-termini, respectively, of the two

tryptic peptides. In this manner the order of the peptide fragments

in their parent polypeptide chain is established.

See the

Animated Figures

JWCL281_c07_163-220.qxd 2/22/10 9:11 PM Page 171

charge).Yet, until about 1985, macromolecules such as pro-

teins and nucleic acids could not be analyzed by MS. This

was because the method by which mass spectrometers pro-

duced gas phase ions destroyed macromolecules: vaporiza-

tion by heating followed by ionization via bombardment

with electrons. However, the development of two tech-

niques has eliminated this roadblock:

1. Electrospray ionization (ESI; Fig. 7-7a), a technique

pioneered by John Fenn in which a solution of a macromol-

ecule such as a peptide is sprayed from a narrow capillary

tube maintained at high voltage (⬃4000 V), forming fine,

highly charged droplets from which the solvent rapidly

evaporates. This yields a series of gas phase macromolecu-

lar ions that typically have ionic charges in the range 0.5

to 2 per kilodalton. For polypeptides, the ionic charges

result from the protonation of basic side chains such as Lys

and Arg [(M  nH)

n

ions].

2. Matrix-assisted laser desorption/ionization (MALDI;

Fig. 7-7b), in which the macromolecule is embedded in a

crystalline matrix of a low molecular mass organic mole-

cule (usually prepared by drying a droplet of solution con-

taining the macromolecule and a large excess of the

organic molecule) and irradiated with intense short

172 Chapter 7. Covalent Structures of Proteins and Nucleic Acids

H. Assignment of Disulfide Bond Positions

The final step in an amino acid sequence analysis is to de-

termine the positions (if any) of the disulfide bonds. This

is done by cleaving a sample of the native protein under

conditions that leave its disulfide bonds intact. The re-

sulting peptide fragments are then separated by reverse-

phase HPLC.The disulfide-linked peptide fragments are

easily identified because, for each such linkage, two of

the original peptide fragments will be replaced by a

larger one. The disulfide-linked fragments are then sub-

jected to Edman degradation. Although such a fragment

yields two PTH-amino acids in each step of this process

(at least initially), their locations within the predeter-

mined amino acid sequence of the protein are readily

surmised, thereby establishing the positions of the disul-

fide bonds.

I. Peptide Characterization and Sequencing

by Mass Spectrometry

Mass spectrometry (MS) has emerged as an important

technique for characterizing and sequencing polypeptides.

MS accurately measures the mass-to-charge (m/z) ratio for

ions in the gas phase (where m is the ion’s mass and z is its

Figure 7-7 Generation of the gas phase ions required for the

mass spectrometric analysis of proteins. (a) By electrospray

ionization (ESI) and (b) by matrix-assisted laser

desorption/ionization (MALDI). In ESI, a stream of dry N

Electrospray ionization (ESI)

ESI mass spectrum

m/z

MALDI mass spectrum

m/z

Detector

Mass analyzer

(a)

4000V

Atmospheric

pressure

Low

vacuum

Laser

pulse

Sample

matrix

30,000 V

High vacuum

Matrix-assisted laser desorption/ionization (MALDI)

Detector

Mass analyzer

(b)

High vacuum

Sample

solution

Capillary

some other gas is used to promote the evaporation of the solvent

from the droplets. [After Fitzgerald, M.C. and Siuzdak, G., Chem.

Biol. 3, 708 (1996).]

JWCL281_c07_163-220.qxd 2/22/10 9:11 PM Page 172

(nanosecond) pulses of laser light at a wavelength ab-

sorbed by the matrix material but not the macromolecule.

The energy absorbed by the matrix ejects the intact

macromolecules from its surface into the gas phase, usu-

ally with a charge of 1 but, with larger molecules, occa-

sionally with charges of 2, 3, etc. For polypeptides, gen-

tisic acid (2,5-dihydroxybenzoic acid) is one of the few

substances found to have satisfactory properties as a ma-

trix.Through the use of MALDI, polypeptides of 400 kD

have been characterized.

In both of these techniques, the gas phase macromolec-

ular ions are directed into the mass spectrometer, which

measures their m/z values with an accuracy of 0.01%.

Consequently, if an ion’s z value can be determined, its mo-

lecular mass can be determined with far greater accuracy

than by any other method. For example, Fig. 7-8 shows the

ESI-based mass spectrum (ESI-MS) of the 16,951-D pro-

tein myoglobin. Note that successive peaks in this spec-

trum differ by a single ionic charge with the rightmost peak

corresponding to an (M  9H)

9

ion. Consequently, for the

mass spectrum of a macromolecule of molecular mass M

containing two adjacent peaks with m/z values of p

and p

arising from ions with charges of z

and z

 1,

[7.1]p



M  z

and

[7.2]

These two linear equations can therefore be readily solved

for their two unknowns, M and z

Since most mass spectrometers are limited to detecting

ions with m/z values less than several thousand, the use of

ESI-MS has the advantage that the high ionic charges of

the ions it produces has permitted the analysis of com-

pounds with molecular masses 100 kD. Another advan-

tage of ESI-MS is that it can be configured to operate in a

continuous flow mode with the effluent of an HPLC or

CE system. ESI-MS is used in this way, for example, to

characterize the tryptic digest of a protein by determining

the molecular masses of its component peptides (Section

7-1J).

a. Peptide Sequencing by Mass Spectrometry

Short polypeptides (25 residues) can be directly se-

quenced though the use of a tandem mass spectrometer

(MS/MS; two mass spectrometers coupled in series; Fig. 7-9).

The first mass spectrometer functions to select the peptide

ion of interest from other peptide ions as well as any con-

taminants that may be present.The selected peptide ion (P

in Fig. 7-9b) is then passed into a collision cell, where it



M  z

 1

Section 7-1. Primary Structure Determination of Proteins 173

Figure 7-8 The ESI-MS spectrum of the 16,951-D horse heart

protein apomyoglobin. The measured m/z ratios and the inferred

charges for most of the peaks are indicated. Note the bell-shaped

distribution of the peaks, which is typical of ESI-MS spectra. The

peaks all have shoulders because the polypeptide’s component

Relative abundance

100

678.6

+25

738.1

+23

808.3

+21

848.7

+20

893.3

+19

998.4

+17

1060.5

+16

1211.8

+14

1414.0

+12

1542.3

+11

1696.3

+10

1884.7

600 800 1000 1200 1400 1600 1800 2000

m/z

elements contain small admixtures of heavier isotopes (e.g.,

naturally abundant carbon consists of 98.9%

C and 1.1%

and naturally abundant sulfur consists of 0.8%

S, 4.2%

S, and

95.0%

S). [After Yates, J.R., Methods Enzymol. 271, 353

(1996).]

JWCL281_c07_163-220.qxd 2/22/10 9:11 PM Page 173

collides with chemically inert atoms such as helium. The

energy thereby imparted to a peptide ion causes it to frag-

ment predominantly at only one of its several peptide

bonds, yielding one or two charged fragments (Fig. 7-10).

The molecular masses of the charged fragments are then

determined by the second mass spectrometer.

By comparing the molecular masses of successively

larger members of a family of fragments, the molecular

masses and therefore the identities of the corresponding

amino acid residues can be determined.The sequence of an

entire polypeptide can thus be elucidated (although MS

cannot distinguish the isomeric residues Ile and Leu be-

cause they have exactly the same mass and cannot always

reliably distinguish Gln and Lys residues because their mo-

lecular masses differ by only 0.036 D). Computerization of

this comparison process has reduced the time required to

sequence a (short) polypeptide to only a few minutes as

compared to the 30 to 50 min required per cycle of Edman

degradation. The reliability of this process has been in-

creased through the computerized matching of a measured

mass spectrum with those of peptides of known sequence

as maintained in databases.

The sequences of several polypeptides in a mixture can

be determined, even in the presence of contaminants, by

sequentially selecting the corresponding polypeptide ions

in the first mass spectrometer of the tandem instrument.

Hence, in separating and purifying the polypeptide frag-

ments of a protein digest in preparation for their sequenc-

ing,less effort need be expended for MS/MS-based as com-

pared to Edman techniques. MS/MS can also be used to

sequence peptides with chemically blocked N-termini (a

common eukaryotic post-translational modification that

prevents Edman degradation) and to characterize other

post-translational modifications such as phosphorylations

(Section 4-3A) and glycosylations (Section 11-3C). Finally,

MS/MS can be used to locate disulfide bonds by chemically

or enzymatically fragmenting both the disulfide-cleaved

protein and the intact protein. From the disappearance of

peaks from the mass spectrum of the latter relative to that

of the former and the appearance of new peaks, the posi-

tions of the disulfide bonds in the protein can be deduced.

Thus, MS/MS has become an indispensable tool for the

characterization of polypeptides.

J. Peptide Mapping

Once the primary structure of a protein has been eluci-

dated, that of a nearly identical protein, such as one arising

from a closely related species, a mutation, or a chemical

modification, can be more easily determined. This was

originally done through the combined paper chromatography

and paper electrophoresis (Section 6-4A) of partial protein

digests, a technique synonymously known as fingerprinting

or peptide mapping. The peptide fragments incorporating

the amino acid variations migrate to different positions on

their fingerprint (peptide map) than do the corresponding

peptides of the original protein (Fig. 7-11).The variant pep-

tides could then be eluted and sequenced, thereby identify-

ing the changes in the protein without the need to

sequence it in its entirety.

In more recent times, peptide mapping has come to

mean any method that fragments a protein in a repro-

174 Chapter 7. Covalent Structures of Proteins and Nucleic Acids

MS-1Electrospray

ionization

source

Electrospray ionization tandem mass spectrometer(a)

(b)

Collision cell

MS-1

MS-2

Collision

cell

Ion

source

Detector

MS-2 Detecto

Figure 7-9 The use of a tandem mass spectrometer (MS/MS)

in amino acid sequencing. (a) An MS/MS consists of an ion

source (here shown as an ESI system), a first mass spectrometer

(MS-1), a collision cell, a second mass spectrometer (MS-2), and

a detector. (b) The ion source generates gas phase peptide ions,

, etc., from a digest of the protein being analyzed.These

peptides are separated by MS-1 according to their m/z values

and one of them, here P

, is directed into the collision cell, where

it collides with helium atoms.This treatment induces the

breakdown of the polypeptide ion to yield the fragments F

etc., which are directed into MS-2, where their m/z values are

determined. [Part a after Yates, J.R., Methods Enzymol. 271, 358

(1996); Part b after Biemann, K. and Scoble, H.A., Science 237,

992 (1987).]

JWCL281_c07_163-220.qxd 2/22/10 9:11 PM Page 174

ducible manner and then separates the resulting peptides

to yield a pattern that can be used to distinguish differ-

ences between related proteins. Thus, peptide mapping can

be carried out by two-dimensional gel electrophoresis or

by such high resolution one-dimensional techniques as

HPLC, SDS–PAGE, IEF, or CE (Sections 6-3 and 6-4).

With any of these methods, variant peptides can be isolated

and sequenced to establish the sequence differences be-

tween the related proteins.

Section 7-1. Primary Structure Determination of Proteins 175

100

Relative abundance

173.7

245.7

332.9

479.1

684.2

784.8

813.0

942.7

1054.5

1171.2

1286.0

1325.1

1396.8

627.8

642.6

AsnValGlyGlu Asp Asn Glu Glu Gly Phe Phe Ser

400

1172

286

1286

187

1385

130

1442

515

1057

629

943

758

813

887

684

944

627

1092

480

1239

333

1326

246

1397

175

Ala Arg

(a)

(b)

200 400 600 800 1000 1200 1400

m/z

(a)

Figure 7-11 Peptide mapping. A comparison of the fingerprints

of trypsin-digested (a) hemoglobin A (HbA) and (b) hemoglobin

S (HbS) shows two peptides that differ in these two forms of

hemoglobin (boxes).These peptides constitute the eight

N-terminal residues of the  subunit of hemoglobin.Their amino

acid sequences are

Figure 7-10 The tandem mass

spectrum of the doubly charged ion of

the 14-residue human [Glu

]

fibrinopeptide B (m/z ⫽ 786). (a) The

peptide’s sequence.The upper and

lower rows of numbers are the

molecular masses of the charged

N-terminal and C-terminal fragments,

respectively, that are formed by the

cleavage indicated by their connecting

diagonal line. (b) The mass spectrum

of the fragmented peptide with the

m/z values of the most abundant

fragments indicated above the

corresponding peaks.The energy of

the collisions in the collision cell has

been adjusted so that each peptide

ion fragments an average of only

once. Note that the predominant

fragments under these conditions

have z  1 and contain the intact

peptide’s C-terminus. [After Yates,

J.R., Methods Enzymol. 271, 354

(1996).]

(b)

Hemoglobin A Val—His—Leu—Thr—Pro—Glu—Glu—Lys

Hemoglobin S Val—His—Leu—Thr—Pro—Val —Glu—Lys

12 3 4 5 6 78

[Courtesy of Corrado Baglioni, State University of New York at

Albany.]

JWCL281_c07_163-220.qxd 2/22/10 9:11 PM Page 175

2 NUCLEIC ACID SEQUENCING

The basic strategy of nucleic acid sequencing is identical to

that of protein sequencing (Section 7-1). It involves

1. The specific degradation and fractionation of the

polynucleotide of interest to fragments small enough to be

fully sequenced.

2. The sequencing of the individual fragments.

3. The ordering of the fragments by repeating the pre-

ceding steps using a degradation procedure that yields a set

of polynucleotide fragments that overlap the cleavage

points in the first such set.

Before about 1975, however, nucleic acid sequencing tech-

niques lagged far behind those of protein sequencing,

largely because there were no available endonucleases that

were specific for sequences greater than a nucleotide.

Rather, RNAs were cleaved into relatively short fragments

by partial digestion with enzymes such as ribonuclease T1

(from Aspergillus oryzae), which cleaves RNA after gua-

nine residues, or pancreatic ribonuclease A, which does so

after pyrimidine residues. Moreover, there is no reliable

polynucleotide reaction analogous to the Edman degrada-

tion for proteins (Section 7-1A). Consequently, the polynu-

cleotide fragments were sequenced by their partial diges-

tion with either of two exonucleases: snake venom

phosphodiesterase, which removes residues from the 3¿

end of polynucleotides (Fig. 7-12), or spleen phosphodi-

esterase, which does so from the 5¿ end. The resulting

oligonucleotide fragments were identified from their chro-

matographic and electrophoretic mobilities. Sequencing

RNA in this manner is a lengthy and painstaking

procedure.

The first biologically significant nucleic acid to be se-

quenced was that of yeast alanine tRNA (Section 32-2A).

The sequencing of this 76-nucleotide molecule by Robert

Holley, a labor of 7 years, was completed in 1965, some 12

years after Frederick Sanger had determined the amino

acid sequence of insulin.This was followed, at an accelerat-

ing pace, by the sequencing of numerous species of tRNAs

and the 5S ribosomal RNAs (Section 32-3A) from several

organisms.The art of RNA sequencing by these techniques

reached its zenith in 1976 with the sequencing, by Walter

Fiers, of the entire 3569-nucleotide genome of the bacterio-

phage MS2. In contrast, DNA sequencing was in a far more

primitive state because of the lack of available DNA en-

donucleases with any sequence specificity.

After 1975, dramatic progress was made in nucleic acid

sequencing technology.Three advances made this possible:

1. The discovery of restriction endonucleases to enable

the cleavage of DNA at specific sequences (Section 5-5A).

2. The development of molecular cloning techniques to

permit the acquisition of almost any identifiable DNA seg-

ment in the amounts required for sequencing (Section 5-5).

3. The development of DNA sequencing techniques.

These procedures are largely responsible for the enormous

advances in our understanding of molecular biology that

have been made over the past three decades and which we

discuss in succeeding chapters. DNA sequencing tech-

niques are the subject of this section.

The pace of nucleic acid sequencing has become so

rapid that directly determining a protein’s amino acid se-

quence is far more time-consuming than determining the

base sequence of its corresponding gene (although amino

acid and base sequences provide complementary informa-

tion; Section 7-2D). There has been such a flood of DNA

sequence data—over 300 billion nucleotides in over 200

million sequences as of 2010, and doubling every ⬃18

months—that only computers can keep track of them. The

first complete genome sequence to be determined, that of

the gram-negative bacterium Haemophilus influenzae, was

reported in 1995 by J. Craig Venter. By 2010, the genome

sequences of over 1000 prokaryotes had been reported

(with many more being determined) as well as sequences

of over 120 eukaryotes (with many more in progress), in-

cluding those of humans and many other vertebrates, in-

sects, worms, plants, and fungi (Table 7-3).

A. The Sanger Method

See Guided Exploration 5: DNA sequence determination by the

chain-terminator method

After 1975, several methods were de-

veloped for the rapid sequencing of long stretches of DNA.

Here we discuss the Sanger method, formulated by Freder-

ick Sanger (the same individual who pioneered the amino

acid sequencing of proteins), which is mainly responsible

for the vast number of DNA sequences that have been elu-

cidated.

The Sanger method (alternatively called the chain-

terminator method and the dideoxy method) utilizes the E.

coli enzyme DNA polymerase I (Section 5-4Cc) to synthe-

size complementary copies of the single-stranded DNA

176 Chapter 7. Covalent Structures of Proteins and Nucleic Acids

Figure 7-12 Sequence determination of an oligonucleotide by

partial digestion with snake venom phosphodiesterase. This

enzyme sequentially cleaves the nucleotides from the 3¿ end of a

polynucleotide that has a free 3¿-OH group. Partial digestion of

an oligonucleotide with snake venom phosphodiesterase yields a

mixture of fragments of all lengths, as indicated, that may be

chromatographically separated. Comparison of the base

compositions of pairs of fragments that differ in length by one

nucleotide establishes the identity of the 3¿-terminal nucleotide

of the larger fragment. In this way the base sequence of the

oligonucleotide may be elucidated.

GCACUUGA

GCACUUG

GCACUU

GCACU

GCAC

GCA

+ Mononucleotides

snake venom

phosphodiesterase

JWCL281_c07_163-220.qxd 2/22/10 9:11 PM Page 176

being sequenced. As we have previously seen, under the

direction of the strand being replicated (the template

strand), DNA polymerase I assembles the four deoxynu-

cleoside triphosphates (dNTPs), dATP, dCTP, dGTP, and

dTTP, into a complementary polynucleotide chain that it

elongates in the 5¿ to 3¿ direction (Fig. 5-31). To initiate

DNA synthesis, DNA polymerase requires a primer in a sta-

ble base paired complex with the template DNA. If the

DNA being sequenced is a restriction fragment, as it usually

is, it begins and ends with a restriction site. The primer can

therefore be a short DNA segment containing this restric-

tion site annealed to the strand being replicated. The tem-

plate DNAs are obtained in sufficient quantity to sequence

by cloning them in M13-based vectors (Section 5-5Bb) or

by PCR (Section 5-5F), both of which yield the required

single-stranded DNAs.

DNA polymerase I’s 5¿S3¿ exonuclease activity (Fig.

5-33) is catalyzed by a separate active site from those

which mediate its polymerase and 3¿ S 5¿ exonuclease

(Fig. 5-36) functions. This is demonstrated by the observa-

tion that on proteolytic cleavage of the enzyme into two

fragments, the larger C-terminal fragment, which is known

as the Klenow fragment, possesses the full polymerase and

3¿ S 5¿ exonuclease activities of the intact enzyme,

whereas the smaller N-terminal fragment has its 5¿ S 3¿

exonuclease activity. Only the Klenow fragment is used in

DNA sequencing to ensure that all replicated chains have

the same 5¿ terminus.

a. The Synthesis of Labeled DNA by DNA

Polymerase Is Terminated after Specific Bases

In the Sanger method (Fig. 7-13), the DNA to be sequenced

is incubated with the Klenow fragment of DNA polymerase I,

a suitable primer, and the four deoxynucleoside triphosphates

(dNTPs). Either at least one dNTP (usually dATP) or the

primer is ␣-

P-labeled. In addition, a small amount of the

2ⴕ,3ⴕ-dideoxynucleoside triphosphate (ddNTP)

of one of the bases is added to the reaction mixture. When

the dideoxy analog is incorporated in the growing polynu-

cleotide in place of the corresponding normal nucleotide,

chain growth is terminated because of the absence of a

3¿-OH group. By using only a small amount of the ddNTP,

a series of truncated chains is generated, each of which is ter-

minated by the dideoxy analog at one of the positions occu-

pied by the corresponding base. Each of the four ddNTPs is

reacted in a separate vessel.

The four reaction mixtures are simultaneously elec-

trophoresed in parallel lanes on a sequencing gel. This is a

2,3-Dideoxynucleoside

triphosphate

OCH

Base

Section 7-2. Nucleic Acid Sequencing 177

Table 7-3 Some Sequenced Genomes

Haploid Genome Number of

Organism Size (kb) Chromosomes

Carsonella ruddii (insect endosymbiont; smallest known cellular genome) 160 1

Mycoplasma genitalium (human parasite) 580 1

Rickettsia prowazekii (bacterium; cause of typhus; putative mitochondial relative) 1,112 1

Haemophilus influenzae (bacterium; human pathogen) 1,830 1

Archaeoglobus fulgidus (hyperthermophilic, sulfate-reducing archaeon) 2,178 1

Synechocystis sp. (cyanobacterium) 3,573 1

Mycobacterium tuberculosis (cause of tuberculosis) 4,412 1

Escherichia coli (bacterium, human symbiont) 4,639 1

Saccharomyces cerevisiae (baker’s or budding yeast) 12,070 16

Plasmodium falciparum (protozoan; cause of malaria) 23,000 14

Caenorhabditis elegans (nematode worm) 97,000 6

Drosophila melanogaster (fruit fly) 180,000 4

Arabidopsis thaliana (flowering plant) 119,200 5

Oryza sativa (rice) 389,000 12

Danio rerio (zebra fish) 1,700,000 25

Gallus gallus (chicken) 1,200,000 40

Ornithorhynchus anatinus (platypus) 1,840,000 31

Mus musculus (mouse) 2,500,000 20

Canis familiaris (dog) 2,400,000 40

Pan troglodytes (chimpanzee) 3,100,000 24

Homo sapiens (human) 3,038,000 23

Source: http://www.ncbi.nlm.nih.gov/sites/entrez?db=genome.

JWCL281_c07_163-220.qxd 2/22/10 9:11 PM Page 177

long, thin (as little as 0.1 mm by up to 100 cm) polyacry-

lamide slab. It contains ⬃7M urea and is run at ⬃70°C so as

to eliminate all hydrogen bonding associations. These con-

ditions ensure that the DNA fragments separate only ac-

cording to their size. The sequence of the DNA that is com-

plementary to the template DNA can then be directly read

off an autoradiogram of the sequencing gel, from bottom to

top, as is indicated in Fig. 7-14. Indeed, computerized de-

vices are available to aid in doing so. However, a single gel

is incapable of resolving much more than 300 to 400 con-

secutive fragments. This limitation is circumvented by gen-

erating two sets of gels, one run for a longer time and per-

haps at a higher voltage than the other, to obtain the

sequence of up to 800-bp DNA fragments.

Improved gels can be obtained through the use of

dNTPs whose ␣-phosphate groups are radioactively la-

beled with

S rather than

This is because the ␤ particles emitted by

S nuclei have

⫺

OP POPO

␣-Thio-[

S]dNTP

Base

⫺

178 Chapter 7. Covalent Structures of Proteins and Nucleic Acids

Figure 7-13 Flow diagram of the

Sanger (chain-terminator or dideoxy)

method of DNA sequencing. The

symbol ddATP represents

dideoxyadenosine triphosphate, etc.

The sequence that is determined by

reading the gel from bottom to top

(from the smallest to the largest

fragment) is complementary to the

sequence of the template DNA.

Figure 7-14 Autoradiograph of a sequencing gel. DNA

fragments were produced by the Sanger method of DNA

sequencing.A second loading of the gel (right) was made 90 min

after the initial loading (left).The deduced sequence of 140

nucleotides is written alongside. [From Hindley, J., DNA

sequencing, in Work, T.S. and Burdon, R.H. (Eds.), Laboratory

Techniques in Biochemistry and Molecular Biology, Vol. 10, p. 82,

Elsevier (1983). Used by permission.]

dATP + ddATP

dCTP

dGTP

dTTP

dATP

dCTP + ddCTP

dGTP

dTTP

dATP

dCTP

dGTP + ddGTP

dTTP

dATP

dCTP

dGTP

dTTP + ddTTP

CCGGTAGCAACT

GGCCA

GGCCATCGTTGA

GGC

GGCC

GGCCATCG

GGCCATCGTTG

GGCCATC

GGCCAT

GGCCATCGT

GGCCATCGTT

CAGT

Template:

Primer:

Sequence complementar

to template DNA

5′

3′

5′ 3′

5′3′

140

GAACT GCT

JWCL281_c07_163-220.qxd 7/20/10 5:30 PM Page 178

less energy and hence shorter path lengths than those of

P, thereby yielding sharper gel bands. More readily inter-

pretable gels may also be obtained by replacement of the

Klenow fragment with DNA polymerases either from bac-

teriophage T7 (T7 DNA polymerase, which is less sensitive

to the presence of ddNTPs than is Klenow fragment and

hence yields gel bands of more even intensities) or from

thermophilic bacteria such as Thermus aquaticus (Taq

polymerase; Section 5-5F) that are stable above 90°C and

hence can be used at the temperatures required to dena-

ture particularly stable segments of DNA.

b. RNA May Be Sequenced through Its Transcription

to cDNA

RNA can be readily sequenced by only a slight modifi-

cation of the above DNA sequencing procedures. The

RNA to be sequenced is transcribed into a complementary

strand of DNA (cDNA) through the action of reverse tran-

scriptase (Section 5-5Fa).The resulting cDNA may then be

sequenced normally.

c. The Sanger Method Has Been Automated

In order to sequence large tracts of DNA such as entire

chromosomes, the Sanger method has been greatly acceler-

ated through automation. This required that the above-

described radiolabeling techniques, which are not readily

automated, be replaced by fluorescence labeling techniques

(with the added benefit of eliminating the health hazards

and storage problems of using radiolabeled nucleotides).

In the most widely used such technique, each of the four

ddNTPs used to terminate chain extension is covalently

linked to a differently fluorescing dye, the chain-extension

reactions are carried out in a single vessel containing all

four of these labeled ddNTPs, and the resulting fragment

mixture is subjected to sequencing gel electrophoresis in a

single lane (Fig. 7-15a). As each fragment exits the gel, its

terminal base is identified according to its characteristic

fluorescence spectrum by a laser-activated fluorescence

detection system (Fig. 7-15b).

Section 7-2. Nucleic Acid Sequencing 179

Figure 7-15 Automated DNA sequencing using the Sanger

method. (a) The primer that is base paired to the template strand

being sequenced is extended in the presence of four differently

fluorescently labeled dideoxynucleoside triphosphates

(ddNTPs).The resulting mixture of dye-labeled DNA strands of

all lengths is subject to gel electrophoresis in a capillary tube,

thereby separating them according to size with smaller

polynucleotides migrating faster than larger ones.As each

polynucleotide passes the detector, it’s 3¿-terminal nucleotide is

identified according to the spectrum of its laser-stimulated

dATP + dCTP + dGTP + dTTP +

ddATP ddCTP++

ddGTP ddTTP+

CCGGTAGCAACT

GGCCATCGTTGA

GGCCATCGTTG

GGCCATCGTT

GGCCATCGT

GGCCATCG

GGCCATC

GGCCAT

GGCCA

GGCC

GGC

Template:

(a)

DNA

polymerase

Dye-labeled

DNA segments

Dye-labeled DNA segments are

subjected to electrophoresis

in a capillary gel

DNA

migration

Primer:

5′ 3′

5′3′

DetectorLaser

G G G G G G GG G G G G GG GG G G G G GG

200 210 220 230 240 250 260 270 280

AAA A A A A A AA A A A A AAA AAA AA A AAAACC

C CCC CCCC C C CC C CC CC CTTT TTTTT T TT T TT TTTTTTT

GGGGGGG N

120 130 140 150 160 170 180 1901100

GG GG G G G G G GGG G G GG G GG G GAAA A AA AA AAA AAA AAC C CC CC C C C C C CCC C C C C C C

CCTT T T T T T TT T T TTT T TT TTT TT T T T T T TTTT

(b)

fluorescence. (b) A portion of the output of an automated DNA

sequencing system. Each of the four differently colored curves

indicates the fluorescence intensity of a particular dye that is

linked to a specific ddNTP in terminating the primer extension

reaction (green, blue, black, and red for ddATP, ddCTP, ddGTP,

and ddTTP, respectively; the letters above the bands identify the

bases and the numbers indicate their positions in the DNA

segment being sequenced). [Courtesy of Mark Adams, The

Institute for Genomic Research, Gaithersburg, Maryland.]

JWCL281_c07_163-220.qxd 2/23/10 8:54 AM Page 179

The fluorescence detectors used in these devices, which

have error rates of ⬃1%, are computer-controlled and

hence data acquisition is automated. In the most advanced

such systems, the sequencing gel is contained in an array of

up to 96 capillary tubes rather than in a slab-shaped appa-

ratus, sample preparation and loading are performed by

robotic systems, and electrophoresis and data analysis are

fully automated. Such systems can simultaneously sequence

96 DNA samples averaging ⬃650 bases each with a turn-

around time of ⬃1 h and hence can identify up to 1.6 mil-

lion bases per day—all with only ⬃15 min of human atten-

tion (vs the ⬃25,000 bases per year that a skilled operator

can identify using the above-described manual methods).

Nevertheless, one such system would require ⬃10 years of

uninterrupted operation to sequence the 3 billion-bp hu-

man genome with only two sets of overlapping fragments.

However, to ensure the complete coverage of a large tract

of DNA (Section 5-5Ea) and to reduce its error rate to

0.01%, at least 10 sets of overlapping segments must be

sequenced (Section 7-2B). Hence, the major sequencing

centers, where most genome sequencing is carried out,

each have over 100 such sequencing systems in factory-like

settings.

B. Genome Sequencing

The major technical challenge in sequencing a genome is

not the DNA sequencing itself but, rather, assembling the

tens of thousands to tens of millions of sequenced seg-

ments (depending on the size of the genome) into contigu-

ous blocks (called contigs) and assigning them to their

proper positions in the genome. One way that contigs

might be ordered is through chromosome walking (Section

5-5Ea). However, to do so for a eukaryotic genome would

be prohibitively time-consuming and expensive (e.g., to

“walk” the 125 million-bp length of an average length hu-

man chromosome using ⬃10-kb inserts from a plasmid li-

brary would require a minimum of 1.25  10

/10,000 

12,500 labor-intensive “steps”).

a. Map-Based Genome Sequencing

A more efficient technique of genome sequencing, the

map-based sequencing strategy (Fig. 7-16a), was developed

in the late 1980s. In this approach, low resolution physical

maps of each chromosome are prepared by identifying

shared landmarks on overlapping ⬃250-kb inserts that are

cloned in yeast artificial chromosomes (YACs). These land-

180 Chapter 7. Covalent Structures of Proteins and Nucleic Acids

Chromosome (250 Mb)

YAC library

1000 kb

Cosmid library

M13 or plasmid library

Sequence 800 clones

and assemble

Choose minimum cosmid overlap

Sequence cosmids in both directions

Cosmid walking

Plasmid

(5–10 kb)

or M13

(1 kb)

Cosmid

Cosmids

YAC

Random or

shotgun

High resolution

map

Low resolution

map

Genome or chromosome

40 kb

Map-based(a)

Chromosome (250 Mb)

BAC library

150 kb

Array 300,000

BACs

Sequence both ends

Fingerprint each BAC

BAC walking

BACs

STCs

M13 or plasmid library

Sequence 3000 clones

and assemble

Find 30 overlapping

BACs from STCs

Fingerprint comparisons

Sequence BACs with minimal

overlap at each end and repeat

Select seed BAC

WGSA(b)

Figure 7-16 Genome sequencing strategies. (a) The map-based

strategy uses three sets of progressively smaller cloned inserts

and assembles the sequenced inserts through cosmid walking and

the use of landmarks such as STSs and ESTs (see text). (b) The

WGSA strategy uses only two levels of cloning and employs

sophisticated computer algorithms as well as STCs to assemble

the sequenced inserts into finished chromosomes. [After Venter,

J.C., Smith, H.O., and Hood, L., Science 381, 365 (1996).]

JWCL281_c07_163-220.qxd 2/22/10 9:11 PM Page 180