https://doi-org.biblioteca.ibt.unam.mx:8080/10.1038/npg.els.

0003147

Abstract
DNA sequencing is the determination of base order in a DNA molecule. Methods for
determining base order involve either chemical degradation or, more commonly, enzymatic
synthesis of the region that is being sequenced. Automation of the DNA sequencing
process is accelerating the progress of the Human Genome Project.

Introduction
The development of methods that allow one to quickly and reliably determine the order of
bases, or the ‘sequence’, in a fragment of DNA is a key technical advance, the importance
of which cannot be overstated. Knowledge of DNA sequence enables a greater
understanding of the molecular basis of life. DNA sequence information provides
information critical to understanding a wide range of biological processes. The order of
bases in DNA specifies the order of bases in RNA, the molecule within the cell that directly
encodes the informational content of proteins. Scientists routinely use the DNA sequence
information to deduce protein sequence information. Base order dictates DNA structure and
its function, and provides a molecular programme that can specify normal development,
manifestation of a genetic disease, or cancer. See also DNA Structure: A‐, B‐ and Z‐DNA
Helix Families, DNAStructure: Sequence Effects, and Cancer Genetics (human)

Knowledge of DNA sequence and the ability to manipulate these sequences has accelerated
the development of biotechnology and has led to the development of molecular techniques
that provide the tools for asking and answering important scientific questions. The
polymerase chain reaction (PCR), an important biotechnique that facilitates sequence‐
specific detection of nucleic acid, relies on sequence information. DNA sequencing
methods allow scientists to determine whether a change has been introduced into the DNA,
and to assay the effect of the change on the biology of the organism, regardless of the type
of organism that is being studied. Ultimately, DNA sequence information may provide a
way to identify individuals uniquely. See also Polymerase Chain Reaction (PCR)

DNA sequencing has become so commonplace that the technique itself is often taken for
granted. However, this has not always been the case. It was, in fact, almost required that
scientists publish or present DNA sequence data before a sequence was considered reliable.
Furthermore, the length of the DNA information that it is possible to obtain and the number
of sequences that are analysed on a single gel have increased by an order of magnitude.
This article provides an overview of DNA sequencing development. See also History of
Biotechnology

To understand the DNA sequencing process, one must recall several facts about DNA.
First, a DNA molecule is composed of four bases, adenine (A), guanine (G), cytosine (C)
and thymine (T). These bases interact with each other in very specific ways through
hydrogen bonds, such that A interacts with T, and G interacts with C. These specific
interactions between the bases are referred to as base pairings. The two strands of a DNA
molecule occur in an antiparallel orientation in which one strand is positioned in the 5′ to 3′
direction and the other strand is positioned in the 3′ to 5′ direction. The terms 5′ and 3′ refer
to the directionality of the DNA backbone, and are critical to describing the order of the
bases. The convention for describing base order in a DNA sequence uses the 5′ to 3′
direction, and is written from left to right. Thus, if one knows the sequence of one DNA
strand, the complementary sequence can be deduced. See also Watson–Crick Base Pairs
and Nucleic Acids Stability, and Nucleic Acids: General Properties

There are two methods that are typically used to determine DNA sequence; the
development of each method resulted in the award of a Nobel Prize. The first uses
chemicals to specifically degrade the DNA strand, and is referred to as Maxam–Gilbert
DNA sequencing in honour of the inventors, A. Maxam and W. Gilbert. The second
method involves specific inhibition of enzymatic DNA synthesis and is referred to as
Sanger sequencing in honor is its inventor, F. Sanger. These two sequencing methods are
described in more detail below. See also Sanger, Frederick, and Gilbert, Walter

Both methods require that the reaction products share a common endpoint. This
requirement stems from the separation method used to visualize the reaction products.
These reaction products are size‐separated by applying an electric current through a gel
matrix (electrophoresis), and a common end is necessary to keep the reaction products in
register with respect to size mobility, so that the smaller products migrate more rapidly on
the gel relative to the larger products. More specifically, either the 5′ or the 3′ end can
define the fragment endpoint in a Maxam–Gilbert sequencing reaction, while only the 5′
end defines the fragment endpoint in a Sanger sequencing reaction. The reason for this
difference is clarified below. See also Gel Electrophoresis: One‐dimensional

Maxam–Gilbert DNA Sequencing (Chemical

Degradation)
In this method, a singly end‐labelled DNA fragment (typically labelled with a radioactive
marker) is exposed to base‐specific damage. The chemicals used to induce DNA damage
are dimethylsulfate (attacks G), sodium hydroxide (attacks A), formic acid (attacks G and
A), hydrazine (attacks C and T), and hydrazine in the presence of sodium chloride (attacks
C). These treatments are limited so that on average only one of the bases in the strand is
damaged. Modification and, ultimately, elimination of the base (but not the sugar) produces
a weak point in the DNA molecule that is susceptible to cleavage. Next, the DNA is
exposed to piperidine at high temperature to break the strand at the weakened position.
Since these reactions are performed on a population of molecules, electrophoresis of the
reaction produces a ladder of cleavage products that correspond to the positions of that base
along the DNA strand. Products from the different chemical modifications are
electrophoresed in adjoining lanes on the gel and read to determine the DNA sequence
(Figure 1). The smaller fragments indicate the identity of bases closest to the labelled end,
and successively larger products indicate the identity of bases farther from this end.
Depending on whether the label is located at the 5′ or the 3′ end of the DNA strand, the
base order is read from the bottom to the top of the autoradiogram as either 5′ to 3′ or 3′ to
5′, respectively.
Figure 1
Open in figure viewerPowerPoint
Schematic view of Maxam–Gilbert reaction products. G, A, R, Y and C represent the
specific chemical reactions that identify the relative positions of guanine (G), adenine (A),
purine (‘R’; G and A), pyrimidine (‘Y’; C and T) and cytosine (C) bases, respectively. In
this example the fragment is labelled at the 5′ end. Reading from the bottom towards the
top of the gel, the banding pattern corresponds to the sequence 5′ GGTACGCCTGA 3′.

Maxam–Gilbert sequencing is not routinely used by most investigators for several reasons.
First, data produced in chemical sequencing reactions are typically more ambiguous than
data produced in enzymatic sequencing reactions. One reason for this is that the chemical
reactivity of the bases is influenced by reaction impurities. Therefore, when one reads the
sequence from this type of reaction, the relative intensities of the reaction products must be
analysed for proper interpretation of base identity. Additionally, this procedure uses
hazardous chemicals and high levels of radioactivity. When compared with enzymatic
DNA sequencing, Maxam–Gilbert sequencing produces relatively shorter sequence
information and the procedures required to generate this information are more labour‐
intensive.

Sanger DNA Sequencing (Enzymatic

Synthesis)
Sanger sequencing is currently the most commonly used method for sequencing DNA. The
method exploits several features of a DNA polymerase: its ability to make an exact copy of
a DNA molecule; its directionality of enzymatic synthesis (5′ to 3′); its requirement for a
DNA strand (a ‘primer’) from which to begin synthesis; and its requirement for a 3′ OH at
the end of the primer. If a 3′ OH is not available, the DNA strand cannot be extended by the
polymerase. If a dideoxynucleotide (ddNTP – ddATP, ddTTP, ddGTP, ddCTP,), a base
analogue lacking a 3′ OH, is added into an enzymatic sequencing reaction, it is incorporated
into the growing strand by the polymerase. However, once the ddNTP is incorporated, the
polymerase is unable to add any additional bases to the end of the strand. Importantly,
ddNTPs are incorporated into the DNA strand by the polymerase using the same base
incorporation rules that dictate incorporation of natural nucleotides, where A specifies
incorporation of T, and G specifies incorporation of C (and vice versa). See
also DNAPolymerase Fidelity Mechanisms
Once the polymerase incorporates a ddNTP, chain extension stops. To determine DNA
sequence, one performs four reactions per template, where each reaction is used to
determine the relative position of a specific base along the DNA strand. This is
accomplished by adding a different dideoxynucleotide into each reaction containing the
DNA polymerase, the DNA primer, and the dNTPs. The ratio between ddNTP and dNTP is
critical for determining how many nucleotides (on average) the polymerase is able to
incorporate into the DNA molecule before incorporating a ddNTP, thereby terminating
chain elongation. The DNA primer, the ddNTPs, or the dNTPs can be either radiolabelled
or otherwise tagged to allow detection of the newly synthesized DNA strands. Since these
reactions are performed on a population of molecules, electrophoresis of the reaction
produces a ladder of extension products that correspond to the positions of that base along
the DNA strand. Products from the different reaction vessels are electrophoresed in
adjoining lanes on a sequencing gel, detected, and read to determine the DNA sequence
(Figure 2). The smaller fragments indicate the identity of bases closest to the 5′ end and,
since the DNA polymerase only incorporates bases in the 5′ to 3′ direction, reading the
identity of the successively larger products provides the 5′ to 3′ sequence of the extended
DNA strand. Typically, approximately 300–400 bases can be determined in a single
(manual) reaction. See also Gel Electrophoresis: One‐dimensional
Figure 2
Open in figure viewerPowerPoint
Data produced using Sanger sequencing reaction. G, A, T and C represent the sequencing
reaction products resulting from inclusion of ddGTP, ddATP, ddTTP or ddCTP. Since
enzymatic synthesis proceeds 5′ to 3′, the smaller fragments identify bases that are closer to
the primer (5′ end of the sequence information). (a) Schematic view of Sanger reaction
products. The DNA sequence identified by this pattern of bands is indicated. (b)
Photograph of corresponding sequence data.

Automated DNA Sequencing

A major advance in determining DNA sequence information occurred with the introduction
of automated DNA sequencing machines. The automated sequencer is used to separate
sequencing reaction products, detect and collect (via computer) the data from the reactions,
and analyse the order of the bases to automatically deduce the base sequence of a DNA
fragment. Automated sequencers detect extension products containing a fluorescent tag,
allowing researchers to eliminate radioactivity from the DNA sequencing process.
Sequence lengths that can be read using an automated sequencer are dependent upon a
variety of parameters, but typically range between 500 and 1000 bases.

As described for Sanger‐type sequencing reactions using (primarily) isotopes to detect the
extension products, some automated sequencers use four lanes to collect the data from the
reactions. However, some machines use differently coloured fluorescent tags to indicate
base identity (Figure 3). This approach enables a single lane to contain the data for a DNA
template and increases fourfold the amount of data contained on a gel. This single‐lane
approach is made possible by the development of fluorescent tags that can be attached
either to the DNA primer or to the ddNTP. Since four‐colour chemistry is used by more
researchers, it is discussed in more detail below.
Figure 3
Open in figure viewerPowerPoint
(a) Raw sequence data collected on an automated DNA sequencer (Perkin‐Elmer ABI
PRISM Model 377). The four colours indicate the relative position of the bases in the DNA
fragment. Each four‐colour vertical line corresponds to a different sequence reaction. The
smaller fragments (nearer the cathode) identify bases that are closer to the primer (5′ end of
the sequence information). (b) Portions of a representative, analysed sequence determined
by the automated sequencer.

Dye primer chemistry

When dye primer chemistry is used to detect the sequencing products, fluorescent tags are
attached to the sequencing primer. With this chemistry, the primer is synthesized four times
and a different tag (corresponding to a different base identity) is attached to the primer
during each synthesis. Subsequently, the researcher assembles four separate reactions, each
containing the DNA template, a specific ddNTP and colour‐coded primer, and the reagents
necessary to produce extension products. The primer defines the beginning (5′ end) of the
extension product, and the incorporated ddNTP defines base identity at the 3′ end of the
molecule. After the reaction is completed, the colour‐coded products are pooled and
prepared for loading into a single lane on an automated sequencer.

Dye terminator chemistry

When dye terminator chemistry is used to detect the sequencing products, base identity is
determined by the fluorescent tag attached to the ddNTP. This type of reaction chemistry is
performed in a single tube that contains the DNA template, the primer, all four
fluorescently labelled ddNTPs, and the reagents necessary to produce extension products.
The primer defines the beginning (5′ end) of the extension product and the incorporated,
colour‐coded ddNTP defines base identity at the 3′ end of the molecule. After the reaction
is complete, the extension products are prepared for loading into a single lane on an
automated sequencer. An advantage of dye terminator chemistry is that extension products
are visualized only if they terminate with a dye‐labelled ddNTP; prematurely terminated
products are not detected. Thus, reduced background signal typically results with this
chemistry.

Genome Sequencing
Very often, a researcher needs to determine the sequence of a DNA fragment that is larger
than the 500–1000 base average sequencing read length. Not surprisingly, strategies to
accomplish this have been developed. These strategies are divided into two major classes,
random or directed. Strategy choice is influenced by the size of the fragment to be
sequenced.

In random, or shotgun, DNA sequencing, a large DNA fragment (typically one larger than
20 000 base pairs) is broken into smaller fragments that are inserted into a cloning vector. It
is assumed that the sum of information contained within these smaller clones is equivalent
to that contained within the original DNA fragment. Numerous smaller clones are randomly
selected, DNA templates are prepared for sequencing reactions, and fluorescently‐labelled
primers that will base‐pair with the vector DNA sequence bordering the insert are used to
begin the sequencing reaction. Subsequently, the sequence of the original DNA fragment is
reconstructed by computer assembly of the sequences obtained from the smaller DNA
fragments. This strategy is being used extensively to determine the sequence of ordered
fragments that represent the entire human genome
[http://www.nhgri.nih.gov.biblioteca.ibt.unam.mx:2048/HGP/]. However, this random
approach is typically not sufficient to complete sequence determination, since gaps in the
sequence often remain after computer assembly. A directed strategy (described below) is
usually used to complete the sequence project.

A directed, or primer‐walking, sequencing strategy can be used to fill gaps remaining after
the random phase of large‐fragment sequencing, and as an efficient approach for
sequencing smaller DNA fragments. This strategy uses DNA primers that anneal to the
template at a single site and act as a start site for chain elongation. This approach requires
knowledge of some sequence information to design the primer. The sequence obtained
from the first reaction is used to design the primer for the next reaction and these steps are
repeated until the complete sequence is determined. Thus, a primer‐based strategy involves
repeated sequencing steps from known into unknown DNA regions; this process minimizes
redundancy, and it does not require additional cloning steps. However, this strategy
requires the synthesis of a new primer for each round of sequencing.

The necessity of designing and synthesizing new primers, coupled with the expense and the
time required for their synthesis, has limited the routine application of primer‐walking for
sequencing large DNA fragments. Researchers have proposed using a library of short
primers to eliminate the requirement for custom primer synthesis. The availability of a
primer library would minimize waste of primer, since each primer could be used to prime
multiple reactions, and would allow immediate access to the next sequencing primer. See
also Genome Mapping, and Genome Sequence Analysis

Prospects
One of the original goals of the Human Genome Project was to complete sequence
determination of the entire human genome by 2005. However, the project is ahead of
schedule and it is expected to produce a ‘working draft’ of the human genome by 2001. The
completed genome sequence is expected by 2003, at least two years ahead of schedule.
Technological advances are responsible for the rapid progress of this ambitious project.
Progress in all aspects involving DNA manipulation (especially manipulation and
propagation of large DNA fragments), evolution of faster and better DNA sequencing
methods, development of computer hardware and software capable of manipulating and
analysing the data (bioinformatics), and automation of procedures associated with
generating and analysing DNA sequences is responsible for this acceleration. See
also Human Genome Project, and The Promise of Whole Genome Sequencing