DNA sequencing means working out the linear sequence of the four bases (A, C, G, and T) in fragments of DNA, usually fragments that have been amplified by PCR or DNA cloning. By working out the base sequence of many short DNA fragments with over lapping DNA sequences, one can get complete sequences for whole genes and whole genomes. The standard method does not distinguish between modified bases and the unmodified version of the base (such as between 5-methylcytosine and cytosine, or between N6-methyladenine and adenine)—more specialized methods can do that job. Note that DNA sequencing is also used to infer the linear sequence of bases in RNA molecules (the RNA must first be converted into a complementary DNA sequence using a reverse transcriptase).
Until quite recently, dideoxy DNA sequencing was essentially the only DNA sequencing method that was used (it was first published in 1997 by Fred Sanger, earning him his second Nobel Prize). It relies on amplifying individual DNA sequences that have been purified by DNA cloning or PCR. For each amplified DNA, nested sets of labeled DNA copies are made and then separated according to size using gel electrophoresis.
In the last few years, completely different technologies have allowed massively- parallel DNA sequencing. No attempt is made to obtain the sequence of just a purified DNA component; instead, millions of DNA fragments present in a complex DNA sample are simultaneously sequenced without the need for gel electrophoresis.
Dideoxy DNA sequencing provides highly accurate DNA sequences and is still widely used for investigating specific DNA sequences, such as testing whether individuals have mutations in a particular gene or confirming a suspected mutation. What the newer DNA sequencing technologies offer is a dramatic increase in sequencing capacity and the ability to sequence complex DNA populations, such as genomic DNA sequences, very rapidly. As a result of fast-developing technology, the running costs of DNA sequencing are plummeting and very rapid sequencing of whole exomes and even whole genomes is becoming routine.
The basics of Sanger dideoxy DNA sequencing
Like PCR, dideoxy DNA sequencing uses primers and a DNA polymerase to make DNA copies of specific DNA sequences of interest. To obtain enough DNA for sequencing, the DNA sequences are amplified using PCR (or sometimes by cloning in cells). The resulting purified DNAs are then sequenced, one after another, in individual reactions. Each reaction begins by denaturing a selected purified DNA. A single oligonucleotide primer is then allowed to bind and used to make labeled DNA copies of the desired sequence (using a provided DNA polymerase and the four dNTPs).
Instead of making full-length copies of the sequence, the DNA synthesis reactions are designed to produce a population of DNA fragments sharing a common 5′ end sequence (defined by the primer sequence) but with variable 3′ ends. This is achieved by simultaneously having present the standard dNTP precursors of DNA plus low concentrations of ddNTPs, dideoxynucleotide analogs that differ from a standard deoxynucleotide only in that they lack an OH group at the 3′ carbon of the sugar as well as at the 2′ carbon (Figure 1A).
Fig1. Principle of dideoxy sequencing. (A) Generalized structure of a 2′,3′ ddNTP. The sugar is dideoxyribose because the hydroxyl groups attached to both carbons 2′ and 3′ of ribose are each replaced by a hydrogen atom (shown by yellow shading). (B) In dideoxy sequencing reactions, a DNA polymerase uses an oligonucleotide primer to make complementary sequences from a purified, single-stranded starting DNA. The sequencing reactions include ddNTPs that compete with the standard dNTPs for insertion of a nucleoside monophosphate into the DNA. Different labeling systems can be used but it is convenient to use labeled ddNTPs that have different fluorescent groups according to the type of base, as shown here by colored circles. The DNA copies will have a common 5′ end (defined by the sequencing primer) but variable 3′ ends, depending on where a labeled dideoxynucleotide has been inserted, producing a nested set of DNA fragments that differ by a single nucleotide in length. A series of nested fragments that incrementally differ by one nucleotide from their common 5′ end are fractionated according to size by gel electrophoresis, and the fluorescence signals are recorded and interpreted to produce a linear base sequence. (C) Example of DNA sequence output, showing a succession of dye-specific (and therefore base-specific) intensity profiles. This example shows a cDNA sequence from the PHC3 polyhomeotic gene, provided by E. Tonkin, Newcastle University.
DNA synthesis continues smoothly when dNTPs are used, but once a dideoxynucleotide is incorporated into a growing DNA chain, chain synthesis is immediately terminated (the dideoxynucleotide lacks a 3′-OH group to form a phosphodiester bond). To keep the balance tilted toward chain elongation, the ratio of each ddNTP to the corresponding dNTP is set to be about 1:100, so that a dideoxynucleotide is incorporated at only about 1% of the available nucleotide positions.
If we consider competition between ddATP and dATP in the example in Figure 6.18B, there are four available positions for nucleotide insertion: opposite the T at nucleotide positions 2, 5, 13, and 16 in the starting DNA. Because the DNA synthesis reaction results in numerous DNA copies, then, by chance, some copies will have a dideoxyA incorporated opposite the T at position 2, some will have a dideoxyA opposite the T at position 5, and so on. Effectively, chain elongation is randomly inhibited, producing sets of DNA strands that have a common 5′ end but variable 3′ ends.
Fluorescent dyes are used to label the DNA. One convenient way of doing this, as shown in Figure 6.18B, is to arrange that the four different ddNTPs are labeled with different fluorescent dyes. The reaction products will therefore consist of DNA strands that have a labeled dideoxynucleotide at the 3′ end, carrying a distinctive fluorophore according to the type of base incorporated.
All that remains is to separate the DNA fragments according to size using electrophoresis (see Box 6.3), and then to detect the fluorescence signals. The latter is usually achieved during gel electrophoresis: as the migrating DNA fragments reach a certain point in the gel, they pass a laser that excites the fluorophores, causing them to emit fluorescence at distinct wavelengths. The fluorescence signals are recorded, and an output is provided in the form of intensity profiles for the differently colored fluorophores (see Figure 1C for an example).
Dideoxy DNA sequencing is disadvantaged by relying on gel electrophoresis. It is not amenable, therefore, to full automation (slab polyacrylamide gels were used initially; more modern machines use semi-automated capillary electrophoresis—see Box 6.3 Figure 1B). Because gel electrophoresis is not suitable for handling large numbers of samples at a time, dideoxy sequencing can generate a limited amount of sequence data only. It is, therefore, not well suited to genome sequencing. It does, however, provide highly accurate sequences over several hundred bases. In modern times it is often used for analyzing variation over small DNA regions, such as regions encompassing individual exons, and in validating some sequences obtained by newer, high-throughput DNA sequencing methods.
