Microarrays and sequencers: Moving forward, in sequencers

When you think of the Human Genome Project, images of banks of sequencing machines pumping out millions of base pairs of information comes to mind. According to Dr John Barlow, Melbourne divisional manager of the Australian Genome Research Facility (AGRF), state-of-the-art sequencing is still largely electrophoresis-based. However, sequencers have come a long way since the days of pouring your own sequencing gel and using radioactive dideoxynucleotides.

While flatbed sequencing is still used, particularly in small-scale occasional research situations, the trend today is to use capillary electrophoresis which allows the use of fluorescent dye terminators, automatic interpretation of the sequence and high throughput capacity.

The AGRF has about 20 sequencers of both varieties, according to Barlow. The latest acquisition is a pair of Applied Biosystems' latest models, the 3730xl DNA Analyser. These have a 96 capillary capacity, but the high speed matrix reduces the runtime significantly, allowing up to 48 runs a day.

The other major competitor in the high throughput sequencing stakes is the Amersham MegaBACE 4000, which has a 384-sample capacity. But capillary sequencers are available in a range of formats and their use is not restricted to sequencing, they are also increasingly used for genotyping and fragment analysis. The Beckman CEQ8000 for example, can perform both sequencing and genotyping functions within the same run.

According to sales manager Stephen Paul, Beckman has seen a jump in sales of sequencers in the Australian market this year, primarily due to increased interest in the genotyping applications available. He notes that a lot of the interest comes from the agricultural and molecular diagnostics sectors.

"Practical applications of genetic analysis appear to be significantly on the increase. It is likely that Australia will see growth in the use of this technology for detection and genotyping of pathogens and for routine diagnostic testing," he claims.

On the horizon though, are several new technologies, which might ultimately supplant electrophoretic sequencing. "There is a possibility that sequencing will become a chip-based technology but it is still a way away," says Barlow.

Affymetrix, for example, has pioneered a lot of the microarray technology used today, and is working on custom-made "resequencing" chips. According to Robert Henke, a product specialist for Millennium Science, which distributes Affymetrix products in Australia, these chips will be available in a couple of months.

These chips will be able to analyse 30 kilobases of DNA in a bidirectional fashion, says Henke. Their major use will be to compare DNA sequences between large numbers of genomes, for example corresponding to a cohort of patients, allowing rapid identification of the individual patterns of mutations along a known stretch of DNA.

The basis of this technique relies on knowing the consensus sequence of the DNA of interest. Essentially, each position in the sequence is "interrogated" by a set of four oligomers, 25 bases in length, differing only at the central position. An overlapping series of these oligomer sets is used, designed so that each set in the series moves up or down the DNA by one base. In this way, the entire sequence is built up.

Henke says that the major use of this kind of technology will be in clinical situations. He notes that Australian researchers may not be big users of the technology, at least at first. But he sees it as an economical way to obtain substantial sequence data from multiple genomes in a short timeframe.

Another company, US Genomics is developing technology to read DNA in a linear fashion, using nanotechnology approaches to directly read the molecule, base by base.

The ultimate goal is to drive the speed of sequencing up and the cost down, so that sequencing of an individual's genome in a doctor's office may become a practical possibility. This idea has been much touted recently by Craig Venter, who claims that individual genome sequences may one day cost as little as $US1000.