The mountain of information that is the draft sequence of the human genome may be impressive, but without interpretation that is all it remains — a mass of data. Gene function is one of the key elements researchers want to extract from the sequence, and the DNA microarray is one of the most important tools at their disposal.

The past few years have seen rapid growth within the microarray field, with the falling price of technology allowing biologists to abandon their home-made equipment in favour of one of an expanding range of commercial instruments now on the market.

Custom-built robot arrayer in action at TIGR can spot 100 slides at a time. Credit: JEREMY HASSEMAN

“Medicine is going molecular in a major way, and microarrays are being used to profile everything from autism and schizophrenia to Alzheimer's and Parkinson's,” says Mark Schena, visiting scholar at TeleChem International/arrayit.com in Sunnyvale, California, which sells microarray reagents and parts for arrayer robots.

Microarrays exploit the preferential binding of complementary single-stranded nucleic-acid sequences. The underlying principle is the same for all microarrays, no matter how they are made — the unknown sample is hybridized to an ordered array of immobilized DNA molecules whose sequence is known. Each array features thousands of different DNA probe sequences arranged in a defined matrix on a glass or silicon support. Unlike conventional nucleic-acid hybridization methods, microarrays can identify thousands of genes simultaneously, which means that genetic analysis can be done on a huge scale.

This has revolutionized the way in which researchers analyse gene expression in cells and tissues. Microarrays — also referred to as DNA arrays, DNA chips, biochips and GeneChips — allow researchers to determine which genes are being expressed in a given cell type at a particular time and under particular conditions. They can be used to compare the gene expression in two different cell types or tissue samples, for example, healthy versus diseased tissue, and to examine changes in gene expression at different stages in the cell cycle or during embryonic development.

Microarrays are also being used in comparative genomic hybridization studies, a molecular cytogenetic approach for genome-wide detection of chromosomal deletions and amplifications, as well as for genotyping individuals for genetic differences, such as single-nucleotide polymorphisms (SNPs), that might be associated with disease.

At a fundamental level, microarrays are also being used in attempts to assign probable functions to newly discovered genes by comparison with the expression patterns of known genes, to identify key players in signalling pathways and to uncover new categories of genes.

But their use is not restricted to basic biology. They are also finding applications in the identification of new targets for therapeutic drugs, in disease diagnosis, and in toxicogenomics, the study of the genetic basis of an individual's response to environmental factors such as drugs and pollutants.

Spot specs

The most commonly used substrate for microarrays is glass — although they can be made of other materials, such as silicon — onto which thousands of spots of single-stranded DNA probes, in the form of cDNAs or oligonucleotides, are placed by a robot arrayer using contact or non-contact printing methods.

Alternatively, oligonucleotides can be synthesized in situ, building up each element of the array nucleotide by nucleotide and using ink-jet printing or photolithographic methods similar to those used in the semiconductor industry.

The spots are typically less than 200 µm in diameter and need to be read by specialized imaging equipment — confocal laser scanners. The spot sizes on 'macroarrays', by contrast, are about 300 µm or more and can be imaged using conventional gel and blot scanners. Contact printing and ink-jetting methods typically give spots of 100 µm in diameter, whereas those produced by photolithography are about 20 µm. This produces microarray densities of 10,000 and 250,000 spots per cm2, respectively.

Industry landscape

Affymetrix (above) leads the high-density chip market; new kid on the block is Motorola's CodeLink chip (inset). Credit: AP PHOTO/RICH PEDRONCELLI/MOTOROLA

Affymetrix of Santa Clara, California, was one of the first commercial microarray companies and still has command over the high-density microarray market. The company uses 25-mer oligonucleotides synthesized in situ using its proprietary process, which combines solid-phase chemical synthesis with photolithography. Its GeneChip — an Affymetrix trademark — Human Genome U133 set of two microarrays contains over 1 million different oligonucleotides, representing more than 33,000 of the best-characterized human genes.

The price of GeneChips has come down by about half, bringing them within the reach of at least some academic researchers. The company shipped over 280,000 GeneChips in 2001 and reported revenues of US$194.9 million, up 12% on 2000.

But recent years have seen a shake-up in the industry. Last year, Incyte Genomics of Palo Alto, California, a leading supplier of microarrays, quit the chip-making business, deciding instead to refocus its efforts on its core information business. By forging strategic collaborations with microarray manufacturers, which get access to the company's extensive database and patent portfolio, Incyte hopes to benefit from microarray sales without having to make them. Incyte may be gone, but some heavy hitters — most notably Agilent Technologies in Palo Alto and Motorola of Northbrook, Illinois — have recently entered the market-place.

It is perhaps not surprising that Motorola is making a play in this area. The company has a keen nose for business opportunities in emerging markets and the deep financial pockets needed to secure some market share. It also has core expertise in manufacturing, microfluidics, miniaturization, software engineering and systems integration.

Its subsidiary, Motorola Life Sciences, launched its first microarray product last summer. The CodeLink bioarray system for gene-expression profiling and SNP genotyping includes off-the-shelf arrays, optimized reagents and software to capture the images and carry out a first-level analysis of the array. Labs can use their own scanners. The company offers human and rat arrays, each representing 10,000 full-length gene sequences, and expects to launch a mouse array next month. Its genotyping array contains 72 SNPs from the P450 cytochrome family. Motorola's agreement with Incyte Genomics allows it to develop microarrays based on Incyte's comprehensive gene databases.

Motorola synthesizes 30-mer oligonucleotides 'off-line' and spots them onto slides coated with a three-dimensional, branched polymeric substrate gel surface, using Hewlett-Packard's non-contact, piezo-dispense technology. The company also produces custom arrays to order and sells 'activated' non-spotted slides for researchers to make their own arrays.

Agilent Technologies, on the other hand, uses proprietary SurePrint ink-jet technology and offers human, mouse and and rat cDNA arrays and custom oligonucleotide arrays. In the latter case the oligonucleotides (either 25- or 60-mer) are synthesized in situ and built up a base at a time on standard 1 × 3-inch glass slides to give arrays of either 8,400 or 22,000 features. Doug Amorese, R&D section manager responsible for chemistry and molecular biology in Agilent's DNA Microarray Program, says the cDNA type of microarray is useful when large numbers of identical arrays are needed, whereas the in situ system provides the flexibility to tailor designs to suit individual needs.

As a subsidiary of Hewlett-Packard, Agilent has access to considerable expertise in ink-jet printing methods and high-end analytical instrumentation — principally high-performance liquid chromatography and mass spectrometry. So the microarray area “seemed like a very good fit” for the company, says Amorese. Hewlett-Packard had been looking for a way into molecular biology, and microarrays “seemed like an area that was going to grow”, he says.

The cross-licensing agreement Agilent signed in 1999 with Oxford Gene Technology (OGT) of Oxford, UK, is seen by the company as key to making this happen. OGT was set up by Edwin Southern and the University of Oxford in 1995 to commercialize Southern's DNA microarray patents. Agilent's other main collaborators are Rosetta Inpharmatics of Kirkland, Washington, and Incyte Genomics.

David and Goliath

Peer Stähler (left) and board members of febit with Geniom one. Credit: FEBIT

As well as the big guns, several smaller companies are seeking to carve out a niche. One example is febit, a young biotechnology company employing some 70 people in Mannheim, Germany. It has developed a prototype DNA analysis device that fully automates and integrates all the steps in the analysis process. Its machine, Geniom one, is designed for both gene-expression analysis and genotyping. It offers “a plug-and-play solution”, says Peer Stähler, febit's vice-president and chief scientific officer, and one of the company's founders. “You don't have to become an expert in surface chemistry, you don't have to optimize the processes. All you need is data,” he says.

At the heart of Geniom one is the programmable DNA processor — a special reaction carrier with a three-dimensional microchannel structure. Both the synthesis of the oligonucleotide probes — which uses a light-dependent technique that does not rely on physical masks — and the hybridization of the labelled samples takes place in the channels. “You insert the reaction carrier and never touch it again until you throw it away,” says Stähler. “If you're efficient you can do two runs a day.”

The current design can produce microarrays containing up to 64,000 different oligonucleotides — it runs eight arrays in parallel, each with 8,000 spots per array. With between one and four spots covering a gene, each array can cover a few thousand genes. This is not as dense a coverage as Affymetrix's GeneChips, but Stähler expects future versions of Geniom to have 10 times as many spots per array.

The prototype is being tested by Jörg Hoheisel and his team at the German Cancer Research Centre in Heidelberg. Stähler expects Geniom one, which has a price tag of a few hundred thousand dollars, to hit the market by the end of the year.

Room for improvement

There is still a lot of room for improvement in microarray technology, say players in the field. TeleChem International/arrayit.com, for example, is exploring the use of reflective substrates. Although still in the development phase, Schena says it seems that printing microarrays on mirrors rather than glass improves the signal-to-noise ratio by as much as 1,000%.

Several companies are pursuing the development of 'active' hybridization technologies. Advalytix, a recent spin-off from the Center for NanoScience at the Ludwig-Maximilians University of Munich, will begin shipping a hybridization device this month, which has no moving parts and is designed to speed up hybridization reactions, as well as to produce more homogeneous reaction conditions than with 'passive' hybridization, eliminating so-called edge effects. The mixer chip uses surface acoustic waves to control the motion of reagents. It is used in a 'sandwich' arrangement, with a conventional DNA microarray slide on the bottom, the mixer chip on top and the hybridization solution in between.

“Microarrays will get better over time and a lot of that will be in content as we better understand which genes are important and, specifically, perhaps which splice variants are most important,” says Amorese. In addition to improvements in the probes themselves, he expects advances in labelling technologies for the sample nucleic acid, allowing researchers to use less starting material. As for chips in the clinic, Schena believes they will be there within five years, and probably a lot sooner on the genetic screening side.

→ http://www.microarray.org

→ http://www.gene-chips.com

→ http://www.lab-on-a-chip.com

→ http://cmgm.stanford.edu/pbrown/mguide/index.html

→ http://www.mged.org/