as oligonucleotides or cDNAs deposited onto a solid surface. The solid support may be
either glass or silicon and currently the arrays are synthesised on or off the chip. They
require complex fabrication methods similar to that used in producing computer micro-
chips. Most commercial productions employ robotic ultrafine microarray deposition
instruments which dispense volumes in the picolitre range. Alternatively on-chip fabri-
cation as used by Affymetrix builds up layers of nucleotides using a process borrowed
from the computer industry termed photolithography. Here wafer-thin masks with holes
allow photoactivation of specific dNTPs which are linked together at specific regions on
the chip. The whole process allows layers of oligonucleotides to be built up with each
nucleotide at each position being defined by computer.
The arrays themselves may represent a variety of nucleic acid material. This may be
mRNA produced in a particular cell type, termedcDNA expression arrays,ormay
alternatively represent coding and regulatory regions of a particular gene or group of
genes. A number of arrays are now available that may determine mutations in DNA,
mRNA transcript levels or other polymorphisms such as SNPs. Sample DNA is placed on
the array and any unhybridised DNA washed off. The array is then analysed and scanned
for patterns of hybridisation by detection of fluorescence signals. Any mutations or
genetic polymorphisms in relevant genes may be rapidly analysed by computer inter-
pretation of the resulting hybridisation pattern and mutation, transcript level or poly-
morphism defined. Indeed the collation and manipulation of data from microarrays
presents as big a problem as fabricating the chips in the first place. The potential of
microarrays appears to be limitless and a number of arrays have been developed for the
detection of various genetic mutations including the cystic fibrosis CFTR gene (cystic
fibrosis transmembrane regulator), the breast cancer gene BRCA1 and in the study of the
human immunodeficiency virus (HIV).
At present microarrays require DNA to be highly purified, which limits their applic-
ability. However as DNA purification becomes automated and microarray technology
develops it is not difficult to envisage numerous laboratory tests on a single DNA
microchip. This could not only be used for analysing single genes but large numbers of
genes or DNA representing microorganisms, viruses, etc. Since the potential for quantita-
tion of gene transcription exists expression arrays could also be used in defining a
particular disease status. This technique may be very significant since it will allow large
amounts of sequence information to be gathered very rapidly and assist in many fields
of molecular biology, especially in large genome sequencing projects or in so-called
resequencingprojects where gene regions such as those containing potentially important
polymorphisms require analysis in a number of samples.
One current application of microarray technology is the generation of a catalogue
of SNPs across the human genome. Estimates indicate that there are approximately
10 million SNPs and importantly 200 000 coding or cSNPs that lie within genes and
may point to the development of certain diseases. SNP analysis is therefore clearly a
candidate for microarray analysis and developments such as Affymetrix Genome
Wide SNP array enables the simultaneous analysis of nearly 1 million SNPs on one
gene chip. In order to simplify the problem of the vast numbers of SNPs that need to
be analysed the HapMap project currently analyses SNPs that are inherited as a block,
and in theory as few as 500 000 SNPs will be required to genotype an individual.253 6.8 Analysing genes and gene expression