Soil Teeming with Life: New Frontiers for Soil Science - Sustainable Management of Soil Organic Matter - page 400

intended to give perspective to the reader for what is needed to conduct a
microarray experiment. To start, careful attention needs to be paid to
developing a research question, and determining the appropriate array
format. Arrays which can address a variety of questions will be most
valuable, since the bulk of time and cost involves array design and preparing
the nucleic acid samples (oligonucleotide synthesis, PCR) for placement on
the array. Array fabrication is largely automated and, other than the initial
cost of the arraying device (see Bowtell, 1999 for recent listing of products
available), this step is quite affordable. The technical details of sample
preparation are outlined briefly, though these are important details that
need to be worked out, particularly for low biomass environmental samples.
The hybridization itself is straightforward; specificity, normalization and
sensitivity of the hybridization reaction can be assessed with internal
controls on the array. Experiments are conducted with dual fluorochrome-
labelled templates, with either gene expression compared under two
experimental conditions, or a reference sample compared with the
experimental sample. Microarrays are visualized with either a confocal
scanning laser or a CCD camera specifically designed for microarrays. The
image file representing the microarray is processed using commercially
available software (see Bassett et al ., 1999) or shareware available on the
web (http://rana.stanford.edu/software/).
Arrays work in much the same way that traditional hybridization
approaches have operated. In a simple case, where the relative amount
of gene expression is to be assessed, the target (labelled nucleic acid in
solution) samples are varied experimentally. For example, DeRisi et al .
(1997) compared mRNA isolated from starved cells with mRNA isolated
from cells grown under nutrient-rich conditions. The two different mRNA
populations were labelled with different fluorochromes (Cy3 and Cy5), and
hybridized together on the same microarray. The scanner delivers two
images (one for each fluorochrome) which are overlaid using the processing
software. Signal intensities of each spot are determined and a ratio of signal
intensities is derived. Using the relative representation of RNA to compare
different samples is the most optimal way to use these data, due to
differences between sample processing, variations in labelling and other
experimental conditions (Eisen and Brown, 1999). The ratio values can be
analysed by a variety of statistical methods to assess relationships between
coexpressed genes.
Microarray uses in environmental microbiology
There are a number of ways in which environmental microbiology will
benefit from microbial genomics. As mentioned earlier, microarrays are
being used in microbial functional genomics research to determine patterns