Biology Reference
In-Depth Information
determined (
Ramsay 1998
). The technology of depositing nucleic acids on glass
slides at very high densities has allowed the miniaturization of nucleic-acid
arrays with dramatic increases in experimental efficiency and information con-
tent. Arrays with
>
250,000 different cDNAs per square centimeter can be pro-
duced (
Lockhart and Winzeler 2000
).
To conduct an experiment in which gene activity of two different tissue types
(or the two sexes, or different developmental stages) are compared, cells are
broken open and then the mRNAs present are extracted and converted into
complementary DNA (cDNA) copies. The cDNAs are labeled with different dyes
(fluorophores) and applied to the microarray. The cDNAs bind to (hybridize
with) the DNA template on the microarray based on complementary-base pairing.
The colors of the spots are analyzed. If, for example, one cDNA sample has a
red label and the other cDNA sample has a green label, there will be four color
options on the microarray: red spots, green spots, no color (where no cDNAs
bound to the spot), and yellow (where both cDNAs bound to the spot). The
microarray is laser scanned, and data are recorded and analyzed with computer
software that can estimate the intensity of the colors. Because the location of
each spot and its sequence are known, scientists can determine which genes are
expressed at a specific stage of development or in specific tissues.
However, scientists are concerned that measuring mRNA levels (as cDNAs) may
not be the best way to measure gene expression. It is not clear how closely levels
of mRNAs relate to levels of their corresponding proteins, indicating that studies
will have to be done to determine how to accurately quantify protein levels and
their differential stability in the organism (
White et al. 1999, Andrews et al. 2000
).
The use of gene chips has provided some surprises and enormous amounts of
data (
Andrews et al. 2000, Quackenbush 2001
). There is almost too much data
produced, and data analysis methods are still under development. For example,
the question needs to be answered, “Does gene expression indicate function?”
At present, the conclusion is that there is a correlation between distinct expres-
sion profiles and function, but expression should not be taken as sufficient evi-
dence for function. For example, not all genes involved in a function such as
DNA replication are expressed periodically during the cell cycle and some genes
that do not need to be cell-cycle regulated are transcribed in a periodic manner
(
Lockhart and Winzeler 2000
).
Gene chip or microarray data require careful analysis because experimen-
tal evidence shows there is a disparity between the relative expression levels
of mRNA and their corresponding proteins. Thus, expression information from
both mRNA and proteins is required to understand a gene network (Dutt and
