Systematic Screening of Gene Expression Using a cDNA Macroarray - Drugs of Abuse: Neurological Reviews and Protocols - page 240

expression. The identity of the genes may remain the same, but their levels of
expression have been changed. Determining which genes' expression levels
change in response to drug administration in various animal models will
provide insights into the epigenetic imprinting by drugs of abuse. Moreover,
gene expression changes identifi ed in these studies may provide important
molecular targets for development of pharmacotherapeutic agents.
1.1. General Approach to DNA Array Technologies
With the completion of the human and other genome projects (and the
identification of more than 30,000 human genes), traditional models of
examining one gene at a time are being supplemented by large-scale screening
technologies. DNA hybridization arrays are a common form of screening
technology and allow the analysis of hundreds to thousands of genes in parallel
(5 , 6) . In the past, Northern blotting, dot blots, in situ hybridization, and
quantitative reverse transcription-polymerase chain reaction (QRT-PCR) were
the common methods for investigating changes in gene expression. Although
these approaches remain in common use, low throughput is a pervading
problem.
Hybridization array technology, on the other hand, offers to bypass many of
the limitations of these techniques by simultaneously creating labeled copies
of multiple RNAs and then hybridizing them to many different, gene-specifi c,
fi xed DNA molecules ( Fig. 1 ). The nomenclature has developed whereby the
labeled sample RNA is termed the target and the individual gene sequences
placed on the array are termed probes.
Although arrays are increasingly used for gene expression analysis, one
limitation to this technology is that arrays only measure relative levels of
mRNA expression. That is, the relative levels of RNAs can be described (e.g.,
sample A has 50% more of the specifi c RNA than sample B), but absolute
amounts (e.g., sample A has 1000 copies of the RNA and sample B has 500 copies
of the transcript) cannot be reported. As well, most hybridization arrays are not
designed to differentiate between alternatively spliced transcripts of the same
gene and, in some cases, between highly homologous members of a gene family.
Finally, a change in messenger RNA does not necessarily correlate with a
change in protein expression, and the translated protein often requires further
modifi cation to realize its full activity. These latter two points are a common and
legitimate criticism of array technology because it measures an intermediate
step (mRNA levels) and not functional product (active protein). However,
until proteomic technologies become universally accessible to the research
community, hybridization arrays are the best opportunity for studying gene
expression on a genome-wide scale.