Gene Chips and Arrays Revealed: A Primer on Their
Power and Their Uses
Stanley J. Watson and Huda Akil
This article provides an overview and general explanation
of the rapidly developing area of gene chips and expres-
sion array technology. These are methods targeted at
allowing the simultaneous study of thousands of genes or
messenger RNAs under various physiological and patho-
logical states. Their technical basis grows from the Hu-
man Genome Project. Both methods place DNA strands on
glass computer chips (or microscope slides). Expression
arrays start with complementary DNA (cDNA) clones
derived from the EST data base, whereas Gene Chips
synthesize oligonucleotides directly on the chip itself. Both
are analyzed using image analysis systems, are capable of
reading values from two different individuals at any one
site, and can yield quantitative data for thousands of genes
or mRNAs per slide. These methods promise to revolution-
ize molecular biology, cell biology, neuroscience and
psychiatry. It is likely that this technology will radically
open up our ability to study the actions and structure of
the multiple genes involved in the complex genetics of
brain disorders. Biol Psychiatry 1999;45:533–543
© 1999 Society of Biological Psychiatry
Key Words: Expression arrays, genes, mRNA, brain
Why Do We Need Gene Arrays or Gene
Chips?
B
etween the mid-1970s and the early 1980s, the world
of biology, including neurobiology, embarked on a
major conceptual revolution, as critical ideas and tools of
molecular biology were radically altered. Over the next 10
to 15 years, the fields of genetics and molecular and
cellular biology matured both theoretically and techni-
cally. A major offshoot of this revolution is the Human
Genome Project, which focuses on the description and
sequencing of the human genome, along with that of a few
other selected species. As a result of these two forces—
rapidly advancing molecular biology and genetics, and the
torrent of information from the Human Genome Project,
biomedical research finds itself at a critically exciting
time. Once all the genes are known and their chromosomal
locations assigned, and once we have a sense of their
allelic variations (see Burmeister’s paper in this issue), the
nature of the questions we ask will change. We will no
longer need to decode the DNA sequences. Rather, we
need to move to a better understanding of structure,
function, and regulation of the resultant proteins. The first
stage of basic information will have been laid out. The
hard work of asking the right questions and attempting to
understand the functioning of these genes in the context of
health and disease will have just begun.
This increase in fundamental knowledge is accompa-
nied by a dawning realization that in order to understand
many genetic disorders, it is critical for us to understand
not only the actions of single genes, but the interactions
between multiple genes. Nowhere is this truer than in the
brain, and no set of diseases as a group exemplify this
necessity better than the psychiatric disorders.
Much of modern cellular and developmental biology
studies how genes modulate each other and are modulated
by internal and external environmental triggers (ranging
from light and sound, to ions, transmitters, hormones,
growth factors, and nutrients) to control developmental
and cellular processes. Neurobiology concerns itself not
only with all these signaling mechanisms, but also with the
functions of cellular networks, which represent an addi-
tional level of complexity and plasticity in the interplay
among the action of various genes. Since all brain func-
tions are dependent on the interactions of countless genes,
it is not surprising that brain-related disorders will reflect
disruptions in multiple genes. Of course, there are many
cases where a defect in a single, critical gene along the
cascade of events is sufficient to cause a clear-cut illness
(e.g., Huntington’s disease). However, many psychiatric
illnesses, when considered closely, appear to represent
problems in “regulation” or “fine tuning” of complex
functions, such as emotional tone or cognitive processing.
For example, no one believes that a patient suffering from
a mood disorder is congenitally incapable of generating or
experiencing normal affective responses such as sadness
or joy. The illness likely reflects a difficulty in dampening
the responses or terminating them when no longer appro-
From the Mental Health Research Institute, and Department of Psychiatry,
University of Michigan, Ann Arbor, Michigan.
Address reprint requests to Stanley J. Watson, 205 Zina Pitcher Place, Ann Arbor,
MI 48109-0720.
Received August 21, 1998; revised October 19, 1998; accepted October 21, 1998.
© 1999 Society of Biological Psychiatry 0006-3223/99/$19.00
PII S0006-3223(98)00339-4