The following paragraphs are taking from a copy of lecture notes from Stanford University.

DNA arrays
Before exploring the details of DNA chips, let’s take a step back and look at the larger significance of this
technology. DNA chips are transforming the way scientists think of cells. Because of this technology, scientists can
now view a cell as an array of expressed genes. This is very different from the way cells were previously viewed
and this incredible amount of detail – at the very gene level – gives much more information about the internal
processes of a cell than was ever previously possible.
From this new point of view, a patient can be thought of as a collection of genetic mutations. Our knowledge of
specific mutation can help in a prognosis, and in further drug selection and design. More importantly, we can think
of disease as a result of inappropriate gene expression. In this way, DNA chips can be instructive in detecting
diseases such as cancer and finding drugs targets for treatment.
Background
We will start by first going into a short background of gene expression. Cells have a genome, which contains genes
(made up of DNA). For example, the human genome consists of 3 billion base pairs of DNA, and is thought to
contain 75,000-100,000 genes. From the DNA, expressed genes are transcribed into messages (RNA). It is
important to realize that although each cell contains a copy of the entire genome, in any given cell, at any given time
point, not all genes are necessarily expressed.
Experimentally, the RNA messages can be isolated from a cell and reversed-transcribed back into DNA, called
cDNA. A collection of cDNAs from cellular RNA makes up a cDNA library of the cell, and two different cDNA
libraries can be compared using a technique called differential hybridization.
There are many possible sources of cDNA libraries. Two different cell populations, such as a normal cell and
cancer cell, can be used as probes – another term for cDNA libraries. Such a comparison could yield valuable
insight into the differences of the two populations. Another option is to look at the same cell as it develops, in this
way seeing which genes are active over time during the different stages of the cell. This can give clues to the gene
function within a cell. Finally, cDNA libraries can be made of a cell as it responds to therapy, which can be used to
measure the effectiveness of a drug.
cDNA hybridization
The mechanism behind hybridization is the basis of DNA chips. DNA molecules are double-stranded, but these two
strands can melt apart at a characteristic melting temperature, Tm, which is usually above 65°C (see diagram
below). As the temperature is reduced and held below the melting temperature, single-stranded molecules hybridize
to their counterparts. In the same way, a RNA molecule (single-stranded, part of probe) can hybridize to a melted
cDNA molecule.
+
Hybridization protocol
DNA chips are two-dimensional arrays of prepared cDNA molecules. To use such an array, we first take a pool of
cDNA molecules and transcribe them into RNA molecules. These RNA molecules are then labeled with a
fluorescent dye (fluorochrome), a different color for each of the two different pools we wish to compare. The two
colors currently used to label molecules are red and green. Because we use two different colored dyes, we can mix
the two different populations and compare their genetic expression at the same time. This is important, because we
want to have nearly the same environment for both populations throughout the measuring process.
The resulting labeled RNA probe is put in contact with the array of targets on the DNA chip in a thin liquid layer for
several hours at Tm, 65°C. Finally, when the labeled RNA molecules have had time to hybridize, the chip is washed
in a solution several times and then scanned. The scanning process uses a laser and a computer to measure the level
of excitation of the fluorochromes on the DNA chip target array. The laser can detect both wavelength and intensity
of the excitation, and produces an image where each pixel corresponds to a single cDNA molecule. The
color of the pixel indicates the level of hybridization intensity ratio between the two probes. For example, a red
color indicates higher hybridization of probe 1, while green indicates a higher hybridization of probe 2.
Because of the measuring technique used, gene expression has a magnitude. In addition, there is a threshold to
response before a gene actually registers as being expressed.