Transcript Document

What are microarrays?
Microarrays consist of thousands of oligonucleotides or cDNAs that
have been synthesized or spotted onto a solid substrate (nylon,
glass, or plastic).
Samples of cell DNA or cDNA are labeled with radioactivity or
fluorescent probes and hybridized to the array.
When the labeled sample DNA binds to its complementary strand on
the array, the specific spot for that gene ‘lights up’ with color or
radioactivity.
Spots that bind a lot of DNA (or cDNA) appear bright, spots that bind
less DNA are dim.
The relative signal for each spot can be measured quantitatively
using an array scanner and image analysis software. This allows
one to simultaneously measure expression of thousands of genes.
Microarrays contain spots that represent
expression of 1000s of genes
How do microarrays measure gene expression?
1. The total RNA is isolated from 2 different biological samples (for
example, RNA from normal cells and cancer cells).
2. The RNA is reversed transcribed to cDNA. Each population of
cDNA is labeled with a different fluorescent dye (ie, red for
cancer and green for normal). The 2 cDNA populations contain
all expressed genes in the cells.
3. The cDNAs are denatured and hybridized to the microarray
(similar to Southern hybridization) under stringent conditions.
The colored cDNAs stick to the complementary sequences on
the array.
4. If equal amounts of RNA (cDNA) are present in normal and
cancer cells, the color of the spot is yellow (equal amounts of
red and green make yellow).
5. If there is more red labeled cDNA (cancer) than green (normal),
the spot is red.
6. If the gene is highly expressed, there is more cDNA and the spot
is bright (either yellow, green, or red). If the gene is expressed
poorly, the spot is dimmer. Genes that are expressed at a very
low level can create problems for array analysis.
7. The intensity and the color of each spot is read and
quantitatively analyzed using an array scanner and image
analysis software.
Why measure gene expression on a global scale?
It is estimated that humans have about 30,000 genes. Microarrays
have been constructed that could simultaneously measure all genes
in a human cell. All spots could fit on several microscope slides.
1. Classification of human diseases: Most pathologists classify
cancer or other illness by looking at cells in the microscope and
describing how they appear. This is sometimes not very precise.
Microarrays have been used to compare normal and disease tissue
for differences in gene expression. This method is very precise and
discriminating. Some types of tumors that were thought to be
similar by microscopic analysis were found to be very different
when examined by microarray.
This led to a different strategy of treatment based on the genes that
were overexpressed, and this was more effective therapy. For
example, different types of human leukemia were found to appear
similar in the microscope, but to be very different when one
examined their pattern of gene expression.
2. Discovery of new drugs: When one can compare normal and
diseased tissue, it is possible to find all of the genes that are
expressed abnormally.
When all of these genes are grouped together, they often suggest
major pathways that are perturbed in a specific disease.
For example, if many genes for apoptosis were upregulated in
diseased tissue, the underlying problem may be that the cells are
dying via too much apoptosis. If many genes for cell cycle
progression are upregulated in disease, then an underlying cause of
disease may be too much cell growth.
If major pathways are discovered, it is easier to develop a drug that
can prevent or reverse the effects of disease. This leads to a more
rational method of drug development.
3. Measuring drug toxicity: the traditional method is to treat animals
(rabbits/mice) with drugs to see whether they become sick or die.
This method is slow, expensive, and harms animals.
Cell death or degeneration can also be examined by using
microarrays to analyze cultured cells. Many of the changes in gene
expression that occur during toxicity testing in animals can be seen
after similar treatment of cell cultures.
Microarrays can provide a rapid and inexpensive first screen for
new drugs. Decreases the need for large scale studies with animals.
4. Discovery of gene function: now that the first draft of the human
gene sequence is complete, the big question is “what do all of these
genes do”? There is information about 6,000 but the other 25,000 or
more are mainly unknown.
Microarrays can provide answers to this question.
1. One method is to introduce a gene into cultured cells using
transfection and then ask how the gene alters cell function.
Microarrays identify many other genes that are regulated by the
original gene and these can suggest activation of major cell
pathways (growth, differentiation, secretion, death, etc).
2. Use microarrays to study transgenic mice or gene knockout
mice to identify the gene’s function in the whole animal.
5. Analysis of genomic DNA for single nucleotide polymorphisms
(SNPs): small differences in the sequence of genes occur naturally
between different individuals. Many of these SNPs have been
carefully described and correlated with increased susceptibility to
specific diseases.
Microarrays containing all known SNPs can be used to rapidly
screen any individual for SNPs and susceptibility.
In the future, children may be screened for SNPs using array based
methods. This would allow physicians to predict which diseases the
person will suffer from and to suggest changes in lifestyle to
minimize future risk.
There are many types of arrays
1. High density oligonucleotide ‘chips’: these are constructed on
glass or silicon wafers by a process similar to photolithography.
The surface is altered to create artificial oligos that can hybridize
with the sample.
Affymatrix has pioneered this technology.
The advantage of this method is that the oligos can be arrayed at a
very high density (20 m apart), thus they can fit on a small substrate.
Whole genomes can be arrayed on several slides.
The disadvantage is that the technology is difficult to perform and
the arrays are very expensive. Only large well funded labs can
participate.
2. Nylon membrane filters: nylon membranes are the oldest array
technology. They are similar to Southern blots in that the DNA
samples are immobilized on nylon membranes for hybridization.
The membranes may be spotted with cDNAs, oligonucleotides, or
ESTs (expressed sequence tags are small bits of cDNA that have
not been characterized for function).
The RNA from the cell is reverse transcribed and labeled with P32.
The cDNA probe is hybridized to the membrane as a typical
Southern blot and is used to determine the level of expression for
each of the genes.
The advantage of this method is that it is cheap and easy. You need
no special equipment to read or scan the arrays. The disadvantage
is that not as many genes can be arrayed on a nylon filter.
3. Spotted oligonucleotide or cDNA arrays: robotic arraying
machines can place microspots of DNA on a glass slide (about 100
nL or the size of 100 m) that are too small to see with the naked
eye.
This is the most popular method.
•
It can be used to place tens of thousands of spots on 2-3
microscope slides (even all genes in humans).
•
It is versatile. The technique can also be used to make smaller
arrays of defined genes (death,growth, cancer genes).
•
Not as expensive as ‘chips’
•
You need a lot of cash to get started (the robot, all of the cDNAs
or oligonucleotides, an array scanner = 100-200K).
Problems with microarrays
1. Validation of results: One microarray experiment can yield
tremendous amounts of data, however, due to the small size of
spots, the data sometimes contains artifacts.
It is important to validate results from microarray experiments using
an additional method.
Northern blotting: is a good way to check for differences in gene
expression. However, it is labor intensive and only a few genes can
be screened at once.
RT-PCR: is also good but not as quantitative as Northern blotting. It
is the method of choice when not much sample is available (clinical
specimens).
Tissue arrays: Pathologists usually section a tissue in large pieces
before examining the cells under the microscope. In tissue arrays,
about 100-300 tiny tissue plugs are arrayed on 1 glass slide. This
allows the pathologist to rapidly analyze multiple samples for both
protein expression and morphology.
2. What to do with all that data? Microarray experiments quickly
generate more data than most scientists can cope with. Software
makes life easier.
•
Self organizing maps: software that looks at similarities and
differences between samples from different arrays. It groups
experiments or groups by a common characteristic.
For example, the self organizing map software was used to
compare different types of tumors and discover that some were
completely unrelated to others (this influences therapy).
•
Cluster analysis: software that can select those genes out of the
1000s expressed that go up or down in unison.
If a cluster analysis showed that cyclins, kinases, and DNA
polymerases were increased coordinately, it clearly indicates
that a specific growth pathway was activated.
Cluster Analysis and Self-Organizing Map Software
In Analysis of HPV-16-Infected Keratinocytes
1
2
3
4 5
cell cycle genes and TGF-b2responsive genes-downregulated
cluster
I
interferon-inducible
genes-down
regulated
cluster II
cell-cycle genes and NF-kB
responsive genes-up regulated
cluster III
genes induced in immortal lines
cluster IV
genes dependent on dsRNA
cluster
V
•
Data mining: maybe it is not necessary for a small lab to actually
perform arrays. Efforts are underway to develop a central
storage area for all data. This would contain billions of pieces of
information.
In the future you might be able to just mine existing data to
answer your questions.