DNA Chips: A Laboratory
In The Palm Of Your Hand
By Ivan Amato
Once you get past kindergarten, the alphabet is pretty much old hat. Yet we use this little set of symbols to express lifetimes of thoughts,
jokes, dreams, and joys. Shakespeare's plays, Sweet Valley High novels, the U.S. Constitution, comic books-all written with the same set of squiggly lines. Pretty amazing. But unless you know how letters work
together to form words, sentences, and ideas, Shakespeare looks a lot like Sweet Valley-just gobbledygook.
A hand-held DNA Chip device, made by Nanogen,
Inc. The circles at the top are sample ports. The wires guide electric fields over the DNA array, located on the light blue diamond. [Photo courtesy Nanogen, Inc.]
Living things have a language, too, coded in the order of the nucleotide letters A, C, T, and G in
their genes. One of the great scientific quests of the past-that's the 20th-century was to understand this "language of life." First, scientists learned that
cells store their instructions for living in their DNA. Then, researchers figured out how cells convert the nucleotide sequences in DNA into the sequences of amino acids that make up proteins. Now, scientists
are busy reading out complete DNA sequences for whole organisms. They already have a rough draft of virtually the entire human genome, all 3 billion nucleotides of it.
Unfortunately, a genome's worth of raw sequence data is about as comprehensible as a shredded encyclopedia. You might pick out individual words, or even a few paragraphs, but you still can't readily
understand how the whole thing fits together.
In the 1990s, scientists developed a new tool for deciphering DNA called a "DNA chip," also known as
a "DNA microarray." It allows one scientist to collect more information about DNA sequences in an afternoon than an army of scientists could collect in several years using earlier techniques.
DNA chips promise to carry the science of understanding genomes to a whole new level, and to bring tools for getting DNA-sequence information out of research labs into doctors' offices, the better to
tailor-fit medical treatments to an individual's particular genetic makeup.
In fact, says Leroy Hood, a molecular biologist at the University of Washington in Seattle, DNA-chip
technology will be key to meeting one of the biggest scientific challenges of the coming century-the analysis of how all the genes in an organism work together as a very complex system.
The brain has about a trillion neurons, and about a quadrillion interconnections, says Hood. What we call "consciousness" somehow "emerges" from how all
these neurons interact. "We could study an individual neuron for 50 years, and that wouldn't tell us one iota more about the brain's emergent properties, because they arise from the network, not a single
cell," says Hood. "If we were to study each gene in isolation, we'd never know how the genome functions as a whole. DNA chips are the prototype global technology for genetics, because they let us look at
the behavior of thousands of genes at once."
How Chips Work
DNA chips come in many varieties. Some are "homemade" in scientists' laboratories, with glass
microscope slides and a robot arm wielding a high-tech fountain pen. Private companies are developing other techniques for mass production. But DNA chips all depend on the same basic principle:
complementary DNA stands stick together.
First, recall that a double-stranded DNA molecule can unzip into two complementary strands. Each of these can zip back together with its complementary
sequence. That could be either its old partner, or a new partner with the same sequence. The trick that makes DNA chips work is that you can tether a "new partner" to a flat surface.
Imagine a standard checkerboard, 8 squares on a side, 64 squares total. In each square, you tie down a different snippet of single-stranded DNA just three nucleotides long. You write down the sequence in
each square. (You can make 64 different sequence variations from three nucleotides-ACG, CGT, GTA, TAC, AAA, and so on- so there's just enough room for all the possibilities.)
Now imagine you have an unknown sequence, also three nucleotides long. To find out what this unknown is, set it loose on the array so that it wanders from square to square. When your unknown
sequence finds its complement, it sticks. To figure out your unknown sequence, all you have to do is find which square your unknown DNA stuck to. Because you know the sequence of the DNA you tied
down to that square, you know that the unknown sequence is the complement. (See diagram “How
What gives DNA chips their power in the real world is their flexibility, compact size, speed, and low cost.
Scientists can put not just a hundred but hundreds of thousands of distinct DNA sequences on a microscopic grid a few centimeters across. Then, using fluorescent molecular tags that light up when a
complementary strand binds to a particular spot, a person (or a robot) can read out which sequences on the chip find their complement in an unknown sample.
DNA chips can gather an incredible variety of data
very quickly. And because chips can be mass-produced, they will likely be very inexpensive in the near future. That will allow easy collection of genetic information from many, many individuals,
opening up all kinds of opportunities to help doctors diagnose and treat their patients.
One way DNA chips allow scientists to observe genes
working together is called "expression analysis." (Remember that to "express" a gene as a protein, cells first transcribe the gene's DNA sequence into a complementary mRNA copy. Then a ribosome
translates the mRNA sequence into the string of amino acids that makes up the protein. Cells constantly switch genes on or off as conditions change. To understand a cell's behavior in response
to a stimulus-the presence of a hormone, say, or a toxin, or some environmental signal-it would be handy to have a minute-to-minute reading of which genes are turned on.
DNA chips are just about perfect for tracking this kind of minute-to-minute change in gene expression. For example, Patrick Brown and his colleagues at Stanford University wanted to find out the details of
how yeast cells make spores. Other scientists had already determined the DNA sequence of every possible mRNA a yeast cell makes. So, Brown and his colleagues put the complements of each of these
possible mRNA sequences onto a chip. Then, they ground up a bunch of resting yeast cells, which of course contained an mRNA corresponding to each gene that was active at the moment the cells hit the
blender. Next, the researchers spread this mixture over the surface of the chip. Only the spots corresponding to genes that were actively churning out their mRNA lit up, because these were the only
spots on the chip that had found their complementary sequence.
This first experiment gave Brown and his colleagues a baseline. Next, they stimulated the yeast to form
spores (by taking away their food) and repeated their chip analysis six times over the next 12 hours. By looking at which genes turned on, and when, Brown and his colleagues got many new insights into how
yeast cells genetically shift gears to make spores.
But the significance of Brown and company's work goes way beyond yeast physiology-it paved the way for using DNA chips to see how dozens of genes work
in concert to change a cell's behavior.
Expression analysis has medical applications, too. For example, a team led by Eric Lander, director of the Whitehead Institute at the Massachusetts
Institute of Technology, announced last October that they used expression analysis-made possible by a DNA chip-to develop a test to classify different types of leukemia. (To choose the best treatment, doctors
need to know exactly what type of cancer a patient has.) These researchers looked at samples from about 50 patients already known to have one of two different kinds of leukemia. Then, using the patterns
of gene expression they found in the two groups, they correctly predicted which type of leukemia several patients had. In the near future, doctors may be able to use this test to decide which is the best
treatment for a new leukemia patient. Researchers also plan to develop similar tests to match treatments to patients for other kinds of cancer, too.
Mapping Our Differences
Pick any two people in the world, and you would find their DNA is 99.9 percent identical. The remaining 0.1 percent is the genetic basis of all of humanity's differences, from the shape of our faces to the way
some people get cancer to the fact that some patients respond to a certain drug while others don't. Scientists are now starting to use DNA chips to map out tiny one-letter variations in the
3-billion-nucleotide human genome. These pinpoint differences are called "single nucleotide polymorphisms," or SNPs. Identifying them will help researchers understand the basis for human variation.
But to map SNPs, you need a different kind of chip. For expression analysis, you use a chip containing all possible genes. For SNP work, you make a chip with many, many possible variations of one gene. Then
you take a DNA sample from the person you want to test, use PCR to make multiple copies of the gene you're interested in, and put this "amplified" sample
on the chip. The spot that lights up will correspond to the particular sequence variant the person has. Because the test is quick and not too expensive, you can do many of them. Then, you correlate different
outcomes-response to a certain drug, for example, or the probability of getting heart disease-with the different genetic variations.
Francis Collins, director of the National Human
Genome Research Institute in Bethesda, Maryland, is enthusiastic about SNP analysis. "There are only about 200,000 functionally important variants [SNPs] in the human genome that have reasonable
frequencies," he says. "Nearly all of the genetic contributions to diabetes and heart disease and hypertension and all of the common illnesses are found in those 200,000 elements." Moreover, says
Collins, once researchers know which SNPs correlate with higher risk for disease, people with these traits will be able to take extra steps to avoid getting sick.
This might allow "medicine to move from its present mode, where we spend most of our resources treating people who are sick, to a preventive strategy, which is individualized," says Collins.
Big Power, Big Responsibility
DNA chips will help scientists make sense of genetic information. Medical applications are on most people's minds, but the same technologies can be
used for everything from confirming lineages of racehorses to teasing out evolutionary relationships between closely related species.
But as the power of chips and genetic science grows,
questions that society must answer pop up right and left. Should employers use genetic information in hiring decisions? How about insurers who may want to avoid insuring people at high risk for certain
diseases? How about a zealous political group trying, say, to portray an opposing candidate as having a high risk of dying of a heart attack? Today, it's impossible even to list all the questions, let alone
answer them. (Do the Social Impact section for insight into one interesting question.) It will take laws, regulations, constraint, and wisdom to ensure that the good consequences of the genetic revolution
outweigh the bad, say many researchers, including Lander. "I know of no other field that is more exciting, or in which it is more important for us all to imagine the future," he says.
More on the Web
There’s plenty to find about DNA Chips on the Web, especially if you are a scientist. Here are a few links to get you started.
Partrick Brown’s lab has a homepage, with a good list of links. http://cmgm.stanford.edu/pbrown/
A graduate student in computer science at Washington University named Jeremy Buhler has created a tutorial about DNA Microarrays that you might find helpful. http://www.cs.washington.edu/homes/jbuhler/resear ch/array/
A magazine called IVD Technologies has published an in-depth, three-part article on DNA chips. http://www.devicelink.com/ivdt/archive/98/09/009.ht ml
Affymetrix, Inc., a leading maker of DNA chips, has a web site. http://www.affymetrix.com/
Nature Genetics, a leading scientific journal, has a good collection of aritcles on DNA Chips. http://www.nature.com/cgi-taf/DynaPage.taf?file=/n
Here’s an animation produced at Davidson University in North Carolina, depicting DNA Microarray technology. http://www.bio.davidson.edu/biology/courses/genomi cs/chip/chip.html
Want to try your hand at making microarrays? Here’s one site with information on a low-cost do-it-yourself system.