The science of genomics faces two major quantitative challenges. First, there are very large numbers of genes in eukaryotic genomes. Second, there are myriad distinct patterns of gene expression in different tissues at different times. For example, the cells of a skin cancer at its early stage may have a unique mRNA "fingerprint" that differs from those of normal skin cells and cells from a more advanced skin cancer. In such a case, the pattern of gene expression could provide invaluable information to a clinician trying to characterize a patient's tumor.
Patterns of gene expression can be analyzed with a thumbnail-sized invention called a DNA microarray ("gene chip"), one of the most powerful new tools to emerge from genome studies. A DNA microarray is made with thousands of DNA sequences attached to the microarray in a grid pattern. The attached sequences act as probes and tell a researcher whether a test sample contains a particular DNA or RNA sequence.
DNA microarray technology permits an investigator to rapidly determine which genes are expressed by a cell or tissue. To understand how a DNA microarray works, let's consider an experiment that compares the pattern of gene expression in tumor tissue compared to normal tissue.
When a gene is expressed, it is transcribed into mRNA. The mRNAs from the two tissues are isolated and converted into complementary strands of DNA (cDNAs) by the enzyme reverse transcriptase. To distinguish between the two pools of cDNAs, the molecules are fluorescently labeled—red for the tumor-derived pool and green for the cDNAs derived from the normal tissue.
After the RNA is removed, DNA microarrays are used to compare the two cDNA samples. DNA microarrays can contain 60,000 or more different DNA sequences attached in microscopic spots to a glass slide. The different DNA sequences are oligonucleotides of about 20 bases in length. The oligonucleotides represent tiny but unique regions of genes in the genome.
The cDNA samples are mixed together and added to the microarray. cDNAs that are complementary to oligonucleotides on the microarray will bind (or hybridize) with the DNA, and thereby stick to that location on the slide. Unbound cDNAs are washed away.
A scanner detects patterns of hybridization by sensing the fluorescent signals. In this example, a red spot indicates that expression of the gene is higher in the tumor tissue compared to normal tissue. In contrast, a green spot indicates that expression of the gene is higher in normal tissue than in the tumor tissue. A yellow spot indicates that the gene is expressed equally in both tissues. If the gene is not expressed in either tissue, the spot will not fluoresce.
Because each area of the microarray contains a known DNA sequence, corresponding to a known gene, the identities of the hybridizing cDNAs can be determined. Using these data, investigators can establish which genes are expressed differently in the cancerous tissue, and thus may be able to design better treatment strategies.
For breast cancer, scientists have identified 70 genes whose expression differs dramatically between tumors from patients whose cancers have recurred and tumors from patients whose cancers did not recur. From this information, gene expression signatures have been identified that are useful in clinical decision-making: patients with a good prognosis can avoid unnecessary chemotherapy, whereas those with a poor prognosis can receive more aggressive treatment.
DNA microarray technology is popping up in a variety of applications around the world. Each time the technology is applied to a biological question, it typically replaces a much slower, more expensive method of obtaining answers, or it simply opens up a totally new avenue of inquiry.
In the accompanying animation, we presented an application in which the gene expression profile of a cancerous tissue was compared to that of a normal tissue. Researchers are currently performing this type of experiment on an enormous range of cancers—cancers of the breast and prostate, for example, and cancers that may be either relatively benign or aggressive.
As researchers are placing cancers into more specific categories (based on gene expression profiles), they are also cataloging the relative success or failure of a patient's treatment strategy. As this bank of information grows, it will become an invaluable resource to patients and doctors. In the future, a doctor may be able to read a cancer's gene expression profile and then know immediately which treatment strategy will work the best.
As another example of this technology, a company in France recently announced that it will use DNA chips to test drinking water safety. The DNA chips for this purpose are made with gene sequences from a variety of disease-causing microbes. To perform this procedure, researchers isolate microbes from a sample of water, extract and label the DNA of the microbes, and then incubate the labeled DNA with the DNA chips. A specific fluorescence pattern on the chip indicates that a particular species of microbe is present in the water. With this technology, the company can quickly test for a number of species at the same time, rather than performing individual, time-consuming tests for each species.