Eric Lander’s group was the first to show that gene expression arrays could be used todistinguish between two types of leukemia, acute lymphoblastic leukemia (ALL) and acutemyeloid leukemia (AML). In a paper published in 1999 in the journal
the scientists usedexpression results from 6,800 human genes measured on arrays to accurately predict whether apatient had AML or ALL. Until then, doctors distinguished between the two cancers by using abattery of expensive tests that could cost critical time. Since that study, a number of studieshave shown similar results with different types of cancers.
DNA Sequence Arrays
Other types of arrays can be used to measure genetic variations among individuals. Forexample, single nucleotide variations, or SNPs (pronounced “snips”), which occur throughoutthe genome, can be detected by using genotyping arrays or SNP chips. These types of arrayscarry all the possible variations of one gene or several genes in a grid pattern. A DNA sample isextracted, multiple copies of the gene or genes of interest are generated using polymerasechain reaction (PCR), and the sample is then applied to the chip. The spots that light upcorrespond to the particular gene variants the individual has.Like the expression arrays, these types of arrays provide basic information, such as the rangeof variation in human genomes. But they can also have clinical applications. By looking atseveral SNPs at once, researchers can identify “SNP signatures” associated with a specificdisease or a response to a drug. Individuals at risk for a particular disease could then be testedfor the telltale signature.A newer technique called microarray comparative genomic hybridization (microarray CGH) hasbeen developed to identify large regions of DNA that are either missing (deletions) or presentin more copies (amplifications). Rearrangements in the DNA, such as deletions andamplifications, are often involved in diseases such as cancer. Other types of arrays are able todetect DNA sequences that bind to proteins or sequences that are chemically modified in thegenome, such as by methylation or acetylation.
A different but conceptually similar approach is being applied directly to proteins. Scientistshave developed microarrays that can be used either to identify and quantify thousands of different proteins at once or to find associations between different kinds of proteins andbetween proteins and other molecules. These types of arrays are collectively referred to asprotein arrays.Protein arrays that are used to identify proteins typically consist of many antibodies arrayed ona glass or plastic slide. Each antibody can bind to a different target protein. When a mixture of proteins—for example, in a blood sample—is applied to the array, the proteins recognized bythe different antibodies will bind to the array. Bound proteins can be detected either by addinga second antibody tagged with a fluorescent molecule or by chemically labeling all the proteinsin the blood sample before adding the sample to the array. Each bound protein can thereforebe detected as a signal on the array, and the intensity of the signal roughly represents theamount of protein in the blood sample. This type of array, like a gene expression array, can beused to generate “signatures” for different cell types and tissues.Another type of protein microarray is used to glean insights into the function of differentproteins by looking at the molecules they bind to. For this application, the proteinsthemselves, rather than their antibodies, are arrayed on a slide. Stuart Schreiber’s group wasone of the first to show that more than 10,000 different proteins could be “stuck” to a singleglass microscope slide and still retain their biological activity. In a typical experiment,