P&S Journal: Winter 1995, Vol.15, No.1
Rare Recipes in the Genomic Cookbook
The entire cookbook of genetic information exists in every cell of the body, but what makes a cell unique is its use of distinct recipes-genes-from the complete collection to perform specific cellular functions. The liver and kidney cells, for example, have the same 50,000 to 100,000 genes in the chromosomes in their respective nuclei and will use many of the same genes, but each cell also chooses different recipes from the whole collection to make, say, enzymes unique to their cell type.
Molecular biologists try to identify those genes, or recipes, that cause cells to be distinct. Dr. Marcello Bento Soares, assistant professor of neurogenetics, and his colleagues in psychiatry, genetics and development, and the New York State Psychiatric Institute have devised a way to make the search for cell-specific genes easier. Their method-called normalization of a cDNA library-should help speed the process whereby all scientists can identify important or disease-causing genes as part of the human genome project.
A cDNA library is an entire collection of recipes, or genes, used by a particular cell type. There are as many kinds of cDNA libraries as there are tissues in the body, including, brain, liver, and kidney versions. A cDNA library is the collection of genes that are used by a given cell type. cDNA is derived from messenger RNA, a nucleic acid that represents the "reading" of a chromosomal gene. An mRNA is an edited chromosomal version of the recipe or gene.
There are different numbers of mRNA and, ultimately, cDNA species in a cell type. The most prevalent class consists of 10 mRNA species, each represented by 5,000 copies per cell. The class of high complexity mRNAs, on the other hand, comprises 15,000 different species represented by one to 15 copies. If a scientist is searching for a gene present only in one to 15 copies, it is hard to find it in a library that contains both the rare and high-copy number genes.
In other words, if a cook made one photocopy of a recipe for chocolate chip cookies and 100 copies of a recipe for oatmeal cookies and placed them randomly in a pile, finding the chocolate chip recipe would be relatively difficult. But if the number of oatmeal cookie recipes could be reduced to 10, the chocolate chip recipe would be easier to find. Columbia scientists have been able to reduce the number of multicopy cDNAs in a cDNA library so it is easier to ferret out low-copy number cDNAs.
The researchers took advantage of the fact that denatured nucleic acids reassociate to form a DNA duplex at different rates. cDNA, like any other nucleic acid, can be disassociated to form two single and complementary strands of DNA. The molecule can be separated like a helical ladder split down its long axis into two halves. But the two halves of a ladder, or the single strands of the DNA, tend to want to make the ladder, or helix, whole.
cDNAs present in higher copy number reassociate faster than those present at a lower frequency. The scientists therefore were able to eliminate from the library cDNAs that reassociate to form the double-stranded DNA helix faster. As a result, they were left with a library enriched with the rarer cDNAs. In this second or normalized cDNA library, the frequency of rare and more common cDNAs equalized, improving the probability of finding low-copy cDNA.
The researchers have been able to reduce by one-thousand fold the number of high-copy number cDNAs in a human infant brain cDNA library. Many laboratories in the country, says Dr. Soares, have used the normalized library with success to find genes. As the method catches on, other laboratories will follow suit.