Many different variations in human metabolism and behavior might result from how many copies of a gene someone has, rather than from mutations or single-letter alterations in the genetic code, a study suggested.

A survey of the genomes of 20 normal individuals, carried out at Cold Spring Harbor Laboratory, showed that there were 221 differences in the numbers of copies of individual genes. In all, 70 genes varied in the number of copies held by different individuals. Those included genes involved in neurological function, regulation of cell growth, regulation of metabolism and several disease genes.

Michael Wigler, a professor at New York-based Cold Spring Harbor, told BioWorld Today: "The extent of this variation in normal humans had not been documented before. This is a new source of variation in humans, and one that people who analyze single nucleotide polymorphisms don't see."

Wigler postulated that 5 percent to 10 percent of genetic disease in humans could be caused by variations in gene copy number. "This means there are many disease-associated lesions that have not been observed properly before," he said.

Wigler and his team reported their findings in the July 23, 2004, issue of Science in a paper titled: "Large-Scale Copy Number Polymorphism in the Human Genome."

The group, which has collaborators at centers elsewhere in the U.S. and Sweden, had been studying gene copy number (also known as copy number polymorphism) in tumors and in people with genetic diseases. But they realized that they needed to establish a baseline by first looking at the DNA of normal people.

They used a method called representational oligonucleotide microarray analysis (ROMA) - a high-resolution technique for measuring the number of copies of genes - to examine the genomes.

Wigler said: "Our study shows that people have inherited different numbers of copies of some genes, and this variation is likely to lead to variations in levels of expression, which is likely to lead to changes in phenotype. These differences in phenotype could be related directly or indirectly to susceptibility to disease or resistance to disease."

The study found that, between any two individuals, there was an average of 11 copy number polymorphisms, with an average length of 465 kb and a median length of 222 kb. At least five had been described previously, but most copy number polymorphisms identified had not been reported before. About half of the copy number polymorphisms occurred in several individuals.

Places where copy number polymorphisms were present were widely distributed throughout the genome, although certain points had clusters, possibly representing "hotspots" of copy number variation, the authors wrote. The study found no copy number polymorphisms on the X chromosome, but the team suggested that might be because 17 of the 20 samples came from males.

Regions of the genome already known to give rise to specific inherited abnormalities when mutated also were home to many copy number polymorphisms. Those polymorphisms were common in areas where chromosomal rearrangements are associated with Prader-Willi and Angelman syndromes, cat eye syndrome, DiGeorge/velocardiofacial syndrome and spinal muscular atrophy.

Wigler and colleagues wrote: "These copy number polymorphisms are not directly implicated in the above diseases, but they may reflect the instability of these genomic regions."

The findings of the study back up much of what is known about some of the genes involved. For example, researchers already knew that some neurological disorders are caused by inheritance of different numbers of copies of certain genes. One of the people analyzed in that instance had a deletion of a gene known to cause an autosomal recessive neurological disease called Cohen syndrome.

Several genes found to be prone to copy number polymorphism in the study also are involved in neurodevelopment. Another is thought to have a role in leukemia.

Some of the genes highlighted, by contrast, are known to affect normal human phenotypes. The authors noted: "For example, we observed triplication of the neuropeptide-Y4 receptor . . . a gene that is directly involved in the regulation of food intake and body weight. Thus, a relationship between copy number polymorphism and susceptibility to health problems, such as neurological disease, cancer and obesity, is an intriguing possibility."

Wigler and his team plan to expand their baseline data by analyzing the genomes of more "normal" people.

"After that, we will be using the ROMA methodology to look at cancers and people with mutations," Wigler said.