By David N. Leff

Groping blindfolded, in a dark room, trying to find a black cat that may or may not be there.

Many a molecular geneticist must evoke this image when pursuing pedigree analysis and positional cloning to identify and isolate a gene in a putative chromosomal region, which may or may not be linked to a disease of interest.

Clinical molecular geneticist Thaddeus Dryja (DRY-zha), on the other hand, starts with the cat. In his approach, a gene is the point of departure rather than the end point.

Dryja directs the Ocular Molecular Genetics Institute at the Harvard University-affiliated Massachusettes Eye and Ear Infirmary, in Boston.

"Our gene-oriented approach," he told BioWorld Today, "is just an alternative to the more conventional phenotype-oriented method. In many cases, we don't care what chromosome the gene is on."

He explained: "You have a gene that you're interested in. You note that this gene is expressed only in the eye. So you reason that someone out there in the world probably has a mutation in this particular gene. And that person must have something wrong with his or her vision."

Search Aided By Thousands Of Samples

"I have in my freezer," Dryja went on, "white blood cells from 6,000 patients with eye problems. If I go through them, there's a good chance I'm going to find what disease, which problem, links up to that gene. So the pedigree analysis, the chromosome mapping, all of that is bypassed. We do that at the end, once we find something."

A paper by Dryja in the current Proceedings of the National Academy of Sciences (PNAS), dated Oct. 28, 1997, lays out his rationale and technical experience with this gene-oriented approach. It is one of the inaugural articles by recently elected members of the academy, and bears the title: "Gene-based approach to human gene-phenotype correlations."

A key feature of that approach is detecting mutations in the gene under scrutiny, then evaluating large numbers of unrelated individuals with a variety of signs or symptoms possibly related to the altered gene sequence.

"Our freezers have 12,000 or 13,000 leukocyte samples from over 6,000 patients," Dryja said, "collected since 1983. Most of them had very bad eye diseases; they were going blind, and needed assistance."

That frozen source of DNA holds genes related to the photoreceptors of the retina. Defects in them cause sight-threatening hereditary and sporadic diseases, such as retinitis pigmentosa (progressive degeneration of rod and cone receptors, with visual loss in early adulthood), congenital amaurosis (blindness before one year of age), and macular degeneration.

This organ-oriented strategy is central to Dryja's gene-focused modus operandi.

At least 50 genes, perhaps 100, he pointed out, can cause hereditary forms of blindness. Congenital deafness involves some 30 genes, known to be expressed in the inner ear. "So," Dryja suggested, "the assumption of one-to-one genotype-to-phenotype mapping is false."

He observed that "although the determination of chromosomal map positions is not an essential part of this gene-oriented approach, if a chosen gene happens to be from a region known to be implicated in a plausibly corresponding phenotype — in our case, the retina — it is wise to screen affected individuals early on."

His laboratory's current technology "can probably get 380 patients done in a week, for every exon [functional gene partial sequence] assayed."

The process begins, he elaborated, "by taking 95 leukocyte samples from as many unrelated patients, say, with dominant retinitis pigmentosa. They all have the same disease, as designated by ophthalmologists, but probably represent dozens of different gene loci." These (plus one blank control) go into 96-well microtiter plates.

DNA arranged on those plates then goes into freezers, sorted by retinal disease entity, awaiting PCR amplification of the gene's exons. Later, lab personnel will screen a given gene of interest, exon by exon, to detect mutations.

"The next step," Dryja recounted, "would be to sequence exons, one at a time. If we find a sequence that looks pretty pathogenic — say, a nonsense mutation or a frame shift — the next thing we'd do is call up that patient and ask if we could get blood samples from all the relatives. Then we'd check to see if the defect co-segregates with the disease."

As for potential clinical application of this gene-based approach, he pointed out, "My lab is geared up to find the gene defects. Every year we discover a couple or more genes that cause various eye diseases. But because retinal degenerations seem to be so heterogeneous, and with dozens and dozens of genes, we're going to have work for quite a few more years.

"I haven't gone intensively into the next step," he added, "how these gene defects are ultimately causing the blindness. For some of them, we've developed transgenic mice, to study the phenotype of animals with inherited or spontaneous mutations in the gene of interest. But that has not been the major emphasis in my lab. Our focus has been on trying to find which gene causes which disease."

Next? Gene-Based Assay Of Brain Disorders

Dryja views his gene-oriented strategy as an alternative to classical gene-mapping, not a replacement. But he predicted that "despite the successes and clear logic of positional cloning, it is beginning to face the law of diminishing returns. Most of the phenotypes that remain to be found," he pointed out, "occur in relatively small pedigrees that do not permit chromosomal mapping with satisfactory precision for a linkage-based approach."

In the future, he foresees wider adoption of the gene-oriented method, particularly in disorders affecting the brain. "Some 30 to 40 percent of our genes," he pointed out, "are used only by the brain — thousands of genes that are important in how we think and feel and act. Defects in them are going to give major problems, like psychiatric diseases or mental retardations.

"Take autism for example," he continued. "Its victims never reproduce because they are so disabled. So we rarely see the genetic pattern, but we know that there are genetic defects that cause autism."

Dryja concluded, "Gene disorders in the brain are one category that I think is ripe for the gene-oriented approach. And I know there are some groups already doing it." *