By David N. Leff

Science Editor

The hourglass-shaped centromere, which pinches the midriff of every human chromosome, carries no genes. Its job is to honcho the proper divvying up of chromosomal DNA during cell division.

"While it's been known since the early years of this century that chromosomes carry genes," said geneticist Huntington Willard, "until now the complexity and size of normal chromosomes has limited our ability to analyze their structure and function."

One of the most complex parts of that overall complexity is the centromere.

"Our successful creation of functional centromeres and incorporation of them into human artificial chromosomes [HAC]," Willard observed, "were the critical achievements enabling the stability and normal behavior of the chromosomes during cell division."

Willard, who chairs the department of genetics at Case Western Reserve University, in Cleveland, is senior author of a paper in the April issue of Nature Genetics, titled: "Formation of de novo centromeres and construction of first-generation human artificial chromosomes."

Molecular geneticists have been frustrated in their efforts to create human artificial chromosomes. Instead, they must make do with yeast artificial chromosomes (YAC) in cloning experiments. These are patterned on the genomic DNA of Saccharomyces cerevisiae — baker's yeast.

"YACs have been instrumental in allowing the Human Genome Project to move forward," observed cancer biologist John Harrington, the article's lead author. "They've been used to date largely as a cloning vector. Until very recently," he pointed out, "YACs have been the only vector that allowed very large pieces of human DNA to be cloned. By large, I mean from several hundred kilobases up to, and perhaps exceeding, a million bases in size."

Why HACs Lagged Behind YACs

Human chromosomes, which range from about 50 to 250 megabases in size, are 100-fold larger than their yeast-cell counterparts.

"One of the major obstacles confronting people interested in cloning human artificial chromosomes," Harrington told BioWorld Today, "has been an inability to clone a functional human centromere." He counted the reasons why:

"First, it's extremely large, unlike the yeast centromere, which is only 125 base pairs in length.

"Second, the human centromere is made up of extremely repetitive sequences of alpha-satellite DNA. This consists of a 171-base-pair, head-to-tail tandem repeat, iterated over and over again for several thousand base pairs long." Harrington pointed out that "the exact function of this class of DNA is largely unknown, because until now we've never had a system that allowed us to dissect the function associated with these centromeric repeats."

A principal goal of Case Western's two-year HAC-building effort was to provide vectors for gene therapy free of the drawbacks that hamper viral and non-viral DNA vehicles. "Viral vectors," Harrington pointed out, "suffer from serious limitations, such as unstable gene expression, unwanted immune responses, occasional generation of infectious viral particles, and stringent size restrictions, which allow only small genes to be placed in the viral plasmid and delivered to the target tissue."

If genes are the "software" of chromosomes, then their "hardware" consists of the gene-free, wasp-wasted centromere near their middle, and telomeres at each end. The latter are far-out (literally) repetitive regions of DNA, capping both tips of every chromosome. Telomeres act like boundary markers or bumpers to protect the chromosome from unwanted hook-ups, end-to-end, with other chromosomes.

As Harrington recounted, to construct their human artificial chromosomes, he and his co-authors first synthesized arrays of alpha satellite DNA. Then, to this centromeric genetic material, they added telomeres and large fragments of random human genomic DNA. The latter presumably contained origins of replication — sequences that regulate the controlled copying of a chromosome's total DNA during cell division. They kick-start the replication process, and thus constitute the business part of a gene-therapy vector.

"This approach," Harrington explained, "simply allowed us, in essence, to make a DNA library within the cells we transfected."

Finally, they inserted this mix into human tumor cell lines. There, like building blocks, these independent elements assembled into miniature versions of human chromosomes.

As for the actual assembly of the HACs inside those cells, Harrington said, "Once we introduced the several DNA ingredients into the cell, they found their way to the nucleus, where they assembled into various HAC combinations. Genes present in both the genomic DNA and in our constructs were then expressed. At 24 to 48 hours after that transfection took place, we put the cells under drug selection, which killed off the vast majority of them.

"Among the survivors," Harrington went on, "were putative human microchromosomes, which persisted inside dividing cells for more than six months of culture. At that point they had all the characteristics of normal human chromosomes, except for size. These HACs are only a tenth as large as native human chromosomes."

Readying HACs For Gene Therapy Vector Duty

To bring their HACs to the clinic, and to market, three of the paper's co-authors founded Cleveland-based Athersys Inc. in 1994.

Harrington is the company's vice president for research and development. "Athersys was a spin-off of Case Western," he explained, "but is now an independent, privately financed company, which has a very good working relationship with the university."

Geneticist Gil Van Bokkelen is the firm's president and CEO and Willard is a member of its scientific advisory board. All three are inventors on an allowed U.S. patent which, Harrington said, "covers methods of cloning long alpha-satellite DNA sequences up to about 200 kilobases in length, then expanding them in vitro to a megabase long."

Two other patent applications are pending.

Once their prototype synthetic microchromosomes (HACs) are refined for use, Harrington said, "the first thing we'd want to demonstrate is that they can be introduced into primary, freshly isolated human cells. Next, to show that they could go into stem cells, which can be used in various ex vivo gene therapy applications."

The project leaders are also "very interested in developing animal models that would allow us to determine the safety and efficacy of our prototype HACs in a non-human system."

Harrington foresees their use "in gene therapy human clinical trials -- the ultimate goal — we hope, in the next couple of years." The first such applications, he concluded, "are expected to be in treating disorders such as AIDS, sickle-cell anemia, beta-thalassemia, cystic fibrosis and muscular dystrophy."

An editorial accompanying the Case Western paper in Nature Genetics called it "an important landmark in terms of constructing HACs. This is the first time that a mitotically and cytogenetically stable artificial chromosome derived from transfected DNA has been generated." *