Enzymes illuminate switch to a DNA world.

In one scenario of how life on Earth began billions of years ago, RNA was a biochemical jack-of-all-trades. The molecule could not only transfer informa- tion in cells, as it still does, it could also catalyze reactions and perhaps even make copies of itself (SN: 8/10/96, p. 93).

At some point, however, various RNA functions got parceled out to other molecules, eventually leading to the strict division of labor that exists today. Now, except in some viruses, DNA acts as the blueprint for genetic information. Enzymes catalyze reactions, and RNA only carries information from DNA to the cellular machinery that synthesizes proteins.

Two studies in the Jan. 21 Proceedings of the National Academy of Sciences suggest what may have happened during that transition from an RNA to a DNA world. By examining important enzymes in present-day organisms, scientists have shed light on what these enzymes may have been like long ago.

In one of the studies, an international team of researchers analyzed an enzyme from Pyrococcus furiosus, an archaebacterium that thrives in boiling- hot, oxygen-poor conditions (SN: 3/11/95, p. 150). That enzyme, ribonucleotide reductase, converts the building blocks of RNA into the building blocks of DNA.

Ribonucleotide reductases from other organisms fall into three distinct categories, based on their amino acid sequences and biochemical functions. The P. furiosus reductase, however, "falls between categories," says Frank T. Robb, acting director of the University of Maryland's Center of Marine Biotechnology in Baltimore. "The reaction mechanism goes one way and the [amino acid sequence similarity] goes another."

Because the enzyme combines characteristics from all three classes of reduc- tase, it appears to be a kind of "missing link," Robb says. "The term is over-used, but it really does apply." The modern forms of the ribonucleotide reductase probably evolved from some- thing similar to the P. furiosus version, he says. That original reductase would have been one of the keys that opened the door to a DNA-centered world. The team's next project is to synthesize enough of the enzyme to determine its three-dimensional structure, Robb says. The enzyme is very difficult to purify from P. furiosus, so he and his colleagues, Marc Fontecave and Joan Riera of the Joseph Fourier University in Grenoble, France, and Robert Weiss of the University of Utah in Salt Lake City, are trying to get a common bacterium to produce it in large quantities.

The determination of another enzyme's three-dimensional structure was essen- tial to the second study, conducted by Stephen P. Goff and his colleagues at Columbia University and at Rutgers University in Piscataway, N.J. They looked at a leukemia virus' DNA polymerase, an enzyme that connects the building blocks of DNA.

A model of the polymerase's structure, put together by Wayne A. Hendrickson of Columbia, showed which part of the enzyme recognizes the DNA building blocks. The researchers found that exchanging just one of the amino acids in that part enabled the enzyme to synthesize RNA as well as DNA. They replaced a bulky phenylalanine with a smaller valine, which gave the RNA building block a more comfortable fit.

"These enzymes [RNA and DNA polymerases] share a lot of similarities in structure, and so rather subtle changes-in this case just one amino acid-are enough to alter that specificity," Goff says. The DNA polymerase is a reverse transcriptase, the kind of enzyme that HIV and other retroviruses use to copy their RNA-encoded genetic information into an infected cell's DNA (SN: 5/9/92, p. 308).

"The origin of reverse transcriptase is murky and controversial," Goff says.

"[Transcriptases] could be indeed very primitive and related to some of the very earliest forms of replication. The alternative view is that they devolved from some more advanced polymerase." The results of the study support the former scenario, suggesting that all polymerases evolved from the same ancestor molecule, the authors assert.

The mutant polymerase the researchers created is "rather poor," Goff says. It makes RNA chains that are only about six building blocks long. "We're trying to make it better through other mutations," he adds.

The researchers are also characterizing dozens of other mutant forms of the polymerase to learn more about its various features.