By David N. Leff

DNA has become the celebrity molecule of the decade.

It solves crimes. It identifies old bones — from the Russian Czar to the mummified Pharaohs of Egypt to woolly mammoths. DNA sequence matching settles paternity suits and traces mutant cancer-causing genes back through generations of afflicted families.

This versatility of deoxyribonucleic acid (DNA) rests on the fact that every human carries his or her unique signature sequence of nucleotides, handed down from parents to children back through time.

How far back in time does this hand-me-down chain of DNA extend?

Certainly, to the origins of life on earth. Some scientists push it even further in the past, to a pre-DNA world chemically micromanaged by RNA — ribonucleic acid.

"There was chemical synthesis going on in that RNA world," observed molecular evolutionary biologist Margaret Saks. "A lot of people think about it as a cell-free time. A time where molecules are being made, but before we really have organisms the way we know them now. That RNA world is a very hard time to envision."

Saks is a post-doctoral fellow at the California Institute of Technology, in Pasadena. Her research focus is tracking the origin of the genetic code and how structure-function relationships affect the evolution of genes. She is lead author of a paper in today's Science, dated March 13, 1998, titled: "Evolution of a transfer RNA gene through a point mutation in the anticodon."

At this point, a brief detour into Biotech 101 is in order.

Living organisms, from viruses to people, consist largely of proteins. Proteins consist of amino acids, of which there are 20 varieties.

When an organism's DNA feels the urge to manufacture a protein, it converts one strand of its double helix into a variety of RNA molecules. Initially, transfer RNAs (tRNAs) and certain enzymes round up the required amino acids teeming in the cell's cytoplasm, and attach them to the tRNAs. Later, these download the amino acids at the ribosome, based on the precise pairing of the tRNA anticodon and the messenger RNA (mRNA) codon. Then the ribosome's protein-making translation machinery adds the amino acid to the growing protein polypeptide chain.

"Given the central role of tRNA in translation," Saks' article points out, "and the near-universality of the genetic code, it seems likely that tRNA genes were among the earliest genes to arise."

To envision how these genomic sequences may have evolved, she studied the evolution of one of the 20 isoaccepting tRNAs. These go out and carry back the nascent amino acids in the cytoplasm, as needed.

Study Employed Genomic Paleontology

As ordered by the genetic code, each amino acid is encoded by a three-base-pair codon in the DNA sequence. There are 45 to 50 different tRNA isoacceptors, which recruit one or another of the 20 amino acids.

"What we came up with," Saks told BioWorld Today, "is a recruitment model based on a point mutation that converts the isoacceptor from one function to another. For example, if an isoacceptor earmarked to bring in the amino acid arginine sustains a point mutation at a site in its molecule, it converts it into a tRNA threonine. So it basically has taken on a new function. It's still a tRNA, but it has a new amino-acid identity."

"That result," she continued, "speaks to the origin of the genetic code — how the amino acid assignments grew during evolution. The tRNA basically recapitulates the code. Its anticodon is complementary to the codon we see in mRNA [which transfers the coding sequence from the cell nucleus to the ribosome]. The amino acid that's being carried by that tRNA must correspond to the anticodon, in order for any gene to be translated with fidelity."

Citing the role of point mutations in this evolutionary process, Saks observed: "People tend to think of mutations as deleterious, or at best silent. That comes from thinking about mRNAs, the structural or functional RNAs that are translated into proteins.

"But here it turns out that a mutation in a very critical region of the molecule is not silent, it's not deleterious. Rather, it's completely acceptable, and it changes the function."

Saks continued, "If our recruitment model is generalizable — and that remains to be seen — what we've done is taken a group of nucleotide sequences that, let's say, on Day One, prior to the mutation, was called the tRNA arginine. Now the mutation happened. On Day Two, if we as scientists simply looked at the sequences, because of that mutation in the anticodon, we would say: 'That's a tRNA threonine,' because it would be identified by its anticodon. But it now has all the sequence characteristics that it had yesterday as a tRNA arginine, so it's dragging sequences from one gene category to another."

To carry out their molecular paleontology, Saks and her co-authors turned to the genome of Escherichia coli. "Fortunately," she observed, "the tRNAs of E. coli had been characterized. Its chromosomal locations and their sequences were known. And we knew that there was on that bacterium a single gene for the tRNA threonine triplet codon — uridine-guanine-uridine."

The team used homolgous recombination to disrupt that gene. "We knocked out its tRNA threonine. And that's a lethal mutation. Without it, the cell is dead."

To restore it, they took their tRNA arginine, in which they had engineered a point mutation, and inserted it via a plasmid into the E. coli cells. "We asked," Saks recounted, "whether the expression of that tRNA variant that we made could compensate for the knockout. The answer was yes.

"If there is any practical utility in this experiment," she suggested, "it's with respect to the basic problem of RNA-protein interactions. These drive a lot of processes in the cells. The guts of our system had to do with the aminoacylation — addition of the amino acid to the 3' terminus — of the tRNA, and that's determined by an RNA-protein interaction.

"It might be possible to use this same logic to eliminate, rather than change, translation interactions, such as getting rid of pathogens." *