Malcolm Simons proposes codon changes to identify DNA.

By Graeme O’Neill

July 14, 2008 | MELBOURNE — In the fall of 2001, five people died in the anthrax attacks in New York and Boca Raton, Florida. Although the crime remains unsolved, the Bacillus anthracis strain used in the attacks was virtually identical to one held at the U.S. Army Medical Research Institute for Infectious Diseases (USAMRIID) in Fort Detrick, Maryland. The episode highlights the need for a means of readily distinguishing modified strains of bacteria, viruses, cell lines, and transgenic plants and animals, from naturally occurring forms.

Australian immunologist Malcolm Simons—who made headlines a few years ago with his controversial “junk DNA” patent (see, “Simons Says”)—is not the first to propose a method for “watermarking” proprietary DNA, but his idea is simple… and waterproof.

An obvious way to watermark a gene or genome is to insert a distinctive, non-coding DNA sequence. J. Craig Venter recently watermarked the name of his research institute into the genome of the prototype of the world’s first synthetic living organism, a genetically minimal Mycloplasma. His group inserted a synthetic sequence of triplet codons to spell out “VenterInstitvte”, along with his own name and the names of several colleagues, using the single-letter abbreviations of the 20 amino acids. (The amino acid “alphabet” lacks six of the 26 letters in the English alphabet, hence Venter’s substitution of a valine (V) in “Institvte”.)

“The issue is conserved sequence,” says Simons. “Everyone understands that it is only possible to hide cryptic information in conserved sequences. There are ultra-conserved sequences in non-coding DNA, but within the protein-coding region of a gene, any change in a single nucleotide may change the function of the protein.”

Simons’ solution, called TriPrint, exploits the variation available in synonymous codons within protein-coding exons to “imprint” the watermark within the gene, without changing the encoded protein. The four-base DNA code yields 64 unique triplet codons, which specify 20 amino acids. Each amino acid has between two and six alternatives, and even the universal “stop” codon that terminates genes has three variants: UAA, UAG and UGA. TriPrint would combine synonymous variants of the stop codon, and serine codons in each protein-coding exon, to create unique watermarks. (Altering just the first or second nucleotide in the triplet provides five alternatives to any wild-type serine codon).

Patented System
“The stop codon is least the mutable codon, and the least mutable amino acid of the 20 has got to be serine,” he said. “You would have to change two nucleotides in a transversion manner and the probability of simultaneous substitutions in the same serine is extremely remote.”

He said that a watermark pattern comprising just the first occurrence of serine in each exon of the gene, plus the stop codon, would provide many more unique permutations than required to distinguish between all naturally occurring variants of the gene. Indeed, even in a small gene with just a handful of exons, the number of potential permutations rapidly approaches Avogadro’s number.

“In practice, it might be sufficient just to modify the first serine codon in each exon of the gene, plus the stop codon,” Simons said. “The particular combination of serine variant and the stop codon could even serve to identify the company that created the gene.”

Simons’ TriPrint patent, filed with the Australian Patent Office on April 17, relates specifically to transversion mutations—changes that substitute a purine (A or G) for a pyrimidine (T or C) at any of the three codon positions. “The improbability of a dual mutation in the first two nucleotides can be regarded as infinitesimally small,” he said. “There is no need for an error-checking mechanism.”

Simons proposes that DNA watermarks would be recorded in a central database. “Given the concerns of some consumers about ‘contamination’ of conventional crops by pollen from genetically modified crops, TriPrint provides a way of individually watermarking every transgenic variety,” he said.

 A DNA watermark in the presumptive Fort Detrick strain of B. anthracis might not have prevented the anthrax attack, but Simons said the ability to rapidly trace a transgenic microbe to its source might give terrorists pause for thought. 

_________________________________________________

This article appeared in Bio-IT World Magazine.

Subscriptions are free for qualifying individuals.  Apply Today.