Variance in Anthrax Strains Could Crack Case
Monday, February 18, 2002
By Rick Weiss,
Washington Post Staff Writer
The pudgy, rectangular bacterium that causes anthrax has long had a reputation for being genetically unimaginative. Unlike other kinds of bacteria, which have thrived over time by mutating, diversifying and trying new things in different parts of the world, Bacillus anthracis has seemed stuck in its ways. Until recently, DNA fingerprints of specimens retrieved from England, the United States, South Africa and other countries were virtually indistinguishable.
Now, however, scientists are finding they just weren't looking closely enough. Newly developed high-resolution tests are revealing subtle genetic variations that are apparently unique to many strains and sub-strains of B. anthracis. And as researchers learn precisely how the bacterium's genetic code varies in tiny ways from one laboratory batch to another, they are growing hopeful that they may be able to help federal authorities identify the source of last fall's bioterrorist attacks, which left five people dead, hobbled the U.S. postal system and panicked the nation.
The new findings are the result of a fresh collaboration between two teams of scientists who had toiled quietly and independently for years with the seemingly academic and arcane goal of understanding B. anthracis genetics. Now those two previously low-profile groups – a Rockville biotech company and a group of Arizona scientists – are suddenly in the national spotlight, flush with federal funds and feeling the pressure to break open an investigation that the FBI has conceded is lacking in hot leads.
At the core of the scientists' investigative approach is a key biological truth: All living things gradually accumulate mutations in their genetic codes. In bacteria, offspring typically inherit those changes and add new ones of their own, creating a cumulative and permanently inscribed genetic record of the organism's evolution. By comparing the genetic codes of different individuals, and seeing which ones have similar mutation patterns, scientists can determine which isolates share common familial roots and therefore are most closely related to each other.
The mutations that scientists are looking at are on the double-stranded molecules of DNA that carry the instructions for survival, growth and reproduction in most cells. In the case of B. anthracis, most of that molecular blueprint for life is encoded on a single circular strand of DNA made of 5.1 million subunits, or "bases." (By comparison, the human DNA blueprint contains about 3.1 billion bases.) Each B. anthracis cell also contains two small loops of DNA – one made of 90,000 bases and the other made of 185,000 bases – that carry the instructions for making the toxins and other products that together make B. anthracis so deadly.
The first hint that some strains, or far-flung family lines, of B. anthracis might vary from one another in subtle but reliable ways arose in 1996, when researchers in North Carolina compared a small region of DNA in two strains of the bacterium and noticed something strange. A sequence of 12 DNA bases was repeated four times in every bacterial cell belonging to one strain, but that 12-base sequence was repeated only twice in cells belonging to the other strain. A year later, Paul Keim of Northern Arizona University in Flagstaff reported that he had discovered seven additional regions of the B. anthracis genome that also differed consistently from strain to strain.
Those eight points of comparison have together proven capable of differentiating among several different strains of B. anthracis, but additional hallmarks will be needed if scientists are to differentiate among various substrains, or "isolates" – bacterial populations that belong to the same strain but have been growing and evolving independently in different laboratories.
Investigators know, for example, that the B. anthracis used in the attacks belongs to the so-called Ames strain. But several laboratories around the country have batches of the Ames strain, and all of those laboratory isolates have the same genetic pattern when only the eight markers are considered. To differentiate among those isolates – and to see if one matches the bacteria used in the attacks – scientists want to find additional regions of variability that differ consistently from one Ames isolate to the next.
That posed a daunting problem, though: Where in all those 5.1 million bases should scientists look for such variation? What a waste of energy it would be to spend months or years comparing genetic regions that simply don't differ from isolate to isolate.
Researchers at the Institute for Genomic Research in Rockville (TIGR) were well positioned to provide a shortcut. They had already spent the past two years placing all 5.1 million bases in order for a single strain of B. anthracis. The sequence – the first ever put together for that species of bacterium – is set to be published within the next few weeks.
In October, after the anthrax attacks, the National Science Foundation quickly contracted with TIGR to complete a quick genetic sequence of a second isolate of B. anthracis – one from the body of Bob Stevens, the Florida photo editor who was the first to die from the bioterror attacks. TIGR recently completed that job, and a comparison of the two has revealed some telltale differences.
Those specific differences may turn out to be useful "name tags" for those two strains. But for now, the primary usefulness of the new findings is that they indicate where in the sprawling landscape of 5.1 million bases scientists should focus their efforts as they search for genetic variations unique to each laboratory's version of the Ames strain. And now Keim and his colleagues at Northern Arizona are doing just that.
By focusing on the regions of variability highlighted by the TIGR team, the Arizona scientists have begun to document specific differences in genetic coding that seem to distinguish among previously indistinguishable isolates of the Ames strain. At a meeting in Las Vegas last week, for example, Keim reported that by counting the number of times a particular base was repeated in a specific part of the B. anthracis genome, he could tell which of five different laboratory isolates of the Ames strain he was looking at.
Focusing on that and other emerging genetic hallmarks, Keim has now begun to compare the bacteria used in the attacks to bacteria retrieved by the FBI from various laboratories around the country. It may be that the Florida sample won't match any laboratory isolate exactly, but even a near match could indicate that the terror sample has family roots there.
"It won't prove who did it," said TIGR's Steven Salzberg, "but it can tell the FBI, 'This is a lab you should look at. The bacteria probably came from this lab.'‚"
Of course, anthrax is only one of several diseases that the federal government has listed as likely agents of biological warfare. And even less is known about the genetics of most of them than was known about B. anthracis, said TIGR President Claire Fraser.
"We need to be collecting this kind of information for lots of other pathogens," Fraser said, "for both lab strains and natural isolates of biological warfare agents, so we'll have it available immediately if we ever need it." |
