American Scientist

What's the oldest living thing in the world? Well, a Madagascar radiated tortoise given to the Tongan royal family by Captain James Cook in the 1770s was at least 188 years old when it died in 1965. And a creosote bush in the California desert clocks in at around 12,000 years. The King's lomatia plant in Tasmania is estimated to be at least 44,000 years old and it's still growing (although it doesn't seem to be able to reproduce sexually). Yet these venerable organisms are infants, day-old newborns compared with a living bacterium that may have been locked inside a salt crystal for the past 250 million years.

In 2000, a trio of scientists led by Russell H. Vreeland at West Chester University of Pennsylvania described in the journal Nature the isolation of a dormant, yet living, bacterium. The microbe came from pockets, or inclusions, of fluid trapped inside 250-million-year-old salt crystals buried half a kilometer deep in New Mexico. They  gave it the unprepossessing name of Bacillus (later, Virgibacillus) species 2-9-3.

The report elicited strong skepticism from many quarters. Biological chemists doubted that nucleic acids could remain pristine over such time periods. Even had the bacterium hibernated as a hardy spore, its DNA surely would have broken down over 250,000 millennia, if not from the barrage of ultraviolet light during its long-ago residence on the surface, then from naturally occurring terrestrial radiation over the Earth's evolution.

Geologists questioned the age of the fluid inclusions, arguing that certain features of the Salado Formation (the source of the halite crystal) suggested that flaws in the rock had permitted the intrusion of more recent fluid (which, by inference, had carried more recent bacteria into the ancient rock).

Geneticists pointed out that one of the bellwether genes that the group had sequenced—one that encodes the so-called 16S ribosomal subunit—was far too similar to its counterpart in another strain of bacteria. According to this critique, either the "ancient" bacterium was actually a contaminant, or its descendants had inexplicably failed to change in the past 250 million years.

Yet Vreeland and an expanding circle of collaborators have followed up the original report with publications that seek to counter each of these criticisms. In 2002, he and two West Chester University colleagues reported in the International Journal of Radiation Biology their calculation that the degree of genetic damage caused by normal traces of radioactive potassium-40 in the surrounding rock was not great enough to rule out a quarter-billion years of bacterial survival. Scratch objection number 1.

In April of 2005, the three authors from the Nature paper teamed with Tim K. Lowenstein, a geologist at Binghamton University in New York, and his student, Cindy L. Satterfield, in publishing a detailed report in the journal Geology. To test the idea put forth by critics that inclusions in the salt crystals were newer than the surrounding rock, they measured the temperature of original crystallization for samples from the same part of the Salado Formation that yielded Virgibacillus sp. 2-9-3. The team reasoned that if microbe-carrying fluid had recently reached the deeply buried salt deposit and recrystallized, the temperatures of those crystallizations would be similar. Instead, they found the opposite: The results ranged from 17 to 37 degrees Celsius, or about 63 to 99 degrees Fahrenheit, a distribution that suggests seasonal climatic variation. In other words, the crystals that formed around pockets of fluid (and presumably bacteria) were created on or near the surface instead of far underground.

A second, more definitive, line of investigation examined the concentrations of various ions in the fluid inclusions. The balance of ions in seawater changes over geological time, so measuring them can provide an approximate date at which the saltwater crystallized. The ion concentrations in the halite inclusions matched those of oceans in the Permian period—a profile that is distinct from the seawater of today and also from larger pockets of trapped brine elsewhere in the Salado. As a final test, the team plans to use an ultrasensitive mass spectrometer to date tiny, individual inclusions by the rubidium-strontium method (⁸⁷Rb decays into ⁸⁷Sr with a half-life of 49 billion years). Scratch objection number 2.

The third criticism, based on DNA similarities, has been harder to dismiss. Despite a protocol of sterilization and controls that even critics describe as "heroic," contamination remained a potential source of the 2-9-3 bacterium based on its molecular resemblance to current strains. Understandably, Vreeland defends the work against charges of contamination. He even views the genetic objections as the least valid, stating that of all the challenges (geologic, chemical and genetic), "this is by far the weakest of the critiques."

Although Vreeland admits that the DNA evidence classifies Virgibacillus species 2-9-3 as one strain of another bacterium isolated from the Dead Sea, he challenges the models that predict that these strains diverged tens of thousands, not millions, of years ago. In summarizing a 2002 letter to the journal Molecular Biology and Evolution, Vreeland explained that such models calculate a phylogenetic history based on specific generation and mutation rates, but "you can pick whatever rate you want." Instead of an intergenerational interval of days or months, he cites evidence that some spore-forming organisms can lie dormant for hundreds or thousands of years between growth episodes. Thus, he suggests that as the 2-9-3 strain lay locked in the salt, its relatives might have grown—and evolved—far less than critics believe.

Skepticism lingers—even among some of Vreeland's collaborators—but that hasn't prevented them from conducting further studies in this highly visible area of science. And of course, they may yet be persuaded by new data. In a paper published online August 30, 2005 in the journal Extremophiles, Vreeland presented evidence that four strains of Permian microbes (2-9-3 and three others that were found later) are different enough from modern relatives in a number of categories that they could not arise from contamination.

Many questions remain, as Vreeland readily admits. Wouldn't organic molecules crucial to life, including DNA, spontaneously degrade in 250 million years even in the absence of ionizing radiation? Vreeland responds, "That is something we have to look at." Do the older, nonviable inclusion pockets contain the remains of expired microbes? Says Vreeland, "I can't answer that." And, perhaps most interestingly of all, what is the mechanism? How do these ancient organisms manage to survive so long? His answer: "I have no idea how that happens."