Willerslev and Hebsgaard's comment on the evidence for the 250 Ma age of the halite, brine inclusion, and Virgibacillus strain 2-9-3 reported in Satterfield et al. (2005) comes at a time when earth scientists are busily searching for signs of microscopic life in ancient samples of permafrost, ice, deep-sea sediments, amber, salt, and chert. In the not too distant future, sedimentary rocks may be returned from Mars. It is critically important that the scientific community agree on the methods used for study of ancient microbes and ancient DNA and that issues of sample age, contamination, sterilization, and replication of results be addressed now, not later. In this regard, Willerslev and Hebsgaard here question the age and origin of Virgibacillus strain 2-9-3 cultured from a brine inclusion in halite from the Permian Salado salts. They do not take issue with the evidence for the Permian age of the brine inclusions discussed in Satterfield et al. (2005), but they imply that Virgibacillus strain 2-9-3, reported by Vreeland et al. (2000), is a modern organism, presumably a laboratory contaminant. That conclusion is based on (1) lack of replication of results in an independent laboratory, (2) degradation of relatively unstable DNA molecules over geological time, and (3) “relative rate tests,” which suggest that Virgibacillus strain 2-9-3 is not geologically ancient.
1. Willerslev and Cooper (2005) and Hebsgaard et al. (2005) reviewed issues of contamination and proposed important guidelines for geobiological studies by offering criteria for the authentication of results for the study of ancient DNA and viable microbial cells. Vreeland et al. (2000) described their laboratory procedures and the sterilization techniques used to avoid contamination by modern organisms. They meet the guidelines of Hebsgaard et al. (2005). Willerslev and Hebsgaard are correct that Vreeland et al. (2000) did not obtain replication of their results by an independent laboratory before publication. Such verification is important because it is not likely that separate laboratories would have common microbial contaminants. We welcomed any geobiologists interested in studying the Salado halites and fluid inclusions in the years since 2000. None have so far taken up the offer.
Satterfield et al. (2005) showed that halites of the Permian Salado salts trapped surface brines as fluid inclusions. This halite is ideal for the study of ancient DNA and microbial cells because samples of brine, once trapped inside crystals as fluid inclusions, can remain completely sealed and isolated from the environment for periods of more than 500 m.y. Such preservation of fluid inclusions has allowed samples of ancient evaporated seawater to be analyzed from halites going back to the late Precambrian (Lowenstein et al., 2001). Careful surface sterilization of similar halite crystals should also allow study of uncontaminated samples of Earth's ancient biosphere in fluid inclusions.
2. Willerslev and Hebsgaard state, “DNA is a relatively unstable molecule compared to other cellular components…. and will degrade with time if not repaired.” Twelve years ago, it was thought that DNA, as short fragments, may survive 104 years (Lindahl, 1993). Now it is reported that, under certain conditions such as in permafrost, preservation of DNA may approach 106 years (Hebsgaard et al., 2005). Clearly, the upper limit of DNA survival has changed and should not therefore be used as evidence against the Permian age of Virgibacillus strain 2-9-3. Willerslev and Hebsgaard state, “The rate of degradation (of DNA) is known to be highly dependent on the environment.” In this regard, fluid inclusions in halite from shallow burial environments (<1 km, <30 °C) offer a “friendly” setting for long-term preservation of microbial cells and DNA. They are low in oxygen and their salty waters are known from experiments to slow down the breakup of DNA by depurination (Lindahl and Nyberg, 1972). There has been no systematic research on DNA preservation in such oxygen-poor, saline systems at low temperatures, although that is changing. We are currently studying ancient microorganisms and DNA in fluid inclusions in halite from Death Valley and Saline Valley salt cores, 103 to 105 years old.
3. Relative rate tests, which suggest that Virgibacillus strain 2-9-3 is not geologically ancient, are based on the assumption that evolution follows a predictable mutation rate and that the rate is known. The rates being used by the researchers cited are based on nucleotide substitutions in laboratory-grown bacteria, which may not be realistic for all organisms. Furthermore, growth rates of microorganisms in nature may, in some cases, be measured on time scales of centuries or longer, as illustrated by Parkes et al. (2000) for bacteria isolated from subseafloor sediments. Such long generation times might explain the similarities in 16S ribosomal DNA of Virgibacillus strain 2-9-3 and its contemporary relatives (Maughan et al., 2002).
As we continue to study microorganisms preserved in ancient halite, perhaps others will participate or conduct the additional independent work that Willerslev and Hebsgaard want to see. We agree that extreme care is required when analyzing small numbers of ancient cells or small amounts of ancient DNA. As the field of geobiology expands, the need for discipline-wide laboratory standards is greater than ever.
This research was supported by the NSF Geosciences Life in Extreme Environments and Biogeosciences Programs.