The rogue's gallery of human pathogens is filled with members of the Bacteria and Eukaryota domains of life. Notably absent is the third domain: Archaea. According to a recent report in the Journal of Clinical Microbiology, however, that may no longer be the case.
When Hans-Peter Horz, an assistant professor at the University Hospital in Aachen, Germany, and colleagues used real-time quantitative PCR to survey the microbial ecology of 20 infected dental root canals, they found that five contained archaeal sequences, representing up to 2.5% of the total microbial load in these samples (J Clin Microbiol, 44:1274-82, April 2006). The predominant archaeal species was Methanobrevibacter oralis.
Healthy root canals are sterile environments, Horz says. "Microbes that are able to penetrate into this area must have some pathogenic features: They must be able to invade into the system and they must be able to evade host immune mechanisms," he says.
But does that make M. oralis a pathogen? No archaeon has ever been definitively linked with human disease. Indeed, a study that looked for Archaea in infected root canals a few months previously failed to detect its intended quarry (J Endodont, 31:719-22, 2005) - a discrepancy Horz attributes to the use of different PCR primer sets. Jóse Siqueira Jr., chair of endodontics at Estácio de Sá University in Rio de Janeiro, who led that earlier study, agrees with Horz's explanation, calling the new study "well done, well conducted."
Still, Siqueira stops short of calling these organisms pathogenic. "It's difficult based on their findings to know whether the Archaea were participating in the disease process, or whether they were only there in the root canals because the disease created environmental conditions [that] were conducive to the establishment of these microorganisms in the canal," he says. In other words, "Which came first, the disease or the Archaea?"
David Relman of Stanford University, who demonstrated in 2004 that M. oralis (or at least, its 16S rDNA) is associated with some cases of moderate to severe periodontitis (Proc Natl Acad Sci, 101:6176-81, 2004), says this organism and related methanogens could merely be supporting players in a larger ecological cooperative: "community as pathogen" as he puts it. "And there may be 101 flavors of the community profile that can act essentially as a pathogen."
In this view of endodontic disease, M. oralis and related microbes serve to reduce molecular hydrogen levels at the site of infection. Hydrogen is a byproduct of the fermentation process that occurs in anaerobic infections, yet it is growth-limiting to the organisms that produce it. Methanogens, though, can use molecular hydrogen to create methane and energy. The two organisms - methanogen and fermenter - thus have what's known as a syntrophic relationship, "a two-way cross-feeding relationship that enhances both members," Relman says.
If methanogens really do fill such a syntrophic niche, then other hydrogen metabolizers might also be expected at sites of disease, and indeed, Relman did identify a greater proportion of treponemes, hydrogen-utilizing spirochetes, in methanogen-negative disease sites in his study. Horz says he has begun investigating competition between methanogens and other hydrogen metabolizers in periodontal disease as well. It's unclear what antibiotics would kill M. oralis, but a related methanogen, M. smithii, is known to be susceptible to metronidazole (Chemotherapy, 47:177-83, 2001).
None of that means that members of such cooperatives would be called pathogens under Koch's postulates. For one thing, the organism alone must be capable of causing disease. Moreover, as most microbes are uncultivable, they could never be subjected to Koch's postulates in the first place. New genomics approaches could help, however. Relman is coprincipal investigator on a study that employs metagenomics - an approach in which total microbial DNA is cloned and sequenced from all organisms en masse (see related story, p. 59) - to understand the nature and dynamics of oral ecology. The study probably won't be completed before 2007, he says.