Objections to the transfer of Francisella novicida to the subspecies rank of Francisella tularensis

We disagree with a recent proposal by Huber et al. to transfer Francisella novicida to the subspecies rank of Francisella tularensis (Huber et al., 2010). We believe that the proposal is not appropriate in light of all currently available knowledge.

In 1989, Hollis et al. (1989) argued that F. novicida and F. tularensis could be considered to be one species as judged from DNA–DNA hybridization experiments (Hollis et al., 1989). Their publication was not valid according to the requirements outlined in the Bacteriological Code (Lapage et al., 1992; Tindall et al., 2006). As a result, the proposed elimination of the species F. novicida and its demotion to a biogroup of F. tularensis was not included among prokaryotic names with standing in nomenclature. Notably, earlier publications considered F. novicida and F. tularensis to be separate species based on differences in phenotype including chemotaxonomic markers, distinct ecological roles, different clinical and epidemiological characteristics, and differing abilities and modes of invasion and mechanisms of tissue damage in mammals (Larson et al., 1955; Olsufiev et al., 1959; Skerman et al., 1980).

From a practical standpoint, separate species names are useful in a microbiological laboratory or a clinical setting and also as a basis for regulations governing the handling of medically important organisms. For example, laboratory handling of F. tularensis, but not F. novicida, is associated with a high risk of airborne laboratory-acquired infection. Importantly, it is fairly easy to distinguish F. novicida and F. tularensis on the basis of their different growth and metabolic requirements on artificial media. Indeed, in Table 2 of Huber et al. (2010) data are provided that contradict their own proposal by presenting 11 metabolic reactions that are distinct between F. novicida and F. tularensis (Huber et al., 2010).

Perhaps most importantly, recent findings from the analysis of multiple genome sequences of F. tularensis versus F. novicida have indicated that the increased host-association of F. tularensis is tied to evolution as a population lineage disconnected from F. novicida, even though genome-wide average nucleotide identities exceeded 97 % (Larsson et al., 2009). We propose that different population structures and otherwise disparate evolutionary patterns in F. tularensis and F. novicida should be considered as arguments for retaining separate species names. A comparison of 17 genomes of members of the genus Francisella has shown that the emergence of F. tularensis, in an evolutionary and population genetic framework, was a speciation event with no signs of reversals. For example, there were no traces of genetic exchange between F. tularensis and F. novicida. The analysis provided genetic information that was more precise than crude DNA–DNA hybridization values for defining the genetic relationships between F. tularensis and F. novicida. Recent intense efforts, including evolutionary and population criteria, have provided a useful theoretical framework for defining prokaryotic species (Achtman & Wagner, 2008; Gevers et al., 2005; Koeppel et al., 2008). We believe that such a framework should be taken into consideration in the taxonomy of the genus Francisella.