Tiedje Lab - Projects


Comparative genomics of the Burkholderia cepacia complex and Shewanella spp (Kostantinos Konstantinidis)

Prokaryotes represent the richest reservoir of biodiversity on the planet, while their biomass probably equals that of terrestrial and marine plants. Yet, several fundamental issues such as what are prokaryotic species and what drives the extensive inter- and intra-species diversity that characterizes prokaryotes are subjects of hot debate and/or unclear. Further, how and what ecological and biochemical factors determine genome structure and gene content evolution in complex ecological settings such as soil remain elusive. Gaining this information is at the heart of understanding the basis for and value of prokaryotic biodiversity, and for understanding the diverse environmental prokaryotes that catalyze much of the biogeochemical cycles that sustain life on earth.

We are employing genomic approaches including comparative sequence analysis and microarray-based comparative genome hybridization to gain novel insights into these questions. Our efforts are split between two areas: I) development of new ways to analyze all publicly available genomic sequences to reveal universal trends and impressions about prokaryotic diversity and II) comparative genomics within two phylogenetically close groups of bacteria, the Burkholderia cepacia complex and the Shewanella spp., for finer-scale resolution and for better understanding of the exceptional capabilities of these groups. Burkholderia is challenging in terms of species description and encompasses among the most metabolically versatile heterotrophic bacteria known while Shewanella is of major environmental importance in areas of remediation, metal reduction, and denitrification. Our efforts on these groups are greatly facilitated by the availability of whole-genome microarrays and several genomic projects (currently, 16 Burkholderia and 15 Shewanella strains are being sequenced), which have been led by a collaborative team for each organism in group.

Examining the current bacterial species definition (Deborah Himes)

The field of taxonomy and systematics is in need of a new definition for bacterial species that is more selective and more predictive of phenotype than the current definition. Currently, two bacteria are classified as the same species if they share one diagnostic trait and exhibit a 70% or greater DNA-DNA reassociation value. Even though the current definition is universally applicable in the prokaryotic world, it leads many bacterial species to be comprised of a heterogenous set of microbes that are isolated from a wide array of environments and/or exhibit a broad array of pathogenicity. For example, Escherichia coli strain O157:H7 Sakai and strain CFT073 exhibit greater that 70% DNA-DNA reassociation values yet differ in as much as 45% of their gene content. Additionally, these two E. coli strains also exhibit different pathogenicity as 0157 Sakai is enterohemorrhagic while CFT073 is uropathogenic in humans. As a means for comparison, if the current definition of a bacterial species were to be applied to eukaryotic species, humans and all other primates would belong to the same species. Moreover, in some cases, two species that show higher than 70% DNA-DNA reassociation values have been classified into two separate species (i.e. E. coli and Shigella sp.). Thus, the current system not only too broad but is also inconsistent.

Ideally, bacteria could be classified by the presence or absence of a species-specific set of genes ("natural species definition") as this could eliminate any arbitrary classification scheme. Alternatively, species-specific expression of a common set of genes could also be used. We will be comparing the genomic sequences of a large number of closely related bacteria using in silico analysis of the genome sequences, comparative genomic hybridizations (CGH), and transcriptome analyses using microarrays. We have chosen to study two groups of bacteria: one group is comprised of members of the Burkholderia cepacia complex (Bcc) and the other group consists of Escherichia coli and Shigella species. We chose to study the Bcc for several reasons: (i) the taxonomic classification of Bcc members has been extensively analyzed using a multi-phasic approach, (ii) representatives of seven strains [including three strains (J2315, AU1054, and HI2424) that have been classified as Burkholderia cenocepacia] have been sequenced and are available for sequence analysis, and (iii) we can examine the effect of niche on gene content as the Bcc have been isolated from a broad range of ecological niches including plant rhizospheres, soil, and the human respiratory tract. E. coli and related bacteria were chosen due the large number of sequenced strains (9) as well as the disparities that have been observed in E. coli classification (see above examples).

We are currently evaluating 60-mer oligonucleotide arrays as platforms for CGH as this has not been reported for bacterial genomic DNA. When we have established the optimal conditions for hybridizations across species, we intend to create a microarray containing probes designed against either all unique open reading frames of a given bacterial group based on genome sequence or all conserved genes in a given bacterial group. We will then attempt to utilize these arrays to determine if species-specific genetic (using the unique open reading frame array) or expression (using the conserved gene array) signatures exist within these groups.