This story originally ran on Sept. 9.
By Tony Fong
Using an approach combining proteomics and genomics, researchers have determined that different strains of the Shewanella bacteria vary much more greatly than had previously been known, suggesting that a new –omics-based approach to microbiology may be necessary for identifying new species.
In the work described in a paper published Sept. 1 in the online edition of the Proceedings of the National Academy of Sciences, the researchers analyzed and compared 10 isolates of the genus Shewanella on the whole-proteome and whole-genome level and found significant diversity on both levels, but especially on the proteome level.
And while they focused their research on Shewanella, the approach they used could – and should, the researchers said in their study – be used for investigations of other microbes as well in order to present a clearer and more complete picture of microbial activity on a functional level.
"Although a high degree of variation in protein expression profiles was anticipated among distantly related species," in Shewanella, "the variation observed among strains of the same species was comparatively much larger than expected," the authors wrote, adding that "these findings have important implications for the correspondence between genotype and phenotype, and hence, for the bacterial species concept."
Shewanella plays an important role in bioremediation and the environment because of its ability to filter certain metals and compounds and convert them into less toxic material. This makes them useful for environmental cleanup work. One strain, Shewanella oneidensis MR-1 has been shown to be especially good at filtering metal oxides from groundwater and turning them into insoluble forms that can then be easily removed. The US Department of Energy is investigating whether this strain can be useful in cleaning up radioactive nuclear weapons sites.
Studies are also being conducted into the applicability of Shewanella as a potential power converter for microbial fuel cells in which it would "eat up" metals and expel electrons as a waste product.
Despite the different potential uses of the microbe, though, it is not easy to tell one strain from another, and using the standard method for identifying newly isolated strains of bacteria – by looking at the ribosomal genes – the varying strains look almost identical, Kostas Konstantinidis, an assistant professor in the School of Civil and Environmental Engineering at the Georgia Institute of Technology, and the first author of the PNAS article, told ProteoMonitor last week.
"If we had used traditional microbiological techniques to identify microbes instead of what [we did], we would have gotten a much more incomplete picture, compared to what we got with this advanced proteomics and whole-cell approach," he said. A traditional approach only would have suggested that the 10 isolates belong to the same species, he noted.
"In microorganisms, in general, [the phenotypic differences are] a little bit unknown. They are so tiny and sometimes it's not so easy to see the phenotypic differences with the classical methods we use, so we tried to take a whole-cell systems biology approach to investigate all kinds of genes or expressed proteins."
Using this approach, they found that while some strains of Shewanella had 98 percent of the same genes, others had only 70 percent of the same genes. And expressed proteins showed even wider differences in genetic content, they said.
That discovery, Konstantinidis said, is especially important because most of the literature on bacteria has focused on their genomic profile. As a result, "we know a lot" about many bacteria on the genomic level, but information about them on the proteomic level is limited, he said.
In Shewanella, the amount of protein information is "a couple of orders of magnitude" less that the amount of genomic information because the bacteria is more expensive and difficult to sample, he said. More than 20 Shewanella genomes have been sequenced.
"The genome is the blueprint, but we want to see what happens at the functional, the phenotypic, the expression level," he added. "That's why we studied the proteins. The genes don't tell us what pathways or genes are getting activated."
[ pagebreak ]
The 10 Shewanella genomes investigated in the PNAS paper were chosen to "represent a range of evolutionary distances, providing for a more unconstrained view of microbial diversity and evolution," he and his co-authors wrote.
The degree of differentiation in protein expression that they found among them "tells us there is much more phenotypic diversity there than we thought based on the genome," Konstantinidis said.
"The big surprise is that even for the genes that they all have in common, when they are grown under identical conditions we found frequently [that] they regulate these genes differently. Some of them are getting expressed or not getting expressed," he said. "And that's very important because now we have to beware of that. We don't have to find only the gene, but [also the protein because] it plays a role whether the gene is active or not, and this was possible only with proteomics."
For the proteomics part of the work, Konstantinidis and his colleagues at Georgia Tech grew the bacteria under identical conditions, then extracted the protein lysates from the cells and then shipped them to researchers at Pacific Northwest National Laboratory where they analyzed the samples with Fourier transform mass spectrometry.
New Path to Identification?
The –omics-based approach described in the PNAS paper represents a new method for identifying new biological species, according to Konstantinidis and his colleagues. The published literature and the researchers' own experimentally derived physiological and growth data could not distinguish between most of the Shewanella strains "or even define general properties for the major clades represented by these strains," they wrote. There was also very low correlation between anaerobic growth characteristics and the "evolutionary relatedness of the strains compared."
The genomic and proteomic data they generated, however, correlated "well" with the phylogeny of the strains and identified congruently strain-specific adaptations "that might be linked to speciation for several of the strains studied.
"These results further corroborate the notion that it is time to start replacing the traditional approaches for defining diagnostic phenotypes for new species or clades with –omics-based procedures," Konstantinidis and his fellow authors wrote.
The approach, he added, comes as the result of improvements in proteomics technology as well as an awakening among microbiologists to the applicability of the technology to their research. While his field has always been interested in the phenotypes of bacteria and other microbes, some "prerequisites" for investigating the functions of genes were not available. But as the number of microbes whose genomes have been sequenced continually increase, scientists can now explore beyond their genomes, Konstantinidis said.
"So I think you're going to see more proteomic studies in the future," he said. "My feeling is that … we didn't have the genome or the resources [to do it before]. And we realize the benefit of [an –omics-based approach] more recently."
While the PNAS study focuses on the team's approach for use in bioremediation and environmental purposes, Konstantinidis said that it also may have clinical ramifications.
One avenue by which microorganisms become so diverse is through horizontal gene transfer in which one organism incorporates genetic material from another organism. That, he said, is how some of the "nasty epidemic pathogens" emerge.
Using the approach developed by he and his colleagues, researchers may be able to gain a better understanding of the phenotypes of such virulent microbes and eventually find more effective ways of combating them, he said.
In ongoing work, he and his collaborators are using the same approach to study other microbes such as Dehalococcoides, and to study the microbial environment in Lake Lanier, which provides drinking water to Atlanta.
They also are working to develop more robust protocols for their method, especially to remove noise from the proteomics data. For about half of the peptides they got from the mass spec, they could not assign them to a gene, which could be due to a number of factors including post-translational modifications and errors.
"I guess this paper is about the system level, looking at the whole cell," Konstantinidis said. "And now based on that, we want to find individual pathways and how they are important, how they are regulated, how they interact with the rest of the genes, in order to eventually be able to then say, 'These genes are important for mercury reduction. This is how they are regulated and this is how we can use them in a biotechnological application.'"