Chemostats allow for the steady-state cultivation of microbes. (Image: Jeffrey Marlow)
Studies of microbiological processes in the natural world have changed dramatically in the ‘omics era, as sequencing tools enable enormous data sets of DNA, RNA, and proteins. It’s a remarkable tool that exposes the intricate working of biological function, but, warns University of Calgary Professor Marc Strous, it’s not a silver bullet.
Strous has made a career of chasing down “unicorns”, as he calls them: elusive scientific problems that conventional wisdom rejects. He’s discovered microbes performing previously unseen – and occasionally unexpected – metabolisms, guided by the principles of thermodynamics and a keen investigatory eye. During a presentation last week at the ISME conference in Seoul, Strous urged the scientific community to run away from the crowd, to embrace, and then explore, uncertainty.
“Paradigms affect how we do experiments in implicit ways,” he said. “We can’t find what we don’t think to look for.” A couple of decades ago, scientists thought they had a pretty good handle on how nitrogen flows through the Earth’s geochemical reservoirs. Nitrogen gas in the atmosphere is a stable molecule, difficult to crack and incorporate into other biological processes. Nitrogen fixing microbes (found prominently in association with plant roots in soils) are able to perform the task, mobilizing the critical element that is required for protein synthesis by all forms of life.
What wasn’t known when Strous entered the fray was how much of that ammonium re-entered the atmosphere as N2. Most scientists believed that oxygen – the most energetic electron acceptor – was the necessary other reactant, that other common options like nitrate or sulfate wouldn’t provide enough energy to rip electrons off of ammonium. But the math didn’t add up: when estimated quantities of N2 formation and removal were tallied, it seemed that substantial quantities of the pervasive atmospheric gas were missing. Strous later identified naturally occurring organisms that were able to make N2 anaerobically, using nitrite as an electron acceptor; it wasn’t as energetically profitable as the oxygen-utilizing process, but it did balance the books. The process is now believed to account for 30-50% of the N2 formed in marine settings, and has been incorporated into engineered wastewater treatment plant ecosystems, minimizing their carbon dioxide emissions.
Strous uses this anecdote as a cautionary tale, a reminder of our continuing ignorance of the natural world, and an exhortation to not rely too heavily on technological advances. “We now know that we don’t know very much,” he says, “and many descriptive studies of imbalances in the natural world could lead to hypotheses that are important. If we had just used metagenomic tools, we probably wouldn’t have discovered this organism.”
Strous also took aim at the common practice of removing outlier data points from experimental datasets. If one data point in an experiment skews wildly from canonical, expected results, many scientists are quick to remove it from further analysis, citing any number of factors that may have gone wrong. It’s classic confirmation bias, and “it’s very difficult to discover new things that way,” he explained.
Sequencing microorganisms is psychically satisfying, as it distills complicated biochemical processes to a tidy string of letters. Interpreting the code, of course, is far from straightforward, and applying such lessons to a real-world environmental context is another challenge altogether. To Strous, a critical arrow in the quiver of modern microbiologists is the chemostat, a continually flushing bioreactor that maintains stable chemical conditions. This tool of cultivation allows the user to tune to a precise biochemical environment, examining a microbial milieu’s “natural” response. In more conventional experiments, ingredients are added to a test tube or vial only at the initial time point, and subsequent biological activity generates a constantly changing microenvironment. It’s difficult to attribute experimental findings to any particular set of conditions.
“Single cell genomics has its place,” Strous allows, “but it need to be supported by other methods, probably cultivation methods. Our rates of discoveries of new processes probably exceed the rates of the past, so we’re doing just fine. But we can certainly do better.”
No comments:
Post a Comment