In just the past decade, the way that we interact with biological systems has advanced dramatically. We can now model diseases with stem cells, turn neurons on and off again, and even edit individual genes. More and more, we are developing technology to precisely control biological systems, for the purposes of both basic research and therapeutic interventions.
Remarkably, much of this research is taking place in very simple organisms; Escherichia coli (E. coli) is one such example (yes, the same bacterial genus that occasionally causes food poisoning – well certain strains of it at least). As a dominant model in biology, E. coli has very well-studied biological pathways and behaviors. In recent years, researchers have taken on the task of inserting synthetic genes into E. coli that have a variety of roles: a genetic switch, a cell cycle counter, or a timer for events happening within the organism. A 2014 study even engineered E. coli that could sense and record events happening in the gut of a mouse, implying that someday we may be able to use synthetic organisms for detection and diagnosis of gut diseases.
Building on these advances in synthetic biology, other researchers are using communities of organisms to study input-output reactions at the macro-scale. As you might imagine, understanding a biological network in a single E. coli is already quite difficult, so unraveling these network interactions between E. coli, millions upon thousands of them, is quite daunting. However, with the combination of synthetic biology and macro-scale studies, we can capitalize on these “biomimetic” systems to understand how even primitive life forms have remarkably complex behaviors.
As an example of such research, in a recent study from Virginia Tech, researchers designed a system that could read the output of E. coli and use it to direct a car to various stimuli. Put simply, they proposed creating a small microbiome out of a community of genetically-modified E. coli and connecting it to a remote control car, which would interpret the activity of the E. coli. A change in the environment would trigger the expression of reporter genes, which a “microchemostat” could convert into voltage, ultimately using this change in voltage to control the robot host. By modeling the genetic switch mentioned above, the researchers were able to switch the food preferences of the microbiome. With switch in one configuration, the microbiome-guided robot preferred lactose; in the other configuration, it preferred arabinose.
The goal of this project was to establish a way to probe host-microbiome interactions and search for unique, emergent properties of their biomimetic system. Although their design is an in silico model, it demonstrated some compelling properties. Interestingly, the researchers noticed that the robot would often move in a pattern where it would slowly approach the food, pause, and then quickly strike, which the authors argue is similar to predatory behavior. Remarkably, even reading out information from a community of E. coli can mimic behaviors typically attributed to more complex organisms.
Ultimately, studying the interactions of the microbiome “brain” and its robot host in this way brings us another step closer to understanding how bacteria, such as the flora that populate our gut, affect our health and mood. A growing body of evidence has linked human conditions such as depression, autism and obesity to our microbiomes, and that perhaps even some behavioral traits can be attributed to them too. Certainly, there are experiments showing that mice exhibit lower stress levels when implanted with probiotics; then there is evidence that the mating behavior of fruit flies can be manipulated using bacteria (creepy and impressive at the same time!)
- John Muir, 1911
Technological innovation often takes tips from nature, trying to build faster, more efficient machines based on principles found in neuroscience or biology. While we’re often modifying nature for our own immediate purposes, such as genetically-modified organisms, we can also use these modifications to study the computations that underlie interesting behaviors. As the authors of the 2015 bacteria robot study conclude, “We expect [our] model system will have implications in fields ranging from synthetic biology and ecology to mobile robotics.” We won’t be traveling around in E. coli-driven cars any time soon, but the computations completed by these small but elaborate organisms may ultimately influence the way we build machines. and, indeed, who we are as individuals.
Ashley Juavinett is a UCSD neurosciences PhD student, an NSF Graduate Research Fellow, and an aspiring science writer. Working at the Salk Institute (La Jolla, CA), Ashley is using in vivo imaging to investigate the neural circuitry underlying visual perception in mice. She currently co-directs a collaborative science writing group, NeuWrite San Diego (http://www.neuwriteSD.org), and writes about neuroscience and society on her own blog (http://scramblingforsignificance.blogspot.com). Follow her on Twitter: @ashleyjthinks