Lily Sharratt-Davidson explores how researchers attempting to bioengineer bomb-sniffing locusts have discovered that the insect’s sense of smell is governed by an arithmetic neuron mechanism.
For several years now a team at Washington University in St. Louis has been researching the odour sensing system of locusts, with the aim of bioengineering locusts that are able to detect explosive devices. A recent experiment involving trained locusts reacting to different smells in a variety of complex situations, allowed the researchers to determine the neuronal mechanisms at play. The neurons in the locusts’ brains use what Baranidharan Raman, a professor on the team, described as a “surprisingly simple” arithmetic method to recognise a smell regardless of the circumstance. This can be likened to human behaviours – we are able to recognise certain distinct smells regardless of the complexity of the environment or what other odours may be present. Raman and his colleagues published their findings in the scientific journal Proceedings of the National Academy of Sciences of the United States of America.
The locusts were subject to Pavlovian-style conditioning – they were trained by repeated instances to associate certain odours with food. After as little as six repetitions whereby the locusts were exposed to particular smells, they would open their palps (sensory appendages near the mouth) in expectation of food. This conditioning allowed the scientists to then study which neurons fired when the same smell was detected in different circumstances. The variety of conditions explored included having the locusts being hungry or recently fed, the environment humid or dry, and testing the smell while other odours were present. The results of these experiments suggested that despite the locusts having varied neuron responses to the different environmental conditions, the locust would open its palps in expectation of food regardless, showing a recognition of the scent. The question the researchers were left with, however, was how a variable neuron response, that is, in different environmental conditions, was still associated with the consistent physical behaviour of the locusts
The answer, it turns out, was found through machine learning. The locusts’ sense of smell is controlled by two types of neurons; the “On” neuron, which is activated when a specific odour is present, and the “Off” neuron, which is activated when the odour is no longer detected. This works through simple addition, if enough of the “On” neurons are activated, the locust will detect the smell and open its palps. The criteria for a locust detecting a smell, therefore, is simply that enough “On” neurons have fired, and most “Off” neurons have not.
This research is a big step in Raman and his colleagues’ work to engineering bomb-sniffing locusts. If the presence of a smell can be determined simply by adding and subtracting the presence of certain neurons then it will be easily identifiable whether a locust has detected an odour and how this is linked to a certain type of behaviour (such as the opening of its palps). Researchers can then determine whether a locust has detected a smell without relying on physical indicators, and use this process to determine with greater accuracy whether a physical indicator is reliable or not. These steps could lead the team towards their goal; being able to use locusts to detect explosives. Not only would this have large military impacts, but it could positively impact security also. Whilst there is still potential debate around the ethics of bioengineering, the use of locusts would be more economically viable than the current use of bomb-sniffing dogs. This is not to say that locusts could replace dogs in all applications, but there are an array of scenarios where locusts would be beneficial. While bomb-sniffing locusts may be far down the line in terms of real world applications, the prospect also has exciting ramifications for bioengineering research. By uncovering the locusts arithmetic neuronal mechanism controlling the detection of odours, and more specifically by showing the capability of the locust to detect specific odours despite external conditions, the question has been raised as to whether similar neuronal mechanisms are present in other insects or animals.
This research at Washington University has provided potential answers to understanding why other odours or environmental factors do not prevent us from distinguishing specific smells. Undoubtedly, the neuroscience of sense perception in locusts cannot be precisely compared to humans or other organisms, but the door has been opened for more research into how the neuron systems in other organisms go about dealing with identifying and distinguishing scents.
Written by Lily Sharratt-Davidson and edited by Diana Barreiros Jorge. Lily Sharratt-Davidson is a second year Physics student. Find her on Linkedin.
