Sniffing out cancer with insect olfactory neural circuits

The metabolism of cancer cells is abnormal compared to that of normal tissues. This leads to the alteration of volatile organic compounds (VOCs) in the exhaled breath of individuals with cancer.

A fascinating new preprint reports the results of examining the VOC composition from individual cell cultures derived from human oral cancers, using an experimental setup employing an insect brain with attached antenna coupled to an electrophysiological monitoring platform, analyzed by computational modeling.

Introduction

Cancer detection methods are urgently required to diagnose this condition early. Non-invasive methods are the focus of much research since these would encourage people to get tested earlier and more regularly and permit more effective intervention.

One such technique is breath analysis, which examines the VOC content of exhaled breath as a reflection of the altered metabolism of the body due to the presence of cancer cells. Earlier studies have identified some potential biomarkers of lung cancer and those of the head, neck, or breast.

The breath is gas. As such, no clinical gas sensing technology exists at present for the detection of cancer. Volatile chemicals are currently detected as individual components of gaseous mixtures by gas chromatography-mass spectrometry (GC-MS). However, it is slow and requires bulky equipment, with sample pre-processing and storage mandatory.

Moreover, and perhaps more importantly, knowing the composition of the gas is not as useful as sensing variations in the gas composition due to host and environmental factors.

Another avenue is the electronic nose (e-nose) devices built biologically. These use, for instance, combinatorial coding to sense VOCs in one step in real-time. However, they suffer from poor specificity and cannot work in actual conditions.

The most sensitive sensors for volatile compounds remain the olfactory receptors. While dogs have been used extensively, among biosensors, to detect drugs, weapons, explosives, and even diseases, they can be trained to detect only one disease at a time.

Again, African giant pouched rats, honeybee proboscis extension reflex (PER), fruit flies, and ants have all shown behavioral changes that correctly identified the presence of specific infections or cancers. These are, however, potentially modified by the animals’ inherent behaviors.

Rather than reverse-engineering biological olfaction, the scientists in this study decided to exploit the marvels of insect biology, combining it with sophisticated recording and computational tools for non-invasive cancer detection.

The current study, published on the preprint server bioRxiv*, is based on the highly sensitive olfactory receptors found in insects. Why insects? The reasons include their exquisite sense of smell, which can detect molecules of gases at extremely low concentrations and identify very tiny changes in the composition of a gas mixture. They are also low maintenance and have a high potential for behavioral training to detect a given volatile compound.

The olfactory sensory system of insects begins with detecting an odorant like a VOC by olfactory receptor neurons (ORNs) in the insect antennae. These nerve cells respond to more than one VOC based on the type of chemical. They use a combinatorial coding process of identification which gives as few as 50 ORNs the ability to detect several trillions of odorant molecules.

Once the ORNs are activated by an odorant, they send electrical signals to the antennal lobe of the insect brain for further processing by an intricate complex of excitatory and inhibitory neurons. Earlier research by these authors confirmed the ability to recognize odors specifically and detect new compounds, both of which are required for a reliable detection system in a natural context.

The researchers, therefore, used three key components: a live locust brain, since the neural firing profiles for different odors have been studied deeply in this species; a connected platform for recording electrophysiological responses – the sensor; precise delivery systems for the VOCs; and advanced data analytics.

How this works is as follows: first, the sensor is calibrated by exposing it to cancer-associated VOCs and recording the neural responses as templates. These are like cancer fingerprints, identifying each type of cancer because of the different and specific VOC composition of each breath. The use of a live insect brain connected to the powerful chemical sensors in the insect antennae allows both to be utilized to their full potential to classify VOC fingerprints for different cancers.

The researchers first got the locust antennal lobe neurons to respond to various odorants via the ORNs, recorded the neural responses, and used them to detect mixed VOCs from human mouth cancer cells. They predicted that the firing pattern would vary with the cancer line and also differ from that of a non-cancer cell line. They also predicted a rapid, robust and sensitive response.

What did the study show?

The investigators found changes evoked by odorants, specifically VOCs, in the excitatory neurons of the locust antennal lobe of the brain. They observed distinctive responses to oral cancer cells, non-cancer cells, and the culture medium, with the three oral cancer cell lines evoking different spiking responses. The spike counts were also different for cancer cells vs. non-cancer cells.

By measuring the agreement between a neural response template and the unknown VOC sample, they identified the odor using neural computational schemes.

Unique neural trajectories corresponding to individual VOC mixtures indicate that oral cancer VOC profiles are distinct from the non-cancer cell line. Moreover, we observed distinctions among the neural trajectories evoked by the three oral cancer cell lines.”

By repeating the process at various time points across the cell culture, they could distinguish each tested stimuli at any point of their growth. In other words, the specific VOC profile emerged in cultured cells early in their growth and remained distinguishable over the following days.

“These results validated our hypothesis that neural response-based classification of cancer VOCs is unaffected by the variations in chemical background caused by evolution of cancer cells in the culture medium.” They rightfully consider this ability to separate all three cancer cell lines and the non-cancer control cell line accurately a unique feat.

They also tested a classification system based on raw neuron voltage response, with root mean square (RMS) filtering since this was relatively less demanding for the computer system, could be carried out without supervision, and has been previously shown to be specific for different odors. This system proved to separate all 7 VOCs tested on it with speed and accuracy.

These sets of results demonstrated that neural response-based cancer classification is fast and only requires 250 ms of neural data from stimulus onset to distinguish oral cancers from controls.”

What are the implications?

The scientists found a chemical sensing system that produces accurate and reliable detection of VOCs independent of the background chemicals, unlike e-noses and GC-MS. They exploited biological neural coding schemes in the insect brain, which depends not on identifying separate components of gas mixtures, but separates gas mixtures as a whole by producing a unique neuronal fingerprint or neural response template when exposed to a VOC mixture by olfactory neuron signaling.

The next step may be to use a whole brain attached to the antenna, as a sensor, for the real-time analysis of breath samples at a rapid pace, enabling the high-throughput screening of numerous samples of VOCs. Such a sensor can be kept viable for a longer time. The final achievement will be, eventually, to develop “a portable, one-shot, point-of-care brain-based VOC sensor.”

*Important notice

bioRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.

Written by

Dr. Liji Thomas

Dr. Liji Thomas is an OB-GYN, who graduated from the Government Medical College, University of Calicut, Kerala, in 2001. Liji practiced as a full-time consultant in obstetrics/gynecology in a private hospital for a few years following her graduation. She has counseled hundreds of patients facing issues from pregnancy-related problems and infertility, and has been in charge of over 2,000 deliveries, striving always to achieve a normal delivery rather than operative.