Honeybees can be trained to identify SARS-CoV-2 infected samples, say scientists

By Susha Cheriyedath, M.Sc.Oct 21 2021

The unprecedented impact of the COVID-19 pandemic has demonstrated the need for fast and reliable testing methods for novel viral diseases in humans and animals. Infections and other pathologies cause detectable physiological changes in the volatile organic compounds (VOCs) profile of animals and humans. Moreover, this profile differs between healthy and infected individuals.

VOCs offer an odor fingerprint based on sex, age, diet, genetic history, and metabolic conditions, thus making the fingerprint unique for every individual. This fingerprint, when analyzed, offers relevant information about the individual’s health, thus prompting the development of a rapid VOC-based testing method.

Study: Bees can be trained to identify SARS-CoV-2 infected samples. Image Credit: Dirk Daniel Mann / Shutterstock

This news article was a review of a preliminary scientific report that had not undergone peer-review at the time of publication. Since its initial publication, the scientific report has now been peer reviewed and accepted for publication in a Scientific Journal. Links to the preliminary and peer-reviewed reports are available in the Sources section at the bottom of this article. View Sources

Using Pavlovian conditioning protocols to train honeybees to identify SARS-CoV-2 infected minks

In a recent study published on the bioRxiv* pre-print server, researchers from the Netherlands successfully trained honeybees (Apis mellifera) to identify severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infected minks (Neovison vison) using Pavlovian conditioning protocols. They were able to quickly condition the bees to specifically respond to infected mink’s odors, which highlights the possibility of making bees part of a broader SARS-CoV-2 diagnostic system.

Diagnostic evaluation trial shows tests using bees had 86% specificity and 92% sensitivity

The researchers tested two different training protocols to assess the performance of bees in terms of rate of learning, memory retention, and accuracy. They developed a non-invasive, rapid test in which multiple bees are parallelly tested on the same samples to obtain reliable results on the health status of a subject. With the help of data obtained from the training experiments, they simulated a diagnostic evaluation trial to calculate the potential efficacy of their diagnostic test. The diagnostic sensitivity of the test was 92%, and specificity was 86%.

“Further research is needed in order to define the duration of their memory.”

Based on the results, the authors recommend a honeybee-based diagnostics system to obtain a reliable and rapid test that is readily available and is a low-input addition to the existing testing methods. According to the authors, a honeybee-based SARS-CoV-2 diagnostic test might be particularly useful in remote and developing communities that lack the infrastructure and resources needed for mainstream testing methods.

“Our results indicate that the VOC profile differs between healthy and SARS-CoV2 infected minks and  that honeybees can recognize these differences and discriminate between them.”

Diagnostic sensitivity and specificity of single bees vs. multiple bees

According to the results of the study, single bees have limited sensitivity and specificity in the diagnosis, as the retention tests for protocol 2 show that 24 hours after conditioning, 67% of the bees correctly identified the infected sample (sensitivity), and 58% of the bees correctly identified the healthy sample (specificity). Hence, the parallel use of multiple bees on the same sample may effectively improve diagnostic performance. In that case, a diagnosis would be based on the reaction of a certain number of bees (called the diagnostic threshold) to the sample tested.

To analyze such an approach, the team performed simulations using groups of 10 bees to test a sample where at least six bees would have to show proboscis extension reaction (PER) for the sample to be considered positive. With this approach, the test’s potential sensitivity was predicted to be about 92%, and the specificity was predicted to be about 86%.

Picture of the conditioning procedure during protocol 1. A single honeybee harnessed inside our custom-made bee-holder. The bee has just been exposed to a positive sample and been provided with a wooden stick soaked in sugar water, which has led her to express the Proboscis Extension Reflex (PER). The red arrow indicates the bee’s proboscis.

Predicted diagnostic accuracy of bees was comparable to point of care tests

The current standard for SARS-CoV-2 diagnosis in the lab is viral RNA detection from respiratory specimens using the real-time, reverse transcription-polymerase chain reaction (qRT-PCR). The accuracy results predicted in this study are comparable to the diagnostic results of point of care (POC) tests, including RT-LAMP tests and rapid antigen tests.

To summarize, the findings of this study suggest that honeybees could be useful in SARS-CoV-2 diagnosis and potentially the diagnosis of other infectious diseases.

“A formal diagnostic validation study is necessary to properly validate the diagnostic approach applied under field conditions and confirm the diagnostic potential predicted in this study.”

This news article was a review of a preliminary scientific report that had not undergone peer-review at the time of publication. Since its initial publication, the scientific report has now been peer reviewed and accepted for publication in a Scientific Journal. Links to the preliminary and peer-reviewed reports are available in the Sources section at the bottom of this article. View Sources