Detection of volatile indicator compounds for biological contamination and deterioration of food

Lead Research Organisation: Rothamsted Research
Department Name: Directorate


When fruit deteriorates because of damage, insect infestation or fungal and bacterial infections, it produces characteristic odour chemicals that are not produced by healthy fruit. If those chemicals were detectable at an early stage of deterioration, before there were any visible signs, it would save a lot of wastage by growers, distributors and retailers. Although the means of detecting traces of chemicals in the air exist there is no existing method for identifying which chemicals are specifically associated with the early stages of deterioration. They are likely to be present in tiny amounts as part of a complex mixture of chemicals, many of which will be present in relatively large quantities. This project will provide a means of identifying those chemicals, using the sensitivity and discriminatory power of insect olfaction, so that their detection can be used to monitor for the early deterioration of fruit. The methods will also be applicable to many other foods and situations. The Mediterranean fruit fly lays its eggs inside healthy fruit, oranges in this project, where the larvae will subsequently develop with no visible external signs of their presence for some time. Honeybees, that we know can respond to fruit and infection related chemicals, will be trained to recognise the specific chemicals associated with such an infestation. This will be done by giving the bees a sugar reward, to which they respond by extending their proboscis or tongue, at the same time as exposing them to the odours from fruit containing the insect larvae and no reward when exposed to the odours of healthy fruit or damaged but uninfested fruit. After training they will extend their proboscis in response to the odour from the contaminated orange even without a reward. The odours from insect-infested oranges will also be collected by adsorption onto a porous polymer and elution with a solvent to provide a solution of the chemicals involved. This will contain many additional chemicals but these will be separated by gas chromatography into the separate component chemicals which will be blown, sequentially but one at a time, across the antennae of a trained honeybee. When the bee is exposed to the particular chemical(s) that it uses to discriminate between infested and healthy fruit it will extend its proboscis, a response that is very easy to observe and record. These particular chemicals will then be identified by standard analytical methods including mass spectrometry and nuclear magnetic resonance spectroscopy. As a check on the methods, samples of the chemicals will be obtained and used to train more honeybees which will then be exposed to infested and uninfested fruit to confirm that they are the correct chemicals. Portable mass spectrometers will then be programmed to look for these chemicals in the odours from fruit, both in the laboratory and in the processing and distribution centres.

Technical Summary

Honeybees will be trained to discriminate between fruit infested with insect larvae, healthy fruit and damaged but uninfested fruit by simple Pavlovian conditioning. They will be given a sugar solution reward at the same time as short exposures to the volatiles emanating from infested fruit. After a number of repetitions they will be tested to check that they are responding to infested fruit only, with a proboscis extension reflex (PER). The headspace volatiles of healthy, damaged and infested fruit will be collected by sorption onto a porous polymer and elution with organic solvent. The resulting solutions will be analysed by gas chromatography (GC) coupled to a PER assay in which the effluent from the GC column is split, half going to the normal detector and half to a conditioned honey bee. In this way component peaks that the bees are using to detect the contamination will be indicated. The solutions will then be analysed by GC-mass spectrometry (MS) and the peaks of interest identified using additional chemical techniques and nuclear magnetic resonance (NMR) when necessary. The compounds identified as markers of infestation will be obtained and as a check that they are indeed markers of contamination will be used to condition bees, which will then be tested on infested fruit, uninfested fruit and damaged but uninfested fruit in a PER assay. To determine the applicability of the results for routine monitoring devices, the identified compounds will be used to tune a monitor such as miniaturised MS or GC-MS which will be tested in research and then commercial situations. To determine the applicability of the results for routine monitoring devices, the identified compounds will be used to tune a monitor such as miniaturised MS or GC-MS which will be tested in research and then commercial situations.


10 25 50