Benchtop NMR spectroscopy for meat authentication

In This Section

June 2016

  • In 2013, undeclared horsemeat was detected in a wide range of processed meat products on supermarket shelves across the United Kingdom and Europe. The crisis exposed shortcomings in testing regimes and highlighted the need for additional analytical approaches suitable for rapid low-cost screening.
  • Many methods for verifying the species present in meat products are DNA-based, but there are other compositional factors amenable to measurement which can also provide means of species confirmation.
  • This article describes the development of a rapid screening approach for authenticating beef based on triglyceride composition. The screening protocol only takes 10 minutes and can be used at key points in the supply chain, such as meat processors or wholesalers, where the incoming raw materials are in the form of frozen blocks of trimmings.

Nuclear magnetic resonance (NMR) spectroscopy is a well-known technique used in laboratories worldwide. Modern research-grade instruments are based on super-cooled electromagnets that are used to generate the high magnetic fields needed. They are expensive to buy and maintain, occupy a large amount of space, and require highly trained personnel to run them. In recent years, a new crop of low-field NMR spectrometers has appeared on the market. In contrast to their high-field cousins, these instruments are small (often referred to as “benchtop”) and have much lower capital and insignificant running costs. Typically operating at field strengths <100MHz, benchtop spectrometers are based on permanent magnets and work without needing any cryogens.

Food sector applications: starting with triglycerides

Since 2012, the Analytical Sciences Unit at the UK’s Institute of Food Research (IFR) has been working in partnership with Oxford Instruments (OI), a leading manufacturer of scientific instrumentation, to develop benchtop NMR spectroscopy for food sector applications. The project received support from the translational science funding agency, InnovateUK, as well as the Biotechnology and Biological Sciences Research Council. The spectrometer at the heart of the project was the PulsarTM, a 60MHz instrument launched by OI in 2013 (Fig. 1).

0616 Fig 1

FIG. 1. The PulsarTM 60MHz benchtop instrument. To collect a spectrum, an NMR tube containing a liquid sample is introduced into the spectrometer. Data collection is initiated by the operator using acquisition software installed on a PC. Here, the machine is being used to carry out quantitative analysis of edible oils, with the results displayed directly following recording of the NMR spectrum.

 

Among the first compound classes targeted were triglycerides, which are the main constituents of the fat component in foods. Triglycerides are ideal samples for study by low-field NMR, as good quality spectra can be obtained quickly and easily. To examine fats and oils, samples can simply be mixed with chloroform to reduce their viscosity, then placed into a standard NMR tube for analysis. 60MHz NMR spectral profiles contain distinct peaks arising from various moieties (Fig. 2), from which accurate and precise quantitative information on the mono- and polyunsaturated contents can be calculated. A first application of this work sought to distinguish between olive and hazelnut oils, which are highly similar with regards to their fatty acid composition [1].

0616 Fig 2 

Fig. 2. A typical low-field NMR spectrum of vegetable oil, which is composed largely of triglycerides. Indicated on the figure are some of the key resonances that provide information on the detailed composition of the sample, in particular with regards to the amount and type of unsaturated fatty acid chains.

 

To analyze solid foods, an extraction step is needed, but this can also be relatively simple: shaking a few grams of homogenized sample in chloroform, vortexing and filtering into the NMR tube is all that is required. Chloroform is an efficient extractor of lipophilic compounds, and using this procedure, high quality spectra can be obtained of the fat component from a wide range of food products and raw materials [2].

Developing a screening method for authenticating raw beef

In 2013, a major incident of food fraud was uncovered, in which undeclared horsemeat was detected in a wide range of processed meat products on supermarket shelves across the UK and Europe. Thousands of tons of food were recalled, and there was substantial brand damage to the companies involved. The crisis exposed shortcomings in testing regimes, and highlighted the need for additional analytical approaches suitable for rapid low-cost screening.

Many methods for verifying the species present in meat products are DNA-based. However, animals do not differ only in their DNA; there are other compositional factors amenable to measurement which can also provide means of species confirmation. It is common knowledge that pork and beef fat, for instance, are very different from one another. This is due to dissimilarities in the animals’ triglyceride compositions, which in turn arise from differences in their diets, metabolism and digestive systems.  With this in mind, the teams at IFR and OI began an extensive study of the fat component extracted from raw meats, with the aim of developing a rapid screening approach for authenticating beef. 60MHz NMR spectra were collected from extracts prepared from fresh red meats, specifically beef and two potential adulterant species: pork and horse. 

Over the course of a year, the method was refined and repeated on hundreds of meat samples in the laboratories at OI and IFR.  The results obtained were compelling: Each of the different meats examined exhibited clearly different spectra. For example, in the case of beef versus horse, spectral profiles were found to be entirely distinct.  Even allowing for natural variation, no overlap between the two types was found; the test was completely accurate in determining whether an extract originated from a piece of horsemeat or a piece of beef. Figure 3 shows a collection of spectra from the three meat types, along with a graphical representation of the statistical model built to characterise the beef group.  The ellipse delineates a confidence interval around the “authentic beef” group. When challenged with a range of pork and horse test samples, this model correctly placed all of these outside the “authentic beef” group.

0616 Fig 3

FIG. 3. Chemometric analysis of key regions of the low-field NMR spectrum (left panel) leads to a simple two-dimensional model in principal component space (right panel), capable of distinguishing beef from pork or horse with complete accuracy across >100 test samples.

Accurate results in 10 minutes

As part of the project, a stand-alone software package was developed to carry out the mathematical analysis of the spectra, thereby providing a complete system for authenticating raw beef in a screening protocol that takes 10 minutes from start to finish. The test is intended for use at key points in the supply chain, such as meat processors or wholesalers, where the incoming raw materials are in the form of frozen blocks of trimmings; it is also suitable for pre-screening ahead of more time-consuming DNA testing. This work was published in an open access paper in Food Chemistry [3], and a patent on this approach to meat species confirmation is pending.

High-field NMR spectroscopy has long been recognized as a powerful analytical tool, but the equipment is too expensive and technically demanding to allow deployment anywhere apart from specialist laboratories. The advent of benchtop NMR looks set to change this landscape.  Through a number of key food sector applications, we have discovered how useful the low-field modality can be, particularly for the analysis of fat-containing samples. The IFR and OI teams are looking forward to further fruitful collaboration on industrially important challenges as the technology of benchtop NMR continues to evolve.

 

E. K. Kemsley, M. Defernez and A. D. Watson are researchers at the Institute of Food Research, Norwich Research Park, Colney Lane, Norwich, NR4 7UA, UK, http://www.ifr.ac.uk/
D. Williamson is a researcher at Oxford Instruments, Tubney Woods, Abingdon, Oxford, OX13 5QX, http://www.oxford-instruments.com/.

Further reading

  1. 60MHz 1H NMR Spectroscopy for the analysis of edible oils. Parker, T., Limer, E., Watson, A., Defernez, M., Williamson, D., and Kemsley, E.K. TRAC— Trends in Analytical Chemistry (2014) 57: 147–158.
  2. 60 MHz 1H NMR Spectroscopy of triglyceride mixtures. Gerdova, A., Defernez, M., Jakes, W., Limer, E., McCallum, C., Nott, K., Parker, T., Rigby, N., Sagidullin, A., Watson, A.D., Williamson, D., and E.K. Kemsley (2015) in Magnetic Resonance in Food Science: Defining Food by Magnetic Resonance (eds: F. Capozzi, L. Laghi, P.S. Belton) Royal Society of Chemistry pp.17–30.
  3. Authentication of beef versus horse meat using 60 MHz 1H NMR spectroscopy. Jakes, W., Gerdova, A., Defernez, M., Watson, A.D., McCallum, C., Limer, E., Colquhoun, I.J., Williamson, D.C, and Kemsley, E.K. Food Chemistry (2015) 175 :1–9.