Counterfeit food

C&I Issue 1, 2015

The presence of ‘rogue’ horsemeat in products sold as beef has focused attention on the shortcomings of European food supply chains, and on analytical testing regimes.

‘If this had been a food safety issue, there would have been a whole infrastructure to call on,’ says Paul Brereton, head of agri-food research at the UK-based Food and Environment Research Agency (Fera) and coordinator of the €12m pan-European Food Integrity research project set up in the wake of the horsemeat scandal. ‘Because it was food authenticity, people did not know where to go for information. We’re looking to support and redress that imbalance.’

One factor that allowed horsemeat to make its way into beef products, says Brereton, was a misplaced confidence in the food traceability systems: many suppliers whose products were adulterated had mature supply chain management systems to track and trace their products.

‘The problem is that these systems track the packaging – but it’s the contents that we’re interested in,’ says Brereton. ‘You must be able to verify that the contents match the package.’

Food can be compromised at any stage of the chain, meaning that analytical testing should be done at each stage. However, this potentially adds enormous cost so there is a constant search for faster, cheaper tests that can verify food quality and authenticity.

At the time of the horsemeat scandal in 2013, labs were using targeted DNA analysis. However, a multi-species platform – or an ‘untargeted’ test such as mass spectrometry, which looks to see whether a product has changed over time – would have been much more useful.

Brereton cites the infamous example of melamine in milk in China in 2008. Although many people – mainly children – were affected, it took some time to identify the reason, because of shortcomings in the test procedure, which involved an indirect method of measuring protein. Melamine, a rich source of nitrogen, was fraudulently added to the milk in order to raise its apparent protein content – and help it to ‘pass’ the test. ‘If untargeted analysis had been used, it would have picked this up,’ says Brereton.

In the UK, a government commissioned review published in July 2014 recommended a food integrity system based on eight equally important ‘pillars’ – such as putting consumers first; working towards a zero tolerance of ‘food crime’; and ensuring that laboratories are able to provide standardised, validated testing.

While testing is a critical weapon in the fight against food fraud, however, it is not the only one. Food Integrity is looking to develop an ‘early warning’ system, using technology similar to that used in the international money markets. It will track events, such as poor harvests, to anticipate where food fraud might happen in future. Another method, called mass balance, looks at inputs and outputs, and is useful for determining whether there is too much of a particular product on the market.

‘There’s always been a question mark over the amount of Manuka honey sold,’ says Brereton, because it commands a very high price. By determining an accurate number of Manuka trees, it is possible to estimate the maximum amount of Manuka honey that could be produced. Mass balance is also useful in the fish sector, because the quota system allows accurate estimation of an input value. ‘Mass balance is under-utilised, and we use it anecdotally,’ he says. ‘We’re trying to develop it in order to identify risk.’

Even with safeguards in place, food needs to be checked for authenticity. One recently developed test can rapidly differentiate between beef and horsemeat using nuclear magnetic resonance (NMR). Rather than relying on DNA analysis – which is accurate, though relatively slow – it analyses the triglyceride content of the meat fat.

The principle of analysing fats to differentiate between species is well established. As far back as 1938, there was a test for horsemeat that relied on its higher levels of linoleic acid. Horsemeat is typically higher in polyunsaturated fatty acids, but lower in the saturated and mono-unsaturated variants, than beef. The UK researchers, from the Institute of Food Research in Norwich, and Oxford Instruments, used NMR to analyse 19 beef and 19 horsemeat samples.

Samples were also frozen, then thawed and analysed at a later date. Carrying out a standard chloroform-extraction on fresh meat produced triglyceride spectra with a 10-minute spectral acquisition time. Three particular signals in the spectrum – bis-allylic, olefinic and terminal CH3 peaks – were particularly useful in differentiating between species. Analysis of thawed meat samples showed that freezing had no appreciable effect on test results.

‘We envisage that our approach will be suitable as a screening technique early in the food supply chain – before cuts or chunks of raw beef are processed into mince or other preparations,’ say the researchers in a paper published in Food Chemistry (doi:10.1016/j.foodchem.2014.11.110). ‘A candidate point for detecting adulteration is in large frozen blocks of meat trimmings.’

The blocks could be core-sampled, and tissue fragments analysed using this approach. The technique could be developed to analyse mixtures of beef and horsemeat, say the researchers, though this was not investigated. However, one potential weakness here is that a mixed sample will contain more triglycerides from beef, because it has a higher fat content than horsemeat.

NMR may also one day form the basis of a test for organic food – seen by many as the ‘holy grail’ of food testing. Organic food might seem tailor-made for food fraud: it commands a higher price than its conventionally grown alternative, yet the two are virtually indistinguishable in tests.

Now, researchers in Germany have used NMR ‘profiling’ to analyse organic and non-organic tomatoes. The researchers, from the Bavarian Health and Food Safety Authority and the University of Wurzburg, analysed 361 samples comprising two tomato cultivars – Mecano and Tastery – that were grown in greenhouses by four different producers, using either organic or conventional methods.

‘Our key aim is the development of a sufficient method for an authentication of organic tomatoes against the background of food surveillance and consumer protection,’ said the researchers in a recent paper in the Journal of Agricultural and Food Chemistry (
The test analysed the balance of many chemicals – such as amino acids – found in tomatoes, using multivariate statistical analysis. Although the test managed to differentiate organic from non-organic tomatoes, there was some data overlap. The researchers also found that the same tomato variant, grown the same way by different producers, showed wide variation in results.

So far, the test proves the concept that NMR might be used to determine whether or not a tomato has been grown organically, but it is nowhere near robust enough to be used commercially.
A potential way of using it in future would be to build a database of NMR spectra from authentic tomatoes, against which new tests could be compared, say the researchers.

Yet another method to verify food authenticity, meanwhile, is to use ‘tags’–  the microscopic equivalent of putting a hologram onto a banknote. Swiss researchers, for example, have come up with a tagging system that combines nanoparticles with a short length of DNA as the tracer.

‘We’re combining nanotechnology with genes; that will be difficult to sell to the public,’ says Robert Grass, who led the development at the functional materials laboratory at ETH in Zurich (ACS Nano, doi: 10.1021/nn4063853).

Despite this, the team believes it could be used to tag expensive organic substances such as extra virgin olive oil, petrol and essential oils. Olive oil is commonly counterfeited due to its high value. Sometimes, it is replaced by a cheaper oil; at other times, it is falsely labelled as originating from a different place, such as Italy.  ‘Counterfeit olive oil is a huge topic in the US,’ Grass says.

The ‘taggant’ consists of three main elements: a magnetic nanoparticle, which allows the particles to be extracted from the olive oil and tested; a sequence of DNA, 100 base pairs long, which is unique to each tagged batch; and a protective silica coating, which protects the DNA. One problem that the team had to solve was how to attach the nanoparticle to the DNA. This was done by functionalising the nanoparticle’s surface with ammonium groups, then absorbing the DNA onto it. The silica is grown over the surface using a sol-gel process. ‘This protects the particle, like a glass wall,’ Grass explains.

Before a sample can be verified, the taggant must be removed from the oil using magnetic separation. Then, using a technique borrowed from silicon chip manufacture, the silica coating is removed using a buffered fluoride solution, leaving the DNA exposed for analysis. The big advantage of using DNA is that modern techniques like polymerase chain reaction (PCR) can detect a single molecule. This means that a concentration of just one part per billion would be needed – or 1mg per tonne of olive oil.

The technology is already partially proven, as it is being used to tag plastic granules that are used to make construction materials. The tags used here would be identical to those used in olive oil, Grass says.

Lou Reade is a freelance science writer based in Kent, UK

Become an SCI Member to receive benefits and discounts

Join SCI