Chemists develop DNA 'barcodes' to scan for infection

Imagine having a printed barcode and getting it scanned by a hospital doctor who could then tell you very quickly whether you might have a serious infection or illness. Thanks to an approach for visualising DNA sequences pioneered by chemist Dr Robert Neely and his research group, it’s a technique that’s already possible.

‘We can study genomes very easily and very quickly,’ explains Rob Neely, a Senior Lecturer in Physical Chemistry. ‘We can take complex mixtures and understand what’s in them in a very simple test. We do that by making the DNA sequence visible on the microscope, using enzymes to fluorescently label DNA molecules. That gives us something akin to a barcode of the DNA molecule.’

A lot more work needs to be done to make it commercially viable, but such a method could see diagnoses made in three to four hours instead of several days.

‘The test could be used, for example, as a way to screen for bacteria in a wound,’ says Rob, whose interdisciplinary research group focuses on understanding the DNA molecule by working at the interface of physical and biological sciences. ‘Using our test, you can ascertain if there’s something that might be problematic for a patient. For instance, you could identify E.coli and, although you wouldn’t know if it was virulent, this would allow doctors to make a more informed decision and run some more tests if necessary.’ This compares to the way infection is diagnosed at the moment: bacteria taken from a sample or swab are grown in litres of culture, which is then tested. It’s a time-consuming process.

Rob’s work on developing fluorescence spectroscopy methods to study biological systems dates back to his PhD studies in Edinburgh in 2005, but it was when he moved to Leuven, Belgium on a Marie Curie Intra-European Fellowship in 2009 that he made a significant breakthrough – discovering a way to read DNA sequencing molecules by scanning the sequence for patterns.

‘The analogy I use is a book: You don’t need to read every letter of every word to identify a book. Skipping through chapter headings or looking at the index might be sufficient. In effect, that’s what we do: we don’t read every letter of the sequence; we highlight words. If you find “and” on every page of any book, it will produce a unique pattern. You can use this pattern to identify the book. It’s that pattern we can see in the DNA molecule. We called these patterns DNA barcodes because when we image them on a microscope, they look just like black and white barcodes.’

When Rob’s first research paper on the work was published in 2010, he was only able to image 20 molecules; today he’s imaging hundreds of thousands, which is the sort of scale necessary to study mixtures of bacteria.

His latest paper, written with six colleagues including Birmingham PhD student Nat Wand and post-doc Darren Smith, entitled ‘DNA barcodes for rapid, whole genome, single-molecule analyses’, is a significant interdisciplinary contribution to the journal Nucleic Acids Research. Computer scientist Dr Iain Styles and biochemist Professor Steve Busby are co-authors.

The paper, which recently won the College’s Paper of the Month award, focuses on sourcing images and finding similarities in imaging data. Using another analogy, Rob explains: ‘If you have a lot of images of CCTV data from across a city and you want to identify a person from those images, you would have hundreds of thousands of grainy images of faces to look through. You could go through them one at a time, but that is slow and unreliable. We look for the similarities in the data set – matching every face to every other face. When you’ve done that, you make an average from all those matches so that you get a high-res image of the face or, in our case, the DNA molecule.’

The same thing can be done with sequencing, but it’s expensive, says Rob, who currently holds an EPSRC Healthcare Technologies Challenge Award. ‘So the paper is proof of concept. There’s a lot of work to do to make it commercial, but it’s better defined now than it was five years ago. I think we understand more now about what it can do and where it fits.’