Engineers develop faster, cheaper DNA sequencing

Boston University biomedical engineers have found a way of carrying out gene sequencing using a great deal less DNA.

The new method could eliminate the expensive, time-consuming and error-prone strategy of DNA amplification.

“The current study shows that we can detect a much smaller amount of DNA sample than previously reported,” said Biomedical Engineering Associate Professor AmitMeller. “When people start to implement genome sequencing or genome profiling using nanopores, they could use our nanopore capture approach to greatly reduce the number of copies used in those measurements.”

The technique is based on detecting DNA molecules as they pass through silicon nanopores. Electrical fields surrounding the mouths of the nanopores are used to attract long, negatively charged strands of DNA and slide them through the nanopore where the DNA sequence can be detected. Since the DNA is drawn to the nanopores from a distance, far fewer copies of the molecule are needed.

Weirdly, the longer the DNA strand, the more quickly it found the pore opening.

“That’s really surprising,” Meller said. “You’d expect that if you have a longer ‘spaghetti,’ then finding the end would be much harder. At the same time this discovery means that the nanopore system is optimized for the detection of long DNA strands – tens of thousands basepairs, or even more. This could dramatically speed future genomic sequencing by allowing analysis of a long DNA strand in one swipe, rather than having to assemble results from many short snippets.

Meller and his team were able to to optimize the effect by using salt gradients to alter the electrical field around the pores. This increased the rate at which DNA molecules were captured and shortened the lag time between molecules, thus reducing the quantity of DNA needed for accurate measurements.

By boosting capture rates and reducing the volume of the sample chamber the researchers reduced the number of DNA molecules required from about 1 billion sample molecules to 100,000.

The study is published in Nature Nanotechnology.