In nanopore sequencing, DNA passes through a tiny pore in a membrane, much like a thread goes through a needle. The pore also contains an electrical current. As each of the four nucleotides pass through the pore, they block the current in individual ways that can be used to identify them. The method is under development since the mid 1990s.
Very recently this method caught the attention of the community including us barcoders as Oxford Nanopore allowed a limited number of users access to the MinION device through their MinION Access Programme. A couple of talks (see here starting at about 32:00) at the last Barcode Conference demonstrated how close we might be to a miniature version of a handheld barcode scanner.
Although powerful, the method suffers from high speed: DNA goes through the pore too quickly to be read with best accuracy. Researchers at the Ecole Polytechnique Fédérale de Lausanne have now developed a way to overcome this problem by using a thick, viscous liquid that slows the passage of DNA by two to three orders of magnitude.
The researchers developed a 0.7 nm thick film made of molybdenum disulfide (MoS2). This is different from the current solid state nanopores that are made with silicon compunds or graphene. DNA is a fairly sticky molecule and MoS2 is considerably less adhesive than graphene. The team then created a nanopore on a membrane, almost 3 nm wide.
The next step was to dissolve DNA in a thick liquid (room-temperature ionic liquid) that contained charged ions and whose molecular structure can be fine-tuned to change its viscosity. The results are quite promising: Our technique, which exploits the high viscosity of room-temperature ionic liquids, provides optimal single nucleotide translocation speeds for DNA sequencing, while maintaining a signal-to-noise ratio higher than 10.
The colleagues also predict that using high-end electronics and control of the viscosity gradient of the liquid could further optimize their system. With their approach they hope to create a cheaper DNA sequencing platform with a better output.
To conclude, we have demonstrated that single-nucleotide identification can be achieved in MoS2 nanopores by using a viscosity gradient to regulate the translocation speed. The viscosity gradient system can not only be used in standard ionic sensing experiments, but can be potentially combined with other schemes of nanopore sensing such as transverse current signal detection. The ultrahigh viscosity of ionic liquids results in reduced capture rates. Therefore, an optimal experimental configuration would capitalize on high-end electronics and the viscosity gradient system presented here with a suitable capture rate. We believe that combining ionic liquids and monolayer MoS2 nanopores, together with the readout of transverse current either using the tunnelling or FET modality, would meet all the necessary requirements for DNA strand-sequencing such as optimal time resolution and signal resolution. This could also be achieved in a platform that allows multiplexing, thereby reducing costs and enhancing signal statistics.