Multiplexed direct quantification of barcoded protein reporters on a MinION

Wednesday's Lightning session was opened by Jeff Nivala (University of Washington), who explained that he wasn't there to discuss DNA, or RNA, or, in fact, sequencing, but proteins (because "DNA and RNA are definitely not as cool as proteins"). Here, Jeff showed a clip demonstrating the unfolding of a protein by an unfoldase motor, which is then passed through a nanopore. However, Jeff noted that the structure of proteins, and their diversity, present more of a technical challenge for sequencing with a nanopore than DNA or RNA. Luckily, scientists are often interested in only a subset of proteins, some of which are intrinsically easier to detect - or can be tagged and engineered to be easier to detect. Reporter proteins, Jeff explained, are molecules that are linked to a cellular activity of interest, and used as a proxy for detection of protein expression or other biological processes. This is typically achieved via fluorescent proteins and optical methods, for example with the use of GFP, but spectral overlap limits multiplexing capacity to ~3-4 samples.

In order to detect many reporters simultaneously, Jeff's team developed NanoporeTERs: orthogonally-barcoded protein tags that are engineered for detection by nanopores. This is achieved with the use of a charged tail that can be electrophoretically captured, plus a  variable barcode region, which can be held statically in the nanopore as the other protein domains sit above it. Jeff demonstrated the raw nanopore signal produced when the variable barcode region of a NanoporeTER is captured by a nanopore; from this, the barcode and so the associated reporter protein can be directly identifed. As these are engineered to be expressed and secreted by the cell, no sample preparation is required, enabling a simple workflow that works even from raw cell culture. 20 barcodes are currently available for multiplexed protein expression detection, for applications such as detection of proteins in cell populations with differential expression, or to track the expression of reporters over time. Detection is sensitive down to the picomolar level, and the number of available barcodes is also rapidly increasing; Jeff notes that the theoretical limit is 20^17.