α-hemolysin nanopore (ribbon diagram) with covalently attached cyclodextrin (teal) transiently binds a base (red) traversing the pore. Credit: Oxford Nanopore Technologies

The quest for the Holy Grail of the $1,000 genome has led to big improvements in sequencing technologies. The original Sanger sequencing brought us the first human genome sequence for a price tag of billions of dollars; then second-generation sequencers—Illumina's Genome Analyzer, Roche's 454 sequencer and Applied Biosystems' SOLiD System—resequenced a human genome for under $100,000. Now a third generation of sequencers, able to determine the base composition of single DNA molecules, has joined the race.

Currently three technology platforms comprise this third generation. Two achieve single-molecule sequencing by incorporating and detecting fluorescently labeled nucleotides: Helicos' Genetic Analysis System is already commercially available, and Pacific Biosciences is ready to launch commercial distribution of their Single Molecule Real Time (SMRT) technology next year. The third, Oxford Nanopore's nanopore sequencing, is not ready for commercial prime time, but is likely to be the cheapest of the three.

“Nanopore sequencing has the advantage that it does not require any labeling of the DNA, no expensive fluorescent reagents or really expensive CCD [charge-coupled device] cameras to record from optical chips,” says Hagan Bayley from Oxford Nanopore Technologies.

In their most recent work, Bayley and his team focused on improving a nanopore that can read individual nucleoside monophosphates. Building on their previous work, they designed the nanopore from α-hemolysin with the molecular adapter cyclodextrin covalently attached to the inside of the pore. An exonuclease digests single-stranded DNA, and as single bases fall into the pore, they transiently interact with the cyclodextrin and block an electrical current that runs through the pore. The current amplitude—characteristic for the individual bases A, G, C and T as well as methylcytosine—is then easily converted to DNA sequence. Each base has a characteristic mean dwell time in the millisecond range, its dissociation rate constant isvoltage-dependent and a potential of +180 mV ensures that the base is swept out of the pore on the other side.

The ability to directly read the fifth base, methylcytosine, without prior bisulfite conversion, is unique to this nanopore technology and, judging by the feedback in the blogosphere, is already creating great excitement among researchers seeking to sequence the epigenome.

Nanopore sequencing promises to meet the needs for most sequencing users: the 99.8% read accuracy is high, and errors will be easier to correct computationally because the amplitude-based recordings may cause an ambivalent call between two but not between all four bases. Homopolymer stretches are recorded without difficulty, as the pore records each base irrespective of what comes before or after. Reads will be long. “It does have the potential to read thousands of bases,” says Bayley, “and there is no degradation of sequence quality. Even if there is a glitch in the middle, it will pick up again.”

But there are two important technical issues that still need to be addressed before Oxford Nanopore's sequencers are ready for prime time. One is the optimal attachment of the exonuclease to the pore. “The real challenge,” says Bayley, “is to get it to drop every single base into the pore.” The other is parallelization, which would be achieved by creating a chip with tens of thousands of pores to ensure a fast overall sequencing process.

With cheap, long and accurate reads, nanopore sequencing may well bring the Holy Grail within grasping distance.