Nanopore sensing is the detection of a molecule coming into contact with a tiny hole, the nanopore. The contact may be the molecule passing through the nanopore, or a transient blocking of the pore. To sense the molecule, the nanopore is set in a electrically-resistant membrane so that an ionic current can pass through the nanopore when a voltage is applied across the membrane. Disruption of the current occurs when the molecule and nanopore come into contact with each other, and this disruption can be measured.

A nanopore may be used to identify a target analyte as follows:

If an analyte passes through the pore or near it, this event creates a characteristic disruption in current. Measurement of that current makes it possible to identify the molecule in question. For example, this system can be used to distinguish between the five standard DNA and RNA bases G, A, T, C and U, and also modified bases. It can also be used to identify target proteins and small molecules.

As the sensing measurements are based on electronics rather than optics, Oxford Nanopore's technology can be scaled to meet the experimental requirement. The technology has been miniaturised in the portable MinION Mk1B device as well as the benchtop instruments GridION and PromethION, which allows users to adjust throughput and sample numbers according to need. The portability of the technology also opens up a range of applicatons in new environments outside of the traditional laboratory.

Oxford Nanopore Technologies sensing platforms are founded on years of research and technological developments.

Over 20 years ago David Deamer and Dan Branton et al., based in Santa Cruz and Harvard, were working on the hypothesis that small holes existed in cells to allow the movement of molecules such as DNA into different compartments. Their work was based on the creation of an electrical current through the nanopore; facilitating the movement of DNA molecules through the pore one strand at a time. At the same time Hagan Bayley at the University of Oxford established that it was possible to detect the translocation of molecules through small holes in the membrane. By engineering proteins with pores at their centre, this would allow for the specific sensing of analytes. Mark Akeson, also at Santa Cruz, was developing methods to control the speed at which the DNA translocated through the pore by incorporating processive enzymes into the system that would slow the DNA through the pore. Through improved ability to control speed he made significant steps in being able to rapidly discriminate signals produced by bases passing through the nanopores.

Oxford Nanopore Technologies Ltd was formed in 2005 as Oxford Nanolabs Ltd with the aim of commercialising nanopore sensing technologies. The company was spun out of the University of Oxford with Professor Hagan Bayley as its academic founder and Dr Gordon Sanghera as CEO. Academic collaborators contributing early research to the project included Daniel Branton and Jene Golovchenko (Harvard), David Deamer and Mark Akeson (University of California Santa Cruz), and Amit Meller (Boston University).

In February 2012 at the AGBT conference, Oxford Nanopore announced the MinION Mk1B sequencing product and described progress on DNA strand sequencing, showing de novo base calls and read lengths of >50 kilobases (kb). The company also disclosed that phi X 174 bacteriophage had been analysed, becoming the first complete organism to be sequenced using a nanopore, sixteen years after the original concept.

Oxford Nanopore's work introduced robustness and functionality into the nanopore sensing system by addressing two types of challenges:

Nanopore sensing:

• Developed a synthetic polymer membrane that was robust enough to be shipped around the world and is resistant to complex biological samples, while maintaining the functioning nanopores
• Created a support that would preserve each individual nanopore and allow it to function independently of any others around it
• Designed the electronic interface which would allow signal to be detected and processed in real-time

Sequencing application:

• Controlled the speed of the DNA translocation through the nanopore by designing specialised molecules known as motors
• Engineered pores that could read the bases going through them
• Developed tethering mechanisms to enrich the molecules being measured at the membrane surface, to minimise the amount of analyte required
• Developed data analysis methods to process raw signal into DNA sequence

Nanopores are tiny holes, a few nanometres in diameter.

They occur naturally in biology as proteins embedded in cell membranes, providing a conducting channel through the centre. Oxford Nanopore Technologies has exploited the natural characteristics of protein nanopores and coupled them with the latest in high-speed, miniaturised electronics to develop versatile platforms capable of sensing many different molecules, including DNA, RNA, proteins, and small molecules. Oxford Nanopore continues to innovate new designs of pores for improved sensing characteristics for a variety of molecules.

The current products offered by Oxford Nanopore use biological nanopores, but the nanopores may be:

• Biological: formed by a pore-forming protein in an amphiphilic membrane such as a lipid bilayer or synthetic equivalent
• Solid-state: formed in synthetic materials such as silicon nitride or graphene
• Hybrid: formed by a pore-forming protein set in synthetic solid-state material

Protein nanopores

Oxford Nanopore's technology uses bespoke, proprietary pore-forming proteins to create pores in membranes. Pore-forming proteins are common in nature and are found naturally in cell membranes, where they act as channels for ions or molecules to be transported in and out of cells.

A protein nanopore can have an inner diameter of 1 nm, about 100,000 times smaller than that of a human hair. This diameter is the same scale as many single molecules, including DNA. The current pore (version R9) is a mutant of the CsgG lipoprotein from E. coli, which translocated polypeptides across the bacterial membrane. The CsgG lipoprotein is made of nine identical subunit which form the nonameric pore with a 36 stranded beta-barrel. The pore has been engineered to allow DNA, instead of peptides, to be translocated through the structure.

Modification of biological nanopores

Protein nanopores can be adapted at atomic-level precision using protein-engineering techniques and targeted chemical modifications. Specific adaptations can be designed to tune the nanopore's sensitivity to individual molecules. The wide range of possible modifications means that the we can create custom nanopore sensors for the detection of a very broad range of analytes; from metal ions and small molecules, to large protein assemblies.

Modifications may include:

• Genetic modifications to produce or enhance specific binding sites within the cavity of the pore
• Chemical or genetic attachment of affinity probes or ligands, such as; DNA oligos, aptamers, biotin, his-tags, sugars...etc
• Attachment of larger species to the outside of the pore, such as motor proteins or receptor sites
• The pore entrance can also be modified to enhance the binding of the molecular motors, such as the helicases used in our DNA sequencing chemistry

Nanopore production

Oxford Nanopore uses recombinant DNA technology to produce the nanopores. The bacteria cells are broken open and the resulting extract purified to generate a high concentration nanopore solution. A further treatment is then performed on the nanopores to ensure the consistency of the material. This end-to-end process can now be performed in a number of days.

The Company has also developed proprietary electronics that enable multiple nanopore sensing experiments to be performed in parallel, the data collected, and analysed in real-time.

For DNA sequencing, Oxford Nanopore uses a strand sequencing method, in which intact DNA strands are processed by the nanopores and analysed in real-time. The nanopore sequences the DNA fragments presented to it, regardless of their length. Oxford Nanopore Technologies provides users with the ability to analyse the native DNA molecule without the need for any amplification, thus avoiding the introduction of bias and loss of valuable information, such as base modifications and methylation sites. Given that the nanopores sequence the full length of fragments presented to them, the user can access very long reads (>100 kb), opening up the opportunity to sequence, for example, repetitive sections of DNA or RNA.

Data from each read is written into a file as soon as a strand has finished passing through the nanopore, and is available for analysis in real-time. As each of the nanopores act independently within the sensor array, as soon as a strand has completed passing through a pore, the pore becomes available for another strand to be analysed.

In Oxford Nanopore's strand sequencing method, the speed of translocation is controlled by the inclusion of a Motor protein. The Motor protein is a ratchet that moves the DNA strand through the nanopore. The speed of translocation can be controlled: the faster it runs, the more data is yielded per second.

The Motor protein is provided on a leader adapter that is introduced to the end of the double-stranded DNA template, and unzips the double strand as it feeds it through the nanopore, sequencing the intact strand. Nanopores can process read lengths of hundreds of kilobases and when a nanopore has processed a complete read, it will start a new one. There is no deterioration of accuracy as the DNA strand is sequenced. Sequencing will continue until there are no more strands, or until sufficient information has been collected and the sequencing experiment can be stopped.

The nanopore has been engineered to have a narrow part to the barrel, which houses the Reader (a combination of amino acid residues that contact and can discriminate between different nucleotides in the pore). As the DNA moves through the pore, the Reader reads the combination of nucleotides which create a characteristic disruption in the electrical current. This information can be used to determine the order of bases in the DNA strand. The Motor protein moves the strand one base at a time, controlling the speed that the strand passes the Reader.

The Ligation Sequencing Kit family prepares DNA or cDNA libraries for nanopore sequencing using a ligation-based method.

The method adapts DNA or cDNA molecules with a Leader adapter ligated on during the reaction. The single strand section of the leader adapter is pulled through the nanopore. The first strand sequenced is known as the template. Usually, the complement strandthen dissociates, and may be captured by the nanopore at a later time.

In 1D^2 chemistry, the complement strand is encouraged to translocate through the pore immediately after the template strand. The combined information from both strands can produce higher accuracy basecalling.

The Rapid Sequencing Kit prepares a genomic DNA sample for nanopore sequencing using a transposase-based method.

A DNA molecule is fragmented and adapted in the transposase step. A Y-adapter is then added on during the ligation step. This method requires only 10 minutes for library preparation and produces linear strands.

To maximise the amount of DNA read by a pore while maintaining input material requirements low, Oxford Nanopore Technologies has developed a method of tethering the DNA molecules to the surface of the membrane. This mechanism concentrates the material of interest near the pore, greatly increasing the sensitivity of the device and sequencing throughput. The tether is incorporated into the materials provided in the sequencing kits.

Oxford Nanopore devices incorporate all the components required for nanopore sensing. The nanopore is inserted into a polymer membrane, which is formed across a microsupport to provide structure. The microsupport positions each nanopore above an electrode for individual sensing.

In the presence of the electrically resistant membrane, a voltage can be applied and the resulting ionic current will only be able to pass through the nanopore. Changes to the ionic current (the nanopore signal) will be detected by the electrode. Each electrode is connected to an individual channel on an ASIC chip (Application Specific Intergrated Circuit), which controls and measures the nanopore signal. This complete unit of membrane, nanopore, electrode and ASIC channel is known as the the nanopore sensor. To complete the system, software is required to carry out control of the nanopore sensor and also to collect and process the signal.

Using methods common in microelectronics multiple nanopore sensors can be brought together - 100s to 1000s - in a sensor array.

The size of the sensor array can be varied to meet the demands of the experiment and are housed in a flow cell to control the environment of the nanopore.

The ASIC is capable of recording through all channels simultaneously, with the 2048 active well electrodes organised in groups of four. The choice of a single well for each of the recording channels is referred to as multiplexing. Multiplexing the nanopore array is used to improve the yield of channels containing a single pore and the sequencing output of the consumable. Is it common to see data provided with ‘channel’ and ‘MUX’ information. Most sequencing scripts perform a ‘MUX selection’ where the best 512 wells are chosen as the first group. Later groups of wells can be switched to during a sequencing run to provide more output.

Oxford Nanopore has developed proprietary electronics to allow hundreds to hundreds of thousands of nanopore sensing processes to be controlled and measured simultaneously. The key elements are the Company's proprietary sensor array, in which nanopores are embedded and a bespoke Application-Specific Integrated Circuit (ASIC).

Oxford Nanopore Technologies has developed custom ASICs for the MinION Mk1B, GridION, and PromethION devices. These ASICs are connected to the sensor arrays and measure the ionic current flow through each nanopore in the sensor array chip. The ASICs have been designed to measure current at tens of kHz per nanopore, while minimising measurement noise and maximising signal. Each ASIC contains a high-density array of low-noise amplifier circuitry, and can be scaled to measure from hundreds to thousands of channels in parallel in real-time.

Using microchip fabrication techniques, Oxford Nanopore has developed devices that enable highly scalable arrays of nanopores to be used for sensing at the single molecule level. These utilise a sensor array chip that consists of a number of microwells, each of which contains its own electrode, membrane and nanopore.

Each microwell electrode is connected to a common counter electrode positioned inside the flow cell, that is constructed over the top of the microwell array. The sensor chips are stored with nanopores embedded in a polymer membrane. The measurement circuitry in the accompanying ASICs allows each step in the process to be monitored for manufacturing and development purposes and selection of active nanopores during an experiment.

Figure: A typical silicone wafer. On the left are numerous sensor array chips in a standard manufacturing format, which can be seen in close-up view on the right.

Once the set-up process is completed and experimental analysis begins, the nanopore array is fully functional for as long as the user wishes to conduct their experiment. As the experiment progresses, each channel on the sensor chip streams experimental data in real-time. In the case of DNA sequencing, this means that full-length reads are processed individually in real-time, and as a read is completed a new DNA strand is acquired and starts to process. Real-time data acquisition and analysis means that users can monitor the results of their experiment as it progresses and can stop analysing the sample when enough data have been collected to answer their experimental question.

Movement of the DNA strand through the pore produces a series of current measurements (squiggles), each of which have a unique current signature and can be decoded to produce the sequence. The squiggles are converted to reads by the MinKNOW™ software, and can then be basecalled to produce the DNA sequence of each strand. The DNA sequences that are generated during the basecalling component of the analysis are made available to the users for further processing.

Please refer to the MinKNOW document, EPI2ME document and Data Analysis document for more information.