Overview: IBM is developing a nanostructured DNA sequencing device to electronically detect individual bases and rapidly sequence genomes. The technology is innovative yet unproven, and faces major technical hurdles. If experiments are successful, the device will be highly competitive. Its device would be a low-cost alternative to on-market other development-stage technologies at Illumina (ILMN), Life Technologies (LIFE), Pacific Biosciences, Ion Torrent, and other leading companies.

Background: IBM plans to build a DNA sequencing device comprised of nanometer pores with internal surfaces of axially-stratified, alternating nanoscale layers of metal and dielectric material. Individual, in-tact, single-stranded DNA molecules are to be alternatively passed through the pores and held in place for electronic interrogation of individual bases. The approach aims to control DNA’s local motion and configuration at an unprecedented level. IBM has carried out computational and theoretical work, but lacks experimental verification[1]. This is not the first time IBM has ventured into biology. For example, IBM is developing methods to use DNA in the assembly of finer-featured silicon chips[2], has an experimental R&D group focused on microfluidics and biopatterning, and has filed patents covering the manipulation of biological molecules related to the scanning tunneling and related atomic force microscopy technologies that its researchers famously discovered[3].

Analysis: IBM’s cost savings would result from fewer reagents, no need for labels and optical instrumentation, and long read lengths. In comparison to other nanopore-based sequencing technologies (e.g. Oxford Nanopore Technologies/Illumina and NABsys), IBM’s device would be more stable and less expensive. IBM faces two critical technical challenges. Its device must: 1) distinguish the signal of a single base from the signals of nearby surrounding bases in the DNA chain; 2) manipulate and control DNA to slow down translocation and to enable optimal base orientation and/or sufficient sampling of a base throughout its range of configurations.

Regarding the challenge of base measurement, IBM will almost necessarily take a different approach to measurement than Oxford Nanopore, likely involving direct measurement of tunneling currents or capacitance. This approach could obviate limitations of spatial resolution and sensitivity associated with the Oxford Nanopore approach[4], and recent simulations have shown that a host of complicating phenomena such as the effect of ionic motion in a nanopore may not obscure the desired signal of an individual base[5]. Oxford Nanopore does not need to achieve nearly the same degree of measurement sensitivity as IBM, as the current Oxford approach involves interrogation of cleaved bases, and not in-tact DNA strands. NABsys is also taking a very different approach with lower measurement sensitivity requirements. A search of the patent literature reveals the first IBM patent covering its DNA transistor technology[6], and indeed IBM is seeking coverage with its technology of tunneling current and capacitance measurements.

Schematic of nanopore (from IBM's US 2008/0187915)

Regarding the challenge of precise DNA translocation, the axially-stratified layers of IBM’s proposed nanopores represent an entirely new approach to the problem. In its initial patent, IBM presents an innovative approach to manipulate nucleic acids, including voltage application to draw charged polymers into nanopores (or advance segments within pores), in combination with electrostatic potential wells that interdigitate with individual bases for trapping and analysis. The electrostatic wells could be used in combination with other electrostatic wells in the same pore[6]. IBM researchers estimate that it is possible to achieve a frequency of around 500 MHz with such a device, which would translate into a theoretical read rate of 0.5 billion bases per second[7]. Although there are polymer structure limitations which make this exceptionally high rate unlikely, it is clear that voltage cycling times will not be a limiting factor. There are several approaches IBM may take in upcoming experiments. It will need to focus on synchronizing DNA movement by voltage application with base-level movement using potential wells, while ensuring that DNA-wall interactions (possibly affected in a sequence-dependent manner), the time-dependent quantity of DNA on either side of a nanopore during sequencing (and hence changing drag force), and inconsistencies in test fluid composition do not reduce read fidelity below acceptable levels. Finally, fabrication of such a nanostructured material is non-trivial and represents a challenge to IBM, as well as serious barrier to other companies and researchers in the field. IBM researchers indicate several possible fabrication methods, including electron and ion beam techniques[7].

For additional information on IBM’s technology or analysis of related technologies and companies, contact: bruce@schiamberggroup.com

Copyright © Bruce A. Schiamberg 2010. All rights reserved.

References:

[1] http://www-03.ibm.com/press/us/en/pressrelease/28558.wss
[2] http://www-03.ibm.com/press/us/en/pressrelease/28185.wss
[3] US 7211789
[4] Branton et. al., Nature Biotechnology, vol. 26, no. 10, p. 1146 (2008)
[5] Krems et. al., Biophysical Journal, vol. 97, no. 7, p.1990 (2009)
[6] US 2008/0187915
[7] Polonsky et. al., IBM Research Report: DNA Transistor, RC24242 (W0704-094), April 2007