Nucleotide discrimination with DNA immobilized in the MspA nanopore

Abstract

Nanopore sequencing has the potential to become a fast and low-cost DNA sequencing platform. An ionic current passing through a small pore would directly map the sequence of single stranded DNA (ssDNA) driven through the constriction. The pore protein, MspA, derived from Mycobacterium smegmatis, has a short and narrow channel constriction ideally suited for nanopore sequencing. To study MspA's ability to resolve nucleotides, we held ssDNA within the pore using a biotin-NeutrAvidin complex. We show that homopolymers of adenine, cytosine, thymine, and guanine in MspA exhibit much larger current differences than in α-hemolysin. Additionally, methylated cytosine is distinguishable from unmethylated cytosine. We establish that single nucleotide substitutions within homopolymer ssDNA can be detected when held in MspA's constriction. Using genomic single nucleotide polymorphisms, we demonstrate that single nucleotides within random DNA can be identified. Our results indicate that MspA has high signal-to-noise ratio and the single nucleotide sensitivity desired for nanopore sequencing devices.

Figure 1. Schematic diagram.

Schematic diagram of MspA (blue) set up in a lipid bilayer (grey). Single stranded DNA (ssDNA) was attached to a bulky NeutrAvidin molecule (green) using a biotin linker (black). A specific nucleotide (red) is designated by it's position, X, from the biotin-NeutrAvidin ‘anchor’. The ssDNA threads through the pore from the cis side of the bilayer until the bulky NeutrAvidin prevents it from further translocation. Residual ion current was recorded as the ssDNA is immobilized within MspA.

Figure 2. Orientation specific homopolymer histograms.

Mean residual ionic current (gray) and fitted Gaussian curves are shown for homopolymer adenine (‘poly-dA’, black), cytosine (‘poly-dC’, red), thymine (‘poly-dT’, green), and guanine (‘poly₃-dG’, blue) for (A) 3′ threading and (B) 5′ threading. With 3′ threading, poly-dA and poly₃-dG overlap by 27% of the area of the two Gaussians. With 5′ threading, poly-dA and poly₃-dG overlap by 2% and poly-dC and poly-dA overlap by 0.1% of the area of the two Gaussians.

Figure 3. Methylated cytosine.

Mean residual ionic currents and the fitted Gaussian curve are shown for ssDNA of homopolymer cytosine (‘poly-dC’, red) and a strand with 4 methylated cytosines (‘poly₄-mC’, black) substituted at the 13^th–16^th nucleotides from the NeutrAvidin anchor in an otherwise homopolymer cytosine strand. Experiments were conducted on three pores with poly₄-mC and poly-dC added sequentially to each pore. The peaks of the Gaussians are separated by 1.1 pA and the curves overlap by ∼2% of the area of the two Gaussians.

Figure 4. Single heteromeric substitutions in homopolymer ssDNA.

Peak values and half width half height values, represented as error bars, of fitted Gaussians of mean residual ionic currents are shown for experiments with a single heteromeric substitution in homopolymer ssDNA. For comparison, the mean residual ionic currents for relevant homopolymer strands are shown as dashed lines. Gaussian curves are included to aid the eye. (A) DNA strands containing a single cytosine (dC) substituted at nucleotide position X from the NeutrAvidin anchor in an otherwise poly-dA strand. Large error bars (X = 13) indicate a wide distribution and the two points at X = 13 and X = 16 indicate two current levels for that substitution. The residual current differs most from that of poly-dA (black dashed line), most resembling that of poly-dC (red dashed line) at X = 14 and 15. (B) DNA strands containing a single cytosine (dA) substituted at nucleotide position X from the NeutrAvidin anchor in an otherwise poly-dC (red) or poly-dT (green) strand. A single dA substitution in homopolymer poly-dC yields a residual current similar to that of poly-dC (red dashed). For a single dA substitution in homopolymer poly-dT, the current deviates most from that of poly-dT (green dashed) towards the level for poly-dA (black-dashed) at X = 14 and 15.

Figure 5. SNP histograms (rs889312).

A segment of DNA containing SNP rs889312 was bound to NeutrAvidin such that the polymorphism is at the 13–16^th position from the NeutrAvidin anchor (denoted dC_X or dA_X where X was the position of the SNP). Part of the surrounding sequence is shown with the SNP highlighted in red (see Table S1 for complete sequence). Histograms of mean residual current levels are shown for each variant and SNP location, X = 13–16. The two variations are most clearly resolved for X = 14 and 15. For reference, the mean residual ionic currents for poly-dC (red dashed) and poly-dA (black dashed) are shown.