Rapid nanopore discrimination between single polynucleotide molecules

Abstract

A variety of different DNA polymers were electrophoretically driven through the nanopore of an alpha-hemolysin channel in a lipid bilayer. Single-channel recording of the translocation duration and current flow during traversal of individual polynucleotides yielded a unique pattern of events for each of the several polymers tested. Statistical data derived from this pattern of events demonstrate that in several cases a nanopore can distinguish between polynucleotides of similar length and composition that differ only in sequence. Studies of temperature effects on the translocation process show that translocation duration scales as approximately T(-2). A strong correlation exists between the temperature dependence of the event characteristics and the tendency of some polymers to form secondary structure. Because nanopores can rapidly discriminate and characterize unlabeled DNA molecules at low copy number, refinements of the experimental approach demonstrated here could eventually provide a low-cost high-throughput method of analyzing DNA polynucleotides.

Figure 1

Typical current trace showing two DNA translocation events (within pairs of facing arrows) when a 120-mV gradient is applied across a lipid membrane containing one α-hemolysin channel. For each event, we measured the translocation duration, t_D, and the normalized blockade level defined as: I_B = 〈I_Event〉/〈I_Open〉, where 〈I_Event〉 denotes current average over the translocation duration, and 〈I_Open〉 denotes the average during 150 μsec before and 150 μsec after each event.

Figure 2

(a) Event diagram showing translocation duration vs. blockade level for poly(dA)₁₀₀ (blue) and poly(dC)₁₀₀ (red) at 20.0°C. The two polymers were examined separately. Each point on this diagram represents the translocation of a single molecule that was characterized by its translocation duration, t_D, and blockade current, I_B. (b) Current histogram projected from the above event diagram; the color codes are the same. The two peaks corresponding to the two groups of events are denoted by I_P1 and I_P2. The solid lines are fits of the data to a sum of two Gaussians. (c) Duration histogram projected from a for the first group of events. Note the well-defined peak locations, t_P, and the exponential temporal decay constants, τ_T. The temporal decay constant associated with the poly(dC) polymers is ∼7 times shorter than the τ_T associated with the poly(dA).

Figure 3

(a) Event diagram at 20.0°C for poly(dA₅₀dC₅₀) (orange) and poly(dAdC)₅₀ (green). As in Fig. 2, the two polymers were examined separately. The ovals contain >95% of the events in group 1 for each of the two polymer types. (Inset) The corresponding current (b) and duration (c and d) histograms, where c is the duration histogram for group 1 events and d, for group 2 events. The two polymers were readily discriminated in spite of their identical base composition. The temporal spread in group 2 of poly(dA₅₀dC₅₀) was much larger than that of poly(dAdC)₅₀ (see Table 1 for the values of τ_T2 for the two polymers).

Figure 4

(a) Event diagram at 25°C for poly(dC₅₀dT₅₀) (light-blue markers) and poly(dCdT)₅₀ (purple markers). As in Fig. 2, the two polymers were examined separately. The ovals contain more than 95% of the total number of events. (Inset) The corresponding current histogram (b) and the translocation duration histogram (c). Note that in contrast to the polymers shown in Figs. 2 and 3, the current histogram has a single peak, suggesting a single group of events for each of these two polymers.

Figure 5

Representative current trace showing 10 events recorded from a mixture of equal molar concentrations of poly(dA)₁₀₀ and poly(dC)₁₀₀. The time between events is truncated. The individual events are identified, on the basis of t_D alone, as traversal of a molecule of poly(dA)₁₀₀ or a molecule of poly(dC)₁₀₀. The confidence in the molecule identification, χ, was estimated by using the predetermined translocation time distributions (see Fig. 2c and text) and is given in percent. Of 999 events recorded in 4 min, the algorithm unambiguously (χ > 90%) identified 98% of the events as either poly(dA)₁₀₀ or poly(dC)₁₀₀.

Figure 6

Representative event diagrams for poly(dA)₁₀₀ (blue) and poly(dC)₁₀₀ (red) at three different temperatures: (a) 15.0°C; (b) 25.0°C; (c) 33.0°C. As in Fig. 2, the two polymers were examined separately. Note that because of the strong temperature dependence of t_D, a different vertical scale is used for each of the three plots. (Inset) The corresponding duration and current histograms from which we extracted I_B, t_P, and τ_T (see text).

Figure 7

Dependence of t_P for group 1 events for poly(dA)₁₀₀ (blue), poly(dC)₁₀₀ (red), poly(dA₅₀dC₅₀) (orange), poly(dAdC)₅₀ (green), and poly(dCdT)₅₀ (purple). Because all poly(dCdT)₅₀ events fall into one group, that group is considered group 1. All measurements were performed at 120 mV. Bars: SEM of more than five groups of measurements. With rising temperature between 15°C and 40°C, there is a 12-fold decrease of t_P1 for the slowest polymer poly(dA) and an 8-fold decrease of t_P1 for the fastest poly(dC). The dashed black line that matches closely to the poly(dA₅₀dC₅₀) data is the algebraic average between t_P1 of poly(dA)₁₀₀ and t_P1 of poly(dC)₁₀₀. Note that the temperature dependence is not exponential; rather, ∼T⁻² scaling (solid lines) yielded the best fit to the data.

Figure 8

Semilogarithmic plot of t_P2 as a function of temperature for poly(dA)₁₀₀ (solid triangles), poly(dA₅₀dC₅₀) (empty circles), and poly(dAdC)₅₀ (crosses). The lines connecting data points were drawn to guide the eye. The t_P2 values for poly(dAdC)₅₀ vary exponentially with temperature from 15°C to 40°, whereas the t_P2 values for poly(dA)₁₀₀ and poly(dA₅₀dC₅₀) show a greater divergence to larger values for the low temperature range (see text).