Remote Activation of a Nanopore for High-Performance Genetic Detection Using a pH Taxis-Mimicking Mechanism

Abstract

Aerolysin protein pore has been widely used for sensing peptides and proteins. However, only a few groups explored this nanopore for nucleic acids detection. The challenge is the extremely low capture efficiency for nucleic acids (>10 bases), which severely lowers the sensitivity of an aerolysin-based genetic biosensor. Here we reported a simple and easy-to-operate approach to noncovalently transform aerolysin into a highly nucleic acids-sensitive nanopore. Through a remote pH-modulation mechanism, we simply lower the pH on one side of the pore, then aerolysin is immediately “activated” and enabled to capture target DNA/RNA efficiently from the opposite side of the pore. This mechanism also decelerates DNA translocation, a desired property for sequencing and gene detection, allowing temporal separation of DNAs in different lengths. This method provides insight into the nanopore engineering for biosensing, making aerolysin applicable in genetic and epigenetic detections of long nucleic acids.

Graphical Abstract

Nanopore is a new-generation biosensing technology. Its single-molecule sensitivity enables a variety of biomedical applications,^1,2 from gene sequencing^3,4 to genetic, epigenetic,^5–10 and proteomic detections.^11–20 Recently, aerolysin has been emerged as a new nanopore system for sensing peptides,^21,22 proteins,^23–27 polysaccharides,²⁸ and biological relevant polymers.²⁹ However, in contrast to most protein pores (e.g., Phi29,¹⁶ MspA,⁴ α−hemolysin,³⁰ and SP1³¹), there are much fewer studies on aerolysin-based nucleic acid detections. Representative studies include Pastoriza-Gallego et al.³² and Payet et al.³³ investigated the translocation of DNA through the aerolysin pore; and Long’s group³⁴ discovered that aerolysin’s conductance variation caused by short oligonucleo-tides (<10 nts) translocation can be used to discriminate the DNA length at the single nucleotide resolution. By now, a main challenge confronted by aerolysin for genetic detection is the extremely low capture efficiency for target nucleic acids (>10 nts). For example, previous studies^32,33 have shown that DNA cannot be captured by aerolysin at voltage <100 mV, and DNA translocation rate at voltage >100 mV is much (~ten fold) lower compared with the widely applied α−hemolysin pore. Such low capture efficiency is impractical for nucleic acids detection, thus lowering the enthusiasm for developing aerolysin-based genetic biosensor. In this letter, we present an interesting noncovalent approach to transforming aerolysin into a highly nucleic acids-sensitive nanopore through a remote pH modulation mechanism. The “activated” aerolysin can not only efficiently capture gene fragments but also decelerate nucleic acids translocation through the nanopore. Both properties are required for sequencing and precise genetic detection.

As shown in Figure 1a, we formed single aerolysin pores from the cis side of the lipid bilayer and presented target DNA D16 (16-nt, Table S1) in the cis solution. A positive voltage (+40 mV and +80 mV) was applied from the trans compartment, with cis grounded, to drive cis DNA toward the pore. The current trace in Figure 1a shows that when both cis and trans solutions were neutral, the nanopore conductance was not affected by D16, indicating that cis DNA cannot access the aerolysin pore. Strikingly, when lowering pH of the trans solution to 3.4 while keeping the cis solution (where DNA was presented) neutral, we immediately observed a large number of current blockades that reduced the current from I₀ of the empty pore to I with I/I₀ = 4.3% (Figure S1). The block frequency increased in proportion to the cis DNA concentration (Figure S2). Without DNA, aerolysin alone in this pH gradient did not generate blocks (Figure S3). These findings together suggest that the blockades appeared at low trans pH are produced by capturing DNA from the cis entrance.

Figure 1.

Activation of aerolysin by acidic solution on one side to capture DNA from the other side capture and translocation. (a) Model (up) and current trace (bottom) showing that DNA D16 (1 μM) on the cis side cannot be captured by aerolysin when both cis and trans solutions were at pH 7.4 (left) and can immediately produce a large number of current blocks when trans solution was changed to pH 3.4 while the cis solution remained at pH 7.4 (right). (b and c) Voltage-dependence of DNA capture rate (k_on) (b) and block duration (τ, c) at various acidic trans pH (see Figure S10a–e for histograms).

This pH effect on DNA capture also applied to DNA in the trans solution (Figure S4). Trans D16 did not affect the pore conductance in neutral solutions on both sides, but produced many blockades as the cis pH was lowered to 2.6 while the trans pH (where DNA was presented) remained neutral, suggesting that low cis pH enables DNA capture from the trans entrance. We also found that an RNA fragment can also be captured from one side (cis) by lowing pH of the opposite side (trans) (Figure S5). Overall, we conclude that, by increasing the acidity on one side, aerolysin can be remotely “activated” and enabled to capture nucleic acids from the opposite side.

Figure 1b shows that for all trans pH tested, the cis DNA capture rate (k_on) can be enhanced by the voltage. For example, k_on for trans pH 3.4 was increased by over 10-fold from 0.21 ±0.03 μM⁻¹ s⁻¹ to 2.8 ± 0.4 μM⁻¹ s⁻¹ as voltage increased from +20 mV to +80 mV. This capture efficiency is superior to that of the α−hemolysin pore, which is at the 1 μM⁻¹ s⁻¹ level at +100 mV³⁵ and close to zero at +80 mV. Studying the voltage-dependence of the capture rate can help to understand the nature of the capture procedure. In a capture event, a DNA molecule first migrates from the bulk solution to the pore opening. This is a diffusive step biased by an electric field outside the pore entrance.^36–38 It is characterized by a linear k_on −V relation.^9,36,37 on Next, the DNA is threaded into the pore for translocation, a step that needs to overcome an energy barrier due to the nanopore confinement of DNA end and/or DNA−pore interactions.^9,36,37 This barrier-limited capture allows k_on to be exponentially changed with the voltage (k_on ~ exp(V)).^9,36,37 The DNA capture by aerolysin can be considered as both voltage-biased diffusion-limited and barrier-limited procedures, but diffusion plays a more dominative role, as shown by better linear fitting of the k_on−V curves as trans pH decreases (section S1 in the Supporting Information). This implies that lowering trans pH can decrease the barrier for capturing DNA from the cis entrance.

Decelerated translocation of nucleic acids through the nanopore is a desirable property for sequencing and high-performance genetic detection.³⁹ This can be achieved by lowering the driving voltage. However, low voltage heavily reduces DNA capture efficiency. Here we found that “activated” aerolysin can prominently decelerate DNA translocation while maintaining high capture efficiency at low voltage (Figure 1a up-right traces and Figure 1c). The DNA block duration (τ) varied from 0.83 to 40 ms at voltage between +10 and ~+ 80 mV and trans pH 3.4−2.1. This time scale is longer than that for DNA translocation in α−hemolysin (~100 μs) by 2 or 3 orders of magnitude. For trans pH 3.4, we observed a hill-shaped τ−V curve, which suggests that DNA translocates through the pore above the peak voltage (+30 to ~+40 mV) but returns to the cis solution at smaller voltages.⁴⁰ For trans pH 2.6 and 2.1, we identified a monotonic τ−V relation, which indicates DNA translocation at all voltages. The cis-to-trans translocation of DNA was verified by PCR amplification of translocated DNA in the trans solution (Figure S6). At the same voltage, τ was consistently prolonged with lowering trans pH. For example, at +80 mV, τ was extended from 0.83 ± 0.1 ms to 1.6 ± 0.2 ms as trans pH was lowered from 3.4 to 2.1. The pH effect on τ suggests that, in addition to voltage, other factors may contribute to translocation deceleration.⁴¹ One of factors could be DNA protonation, which occurs on the N3 group of adenine (pK_a = 3.5) and cytosine (pK_a = 4.2) (Table S2). The resulting reduction of the negative charge on DNA (section S2 in the Supporting Information) may assist to slow down translocation.

Deceleration of DNA translocation allows studying how the DNA translocation time is correlated to the DNA length. The Long’s group has found that 2-, 3-, 4-, 5-, and 10-base short oligonucleotides can be discriminated in the aerolysin pore based on their characteristic blocking levels,³⁴ but characterization of longer oligonucleotide translocation remains limited in the neutral environment. We detected the translocation time of targets D5, D16, and D30, which contain 5, 16, and 30 nucleotides, respectively (Table S1), at cis pH 7.4 vs trans pH 3.4 and monitored at +40 mV (Figure 2). The current traces for the three DNAs show that longer DNA corresponded to longer block duration (Figure 2a). τ was extended from 0.26 ± 0.09 ms for D5, to 2.9 ± 0.5 ms for D16, and 4.1 ± 0.5 ms for D30 (Figure 2b), suggesting that longer DNA spends longer time to pass through the pore. The translocation speeds for D5, D16, and D30 were about 19, 5.5, and 7.5 bases per millisecond. They are in the same scale but not equal, suggesting that not only the oligonucleotide sequence but also other factors such as the nucleic acids−pore interactions may be involved in the complex translocation procedure. This result indicates the potential for gene fragment length discrimination, which is a long-term interest in our follow-up study.

Figure 2.

Discriminating DNA of different lengths from their block durations. (a) Current traces showing DNA blocks in the presence of 1 μM DNAs D5 (5 nts), D16 (16 nts), and D30 (30 nts) (Table S1 for sequences) in the cis solution with cis pH 7.4 vs trans pH 3.4, recorded under +40 mV. (b) Duration of blocks by the three different length DNAs. Histograms for τ are shown in Figure S11.

To understand the remote pH-modulation mechanism, we kept cis pH at 7.4 and observed the continuous increase of cis DNA capture rate (+40 mV) by lowering trans pH from 7.4 to 2.1. As shown in Figure 3a,b, the DNA capture rate increased in three phases: Initially, k_on remained zero from trans pH 7.4 to 5.0 and modestly increased up to trans pH 3.7; next, k_on sharply climbed up by tens of fold within a narrow pH range from 3.7 to 2.6; finally, k_on approximated to saturation as trans pH was continuously lowered. The sharp increase of capture rate corresponds to a 1.5 kcal mol⁻¹ activation energy drop from trans pH 4 to pH 2.6 (RT ln(k_on‑pH2.6/k_on‑pH4). The k_on−trans pH correlation can be described using

$$k_{\text{on}}=k_{\text{on\_S}}/\left(1+{10}^{\text{trans pH}-{\text{pH}}_{50}}\right)$$ (1)

where k_{on_S} is the saturate capture rate, and pH₅₀ is the trans pH at which k_on is increased to 50% of k_{on_S}. The fitting gave pH₅₀ = 3.2. This trans pH-dependent capture rate is consistent with the pH-dependent protonation probability, which is governed by the Henderson−Hasselbalch equation,

$$P_{\text{RH}}=1/\left(1+{10}^{\text{pH}-\text{p}{K}_{\text{a}}}\right)$$ (2)

where P_RH (between 0−1) is the probability of a residue (R) in the protonated state (RH⁺), and pK_a is the pH value at which P_RH = 50%. Comparison of eq 1 and eq 2 suggests that DNA capturing is enhanced by pH-induced protonation.

Figure 3.

Trans pH-dependent capture of cis DNA by the aerolysin pore and mechanistic study. (a) Current traces showing the consistent increase of block frequency for the cis DNA D16 (1 μM) as the trans pH was lowered from 7.4 to 2.6 (cis pH remained 7.4). (b) D16 capture rate (k_on) as the function of the trans pH. The k_on−trans pH curve was fitted using eq 1. (Top) The trans pH-dependent K⁺/Cl⁻ permeability ratio P₊/P₋ was shown with the k_on−trans pH curve (see Figure S7 for P₊/P₋ measurement). (Bottom) trans pH-dependent relative net anion flow J/J₀ and the k_on−trans pH curve. J = J₋−J₊ is the net anion flow and J₀ is the maximal anion flow in which the current is completely carried by anions (J₀ = J₋ + J₊). As P₊/P₋ = J₊/J₋, the permeability ratios were transformed into the relative net ion flow as J/J₀ = (1 − P₊/P₋)/(1 + P₊/P₋). (c) Aerolysin structure, including cross-section (left), top view (cis), bottom view (trans), showing charge distribution at the cis and trans entrance.

To prove the occurrence of protonation, we investigated the ion selectivity of aerolysin, supposing that protonation-induced positive charge on the channel wall can produce a net anion (Cl⁻) flow. Indeed, Figure 3b shows that the ratio of the Cl⁻ vs K⁺ flows through aerolysin dramatically increased from 7-fold (K⁺/Cl⁻ permeability ratio P₊/P₋ = 0.14) in neutral trans solution⁴² to 16-fold (P₊/P₋ = 0.06) at trans pH 3.7, and 50-fold (P₊/P₋ = 0.02) at trans pH 3.2 (Figure S7 for P₊/P₋ measurement). Such great enhancement of the net Cl⁻ flow should be due to the increase of positive charges in the pore, thus verifying the protonation under the pH gradient. Structurally,^43,44 the aerolysin pore offers rich protonatable residues on the channel wall and in particular around both entrances (Figure 3c), such as E237, E252, E254, and E258 close to the trans entrance for protonation at low trans pH, and D209 and D216 close to the cis entrance for protonation at low cis pH. Their protonation would induce a large number of net positive charges contributed by K238, K242, K244, and K246 in the trans barrel and/or R220, R282, and R288 around the cis entrance, which work together to produce a large net anion flow.

The observed pH-dependent anion flow allows us to propose that there exists a protonation-induced electroosmostic flow that drives DNA capturing by aerolysin. Previously, we have found that engineered α−hemolysin pore with different charge polarities can modulate the pore’s ion selectivity; the resulting net ion flow produced a nano-electroosmotic flow to drive neutral molecules binding with the pore.⁴⁵ In the current study, Figure 3b shows that the enhancement of the net Cl⁻ flow was synchronized with the increase of DNA capture rate as trans pH was lowered. This suggests that the protonation-induced Cl⁻ flow generates a cis-to-trans electroosmotic flow at a positive voltage, which can drive DNA into the pore from the cis entrance (Figure 1a model). Similarly, lowering cis pH can result in an opposite trans-to-cis electroosmotic flow under a negative voltage, enabling aerolysin to capture DNA presented in the trans solution (Figure S4 model). The electroosmotic effect is also supported by the observation that the DNA capture rate is linearly increased with the voltage (Figure 1b). This is consistent with the linear voltage-dependence of the electroosmotic flow as determined by the Helmholtz− Smoluchowksi equation⁴⁶ and that the capture rate is proportional to the water flow velocity that is increased linearly with the voltage applied.⁴⁵

Figure 3b (bottom) also shows that the net Cl⁻ flow reached the maximal level around pH 3.0 (J/J₀ = 95% at trans pH 3.2). Beyond pH 3.0, the electroosmotic effect becomes saturated, but the capture rate continued to increase. One possibility is that very low trans pH may even “remotely” protonate residues at the cis entrance (e.g., D209 and D216) through a [H⁺] gradient across the pore. Upon protonation, the local positively charged residues (e.g., R220, R282, and R288) could attract DNA in the bulk solution. To protonate the cis entrance D209 and D216 (pK_a = 3.86), the trans pH must be lower than this pK_a due to a pH gradient along the pore. This is consistent with the experiment result that the k_on−trans pH curve (Figure 3b) gave pH₅₀ = 3.2, lower than pK_a = 3.86. In addition, we lowered cis pH to 3.4 (trans pH remained 7.4) and observed cis DNA capturing blockades (Figure S8), verifying that the protonated cis entrance carrying positive charges can enhance DNA capturing. Moreover, placing charged motif at the protein pore’s entrance has been found to enhance DNA capture efficiency⁴⁷ and is a key design parameter that needs fulfilling to optimize the biosensing performance.⁴⁸

In summary, we have discovered a remote pH modulation mechanism to transform aerolysin into a highly nucleic acids-sensitive nanopore for potential genetic detection. This method is efficient while simple and easy to operate, and the resulting deceleration of DNA translocation is a desired property for both sequencing and precise genetic detection. In addition, this method is also applicable to other nanopore systems (Figure S9). Most importantly, unlike previous work in which pH of both sides were changed (for α−hemolysin),⁴¹ our method retains the target gene fragments in the neutral environment throughout the detection, therefore effectively preventing target nucleic acids from chemical damaging. In particular, this advantage is critical for preserving the structure of nucleic acids (e.g., RNA tertiary structures). This method may ignite new studies on nanopore sensing mechanism. Guided by the structure,⁴³ we can engineer the aerolysin pore with different charge distributions to mimic the protonation mechanism and to improve biosensing performance. We can also adapt this method to detection of genetic alteration such as circulating nucleic acids biomarkers^49–51 and molecular processes like folding and unfolding mechanisms of tertiary nucleic acid structures^52,53 and nucleic acids−ligand interactions.