Simultaneous detection of CMPA and PMPA, hydrolytes of soman and cyclosarin nerve agents, by nanopore analysis

Sensors and Actuators B 176 (2013) 625–631

Contents lists available at SciVerse ScienceDirect

Sensors and Actuators B: Chemical journal homepage: www.elsevier.com/locate/snb

Simultaneous detection of CMPA and PMPA, hydrolytes of soman and cyclosarin nerve agents, by nanopore analysis Jyoti Gupta a , Qitao Zhao a , Guihua Wang a , Xiaofeng Kang b , Xiyun Guan c,∗ a

Department of Chemistry and Biochemistry, The University of Texas at Arlington, 700 Planetarium Place, Arlington, TX 76019-0065, USA College of Chemistry and Materials Science & Ministry of Education Key Laboratory of Synthetic and Natural Functional Molecular Chemistry, Northwest University, Xian 710069, PR China c Department of Biological and Chemical Sciences, Illinois Institute of Technology, Chicago, IL 60616, USA b

a r t i c l e

i n f o

Article history: Received 10 April 2012 Received in revised form 15 September 2012 Accepted 10 October 2012 Available online 22 October 2012 Keywords: Nanopore Stochastic sensing Nerve agents ␣-Hemolysin Degradation products

a b s t r a c t A sensitive nanopore-based analytical method was developed for the detection of cyclohexyl methylphosphonic acid (CMPA) and pinacolyl methylphosphonate (PMPA), hydrolytes of nerve agents soman and cyclosarin, respectively. The method uses a multi-functionalized ␣-hemolysin protein ion channel as the sensing element, with a host molecule ␤-cyclodextrin lodged in the lumen of the channel as a molecular adapter. The capture and release of CMPA/PMPA by the ␤-cyclodextrin host caused current modulations in the nanopore. Since the event residence times and amplitudes were significantly different, CMPA and PMPA could conveniently be differentiated and simultaneously quantitated. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Nanopore-based stochastic sensors can detect analytes at the single-molecule level, offering the potential as a highly sensitive, rapid, and multi-functional sensing platform. By monitoring the ionic current modulations induced by the passage of the analyte of interest through a single nanopore at a fixed applied potential, information about the identity and concentration of the analyte could be revealed, the former from the mean residence time ( off ) of the analyte coupled with the extent of current blockage (amplitude), while the latter from the frequency of occurrence (1/ on ) of the recorded blockage events [1]. In addition to the development of ultrasensitive sensors for a wide variety of substances at the single-molecule level [2–11], these nanometer-sized pores have been used to investigate covalent and non-covalent bonding interactions [12–14], probe enzyme kinetics [15], study biomolecular folding and unfolding [16,17], and analyze DNA molecules [18–29]. Organophosphorus chemical agents, commonly known as nerve gases, are the most toxic group in chemical warfare agents.

∗ Corresponding author. Tel.: +1 312 567 8922; fax: +1 312 567 3494. E-mail address: [email protected] (X. Guan). 0925-4005/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.snb.2012.10.058

These compounds were developed just before and during World War II and are chemically related to the organophosphorus insecticides. They disrupt the nervous system by irreversibly binding to acetylcholine esterase, an enzyme that relaxes the activity of the neurotransmitter acetylcholine. Thus far, although various methods have been developed for the detection of these nerve agents or their simulants [30–42], most of them cannot satisfy our needs in terms of sensitivity, selectivity, portability, low cost, ease of use, and rapid response. A good overview of the current analytical approaches was recently provided by Jenkins and Bae [36]. Whereas techniques such as HPLC and GC/MS offer the requisite sensitivity, they are relatively expensive and not easily portable. Additionally, the complex procedures involved are not conducive to their deployment as early warning detectors. In contrast, real time detection methods, such as those involving QCM and SAW technologies, lack specificity thus generating false positive signals such as those created by chemically similar, but far less toxic compounds (e.g., many common pesticides). Biologically based sensors involving enzymes and antibodies, while providing generally high sensitivity and good specificity, also suffer from time delays as well as stability and one-time use concerns. More recent advances involving Ion Mobility Spectrometers (IMS) represent an interesting approach and one certainly worthy of further evaluation. The IMS approach is sensitive, selective, and portable. However, it has a

626

J. Gupta et al. / Sensors and Actuators B 176 (2013) 625–631

limited linear dynamic range and poor resolution, and is subject to interferences from complex matrices although recent studies suggest that these weaknesses or limitations could potentially be overcome [42]. In our previous study [43], we have developed a stochastic sensing system to detect nerve agent simulants CMPA and PMPA, which are hydrolysis products of GD and GF [44], respectively. The sensing element was an engineered ␣-hemolysin (␣HL) (M113F/K147N)7 protein pore, and the host molecule ␤CD was used as a molecular adapter. Although this sensing system was sensitive (capable of detecting nanomolar PMPA/CMPA) and highly selective (without interference from other organophosphates), PMPA and CMPA produced current modulation events with very similar mean residence times (0.83 ms versus 0.62 ms) and blocking residual currents (0.08 pA versus 0.31 pA), and hence were difficult to be differentiated. It should be noted that, although poisoning by various nerve agents has relatively the same type of symptoms and receives the same emergency treatments (e.g., atropine or oximes), the capability to the rapid differentiation of one nerve agent against another is highly important. It has been reported that some antedotes are ineffective for certain nerve agents. For example, obidoxime is effective with sarin and VX, but ineffective against cyclosarin. Oxime HI 6 has been shown to be able to reactivate acetylcholine esterase inhibited by cyclosarin, but is ineffective with tabun [45]. Recently, we found that by using a multi-functionalized heteroheptameric ␣HL protein (M113K)3 (M113Y-D8)4 pore, the difference between the residence times and amplitudes of PMPA and CMPA events became much larger, permitting the convenient differentiation between PMPA and CMPA [46]. In this work, we investigate whether the enhanced sensor resolution of the multi-functionalized heteroheptameric ␣HL protein nanopore approach could be utilized for the simultaneous detection and quantification of CMPA and PMPA mixture.

2.2. Single-channel recording and data analysis Bilayer experiments have been described previously at 22 ± 1 ◦ C [45]. Briefly, the experiments were carried out in a chamber which was divided by a Teflon septum into two compartments, cis and trans. A bilayer of 1,2-diphytanoylphosphatidylcholine was formed on an aperture in the septum by using the Montal-Mueller method [47] after pretreating the aperture with hexadecane. The experiments were performed under a series of symmetrical conditions with a 2.0 mL electrolyte solution comprising 1 M NaCl and 10 mM NaH2 PO4 , with the pH of the solution adjusted to 3.0 using hydrochloric acid unless otherwise stated. The ␣HL protein (with the final concentration at 0.2– 2.0 ng mL−1 ) was added to the cis compartment. The host compound ␤-cyclodextrin (␤CD) and the analyte PMPA/CMPA were added to the trans compartment. The cis compartment was connected to “ground”. Unless otherwise noted, the applied potential was −120 mV by using a patch-clamp amplifier and head-stage (Axopatch 200B, Molecular Devices; Sunnyvale, CA, USA). A negative potential indicates a lower potential in the trans chamber of the apparatus. The resultant ionic currents were sampled at 25 kHz and the amplified signals were filtered at 5 kHz by a computer equipped with a Digidata 1440 A/D converter (Molecular Devices). For data acquisition and analysis, the pClamp 10.0 software package (Molecular Devices) were used. All the results were reported as mean values ± standard deviation. 2.3. Molecular graphics The software SPOCK 6.3 was used for the derivation of the model of the (M113K)3 (M113Y-D8)4 ␣HL pore. Briefly, the structure of the (M113K)3 (M113Y-D8)4 protein was obtained by reading the new amino acids from the library in the $SP AALIB directory, and superimposing the C␣ C␤ bonds onto the residues of the wild-type ␣HL (PDB: 7AHL) based on Mackay’s quaternion method. The image in Fig. 1 was displayed using PyMol (DeLano Scientific, Palo Alto, CA).

2. Experimental 3. Results and discussion 2.1. Materials and reagents n-pentane was purchased from Burdick & Jackson (Muskegon, MI). 1,2-Diphytanoylphosphatidylcholine was obtained from Avanti Polar Lipids (Alabaster, AL). Teflon film (25 ␮m thick) was bought from Goodfellow (Malvern, PA). Except for CMPA, which was obtained from Cerilliant Corporation (Round Rock, Texas), all other reagents, including PMPA, were purchased from SigmaAldrich (St. Louis, MO). Both PMPA and CMPA were dissolved in HPLC-grade water (ChromAR, Mallinckrodt chemicals) with the concentrations of all the stock solutions were 10 mM each. Both hexadecane [10% (v/v)] and 1,2-diphytanoylphosphatidylcholine (10 mg/mL) were dissolved in n-pentane. The production of the mutant ␣HL genes and the assembly of the heteroheptameric ␣HL pores have been described elsewhere [43,46]. Briefly, M113K and M113Y-D8 genes were constructed by site-directed mutagenesis with a WT ␣HL gene in a T7 vector (pT7-␣HL). The heteroheptameric M113K/M113Y-D8 pores were obtained by co-assembling M113K and M113Y-D8 ␣HL subunits in various ratios (from 1:6 to 6:1) on rabbit erythrocyte membranes. The resulting heteroheptamers, (M113K)1 (M113YD8)6 , (M113K)2 (M113Y-D8)5 , (M113K)3 (M113Y-D8)4 , through to (M113K)6 (M113Y-D8)1 , were separated and purified by SDSpolyacrylamide gel electrophoresis based on their different gel shifts, which were caused by the C-terminal extensions of eight aspartate residues (the “D8” tail). The ␣HL proteins obtained were stored in aliquots at −80 ◦ C.

3.1. The effect of solution pH and applied voltage bias on the nanopore sensor resolution In nanopore analysis, typically a buffer solution containing 1 M NaCl or 1 M KCl at or near physiological pH (i.e., pH 7.4) is used to produce the open channel current which is monitored [43]. However, in a separate study, we found that a reduction in the pH of the electrolyte solution could enhance the nanopore resolution to DNA analysis [48]. To examine whether a low pH value of the electrolyte solution could also provide a better resolution to PMPA/CPMA differentiation than a higher pH solution, initial experiments were performed to detect PMPA/CPMA using the (M113K)3 (M113Y-D8)4 pore (Fig. 1a) and ␤CD host at −120 mV in 1 M NaCl solutions at pH 3.0 and pH 6.5. In such nanopore sensing systems, ␤CD’s binding to the ␣HL pore would result in partial blockade of the channel. The followed capture and release of PMPA/CMPA by the ␤CD host would cause further current modulations (Fig. 1b–d). Our experimental result showed that, at pH 3.0, the PMPA and CPMA events had amplitudes of 26.8 ± 0.5 and 20.7 ± 0.8 pA, respectively, and residence times of 1.18 ± 0.17 and 0.17 ± 0.01 ms, respectively (Figs. 1 and 2). In contrast, at pH 6.5, the event mean blockage amplitudes of PMPA and CPMA were 20.3 ± 0.1 pA, and 18.3 ± 0.8 pA, respectively, while their event mean residence times were 1.40 ± 0.02 ms, and 0.38 ± 0.01 ms, respectively (data not shown). Therefore, both the amplitude difference (∼2 pA) and residence time difference (∼3.7 folds) of PMPA/CPMA events at pH

J. Gupta et al. / Sensors and Actuators B 176 (2013) 625–631

627

Fig. 1. Nanopore detection of nerve agent hydrolytes. (a) Molecular graphics representation of the staphylococcal ␣HL protein (M113K)3 (M113Y-D8)4 pore, where the naturally occurring Met residues at position 113 have been substituted with Lys (red) and Tyr (blue). Other mutant ␣HL heteroheptameric pores used in this work have different ratios of M113K and M113Y-D8 subunits. (b–d) Typical trace segments after addition of 10 ␮M PMPA, 30 ␮M CMPA, and a mixture of 10 ␮M CMPA and 30 ␮M PMPA, respectively to the (M113K)3 (M113Y-D8)4 pore in the presence of 40 ␮M ␤CD. The experiments were performed in a solution containing 1 M NaCl and 10 mM NaH2 PO4 (pH 3.0). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

a

1400

300

1200

Counts

Counts

1000 800 600 400

200

100

200 0 10

15

20

25

30

35

0

40

Amplitude (pA)

b

30 20

2.5

0.5 1.0 1.5 Dwell time (ms)

2.0

2.5

0.5 1.0 1.5 Dwell time (ms)

2.0

2.5

200 100

10 15

20

25

30

35

0

40

Amplitude (pA)

140

600

120

500

100 80

Counts

Counts

2.0

300

40

Counts

Counts

50

60 40

400 300 200 100

20 0 10

1.5

400

60

c

1.0

Dwell time (ms)

70

0 10

0.5

15

20

25

30

Amplitude (pA)

35

40

0

Fig. 2. (Left) Event amplitude histograms; and (right) residence time histograms, showing transient single-channel current blocks of the protein pore by the nerve agent hydrolytes in the presence of 40 ␮M ␤CD: (a) 10 ␮M PMPA; (b) 30 ␮M CMPA; and (c) a mixture of 30 ␮M CMPA and 10 ␮M PMPA. The experiments were performed in the (M113K)3 (M113Y-D8)4 pore in a solution containing 1 M NaCl and 10 mM NaH2 PO4 (pH 3.0). Data analysis was carried out after filtering the traces at 2 kHz.

628

J. Gupta et al. / Sensors and Actuators B 176 (2013) 625–631

6.5 were much smaller than those (∼6.1 pA and ∼6.9 folds, respectively) obtained at pH 3.0. This confirms that a lower pH buffer solution indeed offered an enhanced resolution for the differentiation of PMPA and CPMA. One possible reason is that the change in the pH of the solution affects the charge selectivity of the protein pore [48]. It has been reported that the charge selectivity of the protein pore would become more anion selective as the pH of the solution decreased [49]. Further, at pH 6.5, PMPA/CPMA has a net negative charge, which would decrease with a reduction in the pH of the electrolyte solution. For the sensitive and convenient detection of PMPA/CMPA, it is important that both the analyte and the host molecule ␤CD produce events having large mean residence times and frequencies; further, a large event blockage amplitude for the analyte is also desired. To this end, the detection of 10 ␮M PMPA using the (M113K)3 (M113YD8)4 pore was investigated at various applied potentials ranging from −40 mV to −180 mV in the presence of 40 ␮M ␤CD. Our experimental results (Fig. 3) showed that, as the applied voltage bias increased, the event mean blockage amplitudes and frequencies for both ␤CD and PMPA increased continuously, while their event mean residence times first increased and then decreased. The continuous increase in the event frequency with an increase in the applied voltage bias may be attributed to the enhanced electro-osmotic flow. The result regarding the effect of the applied potential on the event dwell time suggests a binding interaction between the host molecule ␤CD and the guest PMPA. The value  off would decrease monotonically with increasing transmembrane potential biases in the absence of a ␤CD–PMPA interaction. This voltage-dependent phenomenon has been well documented in the nanopore field [50–52]. Since the event signatures (i.e., dwell time, amplitude, and frequency) at −120 mV had relatively large values, −120 mV was used as the optimum potential for our nanopore sensor system. Note that at a higher applied voltage bias, the lipid bilayer used in the experiment would become less stable. 3.2. The effect of mutant ˛HL pores on PMPA/CMPA differentiation The interaction between ␤CD and PMPA/CMPA was then investigated with various ␣HL heteroheptamer pores (Fig. 1a) in an electrolyte solution containing 1 M NaCl and 10 mM NaH2 PO4 (pH 3.0) at −120 mV. The protein nanopores examined included (M113K)1 (M113Y-D8)6 , (M113K)2 (M113YD8)5 , (M113K)3 (M113Y-D8)4 , (M113K)4 (M113Y-D8)3 , (M113K)5 (M113Y-D8)2 , and (M113K)6 (M113Y-D8)1 . The M113YD8 ␣HL subunit was chosen because it has been shown that ␤CD could stay inside its channel for a significantly (more than 10,000 folds) longer time than in the wild-type ␣HL pore [49], thus facilitating the study of host-guest interaction. Although the residence time of ␤CD in the mutant ␣HL (M113K)7 pore is similar to that in the wild-type ␣HL pore, the positively charged Lys residues in the M113K subunit may interact with the polar phosphonate group of CMPA/PMPA. Taken together, the combination of these two residues (Tyr and Lys) inside a single nanopore may provide a feasible approach to improve the organophosphate differentiation resolution. The experimental results obtained with various pores are summarized in Table 1. It is apparent that, with an increase in the M113K/M113Y-D8 subunit ratio, the mean residence time of PMPA/CMPA first increased and then decreased, while the blockage amplitude for both PMPA and CMPA did not change significantly. This result supports our hypothesis that the M113K subunit of the protein pore may interact with CMPA/PMPA. The  off value would not vary much or decrease monotonically if the residence of CMPA/PMPA in the ␤CD’s cavity is only related to the strength of the host-guest interaction or the strength between ␤CD and the different protein pores (note that the M113Y-D8 subunit

binds to ␤CD much more strongly than the M113K subunit) [53,54]. Among the six different mutant heteroheptamer protein pores, PMPA and CMPA produced current modulation events with the largest differences in the (M113K)3 (M113Y-D8)4 pore in terms of the event mean residence time (1.18 ± 0.17 ms versus 0.17 ± 0.01 ms) and current blockage amplitude (26.8 ± 0.5 pA versus 20.7 ± 0.8 pA) (Table 1). Hence, the (M113K)3 (M113YD8)4 pore was chosen as the stochastic sensing element in the remaining experiments. It should be mentioned that, although the heteroheptameric (M113K)3 (M113Y-D8)4 pore has four possible permutations (Fig. S1, Supporting Information), our results of nine replicate experiments showed that the event signatures (e.g., dwell time and bloackage amplitude) of CMPA/PMPA and the open channel current of the protein pore did not change significantly. This suggests that either the binding of CMPA/PMPA to the host ␤CD molecule would not be significantly affected by the pore permutations, or the (M113K)3 (M113Y-D8)4 pore has only one dominant permutation under our experimental condition. 3.3. Simultaneous detection and quantification of PMPA and CMPA The feasibility of utilizing the (M113K)3 (M113Y-D8)4 pore to simultaneously detect PMPA and CMPA was initially carried out with a mixture of 10 ␮M PMPA and 30 ␮M CMPA at −120 mV in 1 M NaCl (pH 3.0). Although single solutions of PMPA and CMPA produced significantly different events with the mean current blockage amplitudes of 26.8 ± 0.5 pA, and 20.7 ± 0.8 pA, respectively, and having the mean residence time of 1.18 ± 0.17 ms, and 0.17 ± 0.01 ms, respectively, two populations of events for the PMPA and CMPA mixture were not well separated if using the amplitude parameter alone (Fig. 2). One of the major reasons is that both PMPA and CMPA events had a wide range of current blockage amplitudes; further, the event frequency of PMPA was ∼4-fold larger than that of CMPA, thus making CMPA events difficult to be observed. Although we could not rule out the possibility that the wide spread of event amplitudes might be due to the relatively small filter employed in this work so that some of the rapid events were missed under the experimental conditions [note that the rise time (=∼0.33/fc) is ∼66 ␮s at 5 kHz], we are leaning toward the interpretation that the various geometries of the ␤CD and PMPA/CMPA host-guest complexes caused events with different blockage amplitudes. In order to detect PMPA accurately from the mixture, the experimental traces were further filtered at 1 kHz (which was the optimum value to achieve the best PMPA/CMPA differentiation) using a low pass Bessel filter. Since PMPA had a much larger residence time than CMPA, this operation will remove more CMPA events than PMPA ones (∼19.2% reduction versus ∼9.7% reduction in their event counts), thus reducing the interference from CMPA. To further minimize the CMPA interference, only the events with the blockage amplitudes more than 22 pA were included in the data analysis (note that ∼80% of PMPA events had blockage amplitudes larger than 22 pA, while less than 50% of CMPA events had such large amplitudes). To obtain the concentration of PMPA, a dose–response curve for PMPA was obtained using the same approach (Fig. 4). The calculated PMPA concentration in the mixture was 11.9 ± 1.5 ␮M, which was in agreement with the theoretical value (10 ␮M). To obtain the concentration of CMPA from the mixture, the number of events contributed by PMPA was first calculated from its concentration (11.9 ␮M in this particular example) using the PMPA dose–response curve obtained including all the events. For convenience and better clarification, the calibration curve obtained with all the events is called dose–response without cutoff, while that obtained using only the events with the blockage amplitudes more than 22 pA is called dose–response with the amplitude cutoff (Fig. 4a and b). Then, the number of

J. Gupta et al. / Sensors and Actuators B 176 (2013) 625–631

629

Fig. 3. The effect of applied potential on the characteristics of (a) ␤CD and (b) PMPA events. (Left) Event frequency; (middle) residence time; and (right) amplitude. The experiments were performed in 1 M NaCl and 10 mM NaH2 PO4 (pH 3.0) in the presence of 40 ␮M ␤CD. The concentration of PMPA was 10 ␮M. Table 1 Effect of mutant ␣HL pores on PMPA/CMPA differentiation. M113K/M113Y-D8 subunit ratio

PMPM 1:6 2:5 3:4 4:3 5:2 6:1

 off (ms)

Amplitude (pA)

24.5 25.4 26.8 24.8 25.3 23.3

± ± ± ± ± ±

CMPA 1.6 2.2 0.5 0.1 0.6 0.5

events of CMPA was obtained by subtracting the number of events contributed by PMPA from the total events. And then, the concentration of CMPA could be calculated from the dose–response curve of CMPA without amplitude cutoff (Fig. 4c). In this way, the calculated concentration of CMPA in the mixture was 29.9 ± 3.4 ␮M, which is in agreement with its theoretical value (30 ␮M). Two

22.3 21.3 20.7 19.3 19.5 19.2

± ± ± ± ± ±

PMPA 0.6 0.1 0.8 2.7 1.7 0.9

0.43 0.52 1.18 0.31 0.26 0.22

± ± ± ± ± ±

CMPA 0.01 0.02 0.17 0.05 0.01 0.02

0.12 0.17 0.17 0.16 0.13 0.12

± ± ± ± ± ±

0.01 0.03 0.01 0.03 0.01 0.01

additional CMPA and PMPA mixture samples were then examined with the (M113K)3 (M113Y-D8)4 pore, and the results were satisfactory (Table 2). It should be noted that similar to the (M113F/K147N)7 pore which we have reported previously, the (M113K)3 (M113Y-D8)4 pore sensing system is also very selective toward PMPA and

Fig. 4. Dose–response curves for (a) PMPA with amplitude cutoff; (b) PMPA without amplitude cutoff; and (c) CMPA without amplitude cutoff. The experiments were performed in the (M113K)3 (M113Y-D8)4 pore at −120 mV in a solution containing 1 M NaCl and 10 mM NaH2 PO4 (pH 3.0). Data analysis was carried out after filtering the traces at 1 kHz.

630

J. Gupta et al. / Sensors and Actuators B 176 (2013) 625–631

Table 2 Recovery of PMPA and CMPA from mixture samples by use of the nanopore stochastic sensing method. Each experimental value represents the mean of three replicate analyses ± one standard deviation. The experiments were performed at −120 mV with the mutant ␣HL (M113K)3 (M113Y-D8)4 pore in the presence of 40 ␮M ␤CD. Sample number

1 2 3

Theoretical value (␮M)

Experimental value (␮M)

PMPA

CMPA

PMPA

CMPA

10.0 10.0 10.0

10.0 20.0 30.0

11.3 ± 1.6 11.6 ± 2.3 11.9 ± 1.5

10.1 ± 2.7 20.8 ± 3.3 29.9 ± 3.4

CMPA. Many other organophosphates have been tested with this nanopore sensor, including isopropyl methylphosphonic acid, diisopropyl methylphosphonic acid, ethyl methylphosphonic acid, methyl phosphonic acid, ethyl hydrogen dimethylamidophosphate, dimethyl methylphosphonate, diazinon, parathion, malathion and amifostine. However, no current blockage events were observed when the concentrations of these analytes were lower than 500 ␮M. It is not unreasonable that in this nanopore sensor system, CMPA/PMPA is detected because the cyclohexyl/pinacolyl group could be captured by the ␤CD host, while the interaction between other compounds and ␤CD are so weak that they could not be captured by ␤CD for long time enough to produce observable events in our single-channel recording experiments. Given that GD/GF also contains the cyclohexyl/pinacolyl group (note that CMPA and PMPA are hydrolysis products of GD and GF, respectively), it is reasonable to believe that the nanopore sensor described in this work should be able to detect GD and GF. 4. Conclusions In conclusion, by introducing two different functional groups inside a single ␣HL protein pore, rapid differentiation and simultaneous quantitation of organophosphorus nerve agent hydrolytes PMPA and CMPA were successfully achieved. Indeed, this is the first time multi-functionalized ␣HL protein pore as the sensing element has been employed in the nanopore stochastic sensing field. This approach provides new strategies for future nanopore sensor design. It should be noted that since the biological pore used in this work is embedded in a fragile lipid bilayer, our present nanopore sensing system could not easily be transitioned to deployable sensors for the analysis of terrorist agents in real-world samples. Fortunately, this limitation could be potentially overcome by sandwiching the protein pore sensor between two agarose gel layers or using a glass nanopore membrane as the support structure for lipid bilayer membranes [55,56]. Alternatively, the multi-functionalized pore approach could be adopted to produce the artificial-nanopores fabricated in robust solid-state [17]. Acknowledgment This work was financially supported by the Defense Advanced Research Projects Agency (FA9550-06-C-0006). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.snb.2012.10.058. References [1] H. Bayley, P.S. Cremer, Stochastic sensors inspired by biology, Nature 413 (2001) 226–230. [2] L.Q. Gu, O. Braha, S. Conlan, S. Cheley, H. Bayley, Stochastic sensing of organic analytes by a pore-forming protein containing a molecular adapter, Nature 398 (1999) 686–690.

[3] O. Braha, L.Q. Gu, L. Zhou, X. Lu, S. Cheley, H. Bayley, Simultaneous stochastic sensing of divalent metal ions, Nature Biotechnology 18 (2000) 1005–1007. [4] S. Cheley, L.Q. Gu, H. Bayley, Stochastic sensing of nanomolar inositol 1,4,5trisphosphate with an engineered pore, Chemistry and Biology 9 (2002) 829–838. [5] X.F. Kang, S. Cheley, X. Guan, H. Bayley, Stochastic detection of enantiomers, Journal of the American Chemical Society 128 (2006) 10684–10685. [6] C. Gao, S. Ding, Q. Tan, L.Q. Gu, Method of creating a nanopore-terminated probe for single-molecule enantiomer discrimination, Analytical Chemistry 81 (2009) 80–86. [7] S.B. Lee, D.T. Mitchell, L. Trofin, T.K. Nevanen, H. Soderlund, C.R. Martin, Antibody-based bio-nanotube membranes for enantiomeric drug separations, Science 296 (2002) 2198–2200. [8] L. Movileanu, S. Howorka, O. Braha, H. Bayley, Detecting protein analytes that modulate transmembrane movement of a polymer chain within a single protein pore, Nature Biotechnology 18 (2000) 1091–1095. [9] S. Ding, C. Gao, L.Q. Gu, Capturing single molecules of immunoglobulin and ricin with an aptamer-encoded glass nanopore, Analytical Chemistry 81 (2009) 6649–6655. [10] Z. Siwy, L. Trofin, P. Kohli, L.A. Baker, C. Trautmann, C.R. Martin, Protein biosensors based on biofunctionalized conical gold nanotubes, Journal of the American Chemical Society 127 (2005) 5000–5001. [11] A. Liu, Q. Zhao, X. Guan, Stochastic nanopore sensors for the detection of terrorist agents: current status and challenges, Analytica Chimica Acta 675 (2010) 106–115. [12] Q. Zhao, D.A. Jayawardhana, X. Guan, Stochastic study of the effect of ionic strength on non-covalent interactions in protein pores, Biophysical Journal 94 (2008) 1267–1275. [13] T. Luchian, S.H. Shin, H. Hagan, Single-molecule covalent chemistry with spatially separated reactants, Angewandte Chemie: International Edition in English 42 (2003) 3766–3771. [14] S.H. Shin, T. Luchian, S. Cheley, O. Braha, H. Bayley, Kinetics of a reversible covalent-bond-forming reaction observed at the single-molecule level, Angewandte Chemie: International Edition in English 41 (2002) 3707–3709. [15] Q. Zhao, R.S. de Zoysa, D. Wang, D.A. Jayawardhana, X. Guan, Real-time monitoring of peptide cleavage using a nanopore probe, Journal of the American Chemical Society 131 (2009) 6324–6325. [16] J.W. Shim, Q. Tan, L.Q. Gu, Single-molecule detection of folding and unfolding of the G-quadruplex aptamer in a nanopore nanocavity, Nucleic Acids Research 37 (2009) 972–982. [17] D.S. Talaga, J. Li, Single-molecule protein unfolding in solid state nanopores, Journal of American Chemical Society 131 (2009) 9287–9297. [18] J.J. Kasianowicz, E. Brandin, D. Branton, D. Deamer, Characterization of individual polynucleotide molecules using a membrane channel, Proceedings of the National Academy of Sciences of the United States of America 93 (1996) 13770–13773. [19] A. Meller, L. Nivon, E. Brandin, J. Golovchenko, D. Branton, Rapid nanopore discrimination between single polynucleotide molecules, Proceedings of the National Academy of Sciences of the United States of America 97 (2000) 1079–1084. [20] S. Howorka, S. Cheley, H. Bayley, Sequence-specific detection of individual DNA strands using engineered nanopores, Nature Biotechnology 19 (2001) 636–639. [21] J. Li, D. Stein, C. McMullan, D. Branton, M.J. Aziz, J.A. Golovchenko, Ion-beam sculpting at nanometre length scales, Nature 412 (2001) 166–169. [22] A.J. Storm, C. Storm, J. Chen, H. Zandbergen, J.F. Joanny, C. Dekker, Fast DNA translocation through a solid-state nanopore, Nano Letters 5 (2005) 1193–1197. [23] G.V. Soni, A. Meller, Progress toward ultrafast DNA sequencing using solid-state nanopores, Clinical Chemistry 53 (2007) 1996–2001. [24] G. Sigalov, J. Comer, G. Timp, A. Aksimentiev, Detection of DNA sequences using an alternating electric field in a nanopore capacitor, Nano Letters 8 (2008) 56–63. [25] D. Stoddart, A.J. Heron, E. Mikhailova, G. Maglia, H. Bayley, Single-nucleotide discrimination in immobilized DNA oligonucleotides with a biological nanopore, Proceedings of the National Academy of Sciences of the United States of America 106 (2009) 7702–7707. [26] J. Clarke, H. Wu, L. Jayasinghe, A. Patel, S. Reid, H. Bayley, Continuous base identification for single-molecule nanopore DNA sequencing, Nature Nanotechnology 4 (2009) 265–270. [27] T. Jovanovic-Talisman, J. Tetenbaum-Novatt, A.S. McKenney, A. Zilman, R. Peters, M.P. Rout, B.T. Chait, Artificial nanopores that mimic the transport selectivity of the nuclear pore complex, Nature 457 (2009) 1023–1027. [28] T.Z. Butler, J.H. Gundlach, M. Troll, Ionic current blockades from DNA and RNA molecules in the alpha-hemolysin nanopore, Biophysical Journal 93 (2007) 3229–3240. [29] R.S. de Zoysa, D.A. Jayawardhana, Q. Zhao, D. Wang, D. Armstrong, X. Guan, Slowing DNA translocation through nanopores using a solution containing organic salts, Journal of Physical Chemistry B 113 (2009) 13332–13336. [30] Z. Witkiewicz, M. Mazurek, J. Szulc, Chromatographic analysis of chemical warfare agents, Journal of Chromatography 503 (1990) 293–357. [31] E.W. Hooijschuur, C.E. Kientz, U.A. Brinkman, Determination of alkylphosphonic acids by microcolumn liquid chromatography with gradient elution coupled on-line with flame photometric detection, Journal of Chromatography A 907 (2001) 165–172.

J. Gupta et al. / Sensors and Actuators B 176 (2013) 625–631 [32] A.F. Kingery, H.E. Allen, Ion chromatographic separation of closely related nerve agent degradation products using an organic modifier to provide selectivity, Analytical Chemistry 66 (1994) 155–159. [33] J.E. Melanson, B.L. Wong, C.A. Boulet, C.A. Lucy, High-sensitivity determination of the degradation products of chemical warfare agents by capillary electrophoresis-indirect UV absorbance detection, Journal of Chromatography A 920 (2001) 359–365. [34] D. Pardasani, A. Mazumder, A.K. Gupta, P.K. Kanaujia, V. Tak, D.K. Dubey, Determination of hydrolytic degradation products of nerve agents by injection port fluorination in gas chromatography/mass spectrometry for the verification of the Chemical Weapons Convention, Rapid Communications in Mass Spectrometry 21 (2007) 3109–3114. [35] A.L. Simonian, T.A. Good, S.S. Wang, J.R. Wild, Nanoparticle-based optical biosensors for the direct detection of organophosphate chemical warfare agents and pesticides, Analytica Chimica Acta 534 (2005) 69–77. [36] A.L. Jenkis, S.Y. Bae, Molecularly imprinted polymers for chemical agent detection in multiple water matrices, Analytica Chimica Acta 542 (2005) 32–37. [37] K.J. Wallace, J. Morey, V.M. Lynch, E.V. Anslyn, Colorimetric detection of chemical warfare simulants, New Journal of Chemistry 29 (2005) 1469–1474. [38] S.W. Zhang, T.M. Swager, Fluorescent detection of chemical warfare agents: functional group specific ratiometric chemosensors, Journal of the American Chemical Society 125 (2003) 3420–3421. [39] I. Levitsky, S.G. Krivoshlykov, J.W. Grate, Rational design of a Nile Red/polymer composite film for fluorescence sensing of organophosphonate vapors using hydrogen bond acidic polymers, Analytical Chemistry 73 (2001) 3441–3448. ˜ F. Sancenón, A.M. Costero, M. Parra, S. Gil, Chro[40] S. Royo, R. Martínez-Mánez, mogenic and fluorogenic reagents for chemical warfare nerve agents’ detection, Chemical Communications 46 (2007) 4839–4847. [41] T.J. Dale, J. Rebek Jr., Fluorescent sensors for organophosphorus nerve agent mimics, Journal of the American Chemical Society 128 (2006) 4500–4501. [42] G.A. Eiceman, J.A. Stone, Ion mobility spectrometers in national defence, Analytical Chemistry 76 (2004) 390A–397A. [43] D. Wang, Q. Zhao, R.S. de Zoysa, X. Guan, Detection of nerve agent hydrolytes in an engineered nanopore, Sensors and Actuators B: Chemical 139 (2009) 440–446. [44] W.E. Steiner, C.S. Harden, F. Hong, S.J. Klopsch, H.H. Hill Jr., V.M. McHugh, Detection of aqueous phase chemical warfare agent degradation products by negative mode ion mobility time-of-flight mass spectrometry [IM(ToF)MS], Journal of the American Society for Mass Spectrometry 17 (2006) 241–245. [45] F. Worek, P. Eyer, N. Aurbek, L. Szinicz, H. Thiermann, Recent advances in evaluation of oxime efficacy in nerve agent poisoning by in vitro analysis, Toxicology and Applied Pharmacology 219 (2007) 226–234. [46] X. Guan, J. Gupta, Q. Zhao, Nanopore detection of nerve agent degradation products, in: Proceedings of Nanotech Conference Expo 2010, vol. 2, Taylor & Francis, New York, NY, 2010, pp. 99–102. [47] M. Montal, P. Mueller, Formation of bimolecular membranes from lipid monolayers and a study of their electrical properties, Proceedings of the National Academy of Sciences of the United States of America 69 (1972) 3561–3566. [48] R.S. de Zoysa, D.M.M. Krishantha, Q. Zhao, J. Gupta, X. Guan, Translocation of single stranded DNA through the alpha-hemolysin nanopore in acidic solutions, Electrophoresis 32 (2011) 3034–3041. [49] L.Q. Gu, S. Cheley, H. Bayley, Prolonged residence time of a noncovalent molecular adapter, beta-cyclodextrin, within the lumen of mutant alpha-hemolysin pores, Journal of General Physiology 118 (2001) 481–494. [50] L. Movileanu, J.P. Schmittschmitt, J.M. Scholtz, H. Bayley, Interactions of peptides with a protein pore, Biophysical Journal 89 (2005) 1030–1045.

631

[51] S. Yang, A.V. Neimark, Adsorption-driven translocation of polymer chain into nanopores, Journal of Chemical Physics 136 (2012) 214901. [52] D.A. Jayawardhana, J.A. Crank, Q. Zhao, D.W. Armstrong, X. Guan, Nanopore stochastic detection of a liquid explosive component and sensitizers using boromycin and an ionic liquid supporting electrolyte, Analytical Chemistry 81 (2009) 460–464. [53] X. Guan, L.Q. Gu, S. Cheley, O. Braha, H. Bayley, Stochastic sensing of TNT with a genetically engineered pore, ChemBioChem 6 (2005) 1875–1881. [54] A. Banerjee, E. Mikhailova, S. Cheley, L.Q. Gu, M. Montoya, Y. Nagaoka, E. Gouaux, H. Bayley, Molecular bases of cyclodextrin adapter interactions with engineered protein nanopores, Proceedings of the National Academy of Sciences of the United States of America 107 (2010) 8165–8170. [55] X.F. Kang, S. Cheley, A.C. Rice-Ficht, H. Bayley, A storable encapsulated bilayer chip containing a single protein nanopore, Journal of the American Chemical Society 129 (2007) 4701–4705. [56] R.J. White, E.N. Ervin, T. Yang, X. Chen, S. Daniel, P.S. Cremer, H.S. White, Single ion-channel recordings using glass nanopore membranes, Journal of the American Chemical Society 129 (2007) 11766–11775.

Biographies Jyoti Gupta received her BS and MS degrees from Lucknow University of India in 2000, and 2002, respectively. She obtained her PhD degree at the Central Drug Research Institute, Lucknow, India in 2008, and did her postdoctoral research at the University of Texas at Arlington, USA from April 2009 to October 2011. Her research interests include nanopore stochastic sensing and organic synthesis. Qitao Zhao received his doctoral degree from Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, P.R. China, in 2004, and did his postdoctoral research at the University of Texas at Arlington, USA from March 2006 to November 2011. His research activities include preparation and characterization of low-dimensional nanostructures as well as study of their physical chemistry properties, and development of nanopore array technique and nanopore microfluidic device. Guihua Wang received her BS degree in Applied Chemistry in 1996, MS degree in Analytical Chemistry in 1998, and PhD degree in Physical Chemistry in 2007 from University of Science and Technology Beijing. Currently, Dr. Wang is a postdoctoral researcher in the Department of Chemistry and Biochemistry at the University of Texas at Arlington. Her research focuses on the development of nanopore stochastic sensing method for the detection of biomolecules. Xiaofeng Kang is a Professor in the College of Chemistry and Materials Science at Northwest University in P.R. China. He received his PhD in Chemistry at Northwest University in 1999. He is interested in the research of nanosensor and nanobiotechnology. Xiyun Guan is an Associate Professor in the Department of Biological and Chemical Sciences at the Illinois Institute of Technology. He obtained his PhD in Chemistry at the University of Kentucky, Lexington in 2002. Then, he worked as a postdoctoral research associate in the Department of Medical Biochemistry and Genetics at Texas A&M Health Science Center, and an Assistant Professor in the Department of Chemistry and Biochemistry at the University of Texas at Arlington. His current research focuses on the development of nanopore technique for applications in biotechnology at the single molecule level.