Polyelectrolyte Entry and Transport through an Asymmetric r-Hemolysin Channel

Gabriel Gibrat,† Manuela Pastoriza-Gallego,† Benedicte Thiebot,‡ Marie-France Breton,‡Loïc Auvray,† and Juan Pelta*,†,‡

Institut de Chimie des Materiaux Paris-Est, Equipe Materiaux Polymeres aux Interfaces, CNRS-UMR 7182,UniVersite d’EVry, Bd F. Mitterrand, 91025 EVry France, and Groupe Micro-enVironnement et ComportementsCellulaires, UniVersite de Cergy-Pontoise, 2 aVenue A. ChauVin, 95302 Cergy-Pontoise Cedex France

ReceiVed: September 11, 2008; ReVised Manuscript ReceiVed: October 17, 2008

We study the entry and transport of a polyelectrolyte, dextran sulfate (DS), through an asymmetric R-hemolysinprotein channel inserted into a planar lipid bilayer. We compare the dynamics of the DS chains as they enterthe channel at the opposite stem or vestibule sides. Experiments are performed at the single-molecule levelby using an electrical method. The frequency of current blockades varies exponentially as a function of appliedvoltage. This frequency is smaller for the stem entrance than for the vestibule one, due to a smaller couplingwith the electric field and a larger activation energy for entry. The value of the activation energy is quantitativelyinterpreted as an entropic effect of chain confinement. The translocation time decreases when the appliedvoltage increases and displays an exponential variation which is independent of the stem or vestibule sides.

Introduction

Lipid membranes act as physical barriers that define the celland compartmentalize biological functions; therefore, the trans-port of macromolecules through channels inserted in themembranes plays an important role in many biological pro-cesses. The electric method is a powerful tool to investigatethe mechanism of molecule transport across nanopores.1 Anelectrical potential difference is applied to a lipid membrane,inducing a current throughout a single pore. A macromoleculepassing through a nanopore induces an abrupt decrease of theelectrical current, which depends on the pore geometry, on theionic strength but also on the length, chemical nature, andconformation of the chain. Biological channels such as theR-hemolysin2 and the aerolysin,3 and a solid state pore,4,5 havebeen used as nanopores.6 Experiments with protein pores,7-16

simulations,17-19 and theoretical studies20-29 have been per-formed (i) to understand the dynamics of confined neutralpolymers7,30 and polyelectrolyte chains8,13,15 or (ii) to study thetransport of peptides16,31 or unfolded proteins14 and theirinteractions with pores.11,12 The electrical detection of themacromolecule transport through a single protein channel hasalso been used in many applications.32-37 The R-hemolysin poreis geometrically38 and electrically asymmetric.39 The pore lengthis around 10 nm, and the opening on the vestibule and stemsides are different, with respective diameters of 2.6 and 2 nm.38

R-Hemolysin exhibits a rather strong ion current rectification.39

This channel is very stable as a function of environment.40-42

We have recently shown that we can control the electricalasymmetry by selectively denaturing the cap domain of theR-hemolysin.42 Up to now, many experimental and theoreticalworks have been devoted to studying the dependence of the

translocation time as a function of the polymer molecularweight,8,10,12,13,18,22,24 but only a few experiments have inves-tigated the way charged polymers enter a pore and the effect ofa pore asymmetry on the entry and transport of chargedmolecules.9,43 As shown in ref 9 and further discussed in ref44, single-stranded DNA enters the R-hemolysin pore morereadily through the vestibule entrance. The effective numberof negative charges coupled to the electric field at the poreentrance (stem or vestibule side) was shown to be smaller thanthe total number of negative charges on the polyelectrolyte.9

Another experiment on the asymmetric maltoporine channel hasalso shown that the maltodextrine entry is easier on the vestibuleside and that the translocation time does not depend on the sidethe sugar is added.43

In the present work, we investigated the dynamics of entryand transport of a polyelectrolyte, dextran sulfate, through thedifferent sides of the R-hemolysin channel as a function ofapplied voltage. The word translocation is defined as the passageof the molecule across a nanopore from one chamber to theother. Data on the vestibule side have been obtained previously.8

Here, we compare the frequency of current blockades, theassociated activation energy, the blockade times, the couplingof the polyelectrolyte with the electric field at the pore entranceand inside the pore, and also the level of normalized currentpore blockade, when the molecules enter by the stem or by thevestibule sides. Our results show a decrease of the efficiencyof entry of molecules at the stem domain of R-hemolysin andan increase of the activation energy. The long blockade timesor translocation times, the effective charge inside the pore, andthe current pore blockade are similar when the polyelectrolyteenters the pore through the wide vestibule, or through the narrowstem.

Experimental Methods

Bilayers and Pore. Lipid bilayers are made using a previ-ously described method.10,45 In brief, a film of a 1% solution of

* To whom correspondence should be addressed. E-mail: juan.pelta@bio.u-cergy.fr.

† Universite d’Evry.‡ Universite de Cergy-Pontoise.

14687

10.1021/jp808088y CCC: $40.75 2008 American Chemical Society

Published on Web 10/31/2008

2008, 112, 14687–14691

diphytanoylphosphatidylcholine-lecithine (Avanti) in decane isspread across a 150 µm wide hole, drilled in a polysulfone wallseparating the two compartments of the cell. Each compartmentcontains a volume of 1 mL of a solution of 1 M KCl, 5 mMHEPES pH 7.4. After thinning of the decane film and formationof a planar bilayer, the channels are inserted by adding 0.30nmol of monomeric R-hemolysin (Sigma) from a stock solutionin the trans compartment. All experiments have been carriedout with R-hemolysin orientated with the vestibule on the transside of the lipid bilayer (see Figure 1a).

Dextran Sulfate. The dextran sulfate sodium (DS) used hasa molecular weight of 8 kDa (Aldrich) corresponding to 31monomers, and the polydispersity is 1.35.46 One monomer bearstwo negative charges. The final concentration used is 0.5% (w/v) in 1 M KCl, 5 mM HEPES pH 7.4. The dextran sulfatesodium is added to the cis and trans compartment, and themolecules enter through the stem side (see Figure 1a).

Data Acquisition. The ionic current through one R-hemolysinchannel is measured with an Axopatch 200B amplifier. Dataare filtered at 10 KHz and acquired at 4 µs intervals with theDigiData 1322A digitizer coupled with Clampex software (AxonInstruments, Union City, CA). The measurements of thetransients are based on the statistical analysis of the currenttraces previously described.15 All of the statistical analysis ofthe data is performed with the software Igor Pro (WaveMetricsInc.).

Results and Discussion

Frequency of Blockade Events and Effective Charge ofthe Polylectrolyte at the Entrance of the Pore. The asymmetryof a channel is classically demonstrated by the variation of thecurrent as a function of applied voltage. In the case ofR-hemolysin, the conductance is higher at negative appliedvoltage than at positive one, with a nonlinear transition aroundzero (Figure 1a), the vestibule domain of the pore being orientedtoward the trans compartment (Figure 1a). The conductance is691 ( 16 pS in accordance with classical measurements.39

Single-channel current traces are obtained between 100 and 150mV in the presence of DS (Figure 1b-e) entering through thestem side. When the applied voltage (V) increases, the meancurrent and the number of current blockades increase (Figure1b-d). A zoom of a current trace, at V ) +130 mV, shows thetime intervals Ti between two blockade events, the time durationof event Tt, the current of the empty pore I0, the current poreblockade IB, and the thresholds (Th1, Th2, Figure 1e). The meantransient frequencies at V ) +105 mV (Figure 2a) and at V )+130 mV (Figure 2b) are, respectively, f ) 0.67 ( 0.5 Hz andf ) 3.3 ( 0.6 Hz. The frequency of blockades f is plotted as afunction of the applied voltage (Figure 2c). As explained inrefs 9 and , the polymer translocation driven by an electric fieldis described as an activated process involving free energybarriers of entropic and electrostatic origin. Data are welldescribed by the following equation: f ) f0 exp (|V|/V0), wheref0 ) pν exp(-U(/kBT)23 is the frequency in the absence ofapplied voltage, p is a probability factor, ν a frequency factor,U( the activation energy, kBT the thermal agitation, and V0 iswritten as V0 ) kBT/ze, where z is defined as the effective chargeof the polyelectrolyte on which the electric field acts at the poreentrance and e is the elementary charge. We have found f0 )0.003 Hz, V0 ) 18.9 ( 0.14 mV, and zstem ) 1.3 ( 0.01 (Figure2c).

It is interesting to compare these observations with ourprevious data (Table 1), concerning DS entry through the

Figure 1. Current-voltage curve of an R-hemolysin pore (a) andsingle-channel current traces of dextran sulfate through the protein poreinserted into the lipid bilayer as a function of applied voltage, V (b, c,d, e). The vestibule of the R-hemolysin pore is orientated on the transside. The current level is asymmetric as a function of positive versusnegative applied voltage (a). Current traces when the molecules areentering by the stem side, for V ) 100 mV (b), V ) 130 mV (c), andV ) 150 mV (d). Detail of a current trace (V ) +130 mV) showingblockade events, time interval between two blockade events Ti, blockadeduration Tt, the ionic current in the empty pore I0, the blockade currentIB, a first threshold Th1, and a second threshold Th2 (e). Experimentsare made at 1 M KCl, 5 mM HEPES pH 7.4, dextran sulfateconcentration 0.5%. The applied voltage is between -125 and 125 mV(a).

Figure 2. Frequency of blockade events versus applied voltage.Semilog scale (a-d). Explanation of the statistical analysis of themeasured current traces, distribution of time intervals Ti for V ) +105mV (a) and V ) +130 mV (b); continuous lines are exponential fits.Frequency of events as a function of applied voltage when the moleculesare entering by the stem side of the pore (c); the line is an exponentialfit: f ) f0 exp(V/V0), with f0 ) 0.003 Hz and V0 ) 18.9 ( 0.14 mV (c).Comparison of polyelectrolyte event frequency when the moleculesare entering by the stem or by the vestibule8 sides of the channel (d).Reprinted from ref 8 with permission. Copyright 2008 AmericanPhysical Society.

14688 J. Phys. Chem. B, Vol. 112, No. 47, 2008 Letters

vestibule side.8 The mean blockade frequencies are higher inthis case (Figure 2d) and are also well described by anexponential dependence on the electrical potential. The fre-quency in the absence of applied voltage is higher, f0 ) 0.5Hz, and the effective charge is zvestibule ) 0.95 ( 0.01. We indeedexpect the electric field to be lower in the wide vestibule regionthan in the narrow stem side. This hypothesis could interpretthe smaller coupling of an entering dextran sulfate moleculewith the field on the vestibule side. A quantitative description,which could probably only be obtained by numerical simula-tions, is lacking.

Activation Energy. In order to evaluate the activation energyassociated with the entry process, the frequency factor (ν) isestimated by a barrier penetration calculation:9 ν ) CDA/L,where C is the bulk concentration of the polyelectrolyte, D isits diffusion coefficient, A is the cross-sectional area of thechannel, and L is the pore length. With the values of ref 9, aDS concentration of 0.5%, and a molecular weight of 8000 Da,we obtain νstem ≈ 11 250 Hz and Ustem

( ≈ 15kBT, to becompared with the value Uvestibule

( ≈ 10kBT8 (Table 1) obtainedat the vestibule entrance. When the radius of a flexible chain islarger than the radius of a pore, the translocation of electricallycharged macromolecules at high ionic strength is mainlycontrolled by an entropic barrier.47 Our experiments showdirectly that the narrowness of the channel at the stem sidereduces the conformational entropy of the chain and the chainfree energy increases with respect to the vestibule side. If theentropic barrier is dominated by the confinement of the chains,we expect the ratio Ustem/Uvestibule to vary as (Dvestibule/Dstem)5/3,where Dstem and Dvestibule are, respectively, the diameter of thestem and vestibule entrance.38 Assuming Dstem ) 2 nm andDvestibule ) 2.6 nm, the value (Dvestibule/Dstem)5/3 ) 1.55 is to becompared to Ustem/Uvestibule ) 1.5 and it is in good agreement.

Transport and Effective Charge of the Polyelectrolyteinside the Pore. We have analyzed the duration of the blockadesas a function of applied voltage (Figure 3). For molecules goingacross the stem side, two kinds of events with different durationsare observed. See the examples at V ) +105 mV (Figure 3a)and V ) +130 mV (Figure 3b); the mean blockade times are,respectively, tshort ) 44 ( 6.5 µs and tlong ) 158 ( 31 µs (Figure3a) and tshort ) 23 ( 5 µs and tlong ) 92 ( 17 µs (Figure 3b).The population of short events is larger than that of long events(Figure 3a and b). The short (Figure 3c) and long (Figure 3d)mean blockade durations decrease when the applied voltageincreases. The dependence is well-described by an exponentialfunction, f(V) ) A exp(V/Vc) (Figure 3c and d), where A ) 350( 35 µs and Vc ) 48 ( 1.5 mV for the short blockades and A) 950 ( 50 µs and Vc ) 58 ( 1.3 mV for the long ones. Thevalues previously obtained for entry through the vestibule sidewere, respectively, A ) 195 ( 33.5 µs and Vc ) 66 ( 1.5 mVand A ) 1035 ( 109 µs and Vc ) 58 ( 3.6 mV.8 For the longblockade times, these values are identical (Figure 3f) whateverthe entry side, but they differ for the short blockades (Figure3e)8 (Table 1). Assuming, as already discussed in ref 8, thatthe long blockades are associated to translocation events, thesefacts directly show that the dynamics of the polyelectrolytechains inside the pore is independent of the entry side even if

the pore is asymmetric. The coupling with the electric fieldinside the pore is measured by the effective charge zpore, definedas zpore ) kBT/Vce and deduced from the voltage dependence ofthe long blockade time. Here, we measure zpore ) 0.44 ( 0.01for the entries at the stem compared to 0.44 ( 0.03 for theentries by the vestibule.8 Thus, the effective charge of thepolyelectrolyte inside the pore is independent of the side of entryas it should be (Table 1). The low value of this effective chargeis not fully explained, and it could be associated with anincreased condensation of the counterions due to the confinementof the charges in the medium of low dielectric constant.29

A diagram of event blockade duration versus current poreblockade for each applied voltage is represented in Figure 4.Each polyelectrolyte chain passing through the pore is charac-terized by the duration of the blockade and by a value of acurrent pore blockade. For all applied voltages, the short eventsare associated with low current pore blockade and the long

TABLE 1: Dextran Sulfate Transport Parameters through an r-Hemolysin Porea

entrance side f0 (Hz) V0 (mV) z at the entrance U ( A1 (µs) Vc1 (mV) A2 (µs) Vc2 (mV) z inside the pore

vestibule 0.5 27.1 ( 0.03 0.95 ( 0.01 10kBT 195 ( 33.5 66 ( 1.5 1035 ( 109 58 ( 3.6 0.44 ( 0.03stem 0.003 18.9 ( 0.14 1.3 ( 0.01 15kBT 350 ( 35 48 ( 1.5 950 ( 50 58 ( 1.3 0.44 ( 0.01

a A1, Vc1: values for short time or straddling time. A2, Vc2: values for long time or translocation time.

Figure 3. Duration of events versus applied voltage. Semilog scale(a-f). Explanation of the statistical analysis of the measured currenttraces, distribution of blockade time Tt for V ) +105 mV (a) and V )+130 mV (b); continuous lines are single exponential fits, and dottedlines correspond to a double exponential fit. Duration of events as afunction of applied voltage: f(V) ) A exp(V/V0), we found for the shorttime A ) 350 ( 25 µs and Vc ) 48 ( 1.5 mV (c) and for the longtime A ) 950 ( 50 µs and Vc ) 58 ( 1.3 mV (d). Comparison ofpolyelectrolyte short (e) and long (f) blockade times when the moleculesare entering by the stem side of the channel (full symbols) or by thevestibule8 side (empty symbols). Reprinted from ref 8 with permission.Copyright 2008 American Physical Society.

Letters J. Phys. Chem. B, Vol. 112, No. 47, 2008 14689

events are correlated to high current pore blockade. When theelectrical potential difference increases, the duration of the longand short events is translated toward a lower value, and thelevel of blockage of the short events diminishes, while the oneof the long events remains constant (Figure 4). The normalizedaverage value of the current is then 66 ( 3.2% (Figure 5). Thevalue of this current pore blockade is discussed in ref 29, andit takes into account the effect of the electrostatic interactions,the screening and confinement of the electric field in the pore.We observe in our experiments that the electric field does notinfluence the conduction process of ions inside the pore in thepresence of a confined DS chain.

Conclusions

We have clearly shown that the dynamics of entry of apolyelectrolyte inside a channel of nanometer size is verysensitive to the asymmetry of the channel and dominated inthe investigated case by the entropic effect associated with theconfinement of the chains. An electrostatic effect is alsoobserved, since the coupling with the electric field at the stemor vestibule entrance is different. On the contrary, the dynamics

of a charged macromolecule inside the channel is independentof the channel asymmetry.

Acknowledgment. This work is supported by a grantfunding, Action Thematique Incitative Genopole and “pro-gramme interdisciplinaire du CNRS, Interface PhysiqueChimie Biologie”.

References and Notes

(1) Zimmerberg, J.; Parsegian, V. A. Nature 1986, 323, 36–39.(2) Song, L.; Hobaugh, M. R.; Shustak, C.; Cheley, S.; Bayley, H.;

Gouaux, J. E. Science 1996, 274, 1859–1866.(3) Parker, M. W.; Buckley, J. T.; Postma, J. P.; Tucker, A. D.; Leonard,

K.; Pattus, F.; Tsernoglou, D. Nature 1994, 367, 292–295.(4) Martin, C. R.; Siwy, Z. S. Science 2007, 317, 331–332.(5) Smeets, R. M.; Keyser, U. F.; Dekker, N. H.; Dekker, C. Proc.

Natl. Acad. Sci. U.S.A. 2008, 105, 417–421.(6) Kasianowicz, J. J.; Robertson, J. W. F.; Chan, E. R.; Reiner, J. E.;

Stanford, V. M. Annu. ReV. Anal. Chem. 2008, 1, 737–766.(7) Bezrukov, S. M.; Vodyanoy, I.; Brutyan, R. A.; Kasianowicz, J. J.

Macromolecules 1996, 29, 8517–8522.(8) Brun, L.; Pastoriza-Gallego, M.; Oukhaled, G.; Mathe, J.; Bacri,

L.; Auvray, L.; Pelta, J. Phys. ReV. Lett. 2008, 100, 158302–158304.(9) Henrickson, S. E.; Misakian, M.; Robertson, B.; Kasianowicz, J. J.

Phys. ReV. Lett. 2000, 85, 3057–3060.(10) Kasianowicz, J. J.; Brandin, E.; Branton, D.; Deamer, D. W. Proc.

Natl. Acad. Sci. U.S.A 1996, 93, 13770–13773.(11) Mohammad, M. M.; Prakash, S.; Matouschek, A.; Movileanu, L.

J. Am. Chem. Soc. 2008, 130, 4081–4088.(12) Movileanu, L.; Schmittschmitt, J. P.; Scholtz, J. M.; Bayley, H.

Biophys. J. 2005, 89, 1030–1045.(13) Murphy, R. J.; Muthukumar, M. J. Chem. Phys. 2007, 126, 051101.(14) Oukhaled, G.; Mathe, J.; Biance, A. L.; Bacri, L.; Betton, J. M.;

Lairez, D.; Pelta, J.; Auvray, L. Phys. ReV. Lett. 2007, 98, 158101–158104.(15) Oukhaled, G.; Bacri, L.; Mathe, J.; Pelta, J.; Auvray, L. Europhys.

Lett. 2008, 82, 48003.(16) Stefureac, R.; Long, Y. T.; Kraatz, H. B.; Howard, P.; Lee, J. S.

Biochemistry 2006, 45, 9172–9179.(17) Huang, L.; Kirmizialtin, S.; Makarov, D. E. J. Chem. Phys. 2005,

123, 124903.(18) Matysiak, S.; Montesi, A.; Pasquali, M.; Kolomeisky, A. B.;

Clementi, C. Phys. ReV. Lett. 2006, 96, 118103.(19) Kong, C. Y.; Muthukumar, M. J. Am. Chem. Soc. 2005, 127, 18252–

18261.(20) Berezhkovskii, A. M.; Hummer, G.; Bezrukov, S. M. Phys. ReV.

Lett. 2006, 97, 020601.(21) de Gennes, P. G. AdV. Polym. Sci. 1999, 138, 91.(22) Flomenbom, O.; Klafter, J. Phys. ReV. E 2003, 68, 041910.(23) Hanggi, P.; Talkner, P.; Borkovec, M. ReV. Mod. Phys. 1990, 62,

251.(24) Kantor, Y.; Kardar, M. Phys. ReV. E 2004, 69, 021806.(25) Mohan, A.; Kolomeisky, A. B.; Pasquali, M. J. Chem. Phys. 2008,

128, 125104.(26) Reguera, D.; Schmid, G.; Burada, P. S.; Rubi, J. M.; Reimann, P.;

Hanggi, P. Phys. ReV. Lett. 2006, 96, 130603.(27) Slonkina, E.; Kolomeisky, A. B. J. Chem. Phys. 2003, 118, 7112–

7118.(28) Sung, W.; Park, P. J. Phys. ReV. Lett. 1996, 77, 783–786.(29) Zhang, J.; Shklovskii, B. I. Phys. ReV. E 2007, 75, 021906.(30) Movileanu, L.; Cheley, S.; Bayley, H. Biophys. J. 2003, 85, 897–

910.(31) Wolfe, A. J.; Mohammad, M. M.; Cheley, S.; Bayley, H.;

Movileanu, L. J. Am. Chem. Soc. 2007, 129, 14034–14041.(32) Akeson, M.; Branton, D.; Kasianowicz, J. J.; Brandin, E.; Deamer,

D. W. Biophys. J. 1999, 77, 3227–3233.(33) Bayley, H.; Cremer, P. S. Nature 2001, 413, 226–230.(34) Benner, S.; Chen, R. J. A.; Wilson, N. A.; Abu-Shumays, R.; Hurt,

N.; Lieberman, K. R.; Deamer, D. W.; Dunbar, W. B.; Akeson, M. Nat.Nanotechnol. 2007, 2, 718–724.

(35) Kang, X. F.; Cheley, S.; Rice-Ficht, A. C.; Bayley, H. J. Am. Chem.Soc. 2007, 129, 4701–4705.

(36) Mathe, J.; Visram, H.; Viasnoff, V.; Rabin, Y.; Meller, A. Biophys.J. 2004, 87, 3205–3212.

(37) Sauer-Budge, A. F.; Nyamwanda, J. A.; Lubensky, D. K.; Branton,D. Phys. ReV. Lett. 2003, 90, 238101.

(38) Gouaux, E. J. Struct. Biol. 1998, 121, 110–122.(39) Menestrina, G. J. Membr. Biol. 1986, 90, 177–190.(40) Bezrukov, S. M.; Kasianowicz, J. J. Eur. Biophys. J 1997, 26, 471–

476.

Figure 4. Scatter plots of blockade duration versus current poreblockade, [(1 - (⟨IB⟩/⟨I0⟩)] × 100, for dextran sulfate as a function ofapplied voltage. Examples of scatter plots for V ) +105 mV, V )+120 mV, V ) +130 mV, and V ) +140 mV.

Figure 5. The current pore blockade, [(1 - (⟨IB⟩/⟨I0⟩)] × 100, as afunction of applied voltage for the long blockade events. The linecorresponds to the average value (1 - (⟨IB⟩/⟨I0⟩)) ) b, with b ) 66 (3.2%.

14690 J. Phys. Chem. B, Vol. 112, No. 47, 2008 Letters

(41) Jung, Y.; Bayley, H.; Movileanu, L. J. Am. Chem. Soc. 2006, 128,15332–15340.

(42) Pastoriza-Gallego, M.; Oukhaled, G.; Mathe, J.; Thiebot, B.; Betton,J. M.; Auvray, L.; Pelta, J. FEBS Lett. 2007, 581, 3371–3376.

(43) Kullman, L.; Winterhalter, M.; Bezrukov, S. M. Biophys. J. 2002,82, 803–812.

(44) Ambjornsson, T.; Apell, S. P.; Konkoli, Z.; Di Marzio, E. A.;Kasianowicz, J. J. J. Chem. Phys. 2002, 117, 4063–4073.

(45) Montal, M.; Mueller, P. Proc. Natl. Acad. Sci. U.S.A. 1972, 69,3561–3566.

(46) Takeuchi, T.; Safni Safni, Miwa T.; Hashimoto, Y.; Moriyama, H.Chromatographia 1999, 50, 70–74.

(47) Muthukumar, M. Annu. ReV. Biophys. Biomol. Struct. 2007, 36,435–450.

JP808088Y

Letters J. Phys. Chem. B, Vol. 112, No. 47, 2008 14691