S. A. Sarles, L. J. Stiltner, C. B. Williams and D. J. Leo, ACS applied materials & ... S. I. Sukharev, W. J. Sigurdson, C. Kung and F. Sachs, The Journal of General.
Mater. Res. Soc. Symp. Proc. Vol. 1621 © 2014 Materials Research Society DOI: 10.1557/opl.2014.64
Mechanosensitive Channels Activity in a Droplet Interface Bilayer System Joseph Najem1, Myles Dunlap1, Sergei Sukharev2, and Donald J. Leo3 1 Biomolecular Materials and Systems Laboratory, Virginia Tech, Blacksburg, VA 24061, U.S.A. 2 Department of Biology, University of Maryland, College Park, MD 20742, U.S.A. 3 College of Engineering, University of Georgia, Athens, GA 30609, U.S.A. ABSTRACT This paper presents the first attempts to study the large conductance mechano-sensitive channel (MscL) activity in an artificial droplet interface bilayer (DIB) system. A novel and simple technique is developed to characterize the behavior of an artificial lipid bilayer interface containing mechano-sensitive (MS) channels. The experimental setup is assembled on an inverted microscope and consists of two micropipettes filled with PEG-DMA hydrogel and containing Ag/AgCl wires, a cylindrical oil reservoir glued on top of a thin acrylic sheet, and a piezoelectric oscillator actuator. By using this technique, dynamic tension can be applied by oscillating axial motion of one droplet, producing deformation of both droplets and area changes of the DIB interface. The tension in the artificial membrane will cause the MS channels to gate, resulting in an increase in the conductance levels of the membrane. The results show that the MS channels are able to gate under an applied dynamic tension. Moreover, it can be concluded that the response of channel activity to mechanical stimuli is voltage-dependent and highly related to the frequency and amplitude of oscillations. INTRODUCTION Biomolecular unit cells can be described as small building blocks whose repetition can form the basis of a novel biomolecular material system1. The biomolecular unit cell consists of a lipid bilayer interface formed at the contact of two aqueous droplets encased in lipid monolayers. The droplets are surrounded by a hydrophobic organic solvent (Hexadecane), and are sitting on fixed silver-silver chloride (Ag/AgCl) electrodes2, 3. Many types of biomolecules, such as ion channels can self-assemble in the lipid bilayer interface. Therefore, the Droplet Interface Bilayer (DIB) has been extensively used to systematically study the activity of various biomolecules including alamethicin4, bacteriorhodopsin5, and many more. Other types of biomolecules such as mechanosensitive (MS) channels self-assembled within the DIB should be able to respond to an expansion in the artificial membrane. Therefore, it can be used as a model system to understand how the structure of the protein and its incorporation into the unit cell affects its transduction properties. MS channels residing in the cytoplasmic membrane of Escherichia coli respond to a mechanical tension in the cell membrane6, and fall under three categories according to their conductance level7. MscL, the mechanosensitive channel of large conductance, has been studied both in vivo and in vitro using patch-clamp methods8. In this paper, the incorporation and activation of MscL is investigated. These channels, usually found in E-Coli bacteria, respond to a mechanical tension in the cell membrane by either opening or closing. Hence, mechanosensitive channels self-assembled within the lipid bilayer interface should be able to respond to any change in the artificial membrane tension. However, when membrane is under tension, excess lipids existing in the aqueous phase will self-insert in
171
the membrane, leading to tension relaxation. Therefore, the MscL channels incorporated within the lipid bilayer are tested by applying a dynamic oscillation of the membrane rather than static. The application of a dynamic oscillation will result in stretching the lipid bilayer membrane at low frequencies of oscillations, where other vibrational modes will appear at high frequencies9, 10 . This phenomenon is also shown and proved in this paper through a series of mechanoelectrical response experiment of the artificial lipid bilayer membrane. Two different types of MS channels are studied in this paper: (1) MscL, a nonselective MS channel with large conductance (3-nS, in 200 mM KCl), and is activated at relatively high tensions (~14 dynes/cm)6, 11. In addition, (2) MscL V23T mutant, which is a gain-of-function mutant of MscL12. V23T mutant activates at lower tension similar to the MS channel with small conductance (MscS), which is around 1.8 times lower than MscL (~7.8 dynes/cm)12. EXPERIMENTAL METHODS AND MATERIALS The technique described in this report has been developed to characterize the behavior of artificial lipid bilayer interface containing mechanosensitive (MS) channels. Using this technique, dynamic tension can be applied axially to the droplet interface bilayer through squeezing and releasing the droplets. The experimental setup is centered on an inverted microscope (AxioSkop-ZEISS), and consists of two micropipettes filled with PEG-DMA hydrogel, and also contains Ag/AgCl wires, a cylindrical oil reservoir glued on top of a thin acrylic sheet, and finally, a piezoelectric oscillator (Figure 1).
Figure 1: The experimental setup which includes the headstage, the micropipette holder, and the piezoelectric oscillator. The flat tip micropipettes are made of borosilicate glass with 1 mm and 0.5 mm outer and inner diameters respectively. The micropipettes are initially filled with a UV curable hydrogel, and then a silver/silver-chloride (Ag/AgCl) wire is fed into the micropipette through the hydrogel. The hydrogel is then cured through free-radical photopolymerization upon exposure to UV light for 3 min at 1 W intensity, 365 nm UV source. The hydrogel solution with a concentration of 40 % (w/v) PEG-DMA contains 0.5 % (w/v) Irgacure 2959, and is mixed with a 500 mM KCl and
172
10 mM MOPS, pH 7 electrolyte solution. A spot UV source (LED-100, Electro-Lie Corp) is used, as well as Poly(ethylene glycol) dimethacrylate polymer (PEG-DMA; MW=1000 g/mol). Irgacure 2959 is used as a photoinitiator in this study and was obtained from Ciba. The micropipettes are then placed horizontally in opposite directions in a custom-made apparatus. The apparatus is made by gluing an acrylic cylinder on a thin acrylic sheet using epoxy. Two opposing holes (1.1 mm in diameter) are drilled on the wall of the cylinder at 1 cm from the bottom. Two other holes, concentric to the previously drilled holes with a diameter of 4 mm, are also drilled. In order to prevent hexadecane oil from leaking, vacuum grease is deposited around the glass micropipettes. Vacuum grease is used because it is soft; hence no vibration will be transmitted from the oscillator to the oil reservoir and consequently to the micropipette connected to the positive lead. Materials The aqueous phases that form the droplets consist of a suspension of phospholipids vesicles and a buffering agent in highly pure deionized water, while the oil phase consists of Hexadecane (99%, Sigma). The lipid vesicle solution is prepared and stored as described in many articles previously published2, 13. The lipids solution contains 2mg/ml solution of 1,2-diphytanoyl-snglycero-3-phosphocholine (DPHPC, Avanti Polar Lipids, Inc.) vesicles in 500mM potassium chloride (KCl, Sigma), 10mM 3-(N-morpholino)propanesulfonic acid (MOPS, Sigma), pH7. MscL and V23T mutant are received form Dr. Sergei Sukharev’s lab. The received material is initially reconstituted in DPhPC liposomes and then diluted in the DPhPC-lipid vesicle solution to yield a final concentration around 0.2μg/ml. Bilayer formation and Electrical recordings Lipid bilayer interface formed within the biomolecular unit cell is characterized through two types of electrical measurements. Electrically, the lipid bilayer interface is modeled as a capacitor and a resistor in parallel. Therefore, capacitance measurements are carried out in order to verify the increase in capacitance resulting from the bilayer formation. Axopatch 200B and Digidata 1440A (Molecular Devices) are used to measure the resulting square-wave current produced by an external, 10 mV triangular voltage waveform at 10 Hz (Hewlett Packard 3314A function generator). The second type of electrical recording is a current measurement of the bilayer interface, which is held under voltage-clamp while mechanically oscillating the bilayer containing the MS channels. All electrical recordings are carried out under a lab-made Faraday cage that serves as an electrical shield. RESULTS AND DISCUSSION Droplet interface bilayer mechanoelectrical response The DIB mechanoelectrical response was observed at a variety of frequencies by using a linear sinusoidal-sweep (chirp signal) from the piezoelectric actuator, as well as single frequency sinusoidal oscillations. Figure 2(a) shows the response of the bilayer before and after the application of the mechanical oscillations while an external 10 Hz triangular voltage signal is applied. When applying the mechanical excitation, the current response exhibits a change in the
173
amplitude as a result of the change in capacitance, which reflects the zipping and unzipping of the bilayer. This observation is clearly portrayed in Figure 2(b), where no external voltage signal is applied and the current variation is only related to the change in capacitance (dC/dt). This experiment served not only to understand the mechanoelectrical response of the DIB, but also as a baseline for the mechanosensitive channels’ experiments. A DC potential with amplitudes ranging from -150 mV to 150 mV is applied, resulting a highlight two major observations: (1) the magnitude of the current does increase proportionally with the applied voltage, and (2) no traces or any gating-like event happened as shown in Figure 2(b).
Figure 2: (a) Bilayer formation and current amplitude change due to membrane oscillation (No protein), (b) Current response of the bilayer (No protein) due to a change in bilayer are resulting from membrane oscillation. Effect of protein concentration on the bilayer formation When reconstituting MscL and V23T mutant in the liposome, it is noticed that the bilayer formation is inhibited as shown in Figure 3(a). Specifically, the bilayer containing MscL when “formed” did not exhibit a significant increase in capacitive current (~15 pA peak-to-peak). The current response displays a triangular shape waveform which reflects a leaky bilayer (i.e. low resistance bilayer). In the course of trying to understand what is leading to such a behavior, it is found that the protein to lipid ratio in our solutions is at least three orders of magnitudes higher than the values reported in the literature14. Therefore, the protein to lipid ratio is reduced by diluting the solutions with DPhPC lipid solution. The formed bilayer seemed more consistent with previous bilayer formation results when no MscL protein existed in the bilayer. The resulting capacitive current (~ 200 Pa) significantly increased upon thinning of the bilayer, and exhibits a square-shape like waveform, which reflects a high resistance bilayer (Figure 3(b)). Note that in this paper the optimum protein concentration is not explored. The concentration is reduced in a manner just to be able to form a functional bilayer to perform the experiments. Mechanosensitive channels activation The diluted solutions containing MscL V23T mutant are used to form artificial lipid bilayer membranes. The membranes are then mechanically oscillated at different frequencies ranging from 0.5 Hz to 75 Hz using sinusoidal and step waveforms. Simultaneously, a DC potential
174
ranging from 0 to 100 mV is applied. The results show that sub-conductance gating events are taking place at the peak of the mechanical oscillation, which corresponds to the larger bilayer formed. These events occurred when a DC potential higher than 80 mV is applied. Three important observations are made from Figures 4 (a) and (b): (1) The gating events happened at the maximum compression of the bilayer. This may be an indication that the membrane is mechanically stretched, leading to channel gating. (2) It is shown that the gating frequency and conductance levels of the V23T mutant are related to the applied DC potential across the membrane. Figure 4 (b) shows what is considered full gating (around second 75) when a potential of 100 mV, which is higher than the one applied in Figure 4(a). Moreover, some subconductance gating event were seen even when the mechanical oscillations are turned off and a DC potential is applied, which emphasizes the role of applied potential on the gating of MS channels. (3) While running the experiments, the importance of the frequency of oscillations was obvious. Mainly, the gating events happened at lower oscillations frequencies and disappeared while the frequency is increased. As a result, it is believed that most of the events seen are sub-conductance states of the V23T mutant.
Figure 3: (a) Bilayer formation inhibited due to high protein concentration (MscL V23T mutant), (b) Bilayer formed normally after reducing the protein (MscL V23T mutant) concentration.
Figure 4: The results presented in the these plots refer to the MscL V23T mutant protein (a) Bilayer current response when a 0.2 Hz sinusoidal waveform in applied along with a 90 mV DC
175
potential, (b) Bilayer current response when a 0.2 Hz sinusoidal waveform in applied along with a 100 mV DC potential. CONCLUSIONS The results presented in this paper highlight the first gating events of MS channels in the DIB system. The MS channels incorporated in the lipid bilayer responded to an applied mechnical oscillation by opening and increasing the conductance level of the membrane. All experiments are conducted using a novel experimental setup that features two aqueous droplets placed at the tip of two horizontally opposing hydrogel filled micropippettes. As expected, it was revealed that the mechanical tension has a major contribution to the gating of the MS channels. In the end, it can be concluded that the response of channel activity to mechanical stimuli is voltage-dependent and is highly related to the frequency and amplitude of oscillations. ACKNOWLEDGMENTS The authors gratefully acknowledge financial support through the Air Force Office of Scientific Research Basic Research Initiative grant number 11157642. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
S. A. Sarles, L. J. Stiltner, C. B. Williams and D. J. Leo, ACS applied materials & interfaces 2 (12), 3654-3663 (2010). S. A. Sarles and D. J. Leo, Analytical chemistry 82 (3), 959-966 (2010). S. A. Sarles, Virginia Polytechnic Institute and State University, 2010. S. A. Sarles and D. J. Leo, Journal of intelligent material systems and structures 20 (10), 1233-1247 (2009). W. L. Hwang, M. Chen, B. Cronin, M. A. Holden and H. Bayley, Journal of the American Chemical Society 130 (18), 5878-5879 (2008). S. I. Sukharev, B. Martinac, V. Y. Arshavsky and C. Kung, Biophysical Journal 65 (1), 177-183 (1993). P. Blount, S. I. Sukharev, P. C. Moe, B. Martinac and C. Kung, in Methods in Enzymology, edited by P. M. Conn (Academic Press, 1999), Vol. Volume 294, pp. 458482. S. I. Sukharev, B. Martinac, P. Blount and C. Kung, Methods 6 (1), 51-59 (1994). A. G. Petrov, Analytica Chimica Acta 568 (1–2), 70-83 (2006). S. A. Sarles, J. D. Madden and D. J. Leo, Soft Matter 7 (10), 4644-4653 (2011). S. I. Sukharev, W. J. Sigurdson, C. Kung and F. Sachs, The Journal of General Physiology 113 (4), 525-540 (1999). B. Akitake, A. Anishkin and S. Sukharev, The Journal of General Physiology 125 (2), 143-154 (2005). G. Szabo, M. C. Gray and E. L. Hewlett, Journal of Biological Chemistry 269 (36), 22496-22499 (1994). N. Saint, J.-J. Lacapère, L.-Q. Gu, A. Ghazi, B. Martinac and J.-L. Rigaud, Journal of Biological Chemistry 273 (24), 14667-14670 (1998).
176
