- Email: [email protected]
Electroformation and electrofusion of giant vesicles in a microfluidic device
Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
Contents lists available at SciVerse ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Electroformation and electrofusion of giant vesicles in a microfluidic device Zhenyu Wang a , Ning Hu a,b,∗ , Li-Hsien Yeh c , Xiaolin Zheng a , Jun Yang a,∗∗ , Sang W. Joo b,∗ ∗ ∗ , Shizhi Qian b,d,∗ ∗ ∗∗ a Key Laboratory of Biorheological Science and Technology (Chongqing University), Ministry of Education, and Key Lab of Visual Damage and Regeneration & Restoration of Chongqing, Chongqing, Bioengineering College, Chongqing University, Chongqing 400030, China b School of Mechanical Engineering, Yeungnam University, Gyongsan 712-749, South Korea c Department of Chemical & Materials Engineering, National Yunlin University of Science and Technology, Yunlin 64002, Taiwan d Institute of Micro/Nanotechnology, Old Dominion University, Norfolk, VA 23529, USA
a r t i c l e
i n f o
Article history: Received 10 January 2013 Received in revised form 10 April 2013 Accepted 16 April 2013 Available online 30 April 2013 Keywords: Giant vesicle Electroformation Electrofusion Drug delivery Microfluidics
a b s t r a c t Electroformation and electrofusion of giant vesicles with diameters of 10–20 m have been performed in a microfluidic device with high-density microelectrodes forming the sidewalls of the microchannel. Electroformation of giant vesicles by a solution mixture of phosphatidylcholine (PC) and cholesterol (Chol) with different concentrations under AC electric field was investigated. Under the conditions of 0.5–12 mg/mL PC and 0.1–2.4 mg/mL Chol, vesicles were electroformed by the AC electric field imposed. About 60% electroformed vesicles were giant (unilamellar) vesicles with diameters 10–20 m. The eletroformed vesicles were collected from the chip, re-suspended in fresh buffer, and then separated by centrifugation to segregate the ones with desired diameters (10–20 m). Electrofusion of the giant vesicles was conducted in the same chip. Vesicles were aligned to form pairs under AC electric field due to positive dielectrophoresis, and the paired vesicles were subsequently fused upon the application of high strength electrical pulses. The alignment and fusion efficiencies were, respectively, about 50% and 20%. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Giant lipid membrane vesicles (giant vesicles) with diameters larger than 10 m have obtained widespread attentions for their potential applications, including drug delivery [1–4], microreactors [5–8], and modeling cytomembrane systems [9]. Since the electroformation method of giant vesicles was first developed by Angelova and Dimitrov [10] in 1986, a growing number of devices have been implemented to improve this technique [11–24]. Owing to relatively easy fabrication, chemically inert nature to various organic solvents, and adjustable distance between two
∗ Corresponding author at: Key Laboratory of Biorheological Science and Technology (Chongqing University), Ministry of Education, and Key Lab of Visual Damage and Regeneration & Restoration of Chongqing, Chongqing, Bioengineering College, Chongqing University, Chongqing 400030, China. Tel.: +86 23 65111931; fax: +86 23 65111931. ∗∗ Corresponding author. Tel.: +86 23 65111931; fax: +86 23 65111931. ∗∗∗ Corresponding author. Tel.: +82 53 810 2568; fax: +82 53 810 2062. ∗∗∗∗ Corresponding author at: Institute of Micro/Nanotechnology, Old Dominion University, Norfolk, VA 23529, USA. Tel.: +1 757 683 3304; fax: +1 757 683 3200. E-mail addresses: [email protected] (N. Hu), [email protected] (J. Yang), [email protected] (S.W. Joo), [email protected] (S. Qian). 0927-7765/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.colsurfb.2013.04.042
parallel electrodes, the device with a pair of parallel platinum electrodes apart has been widely adopted for the electroformation of giant vesicles [11–15]. In such a device, Dimitrov and Angelova [11] investigated some key factors, such as the lipid swelling and the charge and frequency of the imposed electric field, on the electroformation process of giant vesicles, and proposed a possible mechanism [12] for the electroformation process. Giant vesicles were also prepared by Angelova et al. [13] to investigate the enzyme-mediated vesicle transformation by microinjecting reagents. Bucher et al. [14] further electroformed giant vesicles as biochemical compartments, while Okumura et al. [15] electroformed giant vesicles on a non-electroconductive substrate that helped further examination and development of the electroformation by allowing various unconventional setups and substrate surfaces. In these devices, the distance between the two parallel electrodes was 0.5–4 mm, and the diameter of the electrodes was 0.48–0.5 mm, resulting in a non-uniform distribution of the electric field [25]. To induce a sufficiently high electric field for electroforming giant vesicles, an electric potential bias (0.01–17 V) is applied [10–15]. However, the shape of electrodes is typically cylindrical [10–15], making the droplets of lipid solution on the electrodes distributed unevenly and affecting the formation of lipid membranes and vesicles. In addition, the traditional device is not transparent,
82
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
so that it is difficult to observe the vesicles electroformed near two electrodes, especially those under the cylindrical electrode. In order to observe giant vesicles, Angelova et al. [16] used two transparent indium tin oxide (ITO) glass electrodes 0.3 mm apart each other to electroform giant vesicles. To further reduce the distance between the two parallel ITO glass electrodes, Kuribayashi et al. [21] developed a chip consisted of several microfluidic channels sandwiched between two ITO glass electrodes to produce various types of giant vesicles. Estes and Mayer [18–20] electroformed giant vesicles by first spin coating lipids on an ITO glass slide. The electroformation of giant vesicles can be observed in these transparent devices. The electric field strength between two ITO glass electrodes is higher than that between two parallel platinum wires, and the ITO glass electrodes are planar and have larger surface area than the parallel platinum electrodes, resulting in higher electroformation efficiency. In addition, the uniform distribution of the electric field between two ITO glass electrodes also produces a constant electrical field for electroformation, improving the performance of that device. However, precise control of the vesicle formation on ITO glass is challenging because it is not easy to pattern lipid film on ITO glass due to its complex surface properties [22]. In addition, the two parallel flat ITO glass electrodes are placed at different levels, resulting in vesicle chains stacked on top of each other along the direction of the electric field. Therefore, in these transparent devices it is still challenging to observe the electroformation process of individual giant vesicles, and thus hard to enlighten the mechanisms involved. With the rapid development of microfabrication technology, precise control of the vesicle formation with a desired size has attracted considerable attention. Microchips for the electroformation of giant vesicles have been developed. Taylor et al. [17] electroformed giant vesicles with a narrow size distribution on an ITO glass substrate, which had patterned lipid film by using polydimethylsiloxane (PDMS) stamp technique. The size of electroformed vesicles can be controlled by the size of the lipid film pattern. Using a micropatterned silicon dioxide layer with circular holes arranged in a hexagonal array on a Si substrate and an ITO glass electrode, separated by a 1 mm silicone rubber spacer, a microchip was developed by Le Berre et al. [22], and giant vesicles with a narrow size distribution were successfully obtained. Diguet et al. [23] fabricated a microdevice composed of a microstructured silicon wafers (with hexagonal arrays of holes) and an ITO glass slide as a counter electrode separated by 4 mm silicon rubber spacer, and giant vesicles with narrow size distribution were electroformed by applying an AC electric field between the two electrodes. Takeuchi and Kuribayashi [24] designed a microchip with a microaperture array to electroform dome-shaped organic-solvent-free artificial lipid membranes. This device overcomes some disadvantages of the ITO glass based devices, and has high yield to produce giant vesicles with narrow size distribution. However, in these devices, at least one ITO glass slide is used as one electrode, and it is separated from the other electrode by a micro-spacer. Furthermore, in these devices, lipid film must be formed at some given positions and manipulation is complicated [17,22–24]. In this study, a microchip with two types of microelectrode arrays, schematically shown in Fig. 1, is designed, fabricated and tested on a transparent quartz-glass substrate to electroform cell-size giant (unilamellar) vesicles (i.e., 10–20 m in diameter). In contrast to the traditional devices composed of two large parallel electrodes, such as the ITO glass slide, this microchip consists of two parallel chiasm-shaped microelectrode arrays (serpentineshaped microchannel sidewall). Each microelectrode array has many silicon microelectrode strips. In the first design each strip has protruding microelectrodes (Fig. 1b), resulting in spatially non-uniform electric field inside the microchannel. The second design does not protrude microelectrodes on each strip (Fig. 1c), and the resulting electric field between two opposite electrodes
is uniform. After loading the lipid solution into the microchannel, lipid film forms on the electrodes which are also the sidewalls of the microchannel, after the lipid solution is dried. Owing to the small distance between the opposing electrodes (i.e., the width of the microchannel), a strong electric field can be generated under a low electric voltage. Since the direction of the electric field on this chip is horizontal, the formed vesicles will not stack lengthwise, which is helpful for real-time observations. To further test that the electroformed giant vesicles have the capability to enclose other materials for applications, such as drug delivery and gene transferring, the formed giant vesicles are washed out, resuspended, separated by centrifugation, and reloaded into the same device for electrofusion.
2. Experimental methods 2.1. Materials and instruments l-␣-Phosphatidylcholine (1,2-diacyl-sn-glycero-3-phosphocholine), cholesterol (3-hydroxy-5-cholestene), and fluorescent dye (DiI) (1,1 -dihexadecyl-3,3,3 ,3 -tetramethylindocarbo-cyanie perchlorate, ex/em: 549/564 nm, Molecular Probes) were purchased from the Sigma–Aldrich, Inc. Centrifuge 5417R (Eppendorf, Germany) was used for the separation of giant vesicles by centrifugation. A home-made electrical signal generator was used to generate required electrical signal for the electroformation and electrofusion of giant vesicles, and DMI4000 B phase contrast inverted fluorescence microscope (Leica, Germany) was used for experimental observation.
2.2. Chip design and fabrication The microdevice consists of a serpentine-shaped microchannel of 42 m in depth, whose opposite sidewalls are made of two chiasm-shaped microelectrode arrays. Two types of microelectrode arrays, one with protruding microelectrodes to generate spatially non-uniform electric field (Fig. 1b) and the other with planar microelectrodes to generate uniform electric field (Fig. 1c), are designed for testing the effect of the distribution of local electric field on the electroformation of giant vesicles. For the design with protruding microelectrodes, the width of the microchannel is 80 m. Both the length and width of each protruding microelectrode are 20 m, and the distance between two adjacent protruding microelectrodes is 60 m (Fig. 1b). For the design with planar electrodes, the width of the microchannel and the distance between two counter electrodes are 80 m (Fig. 1c). The microchip was fabricated by using the MEMS (microelectro-mechanical systems) fabrication techniques [22–24]. The chiasm-shaped microelectrode arrays were fabricated on a 40 m thick, highly doped silicon wafer, which also worked as the vertical sidewalls of the microchannel. First, the silicon wafer (resistance 7–9 m, crystal orientation 1 0 0) was bonded with a 500 m thick Corning 7740 glass wafer by using the electrostatic-alloy bonding technique under the condition of 400 ◦ C, 600 V, and 1000 N force in the vacuum. Subsequently, a 50 nm Cr film was sputtered onto the silicon wafer, followed by electroplating 2 m Au on the surface of the Cr film. The unwanted Cr/Au was etched away by using KI etching solution (28 g KI; 20 g I2 , and 800 mL H2 O). Finally, the microelectrode arrays were etched on the heavily doped silicon by inductively coupled plasma (ICP) etching technology. The chip was then fixed on a printed circuit board (PCB), and the microelectrode arrays were connected to gold weld spots on the PCB by using gold silks of 75 m in diameter. The ceiling of the microchip was a PDMS layer, which was bonded on the microchip after the organic solvent of the lipid solution evaporated.
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
83
Fig. 1. Schematic diagram of on-chip giant vesicles electroformation process (a), protruding microelectrode array with spatially non-uniform electric field (b), and planar electrode array with uniform electric field (c).
2.3. Giant vesicles electroformation A mixture of phosphatidylcholine (PC) and cholesterol (Chol) was dissolved in ether to prepare lipid solutions with different concentrations. In each experiment, 25 L lipid solution was added on the chiasm-shaped microelectrode arrays by using pipette (Eppendorf, Germany). A lipid film formed on the bottom of the microchannel after the lipid solution was totally dried (volatilization of the solvent) by blowing it with a nitrogen stream for 5 min and then stored under negative pressure (0.04 MPa) for 2 h. Note that the thickness of the lipid film was not uniform due to the effect of surface tension and the rapid volatilization of the solvent. The lipid film was thicker near the sidewalls of the microchannel, and was thinner in the middle of the microchannel. After formation of the lipid film, the PDMS cap was bonded onto the microfluidic chip. Then 50 L buffer solution (100 mM glucose and 10 mM NaCl) was loaded into the microfluidic channel, and an AC electrical signal (sinusoidal wave, 0.05–1 V peak-to-peak, 10 Hz) was applied through the two microelectrode arrays for 1 h. The
experimental process was observed by using the phase contrast inverted fluorescence microscope. 2.4. Electrofusion of giant vesicles Electroformed vesicles suspension was collected and centrifuged (2500 rpm) to screen the giant vesicles with diameters larger than 10 m. Giant vesicles within the deposit were resuspended by using 500 L fresh buffer solution (100 mM glucose, 0.01 mM Ca2+ , 0.02 mM Mg2+ ) for electrofusion experiments. In the electrofusion experiments, 25 L giant vesicles suspension was loaded into the microfluidic device. An AC electrical signal (sinusoidal wave, 3 V peak-to-peak, 300 kHz) was imposed to the opposing microelectrode arrays to generate positive dielectrophoretic forces acting on the giant vesicles, which form pearl chains along the direction of the electric field. After alignment of giant vesicles, high-strength DC electric pulses (amplitude: 60 V, pulse width: 50 s, pulse interval: 200 ms, pulse number: 6) were imposed to the microelectrodes, resulting in reversible
84
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
Fig. 2. Expansion process of the giant vesicle in the electric field. From left to right, these pictures are the fluorescence microscopic pictures for t = 15, 20, 30 min, respectively. The processed fluorescence microscopic picture (the giant vesicle expansion process at the electrode position) is placed at the left-bottom of each fluorescence microscopic picture.
electroporation on the membranes of the paired giant vesicles, culminating in electrofusion. 3. Results and discussion 3.1. Distribution of electric field in the chip with protruding microelectrodes The electric field within the straight microchannel made up of parallel planar microelectrodes, estimated by dividing the imposed voltage bias by the distance of the two electrodes, is nearly uniform. However, the electric field within the microchannel with protruding microelectrodes is spatially non-uniform [26–28]. The distribution of electric field is obtained by numerically solving the Laplace equation using the commercial finite-element package,
COMSOL Multiphysics (www.comsol.com). In the numerical simulation, the electrical resistivities of the solution and the electrode were set to 4 × 104 and 0.1 .m, respectively, and the imposed voltage bias between the two electrodes was 0.25 V [11,14]. Fig. 1b and c depict the spatial distribution of the electric field, showing that the maximum electric field with strength 5791 V/m occurs near the protruding microelectrodes. Note that this electric filed strength is much higher than that in a device without protruding electrodes (i.e., 0.25 V/80 m = 3125 V/m). 3.2. Electroformation of giant vesicles The electroformation of giant vesicles was investigated in the two devices filled with lipid solutions of different mass concentrations of PC and Chol at different electric voltages (Vpp = 0.05–1 V)
Fig. 3. On-chip electroformation of giant vesicles in 1 h in the device with planar electrode array (left column) and protruding microelectrode array (right column).
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
85
Fig. 4. Giant vesicle swelled in a few seconds. (a) A giant vesicle closing the channel wall before swelling, and (b) the giant vesicle swelled to a sphere after 5 s.
with the frequency of the sinusoidal electrical signal kept at 10 Hz. The mass concentrations of PC and Chol affect the thickness of the formed lipid film. When the mass concentrations of the PC and Chol were, respectively, 0.5–12 mg/mL and 0.1–2.4 mg/mL, the electric voltages at Vpp = 0.25 V, giant vesicles were electroformed after imposing an AC electric signal. Fig. 2 depicts the electroformed giant vesicles near one of the protruding microelectrodes after the AC electric field was applied for 15 min, 20 min, and 30 min, while the corresponding images processed by color edge detection and linear combination function in the Matlab Imaging Processing Toolbox (www.mathworks.com) are located on the bottom left of each figure. A giant vesicle with diameter 10–20 m is formed near the protruding electrode. Fig. 3 depicts the electroformed giant vesicles in the device with protruding microelectrodes (the right column) and planar microelectrodes (the left column) after applying an AC signal for 1 h under different mass concentrations of PC and Chol. In the experiments, the mass concentration ratio between PC and Chol was 5. Table S1 in the supporting information compares the results obtained from the two devices under PC and Chol solutions with different mass concentrations. If the mass concentration of the lipid solution is relatively low, both ratio and velocity of electroforming giant (unilamellar) vesicles with diameters of 10–20 m increase with the increase in the concentrations
of a solution mixture of PC and Chol. However, if the concentration exceeds a threshold value (i.e., a solution mixture of 9 mg/mL PC and 1.8 mg/mL Chol), electroformed giant vesicles were stacked on top of another, and appeared as strips. In this case, the formation ratio of giant spherical vesicles stopped growing, and the ratio of giant spherical vesicles keeps a lower level with further increase in the concentration of the lipid solution. Note that in the present device a large ratio of giant (unilamellar) vesicles with diameters 10–20 m were electroformed when the concentrations of PC and Chol were, respectively, 6 mg/mL and 1.2 mg/mL, as shown in the 3rd row of Fig. 3. Comparisons between the corresponding results in the left (planar electrode array) and the right columns (protruding microelectrode array) of Fig. 3 show that the spatial distribution of the electric field has limited effect on giant vesicles electroformation. The influence of the imposed electric voltage bias, Vpp (ranging from 0.05 to 1 V), on the electroformation of giant (unilamellar) vesicles is illustrated in Table S2 of the supporting information, where the concentrations of PC and Chol solution were fixed at 6 mg/mL and 1.2 mg/mL, respectively. Note that for comparison, the results obtained from planar electrode array and interdigital microelectrode array are also presented in Table S2. If the imposed voltage bias is fixed, the results from these two devices are statistically the same, suggesting that the spatial distribution
Fig. 5. On-chip giant vesicle electrofusion. (a) The giant vesicles were contacted each other, (b) the contacted giant vesicles were electroporated within 2 s, (c) the contacted giant vesicles deformed themselves after 4 s and (d) the contacted giant vesicles have been fused after 6 s. All scale bars are 10 m.
86
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
of electric field plays a limited effect on the electroformation of giant vesicles. Experimental observations show that vesicles swell faster as the voltage bias increases. Some vesicles swelled to form spherical shape within a few seconds, as depicted in Fig. 4. However, the ratio of collected giant vesicles with diameters of 10–20 m did not increase significantly with the increase in the imposed voltage bias. During the form stage of giant vesicles, we observed more giant vesicles of diameter larger than 50 m and less 10–20 m giant vesicles formed with the increase in the imposed voltage bias. However, less giant vesicles of diameter larger than 50 m were collected. The reduction of the collected giant vesicles (≥50 m) was probably attributed to change of the membrane from bilayer to monolayer during the swell process. Since the monolayer membrane is very thin, it is easily broken during the collection process. The results show that a very low voltage bias is required in the device to quickly electroform giant vesicles. Although the spatial distribution of the electric field does not have significant effect on the quantity and formation rate of giant vesicles, we observed that the vesicles formed near the protruding microelectrodes are typically larger than those formed far away from them. In addition, the electroformation percentage within the protruding microelectrodes is about 60% (as shown in Fig. S1 in the supporting information), which is also higher than that within the planar electrodes. In the device with planar electrodes, larger vesicles are formed near the electrodes, which might be attributed to thicker lipid film near the electrodes. In the device with protruding microelectrodes, since the electric field along the diagonal of the opposite protruding microelectrodes is stronger than other places, we also observed that vesicles formed swell faster than those in other regions. Since the electric-field strength in the device with protruding microelectrodes is higher than that with planar electrodes and the swelling rate of vesicles increases with the electric field strength, the vesicles formed near the protruding electrodes swell faster than those near the planar electrodes. Electroformation of giant vesicles is influenced by many factors, such as the voltage bias and the frequency of the applied AC field [14,18], the lipid composition [12,14,18,29], the thickness of the lipid bilayer [12,14,18], the membrane fluidity [12,14,29], the properties of the electrode [15,18,22], the osmotic pressure [12,14,19], the duration time [12], and the temperature [29]. However, the mechanism of the electroformation of giant vesicles still remains unclear. Our experimental results showed that in the initial stage, the electroformation of giant vesicles was affected by the property of the lipid film. Vesicles were formed near the electrodes where the lipid film is thicker than that in the center of the microchannel. But most of the vesicles formed were not spherical, and their sizes were small. As the electric field applied continuously, the distribution of the electric field indeed affects the electroformed vesicles. In the region with a stronger electric field, vesicles swelled more quickly, and the adjacent lipid film gradually fuses and bends to form giant spherical vesicles. 3.3. Electrofusion of giant vesicles Electrofusion of giant vesicles is very important in many vesicles related applications [30–32], and thus we further investigate the electrofusion of giant vesicles prepared in the present microfluidic devices. We first collect the giant vesicles electroformed in the device with protruding microelectrodes under 6 mg/mL and 1.2 mg/mL concentrations of PC and Chol, respectively and Vpp = 0.45 V at 10 Hz. The collected giant vesicles were separated by using centrifugation method. The separated giant vesicles with diameters larger than 10 m (most of the giant vesicles collected were ca. 10–20 m) were resuspended by the electrofusion buffer,
and were reloaded into the same device with protruding microelectrodes. The reason to use the same device with protruding electrodes is due to the presence of the spatially non-uniform electric field, which induces pairing of the giant vesicles by the resulting positive dielectrophoresis (DEP) [26]. As shown in Fig. 5a, under an AC electric field (sinusoidal wave, 3 V peak-to-peak, 300 kHz), the positive DEP force [26,27,33–37] generated by the spatially non-uniform electric field drives the giant vesicles toward the protruding microelectrodes forming chains. After the giant vesicles were aligned near the protruding microelectrodes, DC electrical pulses (amplitude: 60 V, pulse width: 50 s, pulse interval: 200 ms, pulse number: 6) were applied to fuse the aligned giant vesicles. Fig. 5b–d shows the process of fusing a pair of giant vesicles aligned on one protruding microelectrode. Our experimental results show that about 50% of giant vesicles were aligned near the protruding microelectrodes due to positive DEP, and about 20% giant vesicles were fused. Note that most of the fused giant vesicles have similar diameters. The electrofusion of giant vesicles is affected by many factors, such as the diameter and component of the giant vesicles and the layers of the giant vesicle membranes. Giant vesicles with different diameters require different electric-field strength to electroporate their membranes [28]. Electrofusion of multilayer giant vesicles is more difficult than the monolayer ones. Components of giant vesicles affect their electrical properties, thus significantly affecting the electrofusion process. In order to yield high electrofusion efficiency, unilamellar giant vesicles with comparable diameters are desired. 4. Conclusions Electroformation and electrofusion of giant vesicles in a microfluidic device with microelectrode arrays have been investigated. Giant vesicles of diameters 10–20 m have been electroformed using two different designs. The electric field is spatially non-uniform in the device with protruding microelectrodes, while it is uniform in that with planar electrodes. As the concentrations of phosphatidylcholine (PC) and cholesterol (Chol) increase, thickness of the lipid film and quantity of the electroformed giant vesicles increase. Thin lipid film electroforms giant vesicles with monolayers, and thick lipid film forms giant vesicles with multilayers. Vesicles swell faster in the device with protruding microelectrodes due to the higher electric field strength. Most giant vesicles are formed near the protruding microelectrodes where the electric field strength is high. The device with protruding microelectrodes can also be used to fuse the electroformed giant vesicles. Giant vesicles are aligned near the protruding microelectrodes due to positive DEP, which arises from the spatially non-uniform electric field. Under high-strength DC electric pulses, the aligned giant vesicles can be fused. Due to the use of the transparent substrate, the devices developed allow observation of the electroformation and electrofusion processes. In addition, owing to the small distance between the opposite microelectrode arrays, a very low voltage is required for both giant vesicle electroformation and electrofusion. The highdensity microelectrode array makes it possible to electroform giant vesicles with a large quantity within a short time. Acknowledgments This work was supported by the National Natural Science Foundation of China (nos. 81071278, 81101168, 31070882), the Program for New Century Excellent Talents in University (no. NCET-09-0842), and the World Class University Grant R32-2008000-20082-0 of the National Research Foundation of Korea.
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
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.colsurfb. 2013.04.042. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]
N. Weiner, F. Martin, M. Riaz, Drug Dev. Ind. Pharm. 15 (1989) 1523. K.J. Ørgensen, J. Davidsen, O.G. Mouritsen, FEBS Lett. 531 (2002) 23. A. Samad, Y. Sultana, M. Aqil, Curr. Drug Deliv. 4 (2007) 297. G. Gregoriadis, B. McCormack, M. Obrenovic, R. Saffie, B. Zadi, Y. Perrie, Methods 19 (2002) 156. K. Tsumoto, S.M. Nomura, Y. Nakatani, K. Yoshikawa, Langmuir 17 (2001) 7225. A. Fischer, A. Franco, T. Oberholzer, ChemBioChem 3 (2002) 409. S. Kulin, R. Kishore, K. Helmerson, L. Locascio, Langmuir 19 (2003) 8206. M. Michel, M. Winterhalter, L. Darbois, J. Hemmerle, J.C. Voegel, P. Schaaf, V. Ball, Langmuir 20 (2004) 6127. M. Vestergaard, T. Hamada, M. Takagi, Biotechnol. Bioeng. 99 (2008) 753. M.I. Angelova, D.S. Dimitrov, Faraday Discuss. Chem. Soc. 81 (1986) 303. D.S. Dimitrov, M.I. Angelova, Prog. Colloid Polym. Sci. 73 (1987) 48. M. Angelova, D.S. Dimitrov, Prog. Colloid Polym. Sci. 76 (1988) 59. R. Wick, M.I. Angelova, P. Walde, P.L. Luisi, Chem. Biol. 3 (1996) 105. P. Bucher, A. Fischer, P.L. Luisi, T. Oberholzer, P. Walde, Langmuir 14 (1998) 2712. Y. Okumura, H. Zhang, T. Sugiyama, Y. Iwata, J. Am. Chem. Soc. 129 (2007) 1490. M.I. Angelova, S. Soléau, P. Méléard, J.F. Faucon, P. Bothorel, Prog. Colloid Polym. Sci. 89 (1992) 127.
87
[17] P. Taylor, C. Xu, P.D.I. Fletcher, V.N. Paunov, Phys. Chem. Chem. Phys. 5 (2003) 4918. [18] D.J. Estes, M. Mayer, Colloids Surf. B 42 (2005) 115. [19] D.J. Estes, M. Mayer, Biochim. Biophys. Acta Biomembr. 1712 (2005) 152. [20] D.J. Estes, S.R. Lopez, A.O. Fuller, M. Mayer, Biophys. J. 91 (2006) 233. [21] K. Kuribayashi, G. Tresset, P. Coquet, H. Fujita, S. Takeuchi, Meas. Sci. Technol. 17 (2006) 3121. [22] M. Le Berre, A. Yamada, L. Reck, Y. Chen, D. Baigl, Langmuir 24 (2008) 2643. [23] A. Diguet, M. Le Berre, Y. Chen, D. Baigl, Small 5 (2009) 1661. [24] S. Takeuchi, K. Kuribayashi, U.S. Patent, 2010307918 A1, 2010. ˇ H. Mekid, L.M. Mir, Biochim. Biophys. Acta 1523 (2000) [25] D. Miklavˇciˇc, D. Semrov, 73. [26] Y. Cao, J. Yang, Z.Q. Yin, H.Y. Luo, M. Yang, N. Hu, J. Yang, D.Q. Huo, C.J. Hou, Z.Z. Jiang, R.Q. Zhang, R. Xu, X.L. Zheng, Microfluid. Nanofluid. 5 (2008) 669. [27] N. Hu, J. Yang, S. Qian, S.W. Joo, X.L. Zheng, Biomicrofluidics 5 (2011) 034121. [28] N. Hu, J. Yang, Z.Q. Yin, Y. Ai, S. Qian, I.B. Svir, B. Xia, J.W. Yan, W.S. Hou, X.L. Zheng, Electrophoresis 32 (2011) 2488. [29] T. Shimanouchi, H. Umakoshi, R. Kuboi, Langmuir 25 (2009) 4835. [30] G. Tresset, S. Takeuchi, Biomed. Microdevices 6 (2004) 213. [31] G. Tresset, S. Takeuchi, Anal. Chem. 77 (2005) 2795. [32] N.G. Stoicheva, S.W. Hui, Biochim. Biophys. Acta Biomembr. 1195 (1994) 31. [33] Y. Cao, J. Yang, Z.Q. Yin, W.S. Hou, X.L. Zheng, N. Hu, J. Yang, R. Xu, R.Q. Zhang, Chin. J. Anal. Chem. 36 (2008) 593. [34] Y. Ai, S. Qian, J. Colloid Interface Sci. 346 (2010) 448. [35] J. Yang, L.P. Zhao, Z.Q. Yin, N. Hu, J. Chen, T.Y. Li, I. Svir, X.L. Zheng, Adv. Eng. Mater. 12 (2010) B398. [36] Y. Ai, S. Qian, S. Liu, S.W. Joo, Biomicrofluidics 4 (2010) 013201. [37] Y. Ai, B. Mauroy, A. Sharma, S. Qian, Electrophoresis 32 (2011) 2282.
Contents lists available at SciVerse ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Electroformation and electrofusion of giant vesicles in a microfluidic device Zhenyu Wang a , Ning Hu a,b,∗ , Li-Hsien Yeh c , Xiaolin Zheng a , Jun Yang a,∗∗ , Sang W. Joo b,∗ ∗ ∗ , Shizhi Qian b,d,∗ ∗ ∗∗ a Key Laboratory of Biorheological Science and Technology (Chongqing University), Ministry of Education, and Key Lab of Visual Damage and Regeneration & Restoration of Chongqing, Chongqing, Bioengineering College, Chongqing University, Chongqing 400030, China b School of Mechanical Engineering, Yeungnam University, Gyongsan 712-749, South Korea c Department of Chemical & Materials Engineering, National Yunlin University of Science and Technology, Yunlin 64002, Taiwan d Institute of Micro/Nanotechnology, Old Dominion University, Norfolk, VA 23529, USA
a r t i c l e
i n f o
Article history: Received 10 January 2013 Received in revised form 10 April 2013 Accepted 16 April 2013 Available online 30 April 2013 Keywords: Giant vesicle Electroformation Electrofusion Drug delivery Microfluidics
a b s t r a c t Electroformation and electrofusion of giant vesicles with diameters of 10–20 m have been performed in a microfluidic device with high-density microelectrodes forming the sidewalls of the microchannel. Electroformation of giant vesicles by a solution mixture of phosphatidylcholine (PC) and cholesterol (Chol) with different concentrations under AC electric field was investigated. Under the conditions of 0.5–12 mg/mL PC and 0.1–2.4 mg/mL Chol, vesicles were electroformed by the AC electric field imposed. About 60% electroformed vesicles were giant (unilamellar) vesicles with diameters 10–20 m. The eletroformed vesicles were collected from the chip, re-suspended in fresh buffer, and then separated by centrifugation to segregate the ones with desired diameters (10–20 m). Electrofusion of the giant vesicles was conducted in the same chip. Vesicles were aligned to form pairs under AC electric field due to positive dielectrophoresis, and the paired vesicles were subsequently fused upon the application of high strength electrical pulses. The alignment and fusion efficiencies were, respectively, about 50% and 20%. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Giant lipid membrane vesicles (giant vesicles) with diameters larger than 10 m have obtained widespread attentions for their potential applications, including drug delivery [1–4], microreactors [5–8], and modeling cytomembrane systems [9]. Since the electroformation method of giant vesicles was first developed by Angelova and Dimitrov [10] in 1986, a growing number of devices have been implemented to improve this technique [11–24]. Owing to relatively easy fabrication, chemically inert nature to various organic solvents, and adjustable distance between two
∗ Corresponding author at: Key Laboratory of Biorheological Science and Technology (Chongqing University), Ministry of Education, and Key Lab of Visual Damage and Regeneration & Restoration of Chongqing, Chongqing, Bioengineering College, Chongqing University, Chongqing 400030, China. Tel.: +86 23 65111931; fax: +86 23 65111931. ∗∗ Corresponding author. Tel.: +86 23 65111931; fax: +86 23 65111931. ∗∗∗ Corresponding author. Tel.: +82 53 810 2568; fax: +82 53 810 2062. ∗∗∗∗ Corresponding author at: Institute of Micro/Nanotechnology, Old Dominion University, Norfolk, VA 23529, USA. Tel.: +1 757 683 3304; fax: +1 757 683 3200. E-mail addresses: [email protected] (N. Hu), [email protected] (J. Yang), [email protected] (S.W. Joo), [email protected] (S. Qian). 0927-7765/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.colsurfb.2013.04.042
parallel electrodes, the device with a pair of parallel platinum electrodes apart has been widely adopted for the electroformation of giant vesicles [11–15]. In such a device, Dimitrov and Angelova [11] investigated some key factors, such as the lipid swelling and the charge and frequency of the imposed electric field, on the electroformation process of giant vesicles, and proposed a possible mechanism [12] for the electroformation process. Giant vesicles were also prepared by Angelova et al. [13] to investigate the enzyme-mediated vesicle transformation by microinjecting reagents. Bucher et al. [14] further electroformed giant vesicles as biochemical compartments, while Okumura et al. [15] electroformed giant vesicles on a non-electroconductive substrate that helped further examination and development of the electroformation by allowing various unconventional setups and substrate surfaces. In these devices, the distance between the two parallel electrodes was 0.5–4 mm, and the diameter of the electrodes was 0.48–0.5 mm, resulting in a non-uniform distribution of the electric field [25]. To induce a sufficiently high electric field for electroforming giant vesicles, an electric potential bias (0.01–17 V) is applied [10–15]. However, the shape of electrodes is typically cylindrical [10–15], making the droplets of lipid solution on the electrodes distributed unevenly and affecting the formation of lipid membranes and vesicles. In addition, the traditional device is not transparent,
82
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
so that it is difficult to observe the vesicles electroformed near two electrodes, especially those under the cylindrical electrode. In order to observe giant vesicles, Angelova et al. [16] used two transparent indium tin oxide (ITO) glass electrodes 0.3 mm apart each other to electroform giant vesicles. To further reduce the distance between the two parallel ITO glass electrodes, Kuribayashi et al. [21] developed a chip consisted of several microfluidic channels sandwiched between two ITO glass electrodes to produce various types of giant vesicles. Estes and Mayer [18–20] electroformed giant vesicles by first spin coating lipids on an ITO glass slide. The electroformation of giant vesicles can be observed in these transparent devices. The electric field strength between two ITO glass electrodes is higher than that between two parallel platinum wires, and the ITO glass electrodes are planar and have larger surface area than the parallel platinum electrodes, resulting in higher electroformation efficiency. In addition, the uniform distribution of the electric field between two ITO glass electrodes also produces a constant electrical field for electroformation, improving the performance of that device. However, precise control of the vesicle formation on ITO glass is challenging because it is not easy to pattern lipid film on ITO glass due to its complex surface properties [22]. In addition, the two parallel flat ITO glass electrodes are placed at different levels, resulting in vesicle chains stacked on top of each other along the direction of the electric field. Therefore, in these transparent devices it is still challenging to observe the electroformation process of individual giant vesicles, and thus hard to enlighten the mechanisms involved. With the rapid development of microfabrication technology, precise control of the vesicle formation with a desired size has attracted considerable attention. Microchips for the electroformation of giant vesicles have been developed. Taylor et al. [17] electroformed giant vesicles with a narrow size distribution on an ITO glass substrate, which had patterned lipid film by using polydimethylsiloxane (PDMS) stamp technique. The size of electroformed vesicles can be controlled by the size of the lipid film pattern. Using a micropatterned silicon dioxide layer with circular holes arranged in a hexagonal array on a Si substrate and an ITO glass electrode, separated by a 1 mm silicone rubber spacer, a microchip was developed by Le Berre et al. [22], and giant vesicles with a narrow size distribution were successfully obtained. Diguet et al. [23] fabricated a microdevice composed of a microstructured silicon wafers (with hexagonal arrays of holes) and an ITO glass slide as a counter electrode separated by 4 mm silicon rubber spacer, and giant vesicles with narrow size distribution were electroformed by applying an AC electric field between the two electrodes. Takeuchi and Kuribayashi [24] designed a microchip with a microaperture array to electroform dome-shaped organic-solvent-free artificial lipid membranes. This device overcomes some disadvantages of the ITO glass based devices, and has high yield to produce giant vesicles with narrow size distribution. However, in these devices, at least one ITO glass slide is used as one electrode, and it is separated from the other electrode by a micro-spacer. Furthermore, in these devices, lipid film must be formed at some given positions and manipulation is complicated [17,22–24]. In this study, a microchip with two types of microelectrode arrays, schematically shown in Fig. 1, is designed, fabricated and tested on a transparent quartz-glass substrate to electroform cell-size giant (unilamellar) vesicles (i.e., 10–20 m in diameter). In contrast to the traditional devices composed of two large parallel electrodes, such as the ITO glass slide, this microchip consists of two parallel chiasm-shaped microelectrode arrays (serpentineshaped microchannel sidewall). Each microelectrode array has many silicon microelectrode strips. In the first design each strip has protruding microelectrodes (Fig. 1b), resulting in spatially non-uniform electric field inside the microchannel. The second design does not protrude microelectrodes on each strip (Fig. 1c), and the resulting electric field between two opposite electrodes
is uniform. After loading the lipid solution into the microchannel, lipid film forms on the electrodes which are also the sidewalls of the microchannel, after the lipid solution is dried. Owing to the small distance between the opposing electrodes (i.e., the width of the microchannel), a strong electric field can be generated under a low electric voltage. Since the direction of the electric field on this chip is horizontal, the formed vesicles will not stack lengthwise, which is helpful for real-time observations. To further test that the electroformed giant vesicles have the capability to enclose other materials for applications, such as drug delivery and gene transferring, the formed giant vesicles are washed out, resuspended, separated by centrifugation, and reloaded into the same device for electrofusion.
2. Experimental methods 2.1. Materials and instruments l-␣-Phosphatidylcholine (1,2-diacyl-sn-glycero-3-phosphocholine), cholesterol (3-hydroxy-5-cholestene), and fluorescent dye (DiI) (1,1 -dihexadecyl-3,3,3 ,3 -tetramethylindocarbo-cyanie perchlorate, ex/em: 549/564 nm, Molecular Probes) were purchased from the Sigma–Aldrich, Inc. Centrifuge 5417R (Eppendorf, Germany) was used for the separation of giant vesicles by centrifugation. A home-made electrical signal generator was used to generate required electrical signal for the electroformation and electrofusion of giant vesicles, and DMI4000 B phase contrast inverted fluorescence microscope (Leica, Germany) was used for experimental observation.
2.2. Chip design and fabrication The microdevice consists of a serpentine-shaped microchannel of 42 m in depth, whose opposite sidewalls are made of two chiasm-shaped microelectrode arrays. Two types of microelectrode arrays, one with protruding microelectrodes to generate spatially non-uniform electric field (Fig. 1b) and the other with planar microelectrodes to generate uniform electric field (Fig. 1c), are designed for testing the effect of the distribution of local electric field on the electroformation of giant vesicles. For the design with protruding microelectrodes, the width of the microchannel is 80 m. Both the length and width of each protruding microelectrode are 20 m, and the distance between two adjacent protruding microelectrodes is 60 m (Fig. 1b). For the design with planar electrodes, the width of the microchannel and the distance between two counter electrodes are 80 m (Fig. 1c). The microchip was fabricated by using the MEMS (microelectro-mechanical systems) fabrication techniques [22–24]. The chiasm-shaped microelectrode arrays were fabricated on a 40 m thick, highly doped silicon wafer, which also worked as the vertical sidewalls of the microchannel. First, the silicon wafer (resistance 7–9 m, crystal orientation 1 0 0) was bonded with a 500 m thick Corning 7740 glass wafer by using the electrostatic-alloy bonding technique under the condition of 400 ◦ C, 600 V, and 1000 N force in the vacuum. Subsequently, a 50 nm Cr film was sputtered onto the silicon wafer, followed by electroplating 2 m Au on the surface of the Cr film. The unwanted Cr/Au was etched away by using KI etching solution (28 g KI; 20 g I2 , and 800 mL H2 O). Finally, the microelectrode arrays were etched on the heavily doped silicon by inductively coupled plasma (ICP) etching technology. The chip was then fixed on a printed circuit board (PCB), and the microelectrode arrays were connected to gold weld spots on the PCB by using gold silks of 75 m in diameter. The ceiling of the microchip was a PDMS layer, which was bonded on the microchip after the organic solvent of the lipid solution evaporated.
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
83
Fig. 1. Schematic diagram of on-chip giant vesicles electroformation process (a), protruding microelectrode array with spatially non-uniform electric field (b), and planar electrode array with uniform electric field (c).
2.3. Giant vesicles electroformation A mixture of phosphatidylcholine (PC) and cholesterol (Chol) was dissolved in ether to prepare lipid solutions with different concentrations. In each experiment, 25 L lipid solution was added on the chiasm-shaped microelectrode arrays by using pipette (Eppendorf, Germany). A lipid film formed on the bottom of the microchannel after the lipid solution was totally dried (volatilization of the solvent) by blowing it with a nitrogen stream for 5 min and then stored under negative pressure (0.04 MPa) for 2 h. Note that the thickness of the lipid film was not uniform due to the effect of surface tension and the rapid volatilization of the solvent. The lipid film was thicker near the sidewalls of the microchannel, and was thinner in the middle of the microchannel. After formation of the lipid film, the PDMS cap was bonded onto the microfluidic chip. Then 50 L buffer solution (100 mM glucose and 10 mM NaCl) was loaded into the microfluidic channel, and an AC electrical signal (sinusoidal wave, 0.05–1 V peak-to-peak, 10 Hz) was applied through the two microelectrode arrays for 1 h. The
experimental process was observed by using the phase contrast inverted fluorescence microscope. 2.4. Electrofusion of giant vesicles Electroformed vesicles suspension was collected and centrifuged (2500 rpm) to screen the giant vesicles with diameters larger than 10 m. Giant vesicles within the deposit were resuspended by using 500 L fresh buffer solution (100 mM glucose, 0.01 mM Ca2+ , 0.02 mM Mg2+ ) for electrofusion experiments. In the electrofusion experiments, 25 L giant vesicles suspension was loaded into the microfluidic device. An AC electrical signal (sinusoidal wave, 3 V peak-to-peak, 300 kHz) was imposed to the opposing microelectrode arrays to generate positive dielectrophoretic forces acting on the giant vesicles, which form pearl chains along the direction of the electric field. After alignment of giant vesicles, high-strength DC electric pulses (amplitude: 60 V, pulse width: 50 s, pulse interval: 200 ms, pulse number: 6) were imposed to the microelectrodes, resulting in reversible
84
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
Fig. 2. Expansion process of the giant vesicle in the electric field. From left to right, these pictures are the fluorescence microscopic pictures for t = 15, 20, 30 min, respectively. The processed fluorescence microscopic picture (the giant vesicle expansion process at the electrode position) is placed at the left-bottom of each fluorescence microscopic picture.
electroporation on the membranes of the paired giant vesicles, culminating in electrofusion. 3. Results and discussion 3.1. Distribution of electric field in the chip with protruding microelectrodes The electric field within the straight microchannel made up of parallel planar microelectrodes, estimated by dividing the imposed voltage bias by the distance of the two electrodes, is nearly uniform. However, the electric field within the microchannel with protruding microelectrodes is spatially non-uniform [26–28]. The distribution of electric field is obtained by numerically solving the Laplace equation using the commercial finite-element package,
COMSOL Multiphysics (www.comsol.com). In the numerical simulation, the electrical resistivities of the solution and the electrode were set to 4 × 104 and 0.1 .m, respectively, and the imposed voltage bias between the two electrodes was 0.25 V [11,14]. Fig. 1b and c depict the spatial distribution of the electric field, showing that the maximum electric field with strength 5791 V/m occurs near the protruding microelectrodes. Note that this electric filed strength is much higher than that in a device without protruding electrodes (i.e., 0.25 V/80 m = 3125 V/m). 3.2. Electroformation of giant vesicles The electroformation of giant vesicles was investigated in the two devices filled with lipid solutions of different mass concentrations of PC and Chol at different electric voltages (Vpp = 0.05–1 V)
Fig. 3. On-chip electroformation of giant vesicles in 1 h in the device with planar electrode array (left column) and protruding microelectrode array (right column).
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
85
Fig. 4. Giant vesicle swelled in a few seconds. (a) A giant vesicle closing the channel wall before swelling, and (b) the giant vesicle swelled to a sphere after 5 s.
with the frequency of the sinusoidal electrical signal kept at 10 Hz. The mass concentrations of PC and Chol affect the thickness of the formed lipid film. When the mass concentrations of the PC and Chol were, respectively, 0.5–12 mg/mL and 0.1–2.4 mg/mL, the electric voltages at Vpp = 0.25 V, giant vesicles were electroformed after imposing an AC electric signal. Fig. 2 depicts the electroformed giant vesicles near one of the protruding microelectrodes after the AC electric field was applied for 15 min, 20 min, and 30 min, while the corresponding images processed by color edge detection and linear combination function in the Matlab Imaging Processing Toolbox (www.mathworks.com) are located on the bottom left of each figure. A giant vesicle with diameter 10–20 m is formed near the protruding electrode. Fig. 3 depicts the electroformed giant vesicles in the device with protruding microelectrodes (the right column) and planar microelectrodes (the left column) after applying an AC signal for 1 h under different mass concentrations of PC and Chol. In the experiments, the mass concentration ratio between PC and Chol was 5. Table S1 in the supporting information compares the results obtained from the two devices under PC and Chol solutions with different mass concentrations. If the mass concentration of the lipid solution is relatively low, both ratio and velocity of electroforming giant (unilamellar) vesicles with diameters of 10–20 m increase with the increase in the concentrations
of a solution mixture of PC and Chol. However, if the concentration exceeds a threshold value (i.e., a solution mixture of 9 mg/mL PC and 1.8 mg/mL Chol), electroformed giant vesicles were stacked on top of another, and appeared as strips. In this case, the formation ratio of giant spherical vesicles stopped growing, and the ratio of giant spherical vesicles keeps a lower level with further increase in the concentration of the lipid solution. Note that in the present device a large ratio of giant (unilamellar) vesicles with diameters 10–20 m were electroformed when the concentrations of PC and Chol were, respectively, 6 mg/mL and 1.2 mg/mL, as shown in the 3rd row of Fig. 3. Comparisons between the corresponding results in the left (planar electrode array) and the right columns (protruding microelectrode array) of Fig. 3 show that the spatial distribution of the electric field has limited effect on giant vesicles electroformation. The influence of the imposed electric voltage bias, Vpp (ranging from 0.05 to 1 V), on the electroformation of giant (unilamellar) vesicles is illustrated in Table S2 of the supporting information, where the concentrations of PC and Chol solution were fixed at 6 mg/mL and 1.2 mg/mL, respectively. Note that for comparison, the results obtained from planar electrode array and interdigital microelectrode array are also presented in Table S2. If the imposed voltage bias is fixed, the results from these two devices are statistically the same, suggesting that the spatial distribution
Fig. 5. On-chip giant vesicle electrofusion. (a) The giant vesicles were contacted each other, (b) the contacted giant vesicles were electroporated within 2 s, (c) the contacted giant vesicles deformed themselves after 4 s and (d) the contacted giant vesicles have been fused after 6 s. All scale bars are 10 m.
86
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
of electric field plays a limited effect on the electroformation of giant vesicles. Experimental observations show that vesicles swell faster as the voltage bias increases. Some vesicles swelled to form spherical shape within a few seconds, as depicted in Fig. 4. However, the ratio of collected giant vesicles with diameters of 10–20 m did not increase significantly with the increase in the imposed voltage bias. During the form stage of giant vesicles, we observed more giant vesicles of diameter larger than 50 m and less 10–20 m giant vesicles formed with the increase in the imposed voltage bias. However, less giant vesicles of diameter larger than 50 m were collected. The reduction of the collected giant vesicles (≥50 m) was probably attributed to change of the membrane from bilayer to monolayer during the swell process. Since the monolayer membrane is very thin, it is easily broken during the collection process. The results show that a very low voltage bias is required in the device to quickly electroform giant vesicles. Although the spatial distribution of the electric field does not have significant effect on the quantity and formation rate of giant vesicles, we observed that the vesicles formed near the protruding microelectrodes are typically larger than those formed far away from them. In addition, the electroformation percentage within the protruding microelectrodes is about 60% (as shown in Fig. S1 in the supporting information), which is also higher than that within the planar electrodes. In the device with planar electrodes, larger vesicles are formed near the electrodes, which might be attributed to thicker lipid film near the electrodes. In the device with protruding microelectrodes, since the electric field along the diagonal of the opposite protruding microelectrodes is stronger than other places, we also observed that vesicles formed swell faster than those in other regions. Since the electric-field strength in the device with protruding microelectrodes is higher than that with planar electrodes and the swelling rate of vesicles increases with the electric field strength, the vesicles formed near the protruding electrodes swell faster than those near the planar electrodes. Electroformation of giant vesicles is influenced by many factors, such as the voltage bias and the frequency of the applied AC field [14,18], the lipid composition [12,14,18,29], the thickness of the lipid bilayer [12,14,18], the membrane fluidity [12,14,29], the properties of the electrode [15,18,22], the osmotic pressure [12,14,19], the duration time [12], and the temperature [29]. However, the mechanism of the electroformation of giant vesicles still remains unclear. Our experimental results showed that in the initial stage, the electroformation of giant vesicles was affected by the property of the lipid film. Vesicles were formed near the electrodes where the lipid film is thicker than that in the center of the microchannel. But most of the vesicles formed were not spherical, and their sizes were small. As the electric field applied continuously, the distribution of the electric field indeed affects the electroformed vesicles. In the region with a stronger electric field, vesicles swelled more quickly, and the adjacent lipid film gradually fuses and bends to form giant spherical vesicles. 3.3. Electrofusion of giant vesicles Electrofusion of giant vesicles is very important in many vesicles related applications [30–32], and thus we further investigate the electrofusion of giant vesicles prepared in the present microfluidic devices. We first collect the giant vesicles electroformed in the device with protruding microelectrodes under 6 mg/mL and 1.2 mg/mL concentrations of PC and Chol, respectively and Vpp = 0.45 V at 10 Hz. The collected giant vesicles were separated by using centrifugation method. The separated giant vesicles with diameters larger than 10 m (most of the giant vesicles collected were ca. 10–20 m) were resuspended by the electrofusion buffer,
and were reloaded into the same device with protruding microelectrodes. The reason to use the same device with protruding electrodes is due to the presence of the spatially non-uniform electric field, which induces pairing of the giant vesicles by the resulting positive dielectrophoresis (DEP) [26]. As shown in Fig. 5a, under an AC electric field (sinusoidal wave, 3 V peak-to-peak, 300 kHz), the positive DEP force [26,27,33–37] generated by the spatially non-uniform electric field drives the giant vesicles toward the protruding microelectrodes forming chains. After the giant vesicles were aligned near the protruding microelectrodes, DC electrical pulses (amplitude: 60 V, pulse width: 50 s, pulse interval: 200 ms, pulse number: 6) were applied to fuse the aligned giant vesicles. Fig. 5b–d shows the process of fusing a pair of giant vesicles aligned on one protruding microelectrode. Our experimental results show that about 50% of giant vesicles were aligned near the protruding microelectrodes due to positive DEP, and about 20% giant vesicles were fused. Note that most of the fused giant vesicles have similar diameters. The electrofusion of giant vesicles is affected by many factors, such as the diameter and component of the giant vesicles and the layers of the giant vesicle membranes. Giant vesicles with different diameters require different electric-field strength to electroporate their membranes [28]. Electrofusion of multilayer giant vesicles is more difficult than the monolayer ones. Components of giant vesicles affect their electrical properties, thus significantly affecting the electrofusion process. In order to yield high electrofusion efficiency, unilamellar giant vesicles with comparable diameters are desired. 4. Conclusions Electroformation and electrofusion of giant vesicles in a microfluidic device with microelectrode arrays have been investigated. Giant vesicles of diameters 10–20 m have been electroformed using two different designs. The electric field is spatially non-uniform in the device with protruding microelectrodes, while it is uniform in that with planar electrodes. As the concentrations of phosphatidylcholine (PC) and cholesterol (Chol) increase, thickness of the lipid film and quantity of the electroformed giant vesicles increase. Thin lipid film electroforms giant vesicles with monolayers, and thick lipid film forms giant vesicles with multilayers. Vesicles swell faster in the device with protruding microelectrodes due to the higher electric field strength. Most giant vesicles are formed near the protruding microelectrodes where the electric field strength is high. The device with protruding microelectrodes can also be used to fuse the electroformed giant vesicles. Giant vesicles are aligned near the protruding microelectrodes due to positive DEP, which arises from the spatially non-uniform electric field. Under high-strength DC electric pulses, the aligned giant vesicles can be fused. Due to the use of the transparent substrate, the devices developed allow observation of the electroformation and electrofusion processes. In addition, owing to the small distance between the opposite microelectrode arrays, a very low voltage is required for both giant vesicle electroformation and electrofusion. The highdensity microelectrode array makes it possible to electroform giant vesicles with a large quantity within a short time. Acknowledgments This work was supported by the National Natural Science Foundation of China (nos. 81071278, 81101168, 31070882), the Program for New Century Excellent Talents in University (no. NCET-09-0842), and the World Class University Grant R32-2008000-20082-0 of the National Research Foundation of Korea.
Z. Wang et al. / Colloids and Surfaces B: Biointerfaces 110 (2013) 81–87
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.colsurfb. 2013.04.042. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]
N. Weiner, F. Martin, M. Riaz, Drug Dev. Ind. Pharm. 15 (1989) 1523. K.J. Ørgensen, J. Davidsen, O.G. Mouritsen, FEBS Lett. 531 (2002) 23. A. Samad, Y. Sultana, M. Aqil, Curr. Drug Deliv. 4 (2007) 297. G. Gregoriadis, B. McCormack, M. Obrenovic, R. Saffie, B. Zadi, Y. Perrie, Methods 19 (2002) 156. K. Tsumoto, S.M. Nomura, Y. Nakatani, K. Yoshikawa, Langmuir 17 (2001) 7225. A. Fischer, A. Franco, T. Oberholzer, ChemBioChem 3 (2002) 409. S. Kulin, R. Kishore, K. Helmerson, L. Locascio, Langmuir 19 (2003) 8206. M. Michel, M. Winterhalter, L. Darbois, J. Hemmerle, J.C. Voegel, P. Schaaf, V. Ball, Langmuir 20 (2004) 6127. M. Vestergaard, T. Hamada, M. Takagi, Biotechnol. Bioeng. 99 (2008) 753. M.I. Angelova, D.S. Dimitrov, Faraday Discuss. Chem. Soc. 81 (1986) 303. D.S. Dimitrov, M.I. Angelova, Prog. Colloid Polym. Sci. 73 (1987) 48. M. Angelova, D.S. Dimitrov, Prog. Colloid Polym. Sci. 76 (1988) 59. R. Wick, M.I. Angelova, P. Walde, P.L. Luisi, Chem. Biol. 3 (1996) 105. P. Bucher, A. Fischer, P.L. Luisi, T. Oberholzer, P. Walde, Langmuir 14 (1998) 2712. Y. Okumura, H. Zhang, T. Sugiyama, Y. Iwata, J. Am. Chem. Soc. 129 (2007) 1490. M.I. Angelova, S. Soléau, P. Méléard, J.F. Faucon, P. Bothorel, Prog. Colloid Polym. Sci. 89 (1992) 127.
87
[17] P. Taylor, C. Xu, P.D.I. Fletcher, V.N. Paunov, Phys. Chem. Chem. Phys. 5 (2003) 4918. [18] D.J. Estes, M. Mayer, Colloids Surf. B 42 (2005) 115. [19] D.J. Estes, M. Mayer, Biochim. Biophys. Acta Biomembr. 1712 (2005) 152. [20] D.J. Estes, S.R. Lopez, A.O. Fuller, M. Mayer, Biophys. J. 91 (2006) 233. [21] K. Kuribayashi, G. Tresset, P. Coquet, H. Fujita, S. Takeuchi, Meas. Sci. Technol. 17 (2006) 3121. [22] M. Le Berre, A. Yamada, L. Reck, Y. Chen, D. Baigl, Langmuir 24 (2008) 2643. [23] A. Diguet, M. Le Berre, Y. Chen, D. Baigl, Small 5 (2009) 1661. [24] S. Takeuchi, K. Kuribayashi, U.S. Patent, 2010307918 A1, 2010. ˇ H. Mekid, L.M. Mir, Biochim. Biophys. Acta 1523 (2000) [25] D. Miklavˇciˇc, D. Semrov, 73. [26] Y. Cao, J. Yang, Z.Q. Yin, H.Y. Luo, M. Yang, N. Hu, J. Yang, D.Q. Huo, C.J. Hou, Z.Z. Jiang, R.Q. Zhang, R. Xu, X.L. Zheng, Microfluid. Nanofluid. 5 (2008) 669. [27] N. Hu, J. Yang, S. Qian, S.W. Joo, X.L. Zheng, Biomicrofluidics 5 (2011) 034121. [28] N. Hu, J. Yang, Z.Q. Yin, Y. Ai, S. Qian, I.B. Svir, B. Xia, J.W. Yan, W.S. Hou, X.L. Zheng, Electrophoresis 32 (2011) 2488. [29] T. Shimanouchi, H. Umakoshi, R. Kuboi, Langmuir 25 (2009) 4835. [30] G. Tresset, S. Takeuchi, Biomed. Microdevices 6 (2004) 213. [31] G. Tresset, S. Takeuchi, Anal. Chem. 77 (2005) 2795. [32] N.G. Stoicheva, S.W. Hui, Biochim. Biophys. Acta Biomembr. 1195 (1994) 31. [33] Y. Cao, J. Yang, Z.Q. Yin, W.S. Hou, X.L. Zheng, N. Hu, J. Yang, R. Xu, R.Q. Zhang, Chin. J. Anal. Chem. 36 (2008) 593. [34] Y. Ai, S. Qian, J. Colloid Interface Sci. 346 (2010) 448. [35] J. Yang, L.P. Zhao, Z.Q. Yin, N. Hu, J. Chen, T.Y. Li, I. Svir, X.L. Zheng, Adv. Eng. Mater. 12 (2010) B398. [36] Y. Ai, S. Qian, S. Liu, S.W. Joo, Biomicrofluidics 4 (2010) 013201. [37] Y. Ai, B. Mauroy, A. Sharma, S. Qian, Electrophoresis 32 (2011) 2282.