Detection of Association and Fusion of Giant Vesicles Using a

Sep 7, 2010 - Department of Bioinformatic Engineering, Graduate School of Information Science and Technology. ∥ Department of Frontier Biosciences, ...
0 downloads 9 Views 2MB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

Detection of Association and Fusion of Giant Vesicles Using a Fluorescence-Activated Cell Sorter )

Takeshi Sunami,†,# Filippo Caschera,‡,# Yuuki Morita,§ Taro Toyota,^ Kazuya Nishimura,§ Tomoaki Matsuura,†,§ Hiroaki Suzuki,†,§ Martin M. Hanczyc,‡ and Tetsuya Yomo*,†,§,

)

† Exploratory Research for Advanced Technology (ERATO), Japan Science and Technology Agency, Yamadaoka 1-5, Suita, Osaka 565-0871, Japan, ‡Center for Fundamental Living Technology (FLinT ), Institute of Physics and Chemistry, University of Southern Denmark, Campusvej 55, DK-5230 Odense M, Denmark, §Department of Bioinformatic Engineering, Graduate School of Information Science and Technology, Department of Frontier Biosciences, Graduate School of Frontier Biosciences, Osaka University, Yamadaoka 1-5, Suita, Osaka 565-0871, Japan, and ^Department of Basic Science, Research Center for Complex Systems Biology, Graduate School of Arts and Sciences, The University of Tokyo, 3-8-1 Komaba, Meguro, Tokyo 153-8902, Japan. # These authors contributed equally to this work.

Received July 6, 2010. Revised Manuscript Received August 30, 2010 We have developed a method to evaluate the fusion process of giant vesicles using a fluorescence-activated cell sorter (FACS). Three fluorescent markers and FACS technology were used to evaluate the extent of association and fusion of giant vesicles. Two fluorescent markers encapsulated in different vesicle populations were used as association markers; when these vesicles associate, the two independent markers should be observed simultaneously in a single detection event. The quenched fluorescent marker and the dequencher, which were encapsulated in separate vesicle populations, were used as the fusion marker. When the internal aqueous solutions mix, the quenched marker is liberated by the dequencher and emits the third fluorescent signal. Although populations of pure POPC vesicles showed no detectable association or fusion, the same populations, oppositely charged by the exogenous addition of charged amphiphiles, showed up to 50% association and 30% fusion upon population analysis of 100 000 giant vesicles. Although a substantial fraction of the vesicles associated in response to a small amount of the charged amphiphiles (5% mole fraction compared to POPC alone), a larger amount of the charged amphiphiles (25%) was needed to induce vesicle fusion. The present methodology also revealed that the association and fusion of giant vesicles was dependent on size, with larger giant vesicles associating and fusing more frequently.

1. Introduction Membrane fusion is one of the fundamental mechanisms of material transfer present in living cells.1-3 This mechanism is important in cell-cell, intracellular, and viral fusion events. Giant vesicles composed of phospholipids have been studied extensively as model cell membranes in investigating the physicochemical basis of membrane fusion. Giant vesicles can be produced in vitro with defined components such as purified phospholipids with defined aqueous phases.4-8 Although numerous protocols to prepare cell-sized giant vesicles have been described,9 the resulting vesicles vary widely in terms of size and number of lamelle.10 The structural organization of giant vesicles is also likely to play a significant role in fusion events; therefore, the characterization of a substantial sample of a diverse vesicle population is needed. *To whom correspondence should be addressed. Tel: þ81-6-6879-4171. Fax: þ81-6-6879-7433. E-mail: [email protected].

(1) Jahn, R.; Grubmuller, H. Curr. Opin. Cell Biol. 2002, 14, 488. (2) Lentz, B. R. Chem. Phys. Lipids 1994, 73, 91. (3) Tamm, L. K.; Crane, J.; Kiessling, V. Curr. Opin. Struct. Biol. 2003, 13, 453. (4) Franzin, C. M.; Macdonald, P. M. Biochemistry 1997, 36, 2360. (5) Pincet, F.; Lebeau, L.; Cribier, S. Eur. Biophys. J. 2001, 30, 91. (6) Toraya, S.; Nagao, T.; Norisada, K.; Tuzi, S.; Saito, H.; Izumi, S.; Naito, A. Biophys. J. 2005, 89, 3214. (7) Lei, G.; MacDonald, R. C. J. Membr. Biol. 2008, 221, 97. (8) Maruyama, T.; Yamamura, H.; Hiraki, M.; Kemori, Y.; Takata, H.; Goto, M. Colloids Surf., B 2008, 66, 119. (9) Walde, P.; Ichikawa, S. Biomol. Eng. 2001, 18, 143. (10) Nishimura, K.; Hosoi, T.; Sunami, T.; Toyota, T.; Fujinami, M.; Oguma, K.; Matsuura, T.; Suzuki, H.; Yomo, T. Langmuir 2009, 25, 10439. (11) Cypionka, A.; Stein, A.; Hernandez, J. M.; Hippchen, H.; Jahn, R.; Walla, P. J. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 18575.

15098 DOI: 10.1021/la102689v

Vesicle fusion is a complex event composed of at least two major steps.2,11-13 First, two or more vesicles come into close contact and become physically associated. Second, the intervening membrane separating these vesicles destabilizes and the internal aqueous contents mix, resulting in a single vesicle. In this letter, we define these steps as association and fusion, respectively. Microscopic observation is the most straightforward method of viewing the association and fusion of giant vesicles in real time. The shape changes and other morphological changes of individual giant vesicles can also be monitored at higher temporal resolution.14-17 However, the number of vesicles that can be observed by microscopy is extremely limited, and it is difficult to evaluate the behavior of heterogeneous giant vesicles statistically. Meanwhile, measuring the intermixing of the internal aqueous contents of vesicles in a bulk suspension is the best method of quantitatively evaluating vesicle fusion.18-21 This approach involves encapsulating two kinds of components in each of two vesicle populations. When the vesicles are fused together, a chemical (12) Lentz, B. R. Eur. Biophys. J. 2007, 36, 315. (13) Luisi, P. L.; de Souza, T. P.; Stano, P. J. Phys. Chem. B 2008, 112, 14655. (14) Pantazatos, D. P.; MacDonald, R. C. J. Membr. Biol. 1999, 170, 27. (15) Tanaka, T.; Yamazaki, M. Langmuir 2004, 20, 5160. (16) Lei, G.; MacDonald, R. C. Biophys. J. 2003, 85, 1585. (17) Haluska, C. K.; Riske, K. A.; Marchi-Artzner, V.; Lehn, J. M.; Lipowsky, R.; Dimova, R. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 15841. (18) Kendall, D. A.; MacDonald, R. C. J. Biol. Chem. 1982, 257, 13892. (19) Pantazatos, D. P.; Pantazatos, S. P.; MacDonald, R. C. J. Membr. Biol. 2003, 194, 129. (20) Struck, D. K.; Hoekstra, D.; Pagano, R. E. Biochemistry 1981, 20, 4093. (21) Wilschut, J.; Papahadjopoulos, D. Nature 1979, 281, 690.

Published on Web 09/07/2010

Langmuir 2010, 26(19), 15098–15103

Sunami et al.

Letter

reaction produces the fluorescence signal from the mixed components, and the total mixed volume fraction is evaluated. However, with this kind of bulk assay, it is not straightforward to resolve the intermediate states in the fusion process, and the state of individual vesicles cannot be evaluated. Here, we used a fluorescence-activated cell sorter (FACS) to study the fusion process of individual giant vesicles in a large sample. FACS is a powerful tool that can quantify the properties of every cell-sized particle using a high-throughput (up to 50 000 events/s) counting technique that can detect multiple colors of fluorescence.22 In our laboratory, FACS has been used to evaluate the structural properties of giant vesicles, including the inner aqueous volume, nominal membrane surface area,10,23 and internal structure24 as well as biochemical reactions such as substrate hydrolysis by enzymes,25 protein synthesis from DNA,24-28 and RNA replication by self-encoded replicase29 within giant vesicles. We devised a three-color fluorescence system to detect association and fusion. In this system, two fluorescent markers are encapsulated in the aqueous phase of each vesicle population to be fused. Thus, vesicles in the associated state (either in contact or fused) are recognized as the detection events that simultaneously emit both of these signals, whereas giant vesicles remaining in the discrete state emit only one of the fluorescence signals. To detect fusion, reagents that emit another fluorescence signal upon mixing were included in the respective vesicle populations. Giant vesicles in the fused state, which is defined as vesicles with internal content mixing, are detected by the appearance of the third fluorescence signal produced by the reaction of the mixed reagents. Using FACS, we can readily identify the states of association and fusion by analyzing the three fluorescence signals in a large fraction of the vesicle population. In this study, we used this fluorescence assay to investigate the effect of surface decoration of neutral giant vesicles with cationic and anionic amphiphiles30 on vesicle association and fusion. By injecting ethanol solutions containing either anionic or cationic amphiphiles into each preformed neutral vesicle suspension, the vesicles become electrically charged (decorated) as the exogenous amphiphiles are incorporated into the outer lipid membrane.30,31 By mixing these two populations, the oppositely charged vesicles associate and fuse.14,19 A quantitative assessment of the effect of the amphiphiles on the vesicle association and fusion ratio revealed that greater decoration with charged amphiphiles was required to induce fusion than the amount necessary for association only. We also evaluated the dependence of vesicle size on the association and fusion ratio, which suggested that larger giant vesicles tend to associate and fuse at higher frequency.

2. Experimental Section Materials. 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) was purchased from Avanti Polar Lipids (Alabaster, AL). (22) Taly, V.; Kelly, B. T.; Griffiths, A. D. ChemBioChem 2007, 8, 263. (23) Sato, K.; Obinata, K.; Sugawara, T.; Urabe, I.; Yomo, T. J. Biosci. Bioeng. 2006, 102, 171. (24) Hosoda, K.; Sunami, T.; Kazuta, Y.; Matsuura, T.; Suzuki, H.; Yomo, T. Langmuir 2008, 24, 13540. (25) Sunami, T.; Hosoda, K.; Suzuki, H.; Matsuura, T.; Yomo, T. Langmuir 2010, 26, 8544. (26) Ishikawa, K.; Sato, K.; Shima, Y.; Urabe, I.; Yomo, T. FEBS Lett. 2004, 576, 387. (27) Sunami, T.; Sato, K.; Matsuura, T.; Tsukada, K.; Urabe, I.; Yomo, T. Anal. Biochem. 2006, 357, 128. (28) Yu, W.; Sato, K.; Wakabayashi, M.; Nakaishi, T.; Ko-Mitamura, E. P.; Shima, Y.; Urabe, I.; Yomo, T. J. Biosci. Bioeng. 2001, 92, 590. (29) Kita, H.; Matsuura, T.; Sunami, T.; Hosoda, K.; Ichihashi, N.; Tsukada, K.; Urabe, I.; Yomo, T. ChemBioChem 2008, 9, 2403. (30) Caschera, F.; Stano, P.; Luisi, P. L. J. Colloid Interface Sci. 2010. (31) Thomas, C.; Luisi, P. J. Phys. Chem. B 2004, 108, 11285.

Langmuir 2010, 26(19), 15098–15103

Oleic acid, didodecyldimethylammonium bromide (DDAB), and cobalt(II) (Co2þ) chloride hexahydrate were purchased from Sigma-Aldrich (St. Louis, MO). Bis[N,N-bis(carboxymethyl)aminomethyl]fluorescein (calcein), R-phycoerythrin (R-PE), transferrin from human serum, and Alexa Fluor 647 conjugate (TA647) were purchased from Invitrogen (Carlsbad, CA). Ethylenediaminetetraacetic acid (EDTA) was purchased from DOJINDO Laboratories (Japan).

Preparation of Freeze-Dried Empty Liposome Membranes. Giant vesicles were prepared by the freeze-dried empty liposome (FDEL) method32 using procedures almost identical to those described in our previous reports.24,25,29 Briefly, POPC was dissolved in dichloromethane/diethyl ether (1:1 v/v), dried using a rotary evaporator, and then hydrated with Milli-Q water (12 mM lipid suspension). This suspension was vortex mixed for 20 s and sonicated for 5 s. After passing through a polycarbonate filter with a pore size of 0.4 μm (Nuclepore track-etched membranes; Whatman, Maidstone, Kent, U.K.), the suspension was dispensed into small aliquots (40 μL each) and freeze-dried overnight (Labconco Corp., Kansas City, MO). After being filled with argon gas, the freeze-dried membranes were stored in a freezer. Fluorescence-Generating System. We used the calceinCo2þ-EDTA system, as described by Kendall and MacDonald,18 to detect fusion. The concentration of Co2þ necessary to quench the fluorescence of calcein was determined as follows. Aliquots (20 μL) of the reaction mixture, consisting of various concentrations of CoCl2, 20 μM calcein, and 50 mM Hepes-KOH (pH 7.6), were prepared on ice. The change in the green fluorescence intensity at 37 C was measured with a real-time PCR system (Mx3005P QPCR system; Stratagene, San Diego, CA) at excitation and emission wavelengths of 492 and 610 nm, respectively. In order to convince the linearity of the fluorescence intensity within the detection range of the PCR system, we selected the emission wavelength of 610 nm which is off the peak of the calcein emission spectrum. Next, the concentration of EDTA required to chelate Co2þ in order to liberate calcein (dequenching of the calcein-Co2þ complex) was determined. Aliquots (20 μL) of the reaction mixture, consisting of various concentrations of EDTA, 2 mM CoCl2, 20 μM calcein, and 50 mM Hepes-KOH (pH 7.6), were prepared on ice. The subsequent procedures were identical to those described for the fluorescence-quenching assay.

Preparation of Giant Vesicles for Fusion Experiments. Two kinds of POPC vesicle populations were prepared for fusion. First, 10 μL of each reaction mixture was added to an aliquot of freeze-dried membranes (48 mM final lipid concentration). One mixture contained 10 mM EDTA, 1 μM TA647, and 50 mM Hepes-KOH (pH 7.6) (mixture 1), and the other contained 2 mM CoCl2, 20 μM calcein, 400 nM R-PE, and 50 mM Hepes-KOH (pH 7.6) (mixture 2). After a 20-fold dilution with 50 mM Hepes-KOH (pH 7.6), the vesicle suspensions were dialyzed with 50 mM Hepes-KOH (pH 7.6) using a microdialyzer (TOR-14K, Nippon Genetics, Japan; 14 000 Da molecular weight cutoff) for 2 h to remove the unencapsulated components (calcein, EDTA, and CoCl2) outside the vesicles. Then, 0.8 μL of the ethanol solution containing the positive (DDAB) or negative (oleic acid) amphiphile was added to 79.2 μL of the POPC vesicle suspension (2.4 mM lipid concentration) containing mixture 1 or 2, respectively. These vesicle suspensions were incubated at 25 C for 20 min to equilibrate the injected amphiphiles. Finally, 20 μL of a 2.5 mM CoCl2 solution was added to inhibit the unwanted dequenching of calcein-Co2þ in the suspension. To induce vesicle association and fusion, the two vesicle suspensions were mixed in a volume ratio of 1:1 (20 μL total) and incubated at 25 C for 30 min. The final vesicle mixtures were subjected to FACS measurement and microscopic observation. FACS Measurement and Data Processing. Before FACS measurement, the mixed vesicle suspension was diluted 5-fold (32) Kikuchi, H.; Suzuki, N.; Ebihara, K.; Morita, H.; Ishii, Y.; Kikuchi, A.; Sugaya, S.; Serikawa, T.; Tanaka, K. J. Controlled Release 1999, 62, 269.

DOI: 10.1021/la102689v

15099

Letter

Sunami et al.

with 50 mM Hepes-KOH (pH 7.6) to reduce the measurement frequency to less than 20 000 events/s. Then, three fluorescence signals (R-PE, TA647, and calcein) from individual vesicles were measured by FACS (FACSAria; Becton Dickinson, San Jose, CA). We obtained 100 000 data samples under each measurement condition. Calcein was excited with a semiconductor laser at 488 nm, and the emission was detected through a 530 ( 15 nm bandpass filter. R-PE was excited with the laser at 488 nm, and the emission was detected through a 575 ( 13 nm bandpass filter. TA647 was excited with a HeNe laser at 633 nm, and the emission was detected through a 660 ( 10 nm bandpass filter. During measurement, all fluorescence intensities (FIR-PE-obs, FITA647-obs, FIcalcein-obs) were compensated to eliminate fluorescence spectral overlap. For details, see the Supporting Information. Because the fluorescent proteins (R-PE and TA647) are uniformly encapsulated in vesicles,24,25 the volume of individual vesicles or aggregates can be determined from their fluorescence intensities. To achieve this, the fluorescence intensity of R-PE was first converted to the number of proteins using a standard curve derived from calibration beads bound to a known amount of R-PE (QuantiBRITE PE quantitation kit; BD Biosciences Clontech, Palo Alto, CA). Then, by assuming that R-PE was encapsulated at the same concentration as the initial mixture (400 nM), the vesicle volume was calculated using the equation VR-PE (fL) = 0.00357 FIR-PE. The number of TA647 molecules was converted from the TA647 fluorescence intensity (FITA647) using the ratio of intensity between TA647 and R-PE. The volume of vesicles encapsulating TA647 was calculated using the equation VTA647 (fL) = 0.00212 FITA647. The volume of the giant vesicles after amphiphile injection ranged from 0.1 to 100 fL, with that of most vesicles ranging from 0.4 to 1.4 fL (Supporting Information Figure S1). Differences in volume distribution may reflect the extent of vesicle aggregation, which can be enhanced or suppressed by decoration with charged amphiphiles. Microscopic Observation. Microscopic observation of the giant vesicles was carried out using an inverted light microscope (IX81; Olympus, Japan) with a 60 oil-immersion objective lens and a digital color charge-coupled-device camera (VB-7000, Keyence, Japan). Bright-field images were obtained through differential interference contrast observation. Images of calcein fluorescence were obtained through the corresponding filter and dichroic mirror unit (NIBA, Olympus, Japan; excitation 470-490 nm, emission 510-550 nm).

3. Results and Discussion Scheme of Vesicle Fusion. We prepared POPC giant vesicles containing fluorescent molecules as markers for association and fusion using the freeze-dried empty liposome (FDEL) method.24,25,29 The practical scheme of the vesicle fusion and detection strategy is illustrated in Figure 1. Two vesicle populations containing different contents were prepared. In one population, EDTA was used as the chelating agent and was encapsulated together with transferrin tagged with Alexa Fluor 647 (TA647), which emits farred fluorescence (mixture 1). In the second population, a mixture of Co2þ and calcein was encapsulated with fluorescent protein R-PE, which emits red fluorescence (mixture 2). Initially, mixture 2 is nonfluorescent within the green emission spectrum because Co2þ quenches the calcein. To induce vesicle association and fusion, oppositely charged amphiphiles were added to each vesicle population.30 DDAB was added as a cationic amphiphile to the former population, and oleic acid was added as an anionic amphiphile to the latter population. We defined the molar percentage of the added amphiphiles to the total amount of POPC and amphiphiles as MP = Mamp/(Mamp þ MPOPC)100(%), where Mamp and MPOPC are moles of added amphiphiles and pre-existing POPC, respectively. When these vesicles associate through electrostatic interaction but the internal aqueous phases do not mix, TA647 15100 DOI: 10.1021/la102689v

Figure 1. Schematic diagram showing the induction of vesicle fusion by the exogenous injection of charged amphiphiles and the detection of fusion using three fluorescent markers. EDTA and TA647 (association marker 1) were encapsulated in one vesicle population, and a nonfluorescent Co2þ-calcein complex (fusion marker) and R-PE (association marker 2) were encapsulated in another population. After adding oppositely charged amphiphiles (DDAB or oleic acid) to each population, the vesicle suspensions were mixed. When the oppositely charged vesicles associate, the fluorescence signals for R-PE and TA647 are detected simultaneously. When the vesicles fuse and the internal components are mixed, EDTA chelates Co2þ and the green fluorescence of the liberated calcein is detected. Plus, minus, crescent, star, and circle symbols represent positively charged amphiphiles (DDAB), negatively charged amphiphiles (oleic acid), EDTA, calcein, and Co2þ, respectively.

and R-PE fluorescence is simultaneously observed whereas the green fluorescence from calcein is not. Thus, TA647 and R-PE serve as markers to detect association (association markers). When the vesicles fuse and the internal aqueous phases mix, EDTA chelates Co2þ to release calcein, which emits green fluorescence.18 Thus, vesicles in the fused state emit all three fluorescence signals, and we refer to calcein as the fusion marker. We determined the concentrations of this calcein-Co2þ-EDTA system in preliminary titration experiments and confirmed that the fluorescence quenching and dequenching of calcein worked properly (Figure S2). Association and Fusion Assay of Giant Vesicles. After mixing two vesicle populations, we observed giant vesicles in the mixture under the microscope. First, we performed a control experiment where vesicles containing EDTA and those containing the Co2þ-calcein complex were mixed after injecting only ethanol (1% v/v). In this case, no charged amphiphiles were incorporated into the vesicles, because only ethanol was added to the vesicle suspension to match the experimental condition (Figure 2 (A1)). In this case, green fluorescence was not observed (Figure 2 (B1)). Next, the two vesicle populations containing EDTA and the Co2þ-calcein complex, which were decorated with oppositely charged amphiphiles at MP = 25%, were mixed. As shown in Figure 2 (A2), an intense green fluorescence signal was observed within the boundary of the vesicles (Figure 2 (B2)) because EDTA chelated the Co2þ ions to release calcein, which emitted an intense green signal as a marker for content mixing. To confirm that the green signal under this condition was not an artifact, we performed another control experiment in which vesicle populations decorated with the amphiphiles were mixed, but EDTA was not included in the cationic vesicles (Figure 2 (A3)). Under this condition, a green fluorescence signal was not observed in the giant vesicles (Figure 2 (B3)), confirming that the green signal is induced by the mixing of the internal aqueous phases. Langmuir 2010, 26(19), 15098–15103

Sunami et al.

Letter

Figure 2. Assay of the association and fusion of vesicles under three conditions. (A) Illustration of the experimental conditions. Plus, minus, crescent, star, and circle symbols represent positively charged amphiphiles (DDAB), negatively charged amphiphiles (oleic acid), EDTA, calcein, and Co2þ, respectively. (B) Microscopic images of vesicles after mixing the two populations. Images on the left were obtained by differential interference contrast observation, and those on the right were obtained by fluorescence microscopy. The scale bar indicates 10 μm. (C) Contour plots for vesicle frequency, as determined by FACS. The horizontal and vertical axes indicate the vesicle volume calculated from TA647 and R-PE fluorescence, respectively. Vesicles that fall into the R1 region were counted as associated vesicles. (D) Contour plots of vesicle frequency, as determined by FACS. The vertical axis shows the vesicle volume calculated from R-PE, and the horizontal axis shows the green fluorescence from calcein, which was used as the fusion marker. Vesicles that fall into the R2 region were counted as fused vesicles. (A1-D1) Results of the condition in which no charged amphiphiles were injected. (A2-D2) Condition under which all of the components required to detect fusion were included. (A3-D3) Control experiment in which EDTA was excluded.

Next, we performed FACS measurement for the same samples and examined the association and fusion of vesicles using the strategy presented in Figure 1. Figure 2C,D show the 2D density maps of the frequency of vesicles for combinations of fluorescence signals. The intensities of the embedded association markers were converted to the vesicle volume (fL). The frequency of vesicles was counted in logarithmic-scaled bins (width = 0.1 in common log units) to show their wide distribution (3 orders of magnitude). Figure 2C1-C3 show the volumes of two populations (horizontal Langmuir 2010, 26(19), 15098–15103

axis, volume from TA647; vertical axis, volume from R-PE). These plots characterize the association of the two vesicle populations (i.e., detected events that simultaneously have two association markers represent vesicles in both association and fusion states). Figure 2D1-D3 show the correlation between the intensity of the green fluorescence of calcein along the horizontal axis and one of the volume markers (R-PE) along the vertical axis. These plots characterize the extent of fusion events because the magnitude of green fluorescence increases upon fusion. The FACS DOI: 10.1021/la102689v

15101

Letter

Sunami et al.

measurement data for each population before mixing are plotted in Supporting Information Figure S3, which shows that the initial level of green fluorescence is low (background level). Figure 2C1,D1 shows the results of the control experiment in the absence of additional charged amphiphiles. In Figure 2C1, two distinct populations containing each of the association markers were apparent. This indicates that the two populations coexist in suspension without interacting with each other. Thus, no intermixing should occur, and the intensity of green fluorescence was limited to the background level (Figure 2D1). Figure 2C2,D2 shows the effect of adding charged amphiphiles at MP = 25%. Figure 2C2 shows that the vesicles became associated because of electrostatic interaction, and Figure 2C3 shows that the vesicles simultaneously emit both signals. As shown in Figure 2D2, there was a substantial increase in the calcein fluorescence intensity. In Figure 2C3, which shows the control without EDTA, the extent of association between the two populations was similar to that in Figure 2C3. As expected, the intensity of the green fluorescence signal did not increase (Figure 2D3). Taken together, these results demonstrate the utility of our vesicle fusion methods and analyses. Only vesicles properly modified with exogenous charged amphiphiles can participate in association and fusion on the timescale that we tested (30 min). Quantification of the Association and Fusion Yield at Various Amphiphile Concentrations. Experimental data obtained by FACS were used to assess the fusion process of vesicle populations quantitatively. We evaluated the dependence of the association and fusion yields on the number of amphiphiles injected for decoration. In the following analysis, we focused on giant vesicles larger than 1 fL (VR-PE g 1 fL) because we cannot resolve the extent of association for subfemtoliter vesicles due to the strong overlap in intensity of the two association markers between the two populations (Figure 2C1). We defined associated vesicles as those included in the R1 region indicated in Figure 2C1. In this region, giant vesicles from both populations aggregated, so that the aggregates contained the vesicles with total volume from each population exceeding 1 fL. Fused vesicles were defined as those included in both the R1 and R2 regions. (R2 is indicated in Figure 2D1.) The detected events that fulfill this criterion are associated vesicles that have a green fluorescence intensity greater than the background level (Figure 2D1). Then, we calculated the association ratio Ra and fusion ratio Rf, defined as Ra ¼

NR1 NR - PE

ð1Þ

Rf ¼

NR1∩R2 NR - PE

ð2Þ

the dependence of association and fusion yields on the size of giant vesicles. Because the volumes of the individual giant vesicles and aggregates are known, we can analyze the volume dependence of the fusion dynamics. We first noted the volume distributions of vesicles before and after mixing the two populations (Supporting Information Figure S5). If a large number of relatively small vesicles (10%. This indicates that, although the association of giant vesicles was induced by a small number of charged amphiphiles, a greater quantity was necessary to induce the fusion of these vesicles. Under the conditions tested, approximately 50% of the vesicles were associated and 30% were fused when MP = 25%. Quantification of the Association and Fusion Yield at Various Amphiphile Concentrations. Finally, we focused on 15102 DOI: 10.1021/la102689v

Figure 3. Association and fusion ratios (Ra and Rf) of giant vesicles. (A) Effect of the number of charged amphiphiles on Ra and Rf. (B) Dependence on the vesicle volume when MP = 25%. Values of Ra and Rf were plotted against the volume calculated from R-PE.

Here, we showed that two stages of the vesicle fusion process (i.e., association and internal content mixing) can be evaluated using FACS analysis in combination with three fluorescent markers (calcein, R-PE, and TA647). We examined the vesicle fusion process induced by the addition of oppositely charged amphiphiles to preformed neutral giant vesicles and found that the number of charged amphiphiles needed to induce vesicle fusion was much greater than that needed to induce vesicle association. In addition, because this approach can analyze the size of the vesicles, we found that larger vesicles tended to associate and fuse at higher frequency, resulting in an automatic size bias in the system. Although our analysis was performed for giant vesicles with V>1 fL, the present approach should also be applicable to submicrometric vesicles by improving the signal-to-noise ratio of the detection. Vesicle fusion is a major mechanism involved in material transport within living systems, and a fundamental understanding of this mechanism can lead to applications in artificially constructed Langmuir 2010, 26(19), 15098–15103

Sunami et al.

biochemical reaction platforms. The detailed analysis of the vesicle fusion process presented in this letter should provide a new tool for the stoichiometric understanding of fusion on the population level. We plan to utilize repeated cycles of vesicle fusion as a critical step in our artificial cell model in which nutrient materials are supplied for continual biochemical reactions within vesicles.29 With the present method, we can also add the charged amphiphiles to the fused and electrically neutralized vesicles to induce the second round of fusion. This study also demonstrates the utility of modifying vesicles after formation to decorate vesicles selectively for subsequent fusion. We believe that the present method of evaluating vesicle association and fusion will

Langmuir 2010, 26(19), 15098–15103

Letter

help to realize the optimum conditions for repetitive cycles of controlled vesicle fusion necessary for both complex nanoscale bioreactor platforms and evolving artificial cell models. Acknowledgment. This research was supported in part by the Global COE (Centers of Excellence) Program of the Japanese Ministry of Education, Culture, Sports, Science and Technology. Supporting Information Available: Compensation for the spectral overlap of the fluorescence intensities of the three markers. This material is available free of charge via the Internet at http://pubs.acs.org.

DOI: 10.1021/la102689v

15103