Temperature-Triggered Fusion of Vesicles Composed of Right

LETTER pubs.acs.org/Langmuir

Temperature-Triggered Fusion of Vesicles Composed of Right-Handed and Left-Handed Amphiphilic Helical Peptides Motoki Ueda,† Akira Makino,† Tomoya Imai,‡ Junji Sugiyama,‡ and Shunsaku Kimura*,† †

Department of Material Chemistry, Graduate School of Engineering, Kyoto University, Kyoto-Daigaku-Katsura, Nishikyo-ku, Kyoto 615-8510, Japan ‡ Research Institute for Sustainable Humanospherical (RISH), Kyoto University, Gokasho, Uji, Kyoto 611-0011, Japan

bS Supporting Information ABSTRACT: Vesicles prepared from a mixture of (Sar)25-b-(LLeu-Aib)6 (SLL) and (Sar)25-b-(D-Leu-Aib)6 (SDL) fused with themselves upon heating to 90 °C. The vesicles also fused with (Sar)28-b-(L-Leu-Aib)8 vesicles upon heating to 90 °C. The temperature-triggered fusion was due to the phase transition of the mixed membrane of SLL and SDL at 90 °C and should be driven by the bending energy stored in the stereocomplex membrane upon taking a vesicular structure.

T

he specific fusion of biological membranes is a central issue in many cellular processes.1 Studies on membrane fusion are important not only for the elucidation of bioprocesses but also for applications to biomaterials, including drug-delivery systems. For example, liposomes have been used as a model for cell membranes, and the fusion mechanism of liposomes has been clarified.2
5 However, the application of simple liposomes as drug carriers is not so feasible because of their weak physical properties. However, vesicles have been also prepared from amphiphilic polymers, which have the advantage of being robust.6
8 Polymer vesicles are, however, considered to be inactive in membrane fusion because membrane fluidity is significantly suppressed by the entanglement of polymer chains in the hydrophobic core of the membrane. It is therefore a challenging task to seek a robust vesicle with membrane-fusion activity. We demonstrate here successful fusion triggered by the high temperature of polypeptide vesicles, which are composed of a sterocomplex of (Sar)n-b-(L-Leu-Aib)n and (Sar)n-b(D-Leu-Aib)n. The amphiphilic block polypeptides comprise hydrophobic helical blocks, which associate regularly in the hydrophobic core of the vesicular membrane without chain entanglement. We have reported on molecular assemblies of amphiphilic peptide molecules, especially those with hydrophobic helical blocks in the hydrophobic core of the molecular assemblies.6
11 Helical peptides have a good ability to be packed regularly in the molecular assembly as shown by the frequent observation of helix bundles in nature.12,13 Indeed, a peptide nanotube with a diameter of ca. 70 nm and a length of ca. 200 nm was obtained from an amphiphilic block polypeptide with a hydrophobic helix {(Sar)25-b-(L-Leu-Aib)6 (SLL) with Mw/Mn (the molecular weight distribution) of 1.02, which was evaluated from MALDI-TOF mass spectroscopy, or (Sar)25-b-(D-Leu-Aib)6 (SDL) (Mw/Mn = 1.02) (Figure 1)}.14,15 Furthermore, we succeeded in preparing a vesicular assembly using r 2011 American Chemical Society

a mixture of amphiphilic polypeptides, SLL and SDL, that have right- and left-handed hydrophobic helices, respectively.15 When the right-handed helix is mixed with the left-handed helix, they form a stereocomplex that probably results from the convex-concave fitness between their surfaces.16,17 This vesicle from an equimolar mixture of SLL and SDL (DL vesicle) has a diameter of 180 nm with a narrow size distribution. A single component of (Sar)28-b-(L-Leu-Aib)8 (S28L16, Mw/ Mn = 1.03, Figure 1) formed round planar sheets with a diameter of ca. 100
200 nm with the concomitant presence of a minor fraction of vesicle in buffer at room temperature by transmission electron microscopy (TEM) observation (Figure 2a). These planar sheets were transformed completely to a vesicular shape of ca. 90 nm diameter (L16 vesicle) upon heat treatment at 90 °C for 1 h (Figure 2b). The L16 vesicle retained its morphology in buffer for at least for 2 months on the basis of TEM observation. We assayed the membrane fusion of these polypeptide vesicles by a fluorescence resonance energy transfer (FRET) technique, TEM observations, and DLS measurements. Two kinds of DL vesicles were prepared: one with 0.25% (mol/mol) 1,2-dipalmitoyl-snglycero-3-phosphoethanolamine-N-(7-nitro-2-1,3-benzoxadiazol4-yl) (NBD-PE; ex, 460 nm; em, 535 nm) and 0.25% (mol/mol) 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(lissamine rhodamine B sulfonyl) (Rho-PE; ex, 560 nm; em, 583 nm) and the other without these fluorophores. In the former double-labeled vesicle, upon photoexcitation of the NBD groups, fluorescence from the rhodamine groups at 583 nm was observed to be stronger than that from the NBD groups at 535 nm because of FRET from NBD to rhodamine in which both fluorophores are condensed in the same Received: December 30, 2010 Revised: March 17, 2011 Published: March 22, 2011 4300

dx.doi.org/10.1021/la105140v | Langmuir 2011, 27, 4300–4304

Langmuir vesicle (Figure 2c). When a mixture of the double-labeled vesicle and the label-free vesicle was incubated at 90 °C, the fluorescence intensity of NBD increased with incubation time whereas that from rhodamine decreased (Figure 2c). The change in the fluorescence pattern suggests the occurrence of membrane fusion, by which the

Figure 1. Molecular structure of amphiphilic polypeptides.

LETTER

fluorophores are diluted in the fused membrane to reduce FRET. The vesicle fusion was completed within 7.5 min after heating to 90 °C, but no fusion was observed even after 2 h when the heating temperature was 80 °C (Figure 2d). There seems to be a critical temperature for the membrane fusion of DL vesicles. The membrane fusion of L16 vesicles was examined by using similar methods. Two kinds of L16 vesicles were prepared: one with 0.25% (mol/mol) NBD-PE and 0.25% (mol/mol) Rho-PE and the other without these fluorophores. When these two kinds of vesicles were mixed and heated at 90 °C, no fluorescence spectral change was observed even after 2 h of heating (Figure 2e), suggesting that L16 vesicles do not fuse together in this temperature range. Morphology changes in DL vesicles and L16 vesicles upon heating to 90 °C were examined by DLS and TEM observations. The hydrodynamic diameter of DL vesicles was ca. 180 nm before heating and ca. 200 nm after heating. L16 vesicles kept their ca. 90 nm diameter before and after heating. TEM images of DL vesicles and L16 vesicles after heating show that the morphologies and sizes were preserved before and after heating (Figure 2f,g), supporting the results of DLS measurements. In the case of the fusion of DL vesicles, the fused vesicles are considered to undergo fission to vesicles with an original size of ca. 180 nm.

Figure 2. TEM images (negative staining with uranyl acetate) of molecular assemblies prepared from L16 (a) before and (b) after heating to 90 °C. Emission spectra from a mixture of DL vesicles and DL vesicles containing of 0.25% NBD-PE and 0.25% Rho-PE upon heating (c) to 90 °C and (d) to 80 °C. (e) Emission spectra from a mixture of L16 vesicles and L16 vesicles containing 0.25% NBD-PE and 0.25% Rho-PE upon heating to 90 °C. The photoexcitation wavelength was 460 nm. The assembly suspensions were prepared in 10 mM Tris-HCl buffer (pH 7.4, 1 mg/mL) by the ethanol injection method. TEM images (negative staining with uranyl acetate) of (f) DL vesicles and (g) L16 vesicles after heating to 90 °C. 4301

dx.doi.org/10.1021/la105140v |Langmuir 2011, 27, 4300–4304

Langmuir

Figure 3. DLS data of transformation from planar sheets to vesicles after holding at 50, 60, 70, 80, or 90 °C for 1 h in the case of (a) a mixture of SLL and SDL and (b) L16. (c) The hydrodynamic diameters of molecular assemblies from L16 are plotted against the variation in the heating period at 90 °C (b) and at 80 °C (O). DSC thermograms of (d) planar sheets prepared from a mixture of SLL and SDL and (e) those from L16.

To understand the reason for the inability of L16 vesicles to undergo membrane fusion, we focus our attention on the chain length of the hydrophilic block, poly(Sar), which should contribute significantly to the stability of vesicles in buffer. (Sar)14-b-(L-LeuAib)8 (S14L16) and (Sar)35-b-(L-Leu-Aib)8 (S35L16) were synthesized; they have shorter and longer hydrophilic blocks, respectively, than S28L16. S14L16 formed planar sheets of ca. 100
250 nm size by the ethanol injection method. Upon being held at 90 °C for 1 h, vesicles were observed but just as minor morphology (10% rate in the total molecular assemblies observed in the TEM images), and the rest retained the shape of the planar sheet (Figure S1 in the Supporting Information (SI)). However, (Sar)35-b-(L-Leu-Aib)8 (S35L16) formed very small sheets of ca. 30 nm2, and these small sheets retained their morphology after heat treatment at 90 °C for 1 h. These observations clearly support our consideration that the chain length of poly(Sar) influences the stability of the molecular assemblies. The planar sheet possesses hydrophobic edges at the periphery of the sheet, which are not stable in buffer. In the case of S35L16, long poly(Sar) chains should easily conceal the hydrophobic edges to stabilize the small sheets, which results in small sheets that are inadequate for vesicle formation. In the case of S14L16, the planar sheets could grow larger, but these sheets were not favorable for vesicle formation, probably because the outer membrane surface of vesicles, if transformation from a planar sheet to a vesicle took place, would generate hydrobphobic defects as a result of the large curvature of the membrane that could not be concealed by the short poly(Sar) chains. Taken together, the chain length of poly(Sar) of S28L16 is

LETTER

suited to taking a vesicular structure, and thus L16 vesicles are so stable that vesicle fusion does not occur. The transformation of morphology from planar sheets to vesicles was studied in detail by varying the incubation temperature. In the case of a mixture of SLL and SDL, the hydrodynamic diameters of the molecular assembly do not change upon increasing the incubation temperature to 80 °C for 1 h, and a sudden downsizing from 600 nm (planar sheet) to ca. 180 nm (vesicle) upon heating to 90 °C was observed (Figure 3a). This temperature coincides with the critical temperature for the fusion of DL vesicles determined by the FRET method. However, in the case of S28L16, the hydrodynamic diameters of the molecular assemblies increase gradually from ca. 45 nm (small planar sheet) to ca. 90 nm (vesicle) with increasing temperature (Figure 3b,c). These observations suggest that DL vesicles have a phase-transition temperature of around 90 °C, which may reflect the sol
gel transition as observed with liposomes. However, S28L16 does not have such a critical temperature. The temperature dependence of the helical conformation was examined by CD measurements. For this purpose, we prepared a nanotube assembly composed of SLL.15 CD spectra of the SLL nanotube remained nearly unchanged in the temperature range from 25 to 95 °C (Figure S2 in SI). The local unfolding of the helical conformation is therefore considered not to occur with thermal fusion and fission. The molecular packing mode of the hydrophobic helices in the membrane should be changed upon heat treatment when retaining the helical structure. To characterize peptide membranes further, vesicles were analyzed by differential scanning calorimetry (DSC). DSC measurement of planar sheets prepared from a mixture of SLL and SDL showed an endothermic peak at 89 °C (Figure 3d), but planar sheets prepared from L16 did not show any endothermic peaks (Figure 3e). The membrane of a mixture of SLL and SDL may therefore become fluidic above 89 °C to trigger vesicle fusion. However, there should be another factor in vesicle fusion because L16 membranes should be fluidic in the temperature region examined here when we consider the fact that the morphology transformation from small planar sheets to vesicles took place more frequently with increasing temperature. The morphology transformation process should require the diffusion or displacement of peptide molecules in the membrane to realize the molecular rearrangement that is suitable for vesicles. Membrane fluidity is therefore not enough for vesicle fusion but rather a prerequisite, even though we cannot completely exclude the possibility that the different degrees of membrane fluidity between DL and L16 vesicles may explain the observations. We speculate that the high bending energy stored in DL vesicles contributes to vesicle fusion in addition to membrane fluidity. A mixture of SLL and SDL yields large planar sheets by the ethanol injection method, which makes a contrast with curved sheets prepared from SLL or SDL. In the mixed membranes of SLL and SDL, the helical chirality in total is cancelled out by stereocomplex formation to generate large planar sheets. However, membranes of SLL or SDL have chirality that depends on the right-handed or the left-handed helix to induce curvature in the membrane. It is therefore considered that the mixed membranes of SLL and SDL should be rigid and possess high bending energy upon taking a vesicular structure. However, the membranes of L16 vesicles do not have such a high bending energy, which is suitable for the spontaneous formation of small vesicles of ca. 90 nm diameter even at room temperature, whereas DL vesicles take a large diameter of ca. 180 nm upon heating to 90 °C. 4302

dx.doi.org/10.1021/la105140v |Langmuir 2011, 27, 4300–4304

Langmuir

LETTER

Figure 4. Emission spectra from a mixture of DL vesicles and L16 vesicles containing of 0.25% NBD-PE and 0.25% Rho-PE upon heating to (a) 90 and (b) 80 °C. TEM images of (c) a mixture of two kinds of vesicles after heating to 90 °C for 1 h and (d) molecular assemblies prepared from a mixture of SLL, SDL, and L16 (1/1/2 w/w/w) after heating to 90 °C for 1 h. (e) Histogram of diameters of L16 vesicles (blue), DL vesicles (yellow), and a mixture of DL vesicles and L16 vesicles after holding at 90 °C for 1 h (green)

Figure 5. Elution profiles of FITC-dextran loading DL vesicles through a Sephacryl S-100 column (a) after the purification and subsequent elution of fraction 8 (in profile a) and (b) before and (c) after heat treatment at 90 °C for 30 min. (d) Schemetic illustration of the fused and fission vesicle without the loaded FITC-dextran in DL vesicles leaking out.

The fusion of DL vesicles and L16 vesicles was examined. L16 vesicles containing 0.25% NBD-PE and 0.25% Rho-PE were

prepared and mixed with DL vesicles without these fluorophores. Upon heating to 90 °C, the fluorescence intensity of NBD at 4303

dx.doi.org/10.1021/la105140v |Langmuir 2011, 27, 4300–4304

Langmuir 535 nm increased and that of rhodamine at 583 nm decreased, suggesting that these two kinds of vesicles fuse together to dilute fluorophores in the membrane, resulting in a reduction of FRET (Figure 4a). Upon holding at 80 °C even for 2 h, no fusion was suggested from the fluorescent spectra (Figure 4b). The critical temperature of 90 °C for the fusion of DL vesicles and L16 vesicles coincides with that for the fusion of DL vesicles, indicating that the physical change in DL membranes at 90 °C triggers the fusion. The DLS measurement of the molecular assemblies after heating to 90 °C revealed the hydrodynamic diameter of ca. 140 nm, which is a value between DL vesicles of ca. 180 nm and L16 vesicles of ca. 90 nm. Furthermore, the TEM image confirms that the fused morphology is of a vesicle with a diameter of ca. 140 nm (Figure 4c,e). The fused vesicles were stable at 25 °C at least for 2 weeks on the basis of DLS measurements. Taken together, DL vesicles and L16 vesicles fuse together upon heating to 90 °C, and the fused vesicles should fission to vesicles with an intermediate size of ca. 140 nm. The fused vesicles of ca. 140 nm should therefore be composed of SLL, SDL, and S28L16. To support this interpretation, a mixture of SLL, SDL, and S28L16 (1/1/2 w/w/w) was injected into buffer and held at 90 °C for 1 h. The TEM image of the molecular assemblies clearly shows the formation of vesicles of ca. 140 nm diameter (Figure 4d), which are exactly the same as those prepared by the fusion of DL and L16 vesicles. It is therefore concluded that DL vesicles and L16 vesicles fuse together at 90 °C, and the fused membrane should be composed of a homogeneous mixture of SLL, SDL, and L16 by the diffusion of peptide molecules in the membrane. With this composition, vesicles of ca. 140 nm diameter should be the most stable morphology after heating to 90 °C. To elucidate the fusion mechanism of DL vesicles, the release behavior of the fluorescent agent encapsulated in DL vesicles was studied during vesicle fusion. Fluorescein-labeled dextran (FITC-dextran) (Mw = 4000) (ex, 480 nm; em, 520 nm) was used as an encapsulating reagent. DL vesicles loaded with FITCdextran were prepared by the ethanol injection method using an FITC-dextran solution (2 mg/mL). The solution was held at 90 °C for 1 h, followed by elution through a Sephacryl S-100 column (1.5  30 cm2, GE healthcare Bio-Sciences) using 10 mM Tris-HCl buffer (pH 7.4) as an eluent (Figure 5a). The peak at fraction number 5
9 represents DL vesicles encapsulating FITC-dextran. Fraction number 11
20 corresponds to free FITC-dextran. The eighth fraction was eluted again through the same column (Figure 5b), showing nearly no leakage from DL vesicles during the column operation. This result showed that DL vesicles encapsulated FITC-dextran, and the amount of physisorbed FITC-dextran was negligible. The eighth fraction was held again at 90 °C for 30 min, which induced vesicle fusion as described before. Then the fraction was eluted through the same column. The elution profile clearly shows nearly no leakage during vesicle fusion (Figure 5c). Taken together, it is concluded that DL vesicles fuse to each other to connect two vesicular membranes, followed by fission into two vesicles of 200 nm diameter without leaking the encapsulated reagents (Figure 5d). The membranes of DL vesicles are therefore speculated to be very flexible for distortions during vesicle fusion.

LETTER

S14L16 and S35L16. CD spectra of the molecular assembly that forms a nanotube structure. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Fax: (þ81)75-383-2401. Tel: (þ81)75-383-2400. E-mail: [email protected].

’ ACKNOWLEDGMENT This study is a part of joint research focusing on the development of the basis of technology for establishing COE for nanomedicine, carried out through the Kyoto City Collaboration of Regional Entities for Advancing Technology Excellence (CREATE) assigned by the Japan Science and Technology Agency (JST). ’ REFERENCES (1) Blumenthal, R. Dimitrov, D. S. Handbook of Physiology; Oxford University Press: New York, 1997; Chapter 14, pp 563
603. (2) Discher, D. E.; Eisenberg, A. Science 2002, 297, 967–973. (3) Jonkheijm, P.; Schoot, P. V. D.; Schenning, A. P. H. J.; Meijer, E. W. Science 2006, 313, 80–83. (4) Engelkamp, H.; Middelbeek, S.; Nolte, R. J. M. Science 1999, 284, 785–788. (5) Percec, V.; Dulcey, A. E.; Balagurusamy, V. S. K.; Miura, Y.; Smidrkal, J.; Peterca, M.; Nummelin, S.; Edlund, U.; Hudson, S. D.; Heiney, P. A.; Duan, H.; Magonov, S. N.; Vinogradov, S. A. Nature 2004, 430, 764–768. (6) Tanisaka, H.; Kizaka-Kondoh, S.; Makino, A.; Tanaka, S.; Hiraoka, M.; Kimura, S. Bioconjugate Chem. 2008, 19, 109–117. (7) Makino, A.; Kizaka-Kondoh, S.; Yamahara, R.; Hara, I; Kanzaki, T.; Ozaki, E.; Hiraoka, M.; Kimura, S. Biomaterials 2009, 30, 5156–5160. (8) Fujita, K.; Kimura, S.; Imanishi, Y. Langmuir 1999, 15, 4377–4379. (9) Kimura, S.; Kim, D.-H.; Sugiyama, J.; Imanishi, Y. Langmuir 1999, 15, 4461–4463. (10) Kimura, S.; Muraji, Y.; Sugiyama, J.; Fujita, K.; Imanishi, Y. J. Colloid Interface Sci. 2000, 222, 265–267. (11) Ueda, M.; Makino, A; Imai, T.; Sugiyama, J.; Kimura, S. J. Pept. Sci. 2010, 17, 94–99. (12) Milburn, M. V.; Prove, G. G.; Milligan, D. L.; Scott, W. G.; Yeh, J.; Jancarik, J; Koshland, D. E.; Kim, S.-H. Science 1991, 254, 1342–1347. (13) Parker, M. W.; Pattus, F.; Tucker, A. D.; Tsernoglou, D. Nature 1989, 337, 93–96. (14) Kanzaki, T.; Horikawa, Y.; Makino, A.; Sugiyama, J.; Kimura, S. Macromol. Biosci. 2008, 8, 1026–1033. (15) Ueda, M.; Makino, A; Imai, T.; Sugiyama, J.; Kimura, S. Chem. Commun. 2011, 47, 3204–3206. (16) Ikada, Y.; Jmshidi, K.; Tsuji, H.; Hyon, S.-H. Macromolecules 1987, 20, 904–906. (17) Brizzolara, D.; Cantow, H.-J.; Diederichs, K.; Keller, E.; Domb, A. J. Macromolecules 1996, 29, 191–197.

’ ASSOCIATED CONTENT

bS

Supporting Information. Materials and methods. TEM images of molecular assemblies from single components of 4304

dx.doi.org/10.1021/la105140v |Langmuir 2011, 27, 4300–4304