Article pubs.acs.org/Langmuir
Long-Time-Scale Interaction Dynamics between a Model Antimicrobial Peptide and Giant Unilamellar Vesicles Matthew G. Burton,† Qi M. Huang,† Mohammed A. Hossain,†,‡ John D. Wade,†,‡ Andrew H. A. Clayton,*,§ and Michelle L. Gee*,† †
School of Chemistry and ‡Florey Department of Neuroscience and Mental Health, Centre for Neuroscience Research Chemistry, University of Melbourne, Parkville, Victoria 3010, Australia § Centre for Micro-Photonics, Swinburne University of Technology, Hawthorn, Victoria 3122, Australia S Supporting Information *
ABSTRACT: The interaction dynamics between a lytic peptide and a biomembrane was studied using time-lapse fluorescence lifetime imaging microscopy. The model membrane was 1,2-dipalmitoyl-sn-glycero-3phosphochloine giant unilamellar vesicles (GUVs), and the peptide was the K14 derivative of melittin, to which the polarity-sensitive fluorescent probe AlexaFluor 430 was grafted. The interaction of the peptide with the GUVs resulted in a progressive quenching of the fluorescence lifetime over a period of minutes. From previous photophysics characterization of the peptide, we were able to deconvolve the contribution of three distinct peptide states to the lifetime trajectory and use this data to develop a kinetics model for the interaction process. It was found that the peptide− membrane interaction was well described by a two-step mechanism: peptide monomer adsorption followed by membrane surface migration, assembly, and insertion to form membrane pores. There was an equilibrium exchange between pore and surface monomers at all lipid/peptide (L/P) concentration ratios, suggesting that the fully inserted phase was reached, even at low peptide concentrations. In contrast to previous studies, there was no evidence of critical behavior; irrespective of L/P ratio, lytic pores were the dominant peptide state at equilibrium and were formed even at very low peptide concentrations. We suggest that this behavior is seen in GUVs because their low curvature means low Laplace pressure. Membrane elasticity is therefore relatively ineffective at damping the thermal fluctuations of lipid molecules that lead to random molecular-level lipid protrusions and membrane undulations. The transient local membrane deformations that result from these thermal fluctuations create the conditions necessary for facile peptide insertion.
■
INTRODUCTION
One approach to probing the mechanism of peptide− membrane interactions is by studying their interaction kinetics. It has been shown that the initial binding of antimicrobial peptides to a membrane occurs on the order of milliseconds.15,16 Longer-time-scale events have been observed, including the reorientation and conformational change of the peptide upon insertion into the membrane,17 exchange between membrane-adsorbed and free peptide,18 and osmotic swelling of a vesicle due to peptide−lipid interactions.19 Cooperativity between peptide monomers has been suggested to be a key step in the mechanism toward membrane pore formation.20−22 More recently, slow insertion kinetics and the formation of peptide oligomers that are transient states in the mechanisms toward pore formation have been observed.23 Much of the work to date on peptide−membrane interactions has utilized model membranes, predominantly small unilamellar vesicles (SUVs), large unilamellar vesicles, or
Antimicrobial peptides are a diverse array of small molecular species found naturally in a wide variety of organisms, including insects, amphibians, humans, and plants.1−3 Many of these peptides form part of an innate host defense system that can preferentially target and kill bacterial cells.4 These peptides have gained significant attention as the basis for the design of new therapeutics to combat the increasing threat of antimicrobial resistance.5,6 There are many factors that influence peptide− membrane and lipid−peptide interactions. These include the lipid membrane composition, degree of lipid hydration/fluidity, membrane curvature, lipid headgroup charge, and lipid-topeptide concentration ratio.3,6−9 Despite extensive investigations on a variety of antimicrobial peptides spanning several decades, a precise mechanistic understanding of their mode of action remains elusive. Two of the more popular models of peptide−membrane interaction are based on either the formation of pore structures within the membrane9−11 or the perturbation and/or solubilization of the membrane in a detergentlike manner.12−14 © 2013 American Chemical Society
Received: August 9, 2013 Revised: October 29, 2013 Published: October 29, 2013 14613
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
Article
methanol mixture (9.8:1). Filtered HEPES buffer was then carefully added to the organic phase, with the organic phase subsequently removed via rotary evaporation for 2 to 3 min at 40 rpm, 40 °C, 100 mbar minimum pressure. The resulting GUV suspension was centrifuged at 15 300g for 10 min to remove multilamellar vesicles and lipid debris. The supernatant was collected and stored in glass vials at room temperature (approximately 23 °C) for up to 2 weeks. Frequency Domain Fluorescence Lifetime Imaging Microscopy Measurements. All fluorescence lifetime imaging microscopy (FLIM) experiments were performed in duplicate using a frequencydomain Lambert Instruments fluorescence lifetime attachment (Lambert Instruments, Leutingwolde, The Netherlands) interfaced with an inverted microscope (TE2000U, Nikon Inc., Japan). GUV samples were observed in situ through a 100× NA 1.2 oil objective (Nikon Plan-Fluor, Nikon Inc., Japan). The fluorescence excitation source was a 470 nm LED with a sinusoidal modulation frequency of 40 MHz. From the resulting phase image stacks (12 phase images per stack), phase and modulation lifetimes, τϕ and τM, respectively, were determined. Note that in frequency domain measurements two fluorescence lifetimes are determined: the phase lifetime, τϕ, is calculated from the phase lag of the fluorescence signal relative to the excitation signal, and the modulation lifetime, τM, is calculated from the amplitude modulation of the fluorescence signal relative to the excitation signal. Photobleaching was corrected by varying the phase angle of the excitation source over 0−360° during acquisition.31 The reference used for all lifetime determinations was rhodamine 6G (lifetime τ = 4.1 ns).32 An experimental sample was prepared by placing a 100 μL aliquot of a GUV solution into a glass observation cell. GUVs were allowed to settle for 45−60 min, after which transmitted light microscopy was used to identify a suitable GUV of interest. The total lipid concentration was estimated from the number and size distribution of GUVs in the field of view over a number of different areas across the entire sample. A sufficient amount of the melittin derivative solution was then injected into the sample solution for global L/P ratios of 100:1, 50:1, and 27:1 (total peptide concentration 0.27−1 μM). Injections were made gently and as close to the point of interest as possible to minimize diffusion effects of the fluorescent peptide. Fluorescence lifetime image stacks were recorded every 1−4 min for the first 15 min after the introduction of peptide and then every 15 min after that for up to 3 h. Transmitted light microscopy images were recorded for every phase image stack. The fluorescence lifetime measurements were represented in two different ways. First, to get an indication of trends, the net fluorescence lifetime was taken and both phase (τϕ) and modulation lifetimes (τM) were plotted. The second approach utilized the AB plot, also referred to as a phasor or polar plot, to display the lifetime experiments graphically.33−37 In such a plot, an experiment is represented by a point in 2D space defined by x = m cos ϕ and y = m sin ϕ, where ϕ is the phase and m is the modulation of the fluorescence signal. A series of lifetime experiments plotted as a series of points and joined by lines is called a trajectory. This graphical approach has the advantage that the type of fluorescence decay (simple, complex, excited-state reaction/solvent relaxation) and the complexity of the system trajectory (binary or more complex) can be deduced visually without further analysis. The phasor of single-exponential-decaying fluorophores lies on a semicircle described by m = cos ϕ that intersects with points (0, 0), (0.5, 0.5,) and (1, 0). Phasors from heterogeneous fluorescence decays lie inside the semicircle such that m < cos ϕ. Excited-state reactions including solvent relaxation have phasors that lie outside the semicircle: m > cos ϕ. The linear combination of two phasors is described by a linear trajectory in AB space, whereas the mixing among three or more species is nonlinear. Note that in frequency domain measurements two lifetimes are determined: the phase lifetime, τϕ, is determined from the phase lag of the fluorescence signal relative to the excitation signal, and the modulation lifetime, τM, is determined from the amplitude modulation of the fluorescence signal relative to the excitation signal.
supported lipid bilayers. In light of the fact that membrane curvature appears to affect the observed behavior of a peptide,8,24 it is imperative that studies employ a model system that is geometrically similar to that of a living bacterial cell. Furthermore, indirect analysis techniques on small and large unilamellar vesicles report on the average behavior of the entire ensemble, which may mask important mechanistic information occurring specifically at the membrane. Giant unilamellar vesicles (GUVs) have the potential to overcome these issues since because of their size (on the order of micrometers) they can be viewed microscopically, allowing the direct extraction of qualitative and quantitative information from regions of interest.25,26 In the present work, we used time-lapse frequency domain fluorescence lifetime imaging microscopy (FLIM) to monitor long-time-scale (minutes to hours) peptide−membrane interaction kinetics. Our model peptide is a derivative of melittin, a 26 amino acid cationic peptide that is predominantly random coil in solution27 but rapidly adopts an α-helical conformation upon association with phospholipid bilayers.16,28 The melittin derivative is based on the parent peptide but with a proline 14− lysine 14 substitution. The polarity-sensitive fluorescence probe AlexaFluor 430 grafted at lysine 14 is near the center of the peptide and is able to sample local micropolarity near the center of the peptide helix. This fluorescently tagged melittin derivative is α-helical in membranes and more effective at membrane lysis than its parent.29 This allows us to monitor directly the interaction kinetics of the peptide based upon the photophysical behavior of the fluorescent tag. We have monitored its interaction with 1,2-dipalmitoyl-sn-glycero-3phosphatidylcholine (DPPC) giant unilamellar vesicles (GUVs). Previously, we observed complex kinetics of the melittin derivative interacting with small unilamllar vesicles (SUVs) during which several peptide states are formed during helix insertion into the vesicle membrane that are intermediate in lytic pore formation.23 We show here that the kinetics of the peptide’s interaction with GUVs are quite different from those seen with SUVs, suggesting a different interaction mechanism. This has important implications for peptide−cell interactions.
■
MATERIALS AND METHODS
Materials. 1,2-Dipalmitoyl-sn-glycero-3-phosphatidylchloine (DPPC) was purchased from Avanti Polar Lipids (Alabaster, AL). UV/vis spectroscopy grade chloroform and methanol were purchased from Merck (Darmstadt, Germany). 4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES, 99% purity) was purchased from Acros Organics (Geel, Belgium). The melittin derivative, melittin K14 AlexaFluor 430, was synthesized on a Fmoc-PAL-EG-PS resin, as detailed previously.29 The AlexaFluor 430 fluorescent label was purchased from Molecular Probes (Carlsbad, CA). All materials were used without further purification. Milli-Q water with a resistivity of 18.2 MΩ was generated by a Milli-Q ultrapure water system or a MilliQ academic system (Millipore Corp., Bedford, CT) and was used in the preparation of all aqueous solutions. HEPES buffer solutions (10 mM HEPES, pH 7.40 ± 0.05) filtered through 0.22 μm hydrophilic poly(1,1,2,2-tetrafluoroethylene) (Teflon) membranes (Millipore Corp., Bedford, CT) were used in the preparation of all peptide and vesicle solutions. The pH was adjusted using small volumes of concentrated sodium hydroxide (8 M). The melittin derivative solutions were prepared at concentrations of 5−20 μM and were stored at 4 °C for no longer than 2 weeks. Preparation of Giant Unilamellar Vesicles. GUVs composed of pure DPPC were prepared using the solvent evaporation method.30 Briefly, a concentrated solution of DPPC powder dissolved in spectroscopy grade chloroform (0.1 M) was added to a chloroform/ 14614
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
■
Article
RESULTS Time-Lapse Fluorescence Lifetime Measurements of Peptide−Membrane Interactions. Time-lapse frequency domain fluorescence lifetime imaging microscopy (FLIM) was performed to monitor the evolution of the lipid−peptide interaction between our fluorescently tagged melittin derivative and DPPC GUVs as a function of time for three lipid/peptide (L/P) concentration ratios. A representative set of FLIM images is shown in Figure 1, together with the corresponding
50:1, and 27:1), in agreement with prior observations that high peptide concentrations (i.e., low global L/P ratios ( k2 at all times t. This, however, resulted in very poor fits to the experimental results (data not shown). We find here for the interaction of peptide with GUVs that the fastest step in the mechanism is the aggregation of peptide on the
step 1: k1
free interfacial peptide → membrane‐adsorbed monomer
step 2, forward: k2
membrane‐adsorbed monomer → lytic pore peptide
step 2, reverse: k3
lytic pore peptide → membrane‐adsorbed monomer
In this reaction scheme, which represents the process followed by individual peptide monomers during peptide− membrane interaction, k1 is the rate constant for the adsorption of a peptide monomer onto the membrane to form a membrane-adsorbed monomer. This is followed by peptide insertion/pore formation, which occurs at a rate determined by the rate constant k2. In the proposed reaction scheme, we allow for the reversibility of peptide aggregation and pore formation at a rate determined by k3. Note that fluorescence lifetime measurements were made on the vesicle of interest to exclude bulk solution. The term “free interfacial peptide” therefore means peptide in the interfacial region but not bound to the 14617
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
Article
Table 1. Rate Constants for Each Step in the Peptide− Membrane Interaction Mechanisma L/P ratio
k1 (min−1)
k2 (min−1)
k3 (min−1)
100:1 50:1 27:1
0.05 ± 0.02 0.12 ± 0.02 0.20 ± 0.04
1.8 ± 0.4 1.03 ± 0.2 0.49 ± 0.06
0.13 ± 0.02 0.08 ± 0.02 0.10 ± 0.02
a
k1 is the rate constant for the adsorption of free interfacial peptide onto the membrane. Membrane-adsorbed monomers then insert into the membrane to form membrane pores at a rate determined by rate constant k2. The reversibility of pore formation occurs at a rate determined by k3. Rate constants are given for each global lipid/ peptide (L/P) concentration ratio studied and were obtained by fitting time-lapse fluorescence lifetime measurements to eqs 6−8. The errors in the rate constants are estimated to be around 20%. This is a conservative estimate that takes into account experimental errors and errors in model fitting.
membrane and insertion to form lytic pores. It is interesting, however, that this step becomes slower the higher the relative peptide concentration. Another interesting result is that the reverse process (i.e., peptide aggregate dissociation) occurs at an intermediate rate that is independent of the L/P concentration ratio. We discuss the implications of this in the Discussion section below. It is a little more intuitive to discuss the kinetics of peptide− membrane interaction in terms of mole fractions rather than fractional fluorescence. We have therefore taken the rate constants in Table 1 obtained from fitting the kinetics model to the fluorescence time-lapse data (Figure 4) and back calculated mole fractions for each peptide state using eqs 6−8. The resulting data are shown in Figure 5 and show that at all global L/P ratios membrane pores are rapidly formed and that the pore peptide is the dominant peptide state.
Figure 5. Time-lapse fractional concentrations during peptide− membrane interaction from each identified peptide state: free interfacial peptide (blue diamonds), membrane-adsorbed peptide monomers (green triangles), and lytic pore peptides (red circles) derived from the kinetics modeling defined by eqs 6−8. Data is shown for global L/P ratios: (a) 100:1, (b) 50:1, and (c) 27:1. High fractional concentrations of pore peptide are observed at all L/P ratios studied, even at peptide concentrations far below that expected for pore formation. Errors in fractional concentration are 12% as per the fractional fluorescence data of Figure 4
■
DISCUSSION We have examined the interaction dynamics of a melittin derivative with giant unilamellar phospholipid vesicles (GUVs) using time-lapse fluorescence lifetime imaging microscopy (FLIM). The peptide derivative has a proline-14 to lysine substitution onto which polarity-sensitive fluorescent probe AlexaFluor 430 has been grafted. Fluorescent tagging at this point on the peptide chain allows an examination of the microenvironment local to the center of the peptide helix. As described above, we have found that the interaction between the melittin derivative and DPPC GUVs is best fit by a modified two-step mechanism. Table 1 contains the rate constants for each step at different L/P ratios, extracted by fitting this model to the time-lapse fluorescence data of Figure 4. The first step, free peptide → membrane-adsorbed monomer with rate constant k1, is rate-limiting but increases in rate as the concentration of peptide increases. This is expected for a simple pseudo-first-order adsorption process that is diffusion-controlled. Recall, however, that membrane-adsorbed monomer has a fluorescence lifetime consistent with the fluorescent probe on the peptide sampling the headgroup region of the membrane, implying that the peptide is oriented with its long axis parallel to the membrane surface. This is consistent with previous studies looking at fluorescently labeled native melittin interacting with lipid membranes.42 This means that there must be surface reorientation of peptide after initial nonspecific adsorption until the AlexaFluor 430 probe is positioned to sample the headgroup region of the membrane. The first-order behavior of this step in the mechanism implies that the
adsorption and surface reorientation of peptide are unhindered and do not involve any cooperativity between neighboring peptide molecules. In contrast, the second step, membrane-adsorbed monomer → lytic pore peptide, has a rate (k2) that suggests that cooperativity is important in this step of the process. This is the fast step in the interaction mechanism, highlighting the ease of peptide insertion into GUVs. But peptide insertion and pore formation become slower as the peptide concentration increases. The assembly of membrane-adsorbed peptide monomers into pore peptide relies on the surface migration and reorientation of peptide molecules. This becomes increasingly hindered as the surface becomes more crowded, presumably through steric and electrostatic repulsion between peptides. The reverse of the second step, lytic pore peptide → membrane-adsorbed monomer, is slow (k3) and independent of peptide concentration within experimental error. This implies that pores do not disassemble once formed but that any migration of pore peptide to the surface is part of an equilibrium exchange or coexistence between pore peptides and peptide monomers. This is consistent with Huang et al.,9 who discussed the critical concentration at which there is a 14618
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
Article
coexistence region where a fraction of peptide is inserted and the rest remains on the surface, with the inserted fraction increasing as the peptide concentration increases. The observation that there is an equilibrium exchange between pore and surface monomers means that we have reached the fully inserted phase even at low peptide concentrations. This is consistent with the measured fractional concentrations of peptide states (Figure 5). It has been shown previously23 that the interaction of the melittin mutant with DPPC SUVs has long-time-scale (30 min−2 h) complex kinetics. Pore formation depends critically on the L/P ratio and does not occur until the L/P ratio is 20:1 (relatively high peptide concentration), consistent with other studies9 and predicted by Huang.43 For the melittin derivative interacting with DPPC SUVs, the mechanism of peptide− membrane interaction involves initial rapid peptide adsorption followed by the formation of several oligomeric peptide states that are inserted into the membrane and sample the alkyl region of the bilayer. These peptide states are intermediates to the slowly forming lytic pores. In contrast to our previous work on peptide−membrane interactions with SUVs,23 we observed that the kinetics of interaction between the melittin derivative and DPPC GUVs are similar across all global L/P ratios (Figure 5) and there is no evidence of critical behavior within the observed L/P ratio range. In fact, lytic pores are the dominant peptide state at equilibrium for all peptide concentrations, even at those far lower than the critical concentration for pore formation found for SUVs (i.e., L/P ratio of around 25:1). This is a surprising result with criticality in lipid−peptide interactions reported to occur around a lipid/peptide ratio of 30:1, consistent with the predictions of Huang9,43 and reported measurements (ref 23 and references therein). This suggests that a very different mechanism of interaction is occurring in GUVs. The absence of criticality is initially surprising because a lower-curvature membrane should lead to closer packing of lipid molecules, hindering peptide insertion.14,44,45 Schwarz et al.46 have shown that the interaction of lytic peptides with large unilamellar vesicles resulted in fewer pores per mole of lipid compared to their interaction with SUVs. On this basis, we might have expected critical behavior with pore formation at higher peptide concentration than seen with SUVs of the same lipid. It is important, however, to consider that a lipid membrane is a fluid, thermally mobile system. Thermal fluctuations of lipid molecules lead to transient molecular-level lipid protrusions,47,48 and the collective thermal fluctuations of lipid molecules result in membrane ripples or undulations.49,50 On the time and length scales of these fluctuations, the membrane is locally deformed, altering the headgroup spacing and creating the conditions for facile peptide insertion. This is illustrated schematically in Figure 6. It is likely to occur in synergy with membrane point defects51 and localized membrane thinning created by the peptide−lipid interaction.43 We observe pore formation in DPPC GUVs even at low peptide concentrations (high L/P ratios) because the low vesicle Laplace pressure (due to low global curvature) results in low membrane elasticity.51 Membrane thermal fluctuations are therefore not damped, relatively speaking. In high-curvature SUVs, the high Laplace pressure renders the membrane more elastic, damping membrane thermal fluctuations and thus hindering peptide insertion.
Figure 6. Visual representation of the two-step peptide−membrane interaction kinetics mechanism for peptides interacting with GUVs, as defined by the reaction scheme. Phospholipids are shown here with a single acyl chain tail for clarity. Peptide free in solution binds to the surface of the membrane and undergoes subsequent reconfiguration such that the fluorescent probe samples the membrane headgroup region. This is facilitated by thermal fluctuations of lipids, creating localized and transient regions where membrane organization is disrupted. This nucleates peptide insertion into the membrane and pore formation. Pores are formed even at low peptide concentrations, and there is coexistence between pore peptide and membrane-bound monomers at all L/P ratios studied.
■
CONCLUSIONS In this study, we examined the interaction dynamics of a melittin derivative with giant unilamellar phospholipid bilayers (GUVs) using time-lapse fluorescence lifetime imaging microscopy. In contrast to our previous work on peptide− membrane interactions with SUVs, the kinetics of interaction between the melittin derivative and DPPC GUVs are similar across all global L/P ratios, and there is no evidence of critical behavior; lytic pores are the dominant peptide state at equilibrium at all L/P ratios. Thermal fluctuations of lipid molecules that lead to random molecular-level lipid protrusions and membrane undulations result in transient local membrane deformations of high curvature, creating the conditions for facile peptide insertion. These fluctuations are relatively undamped in GUVs because the membrane elasticity is low as a result of a low Laplace pressure. The interaction kinetics between the melittin derivative and DPPC GUVs are best fit by a two-step mechanism: free peptide → membrane-adsorbed monomer followed by membraneadsorbed monomer → lytic pore peptide. The second step is 14619
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
Article
reversible: lytic pore peptide → membrane-adsorbed monomer. The first step, which is rate-limiting, is a simple first-order, diffusion-controlled adsorption. The formation of pores after peptide adsorption is a cooperative process that requires surface migration of peptide and assembly. Once pores are formed, there is an equilibrium exchange between pore and surface monomers at all L/P ratios. This suggests that the fully inserted phase is reached even at low peptide concentrations, in contrast to what is observed with SUVs.
■
(11) Matsuzaki, K.; Yoneyama, S.; Miyajima, K. Pore Formation and Translocation of Melittin. Biophys. J. 1997, 73, 831−838. (12) Naito, A.; Nagao, T.; Norisada, K.; Mizuno, T.; Tuzi, S.; Saito, H. Conformation and Dynamics of Melittin Bound to Magnetically Oriented Lipid Bilayers by Solid-State P-31 and C-13 NMR Spectroscopy. Biophys. J. 2000, 78, 2405−2417. (13) Benachir, T.; Lafleur, M. Study of Vesicle Leakage Induced by Melittin. Biochim. Biophys. Acta 1995, 1235, 452−460. (14) Bechinger, B.; Lohner, K. Detergent-like Actions of Linear Amphipathic Cationic Antimicrobial Peptides. Biochim. Biophys. Acta 2006, 1758, 1529−1539. (15) Andersson, A.; Danielsson, J.; Graslund, A.; Maler, L. Kinetic Models for Peptide-Induced Leakage from Vesicles and Cells. Eur. Biophys. J. Biophys. Lett. 2007, 36, 621−635. (16) Dempsey, C. E. The Actions of Melittin on Membranes. Biochim. Biophys. Acta 1990, 1031, 143−161. (17) Ennaceur, S. M.; Hicks, M. R.; Pridmore, C. J.; Dafforn, T. R.; Rodger, A.; Sanderson, J. M. Peptide Adsorption to Lipid Bilayers: Slow Processes Revealed by Linear Dichroism Spectroscopy. Biophys. J. 2009, 96, 1399−1407. (18) Mazzuca, C.; Orioni, B.; Coletta, M.; Formaggio, F.; Toniolo, C.; Maulucci, G.; De Spirito, M.; Pispisa, B.; Venanzi, M.; Stella, L. Fluctuations and the Rate-Limiting Step of Peptide-Induced Membrane Leakage. Biophys. J. 2010, 99, 1791−1800. (19) Mally, M.; Majhenc, J.; Svetina, S.; Zeks, B. The Response of Giant Phospholipid Vesicles to Pore-Forming Peptide Melittin. Biochim. Biophys. Acta 2007, 1768, 1179−1189. (20) Huang, H. W. Molecular Mechanism of Antimicrobial Peptides: The Origin of Cooperativity. Biochim. Biophys. Acta 2006, 1758, 1292−1302. (21) Torrens, F.; Castellano, G.; Campos, A.; Abad, C. Negatively Cooperative Binding of Melittin to Neutral Phospholipid Vesicles. J. Mol. Struct. 2007, 834−836, 216−228. (22) Sekharam, K. M.; Bradrick, T. D.; Georghiou, S. Kinetics of Melittin Binding to Phospholipid Small Unilamellar Vesicles. Biochim. Biophys. Acta 1991, 1063, 171−174. (23) Ningsih, Z.; Hossain, M. A.; Wade, J. D.; Clayton, A. H. A.; Gee, M. L. Slow Insertion Kinetics during Interaction of a Model Antimicrobial Peptide with Unilamellar Phospholipid Vesicles. Langmuir 2012, 28, 2217−2224. (24) Lundquist, A.; Wessman, P.; Rennie, A. R.; Edwards, K. Melittin-Lipid Interaction: A Comparative Study Using Liposomes, Micelles and Bilayer Disks. Biochim. Biophys. Acta 2008, 1778, 2210− 2216. (25) Yamazakit, M.; Tamba, Y. The Single GUV Method for Probing Biomembrane Structure and Function. e-J. Surf. Sci. Nanotech. 2005, 3, 218−227. (26) Tamba, Y.; Ohba, S.; Kubota, M.; Yoshioka, H.; Yamazaki, M. Single GUV Method Reveals Interaction of Tea Catechin (2)Epigallocatechin Gallate with Lipid Membranes. Biophys. J. 2007, 92, 3178−3194. (27) Bernheimer, A. W.; Rudy, B. Interactions between Membranes and Cytolytic Peptides. Biochim. Biophys. Acta 1986, 864, 123−141. (28) Hristova, K.; Dempsey, C. E.; White, S. H. Structure, Location, and Lipid Perturbations of Melittin at the Membrane Interface. Biophys. J. 2001, 80, 801−811. (29) Rapson, A. C.; Hossain, M. A.; Wade, J. D.; Nice, E. C.; Smith, T. A.; Clayton, A. H. A.; Gee, M. L. Structural Dynamics of a Lytic Peptide Interacting with a Supported Lipid Bilayer. Biophys. J. 2011, 100, 1353−1361. (30) Moscho, A.; Orwar, O.; Chiu, D. T.; Modi, B. P.; Zare, R. N. Rapid Preparation of Giant Unilamellar Vesicles. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 11443−11447. (31) Van Munster, E. B.; Gadella, T. W. J. FLIM: A New Method to Avoid Aliasing in Frequency-Domain Fluorescence Lifetime Imaging Microscopy. J. Microsc. (Oxford, U. K.) 2004, 213, 29−38. (32) Magde, D.; Rojas, G. E.; Seybold, P. G. Solvent Dependence of the Fluorescence Lifetimes of Xanthene Dyes. Photochem. Photobiol. 1999, 70 (), 737−744.
ASSOCIATED CONTENT
S Supporting Information *
Alternative two-step kinetic models proposed to describe the experimental data, including fits to fractional fluorescence. This material is available free of charge via the Internet at http:// pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail: [email protected]. *E-mail: [email protected]. Tel: (+61)8344-3949. Fax: (+61)9347-5180. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS M.L.G. and A.H.A.C. gratefully thank the Australian Research Council for their generous financial support of this project in the form of Discovery grants DP0557718 and DP110100164.
■
ABBREVIATIONS GUV, giant unilamellar vesicle; FLIM, fluorescence lifetime imaging microscopy; L/P, lipid-to-peptide concentration ratio; SUV, small unilamellar vesicle; DPPC, 1,2-dipalmitoyl-snglycero-3-phosphochloine; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
■
REFERENCES
(1) Diamond, G.; Beckloff, N.; Weinberg, A.; Kisich, K. O. The Roles of Antimicrobial Peptides in Innate Host Defense. Curr. Pharm. Design 2009, 15, 2377−2392. (2) Zasloff, M. Antimicrobial Peptides of Multicellular Organisms. Nature 2002, 415, 389−395. (3) Tossi, A.; Sandri, L.; Giangaspero, A. Amphipathic, α-Helical Antimicrobial Peptides. Biopolymers 2000, 55, 4−30. (4) Oren, Z.; Shai, Y. Mode of Action of Linear Amphipathic αHelical Antimicrobial Peptides. Biopolymers 1998, 47, 451−463. (5) Neu, H. C. The Crisis in Antibiotic Resistance. Science 1992, 257, 1064−1073. (6) Yeaman, M. R.; Yount, N. Y. Mechanisms of Antimicrobial Peptide Action and Resistance. Pharmacol. Rev. 2003, 55, 27−55. (7) Brogden, K. A. Antimicrobial Peptides: Pore Formers or Metabolic Inhibitors in Bacteria? Nat. Rev. Microbiol. 2005, 3, 238− 250. (8) Gazzara, J. A.; Phillips, M. C.; Lund-Katz, S.; Palgunachari, M. N.; Segrest, J. P.; Anantharamaiah, G. M.; Rodrigueza, W. V.; Snow, J. W. Effect of Vesicle Size on Their Interaction with Class A Amphipathic Helical Peptides. J. Lipid Res. 1997, 38, 2147−2154. (9) Huang, H. W. Action of Antimicrobial Peptides: Two-State Model. Biochemistry 2000, 39, 8347−8352. (10) Yang, L.; Harroun, T. A.; Weiss, T. M.; Ding, L.; Huang, H. W. Barrel-Stave Model or Toroidal Model? A Case Study on Melittin Pores. Biophys. J. 2001, 81, 1475−1485. 14620
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
Article
(33) Hanley, Q. S.; Clayton, A. H. A. AB-Plot Assisted Determination of Fluorophore Mixtures in a Fluorescence Lifetime Microscope Using Spectra or Quenchers. J. Microsc. (Oxford, U. K.) 2005, 218, 62−67. (34) Clayton, A. H. A.; Hanley, Q. S.; Verveer, P. J. Graphical Representation and Multicomponent Analysis of Single-Frequency Fluorescence Lifetime Imaging Microscopy Data. J. Microsc. (Oxford, U. K.) 2004, 213, 1−5. (35) Redford, G. I.; Clegg, R. M. Polar Plot Representation for Frequency-Domain Analysis of Fluorescence Lifetimes. J. Fluoresc. 2005, 15, 805−815. (36) Digman, M. A.; Caiolfa, V. R.; Zamai, M.; Gratton, E. The Phasor Approach to Fluorescence Lifetime Imaging Analysis. Biophys. J. 2008, 94, L14−L16. (37) Gadella, T. W. J.; Jovin, T. M.; Clegg, R. M. Fluorescence Lifetime Imaging Microscopy (FLIM): Spatial Resolution of Microstructures on the Nanosecond Time Scale. Biophys. Chem. 1993, 48, 221−239. (38) Lakowicz, J. R.; Szmacinski, H.; Nowaczyk, K.; Berndt, K. W.; Johnson, M. Fluorescence Lifetime Imaging. Anal. Biochem. 1992, 202, 316−330. (39) Gee, M. L.; Burton, M.; Grevis-James, A.; Hossain, M. A.; McArthur, S.; Palombo, E. A.; Wade, J. D.; Clayton, A. H. A. Imaging the Action of Antimicrobial Peptides on Living Bacterial Cells. Sci. Rep. 2013, 3, 1557. (40) Kreutzberger, A. J.; Pokorny, A. On the Origin of Multiphasic Kinetics in Peptide Binding to Phospholipid Vesicles. J. Phys. Chem. B 2012, 116, 951−957. (41) Rakowski, A. Kinetics of the Consequential Reaction of the First order. Z. Phys. Chem. 1906, 57, 321−340. (42) Haldar, S.; Raghuraman, K.; Chattopadhyay, A. Monitoring Orientation and Dynamics of Membrane-Bound Melittin Utilizing Dansyl Fluorescence. J. Phys. Chem. B 2008, 112, 14075−14082. (43) Huang, H. W. Elasticity of Lipid Bilayer Interacting with Amphiphilic Helical Peptides. J. Phys. II 1995, 5, 1427−1431. (44) Bigay, J.; Gounon, P.; Robineau, S.; Antonny, B. Lipid Packing Sensed by ArfGAP1 Couples COPI Coat Disassembly to Membrane Bilayer Curvature. Nature 2003, 426, 563−566. (45) Ishitsuka, Y.; Pham, D. S.; Waring, A. J.; Lehrer, R. I.; Lee, K. Y. C. Insertion Selectivity of Antimicrobial Peptide Protegrin-1 into Lipid Monolayers: Effect of Head Group Electrostatics and Tail Group Packing. Biochim. Biophys. Acta 2006, 1758, 1450−1460. (46) Schwarz, G.; Zong, R. T.; Popescu, T. Kinetics of Melittin Induced Pore Formation in the Membrane of Lipid Vesicles. Biochim. Biophys. Acta 1992, 1110, 97−104. (47) Israelachvili, J. N.; Wennerstrom, H. Entropic Forces between Amphiphilic Surfaces in Liquids. J. Phys. Chem. 1992, 96, 520−531. (48) Lipowsky, R.; Grotehans, S. Renormalization of Hydration Forces by Collective Protrusion Modes. Biophys. Chem. 1994, 49, 27− 37. (49) Gordeliy, V. I.; Cherezov, V.; Teixeira, J. Strength of Thermal Undulations of Phospholipid Membranes. Phys. Rev. E 2005, 72, 061913. (50) Servuss, R. M.; Helfrich, W. Mutual Adhesion of Lecithin Membranes at Ultralow Tensions. J. Phys. (Paris) 1989, 50, 809−827. (51) Helfrich, W. Elastic Properties of Lipid Bilayers - Theory and Possible Explanations. Z. Naturforsch. 1973, C28, 693−703.
14621
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Long-Time-Scale Interaction Dynamics between a Model Antimicrobial Peptide and Giant Unilamellar Vesicles Matthew G. Burton,† Qi M. Huang,† Mohammed A. Hossain,†,‡ John D. Wade,†,‡ Andrew H. A. Clayton,*,§ and Michelle L. Gee*,† †
School of Chemistry and ‡Florey Department of Neuroscience and Mental Health, Centre for Neuroscience Research Chemistry, University of Melbourne, Parkville, Victoria 3010, Australia § Centre for Micro-Photonics, Swinburne University of Technology, Hawthorn, Victoria 3122, Australia S Supporting Information *
ABSTRACT: The interaction dynamics between a lytic peptide and a biomembrane was studied using time-lapse fluorescence lifetime imaging microscopy. The model membrane was 1,2-dipalmitoyl-sn-glycero-3phosphochloine giant unilamellar vesicles (GUVs), and the peptide was the K14 derivative of melittin, to which the polarity-sensitive fluorescent probe AlexaFluor 430 was grafted. The interaction of the peptide with the GUVs resulted in a progressive quenching of the fluorescence lifetime over a period of minutes. From previous photophysics characterization of the peptide, we were able to deconvolve the contribution of three distinct peptide states to the lifetime trajectory and use this data to develop a kinetics model for the interaction process. It was found that the peptide− membrane interaction was well described by a two-step mechanism: peptide monomer adsorption followed by membrane surface migration, assembly, and insertion to form membrane pores. There was an equilibrium exchange between pore and surface monomers at all lipid/peptide (L/P) concentration ratios, suggesting that the fully inserted phase was reached, even at low peptide concentrations. In contrast to previous studies, there was no evidence of critical behavior; irrespective of L/P ratio, lytic pores were the dominant peptide state at equilibrium and were formed even at very low peptide concentrations. We suggest that this behavior is seen in GUVs because their low curvature means low Laplace pressure. Membrane elasticity is therefore relatively ineffective at damping the thermal fluctuations of lipid molecules that lead to random molecular-level lipid protrusions and membrane undulations. The transient local membrane deformations that result from these thermal fluctuations create the conditions necessary for facile peptide insertion.
■
INTRODUCTION
One approach to probing the mechanism of peptide− membrane interactions is by studying their interaction kinetics. It has been shown that the initial binding of antimicrobial peptides to a membrane occurs on the order of milliseconds.15,16 Longer-time-scale events have been observed, including the reorientation and conformational change of the peptide upon insertion into the membrane,17 exchange between membrane-adsorbed and free peptide,18 and osmotic swelling of a vesicle due to peptide−lipid interactions.19 Cooperativity between peptide monomers has been suggested to be a key step in the mechanism toward membrane pore formation.20−22 More recently, slow insertion kinetics and the formation of peptide oligomers that are transient states in the mechanisms toward pore formation have been observed.23 Much of the work to date on peptide−membrane interactions has utilized model membranes, predominantly small unilamellar vesicles (SUVs), large unilamellar vesicles, or
Antimicrobial peptides are a diverse array of small molecular species found naturally in a wide variety of organisms, including insects, amphibians, humans, and plants.1−3 Many of these peptides form part of an innate host defense system that can preferentially target and kill bacterial cells.4 These peptides have gained significant attention as the basis for the design of new therapeutics to combat the increasing threat of antimicrobial resistance.5,6 There are many factors that influence peptide− membrane and lipid−peptide interactions. These include the lipid membrane composition, degree of lipid hydration/fluidity, membrane curvature, lipid headgroup charge, and lipid-topeptide concentration ratio.3,6−9 Despite extensive investigations on a variety of antimicrobial peptides spanning several decades, a precise mechanistic understanding of their mode of action remains elusive. Two of the more popular models of peptide−membrane interaction are based on either the formation of pore structures within the membrane9−11 or the perturbation and/or solubilization of the membrane in a detergentlike manner.12−14 © 2013 American Chemical Society
Received: August 9, 2013 Revised: October 29, 2013 Published: October 29, 2013 14613
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
Article
methanol mixture (9.8:1). Filtered HEPES buffer was then carefully added to the organic phase, with the organic phase subsequently removed via rotary evaporation for 2 to 3 min at 40 rpm, 40 °C, 100 mbar minimum pressure. The resulting GUV suspension was centrifuged at 15 300g for 10 min to remove multilamellar vesicles and lipid debris. The supernatant was collected and stored in glass vials at room temperature (approximately 23 °C) for up to 2 weeks. Frequency Domain Fluorescence Lifetime Imaging Microscopy Measurements. All fluorescence lifetime imaging microscopy (FLIM) experiments were performed in duplicate using a frequencydomain Lambert Instruments fluorescence lifetime attachment (Lambert Instruments, Leutingwolde, The Netherlands) interfaced with an inverted microscope (TE2000U, Nikon Inc., Japan). GUV samples were observed in situ through a 100× NA 1.2 oil objective (Nikon Plan-Fluor, Nikon Inc., Japan). The fluorescence excitation source was a 470 nm LED with a sinusoidal modulation frequency of 40 MHz. From the resulting phase image stacks (12 phase images per stack), phase and modulation lifetimes, τϕ and τM, respectively, were determined. Note that in frequency domain measurements two fluorescence lifetimes are determined: the phase lifetime, τϕ, is calculated from the phase lag of the fluorescence signal relative to the excitation signal, and the modulation lifetime, τM, is calculated from the amplitude modulation of the fluorescence signal relative to the excitation signal. Photobleaching was corrected by varying the phase angle of the excitation source over 0−360° during acquisition.31 The reference used for all lifetime determinations was rhodamine 6G (lifetime τ = 4.1 ns).32 An experimental sample was prepared by placing a 100 μL aliquot of a GUV solution into a glass observation cell. GUVs were allowed to settle for 45−60 min, after which transmitted light microscopy was used to identify a suitable GUV of interest. The total lipid concentration was estimated from the number and size distribution of GUVs in the field of view over a number of different areas across the entire sample. A sufficient amount of the melittin derivative solution was then injected into the sample solution for global L/P ratios of 100:1, 50:1, and 27:1 (total peptide concentration 0.27−1 μM). Injections were made gently and as close to the point of interest as possible to minimize diffusion effects of the fluorescent peptide. Fluorescence lifetime image stacks were recorded every 1−4 min for the first 15 min after the introduction of peptide and then every 15 min after that for up to 3 h. Transmitted light microscopy images were recorded for every phase image stack. The fluorescence lifetime measurements were represented in two different ways. First, to get an indication of trends, the net fluorescence lifetime was taken and both phase (τϕ) and modulation lifetimes (τM) were plotted. The second approach utilized the AB plot, also referred to as a phasor or polar plot, to display the lifetime experiments graphically.33−37 In such a plot, an experiment is represented by a point in 2D space defined by x = m cos ϕ and y = m sin ϕ, where ϕ is the phase and m is the modulation of the fluorescence signal. A series of lifetime experiments plotted as a series of points and joined by lines is called a trajectory. This graphical approach has the advantage that the type of fluorescence decay (simple, complex, excited-state reaction/solvent relaxation) and the complexity of the system trajectory (binary or more complex) can be deduced visually without further analysis. The phasor of single-exponential-decaying fluorophores lies on a semicircle described by m = cos ϕ that intersects with points (0, 0), (0.5, 0.5,) and (1, 0). Phasors from heterogeneous fluorescence decays lie inside the semicircle such that m < cos ϕ. Excited-state reactions including solvent relaxation have phasors that lie outside the semicircle: m > cos ϕ. The linear combination of two phasors is described by a linear trajectory in AB space, whereas the mixing among three or more species is nonlinear. Note that in frequency domain measurements two lifetimes are determined: the phase lifetime, τϕ, is determined from the phase lag of the fluorescence signal relative to the excitation signal, and the modulation lifetime, τM, is determined from the amplitude modulation of the fluorescence signal relative to the excitation signal.
supported lipid bilayers. In light of the fact that membrane curvature appears to affect the observed behavior of a peptide,8,24 it is imperative that studies employ a model system that is geometrically similar to that of a living bacterial cell. Furthermore, indirect analysis techniques on small and large unilamellar vesicles report on the average behavior of the entire ensemble, which may mask important mechanistic information occurring specifically at the membrane. Giant unilamellar vesicles (GUVs) have the potential to overcome these issues since because of their size (on the order of micrometers) they can be viewed microscopically, allowing the direct extraction of qualitative and quantitative information from regions of interest.25,26 In the present work, we used time-lapse frequency domain fluorescence lifetime imaging microscopy (FLIM) to monitor long-time-scale (minutes to hours) peptide−membrane interaction kinetics. Our model peptide is a derivative of melittin, a 26 amino acid cationic peptide that is predominantly random coil in solution27 but rapidly adopts an α-helical conformation upon association with phospholipid bilayers.16,28 The melittin derivative is based on the parent peptide but with a proline 14− lysine 14 substitution. The polarity-sensitive fluorescence probe AlexaFluor 430 grafted at lysine 14 is near the center of the peptide and is able to sample local micropolarity near the center of the peptide helix. This fluorescently tagged melittin derivative is α-helical in membranes and more effective at membrane lysis than its parent.29 This allows us to monitor directly the interaction kinetics of the peptide based upon the photophysical behavior of the fluorescent tag. We have monitored its interaction with 1,2-dipalmitoyl-sn-glycero-3phosphatidylcholine (DPPC) giant unilamellar vesicles (GUVs). Previously, we observed complex kinetics of the melittin derivative interacting with small unilamllar vesicles (SUVs) during which several peptide states are formed during helix insertion into the vesicle membrane that are intermediate in lytic pore formation.23 We show here that the kinetics of the peptide’s interaction with GUVs are quite different from those seen with SUVs, suggesting a different interaction mechanism. This has important implications for peptide−cell interactions.
■
MATERIALS AND METHODS
Materials. 1,2-Dipalmitoyl-sn-glycero-3-phosphatidylchloine (DPPC) was purchased from Avanti Polar Lipids (Alabaster, AL). UV/vis spectroscopy grade chloroform and methanol were purchased from Merck (Darmstadt, Germany). 4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES, 99% purity) was purchased from Acros Organics (Geel, Belgium). The melittin derivative, melittin K14 AlexaFluor 430, was synthesized on a Fmoc-PAL-EG-PS resin, as detailed previously.29 The AlexaFluor 430 fluorescent label was purchased from Molecular Probes (Carlsbad, CA). All materials were used without further purification. Milli-Q water with a resistivity of 18.2 MΩ was generated by a Milli-Q ultrapure water system or a MilliQ academic system (Millipore Corp., Bedford, CT) and was used in the preparation of all aqueous solutions. HEPES buffer solutions (10 mM HEPES, pH 7.40 ± 0.05) filtered through 0.22 μm hydrophilic poly(1,1,2,2-tetrafluoroethylene) (Teflon) membranes (Millipore Corp., Bedford, CT) were used in the preparation of all peptide and vesicle solutions. The pH was adjusted using small volumes of concentrated sodium hydroxide (8 M). The melittin derivative solutions were prepared at concentrations of 5−20 μM and were stored at 4 °C for no longer than 2 weeks. Preparation of Giant Unilamellar Vesicles. GUVs composed of pure DPPC were prepared using the solvent evaporation method.30 Briefly, a concentrated solution of DPPC powder dissolved in spectroscopy grade chloroform (0.1 M) was added to a chloroform/ 14614
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
■
Article
RESULTS Time-Lapse Fluorescence Lifetime Measurements of Peptide−Membrane Interactions. Time-lapse frequency domain fluorescence lifetime imaging microscopy (FLIM) was performed to monitor the evolution of the lipid−peptide interaction between our fluorescently tagged melittin derivative and DPPC GUVs as a function of time for three lipid/peptide (L/P) concentration ratios. A representative set of FLIM images is shown in Figure 1, together with the corresponding
50:1, and 27:1), in agreement with prior observations that high peptide concentrations (i.e., low global L/P ratios ( k2 at all times t. This, however, resulted in very poor fits to the experimental results (data not shown). We find here for the interaction of peptide with GUVs that the fastest step in the mechanism is the aggregation of peptide on the
step 1: k1
free interfacial peptide → membrane‐adsorbed monomer
step 2, forward: k2
membrane‐adsorbed monomer → lytic pore peptide
step 2, reverse: k3
lytic pore peptide → membrane‐adsorbed monomer
In this reaction scheme, which represents the process followed by individual peptide monomers during peptide− membrane interaction, k1 is the rate constant for the adsorption of a peptide monomer onto the membrane to form a membrane-adsorbed monomer. This is followed by peptide insertion/pore formation, which occurs at a rate determined by the rate constant k2. In the proposed reaction scheme, we allow for the reversibility of peptide aggregation and pore formation at a rate determined by k3. Note that fluorescence lifetime measurements were made on the vesicle of interest to exclude bulk solution. The term “free interfacial peptide” therefore means peptide in the interfacial region but not bound to the 14617
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
Article
Table 1. Rate Constants for Each Step in the Peptide− Membrane Interaction Mechanisma L/P ratio
k1 (min−1)
k2 (min−1)
k3 (min−1)
100:1 50:1 27:1
0.05 ± 0.02 0.12 ± 0.02 0.20 ± 0.04
1.8 ± 0.4 1.03 ± 0.2 0.49 ± 0.06
0.13 ± 0.02 0.08 ± 0.02 0.10 ± 0.02
a
k1 is the rate constant for the adsorption of free interfacial peptide onto the membrane. Membrane-adsorbed monomers then insert into the membrane to form membrane pores at a rate determined by rate constant k2. The reversibility of pore formation occurs at a rate determined by k3. Rate constants are given for each global lipid/ peptide (L/P) concentration ratio studied and were obtained by fitting time-lapse fluorescence lifetime measurements to eqs 6−8. The errors in the rate constants are estimated to be around 20%. This is a conservative estimate that takes into account experimental errors and errors in model fitting.
membrane and insertion to form lytic pores. It is interesting, however, that this step becomes slower the higher the relative peptide concentration. Another interesting result is that the reverse process (i.e., peptide aggregate dissociation) occurs at an intermediate rate that is independent of the L/P concentration ratio. We discuss the implications of this in the Discussion section below. It is a little more intuitive to discuss the kinetics of peptide− membrane interaction in terms of mole fractions rather than fractional fluorescence. We have therefore taken the rate constants in Table 1 obtained from fitting the kinetics model to the fluorescence time-lapse data (Figure 4) and back calculated mole fractions for each peptide state using eqs 6−8. The resulting data are shown in Figure 5 and show that at all global L/P ratios membrane pores are rapidly formed and that the pore peptide is the dominant peptide state.
Figure 5. Time-lapse fractional concentrations during peptide− membrane interaction from each identified peptide state: free interfacial peptide (blue diamonds), membrane-adsorbed peptide monomers (green triangles), and lytic pore peptides (red circles) derived from the kinetics modeling defined by eqs 6−8. Data is shown for global L/P ratios: (a) 100:1, (b) 50:1, and (c) 27:1. High fractional concentrations of pore peptide are observed at all L/P ratios studied, even at peptide concentrations far below that expected for pore formation. Errors in fractional concentration are 12% as per the fractional fluorescence data of Figure 4
■
DISCUSSION We have examined the interaction dynamics of a melittin derivative with giant unilamellar phospholipid vesicles (GUVs) using time-lapse fluorescence lifetime imaging microscopy (FLIM). The peptide derivative has a proline-14 to lysine substitution onto which polarity-sensitive fluorescent probe AlexaFluor 430 has been grafted. Fluorescent tagging at this point on the peptide chain allows an examination of the microenvironment local to the center of the peptide helix. As described above, we have found that the interaction between the melittin derivative and DPPC GUVs is best fit by a modified two-step mechanism. Table 1 contains the rate constants for each step at different L/P ratios, extracted by fitting this model to the time-lapse fluorescence data of Figure 4. The first step, free peptide → membrane-adsorbed monomer with rate constant k1, is rate-limiting but increases in rate as the concentration of peptide increases. This is expected for a simple pseudo-first-order adsorption process that is diffusion-controlled. Recall, however, that membrane-adsorbed monomer has a fluorescence lifetime consistent with the fluorescent probe on the peptide sampling the headgroup region of the membrane, implying that the peptide is oriented with its long axis parallel to the membrane surface. This is consistent with previous studies looking at fluorescently labeled native melittin interacting with lipid membranes.42 This means that there must be surface reorientation of peptide after initial nonspecific adsorption until the AlexaFluor 430 probe is positioned to sample the headgroup region of the membrane. The first-order behavior of this step in the mechanism implies that the
adsorption and surface reorientation of peptide are unhindered and do not involve any cooperativity between neighboring peptide molecules. In contrast, the second step, membrane-adsorbed monomer → lytic pore peptide, has a rate (k2) that suggests that cooperativity is important in this step of the process. This is the fast step in the interaction mechanism, highlighting the ease of peptide insertion into GUVs. But peptide insertion and pore formation become slower as the peptide concentration increases. The assembly of membrane-adsorbed peptide monomers into pore peptide relies on the surface migration and reorientation of peptide molecules. This becomes increasingly hindered as the surface becomes more crowded, presumably through steric and electrostatic repulsion between peptides. The reverse of the second step, lytic pore peptide → membrane-adsorbed monomer, is slow (k3) and independent of peptide concentration within experimental error. This implies that pores do not disassemble once formed but that any migration of pore peptide to the surface is part of an equilibrium exchange or coexistence between pore peptides and peptide monomers. This is consistent with Huang et al.,9 who discussed the critical concentration at which there is a 14618
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
Article
coexistence region where a fraction of peptide is inserted and the rest remains on the surface, with the inserted fraction increasing as the peptide concentration increases. The observation that there is an equilibrium exchange between pore and surface monomers means that we have reached the fully inserted phase even at low peptide concentrations. This is consistent with the measured fractional concentrations of peptide states (Figure 5). It has been shown previously23 that the interaction of the melittin mutant with DPPC SUVs has long-time-scale (30 min−2 h) complex kinetics. Pore formation depends critically on the L/P ratio and does not occur until the L/P ratio is 20:1 (relatively high peptide concentration), consistent with other studies9 and predicted by Huang.43 For the melittin derivative interacting with DPPC SUVs, the mechanism of peptide− membrane interaction involves initial rapid peptide adsorption followed by the formation of several oligomeric peptide states that are inserted into the membrane and sample the alkyl region of the bilayer. These peptide states are intermediates to the slowly forming lytic pores. In contrast to our previous work on peptide−membrane interactions with SUVs,23 we observed that the kinetics of interaction between the melittin derivative and DPPC GUVs are similar across all global L/P ratios (Figure 5) and there is no evidence of critical behavior within the observed L/P ratio range. In fact, lytic pores are the dominant peptide state at equilibrium for all peptide concentrations, even at those far lower than the critical concentration for pore formation found for SUVs (i.e., L/P ratio of around 25:1). This is a surprising result with criticality in lipid−peptide interactions reported to occur around a lipid/peptide ratio of 30:1, consistent with the predictions of Huang9,43 and reported measurements (ref 23 and references therein). This suggests that a very different mechanism of interaction is occurring in GUVs. The absence of criticality is initially surprising because a lower-curvature membrane should lead to closer packing of lipid molecules, hindering peptide insertion.14,44,45 Schwarz et al.46 have shown that the interaction of lytic peptides with large unilamellar vesicles resulted in fewer pores per mole of lipid compared to their interaction with SUVs. On this basis, we might have expected critical behavior with pore formation at higher peptide concentration than seen with SUVs of the same lipid. It is important, however, to consider that a lipid membrane is a fluid, thermally mobile system. Thermal fluctuations of lipid molecules lead to transient molecular-level lipid protrusions,47,48 and the collective thermal fluctuations of lipid molecules result in membrane ripples or undulations.49,50 On the time and length scales of these fluctuations, the membrane is locally deformed, altering the headgroup spacing and creating the conditions for facile peptide insertion. This is illustrated schematically in Figure 6. It is likely to occur in synergy with membrane point defects51 and localized membrane thinning created by the peptide−lipid interaction.43 We observe pore formation in DPPC GUVs even at low peptide concentrations (high L/P ratios) because the low vesicle Laplace pressure (due to low global curvature) results in low membrane elasticity.51 Membrane thermal fluctuations are therefore not damped, relatively speaking. In high-curvature SUVs, the high Laplace pressure renders the membrane more elastic, damping membrane thermal fluctuations and thus hindering peptide insertion.
Figure 6. Visual representation of the two-step peptide−membrane interaction kinetics mechanism for peptides interacting with GUVs, as defined by the reaction scheme. Phospholipids are shown here with a single acyl chain tail for clarity. Peptide free in solution binds to the surface of the membrane and undergoes subsequent reconfiguration such that the fluorescent probe samples the membrane headgroup region. This is facilitated by thermal fluctuations of lipids, creating localized and transient regions where membrane organization is disrupted. This nucleates peptide insertion into the membrane and pore formation. Pores are formed even at low peptide concentrations, and there is coexistence between pore peptide and membrane-bound monomers at all L/P ratios studied.
■
CONCLUSIONS In this study, we examined the interaction dynamics of a melittin derivative with giant unilamellar phospholipid bilayers (GUVs) using time-lapse fluorescence lifetime imaging microscopy. In contrast to our previous work on peptide− membrane interactions with SUVs, the kinetics of interaction between the melittin derivative and DPPC GUVs are similar across all global L/P ratios, and there is no evidence of critical behavior; lytic pores are the dominant peptide state at equilibrium at all L/P ratios. Thermal fluctuations of lipid molecules that lead to random molecular-level lipid protrusions and membrane undulations result in transient local membrane deformations of high curvature, creating the conditions for facile peptide insertion. These fluctuations are relatively undamped in GUVs because the membrane elasticity is low as a result of a low Laplace pressure. The interaction kinetics between the melittin derivative and DPPC GUVs are best fit by a two-step mechanism: free peptide → membrane-adsorbed monomer followed by membraneadsorbed monomer → lytic pore peptide. The second step is 14619
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
Article
reversible: lytic pore peptide → membrane-adsorbed monomer. The first step, which is rate-limiting, is a simple first-order, diffusion-controlled adsorption. The formation of pores after peptide adsorption is a cooperative process that requires surface migration of peptide and assembly. Once pores are formed, there is an equilibrium exchange between pore and surface monomers at all L/P ratios. This suggests that the fully inserted phase is reached even at low peptide concentrations, in contrast to what is observed with SUVs.
■
(11) Matsuzaki, K.; Yoneyama, S.; Miyajima, K. Pore Formation and Translocation of Melittin. Biophys. J. 1997, 73, 831−838. (12) Naito, A.; Nagao, T.; Norisada, K.; Mizuno, T.; Tuzi, S.; Saito, H. Conformation and Dynamics of Melittin Bound to Magnetically Oriented Lipid Bilayers by Solid-State P-31 and C-13 NMR Spectroscopy. Biophys. J. 2000, 78, 2405−2417. (13) Benachir, T.; Lafleur, M. Study of Vesicle Leakage Induced by Melittin. Biochim. Biophys. Acta 1995, 1235, 452−460. (14) Bechinger, B.; Lohner, K. Detergent-like Actions of Linear Amphipathic Cationic Antimicrobial Peptides. Biochim. Biophys. Acta 2006, 1758, 1529−1539. (15) Andersson, A.; Danielsson, J.; Graslund, A.; Maler, L. Kinetic Models for Peptide-Induced Leakage from Vesicles and Cells. Eur. Biophys. J. Biophys. Lett. 2007, 36, 621−635. (16) Dempsey, C. E. The Actions of Melittin on Membranes. Biochim. Biophys. Acta 1990, 1031, 143−161. (17) Ennaceur, S. M.; Hicks, M. R.; Pridmore, C. J.; Dafforn, T. R.; Rodger, A.; Sanderson, J. M. Peptide Adsorption to Lipid Bilayers: Slow Processes Revealed by Linear Dichroism Spectroscopy. Biophys. J. 2009, 96, 1399−1407. (18) Mazzuca, C.; Orioni, B.; Coletta, M.; Formaggio, F.; Toniolo, C.; Maulucci, G.; De Spirito, M.; Pispisa, B.; Venanzi, M.; Stella, L. Fluctuations and the Rate-Limiting Step of Peptide-Induced Membrane Leakage. Biophys. J. 2010, 99, 1791−1800. (19) Mally, M.; Majhenc, J.; Svetina, S.; Zeks, B. The Response of Giant Phospholipid Vesicles to Pore-Forming Peptide Melittin. Biochim. Biophys. Acta 2007, 1768, 1179−1189. (20) Huang, H. W. Molecular Mechanism of Antimicrobial Peptides: The Origin of Cooperativity. Biochim. Biophys. Acta 2006, 1758, 1292−1302. (21) Torrens, F.; Castellano, G.; Campos, A.; Abad, C. Negatively Cooperative Binding of Melittin to Neutral Phospholipid Vesicles. J. Mol. Struct. 2007, 834−836, 216−228. (22) Sekharam, K. M.; Bradrick, T. D.; Georghiou, S. Kinetics of Melittin Binding to Phospholipid Small Unilamellar Vesicles. Biochim. Biophys. Acta 1991, 1063, 171−174. (23) Ningsih, Z.; Hossain, M. A.; Wade, J. D.; Clayton, A. H. A.; Gee, M. L. Slow Insertion Kinetics during Interaction of a Model Antimicrobial Peptide with Unilamellar Phospholipid Vesicles. Langmuir 2012, 28, 2217−2224. (24) Lundquist, A.; Wessman, P.; Rennie, A. R.; Edwards, K. Melittin-Lipid Interaction: A Comparative Study Using Liposomes, Micelles and Bilayer Disks. Biochim. Biophys. Acta 2008, 1778, 2210− 2216. (25) Yamazakit, M.; Tamba, Y. The Single GUV Method for Probing Biomembrane Structure and Function. e-J. Surf. Sci. Nanotech. 2005, 3, 218−227. (26) Tamba, Y.; Ohba, S.; Kubota, M.; Yoshioka, H.; Yamazaki, M. Single GUV Method Reveals Interaction of Tea Catechin (2)Epigallocatechin Gallate with Lipid Membranes. Biophys. J. 2007, 92, 3178−3194. (27) Bernheimer, A. W.; Rudy, B. Interactions between Membranes and Cytolytic Peptides. Biochim. Biophys. Acta 1986, 864, 123−141. (28) Hristova, K.; Dempsey, C. E.; White, S. H. Structure, Location, and Lipid Perturbations of Melittin at the Membrane Interface. Biophys. J. 2001, 80, 801−811. (29) Rapson, A. C.; Hossain, M. A.; Wade, J. D.; Nice, E. C.; Smith, T. A.; Clayton, A. H. A.; Gee, M. L. Structural Dynamics of a Lytic Peptide Interacting with a Supported Lipid Bilayer. Biophys. J. 2011, 100, 1353−1361. (30) Moscho, A.; Orwar, O.; Chiu, D. T.; Modi, B. P.; Zare, R. N. Rapid Preparation of Giant Unilamellar Vesicles. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 11443−11447. (31) Van Munster, E. B.; Gadella, T. W. J. FLIM: A New Method to Avoid Aliasing in Frequency-Domain Fluorescence Lifetime Imaging Microscopy. J. Microsc. (Oxford, U. K.) 2004, 213, 29−38. (32) Magde, D.; Rojas, G. E.; Seybold, P. G. Solvent Dependence of the Fluorescence Lifetimes of Xanthene Dyes. Photochem. Photobiol. 1999, 70 (), 737−744.
ASSOCIATED CONTENT
S Supporting Information *
Alternative two-step kinetic models proposed to describe the experimental data, including fits to fractional fluorescence. This material is available free of charge via the Internet at http:// pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail: [email protected]. *E-mail: [email protected]. Tel: (+61)8344-3949. Fax: (+61)9347-5180. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS M.L.G. and A.H.A.C. gratefully thank the Australian Research Council for their generous financial support of this project in the form of Discovery grants DP0557718 and DP110100164.
■
ABBREVIATIONS GUV, giant unilamellar vesicle; FLIM, fluorescence lifetime imaging microscopy; L/P, lipid-to-peptide concentration ratio; SUV, small unilamellar vesicle; DPPC, 1,2-dipalmitoyl-snglycero-3-phosphochloine; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
■
REFERENCES
(1) Diamond, G.; Beckloff, N.; Weinberg, A.; Kisich, K. O. The Roles of Antimicrobial Peptides in Innate Host Defense. Curr. Pharm. Design 2009, 15, 2377−2392. (2) Zasloff, M. Antimicrobial Peptides of Multicellular Organisms. Nature 2002, 415, 389−395. (3) Tossi, A.; Sandri, L.; Giangaspero, A. Amphipathic, α-Helical Antimicrobial Peptides. Biopolymers 2000, 55, 4−30. (4) Oren, Z.; Shai, Y. Mode of Action of Linear Amphipathic αHelical Antimicrobial Peptides. Biopolymers 1998, 47, 451−463. (5) Neu, H. C. The Crisis in Antibiotic Resistance. Science 1992, 257, 1064−1073. (6) Yeaman, M. R.; Yount, N. Y. Mechanisms of Antimicrobial Peptide Action and Resistance. Pharmacol. Rev. 2003, 55, 27−55. (7) Brogden, K. A. Antimicrobial Peptides: Pore Formers or Metabolic Inhibitors in Bacteria? Nat. Rev. Microbiol. 2005, 3, 238− 250. (8) Gazzara, J. A.; Phillips, M. C.; Lund-Katz, S.; Palgunachari, M. N.; Segrest, J. P.; Anantharamaiah, G. M.; Rodrigueza, W. V.; Snow, J. W. Effect of Vesicle Size on Their Interaction with Class A Amphipathic Helical Peptides. J. Lipid Res. 1997, 38, 2147−2154. (9) Huang, H. W. Action of Antimicrobial Peptides: Two-State Model. Biochemistry 2000, 39, 8347−8352. (10) Yang, L.; Harroun, T. A.; Weiss, T. M.; Ding, L.; Huang, H. W. Barrel-Stave Model or Toroidal Model? A Case Study on Melittin Pores. Biophys. J. 2001, 81, 1475−1485. 14620
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
Langmuir
Article
(33) Hanley, Q. S.; Clayton, A. H. A. AB-Plot Assisted Determination of Fluorophore Mixtures in a Fluorescence Lifetime Microscope Using Spectra or Quenchers. J. Microsc. (Oxford, U. K.) 2005, 218, 62−67. (34) Clayton, A. H. A.; Hanley, Q. S.; Verveer, P. J. Graphical Representation and Multicomponent Analysis of Single-Frequency Fluorescence Lifetime Imaging Microscopy Data. J. Microsc. (Oxford, U. K.) 2004, 213, 1−5. (35) Redford, G. I.; Clegg, R. M. Polar Plot Representation for Frequency-Domain Analysis of Fluorescence Lifetimes. J. Fluoresc. 2005, 15, 805−815. (36) Digman, M. A.; Caiolfa, V. R.; Zamai, M.; Gratton, E. The Phasor Approach to Fluorescence Lifetime Imaging Analysis. Biophys. J. 2008, 94, L14−L16. (37) Gadella, T. W. J.; Jovin, T. M.; Clegg, R. M. Fluorescence Lifetime Imaging Microscopy (FLIM): Spatial Resolution of Microstructures on the Nanosecond Time Scale. Biophys. Chem. 1993, 48, 221−239. (38) Lakowicz, J. R.; Szmacinski, H.; Nowaczyk, K.; Berndt, K. W.; Johnson, M. Fluorescence Lifetime Imaging. Anal. Biochem. 1992, 202, 316−330. (39) Gee, M. L.; Burton, M.; Grevis-James, A.; Hossain, M. A.; McArthur, S.; Palombo, E. A.; Wade, J. D.; Clayton, A. H. A. Imaging the Action of Antimicrobial Peptides on Living Bacterial Cells. Sci. Rep. 2013, 3, 1557. (40) Kreutzberger, A. J.; Pokorny, A. On the Origin of Multiphasic Kinetics in Peptide Binding to Phospholipid Vesicles. J. Phys. Chem. B 2012, 116, 951−957. (41) Rakowski, A. Kinetics of the Consequential Reaction of the First order. Z. Phys. Chem. 1906, 57, 321−340. (42) Haldar, S.; Raghuraman, K.; Chattopadhyay, A. Monitoring Orientation and Dynamics of Membrane-Bound Melittin Utilizing Dansyl Fluorescence. J. Phys. Chem. B 2008, 112, 14075−14082. (43) Huang, H. W. Elasticity of Lipid Bilayer Interacting with Amphiphilic Helical Peptides. J. Phys. II 1995, 5, 1427−1431. (44) Bigay, J.; Gounon, P.; Robineau, S.; Antonny, B. Lipid Packing Sensed by ArfGAP1 Couples COPI Coat Disassembly to Membrane Bilayer Curvature. Nature 2003, 426, 563−566. (45) Ishitsuka, Y.; Pham, D. S.; Waring, A. J.; Lehrer, R. I.; Lee, K. Y. C. Insertion Selectivity of Antimicrobial Peptide Protegrin-1 into Lipid Monolayers: Effect of Head Group Electrostatics and Tail Group Packing. Biochim. Biophys. Acta 2006, 1758, 1450−1460. (46) Schwarz, G.; Zong, R. T.; Popescu, T. Kinetics of Melittin Induced Pore Formation in the Membrane of Lipid Vesicles. Biochim. Biophys. Acta 1992, 1110, 97−104. (47) Israelachvili, J. N.; Wennerstrom, H. Entropic Forces between Amphiphilic Surfaces in Liquids. J. Phys. Chem. 1992, 96, 520−531. (48) Lipowsky, R.; Grotehans, S. Renormalization of Hydration Forces by Collective Protrusion Modes. Biophys. Chem. 1994, 49, 27− 37. (49) Gordeliy, V. I.; Cherezov, V.; Teixeira, J. Strength of Thermal Undulations of Phospholipid Membranes. Phys. Rev. E 2005, 72, 061913. (50) Servuss, R. M.; Helfrich, W. Mutual Adhesion of Lecithin Membranes at Ultralow Tensions. J. Phys. (Paris) 1989, 50, 809−827. (51) Helfrich, W. Elastic Properties of Lipid Bilayers - Theory and Possible Explanations. Z. Naturforsch. 1973, C28, 693−703.
14621
dx.doi.org/10.1021/la403083m | Langmuir 2013, 29, 14613−14621
