Conformations and dynamic transitions of a melittin derivative that

Subscriber access provided by Kaohsiung Medical University

Biological and Environmental Phenomena at the Interface

Conformations and dynamic transitions of a melittin derivative that forms macromolecule-sized pores in lipid bilayers Anna E. Pittman, Brendan P. Marsh, and Gavin M King Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b00804 • Publication Date (Web): 22 Jun 2018 Downloaded from http://pubs.acs.org on June 24, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.

In preparation for Langmuir (2017)

1

2

Conformations and dynamic transitions of a

3

melittin derivative that forms macromolecule-

4

sized pores in lipid bilayers

5

Anna E. Pittman1, Brendan P. Marsh1, and Gavin M. King1,2,*

6 7 1

8

65211, USA

9 10

Department of Physics and Astronomy, University of Missouri-Columbia, Columbia, MO

2

Department of Biochemistry, University of Missouri-Columbia, Columbia, MO 65211, USA

11 12 13

*Corresponding author, email: [email protected]

14

15

16

17

18

1 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al.

Page 2 of 30


19

20

ABSTRACT

21

Systematically evolved from the primary active component of bee venom, MelP5 is a lipophilic

22

peptide with important physical properties that differ from wild type melittin, including the

23

ability to create large equilibrium pores in lipid bilayers at low peptide to lipid ratios. Self-

24

assembly into stable membrane spanning pores makes MelP5 a promising candidate for future

25

applications in the pharmaceutical arena. Despite significant interest, little is known about the

26

mechanism by which MelP5 remodels the lipid bilayer upon binding. We demonstrate by direct

27

atomic force microscope imaging of supported lipid bilayers in solution that MelP5 remodels 1-

28

palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) in one of two ways. It either creates

29

highly localized voids in the bilayer, or diffuse non-localized thinning. Thinning of the bilayer

30

was measured to be 3.0 ± 1.4 Å (mean +/- standard deviation) below the surface of the upper

31

leaflet of the bilayer. Pores, defined as highly localized voids in the bilayer, exhibited several

32

sizes. Approximately 20% of pores exhibited large footprint areas (47 ± 20 nm2) which appear

33

capable of passing bulky macromolecules. The peptide-effected bilayer was observed to

34

reversibly exchange between membrane-thinned and pore states in an apparent dynamic

35

equilibrium. Analysis of time-lapsed images suggested upper and lower bounds (0.2 < τ < 180 s)

36

on the characteristic time scale of transitions between the membrane-thinned and pore states.

37

Moreover, pores were found to colocalize with membrane-thinned regions, a novel observation

38

that is consistent with the notion of cooperativity among membrane-bound peptides when

39

forming pores.

40 41 2 ACS Paragon Plus Environment

Page 3 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.


42 43

INTRODUCTION

44

Polypeptide chains that bind to and subsequently form pores in membranes have garnered

45

significant attention in fields ranging from fundamental biophysics to biotechnology.1 Such pore-

46

forming peptides have significant promise in medicine as direct antimicrobial agents, as well as

47

for drug delivery vehicles, including for targeted cancer therapy.2-4 Notwithstanding their

48

potential, the complex modes of action and dynamic structure of these lipophilic peptides remain

49

poorly understood. To develop more effective therapies and to inform development of novel

50

pore-forming peptides, a more quantitative and mechanistic understanding of peptides interacting

51

with near-native membranes is needed.5-7

52

Pore formation by lipophilic peptides has been investigated by several methods, including

53

circular dichroism, lamellar x-ray diffraction, and surface-induced fluorescence.8,9 A model

54

system for these studies is the melittin peptide, a chain of 26 amino acids that is the primary

55

toxic component of Honey Bee venom. Melittin is widely known to affect cell membranes by

56

forming pores, but its effect is short-lived, owing to the inherent instability of the pores.10-13

57

Wild-type melittin is known to have a bimodal interaction with a lipid bilayer; it can exist either

58

in a surface-bound state, with its backbone approximately parallel to the membrane surface, or in

59

an inserted state, forming pores that span the membrane. An amphipathic helix arises when

60

melittin contacts the membrane.14-16 Furthermore, evidence shows that the percentage of peptides

61

in each state is a function of the peptide-to-lipid ratio.17 At low peptide concentrations, the

62

majority of peptides associate with their backbones parallel to the membrane; this binding can

63

locally displace lipid head groups, causing a thinning of the membrane (a few Angstroms for


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al.

Page 4 of 30


64

wild-type melittin).9 At higher peptide concentrations, the amino acid chains are thought to insert

65

themselves into the membrane and create transmembrane pores.

66

To improve upon the shortcomings of wild-type, a variant of melittin, MelP5, was synthetically

67

evolved to create stable pores with radii large enough to pass macromolecules and to do so at

68

lower peptide concentrations than are required for wild-type melittin.18 Molecular dynamics

69

simulations of the MelP5 mutant have suggested that this gain of function variant works in a

70

similar fashion to the wild-type peptide.19 MelP5-induced membrane pores have been studied by

71

electrochemical impedance spectroscopy, macromolecule leakage assays, and single ion channel

72

recording, but the dynamic topographic structure of MelP5-induced membrane remodeling

73

remains uncharacterized.12,18

74

Atomic force microscopy (AFM) is a powerful single molecule tool that provides a real-time

75

real-space view of membrane topography with potentially sub-nm lateral resolution.20,21 This

76

method allows for direct imaging of peptide-induced remodeling of a fluid lipid bilayer in near-

77

native conditions, and has been used to observe defects induced by wildtype melittin.22 This

78

capability opens the door for studies of nuanced effects, such as the colocalization of the thinning

79

of the bilayer and pores, which can go undetected by other methods.

80

Here we employed AFM imaging to study the interaction of MelP5 with supported POPC lipid

81

bilayers. MelP5 created stable, highly localized pore-like voids in the membrane with diameters

82

out to around 80 Å. Bilayer remodeling was observed in two distinct modes: shallow non-

83

localized thinning, or deep highly localized voids. MelP5 exchanged between these two states

84

reversibly in an apparent dynamic equilibrium. Further, the two states were found to be

85

colocalized.


Page 5 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.


86

RESULTS AND DISCUSSION

87

Overview of atomic force microscope imaging

88

We studied the topographical consequences of MelP5 binding to lipid bilayers. To this end, we

89

performed tapping mode AFM imaging of supported bilayers made from liposomes which had

90

been incubated with peptides prior to fusion onto mica supporting surfaces.23 1-palmitoyl-2-

91

oleoyl-sn-glycero-3-phosphocholine (POPC) was chosen as a model bilayer due to its prevalence

92

in many eukaryotic cellular membranes.7,24,25

93

Previous studies have shown that MelP5 affects lipid bilayers at a variety of concentrations,

94

including significantly lower than those required for wild-type.18 AFM imaging trials were

95

conducted at varying concentrations of MelP5. Conditions were optimized to yield a significant

96

number of topographic features while simultaneously avoiding large-scale destruction of the

97

membrane. Experiments at high peptide-to-lipid ratios (P:L = 1:250 or 1:120) led to widespread

98

membrane distortion (Supporting Information Figure S1), whereas trials at lower concentrations

99

led to too few features for statistical analysis. It was found that P:L = 1:1200 was optimal for our

100

study, which focused on characterizing the topography and dynamics of MelP5-induced

101

membrane remodeling. We note that trials with wild type melittin were also carried out, but none

102

of them resulted in the formation of stable pore-like features (Supporting Information Figure S2).

103

Figure 1A shows a representative image of the effects of MelP5 on POPC (500 × 500 nm2, P:L =

104

1:1200). Numerous punctate depressions in the membrane surface are visible. A line scan

105

through the image (Fig. 1C, white dashed in Fig. 1A) reveals the depth of several depressions

106

and verifies the presence of the lipid bilayer via its characteristic 40 Å thickness.26 As a control,

107

membrane samples prepared in an identical manner, but without MelP5, were shown to be


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al.

Page 6 of 30


108

essentially featureless (Supporting Information Figure S3). Figure 1B is an image of MelP5-

109

treated membrane at increased magnification (290 × 290 nm2). The associated line scan (Fig. 1D)

110

highlights two localized pore-like features (purple shades) as well as a diffuse region of thinned

111

membrane (green). Note, the vertical scale in the images (Fig. 1A,B) is distinct from the line

112

scans (Fig. 1C,D); zero depth in the line scans delineates the top surface in the images, which we

113

take to be the surface of the unmodified upper leaflet of the bilayer. Clearly, MelP5 has a

114

significant effect on membrane topography.

115

To quantify the distribution of topographic distortions further, a histogram from data acquired in

116

the membrane region of the image (excluding the underlying mica surface) was compiled. This

117

histogram of pixel weights (Fig. 1E) reveals several important topographic populations. There

118

appears to be one prominent state exhibiting a depth of -0.3 +/- 0.2 nm (mean +/- σ) below the

119

top of the bilayer, and two other states with depths of -0.6 +/- 0.2 nm, and -1.2 +/- 0.3 nm,

120

respectively. These populations are robust; the histogram (Fig. 1E) was compiled using data

121

from three independent experiments (Supporting Information Figure S4).

122

In terms of relative weights, the 0.6 nm deep state was the most common topographical

123

distortion observed at this P:L = 1:1200. It occurred with a probability per area approximately 3-

124

fold higher than the deeper 1.2 nm state. The 0.3 nm deep state of the membrane was found with

125

an intermediate probability. To summarize, direct inspection of AFM images coupled with

126

histogram analysis of image pixels identified distinct topographic states of MelP5-induced

127

membrane distortion as well as relative weights.

128

Characterizing pore-like voids in the bilayer


Page 7 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.


129

To further elucidate the size and form of the voids, these pore-like features were isolated

130

algorithmically. Images were analyzed using custom software that located voids in a given image

131

with depths greater than 0.3 nm, controlled by the blob strength parameter in the Hessian blob

132

algorithm.27 We explored the relationship between measured void depth and footprint area.

133

By their nature, pores are highly localized voids within an otherwise planar bilayer surface. Both

134

the shallow (-0.6 nm +/- 0.1 nm) and the deep (-1.2 nm +/- 0.1 nm) sub-populations of voids

135

exhibited sharply peaked footprint area histograms (Fig. 2A & B). If these two populations

136

represented a deep mode of non-localized membrane thinning, then one would expect no

137

underlying structure and no preferred footprint area. In contrast, the data indicated that both of

138

these void populations were highly localized, and hence pore-like. Deeper pores exhibited larger

139

areas. Due to their localized nature, different depths could arise from pores with different

140

diameters. This is because the tip would read a shallower depth for a smaller diameter pore (Fig.

141

1E, Inset). Due to sharpness and aspect ratio limitations of the AFM probe (Supporting

142

Information Figure S5), the tip was not able to penetrate deeply into the MelP5 pores. However,

143

the underlying structure of the 0.6 nm and 1.2 nm deep features revealed by area analysis

144

suggests that these two populations are indeed pores.

145

Once identified (Fig. 2C, white contours), a complete histogram of footprint areas was compiled

146

for all pores (Fig. 2D). Assuming a circular geometry, which matched many of these features, a

147

histogram of radii was also created (Fig. 2D, Inset). As seen in Figure 2D, the two most

148

prominent populations of footprint area for MelP5 pores were 8.2 +/- 4.1 nm2 and 14.4 +/- 5.7

149

nm2. These populations give rise to radii of 1.7 +/- 0.4 nm and 2.4 +/- 0.5 nm, respectively. This

150

geometry is in overall agreement with previous results, which put the radius of MelP5 pores at

151

≥2 nm, due to its release of 10 kDa dextran.18 Though these two populations together constituted 7 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al.

Page 8 of 30


152

more than half of all the pore-like features observed, a significant population (23% of total)

153

exhibited a larger radius of 3.8 +/- 0.8 nm. Given the challenges in measuring longitudinal pore

154

geometry via AFM,28 the radii and footprint areas reported here are taken to be an upper limit on

155

the narrowest constriction of the pore. In summary, MelP5 does not create one size of pore, but

156

rather has a variability of pore footprint areas. This observation, which is based on direct

157

imaging of bilayer topography, is consistent with recent electrophysiology results,12 though

158

Fennouri et. al employed bilayers from mixed lipid species and here we used POPC.

159

Characterizing thinned regions of the bilayer

160

Spatially inhomogeneous and de-localized thinned regions of the membrane were frequently

161

observed in the AFM images. The thinned areas of the membrane had shallower depths and

162

dynamic, less definite structure as compared to the highly localized pores. Thinned areas of the

163

membrane were selected algorithmically in order to more precisely define their structure. Then, a

164

histogram was compiled using only the affected segments of the membrane. As seen in Figure 3,

165

the most probable depth of the thinned area was -0.30 +/- 0.14 nm. This agrees with and refines

166

the -0.3 +/- 0.2 nm peak identified in Figure 1. Tip restricted depth readings were less of a

167

concern here, since the thinned portions were not deep and exhibited significantly larger areas as

168

compared to both the pores and the apex area of the AFM tip itself. However, we note that while

169

thinned regions with small areas were rare (see Fig. 4), they could not be differentiated from

170

pores, given the limitations imposed by tip geometry.

171

The thinned features were characterized by their footprint area, to ascertain if they were indeed

172

thinned only—exhibiting no underlying structure—or if they were pores—which exhibit a high

173

degree of localization. Analyzing the area histogram of the thinned features revealed a wide


Page 9 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.


174

variety of possible footprint areas, the histogram for the thinned population has a shoulder much

175

larger than the sharply peaked pore histogram. This difference in underlying structure is seen in

176

Figure 4, where integrals of the two area histograms are plotted. For the pores, the integral

177

reaches unity very quickly—indicating that the vast majority of footprint areas are localized in a

178

distinct peak. However, for the thinned areas, the integral does not saturate nearly as rapidly,

179

owing to a lack of underlying structure.

180

Additionally, for each image there was on average 59 thinned features that accounted for 37% of

181

the total area of the image and 190 pores that accounted for only 3% of the total area of the

182

image (Fig. 4B). The membrane-thinned features were therefore much larger in area than the

183

pore-like features.

184

Colocalization of thinned membrane regions and pores

185

When analyzing the spatial distribution of the pores and the thinned portions of the membrane,

186

we discovered that pores were always surrounded by or immediately adjacent to thinned areas of

187

the membrane. Figure 5 shows representative images from three separate preparations

188

highlighting this observation. Here the pores are circled in white and the thinned areas in red. In

189

a popular model of pore formation,17 pores form in regions where enough peptides have already

190

congregated and thinned the membrane; this membrane-bound pool of peptides can then act in a

191

cooperative manner to form a pore. We emphasize that the observation of colocalization between

192

membrane thinned regions and pores requires a single molecule approach. Using the AFM we

193

were able to directly image the colocalization and thus shed light on the process of cooperativity.

194

Previous studies have shown that wild-type melittin can thin the bilayer,9 but the lateral

195

distribution and location of the thinning, and how it relates to pore location remained unknown.


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al.

Page 10 of 30


196

We observed that MelP5 causes laterally disperse and inhomogeneous thinning of the bilayer,

197

and that these thinned regions have within or abutting them punctate pores.

198

Topographical dynamics: transitions between membrane-thinned states and pore states

199

Continuously imaging the same area of the lipid bilayer surface gave us insight into the complex

200

dynamics that underlie the peptide-lipid interactions. By direct inspection of AFM images we

201

found that the peptides were able to exchange reversibly between the spatially diffuse

202

membrane-thinned state and the highly localized pore state. For example, in Figure 6A & B, the

203

white arrow points to a section of the membrane where the topography changed from a thinned

204

state (panel A) to a pore state (panel B). This change was registered by the membrane going from

205

a shallow depth reading to a deeper depth reading (indicated by the change in color). The circles

206

in Figure 6A-D highlight a region of the bilayer surface that shows that this process is reversible.

207

The circled feature changes its topography from a pore state (panels A & B) to a thinned state

208

(panel C) and then back to a pore state (D). To evaluate this behavior more quantitatively, line

209

scans (Fig. 6A’-D’, image panel A corresponds to line scan A’, etc.) show how the feature’s

210

depth changes as a function of time. The elapsed time between trace and retrace in the AFM

211

raster scan (red and black in Fig. 6A’-D’) was ≤ 200 ms. We note that the circled feature in

212

Figure 6C & C’ is labeled thin for two reasons: (i) its depth falls within one standard deviation of

213

the membrane thinned peak and (ii) topographically, it is not highly localized. More generally,

214

this feature could represent a transitional state between the pore state and the thinned state of the

215

membrane.

216

The time resolution of our AFM (one complete image every 180 seconds) sets an upper limit to

217

the timescale of the observed topographic state changes. Analysis of fifty individual transitions


Page 11 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.


218

similar to that circled in Figure 6 revealed that topographic transitions were not observed when

219

examining sequential trace and retrace line scans over nominally the same area of the membrane.

220

This is consistent with the data shown (Fig. 6A’-D’) and puts an approximate lower limit of 0.2

221

seconds on the time scale to transition between the thinned state and the pore state.

222

CONCLUSIONS

223

We provide novel characterization of membrane remodeling induced by the synthetically

224

evolved gain of function peptide MelP5. Direct observations of lipid bilayer remodeling show

225

that MelP5 interacts with the membrane and causes one of two general outcomes that are in

226

equilibrium: delocalized membrane thinning or highly localized pores. We measured the thinning

227

of the bilayer to be approximately 3 Å below the upper leaflet of the bilayer, in overall

228

agreement with previous studies of wild-type melittin. When analyzing pores formed by MelP5,

229

the most probable footprint area was found to be 8.2 nm2, which gives a most probable radius of

230

1.7 nm. However, the area histogram exhibited a sizeable shoulder and stable pores (on the

231

timescale of image acquisition) were observed with much larger footprint areas, out to about 47

232

nm2. Additionally, we have shown that the peptides exchanged reversibly between the thinned

233

state and the pore state, and have suggested upper and lower bounds on the characteristic time

234

scale of this transition (0.2 < τ < 180 s). At the P:L ratio we utilized (1:1200) pores were always

235

found surrounded by or abutting thinned areas of the membrane. This is consistent with

236

cooperative pore-formation where the local concentration of peptides determines the state—

237

inserted or membrane bound—of the peptide.17 When the local concentration reaches a

238

threshold, a transmembrane pore is created. Under that threshold, the peptides only thin the

239

membrane. In the future it would be interesting to vary the peptide concentration to explore this

240

local thresholding phenomena further. In conclusion, MelP5 has a profound influence on POPC 11 ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al.

Page 12 of 30


241

bilayers and AFM imaging provides a direct visualization of the topographic changes imparted

242

by the peptide in a physiologically relevant setting.

243 244

MATERIALS AND METHODS

245

Peptides. MelP5 was synthesized (BioSynthesis Inc.) with a free N-terminus and C-terminal

246

amidation. Identity and purity were verified with HPLC and mass spectrometry. The sequence of

247

MelP5 is GIGAVLKVLATGLPALISWIKAAQQL. Prior to AFM study, the peptide was diluted

248

to a final concentration of 1 µM in imaging buffer: 100 mM NaCl, 50 mM TrisCl at pH 7.4.

249

Peptide-lipid bilayer preparation. We adopted established methods to form supported lipid

250

bilayers on mica.29 Briefly, POPC (Avanti Polar Lipids) was suspended in 150 mM NaCl, 40

251

mM CaCl2, and 20 mM Hepes with a pH of 7.5 at a concentration of 650 µM. Liposomes were

252

prepared by extrusion of POPC through a polycarbonate membrane (approximately 25 times)

253

with a 100 nm pore diameter. POPC liposomes were incubated for 10 minutes with MelP5 at 25

254

°C to yield a final peptide concentration of 0.33 µM. The liposomes containing peptides were

255

then deposited onto freshly cleaved mica. Supported bilayers were formed by vesicle fusion (30

256

minute incubation, ~30 °C).29 Prior to imaging, samples were rinsed (75 µL, 4 times) with

257

imaging buffer. Unless otherwise specified, experiments were performed at a starting (i.e., prior

258

to rinsing) peptide-to-lipid ratio of 1:1200 and at ~30 °C, significantly above the gel-to-fluid

259

transition temperature of POPC.

260

AFM imaging and analysis. Images were acquired in tapping mode in aqueous buffer solution

261

using a commercial apparatus (Asylum Research, Cypher). Care was taken to control the

262

magnitude of the tip sample force to ≤ 100 pN (estimated by comparing the free amplitude to the


Page 13 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.


263

set point amplitude). As is typical, images were flattened (2nd order) to minimize background.

264

When histograms were compiled, the data was offset so as to align the population having the

265

greatest height (which we assumed to be the unaltered upper surface of the lipid bilayer) to zero.

266

Individual pores were selected using custom software (Igor Pro 7, WaveMetrics), which gave

267

details on the depth and footprint area of each pore27. Histograms were fit with multiple

268

Gaussians (MagicPlot), with x0 defined as the center of the peak.

269 270

ACKNOWLEDGEMENTS

271

The authors are grateful to William C. Wimley and Kalina Hristova for stimulating discussions.

272

Additionally we wish to thank all members of the G.M. King and L.L. Randall research groups at

273

University of Missouri. This work was supported by the National Science Foundation (Award #:

274

1709792) and the MU Research Board.

275 276

REFERENCES

277 278 279 280 281 282 283 284 285 286 287 288 289 290 291

1

2 3

4

5

Gilbert, R. J. C., Bayley, H. & Anderluh, G. Membrane pores: from structure and assembly, to medicine and technology. Philos Trans R Soc Lond B Biol Sci 372, doi:10.1098/rstb.2016.0208 (2017). Brogden, K. A. Antimicrobial peptides: pore formers or metabolic inhibitors in bacteria? Nat Rev Microbiol 3, 238-250, doi:10.1038/nrmicro1098 (2005). Soman, N. R. et al. Molecularly targeted nanocarriers deliver the cytolytic peptide melittin specifically to tumor cells in mice, reducing tumor growth. J Clin Invest 119, 2830-2842, doi:10.1172/JCI38842 (2009). Salomone, F. et al. A novel chimeric cell-penetrating peptide with membrane-disruptive properties for efficient endosomal escape. J Control Release 163, 293-303, doi:10.1016/j.jconrel.2012.09.019 (2012). Rausch, J. M., Marks, J. R., Rathinakumar, R. & Wimley, W. C. Beta-sheet pore-forming peptides selected from a rational combinatorial library: mechanism of pore formation in lipid vesicles and activity in biological membranes. Biochemistry 46, 12124-12139, doi:10.1021/bi700978h (2007).


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al. 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337

6 7

8

9

10 11

12

13

14

15

16

17

18

19

20 21

Page 14 of 30


Krauson, A. J. et al. Conformational Fine-Tuning of Pore-Forming Peptide Potency and Selectivity. J Am Chem Soc 137, 16144-16152, doi:10.1021/jacs.5b10595 (2015). Matin, T. R. et al. Single-Molecule Peptide-Lipid Affinity Assay Reveals Interplay between Solution Structure and Partitioning. Langmuir 33, 4057-4065, doi:10.1021/acs.langmuir.7b00100 (2017). Li, Y. et al. Single-molecule visualization of dynamic transitions of pore-forming peptides among multiple transmembrane positions. Nat Commun 7, 12906, doi:10.1038/ncomms12906 (2016). Chen, F.-Y., Lee, M.-T. & Huang, H. W. Evidence for Membrane Thinning Effect as the Mechanism for Peptide-Induced Pore Formation. Biophysical Journal 84, 3751-3758, doi:http://dx.doi.org/10.1016/S0006-3495(03)75103-0 (2003). Hanke, W. et al. Melittin and a chemically modified trichotoxin form alamethicin-type multi-state pores. Biochim Biophys Acta 727, 108-114 (1983). Wiedman, G., Herman, K., Searson, P., Wimley, W. C. & Hristova, K. The electrical response of bilayers to the bee venom toxin melittin: evidence for transient bilayer permeabilization. Biochim Biophys Acta 1828, 1357-1364, doi:10.1016/j.bbamem.2013.01.021 (2013). Fennouri, A., Mayer, S. F., Schroeder, T. B. H. & Mayer, M. Single channel planar lipid bilayer recordings of the melittin variant MelP5. Biochim Biophys Acta 1859, 2051-2057, doi:10.1016/j.bbamem.2017.07.005 (2017). Yang, Z., Choi, H. & Weisshaar, J. C. Melittin-Induced Permeabilization, Re-sealing, and Re-permeabilization of E. coli Membranes. Biophys J 114, 368-379, doi:10.1016/j.bpj.2017.10.046 (2018). Hristova, K., Dempsey, C. E. & White, S. H. Structure, location, and lipid perturbations of melittin at the membrane interface. Biophys J 80, 801-811, doi:10.1016/S00063495(01)76059-6 (2001). van den Bogaart, G., Guzman, J. V., Mika, J. T. & Poolman, B. On the mechanism of pore formation by melittin. J Biol Chem 283, 33854-33857, doi:10.1074/jbc.M805171200 (2008). Bechinger, B. The structure, dynamics and orientation of antimicrobial peptides in membranes by multidimensional solid-state NMR spectroscopy. Biochim Biophys Acta 1462, 157-183 (1999). Huang, H. W. Molecular mechanism of antimicrobial peptides: the origin of cooperativity. Biochim Biophys Acta 1758, 1292-1302, doi:10.1016/j.bbamem.2006.02.001 (2006). Wiedman, G. et al. Highly Efficient Macromolecule-Sized Poration of Lipid Bilayers by a Synthetically Evolved Peptide. Journal of the American Chemical Society 136, 47244731, doi:10.1021/ja500462s (2014). Woo, S. Y. & Lee, H. Aggregation and insertion of melittin and its analogue MelP5 into lipid bilayers at different concentrations: effects on pore size, bilayer thickness and dynamics. Phys Chem Chem Phys 19, 7195-7203, doi:10.1039/c6cp06834k (2017). Muller, D. J. & Dufrene, Y. F. Atomic force microscopy: a nanoscopic window on the cell surface. Trends Cell Biol 21, 461-469, doi:10.1016/j.tcb.2011.04.008 (2011). Sanganna Gari, R. R., Frey, N. C., Mao, C., Randall, L. L. & King, G. M. Dynamic structure of the translocon SecYEG in membrane: direct single molecule observations. J Biol Chem 288, 16848-16854, doi:10.1074/jbc.M113.471870 (2013). 14 ACS Paragon Plus Environment

Page 15 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al. 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357

22

23

24 25 26 27

28 29


Pan, J. & Khadka, N. K. Kinetic Defects Induced by Melittin in Model Lipid Membranes: A Solution Atomic Force Microscopy Study. J Phys Chem B 120, 4625-4634, doi:10.1021/acs.jpcb.6b02332 (2016). Cremer, P. S. & Boxer, S. G. Formation and spreading of lipid bilayers on planar glass supports. Journal of Physical Chemistry B 103, 2554-2559, doi:DOI 10.1021/jp983996x (1999). Chada, N. et al. Glass is a Viable Substrate for Precision Force Microscopy of Membrane Proteins. Sci Rep 5, 12550, doi:10.1038/srep12550 (2015). van Meer, G., Voelker, D. R. & Feigenson, G. W. Membrane lipids: where they are and how they behave. Nat Rev Mol Cell Biol 9, 112-124, doi:10.1038/nrm2330 (2008). Luckey, M. Membrane Structural Biology. 2nd edn, (Cambridge University Press, 2014). Marsh, B. P., Chada, N., Sanganna Gari, R. R., Sigdel, K. P. & King, G. M. The Hessian Blob Algorithm: Precise Particle Detection in Atomic Force Microscopy Imagery. Sci Rep 8, 978, doi:10.1038/s41598-018-19379-x (2018). King, G. M. & Golovchenko, J. A. Probing nanotube-nanopore interactions. Phys Rev Lett 95, 216103, doi:10.1103/PhysRevLett.95.216103 (2005). Mingeot-Leclercq, M. P., Deleu, M., Brasseur, R. & Dufrene, Y. F. Atomic force microscopy of supported lipid bilayers. Nat Protoc 3, 1654-1659, doi:10.1038/nprot.2008.149 (2008).

358 359 360 361 362 363 364 365 366 367 368 369 370 371 372


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al.

Page 16 of 30


373 374 375

FIGURE CAPTIONS

376

Figure 1. Experimental overview. (A, B): Representative images of the membrane affected by

377

MelP5. (C) Line scan from image in (A), position indicated by dashed line; note the

378

characteristic 4 nm thickness of the bilayer. (D) Line scan taken from image in (B), with certain

379

topographic populations highlighted. (E) Histogram compiled of pixels from representative

380

images taken from three independent experiments. Data is shown (red), the sum fit (black

381

dashed). E, Inset: Cartoon showing an explanation for a shallow reading.

382

Figure 2. Geometric analysis of pores. Histograms of footprint areas constructed from

383

subpopulations of voids exhibiting depths of (A) -0.6 nm +/- 0.1 nm or (B) -1.2 nm +/- 0.1 nm.

384

In both cases, the distributions are sharply peaked, indicating that these features are highly

385

localized. (C) The program picked out exclusively the pores (white contours), and did not

386

analyze the thinned portions of the membrane. (D) Histogram of footprint areas for all pores. The

387

histogram has a major population at an area of 8.2 nm2 but has subpopulations that go out as far

388

as 46.8 nm2. The weight of each peak is listed (grey dashed indicate the Gaussian fits). The data

389

is shown in red, while the dashed black line is the sum fit. Inset: Histogram of pore radii.

390

Figure 3. Analysis of thinned regions of membrane. (A) Thinned regions were identified

391

algorithmically and differentiated from taller regions of the bilayer and deeper pore-like

392

structures. (B) Histogram of the thinned region exhibiting a peak at -0.30 +/- 0.14 nm.

393

Figure 4. Comparison of areas. (A) Integrated area histograms for both the pore-like features

394

(black) and the thinned features (red). Inset: A detailed view highlights the difference between

395

the two populations. (B) Table comparing the pores and thinned features of the membrane. Two 16 ACS Paragon Plus Environment

Page 17 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.


396

quantities are of interest: (i) the average percent area of an image taken up by the pores or the

397

thinned features and (ii) the average total number of each feature type per unit area.

398

Figure 5. Colocalization of pores and membrane-thinned regions. Images of bilayers with MelP5

399

show that the pores—circled in white—fall within the flood mask (red) used to delineate the

400

thinned areas of the membrane.

401

Figure 6. Dynamic transitions of MelP5-induced membrane topography. (A-D) Time lapse

402

images of membrane, arrows and circles highlight interesting features. Line scans from each

403

image (A’-D’, line scan A’ corresponds to image A, etc.) show a detailed view of a transition

404

between a pore state (A,B,D) and a thinned state (C).


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al.

405


FIGURES

406

407

Page 18 of 30

Figure 1.

408


Page 19 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.


409

410

411

412 413

Figure 2.

414

415

416

417

418


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al.


419

420

421

422

423

424

Page 20 of 30

Figure 3.

425

426

427

428

429 20 ACS Paragon Plus Environment

Page 21 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.


430

431

432

433

434

Figure 4.

435

436

437

438

439

440

441

442


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Pittman et al.


443

444

445 446

447

448

449

Page 22 of 30

Figure 5.

450

451

452

453

454

455

456


Page 23 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Pittman et al.


457

458

459

460

461 462

463

Figure 6.

464 465


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table of Contents Graphic 33x14mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 24 of 30

Page 25 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 1 186x199mm (300 x 300 DPI)


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2 110x62mm (300 x 300 DPI)


Page 26 of 30

Page 27 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 3 96x116mm (300 x 300 DPI)


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4 58x24mm (300 x 300 DPI)


Page 28 of 30

Page 29 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 5 40x12mm (300 x 300 DPI)


Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6 114x69mm (300 x 300 DPI)


Page 30 of 30

Conformations and dynamic transitions of a melittin derivative that

Recommend Documents