Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption

BBAMEM-81931; No. of pages: 13; 4C: 3, 5, 6, 8, 9, 10, 11 Biochimica et Biophysica Acta xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbamem

3Q3

Marc-Antoine Sani a,1, Tzong-Hsien Lee b,1, Marie-Isabel Aguilar b,⁎, Frances Separovic a,⁎

4 5

a

6

a r t i c l e

7 8 9 10 11

Article history: Received 24 April 2015 Received in revised form 5 June 2015 Accepted 11 June 2015 Available online xxxx

12 13 14 15 16 17 18

Keywords: Antimicrobial peptide Maculatin 1.1 Model membranes Peptide–lipid interaction Dual polarization interferometry Dye leakage

i n f o

R O

School of Chemistry, Bio21 Institute, University of Melbourne, VIC 3010, Australia Department of Biochemistry and Molecular Biology, Monash University, Clayton, VIC 3800, Australia

a b s t r a c t

T

E

D

P

The membrane interaction of peptides derived from maculatin 1.1 and caerin 1.1, with the sequence motif of N and C termini of maculatin 1.1, was compared in order to understand the role of these common sequence motifs, which encompass critical proline residues, on peptide secondary structure and on membrane binding and disruption in zwitterionic and anionic membranes. The peptides incorporated a single substitution with lysine or deletion of the central region to mimic the length of the antimicrobial peptides, citropin 1.1 and aurein 1.2. The impact of these changes in the sequence, length and physicochemical properties, on lytic activity and structure was assessed by dye-release from lipid vesicles and the change in the bilayer order as a function of membrane-bound peptide mass. All peptides adopted similar degrees of helical structure in both membrane systems. In addition, all peptide analogues were less active than either maculatin 1.1 or caerin 1.1 in dye release assays. The membrane binding was analyzed by dual polarization interferometry and the results showed that membrane binding was significantly affected by changes in the hydrophobic environment of Pro-15. Moreover, changes in the relative distribution of charge and hydrophobicity flanking Pro-15 also caused significant changes to the membrane order. Overall, the proline residue plays an important role in inducing a peptide structure that enhances the activity of these antimicrobial peptides. © 2015 Elsevier B.V. All rights reserved.

C

b

33

E

37 35 34 36

1. Introduction

39 40

The alarming increase in antibiotic resistance presents a new challenge for medicinal chemistry to find potent alternatives that also limit natural selection of resistant strains. Among the innate immune system of higher organisms is found a plethora of relatively short (b50 amino acids) cationic peptides that have attracted significant interest. Their bactericidal mechanism is promising since their mode of action is to compromise the integrity of lipid membranes, which thereby avoids the activation of intracellular defense mechanisms [1,2]. First discovered in the late thirties, antimicrobial peptides (AMPs) are usually short cationic peptides, mainly adopting a helical structure, that induce lipid membrane leakage, either by forming a pore or by removing lipids in a detergent-like manner [3]. Many studies aim to identify the effect of amino acids in the primary sequence, mutating native AMP or trying to reproduce a certain scaffold with synthetic sequences. The main features that have been identified as important for activity are: amphipathicity that is promoted by the distribution of polar versus hydrophobic amino acids along the long axis of the structured peptide,

47 48 49 50 51 52 53 54 55

R

N C O

45 46

U

43 44

R

38

41 42

O

F

2

Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption

1Q2

⁎ Corresponding authors. E-mail addresses: [email protected] (M.-I. Aguilar), [email protected] (F. Separovic). 1 These authors contributed equally.

the net charge and the length of the peptide. Peptides shorter than 15 residues generally act in a detergent-like manner while longer peptides may insert and span the hydrophobic core of the lipid membrane to form pores. Overall, a certain threshold of peptide concentration must be reached before membrane disruption. Yet, despite more than two thousand AMPs having been identified [4], the relationship between potency and AMP sequence is still elusive. This is partly due to the lack of information on the changes in the membrane structure that accompany the binding of AMPs to the target membrane. We have selected two natural AMPs secreted on the skin of Australian tree frogs, maculatin 1.1 (Mac-1) and caerin 1.1 (Cae-1) as starting scaffolds to examine the structure–activity relationship of these two similar peptides. Maculatin 1.1 contains 21 residues and a proline residue which has been shown to play a critical role in the disruption of the bilayer [5,6], while caerin 1.1 contains 25 amino acids and two proline residues. Eight peptides were synthesized based on these two cationic peptides in which amino acid residues were either substituted or deleted (Table 1) in the region surrounding the proline residues. In the first category, an additional charge was added to Mac-1 by substituting Lys at residues 11 and 12. In the second category, either one or two turns of an α-helix were removed. The binding of these AMPs with two lipid membrane models was analyzed: the neutral or zwitterionic system of the major phospholipid found in eukaryotic cell membranes, palmitoyloleoylphosphatidylcholine (POPC); and a negatively charged

http://dx.doi.org/10.1016/j.bbamem.2015.06.013 0005-2736/© 2015 Elsevier B.V. All rights reserved.

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

19 20 21 22 23 24 25 26 27 28 29 30 31 32

56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79

2 t1:1 t1:2 Q1

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

Table 1 Physicochemical properties of Mac-1 and analogues. Sequence

Short name

No. of residues

MW (M + H) Da

Charge (pH 7)

Hydrophobicitya

t1:4 t1:5 t1:6 t1:7 t1:8 t1:9 t1:10 t1:11 t1:12 t1:13 t1:14 t1:15 t1:16 t1:17 t1:18

Maculatin — charge density changes GLFGVLAKVAAHVVPAIAEH F-NH2 GLFGVLAKVAAKVVPAIAEHF-NH2 GLFGVLAKVAKHVVPAIAEHF-NH2

Mac-1 Mac-2 Mac-3

21 21 21

2145.77 2136.44 2202.42

+1 +2 +2

1.3 1.3 1.0

Maculatin → Citropin → Aurein GLFGVLAK HVVPAIAEHF-NH2 GLFGVLAKVAA AIAEHF-NH2 GLFDVIKK VASVIGGL-NH2 GLFGVLAK IAEHF-NH2 GLFDIIKK IAESF-NH2

Mac-4 Mac-5 citropin Mac-6 aurein

18 17 16 13 13

1904.14 1712.10 1614.98 1400.10 1480.79

+1 +1 +1 +1 +1

1.1 1.4 1.3 1.1 1.1

Caerin → Maculatin GLLSVLGSVAKHVLPHVVPVIAEHL-NH2 GLLSVLGSVAKHV VPVIAEHL-NH2

Cae-1 Cae-2

25 21

2583.77 2137.42

+1 +1

t1:19

a

O

2. Material and methods

88

2.1. Materials

89 90

96 97

All peptides were obtained from the Bio21 peptide synthesis facility (Melbourne, Australia) with a N 95% purity. The peptides were washed in 5 mM HCl solution and lyophilised overnight to remove residual trifluoroacetic acid as described [7]. Palmitoyloleoylphosphatidylcholine (POPC) and palmitoyloleoylphosphatidylglycerol (POPG) phospholipids were purchased from Avanti Polar Lipids (Alabaster, USA) and were used without further purification. Calcein, Aprotonin (A3886), Tritonx100 and Sephadex G-100 gel filtration media were purchased from Sigma (St Louis, USA).

98

2.2. CD sample preparation

99

114 115

Peptides were dissolved in 10 mM phosphate and 1 mM buffer solution (pH 7.4), except Mac-5 which showed higher solubility in Milli-Q water. For all peptides, stock solutions containing 1 mg/mL were made. The stock solution was sonicated (10 s) and vortexed prior to each use. To prepare binary vesicles, lipids were co-solubilized in chloroform/methanol (3:1, v/v) before removal of solvents by rotary evaporation. Lipids were hydrated in Milli-Q water and lyophilised overnight. The resultant lipid powders were re-suspended in 10 mM phosphate and 1 mM NaCl buffer solution (pH 7.4). The homogeneous solutions were then extruded 10 times through an Avanti Mini-Extruder (Alabaster, USA) using polycarbonate filters to produce LUV of 200 nm diameter. The size of the LUV was confirmed by dynamic light scattering (DLS) performed on a Nano Zetasizer (Malvern Instruments Ltd, UK). Appropriate volumes of peptide stock solution and lipid vesicle dispersion were mixed to produce 160 μL samples with a fixed peptide concentration of 100 μM and 3 mM lipid.

116

2.3. CD measurements

117

CD spectra were acquired on a Chirascan spectropolarimeter (Applied Photophysics Ltd, UK) between 180 and 260 nm using a

102 103 104 105 106 107 108 109 110 111 112 113

118

2.4. CD spectral deconvolution

122 123

2.5. Calcein solution preparation

131

Calcein was initially insoluble in water and was, therefore, dissolved in 4 eq. NaOH and vortexed until completely dissolved. Appropriate volumes of Tris and NaCl were added to yield an 80 mM calcein stock in 30 mM Tris, 20 mM NaCl before final adjustment to pH 7.3 with HCl.

132 133

2.6. Calcein encapsulation in large unilamellar vesicles

136

To prepare calcein-loaded large unilamellar vesicles (LUVs), lipids were suspended in 250 μL of a dye solution made of 80 mM calcein, 20 mM NaCl and Tris 30 mM (pH 7.3). The solutions were freeze/ thawed five times and then extruded 10 times through an Avanti Mini-Extruder (Alabaster, USA) using 0.2 μm polycarbonate filters to produce LUVs of a theoretical 200 nm diameter. Samples were extruded above the corresponding main gel-to-fluid chain melting temperatures of the used lipids. Separation from free dye was obtained by gel filtration using Sephadex G-100 column media. The gel media and running buffer were carefully degassed before use to prevent air bubbles forming in the column. Samples were eluted under gravity with 30 mM Tris/100 mM NaCl (pH 7.3) running buffer solution at ~ 1 mL/min. Approximately 1 mL fractions were collected, with the two most concentrated fractions pooled prior to lipid concentration determination. Phospholipid concentrations were determined in triplicate using the phosphorus assay of Anderson et al. [10]. In parallel, lipids were also suspended in a calcein-free buffer to produce similar LUV solutions. The diameter of the calcein-free LUV was determined by DLS measurements.

137 138

E

Spectra were zeroed at 260 nm and normalized to give units of mean-residue ellipticity (MRE) according to [θ]MRE = θ/(c × l × Nr), where θ is the recorded ellipticity in milli-degrees, c is the peptide concentration in dmol ∙ L−1, l is the cell path-length in cm, and Nr is the number of residues per peptide. The helicity was calculated from the Luo–Baldwin formula Hα (%) = (θ222nm − θC)/(θ∞222nm − θC). with θC = 2200 − 53 T, θ∞222nm = (−44,000 + 250 T)(1-k/NResidues), and k = 4 as described for unrestricted peptides [8,9].

T

C

E

R

R O

100 101

C

94 95

N

92 93

U

91

R O

87

84 85

0.1 mm pathlength cylindrical quartz cell (Starna, Hainault, UK). Spectra 119 were acquired with 1 nm data intervals, 1 s integration time and 2 scans 120 accumulation. Signal was recorded as milli-degrees at 25 °C. 121

D

86

system of POPC with 30% palmitoyloleoylphosphatidylglycerol (POPG). The change in secondary structure and adsorption on and disruption of lipid bilayers was then investigated in terms of the peptide physicochemical properties versus the membrane surface charge. On the whole, the results revealed that the topological distribution of amino acid residues around the critical proline residue (Fig. 1) strongly influences the membrane binding and disruption by Mac-1.

82 83

1.2 1.3

P

80 81

Kyte and Doolittle hydrophobicity scale [23].

F

t1:3

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

124 125 126 127 128 129 130

134 135

139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155

3

U

N C O

R

R

E

C

T

E

D

P

R O

O

F

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

Fig. 1. Helical wheel and net presentations of the maculatin analogues. The arrow shows direction of, and is proportional to, the strength of hydrophobic moment.

156

2.7. Calcein leakage measurements and curve fitting

157

Measurements were made on a Varian Cary Eclipse spectrophotometer (Melbourne, Australia) using a semi-micro quartz cell (Starna,

158

Hainault, UK). The excitation wavelength was 490 nm and fixed wavelength fluorescence emission intensity was recorded at 515 nm for 0.5 s. Samples were prepared by mixing a 1:1 M ratio of calcein-loaded and calcein-free vesicles and appropriate amount of peptide stock solution

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

159 160 161 162

4

163 164 165 166

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

to produce desired lipid to peptide (L/P) molar ratios. Negative controls were made by substituting peptide stock for buffer-only or Aprotonin protein (soluble protein 6.5 kDa) and positive controls by adding 10 μl of 10% Triton X-100. The data was then fit using a dose–response model:
n y ¼ y0 þ ðymax −yo Þ= 1 þ ð log½Pep=EC 50 Þ

solid support, and the mass of peptide (mpeptide) bound to the lipid 218 bilayer, are calculated as follows: 219 mlipid ¼ d f ðniso −nbuffer Þ=ðdn=dcÞlipid ;

ð1Þ 221

mpeptide ¼ d f ðniso −nbuffer Þ=ðdn=dcÞpeptide ;

ð2Þ

222

where df is the thickness of the adlayer, and niso denotes the average or 224 corresponding isotropic RI of the adlayer:

170

where y0 is the baseline fluorescence, ymax is the fluorescence obtained upon Triton X-100 addition, EC50 is the concentration at which the peptide induces 50% of the maximum fluorescence, and n is the Hill coefficient indicative of the cooperativity between the peptides.

171

2.8. Dual polarization interferometry

226

172 173

192

Dual polarization interferometry (DPI) is an analytical method for analyzing thin films using an alternate dual orthogonal polarization [11]. This allows unique combinations of several opto-geometrical properties, including refractive index (RI), density, thickness, mass and birefringence, to be measured in real time for the formation of biomolecular layers and also to monitor the effect of additives to the properties of the adsorbed layer. The real-time measurements were performed using Analight BIO200 (Farfield Group Ltd, Manchester, UK). The real-time data were acquired with AnaLight200 version 2.1.0 software and processed and analyzed using AnaLight® Explorer software. Data were acquired at 10 Hz and averaged to give an output of one data point per second. The planar-supported unilamellar lipid bilayer was prepared on an unmodified silicon oxynitride FB80 AnaChips (Farfield-Group, UK). The chip surface was cleaned on-line by rinsing two times each with 10% Hellmanex II, 2% sodium dodecyl sulfate (SDS) and absolute ethanol. Prior to measurement, the waveguide chips were calibrated with respect to their optical properties using an 80:20 w/w ethanol/water mixture at 20 °C. All experiments were conducted in 10 mM HEPES pH 7.0, 150 mM NaCl. The flow rate of running buffer was controlled using a Harvard Apparatus PHD2000 programmable syringe pump.

nbuffer is the RI of the HEPES buffer used for these experiments, which was obtained experimentally as nbuffer = 1.3349 (20 °C) and 1.3340 (30 °C); c is concentration, and dn/dc is the specific RI increment of the adlayer. The de Feijter formula assumes that dn/dc remains constant throughout the experiment. For the present analysis, the dn/dc values of 0.135 mL/g and 0.182 mL/g were used for lipids and peptides, respectively [12–14]. The degree of molecular order, S, of the uniaxial lipid bilayer is defined by the ratio of the principal polarizabilities of the bilayer to the molecular polarizabilities [18,19]. This order parameter (S) is proportional to the birefringence (Δnf) values. Thus, the birefringence values represent an averaged measurement of lipid molecular orientation order and the lipid acyl chain order. High Δnf values are obtained for a fully aligned lipid bilayer whereas low Δnf indicating a random and disordered lipid bilayer. The birefringence of the adsorbed planar bilayer can be obtained with DPI through calculating the difference between two effective refractive indices, namely RI of transverse magnetic (TM) waveguide mode (nTM) and RI of transverse electric (TE) waveguide mode (nTE), as described previously [15]. The effective birefringence (nTM–nTE) can only be determined by calculating the two RIs, nTM and nTE, for each waveguide mode by fixing the thickness or RI of the deposited layer and assuming uniform layer coverage. For the purposes of the present study, the RI was fixed at 1.47 for the supported lipid bilayer formation at 20 °C [15].

193

2.9. Formation of supported lipid bilayer (SLB)

2.12. Peptide structure modeling

249

194

204

2.10. Peptide binding to bilayers

205 206

3. Results

265

3.1. Peptide design rationale

266

213 214

Peptide solutions were prepared at concentrations of 1, 2, 5, 10 and 20 μM in 10 mM HEPES, pH 7.4, 150 mM NaCl. 160 μL of each peptide was injected consecutively in order of increasing concentration onto the deposited bilayer at a flow rate of 40 μL/min at 20 °C. At the end of each peptide concentration injection, the layer was equilibrated by running buffer for 30 min prior to injecting the next concentration onto the same bilayer surface. Each concentration measurement was performed on the same lipid bilayer for each analogue. At the end of all peptide binding, the waveguide surface was regenerated with 2% SDS, 10% Hellmanex II and ethanol at 28 °C.

The PDB structures of maculatin 1.1 and caerin 1.1 obtained from the solution NMR experiments in dodecylphosphocholine environment [20, 21] were utilized. The visualization software, Chimera [22], was used to produce the peptide hydrophobicity surface according to the KyteDoolittle scale [23]. The structure of Mac-2 and Mac-3 was produced by mutating residues His-12 to Lys, and Ala-11 to Lys, respectively, and using the Dynameomics rotamer library to recompose the sidechain conformation [24]. The structures of Mac-4, Mac-5 and Mac-6 were obtained by truncating Mac-1 and patching the two segments using a peptide bond of 1.33 Å and the dihedral angles of an α-helix (−57°, −47°). Similarly, Cae-2 was obtained by truncating Cae-1. The structure-based alignment was performed within Chimera using the matchmaker analysis. Briefly, the Needleman–Wunsch alignment algorithm was used with the Blosum-100 matrix, 50% weighted on primary sequence and 50% weighted on secondary structure [22].

250 251

203

Liposomes (100 nm in diameter) were prepared by aliquoting 2 mM POPC and 2 mM POPG stock solution as previously described [12–15]. Liposome solutions (0.1 mg/mL) of POPC and POPC/POPG (7:3) were injected in the presence of 1 mM CaCl 2 for 10 min at 20 mL/min at 28 °C. The adsorption was immediately followed by injection of 1 mM CaCl2 in running buffer for 10 min. The SLB was further equilibrated in running buffer for 20 min before adjusting the temperature to 20 °C. A full lipid bilayer-covered silicon oxynitride chip was confirmed by the absence of BSA (50 mg/mL) binding to the membrane bilayer.

267 268

215

2.11. Calculation of mass and birefringence (Δnf)

216

The mass of an adsorbed molecular layer was calculated using the de Feijter formula [16,17]. The mass of lipid bilayer (mlipid) formed on the

Maculatin 1.1 and caerin 1.1 are two very closely related peptides with similar sequence, length, charge and hydrophobicity (Table 1). We hypothesized that specific segments of the Mac-1 sequence are critical to its antimicrobial activity and that removal of or changes to these specific regions will provide new insight into their role in the activity of these peptides. The overall aim of this study was to selectively modify or remove regions of the Mac-1 or Cae-1 sequences to determine the effect

190 191

195 196 197 198 199 200 201 202

207 208 209 210 211 212

217

q
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  ffi n2TM þ 2n2TE =3

D

P

R O

O

F

ð3Þ

E

188 189

niso ¼

T

186 187

C

184 185

E

182 183

R

180 181

R

178 179

O

176 177

C

174 175

N

169

U

168

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248

252 253 254 255 256 257 258 259 260 261 262 263 264

269 270 271 272 273

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

298 299 300 301 302 303 304 305 306 307 308

C

296 297

E

294 295

−20,799 −23,489 −18,033 −18,631 −21,140 −20,192 −19,002 −17,851

1.00 1.12 0.87 0.90 1.02 0.97 0.92 0.86

−18,305 −24,743 −18,662 −17,871 −20,333 −15,878 −18,677 −21,833

1.00 1.34 1.02 0.98 1.11 0.87 1.02 1.18

1.14 0.95 0.97 1.04 1.04 1.27 1.02 0.82

t2:5 t2:6 t2:7 t2:8 t2:9 t2:10 t2:11 t2:12

F

determined using the ratio of the θ222nm molar ellipticity for both lipid systems. Mac-6 was less helical (0.87) in the presence of POPC/POPG vesicles while the helicity for all other peptides was similar to within ~10% to that of Mac-1. Overall, however, all analogues adopted substantial helical structure in both lipid mixtures.

309

3.3. Membrane disruption analysis by dye-release

314

The potency of Mac-1 and analogues to disrupt the bilayer structure of POPC and POPC/POPG (7:3) vesicles was investigated by monitoring the increase in fluorescence upon addition of increasing amount of peptide to self-quenched calcein-encapsulated LUV. Dynamic light scattering measurements did not show any significant size variation upon Mac-1 titration up to L/P 5:1 (~ 90 nm radius), which indicates that Mac-1 did not cause significant disruption of the membrane and is consistent with formation of a pore-like structure [25]. The fluorescence obtained at each peptide concentration was normalized by using the signal obtained without peptide (baseline) and with addition of Triton X-100 (100% signal). The data was then fitted using a dose–response model to extract the EC50 values of each peptide and relative ratios were calculated using Mac-1 as reference (Fig. 3). A relative EC50 ratio less than one (Table 3) indicated a less potent peptide compared to Mac-1.

315 316

R

292 293

Mac-1 Mac-2 Mac-3 Mac-4 Mac-5 Mac-6 Cae-1 Cae-2

R

290 291

t2:4

N C O

288 289

θPC222nm/θPCPG222nm

U

283 284

t2:3

Relative helicity

O

The secondary structure of Mac-1 and analogues was determined using circular dichroism (CD) spectroscopy at 25 °C. In the absence of lipid vesicles, all peptides except Mac-5 were mainly unstructured, as seen in the single minimum around 200 nm (Fig. 2A). It is worth noting that Mac-5 and Mac-6 were not fully soluble in buffer, and the CD experiments were then performed in Milli-Q water. In the presence of neutral POPC vesicles, all peptides adopted a helical conformation, with the characteristic two minima observed at 222 nm and 208 nm and a strong maximum around 190 nm (Fig. 2B). The helical content was calculated according to the Luo–Baldwin formula using the molar ellipticity at 222 nm (Table 2). Since this method required using a corrective factor (k), estimated to a value of 4 for unblocked peptides [9], we determined a relative helicity based on the wild-type Mac-1 for comparison. The difference in helicity between the 8 peptides was not significant, ranging from + 0.12 to − 0.14. Although Mac-5 was already structured in water, its helicity increased by about 50% in presence of POPC vesicles. In the presence of negatively charged lipid vesicles comprised of POPC/POPG (7:3), all peptides showed typical α-helical CD line shapes (Fig. 2C). The helicities of Mac-2 and Cae-2 were greater relative to Mac-1, while the other peptides had similar values (Table 2). The difference in structure for each peptide in the presence of vesicles was

281 282

POPC vs. POPC/POPG

θ222nm

R O

287

280

POPC/POPC 7:3 Relative helicity

P

3.2. Secondary structure analysis by circular dichroism

278 279

t2:1 t2:2

θ222nm

POPC

D

286

276 277

Table 2 Mean residue ellipticity and helicity relative to Mac-1 in LUV environments.

T

285

of these deletions on membrane disruptive properties. Firstly, the net charge of maculatin was increased from + 1 to + 2 by changing His12 to Lys (Mac-2) and disrupting amphipathicity by changing Ala-11 to Lys (Mac-3). Residues 9–15 in Mac-1 are amino acids that are absent in the closely-related AMP, aurein 1.2 and correspond to individual turns in the α-helical structure of each peptide. We, therefore, prepared Mac-4 and Mac-5 in which residues 9–11 or 12–15 (one turn) were deleted respectively, and Mac-6 in which residues 9–15 (two turns) were deleted. A similar approach was used to determine the relationship between Mac-1 and Cae-1 by removing residues 14–17 (one turn) in Cae-1 to yield Cae-2.

E

274 275

5

Fig. 2. CD spectra of 100 μM Mac-1 and analogues in: [A] buffer composed of 10 mM phosphate and 1 mM NaCl, pH 7.4; [B] POPC large unilamellar vesicles (LUV, 200 nm); and [C] POPC/POPG (7:3) LUV (200 nm) at a lipid to peptide ratio 30:1. Spectra are an average of 3 accumulations recorded at 25 °C.

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

310 311 312 313

317 318 319 320 321 322 323 324 325 326 327 328

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

R

R

E

C

T

E

D

P

R O

O

F

6

329 330

C

O

Fig. 3. Dye-leakage assays showing the lytic activity of Mac-1 and analogues against: (A) POPC LUV (200 nm); and (B) POPC/POPG (7:3) LUV (200 nm). Aliquot of peptides were added to a fixed lipid concentration of 250 μM and the samples were incubated for 15 min prior to measurements. Experiments were performed in duplicate and fitted with a dose–response model.

t3:1 t3:2 t3:3

Table 3 EC50 ratios relative to Mac-1 obtained from dose–response fit of dye-leakage assays in LUV environments.

U

331 332

N

333

In the presence of neutral vesicles, only Cae-1 was as lytic as Mac-1, with the other peptides being less lytic with POPC LUV, especially the shorter analogues Mac-4, Mac-5 and Mac-6 (Table 3). The Mac-2 and Mac-3 analogues also showed reduced activity, indicating that the extra positive charge carried by these peptides lower their insertion ability.

t3:4

POPC

POPC/POPG (7:3)

POPC vs. POPC/POPG

t3:5

relative EC50

relative EC50

PCPG ECPC 50/EC50

1.00 0.37 0.57 0.15 0.19 0.06 0.98 0.25

1.00 0.52 0.48 0.54 0.04 0.06 2.90 0.59

3.03 2.15 3.64 0.85 12.90 3.08 1.01 1.29

t3:6 t3:7 t3:8 t3:9 t3:10 t3:11 t3:12 t3:13

Mac-1 Mac-2 Mac-3 Mac-4 Mac-5 Mac-6 Cae-1 Cae-2

This was also observed for the Cae-2 peptide, although this analogue has the same net charge, number of amino acids and hydrophobicity. With negatively charged vesicles, a similar trend was observed with Mac-5 and Mac-6, which were less lytic, while Cae-1 showed a greater potency than Mac-1. Remarkably all peptides, except Mac-4, were less lytic against charged vesicles than neutral vesicles, as cationic AMP is known to target specifically the negatively charged membrane of bacteria. Even Mac-2 and Mac-3 analogues with an extra Lys substituted for a neutral amino acid, did not disrupt the charged LUV more efficiently. In fact, of all the analogues, only Cae-1 was as potent against POPC/POPG as POPC membranes.

334 335 336 337 338 339 340 341 342 343 344

3.4. Membrane disruption analysis by dual polarization interferometry 345 (DPI) 346 The interaction of each maculatin peptide with model membranes was studied in real time by DPI. Each of the maculatin peptides (1–20 μM) was injected consecutively onto a planar supported POPC and POPC/POPG (7:3) bilayer with the following structural properties:

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

347 348 349 350

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

372

F

O

R O

370 371

P

368 369

D

366 367

E

364 365

T

362 363

C

360 361

E

358 359

R

357

R

355 356

Overall, all peptides dissociated fully with the mass returning to the starting baseline at the end of dissociation for each peptide concentration (Fig. 6B). The binding of each analogue to the anionic POPC/POPG bilayer (Fig. 5) showed small and almost irreversible increases in mass at the lowest peptide concentration (1 μM). As the peptide concentration increased, the mass values started to decrease during the injection of peptide solution, dropping below the starting baseline at the end of dissociation, which indicates mass loss from the surface. The mass loss was observed for the native Mac-1 and Cae-1 and analogue Mac-6 at 2 μM, while for other analogues this occurred at a higher concentration (5 μM). For further increases in peptide concentration, the extent of mass loss varied with the analogue sequence (Table S2). However, this large mass loss during the peptide adsorption was not observed for Mac-5, which showed a small mass drop below the baseline at the end of 10 and 20 μM injections. The lower POPC-bound mass and higher peptide concentration required to induce mass loss during the interaction of POPC/POPG bilayer indicated that the sequence alteration made the analogues less effective binders than the native Mac-1 and Cae-1, consistent with the dye release data. The binding characteristics of each analogue were further evaluated by the relative changes in the molecular organization of the bilayer as a

N C O

353 354

POPC mass/unit area 4.49 ± 0.05 ng/mm2, thickness 4.48 ± 0.04 nm and birefringence (Δnf) 0.0169 ± 0.0014; and POPC/POPG (7:3) mass/ unit area 4.40 ± 0.04 ng/mm2, thickness 4.39 ± 0.04 nm and birefringence (Δnf) 0.018 ± 0.0004, each averaged from 16 independent liposome depositions. The mass per unit area versus time for the binding of each maculatin analogue to POPC and POPC/POPG (7:3) bilayers are shown in Figs. 4 and 5, respectively. The mass bound to and dissociated from POPC is also plotted against concentration for each peptide in Fig. 6. The POPCbound mass of Mac-1 increased with increasing peptide concentration during the association (Fig. 4A). However, the single Lys substitution for His12 and Ala11 in Mac-2 and Mac-3, respectively, resulted in lower mass bound to the POPC bilayer than for Mac-1 (Fig. 6, Table S1). The deletion Mac analogues, Mac-4 and Mac-5 also showed significantly lower mass bound to the POPC than Mac-1. In addition, the deletion of amino acids 12–15 (Mac-5) resulted in a very slow dissociation of peptide. Interestingly, the removal of residues 9–15 (Mac6) partially restored the binding to Mac-1 levels. Cae-1 showed similar mass binding to POPC as Mac-1 at low peptide concentrations (1 and 2 μM) while the mass of Cae-1 bound to POPC was higher than Mac-1 at the higher concentrations (10 and 20 μM). Cae-2, in which residues 14–17 are removed showed a substantial loss of binding to the POPC.

U

351 352

7

Fig. 4. The real-time changes in the bilayer mass for POPC-bound maculatin analogues.

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

R

E

C

T

E

D

P

R O

O

F

8

U

N

C

O

R

Fig. 5. The real-time changes in bilayer mass for POPC/POPG (7:3)-bound maculatin analogues.

Fig. 6. Binding characteristics of maculatin analogues to planar-supported POPC bilayers. (Effects of sequence variation on mass bound to lipid bilayer.) Only POPC is shown as there was significant mass loss with POPC/POPG (see Supp. Info. Tables S1 & S2).

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

9

R

E

C

T

E

D

P

R O

O

F

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416

function of membrane-bound peptide mass (Figs. 7 & 8). The molecular order of a lipid bilayer is measured by the anisotropic optogeometrical parameter, birefringence, and the influence of peptide binding on this parameter can be obtained by DPI. During the binding of Mac-1 to POPC, the birefringence values decreased from 0.0181 in a concentration-dependent manner, with the lowest value of 0.0121 obtained during injection of 20 μM peptide. The bilayer disordering evident from the decrease in birefringence was reversible during the peptide dissociation from the bilayer. However, there was a small decrease in the birefringence values at the end of dissociation and the mass almost returned to the starting base line. Overall, this profile was similar to the pattern previously observed for the binding of Mac-1 to the DMPC bilayer, except the birefringence values varied between 0.0165 and 0.0195 [6]. The binding of Mac-2 to POPC caused much smaller changes in the birefringence than Mac-1 although the POPC-bound Mac-2 mass was similar to that of Mac-1. In contrast, Mac-3 induced similar decreases in POPC birefringence at a lower POPC-bound Mac-3 mass compared to Mac-1. For the truncation analogues, a small decrease in POPC birefringence was observed for Mac-4, which also bound a small amount. However, Mac-5 and Mac-6 induced a significant decrease in POPC birefringence to values lower than those for Mac-1.

U

395 396

N C O

R

Fig. 7. POPC bilayer order as characterized by the birefringence change (Δnf) against the mass of the POPC-bound maculatin analogues.

This indicated that Mac-5 and Mac-6 changed the molecular order of POPC bilayer more than Mac-1 but a lower Mac-6 concentration was required for this large drop. For Cae-1 and Cae-2, small decreases in POPC birefringence were observed. These differences in POPC-birefringence values are related to the depth and orientation of peptides interacting with the POPC molecules. The induction of packing disorder by the accumulation of Mac-1 to the POPC/POPG bilayers was accompanied by mass loss, in contrast to the disorder in POPC, which occurred without peptide-induced mass loss. As shown in Fig. 8, the binding of the charge density analogues Mac-(1–3) to POPC/POPG showed significant decreases in birefringence at low levels of membrane-bound mass. This was accompanied by an overall loss in mass after injection of all 5 concentrations. For the truncation analogues, Mac-4 showed a similar change in birefringence to Mac-1 while Mac-5 and Mac-6 showed larger bilayer disordering with very small levels of peptide binding to the membrane. Although there was no significant overall mass loss for Mac-5, there was a large mass loss for 2 μM Mac-6. Cae-1 and Cae-2 showed larger decreases in POPC/POPG birefringence at lower concentration but the overall extent of POPC/POPG birefringence changes induced by Cae-1 and Cae-2 was similar to Mac-4 and less than those obtained for Mac-1 to Mac-3, Mac5 and Mac-6.

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

R

E

C

T

E

D

P

R O

O

F

10

439

R

Fig. 8. POPC/POPG (7:3) bilayer order as characterized by the birefringence change (Δnf) against the mass of the POPC/POPG (7:3)-bound maculatin analogues.

446

4. Discussion

447 448

AMPs with effective antibacterial activities have long been anticipated as a new hope in combating bacterial infections resistant to multiple generic antibiotics. While the sequence and structural determinants are critical for activity, the highly diverse sequence properties, complex physico-structural parameters, and low target cell selectivity with cytotoxicity have limited the development and of AMPs as antibiotics for clinical use. Thus, understanding the sequence characteristics, structural determinants and physical properties for membrane-destabilization and antibacterial activity is essential for the rational development of effective peptide antibiotics. Previous analysis of Australian amphibian AMP sequences has demonstrated high degree homology in the Nand C-termini of four peptides: aurein 1.2, citropin 1.1, maculatin 1.1 and caerin 1.1 [26]. In particular, correlation of the physical properties

449 450 451 452 453 454 455 456 457 458 459

C

N

U

442 443

O

444 445

Overall, the binding of peptides created a new irreversible state in the anionic POPC/POPG bilayers, which was not observed for POPC as the peptides fully dissociated from the bilayer. Nevertheless, all peptides induced a decrease in POPC/POPG birefringence and the membrane was able to recover to a significant extent during dissociation of all peptides except Mac-5, which reflects the stability of the bilayer against the destructive action of the peptides.

440 441

of these four peptides with their membrane-disrupting behavior showed that an increase in peptide length causes changes from a parallel binding induced membrane lysis to an oblique or transmembrane insertion associated with membrane disruption. Moreover, we recently demonstrated that the proline residue in the sequence of Mac-1 [5,6], which creates a kink in the helical structure, plays an important role in the ability of Mac-1 to insert into a membrane bilayer. In the present study, we have explored the role of the amino acid residues that flank the critical proline residues in maculatin 1.1 and caerin 1.1 and hypothesize that the relative topographic distribution of hydrophobic residues plays a determining role in the insertion and lysis by these peptides. Particularly, the interaction with zwitterionic and anionic membranes of peptides derived from maculatin 1.1 and caerin 1.1, with the sequence motif of N and C termini of maculatin1.1, was compared with the aim of understanding the role of these common sequence motifs, which encompass critical proline residues, on peptide secondary structure and on membrane binding and disruption. These peptides encompass three different sub-groups (Fig. 9) and incorporated single substitution with lysine (Mac-2 and Mac-3), deletion of central region to mimic the length of citropin 1.1 (Mac-4 and Mac-5), and a mimic of aurein 1.2 (Mac-6). Deletion of the central region of caerin 1.1 (Cae-2), results in a peptide with the same length as Mac-1. The impact of these changes in the sequence, length and physicochemical properties

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482

11

N C O

R

R

E

C

T

E

D

P

R O

O

F

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

U

Fig. 9. Left panel: Space filling representation of the maculatin analogues with the hydrophobic residues (red), hydrophilic (blue) and neutral (white) according to the Kyte and Doolittle scale. Right panel: structure-based sequence alignment weighted with 25% of primary sequence analogy with (A) Mac-1 (brown) as reference versus Cae-1 (purple) and Cae-2 (blue); (B) citropin (purple) as reference versus Mac-4 (brown) and Mac-5 (blue); and (C) aurein (brown) as reference versus Mac-6 (blue).

483 484 485

(charge, hydrophobicity, amphipathicity, secondary structure) was assessed by dye-release from lipid vesicles and change in the bilayer order as a function of membrane-bound peptide mass.

486

4.1. Net charge and charge distribution around Pro-15

487

The lysine substitutions are in close proximity and are located at the polar face of the helix for Mac-2 and Mac-3. However, the relative orientation of these residues with the neighboring amino acids resulted in substantially different binding properties particularly with POPC, although there was no effect on the degree of helical structure. In particular, despite having one additional positive charge, neither Mac-2 nor

488 489 490 491 492

Mac-3 bound more than the native Mac-1 to either POPC or POPC/ POPG. In a previous study [6], we demonstrated that the proline in position 15 in maculatin 1.1 plays a critical role in bilayer penetration as analogues without the proline do not contain the kinked helix and are less able to penetrate the bilayer. Maculatin 1.1 is strongly amphipathic along both ends of the helix, and the peptide causes the bilayer surface to indent to accommodate the kink and maximize the interaction with the entire peptide. In the present study, the two lysine substitutions are in close proximity to this proline residue in Mac-2 and Mac-3, and exerted an effect on the ability of maculatin to bind efficiently to either bilayer. In POPC, Mac-2 (His-12 to Lys) could not penetrate the bilayer (small changes in birefringence, low dye release) while the Ala-11 to

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

493 494 495 496 497 498 499 500 501 502 503 504

533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568

C

528 529

E

526 527

R

524 525

R

522 523

O

520 521

C

518 519

N

516 517

U

514 515

569 570 571 572 573 574 575 576 577 578 579 580

4.3. Role of proline and bilayer charge in membrane lysis by maculatin 1.1 581

F

The influence of the central sequence region of maculatin 1.1 on activity was investigated further by deletion of residues 9–11 (Mac-4), 12–15 (Mac-5) and 9–15 (Mac-6). Relative to Mac-1, all three peptides contain the same net charge but deletion of these residues broadened the hydrophilic surface of the helix, thus increasing its amphipathicity. The deletions did not significantly change the degree of helicity but did cause a significant change in the vesicle lysis (dye leakage) and in binding to both POPC and POPC/POPG (DPI). Firstly, Mac-4, which is less hydrophobic, bound weakly to and could not penetrate the POPC bilayer to any significant extent. It was, however, able to bind and penetrate the negatively-charged POPC/POPG bilayer also causing mass loss or bilayer expansion. In contrast, Mac-5, which does not contain the critical proline residue, bound weakly to both POPC and POPC/ POPG at low concentrations and dissociated very slowly, but inserted reversibly at the highest concentration. Most significantly, Mac-5 exerted a very dramatic decrease in the order of the POPC/POPG bilayer at very low levels of binding. Finally, Mac-6 which also does not contain the critical proline and contains the same number of amino acids as aurein 1.2 (which induces membrane lysis/micellisation [27]) caused drastic changes in disordering of both POPC and POPC/POPG bilayers. However, no mass loss was noted when binding to POPC, unlike the membrane lysis effect seen for aurein 1.2 binding to DMPC and DMPC/ DMPG [14,15]. The sequence of aurein 1.2 is GLFDIIKKIAESF and differs from Mac-6 at residues 4 (D → G), 5 (I → V), 7 (K → A) and 12 (S → H). While the overall net charge is still +1, aurein 1.2 has two Lys residues and one Asp residue and hence the differences in the overall charge density clearly influence the lytic ability of these peptides as Mac-6 is unable to cause significant mass loss as has been previously observed for aurein 1.2 [14,15]. The homologous sequence region in caerin 1.1 (Cae-1) was also removed to give an analogue of maculatin 1.1 (Cae-2). Cae-1 contains two proline residues and behaved very similarly to Mac-1 with both POPC and POPC/POPG. Irreversible mass loss together with reversible bilayer disordering of POPC/POPG by native Cae-1 was similar to previous studies with DMPC/DMPG [14]. In addition, the higher mass loss for POPC/POPG indicates greater susceptibility for more fluid bilayers. By comparison, Cae-2 which contains the HVVP motif but only one proline,

512 513

Overall, the peptides in this study were either equally or more lytic towards neutral rather than negatively charged bilayers based on the dye release results with POPC/POPG. By comparison, the DPI data indicated more membrane disruption or insertion/expansion with the negatively charged SLBs. The dye-release assays, however, measure fluorescence release from vesicles after a defined time of exposure to each peptide, while DPI reports the changes in membrane structure of a SLB during binding of a particular peptide. In addition, the dyerelease experiments were performed at lower peptide concentrations (Fig. 3), below the level which caused the structural changes in the bilayers reported by DPI. The effects of Mac-1 on POPC and POPC/POPG bilayers were consistent with those previously reported for DMPC and DMPC/DMPG [14,15]. Mac-1 induced significant but reversible disorder of the lipid acyl chains without substantial mass loss when binding to POPC. However, the extent of mass loss and disordering was significantly enhanced for binding to POPC/POPG compared to DMPC/DMPG. While disordering of POPC/ POPG was also reversible, the birefringence of the bilayer that remained following peptide dissociation and mass loss returned to pre-injection values, which indicates that the fluid phase of POPC/POPG bilayers is more susceptible to Mac-1 disruption but can subsequently recover. The native peptides, Mac-1 and Cae-1, were most active and induced similar effects on both phospholipid bilayers, although the effects on neutral and anionic bilayers were very distinct. Cae-1, the longer and more kinked peptide with two Pro residues, showed similar lytic activity in neutral and anionic LUVs. Mac-5 and Mac-6 were an order of magnitude less lytic with anionic LUV, probably as a result of truncation (bilayer mismatch) together with the loss of the HVVP residues. The question, therefore, is what sequence motifs are responsible for the lytic properties of aurein 1.2, maculatin 1.1 and caerin 1.1. Sequence-activity comparison of each peptide suggests an important role for the hydrophobic and charge distribution around the critical proline residue and further emphasizes the interplay between this proline and the overall amphipathicity in the AMP activity [6,14,15]. We previously demonstrated that Pro-15 is important for the lytic activity of maculatin 1.1 as the kink that it generates in the helical structure facilitates the insertion of the peptide into the membrane bilayer [5,6]. The results of the present study clearly demonstrate that the hydrophobic environment and the specific location of cationic charge with respect to Pro-15 are also central to the ability of maculatin 1.1 to penetrate and lyse a lipid bilayer. More specifically, the results suggest that maculatin 1.1 contains an insertion motif that is comprised of an amphipathic wedge in which Pro-15 is flanked by hydrophobic residues on one surface and positively charged residues on the opposite surface. Furthermore, the different effects on the zwitterionic and anionic bilayers suggests that manipulation of the topography of these residues to differentially inhibit the binding and/or insertion depending on the bilayer properties can be used to control the selectivity between membranes. Overall, our studies have correlated the structural domain/motif of maculatinrelated peptides with the membrane-binding characteristics and

O

532

511

R O

4.2. Truncation of the residues encompassing Pro-15

509 510

P

531

507 508

bound much less and caused less bilayer disordering than Cae-1 and Mac-1 on both POPC and POPC/POPG bilayers at a given concentration. However, the mass-birefringence profiles (Figs. 7 and 8) are similar in shape, which indicates that both Cae-1 and Cae-2 act via a similar disruptive mechanism but Cae-2 are less potent. The sequence of Cae-2 is GLLSVLGSVAKHVVPVIAEHL and contains the same number of residues as Mac-1 but differs from Mac-1 at residues 3 (F → L), 4 (G → S), 7 (A → G), 8 (K → S), 16 (A → V), and 21 (F → L) and has an overall net charge of +1. Thus, removal of the central LPHV residues yields a less active peptide than Mac-1, which demonstrates that the specific sequence in terms of charge distribution and amphipathicity plays a dominant role.

T

530

Lys substitution in Mac-3 caused more change in membrane order than Mac-2 but less than Mac-1. All three peptides were more effective in causing dye leakage in POPC than POPC/POPG LUV. Possibly at low peptide concentrations there may be more peptides interacting with the charged surface but less peptide inserting into the hydrophobic core of the membrane. The extent of mass loss in the POPC/POPG bilayer induced by Mac-2 and Mac-3 was also reduced relative to Mac-l. However, significant irreversible disordering of POPC/POPG by the increased positive charge was observed at low concentration (2 μM). The effect of more charge with similar (Mac-2) or less (Mac-3) hydrophobicity on the bilayer properties showed that the more charged peptides can increase irreversible disordering of the acyl chains and that hydrophobic interactions affect insertion and consequently the extent of bilayer expansion. Mac-1, Mac-2 and Mac-3 all caused large changes in membrane order and subsequent mass loss of the membrane, indicating that the presence of the negatively charged lipid facilitated the penetration of these peptides into the bilayer. In addition, while the different positive charge distribution around the proline residue did not prevent mass loss and disordering of the POPC/POPG bilayer by Mac-2 and Mac-3, higher concentrations of these two peptides were required to induce the changes in membrane order relative to Mac-1. Thus the relative distribution of the positive charge around the proline residue is a critical determinant of maculatin 1.1 activity and suggests that manipulation of this charge distribution may offer additional approaches for the design of more potent and selective AMPs.

D

505 506

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

E

12

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632

M.-A. Sani et al. / Biochimica et Biophysica Acta xxx (2015) xxx–xxx

639

Acknowledgements

640 641 642

The authors gratefully acknowledge the support of the Australian Research Council for award of Discovery grant DP110101866 to MIA and FS.

643

Appendix A. Supplementary data

644 645

Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.bbamem.2015.06.013.

646

References

647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668

[1] M.L. Cohen, Changing patterns of infectious disease, Nature 406 (2000) 762–767. [2] G. Taubes, The bacteria fight back, Science 321 (2008) 356–361. [3] M. Zasloff, Antimicrobial peptides of multicellular organisms, Nature 415 (2002) 389–395. [4] G. Wang, Improved methods for classification, prediction, and design of antimicrobial peptides, Methods Mol. Biol. 1268 (2015) 43–66. [5] D.I. Fernandez, A.P. Le Brun, T.H. Lee, P. Bansal, M.I. Aguilar, M. James, F. Separovic, Structural effects of the antimicrobial peptide maculatin 1.1 on supported lipid bilayers, Eur. Biophys. J. 42 (2013) 47–59. [6] D.I. Fernandez, T.H. Lee, M.A. Sani, M.I. Aguilar, F. Separovic, Proline facilitates membrane insertion of the antimicrobial peptide maculatin 1.1 via surface indentation and subsequent lipid disordering, Biophys. J. 104 (2013) 1495–1507. [7] M.A. Sani, E.J. Dufourc, G. Grobner, How does the Bax-alpha1 targeting sequence interact with mitochondrial membranes? The role of cardiolipin, Biochim. Biophys. Acta 1788 (2009) 623–631. [8] P. Luo, R.L. Baldwin, Mechanism of helix induction by trifluoroethanol: a framework for extrapolating the helix-forming properties of peptides from trifluoroethanol/ water mixtures back to water, Biochemistry 36 (1997) 8413–8421. [9] N.E. Shepherd, H.N. Hoang, G. Abbenante, D.P. Fairlie, Single turn peptide alpha helices with exceptional stability in water, J. Am. Chem. Soc. 127 (2005) 2974–2983. [10] R.L. Anderson, S. Davis, An organic phosphorus assay which avoids the use of hazardous perchloric acid, Clin. Chim. Acta 121 (1982) 111–116.

F

The Transparency document associated with this article can be found, in the online version.

O

638

R O

637

P

Transparency document

C

U

N C O

R

R

E

[11] J. Escorihuela, M.A. Gonzalez-Martinez, J.L. Lopez-Paz, R. Puchades, A. Maquieira, D. Gimenez-Romero, Dual-Polarization Interferometry: a novel technique to light up the nanomolecular world, Chem. Rev. (2014) [12] K. Hall, T.H. Lee, A.I. Mechler, M.J. Swann, M.I. Aguilar, Real-time measurement of membrane conformational states induced by antimicrobial peptides: balance between recovery and lysis, Sci. Rep. 4 (2014) 5479. [13] T.H. Lee, K.N. Hall, M.J. Swann, J.F. Popplewell, S. Unabia, Y. Park, K.S. Hahm, M.I. Aguilar, The membrane insertion of helical antimicrobial peptides from the N-terminus of Helicobacter pylori ribosomal protein L1, Biochim. Biophys. Acta 1798 (2010) 544–557. [14] T.H. Lee, C. Heng, F. Separovic, M.I. Aguilar, Comparison of reversible membrane destabilisation induced by antimicrobial peptides derived from Australian frogs, Biochim. Biophys. Acta 1838 (2014) 2205–2215. [15] T.H. Lee, C. Heng, M.J. Swann, J.D. Gehman, F. Separovic, M.I. Aguilar, Real-time quantitative analysis of lipid disordering by aurein 1.2 during membrane adsorption, destabilisation and lysis, Biochim. Biophys. Acta 1798 (2010) 1977–1986. [16] G.H. Cross, A.A. Reeves, S. Brand, J.F. Popplewell, L.L. Peel, M.J. Swann, N.J. Freeman, A new quantitative optical biosensor for protein characterisation, Biosens. Bioelectron. 19 (2003) 383–390. [17] M.J. Swann, L.L. Peel, S. Carrington, N.J. Freeman, Dual-polarization interferometry: an analytical technique to measure changes in protein structure in real time, to determine the stoichiometry of binding events, and to differentiate between specific and nonspecific interactions, Anal. Biochem. 329 (2004) 190–198. [18] R. Horvath, J.J. Ramsden, Quasi-isotropic analysis of anisotropic thin films on optical waveguides, Langmuir 23 (2007) 9330–9334. [19] J.J. Ramsden, Molecular orientation in lipid bilayers, Philos. Mag. B Phys. Condens. Matter Stat. Mech. Electron. Opt. Magn. Prop. 79 (1999) 381–386. [20] B.C. Chia, J.A. Carver, T.D. Mulhern, J.H. Bowie, Maculatin 1.1, an anti-microbial peptide from the Australian tree frog, Litoria genimaculata solution structure and biological activity, Eur. J. Biochem. 267 (2000) 1894–1908. [21] H. Wong, J.H. Bowie, J.A. Carver, The solution structure and activity of caerin 1.1, an antimicrobial peptide from the Australian green tree frog, Litoria splendida, Eur. J. Biochem. 247 (1997) 545–557. [22] E.F. Pettersen, T.D. Goddard, C.C. Huang, G.S. Couch, D.M. Greenblatt, E.C. Meng, T.E. Ferrin, UCSF Chimera—a visualization system for exploratory research and analysis, J. Comput. Chem. 25 (2004) 1605–1612. [23] J. Kyte, R.F. Doolittle, A simple method for displaying the hydropathic character of a protein, J. Mol. Biol. 157 (1982) 105–132. [24] A.D. Scouras, V. Daggett, The Dynameomics rotamer library: amino acid side chain conformations and dynamics from comprehensive molecular dynamics simulations in water, Protein Sci. 20 (2011) 341–352. [25] M.A. Karal, J.M. Alam, T. Takahashi, V. Levadny, M. Yamazaki, Stretch-activated pore of the antimicrobial Peptide, magainin 2, Langmuir 31 (2015) 3391–3401. [26] M.P. Boland, F. Separovic, Membrane interactions of antimicrobial peptides from Australian tree frogs, Biochim. Biophys. Acta 1758 (2006) 1178–1183. [27] D.I. Fernandez, A.P. Le Brun, T.C. Whitwell, M.A. Sani, M. James, F. Separovic, The antimicrobial peptide aurein 1.2 disrupts model membranes via the carpet mechanism, Phys. Chem. Chem. Phys. 14 (2012) 15739–15751.

D

636 Q4

T

635

provide a more specific basis for the design of highly selective antibacterial therapeutics. Molecular dynamics studies may give further insight into this relationship.

E

633 634

13

Please cite this article as: M.-A. Sani, et al., Proline-15 creates an amphipathic wedge in maculatin 1.1 peptides that drives lipid membrane disruption, Biochim. Biophys. Acta (2015), http://dx.doi.org/10.1016/j.bbamem.2015.06.013

669 670 671 Q6 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715