Investigation of the interaction of antimicrobial peptides with lipids and lipid membranes 3

7.3 Results and discussion 7.3.1 Isotherm studies of V4 interacting with POPG and POPC 7.3.1.1 Isotherms of lipid monolayers The surface pressure π and molecular area A isotherms for P

CHAPTER 7 INTERACTION BETWEEN ANTIMICROBIAL PEPTIDES AND LIPID MEMBRANE LAYERS

7.1 Introduction

V4 was shown to first bind to the membranes and finally induce membrane permeation

by membrane aggregation and disruption However, the intermediate process is not clear

In this chapter, monolayer is mainly used to investigate the process of V4 inserting into membranes and the comparision of V4 to other antimicrobial peptides is also shown

7.2 Materials and methods

Materials

POPG, POPC and DPPG were purchased from Avanti Solvent chloroform (HPLC grade) and methanol (HPLC grade) and antimicrobial peptide magainin 2 (M2), melittin (ME) and polymyxin B (PB) were purchased from Sigma-Aldrich V4 peptide was synthesized

by Genemed The purity of different peptides has been presented in previous chapters All the materials were used without further purification

Instrumentation

Langmuir film balance (model 601M) (NIMA Technology Ltd England) was used for experiments The instrument includes a 105 cm2 trough connected with an external circulator for temperature control, two mechanically coupled barriers, a surface pressure sensor using Wihelmy plate, a sapphire window and a dipper well (25mm stroke) An interface unit (IU4) connected the film balance and the computer An operating software

(version 5.16) provided by NIMA Technology Ltd was used to collect data The cleanness of the trough was checked by closing and opening of the barriers to ensure that

surface pressure did not vary by more than ±0.1 mN/m

Monolayer isotherms

POPG and POPC were dissolved in chloroform with a concentration of 0.2 mM Due to the low solubility of V4 in chloroform, V4 was first dissolved in the minimal volume of methanol to prepare a clear solution Additional chloroform was then added to prepare a solution with V4 concentration of 0.2mM Required volume of POPG or POPC was mixed with V4 to form lipid/V4 mixture with V4 percentage of 0%, 5%, 10%, 20%, 33%, 50% and 100% A syringe was cleaned completely and an appropriate volume of individual lipid/V4 mixed solution was drawn and carefully deposited on the water surface in a drop-wise manner, making sure that the surface pressure did not change after deposition The monolayer of lipid in the absence or in the presence of V4 formed spontaneously on the air-water interface After solvent evaporation for 10 minutes, the monolayer was compressed with a rate of 7 cm2/min and the isotherm curve was record

Each curve was repeated at least twice for reproducibility

Penetration studies

POPG and POPC were dissolved in chloroform and DPPG was dissolved in the mixture

of chloroform and methanol (v/v=3:1) with a final lipid concentration of 0.2 mM All the studied antimicrobial peptides were dissolved in water with high concentrations as stock solution Required volume of lipid solution (usually 60 µl) was drawn by using a clean

syringe and spread on the water surface carefully in a drop-wise manner, making sure that the surface pressure did not change after lipid deposition The lipid monolayer formed spontaneously on the air-water interface After solvent evaporation for 10 minutes, the monolayer was compressed with a rate of 7 cm2/min to a target surface pressure The lipid monolayer was allowed to adjust until a constant molecular area was achieved Afterward an appropriate volume of peptide solution was injected underneath the monolayer into the subphase, generating different peptide concentration in the trough The surface pressure change with time with fixed molecular area was record All the experiments were done at 37 °C and each penetration experiment was repeated at least twice for reproducibility A water subphase but not buffer was used to avoid crystallization of salt on the sample which would interfere with the imaging of sample in the AFM experiments

AFM experiment

A monolayer which was penetrated by antimicrobial peptides was transferred to freshly cleaved mica (Electron Microscopy Sciences, USA) by vertically placing the mica in the water subphase before lipid was spread on the water surface After the penetration experiment, the surface pressure was kept constant at the surface pressure of complete penetration The mica was slowly extracted from the subphase to the air phase with a constant rate of 3 mm/min A lipid monolayer was obtained by compressing the lipid monolayer to a certain surface pressure and extracted from the mica from the subphase at the constant target surface pressure The monolayer with or without peptide on the mica was dried in a desiccator overnight before AFM imaging

AFM experiment was performed in air on the NanoScope IIIa MultiMode Scanning Probe Microscope manufactured by Digital Instruments Inc (Santa Barbara, CA 93117, USA) Topographic images were acquired in tapping mode The typical scan rates ranged from 1 to 1.25 Hz depending on the scan size The monolithic silicon probes (NanoWorld

AG, Switzerland) with a cantilever length of 125 µm and force constant of 42 N/m were used for measurements Images was obtained and analyzed by the Nanoscope software provided by the company Images from at least two different sample prepared on different days with several macroscopically separated areas on each sample were acquired for data reproducibility Representative images are shown

Insertion study of V4 into POPG bilayer

Dual polarization interferometry was used to study the insertion of V4 into solid supported bilayers This technique allows the opto-geometrical properties (density and thickness) of adsorbed layers at a solid-liquid interface to be determined 132 When a peptide is introduced into the thin lipid bilayer, the changes of average thickness and average density (through the refractive index) of the lipid bilayer as well as the mass can yield information of how the peptide interacts with the lipid bilayer Experiments were done on an AnaLight ® Bio 200 system POPG SUVs were prepared and deposited on an amine modified surface sensor chip After obtaining a stable POPG bilayer on the surface, V4 was injected and the density and mass change of the POPG bilayer were recorded

7.3 Results and discussion

7.3.1 Isotherm studies of V4 interacting with POPG and POPC

7.3.1.1 Isotherms of lipid monolayers

The surface pressure (π) and molecular area (A) isotherms for POPG and POPC

monolayer at the air-water interface at 37 °C are shown in Fig 7.1 When the lipid monolayer was compressed, the isotherm for POPG and POPC began to rise at a molecular area of 107 Å and 96 Å, respectively With increasing compression, the surface pressure increased continuously until the collapse pressure of 45.6 mN/m and 44.5 mN/m for POPG and POPC, respectively The shape of isotherms of POPG and POPC monolayer was similar especially at low molecular area, which indicated that the packing

of POPG and POPC lipid molecules was similar POPG and POPC are both unsaturated lipids containing one double bond They have the same hydrophobic alkyl chains and differ in the headgroup The molecular packing of the monolayer is mainly dependent on the hydrophobic interaction between the alkyl chains of the lipid molecules The same alkyl chain of POPG and POPC allowed similar molecular packing so that POPG and POPC showed similar isotherms Thus hydrophobic interaction played a significant role during the compression Although POPG bears a negative charge, which might impose a repelling effect between POPG molecules especially when the molecules were close, there was not much difference between the POPG and POPC isotherms, indicative of the negligible effect of electrostatic interaction

Fig 7.1 Isotherms of POPG and POPC monolayers

7.3.1.2 Isotherms of mixed lipid/V4 monolayers

Fig 7.2 shows the surface pressure (π) and molecular area (A) isotherms of POPG/V4

and POPC/V4 monolayers at the air-water interface at 37 °C The pure V4 showed strong surface activity compared with lipid At a molecular area of 125 Å, the isotherm of V4 began to rise With increasing compression, the surface pressure of V4 monolayer continuously increased to 45.0 mN/m, which was comparable to the collapse pressure of POPG and POPC When V4 was incorporated into the lipids, the isotherms of the mixed POPG/V4 and POPC/V4 monolayer shifted right to the high molecular areas with increasing percentage of V4 The collapse pressure for all isotherms was similar except for the POPG/V4 mixture with 50% V4 incorporation, which was an incomplete isotherm because the two barriers were too close The shape change of isotherms was complicated The isotherms of POPG/V4 with V4 percentage of 5% and 10% were similar to the isotherm of pure POPG When V4 percentage increased to 20% or higher, there was a kink at surface pressure of 30 mN/m, which indicated that V4 peptide might induce a

phase transition Above 30 mN/m, the increase of surface pressure due to the compression slowed down The presence of V4 in the POPC monolayer with V4 percentage of 5% did not induce much change in the isotherm However when the percentage of V4 increased to 10% and higher, the shape of the mixed POPC/V4 isotherms was similar to pure V4 isotherm Comparing all the molecular areas at which the surface pressure began to increase (lift-off area), it was found that with increasing incorporation of V4, the lift-off areas gradually increased for both lipids, which indicated that V4 had an area-expanding effect on the lipid monolayer at low surface pressure

Fig 7.2 Isotherms of mixed POPG/V4 and POPC/V4 monolayers

7.3.1.3 Miscibility analysis of monolayers

When the studied components were mixed to form monolayers on the air-water interface,

it was not easy to determine if the components were really miscible or not from the direct measurement Analysis of the monolayer isotherm may provide useful information to

determine the miscibility of the studied components in the monolayer Each pure component has its own collapse pressure In a two-component system, if the components were immiscible, two collapse pressures would be observed at the corresponding collapse pressure of pure components However if the two components were miscible, only one collapse pressure would be obtained220 Fig 7.2 showed that all mixed monolayers had one collapse pressure for both POPG and POPC, which gave an indication that V4 was miscible with POPG and POPC

According to the phase rule of Defay and Crisp221, in a two-component system, if the components are completely miscible for all the ratios, there is only one degree of freedom

at constant temperature and pressure, assuming no externally imposed electrical potentials Therefore when the percentage of V4 in the mixed lipid/V4 solution varies, the surface pressure will vary correspondingly at fixed molecular area If V4 and lipid mix ideally, the ideal surface pressure can be calculated by

A A

ideal A A

ex

A, π , exp π ,

π = − (7.2) Fig 7.3 shows that the surface pressure varied with different percentage of V4 incorporating into POPG and POPC monolayer at different molecular areas An apparent difference between the different molecular areas was observed for both lipids At high molecular area, which the molecules were loosely packed, the surface pressure increased gradually with the increasing percentage of V4 When there was 50% V4 incorporated into the lipid monolayer maximal surface pressure was obtained With increasing molecular area, the surface pressure difference between the different molecular areas became smaller At low molecular area, the curve became flat (POPG) or the maximal surface pressure shifted to low percentage of V4 incorporation (POPC)

Fig.7.3 Surface pressure of lipid monolayers incorporated with different percentage of V4 Left: POPG; Right: POPC

The excess surface pressure, which indicates the deviation from ideal mixing is shown in Fig 7.4 For the POPG lipid, except at a low molecular area of 50 Å, the surface pressure

at the other molecular areas all showed positive deviation from linearity The positive excess surface pressure indicated that V4 had a surface pressure increasing effect on the monolayer, which was equivalent to an area expanding effect The more V4 was incorporated, the greater the effect was until a maximal value was reached with V4 incorporation of 33% to 50% dependent on different molecular areas At the molecular

area of 50 Å, there was not much excess surface pressure When the percentage of V4 was smaller than 20%, slight negative deviation from the linearity was observed This indicated that at this molecular area, V4 peptide was almost ideally mixed with POPG with a tiny area condensing effect on the monolayer at a certain percentage of V4 Within the all studied molecular areas, V4 induced positive deviation from the linearity for the POPC monolayer Similar to POPG, with increasing V4 incorporated into POPC monolayer, the excess surface pressure increased until it reached a maximal value at the percentage of V4 between 20% and 50% Simultaneously with decreasing molecular area, the maximal excess surface pressure shifted to low V4 percentage In brief, similar

to POPG, V4 also generated an area expanding effect on the POPC monolayer

Fig 7.4 Excess surface pressure of lipid monolayers incorporated with different percentage of V4 Left: POPG; Right: POPC

7.3.1.4 Stability analysis of monolayers

The stability of a monolayer can be investigated by the evaluation of the excess energy caused by interaction between V4 and lipids According to the method developed by Zhao and Feng220,222, the excess Helmholtz energy can be used to determine the energy change The excess Helmholtz energy was defined as the deviation from the ideal value

of Helmholtz energy according to Eq 7.3

dA X

X A

∫

=

∆ (7.3)

A 0 and A are the molecular area where the surface pressure begins to increase from zero

and where the Helmholtz excess energy is calculated, respectively π12, π1 and π2 are the surface pressure of mixed monolayer, pure lipid monolayer and pure V4 monolayer at the

molecular area of A, respectively X 1 and X 2 imply the percentage of lipid and V4 respectively in the mixed monolayer

Fig 7.5 shows the excess Helmholtz energy change at different molecular areas For both POPG and POPC, the presence of V4 in the mixed monolayer induced a negative deviation from ideal mixing The negative excess Helmholtz energy indicated that there was interaction between lipid and V4 peptide and less energy was needed for a stable system Therefore the stability of a system can be determined by the value of excess Helmholtz energy With increasing percentage of V4 incorporated into the lipid monolayer, the excess Helmholtz energy continuously decreased, which indicated an increasing interaction between lipid and V4 The minimal excess energy which corresponded to a stable system was obtained in the presence of 50% V4 in the mixed monolayer Therefore when V4 was incorporated into the monolayer with the percentage from 0% to 50%, V4 increased the system stability Except the percentage of V4, the molecular area also affected the change of excess Helmholtz energy It was shown that at the biggest studied molecular area of 100 Å, V4 created a relatively small energy decrease Decrease in the molecular area induced apparent decrease in excess energy when the percentage of V4 was fixed This result indicated that in the presence of a

particular percentage of V4, the compression increased the interaction between V4 and lipid and consequently the system stability

Fig 7.5 Excess Helmholtz energy of lipid monolayers incorporated with different percentage of V4 Left: POPG; Right: POPC

7.3.1.5 Compressibility analysis of monolayers

Compressibility is a parameter to describe the elastic packing of the monolayer It is defined as

)/(

on the packing of POPG monolayer, which can be attributed to the interaction between V4 and POPG lipid Compared with POPG, V4 induced much less effect on the POPC monolayer packing

Fig 7.6 Compressibilities of lipid monolayers incorporated with different percentage of V4

POPG and POPC showed similar isotherms especially at low molecular area As mentioned before, the determining factor for the isotherm is the hydrophobic interaction between the alkyl chains of lipids and charge nearly has no effect Compared with some saturated lipids, POPG and POPC both have high lift-off molecular areas at which the surface pressure began to rise from zero upon compression223 Because of the unsaturated structure, the kink in the alkyl chains, which is formed by the double bond, makes one molecule occupy a large area than a saturated lipid did Therefore unsaturated lipid monolayers are more compressible than saturated lipid monolayers This was confirmed

by the result of the compressibility shown in Fig 7.6 in which both POPG and POPC monolayers have a relatively high value The presence of V4 further increased the compressibility of the monolayer, indicative of the increased molecular area, which was consistent with the result of excess surface pressure

The incorporation of V4 decreased the system energy as shown in Fig 7.5 With increasing percentage of V4, the energy decreased steadily which indicated an increasing interaction between V4 and lipid This result was consistent with the previous study that V4 had affinity for both POPG and POPC Therefore V4 did increase the system stability Simultaneously, the presence of V4 induced a positive excess surface pressure, which indicated that V4 promoted a surface pressure-increasing effect or an area-expanding effect on both lipid monolayers One of the possible reasons might be the conformational change due to the interaction between V4 and lipid When there was no V4, the alkyl chains of the lipid molecules extended upward to the air phase and the headgroup contacted with water phase The hydrophobic interaction between the hydrophobic part of V4 and alkyl chains of the lipid made the alkyl chains tilt to the interface which increased the area needed for one molecule and led to an area expanding effect When the area was fixed, the increased surface pressure was observed as shown in Fig 7.4 Therefore, V4 changed the conformation of lipid through peptide-lipid interaction and imposed a surface pressure increase effect on the lipid monolayer

7.3.2 Penetration studies of antimicrobial peptides interacting with lipid monolayers 7.3.2.1 Penetration of antimicrobial peptides into POPG monolayers

The presence of antimicrobial peptide in the subphase induces a surface pressure change, which indicates the penetration or insertion of the peptide into the lipid monolayer The interaction between magainin 2, melittin, polymyxin B and V4 with POPG monolayers is shown in Fig 7.7 Before peptide was injected, the POPG monolayer was compressed to

a target surface pressure (initial pressure) of 15 mN/m The addition of antimicrobial

peptide into the subphase resulted in an increase in the surface pressure for all studied peptides Fig 7.7 shows the penetration kinetics of different peptides into POPG monolayers With time increasing, the surface pressure increased continuously When the surface pressure showed no significant increase, the monolayer was saturated and the penetration was thought to reach equilibrium With increasing peptide concentration, the penetration apparently sped up At the lowest studied peptide concentration of 50 nM, the time needed to achieve complete penetration was over 6000 s However, when there was more peptide in the solution, less time was needed to complete penetration At a peptide concentration of 150 nM, magainin 2, melittin and polymyxin B led to a fast penetration

of less than 2000 s V4 took more time with about 4000 s When the peptide concentration further increased, the rate of penetration correspondingly increased This result indicated that the rate of penetration depended on the peptide concentration and is diffusion limited The rate of penetration was related to the peptide concentration, which was close to the surface of the monolayer in the water phase Because concentrated peptide was injected underneath the monolayer in the water phase, the peptide molecules accumulated in several regions These peptide molecules diffused and some of them reached the surface of the monolayer, where peptide molecules might eventually penetrate the monolayer At high peptide concentration more peptide molecules reached the surface of the monolayer and penetrated into the monolayer, which showed a fast penetration compared with low peptide concentration

Fig 7.7 Kinetics of antimicrobial peptides with different concentrations penetrating into POPG monolayers

Fig 7.8 Comparison of antimicrobial peptides penetrating into POPG monolayers

The peptide concentration affected not only the rate of penetration, but also the extent of penetration Fig 7.8 compares the penetration of antimicrobial peptides into POPG monolayers ∆π is used to describe the surface pressure increase from the target surface pressure (or initial surface pressure) When V4 concentration increased from 0.05 to 1

µM, V4 exhibited a gradual increase in ∆π from 4.9 to 19.7 mN/m Further addition of V4 into the subphase did not induce much change in ∆π, which indicated that the monolayer reached saturation at 1 µM Similar to V4, magainin 2 and melittin generated increasing ∆π with increasing peptide concentration up to 0.5 µM More addition of peptide into the subphase did not induce apparent increase in surface pressure Thus the maximal penetration into POPG monolayer due to magainin 2 and melittin was obtained

at 0.5 µM Polymyxin B was different from the other antimicrobial peptides From a peptide concentration of 0.05 µM to 1 µM, polymyxin B generated similar penetration Further addition of peptide induced slight increase in ∆π compared with the lower peptide concentration This result implied that at 0.05 µM, polymyxin B caused monolayer saturation According to the increased surface pressure ∆π at a saturated peptide concentration, for example 1 µM, the ability of peptide penetrating into the POPG monolayer was determined Therefore from Fig 7.8 the penetration ability of these antimicrobial peptide was melittin > magainin 2 > V4 > polymyxin B

7.3.2.2 Penetration of antimicrobial peptides into POPC monolayers

It is known that mammalian cell membranes consist mainly of zwitterionic lipids in contrast to bacterial membranes Thus POPC was chosen in this study to investigate the interaction between antimicrobial peptides and zwitterionic monolayers The penetration

of antimicrobial peptides into POPC monolayers and comparison with POPG monolayers are shown in Fig 7.9 Compared with POPG monolayers, melittin, polymyxin B and V4 all displayed relatively weak penetration ability for POPC One important reason might

be the charge difference between POPG and POPC POPG harbors a negative charge,

which electrostatically interacts with cationic antimicrobial peptides Electrostatic interaction leads to the absorption of higher amount of peptide on the headgroup of POPG, which increased the opportunity of peptides penetrating into the lipid monolayer Once the peptide molecules absorbed on the monolayer, the hydrophobic interaction drove the peptide molecules to penetrate into the monolayer and induced an increase in surface pressure However, due to the lack of electrostatic interactions between the peptides and POPC, less peptide molecules were absorbed on the monolayer surface, which reduced the possibility of penetration Therefore cationic antimicrobial peptides showed higher penetration ability for POPG than for POPC monolayers, which is the basis for selectivity

Fig 7.9 Comparison of melittin, polymyxin B and V4 penetrating into POPG and POPC monolayers

The experiments showed that the penetration of antimicrobial peptides into POPC monolayers is also peptide concentration dependent With increasing melittin concentration from 0.05 to 1 µM, ∆π increased gradually until monolayer saturation The addition of the extra melittin led to a constant surface pressure increase of 6-7 mN/m Compared to melittin, polymyxin B induced much less increase in the surface pressure

which indicated weak penetration into POPC monolayers At the peptide concentration of 0.05 and 0.15 µM, nearly no penetration was observed When the concentration increased

to 0.5 µM, only about 1mN/m surface pressure increase was obtained Additional peptide did not generate a higher surface pressure increase up to a peptide concentration of 3 µM The low value in ∆π caused by polymyxin B indicated that polymyxin B might only absorb on the headgroup of the lipid without real penetration into POPC monolayer V4 displayed relatively strong penetration ability compared with melittin and polymyxin B

At the lowest studied peptide concentration of 0.05 µM, V4 induced a ∆π of 3.3 mN/m Increase in the peptide concentration led to an increase in ∆π At 0.5 µM, V4 caused POPC monolayer saturation with ∆π of 10 mN/m Comparing these three antimicrobial peptides, it is found that V4 had the highest penetration ability into POPC monolayer followed by melittin Polymyxin B showed nearly no penetration ability to this zwitterionic lipid The difference in penetration into POPC monolayer can be attributed

to the hydrophobicity of the peptide V4 is a very hydrophobic antimicrobial peptide, containing 8 valines out of a total of 19 amino acids It was shown that even in detergent, V4 aggregated due to its high hydrophobicity (Chapter 5) The strong hydrophobic interaction between V4 and the alkyl chains of the lipids led to the highest penetration among the studied peptides Correspondingly melittin and polymyxin B are less

hydrophobic which resulted in lower penetration This result is not consistent with the in

vivo lower hemolytic activity and lower cytotoxic activity of V4 compared to polymyxin

B One possible reason is that V4 might strongly penetrate into POPC membranes and interact with hydrophobic alkyl chains However, due to the high hydrophobicity of V4, the possibility to translocate into the inner layer of the membrane, which faces the

hydrophilic cytosol, is much lower than polymyxin B Therefore V4 has shown higher penentration ability into POPC monolayers but less ability to translocate into the membrane inner layer leading to final cell death

Fig 7.10 shows the penetration kinetics of melittin and V4 into POPC monolayers More time was needed to achieve complete penetration into POPC than POPG monolayers, especially at high peptide concentration for both melittin and V4 The slow penetration into POPC monolayers indicated that charge not only affected the extent of penetration, but also the rate Electrostatic interaction allowed peptide molecules to strongly absorb

on the lipid monolayer, not easy to escape, and reach full penetration rapidly However in the absence of electrostatic interaction, it was difficult for the monolayer to capture the peptide molecules, which diffused away easily, and resulted in more time needed for full penetration Thus the electrostatic interaction played a significant role in penetration

Fig 7.10 Kinetics of melittin and V4 with different concentrations penetrating into POPC monolayers

Fig 7.11 Comparison of V4 penetrating into different lipid monolayers

7.3.2.3 Penetration of V4 into different lipid monolayers

The interaction of V4 with different lipid monolayers was investigated to examine the penetration ability of V4 into different membranes Fig 7.11 illustrates the comparison of V4 penetrating into POPG, POPC, DPPG and a mixed monolayer of POPG/POPE (1/2) With increasing peptide concentration, the penetration of V4 into the monolayers gradually increased until a saturation at a certain peptide concentration between 0.5 µM

to 1 µM Similarly according to ∆π at a saturated peptide concentration at which extra peptide did not induce higher increased surface pressure, the ability of V4 penetrating into the different monolayer was determined Thus the sequence of V4 penetrating into different lipid monolayer was POPG > DPPG > POPG/POPE (1/2) > POPC

Among the studied lipid or mixed lipid monolayers, all but POPC are negatively charged Fig 7.11 showed that V4 had a higher ability to penetrate into these negatively charged lipid monolayers than the zwitterionic POPC monolayer, which confirmed the significant role of electrostatic interaction in the penetration This result is also consistent with the

result that V4 shows less affinity for POPC SUVs compared to negatively charged SUVs

in the binding experiments

Lipid mixture of POPG/POPE (1/2) bears part negative charges, so that less penetration was observed than for pure POPG However, this result is different from the result in the previous study that V4 has a similar binding affinity for the mixed lipid vesicles and POPG vesicles The reason for this difference was that in the binding FCS experiments, lipid was present in a much higher concentration than V4 Once one V4 molecule bound

to a lipid vesicle, the vesicle was detected When more peptide molecules bound to same vesicle, the vesicle number did not change although it was more fluorescent Thus the main factor to determine the binding was lipid vesicle without bound peptide compared

to vesicles containing at least one peptide, allowing fluorescent detection In other words,

it was the first V4 molecule binding to the lipid vesicle that determined the binding Therefore the binding of V4 to mixed lipid SUVs and pure POPG SUVs was similar However in the monolayer experiments, the absorption of the peptide molecules was dependent on the charge amount of the lipid monolayer The more charge the monolayer harbored, the more V4 molecules absorbed on the monolayer, which led to a higher penetration Therefore the amount of charge exerted an important influence on the penetration

POPG and DPPG possess the same headgroup, thus the same negative charge, however V4 showed a slightly higher penetration into POPG than DPPG monolayers An explanation would be the difference in alkyl chain of POPG and DPPG The double bond

of POPG increased the fluidity of the monolayer, leading to less dense lipid packing This provided peptide molecules better access to interact with the alkyl chains and increased penetration DPPG is a saturated phospholipid and lipid molecules are closely packed which hindered the penetration to some extent Therefore the saturation of the lipid molecule contributed to the penetration of V4

In the previous study of V4 binding to different lipid vesicles, V4 showed higher affinity for POPG than for DPPG vesicles However, in this case the difference between these two anionic lipids was not large One possible explanation is the curvature of the membrane The unsaturated property made the POPG form smaller vesicles than DPPG

as shown by the fast diffusion time of POPG vesicles This small sized vesicle has a bigger curvature and allows the head group of the POPG molecule to occupy more space Therefore it is easier for V4 to bind to and insert into the POPG vesicle than DPPG vesicle However, in the monolayer, there is no difference in the curvature and the absorption is mainly dependent on the charge Therefore the different penetration into POPG and DPPG was primarily due to the different hydrophobic interaction between V4 and alkyl chain of lipid after absorption

Above results showed that both electrostatic interaction and hydrophobic interaction contributed to antimicrobial peptide penetration The injected antimicrobial peptide first absorbed on the headgroup of the lipid in the water phase through electrostatic interaction followed by penetration into the alkyl chain region of the monolayer due to hydrophobic interaction Especially the electrostatic interaction was of great importance because it

determined how much peptide absorbed on and penetrated into the monolayer These results offer the evidence for the selectivity of antimicrobial peptides Bacterial membranes are negatively charged whereas the mammalian cell membranes are neutral Thus the cationic antimicrobial peptides preferentially bound or absorbed on the negative bacterial membrane and had more opportunity to insert into the membrane leading to final membrane disruption and bacterial lysis Different from the bacterial membranes, the neutral mammalian cell membranes permitted less binding or absorption and thus less possibility of disruption

In the isotherm experiments, V4 showed similar properties for POPG and POPC When V4 was mixed with POPG or POPC individually, similar excess surface pressure and excess Helmholtz energy were induced Therefore it was concluded that the hydrophobic interaction played the main role in the interaction between V4 and POPG or POPC and electrostatic interaction nearly had no effect Different from the isotherm of V4 mixed with POPG and POPC, V4 displayed apparently higher penetration ability into POPG than POPC monolayer This difference can be explained by the different sample preparation In the isotherm experiments, V4 was mixed with POPG or POPC beforehand and spread on the water surface V4 peptide contacted both headgroup and alkyl chain of lipid molecules during the compression There was no absorption process Thus the interaction between V4 and lipid mainly depended on the hydrophobic interaction and electrostatic interaction did not perform any important function during the compression POPG and POPC had the same alkyl chain and thus similar hydrophobic interaction, leading to similar excess surface pressure and excess Helmholtz energy However in the

penetration experiments, peptide was injected underneath the monolayer Absorption of the peptide molecules on the monolayer was an important step leading to penetration and this step was mainly dependent on the electrostatic interaction Therefore electrostatic interaction exerted a significant influence on the penetration process

The calculation of the lipid amount showed that there were 1.2х10-8 mol lipid molecules

on the water surface At peptide concentration of 0.05, 0.15, 0.5, 1 and 3 µM, the peptide amount was 0.6х10-8, 1.8х10-8, 6х10-8, 12х10-8 and 36х10-8 mol respectively in the trough Thus the corresponding peptide/lipid ratio was 1:2, 3:2, 10:2, 20:2 and 60:2, respectively Thus the above calculation shows that in the penetration experiments, peptide was mostly present at higher amounts than lipids Antimicrobial peptides induced monolayer saturation at a certain peptide concentration Extra peptide kept the increased surface pressure constant and did not induce a surface pressure decrease, indicating that the extra peptide did not cause monolayer rupture This was different from the result obtained from vesicles that antimicrobial peptides induced vesicle disruption at high peptide/lipid ratios

A possible reason might be the different membrane structure Vesicles are symmetrical bilayer structures, which allowed antimicrobial peptides to penetrate the outer layer and translocate into the inner layer After peptide molecules reached the inner layer, they enwrapped the vesicle fragments and broke the vesicles However, monolayer is only half

a membrane with only one lipid layer Antimicrobial peptides could penetrate but they were not capable of enwrappping the monolayer Therefore monolayers cannot be disrupted by the peptides

7.3.2.4 Effect of lipid packing on the penetration of V4

The pressure of vesicles is about 30 mN/m198-201 In order to compare the penetration of V4 into vesicle and monolayer and mimic live cell pressure, the target pressure of 30 mN/m was chosen to investigate the penetration of V4 into monolayers Fig 7.12 illustrates the comparison of V4 penetrating into POPG and mixed POPG/POPE (1/2) monolayers at the target surface pressure of 15 and 30 mN/m, respectively At a target surface pressure of 30 mN/m, 0.05 µM V4 did not induce penetration into the POPG monolayer When the peptide concentration increased to 0.15 µM, there was slight penetration with ∆π of 2.6 mN/m With increasing V4 peptide concentration, ∆π slightly increased At the highest studied peptide concentration of 3 µM, ∆π was 5.3 mN/m Compared with the penetration of V4 into POPG monolayers at target surface pressure of

15 mN/m, the penetration at a high target surface pressure of 30 mN/m was very low Similar to POPG monolayers, V4 exhibited low penetration ability into mixed POPG/POPE monolayer At V4 concentration of 0.05 and 0.15 µM, no penetration was observed From peptide concentration of 0.5 µM, ∆π slightly increased with the increasing V4 concentration with the maximal value of 3.9 mN/m This value was much lower than the ∆π induced by V4 at a target surface pressure of 15 mN/m Therefore it can be concluded that the target surface pressure had an important effect on the penetration

Fig 7.12 V4 penetrates into POPG and POPG/POPC monolayers at target surface pressure of 15 and 30 mN/m

Different target surface pressures represented different lipid packing The higher the target surface pressure was, the denser the lipid monolayer packed The dense lipid packing hindered the insertion of peptide, which resulted in a low ∆π Hence ∆π reduced when the target surface pressure increased Similarly the penetration of the V4 into POPG/POPE monolayer was slightly lower than into POPG monolayer at corresponding peptide concentrations, which confirmed the role of electrostatic interaction in the penetration process

7.3.3 AFM studies of antimicrobial peptides interacting with lipid monolayers

7.3.3.1 AFM images of pure lipid monolayers

The morphology of pure POPG, POPC and DPPG monolayers as references is shown in Fig 7.13 In the absence of antimicrobial peptides, the AFM image of the pure lipid monolayer displayed a regular and flat area for all studied lipids, which indicated that these lipid molecules formed a homogenous monolayer organization The section analysis of the pure lipid monolayer as shown in Fig 7.14 provided the cross-section profile of monolayer with regard to height difference It can be observed that the height difference of the pure lipid monolayer was very small The height difference of the pure POPG, POPC and DPPG monolayer illustrated in the figure by the two arrows was 0.291, 0.226 and 0.377 nm, respectively, which confirmed the flatness of the pure lipid monolayers

Fig 7.13 AFM topographic images of pure POPG, POPC and DPPG monolayers

Fig 7.14 Section analysis of pure POPG, POPC and DPPG monolayers

7.3.3.2 AFM images of POPG monolayers penetrated by antimicrobial peptides

Fig 7.15 shows the AFM images of POPG monolayers penetrated by different antimicrobial peptides at peptide concentration of 50 nM and 500 nM The images revealed heterogeneous organization of the monolayer in the presence of antimicrobial peptides When magainin 2 penetrated into POPG monolayer at relatively low peptide concentration of 50 nM, there were a lot of dark regions, which related to hole-like lower vertical height regions These holes were round and most closely packed When peptide concentration increased to 500 nM, the number of holes decreased Simultaneously the holes became deeper

The penetration of melittin into POPG monolayers induced similar monolayer morphology at both 50 nM and 500 nM The whole image was separated into the bright regions, which indicated the higher vertical domains, and the dark regions, which related

to the lower vertical domains These two kinds of regions were mixed homogeneously with irregular shape, indicative of the random insertion of melittin into POPG monolayer Similar to melittin and magainin 2, the presence of polymyxin B in the POPG monolayer, also caused random distribution of higher and lower vertical domains At peptide concentration of 50 nM, the AFM image was similar to that of magainin 2 and melittin with small and irregular shaped different vertical domains However when the polymyxin

B concentration increased to 500 nM, the topography was similar to that of magainin 2 There were some holes with increased size forming on the surface

V4 displayed different modification on the morphology of the POPG monolayer At peptide concentration of 50 nM, besides the small holes, some thin and long filaments, which were randomly distributed protruding over the monolayer surface, were also observed The filaments illustrated in the figure were up to about 200 nm in length and 10

nm in width These protrusions were probably the aggregation of V4, which has penetrated the monolayer, strongly intracted with lipid molecules and accumulated on the hydrophobic part of POPG monolayer When V4 peptide concentration increased to 500

nM, the filaments, which were randomly distributed at low peptide concentration, connected together to form a network of filaments Simultaneously the small holes were enlarged and filled in the space between the filaments Therefore a regular filament network structure was constructed

Fig 7.15 AFM topographic images of POPG monolayers penetrated by antimicrobial peptides Left column indicates the peptide concentration of 50 nM and right column indicates the peptide concentration of 500 nM From top to bottom the peptide sequence

Fig 7.16 Section analysis of POPG monolayers penetrated by magainin 2 and melittin The top two graphs are magainin 2 at peptide concentration of 50 and 500 nM respectively The bottom two graphs are melittin at peptide concentration of 50 and 500

nM respectively