This Article Abstract Full Text (PDF) All Versions of this Article: biophysj.105.067595v1 90/4/1275    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Neville, F. Articles by Gidalevitz, D. Search for Related Content PubMed PubMed Citation Articles by Neville, F. Articles by Gidalevitz, D.

Lipid Headgroup Discrimination by Antimicrobial Peptide LL-37: Insight into Mechanism of Action

Frances Neville ^*, Marjolaine Cahuzac ^*, Oleg Konovalov , Yuji Ishitsuka , Ka Yee C. Lee , Ivan Kuzmenko , Girish M. Kale ^* and David Gidalevitz ^¶

^* Institute for Materials Research, University of Leeds, Leeds, United Kingdom; European Synchrotron Radiation Facility, Grenoble, France; Department of Chemistry, The Institute for Biophysical Dynamics and The James Franck Institute, The University of Chicago, Chicago, Illinois; Advanced Photon Source, Argonne National Laboratory, Argonne, Illinois; and ^¶ Department of Chemical Engineering, Illinois Institute of Technology, Chicago, Illinois

Correspondence: Address reprint requests to David Gidalevitz, Dept. of Chemical Engineering, Illinois Institute of Technology, Chicago, IL 60616. Tel.: 312-567-3534; Fax: 312-567-8874; E-mail: gidalevitz{at}iit.edu.

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES

 
Interaction of the human antimicrobial peptide LL-37 with lipid monolayers has been investigated by a range of complementary techniques including pressure-area isotherms, insertion assay, epifluorescence microscopy, and synchrotron x-ray scattering, to analyze its mechanism of action. Lipid monolayers were formed at the air-liquid interface to mimic the surface of the bacterial cell wall and the outer leaflet of erythrocyte cell membrane by using phosphatidylglycerol (DPPG), phosphatidylcholine (DPPC), and phosphatidylethanolamine (DPPE) lipids. LL-37 is found to readily insert into DPPG monolayers, disrupting their structure and thus indicating bactericidal action. In contrast, DPPC and DPPE monolayers remained virtually unaffected by LL-37, demonstrating its nonhemolytic activity and lipid discrimination. Specular x-ray reflectivity data yielded considerable differences in layer thickness and electron-density profile after addition of the peptide to DPPG monolayers, but little change was seen after peptide injection when probing monolayers composed of DPPC and DPPE. Grazing incidence x-ray diffraction demonstrated significant peptide insertion and lateral packing order disruption of the DPPG monolayer by LL-37 insertion. Epifluorescence microscopy data support these findings.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES

 
A worldwide increase in multidrug-resistant bacteria (1–4) has made it imperative to develop new classes of antibiotics. Certain antimicrobial peptides of the innate immune system can distinguish membranes having different lipid compositions (5–9). This property allows them to lyse bacterial membranes while leaving eukaryotic plasma membranes unaltered. A better understanding of this ability at a molecular level could enhance the design and development of antimicrobial peptides as alternatives to the conventional antibiotics used today.

During the last decade it has been established that antimicrobial peptides act by permeating cell membranes (7,10–12). Two main mechanisms of action of membrane perturbation by antimicrobial peptides have been previously proposed: the barrel stave model (13–17) and the carpet model (11,12,18–22).

In the barrel stave mechanism, peptide monomers bind to the lipid membrane and then assemble together to form bundles, in which the hydrophobic surfaces interact with the acyl chain part of the membrane. The hydrophilic regions align together to form a pore which may be increased in size by additional peptide monomers.

In the carpet model, peptides bind to the outside of the membrane by adsorption to the headgroup region of the lipid membrane, carpeting the surface of the membrane and orienting parallel to it. Membrane disruption may occur due to the formation of pores in the membrane or via membrane micellization. When a critical peptide/lipid ratio is reached, the peptides are thought to change orientation to a direction perpendicular to the membrane (23). It is these reorientated peptides that can then form the toroidal pores (7,22–24). The formation of toroidal pores may then be followed by the complete collapse of the membrane.

It has been proposed that human antimicrobial peptide LL-37 (LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES-NH₂) acts by perturbing membranes via the carpet mechanism of action (19,25,26). Oren et al. (19) suggest that LL-37 carpets the surface of both zwitterionic phosphatidylcholine (PC) and negatively charged PC/phosphatidylserine (PS) vesicles. They suggest that LL-37 oligomerizes in solution, and although it is self-associated when bound to zwitterionic phospholipid vesicles, it dissociates into monomers upon binding to negatively charged vesicles. Henzler-Wildman et al. (25,26) also demonstrate that LL-37 first carpets lipid bilayers before membrane perturbation by using solid-state nuclear magnetic resonance and differential scanning calorimetry experiments. Their results support the carpet mechanism of action of lipid bilayer disruption by LL-37.

One of the most important stages of peptide-membrane interaction is an initial contact of the peptide with the outer leaflet of the plasma membrane. To investigate what defines the propensity of peptides to interact with specific membrane lipids, methods that allow the membrane to be modeled in a fluid environment are needed. One such approach is the use of a Langmuir monolayer to mimic the external leaflet of the cell membrane, coupled with the introduction of peptides into the subphase of a Langmuir trough to represent the extracellular fluid and thus the approach of the peptide toward the cell surface. Different lipid compositions can be used to represent the membranes of different cell types, and changes in membrane structure resulting from its interaction with peptides allow us to suggest a possible mechanism of peptide action.

The outer leaflet of mammalian cell membranes mainly comprises PC, phosphatidylethanolamine (PE), sphingomyelin, and cholesterol, which are charge-neutral at physiological pH.

In this work, we used dipalmitoylphosphatidylcholine (DPPC) and dipalmitoylphosphatidylethanolamine (DPPE) to study the interaction of LL-37 with the human red blood cell membrane. Nouri-Sorkhabi et al. (27) used P-31 NMR spectra of human erythrocyte lysates to find that PE accounts for 33.3% and PC makes up 30.3% of human erythrocyte membrane mass. Independently, Keller et al. (28) found that human red blood cell membrane contains on average 34% PE and 35% PC, although their distribution among the leaflets is asymmetrical: the majority of PC (28% total lipid content) is at the outer leaflet, whereas the majority of PE (28% total lipid content) is within the inner leaflet.

The surfaces of both Gram-negative and Gram-positive bacterial cell walls contain large amounts of negatively charged lipids (29,30). The outer layer of Gram-positive bacteria cell wall is composed of acidic polysaccharides (teichoic acids) and phosphatidylglycerol (PG) (31), whereas the outer leaflet of the outer membrane bilayer of the Gram-negative bacteria is predominantly composed of lipopolysaccharide (LPS), a polyanionic molecule (29,30,32–34).

In this article, we look into the interaction of LL-37 with DPPG lipid. In a consequent article (F. Neville, C. S. Hodges, C. Liu, O. Konovalov, and D. Gidalevitz, unpublished), we use lipid A, the main lipid component of lipopolysaccharides, to study Gram-negative membrane lysis by LL-37, protegrin-1, and SMAP-29 antimicrobial peptides.

A similar approach has been successfully used in conjunction with x-ray surface scattering and epifluorescence techniques to study interactions of phospholipid monolayers with the porcine antimicrobial peptide protegrin-1 (35). Surface x-ray scattering techniques yield information about the packing arrangement of the lipid layer before and after introduction of antimicrobial peptide into the system, whereas epifluorescence microscopy allows direct observation of the changes in lipid monolayer morphology caused by peptides. Langmuir monolayer techniques and surface x-ray scattering have also been successfully applied to study interactions of an amphibian antimicrobial peptide, PGLa (9). More recently, preliminary studies on the human antimicrobial peptide LL-37 using electrochemical techniques and epifluorescence microscopy have been carried out (36).

In this article, the interaction of LL-37 with phospholipid monolayers is further studied using a combination of pressure area isotherms, specular x-ray reflectivity (XR), grazing incidence x-ray diffraction (GIXD), and epifluorescence microscopy (EFM) in a complementary manner, with the purpose of probing the question of antimicrobial peptide selectivity between prokaryotic and eukaryotic cells and its possible mechanism of action. Discrimination of LL-37 between monolayers modeling outer layers of red blood cells and bacterial membranes has been demonstrated. Although no interactions of LL-37 with DPPC or DPPE monolayers were observed, significant insertion of LL-37 into DPPG monolayers was documented with each of the experimental techniques used. The mechanism of LL-37 insertion into DPPG monolayers is consistent with a "carpet" mechanism proposed earlier.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES

 
Lipid monolayers
To study peptide-membrane interactions, Langmuir monolayers composed of lipids representative of membranes were studied. This approach has been utilized repeatedly over the years (37–43). The planar monolayer system allows versatile adjustments of parameters such as monolayer composition, surface pressure, packing density, temperature, and aqueous subphase condition to model the conditions under which peptides approach the outer surface of the membrane from the extracellular medium. The cell membranes of different organisms have characteristic lipid compositions. However, to develop a full understanding of membrane interactions, each component of a membrane should be studied separately to ascertain the contribution of each membrane component to the overall interaction of the peptide with the membrane. DPPC, DPPE, and dipalmitoylphosphatidylglycerol (DPPG) were used to form Langmuir monolayers. All lipids were purchased from Avanti Polar Lipids (Alabaster, AL) and were used without further purification.

Pressure-area compression isotherms
All experiments were performed using Dulbecco's phosphate buffered saline (DPBS) (Invitrogen, Carlsbad, CA) without calcium and magnesium ions. Monolayers of DPPC and DPPE were deposited from chloroform (high-performance liquid chromatography grade, Fisher Scientific, Pittsburgh, PA) solution, whereas DPPG was deposited from 9:1 v/v % chloroform/methanol solution (high-performance liquid chromatography grade, Fisher Scientific). Upon spreading, the lipid film was left undisturbed for 15 min to allow for solvent evaporation. At this point, barrier compression was initiated and the increase in surface pressure of the monolayer was monitored. This gives rise to a surface pressure (mN/m) versus area per lipid molecule (Ų/molecule) isotherm, which can be utilized to deduce the phases and phase transitions associated with the monolayer as a function of lateral lipid packing density.

Insertion experiments
Insertion experiments were carried out to quantify the interaction of LL-37 with the lipid monolayer. Initially, the lipid monolayer was deposited and equilibrated, followed by compression to the surface pressure corresponding to the liquid-condensed phase of lipids (30–40 mN/m), and equivalent to the packing density of the cell membrane (44,45). The surface pressure was kept constant via a built-in proportional-integral-derivative control feedback system by adjusting the surface area. The LL-37 solution (10 µg/ml LL-37 in 0.01% glacial acetic acid w/v) was then evenly injected underneath the monolayer using a microsyringe with an L-shaped needle (VDRL needle; Hamilton, Reno, NV) to make up the final concentration of 0.04 µg/ml or 0.1 µg/ml. The surface pressure immediately increased as a result of peptide incorporation into the lipid monolayer. Injected peptides interact with the lipid monolayer and result in an increase in the surface pressure. To keep the surface pressure constant, the surface area would have to increase. The resulting relative change in area per molecule, A/A, was monitored throughout the experiment to compare the degree of LL-37 insertion into DPPC, DPPE, and DPPG monolayers. Epifluorescence microscopy was used concurrently with insertion experiments to monitor the surface morphology of the monolayers on insertion of LL-37; this involved the incorporation of a small amount of fluorescent dye into the different lipid-spreading solutions.

LL-37 at the interface
LL-37 (90–95% pure) was supplied by Pepceuticals (Nottingham, UK). LL-37 peptide was provided as a solid powder and made up to a working solution (10 µg/ml LL-37 in 0.01% glacial acetic acid w/v). Acetic acid was used to maintain the peptide structure while in solution. LL-37 is soluble in water, but, being amphipathic (46), is expected to be adsorbed at the air-liquid interface. LL-37 was injected into the pure subphase without a lipid monolayer present. The properties of pure LL-37 monolayer were then investigated using pressure-area isotherms in conjunction with x-ray reflectivity and grazing incidence x-ray diffraction.

Langmuir troughs
Pressure-area compression isotherms were performed using a twin-barrier rectangular Teflon micro Langmuir-Blodgett trough equipped with a Wilhelmy plate (Nima Technology, Coventry, UK). Insertion experiment data presented here were obtained using a custom-built trough (35,47) at the University of Chicago. X-ray scattering measurements were taken at the European Synchrotron Radiation Facility (ESRF) (9) and Advanced Photon Source (APS) synchrotrons, both of which utilizing custom-built Langmuir troughs equipped with a single moveable barrier. The subphase temperature was maintained at 22 ± 1°C for lateral compression and x-ray scattering experiments, and at 30 ± 1°C for insertion assays.

Epifluorescence microscopy
The Langmuir trough used for insertion experiments was equipped with an epifluorescence microscope mounted to observe the phase morphology of the lipid monolayer. The epifluorescence microscopy techniques were carried out as previously described (36,47–50). A resistively heated indium tin-oxide-coated glass plate was placed over the trough to minimize contamination and condensation on the microscope objective lens and to minimize. Excitation between 530 and 590 nm and emission between 610 and 690 nm was gathered through the use of an HYQ Texas Red filter cube. Lipid-linked Texas Red dye ((TR-DHPE) Molecular Probes, Eugene, OR; 0.5 mol %) was incorporated into the spreading phospholipid solutions. Due to steric hindrance, the dye partitions into the disordered phase, rendering it bright and the ordered phase dark. This assembly permits the monolayer morphology to be observed over a large lateral area while isotherm data are obtained concurrently.

GIXD and XR
Surface x-ray scattering experiments were carried out at the ID10B (Troïka II) beam line at the ESRF (Grenoble, France), and at the 9-ID (CMC-CAT) beam line at the APS, Argonne National Laboratory (Argonne, IL). The oxygen level was monitored as the vessel was filled with water-saturated helium and was allowed to reach a sufficiently low level before measurements were commenced (<0.1%). Water-saturated helium was used to reduce evaporation and scattering from the air. Control measurements of pure lipid monolayers were followed by subsequent injection of the desired amount of peptide into the subphase under the monolayer-covered area of the surface at 30 mN/m.

GIXD (51) was used to obtain in-plane information concerning the molecular structure of surfaces (52). For GIXD measurements, the angle of incidence (_i) was set to strike the air-aqueous interface at 0.8 _c, where _c is the critical angle for total external reflection. A linear position-sensitive detector that detected the diffracted beam recorded the intensity profile as a function of scattering angle. A Soller collimator with horizontal plates was placed after the detector. Such a configuration enables the incident wave to be reflected, whereas the refractive wave propagates along the surface, making this technique very sensitive to changes at the air-aqueous interface. The incident wavelength used at the ESRF was 1.54 Å, and the wavelength at APS was 0.92 Å.

Specular x-ray reflection measurements reveal information on the electron-density distribution along the surface normal and may be used to determine the density and thickness of thin layers (41,51,53). When specular x-ray reflection occurs, the scattering vector q_z may be calculated from q_z = 4 sin /, where is the grazing angle of the incident beam and the wavelength of the x-ray beam. When the reflectivity is measured as a function of the scattering vector q_z, the reflectivity curve contains information regarding the gradient of the electron-density profile in the direction normal to the surface (54,55). XR measurements were carried out over a range of angles corresponding to q_z values of 0–0.65 Å^–1. A position-sensitive detector was used for detection and the reflected beam intensity was measured as a function of the incident angle.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES

 
Pressure-area compression isotherms
Pressure-area (-A) isotherms are a traditional method for the investigation of Langmuir monolayer phase behavior. As the area decreases at constant temperature, any increase in pressure on compression may be recorded.

Fig. 1 A shows typical -A isotherms of DPPC, DPPE, and DPPG that exhibit trends similar to those described in earlier publications (35,50,56), as well as the -A isotherm of LL-37. As found from the Langmuir isotherm, the area per DPPG molecule in a monolayer compressed at 30 mN/m is 45.6Ų and that for DPPC is 46.4Ų at the same pressure. These values will be compared with area per molecule values found from grazing incidence x-ray diffraction data.

View larger version (18K): [in this window] [in a new window]   FIGURE 1  Pressure-area isotherms of pure lipids DPPC (solid line), DPPE (dashed line), and DPPG (dash-dotted line), and pure LL-37 peptide (dotted line). LL-37 isotherm has been presented at a scale of one-fifth of its real area per molecule; pressure values are those actually recorded.

Constant-pressure insertion isotherms
Constant-pressure isotherm experiments involve compression of the monolayer to the desired pressure followed by peptide injection into the subphase. Any increase or decrease in the surface pressure from the setpoint after peptide insertion will result in the compression or expansion of the barrier(s) to advert the change. Increase in the surface pressure value after injection is due to peptide insertion into the lipid layer and will therefore result in barrier expansion.

Constant-pressure insertion isotherms show little or no LL-37 insertion into DPPC and DPPE monolayers (Fig. 2, A and B), whereas a substantial increase in DPPG area per molecule indicates incorporation of peptide molecules into the monolayer structure (Fig. 2, A and B). However, when DPPG surface pressure is increased to 40 mN/m and peptide is injected to reach a subphase concentration of 0.1 µg/ml (Fig. 2 C), a critical destabilization of the monolayer takes place.

View larger version (16K): [in this window] [in a new window]   FIGURE 2  Insertion isotherms of DPPC and DPPG showing percentage change in area per molecule after injection of LL-37. (A) Systems at concentration 0.04 µg/ml LL-37: DPPG + LL-37 at 30 mN/m (solid line); DPPC + LL-37 at 30 mN/m (dashed line); DPPE + LL-37 at 30 mN/m (dotted line); DPPG + LL-37 at 40 mN/m (dash-dotted line). (B) Systems at concentration 0.1 µg/ml LL-37: DPPG +LL-37 at 30 mN/m (solid line); DPPG + LL-37 at 40 mN/m (dashed line); DPPC + LL-37 at 30 mN/m (dotted line). (C) Critical destabilization of DPPG at 40 mN/m after injection of 0.1 µg/ml LL-37 (dashed line). (Concentration values refer to final concentrations of peptide in solution.)

When the peptide concentration reaches 0.1 µg/ml (Fig. 2 B), a pronounced difference is seen between the insertion isotherms taken at 30 mN/m and those taken at 40 mN/m. A very large increase in area (180%) was seen on injection of 0.1 µg/ml LL-37 under the DPPG monolayer at 30 mN/m. However, when the pressure of the system was kept at 40 mN/m, injection of the peptide resulted in insertion during the first 5–10 min and then a gradual decrease in area to a value below that seen before the introduction of the peptide (Fig. 2 C). This effect is very similar to that of introducing the porcine PG-1 peptide into a lipid A system at 35 mN/m (35). It is likely that at this higher concentration and pressure, a critical threshold has been reached, resulting in a critical destabilization of the DPPG monolayer. This result is consistent with permeabilization of the bacterial membrane by LL-37 via the "carpet" route, which involves a threshold concentration of peptide to carpet the membrane before any pore formation can occur.

Epifluorescence
EFM measurements were performed simultaneously with the insertion isotherms to monitor the effect of peptide binding on the morphology of the monolayer. Fluorescence image contrast arises due to different phase densities and partitioning characteristics of the dye molecules in coexisting phases. Therefore, it is possible to gain insight into the structure of the lipid layer by imaging its lateral fluorescence distribution.

EFM images of the pure lipid monolayers at both 30 and 40 mN/m display an array of very densely packed "dark gray" domains of condensed phase with narrow, "light gray" liquid-phase borders between them. Little or no change in DPPE and DPPC morphology was observed after injection of LL-37 at either 0.04- or 0.1-µg/ml concentration (Fig. 3, A and C). A very slight increase in the amount of bright disordered phase of DPPC at the higher peptide concentration was observed after 19 min (Fig. 3 C3).

View larger version (142K): [in this window] [in a new window]   FIGURE 3  Epifluorescence images of systems at 30 mN/m. (A1) DPPE monolayer at 30 mN/m before injection. (A2) DPPE monolayer 1 min after 0.04 µg/ml LL-37 injection. (A3) DPPE monolayer 2 min after 0.04 µg/ml LL-37 injection. (B1) DPPC monolayer at 30 mN/m before injection. (B2) DPPC monolayer 1 min after 0.04 µg/ml LL-37 injection. (B3) DPPC monolayer 6 min after 0.04 µg/ml LL-37 injection. (C1) DPPC monolayer at 30 mN/m before injection. (C2) DPPC monolayer 18 min after 0.1 µg/ml LL-37 injection. (C3) DPPC monolayer 19 min after 0.1 µg/ml LL-37 injection. (Concentration values 0.04 µg/ml and 0.1 µg/ml refer to final concentrations of peptide in solution.)

View larger version (129K): [in this window] [in a new window]   FIGURE 4  (A1) DPPG monolayer at 30 mN/m before injection and at 2 min (A2), 3 min (A3), 4 min (A4), 10 min (A5), and 20 min (A6) after LL-37 injection. (B1) DPPG monolayer at 30 mN/m before injection and at 4 min (B2) and 7 min (B3) after 0.1 µg/ml LL-37 injection. (Concentration values refer to final concentrations of peptide in solution.)

View larger version (188K): [in this window] [in a new window]   FIGURE 5  Epifluorescence images of DPPG + LL-37 at 40 mN/m. (A1) DPPG monolayer at 40 mN/m before injection, and at 3 min (A2), 6 min (A3),7 min (A4), 10 min (A5), and 13 min (A6) after 0.04 µg/ml LL-37 injection. (B1) DPPG monolayer at 40 mN/m before injection and at 8 min (B2), 14 min (B3), 31 min (B4), 63 min (B5), and 92 min (B6) after 0.1 µg/ml LL-37 injection. (Concentration values 0.04 µg/ml and 0.1 µg/ml refer to final concentrations of peptide in solution.)

Grazing incidence x-ray diffraction
Grazing incidence x-ray diffraction measurements are made with the x-ray momentum transfer in or close to the plane of the air-aqueous interface. The reflections of the Bragg peaks observed with this geometry can be indexed by two Miller indices, hk. Their angular positions 2_hk, corresponding to q_hk = (4/)sin_hk, yield the repeat distances d_hk = 2/q_hk for the two-dimensional lattice structure (40,55).

GIXD was performed in conjunction with insertion isotherms (not shown here but similar to those taken with fluorescence experiments) to monitor the effect of peptide insertion on the molecular packing of the lipid monolayers of different compositions. Bragg peak profiles (intensity against q_xy) were fitted with Gaussians and the peak position values were used to obtain unit-cell dimensions of the lipid lattices, whereas full-width half-maximum values of the peaks were used to determine the coherence length from the Scherrer formula, L = 0.9 x 2/full-width half-maximum (q_xy) (61,62). Bragg rod profiles were measured at Bragg peak positions. Bragg rod profiles were analyzed to determine the tilt of the hydrocarbon chains (52).

The observation of two Bragg peaks in the diffraction pattern of an amphiphilic monolayer is indicative of a distorted hexagonal unit cell (which can be viewed as centered rectangular). Therefore, all unit-cell dimensions in this study were calculated using the centered rectangular unit-cell approximation.

GIXD—DPPC
Pure DPPC monolayer data (Fig. 6 A1) show two Bragg peaks corresponding to d-spacings of 4.60 Å and 4.26 Å. This translates to a centered rectangular unit cell with dimensions of a = 5.47 Å and b = 8.53 Å and an area per DPPC molecule of A = 46.6 Ų, as each unit cell contains two hydrocarbon chain components of the phospholipid molecule. This area per molecule value agrees well with that obtained from the pressure-area isotherm (Fig. 1) and those obtained in previous studies (9,54,57,63–65). The coherence lengths, calculated with the Scherrer formula (52,62), give values of 126 Å and 1061 Å for the {1,1} (L₁₁) and {0,2} (L₀₂) reflections, respectively. Analysis of Bragg rod profile yields a molecular tilt of 30° for the lipid molecules in the condensed phase. These values are again very close to those obtained by other authors (9,54,57,63–65). The area per molecule obtained from GIXD data is valid only for the ordered part of the monolayer. The fact that the area per molecule from the -A isotherm corresponds to values obtained from GIXD indicates that the monolayer is mostly ordered, corroborating our EFM observations.

View larger version (74K): [in this window] [in a new window]   FIGURE 6  Bragg peak/rod combination plot of scattering vectors qxy and qz as a function of their respective intensities. (A) DPPC monolayer at 30 mN/m. (B) DPPC monolayer at 30 mN/m after 0.04 µg/ml LL-37 injection into the subphase. (C) DPPG monolayer at 30 mN/m. (D) DPPG monolayer at 30 mN/m after 0.04 µg/ml LL-37 injection into the subphase. The initial Bragg peaks from scattering of ordered structure of the monolayer vanish upon LL-37 insertion into the DPPG monolayer, whereas the DPPC monolayer was left practically unchanged. (Concentration values 0.04 µg/ml and 0.1 µg/ml refer to final concentrations of peptide in solution.)

GIXD—DPPG
Pure DPPG at 30 mN/m (Fig. 6 B1) yields two peaks corresponding to d-spacings of 4.51 Å and 4.25 Å, centered-rectangular unit-cell dimensions of a = 5.32 Å and b = 8.51 Å, and an area per molecule of A = 45.2 Ų. This value is again in good agreement with that obtained using the area-pressure compression isotherm and with previously published GIXD results (9,35,57). The coherence lengths for a DPPG monolayer at 30 mN/m are calculated to be L₁₁ = 75 Å and L₀₂ = 448 Å. The molecular tilt is found to be 27° for the DPPG monolayer.

After LL-37 insertion, no Bragg peaks or rods were observed, indicating that the ordered structure of the DPPG monolayer has been totally disrupted by LL-37 (Fig. 6 B2). This is again confirmed by insertion isotherms (Fig. 2) and epifluorescence data (Fig. 4), which show increases in area per molecule and an increased area fraction of the disordered bright phase, respectively, demonstrating disruption of the membrane by the peptide. For easy comparison of the GIXD data, results for both the DPPC and DPPC lipid systems are shown in Table 1.

XR—DPPC
Fig. 7 A shows the reflectivity curve normalized to the Fresnel reflectivity of a planar interface of the subphase (R/R_F) against the scattering vector (q) in the z direction (q_z) for the DPPC monolayer at 30 mN/m before and after injection of LL-37 into the subphase. It shows the measured experimental data obtained and the calculated least-square fitted model. Successful fits to the data demonstrated that the DPPC monolayer may be modeled as two slabs with different thicknesses and electron densities. The thickness of the tail layer was found to be 15.2 Å and the headgroup layer thickness 8.8 Å, with tail- and head-layer electron-density values (normalized to the electron density of the subphase) of 0.91 and 1.33, respectively. The roughness of the layers was found to be 3.3 Å. These values agree well with previously published values (54,63,68).

View larger version (18K): [in this window] [in a new window]   FIGURE 7  Reflectivity data and corresponding fits normalized by Fresnel reflectivity plotted against scattering vector (qz). (A) DPPC monolayer reflectivity data (square) and fit (solid line), DPPC after 0.04 µg/ml LL-37 injection, and data (circle) and fit (dashed line) 0.1 min after 0.04 µg/ml LL-37 injection; (A4) 7 min after 0.04 µg/ml LL-37 injection; and (A5) 10 min after 0.04 µg/ml LL-37 injection. (A6) DPPG monolayer 13 min after 0.04 µg/ml LL-37 injection. (B1) DPPG monolayer at 40 mN/m before injection and at 8 min (B2), 14 min (B3), 31 min (B4), 63 min (B5), and 92 min (B6) after 0.1 µg/ml LL-37 injection. (Concentration values 0.04 µg/ml and 0.1 µg/ml refer to final concentrations of peptide in solution.)

XR—DPPG
The R/R_F plot in Fig. 7 B shows the DPPG data obtained at 30 mN/m. The best fit was again obtained using two slabs, as with DPPC, yielding 18.1 Å for the tail layer thickness and 5.9 Å for the headgroup layer thickness. The normalized electron-density values for the tail and headgroup layers were 0.98 and 1.55, respectively, with a roughness value of 3.7 Å for both. These results are in line with previous results (35,77), with small differences most likely attributable to the use of different subphases.

Fig. 7 B shows substantial change in the reflectivity profile upon injection of LL-37 under the DPPG monolayer at 30 mN/m. Since the minima that are observable in R/R_F plots are due to interference of x-ray waves reflected from the interfaces of the slabs in the surface layer, the minima are characteristic of the thickness of the lipid monolayer. After injection of 0.04 µg/ml LL-37, the XR profile is very different from that of the DPPG monolayer alone, with the minimum of the curve being shifted to a greater q_z value by 0.11 Å^–1. This indicates a decrease in monolayer thickness, similar to that observed by Huang et al. (78). Although the amplitude of the second peak was reduced, it was possible to fit the data using a two-slab model (Fig. 8). The top slab has been modeled as the lipid tailgroups, with peptide inserted slightly into the hydrocarbon tail region, and the second slab has been modeled as a mix of headgroups and peptide. The least-square fit parameters gave 7.4 Å for the thickness of the top layer (the layer closest to the air), with a normalized electron density of 1.23, and 8.1 Å for the thickness of the second layer, with a density of 1.33. The roughness obtained was 5.6 Å. The results of the fitting suggest that the peptide has penetrated the headgroup region, as the electron density has decreased by 17%. It is clearly seen that the peptide has also partially penetrated into the tail region, destabilizing it, as its electron density has increased by 26%. All the results of the XR data fitting are presented in Table 2 for clarity.

View larger version (33K): [in this window] [in a new window]   FIGURE 8  Cartoon schematic of possible interactions of LL-37 with bacterial membranes. (A) DPPG. (B) DPPG after insertion of LL-37. Cartoon based on x-ray data (GIXD and XR).

XR—LL-37
X-ray reflectivity data for LL-37 (Fig. 7 C) show slight differences in the structure of the film when the pressure was increased from 0 to 28 mN/m. A small increase in layer thickness was observed between these two pressures, suggesting that the LL-37 molecules become more tightly packed after the film compression and thus the molecular rearrangement produces a thickening of the peptide layer. Moreover, XR analysis indicates a second partial layer of LL-37 forming underneath the top layer. XR data at 0 mN/m was modeled as a one-slab model with an electron density normalized to the electron density of the subphase of 1.18 and a thickness of L = 7.7 Å. This suggests that LL-37 aligns at the air-liquid interface with the molecules oriented parallel to the surface. XR data at = 28 mN/m can no longer be fitted with a single slab and it was necessary to use a two-slab model. It is most likely that at 28 mN/m there is not enough space for every LL-37 molecule to align at the surface in the same way as at 0 mN/m, so some molecules would get forced out from a single layer to make a partial second layer. The fitting of the data using a two-slab model returned values of L₁ = 8.5 Å and L₂ = 4.3 Å for the two layers (going from the air-liquid interface toward the subphase). Corresponding normalized density values for the layers were 1.13 and 0.96, which could correspond to an upper layer and a partial second layer consisting of some peptide molecules that might be partly in the upper layer, with the more hydrophilic regions extending into the subphase. The fact that the LL-37 XR data do not show defined structural features, such as those seen for lipids like DPPC and DPPG, suggests that the LL-37 peptide layer is somewhat amorphous, as is seen with other protein structures at the air-liquid interface (9,79). This is corroborated by GIXD data (not shown), which show no Bragg peaks at either pressure (0 or 28 mN/m) and thus demonstrate that LL-37 does not show long-range lateral two-dimensional structure at the air-aqueous interface.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES

 
This is the first study we know of in which x-ray reflectivity and grazing incidence x-ray diffraction were used to observe the effects of the antimicrobial peptide LL-37 on phospholipid monolayers at the air-liquid interface. Moreover, synchrotron surface x-ray scattering has been combined with epifluorescence microscopy to give a further insight into the interactions of LL-37 with lipid monolayers.

GIXD experiments show that there is very little change in the phosphatidylcholine lipid packing upon addition of LL-37 to the subphase. This indicates that there is no detrimental interaction between the human antimicrobial peptide and the DPPC monolayer at the concentrations used here, contradicting the results of Oren et al. (19) and Henzler-Wildman et al. (26), who observed LL-37 lytic activity toward human red blood cell membrane mimics. However, our working peptide concentrations of 0.04 µg/ml (9 nM) and 0.1 µg/ml (22 nM) were significantly lower, and it is entirely possible that with an increase of LL-37 concentration to MIC levels (1–10 µM) we would observe peptide insertion into DPPC. This is in stark contrast to the DPPG monolayer, whose structure is completely destroyed after injection of LL-37. These results agree well with insertion assay and epifluorescence data, which show LL-37 insertion into DPPG, but not into DPPC, films.

The GIXD data obtained from the LL-37 peptide alone reveal that it does not form an ordered two-dimensional crystalline structure at the air-aqueous interface although it is able to form a film at the surface, as demonstrated by the pressure-area isotherms carried out with the peptide. XR data of the peptide alone corroborate the GIXD findings, suggesting that the peptide molecules self-assemble in-plane with the surface, forming layers of approximately the thickness expected for an -helical peptide. These results prove very useful in suggesting a proposed model for the lipid-peptide interactions of the bacterial membrane with LL-37.

As expected, the x-ray reflectivity model fits suggest that the DPPC monolayers show little change after addition of the LL-37 to the subphase. Even increasing the concentration of the peptide by 2.5 times the original concentration makes little difference in the results obtained. However, the XR data of the DPPG/LL-37 system suggest that LL-37 molecules penetrate in-plane with the monolayer and insert predominantly into the phospholipid headgroup region, partitioning the peptide partly into the hydrocarbon tail region. The peptide/lipid ratio at the interface can be calculated analyzing insertion isotherms and x-ray reflectivity data. The insertion isotherm yields an area per lipid molecule increase caused by peptide absorption at the interface, whereas x-ray reflectivity yields the thickness of the peptide layer, and, thus, the peptide insertion angle. For DPPG at 30 mN/m this gives us a lipid/peptide ratio of 28:1 for peptide concentrations of 0.04 µg/ml and 6:1 for those of 0.1 µg/ml.

This is in agreement with recent work using other techniques (25,26,36), which proposes that the peptide carpets bacterial membranes. The work presented here also suggests that the LL-37 carpets the outside of the membrane before complete membrane disruption.

In summary, the research presented here used Langmuir-trough-generated lipid monolayers to investigate the lipid discrimination of LL-37 and its mechanism of action. By coupling synchrotron x-ray scattering and epifluorescence microscopy techniques with the use of Langmuir films, the interactions of LL-37 with phospholipid monolayers can be directly examined. Our results suggest that LL-37 can differentiate between eukaryotic and bacterial cell membrane lipid types. This work also supports previous work suggesting that the human antimicrobial peptide LL-37 acts against bacterial membranes via the "carpet" mechanism of membrane perturbation. The knowledge of the discrimination of LL-37 for different cell types and its mechanism of action bodes well for the production of future pharmaceutical therapeutic agents in the ongoing battle against antibiotic-resistant bacterial disease.

	ACKNOWLEDGEMENTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES

 
We thank Jeff Keen and Deidre Devine for providing the peptide. We acknowledge the European Synchrotron Radiation Facility for providing synchrotron radiation facilities at the ID10B beam line.

Use of the Advanced Photon Source was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, under Contract No. W-31-109-ENG-38. This project is sponsored by the Engineering and Physical Sciences Research Council, Swindon, UK. Y.I. and K.Y.C.L. are grateful for the support of the Packard Foundation (99-1465).

Submitted on May 27, 2005; accepted for publication September 29, 2005.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS AND DISCUSSION CONCLUSIONS ACKNOWLEDGEMENTS REFERENCES

 
1. Tonks, A. 1994. Drug resistance is a worldwide threat, warns report. Brit. Med. J. 309:1109–1110.[Free Full Text]

2. Hart, C. A., and S. Kariuki. 1998. Antimicrobial resistance in developing countries. BMJ. 317:647–650.[Free Full Text]

3. Davies, J. 1996. Bacteria on the rampage. Nature. 383:219–220.[CrossRef][Medline]

4. Reacher, M. H., A. Shah, D. M. Livermore, M. C. J. Wale, C. Graham, A. P. Johnson, H. Heine, M. A. Monnickendam, K. F. Barker, D. James, and R. C. George. 2000. Bacteraemia and antibiotic resistance of its pathogens reported in England and Wales between 1990 and 1998: trend analysis. Brit. Med. J. 320:213–216.[Abstract/Free Full Text]

5. Andreu, D., and L. Rivas. 1998. Animal antimicrobial peptides: an overview. Biopolymers. 47:415–433.[CrossRef][Medline]

6. Matsuzaki, K., K. Sugishita, N. Fujii, and K. Miyajima. 1995. Molecular basis for membrane selectivity of an antimicrobial peptide, magainin 2. Biochemistry. 34:3423–3429.[CrossRef][Medline]

7. Matsuzaki, K. 1999. Why and how are peptide-lipid interactions utilized for self-defense? Magainins and tachyplesins as archetypes. Biochim. Biophys. Acta. 1462:1–10.[Medline]

8. Zasloff, M. 2002. Antimicrobial peptides of multicellular organisms. Nature. 415:389–395.[CrossRef][Medline]

9. Konovalov, O., I. Myagkov, B. Struth, and K. Lohner. 2002. Lipid discrimination in phospholipid monolayers by the antimicrobial frog skin peptide PGLa. A synchrotron X-ray grazing incidence and reflectivity study. Eur. Biophys. J. 31:428–437.[CrossRef][Medline]

10. Matsuzaki, K. 1998. Magainins as paradigm for the mode of action of pore forming polypeptides. Biochim. Biophys. Acta. 1376:391–400.[Medline]

11. Oren, Z., and Y. Shai. 1998. Mode of action of linear amphipathic alpha-helical antimicrobial peptides. Biopolymers. 47:451–463.[CrossRef][Medline]

12. Shai, Y. 1999. Mechanism of the binding, insertion and destabilization of phospholipid bilayer membranes by -helical antimicrobial and cell non-selective membrane-lytic peptides. Biochim. Biophys. Acta. 1462:55–70. match the journal[Medline]

13. Ehrenstein, G., and H. Lecar. 1977. Electrically gated ionic channels in lipid bilayers. Q. Rev. Biophys. 10:1–34.[Medline]

14. Westerhoff, H. V., D. Juretic, R. W. Hendler, and M. Zasloff. 1989. Magainins and the disruption of membrane-linked free-energy transduction. Proc. Natl. Acad. Sci. USA. 86:6597–6601.[Abstract/Free Full Text]

15. Matsuzaki K., M. Harada, S. Funakoshi, N. Fujii and K. Miyajima. 1991. Physicochemical determinants for the interactions of magainins-1 and magainins-2 with acidic lipid bilayers. Biochim. Biophys. Acta. 1063:162–170.[Medline]

16. Rapaport, D., and Y. Shai. 1991. Interaction of fluorescently labeled pardaxin and its analogs with lipid bilayers J. Biol. Chem. 266:23769–23775.[Abstract/Free Full Text]

17. Shai, Y. 1994. Pardaxin-channel formation by a shark repellant peptide from fish. Toxicology. 87:109–129.[CrossRef][Medline]

18. Pouny, Y., D. Rapaport, A. Mor, P. Nicolas, and Y. Shai. 1992. Interaction of antimicrobial dermaseptin and its fluorescently labeled analogs with phospholipid membranes. Biochemistry. 31:12416–12423.[CrossRef][Medline]

19. Oren, Z., J. C. Lerman, G. H. Gudmundsson, B. Agerberth, and Y. Shai. 1999. Structure and organization of the human antimicrobial peptide LL-37 in phospholipid membranes: relevance to the molecular basis for its non-cell-selective activity. Biochem. J. 341:501–513.[CrossRef][Medline]

20. Wu, M. H., E. Maier, R. Benz, and R. E. W. Hancock. 1999. Mechanism of interaction of different classes of cationic antimicrobial peptides with planar bilayers and with the cytoplasmic membrane of Escherichia coli. Biochemistry. 38:7235–7242.[CrossRef][Medline]

21. Shai, Y., and Z. Oren. 2001. From "carpet" mechanism to de-novo designed diastereomeric cell-selective antimicrobial peptides. Peptides. 22:1629–1641.[CrossRef][Medline]

22. Shai, Y. 2002. Mode of action of membrane active antimicrobial peptides. Biopolymers. 66:236–248.[CrossRef][Medline]

23. Yang, L., T. A. Harroun, T. M. Weiss, L. Ding, and H. W. Huang. 2001. Barrel-stave model or toroidal model? A case study on melittin pores. Biophys. J. 81:1475–1485.[Abstract/Free Full Text]

24. Chen, F. Y., M. T. Lee, and H. W. Huang. 2003. Evidence for membrane thinning effect as the mechanism for peptide induced pore formation. Biophys. J. 84:3751–3758.[Abstract/Free Full Text]

25. Henzler Wildman, K. A., D. K. Lee, and A. Ramamoorthy. 2003. Mechanism of lipid bilayer disruption by the human antimicrobial peptide, LL-37. Biochemistry. 42:6545–6558.[CrossRef][Medline]

26. Henzler-Wildman, K. A., G. V. Martinez, M. F. Brown, and A. Ramamoorthy. 2004. Perturbation of the hydrophobic core of lipid bilayers by the human antimicrobial peptide LL-37. Biochemistry. 43:8459–8469.[CrossRef][Medline]

27. Nouri-Sorkhabi, M. H., L. C. Wright, D. R. Sullivan, and P. W. Kuchel. 1996. Quantitative 31P nuclear magnetic resonance analysis of the phospholipids of erythrocyte membranes using detergent. Lipids. 31:765–770.[CrossRef][Medline]

28. Keller, S. L., W. H. Pitcher, W. H. Huestis, and H. M. Mcconnell. 1998. Red blood cell lipids form immiscible liquids. Phys. Rev. Lett. 81:5019–5022.[CrossRef]

29. Ratledge, C., and S. G. Wilkinson. 1988. Microbial Lipids. Academic Press, London.

30. Graham, J. M. 1997. Membrane Analysis. Springer, New York.

31. Brock, T. D. 1974. Biology of Microorganisms. Prentice-Hall, Englewood Cliffs, NJ.

32. Goldfine, H. 1972. Comparative aspects of bacterial lipids. Adv. Microb. Physiol. 8:1–58.[Medline]

33. Op Den Kamp, J. A. F. 1979. Lipid asymmetry in membranes. Annu. Rev. Biochem. 79:47–71.[CrossRef]

34. Basanez, G., A. E. Shinnar, and J. Zimmerberg. 2002. Interaction of hagfish cathelicidin antimicrobial peptides with model lipid membranes. FEBS Lett. 532:115–120.[CrossRef][Medline]

35. Gidalevitz, D., Y. J. Ishitsuka, A. S. Muresan, O. Konovalov, A. J. Waring, R. I. Lehrer, and K. Y. C. Lee. 2003. Interaction of antimicrobial peptide protegrin with biomembranes. Proc. Natl. Acad. Sci. USA. 100:6302–6307.[Abstract/Free Full Text]

36. Neville, F., M. Cahuzac, A. Nelson, and D. Gidalevitz. 2004. The interaction of antimicrobial peptide LL-37 with artificial biomembranes: epifluorescence and impedance spectroscopy approach. J. Phys. Condens. Matter. 16:S2413–S2420.[CrossRef]

37. Maget-Dana, R. 1999. The monolayer technique: a potent tool for studying the interfacial properties of antimicrobial and membrane-lytic peptides and their interactions with lipid membranes. Biochim. Biophys. Acta. 1462:109–140.[Medline]

38. Brockman, H. 1999. Lipid monolayers: why use half a membrane to characterize protein-membrane interactions? Curr. Opin. Struct. Biol. 9:438–443.[CrossRef][Medline]

39. Castano, S., B. Desbat, and J. Dufourcq. 2000. Ideally amphipathic beta-sheeted peptides at interfaces: structure, orientation, affinities for lipids and hemolytic activity of (KL)(m)K peptides. Biochim. Biophys. Acta. 1463:65–80.[Medline]

40. Jensen, T. R., K. Balashev, T. Bjornholm, and K. Kjaer. 2001. Novel methods for studying lipids and lipases and their mutual interaction at interfaces. Part II. Surface sensitive synchrotron X-ray scattering. Biochimie. 83:399–408.[Medline]

41. Losche, M. 2002. Surface-sensitive X-ray and neutron scattering characterization of planar lipid model membranes and lipid/peptide interactions. Curr. Top. Membr. 52:117–161.

42. Sun, F. 2002. Constant normal pressure, constant surface tension, and constant temperature molecular dynamics simulation of hydrated 1,2-dilignoceroylphosphatidylcholine monolayer. Biophys. J. 82:2511–2519.[Abstract/Free Full Text]

43. Ambroggio, E. E., F. Separovic, J. Bowie, and G. D. Fidelio. 2004. Surface behaviour and peptide-lipid interactions of the antibiotic peptides, Maculatin and Citropin. Biochim. Biophys. Acta. 1664:31–37.[Medline]

44. Demel, R. A., W. S. M. Geurts Van Kessel, R. F. A. Zwaal, B. Roelofsen, and L. L. M. Van Deenen. 1975. Relation between various phospholipase actions on human red cell membranes and the interfacial phospholipid pressure in monolayers. Biochim. Biophys. Acta. 406:97–107.[Medline]

45. Blume, A. 1979. A comparative study of the phase transitions of phospholipid bilayers and monolayers. Biochim. Biophys. Acta. 557:32–44.[Medline]

46. Gennaro, R., and M. Zanetti. 2000. Structural features and biological activities of the cathelicidin-derived antimicrobial peptides. Biopolymers. 55:31–49.[CrossRef][Medline]

47. Gopal, A., and K. Y. C. Lee. 2001. Morphology and collapse transitions in binary phospholipid monolayers. J. Phys. Chem. B. 105:10348–10354.

48. Maskarinec, S. A., J. Hannig, R. C. Lee, and K. Y. C. Lee. 2002. Direct observation of poloxamer 188 insertion into lipid monolayers. Biophys. J. 82:1453–1459.[Abstract/Free Full Text]

49. Maskarinec, S. A., and K. Y. C. Lee. 2003. Comparative study of polaxmer insertion into lipid monolayers. Langmuir. 19:1809–1815.[CrossRef]

50. Ege, C., and K. Y. C. Lee. 2004. Insertion of Alzheimer's Aß40 Peptide into lipid monolayers. Biophys. J. 87:1732–1740.[Abstract/Free Full Text]

51. Als-Nielsen, J., and D. McMorrow. 2001. Elements of Modern X-Ray Physics. John Wiley & Sons, Chichester, UK.

52. Jensen, T. R., and K. Kjaer. 2001. Structural properties and interactions of thin films at the air-liquid interface explored by synchrotron X-ray scattering. In Novel Methods to Study Interfacial Layers. D. Möbius and R. Miller, editors. Elsevier, Amsterdam. 205–254.

53. Schalke, M., and M. Losche. 2000. Structural models of lipid surface monolayers from X-ray and neutron reflectivity measurements. Adv. Colloid Interface Sci. 88:243–274.[CrossRef][Medline]

54. Helm, C. A., H. Mohwald, K. Kjaer, and J. Als-Nielsen. 1987. Phospholipid monolayer density distribution perpendicular to the water-surface: a synchrotron x-ray reflectivity study. Europhys. Lett. 4:697–703.[CrossRef]

55. Als-Nielsen, J., D. Jacquemain, K. Kjaer, F. Leveiller, M. Lahav, and L. Leiserowitz. 1994. Principles and applications of grazing incidence X-ray and neutron scattering from ordered molecular monolayers at the air-water interface. Phys. Rep. 246:251–313.[CrossRef]

56. Dubreil, L., V. Vie, S. Beaufils, D. Marion, and A. Renault. 2003. Aggregation of puroindoline in phospholipid monolayers spread at the air-liquid interface. Biophys. J. 85:2650–2660.[Abstract/Free Full