INVESTIGATION OF THE INTERACTION OF ANTIMICROBIAL PEPTIDES WITH LIPIDS AND LIPID MEMBRANES YU LANLAN (B. Sc.) A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF CHEMISTRY NATIONAL UNIVERSITY OF SINGAPORE 2006 This work was performed in the biophysical fluorescence laboratory, Department of Chemistry, National University of Singapore under the supervision of Asst Prof. Thorsten Wohland. The results have been partly published in: Yu, L., Ding, J. L, Ho, B. and Wohland, T. Investigation of a novel artificial antimicrobial peptide by fluorescence correlation spectroscopy: An amphipathic cationic pattern is sufficient for selective binding to bacterial type membranes and antimicrobial activity. Biochimica et Biophysica Acta (BBA) – Biomembranes 1716, 2939 (2005). Yu, L., Tan, M., Ho, B., Ding, J. L. and Wohland, T. Determination of critical micelle concentrations and aggregation numbers by fluorescence correlation spectroscopy: Aggregation of a lipopolysaccharide. Analytica Chimica Acta 556, 216-225 (2006). Acknowledgements A thesis like this, involving different and varied fields, cannot be undertaken without the help of many people. I would like to take this opportunity to thank the persons who especially provide help in my study. First, I would like to express my deepest gratitude to my supervisor Asst Prof. Thorsten Wohland for giving me the opportunity to work in such an interesting field of antimicrobial peptides. I would like to thank him for always offering the guidance, support and assistance during the course of this work. I would like to thank Prof. Ding Leak Ling and Ho Bow, for giving me many valuable and scientific suggestions and constructive discussion on the project, which is a great help to me. I would like to thank Assoc Prof. Feng Si Shen for his kind help of providing instrument of Langmuir film balance for my study. I am grateful to all the members of biophysical fluorescence laboratory for their enthusiastic help and support. And last but not least I would like to thank my parents and my husband for their kind understanding, continuous support and unconditional love all these years. i Table of Contents Acknowledgements Table of contents Summary List of Tables List of Figures List of Symbols CHAPTER INTRODUCTION i ii v viii ix xiii CHAPTER OVERVIEW OF ANTIMICROBIAL PEPTIDES 2.1 Biological activities of antimicrobial peptides 2.2 Classification of antimicrobial peptides 2.3 Structural features of antimicrobial peptides 2.4 Biological membranes 2.5 Mechanisms of antimicrobial peptides 2.6 Native and designed antimicrobial peptides 2.7 Therapeutic potential of antimicrobial peptides 7 11 17 22 28 31 36 CHAPTER METHODOLOGY 3.1 Techniques used in the study of antimicrobial peptides 3.2 Fluorescence Correlation Spectroscopy (FCS) 3.2.1 Principle of Fluorescence Correlation Spectroscopy (FCS) 3.2.2 Setup of FCS 3.2.3 System calibration 3.3 Langmuir film balance 3.3.1 Comparison of monolayer and bilayer 3.3.2 Principle of surface chemistry 3.3.3 Realization of the surface pressure measurement 3.3.4 Instrumentation 39 39 39 39 46 47 50 50 51 52 53 CHAPTER DETERMINATION OF CRITICAL MICELLE CONCENTRATIONS AND AGGREGATION NUMBERS OF DETERGENT AND LPS 4.1 Introduction 4.2 Materials and methods 4.3 Results and Discussion 4.3.1 Theory development 4.3.1.1 Titration of micelles with fluorophore 4.3.1.2 Disaggregation of FITC-LPS with a detergent 4.3.2 Calculation of titrating a solution of aggregates with a fluorescent probe 4.3.3 Determination of aggregation number of C12E9 using R18 4.3.4 Determination of aggregation number of LPS using R18 56 56 58 61 61 61 65 66 69 74 ii 4.3.5 Determination of aggregation number of LPS using FITC-LPS and Triton X-100 CHAPTER INVESTIGATION OF THE BINDING OF A NOVEL ANTIMICROBIAL PEPTIDE V4 TO MEMBRANE MIMICS 5.1 Introduction 5.2 Materials and methods 5.3 Results 5.3.1 Calibration of the FCS setup 5.3.2 Solubility of V4-TMR 5.3.3 Binding of V4-TMR to LPS 5.3.4 Binding of unlabeled V4 to FITC-LPS 5.3.5 Attachment on the coverslip surface 5.3.6 Comparison of V4-TMR binding to LPS, lipid A and PC 5.3.7 Binding to V4-TMR to SUVs of pure lipids 5.3.8 Interaction of V4-TMR with mixed lipid SUVs 5.4 Discussion 5.4.1 V4-TMR aggregates and is strongly quenched in PBS 5.4.2 LPS and lipids can partly dissolve V4 5.4.3 V4-TMR functions via hydrophobic and electrostatic forces 5.4.4 Saturation of lipids affects the binding with V4-TMR CHAPTER INVESTIGATION OF THE MECHANISMS OF ANTIMICROBIAL PEPTIDES INTERACTING WITH MEMBRANES 6.1 Introduction 6.2 Materials and methods 6.3 Results 6.3.1 Antimicrobial peptides induce leakage of rhodamine 6G entrapped POPG LUVs 6.3.2 Antimicrobial peptides interact with Rho-PE labeled POPG LUVs 6.3.3 Antimicrobial peptides induce leakage of rhodamine 6G entrapped DPPG LUVs 6.3.4 Antimicrobial peptides interact with Rho-PE labeled DPPG LUVs 6.3.5 Visualization of antimicrobial peptides interacting with RhoPE labeled LUVs 6.3.6 Effect of incubation time on the activity of antimicrobial peptides 6.3.7 Real time interaction of V4 with rhodamine 6G entrapped POPG LUVs 6.4 Discussion 6.4.1 Mechanisms of antimicrobial peptides 6.4.2 Comparison of different antimicrobial peptides 75 78 78 79 84 84 84 87 89 90 90 92 94 95 95 97 99 100 101 101 102 105 105 113 119 123 128 131 131 132 132 136 iii CHAPTER INTERACTION BETWEEN ANTIMICROBIAL PEPTIDES AND LIPID MEMBRANE LAYERS 7.1 Introduction 7.2 Materials and methods 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 7.3.1.2 Isotherms of mixed lipid/V4 monolayers 7.3.1.3 Miscibility analysis of monolayers 7.3.1.4 Stability analysis of monolayers 7.3.1.5 Compressibility analysis of monolayers 7.3.2 Penetration studies of antimicrobial peptides interacting with lipid monolayers 7.3.2.1 Penetration of antimicrobial peptides into POPG monolayers 7.3.2.2 Penetration of antimicrobial peptides into POPC monolayers 7.3.2.3 Penetration of V4 into different lipid monolayers 7.3.2.4 Effect of lipid packing on the penetration of V4 7.3.3 AFM studies of antimicrobial peptides interacting with lipid monolayers 7.3.3.1 AFM images of pure lipid monolayers 7.3.3.2 AFM images of POPG monolayers penetrated by antimicrobial peptides 7.3.3.3 AFM images of V4 penetrating into POPC and DPPG monolayers 7.3.4 Insertion of V4 into solid supported bilayer 141 CHAPTER CONCLUSION AND OUTLOOK 8.1 Conclusion 8.2 Outlook 184 184 189 Bibliography 192 141 141 145 145 145 146 148 151 153 155 155 158 163 168 170 170 171 176 182 iv Summary Antimicrobial peptides serve as a significant weapon of organisms to defend against microbial infections. Different from conventional antibiotics, antimicrobial peptides mainly target bacterial membranes, induce permeation and eventually lead to bacterial lysis. Because it is difficult for bacteria to develop resistance, antimicrobial peptides have been considered a promising drug candidate as a substitute or addition for conventional antibiotics. Many native antimicrobial peptides have a broad-spectrum activity against different microorganisms. However, due to the lack of selectivity, their application as therapeutics is limited. Therefore de novo design of antimicrobial peptides has become an interesting method to provide new and efficient drug candidates. V4 which is designed based on some sequences of endotoxin-binding host defense proteins has an amphipathic pattern of HBHPHBH (H: hydrophobic; B: basic; P: polar residue) and displays a good combination of high antimicrobial activity, low cytotoxic activity and low hemolytic activity. Its greatly enhanced antimicrobial potency is worthy of further investigation. This study investigated the interaction between V4 and different membrane components to unravel the mechanism of V4 targeting membranes. Peptide binding, insertion and membrane permeation were examined. The physical property of V4 and the binding affinity for different membrane components, including LPS, lipid A and distinct phospholipids, were studied by fluorescence correlation spectroscopy (FCS). The results show that V4 is a highly hydrophobic peptide, leading to a strong aggregation in solution. Only a small percentage of V4 is active. At low peptide/lipid ratio, V4 shows a higher v binding affinity for anionic membrane components compared to the zwitterionic lipids. This strong interaction of V4 with negatively charged lipids indicates that electrostatic force is a prerequisite for its selective action on bacterial membranes in contrast to mammalian membranes. The insertion of V4 into membranes was examined by Langmuir film balance and atomic force microscopy. V4 shows higher ability to penetrate into negatively charged membranes than neutral membranes, which confirms the significance of electrostatic interaction. Except insertion, the formation of filaments was also observed, indicating a peptide aggregation. At high peptide/lipid ratio, V4 is shown to induce membrane permeation by causing membrane aggregation and disruption. The mechanism of V4 interacting with membrane is also compared to other antimicrobial peptides magainin 2, melittin and polymyxin B. Necessary for the understanding of V4 binding to bacterial membranes is an understanding of the aggregation behavior of the main target, LPS. Therefore, the aggregation property of LPS was examined using a new method based on FCS, which allows the determination of aggregation numbers of complexes from the amplitudes of the correlation function. This method was evaluated by a well-known detergent C12E9 and showed a good agreement with literature values, which confirmed the feasibility of the method. Later the method was applied on LPS to determine the critical micelle concentration and aggregation number. This study should enhance the understanding of the predominant specificity of V4 and the mode of action on membranes. The results could contribute to the rational design of vi novel antimicrobial peptides and provide useful information to develop new antimicrobial drugs. vii List of Tables Table 2.1 Representative antimicrobial peptides with different secondary structure 12,13 Table 4.1 Number N of particles per observation volume of FITC-LPS solutions of concentrations between and 50 µM before and after treatment with Triton X-100 76 Table 5.1 FITC-LPS interacting with unlabeled V4 90 Table 5.2 Comparison of TMR, V4-TMR, V4-TMR: LPS on coverslip 90 Table 5.3 Comparison of interaction of V4-TMR peptide with LPS, lipid A and PC 91 Table 6.1 Comparison of mechanisms of antimicrobial peptides inducing membrane permeation 133 Table 7.1 Comparison of height differences of POPG monolayers penetrated by different antimicrobial peptides 176 viii P(0, m ) = e − m (4.5) and the overall number of micelles with at least one label is m k e−m k! k =1 L N = N mic (1 − P(0, m )) ≈ N mic ∑ (4.6) We have here approximated the Poisson distribution by a finite sum since in FCS each factor in the sum will have to be evaluated individually due to differences in the photon count rates per particle as will be explained below. The cutoff L in this sum was set arbitrarily at the number of labels k for which the probability P(k,m) ≤ 0.01. For instance, for cpr = cmic, i.e. the concentration of labels is times higher than the micelle concentration, L = since the probability to find a micelle with more than labels is smaller than %. The number N is the observable parameter in FCS. In the following we distinguish two different cases depending on the ratio of cpr/cmic. If measuring at probe concentrations much below the micelle concentration cpr < 0.1 cmic then the sum over all P(k,m) with k>1 is smaller than 0.5% and particles with more than one label can be neglected. Then it follows from Eqs. 4.3, 4.5 and 4.6 that − c pr ⎞ ⎛ N = N mic ⋅ ⎜⎜1 − e cmic ⎟⎟ ⎠ ⎝ (4.7) Using Eqs. 4.1 and 4.2 this can be written as a function depending on the probe label, the detergent concentration, the CMC, and Nagg: − c pr N agg ⎛ cdet − ccmc cdet −ccmc ⎞ ⎜ ⎟ N= Vobs N A ⎜1 − e ⎟ N agg ⎝ ⎠ (4.8) 62 In the case for cpr ≥ 0.1 cmic the different P(k,m) cannot be neglected anymore and have to be included. Eqs. 4.1-3 and 4.5-6 yield then the following equation for the number of labeled particles: k N agg c pr ⎛ N agg c pr ⎞ − cdet −ccmc ⎜ ⎟ e ⎟ L ⎜c c det − ccmc det − c cmc ⎠ ⎝ Vobs N A ∑ N= N agg k! k =0 (4.9) Eqs. 4.8 and 4.9 determine the average number of labeled particles in any observation volume. This number can be measured with FCS. According to Eq. 3.8 in Chapter 3, the number of particles N in the FCS observation volume thus is n α k2 Fk ∑ k =1 N= G (0) − G∞ ⎡ n ⎤ ⎢∑ α k Fk ⎥ ⎣ k =1 ⎦ (4.10) For later discussions we introduce here the average photon count rates per particle (cps) given by the overall count rates divided by the particle number N. If only one particle species is present in solution the cps is a direct measure of the fluorescence yield Qk since the amplitude G(0)-G∞ is proportional to 1/N. If several particle species are in solution the cps is an average value for the different species. Changes in this value can indicate whether fluorescence yields change due to incorporation of several labels into one particle or if aggregation happens. Eq. 4.10 represents the general solution for the measured number of fluorescent particles under different conditions and can be used to determine the number of labeled micelles. 63 In the case that only one label is present on each micelle or the αk are all equal, independent of the number of labels, Eq. 4.10 reduces to N= G (0) − G∞ (4.11) In the case for the micelles the different fractions Fk in Eq. 4.10 are given by Fk = P(k , m ) . (4.12) The different αk depend on the particular situation. In general they have to be measured and cannot be easily recovered from fits to the ACF. We will discuss here two exemplary cases applicable to our situation. In the simplest situation all αk are equal, which is e.g. the case if cpr cmic) converges to the number of micelles present. 66 Fig. 4.2 Calculations of the titration curves. On the upper left is a sketch of the micells labeling when going from low to high label concentrations. The curves represent the particle number as measured vs. the concentration of probe labels and are given for different values of the parameter cmic=(cdet-ccmc)/Nagg which describes the concentration of micelles. Secondly, we use a model in which the fluorescence yields per particle are proportional to the number of labels Qk = k Q1 (i.e. αk = k). For this second model we have depicted the cps in Fig. 4.3A. In this case N (Fig. 4.2B) also rises linearly with the label concentration (cpr > cmic) N decreases again. This is due to the fact that the FCS amplitude is proportional to the square of the fluorescence yield of the particles measured. Thus particles with a higher fluorescence yield contribute disproportionally more to the ACF than would be expected from their mole fraction (Eq. 4.10). This leads to an 67 increase in the ACF and thus a decrease in N. The cps in this case rises continuously since the average number of labels per particle increases continuously during the titration (Eqs. 4.3 and 4.4) and the real number of particles is underestimated by the amplitude of the correlation function. Thirdly, we use a model in which we assume that the fluorescence gets progressively quenched for every additional fluorophore incorporated into a micelle. The fluorescence yields per particle are then proportional to the quenching constant β to the power of the number of labels minus one, since quenching happens only for k>1: Qk =β k-1 Q1 (i.e. αk =β k-1 ). In this case N (Fig. 4.2C) is dominated by singly labeled, unquenched particles and the normalized cps (Fig. 4.3B) decreases due to quenching. Again the real number of particles is underestimated by the amplitude of the correlation function. Fig. 4.3 Calculation of the titration curves. Depictes are the cps vs. the concentration of probe labels. In general, it can be seen that for concentrations of micelles larger than the concentration of labels, the different models have only a small influence on the number of particles N, 68 since this parameter is dominated by the almost exclusively singly labeled micelles. In this situation differences between the three models can only be made by the normalized cps which will stay constant, increase or decrease, respectively. If the concentration of micelles is smaller than the concentration of labels, then N will be dominated by the multiple labeled micelles in the case of increasing fluorescence yield, or will be dominated by the singly labeled micelles in the case of quenching. 4.3.3 Determination of aggregation number of C12E9 using R18 R18 is a fluorophore which is extensively used for membrane staining. Due to the long hydrophobic alkyl chain (Fig. 4.1A) R18 can be incorporated into the highly hydrophobic core of micelles. R18 in aqueous solution forms strongly quenched multimers209 and exhibits ACFs with low N, low cps and at least two different diffusion times, consistent with aggregation and quenching. It exhibits high fluorescence only upon incorporation into hydrophobic structures and thus is a good probe for the detection of these structures, including micelles. To evaluate the method, we have studied the detergent C12E9 with a known CMC of 0.08 mM and Nagg of about 120208. First, we tested the transition around the CMC by titrating fixed concentrations of 20 and 40 nM of R18 with C12E9 concentrations between 40 to 200 µM and taking FCS measurements. In Fig. 4.4 the parameters N, cps and τD of the FCS fits are shown. All three parameters show a transition in the solutions between 70 and 120 µM. The number of particles N (Fig. 4.4A) rises because R18, which is usually found in strongly quenched aggregates in solution, is dissolved in the micelles that are 69 formed at concentrations of detergent higher than the CMC. With a known aggregation number of 120, 2.4 µM of extra C12E9 above the CMC will form 20 nM of micelles, enough to dissolve all R18 molecules. Thus the broad transition over more than 40 µM is not an artifact of the R18 label. This is confirmed by the fact that the transition is independent of the concentration of R18, and 20 and 40 nM R18 show the same transition. Thus the general assumption that the transition at the CMC from monomers to micelles is abrupt is an idealization. The other two parameters, cps and τD (Figs. 4.4B and C) show a similar transition. Below a concentration of 90-100 µM these parameters fluctuate and show very large errors. Above this concentration range the parameters stabilize and show much smaller standard deviations, consistent with the idea that in this concentration range micelles dominate. Sample ACFs covering this transition are shown in Fig. 4.5A. 70 Fig. 4.4 Titration of 20 and 40 nM of R18 with increasing concentrations of C12E9 (40200 µM) Fig. 4.5 ACFs taken in the transition region around the CMC (dashed lines). Fits assuming a single particle in solution are shown as solid lines. (A) 20 nM R18 with varying concentrations of C12E9. (B) 20 nM R18 with varying concentrations of LPS. 71 150 µM C12E9, a concentration above the CMC and the transition seen in the experiments is chosen to test the aggregation number by titrating this solution with R18 concentrations from - 450 nM. FCS curves are taken and the number of particles N and the cps are shown in Fig. 4.6A and B. In these measurements the cps increases slowly but steadily with R18 concentration indicating that more than one R18 label per micelle is incorporated on average. The models assuming either a constant fluorescence yield (Eq. 4.8) or a fluorescence yield proportional to the number of labels (Eq. 4.14 with αk~k) are used to fit the data as limiting cases. The third model assuming quenching and a decreasing cps is not applicable. Assuming a constant fluorescence yield the fit yields a CMC of 105 µM and an Nagg of 132. Assuming a fluorescence yield proportional to the number of labels the fit yields a CMC of 93 µM and an Nagg of 112. This is in good agreement with the theoretical values of a CMC of 80 µM and an Nagg of 120 indicating that this titration method can be used to determine not only Nagg but as well the CMC. 72 Fig. 4.6 Titration of C12E9 and LPS solutions with R18. At least 10 FCS measurements were taken and the average fitted parameters are shown. A) The number of particles in the focal volume N for solutions of 0.15 mM C12E9 are shown for concentrations of R18 (10 – 450 nM). The dashed line is a global fit assuming that the fluorescence yield of a particle is directly proportional to the number of attached labels (CMC = 93µM, Nagg = 112). The solid line represents the result of a global fit to the model assuming that all particles have the same fluorescent yield independent of the number of labels attached (CMC = 105 µM, Nagg = 132). B) The cps are shown for the same experiment. C) The number of particles in the focal volume N for solutions of µM LPS are shown for concentrations of R18 (1 – 60 nM). The dashed line is a global fit assuming that the fluorescence yield of a particle is directly proportional to the number of attached labels (CMC = 1.6 µM, Nagg = 43). The solid line represents the result of a global fit to the model assuming that all particles have the same fluorescent yield independent of the number of labels attached (CMC = 1.3 µM, Nagg = 49). D) The cps are shown for the same experiment. 73 4.3.4 Determination of aggregation number of LPS using R18 LPS concentrations from 0.25 - µM were titrated with fixed concentrations of 10 and 20 nM R18. In Fig. 4.7 the parameters N, cps and τD of the FCS fits are shown. A rise in N can be seen, stemming from the progressive integration of R18 into micelles. However, the transition happens over a much smaller concentration range of about µM. Experimental ACFs for this transition region are shown in Fig. 4.5B. The LPS solutions show especially at lower concentrations aggregates and in general the experimental values show higher errors than for the case of C12E9 (see Fig. 4.6). This indicates that the LPS solutions are less homogeneous and the aggregates have a wider distribution of sizes. Fig. 4.7 Titration of 10 and 20 nM of R18 with increasing concentrations of LPS (0.25-5 µM). At least 10 FCS measurements were taken and the average fitted parameters are shown.(Blue: 20 nM R18; Red: 10 nM R18) 74 A LPS concentration of µM which was above the transition was chosen for the CMC investigation. The results of the FCS measurements of µM LPS with increasing concentrations of R18 (0-60 nM) are shown in Fig. 4.6. In Fig. 4.6C the rise in the number of particles N with concentration of R18 can be seen as well as the saturation of this value for large R18 concentrations when all LPS micelles are labeled. Concomitantly, the cps (Fig. 4.6D) rises due to the increase in the average number of labels per LPS micelles. The data was fitted with the models of constant fluorescence yield (Eq. 4.8) and the model of fluorescence yield proportional to the number of labels (Eq. 4.14 with αk ~k). Assuming a constant fluorescence yield the fit yields a CMC of 1.3 µM and an Nagg of 49. Assuming a fluorescence yield proportional to the number of labels the fit yields a CMC of 1.6 µM and an Nagg of 43. 4.3.5 Determination of aggregation number of LPS using FITC-LPS and Triton-X100 In this experiment, varying concentrations of FITC-LPS were dissolved by Triton X-100 (Table 4.1, Fig. 4.8). From to 50 µM FITC-LPS, the number of fluorescent particles N detected in the confocal volume increased linearly from 4.19 to 203.50, respectively. Upon addition of Triton X-100, N increased by a factor between 1.67 and 2.26. Variations in this factor were not systematic and on average N increase by 1.94 ± 0.21. In Fig. 4.8 these values are shown including linear fits to the data (FITC-LPS: y = -1.35 + 3.84 x and FITC-LPS in presence of Triton X-100: y = -0.36 + 7.03 x), since N should rise linearly with concentration. The slopes of the two fitted lines have a ratio of 1.83, consistent with the values 1.94 ± 0.21 determined from the measurements. Therefore, we conclude that on average there were 1.94 ± 0.21 particles per LPS micelle. 75 Table 4.1 Number N of particles per observation volume of FITC-LPS solutions of concentrations between and 50 µM before and after treatment with Triton-X100 FITC-LPS[µM] N N’ N’/N 4.2 ± 0.35 7.2 ± 0.41 1.72 ± 0.239 15.2 ± 1.94 34.5 ± 5.86 2.26 ± 0.673 10 40.2 ± 3.36 84.1 ± 7.58 2.09 ± 0.364 15 59.8 ± 1.82 123.4 ± 23.36 2.06 ± 0.454 20 71.1 ± 3.31 134.1 ± 16.72 1.89 ± 0.323 30 92.6 ± 8.19 154.6 ± 11.69 1.67 ± 0.274 50 203.5 ± 18.05 379.1 ± 31.01 1.86 ± 0.318 Values are given as means ± S. D. of 10 meaaurements. The last column shows the factor by which N increased upon addition of Triton-X100. Fig. 4.8 Dissolution of FITC-LPS micelles by Triton X-100. The number of particles in the focal volume N, as measured by FCS, is shown versus the concentration of FITCLPS for solutions prepared in PBS buffer or in the presence of Triton X-100. The data are fitted with a straight line. According to Sigma, the degree of substitution is µg FITC per mg LPS which corresponds to 6% labeled LPS. Based on the LPS molecular weight of 10 kD, the degree of substitution showed that at 6% substitution, out of every 50 LPS molecules was labeled. Combined with the estimation that one LPS micelle was labeled with 1.94 FITC molecules on average, therefore, there could be 32 LPS molecules per micelle. This value is a lower limit for the LPS aggregation number since the estimation relies on the 76 assumption that LPS is completely dissolved and each Triton X-100 micelle contains at most FITC-LPS molecule. Hence from this experiment, the number of LPS molecules per micelle is at least 32. This result is consistent with the earlier measurements by FCS titrations showing that the LPS micelles contain 43-49 LPS monomers. In general, the LPS solutions are less homogeneous than the C12E9 solutions and from time to time larger aggregates can be seen (Fig. 4.5). Therefore, the aggregation number is an average value around which size of micelles will vary more or less strongly. The transition of detergent molecules from the monomer solution to micelle solution is a successive process covering a certain concentration range in reality as confirmed by the transition of the known detergent C12E9. Therefore at the detergent concentration below the CMC some micelles exist in the solution. When the detergent concentration increases to above the CMC, the micelles dominate, leading to the usual micellar solution. LPS is a weak detergent compared with C12E9. It also displayed a transition in forming micelles. Therefore below the value of CMC, there are also some LPS micelles existing in the solution. This is consistent with the work of Santos et al. who showed that below CMC of LPS (E. coli 026: B6), there are also some aggregates of LPS which they called premicelles113. Santo and colleagues showed that the sizes of micelles and pre-micelles were different, however, the present study did not observe any obvious change in size (Fig. 4.7C). This is probably due to the different bacterial strain used in both studies, which significantly influences the property of forming micelles. 77 [...]... entrapped POPG LUVs 11 1 Figure 6.4 Interaction of antimicrobial peptides with Rho-PE labeled POPG LUVs 11 6 Figure 6.5 Aggregation caused by melittin at peptide /lipid ratio of 1: 2.67 11 7 Figure 6.6 Comparison of Napp and diffusion time of antimicrobial peptides interacting with Rho-PE labeled POPG LUVs 11 8 Figure 6.7 Interaction of antimicrobial peptides with rhodamine 6G entrapped DPPG LUVs 12 0 x Figure 6.8... Comparison of Nrho, Nvesicle, F2 and photon count rates of antimicrobial peptides interacting with rhodamine 6G entrapped DPPG LUVs 12 3 Figure 6.9 Interaction of antimicrobial peptides with Rho-PE labeled DPPG LUVs 12 5 Figure 6 .10 Comparison of Napp and diffusion time of antimicrobial peptides interacting with Rho-PE labeled DPPG LUVs 12 7 Figure 6 .11 Confocal images of Rho-PE labeled LUVs in the absence and. .. Figure 3 .10 Langmuir-Blodgett trough 54 Figure 3 .11 Schematic drawing of antimicrobial peptides penetrating into lipid monolayers 55 Figure 4 .1 Structure of R18 and formula of C12E9 57 Figure 4.2 Calculations of the titration curves 67 Figure 4.3 Calculation of the titration curves (cps vs concentration of probe labels) 68 Figure 4.4 Titration of 20 and 40 nM of R18 with increasing concentrations of C12E9... V4-TMR to SUVs of mixed lipid composition 94 Figure 6 .1 Schematic drawing of the investigation of the mechanisms of antimicrobial peptides by rhodamine 6G entrapped LUVs and Rho-PE labeled LUVs 10 2 Figure 6.2 Interaction of antimicrobial peptides with rhodamine 6G entrapped POPG LUVs 10 8 Figure 6.3 Comparison of Nrho, Nvesicle, F2 and photon count rates of antimicrobial peptides interacting with rhodamine... increase the interaction between antimicrobial peptides and cancer cell membranes and contribute to the selectivity of antimicrobial peptides The large number of microvilli on tumorigenic cells is another possible reason for the different susceptibilities of cancer cells and normal 8 cells to antimicrobial peptides2 3 The microvilli increase the surface area of the tumorigenic cell membranes and allow... examine the ability of V4 to induce membrane permeation and compare with other antimicrobial peptides magainin 2, melittin and polymyxin B 5) To investigate the interaction of V4 with different lipid monolayers including insertion ability and membrane morphological change, and compare with other antimicrobial peptides 6) To elucidate the possible mechanisms of studied antimicrobial peptides interacting with. .. Troton X -10 0 and ACFs of 1 nM rhodamine 6G and 10 0 nM V4-TMR in DMSO 86 Figure 5.3 ACF of 10 0 nM V4-TMR with different concentrations of LPS in PBS 88 Figure 5.4 Titration of 10 0 nM V4-TMR with increasing concentrations of LPS 89 Figure 5.5 Comparison of ACFs of V4-TMR and complexes of V4-TMR with LPS, lipid A and PC 91 Figure 5.6 Comparison of V4-TMR binding to different SUVs 93 Figure 5.7 Binding of V4-TMR... 13 2 Figure 7 .1 Isotherms of POPG and POPC monolayers 14 6 Figure 7.2 Isotherms of mixed POPG/V4 and POPC/V4 monolayers 14 8 Figure 7.3 Surface pressure of lipid monolayers incorporated with different percentage of V4 15 0 Figure 7.4 Excess surface pressure of lipid monolayers incorporated with different percentage of V4 15 1 Figure 7.5 Excess Helmholtz energy of lipid monolayers incorporated with different... Sushi 3 domain of Factor C allows the formation of S3 dimers The S3 dimer is capable of disrupting the LPS micelles whereas the reduction of the dimer into monomer makes S3 lose this ability1 01 L- and D- amino acid composition: The biological activity of antimicrobial peptides is also related to the L- or D- enantiomer However the effect of enantiomer of the amino acid is dependent on whether the replacement... selectivity of antimicrobial peptides killing bacteria, not harming mammalian cells, and to unravel the mechanism of the action of antimicrobial peptides, which are the keys for the rational 2 design of novel antimicrobial drugs A more detailed overview of antimicrobial peptides can be found in Chapter 2 Because many natural antimicrobial peptides kill bacteria and are toxic to mammalian cells, few of them . Introduction 14 1 7.2 Materials and methods 14 1 7.3 Results and Discussion 14 5 7.3 .1 Isotherm studies of V4 interacting with POPG and POPC 14 5 7.3 .1. 1 Isotherms of lipid monolayers 14 5 7.3 .1. 2 Isotherms of. affects the binding with V4-TMR 10 0 CHAPTER 6 INVESTIGATION OF THE MECHANISMS OF ANTIMICROBIAL PEPTIDES INTERACTING WITH MEMBRANES 10 1 6 .1 Introduction 10 1 6.2 Materials and methods 10 2 6.3. INVESTIGATION OF THE INTERACTION OF ANTIMICROBIAL PEPTIDES WITH LIPIDS AND LIPID MEMBRANES YU LANLAN (B. Sc.) A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY