Cover Page. The handle holds various files of this Leiden University dissertation

Transcription

1 Cover Page The handle holds various files of this Leiden University dissertation Author: Mojumdar, Enamul Haque Title: A model membrane approach to elucidate the molecular organization in the skin barrier Issue Date:

2 CHAPTER Unraveling the Characteristic Long Periodicity Phase in the Skin Barrier: Localization of Acyl Ceramide and Cholesterol in the Unit Cell of the Long Periodicity Phase E.H. Mojumdar 1 G.S. Gooris 1 D.J. Barlow 2 M.J. Lawrence 2 B. Deme 3 J.A. Bouwstra 1* 1 Leiden /Academic Center for Drug Research, Department of Drug Delivery Technology, Gorlaeus Laboratories, University of Leiden, Leiden, the Netherlands 2 Pharmaceutical Science Division, King s College London, London, United Kingdom 3 Institute Laue-Langevin, Grenoble, France To be submitted ABSTRACT The extracellular lipid matrix in the skin s outermost layer, the stratum corneum (SC), is crucial for the skin barrier function. The matrix is composed of ceramides (CERs), cholesterol (CHL) and free fatty acids (FFAs) and form two lamellar phases: the short periodicity phase (SPP) and the long periodicity phase (LPP). To understand the skin barrier thoroughly, information about the molecular arrangement in the unit cell of these lamellar phases is paramount. Previously we examined the molecular arrangement in the unit cell of the SPP. In the present study, we used neutron diffraction to obtain more details on the molecular arrangement in the unit cell of the LPP. The diffraction pattern revealed at least 8 diffraction orders of the LPP with a repeating unit of ± 0.5 Å. To obtain information about the location of lipid subclasses in the unit cell, protiated samples and partly deuterated samples were examined. The diffraction data obtained by means of D 2 contrast variation together with a gradual replacement of one particular CER, the acyl CER, by its partly deuterated counterpart, were used to construct the scattering length density profiles. This resulted in the position of the acyl chain of this CER subclass being in the inner head-group regions of the LPP trilayer at a position of ~ 21.4 ± 0.2 Å from the unit cell center. The position of CHL was determined using either a deuterated tail or a partially deuterated head group. CHL is located in the two outer layers of the trilayer arrangement in the unit cell of the LPP with its head-group being at a position of ~ 26 ± 0.2 Å from the center of the unit cell. This offers the possibility of a hydrogen bond with the ester group of the CER located in close proximity. A molecular model is proposed based on these lipid arrangements in the unit cell of the LPP. KEYWRDS Stratum corneum Lipid mixtures, Ceramides Long periodicity phase, Neutron diffraction

3 PART II Molecular organization in the LPP unit cell INTRDUCTIN The skin protects us from the hazardous external environment by providing an essential barrier. However, in several inflammatory skin diseases, this barrier is impaired 1-6. The physical barrier of the skin is located in its outermost layer referred to as stratum corneum (SC). The SC consists of corneocytes (dead flattened cells filled with keratin and water) embedded in the lipid matrix. The corneocytes are surrounded by a cornified envelope which hampers the partitioning of molecules into the corneocytes. As a consequence, the intercellular lipid matrix is an important route for penetration of substances across the skin 8, 9. As the intercellular lipid domains in the SC serve as an important route for penetration of compounds, the lipid matrix plays an important role in the skin barrier function 10. For this reason it is essential to obtain information on its organization. The lipids assemble in two coexisting crystalline lamellar phases. The aim of this study was to provide details on the lipid arrangement on a molecular level in the unit cell in one of these lamellar phases. The main lipid classes in the SC are ceramides (CERs), cholesterol (CHL) and free fatty acids (FFAs) in an approximately equimolar ratio The CERs consists of long acyl chains linked to a sphingoid base through an amide linkage. Currently 14 different CER subclasses are identified in human SC 14, These subclasses vary in molecular architecture of both, the sphingoid base and acyl chain. Four of these CER subclasses have an exceptional molecular structure with a very long ɷ-hydroxy fatty acid chain ester linked to a linoleic acid, referred to as CER E (acyl CERs). In human and pig SC, the CERs together with CHL and FFAs form two coexisting crystalline lamellar phases. These two phases are referred to as the long periodicity phase (LPP) and the short periodicity phase (SPP) with a repeat distance of approximately 130 and 60 Å, respectively Furthermore, it has been reported that the presence of CER E is required for the formation of the LPP 23. As the LPP is considered to be important for the skin barrier, the presence of CER E in the lipid composition is an important factor in maintaining the skin barrier 24. The importance of CER E and the LPP for the skin lipid barrier has been demonstrated by studies performed with model membranes: the absence of CER E hampered the formation of the LPP and reduced the lipid barrier of these lipid membranes 25. As this indicates that the LPP plays a crucial role in the skin barrier, it is thus important to unravel the molecular structure of the LPP. Studies have been reported focusing on the lipid arrangement and molecular interaction of lipids in mixtures consisting of only 3-5 lipid subclasses using neutron diffraction Although these studies are of interest and provide important information on the localization of lipid subclasses in the unit cell, often the lipid phase behavior does not mimic that in SC. In our studies we focus on more complex mixtures that form a single lamellar phase that is also present in SC. In previous studies we examined the lipid arrangement in the unit cell of the SPP. The lipids exhibit a typical bilayer formation very similar as observed in the phospholipid mixtures Recently the arrangement of the most abundant CER subclass (CER NS C24), see figure S1, FFA with a chain length of C24 (lignoceric 155

4 acid) and CHL in the unit cell of the SPP has been determined using neutron diffraction. Based on these results a molecular model for the SPP unit cell was proposed 25, 34. As far as the LPP is concerned, a few studies are also documented in the literature regarding the arrangement of lipids in the LPP 35, 36. In one of these studies an asymmetric arrangement of the lipids in the unit cell is proposed. These studies were performed by cryo-electron microscopy combined with simulations. In the other study, using X-ray diffraction, a two-layer lipid arrangement is proposed in which CHL is asymmetrically located near the outer border of each lipid bilayers in a 130 Å unit cell. In a recent study using X-ray diffraction, a more detailed electron density profile could be constructed showing a symmetric trilayer arrangement of lipids in the unit cell of the LPP 3. In order to obtain more details of this trilayer lipid arrangement in the unit cell of the LPP, recently neutron diffraction experiments were performed as well. In these studies, the water distribution in the LPP unit cell was determined demonstrating a trilayer arrangement of the lipids in the LPP unit cell (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). In that study we determined the localization of two highly abundant lipids in the lipid mixtures forming only the LPP: CER NS C24 and FFA C24 (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). As the CER ES (one subclass of CER E family, see supplemental figure S1 for structure) is crucial for the formation of the LPP, it is important to locate CER ES in the unit cell of the LPP. Therefore, in this study we focused on the localization of CER ES in the unit cell using neutron diffraction in combination with contrast variation. Furthermore, the molecular location of the 3 rd lipid class present in SC, CHL has also been examined. The linoleate moiety (18 carbon atoms (C)) attached to the 30 C acyl chain of the CER ES was perdeuterated. The CHL molecules were deuterated either in the head-group region or in the tail of the molecule. First, the neutron scattering length density profiles of the LPP were constructed using D 2 and CER ES contrast variation. Second, the localization of the CER ES linoleate moiety and the CHL head and tail region were determined from its partially deuterated counterparts. Finally, based on the position of these molecules together with previously determined CER NS C24 and FFA C24 position, a molecular model for the LPP was proposed. MATERIALS AND METHD Materials In this study the following synthetic CERs were used; the ester linked omega-hydroxy acyl chain (abbreviation E, 30 carbons in the acyl chain (C30)) with a sphingosine chain (abbreviation S, C18) referred to as CER ES (C30), a non-hydroxy acyl chain (abbreviation N, C24) linked to a sphingosine base (C18) referred to as CER NS (C24), a non-hydroxy acyl chain (C24 or C16) linked to a phytosphingosine base (abbreviation P) referred to as CER NP (C24) and CER NP (C16) respectively, an alpha-hydroxy chain (abbreviation A) linked to a sphingosine base referred to as CER AS (C24) and an alpha-hydroxy acyl 156

5 PART II Molecular organization in the LPP unit cell chain (C24) linked to a phytosphingosine base referred to as CER AP (C24). The number between parentheses indicates the number of carbon atoms present in the acyl chain of the CERs. All the CERs were generously provided by Evonik (Essen, Germany). Palmitic acid (C16:0), stearic acid (C18:0), arachidic acid (C20:0), behenic acid (C22:0), tricosanoic acid (C23:0), lignoceric acid (C24:0), cerotic acid (C26:0), CHL and deuterated water were obtained from Sigma-Aldrich Chemie GmbH (Schnelldorf, Germany). The CHL with head deuterated in the positions 2, 3, 4 and 6 (total 6 deuterium) and tail deuterated in the positions 25, 26 and 2 (total deuterium) were obtained from Larodan (Malmӧ, Sweden). The CER ES with its deuterated linoleate was costume synthesized by Evonik (Essen, Germany) and kindly provided for this study. The molecular structure of the deuterated CER ES and CHL are presented in figure 1. Silicon substrates (cut from a wafer) were obtained from kmetic (Vantaa, Finland). All solvents used were of analytical grade and supplied by Labscan (Dublin, Ireland). The water was of Millipore quality produced by Milli-Q water filtration system with a resistivity of 18 MΩ.cm at 25 C. Composition of the model lipid mixtures The model lipid mixtures were prepared from synthetic CERs, CHL and FFAs in an equimolar ratio. The CER subclasses consisted of the following CER composition; CER ES C30, CER NS C24, CER NP C24, CER AS C24, CER NP C16 and CER AP C24 in a 13.3:12.0:3.:1.0:2.0:1.3 molar ratio. This ratio resembles very closely the CER composition reported in the pig SC 38, except for CER ES. This CER subclass is present at an increased level (4.9 molar % in case of pig SC vs 13.3 molar % in our present study). The FFA composition was prepared from seven FFA subclasses; C16:0, C18:0, C20:0, C22:0, C23:0, C24:0 and C26:0 at molar ratios of 0.6:1.3:2.6:14.2:1.:11.5:1.4. The composition of FFA mixture is based on the FFA chain length distribution reported for SC 39. For preparing the deuterated lipid mixtures, the protiated CER ES and CHL were replaced by their deuterated counterparts. To perform contrast variation, the level of CER ES deuterated linoleate (ES DLIN) varied at the expense of the protiated CER ES (ES LIN) keeping the total level of CER ES constant. Five different molar ratios were prepared; 0, 3.3, 6.6, 9.9 and 13.3 which corresponds to the 0 %, 25 %, 50 %, 5 % and 100 % of the ES DLIN in the lipid mixtures. The composition of the model lipid mixtures prepared with their molar ratios and the total number of deuterated atoms per lipid molecule are provided in table 1. 15

6 H H H H (A) ω1 6 9 Linoleic acid (C18:2) Fatty acyl chain (C30) Sphingoid base H N (B) (C) H 2 3 H 4 6 H H Figure 1: The molecular structure of the deuterated lipids. A) CER ES with its linoleate moiety deuterated (bold part, total 31 deuterium atoms). B) CHL with its deuterated head-group region at position 2, 3, 4 and 6 (total 6 deuterium atoms). C) CHL deuterated in the tail of the molecule at a position of 25, 26 and 2 (total deuterium atoms). Preparation of the model lipid mixtures The sample preparation method and the equilibration procedure for all the lipid mixtures are the same as described previously (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). Briefly, the appropriate amount of lipids was dissolved in chloroform/methanol (2:1 v/v) solution at 10 mg/ml concentration. Subsequently the lipids were sprayed on a silicon substrate in an area of cm 2 using a Camag Linomat IV sample applicator (Muttenz, Switzerland). After equilibrating the samples at ~ 0 ⁰C, the samples were cooled and subsequently hydrated with D 2 buffer at 100 % relative humidity (RH) for about 15 hours for the first time at 3 C prior to the neutron diffraction measurement. The samples, in which the ES DLIN level varied were all measured at 100/0 D 2. In addition, the samples with 100 % ES DLIN were hydrated at four different D 2 ratios; 8/92, 50/50, 6/33 and 83/1 (v/v). The samples in which either the tail or the head group of CHL was partly deuterated were all measured at three different D 2 namely 8/92, 6/33 and 100/0 (v/v). The sequence of hydration of the samples using the various D 2 ratios was randomized. The hydration period between two measurements was around 12 hours. 158

7 PART II Molecular organization in the LPP unit cell Table 1: The lipid composition and molar ratios, the deuterated molar ratios and the total number of deuterated atoms per lipid molecule for the various lipid mixtures. The repeat distance with its standard error of all the lipid mixtures is also provided. The CERmix is composed of CER ES C30, CER NS C24, CER NP C24, CER AS C24, CER NP C16 and CER AP C24 in a 13.3:12:3.:1:2:1.3 molar ratio. Abbreviation Lipid composition and molar ratio (equimolar) Deuterated molar ratio No. of deuterated atoms per molecule Repeating unit D (Å) CER PR * CERmix / CHL / FFA ± 0.5 CER ES DLIN ** CERmix (CER ES (deuterated linoleate)) / CHL / FFA ± 0.3 CER CHL DH CERmix / CHL (deuterated head) / FFA ± 0.3 CER CHL DT CERmix / CHL (deuterated tail) / FFA ± 0.4 * The neutron diffraction profile of CER PR (only protiated lipids) has been published previously (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted) and used here as a control. ** In the CER ES DLIN mixture the level of CER ES DLIN varied between 0, 3.3, 6., 9.9 to 13.3 molar ratio of the total lipid components, keeping the CER ES LIN + CER ES DLIN at 13.3 molar ratio. This corresponds to a 0 %, 25 %, 50 %, 5 % and 100 % of the CER ES DLIN in the lipid mixtures. Neutron diffraction experiment Neutron diffraction data were collected at the D16 cold neutron diffractometer of the Institute Laue-Langevin (ILL) located in Grenoble, France. The wavelength of the neutrons 4.5 Å was achieved by the reflection of a highly ordered pyrolytic graphite (HPG) monochromator. The sample to detector distance was 0.85 m and all the samples were measured in the reflection mode. The samples were mounted on a goniometer placed in an aluminum chamber. The temperature of the chamber was maintained at 25 C throughout the measurements. During the sample measurements, the bottom of the chamber was filled with the same D 2 ratio as used for hydrating the sample, to maintain a 100 % RH at a constant D 2 ratio. The measurement time per sample varied between 10 to 12 hours depending on the signal/noise ratio. The neutron scattering density was recorded by a position sensitive two dimensional 3 He detector ( mm area with a spatial resolution of 1 1 mm). Data reduction procedure A water calibration measurement was performed to correct for the intensity differences on the detector surface. An empty chamber was measured as background and subtracted from each measurement to increase the signal/noise ratio. For data analysis, the ILL inhouse software LAMP was used 40. During the measurements, the sample was rotated in 159

8 steps of 0.1 degree from 0 to 10.5 degree to cover all the 8/9 diffraction orders and the detector images were taken at each step. The two dimensional detector image was vertically integrated which results in one dimensional diffraction pattern of scattering intensities (I) vs scattering angle (2θ). The scattering angle is later converted to the scattering vector (q): 4. π. sinθ q = (1) λ In this equation, θ is the Bragg angle, λ is the wavelength of the neutron beam. The repeat distance (d) was calculated from the positions of a series of equidistant peaks attributed to the lamellar phase (q h ), in which h is the diffraction order: d 2. π. h = (2) q h All the diffraction orders were fitted with the Gaussian function (f(x)): 2 ( x µ ) 2 2σ ( x) = ae + c (3) Here, c is the offset for the baseline correction. The structure factor amplitude F h of each diffraction order was then calculated from the Gaussian peak height (I h ) by using the formula: F A L. I h = (4) h h The Lorentz correction (L) values were calculated for all the mixtures and applied to correct for the intensities. The sample absorption correction factor (A h ) was calculated using the following equation 31 : A h = 1 sinθ 1 e 2µ l 2µ l sinθ (5) Here µ represents the linear attenuation coefficient and l is the thickness of the lipid 160

9 PART II Molecular organization in the LPP unit cell film. In a previous study the thickness of the lipid film was calculated to be 24 µm 34. The attenuation coefficients were calculated using the wavelength of the neutrons in combination with the lipid density ( ~ 0.83 g/cc) and the chemical composition of the lipid films 41. Using the D 2 contrast variation, previously it was reported that the LPP unit cell is centrosymmetric (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). To determine the phase signs of water located in the unit cell of the LPP, the assumption was made that the water is expected to be located near the hydrophilic head-groups situated at the boundaries of the unit cell. By a systematic variation of the phase signs of various diffraction orders, the most realistic phase combination was selected. This phase sign is -, +, -, +, -, +, -, + and - for the first 9 diffraction orders as reported previously (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). Using these phase signs the localization of the water was determined. The water was not only present at the boundary of the unit cell but also at two other distinct positions inside the unit cell ( ~ 20 Å from the unit cell center). A more detailed explanation for the choice of the phase sign assignments is given in our previous work (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). The water layer structure factors are defined as the structure factors at 100 % D 2 minus the structure factors at 8 % ES DLIN D 2. The structure factor phase signs of the various diffraction orders of the CER (100 % ES DLIN) and deuterated CHL (CER CHL DH and CER CHL DT ) samples were determined from the linear plot of the structure factors vs D 2 ratios such that the difference between 100 % D 2 and 8 % D 2 corresponds to the correct phase sign for the water layer structure factor at that particular diffraction order. A more detailed description is given elsewhere 31, 32, 42. In obtaining additional information on the localization of CER ES, the following procedure was used: the structure factor amplitudes were plotted as a function of ES DLIN/ES LIN ratio between 0 % and 100 % for the various diffraction orders. To determine the slope of these diffraction orders, we used the phase signs of the structure factors of the 100 % ES DLIN that were obtained from the D 2 contrast variation. The structure factor amplitudes for the various diffraction orders of all lipid mixtures measured at different contrast variation levels are provided in the supplemental table 1 with their uncertainties. The SLD profile across the unit cell ρ(x) was calculated by Fourier reconstructions: h 2. π. h. x ( x) F + 2 max F cos ρ (6) = 0 h h= 1 d Here, x is the direction normal to the unit cell surface and x = 0 is the center of the unit cell. The average scattering density per unit volume (F 0 ) was calculated in order to put the data on an absolute scale. Using the chemical composition and the mass density of the sample, the F 0 was calculated 33, 43. The data were then placed on a relative absolute scale 25, ; the known neutron SLDs of deuterium and hydrogen were used to scale the differ- 161

10 ences in such a way so that the area differences between the SLDs is equal to the SLD of the deuterium label, for instance, deuterium difference in case of CHL deuterated tail. The difference density profiles were then constructed by subtracting the protiated profiles from the deuterated profiles. RESULTS The lipid mixtures measured with neutron diffraction contain either deuterated CER ES (18 C linoleate moiety is deuterated) or deuterated CHL (deuterated in the headgroups or in the tail of the molecule) (see figure 1). The repeat distance of all the lipid mixtures measured with neutron diffraction was calculated by a least square fitting method and are provided in table 1. The mean repeat distance of all the samples was calculated to be ± 0.5 Å. A one dimensional intensity vs q diffraction plot for CER ES DLIN is provided in figure 2 showing the various diffraction orders. The pattern exhibit diffraction peaks attributed to the LPP and crystalline phase separated CHL. No additional phases were observed. The protiated lipid mixture (CER PR ) used as a control in our present studies has been reported previously (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). In this lipid mixture the water was not only present at the boundaries of the unit cell but also at two well defined regions inside the unit cell ( ~ 20 Å from the unit cell center). The water profile was obtained by D 2 contrast variation with phase signs of the Fourier transforms of the diffraction peaks of the LPP being -, +, -, +, -, +, -, + and - for the first 9 diffraction orders. Figure 2: Neutron diffraction one dimensional plot of intensity vs q for the CER ES DLIN (100 % ES DLIN) hydrated and measured at 100 % D 2. The higher diffraction orders of the LPP lamellae are indicated by the Arabic numbers and the CHL peaks by means of an asterisk. 162

11 PART II Molecular organization in the LPP unit cell In the present study, we performed neutron diffraction measurements for the 100 % ES DLIN sample as a function of D 2 contrast variation. A linear relationship in the structure factor amplitudes vs D 2 contrast variation at five different D 2 ratios for the various diffraction orders indicate a centrosymmetric lipid arrangement in the unit cell of the LPP (figure 3A). In addition, we also performed measurements using an increasing level of ES DLIN (ES DLIN/ES LIN contrast variation). In these experiments the samples were hydrated and measured at 100 % D 2. A linear relation between the structure factors of these samples as a function of ES DLIN increment was also observed, demonstrating a centrosymmetric structure (figure 3B), that is not only water, but also the linoleate moiety of CER ES is located symmetrically in the unit cell of the LPP. (A) (B) Structurefactor (a.u.) Structurefactor (a.u.) D 2 % CER ES DLIN (%) ES DLIN Figure 3: Relative structure factor amplitudes of the various diffraction orders for the CER mixture A) as a function of D 2 exchange ratio and B) as a function of ES DLIN/ES LIN contrast variation. The numbers in the plot indicate the different diffraction orders. The structure factor phase signs of CER ES DLIN measured at 100 % ES DLIN and 100 % D 2 is used in the D 2 contrast variation (figure 3A) as well as in the ES DLIN/ES LIN contrast variation (figure 3B). As this concerns the same sample, the phase signs of the structure factors at 100 % D 2 and at 100 % ES DLIN/ ES LIN are equal. First the phase signs of the 100 % ES DLIN sample were determined from the linear plot as a function of D 2 contrast variation, using the correct water layer structure factors for the various diffraction orders. These phase signs for the first 8 orders are +, -, -, -, +, +, - and + (figure 3A). Applying these phase signs to the 100 % ES DLIN/ES LIN contrast variation and assuming a straight line for each of the diffraction 163

12 orders in the contrast variation plot in figure 3B results in the phase signs for the ES DLIN/ES LIN contrast variation samples (+, -, -, -, +, +, - and - for the first 8 orders). The structure factor phase signs at 0 % ES DLIN/ES LIN at 100 % D 2 were also determined in our previous study (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). When comparing these phase signs, an excellent matching is observed (-, +, -, +, -, +, - and + for the first 8 diffraction orders) (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted) (figure 3B). Based on the structure factors in combination with the phase signs and using equation 6, the neutron scattering length density (SLD) profiles in the unit cell of the LPP were constructed. Figure 4A (bottom) displays the SLD profiles of the CER ES DLIN constructed at 8 % and 100 % D 2. The difference between 8 % and 100 % D 2 indicate the water profile (figure 4A, top solid line); four elevated scattering density regions (two at the border and two inside the unit cell) are observed, indicating the positions of the hydrophilic head-groups in these regions. The position of the inside water regions in the netto D 2 profile is located at ~ 21 ± 0.3 Å from the unit cell center (determined by Gaussian peak fitting procedure). 164

13 PART II Molecular organization in the LPP unit cell Figure 4: A) The relative SLD profiles for the CER ES DLIN mixture consisted of 100 % ES DLIN hydrated and measured at 8 % (dashed line) and 100 % (solid line) D 2 (bottom). The difference plot between 8 % and 100 % D 2 shows the water profile (top solid line). B) The SLD profiles constructed for 0 % (thin) and 100 % (thick) ES DLIN hydrated and measured at 100 % D 2. The difference between the SLDs indicated by the dashed line shows the position of the linoleate moiety in the LPP unit cell. Figure 4B depicts the SLD profiles generated for the CER ES DLIN mixture at 0 % and 100 % ES DLIN hydrated and measured at 100 % D 2. The difference profile between 0 % and 100 % ES DLIN shows the position of the linoleate moiety. This profile clearly indicates the maximum of the position of the linoleate in the LPP unit cell at ~ 21.4 ± 0.2 Å from the unit cell center. The position of these maxima is very similar as the position of water regions inside the unit cell. In order to obtain further details on the molecular arrangement in the unit cell of the LPP, the localization of CHL (deuterated head-group and the deuterated tail region) in the LPP unit cell was also determined. Figure 5 displays one dimensional diffraction pattern of the scattered intensity vs q plots for CER CHL DH (figure 5A) and CER CHL DT mixture (figure 5B). All the diffraction peaks show slight peak asymmetry due to the large size of the samples in respect to distance to the detector, most evidently at lower q values. The linear relationship between the structure factor amplitudes of various diffraction orders vs D 2 contrast of these two lipid mixtures are shown in supplemental figure S2. This linearity demonstrates that CHL is also symmetrically located in the unit cell. Using the D 2 contrast variation at 3 different ratios (8, 66 and 100 % D 2 ), the structure factor phase signs of the various diffraction orders in the diffraction pattern of the deuterated CHL containing lipid mixtures were determined: the difference between

14 % and 8 % D 2 correspond to the correct phase signs for the water layer structure factor at that particular order. The phase signs for the first 9 diffraction orders of the deuterated CHL containing lipid mixtures are -,+, -, +, -, +, -, + and - respectively. Figure 5: ne dimensional neutron diffraction plot of intensity vs q for the CER CHL DH mixture (figure 5A) and CER CHL DT mixture (figure 5B) hydrated and measured at 100 % D 2. The various diffraction orders of the LPP lamellae are indicated by the Arabic numbers and the CHL peaks by means of an asterisk. Inset; the 1 st order diffraction pattern of the CER CHL DH and CER CHL DT mixtures measured at different detector position. The 8 th order diffraction peak is not visible in the spectra, indicating the continuous form factor is zero at this point. To locate the molecular position of the CHL in the unit cell of the LPP, the SLD profiles of the deuterated CHL (head group or tail) were constructed and are presented in figure 6A (CER CHL DH ) and 6B (CER CHL DT ) along with protiated profiles (CER PR ) measured previously (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). The difference density profiles for the deuterated CHL were obtained by subtracting the protiated profile either from the profile of deuterated CHL head or from the deuterated CHL tail profile (shown in dashed lines). The difference profiles show the position of the CHL in the unit cell of the LPP; CHL head-group resides slightly outward from the inside head-group regions at a position of ~ 26.0 ± 0.2 Å from the unit cell center, with the CHL tail ~ 42.0 ± 0.2 Å from the center of the unit cell (determined after Gaussian peak fitting). Therefore, the total length over which CHL is present in the unit cell is ~ 16 Å, similar to the length of CHL. 166

15 PART II Molecular organization in the LPP unit cell Figure 6: The absolute neutron SLD profiles for the A) CER CHL DH and B) for CER CHL DT mixture hydrated and measured at 100 % D 2 (thick lines). The profiles for CER PR mixture at 100 % D 2 are given in thin lines. The difference profiles indicating the molecular position of the CHL headgroup and tail are shown in the dashed lines. 16

16 DISCUSSIN In our work, we focus on the lipid organization in the SC as the lipids play a prominent role in the skin barrier function. In previous studies, we examined the lipid organization in SC and detected two lamellar phases, the LPP and SPP 19. In subsequent studies using lipid model systems the role the lipids play in the SC lipid organization was studied and it was observed that CERs and CHL are crucial for the formation of lamellar organization 4, 48, while FFAs enhance the formation of an orthorhombic lateral packing 39. Furthermore, when focusing on the CER subclasses, CER ES is crucial for the formation of the LPP 25. In our recent studies, our aim is to elucidate the molecular arrangement of lipids within the unit cell of the two lamellar phases. First, we examined the molecular arrangement in the unit cell of the SPP 25, 34. Using neutron diffraction experiments in combination with contrast variation, it was possible to locate CHL and the most abundant CER and FFA in the SPP unit cell. In the present study we aimed to determine the arrangement of the lipid (sub)classes in the unit cell of the LPP. In previous studies the electron density profile of the unit cell of the LPP was determined. This profile indicated a three layer arrangement of the lipids in the unit cell 3, but the localization of the lipid (sub)classes in this arrangement could not be determined. In a recent study using neutron diffraction we demonstrated a symmetric three layer arrangement very similar to that indicated by the electron density profile (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). In addition, we were able to localize the CER NS and the FFA C24 in the unit cell of the LPP. Both lipids are most abundantly present in the central layer of the unit cell (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). The aim of this study was to determine the localization of CER ES and CHL. These lipids are important building blocks of the LPP as it has been demonstrated that CER ES and CHL are both indispensable for the formation of the LPP ( 24, submitted for publication). In order to be able to form exclusively the LPP, in the present study we increased the level of CER ES in the lipid mixtures from around 5 %, the level approximately present in pig SC to 13.3 % molar fraction 49. Indeed at this CER ES level no SPP is formed and the diffraction peaks attributed to phase separated crystalline CHL does not interfere with the diffraction peaks assigned to the LPP. The rationale behind choosing the CER composition is explained in more details in our previous study (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). Localization of CER ES in the LPP The maximum of the peak attributed to CER ES DLIN in the difference profile of the SLD is located at ~ 21.4 ± 0.2 Å from the unit cell center (evident from Gaussian peak fitting of the SLD profile as shown in figure 4B). This maximum location of linoleate in the unit cell is approximately at the same position as the inner water layer regions in the unit cell, which are ~ 20 Å from the unit cell center. This position also corresponds to the 168

17 PART II Molecular organization in the LPP unit cell inner head group region in the unit cell. The CER ES consists of 30 C in acyl chain, the ɷ-hydroxy moiety of which is ester bound to the carboxyl group of the 18 C linoleate moiety (containing two double bonds). The linoleate in this region exists in a highly disordered state, which is evident from the high CH 2 stretching band frequencies observed in the Fourier transform infrared spectroscopy studies 50. Therefore, the linoleate chain may wiggle around and also partly fold back at the position of the inner head-group regions and can compensate for two acyl chains directing from the opposite head-group regions. When focusing on the profile of CER PR at 100 % D 2 (figure dashed line), the inner head-group regions are less pronounced compared to the outer border, indicating the CER and FFA head-groups are compensated by the linoleate in that region, giving rise to a lower intensity. A schematic drawing of CER ES depicting its arrangement in the LPP unit cell is shown in figure. The importance of unsaturation of the linoleate chain The location of the unsaturated linoleate moiety in the hydrophilic head-group region is unusual and the interactions that takes place in this region is not fully understood. However, some interesting observations have been made. i) Replacing CER ES linoleate by CER ES stearate (no unsaturation) hampers the lipids to form the LPP 51. Therefore, a high degree of mobility of the C18 chain is a prerequisite to form the LPP. Possibly a part of the unsaturated chain is folding back extending to the head group region is a prerequisite for the formation of the LPP. ii) The (partially) folded linoleate chain in the inner head-group regions can compensate for the long chains of CERs and FFAs from the opposite direction, which extends the central region of the unit cell reaching the opposite head group region approximately at a distance of 10 Å, corresponding to a chain length of ~ -8 C atoms. The gap between the localization of the ester bond of CER ES and the CH 3 groups of the extended CERs and FFAs is around 1 Å, while a fully extended linoleate is around 23 Å in length, again indicating conformational disordering or folding back of the linoleate moiety. Localization of CHL in the unit cell of the LPP The head and tail position of the sterol molecule in the SLD profiles of the LPP unit cell demonstrate that CHL spans a distance of ~ 16 Å in length (evident from Gaussian peak fitting of the SLD profiles of CHL), indicating that the molecule is almost oriented perpendicular to the basal plane (a schematic drawing of the CHL is provided in figure ). The CHL head-group is located at a position of ~ 26 Å from the unit cell center and this is around 38 Å from the unit cell boundary. If the head-group region of CER ES is at the unit cell boundary, the 30 C acyl chain of CER ES directing from the unit cell border into the direction of the center of the unit cell expands ~ 3 Å (30 C 1.25 Å) towards the unit cell center ending in the ester bond linking the linoleate. This mean that the ester bond is present at a similar position in the unit cell as the head-group of CHL. This may allow the formation of hydrogen bond between the H group of CHL and 169

18 the carbonyl group (C = ) of the CER ES in this region. In this way the hydrophobic part of CHL is arranged close to the saturated acyl chains of the CERs providing strong van der Waals interactions. In a previous study an asymmetric distribution of CHL near the outer border of the two lipid bilayers in the unit cell of ~ 130 Å has been reported 36 which resembles very similar position of CHL to that observed in this study. ~38 Å ~28 Å ~3 Å Figure : A schematic drawing of the position of CER ES and CHL in the unit cell of the LPP. The positions are based on their difference profiles in the LPP unit cell: thin line, CHL head deuterated and thick line, CHL tail deuterated. The dashed line indicate CER PR profile at 100 % D 2. Note that, the position of the CER ES linoleate moiety is approximately at the same position of inner head-group region of the CER PR profile ( ~ 21 Å from the unit cell center). This is also the location of inner water layers within the unit cell. The impact of CHL in the LPP lipid lamellae CHL plays a key role in the formation of the LPP. ur recent studies demonstrate that in the absence of CHL the lipids do not adopt to the LPP (EH Mojumdar, GS Gooris, JA Bouwstra, manuscript submitted). These studies indicate a minimum 0.2 CHL level in the CER/CHL/FFA (1:0.2:1) mixture is required for the formation of the LPP, resulting in an CHL/CER ES ratio of 1:2. Furthermore, these data also suggest that CHL not only participate in the formation of the LPP, but also increases the packing density of the lipids in the LPP. ur neutron diffraction studies show that CHL is located in the outer layer of the LPP with its head-group close to the inner head-group region of the LPP. The positioning of CHL in this region can fill the gap between the inner head group and the chains of the sphingoid base of CERs and FFAs directing from the unit cell border and may provide stability in the structure. 10

19 PART II Molecular organization in the LPP unit cell Molecular model of the LPP A molecular model for the LPP lipid lamellae is proposed (figure 8) based on the location of the lipid molecules determined in our current (CER ES and CHL) and previous (CER NS and FFA C24) studies (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). As outlined above, in summary these data show that i) the central lipid layer of the unit cell of the LPP ( ~ 42 Å) is dominated by CER NS and FFA C24 with substantial interdigitation of the acyl chains in the unit cell center (Mojumdar EH, Gooris GS, Barlow DJ, Lawrence MJ, Deme B, Bouwstra JA, Biophysical Journal, submitted). The arrangement of these molecules in the central part leaves a gap of around 10 Å corresponding to ~ 8 C atoms between the end of the extended CER and FFA acyl chains and the inner head-group regions. Besides a fraction of the CER NS and FFA C24 is present in the outer lipid layer. ii) As CER NS and FFA C24 occupy a substantial amount of space in the central lipid layer, it was expected that the CER ES head group is present close to the unit cell border. The 30 C acyl chain of CER ES extends from the unit cell border, which results in the localization of its esterfied hydroxy-group at ~ 3 Å from the unit cell border (figure ). iii) The position of CHL head-group region is very similar. ( ~ 38 Å from the border of the unit cell) (figure ). Therefore, it is likely that the H group of CHL may form hydrogen bond with the ester bond of CER ES. iv) Based on the location of the ester bond, CER ES linoleate extends ~ Å ( ~ 5-6 C atoms) in the direction of the unit cell center before it reaches the inner head-group region and the total distance between the ES ester bond and the tails of CER NS and FFA C24 is around 1 Å. Therefore, it is likely that the linoleate partly folds back or shows a high conformational disordering, as is indeed observed by FTIR 50. Besides the results presented in the present studies, we have information from other studies which also support our model. i) Recently, we observed that the LPP can also be formed using a less complex composition. nly 4 lipid (sub) classes are required, that is CER ES C30, CER NS C24 (0.4:0.6 molar ratio), CHL and FFA C24 in which the CER/CHL/ FFA are in an equimolar ratio (EH Mojumdar, GS Gooris, JA Bouwstra, unpublished data). This indicate that a variation in the head-group architecture of the CER subclasses is not required for the formation of the LPP and therefore a model composed of these four lipid classes as in figure 8 is sufficient to show the arrangement. ii) The maximum level of CHL that can be incorporated in the LPP is 0.5 molar ratio in the CER/CHL/FFA (1:0.5:1) mixture (EH Mojumdar, GS Gooris, JA Bouwstra, submitted for publication), while a minimum level of 0.2 molar CHL is required for the formation of the LPP. This reflect a certain flexibility in the amount of CHL that can be incorporated in the crystalline LPP. In addition to the equimolar CER/CHL/FFA lipid mixtures, we observed that a gradual increase in CER ES level results in a gradual increase in the repeat distance of the LPP (unpublished results). This can be explained by the formation of domains of CER ES surrounded by CHL interacting with the ester bonding of CER ES (figure 8). The partially folding back of the linoleate will fill the gap between the hydroxy group of CHL and the terminal CH 3 groups of CER NS or FFA present in the central lipid layer. When the level of CER ES is gradually increased keeping the CHL level constant, the 11

20 domains are enlarged and hardly no additional CHL is required for the formation of the LPP, demonstrating a certain flexibility of the LPP with respect to its CHL/CER ES ratio in agreement with the observations. Furthermore, when increasing the level of CER ES the average additional space for each linoleate moiety provided by the gap between the terminal CH 3 groups of CER NS or FFA and the hydroxy group of CHL will be reduced. Therefore to create sufficient space for the linoleate an increase in repeat distance is required when increasing the level of CER ES. Figure 8: The proposed molecular model of the unit cell of the LPP based on the localization of lipids determined in our current and previous neutron diffraction experiments. The model is presented in fully extended conformation. CNCLUSINS In the present study using neutron diffraction, we examined the molecular arrangement of lipids in the unit cell of the LPP. The neutron diffraction reveals the arrangement of CER ES linoleate and CHL in the ~ 130 Å repeating unit: the linoleate moiety is located at the position of the inner head-group regions of the LPP in a highly disorder state while CHL is located in the outer layer but close to the inner head-group regions of the LPP. The results obtained from our current and previous works demonstrate a detailed molecular arrangement of lipids in the unit cell of the LPP and SPP. This information would be beneficial in the future to examine the molecular localization of other molecules (drug compounds, moisturizers etc.) in the unit cell of these lamellar phases to get more insight about their interaction with the SC lipids. ACKNWLEDGEMENTS We would like to thank the company Evonik (Essen, Germany) for their generous provision of CERs. We also like to thank the personnel at the ILL in Grenoble, France for their assistance and allocation of beam time for neutron diffraction measurements. 12

21 PART II Molecular organization in the LPP unit cell REFERENCES 1. C. Bonnart, C. Deraison, xe, line, M. Lacroix, Y. Uchida, C. Besson, xe, line, A. Robin, xe, lie, A. Briot, xef, M. Gonthier, L. Lamant, P. Dubus, B. Monsarrat and A. Hovnanian. Elastase 2 is expressed in human and mouse epidermis and impairs skin barrier function in Netherton syndrome through filaggrin and lipid misprocessing. The Journal of Clinical Investigation, 120:81-882, J. N. Ishikawa, H.; Kondo, N.; Hotta, M.; Takagi, Y.; Masukawa, Y.; Kitahara, T.; Takema, Y.; Koyano, S.; Yamazaki, S.; Hatamochi, A. Changes in the Ceramide Profile of Atopic Dermatitis Patients. J. Invest. Dermatol., 130: , S. Motta, M. Monti, S. Sesana, L. Mellesi, R. Ghidoni and R. Caputo. Abnormality of water barrier function in psoriasis: Role of ceramide fractions. Archives of Dermatology, 130: , D. G. Paige, N. Morse-Fisher and J. I. Harper. Quantification of stratum corneum ceramides and lipid envelope ceramides in the hereditary ichthyoses. Br. J. Dermatol, 131:23-2, Y.-H. Park, W.-H. Jang, J. A. Seo, M. Park, T. R. Lee, Y.-H. Park, D. K. Kim and K.-M. Lim. Decrease of Ceramides with Very Long-Chain Fatty Acids and Downregulation of Elongases in a Murine Atopic Dermatitis Model. J Invest Dermatol, 132:46-49, T.. Sassa, Y.; Suzuki, S.; Nomura, T.; Nishioka, C.; Kashiwagi, T.; Hirayama, T.; Akiyama, M.; Taguchi, R.; Shimizu, H.; Itohara, S.; Kihara, A. Impaired Epidermal Permeability Barrier in Mice Lacking Elovl1, the Gene Responsible for Very-Long-Chain Fatty Acid Production. Molecular and Cellular Biology 33:28-296, E. Proksch, J. M. Brandner and J.-M. Jensen. The skin: an indispensable barrier. Experimental Dermatology, 1: , P. Talreja, G. Kasting, N. Kleene, W. Pickens and T.-F. Wang. Visualization of the lipid barrier and measurement of lipid pathlength in human stratum corneum. AAPS PharmSci, 3:48-56, H. E. Boddé, I. van den Brink, H. K. Koerten and F. H. N. de Haan. Visualization of in vitro percutaneous penetration of mercuric chloride; transport through intercellular space versus cellular uptake through desmosomes. Journal of Controlled Release, 15:22-236, H. E. Boddé, M. A. M. Kruithof, J. Brussee and H. K. Koerten. Visualisation of normal and enhanced HgCl2 transport through human skin in vitro. International Journal of Pharmaceutics, 53:13-24, P. W. Wertz, M. C. Miethke, S. A. Long, J. S. Strauss and D. T. Downing. The Composition of the Ceramides from Human Stratum Corneum and from Comedones. J Investig Dermatol, 84: , K. J. Robson, M. E. Stewart, S. Michelsen, N. D. Lazo and D. T. Downing. 6-Hydroxy-4-sphingenine in human epidermal ceramides. Journal of Lipid Research, 35: , M. E. Stewart and D. T. Downing. A new 6-hydroxy-4-sphingenine-containing ceramide in human skin. Journal of Lipid Research, 40: , Y. Masukawa, H. Narita, E. Shimizu, N. Kondo, Y. Sugai, T. ba, R. Homma, J. Ishikawa, Y. Takagi, T. Kitahara, Y. Takema and K. Kita. Characterization of overall ceramide species in human stratum corneum. Journal of Lipid Research, 49: , A. Weerheim and M. Ponec. Determination of stratum corneum lipid profile by tape stripping in combination with high-performance thin-layer chromatography. Archives of Dermatological Research, 293: , M. Rabionet, K. Gorgas and R. Sandhoff. Ceramide synthesis in the epidermis. Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids, 1841: , J. van Smeden, L. Hoppel, R. van der Heijden, T. Hankemeier, R. J. Vreeken and J. A. Bouwstra. LC/MS analysis of stratum corneum lipids: ceramide profiling and discovery. Journal of Lipid Research, 52: , M. Ponec, A. Weerheim, P. Lankhorst and P. Wertz. New Acylceramide in Native and Recon- 13

22 structed Epidermis. J Investig Dermatol, 120: , J. A. Bouwstra, G. S. Gooris, J. A. van der Spek and W. Bras. Structural Investigations of Human Stratum Corneum by Small-Angle X-Ray Scattering. J Investig Dermatol, 9: , K. C. Madison, D. C. Swartzendruber, P. W. Wertz and D. T. Downing. Presence of Intact Intercellular Lipid Lamellae in the Upper Layers of the Stratum Corneum. J Investig Dermatol, 88:14-18, S. H. White, D. Mirejovsky and G. I. King. Structure of lamellar lipid domains and corneocyte envelopes of murine stratum corneum. An x-ray diffraction study. Biochemistry, 2: , I. Hatta, N. hta, K. Inoue and N. Yagi. Coexistence of two domains in intercellular lipid matrix of stratum corneum. Biochim. Biophys. Acta, Biomembr., 158: , M. W. de Jager, G. S. Gooris, I. P. Dolbnya, W. Bras, M. Ponec and J. A. Bouwstra. Novel lipid mixtures based on synthetic ceramides reproduce the unique stratum corneum lipid organization. Journal of Lipid Research, 45: , M. Janssens, J. van Smeden, G. S. Gooris, W. Bras, G. Portale, P. J. Caspers, R. J. Vreeken, T. Hankemeier, S. Kezic, R. Wolterbeek, A. P. Lavrijsen and J. A. Bouwstra. Increase in shortchain ceramides correlates with an altered lipid organization and decreased barrier function in atopic eczema patients. Journal of Lipid Research, 53: , E. H. Mojumdar, D. Groen, G. S. Gooris, D. J. Barlow, M. J. Lawrence, B. Deme and J. A. Bouwstra. Localization of Cholesterol and Fatty Acid in a Model Lipid Membrane: A Neutron Diffraction Approach. Biophysical Journal, 105: , A. Schröter, D. Kessner, M. A. Kiselev, T. Hauß, S. Dante and R. H. H. Neubert. Basic Nanostructure of Stratum Corneum Lipid Matrices Based on Ceramides [ES] and [AP]: A Neutron Diffraction Study. Biophysical journal, 9: , D. Kessner, M. Kiselev, S. Dante, T. Hauß, P. Lersch, S. Wartewig and R. H. Neubert. Arrangement of ceramide [ES] in a stratum corneum lipid model matrix: new aspects revealed by neutron diffraction studies. European Biophysics Journal, 3: , A. Ruettinger, M. Kiselev, T. Hauss, S. Dante, A. Balagurov and R. Neubert. Fatty acid interdigitation in stratum corneum model membranes: a neutron diffraction study. European Biophysics Journal, 3:59-1, G. Zaccai, J. K. Blasie and B. P. Schoenborn. Neutron Diffraction Studies on the Location of Water in Lecithin Bilayer Model Membranes. Proceedings of the National Academy of Sciences, 2:36-380, E. Pebay-Peyroula, E. J. Dufourc and A. G. Szabo. Location of diphenyl-hexatriene and trimethylammonium-diphenyl-hexatriene in dipalmitoylphosphatidylcholine bilayers by neutron diffraction. Biophysical Chemistry, 53:45-56, N. P. Franks and W. R. Lieb. The structure of lipid bilayers and the effects of general anaesthetics: An X-ray and neutron diffraction study. Journal of Molecular Biology, 133: , D. L. Worcester and N. P. Franks. Structural analysis of hydrated egg lecithin and cholesterol bilayers II. Neutron diffraction. Journal of Molecular Biology, 100:359-38, T. A. Harroun, J. Katsaras and S. R. Wassall. Cholesterol Hydroxyl Group Is Found To Reside in the Center of a Polyunsaturated Lipid Membrane. Biochemistry, 45: , D. Groen, G. S. Gooris, D. J. Barlow, M. J. Lawrence, J. B. van Mechelen, B. Demé and J. A. Bouwstra. Disposition of Ceramide in Model Lipid Membranes Determined by Neutron Diffraction. Biophysical journal, 100: , I. Iwai, H. Han, L. d. Hollander, S. Svensson, L.-G. fverstedt, J. Anwar, J. Brewer, M. Bloksgaard, A. Laloeuf, D. Nosek, S. Masich, L. A. Bagatolli, U. Skoglund and L. Norlen. The Human Skin Barrier Is rganized as Stacked Bilayers of Fully Extended Ceramides with Cholesterol Molecules Associated with the Ceramide Sphingoid Moiety. J Invest Dermatol, 132: , T. J. McIntosh. rganization of Skin Stratum Corneum Extracellular Lamellae: Diffraction Evidence for Asymmetric Distribution of Cholesterol. Biophysical Journal, 85: ,