3D Molecular ToF-SIMS Imaging of Artificial Lipid ... - ACS Publications

Jul 3, 2018 - components and usually not all substances can be displayed at the same time. ..... were used in order not to exceed the static limit. Hi...
0 downloads 0 Views 1MB Size
Subscriber access provided by University of Winnipeg Library

Interfaces: Adsorption, Reactions, Films, Forces, Measurement Techniques, Charge Transfer, Electrochemistry, Electrocatalysis, Energy Production and Storage

3D Molecular ToF-SIMS Imaging of Artificial Lipid Membranes Using a Discriminant Analysis Based Algorithm Rainer Kassenböhmer, Marcel Heeger, Mridula Dwivedi, Martin Körsgen, Bonnie J. Tyler, Hans-Joachim Galla, and Heinrich F. Arlinghaus Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b01253 • Publication Date (Web): 03 Jul 2018 Downloaded from http://pubs.acs.org on July 12, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

3D Molecular ToF-SIMS Imaging of Artificial Lipid Membranes Using a Discriminant Analysis Based Algorithm Rainer Kassenböhmer†*, Marcel Heeger†, Mridula Dwivedi‡, Martin Körsgen†, Bonnie J. Tyler†, Hans-Joachim Galla‡, Heinrich F. Arlinghaus† †Physikalisches Institut, Westfälische Wilhelms-Universität Münster, Wilhelm-Klemm-Straße 10, 48149 Münster, Germany, ‡Institut für Biochemie, Westfälische Wilhelms-Universität Münster, Wilhelm-Klemm-Straße 2, 48149 Münster Abstract Artificial lipid membranes play a growing role in technical applications such as biosensors, in pharmacological research and as model systems in the investigation of biological lipid films. In the standard procedure for displaying the distribution of membrane components, fluorescence microscopy, the fluorophores used can influence the distribution of the components and usually not all substances can be displayed at the same time. The discriminant analysis (DA) based algorithm used in combination with scanning ToF-SIMS enables marker-free, quantitative, simultaneous recording of all membrane components. These data are used for reconstruction of distribution patterns. In the model system used for this survey, a tear fluid lipid layer, the distribution patterns of all lipids correlate well in calculated ToF-SIMS images and epi-fluorescence microscopic images. All epi-fluorescence microscopically viewable structures are visible when using both positive and negative secondary ions and can be reproduced with high lateral resolution in the sub-micrometer range despite the very low signal intensity and a very low signal-to-noise ratio. In addition, three-dimensional images can be obtained with a sub-nanometer depth resolution. Furthermore, structures and the distribution of substances that cannot be made visible by epi-fluorescence microscopy can be displayed. This enables new insights that cannot be gained by epi-fluorescence microscopy alone.

Introduction Membranes envelop both cells and organelles as the smallest structural and functional units of the organism. In their basic structure they consist of a double layer of phospholipids with interlaced cholesterol.1,2 There are also other differently constructed and composed natural lipid films such as lung surfactant and tear film with important physiological functions. To date, a large number of biochemical processes involving lipid membranes have been insufficiently clarified or are not understood at all.3-5 Many physicochemical and biological properties of natural lipid films can be investigated using artificial lipid membranes as model systems.6-13 In addition, there is an increasing number of technical applications based on lipid 14-19 membranes or liposome platforms, e. g. biosensors in medical diagnostics and environmental monitoring. Artificial 20-22 membranes are also playing an increasingly important role in pharmacological research. Usually the distribution of individual membrane components is visualized with high lateral resolution by fluorescence 23-25 26 microscopy. However, fluorophores can affect the properties of the membrane components themselves. Alternatively, imaging time-of-flight secondary ion mass spectrometry (ToF-SIMS), with which simultaneous sputtered ions over a wide mass range can be detected, provides a label-free high-resolution scanning technique.27-31 In static SIMS mode, ToF-SIMS is highly surface sensitive and solely the outermost monolayer is detected. If dynamic SIMS 32,33 mode is used, even three-dimensional images are possible. Essential limitations of the method are that the often relatively large biological molecules are severely fragmented and the signal intensity and mass resolution decrease significantly with increasing spatial resolution, which makes the imaging of particular substances difficult or impossible. To overcome these limitations, an algorithm which makes it possible to calculate the percentages of all components contained in a mixture from mass spectra with only low signal intensity and high fragmentation can be used.34 To achieve this, the algorithm uses multivariate DA, which allows the consideration of non-characteristic but relative frequent small fragments, provided that all possible constituents are included in DA. This might even allow the threedimensional distribution of each individual component to be visualized. The goal of this study was to validate the algorithm and demonstrate its utility by comparing calculated distribution patterns with epi-fluorescence microscopic images. To accomplish this objective, a well-documented system of artificial lipid films developed by Dwivedi et al.35 was used. The films consisted of binary and ternary lipid mixtures of dipalmitoyl phosphatidylcholine, cholesteryl esters and triacyl-glycerols. The lipid monolayers can be found in two co-existing liquid phases, the liquid-expanded (LE), and the liquid-condensed (LC) phase.36 The dye used for epi-fluorescence microscopy accumulates more in the LE than in the LC phase, thus, the LE phase appears brighter.37 In previous work35 epi-fluorescence microscopy was employed to

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

study the effect of ectoine, a low molecular weight organic solute which increases spacing between lipid headgroups, on shape and composition of the domains. However, as the observed changes may have been caused by ectoine or the fluorophore itself, in this study ToF-SIMS-based imaging of distribution patterns was used to determine the alterations.

Materials and Methods 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) was purchased from Avanti Polar Lipids Inc. (Alabaster, AL). Cholesteryl-3-palmitate (CP) and 1,3-dipalmitoyl-2-oleoylglycerol (DPOG) was purchased from Sigma-Aldrich (Steinheim, Germany). 2-(4, 4-difluoro-5-methyl-4-bora-3a, 4a-diaza-s-indacene-3-dodecanoyl)-1-hexadecanoyl-snglycero-3-phosphocholine (β-BODIPY®500/510C12-HPC, BODIPY-PC) was obtained from Molecular Probes (Eugene, OR). Ectoine ((S)-2-methyl-1, 4, 5, 6-tetrahydropyrimidine-4-carboxylic acid) was provided by Bitop AG (Witten, Germany). Computation of Mixture Ratios. For the calculation of the individual components of the mixtures, a recently introduced algorithm was used34, which requires a DA of the spectra of all pure substances to be carried out in the first step. The original data space consists of the intensity values of the peaks of the mass spectra used for the analysis. By introducing discriminant scores representing linear combinations of the original variables as new variables, the high-dimensional data space can be reduced to  − 1 dimensions, where  is the number of components of the mixture. Assuming that the discriminant scores of a mixture represent a linear combination of the discriminant scores of the pure substances, together with the standardization condition that the proportion of all components must be nonnegative and their sum must be 1, the percentages of all components can be calculated unambiguously by  ⋅  = . The  ×  matrix  contains in each row the mean values of the ( − 1) discriminant scores, which were determined from the pure substances of the k components, and in the last row only ones which originate from the standardization condition. The column vector  contains the  percentages of the mixture components and the column vector  ( − 1) discriminant scores of the mixture and in the last row a 1. Sample preparation. For the pure CP and DPOG preparations dissolved in a chloroform/methanol solution (1/1, v/v), spin coating was used to produce homogeneously thin lipid films on silicon wafers. DPPC and the mixtures were prepared as Langmuir-Blodgett (LB) films with an analytical Wilhelmy film balance (Rieger and Kirstein, Germany). Purified water at 20 °C with or without ectoine was used as subphase. The surface pressure was always set to 7 mN/m. Lipid mixtures consisted of DPPC and CP in a molar ratio of 7/3 and DPPC, DPOG and CP in a molar ratio of 4/3/3. The mixtures were spread in a chloroform/methanol solution onto the subphase without ectoine and the mixture of DPPC, DPOG and CP was additionally applied to a subphase with 100 mM of ectoine. After an equilibration time of 10-15 minutes, the monolayers were compressed at a rate of 2.9 cm2/min and transferred onto ozonized silicon wafers. Epi-fluorescence microscopy. The lipid mixtures were doped with 0.5 mol% BODIPY-PC and visualized with an epifluorescence microscope (Olympus STM5-MJS, Olympus, Germany). ToF-SIMS imaging. ToF-SIMS analyses were performed by means of a gridless reflectron-based ToF-SIMS 38 instrument. ToF-SIMS imaging was performed in burst alignment mode at high lateral resolution and unit mass + 2 resolution with Bi3 primary ions. For each wafer, 60 scans of 3 areas sized 150 × 150 µm each were taken with 512 × 512 pixels. For both the calculations and the ToF-SIMS images only the first 5 scans with primary ion dose densities of approximately 1011 cm-2 were used in order not to exceed the static limit. Higher ion dose densities not only lead to a destruction of the surface by fragmentation of the molecules but also to a removal of superficial layers. In this manner depth profiles were created by calculating images from the scans 1-5, 2-6, ..., 56-60 in dynamic ToF-SIMS mode with pulsed primary ion beam. All peaks were included in the peaklists except for those in which counts were less than 0.1% of the total ion count (TIC) and containing elements that could not be derived from lipids or ectoine (see Tables S1 and S2 in the Supporting Information). The signal intensities were normalized by dividing the signal counts by the corresponding total counts.

Results and Discussion Mass spectrometry. Since very high resolution ToF-SIMS images were to be produced, the ion source, a liquid metal ion gun was operated in burst alignment mode. In this mode small spot sizes of less than 250 nm are possible. An image size of 150 × 150 µm2 and a raster of 512 × 512 resulted in a pixel size of approximately 300 nm without

ACS Paragon Plus Environment

Page 2 of 12

Page 3 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

oversampling. However, this type of scanning resulted in very low signal intensities and only unit mass resolution. The TIC and the sum of counts of the peaks used for the calculation were 30.0 ± 8.0 or 18.2 ± 6.5 per pixel in negative ion mode and 53.6 ± 16.5 or 42.3 ± 12.7 in positive ion mode. This means that on average less than 17% of all 112 peaks selected for the analysis in negative ion mode and less than 26 % of the selected 164 peaks in positive ion mode contained signals at all. For the components, i. e. DPPC, DPOG, CP, and ectoine, characteristic signals were identified whose intensity, normalized to TIC, was at least 10 times higher than its intensity in the other three components. All characteristic signals are listed in Tables 1 and 2. The native ToF-SIMS images obtained by adding the corresponding characteristic signals are shown in Figures 1-3 for comparison with the epi-fluorescence microscopic images and the reconstructed ToF-SIMS images. The reconstruction was carried out as follows: First, a DA was first performed using the pure standards (7 - 14 images with 5122 pixels each). DA led to an almost complete correct classification of pure components, in positive ion mode (99.98%) even better than in negative ion mode (98.15%). The percentages of all four substances DPPC, DPOG, CP, and ectoine were calculated for each individual pixel as described. In the last step, the distribution pattern of each substance was reconstructed from the calculated percentages. In order to estimate the influence of primary ion bombardment on the degree of fragmentation and the calculation of component proportions, sputtering of pure lipids was investigated (see Figure S1 in the Supporting Information). The median of the calculated proportions of the components was between 90.0% and 95.3%, the mean value between 84.4% and 93.2% (see Tables 1 and 2), depending on substance and ion mode. The difference from the expected 100% was relatively large, since only one-sided deviations downwards were possible due to the standardization constraints, because of the very low signal intensity with correspondingly high variation of the spectral composition, and due to the low signal-to-noise ratio. During the sputtering process, a significant proportion of the characteristic fragments of the pure substances further fragmented (see Tables 1 and 2). The calculated fraction decreased significantly less despite the change in the composition of the spectra due to progressive fragmentation during depth profiling. The largest decline of approximately 1/3 of the initial value in negative ion mode and 1/5 in positive ion mode related to the DPOG's percentage, while the CP portion increased commensurately (see Figures S1 c) and d) in the Supporting Information). LB films of DPPC-CP. The DPPC-CP mixtures showed epi-fluorescence microscopically darker domains (see Figure 35 1). As these domains appeared at higher CP concentrations, it can be assumed that CP was enriched in them. Reconstructed ToF-SIMS distribution patterns revealed fine fibrils in the CP domains both in positive and negative ion mode, which were hinted at but not clearly resolved in the epi-fluorescent images. These fibrils were approximately 1.5 – 2.0 µm thick and often formed spoke-like structures. When the positive ion characteristic signals were added together, these structures were less clearly recognizable when compared to the reconstructed distribution patterns. The improvement obtained in the reconstructed image can be quantified using a contrast factor, , defined as

, =

 ,

where  is the mean signal intensity in the fibrils,  is the mean signal intensity in the background and , is the 39 pooled standard deviation . The contrast factor for the reconstructed CP images is 4.41 whereas the contrast obtained using the characteristic ions is 2.40. In the negative ion mode, the characteristic signals provided a contrast of only 0.88 compared to a contrast of 2.40 in the reconstructed image. The fibrils could also be identified as dark areas in the reconstructed DPPC images of both modes. To demonstrate the robustness of the algorithm we included DPOG and ectoine, although not present in this mixture, in the calculation of component proportions. The mean calculated proportion of DPOG and ectoine in positive ion mode was 4.8% and 1.2% respectively, in negative ion mode 6.4% and 2.5%. The incorrectly positively calculated DPPC and ectoine were very homogeneously distributed in both ion modes as would be expected for random noise. LB films of DPPC-DPOG-CP (subphase without ectoine). Epi-fluorescence microscopy revealed large, darker appearing domains that should correspond to LC phase monolayers (see Figure 2). In these areas DPOG was presumably on top of a lipid film consisting of DPPC and CP, since the LC phase is more hydrophobic than the LE phase. While in the native TOF-SIMS images, the distribution of the DPOG could be displayed neither in positive nor negative ion mode, the calculated distribution patterns showed DPOG-enriched domains that were the same in shape and size to the fluorescence microscopic LC domains. Since the yields for fatty acid signals in negative ion mode are 33,40 significantly higher than in positive ion mode , the DPOG-enriched regions were better represented in the images reconstructed from negative ion mode, although all lipids contain palmitate. Only DPOG contains in addition to 2 palmitates a unique fatty acid, oleic acid. In dynamic ToF-SIMS mode it was also apparent that the DPOG actually lies on the surface of the lipid monolayer (see Video S1 in the Supporting Information). The thickness of the DPOG film 35 was estimated to be less than 2 nm. In depth profiling (see Figure 4), the DPOG percentage decreased from approximately 47% to 5% after 50-55 scans, which corresponded to the expected background. Since the calculations were made from the summation spectra of 5 scans each, a total of 50 scans correspond to 10 calculated layers, which of course can overlap. I.e. with 5 scans considerably below 1 nm will be removed. The drop of DPOG was not solely

ACS Paragon Plus Environment

Langmuir

due to the reduction of the DPOG portion in the sputtering process described above, which explained a decline of the DPOG portion by about 1/3 of the initial value, i. e. by only 16% in absolute terms. In areas where the monolayer was in the LE phase and only covered by a thin DPOG layer, the CP percentage increased approximately linearly with increasing cumulative primary ion dose density compared to the DPPC (see Figure S2 in the Supporting Information). This increase could already be observed with the depth profiling of the pure CP (see Figure S1 f) in the Supporting Information) and was not based on a stronger fragmentation of the DPPC in the dynamic SIMS mode. If the percentage change in the dynamic SIMS was corrected accordingly, DPPC and CP remained constant after DPOG sputtering. For the ectoine which was not present in the sample, only a homogeneous background of about 2 - 3% on average was calculated.

Epi-fluorescence Microscopy

Calculated Percentages

Added ToF-SIMS Images

Positive Ion Mode

DPPC

CP

DPPC

Negative Ion Mode

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 12

CP

Figure 1. LB films of DPPC-CP: epi-fluorescence microscopic image, distribution patterns reconstructed from calculated percentages and added ToF-SIMS images from characteristic fragments of DPPC and CP. Scale bars of calculated percentages denote percentages, scale bars of added ToF-SIMS images denote ion counts per pixel. Scale bars are 10 µm.

ACS Paragon Plus Environment

Page 5 of 12

Epi-fluorescence Microscopy

Calculated Percentages

Added ToF-SIMS Images

Positive Ion Mode

DPOG

DPPC

CP

DPOG

Negative Ion Mode

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

DPPC

CP

Figure 2. LB films of DPPC-DPOG-CP without ectoine: epi-fluorescence microscopic image, distribution patterns reconstructed from calculated percentages and added ToF-SIMS images from characteristic fragments of DPPC, DPOG, and CP. Scale bars of calculated percentages denote percentages, scale bars of added ToF-SIMS images denote ion counts per pixel. Scale bars are 10 µm.

ACS Paragon Plus Environment

Langmuir

Positive Ion Mode

Negative Ion Mode

Calculated Percentages

DPOG

DPPC

CP

Ectoine

Added ToF-SIMS Images

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 12

CP

Ectoine

Figure 3. LB films of DPPC-DPOG-CP with ectoine: distribution patterns reconstructed from calculated percentages of DPPC, DPOG, and CP and added ToF-SIMS images from characteristic fragments of CP and ectoine. Scale bars of calculated percentages denote percentages, scale bars of added ToF-SIMS images denote ion counts per pixel. Scale bars are 5 µm.

ACS Paragon Plus Environment

Page 7 of 12

80 70

calculated percentage

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

60 50 DPPC DPOG CP

40 30 20 10 0 0

10

20

30

40

50

60

scan

Figure 4. Depth-dependent percentages of DPPC, DPOG, and CP in DPOG-enriched domains.

LB films of DPPC-DPOG-CP (subphase with ectoine). Both the epi-fluorescence microscopic images and the reconstructed images contained the DPOG enriched domains described above. Compared to the lipid layers without the addition of ectoine, they were on average 20% smaller. This is consistent with the observations of Dwivedi et al. describing a reduced size of the liquid-condensed domains due to increased areas occupied by the headgroups in the presence of ectoine [35]. In reconstructed ToF-SIMS images the most striking difference between layers which were prepared with and without ectoine consisted in ring-shaped conglomerates of CP (see Figure 3). Such structures were rarely and less clearly seen in layers in the absence of ectoine. The rings were often observed in the center of the DPOG enriched domains and were barely covered by DPOG. The wall thickness of the rings was approximately 1 - 2 µm, the outer diameter about 5 - 10 µm. Often two or more of these ring structures were placed together. The images calculated from the positive ion mode spectra showed the structures more clearly than those calculated from negative ion mode data. This was supported by the clear evidence for CP found in the native positive ion mode TOF-SIMS images. Here the structures described were visible in the images, which were created by pixelwise addition of the counts of the characteristic fragments of the CP. The significance of this phenomenon has to be elucidated. The calculated ectoine percentages were approximately 7 - 8% in positive ion mode and 12 - 13% in negative ion mode. Local enrichments were found in areas where the lipid film was not intact and around the CP conglomerates.

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 12

Table 1: Positive ion mode: signal intensities (counts/pixel) and the change in signal intensities during the sputter process of characteristic fragments in pure preparations (see text). m/q

DPPC

counts/pixel 30.03 ([C H4N]+ ) 58.07 ([C3H8N]+ ) 86.10 ([C5H12N]+ ) 133.10 ([C10H13]+ ) 143.08 ([C6H11N 2O2]+ ) 144.08 ([C6 H12N 2O2]+ ) 147.11 ([C11H15]+ ) 161.13 ([C12H17]+ ) 165.06 ([C6H10N 2O2Na]+ ) 184.07 ([C5H15 PNO4 ]+ ) 224.11 ([C8H19PNO4]+ ) 307.14 ([C12H20N4O4Na]+ ) 313.31 368.34 369.35 370.36 371.37 384.34 467.29 549.56

([C20H41O2]+ ) ([C27H44]+ ) ([C27H45]+ ) ([C27H46]+ ) ([C27H47]+ ) ([C27H44O]+ ) ([C27H47O6]+ ) ([C35H65O4]+ )

551.40 552.56 575.53 577.54

([C35H67O4]+ ) ([C35H68O4]+ ) ([C37H67O4]+ ) ([C37H69O4]+ )

578.59 ([C37H70O4]+ ) sum

DPOG

change

counts/pixel

CP

change

counts/pixel

Ectoine

change

counts/pixel 1.00

2.87 2.62

+ 9.1% - 49.4% 0.55

- 25.0% 4.69 1.55

0.93 0.30

- 57.6% - 40.0% 0.97

2.41 0.30

- 50.0% - 42.9% 0.12 0.33

8.20

total ion count

30.75

calculated percentage mean median

93.2% 95.3%

- 29.9%

- 67.4%

0.21 0.12 0.97 0.33 0.20 1.25

-

0.45

- 81.2%

3.86

- 78.9%

- 74.3% - 79.7% - 80.0%

0.07 0.04

- 75.0% - 57.1%

3.86

- 65.1%

90.0% 62.5% 81.3% 82.0% 63.0% 80.0%

62.91 -13.2%

0.47 1.14 0.36

66.93

91.6% 94.6%

-19.2% -19.2%

8.33 27.84

91.7% 94.9%

-9.6% -9.6%

Table 2: Negative ion mode: signal intensities (counts/pixel) and the change in signal intensities during the sputter process of characteristic fragments in pure preparations (see text). m/q

DPPC

counts/pixel 40.00 ([C2O]− ) 62.97 ([PO2]− ) 78.96 ([PO3 ]− ) 141.04 ([C6H9N 2 O2]− ) 152.04 ([C3H7PNO4]− ) 153.04 281.12 282.31 383.21 624.13 673.85 737.74 sum

([C3H8PNO4]− ) ([C18H33O2]− ) ([C18H34O2]− ) ([C27H43O]− ) ([C43H75O2]− ) ([C38H76PNO6]− ) ([C46H89O6]− )

total ion count calculated percentage mean median

DPOG

change

counts/pixel

CP

change

counts/pixel

Ectoine

change

counts/pixel 0.82

2.49 4.71

+ 11.0% - 4.8%

0.19

- 83.3%

0.36

- 70.0%

3.53

1.09 0.24

0.02

- 35.5%

7.77 34.55

- 4.7%

87.8% 90.0%

-5.8%

- 43.8% - 43.1%

0.05 1.37 31.37

- 91.7% - 45.2%

90.4% 91.4%

-34.1%

0.04

- 83.3%

0.03

- 90.0%

0.06 21.80

- 86.3%

88.2% 93.1%

-9.6%

4.34 28.63

Conclusions High lateral resolution ToF-SIMS imaging of single compounds in mixtures using characteristic signals is either not possible or only possible to a limited extent due to the generally low number of counts, even in model systems. In spite of the very low signal intensity and only unit mass resolution, the employed algorithm allows quantitative imaging of all available components with lateral resolution in the sub-micrometer range when all possible constituents are known.

ACS Paragon Plus Environment

Page 9 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Usually, the dynamic ToF-SIMS mode with pulsed primary ion beam is not applied to biological samples because of the loss of molecular information due to increased fragmentation.32,33 The algorithm generally weights more frequent fragments, i. e. usually small fragments, more heavily than less frequent ones when calculating the proportions of ingredients in mixtures. As a result, the calculated values are relatively insensitive to molecular fragmentation caused by accumulated ion beam damage, which is visible in the reduction of characteristic fragments. Therefore, it is no longer necessarily to not exceed the static limit, up to which there is no significant molecular fragmentation in the surface area due to primary ion bombardment. This makes it possible to determine the threedimensional distribution of components of organic model systems with depth resolution in the sub-nanometer range while the resolution of depth profiling with argon clusters is in the range of several nanometers which is not significantly below the thickness of these lipid layers. Therefore, a sequential removal of the membrane with argon cluster sputtering would not be able to detect changes in the film with depth. By using more scans for the calculation, the intensities of the summation spectra can be increased at the expense of depth resolution. Therefore, it seems likely that more complex systems can also be investigated, especially if sophisticated methods like delayed extraction are used. However, in a preliminary study in which we compared delayed extraction and burst alignment mode, no significant differences in the calculated percentages of up to 4 phospholipids were found. Here the benefit of increased mass resolution of delayed extraction compared to burst alignment mode might be cancelled out by the disadvantage of a reduced signal-to-noise ratio. To illustrate the robustness of the algorithm we used burst alignment mode. Another advantage of using small fragments is that the matrix effect could be less severe, as it appears to be more 41 pronounced with larger molecules. The calculated distribution patterns of the model system used, which simulates tear fluid lipid layers, correlate well with patterns observed in epi-fluorescence microscopy. Our data confirm that structural changes which Dwivedi et al. [35] attributed to accumulation of ectoine in the subphase, are in fact due to ectoine and are not artifacts of the fluorescent label. In contrast to epi-fluorescence microscopy, the distribution of all components, including substances that cannot be represented by epi-fluorescence microscopy like ectoine, can be depicted separately. Using the possibility of three-dimensional imaging, it can be determined directly whether a lower fluorescence is caused by a correspondingly reduced concentration of the fluorescent dye or a covering by a layer of other substances. Since the algorithm is based on mass spectrometry, the imaging is free of markers and thus prevents fluorophores from affecting the composition and distribution of the components. All visible structures are displayed in positive and negative ion mode. For substances not present in the mixtures, only low values are calculated which impose like a low, homogeneous noise. The use of the applied algorithm for quantifying and representing the distribution of model systems is not limited to ToF-SIMS of artificial membranes, but should be useful for other surfaces as well as other multispectral images.

Supporting Information Ion peak lists, depth profiles of pure substances, depth profile of lipids in LE domains, and video of changes in the percentages of all lipids in successive depths.

Acknowledgement We would like to thank Richard Peterson for his helpful suggestions and editing the manuscript.

References 1 Singer, S. J.; Nicolson, G. L. The fluid mosaic model of the structure of cell membranes. Science. 1972, 175, 720-731. 2 Yeagle, P. L. Cholesterol and the cell membrane. BBA – Reviews on Biomembranes. 1985, 822, 267 -287. 3 Leventis, P. A.; Grinstein, S. The distribution and function of phosphatidylserine in cellular membranes. Annu. Rev. Biophys. 2010, 39,407-427. 4 Lingwood, D.; Simons, K. Lipid rafts as a membrane-organizing principle. Science. 2010, 327, 46-50. 5 Jacobson, K.; Mouritsen, O. G.; Anderson, R. G. W. Lipid rafts: at a crossroad between cell biology and physics. Nat. Cell Biol. 2007, 9, 7-14,. 6 Winter, R.; Jeworrek, C. Effect of pressure on membranes. Soft Matter. 2009, 5, 3157-3173. 7 Dietrich, C.; Bagatolli, L. A.; Volovyk, Z. N.; Thompson, N. L.; Levi, M.; Jacobson, K.; Gratton, E. Lipid rafts reconstructed in model membranes. Biophys. J. 2001, 80, 1417-1428. 8 Bourdos, N.; Kollmer, F.; Benninghoven, A.; Sieber, M.; Galla, H.-J. Imaging of domain structures in a onecomponent lipid monolayer by time-of-flight secondary ion mass spectrometry. Langmuir 2000, 16, 1481-1484. 9 Breitenstein, D.; Batenburg, J. J.; Hagenhoff, B.; Galla, H.-J. Lipid specificity of surfactant protein B studied by time-of-flight secondary ion mass spectrometry. Biophys. J. 2006, 91, 1347-1356.

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

10 Bourdos, N.; Kollmer, F.; Benninghoven, A.; Ross, M.; Sieber, M.; Galla, H.-J. Analysis of lung surfactant model systems with time-of-flight secondary mass spectrometry. Biophys. J. 2000, 79, 357-369. 11 Ross, M.; Seinem, C.; Galla, H.-J.; Janshoff, A. Visualization of chemical and physical properties of calcium-induced domains in DPPC/DPPS Langmuir-Blodgett layers. Langmuir 2001, 17, 2437-2445. 12 Ohvo-Rekilä, H.; Ramstedt, B.; Leppimäki, P.; Slotte, J. P. Cholesterol interactions with phospholipids in membranes. Prog. Lipid Res. 2002, 41, 66-97. 13 Kooijman, E. E.; King, K. E.; Gangoda, M.; Gericke, A. Ionization properties of phosphatidylinositol polyphosphates in mixed model membranes. Biochemistry 2009, 48, 9360-9371. 14 Murakami, Y.; Zhang, Z.; Taniguchi, T.; Sohgawa, M.; Yamashita, K.; Noda, M. High-sensitive detection of several tens of nM of amyloid-β by cantilever-type biosensor immobilized DPPC liposome incorporated with cholesterol. Proc. Eng. 2016, 168, 565-568. 15 Jiao, T.; Leca-Bouvier, B. D.; Boullanger, P.; Blum, L. J.; Girard-Egrot, A. P. A chemiluminescent Langmuir– Blodgett membrane as the sensing layer for the reagentless monitoring of an immobilized enzyme activity. Colloids Surf. A. 2010, 354, 284-290. 16 Zhang, Z.; Sohgawa, M.; Yamashita, K.; Noda, M. Real-time characterization of fibrillization process of amyloid-beta on phospholipid membrane using a new label-free detection technique based on a cantilever-based liposome biosensor. Sens. Actuators B: Chem. 2016, 236, 893-899. 17 Nikolelis, D. P.; Mitrokotsa, M. Stabilized lipid film based biosensor for atenolol. Biosens. Bioelectron. 2002, 17, 565-572. 18 Nikoleli, G.-P.; Nikolelis, D. P.; Tzamtzis, N.; Psaroudakis, N. A selective immunosensor for D-dimer based on antibody immobilized on a graphene electrode with incorporated lipid films. Electroanalysis 2014, 26, 1522-1527. 19 Luna, D. M. N.; Oliveira, M. D. L.; Nogueira, M. L.; Andrade, C. A. Biosensor based on lectin and lipid membranes for detection of serum glycoproteins in infected patients with dengue. Chem. Phys. Lipids 2014, 180, 7-14. 20 Peetla, C.; Stine, A.; Labhasetwar, V. Biophysical interactions with model lipid membranes: Applications In drug discovery and drug delivery. Mol. Pharm. 2009, 6, 1264-1276. 21 Pavinatto, F.J.; Caseli, L.; Pavinatto, A.; dos Santos, D. S., Jr.; Nobre, T. M.; Zaniquelli, M. E. D.; Silva, H. S.; Miranda, P. B.; de Oliveira, O. N., Jr. Probing chitosan and phospholipid interactions using Langmuir and Langmuir-Blodgett films as cell membrane models. Langmuir 2007, 23, 7666-7671. 22 Bensikaddour, H.; Snoussi, K.; Lins, L.; Van Bambeke, F.; Tulkens, P. M.; Brasseur, R.; Goormaghtigh, E.; Mingeot-Leclercq, M.-P. Interactions of ciprofloxacin with DPPC and DPPG: Fluorescence anisotropy, ATR-FTIR and 31 P NMR spectroscopies and conformational analysis. Biochim. Biophys. Acta 2008 1778, 2535-2543. 23 Nandi, N.; Vollhardt, D. Effect of molecular chirality on the morphology of biomimetic Langmuir monolayers. Chem. Rev. 2003, 103(10), 4033-4075. 24 Rossi, C.; Chopineau, J. Biomimetic tethered lipid membranes designed for membrane-protein interaction studies. Eur. Biophys. J. 2007, 36(8), 955-965. 25 Stottrup, B. L.; Nguyen, A. H.; Tüzel, E. Taking another look with fluorescence microscopy: Image processing techniques in Langmuir monolayers for the twenty-first century. Biochim. Biophys. Acta 2010, 1798(7), 1289-1300. 26 Cruz, A.; Vázquez, L.; Vélez, M.; Pérez-Gil, J. Phospholipid membranes: Dipalmitoylphosphatidylcholine interfacial monolayers. Langmuir. 2005, 21, 5349-5355. 27 Benninghoven, A.; Rudenauer, F. G.; Werner, H. W. Secondary ion mass spectrometry. Basic concepts, instrumental aspects, applications and trends. Wiley: New York, 1987. 28 Gunnarsson, A.; Kollmer, F.; Sohn, S.; Höök, F.; Sjövall, P. Spatial-resolution limits in mass spectrometry imaging of supported lipid bilayers and individual lipid vesicles. Anal. Chem. 2010, 82, 2426-2433. 29 Braun, R. M.; Blenkinsopp, P.; Mullock, S. J.; Corlett, C. Willey, K. F.; Vickerman, J. C.; Winograd, N. Performance characteristics of a chemical imaging time-of-flight mass spectrometer. Rapid Commun. Mass Spectrom. 1998, 12, 1246-1252. 30 Touboul, D.; Laprévote, O.; Brunelle, A. Micrometric molecular histology of lipids by mass spectrometry Imaging. Curr. Opin. Chem. Biol. 2011, 15, 725 -732. 31 Draude, F.; Pelster, A.; Körsgen, M.; Kassenböhmer, R.; Schwerdtle, T.; Müthing, J.; Arlinghaus, H. F. ToFSIMS imaging of plasma membrane lipids with sub-micrometer resolution. Surf. Interface Anal. 2014, 46 (S1), 127-130. 32 Boxer, S. G.; Kraft, M. L.; Weber, P. K. Advances in imaging secondary ion mass spectrometry for biological samples. Annu. Rev. Biophys. 2009, 38, 53-74. 33 Passarelli, M. K.; Winograd, N. Lipid imaging with time-of-flight secondary ion mass spectrometry (ToFSIMS). Biochim. Biophys. Acta 2011, 1811, 976-990. 34 Kassenböhmer, R.; Draude, F.; Körsgen, M.; Pelster, A.; Arlinghaus, H. F. Calculation of membrane lipid ratios using single-pixel time-of-flight secondary ion mass spectrometry analysis. Anal. Chem. 2015, 87(15), 7795-7802. 35 Dwivedi, M.; Brinkkötter, M.; Harishchandra, R. K.; Galla, H.-J. Biophysical investigation of the structure and function of the tear fluid lipid layers and the effect of ectoine. Part B: Artificial lipid films. Biochim. Biophys. Acta 2014, 1838, 2716-2727. 36 Vollhardt, D.; Fainerman, V. B. Progress in characterization of Langmuir monolayers by consideration of compressibility. Adv. Colloid Interface Sci. 2006, 127, 83-97. 37 Tokumasu, F.; Hwang, J.; Dvorak, J. A. Heterogeneous molecular distribution in supported multicomonent lipid bilayers. Langmuir 2004, 20, 614-618. 38 Möller, J.; Beumer, A.; Lipinsky, D.; Arlinghaus, H. F. Introduction of a cryosectioning-TOF-SIMS instrument for analysis of non-dehydrated biological samples. Appl. Surf. Sci. 2006, 252, 6709-6711. 39 Tyler, B. J.; Rayal, G.; Castner, D. G. Multivariate analysis strategies for processing ToF-SIMS images of biomaterials. Biomaterials 2007, 28, 2412-2423.

ACS Paragon Plus Environment

Page 10 of 12

Page 11 of 12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

40 Brulet, M.; Seyer, A.; Edelman, A.; Brunelle, A.; Fritsch, J.; Ollero, M.; Laprévote, O. Lipid mapping of colonic mucosa by cluster TOF-SIMS imaging and multivariate analysis in cftr knockout mice. J. Lipid Res. 51(10), 3034-3045 (2010) 41 Shard, A. G.; Spencer, S. J.; Smith, S. A.; Havelund, R.; Gilmore, I. S. The matrix effect in organic secondary ion mass spectrometry. Int. J. Mass Spectrom. 2015, 377, 599-609.

ACS Paragon Plus Environment

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

TOC graphic

ACS Paragon Plus Environment

Page 12 of 12