Nanomechanics of phospholipid LB film studied layer by layer with AFM
© Li et al.; licensee Springer. 2014
Received: 30 April 2014
Accepted: 11 November 2014
Published: 10 December 2014
Phospholipid, a main component of cell membrane, has been explored as a model system of the cell membrane and temporary scaffold materials in recent studies. The mechanical properties of phospholipid layers are essentially interesting since it is involved in several biological processes.
Here, the nanomechanical properties such as indentation force, adhesion force and DMT (Derjaguin-Müller-Toporov) modulus of 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC) Langmuir-Blodgett (LB) films were analyzed layer by layer with Atomic Force Microscope (AFM) under deionized water condition.
KeywordsPhospholipid Nanomechanics Langmuir-Blogget film Atomic force microscope Indentation force curve
Membranes are essential for life as they are the only means by which biochemical reactions of closed compartments of a living cell can be shielded from the environment . Monolayer and bilayers of lipid not only constitute the matrix of the cell membrane , but also could be used as models of the cell membrane , which makes it convenient to study these models in vitro. Phospholipid has been used as temporary scaffold materials to provide mechanical support for the vessel wall , and their two-dimensional assemblies of bilayers have been explored as biomimetic analogues of cell membranes .
The biological relevant function of lipid layer has aroused interest in exploring its mechanical properties because these mechanical properties govern several biological processes , e.g., osmotic shrinkage and swelling of cells . Interacting with the first lipid layer and further permeating through the second layer of cell bilayer membrane, substances can execute biological functions such as exchanging of nutrients, ions , and signalling  between cells and external environment. Therefore, the mechanics of both the first lipid layer and the second layer are closely related to the biological functions of cells. Nanomechanical properties have been explored by Atomic Force Microscope (AFM), micropipet or other techniques in phospholipid microbubbles , vesicles  and living cells . Recently, studies on nanomechanical properties of phospholipid layers have been focused by experimental and numerical analysis . The environmental effects on the nanomechanics of lipid layers caused by ions , temperature , proteins , molecular chain unsaturation and headgroup type  have been explored in detail with the method of AFM or single molecule force spectroscopy (SMFS) . Moreover, the recent developments in AFM monitoring bilayer alteration and remodeling have been reviewed by Sandrine M. and his coworkers  due to the significance of the lipid layers in modeling of biomimetic systems.
Although there are some literatures about the nanomechanical properties of several phospholipids in salt solution reported, the indentation force curve of monolayer, upper layer or lower layer of bilayer phospholipid prepared on substrate has not been reported yet. The reason might be that the nanomechanics of lipid layers are difficult to be obtained , especially in low ionic strength concentrations ,. Moreover, the properties of the upper layer and lower layer of the bilayer would be different under the influence of ions or substrate, but they have not been analyzed to our knowledge.
In order to address this issue, DSPC was selected in our experiment. DSPC has been explored extensively since it is one of the main components of cell membranes and in solid state at room temperature. Fresh monolayer and bilayer LB films of DSPC were prepared, and their nanomechanical properties were measured layer by layer using AFM probes with smaller spring constant parameter (0.02 N/m) in deionized water. The indentation force curves of DSPC monolayer, upper layer and lower layer LB films were obtained for the first time. The analysis shows that the obtained penetration distances of the indentation force curves are equal to the thickness values of DSPC LB films, while the DMT modulus  of DSPC upper layer is different with that of monolayer films.
1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC) and 1,2-dilauroyl-sn-glycero-3- -phosphocholine (DLPC) were purchased from Avanti polar lipids Inc. (Alabaster, AL). Chloroform was purchased from Tianjin Kermel Chemical Company Ltd. Fresh ultrapure deionized water with resistivity of 18.23 MΩ · cm was produced using GT-8L ion exchange system. All substrates used in the experiment were mica with freshly cleaved surface (Ted Pella, Inc.).
LB film preparation
Unique deposition technique such as LB deposition is generally used to obtain films with controlled packing density of molecules over a wide range, which is not easily accomplished with chemisorbed or other assembling methods . In our experiment, the assembled monolayers and bilayers of phospholipid films were prepared in a computer controlled LB trough (μ trough, Kibron Co., Helsinki, Finland)  which monitors surface pressure using a Whilhelmy wire. 25 μL 0.5 M phospholipid chloroform solutions were applied onto the surface of deionized water subphase (about 70 mL) with a microsyringe. After 20 minutes, the phospholipid film was compressed with two barriers at 3 cm2/min compression rate until surface pressure reached 45 mN/m (compact LB films could be obtained at this surface pressure though it is smaller than the collapse surface pressure of the used phospholipid), and then the two barriers were released to the sides of the trough at the speed of 5 cm2/min. After 3 times of compression, monolayer phospholipid film was deposited onto a newly cleaved mica substrate (1cm × 1 cm, previously immerged into water subphase) with a lifting rate of 1.5 mm/min. The second layer of phospholipid was transferred onto the monolayer film with a dipping rate of 1.5 mm/min when the monolayer film was dried 10 minutes in air. The compensation rate of 3 cm2/min was used to keep a constant pressure of the interface during LB film deposition process by the aid of a Whilhelmy wire to monitor the sample pressure. This compensation speed of 3 cm2/min is merely a maximum speed, which could be auto-tuned to adapt to the sample raising speed in order to keep the sample pressure constant. The configuration of the prepared phospholipid bilayer films is similar to that of cell membrane in nature . In the end, the phospholipid bilayer was brought out of the subphase with the speed of 10 mm/min, and AFM measurements were carried out right after the film preparations.
AFM characterizations were carried out in contact mode under deionized water using an Agilent AFM/STM 5500 microscope with a Multi-purpose small scanner with low coherence laser (N9520A, Agilent Technologies, Inc. USA). The cantilever was a 100 μm long and 106 μm wide silicon nitride triangular (OMCL-TR400PSA from Olympus, Japan), with a nominal spring constant of 0.02 N/m and tip radius less than 20 nm. The imaging frequency was fixed to 1.0 Hz, and the resolution of images was 512 × 512 pixels per image. Monolayer and bilayer films were confirmed with the Agilent AFM/STM 5500 microscopy at first, and then the indentation force curves were recorded at different sites of the phospholipid LB films without defect. The sweep duration of AFM tip approaching and retracting, and pre-sweep delay time were all set as 1.0 second. The adhesion and DMT modulus of the samples were studied with another AFM of MultiMode® 8 from Bruker AXS. All recorded images and indentation force curves were analysed using the commercial Scanning Probe Image Processor (SPIP™) software (image metrology ApS, version 5.13, Lyngby, Denmark). All indentation force curves were processed with force curve analysis module, and Hertz sphere-on-flat model was chosen to fit the force curves. The Poisson’s ratio was assumed to be 0.5 (highly elastic), which is in agreement with related studies.
Results and discussion
Monolayer and bilayer of phospholipid films
Monolayer and bilayer phospholipid films were prepared with LB technique since it is convenient to prepare films with controllable layers, which is different from the experiments where films were prepared by dropping phospholipid solutions onto substrates , or absorbed on AFM probe tips . The monolayer and bilayer phospholipid films were then used to analyze their mechanical properties layer by layer in this paper.
In order to validate the indentation force curves of DSPC monolayer LB film, monolayer film of another phospholipid (DLPC) was researched with the same method as that of DSPC monolayer film in both sample preparation and sample characterization. DLPC has the same basic molecular structure as that of DSPC. DLPC has 12 repeat carbon units (-CH2) while DSPC has 18 repeat carbon units , and then the length of DLPC molecule is shorter than that of DSPC molecule. Then the layers of the two phospholipids were expected to be of distinct mechanical properties at nano (or pico) Newton level.
It will be found that the penetration distances of phospholipid films are smaller than that of films’ theoretical thickness, which is mainly due to the deformation of the phospholipid molecules when they are interacting with AFM tip during the penetrating process in our experiment. Moreover, the deformation of DSPC film (in ordered-phase) would be different from that of DLPC film (in fluid-phase) under experimental temperature of 28°C when their phase transition temperature are considered (phase transition temperature of DSPC is 57.5 ± 0.5°C while that of DLPC is -1°C ,), which could be helpful in understanding the experiment result that no obvious penetration distance difference were obtained between DSPC and DLPC.
The penetration distances measured both at the first yield threshold and the following one, which represent the thicknesses of the upper and lower layers, are close to the width of jump (wm) in the case of DSPC monolayer. This result is reasonable since each of the penetration distance demonstrates the thickness of a single layer of DSPC LB film. The value of wu plus wd is about 4.64 nm, the total penetration distance of DSPC bilayer, which is the detected thickness of DSPC bilayer in our experiment. This detected thickness of DSPC bilayer is not only in agreement with the thickness of DSPC bilayer measured by modified AFM tip (4.4 nm) , but also consistent with the result measured by X-ray (4.7 nm) . Moreover, the D-value (fd-fu) of the break force between the lower and upper layers is about 0.41 nN, which is almost equal to the break through forces of upper layer (fu: 0.38 ± 0.28 nN) of DSPC LB films. This result means that the lower layer is as strong as that of the upper layer to be penetrated even if it is not expected since there is hydrophobic surface upside for lower layer DSPC LB film, while there is hydrophilic surface upside for upper layer. The possible reasons for that might be the water  and the hydration layer (formed between the mica surface and the hydrophilic surface of lipid films)  can affect the penetration forces.
The yield threshold force of DSPC bilayer (lower layer covered with upper layer) obtained in this work (fd =0.79 ± 0.39 nN, this force was used to represent the force needed to penetrate bilayer DSPC LB film since it was the force used to penetrate the lower layer of bilayer DSPC LB film after the AFM tip penetrated the upper layer of bilayer DSPC LB film.) is smaller than that of similar phospholipid film (1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) or 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), Figure Eight A in reference ) at the same temperature of 28°C, which could be explained by the effect of ions. The effects of ions, molecular structure and temperature on the nanomechanics of DLPC, DMPC and DPPC bilayers have been studied by Sergi and his coworkers ,, where all solutions used were set at pH =7.4 with 10 mM Hepes/NaOH. The basic molecular structures of DLPC, DMPC and DPPC  are the same as that of DSPC, which only differ in the number of –CH2 groups in the fatty acid chains. The number of –CH2 groups are 12, 14, 16 and 18 for DLPC, DMPC, DPPC and DSPC respectively. The yield threshold forces of DMPC are about 2 nN (Figure Eight A in reference ) in buffered water solution (fl) at 30°C, which is much smaller than that in buffered high ionic strength solution (about 4.7nN (fh), fl is equal to 0.43 times of fh). The same phenomenon was also found in the case of DPPC and DLPC bilayers at different temperatures by Sergi et. al., which explores the significant impact of ions on the yield threshold force of lipid film. This conclusion could be used to explain the small yield threshold force of DSPC bilayer (lower layer covered with upper layer) obtained in our work (0.79 ± 0.39 nN) compared with that of DMPC or DPPC at the same temperature (Figure Eight A in reference ) considering that our experiment was conducted under the deionized water condition.
Phenomenon should be pointed out that the slopes of the indention force curves changed very little after AFM tip having penetrated phospholipid films both in our experiment (Figures 2, 3 and 4) and in references ,. The major reason for that could be that AFM tips with different spring constant cantilevers fit for samples with different elastic modulus ranges, and softer cantilevers suit for lower elastic modulus samples . It has been reported that the sensitivities of AFM probe cantilevers can be affected by sample hardness , and then, the cantilever with 0.02 N/m spring constant were selected for phospholipid films imaging and not suitable for measurement of stiffer mica in our measurement.
Mechanical property of phospholipid film at different layers
According to Figure 5C, DMT modulus value of monolayer is calculated to be 68 ± 10 MPa (Em), and that of upper layer is 48 ± 12 MPa (Eu), which are in the same order of magnitude as the Young’s modulus of DOPC/DPPC (1:1, mol/mol) membranes (19.3 MPa and 28.1 MPa for the liquid and gel phases respectively) . The higher DMT modulus of DSPC LB films in our experiment, compared with the Young’s modulus of DOPC/DPPC films in liquid or gel phases, are reasonable since DSPC LB films are in solid phase under experimental temperature of 28°C (phase transition temperature of DSPC is 57.5 ± 0.5°C ).
The difference of DMT modulus between DSPC upper layer and monolayer is about 20 MPa, which could be resulted from their different substrates, since they are made of the same kind of molecules. This result indicates that the Young’s modulus of monolayer phospholipid film on soft lower layer or mica (with young’s modulus of 65.7 GPa ) increased with the hardness of substrate, which agrees with literature . Moreover, the adhesion property of the upper layer is almost equal to that of the monolayer analyzed from Figure 5B. These observations imply that the property of the upper layer is similar to that of the monolayer, but their property is not completely the same under the influence of their different substrates. The different property of bilayer and monolayer in our experiment indicates that the nature of DSPC LB film is closer to that of the upper layer, and this is why we focus on them. Especially, ions or other reagent in solvent used in experiment will have an influence on the property of the lipid layers, and the influence on the bilayer might be different from that on the monolayer since the cations or small molecules need more energy to pass through the bilayer than that of the monolayer.
In our experiment, the assembled monolayer and bilayer films composed by DSPC phospholipid molecules were prepared with the assistance of LB technique under the controllable conditions. Then, their nanomechanics properties were characterized layer by layer with AFM in contact mode under deionized water condition for the first time. The nanomechanics of DSPC monolayer, upper layer and lower layer of bilayer were analyzed and compared. It is found that though the adhesion of upper layer phospholipid is consistent with that of monolayer made from the same material, the DMT modulus of upper layer is closer to the nature of DSPC LB film. All obtained data of the penetration distances for DSPC represent the thicknesses of the lipid films as expected, and the mean yield force of DSPC LB bilayer in deionized water obtained in this work is smaller than that of similar lipid bilayers such as DMPC and DPPC in buffered water in the previous work due to the effect of ions. The success of exploring the nanomechanics of the phospholipid in deionized water is helpful to analyze the influence factors (cations or other chemicals) on the phospholipid membranes.
The authors acknowledge financial support from the National Natural Science Foundation of China (No. U1304310 and 21373077), Henan Province Department of Science and Technology Fund (No.122102310096 and 142102210561), and Henan Natural Science Research Office of Education project (No12A180005. and No. 2010A140001). We thank Mr. Kai Liu for force data analysis.
- Steinem C, Janshoff A: Multicomponent membranes on solid substrates: Interfaces for protein binding. Curr Opin Colloid Interface Sci. 2010, 15 (6): 479-488. 10.1016/j.cocis.2010.06.004.View ArticleGoogle Scholar
- Erasmo OG, José JTH, Armando A, Iván OB: Simultaneous determination of the elastic properties of the lipid bilayer by atomic force microscopy: bending, tension, and adhesion. J Phys Chem B. 2011, 115 (16): 4826-4833. 10.1021/jp111985z.View ArticleGoogle Scholar
- Lundquist A, Hansen SB, Nordström H, Danielson UH, Edwards K: Biotinylated lipid bilayer disks as model membranes for biosensor analyses. Anal Biochem. 2010, 405 (2): 153-159. 10.1016/j.ab.2010.06.030.View ArticleGoogle Scholar
- Il Kim H, Takai M, Ishihara K: Bioabsorbable material-containing phosphorylcholine group-rich surfaces for temporary scaffolding of the vessel wall. Tissue Eng Part C-Methods. 2009, 15 (2): 125-133. 10.1089/ten.tec.2008.0307.View ArticleGoogle Scholar
- Goksu EI, Nellis BA, Lin WC, Satcher JH, Groves JJT, Risbud SH, Longo ML: Effect of support corrugation on silica xerogel − supported phase-separated lipid bilayers. Langmuir. 2009, 25 (6): 3713-3717. 10.1021/la803851b.View ArticleGoogle Scholar
- McMahon HT, Gallop JL: Membrane curvature and mechanisms of dynamic cell membrane remodelling. Nature. 2005, 438 (7068): 590-596. 10.1038/nature04396.View ArticleGoogle Scholar
- Dieluweit S, Csiszár A, Rubner W, Fleischhauer J, Houben S, Merkel R: Mechanical properties of bare and protein-coated giant unilamellar phospholipid vesicles. A comparative study of micropipet aspiration and atomic force microscopy. Langmuir. 2010, 26 (13): 11041-11049. 10.1021/la1005242.View ArticleGoogle Scholar
- Lodish H, Berk A, Zipursky SL, Matsudaira P, Baltimore D, Darnell J: Molecular Cell Biology. W H Freeman & Co (Sd); 1995.Google Scholar
- Heldin CH, Miyazono K, Ten Dijke P: TGF- Signalling from Cell Membrane to Nucleus through SMAD Proteins. 1997, NATURE, LONDONGoogle Scholar
- Santos EB, Morris JK, Glynos E, Sboros V, Koutsos V: Nanomechanical properties of phospholipid microbubbles. Langmuir. 2012, 28 (13): 5753-5760. 10.1021/la204801u.View ArticleGoogle Scholar
- Sirghi L: Atomic Force Microscopy Indentation of Living Cells. Formatex. 2010Google Scholar
- Franz V, Loi S, Müller H, Bamberg E, Butt HJ: Tip penetration through lipid bilayers in atomic force microscopy. Colloids Surf B: Biointerfaces. 2002, 23 (2–3): 191-200. 10.1016/S0927-7765(01)00254-5.View ArticleGoogle Scholar
- Garcia-Manyes S, Oncins G, Sanz F: Effect of ion-binding and chemical phospholipid structure on the nanomechanics of lipid bilayers studied by force spectroscopy. Biophys J. 2005, 89 (3): 1812-1826. 10.1529/biophysj.105.064030.View ArticleGoogle Scholar
- Garcia-Manyes S, Oncins G, Sanz F: Effect of temperature on the nanomechanics of lipid bilayers studied by force spectroscopy. Biophys J. 2005, 89 (6): 4261-4274. 10.1529/biophysj.105.065581.View ArticleGoogle Scholar
- Pera I, Stark R, Kappl M, Butt H, Benfenati F: Using the atomic force microscope to study the interaction between two solid supported lipid bilayers and the influence of synapsin I. Biophys J. 2004, 87 (4): 2446-2455. 10.1529/biophysj.104.044214.View ArticleGoogle Scholar
- Zhao L, Feng S: Effects of lipid chain unsaturation and headgroup type on molecular interactions between paclitaxel and phospholipid within model biomembrane. J Colloid Interface Sci. 2005, 285 (1): 326-335. 10.1016/j.jcis.2004.11.032.View ArticleGoogle Scholar
- Dong M, Sahin O: A nanomechanical interface to rapid single-molecule interactions. Nat Commun. 2011, 2: 247-10.1038/ncomms1246.View ArticleGoogle Scholar
- Morandat S, Azouzi S, Beauvais E, Mastouri A, Kirat KE: Atomic force microscopy of model lipid membranes. Anal Bioanal Chem. 2013, 405 (5): 1445-1461. 10.1007/s00216-012-6383-y.View ArticleGoogle Scholar
- Strulson MK, Maurer JA: Microcontact printing for creation of patterned lipid bilayers on tetraethylene glycol self-assembled monolayers. Langmuir. 2011, 27 (19): 12052-12057. 10.1021/la201839w.View ArticleGoogle Scholar
- Goertz MP, Moore NW: Mechanics of soft interfaces studied with displacement-controlled scanning force microscopy. Prog Surf Sci. 2010, 85 (9–12): 347-397. 10.1016/j.progsurf.2010.07.003.View ArticleGoogle Scholar
- Roberts GG: Langmuir-blodgett Films. 1990, Plenum press, New YorkView ArticleGoogle Scholar
- Ren W, Li YL, Chen ML, Liu B, Li X, Dong MD: Oriented growth of single NaCl (100) crystal induced by Langmuir–Blodgett film. J Mater Res. 2011, 26 (02): 230-235. 10.1557/jmr.2010.26.View ArticleGoogle Scholar
- Schneider J, Dufrêne YF, Barger JWR, Lee GU: Atomic force microscope image contrast mechanisms on supported lipid bilayers. Biophys J. 2000, 79 (2): 1107-1118. 10.1016/S0006-3495(00)76364-8.View ArticleGoogle Scholar
- Liu J, Conboy JC: Asymmetric distribution of lipids in a phase segregated phospholipid bilayer observed by sum-frequency vibrational spectroscopy. J Phys Chem C. 2007, 111 (25): 8988-8999. 10.1021/jp0690547.View ArticleGoogle Scholar
- Leidy C, Kaasgaard T, Crowe JH, Mouritsen OG, Jørgensen K: Ripples and the formation of anisotropic lipid domains: imaging two-component supported double bilayers by atomic force microscopy. Biophys J. 2002, 83 (5): 2625-2633. 10.1016/S0006-3495(02)75273-9.View ArticleGoogle Scholar
- Goertz MP, Stottrup BL, Houston JE, Zhu XY: Nanomechanical contrasts of gel and fluid phase supported lipid bilayers. J Phys Chem B. 2009, 113 (27): 9335-9339. 10.1021/jp900866h.View ArticleGoogle Scholar
- Marsh D: CRC Handbook of Lipid Bilayers. 1990, CRC press, Boca Raton, FL, 1Google Scholar
- Wan Oss CJ, Absolom DR, Neumann AW: The “hydrophobic effect”: essentially a van der Waals interaction. Colloid Polym Sci. 1980, 258 (4): 424-427. 10.1007/BF01480835.View ArticleGoogle Scholar
- Yu QK, Qin GT, Darne C, Cai CZ, Wosik W, Pei SS: Fabrication of short and thin silicon cantilevers for AFM with SOI wafers. Sensor Actuat A-Phys. 2006, 126 (2): 369-374. 10.1016/j.sna.2005.10.019.View ArticleGoogle Scholar
- Damircheli M, Korayem MH: Sensitivity of higher mode of rectangular atomic force microscope to surface stiffness in air environment. Micro Nano Lett. 2013, 8 (12): 877-881. 10.1049/mnl.2013.0533.View ArticleGoogle Scholar
- Staedler T, Schiffmann K: Correlation of nanomechanical and nanotribological behavior of thin DLC coatings on different substrates. Surf Sci. 2001, 482: 1125-1129. 10.1016/S0039-6028(01)00726-9.View ArticleGoogle Scholar
- Fang TH, Chang WJ: Nanomechanical properties of copper thin films on different substrates using the nanoindentation technique. Microelectron Eng. 2003, 65 (1): 231-238. 10.1016/S0167-9317(02)00885-7.View ArticleGoogle Scholar
- Shulha H, Kovalev A, Myshkin N, Tsukruk VV: Some aspects of AFM nanomechanical probing of surface polymer films. Eur Polym J. 2004, 40 (5): 949-956. 10.1016/j.eurpolymj.2004.01.021.View ArticleGoogle Scholar
- Schön P, Bagdi K, Molnár K, Markus P, Pukánszky B, Vancso GJ: Quantitative mapping of elastic moduli at the nanoscale in phase separated polyurethanes by AFM. Eur Polym J. 2011, 47 (4): 692-698. 10.1016/j.eurpolymj.2010.09.029.View ArticleGoogle Scholar
- Derjaguin B, Muller V, Toporov YP: Effect of contact deformations on the adhesion of particles. J Colloid Interface Sci. 1975, 53 (2): 314-326. 10.1016/0021-9797(75)90018-1.View ArticleGoogle Scholar
- Dokukin ME, Sokolov I: On the measurements of rigidity modulus of soft materials in nanoindentation experiments at small depth. Macromolecules. 2012, 45 (10): 4277-4288. 10.1021/ma202600b.View ArticleGoogle Scholar
- Balzer BN, Micciulla S, Dodoo S, Zerball M, Gallei M, Rehahn M, Klitzing RV, Hugel T: Adhesion property profiles of supported thin polymer films. ACS Appl Mater Interfaces. 2013, 5 (13): 6300-6306. 10.1021/am4013424.View ArticleGoogle Scholar
- Picas L, Rico F, Scheuring S: Direct measurement of the mechanical properties of lipid phases in supported bilayers. Biophys J. 2012, 102 (1): L01-L03. 10.1016/j.bpj.2011.11.4001.View ArticleGoogle Scholar
- Elbra T, Karlqvist R, Lassila I, Hæggström E, Pesonen L: Laboratory measurements of the seismic velocities and other petrophysical properties of the Outokumpu deep drill core samples, eastern Finland. Geophys J Int. 2011, 184 (1): 405-415. 10.1111/j.1365-246X.2010.04845.x.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.