Confocal Raman Microscopy for In-situ Measurement of Phospholipid

5 hours ago - File failed to load: .... Open Access ... In this work, we employ in-situ confocal Raman microscopy to probe the partitioning of a ... T...
1 downloads 3 Views 1MB Size
Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Article

Confocal Raman Microscopy for In-situ Measurement of Phospholipid-Water Partitioning into Model Phospholipid Bilayers within Individual Chromatographic Particles Jay P. Kitt, David A. Bryce, Shelley D. Minteer, and Joel M. Harris Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b01452 • Publication Date (Web): 14 May 2018 Downloaded from http://pubs.acs.org on May 15, 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 28 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

Analytical Chemistry

Confocal Raman Microscopy for In-situ Measurement of Phospholipid-Water Partitioning into Model Phospholipid Bilayers within Individual Chromatographic Particles Jay P. Kitt, David A. Bryce, Shelley D. Minteer, and Joel M. Harris* Department of Chemistry, University of Utah, 315 South 1400 East, Salt Lake City, UT 84112-0850 USA Abstract The phospholipid-water partition coefficient is a commonly measured parameter that correlates with drug efficacy, small-molecule toxicity, and accumulation of molecules in biological systems in the environment. Despite the utility of this parameter, methods for measuring phospholipid-water partition coefficients are limited. This is due to the difficulty of making quantitative measurements in vesicle membranes or supported phospholipid bilayers, both of which are small-volume phases that challenge the sensitivity of many analytical techniques. In this work, we employ in-situ confocal Raman microscopy to probe the partitioning of a model membrane-active compound, 2-(4-isobutylphenyl) propionic acid or ibuprofen, into both hybrid- and supported-phospholipid bilayers deposited on the pore walls of individual chromatographic particles. The large surface-area-to-volume ratio of chromatographic silica allows interrogation of a significant lipid bilayer area within a very small volume. The local phospholipid concentration within a confocal probe volume inside the particle can be as high as 0.5 M, which overcomes the sensitivity limitations of making measurements in the limited membrane areas of single vesicles or planar supported bilayers. Quantitative determination of ibuprofen partitioning is achieved by using the phospholipid acyl-chains of the within-particle bilayer as an internal standard. This approach is tested for measurements of pH-dependent partitioning of ibuprofen into both hybrid-lipid and supported-lipid bilayers within silica particles, and the results are compared with octanol-water partitioning and with partitioning into individual optically-trapped phospholipid vesicle membranes. Additionally, the impact of ibuprofen partitioning on bilayer structure is evaluated for both within-particle model membranes and compared with the structural impacts of partitioning into vesicle lipid bilayers.

* Corresponding author: [email protected]

ACS Paragon Plus Environment

Analytical Chemistry 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 2 of 28

2 INTRODUCTION Measuring small-molecule partitioning into phospholipid bilayers is important because of its relationship to drug efficacy,1 small-molecule toxicity,2 and the biological impact of molecules in the environment.2,3 Despite its importance, methods for measuring phospholipidwater partition coefficients are limited due to the difficulty of making measurements on the small surface area of lipid-vesicle membranes or supported phospholipid bilayers, which challenge the sensitivity of many analytical techniques. Consequently, small molecule-phospholipid partition coefficients are often predicted based on measurements of partitioning into bulk phases (most commonly octanol)4 or by comparing chromatographic retention of an analyte on hydrophobic stationary phases (such as C18 alkyl chains) to retention of similar molecules for which lipid membrane partitioning has been previously measured.4-7 Unfortunately, small molecule partitioning into a bulk phase such as octanol often correlates poorly with partitioning of the same molecule into a phospholipid membrane.8-10 This problem is often observed in the case of ionizable molecules, where octanol-water partitioning of the charged form of the molecule has been shown to differ by as much as two orders-of-magnitude from partitioning into phospholipid vesicles membranes.11 This is due in part to the fact that a phospholipid bilayer is an interfacial phase, where partitioning can be driven by a combination of headgroup interactions including electrostatic attraction or repulsion and hydrophobic interactions with the lipid acyl-chains.8,9,12 Additionally, bulk liquid phases fail to account for dipole coupling between adjacent phospholipid acyl-chains and for acyl-chain conformation (gauche- vs trans- acyl-chain conformers) that can impact partitioning.13,14 Likewise, predicting membrane partitioning from reversed-phase chromatography is not straightforward.4,15 Reversed-phase retention depends on competing interactions of the analyte in the stationary phase versus the mobile phase, including hydrophilic/hydrophobic interactions that are influenced by mobile-phase composition and the resulting changes in surface tension.16 These factors make prediction of lipid partitioning from reversed-phase retention especially challenging for charged and polar molecules,17 which is

ACS Paragon Plus Environment

Page 3 of 28 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

Analytical Chemistry

3 apparent in the extensive correlation analyses and mobile-phase modifications required to predict partitioning of benzoic acids and phenols.18 To address these challenges, separation scientists have developed a simple modelmembrane system for on-column lipid-partitioning measurements where a hybrid bilayer is formed on the pore walls within porous n-alkane-modified chromatographic particles.19-21 The bilayer comprises a proximal leaflet of n-alkane acyl chains which serve as an interface for sorption of a distal leaflet of phospholipid. Chromatographic measurements at these hybrid bilayers have been used to accurately predict both small molecule-bilayer and peptide-bilayer partition coefficients.19,20 These bilayers show promise as effective membrane models; however, measurements of chromatographic retention can require significant time for sample elution and large sample volumes for injection and detection, which can be prohibitive for high throughput screening of drug candidates that may only be available in sub-nmol quantities. Furthermore, chromatographic measurements provide little insight into the mechanism of partitioning or the impact of partitioning on the bilayer structure, which can provide valuable insight into structurefunction relationships and provide direction for improving drug compound design. Recently, confocal Raman microscopy was used to investigate within-chromatographicparticle hybrid bilayers, providing insight into their structure as a function of temperature.22 This result, combined with the promising results from chromatographic retention experiments motivate the present work wherein confocal Raman microscopy is employed to probe smallmolecule partitioning into model lipid bilayers in individual chromatographic particles. Raman spectroscopy provides both quantitative measurement of small-molecule partitioning and the impact of partitioning on the phospholipid bilayer structure. The confocal probe volume, which has a radius (1/e2) of ~0.6 µm, is well matched to probe the interior of an individual chromatographic particle, reducing sample volumes required for partitioning measurements by more than a million-fold compared to even micro-scale extraction methods.23 Confocal Raman microscopy within individual reversed-phase chromatographic particles has been used to measure solid-phase extraction of pyrene into surface-bound C18 chains23 and liquid-liquid

ACS Paragon Plus Environment

Analytical Chemistry 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 28

4 (octanol-water) partitioning of naphthoic acid over a range of source-phase pH.24 Raman spectroscopy has been utilized previously to measure partitioning of trace organics into polydimethylsiloxane matrices,25 and within-particle confocal Raman microscopy has been used to study silica surface functionalization,26 solvation of the C18 stationary phase environment,27,28 and to investigate the mechanism of surfactant-mediated retention in ion-interaction chromatography.29 Despite correlations between hybrid-bilayer retention in chromatographic experiments and membrane partitioning, confocal-Raman microscopy measurements have shown that hybrid bilayers formed in reversed-phase chromatographic particles are interdigitated, with the acyl chains from the upper leaflet extending into the lower leaflet of C18 chains.22 This result suggests that hybrid bilayers may not be an ideal model for measuring partitioning of solutes into natural phospholipid membranes whose acyl-chains are not commonly interdigitated.30 To address this issue in the current work, partitioning measurements in hybrid bilayers are compared with a more realistic membrane model, a supported phospholipid bilayer deposited on bare porous silica surfaces by vesicle fusion.31 Supported phospholipid bilayers consist of both upper and lower leaflets of phospholipid, where the lower leaflet is localized at the pore walls within bare silica particle through interaction with a thin water layer held to the interface through dipole-dipole interactions with terminal hydroxyl groups on the silica interface.32 Raman microscopy measurements of within-particle supported lipid bilayers suggest that the bilayer structure is much closer to a more natural phospholipid vesicle bilayer than a hybrid bilayer, with lipid headgroup spacing, structural characteristics (from Raman spectra), and melting transition behavior that are similar to free-solution vesicles. To investigate how these two model bilayers compare as models for membrane association measurements, the partitioning of a membrane-active small-molecule drug, 2-(4isobutylphenyl) propionic acid or ibuprofen, into both hybrid- and supported-phospholipid bilayers deposited on the pore walls of chromatographic silica particles is investigated in-situ using confocal Raman microscopy. The results are compared with ibuprofen partitioning into

ACS Paragon Plus Environment

Page 5 of 28 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

Analytical Chemistry

5 individual optically-trapped vesicles to assess the capability of each model system to effectively predict phospholipid partition coefficients. The influence of pH on ibuprofen partitioning is also investigated through measuring partitioning over a range of pH that includes the pKa of the carboxylic-acid group on the compound. Raman spectroscopy also provides information on the impact of ibuprofen partitioning on phospholipid acyl-chain structure, which is compared for both silica supported bilayers as well as phospholipid vesicles membranes.

EXPERIMENTAL SECTION Reagents and materials. Spherical chromatographic silica particles were obtained from YMC America (YMC-Pack ODS-A and YMC Sil, Allentown, PA). Bare silica particles were 10-µm mean diameter, with a specific surface area of 220 m2/g and mean pore diameter of 29 nm as determined by Brunauer-Emmett-Teller (BET) nitrogen analysis and mercury porosimetry, respectively, by Porous Materials Inc. (Ithaca, NY). Octadecylsilane (C18) modified particles were

10-µm

mean

diameter,

monofunctionally

derivatized,

and

end-capped

with

trimethychlorosilane; the particle specific surface area was 93 m2/g with 31.9 nm diameter pores and a pore volume of 0.80 mL/g as reported by the manufacturer. Carbon content (6.9%) and specific surface area were used to compute the C18 surface coverage of 3.7 μmol/m2. Calcium chloride, sodium chloride, potassium chloride, Tris (all ACS reagent, ≥99%), chloroform (Chromasolv Plus, >99.9%), and racemic 2-(4-Isobutylphenyl) propionic acid or ibuprofen (≥98%, GC) were obtained from Sigma-Aldrich (St. Louis, MO). Sodium hydroxide (Macron, ACS Reagent Grade, ≥98%), and hydrocholoric acid (Macron, ACS Reagent Grade, 36.5-38.0%) were obtained from VWR (Radnor, PA). Isopropanol (HPLC Grade, 99%), dibasic sodium phosphate (Certified ACS, ≥99%), and monobasic sodium phosphate (ACS, 99.7%) were obtained from ThermoFisher Scientific (Waltham, MA). Water for all experiments was filtered using a Barnstead GenPure UV filtration system (ThermoFisher Scientific, Waltham, MA) and had

a

minimum

resistivity

of

18.0 MΩ·cm.

DMPC

ACS Paragon Plus Environment

(1,2-dimyristoly-sn-glycero-3-

Analytical Chemistry 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 28

6 phosphocholine) was purchased from Avanti Polar Lipids, Inc. (Alabaster, AL), diluted into chloroform, and stored at -15°C until use. Within-particle and optically trapped vesicle Raman microscopy experiments were carried out in two-port and three-port flow cells, diagrams of which have been published previously.24,33 Bulk solution Raman spectroscopy measurements of ibuprofen and hexadecane used for calibration were carried out in a microscopy well-cell, a diagram of which has been published elsewhere.23 All measurements were carried out at ambient temperature (20±2ºC). Sample preparation and characterization. Ibuprofen samples for all experiments were prepared in phosphate buffer and adjusted to final pH with small amounts of sodium hydroxide or hydrochloric acid prior to dilution to final volume. The final sample composition was 10-mM phosphate and 500-µM ibuprofen. Films of DMPC were formed by drying lipid from chloroform solution under a light stream of nitrogen and removing any remaining solvent by maintaining the film under vacuum for 1 hour. The dried lipid film was rehydrated in buffer, and small unilamellar vesicles were then formed by extruding the vesicle solution through 500-nm pores in a track-etched polycarbonate membrane using a mini-extruder™ (Avanti Polar Lipids, Alabaster, AL) maintained at 40°C (well above the DMPC melting transition). Supported DMPC bilayer particles were formed by vesicle fusion from Tris-buffered saline (TBS: 10 mM Tris, 137 mM NaCl, 2.7 mM KCl) with 5 mM CaCl2 added to facilitate fusion. Dry lipid films were formed, as above, and rehydrated in TBS to a final lipid concentration of 2 mg/mL. Vesicle fusion was carried out by mixing the lipid suspension with silica particles to achieve a concentration ratio of 2-mg lipid/1-mg silica. This slurry was sonicated using a bath sonicator (Branson 1800, Branson-Emerson, St. Louis, MO) held at 40°C (well above the DMPC melting transition)34 for 20 minutes and then stirred at the same temperature for ~12 hours. After vesicle fusion, the particle-vesicle suspension was separated by centrifugation, leaving the particles in a pellet, while excess lipid vesicles could be removed in the supernatant. The particles were then re-suspended in water, and rinsed 3 times following the

ACS Paragon Plus Environment

Page 7 of 28 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

Analytical Chemistry

7 same procedure to ensure that excess lipid in solution was removed. The lipid-filled particles were used immediately or stored refrigerated in water until use. Hybrid DMPC bilayer particles were formed by lipid deposition from isopropanol-water (15% v/v) solution. Phospholipid was first dissolved in isopropanol. C18-silica particles were then added to the isopropanol and gently swirled until the silica particles were fully wetted. The samples were then diluted with water to achieve a final composition of 1-mg lipid/1-mg particles/1-mL isopropanol-water solution. The samples were stirred overnight and rinsed following the centrifugation procedure outlined above. Within-particle and optical-trapping confocal Raman microscopy. The confocal Raman microscope used in this work has been described previously.26 Briefly, an expanded (5025- 4X-647 Special Optics) 647 nm Kr+ laser beam (Coherent, Santa Clara, CA, USA) is directed into the rear entrance port of an inverted fluorescence microscope frame (Nikon TE-200, El Segundo, CA, USA) reflected off of a dichroic mirror (Semrock, Lake Forest, IL, USA) slightly overfilling the back aperture of a 100× oil immersion objective (Nikon Plan Fluor, El Segundo, CA, USA). Scattered light is collected using the same objective, passed through a dichroic mirror and high pass filter (Semrock, Lake Forest, IL, USA) and then focused onto the entrance slit of a grating monochromator (Bruker 500-IS, Billerica, MA). Raman scattered light is detected using a charge-coupled device (CCD) camera (Andor, iDus DU401A, South Windsor, CT, USA). Raman spectra were acquired using a 300 lines/mm diffraction grating blazed at 750 nm with a resolution of 2 cm–1. Confocal microscopy was accomplished by defining a confocal aperture using the entrance slit of the monochromator (50 µm) in the horizontal dimension and by binning three rows of pixels on the CCD camera (78 µm) in the vertical dimension.35 To collect Raman scattering from within individual chromatographic particles, microscope objective was translated in the z dimension until the focused laser beam produced a visible reflected spot at the coverslip-solution interface. The microscope stage was then translated in x and y dimensions, until the focused spot was positioned below the center of an individual particle which had settled onto the coverslip surface. The microscope objective was

ACS Paragon Plus Environment

Analytical Chemistry 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 28

8 then translated upward 5 µm in the z dimension, to center the focused beam and confocal volume inside the particle where Raman scattering was excited and collected. For optical-trapping experiments, the beam was focused into a vesicle dispersion between the first two ports in the three-port microscopy flow cell. The polarizability contrast between the lipid bilayer of the vesicle and surrounding buffer solution allows optical trapping of a single vesicle in the strong electric field at the laser focus.36 Raman scattering was excited with the same laser beam and collected through the same objective. The optically-trapped vesicle was then manipulated from the vesicle dispersion between the first two ports in the flow cell into ibuprofen solution, which was flowed in between the last two ports in the flow cell. Raman spectra were collected until ibuprofen Raman bands were no longer changing in intensity. Raman spectra of ibuprofen-hexadecane standards in chloroform were collected from ~0.5 mL aliquots in a microscopy well cell. The laser was brought to focus at the coverslip-solution interface and then translated 5-µm into solution. A spectrum was acquired from each of three standards. The ratio of the areas of the ibuprofen phenyl C=C stretching mode and the hexadecane CH2-twisting mode were averaged for the three samples to provide relative intensities for quantitative analysis. All spectra were baseline corrected using a custom Matlab (Mathworks, Natik, MA) routine that fits non-peak containing portions of the spectrum to a 5th-order polynomial and subsequently subtracts it from the original data. Following baseline correction, Raman scattering bands were fit with Gaussian or Gauss-Lorentz shapes to determine peak areas for quantitative analysis. RESULTS AND DISCUSSION Quantitative measurement of phospholipid-water partitioning from within-particle bilayers. The phospholipid-water partition coefficient (kpw) is defined as the ratio of the concentration of a solute molecule partitioned into a phospholipid bilayer to its concentration in water at equilibrium between the two phases: [𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆]𝑝𝑝

𝐾𝐾𝑝𝑝𝑝𝑝 = [𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆]

𝑤𝑤

ACS Paragon Plus Environment

(1)

Page 9 of 28 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

Analytical Chemistry

9 Because the magnitudes of phospholipid partition coefficients can vary over orders-ofmagnitude, these partition coefficients are often expressed as log 𝐾𝐾𝑝𝑝𝑝𝑝

Confocal Raman microscopy has been used previously to measure partitioning into

individual optically-trapped phospholipid vesicles.37 However, these measurements are challenging due to the limited quantity of phospholipid (a single vesicle bilayer) within the optical trap being difficult to detect by Raman scattering. In this work, we propose a more sensitive method for measuring phospholipid-water partitioning with confocal Raman microscopy, wherein a model phospholipid bilayer is deposited on the pore walls of porous chromatographic particles. The large specific surface area (~100 m2/g) of the porous particles corresponds to a surface area to volume ratio of ~105 m2/L, and a supported bilayer generates a within-particle phospholipid concentration of nearly 0.5 M, well above the ~1-mM detection limit of confocal Raman microscopy, providing a greater than 50-fold enhancement in bilayer concentration

within

the

confocal-probe

volume when compared with an opticallytrapped phospholipid vesicle. The proposed within-particle method for measuring logkpw was tested for ibuprofen

partitioning into hybrid-phospholipid bilayers and supported-phospholipid bilayers, each

deposited on the interior surfaces of a porous chromatographic particle. This was achieved by collecting Raman scattering from a ~1-fL confocal probe volume23 localized within an individual particle following equilibration with the model solute. Experiments were carried out by injecting a suspension of either hybridbilayer or supported-bilayer particles into a

Figure 1. Raman spectra of hybrid (A) and supported (B) phospholipid bilayers before (black) and after (red) equilibration with 500 µM ibuprofen at pH=3.0.

ACS Paragon Plus Environment

Analytical Chemistry 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 10 of 28

10 microscopy flow cell and allowing the particles settle onto the coverslip which serves as the bottom window of the flow cell. Once particles settled onto the surface, the surrounding solution was changed by flowing a buffered solution of ibuprofen ensuring the ibuprofen concentration surrounding a particle remained constant during an experiment. Within-particle Raman spectra were collected continuously during solute accumulation and equilibrium was confirmed when consecutive within-particle spectra were no longer changing. Example spectra of a withinparticle hybrid bilayer in buffer and a hybrid bilayer that has been equilibrated with 500-μM neutral (protonated) ibuprofen (pH=3) are presented in Figure 1A, while the corresponding spectra of a supported phospholipid bilayer with and without partitioned ibuprofen are presented in Figure 1B. Determination of the concentration of ibuprofen in hybrid- or supported-bilayers was carried out as follows: the ibuprofen peaks in the Raman spectra (see Figure 1) correspond overwhelmingly to ibuprofen that is partitioned into the bilayer, because the 500-µM ibuprofen in solution lies below the detection limit of the Raman microscope. Thus, the intensity of ibuprofen Raman bands can be used to determine the within-bilayer ibuprofen concentration, which is then ratioed to the solution concentration to determine the partition coefficient. Quantification of partitioned ibuprofen is achieved by using Raman scattering from the phospholipid as an internal standard, where the ratio of the intensity (peak area) of the ibuprofen phenyl C=C stretching mode (1611 cm-1) to the intensity of the bilayer acyl-chain CH2-twisting mode (1303 cm-1), which is well-resolved from adjacent spectral modes and has well-known variation in shape and Raman scattering intensity as a function of phospholipid ordering, is compared to solution-phase standards of known ibuprofen and hexadecane concentrations; within-bilayer quantification details are presented in the Supporting Information. Measuring pH dependent phospholipid-water partitioning. When compounds undergo proton-transfer reactions, their protonation state has a significant impact on phospholipid-water partitioning. Changes in molecular charge, as protons are added or removed from the molecule, lead to pH-dependent partitioning between the bilayer and aqueous

ACS Paragon Plus Environment

Page 11 of 28 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

Analytical Chemistry

11 phases.24,38 Phospholipid-water partitioning of an acidic or basic molecule near its pKa, then, depends on the relative concentrations of charged and neutral forms of the molecule in each of the phases, described by a phospholipid-water distribution coefficient logDpw, which for a monoprotic acid is defined as:24 �𝐻𝐻𝐻𝐻𝑝𝑝 �+�𝐴𝐴−𝑝𝑝 �

𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 = log �[𝐻𝐻𝐻𝐻

− 𝑤𝑤 ]+[𝐴𝐴 𝑤𝑤 ]



(2)

The pH-dependence of the distribution coefficient is modeled by combining the expression for simple monoprotic acid-base equilibrium (Eq. 3) with the definition of the distribution coefficient (Eq. 2): 𝐾𝐾𝑎𝑎 =

[𝐴𝐴− 𝑤𝑤 ][𝐻𝐻 + 𝑤𝑤 ]

(3)

[𝐻𝐻𝐻𝐻𝑤𝑤 ]

𝐻𝐻𝐻𝐻 + 𝐾𝐾 𝐴𝐴− �10(𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 ) � 𝐾𝐾𝑝𝑝𝑤𝑤 𝑝𝑝𝑝𝑝

𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 = log �

�1+10(𝑝𝑝𝑝𝑝−𝑝𝑝𝐾𝐾𝑎𝑎 ) �



(4)

𝐻𝐻𝐻𝐻 𝐴𝐴− where 𝐾𝐾𝑝𝑝𝑝𝑝 and 𝐾𝐾𝑝𝑝𝑝𝑝 are the phospholipid-water

partition

coefficients

of

the

protonated

and

deprotonated forms of the molecule, respectively. To test each of the within-particle membrane models (supported- and hybrid-bilayers) for pHdependent phospholipid-water partitioning, Raman scattering of ibuprofen from within-particle bilayers was measured as a function of pH from pH=3.0 to pH=7.0 at 0.5-pH intervals, bracketing the pKa of ibuprofen (Figure 2A and Figure 3A). In both cases, the scattering intensity of the ibuprofen phenyl C=C stretching mode decreases as the aqueous-phase pH increases. This result is expected, as the deprotonated carboxylate form of ibuprofen should be less soluble than neutral ibuprofen in the non-polar bilayer acyl-

Figure 2. pH-dependent ibuprofen partitioning of 500 µM ibuprofen into within-particle hybrid bilayers. A. Raman spectra of ibuprofen partitioning into hybrid bilayers from pH=3.0 to pH=7.0. B. Log(P) values determined from the Raman spectra.

ACS Paragon Plus Environment

Analytical Chemistry 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 12 of 28

12 chains. The peak areas of the pH-dependent Raman spectra were converted to within-bilayer phospholipid concentrations and distribution coefficients (Supporting Information), and the results are plotted in Figures 2B and 3B. The pH-dependence of the distribution coefficient is well fit by the model (Eq. 4) allowing determination of the log partition coefficients of neutral and charged ibuprofen as well as the effective pKa of the proton-transfer reaction. The ibuprofen 𝐻𝐻𝐻𝐻 =3.2±0.1 and partition coefficients thus determined for the hybrid bilayer are log𝐾𝐾𝑃𝑃𝑃𝑃

𝐻𝐻𝐻𝐻 𝐴𝐴− 𝐴𝐴− log𝐾𝐾𝑝𝑝𝑝𝑝 =2.48±0.09 and for the supported bilayer are log𝐾𝐾𝑃𝑃𝑃𝑃 =3.5±0.1 and log𝐾𝐾𝑝𝑝𝑝𝑝 =2.9±0.1; the

fitted values of the pKa are 4.7±0.2 and 4.5±0.2 for hybrid bilayers and supported bilayers, respectively, which lie within uncertainty (95% confidence)39

of

the

literature

value,

pKa=4.45.11 The

most

common

method

for

predicting phospholipid-water partitioning is measurement of octanol-water partitioning, where octanol acts as a lipid-like phase allowing for a simple, predictive measurement of lipophilicity. However, measuring smallmolecule partitioning into a bulk phase such as octanol often poorly predicts partitioning of the same molecule into a phospholipid membrane because octanol lacks the interfacial character necessary to mimic a phospholipid bilayer. For example, when we compare partitioning of the anionic carboxylate form of ibuprofen into the within-particle

hybrid-bilayer

2.48±0.09)

and

𝐴𝐴− (log𝐾𝐾𝑝𝑝𝑝𝑝 =

supported-bilayer

Figure 3. pH-dependent ibuprofen partitioning of 500 µM ibuprofen into within-particle supported bilayers. A. Raman spectra of ibuprofen partitioning into supported bilayers from pH=3.0 to pH=7.0. B. Log(P) values determined from the Raman spectra.

ACS Paragon Plus Environment

Page 13 of 28 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

Analytical Chemistry

13 𝐴𝐴− 𝐴𝐴− (log𝐾𝐾𝑝𝑝𝑝𝑝 =2.9±0.1) to the octanol water partition coefficient (log𝐾𝐾𝑜𝑜𝑜𝑜 = -0.05±0.01)11, they differ

by more than two-orders of magnitude, suggesting that partitioning of charged ibuprofen is likely driven by an interfacial interaction where the hydrophobic phenyl-ring portion of the molecule is “hidden” from water by interacting with the phospholipid acyl chains and the charged, hydrophilic carboxylate interacts with zwitterionic phospholipid headgroups in the hydrated region of the bilayer. The ~2.5 fold greater partitioning of anionic ibuprofen in the supportedversus the hybrid-bilayer is consistent with doubling of possible head-group interactions in the supported bilayer, since the interdigitated hybrid-bilayer lacks a second head-group region in the C18 leaflet proximal to the silica surface. Interestingly, for the neutral form of ibuprofen, phospholipid-water partitioning measured in model membranes is 6-fold smaller for the hybrid-lipid bilayer and 3-fold smaller for the 𝐻𝐻𝐻𝐻 supported bilayer, than the octanol-water partition coefficient (log𝐾𝐾𝑜𝑜𝑜𝑜 =3.97±0.01).11 These

results suggest that octanol provides a better solvation environment for neutral ibuprofen than the phospholipid bilayers, likely due to the uniform polarity of octanol throughout the bulk phase, compared with a bilayer which only provides a polar environment near the head-group region. The 2-fold greater affinity of the supported bilayer compared to the hybrid bilayer is consistent with this argument, where again the interdigitated hybrid-bilayer lacks a second head-group region in the C18 leaflet proximal to the silica surface and should on-average be less polar. While octanol-water partitioning is a common predictor of phospholipid-water partitioning, the results above show that it has limitations in predicting lipid-membrane affiliation, especially for interfacially-active charged molecules. A better test of the validity of within-particle bilayers as models for lipid-membrane partitioning is to compare the results to partitioning into authentic, free-standing vesicle phospholipid bilayers. To accomplish this, phospholipid-water partitioning of ibuprofen into phospholipid vesicle membranes was measured using optical-trapping confocal Raman microscopy.37 Quantification of the ibuprofen concentration in the vesicle membrane (see Supporting Information) was made at the extreme pH values from the within-particle pH dependence (pH=3 and pH=7) where >97% and >99% of the

ACS Paragon Plus Environment

Analytical Chemistry 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 14 of 28

14 ibuprofen in solution is neutral or ionized, respectively, and compared with the model membrane measurements. The partition coefficients of neutral ibuprofen (pH=3) measured in hybrid and supported 𝐻𝐻𝐻𝐻 𝐻𝐻𝐻𝐻 =3.2±0.1 and log𝐾𝐾𝑃𝑃𝑃𝑃 =3.5±0.1, respectively) are in good agreement (95% lipid bilayers (log𝐾𝐾𝑃𝑃𝑃𝑃 𝐻𝐻𝐻𝐻 =3.3±0.1. At high pH, the charged confidence)39 with partitioning into the vesicle bilayer log𝐾𝐾𝑃𝑃𝑃𝑃

carboxylate form of ibuprofen into a lipid vesicle membrane has a log phospholipid-water

𝐴𝐴− partition coefficient of log𝐾𝐾𝑝𝑝𝑝𝑝 =2.1±0.1. This result illustrates the limitations of octanol-water

partition coefficients for predicting lipophilicity, where the vesicle membrane exhibits more than

𝐴𝐴− 100-fold greater affinity for the anionic solute than bulk-phase octanol (log𝐾𝐾𝑜𝑜𝑜𝑜 = -0.05±0.01).11

Partitioning into the lipid vesicle membrane is comparable to the partition coefficients measured

𝐴𝐴− 𝐴𝐴− =2.48±0.09) and supported-bilayer (log𝐾𝐾𝑝𝑝𝑝𝑝 =2.9±0.1) models, for the hybrid bilayer (log𝐾𝐾𝑝𝑝𝑝𝑝

which exhibit 2-fold and 6-fold greater partitioning, respectively. For hybrid bilayers, this

difference is likely due to the interdigitated nature of the bilayer which increases the spacing between headgroups,22 leaving larger portions of hydrophobic interface in contact with the solution phase and contributes to a driving force for partitioning through lowering of interfacial tension. The partitioning into the supported lipid bilayer, with a 6-fold greater affinity for ibuprofen compared to a lipid vesicle, may arise from defects31 due to substrate-bilayer interactions that lower the energy required to accommodate a solute. Additionally, ibuprofen has been observed to permeate lipid-vesicle membranes,37 which could allow the solute access to the silica surface underlying the supported phospholipid bilayer. Adsorption of ibuprofen to the silica surface40 could, therefore, also be contributing to the increased partitioning observed for the silica-supported lipid bilayer. Despite these differences, the phospholipid partitioning measured with the within-particle model bilayers are far more accurate in predicting lipid bilayer partitioning in vesicle membranes than bulk-phase octanol-water partitioning, especially for interfacially-active, charged solutes. Detecting solute-partitioning induced structural changes in model bilayers. Measuring phospholipid-water partitioning using Raman spectroscopy provides a unique

ACS Paragon Plus Environment

Page 15 of 28 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

Analytical Chemistry

15 advantage over ex situ measurement techniques because the Raman spectrum provides both quantitative and structural information about the bilayer and partitioned molecule.37 This information can offer insight into the mechanism of partitioning as well as the impact of the partitioned molecule on the lipid bilayer. To determine whether the within-particle model bilayers are reasonable structural models of free-standing vesicle membranes with respect to detecting solute-interactions, the bilayer structural changes observed in the Raman spectra of within-particle hybrid- and supported-bilayers as a function of ibuprofen partitioning were compared to the structure changes observed in the spectra from vesicle partitioning experiments. To examine the impact of solute partitioning on within-particle hybrid- and supportedphospholipid bilayers, the spectra from the pH-dependent partitioning experiments were analyzed for changes that correlated with neutral and deprotonated ibuprofen partitioning. To discover the spectral variations, the data from these experiments were

analyzed

modeling

using

curve

self-

resolution

(SMCR).41,42 SMCR is a modelfree method of multivariate curve resolution eigenvector resolve

the

independent

which

utilizes

decomposition set

of

spectra

to

linearlythat

contribute to a multivariate data set as well as the set of composition vectors that describe the relative contribution of these

Figure 4. Component spectra determined from self-modeling curve resolution analysis of pH-dependent partitioning of ibuprofen into within-particle hybrid phospholipid bilayers. The spectrum of the high pH, low partitioning, ordered, membrane in black and the spectrum of the low pH, high partitioning, disordered membrane in red.

spectra to the data as a function of the independent variable, in this case pH;41 see Supporting Information. Initial factor analysis of the data43 indicated two independent components, which

ACS Paragon Plus Environment

Analytical Chemistry 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 16 of 28

16 could be resolved by SMCR into component spectra, Figure 4 and 5 for hybrid bilayer and supported bilayer, respectively, and their corresponding pH-dependent compositions (Figures S3 and S-4). The first component in both cases accounts for the spectra at high pH, where the withinbilayer ibuprofen concentration is low as indicated by the modest intensity of the ibuprofen phenyl-ring C=C stretching mode at 1611 cm-1. Interestingly, this spectrum shows that when the within-particle ibuprofen concentration is low, the bilayer is in an ordered gel-phase. This is indicated by the large intensity of the symmetric and antisymmetric C-C stretching modes at 1061 cm-1 and 1126 cm-1, which are assigned to trans-conformers in the bilayer acyl-chains. Ordering of the acyl chains is consistent with sharpness of the CH2-twisting mode at 1296 cm-1, which indicates low heterogeneity in CH conformations, and with the large in intensity of both the symmetric and antisymmetric trans-conformer C-H stretching modes at 2847 cm-1 and 2883 cm-1,

which

are

indicative

of

vibrational coupling between adjacent acyl-chains.44 These spectra are in good agreement with the high-pH vesicle spectrum (Figure 6), where the spectra similarly show low ibuprofen concentrations as well as spectral indicators of ordered acyl-chains in the bilayer. The

second

component

responses in each case account for the spectra at low pH, and correspond to a high

within-bilayer

ibuprofen

concentration consistent with a large

Figure 5. Component spectra determined from selfmodeling curve resolution analysis of pH-dependent partitioning of ibuprofen into within-particle supported phospholipid bilayers. The spectrum of the high pH, low partitioning, ordered, membrane in red and the spectrum of the low pH, high partitioning, disordered membrane in black.

increase in the ibuprofen phenyl-ring

ACS Paragon Plus Environment

Page 17 of 28 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

Analytical Chemistry

17 C=C stretching mode (1611 cm-1). In the low pH spectrum, the acyl-chains appear disordered as indicated by decreases in the symmetric and antisymmetric C-C stretching modes (1061 cm-1 and 1126 cm-1), an increase in broadness and frequency of the CH2-twisting mode reflecting greater

heterogeneity

of

CH

conformations and consistent with disordered acyl chains, and the decrease in intensity of both the symmetric and antisymmetric C-H

Figure 6. Spectra of ibuprofen partitioned into vesicle phospholipid bilayers at pH=3.0 (black) and pH=7.0 (red). The spectrum at high pH shows low partitioning, and an ordered membrane whereas the spectrum at low pH, shows high partitioning, and a disordered membrane.

stretching modes (2847 cm-1 and 2883 cm-1), which indicate a loss of vibrational coupling between adjacent acyl-chains typical of an increase in gauche conformers in disordered bilayers. These spectra are also in good agreement with the low-pH vesicle spectrum (Figure 6), where the spectra similarly show a greatly increased ibuprofen concentration as well as spectral indicators of acyl-chain disordering in the bilayer. Together, these results show that both hybrid- and supported-lipid bilayers respond similarly to ibuprofen partitioning suggesting that both systems are good models of the structural response of the free-standing vesicle bilayer membrane. The composition vectors derived from SMCR are in good agreement with the pHdependent partitioning of ibuprofen fit using the simple pH-dependent partitioning model presented above (Supporting Information, Figures S-3 and S-4). These results show that membrane disordering follows partitioning of ibuprofen in the bilayer, but show no distinction between neutral and charged forms in inducing membrane disorder. The results also agree with previous optical-trapping confocal Raman microscopy results where membrane disordering was

ACS Paragon Plus Environment

Analytical Chemistry 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 18 of 28

18 monitored as a function of solution-phase ibuprofen concentration at pH=7.2;37 in these experiments, at high ibuprofen concentrations (>15 µM), the charged form of ibuprofen completely disorders the membrane. At this total concentration and pH, the neutral form concentration in the membrane lies below 26 µM; in the present work, the equivalent neutral form concentration is found between pH=6.0 and pH=6.5 where little disordering is observed. This observation indicates that vesicle membrane disordering in the previous work must be due to partitioning of charged ibuprofen, which suggests that incorporation of charged ibuprofen into the bilayer is like a surfactant, where the hydrophobic portion of the molecule extends into the bilayer acyl chains causing their disordering.

CONCLUSIONS Phospholipid-water partitioning is a critical molecular property in drug development and environmental chemistry, where partition coefficients are used to make predictions of efficacy, toxicity and bioaccumulation. Given the importance of these measurements, development of measurement techniques and model systems is essential to advancement of these fields. In this work, we demonstrate the use of two different model phospholipid bilayer systems capable of phospholipid-water partition coefficient measurements using confocal Raman microscopy. In these models, either hybrid or supported phospholipid bilayers are deposited onto the interior surfaces of high-surface area porous chromatographic silica particles which localize a high ‘concentration’ of lipid bilayer into a small volume, allowing detection of both the phospholipid and partitioned small-molecule. The phospholipid bilayer partitioning can be probed within a single chromatographic particle, the small volume of which (0.5 pL) allows the lipophilicity of a target compound to be assessed with very small (less than nL) samples volumes. Quantification of within-bilayer solute concentrations is achieved by using the acyl chains from the phospholipid bilayer as an internal standard and comparing the spectra to solution phase standards. Measurement of pH-dependent partitioning of ibuprofen was demonstrated, and the results were compared to octanol-water and vesicle-bilayer partitioning in

ACS Paragon Plus Environment

Page 19 of 28 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

Analytical Chemistry

19 order to assess the how well the proposed models predict phospholipid-water partition coefficients compared to traditional methods. Partitioning into model bilayers provided a much better measure of the phospholipid partition coefficients than octanol-water partitioning, likely due to the model bilayers more effectively mimicking the interfacial nature of a vesicle membrane. The impact of solute partitioning on bilayer acyl-chain structure was also examined and compared with vesicle bilayer structure during partitioning, to determine if the model bilayers reflected comparable structural changes as vesicle membranes. The changes in spectra were in good agreement with vesicle bilayers, supporting the use of these model bilayers in porous particles probed with Raman microscopy to assess the impact of small molecule partitioning on phospholipid bilayers.

ASSOCIATED CONTENT Supporting Information. The Supporting Information is available free of charge on the ACS Publications website. Additional information is provided on quantification of within-bilayer solute concentrations for hybrid-, supported-, and vesicle bilayers, and self-modeling curve resolution methods to resolve component responses from pH-dependent spectra.

ACKNOWLEDGMENTS This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Division of Energy, Chemical Sciences, Geosciences, & Biosciences under Award Number DE-FG03-93ER14333.

ACS Paragon Plus Environment

Analytical Chemistry 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 20 of 28

20 LITERATURE CITED (1)

Leeson, P.D.; Springthorpe, B. Nat. Rev. Drug Discov. 2007, 6, 881-890.

(2)

Cronin, M.T.D.; Livingstone, D.J. Predicting chemical toxicity and fate; CRC Press:

Boca Raton, FL, 2004. (3)

Calabrese, E.J.; Baldwin, L.A. Performing Ecological Risk Assessments; Lewis

Publishers: Chelsea, MI, 1993. (4)

Sangster, J. Octanol-water partition coefficients: fundamentals and physical chemistry;

John Wiley & Sons: New York, 1997. (5)

Sangster, J. J. Phys. Chem. Ref. Data 1989, 18, 1111-1229.

(6)

Partition Coefficient (n-octanol/water), Estimation by Liquid Chromatography, Product

Properties Test Guidelines OPPTS 830.7570, United States Environmental Protection Agency, U.S. Government Printing Office, 1996. (7)

Poole, S.K.; Poole, C.F. J. Chromatogr. B 2003, 797, 3-19.

(8)

Bittermann, K.; Spycher, S.; Endo, S.; Pohler, L.; Huniar, U.; Goss, K.-U.; Klamt, A. J.

Phys. Chem. B 2014, 118, 14833-14842. (9)

Hartmann, T.; Schmitt, J. Drug Discovery Today: Technol. 2004, 1, 431-439.

(10)

Escher, B.I.; Schwarzenbach, R.P. Environ. Sci. Technol. 1996, 30, 260-270.

(11)

Avdeef, A.; Box, K.; Comer, J.; Hibbert, C.; Tam, K. Pharm. Res. 1998, 15, 209-215.

(12)

Gallagher, E.S.; Adem, S.M.; Bright, L.K.; Calderon, I.A.; Mansfield, E.; Aspinwall,

C.A. Electrophoresis 2014, 35, 1099-1105. (13)

Rujanavech, C.; Silbert, D.F. J. Biol. Chem. 1986, 261, 7215-7219.

(14)

Baumgart, T.; Hunt, G.; Farkas, E.R.; Webb, W.W.; Feigenson, G.W. Biochim. Biophys.

Acta, Biomembr. 2007, 1768, 2182-2194. (15)

Sander, L.C.; Wise, S.A. J. Chromatogr. A 1993, 656, 335-351.

(16)

Snyder, L.R.; Kirkland, J.J.; Dolan, J.W. Introduction to Modern Liquid

Chromatography; 3 ed.; Wiley: Hoboken, 2009. (17)

Cecchi, T. Crit. Rev. Anal. Chem. 2008, 38, 161-213.

(18)

Ming, X.; Han, S.-y.; Qi, Z.-c.; Sheng, D.; Lian, H.-z. Talanta 2009, 79, 752-761.

(19)

Krause, E.; Dathe, M.; Wieprecht, T.; Bienert, M. J. Chromatogr. A 1999, 849, 125-133.

(20)

Ollila, F.; Halling, K.; Vuorela, P.; Vuorela, H.; Slotte, J.P. Arch. Biochem. Biophys.

2002, 399, 103-108.

ACS Paragon Plus Environment

Page 21 of 28 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

Analytical Chemistry

21 (21)

Tsirkin, I.; Grushka, E. J. Chromatogr. A 2001, 919, 245-254.

(22)

Kitt, J.P.; Harris, J.M. Langmuir 2016, 32, 9033-9044.

(23)

Kitt, J.P.; Harris, J.M. Anal. Chem. 2014, 86, 1719-1725.

(24)

Kitt, J.P.; Harris, J.M. Anal. Chem. 2015, 87, 5340-5347.

(25)

Ikechukwu, C.N.; Qiuming, Y.; Daniel, T.S. Appl. Spectrosc. 2012, 66, 1487-1491.

(26)

Houlne, M.P.; Sjostrom, C.M.; Uibel, R.H.; Kleimeyer, J.A.; Harris, J.M. Anal. Chem.

2002, 74, 4311-4319. (27)

Gasser-Ramirez, J.L.; Harris, J.M. Anal. Chem. 2009, 81, 2869-2876.

(28)

Gasser-Ramirez, J.L.; Harris, J.M. Anal. Chem. 2009, 81, 7632-7638.

(29)

Gasser-Ramirez, J.L.; Harris, J.M. Anal. Chem. 2010, 82, 5743-5750.

(30)

Yamasaki, M.; Miyazu, M.; Asano, T. Biochim. Biophys. Acta, Biomembr. 1992, 1106,

94-98. (31)

Bryce, D.A.; Kitt, J.P.; Harris, J.M. J. Am. Chem. Soc. 2018, 140, 4071–4078.

(32)

Tamm, L.K.; McConnell, H.M. Biophys. J. 1985, 47, 105-113.

(33)

Kitt, J.P.; Bryce, D.A.; Minteer, S.D.; Harris, J.M. J. Am. Chem. Soc. 2017, 139, 3851-

3860. (34)

Caffrey, M.; Hogan, J. Chem. Phys. Lipids 1992, 61, 1-109.

(35)

Williams, K.P.J.; Pitt, G.D.; Batchelder, D.N.; Kip, B.J. Appl. Spectrosc. 1994, 48, 232-

235. (36)

Cherney, D.P.; Conboy, J.C.; Harris, J.M. Anal. Chem. 2003, 75, 6621-6628.

(37)

Fox, C.B.; Horton, R.A.; Harris, J.M. Anal. Chem. 2006, 78, 4918-4924.

(38)

Huang, D.; Zhao, T.; Xu, W.; Yang, T.; Cremer, P.S. Anal. Chem. 2013, 85, 10240-

10248. (39)

Student Biometrika 1908, 6, 1-25.

(40)

Starek, M.; Krzek, J. J. Chromatogr. Sci. 2010, 48, 825-829.

(41)

Lawton, W.H.; Sylvestre, E.A. Technometrics 1971, 13, 617-633.

(42)

Fox, C.B.; Uibel, R.H.; Harris, J.M. J. Phys. Chem. B 2007, 111, 11428-11436.

(43)

Malinowski, E.R. Factor Analysis in Chemistry (2nd edition); John Wiley and Sons: New

York, 1991. (44)

Schultz, Z.D.; Levin, I.W. Annu. Rev. Anal. Chem. 2011, 4, 343-366.

ACS Paragon Plus Environment

Analytical Chemistry 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

257x138mm (96 x 96 DPI)

ACS Paragon Plus Environment

Page 22 of 28

Normalized Raman Scattering Intensity

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

Normalized Raman Scattering Intensity

Page 23 of 28

Analytical Chemistry

A Antisymmetric C-C Stretch (Trans)

CN Stretch Headgroup

Antisymmetric CH3 bend

Symmetric C-C Stretch CH2 Twist (Gauche)

Antisymmetric C-C Stretch (Trans)

800

CH2 Scissor

Symmetric CH Stretch

Ibuprofen Phenyl Ring C=C Stretch

Symmetric C-C Stretch (Trans)

B

CN Stretch Headgroup

Antisymmetric CH Stretch

Antisymmetric CH Stretch

Antisymmetric CH3 bend

Symmetric C-C Stretch CH2 Twist (Gauche) Symmetric C-C Stretch (Trans)

CH2 Scissor

Symmetric CH Stretch

Ibuprofen Phenyl Ring C=C Stretch

1000 1200 1400 1600 1800 Raman shift (cm-1)

Figure 1

ACS Paragon Plus Environment

2800

3000

Analytical Chemistry

Normalized Raman Scattering Intensity

A

Antisymmetric CH Stretch

Antisymmetric CH3 bend

Antisymmetric Symmetric C-C Stretch C-C Stretch CH Twist (Trans) 2 (Gauche) Symmetric CN Stretch C-C Stretch Headgroup (Trans)

3.4

B 800

1000

1200

Symmetric CH Stretch Ibuprofen Phenyl Ring C=C Stretch

CH2 Scissor

1400

1600

2800

3000

Raman Shift (cm-1)

3.2

3.0

Log(P)

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

Page 24 of 28

2.8

2.6

2.4 2

3

4

5

pH Figure 2

ACS Paragon Plus Environment

6

7

8

Page 25 of 28

Normalized Raman scattering Intensity

A

3.6

Antisymmetric CH Stretch

Antisymmetric C-C Stretch Antisymmetric (Trans) Symmetric CH3 bend C-C Stretch (Gauche)CH Twist

Ibuprofen Phenyl Ring C=C Stretch Symmetric CH2 Scissor CH Stretch

2

Symmetric C-C Stretch (Trans)

B 1200

1400

1600

18002800 -1

3000

Raman shift (cm )

3.4

log(P)

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

Analytical Chemistry

3.2

3.0

2.8

3

4

5

pH Figure 3

ACS Paragon Plus Environment

6

7

Analytical Chemistry

Antisymmetric CH3 bend CH2 Scissor

Antisymmetric CH Stretch

Antisymmetric C-C Stretch CH2 Twist (Trans) Symmetric C-C Stretch (Gauche)

Normalized Raman Scattering Intensity

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

Page 26 of 28

Symmetric CH Stretch

Symmetric C-C Stretch (Trans)

1000

1250

Ibuprofen Phenyl Ring C=C Stretch

1500

1750

Raman Shift (cm-1)

Figure 4

ACS Paragon Plus Environment

3000

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

Analytical Chemistry

Normalized Raman Scattering Intensity

Page 27 of 28

Antisymmetric C-C Stretch (Trans) Symmetric C-C Stretch (Gauche)

Antisymmetric CH3 bend CH2 Scissor CH2 Twist Ibuprofen Phenyl Ring C=C Stretch

Symmetric C-C Stretch (Trans)

1000

1250

Symmetric CH Stretch

Antisymmetric CH Stretch

1500

1750

Raman Shift (cm-1)

Figure 5

ACS Paragon Plus Environment

3000

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

Normalized Raman Scattering Intensity

Analytical Chemistry

Antisymmetric C-C Stretch Symmetric (Trans) C-C Stretch

Page 28 of 28

Antisymmetric CH3 bend CH2 Scissor

(Gauche) Ibuprofen Phenyl Ring C=C Stretch

CH2 Twist Symmetric C-C Stretch (Trans)

1000

1250

Antisymmetric Symmetric CH Stretch CH Stretch

1500

1750

Raman Shift (cm-1)

Figure 6

ACS Paragon Plus Environment

2750

3000