Synthesis and properties of dodecyl trehaloside detergents for membrane protein studies

Article abstract of DOI:10.1021/la3020404

Sugar-based detergents, mostly derived from maltose or glucose, prevail in the extraction, solubilization, stabilization, and crystallization of membrane proteins. Inspired by the broad use of trehalose for protecting biological macromolecules and lipid b

Full text of DOI:10.1021/la3020404

Article
pubs.acs.org/Langmuir
Synthesis and Properties of Dodecyl Trehaloside Detergents for
Membrane Protein Studies
Houchao Tao,† Yu Fu,† Aaron Thompson,† Sung Chang Lee,† Nicholas Mahoney,†,‡
Raymond C. Stevens,† and Qinghai Zhang*,†
†Department of Molecular Biology, The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, California 92037,
United States
‡Department of Chemistry, University of Southern Maine, Portland, Maine 04104, United States
S
* Supporting Information
ABSTRACT: Sugar-based detergents, mostly derived from
maltose or glucose, prevail in the extraction, solubilization,
stabilization, and crystallization of membrane proteins.
Inspired by the broad use of trehalose for protecting biological
macromolecules and lipid bilayer structures, we synthesized
new trehaloside detergents for potential applications in
membrane protein research. We devised an efficient synthesis
of four dodecyl trehalosides, each with the 12-carbon alkyl
chain attached to different hydroxyl groups of trehalose, thus
presenting a structurally diverse but related family of
detergents. The detergent physical properties, including
solubility, hydrophobicity, critical micelle concentration
(CMC), and size of micelles, were evaluated and compared
with the most popular maltoside analogue, β-D-dodecyl maltoside (DDM), which varied from each other due to distinct
molecular geometries and possible polar group interactions in resulting micelles. Crystals of 2-dodecyl trehaloside (2-DDTre)
were also obtained in methanol, and the crystal packing revealed multiple H-bonded interactions among adjacent trehalose
groups. The few trehaloside detergents were tested for the solubilization and stabilization of the nociceptin/orphanin FQ peptide
receptor (ORL1) and MsbA, which belong to the G-protein coupled receptor (GPCR) and ATP-binding cassette transporter
families, respectively. Our results demonstrated the utility of trehaloside detergents as membrane protein solubilization reagents
with the optimal detergents being protein dependent. Continuing development and investigations of trehaloside detergents are
attractive, given their interesting and unique chemical−physical properties and potential interactions with membrane lipids.
association with cellular phospholipid head groups by the
displacement water molecules, possibly via H-bonds between the
trehalose hydroxyl groups and the lipid phosphate groups, which
exert enhanced stability of lipid bilayer structures.5,8,9 Accord-
ingly, trehalose can inhibit the liposome fusion and lipid phase
transitions during dehydration. In addition, trehalose-derived
lipids are naturally present in some cell membranes. For example,
6,6′-diacyl trehalose lipids are the most abundant extractable
lipids from the outer cell wall of Mycobacteria tuberculosis.10,11
Similar diacyl trehalose lipids were recently found specific to
dauer larva in Caenorhabditis elegans in response to harsh
environmental conditions.12 With the alkyl chains of trehalose
lipids anchored into a lipid bilayer, the trehalose moiety was
assumed to have a more effective interaction with phospholipids,
thereby making the bilayer highly stable and impermeable and
resistant to desiccation.13,14
1. INTRODUCTION
Trehalose, also known as mycose, tremalose, or α,α-trehalose, is
an atypical, nonreducing disaccharide composed of two D-
glucoses with an α,α-1,1′-glycosidic linkage. The nonreducing
glycosidic bond of trehalose confers high chemical stability and is
resistant to hydrolysis under mildly acidic conditions and
elevated temperature. Trehalose is widely used in life sciences
as an additive reagent for the protection of protein molecules.1
The exceptional chemical stability of its glycosidic bond prevents
the Maillard reaction with protein side chains. The stabilization
effect of trehalose may also be linked to the noncovalent
interaction with solvent molecules, protein surfaces, and its
peculiar structural and physical properties, such as relatively low
conformational flexibility and high viscosity.1−4
It is also widely acknowledged that high concentrations of
trehalose help maintain the membrane structural and functional
integrity during the dehydration and rehydration process, an
effect unsurmounted by other sugars.5−7 In addition, trehalose is
abundantly found in a variety of anhydrobiotic organisms where
the cumulated amount can be up to 20% of the dry weight. The
protection mechanism of trehalose is partly explained by its
Received: May 18, 2012
Revised: July 10, 2012
Published: July 10, 2012
© 2012 American Chemical Society
11173
dx.doi.org/10.1021/la3020404 | Langmuir 2012, 28, 11173−11181

Langmuir
Article
Given the protective function of trehalose on proteins and
lipid bilayers, we speculate that trehalose-derived detergents are
of special utility in membrane protein studies. The use of
detergents is indispensable in the preparation of membrane
protein samples prior to biochemical and biophysical character-
ization, although the stability of these proteins is often a serious
concern.15 Of note, sugar-based detergents, mostly derived from
the maltose headgroup, prevail in the solubilization, purification,
stabilization, and structural studies of membrane proteins, in
particular for challenging eukaryotic targets.16 The prevalence of
noncharged maltoside detergents can be attributed to their
relatively mild, nondenaturing properties and their increasing
success in highly demanding crystallization efforts. It is
worthwhile to emphasize that the selection of a detergent
molecule, using not only the chain length but also the polar
group, has a profound effect on the stability and crystallization of
a protein target. Nonetheless, how the polar functionality of a
detergent molecule affects the crystallization outcome is poorly
understood. Part of the reason is that there is only a limited
selection of useful detergent molecules, among which a majority
are maltose derived. From this viewpoint, development of
detergents with new polar functionalities, such as trehalose
herein, is of particular interest for membrane protein research.
But to the best of our knowledge, no effort of synthesizing and
testing trehaloside detergents has been exercised in the field.
In this study, we describe our efforts toward the synthesis and
property evaluation of new trehaloside detergents along with our
long-term goal of developing new chemical tools for the study of
membrane proteins. We also report the crystal structure of one
trehaloside detergent (2-DDTre) which, compared to the free
trehalose crystal structures, suggests that the trehaloside polar
head is relatively inflexible in conformation. The crystal packing
also reveals interesting features of polar and nonpolar
interactions, which aid in our understanding of the detergent
physical properties in the solution state. Finally, we evaluate the
new trehaloside detergents for the solubilization and stabilization
of two membrane protein systems, thereby providing clues for
their future applications.
was separated, dried over Na2SO4, filtered, and concentrated in
vacuo. The residue was purified by silica gel chromatography
(eluent: hexanes/EtOAc = 3:1 and 100% acetone), which gave a
nonseparable mixture of 2- and 3-dodecylated products (2 + 3:
3.2 g, 40%). Unreacted starting material (1: 1.0 g, 18%) was also
recovered.
The above mixture of 2 and 3 was dissolved in methanol and
treated with p-toluene sulfonic acid (500 mg, 2.6 mmol) with
stirring for 5 h at room temperature. Upon completion by TLC
examination, methanol was evaporated in vacuo. The crude
product was mixed with dry silica gel powder and further dried
under vacuum. The resulting preabsorbed silica gel was loaded to
silica column (gradient eluent: DCM/MeOH = 20:1 to 4:1), by
which 2-DDTre (1.0 g, 42%) and 3-DDTre (1.1 g, 46%) was
purified individually. 2-DDTre: 1H NMR (500 MHz, CD3OD) δ
5.28 (d, J = 3.4 Hz, 1H), 5.11 (d, J = 3.7 Hz, 1H), 3.89 − 3.65 (m,
9H), 3.58 − 3.53 (m, 1H), 3.48 (dd, J = 9.7, 3.7 Hz, 1H), 3.39 −
3.33 (m, 2H), 3.22 (dd, J = 9.7, 3.7 Hz, 1H), 1.70 − 1.56 (m, 2H),
1.29 (s, 18H), 0.90 (t, J = 6.9 Hz, 3H). 13C NMR (125 MHz,
CD3OD) δ 95.8, 93.4, 81.4, 74.8, 74.0, 73.9, 73.8, 73.4, 72.6, 72.0,
71.7, 62.7, 62.6, 33.2, 31.3, 31.0, 30.9, 30.6, 27.3, 23.9, 14.6. ESI-
MS calcd for C24H46O11Na ([M + Na]+) 533.3, found 533.3. 3-
1
DDTre: H NMR (500 MHz, CD3OD) δ 5.12 (d, J = 3.6 Hz,
1H), 5.10 (d, J = 3.5 Hz, 1H), 3.87 − 3.75 (m, 7H), 3.67 (m, 2H),
3.60 (t, J = 9.2 Hz, 1H), 3.55 − 3.47 (m, 2H), 3.39 − 3.33 (m,
2H), 1.70 − 1.56 (m, 2H), 1.29 (s, 18H), 0.90 (t, J = 6.8 Hz, 3H).
13C NMR (125 MHz, CD3OD) δ 95.1, 94.9, 83.2, 74.6, 74.5,
74.0, 73.9, 73.3, 73.2, 72.1, 71.7, 62.74, 62.69, 33.2, 31.6, 30.93,
30.91, 30.89, 30.87, 30.6, 27.3, 23.9, 14.6. ESI-MS calcd for
C24H46O11Na ([M + Na]+) 533.3, found 533.3.
2.3. Synthesis of 4-DDTre and 6-DDTre. 2,2′,3,3′-Tetra-
O-benzyltrehalose (4)18 was prepared in 80% yield from di-O-
benzylidene trehalose (1) by tetra-O-benzylation and subse-
quent benzylidene deprotection. A solution of the tetrabenzyl-
trehalose 4 (2.1 g, 3.0 mmol) in DMF (18 mL) was treated with
NaH (60% dispersion, 300 mg, 7.5 mmol) at room temperature
for 2 h. To this suspension was added dropwise n-dodecyl
bromide (0.6 mL, 2.4 mmol) over a period of 30 min. The
resulting mixture was stirred for 14 h at room temperature. The
reaction was quenched by 1% HCl aqueous solution and
extracted with ethyl acetate twice. The combined organic phases
were washed with saturated NaHCO3 and NaCl solution, dried
over Na2SO4, filtered, and concentrated in vacuo. The residue
was purified by silica gel chromatography (eluent: hexanes/
acetone = 8:1 to 2:1) to separate 4- and 6-dodecylated products
(5: 520 mg, 20%; 6: 653 mg, 25%). Unreacted starting material 4
(614 mg, 29%) was recovered at the same time. The NMR
characterization data for 5 and 6 are reported in the Supporting
Information (SI). Hydrogenation of 5 and 6 in the presence of
10% Pd/C catalyst was conducted in methanol, affording the
final products 4-DDTre and 6-DDTre, respectively. 4-DDTre:
1H NMR (500 MHz, CD3OD) δ 5.11 (d, J = 3.8 Hz, 1H), 5.09 (d,
J = 3.7 Hz, 1H), 3.90 − 3.84 (m, 2H), 3.84 − 3.74 (m, 5H), 3.69
− 3.65 (m, 2H), 3.59 − 3.56 (m, 1H), 3.49 − 3.45 (m, 2H), 3.35
− 3.33 (m, 1H), 3.20 (dd, J = 9.9, 9.0 Hz, 1H), 1.63 − 1.52 (m,
2H), 1.37 − 1.25 (m, 18H), 0.90 (t, J = 7.0 Hz, 3H). 13C NMR
(125 MHz, CD3OD) δ 95.1, 95.0, 79.8, 74.7, 74.5, 74.1, 73.9,
73.4, 73.2, 73.0, 72.0, 62.7, 62.3, 33.2, 31.5, 30.9, 30.9, 30.9, 30.8,
30.6, 27.3, 23.8, 14.6. ESI-MS calcd for C24H46O11Na ([M +
2. MATERIALS AND METHODS
2.1. General Method. All organic reactions were carried out
under anhydrous conditions and an argon atmosphere, unless
otherwise noted. NMR spectra were recorded using Bruker
DRX-500, or Varian Inova-400 instruments, which were
calibrated using residual undeuterated solvent as an internal
reference. Flash column chromatography was performed using
60 Å silica gel (Acros) as a stationary phase. Thin-layer
chromatography (TLC) was performed using glass-backed silica
gel 60F254 (Merck, 250 μm thickness). The single-crystal X-ray
diffraction studies were carried out on a Bruker D8 Smart 6000
CCD diffractometer equipped with Cu Kα radiation (Bruker FR-
591 Rotating Anode Generator/Montels Optics λ = 1.5478). n-
Dodecyl β-D-maltoside (DDM) was purchased from Anatrace.
2.2. Synthesis of 2-DDTre and 3-DDTre. A solution of
4,6:4′,6′-di-O-benzylidene trehalose (1)17 (6.0 g, 11.5 mmol) in
DMF (100 mL) was treated with NaH (60% dispersion, 1.0 g,
25.0 mmol) at room temperature for 2 h. To this suspension was
added dropwise n-dodecyl bromide (3.0 mL, 12.5 mmol) over a
period of 30 min. The resulting mixture was stirred for 48 h at
room temperature. Then the reaction was quenched by careful
addition of 1% HCl solution and extracted with ethyl acetate
twice. The combined organic phases were washed sequentially
with saturated NaHCO3 and NaCl solutions. The organic layer
1
Na]+) 533.3, found 533.3. 6-DDTre: H NMR (500 MHz,
CD3OD) δ 5.11 (d, J = 4.0 Hz, 1H), 5.10 (d, J = 4.0 Hz, 1H), 3.93
(ddd, J = 10.0, 5.1, 2.1 Hz, 1H), 3.85 − 3.79 (m, 4H), 3.72 − 3.65
(m, 2H), 3.62 (dd, J = 11.1, 5.2 Hz, 1H), 3.57 − 3.43 (m, 4H),
11174
dx.doi.org/10.1021/la3020404 | Langmuir 2012, 28, 11173−11181

Langmuir
Article
3.36 − 3.33 (m, 2H), 1.61 − 1.53 (m, 2H), 1.37 − 1.25 (m, 18H),
0.90 (t, J = 7.0 Hz, 3H). 13C NMR (125 MHz, CD3OD) δ 95.2,
95.1, 74.7, 74.6, 73.9, 73.3, 73.2, 72.9, 72.7, 72.2, 72.0, 71.3, 62.7,
33.2, 30.93, 30.89, 30.82, 30.79, 30.6, 27.3, 23.9, 14.6. ESI-MS
calcd for C24H46O11Na ([M + Na]+) 533.3, found 533.3.
2.4. Evaluation of Detergent Hydrophobicity by
Reverse-Phase HPLC. HPLC was performed on a Shimadzu
instrument equipped with both a UV detector and an evaporative
light-scattering detector (ELSD). The detergent sample was
analyzed using a C18 column (Phenomenex, Cat. 309020-1,
Gemini 5 μ, 150 mm × 4.60 mm) and a mobile phase of 50%
CH3CN/50% H2O at a flow rate of 1.0 mL/min. Assignment of
each detergent peak was confirmed by multiple single and
coinjections. The HPLC retention factor (k′), used for ranking
the hydrophobicity of each detergent, was calculated according
to the equation: k′ = (tr − t0)/t0, where t0 = retention time of the
solvent front and tr = retention time of the detergent molecule.19
2.5. Measurement of Critical Micelle Concentration
(CMC). The CMC value of each detergent was determined by
monitoring the fluorescence (λex = 388 nm, λem = 477 nm) of the
ammonium salt of 8-anilino-1-naphtalenesulfonic acid (ANS),
which undergoes an increase in the fluorescence emission when
incorporated into the hydrophobic micellar environment.20
Solutions containing 10 μM ANS and different concentrations of
detergents were measured at room temperature on a Cary
Eclipse fluorescence spectrophotometer (Varian). All measure-
ments were performed in triplicate using separately prepared
solutions. The CMC was defined as the inflection point in the
plot of fluorescence intensity versus detergent concentration.
2.6. Dynamic Light Scattering (DLS) Measurements.
The hydrodynamic radius (Rh) of detergent micelles was
determined at 25 °C using a DynaPro Plate Reader (Wyatt
Technology Corporation, CA). Each detergent being tested was
solubilized in deionized (di) water at 0.2% (w/v), a
concentration well above the CMCs. Rh values were determined
using the integrated Dynamics software that analyzes the time
scale of the scattered light intensity fluctuations by an
autocorrelation function. The viscosity value of pure water was
used for all analyses. Ten acquisitions were collected for each
sample, and averages for triplicate experiments were used for the
analysis.
2.7. ORL1 Purification. The wild-type human ORL1 gene
(encoded by OPRL1; UniProt accession P41146) was
synthesized by DNA2.0, and then cloned into a modified
pFastBac1 vector (Invitrogen) containing an expression cassette
with a hemagglutinin signal sequence followed by a Flag tag at the
N terminus, and a PreScission protease site followed by a 10 ×
His tag at the C terminus. Thirty-one amino acids were deleted
from the C-terminus (residues 341−370), and 43 residues of the
N-terminus (residues 1−43) of ORL1 were replaced with the
thermostabilized apocytochrome b562RIL from Escherichia coli
(M7W, H102I, and K106L) (BRIL) protein. BRIL-ORL1 was
expressed in Spodoptera frugiperda (Sf 9) insect cells and purified
as described previously.21 Briefly, purified cell membranes were
incubated with a stabilizing antagonist (5 μM C-24: 1-benzyl-N-
{3-[spiroisobenzofuran-1(3H),40-piperidin-1-yl]propyl}-
pyrrolidine-2-carboxamide),22 50 mM HEPES (pH 7.5), 500
mM NaCl, 20 mM KCl, and 5% glycerol (v/v), and incubated at
4 °C for 1 h. Iodoacetamide (Sigma) was then added to the
membranes at a final concentration of 1 mg/mL for another 15
min before solubilization with 0.5% (w/v) each detergent being
tested for 3 h at 4 °C. The supernatant was isolated by
centrifugation at 160,000g for 45 min, supplemented with 25 mM
imidazole (pH 7.5) and incubated with TALON IMAC resin
(Clontech) overnight at 4 °C. After binding, the resin was
divided and washed with 20 column volumes of wash buffer 1 [50
mM HEPES (pH 7.5), 500 mM NaCl, 20 mM KCl, 10 mM
MgCl2, 1 mM ATP, 10% glycerol (v/v), 25 mM imidazole]
containing 5 μM C-24 and 5 × CMC of each detergent being
tested, and 25 column volumes of wash buffer 2 (same as wash
buffer 1, but without ATP and MgCl2), before protein elution
with 4 column volumes of elution buffer [50 mM HEPES (pH
7.5), 500 mM NaCl, 20 mM KCl, 10% glycerol (v/v), 250 mM
imidazole] containing C-24 and 5 × CMC of each detergent
being tested. The eluted protein in each detergent was assayed
for purity by SDS gel electrophoresis.
2.8. Thermal Stability Assay. Thermal stability data were
collected using a modified procedure incorporating a thiol-
reactive fluorophore, N-[4-(7-diethylamino-4-methyl-3-
coumarinyl)phenyl]maleimide (CPM), which undergoes an
increase in the fluorescence emission upon binding with cysteine
residues.23 Reactions contained 5 μM CPM and ∼5 μg purified
ORL1 in a buffer (elution buffer, but without imidazole)
containing 50 mM HEPES (pH 7.5), 500 mM NaCl, 20 mM
KCl, 10% glycerol (v/v), 5 μM C-24, and 5 × CMC of each
detergent being tested. Fluorescence emission was measured on
a Cary Eclipse fluorescence spectrophotometer (λex = 387 nm;
λem = 463 nm) from 20 − 90 °C with 1 °C intervals and a ramp
rate of 2 °C/min, and the background fluorescence of buffer in
the absence of protein was subtracted. Midpoints of the thermal
transitions (Tm) were obtained using a least-squares nonlinear
regression analysis (GraphPad Prism) of fluorescence signal
versus T plots according to the equation described previously.21
2.9. MsbA Purification. The recombinant MsbA construct,
originated from E. coli, was kindly provided by Dr. G. Chang.
EcMsbA was prepared and purified as described previously.24
Briefly, the bacterial cells overexpressing EcMsbA were directly
solubilized in 1% detergent solution containing 20 mM Tris, 20
mM NaCl (pH 8.0), 10% glycerol, 0.1 mg/mL of DNase I, and
proteinase inhibitor cocktail. The supernatant after centrifuga-
tion at 38,000g for 45 min was subject to Ni-affinity column for
purification. Three × CMC of each detergent being tested was
included in the wash and elution steps. MsbA was further purified
by a desalting column using buffer containing 3 × CMC of each
detergent, 20 mM Tris (pH 7.5), and 20 mM NaCl. The protein
purity was assessed by SDS gel electrophoresis.
2.10. ATPase Activity. ATPase activity was measured using
an ATP-regenerating system described by Vogel and Steinhart,25
and modified by Urbatsch et al.26 Briefly, 1−2 μg of MsbA
purified in each detergent being tested was added to 100 μL of
linked enzyme (LE) buffer at 37 °C containing 10 mM ATP, 12
mM MgCl2, 6 mM phosphoenolpyruvate (PEP), 1 mM NADH,
10 units of lactate dehydrogenase (LDH), 10 units of pyruvate
kinase (PK), and 50 mM Tris-HCl (pH 7.5). ATP hydrolysis was
measured as the decrease in absorbance of NADH at 340 nm
using a DXT880 multiplate spectrofluorimeter (Beckman-
Coulter). ATPase activity was calculated using the following
equation: ΔOD*ε/([protein]*time), where ΔOD is the change
in optical density and ε is the extinction coefficient.
3. RESULTS AND DISCUSSION
3.1. Synthesis of Trehaloside Detergents. The industrial
production of trehalose at significantly lowered cost has been
achieved in recent years,27 making possible the medium to large-
scale synthesis of trehaloside detergents at reasonable cost.
Unlike maltoside detergents in which alkyl chains were almost
11175
dx.doi.org/10.1021/la3020404 | Langmuir 2012, 28, 11173−11181

Langmuir
Article
Figure 1. Chemical structures of DDM and trehalose-derived structural isomers: 2-, 3-, 4-, and 6-DDTre.
Scheme 1. Divergent Synthesis of Four n-Dodecyl Trehaloside Detergents, Each with the Alkyl Chain Attached to Different
a
Positions of Trehalose
a
Embedded is the ORTEP drawing of the crystal structure of 2-DDTre in complex with one methanol molecule.
exclusively extended from the anomeric carbon of maltose
through a convenient glycosidic bond formation,28,29 trehaloside
detergents need to be constructed by selectively attaching the
alkyl chain through one of its four hydroxyl groups in the two-
fold symmetric glucose unit using different chemistry. We set out
to make n-dodecyl 2-, 3-, 4-, and 6-trehalosides as the analogue of
DDM (Figure 1), one of the most popular detergents being used
in membrane protein structural biology. The four dodecyl
trehaloside (DDTre) molecules have the chemical composition
and molecular weight identical to DDM yet with apparent
distinction in molecular shape from each other, hence presenting
structural diversity. Obviously, different protection/deprotection
11176
dx.doi.org/10.1021/la3020404 | Langmuir 2012, 28, 11173−11181

Langmuir
Article
operations are necessary to construct 2-, 3-, 4-, and 6-DDTre
separately, and achieving high regioselectivity of certain steps for
the concise synthesis (e.g., for 2- and 3-DDTre) could be
challenging. In this report, we devised a divergent strategy to
reduce synthetic steps, which allowed us to quickly synthesize all
four DDTre detergents in sufficient amounts for subsequent
property evaluations.
(RP)-HPLC retention factor,19 differs substantially from each
other. All DDTre molecules are less hydrophobic than DDM,
ranking in the order of 2-DDTre <3-DDTre <4-DDTre <6-
DDTre < DDM (Figure S1, SI, k′ = 1.92, 2.48, 2.99, 3.67, and
4.07 in the same order). The ranking of hydrophobicity,
however, is unpredicted given their identical chemical
compositions (identical polar and nonpolar segments for
DDTre detergents). Of note, the evaluation is based on the
assumption that each detergent runs as amonomer in the mixture
of organic solvent and water (eluent: 50% acetonitrile in water)
on a hydrophobic stationary phase. Indeed, we detected neither
additional peaks on HPLC for each detergent being tested nor
micelle structures (in the same solvent) by DLS measurements.
It is possible that the decreased hydrophobicities of 2- and 3-
DDTre relative to those of 4- and 6-DDTre and DDM is due to
more effective hydration of the hydrophobic carbons proximal to
the sugar moiety. In 2-DDTre (most hydrophilic), the alkyl chain
is positioned close to the middle of two sugar units, whereas the
alkyl chains are extended further away from the sugar hydroxyl
groups in both 6-DDTre and DDM (most hydrophobic). In both
3- and 4-DDTre, the alkyl chains are instead proximal to two
glucoside hydroxyl groups.
Starting from 4,6:4′,6′-di-O-benzylidene trehalose 1, we
carried out a nonselective alkylation of the free 2- and 3-OH
groups by reacting with 1.1 equiv of n-C12H25Br in the presence
of NaH in DMF at room temperature (Scheme 1). Precursors of
2-DDTre and 3-DDTre (2 and 3) were thus obtained as a ∼1:1
mixture in 40% yield, which ran as a single spot on either silica
plates or column, but were well separated from the less polar
multialkylated byproducts. Meanwhile, 18% of starting material
(1) was recovered. Using more equivalents of n-C12H25Br
resulted in more byproducts rather than higher yield of the
desired products. We also found that running the alkylation
reaction at lower temperature increased the selectivity for 3 (e.g.,
2:3 = 1:4 at 0 °C). For our purpose of accessing both 2- and 3-
DDTre, the nonselective alkylation is ideal. Finally, benzylidene
acetals of 2 and 3 were removed by acidic methanolysis to give 2-
and 3-DDTre in 42% and 46% isolated yield, respectively,
separated by silica chromatography. The close proximity of the
dodecyl chain to the annomeric carbon (C1) in 2-DDTre,
relative to other DDTre analogues, results in obvious separation
of both 1H and 13C NMR signals for the two annomeric carbons
and attached protons, as followed the structure assignment for 2-
DDTre and 3-DDTre. The identity of 2-DDTre was also
unambiguously confirmed by its crystal structure. Overall, 2- and
3-DDTre were each prepared in only three steps with an 18%
total yield based on recovered starting material, and the reactions
can be carried out in a multigram scale in a laboratory setting.
The strategic synthesis of 4- and 6-DDTre in a pair was similar
to the above synthesis of 2- and 3-DDTre (Scheme 1). From the
common intermediate 1, tetrabenzylation of the remaining
2,2′,3,3′−OH groups and subsequent removal of 4,6- and 4′,6′-
benzylidene acetals afforded the compound 4 in 80% yield. 4 was
then reacted with 0.8 equivalent of n-C12H25Br in the presence of
NaH at room temperature to yield the monoalkylated products 5
and 6, which were separated at this stage by silica
chromatography (5:6 = 1:1.3, 20% and 25% isolated yield,
respectively) together with the recovery of 29% unreacted 4.
Subsequent removal of benzyl groups in both 5 and 6 by Pd/C-
catalyzed hydrogenation gave 4-DDTre and 6-DDTre in 90%
yield. Later, we also explored the synthesis of 4-DDTre alone
because of its better performance for the stabilization of ORL1
(vide infra). Six steps, of which five steps for protection and
deprotection operations, were performed to achieve a concise
synthesis of 4-DDTre starting from trehalose (Scheme S1, SI).
This synthesis route is one step longer than the concurrent
synthesis of 4- and 6-DDTre outlined in Scheme 1 and may be
further optimized for higher yield.
The CMC values of 2-DDTre, 3-DDTre, 4-DDTre, and 6-
DDTre, measured in water using a fluorescence dye (ANS)
incorporation assay, are 0.018% (0.35 mM), 0.024% (0.47 mM),
0.009% (0.18 mM), and 0.007% (0.14 mM), respectively (Figure
2). Since the formation of detergent micelles is driven mainly by
Figure 2. Changes of ANS fluorescence (λex = 388 nm, λem = 477 nm)
intensities with detergent concentrations. The CMC values are defined
as the inflection point of fluorescence change.
hydrophobic association, the hydrophobicity is usually inversely
correlated with the CMC value.30 However, this is not exactly the
case for DDTre molecules and DDM. For example, the CMCs
for 4-DDTre, 6-DDTre, and DDM (0.008%, 0.16 mM) are about
the same, but the hydrophobicity measurements differ largely as
shown above. We assume that the interaction between polar
groups also plays a significant role in the micelle formation of
these sugar detergents.31 We also noted that the ANS
fluorescence curves in Figure 2 differ in the slope above the
CMC of each detergent, which might indicate changes in total
concentration and/or microenvironment of ANS molecules that
partition into the different micelle solutions.
The hydrodynamic radius (Rh) of micelles formed by each
detergent was measured using dynamic light scattering (Figure
3). 3-DDTre, 6-DDTre, and DDM micelles are similar in size
with mean radii of 3.3, 3.4, and 3.4 nm, respectively, and their size
distributions are similar as well, which suggest a similar packing of
monomers in the micelles for the three detergents. However, 2-
DDTre forms much smaller micelles (Rh = 2.9 nm) indicative of a
3.2. Micellar Properties of Trehaloside Detergents. The
four structural isomers of DDTre display major differences in
molecular geometry, which should have theoretical implication
on the respective self-assembled micelle structures. We
compared the following physical properties: solubility, hydro-
phobicity, CMC, and micelle size.
All four DDTre molecules are highly soluble in water (>10 wt
%), except that the complete dissolution of 4-DDTre is a little
slower at high concentrations. Interestingly, their hydro-
phobicity, evaluated by k′, the normalized reversed phase
11177
dx.doi.org/10.1021/la3020404 | Langmuir 2012, 28, 11173−11181

Langmuir
Article
trehalose and the trehalose dihydrate form, only small deviations
were observed.34−38 In this work, we have successfully grown 2-
DDTre crystals in methanol at room temperature. Of note, no
crystal structures of long alkyl maltosides and other disaccharide
detergents have been found by searching the Cambridge
Structural Database, which may be related to the flexible nature
of these long-chain detergent structures. Our result is very
intriguing also because methanol, a protic polar solvent like
water, is often a superior solvent for the solubilization of long-
chain hydrophobic detergents based on our experience. Align-
ment of the 2-DDTre structure with known crystal structures of
anhydrous trehalose and trehalose dihydrate revealed relatively
small deviations (Figure S2, SI), with calculated rmsd (root-
mean-square deviation) values of 0.76 Å and 1.03 Å, respectively.
The calculation is based on the distances of all C and O atoms of
trehalose after overlay of the above structures. This analysis
further supports the relatively inflexible conformation of
trehalose, even after the modification of 2-OH group with a
long alky chain that is proximal to the rotatable glycosidic linkage.
In the crystal structure, methanol was found in complex with 2-
DDTre in 1:1 ratio, being packed in a polar channel formed by
adjacent trehalosides in the symmetry-generated crystal packing
(Figure 4). Overall, the crystal lattice displayed a pattern of
alternating polar and nonpolar layers. The molecules approach
from adjacent layers in a head-to-head (trehaloside) fashion with
the alkyl chains angled at ∼65° (Figure 4b). The alkyl chains of 2-
Figure 3. Hydrodynamic radii, Rh, of detergent micelles measured by
dynamic light scattering. The data were an average of triplicate
measurements for each detergent (0.2 wt % in water, 25 °C).
larger surface curvature. Interestingly, 4-DDTre forms the largest
and most polydisperse micelles. The different micellar properties
are a reflection of the distinct molecular geometry of the four
DDTre and DDM molecules, and are potentially affected by
hydrophilic interactions between the sugar units.
3.3. 2-DDTre Crystal Structure and Molecular Inter-
action in Crystalline States. Optical rotation and computa-
tional analysis indicate that the conformation of trehalose is
relatively inflexible compared to most other disaccharides.32,33
When comparing the multiple crystal structures of anhydrous
Figure 4. Crystal packing of 2-DDTre in complex with methanol, viewed along the (a) a-axis and the (b) b-axis of the unit cells (black box, left). On the
right, zoomed-in structures, H-bond interactions among trehalosides are depicted by cyan lines.
11178
dx.doi.org/10.1021/la3020404 | Langmuir 2012, 28, 11173−11181

Langmuir
Article
Figure 5. Thermal stability of ORL1 in the presence of various detergents (5 × CMC), probed by the CPM fluorescence thermal denaturation assay.
Thermal transition temperatures (Tm) and standard errors were calculated by fitting the data to the equation as described previously.21
DDTre detergents adopt fully extended conformations and form
interleaved, parallel sheets in each hydrophobic layer. In the polar
layer, each trehaloside is connected to adjacent trehalosides with
multiple H-bonds, as shown in the right, zoomed-in structures in
Figure 4.
Although 2-DDTre was crystallized here in methanol, the
crystal structure is useful for our understanding of the molecular
interactions in an aqueous solution state.39 We assume that the
H-bonding among trehalosides plays a significant role in the
formation of micelles. Accordingly, the H-bonding pattern for
each DDTre detergent is likely different, thereby affecting the
respective micelle properties as we have measured.
addition, 6-DDTre and DDM have similar CMCs and micelle
sizes, which also gave comparable ORL1 stability.
For MsbA purified in DDTre and maltoside detergents, we
have measured the protein’s enzymatic ATPase activity. The
CPM thermal stability assay was not utilized because both
cysteine residues (C88 and C315) in the wild-type construct of
EcMsbA are not buried based on the X-ray structural model. Of
note, DDM and UDM (β-D-undecylmaltoside), generally used
for the purification and crystallization of MsbA, confer higher
ATPase activity than most commercial detergents.24,41 Our
measurements showed that all four DDTre detergents could
effectively extract and purify MsbA from cell membranes (Figure
S4), and the high ATPase catalytic activity remained in all
preparations (Figure 6). Further, EcMsbA purified in 2-DDTre
3.4. Solubilization and Stabilization of ORL1 and MsbA
Using DDTre Detergents. DDM is the most popular
commercial detergent for the solubilization of many human G-
protein coupled receptors (GPCRs) and transporters. We
evaluated the new DDTre detergents on the solubilization and
stabilization of the nociceptin/orphanin FQ peptide receptor,
ORL1, a member of opioid receptor family, and EcMsbA, a
member of ATP-binding cassette (ABC) transporters. The
structure of ORL1 containing the N-terminal fusion protein
b562RIL (BRIL-ORL) has most recently been determined using
protein purified in DDM-cholesterol hemisuccinate (CHS)
mixture and in meso phase crystallization technique.40 Several
MsbA homologue proteins including EcMsbA have been
crystallized in maltoside detergents and structurally determined
at moderate resolutions.24
In our test, all DDTre detergents behaved similarly as DDM
for the extraction and solubilization of ORL1 from insect cell
membranes (Figure S3, SI). The stability of ORL1 was
subsequently assessed using a CPM fluorescence thermal
denaturation assay that has been standardized in our laboratory
for GPCR quality control measurement.23 With this assay,
cysteine residues buried inside proteins will become accessible to
the CPM dye for covalent labeling upon thermal unfolding, and
the thermal transition temperatures (Tm) are generally used as a
measure of relative protein thermostability. We found that 4-
DDTre was superior to other trehaloside isomers and DDM in its
ability to thermally stabilize the receptor, giving the highest Tm at
61.4 °C (Figure 5). 6-DDTre (Tm = 55.7 °C) was comparable to
DDM (Tm = 53.9 °C), whereas 2-DDTre and 3-DDTre were
inferior (Tm = 41.4 and 50.1 °C, respectively). Similar stabilizing
rank orders of the DDTre detergents were also observed on β2
adrenergic receptor (data not shown). It is interesting to note
here that 4-DDTre, performing the best among the assessed
detergents for stabilizing ORL1, forms the largest micelles with
low CMC value, whereas 2-DDTre that destabilizes ORL1 forms
the smallest micelles with high CMC (Figures 2 and 3). In
Figure 6. ATPase activity of EcMsbA in the presence of various
detergents (3 × CMC), measured by a standard linked enzyme ATPase
assay. The data are an average of three measurements with standard
error bars shown.
gave relatively higher ATPase activity than DDM, UDM, and
other DDTre analogues. No protein precipitation was observed
in all the above detergent preparations after their incubation at 4
°C for more than two weeks.
That the detergent effect is protein dependent is a common
observation. It is therefore not surprising that each DDTre
detergent behaved differently in our tests on different (families
of) proteins. Thus far, we still have very limited understanding of
the detergent structure and function relationship with regards to
their utility in membrane protein studies, and the detergent
selection remains a more or less empirical process.42 Never-
theless, our preliminary solubility and stability test of new
trehaloside detergents on ORL1 and MsbA demonstrates that
they are useful tools for membrane protein studies.
11179
dx.doi.org/10.1021/la3020404 | Langmuir 2012, 28, 11173−11181

Langmuir
Article
(2) Sola-Penna, M.; Meyer-Fernandes, J. R. Stabilization against
thermal inactivation promoted by sugars on enzyme structure and
function: Why is trehalose more effective than other sugars? Arch.
Biochem. Biophys. 1998, 360, 10−14.
(3) Kaushik, J. K.; Bhat, R. Why is trehalose an exceptional protein
stabilizer? An analysis of the thermal stability of proteins in the presence
of the compatible osmolyte trehalose. J. Biol. Chem. 2003, 278, 26458−
26465.
(4) Diamant, S.; Eliahu, N.; Rosenthal, D.; Goloubinoff, P. Chemical
chaperones regulate molecular chaperones in vitro and in cells under
combined salt and heat stresses. J. Biol. Chem. 2001, 276, 39586−39591.
(5) Crowe, J. H. Trehalose as a “chemical chaperone”: Fact and fantasy.
Adv. Exp. Med. Biol. 2007, 594, 143−158.
(6) Crowe, J. H.; Crowe, L. M.; Jackson, S. A. Preservation of structural
and functional activity in lyophilized sarcoplasmic reticulum. Arch.
Biochem. Biophys. 1983, 220, 477−484.
4. CONCLUSION
The dodecyl trehalosides described in this report are structural
isomers of DDM, which all contain identical alkyl chains at
different positions on the sugar head but vary from each other in
molecular shape and the respective micellar properties. Studies of
these structural isomers and the various physicochemical
properties underscore the importance of understanding the
interactions between head groups/apolar tails for molecular
assembly. Our results suggest that H-bond interactions among
trehalosides and hydrophobic association in the apolar region, as
revealed in the 2-DDTre crystal packing, may contribute
together to DDTre micelle formation in aqueous solution. The
forces mediating detergent self-association also likely affect how
DDTre assembles onto membrane protein surfaces.
Our preliminary exploration of new trehaloside detergents
demonstrates their utility as membrane protein solubilization
reagents. The determination of their physicochemical properties
is informational for future applications and possible structure and
activity relationship analysis. In addition, trehalose detergents
display several features that make their further development
attractive. The exceptional chemical stability of trehalose
compared to other disaccharides may be advantageous for the
application of trehalose detergents in crystallization studies
which often require weeks of incubation of proteins over a wide
pH range. The stability and purity of chemical components is
always a factor affecting the outcome of protein crystallization. In
addition, the relative conformational inflexibility of trehalosides
may also be advantageous for membrane protein crystallization.
One of the major reasons that membrane proteins are difficult to
crystallize is due to the high surface entropy of the flexible
protein−detergent complexes. Therefore, limiting the flexibility
of surface-associated detergents may benefit the crystal growth.42
Moreover, given the well-known interactions between trehalose
and phospolipids, investigation of trehalose detergents in lipid-
based membrane protein platforms (e.g., bicelles, nanodiscs) is
warranted.
(7) Crowe, L. M.; Mouradian, R.; Crowe, J. H.; Jackson, S. A.;
Womersley, C. Effects of carbohydrates on membrane stability at low
water activities. Biochim. Biophys. Acta 1984, 769, 141−150.
(8) Pereira, C. S.; Lins, R. D.; Chandrasekhar, I.; Freitas, L. C. G.;
Huenenberger, P. H. Interaction of the disaccharide trehalose with a
phospholipid bilayer: A molecular dynamics study. Biophys. J. 2004, 86,
2273−2285.
(9) Crowe, J. H.; Crowe, L. M.; Chapman, D. Preservation of
membranes in anhydrobiotic organisms: the role of trehalose. Science
1984, 223, 701−703.
(10) Goren, M. B. Sulfolipid I of Mycobacterium tuberculosis, strain
H37Rv. II. Structural studies. Biochim. Biophys. Acta 1970, 210, 127−
138.
(11) Brennan, P. J.; Nikaido, H. The envelope of mycobacteria. Annu.
Rev. Biochem. 1995, 64, 29−63.
(12) Penkov, S.; Mende, F.; Zagoriy, V.; Erkut, C.; Martin, R.; Paessler,
U.; Schuhmann, K.; Schwudke, D.; Gruner, M.; Maentler, J.; Reichert-
Mueller, T.; Shevchenko, A.; Knoelker, H.-J.; Kurzchalia, T. V.
Maradolipids: Diacyltrehalose glycolipids specific to dauer larva in
Caenorhabditis elegans. Angew. Chem., Int. Ed. 2010, 49 (9430−9435),
S9430/1−S9430/10.
(13) Harland, C. W.; Rabuka, D.; Bertozzi, C. R.; Parthasarathy, R. The
Mycobacterium tuberculosis virulence factor trehalose dimycolate
imparts desiccation resistance to model mycobacterial membranes.
Biophys. J. 2008, 94, 4718−4724.
(14) Harland, C. W.; Botyanszki, Z.; Rabuka, D.; Bertozzi, C. R.;
Parthasarathy, R. Synthetic trehalose glycolipids confer desiccation
resistance to supported lipid monolayers. Langmuir 2009, 25, 5193−
5198.
ASSOCIATED CONTENT
■
S
* Supporting Information
Supplemental figures, experimental procedure for the synthesis
of 4-DDTre, NMR spectra for new compounds, and crystallo-
graphic information files (CIF) of 2-DDTre. This material is
available free of charge via the Internet at http://pubs.acs.org.
(15) le Maire, M.; Champeil, P.; Moller, J. V. Interaction of membrane
proteins and lipids with solubilizing detergents. Biochim. Biophys. Acta
2000, 1508, 86−111.
(16) Newstead, S.; Ferrandon, S.; Iwata, S. Rationalizing α-helical
membrane protein crystallization. Protein Sci. 2008, 17, 466−472.
(17) Baddeley, T. C.; Wardell, J. L. Synthesis of per- and poly-
substituted trehalose derivatives: studies of properties relevant to their
use as excipients for controlled drug release. J. Carbohydr. Chem. 2009,
28, 198−221.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: qinghai@scripps.edu.
Notes
The authors declare no competing financial interest.
(18) Hough, L.; Palmer, A. K.; Richards, A. C. Chemical modification
of trehalose. Part XI. 6,6′-dideoxy-6,6′-difluoro-α,α-trehalose and its
galacto-analog. J. Chem. Soc., Perkin Trans. 1 1972, 2513−2517.
(19) Armstrong, M. J.; Carey, M. C. The hydrophobic-hydrophilic
balance of bile salts. Inverse correlation between reverse-phase high
performance liquid chromatographic mobilities and micellar cholester-
ol-solubilizing capacities. J. Lipid Res. 1982, 23, 70−80.
(20) De Vendittis, E.; Palumbo, G.; Parlato, G.; Bocchini, V. A
fluorimetric method for the estimation of the critical micelle
concentration of surfactants. Anal. Biochem. 1981, 115, 278−286.
(21) Thompson, A. A.; Liu, J. J.; Chun, E.; Wacker, D.; Wu, H.;
Cherezov, V.; Stevens, R. C. GPCR stabilization using the bicelle-like
architecture of mixed sterol-detergent micelles. Methods 2011, 55, 310−
307.
ACKNOWLEDGMENTS
■
We are grateful for the financial support by the NIH Roadmap
Grant P50 GM073197. We thank G. Chang for providing
EcMsbA construct, M.G. Finn for use of his fluorescence
spectrophotometer and DLS instrument, and C. Moore in the
UCSD X-ray crystallography facility for collecting the diffraction
data of 2-DDTre crystals.
REFERENCES
■
(1) Ohtake, S.; Wang, Y. J. Trehalose: Current use and future
applications. J. Pharm. Sci. 2011, 100, 2020−2053.
11180
dx.doi.org/10.1021/la3020404 | Langmuir 2012, 28, 11173−11181

Langmuir
Article
(22) Goto, Y.; Arai-Otsuki, S.; Tachibana, Y.; Ichikawa, D.; Ozaki, S.;
Takahashi, H.; Iwasawa, Y.; Okamoto, O.; Okuda, S.; Ohta, H.; Sagara,
T. Identification of a novel spiropiperidine opioid receptor-like 1
antagonist class by a focused library approach featuring 3D-
pharmacophore similarity. J. Med. Chem. 2006, 49, 847−849.
(23) Alexandrov, A. I.; Mileni, M.; Chien, E. Y.; Hanson, M. A.;
Stevens, R. C. Microscale fluorescent thermal stability assay for
membrane proteins. Structure 2008, 16, 351−359.
(24) Ward, A.; Reyes, C. L.; Yu, J.; Roth, C. B.; Chang, G. Flexibility in
the ABC transporter MsbA: Alternating access with a twist. Proc. Natl.
Acad. Sci. U.S.A. 2007, 104, 19005−19010.
(25) Vogel, G.; Steinhart, R. ATPase of Escherichia coli: purification,
dissociation, and reconstitution of the active complex from the isolated
subunits. Biochemistry 1976, 15, 208−216.
(26) Urbatsch, I. L.; Sankaran, B.; Weber, J.; Senior, A. E. P-
glycoprotein is stably inhibited by vanadate-induced trapping of
nucleotide at a single catalytic site. J. Biol. Chem. 1995, 270, 19383−
19390.
(27) Higashiyama, T. Novel functions and applications of trehalose.
Pure Appl. Chem. 2002, 74, 1263−1269.
(28) Rosevear, P.; VanAken, T.; Baxter, J.; Ferguson-Miller, S. Alkyl
glycoside detergents: a simpler synthesis and their effects on kinetic and
physical properties of cytochrome c oxidase. Biochemistry 1980, 19,
4108−4115.
(29) Hong, W. X.; Baker, K. A.; Ma, X.; Stevens, R. C.; Yeager, M.;
Zhang, Q. Design, synthesis, and properties of branch-chained maltoside
detergents for stabilization and crystallization of integral membrane
proteins: human connexin 26. Langmuir 2010, 26, 8690−8696.
(30) Tanford, C. Hydrophobic Effect: Formation of Micelles and
Biological Membranes; John Wiley & Sons: New York, 1973.
(31) Venkatesan, P.; Cheng, Y.; Kahne, D. Hydrogen-bonding in
micelle formation. J. Am. Chem. Soc. 1994, 116, 6955−6956.
(32) Duda, C. A.; Stevens, E. S. Trehalose conformation in aqueous-
solution from optical-rotation. J. Am. Chem. Soc. 1990, 112, 7406−7407.
(33) Dowd, M. K.; Reilly, P. J.; French, A. D. Conformational-analysis
of trehalose disaccharides and analogs using Mm3. J. Comput. Chem.
1992, 13, 102−114.
(34) Brown, G. M.; Rohrer, D. C.; Berking, B.; Beevers, C. A.; Gould,
R. O.; Simpson, R. Crystal-Structure of α,α-trehalose dihydrate from 3
independent X-ray determinations. Acta Crystallogr., Sect. B 1972, 28,
3145−3158.
(35) Taga, T.; Senma, M.; Osaki, K. Crystal and molecular-structure of
trehalose dihydrate. Acta Crystallogr., Sect. B 1972, 28, 3258−3263.
(36) Jeffrey, G. A.; Nanni, R. The crystal-structure of anhydrous α,α-
trehalose at −150-degrees. Carbohydr. Res. 1985, 137, 21−30.
(37) Nagase, H.; Ogawa, N.; Endo, T.; Shiro, M.; Ueda, H.; Sakurai, M.
Crystal structure of an anhydrous form of trehalose: Structure of water
channels of trehalose polymorphism. J. Phys. Chem. B 2008, 112, 9105−
9111.
(38) Stevens, E. D.; Dowd, M. K.; Johnson, G. P.; French, A. D.
Experimental and theoretical electron density distribution of α,α-
trehalose dihydrate. Carbohydr. Res. 2010, 345, 1469−1481.
(39) Chae, P. S.; Guzei, I. A.; Gellman, S. H. Crystallographic
characterization of N-oxide tripod amphiphiles. J. Am. Chem. Soc. 2010,
132, 1953−1959.
(40) Thompson, A. A.; Liu, W.; Chun, E.; Katritch, V.; Wu, H.; Vardy,
E.; Huang, X.-P.; Trapella, C.; Guerrini, R.; Calo, G.; Roth, B. L.;
Cherezov, V.; Stevens, R. C. Structure of the nociceptin/orphanin FQ
receptor in complex with a peptide mimetic. Nature 2012, 485, 395−
399.
(41) Zhang, Q.; Ma, X.; Ward, A.; Hong, W. X.; Jaakola, V. P.; Stevens,
R. C.; Finn, M. G.; Chang, G. Designing facial amphiphiles for the
stabilization of integral membrane proteins. Angew. Chem., Int. Ed. 2007,
46, 7023−7025.
(42) Zhang, Q.; Tao, H.; Hong, W. X. New amphiphiles for membrane
protein structural biology. Methods 2011, 55, 318−323.
11181
dx.doi.org/10.1021/la3020404 | Langmuir 2012, 28, 11173−11181