Nitrated Fatty Acids Modulate the Physical Properties of Model Membranes and the Structure of Transmembrane Proteins

Article abstract of DOI:10.1002/chem.201702041

Nitrated fatty acids (NO₂-FAs) act as anti-inflammatory signal mediators, albeit the molecular mechanisms behind NO₂-FAs’ influence on diverse metabolic and signaling pathways in inflamed tissues are essentially elusive. Here, we combine fluorescence measurements with surface-specific sum frequency generation vibrational spectroscopy and coarse-grained computer simulations to demonstrate that NO₂-FAs alter lipid organization by accumulation at the membrane–water interface. As the function of membrane proteins strongly depends on both, protein structure as well as membrane properties, we consecutively follow the structural dynamics of an integral membrane protein in presence of NO₂-FAs. Based on our results, we suggest a molecular mechanism of the NO₂-FA in vivo activity: Driven by the NO₂-FA-induced lipid layer reorganization, the structure and function of membrane-associated (signaling) proteins is indirectly affected.

Full text of DOI:10.1002/chem.201702041

A Journal of
Accepted Article
Title: Nitrated fatty acids modulate the physical properties of model
membranes and the structure of transmembrane proteins
Authors: Johannes Franz, Tristan Bereau, Stefanie Pannwitt,
Veerappan Anbazhagan, Alexander Lehr, Udo Nubbemeyer,
Ulrich Dietz, Mischa Bonn, Tobias Weidner, and Dirk
Schneider
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201702041
Link to VoR: http://dx.doi.org/10.1002/chem.201702041
Supported by

Chemistry - A European Journal
10.1002/chem.201702041
FULL PAPER
Nitrated fatty acids modulate the physical properties of model
membranes and the structure of transmembrane proteins
Johannes Franz,a,b Tristan Bereau, Stefanie Pannwitt, Veerappan Anbazhagan, Alexander Lehr,c
b
a
a,†
c
d
b
b,
a,
Udo Nubbemeyer, Ulrich Dietz, Mischa Bonn, Tobias Weidner, * Dirk Schneider *
Abstract: Nitrated fatty acids (NO
2
-FAs) act as anti-inflammatory
-FAs’
(Scheme 1), as their bis-allylic hydrogen atoms are relatively
easy to remove by external free radicals. Moreover, nitration is
accelerated in the absence of water and water-soluble
antioxidants.[ The nitro group reduces electron density at the β-
carbon of the double bond, leading to an increased reactivity
with reaction products that have been detected in carcinogenic
signal mediators, albeit the molecular mechanisms behind NO
2
influence on diverse metabolic and signaling pathways in inflamed
tissues are essentially elusive. Here, we combine fluorescence
measurements with surface-specific sum frequency generation
vibrational spectroscopy and coarse-grained computer simulations to
8]
demonstrate that NO
2
-FAs alter lipid organization by accumulation at
tissue, blood and urine.[9] The physiological impact of NO
still elusive although pleiotropic biological functions have been
suggested:[9a, 9c, 10] Interestingly, incorporation of NO
-FAs into
2
-FAs is
the membrane–water interface. As the function of membrane
proteins strongly depends on both, protein structure as well as
membrane properties, we consecutively follow the structural
2
11]
[
micelles and liposomes suppresses its dissociation, rendering
•
dynamics of an integral membrane protein in presence of NO
Based on our results, we suggest a molecular mechanism of the
NO -FA in vivo activity: Driven by the NO -FA-induced lipid layer
reorganization, the structure and function of membrane-associated
signaling) proteins is indirectly affected.
2
-FAs.
biomembranes as potential endogenous NO-storage.
2
2
(
Introduction
Excess production of reactive oxygen species (ROS) and
reactive nitrogen species (RNS), such as nitrogen monoxide
(•
Scheme 1. Schematic illustration of lipid nitration. Reactive nitrogen species
NO) and its oxidative derivatives, has been implicated in the
(RNS) diffuse into biomembranes and attack unsaturated lipid acyl chains.
development of over 200 different pathophysiologies, involving
cancer, AIDS, inflammatory diseases, diabetes and aging.[1] •NO
is a key intracellular messenger molecule, which has major
influence on the vascular tonus, blood coagulation and neuronal
activities.[2] In vivo, NO is produced enzymatically and diffuses
into the hydrophobic core of biomembranes[4] where it starts a
reaction cascade forming various highly reactive molecules.[5]
Under metabolic stress or inflammatory conditions, the levels of
ROS/RNS increase dramatically[6] and cause significant damage
to cellular structures, either directly via oxidation/nitration of
2
Moreover, in hydrophobic environments NO -FAs can form
covalent adducts with proteins,[12] which emphasizes the role of
biomembranes in nitro-lipid signaling.[13] Changes in lipid
composition modulate physical properties of biomembranes and
thereby indirectly influence the structure of transmembrane (TM)
proteins and the dynamics of TM domains.[14] [15] As products of
lipid peroxidation disturb the structure of biological
membranes,[ it is reasonable to expect altered biophysical
properties of lipid layers in presence of membrane embedded
polar nitro groups that, as a consequence, affect membrane
protein structure and function.
•
[3]
16]
lipids, proteins or DNA, or indirectly by acting as signaling
molecules, triggering apoptotic pathways.[7]
Unsaturated lipids are particularly vulnerable to nitration
To determine the impact of fatty acid nitration on model
membranes in greater detail, we first incorporated custom
[a]
[b]
Dr. J. Franz, S. Pannwitt, Prof. Dr. V. Anbazhagan, Prof. Dr. D.
Schneider
Institute for Pharmacy and Biochemistry
Johannes Gutenberg University, 55128 Mainz, Germany
E-mail: dirk.schneider@uni-mainz.de
Dr. J. Franz, Dr. T. Bereau, Prof. Dr. M. Bonn, Prof. Dr. T. Weidner
Max Planck Institute for Polymer Research, Molecular Spectroscopy
Department, 55128 Mainz, Germany
designed and synthesized[17] E-9-nitrooleic acid (NO
2
-OA,
Scheme 2, see experimental section for further synthesis
details) into model membranes. Fluorescence measurements
and surface-specific sum frequency generation (SFG) vibrational
spectroscopy were used to gather information about the impact
E-mail: weidner@mpip-mainz.mpg.de
[
c]
d]
A. Lehr, Prof. Dr. U. Nubbemeyer
Institute for Organic Chemistry
Johannes Gutenberg University, 55128 Mainz, Germany
U. Dietz
[
65191 Wiesbaden, Germany
Supporting information for this article is given via a link at the end of
the document.
2
Scheme 2. Structural formula of E-9-nitrooleic acid (NO -OA).
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201702041
FULL PAPER
of nitro groups on the molecular organization of model
membranes. The results were combined with coarse-grained
The concept of generalized polarization (GP) quantifies the
stokes-shift to reflect changes in membrane phase properties.[
At room temperature, the DSPC membrane is in the gel phase
and shows the expected high GP values (Figure 1a). Passing
through the main phase transition temperature of the lipid, the
GP values drop and saturate, which is indicative of a transition
into the fluid-crystalline phase. In the absence of free fatty acids,
the DSPC membranes show sharp phase transitions with a
phase transition temperature of 54 °C. The addition of OA or
18]
computer simulations to draw
a molecular picture of a
membrane faced with nitrosative stress. Moreover, the impact of
the altered membrane properties on protein-protein interactions
was investigated to understand the biological mode of action of
nitrolipids.
Results and Discussion
2
NO -OA gradually lowers the phase transition temperature of
DSPC liposomes in a concentration-dependent manner, albeit
calculated GP values of the gel as well as the fluid-crystalline
phases are only slightly increasing in OA-containing DSPC
membranes (Figure 1b, lower panel). However, the effects
Densification within the lipid headgroup region
First, the influence of NO -OA on the organization of biological
2
membranes was analyzed in 1,2-disteroyl-sn-glycero-3-
phosphocholine (DSPC) liposomes. In order to delineate the
specific impact of the nitro group, control measurements were
performed with the molecular precursor, oleic acid (OA). All
samples formed stable, equally sized liposomes (see supporting
information, Figure S1).
2
observed upon incorporation of NO -OA into DSPC liposomes
differ significantly: gel phase GP values decrease at higher NO
2
-
OA concentrations, whereas GP values of the fluid-crystalline
phase increase significantly with fatty acid concentration (Figure
1
2
d, lower panel). Thus, NO -OA severely influences the
Membrane integrity in the lipid headgroup region was
monitored using the fluorescent probe Laurdan, whose
fluorescence emission varies with the polarity of its
microenvironment (see experimental section, Figure 6).
molecular organization of a biological membrane by “fluidizing”
the gel phase, leading to a higher headgroup-order in the
physiologically relevant fluid lipid phase.
Disturbance of the alkyl chain order
To further define membrane activity of NO -OAs on a molecular
2
level, the molecular order in lipid monolayers was probed via
surface-specific sum frequency generation (SFG) spectroscopy.
As hydrophobic mismatch is negligible at the air/water interface,
0.6
0.4
0.2
0.0
a
b 54
5
5
4
2
0
8
1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) was used
as a lipid template, which has a phase transition temperature at
24 °C allowing to perform experiments in the phospholipids’
0
.4
DSPC
0.1 mM OA
0.2 mM OA
0.3 mM OA
0.4 mM OA
0.5 mM OA
-
0.2
0.4
liquid-expanded phase at a physiologically more relevant
temperature of 35 °C. Figure 2a shows SFG spectra of a pure
DMPC monolayer as well as DMPC monolayers containing
0.0
-
-0.4
3
3 mol% of OA or NO
assigned to CH (2850 cm ) and CH
stretch, CH asymmetric stretch (2910 cm ), CH
2
-OA, respectively. The peaks can be
2
0
30
40
50
60
70
0.0
0.1
0.2
0.3
0.4
0.5
-1
-1
Temperature (°C)
OA concentration (mM)
2
3
(2870 cm ) symmetric
-1
0
0
0
0
.6
.4
.2
.0
2
3
Fermi
c
d
54
resonance (2940 cm-1) and CH vinyl stretch (3010 cm-1)
vibrations. Interfacial water shows a broad signal between
52
50
48
-1
-1
-1
3100 cm and 3600 cm , with two features around 3250 cm
-1
and 3400 cm . In presence of the free fatty acids the water
signal rises, which indicates increased ordering of the interfacial
water. Interestingly, this effect is more pronounced in OA-
containing membranes, despite equal amounts of negatively
charged headgroups at the lipid-water interface and the lack of
DSPC
.1 mM NO -OA
0
.4
.0
0
2
-
0.2
0.4
0.2 mM NO
2
-OA
0
0
0
0
.3 mM NO -OA
2
.4 mM NO -OA
-
2
-0.4
charged NO
2 3 2
-moieties. The CH /CH ratio is a well-established
.5 mM NO -OA
2
quantity to determine the molecular order in model
2
0
30
40
50
60
70
0.0
0.1
0.2
0.3
0.4
0.5
[19]
membranes. In comparison to the pure DMPC layer, the OA-
or NO -OA-containing lipid mixtures show a significantly different
CH /CH ratio (see supporting information, Table S1, for fitting
results), indicating that the interfacial arrangement of alkyl
chains is affected by the presence of OA and NO -OA moieties,
Temperature (°C)
NO -OA concentration (mM)
2
2
3
2
Figure 1. Lipid order in DSPC liposomes determined by Laurdan
depolarization. Generalized Polarization values plotted against the
temperature in (a) DSPC/OA and (c) DSPC/NO
2
-OA liposomes, containing
2
increasing concentration of the free fatty acids. In (b) and (d) the phase
transition temperatures of the lipid mixtures are summarized in the upper
panels; the GP values in the gel phase (25 °C, open circles) as well as the
fluid phase (70 °C, solid circles) in the lower panels. The concentrations of
DSPC (1 mM) and Laurdan (2 µM) were kept constant. Every data point is the
mean of 3 individual experiments.
possibly due to a condensing effect caused by phospholipid
head-tail mismatch compensations.[20] However, in these
multicomponent systems the observed CH signals are
a
2
convolution of OA or NO -OA acyl chains with DMPC tails and
headgroups.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201702041
FULL PAPER
0
0
0
0
0
0
.5
.4
.3
.2
.1
.0
0.15
a
b
c
0
.4
+
OA
0.10
0.05
0.00
+ OA
0.3
0
.2
.1
+
NO -OA
2
+
NO -OA
2
0
DMPC
3400
d-DMPC
0.0
2
800
3000
3200
2000
2100
2200
2850
2900
2950
3000
-
1
-1
Wavenumber (cm )
Wavenumber (cm )
Figure 2. SFG spectra of lipid monolayers at the air/water interface. (a) Spectra of the CH/OH stretch region of pure DMPC and DMPC/OA or DMPC/NO
2
-OA
lipid mixtures in a molar ratio of 2:1. The same lipid mixtures were also measured with an isotopically labeled d-DMPC phospholipid template. SFG spectra of the
CD (b) and the CH region (c) are shown. The fit components shown in (b) represent the symmetric CD (black) and CD (red) contributions. All spectra were
3 2
collected at a constant surface pressure of 20 mN/m. Spectra are offset for clarity
To disentangle overlapping spectral resonances, and to
Table 1. Quantification of lipid order based on SFG.
determine the structure of the different components separately,
the same experiment was performed with deuterated
phospholipid template. 1,2-dimyristoyl-d54-sn-glycero-3-
phosphocholine (d-DMPC) provides
hydrophobic tail region. Used in OA or NO
the deuterocarbon chains spectrally separate the order of the
fatty acid additive (Figure 2c) and the additives’ impact on the
surrounding phospholipid model membrane (Figure 2b). In the
Model membrane
d-DMPC
Lipid order[a]
0.97 ± 0.02
1.07 ± 0.03
0.75 ± 0.02
a
a
fully deuterated
-OA lipid mixtures,
2
+
+
OA
NO
2
-OA
[a] Based on the methyl:methylene ratio of the symmetric stretch vibrations
CH region (Figure 2c), d-DMPC only shows asymmetric CH
2
and
in Figure 2b.
CH signals arising from the non-labeled lipid headgroup. In
3
presence of OA the spectrum is dominated by the asymmetric
stretch vibrations (Figure 2c). This can be explained by the kink
in the alkyl chain due to the double bond in cis-configuration.[
assemble and arrange themselves at the membrane-water
interface (Figure 3a), whereas the OA molecules are distributed
more evenly (Figure 3b). Accordingly, the radial distribution
21]
The d-DMPC/NO
stretch vibrations, with dominating CH
high disorder of the nitro fatty acids. The CD stretch region
Figure 2b) solely shows signals from the labeled d54-myristoyl
chains. Peaks correspond to the symmetric and asymmetric CD
2
-OA lipid monolayer shows stronger symmetric
function of NO
tight arrangement of molecules close to the membrane
headgroups (Figure 3c). This NO -OA clustering shields the
polar NO -moieties from the hydrophobic membrane
environment and is therefore energetically favorable. The NO
2
-OA shows significant peaks resulting from the
2
signals that implicate
2
(
2
2
-
2
-1
-1
(
2100 and 2200 cm ) and CD
well as CD Fermi resonance (2125 cm ). In comparison to a
pure d-DMPC monolayer, NO -OA significantly reduces the
phospholipid acyl chain order, indicated by the lower CD /CD
3
(2070 and 2225 cm ) stretch as
OA molecules arrange along the bilayer normal because the
aliphatic chains interact preferentially with the hydrocarbon tails
of the phospholipids. The NO groups slightly protrude into the
2
headgroup region but remain located below the phospholipid
phosphates, while the carboxylic groups are closer to the water
-1
3
2
3
2
ratio (Table 1), whereas OA only has a minor impact (see
supporting information, Table S2, for fitting results).
(
Figure 4).
The accumulation at the membrane-water interface locally
Polar nitro groups induce cluster formation
increases molecular packing density as shown by the Laurdan
measurements (Figure 1c). The concomitant deformation of the
NO -OA alkyl chains is in line with the experimentally found NO -
OA chain disorder (Figure 2c) as well as the looser packing of
the DMPC acyl chains (Figure 2b), gaining more space due to
At first glimpse, results from fluorescence (Figure 1) and SFG
measurements (Figure 2b and Table 1) are contradictory.
However, Laurdan measurements indicate an increased order in
the lipid headgroup region in the fluid lipid phase (Figure 1c),
whereas SFG reveals a disordering effect in the acyl-chain
2
2
the accumulated NO -OAs in the interfacial region. This does in
2
region of NO
2
-OA-containing lipid layers (Figure 2b). In fact,
turn explain the less effective condensing effect of NO -OA
2
coarse-grained computer simulations[ of NO
22]
-OA and OA in
2
(
Figure 2a, Table S1). In contrast, OA is more evenly distributed
2
DSPC membranes suggest that NO -OA molecules self-
over the membrane (Figure 3c).
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201702041
FULL PAPER
Modulation of TM protein structural dynamics
Modulation of global lipid bilayer properties, such as lipid
order, has been shown to influence the structure and function of
individual membrane proteins.[15a, 15d, 15e, 23] Hence, it is not only
crucial to understand the effect of lipid nitration on the structure
of
a biological membrane but also to reveal potential
consequences on folding, assembly and structural dynamics of
TM proteins. The TM helix dimer of the human glycophorin A
(
GpA) protein is an excellent probe to monitor the impact of the
[15c,
lipid environment on the structural dynamics of a TM protein.
2
3-24]
2
To investigate the role of NO -FAs on the stability of the
GpA TM helix dimer, the chemically synthesized GpA TM
domain was fluorescently labeled with fluorescein (Fl) or 5-
carboxytetramethylrhodamine
(TAMRA),
which
act
as
fluorescence donor or acceptor, respectively (Figure 5a and
Figure 7 in the experimental section). Dimerization of the TM
helices was monitored in 1-stearoyl-2-oleoyl-sn-glycero-3-
phosphocholine (SOPC) model membranes with increasing
amounts of incorporated OA or NO
2
-OA, respectively. In Figure
5b the calculated dimeric GpA fractions are plotted against the
Figure 3. Coarse-grained simulations of NO
membranes. Cross-sectional cartoon representations of NO
2
-OA and OA in DSPC
-OA (a) and OA
2
-OA and OA are shown as colored spheres (blue and orange
2
OA or NO -OA concentration. Both molecules significantly
reduce the amount of dimeric GpA, which can be caused by
both, altered membrane fluidity or peptide dilution. However,
2
(b) in DSPC. NO
indicate apolar and polar coarse-grained particles, respectively), while the
phospholipids appear in black and white. Water molecules are not shown for
clarity. (c) Corresponding radial distribution functions for NO
molecules. The simulation box contains 128 DSPC and 35 OA or NO
molecules, respectively.
NO
stronger compared to its precursor OA, whereas NO
2
2
-OA suppresses the helix-helix interaction propensity
2
-OA and OA
-OA has no
2
-OA
direct effects on GpA dimerization (see supporting information,
Figure S2). The linear dependency of FRET efficiencies as a
function of the acceptor mole fraction clearly demonstrates the
exclusive formation of dimers (Figure 5c). Peptide aggregation
The differences in the water spectra shown in Figure 2a further
supports NO -cluster formation: Interfacial water molecules only
need to rearrange locally in the vicinity of highly charged NO
OA clusters, whereas in homogeneous OA-containing
and/or non-specific helix-helix interactions can, therefore, be
2
[24a]
excluded.
Far-UV CD spectra of the GpA TM domain are
2
-
typical for TM α-helices, characterized by bands around 190,
04 and 220 nm in SOPC liposomes and both lipid mixtures
Figure 5d). Quantification of the spectra[25] reveals that the
a
2
(
membrane the negative charges are more evenly distributed
over the interfacial region, inducing a more extensive water
ordering. Noteworthy, computer simulations do not indicate any
inter-leaflet cross talk of the formed NO
supporting information, Figure S3).
secondary structure of the GpA TM domain is essentially not
altered in the various membrane environments. Thus, the α–
helical structure of the GpA TM domain is not significantly
2
-OA clusters (see
2
affected by the NO -OA-induced lipid reorganization, and an
altered helix structure can be ruled out as an explanation for the
enhanced loss in dimer fraction.
0
0
0
0
0
0
0
0
0
.16
.14
.12
.10
.08
.06
.04
.02
.00
NO -OA
2
-
COO
Tail
NO2
Conclusion
2
In summary, we here demonstrate that NO -FAs alter lipid
organization by cluster formation at the membrane–water
interface. This clustering affects the structural dynamics of a
DSPC
Phosphate
2
selected integral membrane protein. NO -FAs have already
been shown to affect several physiological processes and
numerous biological pleiotropic effects are described, which are
not clearly connected yet. The presented experimental results
now provide a unifying explanation for the observed effects and
0
.0 0.5 1.0 1.5 2.0 2.5 3.0
proposed
NO
2
-FA
functions.
Although
the
2
NO -OA
Distance from bilayer midplane (nm)
concentrations used in this study are above physiological levels,
the results clearly suggest a general nitrolipid mode of action:
NO -FA-induced lipid layer reorganization (indirectly) modulates
2
the structure and function of membrane-associated (signaling)
proteins, finally resulting in the pleiotropic signaling defects
observed in vivo upon lipid nitration.
Figure 4. Occupation probability of different chemical groups as
function of distance from the bilayer midplane. “COO “, “NO
refer to the carboxylic, nitro and aliphatic parts of NO -OA. The location of the
2
DSPC phosphate sphere is also shown for comparison. The scheme on the
right depicts the corresponding lipid arrangement.
a
-
2
” and “tail”
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201702041
FULL PAPER
2
Figure 5. Effect of OA and NO -OA on the GpA helix-helix association in SOPC liposomes. (a) Emission spectra of membrane incorporated GpA TM helices
labeled with fluorescein (black) or TAMRA (blue), or a 1:1 mixture of both (red). The inlays depict schemes of the dimerization-induced energy transfer (FLTR:
fluorescein- and TAMRA-labeled GpA and the dimer). (b) Dimer fractions plotted against the concentration of oleic acid or nitro-oleic acid, respectively. The
concentrations of DSPC (1 mM) and the peptide-lipid ratio were kept constant. (c) FRET efficiency plotted against the acceptor mole fraction. (d) Far-UV CD
spectra of the GpA TM domain in SOPC (gray), SOPC/OA (black) or SOPC/NO
2
-OA liposomes (red).
Methyl-10-hydroxy-9-nitrooctadecanoate (4): Reaction with methyl-9-
nitrononanoate (4.2 g, 19.3 mmol), nonanal (2.75 g, 3.5 mL, 19.3 mmol)
and potassium tert.-butanolate (110 mg, 0.98 mmol) according to the
literature procedure. Yield: 6.63 g (19.3 mmol, quantitative) crude 10-
hydroxy-9-nitrooctadecanoate (mixture of diastereomers).
Experimental Section
Synthesis of E-9 nitrooleic acid. Regio- and stereo-selective synthesis
of E-9 NO
sequence starting from cyclooctene and nonanal with 21% yield overall.
-OA succeeded as described before[17] conducting a nine step
2
Methyl-E-9-nitrooleate (5): Reaction with crude 10-hydroxy-9-
nitrooctadecanoate (6.63 g, 19.3 mmol), 4-N,N-dimethylaminopyridine
(DMAP, 225.3 mg, 1.84 mmol) and acetanhydride (2.07g, 1.91 mL, 20.3
mmol), then 4-N,N-dimethylaminopyridine (DMAP, 2.7 g, 22.1 mmol) in
dichloromethane (37 mL) according to the literature procedure.
Purification 1. column chromatography on silica gel (Et
ether (1:10), 2. Preparative HPLC (Gemini NX 250x30 mm, 95%
MeOH/5% H O). Yield: 2.82 g (8.3 mmol, 44.7 %) methyl-E-9-nitrooleate
2
O/petroleum
2
and 354 mg (0.88 mmol, 5%) 10-acetoxy-9-nitrooctadecanoate.
E-9-nitrooleic acid (6): Reaction with methyl-E-9-nitrooleate (2.56 g, 7.5
mmol) and aq. HCl (6 M, 75 mL) according to the literature procedure.
Purification via column chromatography on silica gel (Et
ether (1:5). Yield: 1.82 g (5.6 mmol, 68.4 %) E-9-nitrooleic acid. H NMR
2
O/petroleum
1
3
3
(
2
1
400 MHz, CDCl
3
, 23°C): δ = 7.04 (t, J = 7.9 Hz, 1H), 2.53 (t, J = 7.4 Hz,
H), 2.31 (t, 3J = 7.4 Hz, 2H), 2.17 (q, 3J = 7.5 Hz, 2H), 1.59 (m, 2H),
.45 (m, 4H), 1.30 – 1.23 (m, 16H), 0.84 (t, 3J = 6.7 Hz, 3H) ppm. 13
, 23°C): δ = 180.3, 151.7, 136.5, 34.0, 31.8, 29.3,
9.3, 29.1, 29.0, 28.9, 28.5, 28.0, 27.8, 26.3, 24.5, 14.0 ppm.
C
NMR (100 MHz, CDCl
2
3
Chemicals. All solvents, chemicals, and reagents were bought from
commercial sources and used without further purification. All
phospholipids used in this study [1,2-distearoyl-sn-glycero-3-
2
Scheme 3. Synthesis of E-9-nitrooleic acid (NO -OA).
Methyl-8-oxo-octanoate (1): Reaction with cyclooctene (6.61 g, 8.29 mL,
0 mmol), sodium carbonate (1.64 g, 15.48 mmol), then triethylamine (9
phosphocholine
phosphocholine (SOPC), 1,2-dimyristoyl-sn-glycero-3-phosphocholine
DMPC) and 1,2-dimyristoyl-d54-sn-glycero-3-phosphocholine (d-
(DSPC),
1-stearoyl-2-oleoyl-sn-glycero-3-
6
g, 12.4 mL, 89 mmol) and acetanhydride (17.06 g, 15.7 mL, 167 mmol)
according to the literature procedure. Yield: 11.01 g of crude methyl-8-
oxo-octanoate.
Methyl-8-hydroxy-9-nitrononanoate (2): Reaction with crude methyl-8-
oxo-octanoate (5 g, 29 mmol), nitromethane (3.54 g, 3.11 mL, 58 mmol),
potassium tert.-butanolate (164 mg, 1.46 mmol) according to the
literature procedure. Yield: 6 g (25.7 mmol, 88.5 %) methyl-8-hydroxy-9-
nitrononanoate.
(
DMPC)] were purchased from Avanti Polar Lipids (Alabama, USA). 6-
lauroyl-2-N,N-dimethylaminonaphthalene (Laurdan) and oleic acid (99%)
were purchased from Sigma Aldrich.
Laurdan measurements. 1,2-distearoyl-sn-glycero-3-phosphocholine
3
(DSPC) was dissolved in a mixture of CHCl /MeOH (1:1), the free fatty
acids (oleic acid (OA) or E-9 nitro oleic acid (NO
2
-OA)) in MeOH and
Methyl-9-nitrononanoate (3): Reaction with methyl-8-hydroxy-9-
nitrononanoate (5.95 g, 25.5 mmol), 4-N,N-dimethylaminopyridine
Laurdan in EtOH. 0.5 mol% of Laurdan (6-lauroyl-2-N,N-
dimethylaminonaphthalene) was added to the DSPC, DSPC/OA or
(DMAP, 157.3 mg, 1.29 mmol) and acetanhydride (2.87 g, 2.64 mL, 28.1
DSPC/NO
vacuum. The dried lipid films were hydrated by incubating for 30 min with
0 mM HEPES buffer (pH 7,4) containing 150 mM NaCl. The samples
2
-OA lipid mixtures before drying in a nitrogen flow and under
mmol), then sodium borohydride (1.93 g, 51.05 mmol) in ethanol (52 mL)
according to the literature procedure. Yield: 4.24 g (19.5 mmol, 76.6 %)
methyl-9-nitrononanoate.
1
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201702041
FULL PAPER
were then subjected to 5 cycles of freeze-thawing. The measurements
arising due to a random distribution of donor and acceptor peptides in the
were carried out on
a
Horiba Scientific FluoroMax-4 fluorescence
D
lipid bilayers. The dimer fraction f was calculated by assuming a
spectrometer. Fluorescent emission spectra of Laurdan were measured
from 400 to 550 nm after excitation of the Laurdan probe at 350 nm.
Based on the contribution of the red (490 nm) and blue (435 nm)
emission band of Laurdan (Figure 6), generalized polarization (GP)
values were calculated according to GP=(I435-I490)/=(I435+I490).
monomer–dimer equilibrium and a binomial distribution of donors and
acceptors in the dimer, since an oligomer without any acceptor molecules
does not contribute to FRET. Given that [T] is the total peptide mole
fraction, and [D] is the mole fraction of the dimer, the dimer fraction can
D
be calculated by f =2[D]/[T].
Circular Dichroism (CD) spectroscopy. SOPC was dissolved in a
mixture of CHCl /MeOH (1:1), the free fatty acids in MeOH and unlabeled
GpA-peptides in TFE. Samples contained 1 mM SOPC, 500 μM of OA or
NO -OA and 33 μM unlabeled GpA-peptides to reach a phospholipid-to-
peptide ratio of 1:30. The lipid/peptide mixtures were dried in vacuum
and rehydrated in phosphate buffer (10 mM NaH PO pH 7,4).
gel phase
fluid phase
DSPC
3
25 °C
70 °C
2
2
4
,
Background spectra of samples without GpA-peptides were subtracted
from sample spectra. CD spectra were recorded at 22 °C on a Jasco J-
8
15 spectrometer at a scan rate of 20 nm/min using quartz cells with a
path-length of 1 mm. Measurements were done with a response time of
and slit width of 1 nm. Spectra were averaged over three
2
s
a
4
00
425
450
475
500
525
550
consecutive scans. Deconvolution of the spectra to determine GpA
helicity was performed with the DichroWeb service using the CDSSTR
reference database.[25]
Wavelength (nm)
Figure 6. Temperature-dependent Laurdan fluorescence emission
measured in DSPC liposomes. At 25 °C, the DSPC molecules are in the gel
phase and the fluorescent dye emits at 440 nm (black line). Above the phase
transition temperature, i.e. when the DSPC membrane is in the lipid fluid
phase, the fluorescence emission maximum shifts to 490 nm (red line). The
structural formula of laurdan is depicted on the right.
Sum Frequency Generation (SFG) Vibrational Spectroscopy. 1,2-
dimyristoyl-sn-glycero-3-phosphocholine (DMPC) or isotopically labeled
1,2-dimyristoyl-d54-sn-glycero-3-phosphocholine (d-DMPC), either pure
or mixtures with 33 mol% oleic acid (OA) or nitro oleic acid (NO -OA),
2
dissolved in chloroform, were used to spread a lipid monolayer on an
aqueous subphase in a Langmuir trough (MicroTroughX, Kibron Inc.,
Finland). Temperature was set to 35 °C and surface pressure was kept
constant at 20 mN/m by a Teflon barrier (2 mm/min). All experiments
were carried out in an enclosed box purged with a pre-heated nitrogen
flow. To minimize local temperature gradients caused by laser irradiation,
the Langmuir trough was continuously moved during the experiments by
using a homemade stage equipped with two stepper motors.[26] SFG
spectra were obtained by overlapping visible and IR laser radiation in
space and time. Femtosecond seed pulses were provided by a mode-
locked Ti:sapphire laser (Mai Tai SP, Spectra-Physics) and amplified in a
regenerative amplifier (Spitfire Ace, Spectra-Physics) pumped by a
Nd:YLF laser (Empower 45, Spectra-Physics) to generate a 4.82 mJ
pulse at 800 nm with a pulse duration of 37 fs at a repetition rate of 1 kHz.
The beam was split and tunable IR pulses were generated by using one
part of the beam to pump an OPA system (TOPAS C, Light Conversion).
The bandwidth of the other part of the beam was spectrally narrowed to
Förster resonance energy transfer (FRET). Peptides corresponding to
the human GpA TM domain (69SEPEITLIIFGVMAGVIGTILLISY-
GIRRLIKK1 ) were custom-synthesized (Peptide Specialty Laboratories
GmbH, Heidelberg, Germany). The N-termini of the peptides were
fluorescently labeled (Figure 7) with 5-carboxyfluorescein (Fl) or
tetramethyl-5-carboxyrhodamine (TAMRA) (Merck Biosciences).
01
excitation
emission
-1
15 cm . The energy of the IR radiation was 4.6 μJ in the C-H region and
.8 μJ in the C-D region, whereas the VIS beam was at 20 μJ. Both
3
400
450
500
550
600
650
beams were gently focused on the sample surface with a beam diameter
of 0.1 mm. The SFG signal was dispersed into a spectrograph (Acton
SpectraPro 300i, Princeton Instruments) and detected by an EMCCD
camera (Newton, Andor). All spectra were collected in ssp (s-SFG, s-VIS,
p-IR) polarization combination, background subtracted and normalized
with a non-resonant reference spectrum of a gold-coated Si-wafer. The
spectra were fit by assuming Lorentzian lineshapes.
Wavelength (nm)
Figure 7. The fluorescently-labeled GpA TM peptides. Excitation and
emission spectra of fluorescein- (top) and TAMRA-labeled (bottom) GpA TM
peptides. The structural formula of the fluorophors are depicted on the right.
The
phospholipid
(1-stearoyl-2-oleoyl-sn-glycero-3-phosphocholine,
/MeOH (1:1), the free fatty
SOPC) was dissolved in a mixture of CHCl
3
acids in MeOH and the GpA-peptides in TFE. The Fl- and TAMRA-
labeled GpA TM helices were co-reconstituted into the SOPC liposomes
at a 1:1 molar ratio. Furthermore, liposomes containing solely Fl-labeled
GpA TM peptide served as a negative control. Upon removal of the
organic solvents, the dried lipid/peptide films were hydrated by incubating
for 30 min with 10 mM HEPES buffer (pH 7,4) containing 150 mM NaCl.
The samples were then subjected to 5 cycles of freeze-thawing. The
measurements were carried out at room temperature in an Aminco
Coarse-grained simulations. We performed coarse-grained computer
simulations using the MARTINI force field.[22a] The parametrization of OA
22b]
and NO
2
-OA was generated automatically.[
Molecular dynamics
[
22c]
simulations were run using GROMACS 4.5.3
using the standard
MARTINI parameters and polarizable water.[22d] Patches of (6 nm) were
2
generated for pure DSPC membranes, using the insane building tool,[22e]
2
amounting to 128 phospholipids in each setup. 35 OA or NO -OA
molecules were inserted randomly in the simulation box. The systems
were subsequently minimized, heated up, and equilibrated. Production
runs amounted to 1 microsecond. Radial distribution functions were
averaged over the last 800 nanosecond. For the DSPC systems,
simulations at 350K are reported, though we monitored the temperature
dependence to check the robustness of the results (data not shown).
Bowman Series
2 luminescence spectrometer. Fluorescence was
measured from 480 to 650 nm after donor excitation at 439 nm. FRET
arising due to GpA TM domain dimerization, was measured and the
energy transfer efficiency E was calculated from the donor fluorescence
intensity at 525 nm; according to E=(F −FDA)/F , where F and FDA are
D D D
the donor fluorescence intensities in absence or presence of the acceptor,
respectively. The observed FRET efficiency E was corrected for FRET
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201702041
FULL PAPER
Acknowledgements
[
16] A. Negre-Salvayre, C. Coatrieux, C. Ingueneau, R. Salvayre, Br. J.
Pharmacol. 2008, 153, 6.
J.F. and T.W. acknowledge funding from the Deutsche
Forschungsgemeinschaft (WE4478/2-1). D.S. was supported by
the Research Center "Complex Materials" (COMATT) and the
University of Mainz. The authors thank Nadja Hellmann for
scientific support.
[17] a) M. J. Gorczynski, J. Huang, S. B. King, Org. Lett. 2006, 8, 2305; b) S.
R. Woodcock, A. J. Marwitz, P. Bruno, B. P. Branchaud, Org. Lett. 2006,
8, 3931.
[
18] a) T. Parasassi, G. De Stasio, G. Ravagnan, R. M. Rusch, E. Gratton,
Biophys. J. 1991, 60, 179; b) F. M. Harris, K. B. Best, J. D. Bell,
Biochim. Biophys. Acta 2002, 1565, 123.
[19] P. Guyotsionnest, J. H. Hunt, Y. R. Shen, Phys. Rev. Lett. 1987, 59,
1
597.
Author Contributions
[
[
[
20] G. Ma, H. C. Allen, Langmuir 2007, 23, 589.
21] G. Ma, H. C. Allen, Langmuir 2006, 22, 5341.
D.S., J.F., V.A., T.W. and U.D. designed research. A.L. and U.N.
synthesized material. J.F., T.B. and S.P. performed research.
J.F., T.B., D.S, M.B. and T.W. analyzed the data and wrote the
manuscript. All authors discussed results.
22] a) S. J. Marrink, H. J. Risselada, S. Yefimov, D. P. Tieleman, A. H. de
Vries, J. Phys. Chem. B 2007, 111, 7812; b) T. Bereau, K. Kremer, J.
Chem. Theory Comput. 2015, 11, 2783; c) S. Pronk, S. Pall, R. Schulz,
P. Larsson, P. Bjelkmar, R. Apostolov, M. R. Shirts, J. C. Smith, P. M.
Kasson, D. van der Spoel, B. Hess, E. Lindahl, Bioinformatics 2013, 29,
845; d) S. O. Yesylevskyy, L. V. Schafer, D. Sengupta, S. J. Marrink,
Keywords: Fluorescence • Helix-helix interaction • Lipid nitration
PLoS Comput. Biol. 2010, 6, e1000810; e) H. I. Ingolfsson, M. N. Melo,
F. J. van Eerden, C. Arnarez, C. A. Lopez, T. A. Wassenaar, X. Periole,
A. H. de Vries, D. P. Tieleman, S. J. Marrink, J. Am. Chem. Soc. 2014,
•
Membranes • Sum frequency generation
[1]
a) J. T. Coyle, P. Puttfarcken, Science 1993, 262, 689; b) R. A. Roberts,
R. A. Smith, S. Safe, C. Szabo, R. B. Tjalkens, F. M. Robertson,
Toxicology 2010, 276, 85.
1
36, 14554.
23] V. Anbazhagan, C. Munz, L. Tome, D. Schneider, J. Mol. Biol. 2010,
04, 773.
[
[
4
[2]
G. Krauss, Biochemistry of Signal Transduction and Regulation, 4. ed.,
Viley VCH, 2008.
24] a) B. D. Adair, D. M. Engelman, Biochemistry 1994, 33, 5539; b) V.
Anbazhagan, F. Cymer, D. Schneider, Arch. Biochem. Biophys. 2010,
[
3]
4]
C. Nathan, FASEB J. 1992, 6, 3051.
495, 159; c) M. Orzaez, D. Lukovic, C. Abad, E. Perez-Paya, I.
[
M. N. Moller, Q. Li, J. R. Lancaster, Jr., A. Denicola, IUBMB Life 2007,
Mingarro, FEBS Lett. 2005, 579, 1633.
59, 243.
[
25] J. G. Lees, A. J. Miles, F. Wien, B. A. Wallace, Bioinformatics 2006, 22,
[5]
[6]
[7]
G. L. Squadrito, W. A. Pryor, Free Radic. Biol. Med. 1998, 25, 392.
R. Radi, Methods Enzymol. 1996, 269, 354.
1955.
[
26] J. Franz, M.-J. van Zadel, T. Weidner, Rev. Sci. Instrum. 2017, 88.
a) B. A. Freeman, P. R. Baker, F. J. Schopfer, S. R. Woodcock, A.
Napolitano, M. d'Ischia, J. Biol. Chem. 2008, 283, 15515; b) P. Pacher,
J. S. Beckman, L. Liaudet, Physiol. Rev. 2007, 87, 315; c) R. Radi, A.
Cassina, R. Hodara, Biol. Chem. 2002, 383, 401.
[
8]
9]
H. Rubbo, R. Radi, Biochim. Biophys. Acta 2008, 1780, 1318.
a) P. R. Baker, Y. Lin, F. J. Schopfer, S. R. Woodcock, A. L. Groeger, C.
Batthyany, S. Sweeney, M. H. Long, K. E. Iles, L. M. Baker, B. P.
Branchaud, Y. E. Chen, B. A. Freeman, J. Biol. Chem. 2005, 280,
[
42464; b) M. Balazy, T. Iesaki, J. L. Park, H. Jiang, P. M. Kaminski, M.
S. Wolin, J. Pharmacol. Exp. Ther. 2001, 299, 611; c) E. S. Lima, M. G.
Bonini, O. Augusto, H. V. Barbeiro, H. P. Souza, D. S. Abdalla, Free
Radic. Biol. Med. 2005, 39, 532.
[
10] a) A. Trostchansky, H. Rubbo, Free Radic. Biol. Med. 2008, 44, 1887;
b) N. K. Khoo, B. A. Freeman, Curr. Opin. Pharmacol. 2010, 10, 179; c)
T. Cui, F. J. Schopfer, J. Zhang, K. Chen, T. Ichikawa, P. R. Baker, C.
Batthyany, B. K. Chacko, X. Feng, R. P. Patel, A. Agarwal, B. A.
Freeman, Y. E. Chen, J. Biol. Chem. 2006, 281, 35686; d) Y. Li, J.
Zhang, F. J. Schopfer, D. Martynowski, M. T. Garcia-Barrio, A. Kovach,
K. Suino-Powell, P. R. Baker, B. A. Freeman, Y. E. Chen, H. E. Xu, Nat.
Struct. Mol. Biol. 2008, 15, 865; e) L. Villacorta, J. Zhang, M. T. Garcia-
Barrio, X. L. Chen, B. A. Freeman, Y. E. Chen, T. Cui, Am. J. Physiol.
Heart Circ. Physiol. 2007, 293, H770; f) M. M. Wright, F. J. Schopfer, P.
R. Baker, V. Vidyasagar, P. Powell, P. Chumley, K. E. Iles, B. A.
Freeman, A. Agarwal, Proc. Natl. Acad. Sci. USA 2006, 103, 4299.
11] F. J. Schopfer, P. R. Baker, G. Giles, P. Chumley, C. Batthyany, J.
Crawford, R. P. Patel, N. Hogg, B. P. Branchaud, J. R. Lancaster, Jr., B.
A. Freeman, J. Biol. Chem. 2005, 280, 19289.
[
[
12] C. Batthyany, F. J. Schopfer, P. R. Baker, R. Duran, L. M. Baker, Y.
Huang, C. Cervenansky, B. P. Branchaud, B. A. Freeman, J. Biol.
Chem. 2006, 281, 20450.
[
[
[
13] S. Bartesaghi, G. Ferrer-Sueta, G. Peluffo, V. Valez, H. Zhang, B.
Kalyanaraman, R. Radi, Amino Acids 2007, 32, 501.
14] F. Cymer, A. Veerappan, D. Schneider, Biochim. Biophys. Acta 2012,
1818, 963.
15] a) A. G. Lee, Biochim. Biophys. Acta 2004, 1666, 62; b) D. Marsh,
Biochim. Biophys. Acta 2008, 1778, 1545; c) V. Anbazhagan, D.
Schneider, Biochim. Biophys. Acta 2010, 1798, 1899; d) M. Stangl, D.
Schneider, Biochim. Biophys. Acta, Biomembr. 2015, 1848, 1886; e) D.
C. Mitchell, Biochim. Biophys. Acta, Biomembr. 2012, 1818, 951.
This article is protected by copyright. All rights reserved.

Chemistry - A European Journal
10.1002/chem.201702041
FULL PAPER
This article is protected by copyright. All rights reserved.