Synthesis of trifunctional phosphatidylserine probes for identification of lipid-binding proteins

Article abstract of DOI:10.1002/ejoc.201001264

Phosphatidylserine (PS) lipids play a number of roles in cell biology, one of which is to mark apoptotic cells for clearance by macrophages. PS recognition triggers macrophage recognition and clearance, which occurs concomitantly with active suppression o

Full text of DOI:10.1002/ejoc.201001264

FULL PAPER
DOI: 10.1002/ejoc.201001264
Synthesis of Trifunctional Phosphatidylserine Probes for Identification of
Lipid-Binding Proteins
Saibal Bandyopadhyay[a] and Dennis Bong*[a]
Keywords: Lipids / Sensors / Receptors / Proteins / Synthesis design
Phosphatidylserine (PS) lipids play a number of roles in cell
biology, one of which is to mark apoptotic cells for clearance
headgroup for recognition, a benzophenone moiety for pho-
toaffinity cross-linking, and an alkyne for post-labeling read-
by macrophages. PS recognition triggers macrophage re- out. These probes may be useful tools to identify new PS re-
cognition and clearance, which occurs concomitantly with
active suppression of an inflammatory response. A small
number of proteins in different tissues have been identified
as PS receptors, and we hypothesize that many more PS re-
ceptors (PSRs). The PS lipid probes thus have potential im-
pact in the areas of discovery biology, anti-inflammatory
therapeutics, and cellular delivery. We present herein a ver-
satile and robust synthetic approach to trifunctional phos-
ceptors have yet to be found. We have designed and synthe- phatidyl serine lipid mimics for identification of novel PSR
sized a set of phosphatidylserine lipid mimics with a PS lipid
proteins.
process, as only a handful of mutually exclusive receptors
have been found in different cell lines. Phosphatidylserine
affinity probes would provide an unbiased screen for cell-
surface and PS-opsonin receptors across all cell lines and
serum types; this would greatly facilitate advances in under-
standing phagocytosis and related biological processes.
Given the generality of PS-lipid-mediated engulfment, we
believe that PS recognition and the associated biological re-
sponse has been underexploited in biotechnology. PS re-
cognition could provide a cell-entry pathway of therapeutic
importance. Synthetic entry to PS-derivatized materials
would enable exploration along the lines of discovery bio-
logy through the identification of novel PS receptors
(PSRs). We present herein the synthesis of lipid affinity
probes that mimic phosphatidylserine lipid.
Our synthetic approach to mixed 1,2-O-diacyl-sn-glyc-
erol PS lipid photoaffinity probes expands on procedures
for the preparation of phosphoinositol lipid derivatives pre-
viously reported by Prestwich[9] and utilizes the concept of
trifunctional photocross-linking probes developed for ac-
tivity-based protein profiling by Cravatt and co-
workers.[10,11] A protected phosphatidylserine glycerol core
is first synthesized with a protected terminal amino group
on one of the two lipid acyl chains. The PS glycerol core is
then acylated to install a benzophenone photoaffinity mod-
ule and an alkyne group. This general strategy was used
to prepare the set of lipid probes shown in Figure 1. The
phosphatidylserine headgroup is envisioned to bind directly
to the PSRs, which are then captured by photoaffinity label-
ing with the benzophenone group; this labeled PSR is then
identified on a cell lysate gel through in-gel reaction of a
fluorophore azide with the probe alkyne by using azide–
alkyne [2+3] cycloaddition. The alkyne permits post-label-
Introduction
Engulfment of apoptotic cells during tissue remodeling
and concomitant suppression of inflammation are complex
processes that are still not completely understood.[1,2]
A
lynchpin event in the clearance of apoptotic and necrotic
cells is phosphatidylserine (PS) lipid recognition. PS lipids
are primarily found on intracellular membrane surfaces,
modulating electrostatic targeting of proteins to the mem-
brane.[3,4] This membrane asymmetry is lost when the cell
is activated or apoptotic; surface presentation of PS lipid
on apoptotic cells provides a recognition trigger for en-
gulfment by neighboring cells and professional phagocytes
such as macrophages during the course of tissue remodel-
ing. As receptors that directly bind to PS have only recently
been found and validated,[5–8] opsonin receptors[2] have
been the subject of the majority of studies on PS-triggered
phagocytosis. Three macrophage opsonin receptors have
been well studied and have been shown to be important in
some, but not all, cell clearance processes, leaving open the
question of how the remaining processes are mediated. The
distinct PS lipid recognition proteins implicated in these fin-
dings suggest redundant and compensatory roles in apop-
totic cell clearance to ensure stable homeostatic cell popula-
tions. We hypothesize that there are many more unknown
PS opsonizing proteins as well as tissue-specific receptors.
Discovery of the key players in these early stages of phago-
cytosis remains a limiting step in understanding this critical
[a] Department of Chemistry, The Ohio State University,
Columbus, Ohio 43210, USA
E-mail: bong@chem.osu.edu
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201001264.
Eur. J. Org. Chem. 2011, 751–758
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
751

S. Bandyopadhyay, D. Bong
FULL PAPER
Figure 1. Phosphatidylserine lipid mimics synthesized in this study. Compounds with the natural glycerol configuration are denoted with
the suffix “a” (1a, 2a), whereas compounds with the nonnative configuration at glycerol are denoted with the suffix “b” (1–6b).
ing readout of captured lipid-binding protein targets, through enzymatic and synthetic manipulation of lysophos-
thereby excluding contributions of fluorophore or affinity phatidylcholine lipid, an expensive starting material that re-
tag binding to the labeling process. PS lipid mimics 4b and stricts modification to the sn-2 position.[32] Glycidyl deriva-
5b were also prepared for use as competitive photocross- tives have also been used as a starting point to ether phos-
linking probes without alkyne readout; 6b was prepared for pholipids through Lewis acid catalyzed epoxide ring open-
use as a soluble PS lipid competitive binding inhibitor to ing;[12–14] addition to the epoxide with a long-chain carbox-
assay for PS probe binding selectivity (Figure 1). As PSRs ylic acid may yield phospholipid diesters.[33] To our surprise,
important in cell clearance are expected to be surface-ex- regioselective stepwise acylation[34,35] of monoprotected
pressed, cellular uptake is not an issue. Therefore, photoaf- glycerol, a robust approach that has been widely used to
finity tagging by these PS lipid probes may be performed prepare mixed 1,2-O-diacylglycerol phospholipids, has not
on intact, living cells as well as with fractionated cell lysates, been previously used to prepare phosphatidylserine lipid de-
followed by in-gel readout of the labeled fraction by using rivatives. We describe herein our application of this known,
azide–alkyne cycloaddition with a fluorophore azide. To versatile, and scalable strategy for lipid synthesis to produce
date, there have been relatively few phosphatidylserine lipid new PS lipid affinity probes for discovery biology (Fig-
syntheses reported,[12–16] and while the general synthesis ure 1).
and application of lipid photoaffinity probes to identify
biological partners has been well established,[17–30] none of
Results and Discussion
these prior efforts are focused on the use of PS lipid mimics
as affinity capture reagents for PS binding partners.[26] Pho-
toaffinity probes based on PS lipids have been previously
reported by using azido-photoactivated cross-linking pre-
pared through enzyme-catalyzed choline–serine headgroup
Synthesis of a Selectively Protected Phosphatidylserine
Lipid Core
The regioselective glycerol acylation approach allows in-
exchange.[26,31] Phosphatidylserine lipid derivatives for de- stallation of the benzophenone photocross-linking module
livery and imaging applications have also been synthesized and the alkyne readout tag at either the sn-1 or sn-2 ester
752
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 751–758

Trifunctional Phosphatidylserine as Probes for Lipid-Binding Proteins
Scheme 1.
position. We designed a convergent synthetic scheme that in Scheme 2 was then subjected to CBz hydrogenolysis to
begins with protected glycerol 18 (Scheme 1). Both enantio- free the primary amine at the acyl chain terminus. This re-
mers are commercially available, allowing ready access to sulted in partial cyanoethyl deprotection at the phosphorus
the natural lipid stereochemistry (i.e., 18a) as well as its center, which could be suppressed with formic acid during
stereoisomers. We prepared probes with native stereochem- hydrogenolysis or driven to completion with an excess
istry (i.e., 1a, 2a) as well as probes with the inverted glycerol amount of triethylamine (Scheme 3).[39] Acylation of the re-
center (i.e., 1b, 2b) to facilitate examination of stereoselec- sulting zwitterionic intermediate 13 with benzophenone
tivity of probe recognition; the routes to the diastereomers carboxylic acid derivatives gave isolated yields of 55–68%
were found to be identical in terms of isolated yield. Thus, of the serine-protected PS lipid probes. We suspect that
a series of known synthetic manipulations of d-(+)-1,2-iso- these modest yields are due to inefficient recovery from sil-
propylideneglycerol, which has the natural configuration at ica gel of the free phosphate intermediates, as purification
C2, led to 17a, an enantiopure glycerol nucleus[9] with p- on ion-exchange resin instead of silica afforded pure prod-
methoxybenzyl (PMB)-protection on the primary hydroxy ucts with higher recovery (Supporting Information). To
group, free secondary hydroxy group, and the 6(N-Cbz- broaden options for variation at one glycerol ester position,
amino)hexanoate primary ester (Scheme 1). DMAP-cata-
lyzed acylation of the secondary alcohol with hexanoyl
chloride or palmitoyl chloride yielded diacyl glycerol frame-
works for water-soluble and lipid-soluble probes, 1a and 2a,
respectively. Oxidative cleavage of the PMB protecting
group from water-soluble probe precursor 16a freed the pri-
mary hydroxy group of the mixed 1,2-O-diacyl-sn-glycerols
for O–P coupling with preassembled serine phosphoramid-
ite 11 (Scheme 2).[9,36–38] Intermediate 11 was derived by
condensation of Boc-Ser-OtBu and 2-cyanoethyl-N,N-di-
isopropylchlorophosphoramidite, yielding a mixture of epi-
mers at the phosphorus center. Coupling of 11 with glycerol
16a generated the serine glyceryl phosphite, which was oxid-
ized in situ to obtain protected phosphate 14a as a mixture
of diastereomers at the chiral phosphorus center, which
were distinguishable by 31P NMR spectroscopy as a pair of
singlets at –1.6 and –1.7 ppm. The fully constructed and
differentially protected phosphatidylserine nucleus formed Scheme 2.
Scheme 3. Identical separate routes employed for either glycerol stereoisomer at the chiral center marked with *. Structure is shown with
native configuration as found in 1a, 13a, 14a, and 15a, whereas 1b, 4b, 5b, 15b, 21b, and 22b have inverted stereochemistry. 12 =
HO2CC6H4COC6H4-R.
Eur. J. Org. Chem. 2011, 751–758
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
753

S. Bandyopadhyay, D. Bong
FULL PAPER
we combined benzophenone and the alkyne in one module derivatives with straight chain saturated acylation at the
by installation of a propargyl ether substituent the on central glycerol position, these procedures could be used to
benzophenone carboxylic acid (compound 12a, Scheme 4); install other lipid anchors with different biophysical proper-
this approach potentially allows wide variation at the re- ties, such as sterol and sphingosine anchors.[40–42] Probes
maining ester position. Serine deprotection with trifluoro- were prepared with the native glycerol configuration (i.e.,
acetic acid then yielded the free lipid probes. Using this 1a, 2a) as well as probes with inverted glycerol stereochem-
straightforward acylation approach, PS lipid derivatives istry (i.e., 1b, 2b, 4b, and 5b). Compounds with nonnative
were constructed with benzophenone and alkyne groups for glycerol stereochemistry were prepared to test whether or
cross-linking and post-labeling readout. Water-soluble and not native stereochemistry was required at both serine and
lipid-soluble PS lipid probe derivatives were prepared sim- glycerol chiral centers for recognition and function.
ply by varying the chain length of the fatty acid ester at the
While the combination of photocross-linking and read-
2-glycerol position; hexanoates 1a and 1b were water-solu- out at one position is a versatile design, this may also result
ble and nonaggregated, whereas palmitate derivatives 2a in greater steric barriers to labeling and readout. To address
and 2b were not water soluble but partitioned readily into this possibility, we also prepared PS probe molecules with
lipid vesicles. Probes were incubated with lipid vesicles then photolabeling and readout moieties separately displayed at
sedimented at 105 g; 1a concentration in the supernatant two different ester positions (Scheme 5). This approach is
was unchanged as judged by benzophenone absorption, essentially the same, with the exception that Cbz protection
whereas 2a was sedimented completely with the lipid pellet. is cleaved prior to acylation at sn-2, allowing regioselective
Lipid-soluble probes may be desirable to mimic the context amidation with benzophenone carboxylic acid, followed by
in which PSRs for cell-clearance encounter PS lipid on the esterification with hexynoic acid. Reaction with phos-
surfaces of apoptotic cells. Though we have only prepared phoramidite 11 and global deprotection with an base/acid
sequence then yields probe 3b with photoaffinity and al-
kyne tags at separate glycerol positions.
Biotin- and Fluorescein-Derivatized Phosphatidylserine
Lipid Mimics
Lipid-based probes have potential utility as intracellular
probes.[43,44] Derivatization of PS lipid mimics with affinity
and fluorescence tags may also be useful in target en-
richment and diagnostic applications. Intermediate 13b was
subjected to Cbz hydrogenolysis and directly treated with
fluorescein isothiocyanate (FITC) to yield fluorescein deriv-
ative 7b (Scheme 6); this compound may be useful to follow
PSR cellular localization[32] or as a reagent to screen chemi-
Scheme 4.
cal libraries for PS lipid binding elements.
Scheme 5. R1 = C6H4COC6H5, R2 = (CH2)3CCH. Intermediate 17b is the enantiomer of 17a in Scheme 2.
754 Eur. J. Org. Chem. 2011, 751–758
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Trifunctional Phosphatidylserine as Probes for Lipid-Binding Proteins
coupling FITC with 2-azidoethylamine. Biotin azide deriva-
tives were similarly prepared through acylation of amino
azide linkers with biotin active esters to facilitate isolation
and enrichment of labeled species on an avidin support; a
biotin thiolester azide derivative was also prepared to in-
crease the efficiency of elution of support-bound proteins
through thiolester cleavage. Under known copper-catalyzed
azide–alkyne [2+3] cycloaddition conditions in aqueous sol-
vent, each of these azides coupled cleanly to soluble probe
1b within 2 h, as indicated by HPLC and mass spectromet-
ric analysis (Scheme 7). These validated molecular tools are
currently being used in our search for new cell-surface
PSRs.
Scheme 6.
Probe Readout by [2+3] Azide–Alkyne Cycloaddition
Protein-Labeling with Soluble Benzophenone–Alkyne-
Functionalized PS Lipid Probes
Photoaffinity probe derivatization with fluorescent and
enrichment tags is of particular importance to their utility
Though propargyloxybenzophenone carboxylic acid 12a
in discovery biology, as the chemistry for readout of affinity has not been previously explored as a photoaffinity tag for
labeling of cell lysate fractions and subsequent target identi- protein labeling, it is well suited for the trifunctional probe
fication must be robust. While [2+3] azide–alkyne click cy- strategy described by Cravatt and co-workers and is readily
cloaddition has been used extensively for this pur- prepared by using adaptations of known procedures
pose,[32,45–49] the propargyloxybenzophenone probe module (Scheme 4).[50] This is conceptually related to the heterobi-
has not been previously validated as an efficient substrate. functional cross-linking maleimidobenzophenone affinity
To address this issue, we carried out the desired readout probes that combine targeted benzophenone photocross-
and isolation reactions on soluble probe 1b (Scheme 7).
linking with thiol capture on a pendant maleimide group.[51]
In a typical protein-labeling procedure based on pub- Replacement of the maleimide with alkyne results in im-
lished protocols with benzophenone photoaffinity proved chemoselectivity and probe stability. In preliminary
probes,[10,11] the PS probe is incubated with protein or a studies, we examined the ability of soluble PS probes 1a
mixture of proteins and then photocross-linked with UV and 1b to label prothrombin-1 (PT-1), which has a known
light at 0 °C, followed by readout by “click” [2+3] azide– affinity for PS lipids.[52] Soluble PS probe 1a has native
alkyne cycloaddition with an azide-derivatized fluorophore. stereochemistry, whereas 1b has inverted stereochemistry at
Denaturing polyacrylamide gel electrophoresis (SDS- the glycerol center and native l-serine configuration, al-
PAGE) is then used to visualize the profile of probe binding lowing isolation of the effect of phosphoserine recognition
propensities within the protein mixture as a function of from the glycerol backbone. Our studies revealed that 1a
fluorescence intensity of each protein band on SDS-PAGE. and 1b both labeled PT-1 with comparable efficiency in a
We set out to confirm that click functionalization would concentration-dependent manner, as judged by in-gel fluo-
proceed smoothly with relevant azide derivatives under rescence intensity of the protein band on SDS-PAGE (Fig-
these conditions. Both rhodamine azide and fluorescein az- ure 2), whereas no labeling was detected with heat-dena-
ide were prepared: acylation of 3-azidopropylamine with N- tured PT-1 protein. These preliminary labeling studies indi-
hydroxysuccinimidyl ester of rhodamine furnished rhod- cate that protein-labeling with our synthetic probes is in-
amine azide, whereas fluorescein azide was prepared by deed possible under standard photolabeling conditions and
Scheme 7. Conditions a: CuSO4, sodium ascorbate, water; conditions b: CuSO4, TCEP, TBTA, water.
Eur. J. Org. Chem. 2011, 751–758
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
755

S. Bandyopadhyay, D. Bong
FULL PAPER
2 H, iPr-CH), 2.64 (m, 2 H, CH2CN), 1.48 (br., 9 H, CO2tBu-
CH3), 1.46 (br., 4.5 H, Boc-tBu-CH3), 1.45 (br., 4.5 H, Boc-tBu-
CH3), 1.18 (br., 12 H, iPr-CH3) ppm. 31P NMR (161 MHz, CDCl3,
25 °C): δ = 149.4, 149.2 ppm. 13C NMR (100 MHz, CDCl3, 25 °C):
that the glycerol stereochemistry may not be generally re-
quired for lipid recognition, though further examination is
necessary. Exploration of differential labeling profile of this
set of PS-lipid photoaffinity probes in cell lysates of phago-
cytic and nonphagocytic cell lines is currently underway.
δ = 169.2, 155.1, 155.0, 117.5, 81.9, 81.7, 79.4, 64.4 (d, JCP
=
16 Hz), 63.9 (d, JCP = 17 Hz), 58.4 (d, JCP = 18 Hz), 58.2 (d, JCP
= 18 Hz), 55.0 (d, JCP = 7.6 Hz), 54.9 (d, JCP = 8.0 Hz), 43.0 (d,
JCP = 4.1 Hz), 42.8 (d, JCP = 4.1 Hz), 28.2, 27.8, 24.4 (d, JCP
=
7.4 Hz), 20.2 (d, JCP = 6.2 Hz), 20.1 (d, JCP = 6.8 Hz) ppm. HRMS
(ESI): calcd. for C21H40N3O6PNa [M + Na]+ 484.2553; found
484.2555.
Procedure for the Conjugation of 11 with Diacyl Glycerol and Subse-
quent Oxidation (14a): A solution of primary alcohol 16a (0.28 g,
0.64 mmol) and 11 (0.36 g, 0.77 mmol) in anhydrous CH2Cl2
(10 mL) was treated with a solution of 4,5-dicyanoimidazole (DCI;
0.22 g, 1.93 mmol) in anhydrous CH3CN (10 mL). The reaction
was run under a nitrogen atmosphere and judged to be complete
after 3 h by 31P NMR spectroscopy. The reaction mixture was co-
Figure 2. In-gel fluorescent readout of probe-labeled PT-1 using 1a
and 1b and rhodamine azide at the indicated probe concentrations.
Conclusions
We have demonstrated here a straightforward and flexi- oled in an ice bath and subjected to dropwise addition of a solution
of tert-butylhydrogen peroxide (≈5.5 m in decane, 1.42 mL,
3.85 mmol) then warmed to room temperature over 2 h. Complete
oxidation was indicated by 31P NMR spectroscopy, which revealed
ble synthesis of PS lipid photoaffinity probes to aid in the
search for novel cell-surface PSR proteins important in me-
diating cell clearance of apoptotic cells. Preliminary data
has been presented, which demonstrates the ability of these
probe molecules to react with lipid binding proteins. Given
the limited knowledge currently available regarding PSRs,
we anticipate that approaches of this type will be useful in
the elucidation of PSR biology, which has an impact on
cellular delivery strategies as well as control of inflam-
mation.
a
31P chemical shift change from δ = 139.2, 139.1 ppm for the in-
terconverting PIII diastereomers to δ = –1.61, –1.67 ppm for the
oxidized PV diastereomeric products. Dilution with CH2Cl2
(100 mL), acidic workup (3ϫ60 mL of 1 n HCl), followed by 10%
sodium sulfite wash (2ϫ40 mL), drying over anhydrous Na2SO4,
and concentration produced a crude product that was purified by
column chromatography (SiO2; EtOAc/hexanes, 7:3) to yield 14a
as a colorless oil (0.48 g, 93%).
Procedure for Protein Photolabeling and Fluorescence Readout by
Click Conjugation: Human prothrombin-1 (PT-1) solutions (44 μL
of 1.2 mg/mL stock solution in HBSS buffer with 50 μm CaSO4)
were incubated with 4.9 μm PS probe (1 μL of 215 μm stock in
HBSS) for 15 min then cooled with an ice bath and irradiated with
a long-wave UV lamp (centered at 365 nm) for 1 h. After irradia-
tion, each protein sample was treated with rhodamine azide (final
concentration 44 μm, from 40ϫ DMSO stock solution) followed by
a freshly prepared catalyst cocktail of TCEP (aqueous stock solu-
tion), TBTA (DMSO/tBuOH, 1:4 stock solution) and CuSO4
(aqueous stock solution) to yield final concentrations of 1 mm,
100 μm, and 1 mm, respectively. All stock solutions and reaction
volumes were prepared with degassed solvents. The resulting reac-
tion mixture was briefly vortexed and incubated at room tempera-
ture for 1 h, and then a 10 μL aliquot was collected and quenched
with the addition of 6 μL standard 2ϫ SDS-PAGE loading buffer
(reducing) followed by heat denaturation of the protein sample for
10 min at 70 °C and analysis on SDS-PAGE (12 μg of protein/gel
lane). In-gel visualization was performed by using a Typhoon Trio
(GE Healthcare) laser-induced fluorescence scanner followed by
standard coomassie staining and destaining.
Experimental Section
General Procedure for the Purification of Probe Compounds: After
complete removal of the Boc and tert-butyl ester protecting groups
on the headgroup serine, the excess amount of trifluoroacetic acid
was removed under reduced pressure to yield the free phosphatidyl-
serine derivatives as oils. The oil was redissolved in a minimum
amount of methanol and precipitated with an excess amount of
diethyl ether to give white solids that were washed with diethyl
ether (3ϫ) to yield analytically pure PS lipid derivatives 1a–6b (Fig-
ure 1). Compounds 1a and 1b were further purified by RP-HPLC
(C18, gradient run from 0% solvent A to 70% solvent B over
20 min, with elution at 15 min; solvent A = 1% CH3CN in water,
1% TFA; solvent B = 10% CH3CN in water, 0.1% TFA). All inter-
mediates were purified by silica gel chromatography (unless specifi-
cally noted) and all final compounds were judged to be pure by
NMR and HRMS.
Procedure for Synthesis of Serine Phosphoramidite 11: To a solution
of Boc-Ser-OtBu (0.41 g, 1.57 mmol) in anhydrous CH2Cl2 (10 mL)
was added diisopropylethylamine (DIEA; 0.24 g, 1.88 mmol) fol-
lowed by 2-cyanoethyl N,N-diisopropylchlorophosphoramidite
(0.40 g, 1.73 mmol), and the resulting solution was stirred under
an atmosphere of nitrogen. The reaction was judged to be complete
after 2 h by TLC and 31P NMR spectroscopy, and the mixture was
diluted with CH2Cl2 (100 mL), washed with aqueous saturated
NaHCO3 (2ϫ60 mL), and dried with anhydrous Na2SO4. Concen-
tration under reduced pressure and purification by column
chromatography (SiO2; hexanes/EtOAc/Et3N, 65:34:1) yielded 11 as
a colorless oil (0.67 g, 92%). 1H NMR (400 MHz, CDCl3, 25 °C): δ
Phosphatidylserine Probe (1a): Intermediate 15a (0.19 g,
0.22 mmol) was dissolved in TFA (1 mL), and the mixture was
stirred for 4 h at room temperature. After completion of Boc and
tert-butyl deprotection, the excess amount of TFA was removed,
and the resulting phosphatidylserine was triturated with diethyl
1
ether to give 1a (0.16 g) as a white solid in quantitative yield. H
NMR (400 MHz, CD3OD, 25 °C): δ = 7.95 (d, JH,H = 8.5 Hz, 2 H,
3-ArCH), 7.82 (d, JH,H = 8.5 Hz, 2 H, 2-ArCH), 7.79 (d, JH,H
=
8.4 Hz, 2 H, 2Ј-ArCH), 7.12 (d, JH,H = 8.5 Hz, 2 H, 3Ј-ArCH),
= 5.45 (br., 0.5 H, N-H), 5.34 (br., 0.5 H, N-H), 4.30 (br., 1 H, 5.22 (m, 1 H, sn-2-CH), 4.86 (d, JH,H = 2.4 Hz, 2 H, OCH2C) 4.42
CH2α), 4.02 (m, 1 H, CH2β), 3.85 (m, 3 H, O-CH2, CH), 3.59 (m, (dd, JH,H = 3.72, 12 Hz, 1 H, CH), 4.28 (m, 3 H, CH, O-CH2),
756
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 751–758

Trifunctional Phosphatidylserine as Probes for Lipid-Binding Proteins
4.18 (m, 1 H, CH), 4.01 (m, 2 H, CH2), 3.41 [m, 2 H, C(O)NH- (100 MHz, CD3OD, 25 °C): δ = 197.9, 174.9, 174.8, 159.3, 158.9,
CH2], 3.02 (t, JH,H = 2.3 Hz, 1 H, CC-H), 2.35 (m, 4 H, CO-CH2),
141.4, 139.5, 138.5, 134.3, 131.2, 131.1, 129.8, 128.5, 117.6, 114.8,
1.74–1.54 (m, 6 H, CH2), 1.44 (m, 2 H, CH2), 1.31 (m, 4 H, CH2), 71.8 (br.), 65.6, 63.7, 54.6, 54.3, 41.1, 35.1, 34.9, 32.5, 30.2, 27.6,
0.89 (t, JH,H = 6.6 Hz, 3 H, CH3) ppm. 13C NMR (100 MHz,
CD3OD, 25 °C): δ = 197.1, 175.3, 175.1, 169.7, 163.7, 142.4, 139.5,
134.0, 131.9, 131.2, 128.8, 116.3, 79.6, 78.0, 72.2 (d, JC,P = 7.5 Hz),
65.6 (d, JC,P = 5.1 Hz), 64.8 (d, JC,P = 4.8 Hz), 64.0, 57.4, 55.24 (d,
JC,P = 7.3 Hz), 41.4, 35.5, 35.2, 32.8, 30.6, 28.0, 26.1, 23.9,
14.8 ppm. HRMS (ESI): calcd. for C35H45N2O13PNa [M + Na]+
755.2551; found 755.2515. Probe 1b is indistinguishable from 1a
25.8, 25.7, 23.5, 14.4 ppm. HRMS (ESI): calcd. for C32H42N2O12P
[M – H]– 677.2475; found 677.2453.
Phosphatidylserine Probe (5b): Water-soluble phosphatidylserine
probe 5b (0.19 g) was quantitatively obtained from 22b (0.23 g,
Supporting Information) following a procedure analogous to that
1
described for the conversion of 15a into 1a. H NMR (400 MHz,
CD3OD, 25 °C): δ = 7.94 (d, JH,H = 8.2 Hz, 2 H, 3-ArCH), 7.80
(d, JH,H = 8.8 Hz, 2 H, 2-ArCH), 7.77 (d, JH,H = 8.2 Hz, 2 H, 2Ј-
ArCH), 7.05 (d, JH,H = 8.8 Hz, 2 H, 3Ј-ArCH), 5.23 (m, 1 H, CH),
4.41 (dd, JH,H = 3.9, 12.2 Hz, 1 H, CH), 4.31 (m, 3 H, CH, O-
CH2), 4.19 (dd, JH,H = 6.4, 12.0 Hz, 1 H, CH), 4.05 (m, 2 H, CH2),
3.89 (s, 3 H, O-CH3), 3.41 [m, 2 H, C(O)NH-CH2], 2.35 [m, 4 H,
C(O)CH2], 1.71–1.56 (m, 6 H, CH2), 1.44 (m, 2 H, CH2), 1.31 (m,
4 H, CH2), 0.89 (t, JH,H = 6.9 Hz, 3 H, CH3) ppm. 13C NMR
(100 MHz, CD3OD, 25 °C): δ = 195.3, 173.3, 167.8, 167.7, 164.0,
1
with respect to HRMS and the H and 13C NMR spectra.
Phosphatidylserine Probe (2a): Lipid-soluble phosphatidylserine
probe 2a (0.15 g) was quantitatively obtained from 42 (0.18 g, Sup-
porting Information) following a procedure analogous to that de-
scribed for the conversion of 15a into 1a. 1H NMR (400 MHz,
CD3OD, 25 °C): δ = 7.95 (d, JH,H = 8.6 Hz, 2 H, 3-ArCH), 7.81
(d, JH,H = 8.6 Hz, 2 H, 2-ArCH), 7.79 (d, JH,H = 8.4 Hz, 2 H, 2Ј-
ArCH), 7.12 (d, JH,H = 8.6 Hz, 2 H, 3Ј-ArCH), 5.23 (m, 1 H, CH),
4.85 (d, JH,H = 2.3 Hz, 2 H, O-CH2C), 4.41 (dd, JH,H = 3.5, 12 Hz,
1 H, CH), 4.28 (m, 3 H, CH, O-CH2), 4.18 (dd, JH,H = 6.1, 12 Hz,
1 H, CH), 4.02 (m, 2 H, CH2), 3.42 [m, 2 H, C(O)NH-CH2], 3.00
(t, JH,H = 2.1 Hz, 1 H, CC-H), 2.37 [t, JH,H = 7.0 Hz, 2 H, C(O)-
CH2], 2.34 [t, JH,H = 7.0 Hz, 2 H, C(O)CH2], 1.73–1.54 (m, 6 H,
140.7, 137.5, 132.3, 129.3, 129.2, 126.9, 113.5, 70.2 (d, JC,P
7.7 Hz), 64.0, 63.2 (d, JC,P = 5.1 Hz), 62.0, 54.8, 53.3 (d, JC,P
=
=
6.8 Hz), 39.6, 33.6, 33.3, 30.9, 28.7, 26.1, 24.3, 24.2, 22.0, 12.9 ppm.
HRMS (ESI): calcd. for C33H44N2O13P [M – H]– 707.2586; found
707.2551.
CH2), 1.44 (m, 2 H, CH2), 1.26 (br., 24 H, CH2), 0.88 (t, JH,H
=
Phosphatidylserine Mimic (6b): Water-soluble phosphatidylserine
probe 6b (0.31 g) was quantitatively obtained from 37 (0.43 g, Sup-
porting Information) following a procedure analogous to that de-
scribed for the conversion of 15a into 1a. 1H NMR (400 MHz,
CD3OD, 25 °C): δ = 5.23 (m, 1 H, CH), 4.41 (dd, JH,H = 3.6, 12 Hz,
1 H, CH), 4.28 (m, 3 H, CH, O-CH2), 4.20 (dd, JH,H = 6.4, 12 Hz,
1 H, CH), 4.00 (m, 2 H, CH2), 2.32 (m, 4 H, CO-CH2), 1.64 (m, 4
H, CH2), 0.95 (m, 6 H, CH3) ppm. 13C NMR (100 MHz, CD3OD,
25 °C): δ = 175.3, 175.0, 169.7, 72.2 (d, JC,P = 7.1 Hz), 65.6 (d, JC,P
= 5.0 Hz), 64.8 (d, JC,P = 5.5 Hz), 64.0, 54.9 (d, JC,P = 7.2 Hz),
37.4, 37.2, 19.5, 14.4 ppm. HRMS (ESI): calcd. for C14H25NO10P
[M – H]– 398.1222; found 398.1217.
6.7 Hz, 3 H, CH3) ppm. 13C NMR (100 MHz, CD3OD, 25 °C): δ
= 197.1, 175.3, 175.2, 169.6, 163.7, 142.3, 139.5, 134.0, 131.9, 131.2,
128.9, 116.3, 79.5, 78.0, 72.2 (d, JC,P = 7.9 Hz), 65.7 (d, JC,P
=
5.4 Hz), 64.9 (d, JC,P = 4.8 Hz), 64.0, 61.2, (possible impurity) 57.4,
55.2 (d, JC,P = 6.9 Hz), 41.5, 35.6, 35.2, 33.5, 31.3 (br.), 31.2, 31.1,
31.0, 30.7, 30.6, 28.0, 26.5, 26.1, 24.2, 14.9 ppm. HRMS (ESI):
calcd. for C45H64N2O13P [M – H]– 871.4146; found 871.4122. Probe
1
2b is indistinguishable from 2a with respect to HRMS and the H
and 13C NMR spectra.
Phosphatidylserine Probe (3b): Lipid-soluble phosphatidylserine
probe 3b (0.17 g) was quantitatively obtained from 23 (0.22 g, Sup-
porting Information) following a procedure analogous to that de-
scribed the conversion of 15a into 1a. 1H NMR (400 MHz,
CD3OD, 25 °C): δ = 7.95 (m, 2 H, 3-ArCH), 7.83 (d, JH,H = 8.3 Hz,
2 H, 2-ArCH), 7.78 (d, JH,H = 8.2 Hz, 2 H, 2Ј-ArCH), 7.66 (t, JH,H
= 7.4 Hz, 1 H, 4Ј-ArCH), 7.54 (m, 2 H, 3Ј-Ar-CH), 5.24 (m, 1 H,
CH), 4.42 (dd, JH,H = 3.7, 12 Hz, 1 H, CH), 4.28 (m, 3 H), 4.19
(dd, JH,H = 6.1, 12 Hz, 1 H, CH), 4.02 (m, 2 H, CH2), 3.42 [m, 2
H, C(O)NH-CH2], 2.48 (m, 2 H, C-CH2), 2.38 [m, 2 H, C(O)-CH2],
2.25 [m, 2 H, C(O)-CH2], 2.23 (t, JH,H = 1.7 Hz, 1 H, CC-H), 1.80
(m, 2 H, CH2), 1.67 (m, 4 H, CH2), 1.44 (m, 2 H, CH2) ppm. 13C
NMR (100 MHz, CD3OD, 25 °C): δ = 196.3, 173.5, 172.6, 167.8,
139.9, 138.0, 137.0, 132.8, 129.7, 129.6, 128.3, 127.0, 82.7, 70.4
(br.), 69.0, 63.8, 63.0 (br.), 62.0, 59.4, 53.4 (br.), 39.6, 33.3, 32.3,
28.7, 26.1, 24.2, 23.5, 17.0 ppm. HRMS (ESI): calcd. for
C32H38N2O12P [M – H]– 673.2162; found 673.2141.
Phosphatidylserine Probe (7b): Water-soluble phosphatidylserine
probe 7b (0.14 g) was quantitatively obtained from 29 (0.17 g, Sup-
porting Information) following a procedure analogous to that de-
scribed for the conversion of 15a into 1a. 1H NMR (400 MHz,
CD3OD, 25 °C): δ = 8.25 (s, 1 H, Ar-CH), 7.85 (d, JH,H = 8.4 Hz,
1 H, Ar-CH), 7.18 (d, JH,H = 8.4 Hz, 1 H, Ar-CH), 6.86 (d, JH,H
= 8.9 Hz, 2 H, Ar-CH), 6.80 (JH,H = 2.5 Hz, 2 H, Ar-CH), 6.67 (d,
JH,H = 8 Hz, 2 H, Ar-CH), 5.15 (m, 1 H, CH), 4.36 (dd, JH,H
=
3.7, 12 Hz, 1 H, CH), 4.21 (m, 4 H, CH, CH, O-CH2), 3.95 (m, 2
H, CH2), 3.56 [m, 2 H, C(S)NH-CH2], 2.29 (m, 4 H, CO-CH2),
1.67–1.49 (m, 5 H, CH2), 1.39 (m, 2 H, CH2), 1.24 (m, 5 H, CH2),
0.83 (t, JH,H = 6.6 Hz, 3 H, CH3) ppm. 13C NMR (100 MHz,
CD3OD, 25 °C): δ = 171.9, 171.8, 166.3, 152.9, 140.0, 128.4, 126.8,
124.2, 124.1, 112.6, 110.4, 100.6, 68.8 (d, JC,P = 7.0 Hz), 62.2 (br.),
61.4, 60.4, 51.9 (d, JC,P = 5.0 Hz), 42.5, 32.0, 31.9, 29.4, 27.8, 26.7,
24.6, 22.8, 22.7, 20.5, 11.4 ppm. HRMS (ESI-TOF): calcd. for
C39H45N3O15PS [M – H]– 858.2314; found 858.2318.
Phosphatidylserine Probe (4b): Lipid-soluble phosphatidylserine
probe 4b (0.13 g) was quantitatively obtained from 21b (0.16 g,
Supporting Information) following a procedure analogous to that
Supporting Information (see footnote on the first page of this arti-
cle): Detailed experimental procedures, mass spectrometric data,
and 1H and 13C NMR spectral characterization of all new com-
pounds and synthetic intermediates.
1
described for the conversion of 15a into 1a. H NMR (400 MHz,
CD3OD, 25 °C): δ = 7.95 (d, JH,H = 8.1 Hz, 2 H, 3-ArCH), 7.83
(d, JH,H = 8.1 Hz, 2 H, 2-ArCH), 7.79 (d, JH,H = 7.9 Hz, 2 H, 2Ј-
ArCH), 7.66 (t, JH,H = 7.6 Hz, 1 H, 4Ј-ArCH), 7.54 (m, 2 H, 3Ј-
ArCH), 5.22 (m, 1 H, CH), 4.41 (m, 1 H, CH), 4.30 (m, 4 H, CH,
CH, O-CH2), 4.01 (m, 2 H, CH2), 3.41 [m, 2 H, C(O)NH-CH2],
2.37 [t, JH,H = 7.0 Hz, 2 H, C(O)-CH2], 2.33 [t, JH,H = 6.9 Hz, 2
Acknowledgments
H, C(O)-CH2], 1.64 (m, 6 H, CH2), 1.44 (m, 2 H, CH2), 1.31 (m, We thank Dr. Dan Carper for assistance with mass spectroscopy.
4 H, CH2), 0.89 (t, JH,H = 6.5 Hz, 3 H, CH3) ppm. 13C NMR
This research was supported in part by the Institute of Materials
Eur. J. Org. Chem. 2011, 751–758
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
757

S. Bandyopadhyay, D. Bong
FULL PAPER
[24] G. D. Prestwich, G. Dorman, J. T. Elliott, D. M. Marecak, A.
Chaudhary, Photochem. Photobiol. 1997, 65, 222.
[25] G. Dorman, G. D. Prestwich, Trends Biotechnol. 2000, 18, 64.
[26] M. P. Blanton, H. H. Wang, Biochemistry 1990, 29, 1186.
[27] S. K. Arora, E. J. Kattelman, C. T. Lim, G. C. Le Breton, D. L.
Venton, J. Med. Chem. 1987, 30, 918.
[28] D. S. Middlemas, M. A. Raftery, Biochemistry 1987, 26, 1219.
[29] K. B. Delclos, E. Yeh, P. M. Blumberg, Proc. Natl. Acad. Sci.
USA 1983, 80, 3054.
Research at the Ohio State University and a National Science
Foundation – CAREER award to D. B. We thank the Ohio Bio-
products Innovation Center for instrumentation support.
[1] L. P. Erwig, P. M. Henson, Cell Death Differ. 2008, 15, 243.
[2] G. Lemke, C. V. Rothlin, Nat. Rev. Immunol. 2008, 8, 327.
[3] T. Yeung, M. Terebiznik, L. Yu, J. Silvius, W. M. Abidi, M.
Philips, T. Levine, A. Kapus, S. Grinstein, Science 2006, 313,
347.
[4] T. Yeung, G. E. Gilbert, J. Shi, J. Silvius, A. Kapus, S.
Grinstein, Science 2008, 319, 210.
[5] D. Park, A. C. Tosello-Trampont, M. R. Elliott, M. Lu, L. B.
Haney, Z. Ma, A. L. Klibanov, J. W. Mandell, K. S. Ravichan-
dran, Nature 2007, 450, 430.
[6] D. Park, A. Hochreiter-Hufford, K. S. Ravichandran, Curr.
Biol. 2009, 19, 346.
[7] M. Miyanishi, K. Tada, M. Koike, Y. Uchiyama, T. Kitamura,
S. Nagata, Nature 2007, 450, 435.
[8] S. Y. Park, M. Y. Jung, H. J. Kim, S. J. Lee, S. Y. Kim, B. H.
Lee, T. H. Kwon, R. W. Park, I. S. Kim, Cell Death Differ.
2008, 15, 192.
[30] P. Z. Wang, D. H. Blank, T. A. Spencer, J. Org. Chem. 2004,
69, 2693.
[31] A. J. Schroit, J. Madsen, A. E. Ruoho, Biochemistry 1987, 26,
1812.
[32] A. J. Lampkins, E. J. O’Neil, B. D. Smith, J. Org. Chem. 2008,
73, 6053.
[33] J. Lindberg, J. Ekeroth, P. Konradsson, J. Org. Chem. 2002, 67,
194.
[34] S. F. Martin, J. A. Josey, Tetrahedron Lett. 1988, 29, 3631.
[35] S. F. Martin, J. A. Josey, Y.-L. Wong, D. W. Dean, J. Org.
Chem. 1994, 59, 4805.
[36] U. Schlueter, J. Lu, B. Fraser-Reid, Org. Lett. 2003, 5, 255.
[37] J. Chen, L. Feng, G. D. Prestwich, J. Org. Chem. 1998, 63,
6511.
[38] Q.-M. Gu, G. D. Prestwich, J. Org. Chem. 1996, 61, 8642.
[39] H. M. Hsiung, Tetrahedron Lett. 1982, 23, 5119.
[40] S. L. Veatch, S. L. Keller, Biophys. J. 2003, 85, 3074.
[41] M. E. Beattie, S. L. Veatch, B. L. Stottrup, S. L. Keller, Bio-
phys. J. 2005, 89, 1760.
[42] B. L. Stottrup, S. L. Keller, Biophys. J. 2006, 90, 3176.
[43] A. B. Neef, C. Schultz, Angew. Chem. Int. Ed. 2009, 48, 1498.
[44] C. Y. Jao, M. Roth, R. Welti, A. Salic, Proc. Natl. Acad. Sci.
USA 2009, 106, 15332.
[9]
J. Chen, A. A. Profit, G. D. Prestwich, J. Org. Chem. 1996, 61,
6305.
[10]
[11]
G. C. Adam, E. J. Sorensen, B. F. Cravatt, Mol. Cell. Proteom-
ics 2002, 1, 828.
C. M. Salisbury, B. F. Cravatt, J. Am. Chem. Soc. 2008, 130,
2184.
[12] A. B. Kazi, J. Hadju, Tetrahedron Lett. 1992, 33, 2291.
[13] A. B. Kazi, S. Shidmand, J. Hajdu, J. Org. Chem. 1999, 64,
9337.
[14] X. Qiu, S. Ong, C. Bernal, D. Rhee, C. Pidgeon, J. Org. Chem.
1994, 59, 537.
[15] J. M. Delfino, C. J. Stankovic, S. L. Schreiber, F. M. Richards,
Tetrahedron Lett. 1987, 28, 2323.
[45] H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew. Chem. Int.
Ed. 2001, 40, 2004.
[46] H. C. Kolb, K. B. Sharpless, Drug Discovery Today 2003, 8,
1128.
[16] W. M. Loffredo, M. D. Tsai, Bioorg. Chem. 1990, 18, 78.
[17] C. Diaz, K. Balasubramanian, A. J. Schroit, Bioconjugate
Chem. 1998, 9, 250.
[47] A. Krasinski, Z. Radic, R. Manetsch, J. Raushel, P. Taylor,
K. B. Sharpless, H. C. Kolb, J. Am. Chem. Soc. 2005, 127,
6686.
[48] A. E. Speers, G. C. Adam, B. F. Cravatt, J. Am. Chem. Soc.
2003, 125, 4686.
[49] M. D. Best, Biochemistry 2009, 48, 6571.
[50] O. I. A. Salem, M. Frotscher, C. Scherer, A. Neugebauer, K.
Biemel, M. Streiber, R. Maas, R. W. Hartmann, J. Med. Chem.
2006, 49, 748.
[18] M. D. Smith, C. G. Sudhahar, D. Gong, R. V. Stahelin, M. D.
Best, Mol. Biosyst. 2009, 5, 962.
[19] M. Hashimoto, Y. Hatanaka, Bioorg. Med. Chem. Lett. 2008,
18, 650.
[20] A. K. Hamouda, M. Sanghvi, D. C. Chiara, J. B. Cohen, M. P.
Blanton, Biochemistry 2007, 46, 13837.
[21] X. Lu, R. Bittman, J. Org. Chem. 2005, 70, 4746.
[22] C. Visintin, A. E. Aliev, D. Riddall, D. Baker, M. Okuyama,
P. M. Hoi, R. Hiley, D. L. Selwood, Org. Lett. 2005, 7, 1699.
[23] K. A. Chehade, K. Kiegiel, R. J. Isaacs, J. S. Pickett, K. E.
Bowers, C. A. Fierke, D. A. Andres, H. P. Spielmann, J. Am.
Chem. Soc. 2002, 124, 8206.
[51] G. Dorman, G. D. Prestwich, Biochemistry 1994, 33, 5661.
[52] M. Huang, A. C. Rigby, X. Morelli, M. A. Grant, G. Huang,
B. Furie, B. Seaton, B. C. Furie, Nat. Struct. Biol. 2003, 10,
751.
Received: September 8, 2010
Published Online: December 17, 2010
758
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 751–758