Towards a new SPE material for EDCs: Fully automated synthesis of a library of tripodal receptors followed by fast screening by affinity LC

Article abstract of DOI:10.1002/ejoc.200801290

A series of human estrogen receptor (hER) mimics were synthesised. On the basis of the knowledge on the structure of the hormone binding domain of the hER, three different peptide chains were constructed onto a tripodal scaffold. By using a fully automated solid-phase synthesis protocol, 120 members with known identity were synthcsised in only a week. Affinity towards estrogenic compounds was checked by affinity LC. For this purpose, ethinylestradiol was "clicked" onto a modified-silica phase, and the obtained material was packed into an HPLC column. The results stemming from the affinity LC experiments were confirmed by clicking one library member to silica and by using this solid phase to extract different endocrine-disrupting chemicals (EDCs) from aqueous media. Wiley-VCH Verlag GmbH & Co. KGaA.

Full text of DOI:10.1002/ejoc.200801290

FULL PAPER
DOI: 10.1002/ejoc.200801290
Towards a New SPE Material for EDCs: Fully Automated Synthesis of a
Library of Tripodal Receptors Followed by Fast Screening by Affinity LC
Steven E. Van der Plas,[a] Els Van Hoeck,[a] Fréderic Lynen,[a] Pat Sandra,[a] and
Annemieke Madder*[a]
Dedicated to Professor Alain Krief
Keywords: Receptors / Protein models / Solid-phase synthesis / Endocrine-disrupting chemicals / Combinatorial chemistry
A series of human estrogen receptor (hER) mimics were syn-
thesised. On the basis of the knowledge on the structure of
the hormone binding domain of the hER, three different pep-
tide chains were constructed onto a tripodal scaffold. By
using a fully automated solid-phase synthesis protocol, 120
members with known identity were synthesised in only a
week. Affinity towards estrogenic compounds was checked
by affinity LC. For this purpose, ethinylestradiol was
“clicked” onto a modified-silica phase, and the obtained ma-
terial was packed into an HPLC column. The results stem-
ming from the affinity LC experiments were confirmed by
clicking one library member to silica and by using this solid
phase to extract different endocrine-disrupting chemicals
(EDCs) from aqueous media.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2009)
which they are present complicate identification and quan-
tification of EDCs.[3] This problem can be tackled by a pre-
concentration step before the actual analysis. In general, so-
lid-phase extraction (SPE) of aqueous samples is a widely
applied technique to enrich pollutants.[4] With the current
SPE cartridges, the extraction efficiency of the pollutants
depends on their polarity. Polar compounds tend to stay in
the water sample, while apolar compounds are retained on
the solid phase, with poor selectivity of the currently ap-
plied SPE methods as a consequence. A solid-phase mate-
rial that could selectively bind EDCs would greatly simplify
the analysis.
Attempts to address this problem include the develop-
ment of immunoaffinity based systems and synthetic alter-
natives such as molecular imprinted polymers (MIPs) and
linear peptides. Polyclonal antibodies against estradiol, es-
trone and estriol were developed and coupled to Sepharose
to make an immunoaffinity cartridge that is able to pre-
concentrate the three estrogens from women’s urine.[5] Be-
sides the disadvantage of facing a time-consuming, trial-
and-error-based production process, the immunoaffinity
systems are generally too specific and do not allow the pre-
concentration of a whole range of compounds as in the case
of EDCs. Alternatively and more easily produced, molecu-
lar imprinted polymers have been developed by using estra-
diol and diethylstilbesterol as templates.[6] Though good
properties have been obtained in organic solvents, trapping
of analytes with MIPs remains difficult in aqueous media.[7]
Introduction
Worldwide concern has been growing on the increasing
distribution of endocrine-disrupting chemicals (EDCs) over
the last decade. The overall anxiety is caused by the effect
of these substances on the endocrine system of wildlife and
humans. For humans, the most disturbing effects, such as a
decrease in sperm count in men and the early sexual matu-
ration of women, are thought to originate from the binding
of chemicals with the human estrogen receptor (hER).[1]
There are numerous chemicals with estrogenic properties
present in the environment. These include synthetic and
natural hormones, industrial chemicals (insecticides, house-
hold detergents) and phyto-estrogens (plant derived). Struc-
tural inspection of this wide variety of chemicals shows that
a common feature of these estrogenic compounds is the
presence of an aromatic ring, which can be accommodated
through a pincer-like arrangement of amino acids in the
hormone binding cavity of the hER.[2] The establishment of
a causal relationship between the presence of EDCs in the
environment and their possible effects on human health is
a challenging quest. Besides their low physiologically active
concentrations, the complex environmental matrices in
[a] Department of Organic Chemistry, Ghent University,
Krijgslaan 281, 9000 Ghent, Belgium
Fax: +32-9-2644998
E-mail: Annemieke.Madder@UGent.be
Supporting information for this article is available on the
WWW under http://www.eurjoc.org/ or from the author.
1796
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1796–1805

A New Solid-Phase Extraction Material for Endocrine-Disrupting Chemicals
A remarkable contribution towards selective trapping of with estradiol, Arg/Glu/His were selected in order to mimic
EDCs has been made by Tozzi et al. who used small linear the hydrogen bonding at the extremities of the steroid. Ad-
peptides as recognition motifs for estrogenic chemicals.[8] ditionally, Phe/Leu/Met were chosen to provide an apolar
By using the amino acids present in the hormone-binding environment for the hydrophobic skeleton of the ligand.
domain of the human steroid binding protein as building
blocks, a parallel library of dipeptides was generated. After
incubation with a radioactively labelled estradiol ligand, the
dipeptide with the highest affinity towards estradiol was se-
lected and then used as a starting point for the generation
of larger peptides. Though peptides with up to 8 amino ac-
ids were synthesised, it seemed that the tetrapeptide Arg–
Ser–Ser–Val–OH showed the best binding properties, and a
pre-concentration column was constructed with this solid-
phase-bound peptide.
Inspired by these results, we decided to try to develop an
alternative synthetic estrogen receptor for extracting EDCs
from aqueous media. It was believed that an improved and
a more-selective binding of EDCs could be obtained if the
amino acids of the hormone-binding domain of the hER
were selected as building blocks for the synthetic receptors.
Moreover, instead of making linear peptides without a pre-
defined structure, tripodal scaffold 1 was chosen as the
backbone, while allowing the incorporation of 3 different
peptide chains. Earlier molecular modelling results have
indicated the potential of 1 to orient attached peptide
chains in a parallel way.[9] Previous examples in the area of
scaffolded peptides towards the selective binding of small
organic molecules or ions[10] and as mimics of peptidases[11]
show that by pre-organising functionalised groups good re-
sults can be obtained in terms of affinity and selectivity.
When using scaffolded peptides, generally, molecular di-
Figure 1. Schematic representation of (A) hydrogen-bonding net-
versity is obtained by synthesising split-and-mix libraries.
Screening can then be performed by adding a labelled
ligand and selecting the fluorescent or coloured
beads.[10b,10c,12] By using such a combinatorial approach, it
sometimes remains difficult to establish the identity of the
work and (B) hydrophobic interactions between the amino acids
(AAs) of the hormone-binding domain (HBD) of the hER and its
natural ligand estradiol.
The orthogonally protected tripodal scaffold was synthe-
active compound, a process that is often complicated by the sised and attached to a solid phase as described before.[16]
low amount of product synthesised.[13] These problems can A photocleavable linker[17] was first coupled with Tentagel,
be bypassed if a parallel strategy is employed. Indeed, in followed by the incorporation of a spacer, γ-aminobutyric
that case, the identity of the different library members is acid (GABA). The tripodal scaffold was then attached by
known and sufficient amounts can be synthesised for fur- using a PyBOP/DIEA coupling protocol to yield construct
ther, more elaborate screening.[14] This paper presents the 1b (see Scheme 1). The weak nucleophilicity of the aromatic
first fully automated synthesis of a library of tripodal struc- amines was then bypassed by incorporation of differently
tures and demonstrates that in a short period of time, a protected glycines through their symmetrical anhydrides.[18]
considerable amount of candidate receptors can be synthe- The protecting groups of these glycines were chosen in such
sised in a parallel way. Moreover, by using this automated a way that their deprotection could be performed on an
parallel approach, suitable quantities of material are within automated peptide synthesiser. Indeed, the removal of the
reach to allow extensive analysis of affinity properties.
acid labile Boc group and the Pd0 labile Alloc were not
compatible with the automated synthesiser present in our
laboratory.[19] In first instance, the 2-nitrobenzenesulfonyl
(oNBS) derivative of Gly was tested.[20] However, in our
hands upon test coupling of an amino acid to one of the
other scaffold “arms” with the use of PyBOP/DIEA, the
Results and Discussion
Synthesis
In order to mimic the hormone-binding domain of the oNBS sulfonamide linkage was partially acylated.
hER, the amino acids responsible for ligand binding were
Therefore, the Boc and the Alloc group were replaced
selected on the basis of X-ray structures of the hER with with 1-(4,4-dimethyl-2,6-dioxocyclohexylidene)ethyl (ivDde)[21]
estradiol.[15] The important binding interactions are sche- and the azide[22] group. The former is smoothly cleaved by
matically presented in Figure 1. From the different contacts a 2% hydrazine solution in DMF, and the latter can be
Eur. J. Org. Chem. 2009, 1796–1805
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1797

S. E. Van der Plas, E. Van Hoeck, F. Lynen, P. Sandra, A. Madder
FULL PAPER
ber was designed to contain all 6 different, selected AAs.
The size of the library was further substantially decreased
by applying some simplifications. It was decided to omit the
combinations in which one strand seems to be switched
with another strand. A second method to restrict the size
was to insert each AA only once into the tripodal structure.
In this way, a small-sized library of 120 members was de-
signed. Because the synthesiser was equipped with a reac-
tion block that could contain 24 reaction vessels, the whole
library was synthesised in 5 consecutive cycles.
Each cycle started by removing the Fmoc group, after
which two AAs were inserted by using HBTU and DIEA
(see Scheme 2). The first strand was terminated by capping
Scheme 1. Bypassing the weak nucleophilicity of the aromatic
amines by first incorporating glycine derivatives through their sym-
metrical anhydrides. FmocGlyOH, ivDdeGlyOH 2, and azidoacetic
acid (3) were attached.
reduced with Me3P in THF/H2O. As depicted in Scheme 1,
the Fmoc group of 1 was deprotected and subsequently
2 equiv. 2, synthesised from glycine and ivDdeOH, were ac-
tivated in the presence of 1 equiv. DIC and added to the
resin. After overnight reaction, the Boc group was depro-
tected and FmocGlyOH was coupled. Subsequently, the Al-
loc group was deprotected with Pd0 and PhSiH3,[23] and
azidoacetic acid (3), prepared from bromoacetic acid,[24]
was then coupled to give orthogonally protected 4. Con-
struct 4 was then used for the automated synthesis of a
parallel library of 120 members. When 6 AAs are used for
construction in 6 different positions, the size of the library
can rise to 46656 members. Though molecular diversity can
be generated with automated synthesis, it is practically im-
possible to synthesise a library of more then 46000 mem-
Scheme 2. Automated synthesis of 120 library members and subse-
quent deprotection of the side chain protecting groups including
bers. In order to constrain the size of the library, each mem- Met(O).
1798
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1796–1805

A New Solid-Phase Extraction Material for Endocrine-Disrupting Chemicals
with Ac2O. The second strand was accessed by removing
the ivDde group with a 2% hydrazine solution in DMF,
which could be added to the different reactors with a robot
arm. Again, two cycles of coupling and Fmoc deprotection
were followed by a capping step. The last peptide chain was
inserted after the azide was reduced. Reduction with Ph3P
resulted in the formation of the N–P ylide, but the sterically
less-hindered Me3P gave
a smooth conversion after
2ϫ10 min reaction time. For this reason, a 0.5  solution
of Me3P in THF/H2O (1:1) was added to the reactors
through the robot arm. This deprotection step was followed
by the coupling of the last two AAs, and the third strand
was finally terminated by capping with Ac2O. Each batch
of 24 members could be synthesised in 20 h, and conse-
quently, the whole library was synthesised in one week. The
purity of the crude products was easily checked by irradiat-
ing a small sample of the resin with light (λ = 360 nm) for
3 h in EtOH.[17] Analysis of the resulting solution by HPLC
showed the presence of two products. Further analysis with
MALDI-TOF indicated that partial oxidation of the Met
side chain had occurred during the synthesis, which ex-
plains the two different signals in the HPLC chromatogram.
This problem was solved during the manual side chain de-
protection step by adding Me2S and NH4I to the cleavage
cocktail. After 3 h, not only were the different side chain
protecting groups removed, but the sulfoxide was also re-
duced, to yield the desired library in good purity (see Sup-
porting Information for analytical details of library mem-
bers).
Screening
In order to screen the library for affinity towards estra-
diol, it was decided to develop an affinity LC based method
(Scheme 3). By immobilising estradiol on silica and by
using this material to pack an HPLC column, it should be
possible to identify the most-active receptor by monitoring
the retention times of the different library members. Indeed,
if a receptor forms a strong complex with the silica-bound
Scheme 3. Screening of the library in solution with an affinity-
based method. Two new silica materials 11 and 12 were prepared
by using “click” chemistry.
ligand, this receptor should be retained longer on the col- MeOH to target the hydrophobic interactions, the members
umn than weakly binding receptors. finally eluted. However, the compounds all eluted in a sim-
The affinity LC material was prepared by acylating ami- ilar time frame, which suggests that the different library
nopropylsilica (5 µm particle size) with azidohexanoic acid members interact in a similar manner with the silica-bound
9, which was prepared in a similar fashion to azidoacetic estradiol. The selectivity was tested by designing a non-es-
acid 3.[25] Ethinylestradiol was then attached to this mate- trogenic norethindrone column packed with material 12.
rial by using the CuI-mediated cycloaddition between an From these experiments, it became clear that nonselective
azide and an alkyne[26] to yield material 11. This material binding was observed, since all members interacted simi-
was subsequently packed into a column (2.1ϫ150 mm) that larly with the silica-bound norethindrone. In a first attempt
was then integrated into a LC-MS system. In this way, this to evaluate the obtained results, a competition experiment
technique has the additional advantage of confirming the was designed. The same LC-MS setup with the estradiol
identity of each member. Using an autosampler with the column 11 was envisaged, however, in the presence of in-
LC-MS system, the whole library can be screened in a few creasing amounts of estradiol in the eluent. Normally, a
days.
shift in the peaks can be expected because the receptors
Under isocratic conditions, with H2O at pH 7.4, none of should now also interact with the estradiol in solution. Un-
the members eluted from the column. In an effort to disrupt fortunately, the solubility of estradiol in MeOH was not
the hydrogen bonding, the pH was lowered to 3, but no sufficient to obtain the concentrations needed for the com-
improvement was observed. When H2O was replaced with petition experiments.
Eur. J. Org. Chem. 2009, 1796–1805
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1799

S. E. Van der Plas, E. Van Hoeck, F. Lynen, P. Sandra, A. Madder
FULL PAPER
In a second attempt to evaluate the obtained results and pounds, including EDCs, silica was used to immobilise the
as a proof of principle, we decided to construct an SPE library member. Direct synthesis on silica proved to be un-
cartridge consisting of a solid support loaded with one li- successful because cleavage of the side chain protecting
brary member. This cartridge could then be used to extract groups of the AAs was not complete. This is probably due
EDCs from an aqueous matrix, thus the efficiency of the to the oxygen-rich backbone of the silica material, which
synthesised receptor could be evaluated. Because Tentagel can be protonated by TFA, which lowers the effective [H+].
and other commercially available SPPS resins showed a too It was thus confirmed that silica is not a suitable support
high nonselective background absorption of apolar com- for the construction of these tripodal peptides, which is a
fact that is not entirely unexpected. Hence, an alkyne-con-
taining library member was synthesised on Tentagel, which
allows further attachment to azido-functionalised silica
through the CuI azide–alkyne cycloaddition. The synthesis
started from Tentagel modified with the photocleavable
linker to which FmocLys(Boc)OH was coupled (see
Scheme 4). The α-N atom was first deprotected and capped
with pentynoic acid. The side chain of Lys was sub-
sequently used as a spacer arm to couple the tripodal scaf-
fold with PyBOP and DIEA. After insertion of the 3 or-
thogonally protected glycines, construct 17 was used as
starting point for the automated synthesis. The obtained
protected peptide was then photocleaved from the resin, af-
ter which the side chain protecting groups were cleaved. The
resulting crude material was purified by reversed-phase pre-
parative HPLC to give library member 18 in a good purity
with an overall yield of 16%.
Next, this compound was “clicked” to azido-function-
alised silica 19 (50 µm particle size, loading 0.110 mmolg–1)
by using CuI as a catalyst. The reaction was monitored by
UV, and in this way, the amount of reacted compound 18
could be determined. A loading of 52 µmolg–1 for material
20 was then used. This material was incubated with an
aqueous solution of bisphenol-A (BPA), estradiol (E2), ethin-
ylestradiol (EE2), estrone, diethylstilbesterol (DES) and
testosterone (Tes). The concentrations of the analytes after
incubation were determined by HPLC-UV. These values
were correlated with the amount of EDC that was trapped
on the material.
Figure 2. The percentage of EDC trapped on aminopropylsilica
and material 20. While the silica retains little or no EDCs, an in-
crease is clearly obtained for material 20. However, the silica-bound
receptor is not able to differentiate between estrogenic compounds
and testosterone.
Scheme 4. Preparation of alkyne-modified library member 18 and
subsequent attachment to silica.
1800
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1796–1805

A New Solid-Phase Extraction Material for Endocrine-Disrupting Chemicals
an anisaldehyde (5% anisaldehyde in ethanol with 1% sulfuric acid)
From Figure 2, it is clear that the newly synthesised ma-
terial 20 entraps the selected EDCs from an aqueous matrix
and does so to a higher degree than the unmodified ami-
nopropylsilica. As a further control, the earlier mentioned
tetrapeptide Arg–Ser–Ser–Val–OH, developed by Tozzi and
co-workers, was N-terminally treated with pentynoic acid
and, after release, clicked onto azido-functionalised silica
19. This material did not show any EDC retention in this
experiment. These results confirm that there are interac-
or a PMA (5% phosphomolybdic acid in ethanol) solution. Flash
column chromatography was performed by using BIOSOLVE silica
gel (0.063–0.200 mm particle size). NMR spectra were recorded at
500 MHz or 300 MHz for proton and at 125 MHz or 75 MHz for
carbon nuclei in [D]chloroform, [D6]DMSO, [D4]MeOD or [D6]-
acetone. Chemical shifts are reported in units of parts per million
(ppm), referenced relative to the residual 1H or 13C peaks of the
used solvent as internal standards ([D]chloroform: 1H 7.26 and 13
77.16; [D6]DMSO: 1H 2.50 and 13C 39.52; [D4]MeOD: 1H 3.31 and
C
tions between the solid-phase-bound receptor 18 and the 13C 49.00; [D6]acetone: 1H 2.05 and 13C 29.84 and 206.26). Infrared
spectra (IR) were recorded on a Perkin–Elmer 1600 series FTIR
spectrometer and reported in wavenumbers (cm–1). Samples were
prepared as a thin films (neat) on KBr plates. Low-resolution mass
spectra were recorded with an atmospheric pressure electrospray-
ionisation (ESI) Hewlett–Packard 5988 A mass spectrometer.
HPLC analyses were performed on an Agilent 1100 Series instru-
ment with a Phenomenex Luna C18(2) column (250ϫ4.6 mm, 5µ
at 35 °C) by using a flow rate of 1 mL/min and with the following
solvent systems: 0.1% TFA in H2O (A) and MeCN (B). Unless
otherwise stated the column was flushed for 3 min with 100% A,
then a gradient from 0 to 100% B over 15 min was used, followed
by 5 min of flushing with 100% B. Photolyses were carried out with
a 4 W Bioblock Scientific compact UV lamp set at 365 nm. Melting
point ranges were determined with an Electrothermal 9100 melting
point apparatus.
various ligands. However, as can be seen from the percen-
tage of testosterone that is retained, no real selectivity has
yet been obtained.
Conclusions
This paper presents the solid-phase preparation of a li-
brary of 120 members of tripodal-scaffolded peptides that
are designed to mimic the hormone-binding domain of the
hER. The originally designed scaffold was modified to al-
low the fully automated synthesis of a parallel library in a
very short time frame. By using this parallel approach, the
peptide sequences of the members were known (positional
encoding), and, consequently, no time had to be invested in
deciphering the identity of the compounds whilst a signifi-
cant molecular diversity was still reached. The amount of
available material allowed a fast solution screening to be
performed by using a novel type of estradiol column. Be-
sides the establishment of protocols for the efficient synthe-
sis of new column materials with “click” chemistry, the fast
screening of the whole library was possible by using this
approach. Though strong interactions between the recep-
tors and the silica-bound ligand were observed (relative to
the interactions between the linear control peptide and the
silica-bound ligand), no selectivity was obtained as wit-
nessed after screening with a non estrogenic norethindrone
column. These results were then evaluated and confirmed
by “clicking” an alkyne-bearing library member on silica
and by using this material as a solid-phase extraction car-
tridge. With the methodology for fast synthesis and screen-
ing now fully optimised, we currently focus on the design of
new libraries with longer peptide chains in order to develop
compounds with stronger and more selective EDC-ex-
tracting capacities.
Materials: All amino acids and the solid support Tentagel-S-NH2
were purchased from NovaBiochem. DMF extra dry was pur-
chased from Aldrich. DMF peptide grade was purchased from Bi-
osolve. All chemicals were purchased and used without any further
purification, except tetrahydrofuran (THF), which was distilled
from Na/benzophenone prior to use, and dichloromethane, which
was distilled from CaH2.
Standard Operating Procedures
Swelling: The resin is swollen by adding a solvent (10 mL/g) of
choice (usually DMF or DCM) to the beads and shaking the sus-
pension for 5 min. To remove the solvent, the resin is filtered at the
end.
Washing: Unless stated otherwise, the resin is washed with DMF
(3ϫ), MeOH (3ϫ) and DCM (3ϫ). Finally, the resin can be dried
by washing with diethyl ether (3ϫ) or pentane (3ϫ).
Fmoc Deprotection: A 20% piperidine/DMF solution is added to
the resin (10 mL/g resin). The suspension is shaken for 1 min, after
which the suspension is filtered. This step is repeated for 5 min and
10 min. Finally, the resin is washed with DMF (3ϫ), MeOH (3ϫ)
and DCM (3ϫ).
Boc Deprotection: The resin is treated with a 50% TFA/DCM solu-
tion for 5 min. After filtration, the same solution is applied, and
the suspension is shaken for 25 min. Finally, the beads are washed
with DCM (3ϫ), MeOH (3ϫ), 10% DIEA/DCM (3ϫ) and DCM
(3ϫ).
Experimental Section
General Methods: All solid-phase reactions on 20 mg of resin or
less were performed in polypropylene Chromabond columns of
1 mL with a polyethylene frit, closed at the bottom with a B7 sep-
tum from Aldrich. Solid-phase reactions on a bigger scale were
performed in a peptide vessel protected against light with alumin-
ium foil and comprising a sintered glass funnel and a 3-way stop-
cock for easy filtration and washing. All solution-phase reactions
were conducted under an inert atmosphere of argon gas in oven
dried glassware. The reactions were monitored by thin layer
chromatography (TLC) by using SIL G-25 UV254 precoated silica
gel plates (0.25 mm thickness). The TLC plates were visualised with
Capping: The free amines are capped by adding a mixture of Ac2O/
pyridine/DCM (1:3:16) to the beads. After 1 h, the resin is filtered
and washed.
Loading Determination: The resin is treated with a 20% piperidine/
DMF solution for 30 min. After occasional swirling, the resin is
left to settle. Subsequently, a small amount of the solution is trans-
ferred to two UV cuvettes (≈3 mL). By measuring the absorbance
value at λ = 300 nm, the concentration of the piperidine/fulvene
adduct can be determined. This concentration can be correlated to
Eur. J. Org. Chem. 2009, 1796–1805
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1801

S. E. Van der Plas, E. Van Hoeck, F. Lynen, P. Sandra, A. Madder
FULL PAPER
the loading by taken into account the amount of resin used for the
loading test.
of a small quantity of starting material. The coupling was thus
repeated once, after which no more starting material was observed.
LC-MS analysis of 4 after photocleavage in EtOH (1 mg of resin
in 100 µL): calcd. for [M + H]+ 1298.6; found 1299.2.
Synthesis of ivDdeGlyOH (2): Et3N (1.74 mL, 12.5 mmol,
1.5 equiv.) and ivDdeOH (2.36 mL, 10.8 mmol, 1.3 equiv.) were
added to a suspension of glycine (623 mg, 8.30 mmol, 1 equiv.) in
EtOH (16 mL). The suspension was heated to reflux overnight un-
der inert atmosphere. After 18 h, the clear, yellow solution was con-
centrated under reduced pressure to give a yellow residue. After the
residue was redissolved in a 1:1 dioxane/water mixture, a 4  HCl
solution was added dropwise until no more precipitation occurred.
The white precipitate was dissolved in water/MeOH and was sub-
sequently lyophilised to give a white powder (1.86 g, 80%). 1H
NMR (500 MHz, CDCl3, 25 °C): δ = 9.99 (br. s, 1 H), 6.67 (br. s,
1 H), 5.87 (m, 1 H), 5.62 (s, 1 H), 5.27 (d, J = 17.0 Hz, 1 H), 5.18
(d, J = 10.1 Hz, 1 H), 4.56 (d, J = 4.7 Hz, 2 H), 3.96 (s, 2 H) ppm.
Automated Synthesis of Chemset 7b: All reactions were performed
at ambient temperature and without a protective atmosphere. The
resin was weighed into 2 mL syringes, depending on the amount of
resin. In 24 plastic syringes of 2 mL, a small amount of resin 4
(1–3 mg, 0.09–0.27 µmol) was weighed. These reaction vessels were
placed into the reaction block of the SYRO. After inserting the
correct sequences in the software program and preparing the AAx
(0.5  in DMF or NMP), HBTU (0.5  in DMF), DIEA (2  in
NMP), piperidine (40% piperidine in DMF) and capping (Ac2O/
pyridine/NMP 1:3:10) solutions, the sequence was started. Firstly,
the Fmoc group was deprotected. The first AA was coupled by
addition of a solution of AAx (60 µL) to the resin. Each syringe
was then filled with the HBTU solution (60 µL) and the DIEA
solution (30 µL). The suspension was vortexed for 40 min, after
which it was drained and washed. This procedure was repeated
once. The first strand was terminated by addition of the capping
mixture (120 µL) and DMF (120 µL) to the resin. The ivDde group
was removed by addition of a 2% H2NNH2·H2O/DMF (180 µL)
to the reaction vessel. After vortexing for 10 min, the resin was
shortly washed with DMF, and the deprotection was repeated once.
The subsequent AAx were then coupled as already described. The
second strand was terminated by capping the free amine with A2O.
Finally, the azide was reduced by adding 0.5  Me3P in THF/H2O
(180 µL) for 10 min to the reaction vessels. The reduction was re-
peated once, and then the resin was drained and washed thoroughly
with DMF and iPrOH. The subsequent AAx were then coupled as
already described. The synthesis was terminated by capping the
free amine. After the synthesis, the resins were washed with iPrOH
(3ϫ) and pentane (3ϫ). In total, 120 members were synthesised by
repeating this overall procedure over 5 synthesis cycles. MALDI-
TOF analysis of 7b after photocleavage in EtOH: calcd. for [M +
H]+ 2335.1; found 2094.3 [M – Trt + H]+, 2115.5 [M – Trt + Na]
+, 2334.9 [M + H]+, 2350.7 [M + O + H]+, 2357.4 [M + Na]+,
2373.4 [M + O + Na]+ or [M + K]+.
Synthesis of Azidoacetic Acid (3): NaN3 (5.13 g, 78.9 mmol,
2 equiv.) was dissolved in DMSO (210 mL) and stirred for 90 min.
Bromoacetic acid (5.17 g, 37.4 mmol, 1 equiv.) in DMSO (2 mL)
was then added. After stirring for 18 h, the solution was diluted
with H2O (170 mL) and acidified with 6  HCl (33 mL). The aque-
ous phase was then extracted three times with MTBE. The organic
phases were pooled and extracted with a 0.1  HCl solution satu-
rated with NaCl. After drying and subsequent evaporation of the
organic phase, an oil (2.40 g, 64%) was obtained. 1H NMR
(300 MHz, CDCl3, 25 °C): δ = 9.62 (br.s, 1 H), 3.97 (br.s, 1 H),
ppm. IR: ν = 2111 (s) cm–1.
˜
Synthesis of Construct 4: Resin 1 (1.6 g, 0.100 mmol/g, 0.19 mmol,
1 equiv.) was Fmoc deprotected. During Fmoc deprotection,
ivDdeGlyOH (646 mg, 2.3 mmol, 12 equiv.) was dissolved in DCM
(4 mL), and the solution was cooled down until 0 °C. After adding
DIC (170 µL, 1.1 mmol, 6 equiv.), the mixture was stirred for
30 min and subsequently added to the Fmoc-deprotected scaffold.
To ensure complete solubility of the symmetrical anhydride, DMF
(4 mL) was added. The resulting suspension was shaken for 18 h
and after filtering the resin, the beads were thoroughly washed. A
small sample was photocleaved in EtOH and analysed by LC,
which showed the presence of a small amount of starting material.
The coupling was thus repeated once, and no more starting mate-
rial was observed. Next, the Boc group was deprotected. In the
mean while, FmocGlyOH (683 mg, 2.3 mmol, 12 equiv.) was dis-
solved in DCM (4 mL) and cooled down until 0 °C. After adding
DIC (170 µL, 1.1 mmol, 6 equiv.), the mixture was stirred for
30 min and subsequently added to the resin. To ensure complete
solubility of the symmetrical anhydride, DMF (4 mL) was added.
The resulting suspension was shaken for 18 h, and after filtering
the resin, the beads were thoroughly washed. A small sample was
photocleaved in EtOH and analysed by LC, which showed the pres-
ence of a small amount of starting material. The coupling was thus
repeated once, and no more starting material was observed. Finally,
the resin was suspended in DCM. First, the scavenger PhSiH3
(585 µL, 4.75 mmol, 25 equiv.) was added to this suspension fol-
lowed the Pd0 catalyst (22 mg, 19 µmol, 10 mol-%). The resulting
brown mixture was shaken for 1 h, filtered and washed three times
with DCM. After repeating this procedure once, the Alloc group
was fully removed as monitored by HPLC after photocleavage in
EtOH. Next, azidoacetic acid (3, 230 mg, 2.3 mmol, 12 equiv.) was
dissolved in DCM (8 mL) and cooled down until 0 °C. After add-
ing DIC (170 µL, 1.1 mmol, 6 equiv.), the mixture was stirred for
30 min and subsequently added to the Alloc-deprotected resin. The
resulting suspension was shaken for 18 h, and after filtering the
resin, the beads were thoroughly washed. A small sample was pho-
tocleaved in EtOH and analysed by LC, which showed the presence
Side Chain Deprotection and Met(O) Reduction of Chemset 7b: The
TFA cleavage mixture was prepared by mixing TIS (100 µL),
DODT (250 µL), Me2S (200 µL) and thioanisole (500 µL) with
TFA (8.45 mL). For the reduction of Met(O), NH4I (230 mg) was
dissolved in H2O (500 µL). Firstly, NH4Iaq (10 µL) was added to
each member, followed by the addition of the TFA cleavage mixture
(190 µL). The suspension was left for 3 h, after which the beads
were drained and washed thoroughly. ES analysis of 7 after photo-
cleavage in H2O: calcd. for [M + H]+ = 1784.9; found 892.6 [M +
2H]2+, 903.6 [M + Na + H]2+
.
Synthesis of Azidohexanoic Acid (9): Bromohexanoic acid (6 g,
31 mmol, 1 equiv.) was dissolved in DMF (20 mL). After adding
NaN3 (4 g, 62 mmol, 2 equiv.), the solution was heated to reflux
for 18 h with an oil bath with a temperature of 85 °C. After cooling
the solution, DMF was evaporated under vacuum to give an oil,
which was redissolved in DCM. This organic phase was then ex-
tracted three times with a 0.1  HCl solution. After drying the
organic phase on MgSO4 and evaporation, a light yellow oil was
obtained (3.96 g, 81% yield). 1H NMR + COSY (500 MHz,
CDCl3, 25 °C): δ = 10.37 (br. s, 1 H), 3.26 (t, J = 6.9 Hz, 2 H),
2.86 (t, J = 7.4 Hz, 2 H), 1.69–1.64 (m, 2 H), 1.62–1.58 (m, 2 H),
1.45–1.39 (m, 2 H) ppm. IR: ν = 2093 (s) cm–1.
˜
Attachment of 9 to Aminopropylsilica (5 µm): The silica material
(2 g, max. loading 0.73 mmolg–1) was suspended in DMF (10 mL)
1802
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1796–1805

A New Solid-Phase Extraction Material for Endocrine-Disrupting Chemicals
and azidohexanoic acid (9, 691 mg, 4.4 mmol, 3 equiv.) was added
to this suspension. After addition of PyBOP (2.3 g, 4.4 mmol,
3 equiv.) and DIEA (1.53 mL, 8.8 mmol, 6 equiv.), the suspension
was shaken for 18 h. Subsequently, the solution was drained, and
the remaining powder was washed thoroughly. After capping, the
ninhydrine test gave a colourless result, which indicates that there
eous, it was transferred to the pre-swollen resin, and after the ad-
dition of DIEA (181 µL, 1.04 mmol), the suspension was shaken
for 18 h. The resin was drained and washed thoroughly. The beads
turned dark yellow with the TNBS test and red with NF31. The
remaining free spacer amino groups were capped for 1 h with a
0.1  solution of AcOH and PyBOP in the presence of DIEA
(2 equiv.). The ninhydrin test was negative. The total loading was
0.101 mmolg–1, which corresponded (maximal theoretical loading
0.173 mmolg–1) to a yield of 58%.
was no more free amine left. IR: ν = 3414 (br. s), 2104 (s), 1642
˜
(m), 1111 (br. s), 810 (s) cm–1.
Clicking of Ethinylestradiol to Construct 10: Silica material 10 (6 g)
was suspended in an iPrOH/H2O (60 mL, ratio 1:1) solution. Ethin-
ylestradiol (4.45 g, 15 mmol, 1 equiv.) was added to this suspension
followed by CuSO4·5H2O (749 mg, 3 mmol, 20 mol-%) and sodium
ascorbate (1.2 g, 6 mmol, 40 mol-%). The resulting orange solution
was gently stirred in an oil bath with a temperature of 40 °C. The
reaction mixture was left overnight, after which the suspension was
filtered and the remaining silica material was washed with MeOH
(3ϫ), EDTAaq (3ϫ), MeOH (3ϫ) and diethyl ether (3ϫ). All the
azide functionalities had reacted, as shown by the disappearance
of the azide absorption (≈2100 cm–1) in the IR spectrum. Finally,
the silica material 11 was dried overnight at an oven temperature
of 60 °C. The column was packed by using the slurry packing
method with a Haskel air driven pump (Burbobank, CA, USA).[27]
The slurry solvent was THF/water (50:50) and the packing solvent
used was deionised water, both were degassed. The derivatised silica
(approx. 0.7 g) was slurried in slurry solvent (8 mL) in an ultrasonic
bath, followed by packing at 450 bar. The columns were 15 cm in
Incorporation of the Three Differently Protected Glycines: For the
incorporation of the glycine derivatives, the same protocol as that
described for the synthesis of 4 was applied. The amount of resin
16 used was 1.1 g (110 µmol).
Preparation of Compound 18: Automated synthesis was performed
in a manner similar to the protocol described for the preparation
of library 7b. Construct 17 (100 mg, 11 µmol) was subjected to this
protocol. The protected peptide was cleaved from the resin by irra-
diating the beads with light at 360 nm. This was done under con-
tinuous vortexing of the beads in DCM (10 mL) for 18 h. The sus-
pension was then drained, and the beads were washed with DCM
(3ϫ), EtOH (3ϫ) and DCM (3ϫ). The organic fractions were com-
bined, and the solvents evaporated. This whole procedure was re-
peated five times, and the organic fractions were combined in the
same flask. For side-chain deprotection, a mixture of TFA/TIS/
DODT/thioanisole/Me2S (89.5:1:2.5:5:2) was prepared. After add-
ing NH4Iaq (125 µL, 460 mg/mL) to the flask containing the pro-
tected peptide, the TFA mixture (2.5 mL) was added. The solution
was stirred for 3 h, and then it was evaporated under a flow of Ar.
The residue was subsequently dissolved in 10% AcOH/H2O, and
this aqueous fraction was extracted three times with MTBE ether.
The water fraction was then lyophilised. The residue obtained after
lyophilisation was redissolved in 10% AcOH/H2O (1 mL). The de-
sired compound was then purified by injecting 500 µL of this mix-
ture on a preparative HPLC system (C18 column). A gradient from
0–40% MeCN in 40 min was used in combination with a 0.1%
TFAaq solution (from 100–60% in 40 min). After lyophilisation, 18
(3.5 mg, 16%) was obtained. MALDI-TOF analysis: calcd. for
[M – H]– 1904.9; found 1905.1 [M – H]–, 1928.7 [M – H + Na]–,
1942.9 [M – H + K]–.
length and 2.1 mm in diameter. IR: ν = 3399 (br. s), 2936 (s), 1670
˜
(s), 1050 (br. s), 800 (s) cm–1.
“Clicking” of Norethindrone to Construct 10: Silica material 10 (1 g)
was suspended in an iPrOH/H2O (10 mL, 1:1) solution. Norethin-
drone (756 mg, 2.5 mmol, 1 equiv.) was added to this suspension,
followed by CuSO4·5H2O (125 mg, 0.5 mmol, 20 mol-%) and so-
dium ascorbate (198 mg, 1.0 mmol, 40 mol-%). The resulting
orange solution was gently stirred in an oil bath with a temperature
of 40 °C. The reaction mixture was left overnight, after which the
green suspension was filtered and the remaining silica material was
washed with MeOH (3ϫ), EDTAaq (3ϫ), MeOH (3ϫ) and ether
(3ϫ). Finally, the silica material 12 was dried overnight at an oven
temperature of 60 °C. The column was packed in an analogous way
as that described for 11. IR: ν = 3430 (br. s), 2947 (m), 1647 (s),
˜
Preparation of Azido-functionalised Silica 19: The silica particles
(2.1 g) were put in a plastic falcon tube. DMF (6 mL) was added,
1092 (br. s), 802 (s) cm–1.
Synthesis of Construct 15: Resin 13 (1 g, 0.21 mmol, 1 equiv.) was followed by the addition of FmocGlyOH (119 mg, 0.4 mmol,
weighed into a reaction vessel. After the addition of DMF 20 mol-%), AcOH (92 µL, 1.6 mmol, 80 mol-%), HBTU (759 mg,
(4.6 mL), FmocLys(Boc)OH (539 mg, 1.15 mmol, 5.5 equiv.) and 2 mmol, 1 equiv.) and DIEA (700 µL, 4 mmol, 2 equiv.). After
PyBOP (598 mg, 1.15 mmol, 5.5 equiv.) were added to this suspen-
sion. Subsequently, DIEA (400 µL, 2.30 mmol, 11 equiv.) was
added, and the suspension was shaken for 4 h and 30 min, after
which the solution was drained and the beads were thoroughly
washed. The ninhydrine test on a small sample was negative. The
Fmoc group was then deprotected to give construct 14. Next, pen-
shaking for 90 min, the suspension was centrifuged and washed
thoroughly. Determination of the loading gave value of
a
0.110 mmolg–1. Next, the Fmoc group was deprotected. The pres-
ence of free amine groups was confirmed by a positive TNBS test.
The material (1 g, 0.066 mmol, 1 equiv.) was then suspended in
DMF (4 mL). PyBOP (260 mg, 0.5 mmol, 7.6 equiv.), azidohexa-
tynoic acid (68 mg, 0.69 mmol, 3.3 equiv.), PyBOP (359 mg, noic acid (9, 79 mg, 0.5 mmol, 7.6 equiv.) and DIEA (174 µL,
0.69 mmol, 3.3 equiv.) and DIEA (241 µL, 1.38 mmol, 6.6 equiv.)
were added to a suspension of 14 in DMF (4.6 mL). The suspen-
sion was shaken for 3 h and 30 min and drained, and the remaining
beads were washed. The ninhydrine test was negative. ES analysis
after photocleavage in EtOH: calcd. For [M + H]+ 326.4; found
348.0 [M + Na]+.
1 mmol, 15.2 equiv.) were subsequently added to this suspension.
After shaking for 4 h, the solution was drained, and the particles
were thoroughly washed. Though no free amines were detected by
the TNBS test, a capping step was still performed to ensure that
there were no free amines. IR: ν = 3414 (br. s), 2104 (s), 1642 (m),
˜
1111 (br. s), 810 (s) cm–1.
Attachment of the Tripodal Scaffold to Construct 15: Resin 15
(1.1 g, 0.23 mmol, 1 equiv.) was weighed in a reaction vessel and
Boc deprotected. A solution of the tripodal scaffold (345 mg,
0.35 mmol, 1.5 equiv.) and PyBOP (180 mg, 0.35 mmol, 1.5 equiv.)
in DMF (15 mL) was prepared. When the solution was homogen-
“Clicking” of Tripodal Compound 18 to Azido-functionalised 19:
Firstly, a UV-calibration curve was determined by preparing dif-
ferent concentrations (98, 9.8, 4.9, 2.5, 2.0 and 1.0 µ) of 18 in
H2O. By measuring the absorbance of the different concentrations,
the following equation was determined: Abs = 63.082ϫ [18]aq
–
Eur. J. Org. Chem. 2009, 1796–1805
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1803

S. E. Van der Plas, E. Van Hoeck, F. Lynen, P. Sandra, A. Madder
0.0544 (R2 = 0.9999). Next, CuSO4·5H2O (5.5 mg, 22 µmol) and Framework Programme under contract number MEST-CT-2005-
FULL PAPER
sodium ascorbate (8.71 mg, 44.0 µmol) were dissolved in H2O
(20 mL). From this stock solution, 200 µL (16 mol-% CuSO4·5H2O
and 32 mol-% sodium ascorbate) was pipetted in an eppendorf con-
taining 18 (3.5 mg, 1.64 µmol, 8.2 m 1 equiv.) and 19 (20 mg,
2.20 µmol). The resulting mixture was vortexed at 40 °C. After
48 h, 3 µL of the reaction mixture was transferred to a UV cuvette
and 2997 µL H2O was added. For the blank sample, 3 µL from the
CuSO4·5H2O stock solution was diluted up to 3 mL in a UV cu-
vette. The absorbance from this sample was 0.121 mAu, which cor-
relates to a concentration of 2.8 m. From this value, a loading of
≈52 µmol/g was determined. The silica material 20 was isolated by
centrifugation and subsequent decantation of the liquid. The mate-
rial was washed thoroughly water (3ϫ), EDTAaq (3ϫ) and iPrOH
(3ϫ).
020643. S. Van der Plas is indebted to the Institute for the Pro-
motion of Innovation through Science and Technology in Flanders
(IWT Vlaanderen) for a research fellowship. Financial support
from the National Fund of Scientific Research (FWO-Vlaanderen),
(project 1.5.146.06) and Ghent University (GOA 01G01507) are
gratefully acknowledged.
[1] F. Harry, G. Robert, Pure Appl. Chem. 2003, 75, 2181–2193.
[2] a) D. M. Tanenbaum, Y. Wang, S. P. Williams, P. B. Sigler,
Proc. Natl. Acad. Sci. USA 1998, 95, 5998–6003; b) A. C. Pike,
A. M. Brzozowski, R. E. Hubbard, T. Bonn, A. G. Thorsell, O.
Engstrom, J. Ljunggren, J. A. Gustafsson, M. Carlquist,
EMBO J. 1999, 18, 4608–4618.
[3] M. J. L. de Alda, D. Barcelo, J. Chromatogr. A 2001, 938, 145–
153.
Screening by Affinity LC: LC-MS analyses were carried out on an
Alliance 2690 LC system equipped with an online degasser and an
autosampler (Waters Milford, MA, USA). The affinity LC analyses
were carried out on the home-made estradiol or testosterone col-
umn (5 µm, 2.1ϫ150 mm). The column was thermostatted at
37 °C. The mobile phase consisted of 10 m NH4Ac in water and
methanol. The pH of the water was adjusted to 7.4 with NH3.
Different gradients were evaluated, and the best results were ob-
tained using the following linear gradient. Firstly, the mobile phase
consisted of 100% 10 m NH4Ac in water. Thereafter, the compo-
sition was changed to 100% MeOH in 10 min. Finally, this compo-
sition was maintained for 15 min. In the end, the mobile phase
composition was returned to the initial conditions. The flow rate
was 0.2 mL/min, and the injection volume was 10 µL, which corre-
sponds to approximately 250 ppm.
[4] P. T. Holland, Pure Appl. Chem. 2003, 75, 1843–1857.
[5] P. Su, X.-X. Zhang, W.-B. Chang, J. Chromatogr. B 2005, 816,
7–14.
[6] a) J. C. Bravo, P. Fernandez, J. S. Durand, Analyst 2005, 130,
1404–1409; b) Z. H. Meng, W. Chen, A. Mulchandani, Environ.
Sci. Technol. 2005, 39, 8958–8962; c) J. C. Bravo, R. M. Garci-
nuno, P. Fernandez, J. S. Durand, Anal. Bioanal. Chem. 2007,
388, 1039–1045.
[7] a) B. Tse Sum Bui, A. S. Belmont, H. Witters, K. Haupt, Anal.
Bioanal. Chem. 2008, 8, 2081–2088; b) L. Ye, Y. H. Yu, K.
Mosbach, Analyst 2001, 126, 760–765; c) A. Rachkov, S.
McNiven, S. H. Cheong, A. ElЈSkaya, K. Yano, I. Karube, Su-
pramol. Chem. 1998, 9, 317–323; d) G. Giraudi, C. Giovannoli,
C. Tozzi, C. Baggiani, L. Anfossi, Chem. Commun. 2000, 13,
1135–1136.
[8] C. Tozzi, L. Anfossi, G. Giraudi, C. Giovannoli, C. Baggiani,
A. Vanni, J. Chromatogr. A 2002, 966, 71–79.
SPE of an Aqueous Mixture of EDCs: The affinity of the aminopro-
pyl silica and material 20 were evaluated by using the following
procedure. Firstly, the materials were incubated with 900 µL of a
10 m NH4Ac/water solution, adjusted with NH3 to obtain a pH
of 7.4. The cartridge was then loaded with 900 µL of a 10 m
NH4Ac/water solution, adjusted with NH3 to obtain a pH of 7.4,
spiked with the EDCs, which resulted in a concentration of 10 mg/
L (ppm) for each EDC in the water. The flow rate was approxi-
mately 0.5 mL/min. The concentration of the EDCs in the water
effluent was determined with HPLC-UV. After SPE extraction, the
cartridges were rinsed with MeOH to remove the extracted EDCs
from the resin so that the resins can be reused. The effluent ob-
tained after the SPE procedure was separated on a Luna-C18 col-
umn (5 µm, 2.1ϫ150 mm) (Phenomenex, USA). The column was
thermostatted at 20 °C. The mobile phase consisted of acetonitrile
and water. A linear gradient program was used from 40–60% aceto-
nitrile in 15 min. Thereafter, the mobile phase was switched directly
to 100% acetonitrile. This mobile phase composition was then
maintained for 5 min. Finally, the mobile phase composition was
returned to the initial conditions. The flow rate was 0.2 mL/min,
and the injection volume was 10 µL. Ultraviolet detection was car-
ried out by using a Waters 2487 dual λ absorbance detector (Waters
Milford). Analyses were recorded at 230 nm.
[9] A. Gea, N. Farcy, N. R. I. Rossell, J. C. Martins, P. J.
De Clercq, A. Madder, Eur. J. Org. Chem. 2006, 18, 4135–4146.
[10] a) H. Wennemers, M. C. Nold, M. M. Conza, K. J. Kulicke,
M. Neuburger, Chem. Eur. J. 2003, 9, 442–448; b) C.
Chamorro, J. W. Hofman, R. M. J. Liskamp, Tetrahedron 2004,
60, 8691–8697; c) C. Chamorro, R. M. J. Liskamp, Tetrahedron
2004, 60, 11145–11157.
[11] a) H. De Muynck, A. Madder, N. Farcy, P. J. De Clercq, M. N.
Perez-Payan, L. M. Ohberg, A. P. Davis, Angew. Chem. Int. Ed.
2000, 39, 145–148; b) H. B. Albada, R. M. J. Liskamp, J.
Comb. Chem. 2008, 10, 814–824.
[12] A. Madder, L. Li, H. De Muynck, N. Farcy, D. Van Haver, F.
Fant, G. Vanhoenacker, P. Sandra, A. P. Davis, P. J. De Clercq,
J. Comb. Chem. 2002, 4, 552–562.
[13] A. W. Czarnik, Proc. Natl. Acad. Sci. USA 1997, 94, 12738–
12739.
[14] C. Schmuck, P. Wich, New J. Chem. 2006, 30, 1377–1385.
[15] a) D. M. Tanenbaum, Y. Wang, S. P. Williams, P. B. Sigler,
Proc. Natl. Acad. Sci. USA 1998, 95, 5998–6003; b) A. M.
Brzozowski, A. C. W. Pike, Z. Dauter, R. E. Hubbard, T. Bonn,
O. Engstrom, L. Ohman, G. L. Greene, J. A. Gustafsson, M.
Carlquist, Nature 1997, 389 (6652), 753–758; c) D. M. Tanen-
baum, Y. Wang, S. P. Williams, P. B. Sigler, Proc. Natl. Acad.
Sci. USA 1998, 95, 5998–6003.
[16] S. E. Van der Plas, A. Gea, S. Figaroli, P. J. De Clercq, A. Mad-
der, Eur. J. Org. Chem. 2008, 9, 1582–1588.
[17] C. P. Holmes, J. Org. Chem. 1997, 62, 2370–2380.
[18] R. Franke, C. Doll, J. Eichler, Tetrahedron Lett. 2005, 46, 4479–
4482.
[19] A SYRO Multiple Peptide Synthesiser equipped with a vor-
texing unit from Multisyntech, Witten, Germany was used.
This apparatus does not allow use of TFA needed for Boc re-
moval. It was further feared that the Pd0 used for the deprotec-
tion of the Alloc group, would be oxidised during the deprotec-
tion cycle as no inert atmosphere could be assured on our cur-
rent SYRO equipment.
Supporting Information (see footnote on the first page of this arti-
cle): RPHPLC and MS data for cleaved compound 4 and detailed
RPHPLC and MS data for 18 (15%) of the library members are
presented.
Acknowledgments
This research project was supported by a Marie Curie Early Stage
Research Training Fellowship of the European Community’s Sixth
1804
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1796–1805

A New Solid-Phase Extraction Material for Endocrine-Disrupting Chemicals
[20] S. C. Miller, T. S. Scanlan, J. Am. Chem. Soc. 1998, 120, 2690–
2691.
[24] J. M. Dyke, A. P. Groves, A. Morris, J. S. Ogden, A. A. Dias,
A. M. S. Oliveira, M. L. Costa, M. T. Barros, M. H. Cabral,
A. M. C. Moutinho, J. Am. Chem. Soc. 1997, 119, 6883–6887.
[25] J. Grandjean, J. Org. Chem. 2005, 70, 7123–7132.
[26] H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew. Chem. Int.
Ed. 2001, 40, 2004–2008.
[27] HPLC slurry packing method, Data sheet 95551U, Alltech As-
sociates, Inc.
[21] S. R. Chhabra, B. Hothi, D. J. Evans, P. D. White, B. W.
Bycroft, W. C. Chan, Tetrahedron Lett. 1998, 39, 1603–1606.
[22] a) F. Debaene, N. Winssinger, Org. Lett. 2003, 5, 4445–4447;
b) Z. S. Zhang, T. Carter, E. K. Fan, Tetrahedron Lett. 2003,
44, 3063–3066; c) R. B. Yan, F. Yang, Y. F. Wu, L. H. Zhang,
X. S. Ye, Tetrahedron Lett. 2005, 46, 8993–8995.
[23] N. Thieriet, J. Alsina, E. Giralt, F. Guibe, F. Albericio, Tetrahe-
dron Lett. 1997, 38, 7275–7278.
Received: December 27, 2008
Published Online: March 9, 2009
Eur. J. Org. Chem. 2009, 1796–1805
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1805