An enzyme-labile safety catch linker for combinatorial synthesis on a soluble polymeric support

Article abstract of DOI:10.1002/(SICI)1521-3773(20000502)39:9<1629::AID-ANIE1629>3.0.CO;2-E

An enzyme-initiated cyclization allows the release of different target molecules from a soluble polymeric support (see picture) under mild conditions (pH 7, 37°C) and in high yield. X = O, NH, NR.

Full text of DOI:10.1002/(SICI)1521-3773(20000502)39:9<1629::AID-ANIE1629>3.0.CO;2-E

COMMUNICATIONS
An Enzyme-Labile Safety Catch Linker for
Combinatorial Synthesis on a Soluble
Polymeric Support**
O
X
target molecule
MeO
O
H
N
O
N
H
O
Uwe Grether and Herbert Waldmann*
O
enzyme-labile
group
Combinatorial chemistry and parallel synthesis of com-
pound libraries on polymeric supports are efficient methods
for the generation of new substances with a predetermined
profile of properties.[1] Paramount to the success of this
approach is the availability of suitable polymers and widely
applicable linker groups for the attachment of the desired
compounds to the polymeric carrier. Recent research in this
field has focused on and achieved the successful transfer of a
steadily growing number of reaction types to synthesis on
solid supports, and the demands on the linker groups
concerning their stability under different reaction conditions
and selective cleavage have risen accordingly. Therefore, new
and broadly applicable linkers which are stable under a
variety of reaction conditions, but which at the same time
allow release of the synthesized products under the mildest
conditions, are of great interest in organic synthesis and
combinatorial chemistry.[2] Enzymatic methods may open up
advantageous alternatives to classical chemical techniques
since enzyme-catalzyed transformations often proceed under
very mild conditions (pH 5 ± 8, 25 ± 378C) and with very high
chemo-, regio-, and stereoselectivity.[3] We now report on the
development of a new enzyme-labile safety catch linker for
combinatorial chemistry[4] on a soluble polymeric support,
allowing for the release of the target compounds in high yields
and under neutral conditions.
X = O, NH, NR
1
penicillin G
COOH
_
acylase
O
2
X
target molecule
MeO
O
O
NH2
N
H
O
MeO
O
O
O
NH
+
HX target molecule
N
O
H
3
4
Scheme 1. Principle for the development of the enzyme-labile safety catch
linker.
synthesis. Thereby possible steric or electronic interactions of
the biocatalyst with the synthesis products, which might lead
to a reduction of the substrate tolerance of the enzyme, are
minimized. The release of the target molecules proceeds
according to the safety catch principle in a two-step process. In
the first step, the amidase penicillin G acylase hydrolyses the
phenylacetamide with complete chemo- and regioselectivity
and under exceptionally mild conditions (pH 7.0, room
temperature or 378C).[6] Subsequently benzylamine 2, thereby
generated as an activated intermediate, cyclizes to polymer-
bound lactam 3 and releases the corresponding target
molecule 4.
The linker was synthesized from commercially available
homovanilinic acid methyl ester 5 (Scheme 2). After alkyl-
ation of the phenolic hydroxyl group with THP-protected
bromoethanol in high yield, the resulting trisubstituted
aromatic compound was subjected to a completely regiose-
lective electrophilic amidoalkylation with N-hydroxymethyl-
phenylacetic acid amide.[7] Under the acidic reaction condi-
tions the THP ether was simultaneously converted into the
corresponding acetate which was then saponified. By means
of this three-step sequence, linker 6 is accessible in an overall
yield of 67%.
In the design of the new linker groups a biocatalyzed
transformation was combined with a subsequent intramolec-
ular cyclization reaction, according to the principle of
ªassisted removalº[5] of the desired target molecules. To this
end the linker embodies a functional group which is recog-
nized and attacked by the biocatalyst. The enzyme liberates
an intermediate which cyclizes, with release of the target
compounds. Furthermore, the linker carries an additional
functional group for attachment to the polymeric support. The
realization of this principle is shown in Scheme 1. The linker
group is attached to an amino-functionalized carrier as a
urethane (!1). It allows for the attachment of, for example,
alkyl halides, alcohols, or amines as carboxylic acid esters and
amides and their subsequent conversion to product libraries.
The enzyme-labile functional group, a phenylacetamide, is
remote from the structures to be generated by combinatorial
[*] Prof. Dr. H. Waldmann, Dipl.-Chem. U. Grether[‡]
Max-Planck-Institut für molekulare Physiologie
Abteilung Chemische Biologie
Otto-Hahn-Strasse 11, 44227 Dortmund (Germany)
Fax : (‡49)231-133-2499
and
Fachbereich 3, Organische Chemie,
Universität Dortmund, 44227 Dortmund (Germany)
E-mail: herbert.waldmann@mpi-dortmund.mpg.de
The polymeric carrier chosen for attachment to the linker
was a soluble polyethyleneglycol functionalized at both
termini with an amino group and with an average molecular
mass of 6000 Da (POE 6000).[8] POE 6000 is soluble in
numerous organic solvents but can be precipitated, filtered
off, and washed after addition of diethyl ether, thereby
‡
[ ] Institut für Organische Chemie der Universität
Richard-Willstätter-Allee 2, 76128 Karlsruhe (Germany)
[**] This work was supported by the Bundesministerium für Bildung und
Forschung and BASF AG.
Angew. Chem. Int. Ed. 2000, 39, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3909-1629 $ 17.50+.50/0
1629

COMMUNICATIONS
incubated with penicillin G acylase at pH 7 and 378C. After
48hours, benzyl alcohols 13 ± 15 were isolated in high
yields and with a purity of >95% by simple extraction with
diethyl ether (the phenylacetic acid remains in the aqueous
phase).
CH3O
COOMe
HO
5
Br
1)
2)
THPO
90°C, 10h, 94%
, K2CO3, DMF
In a further series of experiments Mitsunobu esterification
and Diels ± Alder reactions were investigated (Scheme 4). For
the Mitsunobu reaction, the polymer was esterified quantita-
tively with 1,4-bis(hydroxymethyl)benzene. The obtained
polymer-bound benzyl alcohol 16 was then treated with
4-acetamidophenol in the presence of the Mitsunobu reagent
to give phenyl ether 17 in quantitative yield. By subsequent
treatment of polymeric substrate 17 with penicillin G acylase,
aryl benzyl ether 18 could be obtained in 81% yield. For the
Diels ± Alder reaction, 4-hydroxybutyl acrylate was coupled
to the linker and the polymer-bound acrylic acid ester 19 was
treated with cyclopentadiene. This reaction was carried out at
608C at ambient pressure, and at 1008C in a closed glass vial
to investigate the pressure and temperature stability of the
polymer-bound linker. According to NMR spectroscopic
analysis cycloaddition product 20 was formed with an endo/
exo ratio of 2.5:1 and with quantitative conversion. Subse-
quent enzymatic release delivered alcohol 21 in high yield and
purity. The examples shown in Scheme 4 impressively prove
the chemoselectivity of penicillin G acylase since the enzyme
H
N
, HOAc/H SO (20:1)
ca. 20°C, 10h, 83%
CH2OH
2
4
O
3) NaOMe, MeOH, 0°C, 1h, 87%
OMe
O
CH3O
O
H
N
HO
O
6
1) COCl2, CH2Cl2, 0°C, 3h, quant.
, HOBt, DIPEA, DMF, ca. 20°C,
2)
NH2
10h, quant.
3) 0.5M LiOH/MeOH (1:1), ca. 20°C, 2h, quant.
O
OH
CH3O
H
N
H
N
O
O
O
O
=
POE 6000
7
Scheme 2. Synthesis of the linker and coupling to the
soluble polymer POE 6000. DIPEAˆdiisopropylethyl-
amine, HOBt ˆ 1-hydroxy-1H-benzotriazole, THP ˆ
tetrahydropyranyl.
+
linker OH
Br
I
7
8
Cs2CO3, DMF, 50 C, 24h
facilitating the separation of surplus reagents
and side products. It allows for NMR spectro-
scopic monitoring of the reactions and, due to
its pronounced solubility in water, for biocata-
lyzed transformations it offers the advantage
that the substrates are accessible to the bio-
catalyst in homogeneous phase. For the attach-
ment of 6 to POE 6000 the primary hydroxyl
group was first converted into the correspond-
ing chloroformic acid ester with phosgene and
then the aminofunctionalized polymer was
acylated with this intermediate quantitative-
ly.[9] Basic hydrolysis of the methyl ester was
also quantitative and yielded functionalized
polymer 7.
The suitability of the polymer± linker con-
jugate 7 for combinatorial syntheses and,
thereby, also the scope of its applicability were
investigated for a variety of different trans-
formations. To this end, first the carboxyl group
of the linker was esterified with m-iodobenzyl-
bromide 8 and the polymer-bound aryl iodide 9
generated was then further transformed by
Heck[10], Suzuki[11] or Sonogashira[12] reactions
(Scheme 3). NMR spectroscopic analysis re-
vealed that these transformations proceeded
quantitatively. Compounds 10 ± 12 were then
linker
O
I
9
O
OtBu
OMe
[Pd(PPh3)2Cl2], CuI
dioxane/Et3N (2:1)
ca. 20C, 24h
Pd(OAc)2, Bu4NBr, PPh3
DMF/Et3N/H2O (9:1:1)
50C, 20h
B(OH)2
[Pd(PPh3)4], K3PO4, DMF
80 C, 20h
OMe
tBuO
O
linker
O
linker
O
linker
O
10
11
12
penicillin G acylase
pH 7.0, 10% MeOH
37C
penicillin G acylase
pH 7.0, 10% MeOH
37C
penicillin G acylase
pH 7.0, 10% MeOH
37C
OMe
tBuO
O
HO
HO
HO
13
75%
14
73%
15
94%
Scheme 3. Palladium-catalyzed reactions on polymer 7 and enzyme-catalyzed release of the
coupling products from the polymeric carrier.
1630
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3909-1630 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 9

COMMUNICATIONS
natorial chemistry our results prove that enzymes, in
general, are valuable reagents for transformations
on polymeric supports.
linker
OH
7
DIC, DMAP, HOBt, DMF, pyridine
0°C, 1h, then ca. 20°C, 24h
DIC, DMAP, HOBt, DMF, pyridine
0°C, 1h, then ca. 20°C, 24h
HO
OH
O
O
Received: December 22, 1999 [Z14448]
HO
O
[1] a) F. Balkenhohl, C. von dem Bussche-Hünnefeld, A.
Lansky, C. Zechel, Angew. Chem. 1996, 108, 2436 ± 2487;
Angew. Chem. Int. Ed. Engl. 1996, 35, 2288 ± 2337; b) J. S.
Früchtel, G. Jung, Angew. Chem. 1996, 108, 19 ± 46;
Angew. Chem. Int. Ed. Engl. 1996, 35, 17 ± 42 ; c) L. A.
Thompson, J. A. Ellman, Chem. Rev. 1996, 96, 555 ± 600;
d) D. Obrecht, J. M. Villalgordo, Solid-Supported Combi-
natorial and Parallel Synthesis of Small-Molecular-Weight
Compound Libraries, Pergamon, New York, 1998.
[2] a) I. W. James, Tetrahedron 1999, 55, 4855 ± 4946, and
references therein; b) Review of new developments, see:
B. J. Backes, J. A. Ellman, Curr. Opin. Chem. Biol. 1997, 1,
86 ± 93; c) F. Stieber, U. Grether, H. Waldmann, Angew.
Chem. 1999, 111, 1142± 1145; Angew. Chem. Int. Ed. 1999,
38, 1073 ± 1077.
linker
O
linker
O
OH
O
16
19
H
N
O
DEAD, PPh3, THF/CH2Cl2 (1:1)
molecular sieves 4Å
THF
60 C, 15h
HO
–5°C, 2h, then ca. 20°C, 20h
H
N
O
linker
O
linker
O
O
O
O
17
20
penicillin G acylase
pH 7.0, 10% MeOH
37°C, 48h
penicillin G acylase
pH 7.0, 10% MeOH
37°C, 48h
[3] Enzyme Catalysis in Organic Synthesis (Eds.: K. Drauz, H.
Waldmann), VCH, Weinheim, 1995.
[4] Enzyme-labile linker groups for syntheses on solid
supports, see: a) B. Sauerbrei, V. Jungmann, H. Wald-
mann, Angew. Chem. 1998, 110, 1187 ± 1190; Angew.
Chem. Int. Ed. 1998, 37, 1143 ± 1146, and references
therein; b) G. Böhm, J. Dowden, D. C. Rice, I. Burgess,
J.-F. Pilard, B. Guilbert, A. Haxton, R. C. Hunter, N. J.
H
N
O
HO
HO
O
O
O
21
75%
endo/exo = 2.5:1
18
81%
Scheme 4. Mitsunobu and Diels ± Alder reactions on polymeric carrier 7 and
penicillin G acylase mediated release of the target molecules. DEAD ˆ diethyl-
azodicarboxylate, DIC ˆ diisopropylcarbodiimide, DMAP ˆ 4-(dimethylamino)-
pyridine.
+
H2N
I
OH
linker
22
7
EDC, HOBt, DMF
ca. 20°C, 12h
attacks only the phenylacetamide unit and not the ester and
amide groups which are also present.
H
N
I
linker
In a third series of experiments we investigated whether the
enzymatic method also gives access to target molecules that
are coupled to the linker as amides. To this end, 4-iodoaniline
22 was coupled quantitatively to linker 7 and the polymer-
fixed aryliodide 23 obtained thereby was subjected to a Suzuki
reaction[11] with 4-methoxyphenyl boronic acid or a Shille
reaction[13] to give enol ether 26 (Scheme 5). Upon treatment
with 0.5 m hydrochloric acid, enol ether 26 yielded acetophe-
none 27, thereby proving the acid stability of the linker. In
order to release the coupling products, polymers 24 and 27
were incubated with penicillin G acylase at pH 7.0 and room
temperature. As expected under these conditions the phenyl-
acetamide was hydrolyzed. Due to the higher stability of
amides relative to esters, however, the desired cyclization did
not occur at room temperature. Upon warming of the reaction
mixture to 608C, the expected lactam was formed and amines
25 and 28 were released from the polymer.
23
EtO
(HO)2B
OMe
Bu3Sn
[Pd(PPh3)4], K3PO4
DMF/H2O (10:1)
80°C, 20h
[Pd2dba3], AsPh3,
dioxane, 60°C, 24h
OEt
N
OMe
N
H
linker
linker
H
24
26
ca. 20°C, 4h
0.5M HCl/THF (1:1)
penicillin G acylase ca. 20°C, 48h,
pH 7.0, 10% MeOH then 60°C, 4h
O
N
H
linker
H2N
OMe
27
25
60%
penicillin G acylase ca. 20°C, 48h,
pH 7.0, 10% MeOH then 60°C, 4h
These results demonstrate that the new enzyme-labile
safety catch linker allows for the synthesis of different classes
of compounds and the executions of various reaction types.
The release of the products proceeds under particularly mild
conditions with pronounced selectivity and delivers the
desired products in high yield and purity. The linker is stable
under different conditions, even at elevated temperature.
Beyond the development of a new linker system for combi-
O
H2N
28
67%
Scheme 5. Palladium-catalyzed synthesis of anilines on the polymeric
carrier and their enyzmatic release. dba ˆ trans,trans-dibenzylidenacetone,
EDC ˆ N'-(3-dimethylaminopropyl)-N-ethylcarbodiimide.
Angew. Chem. Int. Ed. 2000, 39, No. 9
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3909-1631 $ 17.50+.50/0
1631

COMMUNICATIONS
Turner, S. L. Flitsch, Tetrahedron Lett. 1998, 39, 3819 ± 3822. Safety
catch linkers for combinatorial synthesis, see examples in: c) N. J.
Osborn, J. A. Robinson, Tetrahedron 1993, 49, 2873 ± 2884; d) X.-Y.
Xiao, M. P. Nova, A. W. Czarnik, J. Comb. Chem. 1999, 1, 379 ± 382.
[5] a) I. D. Entwistle, Tetrahedron Lett. 1994, 35, 4103 ± 4106; b) F.
Cubain, Rev. Roum. Chim. 1973, 18, 449 ± 461; c) G. Just, G. Rosebery,
Synth. Commun. 1973, 3, 447 ± 451; d) B. F. Cain, J. Org. Chem. 1976,
41, 2029 ± 2031; e) T. W. Greene, P. G. M. Wuts, Protective Groups in
Organic Synthesis, 3. ed., Wiley, New York, 1999, and references
therein.
[6] For the use of the PhAc group in the deprotection of peptides,
carbohydrates and nucleosides, see: a) H. Waldmann, Liebigs. Ann.
Chem. 1988, 1175 ± 1180, and references therein; b) H. Waldmann, A.
Heuser, A. Reidel, Synlett 1994, 65 ± 67; c) M. A. Dineva, B. Galunsky,
V. Kasche, D. D. Petkov, Bioorg. Med. Chem. Lett. 1993, 3, 2781 ±
2784; d) H. Waldmann, A. Reidel, Angew. Chem. 1997, 109, 642± 644;
Angew. Chem. Int. Ed. Engl. 1997, 36, 647 ± 649.
[7] a) D. Ben-Ishai, J. Sataty, N. Peled, R. Goldshare, Tetrahedron 1987,
43, 439 ± 450; b) H. E. Zaugg, Synthesis 1984, 85 ± 110.
[8] a) E. Bayer, M. Mutter, Nature 1972, 237, 512± 513; b) D. J. Gravert,
K. D. Janda, Chem. Rev. 1997, 97, 489 ± 509.
[9] The polymer was heated to boiling in a 1% aqueous ninhydrin
solution without development of any colour.
[10] a) M. Hiroshige, J. R. Hauske, P. Zhou, Tetrahedron Lett. 1995, 36,
4567 ± 4570; b) T. Jeffery, J.-C. Galland, Tetrahedron Lett. 1994, 35,
4103 ± 4106.
ˆ
( C)4RR'] has further been developed, mainly due to the
work by Fischer et al.[4]
Earlier attempts to prepare metallapentatetraenes, that is
compounds with n ˆ 3 in the above-mentioned formula, date
back to Lomprey and Selegue[5] and somewhat later to Bruce
and co-workers.[6] The latter group, by using [(h5-
C5H5)Ru(PPh3)2(thf)]PF6 and buta-1,3-diyne as the starting
materials, generated in situ a cationic complex containing the
ˆ ˆ ˆ ˆ
fragment Ru C C C CH2 and supported the existence of
this species by trapping reactions with nucleophiles such as
NHPh2, PPh3, H2O, and imines. More recently, both the
groups of Dixneuf[7] and Winter[8] reported about the in situ
formation of cationic intermediates with the molecular unit
Ru C C C CHR and revealed that these can be converted
to corresponding acylvinylidene, acylalkynyl, butenynyl, and
allenylidene ruthenium derivatives. In 1999, a dinuclear
ˆ ˆ ˆ ˆ
ˆ ˆ ˆ ˆ
cationic compound with the core fragment [M] C C C
CH[M'] ([M] ˆ (h5-C5Me5)Fe(ꢁP2ꢁ), ꢁP2ꢁ ˆ 1,2-bis(diphenyl-
phosphanyl)ethane (dppe), 1,2-bis(diisopropylphosphanyl)-
ethane (dippe); [M'] ˆ (h5-C5Me5)Fe(CO)2)) was prepared
by Lapinte and co-workers, using the butadiynediyl complex
[M]C CC C[M'] as the precursor.[9] We have now succeeded
with the isolation and structural characterization of the first
stable neutral compound of the type [LxM C C C CR2]
with LxM ˆ IrCl(PiPr3)2, which fills the gap in the system of
the metallacumulenes B ± E (L ˆ PiPr3).
ꢀ
ꢀ
[11] S. R. Piettre, S. Baltzer, Tetrahedron Lett. 1997, 38, 1197 ± 1200.
[12] S. Berteina, S. Wendeborn, W. K.-D. Brill, A. De Mesmaeker, Synlett
1998, 676 ± 678.
[13] K. A. Bearer, A. C. Siegmund, K. L. Spear, Tetrahedron Lett. 1996, 37,
1145 ± 1148.
ˆ ˆ ˆ ˆ
The First Structurally Characterized Metal
Complex with the Molecular Unit
ˆ ˆ ˆ ˆ
M C C C CR2**
Kerstin Ilg and Helmut Werner*
Dedicated to Professor Henri Brunner
on the occasion of his 65th birthday
The chemistry of metallacumulenes of the general compo-
ˆ
ˆ
sition [LxM C( C)nRR'], which could be considered as near
relatives of metal carbenes, continues to receive a great deal
of attention. While species with n ˆ 1 and 2have already been
extensively investigated,[1] not too much is known about the
compounds with n ˆ 3 and 4. In 1994, Dixneuf and co-workers
reported the synthesis of the first cationic complex with
The synthetic route for the iridium complex of type D
follows the methodology which we used for the corres-
ˆ ˆ ˆ
ponding allenylidene derivatives trans-[IrCl( C C CRR')-
(PiPr3)2].[11d, 12] Since the C3 ligand in the compounds of type C
ꢀ
was generated from propargylic alcohols HC CCRR'OH, for
Ru C C C C CPh2 as the building block,[2] and a few
ˆ ˆ ˆ ˆ ˆ
the preparation of a species with a homologous C4 unit we had
to find a precursor with an additional carbon atom in the Cn
chain. The best choice seemed to be the ethynyl ketone
months later we described the isolation of the first neutral
compound with Ir C C C C CPh2 as the core unit.[3] In the
ˆ ˆ ˆ ˆ ˆ
ˆ
meantime, the field of metallahexapentaenes [LxM C-
[13]
ꢀ
HC CC(O)CHPh2. On reaction of this substrate with the
dihydride 1, the alkynyl(hydrido) complex 2 is formed which,
however, is unstable in solution and isomerizes to the
corresponding vinylidene derivative 3. The existence of 2
can be confirmed by trapping the intermediate with pyridine,
which results in the formation of the octahedral complex 4. In
contrast to the 1H NMR spectrum of 3, those of 2 and 4 display
a triplet resonance in the high-field region at d ˆ À40.96 (for
2) and À21.68 (for 4), the chemical shift of which is
[*] Prof. Dr. H. Werner, Dipl.-Chem. K. Ilg
Institut für Anorganische Chemie der Universität
Am Hubland, 97074 Würzburg (Germany)
Fax : (‡49)931-888-4605
E-mail: helmut.werner@mail.uni-wuerzburg.de
[**] This work was supported by the Deutsche Forschungsgemeinschaft
(SFB 347), the Fonds der Chemischen Industrie (Ph. D. grant to K. I.),
and the BASF AG.
1632
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0570-0833/00/3909-1632 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 9