Peptide-embedded heterocycles by mild single and multiple aza-Wittig ring closures

Article abstract of DOI:10.1002/anie.200604507

Ring a ring o'azoles: The aza-Wittig cyclization of amino acids and peptides is extremely mild, selective, and versatile. The reaction of amino acid esters and amino acid thioester azides delivers peptidic 1,3-azolines and 1,3-azoles with unsurpassed func

Full text of DOI:10.1002/anie.200604507

Angewandte
Chemie
DOI: 10.1002/anie.200604507
Heterocycles in Biomolecules
Peptide-Embedded Heterocycles by Mild Single and Multiple
Aza-Wittig Ring Closures**
Matthias Riedrich, Surendra Harkal, and Hans-Dieter Arndt*
In a large number of bioactive natural products and microbial
secondary metabolites, 1,3-azolines and 1,3-azoles as well as
oligomers thereof are found as key structural features.[1]
These heterocycles rigidify otherwise flexible backbones,
contribute to lipophilicity, and their presence frequently
correlates with DNA, RNA, and protein binding properties.
Naturally occurring azolines and azoles are biosynthesized by
dehydration and oxidation of Ser-, Thr-, or Cys-containing
peptide precursors.[1b] For the chemical synthesis of 1,3-
azolines and 1,3-azoles, a considerable spectrum of method-
ologies has been devised.[2,3] In an ideal synthetic approach,
easily available building blocks should be assembled and
transformed in a fashion that is tolerant towards the
Scheme 1. General mechanistic proposition for heterocycle synthesis
cornucopia of functional groups typical for nonribosomal
peptide metabolites. Most current methodologies,[3] however,
utilize strong dehydrating reagents, which limits their applic-
ability to complex substrates. Furthermore, anhydrous reac-
tion conditions that are not easily compatible with complex
molecules are often required. A general, mild, chemoselec-
tive, and stereoretentive procedure would be very desirable.
We envisioned that the (conceptual) intramolecular
dehydration might be the key to overcome limitations
associated with current methods. A suitable azide[4] precursor
1 could be used to localize the cyclization site (Scheme 1). The
azide 1 could easily be transformed into an iminophosphor-
ane 2 (Staudinger reaction),[5] which could attack a proximal
carbonyl in an “aza-Wittig” reaction. The putative transition
state 3 would benefit both from a favorable exo geometry as
by intramolecular aza-Wittig condensation(s); X=(CR2)n, S, O, NR,
etc.
reactions.[7] Less electrophilic esters and thioesters have been
found to undergo aza-Wittig cyclizations in select cases.[8] On
the other hand, the treatment of an azido peptide with
phosphanes was reported to lead to fragmentation of the
peptide chain,[9] and amide carbonyls tend to display low
reactivity.[10]
a-Azido acids would be the most general building blocks
for the realization of aza-Wittig ring closures on peptides.
Accordingly, a-azido derivatives of cysteine (6), serine (7),
and threonine (8) were generated on a multigram scale
through diazo transfer[11] and the introduction of suitable
protecting groups.[12,13]
=
well as the attractive antiparallel arrangement of the P N and
=
The thioesters 10, which were easily obtained from azido
thiol 6 and protected amino acids 9, were initially investigated
(Table 1). Gratifyingly, treatment of the thioesters 10 with
phosphanes induced their clean cyclization to the thiazolines
11. Of all the tested phosphanes (PPh3, PBu3, PMe3,
P(OMe)3), PPh3 gave by far the best results. Both aliphatic
as well as aromatic side chains delivered the desired thiazo-
line in high yields as did the problematic Cys (9e) and the
sterically encumbered substrates. The following oxidation[6d]
proceeded cleanly to give the corresponding thiazoles 12.
Notably, the ee values of the final products remained high
throughout the series of substrates. The magnitude of the ee
values reflects the amount of epimerization during thioester
formation. The de values of the respective thioesters 10 and
the thiazolines 11 were identical, and a resubmission of the
thiazoles 12 to the oxidation conditions did not lead to any
reduction in the ee value.
C O dipoles. Driven by the extrusion of phosphine oxide, the
condensation product 4 would be regioselectively formed.
Further oxidation[6] could then provide the unsaturated
heterocycles 5.
Reactions of this type are well documented for aldehydes
and ketones and are typically referred to as aza-Wittig
[*] Dipl.-Chem. M. Riedrich, Dr. S. Harkal, Dr. H.-D. Arndt
Universität Dortmund
Fachbereich Chemie
44221 Dortmund (Germany)
and
Max-Planck-Institut für molekulare Physiologie
Otto-Hahn-Strasse 11, 44227 Dortmund (Germany)
Fax: (+49)231-133-2498
E-mail: hans-dieter.arndt@mpi-dortmund.mpg.de
[**] Funding by the Fonds der Chemischen Industrie (Liebig scholarship
to HDA) and the DFG (Emmy-Noether grant AR493-1) is gratefully
acknowledged. We thank Prof. Dr. H. Waldmann for support and
constant encouragement.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2007, 46, 2701 –2703
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2701

Communications
Table 1: Formation of thiazoles through thiazolines from amino acid
Table 2: Formation of oxazoles through oxazolines from amino acid
thioester azides.[a]
ester azides.[a]
Amino acid
R’
Yield [%][b]
ee [%]
9
10
11
12
Amino acid
R’
Yield [%][b]
ee [%]
13
14
15
16
a
b
c
d
e
f
Gly
Ala
Phe
Val
Cys(Tr)
Thr(tBu)
Me
Me
Me
Me
Me
Allyl
75
76
96
98
82
78
82
98
80
94
90
87
78
96
–
94[c,d]
a
b
c
d
e
f
Ala
Ala
Phe
Cys(Tr)
Glu(Cy)
His(Ts)
Thr(Bn)
Val
H
Me
H
H
H
H
H
H
99
86
88
87
99
83
92
95
84
78
78
97
65
74
18
18
62
79
78
62
75
31
56
43
89–98[c,d]
74
94[c]
95[c]
>96[c]
95[c]
66[e]
68
88[d]
23[d]
76[f]
>96[g]
>98[d]
92[e]
[a] Conditions: a) diisopropylcarbodiimide (DIC), 1 mol% 4-N,N-dime-
thylaminopyridine (DMAP), CH2Cl2, 0!208C, 0.5–8 h. b) PPh3, THF,
ꢀ20!208C, 2–6 h. c) BrCCl3, 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU),
CH2Cl2, 0!208C, 2–4 h. Boc=tert-butyloxycarbonyl, Tr=trityl. [b] Yield
of isolated product. [c] Determined by 1H NMR spectroscopy after
derivatization. [d] Determined by chiral HPLC (Daicel AD). [e] DIC,
1 mol% DMAP, CH2Cl2, ꢀ20!208C, 4 h. [f] 1-Ethyl-3-(3-dimethylami-
nopropyl)carbodiimide (EDC), 1-hydroxybenzotriazole (HOBt), NEt3,
g
h
>96[c]
94[d]
[a] Conditions: a) DIC, 5 mol% DMAP, CH2Cl2, 0!208C, 2–16 h.
b) PPh3, THF, ꢀ208C!408C, 4–18 h. c) BrCCl3, DBU, CH2Cl2, 0!
208C, 2–4 h. Cy=cyclohexyl, Ts=toluene-p-sulfonyl, Bn=benzyl.
[b] Yield of isolated product. [c] Determined by 1H NMR spectroscopy
after derivatization. [d] Determined by chiral HPLC (Daicel AD).
[e] Determined by HPLC after derivatization.
1
CH2Cl2, 08C. [g] Determined to be diastereomerically pure by H NMR
spectroscopy.
Table 3: Multiple aza-Wittig ring closures on amino acid bis-(thio)-
esters.[a]
Our attention then shifted towards the oxazoles, for which
most known preparations necessitate harsh reaction condi-
tions.[3,8] We were pleased to find that azido esters 14, which
are easily available from suitably protected amino acids 13
and the azido alcohols 7 and 8, could be generally engaged in
aza-Wittig transformations under mild conditions (Table 2).
The rather acid-labile oxazolines 15 were obtained in good to
excellent yields for a large variety of side-chain substituents.
Exceptions are electron-withdrawing side chains and/or b-
position branching side chains, both of which led to substan-
tial retardation (Table 2, entries g and h). The oxidation to
oxazoles 16 proceeded smoothly. The resulting products 16
were obtained with minimal loss of stereochemical integrity at
the C2 exomethine carbon center, with the exception of Cys,
for which we found that the epimerization during esterifica-
tion was difficult to suppress.
X
Y
Yield [%][b]
ee [%]
17
18
19
a
b
c
O
S
S
O
S
O
64
51
47
83
30
82[c]
60[f]
n.d.[d]
n.d.[d]
60[e]
64[e]
>96[c]
[a] Conditions: a) PPh3, THF, 0!208C, 4–18 h. b) BrCCl3, DBU, CH2Cl2,
0!208C, 2–4 h. [b] Yield of isolated product. [c] Determined by 1H NMR
spectroscopy after derivatization. [d] n.d.=not determined. [e] Com-
bined yield for two steps. [f] Determined by HPLC after derivatization.
The ease of this aza-Wittig reaction prompted us to apply
this method to the synthesis of oligomeric azole motifs, which
frequently occur in bioactive natural products.[1d] The (thio)-
ester oligomers 17 were assembled from suitable building
blocks in good yields (Table 3)[12] and could be converted into
the desired bis-heterocycles upon exposure to PPh3. In all
cases investigated, the bis-azolines 18 were found as the
exclusive products, which clearly indicates the preference for
five-membered ring closure in all cases. However, the bis-
thiazoline 18b was found to be very prone to autoxidation
(atmospheric O2), whereas the thiazolino-oxazoline 18c was
isolated as a mixture of tautomers featuring a 4–2’ double
bond. Final oxidation[6d] cleanly gave the bis-azoles 19 in all
cases, but conversion of the bis-oxazoline 18a was slow.
Although the ee values of the final products were indicative of
some incidental stereochemical erosion (especially 19b),
improvements are expected with further optimization of the
esterification conditions. The proof of concept of these mild
and practical multiple aza-Wittig cyclizations nevertheless
opens up a multitude of future applications.
To test if the reaction conditions of these aza-Wittig
cyclizations are mild enough, the two dipeptides 20 and 21
were assembled from the respective amino acid building
blocks[12] and bound together to give the stable peptide
thioester 22 (Scheme 2). Despite the presence of acid-, base-,
and dehydration-sensitive functionalities, compound 22
smoothly cyclized under the action of PPh3 to deliver
thiazoline 23 as the sole product. Most notably, no regioiso-
meric condensation products were observed, and the presence
of aqueous solvent (25% v/v water) did not change the
reaction outcome. Products of premature hydrolysis of the
putative iminophosphorane intermediate were not detected,
nor was fragmentation of the thioester or any other major side
products seen (< 5% by HPLC). Oxidation of 23 then cleanly
delivered the thiazole peptide 24. This example demonstrates
2702 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 2701 –2703

Angewandte
Chemie
[2] For broad reviews on the chemistry of 1,3-azoles, see: a) M. R.
Grimmett in Comprehensive Heterocyclic Chemistry II, Vol. 3
(Eds.: A. R. Katritzky, C. W. Rees, E. F. V. Scriven), Pergamon,
Oxford, 1996, pp. 77 – 220; b) F. W. Hartner in Comprehensive
Heterocyclic Chemistry II, Vol. 3 (Eds.: A. R. Katritzky, C. W.
Rees, E. F. V. Scriven), Pergamon, Oxford, 1996, pp. 261 – 318;
c) A. Dondoni, P. Merino in Comprehensive Heterocyclic
Chemistry II, Vol. 3 (Eds.: A. R. Katritzky, C. W. Rees, E. F. V.
Scriven), Pergamon, Oxford, 1996, pp. 373 – 474.
[3] a) H. Vorbrüggen, K. Krolikiewicz, Tetrahedron Lett. 1981, 22,
4471 – 4474; b) A. I. Meyers, D. Hoyer, Tetrahedron Lett. 1985,
26, 4687 – 4690; c) G. Burrell, J. M. Evans, G. E. Jones, G. Stemp,
Tetrahedron Lett. 1990, 31, 3649 – 3653; d) F. Yokokawa, Y.
Hamada, T. Shioiri, Synlett 1992, 153 – 155; e) P. Wipf, C. P.
Miller, Tetrahedron Lett. 1992, 33, 907 – 910; f) N. GalØotti, C.
Montagne, J. Poncet, P. Jouin, Tetrahedron Lett. 1992, 33, 2807 –
2810; g) M. A. Walker, C. H. Heathcock, J. Org. Chem. 1992, 57,
5566 – 5568; h) A. B. Charette, P. Chua, J. Org. Chem. 1998, 63,
908 – 909; i) A. J. Phillips, Y. Uto, P. Wipf, M. J. Reno, D. R.
Williams, Org. Lett. 2000, 2, 1165 – 1168; j) F. Yokokawa, T.
Shioiri, Tetrahedron Lett. 2002, 43, 8679 – 8682; k) S.-L. You, H.
Razavi, J. W. Kelly, Angew. Chem. 2003, 115, 87 – 89; Angew.
Chem. Int. Ed. 2003, 42, 83 – 85; l) A. Sakakura, R. Kondo, K.
Ishihara, Org. Lett. 2005, 7, 1971 – 1974.
[4] For a general review on azide chemistry, see: S. Brꢀse, C. Gil, K.
Knepper, V. Zimmermann, Angew. Chem. 2005, 117, 5320 –
5374; Angew. Chem. Int. Ed. 2005, 44, 5188 – 5240.
[5] a) H. Staudinger, J. Meyer, Helv. Chim. Acta 1919, 2, 635 – 646;
b) Y. G. Gololobov, I. N. Zhmurova, L. F. Kasukhin, Tetrahedron
1981, 37, 437 – 472; c) Y. G. Gololobov, L. F. Kasukhin, Tetrahe-
dron 1992, 48, 1353 – 1406.
[6] a) D. L. Evans, D. K. Minster, U. Jordis, S. M. Hecht, A. L.
Mazzu, A. I. Meyers, J. Org. Chem. 1979, 44, 497 – 501; b) M.
North, G. Pattenden, Tetrahedron 1990, 46, 8267 – 8290; c) A. I.
Meyers, F. Tavares, Tetrahedron Lett. 1994, 35, 2481 – 2484;
d) D. R. Williams, P. D. Lowder, Y.-G. Yu, D. A. Brooks,
Tetrahedron Lett. 1997, 38, 331 – 334.
[7] a) For an overview, see: P. M. Fresneda, P. Molina, Synlett 2004,
1 – 17. b) A stereoenriching aza-Wittig reaction has been
reported recently: D. Lertpibulpanya, S. P. Marsden, I. Rodri-
guez-Garcia, C. A. Kilner, Angew. Chem. 2006, 118, 5122 – 5124;
Angew. Chem. Int. Ed. 2006, 45, 5000 – 5002.
[8] Thiazoline: a) R. Brossmer, H. Mack, Tetrahedron Lett. 1981, 22,
933 – 936; b) G. B. Kok, M. Campbell, B. Mackey, M. von Itz-
stein, J. Chem. Soc. Perkin Trans. 1 1996, 2811 – 2815; c) J. Chen,
C. J. Forsyth, Org. Lett. 2003, 5, 1281 – 1284; Oxazoline:
d) Yu. G. Gololobov, N. I. Gusar, M. P. Chaus, Tetrahedron
1985, 41, 793 – 799; e) J. Mulzer, A. Meier, J. Buschmann, P.
Luger, Synthesis 1996, 123 – 132; f) H. Kato, K. Ohmori, K.
Suzuki, Synlett 2001, 1003 – 1005; g) K. C. Nicolaou, D. W. Kim,
D. Schlawe, D. E. Lizos, R. G. de Noronha, D. A. Longbottom,
Angew. Chem. 2005, 117, 5005 – 5009; Angew. Chem. Int. Ed.
2005, 44, 4925 – 4929.
Scheme 2. Thiazole synthesis on a peptidic oligomer. a) Dicyclohexyl-
carbodiimde (DCC), HOBt, CH2Cl2, 08C, 2 h; b) PPh3, THF, 90%, or
PPh3, THF/H2O 4:1, 69%; c) DBU, BrCCl3, CH2Cl2, 08C!RT.
the profound potential and applicability of the aza-Wittig
cyclization on highly functionalized linear peptidic precursors.
In conclusion, we have shown that the aza-Wittig reaction
can be employed on peptidic azidoesters and peptidic
thioesters to yield oxazoles, thiazoles, and azole dimers in a
clean, mild, and selective fashion. The outlined procedures
were found to be compatible with a wide range of relevant
functional groups and are capable of providing complex,
multiheterocyclic structures directly from simpler, linear
precursors. Both the building blocks as well as the trans-
formations are amenable to iterative as well as multiple ring-
closure strategies. Notably, this intramolecular aza-Wittig
reaction does tolerate the presence of water. In light of these
experimental findings, we anticipate that the aza-Wittig
cyclization may become a broadly applicable tool for the
synthesis of complex, bioactive molecules from peptidic as
well as nonpeptidic origin.
[9] J. W. Back, O. David, G. Kramer, G. Masson, P. T. Kasper, L. J.
de Koning, L. de Jong, J. H. van Maarseveen, C. G. de Koster,
Angew. Chem. 2005, 117, 8160 – 8164; Angew. Chem. Int. Ed.
2005, 44, 7946 – 7950.
Received: November 3, 2006
Published online: March 5, 2007
[10] N. Kumagai, S. Matsunaga, M. Shibasaki, Angew. Chem. 2004,
116, 484 – 488; Angew. Chem. Int. Ed. 2004, 43, 478 – 482.
[11] P. B. Alper, S.-C. Hung, C.-H. Wong, Tetrahedron Lett. 1996, 37,
6029 – 6032.
Keywords: azides · azoles · cyclization · peptides ·
Wittig reactions
.
[12] Details will be disclosed elsewhere.
[1] Important reviews: a) P. Wipf, Chem. Rev. 1995, 95, 2115 – 2134;
b) R. S. Roy, A. M. Gehring, J. C. Milne, P. J. Belshaw, C. T.
Walsh, Nat. Prod. Rep. 1999, 16, 249 – 263; c) V. S. C. Yeh,
Tetrahedron 2004, 60, 11995 – 12042; d) E. Riego, D. Hernµndez,
F. Albericio, M. lvarez, Synthesis 2005, 1907 – 1922; e) Z. Jin,
Nat. Prod. Rep. 2006, 23, 464 – 496.
[13] Although we have not witnessed any single indication for
instabilities of the azides reported in this work, we strongly
caution the use of appropriate protection measures during
handling, especially when heating and/or concentrating azides of
low molecular weight. See also reference [4].
Angew. Chem. Int. Ed. 2007, 46, 2701 –2703
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
2703