Ring closure to β-turn mimics via copper-catalyzed azide/alkyne cycloadditions

Article abstract of DOI:10.1021/jo0516180

Copper-catalyzed azide alkyne cycloadditions of the linear substrates 1 were used to form the cyclic derivatives 2. Computational, NMR, and CD analyses of these compounds indicate that their most favorable conformational states include type I and type II

Full text of DOI:10.1021/jo0516180

Ring Closure to â-Turn Mimics via
Copper-Catalyzed Azide/Alkyne
Cycloadditions
Yu Angell and Kevin Burgess*
Chemistry Department, Texas A & M University,
P.O. Box 30012, College Station, Texas 77842
burgess@tamu.edu
FIGURE 1. Some â-turn nomenclature, semipeptidic mimics
usually prepared via solid-phase reactions, and the solution
phase click strategy featured in this Note.
Received August 2, 2005
Copper-catalyzed azide alkyne cycloadditions of the linear
substrates 1 were used to form the cyclic derivatives 2.
Computational, NMR, and CD analyses of these compounds
indicate that their most favorable conformational states
include type I and type II â-turn conformations. Selectivity
for the dimeric products 6 in these cyclization reactions is
discussed.
FIGURE 2. Substrate and product structures in this study.
R1′ and R2′ denote protected side chains; R1 and R2 are
deprotected ones.
ods to do this. Solid-phase syntheses were used in all our
previous syntheses of dipeptide-containing peptidomi-
metics. These have the advantage of pseudo-dilution
caused by reactive site isolation on solid supports at
moderate to low loading. Solution phase methods, on the
other hand, require ring closure via efficient reactions.
There have been at least two applications of cycloaddition
reactions to ring-close â-turn peptidomimetics,10,11 but
none featuring the copper-catalyzed azide alkyne cy-
cloadditions (i.e., “click reactions”).12 Consequently, this
Note features application of those reactions as depicted
in Figure 2.
Many groups are interested in syntheses of â-turn
peptidomimetics that have potential applications in
medicinal chemistry.1-4 Our focus in this area has been
to constrain dipeptide fragments in appropriate confor-
mations by using macrocyclizations, particularly to form
14-membered rings.5-9 That ring size facilitates confor-
mational arrangements of edge shared C10 systems, one
featuring the dipeptide and the other comprising a
nonpeptidic fragment (Figure 1).
Scheme 1 outlines the preparation of the starting
materials 1. Side chain protection was necessary for the
Lys (Cbz), Glu (Bn), Thr (Bn), Ser (Bn), and Tyr (Bn)
amino acids. These protecting groups render the inter-
mediates relatively lipophilic, hence use of a water-
soluble diimide reagent in the coupling steps facilitated
separation of the products via aqueous extractions.
Finally, the linear precursors 1 were recrystallized and
isolated in an overall average yield of 70%. Thus the
substrates 1 were conveniently accessible in gram amounts
via syntheses that could be performed in parallel.13
Cyclization of the linear precursors gave the “mono-
meric” and “dimeric” products 5 and 6. Two measures
were taken to favor formation of the protected 14-
membered ring products 5: THF solutions of the azides
1 were added to the copper catalyst over 10-14 h via a
syringe pump, and the final concentration was kept low
(approximately 0.001 M). Nevertheless, dimeric products
Many nonpeptidic fragments can be used in the pep-
tidomimetics featured above. This is a means to incor-
porate diversity and to use different chemical reactions
in the ring-closure step. Our bias is to avoid reactions
that make the macrocycles more peptidic (i.e., amide bond
formation) and to favor ones that give the small hetero-
cyclic fragments found in pharmaceuticals. Moreover, we
are currently interested in scalable solution phase meth-
(1) Blomberg, D.; Hedenstrom, M.; Kreye, P.; Sethson, I.; Brick-
mann, K.; Kihlberg, J. J. Org. Chem. 2004, 69, 3055-3508.
(2) Im, I.; Webb, T. R.; Gong, Y.-D.; Kim, J.-I.; Kim, Y.-C. J. Comb.
Chem. 2004, 6, 207-213.
(3) Gu, X.; Cowell, S.; Ying, J.; Tang, X.; Hruby, V. J. Tetrahedron
Lett. 2003, 44, 5863-5866.
(4) Nagula, G.; Urban, J.; Kim, H.-O.; Lum, C.; Nakanishi, H. WO,
03/030907, 2003.
(5) Reyes, S. J.; Burgess, K. Tetrahedron: Asymmetry 2005, 16,
1061-1069.
(6) Lee, H. B.; Zaccaro, M. C.; Pattarawarapan, M.; Roy, S.; Saragovi,
H. U.; Burgess, K. J. Org. Chem. 2004, 69, 701-713.
(7) Lee, H. B.; Pattarawarapan, M.; Roy, S.; Burgess, K. Chem.
Commun. 2003, 1674-1675.
(11) Kahn, M.; Wilke, S.; Chen, B.; Fujita, K. J. Am. Chem. Soc.
1988, 110, 1638-1639.
(8) Pattarawarapan, M.; Zaccaro, M. C.; Saragovi, U.; Burgess, K.
J. Med. Chem. 2002, 45, 4387-4390.
(9) Burgess, K. Acc. Chem. Res. 2001, 34, 826-835.
(10) Cordero, F. M.; Pisaneschi, F.; Batista, K. M.; Valenza, S.;
Machett, F.; Brandi, A. J. Org. Chem. 2005, 70, 856-867.
(12) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem., Int.
Ed. 2001, 40, 2004-2021.
(13) Cheng, S.; Comer, D. D.; Williams, J. P.; Myers, P. L.; Boger,
D. L. J. Am. Chem. Soc. 1996, 118, 2567-2573.
10.1021/jo0516180 CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/19/2005
J. Org. Chem. 2005, 70, 9595-9598
9595

TABLE 1. Copper-Catalyzed Cyclization Reactions of
Compounds 1
SCHEME 1. Preparation of the Linear Substrates
1a
amino acids
5
i + 1 i + 2 5/6 (monomer/dimer)a isolated yield 5 (%)b
a
b
c
d
e
f
Ile
Lys
Lys
Gly
Lys
Gly
Ser
Gly
Ser
39:61
54:46
29:71
23:77
23:76
39:61
20:79
18:82
12
11
8
14
9
9
8
12
Glu
Thr
Gly
Lys
Gly
Ser
Tyr
a R1′ and R2′ denote protected side chains; R1 and R2 are
deprotected ones (see Table 1 for a-h definitions).
g
h
tended to predominate; data for the cyclization reactions
are shown in Table 1.
a Monomer/dimer ratio was based on the UV peak area of HPLC
traces at 254 nm. b Calculated on the basis of the mass after
preparative HPLC separation.
Even though the crude purities of the monomer/dimer
mixtures were high after the cyclization described above,
relatively low yields were isolated after purification via
HPLC. This is mainly because of the limited solubility
of the products in organic and in aqueous solvents. For
example, the crude yield of compounds 5a and 6a
together (39:61) was 92%, but the isolated yield of 5a was
only 12%.
Deprotection of the products 5 (Pd/C, H2, MeOH, or
EtOH) proceeded cleanly to give the unmasked products
2; any losses of material at this stage probably were
attributable to these relatively insoluble materials stick-
ing to the catalyst. Lysine side chains of the peptidomi-
metics 5d and 5e were converted entirely to N,N-
dimethyl and N-ethyl derivatives, respectively, under
these conditions. Alkylation of amino groups during
hydrogenolysis in methanol and ethanol is well-known.14,15
However, it is curious that this was only observed for
the two peptidomimetics containing Gly/Lys amino acids.
Conformational analyses of 2a-c were performed by
simulating preferred conformations via the quenched
molecular dynamics technique16,17 without any physical
constraints. Bond parameters for low energy conforma-
1
tions were then compared to those measured from H
NMR, ROESY, temperature coefficient studies, and
coupling constants. For instance, interactions of the
triazole CH proton with the NH(i + 1) and NH(i + 2)
were informative. Circular dichroism (CD) spectra were
also recorded. Details of this approach are given else-
where18,19 and in the Supporting Information.
Figure 3 shows the low energy conformers that were
identified for compounds 2a-c that best fit with the
observed physical data. These conformations of com-
pounds 2a and 2b resemble type I â-turns, and for 2c a
type II â-turn is most relevant. These data indicate the
compounds can sample turn conformations. However, the
CD data (Supporting Information) do not closely fit the
shape spectra commonly associated with types I and II
turn conformations,20-22 indicating nonideal matches and/
or other important contributions to the conformational
ensemble.
The studies described above successfully generated
monomeric â-turn mimics, the original goal of this work.
(18) Feng, Y.; Wang, Z.; Jin, S.; Burgess, K. Chem.-Eur. J. 1999,
5, 3273-3278.
(14) Mazaleyrat, J.-P.; Xie, J.; Wakselman, M. Tetrahedron Lett.
1992, 33, 4301-4302.
(19) Feng, Y.; Pattarawarapan, M.; Wang, Z.; Burgess, K. J. Org.
Chem. 1999, 64, 9175-9177.
(15) Filira, F.; Biondi, L.; Gobbo, M.; Rocchi, R. Tetrahedron Lett.
1991, 32, 7463-7464.
(20) Bush, C. A.; Sarkar, S. K.; Kopple, K. D. Biochemistry 1978,
17, 4951-4954.
(16) O’Connor, S. D.; Smith, P. E.; Al-Obeidi, F.; Pettitt, B. M. J.
Med. Chem. 1992, 35, 2870-2881.
(21) Johnson, W. C. Annu. Rev. Biophys. Biophys. Chem. 1988, 17,
145-166.
(17) Pettitt, B. M.; Matsunaga, T.; Al-Obeidi, F.; Gehrig, C.; Hruby,
V. J.; Karplus, M. Biophys. J. 1991, 60, 1540-1544.
(22) Johnson, W. C. Proteins: Struct., Funct., Genet. 1990, 7, 205-
214.
9596 J. Org. Chem., Vol. 70, No. 23, 2005

FIGURE 3. (a-c) Best fit low energy conformers for 2a-c,
respectively, and (d) illustrative ROE cross-peaks.
However, it was surprisingly difficult to do this. Dimers
6 were preferred for seven of the eight cases cited in Table
1. Further, azidoalkynes similar to the linear substrates
1 were also tested as cyclization precursors, but without
success. For instance, when substrates 7 and 8 were
subjected to the copper-catalyzed conditions described
above there was no evidence of the monomeric cyclized
product in the ESI-MS of the crude reaction material;
instead, complex mixtures of compounds formed.
The literature reveals that at least four groups have
attempted to use copper-mediated azide/alkyne cycload-
dition reactions to form macrocyclic rings, but “dimers”
were formed in each case.23-26 Finn et al.23 described
dramatic examples in which 76- and 124-membered rings
were formed in preference to the corresponding mono-
meric ones. This result was unexpected particularly since
solid-phase syntheses were used (i.e., conditions in which
there is an inherent dilution effect).
FIGURE 4. Formation of dimeric products in copper-mediated
azide/alkyne cycloadditions may be associated with less popu-
lated “endo-like” conformations relative to less well-ordered,
more populated, exo-like ones.
the kinetic studies27 suggest it contains two copper atoms
and possibly two alkynes. Intermediates A and B are
possible candidates (Figure 4). For the following discus-
sion, intermediate B is assumed to be the important one
(though this is not critical).
In a simple elaboration of the hypothesis presented by
Finn et al., we suggest that selectivity for dimers in these
copper-mediated cyclizations may be explained as follows.
The perfect regioselectivity of the copper-mediated cy-
cloaddition processes for 1,4-disubstituted triazoles would
require the azido alkyne coils in an endo-like conforma-
tion to form a monomeric cyclization product. In the
absence of extenuating circumstances, this conforma-
tional state would be less populated than exo-like ones
that are intrinsically less ordered. The effects in question
are small, but so are the rate differences leading to
mixtures of monomeric and dimeric products. We suggest
Finn and co-workers hypothesized that preferential
formation of dimeric products could be due to coordina-
tion of two alkynes prior to the cycloaddition and/or to
coordination of the catalyst to the peptidic part of the
molecule. The immediate precursor to cycloaddition in
the copper-mediated transformations is not known, but
(23) Punna, S.; Kuzelka, J.; Wang, Q.; Finn, M. G. Angew. Chem.,
Int. Ed. 2005, 44, 2215-2220.
(24) Billing, J. F.; Nilsson, U. J. J. Org. Chem. 2005, 70, 4847-
4850.
(25) Bodine, K. D.; Gin, D. Y.; Gin, M. S. J. Am. Chem. Soc. 2004,
126, 1638-1639.
(26) van Maarseveen, J. H.; Horne, S. W.; Ghadiri, M. R. Org. Lett.
2005, 7, 4503-4506.
(27) Rodionov, V. O.; Fokin, V. V.; Finn, M. G. Angew. Chem., Int.
Ed. 2005, 44, 2210-2215.
J. Org. Chem, Vol. 70, No. 23, 2005 9597

) 8.5 Hz), 7.85 (s, 1H), 7.79 (d, 1H, J ) 7.5 Hz), 7.69 (m, 1H),
7.61 (t, 1H, J ) 7.5 Hz), 7.53 (d, 1H, J ) 6.0 Hz), 7.38-7.31 (m,
5H), 7.24 (t, 1H, J ) 6.0 Hz), 5.00 (s, 2H), 4.67 (dd, 1H, J ) 15.5
Hz, 8.0 Hz), 4.06 (m, 3H), 2.97 (dd, 2H, J ) 12.5 Hz, 6.0 Hz),
1.77 (m, 1H), 1.59 (m, 1H), 1.40 (m, 2H), 1.24 (m, 5H), 0.87 (m,
6H); MS (ESI, m/z) 576 (M + H)+.
that the preference for the exo-like orientation B prevails
over the conformational biases of the peptide part, or
secondary copper coordination effects, that might some-
times favor reactions via endo-like conformations.
Experimental Section
Acknowledgment. Support for this work was pro-
vided by the NIH (MH 070040 and GM076261) and by
the Robert A. Welch Foundation. TAMU/LBMS-Ap-
plications Laboratory directed by Dr. Shane Tichy is also
thanked.
General Procedure for Cyclization. N,N-Diisopropylethyl-
amine (4.0 mmol, 0.70 mL, 20 equiv) was added to a suspension
of CuI (0.4 mmol, 76 mg, 2 equiv) in 200 mL of dry THF. A
solution of 1a-h (0.2 mmol, 1.0 equiv) in 20 mL of dry THF
was added to the suspension slowly over 10 h and stirred for
another 4 h under nitrogen. The suspension was filtered with a
pad of Celite and concentrated. Flash chromatography on silica
gel (3% MeOH in CH2Cl2) gave a mixture of monomeric and
dimeric compounds. Preparative HPLC experiments were run
using 20-90% B in 30 min (solvent A: H2O with 0.1% TFA,
solvent B: CH3CN with 0.1% TFA) to provide 5a-h as white
powder. Data for cyclic isoleucine-lysine mimic 5a: 1H NMR
(DMSO-d6) 8.62 (d, 1H, J ) 8.5 Hz), 8.32 (s, 1H), 8.01 (d, 1H, J
Supporting Information Available: Procedures for the
preparation of compounds 3a-c, 4a-h, 5a-h, 1a-h, 2a-h,
7, and 8. Data for conformational analysis of compounds 2a-
c. This material is available free of charge via the Internet at
http://pubs.acs.org.
JO0516180
9598 J. Org. Chem., Vol. 70, No. 23, 2005