An access to aza-Freidinger lactams and E-locked analogs

Article abstract of DOI:10.1021/ol3030583

Freidinger lactams, possessing a peptide bond configuration locked to Z, are important key elements of conformationally restricted peptidomimetics. In the present work, the C_αH_i+1 unit has been replaced by N, leading to novel aza-Fre

Full text of DOI:10.1021/ol3030583

ORGANIC
LETTERS
2013
Vol. 15, No. 3
448–451
An Access to Aza-Freidinger Lactams and
E‑Locked Analogs
Philipp A. Ottersbach,† Janina Schmitz,† Gregor Schnakenburg,‡ and
,†
€
Michael Gutschow*
Pharmaceutical Institute, Pharmaceutical Chemistry I, University of Bonn,
An der Immenburg 4, 53121 Bonn, Germany, and Institute of Inorganic Chemistry,
University of Bonn, Gerhard-Domagk-Strasse 1, 53121 Bonn, Germany
guetschow@uni-bonn.de
Received November 6, 2012
ABSTRACT
Freidinger lactams, possessing a peptide bond configuration locked to Z, are important key elements of conformationally restricted peptidomimetics.
In the present work, the CRHiþ1 unit has been replaced by N, leading to novel aza-Freidinger lactams. A synthesis to corresponding building blocks and
their E-locked analogs is introduced. The versatile buildings blocks reported here are expected to serve as useful elements in peptide synthesis.
Peptidomimetics are of great interest for drug develop-
ment. This is due to the opportunity to improve selectivity
and potency and to overcome the disadvantages of natural
peptides, such as proteolytic susceptibility and poor bio-
availability. Azapeptides, peptides in which the CRH unit of
at least one amino acid is replaced by nitrogen,1 emerged as
particularly important structures among peptidomimetics.2
Freidinger lactams (1) possess a linkage from CRHi to
to Z (Figure 1).3 Incorporation of Freidinger lactams is
valued as a major contribution to the concept of confor-
mational constraint in peptides.4 Examples of constrained
peptidomimetics related to Freidinger lactams include bi- to
tetracyclic lactams,5 pyrrolinones,6 and imidazolinones.7
Ring-closing metathesis8 and Mitsunobu cyclization9 allow
access to similar peptide lactams.
Niþ1 and thus a configuration of the peptide bond restricted
† Pharmaceutical Institute.
‡ Institute of Inorganic Chemistry.
(1) For a review on azapeptides, see: Proulx, C.; Sabatino, D.; Hopewell,
R.; Spiegel, J.; Ramos, Y. G.; Lubell, W. D. Future Med. Chem. 2011,3, 1139.
(2) (a) Melendez, R. E.; Lubell, W. D. J. Am. Chem. Soc. 2004, 126,
€
6759. (b) Ekici, O. D.; Li, Z. Z.; Campbell, A. J.; James, K. E.; Asgian,
J. L.; Mikolajczyk, J.; Salvesen, G. S.; Ganesan, R.; Jelakovic, S.;
€
Grutter, M. G.; Powers, J. C. J. Med. Chem. 2006, 49, 5728. (c) Verhelst,
S. H. L.; Witte, M. D.; Arastu-Kapur, S.; Fonovic, M.; Bogyo, M.
€
ChemBioChem 2006, 7, 943. (d) Loser, R.; Frizler, M.; Schilling, K.;
€
Gutschow, M. Angew. Chem., Int. Ed. 2008, 47, 4331. (e) Frizler, M.;
€
€
Lohr, F.; Lulsdorff, M.; Gutschow, M. Chem.;Eur. J. 2011, 17, 11419.
(f) Epinette, C.; Croix, C.; Jaquillard, L.; Marchand-Adam, S.; Kellenberger,
C.; Lalmanach, G.; Cadene, M.; Viaud-Massuard, M. C.; Gauthier, F.;
ꢀ
Korkmaz, B. Biochem. Pharmacol. 2012, 83, 788. (g) Proulx, C.; Picard, E.;
Figure 1. (aza-)Freidinger lactams 1, 2 with Z-locked peptide
bond and E-locked (aza)peptides 3, 4.
Boeglin, D.; Pohankova, P.; Chemtob, S.; Ong, H.; Lubell, W. D. J. Med.
Chem. 2012, 55, 6502. (h) Yang, P.-Y.; Wang, M.; Li, L.; Wu, H.; He, C. Y.;
Yao, S. Q. Chem.;Eur. J. 2012, 18, 6528.
r
10.1021/ol3030583
Published on Web 01/15/2013
2013 American Chemical Society

The formal exchange of CRHiþ1 by N results in
Z-configured aza-Freidinger lactams (2). A linkage from
CRHi to CRHiþ1 provides the 5-amino-6-oxopiperidine-
2-carboxylic acid (Apc) derivatives 3 with a constrained
E configuration,10 and an analogous exchange of CRHiþ1
by N yields the corresponding azapeptides of type 4.
As recently demonstrated, methylation of the hydrazine
fragment in model azapeptides leads tothe E configuration
of the respective COꢀN bonds.11 Inspired by this finding
and by the importance of structural motifs 1 and 3, we
envisaged six-membered rings by installing an alkyl bridge
from CRHi to both N atoms in order to lock the peptide
bond into either a Z configuration in aza-Freidinger
lactams 2 or an E configuration in 4. Moreover, for both
cases, N-methylation was implemented as an important
feature of synthetic and natural peptides.12 To establish
this approach, Cbz-protected azadipeptide amides have
been employed as model compounds. As an example to
demonstrate the biological activity of Z- or E-locked
azapeptides, derived azadipeptide nitriles have been de-
signed and evaluated against a panel of pathologically
relevant cysteine cathepsins.
Starting from Cbz-Gln-OH (5) or Cbz-Asn-OH (6),
alkylation under basic conditions13 smoothly provided
methyl esters Cbz-Gln-OMe (7) or Cbz-Asn-OMe (8)
(Scheme 2). The primary amide function in 7 or 8 was
converted with t-BuONO and hydrolyzed to give Cbz-
Glu-OMe (9) or Cbz-Asp-OMe (10) with unprotected side
chains.14 Selective reduction of the carboxylic acid to the
corresponding alcohol function in Cbz-5-hydroxynorva-
line methyl ester (11) was achieved by transforming 9 into
the mixed anhydride with ClCO2i-Bu and subsequent
addition of NaBH4.15 To obtain Cbz-homoSer-OMe
(12), reduction of 10 with BH3ꢀTHF16 was advantageous
with respect to yield. To produce Cbz-5-bromo-norvaline-
methyl ester (13), required for building blocks 14 and 15,
alcohol 11 was subjected to bromination under Appel
conditions (Scheme 3). Next, 13 was reacted with hydrazine
hydrate and underwent 6-exo-trig cyclization. Gratifyingly,
the Z-locked building block 14 was isolated in excellent
yield and purity. The (mono)methylation of hydrazides has
rarely been investigated.17 However, a recently described
two-step protocol,18 involving reductive alkylation with
The initial aim of this work was to find a synthetic access
to the building blocks 14 and 15 for aza-Freidinger lactams
with a peptide bond configuration restricted to Z as well
as the E-analogous building blocks 19 and 20 (Scheme 1).
Key cyclizations should be achieved by reacting Cbz-
5-bromo-norvaline methyl ester (13) or methyl 2-(benzyl-
oxycarbonylamino)-4-oxobutanoate (16) with hydrazine
or methyl hydrazine, respectively.
Scheme 1. Amino Acid-Derived Z- or E-Locked Building
Blocks for Azapeptides
(3) For reviews on Freidinger lactams, see: (a) Freidinger, R. M.
J. Med. Chem. 2003, 46, 5553. (b) Perdih, A.; Kikelj, D. Curr. Med.
Chem. 2006, 13, 1525.
(4) For a review on conformational constraints in peptidomimetics,
see: Hruby, V. J.; Balse, P. M. Curr. Med. Chem. 2000, 7, 945.
(5) (a) Genin, M. J.; Johnson, R. L. J. Am. Chem. Soc. 1992, 114,
8778. (b) Ndungu, J. M.; Gu, X.; Gross, D. E.; Ying, J.; Hruby, V. J.
Tetrahedron Lett. 2004, 45, 3245. (c) Scott, W. L.; Martynow, J. G.;
Huffman, J. C.; O’Donnell, M. J. J. Am. Chem. Soc. 2007, 129, 7077.
(d) RamanaRao, M. H. V.; Pinyol, E.; Lubell, W. D. J. Org. Chem. 2007,
72, 736. (e) Lawandi, J.; Toumieux, S.; Seyer, V.; Campbell, P.; Thielges,
S.; Juillerat-Jeanneret, L.; Moitessier, N. J. Med. Chem. 2009, 52, 6672.
(6) Smith, A. B., III; Charnley, A. K.; Hirschmann, R. A. Acc. Chem.
Res. 2011, 44, 180.
(7) Proulx, C.; Lubell, W. D. Org. Lett. 2012, 14, 4552.
(8) (a) Fink, B. E.; Kym, P. R.; Katzenellenbogen, J. A. J. Am. Chem.
Soc. 1998, 120, 4334. (b) Hoffmann, T.; Lanig, H.; Waibel, R.; Gmeiner,
P. Angew. Chem., Int. Ed. 2001, 40, 3361. (c) Farran, D.; Parrot, I.;
Toupet, L.; Martinez, J.; Dewynter, G. Org. Biomol. Chem. 2008, 6,
Scheme 2. Synthesis of Side Chain Alcohols 11 and 12
€
3989. (d) Pinsker, A.; Einsiedel, J.; Harterich, S.; Waibel, R.; Gmeiner,
P. Org. Lett. 2011, 13, 3503. (e) For a review, see: de Vega, M. J. P.;
ꢀ ~
´
a-Aranda, M. I.; Gonzalez-Muniz, R. Med. Res. Rev. 2011,
Garcı
31, 677.
(9) (a) Malachowski, W. P.; Tie, C.; Wang, K.; Broadrup, R. L.
J. Org. Chem. 2002, 67, 8962. (b) Broadrup, R. L.; Wang, B.; Malachowski,
W. P. Tetrahedron 2005, 61, 10277.
(10) (a) Kemp, D. S.; Sun, E. T. Tetrahedron Lett. 1982, 23, 3759.
(b) Al-Obeidi, F. A.; Micheli, B. J. M.; Barfield, M.; Padias, A. B.; Wei,
Y.; Hall, H. K., Jr. Macromolecules 1999, 32, 6507.
€
(11) Ottersbach, P. A.; Schnakenburg, G.; Gutschow, M. Chem.
Commun. 2012, 48, 5772.
(12) (a) Doedens, L.; Opperer, F.; Cai, M.; Beck, J. G.; Dedek, M.;
Palmer, E.; Hruby, V. J.; Kessler, H. J. Am. Chem. Soc. 2010, 132, 8115.
(b) For a review, see: Chatterjee, J.; Gilon, C.; Hoffman, A.; Kessler, H.
Acc. Chem. Res. 2008, 41, 1331.
(13) Abe, M.; Akiyama, T.; Umezawa, Y.; Yamamoto, K.; Nagai,
M.; Yamazaki, H.; Ichikawa, Y.; Muraoka, Y. Bioorg. Med. Chem.
2005, 13, 785.
Org. Lett., Vol. 15, No. 3, 2013
449

formaldehyde and NaBH3CN, was successfully applied to
give the monoalkylated building block 15.19
Next, we focused on the building blocks 19 and 20,
bearing a COꢀN bond restricted to E. Aldehyde 16 was
prepared from Cbz-homoSer-OMe (12) via Dess-Martin
oxidation (Scheme 4). Installing an aldehyde function in
driven by hydrazone formation involving the NH2 group
of methyl hydrazine, followed by a 6-exo-trig cyclization,
thus directing the substituted nitrogen to the ester. Selec-
tive CꢀN double bond reduction with NaBH3CN af-
forded the E-locked building blocks 19 and 20 in good
yields. With the four building blocks in hand, the con-
strained azadipeptide amides 21ꢀ24 should be synthesized
in the course of N-carbamoylation reactions. There are no
reports on systematic CONH2 group introductions to
hydrazides in the literature. To produce the E-locked
azadipeptide amides 21 and 22, triphosgene21 was em-
ployed as a carbonyl donor (Scheme 5). To prevent the
formation of symmetrical carbonohydrazides, the building
block 19 or 20 was added slowly to 1 equiv of triphosgene,
corresponding to 3 equiv of phosgene.22 Gaseous NH3 was
then introduced to transform the chloroformylated inter-
mediates to the E-locked azapeptides 21 and 22. However,
when applying a triphosgene/NH3 protocol to Z-locked
building block 14, a mixture of several compounds, in-
cluding the symmetrical carbonohydrazide as the major
product, was formed according to LC/MS, and the desired
aza-Freidinger lactam 23 was not isolated. A successful
carbamoylation in this case was achieved with trimethylsi-
lyl isocyanate. The NH(2) function of 14 or 15, respectively,
reacted to N-TMS carbamoyl derivatives, which were
hydrolyzed to give the Z-locked aza-Freidinger lactams
23 and 24 (Scheme 5).23
Scheme 3. Preparation of Z-Locked Building Blocks 14 and 15
Scheme 4. Synthesis of E-Locked Building Blocks 19 and 20
Aftershowing thatthe configuration of the peptidebond
in azapeptides can be restricted to Z or E, respectively, we
were eager to see, if, in principle, the locked building blocks
are suitable nuclei for the design of bioactive compounds.
To choose one application, a nitrile function, a well-known
electrophilic warhead to interact with cysteine proteases,24
was installed. Forthispurpose, methylatedbuilding blocks
15 and 20 were reacted with cyanogen bromide to obtain
configurationally constrained azadipeptide nitriles 25 and
26 (Table 1). Peptide nitriles interact with the active site
thiol of cysteine proteases to form a reversible thioimidate
complex. 25 and 26 were evaluated against a panel of
pathologically relevant cysteine cathepsins K, S, B, L, and
F. Although their activity was only moderate, both nitriles
exhibited different inhibition profiles. In general, proteases
preferentially bind ligands in an extended conformation.25
16, instead of the alkyl halide moiety in13, should facilitate
both nitrogens of (methyl) hydrazine in being incorporated
into the heterocycle. A similar procedure was reported
but was only executed for an N-unmethylated product.20
Aldehyde 16 was subjected to reaction with hydrazine
hydrate or methyl hydrazine, respectively, to yield, indeed,
dihydro-3(2H)-pyridazinones 17 and 18. With respect to
the synthesis of 18, it was expected that the more nucleo-
philic, methylated nitrogen of methyl hydrazine attacks
the aldehyde rather than the ester carbon in a reversible,
nonproductive step. The generation of 18, however, is
Scheme 5. Carbamoylation to Z/E-Locked Azadipeptides
21ꢀ24 via Triphosgene/NH3- or TMSNCO-Assisted Reactions
(14) Adams, D. R.; Bailey, P. D.; Collier, I. D.; Heffernan, J. H.;
Stokes, S. Chem. Commun. 1996, 349.
(15) For a similar protocol, see: Kokotos, G. Synthesis 1990, 299.
(16) Garrard, E. A.; Borman, E. C.; Cook, B. N.; Pike, E. J.; Alberg,
D. G. Org. Lett. 2000, 2, 3639.
(17) Licandro, E.; Perdicchia, D. Eur. J. Org. Chem. 2004, 665.
(18) Loh, Y.; Shi, H.; Hua, M.; Yao, S. Q. Chem. Commun. 2010, 46,
8407.
(19) The enantiopurity of final products 25, 26 was determined
to be >98% according to chiral HPLC analysis (see Supporting
Information (SI)).
(20) Jungheim, L. N.; Boyd, D. B.; Indelicato, J. M.; Pasini, C. E.;
Preston, D. A.; Alborn, W. E., Jr. J. Med. Chem. 1991, 34, 1732.
450
Org. Lett., Vol. 15, No. 3, 2013

Table 1. Preparation of Z- or E-Locked Azadipeptide Nitriles 25, 26, and Ki Values
Ki ( SE (μM)
compd
Z/E
cat. K
cat. S
cat. B
cat. L
cat. F
25
26
Z
E
0.162 ( 0.018
8.67 ( 1.82
7.77 ( 0.72
283 ( 89
0.29 ( 0.01
1.79 ( 0.06
5.15 ( 0.39
3.87 ( 0.25
3.77 ( 0.29
46.5 ( 5.8
Such a conformation is preformed in the azadipeptide
nitrile 25, but not in 26. Actually, 25 was more potent than
26 toward cathepsins K, S, B, and F. In particular, a more
than 30-fold stronger inhibition of cathepsins K and S by
25 was observed. Thus, using the example of cysteine
cathepsins, it was shown that the incorporation of the
building blocks 15 and 20 into bioactive peptides resulted
in a remarkable impact on proteinꢀligand interaction.
In conclusion, a synthetic access to aza-Freidinger lac-
tams and E-locked analogs (21ꢀ24) was developed. The
possibility of constraining the peptide bond in azapeptides
into the Z or E configuration was demonstrated for the
first time. Cysteine protease inhibition was chosen to prove
that stable Z/E prearrangements of the peptide bond in
azapeptides represent an attractive approach for the design
of bioactive compounds. The corresponding building
blocks (14, 15, 19, and 20) are accessible with moderate
effort.26 The key cyclization steps to obtain these depro-
tectable compounds proceeded in good to excellent yields.
They are expected to be useful elements for incorporation
into various peptides and to be sufficiently stable toward
acidic or basic conditions that are required for protecting
group removal. Moreover, the preparation of the corre-
sponding Boc-protected analogs of 14, 15, 19, and 20
would allow for an application in solid-phase synthesis
whenreacted withresin-bound peptideisocyanates.27 Such
constrained peptide mimetics may serve as valuable motifs
to elucidate bioactive conformations required for the
binding of azapeptides to different targets or to improve
their pharmacokinetic properties.
Acknowledgment. This work has been supported by the
German Research Foundation (fellowship of GRK 804 to
P.A.O.). G.S. thanks Prof. A. C. Filippou (University of
Bonn) for support.
Supporting Information Available. Detailed synthetic
procedures, analytical data, NMR spectra, description of
enzymatic assays and representative kinetic plots, chiral
HPLC analyses of 25, 26, molecular plot of 24, an X-ray
crystallographic file in CIF format, electronic structure
calculations. This material is available free of charge via
the Internet at http://pubs.acs.org.
ꢁ
ꢀ
(21) Cotarca, L.; Delogu, P.; Nardelli, A.; Sunjic, V. Synthesis 1996,
553.
(26) For a rough comparison, the preparation of Freidinger lactams 1
(Figure 1) is synthetically less demanding than that of 2, whereas the
routes to 3 and 4 require similar effort. In contrast to 3, aza analogs 4
lack the second chiral center.
(27) (a) Nowick, J. S.; Holmes, D. L.; Noronha, G.; Smith, E. M.;
Nguyen, T. M.; Huang, S.-L. J. Org. Chem. 1996, 61, 3929. (b) Limal, D.;
Semetey, V.; Dalbon, P.; Jolivet, M.; Briand, J.-P. Tetrahedron Lett.
1999, 40, 2749. (c) Narendra, N.; Vishwanatha, T. M.; Sureshbabu, V. V.
Synthesis 2011, 3247.
(22) Eckert, H.; Forster, B. Angew. Chem., Int. Ed. 1987, 26, 894.
(23) For X-ray crystallographic analysis of 24, see SI and CCDC
896854. Compound 24 occurred as a mixture of diastereomers in NMR
spectra due to slow rotation around the NꢀN bond; see ref 11.ꢀH1owever,
DFT calculations gave a value of ΔG‡
= 83 kJ mol for the
K
298
barrier of rotation, indicating that slow rotation only appears on the
NMR time scale and 24 exists as a single compound; see SI.
(24) For a review, see: Frizler, M.; Stirnberg, M.; Sisay, M. T.;
€
Gutschow, M. Curr. Top. Med. Chem. 2010, 10, 294.
(25) For a review, see: Tyndall, J. D. A.; Nall, T.; Fairlie, D. P. Chem.
Rev. 2005, 105, 973.
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 3, 2013
451