Facing the gem-dialkyl effect in enzyme inhibitor design: Preparation of homocycloleucine-based azadipeptide nitriles

Article abstract of DOI:10.1002/chem.201101350

A tough nut to crack: In contrast to azadipeptide nitriles containing a P2 leucine residue, the synthesis of the corresponding homocycloleucine-based inhibitors was more difficult; possibly due to the gem-dialkyl effect. A synthetic route to the first hom

Full text of DOI:10.1002/chem.201101350

COMMUNICATION
DOI: 10.1002/chem.201101350
Facing the gem-Dialkyl Effect in Enzyme Inhibitor Design: Preparation of
Homocycloleucine-Based Azadipeptide Nitriles
Maxim Frizler,[a, b] Friederike Lohr,[a] Michael Lꢀlsdorff,[a] and Michael Gꢀtschow*[a, b]
Cathepsin K belongs to the subfamily of papain-like cys-
teine proteases and plays an important role in the (patho)-
physiological processes of bone remodeling. Therefore, cath-
epsin K represents a promising target for the development
of new therapeutic strategies for the treatment of osteoporo-
sis and related diseases.[1–3] The known low-molecular-weight
inhibitors of cathepsin K are mainly peptidomimetic struc-
tures containing an electrophilic group to allow covalent in-
teraction with the active-site cysteine. Among such inhibi-
tors, nitrile-derived compounds attracted the most attention
in various drug-development programs.[2] The fact that the
structural combination of a large P3 substituent with 1-
amino-1-cyclohexanecarboxylic acid (homocycloleucine) at
the P2 position is preferable for the selective inhibition of
cathepsin K has also been explored.[2–4]
Recently, azadipeptide nitriles have been introduced as a
new inhibitor class for papain-like cysteine proteases, which
exhibited Ki values toward human cysteine cathepsins L, S
and K in the picomolar range. The nitrogen atoms of the
aza-amino nitrile moiety of such inhibitors are essentially al-
kylated for reasons of synthetic access.[5] Using the example
of cathepsin K, it was further demonstrated that the struc-
tural optimization of azadipeptide nitriles increases the
enzyme–inhibitor association rates and allows the develop-
ment of selective cathepsin inhibitors.[6] Moreover, falcipain-
inhibiting azadipeptide nitriles with modified P1 residues
displayed antimalarial activity against Plasmodium falcipa-
rum.[7] Recently, organelle-specific drug-delivery systems for
azadipeptide nitriles were reported.[8]
tivity of homocycloleucine, complicate the synthesis of the
target compounds. However, herein we present a successful
synthetic route to the first azadipeptide nitriles with homo-
cycloleucine at the P2 position.
The preparation of a first pair of dipeptide and azadipep-
tide inhibitors with homocycloleucine at the P2 position,
that is, compounds 2 and 4, is shown in Scheme 1. The di-
peptide nitrile 2[9] was synthesized by the reaction of the car-
bobenzyloxy (Cbz)-protected homocycloleucine with amino-
acetonitrile by a mixed anhydride. Compound 2 was ob-
tained after chromatographic separation in a moderate yield
of 53%. The carbamate 2.1 was isolated as the main byprod-
uct, as a result of nucleophilic attack of the aminoacetoni-
trile at the carbonate carbon of the mixed anhydride. When
compound 1 was treated, by the same procedure, with 1,2-
dimethylhydrazine, the desired dimethylhydrazide 3 was
formed in a yield of only 18%. The carbazate 3.1 was identi-
Although azadipeptide nitriles containing a leucine resi-
due at the P2 position could be easily obtained,[5] the synthe-
sis of homocycloleucine-derived counterparts is considerably
more difficult. In this study, we discuss the gem-dialkyl
effect as the reason for an enhanced tendency to undergo
cyclization reactions, which, along with the decreased reac-
[a] M. Frizler, F. Lohr, M. Lꢀlsdorff, Prof. Dr. M. Gꢀtschow
Pharmaceutical Chemistry I
Pharmaceutical Institute, University of Bonn
An der Immenburg 4, 53121, Bonn (Germany)
Fax : (+49)228-732567
E-mail: guetschow@uni-bonn.de
[b] M. Frizler, Prof. Dr. M. Gꢀtschow
International Graduate Research School Biotech-Pharma
Biomedical Center, 53105 Bonn (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201101350.
Scheme 1. Synthesis of carbamate-derived compounds 2 and 4; Bn=
benzyl, NMM=N-methylmorpholine.
Chem. Eur. J. 2011, 17, 11419 – 11423
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11419

fied as the major byproduct of the reaction. Obviously, the
presence of the gem-disubstituted carbon in the homocyclo-
leucine-derived mixed anhydride accounts for low yields of
the desired coupling products. This effect was particularly
strong when 1,2-dimethylhydrazine, a rather weak, but steri-
cally demanding nucleophile, was used. For a comparison,
the mixed anhydride coupling of protected glycine or Ca-
monosubstituted amino acids with aminoacetonitrile or 1,2-
dimethylhydrazine occurred with less pronounced formation
of carbamate or carbazate byproducts, respectively.[5,10] The
yield of the dimethylhydrazide 3 was slightly increased when
compound 1 was treated with 1,2-dimethylhydrazine in the
presence of triethylamine (TEA) and N’-(3-dimethylamino-
Scheme 2. Retrosynthetic route to compound 16.
propyl-N-ethylcarbodiimide
(EDC)/4-dimethylaminopyri-
dine (DMAP) at different temperatures. A further coupling
protocol (2-(1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluroni-
um hexafluorophosphonate/hydroxbenzotriazole (HBTU/
HOBt)) was less advantageous. Next, by reaction with cya-
nogen bromide, compound 3 was converted into the product
4, the first homocycloleucine-based azadipeptide nitrile.[11]
The inhibitory activity of the target compounds 2 and 4
was evaluated toward human cathepsins L, S, K, and B
(Table 1). Although the dipeptide nitrile 2 was confirmed to
be a selective,[9] fast-binding cathepsin K inhibitor, the corre-
sponding azadipeptide nitrile 4 showed slow-binding behav-
ior and was more potent (Ki =1.8 nm).[11]
Table 1. Ki values of homocycloleucine-derived compounds 2, 4, 10, and
16, and literature values of leucine-derived compounds 21–24.
Compound
Ki [nm][a]
cath K
cath L
cath S
cath B
>5000
170
>5000
150
2
4
10
16
>5000
400
>5000
280
>5000
130
>5000
78
55
1.8
13
0.35
35
0.064
Scheme 3. Synthesis of the dipeptide nitrile 10.
21[b]
22[d]
23[e]
24[f]
750
130
n.d.[c]
0.43
>22000
0.36
0.90
940
0.22
0.33
140
0.15
2.9
0.032
prepared, activated to the corresponding mixed anhydride,
and allowed to react with aminoacetonitrile. The obtained
dipeptide nitrile 7 was deprotected with methanesulfonic
acid, followed by a basic extraction. The resulting P2 build-
ing block 8, without further purification, was coupled with
the synthetic equivalent 9 to yield the desired dipeptide ni-
trile 10.[11]
[a] The corresponding standard errors for compounds 2, 4, 10, and 16 are
noted in the Supporting Information (Table S1). [b] Compound 21[10] rep-
resents the l-leucine analogue of 2. [c] Not determined. [d] Compound
22[6] represents the l-leucine analogue of 4. [e] Compound 23[6] represents
the l-leucine analogue of 10. [f] Compound 24[6] represents the l-leucine
analogue of 16.
To prepare the aza-analogous compound 16, further at-
tempts were made to improve the coupling of Cbz-protected
homocycloleucine 1 with 1,2-dimethylhydrazine. When com-
pound 1 was treated with oxalyl chloride for activation, the
N-carboxyanhydride (NCA) 11 was obtained in a yield of
83% (Scheme 4).[12] Under the conditions used, that is, in
the absence of a base,[13] a chloride-promoted displacement
of the benzyl group was operative.[14]
Compound 11 was treated with 1,2-dimethylhydrazine car-
bonitrile 12 to obtain the azadipeptide nitrile 19 as a build-
ing block (Scheme 5), which could have been directly cou-
pled with compound 9, leading to the desired product 16 in
only one further step. However, the transformation of 11
To improve the potency and selectivity for cathepsin K,
we decided to replace the carbamate P3–P2 linker with an
amido group and, in turn, to extend the P3 substituent with
a triaryl moiety, which has already been recognized as an
optimized motif for cathepsin K inhibition.[2,6,9] The envis-
aged compound 16 was retrosynthetically divided into three
synthetic equivalents (Scheme 2).
Although equivalent 9 has already been published,[6] the
synthetic route to 17 had to be established. This route, in a
modified form, was at first applied to the carba analogue of
compound 16, that is, dipeptide nitrile 10 (Scheme 3). tert-
Butoxycarbonyl (Boc)-protected homocycloleucine 6 was
11420
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11419 – 11423

Facing the gem-Dialkyl Effect in Enzyme Inhibitor Design
COMMUNICATION
acetonitrile in the presence of DMAP, the 1,2,4-triazinane-
3,6-dione 13 was obtained, instead of the orthogonally pro-
tected azadipeptide ester 20. The attack of the NH nitrogen
at the carbazate carbon operates in this ring closure. The
formation of 13 represents the first example of a direct
1,2,4-triazinane-3,6-dione cyclization of a Cbz-protected aza-
dipeptide tert-butyl ester.[17,18]
Scheme 4. Formation of the homocycloleucine NCA 11 via the 5(4H)-ox-
azolone 18.
The finally successful synthesis of compound 16, the aza
analogue of 10, is outlined in Scheme 7. The Cbz protecting
group of the dimethylhydrazide 3 was hydrogenolytically re-
moved to obtain 14, which was used in the next step without
further purification. The synthetic equivalent 9 (Scheme 2)
was activated with EDC in the presence of DMAP and
treated with 14, leading to the formation of compound 15.
In the final reaction step with cyanogen bromide, the homo-
cycloleucine-based azadipeptide nitrile 16 with a large P3
substituent was obtained.
Scheme 5. Reaction of 11 with different nucleophiles; DIPEA=diisopro-
pylethylamine.
into 19 was not observed under various conditions, ranging
from a reaction in THF at room temperature to a reaction
in DMF at 1308C in a sealed tube. In contrast, the transfor-
mation of 11 with aminoacetonitrile in THF, carried out in
an autoclave at 1008C, provided homocycloleucyl–glycine–
nitrile 8 in a yield of 72%. Thus, two alternative routes from
5 to 8 (via 6 and 7, or via 1 and 11) were employed.[11]
The unexpectedly low reactivity[15] of the homocycloleu-
cine-derived NCA 11 was considered to result from the
gem-dialkyl effect. It has been well established that this
effect promotes cyclization reactions due to an increased
number of gauche interactions in the open-chain substrates
and/or a relief of steric strain in the cyclic products.[16] How-
ever, the gem-dialkyl effect also leads to an enhanced stabil-
ity of the corresponding cyclic compounds, and therefore we
assume that it accounts for the protection of 11 against hy-
drazinolytic cleavage. As noted above, the mixed anhydride
coupling of Cbz-protected homocycloleucine 1 took place
with low yields for steric interference reasons (Scheme 2).
Additionally, these reactions might be impaired by the gem-
dialkyl effect leading to the formation of heterocyclized
compound(s).
Scheme 7. Synthesis of the azadipeptide nitrile 16.
The target product 16, as well as its carba analogue 10,
was tested on the human cathepsins L, S, K, and B
(Table 1). Again, the azadipeptide nitrile 16 showed a slow-
binding behavior and a lower Ki value than the carba ana-
logue 10, a fast-binding inhibitor. A comparison of the two
aza derivatives 16 and 4 revealed that the introduction of a
large substituent in the P3 position, as well as the replace-
ment of the P3–P2 linker, was advantageous and led to an
improved potency and selectivity for cathepsin K. However,
in comparison with homocycloleucine- and leucine-derived
azadipeptide nitriles (Table 1), the corresponding carba ana-
logues tend to result in a greater selectivity for cathepsin K.
Additionally, the analysis of the progress curves in the
case of slow-binding inhibitors 16 and 4 allowed for the de-
termination of the second-order association rate constants
(kon) as shown in Figure 1 (for kon and koff values, see
To produce the synthetic equivalent 17 (Scheme 2), it was
intended to first protect the terminal hydrazide nitrogen of
3, followed by the deprotection of the a-amino group
(Scheme 6). However, after treatment of 3 with (Boc)2O in
AHCTUNGTRENNUNG
Scheme 6. Formation of 1,2,4-triazinane-3,6-dione 13.
Chem. Eur. J. 2011, 17, 11419 – 11423
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11421

M. Gꢀtschow et al.
petroleum ether (1:1) as eluent to obtain 13 (0.34 g, 71%) as a colorless
oil. 1H NMR (500 MHz, [D6]DMSO): d=1.27–1.34 (m, 1H), 1.46–1.51
(m, 1H), 1.59–1.64 (m, 4H), 1.71–1.76 (m, 2H), 1.99–2.05 (m, 2H), 3.12
(s, 3H), 3.20 (s, 3H), 5.16 (s, 2H), 7.31–7.38 ppm (m, 5H); 13C NMR
(125 MHz, [D6]DMSO): d=22.27, 24.52, 30.58, 32.75, 33.93, 61.51, 68.18,
127.95, 128.29, 128.53, 135.51, 151.93, 152.75, 167.42 ppm; MS (ESI): m/z:
346.4 ([M+H]+).
N-{4-[5-(2-Thienyl)-1,2,4-oxadiazol-3-yl]benzoyl}homocycloleucine 1,2-di-
methylhydrazide (15): Compound 3 (0.75 g, 2.35 mmol) was dissolved in
MeOH (30 mL) and treated with Pd/C (0.075 g). The resulting mixture
was hydrogenated under hydrogen flow at 2 bar and RT for 4 h. Pd/C
was filtered off, and the solvent was removed under reduced pressure to
obtain homocycloleucine 1,2-dimethylhydrazide (14, 0.48 g) as a colorless
oil, which was used without further purification. Compound 9 (0.64 g,
2.35 mmol) was suspended in dry CH2Cl2 (20 mL), treated with DMAP
(0.015 g, 0.12 mmol) and EDC (0.36 g, 2.32 mmol), and stirred at RT for
10 min. Homocycloleucine 1,2-dimethylhydrazide (14, 0.48 g) in dry
CH2Cl2 (10 mL) was added, and the resulting solution was stirred for
18 h at RT. The solvent was evaporated under reduced pressure. The resi-
due was suspended in H2O, extracted with ethyl acetate (3ꢂ30 mL) and
washed with saturated aqueous NaHCO3 (30 mL), H2O (30 mL), and
brine (30 mL). The solvent was dried over Na2SO4 and removed under
reduced pressure. The crude oily product was purified by column chro-
matography with CH2Cl2/MeOH 20:1 as eluent to obtain 15 as a white
Figure 1. Plot of the rates of hydrolysis of Z-Leu-Arg-AMC versus in-
creasing concentrations of 16. Nonlinear regression gave an apparent in-
hibition constant Ki’=(1+[S]/Km)Ki =4.9Æ0.6 nm. Inset: Plot of the kobs
values versus concentrations of 16. Linear regression gave an apparent
second-order rate constant kon’=kon/(1+[S]/Km)=(13Æ2)ꢂ103 mÀ1 sÀ1
.
inhibitors. The gem-dialkyl effect by itself is not necessary
for strong cathepsin K inhibition. Replacement of homocy-
cloleucine in compounds 2, 4, 10, and 16 by l-leucine
(Table 1) led to an even stronger cathepsin K inhibi-
tion.[6,9,10] Moreover, a-methylation of Cbz-l-leucyl–glycine–
nitrile reduced the inhibitory activity of the resulting race-
mate.[9] The possible determination of association and disso-
ciation rate constants in the azadipeptide nitrile series ena-
bled us to estimate the influence of a homocycloleucine/l-
leucine replacement on these parameters. Although koff
values were not affected, the l-leucine-containing inhibi-
tors[6] exhibited greater kon values than 4 and 16 (1700ꢂ
103 mÀ1 sÀ1 and 3300ꢂ103 mÀ1 sÀ1 vs. 60ꢂ103 mÀ1 sÀ1 and 180ꢂ
103 mÀ1 sÀ1).
In this study, we confirmed that homocycloleucine is a
particularly suitable building block for the design of cathe-
psin K inhibitors. The incorporation of this amino acid into
the azadipeptide nitrile scaffold afforded remarkable selec-
tivity and, when combined with the triaryl motif at the P3
position, excellent inhibitory potency.
1
solid (0.38 g, 37% from 3). M.p. 1898C; H NMR (500 MHz, CDCl3): d=
1.29–1.36 (m, 1H), 1.49–1.57 (m, 2H), 1.64–1.72 (m, 5H), 1.93–1.98 (m,
2H), 2.44 (s, 3H), 3.10 (s, 3H), 6.42 (s, 1H), 7.22 (dd, 3J=5.1 Hz, 3J=
3.8 Hz, 1H), 7.66 (dd, 3J=5.1 Hz, 4J=1.3 Hz, 1H), 7.91 (d, 3J=8.5 Hz,
3
4
3
2H), 7.96 (dd, J=3.6 Hz, J=1.1 Hz, 1H), 8.20 ppm (d, J=8.5 Hz, 2H);
13C NMR (125 MHz, CDCl3): d=21.84, 25.35, 32.23, 33.54, 35.67, 59.63,
125.59, 127.46, 127.87, 128.59, 129.56, 132.10, 132.25, 137.29, 164.98,
168.11, 171.67 ppm.
N-{4-[5-(2-Thienyl)-1,2,4-oxadiazol-3-yl]benzoyl}homocycloleucylmeth
ACHTUNGTRENNUNGyl-
AHCTUNGERTGaNNUN zalaninenitrile (16): Compound 15 (0.30 g, 0.68 mmol) was suspended in
MeOH (30 mL) and treated with NaOAc (0.17 g, 2.07 mmol) and BrCN
(0.15 g, 1.42 mmol). The resulting mixture was stirred at RT for 22 h. The
solvent was evaporated under reduced pressure. The residue was sus-
pended in H2O and extracted with ethyl acetate (3ꢂ30 mL), washed with
10% aqueous KHSO4 (30 mL), H2O (30 mL), saturated aqueous
NaHCO3 (30 mL), H2O (30 mL), and brine (30 mL). The solvent was
dried over Na2SO4 and removed under reduced pressure. The crude solid
product was purified by column chromatography with ethyl acetate/pe-
troleum ether (1:1) as eluent to obtain 16 as a white solid (0.12 g, 38%).
M.p. 2068C; 1H NMR (500 MHz, CDCl3): d=1.35–1.40 (m, 1H), 1.49–
1.53 (m, 2H), 1.68–1.79 (m, 3H), 1.97–2.02 (m, 2H), 2.94 (s, 3H), 3.22 (s,
3H), 6.52 (s, 1H), 7.22 (dd, 3J=5.1 Hz, 3J=3.8 Hz, 1H), 7.67 (dd, 3J=
5.1 Hz, 4J=1.3 Hz, 1H), 7.91 (d, 3J=8.5 Hz, 2H), 7.96 (dd, 3J=3.8 Hz,
4J=1.3 Hz, 1H), 8.23 ppm (d, 3J=8.6 Hz, 2H); 13C NMR (125 MHz,
CDCl3): d=21.75, 25.02, 31.94, 33.11, 41.07, 59.50, 114.51, 125.53, 127.53,
128.04, 128.60, 130.13, 132.15, 132.31, 136.13, 165.28, 167.96, 171.73,
173.10 ppm; LC-MSACTHUNTRGNE(UNG ESI) (90% H2O to 100% MeOH in 20 min, then
Experimental Section
100% MeOH to 30 min, diode array detection (DAD) 220.0–400.0 nm)
tr =15.37, 97% purity, m/z: 465.2 ([M+H]+).
Homocycloleucine-NCA (11): Compound 1 (4.20 g, 15.1 mmol) was dis-
solved in dry CH2Cl2 (100 mL). Oxalyl chloride (2.88 g, 22.7 mmol) and
DMF (0.5 mL) were added. The resulting reaction mixture was stirred
for 3 h at RT. CH2Cl2 was removed under reduced pressure. The crude
product was purified by column chromatography with ethyl acetate/pe-
troleum ether (1:1) on silica gel to obtain 11 as white solid (2.11 g, 83%).
M.p. 1148C; 1H NMR (500 MHz, [D6]DMSO): d=1.29–1.37 (m, 1H),
1.44–1.54 (m, 3H), 1.62–1.68 (m, 2H), 1.71–1.74 (m, 4H), 9.43 ppm (s,
1H); 13C NMR (125 MHz, [D6]DMSO): d=20.81, 24.21, 33.42, 62.42,
150.96, 173.82 ppm; MS (ESI): m/z: 187.2 ([M+NH4]+).
Acknowledgements
We thank Janina Schmitz for assistance. This work was supported by the
NRW International Research School Biotech-Pharma (Germany).
1-(Benzyloxycarbonyl)-3,4-dimethyl-1,3,4-triazaspiroACTHNUTRGNEU[GN 5.5]undecan-2,5-
Keywords: azapeptides
enzymes · inhibitors
· cathepsin K · drug design ·
dione (13): Compound 3 (0.44 g, 1.38 mmol) was dissolved in MeCN
(10 mL). The resulting solution was treated with DMAP (8.4 mg,
0.069 mmol) and (Boc)2O (0.45 g, 2.06 mmol). It was stirred for 24 h at
RT. The solvent was removed under reduced pressure. The oily residue
was purified by column chromatography on silica gel with ethyl acetate/
[1] F. Lecaille, J. Kaleta, D. Brçmme, Chem. Rev. 2002, 102, 4459–4488.
11422
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 11419 – 11423

Facing the gem-Dialkyl Effect in Enzyme Inhibitor Design
COMMUNICATION
[2] M. Frizler, M. Stirnberg, M. T. Sisay, M. Gꢀtschow, Curr. Top. Med.
Chem. 2010, 10, 294–322.
corresponding 2-benzyloxy-5(4H)-oxazolone was obtained, see J. H.
Jones, M. J. Witty, J. Chem. Soc. Chem. Commun. 1977, 281–282.
[13] For the formation of NCAs from 5(4H)-oxazolones, derived from
Boc-protected amino acids, see a) S. Mobashery, M. Johnston, J.
Org. Chem. 1985, 50, 2200–2202; b) R. Wilder, S. Mobashery, J.
Org. Chem. 1992, 57, 2755–2756.
[14] As a recent example, for the preparation of glycosylated lysine-N-
carboxyanhydrides from glycosylated Cbz-protected lysine educts,
see J. R. Kramer, T. J. Deming, J. Am. Chem. Soc. 2010, 132, 15068–
15071.
[3] W. C. Black, Curr. Top. Med. Chem. 2010, 10, 745–751.
[4] S. Fustero, V. Rodrigo, M. Sꢃnchez-Rosellꢄ, C. Del Pozo, J. Timone-
da, M. Frizler, M. T. Sisay, J. Bajorath, L. P. Calle, F. J. CaÇada, J. Ji-
mꢅnez-Barbero, M. Gꢀtschow, Chem. Eur. J. 2011, 17, 5256–5260.
[5] R. Lçser, M. Frizler, K. Schilling, M Gꢀtschow, Angew. Chem. 2008,
120, 4403–4406; Angew. Chem. Int. Ed. 2008, 47, 4331–4334.
[6] M. Frizler, F. Lohr, N. Furtmann, J. Klꢆs, M. Gꢀtschow, J. Med.
Chem. 2011, 54, 396–400.
[7] R. Lçser, J. Gut, P. J. Rosenthal, M. Frizler, M. Gꢀtschow, K. T. An-
drews, Bioorg. Med. Chem. Lett. 2010, 20, 252–255.
[8] Y. Loh, H. Shi, M. Hu, S. Q. Yao, Chem. Commun. 2010, 46, 8407–
8409.
[9] J. T. Palmer, C. Bryant, D. X. Wang, D. E. Davis, E. L. Setti, R. M.
Rydzewski, S. Venkatraman, Z. Q. Tian, L. C. Burrill, R. V. Men-
donca, E. Springman, J. McCarter, T. Chung, H. Cheung, J. W. Janc,
M. McGrath, J. R. Somoza, P. Enriquez, Z. W. Yu, R. M. Strickley,
L. Liu, M. C. Venuti, M. D. Percival, J. P. Falgueyret, P. Prasit, R.
Oballa, D. Riendeau, R. N. Young, G. Wesolowski, S. B. Rodan, C.
Johnson, D. B. Kimmel, G. Rodan, J. Med. Chem. 2005, 48, 7520–
7534.
[15] For a review on NCAs, see H. R. Kricheldorf, Angew. Chem. 2006,
118, 5884–5917; Angew. Chem. Int. Ed. 2006, 45, 5752–5784.
[16] For reviews, see a) M. E. Jung, G. Piizzi, Chem. Rev. 2005, 105,
1735–1766; b) P. G. Sammes, D. J. Weller, Synthesis 1995, 1205–
1222; c) C. Galli, L. Mandolini, Eur. J. Org. Chem. 2000, 3117–3125;
d) R. Karaman, Tetrahedron Lett. 2009, 50, 6083–6087; e) T. Y. Luh,
Z. Hu, Dalton Trans. 2010, 39, 9185–9192; f) S. P. Verevkin, M.
Kꢀmmerlin, H. D. Beckhaus, C. Gali, C. Rꢀchardt, Eur. J. Org.
Chem. 1998, 579–584.
[17] For the sodium hydride promoted formation of a bicyclic 1,2,4-tria-
zinane-3,6-dione from an activated azadipeptide ester, see F. Pinnen,
G. Luisi, A. Calcagni, G. Lucente, E. Gavuzzo, S. Cerrini, J. Chem.
Soc. Perkin Trans. 1 1994, 1611–1617.
[10] R. Lçser, K. Schilling, E. Dimmig, M. Gꢀtschow, J. Med. Chem.
2005, 48, 7688–7707.
[18] For a review on the reactivity of the N-Boc protecting group, see C.
Agami, F. Couty, Tetrahedron 2002, 58, 2701–2724.
[11] The synthesis and analytical characterization of compounds 1–4 (in-
cluding different coupling procedures to 3), compounds 6–8, 10–13,
15, 16, 2.1, and 3.1, 1H and 13C NMR spectra, cathepsin inhibition
assays and kinetic parameters are given in the Supporting Informa-
tion.
Received: May 3, 2011
Revised: July 18, 2011
Published online: September 5, 2011
[12] When performing the cyclization of Cbz-protected phenylalanine
with phosphorous pentoxide in the presence of triethylamine, the
Chem. Eur. J. 2011, 17, 11419 – 11423
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11423