Toward very potent, non-covalent organophosphonate inhibitors of cathepsin C and related enzymes by 2-amino-1-hydroxy-alkanephosphonates dipeptides

Article abstract of DOI:10.1016/j.biochi.2013.05.006

Cathepsins play an important role in several human disorders and therefore the design and synthesis of their inhibitors attracts considerable interest in current medicinal chemistry approaches. Due to the presence of a strong sulphydryl nucleophile in the active center of the cysteine type cathepsins, most strategies to date have yielded covalent inhibitors. Here we present a series of non-covalent β-amino-α-hydroxyalkanephosphonate dipeptidic inhibitors of cathepsin C, ranking amongst the best low-molecular weight inhibitors of this enzyme. Their binding modes determined by molecular modelling indicate that the hydroxymethyl fragment of the molecule, not the phosphonate moiety, acts as a transition state analogue of peptide bond hydrolysis. These dipeptide mimetics appear also to be potent inhibitors of other cysteine proteases such as papain, cathepsin B and cathepsin K, thus providing new leading structures for these medicinally important enzymes.

Full text of DOI:10.1016/j.biochi.2013.05.006

Biochimie 95 (2013) 1640e1649
Contents lists available at SciVerse ScienceDirect
Biochimie
journal homepage: www.elsevier.com/locate/biochi
Research paper
Toward very potent, non-covalent organophosphonate inhibitors
of cathepsin C and related enzymes by 2-amino-1-hydroxy-
alkanephosphonates dipeptides
,
b
c
a
Marcin Dra˛g a , Ewa Wieczerzak , Ma1gorzata Pawe1czak , qukasz Berlicki ,
*
Zbigniew Grzonka b, Pawe1 Kafarski a
Department of Bioorganic Chemistry, Faculty of Chemistry, Wroclaw University of Technology, Wybrzeze Wyspianskiego 27, 50-370 Wrocław, Poland
a
_
ꢀ
b
ꢀ
Department of Chemistry, Gdansk University, ul. Sobieskiego 18, 80-952 Gdansk, Poland
c Faculty of Chemistry, University of Opole, ul. Oleska 48, 45-052 Opole, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
Cathepsins play an important role in several human disorders and therefore the design and synthesis of
their inhibitors attracts considerable interest in current medicinal chemistry approaches. Due to the
presence of a strong sulphydryl nucleophile in the active center of the cysteine type cathepsins, most
Received 19 November 2012
Accepted 14 May 2013
Available online 24 May 2013
strategies to date have yielded covalent inhibitors. Here we present a series of non-covalent b-amino-a-
hydroxyalkanephosphonate dipeptidic inhibitors of cathepsin C, ranking amongst the best low-
molecular weight inhibitors of this enzyme. Their binding modes determined by molecular modelling
indicate that the hydroxymethyl fragment of the molecule, not the phosphonate moiety, acts as a
transition state analogue of peptide bond hydrolysis. These dipeptide mimetics appear also to be potent
inhibitors of other cysteine proteases such as papain, cathepsin B and cathepsin K, thus providing new
leading structures for these medicinally important enzymes.
Keywords:
Hydroxyphosphonate
Cathepsin
Inhibitor
Cysteine protease
Ó 2013 Elsevier Masson SAS. All rights reserved.
1. Introduction
Cathepsin C (EC 3.4.14.1) known also as dipeptidyl aminopep-
tidase I (DPPI), dipeptidyl transferase, or cathepsin J, is a multifac-
Cathepsins are ubiquitous lysosomal proteases that are classi-
fied according to their structure and proteins they cleave [1]. Most
of them contain an active site cysteine residue that acts as a
nucleophile and a histidine residue being a general base in the
hydrolysis of a target peptide. Cathepsins play an important role in
the turnover of intracellular proteins and extracellular proteins via
endocytosis. Extracellularly, they have been implicated in tumour
invasion and metastasis [2e6] and as positive mediators of
eted protease-processing enzyme [12]. It is a lysosomal cysteine
hydrolase expressed in most of the mammalian cells with the
richest sources of this enzyme being spleen and kidney [13]. It
catalyses the removal of dipeptides fragments from the unblocked
N-terminus of proteins and peptides. Cathepsin C also plays a key
role in the activation of granule serine peptidases involved in
inflammation, cell-mediated apoptosis and tissue remodelling [14e
16]. Mutations in the cathepsins C gene are responsible for immune
defence deficits such as prepubertal aggressive peridontitis,
Papillon-Lefevre or Haim-Munk syndrome [17e19].
Even though well described from biological function and struc-
tural sides, cathepsin C is not well investigated as a target for the
design of low-molecular weight inhibitors or activity based probes.
Reports on existing inhibitors describe classic mechanism based
covalent modifiers possessing diazomethane, fluoromethylketone
or acyloxymethylketone warheads [20,21]. There is only one report
concerning the synthesis of an activity based probe, employing the
known recognition sequence coupled to a vinyl sulfone derivative
with high specificity toward cathepsin C [22].
apoptosis induced by interferon-g, Fas/APO-1, and TNF-a [7e10].
The ability of tumour cells to invade into the extracellular matrix
has been attributed to the activity of cathepsins released by tumour
cells or associated with the plasma membrane of tumour cells.
Thus, cysteine cathepsins are highly upregulated in a wide variety
of cancers by mechanisms ranging from gene amplification to post-
transcriptional modification. Quite high similarity in the substrate
specificity among individual cathepsins presents a challenge in
finding selective and thus effective cathepsin inhibitors [11].
For several years we have been involved in design and prepa-
ration of organophosphonate inhibitors of cathepsin C. These
* Corresponding author. Tel.: þ48 71 320 4526; fax: þ48 71 3202427.
E-mail address: marcin.drag@pwr.wroc.pl (M. Dra˛g).
0300-9084/$ e see front matter Ó 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.biochi.2013.05.006

M. Dra˛g et al. / Biochimie 95 (2013) 1640e1649
1641
efforts, although not unsuccessful [23,24], have provided only
moderate inhibitors of this enzyme. Besides above work, there is
only one more report on the use of dipeptides containing C-ter-
minal dimethyl 2-amino-1-oxoalkanesphonates (designed as bio-
isosteres of peptide ketones) as moderate inhibitors of other
cysteine protease e calpain I [25]. These results are quite surprising
if one considers that tetrahedral geometry of pentavalent phos-
phorus is widely exploited as mimicking the tetrahedral interme-
diate in peptide bond hydrolysis, serving as transition state
analogues of proteases [26,27]. In many instances peptides con-
taining phosphonate or phosphinate groups in place of the scissile
peptide bond rank among the best inhibitors.
In this paper we report the first successful example for the use of
this approach for the synthesis of potent inhibitors of cysteine
proteases. Our strategy is based on the synthesis of phenyl-
alanylglycine analogues, encompassing requirements for competi-
tive inhibitors of dipeptidyl aminopeptidase I reported earlier
solution of ammonium chloride and 100 ml of water were added
and stirred for next 5 min. The mixture was left to reach room
temperature. The organic phase was washed with 5% aqueous so-
lution of citric acid (2 ꢂ 70 ml), water and brine. Standard workup
of the organic phase gave desired product 1 as yellowish solid,
which was immediately used for the next step without further
purification.
1a: 86% yield, 1H NMR
d
¼ 1.27 (d, 3H, 3J(H,H) ¼ 7.4 Hz, CH3), 1.41
(s, 9H, Boc), 3.06 (m, 2H, CH2), 4.37e4.54 (m, 2H, CH), 5.11 (d, 1H,
NH), 7.19e7.32 (m, 5H, phenyl), 9.38 (s, 1H, formyl).
1b: 14% yield, 1H 1.42 (s, 9H, Boc), 3.02 (m, 2H, CH2), 4.32 (m, 2H,
CH2), 4.88 (m, 1H, CH), 7.24e7.40 (m, 5H, phenyl), 9.66 (s, 1H,
formyl).
2.1.2.2. Dialkyl 2-(N-blocked-L-phenylalanylamino)-1-hydroxyalkane
phosphonates (2)
2.1.2.2.1. General procedure. These compounds were prepared
by reacting 10 mmol of N-blocked- -amino aldehydes and
[23,25,27], namely the presence of
L-amino acids (stereochemical
a
specificity), a free amino group at N-terminus of the peptide, and
blocked C-terminal carboxyl group. In our case the latter is replaced
by a phosphonate ester moiety. In our proof of concept approach
we decided to use “non-invasive” amino acids in the P1 position
(Ala, Gly) and hydrophobic amino acid in P2 position (Phe) as
model compounds.
dimethyl or diethyl phosphite (10 mmol) in DMF (50 ml) at room
temperature for 24 h in the presence of potassium fluoride (2.9 g,
50 mmol), according to the procedure described by Texier-Boullet
and Foucaud [28]. Purification by column chromatography using
5% MeOH in CH2Cl2 as eluent provided analytically pure com-
pounds 2.
2. Experimental section
2.1.2.3. Dimethyl 2-(N-Boc-L-phenylalanyl-L-amino)-1-hydroxypro
panephosphonate (2a). 85% yield;
2
diastereoisomers: (SSS)-
¼ 977.9, 1022.6, 1070.60,
2.1. Chemistry
80%; (RSS)-20%; mp ¼ 45e49 ꢁC; IR:
n
1172.8, 1225.1, 1444.9, 1478.8, 1527.3, 1650.9, 1691.4, 2981.6,
2.1.1. General
3338.6 cmꢀ1
;
31P NMR:
d
¼ 25.28 (RSS), 25.48 (SSS) 1H NMR
Materials were obtained from commercial suppliers (Sigmae
Aldrich, Fluka, Merck) and used without purification unless
otherwise stated. Dichloromethane was dried over P2O5 and
distilled. THF was dried over sodium with benzophenone and
distilled. Reaction progress was monitored by thin-layer chro-
matography on Merck 60 F254 0.1 m silica plates. Unless otherwise
specified, extracts were dried over MgSO4 and solvents were
removed with a rotary evaporator. Infrared spectra were recorded
with a PerkineElmer 1600 Series Fourier transform spectrometer
as thin films on KBr pellets, and only partial data are listed. NMR
spectra were recorded on a Bruker Avance DRX 300 MHz instru-
ment, operating at 300.13 MHz (1H), 121.499 MHz (31P) and
75.46 MHz (13C). Measurements were made in CDCl3 (99.5 at. % D)
or D2O (99.8 at. % D) solutions at temperature 300 K, all solvents
were supplied by Dr Glaser AG (Basel, Switzerland). Chemical
shifts are reported in parts per million relative to TMS or 85%
H3PO4 used as external standards, and coupling constants are
reported in Hertz. Melting points were determined on an Elec-
trothermal 9200 apparatus and are reported uncorrected.
Elemental analyses were performed at Chemistry Department of
University of Wroclaw on Perkin Elmer 2400 CHN. Electrospray
mass spectra were recorded at Chemistry Department of Univer-
sity of Wroclaw using Finnigan Mat TSQ 700 Electrospray mass
spectrometer.
d
¼ 1.26 (d, 3H, 3J(H,H) ¼ 5.7 Hz, CH3), 1.37 (s, 9H, Boc), 3.01 (m,
2H, CH2), 3.82 (d, 6H, 3J (P,H) ¼ 10.5 Hz, 2xOCH3), 3.89 (dd, 1H,
3J(H,H) ¼ 3.8 Hz, 2J(H,P) ¼ 8.9, CH), 4.27 (m, 2H, 2xCH), 7.02 (d,
1H, NH), 7.23e7.31 (m, 5H, phenyl); 13C NMR for (SSS) diaste-
reoisomer:
d
¼ 17.6 (d, 3J(C,P) ¼ 9.7 Hz), 28.22(s), 38.50(s),
47.46(s), 53.50 (d, 3J(C,P)
¼
7.0 Hz), 56.50 (s), 70.74 (d,
1J(C,P) ¼ 161.4 Hz), 80.20 (s), 126.8 (s), 128.51 (s), 129.40 (s),
136.67 (s), 155.37 (s), 172.04 (s); Anal. Calcd for C19H31N2O7P: C,
53.02%; H, 7.26%; N, 6.51%. Found: C, 53.21%; H, 7.40%; N, 6.77%;
ESI ¼ 430.5, 452 (M þ Na)þ.
2.1.2.4. Diethyl 2-(N-Boc-
panephosphonate (2b). Yield 82%; 2 diastereoisomers: (SSS)-78%;
(RSS)-22%, IR:
¼ 1032.2, 1171.2, 1250.3, 1367.1, 1392.2, 1455.3,
1528.4, 1655.0, 2978.9, 3326.8 cmꢀ1 31P NMR:
¼ 23.11 (RSS),
23.45 (SSS) 1H NMR
¼ 1.21e1.39 (m, 9H, 3xCH3), 1.41 (s, 9H, Boc),
L-phenylalanyl-L-amino)-1-hydroxypro
n
;
d
d
2.87e3.12 (m, 2H, CH2), 3.83 (dd, 1H, 3J(H,H)
¼
3.8 Hz,
2J(P,H) ¼ 8.6 Hz, CH) 4.08e4.35 (m, 6H, 2xOCH2, 2xCH), 7.03 (d, 1H,
NH), 7.22 (d, 5H, phenyl); 13C NMR for (SSS) diastereoisomer:
d
¼ 16.45 (s), 17.17 (d, 3J(C,P) ¼ 9.8 Hz), 27.99(s), 38.58(s), 47.43(s),
55.59(s), 62.96 (d, J ¼ 6.8 Hz), 70.93 (d, 1J(C,P) ¼ 160.8 Hz), 79.96 (s),
126.78 (s), 128.47 (s), 129.35 (s), 136.73 (s), 155.30 (s), 171.76 (s);
ESI ¼ 458.3, 480.7 (M þ Na)þ.
2.1.2.5. Diethyl 2-(N-Boc-
L
-phenylalanyl-
L-amino)-1-hydroxyethane
2.1.2. Synthesis of aliphatic N-Boc-a-amino aldehydes (1)
phosphonate (2c). Yield 8%; 2
diastereoisomers: 50:50; IR:
2.1.2.1. General procedure. Oxalyl chloride (0.035 mol) was dis-
solved in 100 ml of dry dichloromethane and cooled to ꢀ60 ꢁC. To
this solution dry DMSO (0.105 mol) in 10 ml of dichloromethane
n
¼ 1022.9, 1049.1, 1168.1, 1249.9, 1367.0, 1524.0, 1664.9, 2931.4,
2978.0, 3313.4 cm
.
ꢀ1 31P NMR:
d
¼ 23.28, 23.40; 1H NMR
¼ 1.31 (t,
d
6H, 3J(H,H) ¼ 6.9 Hz, 2xCH3), 1.39 (s, 9H, Boc), 3.06 (m, 2H, CH2),
was added dropwise. After
5 min, N-Boc-a-amino alcohol
3.40 (m, 2H, CH2), 4.16 (q, 4H, 2xOCH2), 4.36-4.56 (m, 2H, 2xCH),
(0.03 mol) in 10 ml of dichloromethane was added dropwise. The
mixture was stirred intensively for 1 h, while the temperature was
kept between ꢀ60 ꢁC and ꢀ50 ꢁC. After completion of the reaction
(controlled by TLC), triethylamine (0.15 mol) was added and the
mixture was stirred for 15 min. Next, 100 ml of saturated aqueous
5.30 (d, 1H, NH), 7.27 (m, 5H, phenyl); 13C NMR:
d
¼ 16.45 (d,
3J(C,P) ¼ 5 Hz), 28.23(s), 38.92 (d, 2J(C,P) ¼ 13 Hz), 41.51 (s), 55.98
(s), 63.0 (d,
J
¼
6.0 Hz), 63.16 (d,
J
¼
6.2 Hz), 67.16 (d,
1J(C,P) ¼ 163.0 Hz), 80.27 (s), 126.90 (s), 128.60 (s),129.31 (s), 136.73
(s), 155.16 (s), 172.14 (s); ESI ¼ 445.3.

1642
M. Dra˛g et al. / Biochimie 95 (2013) 1640e1649
2.1.2.6. Dimethyl 3-(N-Boc-
ypropanephosphonate (2d). Yield 76%; 2 diastereoisomers: 50:50;
L
-phenylalanyl-
L
-amino)-1(RS)-hydrox-
(RSS), 23.83 (SSS) 1H NMR (D2O):
d
¼ 1.07 (d, 3H, 3J(H,H) ¼ 6.4 Hz,
CH3),1.22 (t, 6H, 3J(H,H) ¼ 7.0 Hz, 2xCH3), 2.91 (m, 2H, CH2), 3.94 (m,
31P NMR:
d
¼ 25.56 (s, 50%), 24.70 (s, 50%); 1H NMR:
¼ 1.39 (s, 9H,
d
5H, 2xOCH2, CH), 4.16 (m, 2H, 2xCH), 7.68 (d, 5H, phenyl); 13C NMR
Boc), 2.07 (m, 2H, CH2CHP), 3.02 (m, 2H, CHCH2), 3.57 (m, 2H,
for (SSS) diastereoisomer:
d
¼ 15.53 (d, 3J(C,P) ¼ 6.0 Hz), 16.50 (d, 3J
3
CH2CH2CHP), 3.80 (d, JHP ¼ 10.3 Hz, 4H, 2xOCH3), 3.84 (m, 1H,
(C,P) ¼ 9.0 Hz), 36.45(s), 46.66 (d, 2J(C,P) ¼ 3.6 Hz), 54.12 (s), 63.85
(d, J ¼ 7.0 Hz), 68.50 (d, 1J(C,P) ¼ 165.3 Hz), 125.84 (s), 127.77 (s),
128.80 (s), 133.42 (s), 165.30 (s); Anal. Calcd for C16H28N2O5ClP: C,
48.65%; H, 7.15%; N, 7.10%. Found: C, 48.78%; H, 7.40%; N, 7.27%;
ESI ¼ 359.4.
CHP), 4.35 (m, 1H, CHCH2), 5.30 (bs, 1H, NH), 7.19e7.33 (m, 5H,
arom.); IR (KBr): 1025 cmꢀ1 (PeOeC, vs), 1250 cmꢀ1 (P]O, vs),
1455 cmꢀ1 (C]C, ms), 1527 cmꢀ1 (C]C, ms), 2927e3354 cmꢀ1 (Ce
H, NeH, OeH, bvs, ms).
2.1.2.7. Diethyl 3-(N-Boc-
L
-phenylalanyl-
L
-amino)-1-hydroxypro
2.1.2.13. Diethyl 2-(L-phenylalanylamino)-1-hydroxyethanephos
phonate hydrochloride (3c). Quantitative yield; 2 diastereoiso
panephosphonate (2e). Yield 66%; 2 diastereoisomers: 50:50; 31P
NMR:
d
¼ 22.30 (s, 50%), 22.40 (s, 50%); 1H NMR:
d
¼ 1.31 (t,
mers: 50:50, 31P NMR (D2O):
d
¼ 24.56 and 24.86; 1H NMR (D2O):
3JHH ¼ 7.7 Hz, 6H, 2xOCH2CH3), 1.39 (s, 9H, Boc), 1.82 (m, 2H,
CH2CHP), 2.96 (m, 2H, CHCH2), 3.60 (m, 2H, CH2CH2CHP), 3.81 (m,
1H, CHP), 4.11 (m, 4H, 2xOCH2), 4.32 (m, 1H, CHCH2), 5.66 (bs, 1H,
NH), 7.19e7.32 (m, 5H, arom.); IR (KBr): 1024 cmꢀ1 (PeOeC, vs),
1249 cmꢀ1 (P]O, vs), 1455 cmꢀ1 (C]C, ms), 1498 cmꢀ1 (C]C, ms),
1529 cmꢀ1 (C]C, ms), 2941e3346 cmꢀ1 (CeH, NeH, OeH, bvs, ms).
d
¼ 1.22 (t, 6H, 3J(H,H) ¼ 7.0 Hz, 2xCH3), 3.06e3.44 (m, 5H, 2xCH2,
CH), 4.09 (m, 5H, 2xOCH2, CH), 7.18-7.28 (m, 5H, phenyl); 13C NMR
(D2O):
d
¼ 15.70 (d, 3J(C,P) ¼ 5.3 Hz), 36.90 (s), 41.35 (s), 54.56 (d,
2J(C,P) ¼ 4.6 Hz), 64.33 (d, 2J(C,P) ¼ 7.6 Hz), 64.50 (d, J ¼ 7.2 Hz),
65.40 (d, 1J(C,P) ¼ 166.1 Hz), 127.45 (s), 129.1 (s), 129.26 (s), 133.75
(s), 169.12 (s), Anal. Calcd for C15H26N2O5ClP: C, 47.31%; H, 6.88%; N,
7.36%. Found: C, 47.21%; H, 6.40%; N, 7.57%; ESI ¼ 345.2.
2.1.2.8. Dimethyl 2-(N-Cbz-L-phenylalanyl-L-amino)-1-hydroxypro
panephosphonate (7a). Yield 58%; 2 diastereoisomers: 20:80; 31P
2.1.2.14. Dimethyl 3-(L-phenylalanyl-L-amino)-1-hydroxypropane
phosphonate hydrochloride (3d). Quantitative yield; 2 diastereoiso
NMR:
d
¼ 25.41 (s, 20%), 25.54 (s, 80%); 1H NMR:
d
¼ 1.18 (d,
3J(H,H) ¼ 6.9 Hz, 3H), 2.99e3.09 (m, 2H, Ph-CH2CH), 3.67 (d, 6H,
OCH3), 3.79 (dd,1H, 3J(H,H) ¼ 3.6, 2J(HP) ¼ 8.9 Hz, CHP), 4.10 (m,1H,
CHCH3), 4.37 (m, 1H, CHCH2C6H5), 4.96 (s, 2H, OeCH2C6H5), 7.14e
7.25 (m, 10H, phenyl); Anal. Calcd for C22H29N2O7P: C, 56.89%; H,
6.29%; N, 6.03%. Found: C, 57.71%; H, 6.55%; N, 5.88%; ESI ¼ 464.8,
486.1 (M þ Naþ).
mers: (SS)-50%; (RS)-50%; 31P NMR (D2O):
d
¼ 26.83 ppm ((1R,30S)
e 50%) i 26.90 ppm ((1S,30S) e 50%); 1H NMR (D2O):
d
¼ 1.36 (m, 2H,
CH2CHP), 2.81 (m, 2H, CHCH2), 3.40 (m, 2H, CH2CH2CHP), 3.62 (m,
1H, CHCH2), 3.52 (d, 3JHP ¼ 10.3 Hz, 4H, 2xOCH3), 3.89 (m, 1H, CHP),
6.95e7.18 (m, 5H, arom.); ESI ¼ 331.7.
2.1.2.15. Diethyl 3-(L-phenylalanyl-L-amino)-1-hydroxypropane
2.1.2.9. Diethyl 2-(N-Cbz-
L
-phenylalanyl-
L
-amino)-1-hydroxypro
phosphonate hydrochloride (3e). Quantitative yield; 2 diastereoiso
panephosphonate (7b). Yield 65%; 2 diastereoisomers: 20:80; 31P
mers: (SS)-50%; (RS)-50%; 31P NMR (D2O):
d
¼ 24.22, 24.31; 1H NMR
3
NMR:
d
¼ 23.15 (s, 20%), 23.28 (s, 80%); 1H NMR:
d
¼ 1.13e1.29 (m,
(D2O):
d
¼ 1.07 (t, JHH ¼ 7.3 Hz, 6H, 2xOCH2CH3), 1.41 (m, 2H,
9H, OCH2CH3, CH3), 2.61 (bs, 1H, OH), 2.88e3.09 (m, 2H,
CHCH2C6H5), 3.73 (dd, 1H, CHP, 3J(H,H) ¼ 4.3, 2J(H,P) ¼ 9Hz), 3.92e
4.12 (m, 5H, OCH2CH3, CHCH2C6H5), 4.29e4.37 (m, 1H, CHCH3), 4.97
(s, 2H, OCH2C6H5), 7.12e7.24 (m, 10H, phenyls); Anal. Calcd for
C24H33N2O7P: C, 58.53%; H, 6.75%; N, 5.69%. Found: C, 58.76%; H,
6.58%; N, 5.92%; ESI ¼ 492.7, 513.9 (M þ Naþ).
CH2CHP), 2.87 (m, 2H, CHCH2), 3.12 (m, 2H, CH2CH2CHP), 3.45 (m,
2H, CHP, CHCH2), 4.11 (m, 4H, 2xOCH2), 7.05e7.25 (m, 5H, arom.);
ESI ¼ 359.3.
2.1.2.16. Monomethyl 2-(N-
panephosphonate hydrochloride (4a). (0.43 g,
Dimethyl N-(Boc)- -phenylalanyl- -alanylphosphonate (2a) was
L
-phenylalanyl-
L
-amino)-1-hydroxypro
1
mmol) of
L
L
2.1.2.10. Dialkyl 2-(N-
phosphonate hydrochlorides (3)
L
-phenylalanylamino)-1-hydroxyalkane
refluxed for 6 h in 5 ml of acetone with NaI (0.15 g, 1 mmol). The
precipitated monoester was filtered off and washed successively
with acetone and diethyl ether. The obtained compound was
treated with 6 M HCl (5 ml per 1 mM of the compound) for
20 min. Precipitated sodium chloride was filtered off and the
solvent was removed under reduced pressure, washed with dry
diethyl ether and dried. Monoester was obtained in 65% yield as
a white semisolid in a form of two diastereoisomers: (SSS)-72%;
2.1.2.10.1. General procedure. Compound 2 was treated with 6 M
hydrochloric acid in methanol (5 ml per 1 mM of the compound) for
15 min at room temperature. Then the solvents were evaporated
under reduced pressure and the obtained residue treated with dry
diethyl ether, decanted and dried in vacuum.
2.1.2.11. Dimethyl 2-(
phosphonate hydrochloride (3a). This compound was obtained as
L
-phenylalanyl-
L
-amino)-1-hydroxypropane
(RSS)-28%; IR:
n
¼ 995.3, 1052.1, 1065.7, 1119.6, 1203.4, 1538.6,
1659.6, 2975.1, 3343.8 cmꢀ1
;
31P NMR:
d
¼ 21.35 (RSS), 22.35
white semisolid. Quantitative yield; 2 diastereoisomers: (SSS)-78%;
(SSS); 1H NMR (D2O):
d
¼ 1.13 (d, 3H, 3J(H,H) ¼ 6.7 Hz, CH3),
(RSS)-22%; 31P NMR (D2O):
d
¼ 26.89 (RSS), 27.36 (SSS) 1H NMR
3.20 (m, 2H, CH2), 3.54 (d, 3H, 3J(P,H) ¼ 10.3 Hz, OCH3), 3.66 (m,
(D2O):
d
¼ 1.08 (d, 3H, 3J(H,H) ¼ 6.0 Hz, CH3), 1.15 (d, 3H,
1H, CH), 4.07 (m, 2H, 2xCH), 7.24 (m, 5H,phenyl); 13C NMR for
3J(H,H) ¼ 6.8 Hz, CH3), 2.89 (m, 2H, CH2), 3.70 (d, 6H, 3J
(SSS) diastereoisomer:
d
¼ 16.93 (d, 3J(C,P) ¼ 7.6 Hz), 36.84 (s),
(P,H)
¼
10.5 Hz, 2xOCH3), 3.98 (dd, 1H, 3J(H,H)
¼
4.0 Hz,
47.44 (d, 2J(C,P) ¼ 4.5 Hz), 52.33 (d, 2J(C,P) ¼ 6.5 Hz), 54.65 (s),
69.28 (d, 1J(C,P) ¼ 159.7 Hz), 125.8 (s), 128.21 (s), 129.15 (s),
133.90 (s), 168.30 (s); Anal. Calcd for C13H22N2O5ClP: C, 44.26%;
H, 6.29%; N, 7.94%. Found: C, 47.35%; H, 6.40%; N, 7.77%;
ESI ¼ 317.1.
2J(H,P) ¼ 10.0, CH), 4.09 (m, 2H, 2xCH), 7.24 (m, 5H, phenyl); 13C NMR
for (SSS) diastereoisomer:
d
¼ 16.68 (d, 3J(C,P) ¼ 9.5 Hz), 36.76 (s),
46.45 (d, 2J(C,P) ¼ 3.2 Hz), 53.97 (d, J ¼ 6.0 Hz), 54.06 (d, J ¼ 6.6 Hz),
54.83 (s), 68.32 (d,1J(C,P) ¼ 165.6 Hz),128.06 (s),129.12 (s),129.42 (s),
133.82 (s), 168.26 (s); Anal. Calcd for C14H24N2O5ClP: C, 45.85%; H,
6.60%; N, 7.64%. Found: C, 46.59%; H, 6.19%; N, 7.52%; ESI ¼ 331.3.
2.1.2.17. Monoethyl 2-(
panelphosphonate hydrobromide (4b). (0.458 g, 1 mmol) of diethyl
N-(Boc)- -phenylalanyl- -alanylphosphonate (2b) was dissolved in
L-phenylalanyl-L-amino)-1-hydroxypro
2.1.2.12. Diethyl 2-(
phosphonate hydrochloride (3b). Quantitative yield;
astereoisomers: (SSS)-78%; (RSS)-22%; 31P NMR (D2O):
L
-phenylalanyl-
L
-amino)-1-hydroxypropane
di-
¼ 23.26
L
L
2
5 ml of dry dichloromethane. To this solution bromotrimethylsilane
(2 mmol, 0.306 g) was added and stirred for 24 h. The solvents were
d

M. Dra˛g et al. / Biochimie 95 (2013) 1640e1649
1643
Fig. 1. Reagents and conditions for synthesis of dipeptidyl derivatives of 2-amino-1-hydroxyalkanephosphonates: (a) CO2Cl2, DMSO, CH2Cl2; (b) (R1O)2P(O)H/KF/Al2O3.
removed under reduced pressure and the resulting residue treated
with 1 ml of methanol. The product was precipitated as a white
solid by addition of dry diethyl ether (5 ml). Yield 78%; 2 di-
astereoisomers: (SSS)-92% and (RSS)-8%; mp ¼ 216e219 ꢁC; IR:
2.2. Molecular modelling
The crystal structures of human cathepsin C (with Gly-Phe-
CHN2 bound) and cathepsins K (complex with azepanone-based
inhibitor) [29,30] were obtained from Protein Data Bank refcode
2DJF and 1NL6, respectively [31]. These structures were used as a
starting point for all calculations. Residue numbering is according
to this structure. The hydrogen atoms were added using Insight
2000 program (Insight 2000, Molecular Modelling Program Pack-
age, Accelrys 2000), the protonation states of the amino acid side
chain residues were set to pH 7.0 and the protonated structures
were checked carefully. Molecular mechanics calculations were
performed using the Discover program (DISCOVER, Molecular
Modelling Program Package, Accelrys, 2000) with cff97 force field
n
¼ 1003.8, 1062.9, 1116.8, 1201.5, 1466.9, 1538.4, 1658.3, 2985.3,
3238.9, 3338.9 cmꢀ1
NMR (D2O):
;
31P NMR :
d
¼ 21.75 (RSS), 22.34 (SSS); 1H
d
¼ 1.11e1.21 (m, 6H, 2xCH3), 2.92e3.22 (m, 2H, CH2),
3.69 (dd, 1H, 3J(H,H) ¼ 5.1 Hz, 2J(P,H) ¼ 9.8 Hz, CH), 3.92 (qu., 2H,
OCH2), 4.10 (m, 2H, 2xCH), 7.20-7.29 (d, 5H, phenyl); 13C NMR for
(SSS) diastereoisomer:
d
¼ 15.91 (d, 3J(C,P) ¼ 5.3 Hz), 16.85 (d,
3J(C,P) ¼ 7.5 Hz), 36.70 (s), 47.23 (d, 2J(C,P) ¼ 4.5 Hz), 54.56 (s), 62.72
(d, 2J(C,P) ¼ 6.8 Hz), 69.24 (d, 1J(C, P) ¼ 161.5 Hz), 127.98 (s), 129.18
(s), 129.45 (s), 133.83 (s), 168.30 (s); Anal. Calcd for C14H24N2O5BrP:
C, 40.89%; H, 5.88%; N, 6.81%. Found: C, 39.74%; H 5.44%; N, 6.95%;
ESI ¼ 331.3.
2.1.2.18. 2-(
hydrobromide (5a). (0.43 g, 1 mmol) of dimethyl N-(Boc)-
nylalanyl- -alanylphosphonate (2a) was dissolved in 5 ml of dry
L
-phenylalanyl-
L
-amino)-1-hydroxypropanephosphonate
L
-phe-
L
dichloromethane. To this mixture bromotrimethylsilane (8 mmol,
1.224 g) was added and stirred for 24 h at room temperature. The
solvent was removed in vacuum and the residue treated with 1 ml
of methanol. The product was precipitated as a white solid by
addition of dry diethyl ether (4 ml). Yield 94%; 2 diastereoisomers:
(SSS)-90% and (RSS)-10%; mp ¼ 206e208 ꢁC; IR:
n
¼ 998.8, 1064.1,
1104.0, 1178.1, 1235.1, 1437.7, 1464.3, 1538.6, 1586,5, 1661.1, 2974.0,
3235.3, 3346.5 cmꢀ1
;
31P NMR
d
¼ 19.55 (RSS) and 20.48 (SSS); 1H
NMR (D2O):
d
¼ 1.15 (t, 3H, 3J(H,H) ¼ 6.7 Hz, CH3), 2.93e3.26 (m,
2H, CH2), 3.62 (dd, 1H, 3J(H,H) ¼ 5.4 Hz, 2J(P,H) ¼ 9.7 Hz, CH),
4.05e4.14 (m, 2H, 2xCH), 7.19e7.31 (m, 5H, phenyl); 13C NMR
(D2O):
d
¼ 17.01 (d, 3J(C,P) ¼ 7.6 Hz), 36.66 (s), 47.30 (d,
2J(C,P) ¼ 5.3 Hz), 54.59 (s), 69.68 (d, 1J(C,P) ¼ 159.2 Hz), 127.90 (s),
129.13 (s), 129.36 (s), 133.85 (s), 168.28; Anal. Calcd for
C12H20N2O5BrP: C, 37.61%; H, 5.26%; N, 7.31%. Found: C, 39.67%; H,
5.04%; N, 7.00%; ESI ¼ 302.7.
2.1.2.19. 3-(L-Phenylalanyl-L-amino)-1-hydroxypropanephosphonate
hydrobromide (5d). This compound was obtained from (3d) using
the same methodology as for (5a). Yield 69%; 2 diastereoisomers:
(SS)-50% and (RS)-50%; IR:
n
¼ 1064 cmꢀ1 (P-OH, vs), 1210 cmꢀ1
(P]O, vs), 1455 cmꢀ1 (C]C, ms), 1537 cmꢀ1 (C]C, ms), 2967e
3349 cmꢀ1 (CeH, NeH, OeH, bvs, ms); 31P NMR
d
¼ 18.06 ppm
((RS) e 50%) i 18.12 ppm ((SS) e 50%); 1H NMR (D2O):
d
¼
d
¼ 1.46e
1.85 (m, 2H, CH2CHP), 3.02 (m, 2H, CHCH2), 3.21 (m, 2H,
CH2CH2CHP), 3.45 (m, 1H, CHP), 4.03 (m, 1H, CHCH2), 7.15e7.33 (m,
5H, arom.); ESI ¼ 302.5.
Fig. 2. Derivatisation of dipeptidyl 2-amino-1-hydroxyalkanephosphonates. Reagents
and conditions: (a) HCl, CH3OH; (b) NaI, acetone, reflux; (c) 1 equiv. (CH3)3SiBr, CH2Cl2,
r.t. and then CH3OH; (d) 8 equiv. (CH3)3SiBr, CH2Cl2, r.t. and then CH3OH.

1644
M. Dra˛g et al. / Biochimie 95 (2013) 1640e1649
Table 1
Inhibition of cathepsin C by phosphonate dipeptide mimics. Data represent the mean and S.D. of at least triplicate experiments. K*i represents the value calculated for the
steady-state reaction with the use of Dixon and LineweavereBurk equations.
Compound
Stereoisomer ratio
k1 [Mꢀ1sꢀ1
]
k2 [sꢀ1
]
Ki [m
M]
Ki*
[mM]
3a
3b
3c
4a
4b
5a
6a
6b
6c
50:50
80:20
50:50
78:22
92:8
38.24
30.35
6.70 ꢂ 10ꢀ3
7.60 ꢂ 10ꢀ3
7.92 ꢂ 10ꢀ3
7.30 ꢂ 10ꢀ3
8.88 ꢂ 10ꢀ3
6.27 ꢂ 10ꢀ3
5.8 ꢂ 10ꢀ3
1.08 ꢂ 10ꢀ2
5.1 ꢂ 10ꢀ3
174 ꢃ 4.87
250 ꢃ 60
178 ꢃ 4.09
194 ꢃ 58.2
3.42 ꢂ 105
1.43 ꢂ 105
129.06
74.26
0.023 ꢃ 0.012
0.051 ꢃ 0.016
70.6 ꢃ 17.65
87.1 ꢃ 27.00
499 ꢃ 2.495
1550 ꢃ 3.1
0.012 ꢃ 0.0085
0.076 ꢃ 0.0251
53.7 ꢃ 7.18
90:10
61.1 ꢃ 7.495
501 ꢃ 2.505
1548 ꢃ 6.19
3140 ꢃ 175.84
10.74
6.98
3.88
50:50
3310 ꢃ 198.6
and conjugate gradient minimizer. Minimizations were done up to
an energy change of 0.01 kcal molꢀ1. Compounds present in active
sites of both cathepsins complexes were modified using the Builder
module of Insight package in order to obtain appropriate phos-
phonate inhibitors. Structures were optimized in two steps. First,
the structure of the inhibitor was optimized, while the enzyme was
frozen, second, the positions of all atoms of enzyme active site and
inhibitor were minimized.
P e concentration of the product of the reaction at time; Vs
steady-state velocity; V0 e initial velocity; kapp e apparent first-
e
order rate constant; t e reaction time;
3
e baseline deviations
correction.
kapp, V0 and VS were calculated using “Slow-binding” computer
program, which uses iterative Marquardt algorithm to find optimal
kapp, V0 and VS from Eq. (1).
Papain was assayed by the method adopted from the literature
[34]. Benzoyl-DL-arginine-4-nitroanilide (Bachem) was used as the
substrate in chromogenic assays at pH 6.5 carried out to determine
the active enzyme concentration in solutions of papain (Sigma) by
titration with known amounts of the irreversibly binding papain
inhibitor, E-64 (Sigma). This papain was then used to determine the
amount of papain inhibitory sites in solutions of inhibitors, through
a dilution series and measurement of residual papain activity. Stock
solutions of compounds were prepared by dissolving in 5%
dimethyl sulfoxide (DMSO) in water.
2.3. Enzymatic studies
Cathepsin C was isolated according to the previously described
procedure [32]. Enzymatic reactions were assayed at 37 ꢁC in
0.1 M acetate buffer (pH 5) containing NaCl (30 mM final con-
centration), EDTA-Na2 (1 mM final concentration) and 2-
mercaptoethanol (5 mM final concentration). The assay mixture
contained: 20
m
l of cathepsin C (1.6 mU), 2 ml of glycyl-
L-
Papain, Cathepsin B and Cathepsin K. Continuous-rate assays
with the fluorogenic substrate, benzyloxycarbonyl-phenylalanyl-
arginyl-aminomethyl-coumarin (Bachem) were used to study the
rate of inactivation of papain (above), bovine cathepsin B (Sigma)
and recombinant human cathepsin K (produced in Pichia pas-
toris) at 37 ꢁC [35]. Assays were carried out in 50 mM sodium
phosphate buffer, containing 1 mM dithiothreitol (DTT) and
1 mM ethylenediaminetetraacetic acid (EDTA), adjusting the pH
to 6.5 for papain, and pH 6.0 for cathepsins B and K. A Perkine
Elmer LS50 fluorimeter set at 360 nm excitation and 460 nm
emission wavelengths was used, and the data were analysed with
FLUSYS program [36]. Observed rates of inactivation (kobs) were
plotted against [I] to determine rate constants for association
(kþ1) and dissociation (kꢀ1). Steady-state rates before (V0) and
after (Vi) addition of inhibitor were used to calculate equilibrium
phenylalanine p-nitroanilide (Gly-
L
-Phe-p-nitroanilide HCl)
$
(substrate, 0.3e3 mM final concentration) in acetate buffer and
0.5 ml of inhibitor concentration dependent on inhibitor) in water.
The course of reaction was monitored by recording absorbance at
410 nm [
D
E410 ¼ 9000 Mꢀ1 cmꢀ1] with a Varian Cary 100 UVeVIS
spectrophotometer. Michaelis constant (KM ¼ 2.28 mM) and
inhibitory kinetic data were obtained using the computer pro-
gram “Slow-binding” provided kindly by Dr J. Hurek (University of
Opole). Slow-binding inhibition was described by Eq. (1) accord-
ing to reference [33]:
ꢀ
ꢁ
vo ꢀ vs
app$t
P ¼ Vs$t þ
$ 1 ꢀ eꢀk
þ
3
(1)
kapp
constants for dissociation of enzyme-inhibitor complexes (Ki(app)
)
absence of inhibitor
0,001
0,002
0,003
0,004
0,135
0,2
0,18
0,16
0,14
0,12
0,1
4500
3,16E-03
4000
1,58E-03
3500
3,16E-04
8,53E-04
3000
2500
2000
1500
1000
500
0
0,08
0,06
0,04
0,02
-0,0006
-0,0004
-0,0002
0
0,0002
0,0004
-500
0
0
-1000
200
400
600
800
1000
1200
1400
Time (s)
[I], M
Fig. 3. Glycyl-L-phenyloalanine p-nitroanilide hydrolysis by Cathepsin
increasing concentration of 4a and absence of 4a.
C
at the
Fig. 4. Dixon plot for the hydrolysis reaction of Gly-L-Phe-p-NA substrate catalysed by
bovine cathepsin C in the presence of L-Phe-L-Ala-CHOH-PO3H2 (5a).

M. Dra˛g et al. / Biochimie 95 (2013) 1640e1649
1645
0,016
0,014
0,012
0,010
0,008
0,006
0,004
Fig. 6. Structures and activity toward cathepsin C of the most active dipeptidyl 2-
amino-1-hydroxyalkanephosphonate inhibitors 3c and 4a.
diastereomers several column chromatographic systems were
applied, albeit unsuccessfully.
As shown in Fig. 2, protecting groups were removed by a com-
bination of standard methods of peptide chemistry and those
specifically used in phosphonopeptide chemistry affording the
desired phosphonopeptide inhibitors of cathepsin C. Thus, selective
deprotection of the amino moiety in compounds 2 was easily
achieved by action of 6 M hydrogen chloride in methanol (saturated
solution), a procedure that yielded diesters 3 quantitatively. Mon-
odealkylation of compound 2a with NaI in boiling acetone was
performed according to the literature [41] followed by deprotection
of the amino moiety with 6 M hydrogen chloride in methanol
yielding the desired 4a. However, this method appeared unsuc-
cessful for diethyl ester 2b. We have found that the application of an
equimolar ratio of trimethylbromosilane led selectively to the
desired monodealkylated compound 4b. Moreover, during this re-
action simultaneous deprotection of amino group was also ach-
ieved and therefore total removal of both groups was obtained in
one step by the use of 2 mol of bromomethylsilane per 1 mol of 2b.
Total deprotection of phosphonate moiety in 4a leading to acid 5a
was achieved with excess of trimethylbromosilane (8 equivalents)
with very high (94%) yield (Fig. 2).
0,000
0,005
0,010
[I],
0,015
0,020
µ
M
Fig. 5. kapp versus concentration of phosphonic analogue L-Phe-Gly-CHOH-PO3Et2 3c
(Cath C).
[37]. These values were corrected for substrate competition
using Km values determined for the substrate batch under the
assay conditions employed (42, 55, and 9.1
cathepsin B and cathepsin K, respectively) to obtain true Ki
mM for papain,
values.
3. Results and discussion
3.1. Chemistry
The synthesis of dipeptidyl derivatives of 2-amino-1-hydroxy
alkanephosphonates is illustrated in Fig. 1. The synthesis of start-
ing substrates, dipeptidyl aldehydes 1, was performed by oxidation
of the appropriate N-Boc-dipeptide alcohols by Swern procedure
[38]. In order to avoid undesired decomposition and probable
racemisation of the unstable aldehydes, the crude products were
used immediately in the next step of synthesis without purification.
Their reaction with dimethyl and diethyl phosphites was per-
formed according to Texier-Boullet procedure [28] and resulted in
mixtures of diastereomers of hydroxyphosphonates 2 in yields
3.2. Inhibitory activity
The obtained phosphonate derivatives are a new class of very
promising inhibitors of cathepsin C from bovine spleen (Table 1).
They appeared to be slow-binding (Fig. 3) and competitive (Fig. 4)
inhibitors of the enzyme.
Detailed analysis of kinetic parameters according to previously
established slow-binding inhibition models revealed that com-
pounds 3a, 3b, 3c, 4a, 4b, 5a, 6a, 6b and 6c demonstrate type A
(SB-A) inhibition, in which initial reaction rate is independent
of inhibitor concentration (Table 1) and linear dependance of
kapp from inhibitor concentration is observed (Fig. 5) [42e44]
(Scheme 1).
from 8 to 80%. Efforts to obtain dimethyl 2-(N-Boc-L-phenylalanyl)
amino-1-hydroxymethanephosphonate were unsuccessful because
of the very low yield of the addition reaction (w1%) as determined
by 31P NMR. This resulted probably from low stability of the starting
amino aldehyde in the reaction conditions. The obtained com-
pounds were mixtures of two diastereomers: (SSS) and (SSR) in the
Slope analysis of kapp ¼ f(I) yielded k1, while the intersection
with the y axis yielded k2 (Eq. (2)). Subsequently, the Ki value was
case of derivatives of 2-(
phosphonic acid and (SS) and (SR) in the case of derivatives of 2-
-phenylalanyl)amino-1-hydroxymethanephosphonic acid. Upon
addition of dialkyl phosphite to aldehydes derived from -phenyl-
alanyl- -alanine the formation of (SSS) diastereomer was privileged
L-phenylalanyl)amino-1-hydroxyethane
(L
L
L
3
3
(72e92% de) as determined by analysis of their JPC and JHP
coupling constants according to the previously described proce-
dure [39,40]. In the case of reaction with L-phenylalanylglycinal
the formation of an equimolar ratio of both diastereomers
was observed. In order to separate the resulting mixtures into
k1
E+ I
EI
k2
Fig. 7. L-phenylalanine-b-alanine dipeptidyl derivatives of 2-amino-1-hydroxy- alka-
Scheme 1. Diagram of a slow-binding inhibition according to A mechanism [44].
nephosphonates used in the studies.

1646
M. Dra˛g et al. / Biochimie 95 (2013) 1640e1649
Table 2
Inhibition of cathepsin C by phosphonate dipeptide mimics with beta alanine in the P1 position. Data represent the mean and S.D. of at least triplicate experiments. Ki*
represents the value calculated for the steady-state reaction with the use of Dixon and LineweavereBurk equation.
Compound
Stereoisomer Ratio
k1 [sꢀ1
]
k2 [sꢀ1
]
k3 [Mꢀ1sꢀ1
]
k4 [sꢀ1
]
Ki [m
M] ꢃ SD
Ki*
[m
M] ꢃ SD
3e
3d
5d
50:50
50:50
50:50
9.21 ꢂ 10ꢀ3
5.13 ꢂ 10ꢀ3
3.81 ꢂ 10ꢀ3
7.76 ꢂ 10ꢀ2
8.90 ꢂ 10ꢀ2
9.45 ꢂ 10ꢀ2
184
115
85
3.75 ꢂ 10ꢀ3
1.71 ꢂ 10ꢀ2
2.49 ꢂ 10ꢀ2
20 ꢃ 1.8
21.03 ꢃ 1.05
402 ꢃ 14.99
355 ꢃ 24.85
416 ꢃ 12.5
379 ꢃ 25.77
vs
calculated from Eq. (3). Calculated Ki values correlate very well
with K*i values calculated using LineweavereBurk and Dixon
plot analysis (example in Fig. 4) for the reaction in steady-state
(Table 1).
k4 ¼ kapp
$
(5)
v0
k4
k3
¼
(6)
KI
k1
kapp
¼
$½Iꢄ þ k2
(2)
Calculated Ki values for 3d, 3e and 5d correlate very well with Ki*
values calculated using LineweavereBurk and Dixon plot analysis
for the reaction in steady-state (Table 2).
½Sꢄ
1 þ
kM
The inhibitory action of dipeptides containing 2-amino-1-
hydroxyethanephosphonic acids was also compared with phos-
phonodipeptides composed of L-phenylalanine and aminomethane
k2
k1
KI ¼
(3)
phosphonic (compounds 6a and 6b) or 1-aminoethanephosphonic
(compound 6c) acids, compounds lacking hydroxymethylene frag-
ment (Fig. 9). These derivatives appeared to be very weak inhibitors
with inhibitory constants in the millimolar range, thus demon-
strating the importance of the hydroxygroup for the observed strong
inhibitory effect (Table 1).
To the best of our knowledge compounds 3c and 4a towards
cathepsin C rank amongst the highest binding affinities of low-
molecular inhibitors of this enzyme (Fig. 6).
In order to study how the extension of the length of peptide
chain by one methylene group influences inhibitory activity we also
prepared compounds 3d, 3e and 5d (Fig. 7). These were obtained in
In order to determine if the dipeptide mimetics containing 1-
hydroxy-2-aminoethanephosphonic acids might represent
a
a similar manner as those described above using Boc-
L-phenyl-
general class of inhibitors of cysteine proteases, their inhibitory
potency towards papain, bovine cathepsin B and human cathepsin
K was studied. Most of the compounds studied were N-protected,
because these proteases do not tolerate a free amino group at N-
terminus of the peptide for their activity (Fig. 10). Results collected
in Table 3 clearly indicate that all the compounds studied
appeared to act as very efficient inhibitors of the three enzymes
investigated. Remarkably, the most active of them, compounds 7a
and 7b, were one order of magnitude more potent than the most
active inhibitors of cathepsin C with their inhibitory constants
being within nanomolar range for papain and cathepsin K. Inter-
estingly, compound 8, representing the simple phosphono
dipeptide, appeared to be a quite potent and selective inhibitor of
cathepsin K.
alanyl- -alanine as starting substrate. These compounds
b
a
appeared to be moderate inhibitors of cathepsin C of potency
similar to most of the studied compounds (with exception of 3c and
4a) (Table 2).
Detailed analysis revealed that these compounds display type C
(SB-C) inhibition, in which kapp versus inhibitor concentration has
hyperbolic character (Fig. 8) [42e44] (Scheme 2).
k1, k2, k3 and KI values were calculated using “Slow-binding in-
hibition” and KI was calculated from Eq. (4):
k1
½Sꢄ
k2
kapp
¼
þ
(4)
½Iꢄ
1 þ
1 þ
kM
kI
The findings described above indicate that the application of
phosphonate mimetics of dipeptides might be also considered as a
promising method for designing new inhibitors of cysteine pro-
teases on the papain family (MEROPS Clan CA) [46].
k4 was calculated from Eq. (5), while k3 was calculated from
Eq. (6) [45].
0.025
0.020
0.015
0.010
0.005
0.000
3.3. Molecular modelling
In order to determine if the phosphonate moiety of these in-
hibitors could act as a transition state analogue of hydrolysed
peptide bond, their binding mode was studied by means of mo-
lecular modelling. Our results did not allow discerning simple
structureeactivity relationship for inhibitors of cathepsin C, since
minute changes in their chemical structure are reflected in quite
k1
k2
k3
k4
0
200
400
600
[I],
800
1000
1200
E* + I
E + I
E*I
M
Fig. 8. kapp versus concentration of phosphonic analogue L-Phe-b-Ala-CHOH-PO3Et2 3e
(Cath C).
Scheme 2. Diagram of a slow-binding inhibition according to C mechanism.

M. Dra˛g et al. / Biochimie 95 (2013) 1640e1649
1647
Fig. 9. Dipeptidyl derivatives of 2-aminoalkanephosphonates used in the studies.
Fig. 10. N-terminally protected dipeptidyl derivatives of 2-amino-1-hydroxyalkanephosphonates and 2-aminoalkanephosphonates used in the studies.
significant changes in their affinities towards the enzyme. This
analysis is further complicated by the fact that introduction of hy-
droxyl moiety into the molecule is accompanied by the introduc-
tion of an additional center of the chirality. Moreover, the formation
of some of the stereoisomers (including one of the most active
compound 4a) was not equimolar. Therefore, we decided to use
molecular modelling in order to establish the modes of binding of
the inhibitors to cathepsins C and K.
Starting from the crystal structure of human cathepsin C bound
with inhibitor [29], the models were obtained after appropriate
modification of the inhibitor structure and optimization of the
resulting complexes between the enzyme and both stereoisomers
and 3c. It appeared that isomer of (S,S) diastereomer (corre-
sponding to L,D configuration) is bound more effectively
comparing to its epimer (S,R). The structure of the inhibitor fits
tightly to the enzyme active site potentially forming several
hydrogen bonds and hydrophobic interactions (Fig. 11). The pep-
tide backbone of 3c interacts in a similar manner to that observed
for the analogous fragment in the crystal structure of Gly-Phe-
CHN2-cathepsin C complex. The terminal ammonium group of the
inhibitor shares two strong hydrogen bonds with Asp1 and Gly277
residues (the first interaction is charge assisted). The amide NH
moiety of 3c interacts with the carbonyl of Asn380. The aromatic
fragment of the inhibitor is located in a deep hydrophobic cleft
formed by Phe278, Tyr64, Thr379, Pro279, Ala382 and Ala349
residues. However, the mode of binding of the hydrox-
ymethanephosphonate fragment is the most interesting. The ox-
ygen atom from the OH group is located similarly to the oxygen
atom in C]O group of phenylalanine residue in Gly-Phe-CHN2-
cathepsin C complex. This atom forms a hydrogen bond with NH
of Cys234. Thus, this fragment of molecule, with an sp3 carbon
atom bearing hydroxyl moiety, most likely represents transition
state analogy to peptide bond hydrolysis. The phosphonate group,
which is commonly used as actransition state mimic, does not play
this role here and only forms two subsequent hydrogen bonds
with Gln228 and His381 side chains.
The second highly active hydroxyphosphonate cathepsin C in-
hibitor e compound 4a interacts with the enzyme in a similar
manner as 3c (Fig. 11C and D). The main peptide chain interactions,
as well as the aromatic moiety position within the binding site, are
nearly identical in both cases. The phosphonate monomethyl group
forms two strong hydrogen bonds with His381 and Gln228 and a
weak interaction of this type with Trp405 side chain. Analogously
to inhibitor 3c, the hydroxy function of inhibitor 4a of (S,S,S) dia-
stereoisomer forms a hydrogen bond with the Cys234 amide group.
Moreover, the inhibitor OH is a weak hydrogen bond donor to the
Gln228 side chain carbonyl.
Subsequently, the mode of binding of hydroxyphosphonate in-
hibitors to cathepsin K was studied starting from the known crys-
tallographic structure of the enzyme bound with an azepanone
inhibitor [30]. The complex of the most active compound 7a with
the enzyme was optimized and intermolecular interactions were
found (Fig. 12). Both aromatic fragments of the molecule exhibit
important hydrophobic interactions e a phenylalanine side chain
fits into S2 binding pocket formed by Ala134, Leu160, Leu209 and
Tyr67, while the benzyl moiety of N-benzyloxycarbonyl group is
located in the cleft formed by Asp61, Gly64, Gly65, Gly66 and Tyr67.
The peptide backbone chain of the inhibitor forms a hydrogen bond
with Gly66. Interestingly, the existence of an intramolecular
hydrogen bond of the NHAla/O]CCbz type is suggested strongly by
the computational results. The hydroxyphosphonate fragment in-
teracts with the enzyme analogously to the interactions found in
3c-cathepsin C complex. Thus, the hydroxy group oxygen atom is
an acceptor of a hydrogen bond from amide NH of Cys25, while
phosphonate P]O forms hydrogen bond with His162 side chain.
This suggests once more that the hydroxymethyl fragment of the
Table 3
Inhibition of papain, bovine cathepsin B and recombinant human cathepsin K by
phosphonate dipeptide derivatives. Ki values were calculated as detailed in
Experimental section, from results of 3e5 continuous-rate assays if not otherwise
stated, and presented as mean value ꢃ SD.
Compound Stereoisomer Ki [mM]
ratio
Papain
Cathepsin B
Cathepsin K
7a
7b
2d
2e
5d
8
80:20
80:20
50:50
50:50
50:50
0.0012 ꢃ 0.0001 0.021 ꢃ 0.006 0.00032 ꢃ 0.00003
0.0020 ꢃ 0.0002 0.060 ꢃ 0.005 0.0012 ꢃ 0.0001
NI
0.021a
0.80 ꢃ 0.20
2.33a
0.050a
0.075
33.4a
0.010a
NI
NI
>3a
0.081a
a
Based on single assay.

1648
M. Dra˛g et al. / Biochimie 95 (2013) 1640e1649
Fig. 11. Modelled mode of binding of inhibitors 3c and 4a to human cathepsin C (panels A, B and C, D respectively). The inhibitor is represented as a ball and stick model, while the
active site residues are shown as sticks. The molecular surface of the enzyme (panels A and C) is colored according to electrostatic potential (red e negative, grey e neutral, and blue
e positive). Inhibitor-enzyme hydrogen bond interactions are marked as green solid lines (panels B and D).
Fig. 12. Optimized structure of 7a-cathepsin K complex. The inhibitor is represented as ball and stick model, while active site residues as sticks. The molecular surface of the enzyme
(panel A) is colored according to electrostatic potential (red e negative, grey e neutral, and blue e positive. Inhibitor-enzyme hydrogen bond interactions are marked as green solid
lines (panel B).
molecule, rather than the phosphonate moiety, acts as a transition
state analogue.
phenylalanyl)amino-1-hydroxyalkanephosphonates appeared to be a
class of very potent inhibitors of papain, cathepsin B and cathepsin K.
These compounds might be considered as leading compounds for the
design and synthesis of new inhibitors, in the expectation that they
might act as lead structures in future studies.
4. Conclusions
Diethyl 2-(L-phenylalanyl)amino-1-hydroxymethanephosphonate
(3c) and monomethyl 2-(L-phenylalanyl)amino-1-hydroxyethane
Acknowledgements
phosphonate (4a) appeared to be potent inhibitors of cathepsin C,
thus providing new leads for the design of inhibitors of other cysteine
proteases of Clan CA. Moreover, (N-benzyloxycarbonyl-L-
This work was supported by grant NN 302 276437. Marcin Drag
is grateful to Foundation for Polish Science for support. Calculations

M. Dra˛g et al. / Biochimie 95 (2013) 1640e1649
1649
[22] F. Yuan, S.H. Verhelst, G. Blum, L.M. Coussens, M. Bogyo, A selective activity-
based probe for the papain family cysteine protease dipeptidyl peptidase I/
cathepsin C, J. Am. Chem. Soc. 128 (2006) 5616e5617.
[23] M. Pawelczak, K. Nowak, P. Kafarski, Synthesis of phosphono dipeptides, in-
hibitors of cathepsin C, Phosphorus Sulfur 132 (1998) 65e71.
[24] A. Mucha, M. Pawel, J. Hurek, P. Kafarski, Synthesis and activity of phosphinic
tripeptide inhibitors of cathepsin C, Bioorg. Med. Chem. Lett. 14 (2004) 3113e
3116.
[25] M. Tao, R. Bihovsky, G.J. Wells, J.P. Mallamo, Novel peptidyl phosphorus de-
rivatives as inhibitors of human calpain I, J. Med. Chem. 41 (1998) 3912e3916.
[26] B. Lejczak, P. Kafarski, Biological activity of aminophosphonic acids and their
short peptides, Top. Heterocycl. Chem. 20 (2009) 31e63.
were done using hardware and software resources, including
Accelrys programs of the Supercomputing and Networking Center
in Wroclaw. We would like to thank Guy S. Salvesen for critical
reading of the manuscript.
References
[1] N.D. Rawlings, F.R. Morton, C.Y. Kok, J. Kong, A.J. Barrett, MEROPS: the
peptidase database, Nucleic Acids Res. 36 (2008) D320eD325.
[2] M.M. Mohamed, B.F. Sloane, Cysteine cathepsins: multifunctional enzymes in
cancer, Nat. Rev. Cancer 6 (2006) 764e775.
[3] V. Turk, J. Kos, B. Turk, Cysteine cathepsins (proteases) e on the main stage of
cancer? Cancer Cell 5 (2004) 409e410.
[27] R.M. Metrione, N.L. MacGeorge, The mechanism of action of dipeptidyl
aminopeptidase. Inhibition by amino acid derivatives and amines; activation
by aromatic compounds, Biochemistry 14 (1975) 5249e5252.
[28] F. Texier-Boullet, A.A. Foucaud,
A convenient synthesis of dialkyl 1-
hydroxyalkane-phosphonates using potassium or cesium fluoride without
solvent, Synthesis (1982) 165e166.
[4] T. Nomura, N. Katunuma, Involvement of cathepsins in the invasion, metas-
tasis and proliferation of cancer cells, J. Med. Invest. 52 (2005) 1e9.
[5] C. Palermo, J.A. Joyce, Cysteine cathepsin proteases as pharmacological targets
in cancer, Trends Pharmacol. Sci. 29 (2008) 22e28.
[6] M. Drag, G.S. Salvesen, Emerging principles in protease-based drug discovery,
Nat. Rev. Drug Discov. 9 (2010) 690e701.
[7] C. Beers, K. Honey, S. Fink, K. Forbush, A. Rudensky, Differential regulation of
cathepsin S and cathepsin L in interferon gamma-treated macrophages, J. Exp.
Med. 197 (2003) 169e179.
[8] L.P. Deiss, H. Galinka, H. Berissi, O. Cohen, A. Kimchi, Cathepsin D protease
mediates programmed cell death induced by interferon-gamma, Fas/APO-1
and TNF-alpha, EMBO J. 15 (1996) 3861e3870.
[9] S. Kamolmatyakul, W. Chen, Y.P. Li, Interferon-gamma down-regulates gene
expression of cathepsin K in osteoclasts and inhibits osteoclast formation,
J. Dent. Res. 80 (2001) 351e355.
[10] C.E. Chwieralski, T. Welte, F. Buhling, Cathepsin-regulated apoptosis,
Apoptosis 11 (2006) 143e149.
[29] A. Molgaard, J. Arnau, C. Lauritzen, S. Larsen, G. Petersen, J. Pedersen, The
crystal structure of human dipeptidyl peptidase I (cathepsin C) in complex
with the inhibitor Gly-Phe-CHN2, Biochem. J. 401 (2007) 645e650.
[30] R.W. Marquis, Y. Ru, S.M. LoCastro, J. Zeng, D.S. Yamashita, H.J. Oh, K.F. Erhard,
L.D. Davis, T.A. Tomaszek, D. Tew, K. Salyers, J. Proksch, K. Ward, B. Smith,
M. Levy, M.D. Cummings, R.C. Haltiwanger, G. Trescher, B. Wang,
M.E. Hemling, C.J. Quinn, H.Y. Cheng, F. Lin, W.W. Smith, C.A. Janson, B. Zhao,
M.S. McQueney, K. D’Alessio, C.P. Lee, A. Marzulli, R.A. Dodds, S. Blake,
S.M. Hwang, I.E. James, C.J. Gress, B.R. Bradley, M.W. Lark, M. Gowen,
D.F. Veber, Azepanone-based inhibitors of human and rat cathepsin K, J. Med.
Chem. 44 (2001) 1380e1395.
[31] H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H. Weissig,
I.N. Shindyalov, P.E. Bourne, The protein data bank, Nucleic Acids Res. 28
(2000) 235e242.
[32] J. Ken McDonald, P.X. Callahan, S. Ellis, [22] Preparation and specificity of
dipeptidyl aminopeptidase I, Methods Enzymol. 25C (1972) 272e281.
[33] M. Golicnik, J. Stojan, Slow-binding inhibition: a theoretical and practical
course for students, Biochem. Mol. Biol. Educ. A Bimonth. Publ. Int. Union
Biochem. Mol. Biol. 32 (2004) 228e235.
[11] Y. Choe, F. Leonetti, D.C. Greenbaum, F. Lecaille, M. Bogyo, D. Bromme,
J.A. Ellman, C.S. Craik, Substrate profiling of cysteine proteases using
a
combinatorial peptide library identifies functionally unique specificities,
J. Biol. Chem. 281 (2006) 12824e12832.
[34] M. Abrahamson, Cystatins, Methods Enzymol. 244 (1994) 685e700.
[35] C.J. Linnevers, M.E. McGrath, R. Armstrong, F.R. Mistry, M.G. Barnes, J.L. Klaus,
J.T. Palmer, B.A. Katz, D. Bromme, Expression of human cathepsin K in Pichia
pastoris and preliminary crystallographic studies of an inhibitor complex,
Protein Sci. 6 (1997) 919e921.
[12] D. Turk, V. Janjic, I. Stern, M. Podobnik, D. Lamba, S.W. Dahl, C. Lauritzen,
J. Pedersen, V. Turk, B. Turk, Structure of human dipeptidyl peptidase I
(cathepsin C): exclusion domain added to an endopeptidase framework cre-
ates the machine for activation of granular serine proteases, EMBO J. 20
(2001) 6570e6582.
[36] N.D. Rawlings, A.J. Barrett, FLUSYS: a software package for the collection and
analysis of kinetic and scanning data from PerkineElmer fluorimeters, Com-
put. Appl. Biosci. 6 (1990) 118e119.
[37] P.J. Henderson, A linear equation that describes the steady-state kinetics of
enzymes and subcellular particles interacting with tightly bound inhibitors,
Biochem. J. 127 (1972) 321e333.
[38] A. Mancuso, D. Swern, Activated dimethyl sulfoxide: useful reagents for
synthesis, Synthesis (1981) 165e185.
[39] R. Latajka, M. Drag, R. Gancarz, P. Kafarski, Solution conformations of 2-
amino-1-hydroxy-2-aryl ethylphosphonic acids and their diethyl esters,
J. Mol. Struct. 605 (2002) 79e84.
[40] M. Drag, R. Latajka, E. Gumienna-Kontecka, H. Kozlowski, P. Kafarski, Ster-
eoselective synthesis, solution structure and metal complexes of (1S,2S)-2-
amino-1-hydroxyalkylphosphonic acids, Tetrahedron-Asymmetr. 14 (2003)
1837e1845.
[41] J. Jacques, M. Leclercq, M.-J. Brienne, La Formation de Sels Augmente-T-Elle La
Fréquence des Déboublements Spontanés? Tetrahedron 37 (1981) 1727e
1733.
[13] S.W. Dahl, T. Halkier, C. Lauritzen, I. Dolenc, J. Pedersen, V. Turk, B. Turk,
Human recombinant pro-dipeptidyl peptidase I (cathepsin C) can be activated
by cathepsins L and S but not by autocatalytic processing, Biochemistry 40
(2001) 1671e1678.
[14] M.J. McGuire, P.E. Lipsky, D.L. Thiele, Generation of active myeloid and
lymphoid granule serine proteases requires processing by the granule thiol
protease dipeptidyl peptidase I, J. Biol. Chem. 268 (1993) 2458e2467.
[15] C.T. Pham, T.J. Ley, Dipeptidyl peptidase I is required for the processing and
activation of granzymes A and B in vivo, Proc. Natl. Acad. Sci. U.S.A. 96 (1999)
8627e8632.
[16] K. Sakai, S. Ren, L.B. Schwartz, A novel heparin-dependent processing pathway
for human tryptase. Autocatalysis followed by activation with dipeptidyl
peptidase I, J. Clin. Invest. 97 (1996) 988e995.
[17] B. Noack, H. Gorgens, T. Hoffmann, J. Fanghanel, T. Kocher, P. Eickholz,
H.K. Schackert, Novel mutations in the cathepsin C gene in patients with pre-
pubertal aggressive periodontitis and PapilloneLefevre syndrome, J. Dent. Res.
83 (2004) 368e370.
[18] T.C. Hart, P.S. Hart, M.D. Michalec, Y. Zhang, E. Firatli, T.E. Van Dyke,
A. Stabholz, A. Zlotogorski, L. Shapira, W.A. Soskolne, HaimeMunk syndrome
and PapilloneLefevre syndrome are allelic mutations in cathepsin C, J. Med.
Genet. 37 (2000) 88e94.
[42] R.G. Duggleby, P.V. Attwood, J.C. Wallace, D.B. Keech, Avidin is
a slow-
binding inhibitor of pyruvate carboxylase, Biochemistry 21 (1982) 3364e
3370.
[43] P. Kuzmic, A steady state mathematical model for stepwise “slow-binding”
reversible enzyme inhibition, Anal. Biochem. 380 (2008) 5e12.
[44] B. Lejczak, P. Kafarski, J. Zygmunt, Inhibition of aminopeptidases by amino-
phosphonates, Biochemistry 28 (1989) 3549e3555.
[19] C.T. Pham, J.L. Ivanovich, S.Z. Raptis, B. Zehnbauer, T.J. Ley, PapilloneLefevre
syndrome: correlating the molecular, cellular, and clinical consequences of
cathepsin C/dipeptidyl peptidase I deficiency in humans, J. Immunol. 173
(2004) 7277e7281.
[45] J.F. Morrison, S.R. Stone, Approaches to the study and analysis of the inhibition
of enzymes by slow- and tight-binding inhibitors, Comments Mol. Cell. Bio-
phys. 2 (1985) 347e368.
[20] H. Angliker, P. Wikstrom, H. Kirschke, E. Shaw, The inactivation of the cys-
teinyl exopeptidases cathepsin H and C by affinity-labelling reagents, Bio-
chem. J. 262 (1989) 63e68.
[46] N.D. Rawlings, A.J. Barrett, A. Bateman, MEROPS: the database of proteolytic
enzymes, their substrates and inhibitors, Nucleic Acids Res. 40 (2012) D343e
D350.
[21] C.M. Kam, M.G. Gotz, G. Koot, M. McGuire, D. Thiele, D. Hudig, J.C. Powers,
Design and evaluation of inhibitors for dipeptidyl peptidase I (Cathepsin C),
Arch. Biochem. Biophys. 427 (2004) 123e134.