Design and synthesis of malonic acid-based inhibitors of human neutrophil collagenase (MMP8)

Article abstract of DOI:10.1021/jm9706426

For most of the known synthetic inhibitors of matrix metalloproteinases (MMPs), a substrate-like binding mode was postulated on the basis of X-ray crystallographic structures of MMP/ inhibitor complexes. Conversely, the malonic acid-based inhibitor (2R, S

Full text of DOI:10.1021/jm9706426

J . Med. Chem. 1998, 41, 339-345
339
Design a n d Syn th esis of Ma lon ic Acid -Ba sed In h ibitor s of Hu m a n Neu tr op h il
Colla gen a se (MMP 8)
Erich Graf von Roedern,†,‡ Frank Grams,*,§ Hans Brandstetter,†,| and Luis Moroder*,†
Max-Planck-Institut fu¨r Biochemie, D-82152 Martinsried, Germany, and Boehringer Mannheim,
Department Chemical Research, D-68305 Mannheim, Germany
Received September 25, 1997
For most of the known synthetic inhibitors of matrix metalloproteinases (MMPs), a substrate-
like binding mode was postulated on the basis of X-ray crystallographic structures of MMP/
inhibitor complexes. Conversely, the malonic acid-based inhibitor (2R,S)-HONH-CO-CH(i-
Bu)-CO-Ala-Gly-NH2 was found to bind in a surprisingly different manner. Using this
compound as a new lead structure, the interaction sites with human neutrophil collagenase
(MMP8) were optimized with a series of iteratively designed analogues and with the help of
X-ray structural analysis of selected inhibitors to finally produce low molecular weight
nonpeptidic compounds of 500-1000-fold improved inhibitory potency.
In tr od u ction
substrate-like binding modes of most of the inhibitors
in extended conformations, whereas the malonic acid-
based inhibitor (2R,S)-HONH-Mal(i-Bu)-Ala-Gly-NH2
was found to bind in a nonsubstrate-like mode to human
neutrophil collagenase (MMP8).11 This type of inhibi-
tors containing a malonic acid hydroxamate moiety has
originally been designed for the metalloproteinase ther-
molysin20 and was reported in early stages of the MMP
inhibitor research.21
The binding mode of (2R,S)-HONH-Mal(i-Bu)-Ala-
Gly-NH2 to MMP8 is schematically outlined in Figure
1. The hydroxamate acts as bidentated chelator with
each oxygen in optimal distance (2.2-2.3 Å) from the
active-site Zn2+ ion. The hydrophobic substituent at the
malonic acid moiety in the S configuration is pointing
toward the edge strand of the enzyme active-site cleft.
In contrast to the binding mode of the most potent
inhibitors in clinical trials such as batimastat or ma-
rimastat which contain the succinic acid hydroxamate
moiety,15,22 in this malonic acid-based inhibitors there
is no spacer between the chelating and the hydrophobic
group. Correspondingly, the position of the side chain
of the subsequent amino acid residue becomes similar,
but not identical with that of P2′ residues of substrate-
like binding inhibitors, while the residual peptidic tail
is inserted into the S1′ pocket, thus leading to an overall
â turn-like conformation of the inhibitor in the enzyme-
bound state.
Matrix metalloproteinases (MMPs) are a family of
zinc endopeptidases involved in tissue remodeling and
thus in various disease processes including tumor
invasion and joint destruction. Correspondingly, these
enzymes represent attractive targets for inhibitor design
and drug development.1-5 Considerable insight into
MMP-ligand interactions has been recently obtained
from X-ray crystallography and NMR analysis,6-19 and
many new approaches to optimize MMP inhibitors have
been guided by such structure-based knowledge.
A
central issue in this field is whether broad-spectrum
MMP inhibitors or selective inhibitors targeted against
individual MMP enzymes represent the most appropri-
ate pharmacological strategy. This aspect is even more
decisive if an enzyme or a group of enzymes can
unequivocally be identified as the cause of a particular
disease. Indeed it could be shown that certain MMP
family members are coexpressed in higher amounts in
certain diseases.2
Inspection of the known three-dimensional structures
of MMPs shows that the S1′ pocket offers among the
various subsites the greatest opportunity for a selective
inhibitor design because there is considerable variation
between the MMPs in the residues forming this sub-
strate-binding subsite in the active-site cleft.
A large variety of more or less efficient MMP inhibi-
tors including hydroxamic acids, phosphinic acids, thi-
ols, etc. has been described,1-5 and most of them
contain a peptide-like moiety. The X-ray analyses of
enzyme/inhibitor complexes, reported so far, revealed
The discovery of this unique binding mode led us to
explore the new lead structure in terms of stabilization
of the bended conformation and modifications at the
various interaction sites to improve the inhibitory
potency of such malonic acid hydroxamate-based com-
pounds.
* Addresses for correspondence: Prof. Luis Moroder, Max-Planck-
Institut fu¨r Biochemie, Am Klopferspitz 18A, D-82152 Martinsried,
Germany. Tel: 49-89-8578-3905. Fax: 49-89-8578-2847. E-mail:
moroder@biochem.mpg.de. Dr. Frank Grams, Department Chemical
Research, Boehringer Mannheim GmbH, Sandhofer Strasse 116,
D-68298 Mannheim, Germany. Tel: 49-621-7591220. Fax: 49-621-
7594695.
Resu lts a n d Discu ssion
Ch em istr y. The malonic acid-based inhibitors are
composed of two parts, i.e., the malonic acid hydroxam-
ate moiety and the residue R2 as shown in Figure 2.
The R1-substituted diethyl malonates were prepared by
standard procedures and then hydrolyzed with 1 equiv
of KOH to produce the malonate monoethyl esters which
† Max-Planck-Institut fu¨r Biochemie.
‡ Current address: Hoechst Marion Roussel, HMR Forschung
Chemie, 65929 Frankfurt, Germany.
§ Boehringer Mannheim.
| Current address: Department of Chemistry, MIT, 77 Massachu-
setts Avenue, Cambridge, MA 02139.
S0022-2623(97)00642-0 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/06/1998

340 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Graf von Roedern et al.
F igu r e 1. Schematic presentation of the substrate-like binding mode of batimastat (left)15 and the nonsubstrate-like binding
mode of HONH-Mal(i-Bu)-Ala-Gly-NH2 (8) (right)11 to MMP8.
2
F igu r e 2. General scheme of synthesis used for the preparation of the malonic acid hydroxamates.
were condensed with (benzyloxy)amine following es-
sentially known procedures.23-25 With the use of EDCI/
HOBT as coupling reagent, the guanidine adduct as
observed with the DCC/HOBt method23 was formed to
negligible extents, and because of its water solubility,
it was easily separated together with the resulting urea
from the N-(benzyloxy)malonamic acid ethyl esters.
After an additional saponification step with excesses
NaOH, the resulting benzyl-protected malonic acid
hydroxamates were amidated with the R2-NH2 moieties
acid hydroxamates.
consisting of amino acid derivatives or aryl- and alkyl-
F igu r e 3. Positions A-D of the lead structure HONH-Mal-
(i-Bu)-Ala-Gly-NH2 (8) were the synthetic targets in the
present study for improving inhibition of MMP8 by malonic
Ta ble 1. Inhibition of MMP8 with Peptidic Malonic
Acid-Based Hydroxamates
amines again by the EDCI/HOBt procedure. Finally,
acidolytic removal of side-chain protecting groups, when
required, followed by hydrogenolytic cleavage of the
benzyl group from the hydroxamate over Pd/C catalyst
gave the target hydroxamic acids. Performing the
hydrogenolysis at a low stream of hydrogen and moni-
toring the product formation, reduction of the hydrox-
amic acids to the amide23 was largely suppressed.
Among all compounds only HONH-Mal(i-Bu)-Asp-
NHBn (12b) was obtained as a pure diastereomer by
crystallization of the protected precursor from ether.
There is strong support for the S configuration of the
2-isobutylmalonic acid residue of this diastereomer,
since it is a more potent inhibitor than the diastereo-
meric mixture 12, and in crystals of the complexes
8/MMP811 and 17/MMP8,26 the enzyme-bound malonic
acid moiety is unequivocally in the S-configuration,
whereas in crystals of the 12/MMP8 complex26 an
unambiguous assignment of the configuration was not
possible. For the remaining compounds, the composi-
1
tion of the diastereomeric ratios was determined by H
NMR spectroscopy.
P ep tid ic Ma lon ic Acid -Ba sed In h ibitor s. Using
the binding mode of (2R,S)-HOHN-Mal(i-Bu)-Ala-Gly-
NH2 (8) as lead structure, positions A-D outlined in
Figure 3 were the targets of our attempts to optimize
the inhibition of MMP8 by such malonic acid-based
hydroxamic acids.
Mod ifica tion s in P osition s A a n d B of th e Lea d
Str u ctu r e. Replacement of the Gly-NH2 tail of 8 with
the more hydrophobic benzylamide led to improvement
of the inhibitory potency by a factor of about 3-4 (Table
1). Therefore, in a first series of malonic acid hydrox-
amates benzylamide derivatives were synthesized with
replacements of the Ala residue for exploring the
position B of Figure 3. The side chain of Ala in the
8/MMP8 complex is pointing toward the S2′ subsite of
the enzyme without apparent interactions. A change
of the chirality of the Ala residue (10) was found to
completely reverse the beneficial effects of the hydro-
phobic tail leading to a Ki value identical with that of
just the malonic acid moiety, i.e., of (2R,S)-HOHN-Mal-

Malonic Acid-Based Inhibitors of MMP8
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 341
Ta ble 2. Inhibition of MMP8 with Nonpeptidic Malonic Acid
Hydroxamates
methyl ester function (20), to possibly favor hydrogen
bond or salt bridge interactions with the Arg222 residue
on the bottom of the S1′ pocket, was not achieved as
evidenced by the 10-20-fold reduced inhibitory poten-
cies.
Mod ifica tion s in P osition D. In the 8/MMP8 X-ray
structure the isobutyl group is occupying the S1 subsite
of the enzyme in an hydrophobic surface-type interac-
tion.11 Since omission of the isobutyl group in the Asp
derivative 12 led to a 10-fold increase of the Ki to about
1 mM, we have replaced the isobutyl moiety in com-
pound 18 with the benzyl or 2-phenylethyl group
(compounds 22 and 23); both compounds showed a
reduced inhibition of MMP8 (Table 1). On the other
hand, by substituting the hydrophobic isobutyl group
in 18 with an hydroxyl function (24), practically no effect
on the inhibitory potency was observed.
R1
R2
O-ethyl
O-ethyl
O-ethyl
compd
Ki (µM)
isobutyl
Ph
25
26
27
28
29
30
31
32
33
34
35
36
190
98
53
189
50
3.1
CH2Ph
(CH2)2Ph
CH2Ph
CH2Ph
(CH2)2Ph
isobutyl
CH2Ph
(CH2)2Ph
OH
O-ethyl
N-morpholide
NHCH2Ph
NHCH2Ph
NH(CH2)3Ph
NH(CH2)3Ph
NH(CH2)3Ph
NH(CH2)3Ph
NH-n-octyl
2.3
0.54
0.56
0.49
2.3
isobutyl
0.24
Non p ep tid ic Ma lon ic Acid -Ba sed In h ibitor s. In
the first series of MMP8 inhibitors 8-24 the pseudo P1
residue is linked to the pseudo P1′ residue via a three-
and four-atom (compound 15) spacer formed by an R-
or â-amino acid residue. The results obtained with the
differently C2-substituted malonic acid derivatives 25-
28 (Table 2) were compelling to attempt direct linkage
of the P1 and P1′ groups. Correspondingly, the malonic
acid derivatives 27 and 28 were converted to the
morpholide and benzylamides. While with the mor-
pholide 29 no improvement in the Ki values was
observed, the benzylamides 30 and 31 were found to
exhibit inhibitory potencies comparable to those of the
best peptidic inhibitors with Ki values of 3.1 and 2.3 µM,
respectively (Table 2).
To allow for deeper insertion of the C-terminal
aromatic group into the S1′ subsite, the benzyl group
was replaced by the 3-phenylpropyl group. The related
derivatives of the 2-isobutyl- (32), 2-benzyl- (33), and
2-(2-phenylethyl)- (34) malonic acid hydroxamates
showed very similar Ki values of about 0.5 µM which
were therefore independent of the P1 mimetic group as
long as hydrophobic interactions at the S1 subsite were
retained. By canceling these interactions, e.g., with the
hydroxyl derivative 35, again a 4-5-fold lower affinity
for MMP8 was measured. Finally, mimicking the total
number of backbone atoms of the peptidic inhibitors
with an aliphatic C8 amide substituent (36) a further
decrease of the Ki to 0.24 M was obtained, suggesting a
certain degree of flexibility for the substituents as an
important feature for optimal occupancy of the deep
hydrophobic S1′ subsite of MMP8.
(i-Bu)-OEt (25) (Table 2). Correspondingly, in all
subsequent modifications the L-configuration was re-
tained for this residue, and Ala was replaced by the
amino acids Ser (11), Asp (12), and Asn (13), which
according to modeling experiments were expected to
form hydrogen-bonding networks with Tyr219N and
Glu158O of the active-site cleft of MMP8. Although the
X-ray structure of the 12/MMP8 complex26 revealed that
additional hydrogen bonds were indeed formed as listed
in Table 3, a significant improvement in the inhibitory
potency was not achieved. This can only be explained
if the favored binding energy is attenuated by opposite
effects, e.g., by the strong solvation of these hydrophilic
side chain. However, by incorporating into the Asn
analogue 13 an amino group at the meta position of
benzylamide as potential hydrogen-bonding donor a
2-fold enhanced inhibition of the enzyme by compound
17 in comparison to 13 was obtained. In fact, the X-ray
structure of the 17/MMP complex26 revealed that this
amino group in the crystal is located in hydrogen-
bonding distance to Glu158CO and Leu193CO, as
shown in Table 3 in comparison to other inhibitors.
Con for m a tion a l Restr a in ts in P osition C of th e
Lea d Str u ctu r e. To possibly stabilize the bended
structure of these malonic acid-based inhibitors, the
Ala residue in compound 8 was replaced with the
conformationally restrained residues Pro (14) and the
â-amino acid 2-aminobenzoic acid (Abz) (15), however,
without achieving significant improvements of the
inhibitory potency (Table 1). Nevertheless, the Ki value
of compound 15 clearly indicated that the hydrophobic
character of the S2′ subsite could represent a promising
interaction site. Correspondingly, the Phe analogue 16
was synthesized which exhibited almost the identical
Ki value as that of the Abz analogue 15. Increased
hydrophobic interactions both in the S1′ and S2′ subsites
were attempted with the hPhe analogue 18 where the
C-terminal tail was further extended, i.e., as [2-(4-
methyl)phenyl]ethyl amide derivative, as suggested by
modeling experiments. Indeed a significant improve-
ment was obtained with a Ki of 1.4 µM that suggests
additional hydrophobic interactions of the hPhe side
chain with the S2′ subsite and a good fit of the tail into
the S1′ pocket. Further attempts to improve the binding
affinities of the hPhe analogue 18 by introduction of
hydrophilic moieties in the para position of the 2-phen-
ylethyl group, like the hydroxyl (19), carboxyl (21), or
To conclude, taking advantage of the information
gained with the X-ray structures of a few selected
hydroxamate inhibitors on malonic acid basis, about
500-fold improved inhibition of MMP8 to submicromolar
Ki values could be achieved in comparison to the lead
structure 8 by stepwise optimization of the interaction
sites and finally conversion of the peptidic structures
into low molecular weight nonpeptidic malonic acid
derivatives.
Exp er im en ta l Section
Ma ter ia ls a n d Meth od s. The amino acid derivatives used
in the synthesis were prepared according to standard proce-
dures.27 Solvents and reagents were of the highest quality
commercially available. Dnp-Pro-Leu-Gly-Leu-Trp-Ala-D-Arg-
NH2 was purchased from Bachem (Heidelberg, Germany). TLC

342 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Graf von Roedern et al.
Ta ble 3. Hydrogen Bond Interactions of the MMP8 Active Site with the Inhibitors Batimastat,15 HONH-Mal(i-Bu)-Ala-Gly-NH2 (8),11
HONH-Mal(i-Bu)-Asp-NHBzl (12),26 and HONH-Mal(i-Bu)-Asn-NHBzl-(m-NH2) (17)26
HONH-Mal(i-Bu)-
Ala-Gly-NH2 (8)
HONH-Mal(i-Bu)-
Asp-NHBn (12)
HONH-Mal(i-Bu)-
Asn-NHBn-(m-NH2) (17)
batimastat
OH(NHOH)-Glu198O 2.9/2.4 Å OH(NHOH)-Glu198Oꢀ1/2 3.1/3.2 Å OH(NHOH)-Glu198O 2.9/2.8 Å OH(NHOH)-Glu198O 2.5/3.1 Å
NH(NHOH)-Ala161O 3.0 Å
NH(NHOH)-Ala161O
O(I1)-Leu160N
O(I1)-Ala161N
2.9 Å
2.7 Å
3.5 Å
NH(NHOH)-Ala161O 3.1 Å
NH(NHOH)-Ala161O 2.8 Å
O(I1)-Leu160N
N(I2)-Pro217O
O(I2)-Tyr219N
N1(I2)-Gly158O
2.8 Å
3.3 Å
3.0 Å
3.1 Å
O(I1)-Leu160N
O(I1)-Ala161N
Oδ1(I2)-Tyr219N
Nδ2(I2)-Glu158O
O(I2)-Tyr219N
Oδ2(I2)-Gly158O
3.0 Å
3.4 Å
3.0 Å
O(I1)-Leu160N
O(I1)-Ala161N
Oδ1(I2)-Tyr219N
Nδ2(I2)-Glu158O
O(I2)-Tyr219N
Oδ2(I2)-Gly158O
2.7 Å
3.4 Å
3.0 Å
2.9 Å
3.4 Å
3.1 Å
3.0 Å
silica gel 60 plates were from Merck AG (Darmstadt, Ger-
many), and compounds were visualized with the chlorine/
tolidine or permanganate reagent. FAB-MS spectra were
recorded on a Finnigan MAT 900 and the NMR spectra on
AMX 400 and AMX 500 (Bruker).
described for 2: yield 95%; the analytical data were consistent
with those reported by Fournie-Zaluski et al.24
(2R,S)-Bn ONH-Ma l(CH2CH2P h )-OH (6). (2R,S)-BnONH-
Mal(CH2-CH2-Ph)-OEt (5) was obtained from (2R,S)-EtO-Mal-
(CH2-CH2-Ph)-OH31 as described for 1: yield 55%; TLC (AcOEt/
E n zym e Assa y. The catalytic domain of MMP8 (Phe79
,
n-hexane, 1:1) Rf 0.4; FAB-MS m/z 342.2 [M + H+]; Mr
)
Gly242-MMP8) was used for the inhibition experiments. En-
zyme assays were performed at 25 °C in 10 mM CaCl2, 100
mM NaCl, 50 mM Tris/HCl (pH 7.6) using 8 nM enzyme
concentration and the fluorogenic substrate Dnp-Pro-Leu-Gly-
Leu-Trp-Ala-D-Arg-NH2 (1 × 10-5 M). Enzyme kinetics were
performed essentially as described by Stack et al.,28 and the
increase in fluorescence at 350 nm was monitored over a period
of 100 s with a luminescence spectrometer LS 50B (Perkin-
Elmer) to determine initial rates of hydrolysis. Evaluation of
the kinetic data was performed as reported by Copeland et
al.29 Assuming Michaelis-Menten kinetics, the Ki values were
then calculated with the program Enzfitter (Version 1.05,
1987, Elsevier-Biosoft).
341.19 calcd for C20H23NO4; 1H NMR (DMSO-d6) the spectrum
is consistent with the structure. Compound 5 was saponified
to 6: yield 96%; TLC (CH3CN/H2O, 4:1) Rf 0.65; 1H NMR
(DMSO-d6) δ 13.2 (s (br), 1H, COOH), 11.32 (s, 1H, NH), 7.3-
7.4 (m, 10H, 2 C6H5), 3.09 (t, 1H, H-C2 (Mal)), 2.50 (t, 2H,
H-C4 (Mal)), 2.02 (m, 2H, H-C3 (Mal)).
(2R,S)-Bn ONH-Ma l(OBn )-OH (7). EtO-Mal(OBn)-OEt32
was converted to the title compound as described for 2: yield
58% over the two steps; colorless powder; TLC (CH3CN/H2O,
1
4:1) Rf 0.25; H NMR (DMSO-d6) δ 13.1 (s (br), 1H, COOH),
11.50 (s, 1H, NH), 7.3-7.4 (m, 10H, 2 C6H5), 4.83 (s, 2H, CH2
(BnONH)), 4.53 (2 d, 2H, CH2 (OBn)), 4.34 (s, 1H, H-C2 (Mal)).
(2R,S)-HONH-Ma l(i-Bu )-Ala -Gly-NH2 (8). The compound
was prepared from 2 and H-Ala-Gly-NH2 as described for 9.
The analytical data were consistent with those reported by
by Nishino and Powers;25 diastereomeric ratio was 36:64.
(2R,S)-HONH-Ma l(i-Bu )-Ala -NHBn (9). To H-Ala-NH-
Bn33 (0.18 g, 1.0 mmol) in THF at 0 °C were added compound
2 (0.27 g, 1.0 mmol), HOBt (0.14 g, 1.0 mmol), and EDCI (0.20
g, 1.05 mmol). The reaction mixture was stirred for 12 h,
allowing room temperature to be reached. The solvent was
then evaporated and the residue distributed between AcOEt
and water. The organic phase was washed with 5% KHSO4,
5% NaHCO3, and water, dried over MgSO4, and evaporated
to small volume. The product was precipitated with ether and
then hydrogenated in MeOH over 10% Pd/C catalyst for 1 h.
The catalyst was filtered off, the filtrate was evaporated to
small volume, and the product was precipitated with ether:
yield 187 mg (56%); TLC (CHCl3/MeOH, 9:1) Rf 0.4; FAB-MS
m/z 336.4 [M + H+]; Mr ) 335.41 calcd for C17H25N3O4; 1H
NMR (DMSO-d6) δ 10.57 + 10.51 (2 s, 1H, NHOH), 8.97 (s,
1H, OH), 8.44-8.52 (m, 1H, NHBn), 7.84 + 7.94 (2 d, 1H, NH
(Ala)), 7.20-7.39 (m, 5H, C6H5), 4.30 (m, 3H, HR (Ala) + CH2
(Bn)), 3.00-3.10 (m, 1H, H-C2 (Mal)), 1.37-1.65 (m, 3H, H-C3
(Mal) + H-C4 (Mal)), 1.27 (d, 3H, CH3 (Ala)), 0.85 (2 d, 6H, 2
CH3); diastereomeric ratio, 40:60.
Com p u t a t ion a l Met h od s. Docking calculations were
performed on the crystal structure of MMP811 using the
program package INSIGHT/DISCOVER (version 2.96, MSI
Technologies, San Diego, CA). By keeping the protein struc-
ture fixed and the hydroxamate function as ligand of the
active-site Zn2+ as determined in the 8/MMP8 complex as
constraint, alternative orientations and conformations of
potential new inhibitors were generated and energy-minimized
to convergence. For inspection of the possible binding mode
of newly designed ligands, interaction sites have been gener-
ated with the program LUDI.30
Syn th esis: (2R,S)-Bn ONH-Ma l(i-Bu )-OH (2). To a sus-
pension of (benzyloxy)amine hydrochloride (4.79 g, 30 mmol)
in THF was added sodium methylate (1.62 g; 30 mmol) under
stirring, and after 10 min the solvent was evaporated. The
residue was resuspended in THF and reacted at 0 °C with
(2R,S)-EtO-Mal(i-Bu)-OK23 (6.78 g, 30 mmol), EDCI (6.34 g,
33 mmol), and HOBt (4.6 g, 33 mmol). The reaction mixture
was allowed to reach room temperature, and after 12 h it was
worked up by washing the crude product in AcOEt with 5%
KHSO4, 5% NaHCO3, and water. The organic layer was dried
over MgSO4 and evaporated; yield of (2R,S)-BnONH-Mal(i-
Bu)-OEt (1): 7.45 g (84.6%); colorless oil; TLC (AcOEt/n-
hexane, 1:2) Rf 0.3.
Compound 1 (3.53 g, 12 mmol) was saponified in THF/
MeOH (1:1) with NaOH (1.44 g, 36 mmol) in 2 mL of water at
50 °C for 1 h. The reaction mixture was diluted with MeOH
and treated with 10 g of amberlyst 15 (H+ form; 4.6 mmol/g)
under ice cooling. The resin was filtered off and washed with
MeOH, and the combined filtrates were evaporated to dry-
ness: yield 3.2 g (85%), needles; TLC (CH3CN/H2O, 4:1) Rf 0.6;
FAB-MS m/z 266.1 [M + H+]; Mr ) 265.1 calcd for C14H19NO4;
1H NMR (DMSO-d6) δ 11.26 (s, 1H, NHOBn), 7.34-7.39 (m,
5H, C6H5), 4.78 (s, 2H, CH2 (Bn)), 3.03 (dd, 1H, H-C2 (Mal)),
1.37-1.65 (m, 3H, H-C3 (Mal) + H-C4 (Mal)), 0.85 (2 d, 6H, 2
CH3); the analytical data are consistent with those reported
by Bihovsky et al.23
The following HONH-Mal(i-Bu)-Xxx-NHBn analogues were
prepared by identical procedures via coupling of the malonic
acid derivative 2 with H-Xxx-NHBn using the EDCI/HOBt
procedure followed by deprotection via hydrogenolysis over
10% Pd/C.
(2R,S)-HONH-Ma l(i-Bu )-D-Ala -NHBn (10): yield 50%
over the two steps; TLC (CHCl3/MeOH, 9:1) Rf 0.4; FAB-MS
m/z 336.4 [M + H+]; Mr ) 335.41 calcd for C17H25N3O4; 1H
NMR (DMSO-d6) δ 10.57 + 10.51 (2 s, 1H, NHOH), 8.97 (br,
1H, OH), 8.44-8.52 (m, 1H, NHBn), 7.84 + 7.94 (2 d, 1H, NH
(Ala)), 7.20-7.39 (m, 5H, C6H5), 4.30 (m, 3H, HR (Ala) + CH2
(Bn)), 3.00-3.10 (m, 1H, H-C2 (Mal)), 1.37-1.65 (m, 3H, H-C3
(Mal) + H-C4 (Mal)), 1.27 (d, 3H, CH3 (Ala)), 0.85 (d, 6H, 2
CH3); diastereomeric ratio, 25:75.
(2R,S)-Bn ONH-Ma l(Bn )-OH (4). (2R,S)-EtO-Mal(Bn)-OK
was converted to (2R,S)-BnONH-Mal(Bn)-OEt (3) as described
for compound 1: yield 88%; TLC (AcOEt/n-hexane, 1:1) Rf 0.4;
1H NMR (DMSO-d6) the spectrum is consistent with the
structure. Compound 3 was saponified and worked up as
(2R,S)-HONH-Ma l(i-Bu )-Ser -NHBn (11): yield 78% over
the two steps; TLC (CHCl3/MeOH, 9:1) Rf 0.2; FAB-MS m/z
1
352.2 [M + H+]; Mr ) 351.18 calcd for C17H25N3O5; H NMR
(DMSO-d6) δ 10.50 (br, 1H, NHOH), 9.00 (br, 1H, OH), 8.45-
8.56 (m, 1H, NHBn), 7.80 + 7.96 (2 d, 1H, NH (Ser)), 7.20-

Malonic Acid-Based Inhibitors of MMP8
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 343
7.39 (m, 5H, C6H5), 5.00 (br, 1H, OH (Ser)), 4.30 (m, 3H, HR
(Ser) + CH2 (Bn)), 3.55 + 3.65 (2 ddd, 2H, CH2 (Ser)), 3.00-
3.10 (m, 1H, H-C2 (Mal)), 1.37-1.65 (m, 3H, H-C3 (Mal) +
H-C4 (Mal)), 0.85 (d, 6H, 2 CH3); diastereomeric ratio, 28:72.
(2R,S)-HONH-Ma l(i-Bu )-Asp -NHBn (12). Prior to the
hydrogenation step, exposure to TFA (1 h) was used to
deprotect the Asp(OtBu) residue: yield 70% over the three
steps; TLC (CH3CN/H2O, 4:1) Rf 0.55; FAB-MS m/z 380.2 [M
8.10 + 8.25 (m, 1H, NH (Asn)), 6.85-7.45 (2 s, 6H, NH2 (Asn),
C6H4), 4.62 (m, 1H, HR (Asn)), 4,30 (m, 2H, CH2 (Bn)), 3.10-
3.20 (2 dd, 1H, H-C2 (Mal)), 2.50 (m, 2H, CH2 (Asn)), 1.40-
1.65 (m, 3H, H-C3 (Mal) + H-C4 (Mal)), 0.85 (d, 6H, 2 CH3);
diastereomeric ratio, 44:56.
(2R,S)-H ONH -Ma l(i-Bu )-h P h e-NH -CH 2CH 2-P h -(p -Me)
(18). To Z-hPhe-OH (160 mg, 0.5 mmol) in THF at 0 °C were
added 2-(p-tolyl)ethylamine (68 mg, 0.5 mmol), HOBt (70 mg,
0.5 mmol), and EDCI (100 mg, 0.53 mmol). After 12 h the
reaction mixture was diluted with AcOEt and then washed
with 5% KHSO4, 5% NaHCO3, and water. The organic layer
was dried over MgSO4 and evaporated to dryness: yield 0.2 g
(93%); TLC (AcOEt/n-hexane, 2:1) Rf 0.2. Upon hydrogenation
in MeOH over Pd/C the resulting H-hPhe-NH-CH2CH2-Phe-
(p-Me) was coupled with the malonic acid derivative 2 by the
EDCI/HOBt method and worked up as reported for 9; after
hydrogenation the title compound was precipitated from
MeOH with n-hexane: yield 70% over the two steps; TLC
(CHCl3/MeOH, 9:1) Rf 0.4; FAB-MS m/z 454.2 [M + H+]; Mr )
1
+ H+]; Mr ) 379.18 calcd for C18H25N3O6; H NMR (DMSO-
d6) δ 10.52 (s, 1H, NHOH), 9.00 (br, 1H, OH), 8.35-8.45 (m,
1H, NHBn), 8.05 + 8.13 (2 d, 1H, NH (Asp)), 7.20-7.35 (m,
5H, C6H5), 4.60 (m, 1H, HR (Asp)), 4,30 (m, 2H, CH2 (Bn)),
3.00-3.15 (m, 1H, H-C2 (Mal)), 2.58 + 2.69 (2 dd, 2H, CH2
(Asp)), 1.40-1.65 (m, 3H, H-C3 (Mal) + H-C4 (Mal)), 0.85 (d,
6H, 2 CH3); diastereomeric ratio, 43:57. One pure diastere-
omer (12b) could be isolated by crystallization of the protected
precursor (2R,S)-BnONH-Mal(i-Bu)-Asp(OtBu)-NHBn from
ether.
(2R,S)-HONH-Ma l(i-Bu )-Asn -NHBn (13): yield 59% over
the two steps; TLC (CH3CN/H2O, 4:1) Rf 0.6; FAB-MS m/z
1
453.26 calcd for C26H35N3O4; H NMR (DMSO-d6) δ 10.45 (s,
1
379.2 [M + H+]; Mr ) 378.18 calcd for C18H26N4O5; H NMR
1H, NHOH), 8.92 (s, 1H, OH), 8.05 (m, 1H, NH-CH2CH2-Ph),
7.80 + 7.95 (2 d, 1H, NH (hPhe)), 7.05-7.30 (m, 9H, C6H5,
C6H4), 4.23 (m, 1H, HR (hPhe)), 3.20-3.40 (m, 2H, NH-CH2-
CH2-Ph), 3.05 + 3.15 (2 dd, 1H, H-C2 (Mal)), 2.45 + 2.68 (2
m, 4H, NH-CH2CH2-Ph, Hγ (hPhe)), 2.22 (s, 3H, CH3 (aryl)),
1.75 + 1.85 (2 m, 2H, Hâ (hPhe)), 1.45-1.60 (m, 3H, H-C3
(Mal) + H-C4 (Mal)), 0.85 (m, 6H, 2 CH3); diastereomeric ratio,
23:77.
(DMSO-d6) δ 8.42-8.55 (m, 1H, NHBn), 8.20 (m, 1H, NH
(Asn)), 7.45 + 6.88 (2 s, 2H, NH2 (Asn)), 7.15-7.35 (m, 5H,
C6H5), 4.57 (m, 1H, HR (Asn)), 4,25 (m, 2H, CH2 (Bn)), 3.05-
3.15 (m, 1H, H-C2 (Mal)), 2.50 (m, 2H, CH2 (Asn)), 1.40-1.65
(m, 3H, H-C3 (Mal) + H-C4 (Mal)), 0.85 (d, 6H, 2 CH3);
diastereomeric ratio, 34:66.
(2R,S)-HONH-Ma l(i-Bu )-P r o-NHBn (14): yield 22% over
the two steps; TLC (CHCl3/MeOH, 9:1) Rf 0.3 and 0.4 for the
(2R,S)-HONH-Ma l(i-Bu )-h P h e-NH-CH2CH2-P h -(p-OH)
(19). Z-hPhe-OH (160 mg, 0.5 mmol) was reacted with
tyramine (100 mg, 0.75 mmol), and the reaction mixture was
worked up as described for 18: yield 0.21 g (91%); TLC (CHCl3/
MeOH, 9:1) Rf 0.5. Upon hydrogenation in MeOH over Pd/C
the resulting H-hPhe-NH-(CH2)2-C6H4-(p-OH) was coupled
with 2 by the EDCI/HOBt method and worked up as reported
for compound 9; after hydrogenation the title compound was
precipitated from MeOH with n-hexane: yield 81% over the
two steps; TLC (CHCl3/MeOH, 9:1) Rf 0.25; FAB-MS m/z 456.1
[M + H+]; Mr ) 455.24, calcd for C25H33N3O5; 1H NMR (DMSO-
d6) δ 10.55 (s, 1H, NHOH), 9.13 (s, 1H, OH (aryl)), 8.92 (s,
1H, OH), 8.05 (m, 1H, NH-CH2CH2-Ph), 7.80 + 7.95 (2 d, 1H,
NH (hPhe)), 6.75-7.30 (m, 9H, C6H5 + C6H4), 4.23 (m, 1H,
HR (hPhe)), 3.20-3.40 (m, 2H, NH-CH2CH2-Ph), 3.05 + 3.15
(2 dd, 1H, H-C2 (Mal)), 2.45 + 2.68 (2 m, 4H, NH-CH2CH2-
Ph, Hγ (hPhe)), 1.75 + 1.85 (2 m, 2H, Hâ (hPhe)), 1.45-1.60
(m, 3H, H-C3 (Mal) + H-C4 (Mal)), 0.85 (m, 6H, 2 CH3);
diastereomeric ratio, 24:76.
two diastereoisomers; FAB-MS m/z 362.3 [M + H+]; Mr
)
1
361.18 calcd for C19H27N3O4; H NMR (DMSO-d6) δ 10.60 +
10.65 (2 s, 1H, NHOH), 8.90 + 8.96 (2 s, 1H, OH), 8.30 (m,
1H, NHBn), 7.20-7.35 (m, 5H, C6H5), 4.20-4.35 (m, 3H, HR
(Pro), CH2 (Bn)), 3.50-3.70 (m, 2H, Hδ (Pro)), 3.35 (m, 1H,
H-C2 (Mal)), 1.4-2.2 (m, 7H, Hâ (Pro)), Hγ (Pro)), H-C3 (Mal),
H-C4 (Mal)), 0.90 (m, 6H, 2 CH3); diastereomeric ratio, 50:50.
(2R,S)-HONH-Ma l(i-Bu )-Abz-NHBn (15): yield 40% over
the two steps; FAB-MS m/z 384.1 [M + H+]; Mr ) 385.18 calcd
for C21H25N3O4; 1H NMR (DMSO-d6) δ 11.28 (s, 1H, NH(Abz)),
10.77 (2 s, 1H, NHOH), 8.96 (s, 1H, OH), 9.25 (m, 1H, NHBn),
8.35 (d, 1H, aryl (Abz)), 7.77 (d, 1H, aryl (Abz)), 7.48 (dd, 1H,
aryl (Abz)), 7.15 (dd, 1H, aryl (Abz)), 7.20-7.35 (m, 5H, C6H5),
4.50 (m, 2H, CH2 (Bn)), 3.17 (dd, 1H, H-C2 (Mal)), 1.4-1.75
(m, 3H, H-C3 (Mal), H-C4 (Mal)), 0.87 (d, 6H, 2 CH3); racemic
mixture.
(2R,S)-HONH-Ma l(i-Bu )-P h e-NHBn (16): yield 80% over
the two steps; TLC (CHCl3/MeOH, 9:1) Rf 0.4; FAB-MS m/z
1
412.2 [M + H+]; Mr ) 411.21 calcd for C23H29N3O4; H NMR
(2R ,S )-H O N H -Ma l(i -B u )-h P h e -N H -C H 2 C H 2 -P h -(p -
COOMe) (20). Z-hPhe-OH (160 mg, 0.5 mmol) was reacted
with NH2-CH2CH2-Ph-(p-COOMe) (108 mg, 0.75 mmol) under
the conditions described for 18: yield 170 mg (72%); TLC
(AcOEt/n-hexane, 2:1) Rf 0.35. Upon hydrogenation the
resulting H-hPhe-NH-CH2CH2-Ph(p-COOMe) was coupled
with 2 by the EDCI/HOBt method, and the reaction mixture
was worked up as reported for the Ala analogue 9; after
hydrogenation the title compound was precipitated from
MeOH with n-hexane: yield 63% over the two steps; TLC
(CHCl3/MeOH, 9:1) Rf 0.4; FAB-MS m/z 498.4 [M + H+]; Mr )
(DMSO-d6) δ 10.45 (s, 1H, NHOH), 8.92 (s, 1H, OH), 8.56 (m,
1H, NHBn), 7.85 + 8.03 (2 d, 1H, NH (Phe)), 7.20-7.39 (m,
10H, 2 C6H5), 4.57 (m, 1H, HR (Phe), 4,30 (m, 1H, CH2 (Bn)),
2.75-3.10 (m, 3H, CH2 (Phe), H-C2 (Mal)), 1.20-1.50 (m, 3H,
H-C3 (Mal) + H-C4 (Mal)), 0.75 (m, 6H, 2 CH3); diastereoiso-
meric ratio, 37:63.
(2R,S)-HONH-Ma l(i-Bu )-Asn -NHBn -(m -NH2) (17). 3-Ni-
trobenzylamine was reacted with Boc-Asn-ONp and worked
up by standard procedures: yield 63%; TLC (CHCl3/MeOH,
9:1) Rf 0.35; mp 190-191 °C. Boc-Asn-NHBn-(m-NO2) was
deprotected with 4 N HCl in dioxane. The solution was
concentrated to a small volume, and H-Asn-NHBn-(m-NO2)‚
HCl was precipitated with ether and then coupled with the
malonic acid derivative 2 by the EDCI/HOBt procedure in the
presence of triethylamine as auxiliary base. The reaction
mixture was worked up as described for 9: yield 40%; TLC
1
497.25 calcd for C27H35N3O6; H NMR (DMSO-d6) δ 10.58 (s,
1H, NHOH), 8.98 (s, 1H, OH), 8.15 (m, 1H, NH-CH2CH2-Ph),
7.95 + 8.05 (2 d, 1H, NH (hPhe)), 7.00-7.80 (m, 9H, C6H5
+
C6H4), 4.23 (m, 1H, HR (Phe), 3.80 (s, 3H, COOCH3), 3.20-
3.40 (m, 2H, NH-CH2CH2-Ph), 3.05 + 3.15 (2 dd, 1H, H-C2
(Mal)), 2.45 + 2.80 (2 m, 4H, NH-CH2CH2-Ph, Hγ (hPhe)), 1.75
+ 1.85 (2 m, 2H, Hâ (hPhe)), 1.45-1.60 (m, 3H, H-C3 (Mal) +
H-C4 (Mal)), 0.85 (m, 6H, 2 CH3); diastereomeric ratio, 22:78.
1
(CHCl3/MeOH, 9:1) Rf 0.3; H NMR (DMSO-d6) the spectrum
is consistent with the assigned structure.
(2R,S)-BnONH-Mal(i-Bu)-Asn-NHBn-(m-NO2) was hydro-
genated in MeOH in the presence of equivalent amounts of 1
N HCl over 10% Pd/C catalyst for 10 h. The catalyst was
filtered off, and the filtrate was concentrated to a small
volume. The product was precipitated with ether: yield 92%;
TLC (CHCl3/MeOH, 4:1) Rf 0.15; FAB-MS m/z 394.2 [M + H+];
Mr ) 393.20 calcd for C18H27N5O5; 1H NMR (DMSO-d6) δ 10.75
(s, 1H, NHOH), 10.0 (br, 3H, NH3 (aryl)), 8.45 (m, 1H, NHBn),
(2R ,S )-H O N H -Ma l(i -B u )-h P h e -N H -C H 2 C H 2 -P h -(p -
COOH) (21). Compound 20 (50 mg, 0.1 mmol) was saponified
with 1 N NaOH (1 mL) in THF and then diluted with MeOH.
The reaction mixture was treated with amberlyst 15 (H+-form)
and evaporated to a small volume, and the product was
precipitated with ether: yield 30 mg (60%); TLC (CHCl3/
MeOH, 4:1) Rf 0.4; FAB-MS m/z 484.2 [M + H+]; Mr ) 483.24
1
calcd for C26H33N3O6; H NMR (DMSO-d6) δ 10.56 + 10.53 (2

344 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Graf von Roedern et al.
s, 1H, NHOH), 8.98 (br, 1H, OH), 8.15 (m, 1H, NH-CH2CH2-
Ph), 7.95 (d, 1H, NH (hPhe)), 7.00-7.90 (m, 9H, C6H5 + C6H4),
4.23 (m, 1H, HR (hPhe), 3.20-3.40 (m, 2H, NH-CH2CH2-Ph),
3.05 + 3.15 (2 dd, 1H, H-C2 (Mal)), 2.45 + 2.80 (2 m, 4H, NH-
CH2CH2-Ph, Hγ (hPhe)), 1.75 + 1.85 (2 m, 2H, Hâ (hPhe)),
1.45-1.60 (m, 3H, H-C3 (Mal) + H-C4 (Mal)), 0.85 (m, 6H, 2
CH3); diastereomeric ratio, 50:50.
(dd, 1H, H-C2 (Mal)), 2.00 + 2.50 (2 m, 4H, H-C3 (Mal)), H-C4
(Mal), 1.15 (t, 3H, CH2CH3).
(2R,S)-HONH-Mal(Bn )-m or ph olide (29). (2R,S)-BnONH-
Mal(Bn)-OH (4) was coupled with morpholine by the EDCI/
HOBt method, and the crude product was dissolved in AcOEt
and washed with 5% KHSO4, 5% NaHCO3, and water. The
organic phase was dried over MgSO4 and concentrated to small
volume. The product was precipitated with pentane and then
hydrogenated in MeOH over 10% Pd/C catalyst. The catalyst
was filtered off and the filtrate evaporated to dryness: yield
81% over the two steps; TLC (CHCl3/MeOH, 9:1) Rf 0.4; FAB-
MS m/z 279.2 [M + H+]; Mr ) 278.13 calcd for C14H18N2O4; 1H
NMR (DMSO-d6) δ 10.47 (s, 1H, NHOH), 8.85 (s, 1H, NHOH),
7.15-7.30 (m, 5H, C6H5), 3.73 (t, 1H, H-C2 (Mal)), 2.90-3.50
(m, 10H, H-C3 (Mal), CH2(morpholide)).
(2R,S)-HONH-Mal(Bn )-h P h e-NH-CH2CH2-P h -(p-Me) (22).
The malonic acid derivative 4 (90 mg, 0.3 mmol) was condensed
with H-hPhe-NH-CH2CH2-Ph-(p-Me) (80 mg, 0.2 mmol), and
the product was isolated as described for 18: TLC (CHCl3/
MeOH, 9:1) Rf 0.75. Upon hydrogenation over Pd/C in MeOH
the product was precipitated from MeOH with n-hexane: yield
41 mg (52%) over the two steps; TLC (CHCl3/MeOH, 9:1) Rf
0.45; FAB-MS m/z 488.2 [M + H+]; Mr ) 487.24 calcd for
C
29H33N3O4; 1H NMR (DMSO-d6) δ 10.54 + 10.51(2 s, 1H,
(2R,S)-HONH-Ma l(Bn )-NHBn (30). Compound 4 was
converted to the benzyl amide and then deprotected by
hydrogenolysis as described for 29: yield 76% over the two
steps; TLC (CHCl3/MeOH, 9:1) Rf 0.3; FAB-MS m/z 299.2 [M
NHOH), 8.98 (br, 1H, OH), 7.90-8.05 (m, 2H, NH-CH2CH2-
Ph, NH (hPhe)), 6.95-7.30 (m, 14H, 2 C6H5 + C6H4), 4.15 +
4.23 (2 m, 1H, HR (hPhe)), 1.6-3.40 (m, 10H, 5 CH2), 2.22 (s,
3H, CH3 (aryl)); diastereomeric ratio, 43:57.
1
+ H+]; Mr ) 298.13 calcd for C17H18N2O3; H NMR (DMSO-
(2R,S)-HONH-Ma l(CH2CH2-P h )-h P h e-NH-CH2CH2-P h -
(p-Me) (23). The malonic acid derivative 6 (95 mg, 0.3 mmol)
was reacted with H-hPhe-NH-CH2CH2-Ph-(p-Me) (130 mg, 0.3
mmol) as described for 18. Upon hydrogenation, the product
was precipitated from MeOH with n-hexane: yield 81 mg
(72%) over the two steps; TLC (CHCl3/MeOH, 9:1) Rf 0.4; FAB-
MS m/z 502.1 [M + H+]; Mr ) 501.24 calcd for C30H35N3O4; 1H
NMR (DMSO-d6) δ 10.54 + 10.51 (2 s, 1H, NHOH), 8.98 (s,
1H, OH), 7.90-8.05 (m, 2H, NH-CH2CH2-Ph, NH (hPhe)),
7.00-7.35 (m, 14H, 2 C6H5 + C6H4), 4.23 (m, 1H, HR (hPhe)),
1.6-3.40 (m, 12H, 6 CH2), 2.22 (s, 3H, CH3 (aryl)); diastere-
omeric ratio, 25:75.
(2R,S)-H ONH -Ma l(OH )-h P h e-NH -CH 2CH 2-P h -(p -Me)
(24). The malonic acid derivative 7 (105 mg, 0.33 mmol) was
reacted with H-hPhe-NH-CH2CH2-Ph-(p-Me) (172 mg, 0.4
mmol) under the conditions described for 18. After hydro-
genolysis of both O-benzyl groups over Pd/C, the product was
precipitated from MeOH with n-hexane: yield 110 mg (53%)
over the two steps; TLC (CHCl3/MeOH, 9:1) Rf 0.2; FAB-MS
m/z 414.0 [M + H+]; Mr ) 413.19 calcd for C22H27N3O5; 1H
NMR (DMSO-d6) δ 10.70 (s, 1H, NHOH), 8.12 + 8.05 (2 dd,
1H, NH-CH2CH2-Ph), 7.95 (d, 1H, NH (hPhe)), 7.00-7.35 (m,
9H, C6H5 + C6H4), 4.38 (s, 1H, H-C2 (Mal)), 4.28 (m, 1H, HR
(hPhe), 1.6-3.40 (m, 8H, 4 CH2), 2.22 (s, 3H, CH3 (aryl));
diastereomeric ratio, 45:55.
d6) δ 10.41 (s, 1H, NHOH), 8.89 (s, 1H, NHOH), 8.17 (dd, 1H,
NHBn)), 7.10-7.30 (m, 10H, 2 C6H5), 4.30 (dd, 1H, CHH (NH-
Bn)), 4.20 (dd, 1H, CHH (NH-Bn)), 3.33 (t, 1H, H-C2 (Mal)),
3.05 (m, 2H, H-C3 (Mal)).
(2R,S)-HONH-Ma l(CH2CH2-P h )-NHBn (31). Compound
6 was converted to the benzylamide and deprotected by
hydrogenolysis as described for 29: yield 75% over the two
steps; TLC (CHCl3/MeOH, 9:1) Rf 0.4; mp 159 °C; FAB-MS
m/z 313.2 [M + H+]; Mr ) 312.14 calcd for C18H20N2O3; 1H
NMR (DMSO-d6) δ 10.47 (s, 1H, NHOH), 8.93 (s, 1H, NHOH),
8.14 (dd, 1H, NHBn)), 7.13-7.33 (m, 10H, 2 C6H5), 4.30 (dd,
1H, CHH (NHBn)), 4.28 (dd, 1H, CHH (NHBn)), 2.99 (t, 1H,
H-C2 (Mal)), 2.49 (t, 2H, H-C4 (Mal)), 2.03 (q, 2H, H-C3 (Mal)).
(2R,S)-H ONH -Ma l(i-Bu )-NH -(CH 2)3-P h (32). (2R,S)-
BnONH-Mal(i-Bu)-OH (2) was converted to the 3-phenylpro-
pylamide and deprotected by hydrogenolysis as described for
29: yield 62% over the two steps; mp 148 °C; FAB-MS m/z
1
293.2 [M + H+]; Mr ) 292.17 calcd for C16H24N2O3; H NMR
(DMSO-d6) δ 10.46 (s, 1H, NHOH), 8.86 (s, 1H, NHOH), 7.69
(dd, 1H, NH-(CH2)3-Ph), 7.15-7.30 (m, 5H, C6H5), 3.09 (dd,
2H, NH-CH2-CH2-CH2-Ph), 3.01 (t, 1H, H-C2 (Mal)), 2.55 (dd,
2H, NH-CH2-CH2-CH2-Ph), 1.69 (q, 2H, NH-CH2-CH2-CH2-Ph),
1.58 (m, 2H, H-C3 (Mal)), 1.43 (n, 1H, H-C4 (Mal)), 0.86 +
0.84 (2 d, 6H, 2 CH3).
(2R,S)-H ONH -Ma l(Bn )-NH -(CH 2)3-P h (33). (2R,S)-
BnONH-Mal(Bn)-OH (4) was converted to the 3-phenylpropy-
lamide and deprotected by hydrogenolysis as described 29:
yield 60% over the two steps; mp 143 °C; FAB-MS m/z 327.3
[M + H+]; Mr ) 326.16 calcd for C19H22N2O3; 1H NMR (DMSO-
d6) δ 10.41 (s, 1H, NHOH), 8.87 (s, 1H, NHOH), 7.73 (dd, 1H,
NH-(CH2)3-Bn)), 7.10-7.30 (m, 10H, 2 C6H5), 3.05 (t, 1H, H-C2
(Mal)), 3.04 (m, 4H, NH-CH2-CH2-CH2-Ph, H-C3 (Mal)), 2.50
(dd, 2H, NH-CH2-CH2-CH2-Ph), 1.71 (q, 2H, NH-CH2-CH2-CH2-
Ph).
(2R,S)-HONH-Ma l(i-Bu )-OEt (25). Compound 1 (0.59 g,
2 mmol) was hydrogenated in MeOH over 10% Pd/C: yield
390 mg (95%) of colorless crystals; TLC (CHCl3/MeOH, 9:1)
Rf 0.4; FAB-MS m/z 204.1 [M + H+]; Mr ) 203.15 calcd for
1
C9H17NO4; H NMR (DMSO-d6) δ 10.70 (s, 1H, NHOH), 8.95
(s, 1H, OH), 4.05 (m, 2H, CH2CH3), 3.13 (dd, 1H, H-C2 (Mal)),
1.4-1.7 (m, 3H, H-C3 (Mal), H-C4 (Mal)), 1.15 (t, 3H,
CH2CH3(Et)), 0.83 (d, 6H, 2 CH3 (Mal)).
(2R,S)-HONH-Ma l(P h )-OEt (26). (2R,S)-BnONH-Mal-
(Ph)-OEt was prepared as described for 1 from (2R,S)-EtO-
Mal(Ph)-OK34 and then hydrogenated in MeOH over Pd/C:
yield 82%; TLC (AcOEt/n-hexane, 1:1) Rf 0.15; FAB-MS m/z
224.2 [M + H+]; Mr ) 223.1 calcd for C11H13NO4; 1H NMR
(DMSO-d6) δ 10.70 (s, 1H, NHOH), 9.00 (s, 1H, OH), 7.15-
7.32 (m, 5H, C6H5), 4.05-4.15 (m, 2H, CH2CH3), 3.10 (dd, 1H,
H-C2 (Mal)), 1.18 (t, 3H, CH2CH3).
(2R,S)-HONH-Mal(CH2CH2-P h )-NH-(CH2)3-P h (34). Com-
pound 6 was condensed with 3-phenylpropylamine and the
resulting derivative was deprotected by hydrogenolysis as
described for 29: yield 70% over the two steps; mp 134 °C;
FAB-MS m/z 341.2 [M + H+]; Mr ) 340.17 calcd for C20H24N2O3;
1H NMR (DMSO-d6) δ 10.46 (s, 1H, NHOH), 8.91 (s, 1H,
NHOH), 7.72 (dd, 1H, NH-(CH2)3-Ph), 7.10-7.30 (m, 10H, 2
C6H5), 3.09 (dd, 2H, NH-CH2-CH2-CH2-Ph), 2.95 (t, 1H, H-C2
(Mal)), 2.58 (t, 1H, H-C4 (Mal)), 2.55 (t, 1H, H-C4 (Mal)), 2.49
(dd, 2H, NH-CH2-CH2-CH2-Ph),1.99 (q, 2H, H-C3 (Mal)), 1.71
(q, 2H, NH-CH2-CH2-CH2-Ph).
(2R,S)-HONH-Ma l(Bn )-OEt (27). Compound 3 (165 mg,
0.5 mmol) was hydrogenated in MeOH over 10% Pd/C: yield
86 mg, 72%; TLC (AcOEt/n-hexane, 1:1) Rf 0.15; FAB-MS m/z
1
238.1 [M + H+]; Mr ) 237.25 calcd for C12H15NO4; H NMR
(DMSO-d6) δ 10.63 (s, 1H, NHOH), 8.95 (s, 1H, OH), 7.15-
7.30 (m, 5H, C6H5), 4.05 (m, 2H, CH2CH3), 3.13 (dd, 1H, H-C2
(Mal)), 3.08 + 2.97 (2 dd, 2H, C3 (Mal)), 1.15 (t, 3H, CH2CH3).
(2R,S)-HONH-Ma l(CH2CH2-P h e)-OEt (28). Compound 5
(170 mg, 0.5 mmol) was hydrogenated in MeOH over 10% Pd/
C: yield 100 mg (79%); TLC (AcOEt/n-hexane, 1:1) Rf 0.10;
(2R,S)-HONH-Ma l(OH)-NH-(CH2)3-P h (35). Compound
7 was amidated with 3-phenylpropylamine, and the amide
derivative was deprotected by hydrogenolysis as described for
29: yield 81% over the two steps; mp 82-86 °C; FAB-MS m/z
1
253.1 [M + H+]; Mr ) 252.11 calcd for C12H16N2O4; H NMR
FAB-MS m/z 252.1 [M + H+]; Mr ) 251.15 calcd for C13H17
-
(DMSO-d6) δ 10.59 (s, 1H, NHOH), 8.91 (s, 1H, NHOH), 7.86
(dd, 1H, NH-(CH2)3-Ph), 7.10-7.30 (m, 5H, C6H5), 5.90 (d, 1H,
OH), 4.25 (d, 1H, H-C2 (Mal)), 3.11 (q, 2H, NH-CH2-CH2-CH2-
NO4; 1H NMR (DMSO-d6) δ 10.70 (s, 1H, NHOH), 8.95 (s, 1H,
NHOH), 7.15-7.30 (m, 5H, C6H5), 4.10 (m, 2H, CH2CH3), 3.11

Malonic Acid-Based Inhibitors of MMP8
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 345
(14) Lovejoy, B.; Hassell, A. M.; Luther, A. M.; Weigl, D.; J ordan, S.
R. Crystal-Structures of Recombinant 19-kDa Human Fibroblast
Collagenase Complexed to Itself. Biochemistry 1994, 33, 8207-
8217.
Ph), 2.57 (dd, 2H, NH-CH2-CH2-CH2-Ph), 1.73 (quint, 2H, NH-
CH2-CH2-CH2-Ph).
(2R,S)-HONH-Mal(i-Bu )-NH-(CH2)7-CH3 (36). Compound
2 was condensed with n-octylamine, and the resulting amide
was deprotected by hydrogenolysis as described for 29: yield
68.5%; mp 151 °C; FAB-MS m/z 287.3 [M + H+]; Mr ) 286.22
(15) Grams, F.; Crimmin, M.; Hinnes, P.; Huxley, P.; Pieper, M.;
Tschesche, H.; Bode, W. Structure Determination and Analysis
of Human Neutrophil Collagenase Complexed with a Hydrox-
amate Inhibitor. Biochemistry 1995, 34, 14012-14020.
(16) Van Doren, S. R.; Kurochkin, A. V.; Hu, W. D.; Ye, Q. Z.; J ohnson,
L. L.; Hupe, D. J .; Zuiderweg, E. R. P. Solution Structure of the
Catalytic Domain of Human Stromelysin Complexed with a
Hydrophobic Inhibitor. Protein Sci. 1995, 4, 2487-2498.
(17) Browner, M. F.; Smith, W. W.; Castelhano, A. L. Matrilysin-
Inhibitor Complexes: Common Themes among Metalloproteases.
Biochemistry 1995, 34, 6602-6610.
1
calcd for C15H30N2O3; H NMR (DMSO-d6, 400 MHz) δ 10.42
(s, 1H, NHOH), 8.86 (s, 1H, NHOH), 7.55 (dd, 1H, NH-C8H17),
3.03 (m, 2H, NH-CH2-C7H15), 2.95 (t, 1H, H-C2 (Mal)), 1.20-
1.70 (m, 15H, 2 H-C3 (Mal), H-C4 (Mal), NH-CH2-C6H12-Me)),
0.86 + 0.84 (m, 9H, NH-CH2-C6H12-CH3 + 2 CH3).
Abbr evia tion s: MMP, matrix metalloproteinase; MMP8,
human neutrophil collagenase; hPhe, L-(2-phenyl)ethylglycine
(homo-phenylalanine); EDCI, 1-ethyl-3-(3-(dimethylamino)-
propyl)carbodiimide hydrochloride; HOBT, 1-hydroxybenzo-
triazole; Mal(R), 2-substituted malonic acid, R ) isobutyl (i-
Bu), phenyl (Ph), benzyl (Bn).
(18) Dhanaraj, V.; Ye, Q.-Z.; J ohnson, L. L.; Hupe, D. J .; Ortwine,
D. F.; Dunbar, J . B.; Rubin, J . R.; Pavlovsky, A.; Humblet, C.;
Blundell, T. L. X-Ray Structure of a Hydroxamate Inhibitor
Complex of Stromelysin Catalytic Domain and its Comparison
with Members of the Zinc Metalloproteinase Superfamily.
Nature Struct. Biol. 1996, 4, 375-386.
(19) Becker, J . W.; Marcy, A. I.; Rokosz, L. L.; Axel, M. G.; Burbaum,
J . J .; Fitzgerald, P. M. D.; Cameron, P. M.; Esser, C. K.;
Hagmann, W. K.; Hermes, J . D.; Springer, J . P. Stromelysin-
1-3-Dimensional Structure of the Inhibited Catalytic Domain
and of the C-Truncated Proenzyme. Protein Sci. 1995, 4, 1966-
1976.
(20) Nishino, N.; Powers, J . C. Peptide Hydroxamic Acids as Inhibi-
tors of Thermolysin. Biochemistry 1978, 17, 2846-2850.
(21) J ohnson, W. H.; Roberts, N. A.; Borkakoti, N. Collagenase Inhib-
itors: Their Design and Potential Therapeutic Use. J . Enzymol.
Inhib. 1987, 2, 1-22.
(22) Betz, M.; Huxley, P.; Davies, S. J .; Mushtaq, Y.; Pieper, M.;
Tschesche, H.; Bode, W.; Gomis-Ru¨th, F. X. 1.8-Å Crystal
Structure of the Catalytic Domain of Human Neutrophil Colla-
genase (Matrix-Metalloproteinase-8) Complexed with a Pepti-
domimetic Hydroxamate Primed-Side Inhibitor with a Distinct
Selectivity Profile. Eur. J . Biochem. 1997, 247, 356-363.
(23) Bihovsky, R.; Levinson, B. L.; Loewi, R. C.; Erhardt, P. W.;
Polokoff, M. A. Hydroxamic Acids as Potent Inhibitors of
Endothelin-Converting Enzyme from Human Bronchiolar Smooth
Muscle. J . Med. Chem. 1995, 38, 2119-2129.
(24) Fournie-Zaluski, M.-C.; Coulaud, A.; Bouboutou, R.; Chaillet, P.;
Devin, J .; Waksman, G.; Costentin, J .; Roques, B. P. New
Bidantates as Full Inhibitors of Enkephalin-Degrading En-
zymes: Synthesis and Analgesic Properties. J . Med. Chem. 1985,
28, 1158-1168.
(25) Nishino, N.; Powers, J . C. Design of Potent Reversible Inhibitors
for Thermolysin. Peptides Containing Zinc Coordinating Ligands
and their Use in Affinity Chromatography. Biochemistry 1979,
18, 4340-4347.
(26) Brandstetter, H.; Engh, R. A.; Graf von Roedern, E.; Moroder,
L.; Huber, R.; Bode, W.; Grams, F. Structure of Malonic-Acid-
Based Inhibitors Bound to Human Neutrophil Collagenase.
Protein Sci., submitted.
(27) Wu¨nsch, E. Houben-Weyl, Methoden der Organischen Chemie,
Synthese von Peptiden; Springer-Verlag: Stuttgart, 1974; Vols.
15/I and 15/II.
(28) Stack, M. S.; Gray, R. D. Comparison of Vertebrate Collagenase
and Gelatinase Using a New Fluorogenic Substrate Peptide. J .
Biol. Chem. 1989, 264, 4277-4281.
(29) Copeland, R. A.; Lombardo, D.; Giannaras, J .; Decicco, C. P.
Estimating Ki Values for Tight Binding Inhibitors from Dose
Response Plots. Bioorg. Med. Chem. Lett. 1995, 5, 1947-1952.
(30) Bo¨hm, H. J . LUDI: Rule-Based Automatic Design of New
Substituents for Enzyme Inhibitor Leads. J . Comput.-Aided Mol.
Des. 1992, 6, 593-606.
(31) Boettcher, M.; Henklein, P.; Hoffmann, K.; Heder, G.; Siems,
W. E.; Niedrich, H. Synthese von N-substituierten Aminosa¨ure-
p-nitrobenzylestern als potentielle ACE-Inhibitoren. (Synthesis
of N-substituted amino acid p-nitrobenzyl esters as potential
ACE-inhibitors.) Pharmazie 1986, 41, 140.
(32) Hammond, K. M.; Fisher, N.; Morgan, E. N.; Tanner, E. M.;
Franklin, C. S. 3: 5-Dioxo-1: 2-diphenylpyrazolidines. The
4-Hydroxy- and Certain 4-Alkoxy- and 4-Alkylamino-Analogues.
J . Chem. Soc. 1957, 1062-1067.
Ack n ow led gm en t. The authors thank Prof. H.
Tschesche (University of Bielefeld, Germany) for provid-
ing the catalytic domain of MMP8, Ms. I. Buhrow for
the mass spectra, and Dr. I. Sonnenbichler for the NMR
spectra. The study was partly supported by Boehringer
Mannheim AG with a postdoctoral fellowship for E.G.V.R.
Refer en ces
(1) Schwartz, M. A.; Van Wart, H. E. Synthetic Inhibitors of
Bacterial and Mammalian Interstitial Collagenases. Prog. Med.
Chem. 1992, 29, 271-334.
(2) Beckett, R. P.; Davidson, A. H.; Drummond, A. H.; Huxley, P.;
Whittaker, M. Recent Advances in Matrix Metalloproteinase
Inhibitor Research. Drug Discuss. Today 1996, 1, 16-26.
(3) Hagmann, W. K.; Lark, M. W.; Becker, J . W. Inhibition of Matrix
Metalloproteinases. Annu. Rep. Med. Chem. 1996, 31, 231-240.
(4) Murphy, G. Matrix Metalloproteinases and their Inhibitors. Acta
Orthop. Scand., (Suppl. 266), 1995, 66, 55-60.
(5) Beckett, R. P. Recent Advances in the Field of Matrix Metallo-
proteinase Inhibitors. Exp. Opin. Ther. Pat. 1996, 6, 1305-1315.
(6) Lovejoy, B.; Cleasby, A.; Hassell, A. M.; Longley, K.; Luther, M.
A.; Weigl, D.; McGeehan, G.; McElroy, A. B.; Drewry, D.;
Lambert, M. H.; J ordan, S. R. Structure of the Catalytic Domain
of Fibroblast Collagenase Complexed with an Inhibitor. Science
1994, 263, 375-377.
(7) Borkakoti, N.; Winkler, F. K.; Williams, D. H.; D’Arcy, A.;
Broadhurst, M. J .; Brown, P. A.; J ohnson, W. H.; Murray, E. J .
Structure of the Catalytic Domain of Human Fibroblast Colla-
genase Complexed with an Inhibitor. Nature Struct. Biol. 1994,
1, 106-110.
(8) Stams, T.; Spurlino, J . C.; Smith, D. L.; Wahl, R. C.; Ho, T. F.;
Qoronfleh, M. W.; Banks, T. M.; Rubin, B. Structure of Human
Neutrophil Collagenase Reveals Large S1′ Specificity Pocket.
Nature Struct. Biol. 1994, 1, 119-123.
(9) Bode, W.; Reinemer, P.; Huber, R.; Kleine, T.; Schnierer, S.;
Tschesche, H. The X-Ray Crystal Structure of the Catalytic
Domain of Human Neutrophil Collagenase Inhibited by
a
Substrate Analogue Reveals the Essentials for Catalysis and
Specificity. EMBO J . 1994, 13, 1263-1269.
(10) Spurlino, J . C.; Smallwood, A. M.; Carlton, D. D.; Banks, T. M.;
Vavra, K. J .; J ohnson, J . S.; Cook, E. R.; Falvo, J .; Wahl, R. C.;
Pulvino, T. A.; Wendoloski, J . J .; Smith, D. L. 1.56 Å Structure
of Mature Truncated Human Fibroblast Collagenase. Proteins
Struct. Funct. Genet. 1994, 19, 98-109.
(11) Grams, F.; Reinemer, P.; Powers, F. C.; Kleine, T.; Pieper, M.;
Tschesche, H.; Huber, R.; Bode, W. X-ray Structures of Human
Neutrophil Collagenase Complexed with Peptide Hydroxamate
and Peptide Thiol Inhibitors. Eur. J . Biochem. 1995, 228, 830-
841.
(12) Reinemer, P.; Grams, F.; Huber, R.; Kleine, T.; Schnierer, S.;
Pieper, M.; Tschesche, H.; Bode, W. Structural Implications for
the Role of the N-Terminus in the “Superactivation” of Colla-
genases. FEBS Lett. 1994, 338, 227-233.
(13) Gooley, P. R.; O’Connell, J . F.; Marcy, A. I.; Cuca, G. C.; Salowe,
S. P.; Bush, B. L.; Hermes, J . D.; Esser, C. K.; Hagmann, W. K.;
Springer, J . P.; J ohnson, B. A. The NMR Stucture of the
Inhibited Catalytic Domain of Human Stromelysin-1. Nature
Struct. Biol. 1994, 1, 111-118.
(33) Cignarella, G. Synthesis and Configuration of trans-1-Amino-
4-benzyl-2,6-dimethylpiperazine as an Intermediate of Semi-
synthetic Rifamycins. J . Heterocycl. Chem. 1974, 11, 985-988.
(34) Corey, E. J . The Mechanism of the Decarboxylation of Râ- and
âγ-Unsaturated Malonic Acid Derivatives. J . Am. Chem. Soc.
1952, 74, 5897-5905.
J M9706426