Design and synthesis of phosphinamide-based hydroxamic acids as inhibitors of matrix metalloproteinases

Article abstract of DOI:10.1021/jm980142s

A new series of hydroxamic acid-based matrix metalloproteinase (MMP) inhibitors containing a unique phosphinamide motif derived from D-amino acid was designed, synthesized, and tested for enzyme inhibition. Compounds with an R configuration at phosphorus

Full text of DOI:10.1021/jm980142s

J . Med. Chem. 1999, 42, 87-94
87
Design a n d Syn th esis of P h osp h in a m id e-Ba sed Hyd r oxa m ic Acid s a s In h ibitor s
of Ma tr ix Meta llop r otein a ses
Stanislaw Pikul,*,† Kelly L. McDow Dunham,† Neil G. Almstead,† Biswanath De,† Michael G. Natchus,†
Melanie V. Anastasio,‡,§ Sara J . McPhail,‡ Catherine E. Snider,‡ Yetunde O. Taiwo,† Longyin Chen,†,|
C. Michelle Dunaway,‡,§ Fei Gu,† and Glen E. Mieling†
Procter and Gamble Pharmaceuticals, Health Care Research Center, 8700 Mason-Montgomery Road, Mason, Ohio 45040,
Procter and Gamble, Corporate Research Division, Miami Valley Laboratories, Cincinnati, Ohio 45253
Received March 6, 1998
A new series of hydroxamic acid-based matrix metalloproteinase (MMP) inhibitors containing
a unique phosphinamide motif derived from D-amino acid was designed, synthesized, and tested
for enzyme inhibition. Compounds with an R configuration at phosphorus were found to be
potent MMP inhibitors while molecules with the S configuration were almost inactive.
Structure-activity relationship studies of the series led to the discovery of the potent inhibitor
16 with IC50 ) 20.5 nM and 24.4 nM against fibroblast collagenase (MMP-1) and stromelysin
(MMP-3), respectively. The binding mode of this novel phosphinamide-based series of MMP
inhibitors was established based on X-ray crystallography of the complex of stromelysin and
16.
In tr od u ction
The matrix metalloproteinases (MMPs) are a family
of zinc-containing enzymes that are capable of degrading
many proteinaceous components of the extracellular
matrix.1 Important members of this family are collage-
nase (MMP-1), which cleaves all three R-chains of native
interstitial collagens; 72-kDa gelatinase (MMP-2) and
92-kDa gelatinase (MMP-9), which cleave denatured
collagen (gelatin) and basement membrane; and stromel-
ysin (MMP-3), which degrades a wider variety of protein
substrates including gelatin, fibronectin, and laminin
as well as the core protein of cartilage proteoglycans.
Matrix metalloproteinases have been implicated in
several pathological processes including arthritis,2,3
F igu r e 1. Design of phosphinamide-based MMP inhibitors.
tumor growth, metastasis,4 periodontal disease,5 and
multiple sclerosis;6 thus, the discovery of potent and
inhibition of stromelysin. Introduction of the amide
selective inhibitors of the enzymes is a highly attractive
group in place of the sulfonamide linkage, as in molecule
endeavor both scientifically and commercially. The
A, eliminated all in vitro activity. On the basis of this
observation, the authors suggested11 that the oxygen
growing interest in this field is reflected by an increas-
ing number of inhibitors that are being discovered in
various industrial and academic laboratories.7 Data
from phase II pancreatic cancer trial of marimastat, the
most clinically advanced MMP inhibitor, have also been
discussed.8,9
atoms of the sulfonamide group might develop two
distinct hydrogen bonds leading to strong binding with
the enzyme. The authors also contemplated that the
increased conformational freedom of the sulfonamide
bond could account for the observed increase in potency.
We thought that the tetrahedral geometry at the sulfur
atom, as opposed to planarity of the amide linkage,
could also be a factor playing an important role in
determining potency. Searching for experimental evi-
dence, we have designed analogues of CGS27023A and
molecule A that mimic the geometry and conformational
freedom of the sulfonamide bond yet provide only one
oxygen atom for possible interactions with the enzyme.
We proposed that a phosphinamide linkage, as repre-
sented by molecule B, might be a good candidate to
provide the required structural properties. Indeed, this
idea has led to the discovery of a novel series of MMP
inhibitors with unique SAR.
Design
CGS27023A (Ciba-Geigy, Figure 1) was the first
potent, nonpeptidic MMP inhibitor to be described.10
Structure-activity relationship (SAR) data11,12 obtained
from this novel series of hydroxamic acids suggested
that the sulfonamide group is essential for potent
* To whom correspondence should be addressed.
† Procter and Gamble Pharmaceuticals, Health Care Research
Center.
‡ Procter and Gamble, Corporate Research Division.
§ Present address: Procter and Gamble Pharmaceuticals, Health
Care Research Center.
| Present address: Department of Molecular Pharmacology
&
Biological Chemistry, Northwestern University Medical School, Chi-
cago, IL 60611-3008.
10.1021/jm980142s CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/18/1998

88 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 1
Pikul et al.
Sch em e 1a
Sch em e 2a
a
Reagents: (a) see Scheme 1, steps (a) and (b); (b) 1. H2, Pd/C,
2. BnONH2, HOBT, EDAC, NMM; (c) H2, Pd/C.
a
Reagents: (a) R2CHO, NaBH3(CN); (b) R3R4P(O)Cl, Et3N,
silica gel chromatography; (c) NH2OH, KOH, MeOH; (d) R2NH2,
Et3N, DMF.
Ch em istr y
The preparation of hydroxamic acids 5 and 6 was
carried out as shown in Scheme 1. The D-amino acid
methyl ester 1 was subjected to the reductive alkylation
procedure using an appropriate aldehyde and sodium
cyanoborohydride. The resulting N-alkylated amino acid
2 was then allowed to react with alkylaryl (diaryl)
phosphinyl chloride in the presence of triethylamine to
produce the diastereomeric phosphinamides 3 and 4
which were separated by silica gel column chromatog-
raphy. The diastereomerically pure esters were con-
verted to the corresponding hydroxamic acids 5 and 6
using methanolic hydroxylamine generated by the treat-
ment of its hydrochloride salt with methanolic potas-
sium hydroxide.13
F igu r e 2. An ORTEP view of phosphinamide 16.
prepared by palladium-catalyzed hydrogenolysis of the
benzyl hydroxamate.
The absolute configuration at the phosphorus center
was established using high-resolution X-ray data ob-
tained for compound 16 shown in Figure 2.14 Because
16 and the corresponding benzyl hydroxamate showed
higher mobility on silica gel chromatography plates
compared to their diastereomers, the R configuration
on phosphorus was assigned by analogy to the more
mobile compounds for all diastereomeric pairs in the
series. This assignment was also supported by a char-
In the case of the glycine-derived inhibitors (R1 ) H),
intermediate 2 was prepared using methyl bromoacetate
(7) as the starting material. The reaction of 7 with the
corresponding primary amine in the presence of tri-
ethylamine in N,N-dimethylformamide proceeded cleanly
to give the corresponding N-alkylated product 2. Fur-
ther steps toward the target hydroxamic acid were
identical to those described above.
1
acteristic signal pattern in H NMR which was similar
for all diastereomeric pairs.
All hydroxamic acids were found to be stable in the
solid state and in aqueous solutions at pH > 4.0.
However, at low pH we observed a slow decomposition
due to hydrolysis of the phosphinamide bond. The
results of detailed studies of this process will be
published separately.
An alternative synthetic sequence involving the step-
wise introduction of the hydroxamic acid was also
developed and is shown in Scheme 2. The amino acid
benzyl ester 8 was converted into a mixture of diaster-
eomeric phosphinamides 9 using the methodology de-
scribed above (see Scheme 1). The free acids, obtained
after removal of the benzyl protecting group by hydro-
genolysis, were then coupled with O-benzylhydroxy-
lamine. The diastereomerically pure benzyl hydroxam-
ates 10 and 11 were obtained after silica gel chromatog-
raphy. The target hydroxamic acids 5 and 6 were
Resu lts a n d Discu ssion
SAR of R1 a n d R2 P osition . All compounds were
tested for the inhibition of truncated collagenase-1
(MMP-115) and stromelysin (MMP-316), and the data are
summarized in Tables 1, 2, and 3. The first molecule
prepared to test the concept of the sulfonamide-to-
phosphinamide switch was the simple glycine derivative
12. This compound, as a racemic mixture, showed

Synthesis of Phosphinamide-Based Hydroxamic Acids
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 1 89
Ta ble 3. SAR at the R3 and R4 Position
Ta ble 1. SAR at R1 and R2 Position in the Series with R
Configuration on the Phosphorus
IC50 (nM)a
IC50 (nM)a
compd
R1
R2
(CH2)2C6H5 Me Ph
(CH2)2C6H5 Ph Ph
Me Ph
R3
R4 MMP-1 MMP-3
12b
21b
13
19
22
H
H
525
854
700
1750
compd
12b
13
14
15
16
R1
R2
MMP-1
MMP-3
700
67.9
1600
2550
24.4
16.9 (43)c
H
(CH2)2C6H5
CH2C6H5
n-C6H13
H
525
120
Me
Me
Me
CH2C6H5
CH2C6H5
CH2C6H5
120
67.9
CH3
CH3
CH2i-Pr
CH2i-Pr
Et
Ph
Ph
Ph
608
700
1290
2510
20.5
49.5
6790
10300
16
23
CH2i-Pr CH2C6H5
CH2i-Pr CH2C6H5
Me Ph
Me Me
20.5
518
24.4
1040
CH2C6H5
CGS27023A
a
See Experimental Section for details of the enzyme assays.
Racemic mixture.
a
See Experimental Section for details of the enzyme assays.
b Racemic mixture. c Ki values reported12 for CGS27023A are given
in parentheses.
b
Both enzymes seem to have little tolerance for changes
in the R3 position. In the case of MMP-3, substituting
the methyl group (compound 13) by a slightly larger
ethyl group (compound 19) decreased the potency by 1
order of magnitude.
Ta ble 2. Effect of Configuration at the Phosphorus Atom on
Inhibition of Collagenase and Stromelysin
The presence of the R-phenyl group at position R4 is
essential for the high potency observed with this series
of MMP inhibitors. This is illustrated by a 20-30-fold
drop in potency observed for compound 23 (versus 16)
where the phenyl group is replaced by a small methyl
substituent. However, it is remarkable that 23 is still a
high nanomolar inhibitor with such a small P1′ sub-
stituent.
In h ibition of Gela tin a se, Ma tr ilysin , Neu tr op h il
Colla gen a se, Gela tin a se B a n d Colla gen a se-3.15
Selected molecules from the phosphinamide series have
also been tested for activity against several other matrix
metalloproteinases: gelatinase A (MMP-2), matrilysin
(MMP-7), neutrophil collagenase (MMP-8), gelatinase
B (MMP-9), and collagenase-3 (MMP-13) (Table 4). They
inhibited other matrix metalloproteinases with poten-
cies matching those observed for collagenase and stromel-
ysin. The only exception was matrilysin for which
considerably weaker inhibition at a low micromolar level
was observed.
IC50 (nM)b
compda
R1
R2
R3
MMP-1
MMP-3
15 (R)
17 (S)
16 (R)
18 (S)
19 (R)
20 (S)
CH2i-Pr
CH2i-Pr
CH2i-Pr CH2C6H5 CH3
CH2i-Pr CH2C6H5 CH3
CH3
CH3
H
H
CH3
CH3
2510
2550
130500
24.4
>100000
20.5
7120
608
9170
CH2C6H5 C2H5
CH2C6H5 C2H5
700
33300
49300
a
b
Configuration at the phosphorus center. See Experimental
Section for details of the enzyme assays.
promising potency of less than 1 µM against both
collagenase and stromelysin. A brief SAR study of a
series of inhibitors with the R configuration at the
phosphorus center (Table 1) pointed to the positive effect
of a bulky substituent R1 on the inhibitory potency (e.g.
compare compounds 12, 13, and 16). Appropriate sub-
stitution at R2 was also very important for in vitro
activity, with alkylaryl substituents giving the best
results (e.g. 13 versus 14 and 16 versus 15).
As shown in Table 2, the stereochemistry at the
phosphorus atom has a major influence on the potency
of the phosphinamide-based MMP inhibitors. A decrease
in potency of about 2 orders of magnitude was observed
when switching the configuration at phosphorus from
R to S. Some insight into the conformational require-
ments for such an effect was gained from X-ray studies
of the stromelysin-16 complex and is discussed below.
SAR of th e R3 a n d R4 P osition s. After establishing
that the correct absolute configuration of the phospho-
rus atom is essential for inhibitory activity, we briefly
studied the SAR of substituents R3 and R4 (see Table
3). Clearly an increase in size of R3 has a negative effect
on activity. This effect is especially pronounced for
compounds with substitution at the R-position to the
hydroxamic acid (R1 * H). Exchanging the methyl group
at R3 for a bulkier phenyl led to a 2-fold decrease in
potency in the glycine series (compare compound 12
versus 21), but a 2 orders of magnitude difference was
observed in the alanine series (compound 13 versus 22).
Str u ctu r e of Str om elysin -In h ibitor Com p lex.17
To better understand the binding interactions between
the phosphinamide-based hydroxamic acids and the
MMP enzymes we obtained a crystal of the truncated
stromelysin-16 complex and solved its structure using
the molecular replacement method. The catalytic site
of stromelysin with the bound inhibitor is shown
schematically in Figure 3. In general, interactions of the
hydroxamic acid with stromelysin were consistent with
those already reported in the literature.18 The hydrox-
amic acid acted as a bidentate ligand with the two
oxygen atoms in binding distance of 1.83 and 2.07 Å
from the active site Zn2+ ion. The terminal OH oxygen
and the nitrogen atom of the hydroxamic acid formed
additional hydrogen bonds with the oxygen Oꢀ2 (2.80
Å) of the carboxyl group of Glu-202 and the carbonyl
group (3.09 Å) of Ala-165, respectively (see Figures 3
and 4).
The most interesting feature of the enzyme-inhibitor
interactions was revealed by analysis of the environ-
ment around the phosphorus atom. The phosphinic acid
oxygen was clearly within the hydrogen-bonding dis-
tance to the N-H of Leu-164 (2.70 Å) and Ala-165 (3.19

90 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 1
Pikul et al.
Ta ble 4. Inhibition of MMP-2, MMP-7, MMP-8, MMP-9, and MMP-13
IC50 (nM)a
MMP-7
compd
MMP-1
120
MMP-2
MMP-3
MMP-8
MMP-9
MMP-13
13
16
nd
67.9
1860
886
106
nd
5.3
4.4
32.5
nd
7.4
4.3
20.5
13.3
24.4
20.6
CGS27023A
49.5 (33)b
9.1 (20)b
16.9 (43)b
4.3 (8)b
a
b
See Experimental Section for details of the enzyme assays. nd ) not determined. Ki values reported12 for CGS27023A are given in
parentheses.
der Waals distance to the hydrophobic side chains of
Val-163, Leu-164, and His-166, providing additional
positive interactions of the inhibitor with stromelysin.
Con clu sion s
In summary, we have designed a phosphinamide-
based compounds group as structural probes of enzyme-
inhibitor binding interactions in the region of Leu-164
and Ala-165 of stromelysin. The phosphinamide linkage
was introduced as a mimic of tetrahedral geometry and
conformational freedom of the sulfonamide group present
in a series of potent MMP inhibitors. It was also a mimic
of inactive amides by providing only one oxygen atom
(bound to phosphorus) capable of developing hydrogen
bond interactions with Leu-164 and/or Ala-165. Follow-
ing this design a series of phosphinamide-based hydrox-
amic acids was synthesized and tested for inhibitory
activity of several matrix metalloproteinases. Many of
the analogues were found to be potent inhibitors of
fibroblast collagenase (MMP-1), gelatinase A (MMP-2),
stromelysin (MMP-3), neutrophil collagenase (MMP-8),
gelatinase B (MMP-9), and collagenase-3 (MMP-13) and
moderate inhibitors of matrilysin (MMP-7). Optimum
enzyme inhibitory activity in the series was associated
with the R configuration at the phosphorus chiral
center, allowing for optimal hydrogen bond interactions
between the phosphinic acid oxygen atom and Leu-164
and Ala-165 of stromelysin. The presence of this inter-
action was confirmed by X-ray crystallography of an
inhibitor-enzyme complex. All the SAR and the X-ray
crystallography data show that (1) one oxygen atom can
provide sufficient binding interaction with Leu-164 and
Ala-165 of stromelysin and (2) the tetrahedral geometry
of the sulfonamide or the phosphinamide linkage is
critical for the potent inhibition of matrix metallopro-
teinases.
F igu r e 3. Binding interactions between 16 and stromelysin.
F igu r e 4. Active site of stromelysin with bound 16.
Å). Additionally, the phosphorus bound methyl group
was within van der Waals distance to Leu-164. Such
an arrangement of substituents at the phosphorus
center seemed to be quite optimal since the methyl-to-
ethyl switch at the phosphorus center resulted in a
6-10-fold decrease of potency against both target
enzymes (compare compound 13 versus 19). The en-
zyme’s sensitivity to changes in the phosphorus envi-
ronment was further confirmed by a loss in potency of
about 2 orders of magnitude upon the inversion of
configuration (see Table 2). In fact this is the area where
CGS27023A may experience some unfavorable interac-
tions with MMP enzymes by placing one of the sulfonyl
oxygen atoms close to the hydrophobic isobutyl chain
of Leu-164.
The phosphorus bound phenyl group was comfortably
placed in the S1′ pocket of stromelysin, similar to that
of CGS27023A,18 and no additional substituent on the
aryl ring was necessary to achieve low nanomolar
potency. Interestingly, replacement of the entire aro-
matic system bound to phosphorus by a methyl group
resulted in only a 20-30-fold drop in potency. The
N-benzyl and the isobutyl groups of 16 were within van
Exp er im en ta l Section
Gen er a l Meth od s a n d Ma ter ia ls. 1H and 13C NMR
spectra were measured at 300 and 75.4 MHz, respectively, on
a Bruker or GE instrument using chloroform as solvent unless
otherwise indicated. Mass spectra were measured at 70 eV (CI
mode). Analytical thin-layer chromatography (TLC) analysis
was performed using Merck DC-F254 silica gel plates. Flash
chromatography was performed as described by Still19 using
kieselgel 60 (230-400 mesh) silica gel. All organic extracts
were dried with anhydrous MgSO4 or Na2SO4 prior to solvent
removal on a rotary evaporator under reduced pressure.
Human synovial proMMP-3 was obtained from Dr. Hideaki
Nagase, University of Kansas Medical Center, Kansas City,
KS. Human fibroblast proMMP-1, human MMP-9, and human
recombinant MMP-7 catalytic domain were obtained from Dr.
Howard Welgus, J ewish Hospital, St. Louis, MO. Human
recombinant MMP-8 catalytic domain was obtained from Dr.
Harald Tschesche, University Bielefeld, Bielefeld, Germany.
Human recombinant proMMP-2 was purified from CHO cells
as described below. Human recombinant truncated MMP-3
and truncated MMP-1 were purified from Escherichia coli cells

Synthesis of Phosphinamide-Based Hydroxamic Acids
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 1 91
as described below. Mca-Pro-Leu-Gly-Leu-Dpa-Ala-Arg-NH2
was purchased from Bachem Bioscience, King of Prussia, PA.
(m, 2H), 3.84 (m, 1H), 4.09 (m, 1H), 7.43 (m, 3H), 7.78 (m,
2H); 13C NMR (CDCl3) δ 5.65, 16.19, 16.28, 21.85, 23.20, 60.27,
128.23, 128.38, 129.60, 131.39, 131.90; 31P NMR (CD3OD) δ
46.72.
Eth ylp h en ylp h osp h in ic Ch lor id e. To a solution of ethyl
ethylphenylphosphinate (2 g, 10 mmol) in benzene (200 mL)
was added oxalyl chloride (1.3 mL, 15 mmol). The mixture was
stirred for 3 h at room temperature and concentrated on a
rotary evaporator, and the product was dried under vacuum
for 12 h to give ethylphenylphosphinic chloride as an oil: 1H
NMR (CDCl3) δ 1.23 (m, 3H), 2.38 (m, 2H), 2.59 (m, 3H), 7.88
(m, 2H).
N-Ben zyl D-Ala n in e Meth yl Ester . D-Alanine methyl
ester (4 g, 28.66 mmol) was dissolved in methanol (100 mL),
and sodium acetate (5.88 g, 71.65 mmol) and benzaldehyde
(2.9 mL, 28.66 mmol) were added. The mixture was stirred
for 15 min, and then a solution of sodium cyanoborohydride
(1.08 g, 17.2 mmol) in methanol (5 mL) was added dropwise.
After the mixture was stirred for 2 h, the methanol was
evaporated under reduced pressure, and the product was
extracted into ether. The organic phase was washed with water
(2×). The crude product was purified by flash silica gel
chromatography (8:2 hexane:ethyl acetate) to give 3.3 g (44%
yield) of N-benzyl D-alanine methyl ester as an oil: 1H NMR
(CD3OD) δ 1.33 (d, 3H, J ) 7.1 Hz), 3.40 (q, 1H, J ) 7.0 Hz),
3.73 (s, 3H), 3.74 (AB, 2H, J ) 13.0, 39.7 Hz), 7.23-7.56 (m,
5H); 13C NMR (CD3OD) δ 18.95, 51.71, 51.88, 55.85, 127.08,
128.22, 128.38, 139.98, 176.52.
N-((R a n d S)-Eth ylp h en ylp h osp h in yl)-N-ben zyl-D-a la -
n in e Meth yl Ester . To a solution of ethylphenylphosphinic
chloride (1.04 g, 5.5 mmol) in dichloromethane (15 mL) was
added a solution of N-benzyl D-alanine methyl ester (1.37 g,
7.1 mmol) and N-methylmorpholine (1.36 mL, 12.4 mmol) in
dichloromethane (15 mL). A catalytic amount of 4-(dimethy-
lamino)pyridine was added, and the reaction was stirred 90 h
at room temperature. Two spots were observed on TLC. These
compounds were separated by flash silica gel chromatography
(95:5 ethyl acetate:methanol) to give two diastereomeric
products.
N-(2-P h en eth yl)glycin e Meth yl Ester Hyd r och lor id e.
A solution of phenylethylalanine (6.63 mL, 52.8 mmol) and
triethylamine (7.39 mL, 53 mmol) in anhydrous N,N-dimeth-
ylformamide (80 mL) was cooled to 0 °C, and to this mixture
was dropwise added a solution of methylbromoacetate (5 mL,
52.8 mmol) in anhydrous N,N-dimethylformamide (40 mL).
The reaction was stirred for 20 min at 0 °C. The mixture was
poured into 250 mL of ethyl acetate, washed with water (3×),
dried over sodium sulfate, and evaporated to give a colorless
oil. To prepare the hydrochloride the crude oil was dissolved
in ether (75 mL). In a separate flask, acetyl chloride (3.8 mL)
was added dropwise to 2.5 mL of methanol at 0 °C. This
solution was added dropwise to the ether solution. The
precipitated solids were collected by filtration to give 9.2 g (76%
yield) of N-(2-phenethyl)glycine methyl ester hydrochloride as
a colorless solid: 1H NMR (CDCl3) δ 3.31 (m, 4H), 3.89 (t, 2H,
J ) 5.5 Hz), 3.76 (s, 3H), 7.27 (m, 5H); 13C NMR (CDCl3) δ
36.23, 50.55, 51.54, 54.83, 126.03, 128.27, 128.47, 139.45,
172.55.
N -(Dip h e n ylp h osp h in yl)-N -(2-p h e n yle t h yl)glycin e
Meth yl Ester . Diphenylphosphinic chloride (0.42 mL, 2.2
mmol) was dissolved in dichloromethane (5 mL) and cooled to
0 °C. To this mixture was added a solution of ester N-(2-
phenethyl)glycine methyl ester hydrochloride (500 mg, 2.2
mmol) and N-methylmorpholine (0.73 mL, 6.6 mmol) in
dichloromethane (5 mL). The reaction was stirred for 16 h at
room temperature, washed with water and brine, dried over
sodium sulfate, and concentrated to give N-(diphenylphosphi-
nyl)-N-(2-phenylethyl)glycine methyl ester as
a colorless
solid: 1H NMR (CDCl3) δ 2.72 (m, 2H), 3.19 (m, 2H), 3.61 (s,
3H), 3.78 (d, 2H, J ) 6.0 Hz), 6.87 (m, 2H), 7.13 (m, 3H), 7.42
(m, 5H), 7.70-7.95 (m, 5H); 13C NMR (CDCl3) δ 34.97, 47.27,
49.67, 51.86, 126.26, 128.42, 128.58, 131.97, 132.46, 132.59,
138.62, 176.56.
N-H yd r oxy-2-[[d ip h en ylp h osp h in yl](2-p h en ylet h yl)-
a m in o]-a ceta m id e (21). N-(Diphenylphosphinyl)-N-(2-phe-
nylethyl)glycine methyl ester (160 mg, 0.41 mmol) was dis-
solved in methanol (2.5 mL). To this mixture was added
hydroxylamine hydrochloride (57 mg, 0.81 mmol), followed by
2 mmol of a 25% methanolic solution of sodium methoxide.
The reaction was stirred for 16 h, neutralized with 1 N
hydrochloric acid, and concentrated. The crude product was
purified by silica gel flash chromatography to give 41.6 mg
(26% yield) of hydroxamic acid 21 as a colorless solid: 1H NMR
(CD3OD) δ 2.83 (m, 2H), 3.18 (m, 2H), 3.68 (d, 2H, J ) 9.9
Hz), 6.96 (m, 2H), 7.16 (m, 3H), 7.52-7.65 (m, 7H), 7.94 (m,
3H); 13C NMR (CD3OD) δ 35.92, 47.67, 51.00, 127.47, 129.56,
129.95, 130.12, 130.59, 132.38, 133.67, 133.80, 139.91, 168.83;
31P NMR (CD3OD) δ 35.19; MS-IS m/z 395 [M + H]+, 417 [M
+ Na]+, 433 [M + K]+; HRMS calcd for C22H24N2O3P (M + H)+,
395.1525; Found, 395.1527. Anal. (C22H23N2O3P‚0.25H2O) C,
H, N.
N-((R)-E t h ylp h en ylp h osp h in yl)-N-b en zyl-D-a la n in e
Meth yl Ester : 41% yield; Rf 0.35, 100% ethyl acetate; 1H NMR
(CD3OD) δ 1.04 (m, 3H), 1.31 (d, 3H, J ) 7.3 Hz), 1.93 (m,
2H), 3.43 (s, 3H), 4.19-4.33 (m, 3H), 7.21-7.35 (m, 5H), 7.48
(m, 3H), 7.80 (m, 2H); 13C NMR (CDCl3) δ 5.78, 16.68, 20.14,
21.33, 47.45, 51.50, 53.22, 126.99, 127.54, 128.16, 128.28,
131.31, 131.43, 139.10 173.98; 31P NMR (CDCl3) δ 43.40.
N -((S )-E t h ylp h e n ylp h osp h in yl)-N -b e n zyl-D-a la n in e
Meth yl Ester : 25% yield; Rf 0.25, 100% ethyl acetate; 1H NMR
(CD3OD) δ 1.12 (m, 3H), 1.32 (d, 3H, J ) 7.3 Hz), 2.07 (m,
2H), 3.55 (s, 3H), 4.24 (AB, 2H, J AB ) 9.9 Hz, J AX ) J BX ) 3.7
Hz), 4.37 (m, 1H), 7.26 (m, 5H), 7.50 (m, 3H), 7.82 (m, 2H);
13C NMR (CDCl3) δ 6.06, 16.80, 20.04, 21.25, 47.53, 51.66,
53.04, 126.94, 127.60, 128.03, 128.19, 131.24, 131.36, 131.50,
138.80, 173.29; 31P NMR (CDCl3) δ 43.35.
N-Hydr oxy-2-[[m eth ylph en ylph osph in yl](2-ph en yleth -
yl)-a m in o]-a cet a m id e (12). Following the route described
above and using methylphenylphosphinic chloride compound
12 was obtained with 47% yield as a colorless solid: 1H NMR
(CD3OD) δ 1.80 (d, 3H, J ) 13.9 Hz), 2.75 (t, 2H, J ) 7.0 Hz),
N-Hyd r oxy-2(R)-[[(R)-eth ylp h en ylp h osp h in yl]a m in o]-
p r op ion a m id e (19). N-((R)-Ethylphenylphosphinyl)-N-ben-
zyl-D-alanine methyl ester (333 mg, 0.96 mmol) was treated
with a solution of NH2OH13 (3.3 mL, 1.76 M in methanol), and
the reaction was stirred at room temperature for 16 h. The
reaction mixture was neutralized with 1 M aqueous HCl and
concentrated on a rotary evaporator. The residue was purified
by flash silica gel chromatography (95:5 ethyl acetate:metha-
nol) to give 110 mg (33% yield) of 19 as a colorless solid: 1H
NMR (CD3OD) δ 1.06 (m, 3H), 1.27 (d, 3H, J ) 7.0 Hz), 2.03
(m, 2H), 3.95 (m, 1H), 4.52 (m, 2H), 7.25 (m, 5H), 7.59 (m,
3H), 7.82 (m, 2H); 13C NMR (CD3OD) δ 6.26, 18.12, 21.11,
22.31, 53.63, 128.11, 128.25, 128.49, 129.41, 129.79, 129.95,
132.74, 132.87, 133.50, 141.32, 171.98; 31P NMR (CD3OD) δ
48.14; MS-IS m/z 347 [M + H]+, 369 [M + Na]+. Anal.
(C18H23N2O3P‚0.15H2O) C, H, N.
3.13 (m, 2H), 3.71 (ABX, 2H, J AB ) 10 Hz, J AX ) 13 Hz, J BX
)
13 Hz), 7.03 (m, 2H), 7.17 (m, 3H), 7.54 (m, 3H), 7.82 (m, 3H);
13C NMR (CD3OD) δ 14.05, 15.28, 35.79, 47.41, 50.25, 127.48,
129.56, 129.80, 130.00, 132.49, 132.63, 133.59, 140.05; 31P
NMR (CD3OD) δ 43.85; MS-IS m/z 333 [M + H]+, 350 [M +
NH4]+, 355 [M + Na]+, 371 [M + K]+. Anal. (C17H21N2O3P) C,
H, N.
Eth yl Eth ylp h en ylp h osp h in a te. A mixture of diethyl
phenylphosphonite (4.5 g, 22.70 mmol), ethyl iodide (0.24 mL,
3 mmol), and benzene (100 mL) was stirred and heated at 85
°C for 24 h. Another portion of ethyl iodide (0.30 mL, 3.75
mmol) was added, and the reaction was stirred for an ad-
ditional 36 h at 85 °C. The volatiles were removed on a rotary
evaporator to give ethyl ethylphenylphosphinate as an oil: 1H
NMR (CD3OD) δ 1.08 (m, 3H), 1.27 (t, 3H, J ) 6.0 Hz), 1.88
N-Hyd r oxy-2(R)-[[(S)-eth ylp h en ylp h osp h in yl]a m in o]-
p r op ion a m id e (20). Following the method described for the
preparation of 19, compound 13 was obtained from the
corresponding methyl ester in 54% yield as a colorless solid:

92 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 1
Pikul et al.
1H NMR (CD3OD) δ 0.94 (m, 3H), 1.17 (d, 3H, J ) 7.0 Hz),
2.02 (m, 2H), 4.04 (m, 1H), 4.43 (ABX, 2H, J AB ) 13.4 Hz, J AX
) J BX ) 17.0 Hz); 13C NMR (CD3OD) δ 6.25, 16.92, 20.92,
22.35, 53.87, 128.27, 129.54, 129.94, 130.10, 132.93, 133.06,
133.68, 140.86, 173.10;31P NMR (CD3OD) δ 52.58; MS-IS m/z
347 [M + H]+, 369 [M + Na]+. Anal. (C18H23N2O3P‚0.2H2O) C,
H, N.
The reaction was stirred for 3 h, 10% aqueous NaHCO3 was
added with stirring, the reaction mixture was concentrated
using a rotary evaporator, and the product was extracted into
ether (3×). The combined organic phases were washed with
water (2×), dried over sodium sulfate, and evaporated to give
2.15 g (51% yield) of N-benzyl D-leucine benzyl ester as a
colorless oil: 1H NMR (CDCl3) δ 0.84 (d, 3H, J ) 6.6 Hz), 0.91
(d, 3H, J ) 6.6 Hz), 1.50 (t, 2H, J ) 7.3 Hz), 1.79 (m, 1H),
3.35 (t, 1H, J ) 7.2 Hz), 3.70 (AB, 2H, J ) 12.8, 56.6 Hz), 5.17
(s, 2H), 7.22-7.47(m, 10H); 13C NMR (CDCl3) δ 22.02, 22.59,
24.69, 42.56, 51.94, 59.18, 66.15, 126.84, 127.41, 128.13,
128.39, 128.81, 129.08, 139.66, 175.72.
N -Be n zyloxy-2(R )-[[(R )-m e t h ylp h e n ylp h osp h in yl]-
ben zyla m in o]-4-m eth ylp en ta n a m id e a n d N-Ben zyloxy-2-
(R)-[[(S)-methylphenylphosphinyl] benzylamino]-4-methyl-
p en ta n a m id e. Methyl phenyl phosphinic chloride (0.89 mL,
6.42 mmol) was dissolved in dichloromethane and cooled to 0
°C. To this mixture was added a solution of N-benzyl D-leucine
benzyl ester (2 g, 6.42 mmol), and N-methyl morpholine (1.5
mL, 13.48 mmol) in dichloromethane. A catalytic amount of
4-(dimethylamino)pyridine was added, and the reaction was
stirred at room temperature for 22 h. The mixture was
concentrated using a rotary evaporator, and the residue was
dissolved with ethyl acetate. This mixture was washed with
water and brine, dried over sodium sulfate, and concentrated.
The product was isolated by flash silica gel chromatography
(100% ethyl acetate) in 76% yield as a mixture of diastereo-
mers.
The following compounds were prepared according to the
procedure described above for 19 and 20.
N -H y d r o x y -2(R )-[[(R )-m e t h y lp h e n y lp h o s p h in y l]-
ben zyla m in o]-pr op ion a m id e (13): 42% yield; colorless solid;
1H NMR (CD3OD) δ 1.28 (d, 3H, J ) 10.3 Hz), 1.77 (d, 3H, J
) 13.9 Hz), 3.95 (m, 1H), 4.50 (m, 2H), 7.20-7.35 (m, 5H),
7.47-7.60 (m, 3H), 7.8 (m, 2H); 13C NMR (CD3OD) δ 14.81,
16.02, 18.19, 53.63, 128.13, 128.46, 129.44, 129.79, 129.69,
132.26, 132.40, 133.50, 141.43, 171.74; 31P NMR (CD3OD) δ
44.13; MS-IS m/z 333 [M + H]+, 355 [M + Na]+. Anal.
(C17H21N2O3P‚0.25H2O) C, H, N.
N -H y d r o x y -2(R )-[[(R )-m e t h y lp h e n y lp h o s p h in y l]-
h exyla m in o]-p r op ion a m id e (14): 56% yield; colorless solid;
1H NMR (CD3OD) δ 0.86 (t, 3H, J ) 7.0 Hz), 1.16 (m, 6H),
1.39 (d, 3H, J ) 7.1 Hz), 1.47 (m, 2H), 1.87 (d, 3H, J ) 13.9
Hz), 3.06 (m, 2H), 3.99 (m, 1H), 7.60 (m, 3H), 7.84 (m, 2H);
13C NMR (CD3OD) δ 14.31, 14.52, 16.66, 17.31, 23.59, 27.82,
32.46, 32.59, 45.50, 53.60, 129.85, 130.02, 132.27, 132.41,
133.55, 172.45; 31P NMR (CD3OD) δ 43.23; MS-IS m/z 327 [M
+ H]+, 349 [M + Na]+. Anal. (C16H27N2O3P) C, H, N.
N-Hydr oxy-2(R)-[[(R)-m eth ylph en ylph osph in yl]am in o]-
4-m eth ylp en ta n a m id e (15). Compound 15 was prepared
from the corresponding methyl ester (Rf ) 0.25, 100% ethyl
A mixture of diastereomeric benzyl esters (2.04 g, 4.53
mmol) was dissolved in methanol, 10% Pd/C (500 mg) was
added, and the reaction mixture was stirred under a hydrogen
atmosphere for 45 min. The mixture was filtered through
Celite, and the filtrate was collected and concentrated to
quantitatively give N-(R/S-methylphenylphosphinyl)-N-benzyl-
D-leucine as a white glassy substance.
1
acetate): 49% yield; colorless solid; H NMR (CD3OD) δ 0.95
(d, 3H, J ) 6 Hz), 0.98 (d, 3H, J ) 6.0 Hz), 1.20-1.41 (m, 3H),
1.61 (d, 3H, J ) 14.0 Hz), 3.39 (m, 1H), 7.60 (m, 3H), 7.82 (m,
2H); 13C NMR (CD3OD) δ 15.86, 17.10, 22.63, 23.13, 25.54,
45.53, 64.43, 129.71, 129.88, 132.52, 132.66, 133.25, 177.91;
31P NMR (CD3OD) δ 35.38; MS-IS m/z 283 [M + H]+. Anal.
(C13H19N2O3P) C, H, N.
A portion of this material (1.5 g, 4.17 mmol) was dissolved
in N,N-dimethylformamide and cooled to 0 °C. To this mixture
was added sequentially hydroxybenzotriazole hydrate (1.69 g,
12.5 mmol), N-methylmorpholine (1.37 mL, 12.5 mmol), and
1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDAC, 959
mg, 5 mmol). After the mixture was stirred for 10 min,
O-benzylhydroxylamine hydrochloride (666 mg, 4.17 mmol)
was added, and the reaction was stirred for 3 h, warming to
room temperature. Two diastereomers were observed by TLC.
To the mixture was added water, and the mixture was
extracted with ethyl acetate. The organic phases were com-
bined, washed with water and brine, dried over sodium sulfate,
and concentrated to give an oil. The diastereomers were then
isolated by flash silica gel chromatography (1:1 hexane:ethyl
acetate) to give N-benzyloxy-2(R)-[[(R)-methylphenylphosphi-
nyl]benzylamino]-4-methylpentanamide [37% yield; Rf ) 0.25
(1:1 hexane:ethyl acetate); 31P NMR (CD3OD) δ 43.89] and
N-benzyloxy-2(R)-[[(S)-methylphenylphosphinyl]benzylamino]-
4-methylpentanamide [22% yield; Rf ) 0.15 (1:1 hexane:ethyl
acetate); 31P NMR (CD3OD) δ 43.93].
N -H y d r o x y -2(R )-[[(R )-m e t h y lp h e n y lp h o s p h in y l]-
ben zyla m in o]-4-m eth ylp en ta n a m id e (16). N-Benzyloxy-
2(R)-[[(R)-methylphenylphosphinyl]benzylamino]-4-methyl-
pentanamide (460 mg, 0.99 mmol) was dissolved in methanol
(10 mL), 10% Pd/C (100 mg) was added, and the reaction
mixture was stirred under a hydrogen atmosphere for 2 h. The
mixture was filtered through Celite, and the filtrate was
collected and concentrated to give a white glassy solid. The
product was crystallized from ethyl acetate-hexane to give
278 mg (75% yield) of 16 as a white crystalline solid: 1H NMR
(CD3OD) δ 0.61 (d, 3H, J ) 6.4 Hz), 0.68 (d, 3H, J ) 6.4 Hz),
1.36-1.60 (m, 2H), 1.78 (m, 1H), 1.79 (d, 3H, J ) 13.7 Hz),
3.75 (m, 1H), 4.50 (ABX, 2H, J AB ) 12.4 Hz, J AX ) J BX ) 16.7
Hz), 7.19-7.33 (m, 5H), 7.54 (m, 3H), 7.78 (m, 2H); 13C NMR
(CD3OD) δ 14.92, 26.62, 22.35, 22.94, 25.87, 41.48, 56.19,
128.12, 128.87, 129.36, 129.83, 130.00, 132.17, 132.31, 133.50,
141.00, 171.29; 31P NMR (CD3OD) δ 44.28; MS-IS m/z 375 [M
+ H]+, 397 [M + Na]+, 413 [M + K]+. Anal. (C20H27N2O3P‚
0.5H2O) C, H, N.
N-Hydr oxy-2(R)-[[(S)-m eth ylph en ylph osph in yl]am in o]-
4-m eth ylp en ta n a m id e (17). Compound 17 was prepared
from the corresponding methyl ester (Rf ) 0.14, 100% ethyl
1
acetate): 53% yield; colorless solid; H NMR (CD3OD) δ 1.64
(d, 3H, 6.6 Hz), 1.81 (d, 3H, 6.6 Hz), 1.20-1.42 (m, 3H), 1.64
(d, 3H, J ) 14.7 Hz), 3.35 (m, 1H), 7.54 (m, 3H), 7.86 (m, 2H);
13C NMR (CD3OD) δ 15.32, 16.98, 22.35, 22.93, 25.46, 51.49,
129.67, 129.84, 132.53, 132.67, 133.34, 176.50; 31P NMR (CD3-
OD) δ 34.98; MS-IS m/z 283 [M + H]+. Anal. (C13H19N2O3P)
C, H, N.
N-Hyd r oxy-2(R)-[[d ip h en ylp h osp h in yl]ben zyla m in o]-
p r op ion a m id e (22): 39% yield; oil; 1H NMR (CD3OD) δ 1.25
(d, 3H, J ) 7.0 Hz), 4.04 (m, 1H), 4.39 (m, 2H), 7.16 (m, 4H),
7.40-7.61 (m, 6H), 7.80 (m, 3H), 7.92 (m, 2H); 13C NMR (CD3-
OD) δ 18.01, 54.59, 128.01, 128.72, 129.21, 129.91, 130.04,
132.60, 132.75, 133.49, 133.62, 133.71, 140.59, 171.43; 31P
NMR (CD3OD) δ 36.52; MS-IS m/z 395 [M + H]+, 417 [M +
Na]+; HRMS calcd for C22H24N2O3P (M + H)+, 395.1525; found,
395.1525. Anal. (C22H23N2O3P) C, H, N.
N-Hyd r oxy-2(R)-[[d im eth ylp h osp h in yl]ben zyla m in o]-
4-m et h ylp en t a n a m id e (23): 62% yield; colorless solid; 1H
NMR (DMSO) δ 0.58 (d, 3H, 6.4 Hz), 0.70 (d, 3H, J ) 6.4 Hz),
1.20-140 (m, 2H), 1.33 (d, 3H, J ) 12.5 Hz), 1.38 (d, 3H, J )
12.5 Hz), 1.70 (m, 1H), 3.78 (m, 1H), 4.20-4.50 (m, 2H), 7.10-
7.40 (m, 5H); 13C NMR (DMSO) δ 15.61, 16.05, 16.52, 17.21,
22.11, 22.56, 24.09, 45.57, 53.20, 126.42, 127.23, 128.03,
141.85, 168.41; 31P NMR (DMSO) δ 47.78; MS-IS m/z 313 [M
+ H]+, 335 [M + Na]+, 357 [M + K]+; HRMS calcd for
C15H26N2O3P (M + H)+, 313.1681; found, 313.1688. Anal.
(C15H25N2O3P) C, H, N.
N-Ben zyl D-Leu cin e Ben zyl Ester . D-Leucine benzyl ester
(3 g, 11.66 mmol) was dissolved in methanol, and to this
mixture was added sodium acetate (1.9 g, 23.3 mmol) followed
by benzaldehyde (1.2 mL, 11.66 mmol). The mixture was
stirred for 10 min, and a solution of sodium cyanoborohydride
(427 mg, 6.8 mmol) in methanol (4 mL) was added dropwise.

Synthesis of Phosphinamide-Based Hydroxamic Acids
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 1 93
N-Hydr oxy-2(R)-[[(S)-m eth ylph en ylph osph in yl]ben zyl-
a m in o]-4-m eth ylp en ta n a m id e (18). Following the method
described for the preparation of 16, compound 18 was obtained
in 72% yield from the corresponding benzyl ester as a colorless
solid: 1H NMR (CD3OD) δ 0.47 (d, 3H, J ) 6.6 Hz), 0.64 (d,
3H, J ) 6.6 Hz), 1.10 (m, 1H), 1.29 (m, 1H), 1.70 (m, 1H), 1.71
(d, 3H, J ) 13.7 Hz), 3.84 (m, 1H), 4.35-4.87 (m, 2H), 7.24
(m, 5H), 7.56 (m, 3H), 7.83 (m, 2H); 13C NMR (CD3OD) δ 12.73,
13.94, 20.47, 21.50, 24.22, 38.68, 54.09, 126.60, 126.82, 127.94,
128.30, 128.47, 130.92, 131.05, 132.04, 139.75, 169.23; 31P
NMR (CD3OD) δ 43.02; MS-IS m/z 375 [M + H]+, 397 [M +
Na]+, 413 [M + K]+. Anal. (C20H27N2O3P‚0.25H2O) C, H, N.
Sepharose 4B column equilibrated in MNC buffer, washed
with equilibration buffer to remove nonbinding proteins, and
eluted with MNC buffer containing 0.3 M NaCl and 10% (by
vol) DMSO. Elution fractions containing progelatinase A were
pooled, diluted, and loaded onto an S-Sepharose Fast Flow
column equilibrated in MNC buffer and eluted with MNC
buffer containing 0.3 M NaCl. The concentrated fractions of
purified proMMP-2 were combined and stored in MNC buffer
containing 0.3 M NaCl at -70 °C. N-Terminal sequence, amino
acid, and mass spectrophotometric analysis confirmed the
identity of the purified protein with the expected sequence for
progelatinase A.
Exp r ession a n d P u r ifica tion of Hu m a n Recom bin a n t
Exp r ession a n d P u r ifica tion of Hu m a n Recom bin a n t
Tr u n ca ted MMP -1. Briefly, DNA sequence coding for Val82-
Pro249 of proMMP-1 was amplified by polymerase chain
reaction from a commercially available plasmid, p35-1 (ATCC,
Rockville, MD; N. S. Templeton et al. Cancer Res. 1990 50,
5431-5437) encoding human interstitial MMP-1. The PCR
fragment was ligated into the expression vector, pET-11a
(Novagen, Madison, WI), and expressed in E. coli BL21(DE3)
cells. The protein was solubilized from inclusion bodies in 6
M urea and 0.15 M NaCl (pH 7.5); refolded in 50 mM Tris-
HCl (pH 7.5), 10 mM CaCl2, and 0.1 mM zinc acetate; and
then purified to homogeneity over a hydroxamic acid inhibitor
affinity column as previously described (Moore, W. A.; Spilburg,
C. A. Biochemistry 1986, 25, 5189-5195). N-Terminal se-
quence analysis confirmed the presence of three N-terminal
sequences of Val-Leu-Thr-Glu-Gly-Asn, Met-Val-Leu-Thr-Glu-
Gly-Asn, and Leu-Thr-Glu-Gly-Asn (minor).
Tr u n ca ted MMP -3. Briefly, DNA sequence coding for Ala-1
-
Thr255 of proMMP-3 was amplified by polymerase chain
reaction from a recombinant plasmid encoding human synovial
preproMMP-3 kindly provided by Dr. Hideaki Nagase (Uni-
versity of Kansas Medical Center, Kansas City, KS) and Dr.
Markku Kurkinen (Department of Medicine, UMDNJ -Robert
Wood J ohnson Medical School). After DNA sequence verifica-
tion, this DNA fragment was ligated into the expression vector,
pET-3d (Novagen, Madison, WI), and expressed in E. coli BL21
(DE3) cells as previously described (Marcy, A. I. et al.
Biochemistry 1991, 30, 6476-6483). ProMMP-3 was solubi-
lized form inclusion bodies in 8 M guanidine-HCl, refolded
in 100 mM Tris-HCl (pH 7.5), 10 mM CaCl2, 0.5 mM zinc
acetate, and activated overnight at 37 °C with 1.5 mM
p-aminophenylmercuric acetate. Active, truncated MMP-3 was
purified to homogeneity over a hydroxamic acid inhibitor
affinity column as described (Moore, W. A.; Spilburg, C. A.
Biochemistry 1986, 25, 5189-5195). N-Terminal sequence
analysis confirmed the sequence to be consistent with the
catalytic domain, Phe83-Thr255, of proMMP-3.
Exp r ession a n d P u r ifica tion of Hu m a n Recom bin a n t
p r oMMP -2. A partial cDNA clone for MMP-2 known as K-121
(Tryggvason, K. J . Biol. Chem. 1990, 265, 11077-11082.), was
obtained from ATCC and subcloned into the pBlueScript SK-
(pBS) plasmid. Sanger dideoxy sequencing revealed that the
first 134 bp of the coding sequence were missing from the 5′
end of K-121. To restore the missing sequence, two overlapping
90+-mer oligonucleotides were designed and synthesized.
These, along with a 3′ antisense oligonucleotide, were used as
primers to the K-121 template in a series of polymerase chain
reaction (PCR) experiments to synthesize a full-length MMP-2
cDNA. The PCR product was then subcloned into pBS. To
express MMP-2 in the mammalian CHO D- cell system, it was
first subcloned into the mammalian expression vector pJ T1
(J . Ting, CRD), which contains the DHFR gene. Recombinant
MMP-2/pJ T1 was Polybrene transfected into CHO D- cells.
Clonal cell populations were isolated and screened for produc-
tion of MMP-2 mRNA using specific oligonucleotide primers
and reverse transcription PCR. Ten clones producing the
highest levels of MMP-2 mRNA were selected, and the DHFR/
MMP-2 construct was amplified by gradually increasing media
methotrexate (MTX, a DHFR inhibitor) concentration. Clonal
selection was further narrowed after assessing MMP bioac-
tivity in an MMP fluorescence assay (M. Anastasio, CRD) and
on zymogram gels. Those clones showing activity were tested
for MMP-2 protein production by sequential Edman degrada-
tion protein sequencing (F. Wang, P&GP Cell & Molecular
Biology Core) and Western blot. On the basis of all results,
one clone was expanded for growth in roller bottles. Ap-
proximately 9 L of serum-free conditioned media with an
MMP-2 concentration of ∼35 mG/L were generated. ProMMP-2
was purified from conditioned media as previously described
(T. Crabbe et al. Eur. J . Biochem. 1993 218, 431-438) with
the following modifications. Conditioned serum-free media was
reduced in volume with an Amicon S1Y30 spiral-wound
cartridge prior to chromatography on a gelatin-Sepharose 4B
column equilibrated in 25 mM Tris/HCl, 30 mM NaCl, 10 mM
CaCl2, pH 7.5 (TNC buffer). The column was washed with
equilibration buffer before elution of the bound protein by TNC
buffer containing 1 M NaCl followed by 1 M NaCl and 10%
(by vol) dimethyl sulfoxide (DMSO) in TNC buffer. ProMMP-2
fractions were concentrated in an Amicon ultrafiltration cell
with a YM30 membrane and diafiltered against 25 mM MES/
NaOH, 30 mM NaCl, 10 mM CaCl2, pH 6.0 (MNC buffer). The
concentrate was then chromatographed over a second gelatin-
Colla gen a se (MMP -1) In h ibition Assa y. ProMMP-1 was
activated prior to assay by treatment with trypsin. MMP-1
activity was monitored using a fluorescence assay previously
described.15 In a Dynatech MicroFLUOR plate, 10 nM of
activated collagenase was incubated with 8 µM Mca-Pro-Leu-
Gly-Leu-Dpa-Ala-Arg-NH2 in 50 mM Tris-HCl, 10 mM CaCl2,
0.15 M NaCl, 0.05% Brij for 20-30 min at 37 °C in the
presence of varying concentrations of inhibitor. The reaction
was then quenched with 50 mM EDTA and the fluorescence
increase monitored on a Perkin-Elmer LS50B spectrofluorom-
eter (λex 328 nm, λem 393 nM). Activity was measured as a
percentage of control activity in the absence of inhibitor.
Inhibitor concentrations were run in triplicate, and IC50
determinations were calculated from a four-parameter logistic
fit of the data within a single experiment.
Str om elysin (MMP -3) In h ibition Assa y. ProMMP-3 was
activated prior to assay by treatment with p-aminophenylm-
ercuric acetate or trypsin. MMP-3 activity was measured by
following the degradation of 3[H]-reduced, carboxymethylated
transferrin.16 In a Multiscreen DP filtration plate (Millipore),
50 ng of activated MMP-3, 30 µg of 3[H]-transferrin, and
varying concentrations of inhibitor were incubated in a buffer
of 50 mM Tris-HCl, 10 mM CaCl2, 0.15 M NaCl, 0.05% Brij
(pH 7.5), at 37 °C for 3 h. The reaction was quenched by
addition of 4.4% TCA, and TCA-soluble fragments were
counted for radioactivity. Activity was measured as a percent-
age of control activity in the absence of inhibitor. Inhibitor
concentrations were run in triplicate, and IC50 determinations
were calculated from a four-parameter logistic fit of the data
within a single experiment.
Gela tin a se A (MMP -2), Ma tr ilysin (MMP -7), Neu tr o-
p h il Colla gen a se (MMP -8), Gela tin a se B (MMP -9), a n d
Ra t Colla gen a se-3 (MMP -13) In h ibition Assa y. Inhibition
of all enzymes was measured according to the representative
procedure described below for MMP-2. ProMMP-2 was acti-
vated prior to assay by treatment with 1 mM p-aminophe-
nylmercuric acetate for 45 min at 37 °C. ProMMP-9 was
activated with MMP-3 (1:20 MMP-3:MMP-9) and stored at -80
°C until use. MMP-7, MMP-8, and rat collagenase were all
supplied as active enzymes and stored frozen until use. MMP
activity was monitored using a fluorescence assay previously

94 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 1
described,15 modified for
microtiter plate format. In a
Pikul et al.
(8) Rasmussen, H. S.; McCann, P. P. Matrix Metalloproteinase
Inhibition as a Novel Anticancer Strategy: A Review with
Special Focus on Batimastat and Marimastat. Pharmacol. Ther.
1997, 75, 69-75.
(9) Bramhall, S. R. The Matrix Metalloproteinases and Their
Inhibitors in Pancreatic Cancer. Int. J . Pancreatol. 1997, 21,
1-12.
(10) MacPherson, L. J .; Parker, D. T. New Aryl-sulphonyl-amino-
aceto-hydroxamic Acid Derivatives - Useful as Metalloprotease
Inhibitors for Treating Arthritis, Ulcers, Periodontal Disease,
Tumour Metastasis, HIV Infections, etc. European Patent Ap-
plication, EP606046A, 1993.
(11) Parker, D. T.; MacPherson, L. J .; Goldstein, R.; J ustice, M. R.;
Ziu, L. J .; Capparelli, M.; Whaley, L. W.; Boehm, C.; O’Byrne,
E. M.; Goldberg, R. L.; Ganu, V. S. CGS27023A: A Novel, Potent,
and Orally Active Matrix Metalloprotease Inhibitor. P 73 7th
International Conference of the Inflammation Research Associa-
tion, White Haven, PA, Sept 25-29, 1994.
(12) MacPherson, L. J .; Bayburt, E. K.; Capparelli, M. P.; Carroll,
B. J .; Goldstein, R.; J ustice, M. R.; Ziu, L. J .; Hu, S-i.; Melton,
R. A.; Fryer, L.; Goldberg, R. L.; Doughty, J . R.; Spirito, S.;
Blancuzzi, V.; Wilson, D.; O’Byrne, E. M.; Ganu, V. S.; Parker,
D. T. Discovery of CGS 27023A, a Non-Peptidic, Potent, and
Orally Active Stromelysin Inhibitor That Blocks Cartilage
Degradation in Rabbits. J . Med. Chem. 1997, 40, 2525-2532.
(13) Fieser, L. F.; Fieser, M. Reagents for Organic Synthesis; J ohn
Willey and Sons: New York, 1967; Vol. 1, pp 478-479.
(14) We thank Dr. Fred C. Wireko of Corporate Research Division,
Procter & Gamble Co., for providing us with the high-resolution
X-ray structure of 16.
(15) Knight, C. G.; Willenbrock, F.; Murphy, G. A Novel Coumarin-
labeled Peptide for Sensitive Continuous Assays of the Matrix
Metalloproteinases. FEBS Lett. 1992, 296, 263-266.
(16) Okada, Y.; Nagase, H.; Harris, E. D. A Metalloproteinase from
Human Rheumatoid Synovial Fibroblasts that Digests Connec-
tive Tissue Matrix Components. Purification and Characteriza-
tion. J . Biol. Chem. 1986, 261, 14245-14255.
a
Dynatech MicroFLUOR plate, active enzyme was incubated
with 8 µM Mca-Pro-Leu-Dpa-Ala-Arg-NH2 in 50 mM Tris-HCl
(pH 7.5), 10 mM CaCl2, 0.15 M NaCl, 0.05% Brij, for 20-30
min at 37 °C in the presence of varying concentrations of
inhibitor. The reaction was then quenched with 50 mM EDTA,
and the relative fluorescence was monitored on a Perkin-Elmer
LS50B spectrofluorometer (λex 328 nm, λem 393 nM) fitted
with a microplate reader attachment. Activity was measured
as a percentage of control activity in the absence of inhibitor.
Inhibitor concentrations were run in triplicate, and IC50
determinations were calculated from a four-parameter logistic
fit of the data within a single experiment.
Ack n ow led gm en t. The authors thank the numer-
ous P&GP scientists associated with the MMP project
whose hard work and dedication over many years have
contributed toward the completion of this work.
Su p p or tin g In for m a tion Ava ila ble: X-ray crystallo-
graphic data for 16 (8 pages). Ordering information is given
on any current masthead page.
Refer en ces
(1) Woessner, J . F. Matrix Metalloproteinases and their Inhibitors
in Connective Tissue Remodeling. FASEB J . 1991, 5, 2145-
2154.
(2) Firestein, G. S.; Paine, M. M.; Littman, B. H. Gene Expresion
(Collagenase, Tissue Inhibitor of Metalloproteinase, Complement
and HLA-DR) in Rheumatoid and Osteoarthritic Synovium.
Arthritis Rheum. 1991, 34, 1094-1105.
(3) Walakovits, L. A.; Bhardwaj, N.; Gallick, G. S.; Lark, M. W.
Detection of High Levels of Stromelysin and Collagenase in
Synovial Fluid in Patients with Rheumatoid Arthritis and Post-
Traumatic Knee Injury. Arthritis Rheum. 1992, 35, 35-42.
(4) Pyke, C.; Ralfkiaer, E.; Huhtala, P.; Hurskainen, T.; Danoe, K.;
Tryggvason, K. Localization of Messenger RNA for Mr 72,000
and 92,000 Type IV Collagenases in Human Skin Cancers by in
situ Hybridization. Cancer Res. 1992, 52, 1336-1341.
(5) Overall, C. M.; Wiebkin, O. W.; Thonard, J . C. Demonstration
of Tissue Collagenase Activity in vivo and its Relationship to
Inflammation Severity in Human Gingiva. J . Periodontal Res.
1987, 22, 81-88.
(17) Protein Data Bank file name: 163d.
(18) (a) Babine, E. R.; Bender, S. L. Molecular Recognition of
Protein-Ligand Complexes: Application to Drug Design. Chem.
Rev. 1997, 97, 1359-1472. (b) Gonella, N. C.; Li, Y.-C.; Zhang,
X.; Paris, C. G. Bioactive Conformation of a Potent Stromelysin
Inhibitor Determined by X-nucleus Filtered and Multidimen-
sional NMR Spectroscopy. Bioorg. Med. Chem. 1997, 5, 2193-
2201.
(19) Still, W. C.; Kahn, M.; Mitra, A. Rapid Chromatographic
Technique for Preparative Separations with Moderate Resolu-
tion. J . Org. Chem. 1978, 43, 2923-5.
(6) Gijbels, K.; Galardy, R. E.; Steinman, L. Reversal of Experi-
mental Autoimmune Encephalomyelitis with a Hydroxamate
Inhibitor of Matrix Metalloproteinases. J . Clin. Invest. 1994, 94,
2177-2182.
(7) Zask, A.; Levin, J . I.; Killar, L. M.; Skotnicki, J . S. Inhibition of
Matrix Metalloproteinases: Structure Based Design. Curr.
Pharm. Des. 1996, 2, 624 and references therein.
J M980142S