A peptide hydroxamate library for enrichment of metalloproteinases: Towards an affinity-based metalloproteinase profiling protocol

Article abstract of DOI:10.1039/b718352f

A compound library of 96 enantiopure N-terminal succinyl hydroxamate functionalized peptides was synthesized on solid phase. All compounds were tested for their inhibitory potential towards MMP-9, MMP-12 and ADAM-17, which led to the identification of both broad spectrum inhibitors and metalloproteinase-selective ones. Eight potent and less potent inhibitors were immobilized on Sepharose beads and evaluated in solid-phase enrichment of active MMP-9, MMP-12 and ADAM-17. In addition, one of these inhibitors was used for solid-phase enrichment of endogenous ADAM-17 from a complex proteome (a lysate prepared from cultured A549 cells). This journal is The Royal Society of Chemistry.

Full text of DOI:10.1039/b718352f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
A peptide hydroxamate library for enrichment of metalloproteinases: towards
an affinity-based metalloproteinase profiling protocol†
Paul Geurink,‡a Theo Klein,‡b Michiel Leeuwenburgh,a Gijs van der Marel,a Henk Kauffman,c
Rainer Bischoff*b and Herman Overkleeft*a
Received 28th November 2007, Accepted 18th January 2008
First published as an Advance Article on the web 22nd February 2008
DOI: 10.1039/b718352f
A compound library of 96 enantiopure N-terminal succinyl hydroxamate functionalized peptides was
synthesized on solid phase. All compounds were tested for their inhibitory potential towards MMP-9,
MMP-12 and ADAM-17, which led to the identification of both broad spectrum inhibitors and
metalloproteinase-selective ones. Eight potent and less potent inhibitors were immobilized on
Sepharose beads and evaluated in solid-phase enrichment of active MMP-9, MMP-12 and ADAM-17.
In addition, one of these inhibitors was used for solid-phase enrichment of endogenous ADAM-17
from a complex proteome (a lysate prepared from cultured A549 cells).
in the active site.7,12 As a result, a requirement for potent MMP
Introduction
or ADAM inhibitors is that they contain a good zinc binding
Matrix metalloproteinases (MMPs) are involved in numerous
group. A large number of MMP and ADAM inhibitors that have
biological processes such as cell migration, wound repair and
appeared in the literature consist of an oligopeptide sequence that
tissue remodeling. MMPs exert their role by the processing
is equipped with a hydroxamate moiety at either the C- or the
of extracellular matrix proteins including gelatin, elastin, and
N-terminus. In these structures the oligopeptide portion ensures
collagen and the release of growth factors. ADAMs (a disintegrin
recognition by the metalloproteinases, whereas the hydroxamate
and metalloproteinase) are metalloproteinases that contain a
acts as a zinc chelator.
membrane-spanning and a disintegrin (integrin-binding) domain.
C-terminal peptide hydroxamic acids are readily avail-
These membrane-bound enzymes are involved in membrane
able through modified solid-phase peptide synthesis (SPPS)
protocols.13–18 In contrast, there are very few synthetic procedures
towards N-terminal peptide hydroxamates,19–21 which obviate a
non-SPPS step during synthesis.22–26 The preparation of compound
libraries containing N-terminal peptide hydroxamates would be
greatly facilitated by the existence of suitable, complete SPPS
methods. With this aim in mind, we recently reported the
synthesis of an enantiomerically pure N,O-diprotected succinyl
hydroxamate building block 1 (Fig. 1) and demonstrated that N-
terminal peptide hydroxamates can be prepared by SPPS using
compound 1 in the penultimate step, prior to acid cleavage and
deprotection.27 Here we report the application of building block 1
in the preparation of a library containing 96 enantiopure peptide
hydroxamates 2 (Fig. 1). We further demonstrate the use of several
members of our library in the solid-phase extraction of active
metalloproteinases from complex biological samples.28,29
fusion, cytokine and growth factor shedding, cell migration,
muscle development, fertilization, cellular differentiation, cell–
cell interactions and cell–matrix interactions.1–3 The best known
ADAM is ADAM-17, also known as TACE or tumor necrosis
factor a (TNFa) converting enzyme, which was discovered based
on its sheddase activity with respect to membrane-bound TNFa.4,5
The expression of MMPs and ADAMs is regulated by transcrip-
tion factors and activity is controlled by natural inhibitors, the
tissue inhibitors of metalloproteinases (TIMPs). Disturbances
in these regulatory mechanisms are believed to cause, or be
involved in, a wide range of pathological states. These include
cancer metastasis, rheumatoid arthritis and autoimmune diseases.
Deregulation of ADAM expression or activity has also been
linked to asthma, Alzheimer’s disease, bacterial lung infections and
allergies of the airways.1,6–11 MMPs and ADAMs contain a Zn2+
ion in their active site, which forms a complex with the carbonyl
group of the scissile amide bond, thereby enhancing its reactivity
towards nucleophilic attack of the water molecule that is present
aDepartment of Bio-organic Synthesis (BioSyn), Leiden University, P.O. Box
9502, 2300 RA Leiden, The Netherlands. E-mail: h.s.overkleeft@chem.
leidenuniv.nl; Fax: +31 71-527-4307
bCenter of Pharmacy, Analytical Biochemistry, University of Groningen,
Antonius Deusinglaan 1, 9713 GZ Groningen, The Netherlands. E-mail:
r.p.h.bischoff@rug.nl; Fax: +31 50-363-7582
cDepartment of Pathology, University Medical Center Groningen,
Hanzeplein 1, 9731 GZ Groningen, The Netherlands
† Electronic supplementary information (ESI) available: Experimental de-
tails for building block 1 and compound library 2 (including spectroscopic
data) and all biological experiments. See DOI: 10.1039/b718352f
Fig. 1 Structure of the chiral succinylhydroxamate building block 1
for SPPS of a library of 96 compounds with the general structure 2.
P1ꢀ, P2ꢀ and P3ꢀ refer to the binding pockets in the metalloproteinase.
Boc: tert-butyloxycarbonyl, TBS: tert-butyldimethylsilyl, PFP: pentafluo-
rophenyl. R1 and R2 represent amino acid side chains.
‡ These authors contributed equally to this work.
1244 | Org. Biomol. Chem., 2008, 6, 1244–1250
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Scheme 1 Solid-phase synthesis of a succinylhydroxamate library containing 96 compounds. Reagents and conditions: (a) 20% piperidine–DMF, 15 min.;
(b) Fmoc-Aa1-OH (5 equiv.), HCTU (5 equiv.), DIPEA (10 equiv.), NMP, 1 h; (c) 20% piperidine–DMF, 15 min.; (d) Fmoc-Aa2-OH (5 equiv.), HCTU
(5 equiv.), DIPEA (10 equiv.), NMP, 1 h; (e) 20% piperidine–DMF, 15 min.; (f) compound 1 (5 equiv.), DIPEA (2 equiv.), NMP, 2 h; (g) 95% TFA–H2O,
1 h (2.5 h for Aa1 = Arg(Pmc)). Aa1 = D, E, F, H, L, P, Q, R, S, T, W or Y; Aa2 = A, D, F, H, L, V, W or Y.
40%. The amount of side products formed differed considerably
between the compounds. Hydrolysis of the hydroxamic acid to
Results and discussion
the carboxylic acid in the final step appeared in most cases to be
the major side reaction. The formation of this side product was
apparent from the LC-MS analyses of the crude mixtures by a
15 Da decrease in molecular weight. In some cases condensation
with the activated hydroxamate ester was incomplete. In general,
the best results in terms of yield and side product formation
were obtained for compounds containing an amino acid with
an aliphatic side chain at the Aa2 position. Representative LC-
MS analyses of crude peptides with and without high levels of
side products together with their HPLC-purified counterparts are
shown in the supporting information.†
The results of the inhibitory potential of the 96 compounds
against MMP-9, MMP-12 and ADAM-17 are depicted in Fig. 2
(heat map representation, the compounds were screened at
100 nM). We performed this initial screen to obtain qualitative
insight into the difference in inhibitory potential of the 96 peptide
The preparation of the target compound library (see Scheme 1)
commenced with a-NHFmoc-, e-NHBoc-protected lysine on Rink
amide resin 3. After removal of the Fmoc protecting group, the
first set of amino acids (Aa1) was coupled in a parallel fashion
under standard SPPS coupling conditions giving 12 different
peptides. These resin bound peptides were divided into 8 equal
portions. Removal of the next Fmoc group and coupling of the
second amino acid (Aa2) gave 96 immobilized peptides with the
general structure 4. Final Fmoc deprotection and condensation
with building block 1 (see Fig. 1) in the presence of 2 equivalents
of DIPEA resulted in the immobilized and fully protected peptide
hydroxamates 5. Acidic cleavage from the resin and concomitant
deprotection of the TBS and Boc protecting groups resulted in a
96-membered library of crude compounds 2, which were purified
by HPLC. The yields of the pure peptides based on 3 (purity
>95% as determined by LC-MS analysis) varied between 3% and
Fig. 2 Remaining enzymatic activity of three recombinant metalloproteinases (5 ng) at 100 nM of inhibitor (black: no inhibition; white: complete
inhibition). Each value is the average of three individual experiments: (a) MMP-12 (0.25 pmol); (b) MMP-9 (0.25 pmol); (c) ADAM-17 (66 fmol).
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 1244–1250 | 1245
©

View Article Online
Table 1 IC50 values (nM) of eight selected inhibitors for MMP-9, MMP-
12 and ADAM-17. Each value represents the mean of three independent
inhibition curves (standard deviation in parentheses)
hydroxamates. It is apparent that the efficacy of the inhibitors
towards MMP-12 is generally higher than for the other two
enzymes. Introduction of a proline residue at the Aa1-position
greatly decreases the activity of the inhibitor with respect to
both MMPs. This effect appears to be strongest for MMP-12
but inhibition of ADAM-17 appears to be less affected. This
observation can be explained by the fact that MMPs contain a
straight horizontal cleft and therefore a proline would result in a
large steric hindrance within the active site. ADAMs, however,
do not contain such a rigid cleft and the inhibitor activity
is thus less affected by proline.30,31 It is also obvious that the
presence of acidic residues (D and E) in either position greatly
reduces the efficacy of the inhibitors for MMP-9 and ADAM-
17 and to a somewhat lesser extent for MMP-12. The inhibitors
with the highest efficacy towards MMP-12 are those with the
aromatic amino acids phenylalanine, tryptophan or tyrosine in
either position. These results are in line with earlier observations
by Lang and co-workers.31
The beneficial effect of incorporating aromatic moieties also
holds true for MMP-9, especially if both positions are occupied
by phenylalanine, tryptophan or tyrosine. Interestingly, a serine
residue in the Aa1 position yields very active MMP-9 inhibitors,
whereas threonine at Aa1 has a much weaker beneficial effect.
MMP-12 and ADAM-17 show a similar, albeit not so strong
tendency. The presence of an aliphatic amino acid (A, L or
V) in the Aa2 position decreases the efficacy against MMP-9
in a more pronounced way than for the other two enzymes.
Netzel-Arnett and co-workers reported an extensive study on the
substrate preference of MMP-9 towards a set of oligopeptides.32
Their findings corroborate our results with respect to a positive
effect of aromatic moieties in both positions (Aa1 and Aa2) or
a serine residue at Aa1 on inhibitory potential towards MMP-9.
Interestingly, our results are in disagreement with their findings
that leucine, and to a lesser extent alanine, at Aa2 have a beneficial
effect on inhibitory potential, since we observe a detrimental effect
for both residues at this position. Incorporation of arginine in
position Aa1 improves efficacy towards MMP-12 and ADAM-17
but highly reduces the efficacy towards MMP-9, as was also shown
by Netzel-Arnett and co-workers.32 The positive effect of aromatic
amino acids also holds true for ADAM-17 but to a lesser extent
than for the tested MMPs. In addition it is found that heteroaryl
moieties (His and Trp) or a serine at the Aa1 position improves the
potency towards ADAM-17. These observations are consistent
with reports in the literature.7,33
MMP-9
MMP-12
ADAM-17
DV
FF
FW
PD
PL
QY
SF
905 (221)
23.2 (3.9)
6.69 (0.66)
>10,000a
3624 (328)
9.92 (0.79)
9.93 (1.3)
6.71 (0.96)
10.5 (4.0)
0.92 (0.22)
2.57 (0.80)
2788 (392)
147 (12)
0.85 (0.020)
7.70 (1.3)
4.03 (0.95)
2241 (250)
16.0 (6.4)
29.6 (9.1)
5998 (2555)
92.1 (28)
18.9 (2.0)
11.1 (2.3)
36.0 (3.4)
YW
a Activity of enzyme greater than 50% at 10 lM inhibitor.
exhibits no non-specific interaction with proteins in the sample
is practically impossible, good controls are required. For these
experiments two control materials were used: (a) NHS-Sepharose
that was reacted with ethanolamine instead of the inhibitors
to study non-specific interaction with the Sepharose itself, and
(b) a low-affinity inhibitor (PD) was immobilized to assess the
importance of a good fit with the enzyme’s active site.
Table 2 shows the results of extraction studies with the eight
selected inhibitors. Quantitative extraction of ADAM-17 proves
much more challenging than extraction of the MMPs, which may
be caused by the fact that the recombinant ADAM-17 ectodomain
is substantially larger in size than the catalytic domains of the
MMPs used in this study, possibly giving rise to steric effects that
decrease extraction yield. MMP-9 and MMP-12 can be extracted
by a number of inhibitors with yields above 99% at both 5 and
0.5 nM. Inhibitors QY, SF, FF and YW show enrichment of
ADAM-17 at 5 nM enzyme concentration (extraction yields of
70% or higher). The efficiency drops significantly at 0.5 nM for
inhibitors QY, SF and YW. The negative control with immobi-
lized ethanolamine shows no detectable non-specific binding to
ADAM-17 at both concentrations which is also the case for the
low-affinity inhibitor PD. The results (summarized in Table 2)
show that it is difficult to classify immobilized inhibitors according
to their inhibition efficacy by affinity-SPE (see Tables 1 and 2)
and that it may be misleading to assess inhibitor selectivity in
this manner.28 One interesting inhibitor in this respect is DV,
which gives extraction yields of >99% for MMP-12 at both
concentrations while its IC50 value is 10.5 nM. This value is higher
than, for example, inhibitor YW, which extracts between 96% and
99%. Inhibitor DV also loses its selectivity towards MMP-12 after
immobilization, since both MMP-12 and MMP-9 were almost
completely extracted even at a low concentration. In contrast
ADAM-17 is not extracted at all although the IC50 values for
MMP-9 and ADAM-17 are in the same range.
To assess the applicability of the novel inhibitors for activity-
based extraction34 eight inhibitors were selected for further
experiments. The IC50 values (see Table 1) of these inhibitors for
the target enzymes span the entire range from sub-nanomolar
to over 10 lM (see for instance FF and PD) and some of
them show considerable selectivity towards one or two of the
three tested enzymes (for example YW towards both MMPs and
PL towards ADAM-17). Activity-based extraction of the three
metalloproteinases was performed at both 5 and 0.5 nM enzyme
concentrations.35 Especially at lower concentrations of active
enzyme, highly efficient interaction of the immobilized ligand
with metalloproteinases in the sample is likely to be critical for
efficient extraction. Next to achieving high affinity, it is important
to minimize non-specific binding to the carrier material, especially
at low enzyme concentrations. Since finding a carrier material that
Fig. 3 shows activity-dependent enrichment of ADAM-17 from
a complex proteome (a lysate prepared from cultured A549
cells) using immobilized inhibitor FF. The extraction was almost
complete, with no loss of mature 70 kDa ADAM-17 in the flow-
through fraction (FT) and minor loss in the washing fractions.
Active ADAM-17 could be eluted with a competitive inhibitor
with a similar IC50 value (SF; see Table 1), further confirming
that the interaction was inhibitor-mediated. We have shown
in previous work that inhibitor–metalloproteinase interactions
are strictly dependent on a functional active site and that
1246 | Org. Biomol. Chem., 2008, 6, 1244–1250
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 2 Extraction yield (%) of three recombinant active metalloproteinases by using the immobilized inhibitors as affinity ligand (standard deviation
in parentheses)
Inhibitor
Control
Enzyme
Conc./nM DV
FF
98.9 (0.85)
98.8 (0.99)
98.9 (0.92)
>99
72.9 (1.8)
73.7 (9.8)
FW
PD
7.2 (9.4)
31.7 (24)
96.6 (0.071)
75.7 (11)
PL
98.9 (1.1)
96.9 (0.92)
97.8 (0.49)
80.8 (20)
QY
SF
YW
EAa
MMP-9
5.0
0.5
5.0
0.5
5.0
0.5
98.7 (1.2)
99.0 (0.78)
>99
95.1 (0.14)
>99
35.7 (24)
43.3 (10)
98.4 (0.57)
98.6 (1.3)
>99
>99
88.2 (12)
4.60 (5.8)
>99
>99
>99
97.1 (1.3)
96.0 (5.0)
70.0 (6.4)
2.15 (2.3) <1
<1
<1
<1
<1
<1
>99
>99
>99
<1
98.5 (0.21)
98.9 (0.92)
>99
85.7 (1.3)
49.6 (22)
MMP-12
ADAM-17
<1
<1
65.8 (0.85)
<1
<1
a EA: ethanolamine.
enzyme–inhibitor complexes or pro-enzymes are not bound.29
The minimal losses in the flow-through and washing fractions
indicate that there is little inactive ADAM-17 in non-stimulated
A549 cells and that the interaction is relatively tight. We anticipate
that enrichment of active ADAM-17 from larger sample volumes
will be possible allowing detection of low levels of active ADAM-
17. Extraction of the same sample on a control material that was
derivatized with ethanolamine (EA cartridge) did not result in any
ADAM-17 enrichment.
inhibitors or by inactivation of ADAM-17 by another, unknown
mechanism. After two hours of exposure the overall level of
ADAM-17 has decreased significantly (Fig. 4, lower panel, lane S).
Down-regulation of ADAM-17 after cell stimulation has been
described.37 Our results show, however, that there is still a small
amount of active ADAM-17.
Fig. 4 Activity-dependent enrichment of ADAM-17 from a cell lysate
of PMA-stimulated cultured A549 cells using immobilized inhibitor
FF. S: original lysate, FT: flow-through, W: wash fractions, E: elution
with 100 lM competitive inhibitor SF, SDS: final elution step with 1%
sodium dodecyl sulfate. Fractions were analyzed by electrophoresis on 8%
polyacrylamide gels and transferred to PVDF membrane. ADAM-17 was
detected by Western blotting.
Fig. 3 Activity-dependent enrichment of ADAM-17 from a cell lysate of
non-stimulated cultured A549 cells using immobilized inhibitor FF and
control ethanolamine–Sepharose (EA) cartridges. S: original lysate, FT:
flow-through, W: washing fractions, E: elution with 100 lM competitive
inhibitor SF, SDS: final elution with 1% sodium dodecyl sulfate. Fractions
were analyzed by electrophoresis on 8% polyacrylamide gels and trans-
ferred to PVDF membranes. ADAM-17 was detected by Western blotting.
Conclusion
In summary we prepared a library of 96 enantiopure peptide
hydroxamates which were tested with respect to their inhibitory
efficacy towards three metalloproteinases.38 Our results show that
different amino acids at positions Aa1 and Aa2 (see Scheme 1) have
a substantial influence on inhibitory capacity. This is in contrast
to a report by Whittaker et al. stating that amino acids at position
Aa2 do not enter the active site of the enzyme and therefore
have ‘a modest effect’ on potency.7 Several potent inhibitors were
immobilized on Sepharose beads and evaluated in solid-phase
enrichment of active metalloproteinases. Experiments showed
complete enrichment of recombinant MMP-9 and MMP-12 even
with lower-affinity inhibitors. Enrichment of active ADAM-17
was more challenging, but two of the tested inhibitors proved
to be efficient. Indeed, one of the inhibitors, FF, could be
used for enrichment of endogenous ADAM-17 from a complex
biological sample. Our data are complementary to the results
Phorbol esters have been described to enhance ADAM-
mediated shedding of membrane anchored compounds.36 It is
therefore assumed that ADAM-17 becomes activated upon PMA
(phorbol-12-myristate-13-acetate) stimulation. Fig. 4 shows the
effect of PMA on A549 cells with respect to extraction of active
ADAM-17. Interestingly, short-term exposure (30 min) to PMA
results in the appearance of a non-active form of ADAM-17
with an apparent molecular weight corresponding to the 93 kDa
pro-ADAM-17 zymogen (marked with *). Extraction efficiency
decreases, as shown by the recovery of a substantial fraction of
ADAM-17 in the flow-through and washing fractions, although
the major portion still binds the immobilized inhibitor. Decreased
extraction might be explained by mobilization of endogenous
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 1244–1250 | 1247
©

View Article Online
recently published by Cravatt and co-workers on ADAM-17
enrichment from biological samples.39 The strategy employed by
Cravatt and co-workers makes use of a photo-activatable probe
that is biotinylated in a two-step bioorthogonal fashion after
irreversible enzyme modification, allowing streptavidin-mediated
enrichment. Our approach is based on a straightforward affinity-
based enrichment protocol. Of interest is our finding that the
correlation between affinity (i.e. IC50 value) of free inhibitor and
suitability for activity-based enrichment is not as obvious as we
originally expected. We are currently evaluating the suitability of
our library, in combination with solid-phase enrichment, as a tool
to monitor MMP/ADAM activity in patient-derived biological
samples in relation to aberrant MMP/ADAM activity. With
respect to the enrichment protocol we foresee that our current
constructs can possibly be optimized, for instance by variation of
the linker connecting the bait to the matrix.
human ADAM-17 ectodomain was from R & D Systems. The
fluorogenic MMP substrate Mca-PLGL-Dpa-AR-NH2 (where
Mca = (7-methoxycoumarin-4-yl)acetyl and Dpa = N-3-(2,4-
dinitrophenyl)-L-2,3-diaminopropionyl) was from Bachem, the
ADAM substrate Mca-PLAQAV-Dpa-RSSSR-NH2 was from R
& D Systems. N-Hydroxysuccinimide (NHS)-Sepharose was from
Amersham Bioscience. Acetonitrile (gradient grade) was from
Biosolve, ultra-pure water was produced in-house by an Elga
water purifying system and used for all mobile phase and buffer
preparations. Other chemicals were purchased from Sigma.
Preparation of the library
Rink amide resin was rinsed with DCM, MeOH and DMF
(2× each), then deprotected by shaking with 20% piperidine in
DMF for 10 minutes (2×), and rinsed with DMF and DCM
(2× each). Loading of the resin was effected by shaking with
FmocLys(Boc)OH (5 equiv. relative to the stated loading), HCTU
(5 equiv.) and 0.45 M DIPEA (10 equiv.) in NMP for 2 h. The
resin was filtered, then rinsed with DMF and DCM (2× each). Any
non-reacted amines were capped by shaking the resin with Ac2O (5
equiv.) in 0.45 M DIPEA (10 equiv.) in NMP for 5 minutes. After
rinsing the resin with DMF, DCM and Et2O (2× each) and drying
in vacuo, Fmoc determination (UV measurement at 300 nm) gave
a loading of 0.47 mmol g−1.
Twelve portions of 80 lmol (170 mg resin) were rinsed with
DCM and DMF (2× each), deprotected using 20% piperidine
in DMF (10 min) and shaken with preactivated solutions of the
appropriate Fmoc-amino acids Aa1 (400 lmol), HCTU (400 lmol,
165 mg) and DIPEA (800 lmol, 1.8 mL, 0.45 M in NMP) for
1 h. After rinsing with DMF, DCM and Et2O (2× each), and
drying in vacuo, each portion was in turn divided into eight equal
portions. Every portion (∼10 lmol) was respectively reacted with
preactivated solutions of the eight amino acids Aa2 (50 lmol
each), HCTU (50 lmol) and DIPEA (100 lmol as a 0.45 M
solution in NMP) and shaken for 1 h. After rinsing the resins
with DMF and DCM (2× each), all portions were deprotected
using 20% piperidine in DMF for 10 min, then filtered and
rinsed with DMF and DCM (2× each). Finally, to all 96 resins
were added compound 1 (250 lL, 0.2 M in NMP) and DIPEA
(44 lL, 0.45 M in NMP). After shaking for 2 h, the resins were
filtered, rinsed with DMF, MeOH and DCM (2× each) and treated
with 95% TFA–water (0.5 mL) for 1 h (with the exception of
resins containing Arg(Pmc), these were reacted for 2.5 h). The
filtrates, as well as small portions of 95% TFA–water used for
rinsing the resins, were collected in tubes containing chilled Et2O–
petroleum ether (1 : 1, ∼5 mL) and left at −20 ◦C overnight. The
tubes were then centrifuged, and the filtrates were decanted. The
crude products were analysed by LC-MS, then purified by semi-
preparative HPLC. The yields of the pure peptides based on 3
(purity >95% as determined by LC-MS analysis) varied between
3% and 40%
Experimental
General
Tetrahydrofuran was distilled over LiAlH4 before use. Acetonitrile
(MeCN), dichloromethane (DCM), N,N-dimethylformamide
(DMF), N-methyl-2-pyrrolidone (NMP), methanol (MeOH),
piperidine, diisopropylethylamine (DIPEA) and trifluoroacetic
acid (TFA) were of peptide synthesis grade, purchased at
Biosolve, and used as received. All general chemicals (Fluka,
Acros, Merck, Aldrich, Sigma) were used as received. Rink amide
MBHA resin (0.64 mmol g−1) was purchased at Novabiochem,
as well as all appropriately protected amino acids (FmocAlaOH,
FmocArg(Pmc)OH, FmocAsp(tBu)OH, FmocGln(Trt)OH,
FmocGlu(tBu)OH, FmocHis(Trt)OH, FmocLeuOH, Fmo-
cLys(Boc)OH, FmocPheOH, FmocProOH, FmocSer(tBu)OH,
FmocThr(tBu)OH, FmocTrp(Boc)OH, FmocTyr(tBu)OH and
FmocValOH).
O-(1H-6-Chlorobenzotriazolyl)-1,1,3,3-tetra-
methyluronium hexafluorophosphate (HCTU) was purchased
at Iris Biotech (Marktrewitz, Germany). Traces of water were
removed from reagents used in reactions that require anhydrous
conditions by co-evaporation with toluene. Solvents that were
˚
used in reactions were stored over 4A molecular sieves, except
˚
methanol and acetonitrile which were stored over 3A molecular
sieves. Molecular sieves were flame dried before use. Unless noted
otherwise all reactions were performed under an argon or nitrogen
atmosphere. LC-MS analysis was performed on a Jasco HPLC
system with a Phenomenex Gemini 3 lm C18 50 × 4.60 mm
column (detection simultaneously at 214 and 254 nm), coupled to
a PE Sciex API 165 mass spectrometer with ESI. HPLC gradients
were 10–90%, 0–50% or 10–50% acetonitrile in 0.1% TFA–water.
The compounds were purified on a Gilson HPLC system coupled
to a Phenomenex Gemini 5 lm 250 × 10 mm column and
a GX281 fraction collector. The used gradients were either
0–30% or 10–40% acetonitrile in 0.1% TFA–water, depending
on the lipophilicity of the product. Appropriate fractions were
pooled, and concentrated in a Christ rotary vacuum concentrator
overnight at room temperature at 0.1 mbar.
Screening of inhibitor efficacy
Recombinant human MMP-12 catalytic domain and MMP-9
catalytic domain (without fibronectin type II inserts) were a gift
from AstraZeneca R & D (Lund & Moelndal, Sweden) and were
produced in E. coli (Parkar 2000, Shipley 1996). Recombinant
Efficacy of the inhibitors was tested by evaluating their ability
to inhibit proteolytic conversion of a fluorogenic substrate by
recombinant metalloproteinases. A fixed concentration of each
inhibitor (final concentration 100 nM) was incubated with 5 ng
1248 | Org. Biomol. Chem., 2008, 6, 1244–1250
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
of either MMP-9 catalytic domain, MMP-12 catalytic domain
or ADAM-17 ectodomain in assay buffer (for MMP-9 and -12:
50 mM Tris pH 7.4, 0.1 M NaCl, 10 mM CaCl2, 0.05% w/v Brij-
35; for ADAM-17: 25 mM Tris pH 9.0, 2.5 lM ZnCl2, 0.005%
w/v Brij-35) in 96-well plates (Costar White). The appropriate
fluorogenic substrate was added to a final concentration of 2 lM
and proteolysis rates were determined by measuring fluorescence
(kex,em = 320, 440 nm) increase using a Fluostar Optima plate reader
(BMG Labtech) at 28 ◦C. The remaining catalytic activity was
calculated by comparing with proteolysis rates of 5 ng uninhibited
enzyme (see supporting information†) and plotted in heat maps
generated in Matlab.
on each material. Each batch of inhibitor-beads was slurry packed
into Prospekt SPE cartridges fitted on one side with a 0.2 lm
stainless steel frit (Spark Holland, 2 mm ID × 10 mm) in MMP
assay buffer. The packed cartridges were placed in a clamp and
attached to a syringe pump (KD Scientific). Extractions were
performed in the assay buffers described above. Before extraction
the cartridges were conditioned by flushing two times with 250 lL
of the appropriate assay buffer. All steps were performed at
50 lL min−1. Samples containing recombinant enzyme (200 lL
of 0.5 or 5 nM solution) were applied to the cartridges and
the cartridges were washed four times with 200 lL assay buffer
to determine breakthrough. All eluting fractions were collected
separately and immediately placed on ice. Between extractions the
cartridges were regenerated by eluting bound enzyme with 10 mM
ethylenediaminetetraacetic acid, disodium salt (EDTA) in assay
buffer and reconditioning two times with 200 lL assay buffer.
Enzyme activity in the collected fractions was determined by the
activity assay described above and related to the activity in the
original sample.
Determination of IC50 values
The IC50 values of eight selected novel inhibitors were determined
in a competitive enzyme activity assay monitoring conversion
of the same fluorogenic substrates by recombinant metallopro-
teinases in presence of increasing concentrations of inhibitor.
Measurements were performed in 96-well plates (Costar White),
where each well contained 5 ng of either MMP-9 catalytic domain,
MMP-12 catalytic domain or ADAM-17 and a final concentration
of 2 lM of the appropriate substrate in a final volume of 100 lL
assay buffer. Proteolysis rates were determined by measuring
fluorescence increase. Seven-point inhibition curves (0–10 lM)
were plotted in Origin 7.0 (Micronal) and IC50 values were
determined by sigmoidal fitting.
Extraction of active ADAM-17 from a lung carcinoma cell line
Human alveolar carcinoma cell line A549 (ATCC nr. CCL-185)
was grown to 90% confluency in RPMI-1640 with l-glutamine
(Cambrex) supplemented with 10% (v/v) fetal bovine serum
(Cambrex) and 20 lg mL−1 gentamycin (Centafarm) in 25 cm2
culture plates at 37 ◦C, 5% CO2. The cells were serum starved
for 24 h and stimulated with 100 nM phorbol-12-myristate-13-
acetate (PMA, Sigma-Aldrich) for different time periods. After
stimulation, the cells were harvested in 1 mL ice-cold lysis buffer
(25 mM Tris pH 9.0, 2.5 lM ZnCl2, 1% v/v Triton X-100). After
cell lysis (1 h on ice) the lysates were centrifuged in a cooled (4 ◦C)
Eppendorf centrifuge at 20 000 × g for 15 minutes to remove cell
debris. The supernatant was kept on ice until the extraction. Spark
cartridges (2 mm ID × 10 mm) packed with either Sepharose
immobilized inhibitor FF or control ethanolamine–Sepharose
were equilibrated with lysis buffer (2 × 250 lL pumped at
50 lL min−1). The lysates were applied to the cartridge (250 lL,
25 lL min−1) and the flowthrough was collected. The cartridges
were washed once with lysis buffer (250 lL, 50 lL min−1) and
three times with wash buffer (25 mM Tris pH 9.0, 2.5 lM ZnCl2,
0.005% v/v Brij-35). Elution was carried out by elution with a
competitive inhibitor in wash buffer (100 lM inhibitor SF, 3 ×
60 lL, 10 lL min−1) and a non-specific elution step with 1% sodium
dodecyl sulfate (SDS) in wash buffer (100 lL, 20 lL min−1).
60 lL of each sample was diluted with 20 lL non-reducing sample
buffer (Bio-Rad) and, after heating (5 min at 95◦C), separated by
discontinuous SDS-PAGE according to Laemmli40 on 0.75 mm
thick slabs (8% polyacrylamide). Electrophoresis was carried out
in a Mini-Protean III cell (Bio-Rad). The proteins were transferred
to Immun-blot polyvinylidene fluoride (PVDF) membranes (Bio-
Rad) by wet Western blotting in a mini Trans-blot cell (Bio-Rad)
at 350 mA for 60 minutes in 25 mM Tris, 190 mM glycine with
20% v/v methanol. The membranes were blocked overnight in
TBS-T (10 mM Tris pH 7.4, 137 mM NaCl, 0.05% (v/v) Tween-
20) supplemented with 5% w/v non-fat dried milk (Protifar,
Nutricia) at 4 ◦C. Incubation with the primary anti-ADAM17
ectodomain monoclonal antibody (R & D Systems, clone 111636)
was performed at room temperature for 2 h at 0.5 lg mL−1 in
Immobilization of inhibitors on Sepharose beads
The eight selected inhibitors were linked to Sepharose as described
earlier.28 Briefly, NHS-activated Sepharose beads were washed at
4
◦C with several volumes of 1 mM HCl and several volumes
of coupling buffer (0.1 M K2HPO4, pH 7.5). The washed beads
were then incubated with an equal volume of a 5 mM solution
of inhibitor in coupling buffer for 2 h in a shaking incubator
(Eppendorf thermomixer, 1200 rpm, 25 ◦C). After removal of
the supernatant the unreacted NHS-groups were quenched by
incubation with 10 bead volumes of blocking buffer (0.5 M
ethanolamine in coupling buffer). The immobilization was mon-
itored by HPLC analysis of each inhibitor-containing coupling
buffer before and after the procedure. Analysis was performed
on a Merck-Hitachi system fitted with a UV detector at 214 nm.
Samples were separated on a Zorbax Eclipse XDB-C8 column
(4.6 × 150 mm, 5 lm particles, Agilent Technologies). Mobile
phase A was 0.1% v/v TFA in water, mobile phase B was 0.1%
TFA in acetonitrile. Samples were diluted by ten times in mobile
phase A and 10 lL was injected. A linear gradient separation
was performed by increasing the percentage B from 0 to 60% in
30 minutes. The theoretical ligand density for each immobilized
inhibitor was calculated by evaluation of the peak area for the
free inhibitor before and after the immobilization (see supporting
information†).
Determination of extraction yield
The efficacy of the immobilized inhibitors for use as activity-based
affinity ligands in solid-phase extraction (SPE) of MMPs and
ADAMs was determined by performing column-based extractions
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 1244–1250 | 1249
©

View Article Online
TBS-T with 1% w/v non-fat milk, followed by 1.5 h exposure to
a secondary rabbit anti-mouse antibody conjugated with alkaline
phosphatase (Sigma-Aldrich) at a 1 : 15 000 dilution. Bands were
visualized with 5-bromo-4-chloro-3-indolyl phosphate–nitroblue
tetrazolium (Duchefa).
20 K. Jenssen, K. Sewald and N. Sewald, Bioconjugate Chem., 2004, 15,
594.
21 J. Wang, M. Uttamchandani, L. Ping Sun and S. Q. Yao, Chem.
Commun., 2006, 717.
22 S. Reddy, M. S. Kumar and G. R. Reddy, Tetrahedron Lett., 2000, 41,
6285.
23 G. Giacomelli, A. Porcheddu and M. Salaris, Org. Lett., 2003, 5, 2715.
24 D. E. Levy, F. Lapierre, W. Liang, W. Ye, C. W. Lange, X. Li, D.
Grobelny, M. Casabonne, D. Tyrell, K. Holme, A. Nadzan and R. E.
Galardy, J. Med. Chem., 1998, 41, 199.
Acknowledgements
25 C. Y. Ho, E. Strobel, J. Ralbovsky and R. A. Galemmo Jr., J. Org.
Chem., 2005, 70, 4873.
26 J. Wang, M. Uttamchandani, J. Li, M. Hu and S. Q. Yao, Org. Lett.,
2006, 8, 3821.
27 M. A. Leeuwenburgh, P. P. Geurink, T. Klein, H. F. Kauffman, G. A.
van der Marel, R. Bischoff and H. S. Overkleeft, Org. Lett., 2006, 8,
1705.
We thank the Dutch Technology Foundation STW (project
GPC 6150) and The Netherlands Proteomics Centre (NPC) for
financial support of this work. We thank Dr Peter Horvatovich
for assistance in preparing the heat maps (Fig. 2) in Matlab.
28 J. R. Freije, T. Klein, J. A. Ooms, J. P. Franke and R. Bischoff,
J. Proteome Res., 2006, 5, 1186.
29 J. R. Freije and R. Bischoff, J. Chromatogr., A, 2003, 1009, 155.
30 W. Bode, C. Fernandez-Catalan, H. Tschesche, F. Grams, H. Nagase
and K. Maskos, Cell. Mol. Life Sci., 1999, 55, 639.
31 R. Lang, A. Kocourek, M. Braun, H. Tschesche, R. Huber, W. Bode
and K. Maskos, J. Mol. Biol., 2001, 312, 731.
Notes and references
1 D. F. Seals and S. A. Courtneidge, Genes Dev., 2003, 17, 7.
2 R. A. Black and J. White, Curr. Opin. Cell Biol., 1998, 10, 654.
3 M. L. Marcia and J. W. Bartsch, Biochemistry, 2004, 43, 7227.
4 P. J. Huovila, A. J. Turner, M. Pelto-Huikko, I. Ka¨rkka¨inen and R. M.
Ortiz, Trends Biochem. Sci., 2005, 30, 413.
5 R. A. Black, C. T. Rauch, C. J. Kozlosky, J. J. Peschon, J. L. Slack,
M. F. Wolfson, B. J. Castner, K. L. Stocking, P. Reddy, S. Srinivasan,
N. Nelson, N. Boiani, K. A. Schooley, M. Gerhart, R. Davis, J. N.
Fritzner, R. S. Johnson, R. J. Paxton, C. J. March and D. P. A. Cerretti,
Nature, 1997, 385, 729.
6 L. L. Johnson, R. Dyer and D. J. Hupe, Curr. Opin. Chem. Biol., 1998,
2, 466.
32 S. Netzel-Arnett, Q. X. Sang, W. G. I. Moore, M. Navre, H. Birkedal-
Hansen and H. E. van Wart, Biochemistry, 1993, 32, 6427.
33 K. Maskos, C. Fernadez-Catalan, R. Huber, G. P. Bourenkov, H.
Bartunik, G. A. Ellestad, P. Reddy, M. F. Wolfson, C. T. Rauch, B. J.
Castner, R. Davis, H. R. G. Clarke, M. Petersen, J. N. Fitzner, D. P.
Cerritti, C. J. March, R. J. Paxton, R. A. Black and W. Bode, Proc.
Natl. Acad. Sci. U. S. A., 1998, 95, 3408.
7 M. Whittaker, C. D. Floyd, P. Brown and J. H. Gearing, Chem. Rev.,
1999, 99, 2735; S. P. Gupta, Chem. Rev., 2007, 107, 3042.
8 H. F. Kauffman, J. F. Tomee, M. A. van de Riet, A. J. Timmerman and
P. Borger, J. Allergy Clin. Immunol., 2000, 105, 1185.
9 E. Breuer, J. Frant and R. Reich, Expert Opin. Ther. Pat., 2005, 15, 253.
10 M. Hidalgo and S. G. Eckhardt, J. Natl. Cancer Inst., 2001, 93, 178.
11 W. P. Steward and A. L. Thomas, Expert Opin. Invest. Drugs, 2000, 9,
2913.
12 F. Grams, P. Reinemer, J. C. Powers, T. Kleine, M. Pieper, H. Tschesche,
R. Huber and W. Bode, Eur. J. Biochem., 1995, 228, 830.
13 E. W. S. Chan, S. Chattopadhaya, R. C. Panicker, X. Huang and S. Q.
Yao, J. Am. Chem. Soc., 2004, 126, 14435.
34 An interesting application of high-affinity reversible metalloprotease
inhibitors is their use as immobilized ligands for activity-dependent
enzyme enrichment. For examples see references 28 and 29 and: D.
Hesek, M. Toth, S. E. Meroueh, S. Brown, H. Zhoa, W. Sakr, R.
Fridman and S. Mobashery, Chem. Biol., 2006, 13, 379.
35 In order to relate the extraction yields of active metalloproteases to
the measured IC50 values, it is important to know the ligand densities
of the immobilized inhibitors. The determined ligand densities of the
eight selected inhibitors (see Table 1) are shown in the supporting
information†.
36 N. L. Tsakadze, S. D. Sithu, U. Sen, W. R. English, G. Murphy and
S. E. DꢀSouza, J. Biol. Chem., 2006, 281, 3157.
14 K. Ngu and D. V. Patel, J. Org. Chem., 1997, 62, 7088.
15 S. L. Mellor, C. McGuire and W. C. Chan, Tetrahedron Lett., 1997, 38,
3311.
16 R. Sasubilli and W. G. Gutheil, J. Comb. Chem., 2004, 6, 911.
17 S. Gazal, L. R. Masterson and G. Barany, J. Pept. Res., 2005, 66,
324.
37 J. R. Doedens and R. A. Black, J. Biol. Chem., 2000, 275, 14598.
38 For the combinatorial synthesis and application of a similar compound
library containing enantiomeric mixtures see: M. Uttamchandani, J.
Wang, J. Li, M. Hu, H. Sun, K. Y.-T. Chen, K. Liu and S. Q. Yao,
J. Am. Chem. Soc., 2007, 129, 7848.
39 S. A. Sieber, S. Niesen, H. S. Hoover and B. F. Cravatt, Nat. Chem.
Biol., 2006, 2, 274.
40 U. K. Laemmlie, Nature, 1970, 227, 680.
18 Z. Yin, K. S. Low and P. L. Lye, Synth. Commun., 2005, 35, 2945.
19 B. BarlAam, P. Koza and J. Berriot, Tetrahedron, 1999, 55, 7221.
1250 | Org. Biomol. Chem., 2008, 6, 1244–1250
This journal is
The Royal Society of Chemistry 2008
©