Communications
As shown in Figure 1A, incubation of compound 1 with
Finally, the yields of covalent modification of eight MMPs
also possessing a deep S1’ cavity were determined with
compound 1. Experiments were performed by using the same
number of active MMP catalytic units, as determined for each
MMP by careful titration experiments with a highly potent
MMP inhibitor. The results (Figure 3) demonstrate that
compound 1 cross-links with all of the examined MMPs, as
reflected by the radioactivity incorporation, but with a 40-fold
difference in efficacy.
MMP-12, followed by UV irradiation, led to radioactive
labeling of MMP-12. By contrast, no labeling was observed
without the UV irradiation. These results indicate that the
Figure 1. Covalent labeling of MMP-12 by compound 1. A) MMP-12
(50 ng) was incubated with compound 1 (2 mm) for 10 min, either with
or without compound 2 (11 mm), before UV irradiation (hn, 2 min).
The MMP-12 complex was resolved by 1D sodium dodecylsulfate
PAGE electrophoresis, then proteins were transferred onto a polyvinyl-
idene fluoride (PVDF) membrane that was analyzed with a radio-
imager. B) Results of the same experiments but with the protein
visualized on the gel by silver staining (MMP-12, 100 ng).
Figure 3. Covalent modification of various MMPs (1 pmol of active
MMP) by compound 1 (2 mm). The percentage of modification was
calculated by counting the radioactivity incorporated in each MMP,
with the 42% yield determined for MMP-12 as a reference value.
observed radioactive signal reflects only the presence of a
covalent enzyme–inhibitor complex on the PVDF membrane,
after electrophoresis separation. To demonstrate that this
covalent modification only concerns the active site, MMP-12
was first preincubated with compound 2, and this was
followed by addition of compound 1 and UV irradiation.
Under these conditions, no radioactive signal can be detected,
a result indicating that compound 1 only labels the active site
of MMP-12 when it is free of inhibitor. Interestingly, silver
staining of a gel loaded with UV-irradiated MMP-12 in
complex with compound 1 revealed two bands (Figure 1B),
whereas a single band was observed in the absence of
irradiation. Comparison of these results with those obtained
by the radioimagery analysis of the same samples indicates
that the upper band visualized with the silver staining
corresponds to the MMP-12 form covalently modified by
compound 1, while the lower band is the unmodified form of
MMP-12. This interpretation is also supported by competition
experiments. Indeed, when MMP-12 was preincubated with
compound 2 before compound 1 was added and the photo-
irradiation took place, a single band was observed by silver
staining (Figure 1B). As a consequence, the ratio of the band
intensities visualized by silver staining reflects the yield of
MMP-12 cross-linking by compound 1. This approach pro-
vides a cross-linking yield of (42 Æ 5)% for compound 1 with
MMP-12 (average of three replicates).
From the above results, comparable sensitivity in detec-
tion can be expected for MMP-12, -13, -2, and -14 with
compound 1, but depending on their expression levels in
complex proteomes, MMP-3 and MMP-8 might be hardly
detected with the present probe. It is worth mentioning that
cross-linking yields were not determined in previous reports
on the development of ABPP probes to profile MMPs.[7–9]
From the structure of these probes, it can be anticipated that
the MMP residues possibly modified by these probes belong
to surface solvent-exposed subactive sites,[7–9] a situation that
should result in higher variability of the probe cross-linking
yield with different MMPs. To avoid huge variation in cross-
linking yields and possible failure to detect some MMPs, it
seems mandatory to more systematically evaluate the influ-
ence of the photolabile-group type[15,16] and its position within
the probe structure, in order to deliver a single probe able to
modify each MMP with high yield. Whether such an ideal
probe exists remains uncertain; however, an alternative
possibility to resolve this hurdle would be to use a cocktail
of MMP-optimized probes, instead a single one. Given the
size of the MMP S1’ cavity, the development of compound 1
analogues containing benzophenone or diazirine photolabile
groups may lead to a set of ABPP probes that will allow the
sensitive detection of all MMP members.
In conclusion, we have developed a new MMP photo-
affinity probe, which displays high potency towards MMPs
and is able to covalently modify several members of this
family. With a detection limit of 2.5 fmol for MMP-12,
compound 1 is one of the most sensitive probes reported to
date for MMPs. (A 60-fmol detection limit has been reported
for MMP-2.[8]) MMP-12 has been implicated in several human
diseases.[17–19] Thus the ABPP probe reported in this paper
offers a simple and efficient strategy to detect the expression
of MMP-12 active forms in biopsies or fluids relevant for
these pathologies. Other MMPs may also be detected with
The sensitivity of compound 1 for the detection of MMP-
12 was determined by incubation of this compound with serial
dilutions of MMP-12. As shown in Figure 2, as few as 50 pg of
MMP-12 can be detected with compound 1 (2.5 fmol of
MMP-12 at 100 pm concentration).
Figure 2. Sensitivity of compound 1 (2 mm) for the detection of
MMP-12.
3276
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 3275 –3277
David, Arnaud
