Microarray-guided discovery of two-photon (2P) small molecule probes for live-cell imaging of cysteinyl cathepsin activities

Article abstract of DOI:10.1039/c2cc33476c

A microarray immobilized with 105 aldehyde-containing small molecules was screened against mammalian cell lysates over-expressing cathepsin L to identify two potent inhibitors, which were subsequently converted into cell-permeable probes capable of live-cell imaging of endogenous cysteinyl cathepsin activities by two-photon fluorescence microscopy.

Full text of DOI:10.1039/c2cc33476c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 7304–7306
www.rsc.org/chemcomm
COMMUNICATION
Microarray-guided discovery of two-photon (2P) small molecule probes
for live-cell imaging of cysteinyl cathepsin activitiesw
Zhenkun Na,a Lin Li,a Mahesh Uttamchandani*ab and Shao Q. Yao*a
Received 14th May 2012, Accepted 31st May 2012
DOI: 10.1039/c2cc33476c
A microarray immobilized with 105 aldehyde-containing small
molecules was screened against mammalian cell lysates over-
expressing cathepsin L to identify two potent inhibitors, which
were subsequently converted into cell-permeable probes capable
of live-cell imaging of endogenous cysteinyl cathepsin activities
by two-photon fluorescence microscopy.
interrogation with suitable proteins and other biological targets.
The platform offers a cheap and convenient method for rapid
screening of thousands of compounds and has been successfully
used in ligand/inhibitor identification, substrate fingerprinting
and the profiling of cellular events.8 More recently, it has also
been used for high-throughput discovery of activity-based
probes.9 Herein, a microarray immobilized with 105 different
aldehyde-containing small molecules were fabricated, and
screened against lysates of mammalian cells over-expressing
green fluorescent protein (GFP)-labeled cathepsin L (a key
endolysosomal cysteinyl cathepsin which is known to be over-
expressed in many tumor cells and tissues2). Two potent
inhibitors were identified and subsequently converted into
the corresponding cell-permeable, 2P small molecule probes,
which were then used to image endogenous cysteinyl cathepsin
activities in live HepG2 cancer cells (Fig. 1). The general
structure of our small molecule inhibitors is shown in
Fig. 1B, in which an aldehyde warhead (WH) was strategically
chosen because 1) it is a well-known reversible but tight-
binding inhibitor motif of cysteine proteases, 2) could be
readily synthesized using solid-phase chemistry, and 3) later
facilitate the inhibitor-to-probe conversion (e.g. by linking a
quencher molecule via an amide bond; see Fig. 1A). From
previous literatures,3 it is known that cysteinyl cathespins
prefer hydrophobic/aromatic P2 and P3 residues, while varia-
tions at the P1 position are well-tolerated. Consequently in our
small molecule library, a biotin-containing immobilization
handle, Phe/Leu/Val residues and 35 different commercially
available acid building blocks (aliphatic and aromatic) were
installed at P1, P2 and P3 positions, respectively. Both P2
and P3 building blocks were so chosen that the resulting
compounds would possess reasonable cell permeability. We
anticipated that, after SMM-guided identification of ‘‘hits’’,
the biotin moiety in the inhibitor moiety may be conveniently
replaced with a suitable fluorophore without compromising
the probe/cathepsin recognition.
Proteases are critically involved in a multitude of vital biological
processes, ranging from cellular signaling to tissue homeostasis.1
Cathepsins, of which eleven are cysteine proteases, reside mostly
in endolysosomal vesicles and hydrolyze peptides and proteins
destined for these compartments.2,3 The over-expression of these
enzymes has been implicated in a number of pathological
conditions. Consequently, the development of cell-permeable,
small molecule probes capable imaging endogenous protease
activities has gained considerable interest in recent years.4 Of
the various enzyme-detecting, small molecule probes available,
those amenable to two-photon fluorescence microscopy
(TPFM) provide additional key advantages over the usual
one-photon imaging techniques, as they can afford superior
imaging results with increased penetration depth, lower tissue
autofluorescence and reduced photodamage/photobleaching.5
Therefore TPFM is especially amenable for imaging of thick
tissues. In the context of live-cell bioimaging of cysteinyl
cathepsins using small molecule probes, there are several
reported examples.6 Their design principle is typically rational
and involves taking a well-known substrate/inhibitor scaffold of
a cysteinyl cathespin and introducing a suitable FRET (Forster
resonance energy transfer) donor/acceptor pair or fluorogenic
molecule. Two-photon (2P) small molecule probes for live-cell
imaging of endogenous cysteinyl cathespins, however, are not
presently available.5,7 We sought to develop an alternative
approach for high-throughput discovery of 2P small-molecule
imaging probes against cysteinyl cathepsins using small molecule
microarrays (SMMs).
SMMs are miniaturized assemblies of which compounds are
spatially addressed in high-density grids, enabling simultaneous
The solid-phase synthesis of the 105-member aldehyde library
was carried out as shown in Scheme 1.8d Briefly, Fmoc-Lys(Mtt)
aldehyde was loaded onto a theonine-functionalized resin by acid-
catalyzed oxazolidine formation. Subsequently, Mtt was removed
with 1% TFA followed by biotin coupling. Next, solid-phase
combinatorial library synthesis was carried out by using standard
solid-phase peptide synthesis (SPPS) using Fmoc chemistry
protocols. At the end, compounds were cleaved off the resin,
a Department of Chemistry, National University of Singapore,
3 Science Drive 3, 117543, Singapore.
E-mail: chmyaosq@nus.edu.sg; Fax: 65 6770169; Tel: 65 65162925
b Defence Medical and Environmental Research Institute DSO
National Laboratories, 27 Medical Drive, 117510, Singapore.
w Electronic supplementary information (ESI) available: Experimental
details. See DOI: 10.1039/c2cc33476c
c
7304 Chem. Commun., 2012, 48, 7304–7306
This journal is The Royal Society of Chemistry 2012

View Article Online
results, two hits (D02 and D17) were identified, both of which
showed significantly stronger binding toward cathepsin L than
any other immobilized ligands. Both D02 and D17 carried a
Phe residue at the P2 position, and a hydrophobilic aromatic
moiety at the P3 position. To confirm that both compounds
were indeed genuine hits, their IC50 values against recombinant
cathepsin L were measured; with values of 14.5 Æ 0.08 nM
(for D02) and 20.8 Æ 0.22 nM (for D17) under our assay
conditions, both were shown to be potent inhibitors of the
enzyme. Our results also corroborated well with other known
cathepsin inhibitors, most of which also contain hydropholic/
aromatic P2/P3 residues. It should be noted that, although
cathepsin L was used as the representative enzyme in our
SMM screening, both D02/D17 are likely potent inhibitors of
other cysteinyl cathespins.
Subsequently, these two potent small molecule inhibitors
were converted into imaging probes based on our earlier
described design principle (Fig. 1A). As the P1 residue in the
inhibitors did not play a significant role in binding to cathepsin
L, and the aldehyde WH sat where the scissile bond of a
cysteinyl cathepsin substrate would have been located, we
replaced the biotin moiety in D02/D17 with a two-photon
dye, DL-1 (an 4,6-bis(4-hydroxystyryl)pyrimidine derivative10),
and the aldehyde WH with Disperse Red 1 (DR-1; a fluores-
cence quencher) linked to the P1 residue via an amide bond,
giving ZK-1 and ZK-2, respectively (Fig. 2). In the absence of a
cysteinyl cathepsin, the intrinsic fluorescence of these two
imaging probes would be mostly quenched due to the intra-
molecular FRET effect between DL-1 and DR-1. Upon binding to
active enzymes, however, the probes, with the newly introduced
amide bond serving as the scissile bond, would be proteolytically
processed (to remove DR-1) and emit strong fluorescence.
Synthesis of the probes were shown in Fig. 2 (right), with the
key step involving a highly efficient and modular assembly of
the compounds using click chemistry.11 The final probes were
purified to homogeneity and fully characterized (see ESIw).
We next evaluated whether the proteolytic cleavage and
fluorescence increase of ZK-1 and ZK-2 in the presence of
cysteinyl cathepsins could be monitored spectroscopically.
Mammalian lysates from HepG2 cells (a liver cancer cell line
known to over-express endolysosomal cysteinyl cathepsins)
were used as the enzyme source. As shown in Fig. 3A, a time-
Fig. 1 (A) Overall strategy of the SMM-guided, high-throughput
discovery of cell-permeable, 2P small molecular probes for imaging
of endogenous cysteinyl cathepsins. (B) General structure of the
105-member small molecule library. Biotin is located at the P1 position.
P2 consists of Phe/Leu/Val residues. P3 consists of 35 different aromatic/
aliphatic building blocks. Highlighted (in red) are the P2/P3 structures of
the two hits identified from the SMM screening results.
dependent proportional increase in fluorescence (max lem
=
Scheme 1 Reagents and conditions. (a) i: Fmoc-Gly-OH/HBTU/
HOBt/DMF, 4 h; ii: 20% piperidine/DMF, 30 min; iii: Fmoc-
Thr(OtBu)-OH/HBTU/DIEA/DMF, 4 h; (b) i: 20% piperidine/
DMF, 30 min; ii: TFA/DCM (1 : 1) 1 h; iii: 10% DIEA/DCM.
(c) 1% DIEA/MeOH, 2 h, 60 1C. (d) Boc2O, 1% DIEA/DCM, 2 h.
(e) i: 1% TFA/DCM, 30 min; ii: Biotin/HBTU/DIEA/DMF, 4 h.
(f) i: TFA/TIS/EDT (95 : 5 : 0.1); ii: ACN/H2O/TFA (60 : 40 : 0.1),
30 min, 60 1C.
522 nm) was observed, evidently due to successful proteolytic
cleavage of the probes and release of the quencher. Finally, we
examined whether the probes could be used in live-cell imaging
experiments to detect endogenous cysteinyl cathepsin activities
(Fig. 3B); HepG2 cells treated with either ZK-1 or ZK-2 for
2 h showed exclusive fluorescence (green channel) in endoly-
sosomal compartments which were readily stained by Lyso
Trackert (red channel).12 The same cells pre-treated with E-64
(a broad-spectrum cysteinyl cathepsin inhibitor) showed no
detectable fluorescence in the probe channel, indicating these
fluorescence signals were a direct result of endogenous cysteinyl
cathepsin activities.
precipitated, and immobilized directly onto avidin-
functionalized glass slides to generate the corresponding SMM.
We screened the SMM with mammalian cell lysates over-
expressing GFP-fused cathepsin L. Direct application of total
mammalian lysates, rather than recombinantly purified proteins,
saved substantial time and costs by avoiding the tedious and
often problematic protein purification processes. This method
should also be amenable for future high-throughput screening of
other mammalian proteins. As shown in Fig. 2, from the SMM
In conclusion, by using a small molecule microarray as a
high-throughput screening platform, we have successfully
discovered two potent inhibitors of cathepsin L which were
converted strategically into the corresponding two-photon
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 7304–7306 7305

View Article Online
Fig. 2 (Left) SMM image screened against mammalian cell lysates over-expressing GFP-fused cathepsin L. All compounds on SMM were
spotted in duplicate. The green spots were dye references. See spotting pattern and ID in ESI.w (Right) Structures of hits identified and the
corresponding cell-permeable imaging probes (ZK-1 and ZK-2; boxed) and their synthetic scheme. (Red) DL-1. (Blue) DR-1.
2P small molecule probes to directly image cysteinyl cathepsin
activities in thick tissues.
Funding support was provided by the Ministry of Education
(R-143-000-394-112) and the Agency for Science, Technology
and Research (A*Star) (R143-000-391-305).
Notes and references
1 B. Turk, Nat. Rev. Drug Discovery, 2006, 5, 785–799.
2 M. M. Mohamed and B. F. Sloane, Nat. Rev. Cancer, 2006, 6, 764–775.
3 H. H. Otto and T. Schirmeister, Chem. Rev., 1997, 97, 133–171.
4 L. E. Edgington, M. Verdoes and M. Bogyo, Curr. Opin. Chem.
Biol., 2012, 15, 798–805.
5 (a) H. M. Kim, B. R. Kim, J. Hong and B. R. Cho, Angew. Chem.,
Int. Ed., 2007, 46, 7445–7448; (b) X. Wang, D. Nguyen,
C. O. Yanez, L. Rodriguez, H.-Y. Ahn, M. V. Bondar and
K. D. Belfield, J. Am. Chem. Soc., 2010, 132, 12237–12239.
6 (a) F. Yuan, S. H. Verhelst, G. Blum, L. M. Coussens and M. Bogyo,
J. Am. Chem. Soc., 2006, 128, 5616–5617; (b) G. Blum, G. V. Degenfeld,
M. J. Merchant, H. M. Blau and M. Bogyo, Nat. Chem. Biol., 2007, 3,
668–677; (c) A. Watzke, G. Kosec, M. Kindermann, B. Turk and
K. U. Wendt, Angew. Chem., Int. Ed., 2008, 47, 406–409.
7 M. Hu, L. Li, H. Wu, Y. Su, M. Uttamchandani, Q.-H. Xu and
S. Q. Yao, J. Am. Chem. Soc., 2011, 133, 12009–12020.
8 For a recent review, see: (a) H. Wu, J. Ge, M. Uttamchandani and
S. Q. Yao, Chem. Commun., 2011, 47, 5664–5670; (b) M.
Uttamchandani, W. L. Lee, J. Wang and S. Q. Yao, J. Am. Chem.
Soc., 2007, 129, 13110–13117; (c) C. H. S. Lu, H. Sun, F. B. A. Bakar,
M. Uttamchandani, W. Zhou, Y.-C. Liou and S. Q. Yao, Angew.
Chem., Int. Ed., 2008, 47, 7438–7441; (d) H. Wu, J. Ge and S. Q. Yao,
Angew. Chem., Int. Ed., 2010, 49, 6528–6532; (e) H. Wu, J. Ge, P. Yang,
M. Uttamchandani and S. Q. Yao, J. Am. Chem. Soc., 2011, 133,
1946–1954.
9 (a) H. Shi, K. Liu, A. Xu and S. Q. Yao, Chem. Commun.,
2009, 5030–5032; (b) H. Shi, M. Uttamchandani and S. Q. Yao,
Chem.–Asian. J., 2011, 6, 2803–2815.
10 L. Li, Y. P. Tian, J. X. Yang, P. P. Sun, L. Kong, J. Y. Wu,
H. P. Zhou, S. Y. Zhang, B. K. Jin, X. T. Tao and M. H. Jiang,
Chem. Commun., 2010, 46, 1673–1675.
11 (a) H. C. Kolb and K. B. Sharpless, Drug Discovery Today, 2003, 8,
1128–1137; (b) M. M. Meldal and C. W. Tornøe, Chem. Rev., 2008,
108, 2952–3015; (c) K. A. Kalesh, H. Shi, J. Ge and S. Q. Yao,
Org. Biomol. Chem., 2010, 8, 1749–1762.
12 Fluorescent signals from ZK-1/-2 were not expected to completely
overlap with the tracker signals, as the former represented only
endolysosomal cysteinyl cathepsin activities which might not be
distributed throughout the entire organelles.
Fig. 3 (A) Time-dependent emission spectra (lem = 522 nm) of
ZK-1/-2 in HepG2 lysates (Hepes buffer, pH 5.5) at room temperature.
(B) Two-photon confocal images of live HepG2 cells upon treatment
with ZK-1/-2 (2 mM) for 2 h. Probe channel (pseudo-colored in green):
lem = 550 nm. Tracker channel (pseudo-colored in red): lem = 590 nm
channel. Endolysosomal tracker (LysoTrackers Red DND-99; Invitrogen)
was used. In panels where a general cysteine protease inhibitor (E64)
was used, HepG2 cells were pre-incubated with 15 mM of E64 before
probe treatment and imaging. All images were acquired under the
same settings.
small molecule imaging probes, both of which were found to
be cell-permeable and could detect endogenous cysteinyl
cathepsin activies from cell lysates or live mammalian cells
of HepG2 cancer cells. Future efforts will focus on using these
c
7306 Chem. Commun., 2012, 48, 7304–7306
This journal is The Royal Society of Chemistry 2012