Activity-based fluorescent probes that target phosphatases

Article abstract of DOI:10.1016/S0040-4039(03)00348-4

We have successfully designed and synthesized two fluorescently-labeled, activity-based probes, Probe 1 and Probe 2, which were shown to label protein tyrosine phosphatases specifically, as well as other types of phosphatases. The probes were not reactive towards the other non-phosphatase enzymes tested. These probes may find potential applications in large-scale proteomic experiments whereby subclasses of proteins may be selectively identified.

Full text of DOI:10.1016/S0040-4039(03)00348-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 2669–2672
Activity-based fluorescent probes that target phosphatases
Qing Zhu,a Xuan Huang,a Grace Y. J. Chena,b and Shao Q. Yaoa,b,
*
aDepartment of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore 117543, Singapore
bDepartment of Biological Sciences, National University of Singapore, 3 Science Drive 3, Singapore 117543, Singapore
Received 23 November 2002; revised 31 January 2003; accepted 7 February 2003
Abstract—We have successfully designed and synthesized two fluorescently-labeled, activity-based probes, Probe 1 and Probe 2,
which were shown to label protein tyrosine phosphatases specifically, as well as other types of phosphatases. The probes were not
reactive towards the other non-phosphatase enzymes tested. These probes may find potential applications in large-scale proteomic
experiments whereby subclasses of proteins may be selectively identified. © 2003 Elsevier Science Ltd. All rights reserved.
The human genome project has recently been com-
pleted.1a It opens up new and exciting possibilities in
the understanding of the working of many different
human diseases at the cellular and molecular levels.
However, genes alone do not tell the whole story of
cellular functions. Proteins are ultimately responsible
for most processes that take place within the cell.
Therefore, emphasis has now shifted from research of
the genome to that of the proteome, which aims to
understand the function of every protein in an
organism.1b
In order to accelerate the functional analysis of a cell’s
complete protein repertoire, new and rapid techniques
capable of genome-wide studies of proteins are essen-
tial. High-throughput screening methods, in most cases
aided by automation, have enabled scientists to analyze
proteins on a global scale quickly and efficiently.2
Recently, the activity-based profiling of proteins has
proven to be a powerful tool in proteomic studies,
whereby subclasses of enzymatic proteins could be
selectively identified.3 This strategy takes advantage of
mechanism-based probes that react with different
classes of enzymes, leading to the formation of covalent
probe–protein complexes which are readily distin-
guished from other non-reactive proteins in a crude
proteome mixture. Thus far, a number of research
groups have successfully designed and tested a handful
of probes against different classes of enzymes.3 Among
them, Cravatt et al. developed fluorophosphonate/
fluorophosphate derivatives conjugated to either a
biotin moiety or a fluorescent dye, and used them to
label serine hydrolases selectively in a crude protein
mixture.3a The same group developed sulfonate ester-
based probes that target proteins having nucleophilic
active-site residues.3b Bogyo et al. synthesized vinyl-sul-
fone containing peptides and used them to label
eukaryotic proteosomes.3c Lo et al. generated p-hydroxy-
mandelic acid-containing probes that specifically label
protein tyrosine phosphatases in an activity-based fash-
ion.3d We recently synthesized an aspartic acid analog
that contains a fluoromethylketone (fmk) moiety. By
conjugation of this molecule to a fluorescent dye, Cy3,
via a simple alkyl linker, the resulting probe was able to
label selectively caspases over other enzymes.3e
Figure 1. Structures of known mechanism-based inhibitors of
protein phosphatases, and the two probes synthesized in our
studies.
* Corresponding author. Tel.: +65-68741683; fax: +65-67791691;
e-mail: chmyaosq@nus.edu.sg
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)00348-4

2670
Q. Zhu et al. / Tetrahedron Letters 44 (2003) 2669–2672
In order to extend the potential of this activity-based
approach into studies of other proteins, we would like
to develop new activity-based probes capable of detect-
ing other major classes of enzymes, including kinases,
phosphatases, and other types of proteases such as
metalloproteases and aspartic proteases. We reason
that, in order for these probes to be readily applicable
in large-scale, high-throughput proteomic experiments,
they should preferably be conjugated to a fluorescent
dye that not only possesses good fluorescence proper-
ties, but also is compatible with most commercially
available fluorescence-based gel scanners. Therefore,
the cyanine dye, Cy3, is an ideal choice. Herein, we
report two Cy3-conjugated, activity-based probes,
which are designed to potentially target not only
protein tyrosine phosphatases, but also other types of
phosphatases. Ultimately, we would like to generate a
whole array of probes containing unique reactivity
towards different classes of enzymes. These probes,
combined with standard gel-based proteomic experi-
ments, may serve as a high-throughput tool to study the
human proteome.
conjugated probes containing these moieties and used
them to probe PTPs in a proteomic experiment.3d How-
ever, their probes were not conjugated to dyes such as
Cy3, making them unfeasible for sensitive and quantita-
tive detection of labeled proteins. Furthermore, no
experiments were done to assess whether these probes
could be used to detect other phosphatases, such as
acid/alkaline phosphatases. Therefore, we designed
Probe 1 by conjugation of the p-hydroxymandelic acid
derivative, shown in Figure 1, to Cy3 and tested its
feasibility in detecting different types of phosphatases
in a standard activity-based proteomic experiment. We
also designed Probe 2, by conjugation of Cy3 to a
different reactive unit, 2-difluoromethylphenyl phos-
phate. 2-Difluoromethylphenyl phosphate was shown
previously to be a general mechanism-based phos-
phatase inhibitor against a broad spectrum of different
phosphatases, including acid and alkaline phos-
phatases.5 However, activity-based probes based on this
moiety have not been realized thus far to assess their
feasibility in detecting phosphatases in a proteomic
experiment.
The probes were designed based on principles previ-
ously described (Fig. 1).3e Briefly, the structure of each
probe consisted of three units: a Cy3-containing
fluorescence unit, a linker and a reactive unit. The
fluorescence unit serves as a sensitive means to detect
proteins upon labeling with the probe. The reactive unit
is made of a mechanism-based group that would cova-
lently react with an enzyme in an activity-dependent
fashion, and could be fine tuned to accommodate dif-
ferent enzymes. In this particular study, we were inter-
ested in developing probes that specifically target
phosphatases. It was shown previously that p-hydroxy-
mandelic acid derivatives were good mechanism-based
inhibitors of protein tyrosine phosphatases (PTPs).4
More recently, Lo et al. synthesized dansyl- and biotin-
First, Probe 1 was synthesized, as shown in Scheme 1,
by the modification of procedures reported by Lo et al.
in order to accommodate the introduction of the Cy3
dye.3d Briefly, Boc-protected diamine 2 was coupled
with 4-hydroxylmandelic acid 3, which was obtained
from phenol, to form compound 4 in 80% yield. Phos-
phorylation of 4 was achieved by treatment with diethyl
phosphochloridate to obtain 5 in 80% yield. The
hydroxyl group of 5 was converted to fluoride, resulting
in 6 in 80% yield using DAST. Following deprotection
of the Boc group, compound 7 was directly coupled
with the NHS ester of Cy3 to obtain 8 in 40% yield.
The final product, Probe 1, was obtained by hydrolysis
of 8. Overall, the probe was obtained in seven steps in
a yield of ꢀ4.4%.6
Scheme 1.

Q. Zhu et al. / Tetrahedron Letters 44 (2003) 2669–2672
2671
Probe 2 was synthesized from the intermediate 12,
different classes of proteases and lipases, were not
labeled by either probe, indicating the high specificity of
both probes towards not only protein tyrosine phos-
phatases (PTPs), but also possibly other types of phos-
phatases (e.g. alkaline phosphatases). More experiments
are underway to test the probes against other classes of
phosphatases. The activity-dependent nature of the
probes was confirmed by first denaturing the protein
samples with heat, followed by treatment with the
probes; no labeling of the phosphatase was observed,
indicating the enzymatic activity of the phosphatase
was a prerequisite for the reaction to occur. Under the
same labeling conditions, both probes selectively
labeled phosphatases in a similar fashion, with Probe 1
consistently giving stronger bands on all proteins
tested, indicating that the mechanism-based reactive
unit (e.g. p-hydroxymandelic acid) in this probe may be
intrinsically more reactive towards phosphatases.
which was prepared from commercially available start-
ing material
9
based on procedures previously
reported.5 Briefly, 10 was obtained by phosphorylation
of 9 with diethyl phosphochloridate in 95% yield.
Treatment of 10 with DAST afforded 11 in 80% yield,
which was then converted to 12 under reductive condi-
tions in 70% yield. Treatment of 12 with succinic anhy-
dride afforded 13 in moderate yield (50%). Compound
14 was obtained by conjugation of 13 with Cy3-con-
taining 15, in the presence of HOBt and TBTU to form
14, followed by hydrolysis to generate Probe 2 (Scheme
2).7
Next, the probes were tested for selective labeling of
proteins based on their enzymatic activity. We first
tested their reactivity towards protein tyrosine phos-
phatases, as Probe 1 was designed based on a probe
previously shown to label PTPs selectively.3d A known
protein tyrosine phosphatase from yeast, YBR267W,
was used in our experiments.8 The glutathione-S-trans-
ferase (GST) fusion of the tyrosine phosphatase was
obtained by conventional protein expression protocols,8
and labeled with both Probe 1 and 2 (Fig. 2).9 It was
found that 1 h incubation was sufficient for the two
probes to label the protein with high efficiency. The
labeled proteins were readily separated on a SDS-
PAGE gel, and visualized using a commercial fluores-
cence gel scanner.
In conclusion, we have successfully designed, synthe-
sized and tested two fluorescent small molecule probes
capable of labeling phosphatases in a highly specific,
activity-based fashion. It was found that both probes
selectively labeled not only protein tyrosine phos-
phatases (PTPs), but also other types of phosphatases,
such as alkaline phosphatases. Further experiments are
underway which involve testing these two probes
against a variety of other types of phosphatases, as well
as their feasibility in in vivo experiments to profile
phosphatase activity in live cells and tissues.
We subsequently tested the probes against other types
of phosphatases, as well as other non-phosphatase
proteins. Many known alkaline phosphatases are
known to be involved in a number of critical cellular
processes, and have a broad spectrum of substrate
specificity, thus making them ideal candidates to test
the specificity of our probes. A panel of 10 commer-
cially available proteins was used in our studies; three
of them were alkaline phosphatases isolated from dif-
ferent sources. The remainder were non-phosphatase
enzymes, including proteases and lipases. Upon incuba-
tion with each probe for 1 h (Fig. 3a and b for Probes
1 and 2, respectively), the proteins were separated by
SDS-PAGE, followed by detection of the labeling reac-
tion with a fluorescence-based gel scanner.9 It was
found that both probes selectively labeled ONLY phos-
phatases (Lanes 1–3, Fig. 3). Other proteins, including
Figure 2. Protein tyrosine phosphatase (PTP) from yeast
(YBR267W) labeled by the probes. Lane 1: unlabeled protein
detected with anti-GST by Western Blotting. Lane 2: protein
labeled with Probe 1; Lane 3: protein labeled with Probe 2
followed by detection on a fluorescence gel scanner.9
Scheme 2.

2672
Q. Zhu et al. / Tetrahedron Letters 44 (2003) 2669–2672
Figure 3. Other proteins labeled with (a) Probe 1 and (b) Probe 2. Lane 1: calf alkaline phosphatase (Promega, M182A,); Lane
2: shrimp alkaline phosphatase (Promega, M8201); Lane 3: alkaline phosphatase (Sigma, P-7640); Lane 4: chymopapain (Sigma,
C-8526); Lane 5: papain (Sigma, P-4879); Lane 6: a-chymotrypsin (Sigma, C-4129); Lane 7: proteinase K (Sigma, P-6556); Lane
8: proteinase (Sigma, P-5380); Lane 9: lipase (Sigma, L-1754); Lane 10: lipase (Sigma, L-3001). Proteins were detected by a
fluorescence gel scanner.9
Acknowledgements
15/5 to 15/85/5 gradients) to afford Probe 1 (3.3 mg; 30%
yield). NMR (CDCl3, TMS): H, l 1.76 (s, 12H), 1.94 (m,
1
4H), 2.50 (br, 2H), 3.59 (m, 11H), 4.15 (m, 6H), 5.80 (d,
J=48 Hz, 1H), 7.31 (m, 14H), 8.43 (m, 1H). 19F NMR
(282.2 MHz, CDCl3, 25°C) l −94.89 (d, J=48.6 Hz) ppm;
ESI-MS: m/z 805.4 [M−I]+.
Funding support from the National University of Sin-
gapore (NUS) and the Agency for Science, Technology
and Research (A*STAR) of Singapore are acknowl-
edged. S.Q.Y. is the recipient of the 2002 Young Inves-
tigator Award (YIA) from the Biomedical Research
Council (BMRC).
7. To a cooled solution of 14 (12 mg, 0.012 mmol) in 1 mL
of CH2Cl2 was added TMSI (6 mL, 0.036 mmol). The
reaction was allowed to warm to room temperature and
stirred further for 1 h. The reaction was stopped by
quenching with 10% triethylamine in H2O (2 mL) and
extracted with CHCl3 (3×2 mL). The aqueous layer was
separated and concentrated under reduced pressure, fol-
lowed by purification by flash chromatography (CH2Cl2/
MeOH/TEA; 85/15/5 to 15/85/5 gradients) to afford Probe
References
1. (a) Venter, J. C.; et al. Science 2001, 291, 1304–1351; (b)
Kodadek, T. Chem. Biol. 2001, 8, 105–115.
1
2 (3.4 mg; 35% yield). NMR (CD3OD, TMS): H, l 1.74
2. (a) MacBeath, G.; Schreiber, S. L. Science 2000, 289,
1760–1763; (b) Zhu, H.; Bilgin, M.; Bangham, R.; Hall,
D.; Casamayor, A.; Bertone, P.; Lan, N.; Jansen, R.;
Dinlingmaier, S.; Houfek, T.; Mitchell, T.; Miller, P.;
Dean, R. A.; Gerstein, M.; Snyder, M. Science 2001, 293,
2101–2105; (c) Lesaicherre, M. L.; Lue, R. Y. P.; Chen, G.
Y. J.; Zhu, Q.; Yao, S. Q. J. Am. Chem. Soc. 2002, 124,
8768–8769.
3. (a) Cravatt, B. F.; Sorensen, E. J. Curr. Opin. Chem. Biol.
2000, 4, 663–668; (b) Adam, G. C.; Sorensen, E. J.;
Cravatt, B. F. Nat. Biotechnol. 2002, 20, 805–809; (c)
Greenbaum, D.; Medzihradszky, K. F.; Burlingame, A.;
Bogyo, M. Chem. Biol. 2000, 7, 569–581; (d) Lo, L. C.;
Pang, T. L.; Kuo, C. H.; Chiang, Y. L.; Wang, H. Y.; Lin,
J. J. J. Proteome Res. 2002, 1, 35–40; (e) Liau, M. L.;
Panicker, R. C.; Yao, S. Q. Tetrahedron Lett. 2003, 44,
1043–1046.
(s, 12H), 2.19 (m, 4H), 2.50 (m, 4H), 2.67 (br, 2H), 2.86
(br, 2H), 3.66 (s, 3H), 4.21 (m, 2H), 7.11 (t, J=54 Hz,
1H), 7.30 (m, 2H), 7.45 (m, 4H), 7.62 (m, 1H), 7.75 (m,
1H), 7.85 (m, 1H), 8.50 (m, 1H). 19F NMR (282.2 MHz,
CD3OD, 25°C) l −39.16 (d, J=56.8 Hz) ppm; ESI-MS:
m/z 806.3 [M−I]+.
8. A plasmid containing the GST fusion of the protein
tyrosine phosphatase from yeast, YBR267W, was pur-
chased from Invitrogen (San Diego, USA), in the form of
a Yeast ExClone™. Yeast cells transformed with the
above plasmid were grown in SD-URA-LEU media to
A600=0.8, as recommended by the supplier, followed by
supplementation with 0.5 mM copper sulphate for 2 h to
induce expression of the GST-ORF. Cells were harvested,
washed with cold water, then lysed by ultrasonication. The
resulting crude protein extract was purified with
a
MicroSpin™ GST column (Amersham, USA) to generate
the desired pure GST-fusion protein.
4. Myers, J. K.; Widlanski, T. S. Science 1993, 262, 1451–
1454.
9. SDS-PAGE experiments: 20 ml of the protein (50–500
mg/mL) dissolved in 5 mM Tris (pH 8) was incubated at
room temperature with 0.2 mL of each probe (200 mM
stocks in DMSO). After 1 h, the reaction was stopped by
addition of 6×SDS loading dye, and then heated at 90°C
for 5 min. The samples were run on 12% SDS-PAGE gels,
then scanned using the Typhoon™ 9200 scanner (Amer-
sham, USA) at uex=532 nm. Western blots were per-
formed using anti-GST monoclonal antibody with the
ECL™ kit (Amersham, USA) following the protocol pro-
vided by the vendor.
5. Betley, J. R.; Sandro, C. T.; Mekhalfia, A.; Rickard, J. H.;
Denham, H.; Partridge, L. J.; Pluckthun, A.; Blackburn,
G. M. Angew. Chem., Int. Ed. 2002, 42, 775–777.
6. To a cooled solution of 8 (12 mg, 0.012 mmol) in 1 mL of
CH2Cl2 was added TMSI (6 mL, 0.036 mmol). The reac-
tion was allowed to warm to room temperature and stirred
further for 1 h. The reaction was stopped by quenching
with 10% triethylamine in H2O (2 mL) and extracted with
CHCl3 (3×2 mL). The aqueous layer was separated and
concentrated under reduced pressure, followed by purifica-
tion by flash chromatography (CH2Cl2/MeOH/TEA; 85/