Diced electrophoresis gel assay for screening enzymes with specified activities

Article abstract of DOI:10.1021/ja401792d

We have established the diced electrophoresis gel (DEG) assay as a proteome-wide screening tool to identify enzymes with activities of interest using turnover-based fluorescent substrates. The method utilizes the combination of native polyacrylamide gel electrophoresis (PAGE) with a multiwell-plate- based fluorometric assay to find protein spots with the specified activity. By developing fluorescent substrates that mimic the structure of neutrophil chemoattractants, we could identify enzymes involved in metabolic inactivation of the chemoattractants.

Full text of DOI:10.1021/ja401792d

Communication
pubs.acs.org/JACS
Diced Electrophoresis Gel Assay for Screening Enzymes with
Specified Activities
Toru Komatsu,† Kenjiro Hanaoka,† Alexander Adibekian,‡ Kentaro Yoshioka,† Takuya Terai,†
Tasuku Ueno,† Mitsuyasu Kawaguchi,† Benjamin F. Cravatt,‡ and Tetsuo Nagano*,†
†Graduate School of Pharmaceutical Sciences, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
‡Department of Chemical Physiology, The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, California 92037,
United States
S
* Supporting Information
limit of β-galactosidase activity with the colorimetric substrate
X-gal5 is only 100 ng. Greater sensitivity is important, since
ABSTRACT: We have established the diced electro-
phoresis gel (DEG) assay as a proteome-wide screening
tool to identify enzymes with activities of interest using
turnover-based fluorescent substrates. The method utilizes
the combination of native polyacrylamide gel electro-
phoresis (PAGE) with a multiwell-plate-based fluorometric
assay to find protein spots with the specified activity. By
developing fluorescent substrates that mimic the structure
of neutrophil chemoattractants, we could identify enzymes
involved in metabolic inactivation of the chemoattractants.
lower protein loading on the gel is necessary for the good
separation of proteins. Fluorescent substrates would increase
the sensitivity,7 but in the conventional system, the fluorescent
product readily diffuses in the gel, making it difficult to
determine the precise location of the target protein (Figure
1c).8
Here we describe a simple but effective way to overcome the
problem of diffusion by dicing the electrophoresis gel into small
pieces that are separately loaded into wells of multiwell plates.
The fluorometric assay can then easily identify wells containing
active proteins (Figure 1d). Although the idea is simple, it
appears to be new, and we have coined the term diced
electrophoresis gel (DEG) assay. The method not only enables
fluorescent substrates to be used in zymography to achieve
much higher sensitivity but also allows the use of various
detection platforms, including LC−MS-based analysis, as
shown below. We first tested the feasibility of the DEG assay
by detecting β-galactosidase on the gel after blue native PAGE6
with the fluorescent substrate TG β-Gal7 (Figure 1e,f). As
expected, in-well fluorescence assays provided increased signal-
to-noise ratios after longer incubation, giving a detectable signal
with only 0.03 ng (∼0.2 fmol) of the enzyme, which was hardly
detectable with conventional methods (Figure 1b,c). The
usefulness of the method was confirmed by detection of two
glycosidases (β-glucosidase and β-galactosidase), two esterases
(liver esterase and alkaline phosphatase), and two peptidases
(leucine aminopeptidase and elastase) with appropriate
fluorescent substrates. The detection limits ranged from 0.1
to 10 ng [Table 1 and Figures S1−S4 in the Supporting
Information (SI)]. We also confirmed that multiple electro-
phoretic platforms can be used in this assay (Figure S5). The
high sensitivity of the method is especially valuable for analysis
of proteins in cell lysates. In blue native PAGE of cell lysates,
tailing was minimal because of the low protein loading (Figure
S6). These proof-of-concept studies confirmed the reliability of
our method for functional proteomic analysis. Next, to examine
the practical utility of the method, we used it to identify
enzymes involved in metabolic inactivation of N-formylated
chemotactic peptides.
ince many proteins remain uncharacterized in terms of
S
their functions in health and disease states,1 annotation of
protein function is an important task in current biological and
medicinal study. Even if we limit our consideration to enzymes,
it remains the case that connecting enzymatic functions to
changes in metabolites, in other words, annotation of enzymes
responsible for certain enzymatic reactions,2 can yield huge
progress in understanding the roles of proteins in living
systems. For example, the discovery of the physiological
substrate of dipeptidyl peptidase IV (DPPIV) made the enzyme
a promising target for treatment of diabetes.3 However, despite
the great potential value of such studies, there has been little
progress in methods to annotate proteins on the basis of their
biochemical activities. A conventional approach for this purpose
would be column-chromatography-based proteome separation
coupled with enzymatic assays;4 however, this is slow and
requires large amounts of sample, and clear-cut separation is
often difficult. Since plural enzymes may exhibit a particular
activity in physiological or pathological states, it is important to
identify them in a comprehensive manner to select potential
targets for drug development.1−3 Therefore, we need a method
for sensitive, reliable, rapid, and comprehensive detection of
proteins with specified activities.
One platform for this purpose is zymography,5 in which the
proteome is separated by nondenatured polyacrylamide gel
electrophoresis (PAGE)6 and colorimetric or fluorometric
assays are performed on the gel to visualize protein spots
with desired activities (Figure 1a). The method requires only
small amounts of sample and gives sharp separation, and two-
dimensional electrophoresis can give a comprehensive “activity
map”. However, the sensitivity is low, especially with
colorimetric substrates (Figure 1b). For example, the detection
Received: February 24, 2013
Published: April 12, 2013
© 2013 American Chemical Society
6002
dx.doi.org/10.1021/ja401792d | J. Am. Chem. Soc. 2013, 135, 6002−6005

Journal of the American Chemical Society
Communication
Figure 1. (a) Schematic illustration of the zymography method. (b) Detection of β-galactosidase with the colorimetric substrate X-gal after blue
native PAGE. The black arrow indicates the smallest detectable colorimetric spot. (c) Detection of β-galactosidase with the fluorescent substrate TG
β-gal after blue native PAGE. The white arrow indicates the smallest detectable fluorescence spot. (d) Schematic illustration of the diced
electrophoresis gel (DEG) assay. (e) Detection of β-galactosidase by one-dimensional DEG assay with TG β-gal. (f) Time-dependent increase in the
signal-to-noise ratio for the spot containing 0.1 ng of β-galactosidase in (e).
performed using LC−MS-based reaction analysis as an output,
and it was confirmed that high hydrolytic activity of fMLF was
present in the same well (Figure 2b). Therefore, a single
enzyme is responsible for the cleavage of formylmethionyl
amide in fMLF.
Table 1. Enzymes and Fluorescent Probes Tested to
Determine the Detection Limit of the DEG Assay
detection
limit (ng)
enzyme
probe
TG β-gal
Umb β-Glc
glycosidases β-galactosidase
β-glucosidase
0.03
1
Next, to identify this enzyme, we performed the DEG assay
using two-dimensional gel electrophoresis with isoelectric
focusing (IEF) and blue native PAGE (Figure 2c,d). The
candidate protein was electrophoresed to a single well, and
peptide mass fingerprinting (PMF) analysis indicated that the
candidate was acylamino acid-releasing enzyme (APEH)
(Figure S11).4,15,16 The involvement of this enzyme has been
suggested previously,4 so the present result is a rediscovery of
the activity of this enzyme with our method. However, our
method provides a comprehensive overview of the activity and
indicates that APEH is the predominant contributor to fMLF
metabolism as well as the primary target of the synthesized
fluorescent substrates. We confirmed this by means of
biochemical experiments with enzymes and inhibitors (ebe-
lactone A17 and AA74-115). First, overexpression of mouse
APEH in mammalian cells resulted in an increase of fMet-AMC
cleaving activity in cell lysate (Figure 3a). The enzyme was
electrophoresed to the same spot as in the case of liver lysate
(Figures S12 and S13). Furthermore, treatment of mouse liver
lysate with AA74-1, a selective inhibitor of APEH, almost
completely blocked the fluorescence increase of fMet-AMC
(Figure 3b).
We then analyzed the reactivity of APEH toward the original
substrate, fMLF, and confirmed that fMLF also acts as a
substrate of the enzyme (Figure S14). In mouse liver lysate,
APEH inhibition led to a >50% decrease in fMLF-cleaving
activity (Figure 3c). The remaining activity generated not fMet
but rather formylmethionyl-leucine (fML) (Figure S15). This
result indicated the presence of a second pathway for fMLF
cleavage that generates fML. To identify the enzyme(s)
involved in this pathway, we tried to characterize this activity
with fMet-Leu-AMC in the absence of LAP (Scheme S2). In a
one-dimensional DEG assay, we found a well that exhibited a
fluorescence increase of fMet-Leu-AMC independently of LAP
activity (Figure S10), and PMF analysis identified this target as
esterases
alkaline phosphatase Umb Phos
0.3
10
3
liver esterase
elastase
FDBu
peptidases
Meo-Suc-Ala-Ala-
Pro-Ile-AMC
leucine
Leu-AMC
0.3
aminopeptidase
(LAP)
N-Terminal formylation of proteins is a unique protein
modification observed in prokaryotes and mitochondria.9 N-
Formylated peptides, such as formylmethionyl-leucyl-phenyl-
alanine (fMLF) from Escherichia coli,10 are markers of infection
or tissue damage, and cells exhibit immune responses via
activation of formyl peptide receptors.11 The receptors and
downstream signaling have been studied intensively, but the
mechanism of signal termination is poorly understood. Several
enzymes have been proposed as candidates for metabolic
inactivation of fMLF,4,12,13 but the major player has not been
identified. To address this question, we developed fluorescent
substrates mimicking the structure of fMLF. fMLF is cleaved in
mouse liver lysate mainly at the C-terminus of formylmethio-
nine (fMet) (Figure S7), and we designed two substrates based
on the 7-amino-4-methylcoumarin (AMC) scaffold,14 fMet-
AMC and fMet-Leu-AMC, to report this activity (Figure 2a and
Schemes S1 and S2 in the SI). fMet-AMC was designed for
direct observation of the cleavage of formylmethionyl amide to
release fluorescent AMC. fMet-Leu-AMC more closely
resembles fMLF and was designed to detect formylmethionyl
amide cleavage by a coupled assay with leucine aminopeptidase
(LAP). Incubation of both probes with mouse tissue lysates
resulted in a rapid fluorescence increase (Figure S8). A one-
dimensional DEG assay revealed a single well with high fMet-
AMC-metabolizing activity (Figure S9), and this well exhibited
a fluorescence increase of fMet-Leu-AMC only when LAP was
added to the well (Figure S10). The DEG assay was then
6003
dx.doi.org/10.1021/ja401792d | J. Am. Chem. Soc. 2013, 135, 6002−6005

Journal of the American Chemical Society
Communication
Figure 2. (a) Design of fMet-AMC based on the fMLF-cleaving reaction at the C-terminus of formylmethionine. (b) Mouse liver lysate was tested
for fMet-AMC-cleaving activity (fluorescence assay) and fMLF-cleaving activity (LC−MS) using the DEG assay platform. (c) CBB staining image of
a two-dimensional electrophoresis gel (IEF and blue native PAGE) of mouse liver lysate. (d) Result of the DEG assay of mouse liver lysate (total
protein = 25 μg) with fMet-AMC. The protein in the well with the strongest activity was analyzed by LC−MS/MS after trypsinization and identified
as APEH. See Figure S11 for details.
fMLF cleavage in mouse liver lysate independently of APEH
activity (Figure S19). Therefore, our results show that
proteasome is a second enzyme involved in fMLF metabolism.
However, we think that the major contributor to termination of
inflammatory signals mediated by formyl peptides is APEH
rather than proteasome, since fML, generated by proteasome, is
also a chemotactic peptide11 and fML was mainly metabolized
by APEH too (Figures 3d and S15). Indeed, we found that
APEH decreased the chemotactic activity of fMLF for model
neutrophils (Figure S20). It has been reported that altered
APEH activity is involved in the progression of several diseases,
including tumors,15,19 diabetes,20 and neurodegenerative
diseases.21 The role of formyl peptide-mediated inflammatory
signals in such diseases is still unclear. Thus, a fluorescent
substrate that would enable selective imaging of APEH activity
(i.e., formylmethionyl amide-cleaving activity) in living systems
would be useful to study this issue. Therefore, we used the
formylmethionyl motif to prepare a fluorescent probe based on
the rhodamine scaffold (Figure 4a and Scheme S1), whose
longer fluorescence wavelength is more suitable for fluores-
cence imaging of APEH activity in living systems. In this
Figure 3. (a) Rate of fluorescence increase for fMet-AMC (10 μM)
context, it is extremely important that the substrates have strict
mixed with cell lysate of HeLa cells (30 μg of protein/mL) transfected
specificity for the target enzyme, though this is difficult to
with or without mouse APEH plasmid. (b) Rate of fluorescence
achieve.14 We examined the specificity of the synthesized probe,
increase for fMet-AMC (10 μM) mixed with mouse liver lysate (500
μg of protein/mL) with or without the APEH inhibitor AA74-1 (100
nM). (c) LC−MS-based analysis of the decrease in fMLF (20 μM)
after incubation with mouse liver lysate (50 μg of protein/mL) for 6 h.
(d) Proposed metabolic pathway of fMLF in mouse liver lysate.
Enzymes indicated in red are those studied in the present research.
fMet-Rhod, for APEH using the DEG assay and found no
detectable off-target activity, confirming the high selectivity of
the probe for APEH over the whole proteome (Figure S13).
We therefore applied fMet-Rhod to study APEH activity in live
cells (Figure S21) and in animals in vivo (Figure 4b). The
probe successfully visualized APEH activity in both cases and
could be used to evaluate the efficacy of enzyme inhibitors. We
think that this probe will be useful for studying regulatory
mechanisms of diseases associated with altered APEH activity
proteasome18 (Figures S16−S18). We confirmed that fMLF
could act as a substrate of proteasome and also that the
proteasome inhibitors bortezomib and MG-13218 blocked
6004
dx.doi.org/10.1021/ja401792d | J. Am. Chem. Soc. 2013, 135, 6002−6005

Journal of the American Chemical Society
Communication
REFERENCES
■
(1) Kubinyi, H. Nat. Rev. Drug Discovery 2003, 2, 665.
(2) Saghatelian, A.; Cravatt, B. F. Nat. Chem. Biol. 2005, 1, 130.
(3) Drucker, D.; Easley, C.; Kirkpatrick, P. Nat. Rev. Drug. Discovery
2007, 6, 109.
(4) Suda, H.; Yamamoto, K.; Aoyagi, T.; Umezawa, H. Biochim.
Biophys. Acta 1980, 616, 60.
(5) Manchenko, G. P. Handbook of Detection of Enzymes on
Electrophoretic Gels, 2nd ed.; CRC Press: Boca Raton, FL, 2003.
(6) Wittig, I.; Braun, H. P.; Schagger, H. Nat. Protoc. 2006, 1, 418.
(7) Urano, Y.; Kamiya, M.; Kanda, K.; Ueno, T.; Hirose, K.; Nagano,
T. J. Am. Chem. Soc. 2005, 127, 4888.
(8) Lomolino, G.; Lante, A.; Crapisi, A.; Spettoli, P.; Curioni, A.
Electrophoresis 2001, 22, 1021.
(9) Zhang, Q.; Raoof, M.; Chen, Y.; Sumi, Y.; Sursal, T.; Junger, W.;
Brohi, K.; Itagaki, K.; Hauser, C. J. Nature 2010, 464, 104.
(10) Marasco, W. A.; Phan, S. H.; Krutzsch, H.; Showell, H. J.;
Feltner, D. E.; Nairn, R.; Becker, E. L.; Ward, P. A. J. Biol. Chem. 1984,
259, 5430.
(11) Pike, M. C.; Fischer, D. G.; Koren, H. S.; Snyderman, R. J. Exp.
Med. 1980, 152, 31.
(12) Orning, L.; Fitzpatrick, F. A. Biochemistry 1992, 31, 4218.
(13) Bauvois, B.; Dauzonne, D. Med. Res. Rev. 2006, 26, 88.
(14) Backes, B. J.; Harris, J. L.; Leonetti, F.; Craik, C. S.; Ellman, J. A.
Nat. Biotechnol. 2000, 18, 187.
Figure 4. (a) Design of fMet-Rhod. (b) Fluorescence images of mouse
liver (surrounded by white lines) after intravascular injection of fMet-
Rhod (100 μM, 100 μL/mouse). AA74-1 (650 μM, 100 μL/mouse)
was used as an inhibitor. The fluorescence of the liver was plotted
against time. n = 3 for each condition.
(15) Adibekian, A.; Martin, B. R.; Wang, C.; Hsu, K. L.; Bachovchin,
D. A.; Niessen, S.; Hoover, H.; Cravatt, B. F. Nat. Chem. Biol. 2011, 7,
469.
(16) Sandomenico, A.; Russo, A.; Palmieri, G.; Bergamo, P.;
Gogliettino, M.; Falcigno, L.; Ruvo, M. J. Med. Chem. 2012, 55, 2102.
(17) Scaloni, A.; Jones, W. M.; Barra, D.; Pospischil, M.; Sassa, S.;
Popowicz, A.; Manning, L. R.; Schneewind, O.; Manning, J. M. J. Biol.
Chem. 1992, 267, 3811.
(18) Berkers, C. R.; Verdoes, M.; Lichtman, E.; Fiebiger, E.; Kessler,
B. M.; Anderson, K. C.; Ploegh, H. L.; Ovaa, H.; Galardy, P. J. Nat.
Methods 2005, 2, 357.
(19) Scaloni, A.; Jones, W.; Pospischil, M.; Sassa, S.; Schneewind, O.;
Popowicz, A. M.; Bossa, F.; Graziano, S. L.; Manning, J. M. J. Lab. Clin.
Med. 1992, 120, 546.
and also as a tool for developing diagnostic and therapeutic
approaches.
In conclusion, we have reported herein the development of
the DEG assay as an efficient, high-throughput tool for
proteome-wide screening of specified biochemical activities
via the use of fluorescent substrates as probes. The principle of
the DEG assay is simple, but the assay is highly sensitive and
general. The major limitation might be the requirement of
tailor-made reporter substrates, but nevertheless, the method is
expected to have broad utility in that multiple detection
platforms can be employed, including coupled assays and LC−
MS-based reaction monitoring. Tools to analyze specific
functions of proteins are increasingly important in chemical
biology,2,7,15 and we believe that the DEG assay represents an
efficient platform for identifying proteins with biologically and
pharmacologically important activities and evaluating the
relative contributions of multiple proteins that exhibit a
particular activity.
(20) Shimizu, K.; Ikegami-Kawai, M.; Takahashi, T. Biol. Pharm. Bull.
2009, 32, 1632.
(21) Yamin, R.; Bagchi, S.; Hildebrant, R.; Scaloni, A.; Widom, R. L.;
Abraham, C. R. J. Neurochem. 2007, 100, 458.
ASSOCIATED CONTENT
* Supporting Information
Experimental methods and additional data. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
tlong@mol.f.u-tokyo.ac.jp
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
LC−MS/MS data were collected by APRO Life Science
Institute under contract. This research was financially
supported by the Ministry of Education, Culture, Sports,
Science and Technology of Japan (22000006 to T.N. and
40599172 to T.K.). T.K. is the recipient of a research grant
from the Mochida Memorial Foundation for Medical and
Pharmaceutical Research.
6005
dx.doi.org/10.1021/ja401792d | J. Am. Chem. Soc. 2013, 135, 6002−6005