Cinnamic aldehyde derived probes for the active site labelling of pathogenesis associated enzymes

Article abstract of DOI:10.1039/b905527d

Michael acceptor based natural product derived probes are selective and sensitive chemical tools for the identification and characterization of pathologically relevant enzymes in MRSA.

Full text of DOI:10.1039/b905527d

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Cinnamic aldehyde derived probes for the active site labelling
of pathogenesis associated enzymesw
Maximilian Pitscheider and Stephan A. Sieber*
Received (in Cambridge, UK) 18th March 2009, Accepted 30th April 2009
First published as an Advance Article on the web 19th May 2009
DOI: 10.1039/b905527d
Michael acceptor based natural product derived probes are
selective and sensitive chemical tools for the identification and
characterization of pathologically relevant enzymes in MRSA.
genesis, such as PTPs and cysteine proteases.12 The general
probe design of our library consisted of the CA core scaffold
that was appended by a short alkyne handle as a benign tag for
visualization and enrichment of labelled proteins (Fig. 1A).
The modification of the alkyne tag via the 1,3-dipolar Huisgen
cycloaddition (click chemistry, CC) allows us to introduce the
bulky reporter group (e.g. rhodamine) after enzyme binding
and cell preparation (Fig. 1B).13 Labelled enzymes can be
run on SDS gels, visualized by fluorescent scanning and
subsequently identified by mass spectrometry.
With the emergence of multiresistant bacterial strains, infectious
diseases once again pose a major threat to public health.1 One
fundamental challenge is the identification and functional
characterization of resistance free enzyme targets that are
essential for bacterial viability or pathogenesis and which are
not addressed by currently described antibiotics. We recently
introduced two strategies for the identification and functional
characterization of bacterial enzymes based on bioactive
b-lactam2 and b-lactone3,4 compound libraries. b-Lactam
and b-lactone scaffolds represent electrophilic entities that
are prone to react with nucleophilic enzyme active sites
forming a covalent bond. The b-lactam and b-lactone probe
libraries were successfully applied in activity based protein
profiling (ABPP) experiments (pioneered by Cravatt, Bogyo
and co-workers)5,6 to identify enzymes that play crucial roles
in resistance and virulence. Inspired by the success of this
strategy we here explore cinnamic aldehyde (CA) as an
electrophilic scaffold for bacterial target identification.
Cinnamic aldehyde is the key flavour compound in cinnamon
and an important FDA approved food additive that contains
an a,b-unsaturated carbonyl Michael pharmacophore.7
Previous studies with a,b-unsaturated carbonyl Michael acceptor
systems revealed that these compounds exhibit a preference for
the modification of cysteine active site residues.8,9 The high
reactivity of many Michael acceptor systems compromises
their application for selective biological studies. Peptide
scaffolds that are attached to CA, however, have been reported
to be selective inhibitors of protein tyrosine phosphatases
(PTPs).10 Unmodified CA displays potent anti-cancer
activities, suggesting a promising biological potential of this
compound in the binding and inhibition of pharmacologically
relevant enzyme classes.11
Our general synthetic procedure is shown in Fig. S1, ESIw.
All amino acid positions in the CA tripeptides were varied by
charged, aromatic or hydrophobic residues in order to increase
the chance of specific probe–protein interactions. The syn-
thesis of the peptide scaffold was carried out on a Rink-amide
resin and followed standard Fmoc solid phase chemistry
procedures. The N-terminal end of the peptides was coupled
with the free acid of the reactive CA binding group in the last
step of synthesis. In addition to CA we also attached the
corresponding cinnamic acid (CAC) moiety. In total the
library consists of three CAC (please refer for structures and
compositions to Fig. S1, ESIw) and 11 CA peptide probes
(Fig. 1A).
Previous studies of tripeptide CA and CAC scaffolds
revealed potent inhibition of PTP1B, with Ki values ranging
Following our goal to identify novel therapeutic targets in
pathogenic bacteria we synthesized a CA peptide probe library
to monitor the activity and function of cysteine active site
containing enzymes that play crucial roles in bacterial patho-
Center for Integrated Protein Science Munich CIPSM, Department of
Chemistry and Biochemistry, Ludwig-Maximilians-Universitat
Munchen, Butenandtstr. 5-13, 81377 Munich, Germany.
¨
¨
E-mail: Stephan.sieber@cup.uni-muenchen.de; Fax: +49 2180 77756;
Tel: +49 2180 77660
w Electronic supplementary information (ESI) available: Synthetic
procedures, characterization, biological methods and figures. See
DOI: 10.1039/b905527d
Fig. 1 Structures and applications of CA peptide probes. (A) Struc-
tures of all CA probes. (B) Proteomes are first treated with the CA
probes and subsequently appended with a fluorescent dye via CC.
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 3741–3743 | 3741

View Article Online
and added 50 mM of all probes (Fig. 2C and D). Interestingly,
several probes exhibited strong labelling intensities of both
enzymes in an activity dependent manner as visualized by the
corresponding heat control (Fig. S2, ESIw). Peptide
sequences of probes which performed best in PTPA labelling
contained no charged residues, which indicates that this
enzyme prefers small hydrophobic (Gly) or aromatic (Phe)
residues in its active site. A similar labelling profile was
obtained for PTPB, which reflects a close relationship between
these two enzymes (Fig. 2D).18 In contrast, our three CAC
probes in which the aldehyde moiety is replaced by a free
carboxylic acid did not show any enzyme labelling, indicating
that the a,b unsaturated acid is not electrophilic enough to
acylate the enzyme active site.
Many covalent PTP inhibitors attach to a conserved active
site cysteine residue which functions as the catalytic nucleo-
phile that attacks the tyrosine phosphate to form a transient
phosphocysteinyl intermediate.19 The cinnamic aldehyde
probes most likely attach via this cysteine residue and inhibit
enzyme activity. To test this hypothesis and to figure out
which of the three cysteines in uncharacterized MRSA PTPA
is relevant for catalysis we individually mutated all three
cysteines (C8, C13 and C138) to alanine and checked the
labelling efficiency of the corresponding mutants with probe
10 (Fig. 2E). It turned out that C13A and C138A were
strongly labelled, however, C8A revealed no labelling inten-
sity, indicating that this residue functions as the catalytically
active site.
Fig. 2 Labelling of protein tyrosine phosphatases. (A) Labelling of
recombinant human PTP1B with 50 mM of individual probes. (B) Heat
control (DT) and active site competition by a pre-incubation of the
enzyme with a 40-fold excess of CAC probes 12, 13 and 14 and
subsequent labelling with 50 mM 3. (C) Labelling of recombinant
MRSA PTPA in E. coli proteome with the probe library (50 mM).
(D) Labelling of MRSA PTPB with the probe library (50 mM).
(E) Labelling of PTPA mutants with 50 mM 10.
from 5.4 mM to 79 nM, respectively.14,15 PTP1B is an
important drug target for the treatment of both type 2 diabetes
and obesity.16 To test whether the concept of utilizing tripeptide
CA probes is effective in labelling this PTP we incubated the
commercially available recombinant enzyme with a selection
from our probe library.
We next evaluated if PTP active site labelling also results in
inhibition of catalytic activity. PTPA activity was measured by
the catalysed dephosphorylation reaction of p-nitrophenyl-
phosphate that resulted in the generation of p-nitrophenol as
a chromogenic product that can be measured at 410 nm on a
UV-Vis spectrometer. Several CA probes revealed inhibition
of substrate turnover, with IC50 values ranging from 5 mM for
probe 10 to 271 mM for probe 4 (Fig. 3A, Fig. S3, ESIw).
Interestingly, probe 10 which had already revealed the most
intense labelling also exhibited the lowest IC50 value of all
tested compounds, while CA probe 4, which weakly labelled
PTPA, also weakly inhibited its activity. However, the hydro-
phobic nature of 10 caused solubility problems above 100 mM
in the assay buffer and restricted the evaluation of full PTP
inhibition. As mentioned previously, CAC based probes were
not capable of PTPA labelling due to their low reactivity.
In fact, several probes exhibited strong labelling intensities
(50 mM probe concentration) of the recombinant enzyme
(250 ng) with 3 being one of the best (Fig. 2A). These results
indicate that PTP1B prefers aromatic (3 and 7) or hydro-
phobic (10) residues in its active site. Heat denaturation of
PTP1B prior to probe labelling resulted in a strongly reduced
fluorescent signal which shows that the binding of probes to
the enzyme occurred specifically only with an intact and
properly folded enzyme (Fig. 2B). Since CAC tripeptide scaf-
folds, with a free acid moiety, have been reported to be potent
PTP1B inhibitors we tested labelling of the recombinant
enzyme with probe 12 which closely mimics a published
Gly-Glu-Glu-CAC inhibitor.14 No labelling could be observed
indicating that a,b unsaturated acids may not be reactive
enough to covalently attach to the enzyme active site. To test
whether the CAC probes bind to the PTP1B active site we
pre-incubated the enzyme with a 40-fold excess of 12, 13 and
14 and subsequently added 50 mM of the reactive CA probe 3
(Fig. 2B). In fact, labelling was efficiently reduced, emphasiz-
ing that the CAC based inhibitor exerts its activity via
reversible binding and competes for the same active site as
the CA probe.
With these initial experiments that validated the general
utility of our probes in PTP labelling we next tested the
binding and inhibition of the probe library against two
PTPs from the pathogenic bacteria methicillin resistant
Staphylococcus aureus (MRSA). The precise role of these PTPs
remains elusive so far but it is assumed that they are involved
in virulence mechanisms.17,18 We cloned and overexpressed
PTPA and PTPB from MRSA recombinantly in Escherichia coli
Fig. 3 Inhibition of PTP activity. (A) IC50 values of PTPA inhibition
with CA and CAC compounds. (B) Active site competition of PTPA
and PTPB by a pre-incubation of the enzyme with a 100-fold excess of
CAC probe 12 and subsequent labelling with 50 mM probe 10.
ꢀc
This journal is The Royal Society of Chemistry 2009
3742 | Chem. Commun., 2009, 3741–3743

View Article Online
However, we were able to obtain good inhibition of PTPA
with an IC50 value of 39 mM (12), indicating that cinnamic acid
based tripeptides are promising reversible inhibitors for
bacterial PTPs. In addition, pre-incubation of PTPA and
PTPB with a 100-fold excess of CAC compound 12 and
subsequent incubation with 10 mM CA probe 10 resulted in
strongly reduced labelling for PTPB and a significant reduc-
tion for PTPA, providing further confirmation that both
probes compete for the same active site (Fig. 3B).
mutation of Cys171 to alanine. As expected, the mutant
enzyme revealed no probe labelling (Fig. 4C). Probes 7 and
10 exhibited the best labelling intensities of SsaA2 while probe
1 showed no labelling, indicating that the enzyme does not
prefer two adjacent Gly residues next to the CA group. This
information could be useful to gain insight into the native
substrate preferences of this uncharacterized virulence
associated protein. To estimate the strength of probe–protein
interactions, we compared the reactivity profile of 10 with
SsaA2 across a concentration range from 100 to 0.2 mM
(Fig. 4D). The protein can still be detected at concentra-
tions as low as 1 mM, which indicates an affinity suitable
for monitoring the function and activity of this putatively
important enzyme in future studies.
The previous experiments validated the utility of CA probes
to label and inhibit recombinant bacterial phosphatases. Based
on these results we wanted to test if the probes were capable of
labelling enzymes in whole bacterial proteome lysates. For an
initial evaluation we chose to analyse the labelling pattern of
our library in a proteome of the antibiotic resistant MRSA
strain. Enzymes which are labelled in MRSA proteomes could
be involved in pathogenesis and resistance and may represent
valuable targets for therapeutic interventions.
In conclusion, we utilized cinnamic aldehyde as a proto-
type Michael acceptor system for the activity based labelling
of pathogenesis associated enzymes including phosphatases
as well as a secretory antigen from MRSA. Based on our
library approach we were able to identify substrate preferences
of individual uncharacterized enzymes which may help to
guide the design of specific inhibitors. These findings illustrate
that our Michael acceptor based natural product derived
probes are useful, selective and sensitive chemical tools for
the identification and characterization of physiologically and
pathologically relevant enzymes in complex proteomes.
We thank Prof. T. Carell for his generous support and
acknowledge funding by the DFG Emmy Noether program
We therefore used a selection of our best probes to profile
the MRSA membrane proteome. Although we were not able
to detect PTPA and PTPB which may be less abundant or
inactive under the test conditions, one new protein target was
selectively labelled (Fig. 4A). Subsequent identification by
LC-MS analysis revealed the identity of this protein as a
virulence associated Staphylococcal secretory antigen (SsaA2).
To validate the MS result, we recombinantly expressed SsaA2
and subsequently confirmed its labelling by the corresponding
probes (Fig. 4B). SsaA2 is an uncharacterized protein that has
been previously reported to be involved in Staphylococcus
epidermidis sepsis and infective endocarditis.20 The protein
contains a cysteine protease sub-domain that is supposed to
process the protein into its mature form. Most likely SsaA2 is
labelled by the attachment of the active site cysteine residue
onto the CA tripeptide probe. To support this notion, we
pre-incubated the proteomic mixture with 5 mM N-ethyl-
maleimide (NEM) to alkylate all free cysteine residues.
Subsequent labelling with probe 10 revealed no residual
labelling, indicating that a reactive cysteine was blocked and
was no longer susceptible to CA probe alkylation (Fig. 4B). In
addition, we identified the reactive cysteine by an active site
and SFB 749, Romer-Stiftung, Fonds der chem. Industrie and
¨
CiPSM. We thank K. Kurz for excellent assistance.
Notes and references
1 G. Taubes, Science, 2008, 321, 356.
2 I. Staub and S. A. Sieber, J. Am. Chem. Soc., 2008, 130, 13400.
3 T. Bottcher and S. A. Sieber, J. Am. Chem. Soc., 2008, 130,
¨
14400.
4 T. Bottcher and S. A. Sieber, Angew. Chem., Int. Ed., 2008, 47,
4600.
¨
5 M. J. Evans and B. F. Cravatt, Chem. Rev., 2006, 106, 3279.
6 M. Fonovic and M. Bogyo, Expert Rev. Proteomics, 2008, 5, 721.
7 C. M. Cabello, W. B. Bair, 3rd, S. D. Lamore, S. Ley, A. S. Bause,
S. Azimian and G. T. Wondrak, Free Radical Biol. Med., 2009, 46,
220.
8 E. Weerapana, G. M. Simon and B. F. Cravatt, Nat. Chem. Biol.,
2008, 4, 405.
9 S. Liu, B. Zhou, H. Yang, Y. He, Z. X. Jiang, S. Kumar, L. Wu
and Z. Y. Zhang, J. Am. Chem. Soc., 2008, 130, 8251.
10 H. Fu, J. Park and D. Pei, Biochemistry, 2002, 41, 10700.
11 J. Park, H. Fu and D. Pei, Bioorg. Med. Chem. Lett., 2004, 14,
685.
12 R. DeVinney, O. Steele-Mortimer and B. B. Finlay, Trends
Microbiol., 2000, 8, 29.
13 A. E. Speers, G. C. Adam and B. F. Cravatt, J. Am. Chem. Soc.,
2003, 125, 4686.
14 L. Bialy and H. Waldmann, Angew. Chem., Int. Ed., 2005, 44,
3814.
15 E. J. Moran, S. Sarshar, J. F. Cargill, M. M. Shahbaz, A. Lio, A.
M. M. Mjalli and R. W. Armstrong, J. Am. Chem. Soc., 1995, 117,
10787.
16 S. Zhang and Z. Y. Zhang, Drug Discovery Today, 2007, 12, 373.
17 M. H. Bender and J. Yother, J. Biol. Chem., 2001, 276, 47966.
18 O. Ilan, Y. Bloch, G. Frankel, H. Ullrich, K. Geider and
I. Rosenshine, EMBO J., 1999, 18, 3241.
19 Z. Y. Zhang and J. E. Dixon, Adv. Enzymol. Relat. Areas Mol.
Biol., 1994, 68, 1.
Fig. 4 Labelling of the MRSA proteome. (A) Labelling of the MRSA
membrane proteome with a selection of CA and CAC probes (50 mM).
SsaA2 = Staphylococcal secretory antigen. (B) Labelling of recombi-
nant SsaA2 (À before induction, + after induction, DT heat control).
Blocking of free cysteine residues in native SsaA2 with NEM.
(C) SsaA2 active site cysteine mutation (À before induction, + after
induction). (D) Concentration dependence of SsaA2 labelling.
20 S. Lang, M. A. Livesley, P. A. Lambert, W. A. Littler and
T. S. Elliott, FEMS Immunol. Med. Microbiol., 2000, 29, 213.
ꢀc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 3741–3743 | 3743