Substrate mimicry in an activity-based probe that targets the nitrilase family of enzymes

Article abstract of DOI:10.1002/anie.200603187

(Chemical Equation Presented) Up and at it: A set of activity-based proteomics probes containing a dipeptide α-chloroacetamide scaffold that targets the nitrilase family of enzymes is described. One member of the nitrilase class, ureidopropionase-β, was found to react selectively with a single probe that mimics the enzyme's endogenous substrate N-carbamoyl β-alanine.

Full text of DOI:10.1002/anie.200603187

Communications
Activity-Based Proteomics
DOI: 10.1002/anie.200603187
Substrate Mimicry in an Activity-Based
Probe That Targets the Nitrilase Family
of Enzymes**
Katherine T. Barglow and
Benjamin F. Cravatt*
A principal goal of proteomics is to assign
functions to each of the proteins encoded by
prokaryotic and eukaryotic genomes.[1,2] A pro-
teinꢀs function can be defined on multiple levels
that reflect its coincident impact on biochemical,
cellular, and physiological processes. Activity-
based protein profiling (ABPP) is a functional
proteomics method that utilizes active-site-
directed chemical probes to characterize enzymes
in native biological samples.[3–5] ABPP probes
have proven useful for relating the activity state
of enzymes to higher-order physiological and
pathological processes, including cancer patho-
genesis,[6–8] parasite invasiveness,[9] and obesity.[10]
The extent to which these chemical tools might
also offer insights into the structures of endoge-
nous substrates of enzymes remains a pertinent,
but largely unanswered question. Herein we show
that profiling proteomes with a library of protein-
reactive chemical probes can generate structure–
activity data that are reflective of the natural
substrates of enzymes. Thus, ABPP can facilitate
the annotation of protein function at multiple
levels.
Figure 1. A library of dipeptide a-CA probes for ABPP. a) Structure of the dipeptide
a-CA, where R1 and R2 are 15 different natural amino acids, phospho-serine,
phospho-tyrosine, d-leucine, d-proline, or d-aspartate (see the Supporting Informa-
tion for a complete summary of the dipeptide a-CA probes). For proteome
screening, X=H; for protein identification, X=lysine-biotin. The a-CA reactive
group and rhodamine tag are colored in blue and red, respectively. b) Proteome
reactivity profiles for representative members of the a-CA probe library. Probes
(20 mm) were allowed to react with mouse-liver proteome (2 mg proteinmLÀ1) for
1 h at room temperature and the reactions analyzed by SDS-PAGE and in-gel
fluorescence scanning. A 44-kDa protein that reacted selectively with the Leu-Asp a-
CA probe is highlighted by a red box. See the Supporting Information for
comparison of the reactivity of this protein with Leu-Asp and Leu-Asn a-CA.
Fluorescent gel images are shown in grayscale.
Previous ABPP studies revealed markedly
distinct proteome reactivity profiles for members
of a library of rhodamine-tagged dipeptide a-
chloroacetamide (a-CA) probes.[10] Some protein
targets showed exclusive or preferential reactivity
with a single a-CA, whereas others showed a more
promiscuous profile, reacting with several mem-
bers of the probe library. Among the more intriguing proteins
that fell into the former class was a 44-kDa liver protein that
exclusively reacted with Leu-Asp a-CA, where Leu and Asp
represent R2 and R1, respectively (Figure 1a). We sought to
further characterize the probe-labeling profile of this protein
by testing its reactivity with an expanded library of a-CA
probes in which either the R1 or R2 position was held constant
as leucine and the complementary site varied with one of
twenty amino acid side chains (Figure 1a). The 44-kDa
protein was again found to react specifically with Leu-Asp
a-CA, and showed negligible labeling with other probes
including the structurally related Leu-Glu, Leu-Gln, and Asp-
Leu a-CAs (Figure 1b and the Supporting Information).
The 44-kDa protein was enriched from mouse liver
proteomes by labeling with a trifunctional Leu-Asp a-CA
[*] K. T. Barglow, Prof. B. F. Cravatt
The Skaggs Institute for Chemical Biology and
Departments of Cell Biology and Chemistry
The Scripps Research Institute
10550 North Torrey Pines Road, La Jolla, CA 92037 (USA)
Fax: (+1)858-784-8023
E-mail: cravatt@scripps.edu
[**] This research was supported by the NIH (CA087660), the Howard
Hughes Medical Institute (predoctoral fellowship to K.T.B.), and the
Skaggs Institute for Chemical Biology. We thank M. Patricelli (Activx
Biosciences) for assistance with the identification of probe sites of
labeling for nitrilase enzymes and the members of the Cravatt lab for
helpful discussions and critical reading of the manuscript.
Supporting information for this article, which also contains details
of the materials and methods, is available on the WWW under
http://www.angewandte.org or from the author.
7408
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7408 –7411

Angewandte
Chemie
probe coupled to both biotin and rhodamine, followed by
avidin affinity chromatography and SDS-PAGE. Liquid
chromatography (LC)–MS analysis of a trypsin digest of the
gel-excised 44-kDa protein identified it as ureidopropionase-
b (Upb), a 43.9-kDa member of the nitrilase superfamily (see
the Supporting Information). To confirm this identification, a
cDNA (complementary DNA) encoding the full-length
mouse Upb (appended with a C-terminal myc epitope tag)
was subcloned into the pcDNA3.1 expression vector and
transiently transfected into COS-7 cells. A strong Leu-Asp a-
CA-reactive 44-kDa protein was detected in Upb-transfected
cells, but not mock-transfected cells (Figure 2a, left). Like
native liver Upb, the recombinantly expressed form of this
enzyme showed nearly exclusive labeling with Leu-Asp a-CA
(Figure 2b).
(CarbAla) to b-alanine, which is a central step in the
pyrimidine catabolic pathway.[13] Mutation of Upb in
humans leads to a buildup of CarbAla, which exerts neuro-
toxic effects resulting in severe neuropathology and move-
ment disorders.[13] In considering possible explanations for the
high selectivity displayed by Upb for Leu-Asp a-CA, we
noted a striking structural similarity between this probe and
Upbꢀs endogenous substrate CarbAla (Figure 3). This struc-
Figure 3. Leu-Asp a-CA is structurally similar to the natural Upb
substrate CarbAla. Top: structural comparison of Leu-Asp a-CA and
CarbAla. Bottom: a comparison of the rates of inactivation of Upb by
analogues of Leu-Asp a-CA. The structure of the rhodamine-tagged
Leu-Asp probe can be inferred from Figure 1.
tural relation suggested that Leu-Asp a-CA bound and
labeled Upb at the active site. This premise was confirmed
by mapping the site of probe modification, which was
identified by tandem MS analysis as cysteine 233, the catalytic
nucleophile of Upb (Table 1).[11] Mutation of this residue to
alanine produced a Upb variant (C233A) that was expressed
at wild-type levels in transfected COS-7 cells but failed to
react with Leu-Asp a-CA (Figure 2a, right).
Figure 2. Identification and characterization of Upb as a target of Leu-
Asp a-CA. a) Left: Upb-transfected, but not mock-transfected, COS-7
cells possess a 44-kDa protein that reacted with Leu-Asp a-CA; right:
mutation of the cysteine nucleophile of Upb to alanine (C233A)
eliminated labeling by Leu-Asp a-CA. b) Native (liver) and recombinant
(transfected COS-7) Upb enzymes selectively reacted with Leu-Asp a-
CA compared with other dipeptide a-CA probes. Note that low signals
were also observed in the molecular-mass range of Upb for the Leu-Tyr
and Leu-His probes; however, these probes showed much higher
overall background proteome reactivity (see Figure 1b), suggesting
that these signals may be nonspecific. Top: fluorescent gel images
shown in grayscale. Bottom: quantification of fluorescent signals. Data
represent the average Æstandard deviation for four independent
experiments. If =relative fluorescence, WT=wild type.
Table 1: Sites of a-CA probe labeling for members of nitrilase family.
Enzyme
Labeled peptide[a]
Site of labeling[a]
Upb
Nit1
IAVNIC*YGR
VGLAIC*YDMRFPELSLK
C233
C199
[a] Probe-labeled peptides and sites of labeling were determined by using
an LC–MS platform for ABPP as described previously.[20,21] Proteomes
were treated with an equimolar mixture of Leu-Asp, Leu-d-Asp, and d-
Leu-d-Asp a-CA probes (20 mm of each probe).
The nitrilase family is a phylogenetically ancient class of
enzymes with multiple mammalian members that utilize a
Further evidence that Leu-Asp a-CA utilized substrate
mimicry to achieve active-site modification of Upb was
sought by screening analogues of this probe for their ability to
block enzyme labeling. Such competitive ABPP studies, when
performed in a concentration- and time-dependent manner,
À
conserved Cys-Lys-Glu catalytic triad to cleave C N bonds,
including nitriles, carbamates, ureas, and amides.[11,12] Upb, in
particular, catalyzes the hydrolysis of N-carbamoyl b-alanine
Angew. Chem. Int. Ed. 2006, 45, 7408 –7411
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 7409

Communications
can provide an estimate of the relative potency with which
enzymes. Nitrilases perform important functions in a wide
range of organisms, including humans. For example, Upb is a
key component of the pyrimidine catabolic pathway, catalyz-
ing the conversion of CarbAla to b-alanine.[13] This reaction
appears to constitute the dedicated endogenous function of
Upb, as is reflected in the enzymeꢀs restricted substrate
selectivity, probe reactivity, and tissue distribution profiles
(specific expression in liver; http://symatlas.gnf.org/SymAt-
las/), as well as the metabolic defects that arise from its
genetic mutation.[13] On the other hand, Nit1 is more
ubiquitously expressed in mammalian tissues,[18] and the
disruption of this enzyme leads to multiple phenotypes in
mice, including accelerated cell growth and increased inci-
dence of chemically induced tumors.[18] These physiological
findings, coupled with the more promiscuous probe reactivity
profile of Nit1, suggest that it may display a broader
metabolic function than Upb. Confirmation of this hypothesis
will require identification of endogenous substrates for Nit1.
The additional finding that optimal probes for Upb
mimicked the structure of the enzymeꢀs natural substrate
has broader implications for enzyme annotation by using
ABPP methods. Although other ABPP probes have been
designed based on the substrate preferences of their target
enzyme classes (e.g., negatively charged acyloxymethyl
ketones for caspases[19]), our data with Upb suggest provoca-
tively that the relationships between enzymes and substrates
can be “discovered” de novo by screening libraries of
structurally diverse probes. By extension, we speculate that
the probe-reactivity profiles of uncharacterized enzymes may
also contain examples of substrate mimicry and, thus, provide
insights into the endogenous biochemical functions of these
proteins.
inhibitors target the active site of an enzyme.[14] An a-CA was
designed that most closely resembled the Upb substrate
CarbAla, and this agent (1) was found to block probe labeling
of Upb with a kobs/[I] value of 98 Æ 22mÀ1 minÀ1 (I = compet-
itive inhibitor, Figure 3). The addition of a leucine group
(agent 2) accelerated the rate of inactivation of Upb,
providing an agent that displayed a kobs/[I] value of 470 Æ
73mÀ1 minÀ1 (Figure 3). The introduction of a third amino acid
derivative possessing an extended alkynyl side chain (agent 3)
further increased the rate of Upb inactivation (kobs/[I] =
1450 Æ 255mÀ1 minÀ1, Figure 3) and enabled confirmation of
covalent labeling of the enzyme by a click-chemistry reaction
with an azide-rhodamine reporter group[15,16] (see the Sup-
porting Information). These results indicate that an a-CA
probe containing the core features of the CarbAla substrate is
capable of covalently labeling the active site of Upb, but that
additional features of Leu-Asp a-CA further enhance this
reaction.
Given that labeling of Upb by Leu-Asp a-CA occurred on
the enzymeꢀs cysteine nucleophile, an essential catalytic
residue conserved among all nitrilases,[11,12] we next tested
whether a-CA probes might label other members of this
enzyme family. Analysis of mouse-liver proteomes by using
the MS-based platform ABPP-MudPIT (Multidimensional
Protein Identification Technology)[17] identified nitrilase 2
(Nit2), along with Upb, as specific targets of Leu-Asp a-CA
(see the Supporting Information). Mouse nitrilase 1 (Nit1)
was recombinantly expressed as a myc-tagged fusion protein
in COS7 cells and treated with an expanded panel of a-CA
probes that included agents where the chirality and order of
the Leu-Asp dipeptide scaffold were systematically varied.
Strong labeling was observed for Nit1 with several of the
dipeptide a-CA probes (Figure 4 a and b), which contrasted
with the more selective labeling pattern of Upb (Figures 2b
and 4b). Tandem MS analysis revealed that probe labeling of
Nit1 occurred on the enzymeꢀs cysteine nucleophile (C199;
Table 1). These data indicate that dipeptide a-CAs serve as
effective ABPP probes for multiple members of the nitrilase
enzyme class.
Received: August 5, 2006
Published online: October 11, 2006
Keywords: amino acids · biological activity · nitrilases ·
.
proteomics · screening
In summary, we report the first example of active-site-
directed proteomic probes that target the nitrilase family of
[1] P. C. Babbitt, Curr. Opin. Chem. Biol. 2003, 7, 230.
[2] A. Saghatelian, B. F. Cravatt, Nat. Chem. Biol. 2005, 1, 130.
[3] Y. Liu, M. P. Patricelli, B. F. Cravatt, Proc. Natl. Acad. Sci. USA
1999, 96, 14694.
[4] A. E. Speers, B. F. Cravatt, ChemBioChem 2004, 5, 41.
[5] A. B. Berger, P. M. Vitorino, M. Bogyo, Am. J. Pharmacoge-
nomics 2004, 4, 371.
[6] N. Jessani, Y. Liu, M. Humphrey, B. F. Cravatt, Proc. Natl. Acad.
Sci. USA 2002, 99, 10335.
[7] N. Jessani, M. Humphrey, W. H. McDonald, S. Niessen, K.
Masuda, B. Gangadharan, J. R. Yates, 3rd, B. M. Mueller, B. F.
Cravatt, Proc. Natl. Acad. Sci. USA 2004, 101, 13756.
[8] J. A. Joyce, A. Baruch, K. Chehade, N. Meyer-Morse, E.
Giraudo, F. Y. Tsai, D. C. Greenbaum, J. H. Hager, M. Bogyo,
D. Hanahan, Cancer Cell 2004, 5, 443.
[9] D. C. Greenbaum, A. Baruch, M. Grainger, Z. Bozdech, K. F.
Medzihradszky, J. Engel, J. DeRisi, A. A. Holder, M. Bogyo,
Science 2002, 298, 2002.
Figure 4. Identification and characterization of Nit1 as a target for
dipeptide a-CA probes. a) Nit1-transfected, but not mock-transfected,
COS7 cells possess a 35-kDa protein that reacted with several
members of the a-CA probe library (mock-transfected proteome shown
was allowed to react with Leu-Tyr a-CA). b) Comparison of the
reactivity profiles of Upb and Nit with a chiral library of Leu-Asp and
Asp-Leu a-CA probes.
[10] K. T. Barglow, B. F. Cravatt, Chem. Biol. 2004, 11, 1523.
[11] P. Bork, E. V. Koonin, Protein Sci. 1994, 3, 1344.
[12] C. Brenner, Curr. Opin. Struct. Biol. 2002, 12, 775.
7410
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7408 –7411

Angewandte
Chemie
[13] A. B. van Kuilenburg, R. Meinsma, E. Beke, B. Assmann, A.
Ribes, I. Lorente, R. Busch, E. Mayatepek, N. G. Abeling, A.
van Cruchten, A. E. Stroomer, H. van Lenthe, L. Zoetekouw, W.
Kulik, G. F. Hoffmann, T. Voit, R. A. Wevers, F. Rutsch, A. H.
van Gennip, Hum. Mol. Genet. 2004, 13, 2793.
[14] D. Leung, C. Hardouin, D. L. Boger, B. F. Cravatt, Nat.
Biotechnol. 2003, 21, 687.
[15] A. E. Speers, G. C. Adam, B. F. Cravatt, J. Am. Chem. Soc. 2003,
125, 4686.
[16] A. E. Speers, B. F. Cravatt, Chem. Biol. 2004, 11, 535.
[17] N. Jessani, S. Niessen, B. Q. Wei, M. Nicolau, M. Humphrey, Y.
Ji, W. Han, D. Y. Noh, J. R. Yates III, S. S. Jeffrey, B. F. Cravatt,
Nat. Methods 2005, 2, 691.
[18] S. Semba, S. Y. Han, H. R. Qin, K. A. McCorkell, D. Iliopoulos,
Y. Pekarsky, T. Druck, F. Trapasso, C. M. Croce, K. Huebner, J.
Biol. Chem. 2006, 281, 28244.
[19] D. Kato, K. M. Boatright, A. B. Berger, T. Nazif, G. Blum, C.
Ryan, K. Chehade, G. S. Salvensen, M. Bogyo, Nat. Chem. Biol.
2005, 1, 33.
[20] G. C. Adam, J. J. Burbaum, J. W. Kozarich, M. P. Patricelli, B. F.
Cravatt, J. Am. Chem. Soc. 2004, 126, 1363.
[21] E. S. Okerberg, J. Wu, B. Zhang, B. Samii, K. Blackford, D. T.
Winn, K. R. Shreder, J. J. Burbaum, M. P. Patricelli, Proc. Natl.
Acad. Sci. USA 2005, 102, 4996.
Angew. Chem. Int. Ed. 2006, 45, 7408 –7411
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7411