1,3,5-triazine as a modular scaffold for covalent inhibitors with streamlined target identification

Article abstract of DOI:10.1021/ja400427e

Small-molecule inhibitors can accelerate the functional annotation and validate the therapeutic potential of proteins implicated in disease. Phenotypic screens provide an effective platform to identify such pharmacological agents but are often hindered by challenges associated with target identification. For many protein targets, these bottlenecks can be overcome by incorporating electrophiles into small molecules to covalently trap interactions in vivo and by employing bioorthogonal handles to enrich the protein targets directly from a complex proteome. Here we present the trifunctionalized 1,3,5-triazine as an ideal modular scaffold for generating libraries of irreversible inhibitors with diverse target specificities. A divergent synthetic scheme was developed to derivatize the triazine with an electrophile for covalent modification of target proteins, an alkyne as a click-chemistry handle for target identification, and a diversity element to direct the compounds toward distinct subsets of the proteome. We specifically targeted our initial library toward cysteine-mediated protein activities through incorporation of thiol-specific electrophiles. From this initial screen we identified two compounds, RB-2-cb and RB-11-ca, which are cell permeable and highly selective covalent modifiers for Cys239 of β-tubulin (TUBB) and Cys53 of protein disulfide isomerase (PDI) respectively. These compounds demonstrate in vitro and cellular potencies that are comparable to currently available modulators of tubulin polymerization and PDI activity. Our studies demonstrate the versatility of the triazine as a modular scaffold to generate potent and selective covalent modifiers of diverse protein families for chemical genetics applications.

Full text of DOI:10.1021/ja400427e

Communication
pubs.acs.org/JACS
1,3,5-Triazine as a Modular Scaffold for Covalent Inhibitors with
Streamlined Target Identification
Ranjan Banerjee, Nicholas J. Pace, Douglas R. Brown, and Eranthie Weerapana*
Department of Chemistry, Merkert Chemistry Center, Boston College, Chestnut Hill, Massachusetts 02467, United States
S
* Supporting Information
protein. This is exemplified by the kinase family, where
ABSTRACT: Small-molecule inhibitors can accelerate the
functional annotation and validate the therapeutic
potential of proteins implicated in disease. Phenotypic
screens provide an effective platform to identify such
pharmacological agents but are often hindered by
challenges associated with target identification. For many
protein targets, these bottlenecks can be overcome by
incorporating electrophiles into small molecules to
covalently trap interactions in vivo and by employing
bioorthogonal handles to enrich the protein targets
directly from a complex proteome. Here we present the
trifunctionalized 1,3,5-triazine as an ideal modular scaffold
for generating libraries of irreversible inhibitors with
diverse target specificities. A divergent synthetic scheme
was developed to derivatize the triazine with an electro-
phile for covalent modification of target proteins, an alkyne
as a click-chemistry handle for target identification, and a
diversity element to direct the compounds toward distinct
subsets of the proteome. We specifically targeted our initial
library toward cysteine-mediated protein activities through
incorporation of thiol-specific electrophiles. From this
initial screen we identified two compounds, RB-2-cb and
RB-11-ca, which are cell permeable and highly selective
covalent modifiers for Cys239 of β-tubulin (TUBB) and
Cys53 of protein disulfide isomerase (PDI) respectively.
These compounds demonstrate in vitro and cellular
potencies that are comparable to currently available
modulators of tubulin polymerization and PDI activity.
Our studies demonstrate the versatility of the triazine as a
modular scaffold to generate potent and selective covalent
modifiers of diverse protein families for chemical genetics
applications.
nonconserved, yet functionally relevant, cysteine and lysine
residues have proved invaluable in the development of selective
inhibitors and activity-based probes, respectively.2,3 Developing
and screening libraries of small molecules targeted toward these
reactive amino acids has potential to reveal pharmacological
modulators for diverse protein classes.
One such reactive amino acid found within protein
frameworks is cysteine. Cysteine residues are vital for redox
and nucleophilic catalysis and allosteric regulation in proteins
from diverse classes.4 These functional cysteine residues display
significantly enhanced reactivity with thiol-targeted electro-
philes, such as haloacetamides and acrylamides.5,6 Here we
exploit this elevated reactivity to develop a small-molecule
library to covalently modify functionally relevant cysteines. The
covalent mode of action of these small molecules helps to
overcome one of the largest bottlenecks that currently exist in
high-throughput screening efforts: target identification. Incor-
poration of protein-reactive electrophiles into small-molecule
scaffolds serve to trap the protein target within a physiologically
relevant system. Here, we further expedite this target
identification step by incorporating a bioorthogonal alkyne
handle into our small molecules, thereby enabling the use of
click-chemistry7 to enrich, visualize and identify protein targets.
The potency and selectivity of our small molecules can be
directly assessed within the context of an entire cellular
proteome, and target identification is streamlined.
Our platform is centered upon a small-molecule library that
possesses the following elements: (1) a rigid triazine scaffold,
(2) chloroacetamide (CA), chloronitrobenzyl (CNB) or
acrylamide (Acr) reactive groups (R2) known to covalently
modify reactive cysteine residues,5,8,9 (3) an alkyne handle for
downstream conjugation to reporter tags (R1), and (4) a
variable diversity element (R3) to direct reactivity to distinct
subsets of proteins (Figure 1A). The heterocyclic triazine
scaffold has recently been utilized in multiple large-scale
combinatorial efforts, due to the ease of functionalization and
structural similarity to bioactive endogenous heterocycles such
as purines and pyrimidines.10−14 The triazine has three-fold
symmetry, allowing positional modification with the required
R1, R2 and R3 groups. Here we report the synthesis and
evaluation of 20 compounds incorporating the CA, CNB and
Acr electrophiles, together with a variety of diversity elements
(R3 groups) (Figure 1B). The R3 group was varied across the
library and encompasses a multitude of physicochemical
mall-molecule modulators of protein activity can serve as
Stools to annotate biochemical pathways and validate
therapeutic targets relevant to disease. Despite the emergence
of high-throughput screening platforms and cellular assays to
identify bioactive compounds, a vast subsection of the
proteome remains functionally unannotated with no available
pharmacological tools.1 Proteins from diverse functional
families contain reactive amino acids that can be exploited in
the development of covalent modulators. This subset of
potential protein targets is not restricted to enzymes that
utilize a catalytic nucleophile, but encompasses the myriad of
proteins that contain reactive amino acids within either the
active site or distal, yet functionally relevant, regions of the
Received: January 14, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja400427e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
The synthesis of the trifunctionalized triazines was performed
using a solution-phase scheme adapted from previously
reported solid-phase synthetic routes.11 This synthetic scheme
allows for the large-scale divergent synthesis of the final
compounds from common intermediates (Scheme S1,
Supporting Information). Our first goal was to assess the cell
permeability of these trifunctionalized triazines and furthermore
to determine if each compound targets a unique protein or
subset of protein targets. To achieve this, we incubated HeLa
cells with the triazines for 1 h and evaluated protein labeling
after cell lysis, click chemistry to incorporate a fluorescent
rhodamine moiety, and in-gel fluorescence.15,16 These studies
confirmed the ability of these small molecules to permeate cell
membranes and covalently modify a disparate set of protein
targets (Figure 2A and B, and Figures S1, S2, and S3 in
Supporting Information). As demonstrated for the CA
electrophile (Figure 2A), varying the R3 group on CA-
containing triazine molecules results in varied protein-labeling
patterns. Subtle changes such as a m-methoxybenzyl versus a p-
fluorobenzyl (RB-2-ca vs RB-3-ca; Figure 2A) resulted in a
noticeable change in the protein targets, serving to exemplify
the target diversity we can access through minor changes in R3
structure. In addition to changes in the diversity element, we
can also modulate the protein targets by varying the
electrophile. For three R3 groups, we illustrate that varying
the electrophile results in significant diversification of the
protein targets (Figure 2B). These results highlight the
modularity of the trifunctionalized triazine scaffold, which
allows for varied targeting of proteins driven by modifications
to either diversity element or electrophile structure.
We followed up on target identification and mechanistic
characterization for two of the compounds in our initial library,
RB-2-cb and RB-11-ca, which demonstrated high selectivity
toward a single protein target (Figure 2C). We proceeded to
identify the targeted proteins using mass spectrometry. A biotin
group was incorporated onto labeled proteins using click
chemistry, followed by avidin enrichment and subsequent
identification using LC/LC−MS/MS.17 Spectral counts for
proteins identified in RB-2-cb- and RB-11-ca-treated proteomes
were compared to a DMSO-treated control to identify the
direct targets of these compounds (Table S2 and S2,
Supporting Information). These mass-spectrometry studies
identified the ∼50 kD target of RB-2-cb as a tubulin β chain
Figure 1. 1,3,5-Triazines provide a modular scaffold for covalent
inhibitor design. (A) Trifunctionalization of the triazine scaffold to
present an alkyne (R1), cysteine-reactive electrophile (R2) and a
diversity element (R3). (B) R1, R2 and R3 groups for the 20
compounds evaluated in this study. (C). Structures of the diversity
elements (R3) incorporated in this study.
properties to direct the reactivity of these small molecules
toward distinct subsets of proteins (Figure 1C).
Figure 2. Protein targets of the triazine-based compounds. Library members were incubated with HeLa cells, and protein labeling by each compound
was evaluated after cell lysis, click-chemistry-based incorporation of rhodamine, SDS-PAGE and in-gel fluorescence. (A). CA-containing probes with
seven different R3 groups demonstrate that modifications to the diversity element results in variations of the protein targets. (B) CA, CNB and Acr
electrophiles demonstrated diverse labeling patterns even though the same R3 group was employed. (C) In-gel fluorescence studies of RB-2-cb and
RB-11-ca show labeling of one-major protein identified using LC/LC−MS/MS as TUBB and PDIA1 respectively.
B
dx.doi.org/10.1021/ja400427e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
and the ∼57 kD target of RB-11-ca as protein disulfide
isomerase (PDI).
To verify the protein targets identified by mass spectrometry
and to determine the specific cysteine on each of these proteins
that is covalently modified by these compounds, we overex-
pressed isoforms of tubulin and PDI (TUBB and PDIA1) with
a C-terminal Myc/His tag using transient transfection in
HEK293T cells. We confirmed labeling of the wild-type
proteins by RB-2-cb and RB-11-ca (Figure 3A and B). The
Figure 4. Effects of RB-2-cb and RB-11-ca on protein activity and cell
proliferation. (A) Tubulin polymerization assay comparing the effects
of RB-2-cb with colchicine. (B) Effects of RB-2-cb on cell proliferation
of HeLa cells. (C) PDI activity assay comparing RB-11-ca to the
commercially available PDI inhibitor, 16F16. (D) Effects of RB-11-ca
on HeLa cell proliferation.
Figure 3. Target verification and identification of site of labeling. Wild-
type and mutant TUBB (A) and PDIA1 (B) were overexpressed and
labeled with RB-2-cb and RB-11-ca, respectively, to identify C239 and
C53 as the sites of modification (top panel). The lower band * is the
endogenous protein, and the upper band represents the overexpressed
Myc/His-tagged variant. Western blotting with an α-Myc antibody
verified equal protein expression for wild-type and mutant proteins
(bottom panel).
two irreversible inhibitors of PDI, 16F16 and PACMA 31, have
been shown to suppress apoptosis in a model of Huntington
disease,30 and decrease ovarian tumor growth,31 respectively.
These previous studies underscore the importance of
developing selective small-molecule inhibitors for PDI. To
study the effect of RB-11-ca on PDI activity, we performed an
in vitro fluorescence-based activity assay and compared RB-11-
ca to the commercially available PDI inhibitor 16F16 (Figure
4c).30 RB-11-ca demonstrated comparable inhibition to 16F16,
adding to the arsenal of small molecules available to perturb
PDI activity in vivo. Given the intriguing selectivity of RB-11-ca
for one of the PDI active sites, this small molecule could
provide insight into the regulation and activity of each of the
two PDI active sites. Furthermore, given the report that
targeted PDI therapy was effective in ovarian cancer cells,31 we
sought to evaluate the effect of RB-11-ca on HeLa cell
proliferation and demonstrate that our compound shows
micromolar inhibition (23.9 μM) of cell proliferation (Figure
4D).
In summary, we explore the trifunctionalized 1,3,5-triazine as
an ideal modular scaffold for development of libraries of
covalent inhibitors with streamlined target identification. By
modifying either the diversity element or the electrophile in the
triazine, we can modulate the protein targets of our small-
molecule library. Using a preliminary set of 20 compounds, we
demonstrate the modularity, cell permeability and target
diversity of these triazine-based compounds. Furthermore, we
identified highly selective covalent modifiers for β-tubulin and
protein disulfide isomerase, with comparable efficacy to existing
inhibitors of tubulin polymerization and PDI activity,
respectively. These compounds add to the armory of small-
molecule probes available for these protein activities, and
furthermore exemplify the potential of similar libraries directed
toward cysteines or other reactive amino acids to yield selective
pharmacological agents. Such agents will hold great promise in
chemical genetic screens to identify novel bioactive agents. This
triazine-based platform is an attractive alternative to traditional
small-molecule libraries due to the ease of target identification
that is built into the library. Our preliminary studies lay the
groundwork to support the future generation of larger triazine-
lower band in both gels (*) represents the endogenous protein,
whereas the upper band denotes the overexpressed variant. In
addition to the wild-type proteins, we also overexpressed
cysteine to alanine mutants of several known reactive cysteines
on these proteins.6 This gel-based analysis of wild-type and
mutant proteins confirmed that RB-2-cb targets Cys239 on
TUBB and RB-11-ca targets Cys53 of PDIA1.
Microtubule disruptors that function through inhibition of
tubulin polymerization hold great promise as chemotherapeutic
agents.18 These microtubule-targeting agents include tubulin
stabilizers (e.g., taxanes and epothilones) as well as tubulin
depolymerizers (e.g., colchicine).18 β-Tubulin is a major
component of microtubules and Cys239 in β-tubulin has
been shown to be targeted by electrophilic small molecules and
nitrosylating agents, leading to inhibition of tubulin polymer-
ization.19−22 To assess the effect of Cys239 modification by RB-
2-cb, we used an in vitro assay to monitor tubulin polymer-
ization in the presence of our compound.23 We compared the
effect of RB-2-cb to colchicine,24 a natural product known to
disrupt microtubule formation by reversible binding to β-
tubulin (Figure 4A). Similar to colchicine, RB-2-cb is a potent
inhibitor of tubulin polymerization in vitro, with a >60%
inhibitory effect at 2 μM. Furthermore, given that colchicine is
a potent inhibitor of mitotic growth in cancer cells, we
examined the ability of RB-2-cb to inhibit the growth of HeLa
cells using an MTT cell viability assay. As expected, RB-2-cb is a
potent inhibitor of cell growth with an EC50 value of 2.7 μM
(Figure 4B).
PDIs are primarily localized to the endoplasmic reticulum,
and act as chaperone proteins that catalyze the isomerization
and rearrangement of disulfide bonds.25 Increased PDI levels
have been documented in various cancers, including ovarian,26
lung27 and prostate28 cancers. PDIs contain two active sites for
redox catalysis, denoted as α and α′ domains.29 Each active site
contains a pair of cysteine residues within a CGHC motif
(Cys53/Cys56 and Cys397/Cys400 in PDIA1).29 Recently,
C
dx.doi.org/10.1021/ja400427e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(20) Bai, R. L.; Lin, C. M.; Nguyen, N. Y.; Liu, T. Y.; Hamel, E.
Biochemistry 1989, 28, 5606.
(21) Mi, L.; Xiao, Z.; Hood, B. L.; Dakshanamurthy, S.; Wang, X.;
Govind, S.; Conrads, T. P.; Veenstra, T. D.; Chung, F. L. J. Biol. Chem.
2008, 283, 22136.
(22) Landino, L. M.; Koumas, M. T.; Mason, C. E.; Alston, J. A.
Chem. Res. Toxicol. 2007, 20, 1693.
(23) Li, J.; Risinger, A. L.; Peng, J.; Chen, Z.; Hu, L.; Mooberry, S. L.
J. Am. Chem. Soc. 2011, 133, 19064.
(24) Bhattacharyya, B.; Panda, D.; Gupta, S.; Banerjee, M. Med. Res.
Rev. 2008, 28, 155.
based covalent libraries for cell-lysate or phenotypic screens
with the goal of identifying selective covalent inhibitors for
diverse proteins.
ASSOCIATED CONTENT
■
S
* Supporting Information
Supplementary figures, experimental protocol, synthetic scheme
and compound characterization. This material is available free
of charge via the Internet at http://pubs.acs.org.
(25) Hatahet, F.; Ruddock, L. W. Antioxid. Redox Signal. 2009, 11,
2807.
AUTHOR INFORMATION
■
(26) Bonome, T.; Levine, D. A.; Shih, J.; Randonovich, M.; Pise-
Masison, C. A.; Bogomolniy, F.; Ozbun, L.; Brady, J.; Barrett, J. C.;
Boyd, J.; Birrer, M. J. Cancer Res. 2008, 68, 5478.
(27) Beer, D. G.; Kardia, S. L.; Huang, C. C.; Giordano, T. J.; Levin,
A. M.; Misek, D. E.; Lin, L.; Chen, G.; Gharib, T. G.; Thomas, D. G.;
Lizyness, M. L.; Kuick, R.; Hayasaka, S.; Taylor, J. M.; Iannettoni, M.
D.; Orringer, M. B.; Hanash, S. Nat. Med. 2002, 8, 816.
(28) Welsh, J. B.; Sapinoso, L. M.; Su, A. I.; Kern, S. G.; Wang-
Rodriguez, J.; Moskaluk, C. A.; Frierson, H. F., Jr.; Hampton, G. M.
Cancer Res. 2001, 61, 5974.
Corresponding Author
eranthie@bc.edu
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
E. W. is a Damon Runyon-Rachleff Innovator supported (in
part) by the Damon Runyon Cancer Research Foundation
(DRR-18-12). We are also grateful for financial support from
the Smith Family Foundation and Boston College. We thank
members of the Weerapana Lab for comments and critical
reading of the manuscript.
(29) Gruber, C. W.; Cemazar, M.; Heras, B.; Martin, J. L.; Craik, D. J.
Trends Biochem. Sci. 2006, 31, 455.
(30) Hoffstrom, B. G.; Kaplan, A.; Letso, R.; Schmid, R. S.; Turmel,
G. J.; Lo, D. C.; Stockwell, B. R. Nat. Chem. Biol. 2010, 6, 900.
(31) Xu, S.; Butkevich, A. N.; Yamada, R.; Zhou, Y.; Debnath, B.;
Duncan, R.; Zandi, E.; Petasis, N. A.; Neamati, N. Proc. Natl. Acad. Sci.
U.S.A. 2012, 109, 16348.
REFERENCES
■
(1) Paolini, G. V.; Shapland, R. H.; van Hoorn, W. P.; Mason, J. S.;
Hopkins, A. L. Nat. Biotechnol. 2006, 24, 805.
(2) Barf, T.; Kaptein, A. J. Med. Chem. 2012, 55, 6243.
(3) Patricelli, M. P.; Szardenings, A. K.; Liyanage, M.; Nomanbhoy,
T. K.; Wu, M.; Weissig, H.; Aban, A.; Chun, D.; Tanner, S.; Kozarich,
J. W. Biochemistry 2007, 46, 350.
(4) Pace, N. J.; Weerapana, E. ACS Chem. Biol. 2012, DOI: 10.1021/
cb3005269.
(5) Weerapana, E.; Simon, G. M.; Cravatt, B. F. Nat. Chem. Biol.
2008, 4, 405.
(6) Weerapana, E.; Wang, C.; Simon, G. M.; Richter, F.; Khare, S.;
Dillon, M. B.; Bachovchin, D. A.; Mowen, K.; Baker, D.; Cravatt, B. F.
Nature 2010, 468, 790.
(7) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B.
Angew. Chem., Int. Ed. 2002, 41, 2596.
(8) Leesnitzer, L. M.; Parks, D. J.; Bledsoe, R. K.; Cobb, J. E.; Collins,
J. L.; Consler, T. G.; Davis, R. G.; Hull-Ryde, E. A.; Lenhard, J. M.;
Patel, L.; Plunket, K. D.; Shenk, J. L.; Stimmel, J. B.; Therapontos, C.;
Willson, T. M.; Blanchard, S. G. Biochemistry 2002, 41, 6640.
(9) Sechi, S.; Chait, B. T. Anal. Chem. 1998, 70, 5150.
(10) Walsh, D. P.; Chang, Y. T. Chem. Rev. 2006, 106, 2476.
(11) Zhou, C.; Min, J.; Liu, Z.; Young, A.; Deshazer, H.; Gao, T.;
Chang, Y. T.; Kallenbach, N. R. Bioorg. Med. Chem. Lett. 2008, 18,
1308.
(12) Ni-Komatsu, L.; Leung, J. K.; Williams, D.; Min, J.; Khersonsky,
S. M.; Chang, Y. T.; Orlow, S. J. Pigm. Cell Res. 2005, 18, 447.
(13) Uttamchandani, M.; Walsh, D. P.; Khersonsky, S. M.; Huang,
X.; Yao, S. Q.; Chang, Y. T. J. Comb. Chem. 2004, 6, 862.
(14) Khersonsky, S. M.; Jung, D. W.; Kang, T. W.; Walsh, D. P.;
Moon, H. S.; Jo, H.; Jacobson, E. M.; Shetty, V.; Neubert, T. A.;
Chang, Y. T. J. Am. Chem. Soc. 2003, 125, 11804.
(15) Speers, A. E.; Adam, G. C.; Cravatt, B. F. J. Am. Chem. Soc. 2003,
125, 4686.
(16) Speers, A. E.; Cravatt, B. F. Chem. Biol. 2004, 11, 535.
(17) Speers, A. E.; Cravatt, B. F. Curr. Protoc. Chem. Biol. 2009, 1, 29.
(18) Hamel, E. Med. Res. Rev. 1996, 16, 207.
(19) Shan, B.; Medina, J. C.; Santha, E.; Frankmoelle, W. P.; Chou,
T. C.; Learned, R. M.; Narbut, M. R.; Stott, D.; Wu, P.; Jaen, J. C.;
Rosen, T.; Timmermans, P. B.; Beckmann, H. Proc. Natl. Acad. Sci.
U.S.A. 1999, 96, 5686.
D
dx.doi.org/10.1021/ja400427e | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX