Strategic Design of Catalytic Lysine-Targeting Reversible Covalent BCR-ABL Inhibitors**

Article abstract of DOI:10.1002/anie.202105383

Targeted covalent inhibitors have re-emerged as validated drugs to overcome acquired resistance in cancer treatment. Herein, by using a carbonyl boronic acid (CBA) warhead, we report the structure-based design of BCR-ABL inhibitors via reversible covalent targeting of the catalytic lysine with improved potency against both wild-type and mutant ABL kinases, especially ABL^T315I bearing the gatekeeper residue mutation. We show the evolutionarily conserved lysine can be targeted selectively, and the selectivity depends largely on molecular recognition of the non-covalent pharmacophore in this class of inhibitors, probably due to the moderate reactivity of the warhead. We report the first co-crystal structures of covalent inhibitor-ABL kinase domain complexes, providing insights into the interaction of this warhead with the catalytic lysine. We also employed label-free mass spectrometry to evaluate off-targets of our compounds at proteome-wide level in different mammalian cells.

Full text of DOI:10.1002/anie.202105383

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Strategic Design of Catalytic Lysine-Targeting Reversible
Covalent BCR-ABL Inhibitors
Authors: David Quach, Guanghui Tang, Jothi Anantharajan, Nithya
Baburajendran, Anders Poulsen, John Wee, Priya Retna,
Rong Li, Boping Liu, Doris Tee, Perlyn Kwek, Joma Joy,
Wang-Qi Yang, Chong-Jing Zhang, Klement Foo, Thomas
Keller, and Shao Q. Yao
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202105383
Link to VoR: https://doi.org/10.1002/anie.202105383

10.1002/anie.202105383
Angewandte Chemie International Edition
RESEARCH ARTICLE
Strategic Design of Catalytic Lysine-Targeting Reversible Covalent
BCR-ABL Inhibitors
David Quach,[a][b] Guanghui Tang,[c] Jothi Anantharajan,[b] Nithya Baburajendran,[b] Anders
Poulsen,[b] John L. K. Wee,[b] Priya Retna,[b] Rong Li,[b] Boping Liu,[b] Doris H. Y. Tee,[b] Perlyn
Z. Kwek,[b] Joma K. Joy,[b] Wan-Qi Yang,[d] Chong-Jing Zhang,[d] Klement Foo,*[b] Thomas H.
Keller,*[b] Shao Q. Yao*[a][c]
[a]
Dr. D. Quach, Prof. Dr. S. Q. Yao
NUS Graduate School for Integrative Sciences and Engineering, 21 Lower Kent Ridge, University Hall, Tan China Tuan Wing, #04-02, Singapore 119077
E-mail: chmyaosq@nus.edu.sg
[b]
Dr. D. Quach, Dr. J. Anantharajan, Dr. N. Baburajendran, Dr. A. Poulsen, J. L. K. Wee, P. Retna, R. Li, B. Liu, D. H. Y. Tee, P. Z. Kwek, Dr. J. K. Joy, Dr.
K. Foo, Dr. T. H. Keller
Experimental Drug Development Centre, 10 Biopolis Road, Chromos, #05-01, Singapore 138670
E-mail: thkeller@eddc.a-star.edu.sg, klementfoo@gmail.com
[c]
[d]
Dr. G. Tang, Prof. Dr. S. Q. Yao
Department of Chemistry, National University of Singapore, Singapore 117543
W-.Q. Yang, Dr. C-.J. Zhang
State Key Laboratory of Bioactive Substances and Functions of Natural Medicines and Beijing Key Laboratory of Active Substances Discovery and
Druggability Evaluation, Institute of Materia Medica Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing, China, 100050
Supporting information for this article is given via a link at the end of the document.
Abstract: Targeted covalent inhibitors have re-emerged
as validated drugs to overcome acquired resistance in
cancer treatment. Herein, by using a carbonyl boronic
acid (CBA) warhead, we report the structure-based
design of BCR-ABL inhibitors via reversible covalent
targeting of the catalytic lysine with improved potency
against both wild-type and mutant ABL kinases,
especially ABLT315I bearing the gatekeeper residue
mutation. We show the evolutionarily conserved lysine
can be targeted selectively, and the selectivity depends
largely on molecular recognition of the non-covalent
pharmacophore in this class of inhibitors, probably due
to the moderate reactivity of the warhead. We report the
first co-crystal structures of covalent inhibitor-ABL
kinase domain complexes, providing insights into the
interaction of this warhead with the catalytic lysine. We
also employed label-free mass spectrometry to evaluate
off-targets of our compounds at proteome-wide level in
different mammalian cells.
Introduction
Chronic myeloid leukemia (CML) arises from a genetic
abnormality in human chromosome 22, which is
unusually short and defective because of the reciprocal
translocation of genetic material from chromosome 9.[1]
Figure 1. (A) A previously reported sulfonyl fluoride probe for
covalent lysine labeling of kinases in live cells.[22] (B) Lysine
modification based on the formation of a stable iminoboronate with
2-formylbenzene boronic acid.[30] (C) Model study of various
lysine-targeting probes by using bovine serum albumin (BSA). (D)
Structure-based design of a BCR-ABL inhibitor (12) targeting the
catalytic lysine K271, based on the previously reported inhibitor
PPY-A (7).[32]
Gene expression leads to the formation of
a
constitutively active BCR-ABL1 kinase, which
aberrantly activates multiple signaling pathways that
bring about uncontrollable cell growth and
differentiation.[2] Despite the clinical success of ATP-
competitive inhibitors,[3] a significant number of patients
have suffered from relapse due to drug resistance,
which can arise from point mutations that severely
reduce the effect of such inhibitors.[4,5] An important
mutant in CML is BCR-ABLT315I which can only be
inhibited by ponatinib (Iclusig).[6] Targeted covalent
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202105383
Angewandte Chemie International Edition
RESEARCH ARTICLE
Scheme 1. Synthesis of key inhibitors used in this work. (a) NIS, acetone, rt, 1 h, 98%; (b) TsCl, NaH, THF, 0 C, 3 h, 96%; (c) Pd(dppf)Cl2,
K2CO3, 1,4-dioxane/H2O (9:1), 70 C, 7 h, 60-80%; (d) B2pin2, Pd(dppf)Cl2, KOAc, 1,4-dioxane, 100 C, 7 h, 70-90%; (e) 5-bromo-N,N-
dimethylnicotinamide, Pd(dppf)Cl2, K2CO3, 1,4-dioxane/H2O (9:1), 95 C, 7 h, 70-80%; (f) LiOH, MeOH/dioxane/H2O (2:2:1), rt, 1 h, 80-90%;
(g) CF3SO2Cl, K2CO3, DMF, rt, 2-4 h, 80-90%; (h) B2pin2, Pd(dppf)Cl2, KOAc, 1,4-dioxane, 100 C, 7 h; (i) Cs2CO3, THF/MeOH (2:1), 40-50
C, 1 h, 10-20% (2-step yields from (h) and (i)); (j) 5-bromo-N-methyl-N-(3-(triisopropylsilyl)prop-2-yn-1-yl)nicotinamide, Pd(dppf)Cl2, K2CO3,
1,4-dioxane/H2O (9:1), 95 C, 7 h, 80-90%; (k) TBAF, THF/H2O, rt, 2.5% (3-step yield from (h), (i) and (k)).
inhibitors offer advantages such as greater potency and toxicity is a major concern for both tool compounds and
prolonged duration of action over non-covalent drug candidates, many research groups aspire to tune
inhibitors, and they have re-emerged in recent years as the reactivity of covalent warheads.[24-28] For example,
demonstrated by the clinical success of, for example, Taunton and co-workers used electron-deficient
osimertinib.[7-11] The standard strategy to design Michael acceptors for the design of cysteine-targeting
irreversible kinase inhibitors uses Michael acceptors to reversible covalent kinase inhibitors.[29] Carbonyl
target poorly conserved cysteine residues near the boronic acids (CBAs) have recently been shown to
active site of kinases.[12] This was thought to provide a reversibly but covalently modify amino groups in
measure of selectivity; however, Cravatt and coworkers proteins (Figure 1B);[30] however, this chemistry, to the
have demonstrated that even carefully designed best of our knowledge, has not been used to develop
cysteine-targeting covalent inhibitors have off-target covalent kinase inhibitors.[31] Our aim was therefore to
effects.[13] Furthermore, resistance mechanisms design reversible, covalent inhibitors targeting the
including cysteine point mutations (EGFRC797S and catalytic lysine residue in kinases as a general
BTKC481S) often render cysteine-targeting covalent approach to combat drug resistance. We report herein
drugs ineffective.[11] To our knowledge, no covalent the first successful examples of lysine-targeting
drugs against any of the known BCR-ABL mutants have reversible covalent kinase inhibitors based on CBAs
been reported up to date due to a lack of targetable and the corresponding iminoboronate chemistry (Figure
cysteine residues.[14]
1C/D); our results showed that such compounds
possessed potent and long-lasting inhibition against
BCR-ABL wild type and mutants by targeting the
catalytic lysine residue K271 in this kinase.
Since the catalytic lysine residue is essential for the
enzymatic activity of all protein kinases and therefore,
considered less prone to mutation,[15,16] we have been
interested to study this evolutionarily conserved residue
in the ATP pocket with the aim to produce covalent
kinase inhibitors as an alternative strategy in drug
design. Many non-selective lysine-modifying probes,
including the use of sulfonyl fluorides and activated
esters, have been reported.[17-23] Taunton and co-
workers recently reported sulfonyl fluoride-containing
probes for lysine-targeting in kinases (Figure 1A).[22]
Campos et al later identified lysine-targeting kinase
inhibitors that used an activated ester.[23] Since cellular
Results and Discussion
At the outset, a mechanistic study was carried out to
compare the relative reactivity of model probes 3 and 5
with an NHS probe (6) and other controls by using BSA
as reference in the presence of a rhodamine-azide
“click“ reporter (Figure 1C). Removal of either carbonyl
or boronic acid (or both; e.g. 1, 2 and 4) caused
complete abolishment in fluorescent labeling of BSA in
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202105383
Angewandte Chemie International Edition
RESEARCH ARTICLE
Table 1. Compounds 7-16 in the Biochemical Assay against ABLWT, ABLT315I and ABLE255K
.
IC (WT) (nM)
IC (T315I) (nM)
IC (E255K) (nM)
50
50
50
R1
R2
R3
Compound
T = 0 h
T = 0.25 h
2.0 ± 0.2
43 ± 5
T = 12 h
T = 0 h
T = 12 h
T = 0 h
T = 12 h
7
OMe
H
H
H
H
H
2.0 ± 0.1
2.0 ± 0.1
1.0 ± 0.1
1.0 ± 0.1
9.0 ± 0.5
10.0 ± 0.4
8
-
-
-
-
-
-
9
H
CHO
H
H
-
75 ± 3
-
-
-
-
-
10
11
12
13
14
15
16
H
CHO
B(OH)2
CHO
CHO
CHO
COCH3
CHO
7.0 ± 0.3
8.0 ± 0.1
414 ± 18
59 ± 1
-
-
-
-
-
H
CHO
B(OH)2
H
-
52 ± 2
5.0 ± 0.4
3.0 ± 0.1
1.7 ± 0.2
-
-
-
-
-
H
83 ± 2
-
-
-
-
OMe
OMe
OMe
OMe
3.0 ± 0.1
3.0 ± 0.2
12.0 ± 0.4
7.0 ± 0.3
10.0 ± 0.4
4.0 ± 0.2
3.0 ± 0.1
9.0 ± 0.1
8.0 ± 0.1
B(OH)2
H
13.0 ± 0.3
25 ± 1
0.50 ± 0.02
43 ± 2
0.50 ± 0.03
-
-
-
-
-
-
-
-
-
-
B(OH)2
-
3 and 5. Between 3 and 5, the former consistently
By using a mobility shift assay based on Caliper’s
labeled BSA more strongly, indicating the aldehyde was microfluidics capillary electrophoresis, the IC50 value of
more reactive than the ketone. As expected, both 3 and PPY-A (7) against wildtype ABL was determined to be
5 produced much weaker fluorescence labeling of BSA 2 nM after 15 minutes of incubation (Table 1, Figure S1).
compared to 6, suggesting CBAs have attenuated Removal of the o-methoxy group reduced the potency
lysine reactivity compared to the highly reactive NHS- by 20 folds (compound 8). Introduction of m-aldehyde
based irreversible lysine modifier.[18] We next designed (compound 10), however, improved the IC50 to 8.0 nM.
suitable CBA-containing BCR-ABL kinase inhibitors Shifting the aldehyde functional group to the para
(Figure 1D). Molecular modeling studies were first position was not tolerated (compound 9). Introduction of
carried out by incorporating a CBA warhead into the a boronic acid functionality led to inhibitors 11 and 12.
previously reported ABL1 inhibitor PPY-A (7) (PDB ID: Simialr to 9 and 10, 11 was about 10-fold less potent
2QOH);[32] results showed that introducing the required than 12 and therefore, the position of the aldehyde
functionality would not disrupt the binding to the protein, functional group played an important role in the
and with the distance between the proposed CBA inhibitory activity. As expected of covalent inhibitors, the
moiety and the highly flexible catalytic lysine K271 being enzyme inhibition of compound 12 improved from 83 nM
~3.5 Å, formation of an iminoboronate in the (T = 0 h) to 5.0 nM (T = 12 h) as the incubation time was
kinase/inhibitor complex was indeed possible.
increased (Figure S2). Compound 10 did not show time-
dependence of the IC50, suggesting that the presence
of the boronic acid in 12 may have led to the formation
of an iminoboronate.[33]
A library of analogs was synthesized in order to
establish structure-activity relationship (Scheme 1,
Table 1). The synthesis of a common intermediate 24
was done in two steps via iodination and tosylation of 5-
To determine whether 12 was selective towards the
bromoazaindole. 7-10 were obtained in four steps via catalytic lysine in ABL, mass spectrometric analysis was
Suzuki-Miyaura coupling reactions involving 24 to performed. MALDI-TOF analysis suggested that a
generate 26a-d, which underwent Miyaura borylation single lysine-modified covalent adduct was formed with
that led to 27a-d. The Suzuki-Miyaura cross coupling an observed m/z of 33156.38 Da (Figure 2A bottom;
compared to calculated mass of 33156.59 Da). The
reaction did not reach completion regardless of the
concentration (up to 1 mM) and incubation time (up to
24 h) of 12, indicating that an equilibrium was
reaction was performed at a lower temperature of 60 C
in order to chemoselectively differentiate Br and I. 27a-
d then underwent a second Suzuki-Miyaura coupling
reaction
with
5-bromo-N,N-dimethylnicotinamide
established.[30,31]
On the other hand, incubating the
followed by subsequent removal of tosyl group with
LiOH, leading to the formation of the desired inhibitors
(7-10). The synthesis of CBA inhibitors 11 and 12
occured via a different route in which an additional step
involving the conversion of OH to OTf, led to
intermediates 34a and 34b; CF3SO2Cl and K2CO3 were
shown to be the optimal choice. Other electrophiles
such as PhNTf2 and (CF3SO2)2O in the presence of
bases such as triethylamine, NaH and pyridine did not
work well despite heating. 34a and 34b underwent
borylation and subsequent deprotection of the tosyl
group by using Cs2CO3 was carried out to generate
covalent inhibitors 11 and 12.
ABL kinase domain with 10 of the same concentration
did not lead to the formation of any detectable adduct,
indicating again that the imine formation between 10
and ABL kinase domain, in the absence of boronic,
cannot be detected in our experimental setup compared
to a control using DMSO (Figure 2A top & Figure S3).
Since the OMe group in 7 was important for ABL
inhibition, the same functionality was added to 12,
providing compound 14 which was synthesized via a
similar route (Scheme 1). As expected, 14 showed time-
dependent IC50 values against ABL from 13 nM at 0 h to
1.7 nM after 12 h (Table 1, Figure S2); further testing of
7 and 14 against two ABL mutants indicated that 14
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202105383
Angewandte Chemie International Edition
RESEARCH ARTICLE
Figure 2. (A) Mass spectrometric analysis of the ABL kinase domain-12 complex (bottom). (top): ABL kinase domain-10 complex as control.
(B) Cocrystal structures of the ABL kinase domain with 12 (left; PDB ID: 7CC2) and 14 (right; PDB ID: 7DT2) showing the composite omit map
(2Fo-Fc) contoured to 1휎 in green mesh. The front loop was removed for better visibility. (C) Comparison of ABL inhibition (WT and mutants) of
PPY-A (7), 12 and 14 after 12 h. See Table 1 and Figure S2 for details. (D) Modeled structure of ABLT315I kinase domain-14 complex based on
the obtained cocrystal structure of ABLWT kinase domain-14 complex. (E) Dendrograms showing Kinome Scan™ of 7 (left) and 14 (right) at
1000 nM against 90 different kinases. (F) Anti-proliferative activity of 7, 13, 14 and 15 against K562 cells determined by CellTiter-Glo® viability
assay.
showed time-dependent inhibition, with improved however, observe the formation of the expected dative
potency upon prolonged incubation, against both bond between the imine nitrogen and the boron atom.
ABLT315I and ABLE255K when compared to 7 (Table 1, In fact, the lone pair of the imine appeared to be
Figures 2C). In particular, 14 demonstrated an improved orientated away from the boron atom in both crystal
potency from 25 nM (T = 0 h) to 0.50 nM (T = 12 h) structures. One of the reasons could be due to the vast
against ABLT315I (50-fold improvement) and from 43 nM structural difference between free small molecules and
(T = 0 h) to 0.5 nM (T = 12 h) against ABLE255K (100-fold the macromolecule-inhibitor complex formed within the
improvement). The effect of point mutation on restricting tight binding pocket of a protein target. The latter might
access to the binding pocket or stabilizing certain have significantly influenced the orientation of the
protein conformations has been shown to adversely bound small molecule. Indeed, in a model experiment
affect drug binding which tends to favor only a very (Figure S8), we were able to show that an
specific target conformation;[10] however, our examples iminoboronate adduct could be successfully captured
showed that targeting the catalytic lysine residues gave by 11B NMR. Since the biochemical assays and the
an advantage in this context.
MALDI-TOF analysis suggested that the boronic acid
plays an important role in the formation of the adduct,
we propose that the obtained cocrystal structures had
successfully captured the key intermediate during the
formation of the iminoboronates. Given the fact that 14
was able to potently inhibit ABLT315I while imatinib and
second-generation kinase inhibitors such as nilotinib,
dasatinib failed to do so,[32] we next rationalized this
observation by using the newly obtained structural data
(Figure 2D); by building a modeled structural complex
The X-ray cocrystal structures of 12 and 14 with the
ABL kinase domain (229-510) were solved up to 2.7 Å
and 2.3 Å resolution, respectively, giving more insights
into how this particular warhead interacts with the
catalytic lysine residue (Figures 2B & S4-7; PDB IDs:
7CC2 and 7DT2); the continuous electron density from
K271 in the ABL kinase domain to 12 and 14 suggests
that the imine product was formed. We did not,
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202105383
Angewandte Chemie International Edition
RESEARCH ARTICLE
of 14 and ABL kinase domain in which the T315 residue bonded interaction of 12 and 14 with the kinase (Figures
was artificially changed to I315, we observed no steric S9-10).[34-36] Both known irreversible and slow-binding
clash upon inhibitor binding. Imatinib, nilotinib and reversible models were used to fit our inhibition data,
dasatinib possess crucial elongated structures that are and results indicated that the latter was the more
extended to the back cleft of the ATP binding pocket in suitable one, truly reflecting the slow binding modes of
ABL, and T315I point mutation was expected to restrict our potent inhibitors. 14 had a smaller KI compared to
access to the hydrophobic region at the rear of this that of 12, thus confirming the critical role of OMe in 14
pocket.[32] This is not the case for 14 which does not for increased affinity between the inhibitor and the ABL
have any substituent that occupies this pocket. Finally, kinase domain. The conclusion was further
14 demonstrated better biochemical activity than that of strengthened by time-dependent IC50 biochemical
7 due to apparent covalent modification (Figure 2C),[34] assays in which the incubation time against ABLWT was
and the incorporation of both boronic acid and aldehyde fixed at 15 min for 7, 12 and 14 (Table 1); 14 (IC50 = 12
did not appear to cause any steric clash with the I315 nM) was shown to be about 5-fold more potent than 12
residue.
(IC50 = 59 nM).
The kinetic parameters of binding and inactivation
were next determined to better understand the non-
Figure 3. (A) Proteome reactivity profiles of compound 16 compared to 3, 6 and 17 in K562 cell lysates (in PBS with 0.1% Triton, pH 7.5). (Top):
in-gel fluorescence scanning; (Bottom): Coomassie staining (CBB). (B) Concentration-dependent proteomic reactivity profiles of 16 (0-100 휇M)
in Ba/F3 cells overexpressing BCR-ABLWT. (C) Effect of washing by cold acetone and cold methanol (lane 4, highlighted in red; lane 3: no
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202105383
Angewandte Chemie International Edition
RESEARCH ARTICLE
washing) of probe-labeled K562 (see Figure S13 for corresponding Ba/F3 results) lysates in the absence and presence of excess NaBH3CN
(top); (bottom) Western blotting (WB) detection of BCR-ABLWT in lysates of both K562 and Ba/F3 cells following pull-down (PD) assays. GAPDH
= loading control. [compound 16] = 10 µM, and [compound 14] = 500 µM. Incubation time = 1 h. (D/E) Scatter plots showing potential cellular
targets of 16 from probe-treated K562 and Ba/F3 cell lysates, respectively. x-axis shows “Enrichment ratio“ between 16- and DMSO-treated
samples. y-axis shows “Competition ratio“ between 16- and 14-treated samples. Results were further filtered to provide ’high-confidence
hits“ (right graphs). (F) Data analysis of shared targets identified from chemoproteomic experiments in (D/E). The 40 shared “high-confidence
hits“ were highlighted in red and further analyzed. See Supporting Information for details. (G) Venn diagram showing the shared 523 targets
(including 18 kinases) identified in both K562 and Ba/F3 cells.
One of the potential concerns for covalent inhibitors increase in the fluorescence intensity of the 16-labeled
like 14, which target a key amino acid in the active site proteome was observed. In contrast, in the absence of
of kinases, is the selectivity. We therefore carried out NaBH3CN, washing the 16-labeled proteome with cold
Kinome Scan™ with a panel of ~100 protein kinases acetone and methanol significantly reduced or
(Figure 2E); results showed highly similar interaction abolished the labeling (compare lanes 3 and 4). This
maps for compounds 7 (left) and 14 (right), suggesting thus confirms the reversibility of iminoboronate bond.
that
the
degree
of
compound
selectivity Further evidence of this reversibility was obtained by
1
largely depended on the initial step of molecular using H NMR upon dilution with a model complex
recognition presumably due to the moderately reactive between 2-formylphenyl boronic acid and Ac-Lys-NHMe
2-carbonyl boronic acid warhead in 14. The conclusion (Figure S12). Upon further enrichment of the labeled
was further strengthened by the selectivity scores proteomes by pull-down (PD) experiments followed by
(Table S4). To evaluate the anti-proliferative activity of Western blotting (WB) analysis, we confirmed
this class of inhibitors, compounds 7 and 14, as well as successful labeling of endogenous BCR-ABTWT and
reference compounds 13 and 15 (boronic acid-free BCR-ABLT315I from both K562 and Ba/F3 cell lysates
versions of 14; see Table 1), were tested in K562 cells. (Figure 3C bottom, Figure S11B). By using DMSO-
As shown in Figure 2F, compound 7 showed GI50 of treated and 14-competed samples as controls, we next
0.114 휇M whereas compounds 13, 14 and 15 were less carried out large-scale LC-MS analysis (Figures 3D-G,
S13-14) to identify potential off-targets of 16.[22,38,39]
Consistent with strong labeling shown by in-gel
fluorescence scanning (Figure 3C), probe 16 captured
a number of off-targets in addition to the expected BCR-
ABL (highlighted in red circles in Figure 3D/E). By
setting stringent criteria to remove non-specific binders
(i.e. those shown in CRAPome;[42] See supporting
Information for details) and identify only “high-
confidence hits” (right graphs in Figure 3D/E), we
sucessfully identified 40 putative targets of 14 (Figure
3F); upon further analysis, ABL1 emerged as the only
kinase identified from our experiments in both cell lines.
These results are consistent with our earlier Kinome
Scan™ data of 14 (e.g. Figure 2E) and indicate its
alkyne-containing analog 16 was a selective ABL-
targeting probe. Interestingly in K562 but not in Ba/F3
cells, three additional kinases were identified as ”high-
confidence hits” (shown in green circles, Figure 3D). By
lowering our data filter criteria to include all shared
targets identified in both K562 and Ba/F3 cells (523 in
total; see Figure 3G), including those that appear in
CRAPome, we were able to further identify additional
potential kinase off-targets (Figure S13).
active with GI50 values of 3.85 휇M, 1.19 휇M and 0.384
휇M, respectively. In an attempt to understand the loss
of activity, the cell permeability of three compounds
were tested (Tables S5 and S6); Compound 13 showed
poor recovery rates which could be due to metabolism
of the aldehyde functional group for this particular
pharmacophore.[37] The 10-fold improvement in the anti-
proliferative activity of its ketone counterpart (e.g. 15;
GI50 = 0.384 휇M) suggests that the aldehyde might not
be the optimal choice for this scaffold in the cellular
systems, and a ketone could be a more appropriate
option in future studies.
Finally, to evaluate the proteome-wide reactivity and
potential off-targets of 14 in live mammalian cells, we
synthesized its alkyne-containing analog (16; Scheme
1) and carried out large-scale chemoproteomic studies
(Figures 3, S11).[38] We first compared the proteome
reactivity of the aldehyde-boronic acid moiety to other
well-known lysine-targeting functionalities by using
lysates from K562 cells (Figure 3A);[17,18] in-gel
fluorescence scanning analysis of probe-labeled
lysates, followed by CuAAC with a rhodamine azide,[39]
showed that 3 demonstrated better selectivity at both 1
휇M and 10 휇M when compared to 6 and 17 which are
known lysine-reactive electrophiles, but was predictably
less selective compared to the kinase-targeting 16,
suggesting that molecular recognition was the dominant
factor for similar compounds that contain low-reactivity
electrophiles such as CBAs.[40,41] In 16-treated samples,
we observed a concentration-dependent labeling of
proteomes, with saturated fluorescence signals at ~10
휇M of the probe (Figure 3B). We next repeated the
labeling experiment in the presence of NaBH3CN which
helped trap the reversible covalent iminoboronates into
more stable amine adducts (Figure 3C); a concomitant
Conclusion
In summary, we have successfully demonstrated, for
the first time, that the catalytic lysine of ABL can be
selectively targeted by inhibitors bearing a carbonyl
boronic acid moiety, leading to a reversible covalent
adduct. The incorporation of the two substituents
required for covalent bond formation reduced the affinity
to the ABL active site; however, the slow formation of
the iminoboronate led to highly potent inhibitors of ABL
kinase and its mutants. We also showed that the
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202105383
Angewandte Chemie International Edition
RESEARCH ARTICLE
R. Ferre, M. R. Lazear, M. M. Hayward, J. C. Kath, B. F. Cravatt,
Cell Chem. Bio. 2017, 24, 1388−1400.
[14] F. Rossari, F. Minutolo, E. Orciuolo, J. Hematol. Oncol. 2018,
11, 1−14.
[15] A. C. Carrera, K. Alexandrov, T. M. Roberts, Proc. Natl. Acad.
Sci. U. S. A. 1993, 90, 442−446.
[16] R. Liu, Z. Yue, C. C. Tsai, J. Shen, J. Am. Chem. Soc. 2019,
141, 6553−6560.
[17] D. A. Shannon, R. Banerjee, E. R. Webster, D. W. Bak, C. Wang,
E. Weerapana, J. Am. Chem. Soc. 2014, 136, 3330−3333.
[18] C. C. Ward, J. L. Kleinman, D. K. Nomura, ACS Chem. Biol.
2017, 12, 1478−1483.
[19] A. Cuesta, J. Taunton, Annu. Rev. Biochem. 2019, 88, 365−381.
[20] M. Gehringer, S. A. Laufer, J. Med. Chem. 2019, 62, 5673−5724.
[21] S. M. Hacker, K. M. Backus, M. R. Lazear, S. Forli, B. E. Correia,
B. F. Cravatt, Nat. Chem. 2017, 9, 1181−1190.
carbonyl boronic acid, which is a low-reactivity
electrophile, can be used to design highly selective
kinase inhibitors by maximizing molecular recognition.
Such compounds might be attractive tools for chemical
biology studies given recent interests in the
development of reversible covalent inhibitors,[29,31,43] but
could also serve as potential drug candidates in cases
where improved potency might be desirable once they
are fully optimized. We have also employed label-free
mass spectrometry to evaluate potential off-targets of
our compounds at proteome-wide level in different
mammalian cell lines. In addition to the expected target,
we also identified a few additional kinases as well as
some non-kinases as potnetial off-targets. Unlike
protein kinases which have a known catalytic lysine
residue in their kinase active sites, non-kinase targets
possess solvent-exposed lysine residues, thus
rendering them potentially susceptible to probe labeling.
[22] Q. Zhao, X. Ouyang, X. Wan, K. S. Gajiwala, J. C. Kath, L. H.
Jones, A. L. Burlingame, J. Taunton, J. Am. Chem. Soc. 2017,
139, 680−685.
[23] S. E. Dalton, L. Dittus, D. A. Thomas, M. A. Convery, J. Nunes,
J. T. Bush, J. P. Evans, T. Werner, M. Bantscheff, J. A. Murphy,
S. Campos, J. Chem. Am. Soc. 2018, 140, 932−939.
[24] U. P. Dahal, A. M. Gilbert, R. S. Obach, M. E. Flanagan, J. M.
Chen, C. Garcia-Irizarry, J. T. Starr, B. Schuff, D. P. Uccello, J.
A. Young, Med. Chem. Commun. 2016, 7, 864−872.
[25] J. Pettinger, K. Jones, M. D. Cheeseman, Angew. Chem. Int.
Ed. 2017, 56, 15200−15209.
Acknowledgements
Financial support was provided by the Synthetic
Biology Research & Development Programme (SBP) of
National Research Foundation (SBP-P4 and SBP-P8)
for Shao Q. Yao, the National Medical Research Council
(NMRC) via the Open Fund – Young Individual
Research Grant for Klement Foo and by the Agency for
Science, Technology and Research (A*STAR) via the
A*STAR Graduate Scholarship (AGS) for David Quach.
Financial support from CAMS Innovation Fund for
Medical Sciences (CIFMS) (2017-I2M-4−005) of China
is also acknowledged. The ABL protein construct and
glycerol stock were prepared by Yvonne Y. W. Tan and
Dario B. Heymann. We thank Zi Ye for support in MS
data analysis.
[26] P. Martín-Gago, C. A. Olsen, Angew. Chem. Int. Ed. 2019, 58,
957−966.
[27] L. H. Jones, Angew. Chem. Int. Ed. 2018, 57, 9220−9223.
[28] D. E. Mortenson, G. J. Brighty, L. Plate, G. Bare, W. Chen, S.
Li, H. Wang, B. F. Cravatt, S. Forli, E. T. Powers, K. B.
Sharpless, I. A. Wilson, J. W. Kelly, J. Am. Chem. Soc. 2018,
140, 200−210.
[29] I. M. Serafimova, M. A. Pufall, S. Krishnan, K. Duda, M. S.
Cohen, R. L. Maglathlin, J. M. McFarland, R. M. Miller, M. Frödin,
J. Taunton, Nat. Chem. Biol. 2012, 8, 471−476.
[30] S. Cambray, J. Gao, Acc. Chem. Res. 2018, 51, 2198-2206.
[31] G. Akçay, M. A. Belmonte, B. Aquila, C. Chuaqui, A.W. Hird, M.
L. Lamb, P. B. Rawlins, N. Su, S. Tentarelli, N. P. Grimster, Q.
Su, Nat. Chem. Biol. 2016, 12, 931−936.
[32] T. Zhou, L. Parillon, F. Li, Y. Wang, J. Keats, S. Lamore, Q. Xu,
W. Shakespeare, D. Dalgarno, X. Zhu, Chem. Biol. Drug Des.
2007, 70, 171−181.
Keywords: cancer • lysine • covalent • reversible • proteomics
[1]
[2]
R. Kurzrock, J. U. Gutterman, M. Talpaz, N. Engl. J. Med. 1988,
319, 990−998.
T. G. Lugo, A. M. Pendergast, A. J. Muller, O. N. Witte, Science
1990, 247, 1079−1082.
E. Jabbour, H. Kantarjian, Am. J. Hematol. 2016, 91, 252−265.
H. M. Kantarjian, M. Baccarani, E. Jabbour, G. Saglio, J. E.
Cortes, Clin. Cancer Res. 2011, 17, 1674−1683.
Z. Iqbal, A. Aleem, M. Iqbal, M. I. Naqvi, A. Gill, A. S. Taj, A.
Qayyum, N. ur-Rehman, A. M. Khalid, I. H. Shah, M. Khalid, R.
Haq, M. Khan, S. M. Baig, A. Jamil, M. N. Abbas, M. Absar, A.
Mahmood, M. Rasool, T. Akhtar, PLos One 2013, 8, e55717.
T. Anagnostou, M. R. Litzow, Blood Lymphat. Cancer 2017,
2018, 1−9.
[33] P. M. S. D. Cal, J. B. Vicente, E. Pires, A. V. Coelho, L. F. Veiros,
C. Cordeiro, P. M. P. Gois, J. Am. Chem. Soc. 2012, 134,
10299−10305.
[34] R. A. Bauer, Drug Discov. Today 2015, 20, 1061−1073.
[35] J. M. Bradshaw, J. M. McFarland, V. O. Paavilainen, A.
Bisconte, D. Tam, V. T. Phan, S. Romanov, D. Finkle, J. Shu,
V. Patel, T. Ton, X. Li, D. G. Loughhead, P. A. Nunn, D. E. Karr,
M. E. Gerritsen, J. O. Funk, T. D. Owens, E. Verner, K. A.
Brameld, R. J. Hill, D. M. Goldstein, J. Taunton, Nat. Chem. Biol.
2015, 11, 525−531.
[3]
[4]
[5]
[36] M. Goličnik, J. Stojan, Biochem. Mol. Biol. Educ. 2004, 32,
228−235.
[6]
[37] L. A. Admed, H. Younus, Drug Metab. Rev. 2019, 51, 42−64.
[38] B. R. Lanning, L. R. Whitby, M. M. Dix, J. Douhan, A. M. Gilbert,
E. C. Hett, T. O. Johnson, C. Joslyn, J. C. Kath, S. Niessen, L.
R. Roberts, M. E. Schnute, C. Wang, J. J. Hulce, B. Wei, L. O.
Whiteley, M. M. Hayward, B. F. Cravatt, Nat. Chem. Biol. 2014,
10, 760−767.
[7] D. S. Johnson, E. Weerapana, B. F. Cravatt, Future Med. Chem.
2010, 2, 949−964.
[8]
A. Chaikuad, P. Koch, S. A. Laufer, S. Knapp, Angew. Chem.
Int. Ed. 2018, 57, 4372−4385.
Z. Zhao, P. E. Bourne, Drug Discov. Today 2018, 23, 727−735.
[9]
[10] J. Engel, J. Lategahn, D. Rauh, ACS Med. Chem. Lett. 2016, 7,
2−5.
[39] H. Shi, C.-J. Zhang, G. Y. J. Chen, S. Q. Yao, J. Am. Chem.
Soc. 2012, 134, 3001−3014.
[11] S-H. I. Ou, Crit. Rev. Oncol. Hematol. 2012, 83, 407−421.
[12] Q. S. Liu, Y. Sabnis, Z. Zhao, T. H. Zhang, S. J. Buhrlage, L. H.
Jones, N. S. Gray, Chem. Biol. 2013, 20, 146−159.
[13] S. Niessen, M. M. Dix, S. Barbas, Z. E. Potter, S. Lu, O. Brodsky,
S. Planken, D. Behenna, C. Almaden, K. S. Gajiwala, K. Ryan,
[40] K. M. Backus, B. E. Correia, K. M. Lum, S. Forli, B. D. Horning,
G. E. González-Páez, S. Chatterjee, B. R. Lanning, J. R. Teijaro,
A. J. Oison, D. W. Wolan, B. F. Cravatt, Nature 2016, 534,
570−574.
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202105383
Angewandte Chemie International Edition
RESEARCH ARTICLE
[41] S. Pan, S. Y. Jang, S. S. Liew, J. Fu, D. Wang, J. S. Lee, S. Q.
Yao, Angew. Chem. Int. Ed. 2018, 57, 579−583.
[42] D. Mellacheruvu, Z. Wright, A. L. Couzens, J. P. Lambert, N. A.
St-Denis, T. Li, Y. V. Miteva, S. Hauri, M. E. Sardiu, T. Y. Low,
V. A. Halim, R. D. Bagshaw, N. C. Hubner, A. Al-Hakim, A.
Bouchard, D. Faubert, D. Fermin, W. H. Dunham, M.
Goudreault, Z. Y. Lin, B. G. Badillo, T. Pawson, D. Durocher, B.
Coulombe, R. Aebersold, G. Superti-Furga, J. Colinge, A. J.
Heck, H. Choi, M. Gstaiger, S. Mohammed, I. M. Cristea, K. L.
Bennett, M. P. Washburn, B. Raught, R. M. Ewing, A. C.
Gingras, A. I. Nesvizhskii, Nat. Methods 2013, 10, 730−736.
[43] a) W.-H. Guo, X. Qi, X. Yu, Y. Liu, C.-I. Chung, F. Bai, X. Lin, D.
Lu, L. Wang, J. Chen, L. H. Su, K. J. Nomie, F. Li, M. C. Wang,
X. Shu, J. N. Onuchic, J. A. Woyach, M. L. Wang, J. Wang, Nat.
Commun. 2020, 11, 4268; b) R. Gabizon, A. Shraga, P. Gehrtz,
E. Livnah, Y. Shorer, N. Gurwicz, L. Avram, T. Unger, H.
Aharoni, S. Albeck, A. Brandis, Z. Shulman, B.-Z. Katz, Y.
Herishanu, N. London, J. Am. Chem. Soc. 2020, 142,
11734−11742.
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202105383
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
Using iminoboronate chemistry, we report the first successful examples of catalytic lysine-targeting reversible covalent BCR-
ABL inhibitors which inhibited both ABLWT and ABLT315I at nanomolar potency. We also demonstrated how the study of off-
targets for this class of compounds could be performed using activity-based protein profiling and mass spectrometry.
9
This article is protected by copyright. All rights reserved.