Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering

Full text of DOI:10.1016/j.chembiol.2020.04.004

Article
Identification of a Covalent Molecular Inhibitor of
Anti-apoptotic BFL-1 by Disulfide Tethering
Graphical Abstract
Authors
Edward P. Harvey,
Zachary J. Hauseman,
BFL-1
Daniel T. Cohen, ..., John R. Engen,
James A. Wells, Loren D. Walensky
Mitochondrion
tBID
Correspondence
loren_walensky@dfci.harvard.edu
X
In Brief
BAX
NH
Harvey et al. perform a disulfide tethering
screen of BFL-1, a BCL-2 protein
O
H
N
S
N
S
N
H
O
implicated in cancer, in order to identify
molecules capable of covalently targeting
a unique C55 residue of the anti-apoptotic
groove. 4E14 derivatization (304 Da
moiety) effectively blocks BFL-1,
4E14
Oligomerization
Cyto c
Release
informing the design of cysteine-reactive
drugs to selectively target BFL-1.
Apoptosis
Highlights
d
d
d
d
A disulfide tethering screen identifies small molecules that
target BFL-1 C55
Structural analyses reveal the conformational consequences
of disulfide formation
Lead molecule 4E14 effectively competes with BH3-binding
at the BFL-1 groove
Covalent 4E14 targeting of C55 inhibits BFL-1 suppression of
mitochondrial apoptosis
Harvey et al., 2020, Cell Chemical Biology 27, 1–10
June 18, 2020 ª 2020 Elsevier Ltd.
https://doi.org/10.1016/j.chembiol.2020.04.004

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
Cell Chemical Biology
Article
Identification of a Covalent Molecular
Inhibitor of Anti-apoptotic
BFL-1 by Disulfide Tethering
Edward P. Harvey,1,2,6 Zachary J. Hauseman,1,2,6 Daniel T. Cohen,1,2 T. Justin Rettenmaier,3 Susan Lee,1,2
Annissa J. Huhn,1,2 Thomas E. Wales,4 Hyuk-Soo Seo,5 James Luccarelli,1,2 Catherine E. Newman,1,2 Rachel M. Guerra,1,2
Gregory H. Bird,1,2 Sirano Dhe-Paganon,5 John R. Engen,4 James A. Wells,3 and Loren D. Walensky1,2,7,
1Department of Pediatric Oncology, Dana-Farber Cancer Institute, 450 Brookline Avenue, Boston, MA 02215, USA
*
2Linde Program in Cancer Chemical Biology, Dana-Farber Cancer Institute, 450 Brookline Avenue, Boston, MA 02215, USA
3Departments of Pharmaceutical Chemistry and Cellular and Molecular Pharmacology, University of California, 1700 Fourth Street, San
Francisco, CA 94143, USA
4Department of Chemistry and Chemical Biology, Northeastern University, 412 The Fenway, Boston, MA 02115, USA
5Department of Cancer Biology, Dana-Farber Cancer Institute, 450 Brookline Avenue, Boston, MA 02215, USA
6These authors contributed equally
7Lead Contact
*Correspondence: loren_walensky@dfci.harvard.edu
https://doi.org/10.1016/j.chembiol.2020.04.004
SUMMARY
that pierce the MOM and release apoptogenic factors into
the cytosol (Walensky and Gavathiotis, 2011). Anti-apoptotic
The BCL-2 family is composed of anti- and pro- BCL-2 proteins contain a surface groove that can bind and
trap the exposed BH3 helices of BAX and BAK, and thereby
neutralize the killing mechanism (Sattler et al., 1997). An addi-
apoptotic members that respectively protect or
disrupt mitochondrial integrity. Anti-apoptotic over-
expression can promote oncogenesis by trapping
the BCL-2 homology 3 (BH3) ‘‘killer domains’’ of
pro-apoptotic proteins in a surface groove, blocking
apoptosis. Groove inhibitors, such as the relatively
large BCL-2 drug venetoclax (868 Da), have emerged
as cancer therapies. BFL-1 remains an undrugged
tional layer of regulation involves the BH3-only members, a
heterogeneous subclass of pro-apoptotic BCL-2 proteins that
only share homology in the BH3 region (O’Connor et al., 1998;
Wang et al., 1996; Yang et al., 1995). BH3-only proteins can
deploy their conserved helices to directly trigger pro-apoptotic
BAX and BAK or serve as decoys to block the anti-apoptotic
grooves (Cheng et al., 2001). The latter inhibit-the-inhibitor
mechanism of BH3-only pro-apoptotic activity provided a blue-
print for the development of small-molecule drugs to reactivate
oncogenic protein and can cause venetoclax resis-
tance. Having identified a unique C55 residue in the
BFL-1 groove, we performed a disulfide tethering apoptosis in cancer by targeting the anti-apoptotic groove
(Oltersdorf et al., 2005).
Of the six anti-apoptotic BCL-2 family targets, small molecules
screen to determine if C55 reactivity could enable
smaller molecules to block BFL-1’s BH3-binding
functionality. We found that a disulfide-bearing N-
have been developed to effectively block BCL-2, BCL-XL, and
MCL-1.
A selective BCL-2 inhibitor, venetoclax, is Food
acetyltryptophan analog (304 Da adduct) effectively
targeted BFL-1 C55 and reversed BFL-1-mediated
suppression of mitochondrial apoptosis. Structural
analyses implicated the conserved leucine-binding
pocket of BFL-1 as the interaction site, resulting in
conformational remodeling. Thus, therapeutic tar-
geting of BFL-1 may be achievable through the
design of small, cysteine-reactive drugs.
and Drug Administration-approved for the treatment of CLL,
and other BCL-XL and MCL-1 inhibitors are currently in clinical
trials (Kotschy et al., 2016; Souers et al., 2013; Tse et al.,
2008). Such molecules are mimics of BH3 peptide helices, which
engage a large, flat, and shallow groove on the surface of anti-
apoptotic BCL-2 proteins. Correspondingly, the inhibitory mole-
cules are relatively large (e.g., venetoclax, molecular weight
[MW] 868; navitoclax, MW 975; S63845, MW 829), which can
pose ‘‘beyond-the-rule-of-5’’ (bRo5) pharmacologic challenges
during development (DeGoey et al., 2018). BFL-1 (also known
as BCL2A1 or A1) is another BCL-2 family anti-apoptotic protein
that has been implicated in the pathogenesis of human cancers,
including leukemia, lymphoma, melanoma, and gastric cancer
(Choi et al., 1995; Fan et al., 2010; Haq et al., 2013; Morales
et al., 2005). BH3 helices have been shown to engage individual,
dual, or multiple anti-apoptotic targets (Certo et al., 2006; Chen
et al., 2005; Stewart et al., 2010) and, therefore, a tunable spec-
trum of molecular targeting capability may be achievable. Never-
theless, the current trend favors selectivity to avoid the potential
INTRODUCTION
The response to physiologic or pathologic stress is regulated
by BCL-2 family proteins, which render a cellular life or death
decision that is processed at the mitochondrial outer membrane
(MOM) (Kalkavan and Green, 2018). Stress-induced activation of
pro-apoptotic proteins, such as BAX or BAK, leads to transient
exposure of their BCL-2 homology 3 (BH3) helices, which partic-
ipate in converting BAX/BAK monomers into toxic oligomers
Cell Chemical Biology 27, 1–10, June 18, 2020 ª 2020 Elsevier Ltd.
1

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
toxicity of broad targeting. For example, the dual BCL-2/BCL-XL verify the hits and identify the most potent molecular competitors
inhibitor, nativoclax, was found to induce rapid thrombocyto- of BH3/BFL-1 interaction, we performed a fluorescence polari-
penia as a result of on-mechanism toxicity associated with zation (FP) secondary screen based on the comparative binding
blockade of BCL-XL in platelets, motivating the development of of a fluoresceinated BID BH3 peptide to apo versus fragment-
selective BCL-2 inhibitors to avoid the risk of hemorrhage in conjugated BFL-1 protein. Whereas all of the identified com-
cancer patients (Mason et al., 2007; Souers et al., 2013). How- pounds shifted the fluorescein isothiocyanate (FITC)-BID BH3
ever, resistance against precision anti-apoptotic inhibitors like binding isotherm to the right upon BFL-1 conjugation, 4E14, an
venetoclax has already emerged, including BCL-2 protein N-acetyltryptophan analog of 408 Da, emerged as the most
mutations and alternative anti-apoptotic expression, such as effective competitor (ꢀ56-fold shift in EC50) (Figure 1C). Intact
MCL-1 and BFL-1, portending a ‘‘whack-a-mole’’ challenge for mass spectrometry documented complete conjugation of 4E14
the field (Birkinshaw et al., 2019; Yecies et al., 2010). Thus, to BFL-1DC (M + 304) under the experimental conditions (Fig-
developing molecular inhibitors for the complete array of anti- ure 1D). To evaluate the C55 dependence of 4E14 binding
apoptotic proteins, particularly those like BFL-1 that remain activity and compound selectivity, we repeated the competitive
undrugged, is a high priority goal.
FP using BFL-1 protein that lacked C55 but retained the native
In surveying design opportunities for selective BFL-1 target- C4 and C19 residues (BFL-1DC C55S), and alternative anti-
ing, we noted a cysteine within its BH3-binding groove (Huhn apoptotic members, BCL-XLDC and MCL-1DNDC, which
et al., 2016). Strikingly, no other BCL-2 family protein has an contain cysteine residues but not within their BH3-binding
analogous cysteine at the critical regulatory groove. Guided by grooves. In each case, exposing the proteins to 4E14 under
a native cysteine pair in the complex between NOXA BH3 and conjugating conditions had no effect on the capacity of FITC-
BFL-1 (PDB: 3MQP), we generated a series of cysteine-reactive BID BH3 to bind to the respective anti-apoptotic targets (Figures
stabilized a helices of BCL-2 domains (SAHBs), which were 1E and 1F). Thus, disulfide tethering yielded a lead molecular
shown by biochemical, structural, and cellular means to target fragment capable of direct and selective C55 derivatization
BFL-1 through combined covalent and non-covalent interaction, under reducing conditions, resulting in impaired BH3-binding
and thereby reactivate apoptosis in BFL-1-driven cancers interaction at the canonical BFL-1 groove.
(Guerra et al., 2018; Harvey et al., 2018; Huhn et al., 2016). To
investigate if small molecules could do the same, we turned to Conformational Consequences of the Covalent 4E14/
disulfide tethering, which has emerged as a powerful screening BFL-1 Interaction
method for identifying molecular fragments that engage a target To assess the impact of disulfide bond formation between 4E14
binding site in the vicinity of a natural or installed cysteine residue and C55 on the structural dynamics of BFL-1DC C4S/C19S, we
(Erlanson et al., 2000; Ostrem et al., 2013). We were especially performed hydrogen/deuterium exchange mass spectrometry
interested in determining if much smaller compounds, when en- (HXMS) analyses (Barclay et al., 2015; Engen, 2009; Lee et al.,
dowed with a cysteine-reactive moiety, could achieve BH3- 2016). HXMS interrogates the conformation of proteins by
mimetic functionality. Here, we surprisingly identified a disul- measuring the relative deuterium uptake of backbone amides.
fide-bearing N-acetyltryptophan analog that forms an adduct Protein regions involved in ligand interaction can become pro-
of only 304 Da in size upon C55 targeting yet exerts structural tected or structured, resulting in decreased deuterium exchange
and biochemical consequences analogous to BH3-mediated of the corresponding backbone amide hydrogens. In contrast,
BFL-1 inhibition.
deuterium exchange increases in areas of a protein that become
deprotected, unstructured, or more dynamic upon ligand inter-
action. Here, we incubated BFL-1DC C4S/C19S with 4E14 or
vehicle at room temperature for 1 h and then performed HXMS
analyses at 10 s and 10 min of deuteration. A reducing agent
(tris(2-carboxyethyl)phosphine) was included in the quench
RESULTS
A Disulfide Tethering Screen Identifies Molecular
Fragments that Conjugate to BFL-1 C55
We performed a disulfide tethering screen to identify molecules buffer to remove C55-bound 4E14, thereby allowing for peptide
capable of engaging the BH3-binding groove of BFL-1 in the comparisons between the vehicle and 4E14 conditions. In the
context of disulfide formation with BFL-1 C55 (Figure 1A). Spe- presence of 4E14, the distal region of a2 and proximal portion
cifically, we tested a 1,600-member library (Burlingame et al., of a3 were strongly protected from deuterium exchange at the
2011; Hallenbeck et al., 2018) for interaction with BFL-1DC 10 s labeling time point, indicative of molecular engagement of
C4S/C19S, a construct lacking the C-terminal a9 helix and the the very structures that comprise the upper portion of the canon-
two native cysteine residues located outside of the BH3-binding ical groove, including C55 (Figures 1A, 2A, and 2B; Table S2;
groove. Compounds were incubated with BFL-1 protein at room Data S1). At 10 min of deuterium labeling, additional adjacent
temperature for 1 hour in the presence of 500 mM b-mercaptoe- regions become protected from deuterium exchange, including
thanol and then percent tethering was determined by intact a portion of the a1-a2 loop, proximal a2, and the middle of a4
mass spectrometry. We identified 31 compounds with percent (Figures 2A and 2C; Table S2; Data S1). These data are consis-
tethering values of two standard deviations above the mean, re- tent with 4E14 targeting the canonical groove and reducing the
flecting 51%–73% BFL-1 derivatization under the reducing conformational dynamics of the associated components, span-
experimental conditions (Figure 1B). A common theme among ning from the a1-a2 loop to a4 (Figures 2B and 2C; Table S2;
the chemical structures of the hits is the presence of an aryl, Data S1).
benzyl, pyridyl, or indolyl group separated from the reactive Next, to distinguish between the influence of 4E14 interaction
cysteine by a linker of four to six bond lengths (Table S1). To with BFL-1DC in the presence and absence of disulfide bond
2
Cell Chemical Biology 27, 1–10, June 18, 2020

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
Figure 1. A Disulfide Tethering Screen Iden-
A
B
tifies Covalent BFL-1 Inhibitor Molecules
that Disrupt BH3-Binding Activity
100
80
60
40
20
0
D-NA-NOXA
SAHB
(A) Crystal structure of the complex between anti-
apoptotic BFL-1DC (gray) and a cysteine-reactive
stapled NOXA BH3 peptide (cyan) (PDB: 5WHH),
highlighting the site of covalent attachment to C55
(orange) and interaction at the canonical BH3-
binding groove (purple).
C55
Top 31 Hits
Canonical
Binding Pocket
(B) A disulfide tethering screen of 1,600 disulfide
fragments against BFL-1DC C4S/C19S yielded 31
fragments with percent tethering of at least two
standard deviations above the mean, as measured
by mass spectrometry.
BFL-1'C
Fragment (Rank Order)
(C) Fluorescence polarization (FP) assays
comparing the relative capacity of the 31 identified
fragments to compete with FITC-BID BH3 peptide
for interaction at the BFL-1DC C4S/C19S groove.
4E14 emerged as the most potent competitive in-
C
D
240
O
BFL-1'C
BFL-1'C / 4E14
H
N
Vehicle
4E14
Other Hits
N
S
S
N
H
O
100
80
HN
4E14
200
160
120
80
hibitor of the FITC-BID BH3/BFL-1 interaction (EC50
:
FITC-BID BH3/BFL-1, 23 nM; FITC-BID BH3/BFL-
1–4E14, 1.3 mM). Data are mean SD for experi-
ments performed in technical triplicate and repeated
two times with independent preparations of ligand
and protein with similar results.
60
40
20
0
(D) Complete derivatization of His6-BFL-1DC
by 4E14, as measured by intact mass spectrom-
etry (molecule:protein, 20:1). Left peak (black):
18,473 Da (BFL-1 minus Met plus H2O); right peak
(red): 18,777 Da (M + 304).
40
10-9
10-8
10-7
10-6
17500 18000 18500 19000 19500
Mass (Da)
[BFL-1'C C4S/C19S], M
E
F
FITC-BID BH3 +
FITC-BID BH3 +
BFL-1'C C55S + Vehicle
BFL-1'C C55S / 4E14
(E and F) FP analysis evaluating the capacity of 4E14
to compete with FITC-BID BH3 for interaction with
BFL-1DC C55S (E) and BCL-XLDC and MCL-
1DNDC (F). Data are mean SEM for experiments
performed in technical triplicate and repeated two
times with independent preparations of ligand and
protein with similar results.
BCL-XL'C + Vehicle
BCL-XL'C / 4E14
MCL-1'Ν'C + Vehicle
MCL-1'Ν'C / 4E14
100
80
60
40
20
0
100
80
60
40
20
0
See also Table S1.
10-9
10-8
10-7
10-6
10-9
10-8
[BFL-1
10-7
C C55S], M
10-6
'
[Anti-apoptotic Protein], M
formation, we repeated the HXMS analysis with BFL-1DC C55S. complex between BFL-1 and the cysteine-reactive stapled
In contrast to the prominent protection of the C55 region of peptide D-NA-NOXA SAHB (PDB: 5WHH) revealed that the
the canonical groove upon disulfide bond formation, 4E14 positioning of a2 in the 4E14/BFL-1 complex is intermediate be-
engagement of BFL-1DC in the absence of disulfide formation tween unbound and BH3-bound BFL-1 (Figure 3B). This sug-
caused subtle deprotection of the proximal portion of a2 (Fig- gests that 4E14 is capable of inducing a conformational opening
ure S1; Table S2; Data S1). Since initial non-covalent interaction of the BFL-1 canonical pocket in a manner similar to that
at the canonical groove by individual BH3 helices and small observed for a BH3 helix (Harvey et al., 2018). Integrating the
molecules is believed to open the groove to allow for ligand ac- crystallography and HXMS results, it appears that non-covalent
commodation (Harvey et al., 2018; Lee et al., 2007), it is plausible 4E14 binding may remodel and transiently expose a2 to ‘‘open’’
that the subtle a2 deprotection observed by HXMS could reflect the groove, followed by conformational stabilization of the
an analogous structural perturbation upon weak non-covalent region upon disulfide bond formation between 4E14 and C55
binding of 4E14 in the absence of C55.
and thus reinforcement of molecular interaction at the groove
(Figures 2, S1, 3A, and 3B).
To calculate a model structure of the 4E14 binding interaction,
Structure of the 4E14/BFL-1 Complex
In an effort to determine the binding mode of the 4E14/BFL-1 we performed a covalent docking analysis using the structure
interaction at the canonical groove, we solved a crystal structure of BFL-1 bound to NOXA BH3 (PDB: 3MQP). The preferred
˚
of the complex at 1.74 A resolution (Figure 3A; Table 1, PDB: pose, in addition to molecular dynamics simulations based on
6VO4). Notably, BFL-1 a3 residues 55–57 and the molecular the docking analysis, positioned the indole moiety of 4E14 in
fragment itself were not visualized, likely due to flexibility of the the ‘‘p2’’ binding pocket of BFL-1, where the universally
disulfide linkage and/or 4E14 binding mode. Nevertheless, conserved leucine of BH3 domain helices engages anti-
alignment with the structure of apo BFL-1 (PDB: 5WHI) and the apoptotic targets (Acoca et al., 2011). 4E14 is predicted to
Cell Chemical Biology 27, 1–10, June 18, 2020
3

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
A
BFL-1'C
D1
D2 D3 D4 D5
D6 D7D8
10 sec deuteration
10 min deuteration
DBFL-1'C + 4E14 (Da) - DBFL-1'C (Da)
1
0
-1
-2
-3
BFL-1
'
C Peptide (Residue #)
B
C
D1-D2
Loop
D1-D2
Loop
D2
D2
D2
D3
D3
D2
C55
C55
C55
C55
0.5
0
-0.5
-1.0
-1.5
-2.0
No
Change
D4
D4
Protection
90°
90°
D4
D4
Figure 2. Conformational Consequences of 4E14 Derivatization of BFL-1
(A) A deuterium difference plot showing the relative deuterium incorporation of BFL-1DC C4S/C19S conjugated to 4E14 minus that of apo BFL-1DC, as measured
at the indicated time points. Each identified peptide is listed in order on the x axis with the relative protection displayed on the y axis. The changes outside of the
gray shaded region are above the significance threshold of 0.5 Da. Data are representative of three biological replicates. All HXMS data used to create this figure
can be found in Data S1.
(B and C) The regions of relative change in deuterium uptake for BFL-1DC/4E14 versus apo BFL-1DC are mapped onto the structure of BFL-1DC for the indicated
10 s (B) and 10 min (C) time points. The relative level of protection is indicated by the color scale.
See also Figure S1, Table S2, and Data S1.
interact with hydrophobic BFL-1 residues V48, L52, V74, and F95 with a thiol bearing a linker of desired length. The monophore
of the p2 region, with the indole hydrogen forming a hydrogen core of interest was then attached to the linker by HATU coupling
bond with E78 (Figures 3C–3E and S2). With the exception of (step 7), yielding the disulfide fragment after cleavage from
F95 of a5, peptides that contain these a2-a4 residues were the resin by disulfide exchange (step 8) and LC-MS purification.
all observed to have reduced deuterium exchange by HXMS Using this scheme, we generated 4E14 and a series of deriva-
analysis (Figure 2), supporting the calculated model structure. tives, including N-acetyl-D-tryptophan (D-4E14), N-acetyl-5-
Taken together, our HXMS, crystallography, and computational methyl-tryptophan (5-Me-4E14), and N-acetyl-D-2-naphthylala-
studies suggest that upon disulfide bond formation with C55, nine (D-Nal-4E14) (Figure 4B).
4E14 engages the adjacent p2 binding pocket of BFL-1, resulting
To determine if the covalent 4E14 interaction with BFL-1
in focal conformational stabilization of select canonical groove had functional consequences beyond mitigating BH3-domain
helices, with a2 maintaining an intermediate opening position be- binding, as detected by competitive FP (Figure 1C), we evalu-
tween that of the apo and BH3-liganded conformations of BFL-1. ated the effect of 4E14 and its analogs on mitochondrial cyto-
chrome c release. We purified BAX/BAK-deficient mitochondria
Functional Activity of the Covalent 4E14 Interaction
from the livers of AlbCreposBaxfl/flBakꢁ/ꢁ mice (Walensky et al.,
To generate a series of 4E14 analogs for functional testing, we 2006) and treated them with monomeric full-length BAX, tBID,
developed a facile solid phase synthetic scheme involving four or the BAX/tBID combination in the presence or absence of
reaction phases, namely resin transformation, linker addition, BFL-1DC C4S/C19S. Whereas monomeric BAX or tBID alone
monophore core coupling, and cleavage from the resin by had little to no effect on the mitochondria, the combination
disulfide exchange (Figure 4A). The first four synthetic steps induced robust cytochrome c release, which was significantly
transformed the resin’s amine functional group into a thiol that suppressed by the addition of BFL-1 (Figure 4C). Conditions
remained on the resin for the duration of the synthesis. The that support complete conversion of BFL-1 to the fully 4E14-
desired disulfide tethering linker was added in steps five and conjugated form effectively blocked the anti-apoptotic function-
six by first activating the thiol by disulfide exchange with 2,20-di- ality of BFL-1 (Figure 4C), allowing freed catalytic tBID to trigger
pyridyldisulfide and then exchanging the activated disulfide BAX, which propagates the signal through autoactivation and
4
Cell Chemical Biology 27, 1–10, June 18, 2020

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
A
B
C
D
E
Figure 3. Structural Analyses of the 4E14/BFL-1DC Interaction
(A) Structure of His6-BFL-1DC C4S/C19S when conjugated to 4E14 via C55 (PDB: 6VO4). Amino acid residues 55–57 of a3, including the 4E14-C55 conjugate, are
not resolved in the structure and instead represented by the green dotted line.
(B) Overlay of the a2 region in the crystal structures of apo BFL-1DC (PDB: 5WHI), and the complexes between BFL-1DC and D-NA-NOXA SAHB (PDB: 5WHH)
and 4E14 (PDB: 6VO4), highlighting the dynamic nature of this region upon small molecule and peptide helix engagement.
(C) Calculated model structure of the 4E14/BFL-1DC interaction as determined by covalent docking and molecular dynamics simulations.
(D) Molecular interactions of 4E14 at the docked site on BFL-1DC, highlighting engagement of the indole moiety at the hydrophobic p2 region of BFL-1 and
proximity of E78 to the indole NH, suggestive of a hydrogen-bonding interaction.
(E) Side views of the interaction between the BH3-binding groove of BFL-1 and 4E14 (top) and D-NA-NOXA SAHB (PDB: 5WHH) (bottom), demonstrating the
covalent bond between C55 and the respective compounds, in addition to engagement of the disulfide linker (top) and D-nipecotic acid moiety (bottom) with the
cryptic site that forms upon ligand-target interaction.
See also Figure S2.
resultant cytochrome c release. Importantly, the observed inhib- ment. Intriguingly, the D-enantiomer of 4E14 fully conjugated
itory activity of 4E14 was abrogated upon C55S mutagenesis of to BFL-1 but was ꢀ30% less effective at inhibiting BFL-1 in the
BFL-1DC C4S/C19S (Figure S3). Taken together, these data cytochrome c release assay compared with the L-enantiomer
suggest that a small molecular fragment that engages the (Figure 4C), perhaps due to a less effective binding mode for
groove and is reinforced by covalent attachment to C55 can BH3 competition, as suggested by docking analysis (Figures
inhibit BFL-1 function.
S4A–S4D). A 5-methylated analog of 4E14 likewise conjugated
Finally, we tested several analogs of 4E14 to evaluate potential fully to BFL-1 but was even less active (Figure 4C), indicating
differences in conjugation and selectivity of functional engage- that the added methyl group may have altered the indole
Cell Chemical Biology 27, 1–10, June 18, 2020
5

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
high-priority therapeutic target. Whereas relatively large small
molecules have been developed to inhibit the canonical BH3-
Table 1. Data Collection and Refinement Statistics
Structure Name
RCSB accession code
Data Collectiona
Space group
BFL-1/4E14
6VO4
binding grooves of BCL-2, BCL-XL, and MCL-1 by non-covalent
targeting, here we explored whether incorporating a covalent re-
action with the unique C55 of the BFL-1 groove could enable
smaller molecules to effectively block anti-apoptotic activity.
Indeed, a disulfide tethering screen identified a series of small
molecules that effectively competed with BID BH3 for binding
to the BFL-1 groove. 4E14, a small N-acetyltryptophan analog
of 408 Da that derivatized BFL-1 with a 304 Da moiety, surpris-
ingly emerged as a lead fragment capable of effective BFL-1
blockade.
P21
Cell dimensions
˚
a, b, c (A)
39.55, 43.24, 43.49
90.0, 104.9, 90.0
32.78–1.74 (1.80–1.74)b
0.03546 (1.003)
11.20 (0.73)
a, b, g (ꢂ)
˚
Resolution (A)
Rpim
I/sI
From a structural standpoint, 4E14 induced conformational
changes in the BFL-1 groove that resembled those induced by
a cysteine-reactive stapled peptide modeled after the NOXA
BH3 domain (D-NA-NOXA SAHB), namely displacement of a2
to open and inhibit the groove. The amphipathic BH3 helices
bind tightly to the canonical anti-apoptotic grooves by a series
of hydrophobic contacts, which are reinforced by a perimeter
of complementary charge-charge and hydrophilic interactions
(Sattler et al., 1997; Stewart et al., 2010). One of the key hydro-
phobic interactions involves a leucine residue that is conserved
across all BCL-2 family BH3 domains. Our calculated model
structure of the 4E14/BFL-1 complex suggests that the indole
moiety of 4E14 is positioned to both engage the conserved
leucine-binding or p2 pocket and participate in a hydrogen-
bonding interaction with E78 of BFL-1 a4. Interestingly, the
installation of an indole moiety in venetoclax (based on the
fortuitous interaction of W30 from an adjacent BCL-2 molecule
with the p4 pocket seen in crystal packing) tailored its selectivity
for BCL-2 by a hydrogen-bonding interaction between the
indole moiety and D103 of BCL-2 a2 (Souers et al., 2013). In
addition to the 4E14 indole interaction at the BFL-1 p2 pocket,
the disulfide linker is observed to engage an adjacent cryptic
site (involving L52, L56, V74, and F95) that forms below C55
upon covalent (Harvey et al., 2018) or non-covalent (PDB:
3MPQ) NOXA BH3 engagement of BFL-1. Importantly, this
newfound hydrophobic pocket does not form upon NOXA BH3
interaction with MCL-1 (Harvey et al., 2018) and thus provides
a further design opportunity to achieve selective targeting of
Completeness (%)
Redundancy
Structure Solution
91.91 (87.60)
3.2 (3.2)
PDB entries used for molecular
replacement
3WHH
Refinement
No. reflections, unique
13,520 (1,286)
0.2383/0.2830
1,089
R
work/Rfree
No. of atoms
Protein
1,069
Water
20
B factors
60.06
Protein
60.09
Water
58.46
Root-mean-square deviation
˚
Bond lengths (A)
0.007
0.86
Bond angles (ꢂ)
Ramachandran plot (%)
Preferred
96.2
3.8
Allowed
Not allowed
0.0
aSingle crystal was used to collect data for the reported structure.
bValues in parentheses are for highest-resolution shell.
interaction in a manner that further impaired BH3 competition BFL-1 by cysteine-reactive small molecules.
(Figures S4E and S4F). Of note, aside from 4E14, no fused ring
The incorporation of a cysteine-reactive moiety into a stapled
structures or substituted analogs thereof were found among NOXA BH3 peptide enabled truncation of the original template
the 31 hits from the screen (Table S1), consistent with these sequence from 22 amino acids (aa: 22–43) to 15 amino acids
data. To evaluate the effect of increased hydrophobic bulk and (aa: 26–40) (Guerra et al., 2018; Huhn et al., 2016). Whereas
elimination of the indole hydrogen, we replaced the indole moiety the shortened stapled sequence showed little to no non-covalent
of the 4E14 D-enantiomer with naphthalene, which resulted in interaction with MCL-1 or BFL-1, incorporation of an acrylamide
complete abrogation of conjugation and BFL-1-inhibitory activity warhead conferred robust and BFL-1-selective binding activity
(Figures 4C and S4G). These data support the relative specificity (Guerra et al., 2018; Harvey et al., 2018). Here, we find that
of 4E14 for functional BFL-1 interaction, with impairment of 4E14 reaction with BFL-1 C55 produces an M + 304 Da adduct
activity by only minor modifications of the chemical structure that is capable of blocking both BH3 interaction and the capacity
indicative of a discrete structure-activity relationship.
of BFL-1 to suppress BAX-mediated cytochrome c release.
Thus, our current dataset suggests that incorporation of a
cysteine-reactive moiety into a matured non-covalent inhibitor
of the anti-apoptotic groove could yield potent and selective
DISCUSSION
BFL-1 is an undrugged BCL-2 family protein implicated in both molecular inhibitors of BFL-1 that are appreciably smaller
the pathogenesis and chemoresistance of human cancers (Fan than venetoclax, navitoclax, and S63845 molecules and poten-
et al., 2010; Haq et al., 2013; Mahadevan et al., 2005; Yecies tially bypass the challenges associated with bRo5 compounds
et al., 2010). Thus, like BCL-2, BCL-XL, and MCL-1, BFL-1 is a (DeGoey et al., 2018).
6
Cell Chemical Biology 27, 1–10, June 18, 2020

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
A
B
H
N
H
N
3. HATU, DIEA
Cysteamine HCl
NH2
1. 20% Piperidine
DMF
2. Glutaric Anhydride
DMF
OH
O
O
H
N
Fmoc
H
N
S
S
DMF, H2O
O
O
R
N
H
R
N
H
O
O
HN
HN
4E14
D-4E14
4. TCEP
1:1 DMF:H2O
7. N-Ac-Trp
HATU,DIEA
H
N
H
N
H
N
H
N
S
SH
S
NH2
O
O
DMF
5. (PyS)2, DMF
H
N
H
N
O
O
O
O
S
S
6. Cysteamine HCl
DMF, H2O
R
N
H
R
N
H
O
O
HN
O
H
O
H
N
H
N
H
S
N
N
S
N
N
H
S
S
N
H
8.TCEP, 1:1 DMF:H2O
5-Me-4E14
D-Nal-4E14
O
O
O
O
HN
HN
Bis(2-dimethylaminoethyl)
disulfide dihydrochloride,TEA
R =
N
S
4E14
C
100
80
60
40
20
***
0
BAX
-
+
+
+
+
+
+
+
tBID
-
+
+
-
+
+
-
+
+
+
-
+
+
-
+
+
-
+
+
-
BFL-1'C C4S/C19S
4E14
-
-
-
-
-
-
-
-
D-4E14
5-Me-4E14
D-Nal-4E14
-
-
-
+
-
-
-
-
-
-
-
-
+
-
-
-
-
-
-
+
Figure 4. Covalent Targeting by 4E14 Blocks BFL-1 Suppression of BAX-Mediated Mitochondrial Apoptosis
(A) Solid-phase synthetic scheme for the production of diverse disulfide tethering molecular fragments on resin.
(B) Chemical structures of 4E14 and analogs.
(C) Percent cytochrome c release from BAX/BAK-deficient mouse liver mitochondria treated with tBID, BAX, and/or BFL-1DC C4S/C19S in the presence or
absence of 4E14 and its analogs. Data are mean SEM for experiments performed in technical triplicate and repeated (biological duplicate) with independent
preparations of compounds, protein, and mitochondria with similar results. BAX, 100 nM; BFL-1DC C4S/C19S, 1 mM; tBID, 40 nM; 4E14 and analogs conjugated
to BFL-1 at molecule:protein ratio of 5:1. ***, p < 0.0001 by unpaired Student’s t test.
See also Figures S3 and S4.
Disulfide tethering rapidly provided a starting point for mance liquid chromatography purification. Furthermore, the
proof-of-concept experiments to determine the feasibility of method was amenable to automation, with mixing and
small-molecule covalent targeting of BFL-1. To expand the di- washing steps performed on an automated peptide synthe-
versity of analogs, we developed a synthetic scheme for pro- sizer, which allowed for multiple syntheses to be readily per-
ducing monophores using the solid phase, which enables formed in parallel. Thus, incorporating solid-phase methods
high-throughput production of analogs, robust molecular to further diversify disulfide tethering libraries may help
diversification, and minimal purification requirements. For expand the discovery potential of cysteine-reactive fragments
example, our use of solid-phase synthetic methods eliminated that engage BFL-1 and the myriad of other targets that harbor
the need for aqueous workups and column chromatography, cysteines in or around functionally relevant sites of protein
with residual solvents and salts removed upon high-perfor- interaction.
Cell Chemical Biology 27, 1–10, June 18, 2020
7

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
5T32GM007306 and 5T32GM095450; NIH grant 5F31CA210590 to Z.J.H.; a
SIGNIFICANCE
National Science Foundation pre-doctoral research fellowship to R.M.G.;
NIH grant R50CA211399 to G.H.B.; NIH grant R01CA191018 and the Harry
BCL-2 family proteins maintain the critical balance between
and Dianna Professorship in Pharmaceutical Sciences to J.A.W.; NIH grant
cellular life and death and, when deregulated, can contribute
F31CA180378 and a Krevan predoctoral fellowship to T.J.R.; and support to
to the development, maintenance, and chemoresistance of
C.E.N. and J.L. from NIH grant T32GM007753. Structural analyses were based
human cancer. Anti-apoptotic BCL-2 proteins suppress
cell death by capturing the essential BH3 helix of pro-
apoptotic proteins in a surface groove. Relatively large small
molecules (800–1,000 Da) effectively target the groove but
necessitated overcoming beyond-the-rule-of-5 challenges
during clinical development. Harnessing a unique cysteine
in the canonical groove of anti-apoptotic BFL-1, we per-
formed a disulfide tethering screen that identified a lead
small-molecule fragment capable of derivatizing BFL-1
with a 304-Da moiety, effectively blocking its BH3-binding
and anti-apoptotic functionality. Thus, we demonstrate
that disulfide tethering and a solid-phase synthetic scheme
to expand disulfide tethering libraries can provide a starting
point for the development of selective, covalent inhibitors of
BFL-1.
on X-ray data acquired at the Advanced Photon Source on the Northeastern
Collaborative Access Team beamlines (NIGMS P41 GM103403). We are also
indebted to the Wolpoff Family Foundation, Jim and Lisa LaTorre, the family
of Ivo Coll, and the Todd J. Schwartz Memorial Fund for their steadfast finan-
cial contributions to our cancer chemical biology research.
AUTHOR CONTRIBUTIONS
E.P.H., Z.J.H., and L.D.W. conceived of and designed the study. T.J.R. and
S.L. conducted the disulfide tethering fragment screen in the laboratory of
J.A.W. E.P.H. and Z.J.H. generated recombinant proteins and performed the
biochemical experiments with assistance from D.T.C., S.L., A.J.H., C.E.N.,
and R.M.G. E.P.H. and Z.J.H. synthesized compounds using the synthetic
scheme developed by D.T.C. and G.H.B. E.P.H., Z.J.H., and C.E.N. conducted
the HXMS experiments, which were analyzed and reviewed by T.E.W. under
the supervision of J.R.E. E.P.H and Z.J.H. generated protein conjugates for
structural analysis, which was performed by H.-S.S. and S.D.-P. D.T.C. and
J.L. conducted docking and molecular dynamics analyses. The paper was
written by L.D.W., E.P.H., and Z.J.H. and reviewed by all co-authors.
STAR+METHODS
DECLARATION OF INTERESTS
Detailed methods are provided in the online version of this paper
and include the following:
The authors have no competing interests related to this work.
Received: February 9, 2020
Revised: March 13, 2020
Accepted: April 3, 2020
Published: May 14, 2020
d KEY RESOURCES TABLE
d RESOURCE AVAILABILITY
B Lead Contact
B Materials Availability
B Data and Code Availability
REFERENCES
d EXPERIMENTAL MODEL AND SUBJECT DETAILS
B Microbe Strains
Acoca, S., Cui, Q., Shore, G.C., and Purisima, E.O. (2011). Molecular dynamics
study of small molecule inhibitors of the Bcl-2 family. Proteins 79, 2624–2636.
d METHOD DETAILS
Adams, P.D., Afonine, P.V., Bunkoczi, G., Chen, V.B., Davis, I.W., Echols, N.,
Headd, J.J., Hung, L.W., Kapral, G.J., Grosse-Kunstleve, R.W., et al. (2010).
PHENIX: a comprehensive Python-based system for macromolecular struc-
ture solution. Acta Crystallogr. D Biol. Crystallogr. 66, 213–221.
B Peptide Synthesis
B Recombinant Protein Expression and Purification
B Disulfide Tethering Fragment Screening by Mass
Spectrometry
Barclay, L.A., Wales, T.E., Garner, T.P., Wachter, F., Lee, S., Guerra, R.M.,
Stewart, M.L., Braun, C.R., Bird, G.H., Gavathiotis, E., et al. (2015). Inhibition
of pro-apoptotic BAX by a noncanonical interaction mechanism. Mol. Cell
57, 873–886.
B Solid Phase Synthesis of Disulfide Tethering
Fragments
B Synthesis of 4E14-2-thiopyridine
B Fluorescence Polarization Assays
B Mitochondrial Cytochrome c Release Assays
B X-ray Crystallography
Bird, G.H., Bernal, F., Pitter, K., and Walensky, L.D. (2008). Synthesis and bio-
physical characterization of stabilized alpha-helices of BCL-2 domains.
Methods Enzymol. 446, 369–386.
Birkinshaw, R.W., Gong, J.N., Luo, C.S., Lio, D., White, C.A., Anderson, M.A.,
Blombery, P., Lessene, G., Majewski, I.J., Thijssen, R., et al. (2019). Structures
of BCL-2 in complex with venetoclax reveal the molecular basis of resistance
mutations. Nat. Commun. 10, 2385.
B Molecular Dynamics Simulation and Docking
B Hydrogen Deuterium Exchange Mass Spectrometry
d QUANTIFICATION AND STATISTICAL ANALYSIS
Burlingame, M.A., Tom, C.T., and Renslo, A.R. (2011). Simple one-pot synthe-
sis of disulfide fragments for use in disulfide-exchange screening. ACS Comb.
Sci. 13, 205–208.
SUPPLEMENTAL INFORMATION
Supplemental Information can be found online at https://doi.org/10.1016/j.
chembiol.2020.04.004.
Certo, M., Del Gaizo Moore, V., Nishino, M., Wei, G., Korsmeyer, S.,
Armstrong, S.A., and Letai, A. (2006). Mitochondria primed by death signals
determine cellular addiction to antiapoptotic BCL-2 family members. Cancer
Cell 9, 351–365.
ACKNOWLEDGMENTS
Chen, L., Willis, S.N., Wei, A., Smith, B.J., Fletcher, J.I., Hinds, M.G., Colman,
P.M., Day, C.L., Adams, J.M., and Huang, D.C. (2005). Differential targeting of
prosurvival Bcl-2 proteins by their BH3-only ligands allows complementary
apoptotic function. Mol. Cell 17, 393–403.
We thank M. Godes for technical assistance, K. Korshavn for scientific discus-
sions, and M. Tye for assistance with compound characterization by NMR.
This research was supported by NIH grant R35CA197583 and a Leukemia
and Lymphoma Society (LLS) Translational Research Program award to
L.D.W.; a research collaboration between J.R.E. and the Waters Corporation;
NIH grant 5F31CA210592 to E.P.H., with additional support from
Cheng, E.H., Wei, M.C., Weiler, S., Flavell, R.A., Mak, T.W., Lindsten, T., and
Korsmeyer, S.J. (2001). BCL-2, BCL-X(L) sequester BH3 domain-only
8
Cell Chemical Biology 27, 1–10, June 18, 2020

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
molecules preventing BAX- and BAK-mediated mitochondrial apoptosis. Mol.
implications for selectivity of antagonists of the Bcl-2 family. Cell Death Differ.
Cell 8, 705–711.
14, 1711–1713.
Choi, S.S., Park, I.C., Yun, J.W., Sung, Y.C., Hong, S.I., and Shin, H.S. (1995).
A novel Bcl-2 related gene, Bfl-1, is overexpressed in stomach cancer and
preferentially expressed in bone marrow. Oncogene 11, 1693–1698.
Lee, S., Wales, T.E., Escudero, S., Cohen, D.T., Luccarelli, J., Gallagher, C.G.,
Cohen, N.A., Huhn, A.J., Bird, G.H., Engen, J.R., et al. (2016). Allosteric inhibi-
tion of antiapoptotic MCL-1. Nat. Struct. Mol. Biol. 23, 600–607.
Lelle, M., Freidel, C., Kaloyanova, S., Tabujew, I., Schramm, A., Musheev, M.,
Niehrs, C., Mullen, K., and Peneva, K. (2017). Overcoming drug resistance by
cell-penetrating peptide-mediated delivery of a doxorubicin dimer with high
DNA-binding affinity. Eur. J. Med. Chem. 130, 336–345.
DeGoey, D.A., Chen, H.J., Cox, P.B., and Wendt, M.D. (2018). Beyond the rule
of 5: lessons learned from AbbVie’s drugs and compound collection. J. Med.
Chem. 61, 2636–2651.
Emsley, P., and Cowtan, K. (2004). Coot: model-building tools for molecular
Mahadevan, D., Spier, C., Della Croce, K., Miller, S., George, B., Riley, C.,
Warner, S., Grogan, T.M., and Miller, T.P. (2005). Transcript profiling in periph-
eral T-cell lymphoma, not otherwise specified, and diffuse large B-cell lym-
phoma identifies distinct tumor profile signatures. Mol. Cancer Ther. 4,
1867–1879.
graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132.
Engen, J.R. (2009). Analysis of protein conformation and dynamics by
hydrogen/deuterium exchange MS. Anal. Chem. 81, 7870–7875.
Erlanson, D.A., Braisted, A.C., Raphael, D.R., Randal, M., Stroud, R.M.,
Gordon, E.M., and Wells, J.A. (2000). Site-directed ligand discovery. Proc.
Natl. Acad. Sci. U S A 97, 9367–9372.
Martyna, G.J., Tobias, D.J., and Klein, M.L. (1994). Constant pressure molec-
ular dynamics algorithms. J. Chem. Phys. 101, 4177–4189.
Essmann, U., Perera, L., and Berkowitz, M.L. (1995). A smooth particle mesh
Mason, K.D., Carpinelli, M.R., Fletcher, J.I., Collinge, J.E., Hilton, A.A., Ellis, S.,
Kelly, P.N., Ekert, P.G., Metcalf, D., Roberts, A.W., et al. (2007). Programmed
anuclear cell death delimits platelet life span. Cell 128, 1173–1186.
Ewald method. J. Chem. Phys. 103, 8577.
Evans, P. (2006). Scaling and assessment of data quality. Acta Crystallogr. D
Biol. Crystallogr. 62, 72–82.
McCoy, A.J., Grosse-Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C.,
and Read, R.J. (2007). Phaser crystallographic software. J. Appl. Crystallogr.
40, 658–674.
Fan, G., Simmons, M.J., Ge, S., Dutta-Simmons, J., Kucharczak, J., Ron, Y.,
Weissmann, D., Chen, C.C., Mukherjee, C., White, E., et al. (2010). Defective
ubiquitin-mediated degradation of antiapoptotic Bfl-1 predisposes to lym-
phoma. Blood 115, 3559–3569.
Morales, A.A., Olsson, A., Celsing, F., Osterborg, A., Jondal, M., and Osorio,
L.M. (2005). High expression of bfl-1 contributes to the apoptosis resistant
phenotype in B-cell chronic lymphocytic leukemia. Int. J. Cancer 113,
730–737.
Guerra, R.M., Bird, G.H., Harvey, E.P., Dharia, N.V., Korshavn, K.J., Prew,
M.S., Stegmaier, K., and Walensky, L.D. (2018). Precision targeting of BFL-
1/A1 and an ATM Co-dependency in human cancer. Cell Rep. 24, 3393–3403.
O’Connor, L., Strasser, A., O’Reilly, L.A., Hausmann, G., Adams, J.M., Cory,
S., and Huang, D.C. (1998). Bim: a novel member of the Bcl-2 family that pro-
motes apoptosis. EMBO J. 17, 384–395.
Guo, Z., Mohanty, U., Noehre, J., Sawyer, T.K., Sherman, W., and Krilov, G.
(2010). Probing the alpha-helical structural stability of stapled p53 peptides:
molecular dynamics simulations and analysis. Chem. Biol. Drug Des. 75,
348–359.
Oltersdorf, T., Elmore, S.W., Shoemaker, A.R., Armstrong, R.C., Augeri, D.J.,
Belli, B.A., Bruncko, M., Deckwerth, T.L., Dinges, J., Hajduk, P.J., et al.
(2005). An inhibitor of Bcl-2 family proteins induces regression of solid tu-
mours. Nature 435, 677–681.
Hallenbeck, K.K., Davies, J.L., Merron, C., Ogden, P., Sijbesma, E., Ottmann,
C., Renslo, A.R., Wilson, C., and Arkin, M.R. (2018). A liquid chromatography/
mass spectrometry method for screening disulfide tethering fragments. SLAS
Discov. 23, 183–192.
Ostrem, J.M., Peters, U., Sos, M.L., Wells, J.A., and Shokat, K.M. (2013). K-
Ras(G12C) inhibitors allosterically control GTP affinity and effector interac-
tions. Nature 503, 548–551.
Haq, R., Yokoyama, S., Hawryluk, E.B., Jonsson, G.B., Frederick, D.T.,
McHenry, K., Porter, D., Tran, T.N., Love, K.T., Langer, R., et al. (2013).
BCL2A1 is a lineage-specific antiapoptotic melanoma oncogene that confers
resistance to BRAF inhibition. Proc. Natl. Acad. Sci. U S A 110, 4321–4326.
Pitter, K., Bernal, F., Labelle, J., and Walensky, L.D. (2008). Dissection of the
BCL-2 family signaling network with stabilized alpha-helices of BCL-2 do-
mains. Methods Enzymol. 446, 387–408.
Harder, E., Damm, W., Maple, J., Wu, C., Reboul, M., Xiang, J.Y., Wang, L.,
Lupyan, D., Dahlgren, M.K., Knight, J.L., et al. (2016). OPLS3: a force field
providing broad coverage of drug-like small molecules and proteins.
J. Chem. Theor. Comput. 12, 281–296.
Sastry, G.M., Adzhigirey, M., Day, T., Annabhimoju, R., and Sherman, W.
(2013). Protein and ligand preparation: parameters, protocols, and influence
on virtual screening enrichments. J. Comput. Aided Mol. Des. 27, 221–234.
Sattler, M., Liang, H., Nettesheim, D., Meadows, R.P., Harlan, J.E., Eberstadt,
M., Yoon, H.S., Shuker, S.B., Chang, B.S., Minn, A.J., et al. (1997). Structure of
Bcl-xL-Bak peptide complex: recognition between regulators of apoptosis.
Science 275, 983–986.
Harvey, E.P., Seo, H.S., Guerra, R.M., Bird, G.H., Dhe-Paganon, S., and
Walensky, L.D. (2018). Crystal structures of anti-apoptotic BFL-1 and its com-
plex with a covalent stapled peptide inhibitor. Structure 26, 153–160.
Hoover, W.G. (1985). Canonical dynamics: equilibrium phase-space distribu-
Shelley, J.C., Cholleti, A., Frye, L.L., Greenwood, J.R., Timlin, M.R., and
Uchimaya, M. (2007). Epik: a software program for pK(a) prediction and pro-
tonation state generation for drug-like molecules. J. Comput. Aided Mol.
Des. 21, 681–691.
tions. Phys. Rev. A Gen. Phys. 31, 1695–1697.
Huhn, A.J., Guerra, R.M., Harvey, E.P., Bird, G.H., and Walensky, L.D. (2016).
Selective covalent targeting of anti-apoptotic BFL-1 by cysteine-reactive sta-
pled peptide inhibitors. Cell Chem. Biol. 23, 1123–1134.
Souers, A.J., Leverson, J.D., Boghaert, E.R., Ackler, S.L., Catron, N.D., Chen,
J., Dayton, B.D., Ding, H., Enschede, S.H., Fairbrother, W.J., et al. (2013). ABT-
199, a potent and selective BCL-2 inhibitor, achieves antitumor activity while
sparing platelets. Nat. Med. 19, 202–208.
Jorgensen, W.L., Chandrasekhar, J., Madura, J.D., Impey, R.W., and Klein,
M.L. (1983). Comparison of simple potential functions for simulating liquid wa-
ter. J. Chem. Phys. 79, 926–935.
Kabsch, W. (2010). Integration, scaling, space-group assignment and post-
Stewart, M.L., Fire, E., Keating, A.E., and Walensky, L.D. (2010). The MCL-1
BH3 helix is an exclusive MCL-1 inhibitor and apoptosis sensitizer. Nat.
Chem. Biol. 6, 595–601.
refinement. Acta Crystallogr. D Biol. Crystallogr. 66, 133–144.
Kalkavan, H., and Green, D.R. (2018). MOMP, cell suicide as a BCL-2 family
business. Cell Death Differ. 25, 46–55.
Tse, C., Shoemaker, A.R., Adickes, J., Anderson, M.G., Chen, J., Jin, S.,
Johnson, E.F., Marsh, K.C., Mitten, M.J., Nimmer, P., et al. (2008). ABT-263:
a potent and orally bioavailable Bcl-2 family inhibitor. Cancer Res. 68,
3421–3428.
Kotschy, A., Szlavik, Z., Murray, J., Davidson, J., Maragno, A.L., Le Toumelin-
Braizat, G., Chanrion, M., Kelly, G.L., Gong, J.N., Moujalled, D.M., et al. (2016).
The MCL1 inhibitor S63845 is tolerable and effective in diverse cancer models.
Nature 538, 477–482.
Walensky, L.D., and Gavathiotis, E. (2011). BAX unleashed: the biochemical
transformation of an inactive cytosolic monomer into a toxic mitochondrial
pore. Trends Biochem. Sci. 36, 642–652.
Lee, E.F., Czabotar, P.E., Smith, B.J., Deshayes, K., Zobel, K., Colman, P.M.,
and Fairlie, W.D. (2007). Crystal structure of ABT-737 complexed with Bcl-xL:
Cell Chemical Biology 27, 1–10, June 18, 2020
9

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
Walensky, L.D., Pitter, K., Morash, J., Oh, K.J., Barbuto, S., Fisher, J., Smith,
E., Verdine, G.L., and Korsmeyer, S.J. (2006). A stapled BID BH3 helix directly
binds and activates BAX. Mol. Cell 24, 199–210.
analysis of the [NiFe]-hydrogenase maturation factor HypF1 from Ralstonia
eutropha H16. Acta Crystallogr. Sect. F Struct. Biol. Cryst. Commun. 66,
452–455.
Wales, T.E., and Engen, J.R. (2006). Hydrogen exchange mass spectrometry
Yang, E., Zha, J., Jockel, J., Boise, L.H., Thompson, C.B., and Korsmeyer, S.J.
(1995). Bad, a heterodimeric partner for Bcl-XL and Bcl-2, displaces Bax and
promotes cell death. Cell 80, 285–291.
for the analysis of protein dynamics. Mass Spectrom. Rev. 25, 158–170.
Wales, T.E., Fadgen, K.E., Gerhardt, G.C., and Engen, J.R. (2008). High-speed
and high-resolution UPLC separation at zero degrees Celsius. Anal. Chem. 80,
6815–6820.
Yecies, D., Carlson, N.E., Deng, J., and Letai, A. (2010). Acquired resistance to
ABT-737 in lymphoma cells that up-regulate MCL-1 and BFL-1. Blood 115,
3304–3313.
Wang, K., Yin, X.M., Chao, D.T., Milliman, C.L., and Korsmeyer, S.J. (1996).
BID: a novel BH3 domain-only death agonist. Genes Dev. 10, 2859–2869.
Zhu, K., Borrelli, K.W., Greenwood, J.R., Day, T., Abel, R., Farid, R.S., and
Harder, E. (2014). Docking covalent inhibitors: a parameter free approach to
pose prediction and scoring. J. Chem. Inf. Model. 54, 1932–1940.
Winter, G., Dokel, S., Jones, A.K., Scheerer, P., Krauss, N., Hohne, W., and
Friedrich, B. (2010). Crystallization and preliminary X-ray crystallographic
10 Cell Chemical Biology 27, 1–10, June 18, 2020

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
STAR+METHODS
KEY RESOURCES TABLE
REAGENT OR RESOURCE
Antibodies
SOURCE
IDENTIFIER
Mouse monoclonal anti-His6
Abcam
Cat# Ab18184;
RRID:AB_444306
Sheep anti-mouse IgG:HRP
Goat polyclonal anti-GST
Donkey anti-goat IgG:HRP
Biorad
Cat# AAC10P;
RRID:AB_321929
GE Healthcare
Santa Cruz Biotech
Cat# 27-4577-01;
RRID:AB_771432
Cat# Sc-2020;
RRID:AB_631728
Bacterial and Virus Strains
LOBSTR BL21(DE3) Competent Cells
One Shot BL21(DE3) Competent Cells
Chemicals, Peptides, and Recombinant Proteins
Kerafast
Cat# EC001
Invitrogen
Cat# C600003
His6 BFL-1DC WT, C55S, C4S/C19S, C4S/
Walensky Lab
N/A
C19S/C55S
GST MCL-1DNDC
GST BCL-XLDC
Walensky Lab
Walensky Lab
Sigma-Aldrich
Gold Biotechnology
Thermo Fisher Scientific
Walensky Lab
Walensky Lab
Walensky Lab
Walensky Lab
Walensky Lab
Wells Lab
N/A
N/A
Complete Protease Inhibitor Tablet Cocktail
IPTG
Cat# 16829800
Cat# I2481C
Cat# PI35602
N/A
b-mercaptoethanol (BME)
4E14
4E14-2-thiopyridine
D-4E14
N/A
N/A
5-Me-4E14
N/A
D-Nal-4E14
N/A
Disulfide Tethering Fragment Library
Rink Amide AM Resin 200-400 mesh
Glutaric Anhydride
N,N-Dimethylformamide
N,N-Diisopropylethylamine
Dichloromethane
N/A
Novabiochem
Sigma Aldrich
Pharmco
Cat# 855004
Cat# G3806
Cat# zh324UVHPLC
Cat# 387649
Cat# D65100
Cat# A7022
Cat# M6500
Cat# C4706
Sigma-Aldrich
Sigma-Aldrich
ApexBio Tech LLC
Sigma Aldrich
Sigma Aldrich
HATU
Cysteamine Hydrochloride
Tris(2-carboxyethyl)phosphine
hydrochloride
Aldrithiol-2
Sigma Aldrich
TCI
Cat# 143049
Cat# B2272
Bis(2-dimethylaminoethyl) Disulfide
Dihydrochloride
Critical Commercial Assays
Q5 Site Directed Mutagenesis Kit
New England Biolabs
R&D Systems
Cat# E0554S
Cat# MCTC0
Rat/Mouse Cytochrome c Quantikine
ELISA Kit
Deposited Data
BFL-1/4E14 Crystal Structure
BFL-1/D-NA-NOXA SAHB Crystal Structure
Apo BFL-1 Crystal Structure
BFL-1/NOXA BH3 Crystal Structure
This Paper
PDB: 6VO4
PDB: 5WHH
PDB: 5WHI
PDB: 3MQP
Harvey et al., 2018
Harvey et al., 2018
Protein Data Bank
(Continued on next page)
Cell Chemical Biology 27, 1–10.e1–e6, June 18, 2020 e1

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
Continued
REAGENT OR RESOURCE
Recombinant DNA
SOURCE
IDENTIFIER
Plasmid PET17b
Novagen
Cat# 69663
N/A
Plasmid PET17b_BFL1DC_WT
Plasmid PET17b_BFL1DC_C4S/C19S
Walensky Lab
Walensky Lab
Walensky Lab
N/A
Plasmid PET17b_BFL1DC_C4S/
N/A
C19S/C55S
Plasmid PET17b_BFL1DC_C55S
Plasmid PGEX-4T-1
Walensky Lab
GE Healthcare
Walensky Lab
Walensky Lab
N/A
Cat# 28-9545-49
Plasmid PGEX-4T-1_MCL-1DNDC
Plasmid PGEX-4T-1_BCL-XLDC
Software and Algorithms
GraphPad Prism
N/A
N/A
Graphpad Software Inc.
https://www.graphpad.com/scientific-
software/prism/
XIA2
Winter et al., 2010
Evans, 2006
https://xia2.github.io
POINTLESS
XDS
http://www.ccp4.ac.uk
Kabsch, 2010
http://xds.mpimf-heidelberg.mpg.de/
Phaser
McCoy et al., 2007
https://www.phenix-online.org/
documentation/reference/phaser.html
Phenix
Coot
Adams et al., 2010
https://www.phenix-online.org/
Emsley and Cowtan, 2004
https://www2.mrc-lmb.cam.ac.uk/
personal/pemsley/coot/
Schrodinger Maestro
Pymol
Schrodinger,
Schrodinger, LLC
Version 2016-2
The PyMol Molecular Graphics
System, Version 1.7.4.0
Schrodinger, LLC
ProteinLynx Global Server (PLGS) 3.0.1
Waters Corporation
http://www.waters.com/waters/en_US/
ProteinLynx-Global-SERVER-(PLGS)/
nav.htm?cid=513821&locale=en_US
DynamX 3.0
Waters Corporation
http://www.waters.com/waters/
library.htm?cid=511436&lid=
134832928&locale=en_US
Other
Superdex 75 10/300 GL size exclusion
column
GE Healthcare Life Sciences
Cat# 29148721
RESOURCE AVAILABILITY
Lead Contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact Loren
Walensky (loren_walensky@dfci.harvard.edu).
Materials Availability
Plasmids and compounds generated in this study are available upon request to the lead contact.
Data and Code Availability
The data supporting the findings of this study are available within the article and its supplementary materials. X-ray crystallography
data was deposited to the PDB under accession number 6VO4 (BFL-1DC C4S/C19S / 4E14 complex).
EXPERIMENTAL MODEL AND SUBJECT DETAILS
Microbe Strains
Recombinant proteins were expressed in either E. coli LOBSTR BL21(DE3) or E. coli BL21(DE3) bacteria, which were grown in Luria
Broth (LB) and respectively maintained at 16ꢂC and 37ꢂC after induction, with shaking at 220 rpm.
e2 Cell Chemical Biology 27, 1–10.e1–e6, June 18, 2020

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
METHOD DETAILS
Peptide Synthesis
Synthetic BID BH3 (DIIRNIARHLAQVGDSBDRSI) peptide derivatized at the N-terminus with FITC-b-Ala was synthesized, purified,
and quantified as previously described in detail (Bird et al., 2008; Lee et al., 2016). Briefly, FITC-BID BH3 peptide was synthesized
by sequential amino acid addition to Rink Amide AM resin (EMD Millipore) using Fmoc chemistry and N-terminal derivatization
with FITC, followed by peptide deprotection, cleavage from the resin, purification by reverse phase high performance liquid chroma-
tography-mass spectrometry (LC-MS), and quantitation by amino acid analysis.
Recombinant Protein Expression and Purification
Recombinant, N-terminal hexahistidine-tagged BFL-1DC (aa 1-151) and the indicated cysteine-mutant analogs generated by Q5
site-directed mutagenesis (New England Biolabs) were cloned into pET17b (Novagen), expressed in E. coli LOBSTR BL21(DE3) cells
(Kerafast) overnight at 16ꢂC, and purified by sequential Ni-affinity and size-exclusion chromatography (Harvey et al., 2018). Protein
expression was induced using 0.5 mM isopropyl b-D-1-thiogalactopyranoside (IPTG, Gold Biotechnology). Bacterial pellets were re-
suspended in lysis buffer (20 mM Tris pH 7.5, 250 mM NaCl) containing two complete protease inhibitor tablets (Roche). Bacteria
were lysed using a microfluidizer (M-110L, Microfluidics) and centrifuged at 20,000 x RPM for 45 minutes to remove insoluble debris.
Clarified lysate was passed over a Ni-NTA (Qiagen) column equilibrated with lysis buffer. The column was sequentially washed with
lysis buffer and then lysis buffer containing 10 mM, 20 mM, 35 mM, and 50 mM imidazole. His-BFL-1DC was eluted in buffer con-
taining 150 mM imidazole. The BFL-1 containing fraction was then dialyzed overnight in lysis buffer, concentrated, and purified by
size exclusion chromatography using a Superdex S-75 (GE Healthcare) gel filtration system. Protein purity and identity was confirmed
by Coomassie staining and western blot analysis using a mouse monoclonal anti-His6 tag antibody (Abcam, RRID: Ab_444306) and
sheep anti-mouse secondary antibody (Biorad, RRID: 321929).
Recombinant anti-apoptotic BCL-XLDC (aa 1-212) and MCL-1DNDC (aa 172-329) were cloned into pGEX-4T-1 (GE Healthcare),
expressed in E. coli BL21(DE3) cells (Invitrogen) for four hours at 37ꢂC, and purified by sequential glutathione affinity and size exclu-
sion chromatography (Harvey et al., 2018; Huhn et al., 2016; Pitter et al., 2008). After bacterial lysis and centrifugation, GST-tagged
BCL-XLDC and MCL-1DNDC were purified using a glutathione chromatography column equilibrated with PBS containing 1% Triton
X-100. The column was washed twice with PBS/1% Triton X-100 and then two more times with PBS alone. BCL-XLDC and
MCL-1DNDC were cleaved from their GST tags using thrombin (12-15 units) dissolved in PBS. Eluted protein was concentrated
and purified by size exclusion chromatography using a Superdex S-75 (GE Healthcare) gel filtration system. Protein purity and
identity was confirmed by Coomassie staining and western blot analysis using a goat polyclonal anti-GST antibody (GE Healthcare,
RRID: 771432) and donkey anti-goat secondary antibody (Santa Cruz Biotechnology, RRID: Ab_631728).
Disulfide Tethering Fragment Screening by Mass Spectrometry
A 1600-member disulfide fragment library (Burlingame et al., 2011; Hallenbeck et al., 2018) was screened by incubating BFL-1DC
C4S/C19S protein (1 mM) with each compound (100 mM) for 1 hour at room temperature in the presence of 500 mM BME (Thermo
Fisher Scientific). Each well was then screened by UPLC-MS for a change in mass that corresponded with the individual fragment.
The compounds were then ranked based on percent tethering to C55 by comparing the relative abundance of conjugated vs. uncon-
jugated BFL-1DC C4S/C19S protein. The criteria for pursuing a hit was set at a percent tethering value of at least two standard
deviations above the mean for the 1600 fragments tested.
Solid Phase Synthesis of Disulfide Tethering Fragments
Disulfide tethering fragments were synthesized in accordance with the scheme presented in Figure 4A. Briefly, 0.5 mmol Rink Amide
AM resin was rinsed once with dimethylformamide (DMF) and then dichloromethane (DCM), deprotected with 20% piperidine in
DMF for 10 minutes, and then washed three times with DMF. Glutarylation was performed by adding 2.5 mmol (5 eq) glutaric anhy-
dride dissolved in 5 mL DMF to the deprotected resin, stirring the mixture for 45 minutes, and then washing the resin three times with
DMF. Cysteamine was then used to amidate the glutarylated resin as follows: N,N-diisopropylethylamine (DIEA; 1.5 mmol [3 eq]) in
5 mL DCM was added to the resin, stirred for 5 minutes, and removed by vacuum filtration; the resin was then mixed with HATU
(1.5 mmol [3 eq]) in 5 mL DMF, the solution removed after 10 min, and then cysteamine HCl (2.5 mmol [5 eq]) added with 4 mL
DMF, 0.33 mmol DIEA, and water (dropwise) until the solids dissolved. The mixture was stirred for 45 minutes and the resin washed
three times with DMF. To reverse any oxidation of the nascent free thiol, TCEP reduction of the resin was performed for 30 minutes
using TCEP HCl (0.5 mmol) dissolved in a degassed solution of 1:1 DMF:water. To remove residual TCEP and cysteamine, the resin
was washed three times with the following sequence of solvents: DMF, methanol, water, methanol, DMF. Next, the thiol was con-
verted to an activated sulfide by mixing the resin with 2,2’dithiodipyridine (2.5 mmol [5 eq]) dissolved in 5 mL DMF for 20 minutes
to afford the 2-thiopyridyl disulfide adduct. To remove excess 2-thiopyridyl disulfide, the resin was washed 5 times with a sequence
of DMF and DCM. Next, cysteamine HCl (1 mmol [2 eq]) dissolved in 5 mL of DMF, with minimal water (dropwise addition) for sol-
ubility, was used to exchange disulfides for 30 minutes. Following disulfide exchange, the resin was washed 5 times with a sequence
of DMF and DCM. At this step of the synthesis, the resin was divided for monophore diversification. Monophore core conjugation to
the tether was carried out by adding 3 equivalents of the appropriate carboxylic acid ((S)-2-acetamido-3-(1H-indol-3-yl)propanoic
acid for 4E14; (R)-2-acetamido-3-(1H-indol-3-yl)propanoic acid for D-4E14; 2-acetamido-3-(5-methyl-1H-indol-3-yl)propanoic
Cell Chemical Biology 27, 1–10.e1–e6, June 18, 2020 e3

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
acid for 5-Me-4E14; and (R)-2-acetamido-3-(naphthalen-2-yl) propanoic acid for D-Nal-4E14) and 3.5 eq of HATU and 3.5 eq of DIEA
in 5 mL DMF to the resin, followed by mixing overnight for 16 hours. Cleavage of disulfide tethering fragments from the resin was
performed using Bis(2-dimethylaminoethyl) disulfide dihydrochloride (1.75 mmol [3.5 eq]) dissolved in a degassed solution of
DMF:water (5 mL, 1:1), triethylamine (TEA) (3.5 mmol [7 eq]), and TCEP HCl (0.25 mmol [0.5 eq]). Compounds were purified on an
Agilent 1260 Infinity HPLC-MS using a C18 column and a 5-95% acetonitrile gradient in water with 0.1% formic acid, and identities
confirmed by high-resolution MS and proton (1H) NMR. 1H NMR was performed using a 500 MHz Bruker Avance III instrument
outfitted with a BBFO room-temperature probe. 1H frequencies were referenced to tetramethylsilane (TMS). Proton spectra were ac-
quired with a 45ꢂ pulse and a 4.3-s recycle delay at 0.3 Hz per point resolution.
(S)-2-acetamido-N-(2-((2-(dimethylamino)ethyl)disulfanyl)ethyl)-3-(1H-indol-3-yl)propanamide (4E14). 1H NMR (500MHz,
MeCN-d3) d = 9.17 (br. s., 1 H), 7.52 (d, J = 7.9 Hz, 1 H), 7.31 (d, J = 8.2 Hz, 1 H), 7.06 - 7.03 (m, 2 H), 6.99 - 6.95 (m, 1 H), 6.85
(br. s., 1 H), 6.66 (d, J = 7.3 Hz, 1 H), 4.48 - 4.43 (m, 1 H), 3.35 - 3.26 (m, 2 H), 3.11 (dd, J = 6.0, 14.5 Hz, 1 H), 3.00 (dd, J = 7.3,
14.6 Hz, 1 H), 2.84 (d, J = 9.8 Hz, 4 H), 2.68 - 2.57 (m, 2 H), 2.39 (s, 6 H), 1.77 (s, 3 H).
(R)-2-acetamido-N-(2-((2-(dimethylamino)ethyl)disulfanyl)ethyl)-3-(1H-indol-3-yl)propanamide (D-4E14). 1H NMR (500MHz,
MeCN-d3) d = 9.14 (br. s., 1 H), 7.51 (d, J = 7.9 Hz, 1 H), 7.30 (d, J = 7.9 Hz, 1 H), 7.06 - 7.01 (m, 2 H), 6.98 - 6.94 (m, 1 H), 6.78
(br. s., 1 H), 6.60 (d, J = 7.3 Hz, 1 H), 4.46 - 4.42 (m, 1 H), 3.35 - 3.23 (m, 2 H), 3.09 (dd, J = 6.0, 14.5 Hz, 1 H), 2.98 (dd, J = 7.5,
14.8 Hz, 1 H), 2.82 - 2.76 (m, 2 H), 2.67 - 2.62 (m, 2 H), 2.62 - 2.53 (m, 2 H), 2.26 (s, 6 H), 1.76 (s, 3 H).
2-acetamido-N-(2-((2-(dimethylamino)ethyl)disulfanyl)ethyl)-3-(5-methyl-1H-indol-3-yl)propanamide (5-Me-4E14). 1H NMR (500
MHz, DMSO-d6) d = 10.65 (d, J = 2.4 Hz, 1H), 8.30 (s, 1H), 8.13 (t, J = 5.7 Hz, 1H), 8.02 (d, J = 8.2 Hz, 1H), 7.35 (dd, J = 1.7,
0.9 Hz, 1H), 7.20 (d, J = 8.1 Hz, 1H), 7.06 (d, J = 2.3 Hz, 1H), 6.88 (dd, J = 8.3, 1.6 Hz, 1H), 4.44 (td, J = 8.5, 5.4 Hz, 1H), 3.39 -
3.25 (m, 2H), 3.05 (dd, J = 14.5, 5.4 Hz, 1H), 2.87 (d, J = 8.7 Hz, 1H), 2.86 - 2.80 (m, 2H), 2.75 - 2.63 (m, 2H), 2.48 (s, 1H), 2.38 (s,
3H), 2.15 (s, 6H), 1.79 (s, 3H).
1
(R)-2-acetamido-N-(2-((2-(dimethylamino)ethyl)disulfanyl)ethyl)-3-(naphthalen-2-yl)propanamide (D-Nal-4E14). H NMR (500 MHz,
DMSO-d6) d = 8.37 (s, 3H), 8.28 - 8.14 (m, 1H), 7.90 - 7.77 (m, 3H), 7.75 - 7.68 (m, 1H), 7.53 - 7.38 (m, 3H), 6.28 (s, 1H), 4.55 (td,
J = 8.9, 5.3 Hz, 1H), 3.18 (s, 2H), 3.12 (dd, J = 13.7, 5.3 Hz, 3H), 2.95 - 2.87 (m, 2H), 2.87 - 2.77 (m, 2H), 2.74 - 2.62 (m, 2H), 2.47
(d, J = 6.9 Hz, 2H), 2.39 - 2.31 (m, 1H), 2.18 - 2.09 (m, 6H), 1.80 - 1.73 (m, 3H).
Synthesis of 4E14-2-thiopyridine
4E14-2-thiopyridine was generated to maximize 4E14 conjugation to BFL-1DC C4S/C19S for X-ray crystallography by replacing the
N,N-dimethylcysteamine leaving group with the more efficient 2-mercaptopyridyl leaving group (i.e. molecule:protein ratio reduced
to 4:1, allowing for decreased DMSO content in droplets). The synthesis of 2-(pyridin-2-yldisulfanyl)ethanamine was performed as
previously reported (Lelle et al., 2017). Briefly, 2.5 g of 2,2’-dithiodipyridine (11.5 mmol, 5.75 eq) in 8 mL of methanol was degassed
followed by dropwise addition under nitrogen of a cysteine solution, which was generated by adding 225 mg of cysteamine hydro-
chloride (2 mmol) to 8 mL of degassed methanol. The mixture was stirred at room temperature overnight. The solution was concen-
trated in vacuo, resuspended in 5 mL methanol, and then precipitated three times from 100 mL ice cold ether. The product, an off-
white solid, was collected via filtration (260 mg, 62% yield). The NMR spectrum matched the published data (Lelle et al., 2017). To
generate (S)-2-acetamido-3-(1H-indol-3-yl)-N-(2-(pyridin-2-yldisulfanyl)ethyl) propanamide (4E14-2-thiopyridine), 100 mg of 2-(pyr-
idin-2-yldisulfanyl)ethanamine (537 mmol, 1 eq) was dissolved in 5 mL DMF, followed by 100 mg N-acetyltryptophan (400 mmol, 0.75
eq), 220 mg HATU (580 mmol, 0.9 eq), and 0.25 mL of N,N-diisopropylethylamine (1.5 mmol, 3 eq). The mixture was stirred overnight
at room temperature. The DMF solution was poured into 50 mL of ethyl acetate and washed 3 times with 50 mL HCl (0.2 M). The
organic layer was dried over sodium sulfate, concentrated, and HPLC purified as described above to afford 46 mg of yellow oil
(25% yield). 1H NMR (500 MHz ,DMSO-d6) d = 10.81 (br. s., 1 H), 8.47 (d, J = 4.3 Hz, 1 H), 8.19 (t, J = 5.5 Hz, 1 H), 8.07 (d, J =
7.9 Hz, 1 H), 7.81 (dd, J = 1.7, 7.5 Hz, 1 H), 7.78 - 7.72 (m, 1 H), 7.58 (d, J = 7.6 Hz, 1 H), 7.32 (d, J = 8.2 Hz, 1 H), 7.25 (dd, J =
4.9, 6.7 Hz, 1 H), 7.14 (d, J = 2.1 Hz, 1 H), 7.09 - 6.92 (m, 2 H), 4.47 (d, J = 5.5 Hz, 1 H), 3.38 - 3.25 (m, 2 H), 3.10 (dd, J = 5.3,
14.5 Hz, 1 H), 2.96 - 2.87 (m, 1 H), 2.84 - 2.72 (m, 2 H), 1.80 (s, 3 H). 13C NMR (126 MHz, DMSO-d6) d = 172.3, 169.6, 159.6,
150.1, 138.3, 136.5, 127.8, 124.0, 121.7, 121.3, 119.7, 118.9, 118.6, 111.8, 110.7, 53.8, 38.4, 37.6, 28.4, 23.1
Fluorescence Polarization Assays
The indicated BFL-1DC constructs, MCL-1DNDC, or BCL-XLDC (1 mM) were incubated with 25x 4E14 for one hour at room temper-
ature. The proteins were then serially diluted into fluorescence polarization assay buffer (50 mM Tris pH 8, 100 mM NaCl) and incu-
bated with FITC-peptide (60 nM). Fluorescence polarization was measured at equilibrium using a SpectraMax M5 microplate reader,
and nonlinear regression analysis of dose-response curves was performed using Prism software 7 (GraphPad).
Mitochondrial Cytochrome c Release Assays
Liver mitochondria from AlbCreBaxf/fBak-/- mice were isolated by dounce homogenization and release assays performed as described
(Walensky et al., 2006). Briefly, the indicated BFL-1DC construct was incubated with 4E14 or its analogs (molecule: protein, 5:1) or
vehicle for 1 hour and then desalted into BFL-1 FPLC buffer using Micro Bio-Spin 6 columns (Biorad) to remove excess 4E14. Mito-
chondria (1 mg/mL) were incubated with recombinant BAX (100 nM) in the presence or absence of tBID (40 nM) and BFL-1 protein
(1 mM) for 45 minutes at room temperature in experimental buffer (200 mM mannitol, 68 mM sucrose, 10 mM HEPES-KOH pH 7.4,
110 mM KCl, 1 mM EDTA, protease inhibitor). The pellet and supernatant fractions were isolated by centrifugation, and cytochrome c
e4 Cell Chemical Biology 27, 1–10.e1–e6, June 18, 2020

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
was quantitated using a colorimetric Rat/Mouse Cytochrome c Quantikine ELISA assay (R&D Systems), per the manufacturer’s
protocol. Briefly, cytochrome c conjugate (75 mL) and sample (50 mL) were added to the ELISA plate and incubated at room temper-
ature for 2 hours. Each well was then aspirated and washed four times using the manufacturer-supplied wash buffer. Substrate
solution (100 mL) was added to the plate, which was incubated in the dark for 30 minutes. Stop solution (100 mL) was then added
and absorbance read at 450 nm with wavelength correction at 540 nm using a SpectraMax M5 microplate reader. Percent
cytochrome c released into the supernatant (%cyto c release) was calculated according to the following equation: %cyto c release =
[cyto csup]/ [cyto cmax]*100, where cyto csup and cyto cmax represent the amount of cytochrome c detected in the supernatant upon
treatment with the indicated conditions or 1% (v/v) Triton X-100, respectively.
X-ray Crystallography
Apo His6-BFL-1DC C4S/C19S was expressed and purified as described (Harvey et al., 2018) and buffer exchanged into 20 mM
HEPES pH 7.5, 300 mM NaCl, 10% glycerol, and 50 mM arginine. His6-BFL-1DC C4S/C19 in complex with 4E14 was generated
by incubating 4E14-2-thiopyridine with His-BFL-1DC C4S/C19 (250 mM) at room temperature for 1 hour. Intact mass spectrometry
was used to confirm complete BFL-1 labeling by 4E14. An equal volume (100 nL) of 4.65 mg/mL (250 mM) His6-BFL-1DC C4S/C19S–
4E14 complex was mixed with reservoir buffer (2.8 M sodium acetate trihydrate, pH 7.0), and crystals were prepared in hanging
drops at 20ꢂC. The crystals were transferred into crystallization buffer containing 25% glycerol prior to flash-freezing in liquid nitro-
gen. Diffraction data from crystals of His6-BFL-1DC C4S/C19S conjugated to fragment 4E14 were collected at beamline 24ID-C of
the NE-CAT at the Advanced Photon Source (Argonne National Laboratory), and data sets were integrated and scaled with the XIA2
package, which uses the programs Pointless and XDS (Evans, 2006; Kabsch, 2010; Winter et al., 2010). The structure was solved by
molecular replacement using the program Phaser (McCoy et al., 2007) and the search model PDB: 5WHH. Iterative manual model
building and refinement using Phenix (Adams et al., 2010) and Coot (Emsley and Cowtan, 2004) led to a model with excellent statis-
˚
tics, including maximum diffraction of 1.74 A (Table 1, PDB: 6VO4).
Molecular Dynamics Simulation and Docking
Computational analyses were performed using the Schrodinger software suite (Version 2016-2). The 4E14 molecule was built
¨
and conformations were generated using MacroModel and the OPLS3 forcefield (Harder et al., 2016). BFL-1DC (PDB: 3MQP)
was prepared using default parameters in the PrepWiz wizard in Maestro (Sastry et al., 2013). Briefly, the bound peptide and all water
molecules were removed, the H-bond network of the protein was optimized, and the protein was subjected to an impref minimization
using the OPLS3 forcefield. Protonation states were those predicted to occur at pH 7.0 using the Epik module (Shelley et al., 2007).
The receptor grid (radius 1 nm) was defined at the center of C55. Docking was performed using the pose prediction mode of CovDock
¨
in Schrodinger, with the protein structure minimized in a 0.3 nm radius surrounding the docked molecule (Zhu et al., 2014).
Molecular dynamics calculations were performed using the 4E14/BFL-1 complex generated by molecular docking. The protein
¨
was prepared using the default parameters of the Protein Preparation Workflow in Maestro software (Schrodinger version 2016-
2). Protonation states were those predicted to occur at pH 7.0 using the Epik module. Each protein was pre-soaked in a cubic
box of TIP3P water molecules using the System Builder workflow in Desmond. The box was sized to leave all peptide atoms at least
1 nm from the boundaries. All overlapping solvent molecules were removed, the system was charge neutralized with appropriate
counterions and 150 mM NaCl was added to simulate buffer conditions. All MD simulations were performed using the Desmond
package (Jorgensen et al., 1983), with the OPLS3 forcefield used to model all interactions. Periodic boundary conditions were main-
tained throughout. Long-range electrostatic interactions were calculated using the particle-mesh Ewald method (Essmann et al.,
1995), and van der Waals and short-range electrostatic interactions were smoothly truncated at 0.9 nm. Constant system temper-
ature of 300 K was maintained using Nose-Hoover thermostats (Hoover, 1985), and system pressure was maintained at 1 atm using
the Martyna-Tobias-Klein method (Martyna et al., 1994). The equations of motion were integrated using the RESPA integrator, with a
2.0 fs timestep for bonded and short-range interactions and a 6.0 fs timestep for non-bonded interactions beyond the 0.9 nm cutoff.
The default parameters in Desmond were used to relax the system prior to simulation (Guo et al., 2010) and a 100 ns production simu-
lation was then run. All simulations were judged to have converged on the basis of radius of gyration calculations and RMSD.
Hydrogen Deuterium Exchange Mass Spectrometry
His-BFL-1DC C4S/C19S or His-BFL-1DC C55S (25 mM) were conjugated with 4E14 (20x) or vehicle for 1 hour at room temperature.
Samples (25 pmol, 1 mL) were deuterated for the indicated time points using an 18-fold excess of labeling buffer (50 mM Tris, 150 mM
NaCl, pD 7.6), and labeling was stopped with an equal volume of quench buffer (4M guanidinium chloride, 200 mM potassium phos-
phate, 0.72 M TCEP, H2O, pH 2.34) (Table S2 and Data S1). Proteolysis of BFL-1 protein was conducted using an online pepsin diges-
tion column. Digested protein was injected into a Waters UPLC HDX system (Wales et al., 2008) maintained at 0ꢂC to minimize back-
exchange. Peptic peptides underwent a 3 min trap and desalting step using a VanGuard pre-column trap (2.1 x 5 mm, ACQUITY
UPLC BEH C18, 1.7 mm) flowing at 100 mL/min. Peptides were then separated using an ACQUITY UPLC HSS T3, 1.8 mm, 1.0 x
50 mm analytical column (Waters Corporation) with a 5-35% gradient of increasing acetonitrile (0.1% formic acid) over 6 min at a
flow rate of 100 mL/min. Mass spectra were acquired using a Synapt G2-Si (Waters) in MSE (data independent acquisition) mode
with 0.4 scans/sec over a 50-2000 m/z range with ion mobility engaged. BFL-1 peptides were identified using Protein Lynx Global
Server (PLGS 3.0.1, Waters Corporation, RRID: SCR_016664) with undeuterated BFL-1 protein as a reference. Data analysis was
performed using DynamX 3.0 (Waters Corporation), which included determination of centroid masses for each isotopic distribution,
Cell Chemical Biology 27, 1–10.e1–e6, June 18, 2020 e5

Please cite this article in press as: Harvey et al., Identification of a Covalent Molecular Inhibitor of Anti-apoptotic BFL-1 by Disulfide Tethering, Cell
Chemical Biology (2020), https://doi.org/10.1016/j.chembiol.2020.04.004
generation of deuterium uptake plots, and comparison between experimental states. Relative deuterium uptake for each
peptide was determined by subtracting the average mass of each undeuterated peptide from the average mass of each deuterated
peptide. Deuterium exchange was not corrected for back-exchange and thus expressed as relative deuterium uptake (Wales and
Engen, 2006).
QUANTIFICATION AND STATISTICAL ANALYSIS
The number of technical and biological replicates for each experiment are indicated in the corresponding figure legend. Mean S.D.
or S.E.M. values were calculated using Prism software (Graphpad).
e6 Cell Chemical Biology 27, 1–10.e1–e6, June 18, 2020