Rationally Designed Polypharmacology: α-Helix Mimetics as Dual Inhibitors of the Oncoproteins Mcl-1 and HDM2

Article abstract of DOI:10.1002/cmdc.202000278

Protein–protein interactions (PPIs), many of which are dominated by α-helical recognition domains, play key roles in many essential cellular processes, and the dysregulation of these interactions can cause detrimental effects. For instance, aberrant PPIs involving the Bcl-2 protein family can lead to several diseases including cancer, neurodegenerative diseases, and diabetes. Interactions between Bcl-2 pro-life proteins, such as Mcl-1, and pro-death proteins, such as Bim, regulate the intrinsic pathway of apoptosis. p53, a tumor-suppressor protein, also has a pivotal role in apoptosis and is negatively regulated by its E3 ubiquitin ligase HDM2. Both Mcl-1 and HDM2 are upregulated in numerous cancers, and, interestingly, there is crosstalk between both protein pathways. Recently, synergy has been observed between Mcl-1 and HDM2 inhibitors. Towards the development of new anticancer drugs, we herein describe a polypharmacology approach for the dual inhibition of Mcl-1 and HDM2 by employing three densely functionalized isoxazoles, pyrazoles, and thiazoles as mimetics of key α-helical domains of their partner proteins.

Full text of DOI:10.1002/cmdc.202000278

Communications
doi.org/10.1002/cmdc.202000278
ChemMedChem
1
2
3
4
5
6
7
8
9
Rationally Designed Polypharmacology: α-Helix Mimetics
as Dual Inhibitors of the Oncoproteins Mcl-1and HDM2
Ivie L. Conlon,[a] Brandon Drennen,[a] Maryanna E. Lanning,[a] Samuel Hughes,[b]
Rebecca Rothhaas,[a] Paul T. Wilder,[c] Alexander D. MacKerell, Jr.,[a] and Steven Fletcher*[a]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
molecules, several of which in various stages of clinical
trials.[10–13] Tools such as alanine scanning mutagenesis have
been instrumental in identifying key residues located in “hot
spots” that account for a large degree of the energetic stability
of the complexes. In α-helix-mediated PPIs, these “hot spots”
often involve the i, i+3/4, and i+7 residues that are oriented
on one side of the helix.[14–16] The B-cell lymphoma-2 (Bcl-2)
family and human double minute 2 (HDM2) engage in such α-
helix-mediated PPIs and their aberrant regulations are directly
associated with tumorigenesis.[17,18]
The Bcl-2 family of proteins are key regulators of cell life
and death, and constitute the intrinsic pathway of
apoptosis.[17,19] The family comprises three subgroups: the anti-
apoptotic proteins myeloid cell leukemia-1 (Mcl-1), Bcl-2, Bcl-xL,
Bcl-w and Bfl-1; multidomain pro-apoptotic proteins Bak and
Bax; and BH3-only pro-apoptotic proteins including Bim, Bid,
PUMA, and NOXA. When the cell receives apoptotic stimuli due
to different stressors such as DNA damage or growth-factor
deprivation, BH3-only proteins are upregulated and, through
their α-helical BH3 “death” domains (Figure 1A), bind the anti-
apoptotic proteins to release Bak and Bax, leading to homo-
dimerization of these proteins.[17,19] In turn, this leads to
mitochondrial outer-membrane permeabilization (MOMP), ulti-
mately allowing the release of apoptosis-inducing proteins such
as cytochrome c.[17,19] During the formation of malignancies
dependent on Bcl-2 proteins, the anti-apoptotic proteins are
overexpressed and the BH3-only proteins are unable to
compensate, causing cell immortality.[17–20]
HDM2 is an E3 ubiquitin protein ligase responsible for the
degradation of the tumor-suppressor protein p53 through
binding its α-helical transactivation domain (TAD; Fig-
ure 1C).[21,22] p53 is a short-lived protein whose concentration at
any given time is controlled by the rate at which it is
degraded.[23] It is coined the “guardian of the genome” due to
its involvement in a variety of cellular pathways, notably the
transcription of genes during cell stress events involved in
apoptosis, DNA repair, senescence, and metabolic homeostasis.
In turn it is negatively regulated by HDM2.[22,24] During tumori-
genesis, HDM2 can be overexpressed, thus inhibiting the
activation of p53 and decreasing cell death.[25] p53 regulates
apoptosis with both the transcription-dependent and -inde-
pendent pathways, crossing paths with Bcl-2 proteins.[23] Within
the transcription-dependent pathway, it transcribes genes
necessary for cell senescence and apoptosis such as p21 and
the BH3-only protein PUMA.[18] Importantly, this pathway
involves the translocation of p53 from the nucleus to the
mitochondria, allowing interactions with various members of
Protein–protein interactions (PPIs), many of which are domi-
nated by α-helical recognition domains, play key roles in many
essential cellular processes, and the dysregulation of these
interactions can cause detrimental effects. For instance, aber-
rant PPIs involving the Bcl-2 protein family can lead to several
diseases including cancer, neurodegenerative diseases, and
diabetes. Interactions between Bcl-2 pro-life proteins, such as
Mcl-1, and pro-death proteins, such as Bim, regulate the
intrinsic pathway of apoptosis. p53, a tumor-suppressor protein,
also has a pivotal role in apoptosis and is negatively regulated
by its E3 ubiquitin ligase HDM2. Both Mcl-1 and HDM2 are
upregulated in numerous cancers, and, interestingly, there is
crosstalk between both protein pathways. Recently, synergy has
been observed between Mcl-1 and HDM2 inhibitors. Towards
the development of new anticancer drugs, we herein describe a
polypharmacology approach for the dual inhibition of Mcl-1
and HDM2 by employing three densely functionalized isoxa-
zoles, pyrazoles, and thiazoles as mimetics of key α-helical
domains of their partner proteins.
Introduction
Protein–protein interactions (PPIs) are involved in crucial cellular
pathways including proliferation, differentiation, and
apoptosis.[1,2] Dysregulations of PPIs cause a host of different
diseases, including cancer, neurodegenerative disorders, auto-
immunity complications and diabetes.[1–5] Typically, PPI inter-
faces cover large surface areas and consist of flat hydrophobic
regions with noncontiguous contact points, rendering drug
design challenging.[6–9] However, many academic and industrial
groups have successfully targeted these interactions with small
[a] I. L. Conlon, Dr. B. Drennen, Dr. M. E. Lanning, R. Rothhaas,
Dr. A. D. MacKerell, Jr., Dr. S. Fletcher
Department of Pharmaceutical Sciences
University of Maryland School of Pharmacy
Baltimore, MD 21201 (USA)
E-mail: steven.fletcher@rx.umaryland.edu
[b] S. Hughes
School of Chemistry, Cardiff University
Cardiff, CF10 3AT (UK)
[c] Dr. P. T. Wilder
Department of Biochemistry and Molecular Biology
Center for Biomolecular Therapeutics
University of Maryland School of Medicine
Baltimore, MD 21201 (USA)
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/cmdc.202000278
ChemMedChem 2020, 15, 1–9
1
© 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
doi.org/10.1002/cmdc.202000278
ChemMedChem
Additionally, a recent study involving the Mcl-1 selective
1
2
3
4
5
6
7
8
9
inhibitor S63845 and the HDM2 inhibitor HDM201 evaluated
the cell viability of combination treatments in AML cells.[41]
MOLM-13 and OCI-AML-3 cells were susceptible to S63845
monotherapy with 60–70% viability, while OCI-AML-2 cells were
95% viable. In contrast, OCI-AML-2 AND MOLM-13 cells were
susceptible to HDM201 monotherapy with 80% viability, where-
as OCI-AML-3 showed no reduction. All three cell lines showed
synergistic effects, with the combination index (CI)=0.2–0.5
when treated with both drugs, while OCI-AML-3 exhibited a
strong synergy effect with CI=0.2–0.3.[41] These studies repre-
sent the efficacy of combination therapies of an anti-apoptotic
protein inhibitor with an HDM2 inhibitor.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
An alternative strategy to these polypharmacy regimens is
polypharmacology, in which one drug is fashioned to recognize
multiple targets. Polypharmacology is an emerging field of
research and could provide multiple benefits to the patient,
including increased patient compliance, eliminated drug-drug
interactions, and reduced side effects, not only through a
reduction in drug cocktail complexity but also broadened
therapeutic windows through greater therapeutic efficacies
with reduced doses.[42,43] In light of these studies, coupled with
the multifactorial nature of cancer that suggests multiple drugs,
or a single drug presenting multiple pharmacophores, will be
required for an effective pharmacotherapy, we herein describe
a polypharmacology approach to address the unmet need of
Mcl-1 and HDM2 inhibitors in the clinic using synthetic α-helix
mimicry.
Figure 1. A) Co-crystal structure of Bim-BH3 bound to Mcl-1 (PDB ID: 2NL9).
B) Hot spot residues highlighted in Bim-BH3 peptide. C) Co-crystal structure
of p53TAD bound to HDM2 (PDB ID: 1YCR). D) Hot spot residues highlighted
in p53TAD.
the Bcl-2 family.[26] p53 can bind anti-apoptotic proteins directly
or disrupt pro- and anti-apoptotic interactions such as the Bak/
Mcl-1.[27] Additionally, it can directly bind Bax, which can then
translocate and homodimerize on the mitochondrial outer
membrane, leading to MOMP.[17,23,27–29] Together these interac-
tions indirectly cause apoptosis within the cell.[20,23,24]
Consequently, these proteins have been the target of
intense medicinal chemistry efforts towards the discovery of
new treatments for cancer.[22,30–35] Although clinical trials are on-
going, to date there are no FDA-approved Mcl-1 or HDM2 Results and Discussion
inhibitors, with only one Bcl-2 selective inhibitor, venetoclax,
that has reached the clinic.[36] However, studies have shown that
venetoclax-treated cancer cells develop chemoresistance by
overexpressing other anti-apoptotic proteins, most notably Mcl-
1, re-affirming the unmet medical need of developing Mcl-1
inhibitors.[37,38]
Both the Bim-BH3 and p53TAD α-helices project similar hydro-
phobic character at “hot spot” residues i, i+3/4, and i+7:
Leu62, Ile65, and Phe69 in Bim (Figure 1B), and Phe19, Trp23,
and Leu26 in p53TAD (Figure 1D).[44–46] Importantly, there is a
conserved arginine residue on pro-life proteins (Arg263 in Mcl-
1) that forms a salt bridge with Asp67 on the Bim α-helix at the
i+5 position.[47] Additionally, there is an aspartate residue in an
analogous location (i+2) in p53TAD (Asp21), that helps
maintain the integrity of the helix rather than engage in
recognition.[48] We set out to capitalize on these similarities
across both α-helices towards the discovery of dual inhibitors of
Mcl-1 and HDM2.
Monocyclic scaffolds such as terphenyl, imidazoline, and
pyrazole scaffolds have been used as scaffolds to effectively
mimic either the BH3 or the p53 α-helices.[4,13,49] Roche
introduced small molecules based on cis-imidazolines as
inhibitors of HDM2, leading to the discovery of the nanomolar
inhibitor Nutlin-3a.[50] As evidenced by a co-crystal structure of
Nutlin-3a and HDM2, the isopropoxyphenyl group mimics
Phe19, while the two 4-chlorophenyl groups projected in a cis-
1,2-fuctionalization from the imidazoline scaffold mimic Trp23
and Leu26 (PDB ID: 4 J3E; Figure 2).[22]
In a phase II study evaluating venetoclax in patients with
relapsed/refractory acute myeloid leukemia (AML), a 19%
response rate was observed. 34% of patients in this study
escalated to 1200 mg of venetoclax due to lack of response at
the initial 800 mg dose, which did not achieve additional
response.[39] Next, a phase Ib clinical trial was employed to
assess a combination therapy of venetoclax and idasanutlin, an
HDM2 inhibitor, in relapsed/refractory acute myeloid leukemia
(AML) due to the modest monotherapy response rate. This
combination was explored due to the overexpression of Mcl-1
in venetoclax-resistant cells, compounded with the knowledge
that the inhibition of HDM2 and subsequent activation of p53
can result in Mcl-1 degradation. They reported 35.9% of
patients responded to treatment. Of 39 patients treated, 14
exhibited antileukemic response which include complete and
partial remission.[40] Currently, a phase I/II study is ongoing to
determine the safety, tolerability, pharmacokinetics of the
combination therapy in pediatric and young adult patients with
relapsed/refractory acute leukemias or solid tumors
(NCT04029688).
Given Nutlin 3a’s effective mimicry of the p53TAD helix,
which resembles the Bim-BH3 helix, perhaps it is unsurprising
that it also exhibits low micromolar inhibition of Bcl-2.[22] In
ChemMedChem 2020, 15, 1–9
www.chemmedchem.org
2
© 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
doi.org/10.1002/cmdc.202000278
ChemMedChem
including imidazoles, pyrazoles, and pyrroles. The most potent
1
2
3
4
5
6
7
8
9
compounds project aryl functional groups in a 1,2,3-substitution
pattern, indicating a clustered projection appears optimal.[51]
PDB ID 4OQ3 illustrates the co-crystal structure of one of these
imidazole-based inhibitors with HDM2 (Figure 3). The three
functional groups engage in multiple contacts with the protein
with each functional group projected into the respective
pockets: 3-chlorophenyl in a π–π interaction with His96 while
mimicking Leu26, 2-methyl, 5-chlorophenyl sitting in the Trp23
pocket, and the 3-methylphenyl functional group mimicking
Phe19.[46] More recently, Zhang et al. developed a dual Bcl-2/
HDM2 inhibitor based on a pyrazole scaffold (Figure 4, center
structure). As shown in Figure 3, their work is very closely
related to the corresponding pyrazoles from Novartis (Figure 4,
left structure) with a noticeable departure being the incorpo-
ration of an amide group between the pyrazole core and one of
the aryl side chains. The authors reasoned this flexible amide
bond was incorporated to promote mimicry of the less-
structured p53TAD α-helix, while retaining effective mimicry of
the BH3 α-helix. The most potent inhibitor of Bcl-2, Mcl-1, and
HDM2 achieved affinities of Ki =0.140, 0.161, and 0.107 μM,
respectively.[45]
Figure 2. Nutlin 3a co-crystallized with HDM2 (PDB ID: 4 J3E).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
patent CA2771936 A1, Novartis describe sub-nanomolar inhib-
itors of HDM2 based on tetra-substituted heteroaryl scaffolds
Leveraging the successes from Novartis and Zhang with
heavily functionalized pyrazoles to inhibit the Bcl-2 family of
proteins and HDM2, coupled with the effective deployment of
the acyl sulfonamide functional group as a carboxylic acid
bioisostere in Mcl-1 inhibitors elsewhere, we designed a library
of three novel, densely functionalized scaffolds to mimic the
BH3 binding domain and p53TAD: isoxazoles, pyrazoles, and
thiazoles (Figure 4, right structure), which are all present in
Figure 3. Co-crystal structure of Novartis compound and HDM2 (PDB ID:
4OQ3). Residue labels indicate the binding locations of the key residues in
the p53TAD helix.
Figure 4. Representative structures of Novartis patent CA2771936 A1 (left), Zhang et al. Dual Bcl-2/HDM2 inhibitor (center), and our work presented in this
manuscript (right). Normal font represents p53TAD residues; bold, italic font represents Bim-BH3 residues.
ChemMedChem 2020, 15, 1–9
www.chemmedchem.org
3
© 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
doi.org/10.1002/cmdc.202000278
ChemMedChem
pharmacologically active drug molecules.[52–55] Each scaffold will
allow us to explore the protein binding interfaces with different
substitution patterns: isoxazoles represent a 1,2,3-functionaliza-
tion (with the numbering starting with the acyl substituent),
pyrazoles represent a 1,3,4-functionalization, and thiazoles a
1,2,4-functionalization. In addition to sustaining the acidity
needed to engage Arg263 of Mcl-1, incorporation of the acyl
sulfonamide functional group (pKa 5) – akin to the amide in
Zhang’s work – was hypothesized to ensure delivery of the third
aryl group to both Mcl-1 and HDM2 by providing additional
flexibility that permits effective mimicry of the more rigid BH3
α-helix and equally the less-structured p53 α-helix. Figure 5
shows the overlay of an energy-minimized conformation of acyl
sulfonamide isoxazole OX0 in which all R groups were fixed as
isopropyl groups. Good mimicry of the key side chains (high-
lighted) of both helical peptides is suggested, which is expected
to translate into potent inhibition of Mcl-1 and HDM2. Since the
acyl sulfonamide portion is intended to emulate the carboxylic
acid on one face of the helix as well as an hydrophobic group
on the opposing face, while the R1 and R2 are intended to
emulate adjacent hydrophobic groups on the same face of the
helix, we hypothesize that the isoxazoles and pyrazoles will be
more potent inhibitors than the corresponding thiazoles which
carry an inferior substitution pattern.
energies (GFEs).[59] Visualization of the FragMaps allows for the
identification of synthetically accessible specific functional
groups that will maximize interactions with proteins. SILCS
Monte-Carlo (SILCS MC) sampling can be used in conjunction
with the FragMaps for ligand docking from which Ligand-GFE
(LGFE) scores are obtained, providing a quantitative ranking of
different ligands.[60] This aids in the design of functional group
modifications on the inhibitors to increase affinity for the target
protein. In the present study the SILCS approach is used
simultaneously on Mcl-1 and HDM2 to identify functional
groups that will maximize affinity for the two proteins in the
context of a rational polypharmacology design approach.
Next, OX0 was docked into Mcl-1 (Figure 6A) and HDM2
(Figure 6B) by using SILCS to show the binding mode of our
compounds and determine their ability to occupy regions that
are favorable for binding of the design functional groups.
Figure 7 shows the SILCS-MC docked orientations of OX0 in
Mcl-1 (left) and HMD2 (right), along with FragMaps for aliphatic
(green) and aromatic (purple) functional groups. This informs us
where additional moieties should be grafted onto our inhibitors
to ensure more favorable contacts with the proteins.[13] In both
proteins, the hydrophobic functional groups projected from the
core scaffold are encased almost entirely within the FragMaps
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
Computer-aided drug design
We employed computer-aided drug design (CADD) using SILCS,
site identification by ligand competitive saturation, to assist in
compound library design efforts.[56–58] Unlike other CADD
methods, SILCS takes protein flexibility and desolvation into
account. SILCS initially involves molecular dynamics (MD)
simulations of the target protein in the presence of organic
solutes immersed in an aqueous environment, which compete
for binding sites on the protein. Analysis of the distribution of
organic solutes yields 3D residence fragment probability
distributions that are used to map functional group affinity
patterns, termed FragMaps. These FragMaps include aliphatic,
aromatic, hydrogen bond donors and acceptors, heterocycles
and charged functional groups that are converted into grid free
Figure 6. SILCS MC docking with OX0 in A) Mcl-1 and B) HDM2.
Figure 5. Energy minimization of OX0 in ChemDraw3D and overlaid with
Bim-BH3 (center) and p53TAD (right) α-helices.
Figure 7. OX0 docked in Mcl-1 (left) and HDM2 (right) with SILCS.
ChemMedChem 2020, 15, 1–9
www.chemmedchem.org
4
© 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
doi.org/10.1002/cmdc.202000278
ChemMedChem
1
2
3
4
5
6
7
8
°
°
Scheme 1. i) NaN3, DMF, RT, 18 h; ii) (4-OiPr)benzaldehyde, NaH, EtOH, À 10 C, 4 h; iii) NH2OH·HCl, pyridine, reflux, 1 h; iv) Et3N, NCS, DMF, RT to 90 C, 18 h; v)
9
°
°
LiOH·H2O, THF/MeOH/H2O, RT,18 h; vi) corresponding sulfonamide, isobutyl chloroformate, NMM, NaH, THF, À 10 C to 0 C to RT, 18 h.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
°
°
°
°
Scheme 2. i) a) 2-iodopropane, K2CO3, DMF, RT to 60 C, 18 h; i) b) SnCl2 ·2H2O, EtOAc, 50 C, 18 h; ii) a) NaNO2, H2O, 0 C, 18 h; ii) b) SnCl2 ·2H2O, EtOAc, 50 C, o/
n; iii) corresponding alcohol, K2CO3, DMF, 60 C, 18 h; iv) diethyl oxalate, NaH, THF, 0 C to RT, 18 h; v) AcOH, reflux, 18 h; vi) LiOH·H2O, THF/MeOH/H2O, RT,
18 h; vii) corresponding sulfonamide, isobutyl chloroformate, NMM, NaH, THF, À 10 C to 0 C to RT, 18 h.
°
°
°
°
indicating that they are in the correct orientation to mimic the
native ligand and thereby improve the binding affinity.
were restricted to a focus set of hydrophobic side chains,
including isopropyl and phenyl.
Isoxazoles: Azidation of ethylbromoacetate (1) followed by a
Knoevenagel condensation with 4-isopropxybenzaldehyde (2)
delivered ethyl (Z)-2-azido-3-(4-isopropoxyphenyl) acrylate (3).
Meanwhile, various 4-substituted benzaldehydes 4 were trans-
formed into (E)-benzaldehyde oximes 5 under standard con-
ditions. Subsequently, N-chlorosuccinimide mediated the cycli-
zation of acrylates 3 with oximes 5 to construct the isoxazole
scaffold in compounds 6. Saponification of the ethyl esters
revealed the carboxylic acids 7, which were finally coupled with
various sulfonamides to afford the 1,2,3-functionalized isoxazole
acyl sulfonamides 8.
Pyrazoles: Isopropylation of 4-nitrophenol (9) followed by
reduction with tin(II) chloride yielded 4-isopropoxyaniline (10),
which was subsequently transformed into hydrazine 11. Mean-
while, 4-fluoroacetophenone (12) underwent SNAr reactions
with isopropanol or phenol to furnish the 4-substituted phenol
ethers 13, which underwent Claisen ester condensations with
diethyl oxalate to deliver β-keto esters 14. In an application of
the Knorr pyrazole synthesis, hydrazines 11 were condensed
with β-keto esters to yield the 1,2,4-trisubstituted pyrazoles 15.
Saponification as before delivered the corresponding carboxylic
acids 16, which were subsequently coupled to the library of
sulfonamides to furnish the 1,3,4-trisubstituted pyrazole acyl
sulfonamides 17.
Synthesis
Target molecules based on the generic structures shown in
Figure 4 (right), were accessed as depicted in Scheme 1–3. In
order to evaluate our design strategy, the R1, R2 and R3 groups
Thiazoles: Briefly, ethyl-2-aminothiazole-4-carboxylate (18)
was chlorinated in the remaining aromatic position and then
subjected to a Sandmeyer transformation to deliver the 2-
bromo-5-chlorothiazole-2-carboxylate derivative 20. Both halo-
gens were displaced in a double Suzuki reaction employing an
excess of 4-hydroxybenzeneboronic acid, and then both
°
°
Scheme 3. i) NCS, ACN, 80 C, 18 h; ii) tBuONO, CuBr2, ACN, 80 C, 18 h; iii) 4-
hydroxyphenylboronic acid, CsF, tetrakis((triphenylphosphine)palladium(0)),
DME/MeOH, 80 C, 18 h; iv) 2-iodopropane, K2CO3, DMF, 50 C, 18 h; v)
LiOH·H2O, THF/MeOH/H2O, RT, 18 h; vi) corresponding sulfonamide, isobutyl
°
°
°
chloroformate, NMM, NaH, THF, À 10 to 0 C to RT, 18 h.
ChemMedChem 2020, 15, 1–9
www.chemmedchem.org
5
© 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
doi.org/10.1002/cmdc.202000278
ChemMedChem
phenols were alkylated with isopropyl iodide to yield 22.
Saponification as before delivered the carboxylic acid 23, and
conjugation to our library of sulfonamides furnished the target
molecules, 1,2,4-tri-substituted thiazole acyl sulfonamides 24.
compounds were more potent inhibitors of Mcl-1 than HDM2,
ranging from around three-fold to >100-fold better. Further-
more, the isoxazoles and pyrazoles were, generally, the most
potent dual inhibitors, and, likely owing to the inferior
substitution pattern that was referred to earlier, the thiazoles
were the worst inhibitors. Of course, we cannot rule out that
the identity and location of the heteroatoms in each core did
not impact binding affinity, as well, and this will be investigated
at a later date. It is noteworthy that the more potent isoxazoles
and pyrazoles have a more clustered projection of “side chains”
similar to the potent HDM2 inhibitor Nutlin-3a, and a co-crystal
structure of Nutlin-3a/HDM2 reveals the three sub-pockets on
HDM2 are close together (PDB ID: 4 J3E). This superior
functionality is present in both the Novartis patent and the
Zhang et al. dual inhibitor.[45,51]
1
2
3
4
5
6
7
8
9
Results
Binding affinities of target molecules to Mcl-1 and HDM2 were
determined by a standard fluorescence anisotropy competition
assay using FITC (fluorescein isothiocyanate)-labeled Bak-BH3
for Mcl-1 or TAMRA (tetramethylrhodamine)-labeled p53 for
HDM2; these data are presented in Table 1. As we prepared a
focused set of inhibitors, it would be premature at this stage to
delve into a detailed discussion on R1, R2 and R4 side chains, and
that will be reserved for a follow-up full paper with a larger
library of compounds. However, it is clear that in every case,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
We hypothesized that the conversion of the carboxylic acid
to an acyl sulfonamide would offer greater inhibition, accom-
Table 1. Inhibition of Mcl-1 and HDM2 with isoxazoles, pyrazoles, and thiazoles using fluorescence anisotropy competition assay (FACA).
NA: no activity. Data are represented as the average of experiments performed in triplicate�SD. IC50 values were converted into Ki values by using the
Nikolovska-Coleska equation.[61]
ChemMedChem 2020, 15, 1–9
www.chemmedchem.org
6
© 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
doi.org/10.1002/cmdc.202000278
ChemMedChem
plishing mimicry of the third hydrophobic side chain. While the Conclusion
poor activity of the thiazole acid 23 could not be rescued in this
1
2
3
4
5
6
7
8
9
way, the data for the isoxazoles and pyrazoles proved this
hypothesis true in almost every case with Mcl-1 and in some of
the cases with HDM2. This can be seen, for example, by
comparing acid 7a (Ki =11.4 μM (Mcl-1), 85.8 μM (HDM2)) with
its corresponding acyl sulfonamide 8a Ki =2.91 μM (Mcl-1), 41.5
μM (HDM2)), as well as acid 16a Ki =10.6 μM (Mcl-1), 84.4 μM
(HDM2)) with the analogous acyl sulfonamide 17a Ki =5.89 μM
(Mcl-1), 41.0 μM (HDM2)). We attribute this to a combination of
the retention of an acidic functionality needed to engage with
Arg263 of Mcl-1, but also the excellent delivery of the third
functional group R4 that is seen to overlap with the FragMaps in
OX0 (Figure 6) in the modeling studies with both proteins,
indicating their contribution to improved affinity. It should be
noted that although OX0 was not synthesized, it is very similar
to 8a, and the intention of the FragMaps was to suggest
analogues based on the modeled compound, such as 8a. Closer
inspection of the data suggests that the strategy of converting
a carboxylic acid into a structurally more elaborate acyl
sulfonamide had a greater beneficial impact on Mcl-1 inhibition
than on HDM2 inhibition, and in some cases – predominantly
isoxazoles – this actually proved detrimental to HDM2 inhib-
ition. Of all compounds tested, our most potent dual inhibitor
was pyrazole 17b with Ki values of 433 nM and 11.5 μM for Mcl-
1 and HDM2, respectively. Inadvertently, we may have identified
The discovery of targeted anticancer agents is an intense field
of research that is constantly being met with defeat due to the
development of resistance. Polypharmacology represents a
paradigm shift in the treatment of multi-factorial diseases, such
as cancer. Recognizing that Mcl-1 and HDM2 are both
upregulated in similar cancers and their anticancer activities are
regulated by similarly functionalized α-helical domains in their
partner proteins, we hypothesized that suitably functionalized
heterocycles could be designed to simultaneously inhibit both
Mcl-1 and HDM2. In summary, we have demonstrated that acyl
sulfonamides of triply substituted isoxazoles and pyrazoles are
effective dual α-helix mimetics of the Bim-BH3 and p53TAD
domains, inhibiting Mcl-1 and HDM2, respectively. Although
our data presently indicate that the isoxazoles might be better
suited to the development of selective Mcl-1 inhibitors, further
optimization of the pyrazole core may lead to more potent dual
inhibitors. Particularly, since conversion of the pyrazole acids to
their corresponding acyl sulfonamides was more impactful for
Mcl-1 than HDM2, we intend to first optimize the pyrazole acid
predominantly to HDM2, and then anticipate a greater improve-
ment in Mcl-1 affinity upon converting the acids to acyl
sulfonamides.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
a new scaffold for the development of selective Mcl-1 inhibitors, Experimental Section
since some of our isoxazoles were potent against Mcl-1 with
limited or no activity against HDM2, for example 8b with a Ki of
263 nM for Mcl-1 and inactive against HDM2.
Complete protocols for both chemical syntheses and biological
methods together with characterization data are presented in the
Supporting Information.
One of the challenges with developing dual inhibitors –
through the same recognition motif – of Mcl-1 and HDM2 by α-
helix mimicry is the size mismatch between analogous side
chains. Particularly, the middle side chain of the region of Bim- Conflict of Interest
BH3 that is being emulated is relatively small (Ile65), whereas
the corresponding residue in p53TAD is much larger (Trp23),
and the respective pockets that bind these side chains do not
appear malleable from crystal structures. On the other hand,
Fesik’s group has demonstrated that the p2 pocket on Mcl-1
A.D.M. Jr. is co-founder and Chief Scientific Officer of SilcsBio
LLC.
that binds Leu62 of Bim-BH3 is somewhat plastic, and can Acknowledgements
accommodate larger groups;[37] indeed, this appears to be a
significant source of ligand affinity. Thus, for an effective dual
inhibitor, a large R1 group is required for Mcl-1 inhibition, but a
large R2 group is required for HDM2 inhibition. The isoxazole
scaffold cannot effectively meet these criteria simultaneously.
However, the pyrazole scaffold can because the R1 and R2
groups are interchangeable with respect to the third substitu-
ent at the acyl sulfonamide functionality through rotation about
the pyrazole-acyl bond, and this may be one reason why, in
general, the pyrazole acyl sulfonamides fared better as dual
inhibitors than the corresponding isoxazole acyl sulfonamides.
In other words, the large R1 phenyl group in the pyrazole series
can serve as an effective mimic of Leu62 at the located at the
periphery of the BH3 α-helix and bind the plastic p2 pocket of
Mcl-1, yet, on the other hand through pyrazole-acyl bond
rotation, is also able to mimic the Trp23 in the middle of the
p53TAD α-helix.
We thank the National Institutes of Health (T32GM066706), and
the University of Maryland School of Pharmacy for financial
support of this research.
Keywords: apoptosis
polypharmacology · protein–protein interactions
·
cancer
·
heterocycles
·
[1] A. L. G. Janda, Curr. Top. Med. Chem. 2011, 11, 258–280.
[2] L. Mabonga, A. P. Kappo, Biophys. Rev. Lett. 2019, 11, 559–581.
[3] R. S. Akhtar, J. M. Ness, K. A. Roth, Biochim. Biophys. Acta Mol. Cell Res.
2004, 1644, 189–203.
[4] J. Sassone, A. Maraschi, F. Sassone, V. Silani, A. Ciammola, Cell Death Dis.
2013, 4, 772–772.
[5] D. S. Luciani, S. A. White, S. B. Widenmaier, V. V. Saran, F. Taghizadeh, X.
Hu, M. F. Allard, J. D. Johnson, Diabetes 2013, 62, 170–182.
[6] S. Fletcher, A. D. Hamilton, Curr. Opin. Chem. Biol. 2005, 9, 632–638.
[7] M. E. Lanning, S. Fletcher, Future Med. Chem. 2013, 5, 2157–2174.
ChemMedChem 2020, 15, 1–9
www.chemmedchem.org
7
© 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Communications
doi.org/10.1002/cmdc.202000278
ChemMedChem
[8] V. Azzarito, K. Long, N. S. Murphy, A. J. Wilson, Nat. Chem. 2013, 5, 161–
173.
[9] M. Arkin, Curr. Opin. Chem. Biol. 2005, 9, 317–324.
[10] B. L. Lampson, M. S. Davids, Curr Hematol Malig Rep 2017, 12, 11–19.
[11] V. Tisato, R. Voltan, A. Gonelli, P. Secchiero, G. Zauli, J. Hematol. &
Oncology 2017, 10, 133.
[12] “Phase I Study of MIK665, a) Mcl-1 Inhibitor, in Patients With Refractory
or Relapsed Lymphoma or Multiple Myeloma”, https://clinicaltrials.gov/
ct2/show/NCT02992483, 2020.
[13] “Study of Safety and Efficacy of HDM201 in Combination With LEE011
in Patients With Liposarcoma”, https://clinicaltrials.gov/ct2/show/
NCT02343172.
[14] M. E. Lanning, P. T. Wilder, H. Bailey, B. Drennen, M. Cavalier, L. Chen,
J. L. Yap, M. Raje, S. Fletcher, Org. Biomol. Chem. 2015, 13, 8642–8646.
[15] A. M. Petros, E. T. Olejniczak, S. W. Fesik, Biochim. Biophys. Acta 2004,
1644, 83–94.
[16] A. Barnard, K. Long, D. J. Yeo, J. A. Miles, V. Azzarito, G. M. Burslem, P. A.
Prabhakaran, T. Edwards, A. J. Wilson, Org. Biomol. Chem. 2014, 12,
6794–6799.
[17] P. E. Czabotar, G. Lessene, A. Strasser, J. M. Adams, Nat. Rev. Mol. Cell
Biol. 2014, 15, 49–63.
[18] K. H. Vousden, D. P. Lane, Nat. Rev. Mol. Cell Biol. 2007, 8, 275–283.
[19] R. J. Youle, A. Strasser, Nat. Rev. Mol. Cell Biol. 2008, 9, 47–59.
[20] C. M. Adams, S. Clark-Garvey, P. Porcu, C. M. Eischen, Front. Oncol. 2019,
8. DOI: 10.3389/fonc.2018.00636.
[21] M. S. Lee, J. H. Ha, H. S. Yoon, C. K. Lee, S. W. Chi Biochemical, Biophysical
Research Communications 2014, 445, 120–125.
[22] B. Vu, P. Wovkulich, G. Pizzolato, A. Lovey, Q. Ding, N. Jiang, J.-J. Liu, C.
Zhao, K. Glenn, Y. Wen, C. Tovar, K. Packman, L. Vassilev, B. Graves, ACS
Med. Chem. Lett. 2013, 4, 466–469.
[23] M. N. Saha, H. Jiang, H. Chang, Cancer Biol. Ther. 2010, 10, 567–578.
[24] E. Toufektchan, F. Toledo, Cancers (Basel) 2018, 10, DOI: 10.3390/
cancers10050135.
[25] G. Teoh, M. Urashima, A. Ogata, D. Chauhan, J. A. DeCaprio, S. P. Treon,
R. L. Schlossman, K. C. Anderson, Blood 1997, 90, 1982–1992.
[26] K. Kojima, M. Konopleva, T. McQueen, S. O’Brien, W. Plunkett, M.
Andreeff, Blood 2006, 108, 993–1000.
[27] J. I.-J. Leu, P. Dumont, M. Hafey, M. E. Murphy, D. L. George, Nat. Cell
Biol. 2004, 6, 443–450.
[28] K. Kojima, T. McQueen, Y. Chen, R. Jacamo, M. Konopleva, N. Shinojima,
E. Shpall, X. Huang, M. Andreeff, Blood 2011, 118, 4431–4439.
[29] J. E. Chipuk, T. Kuwana, L. Bouchier-Hayes, N. M. Droin, D. D. Newmeyer,
M. Schuler, D. R. Green, Science 2004, 303, 1010–1014.
[36] “AbbVie Announces US FDA Approval of Venclexta (venetoclax) as a
Chemotherapy-Free Combination Regimen for Previously Untreated
Chronic Lymphocytic Leukemia Patients”, https://www.drugs.com/new-
drugs/abbvie-announces-us-fda-approval-venclexta-venetoclax-chemo-
therapy-free-combination-regimen-4974.html.
[37] A. Friberg, D. Vigil, B. Zhao, R. N. Daniels, J. P. Burke, P. M. Garcia-
Barrantes, D. Camper, B. A. Chauder, T. Lee, E. T. Olejniczak, S. W. Fesik,
J. Med. Chem. 2013, 56, 15–30.
[38] E. A. Punnoose, J. D. Leverson, F. Peale, E. R. Boghaert, L. D. Belmont, N.
Tan, A. Young, M. Mitten, E. Ingalla, W. C. Darbonne, A. Oleksijew, P.
Tapang, P. Yue, J. Oeh, L. Lee, S. Maiga, W. J. Fairbrother, M. Amiot, A. J.
Souers, D. Sampath, Mol. Cancer Ther. 2016, 15, 1132–1144.
[39] M. Konopleva, D. A. Pollyea, J. Potluri, B. Chyla, L. Hogdal, T. Busman, E.
McKeegan, A. H. Salem, M. Zhu, J. L. Ricker, W. Blum, C. D. DiNardo, T.
Kadia, M. Dunbar, R. Kirby, N. Falotico, J. Leverson, R. Humerickhouse,
M. Mabry, R. Stone, H. Kantarjian, A. Letai, Cancer Discov 2016, 6, 1106–
1117.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
[40] C. Caruso, Cancer Discov 2019, 9, 156.
[41] K. Seipel, K. Schmitter, U. Bacher, T. Pabst, Cancers (Basel) 2019, 11,
1779.
[42] A. S. Reddy, S. Zhang, Expert Rev Clin Pharmacol 2013, 6, 41–47.
[43] A. Anighoro, J. Bajorath, G. Rastelli, J. Med. Chem. 2014, 57, 7874–7887.
[44] G. J. P. Rautureau, M. Yabal, H. Yang, D. C. S. Huang, M. Kvansakul, M. G.
Hinds, Cell Death Dis. 2012, 3, e443.
[45] Z. Wang, T. Song, Y. Feng, Z. Guo, Y. Fan, W. Xu, L. Liu, A. Wang, Z.
Zhang, J. Med. Chem. 2016, 59, 3152–3162.
[46] A. Vaupel, G. Bold, A. De Pover, T. Stachyra-Valat, J. Hergovich Lisztwan,
J. Kallen, K. Masuya, Furet. Bioorganic & Medicinal Chemistry Letters 2014,
24, 2110–2114; Medicinal Chemistry Letters 2014, 24, 2110–2114.
[47] A. S. A. Mady, C. Liao, N. Bajwa, K. J. Kump, F. A. Abulwerdi, K. L. Lev, L.
Miao, S. M. Grigsby, A. Perdih, J. A. Stuckey, Y. Du, H. Fu, Z. Nikolovska-
Coleska, Sci. Rep. 2018, 8, 10210.
[48] C. Li, M. Pazgier, C. Li, W. Yuan, M. Liu, G. Wei, W.-Y. Lu, W. Lu, J. Mol.
Biol. 2010, 398, 200–213.
[49] H. Yin, G.-I. Lee, K. A. Sedey, O. Kutzki, H. S. Park, B. P. Orner, J. T. Ernst,
H.-G. Wang, S. M. Sebti, A. D. Hamilton, J. Am. Chem. Soc. 2005, 127,
10191–10196.
[50] L. T. Vassilev, B. T. Vu, B. Graves, D. Carvajal, F. Podlaski, Z. Filipovic, N.
Kong, U. Kammlott, C. Lukacs, C. Klein, N. Fotouhi, E. A. Liu, Science
2004, 303, 844–848.
[51] G. Bold, P. Furet, F. Gessier, J. Kallen, J. H. Lisztwan, K. Masuya, A. Vaupel,
CA2771936A1, March 3, 2011.
[52] N. F. Pelz, Z. Bian, B. Zhao, S. Shaw, J. C. Tarr, J. Belmar, C. Gregg, D. V.
Camper, C. M. Goodwin, A. L. Arnold, J. L. Sensintaffar, A. Friberg, O. W.
Rossanese, T. Lee, E. T. Olejniczak, S. W. Fesik, J. Med. Chem. 2016, 59,
2054–2066.
[53] K. Kapadiya, K. Kavadia, P. Manvar, R. Kotadiya, R. Kothari, K. Ranjan,
Chemistry & Biology Interface 2015, 5, 258–266.
[54] D. K. Miyamoto, H. A. Flaxman, H.-Y. Wu, J. Gao, C. M. Woo, ACS Chem.
Biol. 2019, 14, 2527–2532.
[55] J. K. Gierse, Y. Zhang, W. F. Hood, M. C. Walker, J. S. Trigg, T. J. Maziasz,
C. M. Koboldt, J. L. Muhammad, B. S. Zweifel, J. L. Masferrer, P. C.
Isakson, K. Seibert, J. Pharmacol. Exp. Ther. 2005, 312, 1206–1212.
[56] O. Guvench Jr., A. D. MacKerell, PLoS Comput. Biol. 2009, 5. DOI:
10.1371/journal.pcbi.1000435.
[30] S. Besbes, C. Billard, Cell Death Dis. 2015, 6, e1810.
[31] A. Kotschy, Z. Szlavik, J. Murray, J. Davidson, A. L. Maragno, G.
Le Toumelin-Braizat, M. Chanrion, G. L. Kelly, J.-N. Gong, D. M. Moujalled,
A. Bruno, M. Csekei, A. Paczal, Z. B. Szabo, S. Sipos, G. Radics, A.
Proszenyak, B. Balint, L. Ondi, G. Blasko, A. Robertson, A. Surgenor, P.
Dokurno, I. Chen, N. Matassova, J. Smith, C. Pedder, C. Graham, A.
Studeny, G. Lysiak-Auvity, A.-M. Girard, F. Gravé, D. Segal, C. D. Riffkin,
G. Pomilio, L. C. A. Galbraith, B. J. Aubrey, M. S. Brennan, M. J. Herold, C.
Chang, G. Guasconi, N. Cauquil, F. Melchiore, N. Guigal-Stephan, B.
Lockhart, F. Colland, J. A. Hickman, A. W. Roberts, D. C. S. Huang, A. H.
Wei, A. Strasser, G. Lessene, O. Geneste, Nature 2016, 538, 477–482.
[32] J.-C. Carry, C. Garcia-Echeverria, Bioorg. Med. Chem. Lett. 2013, 23, 2480–
2485.
[33] E. Whiting, M. R. Raje, J. Chauhan, P. T. Wilder, D. Van Eker, S. J. Hughes,
N. G. Bowen, G. E. A. Vickers, I. C. Fenimore, S. Fletcher, Bioorg. Med.
Chem. Lett. 2018, 28, 523–528; I. L. Conlon, D. Van Eker, S. Abdelmalak,
W. A. Murphy, H. Bashir, M. Sun, J. Chauhan, K. M. Varney, R. Godoy-
Ruiz, P. T. Wilder, S. Fletcher, Bioorg. Med. Chem. Lett. 2018, 28, 1949–
1953; M. E. Lanning, W. Yu, J. L. Yap, J. Chauhan, L. Chen, E. Whiting,
L. S. Pidugu, T. Atkinson, H. Bailey, W. Li, B. M. Roth, L. Hynicka, K.
Chesko, E. A. Toth, P. Shapiro, A. D. MacKerell Jr, P. T. Wilder, S. Fletcher,
Eur. J. Med. Chem. 2016, 113, 273–292; B. Drennen, J. A. Scheenstra, J. L.
Yap, L. Chen, M. E. Lanning, B. M. Roth, P. T. Wilder, S. Fletcher,
ChemMedChem 2016, 11, 827–833; M. E. Lanning, P. T. Wilder, H. Bailey,
B. Drennen, M. Cavalier, L. Chen, J. L. Yap, M. Raje, S. Fletcher, Org.
Biomol. Chem. 2015, 13, 8642–8646.
[57] E. P. Raman, W. Yu, S. K. Lakkaraju, A. D. MacKerell, J. Chem. Inf. Model.
2013, 53, 3384–3398.
[58] V. D. Ustach, S. K. Lakkaraju, S. Jo, W. Yu, W. Jiang, A. D. MacKerell, J.
Chem. Inf. Model. 2019, 59, 3018–3035.
[59] E. P. Raman, W. Yu, O. Guvench, A. D. MacKerell, J. Chem. Inf. Model.
2011, 51, 877–896.
[60] M. Small, A. D. MacKerell in In Silico Drug Discovery and Design: Theory,
Methods, Challenges, and Applications. (Ed.: C. Cavasotto), CRC Press,
2015, 45–78.
[61] Z. Nikolovska-Coleska, R. Wang, X. Fang, H. Pan, Y. Tomita, P. Li, P. P.
Roller, K. Krajewski, N. G. Saito, J. A. Stuckey, S. Wang, Anal. Biochem.
2004, 332, 261-273..
[34] J. L. Yap, L. Chen, M. E. Lanning, S. Fletcher, J. Med. Chem. 2017, 60,
821–838; S. Fletcher, Expert Opin. Ther. Pat. 2019, 29, 909–919.
[35] A. Beloglazkina, N. Zyk, A. Majouga, E. Beloglazkina, Molecules. 2020, 25,
12.
Manuscript received: April 27, 2020
Revised manuscript received: May 19, 2020
Version of record online: ■■■, ■■■■
ChemMedChem 2020, 15, 1–9
www.chemmedchem.org
8
© 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

COMMUNICATIONS
1
2
3
4
5
6
7
8
9
The anti-apoptotic protein Mcl-1
and E3 ubiquitin ligase HDM2 are
both frequently upregulated in
cancer. A small-molecule inhibitor
can be developed by mimicking their
protein binding partners, Bim and
p53, to restore apoptosis in
oncogenic cells. Through a polyphar-
macology approach, we have
developed small-molecule α-helix
mimetics with isoxazole, pyrazole,
and thiazole scaffolds that function
as dual inhibitors of Mcl-1 and
HDM2.
I. L. Conlon, Dr. B. Drennen, Dr. M. E.
Lanning, S. Hughes, R. Rothhaas,
Dr. P. T. Wilder, Dr. A. D. MacKerell, Jr.,
Dr. S. Fletcher*
1 – 9
Rationally Designed Polypharmacol-
ogy: α-Helix Mimetics as Dual Inhib-
itors of the Oncoproteins Mcl-1and
HDM2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57