Systematic Investigation of the Permeability of Androgen Receptor PROTACs

Article abstract of DOI:10.1021/acsmedchemlett.0c00194

Bifunctional molecules known as PROTACs simultaneously bind an E3 ligase and a protein of interest to direct ubiquitination and clearance of that protein, and they have emerged in the past decade as an exciting new paradigm in drug discovery. In order to investigate the permeability and properties of these large molecules, we synthesized two panels of PROTAC molecules, constructed from a range of protein-target ligands, linkers, and E3 ligase ligands. The androgen receptor, which is a well-studied protein in the PROTAC field was used as a model system. The physicochemical properties and permeability of PROTACs are discussed.

Full text of DOI:10.1021/acsmedchemlett.0c00194

pubs.acs.org/acsmedchemlett
Letter
Systematic Investigation of the Permeability of Androgen Receptor
PROTACs
Duncan E. Scott, Timothy P. C. Rooney, Elliott D. Bayle, Tashfina Mirza, Henriette M. G. Willems,
Jonathan H. Clarke, Stephen P. Andrews, and John Skidmore*
Cite This: https://dx.doi.org/10.1021/acsmedchemlett.0c00194
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Bifunctional molecules known as PROTACs simultaneously bind an E3
ligase and a protein of interest to direct ubiquitination and clearance of that protein,
and they have emerged in the past decade as an exciting new paradigm in drug
discovery. In order to investigate the permeability and properties of these large
molecules, we synthesized two panels of PROTAC molecules, constructed from a
range of protein-target ligands, linkers, and E3 ligase ligands. The androgen receptor,
which is a well-studied protein in the PROTAC field was used as a model system. The
physicochemical properties and permeability of PROTACs are discussed.
KEYWORDS: PROTACs, drug discovery, androgen receptor, medicinal chemistry
present a greater challenge to PROTACs than other biological
membranes.7
hemically induced, targeted protein degradation is an
Cemerging pharmacological strategy for manipulation of
intracellular protein levels.1 One way this has been achieved is
through the design of bifunctional molecules that simulta-
neously bind to a protein of interest and an E3 ubiquitin ligase.
By bringing these two proteins into proximity, PROteolysis
TArgeting Chimeras (PROTACs) facilitate the ubiquitinyla-
tion and proteasomal degradation of the protein of interest.
Early PROTACs relied on short peptide sequences to bind the
E3 ligase protein, and their low permeability necessitated
microinjection into cells;2 however, recent identification of
small molecule ligands for E3 ligases has opened the possibility
of cell-permeable nonpeptidic PROTACs. The ability to
degrade the entire protein in a cell, rather than simply inhibit
the catalytic domain, distinguishes PROTACs from traditional
small-molecule inhibitors. Further, the removal of additional
roles of the protein, such as complex formation and scaffolding
with other proteins, means that PROTACs can access unique
pharmacology. PROTACs are also potentially catalytic; each
molecule can direct degradation of multiple target proteins.
PROTACs have been developed for numerous protein classes,
including nuclear receptors,3 epigenetic factors,4 and kinases.5
These compounds have been used to study the effect of
enforced degradation of the target protein in a number of cell
lines and also in in vivo studies.6,7 While this work has the
potential to offer a new pharmacological paradigm, clinical
utility is yet to be demonstrated and a number of challenges
still remain.8 A recent in vivo study of a potent PROTAC
(DC50 0.3 nM) indicated that despite an impressive systemic
knock-down of the desired FKBP12 protein, the brain was the
only tissue analyzed where protein levels were unaffected,
indicating that the blood−brain barrier may intrinsically
PROTAC molecules are rule-breaking in respect to
traditional medicinal chemistry guidelines;8 they possess high
molecular weight and high TPSA.9 Permeability is key for
activity against intracellular targets, and in general, perme-
ability drops off drastically with increasing molecular weight.
Yet despite this, some impressive cell potencies for PROTACs
have been reported.6 Other classes of small-molecules, also
beyond the rule of 5, have been reported to have high cell
permeability and oral bioavailability.10 Notably, cyclic peptides
and natural products such as rifamycins are thought to adopt
conformations that allow formation of intramolecular hydro-
gen bonds, reducing the effective polarity of the molecules and
allowing passage through membranes.11,12 We hypothesized
that in a similar fashion, it might be possible for PROTAC
molecules to possess rule-breaking permeability, either by
linker-mediated conformational collapse in water or by the two
functional ends of the PROTAC migrating through a
membrane in a pseudoindependent fashion by virtue of the
length and flexibility of the linker so disguising their high
molecular volume. PROTAC molecules are typically composed
of two rigid protein-binding small molecules, connected by a
linker. As we focus on flexible linkers in this work, further
linker types with more rigidity remain to be explored. Recently
Received: April 15, 2020
Accepted: June 1, 2020
https://dx.doi.org/10.1021/acsmedchemlett.0c00194
© XXXX American Chemical Society
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX
A

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
Figure 1. (A) AR ligand 2 in complex with AR ligand binding domain (pdb id: 3v49). The linker attachment point is circled in red. (B) Structures
of compounds 1−5, derived from AR ligands, for incorporation into PROTACs. The linker attachment point is indicated in red. (C) Proteolysis
directing groups 6−9; the attachment point when incorporated into PROTACs is indicated in blue. (D) Lenalidomide bound to Cereblon E3
ubiquitin ligase (pdb id: 4ci2). The linker attachment point is circled in blue. (E) General AR PROTAC structure of PROTAC Sets 1 and 2.
Foley et al. have published observations on permeability using
a chloroalkane penetration assay, focusing on the permeability
of a chloroalkane modified BRD4 PROTAC.13 A wider study
to evaluate the impact of linker, protein−ligand, and E3 ligase
ligand on the permeability of PROTACs is warranted.
AR binding site, in a position that has been successfully
exploited by previously reported AR PROTACs.3 The
structures of AR ligands 1−5 are shown in Figure 1B. SARM
1 is a close analogue of the clinical compounds andarine and
ostarine. The hydantoin SARM 2 possesses high affinity for AR
(EC50 1.6 nM).19 The thiohydantoin derivative 3 has been
developed into the clinical compound enzalutamide and
incorporated into PROTACs.20 Pyrrolidine-based SARM 4
represents a very ligand-efficient class of modulators, of which
Ligandrol is a compound in clinical development.21 The
cyclobutane SARM 5 has also been previously reported.22
X-ray crystal structures of SARMs bound to AR reveal the
ligand completely enclosed by helix 12 (Figure 2A). In order
for AR PROTACs to bind and the linker to exit the AR
protein, some protein conformational change must occur.
Gryder et al. have previously reported a set of AR-binding
bifunctional molecules that also bind histone deacetylases
(HDACs) and suggest that helix 12 (H12) forms a “lid” over
SARMs, that can move to accommodate extended AR
ligands.23 Indeed, the related estrogen receptor (ER)
demonstrates large changes in the conformation of H12
when bound to an antagonist;24 hence, we created a homology
model of the AR with an H12-open conformation, based upon
an H12-open ER protein complex (Figure 2B and Supporting
Information). Docking PROTACs into this model successfully
reproduced the binding mode of the SARM and allowed the
linker to extend past H12 into the solvent, permitting a rational
design of linker length and linker−SARM attachment points.
Only a limited number of E3 ligases have been targeted by
PROTAC molecules, though recent studies have focused on
In our study, the androgen receptor (AR) was selected as a
model system. A panel of selective AR modulators (SARMs)
have previously been developed, and many PROTACs for the
AR protein have also been reported, employing a variety of E3
ligase ligands.2,3,14,15 As such it is a derisked, well-studied
system. The AR is a DNA-binding transcription factor nuclear
hormone receptor that, through binding to the endogenous
ligand testosterone, mediates growth factor and cell-signaling
dependent gene expression. The ligand binding domain (LBD)
provides a defined binding pocket for endogenous steroidal
ligands including testosterone. A number of SARMs are in
clinical trials for a variety of indications.14,16 Furthermore,
polyglutamine expansion of the AR causes the neuromuscular
disease spinal and bulbar muscular atrophy (SBMA or
Kennedy’s disease).17 A PROTAC designed to catalyze the
degradation of this neurotoxic form of AR could offer a
strategy for intervention in this disease.
In designing a set of PROTACs, the three components of
the PROTAC molecule were systematically varied; the AR
ligand, the E3 ubiquitin ligase ligand, and the linker region. To
select suitable ligands for the AR, SARMs were identified that
were predicted to have good CNS permeability using MPO
scoring.18 SARMs 1−5 all bind in the LBD of the AR (Figure
1A). An analysis of the binding modes of available crystal
structures reveals SARMs present a similar vector to exit the
B
https://dx.doi.org/10.1021/acsmedchemlett.0c00194
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
successfully incorporated into PROTACs.26 Small-molecule
ligands of the von Hippel Lindau (VHL) E3 ligase, such as 9,
have also been developed and incorporated into PROTACs.4−6
Another approach to chemically target protein degradation has
been to covalently27 or noncovalently28 associate a hydro-
phobic tag with a protein and thus promote its clearance. It has
been suggested in the context of AR clearance that the
presentation of the hydrophobic adamantane moiety 8 on the
AR surface is recognized as unfolded protein by the Hsp70/
CHIP system, leading to AR clearance.28
For the first half of this work, a focused set of PROTACs
(PROTAC Set 1, Figures 1E and S2) was synthesized by
linking AR ligands and smaller E3 ligands with a PEG linker
(10−19, Schemes S3 and S4), predicted by calculations to
have a neutral effect on the logD (Figure S4). In the second
half of the study, a further set of compounds (PROTAC Set 2,
Figures 1E and S3) focused on the VHL E3 ligase was
synthesized (20a−i, Scheme S6). In this set, the linker
composition was varied to evaluate the impact of linker on
permeability and functional activity. Fluorine was also
incorporated into the linker, which has not been reported in
PROTAC linkers previously. The effects of addition of fluorine
may be wide-ranging, potentially altering properties such as the
Figure 2. (A) Helix 12 (H12, red) of AR protein (pdb id: 2pnu)
occludes the entrance to the ligand-binding domain. The docked
small molecule 3b (Figure S1) is indicated by green spheres. (B) A
homology model of AR with helix-12 in an open conformation (red
helix) was generated. The remodeled helix 12 allows access from the
ligand-binding domain to the bulk solvent and accommodation of
PROTAC linkers.
expanding the repertoire.25 We employed four different
proteolysis targeting groups 6−9, with distinct physicochem-
ical properties (Figure 1C). The immunomodulatory drugs
thalidomide, and analogues pomalidamide and lenalidomide,
have been identified as ligands of the E3 ligase Cereblon
(Figure 1D), and structural analogues 6 and 7 have been
Figure 3. Structure, biophysical properties and biological activities of PEG-linked PROTAC Set 1. Coloring is added to aid visualization. (A) The
general structure of a set of PROTACs (10−19, Set 1) is shown in the top left, comprised of AR ligands 1−4, proteolysis directing groups 6−8, and
a PEG-linker. The PAMPA permeability for compounds 1−4 and 6−8 is shown in the outer edge of the table. The main body of the table shows
the PROTAC compound number in bold (10−19), composed from AR ligand 1−4 and proteolysis directing group 6−8, followed by PAMPA
permeability. The same table format is used in (B−C). BLQ = Below limit of quantification. (B) HSA % free determined by HPLC and as
described in methods section. (C) chromLogD7.4 determined by HPLC as described in the methods section.
C
https://dx.doi.org/10.1021/acsmedchemlett.0c00194
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
Table 1. Summary of PAMPA and Caco-2 Permeability of Selected PROTACs with Varying Proteolysis Directing Group
a
b
Permeability/10−6 cm s−1. Efflux ratio.
electrostatics of the linker, interaction with proteins, linker
conformation,29 permeability and microsomal stability. All
PROTAC structures are shown in full in the Supporting
Information (Figures S2 and S3).
3A). The high logD for this PROTAC, a feature of PROTACs
with the lipophilic adamantyl headgroup, presumably contrib-
utes to the moderate permeability (Figure 3C).
Given the low passive permeabilities measured in the
PAMPA assay, we also tested a selection of PROTAC
molecules for bidirectional permeability in Caco-2 cells
(Table 1 and Table 2). Briefly, this more biologically relevant
assay uses a polarized monolayer of Caco-2 cells and hence
allows assessment of permeability in two opposing directions
(“A2B” and “B2A”), also permitting an assessment of active
transport from the ratio of directional permeabilities.30 Typical
thresholds for permeability classification are low, <1.0 × 10−6
cm s−1, medium, (1−5) × 10−6 cm s−1, and high, >5 × 10−6 cm
s−1.31 While Caco-2 permeability is a useful measurement, in
particular to predict oral bioavailability, inference of efflux
ratios in other cell types may be complicated by different
transporter profiles in different cell types.32
From the Caco-2 data it is possible to make comparisons to
discern the effect of the proteolysis directing group in the
context of AR ligand 3 and a PEG-linker (Table 1).
Interestingly, cereblon ligand-containing PROTAC 14 has
the best A2B permeability measured at 1.7 × 10−6 cm s−1. The
B2A rate is high for this compound, 14.1 × 10−6 cm s−1, clearly
indicating that transporter efflux is an issue (efflux ratio = 8.4).
Replacing the cereblon ligand with the small, hydrophobic
adamantane degron in PROTAC 18 leads to a 10-fold decrease
in A2B and a 64-fold decrease in B2A (0.15 and 0.22 × 10−6
cm s−1, respectively) and no evidence of transporter efflux (ER
= 1.5). Substituting with the larger, more polar VHL ligand in
PROTAC 20d gives an A2B permeability below the limit of
quantification for this compound, but as the B2A rate is 9.6 ×
10−6 cm s−1, it is evident that an efflux issue is present (ER >
12).
With both sets of PROTACs in hand, the PEG-linked
PROTACs (Set 1: 10−19, Figure 3) and the linker-varied
VHL set (Set 2: 20a−i, Table 2), a panel of ADMET
parameters was selected for their potential to impact cell
activity; passive permeability (PAMPA), human serum
albumin (HSA) binding, and chromLogD7.4. PROTACs 10−
19 (Set 1, Figure 3) represent a model set of PROTACs with
physicochemical properties that minimize molecular weight
and TPSA and are not atypical to other PROTACs in terms of
these parameters. The chosen AR ligands are small and potent,
the proteolysis directing groups have low molecular weight,
and the PEG-linker represents a typical reasonable linker
length and distribution of polar atoms.3,22 The PAMPA Papp
for the AR ligands gives a wide range of permeability: low for
ligand 1 (1.4 × 10−6 cm s−1) and high for ligand 4 (13.3 ×
10−6 cm s−1) (Figure 3A). Given the reported activity of
PROTACs in the literature, reasonable permeability for this
PROTAC set might be expected, especially when combining
the most permeable AR ligand and E3 ligase ligand. However,
almost without exception, the permeability of the final
PROTAC molecules in the PAMPA assay was very low and
below the limit of quantification (Figure 3A and Table 2). It is
apparent that the high molecular weight and high TPSA, which
is intrinsic to PROTAC molecules, is driving poor permeability
for this class of molecules in this assay. Recovery in the
PAMPA assay was moderate, i.e. 40−80% in most cases (Table
S2). Of note is PROTAC 19, which combines ligandrol
derivative 4 and adamantyl degron 8 and has a PAMPA
permeability of 2.3 × 10−6 cm s−1, the highest PAMPA
permeability measured for our bifunctional molecules (Figure
D
https://dx.doi.org/10.1021/acsmedchemlett.0c00194
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
Table 2. Structure, Biophysical Properties, and Permeability of VHL-Targeting PROTACs (PROTAC Set 2)
a
b
c
d
HSA % free, determined by HPLC method. chromLogD7.4, determined by HPLC method. PAMPA permeability/10−6 cm s−1. Permeability/
e
10−6 cm s−1. Efflux Ratio. ND = Not determined.
The effect of the linker on Caco-2 permeability was explored
with a set of PROTACs from Set 2 (VHL ligand containing
PROTACs) (Table 2). For the PROTACs studied, A2B either
could be quantified as low (<1.0 × 10−6 cm s−1) or could not
be measured at all. Interestingly, a range of B2A rates were
observed. For PROTAC 20b, A2B and B2A were measured as
0.35 and 0.24 × 10−6 cm s−1, respectively, indicating no
significant transporter efflux. It is noteworthy that the linkers of
PROTACs 20b, 20c, and 20g bearing different numbers of
fluorine atoms lead to PROTACs that are all rather similar in
terms of permeability; A2B and B2A rates are low.
Contrastingly, PROTAC 20d demonstrates a high sensitivity
to the change to a PEG-linker, B2A is high (8.6 × 10−6 cm s−1)
and the efflux ratio is high (>12). In summary, low passive
permeability in PAMPA assays is typically observed for
PROTACs, but a marked structural influence of the linker
E
https://dx.doi.org/10.1021/acsmedchemlett.0c00194
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
Figure 4. (A) AR clearance activity of previously reported22 PROTAC 20a. (B) AR clearance activity of PROTACs 14 and 15. Plots show
normalized AR protein levels as a function of PROTAC concentration. PROTAC 14 is inactive in this cell assay. A minor change of the Cereblon
ligand leads to higher activity of PROTAC 15: a DCMAX of 33% and a DC50 of 10 nM. N = 5 (or more) independent experiments performed in
triplicate (individual data points shown). Statistical significance by ANOVA (Dunnett’s) **** = p < 0.0001.
and E3 ligase ligand on A2B, B2A, and efflux ratio in Caco-2
cells can be observed. The recovery of compounds in the
Caco-2 assay was reasonable, 60−100% (Table S3). Of note,
all PROTACs were relatively unbound to HSA protein,
between 6 and 15% (Figure 3B and Table 2). More work is
required to investigate this beyond the compound set studied
here.
The biological efficacy of the PROTACs to degrade
endogenous levels of the AR receptor was assessed using a
Western blotting technique in LNCaP cells. LNCaP
(androgen-dependent human prostate adenocarcinoma) cells
are a well-characterized model for AR expression and are
commonly used in AR PROTAC studies. For effective
clearance, a stable tertiary complex must form and surface
lysines on the target protein must be presented in the correct
manner for efficient ubiquitin transfer.25 Briefly, PROTAC
activities can generally be described by two numbers: DCMAX is
the percentage of protein clearance observed, and the DC50 is
the concentration required to clear half of the DCMAX.
Previously reported22 PROTAC 20a was chosen as a positive
control for the clearance assay (Figure 4A). In general our
PROTACs displayed mostly weak (DCMAX < 30%) or no
clearance effects in the cell assay (not shown). To disperse the
concern that the generally poor activity observed in cells might
be due to the chosen attachment points in the PROTACs
preventing AR binding, representative PROTACs 20a and 20i
were tested for AR binding using a radioligand displacement
assay (Figure S5). Both PROTACs retained low nM affinity for
the androgen receptor, despite the different AR ligands, the
different linker, and the attachment of the VHL ligand. This is
supportive of our docking model (Figure 2B) and indicates
that the phenolic attachment point of the AR ligands in our
PROTACs is compatible with AR binding in vitro. The amide
attachment point of the VHL ligand has been well validated in
other PROTAC studies.4−6
F
https://dx.doi.org/10.1021/acsmedchemlett.0c00194
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
The activity data for two structurally related PROTACs, 14
and 15, are shown in Figure 4B and serve to exemplify how
small changes in PROTACs can lead to profound changes in
cell activity. PROTAC 15 is active (DCMAX = 33%, DC50 = 10
nM), but PROTAC 14 is inactive, despite possessing the
highest A2B permeability measured, albeit with a high efflux
ratio. The only structural difference between the two
PROTACs is whether the E3 ligase ligand possesses a carbonyl
group in the cereblon-binding moiety. Interestingly, Wang et
al. have reported similar large activity differences regarding this
carbonyl group in a set of BET PROTACs.33 There are many
reasons why PROTAC AR activity data may show idiosyn-
cratic SAR, beyond permeability. In particular, recent papers
demonstrate that in the case of cereblon-binding PROTACs,
small changes in PROTAC chemical structure can significantly
reduce the intended activity of a bona fide PROTAC and
induce off-target GSPT1 degradation, thereby altering the
PROTAC’s mode of action.34
AUTHOR INFORMATION
Corresponding Author
■
John Skidmore − ALBORADA Drug Discovery Institute,
University of Cambridge, Cambridge CB2 0AH, United
Kingdom; orcid.org/0000-0001-9108-7858;
Email: js930@cam.ac.uk
Authors
Duncan E. Scott − ALBORADA Drug Discovery Institute,
University of Cambridge, Cambridge CB2 0AH, United
Kingdom; orcid.org/0000-0003-1917-9576
Timothy P. C. Rooney − ALBORADA Drug Discovery
Institute, University of Cambridge, Cambridge CB2 0AH,
United Kingdom; orcid.org/0000-0001-6788-5526
Elliott D. Bayle − Alzheimer’s Research UK UCL Drug
Discovery Institute, The Cruciform Building, University College
London, London WC1E 6BT, United Kingdom; The Francis
Crick Institute, London NW1 1AT, United Kingdom;
orcid.org/0000-0002-7089-3858
In summary, even our simplest PROTAC molecules formed
from moderately sized linkers connecting relatively small AR
ligands and E3 ligase ligands possess low PAMPA and Caco-2
A2B permeability. If this is a general observation, it is clear that
PROTAC permeability is not “rule-breaking”. Caco-2 perme-
ability data sheds a little more light on the permeability profiles
of some PROTACs, and evidence of structure-dependent
engagement with transport efflux proteins is observed. We
recommend that Caco-2 permeability therefore, rather than
PAMPA permeability, might be a more useful measurement. A
recent publication from Cantrill et al. supports these
conclusions.35 Broadly, a catalytic mode of action is supported
by our permeability data set, that substoichiometric levels of
PROTACs in a cell can catalyze the clearance of a target
protein population. Some minimal permeability threshold for
PROTACs may exist but is likely to be much lower than is
usually recognized for Rule of 5-compliant small molecules.
The chemical stability of a PROTAC in a cell over time will
also be important for the continued degradation of protein.
Recently, methodology to measure concentrations of small
molecules in cellular compartments has become increasingly
utilized and has already been applied to PROTACs.36 The use
of this technology has emerged as a potentially useful tool to
provide insights into PROTAC uptake into cells, though
interpretation might be complicated by compound adhering to
the outer cell membrane. To investigate the real time clearance
of proteins, Riching et al. have recently reported methods to
monitor PROTAC-induced protein degradation.37 Under-
standing the cellular uptake of a PROTAC, its stability in
the cell over time, and the kinetics of protein clearance will be
critical in informing design of better PROTACs. By better
understanding the connection between PROTAC structure
and cellular efficacy, we will be able to rationally design better
molecules and translate PROTAC molecules into the clinic
more efficiently.
Tashfina Mirza − ALBORADA Drug Discovery Institute,
University of Cambridge, Cambridge CB2 0AH, United
Kingdom; orcid.org/0000-0002-5780-9789
Henriette M. G. Willems − ALBORADA Drug Discovery
Institute, University of Cambridge, Cambridge CB2 0AH,
United Kingdom; orcid.org/0000-0001-7196-5975
Jonathan H. Clarke − ALBORADA Drug Discovery Institute,
University of Cambridge, Cambridge CB2 0AH, United
Kingdom; orcid.org/0000-0002-4079-5333
Stephen P. Andrews − ALBORADA Drug Discovery Institute,
University of Cambridge, Cambridge CB2 0AH, United
Kingdom
Complete contact information is available at:
https://pubs.acs.org/10.1021/acsmedchemlett.0c00194
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Funding
This work was funded by Alzheimer’s Research UK (grant:
ARUK-2015DDI-CAM), with support from the ALBORADA
Trust. The ALBORADA Drug Discovery Institute is core
funded by Alzheimer’s Research UK (registered charity No.
1077089 and SC042474).
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
The authors wish to thank Simon Edwards for synthesis of
compound 2 and David Rubinsztein for helpful discussions.
■
ABBREVIATIONS
■
AR, androgen receptor; PROTACs, PROteolysis TArgeting
Chimeras; PAMPA, parallel artificial membrane permeability;
TPSA, total polar surface area; A2B, Apical to Basolateral;
B2A, Basolateral to Apical; BLQ, Below Limit of Quantifica-
tion
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acsmedchemlett.0c00194.
General experimental information; organic synthesis and
compound characterization, AR affinity data, logD and
HSA data and protocol, cell activity protocol, and AR
binding assay (PDF)
REFERENCES
■
(1) Lai, A. C.; Crews, C. M. Induced protein degradation: an
emerging drug discovery paradigm. Nat. Rev. Drug Discovery 2017, 16
(2), 101−114.
G
https://dx.doi.org/10.1021/acsmedchemlett.0c00194
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
(2) Schneekloth, J. S., Jr.; Fonseca, F. N.; Koldobskiy, M.; Mandal,
A.; Deshaies, R.; Sakamoto, K.; Crews, C. M. Chemical genetic
control of protein levels: selective in vivo targeted degradation. J. Am.
Chem. Soc. 2004, 126 (12), 3748−54.
(3) Schneekloth, A. R.; Pucheault, M.; Tae, H. S.; Crews, C. M.
Targeted intracellular protein degradation induced by a small
molecule: En route to chemical proteomics. Bioorg. Med. Chem.
Lett. 2008, 18 (22), 5904−8.
(4) Zengerle, M.; Chan, K. H.; Ciulli, A. Selective Small Molecule
Induced Degradation of the BET Bromodomain Protein BRD4. ACS
Chem. Biol. 2015, 10 (8), 1770−7.
(5) Tan, L.; Gray, N. S. When Kinases Meet PROTACs. Chin. J.
Chem. 2018, 36 (10), 971−977.
(6) Bondeson, D. P.; Mares, A.; Smith, I. E.; Ko, E.; Campos, S.;
Miah, A. H.; Mulholland, K. E.; Routly, N.; Buckley, D. L.; Gustafson,
J. L.; Zinn, N.; Grandi, P.; Shimamura, S.; Bergamini, G.; Faelth-
Savitski, M.; Bantscheff, M.; Cox, C.; Gordon, D. A.; Willard, R. R.;
Flanagan, J. J.; Casillas, L. N.; Votta, B. J.; den Besten, W.; Famm, K.;
Kruidenier, L.; Carter, P. S.; Harling, J. D.; Churcher, I.; Crews, C. M.
Catalytic in vivo protein knockdown by small-molecule PROTACs.
Nat. Chem. Biol. 2015, 11 (8), 611−7.
(7) Sun, X. Y.; Wang, J.; Yao, X.; Zheng, W.; Mao, Y.; Lan, T. L.;
Wang, L. G.; Sun, Y. H.; Zhang, X. Y.; Zao, Q. Y.; Zhao, J. G.; Xiao, R.
P.; Zhang, X. Q.; Ji, G. J.; Rao, Y. A chemical approach for global
protein knockdown from mice to non-human primates. Cell Discovery
2019, 5, DOI: 10.1038/s41421-018-0079-1
(8) Churcher, I. Protac-Induced Protein Degradation in Drug
Discovery: Breaking the Rules or Just Making New Ones? J. Med.
Chem. 2018, 61 (2), 444−452.
(9) Edmondson, S. D.; Yang, B.; Fallan, C. Proteolysis targeting
chimeras (PROTACs) in ’beyond rule-of-five’ chemical space: Recent
progress and future challenges. Bioorg. Med. Chem. Lett. 2019, 29
(13), 1555−1564.
Jagerschmidt, C.; Fleury, D.; Guedin, D.; Deprez, P. Discovery of
diarylhydantoins as new selective androgen receptor modulators. J.
Med. Chem. 2012, 55 (19), 8225−35.
(20) Crews, C. M.; Crew, A. P.; Dong, H.; Wang, J.; Qian, Y.; Siu,
K.; Jin, M. Imide-based modulators of proteolysis and associated
methods of use. WO/2016/197032, 2016.
(21) Basaria, S.; Collins, L.; Dillon, E. L.; Orwoll, K.; Storer, T. W.;
Miciek, R.; Ulloor, J.; Zhang, A.; Eder, R.; Zientek, H.; Gordon, G.;
Kazmi, S.; Sheffield-Moore, M.; Bhasin, S. The safety, pharmacoki-
netics, and effects of LGD-4033, a novel nonsteroidal oral, selective
androgen receptor modulator, in healthy young men. J. Gerontol., Ser.
A 2013, 68 (1), 87−95.
(22) Meizhong, J.; Crew, A. P.; Dong, H.; Wang, J.; Siu, K.; Ferraro,
C.; Chen, X.; Qian, Y. Compounds and methods for the targeted
degradation of the androgen receptor. WO/2016/118666, 2016.
(23) Gryder, B. E.; Akbashev, M. J.; Rood, M. K.; Raftery, E. D.;
Meyers, W. M.; Dillard, P.; Khan, S.; Oyelere, A. K. Selectively
targeting prostate cancer with antiandrogen equipped histone
deacetylase inhibitors. ACS Chem. Biol. 2013, 8 (11), 2550−60.
(24) Shiau, A. K.; Barstad, D.; Loria, P. M.; Cheng, L.; Kushner, P.
J.; Agard, D. A.; Greene, G. L. The structural basis of estrogen
receptor/coactivator recognition and the antagonism of this
interaction by tamoxifen. Cell 1998, 95 (7), 927−37.
(25) Ottis, P.; Toure, M.; Cromm, P. M.; Ko, E.; Gustafson, J. L.;
Crews, C. M. Assessing Different E3 Ligases for Small Molecule
Induced Protein Ubiquitination and Degradation. ACS Chem. Biol.
2017, 12 (10), 2570−2578.
(26) Chessum, N. E. A.; Sharp, S. Y.; Caldwell, J. J.; Pasqua, A. E.;
Wilding, B.; Colombano, G.; Collins, I.; Ozer, B.; Richards, M.;
Rowlands, M.; Stubbs, M.; Burke, R.; McAndrew, P. C.; Clarke, P. A.;
Workman, P.; Cheeseman, M. D.; Jones, K. Demonstrating In-Cell
Target Engagement Using a Pirin Protein Degradation Probe
(CCT367766). J. Med. Chem. 2018, 61 (3), 918−933.
(27) Tae, H. S.; Sundberg, T. B.; Neklesa, T. K.; Noblin, D. J.;
Gustafson, J. L.; Roth, A. G.; Raina, K.; Crews, C. M. Identification of
hydrophobic tags for the degradation of stabilized proteins.
ChemBioChem 2012, 13 (4), 538−41.
(28) Gustafson, J. L.; Neklesa, T. K.; Cox, C. S.; Roth, A. G.;
Buckley, D. L.; Tae, H. S.; Sundberg, T. B.; Stagg, D. B.; Hines, J.;
McDonnell, D. P.; Norris, J. D.; Crews, C. M. Small-Molecule-
Mediated Degradation of the Androgen Receptor through Hydro-
phobic Tagging. Angew. Chem., Int. Ed. 2015, 54 (33), 9659−62.
(29) Thiehoff, C.; Rey, Y.; Gilmour, R. The Fluorine Gauche Effect:
A Brief History. Isr. J. Chem. 2017, 57 (1−2), 92−100.
(30) van Breemen, R. B.; Li, Y. Caco-2 cell permeability assays to
measure drug absorption. Expert Opin. Drug Metab. Toxicol. 2005, 1
(2), 175−85.
(10) Doak, B. C.; Over, B.; Giordanetto, F.; Kihlberg, J. Oral
druggable space beyond the rule of 5: insights from drugs and clinical
candidates. Chem. Biol. 2014, 21 (9), 1115−42.
(11) Agrawal, S.; Ashokraj, Y.; Bharatam, P. V.; Pillai, O.;
Panchagnula, R. Solid-state characterization of rifampicin samples
and its biopharmaceutic relevance. Eur. J. Pharm. Sci. 2004, 22 (2−3),
127−44.
(12) Matsson, P.; Kihlberg, J. How Big Is Too Big for Cell
Permeability? J. Med. Chem. 2017, 60 (5), 1662−1664.
(13) Foley, C. A.; Potjewyd, F.; Lamb, K. N.; James, L. I.; Frye, S. V.
Assessing the Cell Permeability of Bivalent Chemical Degraders Using
the Chloroalkane Penetration Assay. ACS Chem. Biol. 2020, 15 (1),
290−295.
(14) Narayanan, R.; Mohler, M. L.; Bohl, C. E.; Miller, D. D.;
Dalton, J. T. Selective androgen receptor modulators in preclinical
and clinical development. Nucl. Recept. Signaling 2008, 6, No. e010.
(15) Han, X.; Wang, C.; Qin, C.; Xiang, W.; Fernandez-Salas, E.;
Yang, C. Y.; Wang, M.; Zhao, L.; Xu, T.; Chinnaswamy, K.;
Delproposto, J.; Stuckey, J.; Wang, S. Discovery of ARD-69 as a
Highly Potent Proteolysis Targeting Chimera (PROTAC) Degrader
of Androgen Receptor (AR) for the Treatment of Prostate Cancer. J.
Med. Chem. 2019, 62 (2), 941−964.
(31) Varma, M. V.; Gardner, I.; Steyn, S. J.; Nkansah, P.; Rotter, C.
J.; Whitney-Pickett, C.; Zhang, H.; Di, L.; Cram, M.; Fenner, K. S.; El-
Kattan, A. F. pH-Dependent solubility and permeability criteria for
provisional biopharmaceutics classification (BCS and BDDCS) in
early drug discovery. Mol. Pharmaceutics 2012, 9 (5), 1199−212.
(32) Hilgendorf, C.; Ahlin, G.; Seithel, A.; Artursson, P.; Ungell, A.
L.; Karlsson, J. Expression of thirty-six drug transporter genes in
human intestine, liver, kidney, and organotypic cell lines. Drug Metab.
Dispos. 2007, 35 (8), 1333−40.
(16) Cohen, S.; Nathan, J. A.; Goldberg, A. L. Muscle wasting in
disease: molecular mechanisms and promising therapies. Nat. Rev.
Drug Discovery 2015, 14 (1), 58−74.
(33) Qin, C.; Hu, Y.; Zhou, B.; Fernandez-Salas, E.; Yang, C. Y.; Liu,
L.; McEachern, D.; Przybranowski, S.; Wang, M.; Stuckey, J.;
Meagher, J.; Bai, L.; Chen, Z.; Lin, M.; Yang, J.; Ziazadeh, D. N.;
Xu, F.; Hu, J.; Xiang, W.; Huang, L.; Li, S.; Wen, B.; Sun, D.; Wang, S.
Discovery of QCA570 as an Exceptionally Potent and Efficacious
Proteolysis Targeting Chimera (PROTAC) Degrader of the
Bromodomain and Extra-Terminal (BET) Proteins Capable of
Inducing Complete and Durable Tumor Regression. J. Med. Chem.
2018, 61 (15), 6685−6704.
(17) La Spada, A. R.; Wilson, E. M.; Lubahn, D. B.; Harding, A. E.;
Fischbeck, K. H. Androgen receptor gene mutations in X-linked spinal
and bulbar muscular atrophy. Nature 1991, 352 (6330), 77−9.
(18) Wager, T. T.; Hou, X.; Verhoest, P. R.; Villalobos, A. Moving
beyond rules: the development of a central nervous system
multiparameter optimization (CNS MPO) approach to enable
alignment of druglike properties. ACS Chem. Neurosci. 2010, 1 (6),
435−49.
(34) Yang, J.; Li, Y.; Aguilar, A.; Liu, Z.; Yang, C. Y.; Wang, S.
Simple Structural Modifications Converting a Bona fide MDM2
PROTAC Degrader into a Molecular Glue Molecule: A Cautionary
(19) Nique, F.; Hebbe, S.; Peixoto, C.; Annoot, D.; Lefrancois, J. M.;
Duval, E.; Michoux, L.; Triballeau, N.; Lemoullec, J. M.; Mollat, P.;
Thauvin, M.; Prange, T.; Minet, D.; Clement-Lacroix, P.; Robin-
H
https://dx.doi.org/10.1021/acsmedchemlett.0c00194
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
pubs.acs.org/acsmedchemlett
Letter
Tale in the Design of PROTAC Degraders. J. Med. Chem. 2019, 62
(21), 9471−9487.
(35) Cantrill, C.; Chaturvedi, P.; Rynn, C.; Petrig Schaffland, J.;
Walter, I.; Wittwer, M. B. Fundamental aspects of DMPK
optimization of targeted protein degraders. Drug Discovery Today,
published online 13th March 2020; DOI: 10.1016/j.dru-
dis.2020.03.012.
(36) McCoull, W.; Cheung, T.; Anderson, E.; Barton, P.; Burgess, J.;
Byth, K.; Cao, Q.; Castaldi, M. P.; Chen, H.; Chiarparin, E.; Carbajo,
R. J.; Code, E.; Cowan, S.; Davey, P. R.; Ferguson, A. D.; Fillery, S.;
Fuller, N. O.; Gao, N.; Hargreaves, D.; Howard, M. R.; Hu, J.;
Kawatkar, A.; Kemmitt, P. D.; Leo, E.; Molina, D. M.; O’Connell, N.;
Petteruti, P.; Rasmusson, T.; Raubo, P.; Rawlins, P. B.; Ricchiuto, P.;
Robb, G. R.; Schenone, M.; Waring, M. J.; Zinda, M.; Fawell, S.;
Wilson, D. M. Development of a Novel B-Cell Lymphoma 6 (BCL6)
PROTAC To Provide Insight into Small Molecule Targeting of
BCL6. ACS Chem. Biol. 2018, 13 (11), 3131−3141.
(37) Riching, K. M.; Mahan, S.; Corona, C. R.; McDougall, M.;
Vasta, J. D.; Robers, M. B.; Urh, M.; Daniels, D. L. Quantitative Live-
Cell Kinetic Degradation and Mechanistic Profiling of PROTAC
Mode of Action. ACS Chem. Biol. 2018, 13 (9), 2758−2770.
I
https://dx.doi.org/10.1021/acsmedchemlett.0c00194
ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX