Identification of GDC-0810 (ARN-810), an Orally Bioavailable Selective Estrogen Receptor Degrader (SERD) that Demonstrates Robust Activity in Tamoxifen-Resistant Breast Cancer Xenografts

Article abstract of DOI:10.1021/acs.jmedchem.5b00054

Approximately 80% of breast cancers are estrogen receptor alpha (ER-α) positive, and although women typically initially respond well to antihormonal therapies such as tamoxifen and aromatase inhibitors, resistance often emerges. Although a variety of resistance mechanism may be at play in this state, there is evidence that in many cases the ER still plays a central role, including mutations in the ER leading to constitutively active receptor. Fulvestrant is a steroid-based, selective estrogen receptor degrader (SERD) that both antagonizes and degrades ER-α and is active in patients who have progressed on antihormonal agents. However, fulvestrant suffers from poor pharmaceutical properties and must be administered by intramuscular injections that limit the total amount of drug that can be administered and hence lead to the potential for incomplete receptor blockade. We describe the identification and characterization of a series of small-molecule, orally bioavailable SERDs which are potent antagonists and degraders of ER-α and in which the ER-α degrading properties were prospectively optimized. The lead compound 11l (GDC-0810 or ARN-810) demonstrates robust activity in models of tamoxifen-sensitive and tamoxifen-resistant breast cancer, and is currently in clinical trials in women with locally advanced or metastatic estrogen receptor-positive breast cancer.

Full text of DOI:10.1021/acs.jmedchem.5b00054

Subscriber access provided by NEW YORK UNIV
Article
Identification of GDC-0810 (ARN-810), an orally bioavailable
Selective Estrogen Receptor Degrader (SERD) that demonstrates
robust activity in tamoxifen-resistant breast cancer xenografts
Andiliy Lai, Mehmet Kahraman, Steven Govek, Johnny Nagasawa, Celine Bonnefous, Jackie Julien,
Karensa Douglas, John Sensintaffar, Nhin Lu, Kyoung-jin Lee, Anna Aparicio, Josh Kaufman,
Jing Qian, Gang Shao, Rene Prudente, Michael J Moon, James D Joseph, Beatrice Darimont,
Daniel Brigham, Kate Grillot, Richard Heyman, Peter J Rix, Jeffrey H Hager, and Nicholas D Smith
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b00054 • Publication Date (Web): 16 Apr 2015
Downloaded from http://pubs.acs.org on April 16, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Identification of GDCꢀ0810 (ARNꢀ810), an orally
bioavailable Selective Estrogen Receptor Degrader
(SERD) that demonstrates robust activity in
tamoxifenꢀresistant breast cancer xenografts
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
58
59
60
a
a
a
a
a
Andiliy Lai, Mehmet Kahraman, Steven Govek , Johnny Nagasawa, Celine Bonnefous,
a
a
b
b
c
Jackie Julien, Karensa Douglas, John Sensintaffar, Nhin Lu, Kyoungꢀjin Lee, Anna
c
c
b
b
b
b
Aparicio, Josh Kaufman, Jing Qian, Gang Shao, Rene Prudente, Michael J. Moon, James
b
b
b
b
b
D. Joseph, Beatrice Darimont, Daniel Brigham, Kate Grillot, Richard Heyman, Peter J.
c
b
a*
Rix, Jeffrey H. Hager, Nicholas D. Smith
Seragon Pharmaceuticals, 12780 El Camino Real, Suite 302, San Diego, California 92130.
a
b
c
Department of Chemistry, Department of Biology,
Department of Drug Safety and
Disposition
ABSTRACT
Approximately eighty percent of breast cancers are estrogen receptor alpha (ERꢀα) positive, and
although women typically initially respond well to antiꢀhormonal therapies such as tamoxifen
and aromatase inhibitors, resistance often emerges. Although a variety of resistance mechanism
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
may be at play in this state, there is evidence that in many cases the ER still plays a central role,
including mutations in the ER leading to constitutively active receptor. Fulvestrant is a steroidꢀ
based, selective estrogen receptor degrader (SERD) that both antagonizes and degrades ERꢀα,
and is active in patients who have progressed on antiꢀhormonal agents. However fulvestrant
suffers from poor pharmaceutical properties and must be administered by intraꢀmuscular
injections that limit the total amount of drug that can be administered, and hence lead to the
potential for incomplete receptor blockade. We describe the identification and characterization of
a series of smallꢀmolecule, orally bioavailable SERDs which are potent antagonists and
degraders of ERꢀα, and in which the ERꢀα degrading properties were prospectively optimized.
The lead compound 11l (GDCꢀ0810 or ARNꢀ810) demonstrates robust activity in models of
tamoxifenꢀsensitive and tamoxifenꢀresistant breast cancer, and is currently in clinical trials in
women with locally advanced or metastatic estrogen receptorꢀpositive breast cancer.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
KEYWORDS: Estrogen receptor, estrogen receptor degrader, SERD, antagonist, tamoxifenꢀ
resistant, breast cancer, indazole, GDCꢀ0810, ARNꢀ810.
INTRODUCTION
The estrogen receptors ERꢀα and ERꢀβ are members of the nuclear hormone receptor
superfamily, and are ligandꢀregulated transcription factors that mediate the activity of estrogens
in a number of important physiological processes, including reproduction, cardiovascular
1
maintenance and bone density/remodeling. ERꢀα has been an important target in the
pharmaceutical industry for many years, with ER antagonists such as tamoxifen (1), and
aromatase inhibitors (that block the production of estrogen) such as anastrozole being a mainstay
2
in the treatment of ERꢀpositive breast cancer. Tamoxifen (1) and its active metabolite 4ꢀ
ACS Paragon Plus Environment
2

Page 3 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
hydroxytamoxifen (2) (Figure 1) are selective estrogen receptor modulators (SERMs), in that
1
they can behave as agonists or antagonists depending on the tissue context. Unfortunately,
despite many women with breast cancer initially responding well to tamoxifen, resistance often
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
58
59
60
3
emerges. Although a variety of resistance mechanism may be at play in this state, there is
evidence that in many cases ER still plays a central role, and that tumor cells still utilize proꢀ
4
growth signaling pathways downꢀstream of ER. Recently, there has been increasing clinical
evidence following treatment with aromatase inhibitors, of the involvement of mutations in the
ligandꢀbinding domain of ERꢀα rendering it transcriptionally active in the absence of ligand,
5
leading to resistance to antiꢀhormonal therapy. Based on this observation it might be expected
that a nextꢀgeneration ERꢀα ligand that both antagonizes and degrades (i.e. removes) the ERꢀα
would be of benefit to patients that have progressed on antiꢀhormonal therapy such as tamoxifen
or aromatase inhibitors.
Figure 1: Estrogen receptor ligands
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Fulvestrant (3) is a steroidꢀbased estrogen receptor antagonist originally selected for
development due its lack of agonism across all tissue types tested; however, it was
retrospectively found to also be a selective estrogen receptor degrader (SERD) as well as
6
antagonist. Fulvestrant is approved for the treatment of ERꢀpositive, metastatic breast cancer in
postmenopausal women whose disease has progressed following treatment with endocrine
therapy (i.e. tamoxifen or an aromatase inhibitor). However, fulvestrant is limited by poor
pharmaceutical properties and is being administered by intraꢀmuscular injections (500 mg total
dose in 2 x 5mL injections), that limits the total amount of drug that may be administered and
hence circulating levels. Even at the recently approved higher dose of 500 mg, FESꢀPET imaging
data supports the notion that the limitations of fulvestrant intraꢀmuscular dosing leads to
7
incomplete estrogen receptor engagement in the clinic. GWꢀ5638 (4) is a nonꢀsteroidal ER
antagonist originally selected for development for its lack of undesired uterotrophic activity in
tamoxifen treated animals and patients, but like fulvestrant, was also retrospectively determined
ACS Paragon Plus Environment
4

Page 5 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
8
to be a SERD. Like tamoxifen, GWꢀ5638 may be considered a prodrug as it is metabolized to
produce the active metabolite GWꢀ7604 (5) inꢀvivo. Importantly, tamoxifenꢀresistant breast
cancer cell lines were not crossꢀresistant to GWꢀ5638 treatment, and the compound was
ultimately progressed to a Phase 1 clinical trial, although development was discontinued for
unknown reasons. Thus the activity of fulvestrant in relapsed patients in the clinic, and GWꢀ5638
in models of tamoxifenꢀresistance, support the notion that ER still plays a critical role in
tamoxifenꢀresistant breast cancer. We therefore set out to identify nonꢀsteroidal SERDs where
the ERꢀα degradation efficacy of the compounds was optimized prospectively, and were orally
bioavailable allowing rapid achievement of robust, therapeutic levels of exposure.
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
58
59
60
CHEMISTRY
The compounds presented in this manuscript were prepared as outlined in Schemes 1 to 6. The
synthetic strategy involved preparation of an appropriately substituted alkyne which was
converted in the key synthetic step to the desired tetraꢀsubstituted alkene with regioꢀ and stereoꢀ
control. Heteroaryl alkynes 17cꢀj were synthesized from commercial heteroaryl halides 13cꢀj as
shown in Scheme 1. Heteroaryl halides 13cꢀj, containing a free NH gave low yields in the
Sonogashira coupling and so a variety of protecting groups were employed. Benzothiazolone
13c, benzoxazolone 13d and indole 13e were protected with MOM, trityl and acetyl respectively
to give 14c, 14d and 14e. Meanwhile benzimidazole 13f, benzotriazole 13g and indazoles 13hꢀj
were protected with THP using 3,4ꢀdihydroꢀ2Hꢀpyran and catalytic pyridinium pꢀ
9
toluenesulfonate to afford 14 hꢀj in 56ꢀ89% yield. In the case of benzimidazole 14f and
benzotriazole 14g, a 1:1 mixture of THPꢀregioisomers was obtained. This mixture was used
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
without separation in the Sonagashira reaction, and then subsequently separated. Having suitably
protected heteroaryl halides in hand, compounds 14cꢀj were converted to alkynes 17cꢀj via one
of two routes. Initially for 17gꢀj, a three step procedure was utilized. Coupling of 14gꢀj and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
10
trimethylsilylacetylene under Sonogashira crossꢀcoupling conditions afforded alkynes 15gꢀj
which were then desilylated using potassium carbonate. Then, a baseꢀmediated alkylation of
terminal alkynes 16gꢀj with ethyl iodide gave the desired internal alkynes 17gꢀj. Later, a more
efficient one pot procedure was used: halides 14cꢀf were coupled with butꢀ1ꢀynꢀ1ꢀ
yltrimethylsilane in the presence of palladium (II) acetate catalyst, 1,1’ꢀ
bis(diphenylphosphino)ferrocene, copper iodide and cesium carbonate to afford the desired
internal alkynes 17cꢀf directly in 47ꢀ93% yields.
Scheme 1
ACS Paragon Plus Environment
6

Page 7 of 71
Journal of Medicinal Chemistry
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
58
59
60
In a similar way, alkynes 17a and 17b were synthesized from heteroaryl bromide 13a or aryl
bromide 13b as shown in Scheme 2. Thus, bromopyridine 13a was protected as the NꢀMOM
intermediate in 40% yield, while sulfonamide 13b was protected as the carbamate in 83% yield.
Using Sonagashira crossꢀcoupling conditions, both intermediates were converted to the alkynes
17a and 17b in 72ꢀ77% yield.
Scheme 2:
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
To efficiently carry out structureꢀactivityꢀrelationship (SAR) studies, we utilized a regioselective,
stereospecific, threeꢀcomponent coupling followed by deprotection and saponification to prepare
11
final compounds 9cꢀh (Scheme 3). To this end, heteroaryl alkynes 17cꢀh, iodobenzene and
boronic acid 19 were coupled in 13ꢀ82% yield using catalytic Pd(PhCN) Cl and K CO in
2
2
2
3
DMFꢀwater. The tetraꢀsubstituted olefins 20cꢀh were formed with good regioselectivity
typically >10: 1), and any undesired regioisomer and biaryl (derived from Suzuki coupling of 18
(
and 19) sideꢀproducts were removed by silica gel chromatography. The fully constructed tetraꢀ
substituted alkenes 20cꢀh were then deꢀprotected under acidic condition to give acrylic esters
21cꢀh in high yield. Finally, the acrylic ester was hydrolyzed using lithium hydroxide to afford
the acrylic acids 9cꢀh. In an analogous way, alkynes 17aꢀb and 17iꢀj were converted to the
respective final compounds 9aꢀb, 9iꢀj.
Scheme 3
ACS Paragon Plus Environment
8

Page 9 of 71
Journal of Medicinal Chemistry
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
58
59
60
To investigate substitution on the phenylꢀlinker region, compounds 10aꢀf were prepared (Scheme
4) employing the same threeꢀcomponent coupling described in Scheme 3. Boronic acids 22 were
used as coupling partners and afforded aryl aldehydes 23aꢀf with good regioselectivity and
moderate yields (30ꢀ68%). Aldehydes 23aꢀf were then converted to (E)ꢀacrylic esters 24aꢀf via a
HornerꢀWadsworthꢀEmmons (HWE) reaction with triethylphosphonoacetate. Finally,
compounds 10aꢀf were obtained after removal of the THP protecting group and hydrolysis of the
acrylic ester under standard conditions.
Scheme 4:
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compounds investigating substitution of the pendant phenyl ring were initially prepared
(Scheme 5) using the threeꢀcomponent coupling described above. However for systems with an
orthoꢀsubtituent (i.e.11a) and/or an electron withꢀdrawing substituent (i.e. 11e), these coupling
conditions gave poor yields (< 30%), or in some cases did not work due to the competing Suzuki
crossꢀcoupling of aryl iodide 25 and boronic acid 19. Despite its inefficiency this chemistry was
used to produce intermediates 26aꢀi in 8ꢀ62% yield. Removal of the THP group and hydrolysis
of the ester gave the desired compounds 11aꢀi in good yields.
Scheme 5:
ACS Paragon Plus Environment
10

Page 11 of 71
Journal of Medicinal Chemistry
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
58
59
60
Alternate chemistry was used for the efficient synthesis of tetraꢀsubstituted alkenes containing
orthoꢀsubstituted and/or electronꢀdeficient aryls and heteroaryls as outlined in Scheme 6. Alkyne
1
2
17h was converted to bis(pinacolato)diborylꢀalkene intermediate 28 in high yield, and this
intermediate was sufficiently stable to be stored for future synthesis. More typically, it was used
immediately in a sequential bisꢀarylation to prepare the desired tetraꢀsubstituted alkene with high
1
3
regioꢀ and stereoꢀcontrol and in good yields. Thus, in a one pot reaction from 17h, intermediate
8 (not isolated) was reacted with 4ꢀiodo benzaldehyde 29 to give the intermediate aldehyde 30,
which was coupled with aryl halide 32 in the presence of catalytic PdCl (PPh ) and aqueous
2
2
3 2
KOH to give intermediates 33aꢀe and 33l in 21ꢀ99% yield overall. These were then homologated
using a HWE reaction to give 34aꢀe and 34l. It was observed that lower temperatures (0ꢀ25 ºC)
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
resulted in higher regioꢀcontrol in the key arylation reaction, and that intermediate 30 could also
be isolated and stored for future use. Alternatively, intermediate 28 was coupled with ethyl 3ꢀ(4ꢀ
iodophenyl)acrylate 31 in the presence of PdCl (PPh ) and Cs CO , followed by coupling with
2
3 2
2
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
2 in the presence of PdCl (PPh ) and aqueous potassium carbonate, to provide tetraꢀsubstituted
2 3 2
alkenes 34jꢀk. Lastly, the protecting groups of compounds 34aꢀe and 34jꢀl were removed under
standard conditions to afford the targeted acrylic acid analogs 11jꢀl and 12aꢀe.
Scheme 6:
I
O
O
O
O
O
B
O
B
29
B
B
O
O
O
O
N
+
B
a
N
N
b
N
O
O
THP
N
N
17h
28
30
THP
THP
d
c, d
Ar
Ar
Ar
f, g
e
N
N
N
OH
OEt
O
N
N
N
H
11j-l, 12a-e
O
THP
34j-l, 34a-e
O
I
THP
33a-e, 33l
CN Cl
CN
Cl
F
Ar =
OEt
31
11j
11k
11l
O
N
Cl
N
N
S
N
12a
12b
12c
12d
12e
aReagents and conditions: (a) cat. Pt(PPh3)4, bis(pinacolato)diboron, 2-MeTHF, 80 °C; (b) 29, cat. PdCl2(PPh3)2,
Cs2CO3, 2-MeTHF, 0 oC-rt; (c) 31, cat. PdCl2(PPh3)2, Cs2CO3, 2-MeTHF, 0 oC-rt; (d) Ar-X (32), cat. PdCl2(PPh3)2,
KOH (or K2CO3 in 1:1 2-MeTHF/DMSO), 2-MeTHF, reflux 21-99% (3 steps); (e) Triethyl-phosphonoacetate, LiCl,
DBU, MeCN, rt, 56-96%; (f) HCl, EtOH; (g) LiOH, THF-EtOH, 26-93% (2 steps)
ACS Paragon Plus Environment
12

Page 13 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RESULTS AND DISCUSSION
In addition to more typical assays used in nuclear hormone receptor drug discovery programs
such as binding and transcriptional assays, we also developed a cell based immunofluorescence
assay (inꢀcell western assay, ‘ICW’) which monitored total ERꢀα protein levels in an MCFꢀ7
breast cancer cell line. This assay was central to the program and was used to drive SAR to
maximize ERꢀα degradation (SERD efficacy) of our ligands. Figure 2 shows an example of the
assay with benchmark compounds, including fulvestrant, which was used as an internal control
in each run of the assay, and we report degradation efficacy of compounds as a percent of this
fulvestrant control. It should be noted that a small percentage of the ERꢀα still remains after
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
58
59
60
fulvestrant treatment. This may reflect an inaccessible ERꢀα population resistant to degradation
such as a nuclear insoluble fraction that is revealed upon cell permeabilization and antibody
1
4
interrogation.
a
Figure 2: MCFꢀ7 ERꢀα degradation assay
a: ERꢀα inꢀcell western in MCFꢀ7 breast cancer cells in phenolꢀred free RPMI medium containing 5% charcoal dextran treated FBS; readings
taken after 4 hours incubation using a SPꢀ1 antiꢀER rabbit monoclonal antibody
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 1 shows the profiles of the SERMs arzoxifene (discontinued following a phase 3 in
1
5
patients with metastatic breast cancer) and pipendoxifene (phase 2, discontinued for unknown
1
6
reasons), and the SERDs fulvestrant and GWꢀ5638/7604 in the key ERꢀα degradation assay. A
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
number of points are worth highlighting, first that arzoxifene, although potent in the ERꢀα
degradation assay, has relatively poor degradation efficacy at 71% of the fulvestrant control.
Secondly, GWꢀ7604, the active metabolite of GWꢀ5638, is much more potent than the parent in
this ERꢀα degradation assay. At 91% of the fulvestrant control, GWꢀ7604 is also an efficacious
degrader of ERꢀα.
Table 1: MCFꢀ7 ERꢀα degradation profile of ER modulators
a
MCFꢀ7 ERꢀα Degradation
Compound
Maximum Efficacy %
EC50
ꢀM)
b
Efficacy
of control
(
Fulvestrant (3)
GWꢀ5638 (4)
GWꢀ7604 (5)
Pipendoxifene (6)
Arzoxifene (8)
0.0004
0.39
0.0017
0.0001
0.0002
94%
82%
86%
83%
67%
100%
90%
91%
87%
71%
a: ERꢀα inꢀcell western in MCFꢀ7 breast cancer cells in phenolꢀred free RPMI medium containing 5% charcoal dextran treated FBS; readings
taken after 4 hours incubation using a SPꢀ1 antiꢀER rabbit monoclonal antibody b: Efficacy recorded as percent of efficacy of fulvestrant control.
For a prodrug strategy to be maximally effective, the active species should be efficiently
produced in humans. When we examined the crossꢀspecies microsome profile of GWꢀ5638, we
found that in all species, the active metabolite GWꢀ7604 was indeed produced, but the ratio
varied between species (Table 2). For example in rat liver microsomes, 82% of GWꢀ7604 was
produced after a 4 hour incubation, while in human liver microsomes, only 16% of GWꢀ7604
was produced. In addition, we found that another oxidative metabolite 5c was produced in
human liver microsomes (21% after 4 hours), and that it was 100ꢀfold less potent an ERꢀα
ACS Paragon Plus Environment
14

Page 15 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
degrader than GWꢀ7604 (ERαꢀdegradation EC = 0.113 ꢀM). Further the exposure of GWꢀ7604
5
0
in mice following dosing of the parent GWꢀ5638 was found to be relatively low (dosing of 100
mg/kg of GWꢀ5638, gave: AUC (GWꢀ5638) = 137 ꢀg.hr/mL, AUC (GWꢀ7604) = 13 ꢀg.hr/mL).
Thus metabolism of GWꢀ5638 to the desired active species GWꢀ7604 is inefficient across
species inꢀvitro and in mouse PK studies.
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
58
59
60
Table 2: Crossꢀspecies inꢀvitro liver microsome profile of GWꢀ5638
a
Species
Compound detected (%)
GWꢀ5638 (4) GWꢀ7604 (5) 5a 5b 5c
Mouse
Rat
Dog
50
16
92
63
21
82
5
0
0
0
0
0
0
0
0
30
1
3
Human
16
21
a: Following 4 hours incubation in liver microsomes
We investigated a variety of scaffolds toward our goal of identifying fully efficacious ERꢀα
degraders that had good exposure following oral dosing, and this paper will describe our
investigation of the GWꢀ5638 triphenylalkene scaffold. We initially explored the SAR in the
region of the phenyl ring that is the site of metabolism of GWꢀ5638 to the active 4ꢀ
hydroxyphenyl metabolite, GWꢀ7604 (Table 3).
Table 3: SAR of the phenol region of the triphenylalkene scaffold
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
MCFꢀ7
Viability
a
ERꢀα Degradation
b
1
Compound
R
c
EC (ꢀM)
Efficacy
IC (ꢀM)
5
0
50
4
(GWꢀ5638)
0.39
0.0017
6.28
90%
0.985
0.005
ND
5 (GWꢀ7604)
91%
d
9
a
ND
d
9
b
1.05
0.101
0.023
ND
2.31
0.158
0.056
9
c
92%
90%
9
d
9
e
0.003
0.957
0.014
0.003
93%
0.009
1.18
d
9
f
ND
9
g
92%
90%
92%
0.053
0.006
9h
9i
0.006
0.357
0.021
0.57
d
9
j
ND
a: ERꢀα inꢀcell western in MCFꢀ7 breast cancer cells in phenolꢀred free RPMI medium containing 5% charcoal dextran treated FBS; readings
taken after 4 hours incubation using a SPꢀ1 antiꢀER rabbit monoclonal antibody b: MCFꢀ7 viability assay in RPMI medium containing 10% FBS;
ACS Paragon Plus Environment
16

Page 17 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
day incubation c: Efficacy recorded as percent of efficacy of fulvestrant control d: Not determined due to incomplete floor of dose response
curve because of relative low potency of compound
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
58
59
60
Single ring systems such as pyridyl 9a and sulfonamide 9b, both lost a considerable amount of
activity (> 1ꢀM in the ERꢀα degradation assay) compared to GWꢀ7604 (EC = 0.0017 ꢀM),
50
potentially due to the high polarity of these moieties and/or the pyridone nature of 9a. However
when we examined bicyclic heterocycles such as the benzothiazolone 9c and benzoxazolone 9d,
1
7
where a weakly acidic NꢀH exists in the cyclic structure (pKa 8.7), improved activity was
observed. Exploration of [6,5] systems improved the potency further; thus indole 9e (EC =
5
0
0
.003 ꢀM), benzotriazole 9g (EC = 0.014 ꢀM) and indazole 9h (EC = 0.003 ꢀM) were high
5
0
50
potency compounds, with the indole and indazole effectively equipotent to GWꢀ7604. In contrast
to these systems, basic benzimidazole 9f was considerably less active (EC = 0.957 ꢀM). In
5
0
addition to good potency, indazole 9h was also an efficacious ERꢀα degrader (90% of the
fulvestrant control), and importantly delivered good exposure in mice following oral dosing (100
mpk po: AUC = 240 ꢀg.hr/mL versus GWꢀ7604 AUC = 13 ꢀg.hr/mL, from 100 mpk po dosing
of GWꢀ5638).
The importance of the directionality of the indazole NꢀH was highlighted when different isomers
of 9h were prepared. Thus the 4ꢀindazole 9i (EC = 0.006 ꢀM), maintained potency compared
5
0
to 9h, while the 6ꢀindazole isomer 9j lost >100x potency (EC = 0.357 ꢀM), presumably as it is
5
0
unable to form the key hydrogen bond in the ligand binding domain of the estrogen receptor that
h is able to. Of note, all potent compounds were efficacious ERꢀα degraders, achieving ≥ 90%
9
of the efficacy of the fulvestrant control. This was found to be the case in general for the
triphenylalkene scaffold derivatives that contained the acrylic acid sideꢀchain. The potency in an
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
MCFꢀ7 breast cancer cell viability assay generally tracked with the ERꢀα degradation EC , with
5
0
9h and 9e both having an IC < 0.010 ꢀM in this viability assay. With the identification of
5
0
indazole 9h as a potent, efficacious SERD, we proceeded to examine the SAR of the phenyl
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
linker region of the scaffold (Table 4).
Table 4: SAR in the phenyl linker region of the triphenylalkene scaffold
MCFꢀ7
Viability
a
ERꢀα Degradation
b
Compound
Ar
c
EC (ꢀM)
Efficacy
IC (ꢀM)
50
50
9h
0.003
90%
92%
0.006
0.036
10a
0.015
0.006
0.004
0.011
1
0b
90%
90%
85%
85%
0.014
0.012
0.031
1
0c
1
0d
1
0e
0f
0.11
1.13
0.21
2.9
d
1
ND
a: ERꢀα inꢀcell western in MCFꢀ7 breast cancer cells in phenolꢀred free RPMI medium containing 5% charcoal dextran treated FBS; readings
taken after 4 hours incubation using a SPꢀ1 antiꢀER rabbit monoclonal antibody b: MCFꢀ7 viability assay RPMI medium containing 10% FBS; 5
ACS Paragon Plus Environment
18

Page 19 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
day incubation c: Efficacy recorded as percent of efficacy of fulvestrant control d: Not determined due to incomplete floor of dose response
curve because of relative low potency of compound
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
58
59
60
Although introduction of a methyl or methoxy substituent at either the ‘internal’ (10a, 10c) or
‘
external’ (10b, 10d) positions relative to the core was tolerated, it did not lead to an increase in
potency. Replacement of the phenyl ring with a heterocycle such as pyridine 10e and thiophene
0f led to a substantial dropꢀoff in potency, indicating that these heterocyclic rings are not well
1
tolerated in this region. It is interesting that 10d is one of the few potent acrylic acid examples
where a reduction of degradation efficacy was observed in the ERꢀα inꢀcell western assay. This
suggests the importance of this region for maximal ERꢀα degradation efficacy, and may reflect
the ability of the methoxy substituent to modulate the pKa of the key acrylic acid moiety.
1
8
In general, nuclear hormone receptor ligands, such as estradiol are lipophilic in nature, and so
introduction of polar groups while maintaining potency can be difficult. Throughout the program
we explored typical lipophilic groups as well as trying to incorporate heteroatoms in the series,
and Tables 5 and 6 show some of the SAR that was carried out on the pendent phenyl ring of
indazole 9h toward this goal.
Table 5: SAR of the pendent phenyl ring
MCFꢀ7
Viability
a
2
ERꢀα Degradation
b
Compound
R
ACS Paragon Plus Environment
19

EC (ꢀ
M)
Efficacy
IC (ꢀM)
Journal of Medicinal Chemistry
Page 20 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
c
50
50
9
h
H
oꢀCl
0.003
0.0007
0.002
0.010
0.003
0.028
0.036
0.0008
0.049
0.004
0.001
0.0009
0.0007
90%
92%
91%
93%
93%
87%
93%
92%
91%
93%
89%
92%
91%
0.006
0.002
0.012
0.015
0.012
0.100
0.047
0.004
0.070
0.025
0.005
0.004
0.002
1
1a
1b
1c
11d
1e
1f
1g
11h
1
mꢀCl
pꢀCl
oꢀCN
mꢀCN
pꢀCN
oꢀMe
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
1
1
oꢀSO
Me
2
1
1
1i
1j
oꢀOMe
oꢀMe, pꢀCN
oꢀCl, pꢀCN
oꢀCl, pꢀF
1
1
1k
1l
a: ERꢀα inꢀcell western in MCFꢀ7 breast cancer cells in phenolꢀred free RPMI medium containing 5% charcoal dextran treated FBS; readings
taken after 4 hours incubation using a SPꢀ1 antiꢀER rabbit monoclonal antibody b: MCFꢀ7 viability assay in RPMI medium containing 10% FBS;
5 day incubation c: Efficacy recorded as percent of efficacy of fulvestrant control
Table 5 shows that substitution at the ortho position (11a, 11d) is preferred over the metaꢀ or
paraꢀposition (11c, 11e, 11f) for both a more polar substituent (ꢀCN) or a lipophilic substituent (ꢀ
Cl). Of note, orthoꢀchloro 11a, gives a 3ꢀ5 fold increase in potency over the unsubstituted phenyl
9
h, while orthoꢀnitrile 11d is equipotent to 9h. Additional substituents were examined at the
orthoꢀposition, with a strongly polar orthoꢀsulfone substituent leading to a drop off in activity i.e.
compare 11h (ERꢀα degradation EC = 0.049 ꢀM) to orthoꢀmethyl 11g (ERꢀα degradation EC
5
0
50
=
0.0008 ꢀM). We found that adding an orthoꢀmethyl or orthoꢀchloro substituent to otherwise
less active derivatives (such as the paraꢀcyano derivative 11f; ERꢀα degradation EC = 0.036
50
ꢀ
M), resulting in disubstituted systems such as 11j and 11k, gave a substantial 30ꢀfold increase
in potency (11j and 11k: ERꢀα degradation EC ≤ 0.001 ꢀM). The orthoꢀchloro, paraꢀfluoro
5
0
derivative 11l was also a high potency compound. We also found that heterocycles could be
incorporated in this region (Table 6). The 4ꢀpyridyl 12b (ERꢀα degradation EC = 0.008 ꢀM)
5
0
ACS Paragon Plus Environment
20

Page 21 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
being the preferred substitution pattern compared to 12a and 12c (ERꢀα degradation EC > 0.10
5
0
ꢀM). Chloroꢀpyridyl 12d and thiophene 12e were also potent heterocyclic compounds in this
region. As with SAR on the indazole region, potency in the ERꢀα degradation assay generally
tracked with the MCFꢀ7 breast cancer cell viability IC , with compounds such as 11a, 11k, 11l
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
58
59
60
5
0
and 12d all having an IC < 0.005 ꢀM.
5
0
Table 6: Select pendant heterocycles are tolerated
MCFꢀ7
Viability
a
ERꢀα Degradation
b
3
Compound
R
c
EC (ꢀM)
Efficacy
IC (ꢀM)
5
0
50
1
1
1g
2a
0.0008
0.16
92%
0.004
0.20
N
d
ND
1
2b
2c
2d
2e
0.008
0.11
89%
91%
90%
0.016
0.15
1
1
0.002
0.004
1
0.0005
90%
0.002
a: ERꢀα inꢀcell western in MCFꢀ7 breast cancer cells in phenolꢀred free RPMI medium containing 5% charcoal dextran treated FBS; readings
taken after 4 hours incubation using a SPꢀ1 antiꢀER rabbit monoclonal antibody b: MCFꢀ7 viability assay in RPMI medium containing 10% FBS;
5
day incubation c: Efficacy recorded as percent of efficacy of fulvestrant control d: Not determined due to incomplete floor of dose response
curve because of relative low potency of compound
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
As illustrated in Tables 3ꢀ6, a number of potent and efficacious ERꢀα degraders were identified.
In order to quickly determine those that had good exposure and activity following oral dosing,
we heavily utilized a 4ꢀday immature rat uterine wet weight assay, which allowed rapid
determination of compound antagonist activity in an ER responsive tissue (immature rat uterus)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
19
while competing against the native ER ligand estradiol (i.e. antagonist mode). Drug plasma
levels were also obtained from this assay. Table 7 shows the reduction in immature rat uterine
weight for select compounds when compared to estradiolꢀonly treated animals (the table also
includes exposure in mice from separate PK studies).
Table 7: Immature rat uterine wet weight assay (antagonist mode) and mouse pharmacokinetics
a
Rat Uterine Wet Weight
Mouse PK
ERꢀα
Percent
Drug levels
po AUC
Degradation
Cpd
b
c
inhibition
(nM)
(ꢀg.hr/mL)
EC (ꢀM)
50
9
d
0.023
0.003
0.014
0.003
28%
47%
15%
69%
0.2
ND
ND
3.8
3.0
9.4
2.3
5.8
9
e
9
g
9h
ACS Paragon Plus Environment
22

Page 23 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
9
i
0.006
0.0007
0.002
35%
82%
81%
27%
72%
48%
60%
81%
7%
0.1
2.8
15
11a
3.7
11
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
58
59
60
11b
11
11d
0.003
ND
1.7
0.2
11g
0.0008
0.001
7.2
d
11j
14
4.3
d
11k
0.0009
0.0007
0.008
10
2.9
11l
5.2
7.6
0.5
10
8.8
12b
0.081
0.11
3.8
12d
0.002
66%
e
12e
0.0005
105%
a: Compounds administered orally (1 mg/kg po) followed by an oral dose of 0.1 mg/kg ethynyl estradiol each day for 3 days. On the fourth day,
24 hours post dose, the animals were euthanized and the uterus was removed and weighed. Plasma was taken at the time point indicated. b: 100 x
[
(VehicleEE – Cpd)/(VehicleEE – Vehicle)], where ‘VehicleEE’ is the uterine wet weight of the of the vehicle animals dosed with ethynyl estradiol
(0.1 mg/kg); ‘Cpd’ is the uterine wet weight of the of the compound treated animals dosed with ethynyl estradiol (0.1 mg/kg); ‘Vehicle’ is the
uterine wet weight of the of the vehicle animals c: Measured at 24 hrs after last dose unless otherwise indicated d: Measured at 6 hrs after last
dose e: In same assay, 11l = 112%
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In general, compounds that effectively antagonized estradiolꢀinduced uterine weight gain (>75%
compared to vehicle) when dosed at 1 mg/kg orally, had good potency in the ERꢀα degradation
assay (EC <0.001 ꢀM) and also had measurable drug levels at the end of the study (C > 3 nM
5
0
24
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
following a 1mg/kg po dose). In this way, the uterine assay identified promising compounds such
as 11a, 11l and 12e, while deꢀemphasizing otherwise potent inꢀvitro compounds such as indole
9e or indazole 9i. Although 12e demonstrated robust activity in the UWW model, it had lower
exposure in mouse PK (mouse po AUC @ 10 mg/kg = 3.8 ꢀg.hr/mL) than 11a and 11l (mouse
po AUC @ 10 mg/kg = 15 ꢀg.hr/mL and 8.8 ꢀg.hr/mL respectively). In addition, the presence of
a thiophene in 12e was of concern due to the potential for reactive metabolite formation inꢀvivo,
20
and so we did not pursue 12e further. Compounds 11a and 11l had a similar inꢀvitro and uterine
assay profile, and were subsequently profiled in additional assays including a tamoxifenꢀresistant
MCFꢀ7 xenograft model. Ultimately 11l was selected for development and Table 8 shows a
comparison to clinical benchmarks. 11l is a potent ERꢀα binder (IC = 6.1 nM), a full
5
0
transcriptional antagonist with no agonism (3x ERE, IC = 2 nM), and displays good potency
5
0
and efficacy in ERꢀα degradation (EC = 0.7 nM) and MCFꢀ7 breast cancer cell viability (IC =
5
0
50
2.5 nM) assays.
Table 8: 11l shows a favorable inꢀvitro profile compared to benchmark compounds
a
c
Binding (nM)
Transcriptional
ERꢀα Degradation
Compound
d
MCFꢀ7 Viability
b
Antagonism
e
e
ERꢀα IC50 ERꢀβ IC50
IC50 (nM)
EC50 (nM) Efficacy
IC50 (nM)
2.5
Efficacy
11l
6.1
8.8
2.0
0.6
18
0.6
0.3
0.7
0.4
1.8
0.1
0.2
91%
100%
91%
87%
71%
99%
100%
98%
89%
73%
Fulvestrant (3)
GWꢀ7604 (5)
Pipendoxifene (6)
Arzoxifene (8)
24
21
0.6
5.0
0.4
0.4
6.0
ND
42
3.0
ND
240
ACS Paragon Plus Environment
24

Page 25 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
a: Competitive radiometric binding in TEGM ; 2.5 nM ERꢀα or 1.5 nM ERꢀβ, 1.5 nM HꢀE2. b: 3X EREꢀTKꢀLuc reporter vector, phenolꢀred free
RPMI containing 10% charcoal dextran treated FBS plus 17βꢀestradiol (0.1 nM), 24 hrs. c: ERꢀα inꢀcell western in MCFꢀ7 breast cancer cells in
phenolꢀred free RPMI medium containing 5% charcoal dextran treated FBS; readings taken after 4 hours incubation using a SPꢀ1 antiꢀER rabbit
monoclonal antibody d: MCFꢀ7 viability assay in RPMI medium containing 10% FBS; 5 day incubation e: Efficacy recorded as percent of
efficacy of fulvestrant control
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
58
59
60
Of particular note are the optimized levels of ERꢀα degradation for 11l (91%) compared to
arzoxifene (8) (71%) and pipendoxifene (6) (87%) and that these efficacy levels translate into
increased efficacy in the MCFꢀ7 viability assay (11l: 99% efficacy; 4; arzoxifene (8): 73%
efficacy; pipendoxifene (6): 89% efficacy). Indeed, we observed a correlation between ERꢀα
degradation efficacy and MCF7ꢀviability efficacy which extended across chemical scaffolds and
is captured graphically in Figure 3. Although not definitively causative, this correlation indicates
that optimization of ERꢀα degradation efficacy leads to compounds with improved efficacy in a
breast cancer cell viability assay. As will be presented below, this increase in inꢀvitro efficacy
also led to improved efficacy in a xenograft model of tamoxifenꢀresistance.
Figure 3: Optimization of efficacy in the ERꢀα degradation assay correlates with increased
efficacy in the MCFꢀ7 breast cancer cell line viability assay
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
21
b
22
:
SERM for osteoporosis; : SERM for osteoporosis
The robust degradation of 11l in the in cell western assay was confirmed in a western blot assay
as shown in Figure 4. Similar to fulvestrant, 11l displays robust degradation of ERꢀα in this
assay format. Of note is that, although an agonist, estradiol also leads to degradation of ERꢀα,
while 4ꢀhydroxytamoxifen (2) as is known the literature, leads to stabilization of ERꢀα (an
increase in protein levels).
a
Figure 4: Western blot analysis confirms 11l is a robust degrader of ERꢀα
ACS Paragon Plus Environment
26

Page 27 of 71
Journal of Medicinal Chemistry
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
58
59
60
a: MCFꢀ7 cells in RPMI supplemented with 10% charcoal dextran treated FBS; compound treatment for 20 hours, cell lysates
separated electrophoretically using NuPAGE 4ꢀ12% Bis Tris Gels; SPꢀ1 antiꢀERα antibody and imaged using a LIꢀCOR Odyssey
infrared imaging system.
The pharmacokinetic profile of 11l (Table 9) shows it is a low clearance molecule across species,
with good bioavailability (40ꢀ60%). As would be expected for a lipophilic carboxylic acid, the
compound is highly bound to plasma proteins (>99.5% across species), and has a low to
moderate volume of distribution (V = 0.2 to 2.0 L/Kg across species).
ss
a
Table 9: Cross species pharmacokinetic of 11l
Species
CL
Vss
t1/2
Cmax
AUC
(Oral)
Oral
%F
(
mL/min/kg)
(
L/kg)
(iv)
(Oral)
(µg/mL)
4.4
(
µg·hr/mL)
(hr)
Mouse
Rat
11
14
1.2
1.9
0.2
0.5
4.2
8.8
5.8
69
61
49
61
42
3.9
11
1.8
20
Dog
1.6
7.0
Cyno
13
6.1
10
a: Dosed intravenously at 3 mg/kg and per orally at 10 mg/kg in PEG400 / PVP / TW80 / 0.5% CMC in water, 9:0.5:0.5:90
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 71
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Metabolism profiling of 11l conducted in microsomes, with phase 1 and phase 2 cofactors added,
showed the acyl glucuronide to be the major metabolite across species. The stability of this acyl
glucuronide in physiological medium (pH 7.4 phosphate buffer at 37ºC) was examined to
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
23
understand its potential for reactivity (Figure 5).
Figure 5: Percent remaining of 11l acyl glucuronide in pH 7.4 phosphate buffer at 37 °C
5
4
3
2
1
0
11l acyl glucuronide
Diclofenac acyl
glucuronide
0
4
8
12
Time (hr)
Compared to the diclofenac Oꢀglucuronide positive control (t1/2 = 0.9 hr), the stability of the 11l
Oꢀglucuronide was good at t = 8.1 hr, leading to the conclusion that the potential risk for
1
/2
idiosyncratic toxicity of 11l was low. Cytochrome P450 inhibition profiling of 11l indicated it
had little to no inhibition against CYP1A2, CYP2D6 or CYP3A4 (IC > 20 ꢁM), modest
5
0
inhibitory effect on CYP2C9 and CYP2C19 (IC = 2.2 ꢁM and 3.3 ꢁM respectively), and potent
50
inhibition of CYP2C8 (IC of < 0.1 ꢁM). Inhibition of CYP2C8 was not considered a major
50
liability as few therapeutics are metabolized exclusively by CYP2C8, suggesting the potential for
24
drugꢀdrug interactions is low
.
ACS Paragon Plus Environment
28

Page 29 of 71
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Selectivity of 11l over other nuclear hormone receptors was found to be good. In transcriptional
reporter assays for the mineralocorticoid (MR), progesteroneꢀA (PRꢀA), progesteroneꢀB (PRꢀB)
and glucocorticoid (GR) receptors, 11l had minimal activity (IC > 1 µM). While in binding
5
0
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
58
59
60
assays, 11l displayed little activity towards the androgen receptor (AR; IC > 4 µM) and GR
5
0
(IC = 0.99 µM). In a CEREP panel of radioligand binding assays for 54 diverse targets
5
0
(proteinꢀfree conditions), 11l at 10 µM displayed >50% binding to 7 targets. However, the only
appreciable activity seen in cellꢀbased functional assays for these 7 targets was for the GABAꢀ
gated chloride channel (IC = 1.1 ꢀM) and the dopamine transporter (IC = 3.4 ꢀM). In light of
50
50
the high plasma protein binding of 11l, these low micromolar activities were not considered an
issue as selectivity for the estrogen receptor is >500 fold (MCFꢀ7 viability = 2 nM; ERꢀα
degradation = 0.7 nM). 11l produced no significant inhibition of hERG current at doses up to 30
µM in a patch clamp assay, and was Ames negative in the TAꢀ98 and TAꢀ100 tester strains.
Compound 11l was extensively characterized in xenograft models of breast cancer. In these
models supplemental estradiol (E2) is used to drive xenograft tumor growth inꢀvivo (in the form
of an implanted slowꢀrelease pellet), and these high E2 ligand concentrations (300ꢀ400 pg/ml)
require higher concentrations of 11l in order to outꢀcompete the native ligand (E2) and so
mediate an inhibitory effect. This is of relevance as plasma E2 levels in the target patient
2
5
population of postꢀmenopausal women (~5 pg/mL) are substantially lower than those of the
various preꢀclinical models. Thus preꢀclinical models run in the context of high E2 plasma
concentrations likely significantly overestimate 11l dose and plasma levels necessary to drive
efficacy in a postꢀmenopausal setting.
ACS Paragon Plus Environment
29