Development of the First Low Nanomolar Liver Receptor Homolog-1 Agonist through Structure-guided Design

Article abstract of DOI:10.1021/acs.jmedchem.9b00753

As a key regulator of metabolism and inflammation, the orphan nuclear hormone receptor, liver receptor homolog-1 (LRH-1), has potential as a therapeutic target for diabetes, nonalcoholic fatty liver disease, and inflammatory bowel diseases (IBD). Discovery of LRH-1 modulators has been difficult, in part due to the tendency for synthetic compounds to bind unpredictably within the lipophilic binding pocket. Using a structure-guided approach, we exploited a newly discovered polar interaction to lock agonists in a consistent orientation. This enabled the discovery of the first low nanomolar LRH-1 agonist, one hundred times more potent than the best previous modulator. We elucidate a novel mechanism of action that relies upon specific polar interactions deep in the LRH-1 binding pocket. In an organoid model of IBD, the new agonist increases expression of LRH-1-controlled steroidogenic genes and promotes anti-inflammatory gene expression changes. These studies constitute major progress in developing LRH-1 modulators with potential clinical utility.

Full text of DOI:10.1021/acs.jmedchem.9b00753

Subscriber access provided by Nottingham Trent University
Article
Development of the first low nanomolar Liver Receptor
Homolog-1 agonist through structure-guided design
Suzanne Mays, Autumn R. Flynn, Jeffery Cornelison, C. Denise Okafor, Hongtao
Wang, Guohui Wang, Xiangsheng Huang, Heather Donaldson, Elizabeth Millings,
Rohini Polavarapu, David Moore, John Calvert, Nathan T. Jui, and Eric A. Ortlund
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00753 • Publication Date (Web): 16 Aug 2019
Downloaded from pubs.acs.org on August 16, 2019
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 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 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.
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 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Development of the first low nanomolar Liver
Receptor Homolog-1 agonist through structure-
guided design
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
Suzanne G. Mays1, Autumn R. Flynn2, Jeffery L. Cornelison2, C. Denise Okafor1, Hongtao
Wang4, Guohui Wang4, Xiangsheng Huang4, Heather N. Donaldson1, Elizabeth J. Millings1,3,
Rohini Polavarapu3, David D. Moore5, John W. Calvert3, Nathan T. Jui2, and Eric A. Ortlund1*
1. Department of Biochemistry, Emory University School of Medicine, Atlanta, Georgia 30322,
USA
2. Department of Chemistry, Emory University, Atlanta, Georgia 30322, USA
3. Department of Surgery, Carlyle Fraser Heart Center, Emory University, Atlanta, Georgia
30322, USA
4. Department of Pediatrics, Section of Gastroenterology, Baylor College of Medicine and Texas
Children’s Hospital, Houston, Texas 77030, USA
5. Department of Molecular and Cell Biology, Baylor College of Medicine, Houston, Texas 77030,
USA
KEYWORDS Liver Receptor Homolog-1 (LRH-1, NR5A2), agonist, nuclear receptor, x-ray
crystallography, drug discovery
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 41
1
2
3
4
ABSTRACT
5
6
7
8
9
As a key regulator of metabolism and inflammation, the orphan nuclear hormone receptor,
Liver Receptor Homolog-1 (LRH-1), has potential as a therapeutic target for diabetes,
nonalcoholic fatty liver disease, and inflammatory bowel diseases. Discovery of LRH-1
modulators has been difficult, in part due to the tendency for synthetic compounds to bind
unpredictably within the lipophilic binding pocket. Using a structure-guided approach, we
exploited a newly-discovered polar interaction to lock agonists in a consistent orientation. This
enabled the discovery of the first low nanomolar LRH-1 agonist, one hundred times more potent
than the best previous modulator. We elucidate a novel mechanism of action that relies upon
specific polar interactions deep in the LRH-1 binding pocket. In an organoid model of
inflammatory bowel disease, the new agonist increases expression of LRH-1-conrolled
steroidogenic genes and promotes anti-inflammatory gene expression changes. These studies
constitute major progress in developing LRH-1 modulators with potential clinical utility.
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
INTRODUCTION
Liver Receptor Homolog-1 (LRH-1; NR5A2) is a nuclear hormone receptor (NR) that is
highly expressed in the liver and tissues of endodermal origin. It is indispensable during
embryonic development, where it plays a role in maintenance of pluripotency1, as well as in the
development of the liver and pancreas2. In adults, LRH-1 controls diverse transcriptional programs
in different tissues related to metabolism, inflammation, and cellular proliferation. Targeted
metabolic pathways include bile acid biosynthesis3, reverse cholesterol transport4-5, de novo
lipogenesis6-7, and glucose phosphorylation and transport8-9. The ability to modulate lipid and
glucose metabolism suggests therapeutic potential for LRH-1 agonists in metabolic diseases such
2
ACS Paragon Plus Environment

Page 3 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
as nonalcoholic fatty liver disease, type II diabetes, and cardiovascular disease. Indeed, the
phospholipid LRH-1 agonist dilauroylphosphatidylcholine (DLPC; PC 12:0/12:0) improves
glucose tolerance, insulin sensitivity, and triglyceride levels in obese mice6. These anti-diabetic
effects occur in an LRH-1-dependent manner and have been primarily attributed to a reduction of
de novo lipogenesis6. In addition, targeting LRH-1 in the gut has therapeutic potential for
inflammatory bowel disease, where LRH-1 overexpression ameliorates disease-associated
inflammation and cell death10.
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
While small molecule LRH-1 modulators are highly sought, the large and lipophilic LRH-
1 binding pocket has been extremely challenging to target. A promising class of agonists developed
by Whitby and colleagues features a bicyclic hexahydropentalene core scaffold11-12. The best-
studied of this class, named RJW100, was discovered as a part of an extensive synthetic effort to
improve acid stability and efficacy of a related compound, GSK847012 (Figure 1A). We recently
determined the crystal structure of LRH-1 bound to RJW100 and made a surprising discovery: it
exhibits a completely different binding mode than GSK8470, such that the bicyclic cores of the
two agonists are perpendicular to each other (Figure 1A)13. As a result, the two compounds use
different mechanisms to activate LRH-1 but exhibit similar activation profiles in luciferase
reporter assays13. A tendency for ligands in this class to bind unpredictably in the hydrophobic
pocket has likely been a confounding factor in agonist design. However, insights from the LRH-
1-RJW100 structure have provided new strategies to improve activity.
In the LRH-1-RJW100 crystal structure, the ligand exo hydroxyl group contacts a network
of water molecules deep in the ligand binding pocket (Figure 1B). This water network coordinates
a small group of polar residues (e.g. Thr352, His390, and Arg393) in an otherwise predominantly
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 41
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
Figure 1. Structure-based design of LRH-1 agonists. A. Top, chemical structures of the agonists GSK8470 and
RJW100. Bottom, superposition of GSK8470 and RJW100 (from PDB 3PLZ and 5L11, respectively)12-13 show the
very different binding modes for these similar agonists. B. RJW100 interacts with LRH-1 residue Thr352 via water.
The four water molecules shown coordinate a group of polar residues deep in the binding pocket. The colored circles
indicate the areas targeted by modifications to the RJW100 scaffold in this work.
hydrophobic pocket. The endo RJW100 diastereomer adopts a nearly identical pose and makes the
same water-mediated contact with Thr352, supporting the idea that this interaction is a primary
driver of ligand orientation13. Using both a RJW100 analog lacking a hydroxyl group and a LRH-
1 Thr352Val mutation, we demonstrated that this interaction is required for RJW100-mediated
activation of LRH-113. As the basis for the current studies, we hypothesized that strengthening
this and other polar interactions in the vicinity could anchor ligand conformation, enabling more
predictable targeting of desired parts of the pocket. We designed, synthesized, and evaluated novel
compounds around the hexahydropentalene scaffold with the primary aim of strengthening polar
contacts in the deep part of the binding pocket (the deep part of the pocket is hereafter abbreviated
“DPP”). This systematic, structure-guided approach enabled the discovery of an agonist more
4
ACS Paragon Plus Environment

Page 5 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
potent than RJW100 by two orders of magnitude in luciferase reporter assays. We present three
crystal structures of LRH-1 bound to novel agonists, which depict the modified polar groups
projecting into the DPP. The best new agonist modulates expression of LRH-1-controlled anti-
inflammatory genes in intestinal organoids, suggesting therapeutic potential for treating
inflammatory bowel diseases. This breakthrough in LRH-1 agonist development is a crucial step
in developing potential new treatments for metabolic and inflammatory diseases.
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
RESULTS
Locking the agonist in place with polar interactions. Our structural studies have revealed that
highly similar LRH-1 synthetic agonists can bind unpredictably within the hydrophobic binding
pocket, which has presented a challenge for improving agonist design in a rational manner13. We
reasoned that strengthening contacts within the DPP may anchor synthetic compounds in a
consistent orientation and improve potency. To evaluate this hypothesis, we synthesized
RJW100 analogs with bulkier polar groups in place of the RJW100 hydroxyl (R1), aiming to
displace bridging waters and to generate direct interactions with Thr352 or other nearby polar
residues (Figure 1B). In parallel, we synthesized compounds designed to interact with other sites
in the DPP by (1) modifying the external styrene (R2) to promote interactions with helix 3 or to
fill a hydrophobic pocket in the vicinity or (2) incorporating hydrogen bond donors at the meta
position of the internal styrene (R3) to promote hydrogen bonding with His390 (also via water
displacement) (Figure 1B). To prepare this compound library, we utilized a diastereoselective
variant of Whitby’s zirconecene-mediated Pauson-Khand-type cyclization14. This highly
modular approach unites three readily available precursors (an enyne, an alkyne, and 1,1
dibromoheptane) to generate all-carbon bridgehead [3.3.0]-bicyclic systems with varying
functionality at positions R1, R2, and R3 (Figure 2A). R1 was most conveniently varied through
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 41
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
Figure 2. Synthesis of LRH-1-targeted compounds. A. Overview of synthetic strategy used to generate agonists
based on modification of the [3.3.0]-bicyclic hexahydropentalene scaffold. B. Modifications to the scaffold
evaluated in this study, grouped by site of modification by colored boxes. C. Summary of EC50 and efficacy relative
to RJW100 (RE). RE was calculated as described in the methods section. RJW100 RE = 1.0 and EC50 = 1.5 +/- 0.4
µM. The abbreviation “i.a.” refers to inactive compounds for which EC50 values could not be calculated.
modification of the RJW100 alcohol to yield derivatives 1–8, which were synthesized separately
as both the endo (N) or exo (X) diastereomers (Figure 2B). Oxygen-linked analogs 3 and 5 were
6
ACS Paragon Plus Environment

Page 7 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
formed directly from the diastereomerically-appropriate parent alcohol. Nitrogen-linked analogs
1, 2, 4, 6, and 8 were prepared through alcohol activation (mesylate S4) and substitution (azide
S2, nitrile S3) (Figure 2B and online supplementary materials). Alteration of R2 was
accomplished by introducing phenylacetylene derivatives as the alkyne in the cyclization step
(Figure 2A), generating 9–15 (Figure 2B). R3 variants 16–23 were prepared using functionalized
enyne starting materials. Detailed chemical syntheses of all intermediates and tested compounds
are provided in the online supplementary materials.
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
Discovery of the first low nanomolar LRH-1 agonist. We evaluated the new compounds using
differential scanning fluorimetry (DSF), since entropic gain from displacement of buried water
molecules or favorable energetics from bond formation would result in global stabilization of the
LRH-1-agonist complex. DSF assays were paired with cellular luciferase reporter assays to
determine effects on LRH-1 transcriptional activity. Luciferase data are summarized in Figure 2C
and dose-response curves are shown in Figure S1.
As previously observed13, 15, RJW100 stabilizes the LRH-1 ligand binding domain (LBD) by
around 3 ºC relative to a PL ligand in DSF assays (Figure 3A). While the R2- and R3-modified
compounds (9-23) destabilize the receptor relative to RJW100 (Figure 3A) and tend to be poor
activators (Figure 2C, S1), certain R1 modifications are highly stabilizing, with Tm values 3-8 °C
higher than RJW100 (Figure 3A and online supplementary material). There is a striking
correlation between potency in luciferase reporter assays and LRH-1 stabilization by DSF for the
R1-modified compounds, where lower EC50s are associated with higher Tm values (Pearson
correlation coefficient = -0.71; p = 0.0009, Figure 3B). This correlation provides a direct link
between cellular activity and receptor stabilization. There is no correlation between Tm and EC50
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 41
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
Figure 3. Optimization of R1 modification improves potency by two orders of magnitude. A. DSF assays
demonstrate that the site of modification, R1 substituent size, and stereochemistry affect global LRH-1 stabilization.
Colored bars represent EC50s relative to RJW100 as indicated in the legend. Each bar represents three experiments
conducted in triplicate. *, p< 0.05 for Tm decrease versus RJW100. #, p< 0.05 Tm increase versus RJW100.
Significance was assessed by one-way ANOVA followed by Dunnett’s multiple comparison’s test. Dotted line
indicates the Tm change induced by RJW100 relative to the phospholipid agonist, DLPC. B. Scatter plot showing
the correlation between Tm. shift in DSF assay (x-axis) and EC50 from luciferase reporter assays (y-axis) for the R1-
modified compounds. Data were analyzed by linear regression (curved lines are the 95% confidence interval. C.
Scatter plot comparing potency (EC50) and efficacy relative to RJW100 (relative efficacy, RE) for all compounds for
which EC50 values could be calculated. Dots are color-coded by site of modification (as indicated in Figure 2). The
black dot is RJW100. The EC50 values and efficacies of compounds 2N, 5N, and 6N are indicated. Relative efficacy
was calculated as described in the methods section. D. Dose response curves comparing 6N and RJW100 in
luciferase reporter assays. Each point represents the mean ± SEM for three experiments conducted in triplicate. E.
The significance of difference in potency for 6N versus RJW100 was determined by a two-tailed, paired Student’s t-
test from parallel experiments.
8
ACS Paragon Plus Environment

Page 9 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
for the R2- and R3-modified compounds (data not shown), suggesting that improved potency is
due to specific polar interactions mediated by the R1 group.
7
8
9
The R1 modifications are diverse, ranging from small to large polar groups, including
hydrogen bond donors and acceptors and endo and exo diastereomers (Figure 2B). Both size and
stereochemistry of the R1 group are important for activity. Mid-sized polar groups, mainly
tetrahedral in geometry, tend to increase potency relative to RJW100 (Figure 2C). The close
relationship between R1 size, agonist potency, and LRH-1 stabilization is evident looking at DSF
results, where a strong peak in stabilization occurs for compounds 5 – 6 and 8N (Figure 3A).
Another strong trend among the data is that endo diastereomers are better activators (and more
stabilizing) than the corresponding exo diastereomers (as seen for the triazoles 7, sulfamides 6,
and acetamides 2, Figure 2–3). While the compounds display a wide range of potencies and
efficacies, the endo sulfamide (6N) stands out as being the most potent (Figure 3C). With an EC50
of 15 nM, 6N is two orders of magnitude more potent than RJW100 (Figure 3D-E). This is the
first discovery of a low-nanomolar LRH-1 modulator, representing a leap forward in developing
agonists for this challenging target.
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
DPP contacts drive LRH-1 activation by 6N. The improved potency of 6N is particularly striking
considering that a very similar, highly stabilizing compound (5N) is not much more potent or
effective for transcriptional activation than RJW100 (Figure 3C). The dramatic increase in
potency for 6N relative to 5N is driven by replacement of oxygen with nitrogen in the R1 linker,
as this is the only difference between the two compounds. Remarkably, a nitrogen-containing
linker improves potency relative to an oxygen linker for several pairs of compounds that differ
only at this site (Figure 4A). The NH linker also contributes to selectivity for LRH-1 over its
closest homolog, Steroidogenic Factor-1 (SF-1). Compound 6N is a weaker activator of SF-1 than
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 41
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
Figure 4. A hydrogen-bond donating nitrogen linker in the R1 group improves potency and selectivity. A-B.
Comparison of potencies and efficacies for four sets of compounds that are identical except for the presence of a R1
linker containing an oxygen (red dots) or nitrogen (blue dots). Compounds with no activity at doses up to 30 μM are
(3N and 3X) are plotted as having EC50s of 30 μM and Relative Efficacy of 0 for illustrative purposes. C. Dose
response curves comparing activation of LRH-1 and SF-1 by select compounds. Significance of differences in
activities of each compound for LRH-1 versus SF-1 was determined by two-way ANOVA followed by Sidak’s
multiple comparison’s test. *, p < 0.05.
LRH-1, and 2N (the endo acetamide) displays no activity against SF-1 while strongly activating
LRH-1 (Figure 4C). In contrast, 5N and RJW100 equally activate both receptors (Figure 4C).
10
ACS Paragon Plus Environment

Page 11 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To investigate the role of the R1 linker in agonist activity and to gain insights into
mechanisms underlying the potency of 6N, we determined the X-ray crystal structure of 6N bound
to the LRH-1 LBD at a resolution of 2.23 Å (Figure 5A, Table S1). For comparison and to
delineate the function of the NH-containing linker, we also determined structures of LRH-1 bound
to 2N (with an NH-linker, 2.2 Å) and 5N (with an oxygen linker, 2.0 Å) (Table S1). The complexes
were crystallized with a fragment of the coactivator protein, Transcriptional
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
Intermediary Factor 2 (Tif2), which is bound at the AF-2 activation function surface (AFS) at the
interface between helices 3, 4, and the activation function helix (AF-H, Figure 5A). Overall
protein conformation does not differ greatly and is similar to the LRH-1-RJW100 structure (root
mean square deviations are within 0.2 Å). The ligands are well-defined by the electron density,
with the exception of the alkyl “tails” (Figure 5B). Disorder in the tail is also seen in the endo
RJW100 structure12 and may be a general feature of endo agonists with this scaffold.
One of the main goals for these studies was to develop ligands that bind with consistent
positions of the bicyclic cores. These structures demonstrate that this strategy was successful.
Superposition of RJW100, 2N 5N, and 6N from the crystal structures shows nearly identical
conformation of the agonists’ cores and phenyl groups, with slight variation in the positions of the
R1 headgroups (Figure 5C). All three headgroups protrude into the DPP, filling space typically
occupied by one or more water molecules and making several polar contacts (Figure 5D). For
both 5N and 6N, there is strong tetrahedral density indicating the position of the R1 groups;
however, analysis of structure B factors and ensemble refinement16 suggest that the R1 groups are
somewhat mobile and capable of making transient interactions in the pocket that differ from the
modeled states (Figure S3). Studies with LRH-1 mutants helped to elucidate mechanistic
differences between these agonists.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 41
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
Figure 5. Crystal structures of LRH-1 bound to novel agonists. A. Overall structure of the LRH-1 LBD (grey) bound
to 6N (blue sticks). Tif2 is shown in green. The dotted line indicates a disordered region that could not be modeled.
B. Omit maps for 2N, 5N, and 6N. Maps are FO-FC, contoured at 2.5σ. C. Superposition of ligands from the crystal
structures showing a consistent position of the cores of the modified agonists compared to RJW100. D. Close view
of LRH-1 binding pocket with either 5N, 2N, or 6N bound showing a subset of interactions made by the R1 groups.
ACS Paragon Plus Environment
12

Page 13 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Colored circles highlight interactions that are important for LRH-1 activation by each agonist. Red spheres are water
molecules (the grey sphere in the LRH-1-6N structure is a water molecule typically present in the LRH-1 pocket that
could not be modeled due to poor crystallographic order). Red dotted lines indicate hydrogen bonds, and black dotted
lines indicate hydrophobic contacts. The interaction indicated by the grey dotted line in the LRH-1-5N structure is
outside of hydrogen-bonding distance in the structure but important for activity in mutagenesis studies. E. Luciferase
reporter assays showing how the interactions made by the agonists affect LRH-1 activity. The A349F mutation
occludes the DPP and was used as a negative control. Each bar represents the mean ± SEM for three independent
experiments conducted in triplicate. Cells were treated with 10 μM 2N, 10 µM 5N, or 0.3 μM 6N for 24 hours
(concentrations chosen based on agonist EC50 toward wild-type LRH-1). *, p< 0.05 by two-way ANOVA followed
by Dunnett’s multiple comparisons test. PDB codes for the structures compared in this figure are as follows: LRH-
1-RJW100, 5L11; LRH-1-5N, 6OQX; LRH-1-6N, 6OQY; LRH-1-2N, 6OR1.
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
While the binding modes of the three agonists are similar, mutagenesis studies show that
they activate LRH-1 through different mechanisms (Figure 5E). The first major difference is with
the Thr352 interaction. Both 5N and 6N directly interact with Thr352, but the differential impact
of a Thr352Val mutation shows that this interaction only contributes to agonist-mediated LRH-1
activity in the case of 6N (Figure 5E). Compound 2N is not well-positioned to interact with the
water coordinating Thr352 due to the planar geometry of the R1 acetamide group, and the
Thr352Val mutation has no effect on LRH-1 activity (Figure 5E). The agonists also demonstrate
a differential reliance on the interaction with M345: 5N is unable to activate a Met345L LRH-1
mutant, but 6N and 2N activate it significantly above basal levels (Figure 5E).
We were particularly interested in how interactions made by the NH linker contribute to
agonist activity. All three agonists are positioned to make water-mediated hydrogen bonds with
LRH-1 residue His390 via the R1 linkers (Figure 5D). In the case of 6N, we were unable to model
the bridging water molecule seen in the other two structures (and in other published LRH-1 LBD
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 41
1
2
3
4
5
6
7
8
9
structures13, 15, 17-20) due to weak electron density. The weak density for the water molecule is likely
a consequence of poor crystallographic order, since very few waters could be modeled in this
structure (24 total, unusual for a 2.2 Å structure). However, luciferase reporter assays with LRH-
1 mutants indicate that 6N interacts with His390 and that the interaction is critical for
transcriptional activity (Figure 5E). Compound 2N is also unable to activate the LRH-1 His390Ala
mutant, supporting the idea that a productive water-mediated interaction with His390 is made by
the NH-linker, (Figure 5E). Compound 5N, with an oxygen linker, interacts with His390 with
both the linker and sulfonyl oxygens (Figure 5D). However, 5N does not utilize the His390
interaction for activation, since mutating His390 to alanine has no effect on its ability to activate
LRH-1 (Figure 5E). Therefore, while 5N and 6N make very similar contacts, the presence of a
hydrogen bond donor in the R1 linker is uniquely able to drive activation of LRH-1 via His390.
This provides a potential mechanism through which a nitrogen linker increases agonist potency.
Compound 6N stabilizes the AFS, strengthens allosteric signaling, and promotes coactivator
recruitment. To investigate how 6N alters LRH-1 dynamics to drive receptor activation, we
determined its effects of LRH-1 conformation in solution using hydrogen-deuterium exchange
(HDX) mass spectrometry. The most significant changes occur at sites involved in ligand-driven
recruitment of coregulators: (1) the activation function surface (AFS) and (2) a region of the
receptor involved in allosteric signaling to the AFS, located near helix 6 and the beta sheets18-19
(this site is called activation function B and abbreviated “AF-B”. Relative to RJW100, 6N impacts
the conformation of AF-B by destabilizing the N-terminal portion of helix 7 and stabilizing the
loop between helices 6 and 7 (Figure 6A). Rigidification of the loop between these helices may
induce pressure to unwind helix 7, which could explain this pattern of motion. In addition to these
changes near AF-B, 6N strongly stabilizes a portion of helix 4 near the AFS (Figure 6A).
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
ACS Paragon Plus Environment
14

Page 15 of 41
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
Figure 6. Compound 6N promotes allosteric communication to the AFS and coactivator recruitment. A-B.
Differential HDX comparing 5N to RJW100 (A) or 6N to RJW100 (B). Color bar indicates the percent difference in
deuterium uptake when 5N or 6N is bound compared to RJW100. A positive number indicates more deuterium
exchange, indicating relative destabilization. A negative number indicates relative stabilization. C. MDS results
showing the strongest suboptimal paths (blue lines) between AF-B and the Tif2 coactivator (green) when the
indicated agonists are bound. The AFS is highlighted in light blue in panels A-C, and the position of AF-B is
indicated with brackets. The PBD codes for the starting models used in MD are: LRH-1-RJW100, 5L11; LRH-1-
5N, 6OQX; LRH-1-6N, 6OQY; LRH-1-2N, 6OR1. D. Compound 6N promotes recruitment of the Tif2 coactivator
to purified LRH-1 LBD in a fluorescence polarization-based binding assay. Each point represents the mean +/-
SEM for three independent experiments conducted in triplicate. *, p< 0.05 by two-way ANOVA followed by
Sidak’s multiple comparison’s test.
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 41
1
2
3
4
5
6
7
8
Compound 5N stabilizes the same region of helix 4 relative to RJW100 and 6N, but it also
destabilizes the AF-H, a critical part of the AFS that tunes coregulator associations through subtle
changes to its conformation19. (Figure 6B).
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
Since 6N alters LRH-1 conformation at AF-B and the AFS, we hypothesized that it
increases communication between these two sites. To quantify the predicted strength of agonist-
driven communication between AF-B and the AFS, we conducted 1 µs molecular dynamics
simulations (MDS) using the crystal structures as starting models. Correlated motions of residues
within a protein facilitate allosteric coupling between distant sites21-23. Communication paths
can traverse thousands of possible routes through the receptor, and the chains of residues with
the strongest patterns of correlated motion—the optimal path and a subset of suboptimal paths—
are thought to convey the most information24-25. We therefore constructed dynamical networks of
LRH-1-agonist complexes, using calculated covariance to weight the strength of communication
between pairs of residues. The resulting covariance matrices were used to identify the strongest
suboptimal paths facilitating communication between AF-B and Tif2 coactivator (bound at the
AFS). The number of strong paths markedly increases when 2N, 5N, or 6N are bound compared
to RJW100, with 6N exhibiting the strongest communication between these sites (Figure
6C). There are also significant differences in the directionality of the paths promoted by each
agonist. Although all paths traverse helix 5, indicating that correlated motion is induced in this
region, compounds 2N, 5N, and 6N also induce strong communication along helix 3. Compound
6N also induces highly interconnected communication within the AFS and the Tif2 coactivator,
including significant involvement of the AF-H. This important helix in the AFS is notably
excluded from the paths when the other agonists are bound (Figure 6C).
ACS Paragon Plus Environment
16

Page 17 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The stabilization of the AFS by 6N is associated with enhanced coactivator recruitment.
In a fluorescence-polarization based coregulator binding assay, RJW100, 5N, and 6N dose-
dependently recruit fluorescein-labeled Tif2 peptide to LRH-1 and exhibit similar EC50s (50% of
maximum Tif2 binding occurs with ~600-700 nM agonist, Figure 6D). Each curve reaches a
well-defined plateau that indicates the maximum response with saturating concentrations of
agonist; however, curve maxima are lower for RJW100 and 5N than 6N by 50–60%, which is
characteristic of partial agonists. Although the endogenous ligand has not been identified for
comparison, 6N behaves more like a full agonist than 5N or RJW100 in this assay. Therefore, we
have elucidated a novel mechanism of action utilized by 6N, whereby specific interactions by the
sulfamide and R1 linker promote allosteric signaling to the AFS, stabilizing the site of
coactivator interaction and increasing Tif2 association.
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
Compound 6N promotes expression of intestinal epithelial steroidogenic genes in humanized
LRH-1 mouse enteroids. The discovery of the first highly potent LRH-1 agonist provides the
opportunity to elucidate ligand-regulated transcriptional pathways controlled by this receptor.
LRH-1 controls local steroid hormone production in the gut epithelium26-27, and overexpression of
LRH-1 reduces inflammatory damage in immunologic mouse models of enterocolitis10. These
findings suggest therapeutic potential for LRH-1 agonists in inflammatory bowel diseases (IBD).
The recent development of methods to culture organoids of intestinal crypts (enteroids, Figure 7A)
has provided an excellent research tool for drug discovery for IBD28. When stimulated with
inflammatory cytokines, enteroids mimic features of gut epithelia in IBD10, 28. To investigate anti-
inflammatory properties of 6N, we measured the effects of the new agonist on gene expression in
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 41
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
Figure 7. Compound 6N induces intestinal epithelial steroidogenesis in humanized LRH-1 mouse enteroids. A.
Enteroids were generated through isolation and culture of intestinal crypts from mice expressing human LRH-1. B.
mouse or human (m/h) LRH-1 mRNA expression in the intestinal enteroids of Lrh-1f/f, Lrh-1KO and LRH-1h mouse
lines. C. Compound 6N induces mRNA expression of steroidogenic enzyme Cyp11a1 and Cyp11b1. D. Compound
6N induces anti-inflammatory cytokine IL-10. E. Compound 6N reduces inflammatory cytokine IL-1b (Left) and
TNFa (Right). Error bars represent the standard deviation from six biological replicates for mouse enteroids
expressing human LRH-1 (hL) and three biological replicates from Lrh-1 knockout (KO) mouse enteroids. *, p <
0.01 (One Way ANOVA followed by Tukey’s multiple comparison’s test).
ACS Paragon Plus Environment
18

Page 19 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
humanized LRH-1 mouse enteroids in the context of Tumor Necrosis Factor alpha (TNF-a)-
induced inflammation. Expression of human LRH-1 in the enteroids was verified by qRT-PCR
(Figure 7B). Treatment with 1 µM 6N in hLRH-1-expressing enteroids (but not knockout
enteroids) significantly increased mRNA expression of the steroidogenic enzymes Cyp11a1 and
Cyp11b1, which are Lrh-1 transcriptional targets (Figure 7C). There was a concomitant increase
in expression of the anti-inflammatory cytokine IL-10 and decreases in expression of the
inflammatory cytokines IL-1b and TNFa (Figure 7D-E). These data suggest a role for 6N in
reducing inflammation in the gut via upregulation of steroidogenesis. These findings are in stark
contrast with previous enteroid studies with RJW100, which was inactive at doses up to 20 µM10
(dosage information not reported but obtained by personal communication with authors) Although
the involvement of LRH-1 in IBD is clear from gain- and loss- of function studies10, the finding
that epithelial steroidogenesis can be stimulated by an agonist demonstrates the tremendous
potential for LRH-1 as a drug target for this disease.
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
DISCUSSION AND CONCLUSION
While the therapeutic potential of LRH-1 is widely recognized, this receptor has been
difficult to target with synthetic modulators. Agonists with the hexahydropentalene scaffold11-12
(such as RJW100) are promising and have been used in several studies to probe LRH-1
biology29-33. However, we have shown that small modifications to this scaffold can greatly affect
binding mode13. By exploiting a novel polar interaction in the LRH-1 DPP, we have overcome
this challenge and have made substantial progress in agonist development. Systematic variation
of three sites on the RJW100 scaffold has revealed a robust structure-activity relationship. The
modifications to the styrene sites that we examined (R2 and R3) do not significantly improve
performance and often ablate activity; however, modifications at R1 increase potency in
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 41
1
2
3
4
5
6
7
8
9
transactivation assays (Figure 2C, S1). The increased potency is associated with global receptor
stabilization by DSF promoted by tetrahedral, polar R1 substituents with endo stereochemistry
(Figure 3). In addition, the composition of the R1 group, particularly the linker, is critical for
activity. This is exemplified through the comparison of 5N and 6N, which differ only at the R1
linker. Compound 6N utilizes interactions with both Thr352 and His390 to activate LRH-1, the
latter of which is likely mediated by the linker nitrogen (Figure 5D). This novel binding mode
leads to a distinct mechanism of action for 6N compared to similar, less potent compounds,
inducing conformational changes at AF-B, stabilization of the AFS, and increasing coactivator
association (Figure 6). Results from MDS support the idea that 6N promotes very strong
allostery to the AFS, evidenced in the strong communication between the AF-B and the AFS
predicted to occur when 6N is bound compared to less potent agonists (Figure 6).
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
With three separate crystal structures, we demonstrated that polar modifications at the
RJW100 R1 group do not cause major repositioning of the scaffold (Figure 5), supporting our
hypothesis that this polar group acts as an important anchor point. This finding was not only key
to the success of the current study, but it will also greatly benefit future work. The ability to
anchor the scaffold consistently provides an opportunity to tune for additional desired effects,
such as solubility or selectivity. Moreover, the trajectory of the alkyl “tails” of these molecules
is amenable for introduction of modifications that could engage residues near the mouth of the
pocket in a PL-like manner18, 20. Initial studies in this vein have been fruitful, leading to the
discovery of highly active compounds34. Finally, the establishment of a predictable binding
mode may open avenues for antagonist design; for example, by modifying the scaffold to
promote displacement of the AF-H and recruitment of corepressors. This approach has been
successful for other nuclear receptors35-36 and could generate LRH-1 antagonists useful as
ACS Paragon Plus Environment
20

Page 21 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
therapeutics for certain cancers in which LRH-1 is aberrantly active37-43. This is an active area of
research in our laboratory.
7
8
9
In conclusion, a systematic, structure-guided approach has resulted in the discovery of the
first low nanomolar LRH-1 agonist and elucidated a novel mechanism of action. This agonist
has great potential as a tool to uncover novel aspects of LRH-1 biology and as a therapeutic for
IBD and obesity-associated metabolic diseases. Equally important, the discovery of elements
that stabilize the orientation of the hexahydropentalene scaffold and drive activation of LRH-1 is
invaluable for understanding ligand-regulation of this receptor and for future modulator design.
EXPERIMENTAL SECTION.
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
General Chemical Methods. All reactions were carried out in oven-dried glassware, equipped
with a stir bar and under a nitrogen atmosphere with dry solvents under anhydrous conditions,
unless otherwise noted. Solvents used in anhydrous reactions were purified by passing over
activated alumina and storing under argon. Yields refer to chromatographically and
spectroscopically (1H NMR) homogenous materials, unless otherwise stated. Reagents were
purchased at the highest commercial quality and used without further purification, unless otherwise
stated. n-Butyllithium (n-BuLi) was used as a 1.6 M or a 2.5 M solution in hexanes (Aldrich), was
stored at 4°C and titrated prior to use. Organic solutions were concentrated under reduced pressure
on a rotary evaporator using a water bath. Chromatographic purification of products was
accomplished using forced-flow chromatography on 230-400 mesh silica gel. Thin-layer
chromatography (TLC) was performed on 250 μm SiliCycle silica gel F-254 plates. Visualization
of the developed chromatogram was performed by fluorescence quenching or by staining using
1
13
KMnO4, p-anisaldehyde, or ninhydrin stains. H and C NMR spectra were obtained from the
Emory University NMR facility and recorded on a Bruker Avance III HD 600 equipped with cryo-
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 41
1
2
3
4
5
6
7
8
9
probe (600 MHz), INOVA 600 (600 MHz), INOVA 500 (500 MHz), INOVA 400 (400 MHz),
VNMR 400 (400 MHz), or Mercury 300 (300 MHz), and are internally referenced to residual
protio solvent signals. Data for 1H NMR are reported as follows: chemical shift (ppm), multiplicity
(s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, dd = doublet of doublets, dt =
doublet of triplets, ddd= doublet of doublet of doublets, dtd= doublet of triplet of doublets, b =
broad, etc.), coupling constant (Hz), integration, and assignment, when applicable. Data for
decoupled 13C NMR are reported in terms of chemical shift and multiplicity when applicable. High
Resolution mass spectra (HRMS) were obtained from the Emory University Mass Spectral facility.
Gas Chromatography Mass Spectrometry (GC-MS) was performed on an Agilent 5977A mass
spectrometer with an Agilent 7890A gas chromatography inlet. Liquid Chromatography Mass
Spectrometry (LC-MS) was used to obtain low-resolution mass spectra (LRMS) and was
performed on an Agilent 6120 mass spectrometer with an Agilent 1220 Infinity liquid
chromatography inlet. Purity of all tested compounds was determined by HPLC analysis, using
the methods given below (as indicated for each compound). All key target compounds possess
>95% purity as determined by HPLC.
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
Method A: A linear gradient using water and 0.1 % formic acid (FA) (Solvent A) and MeCN and
0.1% FA (Solvent B); t = 0 min, 30% B, t = 4 min, 99% B (held for 1 min), then 50% B for 1 min,
was employed on an Agilent Poroshell 120 EC-C18 2.7 micron, 3.0 mm x 50 mm column (flow
rate 1 mL/min) or an Agilent Zorbax SB-C18 1.8 micron, 2.1 mm x 50 mm column (flow rate 0.8
mL/min). The UV detection was set to 254 nm. The LC column was maintained at ambient
temperature.
Method B: A linear gradient using water and 0.1 % formic acid (FA) (Solvent A) and MeCN and
0.1% FA (Solvent B); t = 0 min, 70% B, t = 4 min, 99% B (held for 1 min), then 50% B for 1 min,
ACS Paragon Plus Environment
22

Page 23 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
was employed on an Agilent Poroshell 120 EC-C18 2.7 micron, 3.0 mm x 50 mm column (flow
rate 1 mL/min) or an Agilent Zorbax SB-C18 1.8 micron, 2.1 mm x 50 mm column (flow rate 0.8
mL/min). The UV detection was set to 254 nm. The LC column was maintained at ambient
temperature.
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
Method C: An isocratic method using 75% MeCN, 35% water, and 0.1 % FA was employed on
an Agilent Poroshell 120 EC-C18 2.7 micron, 3.0 mm x 50 mm column (flow rate 1 mL/min) or
an Agilent Zorbax SB-C18 1.8 micron, 2.1 mm x 50 mm column (flow rate 0.8 mL/min). The UV
detection was set to 254 nm. The LC column was maintained at ambient temperature.
Method D: An isocratic method using 85% MeCN, 15% water, and 0.1% FA was employed on an
Agilent Poroshell 120 EC-C18 2.7 micron, 3.0 mm x 50 mm column (flow rate 1 mL/min) or an
Agilent Zorbax SB-C C18 1.8 micron, 2.1 mm x 50 mm column (flow rate 0.8 mL/min). The UV
detection was set to 254 nm. The LC column was maintained at ambient temperature.
Chemical Synthesis of 2N, 5N, and 6N. The synthesis and characterization of key target
compounds 2N, 5N, and 6N are outlined below. Detailed synthetic procedures and characterization
data for all new compounds are provided in the supplemental section.
Endo 5-hexyl-4-phenyl-3a-(1-phenylvinyl)-1,2,3,3a,6,6a-hexahydropentalen-1-amine (1N):
Under nitrogen, a solution of azide S3N (54 mg, 0.13 mmol, 1.0 equiv) in anhydrous Et2O was
cooled to 0 °C and treated dropwise with LiAlH4 (4.0M in Et2O, 10.0 equiv). The reaction was
stirred at ambient temperature until the reaction was complete by TLC (ca. 1 h). The reaction was
cooled to 0 °C, diluted with anhydrous Et2O, and slowly treated with water (1mL/g LiAlH4).
Excess 4 M NaOH was added slowly and the solution was extracted with EtOAc three times. The
combined organic layers were washed with Rochelle’s salt and brine, dried over MgSO4, and
concentrated. The crude oil was purified by silica gel chromatography in 50% EtOAc/Hexanes
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 41
1
2
3
4
eluent (1% triethylamine) to afford the title compounds as a colorless oil. 1N: 47.9 mg, 95% yield.
5
6
1
Purity was established by Method C: tr = 0.302 min, 98.6%. H NMR (600 MHz, CDCl3) δ 7.37
7
8
9
– 7.19 (m, 10H), 5.08 (d, J = 1.4 Hz, 1H), 4.94 (d, J = 1.5 Hz, 1H), 3.30 (ddd, J = 11.0, 8.8, 5.7
Hz, 1H), 2.48 (d, J = 17.4 Hz, 1H), 2.42 (t, J = 9.0 Hz, 1H), 2.12 – 2.00 (m, 2H), 1.83 – 1.78 (m,
1H), 1.73 – 1.68 (m, 2H), 1.46 – 1.37 (m, 2H), 1.35 – 1.20 (m, 8H), 0.88 (t, J = 7.1 Hz, 3H).
N-((endo)-5-hexyl-4-phenyl-3a-(1-phenylvinyl)-1,2,3,3a,6,6a-hexahydropentalen-1-yl)acetamide
(2N) A solution of 1N (23 mg, 0.06 mmol, 1.0 equiv) in DCM was cooled to 0 °C and treated
with acetyl chloride (1.5 equiv) and triethylamine (3.0 equiv), then stirred for 1 h. The solution
was diluted with water and extracted with DCM three times. The combined organic layers were
washed with water and brine, dried with Na2SO4, filtered, and concentrated. The crude oil was
purified on silica gel in 35% EtOAc/Hexanes eluent to afford the title compound as a colorless
oil. 2N: 21.1 mg, 83% yield. Purity was established by Method D: tR = 1.00 min, 96.3 %. 1H
NMR (500 MHz, CDCl3) δ 7.35 – 7.27 (m, 5H), 7.25 – 7.22 (m, 5H), 5.35 (d, J = 8.1 Hz, 1H),
5.06 (d, J = 1.5 Hz, 1H), 5.02 (d, J = 1.5 Hz, 1H), 4.25 (dtd, J = 10.5, 8.6, 6.2 Hz, 1H), 2.66
(ddd, J = 16.9, 8.4, 1.6 Hz, 1H), 2.14 – 2.00 (m, 4H), 1.99 (s, 3H), 1.87 (dtd, J = 11.7, 6.0, 2.3
Hz, 1H), 1.76 (td, J = 12.2, 11.7, 5.8 Hz, 1H), 1.66 (ddd, J = 12.7, 5.9, 2.3 Hz, 1H), 1.43 – 1.26
(m, 1H), 1.30 – 1.18 (m, 8H), 0.87 (t, J = 7.0 Hz, 3H). 13C NMR (126 MHz, CDCl3) δ 169.3,
154.6, 143.6, 141.7, 138.9, 137.2, 129.6, 127.9, 127.8, 127.7, 126.9, 126.6, 114.8, 69.0, 59.5,
54.4, 40.9, 33.0, 32.1, 31.6, 29.8, 29.4, 27.8, 26.0, 23.6, 22.6, 14.1. LRMS (ESI, APCI) m/z:
calc’d for C30H39NO [M+H]+ 430.3, found 430.3
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
(endo)-5-hexyl-4-phenyl-3a-(1-phenylvinyl)-1,2,3,3a,6,6a-hexahydropentalen-1-yl sulfamate
(5N): A 1M solution of sulfamoyl chloride (2.5 equiv) in DMA was cooled to 0ºC. A solution of
the appropriate RJW100 alcohol isomer (endo (224.3 mg, 0.6 mmol 1.0 equiv) in DMA was
ACS Paragon Plus Environment
24

Page 25 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
added slowly, followed by triethylamine (excess, ca. 5 equiv); the resulting solution was stirred
for one hour. The solution was then diluted with water and extracted with EtOAc three times.
The combined organic layers were washed with water and brine, dried with MgSO4, filtered, and
concentrated. The crude oil was purified by silica gel chromatography in 20% EtOAc/Hexanes
eluent (with 0.5% triethylamine), to afford the title compound as a clear oil. 5N: 182 mg, 67%
yield. Purity was established by Method D: tR = 1.15 min, 95.3%. 1H NMR (500 MHz, CDCl3) δ
7.35 – 7.24 (m, 8H), 7.23 – 7.15 (m, 2H), 5.11 (s, 1H), 4.92 (s, 1H), 4.87 (td, J = 9.1, 5.2 Hz,
1H), 4.64 (s, 2H), 2.71 (d, J = 9.0 Hz, 1H), 2.60 (d, J = 17.5 Hz, 1H), 2.17 (dd, J = 17.7, 9.3 Hz,
1H), 2.10 – 2.01 (m, 3H), 1.92 – 1.83 (m, 1H), 1.83 – 1.76 (m, 1H), 1.68 (td, J = 12.6, 5.6 Hz,
1H), 1.45 – 1.35 (m, 2H), 1.32 – 1.16 (m, 6H), 0.86 (t, J = 7.1 Hz, 3H). 13C NMR (126 MHz,
CDCl3) δ 153.8, 143.5, 143.2, 138.5, 136.5, 129.8, 127.9, 127.7, 127.6, 127.0, 126.8, 115.7, 84.1,
68.2, 47.1, 34.9, 31.6, 31.2, 30.5, 29.8, 29.4, 27.7, 22.6, 14.1. LRMS (ESI, APCI) m/z: calc’d for
C28H36NO3S [M-H]- 465.3, found 465.4
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
endo 5-hexyl-4-phenyl-3a-(1-phenylvinyl)-1,2,3,3a,6,6a-hexahydropentalen-1-yl sulfamide (6N)
Asolution of 1N (30 mg, 0.08 mmol, 1.1 equiv) in DCM was treated with triethylamine (2.0 equiv.)
and solution of 2-oxo-1,3-oxazolidine-3-sulfonyl chloride (0.5 M in DCM, 1.0 equiv) (prepared
according to the procedure of Borghese et al)44. The reaction was stirred at room temperature for
3 h then concentrated. The residue was treated with ammonia (0.5 M in dioxane, 1.5 equiv) and
triethylamine (3.0 equiv). The solution was heated in a sealed tube at 85°C for 16 h behind a blast
shield. After cooling to ambient temperature, the reaction was diluted with 3:3:94
MeOH:Et3N:EtOAc and passed through a pad of silica. The eluent was concentrated, and the crude
oil was purified on silica in 20-30% EtOAc/hexanes eluent to afford the title compound as a
colorless oil. 6N: 21.6 mg, 60% yield. Purity was established by Method C: tR = 2.0 min, 96.6 %.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 41
1
2
3
4
5
6
7
8
9
1H NMR (600 MHz, CDCl3) δ 7.33 – 7.23 (m, 8H), 7.20 – 7.17 (m, 2H), 5.09 (d, J = 1.3 Hz, 1H),
4.96 (d, J = 1.3 Hz, 1H), 4.44 (s, 2H), 4.36 (d, J = 8.0 Hz, 1H), 3.84 – 3.77 (m, 1H), 2.62 (td, J =
8.9, 2.0 Hz, 1H), 2.38 (dd, J = 17.5, 2.0 Hz, 1H), 2.20 – 2.13 (m, 1H), 2.08 – 2.04 (m, 2H), 2.00 –
1.95 (m, 1H), 1.74 – 1.70 (m, 2H), 1.50 – 1.43 (m, 1H), 1.42 – 1.16 (m, 8H), 0.86 (t, J = 7.1 Hz,
3H). 13C NMR (126 MHz, CDCl3) δ 154.1, 143.6, 142.8, 139.3, 136.6, 129.6, 127.8, 127.7, 126.9,
126.8, 115.5, 68.8, 57.2, 47.4, 35.4, 32.3, 32.0, 31.6, 29.8, 29.5, 27.9, 22.6, 14.1. LRMS (ESI,
APCI) m/z: calc’d for C28H37N2O2S [M+H]+ 465.7, found 464.8
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
Biology: materials and reagents. pCI empty vector was purchased from Promega. The SHP-luc
and Renilla reporters, as well as pCI LRH-1, have been previously described18. The vector for His-
tagged tobacco etch virus (TEV) was a gift from John Tesmer (University of Texas at Austin). The
pMSC7 (LIC-HIS) vector was provided by John Sondek (University of North Carolina at Chapel
Hill). The Tif2 NR Box 3 peptide was purchased from RS Synthesis. DNA oligonucleotide primers
were synthesized by Integrated DNA Technologies.
Protein purification. Purification of human LRH-1 ligand binding domain (residues 300-537) in
a pMCSG7 expression vector was performed as described13. Briefly, protein was expressed in
BL21 PLysS E. coli, using 1 mM IPTG for 4 hours (30°C) to induce expression. Protein was
purified by nickel affinity chromatography. For DSF assays, protein eluted from the nickel column
was exchanged with DLPC (5-fold molar excess overnight at 4 °C), followed by repurification by
size exclusion to remove displaced lipids. The assay buffer was 20 mM Tris-HCl, pH 7.5, 150
mM sodium chloride, and 5% glycerol. Cleaved LRH-1 was then incubated with ligands overnight
at 4 °C prior to repurification by size exclusion, using the same assay buffer as for DSF. Protein
used for crystallography was prepared as for coregulator recruitment, except that it was sized into
ACS Paragon Plus Environment
26

Page 27 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
a buffer of 100 mM ammonium acetate (pH 7.5), 150 mM sodium chloride, 1 mM DTT, 1 mM
EDTA, and 2 mM CHAPS.
7
8
9
Differential scanning fluorimetry (DSF). DSF assays were conducted on a StepOne Plus
thermocycler as previously described13, 15. Briefly, aliquots of purified LRH-1 LBD protein (0.2
mg/ ml) were incubated with saturating concentrations of ligand overnight at 4 °C. Protein-ligand
complexes were heated in the presence of SYPRO orange dye at a rate of 0.5 degree/ minute.
Complexes were excited at 488 nm, and fluorescence emissions at each degree Celsius were
measured using the ROX filter (~600 nm). Tm values were calculated using the Bolzmann
equation in GraphPad Prism, v7.
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
Crystallography. Compounds 5N, 6N, or 2N were incubated with purified LRH-1 LBD (His tag
removed) at 5-fold molar excess overnight at 4°C. The complexes were re-purified by size
exclusion chromatography into the crystallization buffer (see above). Protein was concentrated to
5-6 mg/ ml and combined with a peptide from human Tif2 NR box 3 (H3N-
KENALLRYLLDKDDT-CO2) at four-fold molar excess. Crystals were generated by hanging
drop vapor diffusion at 18 °C, using a crystallant of 0.05 M sodium acetate (pH 4.6), 5-11% PEG
4000, and 0-10% glycerol. Crystals of 2N with LRH-1 were generated by microseeding, using
RJW100-LRH-1 crystals as the seed stocks (crystals used for seeding were grown as described)13.
Structure Determination. Crystals were flash-frozen in liquid nitrogen, using a cryoprotectant
of crystallant plus 30% glycerol. Diffraction data were collected remotely from Argonne National
Laboratory, Southeast Regional Collaborative Access Team, Beamline 22ID. Data were processed
and scaled using HKL200045. Structures were phased by molecular replacement using Phenix46,
with PBD 5L11 used as the search model. The structure was refined using phenix.refine46 and
Coot47, with some additional refinement done using the PDB Redo web server48.
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 41
1
2
3
4
5
6
7
8
9
Tissue culture. Hela cells were purchased from Atlantic Type Culture Collection and cultured in
phenol red-free MEMα media supplemented with 10% charcoal-dextran-stripped fetal bovine
serum. Cells were maintained under standard culture conditions.
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
Reporter gene assays. Hela cells were reverse-transfected with three vectors: (1) full-length,
human LRH-1 in a pCI vector, (2) a firefly reporter (pGL3 Basic) with a portion of the SHP
promoter cloned upstream of the firefly luciferase gene, and (3) a constitutively active vector
expressing Renilla luciferase under control of the CMV promoter. To study SF-1 activity, cells
were transfected with the same constructs, except that full-length SF-1 (in a pcDNA3.1 vector)
was overexpressed instead of LRH-1, with empty pcDNA3.1 used as the negative control.
Transfections utilized the Fugene HD transfection reagent at a ratio of 5 μl per 2 μg DNA. To
perform the reverse transfections, cells were trypsinized, combined with the transfection mixture,
and plated at densities of 7,500 cells per well in white-walled 96-well plates. The following day,
cells were treated with each compound (or DMSO control) for 24 hours. In most cases, six points
in the concentration range of 0.03 – 30 μM were used (exceptions noted in figures), with a final
DMSO concentration of 0.3% in all wells. Luciferase expression was measured using the DualGlo
Kit (Promega). Firefly luciferase signal was normalized to Renilla luciferase signal in each well.
EC50 values were calculated using three-parameter curve-fitting (GraphPad Prism, v.7). Assays
were conducted in triplicate with at least two independent biological replicates. Experiments
involving SF-1 activation were conducted in an identical manner, except full-length human SF-1
(in a pcDNA3.1+ vector) was overexpressed instead of LRH-1. Significance of differences in
luminescence signal for LRH-1 versus SF-1 promoted by particular agonists was determined using
two-way ANOVA followed by Sidak’s multiple comparisons test.
ACS Paragon Plus Environment
28

Page 29 of 41
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Calculation of Relative Efficacy (RE). This value was calculated from curve-fitting to data from
luciferase reporter assays. To compare the maximum activities of the new compounds to RJW100,
we used the formula (Maxcpd – Mincpd) / (MaxRJW100 – MinRJW100), where “Max” and “Min” denote
the dose response curve maximum and minimum, respectively. A RE of 0 indicates a completely
inactive compound, a value of 1 indicates equal activity to RJW100, and values above 1 indicate
greater activity.
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
Mutagenesis. Mutations were introduced to LRH-1 in the pCI vector using the Quikchange
Lightning site-directed mutagenesis kit (Ambion). Constructs were sequenced prior to use in
reporter gene assays as described above.
Model Construction for Molecular Dynamics Simulations. Four LRH-1 LBD complexes were
prepared for molecular dynamics simulations. 1) LRH-1-Tif2-RJW100 (PDB 5L11), 2) LRH-1-
Tif2-5N. 3LRH-1-Tif2-2N, LRH-1-Tif2-6N. For consistency, all structures contained LRH-1
residues 300–540. Missing residues (i.e., that could not be modeled in the structures) were added
to the models used in the simulations.
Molecular Dynamics Simulations. The complexes were solvated in an octahedral box of TIP3P
water with a 10-Å buffer around the protein complex. Na+ and Cl- ions were added to neutralize
the protein and achieve physiological buffer conditions. All systems were set up using the xleap
tool in AmberTools1749 with the ff14SB forcefield50. Parameters for the agonist ligands 6N, 2N
and 5N were obtained using Antechamber51 also in AmberTools17. All minimizations and
simulations were performed with Amber1649. Systems were minimized with 5000 steps of steepest
decent followed by 5000 steps of conjugate gradient minimization with 500-kcal/mol∙Å2 restraints
on all solute atoms. Restraints were removed excluding the atoms in both the ligand and the Tif2
peptide, and the previous minimization was repeated. This minimization was repeated with
ACS Paragon Plus Environment
29