Structure–Activity Relationship Studies of Vitamin D3 Analogues Containing an Ether or Thioether Linker as Hedgehog Pathway Inhibitors

Article abstract of DOI:10.1002/cmdc.201700792

The Hedgehog (Hh) signaling pathway is critical for embryonic patterning and postembryonic tissue regeneration. Constitutive pathway activation has also been linked to human malignancies such as basal cell carcinoma (BCC) and medulloblastoma; therefore, multiple small-molecule scaffolds that inhibit Hh signaling are in development. Previously, Grundmann's alcohol, also known as the “northern region” of vitamin D3 (VD3), has been identified as a moderate Hh pathway inhibitor. In this study, isomers of Grundmann's alcohol with different orientations of the C4 hydroxy group and C3α proton were investigated to determine the optimal configuration for this hexahydroindane scaffold with respect to Hh inhibition. A series of analogues containing Grundmann's alcohol linked to a substituted phenyl or benzyl ring through an ether or thioether linker were synthesized and evaluated for their anti-Hh activity. Of these, analogue 17 ((1R,3aR,4R,7aR)-1-[(R)-1,5-dimethylhexyl]-4-(4-aminophenoxy)-7a-methyloctahydro-1H-indene) demonstrated potent anti-Hh activity in Hh-dependent BCC cells and did not activate canonical vitamin D receptor signaling, demonstrating its selective nature for the Hh signaling pathway.

Full text of DOI:10.1002/cmdc.201700792

DOI: 10.1002/cmdc.201700792
Full Papers
Structure–Activity Relationship Studies of Vitamin D3
Analogues Containing an Ether or Thioether Linker as
Hedgehog Pathway Inhibitors
Jiachen Wen and M. Kyle Hadden*[a]
The Hedgehog (Hh) signaling pathway is critical for embryonic
patterning and postembryonic tissue regeneration. Constitu-
tive pathway activation has also been linked to human malig-
nancies such as basal cell carcinoma (BCC) and medulloblasto-
ma; therefore, multiple small-molecule scaffolds that inhibit Hh
signaling are in development. Previously, Grundmann’s alcohol,
also known as the “northern region” of vitamin D3 (VD3), has
been identified as a moderate Hh pathway inhibitor. In this
study, isomers of Grundmann’s alcohol with different orienta-
tions of the C4 hydroxy group and C3a proton were investi-
gated to determine the optimal configuration for this hexahy-
droindane scaffold with respect to Hh inhibition. A series of an-
alogues containing Grundmann’s alcohol linked to a substitut-
ed phenyl or benzyl ring through an ether or thioether linker
were synthesized and evaluated for their anti-Hh activity. Of
these, analogue 17 ((1R,3aR,4R,7aR)-1-[(R)-1,5-dimethylhexyl]-4-
(4-aminophenoxy)-7a-methyloctahydro-1H-indene) demonstrat-
ed potent anti-Hh activity in Hh-dependent BCC cells and did
not activate canonical vitamin D receptor signaling, demon-
strating its selective nature for the Hh signaling pathway.
Introduction
The Hedgehog (Hh) signaling pathway is critical for embryonic
patterning and postembryonic tissue regeneration.[1] Uncon-
trolled pathway activation has been linked to human malig-
nancies such as basal cell carcinoma (BCC), gastric carcinoma
and medulloblastoma (MB).[2] Selective inhibition of this path-
way prevents tumor growth, indicating that active Hh signal-
ing is essential for tumor survival in these cancers.[1] FDA ap-
proval of vismodegib and sonidegib as treatments for ad-
vanced BCC further confirmed the clinical efficiency of Hh
pathway inhibitors for the treatment of certain cancers.[3] Sev-
eral recent studies have demonstrated that Hh-dependent can-
cers can develop acquired resistance to these therapies, high-
lighting the continued need to identify improved small mole-
cule inhibitors of Hh signaling.[4–6]
amides.[11–13] Several compounds with improved anti-Hh activity
have been identified through these studies (Figure 1, 3–6). Our
SAR for these analogues clearly demonstrated that a phenyl
ring incorporating a hydrophilic moiety at the meta- or para-
position is necessary for potent Hh inhibition.[12] With respect
to the composition of the tether, a 1- to 3-atom distance be-
tween the northern region and the aromatic A-ring were opti-
mal.[13] To further our understanding of SAR for VD3-mediated
inhibition of Hh signaling, we synthesized and evaluated a
new series of analogues linked through an ether or thioether.
In addition, we performed a more rigorous investigation of the
optimal stereochemistry of the hexahydroindane region of the
scaffold.
The seco-steroid vitamin D3 (VD3, 1) has been recognized as
a moderate Hh pathway inhibitor in recent years.[7–9] Initial Results and Discussion
structure–activity relationship (SAR) studies on this scaffold led
Chemistry
to the identification of Grundmann’s alcohol (GA, 2), also
termed the “northern region” of VD3, as the pharmacophore
responsible for VD3-mediated Hh pathway inhibition.[10] Since
then, several series of VD3 analogues based on this hexahy-
droindane scaffold were synthesized and evaluated for their
Hh inhibitory activity. The majority of these compounds incor-
porate an aromatic A-ring mimic linked to the northern region
through a variety of tethers, including esters, amines, or
Using the standard one-pot, two-step global ozonolysis/reduc-
tion of VD3, we isolated 2 as the major product (68%) along
with a minimal amount of GA isomer 9 (3aR,4R, 22%).[9] In the
presence of sodium metal, Grundmann’s ketone 7 (3aR) was
converted into 8 (3aS) through a keto-enol tautomerization
mechanism (Scheme 1). Reduction of 8 with sodium borohy-
dride provided GA isomers 10 (3aS,4S, 41%) and 11 (3aS,4R,
43%). To the best of our knowledge, this is the first disclosure
of a GA isomer with the 3aR,4S configuration. We used a 2D-
NMR strategy to confirm the stereochemistry at the C3a and
C4 positions, with the most informative correlations shown in
Figure 2. Taking 11 as an example, a strong correlation be-
tween the C7a-CH3 (d=1.06 ppm) and the C4 proton (d=
3.94 ppm) can be observed in the NOESY spectrum. These cor-
[a] Dr. J. Wen, Prof. M. K. Hadden
Department of Pharmaceutical Sciences, University of Connecticut, 69
North Eagleville Road, Unit 3092, Storrs, CT 06029-3092 (USA)
E-mail: kyle.hadden@uconn.edu
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under:
https://doi.org/10.1002/cmdc.201700792.
ChemMedChem 2018, 13, 1 – 7
1
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Figure 1. Structures and Hh inhibitory activity for vitamin D3-based compounds.
Scheme 1. Synthetic route to prepare GA-isomers and their analogues: a) O3, CH2Cl2, 08C!RT, 96%; b) Na, EtOH, 08C!RT, 36%; c) NaBH4, CH2Cl2, MeOH, RT,
22–68%; d) 4-fluronitrobenzene, NaH, DMF, 08C!1208C, 73–92%; e) Pd/C, NH2NH2, EtOH, 608C, 60–90%.
relations are only visible when the C4 proton adopts the same
orientation with C7a-CH3 and is at a proper distance. Once the
C4 proton was assigned, the C3a proton could be determined
as the S-orientation, as evidenced by correlations between the
C4 proton (d=3.94 ppm) and C3a proton (d=2.02 ppm) in
both COSY and NOESY spectra. As an additional verification, a
clear correlation between the C3a proton and the C7a-CH3 was
observed in the NOESY spectrum. Similar correlations were
also observed in corresponding ether analogues. More details
and NMR spectra for each isomer can be found in the Support-
ing Information. To probe the anti-Hh activity of the individual
GA isomers, we prepared a series of ether linked analogues in-
corporating a para-aniline. Taking sodium hydride as base,
each of the GA isomers was readily coupled with 4-fluoronitro-
benzene in excellent yield (73–92%). The resulting 4-nitro-
phenyl ethers (12–15) were reduced to the corresponding ani-
lines 16–19 in the present of palladium on carbon (Scheme 1).
In addition to the para-aniline series, we also sought to ex-
amine a broader range of ether linked analogues, which were
generally based on our previous SAR for this scaffold. Allyl pro-
tection and chlorination of commercially available benzyl alco-
hols 20 or 21 provided the corresponding benzyl chlorides 22
and 23, which were coupled to a GA isomer in the following
step (Scheme 2). Through the Williamson ether synthesis
method, GA isomers 2 and 9 were coupled with the appropri-
ate phenyl or benzyl halide to provide several ether intermedi-
&
ChemMedChem 2018, 13, 1 – 7
www.chemmedchem.org
2
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Figure 2. The structures of GA-isomers and their most informative correla-
tions from 2D-NMR.
ates and final analogues (24–35). Following the formation of
ethers 25 and 29, removal of the allyl group was accomplished
with the treatment of a sodium borohydride/iodine system to
provide analogues 26 and 30, respectively, which were subse-
quently acetylated under standard conditions. Nitrophenyl
ether 32 was reduced to aniline 33 in the present of palladium
on carbon. Structures of each of these ether analogues are
given in Scheme 2.
Generation of 4-hydroxyphenyl ether 36 was unsuccessful in
our standard ether synthesis method described above. Alterna-
tively, we used a diazo-hydrolysis reaction to transform aniline
16 into phenol 36 (Scheme 3). The de-amination product 37
was formed simultaneously as an unexpected, but major, by-
product (Scheme 3).
Scheme 2. Synthetic route for A-ring-modified ethers: a) allyl bromide,
K2CO3, CH3CN, reflux; b) SOCl2, CH2Cl2, RT, 76–84% for two steps; c) NaH,
DMF, 60–1408C, 57–91%; d) NaBH4, I2, THF, RT, 95–98%; e) AcCl, K2CO3,
CH2Cl2, RT, 94–97%; f) Pd/C, NH2NH2, EtOH, 608C, 99%.
We also synthesized para-aniline 42 with a thioether linker
to determine whether modification to the bulkier sulfur atom
affects Hh inhibition of the scaffold (Scheme 4). The hydroxy
group of GA 2 was pre-activated by mesyl chloride, followed
by substitution with deprotonated thioacetate, providing 38
(41%) and 39 (43%) simultaneously. The thioesters were sub-
sequently hydrolyzed into corresponding thiols 40 and 41 in
quantitative yield. The stereochemistry of the C4 proton was
determined by 2D-NMR. As exemplified by 41, a correlation be-
tween the C7a-CH3 (d=0.73 ppm) and the C4 proton (d=
3.64 ppm) can be observed in the NOESY spectrum, indicating
the R-configuration of the thiol. Aniline 42 was prepared by
the coupling of thiol 39 and 4-fluoronitrobenzene in the same
method described earlier.
Biological evaluation
Expression of Gli1 mRNA in the Hh-dependent mouse embry-
onic fibroblast (MEF) cell line C3H10T1/2 is a well-characterized
marker of Hh signaling. Gli1 is up-regulated in this cell line fol-
lowing Hh activation with an exogenous Hh ligand or pathway
agonist, which can be abrogated with the concomitant addi-
tion of an Hh pathway inhibitor.[14] For our studies, we activat-
ed Hh signaling (and up-regulated Gli1 expression) with a 1:1
solution of 20S-hydroxycholesterol and 22S-hydroxycholesterol
(5 mm each, OHCs) following our previously described proto-
col.[9,13] Each analogue was initially evaluated for its anti-Hh ac-
tivity in this cell line. Selected compounds were further pro-
Scheme 3. Synthesis of 4-hydroxyphenyl ether: a) NaNO2, HCl, H2O, ꢀ208C!RT; b) CuSO4, H2O, RT, 19–67% for two steps.
ChemMedChem 2018, 13, 1 – 7
www.chemmedchem.org
3
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
values 0.5–0.9 mm) when compared with the mouse
embryonic fibroblasts (MEFs). Of note, analogues of
itraconazole have demonstrated the same trend in
our hands, but previous series of VD3 analogues
have maintained similar inhibition between the two
distinct cell types.[12,13,15,16] The most potent ana-
logue in this series was 17, which contains the
3aR,4R hexahydroindane scaffold (IC50 =0.5 mm). The
negligible up-regulation of Cyp24A1 mRNA induced
by each of the anilines (1.5–4.6-fold) relative to VD3
(>20000-fold) highlights the selective nature of
these ether-linked analogues. It is clear from these
results that the stereochemistry at the C3a and C4
positions in the VD3 scaffold plays a key role for
potent Hh inhibition when the northern region is
linked to an aromatic A-ring through an ether link-
age.
Scheme 4. Synthetic route to VD3-based thioether: a) methanesulfonyl chloride, pyridine,
RT; b) potassium thioacetate, DMF, reflux, 41–43% for two steps; c) NaOH, EtOH, 608C,
quant.; d) NaH, DMF, 1208C; e) Pd/C, NH2NH2, EtOH, 608C, 59% for two steps.
filed in ASZ001 cells, an Hh-dependent murine BCC cell line,
which does not require the addition of exogenous OHCs to ac-
tivate the Hh pathway. In addition, selectivity for Hh inhibition
compared with canonical VDR activation was determined in
the ASZ001 by monitoring up-regulation of Cyp24A1 mRNA, a
well-characterized target gene of the vitamin D receptor
(VDR).[8]
In parallel with the SAR studies exploring the importance of
the stereochemistry of the hexahydroindane scaffold for Hh in-
hibition, a series of ethers with modified aromatic A-rings was
also synthesized and evaluated (Table 2). Because GA (2,
3aR,4S) is the most abundant isomer prepared from the stan-
dard synthetic route, we used this as our primary northern
region (compounds 24–35); however, the cellular results for
analogue 17 prompted us to prepare two ether-linked com-
pounds with the analogous 3aR,4R configuration (36–37). Not
surprisingly based on our previous SAR for the VD3 scaffold,
the initial screening at 5 mm in C3H10T1/2 cells identified phe-
nols 26 and 30, as well as aniline 33, as promising analogues
with potent Hh inhibitory activity (Gli1 expression: 7–12%). By
comparing 24–27, we demonstrate that a phenol (26, Gli1 ex-
pression=7%) is key to maintaining the potent anti-Hh activi-
ty of 1 or 2. The addition of a methyl (24), allyl (25), or acetyl
(27) group to the phenol resulted in inactive analogues. The
same SAR trends were observed in ethers 28–31, where the
substitution is in the para-position. For a more comprehensive
comparison of phenols 26 (3-OH-benzyl) and 30 (4-OH-benzyl),
we determined their IC50 values in both cell lines. While both
compounds were slightly more active in the C3H10T1/2 (IC50:
1.4–1.6 mm) than the ASZ001 (IC50: 5.4–5.8 mm) cells, there was
no significant difference in anti-Hh activity between 26 and 30,
suggesting the tolerance of para- and meta-phenols in the
benzyl ether scaffold. The same results were apparent for the
meta-aniline (33), which in combination with analogue 16 (4-
The GA isomers and corresponding ether linked para-aniline
analogues were initially evaluated in C3H10T1/2 cells at a
single concentration (5 mm). Compounds demonstrating
potent anti-Hh activity were further evaluated for concentra-
tion-dependent inhibition of Hh signaling in these cells. As
shown in Table 1, all GA isomers maintained modest anti-Hh
activity at 5 mm (Gli1 expression=28–32%) that was similar to
VD3 (Gli1 expression=46%). Unfortunately, when evaluated
for their concentration-dependent effects, all four GA isomers
with different orientations at the C8 and C14 positions demon-
strated almost identical anti-Hh activity (IC50 values=2.2–
2.4 mm). Modest differences in anti-Hh activity were observed
for the corresponding ether-linked para- anilines (16–19); how-
ever, their IC50 values were all in the low micromolar range
(1.1–3.7 mm). The 4-nitro precursors 12–15 were completely in-
active (data not shown), revealing the importance of the hy-
drogen bond donors of the aniline.
We also evaluated anilines 16–19 in ASZ001 cells to deter-
mine their activity in an Hh-dependent cancer cell line. Inter-
estingly, the anilines were all more potent in ASZ001 cells (IC50
Table 1. Anti-Hh activity of the GA isomers and their corresponding ether-linked para-anilines.[a]
Compd
C3H10T1/2
Compd
C3H10T1/2
ASZ001[d]
Gli1 [%][b]
IC50 [mm][c]
Gli1 [%][b]
IC50 [mm][c]
IC50 [mm][c]
CYP24A1[e]
DMSO
2
9
10
11
1.0
NA
1
46.4ꢁ3.5
10.1ꢁ1.8
25.9ꢁ4.2
39.1ꢁ7.2
9.2ꢁ2.0
4.1ꢁ0.3
1.1ꢁ0.1
2.2ꢁ0.8
3.7ꢁ1.1
1.4ꢁ0.2
2.1ꢁ0.1
0.7ꢁ0.4
0.5ꢁ0.1
0.9ꢁ0.1
0.9ꢁ0.1
21496ꢁ4622
2.4ꢁ0.9
28.6ꢁ2.1
28.2ꢁ5.2
32.2ꢁ1.7
30.8ꢁ2.7
2.3ꢁ0.1
2.2ꢁ0.2
2.4ꢁ0.1
2.4ꢁ0.4
16
17
18
19
4.6ꢁ0.9
4.5ꢁ0.3
1.5ꢁ0.4
[a] Experiments were performed in triplicate at least two independent times. [b] Values represent percent Gli1 mRNA expression relative to OHCs control
(set as 100%) at 5 mm. [c] IC50 values represent the meanꢁSEM. [d] In ASZ001 cells, all analogues evaluated following 48 h incubation. [e] VD3 was evaluat-
ed at 2.5 mm, others were evaluated at 5 mm. Values represent fold up-regulation of Cyp24A1 mRNA normalized to DMSO (set to 1.0).
&
ChemMedChem 2018, 13, 1 – 7
www.chemmedchem.org
4
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Select compounds were further profiled for stability (human
liver microsomes) and solubility. As shown in Table 3, ethers
with 3aR,4S configuration (26, 30, 34) were extremely stable in
the experimental conditions, with a half-life outside the assay
Table 2. Biological activity of GA ethers and thioether with modified A-ring
mimics.[a]
Table 3. Liver microsomal stability profile of selected compounds.[a]
Compd
t1/2 HLM [min]
Intrinsic
clearance[b]
PBS solubility [mm]
Compd
R group
C3H10T1/2
ASZ001[d]
IC50 [mm][c]
Gli1 [%][b]
IC50 [mm][c]
pH 4.0
pH 7.4
16
26
30
34
36
>231
>231
>231
>231
<3
<3
<3
<3
1.46
0.70
1
/
46.4ꢁ3.5
102ꢁ14
113ꢁ7
7.0ꢁ1.2
104ꢁ8
102ꢁ4
111ꢁ6
12.2ꢁ3.6
121ꢁ17
87ꢁ9
4.1ꢁ0.3
ND[e]
ND
1.6ꢁ0.2
ND
ND
ND
1.4ꢁ0.1
ND
2.1ꢁ0.1
ND
ND
5.4ꢁ0.8
ND
ND
ND
5.8ꢁ0.6
ND
ND
0.9ꢁ0.2
ND
0.051
0.0076
0.012
0.011
0.038
0.0058
0.010
0.033
24
25
26
27
28
29
30
31
32
33
34
35
36
37
40
41
42
3-OMe-Bn
3-Alloc-Bn
3-OH-Bn
3-OAc-Bn
4-OMe-Bn
4-Alloc-Bn
4-OH-Bn
4-OAc-Bn
3-NO2-C6H4
3-NH2-C6H4
4-OH-C6H4
Ph
86.4ꢁ11.7
8.2ꢁ1.1
[a] All assays performed as a contract service by Pharmaron Inc. [b] Repre-
sented in CLint [mLminꢀ1 (mg protein)ꢀ1].
ND
12.1ꢁ6.8
72ꢁ3
1.3ꢁ0.3
7.5ꢁ0.4
ND
1.6ꢁ0.2
1.2ꢁ0.1
ND
ND
10.8ꢁ2.0
window (>231 min) and an intrinsic clearance rate as low as
3 mLminꢀ1 mgꢀ1. Analogue 36 (t1/2 =86.4 min) with 3aR,4R con-
figuration was less stable than its epimer 26 (t1/2 >231 min).
The stability difference between these two compounds was
unexpected, but does provide us with additional information
to apply to the further development of this scaffold of Hh
pathway inhibitors. By contrast, the half-life of VD3 in a similar
microsomal stability assays was 34 min, demonstrating the in-
creased stability of our analogues.[17] The only analogue evalu-
ated in these assays that incorporates an aniline on the aro-
matic A-ring (16) was also the most soluble, further suggesting
this moiety as key for improving both potency and drug-like
properties of the scaffold.
92ꢁ3
ND
3-OH-Bn
4-OH-Bn
/
1.9ꢁ0.3
3.3ꢁ0.2
94ꢁ5
2.1ꢁ0.2
2.1ꢁ0.1
ND
ND
ND
/
92ꢁ2
4-NH2-C6H4
91ꢁ3
[a] Experiments were performed in triplicate at least two independent times.
[b] Values represent percent Gli1 mRNA expression relative to OHCs control
(set as 100%) at 5 mm. [c] IC50 values represent the meanꢁSEM of all inde-
pendent experiments. [d] In ASZ001 cells, all analogues were evaluated fol-
lowing 48 h incubation. [e] ND: not determined.
aminophenyl), further supports the SAR that para- and meta-
substituents are well tolerated across the phenyl ether class of
VD3 analogues. Replacing the aniline (33, Gli1 expression=
12%) with a nitro (32, Gli1 expression=87%) led to almost
complete loss of anti-Hh activity, once again highlighting the
importance of a hydrogen bond donor as critical in this scaf-
fold. To our surprise, the anti-Hh activity decreased when the
para-aniline (16, IC50 =1.1 mm) was replaced with a para-
phenol (34, IC50 =7.5 mm), as this substituent was very potent
in our previous amine-linked analogues (4 and 5). As expected,
unsubstituted analogue 35 was inactive. Finally, the meta- and
para-phenols with the 3aR,4R orientation 36 (IC50 =1.6 mm) and
37 (IC50 =1.2 mm) were similar in activity to the corresponding
3aR,4S, but less active than the aniline (17).
To determine whether the C4 hydroxy group had any effects
on Hh inhibition, Grundmann’s thiols 40, 41, and thioether 42
were synthesized. Surprisingly, the anti-Hh activity of 40 (Gli1
expression=94%) and 41 (Gli1 expression=92%) were totally
abolished at a single concentration (5 mm). Similar results were
obtained for thioether 42. Even though 42 showed a slight im-
provement over its precursor thiol 40 (56 vs. 74%, respectively;
Gli1 expression at 10 mm), the activity is still 10 times weaker
than its corresponding ether 16, demonstrating that the C4 hy-
droxy group is critical for the anti-Hh activity mediated by GA
and its analogues.
Conclusions
In our ongoing analysis of the Hh inhibitory activity of the vita-
min D3 scaffold, we synthesized and evaluated two new ana-
logue series: 1) GA isomers and 2) ether or thioether linked ar-
omatic A-ring analogues. There was no difference in activity
for the GA isomers that have different orientations of the C3a
proton and C4 hydroxy group and minimal differences in anti-
Hh activity when an aromatic A-ring was appended. Interest-
ingly, ether-linked analogues containing a para-aniline substi-
tuted A-ring were more potent in the Hh-dependent murine
BCC cells (ASZ001) than in the Hh-dependent MEFs (C3H10T1/
2). Analogue 17, which contains the 3aR,4R-hexahydroindane
scaffold represented the most potent compound in these two
series with an IC50 value of 0.5 mm in ASZ001 cells. Analogue
17 did not induce the expression of Cyp24A1 mRNA, revealing
its selective nature for the Hh signaling pathway. Preliminary in
vitro PK evaluation of select compounds suggest that the
ether-linked analogues are very stable and demonstrate that
an aniline on the aromatic A-ring can provide enhance solubili-
ty. Overall, these results further our understanding of SAR relat-
ed to VD3 inhibition of the Hh pathway and provide useful in-
formation for the development of next generation VD3-based
analogues.
ChemMedChem 2018, 13, 1 – 7
www.chemmedchem.org
5
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Experimental Section
Acknowledgements
General chemistry
The authors gratefully acknowledge support of this work by the
University of Connecticut Research Foundation and the American
Cancer Society (RSG-13-131-01). ASZ001 cells were provided by
Dr. Ervin Epstein (Children’s Hospital Oakland Research Institute).
VD3 used for chemical synthesis was purchased from HBC Chem,
Inc. Other chemicals were purchased from Sigma–Aldrich or Fisher
Scientific. Reactions were monitored by thin layer chromatography
(G/UV254). Column chromatography silica gel was purchased from
Sorbent Technologies. NMR data was performed on a Bruker AV-
500 MHz spectrometer and analysis with MestReNova software.
HRMS data was analyzed at the Mass Spectrometry Facility at the
University of Connecticut. All yields reported are non-optimized
one-time yields. Please refer to the Supporting Information for de-
tailed synthesis procedures and spectral information.
Conflict of interest
The authors declare no conflict of interest.
Keywords: Grundmann’s alcohol
·
Hedgehog pathway
inhibitors
·
structure–activity relationships thioethers ·
·
vitamin D3
Biological assay protocols
[1] A. P. McMahon, P. W. Ingham, C. J. Tabin, Curr. Top. Dev. Biol. 2003, 53,
1–114.
[2] M. P. di Magliano, M. Hebrok, Nat. Rev. Cancer 2003, 3, 903–911.
[3] Summaries of the information that led to the FDA approval of vismode-
gib and sonidegib can be found at: https://www.cancer.gov/about-
General protocols for cell culture, qPCR, and Hh inhibition in
C3H10T1/2 and ASZ001 cells are as previously described.[9–13,15,16]
Data were analyzed with GraphPad Prism 5 and reported values
represent the meanꢁSEM for at least two separate experiments
carried out in triplicate.
cancer/treatment/drugs/vismodegib
and
https://www.cancer.gov/
about-cancer/treatment/drugs/sonidegib.
[4] R. L. Yauch, G. J. Dijkgraaf, B. Alicke, T. Januario, C. P. Ahn, T. Holcomb, K.
Pujara, J. Stinson, C. A. Callahan, T. Tang, J. F. Bazan, Science 2009, 326,
572–574.
[5] A. L. S. Chang, A. E. Oro, Arch. Dermatol. 2012, 148, 1324–1325.
[6] J. Y. Tang, J. M. Mackay-Wiggan, M. Aszterbaum, R. L. Yauch, J. Lindgren,
K. Chang, C. Coppola, A. M. Chanana, J. Marji, D. R. Bickers, E. H. Ep-
stein, Jr., N. Engl. J. Med. 2012, 366, 2180–2188.
[7] L. W. Brꢁggemann, K. C. Queiroz, K. Zamani, A. V. Straaten, C. A. Spek,
M. F. Bijlsma, Cancer Biol. Ther. 2010, 10, 79–88.
[8] J. Y. Tang, T. Z. Xiao, Y. Oda, K. S. Chang, E. Shpall, A. Wu, P. L. So, J.
Hebert, D. Bikle, E. H. Epstein, Cancer Prev. Res. 2011, 4, 744–751.
[9] U. Banerjee, M. Ghosh, M. K. Hadden, Bioorg. Med. Chem. Lett. 2012, 22,
1330–1334.
[10] A. M. DeBerardinis, S. Lemieux, M. K. Hadden, Bioorg. Med. Chem. Lett.
2013, 23, 5367–5370.
[11] A. M. DeBerardinis, U. Banerjee, M. K. Hadden, ACS Med. Chem. Lett.
2013, 4, 590–595.
[12] A. M. DeBerardinis, D. J. Madden, U. Banerjee, V. Sail, D. S. Raccuia, D. De
Carlo, S. M. Lemieux, A. Meares, M. K. Hadden, J. Med. Chem. 2014, 57,
3724–3736.
PBS solubility and metabolic stability
The PBS solubility assay and human liver microsomal stability assay
were performed as a contract service by Pharmaron Inc., and fol-
lowed their standard procedures. Briefly, for the solubility assay,
excess amount of the test compound was placed in PBS (pH 4.0 or
7.4). The suspension was appropriately vortexed and kept in 378C.
At appropriate intervals, sample was transferred to a filter inserted
tube. After centrifuge, the filtrate was diluted with acetonitrile and
then analyzed by HPLC. For the microsomal stability assay, com-
pound (1 mm), human liver microsomes (1 mgmLꢀ1), and NADPH
(1 mm) were incubated in phosphate buffer at 378C. Aliquots were
taken at a variety of time points and quenched with acetonitrile.
Each sample was centrifuged and analyzed via LC–MS/MS on a
Waters XEVO TQ-D equipped with Acquity UPLC. Peak areas were
determined from extracted ion chromatograms and compared
with the standard curve to determine the slope value, k.
[13] A. M. DeBerardinis, D. S. Raccuia, E. N. Thompson, C. A. Maschinot, M. K.
Hadden, Euro. J. Med. Chem. 2015, 93, 156–171.
[14] J. K. Chen, J. Taipale, M. K. Cooper, P. A. Beachy, Genes Dev. 2002, 16,
2743–2748.
[15] J. R. Pace, A. M. DeBerardinis, V. Sail, S. K. Tacheva-Grigorova, K. A. Chan,
R. Tran, D. S. Raccuia, R. J. Wechsler-Reya, M. K. Hadden, J. Med. Chem.
2016, 59, 3635–3649.
The in vitro half-life (t1/2) was determined from the slope value by
Equation (1):
t1=2 ¼ ꢀð0:693=kÞ
ð1Þ
[16] C. A. Maschinot, M. K. Hadden, Bioorg. Med. Chem. Lett. 2017, 27, 4011–
4014.
Conversion of t1/2 into in vitro intrinsic clearance (CLint,
mLminꢀ1 mgꢀ1) was performed with Equation (2):
[17] J. Chen, J. Wang, T.-K. Kim, E. W. Tieu, E. K. Y. Tang, Z. Lin, D. Kovacic,
D. D. Miller, A. Postlewaite, R. C. Tuckey, A. T. Slomisnki, W. Li, Anticancer
Res. 2014, 34, 2153–2164.
CLint
Manuscript received: December 15, 2017
¼ ½0:693=t1=2ꢂ½volume of incubation ðmLÞ=amount of protein ðmgÞꢂ
ð2Þ
Accepted manuscript online: February 6, 2018
Version of record online: && &&, 0000
&
ChemMedChem 2018, 13, 1 – 7
www.chemmedchem.org
6
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
J. Wen, M. K. Hadden*
&& – &&
Structure–Activity Relationship
Studies of Vitamin D3 Analogues
Containing an Ether or Thioether
Linker as Hedgehog Pathway
Inhibitors
Northern explorations: In our ongoing
study of the anti-Hedgehog (Hh) activity
of the vitamin D3 scaffold, a new series
of analogues focused on structure–ac-
tivity relationships for the central hexa-
hydroindane backbone (C3a and C4 po-
sitions) were synthesized and evaluated.
Of these, analogue 17 showed potent
anti-Hh activity in Hh-dependent basal
cell carcinoma cells and did not activate
canonical vitamin D receptor signaling,
demonstrating its selective nature for
the Hh signaling pathway.
ChemMedChem 2018, 13, 1 – 7
www.chemmedchem.org
7
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ