Structure-activity relationship studies of Niemann-Pick type C1-like 1 (NPC1L1) ligands identified by screening assay monitoring pharmacological chaperone effect

Article abstract of DOI:10.1016/j.bmc.2013.06.022

A number of hereditary diseases are caused by defective protein trafficking due to a folding defect resulting from point mutations in proteins. Ligands that bind to the folding intermediates of such mutant proteins and rescue their trafficking defects, known as pharmacological chaperones, have promise for the treatment of certain genetic diseases, including Fabry disease, cystic fibrosis, and Niemann-Pick disease type C. Here we show that this pharmacological chaperone effect can be used for ligand screening, that is, binding of candidate ligands can be detected by monitoring the ligand-mediated correction of a localization defect caused by artificially introduced point mutations of the protein of interest. Using this method, we discovered novel steroidal ligands of Niemann-Pick type C1-like 1 (NPC1L1), an intestinal cholesterol transporter that is the target of the cholesterol absorption inhibitor ezetimibe, and conducted structure-activity relationship studies. We also present data indicating that the binding site of the new ligands is distinct from both the N-terminal sterol-binding domain and the ezetimibe-binding site.

Full text of DOI:10.1016/j.bmc.2013.06.022

Bioorganic & Medicinal Chemistry 21 (2013) 5297–5309
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Structure–activity relationship studies of Niemann-Pick type C1-like
1
(NPC1L1) ligands identified by screening assay monitoring
pharmacological chaperone effect
a
a,
⇑
b
a
Fumika Karaki , Kenji Ohgane , Kosuke Dodo , Yuichi Hashimoto
a
Institute of Molecular and Cellular Biosciences, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-0032, Japan
RIKEN, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
A number of hereditary diseases are caused by defective protein trafficking due to a folding defect result-
ing from point mutations in proteins. Ligands that bind to the folding intermediates of such mutant pro-
teins and rescue their trafficking defects, known as pharmacological chaperones, have promise for the
treatment of certain genetic diseases, including Fabry disease, cystic fibrosis, and Niemann-Pick disease
type C. Here we show that this pharmacological chaperone effect can be used for ligand screening, that is,
binding of candidate ligands can be detected by monitoring the ligand-mediated correction of a localiza-
tion defect caused by artificially introduced point mutations of the protein of interest. Using this method,
we discovered novel steroidal ligands of Niemann-Pick type C1-like 1 (NPC1L1), an intestinal cholesterol
transporter that is the target of the cholesterol absorption inhibitor ezetimibe, and conducted structure–
activity relationship studies. We also present data indicating that the binding site of the new ligands is
distinct from both the N-terminal sterol-binding domain and the ezetimibe-binding site.
Ó 2013 Elsevier Ltd. All rights reserved.
Received 1 May 2013
Revised 7 June 2013
Accepted 8 June 2013
Available online 17 June 2013
Keywords:
Niemann-Pick type C1-like 1
Niemann-Pick disease type C
Cholesterol
Transporter
Protein trafficking
1
. Introduction
molecular mechanism of loss of function has been proposed to in-
volve a folding and localization defect, rather than an intrinsic loss
2
3
NPC1L1 is a 13-pass transmembrane protein expressed on the
of function. In such cases, an effective therapeutic strategy might
be to use pharmacological chaperones, that is, small-molecular li-
gands that correct the trafficking defect resulting from reduced
folding efficiency, presumably by binding to the folding intermedi-
ate in the endoplasmic reticulum (ER) and promoting its escape
apical membrane of enterocytes and hepatocytes, and plays a crit-
ical role in the intestinal uptake of dietary cholesterol.1 NPC1L1
has three large extracellular loops, and the first N-terminal loop,
which is generally designated as the N-terminal domain (NTD),
has been demonstrated to bind cholesterol and several oxyster-
–6
2
4–28
from the ER quality control system.
Recently, we reported that
7
–10
ols.
Zetia),
Extracellular loop 2 is important for binding of ezetimibe
oxysterols and their derivatives can be used as pharmacological
1
1–16
I1061T
(
a
widely used cholesterol absorption inhibitor.
chaperones to correct the localization defect of NPC1
mutant
2
9
Although much effort has been directed to discovery of novel
NPC1L1 inhibitors, all the compounds reported so far have been
discovered by means of competitive binding assay using fluores-
cent or radioisotope-labeled ezetimibe analogues, and all of them
are considered to bind at the same site as ezetimibe.1
protein. We also demonstrated the existence of a second, non-
NTD sterol-binding site in NPC1 and suggested that this sterol-
binding site might be functionally important.
Based on these findings, we hypothesized that NPC1L1 also has
a non-NTD sterol-binding site and ligands that bind at this site
might be novel NPC1L1 inhibitors. To test this idea, we developed
a screening system based on the pharmacological chaperone effect
as a tool for detecting ligand binding to NPC1L1 protein. We artifi-
cially generated folding-defective NPC1L1 mutants, and by moni-
toring ligand-mediated correction of the localization defect of the
mutant proteins, we were able to discover novel steroidal ligands
for NPC1L1. Here, we report the identification and structure–activ-
ity relationships of these steroidal NPC1L1 ligands, and provide
evidence that the binding site of this class of NPC1L1 ligands is dis-
tinct from both the NTD and the ezetimibe-binding site.
7–21
NPC1 (Niemann-Pick type C1) is a close homologue of NPC1L1
(
42% identity, 51% similarity), and is known to be required for
22
intracellular transport of cholesterol. Loss of function of NPC1
resulting from mutations in the protein is a major cause of Nie-
mann-Pick disease type C, a fatal progressive neurodegenerative
disease. In the case of the most prevalent mutation, I1061T, the
⇑
Corresponding author. Tel.: +81 3 5841 7847; fax: +81 3 5841 8495.
E-mail address: ohgane@me.com (K. Ohgane).
0
968-0896/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2013.06.022

5
298
F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
2
2
. Results and discussion
NPC1L1-GFP showed normal plasma membrane localization (data
not shown), we obtained other mutants (including L1072S and
L1072T/L1168I) that localize predominantly in the ER, a character-
istic localization for folding-defective membrane proteins (Fig. 1a).
To exclude terminally misfolded mutants and to select folding-
defective mutants amenable to ligand-mediated defect correction,
we examined the temperature sensitivity of the ER-retained mu-
tants, because it has been demonstrated that the localization defect
of folding-defective mutants can be corrected if they are expressed
.1. Cell-based ligand screening utilizing folding-defective
NPC1L1 mutants
In our previous study,29 we reported that oxysterols directly
bind to the second, non-NTD sterol-binding site of NPC1, and cor-
rect the trafficking defect of folding-defective NPC1I1061T mutant.
Here, our strategy to identify NPC1L1 ligands directed to the puta-
tive second sterol-binding site of NPC1L1 was to screen sterols by
monitoring ligand-mediated rescue of the localization defect of
folding-defective NPC1L1 mutants. Based on the high homology
between NPC1 and NPC1L1, we hypothesized that the second ste-
rol-binding site of NPC1L1 would accept a subset of sterols, and we
envisioned that introduction of mutation(s) at the position corre-
sponding to the I1061T mutation of NPC1 would yield the required
folding-defective NPC1L1 mutant.
3
0,31
at reduced temperature.
As shown in Figure 1b, the localization
of L1072T/L1168I mutant NPC1L1 changed from ER to plasma
membrane at reduced temperature, indicating the suitability of
this mutant NPC1L1 for our ligand screening assay.
2.2. Identification of oxysterols as lead compounds for further
optimization
First, to obtain the folding-defective NPC1L1 mutant, we intro-
duced mutations around L1072 of NPC1L1, which corresponds to
I1061 of NPC1. Although the L1072T and L1072M mutants of
Next, we tested a variety of sterols for the ability to correct the
localization defect of the L1072T/L1168I mutant, that is, for the
ability to act as pharmacological chaperones. Among the oxysterols
Figure 1. Correction of subcellular localization of NPC1L1L1072T/L1168I mutant. (a) Colocalization of L1072T/L1168I (LT/LI) mutant with a ER marker, calnexin. (b) Subcellular
localization of WT or LT/LI mutant NPC1L1-GFP. The ER retention of the LT/LI mutant was corrected to plasma membrane localization when the mutant was expressed at
reduced temperature. (c) Subcellular localization of WT or LT/LI mutant NPC1L1-GFP, and the effect of the steroidal carbamate 11 on the localization of LT/LI mutant. HEK293
cells stably expressing the indicated construct were incubated with 11 or vehicle for 24 h. The cells were stained with CellMask Orange plasma membrane stain and the
localization of NPC1L1-GFP was examined by fluorescence microscopy. (d) The extent of colocalization was quantified, and the colocalization coefficient (thresholded
Manders’ coefficient) is shown with standard deviations depicted by error bars (n = 13). (e) Effects of the steroidal carbamate 30 and inactive oxysterol 7 on steady-state
L1072T/L1168I
expression levels of LT/LI. The steady-state expression levels of NPC1L1
-GFP were quantified by measuring GFP fluorescence in the lysate of cells treated with or
without the indicated compounds. The GFP fluorescence was normalized with respect to total protein concentration determined by bicinchoninic acid assay. Data points
represent the averages (n = 3) with the standard error of the means depicted by error bars.

F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
5299
tested, 25-hydroxycholesterol (25HC, 11) corrected the trafficking
defect of the mutant protein (Fig. 1c and d). The extent of NPC1L1
localization was quantified as described in Figure 2 and Experi-
known that sufficient membrane permeability to reach the ER is
a prerequisite for pharmacological chaperones, so the above result
supports the view that 25HC is indeed acting as a pharmacological
2
6,34
mental, and EC50 value of 25HC was calculated to be 2.3
lM
chaperone.
(
Fig. 2b, c, and Table 1). As in the case of NPC1, cholesterol itself
While B-ring monohydroxylated sterols, including 6- or 7-
hydroxylated sterols (2, 3, 5), were weakly active, 19-, 20-, and
22-hydroxylated sterols (7, 8, 9, 10) proved to be completely inac-
tive, implying that the corresponding binding site of NPC1L1 has a
was not active toward NPC1L1, possibly because cholesterol di-
rectly added to cell culture does not reach the ER or is not available
for binding to proteins (unpublished observation).3
2,33
It is well
Figure 2. Quantification of localization change of NPC1L1 mutant. (a) Representative images of cells with ER, near ER, mixed, near PM, and PM localization of NPC1L1-GFP. (b
and c) A representative result of quantification showing the dose-dependent correction of NPC1L1 mislocalization by steroidal carbamate 30 (Table 3). Cells expressing
L1072T/L1168I
NPC1L1
-GFP were treated with steroidal carbamate 30, and images were acquired. The images were quantified as described in the Section 4.
Table 1
L1072T/L1168I
Pharmacological chaperone effect of oxysterols and their derivatives on NPC1L1
mutant
No. Compound
EC50
(
l
M) No. Compound
EC50
9.7
(l
M) No. Compound
EC50
2.3
(lM)
OH
H
H
H
H
H
H
H
H
H
H
H
H
H
1
NA
6
11
H
H
H
H
H
H
H
H
O
H
H
H
H
HO
HO
HO
OH
OH
O
HO
H
2
8.5
9.7
2.8
10
7
NA
NA
NA
NA
12
NA
NA
H
H
HO
HO
H
HO
OH
OH
H
H
H
H
3
8
H
13
H
H
H
H
H
HO
H
HO
HO
OH
OH
H
H
4
9
H
H
H
O
HO
H
H
H
H
H
HO
OH
H
H
5
10
H
H
HO
OH
HO
NA: no activity (3 lM for 1 and 10 lM for the others). The EC50 values were determined by phenotypic analysis as described in Figure 2 and the Section 4.

5
300
F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
hydrophobic character. Notably, 5
a
,6b-dihydroxylated sterol (1)
tive, a hydrogen bond donor at this position seems to be disfa-
vored, and this may in part explain why dimethyl amide (19)
was superior to 25HC.
did not correct the localization defect of the mutant NPC1L1 pro-
tein, further supporting the idea that the introduction of polar
functionalities into the B ring is disfavored. These results are in
marked contrast to the structure–activity relationships obtained
for NPC1; that is, 22(S)-hydroxylated sterol (10) was as active as
2.4. Optimization of substituents at the 3 position
2
2
5HC (11) and 5
5HC (11) as a pharmacological chaperone for NPC1, whereas
a
,6b-dihydroxylated sterol (1) was superior to
In the case of NPC1, introduction of a morpholino amide group
at the 3 position improved the activity by nearly one order of mag-
2
9
these oxysterols are inactive in the case of NPC1L1. The implication
is that compounds selective for NPC1 or NPC1L1 may be obtainable
by appropriately modifying the positions of hydroxyl group(s) and/
or introducing substituents at appropriate positions.
nitude. Therefore, we envisioned that modification at the 3 posi-
tion might result in more potent pharmacological chaperones for
NPC1L1, and we synthesized a series of 3-OH-modified steroidal
carbamates (Scheme 1). As shown in Table 3, N,N-dialkylated car-
bamates (24, 25, 26) showed increased potency compared to N-
non-substituted carbamate (23). As the alkyl chains were elon-
gated, however, the potency decreased and the N,N-dibutyl deriv-
ative (27) was completely inactive. Replacement of the terminal –
2
.3. Structure–activity relationship of the isooctyl side chain
moiety
The isooctyl side chain of sterols is important for the biological
3
CH groups of (26) with –OH (30) increased the potency (Table 3
activity, as demonstrated by the fact that sterol (12), which lacks
the side chain, is inactive. In the light of this observation, we next
synthesized side-chain-modified sterols, and examined the struc-
ture–activity relationship of the side-chain moiety (Scheme 1).
The one-carbon-truncated analogue of 25HC (14), its analogue
without the dimethyl group (16), and the corresponding carboxylic
acid (15) were found to be inactive, supporting the importance of
side-chain structure for activity (Table 2). Among the side-chain
amide derivatives tested, N,N-dimethyl amide (19) showed in-
creased potency compared to 25HC, but larger substitutents on
nitrogen decreased the potency, as exemplified by N,N-dibutyl
amide (22). Combined with the finding that N-monomethyl amide
and Fig. 2), implying a favorable interaction in the binding site of
NPC1L1. On the other hand, polar carbamate derivatives with a
more rigid ring structure, such as morpholine or 1-methylpipera-
zine (28 and 29, respectively), were less potent than bis(2-
hydroxyethyl)carbamate (30), indicating the importance of appro-
priately positioned polar groups. Although modification of the 3-
position resulted in increased potency, the preferred substituent
was different for NPC1L1 and NPC1. Thus, the steroidal carbamate
(30) showed 2.5-fold selectivity for NPC1L1 (EC50 0.65
NPC1 (EC50 1.6 M).
lM) over
l
As the compound with bis(2-hydroxyethyl)carbamate at the 3-
position (30, Table 3) was more potent than the parental non-
substituted sterol (19, Table 2), we next tested whether this type
(18) and N-non-substituted amide (17) were only marginally ac-
O
O
O
O
R
Me
OH
N
Me
N
N
H
a
H
b
H
Me
R'
H
Me
H
H
H
H
O
H
H
H
H
H
H
H
H
R
HO
HO
N
O
HO
15
R'
1
1
1
2
2
2
7 R=R'=H
19
23 R=R'=H
8 R=H, R'=Me
9 R=R'=Me
0 R=R'=Et
2
2
2
2
2
2
3
4 R=R'=Me
5 R=R'=Et
6 R=R'=n-Pr
7 R=R'=n-Bu
1 R=R'=n-Pr
2 R=R'=n-Bu
8 NRR'=morpholyl
9 NRR'=N-methylpiperazyl
0 R=R'=CH2CH2OH
0
Scheme 1. Synthetic route to the side chain-modified sterols and their carbamate derivatives. Reagents and conditions: (a) isobutyl chloroformate, RR NH (for 17, 28%
0
aqueous ammonia solution), Et
3
N, THF, rt; (b) triphosgene, pyridine, RR NH (for 23, 28% aqueous ammonia solution), Et
3
N, CH
2
Cl
2
, rt.
Table 2
Pharmacological chaperone effect of the side-chain-modified sterols on NPC1L1L1072T/L1168I mutant
R
H
H
H
H
HO
No.
R
EC50
>10
(
lM)
No.
R
EC50
>10
(lM)
No.
R
EC50 (lM)
OH
O
O
O
O
O
O
H
Et
14
15
16
17
N
H
20
N
(Toxic)
(Toxic)
>3
Et
O
Me
Me
n-Pr
OH
OH
>10
NA
18
19
N
H
>10
21
22
N
n-Pr
n-Bu
N
0.80
N
Me
n-Bu
EC50s of 20 and 21 could not be determined because of their cytotoxicity.

F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
5301
Table 3
Pharmacological chaperone effect of 3-OH-modified steroidal carbamates on LT/LI and
D
NTD-LT/LI mutants
O
Me
N
H
Me
H
H
H
R
No.
R
EC50
(l
M)
NTD-LT/L1
No.
R
EC50 (lM)
LT/L1
8.5
D
LT/L1
NA
D
NTD-LT/L1
O
O
H
n-Bu
23
24
25
26
N
H
O
>10
4.7
4.2
8.0
27
28
29
30
N
O
NA
n-Bu
O
O
Me
N
O
1.1
1.8
3.0
N
O
O
2.3
6.2
>10
1.7
Me
O
O
Et
N
O
N
O
2.9
Et
MeN
HO
O
O
n-Pr
N
O
N
O
0.65
n-Pr
HO
NA: no activity (10 lM).
H
H
H
a
b
H
H
H
O
O
H
H
H
H
H
H
HO
HO
HO
HO
HO
N
O
N
O
cholesterol
37
HO
31
OH
H
H
H
H
H
H
f
d
O
O
O
H
H
H
H
O
H
H
RO
RO
HO
HO
N
O
N
O
N
O
OH
RO
RO
37 R=H
39 R=TBS
35
c
e
38 R=TBS
36 R=H
Scheme 2. Synthetic route to 31, 35, and 36. Reagents and conditions: (a) triphosgene, pyridine, diethanolamine, THF, CH
2
Cl
2
, rt; (b) m-CPBA, CH
2
Cl
2
, rt then HIO
4
ꢀ2H
2
O, THF,
rt; (c) TBSCl, imidazole, DMF, CH Cl , rt; (d) RuCl , TBHP, 1,2-dichloroethane, rt; (e) TBAF, THF, rt; (f) NaBH , CeCl , MeOH, THF, rt.
2
2
3
4
3
of substitution can also increase the potency of other sterols. The
synthetic routes to oxysterol derivatives with bis(2-hydroxy-
ethyl)carbamate at the 3-position are shown in Schemes 2 and 3.
As shown in Table 4, all the carbamate derivatives shown here
mutant protein, the most active steroidal carbamate (30) dose-
dependently increased the level of the NPC1L1 mutant (Fig. 1e).
These data support our hypothesis that the observed correction
of the localization defect was mediated through binding of the ste-
rol derivatives to the NPC1L1 mutant.
(
compounds 31–37, Table 4) were more potent than their non-
substituted counterparts (compounds 1–6, 13, Table 1). They were,
however, less potent than 30, which is in harmony with the fact
that 19 (Table 2) was more potent than B-ring hydroxylated cho-
lesterols (1, 2, 3, 5) and B-ring keto-cholesterols (4, 7) (Table 1).
2.6. The steroidal carbamates act independently of the N-
terminal sterol-binding domain
To test whether the NTD is required for ligand-mediated rescue
L1072T/L1168I
2
.5. Ligand-mediated stabilization of NPC1L1 mutant
of NPC1L1
tives
NPC1L1
, we examined the effect of the sterol deriva-
L1072T/L1168I
on
NTD-deleted
NPC1L1
-GFP
NTD-NPC1L1
(DNTD-
L1072T/L1168I
L1072T/L1168I
Since we had obtained compounds that could correct the local-
-GFP, Fig. 3a). While
D
L1072T/L1168I
ization of NPC1L1
was indeed mediated through binding of the compounds to the
NPC1L1 mutant. It has been shown that ligands added to cell cul-
, we next examined whether the effect
was localized predominantly in the ER in the absence of the li-
gands, the mutant protein showed plasma membrane localization
in the presence of the ligands (Fig. 3b). Additionally, the deletion
of NTD did not affect the rank-order of EC50 value among the sterol
derivatives (Table 3). These results indicate that the NTD is not re-
quired for the effect of our sterol derivatives. Furthermore, the ste-
ture can stabilize destabilized mutant proteins and increase the
steady-state levels of the mutant proteins.3
5–38
To test whether
the steroidal carbamate (30) could stabilize the NPC1L1 mutant
protein, we examined the steady-state level of the mutant protein
by measuring the GFP fluorescence of the cell lysate. Though the
inactive oxysterol (7) did not affect the steady-state level of the
roidal carbamate (30) also increased the steady-state level of
L1072T/L1168I
D
NTD-NPC1L1
(Fig. 3c). Therefore, it is likely that there
is a second sterol-binding site on NPC1L1, as in NPC1, and that the

5
302
F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
H
H
H
b
e
H
H
H
O
H
H
H
H
H
H
R1O
RO
1
R O
N
O
H
H
cholesterol R=H
0 R=TBS
OR2
R O
1
OR2
4 R1=OH, R2=Ac
45 R =TBS, R =Ac
a
4
1 2
41 R =TBS, R =H
4
c
f
4
2 R1=TBS, R2=Ac
1
2
d
g
h
43 R1=OH, R2=Ac
4
6 R1=TBS, R2=OH
32 R1=R2=OH
H
H
H
i
k
H
H
H
O
O
O
H
H
H
O
H
H
H
TBSO
RO
N
HO
HO
N
O
O
N
O
H
H
H
46
33
TBSO
OH
RO
OH
4
7 R=TBS
j
3
4 R=H
Scheme 3. Synthetic route to 32, 33 and 34. Reagents and conditions: (a) TBSCl, imidazole, DMF, CH
pyridine, rt; (d) TBAF, THF, rt; (e) triphosgene, pyridine, diethanolamine, CH Cl
i)PCC, AcONa, CH Cl , rt; (j) TBAF, THF, 45 °C; (k) NaBH , MeOH, rt.
2
Cl
2
, rt; (b) BH
3
ꢀTHF, THF, rt then NaOH aq and H
2
O
2
aq rt; (c) Ac
2
O, DMAP,
2
2
, rt; (f) TBSCl, imidazole, DMF, CH
2
Cl
2
, rt; (g) K
2
CO
3
, MeOH, CH Cl , rt; (h) TBAF, THF, 45 °C;
2 2
(
2
2
4
Table 4
Pharmacological chaperone effect of bis(2-hydroxyethyl)carbamate derivatives of oxysterols on LT/LI
No.
Compound
EC50
(lM)
No.
Compound
EC50
2.0
(lM)
H
H
H
H
H
H
H
H
H
31
32
33
34
8.1
35
O
O
O
O
O
O
O
H
H
H
H
H
H
H
HO
N
HO
N
O
O
O
O
O
O
O
OH
OH
OH
HO
HO
H
H
H
1.1
3.8
2.0
36
37
3.8
NA
H
H
H
H
O
HO
N
HO
N
H
H
H
HO
OH
HO
H
H
H
HO
N
HO
N
HO
OH
HO
H
H
O
HO
N
HO
steroidal carbamates developed in this study bind to this putative
second sterol-binding site.
steroidal carbamate (30). For other steroidal carbamate deriva-
tives, any detectable level of competition by ezetimibe was not ob-
served either (data not shown). These results clearly indicate that
the binding site of our steroidal carbamate is different from that
of ezetimibe.
2
.7. The binding site of the steroidal carbamates is distinct from
that of ezetimibe
The NPC1L1 inhibitor ezetimibe is reported to bind to the
extracellular loop 2 of NPC1L1.14 So, we next examined whether
ezetimibe accesses the same binding site as our steroidal carba-
mates. As shown in Figure 4, ezetimibe did not correct the mislo-
calization of the NPC1L1 mutant, nor did it compete with the
3. Discussion
In this study, we employed ligand-mediated correction of the
localization defect of folding-defective mutant protein as a screen-
ing tool to identify novel NPC1L1 ligands. This is the first time that

F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
5303
Figure 3. The steroidal carbamate acts independently of the NTD. (a) Schematic representation of NTD-deleted NPC1L1L1072T/L1168I mutant (
D
NTD-LT/LI). (b) Effects of 30 on
full-length or NTD-deleted LT/LI mutant NPC1L1-GFP. HEK293 cells stably expressing the indicated construct were incubated with 3
lM 30 or vehicle for 24 h. The samples
were fixed and the localization of NPC1L1-GFP was examined by fluorescence microscopy. (c) Effects of 30 and 7 on steady-state expression levels of
D
NTD-LT/LI.
Figure 4. (a) Effect of ezetimibe on the defective localization of LT/LI mutant. Ezetimibe did not correct the trafficking defect of the mutant NPC1L1. (b) Dose-dependent
correction of LT/LI localization defect by the steroidal carbamate 30, with (dashed line) or without (solid line) 10 M ezetimibe.
l
the pharmacological chaperone effect has been utilized for cell-
based ligand screening, and this strategy offers several advantages.
First, this assay obviates the need to purify the target protein,
which is notoriously difficult for multi-pass membrane proteins.
Second, this method does not require the use of a highly specific
radioisotope- or fluorescent-labeled ligand. Third, this assay can
be performed for proteins with unknown function or for which
no functional assay suitable for ligand screening is available, since
it is only necessary to know the localization of the protein of inter-
est. Thus, this assay can be complementary to conventional ligand-
binding assays or functional screening assays.
4. Experimental
4.1. Biology
4.1.1. Materials
Commercially available oxysterols were from Sigma and Santa
Cruz Biotechnology; Lipofectamine LTX and CellMask Orange plas-
ma membrane stain were from Invitrogen; poly-
D-lysine was from
Millipore.
4.1.2. Cell culture and transfection
Based on our previous study on NPC1, we hypothesized the
presence of a second sterol-binding site distinct from the known
N-terminal sterol-binding domain on NPC1L1. In this report, we
present several lines of evidence in support of this idea, indicating
that NPC1L1 contains a second sterol-binding site with a different
sterol specificity from that in the case of NPC1. We further show
that our steroidal carbamates are likely to exert their effect
through binding to this site. Thus, these compounds represent a
novel class of NPC1L1 inhibitor, since their binding site on NPC1L1
is different from that of ezetimibe. Studies to examine the inhibi-
tory activity of these compounds on NPC1L1-mediated cholesterol
absorption are in progress.
HEK293 cells were cultured in DMEM supplemented with 5%
FBS and penicillin and streptomycin at 37 °C in a humidified incu-
bator (5% CO ). Cells were transfected with Lipofectamine LTX
2
(Invitrogen). To obtain cell lines stably expressing NPC1L1 con-
struct, G418-resistant colonies were isolated and maintained in
the presence of G418 (0.4 mg/mL).
4.1.3. DNA constructs
The cDNA of human NPC1L1 (GenBank BC117178) in
pCR-XL-TOPO vector was obtained from Mammalian Gene
Collection (IMAGE id 40125729). The coding sequence of human
NPC1L1 was amplified from this construct, and subcloned into

5
304
F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
the pCMV6-AC-turboGFP vector (from Origene) at the SgfI and
NotI sites, and the linker and turboGFP were replaced with a
FLAG tag-containing linker (DYKDDDDKGGRDPPVAT) and EGFP.
Site-directed mutagenesis was carried out using a KOD—Plus—
Mutagenesis Kit (from Toyobo).
on a JEOL JMA-HX110 mass spectrometer. Steroids and reagents
were purchased from Aldrich, TCI, Wako Pure Chemical Industry,
and Kanto Kagaku and used without purification. Open column
and flash column chromatography were performed using silica
gel 60 (particle size 0.060–0.210 mm) supplied by Kanto Kagaku.
3
9
4
.1.4. CellMask Orange plasma membrane stain
At day 1, cells were seeded onto poly- -lysine-coated glass-bot-
tomed dishes, and treated as indicated at day 2. At day 3, CellMask
Orange plasma membrane stain (0.25 g/mL, pre-diluted with
4.2.2. General procedure for synthesis of amide derivatives of
5
D
3b-hydroxy-
D
-cholenic acid
To a solution of 3b-hydroxy-D5-cholenic acid (0.1 mmol) and
N (20 L, 0.144 mmol) in THF (8 mL) was added isobutyl chloro-
formate (20 L, 0.154 mmol) at 0 °C. The mixture was stirred at
0 °C for 10 min, and then amine (200 L) was added at 0 °C. The
l
Et
3
l
DMEM supplemented with 1% FBS) was added to the dishes and
incubation was continued at 37 °C for 5 min. Then the cells were
fixed (4% paraformaldehyde-PBS, 37 °C, 10 min) and washed three
times with PBS. Images were immediately acquired, and the extent
of colocalization was quantified as described previously.29
l
l
whole was stirred at ambient temperature for 2 h to overnight,
and then diluted with AcOEt. The organic layer was washed with
saturated aqueous solution of NH Cl and brine, dried (Na SO ), fil-
4 2 4
tered and evaporated. The desired compounds were purified as
indicated below.
4
.1.5. Immunocytochemical staining of calnexin
Cells were fixed (3.7% formaldehyde in PBS, rt, 30 min), perme-
abilized (0.1% Triton X-100 in PBS, rt, 10 min), and blocked in 1%
BSA-PBST for 1 h. The samples were incubated with goat anti-caln-
exin (C-20) antibody (1:100 in blocking buffer, from Santa Cruz) for
4.2.2.1. 3b-Hydroxy-D5-cholenamide (17).
chromatography (n-hexane/AcOEt = 1:3 to 1:7) gave 17 (39.2 mg,
Open column
0.105 mmol, 93% yield) as colorless plates (recrystallised from CH2-
1
2
h, and then washed three times with PBS. The samples were incu-
Cl
2
/n-hexane/MeOH, mp 252.5–255.0 °C). H NMR (CDCl
3 3
–CD OD)
bated with anti-goat IgG AlexaFluor 546 (1:250 in blocking buffer)
for 1 h, then washed with PBS, and images were acquired.
d: 5.36–5.33 (1H, m), 3.51–3.41 (1H, m), 3.37–3.32 (2H, m), 1.02
(3H, s), 0.97 (3H, d, J = 6.71 Hz), 0.70 (3H, s). C NMR (CDCl –CD3-
3
1
3
OD) d: 182.16, 144.77, 125.30, 75.10, 60.64, 59.73, 54.04, 46.24,
4
.1.6. NPC1L1L1072T/L1168I localization assay
45.67, 43.64, 41.15, 40.36, 39.45, 36.45, 35.71 (2C), 34.96, 31.93,
At day 1, HEK293 cells stably expressing NPC1L1L1072T/L11168I
13
6
28.07, 24.91, 23.06, 22.00, 15.55. C NMR (DMSO-d ) d: 174.66,
under control of the CMV promoter were plated at a density of
6,000 cells/well in DMEM supplemented with 1% FBS onto a
poly- -lysine-coated glass-bottomed plate. At day 2, cells were
141.24, 120.39, 69.98, 56.20, 55.37, 49.59, 42.21, 41.84, 36.91,
36.05, 34.92, 31.97, 31.44, 31.40, 31.34 (2C), 27.62, 23.86, 20.61,
1
+
+
D
19.13, 18.28, 11.69. FAB-MS m/z: 374 (MH ), 356 ([MꢂOH] ).
+
treated with the test compounds pre-diluted with DMEM (final
DMSO concentration was 0.1%). After 24 h, cells were fixed (3.7%
formaldehyde-PBS, rt, 30 min) and washed with PBS. Images were
obtained using an IX70 inverted fluorescence microscope (Olym-
pus) with a ꢁ100 objective. For quantitative assessment of NPC1L1
localization to the plasma membrane, 25 images were acquired
and analyzed. Cells with ER localization, near-ER localization,
mixed localization, near-PM localization, and PM localization were
counted as 0, 0.25, 0.5, 0.75, and 1, respectively, and the average
value was presented as a percentage (% plasma membrane localiza-
tion). Images of typical cells and a representative dose–response
curve obtained from the quantification procedure are shown in
Figure 2.
HRMS (FAB, [M+H] ) calcd for
374.3053.
24 2
C H40NO 374.3054, found
4.2.2.2. N-Methyl-3b-hydroxy-D5-cholenamide (18).
Open
column chromatography (n-hexane/AcOEt = 1:4 to 1:5) gave 18
(32.1 mg, 0.0828 mmol, 81% yield) as a colorless powder (recrystal-
1
lised from CH
CD
2.77 (3H, s), 1.00 (3H, s), 0.93 (3H, d, J = 6.71 Hz), 0.68 (3H, s).
NMR (CDCl –CD OD) d: 175.21, 141.09, 121.83, 71.73, 56.98,
2
Cl
2
/n-hexane, mp 218.4–220.2 °C). H NMR (CDCl
3
–
3
OD) d: 6.05 (1H, br s), 5.36–5.32 (1H, m), 3.54–3.45 (1H, m),
13
C
3
3
56.07, 50.35, 42.64, 42.29, 40.01, 37.51, 36.74, 35.82, 33.70, 32.13
(3C), 31.61, 28.39, 26.40, 24.50, 21.31, 19.63, 18.61, 12.10. FAB-
+
+
+
MS m/z: 410 (MNa ), 388 (MH ), 370 ([MꢂOH] ). Anal. Calcd for
ꢀ1/2H O: C, 75.71; H, 10.67; N, 3.53. Found: C, 75.45;
H, 10.44; N, 3.20.
C
25
H
41NO
2
2
4
.1.7. Quantification of GFP fluorescence in lysates
For evaluation of steady-state NPC1L1-GFP expression levels,
GFP fluorescence of lysates was measured and normalized by total
protein concentration. In a 12-well plate, cells were cultured to
4.2.2.3. N,N-(Dimethyl)-3b-hydroxy-D5-cholenamide (19).
This reaction was performed on 0.4 mmol scale. Open column
chromatography (n-hexane/AcOEt = 1:3 to 1:4) gave 19
7
0% confluency in medium without phenol red and treated with
1
vehicle or test compounds. After 24 h, the medium was removed
and the cells were lysed in TNET buffer (1% Triton X-100, 25 mM
Tris, 150 mM NaCl, 5 mM EDTA, pH 7.5) for 1 h on ice. The debris
was removed by centrifugation (13,500 rpm at 4 °C for 5 min),
and the fluorescence of the supernatant was measured with a Wal-
lac 1420 multilabel counter (PerkinElmer Life Science). The total
protein concentrations were determined by BCA assay, and fluores-
cence intensities were normalized.
(141.1 mg, 0.351 mmol, 84% yield) as a colorless solid. H NMR
(CDCl
3
) d: 5.29–5.27 (1H, m), 3.49–3.42 (1H, m), 2.91 (6H, br s),
1
3
0.94 (3H, s), 0.88 (3H, d, J = 6.71 Hz), 0.62 (3H, s). C NMR (CDCl
3
)
d: 173.82, 140.76, 121.61, 71.71, 56.70, 55.89, 50.07, 42.35, 42.26,
39.74, 38.69, 37.24, 36.47, 35.65, 35.60, 31.86, 31.85, 31.61,
31.23, 30.33, 28.16, 24.26, 21.05, 19.37, 18.51, 11.86. FAB-MS m/
+
z: 402 (MH ).
4
.2.2.4. N,N-(Diethyl)-3b-hydroxy-D5-cholenamide (20).
4
4
.2. Chemistry
Flash column chromatography (n-hexane/ethyl acetate 2:1 to 1:1)
gave 20 (36.7 mg, 0.085 mmol, 88%) as colorless plates (recrystal-
1
.2.1. General
lised from CH
2
Cl
2
/n-hexane, mp 171.7–172.9 °C). H NMR (CDCl
3
)
1
Proton and carbon nuclear magnetic resonance spectra ( H and
C NMR) were recorded on a JEOL JNM-ECA 500 (500 MHz) spec-
d: 5.35 (1H, br s), 3.56–3.49 (1H, m), 3.42–3.25 (4H, m), 2.36–
2.16 (4H, m), 1.18 (3H, t, J = 7.0 Hz), 1.10 (3H, t, J = 7.0 Hz), 1.01
1
3
1
3
trometer in the indicated solvent. Chemical shifts (d) are reported
in parts per million relative to the internal standard, tetramethyl-
silane. Fast atom bombardment (FAB) mass spectra were recorded
3
(3H, s), 0.95 (3H, d, J = 6.7 Hz), 0.68 (3H, s). C NMR (CDCl ) d:
173.59, 141.63, 122.53, 72.63, 57.60, 56.86, 50.98, 43.25, 43.16,
42.84, 40.88, 40.64, 38.13, 37.36, 36.57, 32.75 (2C), 32.52, 32.48,

F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
5305
3
1.02, 29.06, 25.16, 21.95, 20.27, 19.43, 15.32, 13.98, 12.75. FAB-
4.2.3.3. N,N-(Dimethyl)-D5-cholenamid-3b-yl N,N-diethylcarba-
mate (25). Open column chromatography (n-hexane/
AcOEt = 1:1) gave 25 (13.3 mg, 0.0266 mmol, 29% yield) as color-
less cubes (recrystallised from CH Cl /n-hexane, mp 145.1–
) d: 5.40–5.35 (1H, m), 4.56–4.47 (1H,
m), 3.26 (4H, br s), 3.01 (3H, s), 2.94 (3H, s), 1.11 (6H, t,
J = 7.02 Hz), 1.02 (3H, s), 0.95 (3H, d, J = 6.71 Hz), 0.68 (3H, s). 13
NMR (CDCl ) d: 173.64, 155.53, 140.00, 122.22, 74.22, 56.62,
+
+
MS (glycerol-mNBA 2:1 + NaCl) m/z: 430 (MH ), 452 (MNa ). Anal.
Calcd for C28 : C, 78.27; H, 11.03; N, 3.26. Found: C, 77.99; H,
1
H47NO
2
1.05; N, 3.26.
2
2
1
1
47.8 °C). H NMR (CDCl
3
4
.2.2.5. N,N-(Di-n-propyl)-3b-hydroxy-D5-cholenamide (21).
Open column chromatography (n-hexane/AcOEt = 3:1 to 2:1)
gave 21 (41.0 mg, 0.0896 mmol, 82% yield) as colorless plates
C
3
1
(
recrystallised from CH
2
Cl
2
/n-hexane, mp 117.1–118.5 °C).
H
55.84, 49.94, 42.35, 41.13 (2C, br s), 39.69, 38.69, 37.31, 37.03,
36.55, 35.58, 35.38, 31.85 (2C), 31.20, 30.30 (2C), 30.05, 28.29,
NMR (CDCl
3
) d: 5.33–5.30 (1H, m), 3.55–3.47 (1H, m), 3.32–3.10
1
3
+
(
(
4
3
1
4H, m), 0.99 (3H, s), 0.94 (3H, d, J = 6.9 Hz), 0.67 (3H, s). C NMR
CDCl ) d: 173.61, 140.87, 121.68, 71.78, 56.79, 56.06, 50.17,
9.83, 47.68, 42.45, 42.34, 39.83, 37.34, 36.55, 35.79, 31.94 (2C),
1.83, 31.69, 30.16, 28.26, 24.33, 22.38, 21.15, 20.99, 19.46, 18.61,
1.94, 11.38 (2C). FAB-MS m/z: 480 (MNa ), 458 (MH ). Anal.
: C, 78.72; H, 11.23; N, 3.06. Found: C, 78.50;
28.13, 24.26, 21.01, 19.36, 18.52, 11.85. FAB-MS m/z: 523 (MNa ),
+
3
501 (MH ). Anal. Calcd for C31
H
52
N
2
5
O : C, 74.35; H, 10.47; N,
5.59. Found: C, 74.09; H, 10.18; N, 5.49.
+
+
4.2.3.4. N,N-(Dimethyl)-D5-cholenamid-3b-yl N,N-di-n-propylc-
Calcd for C30
H
51NO
2
arbamate (26). Open column chromatography (n-hexane/
AcOEt = 1:2) gave 26 (30.0 mg, 0.0567 mmol, 53% yield) as color-
less cubes (recrystallised from CH Cl /n-hexane, mp 81.4–
) d: 5.37–5.32 (1H, m), 4.52–4.44 (1H,
m), 3.14 (4H, s), 2.96 (6H, s), 1.00 (3H, s), 0.93 (3H, d,
H, 11.03; N, 3.22.
2
2
.2.2.6. N,N-(Di-n-butyl)-3b-hydroxy-D5-cholenamide (22).
Flash column chromatography (n-hexane/ethyl acetate = 3:1 to
:1) gave 22 (35.8 mg, 0.074 mmol, 85%) as a colorless powder
83.2 °C). H NMR (CDCl
3
1
4
1
3
2
3
J = 6.87 Hz), 0.86 (6H, t, J = 7.45 Hz), 0.66 (3H, s). C NMR (CDCl )
1
(
recrystallised from CH
2
Cl
2
/n-hexane, mp 112.1–114.0 °C).
H
d: 173.67, 156.01, 140.02, 122.25, 74.30, 56.64, 55.85, 49.95,
49.03, 48.50, 42.36, 39.71, 38.68, 37.34, 37.05, 36.58, 35.61,
35.41, 31.88 (2C), 31.22, 30.33, 28.30, 28.16, 24.28, 21.87, 21.38,
NMR (CDCl
3
) d: 5.32 (1H, d, J = 5.5 Hz), 3.53–3.46 (1H, m), 3.32–
3
.17 (4H, m), 0.98 (3H, s), 0.94–0.87 (total 9H, m), 0.66 (3H, s).
1
3
+
C NMR (CDCl
3
) d: 173.07, 140.52, 121.43, 71.52, 56.49, 55.74,
21.04, 19.40, 18.54, 11.88, 11.30 (2C). FAB-MS m/z: 551 (MNa ),
+
4
3
2
9.86, 47.59, 45.46, 42.14, 42.05, 39.52, 37.02, 36.25, 35.47,
1.64, 31.50, 31.40, 31.05, 29.87, 29.68, 27.96, 24.04, 20.84,
529 (MH ). Anal. Calcd for C33
H
56
N
2
O
3
ꢀ1/2 H
2
O: C, 73.70; H,
10.68; N, 5.21. Found: C, 73.57; H, 10.95; N, 5.22.
0.02, 19.90, 19.16, 18.33, 13.64, 11.64. FAB-MS (glycerol-
+
4.2.3.5. N,N-(Dimethyl)-D5-cholenamid-3b-yl N,N-di-n-butylcar-
bamate (27). Open column chromatography (n-hexane/
AcOEt = 1:1) gave 27 (36.5 mg, 0.0656 mmol, 71% yield) as a color-
mNBA 2:1) m/z: 486 (MH ). Anal. Calcd for C32
H, 11.41; N, 2.88. Found: C, 79.13; H, 11.43; N, 2.97.
H
55NO
2
: C, 79.12:
1
4
.2.3. General procedure for synthesis of carbamate derivatives
3
less oil. H NMR (CDCl ) d: 5.39–5.34 (1H, m), 4.54–4.45 (1H, m),
of compound 19
3.25–3.13 (4H, m), 3.01 (3H, s), 2.94 (3H, s), 1.02 (3H, s), 0.95
To a solution of 19 (0.1 mmol) and pyridine (80
lL, 0.991 mmol)
(3H, d, J = 6.71 Hz), 0.92 (6H, t, J = 7.32 Hz), 0.68 (3H, s). 13C NMR
in CH
°C. The mixture was stirred for 20 min at an ambient tempera-
ture. Then amine (200–500 L) was added and stirring was contin-
ued at ambient temperature overnight. The whole was diluted
with CH Cl . The organic layer was washed with water, saturated
2
Cl
2
(8 mL) was added triphosgene (33.3 mg, 0.112 mmol) at
3
(CDCl ) d: 173.61, 155.88, 139.96, 122.19, 74.26, 56.59, 55.81,
0
49.91, 46.98, 46.31, 42.31, 39.66, 38.63, 37.29, 37.00, 36.53,
35.55, 35.35, 31.83, 31.17, 30.70, 30.27 (2C), 28.24, 28.11, 24.23,
20.98, 20.01(2C), 19.33, 18.49, 13.84, 11.84. FAB-MS m/z: 579
l
+
+
+
2
2
(MNa ), 557 (MH ). HRMS (FAB, [M+Na] ) calcd for C35
579.4496, found 579.4449.
60 2 3
H N NaO
4 2 4
aqueous solution of NH Cl and brine, dried (Na SO ), filtered and
evaporated.
4
.2.3.6. N,N-(Dimethyl)-D5-cholenamid-3b-yl morpholinecarba-
4
.2.3.1. N,N-(Dimethyl)-D5-cholenamid-3b-yl carbamate (23).
Open column chromatography (n-hexane/AcOEt = 1:2) gave 23
25.6 mg, 0.0576 mmol, 62% yield) as colorless cubes (recrystal-
mate (28).
Open column chromatography (n-hexane/
AcOEt = 1:2 to 1:3) gave 28 (34.6 mg, 0.0672 mmol, 74% yield) as
colorless cubes (recrystallised from CH Cl /n-hexane, mp 175.8–
3
177.2 °C). H NMR (CDCl ) d: 5.40–5.35 (1H, m), 4.58–4.49 (1H,
(
2
2
1
1
lised from CH
2
Cl
2
/n-hexane, mp 199.8–202.5 °C). H NMR (CDCl
3
)
d: 5.41–5.36 (1H, m), 4.60 (2H, br s), 4.53–4.45 (1H, m), 3.01 (3H,
s), 2.94 (3H, s), 1.01 (3H, s), 0.95 (3H, d, J = 6.10 Hz), 0.68 (3H, s).
m), 3.68–3.62 (4H, m), 3.48–3.43 (4H, m), 3.01 (3H, s), 2.94 (3H,
s), 1.02 (3H, s), 0.95 (3H, d, J = 6.71 Hz), 0.68 (3H, s). 13C NMR
1
3
C NMR (CDCl
3
) d: 173.66, 156.40, 139.61, 122.56, 74.66, 56.60,
3
(CDCl ) d: 173.61, 155.01, 139.70, 122.48, 75.01, 66.58 (2C),
5
3
1
5.84, 49.95, 42.34, 39.67, 38.35, 37.31, 36.93, 36.51, 35.58, 35.38,
1.83 (2C), 31.20, 30.31, 28.12, 27.98, 24.25, 21.00, 19.29, 18.51,
56.58, 55.82, 49.90, 43.93 (2C, br s), 42.31, 39.65, 38.55, 37.29,
36.94, 36.51, 35.56, 35.35, 31.82 (2C), 31.17, 30.28, 28.14, 28.10,
+
+
+
+
1.85. FAB-MS m/z: 467 (MNa ), 445 (MH ). HRMS (FAB, [M+H] )
445.3425, found 445.3383.
24.23, 20.98, 19.30, 18.49, 11.83. FAB-MS m/z: 537 (MNa ), 515
+
calcd for C27
H
45
N
2
O
3
(MH ). Anal. Calcd for C31
50
H N
2
O
4
: C, 72.33; H, 9.79; N, 5.44.
Found: C, 72.07; H, 9.77; N, 5.34.
4
.2.3.2. N,N-(Dimethyl)-D5-cholenamid-3b-yl N,N-dimethylcar-
bamate (24). Open column chromatography (n-hexane/
AcOEt = 1:2) gave 24 (24.2 mg, 0.0512 mmol, 48% yield) as color-
less cubes (recrystallised from CH Cl /n-hexane, mp 170.9–
) d: 5.38–5.34 (1H, m), 4.57–4.44 (1H,
m), 2.97 (6H, s), 2.88 (6H, s), 1.01 (3H, s), 0.94 (3H, d,
4.2.3.7. N,N-(Dimethyl)-D5-cholenamid-3b-yl 4-methylpiperazi-
necarbamate (29). Open column chromatography (CHCl
MeOH = 100/0 to 100/3) gave 29 (30.0 mg, 0.0568 mmol, 60% yield)
as as a colorless powder (recrystallised from CH Cl /n-hexane, mp
) d: 5.39–5.36 (1H, m), 4.55–4.47
3
/
2
2
1
1
73.1 °C). H NMR (CDCl
3
2
2
1
158.9–160.0 °C). H NMR (CDCl
3
13
J = 6.87 Hz), 0.67 (3H, s). C NMR (CDCl
3
) d: 173.67, 156.30,
(1H, m), 3.51–3.46 (4H, m), 3.01 (3H, s), 2.94 (3H, s), 2.37–2.35
1
3
3
39.98, 122.33, 74.67, 56.64, 55.85, 49.96, 42.37, 39.71, 38.73,
7.34, 37.03, 36.57, 36.26, 35.85, 35.61, 35.40, 31.88 (2C), 31.22,
(4H, m), 2.30 (3H, s), 1.02 (3H, s), 0.95 (3H, d, J = 6.71 Hz), 0.68
(3H, s). C NMR (CDCl ) d: 173.62, 154.97, 139.82, 122.39, 74.81,
3
56.59, 55.82, 54.72 (2C), 49.91, 46.16, 43.49 (2C, br s), 42.33,
39.66, 38.59, 37.30, 36.97, 36.53, 35.57, 35.36, 31.83 (2C), 31.18,
30.29, 28.17, 28.12, 24.24, 20.99, 19.32, 18.51, 11.84. FAB-MS m/
1
3
0.33, 28.30, 28.16, 24.28, 21.03, 19.37, 18.54, 11.88. FAB-MS m/
+
+
z: 495 (MNa ), 473 (MH ). Anal. Calcd for C29
H, 10.23; N, 5.93. Found: C, 73.98; H, 10.32; N, 5.89.
43 2 3
H N O : C, 73.68;

5
306
F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
+
+
z: 550 (MNa ), 528 (MH ). Anal. Calcd for C32
7
H
53
N
3
O
3
ꢀ1/3 H
2
O: C,
51.62, 45.04, 42.62, 39.84, 39.39, 38.12, 37.02, 36.06, 35.72,
2.00; H, 10.13; N, 7.87. Found: C, 72.22; H, 9.75; N, 7.98.
34.29, 32.07, 30.19, 28.12, 27.88, 27.02, 24.04, 23.79, 22.66,
+
2
2.40, 20.98, 18.53, 16.40, 11.98. FAB-MS m/z: 574 (MNa ).
4
.2.3.8.
N,N-(Dimethyl)-D5-cholenamid-3b-yl
Open column chromatogra-
/MeOH = 20/1) gave 30 (20.5 mg, 0.0385 mmol, 49%
yield) as a colorless powder (recrystallised from CH Cl /n-hexane,
) d: 5.40–5.35 (1H, m), 4.57–
.49 (1H, m), 3.88–3.75 (4H, m), 3.57 (2H, br s), 3.52–3.42 (4H,
m), 3.01 (3H, s), 2.93 (3H, s), 1.02 (3H, s), 0.95 (3H, d,
N,N-bis(2-
hydroxyethyl)carbamate (30).
phy (CHCl
4.2.4.3. Cholest-5-en-3b-yl N,N-bis[2-(tert-butyldimethylsilyl-
oxy)ethyl]carbamate (38). To a solution of 37 (395.3 mg,
0.764 mmol) in CH Cl (4.5 mL) and DMF (4.5 mL) was added imid-
3
2
2
2
2
1
mp 149.1–151.2 °C). H NMR (CDCl
4
3
azole (423.4 mg, 6.22 mmol) and TBSCl (441.1 mg, 2.93 mmol) at
0 °C. The mixture was stirred for 16 h at ambient temperature
and then diluted with AcOEt. The organic layer was washed (water,
brine), dried (Na SO ), and concentrated. The residue was purified
2 4
by open column chromatography (n-hexane/AcOEt = 100/0 to 95/
13
J = 6.71 Hz), 0.68 (3H, s). C NMR (CDCl
3
) d: 173.77, 156.73,
39.62, 122.59, 75.22, 61.93 (2C), 56.60, 55.86, 52.45, 51.94,
9.93, 42.35, 39.68, 38.59, 37.35, 36.97, 36.54, 35.61, 35.42, 31.83
1
4
5) to afford the title compound (560.9 mg, 0.752 mmol, 98%) as a
1
(
2C), 31.21, 30.35, 28.23, 28.13, 24.26, 21.02, 19.34, 18.52, 11.85.
colorless oil. H NMR (CDCl
3
) d: 5.34–5.30 (1H, m), 4.52–4.43
+
FAB-MS m/z: 555 (MNa ). Anal. Calcd for C31
H, 9.84; N, 5.26. Found: C, 69.59; H, 9.97; N, 5.09.
H
52
N
2
O
5
: C, 69.89;
(1H, m), 3.68 (2H, t, J = 5.49 Hz), 3.63 (2H, t, J = 6.41 Hz), 3.39–
3.32 (4H, m), 0.97 (3H, s), 0.86 (3H, d, J = 6.71 Hz), 0.84 (9H, s),
0
.83 (9H, s), 0.81 (3H, d, J = 6.71 Hz), 0.81 (3H, d, J = 6.71 Hz),
1
3
4
.2.4. Synthesis of bis(2-hydroxyethyl)carbamate derivatives of
oxysterols
.2.4.1. Cholest-5-en-3b-yl N,N-bis(2-hydroxyethyl)carbamate
0.62 (3H, s), 0.00 (6H, s), ꢂ0.01 (6H, s). C NMR (CDCl
3
) d:
155.65, 139.84, 122.38, 74.49, 61.52 (2C), 56.66, 56.12, 51.14,
50.73, 50.00, 42.29, 39.73, 39.50, 38.67, 37.02, 36.55, 36.17,
35.78, 31.87 (2C), 28.27, 28.21, 27.98, 25.88 (6C), 24.27, 23.82,
22.79, 22.54, 21.02, 19.31, 18.70, 18.21, 14.09, 11.82, ꢂ5.38 (2C),
ꢂ5.44 (2C).
4
(
CH
37).
To a solution of cholesterol (779.2 mg, 2.02 mmol) in
(6 mL) and pyridine (20 mL) was added triphosgene
2
Cl
2
(
507.8 mg, 1.71 mmol) at 0 °C. The mixture was stirred for
2
0 min at 0 °C. In
a separate flask, diethanolamine (4 mL,
4
1.5 mmol) was dissolved in THF (60 mL). The former solution
4.2.4.4. 7-Oxocholest-5-en-3b-yl N,N-bis[2-(tert-butyldimethyl-
was added dropwise to the latter solution at 0 °C, and the mixture
was stirred overnight at ambient temperature. The reaction was
quenched by adding water and the whole was extracted with CH2-
silyloxy)ethyl]carbamate (39).
formed following a reported procedure. To a stirred mixture of 38
(447.8 mg, 0.600 mmol) in 1,2-dichloroethane (3 mL) and RuCl3-
The allylic oxidation was per-
4
1
Cl
chloride, brine), dried (Na
was purified by open column chromatography (hexane/
AcOEt = 2:1 then CHCl /MeOH = 94/6) to afford the title compound
795.8 mg, 1.54 mmol, 76%) as colorless cubes (recrystallised from
2
. The organic layer was washed (water, aqueous ammonium
ꢀnH
(800
ambient temperature. The reaction was quenched by adding ace-
tone (3 mL) and saturated aqueous Na (3 mL). The whole
was stirred for 2 h at 60 °C and then diluted with AcOEt. The organ-
ic layer was washed (water, brine), dried (Na SO ), and concen-
2
O (2.2 mg) in water (200
lL) was added dropwise TBHP
2
SO ), and concentrated. The residue
4
lL), in a water bath. The mixture was stirred overnight at
3
2 2 3
S O
(
1
2
hexane/CH Cl
2
, mp 161.3–164.1 °C). H NMR (CDCl
3
) d: 5.40–
2
4
5
(
.35(1H, m), 4.56–4.46 (1H, m), 4.45–4.16 (2H, br m), 3.83–3.73
4H, m), 3.51–3.39 (4H, m), 1.02 (3H, s), 0.92 (3H, d, J = 6.71 Hz),
trated. The residue was purified by open column chromatography
(n-hexane/AcOEt = 10/0 to 9/1) to afford the title compound
1
3
1
0
.87 (3H, d, J = 6.71 Hz), 0.86 (3H, d, J = 6.71 Hz), 0.68 (3H, s).
) d: 156.57, 139.58, 122.56, 75.22, 61.68, 61.54,
6.61, 56.12, 52.47, 51.99, 49.93, 42.25, 39.67, 39.44, 38.55,
C
(151.1 mg, 0.199 mmol, 33%) as a colorless oil. H NMR (CDCl
3
) d:
NMR (CDCl
3
5.68–5.64 (1H, m), 4.65–4.55 (1H, m), 3.69 (2H, t, J = 5.49 Hz),
3.64 (2H, t, J = 6.41 Hz), 3.37 (4H, t, J = 5.80 Hz), 1.16 (3H, s), 0.88
(3H, d, J = 6.71 Hz), 0.85 (9H, s), 0.84 (9H, s), 0.82 (3H, d,
5
3
2
6.93, 36.50, 36.13, 35.75, 31.84, 31.80, 28.17 (2C), 27.93, 24.22,
3.82, 22.75, 22.51, 20.99, 19.30, 18.66, 11.79. FAB-MS m/z: 540
J = 6.71 Hz), 0.82 (3H, d, J = 6.71 Hz), 0.64 (3H, s), 0.01 (6H, s),
+
13
(
MNa ). Anal. Calcd for C32
H55NO
4
: C, 74.23; H, 10.71; N, 2.71.
3
0.00 (6H, s). C NMR (CDCl ) d: 201.88, 164.16, 155.28, 126.58,
Found: C, 74.01; H, 10.71; N, 2.68.
72.80, 61.43, 61.39, 54.74, 51.08, 50.65, 49.95, 49.77, 45.36,
3.09, 39.43, 38.64, 38.28, 38.25, 36.14, 36.00, 35.67, 28.50,
4
4
.2.4.2.
ethyl)carbamate (31).
according to reported procedures.29,40 To
5
a
,6b-Dihydroxy-cholestan-3b-yl N,N-bis(2-hydroxy-
The title compound was prepared
solution of 37
(6 mL) was added mCPBA
186.0 mg, 0.753 mmol, ca. 70%). The mixture was stirred over-
27.94, 27.86, 26.27, 25.58 (6C), 23.79, 22.76, 22.51, 21.14, 18.83,
18.22, 18.17, 17.23, 11.92, -5.40 (2C), ꢂ5.45 (2C).
a
(
146.0 mg, 0.282 mmol) in CH
2
Cl
2
4.2.4.5. 7-Oxocholest-5-en-3b-yl N,N-bis(2-hydroxyethyl)carba-
(
mate (36).
(1 M in THF, 360 l
A solution of 39 (151.1 mg, 0.199 mmol) and TBAF
L, 0.36 mmol) in THF (4 mL) was stirred for 15 h
night at ambient temperature and then diluted with ethyl acetate.
The organic layer was washed with saturated aqueous solution of
at ambient temperature, and then diluted with AcOEt. The organic
3 4
NaHCO and brine, dried (MgSO ) and concentrated. The crude
layer was washed (water, aqueous ammonium chloride, brine),
epoxide was subjected to the next reaction without further purifi-
cation. To a solution of the epoxide in THF (8 mL) was added HIO4-
dried (Na
open column chromatography (n-hexane/AcOEt = 1:1 then CHCl
MeOH = 9/1) to afford the title compound (58.1 mg, 0.109 mmol,
55%) as colorless plates (recrystallised from n-hexane/CH Cl , mp
) d: 5.71–5.69 (1H, m), 4.68–4.59
2 4
SO ), and concentrated. The residue was purified by
3
/
ꢀ
2
2H O (130.0 mg, 0.5727 mmol) at 0 °C. The solution was stirred at
ambient temperature for 2 h and then diluted with ethyl acetate.
2
2
1
The organic layer was washed with saturated aqueous solution of
144.5–146.8 °C). H NMR (CDCl
3
NaHCO
was purified by flash column chromatography (AcOEt/CH
MeOH = 10/10/1 to 10/10/2) to afford the title compound
3
and brine, dried (MgSO
4
) and concentrated. The residue
(1H, m), 4.16 (2H, br s), 3.86–3.74 (4H, m), 3.52–3.42 (4H, m),
2
Cl
2
/
1.21 (3H, s), 0.92 (3H, d, J = 6.10 Hz), 0.87 (3H, d, J = 6.71 Hz),
1
3
3
0.86 (3H, d, J = 6.71 Hz), 0.68 (3H, s). C NMR (CDCl ) d: 202.12,
1
(
27.0 mg, 0.049 mmol, 17%) as a colorless powder. H NMR (CDCl
3
/
164.22, 156.15, 126.51, 73.40, 61.56, 61.41, 54.75, 52.41, 51.88,
49.89, 49.69, 45.35, 43.05, 39.39, 38.58, 38.26, 38.18, 36.11,
35.92, 35.66, 28.47, 27.92, 27.76, 36.23, 23.81, 22.74, 22.49,
CD OD) d: 5.04–4.97 (1H, m), 3.75 (4H, br s), 3.55–3.33 (5H, m),
3
2
1
0
1
.10 (1H, dd, J = 12.20, 12.20 Hz), 1.98 (1H, dt, J = 12.41, 2.59 Hz),
+
.17 (3H, s), 0.90 (3H, d, J = 6.71 Hz), 0.87 (3H, d, J = 6.71 Hz),
21.12, 18.80, 17.22, 11.89. FAB-MS m/z: 554 (MNa ). Anal. Calcd
.86 (3H, d, J = 6.71 Hz), 0.68 (3H, s). 13C NMR (CDCl
/CD
OD) d:
3
3
5
for C32H53NO : C, 72.28; H, 10.05; N, 2.63. Found: C, 71.98; H,
9.99; N, 2.61.
57.01, 75.80, 75.12, 72.90, 61.07, 60.96, 56.16, 55.76, 52.17,

F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
5307
4
.2.4.6.
7b-Hydroxycholest-5-en-3b-yl,
N,N-bis(2-hydroxy-
(3H, s). 13C NMR (CDCl
3
) d: 170.71, 72.45, 70.62, 56.08, 56.05,
ethyl)carbamate (35).
The title compound was prepared
53.57, 48.57, 42.46, 39.63, 39.33, 37.53, 37.06, 36.42, 35.99,
35.59, 33.97, 31.99, 30.90, 28.00, 27.82, 23.94, 23.71, 22.65,
22.41, 21.07, 20.98, 18.52, 13.19, 11.86. FAB-MS m/z: 469 (MNa ).
according to a reported reaction condition.42,43 To a mixture of
+
3
6 (36.0 mg, 0.0677 mmol) and CeCl
in THF (2 mL) and MeOH (1 mL) was added NaBH
.426 mmol) at 0 °C. The mixture was stirred at ambient tempera-
3
ꢀ7H
2
O (76.2 mg, 0.205 mmol)
4
(16.1 mg,
0
4.2.4.9.
ethyl)carbamate (44).
0.809 mmol) in CH Cl (25 mL) and pyridine (5 mL) was added tri-
6a
-Acetoxy-5a-cholestan-3b-yl
N,N-bis(2-hydroxy-
ture for 14 h and then diluted with AcOEt. The organic layer was
washed (water, aqueous ammonium chloride, brine), dried (Na2-
To solution of 43 (361.5 mg,
a
2
2
SO
chromatography (CHCl
compound (30.7 mg, 0.0575 mmol, 85%) as a colorless powder
4
), and concentrated. The residue was purified by open column
phosgene (258.5 mg, 0.872 mmol) at 0 °C. The mixture was stirred
for 20 min at 0 °C. In a separate flask, diethanolamine (2 mL,
20.7 mmol) was dissolved in CH Cl (25 mL). The former solution
2 2
was added dropwise to the latter solution at 0 °C, and the mixture
was stirred overnight at ambient temperature. The reaction was
quenched by adding water and the whole was extracted with CH2-
3
/MeOH = 10/0 to 9/1) to afford the title
1
(
recrystallised from n-hexane/CH
2
Cl
2
, mp 147.8–150.2 °C).
H
NMR (CDCl ) d: 5.31 (1H, s), 4.59–4.50 (1H, m), 3.85–3.72 (1H,
3
m), 3.84–3.76 (4H, m), 3.53–3.40 (4H, m), 1.06 (3H, s), 0.92 (3H,
d, J = 6.10 Hz), 0.87 (3H, d, J = 6.71 Hz), 0.86 (3H, d, J = 6.71 Hz),
Cl
2
. The organic layer was washed (water, aqueous ammonium
SO ), and concentrated. The residue
1
3
0
7
3
2
.69 (3H, s). C NMR (CDCl
3.17, 61.74 (2C), 55.89, 55.46, 52.41, 51.91, 48.15, 42.90, 40.68,
9.48 (2C), 38.08, 36.66, 36.48, 36.18, 35.72, 28.52, 28.18, 27.99,
3
) d: 156.62, 142.31, 126.34, 74.78,
chloride, brine), dried (Na
was purified by open column chromatography (n-hexane/
AcOEt = 1:1 to 1:5) to afford the title compound (259.5 mg,
0.449 mmol, 55%) as colorless crystals. H NMR (CDCl
4.64 (1H, m), 4.62–4.54 (1H, m), 4.10 (2H, br s), 3.86–3.72 (4H,
m), 3.57–3.33 (4H, m), 2.03 (3H, s), 0.90 (3H, d, J = 6.71 Hz), 0.88
2
4
1
6.35, 23.85, 22.79, 22.54, 21.03, 19.11, 18.76, 11.80. FAB-MS m/
3
) d: 4.71–
+
z: 556 (MNa ). Anal. Calcd for C32
5 2
H55NO ꢀ1/2 H O: C, 70.81; H,
1
0.40; N, 2.58. Found: C, 70.66; H, 10.32; N, 2.51.
(
3H, s), 0.86 (3H, d, J = 6.10 Hz), 0.86 (3H, d, J = 7.32 Hz), 0.65
1
3
4
.2.4.7.
6a
-Hydroxycholesterol tert-butyldimethylsilyl ether
The 6 -hydroxylation was conducted by employing a
reported reaction condition.43 To
solution of cholesterol
1180.0 mg, 3.052 mmol) in CH Cl (15 mL) and DMF (15 mL)
was added imidazole (803.1 mg, 11.80 mmol) and TBSCl
896.6 mg, 5.949 mmol) at 0 °C. The mixture was stirred for 4 h
at ambient temperature and then concentrated to remove CH Cl
3
(3H, s). C NMR (CDCl ) d: 170.99, 156.58, 74.38, 72.34, 61.71,
(
41).
a
61.66, 56.16, 56.04, 53.45, 52.51, 51.91, 48.46, 42.54, 39.66,
39.41, 37.57, 36.83, 36.50, 36.06, 35.69, 34.04, 28.74, 28.08,
27.92, 27.59, 24.02, 23.79, 22.74, 22.49, 21.27, 21.03, 18.61,
a
(
2
2
+
+
13.24, 11.94. FAB-MS m/z: 600 (MNa ), 578 (MH ).
(
2
2
.
4.2.4.10. 6
dimethylsilyloxy)ethyl]carbamate (46).
(198.1 mg, 0.343 mmol) in CH Cl (1 mL) and DMF (3 mL) was
a
-Hydroxy-5
a-cholestan-3b-yl N,N-bis[2-(tert-butyl-
Hexane was added to the residue, and then the hexane layer was
washed (water, MeOH) and concentrated. The resultant silyl ether
To a solution of 44
2
2
(
40) was used for the next reaction without further purification. A
solution of the silyl ether in THF (30 mL) was treated with BH
ꢀTHF
1.03 M in THF, 4 mL, 4.12 mmol) at 0 °C. The whole was stirred for
added imidazole (220.2 mg, 3.234 mmol) and TBSCl (232.0 mg,
1.539 mmol) at 0 °C. The mixture was stirred overnight at ambient
temperature and then diluted with AcOEt. The organic layer was
washed (water, aqueous ammonium chloride, brine) and concen-
trated. The resultant silyl ether (45) was used for the next reaction
without further purification. A mixture of the silyl ether and potas-
sium carbonate (956.1 mg, 6.918 mmol) in MeOH (12 mL) and CH2-
3
(
5
h at ambient temperature, and then 2 M aqueous NaOH (5 mL)
2 2
and 30% aqueous H O
The whole was stirred at ambient temperature for 17 h and then
diluted with AcOEt. The organic layer was washed (aqueous
(5 mL) were added to the solution at 0 °C.
ammonium chloride, water, brine), dried (Na
trated. The residue was purified by open column chromatography
n-hexane/AcOEt = 20/1) to afford the title compound (468.8 mg,
2
SO
4
), and concen-
Cl
diluted with AcOEt. The organic layer was washed (aqueous
ammonium chloride, brine), dried (Na SO ), and concentrated.
2
(2 mL) was stirred overnight at ambient temperature and then
(
2
4
1
0
.903 mmol, 30% for 2 steps) as colorless crystals. H NMR (CDCl
3
)
The residue was purified by open column chromatography (n-hex-
d: 3.57–3.49 (1H, m), 3.44–3.36 (1H, m), 0.91 (3H, d, J = 4.88 Hz),
ane/AcOEt = 7/1) to afford the title compound (179.8 mg,
0
0
6
3
2
.88 (9H, s), 0.86 (3H, d, J = 6.10 Hz), 0.86 (3H, d, J = 0.86 Hz),
0.235 mmol, 69% for 2 steps) as a colorless amorphous solid. 1
H
.80 (3H, s), 0.64 (3H, s), 0.05 (6H, s). 13C NMR (CDCl
) d: 72.05,
3
3
NMR (CDCl ) d: 4.59–4.51 (1H, m), 3.70 (2H, t, J = 5.49 Hz), 3.65
(2H, d, J = 6.41 Hz), 3.42–3.31 (4H + 1H, m), 0.89–0.88 (3H, m),
0.87 (9H, s), 0.86 (9H, s), 0.84 (3H, d, J = 6.71 Hz), 0.84 (3H, d,
9.61, 56.21 (2C), 53.89, 51.85, 42.59, 41.64, 39.87, 39.51, 37.44,
6.28, 36.14, 35.75, 34.30, 32.57, 31.67, 28.18, 28.01, 25.94 (3C),
4.20, 23.80, 22.80, 22.55, 21.15, 18.66, 18.21, 13.49, 12.03,
J = 6.10 Hz), 0.80 (3H, s), 0.63 (3H, s), 0.03 (6H, s), 0.02 (6H, s).
+
13
ꢂ4.54 (2C). FAB-MS m/z: 541 (MNa ).
3
C NMR (CDCl ) d: 155.70, 72.24, 69.42, 61.49 (2C), 56.16, 56.08,
5
3.66, 51.69, 51.11, 50.67, 42.52, 41.76, 39.75, 39.44, 37.06,
4
(
(
.2.4.8. Cholesterol-6
468.8 mg, 0.903 mmol), Ac
23.4 mg, 0.192 mmol) in pyridine (5 mL) was stirred at ambient
a
-yl acetate (43).
A solution of 41
36.23, 36.08, 35.71, 34.28, 28.76, 28.12, 27.93, 27.72, 25.85 (6C),
24.14, 23.77, 22.75, 22.50, 21.07, 18.60, 18.19, 18.15, 13.32,
11.97, ꢂ5.40 (2C), ꢂ5.46 (2C).
2
O (500 L, 5.29 mmol) and DMAP
l
temperature for 19 h and then diluted with AcOEt. The organic
layer was washed (aqueous ammonium chloride, aqueous sodium
bicarbonate, brine), dried (Na
4.2.4.11.
ethyl)carbamate (32).
0.0745 mmol) and TBAF (1 M in THF, 200
6a
-Hydroxy-5
a
-cholestan-3b-yl N,N-bis(2-hydroxy-
solution of 46 (56.9 mg,
L, 0.200 mmol) in THF
SO
2 4
), and concentrated. The resul-
A
tant acetate (42) was used for the next reaction without further
l
purification. A solution of the acetate and TBAF (1 M in THF,
(2 mL) was stirred at 45 °C for 2 h. Then the mixture was cooled
2
mL, 2.00 mmol) in THF (20 mL) was stirred at ambient tempera-
ture overnight and then diluted with AcOEt. The organic layer was
washed (water, aqueous ammonium, brine), dried (Na SO ), and
to ambient temperature and diluted with AcOEt. The organic layer
was washed (aqueous ammonium, brine), dried (Na SO ), and con-
2 4
centrated. The residue was purified by open column chromatogra-
2
4
concentrated. The residue was purified by open column chroma-
phy (n-hexane/AcOEt = 1:5 then AcOEt/MeOH = 10/1) to afford the
tography (n-hexane/AcOEt = 3:1) to afford the title compound
title compound (34.4 mg, 0.0642 mmol, 86%) as colorless needles
1
1
(
361.5 mg, 0.809 mmol, 90% for 2 steps) as colorless crystals.
NMR (CDCl
s), 0.90 (3H, d, J = 6.10 Hz), 0.88–0.84 (6H, m), 0.87 (3H, s), 0.65
H
(recrystallised from n-hexane/CH
NMR (CDCl
m), 3.58–3.52 (2H, m), 3.41–3.27 (2H + 1H, m), 0.90 (3H, d,
2
2
Cl , mp 198.3–200.1 °C).
H
3
) d: 4.71–4.64 (1H, m), 3.56–3.48 (1H, m), 2.02 (3H,
3
) d: 4.62–4.53 (1H, m), 4.25 (2H, br s), 3.87–3.68 (4H,

5
308
F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
J = 6.10 Hz), 0.87 (3H, d, J = 6.71 Hz), 0.86 (3H, d, J = 6.71 Hz), 0.83
C
32
H
57NO
5
1/4 H
2
O:C,71.13; H, 10.73; N, 2.59. Found: C, 71.11; H,
1
3
(
3H, s), 0.65 (3H, s). C NMR (CDCl
3
) d: 156.79, 75.15, 69.37,
10.35; N, 2.53.
6
4
2
1
1.54, 61.46, 56.27, 56.16, 53.69, 52.38, 51.88, 51.69, 42.58,
1.65, 39.80, 39.47, 37.09, 36.31, 36.14, 35.78, 34.33, 28.79,
8.16, 27.98, 27.72, 24.19, 23.89, 22.78, 22.54, 21.13, 18.64,
Acknowledgments
+
+
3.39, 12.02. FAB-MS m/z: 558 (MNa ), 536 (MH ). Anal. Calcd
: C, 71.73; H, 10.72; N, 2.61. Found: C, 71.50; H,
0.59; N, 2.54.
We thank Dr. Tappei Takada for helpful discussion, Dr. Tomomi
Noguchi–Yachide for performing analysis of our data in blind fash-
ion, and Shusuke Tomoshige for his insightful comments. This
work was supported in part by Grants–in–Aid for JSPS Fellows
and for Scientific Research (A) (JSPS KAKENHI Grant Numbers
for C32H57NO
5
1
4
.2.4.12. 6-Oxo-5
a
-cholestan-3b-yl N,N-bis[2-(tert-butyldimeth-
To a suspension of Celite
117.8 mg), sodium acetate (19.9 mg, 0.243 mmol) and PCC
68.2 mg, 0.316 mmol) in CH Cl (1 mL) was added a solution of
6 (122.9 mg, 0.161 mmol) in CH Cl (2 mL) at ambient tempera-
ylsilyloxy)ethyl]carbamate (47).
(
(
2
210583, 256522, and 22249006).
2
2
References and notes
4
2
2
ture. The whole was stirred at ambient temperature for 13 h, fil-
tered through a pad of Celite, and concentrated. The residue was
purified by open column chromatography (n-hexane/AcOEt = 9/1)
1
2
.
.
Davies, J. P.; Levy, B.; Ioannou, Y. A. Genomics 2000, 65, 137.
Altmann, S. W.; Davis, H. R.; Zhu, L.-J.; Yao, X.; Hoos, L. M.; Tetzloff, G.; Iyer, S. P.
N.; Maguire, M.; Golovko, A.; Zeng, M.; Wang, L.; Murgolo, N.; Graziano, M. P.
Science 2004, 303, 1201.
to afford the title compound (59.6 mg, 0.0782 mmol, 49%) as a col-
1
3. Davis, H. R.; Zhu, L.-J.; Hoos, L. M.; Tetzloff, G.; Maguire, M.; Liu, J.; Yao, X.; Iyer,
S. P. N.; Lam, M.-H.; Lund, E. G.; Detmers, P. A.; Graziano, M. P.; Altmann, S. W. J.
Biol. Chem. 2004, 279, 33586.
orless oil. H NMR (CDCl
3
) d: 4.61–4.54 (1H, m), 3.71 (2H, t,
J = 5.80 Hz), 3.67 (2H, t, J = 6.10 Hz), 3.42–3.36 (4H, m), 0.90 (3H,
d, J = 6.71 Hz), 0.88 (9H, s), 0.87 (9H, s), 0.86 (3H, d, J = 6.71 Hz),
4. Davies, J. P.; Scott, C.; Oishi, K.; Liapis, A.; Ioannou, Y. A. J. Biol. Chem. 2005, 280,
12710.
0
0
6
4
2
1
.85 (3H, d, J = 6.71 Hz), 0.75 (3H, s), 0.65 (3H, s), 0.04 (6H, s),
.03 (6H, s). C NMR (CDCl ) d: 210.56, 155.62, 73.45, 61.62,
3
1.40, 56.66, 56.60, 56.11, 53.84, 51.20, 50.71, 46.66, 42.97,
0.95, 39.44 (2C), 37.98, 36.47, 36.06, 35.67, 28.00, 27.98, 27.39,
6.59, 25.87 (7C), 23.94, 23.79, 22.77, 22.52, 21.48, 18.61, 18.20,
3.04, 11.98, ꢂ5.40 (2C), ꢂ5.43 (2C).
5
6
.
.
Davis, H. R.; Altmann, S. W. Biochim. Biophys. Acta 2009, 1791, 679.
Wang, J.; Chu, B.-B.; Ge, L.; Li, B.-L.; Yan, Y.; Song, B.-L. J. Lipid Res. 2009, 50,
1653.
Infante, R. E.; Abi-Mosleh, L.; Radhakrishnan, A.; Dale, J. D.; Brown, M. S.;
Goldstein, J. L. J. Biol. Chem. 2008, 283, 1052.
Infante, R. E.; Radhakrishnan, A.; Abi-Mosleh, L.; Kinch, L. N.; Wang, M. L.;
Grishin, N. V.; Goldstein, J. L.; Brown, M. S. J. Biol. Chem. 2008, 283, 1064.
Zhang, J.-H.; Ge, L.; Qi, W.; Zhang, L.; Miao, H.-H.; Li, B.-L.; Yang, M.; Song, B.-L. J.
Biol. Chem. 2011, 286, 25088.
1
3
7
8
9
.
.
.
4
.2.4.13. 6-Oxo-5
a
-cholestan-3b-yl N,N-bis(2-hydroxyethyl)car-
A solution of 47 (59.6 mg, 0.0782 mmol) and
10. Kwon, H. J.; Palnitkar, M.; Deisenhofer, J. PLoS ONE 2011, 6, e18722.
bamate (34).
11. Burnett, D. A.; Caplen, M. A.; Davis, H. R.; Burrier, R. E.; Clader, J. W. J. Med.
Chem. 1994, 37, 1733.
TBAF (1 M in THF, 200
lL, 0.200 mmol) in THF (2 mL) was stirred
12. Rosenblum, S. B.; Huynh, T.; Afonso, A.; Davis, H. R.; Yumibe, N.; Clader, J. W.;
at 45 °C for 3 h, then cooled to ambient temperature and diluted
with AcOEt. The organic layer was washed (aqueous ammonium,
brine), dried (Na SO ), and concentrated. The residue was purified
2 4
by open column chromatography (n-hexane/AcOEt = 1:5 to 1:9) to
Burnett, D. A. J. Med. Chem. 1998, 41, 973.
13. Garcia-Calvo, M.; Lisnock, J.; Bull, H. G.; Hawes, B. E.; Burnett, D. A.; Braun, M.
P.; Crona, J. H.; Davis, H. R.; Dean, D. C.; Detmers, P. A.; Graziano, M. P.; Hughes,
M.; Macintyre, D. E.; Ogawa, A.; O’neill, K. A.; Iyer, S. P. N.; Shevell, D. E.; Smith,
M. M.; Tang, Y. S.; Makarewicz, A. M.; Ujjainwalla, F.; Altmann, S. W.; Chapman,
K. T.; Thornberry, N. A. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 8132.
afford the title compound (28.4 mg, 0.0532 mmol, 68%) as a color-
less powder (recrystallised from n-hexane/CH
2
Cl
) d: 4.65–4.54 (1H, m), 3.86–3.74 (4H,
m), 3.63 (2H, br s), 3.52–3.39 (4H, m), 0.91 (3H, d, J = 6.71 Hz),
2
, mp 138.9–
14. Weinglass, A. B.; Kohler, M.; Schulte, U.; Liu, J.; Nketiah, E. O.; Thomas, A.;
Schmalhofer, W.; Williams, B.; Bildl, W.; McMasters, D. R.; Dai, K.; Beers, L.;
McCann, M. E.; Kaczorowski, G. J.; Garcia, M. L. Proc. Natl. Acad. Sci. U.S.A. 2008,
1
1
40.1 °C). H NMR (CDCl
3
1
05, 11140.
0
0
.87 (3H, d, J = 6.71 Hz), 0.86 (3H, d, J = 6.71 Hz), 0.77 (3H, s),
.66 (3H, s). C NMR (CDCl ) d: 210.82, 156.59, 74.14, 61.84
3
15. Yamanashi, Y.; Takada, T.; Suzuki, H. J. Pharmacol. Exp. Ther. 2007, 320, 559.
16. Temel, R. E.; Tang, W.; Ma, Y.; Rudel, L. L.; Willingham, M. C.; Ioannou, Y. A.;
Davies, J. P.; Nilsson, L.-M.; Yu, L. J. Clin. Invest. 1968, 2007, 117.
1
3
(
2C), 56.65, 56.52, 56.11, 53.79, 52.53, 51.92, 46.64, 42.98, 40.99,
17. Hawes, B. E.; O’neill, K. A.; Yao, X.; Crona, J. H.; Davis, H. R.; Graziano, M. P.;
Altmann, S. W. Mol. Pharmacol. 2007, 71, 19.
3
2
9.43 (2C), 37.99, 36.40, 36.07, 35.67, 27.98 (2C), 27.30, 26.56,
1
8. Pfefferkorn, J. A.; Larsen, S. D.; Van Huis, C.; Sorenson, R.; Barton, T.; Winters, T.;
Auerbach, B.; Wu, C.; Wolfram, T. J.; Cai, H.; Welch, K.; Esmaiel, N.; Davis, J.;
Bousley, R.; Olsen, K.; Mueller, S. B.; Mertz, T. Bioorg. Med. Chem. Lett. 2008, 18,
546.
3.94, 23.80, 22.77, 22.52, 21.48, 18.61, 13.07, 11.98. FAB-MS m/
+
+
z: 556 (MNa ), 534 (MH ). Anal. Calcd for C32
7
H55NO
5
ꢀ1/3 H
2
O: C,
1.20; H, 10.39; N, 2.59. Found: C, 71.34; H, 10.24; N, 2.53.
19. McMasters, D. R.; Garcia-Calvo, M.; Maiorov, V.; McCann, M. E.; Meurer, R. D.;
Bull, H. G.; Lisnock, J.; Howell, K. L.; Devita, R. J. Bioorg. Med. Chem. Lett. 2009,
4
.2.4.14. 6b-Hydroxy-5
a
-cholestan-3b-yl N,N-bis(2-hydroxy-
Following a reported protocol, the 6-
1
9, 2965.
ethyl)carbamate (33).
20. Howell, K. L.; Devita, R. J.; Garcia-Calvo, M.; Meurer, R. D.; Lisnock, J.; Bull, H.
keto group was reduced to 6b-hydroxyl group.44 To a solution of
G.; McMasters, D. R.; McCann, M. E.; Mills, S. G. Bioorg. Med. Chem. Lett. 2010,
2
0, 6929.
3
4 (20.8 mg, 0.0390 mmol) in MeOH (1 mL) was added NaBH
4
2
1. Simeone, J. P.; Braun, M. P.; Leone, J. F.; Lin, P.; DeVita, R. J.; Garcia-Calvo, M.;
Bull, H. G.; Lisnock, J.; Dean, D. C. Bioorg. Med. Chem. Lett. 2009, 19, 5033.
22. Davies, J. P.; Levy, B.; Ioannou, Y. A. Genomics 2000, 65, 137.
3. Gelsthorpe, M. E.; Baumann, N.; Millard, E.; Gale, S. E.; Langmade, S. J.; Schaffer,
J. E.; Ory, D. S. J. Biol. Chem. 2008, 283, 8229.
4. Loo, T. W.; Clarke, D. M. J. Biol. Chem. 1997, 272, 709.
(
4.4 mg, 0.12 mmol) at 0 °C. The mixture was stirred at ambient
temperature overnight and then diluted with AcOEt. The organic
layer was washed (water, aqueous ammonium chloride, brine),
dried (Na SO ), and concentrated. The residue was purified by
2 4
2
2
open column chromatography (n-hexane/AcOEt = 1:5 to 1:10) to
afford the title compound (14.5 mg, 0.0271 mmol, 69%) as color-
25. Fan, J. Q.; Ishii, S.; Asano, N.; Suzuki, Y. Nat. Med. 1999, 5, 112.
2
6. Morello, J. P.; Salahpour, A.; Laperrière, A.; Bernier, V.; Arthus, M. F.; Lonergan,
M.; Petäjä-Repo, U.; Angers, S.; Morin, D.; Bichet, D. G.; Bouvier, M. J. Clin.
Invest. 2000, 105, 887.
less plates (recrystallised from n-hexane/CH
2
Cl
OD) d: 4.69–4.60 (1H, m),
.81–3.70 (4H + 1H, m), 3.56–3.33 (4H, m), 1.04 (3H, s), 0.91 (3H,
2
/MeOH, mp
1
1
94.0–196.2 °C). H NMR (CDCl
3
/CD
3
27. Morello, J. P.; Petäjä-Repo, U. E.; Bichet, D. G.; Bouvier, M. Trends Pharmacol. Sci.
000, 21, 466.
2
3
2
2
8. Loo, T. W.; Clarke, D. M. Expert Rev. Mol. Med. 2007, 9, 1.
9. Ohgane, K.; Karaki, F.; Dodo, K.; Hashimoto, Y. Chem. Biol. 2013, 20, 391.
d, J = 6.71 Hz), 0.87 (3H, d, J = 6.71 Hz), 0.86 (3H, d, J = 6.71 Hz),
1
3
0
5
3
2
1
.69 (3H, s). C NMR (CDCl
3
) d: 156.83, 75.31, 71.51, 61.30,
30. Denning, G. M.; Anderson, M. P.; Amara, J. F.; Marshall, J.; Smith, A. E.; Welsh,
M. J. Nature 1992, 358, 761.
6.20, 56.03, 54.02, 52.29, 51.78, 47.08, 42.58, 39.80, 39.48,
8.43, 38.18, 36.08, 35.70, 35.30, 31.68, 30.30, 29.60, 28.11,
7.92, 27.76, 24.13, 23.75, 22.69, 22.44, 20.94, 18.58, 15.52,
31. Wang, X.; Koulov, A. V.; Kellner, W. A.; Riordan, J. R.; Balch, W. E. Traffic 2008, 9,
1878.
32. Adams, C. M.; Reitz, J.; De Brabander, J. K.; Feramisco, J. D.; Li, L.; Brown, M. S.;
+
+
Goldstein, J. L. J. Biol. Chem. 2004, 279, 52772.
1.98. FAB-MS m/z: 558 (MNa ), 536 (MH ). Anal. Calcd for

F. Karaki et al. / Bioorg. Med. Chem. 21 (2013) 5297–5309
5309
3
3
3. Bielska, A. A.; Schlesinger, P.; Covey, D. F.; Ory, D. S. Trends Endocrinol. Metab.
012, 23, 99.
4. Chen, Y.; Chen, C.; Wang, Y.; Liu-Chen, L.-Y. J. Pharmacol. Exp. Ther. 2006, 319,
65.
5. Thompson, J. F.; Hayes, L. S.; Lloyd, D. B. Gene 1991, 103, 171.
6. MacEwan, D. J.; Milligan, G. Mol. Pharmacol. 1996, 50, 1479.
39. Yu, L.; Bharadwaj, S.; Brown, J. M.; Ma, Y.; Du, W.; Davis, M. A.; Michaely, P.;
Liu, P.; Willingham, M. C.; Rudel, L. L. J. Biol. Chem. 2006, 281, 6616.
40. Yates, P.; Stiver, S. Can. J. Chem. 1987, 65, 2203.
41. Miller, R. A.; Li, W.; Humphrey, G. R. Tetrahedron Lett. 1996, 37, 3429.
42. Carvalho, J. F. S.; Cruz Silva, M. M.; Moreira, J. N.; Simões, S.; Sá E Melo, M. L. J.
Med. Chem. 2009, 52, 4007.
43. Jung, M. E.; Johnson, T. W. Org. Lett. 1999, 1, 1671.
44. Schmidt, A. W.; Doert, T.; Goutal, S.; Gruner, M.; Mende, F.; Kurzchalia, T. V.;
Knölker, H.-J. Eur. J. Org. Chem. 2006, 2006, 3687.
2
7
3
3
3
3
7. Ramsay, D.; Bevan, N.; Rees, S.; Milligan, G. Br. J. Pharmacol. 2001, 133, 315.
8. Auld, D. S.; Thorne, N.; Nguyen, D.-T.; Inglese, J. ACS Chem. Biol. 2008, 3,
4
63.