Regulation of liver X receptor target genes by 22-functionalized oxysterols. Synthesis, in silico and in vitro evaluations

Article abstract of DOI:10.1016/j.steroids.2016.12.003

The endogenous oxysterol 22(R)-hydroxycholesterol (22RHC, 1) is an LXR agonist which upregulates genes of critical involvement in human cholesterol- and lipid metabolism. In contrast, its synthetic epimer 22(S)-hydroxycholesterol (22SHC, 8) has shown specific antagonistic effects in recent studies, avoiding unwanted side effects provided by potent LXR agonists. In terms of LXR modulation, the aim of this study was to compare 22SHC (8), 22RHC (1) and synthesized ligands with keto- and amide functionality in the 22nd position of the cholesterol scaffold. 22SHC (8) and 22RHC (1) performed as expected while 22-ketocholesterol (22KC, 10) revealed an attractive in vitro profile for further investigation in terms of anti-atherosclerotic properties as selective upregulation of the ATP-binding cassette transporter ABCA1 was observed. A new synthesized amide derivate, Fernholtz cyclohexylamide (13) was shown to reduce lipogenesis in a dose-responsive manner and abolish the effect of the potent LXR agonist T0901317 when administered simultaneously.

Full text of DOI:10.1016/j.steroids.2016.12.003

Accepted Manuscript
Regulation of liver X receptor target genes by 22-functionalized oxysterols.
Synthesis, in silico and in vitro evaluations
Elvar Örn Viktorsson, Mari Gabrielsen, Nugalya Kumarachandran, Ingebrigt
Sylte, Pål Rongved, Ove Alexander Høgmoen Åstrand, Eili Tranheim Kase
PII:
S0039-128X(16)30182-9
http://dx.doi.org/10.1016/j.steroids.2016.12.003
STE 8059
DOI:
Reference:
Steroids
To appear in:
Received Date:
Revised Date:
Accepted Date:
15 September 2016
5 December 2016
7 December 2016
Please cite this article as: Örn Viktorsson, E., Gabrielsen, M., Kumarachandran, N., Sylte, I., Rongved, P., Alexander
Høgmoen Åstrand, O., Tranheim Kase, E., Regulation of liver X receptor target genes by 22-functionalized
oxysterols. Synthesis, in silico and in vitro evaluations, Steroids (2016), doi: http://dx.doi.org/10.1016/j.steroids.
2016.12.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Regulation of liver X receptor target genes by 22-functionalized
oxysterols. Synthesis, in silico and in vitro evaluations.
Elvar Örn Viktorsson,a Mari Gabrielsen,b Nugalya Kumarachandran,c Ingebrigt Sylte,b Pål
Rongved,a Ove Alexander Høgmoen Åstrand, *#a and Eili Tranheim Kase#c
a School of Pharmacy, Department of Pharmaceutical Chemistry, University of Oslo, PO Box 1068 Blindern, N0316 Oslo, Norway.
E-mail: o.a.h.astrand@farmasi.uio.no
b Department of Medical Biology, Faculty of Health Sciences, University of Tromsø, Tromsø, Norway
c Department of Pharmaceutical Biosciences, School of Pharmacy, University of Oslo, Oslo, Norway
*Corresponding author
#Contributed equally.
Keywords: Liver X receptors; Oxysterols; Steroid synthesis; Molecular modelling; Biological
evaluation
Abstract:
The endogenous oxysterol 22(R)-hydroxycholesterol (22RHC, 1) is an LXR agonist which upregulates genes of critical
involvement in human cholesterol- and lipid metabolism. In contrast, its synthetic epimer 22(S)-hydroxycholesterol (22SHC,
8) has shown specific antagonistic effects in recent studies, avoiding unwanted side effects provided by potent LXR agonists.
In terms of LXR modulation, the aim of this study was to compare 22SHC (8), 22RHC (1) and synthesized ligands with keto-
and amide functionality in the 22nd position of the cholesterol scaffold. 22SHC (8) and 22RHC (1) performed as expected
while 22-ketocholesterol (22KC, 10) revealed an attractive in vitro profile for further investigation in terms of anti-
atherosclerotic properties as selective upregulation of the ATP-binding cassette transporter ABCA1 was observed. A new
synthesized amide derivate, Fernholtz cyclohexylamide (13) was shown to reduce lipogenesis in a dose-responsive manner
and abolish the effect of the potent LXR agonist T0901317 when administered simultaneously.
1. Introduction
The liver X receptor (LXR) is a ligand-activated transcription factor and a member of the nuclear
receptor (NR) superfamily.[1] Two isoforms are known, LXRα (NR1H3) and –β (NR1H2) and the
pathways they regulate in vivo rely on activation by endogenous oxysterols.[2, 3] LXRs are involved
in regulating expression of cohorts of genes by interacting with LXR-responsive elements.[4, 5]
Evidences collected over the past decades confirm the crucial role of LXRs in the metabolism of
cholesterol, lipids, bile acids and carbohydrates.[6-9] LXR-induced gene expression is tissue specific
as LXRα is highly expressed in liver, kidneys, small intestine, macrophages and skeletal muscle, while
the β-isoform is expressed more widely and has been identified in most tissues throughout the human
body.[2, 10, 11] The importance of LXRs as pharmacological targets has been considered in several
disease states, such as atherosclerosis,[12] diabetes mellitus type 2 (T2DM) and obesity,[13, 14]
cancer,[15, 16] inflammation[17], skin disorders and Alzheimer’s disease.[19, 20] Moreover, novel
Fig. 1 Two documented LXR-agonists. On the left is the endogenous LXR agonist 22(R)-Hydroxycholesterol (1) with key positions of the
cholesterol scaffold numbered. On the right, T0901317 (2), a potent synthetic LXR agonist is shown.

genes expressed upon LXR activation are still discovered.[1, 4] The LXRs have relatively recently
been considered as a drug targets, as it still was unknown in 1995 whether these receptors had
endogenous ligands.[21] In the active state, LXR bound to its ligand forms an obligatory heterodimer
with retinoid X receptor (RXR) that assistantly can be activated by retinoic acid. The activated
complex is bound to DNA responsive elements in order to initiate gene expression.[21] Thus, the
target is of high complexity. Advances in drug development of LXR-modulators has suffered from the
lack of options for targeting either isoform of the receptor selectively as the α and –β isoforms share
74% of the same amino acid sequence in the ligand binding domain (LBD) and only differ in 2
residues at the ligand binding site.[22] Further, the ligand binding pocket is of large volume compared
with other NRs[23] and high flexibility can be seen observing the great chemical diversity of ligands
known to bind and activate LXRs. An array of ligands, natural and synthetic ones have been explored
in order to address their influence on LXR-induced gene expression.[22, 24] LXR agonists such as
T0901317 (2)(fig. 1) and GW3965 (structure not shown) are potent inducers of gene expression of key
regulatory genes involved in certain metabolic processes.[25, 26] Central target genes of LXR
transcription involve the ATP binding cassette transporters (ABC transporters ABCA1/G1 and
ABBCG5/G8), sterol regulatory element-binding protein-1c gene (SREBPC1c) and others of known
importance in lipogenesis including fatty acid synthase (FAS) and stearoyl-CoA desaturase (SCD).
Additionally, lipidation of Apolipoprotein E (ApoE) which is highly expressed in the human brain and
connected to the pathogenesis of Alzheimer´s disease is LXR regulated.[19] To date, LXR activation
(LXR-agonism) has particularly been of interest due to desirable up-regulation of ABCA1 expression
in the light of anti-atherosclerotic properties.[27] However, their therapeutic values are largely
hampered by undesirable side effects, such as hypertriglyceridemia, enhanced lipogenesis and hepatic
steatosis, which have shown to be hard to avoid, leaving full LXR agonists only as research tools.[13,
28]
The effect of the synthetic derivative 22(S)-hydroxycholesterol (22SHC, 8) on the LXR receptors
and lipid metabolism has been discussed and is reported in prior to be inactive.[3, 5, 29] In contrast,
recent studies report that 22SHC (8) may possess antagonistic effects on LXRs and have the ability to
selectively down-regulate the expression of genes involved in lipid formation, reduce lipid metabolism
and increase glucose uptake.[6, 30] The corresponding epimer 22(R)-hydroxycholesterol (22RHC, 1),
is a well-known endogenous agonist for lipid production.[1] These biological findings are of interest,
since they suggest that a single change in stereochemistry of the steroidal side chain manipulates gene
expression of key LXR target genes oppositely. The literature compound 22-ketocholesterol (22KC,
10) has earlier been shown[31] to be oxidized from 22SHC (1) by NADPH dependent oxygenase [32]
and has been described as a potent inhibitor of P-450-driven cleavage of cholesterol.[33] To our best
knowledge, 22KC (10) has not been evaluated in terms of LXR-modulation.
We hypothesized that further investigation of the cholesterol side-chain functionality specifically at
the 22-position could lead to new therapeutic candidates beneficial for diabetes and obesity. The first
step further was to investigate 22-oxo functionalized cholesterol derivatives. To achieve this, our
recently developed synthetic pathway from Fernholtz acid (3) to TBS-protected Fernholtz Weinreb
amide (5) proved resourceful in order to provide 22-carbonyl-cholesterol analogs such as the TBS-
protected Fernholtz aldehyde (6) and TBS-protected 22KC (9). Molecular modelling studies were
undertaken to explore ligand docking scores and interaction modes for LXRβ. To evaluate
performance of ligands in biological systems, their modulating effects on LXR controlled gene
expression was measured by several assays using both human skeletal muscle cells and HepG2 cells.

2. Experimental procedures
2.1. Materials
General information regarding the materals for the synthetic part of this work is found within the
supplementary information. Dulbecco‘s modified Eagle’s medium (DMEM-Glutamax™, 5.5 mM),
DMEM, foetal bovine serum, Ultroser G, penicillin-streptomycin-amphotericin B, and trypsin-EDTA
were obtained from Gibco, Life Technologies (Paisley, UK). [1-14C]acetic acid (54 mCi/mmol) and
D-[14C(U)]deoxy-D-glucose (6.0 Ci/mmol) were purchased from ARC (American Radiolabeled
Chemicals, St. Louis, MO, USA). Insulin Actrapid was from Novo Nordisk (Bagsvaerd, Denmark).
Bovine serum albumin (BSA) (essentially fatty acid-free) and 22(S)-hydroxycholesterol (8) (and
22(R)-hydroxycholesterol (1) were purchased from Sigma-Aldrich (St. Louis, MO, USA). RNeasy
Mini kit and RNase-free DNase were purchased from Qiagen Sciences (Oslo, Norway). Agilent Total
RNA isolation kit was from Agilent Technologies (Santa Clara, CA, USA). The primers were
purchased from Invitrogen (Paisley, Scotland, UK), while SYBR® Green and TaqMan® reverse-
transcription reagents kit were from Applied Biosystems (Foster City, Canada). T0901317 (2) was
obtained from Cayman Chemical Company (Ann Arbor, MI, USA). Hydrophobic MultiScreen® HTS
plates were from Millipore (Billerica, MA, USA). Corning® CellBIND® tissue culture plates were
obtained from Corning Life-Sciences (Schiphol-Rijk, The Netherlands). The protein assay reagent was
obtained from BioRad (Copenhagen, Denmark). All other chemicals used were used as received and
of high quality.
2.2. Docking experiments
The ligands were docked into the crystal structure of the LXRβ ligand binding domain in complex
with 24(S),25-epoxycholesterol (PDB id 1P8D) using Internal Coordinate Mechanics (ICM) software
version 3.8-4.[35] To set up the receptor grid maps for docking, amino acids within 5 Å of the co-
crystallized ligand were selected. The ligands were charged using ICM auto pKa macro (pH 7) and
converted to 3D before docking. Three parallel docking simulations were performed and the best-
scored ligand from the parallels was selected as the best orientation. The ICM virtual ligand screening
(VLS) scoring function was used.
2.3. Synthetic procedures
2.3.1. Fernholtz Weinreb amide (4)
Fernholtz acid 3 (400 mg, 1.15 mmol, 1 eq) was dispersed in anhydrous DMF (20 mL) stirring at r.t.
under argon atm. To the suspension, HATU (483 mg, 1.27 mmol, 1.1 eq) and HOBt monohydrate (195
mg, 1.27 mmol, 1.1 eq) were added followed by drop wise addition af DIPEA (0.24 mL, 1.38 mmol,
1.2 eq) upon which a transparent solution was obtained. The resulting solution was left stirring for 2
hours or until no starting material was observed by TLC. Then, N,O-dimethylhydroxylamine
hydrochloride (561 mg, 5.75 mmol, 5 eq) was added, followed by another drop wise addition of
DIPEA (0.98 mL, 5.75mmol, 5 eq). The resulting mixture was stirred overnight. The reaction mixture
was transferred to a separatory funnel, diluted by H2O (400 mL) and extracted by EtOAc (4x25 mL).
The organic phases were pooled, washed with brine (3x300 mL), dried over MgSO4 and filtered before
solvents were removed in vacuo. The crude product was further purified by flash column
chromatography using 50% EtOAc/Heptane as eluent yielding 348 mg (77%) of the desired white
1
solid. Rf: 0.40 (EtOAc/Heptane 8:2). [α]20 = -34.2 (1.05 mg/mL, CHCl3). H NMR (300 MHz,
D

CDCl3) δ 5.40 – 5.24 (m, 1H), 3.66 (s, 3H), 3.59 – 3.41 (m, 1H), 3.14 (s, 3H), 3.03 – 2.75 (m, 1H),
2.34 – 0.76 (m, 27H), 0.71 (s, 3H). 13C NMR (101 MHz, CDCl3) δ 178.3, 141.0, 121.6, 71.8, 61.5,
56.3, 52.7, 50.2, 42.4, 42.4, 39.7, 37.4, 37.1, 36.6, 32.3, 32.0, 31.9, 31.7, 27.2, 24.4, 21.2, 19.5, 17.4,
12.3. HRMS: e/z calculated for C24H39NO3 389.2930, found 389.2924 (1.5 ppm).
2.3.2. TBS-Fernholtz Weinreb amide (5)
2,6-Lutidine (0.21 mL, 1.79 mmol, 1.8 eq) and tert-butyldimethylsilyl triflate (0.26 mL, 1,13 mmol,
1.2 eq) were added drop wise to a stirring solution of Fernholtz Weinreb amide 4 (383 mg, 1.06 mmol)
in anhydrous DCM at -40°C under Ar-atm. The reaction mixture was stirred for 30 minutes until no
starting material was observed by TLC and gradually allowed to reach rt and stirred for 1 additional
hour. The reaction mixture was absorbed on silica gel and solvents removed in vacuo. The crude
product was purified by flash column chromatography on silica using 20% EtOAc/heptane as the
1
eluent yielding 384 mg (78%) of white solid. Rf: 0.52 (EtOAc/heptane 1:4). H NMR (400 MHz,
CDCl3) δ 5.40 – 5.25 (m, 1H), 3.69 (s, 3H), 3.57 – 3.41 (m, 1H), 3.16 (s, 3H), 3.05 – 2.84 (m, 1H),
13
2.36 – 2.07 (m, 2H), 2.07 – 0.79 (m, 33H), 0.73 (s, 3H), 0.06 (s, 6H). C NMR (101 MHz, CDCl3) δ
178.4, 141.6, 121.2, 72.8, 61.5, 56.4, 52.8, 50.3, 43.0, 42.4, 39.8, 37.5, 37.2, 36.7, 32.3, 32.2, 32.1,
32.0, 27.2, 26.1, 24.5, 21.2, 19.6, 18.4, 17.4, 12.4, -4.4. HRMS (TOF MS ES+): Exact mass calculated
for C30H53NO3SiNa [M+Na]+ : 526.3692, found 526.3696 (0.68 ppm).
2.3.3. TBS-Fernholtz aldehyde (6)
1M DIBALH solution in THF (4.1 mL, 4.1 mmol) was added drop wise to a stirring solution of TBS-
Weinreb amide 5 (0.50 g, 1.02 mmol) in DCM (20 mL, anhydrous) at -78 °C under argon atm. After 2
h of stirring, the reaction mixture was gradually allowed to reach rt and consequently quenched
(NaHCO3 and Na-K-tartrate saturated aqueous sol., 20 mL of each) and left stirring for 1 h. The crude
mixture was transferred to a separatory funnel and further diluted with H2O (200 mL). The aqueous
phase was extracted by EtOAc (4x25 mL), dried over MgSO4, filtered and concentrated in vacuo. The
resulting crude solution was absorbed on silica gel and further purified by flash column
chromatography on silica (EtOAc/heptane 1:9) affording 375 mg (83%) of white solid. Rf: 0.68 (1:1
1
EtOAc/heptane v/v). H NMR (300 MHz, CDCl3) δ 9.57 (d, J = 3.2 Hz, 1H), 5.31 (m, 1H), 3.48 (m,
1H), 2.43 – 2.10 (m, 3H), 2.08 – 0.77 (m, 34H), 0.72 (s, 3H), 0.05 (s, 6H). 13C NMR (75 MHz, CDCl3)
δ 205.2, 141.7, 121.1, 72.7, 56.2, 51.2, 50.3, 49.7, 43.1, 43.0, 39.7, 37.5, 36.7, 32.2, 32.1, 32.0, 27.2,
1
26.1, 24.8, 21.2, 19.6, 18.4, 13.6, 12.4, -4.4. H and 13C NMR data are in accordance with
literature.[36]
2.3.4. TBS-22(S)-hydroxycholesterol (7)
2M Isopentyl MgBr solution in THF (0.68mL, 1.35mmol) was added drop wise to a stirring
solution of TBS-Fernholtz aldehyde 6 (120 mg, 0.27 mmol) in anhydrous THF (3 mL) at -78°
C under N2 atm. The reaction mixture was left stirring for 3.5 hrs, and then quenched with
NH4Cl (10 mL of 10% aqueous sol.). The aqueous phase was extracted with EtOAc (3x20
mL). Organic phases were pooled, dried over MgSO4 and filtered. The crude solution was
absorbed onto silica gel. Flash column chromatography (EtOAc/heptane 1:9) yielded 73 mg
(53%) of a white solid. Rf: 0.29 (EtOAc/heptane 1:9) 1H NMR (400 MHz, CDCl3) δ 5.36 – 5.27
(m, 1H), 3.68 – 3.55 (m, 1H), 3.54 – 3.42 (m, 1H), 2.33 – 2.10 (m, 2H), 2.09 – 0.76 (m, 46H),

0.69 (s, 3H), 0.05 (s, 6H). 13C NMR (101 MHz, CDCl3) δ 141.7, 121.2, 74.1, 72.8, 56.9, 52.8,
50.3, 43.00, 42.4, 40.3, 40.00, 37.5, 36.7, 35.8, 33.4, 32.3, 32.1, 32.1, 28.3, 27.9, 26.1, 24.4,
22.9, 22.7, 21.3, 19.6, 18.4, 11.9, 11.7, -4.4.
2.3.5. 22-Hydroxycholesterol (22SHC, 8)
TBAF (215 mg, 0.68 mmol) was added to a stirring solution of TBS-22SHC 7 (70.4 mg, 0.14 mmol)
in THF (2 mL). The reaction mixture was stirred at rt for 3 h until no starting material was observed
by TLC. The reaction mixture was concentrated in vacuo before the solid crude material was dispersed
in H2O and extracted by EtOAc (3x25 mL). The pooled organic phases were dried over MgSO4,
filtered and concentrated in vacuo. The crude compound was asbsorbed onto silica gel and further
purified by flash column chromatography on silica (20-50% EtOAc in Heptane) to afford 37 mg
1
(67%) as a white solid. Rf: 0.55 (50% EtOAc/Heptane). H NMR (600 MHz, CDCl3) δ 5.37 – 5.32 (m,
1H), 3.62 (t, J = 6.3 Hz, 1H), 3.56 – 3.47 (m, 1H), 2.34 – 2.18 (m, 2H), 2.04 – 1.79 (m, 5H), 1.68 –
13
0.81 (m, 33H), 0.69 (s, 3H). C NMR (101 MHz, CDCl3) δ 140.9, 121.8, 74.8, 71.9, 56.8, 52.7, 50.3,
42.4, 42.4, 40.3, 40.0, 37.4, 36.6, 35.8, 33.4, 32.1, 32.0, 31.8, 28.3, 27.9, 24.4, 22.8, 22.7, 21.3, 19.6,
11.9, 11.7.
2.3.6. TBS-22-keto cholesterol (9)
Dess-Martin periodinane (127 mg, 0.30 mmol, 1.5 eq) was added to a stirring solution of TBS-22SHC
7 (103 mg, 0.30 mmol, 1eq) in DCM (20 mL) at r.t. The reaction mixture was stirred for 20 h, and
then diluted by 0.5M NaHCO3 (30 mL) and extracted by DCM (3x20 mL). Organic phases were
pooled, dried over MgSO4 and filtered before solvent was removed in vacuo. The resulting crude
compound was purified by flash column chromatography (EtOAc/heptane 1:9) yielding 87 mg (85%)
1
of the white solid. [α]20 = -46.2 (c = 1.04 mg/mL, CHCl3). Rf: 0.51 (EtOAc/heptane 1:1). H NMR
(400 MHz, CDCl3) δ 5.36 – 5.27 (m, 1H), 3.54 – 3.42 (m, 1H), 2.62 – 2.06 (m, 5H), 2.06 – 0.76 (m,
13
42H), 0.70 (s, 3H), 0.05 (s, 6H). C NMR (101 MHz, CDCl3) δ 215.1, 141.7, 121.2, 72.8, 56.3, 52.2,
50.3, 49.6, 43.0, 42.7, 39.8, 37.5, 36.7, 32.6, 32.2, 32.1, 32.0, 27.9, 27.7, 26.1, 24.7, 22.6, 22.5, 21.2,
19.6, 18.4, 16.8, 12.2, -4.4. HRMS (TOF MS ES+): Exact mass calculated for C33H58O2SiNa [M+Na]+:
537.4103, found 537.4109 (0.97 ppm).
2.3.7. 22-ketocholesterol (10)
TBAF (237 mg, 0.75 mmol) was added to a stirring solution of TBS-22-Ketone 9 (77mg, 0.15 mmol)
in THF (3 mL). The reaction mixture was stirred at r.t. for 24 hrs. THF was removed in vacuo and the
resulting crude dispersed in 10% NH4Cl (30 mL). The aqueous layer was extracted by DCM (3x30
mL). The collected organic phases were combined and solvents removed under reduced pressure
before further purification by flash column chromatography (20-50% EtOAc/heptane) to yield 42.2
1
mg (70%) of a white solid. Rf: 0.18 (30% EtOAc/heptane). [α]20 = -26.5 (0.83 mg/mL, CHCl3). H
D
NMR (400 MHz, CDCl3) δ 5.37 – 5.28 (m, 1H), 3.58 – 3.43 (m, 1H), 2.58 – 2.12 (m, 5H), 2.02 – 1.89
(m, 1H), 1.87 – 1.78 (m, 3H), 1.72 – 1.34 (m, 9H), 1.33 – 1.21 (m, 1H), 1.19 – 0.91 (m, 8H), 0.87 (d, J
= 6.5 Hz, 8H), 0.69 (s, 3H). 13C NMR (101 MHz, CDCl3) δ 215.1, 140.9, 121.6, 71.8, 56.2, 52.2, 50.2,
49.6, 42.6, 42.4, 39.8, 39.8, 37.4, 36.6, 32.5, 32.0, 32.0, 31.6, 27.8, 27.7, 24.6, 22.5, 22.5, 21.2, 19.5,
16.7, 12.2. HRMS e/z calculated for C27H44O2: 400.3341, found 400.3348 (-1.7 ppm).

2.3.8. General method for the synthesis of amides (11-18)
Fernholtz acid 3 (0.50 mmol, 1.0 eq) was dispersed in DMF (15 mL) stirring at rt under Ar-atm. To
this suspension, HBTU (0.55 mmol), HOBt monohydrate (0.55 mmol) and the corresponding amine
(0.50-2.5 mmol) were added. The resulting solution was stirred for 5 minutes before DIPEA (0.50
mmol) was added drop wise. The resulting mixture left stirring overnight. The resulting reaction
mixture was diluted with H2O (200 mL) and extracted by EtOAc (4x30 mL). Organic phases were
pooled and washed with 0.5M NaHCO3 (200 mL), brine (2x200 mL), dried over MgSO4 and filtered
before solvents were removed in vacuo. The resulting crude compound was dry loaded on silica gel
and further purified by flash column chromatography. Eluent systems are stated individually below for
compounds 11-18.
2.3.9. Fernholtz isobutyl amide (11)
Flash column chromatography (Heptane/EtOAc 1:4 - 1:1) gave 155mg (75%) of the cream-white
1
solid. [α]20 = -45.3 (c = 1.14 mg/mL, CHCl3). Rf: 0.38 (80% EtOAc/heptane). H NMR (400 MHz,
D
CDCl3) δ 5.45 (t, J = 6.1 Hz, 1H), 5.38 – 5.28 (m, 1H), 3.59 – 3.44 (m, 1H), 3.19 – 2.92 (m, 2H), 2.34
– 2.16 (m, 2H), 2.13 – 0.75 (m, 35H), 0.69 (s, 3H). 13C NMR (101 MHz, CDCl3) δ 176.7, 140.9,
121.7, 71.9, 56.5, 52.9, 50.2, 46.8, 45.3, 42.4, 42.4, 39.7, 37.4, 36.6, 32.1, 31.9, 31.8, 28.7, 27.9, 24.5,
21.2, 20.3, 19.5, 17.9, 12.2. HRMS: e/z calculated for C26H43NO2 401.3294, found 401.3300 (-1.6
ppm)
2.3.10. Fernholtz n-butyl amide (12)
Flash column chromatography on silica (heptane/EtOAc 1:4) gave 157 mg (82%) of the cream-white
solid. Rf: 0.53 (EtOAc/heptane 8:2). [α]20 = -58.0 (1.90 mg/mL, CHCl3). 1H NMR (400 MHz, CDCl3)
δ 5.79 – 5.39 (m, 1H), 5.39 – 5.26 (m, 1H), 3.61 – 3.42 (m, 1H), 3.36 – 3.10 (m, 2H), 2.65 – 0.76 (m,
13
44H), 0.68 (s, 3H). C NMR (101 MHz, CDCl3) δ 176.8, 140.9, 121.6, 71.8, 56.5, 52.9, 50.2, 45.0,
42.5, 42.4, 39.7, 39.2, 37.4, 36.6, 32.1, 31.9, 31.9, 31.8, 27.7, 24.5, 21.2, 20.2, 19.5, 17.8, 13.9, 12.2.
HRMS: e/z calculated for C26H43NO2 401.3294, found 401.3291 (0.7 ppm).
2.3.11. Fernholtz cyclohexyl amide (13)
Flash column chromatography (Heptane/EtOAc 1:4) gave 120 mg (64%) of cream-white solid. Rf:
0.48 (EtOAc/heptane 8:2). [α]20D = -46.9 (1.02 mg/mL, CHCl3). 1H NMR (400 MHz, CDCl3) δ 5.56 –
5.13 (m, 1H), 3.83 – 3.67 (m, 1H), 3.60 – 3.43 (m, 1H), 2.35 – 2.14 (m, 3H), 2.14 – 0.88 (m, 43H),
0.69 (s, 3H). 13C NMR (101 MHz, CDCl3) δ 175.8, 140.9, 121.7, 71.9, 56.5, 53.0, 50.2, 48.0, 45.1,
42.5, 42.4, 39.7, 37.4, 36.6, 33.4, 33.3, 32.1, 32.0, 31.8, 27.6, 25.7, 25.0, 25.0, 24.5, 21.2, 19.5, 17.8,
12.3. HRMS: e/z calculated for C28H45NO2 427.3450, found 427.3438 (2.9 ppm).
2.3.12. Fernholtz pentafluoropropan amide (14)
Flash column chromatography on silica (50% Heptane/EtOAc) gave 123mg of the desired white solid
1
(45%). Rf: 0.38 (EtOAc/heptane 1:1). [α]20 = -39.4 (1.27 mg/mL, CHCl3). H NMR (400 MHz,
D
Methanol-d4) δ 5.39 – 5.30 (m, 1H), 4.04 – 3.81 (m, 2H), 3.47 – 3.34 (m, 1H), 2.45 – 2.32 (m, 1H),
13
2.30 – 2.12 (m, 2H), 2.09 – 0.81 (m, 26H), 0.76 (s, 3H). C NMR (101 MHz, Methanol-d4) δ 180.1,
142.2, 122.3, 120.3 (qt, J = 285.8, 36.0 Hz), 114.6 (tq, J = 253.2, 36.0 Hz), 72.4, 57.8, 53.9, 51.7, 44.9,

43.5, 43.0, 41.0, 39.1 (t, J = 24.5 Hz), 38.6, 37.7, 33.3, 33.0, 32.3, 28.2, 25.4, 22.2, 19.9, 17.9, 12.5.
HRMS: e/z calculated for C25H36F5NO2 477.2666, found 477.2658 (1.6 ppm).
2.3.13. Fernholtz glycine ethyl ester amide (15)
Flash column chromatography on silica using 1:1 EtOAc/heptane as eluent yielded 140 mg of the
1
purified white solid (56%). Rf: 0.42 (EtOAc/heptane 8:2). [α]20 = -46.2 (1.17 mg/mL, CHCl3). H
D
NMR (300 MHz, CDCl3) δ 5.96 (t, J = 5.2 Hz, 1H), 5.41 – 5.27 (m, 1H), 4.20 (q, J = 7.1 Hz, 2H), 3.99
(d, J = 5.2 Hz, 2H), 3.61 – 3.41 (m, 1H), 2.36 – 2.09 (m, 3H), 2.09 – 0.75 (m, 28H), 0.69 (s, 3H).13C
NMR (101 MHz, CDCl3) δ 176.9, 170.3, 140.9, 121.6, 71.8, 61.6, 56.5, 52.9, 50.2, 44.7, 42.5, 42.4,
41.3, 39.7, 37.4, 36.6, 32.0, 31.9, 31.8, 27.7, 24.5, 21.2, 19.5, 17.7, 14.3, 12.2. HRMS: e/z calculated
for C26H41NO4 431.3036, found 431.3032 (0.7 ppm).
2.3.14. Fernholtz dodecyl amide (16)
Flash column chromatography using 1:1 EtOAc/heptane as eluent yielded 119 mg (40%) of the
1
purified cream white solid. Rf: 0.33 (EtOAc/heptane 1:1). [α]20 = -41.4 (1.11 mg/mL, CHCl3). H
D
NMR (400 MHz, CDCl3) δ 5.41 (t, J = 5.8 Hz, 1H), 5.37 – 5.29 (m, 1H), 3.57 – 3.45 (m, 1H), 3.31 –
3.03 (m, 2H), 2.34 – 2.15 (m, 2H), 2.14 – 1.37 (m, 17H), 1.37 – 1.20 (m, 21H), 1.17 (d, J = 6.7 Hz,
13
3H), 1.13 – 1.02 (m, 3H), 1.00 (s, 3H), 0.97 – 0.91 (m, 1H), 0.91 – 0.82 (m, 3H), 0.68 (s, 3H). C
NMR (101 MHz, CDCl3) δ 176.6, 140.9, 121.6, 71.8, 56.5, 53.0, 50.2, 45.2, 42.4, 42.4, 39.7, 39.4,
37.4, 36.6, 32.1, 31.9, 31.8, 29.8, 29.8, 29.8, 29.7, 29.7, 29.5, 29.4, 27.7, 27.1, 24.5, 22.8, 21.2, 19.5,
17.8, 14.3, 12.2. HRMS: e/z calculated for C34H59NO2 513.4546, found 513.4550 (-0.7 ppm).
2.3.15. Fernholtz benzyl amide (17)
The resulting crude product was dry loaded on silica and purified by flash column chromatography on
silica using EtOAc/heptane 1:1 as eluent yielding 164 mg of white solid (65%). Rf: 0.57 (50%
1
EtOAc/heptane). [α]20 = -57.4 (0.94 mg/mL, CHCl3). H NMR (300 MHz, CDCl3) δ 7.40 – 7.20 (m,
D
5H), 5.73 (t, J = 5.8 Hz, 1H), 5.40 – 5.28 (m, 1H), 4.52 – 4.31 (m, 2H), 3.59 – 3.44 (m, 1H), 2.37 –
13
0.73 (m, 31H), 0.68 (s, 3H). C NMR (75 MHz, CDCl3) δ 176.5, 140.9, 138.7, 128.8, 128.0, 127.6,
121.6, 71.8, 56.4, 53.0, 50.2, 45.1, 43.5, 42.5, 42.4, 39.7, 37.4, 36.6, 32.0, 31.9, 31.8, 27.8, 24.5, 21.2,
19.5, 17.9, 12.2. HRMS: e/z calculated for C29H41NO2 435.3137, found 435.3137 (0.2 ppm).
2.3.16. Fernholtz picoyl amide (18)
Flash column chromatography using EtOAc/heptane (8:2-10:0) as eluent yielding 127mg of the
1
desired white solid (50%). Rf: 0.09 (EtOAc/heptane 8:2). [α]20 = -44.0 (1.00 mg/mL, CHCl3). H
D
NMR (300 MHz, CDCl3) δ 8.60 – 8.48 (m, 1H), 7.65 (td, J = 7.8, 1.8 Hz, 1H), 7.26 (d, J = 7.8 Hz,
1H), 7.23 – 7.13 (m, 1H), 6.67 (t, J = 5.0 Hz, 1H), 5.39 – 5.28 (m, 1H), 4.53 (d, J = 5.0 Hz, 2H), 3.60
– 3.44 (m, 1H), 2.36 – 2.12 (m, 3H), 2.11 – 0.76 (m, 25H), 0.71 (s, 3H). 13C NMR (101 MHz, CDCl3)
δ 176.8, 156.7, 149.1, 140.9, 136.9, 122.5, 122.4, 121.7, 71.9, 56.5, 53.0, 50.2, 45.0, 44.4, 42.5, 42.5,
39.7, 37.4, 36.6, 32.1, 32.0, 31.8, 27.7, 24.5, 21.2, 19.5, 17.8, 12.2. HRMS: e/z calculated for
C28H40N2O2 436.3090, found 436.3078 (2.7 ppm).

2.4. Culturing of human myotubes
Satellite cells were isolated as previously described[37] from the M. obliquus internus abdominis of 6
healthy donors, age 39.9 (± 2.9) years, body mass index 23.5 (± 1.4) kg/m2, fasting glucose 5.3 (± 0.2)
mM, insulin, plasma lipids and blood pressure within normal range and no family history of diabetes.
The muscle biopsies were obtained with informed consent and approval by the National Committee
for Research Ethics, Oslo, Norway. The cells were cultured in DMEM-Glutamax™ (5.5 mM glucose),
2 % foetal bovine serum, 2 % Ultroser G, penicillin (100 units/ml), streptomycin (100 µg/ml), and
amphotericin B (1.25 µg/ml) for proliferation. At 70-80 % confluence the growth medium was
replaced by DMEM-Glutamax™ (5.5 mM glucose) supplemented with 2 % foetal bovine serum,
penicillin (100 units/ml), streptomycin (100 µg/ml), amphotericin B (1.25 µg/ml), and insulin (25 pM)
to induce differentiation. The cells were cultured in humidified 5 % CO2 atmosphere at 37°C, and the
medium was changed every 2–3 days. Experiments were performed after 7 days of differentiation.
2.5. Culturing of HepG2 cells
The human hepatoblastoma cell line HepG2 (HB-8065, ATCC, Manassas, VA, USA) was cultured in
DMEM-Glutamax™ (5.5 mM glucose) supplemented with 10 % foetal bovine serum, streptomycin
(100 µg/ml) and penicillin (100 units/ml) at 37°C in 5 % CO2.
2.6. RNA isolation and analysis of gene expression by TaqMan® real-time qPCR
Myotubes were treated with DMSO (0.1 %), 1 µM T0901317 (2), 10 µM 22SHC (8), 22RHC (1) and
compounds 10 - 18 for 4 days, harvested and total RNA was isolated by Agilent Total RNA isolation
kit (Agilent Technologies, Santa Clara, CA, USA) according to the supplier’s total RNA isolation
protocol. Total RNA (1 µg/µl) was reversely transcribed with hexamere primers using a Perkin-Elmer
Thermal Cycler 9600 (25C for 10 min, 37C for 1 h, 99C for 5 min) and a TaqMan reverse-
transcription reagents kit (Applied Biosystems). DNA expression was determined by SYBR® Green
(Applied Biosystems). Primers (36B4, ABCA1, FASN, GAPDH and SCD1) were designed using
Primer Express® (Applied Biosystems). Primer sequences are available upon request. Each target gene
were quantified in duplicates and carried out in a 25 µl reaction volume according to the supplier’s
protocol. All assays were run for 40 cycles (95C for 12 s followed by 60C for 60 s). The
transcription levels were normalized to the reference control genes 36B4 and GAPDH.
2.7. De novo lipogenesis
Cells were treated with DMSO (0.1 %), 1 µM T0901317 (2), 10 µM 22SHC (8), 22RHC (1) and
compounds 4, 10 – 18 for 4 days for myotubes and 24 h for HepG2 cells, before exposure to DMEM
supplemented with [1-14C]acetic acid (1 µCi/ml, 100 µM) for 24 h for myotubes and 4 h for HepG2
cells. Thereafter, cells were harvested in 0.1 M NaOH, assayed for protein[38] and total lipids were
isolated by filtration of the cell lysate through a hydrophobic MultiScreen® HTS plate (Millipore,
Billerica, MA, USA). The levels of lipids were determined by scintillation counting, and lipogenesis
from acetate was calculated by use of protein levels for standardization.

2.8. Glucose uptake
Myotubes were treated with DMSO (0.1 %), 1 µM T0901317 (2), 10 µM 22SHC (8), 22RHC (1) and
compounds 10 and 11 for 4 days. Thereafter, cells were exposed to D-[14C(U)]deoxy-D-glucose (1
µCi/ml, 100 µM) for 4 h. After incubation the cells were washed two times with ice-cold phosphate
buffered saline (PBS), lysed in 0.1 M NaOH, and radioactivity was counted by scintillation counting.
The protein content of each sample was determined[38] and glucose uptake was calculated using
protein levels for standardization.
2.9. Presentation of data and statistical analysis
Data in text and figures are given as mean (± SEM) from n = number of separate experiments. At least
3 parallels were included in each experiment. Comparisons of different treatments were evaluated by
two-tailed, paired Student’s t-test, and P < 0.05 was considered significant.
3. Results and discussion
3.1. Molecular modelling
Currently, a number of LXRα and –β crystal structures have been resolved and are accessible within
protein databanks, all of which are co-crystallized with either a steroidal- or non-steroidal LXR-
agonist.[22] In the present work, molecular modeling experiments were performed on LXRβ (PDB
Fig 2. 22-ketocholesterol (10) docked into the ligand binding domain of LXRβ (PDB 1P8D) where relevant amino acidic side chains are
depicted. The left side shows a polar cavity where the 3rd positioned secondary alcohol of 22KC (10) forms a hydrogen bond (shown in green)
with Glu281. Oxygen atoms are colored in red and carbon atoms of the ligand in yellow. Carbon atoms for adjacent amino acids in grey and
nitrogen atoms blue. The ligand surface (mesh) is displayed in colors by binding property. White – neutral surface, green - hydrophobic, blue –
hydrogen bond acceptor potential and red – hydrogen bond donor potential.
code: 1P8D)[23] complexed with the endogenous oxysterol agonist 24(S),25-epoxycholesterol (eHC)
within the LBD. The rationale for this choice relates to the structural similarity between eHC and the
compounds synthesized in the present work. The molecular structures of 22KC (10) and 22-Fernholtz-
22-isobutylamide (11) along with 22RHC (1) and 22SHC (8) were docked into the LXRβ structure
after removal of 24(S),25-epoxycholesterol. Further, the amide derivatives 4 and 11-18 were plotted
and docked as well. All docking scores are available in table S-1 in the supporting information. The
docking scores included are the best score obtained during three docking simulations. The Fernholtz-

benzyl amide (17) (see table 1) resulted in highest score, most likely as result of π-π interactions
between the terminal aromatic side chain of the synthetic steroid side chain and aromatic moieties of
Trp457/His435 which are currently considered as an essential “switch” activating the receptor upon
LXR-agonism.[23] Since the two epimers 22RHC (1) and 22SHC (8) promote gene expression
oppositely for certain genes, a special interest relates to the performance of 22KC (10) in vitro and in
silico. Fig. 2 shows 22KC (10) in the LBD of LXRβ. 22KC (10) scored better in silico than both
22RHC (1) and 22SHC (8).
3.2. Synthesis of new modulators
Aldehyde 6 was synthesized from Fernholtz acid 3 (scheme 1) in up to 67% yield over 4 steps in
accordance to described methods[36] (Synthetic procedures, NMR spectra and experimental data for
intermediate compounds of this route (19-21) are given in supporting information). Aldehyde 6
(regardless of hydroxyl protection in the 3rd position) has been utilized for synthesis of multiple steroid
compounds for various purposes.[39-42] Our recently developed method transforming Fernholtz acid
3 to its corresponding Weinreb amide (4) proved favorable.[34, 43] Consequent alcohol protection of
position 3 to form 5 and DIBALH reduction gave aldehyde 6 in 50% overall yield over 3 steps. TBS-
protected 22SHC (7) was formed under standard Grignard conditions using isopentylmagnesium
bromide followed by TBAF mediated TBS-deprotection which afforded 22SHC (8) in moderate
yields. Dess-Martin oxidation of 7 followed by TBAF mediated TBS-deprotection of 9 gave 22KC
Scheme 1: Synthesis outline and reaction conditions a) MeNHOMe∙HCl, HATU, HOBt, DIPEA, DMF, rt, overnight, 77% b) TBS-OTf, 2.6-
lutidine, DCM, -40 °C, 78% c) DIBALH, THF, -78 °C - rt, 83 % d) Isopentyl-MgBr, THF, -78 °C,52% e) TBAF, THF, rt , 67% f) DMP,
DCM, rt, 78% g) TBAF, THF, rt , 70% h) Isopentyl-MgBr, THF, -0 °C - rt, 29%
(10) in moderate yields as well. TBS-protected 22KC (9) was also formed directly from the TBS-
protected Weinreb amide (5). However, this method (provided in supporting information) only
afforded 29% yield using 10 molar equivalents of the Grignard reagent at relatively high temperature
(0 °C).

Amide derivatives of Fernholtz acid (FA, 3) were prepared by nucleophilic substitution on the
activated ester formed upon reaction between FA (3) and HBTU in DMF. The poor solubility of FA
(3) in common organic solvents including DMF was greatly enhanced upon instant transformation to
activated benzotriazole ester. By this general method, nucleophilic substitutions were carried out using
different amine nucleophiles to afford amides 11-18. Structures and yields of isolated products are
given in table 1.
Table 1: Synthesis of LXR-amides 11-18.
Product
Yield†
Product
Yield†
75%
56%
82%
65%
45%
40%
65%
50%
†Isolated yields from non-optimized reactions
3.3. In vitro evaluations
Previous work in human skeletal muscle cells has shown that 22SHC (8) acts as an antagonist on
lipid metabolism, while its epimer 22RHC (1) acts as a weak agonist on the same metabolic
pathways.[44] In this work, we have studied expression of genes in human skeletal muscle cells and
hepatocytes (HepG2 cells) important for de novo lipogenesis (FASN and SCD1) and reverse
cholesterol transport (ABCA1) (fig. 3).
The results confirm that 22SHC (8) acts as an antagonist by downregulating genes important for
lipogenesis (fig. 3B and C) while it does not affect the expression of another major LXR target gene
ABCA1 (fig. 2A).[6, 44] From a physiological point of view, it is important that ABCA1 is not
reduced since this could affect reverse cholesterol transport and ultimately increase the risk of
atherosclerosis in vivo.[45] However, the mechanism for the different regulation of FAS and SCD1
versus ABCA1 by 22SHC (8) is still unknown. A possible explanation offered is that different co-
regulators are involved in regulating the activity of LXR in the regulation of genes involved in
different metabolic pathways, but this requires further investigation to unravel. We observe that the
effect of the potent LXR agonist T0901317 (2) on SCD1 and FAS is abolished by 22SHC (8) as
expected by an antagonist, but this is not observed for the expression of ABCA1, supporting the theory

that 22SHC (8) does not regulate gene expression exclusively through activation of LXR. The
reduction of de novo lipogenesis by 22SHC (8) confirmed the functional relevance of the
downregulation of FAS and SCD1 in HepG2-cells (fig. 3D and E). The epimer 22RHC (1) on the
other hand, up-regulates all three LXR target genes tested, but only partly reduces the effect of
T0901317 (2) on the expression of FAS and SCD1 (fig. 3A-C). This can be explained by that 22RHC
(1) is a weaker agonist than T0901317 (2), and therefore will the expression of the target gene be
reduced when 22RHC (1) partly replaces T0901317 (2) through competition for the same seat on the
LXR receptor. However, 22RHC (1) does not seem to up-regulate de novo lipogenesis as expected by
the up-regulation of FAS and SCD1 (fig. 3D). To further explore the chemical functionality of the 22nd
position, we also tested the effects of 22KC (10) and Fernholtz-isobutyl amide (11) on the same
A
B
C
40
30
20
100
12
Basal
T0901317
*
*
80
60
40
20
*
*
#
8
4
0
10
8
*
3
2
1
0
6
#
#
4
#
*
*
#
*
2
*
*
*
0
Control 22SHC 22RHC
10
11
Control 22SHC 22RHC
10
11
Control 22SHC 22RHC
10
11
D
E
F
Basal
T0901317
2.0
1.5
1.0
0.5
0.0
1.5
1.0
0.5
0.0
2.0
1.5
1.0
0.5
0.0
*
#
*
*
*
#
*
*
*
*
22SHC
10
+ T0901317
+ T0901317
Control 22SHC 22RHC
10
11
0
10-8
22SHC (M)
10-7
10-6
10-5
0
10-8 10-7 10-6 10-5 10-4
10 (M)
Fig. 3. Effects of 22SHC (8) and analogs on basal and T0901317-induced gene expression in skeletal muscle cells and de novo lipogenesis in HepG2
cells. Human skeletal muscle and HepG2 cells were treated for 4 days and 24h, respectively, with DMSO (0.1%) control, 1 µM T0901317 (2) and 10
µM 22SHC (8), 22RHC (1), 22KC (10), Fernholtz-isobutyl amide (11)  1 µM T0901317 (2). Total RNA was then isolated from human skeletal
muscle cells and analyzed by qPCR as described in experimental procedures. Gene expressions were normalized to 36B4. For lipogenesis, HepG2
cells were incubated with 1-14Cacetate (1 µCi/ml, 100 µM) for 4 h before lipids were isolated by filtration through hydrophobic MultiScreen® HTS
plate. The levels of lipids were determined by scintillation counting. Values represent fold change relative to control given as means ± SEM (n=3-6).
Analyzed LXR target genes were (A) ATP-binding cassette transporter A1 (ABCA1), (B) fatty acid synthase (FASN) and (C) stearoyl-CoA desaturase
1 (SCD1), while (D) represent lipogenesis in HepG2 cells, (E) dose-response for 22SHC (8)  T0901317 and (F) dose-response for 22KC (10) 
T0901317 (2). *P < 0.05 vs. control and #P < 0.05 for T0901317 (2) vs. treatment + T0901317 (2).
parameters. 22KC (10) up-regulated the ABCA1 expression (fig. 3A) as 22RHC (1) while behaved
similar as 22SHC (8) on the FAS gene expression. 22KC (10) reduced the expression of FAS and
abolished the T0901317-effect significantly (fig. 3B). The T0901317-effect was also abolished by
22KC (10) on the expression of SCD1, but slightly increased the expression of SCD1 from basal.
However, it does not seem like 22KC (10) or Fernholtz-isobutyl amide (11) have any effects on de
novo lipogenesis as 22SHC (8) (fig. 3D and F).
The upregulation of ABCA1 by 22KC (10) shown in Fig. 3A, where expression of the other studied
LXR-target genes mostly remains uninfluenced (Fig. 3B and C), point towards an attractive in vitro
profile. Further investigation of 22KC (10) in terms of anti-atherosclerotic properties should be
performed as the compound could bias unwanted side effects provided by potent LXR-agonists, such
as T0901317 (2) which upregulates both FAS, SCD1 and lipogenesis potently.

Previously, we have also seen an increased glucose uptake in skeletal muscle cells after treatment
with 22SHC (8),[6] but this was not observed for neither 22RHC (1), 22KC (10) nor Fernholtz-22-
isobutyl amide (11) (data not shown).
To further explore the potential of preserving the 22-oxo-23-aza group with several different side
chains (Table 1), compounds 4 and 12-18 were screened for effects on de novo lipogenesis in HepG2
cells with 22SHC (8) and T0901317 (2) as control compounds (fig. 4A). As expected, T0901317 (2)
increased de novo lipogenesis while 22SHC (8) decreased the synthesis and abolished the T0901317-
effect. Fernholtz-cyclohexyl amide (13), seemed to have similar effect as 22SHC (8) on both reducing
de novo lipogenesis and abolish the T0901317-upregulation. None of the other compounds showed
any effect. The dose-response curve of Fernholtz-cyclohexyl amide (13) on lipogenesis in HepG2 cells
confirms that the compound indeed is able to suppress the lipogenesis in a dose-dependent manner
(fig. 4B).
To confirm that the effect of Fernholtz-cyclohexyl amide (13) on lipogenesis most likely is
regulated through LXR, changes in expression of ABCA1, FAS and SCD1 were studied (fig. 4C-E).
The effects of the control substances 22SHC (8) and T0901317 (2) were as expected, while Fernholtz-
N-butyl amide (12) 12 and Fernholtz-cyclohexyl amide (13) counteracted the effect of T0901317 (2)
significantly on FAS, an indication that the mode of action on reduction of lipogenesis is LXR
regulated. None of the tested compounds reduced the basal gene expression of neither FAS nor SCD1
A
3
B
1.2
Basal
T0901317
*
*
1.0
0.8
0.6
*
2
1
0
*
*
#
*
#
*
0
Control 22SHC
4
12
13
14
15
16
17
18
10-8
10-7
10-6
10-5
10-4
Compound 13 (M)
C
D
E
12
40
80
60
40
20
*
Basal
T0901317
*
30
20
*
8
4
0
8
6
4
2
0
#
10
8
6
4
2
#
#
#
#
#
*
0
Control 22SHC
12
13
14
Control 22SHC
12
13
14
Control 22SHC
12
13
14
Fig. 4. De novo lipogenesis and effects of 22SHC (8) and amide analogs on basal and T0901317-induced gene expression. Human skeletal muscle
and HepG2 cells were treated for 4 days and 24h, respectively, with DMSO (0.1%) control, 1 µM T0901317 (2) and 10 µM 22SHC (8), 4, 12 - 18  1
µM T0901317 (2). Total RNA was then isolated from human skeletal muscle cells and analyzed by qPCR as described in experimental procedures. For
lipogenesis, HepG2 cells were incubated with 1-14Cacetate (1 µCi/ml, 100 µM) for 4 h before lipids were isolated by filtration through hydrophobic
MultiScreen® HTS plate. The levels of lipids were determined by scintillation counting. Values represent fold change relative to control for total lipid
synthesis given as means ± SEM n=4 separate experiments from (A) lipogenesis (B) dose-response for Fernholtz-cyclohexyl amide (13). Analyzed
LXR target genes were (C) ATP-binding cassette transporter A1 (ABCA1), (D) fatty acid synthase (FASN) and (E) stearoyl-CoA desaturase 1 (SCD1).
Values represent fold change relative to control for gene expression is given as means ± SEM (n=3-6). Gene expressions were normalized to 36B4. *P
< 0.05 vs. control and #P < 0.05 for T0901317 (2) vs. treatment + T0901317 (2).

(fig. 4B-C).
4. Conclusions
In summary, we have shown that LXR-expression of key target genes involved in human lipid
metabolism is affected differently depending on the stereochemistry and functionality of the 22nd
position on the cholesterol scaffold. Moreover, 22/23 oxo-aza amide compounds 4 and 11-18 were
evaluated in silico and in vitro as well. Several of the ligands tested in silico performed well in terms
of docking scores (table S-1, supporting information) using the LXRβ complex (PDB code:
1P8D).[23] New synthetic methods were established to provide new shorter and robust pathway to
TBS-protected aldehyde 6, a common intermediate for a large variety of different steroidal structures.
In addition, one step method for the preparation of various Fernholtz amides 4 and 11-18 was
established.
In vitro experiments were conducted in human hepatocytes and skeletal muscle cells. The
experiments showed 22SHC (8) to downregulate gene expression of FAS, SCD1 and de novo
lipogenesis. The endogenous agonist 22RHC (1) upregulated expression of all target genes tested.
Interestingly, 22KC (10) behaved as 22RHC (1) (an agonist) in terms of ABCA1 upregulation. This is
important as reduced expression of ABCA1 would affect reverse cholesterol transport and ultimately
increase the risk for development of atherosclerosis. Moreover, the 22KC (10) had only minor
influence on gene expression of FAS, SCD1 and no effect on lipogenesis. The in vitro profile of 22KC
(10) presented in this work therefore suggests that 22KC (10) should be investigated further in terms
of anti-atherosclerotic features as unwanted side-effects provided by potent agonists like T0901317 (2)
could be biased. Moreover, agonistic effects of T0901317 (2) were mostly abolished by 22KC (10).
Of the new amides tested in vitro (4 and 11-18), Fernholtz-cyclohexyl amide (13) downregulated de
novo lipogenesis. No other Fernholtz amides were shown to possess biological activity in terms of
LXR-modulation.
Acknowledgements
The work with the project and this paper was supported by Oslo University Hospital Innovation
Grant. School of Pharmacy, University of Oslo is gratefully acknowledged for Ph.D.-scholarships to
EÖV and OAHÅ. We thank Rasha S. Haseeb for contributions to synthetic experiments. The in silico
studies were allocated national E-infrastructure resources (project NN2987k).
References
[1] E. Viennois, K. Mouzat, J. Dufour, L. Morel, J.-M. Lobaccaro, S. Baron, Selective liver X receptor
modulators (SLiMs): What use in human health?, Molecular and Cellular Endocrinology 351 (2012)
129-141.
[2] T. Jakobsson, E. Treuter, J.-Å. Gustafsson, K.R. Steffensen, Liver X receptor biology and
pharmacology: new pathways, challenges and opportunities, Trends in Pharmacological Sciences
33(7) (2012) 394-404.
[3] B.A. Janowski, P.J. Willy, T.R. Devi, J.R. Falck, D.J. Mangelsdorf, An oxysterol signalling
pathway mediated by the nuclear receptor LXR[alpha], Nature 383(6602) (1996) 728-731.
[4] C. Hong, P. Tontonoz, Liver X receptors in lipid metabolism: opportunities for drug discovery, Nat
Rev Drug Discov 13(6) (2014) 433-444.

[5] B.A. Janowski, M.J. Grogan, S.A. Jones, G.B. Wisely, S.A. Kliewer, E.J. Corey, D.J. Mangelsdorf,
Structural requirements of ligands for the oxysterol liver X receptors LXRα and LXRβ, Proceedings of
the National Academy of Sciences 96(1) (1999) 266-271.
[6] E.T. Kase, G.H. Thoresen, S. Westerlund, K. Højlund, A.C. Rustan, M. Gaster, Liver X receptor
antagonist reduces lipid formation and increases glucose metabolism in myotubes from lean, obese
and type 2 diabetic individuals, Diabetologia 50(10) (2007) 2171-2180.
[7] A.C. Calkin, P. Tontonoz, Transcriptional integration of metabolism by the nuclear sterol-activated
receptors LXR and FXR, Nat Rev Mol Cell Biol 13(4) (2012) 213-224.
[8] N. Mitro, P.A. Mak, L. Vargas, C. Godio, E. Hampton, V. Molteni, A. Kreusch, E. Saez, The
nuclear receptor LXR is a glucose sensor, Nature 445(7124) (2007) 219-223.
[9] D.J. Peet, S.D. Turley, W. Ma, B.A. Janowski, J.-M.A. Lobaccaro, R.E. Hammer, D.J.
Mangelsdorf, Cholesterol and Bile Acid Metabolism Are Impaired in Mice Lacking the Nuclear
Oxysterol Receptor LXRα, Cell 93(5) (1998) 693-704.
[10] E. Viennois, A.J.C. Pommier, K. Mouzat, A. Oumeddour, F.-Z. El Hajjaji, J. Dufour, F. Caira,
D.H. Volle, S. Baron, J.-M.A. Lobaccaro, Targeting liver X receptors in human health: Deadlock or
promising trail?, Expert Opinion on Therapeutic Targets 15(2) (2011) 219-232.
[11] H. de Boussac, A.J.C. Pommier, J. Dufour, A. Trousson, F. Caira, D.H. Volle, S. Baron, J.-M.A.
Lobaccaro, LXR, prostate cancer and cholesterol: the Good, the Bad and the Ugly, American Journal
of Cancer Research 3(1) (2013) 58-69.
[12] A.C. Calkin, P. Tontonoz, Liver X Receptor Signaling Pathways and Atherosclerosis,
Arteriosclerosis, Thrombosis, and Vascular Biology 30(8) (2010) 1513-1518.
[13] N.P. Hessvik, S.S. Bakke, R. Smith, A.W. Ravna, I. Sylte, A.C. Rustan, G.H. Thoresen, E.T.
Kase, The liver X receptor modulator 22(S)-hydroxycholesterol exerts cell-type specific effects on
lipid and glucose metabolism, J. Steroid Biochem. Mol. Biol. 128 (2012) 154-164.
[14] E.T. Kase, N. Nikolic, N.P. Hessvik, A.K. Fjeldheim, J. Jensen, G.H. Thoresen, A.C. Rustan,
Dietary Supplementation with 22-S-Hydroxycholesterol to Rats Reduces Body Weight Gain and the
Accumulation of Liver Triacylglycerol, Lipids 47(5) (2012) 483-493.
[15] A.J.C. Pommier, G. Alves, E. Viennois, S. Bernard, Y. Communal, B. Sion, G. Marceau, C.
Damon, K. Mouzat, F. Caira, S. Baron, J.M.A. Lobaccaro, Liver X Receptor activation downregulates
AKT survival signaling in lipid rafts and induces apoptosis of prostate cancer cells, Oncogene 29(18)
(2010) 2712-2723.
[16] C.-Y. Lin, J.-A. Gustafsson, Targeting liver X receptors in cancer therapeutics, Nat Rev Cancer
15(4) (2015) 216-224.
[17] A. Castrillo, P. Tontonoz, Nuclear receptors in macrophage biology: At the Crossroads of Lipid
Metabolism and Inflammation, Annual Review of Cell and Developmental Biology 20(1) (2004) 455-
480.
[18] A.J. Fowler, M.Y. Sheu, M. Schmuth, J. Kao, J.W. Fluhr, L. Rhein, J.L. Collins, T.M. Willson,
D.J. Mangelsdorf, P.M. Elias, K.R. Feingold, Liver X receptor activators display anti-inflammatory
activity in irritant and allergic contact dermatitis models: Liver-X-receptor-specific inhibition of
inflammation and primary cytokine production, J Invest Dermatol 120(2) (2003) 246-255.
[19] Q. Jiang, C.Y.D. Lee, S. Mandrekar, B. Wilkinson, P. Cramer, N. Zelcer, K. Mann, B. Lamb,
T.M. Willson, J.L. Collins, J.C. Richardson, J.D. Smith, T.A. Comery, D. Riddell, D.M. Holtzman, P.
Tontonoz, G.E. Landreth, ApoE Promotes the Proteolytic Degradation of Aβ, Neuron 58(5) (2008)
681-693.
[20] D.R. Riddell, D.J. O'Neill, Chapter 24 Targeting ApoE in Alzheimer's Disease: Liver X Receptor
Agonists as Potential Therapeutics, Emerging Drugs and Targets for Alzheimer's Disease: Volume 2:
Neuronal Plasticity, The Royal Society of Chemistry2010, pp. 191-212.

[21] P.J. Willy, K. Umesono, E.S. Ong, R.M. Evans, R.A. Heyman, D.J. Mangelsdorf, LXR, a nuclear
receptor that defines a distinct retinoid response pathway, Genes & Development 9(9) (1995) 1033-
1045.
[22] C.M. Tice, P.B. Noto, K.Y. Fan, L. Zhuang, D.S. Lala, S.B. Singh, The Medicinal Chemistry of
Liver X Receptor (LXR) Modulators, Journal of Medicinal Chemistry 57(17) (2014) 7182-7205.
[23] S. Williams, R.K. Bledsoe, J.L. Collins, S. Boggs, M.H. Lambert, A.B. Miller, J. Moore, D.D.
McKee, L. Moore, J. Nichols, D. Parks, M. Watson, B. Wisely, T.M. Willson, X-ray crystal structure
of the liver X receptor beta ligand binding domain: regulation by a histidine-tryptophan switch, J Biol
Chem 278(29) (2003) 27138-43.
[24] J. Loren, Z. Huang, B.A. Laffitte, V. Molteni, Liver X receptor modulators: a review of recently
patented compounds (2009 – 2012), Expert Opinion on Therapeutic Patents 23(10) (2013) 1317-1335.
[25] J. Beltowski, Liver X receptors (LXR) as therapeutic targets in dyslipidemia, Cardiovascular
Therapeutics 26(4) (2008) 297-316.
[26] K.W. Christopherson, II, A. Landay, Liver X receptor α (LXRα) as a therapeutic target in chronic
lymphocytic leukemia (CLL), Journal of Leukocyte Biology 86(5) (2009) 1019-1021.
[27] G. Schmitz, T. Langmann, Transcriptional regulatory networks in lipid metabolism control
ABCA1 expression, Biochimica et Biophysica Acta, Molecular and Cell Biology of Lipids 1735(1)
(2005) 1-19.
[28] N.P. Hessvik, S.S. Bakke, K. Fredriksson, M.V. Boekschoten, A. Fjoerkenstad, G. Koster, M.K.
Hesselink, S. Kersten, E.T. Kase, A.C. Rustan, G.H. Thoresen, Metabolic switching of human
myotubes is improved by n-3 fatty acids, J Lipid Res 51 (2010) 2090-2104.
[29] J.M. Lehmann, S.A. Kliewer, L.B. Moore, T.A. Smith-oliver, B.B. Oliver, J.-l. Su, S. Scott, D.A.
Winegar, E. Daniel, T.A. Spencer, M. Willson, M. Lehmann, S.S. Sundseth, D.E. Blanchard, T.M.
Willson, Activation of the Nuclear Receptor LXR by Oxysterols Defines a New Hormone Response
Pathway, J Biol Chem 272 (1997) 3137-3140.
[30] O.A.H. Åstrand, M. Sandberg, I. Sylte, C.H. Görbitz, G.H. Thoresen, E.T. Kase, P. Rongved,
Synthesis and initial biological evaluation of new mimics of the LXR-modulator 22(S)-
hydroxycholesterol, Bioorganic & Medicinal Chemistry 22(1) (2014) 643-650.
[31] Q. Khuong Huu, Y. Letourneux, M. Gut, R. Goutarel, Steroidal alkaloids. CLXI. Stereospecific
synthesis of (22R)- and (22S)-22-Aminocholesterol, Journal of Organic Chemistry 39(8) (1974) 1065-
8.
[32] G.J. Alsema, H.J. Degenhart, J. Hoogerbrugge, Formation of 22-keto-cholesterol from 22S-
hydroxy-cholesterol by a NADPH and O2 dependent, co-insensitive enzyme in bovine adrenal cortex
mitochondria, J Steroid Biochem 17(1) (1982) 37-40.
[33] J.D. Lambeth, 22-Ketocholesterol. A potent competitive inhibitor of cytochrome P-450scc-
dependent side-chain cleavage of cholesterol, Molecular Pharmacology 23(3) (1983) 743-7.
[34] O.A. Astrand, E.O. Viktorsson, A.L. Kristensen, D. Ekeberg, H. Roberg-Larsen, S.R. Wilson, M.
Gabrielsen, I. Sylte, A.C. Rustan, G.H. Thoresen, P. Rongved, E.T. Kase, Synthesis, in vitro and in
vivo biological evaluation of new oxysterols as modulators of the liver X receptors, J Steroid Biochem
Mol Biol, in press (2016).
[35] M.C. Neves, M. Totrov, R. Abagyan, Docking and scoring with ICM: the benchmarking results
and strategies for improvement, J. Comput. Aided Mol. Des. 26(6) (2012) 675-686.
[36] G. Guerlet, T. Spangenberg, A. Mann, H. Faure, M. Ruat, Synthesis and biological evaluation of
desmethylveramiline, a micromolar Hedgehog inhibitor, Bioorganic & Medicinal Chemistry Letters
21(12) (2011) 3608-3612.
[37] M. Gaster, S.R. Kristensen, H. Beck-Nielsen, H.D. SchrØDer, A cellular model system of
differentiated human myotubes, APMIS 109(11) (2001) 735-744.

[38] M.M. Bradford, A rapid and sensitive method for the quantitation of microgram quantities of
protein utilizing the principle of protein-dye binding, Anal Biochem 72 (1976) 248-54.
[39] K. Mikami, K. Kawamoto, T. Nakai, Stereocontrolled synthesis of either (22S)- or (22R)-
hydroxy-23-acetylenic steroid side chains via [2,3]-Wittig sigmatropic rearrangement, Tetrahedron
Letters 26(47) (1985) 5799-5802.
[40] H. Nagano, M. Matsuda, T. Yajima, Stereoselective synthesis of 24-alkyl-22-hydroxysterols
based on chelation-controlled radical reactions, Journal of the Chemical Society, Perkin Transactions 1
(2) (2001) 174-182.
[41] J.G. Cui, C.W. Lin, L.M. Zeng, J.Y. Su, Synthesis of polyhydroxysterols (III): synthesis and
structural elucidation of 24-methylenecholest-4-en-3β,6α-diol, Steroids 67(13–14) (2002) 1015-1019.
[42] L. Gros, S.O. Lorente, Jimenez, V. Yardley, L. Rattray, H. Wharton, S. Little, S.L. Croft, L.M.
Ruiz-Perez, D. Gonzalez-Pacanowska, I.H. Gilbert, Evaluation of Azasterols as Anti-Parasitics,
Journal of Medicinal Chemistry 49(20) (2006) 6094-6103.
[43] E.O. Viktorsson, O.A.H. Astrand, R.S. Haseeb, C.H. Gorbitz, P. Rongved, Crystal structure of
(S)-2-[(3S,8S,9S,10R,13S,14S,17R)-3-hydroxy-10,13-dimethyl-2,3,4,7,8,9,10,11,12,13,14,15,16,17-
tetradecahydro-1H-cyclopenta[a]phenanthren-17-yl]-N-methoxy-N-methylpropanamide (Fernholz
Weinreb amide), Acta Crystallographica Section E 71(3) (2015) 275-277.
[44] E. Tranheim Kase, B. Andersen, H.I. Nebb, A.C. Rustan, G. Hege Thoresen, 22-
Hydroxycholesterols regulate lipid metabolism differently than T0901317 in human myotubes,
Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of Lipids 1761(12) (2006) 1515-
1522.
[45] G.H. Tomkin, D. Owens, Investigational therapies for the treatment of atherosclerosis, Expert
Opinion on Investigational Drugs 23(10) (2014) 1411-1421.

Graphical abstract

Highlights for
Regulation of liver X receptor target genes by 22-functionalized
oxysterols. Synthesis, in silico and in vitro evaluations.
New version:





New oxysterol based LXR-modulators have been synthesized
Importance of the 22th position of the oxysterol scaffold has been investigated
Compounds were evaluated in silico and in vitro for regulation of key LXR-target genes
22-Ketocholesterol downregulates ABCA1 selectively
A new oxysterol analog was found to reduce lipogenesis
Old version:
From abstract:
The endogenous oxysterol 22(R)-hydroxycholesterol (22RHC) is an LXR agonist which upregulates genes of critical involvement in
human cholesterol- and lipid metabolism. In contrast, its synthetic epimer 22(S)-hydroxycholesterol (22SHC) has shown specific
antagonistic effects in recent studies, avoiding unwanted side effects provided by potent LXR agonists.
Highlights


We have synthesized different oxysterol LXR-modulators which differ in their functional moiety at the 22th position on
the steroidal scaffold.
We have also synthesized a new class of compounds including an amide functionality at the 22-23 position of the
cholesterol scaffold


All compounds have been evaluated in silico and moreover, tested for regulation of LXR target genes in vitro.
22-ketocholesterol which to our knowledge has not been evaluated in such studies before, shows an attractive profile to
investigate further in terms of anti-atherosclerotic properties as it upregulates the ATP-binding cassette transporter
ABCA1 selectively and might therefore bias unwanted side effects provided by classic LXR-full agonists as discussed
within the manuscript. Another compound of the new synthesized class, cyclohexylamide 13 was shown to reduce
lipogenesis in a dose responsive manner.