A genomic DNA reporter screen identifies squalene synthase inhibitors that act cooperatively with statins to upregulate the low-density lipoprotein receptor

Article abstract of DOI:10.1124/jpet.116.239574

Hypercholesterolemia remains one of the leading risk factors for the development of cardiovascular disease. Many large double-blind studies have demonstrated that lowering low-density lipoprotein (LDL) cholesterol using a statin can reduce the risk of having a cardiovascular event by approximately 30%. However, despite the success of statins, some patient populations are unable to lower their LDL cholesterol to meet the targeted lipid levels, due to compliance or potency issues. This is especially true for patients with heterozygous familial hypercholesterolemia who may require additional upregulation of the low-density lipoprotein receptor (LDLR) to reduce LDL cholesterol levels below those achievable with maximal dosing of statins. Here we identify a series of small molecules from a genomic DNA reporter screen that upregulate the LDLR in mouse and human liver cell lines at nanomolar potencies (EC₅₀ = 39 nM). Structure-activity relationship studies carried out on the lead compound, OX03771 [(E)-N,N-dimethyl-3-(4-styrylphenoxy)propan-1-amine], led to the identification of compound OX03050 [(E)-3-(4-styrylphenoxy)propan-1-ol], which had similar potency (EC₅₀ = 26 nM) but a much-improved pharmacokinetic profile and showed in vivo efficacy. Compounds OX03050 and OX03771 were found to inhibit squalene synthase, the first committed step in cholesterol biosynthesis. These squalene synthase inhibitors were shown to act cooperatively with statins to increase LDLR expression in vitro. Overall, we demonstrated here a novel series of small molecules with the potential to be further developed to treat patients either alone or in combination with statins.

Full text of DOI:10.1124/jpet.116.239574

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
A genomic DNA reporter screen identifies squalene synthase inhibitors which
act cooperatively with statins to upregulate the low-density lipoprotein receptor
Alastair G. Kerr, Lawrence C.S. Tam, Ashley B. Hale, Milena Cioroch, Gillian Douglas,
Sarina Agkatsev, Olivia Hibbitt, Joseph Mason, James Holt-Martyn, Carole J. R.
Bataille, Graham M. Wynne, Keith M. Channon, Angela J. Russell, and Richard Wade-
Martins
Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, UK;
(AK., LT., MC., SA., OH., JH-M., RW-M)
Division of Cardiovascular Medicine, British Heart Foundation Centre of Research
Excellence, University of Oxford, John Radcliffe Hospital, Oxford OX3 9DU, UK (AH.,
GD., KC)
Department of Chemistry, Chemistry Research Laboratory, University of Oxford,
Mansfield Road, Oxford, UK (JM., CB., GW., AR)
Department of Pharmacology, University of Oxford, Mansfield Road, Oxford, UK (AR)
1

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Novel small molecules which upregulate the LDLR
Joint corresponding authors
Correspondence should be addressed to Richard Wade-Martins, MA, DPhil, Le Gros Clark
Building, Department of Physiology, Anatomy and Genetics, University of Oxford, South Parks
Road, Oxford, OX1 3QX United Kingdom. E-mail: richard.wade-martins@dpag.ox.ac.uk
Telephone number: +441865 282837
Other corresponding author
Angela Russell, MChem, DPhil, Department of Chemistry, Chemistry Research Laboratory,
University of Oxford, Mansfield Road, Oxford OX1 3TA, UK; Department of Pharmacology,
University
of
Oxford,
Mansfield
Road,
Oxford
OX1
3QT,
UK.
Email:
angela.russell@chem.ox.ac.uk. Telephone number: +441865 275 643 (Chemistry) /271 883
(Pharmacology)
Text pages: 34
Number of tables: 0
Number of figures: 6
References: 32
Abstract word count: 225
Introduction word count: 649
Discussion word count: 1079
Section: Drug Discovery and Translational Medicine
Non-standard abbreviations – Familial Hypercholesterolaemia (FH), 3-hydroxy-3-
methylglutaryl coenzyme A reductase (HMGCR), Low-density Lipoprotein Receptor
(LDLR), Squalene Synthase (SQS)
2

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Abstract:
Hypercholesterolaemia remains one of the leading risk factors for the development of
cardiovascular disease. Many large double-blind studies have demonstrated that
lowering LDL-cholesterol using a statin can reduce the risk of having a cardiovascular
event by ~30%. However, despite the success of statins, some patient populations are
unable to lower their LDL-cholesterol to meet the targeted lipid levels, due to
compliance or potency issues. This is especially true for heterozygous familial
hypercholesterolaemia (heFH) patients who may require additional upregulation of the
Low-Density Lipoprotein Receptor (LDLR) to reduce LDL-cholesterol levels below
those achievable with maximal dosing of statins. Here we identify a series of small
molecules from a genomic DNA reporter screen which upregulate the LDLR in mouse
and human liver cell lines at nanomolar potencies (EC50: 39 nM). Structure-activity
relationship studies carried out on the lead compound (compound OX03771) led to
the identification of compound OX03050, which had similar potency (EC50: 26 nM), but
a much-improved pharmacokinetic profile and showed in vivo efficacy. Compound
OX03050 and OX03771 were found to inhibit squalene synthase, the first committed
step in cholesterol biosynthesis. These squalene synthase inhibitors were shown to
act cooperatively with statins to increase LDLR expression in vitro. Overall, we have
demonstrated here a novel series of small molecules with the potential to be further
developed to treat patients either alone, or in combination with statins.
3

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Introduction
Cardiovascular disease remains one of the largest health and economic burdens in
the world. High circulating levels of low-density lipoprotein cholesterol (LDL-
cholesterol) remains one of the biggest risk factors for cardiovascular disease, as it
leads to the accelerated development of atherosclerosis and progression into coronary
heart disease (Lusis 2000). The front-line pharmacotherapies to lower LDL-cholesterol
are the 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMGCR) inhibitors, known
as statins. Statins have been very successful in primary and secondary prevention of
coronary heart disease (Pedersen 1994; Shepherd et al. 1995), with many large
randomised control trials demonstrating a reduction in cardiovascular disease directly
correlated with lowering of LDL-cholesterol (Sniderman et al. 2012). Despite their
success some patients either fail to achieve their lipid goal or experience intolerable
side effects such as myotoxicity (Moßhammer et al. 2014), leading to them
discontinuing taking statin medication. However, there has been some controversy
regarding the rate of adverse events related to statins, with some large randomised
controlled trials describing no difference in reported side effects compared with
placebo (Ridker et al. 2008). The rate of myopathy in these trials is usually around 3%
(Bays 2006), although the rate of statin intolerance in the general population may be
as high as 10-15% (Grundy 2002; Joy and Hegele 2009). These side effects are much
more common in patients on the maximum dose of statins, so therapies which could
be used in conjunction with a reduced dose of statins to lower LDL-cholesterol, may
increase compliance and help more patients achieve their targeted lipid levels.
Squalene synthase is involved in the same cholesterol biosynthesis pathway as
HMGCR but represents the first committed step in cholesterol synthesis. Statins act
4

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
upstream of squalene synthase and so also inhibit the production of other non-
steroidal isoprenoid molecules, such as isopentenyl adenine (protein synthesis),
coenzyme Q10 (mitochondrial respiration) and dolichol (glycosylation), a consequence
that has been linked to their side effects (Bitzur et al. 2013) (Marcoff and Thompson
2007). Squalene synthase inhibitors have been shown to upregulate the low-density
lipoprotein receptor (LDLR) like statins, but without inhibiting the production of these
other intermediates (Davidson 2007). In fact squalene synthase inhibitors
administered with statins were shown to have a protective effect on statin-induced
myotoxicity (Nishimoto et al. 2007). Many squalene synthase inhibitors have shown
similar or greater anti-hyperlipidemic effects than statins (Ugawa et al.
2000)(Nishimoto et al. 2003) and they continue to be developed as promising lipid
lowering agents to complement statins (Ichikawa et al. 2013). The squalene synthase
inhibitor lapaquistat acetate reached phase II/III clinical trials (Stein et al. 2011) where
it was generally well tolerated and significantly reduced LDL-cholesterol. When used
in conjunction with statins, lapaquistat acetate reduced LDL-cholesterol by a further
19%, when compared with statin monotherapy. Unfortunately, lapaquistat acetate was
withdrawn due to hepatotoxicity seen in two patients receiving a high dose. It was
never established whether the hepatotoxicity was a result of off-target effects of
lapaquistat acetate, or through inhibition of squalene synthase. Interestingly, squalene
synthase knockout mice exhibit only a modest liver injury despite complete loss of
enzyme function, thought to result from a build-up of the squalene synthase substrate
farnesyl diphosphate (FPP) (Nagashima et al. 2015). Treatment with a statin would
alleviate this build up, similar to that already observed using this combination of
inhibitors to prevent non-sterol isoprenoid depletion associated with statins (Wasko et
al. 2011).
5

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Here we present a set of novel stilbenoid-derived small molecules which can
upregulate the LDLR at the mRNA and protein level with nanomolar potencies. We
carried out preliminary structure activity relationship studies to improve the efficacy
and pharmacokinetic profile. The mechanism of action was established through
enzymatic assays and supported by in silico modelling. We demonstrate that these
small molecules are able to drive expression of the human LDLR promoter in vivo and
when dosed in combination with statins give a much greater effect than can be seen
with either inhibitor alone.
6

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Methods
Cell culture
Human Hep3B cells were a kind gift from Dr Zoe Holloway. Mouse hepatoma Hepa
1–6 cells were a kind gift from Dr Natalia Sacilotto. Both cell lines were grown in
Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum
(FBS), 1% penicillin/ streptomycin, 1% L-glutamine, in a 5% CO2 incubator at 37oC.
For mRNA analysis Hep3B cells were seeded in 24-well plates. For Western blot
analysis Hepa1-6 cells were seeded in 6-well plates. After 24 h cells were changed to
Dulbecco’s modified Eagle’s medium supplemented with 5% lipoprotein deficient
serum (LDPS), 1% penicillin/ streptomycin, 1% L-glutamine.. CHO WT cells
transfected with pLDLR-Luc (CHO-pLDLR-Luc) establishing a clonal cell line, were
grown in HamsF-12 medium supplemented with 10% fetal bovine serum (FBS), 1%
penicillin/ streptomycin, 1% L-glutamine, in a 5% CO2 incubator at 37°C. CHO-pLDLR-
Luc were seeded in 24-well plates. After 24 h, cells were changed to HamsF-12
supplemented with 5% LDPS, 1% penicillin/ streptomycin and 1% L-glutamine.
compounds, Simvastatin (Sigma, St Louis, MO) and Pravastatin (Sigma) dissolved in
DMSO and Cholesterol (Sigma) and 25-Hydroxycholesterol (Sigma) dissolved in
ethanol were added 24 h after.
Luciferase assay
CHO-pLDLR-Luc cells were lysed 48 h after compound treatment using lysis buffer
containing 1% Triton X-100. 2 mM ATP, 2 mM DTT and 1 mM D-Luciferin were added
to the lysate in luciferase assay buffer containing 15 mM MgSO4, 15 mM KPO4 and
4 mM EGTA at pH 7.8. Luciferase activity was quantified on a Dynex Technologies
MLX 96 well plate luminometer (Dynex Technologies, Chantilly, VA). Protein
concentration was determined using a BCA Protein assay kit with Bovine Serum
7

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Albumin to generate a standard curve. Luciferase activity was normalized to total
protein within each well.
qRT-PCR
RNA was extracted from Hep3B cells 24 h after compound treatment, using RNeasy
mini kit (Qiagen, Valencia, CA) according to manufacturer's protocol. cDNA was
reverse transcribed from 1 µg of total RNA using random primers, SuperScript III
Reverse Transcriptase and RNaseOUT™ recombinant RNase inhibitor. Quantitative
Real-time PCR was performed on a StepOnePlus Real-Time PCR system, with SYBR
Green PCR Master Mix according to the manufacturer’s protocol. Gene and species
specific primers were used for LDLR: Forward gacagatgcgaaagaaacga, Reverse
acagacaagcacgtctcctg. Β-Actin:
Forward agcgcggctacagcttca, Reverse
cgtagcacagcttctccttaatgtc. Β-Actin was used as a housekeeper gene and all samples
were run in triplicates. ∆∆Ct was calculated and used for quantification.
Western blot analysis
Hepa1-6 cells were lysed 48 h after compound treatment using lysis buffer containing
0.5% NP40. Protein concentration was determined using a BCA Protein assay kit with
Bovine Serum Albumin to generate a standard curve. 15 µg total protein was heated
at 90ºC for 5 minutes then loaded into each well and run on a 10% SDS-
polyacrylamide. Samples were transferred to a PVDF membrane cut and stained with
LDLR (ab30532, Abcam, Cambridge, MA) or β-actin (ab8226, Abcam) primary
antibody overnight at 4ºC. Membranes were then incubated with horseradish
peroxidase-conjugated polyclonal rabbit IgG secondary antibody (Abcam). All blots
were developed using enhanced chemiluminescent substrate kit (Pierce, Thermo
Scientific, Waltham, MA) and exposed to Fugi X-ray films in a dark-room facility.
8

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Protein band intensities were quantified by scanning with Canon M550 Scanner and
analysed using ImageJ.
Adenylate kinase assay
Cell media was aspirated from wells of CHO-pLDLR-Luc cells treated with compounds
for 48 hours. Adenylate kinase concentration present in the media was quantified
using Bioluminescence Cytotoxicity Assay Kit (MBL, Woburn, MA) as per
manufactures instructions. Luciferase activity was quantified on a Dynex Technologies
MLX 96 well plate luminometer.
Squalene synthase activity assay
Squalene synthase activity was assessed as described earlier (Amin et al. 1992).
Briefly each assay was in 1 mL of assay buffer (50 mM phosphate buffer, pH 7.4
containing 10 mM MgCl2) containing 0.5 mM NADPH, 12 µg of human liver
microsomes and compound or vehicle (DMSO) alone in 16x100mm glass screw cap
tubes. All components were allowed to equilibrate for 10 minutes at 37oC before
addition of [3H]-FPP (50 nM 0.045 Ci/mmol) (Perkin Elmer, Wokingham, United
Kingdom) for a further 10 minutes at 37°C. The reaction was stopped by the addition
of 1 mL of 15% KOH dissolved in EtOH. Tubes were incubated at 65oC for 30 minutes
then 5 mL of petroleum ether was added and shaken for 10 minutes. The lower
aqueous phase was frozen and the upper organic phase transferred to clean glass
tubes containing 2 mL distilled water. 1.5 mL of the upper organic phase was removed
and counted with 3 mL of scintillation liquid on a Tri-Carb 2800 TR Liquid Scintillation
Analyzer.
9

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Animals and treatment protocols
Maintenance of animals and animal experiments were carried out in accordance with
the UK Home Office regulations under the Animals (Scientific Procedures) Act (ASPA)
of 1986. Animals had access to food and water ad libitum. All mice were fed a chow
diet (Global 16% Protein Rodent Diet, TD.2016, Harlan-Teklad) and maintained at
22°C and 60 – 70% humidity, with a regular 12 h light-dark cycle. Male MF-1 and CD-
1 mice were ordered in from Charles River between 21-25 g.
Plasma concentration analysis of compound OX03050 using HPLC
CD-1 male mice were administered with compound OX03050 (40 mg/kg) or vehicle
via the intraperitoneal route. Mice were then euthanized at 5, 10, 30, 180 and 360
minutes post compound delivery and plasma was harvested for HPLC analysis. A
standard curve was made up of known concentrations spiked into plasma and ran
alongside unknown drug concentration samples each time. To make the standard
curve known compound OX03050 concentrations were spiked into drug free plasma.
Acetonitrile was added to each sample in a 1:8 ratio, samples were shaken for 15
minutes, before being spun at 13,000 rpm for 15 minutes at 4°C. 200 µL of the
supernatant was run on a Waters 600 controller HPLC machine with a C18 reverse
phase column (Waters), with a linear gradient mobile phase starting at 20:1 H2O with
0.1% TFA: Acetonitrile with 0.1% TFA rising to 1:20 H2O with 0.1% TFA: Acetonitrile
with 0.1% TFA over 7 minutes at a flow rate of 1.5 mL/min.
Hydrodynamic delivery
Mice weighing between 18-35 g received hydrodynamic tail-vein injections of plasmid
DNA as described (Liu, Song and Liu 1999). Animals were anaesthetized with
isoflurane and body temperature was maintained using a heating pad. 50 µg of
10

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
plasmid DNA was resuspended in TransIT-EE Hydrodynamic Delivery Solution (Mirus,
Madison, WI), to a total volume equivalent to 10% of the mouse’s total body weight.
Delivery of the plasmid was administered via the tail-vein with an injection time
between 8-14 seconds depending on the weight of the mouse. Animals were allowed
to recover and left for the appropriate amount of time before sacrifice.
In vivo luciferase assay to assess efficacy of compounds
All mice were blinded, separated into treatment groups and assigned a random
number, images were taken not knowing which number belonged to which
experimental group. Mice 5 days post hydrodynamic delivery had 150 µL of a
15mg/mL D-Luciferin solution delivered via the intraperitoneal (I.P) route, before being
anaesthetized using Isoflurane. Following a 5 minute incubation period, mice were
placed inside the chamber of an IVIS-100 luciferase imaging camera (Caliper Life
Sciences, Waltham, MA) and imaged with a 2 minute exposure time. Mice were dosed
8, 24 and 32 h after acquiring the baseline image with either compound OX03050 (40
mg/kg), vehicle alone or Pravastatin (600 mg/kg), an uninjected group was also
included as a control. 48 h on from the first image the mice were imaged under the
exact same conditions as before. Images were analysed using LivingImage software
(Caliper Life Sciences).
11

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Results
Identification of novel molecules that drive expression of the human LDLR
promoter via a mechanism distinct from statins
A Chinese hamster ovary (CHO) clonal cell line was developed (CHO-pLDLR-Luc)
which contains a 10 kb genomic DNA human low-density lipoprotein receptor (LDLR)
promoter element driving luciferase. The pLDLR-Luc construct has been previously
shown to contain the necessary elements for physiological regulation of expression of
the LDLR locus (Hibbitt et al. 2010). To identify compounds which could upregulate
expression of the LDLR we focused our screening efforts by selecting compounds in
silico with scaffolds reminiscent of known gene transcription modulators, for example
sterols and steroid-like structures. We refined our in-house compound collection to
216 small molecules which we screened at a single concentration (20 µM) in the CHO-
pLDLR-Luc cell line. Compounds 49 (OX03771), 50 and 51 gave the greatest increase
in luciferase activity compared to the vehicle treated control (Figure 1A). This series
of compounds was of interest due to their similarity in structure to sterols (Figure 1B),
as shown in the overlay of compound OX03771 with cholesterol (Figure 1C). Authentic
samples of the three initial hits were synthesized and tested for a dose-dependent
response: compound OX03771 was determined to be the most potent (EC50 = 39 nM
± 29 SEM, Figure 2A). Hep3B cells were transfected with a pCMV-Luc plasmid and
treated with compound OX03771 (Supplementary Figure S1A). No significant
difference was seen between vehicle-treated or compound OX03771-treated cells
expressing pCMV-Luc, ruling out compound OX03771 causing an increase in
luciferase activity through stabilising or interacting with -luciferin or luciferase, a
D
commonly observed artefact in these reporter assays (Thorne et al. 2012). To confirm
compound OX03771 was not itself contributing to the luminescent readout,
untransfected Hep3B cells were treated with vehicle or compounds, with no increase
12

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
in luminescence detected above untreated cells (Supplementary Figure S1B). An
adenylate kinase activity assay carried out on cell culture medium, a measure of
cytotoxicity, demonstrated a favourable safety profile of compound OX03771 with only
the 10 µM dose showing an increase compared to vehicle treated control cells
(Supplementary Figure S1C). The cytotoxicity observed at 10 µM with compound
OX03771 correlates with the drop off in luciferase activity observed in the dose-
response curve at the same concentration.
Compound OX03771 upregulates the LDLR at the mRNA and protein level
Human and mouse hepatocyte cell lines were treated with compound OX03771 to
determine if it could upregulate LDLR mRNA and protein. The EC50 dose established
in the CHO-pLDLR-Luc assay (39 nM) was chosen along with doses 10-fold above
and below (Figure 2A). Hep3B cells were treated with compound OX03771,
cholesterol (negative control) or a statin (positive control). Compound OX03771 gave
a dose-dependent increase at the mRNA level with a significant increase at 390 nM
(Figure 2B). To confirm the increase in LDLR mRNA resulted in upregulation of the
LDLR protein, Hepa1-6 cells were treated with the same concentrations of compound
OX03771 (Figure 2C). Compound OX03771 caused a significant upregulation of the
LDLR protein compared to vehicle treated control cells at 390 nM similar to that of
Simvastatin (Figure 2D).
Compound OX03050 is a more potent analogue of compound OX03771 with
improved pharmacokinetic properties
Compound OX03771 was found to have a half-life of less than 5 minutes when
incubated with mouse liver microsomes and quantified using liquid chromatography–
mass spectrometry (LC-MS) (Supplementary Figure 2A-B). To improve the efficacy,
in vitro ADME and pharmacokinetic profile of compound OX03771, analogues were
13

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
made and their efficacy tested in the CHO-pLDLR-Luc clonal cell line (Figure 3).
Microsomal incubation of selected analogues was also undertaken to assess
metabolic stability. Compound OX03771 was conceptually split into four key regions
(Figure 3), the two aromatic rings in the stilbene (Red (A) and Pink (C) rectangles) the
carbon-carbon double bond linker (Blue (B) rectangles) and the N,N-
dimethylaminopropyloxy chain (Black (D) rectangle). Each of the four regions was
modified systematically in order to assess structure-activity relationships (SARs). The
compounds developed in each series were tested to obtain EC50 and Emax values. All
modifications to aromatic ring A tested (Figure 3A), including the introduction of
electron withdrawing or electron donating substituents, led to a reduction in activity
compared to compound OX03771 (EC50: 39 nM, Emax: 5).
To alter the three-dimensional shape of the molecule the alkane, alkyne and (E,E)-
1,3-butadiene linker analogues were synthesised (Figure 3B), thereby varying either
the compound’s conformational flexibility, the relative orientation of or distance
between the two aryl units. A secondary amide analogue was also prepared as an
isostere of the (E)-alkene. Compound OX03384, which contains a fully saturated alkyl
linker group, alkyne OX03383 and butadiene OX03385 all resulted in lower potency
suggesting that the geometrical constraints imparted by the (E)-olefin are important.
Compound OX03386, containing a secondary amide linker also showed much lower
potency. Given the previously reported propensity of olefins to undergo metabolic
oxidation reactions (Dharwadkar 2016), it was also of interest to investigate bicyclic
isosteres OX03390-93, some of which may be anticipated to be more metabolically
stable. Quinoline OX03390, quinoxaline OX03391 and benzofuran OX03392 all
showed a reduction in potency compared to compound OX03771. Encouragingly
14

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
however, benzothiazole OX03393 showed only a modest drop in potency compared
to compound OX03771, and a similar Emax
.
Altering the regiochemistry within aromatic ring C was found to have a profound effect
on the potency: both ortho and meta N,N-dimethylaminopropyloxy substituted
analogues, compounds OX03394 and OX03373, showed a complete loss of activity
(Figure 3C). Compound OX03387 suggested that reducing the electron density of
aromatic ring C through introduction of an additional fluoro substituent resulted in
reduced activity.
Finally the N,N-dimethylaminopropyloxy chain was modified as shown in Figure 5D.
Hydroxyaryl derivative, compound OX03395, lacking the N,N-dimethylaminopropyl
side chain showed a reduction in potency (EC50 = 3000 nM). The mono-methyl
substituted counterpart (compound OX03389) gave an apparent modest drop in
potency (EC50 = 320 nM) compared to compound OX03771, whereas the truncated
homologue (compound OX03388) showed a significant reduction in potency (EC50 =
2000 nM). Compound OX03050, containing a propyl alcohol chain gave a comparable
EC50 of 26 nM compared to 39 nM (compound OX03771) and a six-fold higher Emax
,
while both the O-methyl ether analogue (compound OX03374, EC50 = 1500 nM) and
the alkyl substituted analogue (compound OX03373, EC50 = 1000 nM) showed a
reduction in activity. Given that, under the assay conditions, the amines would exist
predominantly as their corresponding ammonium salts, taken together these data
suggest that the presence of a hydrogen bond donor within the side chain may confer
higher levels of potency.
15

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Compound OX03050 alongside other analogues that were thought may improve the
in vitro ADME profile were incubated with mouse liver microsomes and their metabolic
conversion measured over time (Figure 3E). Of these, compound OX03050
demonstrated the greatest metabolic stability, having a half-life greater than 45
minutes. All the other compounds tested had dropped below 5% of their starting
amount by 45 minutes, so with the greatest potency and best in vitro metabolic stability
profile compound OX03050 was taken forward for in vivo testing.
Compound OX03050 is able to drive the human LDLR promoter in vivo
To test whether the ADME profile seen in mouse liver microsomes was confirmed in
vivo, an HPLC assay was developed to measure compound OX03050 in mouse
plasma. 40 mg/kg of compound OX03050 was delivered via the intraperitoneal route
(I.P) to CD-1 male mice and plasma taken at various time points up to 6 hours (Figure
4A). Compound OX03050 demonstrated a half-life of 84 minutes in vivo, consistent
with the in vitro stability data, and could be detected at least up to 6 hours after dosing.
To test the pharmacodynamic effects of Compound OX03050 in vivo, MF-1 male mice
had a hydrodynamic injection of the pLDLR-Luc plasmid, in order to quantify LDLR
promoter activity in vivo by non-invasive bioluminescence imaging. Baseline images
were obtained five days after injection of the pLDLR-Luc plasmid (Figure 4C). Mice
then received three doses of either vehicle, compound OX03050 (40 mg/kg) or
pravastatin (600 mg/kg) spaced over 48 hours before final luciferase imaging (Figure
4C). An uninjected group which received the pLDLR-Luc plasmid but no I.P injections
was used as a negative control. Compound OX03050 injected mice had significantly
higher luciferase expression post compound delivery compared with vehicle treated
mice (Figure 4B). Pravastatin injected mice also had significantly higher luciferase
16

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
expression than vehicle injected mice, as would be expected and has previously been
shown (Hibbitt et al. 2010). No difference was observed between uninjected mice and
mice receiving the vehicle treatment. Here we demonstrate that compound OX03050
is able to drive expression of the human LDLR promoter in vivo.
Compound OX03050 and compound OX03771 inhibit squalene synthase
We next confirmed that compound OX03771 upregulates the LDLR through a
mechanism distinct from statins (Figure 5A). The activity of purified HMGCR is
measured indirectly through measuring the absorption of NADPH at 340 nm over time.
As HMGCR converts HMG-CoA to mevalonate, NADPH, a co-substrate in the reaction
is oxidised to NADP+, leading to a decrease in the absorption at 340 nm. The activity
of the enzyme was confirmed, as there is a significant decrease in absorption between
blank (no enzyme added) and vehicle treated samples (enzyme added). Pravastatin
was used a positive control, which significantly inhibited the decrease in absorption
compared to a vehicle treated sample after 15 minutes. No significant difference was
observed between compound OX03771 and vehicle treated samples (Figure 5A),
demonstrating that compound OX03771 does not exert its effects on LDLR expression
via inhibition of HMGCR.
Squalene synthase activity was then measured in human liver microsomes in the
presence of compounds OX03050 and OX03771. The compounds were pre-incubated
with the microsomes for 10 minutes before the reaction was started by adding [3H]-
FPP. The reaction went on for 10 minutes before inhibition was assessed. The EC50
values for compound OX03050 (26 nM) and compound OX03771 (39 nM) established
in the CHO-pLDLR-Luc assay and concentrations 10-fold above and below this were
17

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
used. Compounds OX03050 and OX03771 both inhibited the conversion of FPP to
squalene in a dose-dependent manner, with both showing ~60% inhibition at the
higher dose (Figure 5B). This suggests that the increase in LDLR expression observed
with both compounds OX03050 and OX03771 is mediated, at least in part, by inhibition
of squalene synthase.
In silico modelling was carried out on these compounds in an effort to rationalise their
inhibitory potencies against squalene synthase. MOE software was used with the
available human squalene synthase crystal structure (PDB: 1EZF), which had been
co-crystallised with the ligand CP-320473 (Pandit et al. 2000). Compound OX03771
was docked in both its neutral state and as the corresponding protonated ammonium
species for comparison. It was anticipated that OX03371 would exist predominantly
as the protonated form under physiological conditions, but it was unclear which form
would preferentially bind to squalene synthase. Docking studies were conducted as
30 independent experiments. For the ammonium salt of OX03771 no consistent
docking prediction was found. In contrast, all solutions for the free base of OX03771
were predicted to bind in a similar manner to the ligand CP-320473 in the active site,
with the hydrophobic aryl group sitting deep in the pocket while the chain bearing the
hydrogen bond acceptor amino group was predicted to protrude out to interact with
Arg-77 (Figure 5C). To further explore this putative interaction, compound OX03994
was docked into the active site, a molecule which demonstrated no activity in the CHO-
pLDLR-Luc assay (Supplementary Figure 3). As with OX03771, the amino group
within OX03994 is predicted to interact with Arg-77 however the molecule is predicted
to adopt an alternative orientation within the active site as the regiochemistry of the
N,N-dimethylaminopropyloxy substituent is switched from the para to meta. It is
18

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
possible that this leads to greater instability of the complex, which could explain the
lack of activity observed for this compound. Compound OX03050 was predicted to
bind in a similar manner to compound OX03771 (Figure 4D). As with the previous
compounds the hydrophobic aryl group was predicted to insert into the active site
pocket, however in a slightly different conformation with the terminal hydroxyl group
predicted to interact with Arg77 and Tyr 72. The squalene synthase activity assay
coupled with the in silico modelling provides a potential rationale to support compound
OX03050 and OX03771 as inhibitors of squalene synthase.
A squalene synthase inhibitor and a statin in combination causes a synergistic
drive in expression of the LDLR
We sought to determine if a squalene synthase inhibitor and statin used in combination
are able to drive higher levels human LDLR promoter expression in a synergistic
manner. The CHO-pLDLR-Luc cell line was treated with compounds for 48 hours to
evaluate the effect of these small molecules in combination to drive the human LDLR
promoter. A dose-response curve was generated for compound OX03050 (Figure 6A)
and Simvastatin (Figure 6B) alone, giving EC50 values of 26 and 128 nM, respectively.
A dose-response of compound OX03771 was carried out in the presence of compound
OX03050 dosed at its EC50 (Figure 6C). There was no enhancement in luciferase
activity using two squalene synthase inhibitors (Emax: 2.43 RLU/mg/mL) compared to
the dose response of compound OX03771 alone (Emax: 3.28 RLU/mg/mL) (Figure 2A).
The same effect was seen when two statins were dosed together, a dose response of
pravastatin was carried out in the presence of Simvastatin dosed at its EC50 (Figure
6D). No increases in the luciferase activity were seen using two statins (Emax: 2.04
RLU/mg/mL) compared to Simvastatin alone (Emax: 2.76 RLU/mg/mL).
19

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
However, dose-response curves for Simvastatin in the presence of compound
OX03050 dosed at its EC50 (Figure 6E) and EC80 (Figure 6F) gave a much greater
increase in luciferase activity, giving Emax values of 6.35 and 11.77 RLU/mg/mL,
respectively, than compared to two statins alone. The same was also true for the
reverse experiment where a dose-response of compound OX03050 was carried out in
the presence of Simvastatin dosed at its EC50 (Figure 6G). The Emax value was 16.77
RLU/mg/mL, a much larger increase in luciferase activity than seen using two
squalene synthase inhibitors in combination. The fold increases in activity given when
two squalene synthase inhibitors or two statins were dosed compared to vehicle
treated cells was 3.9 and 1.1, respectively. In comparison, the fold increase when a
statin was dosed alongside a squalene synthase inhibitor was 10- and 16.1-fold
increase compared to vehicle treated cells. This is consistent with a synergistic effect
of inhibiting the cholesterol biosynthesis pathway at two distinct parts of the pathway
can greatly enhance the expression of the human LDLR.
20

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Discussion
Here we present a novel series of small molecules able to upregulate LDL receptor
expression (LDLR) through inhibition of squalene synthase. We first identified
compound OX03771 from a focused small compound screen (216 molecules) in a
reporter cell line produced in our laboratory. The reporter line contains a 10 kb human
LDLR promoter element, including all components required for physiological
regulation of expression, driving luciferase. The initial hit OX03771 was then confirmed
in two secondary assays, and shown to increase LDLR expression at both the mRNA
and protein level. Through systematic SAR studies we identified an even more potent
analogue of compound OX03771 with improved in vitro metabolic stability, compound
OX03050. We dosed compound OX03050 in mouse, establishing a pharmacokinetic
profile of the drug and demonstrated that the efficacy seen in vitro translates to an in
vivo system. Furthermore, we utilised a unique approach to quickly assess if
compound OX03050 can drive the human LDLR promoter in a mammalian system.
Delivery of the reporter vector via hydrodynamic delivery (a well-established method
to deliver plasmid DNA directly to the liver (Liu, Song and Liu 1999)) resulted in an
efficient and quick way to assess in vivo efficacy of our small molecules.
We next elucidated the mechanism of action of compounds OX03771 and OX03050
as inhibiting the enzyme squalene synthase, the first committed step in cholesterol
biosynthesis. This suggests that the increase in LDLR expression observed with both
compounds OX03050 and OX03771 is mediated, at least in part, by inhibition of
squalene synthase. Other inhibitors of squalene synthase have previously been
21

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
reported and investigated for their effects on lipid-lowering (Elsayed and Evans 2008).
Some of our small molecules resemble a series of stilbene derivatives reported several
years ago as inhibitors of squalene synthase with modest potencies (Harwood et al.
1997). All the examples described by Harwood et al incorporate a basic amino group
within the side chain appended to the stilbene and in their study and propose that the
corresponding ammonium derivatives act as structural mimics of the cationic
intermediate produced within the enzyme active site upon farnesyl pyrophosphate
dimerization. Intriguingly, our compounds retain inhibitory potency even in the absence
of an amino group in the side chain, e.g. OX03050 shows similar potency to OX03371,
and our modelling studies predict similar interactions between protein and inhibitor for
both compounds. Our small molecules share some similarity in structure to cholesterol
so it is possible that they may also interact with other enzymes downstream of
squalene synthase; however, this would have to be further investigated. Finally, we
provided data to support that a squalene synthase inhibitor used in conjunction with a
statin acts synergistically to increase LDLR expression.
In the future, our in vivo reporter model could be used to screen further arrays of small
molecules for their ability to drive the human LDLR promoter in a mammalian system.
This platform provides a quick and simple method to eliminate any compounds which
fail to carry in vitro efficacy to an in vivo system. It is well known that statins do not
have the same effect in rodents as they show in man, with studies showing a decrease
in LDLR expression and no change in total cholesterol (Rashid et al. 2005) after statin
administration. This in part could be attributed to the difference in how rodents and
humans carry their cholesterol (Camus et al. 1983), with mice carrying most of their
cholesterol in the high-density lipoprotein (HDL) fraction. Other differences due to the
22

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
absence of certain proteins such as cholesteryl ester transfer protein (Guyard-
Dangremont et al. 1998) in mouse or inclusion of Apolipoprotein-4 8 (Greeve et al.
1993) into LDL particles result in a decreased shuttling of cholesterol ester to the LDL
particles and enhanced clearance of LDL particles respectively. Inhibitors of
cholesterol biosynthesis only consistently show efficacy in larger mammals, such as
guinea pig (Matsunaga et al. 1991) or rabbit (Krause and Newton 1995). Our reporter
system could provide a more economical and higher throughput platform to assess in
vivo efficacy before embarking on a longer-term efficacy study in a larger animal
model.
As lipid goals become stricter for patients who have dyslipidaemia, especially in high
risk patients, such as those with heterozygous familial hypercholesterolaemia (heFH),
high dose statins may become intolerable. Many studies have shown that only a small
percentage of heFH patients reach their lipid goal (Pijlman et al. 2010). The
emergence of the proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors are
a welcome addition to try and further lower the LDL-cholesterol of heFH patients
(Robinson et al. 2015). The excellent lipid lowering seen with this class of biologics
are surely promising however as they need to be administered via a subcutaneous
injection every two weeks and are expected to have a high cost (Zimmerman 2015),
an oral medication which can be taken alongside lower dose statins may still be of
great use. We demonstrated that a squalene synthase inhibitor with a statin could
cause a greater upregulation of LDLR promoter activity than can be achieved with
either class of drug alone. Therefore, a lower dose of both classes of drugs could be
used in high risk patients who are unable to tolerate maximal dose statins, hopefully
alleviating the side effects they are experiencing. This enhanced efficacy has been
23

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
shown in the clinic previously (Stein et al. 2011) however the liver side effects seen
with the squalene synthase inhibitor may mean a statin will have to be used alongside
any squalene synthase inhibitor to inhibit a build-up of FPP (Wasko et al. 2011).
Squalene synthase inhibitors have also been shown to lower triglycerides (Ugawa et
al. 2000) which may be explained by recent studies investigating the orthologs
differentially affected by squalene synthase inhibitors and statins (Rondini et al. 2016).
Only squalene synthase inhibitors were shown to alter cellular lipid metabolism and so
may provide another added benefit to using them alongside a statin.
Here, we have characterised a novel series of small molecules, the most active of
which have potency in the nanomolar range. To further evaluate whether these
compounds can lower LDL-cholesterol in an animal model, dosing a large animal
model such as the guinea pig or rabbit model. These larger species share a similar
lipid profile to humans, in which the majority of cholesterol is carried in the LDL-fraction
and cholesterol levels respond to compounds which inhibit the cholesterol
biosynthesis pathway. This further assessment will be critical for the progression of
this series of molecules towards any future therapy.
24

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Authorship contributions
Participated in research design: Kerr, Tam, Hibbitt, Hale, Douglas, Bataille, Wynne,
Channon, Russell and Wade-Martins.
Conducted experiments: Kerr, Tam, Hale, Cioroch, Douglas, Agkatsev, Hibbit,
Mason, Holt-Martyn and Bataille.
Performed data analysis: Kerr, Tam, Agkatsev, Mason, Holt-Martyn, Bataille, Wynne
Russell, Channon and Wade-Martins.
Wrote or contributed to the writing of the manuscript: Kerr, Douglas, Wynne,
Channon, Russell and Wade-Martins
25

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
References
Amin D, Cornell S a, Gustafson SK et al. Bisphosphonates used for the treatment of
bone disorders inhibit squalene synthase and cholesterol biosynthesis. J Lipid
Res 1992;33:1657–63.
Bitzur R, Cohen H, Kamari Y et al. Intolerance to Statins: Mechanisms and
Management. Diabetes Care 2013;36:S325–30.
Camus M, Chapman MJ, Forgez P et al. Lipoproteins and Apoproteins in the Mouse
,. 1983;24.
Davidson MH. Squalene synthase inhibition: A novel target for the management of
dyslipidemia. Curr Atheroscler Rep 2007;9:78–80.
Dharwadkar SS. Review Implications of Drug Metabolizing Enz Chemistry & Biology
Interface. 2016.
Elsayed RK, Evans JD. Emerging lipid-lowering drugs: squalene synthase inhibitors.
Expert Opin Emerg Drugs 2008;13:309–22.
Greeve J, Altkemper I, Dieterich JH et al. Apolipoprotein B mRNA editing in 12
different mammalian species: hepatic expression is reflected in low
concentrations of apoB-containing plasma lipoproteins. J Lipid Res
1993;34:1367–83.
Guyard-Dangremont V, Desrumaux C, Gambert P et al. Phospholipid and cholesteryl
ester transfer activities in plasma from 14 vertebrate species. Relation to
atherogenesis susceptibility. Comp Biochem Physiol - B Biochem Mol Biol
1998;120:517–25.
Harwood HJ, Barbacci-Tobin EG, Petras SF et al. 3-(4-chlorophenyl)-2-(4-
diethylaminoethoxyphenyl)-A-pentenonitrile monohydrogen citrate and related
analogs. Reversible, competitive, first half-reaction squalene synthetase
inhibitors. Biochem Pharmacol 1997;53:839–64.
Hibbitt OC, McNeil E, Lufino MM et al. Long-term physiologically regulated
expression of the low-density lipoprotein receptor in vivo using genomic DNA
mini-gene constructs. Mol Ther 2010;18:317–26.
Ichikawa M, Ohtsuka M, Ohki H et al. Discovery of DF-461, a potent squalene
synthase inhibitor. ACS Med Chem Lett 2013;4:932–6.
Krause BR, Newton RS. Lipid-lowering activity of atorvastatin and lovastatin in
rodent species: triglyceride-lowering in rats correlates with efficacy in LDL
animal models. Atherosclerosis 1995;117:237–44.
Liu F, Song Y, Liu D. Hydrodynamics-based transfection in animals by systemic
administration of plasmid DNA. Gene Ther 1999;6:1258–66.
Lusis AJ. Insight Review Articles. Nature 2000;407:233–41.
Marcoff L, Thompson PD. The Role of Coenzyme Q10 in Statin-Associated
Myopathy. J Am Coll Cardiol 2007;49:2231–7.
26

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Matsunaga A, Sasaki J, Takada Y et al. Effect of simvastatin on receptor mediated
metabolism of low density lipoprotein in guinea pigs. 1991;90:31–7.
Moßhammer D, Schaeffeler E, Schwab M et al. Mechanisms and Assessment of
Statin-related Muscular Adverse Effects. Br J Clin Pharmacol 2014:n/a-n/a.
Nagashima S, Yagyu H, Tozawa R et al. Plasma cholesterol-lowering and transient
liver dysfunction in mice lacking squalene synthase in the liver. J Lipid Res
2015;56:998–1005.
Nishimoto T, Amano Y, Tozawa R et al. Lipid-lowering properties of TAK-475, a
squalene synthase inhibitor, in vivo and in vitro. Br J Pharmacol 2003;139:911–
8.
Nishimoto T, Ishikawa E, Anayama H et al. Protective effects of a squalene synthase
inhibitor, lapaquistat acetate (TAK-475), on statin-induced myotoxicity in guinea
pigs. Toxicol Appl Pharmacol 2007;223:39–45.
Pandit J, Danley DE, Schulte GK et al. Crystal Structure of Human Squalene
Synthase. Biochemistry 2000;275:30610–7.
Pedersen TR. Randomised trial of cholesterol lowering in 4444 patients with
coronary heart disease: the Scandinavian Simvastatin Survival Study (4S).
Lancet 1994;344:1383–9.
Pijlman AH, Huijgen R, Verhagen SN et al. Evaluation of cholesterol lowering
treatment of patients with familial hypercholesterolemia: a large cross-sectional
study in The Netherlands. Atherosclerosis 2010;209:189–94.
Rashid S, Curtis DE, Garuti R et al. Decreased plasma cholesterol and
hypersensitivity to statins in mice lacking Pcsk9. Proc Natl Acad Sci U S A
2005;102:5374–9.
Robinson JG, Farnier M, Krempf M et al. Efficacy and safety of alirocumab in
reducing lipids and cardiovascular events. N Engl J Med 2015;372:1489–99.
Rondini EA, Duniec-Dmuchowski Z, Cukovic D et al. Differential regulation of gene
expression by cholesterol biosynthesis inhibitors that reduce (pravastatin) or
enhance (squalestatin 1) nonsterol isoprenoid levels in primary cultured mouse
and rat hepatocytes. J Pharmacol Exp Ther 2016;1:216–29.
Shepherd J, Cobbe S, Ford I et al. Prevention of Coronary Heart Disease With
Pravastatin in Men With. Lancet (London, England) 1995;333.
Sniderman AD, Lawler PR, Williams K et al. The causal exposure model of vascular
disease. Clin Sci (Lond) 2012;122:369–73.
Stein E a., Bays H, O’Brien D et al. Lapaquistat acetate: Development of a squalene
synthase inhibitor for the treatment of hypercholesterolemia. Circulation
2011;123:1974–85.
Thorne N, Shen M, Lea WA et al. Firefly luciferase in chemical biology: a
compendium of inhibitors, mechanistic evaluation of chemotypes, and
suggested use as a reporter. Chem Biol 2012;19:1060–72.
27

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Ugawa T, Kakuta H, Moritani H et al. YM-53601, a novel squalene synthase inhibitor,
reduces plasma cholesterol and triglyceride levels in several animal species. Br
J Pharmacol 2000;131:63–70.
Wasko BM, Smits JP, Shull LW et al. A novel bisphosphonate inhibitor of squalene
synthase combined with a statin or a nitrogenous bisphosphonate in vitro. J
Lipid Res 2011;52:1957–64.
Zimmerman MP. How do pcsk9 inhibitors stack up to statins for low-density
lipoprotein cholesterol control? Am Heal Drug Benefits 2015;8:436–42.
Footnotes
The work was funded by the British Heart Foundation (BHF) through Project Grant
PG/10/54/28460 to RW-M and KC, a Research Councils’ UK Fellowship awarded to
AR, a BHF Studentship to AK, and support from the Oxford BHF Centre for
Research Excellence (RE/08/004)
28

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
Figure legends
Figure 1: Identification of novel small molecules which upregulate LDLR
genomic DNA promoter activity.
A compound library of 216 small molecules was screened at a single concentration
(20 µM) in a CHO-pLDLR-Luc cell line. Luciferase is under the control of 10 kb of
genomic DNA upstream of the LDLR locus, including the promoter and elements
essential for physiological regulation. (A) Three initial hits appeared to give an increase
in luciferase expression compared to DMSO (0.1%) control treated cells. (B) The
structure of compound 49 (OX03771), the most potent of the initial hits. (C) The
structure of compound 49 (OX03771) and cholesterol showing their similarity in
structure.
Figure 2: Compound OX03771 dose dependently increases the LDLR at the
mRNA and protein level with an EC50 in the nanomolar range
(A) CHO-pLDLR-Luc cells were treated with compound OX03771, or vehicle control
(0.1% DMSO) for 48 hours before luciferase expression was measured. Compound
OX03771 gave a dose dependent increase in luciferase expression compared with
vehicle treated cells and had an EC50 in the nanomolar range. Luciferase expression
was normalised to total protein; n=4 (B) Hep3B cells were treated with increasing
doses of compound OX03771 for 24 hours before mRNA expression was analysed.
Cholesterol treated (25.8 µM) and Simvastatin treated (500 nM) cells acted as a
negative and positive control respectively. Compound OX03771 gave a significant
increase in LDLR mRNA expression compared with vehicle treated cells; n=5. (C)
Representative western blot and (D) quantification of mouse Hepa1-6 cells treated
with increasing concentrations of compound OX03771 for 48 hours before Ldlr protein
29

JPET Fast Forward. Published on March 30, 2017 as DOI: 10.1124/jpet.116.239574
This article has not been copyedited and formatted. The final version may differ from this version.
JPET #239574
expression was analysed. Cholesterol (25.8 µM) and Simvastatin (500 nM) were used
as negative and positive controls respectively. Compound OX03771 gave a significant
increase in Ldlr protein expression compared with vehicle cells; n=3. Error bars denote
SD. Significance represents treatment compared to vehicle treated control. *P<0.05,
**P<0.01, ***P<0.001. One-way ANOVA with Dunnett’s post-hoc analysis.
Figure 3: Identification of compound OX03050 a more potent analogue of
compound OX03771 with improved metabolic stability
Systematic SAR studies were carried out on compound OX03771, whereby each
component of the molecule was altered individually and screened in the CHO-pLDLR-
Luc at 1, 10, 30, 100, 300, 1000 and 10000 nM concentrations. The (A) red squared
aromatic ring, (B) blue squared double bond linker, (C) purple squared aromatic ring
and (D) N,N-dimethylaminopropyloxy chain modifications lead to the identification of
compound OX03050, a compound with improved potency. All compounds were
screened in duplicate. (E) Listed compounds were spiked into mouse liver microsomes
and after given time points the remaining amounts were quantified through LC/MS.
The amount is given in the table as a percentage in comparison with the amount
detected at the 0 minute time point. Of these compound OX03050 was the most stable,
with over half still present after 45 minutes.
Figure 4: Compound OX03050 is able to increase expression of the human LDLR
promoter in a mammalian system
(A) Compound OX03050 (40 mg/kg) was detectable up to 6 hours post I.P
administration in mouse plasma. (B) Timeline of in vivo efficacy study, pLDLR-Luc
plasmid was delivered to MF-1 male mice via hydrodynamic injection on day 0. 5 days
post hydrodynamic injection a (C) baseline image was taken for each mouse. Three
30