Discovery of potent and selective PPARα/δ dual antagonists and initial biological studies

Article abstract of DOI:10.1016/j.bmcl.2018.12.045

We previously published on the design and synthesis of novel, potent and selective PPARα antagonists suitable for either i.p. or oral in vivo administration for the potential treatment of cancer. Described herein is SAR for a subsequent program, where we set out to identify selective and potent PPARα/δ dual antagonist molecules. Emerging literature indicates that both PPARα and PPARδ antagonism may be helpful in curbing the proliferation of certain types of cancer. This dual antagonism could also be used to study PPARs in other settings. After testing for selective and dual potency, off-target counter screening, metabolic stability, oral bioavailability and associated toxicity, compound 11, the first reported PPARα/δ dual antagonist was chosen for more advanced preclinical evaluation.

Full text of DOI:10.1016/j.bmcl.2018.12.045

Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of potent and selective PPARα/δ dual antagonists and initial
biological studies
Jason D. Jacintho⁎, Christopher S. Baccei, Yalda Bravo, Alex Broadhead, Austin Chen,
Lucia Correa, Kim Fischer, Bryan Laffitte, Catherine Lee, Daniel S. Lorrain, Davorka Messmer,
Peppi Prasit, Karin J. Stebbins, Nicholas S. Stock
Inception Sciences, Suite 100, 5871 Oberlin Drive, San Diego, CA 92121, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
We previously published on the design and synthesis of novel, potent and selective PPARα antagonists suitable
for either i.p. or oral in vivo administration for the potential treatment of cancer. Described herein is SAR for a
subsequent program, where we set out to identify selective and potent PPARα/δ dual antagonist molecules.
Emerging literature indicates that both PPARα and PPARδ antagonism may be helpful in curbing the pro-
liferation of certain types of cancer. This dual antagonism could also be used to study PPARs in other settings.
After testing for selective and dual potency, off-target counter screening, metabolic stability, oral bioavailability
and associated toxicity, compound 11, the first reported PPARα/δ dual antagonist was chosen for more ad-
vanced preclinical evaluation.
PPAR alpha
PPAR delta
Antagonist
Small molecule
Cancer
Introduction
metabolism, is implicated in other processes involving survival, pro-
In our previous publications,1,2 we identified, to the best of our
knowledge, the first selective PPARα antagonists suitable for advanced
in vivo preclinical studies. PPARα, one of three distinct PPAR (peroxi-
some proliferator-activated receptor) isoforms (α, δ and γ) is largely
responsible for the regulation of lipid metabolism. Agonism of PPARα is
known to facilitate fatty acid oxidation (FAO), so we hypothesized that
antagonism of PPARα would be a useful probe to study in certain fatty
acid (FA) metabolizing cancer cells, including: ovarian,3 renal cell
carcinoma,4 and certain types of detached metastatic cells.5–7
liferative genes,5 and angiogenesis.10 Publications have shown that
PPARδ antagonism,12 and PPARδ murine knockout models.13 slow
down certain kinds of tumor progression. Corroborating evidence in-
dicates that the converse is true, namely that PPARδ agonists promote
cancer cell proliferation and inhibit apoptosis.12 More recent literature
has suggested that suppression of PPARδ could be useful against me-
tastatic melanoma,14 and colon cancer.15 Finally, there is additional
evidence that implicates PPARs in the regulation of multi-drug re-
sistance proteins.16 It is possible that antagonizing one or more of the
PPAR isoforms could reduce drug efflux and subsequent chemo-re-
sistance.
PPARδ, together with PPARα, are critical in the sensing of fatty acid
levels and each is central to the cascade of FA-induced gene expres-
sion.8 All three PPAR isoforms are highly homologous, but they differ in
terms of their respective tissue distribution as well as ligand specificity.
Whereas PPARα is primarily expressed in tissues with a propensity for
extensive FA catabolism, e.g, liver, heart, kidney and skeletal muscle,
PPARδ is ubiquitously expressed.9 We hypothesized that a dual PPARα/
δ antagonist may prove superior to PPARα inhibition alone by virtue of
potential overlap in their inherent functions. It was recently revealed
that inhibitors of the PPAR-mediated FAO pathway could be useful in
treating resistant breast cancer,10 glioblastoma,11 and the clearance of
quiescent leukemia initiating cells.8 Inhibition of PPARδ initiates cel-
lular symmetric commitment pathways that result in the exhaustion of
cancer stem cells. PPARδ, in addition to its role in lipid and energy
It was our plan to identify a single molecule that would selectively
antagonize both PPARα and δ, in the hope that we would recruit an
additive biological effect. It is worth mentioning that, as far as we are
aware, no dual PPARα/ δ, single-agent antagonists have been reported
to date. A challenging aspect of this endeavor is that the PPARα and
PPARδ isoforms often bind to natural ligands of dissimilar structure.17
However, dual agonists of PPARα/ δ have been identified and proved to
be good starting points for our SAR initiatives.18 Compound 1 (Fig. 1), a
dual PPARα/ δ agonist, was our reference scaffold from which we based
our initial derivatives. We tested whether a replacement of the fibrate
carboxylic acid with an acylsulfonamide would trigger a literature-
precedented PPAR agonist-to-antagonist switch (compounds 3 and 4,
⁎ Corresponding author.
E-mail address: jjacintho@inceptionsci.com (J.D. Jacintho).
https://doi.org/10.1016/j.bmcl.2018.12.045
Received 31 August 2018; Received in revised form 28 November 2018; Accepted 3 December 2018
0960-894X/©2018ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:Jacintho,J.D.,Bioorganic&MedicinalChemistryLetters,https://doi.org/10.1016/j.bmcl.2018.12.045

J.D. Jacintho et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 2
PPARα/ δ dual antagonism: Replacement of the methoxy linker.
a
a
Compound
X
PPARα IC50 (µM)
PPARδ IC50 (µM)
0.319
8
9
1.10
0.424
0.667
10
23.12b
27.73
0.025
Fig. 1. Initial Probe of Fibrate Agonist to Acylsulfonamide Antagonist Switch.
11
0.113
Fig. 1).1,9
We synthesized compound 2 with the hope of quickly achieving
dual antagonism. This effort was only partly successful in that 2 is a
moderately potent PPARδ antagonist (IC50 = 820 nM), but had little
effect on the PPARα isoform. It was at this point that we decided to
completely replace the fibrate residue of 1. We have previously iden-
tified and described the chemistry of several chemical functionalities
which induce potent PPARα antagonism when grafted on to known
agonist scaffolds.1,2 We synthesized several of these antagonism-indu-
cing fibrate replacements and their dual PPARα and δ inhibitory po-
tencies are listed in Table 1.
a
b
Values are the mean of at least three experiments except where noted.
Values are the mean of two experiments.
Next, we turned our attention toward the ether linker between the
triazolone ring and the central phenyl ring. Compounds 8–11 (Table 2)
were prepared in order to vary the length and composition of this linker
moiety. For purposes of comparison, we kept the 2-ethoxyphenylacetic
acid fibrate replacement found in compound 6 because it provided a
balanced inhibition of both PPARα and δ. To our gratification, ex-
tending the two-atom, methoxy linker to a three atom linker improved
PPARα/ δ dual antagonism tremendously. In particular, compound 11,
with a three-carbon propyl chain, showed superior dual antagonism.
We built upon this discovery by making single-point substitutions
on the triazolone ring, starting with the nitrogen at the 1-position. The
general synthesis for making these derivatives is illustrated in Scheme
1. Commercially available butanoic acid 12, was converted to acylhy-
drazide 13 via CDI coupling with hydrazine. Subsequent treatment with
2-fluorophenyl isocyanate afforded semicarbazide 14, which was then
cyclized to triazolone 15 with TMSOTf and triethylamine in refluxing
toluene. The 1-position can then be diversified by alkylation or aryla-
tion using the desired alkyl bromide and cesium carbonate in acetoni-
trile or Ullmann conditions with the desired aryl iodide, respectively.
These conditions provided aryl bromide intermediate 16 which was
then coupled to methyl 2-(2-ethoxy-5-(4,4,5,5-tetramethyl-1,3,2-diox-
aborolan-2-yl)phenyl)acetate under palladium-mediated Suzuki-
Miyaura borylation conditions to provide the phenylacetic acid methyl
ester 17. Finally, hydrolysis of the ester provided the phenyl acetic
acids 18–25 (Table 3).
This initial effort (compounds 3–7), in which potencies were ob-
tained from standard PPARα and δ luciferase cell-based functional as-
says,19 failed to identify a PPARα/ δ dual antagonist of ample potency.
We did, however, succeed in making a potent and selective PPARα
inhibitor in compound 5.
Table 1
PPARα/ δ dual antagonism of various fibrate replacements.
a
a
Compound
R
PPARα IC50 (µM)
PPARδ IC50 (µM)
Upon examination of the SAR in Table 3, it should be stated that our
effort here was not exhaustive by any means. We were working with an
established SAR from a previous, related program.2 In that work, most
changes to the 1-position were detrimental to potency. We wanted to
justify this bias with a representative, yet small number of analogs.
Analysis of PPARα/ δ potencies in Table 3 reveal that no replacements
were identified which were superior to the 4-trifluoromethylphenyl
group found in 11. Compound 18, containing a small ethyl group at the
1-position, completely abrogated PPARα antagonism and decreased
that of PPARδ by 280 fold. Moving the trifluoromethyl group to the 3-
position, as in 19, caused significant loss in potency across the two
isoforms. Interestingly, compound 20 and 21, each bearing substitu-
tions similar to, yet slightly > 11, had better potency on PPARα, but
were inferior on PPARδ. We then screened a few sulfur-containing
compounds, 22 and 23, focusing in on what we assumed to be the
privileged spot on the phenyl ring, the para-position. In general, neither
of these compounds qualified as a dual antagonist of sufficient potency,
but it was interesting that 22 appeared to be our first moderately
3
31.44
1.97
4
5
6
38.14
3.74
0.018b
11.44
> 100.0c
4.49
7
30.99
48.80
a
b
c
Values are the mean of at least three experiments except where noted.
Values are the mean of two experiments.
One experiment.
2

J.D. Jacintho et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Scheme 1. (a) CDI, THF, RT then NH2NH2 H2O, RT, 95%. (b) 2-Fluorophenyl NCO, THF, RT, > 99%. (c) TMSOTf, TEA, Tol, reflux. (d) ArI, CuI, trans-1,2-di-
methylaminocyclohexane, K2CO3, dioxane, reflux, ∼30–90% (two step yield) or RBr, Cs2CO3, CH3CN, 50 °C, ∼50–90% (two step yield). (e) Tetrakis(triphenyl-
phosphine)palladium(0), methyl 2-(2-ethoxy-5-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)acetate, dioxane, sat. aq. NaHCO3, 85 °C, ∼40–80%. (f) LiOH,
THF, MeOH, H2O, RT, ∼90%.
At this stage, we were satisfied that the 4-trifluoromethylphenyl
group at position 1 of the triazolone was optimal. We then initiated
what would be the most extensive portion of our PPARα/ δ dual an-
tagonist SAR campaign. We hypothesized that replacement of the 2-
fluorophenyl group at position 4 of the triazolone would give us the
best chance to improve on compound 11. The general synthesis for
compounds in Table 4 mirror the synthesis found in Scheme 1, where
we simply used a variety of isocyanates (most of them available com-
mercially) in the second step to introduce novel functionalities at po-
sition 4. We began by synthesizing alkyl groups of various sizes, as
exemplified in compounds 26–29. The general trend seen here, in terms
of dual potency, indicates that smaller groups boosted PPARα inhibi-
tion to the detriment of PPARδ potency. Compound 29 showed pro-
mising dual PPARα/ δ potency but was not selected for further eva-
luation due to borderline CYP3A4 induction potential (Table 5). Next,
we made a variety of phenyl analogs, varying both the functionality and
position of substitution. Compounds 30–36 were prepared and tested
for dual potency. We were pleased to see that two of the compounds
were either superior or better balanced, in terms of dual potency, when
compared to compound 11. Whereas compound 30 was superior for
inhibition of both PPAR isoforms, compound 36 was approximately five
times more potent on PPARα with only a small sacrifice in PPARδ po-
tency. Overall, it can be argued that both compound 30 and 36 were
more balanced in terms of dual potency compared to 11. This excite-
ment was proven premature as compound 30 showed significant
CYP3A4 induction (Table 5) and compound 36 showed disappointing
oral exposure in rat (data not shown) when compared to compound 11.
Finally, we prepared several pyridyl analogs as replacements for the 2-
fluorophenyl group (37–39) as well as a slightly larger, 2-fluor-
obenzylic homolog (40). All three pyridine replacements were tolerated
and showed appreciable dual antagonism, but none of them proved
superior to compound 11 in terms of potency. Compound 40 demon-
strated to us that larger substitutions at this position would most likely
prove detrimental.
Table 3
N-1 substitution: Effect on PPARα/ δ dual antagonism.
Compound
Y
PPARα IC50 (µM)a
89.30b
PPARδ IC50 (µM)a
18
19
24.97
3.59
0.678
20
0.033
0.610
21
22
0.018
1.24
0.197
0.079
23
5.04
2.78
24
25
0.200
0.394
25.55
22.23
a
b
Values are the mean of at least three experiments except where noted.
Value was obtained from one experiment.
selective PPARδ antagonist. Finally, compounds 24 and 25 were syn-
thesized to see what impact larger, single-carbon homologs had on
potency. As the data reveals, neither of these compounds qualified as a
dual inhibitor.
Compounds 41 and 42 were synthesized and tested for PPARα/ δ
dual antagonism (Fig. 2). Compound 41 was prepared in order assess
whether a simple switch of the two phenyl groups found on 11 would
be tolerated. As the potency numbers show, this switch was detrimental
3

J.D. Jacintho et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 4
N-4 substitution: Effect on PPARα/δ dual antagonism.
Compound
Z
PPARα IC50 (µM)a
0.051
PPARδ IC50 (µM)a
0.438
Fig. 2. Additional analogs: Swapping of triazolone substituents (41). Benzoic
acid analog (42).
26
27
28
0.032
0.055
0.439
0.124
Table 6
Screening summary for Compound 11.
Assay Target
IC50 (µM)a
PPARα
0.113
0.025
21.3
29
0.066
0.066
PPARδ
PPARγ
PPARα agonism
PPARδ agonism
PPARγ agonism
mousePPARα
mousePPARδ
mousePPARγ
Estrogen Receptor β
Glucocorticoid Receptor
Thyroid Receptor
Retinoid X Receptor
PXR activation
CYP3A4
> 30b
> 30b
> 30
30
31
32
0.023
0.216
0.021
0.018
0.030
0.138
1.28
0.96
18.3
10.4
22.3
32.4
24.4
Testedc
> 30
33
34
0.080
0.420
0.040
0.083
CHO Cell Viability
CYP2C9
> 30
8.2
CYP2D6
> 30
CYP2C19
> 30
35
36
0.433
0.017
0.515
0.034
CYP1A2
> 30
CYP3A4 TDI
hERG
0.004 min−1
−4.1% @ 10 µM
a
Values are the mean of at least three experiments except where
noted.
b
c
Values are the mean of at least two experiments.
37
38
39
0.569
0.126
0.122
0.047
0.061
0.225
Moderate PXR activation at 10 µM (56% of positive control
Rifampicin) and no observed activation at 1 µM.
the individual PPARα and δ potencies were now switched in magni-
tude. Further consideration of compound 42 was terminated due to
poor oral exposure in rats (data not shown).
40
0.608
1.62
Though we succeeded in synthesizing compounds which were su-
perior or at least competitive with compound 11 in terms of dual
PPARα/ δ inhibition, inferior DMPK attributes precluded their further
development. Compound 11 was chosen for more advanced evaluation
and proof-of-concept studies because of its superior profile (Table 6).
When the sodium salt of compound 11 was orally administered as a
0.5% methocel solution at a dose of 10 mpk in rat, we observed good
drug exposure with a calculated plasma AUC of 14.2 µg*h/mL. Detailed
information for rat and dog oral and i.v. dosing is presented in Table 7.
a
Values are the mean of at least three experiments.
Table 5
CYP induction of selected compounds.
Compound
CYP Induction @ 10 µM; % activity
CYP3A4 mRNA increase vs.
Rifampicinb
vs. 10 µM Rifampicina
11
29
30
19.7
7.7
21%
83%
204%
Table 7
55.7
Pharmacokinetica Profile of Compound 11 in rats and dogs.
a
b
Average of 3 different donors.
Species %F
T1/2 i.v.
Cmax
AUCpo (h*µg/
Cl (mL/
VDss
(h)
(µg/mL)
mL)
min/kg)
(L/kg)
10 µM solutions of each compound were compared to 10 µM of Rifampicin
and values are expressed as percentage vs. Rifampicin-induced mRNA increase.
Rat
33.1 1.7
57.6 2.1
2.4
4.8
14.2
11.9
4.0
4.1
0.5
0.5
Dog
to potency on both isoforms. Benzoic acid 42 was prepared because this
class had proven to be very potent in previous stages of the project. As a
dual inhibitor, 42 was nearly as potent as 11, with the exception that
a
Compound was dosed at 2 mg/kg i.v. (water) and 10 mg/kg p.o. (0.5%)
methylcellulose in rat and 2 mg/kg i.v. (saline) and 5 mg/kg p.o. (0.5%) me-
thylcellulose in dog.
4

J.D. Jacintho et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
6
**
****
4
2
0
Vehicle
(Fed)
Vehicle
(Fasted)
11
11
(10 mg/kg)
(30 mg/kg)
FASTED
4
3
2
1
0
****
***
****
Fig. 4. B16F10 experimental metastasis model – Compound 11.
Vehicle
(Fed)
Vehicle
(Fasted)
11
11
(10 mg/kg)
(30 mg/kg)
FASTED
Fig. 3. PPAR gene attenuation in mouse livers. ACOT2 and CPT1a for
Compound 11.
These exposures will be used to guide future toxicity studies.
Given the favorable in vitro profile of 11 and suitable pharmaco-
kinetics, we wanted to evaluate its role in certain murine cancer
models. In order to determine the efficacious dose, an in vivo target
engagement study was designed.20 Compound 11 was administered
orally to mice at 10 and 30 mg/kg for 4 days. Mice were fasted for 48 h
prior to sacrifice to induce PPAR regulated genes. Fasting significantly
induced liver gene expression (Fig. 3) of ACOT2 (4.3-fold, p < .0001)
and CPT1a (2.7-fold, p < .0001). This increase was dose responsively
attenuated in animals treated with compound 11 at the 30 mg/kg dose.
Relative to higher species, where we saw good systemic exposure upon
oral dosing, higher dosing of compound 11 was anticipated in mice due
to relatively low systemic exposures (AUCpo ∼ 2.9 h*µg/mL, 10 mg/kg
in 0.5% methylcellulose) in addition to poorer intrinsic potency in
mouse PPARs (Table 6).
Fig. 5. SKOV-3 experimental xenograft model – Compound 11.
study period.24 These two preliminary in-vivo studies are certainly
limited in scope as we were unable to determine the differences be-
tween engaging the host (mouse) PPARs versus human tumor cell
PPARs. There are future plans for testing the differences between these
two. We also plan to test compound 11 in other cancer models.
In summary, we have, to the best of our knowledge, designed and
synthesized the first series of potent and selective PPARα/ δ dual an-
tagonists that are suitable for in vivo proof-of-concept experiments. Use
of previous SAR, which uncovered PPAR antagonist fibrate surrogates
enabled us to rapidly identify compound 11 while synthesizing only a
small library of compounds. With this selective tool in hand, it is now
possible to more clearly elucidate the impact of simultaneous antag-
onism of PPARα and PPARδ in various assays including certain types of
cancer. We will be reporting additional findings in this regard in due
course.
With target engagement and related systemic exposures in hand, we
next evaluated the ability of compound 11 to impart protective benefits
in two different tumor models at 30 mg/kg. First, we employed an ex-
perimental model of metastasis using the B16F10 cell line. It is postu-
lated that detached metastatic cells, resistant to anoikis, undergo a
radical shift in metabolism in order to survive.21 Non-glycolytic path-
ways, especially FAO, become important energy alternatives. Pertur-
bation of FAO via PPAR antagonism should reduce the survivability of
cancer cells in this model. As is evident from Fig. 4, compound 11
significantly attenuated the number of lung nodules compared to ve-
hicle treated animals when orally dosed.22
Acknowledgements
We would like to thank Lisa Rahbaek and Diem Vo for their con-
tributions during the manuscript review process.
We also analyzed the ability of compound 11 to inhibit the growth
of SKOV-3, a human ovarian cancer cell line (Fig. 5). It has been es-
tablished that ovarian cancer cells rely extensively on FAO for energy
due to an adipocyte-rich microenvironment.23 Oral administration of
11 significantly inhibited the growth of these tumors over a 28 day
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2018.12.045.
5

J.D. Jacintho et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
References
an overnight incubation, the treatment media was discarded and Luciferase Detection
Reagent (LDR) added. The luminescence intensity of light emission from the ensuing
luciferase reaction is directly proportional to the relative level of PPARα and δ ac-
tivation in the reporter cells. Luminescence was read using a Molecular Devices
FlexStation 3.
1. Bravo Y, Baccei CS, Broadhead A, et al. Bioorg Med Chem Lett. 2014;24:2267.
2. Stock, N.S.; Chen, A.C.; Bravo, Y.M.; Jacintho, J.D.; Baccei, J.M.; Stearns, B.A.; Clark,
R.C. WO2014/099503 A1.
20. Female C57Bl/6 mice, 6-7 weeks old, were treated with compound 11 at 10 and
30mg/kg. Compound 11 was formulated in .5% methylcellulose to yield a dosing
volume of 10 ml/kg and given 2 times per day for 3 consecutive days. A final dose
was given in the morning of day 4 and animals were sacrificed 4 hours post last dose.
Prior to sacrifice, mice were fasted for 48 hours beginning on day 2. Livers were
collected and RNA prepared using Qiazol and RNeasy RNA extraction kit and cDNA
was prepared using BioRad iScript cDNA synthesis kit (catalog #170-8891) according
to the manufacturer’s protocols. Finally gene expression was quantitated using ABI
Fast Advanced Master Mix (catalog #4444554) and predesigned primer probe sets
from IDT were used to evaluate gene expression changes. Data were normalized to
HPRT housekeeping gene.
3. Caro P, Kishan AU, Norberg E, et al. Cancer Cell. 2012;22:547.
4. Aboud OA, Wettersten HI, Weiss RH. PLoS One. 2013;8:e71115.
5. Schafer ZT, Grassian AR, Song L, Jiang Z, Gerhart-Hines Z, Irie HY, Gao S, Puigserver
P, Brugge JS. Nature. 2009;461:109 and references cited therein.
6. Biswas S, Lunec J, Bartlett K. Cancer Metastasis Rev. 2012;31:689.
7. Pang T, Kaufman A, Choi J, et al. Mol Clin Oncol. 2015;3:308.
8. Ito K, Carracedo A, Weiss D, et al. Nat Med. 2012;18:1350 and references cited
therein.
9. Etgen GJ, Mantlo N. Curr Top Med Chem. 2003;3:1649–1661.
10. Thulasiraman P, McAndrews DJ, Mohiudddin IQ. BMC Cancer. 2014;14:724 and
references cited therein.
21. Mason JA, Hagel KR, Hawk MA, Schafer ZT. Trends Cancer. 2017;3:475.
22. B16F10 experimental metastasis model: On day 0, female C57BL/6N mice were in-
jected intravenously with 50,000 B16F10 cells. Beginning on day 1, mice were orally
gavaged twice daily with either 11 (30 mg/kg) or vehicle control (0.5% methocel).
On day 21, mice were euthanized with isoflurane inhalation. Lungs were removed
and placed into 5 ml Fekete’s solution. After at least 24 hours in Fekete’s solution,
pulmonary nodules were enumerated. Nodules were scored according to size using
the following formula: Total score = 1a + 2b + 3c where R = diameter of tumor
nodule and a = number of nodules with R < 1mm, b = number of nodules where 1
mm ≤ R < 2 mm and c= number of nodules with R ≥ 2 mm.
11. De Rosa A, Pellegatta S, Rossil M, Tunici1 P, Magnoni L, Speranza MC, Malusa F,
Miragliotta V, Moril E, Finocchiaro G, Bakker A. PLoS One. 2012;7:e52113.
12. Zaveri NT, Sato BG, Jiang F, Calaoagan J, Laderoute KR, Murphy BJ. Cancer Biol
Ther. 2009;8:1252.
13. Elikkottil J, Kohli DR, Gupta K. Cancer Biol Ther. 2009;8:1262.
14. Ham SA, Yoo T, Hwang JS, et al. Am J Cancer Res. 2014;4:674.
15. Jeong E, Koo JE, Yeon SH, Kwak M-K, Hwang DH, Lee JY. Mol Carcinog.
2014;53:926.
16. Apostoli AJ, Nicol CJB. PPAR Res. 2012;2012:504918.
17. Pirat C, Farce A, Lebègue N, et al. J Med Chem. 2012;55:4027.
18. Pfeifer, L.A.; Ma, T.; Mantlo, N.B.; Michael, L.F.; Mokube, F.M.; Montrose-Rafizadeh,
C. U.S. Patent 2010/0099725.
23. Nieman KM, Kenny HA, Penicka CV, et al. Nat Med. 2011;17:1498.
24. SKOV-3 experimental xenograft model: Female athymic Fox1 nude mice were im-
planted with 2e6 SKOV-3 cells in a 1:1 mixture of SKOV-3 cells: matrigel. Cells were
injected subcutaneously into the left hind flank. Tumors were measured with calipers
and tumor volumes calculated using the formula: tumor volume= (length x width2)/
2. When tumor volumes were ≥ 60 mm3, mice were randomized into groups and
treatment with either vehicle (0.5% methocel PO, bid) or 11 (30 mg/kg PO, bid) was
initiated. Tumor volumes were determined twice weekly until study termination.
19. The screening of test compounds for activities against human PPARα and PPARδ
receptors were analyzed using commercial kits, Human PPARα and δ Reporter Assay
Systems (Indigo Biosciences, Cat. #IB00111 and IB00121). PPARα and PPARδ
Reporter cells were dispensed into 96-well assay plates and test compounds were then
immediately added to the wells. For the separate antagonist assays, PPARα agonist
GW7647 (20 nM) and PPARδ agonist GW0742 (3.3 nM) were also added. Following
6