An SAR study of hydroxy-trifluoromethylpyrazolines as inhibitors of Orai1-mediated store operated Ca2+ entry in MDA-MB-231 breast cancer cells using a convenient Fluorescence Imaging Plate Reader assay

Article abstract of DOI:10.1016/j.bmc.2018.05.012

The proteins Orai1 and STIM1 control store-operated Ca²⁺ entry (SOCE) into cells. SOCE is important for migration, invasion and metastasis of MDA-MB-231 human triple negative breast cancer (TNBC) cells and has been proposed as a target for cancer drug discovery. Two hit compounds from a medium throughput screen, displayed encouraging inhibition of SOCE in MDA-MB-231 cells, as measured by a Fluorescence Imaging Plate Reader (FLIPR) Ca²⁺ assay. Following NMR spectroscopic analysis of these hits and reassignment of their structures as 5-hydroxy-5-trifluoromethylpyrazolines, a series of analogues was prepared via thermal condensation reactions between substituted acylhydrazones and trifluoromethyl 1,3-dicarbonyl arenes. Structure-activity relationship (SAR) studies showed that small lipophilic substituents at the 2- and 3-positions of the RHS and 2-, 3- and 4-postions of the LHS terminal benzene rings improved activity, resulting in a novel class of potent and selective inhibitors of SOCE.

Full text of DOI:10.1016/j.bmc.2018.05.012

Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
An SAR study of hydroxy-trifluoromethylpyrazolines as inhibitors of
Orai1-mediated store operated Ca2+ entry in MDA-MB-231 breast cancer
cells using a convenient Fluorescence Imaging Plate Reader assay
Ralph J. Stevensona,⁎, Iman Azimib,c,d, Jack U. Flanagana,e, Marco Inserraf,g, Irina Vetterf,g,
⁎
Gregory R. Monteithb,c, William A. Dennya,e,
a
Auckland Cancer Society Research Centre, School of Medical Sciences, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand
The School of Pharmacy, The University of Queensland, Brisbane, Queensland, Australia
Mater Research Institute, The University of Queensland, Translational Research Institute, Brisbane, Queensland, Australia
Division of Pharmacy, College of Health and Medicine, University of Tasmania, Hobart, Tasmania, Australia
Maurice Wilkins Centre for Molecular Biodiscovery, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand
Institute for Molecular Bioscience, The University of Queensland, St Lucia, Queensland 4072, Australia
School of Pharmacy, The University of Queensland, Woolloongabba, Queensland 4102, Australia
b
c
d
e
f
g
A R T I C L E I N F O
A B S T R A C T
Keywords:
Orai1
The proteins Orai1 and STIM1 control store-operated Ca2+ entry (SOCE) into cells. SOCE is important for mi-
gration, invasion and metastasis of MDA-MB-231 human triple negative breast cancer (TNBC) cells and has been
proposed as a target for cancer drug discovery. Two hit compounds from a medium throughput screen, displayed
encouraging inhibition of SOCE in MDA-MB-231 cells, as measured by a Fluorescence Imaging Plate Reader
(FLIPR) Ca2+ assay. Following NMR spectroscopic analysis of these hits and reassignment of their structures as
5-hydroxy-5-trifluoromethylpyrazolines, a series of analogues was prepared via thermal condensation reactions
between substituted acylhydrazones and trifluoromethyl 1,3-dicarbonyl arenes. Structure-activity relationship
(SAR) studies showed that small lipophilic substituents at the 2- and 3-positions of the RHS and 2-, 3- and 4-
postions of the LHS terminal benzene rings improved activity, resulting in a novel class of potent and selective
inhibitors of SOCE.
Store-operated Ca2+entry (SOCE)
Ca2+ release-activated Ca2+ (CRAC) channel
Triple negative breast cancer
Inhibitor
Hydroxy-trifluoromethylpyrazoline
Structure-activity relationship (SAR)
1. Introduction
physiological functions. Orai1-mediated Ca2+ influx plays a key role in
the activation, differentiation and proliferation of T lymphocytes,1,10
The Orai family (Orai1, Orai2 and Orai3) of proteins, first identified
in 2006,1–5 are plasma membrane proteins which share no homology
with any other known calcium channel family. The Orai1 isoform is the
key protein of the pore which forms the Ca2+ release-activated Ca2+
(CRAC) channel, which is activated upon depletion of endoplasmic
reticulum (ER) Ca2+ stores. Reduced levels of ER Ca2+ are sensed by
stromal interaction molecule 1 (STIM1 and its homologue STIM2), and
prompt STIM1 to relocate and form puncta6 near the plasma mem-
brane. Protein-protein interactions with Orai1 then promote Ca2+ in-
flux via the pathway referred to as store-operated Ca2+ entry (SOCE).7
Elevation of intracellular Ca2+ rapidly induces dissociation of the
Orai1/STIM1 complex, preventing Ca2+ overloading and may represent
a SOCE gating mechanism.8,9 Orai1 is involved in a number of
lactation,11,12 cardiovascular/respiratory processes13 and pancreatic
cell function/dysfunction.14 However, Orai1, its homologue Orai3 and
activators STIM1 and STIM2 have also been implicated in tumor cell
proliferation, invasion, survival, migration and therapy resistance in
such cancers as prostrate, colorectal, melanoma, ovarian, cervical,
leukemia, nasopharyngeal, pancreatic, lung, renal, glioblastoma, neu-
roblastoma, epidermoid, hepatoma and pertinent to the current study -
breast cancer.15–19 Elevated levels of Orai1 and alterations in the re-
lative levels of STIM1/STIM2 are a feature of the poor prognosis of the
basal breast cancer subtype.12 It has been shown that blockage or
knockdown of Orai1 or STIM1 reduces enolase-1 dependent migration
of highly metastatic MDA-MB-231 triple negative basal-type breast
cancer (TNBC)20 cells, indicating a crucial role of SOCE in the
Abbreviations: FLIPR, Fluorescence Imaging Plate Reader; STIM1, stromal interaction molecule 1; ER, endoplasmic reticulum; TNBC, triple negative breast cancer; SAR, structure
activity relationship; LHS, left hand side; RHS, right hand side; THF, tetrahydrofuran; i-PrOH, isopropanol
⁎
Corresponding authors at: Auckland Cancer Society Research Centre, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand (R.J. Stevenson, W.A. Denny).
E-mail addresses: r.stevenson@auckland.ac.nz (R.J. Stevenson), b.denny@auckland.ac.nz (W.A. Denny).
https://doi.org/10.1016/j.bmc.2018.05.012
Received 29 March 2018; Received in revised form 1 May 2018; Accepted 8 May 2018
0968-0896/©2018ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:Stevenson,R.J.,Bioorganic&MedicinalChemistry(2018),https://doi.org/10.1016/j.bmc.2018.05.012

R.J. Stevenson et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
Fig. 1. Small molecule inhibitors of CRAC channels or SOCE reported in studies and new hydroxy-trifluoromethylpyrazolines.
modulation of cancer cell migratory and invasion properties.21
Orai1siRNA and STIM1 siRNA treatments or blockage of SOCE by a
small molecule inhibitor (SFK-96365, Fig. 1) in a xenograft mouse
model with MDA-MB-231 cells produced a significant antimetastatic
effect indicating the potential for SOCE as a therapeutic target in breast
cancer.22 Whilst Orai1 and STIM1 are widely expressed, CRAC channel
disorders appear to be limited with effects primarily in the immune
system, skeletal muscle and ectodermal- derived tissue, providing a
possible window of high selectivity and low toxicity for SOCE in-
hibitors.23 With emerging evidence of SOCE implicated in a variety of
diseases, the search and discovery of SOCE inhibitors is increasing and a
variety of small molecule inhibitors have now been reported19,24–26
opening the door for new therapeutic opportunities. Several small
molecule SOCE inhibitors have shown promise in both in vitro and in
vivo disease models (Fig. 1).
breast cancers.46 Patients with TNBC are characterised by absence of
the estrogen receptor, the progesterone receptor and HER2 and have a
poor prognosis compared to other subtypes.46 Thus there is an urgent
need for new selective targeted therapies for TNBC.47,48 Recently we
reported49 the use of a Fluorescence Imaging Plate Reader (FLIPR)
assay in the evaluation of 13 known compounds and a small series of
novel iminotriazoles as selective SOCE inhibitors in MDA-MB-231
TNBC cells. We now outline the synthesis and an extensive struc-
ture–activity relationship (SAR) study of a series of hydroxy-tri-
fluoromethylpyrazolines (Fig. 1) using this method.
2. Results and discussion
2.1. Chemistry
Although the aforementioned SFK-96365 is not selective for CRAC
channels with potential to block other Ca2+ channels,27 it does block
SOCE in mast,26 rat basophilic leukemia,28 Jurkat29 and cervical cancer
cells.30 2-APB (Fig. 1), although also nonselective, inhibits the pro-
liferation or migration of hepatoma,31 cervical,30 gastric32 and color-
ectal cells.33 YM-58483 (Fig. 1) in one study appears to be relatively
selective as it specifically inhibits SOCE and not other Ca2+ pathways in
T lymphocytes with an IC50 of 10 nM against SOCE in Jurkat T cells.34 It
has also been shown to inhibit cytokine secretion in mast cells.35 RO-
2959 (Fig. 1) is also thought to be relatively selective and has an IC50 of
400 nM against SOCE in rat basophilic leukemia-2H3 cells.36 5D (Fig. 1)
has been shown to inhibit CRAC channel activity by blocking ion per-
meation and diminishing the severity of experimental autoimmune
encephalomyelitis in mice by inhibition of differentiation of in-
flammatory T cells.37 The CRAC channel blocker Synta-66 (Fig. 1)
produced an IC50 of 1.4 μM in rat basophilic leukemia cells and Jurket T
cells, with demonstrated good selectivity and an absence of activity
against a range of other enzyme, receptor and ion channel targets.38
It has been demonstrated that Ca2+ is involved in a variety of cancer
cell signalling pathways39–41 and a recent review outlined the crucial
role of Orai and STIM proteins in the hallmarks of cancer.42 However,
there are currently no CRAC channel inhibitors in the marketplace for
treating cancer,43 although carboxyamidotriazole CAI (Fig. 1) did reach
a phase II clinical trial for patients with relapsed epithelial ovarian
cancer.44 CM2489, produced by CalciMedica with structure undi-
sclosed and in phase I clinical trials for plaque psoriasis, is claimed to be
the only specific CRAC channel inhibitor tested in patients.45
Although 1,3-diazoles and imidazoles are widely observed in nat-
ural products, the isomeric pyrazoles are somewhat rare.50 Never-
theless, there are a number of manuscripts outlining the potential use of
pyrazoles as pharmaceuticals: riboflavin synthase inhibitors,51 COX-2
inhibitors,52,53 sEH inhibitors,54 HIV-1 reverse transcriptase in-
hibitors,55 and NHE-1 inhibitors.56 The thermal condensation of hy-
drazines and 1,3-dicarbonyls produces pyrazoles and in some cases,
intermediate hydroxy-dihydropyrazolines can be isolated.50,51,57 From
a screen of 2140 purchased compounds,49 34 hydroxy-pyrazolines were
tested. The regio-isomers 5648159 and 6068529, based on the ven-
dor’s structures, were selected as hits as they displayed propitious in-
hibition of SOCE using the FLIPR assay in MDA-MB-231 cells (Fig. 2).
However, the 1H NMR (CDCl3) spectra for both 5648159 and 6068529
were absolutely identical, so a detailed NMR spectroscopy study was
undertaken in order to identify the correct structure. The diagnostic 5-
OH proton appeared at δ 6.14 ppm. Previous mechanistic and synthetic
Fig. 2. Vendor's provided structures for 5648159 and 6068529. 1H, 13C and 19
F
NMR spectral analysis revealed that both compounds have the structure given
for 6068529.
TNBC comprises 12–20% (of these ∼75% are basal-type) of total
2

R.J. Stevenson et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
studies by several groups50,51,57 on the regioisomeric formation of hy-
droxy-pyrazolines from condensation of acylhydrazines and 1,3-dike-
tones demonstrated that if the 5-OH proton was adjacent to a strong
electron withdrawing group (EWG) (e.g. CF3), there was a significant
downfield shift (5-OH appearing at δ 6.59–6.87 ppm) compared with
the OH moiety not adjacent to such a group (5-OH appearing at δ
5.20–5.58 ppm). Thus in the present case, the observed chemical shift
supports 6068529. In the 13C NMR (CDCl3) spectra, the observed signal
from C5 was a quartet at δ 92.2 ppm (JCF = 34.3 Hz) and the C3 atom
gave a singlet at δ 153.3 ppm. This also supports 6068529 for if the
structure was 5648159, the signal for the sp2-hybridised C3 atom
would be a quartet with a significant downfield shift (δ 140–150 ppm)51
from the value observed at 92.2 ppm, i.e. the CF3 group is attached to an
sp3-hybridised carbon rather than the C]N carbon atom. In the 19F
NMR (CDCl3) spectra, the signal at δ −6.02 ppm (CF3) upfield from
CF3COOH (external standard) also supports 6068529 as the structure.
It has been previously demonstrated that δ −4 to −3 ppm (CF3) re-
lative to CF3COOH implies that a CF3 moiety is attached to an sp3-
hybridised carbon atom.51 If CF3 is attached to an sp2-hybridised
carbon atom then δ > +10 ppm (CF3) downfield from CF3COOH
would be observed.51 It was concluded that both 5648159 and
6068529 had the given 6068529 structure (Fig. 2).
The new hydroxy-5-trifluoromethylpyrazolines and hydroxy-3-tri-
fluoromethylpyrazolines were prepared by the thermal condensa-
tion50,51,57 of 2-phenylacetohydrazide (1) or modified hydrazides
(44–46, 69)58 with phenyl- substituted 4,4,4-trifluoro-1-phenylbutane-
1,3-diones (3) (Schemes 1–4), which in turn were prepared from sub-
stituted acetophenones (2) and ethyl trifluoroacetate in the presence of
NaH (Scheme 1).50–52,57,59 In the majority of reactions, the only re-
gioisomer isolated was the one with a CF3 moiety attached to an sp3-
hybridised carbon atom and hence geminal to an OH group (Tables
1–3). In our attempts to synthesis both hit compounds, only a com-
pound (14) with 1H NMR spectrum identical in all respects to 6068529
was isolated. However, in a few cases the alternative regioisomer where
the CF3 group is attached to an sp2-hybridised carbon was either solely
or also isolated (Table 4). The mechanism and regioselectivity of the
condensation reaction between 1,3-diketones and hydrazines resulting
in hydroxy-pyrazolines and pyrazoles has been previously stu-
died.50,51,57,60 It is well documented that in numerous instances, such
reactions do not proceed with a defined regio-chemistry. In a number of
examples, it was demonstrated that in reactions of acylhydrazones with
1,3-diketones F3CCOCH2COAr under neutral conditions, stable hy-
droxy-pyrazolines can be isolated.50,51,57,60 Under acidic conditions or
in the absence of a strong terminal EWG (e.g. CF3), the reactions gen-
erally proceed to pyrazoles. Proposed partial mechanisms for the for-
mation of regio-isomers hydroxy-5-trifluoromethylpyrazolines and hy-
droxy-3-trifluoromethylpyrazolines are outlined in Fig. 3. It has
previously been discovered in such condensation reactions that aryl
Scheme 2. Synthesis of the 5-hydroxy-5-trifluoromethylpyrazolines of Table 2.
Reagents and conditions: (i) i-PrOH (90 °C over 48–96 h).
Scheme 3. Synthesis of the 5-hydroxy-5-trifluoromethylpyrazolines of Table 3.
Reagents and conditions: (i) i-PrOH (90 °C over 48–96 h).
Scheme 4. Synthesis of the 5-hydroxy-3-trifluoromethylpyrazolines of Table 4.
Reagents and conditions: (i) i-PrOH (90 °C over 48–96 h).
electrophilic and hence more reactive carbonyl group (possibility B in
Fig. 3) due to the strongly electron withdrawing CF3 moiety. This may
explain why in some of our reactions, hydroxy-3-tri-
fluoromethylpyrazolines were isolated (Scheme 4). Singh and co-
workers62 have noted that aryl trifluoromethyl-β-diketones exist sub-
stantially in the enol form providing an AreC]CeC]O extended
conjugated system. Thus they have postulated that the first step is
possibly conjugate addition of nitrogen onto the enol, eventually
leading to the isolated products. It has previously been observed that in
the crystalline state, products of possibility B exist entirely as hydroxy-
3-trifluoromethylpyrazolines. However in CDCl3, they exist as equili-
brium mixtures of open hydrazones and cyclic tautomers57 (possibility
1,3-diketones possessing
a COCF3 moiety exist predominantly as
hemiketals due to nucleophilic addition of solvent,61 in our examples i-
PrOH (possibility A in Fig. 3). Formation of a trifluoromethylated
hemiketal may explain why the majority of final products isolated were
hydroxy-5-trifluoromethylpyrazolines (Schemes 1–3). However, the
trifluoromethylated carbonyl group might be expected to be the more
B
in Fig. 3) and in a
few 1H NMR spectra of hydroxy-3-tri-
fluoromethylpyrazolines, peaks appeared noticeably broader. Both
possible covalently hydrated pyrazoles are stabilized by the electron
withdrawing amide and CF3 moieties. Hence, hydroxy-5-tri-
fluoromethylpyrazolines and hydroxy-3-trifluoromethylpyrazolines can
be isolated as stable entities.
In the present work, we first synthesised the hit compound (14) and
then modified the RHS of the scaffold (Scheme 1, Table 1). The left
hand side (LHS) was then modified whilst the RHS was kept relatively
constant by retaining 2-CH3 or 3-CH3 moieties (Scheme 2, Table 2). The
SAR was extended further by synthesising a series of pyrazolines where
Scheme 1. Synthesis of the 5-hydroxy-5-trifluoromethylpyrazolines of Table 1.
Reagents and conditions: (i) NaH/THF, then ethyl trifluoroacetate (0 °C to r.t.
over 24 h); (ii) i-PrOH (90 °C over 48–96 h).
3

R.J. Stevenson et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
Table 1
Table 2
Yields and biological properties of 5-hydroxy-5-trifluoromethylpyrazolines.
Yields and biological properties of 5-hydroxy-5-trifluoromethylpyrazolines;
LHS modifications.
a
a
a
No X or LHS
R
Yield % IC50a (µM) ER release IC50 (CI) (µM) SOCE
No
R
Yield % IC50 (µM) ER release
IC50 (CI) (µM) SOCE
(Peak 1)
(Peak2)
(Peak 1)
(Peak2)
47 2-OCH3
48 2-OCH3
49 3-OCH3
50 3-OCH3
51 4-OCH3
52 4-OCH3
53 2-CH3
54 2-CH3
55 3-CH3
56 3-CH3
57 4-CH3
58 4-CH3
59 2-Cl
60 2-Cl
61 3-Cl
62 3-Cl
63 4-Cl
64 4-Cl
65 45
66 45
67 46
2-CH3 12
3-CH3 32
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
nab
60% at 100 µMc
50% at 100 µMc
60% at 100 µMc
32 (24–43)
30% at 100 µMc
33 (30–37)
20% at 100 µMc
nab
4
5
6
7
8
9
H
71
15
14
16
33
19
37
8
58
94
48
57
92
74
71
97
92
88
74
82
nab
nab
2-CH3
2-OCH3
2-F
2-Cl
2-Br
nab
50% at 30 µMc
25% at 100 µMc
30% at 100 µMc
33 (26–41)
30 (26–36)
nab
nab
2-CH3
7
nab
3-CH3 48
2-CH3 30
3-CH3 16
2-CH3 28
3-CH3 50
2-CH3 14
3-CH3 72
2-CH3 20
3-CH3 50
2-CH3 49
3-CH3 73
2-CH3 19
3-CH3 49
2-CH3 44
3-CH3 63
2-CH3 18
3-CH3 50
2-CH3 28
3-CH3 57
50% at 100 µMc
nab
10 2-OCF3
11 2-NO2
12 2-aza
13 3-CH3
14 3-OCH3
15 3-O(CH2)2W
16 3-F
17 3-Cl
18 3-Br
19 3-OCF3
20 3-NO2
21 3-aza
22 3-CN
23 3-CF3
24 3-NHSO2CH3 36
25 3-NHCOCF3
26 4-CH3
27 4-OCH3
28 4-F
29 4-Cl
30 4-Br
31 4-OCF3
32 4-NO2
33 4-aza
nab
nab
40% at 100 µMc
20% at 100 µMc
36 (31–43)
37% at 100 µMc
50% at 100 µMc
60% at 30 µMc
48 (34–66)
47 (23–100)
50% at 100 µMc
89 (55–142)
nab
nab
42 (32–55)
45 (38–54)
55 (43–71)
42 (32–54)
40 (35–45)
42 (35–50)
79 (69–91)
75 (64–88)
47 (38–59)
60 (40–91)
105 (74–149)
nab
nab
nab
50% at 100 µMc
nab
nab
nab
nab
nab
nab
nab
nab
nab
29 (24–35)
20% at 30 µMc
60% at 30 µMc
nab
nab
nab
68 46
nab
82
69
10
31
69
83
43
55
49
70
30
33
68
14
43
34
45
47
32
nab
nab
a
nab
nab
IC50 is drug concentration for a half-maximal response.
Less than 10% inhibition at 100 µM; na = not active.
Estimate of activity where 100% inhibition not reached. CI = 95%
nab
nab
b
c
nab
nab
nab
nab
Confidence Intervals (n = 4).
nab
nab
nab
nab
nab
nab
Table 3
34 4-CN
35 4-NHCOCH3
36 4-W
37 2,6-diOCH3
38 2,3-diF
39 2,6-diF
40 2,4-diCl
41 2,5-diCl
42 3,4-(CH2)4
nab
nab
Yields and biological properties of 5-hydroxy-5-trifluoromethylpyrazolines;
LHS modifications.
nab
nab
nab
nab
nab
nab
nab
80% at 100 µMc
nab
nab
nab
58 (47–71)
54 (44–66)
89 (62–129)
nab
nab
nab
43
X
nab
a
b
c
a
a
IC50 is drug concentration for a half-maximal response.
Less than 10% inhibition at 100 µM; na = not active.
Estimate of activity where 100% inhibition not reached. CI = 95%
No
X
R
Yield % IC50 (µM) ER release
IC50 (µM) SOCE
(Peak 1)
(Peak2)
70
71
H
H
2-CH3 18
3-CH3 37
nab
nab
ntd
20% at 100 µMc
30% at 100 µMc
ntd
Confidence Intervals (n = 4). W = morpholide, X = 2-thiophene.
72 2-I 3-CH3 17
the LHS had been modified producing phenylmethanones (Scheme 3,
Table 3). Although we were unable to isolate
fluoromethylpyrazoline corresponding to the structure for 5648159
(Fig. 2) assigned by the vendor, from some reactions we did isolate and
characterise several hydroxy-3-trifluoromethylpyrazolines (Scheme 4,
Table 4). Hydroxy-3-trifluoromethylpyrazolines were able to be dis-
a
a hydroxy-3-tri-
IC50 is drug concentration for a half-maximal response.
Less than 10% inhibition at 100 µM; na = not active.
Estimate of activity where 100% inhibition not reached.
nt = not tested.
b
c
d
tinguished
from
their
regio-isomeric
hydroxy-5-tri-
64% yield versus 17% for the 5-isomer 72.
fluoromethylpyrazolines by the diagnostic 1H NMR 5-OH proton signal
appearing upfield at δ 4.95–5.30 ppm. In fact, in the reaction of 1 with
3 containing a 2-CF3 substituted phenyl group the only product isolated
was a hydroxy-3-trifluoromethylpyrazoline (73) (Scheme 4, Table 4).
The major product of the reaction of 1 with 3 containing a 2-NO2
substituted phenyl group was a hydroxy-3-trifluoromethylpyrazoline
Overall, the yields of products varied from 7 to 97% (Tables 1–4),
and appeared to be influenced by the nature and regio-chemistry of the
aryl substituents at both ends of the pyrazolines. None of the reactions
involving 2-substituted phenyl groups on the RHS provided high
yielding products. This may be a steric hindrance issue as a number of
products containing both 3-substituted and 4-substituted phenyl groups
were isolated in high yield. Most notably, lower yields could sometimes
be attributed to thermally-promoted dehydration reactions producing
(74)
(Scheme
4,
Table
4).
Strikingly,
hydroxy-3-tri-
fluoromethylpyrazoline 78 (3-CH3, RHS; 2-I, LHS) was produced in
4

R.J. Stevenson et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
Table 4
internal stores and we designated this as peak 1 (red) in the normalised
2-peak FLIPR plots (% inhibition vs log10[M]). Ideally peak 1 should
remain flat with increasing concentration of introduced compounds
implying such compounds do not have pronounced non-specific direct
effects on ER Ca2+ channels. The “2-peak assessment” allowed us to
promptly identify analogues that were clearly not SOCE selective by
observing if there was appreciable potency against peak 1.51 Store de-
pletion was promoted by selective chelation of extracellular Ca2+ with
1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid) (BAPTA).
Addition of extracellular Ca2+ resulted in SOCE which we designated as
peak 2 (black). A potent and selective SOCE inhibitor would cause peak
2 to reach 100% inhibition at a low concentration whilst peak 1 would
remain flat in the 2-peak plots. We have previously demonstrated that
peak 2 is reduced by silencing of Orai1 but not the related isoforms
Orai2 and Orai3.12,49
The purchased hit compound 5648159 provided an estimated 37%
inhibition of SOCE at 100 μM in MDA-MB-231 cells as indicated in peak
2 (Fig. 2) and importantly it was selective as peak 1 remained un-
changed. Having the same structure (3-OCH3, RHS), the purchased
6068529 (Fig. 2) and synthesised 14 gave identical 2-peak plots
(Table 1). A variety of related structures were also evaluated. If a po-
tential inhibitor did not reach 80–100% in either peak 1 or peak 2 an
IC50 was not assigned for either peak 1 or peak 2 and if a compound
showed less than 10% inhibition at 100 μM (even if it exhibited activity
above 100 μM) it was designated as inactive against the relevant peak.
Compound 14 did not reach the threshold in peak 2 and was therefore
Yields and biological properties of 5-hydroxy-3-trifluoromethylpyrazolines; al-
ternative isomers.
a
a
No LHS
R
Yield% IC50 (CI) (µM) ER
IC50 (CI) (µM)
release (Peak 1)
SOCE (Peak2)
73
74
1
1
2-CF3
34
80 (59–108)
49 (40–60)
30% at 100 µMc
20% at 100 µMc
20% at 10 µMc
nab
2-NO2 40
2-CH3 25
2-CH3 10
3-CH3 18
20% at 100 µMc
75 46
76 69 (X = H)
77 69 (X = H)
nab
nab
nab
ntd
78 69 (X = 2-I) 3-CH3 64
ntd
a
IC50 is drug concentration for a half-maximal response.
Less than 10% inhibition at 100 µM; na = not active.
Estimate of activity where 100% inhibition not reached. CI = 95%
b
c
Confidence Intervals (n = 4).
d
nt = not tested.
aromatic pyrazoles which rapidly lose the LHS acyl groups, due to the
pyrazole moiety being a powerful leaving group in the presence of the
nucleophilic solvent (i-PrOH).63
not allocated an IC50
.
Several of the thirteen 3-substituted aryl (RHS) synthesised analo-
gues inhibited SOCE (peak 2) without non-specific effects on peak 1
(Table 1). Most notably, compound 23 (3-CF3, RHS) provided an IC50 of
29 μM and 100% inhibition at 100 μM against peak 2 and was inactive
against peak 1 (Figs. 4 and S1). Two 3-substituted halides, compound
17 (3-Cl, RHS) and compound 18 (3-Br, RHS) were both selective and
active against SOCE providing IC50s of 48 and 47 μM respectively and
100% inhibition at 100 μM against peak 2 and were inactive against
peak 1 (Figs. 4 and S1). Compound 13 (3-CH3, RHS) was also active
against SOCE (IC50 36, 100% inhibition at 100 μM against peak 2) but
2.2. Biology and structure-activity relationships
The FLIPR Ca2+ assay is a convenient procedure for measuring in-
tracellular free Ca2+ and we have previously validated this assay for
assessment of Orai1-mediated Ca2+ influx in MDA-MB-231 cells.12,49
The detailed procedure is described in the experimental section. In
brief, addition of the sarco/endoplasmic reticulum ATPase (SERCA)
inhibitor cyclopiazonic acid (CPA) induced a gradual loss of Ca2+ from
Fig. 3. Proposed partial mechanisms for formation of two possible hydroxy-trifluoromethylpyrazoline regio-isomers.
5

R.J. Stevenson et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
Fig. 4. Normalised FLIPR plots of compounds 17, 18, 23, 56, 60 and 62 demonstrating SOCE inhibition (black) and selectivity against ER release (red). (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
with some inhibition of peak 1 at 100 μM. Compounds 16 (3-F, RHS),
19 (3-OCF3, RHS), 20 (3-NO2, RHS) and 25 (3-NHCOCF3, RHS)
showed ∼ 60% inhibition selectively at 100 μM against peak 2. Com-
pound 15 (3-O(CH2)2morph, RHS) also provided 60% inhibition at
100 μM against peak 2 but was not selective.
Amongst the eight 2-substituted aryl (RHS) synthesised analogues,
compounds 5 (2-CH3, RHS) and 9 (2-Br, RHS) were the most selectively
potent showing ∼ 60% inhibition at 100 μM against peak 2. Compound
8 (2-Cl, RHS) showed 80% inhibition at 100 μM against peak 2 but was
clearly not selective. Compounds 6 (2-OCH3, RHS), 7 (2-F, RHS) and 11
(2-NO2, RHS) provided ∼25% inhibition selectively at 100 μM against
peak 2.
None of the eleven 4-substituted aryl (RHS) compounds demon-
strated any significant SOCE inhibition (Table 1). The best were com-
pounds 33 (4-aza, RHS) and 36 (4-morph, RHS) showing 50% selective
SOCE inhibition at 300 μM but overall their activity was poor. Since
lack of activity is not due to Log P or solubility (Table S1), we postulate
that steric bulk of 4-substituents may be the cause: 4-substitution is not
well tolerated for SOCE inhibition unlike 2- and 3-substitutions where
activity is improved.
RHS; 2-CH3, LHS), 56 (3-CH3, RHS; 3-CH3, LHS, Fig. 4), 57 (2-CH3,
RHS; 4-CH3, LHS), 58 (3-CH3, RHS; 4-CH3, LHS), 59 (2-CH3, RHS; 2-Cl,
LHS), 60 (3-CH3, RHS; 3-Cl, LHS, Fig. 4), 63 (2-CH3, RHS; 4-Cl, LHS)
and 64 (3-CH3, RHS; 4-Cl, LHS) all provided respectable IC50’s
(33–60 μM) although some were less selective at higher concentrations
(Table 2).
None of the compounds listed in Table 3 (70–72) or Table 4 (73–78)
appeared to be particularly effective SOCE inhibitors in the FLIPR
assay.
The original hit compound 5648159 was also screened in the FLIPR
assay for its ability to inhibit the TRPV1, TRPM8, CaV1.2/ CaV1.3 (L-
type) and CaV2.2 (N-type) calcium channels. This compound showed no
activity up to 100 µM against these channels, implying some selectivity
for SOCE inhibition. 5648159 exhibited low inhibition of hERG (9.1%
at 10 µM), T1/2 (min) 8.2 in human liver microsomes, T1/2 (min) 3.3 in
mouse liver microsomes, a solubility of M(aMEM) = 16.8 µg/mL, a
stability of M1/M0 (H2O) = 93.2% and log P = 4.4.
2.3. Conclusions
The outcomes of the 7 di-substituted and miscellaneous (RHS) were
more promising (Table 1). Compounds 38 (2,3-diF, RHS), 40 (2,4-diCl,
RHS) and 41 (2,5-diCl, RHS) all provided IC50’s of ∼55 μM and ∼60%
inhibition at100 μM against peak 2. Compound 42 (3,4-(CH2)4, RHS)
provided and an IC50 of 89 μM and 50% inhibition at100 μM selectively
against peak 1.
Modification of the LHS (Table 2) proved fruitful as 12 of the 22
compounds synthesised provided > 80% inhibition of SOCE as evi-
denced by peak 2 in the FLIPR plots. Compounds 61 (2-CH3, RHS; 3-Cl,
LHS) and 62 (3-CH3, RHS; 3-Cl, LHS, Fig. 4) demonstrated good se-
lectivity and IC50’s of 79 μM and 75 μM respectively. Compounds 50 (3-
CH3, RHS; 3-OCH3, LHS), 52 (3-CH3, RHS; 4-OCH3, LHS), 55 (2-CH3,
With an urgent need for new selective targeted therapies for TNBC
and the potential use of SOCE inhibitors in a variety of diseases, the
discovery of a number of new active and potentially selective SOCE
inhibitors (e.g. 18, 23, 60 and 63) using our FLIPR Ca2+ assay in MDA-
MB-231 cells is significant. Seizing on hits from a medium throughput
screen and using a relatively simple synthetic strategy, we were able to
observe SAR on a series of hydroxy-5-trifluoromethylpyrazolines and
evaluate them for SOCE inhibition with the convenient FLIPR assay.
Regarding the RHS of the hydroxy-5-trifluoromethylpyrazoline scaf-
fold, it was clear that 2- (e.g. 9) and 3-substitutions (e.g. 18) on the aryl
moiety provided superior activity/selectivity. Modifying the LHS
proved profitable with activity/selectivity maintained with a variety of
6

R.J. Stevenson et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
substitutions round the aryl ring (e.g. 56, 60 and 63). Future studies
will involve expanding the SAR and selection of the most active/se-
lective analogues for patch-clamping studies (ICRAC) and in vivo eva-
luations in tumor-containing mouse models. Given that the potency of
YM-58483 on inhibition of SOCE is greatly increased by extending the
period of pre-incubation,34 the effect of increased pre-incubation and
also the reversibility of the SOCE inhibitors developed here should be
assessed in future studies.
NaHCO3, 140 mM NaCl, 11.5 mM glucose, pH 7.3). Cells were then
treated for 15 min at room temperature with different concentrations of
compound in a solution containing 5% (v/v) PBX Signal Enhancer and
500 μM probenecid in PSS. For assessment of store-operated Ca2+ entry
(SOCE), which has been previously shown to be mediated by Orai1
proteins in MDA-MB-231 cells,66 the following solutions in PSS were
added in order inside the FLIPRTETRA using a robotic arm: 500 µM
BAPTA (Invitrogen, Carlsbad, CA, USA) for chelation of extracellular
Ca2+; 10 µM cyclopiazonic acid (CPA; Sigma-Aldrich, St Louis, MO) for
inhibition of sarco/endoplasmic reticulum Ca2+-ATPases (SERCA)
pump and depletion of endoplasmic reticulum (ER);67 0.5 mM CaCl2
(700 s after the addition of CPA) for assessment of store-operated Ca2+
influx. Fluorescence was measured at 470–495 nm excitation and
515–575 nm emission. ScreenWorks Software (v2.0.0.27, Molecular
Devices) was used for data analyses. Ca2+ levels were assessed through
the change in relative fluorescence of the Ca2+ dye. The percentage
inhibition of the maximum peak height for each concentration of each
compound normalized to its corresponding DMSO control was calcu-
lated and plotted separately for peak 1 (a measure of the release of
endoplasmic reticulum calcium store) by addition of CPA to assess
potential non-specific effects on Ca2+ homeostasis and peak 2 (a
measure of store-operated Ca2+ influx).26,68 Where the% inhibition of
peak 2 (SOCE) did not exceed 10% at 100 µM in initial assessments,
these compounds were defined as not active.
3. Experimental
3.1. Chemistry
Final products were analysed by reverse-phase HPLC (Alltima C18
5 µm column, 150 × 3.2 mm; Alltech Associated, Inc., Deerfield, IL)
using an Agilent HP1100 equipped with a diode-array detector. Mobile
phases were gradients of 80% CH3CN/20% H2O (v/v) in 45 mM
NH4HCO2 at pH 3.5 and 0.5 mL/min. Purity was determined by mon-
itoring at 330
50 nm and was ≥95% for all final products.
Combustion analyses were carried out in the Campbell Microanalytical
Laboratory, University of Otago, Dunedin, New Zealand. High resolu-
tion mass spectra (HRMS) were measured on an Agilent Technologies
6530 Accurate-Mass Quadrupole Time of Flight (Q-TOF) LCMS inter-
faced with an Agilent Jet Stream Electrospray Ionization (ESI) source
allowing positive or negative ions detection. Melting points were de-
termined on an Electrothermal 2300 melting point apparatus. NMR
spectra were obtained on a Bruker Avance 400 spectrometer at
400 MHz for 1H and 100 MHz for 13C spectra. The 19F chemical shifts
were reported relative to CF3COOH (external standard).
3.2.3. Selectivity screen
TRPV1 and TRPM8 responses were assessed in HEK293 cells
(American Tissue Culture Collection, Manassas, VA, USA) 48 h after
transfection with plasmid DNA of rTRPV1 (D. Julius, Department of
Physiology, University of California, Berkeley, CA, USA) or rTRPM8 (P.
Reeh, Department of Anesthesiology, Friedrich-Alexander-University,
Erlangen-Nuremberg, Erlangen, Germany) using Lipofectamine 2000 as
previously described.69 CaV2.2 responses were assessed in SH-SY5Y
neuroblastoma cells in the presence of nifedipine (10 µM) according to
established protocols.70 HEK293 cells were routinely maintained in
DMEM containing 10% foetal bovine serum, 2 mM l-glutamine, pyr-
idoxine and 110 mg/ml sodium pyruvate. SH-SY5Y cells (European
Collection of Authenticated Cell Cultures, Salisbury, UK) were cultured
in RPMI 1640 antibiotic-free medium (Invitrogen) supplemented with
15% heat-inactivated FBS and 2 mM GlutaMAX™ (Invitrogen). Cells
were split every 3–6 days in a ratio of 1:5 using 0.25% trypsin/EDTA.
Cells were plated on 384-well black-walled imaging plates (Corning) at
a density of 10,000 cells/well (HEK293) or 50,000 cells/well (SH-SY5Y)
and used for Ca2+ experiments 48 h after plating. Growth media was
removed and replaced with 20 µl/well Calcium 4 No-Wash dye diluted
according to the manufacturer’s instructions in physiological salt solu-
tion (PSS; NaCl 140 mM, glucose 11.5 mM, KCl 5.9 mM, MgCl2 1.4 mM,
NaH2PO4 1.2 mM, NaHCO3 5 mM, CaCl2 1.8 mM, HEPES 10 mM) and
incubated for 30 min at 37 °C/5% CO2. Ca2+ responses were measured
using a FLIPRTETRA (Molecular Devices, Sunnyvale, CA, USA) fluor-
escent plate reader with excitation at 470–495 nM and emission at
515–575 nM. Camera gain and intensity were adjusted for each plate to
yield a minimum of 1500–2000 arbitrary fluorescence units (AFU)
baseline fluorescence. Test compounds were added 300 s prior to sti-
mulation with capsaicin (100 nM; TRPV1), menthol (100 µM, TRPM8)
and KCl (90 mM)/CaCl2 (5 mM; CaV2.2). Data was analysed using
Screenworks 3.2 and FLIPRTETRA data was plotted using GraphPad
Prism™ software (Version 6.00).
General
method
for
the
synthesis
of
hydroxy-tri-
fluoromethylpyrazolines (chemical data for all compounds are located
in Supporting Information).
3.1.1. 1-(5-Hydroxy-3-phenyl-5-(trifluoromethyl)-4,5-dihydro-1H-
pyrazol-1-yl)-2-phenylethan-1-one (4)
A mixture of 2-phenylacetohydrazide (1) (0.10 g, 0.67 mmol) and
1,1,1-trifluoro-5-phenylpentane-2,4-dione (3a) (0.14 g, 0.67 mmol) in a
solution of i-PrOH (5 mL) was heated at 90 °C for 48 h. After cooling to
room temperature, EtOAc and water were added. The EtOAc extract
was washed with water, brine and dried (Na2SO4). Flash chromato-
graphy (petroleum ether/EtOAc; 100:0 to 93:7) followed by re-
crystallization from Et2O/petroleum ether gave 4 (0.17 g, 71%), mp
122–123 °C (Et2O/petroleum ether); 1H NMR (CDCl3) δ 7.74–7.69 (m,
2H), 7.52–7.43 (m, 3H), 7.37–7.31 (m, 4H), 7.29–7.24 (m, 1H), 6.16 (s,
1H), 4.17 (d, J = 14.9 Hz, 1H), 4.11 (d, J = 14.9 Hz, 1H), 3.68 (d,
J = 18.6 Hz, 1H), 3.53 (dd, J = 18.6, 1.3 Hz, 1H). Anal. calcd. for
C
18H15F3N2O2: C, 60.07; H, 4.34; N, 8.04. Found: C, 61.82; H, 4.34; N,
8.02.
3.2. Biology
3.2.1. Compound library development
A ligand based virtual screening approach used in two phases for the
discovery of SOCE inhibitors was previously described.49,64
3.2.2. Measurement of intracellular free Ca2+ by FLIPR assay
MDA-MB-231 TNBC cells were plated at a density of 2 × 103 cells
per well in 384-well black plates (Corning Costar, Cambridge, MA, USA).
Three days post seeding, intracellular free Ca2+ levels were measured
in a fluorescence imaging plate reader (FLIPRTETRA, Molecular Devices,
Sunnyvale, CA, USA) using the PBX Calcium Assay Kit (640175, BD
Biosciences, Franklin Lakes, NJ, USA) as described previously.65 Briefly,
cells were first loaded for 1 h at 37 °C with a dye-loading solution
comprising of 2 µM PBX Calcium Assay dye, 5% (v/v) PBX Signal En-
hancer and 500 μM probenecid in physiological salt solution (PSS,
5.9 mM KCl, 1.4 mM MgCl2, 10 mM HEPES, 1.2 mM NaH2PO4, 5 mM
Acknowledgements
The authors gratefully acknowledge funding from QUE Oncology
Inc and the University of Auckland Biopharma Initiative. This work was
supported by an Australian Research Council Future Fellowship
FT130101215 awarded to I.V. We thank Karin Tan and Sisira Kumara
for HPLC analysis.
7

R.J. Stevenson et al.
Bioorganic&MedicinalChemistryxxx(xxxx)xxx–xxx
Conflict of interests
32. Sakakura C, Miyagawa K, Fukuda K, et al. Gan Kagaku Ryoho. 2003;30:1784.
33. Wang JY, Sun J, Huang MY, et al. Oncogene. 2015;34:4358.
34. Zitt C, Strauss B, Schwarz EC, et al. J Biol Chem. 2004;279:12427.
35. Law M, Morales JL, Mottram LF, et al. Int J Biochem Cell Biol. 2011;43:1228.
36. Chen G, Panicker S, Lau K-Y, et al. Mol Immunol. 2013;54:355.
37. Kim KD, Srikanth S, Tan YV, et al. J Immunol. 2014;192:110.
38. Di Sabatino A, Rovedatti L, Kaur R, et al. J Immunol. 2009;183:3454.
39. Monteith GR, McAndrew D, Faddy HM, Roberts-Thomson SJ. Nat Rev Cancer.
2007;7:519.
40. Prevaskaya N, Skryma R, Shuba Y. Nat Rev Cancer. 2011;11:609.
41. Monteith GR, Davis FM, Roberts-Thomson SJ. J Biol Chem. 2012;287:31666.
42. Fiorio Pla A, Kondratska K, Prevaskaya N. Am J Physiol-Cell Ph. 2016;310:C509.
43. Tian C, Du L, Zhou Y, Li M. Future. Med Chem. 2016;8:817.
44. Hussain MM, Kotz H, Minasian L, et al. J Clin Oncol. 2003;21:4356.
45. Peverello P, Cainarca S, Liberati C, Tarroni P, Piscitelli F, Severi E. Pharm Pat Anal.
2014;3:171.
G.R.M and W.A.D are associated with QUE Oncology Inc.
Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.bmc.2018.05.012.
References
1. Feske S, Gwack Y, Prakriya M, et al. Nature. 2006;441:179.
2. Vig M, Peinelt C, Beck A, et al. Science. 2006;312:1220.
3. Taylor CW. Trends Biochem Sci. 2006;31:597.
46. Anders CK, Carey LA. Clin Breast Cancer. 2009;9:S73.
47. Crown J, O’Shaughnessy J, Gullo G. Ann Oncol. 2012;23:6.
48. Wahba HA, El-Hadaad HA. Cancer. Biol Med. 2015;12:106.
49. Azimi I, Flanagan JU, Stevenson RJ, et al. Bioorg Med Chem. 2017;25:440.
50. Norris T, Colon-Cruz R, Ripin DHB. Org Biomol Chem. 2005;3:1884.
51. Zhao Y, Bacher A, Illarionov B, et al. J Org Chem. 2009;74:5297.
52. Gao M, Wang M, Miller KD, Zheng Q-H. Euro J Med Chem. 2011;46:4760.
53. Lin J, Lundy deMello KM, et al. Bioorg Med Chem Lett. 2010;20:6379.
54. Lo YH, Man CC, et al. Bioorg Med Chem Lett. 2004;14:95.
55. Genin MJ, Biles C, Keiser BJ, et al. J Med Chem. 2000;43:1034.
56. Guzman-Perez A, Wester RT, et al. Bioorg Med Chem Lett. 2001;11:803.
57. Pakal'nis VV, Zerova IV, Yakimovich SI. Russ J Gen Chem. 2007;77:1732.
58. Hanif Khan M, Rama NH, Noreen S, Choudhary MI, Jones PG, Iqbal M. Med Chem
Res. 2012;21:3885.
4. Zhang SL, Yeromin AV, Zhang XH, et al. Proc Natl Acad Sci USA. 2006;103:9357.
5. Yeromin AV, Zhang SL, Jiang W, Yu Y, Safrina O, Cahalan MD. Nature. 2006;443:226.
6. Zhang SL, Yu Y, Roos J, et al. Nature. 2005;437:902.
7. Luik RM, Wu MM, Buchanan J, Lewis RS. J Cell Biol. 2006;174:815.
8. Palty R, Isacoff EY. J Biol Chem. 2016;291:334.
9. Shim AH-R, Tirado-Lee L, Prakriya M. J Mol Biol. 2015;427:77.
10. Vaeth M, Yang J, Yamashita M, et al. Nature Commun. 2017. http://dx.doi.org/10.
1038/ncomms14714.
11. Davis FM, Janoshazi A, Janardhan KS, et al. Proc Natl Acad Sci USA. 2015;112:5827.
12. McAndrew D, Grice DM, Peters AA, et al. Mol Cancer Ther. 2011;10:448.
13. Spinelli AM, Trebak M. Am J Physiol-Cell Ph. 2016;310:C402.
14. Son A, Park S, Shin DM, Muallem S. Am J Physiol-Cell Ph. 2016;310:C414.
15. Vashisht A, Trebak M, Motiani RK. Am J Physiol-Cell Ph. 2015;309:C457.
16. Chen Y-F, Hsu K-F, Shen M-R. Biochim Biophys Acta. 2016;1863:1427.
17. Jardin I, Rosado JA. Biochim Biophys Acta. 2016;1863:1418.
18. Hoth M. Biochim Biophys Acta. 2016;1863:1408.
19. Xie J, Pan H, Yao J, Zhou Y, Han W. In J Cancer. 2016;138:2067.
20. Chavez KJ, Garimella SV, Lipkowitz S. Breast Dis. 2010;32:35.
21. Didiasova M, Zakrzewicz D, Magdolen V, et al. J Biol Chem. 2015;290:11983.
22. Yang S, Zhang JJ, Huang X-Y. Cancer Cell. 2009;15:124.
59. Enthaler S, Weidauer M, Irran E, Epping JD, Kretschmer R, Someya CI. J Organomet
Chem. 2013;745–746:262.
60. Zelenin KN, Alekseyev VV, Tygsheva AR. Tetrahedron. 1995;51:11251.
61. Fustero S, Roman R, Sanz-Cervera JF, et al. J Org Chem. 2008;73:3523.
62. Singh SP, Kapoor JK, Kumar D, Threadgill MD. J Fluorine Chem. 1997;83:73.
63. Bonacorso HG, Oliveira MR, Costa MB, et al. Heterocyclic Chem. 2005;42:631.
64. Irwin JJ, Sterling TT, Mysinger MM, Bolstad ES, Coleman RG. J Chem Inf Model.
2012;52:1757.
65. Azimi I, Beilby H, Davis FM, et al. Mol Oncol. 2016;10:166.
66. McAndrew D, Grice DM, Peters AA, Davis FM, Stewart T, Rice M, Smart CE, Brown
MA, Kenny PA, Roberts-Thomson SJ, Monteith GR. Mol Cancer Ther. 2011;10:448.
67. Monteith GR, Bird GS. Trends Pharmacol Sci. 2005;26:218.
68. Rice LV, Bax HJ, Russell LJ, et al. Eur J Pharmacol. 2013;704:49.
69. Vetter I, Touska F, Hess A, et al. EMBO J. 2012;31:3795.
23. Hogan PC, Lewis RS, Rao A. Annu Rev Immunol. 2010;28:491.
24. Putney JW. Mol Interv. 2010;10:209.
25. Sweeney ZK, Minatti A, Button DC, Patrick S. ChemMedChem. 2009;4:707.
26. Zhang H-Z, Xu X-I, Chen H-Y, et al. Acta Pharmacol Sin. 2015;36:1137.
27. Franzius D, Hoth M, Penner R. Pflugers Arch. 1994;428:433.
28. Kozak JA, Kerschbaum HH, Cahalan MD. J Gen Physiol. 2002;120:221.
29. Chung SC, McDonald TV, Gardner P. Br J Pharmacol. 1994;113:861.
30. Chen YF, Chiu WT, Chen YT. Proc Natl Acad Sci USA. 2011;108:15225.
31. Enfissi A, Prigent S, Colosetti P, et al. Cell Calcium. 2004;36:459.
70. Sousa SR, Vetter I, Ragnarsson L, Lewis RJ. PLoS One. 2003;8:e59293.
8