Potent antiproliferative activity of bradykinin B2 receptor selective agonist FR-190997 and analogue structures thereof: A paradox resolved?

Article abstract of DOI:10.1016/j.ejmech.2020.112948

Βradykinin stimulation of B2 receptor is known to activate the oncogenic ERK pathway and overexpression of bradykinin receptors B1 and B2 has been reported to occur in glioma, colorectal and cervical cancers. B1R and B2R antagonists have been shown to reverse tumor proliferation and invasion. Paradoxically, B1R and B2R agonism has also been reported to elicit antiproliferative benefits. In order to complement the data accumulated to date with the natural substrate bradykinin and peptidic B2R antagonists, we decided to examine for the first time the response elicited by B2R stimulation in breast cancer lines with a non-peptidic small molecule B2R agonist. We synthesized and assessed the highly selective and potent B2R partial agonist FR-190997 in MCF-7 and MDA-MBA-231 breast cancer lines and found it possessed significant antiproliferative activity (IC₅₀ 2.14 and 0.08 μΜ, respectively). The modular nature of FR-190997 allowed us to conduct a focused SAR study and discover compound 10 which exhibits subnanomolar antiproliferative activity (IC ₅₀ 0.06 nΜ) in the TNBC MDA-MBA-231 cell line. This performance surpasses, in most cases by several orders of magnitude, those of established anticancer agents and FDA-approved breast cancer drugs. In line with the established literature we suggest that this remarkable activity precipitates from a dual mode of action involving agonist-induced receptor internalization/degradation combined with sequestration of functional intracellular B2 receptors and inhibition of the associated endosomal signaling. The latter mode may be realized by appropriate ligands regardless of B2R agonist/antagonist designation which only relates to membrane residing GCPRs. Under this prism the controversy over the antiproliferative effects of B2 agonists and antagonists is potentially neutralized.

Full text of DOI:10.1016/j.ejmech.2020.112948

Journal Pre-proof
Potent antiproliferative activity of bradykinin B2 receptor selective agonist FR-190997
and analogue structures thereof: A Paradox Resolved?
Gerasimos Rassias, Sofia Leonardi, Dionisia Rigopoulou, Eleanna Vachlioti,
Konstantinos Afratis, Zoi Piperigkou, Christos Koutsakis, Nikos K. Karamanos,
Haralambos Gavras, Dionissios Papaioannou
PII:
S0223-5234(20)30920-X
DOI:
https://doi.org/10.1016/j.ejmech.2020.112948
EJMECH 112948
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 27 August 2020
Revised Date: 24 September 2020
Accepted Date: 14 October 2020
Please cite this article as: G. Rassias, S. Leonardi, D. Rigopoulou, E. Vachlioti, K. Afratis, Z. Piperigkou,
C. Koutsakis, N.K. Karamanos, H. Gavras, D. Papaioannou, Potent antiproliferative activity of bradykinin
B2 receptor selective agonist FR-190997 and analogue structures thereof: A Paradox Resolved?,
European Journal of Medicinal Chemistry, https://doi.org/10.1016/j.ejmech.2020.112948.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2020 Elsevier Masson SAS. All rights reserved.

Graphical abstract

1
Potent antiproliferative activity of bradykinin B2 receptor
selective agonist FR-190997 and analogue structures
thereof: A Paradox Resolved?
a*
a
a
Gerasimos Rassias , Sofia Leonardi , Dionisia Rigopoulou , Eleanna
a
a,b
c,d
Vachlioti , Konstantinos Afratis , Zoi Piperigkou , Christos
Koutsakis , Nikos K. Karamanos , Haralambos Gavras , Dionissios
c
c,d
e*
a,*
Papaioannou
a
Laboratory of Synthetic Organic Chemistry, Department of Chemistry, University of
Patras, 26504 Patras, Greece
b
Current address: Organic Chemistry Research Laboratory, University of Oxford, 12
Mansfield Rd, OX1 3TA, Oxford, UK
c
Biochemistry, Biochemical Analysis & Matrix Pathobiology Research Group,
Laboratory of Biochemistry, Department of Chemistry, University of Patras, 26504
Patras, Greece
d
Foundation for Research and Technology-Hellas (FORTH)/Institute of Chemical
Engineering Sciences (ICE-HT), Patras, Greece
e
Hypertension and Atherosclerosis Section, Department of Medicine, Boston
University School of Medicine, Boston, Massachusetts 02118, USA
_
___________________________
*
Corresponding author.
E-mail addresses: rassiasg@upatras.gr (G. Rassias), dapapaio@upatras.gr (D.
Papaioannou).
Highlights
•
•
•
•
Significant antiproliferative activity is demonstrated for the first time with a
B2R agonist (FR-190997)
SAR on FR-190997 led to novel analogue 10 demonstrating subnanomolar
inhibition in the MDA-MB-231 assay.
The antiproliferative performance of analogue 10 in MDA-MB-231 cells is
superior to that exhibited by FDA-approved breast cancer drugs
The
mechanism
probably
involves
agonism-induced
B2R
internalisation/degradation and inhibition of endosomal B2R signalling

2
1
. Introduction
Kinins exert their biological effects by activating two types of G protein-coupled
receptors (GPCRs), traditionally known as B1 (B1R) and B2 (B2R) which are
coupled to Gaq/11 and Gai/o G proteins. These receptors have been shown to
modulate diverse cellular functions leading to pro-inflammatory activity,
vasodilatation, vascular permeability, nociception, chemotaxis and sodium and
chloride excretion by the renal tubules [1-3].
In general, B1 and B2 activation triggers essentially the same signalling cascades
nevertheless the observed responses stemming from B2R are acute and of short
duration whereas those from B1R are more sustained and of lower intensity. The B2R
is constitutively expressed in a variety of cell types whereas the physiological
expression of B1R is low but up-regulated significantly in cases of tissue injury,
cytokine stimulation, inflammation and cancer.
Accumulating data with bradykinin (BK) and its active metabolite des-Arg9-BK
(DABK), selective activators of B2 and B1 receptors respectively, suggests that
activation of these receptors enhance proliferation and invasion of gliomas,
colorectal, breast, and cervical cancer cells and that these trends are reversed by B1/2
receptor antagonists [4-11].
In contrast, the protective effects of B1 agonism with des-Arg9-BK in mice with
tumors derived from TM5 melanoma cells has also been demostrated [12]. Equally
one cannot ignore the reduced metastatic potential in B16F10 murine melanoma
model following B1 agonism with DABK reported recently [13].
In the long and significant contribution of the Marceau group in this area,
compound B-9870 (CU201) was found to inhibit tumour cell growth both in vitro and
in vivo. In particular it activated c-Jun kinase and caspase-3 while antagonizing BK-
induced calcium signalling causing a debate as to whether this compound is a B2R
antagonist or biased agonist [14].
It is yet unclear whether these controversies originate from different cell lines,
cancer type and microenvironment or there is a subtle yet important factor in
determining the ultimate cell response upon B1/2R stimulation.
In an attempt to shed light into this, we decided to explore the effect of B2
agonism in the context of cancer using the highly selective and potent non-peptidic
partial agonist FR-190997 (Figure 1) developed by Fujisawa (legacy company of
Astellas) [15]. We reasoned that this approach would complement the established

3
data which almost invariably has been generated using either the natural peptidic
substrates of B1/2 receptor or synthetic peptidic and non-peptidic antagonists. To the
best of our knowledge, a B2 selective non-peptidic small molecule partial agonist has
not been utilized in antineoplasmatic studies.
2
2
. Results and Discussion
.1 Scaffold selection and SAR design
Chronic stimulation of B2R has been proposed as a promising approach for the
prevention and treatment of cardiovascular disease [1-3,16,17]. Nevertheless, BK
suffers from rapid breakdown in vivo, thus a research team from Fujisawa developed
a series of novel and stable non-peptidic B2R agonists based on a quinoline warhead.
In that work it was established that the quinoline substitution at position 4 is crucial
for exhibiting either selective antagonism (no substitution or amino-substitution) or
agonism (alkoxy substitution) at B2R [15, 18-20]. Optimised agonists of this class
include FR-190997 and 1A (IC 0.41 and 0.36 nm, respectively) while the other
5
0
warhead that imparts agonistic profile is an appropriately substituted benzimidazole
such as the one in full agonist 1B (IC 0.68 nm). Changes in the terminal amide part
5
0
of the molecule only fine tune solubility and the magnitude of the functional profile
determined by the warhead. Fujisawa's FR-173657 [21,22] (Figure 1) is a B2R
antagonist due to the absence of the agonist-imparting substituent at the quinoline
warhead despite sharing the same terminal amide as agonist 1B.
Figure 1. Structures of B2R ligands FR-190997, FR-173657, 1A and 1B.

4
The best characterized member of this class of non-peptide B2R agonists is FR-
90997 (Figure 1); a highly potent and subtype selective partial agonist of B2R (in
1
terms of inositol phosphates, prostagladin production and calcium mobilization in
comparison with bradykinin). FR-190997 was originally developed and progressed
through rigorous preclinical studies for cardiovascular indications although it was
subsequently repurposed to the treatment of glaucoma/ocular hypertension [23]. We
selected FR-190997 to interrogate the significance of B2 agonism in the context of
cancer primarily because of its high B2R potency and complete lack of affinity for
the B1 receptor or other targets for that matter, hence potential antiproliferative
activity would not be associated with off-target effects. Another reason for choosing
FR-190997 for the present study was the modular nature of its structure allowing for
a focused SAR on its respective fragments which in turn are amenable to independent
modification.
FR-190997 consists of four modules, namely a 2-methyl-4-(2-pyridylmethoxy)-
substituted quinoline, a 4-substituted cinnamic acid, a 2,4-dichloro-3-hydroxymethyl-
aniline and glycine. The latter two domains are connected through an N-methylated
amide bond which has been shown to adopt preferentially the cis-conformation, with
the planes of the amide bond and the phenyl ring almost perpendicular to each other
[
17,20]. This central part of the structure seems to play a crucial role in defining the
active conformation of the molecule. In addition, the structure of the acid component
in the terminal amide appears to fine-tune the magnitude of activity and solubility
properties in these series [15, 18-22]. Inspection of these structural components led us
to recognize two privileged scaffolds in cancer drug discovery embedded within FR-
1
1
90997, namely quinoline [24,25] and cinnamic acid [26,27], and reasoned that FR-
90997 could act perhaps as a quinoline-cinnamic acid conjugate.
In terms of cancer cell lines we decided to investigate the response elicited by FR-
90997 initially in the MCF-7 breast cancer cells as it has been reported that
1
bradykinin agonism at B2R induces proliferation in this and related breast cancer
assays by activation of the oncogenic pathway PKCδ/ε, PKB/Akt, and ERK1/2 [28-
3
0]. We also chose to examine the response elicited by the B2R agonist FR-190997
in the triple negative breast cancer (TNBC) cell line MDA-MB-231 because of recent
work [31] demonstrating benefits with the B2R antagonist FR-173657, a close
analogue of FR-190997 (Figure 1) also from the Fujisawa collection. Our in vitro
tests using FR-190997 in these two breast cancer lines, MCF-7 (low invasive

5
potential) and MDA-MB-231 (high invasive potential), presented substantial
antiproliferative activity with IC values of 2.14 µM and 0.08 µΜ respectively for
5
0
the two cell lines.
Figure 2. Structures of B2R selective agonist FR-190997 and its fragments 2-5.
We therefore decided to (a) identify which parts of the molecule were responsible
for the observed biological effect (see compounds 2-5, Figure 2) and (b) synthesize a
series of novel analogs of FR-190997 and key-fragments thereof (Figure 3) in order
to compare their potential antiproliferative activity to that of the prototype compound
FR-190997 with the view of simplifying the active structure. For the latter we
considered modifications in (i) the linker connecting the quinoline with the 2,6-
dichlorophenyl ring (compounds 9-11), (ii) the N-methylated amide bond
(compounds 8 and 9), which is crucial for the overall shape of the molecule, and (iii)
the acid residue in the terminal amide (6 and 7). The compounds thus conceived,
synthesized and evaluated as potential antiproliferative agents are depicted in Figure
3
. It should be noted that compound 12 was designed to assess the impact of the
hydrogen bond donating/accepting capacity of the bis-amide terminal domain in FR-
90997, whereas compound 7, an FR-190997 hybrid with all-trans-retinoic acid
atRA), was designed to mimic the extended conjugated system of analogue 1A.
1
(
Additionally, atRA is itself an anticancer drug for acute promyelotic leukemia (APL).
Furthermore, compounds 13 and 14 are analogues of the FR-190997-fragments 2 and
3
, respectively.

6
Figure 3. Structures of new analogues of FR-190997 (6-11) and fragments (12-14)
thereof synthesized in the context of the present work.

7
2
.2 Chemistry
Τhe synthesis of fragments 2-5 en route to the synthesis of the target molecule FR-
90997 have been described by Fujisawa's researchers in a series of articles [15, 18-
2] and summarized in reference 17. For the needs of the present study, we
1
2
synthesized these compounds based on the Fujisawa routes and implemented some
minor modifications (see Supplementary Material (SM), Figures S1-S4 and related
experimental procedures). Incorporation of the quinoline substructure in the ether-
type analogs 6-8 and 12 was effected using 8-hydroxyquinoline derivative 2, whereas
for the amino-type analogs 9-11 and, alternatively, for fragment analog 12, 8-
aminoquinoline derivative 13 was used as the key precursor. This compound was
synthesized from the commercially available 4-chloro-2-methylquinoline (15) and
(pyridin-2-yl)methanol using the three-step procedure outlined in Scheme 1.
Selective nitration of 15 produced the 8-nitro derivative 16 [32] in 49% yield.
Nucleophilic aromatic substitution by (pyridin-2-yl)methanol in the presence of
t
BuOK was accomplished, as expected, under much milder conditions than the
corresponding reaction with the hydroxyl group at position 8 of the quinoline group
(see Scheme S1), and gave intermediate 17 in 43% yield. Finally, the nitro group was
reduced with hydrazine monohydrate, in the presence of a catalytic quantity of
FeCl 6H 0, to provide fragment 13 in 72% yield. On the other hand, diazotization of
3
˙
2
1
3 followed by KI treatment produced the iodide 18 in 74% yield. Initially, this
compound was considered as a key-intermediate for the assembly of the FR-190997
molecule via the Ullmann reaction. Attempted condensation of iodide 18 with alcohol
5
in the presence of Cs CO , CuI and 1,10-phenanthroline failed however to produce
2 3
the target molecule FR-190997. Interestingly, Ullmann reaction of iodide 18 with the
simpler 3-phenoxybenzyl alcohol under identical reaction conditions, produced
fragment 12 in low yield (18%) suggesting that the reason for the failure of the
Ullmann coupling of 5 is not entirely steric. In both reactions the alcohol partner was
consumed and reduction of iodide 18 was also observed indicating that the stability of
the substrates may be responsible for the poor performance of the Ullmann reactions.
Alternatively, fragment 12 could be obtained in a much better yield (60%) through
the Mitsunobu reaction of fragment 2 with m-phenoxybenzyl alcohol.

8
Scheme 1. Synthesis of fragments 12 and 13. Reagents and conditions: (i) KNO3,
9
6% H SO , -10 °C for 50 min then 25 °C,12 h, 49%; (ii) (Pyridin-2-yl)methanol,
2 4
t
BuOK, DMI, 0 °C then -10 °C for 50 min and finally 25 °C, 12 h, 43%; (iii)
N H ·H O, FeCl ·6H O (cat.), activated C, 80% aq. MeOH, 75 °C, 3 h, 72%; (iv)
2
4
2
3
2
NaNO 37% HCl,H O, 5 °C for 50 min, then 3M aq. KI, 5 °C for 10 min then 25
2
,
2
°
C, 1 h, 74%; (v) CuI, 1,10-phenathroline, Cs CO , m-phenoxybenzyl alcohol, 1,3-
2 3
dimethyl-2-imidazolidinone (DMI), 110 °C, 12 h, 18%; (vi) m-phenoxybenzyl
alcohol, TPP, DIAD, dry THF, 0 °C for 30 min then 25 °C, 48 h, 60%; (vii) (Pyridin-
2
-yl)methanol, NaH, DMI, 0 °C for 1 h then 25 °C and finally at 100 ºC, 12 h, 83%.
We envisaged that replacing the cinnamic acid moiety of FR-190997 with
hydroxycinnamic acid (HCA) ferulic acid (FerA) or the retinoid acid atRA would
require the availability of suitable activated esters, e.g. the succinimidyl esters.
Isolable succinimidyl esters of α,β-unsaturated carboxylic acids are known to
selectively acylate primary amino groups in the presence of secondary ones [33,34].
We therefore selected the succinimidyl 'active' ester (21) of O-(tert-
butyl)diphenylsilylferulic acid (20) and the known isolable succinimidyl 'active' ester
(22) of atRA [34] for generating analogues 6 and 11, respectively. The synthesis of
the former was realized in 42% overall yield by fully protecting FerA with the (tert-
butyl)diphenylsilyl (TBDPS) group to give intermediate 19, followed by selective

9
carboxyl group deprotection providing acid 20, and finally activation of the latter
with N-hydroxysuccinimide (HOSu) and N,N'-dicyclohexylcarbodiimide (DCC)
(
Scheme 2). On the other hand, the also isolable crystalline succinimidyl 'active' ester
2 of atRA was prepared as described in the literature [34], using the same system
HOSu/DCC) for activation.
2
(
Scheme 2. Synthesis of isolable 'active' esters 21 and 22. Reagents and conditions: (i)
TBDPSCl, imidazole, dry DMF, 25 °C, 3 h, 81%; (ii) 10% aq. K CO , MeOH, THF,
2
3
o
0
C then 25 °C, 1 h, 61%; (iii) HOSu, DCC, dry THF, 0 °C for 30 min then 25 °C 12
h, 85% (21), 73% (22).
The synthesis of FR-190997 analogs 8 and 9 required the prior synthesis of
compound 14, which is an analog of fragment 3, in which the cis-amide bond
functionality has been replaced with a triazole ring. Compound 14 was synthesized as
depicted in Scheme 3 and involved the reduction of intermediate S11 (see Scheme S3
.
of SM) with hydrazine monohydrate and a catalytic quantity of FeCl 6H O to
3
2
produce aniline derivative 23 in 81% yield, followed by diazotization and reaction
with sodium azide to provide aryl azide 24 in 79% yield. The latter was subjected to a
Huisgen 1,3-dipolar cycloaddition reaction with the N-phthalylated propargylamine
2
5, to give the desired triazole derivative 14 in 70% yield.

1
0
Scheme 3. Synthesis of fragment 14. Reagents and conditions:(i) N H ·H O,
2
4
2
FeCl ·6H O (cat.), activated C, 80% aq. MeOH, 75 °C, 4 h, 81%; (ii) NaNO ,
3
2
2
CF CO H (TFA), 0 °C, 45 min, then NaN , 0 °C, 5 min then 25 °C for 30 min, 79%;
3
2
3
i
i
(
iii) Phthalimide, Ph P (TPP), PrO CN=NCO Pr (DIAD), dry THF, 0 °C then 25 °C,
3 2 2
2
h, 60%; (iv) CuSO ·5H O, sodium ascorbate, H O, DCM, 25 °C, 3 h, 70%.
4 2 2
Having established the availability of all required intermediates, the synthesis of
FR-190997 analogs 6-11 was realized as follows. The key-intermediate S15 was
subjected to hydrazinolysis to provide amine S16 (Scheme S4, SM), which was then
treated with tetrabutylammonium fluoride (TBAF) giving the aminoalcohol 26
(Scheme 4) in 48% yield for the two steps. Selective N-acylation of intermediate 26
with ‘active' ester 21 gave alcohol 27 in 85% yield. Alcohol 27 was first converted to
the corresponding bromide 28 in 72% yield using Appel reaction, and this was further
used to alkylate fragment 2, providing intermediate 29 in 44% yield. From compound
2
9, the projected FR-190997 analog 6 was finally obtained in 64% yield after
treatment with TBAF (Scheme 4). On the other hand, selective O-deprotection of
key-intermediate S15 gave fragment 3 in 55% yield (see also Scheme S3 of SM)
which was condensed with fragment 2, under Mitsunobu reaction conditions, to give
the new intermediate 30 in 70% yield. The latter, upon hydrazinolysis and coupling
of the thus obtained, in 40% yield, amine 31 with atRA, in the presence of O-
(benzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate (HBTU),
provided the FR-190997 analog 7 in 50% yield.

1
1
Scheme 4. Synthesis of FR-190997 analogs 6 and 7. Reagents and conditions:(i)
N H ·H O, EtOH, reflux, 1-2 h, 88% (S16), 40% (31); (ii) TBAF·3H O, THF, 25 °C,
2
4
2
2
3
h, 57% (3), 56% (26),64% (6); (iii) dry Et N, dry DMF, 25 °C, 1 h, 85%; (iv) CBr ,
3
4
TPP, dry THF, 0 °C then 25 °C, 3 h, 72%; (v) K CO , dry DMF, 25 °C, 3 h, 44%;
2
3
o
o
(
vi) TPP, DIAD, dry DMF, 25 C, 1 h then 65-70 C, 12 h, 70%; (vii) HBTU, dry
Et N, dry DMF, 25 °C, 1 h, 50%.
3
For the synthesis of analog 8, fragment 2 was first condensed with fragment 14
under Mitsunobu reaction conditions. However, the reaction was sluggish and gave

1
2
the product 33 in low yield (15%). Alternatively, fragment 14 was first converted to
the corresponding bromide 32 in 47% yield, under Appel conditions, and this was
then used to alkylate fragment 2 in the presence of K CO (Scheme 5). Product 33
2
3
was then received in 84% yield. Hydrozinolysis of intermediate 33 gave the free
amine 34 in 45% yield, which was then coupled with fragment 4, in the presence of
the coupling agent HBTU and Et N, to give analog 8 in 76% yield.
3
Scheme 5. Synthesis of FR-190997 analog 8. Reagents and conditions: (i) CBr ,
4
o
o
TPP, dry THF, 0 C then 25 °C, 3 h, 47%; (ii) TPP, DIAD, dry DMF, 25 C to 60-70
C,48 h, 15%; (iii) K CO , dry DMF, 25 °C, 3 h, 84%; (iv) N H ·H O, DMF, MeOH,
o
2
3
2
4
2
7
0–80 °C, 1 h, 45%; (v) HBTU, dry Et N, dry DMF, 25 °C, 1 h, 76%.
3
Fragment 13 was envisaged as the key-intermediate for the synthesis of analogs 9-
1 and was initially converted to the N-nosylated derivative 35 in 91% yield (Scheme
). The nosylamide was subsequently coupled, under Mitsunobu reaction conditions
1
6
[
35], with alcohol 3 to give intermediate 36. From this intermediate, the Ns group
was removed with PhSH in the presence of Na CO to give intermediate 37 in 52%
2
3
yield for the two steps. Hydrazinolysis of compound 37 produced the diamine 38 in
7% yield, which was finally coupled with 'active' ester S7 (Scheme S2 of SM) to
give FR-190997 analog 10 in 68% yield. Similarly, diamine 38 was coupled with
active' ester 22 giving the FR-190997 analog 11 in 43% yield. In an analogous
9
'
manner, compound 35 was alkylated with alcohol 14, under Mitsunobu reaction
conditions, to produce intermediate 39. The nosyl group was then removed from
compound 39 to provide the intermediate 40 in 43% yield for the two steps. Finally,
removal of the Phth protecting group with hydrazinolysis initially produced the N-
formylated derivative 41 which, upon deformylation with hydrazine in refluxing

1
3
EtOH and coupling of the thus obtained diamine 42 with 'active' ester S7, produced
the desired FR-190997 analog 9 in 37% yield (combined for the last three reactions).
Compound 41 was formed as an intermediate in the hydrazinolysis reaction because
DMF had to be used as cosolvent in the first step due to the insolubility of compound
4
0 in EtOH.
Scheme 6. Synthesis of FR-190997 analogs 9-11. Reagents and conditions: (i) NsCl,
dry Et N, dry DCM, 0 °C for 50 min then 25 °C for 48 h, 91%; (ii) TPP, DIAD, THF
3
o
o
dry, 0 C then 40-50°C, 4 h (36) or 0 C to 40-50 °C for 3.5 h, then 25 °C, 12 h (39);
iii) PhSH, Na CO , dry DMF, 40–50 °C, 2 h, 52% (37) or 25 °C, 2 h then 40–50 °C
(
2
3
for 1 h and then 25 °C, 48 h, 43% (40); (iv) N H ·H O, EtOH, reflux, 1h, 97% (38) or
2
4
2
3
4
6 h 72% (42); (v) dry Et N, dry DMF, 25 °C, 2 h, 51%(9) or 1.5 h,68% (10) or 12 h,
3
o
2% (11);(vi) N H ·H O, DMF/EtOH, 80 C, 12h.
2
4
2

1
4
2
.3 Structure-Antiproliferative activity of FR-190997 and analogue structures
All synthesized compounds were tested for their antiproliferative potential using
two breast cancer cell lines: the ERα-positive, epithelial MCF-7 cells with low
metastatic potential and the ERβ-positive, mesenchymal MDA-MB-231 cells with
high metastatic potential. For comparison purposes, the commercially available
compounds FerA and atRA were also included in this assessment. The IC values of
50
the tested compounds are summarized in Table 1.
Τable 1. IC (µM) values of the synthesized compounds in MCF-7 and MDA-
5
0
MB-231 breast cancer cells.
LIGAND
MCF-7
MDA-MB-231
0.08 ± 0.089
450 ± 12.169
0.78 ± 0.039
86 ± 1.049
FR-190997
2.14 ± 0.034
0.16 ± 0.06
53 ± 4.026
2
3
4
5
6
7
8
9
2.12 ± 0.017
0.93 ± 0.062
0.33 ± 0.05
65 ± 2.031
650 ± 12.427
5.9 ± 0.028
193 ± 9.048
200 ± 10.012
0.43 ± 0.046
91 ± 7.061
0.73 ± 0.043
1.21 ± 0.071
64 ± 5.031
-5
1
1
1
1
1
0
1
2
3
4
5.94*10 ± 0.00005
-3
5*10 ± 0.0014
6.97 ± 0.103
22 ± 0.57
407 ± 12.073
0.86 ± 0.076
0.07 ± 1.311
0.36 ± 0.017
6.14 ± 0.38
520 ± 10.227
15 ± 1.156
FerA
atRA
16 ± 1.106
In contrast to the hypothesis made earlier in the literature [4] that B2R agonists
may only promote tumor proliferation, we found that B2R agonist FR-190997
inhibited cell proliferation of the MCF-7 cell line significantly with an IC of 2.14
5
0
µΜ whereas the antiproliferative effect in the MDA-MB-231 cell line was even more
pronounced with an IC of 0.08 µM. The SAR results with FR-190997 analogue
5
0
structures and sub-domains are discussed separately for each cell line.
MCF-7 cell line
Careful inspection of Table 1 reveals that the efficacy of FR-190997 in the MCF-7
cell line stems from the quinoline (compound 2) and cinnamic acid anilide
(compound 5) domains which independently produce better antiproliferative activity

1
5
(
IC 0.16 and 0.93 µM respectively). Connecting these two privileged structures into
50
a single molecular entity (FR-190997) does not however produce an additive effect.
Interestingly, examination of the individual components of amide 5 reveals that
the p-substituted cinnamic acid 4 alone exhibits essentially the same activity as FR-
1
90997 whereas the N-Phth-protected dichloroanilino glycinamide core 3 possesses
much reduced activity (IC 53 µM). This suggests that the acid component in the
50
terminal amide of FR-190997 imparts significant activity to the glycinamide core
hence we tested FR-190997 analogues with other carboxylic acids of established
antitumor potential such as ferulic acid (FerA) and all-trans-retinoic acid (atRA).
Indeed, FerA derivative 6 exhibited a 7-fold increase antiproliferative activity with
respect to FR-190997 in the MCF-7 assay whereas atRA derivative 7 was less active
by 30-fold. This implies that extended conjugation in the terminal acid does not
necessarily impart significant antiproliferative activity in the combined structure and
that probably cinnamate derivatives possessing a hydrogen bond donor/acceptor at
the 4-position (CONH for FR-190997 and OH for 6) is a much more important
feature in this context.
Next, we interrogated modification at the N-methylanilide region of FR-190997,
by replacing the cis-N-methylglycinamide moiety with a triazole ring (compound 8)
in an attempt to restrict the conformational freedom in this part of the molecule. The
incorporation of the triazole ring in the structure of FR-190997 had a detrimental
effect on antiproliferative activity (IC 91 µM) although when this modification was
5
0
applied to the dichloroaryl core alone (compound 14), a direct analogue of 3,
produced the most active derivative of the present series with nanomolar activity in
the MCF-7 cell line (IC 0.07 µM).
5
0
Further modification of 8 by switching the ether to a secondary amine linker
compound 9), joining the quinoline core with the 2,6-dichloro-3-triazolobenzyl
(
domain of 8, imparted nanomolar antiproliferative activity in the FR-190997 daughter
triazole scaffold. Applying this modification directly on FR-190997 produced
analogue 10 which showed 2-fold improved activity over FR-190997 yet slightly
reduced activity with respect to its triazole counterpart 9.
Consistent with the trend observed in the O-linked compounds, replacement of the
substituted cinnamic acid residue in 10 with atRA (compound 11), led to 50-fold
reduction of the antiproliferative potency in the MCF-7 assay thus corroborating
further the negative impact of the atRA-derived amides.

1
6
As mentioned above, the 8-hydroxy quinoline fragment (compound 2) possesses
significant activity against the MCF-7 cell line (IC 0.16 µM) which is retained to a
50
large degree in its anilino-analogue 13 (IC of 0.86 µM) whereas the corresponding
5
0
quinoline ether derivative 12 lacking the hydrogen bond donor capacity of 2 and 13,
was essentially inactive in the MCF-7 cell line.
Overall, in the context of the MCF-7 cell line, these results support that the choice
of the terminal carboxylic acid residue is of prime importance for antiproliferative
activity with cinnamate derivatives possessing a hydrogen bond donor/acceptor at the
4
-position being particularly active (compounds FR-190997, 6, 8, 9, 10 versus 7 and
1). Second in importance is the presence of a hydrogen bond donor in the linker
1
region as derivatives with an NH linker are more active than their O-linked
counterparts (compounds 9 and 10 versus 8 and FR-190997 respectively). A subtle
yet noticeable feature that further increases the antiproliferative activity of the NH-
linked FR-190997 analogue structures, involves the regidification around the amine
component in the terminal amide domain by the use of 4-aminomethyltriazole in
place of the conformationaly more labile glycine skeleton (compound 9 versus 10).
Based on the above, it is projected that the FerA analogue structures of 9 and 10
should be even more potent in the MCF-7 cell line.
MDA-MB-231 cell line
In contrast to the MCF-7 assay, the high cellular efficacy of FR-190997 in the
MDA-MB-231 cell line appears to emanate largely from the 3-glycinamido-2,6-
dichlorobenzyl core (compound 3) rather than the quinoline domain (compound 2) or
terminal acid residue (compound 4) (IC 0.78 versus 450 and 86 µM respectively in
5
0
this assay, Table 1). Connecting the cinnamic acid in the form of terminal amide on
core fragment 3 produces the essentially inactive derivative 5 (IC50 650 µM).
Interestingly, attaching the inactive quinoline domain 2 through an ether linker to the
also inactive 5, forms FR-190997 which possess nanomolar activity in the MDA-
MB-231 assay. This unpredicted positive synergy of domains 2, and 5 (in turn formed
by 3 and 4) is in stark contrast with what was observed in the MCF-7 assay where the
same fragments exhibited significant antiproliferative activity yet their union into FR-
1
90997 had a negative impact. Switching the terminal carboxylic acid residue from
the 4-substituted cinnamate in FR-190997 to FerA and atRA in analogue structures 6
and 7 led to inferior potency in the MDA-MB-231 cell line, particularly so in the
latter case.

1
7
Restricting the conformational freedom of the cis-N-methylglycinamide moiety in
FR-190997 by replacing this with a triazole ring (compound 8) also causes a sharp
drop in the antiproliferative activity in the MDA-MB-231assay. The negative impact
of the triazole modification is also consistent in the respective fragments 3 and 14
(
IC 0.78 versus 520 µM respectively).
50
Finally, applying the O to NH switch in the linker region of FR-190997 produced
analogue 10 which proved the most active compound in the present series with
-5
subnanomolar antiproliferative activity in the MDA-MB-231 cell line (IC 5.94*10
5
0
µM). In compound 9, the O to NH replacement at the linker of the inactive FR-
90997 triazole analogue 8 also rescues the antiproliferative activity precipitating
nanomolar potency (IC 0.43 µM). The profound impact of this modification is
1
5
0
clearly seen in analogue 11 where one of the consistently inactive atRA analogues is
-3
rendered a potent antiproliferative agent with low nanomolar potency (IC 5*10
50
µM). The evidently key 8-aminoquinoline fragment 13, does possess antiproliferative
capacity (IC 22 µM) but this is several orders of magnitude lower than those
5
0
achieved by structures encompassing all three domains such as FR-190997 and
analogues 9, 10 and 11 which further corroborates the beneficial synergy among the
quinoline, the 2,6-dichloro-3-glycinamido-benzyl and the terminal acid domains.
It is thus established for the MDA-MB-231 cell line that the NH linker in the FR-
1
90997 analogue structures is of paramount importance for antiproliferative activity
followed by the conformational flexibility of the 3-glycinamide side chain in the 2,6-
dichlorobenzyl core while several carboxylic acids with extended conjugated systems
may be tolerated in the terminal amide domain.
Overall, it is worth noticing that although the SARs for the MCF-7 and MDA-
MB-231 cell lines follow opposite trends with respect to the triazole and terminal
acid modifications, the O to N switch in the linker region is beneficial across both
assays and that respective optimum potencies converge to analogue 10.
It must be also emphasised that whatever the antiproliferative activity of each of
fragments 2, 3, 4, 5, 12, 13 and 14, it is not anticipated to stem from B2R agonism
alone, if at all, since molecules of this size tend to be promiscuous ligands and the
origin of their activity could be attributed to association with more than one targets.
In contrast, compounds 6, 7, 8, 9, 10 and 11 can be safely claimed as B2R agonists or
partial agonists not simply due to their extensive structural resemblance to FR-
1
90997 but mostly because of Fujisawa's extensive SAR and optimisation studies

1
8
which have shown that the agonist/antagonist profile is determined solely by the
substitution at the 4-position of the quinoline domain [15, 18-22] (see also Figure 1).
The same work established that the dichloro benzyl core is important for overall
activity whereas the glycinamide and terminal acid portions modulate only the
magnitude of the agonistic or antagonistic effect and does not change the functional
aspect of the molecule.
2
.4 Proposed mechanism of action based on the established knowledge on B2R
As mentioned above, for the two cancer lines examined, bradykinin activation of
B2R is reported to promote cancer proliferation in the MCF-7 assay [28] and B2R
antagonist FR-173657 showed antiproliferative effects in the MDA-MB-231 assay
[
31] therefore our antiproliferative results with a B2R agonist may appear to create a
paradox. This apparent controversy may be resolved if certain important points are
taken into consideration.
First, all work supporting the oncogenic fate of B2 agonism is based on
stimulation with bradykinin [28-30, 36-40]. B2R stimulation with other types of
agonists does not share the same characteristics with bradykinin simulation hence the
oncogenicity of the latter may not apply to all types of agonists. Synthetic peptidic
and non-peptidic B2R agonists do not actually mirror the functional
aspects/selectivity of the natural hormone bradykinin post B2 stimulation neither in
terms of type nor magnitude. For example, the highly potent and full agonist 1B
(Figure 1) stimulates phosphorylation of ERK (the oncogenic pathway) quicker and
with much reduced efficacy in comparison to bradykinin (1 µΜ vs 10 nM,
respectively) [41,42].
Second, bradykinin is rapidly degraded under physiological conditions whereas
FR-190997 belongs to a class of stable and inactivation-resistant B2R ligands [41].
Third, the fate of B2 receptor itself varies after ligand binding/stimulation and this
fate is dependent on ligand type [43]. Agonist binding to B2R causes internalization
of the receptor via association with β-arrestin . This is a reversible process that with
1
/2
the short lived bradykinin the receptor is quickly released and reinstated at the cell
membrane leading to a vicious circle of potent bursts of ERK activation. Conversely,
as elegantly shown by Marceau, when degradation resistant (long lived) synthetic
peptidic or non-peptidic agonists such as 1B (and FR-190997) are bound to B2R, the
internalization process is predominately followed by receptor degradation [44].

1
9
Fujisawa in collaboration with the Gobeil group have shown this is also the case for
FR-190997 (from the same Fujisawa series as 1B) as it induced desensitization of
B2R [45] (by default driven by receptor internalization).
It is precisely this phenomenon that we believe contributes to the observed
antiproliferative properties of FR-190997. Based on this well-established process for
GPCRs, prolonged engagement of B2 receptor with a high affinity (partial) agonist
would lead to receptor internalization/downregulation thus attenuation of the
membrane-bound B2R oncogenic signalling initiated by over-expressed extracellular
bradykinin.
This agonist-induced inhibitory effect mirrors the mechanism of action of
fingolimod. Phoshorylated fingolimod (fingolimod is actually a prodrug) binds to
S1P1 on the T cell surface and causes receptor internalization and degradation
thereby rendering the cell unresponsive to the egress signal of membrane S1P thus
allowing the lymph node signal to be retained [46-48]. Like fingolimode, FR-190997,
related B2R agonists and derivatives thereof may also be designated as functional
antagonists by virtue of their biased agonism towards β-arrestin1/2 which perturbs B2
receptor homeostasis, leading to downregulation of B2R [49-51].
Furthermore, internalized B2Rs escaping degradation or recycling have been
shown to be subject of further endosomal trafficking leading to nuclear translocation.
Takano has shown that the B2 receptor is transported in the nucleus by forming
heterodimers with lamin C where it is suggested to function as a transcriptional
regulator of specific genes [52,53].
More importantly, the group of Gobeil have found functional B2 receptors in the
nuclei of rat hepatocytes corroborating earlier reports of B2Rs residing in the nucleus.
Bradykinin stimulation of isolated nuclei induced concentration-dependent
mobilization of nucleoplasmic calcium activation/phosphorylation of Akt, acetylation
of histone H3 and pro-inflammatory iNOS gene induction [54]. This intracrine B2R-
mediated signalling was blocked by direct exposure of the nuclei to the peptidic B2R
antagonist HOE-140 but not by the B1R antagonist R715.
The Gobeil team followed this with a landmark article describing for the first time
nuclearly expressed B2R in the human triple-negative breast cancer (TNBC) cell line
MDA-MB-231 (cell line of our study) including human TNBC specimens. Treatment
with cell-permeable peptidic or small molecule B2R antagonists that could reach
endosomal domains elicited far superior anticancer effects to those observed with

2
0
non-permeable ones that could only act at membrane-bound B2R [31]. The
antiproliferating and apoptosis-promoting effects on MDA-MB-231 cells by these
agents was not manifested in B2R shRNA-knockdown or non-B2R expressing (COS-
1
) cells. Since GPCR antagonists do not cause GPCR internalization, Gobeil's work
has established that the most efficient antiproliferative mechanism for B2R
antagonists is to permeate through the cell and inhibit the intracellular/nuclear B2R
signalling rather than the cascade originating at membrane-bound B2R.
The classical definitions of GPCR agonists/antagonists are associated to specific
signal transduction originating from extracellular ligand interactions with membrane-
bound GPCRs. It is therefore apparent that the usual agonist/antagonist designations
are not valid for a ligand which permeates the cell, binds to an intracellular GPCR
and inhibits its endosomal signalling (different from the membrane-originating
cascade).
The apparent paradox of having both B2R agonists and antagonists exhibiting
antiproliferative efficacy is potentially resolved by this assertion because it
disentangles membrane-bound from intracellular GPCR signalling as these do not
necessarily involve the same receptor organisation, ligand interaction, and
effectors/cascades [55-58]. It is fairer to say that permeable ligands capable of
binding to intracellular GPCR and disrupting its endosomal signalling are best
viewed as "interceptors" or "sequestrators" and this property is independent of their
agonistic/antagonistic capacity at the same GCPR residing on the membrane.
As evidenced by Gobeil's work, cell permeability and high GPCR affinity such as
those associated with the B2R antagonist used in that work, FR-173657 (IC 1.4 nm
50
[
21,22], ALogP 5.57, CXLogD (pH 7.4) 4.0 [59]) apparently suffice to precipitate
inhibition of intracellular B2R signalling by entering the cell and sequestering
intracellular B2Rs. The established FR-173657 type of action may be also manifested
by the high affinity B2R ligand FR-190997 (IC 0.41 nm) which is orally active in
5
0
several animal species [22,45] attribute for which cell permeability is a prerequisite
and further supported by its logP value(s) (ALogP 5.95, CXLogD (pH 7.4) 4.07
[
60]). Based on the B2R affinity and permeability profiles of FR-173657 and FR-
90997 one would expect the two Fujisawa molecules to exert similar potencies via
the sequestration mechanism. Nevertheless, in the triple-negative breast cancer
TNBC) cell line MDA-MB-231, agonist FR-190997 exhibits far superior
antiproliferative activity than FR-173657 (IC50 0.08 versus 20 µm respectively)
1
(

2
1
implying that agonist induced B2R internalisation/degradation mechanism (only
available to agonists) is much more efficient than the sequestration mechanism
(available to both agonists and antagonists of appropriate profile) in suppressing
tumorous cell proliferation, at least in this assay.
To put things into perspective a comparison of the antiproliferative activity of
optimised B2R ligands, approved breast cancer drugs and established anticancer
agents against the (TNBC) cell line MDA-MB-231 is presented in Table 2.
Table 2: Antiproliferative activity of B2R antagonists, agonists and
established chemotherapeutic agents against MDA-MB-231 cells
Ligand
FR-173657
NG68
FR-190997
Compound 10
Compound 11
Mode of Action
B2R antagonist
B2R antagonist
B2R agonist
B2R agonist*
B2R agonist*
IC / µM
Ref.
5
0
20
45
3
1
0.08 -
5.94*10
0.005
5
This
work
Approved breast cancer drugs
PI3Kα
Alpesilib
62.9
Cyclophosphamide Alkylating agent
195.5
Docetaxel
Epirubicin
Microtubule stabiliser
DNA intercalator
Antimetabolite
0.01-9.6
0.15
66.6
196.3
2.1
10.6
62.5
0.9
14.6
27.3
69.7
0.024
5-Fluorouracil
Fulvestrant
Gemcitabine
Lapatinib
ESR
Pyrimidine antimetabolite
EGFR, ERBB2
PARP1/2
Microtubule stabiliser
CDK4/6
CDK4/6
6
1
Olaparib
Paclitaxel
Paldociclib
Ribociclib
Tamoxifen
Vinblastine
ESR1
Microtubule destabiliser
Other established anticancer agents
Proteasome inhibitor
PI3K, mTOR
DNA crosslinker
Broad spectrum kinase inhibitor
DNA intercalator
Bortezomib
Rapamycin
Cisplatin
Staurosporin
Doxorubicin
0.003
0.9
36.2
0.01
0.1-1.3
61
*Based on Fujisawa's Structure-Activity/Functional-Relationship work
17-22
The significant antiproliferative activity of FR-190997 arguably stems from both
mechanisms (B2 receptor internalisation/degradation and sequestration of
intracellular/nuclear B2 receptors) thus rendered potentially a multimodal functional
inhibitor of B2R-mediated signalling. Cognizant of the fact that we have not
confirmed B2R agonism or off-target effects for the best analogue structure 10 we

2
2
posit that its exceptional (subnanomolar) antiproliferative performance is due to the
same multimodal capacity as of FR-190997 since these are structurally identical bar
one atom and is consistent with the agonist profile established in Fujisawa's extensive
Structure-Activity/Functional-Relationship work [15, 18-22]. Having gained insight
into the structural elements of FR-190997 responsible for imparting high
antiproliferative potency, our ongoing efforts are focused on preparing additional
analogues that may achieve broader spectrum antitumorigenic properties. In addition,
we intend to gather further evidence to support that the established agonism-induced
B2 receptor internalization and degradation applies also to partial agonist FR-190997
and that the latter is capable of sequestering intracellular/nuclear B2 receptors.
Unraveling the significance of subcellular GPCR trafficking and endosomal
signalling is one of the most exciting topics in current biology research as it holds
promise for novel therapeutic interventions [55-58].
3
. Conclusions
We have demonstrated for the first time that the non-peptidic, cell-permeable,
B2R partial agonist FR-190997 and related analogue structures exhibit exceptional
antiproliferating activities particularly in the TNBC cell line MDA-MB-231 where
our analogue 10 with subnanomolar activity, surpasses popular anticancer agents and
approved breast cancer drugs. This work represents the latest piece in the B2R-cancer
puzzle which we believe reconciles existing controversies by considering how the
action of B2R agonists/antagonists in the classical sense, transcends into the context
of endosomal B2R signalling.

2
3
4
4
4
. Materials and methods
.1 Chemistry
.1.1 General
Melting points were determined with an Electrothermal apparatus and are
1
13
uncorrected. H-NMR and C-NMR spectra were obtained at 600 and 150 MHz,
respectively, on Bruker AvanceIII HD spectrometer. Chemical shifts (δ) are reported
for CDCl solutions in parts per million (ppm) downfield from tetramethylsilane
3
(
TMS), used as internal standard, unless otherwise stated. Electron-spray ionization
ESI) mass spectra were recorded at 30 eV, on a Waters Micromass ZQ spectrometer
(
using MeOH as solvent. Elemental analyses were determined on a Carlo Erba EA
1
108 CHNS elemental analyzer.
Flash column chromatography (FCC) was performed on Acros Organics silica gel
.035-0.070 mm, 60 Å and TLC on Merck silica gel 60 F254 films (0.2 mm) precoated
0
on aluminium foil. Spots were visualized with UV light at 254 nm. The solvents or
solvent systems used for the development of TLC or FCC are the following: (A)
CHCl /MeOH (96:4), (B) CHCl /MeOH (95:5), (C) CHCl /MeOH (9:1), (D)
3
3
3
CHCl /MeOH(8:2), (E) CH Cl /MeOH (9:1), (F) PhMe, (G) EtOAc, (I) PhMe/EtOAc
3
2
2
(9:1), (J) PhMe/EtOAc (8:2), (K) PhMe/EtOAc (6:4), (L) PhMe/EtOAc (1:1), (M)
PhMe/EtOAc (2:8), (N) PhMe/EtOAc (3:7), (O) Hexane/EtOAc (9:1), (P)
Hexane/EtOAc (1:1), (R)Hexane/EtOAc (8:2), (S) PhMe/EtOAc (7:3), (T)
CHCl /MeOH (98:2), (U) CHCl /MeOH (97:3), (V) CHCl /MeOH (92:8), (W)
3
3
3
CHCl /MeOH (93:7).
3
All solvents were dried and/or purified according to standard procedures prior to
use. Solvents were routinely removed at ca. 40 ºC under reduced pressure on Büchi
Rotavapor RE 111 apparatus. Air-sensitive reagents were handled under inert
atmosphere (Ar). All reagents employed in the present work were purchased from
either Sigma-Aldrich or Alfa Aesar or Merck or Acros Organics or Fluorochem and
were used without further purification. For the needs of the present work, 'active'
ester 22 was prepared according to a published procedure [15].
4
.1.2 4-Chloro-2-methyl-8-nitroquinoline (16)
To a cooled to -10 °C solution of 4-chloro-2-methylquinoline (2.5 g, 14.1 mmol) in
6% H SO (8.9 mL) was added KNO (2.2 g, 21.8 mmol) portion-wise over 50 min
9
2
4
3

2
4
and the mixture was stirred at room temperature (rt) overnight. The reaction mixture
was poured carefully on ice (40 g), then adjusted to pH 7-8 with a saturated aq.
solution of NaOH and finally extracted thrice with EtOAc. The organic phases were
combined and washed twice with brine, dried over Na SO and finally evaporated to
2
4
dryness under reduced pressure. The residue was subjected to silica gel flash column
chromatography (FCC), using PhMe as eluant, to afford pure compound 16.
o
Yield: 1.54 g (49%); yellow solid, m.p. 113-114 °C, lit. [13] m.p. 111-113 C; R (F)
f
+
+ 1
0
.26; MS (ESI) m/z: 245.24 (M+Na) , 223.13 (M+H) ; H-NMR: δ 8.38 (dd, J = 8.4
and 0.9 Hz, 1H), 7.98 (dd, J = 8.4 and 0.9 Hz, 1H), 7.63 (t, J = 8.4 Hz, 1H), 7.52 (s,
13
1
1
H), 2.74 (s, 3H) ppm; C-NMR: δ 161.9, 148.3, 142.5, 140.0, 127.7, 125.7, 125.2,
23.9, 123.7, 25.6 ppm.
4
.1.3 2-Methyl-8-nitro-4-(pyridyl-2-ylmethoxy)quinoline (17)
To an ice-cold solution of (pyridin-2-yl)methanol (3.8 mL, 39.3 mmol) in DMI (3.8
t
mL) was added BuOK (0.71 g, 6.3 mmol) and, following its dissolution, the
temperature was lowered to-10 °C. Then, compound 16 (1.4 g, 6.3 mmol) was added
portion-wise over 50 minutes and the mixture was stirred at room temperature
overnight. The reaction mixture was diluted with water (50 mL) and the thus
resulting suspension was extracted thrice with EtOAc. The organic phases were
combined and washed twice with brine, dried over Na SO and evaporated to dryness
2
4
under reduced pressure. The residue was subjected to FCC, using the solvent system
L as eluant, to give pure compound 17.
Yield: 0.8 g (43%); orange solid, m.p. 159.5–160.5 °C; R (L) 0.25; MS (ESI) m/z:
f
+
+
+ 1
3
34.42 (M+K) , 318.42 (M+Na) , 296.50 (M+H) ; H-NMR: δ 8.65 (d, J = 4.3 Hz,
H), 8.43 (d, J = 8.4 Hz, 1H), 7.92 (d, J = 8.4 Hz, 1H), 7.76 (t, J = 8.4 Hz, 1H), 7.53
1
(d, J = 7.8 Hz, 1H), 7.48 (t, J = 7.8 Hz, 1H), 7.30 (dd, J = 7.8 and 4.3 Hz, 1H), 6.80
13
(
s, 1H), 5.43 (s, 2H), 2.67 (s, 3H) ppm; C-NMR: δ 163.4, 160.6, 155.4, 149.6,
47.9, 140.3, 137.1, 125.8, 123.7, 123.4, 123.3, 121.4, 121.2, 103.4, 71.2, 26.4 ppm;
Anal. Calcd for C H N O : C, 65.08; H, 4.44; N, 14.23. Found: C, 65.38; H, 4.23;
1
1
6
13
3
3
N, 14.02.
4
.1.4 2-Methyl-4-(pyridyl-2-ylmethoxy)quinoline-8-amine (13)
To a solution of compound 17 (0.8 g, 2.71 mmol) in aq. MeOH (80%) (10.7 mL)
were added FeCl ·6H O (32.4 mg) and activated carbon (32.4 mg) and the resulting
3
2

2
5
mixture was heated to reflux. Then, N H ·H O (0.33 mL, 6.8 mmol) was added
2
4
2
dropwise and the mixture was stirred at the same temperature for 2 hours. TLC
analysis indicated that the reaction was not complete. An additional amount of
N H ·H O (0.33 mL, 6.8 mmol) was then added and 1 h after the last addition, the
2
4
2
reaction mixture was left to attain rt and diluted with EtOAc (30 mL). The organic
phase was washed once with an aq. solution of 5% NaHCO and twice with brine,
3
dried over Na SO and evaporated to dryness under reduced pressure. The residue
2
4
was subjected to FCC, using the solvent system K as eluant, to afford pure compound
3.
Yield: 0.52 g (72%); brown solid, m.p. 129.5-130 °C; R (L) 0.23; MS (ESI) m/z:
1
f
+
1
2
1
6
3
1
7
66.40 (M+H) ; H NMR: δ8.63 (d, J = 4.8 Hz, 1H), 7.74 (td, J = 8.4 and 1.2 Hz,
H), 7.59 (d, J = 8.4 Hz, 1H), 7.57 (dd, J = 8.4 and 1.2 Hz, 1H), 7.28-7.22 (m, 2H),
.91(dd, J = 8.4 and 1.2 Hz, 1H), 6.66 (s, 1H) 5.39 (s, 2H), 4.93 (br.s, 2H), 2.63 (s,
1
3
H) ppm; C NMR:δ 157.2, 156.4, 149.3, 143.2, 137.0, 129.0, 128.2, 125.6, 122.9,
21.1, 120.2, 110.9, 109.9, 102.0, 70.6, 25.8 ppm; Anal. Calcd for C H N O: C,
1
6
15
3
2.43; H, 5.70; N, 15.84. Found: C, 72.13; H, 5.88; N, 16.12.
4
.1.5 8-Iodo-2-Methyl-4-(pyridyl-2-ylmethoxy)quinoline (18)
To a suspension of 13 (0.13 g, 0.5 mmol) in H O (1.5 mL) and 37% aq. HCl (0.2 mL,
2
2
.4 mmol) at 5 °C was added an aq. solution of 1.6 M NaNO (0.34 mL, 0.55 mmol)
2
dropwise. After 30 min of stirring at 5 °C, TLC analysis confirmedthe complete
conversion of compound 13 to its corresponding diazonium salt and thus an ice cold
aq. solution of 3 M KI (0.35 mL, 1.05 mmol) was added dropwise. The reaction
mixture was stirred at 5 °C for 10 min and at 25 °C for 1 h. Then, it was adjusted to
pH 9 with a saturated aq. solution of NaHCO and extracted thrice with EtOAc. The
3
organic phases were combined and washed once with a saturated aq. solution of
Na S O and twice with brine, dried over Na SO and evaporated to dryness under
2
2
3
2
4
reduced pressure. The residue was subjected to FCC, using the solvent system L, to
afford pure compound 18.
Yield: 0.14 g (74%); brown solid; mp 95.5-97.5 °C; R (L) 0.35; MS (ESI)
f
+
+ 1
m/z:377.31 (M+H) , 399.29 (M+Na) ; H-NMR: δ 8.64 (ddd, J = 4.8, 1.8 and 1.2 Hz,
H), 8.29 (dd, J = 7.5 and 1.2 Hz, 1H), 8.26 (dd, J = 7.5 and 1.2 Hz, 1H), 7.75 (td, J
7.8 & 1.8 Hz, 1H), 7.55 (unresolv. dt, J = 7.8 Hz, 1H), 7.28 (ddt, J = 7.8, 5.4 and
.6 Hz, 1H), 7.17 (dd, J = 8.4 and 7.8 Hz, 1H), 6.74 (s, 1H), 5.41 (s, 2H), 2.72 (s, 3H)
1
=
0

2
6
13
ppm; C-NMR: δ 161.4, 160.9, 155.9, 149.4, 147.6, 140.4, 137.1, 126.0, 123.1,
1
22.5, 121.3, 120.3, 102.5,102.4,71.0, 26.2 ppm; Anal. Calcd for C H IN O: C,
16 13 2
5
1.08; H, 3.48; N, 7.45. Found: C, 51.32; H, 3.20; N, 7.32.
4
.1.6 2-Methyl-8-(3-phenoxybenzyloxy)-4-(pyridyl-2-ylmethoxy)quinoline (12)
Method A (Ullmann reaction). To a solution of CuI (3.6 mg, 0.019 mmol), 1,10-
phenathroline (6.8 mg, 0.038 mmol) and Cs CO (124.0 mg, 0.38 mmol) in DMI (0.1
2
3
mL) were added 18 (71 mg, 0.19 mmol) and m-phenoxybenzyl alcohol (66 µL, 0.38
mmol) and the reaction mixture was stirred at 110 °Covernight. The mixture was then
directly subjected to FCC, using the solvent system L as eluant, to afford pure
compound 12.
Yield: 15 mg (18%); beige solid; mp 120-123.5 °C; R (L) 0.15; MS (ESI) m/z: 449.29
f
+
1
(
M+H) ; H-NMR (600 MHz, CDCl ): δ 8.64 (d, J = 4.2 Hz, 1H),7.84 (dd, J = 8.4
3
and 0.6 Hz, 1H), 7.75 (td, J = 7.8 and 1.8 Hz, 1H), 7.58 (d, J = 7.8 Hz, 1H), 7.33-7.24
m, 6H), 7.18 (unresolv. t, 1H), 7.07 (unresolv. tt, J = 7.2 Hz, 1H), 7.01 (d, J = 7.2
Hz, 1H), 6.99-6.96 (m, 2H), 6.92 (dd, J = 7.8 and 1.8 Hz, 1H), 6.73 (s, 1H), 5.41 (s,
(
1
3
4
1
1
H), 2.71 (s, 3H) ppm. C-NMR: 160.9, 160.8, 159.2, 157.5, 157.1, 156.2, 153.6,
49.4, 139.5, 137.0, 129.9, 129.7 (2C), 124.7, 123.2, 123.0, 121.6, 121.2, 121.1,
18.9 (2C), 118.0, 117.3, 114.0, 111.3, 102.5, 70.7, 70.6, 26.3 ppm; Anal. Calcd for
C H N O : C, 77.66; H, 5.39; N, 6.25. Found: C, 77.48; H, 5.60; N, 6.45.
29
24
2
3
Method B (Mitsunobu reaction). To a solution of quinolinol 2 (100 mg, 0.38 mmol)
and m-phenoxybenzyl alcohol (73 µL, 0.42 mmol) in dry THF (0.5 mL) was added
TPP (121 mg, 0.46 mmol) under argon and the mixture was cooled to 0 °C. Then,
DIAD (91 µL, 0.46 mmol) was added and the reaction mixture was stirred at rt for
1
.5 h. TLC analysis indicated that the reaction was not complete. Therefore,
additional m-phenoxybenzyl alcohol (73 µL, 0.42 mmol), TPP (121 mg, 0.46 mmol)
and DIAD (91 µL, 0.46 mmol) were added in two portions over 48 h. The reaction
mixture was evaporated to dryness and the residue was subjected to FCC using the
solvent system N as eluant, to afford compound 12 containing TPPO in the molar
1
ratio 1:3 (by H- NMR). This mixture was portioned between EtOAc and 2N HCl, the
aqueous layer was neutralized with NaOH and extracted with EtOAc. The organic
layer was dried and evaporated to dryness to leave pure compound 12 which was
crystallized on standing. Yield: 102 mg (60%).

2
7
4
.1.7 tert-Butyldiphenylsilyl (E)-3-(4-(tert-butyldiphenylsilyloxy)-3-methoxyphenyl)-
acrylate (19)
To a solution of FerA (0.19 g, 1 mmol) in dry DMF (1.5 mL) were added
sequentially, under argon, imidazole (0.16 g, 2.4 mmol) and TBDPSCl (0.6 mL, 2.3
mmol) and the mixture was stirred at rt for 3 h. The reaction mixture was then diluted
with water and extracted thrice with Et O. The organic phases were combined and
2
washed thrice with water and once with brine, dried over Na SO and evaporated
2
4
dryness under reduced pressure. The residue was subjected to FCC, using the solvent
system O as eluant, to afford pure ester 19.
+
1
Yield: 0.54 g (81%); pale-yellow oil; R (O) 0.33; MS (ESI) m/z 693.31 (M+Na) ; H-
f
NMR: δ 7.74-7.70 (m, 8H), 7.64 (d, J = 15.6 Hz, 1H), 7.46-7.35 (m, 12H), 6.97 (d, J
=
=
1.8 Hz, 1H), 6.88 (dd, J = 7.8 and 1.8 Hz, 1H), 6.71 (d, J = 7.8 Hz, 1H), 6.35 (d, J
13
15.6 Hz, 1H), 3.63 (s, 3H), 1.16 (s, 9 H), 1.14 (s, 9H) ppm; C-NMR: δ 166.1,
1
1
50.9, 147.6, 146.0, 135.3 (2C), 135.2 (2C), 133.1 (2C), 132.1 (2C), 130.0 (4C),
29.8 (4C), 128.1, 127.7 (4C), 127.6 (4C), 122.4, 120.4, 117.3, 111.0, 55.4, 27.0
(3C), 26.6 (3C), 19.8, 19.2 ppm.
4
.1.8 (E)-3-(4-(tert-Butyldiphenylsilyloxy)-3-methoxyphenyl)acrylic acid (20)
To an ice-cold solution of ester 19 (0.37 g, 0.55 mmol) in MeOH (7 mL) and THF
2.5 mL) was added an aq. solution of 10% K CO (2.3 mL) and the mixture was
(
2
3
stirred at rt for 1 h. The reaction mixture was diluted with cold brine (7 mL), cooled
to 0 °C, adjusted to pH 4-5 with a cold aq. solution of 5% citric acid and extracted
thrice with EtOAc. It should be noted that during acidification ferulic acid was also
produced. The organic phases were combined and washed twice with brine, dried
over Na SO and evaporated to dryness under reduced pressure at a bath temperature
2
4
not exceeding 30 °C. The residue was subjected to FCC, using the solvent system P
for elution, to afford pure acid 20.
+
1
Yield: 0.14 g (61%); yellow oil; R (P) 0.20;MS (ESI) m/z:433.21 (M+H) ; H-NMR:
f
δ 7.80-7.76 (m, 4H), 7.72 (d, J = 16.2 Hz, 1H), 7.50 (tt, J = 7.2 and 1.4 Hz, 2H), 7.44
(t, J = 7.2 Hz, 4H), 7.04 (d, J = 1.5 Hz, 1H), 6.95 (dd, J= 8.4 and 1.5 Hz, 1H), 6.78
1
3
(
d, J= 8.4 Hz, 1H), 6.32 (d, J= 16.2 Hz, 1H), 3.69 (s, 3H), 1.20 (s, 9H) ppm. C-
NMR: δ 172.1, 150.9, 149.7, 149.2, 135.3 (4C), 133.1 (2C), 129.8 (2C), 127.7, 127.6
4C), 122.5, 120.4, 114.7, 111.2, 55.3, 26.6 (3C), 19.8 ppm.
(

2
8
4
.1.9 2,5-Dioxopyrrolidin-1-yl (E)-3-(4-(tert-butyldiphenylsilyloxy)-3-methoxyphe-
nyl)acrylate (21)
To an ice-cold solution of acid 20 (0.12 g, 0.28 mmol) in dry THF (1 mL) were added
HOSu (0.08 g, 0.72 mmol) and DCC (0.08 g, 0.4 mmol) under argon and the mixture
was stirred for 30 min at 0 °C and 5 h at rt. TLC analysis indicated that the reaction
was not completed, therefore, additional DCC (0.01 g, 0.05 mmol) was added in the
same manner and the reaction mixture was further stirred overnight at rt. Then, water
(0.15 mL) and gl. AcOH (0.05 mL) were added and after 30 min of stirring, the
precipitate was filtered off and washed on the filter with cold EtOAc. The filtrate was
diluted with EtOAc, washed thrice with an aq. solution of 5% NaHCO , twice with
3
water and once with brine, dried over Na SO and evaporated to dryness under
2
4
reduced pressure. The residue was subjected to FCC, using the solvent system P as
eluant to afford pure 'active' ester 21.
+
Yield: 126 mg (85%); yellow oil; R (P); 0.34; MS (ESI) m/z: 1081.23 (2M+Na) ,
f
+
+ 1
5
68.31 (M+K) , 552.50 (M+Na) ; H-NMR: δ 7.77 (d, J = 16.2 Hz, 1H), 7.70-7.67
m, 4H), 7.43-7.39 (m, 2H), 7.37-7.33 (m, 4H), 6.95 (d, J = 1.8 Hz, 1H), 6.89 (dd, J =
.8 and 1.8 Hz, 1H), 6.70 (d, J = 8.4 Hz, 1H), 6.36 (d, J = 16.2 Hz, 1H), 3.60 (s, 3H),
(
7
2
1
5
5
1
3
.86 (br.s, 4H), 1.12 (s, 9H) ppm; C-NMR: δ 169.4 (2C), 162.3, 151.0, 150.2, 148.8,
35.3 (4C), 132.9 (2C), 129.9 (2C), 127.6 (4C), 127.2, 123.1, 120.5, 111.3, 108.8,
5.3, 26.6 (3C), 25.6 (2C), 19.8 ppm; Anal. Calcd for C H NO Si: C, 68.03; H,
3
0
31
6
.90; N, 2.64. Found: C, 68.34; H, 5.49; N, 2.41.
4
.1.10 (3-Amino-2,6-dichlorophenyl)methanol (23)
To a solution of nitro compound S11 (1.0 g, 4.5 mmol) in aq. MeOH (80%) (16.2
mL) were added FeCl ·6H O (36.0 mg) and activated carbon (36.0 mg) and the
3
2
mixture was heated to reflux. Then, N H ·H O (0.53 mL, 10.7 mmol) was added
2
4
2
drop-wise and the reaction mixture was stirred at the same temperature for 2 h. TLC
analysis indicated that the reaction was not completed, therefore, additional
N H ·H O (0.53 mL, 10.7 mmol) was added and after 2 more hours at the same
2
4
2
temperature, the reaction was completed. The reaction mixture was left to attain rt
and then diluted with EtOAc. The organic phase was washed once with an aq.
solution of 5% NaHCO and twice with brine, dried over Na SO and evaporated to
3
2
4
dryness under reduced pressure. The residue was subjected to FCC, using the solvent
system S as eluant, to afford pure aniline derivative 23.