The Antiangiogenic Activity of Naturally Occurring and Synthetic Homoisoflavonoids from the Hyacinthaceae (sensu APGII)

Article abstract of DOI:10.1021/acs.jnatprod.8b00989

Excessive blood vessel formation in the eye is implicated in wet age-related macular degeneration, proliferative diabetic retinopathy, neovascular glaucoma, and retinopathy of prematurity, which are major causes of blindness. Small molecule antiangiogenic

Full text of DOI:10.1021/acs.jnatprod.8b00989

Article
Cite This: J. Nat. Prod. XXXX, XXX, XXX−XXX
pubs.acs.org/jnp
The Antiangiogenic Activity of Naturally Occurring and Synthetic
Homoisoflavonoids from the Hyacinthaceae (sensu APGII)
Sianne Schwikkard,†,‡ Hannah Whitmore, Kamakshi Sishtla, Rania S. Sulaiman,
‡
§
§,∥,#
Trupti Shetty,
§,∥
§
,⊥
‡
‡
†
Halesha D. Basavarajappa, Catherine Waller, Alaa Alqahtani, Lennart Frankemoelle,
†
∇,%
○
○
&
Andy Chapman, Neil Crouch,
Wolfgang Wetschnig, Walter Knirsch, Jacky Andriantiana,
‡
‡,%
,‡,%
,§,∥,⊥
Eduard Mas-Claret, Moses K. Langat, Dulcie Mulholland,*
and Timothy W. Corson*
†
School of Life Sciences, Pharmacy and Chemistry, Kingston University, Kingston-upon-Thames, KT1 2EE, U.K.
‡
Natural Products Research Group, Department of Chemistry, Faculty of Engineering and Physical Sciences, University of Surrey,
Guildford, GU2 7XH, U.K.
§
∥
Eugene and Marilyn Glick Eye Institute, Department of Ophthalmology; Department of Pharmacology and Toxicology;
⊥
Department of Biochemistry and Molecular Biology, Indiana University School of Medicine, 1160 West Michigan Street,
Indianapolis, Indiana 46202, United States
#
Department of Biochemistry, Faculty of Pharmacy, Cairo University, Cairo, Egypt
∇
Biodiversity Economy, South African National Biodiversity Institute, P.O. Box 52099, 4007 Berea Road, Durban, South Africa
○
Institute of Biology, NAWI Graz, University of Graz, 8010 Graz, Austria
&
%
Parc Botanique et Zoologique de Tsimbazaza, Rue Fernand Kassanga, Antananarivo 101, Madagascar
School of Chemistry and Physics, University of KwaZulu-Natal, Durban, 4041, South Africa
*
S Supporting Information
ABSTRACT: Excessive blood vessel formation in the eye is
implicated in wet age-related macular degeneration, proliferative
diabetic retinopathy, neovascular glaucoma, and retinopathy of
prematurity, which are major causes of blindness. Small molecule
antiangiogenic drugs are strongly needed to supplement existing
biologics. Homoisoflavonoids have been previously shown to have
potent antiproliferative activities in endothelial cells over other cell
types. Moreover, they demonstrated a strong antiangiogenic potential
in vitro and in vivo in animal models of ocular neovascularization.
Here, we tested the antiangiogenic activity of a group of naturally
occurring homoisoflavonoids isolated from the family Hyacinthaceae and related synthetic compounds, chosen for synthesis
based on structure−activity relationship observations. Several compounds showed interesting antiproliferative and
antiangiogenic activities in vitro on retinal microvascular endothelial cells, a disease-relevant cell type, with the synthetic
−
4
chromane, 46, showing the best activity (GI50 of 2.3 × 10 μM).
4
,5
bnormal formation of new blood vessels in the eye is
associated with blindness in many ocular diseases such as
required for the intravitreal injections. Moreover, resistance
is a problem: about 30% of wet AMD patients are resistant to
A
6
retinopathy of prematurity (ROP) affecting children, pro-
liferative diabetic retinopathy (PDR), the wet form of age-
related macular degeneration (AMD) and neovascular
these biologics. Currently, there is no FDA approved small
molecule for the treatment of ocular neovascularization.
Therefore, there is an unmet need to develop novel and
specific antiangiogenic small molecule therapies to comple-
ment and combine with existing drugs for ROP, PDR, and wet
AMD.
Natural products continue to provide appealing lead
compounds for treatment of diseases. Naturally occurring
homoisoflavonoids are reported to have diverse biological
activities including antioxidant, anti-inflammatory, and anti-
1
glaucoma affecting adults and elderly people, respectively.
The newly formed vasculature is fragile and leaky, causing
hemorrhage and accumulation of fluids in the retina. If left
untreated, the resulting edema and fibrotic scarring can lead to
2
irreversible vision loss. The current pharmacotherapeutic
7
mainstays for these diseases are biologics targeting the vascular
endothelial growth factor (VEGF) such as bevacizumab,
3
ranibizumab, and aflibercept. Despite being successful in
suppressing disease progression, these large molecule therapies
are associated with some undesirable ocular and systemic side
effects as well as the time-consuming regular visits to hospitals
Received: November 23, 2018
©
XXXX American Chemical Society and
American Society of Pharmacognosy
A
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
8
angiogenic activities. The Hyacinthaceae is one of the families
available from a commercial supplier enabling us to isolate
large amounts of compounds for screening purposes.
rich in homoisoflavonoids distributed in different parts of the
plants. The Hyacinthaceae (sensu APGII) consists of
approximately 900 species, in about 70 genera. The family
can be divided into four subfamilies, the widespread
Hyacinthoideae, Urgineoideae, and Ornithogaloideae and the
In addition, on the basis of our cell-based structure−activity
observations, a number of known (44, 47, 48) and previously
undescribed homoisoflavonoids (42, 43, 45, 46, 49, and 50)
were synthesized and investigated for antiangiogenic activity,
including the enantioselective synthesis of compound 45 with
the 3R configuration. Compound 46, a previously undescribed
compound with a chromane skeleton, showed significantly
greater activity than the related chromanones (45 and 47).
small subfamily Oziroeoideae which is restricted to South
̈
America. For the most part, the compounds found are
subfamily restricted. Homoisoflavonoids and spirocyclic
nortriterpenoids characterize the Hyacinthoideae, bufadieno-
lides the Urgineoideae, and cardenolides and steroidal
glycosides the Ornithogaloideae. To date, no phytochemical
̈
investigations have been reported of the Oziroeoideae.
RESULTS AND DISCUSSION
■
9
The EtOH extract of the bulbs of Chionodoxa luciliae Boiss
yielded compound 16, para-hydroxybenzaldehyde, and homo-
isoflavonoids 15, 17 and 21-24. Compound 17 was acetylated
and yielded the 7-acetate (19) and 3,7-diacetate (20).
Compound 16 is likely to be a degradation product of
Members of the Hyacinthoideae subfamily have a long
tradition of being used by traditional healers to treat a wide
range of complaints. Phytochemical investigations of the genus
1
0
11
Eucomis L’Her
bouria ovatifolia (Baker) Jessop, Merwilla plumbea (Lindl.)
́
., Ledebouria revoluta (L.F.) Jessop, Lede-
12
compound 21. All compounds have been reported previously
1
3
Speta, all widely used in African traditional medicine, have
9,25
from several sources
but this is the first report of the
been reported.
isolation of compounds 17, 22 and 23 from a Chionodoxa
species. Structures were determined using NMR techniques
and were confirmed by comparison against literature data; the
configurations at C-3 were confirmed using circular dichroism
spectroscopy and optical rotation measurements.
Previously, a homoisoflavanone, cremastranone, isolated
from the orchid Cremastra appendiculata (D.Don) Makino,
was shown to have antiangiogenic activity in the ocular
context, but unfortunately, the configuration at C-3 was not
1
4−16
reported.
This compound was also previously isolated
The MeOH and butanol extracts of the bulbs of Eucomis
autumnalis yielded compounds 22 and 30−39, and the EtOH
extract of the bulbs of E. bicolor yielded compounds 40 and 41.
Compounds 31 and 32 have been isolated previously from
17
from the Hyacinthoideae genera Muscari Mill. and Merwilla
1
3
Speta. More recently, a synthetic racemate of this compound
showed strong antiangiogenic activity against human retinal
microvascular endothelial cells (HRECs), with lesser effects on
26
both E. autumnalis and E. bicolor, compound 36 from E.
1
8
2
7−29
nonendothelial cell types. A synthetic regioisomer of
cremastranone (SH-11052) also had antiangiogenic activities
autumnalis,
bicolor,
compounds 30, 40, and 41 from E.
2
7,30,31
32
compound 33 from E. montana, compound
1
9
17
in vitro. Subsequently, our SAR campaign identified a
38 from Muscari armeniacum, compound 39 from Scilla
33 23
promising synthetic analogue of cremastranone, named SH-
persica, and compound 37 from Ledebouria ovatifolia.
1
1037, with more potent and selective antiproliferative effects
Compound 35, 3R-(4′-hydroxybenzyl)-6,8-dihydroxy-5,7-
dimethoxy-4-chromanone, is a homoisoflavonoid that has not
been reported previously with a fully substituted ring A. HRMS
showed a protonated molecular ion at m/z 347.11254
corresponding to a molecular formula of C H O . Character-
on HRECs compared to endothelial cells of macrovascular
2
0
origin. Moreover, SH-11037 demonstrated strong antiangio-
20
genic potential in vivo in animal models of ROP and
21
choroidal neovascularization. We have also isolated known
and novel homoisoflavonoids from Massonia Houtt. species,
some of which showed selective antiproliferative effects on
1
8
18
7
istic peaks at δ 4.37 (H-2α, dd, 11.4, 4.1 Hz), δ 4.20 (H-2β,
H
H
dd, 11.4, 7.2 Hz), δ
2.79 (H-3, m), δ 3.17 (H-9A, dd, 14.0,
H
H
2
2
HRECs. Given the selective and potent antiangiogenic
activity of cremastranone and derivatives, the search for
further antiangiogenic homoisoflavonoids is warranted. In
4.4 Hz), and δ
H
2.70 (H-9B, dd, 14.0, 10.6 Hz) were indicative
of a 3-benzyl-4-chromanone structure, and resonances at δ
H
6.79 (2H, 3′ and 5′, 8.1 Hz) and δ 7.11 (2H, 2′ and 6′, 8.1
H
20
parallel with our medicinal chemistry approach, we have here
further explored natural-source and related synthetic homoiso-
flavonoids as antiangiogenic agents.
Hz) showed that the B ring was para- disubstituted. The two
H-2 and two H-9 resonances showed correlations with the C-4
keto group carbon resonance at δ 192.0. Two methoxy group
C
proton resonances and three exchangeable hydroxy group
proton resonances (δH 5.59, 6-OH; δH 5.39, 8-OH and δH
Homoisoflavonoids and related compounds isolated from
seven species of Hyacinthaceae have been investigated in this
study for antiangiogenic activity. The isolation and structure
determination of compounds from Ledebouria socialis (Baker)
Jessop (1, 2), Ledebouria ovatifolia (Baker) Jessop (3−14),
4
.74, 4′-OH) could be assigned from correlations seen in the
HMBC spectrum (Figure 1). A specific rotation of −55.6 and
negative Cotton effect at 290 nm indicated a 3R-config-
11
uration. Homoisoflavonoids with a fully substituted ring A
Rhodocodon campanulatus Knirsch (25, 26), Mart-Azorın and
́
Wetschnig, Rhodocodon aff. intermedius Knirsch, Mart-Azorın
́
and Wetschnig (27−29), have been reported by us previously,
and compounds from these sources were screened as part of
2
3,24
this work.
In addition, Chionodoxa luciliae Boiss, Eucomis
autumnalis (Mill.) Chitt., and Eucomis bicolor Baker were
reinvestigated to provide material for screening and a
previously unreported homoisoflavonoid, 35, along with
previously reported ones were isolated, some for the first
time from these sources. These species were chosen for re-
extraction due to large amounts of bulb material being
Figure 1. HMBC correlations (H → C) for compound 35.
B
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Scheme 1. General Scheme for Synthesis of 3-Benzylidene Homoisoflavonoids
a
Scheme 2. Reactions of Compound 44
a
Iridium catalyst = (S)-(−)-2-[2-(diphenylphosphino)phenyl]-4-isopropyl-2-oxazolium-(1,5)-cyclooctadiene iridium(I) tetrakis(3,5-bis-
trifluoromethyl)phenyl borate).
(
are rare with only two similar compounds being reported
previously: (4′-hydroxybenzyl)-5,6-dihydroxy-7,8-dimethoxy-
hydrogenate compound 44 to form 45, was synthesized
37
using a known methodology and is shown in Scheme 2.
34
4
-chromanone, from Bellevalia eigii, and (4′-hydroxy)-5,8-
Chalcone 50 was synthesized via a Claisen−Schmidt
38
dihydroxy-6,7-dimethoxy-4-chromanone, from Muscari como-
condensation, using known methods.
3
5
sum.
The syntheses of compounds 44, 47, and 48 have been
20,39
Previous work on synthetic analogues of cremastranone
demonstrated that much potential remained for the improve-
ment of activity. Previously, 3-benzyl analogues had been
synthesized as racemic mixtures and we wished to investigate
the possibility of synthesizing a single enantiomer via selective
hydrogenation of the 3,9-double bond of a 3-benzylidene
analogue. In addition, the effect of producing the chromane
derivative of 5,6,7-trimethoxy-3-(3′-hydroxy-4′-methoxybenzy-
lidene)-4-chromanone was of interest. Homoisoflavonoids
reported,
while the synthesis of compounds 42, 43, 45, 46,
49, and 50 has not been reported before. Compounds 42 and
43 were both synthesized in low yield (3 and 3.5%,
respectively). The first two steps proceeded well, with the
low yield being a result of poor reaction completion in step
three and extensive chromatography to produce the pure
product. The structures were confirmed by high resolution
mass spectrometry and NMR analysis.
Compound 44 was used as a starting point for our
investigation into the asymmetric hydrogenation of the 3,9-
double bond in the benzylidene-type homoisoflavonoids using
the iridium catalyst (Scheme 2) for the hydrogenation. The
reaction proceeded quantitatively to give compound 45.
3
6
4
(
2
2−49 were synthesized following known methods
Schemes 1 and 2). An iridium catalyst, the (S)-4-tert-butyl-
-[2-(diphenylphosphino)phenyl]-2-oxazoline complex, also
known as iridium-phosphino-oxazoline or phox, used to
C
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Table 1. Antiproliferative Activity of Homoisoflavonoids and Co-Isolated Compounds against Endothelial and Ocular Tumor
a
Cells (Growth Inhibitory Concentration, GI , Shown in μM)
50
a
ND, no data.
D
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Table 2. Antiproliferative Activity of Natural-Source and Synthetic Homoisoflavonoids and Co-Isolated Compounds against
HRECs and Other Ocular Cell Types (Growth Inhibitory Concentration, GI , Shown in μM)
50
A systematic conformational search of 3R- and 3S-isomers of
,6,7-trimethoxy-(3′-hydroxy-4′-methoxybenzyl)-4-chroma-
it has been applied in the past by our group with satisfactory
41
5
results.
none was performed via molecular mechanics force field
calculation (MMFF) built in the software Spartan’16. Con-
formers under 3.0 kJ/mol relative energy cutoff were selected
and optimized using density functional theory (DFT)
calculations at the B3LYP/6-31G+(d,p) level built into
Gaussian 09. Once optimized, their ECD spectra were
simulated using time dependent density functional theory
The measured ECD spectrum of 45 showed a positive
Cotton effect at λ (Δε) 326 (+1.3) and a negative Cotton
effect at λ (Δε) 284 (−1.9), which is in accordance with the
theoretical calculated spectrum for the 3R-enantiomer,
indicating the 3R configuration for compound 45 (Supporting
42
Information). For comparative purposes, compound 44 was
hydrogenated using Pd/C to give the racemic mixture, 46. This
reaction resulted in the reduction of both the 3,9-double bond
as well as the carbonyl group at C-4, as shown by the loss of
(
TDDFT) at the B3LYP/6-31+G(d,p) level. A polarizable
continuum model (IEFPCM) was applied during the TDDFT
calculation to simulate the presence of acetonitrile, which was
the solvent used in the experimental ECD. The ECD spectra
obtained for each conformer were Boltzmann weighted and
the fully substituted carbon resonance at δ 179.8 (C-4) and
C
the new methylene resonance at δ 37.5 (C-4). Coupling could
C
clearly be seen in the COSY spectrum between the resonance
at δ 2.14 (1H, m, H-3) and the resonance at δ 2.53 (2H, dd,
H
H
40
compared to the experimental ECD spectrum of 45. The 6-
1+G(d,p) basis set was applied, as it provides a fine balance
between accuracy and computational time needed. In addition,
J = 4.1 and 7.4 Hz, H-4). Compound 44 was then successfully
hydrogenated using Wilkinson’s catalyst (tris-
(triphenylphosphine)rhodium(I) chloride) to give compound
3
E
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 2. Homoisoflavonoids block the proliferation of HRECs. Dose−response curves for representative active compounds’ effects on
proliferation of the indicated cell types as assessed by alamarBlue fluorescence. Mean ± SEM, n = 3−6 wells.
2
0
4
7, the previously synthesized racemate. Compound 48 was
types, we also employed two ocular cancer cell lines, uveal
melanoma 92-1 and retinoblastoma Y79. Again, ideal
candidates for further investigation as antiangiogenic homoiso-
flavonoids should show lower activity against these other
ocular cell lines derived from uveal melanocytes and
photoreceptor precursors, respectively.
synthesized to determine whether the presence of a 5-OH
group was significant for activity. Many natural homoiso-
flavonoids have a hydroxyl group at C-5. The selective
demethylation was confirmed by the appearance of a sharp
singlet at δ 12.45, due to the proton of a H-bonded hydroxyl
H
group at C-5. In addition, three methoxy groups were noted
instead of four.
Compound 49 was synthesized from 3,4,5-trimethoxyphenol
and 4-chloro-3-hydroxybenzaldehyde, and the expected
While many of the compounds lacked activity against the
cell types, some compounds were reasonably potent and
selective for the endothelial cell types (Tables 1 and 2; Figure
2). The results would suggest some structural features are
significant in determining the activity of the homoisoflavonoid.
Optimal activity was observed when the A-ring was either
5,6,7-trisubstituted or 5,7-disubstituted. If only two substitu-
ents were present, at least one of them would need to be an
acetyl rather than a hydroxy group. 3′,4′-Disubstitution in ring
B produced the highest activity, in particular when combined
with the presence of a 3/9 double bond. Homoisoflavonoids
with the bulkier 3′-acetyl group were more active than ones
with the 3′-hydroxyl group. The presence of a cyclobutane ring
rendered the homoisoflavonoid inactive, regardless of the
substituents on rings A and B. The importance of a 5,6,7-
trisubstituted A-ring was confirmed by the synthetic samples,
with the trisubstituted compounds (44, 45, 46, 47, 48, and 49)
showing better activity than the disubstituted compounds (42
and 43). Interestingly, the 3R-enantiomer (45) showed greater
activity than the precursor with a double bond in the 3,9-
position (44) as well as greater activity than the racemic
mixture (47). This serves to emphasize the importance of
being able to synthesize these homoisoflavonoids enantiose-
lectively. The complete reduction of both the 3,9-double bond
and the carbonyl group at C-4 produced a highly active
compound (46) (Table 2, GI (HREC) of 2.3 × 10− μM,
product was formed. A HRESIMS value of m/z 377.0786
+
[
M + H] indicated a molecular formula of C H O Cl. The
19
17
6
1
structure of the product was confirmed by key H NMR
resonances at δ 4.99 (2H, d, J = 1.8 Hz, H-2), 7.75 (1H, bs,
H
H-9), and 6.16 (1H, s, H-8) and three methoxy group three-
proton resonances at δ 3.75 (3H, s), 3.82 (3H, s), and 3.91
H
(
8
8
3H, s). The B-ring resonances at δ 7.15 (1H, dd, J = 2.4 and
H
.6 Hz, H-2′), 6.91 (1H, d, J = 2.4, H-6′), and 6.83 (1H, d, J =
.6, H-3′) supported the 4′,5′-substitution pattern.
Compound 50, a chalcone, was synthesized from 3,4-
dimethoxybenzaldehyde and 4′-ethoxy-2′-hydroxy-3′-methyl-
acetophenone via a Claisen−Schmidt condensation. The
+
HRESIMS gave a [M + H] ion at m/z 343.1545, giving a
molecular formula of C H O for the compound. The
20
22
5
chalcone structure was confirmed by a pair of doublets in the
1
H NMR spectrum at δ 7.47 (1H, d, J = 15.2 Hz, H-2) and
H
7
.84 (1H, d, J = 15.2 Hz, H-3) indicative of E geometry.
Antiproliferative Activities of Homoisoflavonoids. To
find novel antiangiogenic homoisoflavonoids, compounds were
tested for their antiproliferative activities on HRECs. To assess
effects on other normal cell types, we tested compounds
against primary macrovascular human umbilical vein endothe-
lial cells (HUVECs; a nontarget endothelial cell type) or the
retinal pigment epithelial cell line, ARPE-19, a nontarget ocular
cell type. Ideal antiangiogenic candidates should have lesser
effects on these nontarget cell types than on HRECs. To
further assess potential cytotoxic effects against ocular cell
4
5
0
−
4
GI (ARPE-19) of 3.4 × 10 μM, and GI (92-1) of 52 μM).
5
0
50
This significant increase in activity with the loss of the carbonyl
group at C-4 has not been reported before, although this
compound was synthesized simultaneously by the Seo lab
(personal communication). Compound 49, with a 4′-chloro-
F
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 3. Potent natural-source homoisoflavonoids significantly block the ability of HRECs to form tube-like structures. Top: Representative
images of the tube formation of HRECs on Matrigel at the highest concentrations of each compound tested plus DMSO control, scale bars = 250
μm. Bottom: Quantification of the extent of tube formation was measured as the tubule length of HRECs treated with compounds compared to
DMSO control. Mean ± SEM, n = 3−6 wells. Representative results from duplicate experiments. *P < 0.05, ***P < 0.001 compared to DMSO
control (one-way ANOVA with Dunnett’s posthoc tests).
Figure 4. Potent synthetic homoisoflavonoids significantly block the ability of HRECs to form tube-like structures. Top: Representative images of
the tube formation of HRECs on Matrigel at the highest concentrations of each compound tested plus DMSO control, scale bars = 500 μm.
Bottom: Quantification of the extent of tube formation was measured as the tubule length of HRECs treated with each compound compared to
DMSO control. SH-11037 is a positive control. Mean ± SEM, n = 6 wells. Representative results from duplicate experiments. *P < 0.05, ***P <
0.001 compared to DMSO control (one-way ANOVA with Dunnett’s posthoc tests).
3
′-hydroxyphenylbenzylidene group at C-3, produced good
activity with selectivity for HRECs (Table 2, GI50 (HREC) of
.11 μM, GI50 (92-1) of 3.6 μM), compared to that of
compound 44 (Table 2, GI (HREC) of 2.2 μM, GI (92-1)
In Vitro Antiangiogenic Efficacy of the Potent
Homoisoflavonoids. After identifying some of the most
potent antiproliferative compounds with some selectivity for
HRECscompounds 13, 27, 28, 45, 46, 48, and 50we
further investigated their potential in inhibiting the ability of
HRECs to form tubes. Matrigel based tube formation is the
gold-standard in vitro assay that recapitulates most of the
physiological angiogenesis events such as cell proliferation,
migration, and cell−cell adhesion, as endothelial cells form
0
50
50
of 8.2 μM). By way of comparison, cremastranone as reported
was substantially less potent but still selective (GI50 (HREC)
of 0.22 μM, GI (HUVEC) of 0.38 μM, GI (92-1) of 48 μM,
GI50 (Y79) of 9.8 μM).
5
0
50
20
G
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
4
3
three-dimensional tube-like structures. Interestingly, all three
of the natural-source compounds (13, 27, and 28), inhibited
the ability of HRECs to form tubes in a concentration-
dependent manner (Figure 3). Three potent synthetic
compounds, 48, 45, and 50, were also effective (Figure 4),
while the most potent antiproliferative compound, 46,
maintained this low-dose efficacy in the tube formation assay
ESI mass spectra were either recorded on a Bruker MicroToF mass
spectrometer using an Agilent 1100 HPLC to introduce samples
(
University of Oxford), a Micromass Quattro Ultima mass
spectrometer using a Waters Alliance HPLC to introduce samples
University of Surrey), or a Waters Xevo G2-S (National Mass
(
Spectrometry Facility, Swansea). Optical rotations were recorded in
CDCl on a JASCO P-1020 polarimeter (University of Surrey), IR
3
spectra were recorded on a PerkinElmer 2000 spectrometer
(University of Surrey), and UV spectra were recorded using a
Biochrom libraS60 in MeOH or acetonitrile in a 1 cm cell (University
of Surrey). CD spectra were recorded using a Chirascan
spectropolarimeter at room temperature in a 1 cm cell in MeOH
(Figure 5).
(
University of Surrey). Reagents were purchased from Sigma-Aldrich
Company Ltd., Gillingham, U.K., and used without further
purification.
Plant Material. The sourcing of plant material, extraction
methodology, and isolation of compounds from the bulbs of L.
ovatifolia (Bak.) Jessop, L. socialis (Bak.) Jessop, Rhodocodon
campanulatus Knirsch, Mart-Azorın
́
and Wetschnig and Rhodocodon
24,25
aff. intermedius have been described before.
Bulbs of Chionodoxa
luciliae Boiss. (Gigantea Hort), Eucomis autumnalis (Mill.) Chitt., and
E. bicolor Baker were purchased from the Dix Export Company
(Netherlands) in 2012 and grown to produce a flowering specimen to
confirm identification. Voucher specimens (DAM 2012 CL, EA and
EB) are retained at the University of Surrey.
Extraction and Isolation. Dried bulbs of Chionodoxa luciliae (3.0
kg) were extracted in EtOH at room temperature for 24 h on a shaker.
The resulting extract (119 g) was separated using gravity column
chromatography over silica gel (Merck 9385) and solvent mixtures of
Figure 5. Highly potent synthetic homoisoflavonoid 46 significantly
blocks the ability of HRECs to form tube-like structures. Top:
Representative images of the tube formation of HRECs on Matrigel,
scale bars = 500 μm. Bottom: Quantification of the extent of tube
formation was measured as the tubule length of HRECs treated with
compound compared to DMSO control. Mean ± SEM, n = 6 wells.
Representative results from duplicate experiments. ***P < 0.001
compared to DMSO control (one-way ANOVA with Dunnett’s
posthoc tests).
−
3
8
0:20 CH Cl :EtOAc (75 cm , 75 fractions) and 70:30
2 2
−
3
CH Cl :EtOAc (75 cm , 10 fractions). Further purification was
2
2
necessary in some cases using PTLC (Macherey-Nagel 0.25 mm, silica
gel 60 with fluorescent indicator UV254), further column chromatog-
raphy or acetylation of complex mixtures to obtain pure compounds.
Fraction 68 yielded 15 (11 mg), fraction 70 yielded, after further
purification, 16 (2 mg) and 17 (10 mg), fraction 72 yielded 18 (17
mg) and after acetylation of a mixture containing 17, the
monoacetate, 19 (15 mg), and diacetate (12 mg), 20. Fraction 74
yielded, after further purification, 21 (20 mg) and 22 (20 mg),
fraction 77 yielded 23 (50 mg), and fraction 80 yielded 24 (15 mg).
Compounds were identified using 2D NMR and CD methods and by
comparison against literature data as referenced earlier.
CONCLUSIONS
■
A group of naturally occurring homoisoflavonoids isolated
from the Hyacinthaceae as well as compounds synthesized
have been tested for their antiangiogenic activities. The most
potent compounds, 13, 28, 29, 42, 43, 45, 46, 48 and the
chalcone 50, demonstrated selective antiproliferative activities
on endothelial cells compared to nonendothelial cell types,
with concentration-dependent antiangiogenic effects in vitro
on HRECs, a disease-relevant cell type. Recently, intracellular
target proteins for some other bioactive homoisoflavonoids
have been identified: ferrochelatase as a target of cremas-
Chopped bulbs of E. autumnalis (4 kg) were extracted with MeOH
(3 L) for 48 h, yielding an extract which was partitioned with CH Cl
2
2
(300 mL) to yield extract 1 (5.1 g) after solvent evaporation. Bulb
material was further extracted in 1-butanol (2.5 L) for 24 h, yielding
extract 2 (26.9 g) after solvent evaporation. Extract 1 was separated
using gravity column chromatography over silica gel (Merck 9385)
−
3
and a gradient solvent elution method (40 cm , 63 fractions) using
−
3
solvent mixtures of 50:50 hexane:CH Cl (40 cm , 6 fractions),
2
2
−3
−3
2
4
0:80 hexane:CH Cl (40 cm , 9 fractions), 100% CH Cl (40 cm ,
2
2
2
2
−3
fractions), 50:50 CH Cl :EtOAc (40 cm , 27 fractions), and 50:50
CH Cl :MeOH (20 cm , 17 fractions). Fraction 24 yielded, after
2
2
4
4
−
3
tranone, soluble epoxide hydrolase as a target of SH-11037
2
2
45
(
a synthetic Boc-Phe-derivatized homoisoflavonoid), and
further purification, 30 (3.5 mg), and fractions 26 and 27 yielded,
after further purification 31 (139.5 mg) and 32 (1.9 mg). Fractions
34−38 yielded, after further purification, 22 (101 mg), fractions 41−
inosine monophosphate dehydrogenase 2 as a target of
4
6
sappanone A. The relative effects of the natural products
described here on each of these enzymes will be a valuable
topic for future exploration. Our results will open the doors to
the development of further synthetic analogues with higher
potency and better antiangiogenic activities to treat blinding
eye diseases caused by pathological neovascularization.
43 yielded, after further purification, 33 (15.7 mg) and 34 (7.1 mg).
Fraction 45 yielded, after further purification, the previously
2
7
unreported compound 35 (37.5 mg). Red-brown oil; [α]
−55.6
D
(
c 0.03, CH OH); ECD (c 0.03, MeOH) 290 nm (−5), 230 nm
3
−1 1
(
−2); IR (NaCl) ν 3327, 3007, 2938, 2400, 2000 1668 cm ; H
max
13
NMR (CDCl , 500 MHz), C NMR (CDCl , 125 MHz) Table 3;
3
3
+
+
HRESIMS m/z 347.1125 [M + H] (calcd for [C H O + H ],
1
8
18
7
EXPERIMENTAL SECTION
■
347.1130).
General Experimental Procedures. NMR spectra were
recorded on a 500 MHz Bruker AVANCE NMR instrument in
either CDCl or CD OD at room temperature. All chemical shifts (δ)
Extract 2 was separated using gravity column chromatography over
silica gel (Merck 9385) and a gradient solvent elution method (40
−
3
−3
cm , 54 fractions) using solvent mixtures of 100% CH Cl (40 cm ,
3
3
2
2
−
3
are in ppm and referenced to the relevant solvent references, 7.26
20 fractions), 50:50 CH Cl :EtOAc (40 cm , 20 fractions), and
50:50 CH Cl :MeOH (20 cm , 14 fractions). Fractions 5−11
yielded after further purification 36 (1.5 mg). Fractions 31−33
2 2
1
−3
ppm (CDCl ) and 4.87 ppm (CD OD) for H NMR spectra and
3
3
2
2
1
3
7
7.23 ppm (CDCl ) and 49.15 ppm (CD OD) for C NMR spectra.
3
3
H
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Table 3. NMR Data ( H 500 MHz, 13C 125 MHz) for
1
for 2 h. The mixture was poured into ice water and extracted into
diethyl ether. The extract was washed with 3 N NaOH (30 mL) and
Compound 35 (in CDCl , J in Hz)
3
water (50 mL) and dried over MgSO . The product (3,5-
4
position
2
1H
13C
dimethoxychromone) was heated at 80 °C with 4-methylthiobenzal-
dehyde (1:1 mol equiv) and piperidine (4−5 drops), and the reaction
was monitored by TLC (3:7, EtOAc:n-hexane). The mixture was
diluted with water and extracted into EtOAc. This was washed with
(α) 4.37 (dd, 11.4, 4.1)
β) 4.20 (dd, 11.4, 7.2)
2.79 m
69.9, CH2
(
3
4
49.3, CH
192.0, C
110.0, C
139.5, C
137.2, C
140.6, C
134.1, C
143.3, C
32.2, CH2
water (30 mL) and brine (30 mL), before drying over MgSO . The
4
resulting product was purified by flash column chromatography (3:7,
EtOAc:n-hexane) to give (E)-5,7-dimethoxy-3-(4′-methylthiobenzyli-
dene)-4-chromanone (42) as a yellow gum (61 mg, 3.0% overall
yield).
4
5
6
7
8
8
9
a
a
UV (MeOH) λmax(log ε) 346 (−3.78); IR (ATR) ν 1659, 1583,
max
−
1
1
1
495, 1463, 1343, 1198, 1093, 830 cm ; H NMR (CDCl , 400
3
1
3
MHz) Table 4; C NMR (CDCl , 100 MHz) Table 5; HRESIMS m/
3
+
+
(a) 3.17 (dd, 14.0 4.4)
z 343.0989 [M + H] (calcd for [C H O SH + H] , 343.0999).
19 18 4
(
b) 2.70 (dd, 14.0, 10.6)
Preparation of (E)-7,8-Dimethoxy-3-(4′-methylthiobenzylidene)-
-chromanone (43). Under inert conditions, NaH (60% in mineral
4
1
2
3
4
5
6
5
6
7
8
4
′
′
′
′
′
′
130.5 C
130.6 CH
115.8 CH
154.6 C
115.8 CH
130.6 CH
62.3
oil, 13 mmol) was added to 2,3-dimethoxyphenol (1 g, 6 mmol) in
dry DMF (10 mL) at 10 °C. After stirring for 60 min, 3-
bromopropionic acid (1.1 g, 7.2 mmol) was added and the reaction
was further stirred at room temperature for 60 min. The mixture was
acidified with HCl to pH 3−5, extracted into EtOAc, and washed with
brine (1 × 30 mL) and water (1 × 30 mL). The product (3-(2,3-
dimethoxyphenoxy)-propanoic acid) was used without further
purification.
Polyphosphoric acid (10 g) was added to 3-(2,3-dimethoxyphe-
noxy)-propanoic acid (from above), and the resulting mixture was
heated at 80 °C for 2 h. The mixture was poured into ice water and
extracted into diethyl ether. The extract was washed with 3 N NaOH
7.11 (d, 8.1)
6.79 (d, 8.1)
6.79 (d, 8.1)
7.11 (d, 8.1)
3.88 s
5.59 br s
4.11 s
-OCH3
-OH
-OCH3
-OH
61.3
5.39 br s
4.74 s
′-OH
(
30 mL) and water (50 mL) and dried over magnesium sulfate. The
yielded, after further purification, 37 (1.2 mg), 38 (6.4 mg), and 39
1.2 mg). Compounds were identified using 2D NMR and CD
methods and by comparison against literature data as referenced
earlier.
product (2,3-dimethoxychromone) was heated at 80 °C with 4-
methylthiobenzaldehyde (1:1 mol equiv) and piperidine (4−5 drops),
and the reaction was monitored by TLC (3:7, EtOAc:n-hexane). The
mixture was diluted with water, extracted into EtOAc, and washed
with water (30 mL) and brine (30 mL), before drying over MgSO4.
The product was purified by flash column chromatography (3:7,
EtOAc:n-hexane) to give (E)-7,8-dimethoxy-3-(4′-methylthiobenzyli-
dene)-4-chromanone (43) as a yellow gum (72 mg, 3.5% overall
yield).
(
Bulbs of E. bicolor (5 kg) were chopped and extracted EtOH (3 L)
for 48 h. The EtOH extract was partitioned with CH Cl (300 mL),
2
2
yielding extract 3 (12.7 g) after evaporation of the CH Cl and the
2
2
remaining EtOH extract, extract 4 (12.0 g). The bulbs were re-
extracted in 1-butanol (2.5 L) for 24 h, yielding extract 5 (14.8 g)
after solvent evaporation. Extract 3 (12.7 g) was coated onto silica,
loaded into a column, and separated using flash column
UV (MeOH) λmax(log ε) 343 (−4.02); IR (NaCl) ν 2921, 1665,
max
−1
1
1
580, 1572, 1493, 1425, 1306, 1264, 1159, 820 cm ; H NMR
−
3
chromatography using a gradient solvent elution method (20 cm ,
13
(CDCl , 400 MHz) Table 4; C NMR (CDCl , 100 MHz) Table 5;
−
3
3
3
2
82 fractions) using solvent mixtures of 100% hexane (20 cm , 50
+
+
HRESIMS m/z 343.0995 [M + H] (calcd for [C H O SH + H] ,
−
3
19 18
4
fractions), 50:50 hexane:CH Cl (20 cm , 50 fractions), 100%
2
2
3
43.0999).
Preparation of (E)-5,6,7-Trimethoxy-3-(3′-hydroxy-4′-methoxy-
benzylidene)-4-chromanone (44). Under inert conditions, NaH
−
3
−3
CH Cl (20 cm , 50 fractions), 50:50 CH Cl :EtOAc (20 cm , 100
2
2
2
2
−
3
fractions), and 50:50 CH Cl :MeOH (20 cm , 32 fractions).
2
2
Fractions 160−170 yielded 40 (193.8 mg), and fractions 224−225
yielded 41 (321.9 mg). Extract 4 (12.0 g) was separated using gravity
column chromatography over silica gel (Merck 9385) and a solvent
(
(
60% in mineral oil, 13 mmol) was added to 3,4,5-trimethoxyphenol
1 g, 5.4 mmol) in dry DMF (10 mL) at 10 °C. After stirring for 60
min, 3-bromopropionic acid (1.1 g, 7.2 mmol) was added and the
reaction was further stirred at room temperature for 60 min. The
mixture was diluted with methanol (20 mL) acidified with HCl to pH
−3
mixture of 80:20 CH Cl :EtOAc (100 cm , 9 fractions). Fraction 2
2
2
yielded 40 (60.5 mg), and fractions 4−6 yielded 41 (322.7 mg).
Extract 5 (14.8 g) was separated using gravity column chromatog-
raphy over Sephadex using a solvent mixture of 50:50 CH Cl :MeOH
3
−5, extracted into EtOAc, and washed with brine (1 × 30 mL) and
2
2
−
3
water (1 × 30 mL). The product (3-(3,4,5-trimethoxyphenoxy)-
propanoic acid) was used without further purification.
Polyphosphoric acid (10 g) was added to 3-(3,4,5-trimethoxy-
phenoxy)-propanoic acid (from above), and the resulting mixture was
heated at 80 °C for 2 h. The mixture was poured into ice water and
extracted into diethyl ether. The extract was washed with 3 M NaOH
(
150 cm , 6 fractions). Fractions 5 and 6 yielded, after further
purification, 41 (186.7 mg) and 32 (28.2 mg). Compounds were
identified using 2D NMR and CD methods and by comparison
against literature data as referenced earlier.
Synthesis of Compounds 42−50. Preparation of (E)-5,7-
Dimethoxy-3-(4′-methylthiobenzylidene)-4-chromanone (42).
Under inert conditions, NaH (60% in mineral oil, 13 mmol) was
added to 3,5-dimethoxyphenol (1 g, 6 mmol) in dry DMF (10 mL) at
(30 mL) and water (50 mL) and dried over MgSO . The product
4
(3,4,5-trimethoxychromone) was heated at 80 °C with 3-hydroxy-4-
methoxybenzaldehyde (1:1 mol equiv) and piperidine (4−5 drops),
and the reaction was monitored by TLC (3:7, EtOAc:n-hexane). The
mixture was diluted with water, extracted into EtOAc, and washed
1
0 °C. After stirring for 60 min, 3-bromopropionic acid (1.1 g, 7.2
mmol) was added and the reaction was further stirred at room
temperature for 60 min. The mixture was acidified with HCl to pH
3
−5, extracted into EtOAc, and washed with brine (1 × 30 mL) and
with water (30 mL) and brine (30 mL), before drying over MgSO .
4
water (1 × 30 mL). The product (3-(3,5-dimethoxyphenoxy)-
propanoic acid) was used without further purification.
The resultant product was purified by flash column chromatography
(3:7, EtOAc:n-hexane) to give (E)-5,6,7-trimethoxy-3-(3′-hydroxy-4′-
methoxybenzylidene)-4-chromanone (44) as a yellow gum (0.1053 g,
5.2% yield).
Polyphosphoric acid (10 g) was added to 3-(3,5-dimethoxyphe-
noxy)-propanoic acid, and the resulting mixture was heated at 80 °C
I
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
J
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Table 5. 13C NMR (100 MHz) Data for Compounds 42−46 and 48−50 (in CDCl3)
position
42
43
44
45
46
48
49
50
2
67.6, CH2
131.4, C
179.4, C
162.8, C
164.5, C
93.6, CH
107.3,C
93.5, CH
165.7, C
136.3, CH
131.3, C
131.3, CH
128.5, CH
140.5, C
128.5, CH
131.3, CH
57.0, CH3
56.4, CH3
68.2, CH2
130.3, C
181.2, C
67.8, CH2
130.2, C
179.8, C
154.8, C
159.3, C
147.9, C
137.9, C
96.3, CH
159.5, C
136.5, CH
128.1, C
123.3, CH
110.8, CH
147.9, C
145.8, C
116., CH
56.1, CH3
56.2, CH3
61.4, CH3
61.7, CH3
69.1, CH2
48.3, CH
191.4, C
154.4, C
159.5, C
154.7, C
137.8, C
96.0, CH
159.2, C
32.2, CH2
131.6, C
120.5, CH
110.8, CH
145.6, C
145.3, C
116.2, CH
56.0, CH3
56.1, CH3
61.3, CH3
61.6, CH3
69.6, CH2
33.2, CH
37.5, CH2
107.2, C
132.7, C
151.6, C
135.8, C
96.2, CH
152.2, C
25.5, CH2
132.7, C
120.4, CH
110.8, CH
152.3, C
67.4, CH2
127.8, C
185.8, C
155.9, C
160.8, C
130.7, C
156.6, C
91.9, CH
158.1, C
138.1, CH
127.6, C
123.3, CH
110.7, CH
145.9, C
145.6, C
115.7, CH
56.1, CH3
56.2, CH3
60.9, CH3
67.2, CH2
129.9, C
179.7, C
159.7, C
154.9, C
153.3, C
138.3, C
96.5, CH
159.9, C
129.8, CH
133.4, C
130.4, CH
118.1, CH
125.3, C
125.4, C
129.2, CH
56.1, CH3
61.2, CH3
61.6, CH3
118.4, CH
144.2, CH
192.3, C
163.2, C
162.5, C
114.0, C
162.9, C
102.5, CH
128.9, CH
3
4
4
5
6
7
8
8
9
1
2
3
4
5
6
a
158.4, C
124.0, CH
106.1, CH
117.3, C
136.6, C
154.7, C
136.3, CH
129.1, C
131.3, CH
128.5, CH
140.5, C
128.5, CH
131.3, CH
56.2, CH3
61.2, CH3
a
′
′
′
′
′
′
127.5, C
123.7, CH
111.3, CH
148.6, C
145.5, C
152.1, C
115.4, CH
55.8, CH3
56.0, CH3
60.6, CH3
61.0, CH3
109.9, CH
56.1, CH3
56.1, CH3
OCH3
OCH3
OCH3
OCH3
SCH3
15.3, CH3
15.2, CH3
Ph-CH3
7.7, CH3
64.0, CH2
13.6, CH3
CH CH3
2
CH CH3
2
1
UV (MeOH) λmax(log ε) 360 (−3.77), 348 (−3.76), IR ν
(MeOH) λmax (log ε) 278 (−3.02); H NMR (CDCl , 400 MHz)
max
3
−1
1
+
(
4
NaCl) 3650−3590, 3376, 2936, 2841, 1596 cm ; H NMR (CDCl ,
Table 4; HRESIMS m/z 397.1254 [M + Na] (calcd for [C H O +
3
20 22
7
1
3
+
00 MHz) Table 4; C NMR (CDCl , 100 MHz) Table 5;
Na] , 397.1263).
3
+
+
HRESIMS m/z 395.1096 [M + Na] (calcd for [C H O + Na] ,
Preparation of (E)-5-Hydroxy-6,7-trimethoxy-3-(3′-hydroxy-4′-
2
0
20
7
3
95.1107).
methoxybenzylidene)-4-chromanone (48). Compound 44 (7.9
Preparation of 5,6,7-Trimethoxy-3(R)-(3′hydroxy-4′-methoxy-
mg, 0.021 mmol) was dissolved in dry CH Cl (2 mL) and cooled
2 2
benzyl)-4-chromanone (45). Compound 44 (34.6 mg, 0.093
to 0 °C. TMSCl (8 equiv) and NaI (8 equiv) were added, and the
mixture was heated to 60 °C for 2 h. The solvent was removed under
reduced pressure and the product purified using column chromatog-
raphy over silica gel (2:8, EtOAc:CH Cl ) to give 5-hydroxy-6,7-
mmol) was dissolved in deuterated toluene (1 mL). Iridium catalyst
1
, (S)-(−)-2-[2-(diphenylphosphino)phenyl]-4-isopropyl-2-oxazo-
lium-(1,5)-cyclooctadiene iridium(I) tetrakis(3,5-bis-
trifluoromethyl)phenyl borate (Scheme 2), was added (1%), and
2
2
(
trimethoxy-3-(3′hydroxy-4′-methoxybenzylidene)-4-chromanone
the reactor was pressurized to 5 bar with hydrogen. The reaction was
stirred overnight at room temperature. The mixture was filtered to
remove the catalyst, and the NMR data were obtained directly in
deuterated toluene (quantitative yield).
(48) (4.5 mg, 60% yield).
Yellow gum; UV (MeOH) λmax (log ε) 368 (−4.03); IR (NaCl)
max 3525, 2935, 2848, 1643, 1579, 1511, 1455, 1359, 1272, 1202,
1106, 813, 736 cm ; H NMR (CDCl
NMR (CDCl , 100 MHz) Table 5; HRESIMS m/z 381.0946 [M +
ν
−
1
1
13
, 400 MHz) Table 4;
3
C
2
3
Yellow gum; [α] −43.8 (c 0.9, CH OH); ECD (c 0.03, ACN)
3
D
3
+
+
3
(
1
26 nm (+1.3), 284 nm (−1.9); UV (MeOH) λmax (log ε) 277
Na] (calcd for [C19H O + Na] , 381.0950).
18 7
−4.20), 277 (−3.66); IR (NaCl) ν 2936, 1674, 1580, 1567, 1485,
Preparation of (E)-5,6,7-Trimethoxy-3-(4′-chloro-3′-hydroxyben-
zylidene)-4-chromanone (49). NaH (60% in mineral oil, 13 mmol)
was added to 3,4,5-trimethoxyphenol (1 g, 5.4 mmol) in dry DMF
max
−
1
1
454, 1412, 1200, 1179, 1116, 822 cm ; H NMR (CDCl , 400
3
13
MHz) Table 4; C NMR (CDCl , 100 MHz) Table 5; HRESIMS m/
z 397.1259 [M + Na] (calcd for [C H O + Na] , 397.1263).
3
+
+
(
10 mL) at 10 °C under inert conditions. After stirring for 60 min, 3-
2
0
22
7
bromopropionic acid (1.1 g, 7.2 mmol) was added and the reaction
was stirred at room temperature for a further 60 min. The mixture was
acidified with HCl to pH 3−5, extracted into EtOAc, and washed with
brine (1 × 30 mL) and water (1 × 30 mL). The product (3-(3,4,5-
trimethoxyphenoxy)-propanoic acid) was used without further
purification.
Polyphosphoric acid (10 g) was added to 3-(3,4,5-trimethox-
yphenoxy)-propanoic acid (from above), and the resulting mixture
was heated at 80 °C for 2 h. The mixture was poured into ice water
and extracted with diethyl ether. The extract was washed with 3 M
NaOH (30 mL) and water (50 mL) and dried over magnesium
sulfate. The product (3,4,5-trimethoxychromone) was heated at 80 °C
with 4-chloro-3-hydroxybenzaldehyde (1:1 mol equiv) and piperidine
(4−5 drops), and the reaction was monitored by TLC (3:7, EtOAc:n-
hexane) and was found to be complete after 24 h. The mixture was
diluted with water and extracted into EtOAc, then washed with water
Preparation of 5,6,7-Trimethoxy-3-(3′-hydroxy-4′-methoxyben-
zyl)-chromane (46). Compound 44 (50.0 mg, 0.134 mmol) was
dissolved in methanol (2 mL). Pd/C (10%) was added, the flask was
saturated with H (1 atm), and the mixture was stirred overnight at
2
room temperature. The catalyst was filtered and the solvent removed
under reduced pressure to yield 46 (quantitative yield).
1
Yellow gum; UV (MeOH) λ (log ε) 282 (−3.14); H NMR
max
1
3
(
CDCl , 400 MHz) Table 4; C NMR (CDCl , 100 MHz) Table 5;
3
3
+
+
HRESIMS m/z 383.1462 [M + Na] (calcd for [C H O + Na] ,
2
0
24
6
3
83.1471).
Preparation of 5,6,7-Trimethoxy-3-(3′-hydroxy-4′-methoxyben-
zyl)-4-chromanone (47). Compound 44 (5.9 mg, 0.0158 mmol) was
dissolved in toluene (2 mL). Wilkinson’s catalyst (1%) was added, the
flask was saturated with H (1 atm), and the mixture was stirred
2
overnight at room temperature. The catalyst was filtered and the
solvent removed under reduced pressure. The mixture obtained was
subject to column chromatography over silica gel (8:2
EtOAc:CH Cl ) to yield 47 (2.4 mg, 40% yield). Yellow gum; UV
(30 mL) and brine (30 mL), before drying over MgSO . The product
4
was purified by flash column chromatography (3:7, EtOAc:n-hexane)
2
2
K
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
to give 5,6,7-trimethoxy-3-(4′-chloro-3′- hydroxybenzylidene)-4-
chromanone (49) as a yellow powder (155 mg, 7.8% yield).
UV (MeOH) λmax (log ε) 295 (−3.94), 352 (−3.77); IR (NaCl)
νmax 3328, 2939, 1661, 1650, 1488, 1470, 1415, 1391, 1258, 1201,
Statistical Analysis. The data obtained from tube formation
experiments were analyzed by one-way ANOVA with Dunnett’s post
hoc tests for comparisons between compounds’ treatments and
control. All analyses were performed using GraphPad Prism software
(v. 7.0). A p value of <0.05 was considered statistically significant.
−1
1
13
1
171, 1100, 818 cm ; H NMR (CDCl , 400 MHz) Table 4;
C
3
NMR (CDCl , 100 MHz) Table 5; HRESIMS m/z 377.0786 [M +
3
+
+
H] [calcd for (C H O Cl + H) , 377.0792].
19
17
6
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jnat-
prod.8b00989.
■
Preparation of (E)-3-(3,4-Dimethoxylphenyl)-1-(4-ethoxy-2-hy-
*
S
droxy-3-methyl-phenyl)prop-2-en-1-one (50). 4-Ethoxy-2-hydroxy-
3
-methyl acetophenone (0.1 g, 0.51 mmol) was added to 3,4-
dimethoxybenzaldehyde (0.85 g, 0.51 mmol) and 10% NaOH (1 mL)
in methanol (5 mL) and stirred at room temperature for 2 h. Upon
standing, (E)-3-(3,4-dimethoxylphenyl)-1-(4-ethoxy-2-hydroxy-3-
methyl-phenyl)prop-2-en-1-one (50) precipitated out as a yellow
powder and was isolated by vacuum filtration (25.9 mg, 15% yield).
Flowcharts showing how the compounds were isolated
and NMR spectra of isolated and synthetic compounds
(PDF)
UV (MeOH) λmax (log ε) 368 (−4.62); IR (ATR) ν 1650, 1570,
max
−
1
1
1
(
550, 1499, 1419, 1364, 1265, 1221, 1080, 847 cm ; H NMR
1
3
CDCl , 400 MHz) Table 4; C NMR (CDCl , 100 MHz) Table 5;
3 3
HRESIMS m/z 343.1545 [M + H] [calcd for (C H O + H) ,
+
+
AUTHOR INFORMATION
Corresponding Authors
Phone: +44 (0)1483301080. E-mail: d.mulholland@surrey.
ac.uk.
*Phone: +1 (317) 274-3305. E-mail: tcorson@iu.edu.
ORCID
Sianne Schwikkard: 0000-0002-7200-927X
Dulcie Mulholland: 0000-0001-7778-5617
Timothy W. Corson: 0000-0002-1402-7875
■
*
2
0
22
5
3
43.1545].
BIOLOGICAL METHODS
Materials. Human retinal microvascular endothelial cells (HRECs;
Cell Systems, Seattle, WA) and human umbilical vein endothelial cells
■
(
HUVECs; Lonza, Walkersville, MD) were used between passage 5
and 8. Endothelial growth medium (EGM-2) was prepared by mixing
the contents of an EGM-2 “Bullet Kit” (Cat no. CC-4147) with
endothelial basal medium (EBM) (Cat no. CC-3156) (Lonza). 92-1
uveal melanoma cells (a kind gift of Dr. Martine Jager, University of
Leiden) were grown in RPMI medium containing 10% FBS and 1%
penicillin−streptomycin (pen−strep). Y-79 retinoblastoma cells (a
kind gift of Dr. Brenda L. Gallie, Ontario Cancer Institute) were
grown in RB medium (IMDM + 10% FBS + 55 μM β-
mercaptoethanol + 10 μg/mL insulin + 1% pen−strep). ARPE-19
retinal pigment epithelial cells (ATCC, Manassas, VA) were grown in
DMEM containing 10% FBS and 1% pen−strep. AlamarBlue reagent
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
H.W. acknowledges a Ph.D. scholarship from the University of
Surrey. S.S. acknowledges support from the Daphne Jackson
Trust, the University of Surrey, and the Royal Society of
Chemistry for funding. T.W.C. acknowledges support from
NIH/NEI R01EY025641, the International Retinal Research
Foundation, and an unrestricted grant from Research to
Prevent Blindness, Inc. S.S. would also like to acknowledge and
thank the EPSRC UK National Mass Spectrometry Facility,
Swansea, for HRMS analysis.
(
product code BUF012B) was from AbD Serotec (Raleigh, NC).
Matrigel matrix basement membrane (Cat no. 354234) was
purchased from Corning (Corning, NY).
Cell Proliferation Assay. The antiproliferative activities of
different compounds were monitored by an alamarBlue based
20
fluorescence assay as described previously. Five cell types were
used: the target endothelial cell type (HRECs), a nontarget
endothelial cell type (HUVECs), a nontarget normal retinal pigment
epithelial cell line (ARPE-19), a nontarget uveal melanoma cell line
REFERENCES
■
(
92-1), and a nontarget retinoblastoma cell line (Y79). Briefly, 2500
cells in 100 μL of growth medium were incubated in 96-well clear
(
1) Penn, J. S.; Madan, A.; Caldwell, R. B.; Bartoli, M.; Caldwell, R.
W.; Hartnett, M. Prog. Retinal Eye Res. 2008, 27, 331−371.
bottom black plates for 24 h at 37 °C, 5% CO ; different
2
(
(
(
(
2
(
2) Zhang, S. X.; Ma, J. X. Prog. Retinal Eye Res. 2007, 26, 1−37.
3) Folk, J. C.; Stone, E. M. N. Engl. J. Med. 2010, 363, 1648−1655.
4) Stewart, M. W. Mayo Clin. Proc. 2012, 87, 77−88.
concentrations of each test compound were then added (range: 0.5
nM to 500 μM) followed by 44 h of incubation. A 11.1 μL portion of
alamarBlue reagent was then added, and after 4 h, fluorescence
readings were taken on a Synergy H1 plate reader (Biotek, Winooski,
VT) with excitation and emission wavelengths of 560 and 590 nm,
respectively. GraphPad Prism software (v. 7.0) was used for data
analysis. Compounds that reduced the cell number by 50% or more at
the highest concentration tested (relative to DMSO control) were
reported as having a value of GI50 < 100 μM.
5) Ghasemi Falavarjani, K. G.; Nguyen, Q. D. Eye (London, U. K.)
013, 27, 787−794.
6) Lux, A.; Llacer, H.; Heussen, F. M.; Joussen, A. M. Br. J.
Ophthalmol. 2007, 91, 1318−1322.
(7) Sulaiman, R. S.; Basavarajappa, H. D.; Corson, T. W. Exp. Eye
Res. 2014, 129, 161−171.
In Vitro Tube Formation Assay. A Matrigel based tube
formation assay was performed to monitor the ability of HRECs to
form tube-like structures in the presence of different concentrations of
(8) Lin, L. G.; Liu, Q. Y.; Ye, Y. Planta Med. 2014, 80, 1053−1066.
(9) Mulholland, D. A.; Schwikkard, S. L.; Crouch, N. R. Nat. Prod.
Rep. 2013, 30, 1153−1266.
20
1
1
3, 27, 28, 45, 46, 48, and 50, as described previously. Briefly,
(10) Pohl, T. S.; Crouch, N. R.; Mulholland, D. A. Curr. Org. Chem.
2000, 4, 1287−1324.
5,000 cells in 100 μL of EGM-2 medium were incubated in the
absence (DMSO treated) and presence of different concentrations of
each compound in 96-well clear plates coated with 50 μL of Matrigel
basement membrane. Known antiangiogenic homoisoflavonoid SH-
(11) Moodley, N.; Crouch, N. R.; Mulholland, D. A.; Slade, D.;
Ferreira, D. S. Afr. J. Bot. 2006, 72, 517−520.
(12) Sparg, S. G.; van Staden, J.; Jager, A. K. J. Ethnopharmacol.
2002, 80, 95−101.
21
1
1037 was a positive control. After 8 h, the images were recorded
using an EVOS FL microscope (AMG, Mill Creek, WA) and the tube
length was measured using Angiogenesis Analyzer macros in ImageJ
(13) Crouch, N. R.; Bangani, V.; Mulholland, D. A. Phytochemistry
1999, 51, 943−946.
(
http://image.bio.methods.free.fr/ImageJ/?Angiogenesis-Analyzer-
(14) Shim, J. S.; Kim, J. H.; Lee, J.; Kim, S. N.; Kwon, H. J. Planta
Med. 2004, 70, 171−173.
for-ImageJ).
L
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
(
15) Kim, J. H.; Kim, K. H.; Yu, Y. S.; Kim, Y. M.; Kim, K. W.;
Kwon, H. J. Biochem. Biophys. Res. Commun. 2007, 362, 848−852.
16) Kim, J. H.; Yu, Y. S.; Jun, H. O.; Kwon, H. J.; Park, K. H.; Kim,
K. W. Mol. Vis. 2008, 14, 556−561.
17) Adinolfi, M.; Corsaro, M. M.; Lanzetta, R.; Laonigro, G.;
Mangoni, L.; Parrilli, M. Phytochemistry 1986, 26, 285−290.
18) Lee, B.; Basavarajappa, H. D.; Sulaiman, R. S.; Fei, X.; Seo, S.
Y.; Corson, T. W. Org. Biomol. Chem. 2014, 12, 7673−7677.
19) Basavarajappa, H. D.; Lee, B.; Fei, X.; Lim, D.; Callaghan, B.;
(45) Sulaiman, R. S.; Park, B.; Sheik, P. B. S.; Si, Y.; Kharwadkar, R.;
Mitter, S. K.; Lee, B.; Sun, W.; Qi, X.; Boulton, M. E.; Meroueh, S. O.;
Fei, X.; Seo, S. Y.; Corson, T. W. ACS Chem. Biol. 2018, 13, 45−52.
(46) Liao, L. X.; Song, X. M.; Wang, L. C.; Lv, H. N.; Chen, J. F.;
Liu, D.; Fu, G.; Zhao, M. B.; Jiang, Y.; Zeng, K. W.; Tu, P. F. Proc.
Natl. Acad. Sci. U. S. A. 2017, 114, E5986−E5994.
(
(
(
(
Mund, J. A.; Case, J.; Rajashekhar, G.; Seo, S. Y.; Corson, T. W. PLoS
One 2014, 9, No. e95694.
(20) Basavarajappa, H. D.; Lee, B.; Lee, H.; Sulaiman, R. S.; An, H.;
Magana, C.; Shadmand, M.; Vayl, A.; Rajashekhar, G.; Kim, E. Y.;
Suh, Y. G.; Lee, K.; Seo, S. Y.; Corson, T. W. J. Med. Chem. 2015, 58,
5
(
015−5027.
21) Sulaiman, R. S.; Merrigan, S.; Quigley, J.; Qi, X.; Lee, B.;
Boulton, M. E.; Kennedy, B.; Seo, S. Y.; Corson, T. W. Sci. Rep. 2016,
, 25509.
22) Schwikkard, S. L.; Whitmore, H.; Corson, T. W.; Sishtla, K.;
Langat, M. K.; Carew, M.; Mulholland, D. A. Planta Med. 2018, 84,
38.
23) Waller, C. P.; Thumser, A. E.; Langat, M. K.; Crouch, N. R.;
Mulholland, D. A. Phytochemistry 2013, 95, 284−290.
24) Schwikkard, S.; Alqahtani, A.; Knirsch, W.; Wetschnig, W.;
6
(
6
(
(
Jaksevicius, A.; Opara, E. I.; Langat, M. K.; Andriantiana, J. L.;
Mulholland, D. A. J. Nat. Prod. 2017, 80, 30−37.
(25) Corsaro, M. M.; Lanzetta, R.; Mancino, A.; Parrilli, M.
Phytochemistry 1992, 31, 1395−1397.
26) Boehler, P.; Tamm, C. H. Tetrahedron Lett. 1967, 8, 3479−
483.
(
3
(
(
27) Tamm, C. H. Arzneim. Forsch. 1972, 22, 1776−1784.
28) Sidwell, W. T. L.; Tamm, C. H. Tetrahedron Lett. 1970, 11,
4
(
1
(
75−478.
29) Farkas, L.; Gottsegen, A.; Nogradi, M.; Strelisky, J. Tetrahedron
971, 27, 5049−5056.
30) Sihra, J. K.; Mulholland, D. A.; Langat, M. K.; Crouch, N. R.;
Nuzillard, J.-M. RSC Adv. 2017, 7, 15416−15427.
(
(
31) Finckh, R. E.; Tamm, C. H. Experientia 1970, 26, 472−473.
32) Koorbanally, N. A.; Crouch, N. R.; Harilal, A.; Pillay, B.;
Mulholland, D. A. Biochem. Syst. Ecol. 2006, 34, 114−118.
33) Hafez-Ghoran, S.; Ebrahimi, P.; Mighani, H.; Saeidnia, S. Braz.
J. Pharm. Sci. 2015, 51, 949−955.
34) Alali, F.; El-Elimat, T.; Albataineh, H.; Al-Balas, Q.; Al-
(
(
Gharaibeh, M.; Falkinham, J.; Chen, W.; Swanson, S.; Oberlies, N. J.
Nat. Prod. 2015, 78, 1708−15.
(35) Adinolfi, M.; Barone, G.; Belardini, M.; Lanzetta, R.; Laonigro,
G.; Parrilli, M. Phytochemistry 1984, 23, 2091−2093.
36) Shaikh, M. M.; Kruger, H. G.; Bodenstein, J.; Smith, P.; du
Toit, K. Nat. Prod. Res. 2012, 26, 1473−1482.
37) Lu, S.-M.; Bolm, C. Angew. Chem., Int. Ed. 2008, 47, 8920−
923.
38) Sharma, P.; Kumar, S.; Ali, F.; Anthal, S.; Gupta, V. K.; Khan, I.
(
(
8
(
A.; Singh, S.; Sangwan, P. L.; Suri, K. A.; Gupta, D. K.; Dutt, P.;
Vishwakarma, R. A.; Satti, N. Med. Chem. Res. 2013, 22, 3969−3983.
(39) Corson, T. W.; Basavarajappa, H. D.; Seo, S.-Y.; Lee, B.; Fei, X.
Indiana University Research and Technology Corporation; Gachon
University of Industry Academic Cooperation Foundation; US
1
(
(
0,238,627, 2019.
40) Nugroho, A. E.; Morita, H. J. Nat. Med. 2014, 68, 1−10.
41) Aziz, A. N.; Ismail, N. H.; Halim, S. N. A.; Looi, C. Y.; Anouar,
E. H.; Langat, M. K.; Mulholland, D.; Awang, K.; Laevifins, A.-G.
Phytochemistry 2018, 156, 193−200.
(42) Adinolfi, M.; Barone, G.; Corsaro, M. M.; Mangoni, L.;
Lanzetta, R.; Parrilli, M. Tetrahedron 1988, 44, 4981−4988.
(43) Arnaoutova, I.; Kleinman, H. K. Nat. Protoc. 2010, 5, 628−635.
(44) Basavarajappa, H. D.; Sulaiman, R. S.; Qi, X.; Shetty, T.; Sheik,
P. B. S.; Sishtla, K. L.; Lee, B.; Quigley, J.; Alkhairy, S.; Briggs, C. M.;
Gupta, K.; Tang, B.; Shadmand, M.; Grant, M. B.; Boulton, M. E.;
Seo, S. Y.; Corson, T. W. EMBO Mol. Med. 2017, 9, 786−801.
M
DOI: 10.1021/acs.jnatprod.8b00989
J. Nat. Prod. XXXX, XXX, XXX−XXX