Curcumin analogues inhibit phosphodiesterase-5 and dilate rat pulmonary arteries

Article abstract of DOI:10.1111/jphp.12302

Objectives Phosphodiesterase (PDE)-5 inhibitors are useful as vasodilators for the treatment of pulmonary arterial hypertension. We aimed to study curcumin analogues for PDE5 inhibitory activity and vasorelaxation of rat pulmonary arteries. Methods Three natural curcuminoids (1-3) and six synthetic analogues (4-9) were tested for PDE5 and PDE6 inhibitory activities using enzymatic radioassay. Their vasorelaxation was measured using freshly isolated segments of rat pulmonary artery and aorta. Key findings Curcuminoids (1-3) mildly inhibited PDE5 (half maximal inhibitory concentration (IC₅₀) = 18 μm): the metamethoxyl of curcumin was important for PDE5 inhibition. But hydroxyl rearrangements, removing both methoxyls and one ketomethylene, yielded the potent 7 and 9 (IC₅₀ = 4 μm) (compared with sildenafil, IC₅₀ = 0.03 μm). Only 1, 3 and 4 were PDE5 selective over PDE6. Triazole-carboxylic addition provided water-solubility while preserving potency. All analogues possessed concentration-dependent vasorelaxant activity on pulmonary arteries (40% of maximal effective concentration (EC₄₀) = 29-90 μm, maximum response = 60-90% at 300 μm), while compounds (1-8) were weakly acting in aorta (maximum response <40%). Only demethoxycurcumin (2) and analogues 5, 8, 9 had endothelium-dependent actions. Sildenafil was highly potent (EC₄₀ = 0.04 μm) and highly endothelium dependent in pulmonary artery but weak on intact aorta (EC₄₀ = 1.8 μm). Activity profiles suggest actions through additional cell pathways for promoting vasorelaxation. Conclusions Curcumin analogues are potential leads for developing efficacious and selective PDE5 inhibitors and other pathologies of pulmonary hypertension.

Full text of DOI:10.1111/jphp.12302

bs_bs_banner
Research Paper
Journal of Pharmacy
And Pharmacology
Curcumin analogues inhibit phosphodiesterase-5 and dilate
rat pulmonary arteries
Oraya Kruangtipa, Krongkarn Chootipb, Prapapan Temkitthawona, Kanokwan Changwichita,
Thipphawan Chuprajobc, Chatchawan Changtamd, Apichart Suksamrarnc, Nantaka Khoranaa,
C. Norman Scholfielda and Kornkanok Ingkaninana
aDepartment of Pharmaceutical Chemistry and Pharmacognosy, Faculty of Pharmaceutical Sciences and Center of Excellence for Innovation in
b
c
Chemistry, Department of Physiology, Faculty of Medical Sciences, Naresuan University, Phitsanulok, Department of Chemistry and Center of
Excellence for Innovation in Chemistry, Faculty of Science, Ramkhamhaeng University, Bangkok, dDivision of Physical Sciences, Faculty of Science
and Technology, Huachiew Chalermprakiet University, Samutprakarn, Thailand
Keywords
curcumin analogues; phosphodiesterase-5;
Abstract
Objectives Phosphodiesterase (PDE)-5 inhibitors are useful as vasodilators for
the treatment of pulmonary arterial hypertension. We aimed to study curcumin
analogues for PDE5 inhibitory activity and vasorelaxation of rat pulmonary
arteries.
Methods Three natural curcuminoids (1–3) and six synthetic analogues (4–9)
were tested for PDE5 and PDE6 inhibitory activities using enzymatic radioassay.
Their vasorelaxation was measured using freshly isolated segments of rat pulmo-
nary artery and aorta.
pulmonary artery; pulmonary artery
hypertension
Correspondence
Kornkanok Ingkaninan, Faculty of
Pharmaceutical Sciences and Center of
Excellence for Innovation in Chemistry,
Naresuan University, Phitsanulok 65000,
Thailand.
E-mail: k_ingkaninan@yahoo.com
Krongkarn Chootip, Department of
Physiology, Faculty of Medical Sciences,
Naresuan University, Phitsanulok 65000,
Thailand.
Key findings Curcuminoids (1–3) mildly inhibited PDE5 (half maximal inhibi-
tory concentration (IC50) = 18 ꢀm): the metamethoxyl of curcumin was important
for PDE5 inhibition. But hydroxyl rearrangements, removing both methoxyls and
one ketomethylene, yielded the potent 7 and 9 (IC50 = 4 ꢀm) (compared with
sildenafil, IC50 = 0.03 ꢀm). Only 1, 3 and 4 were PDE5 selective over PDE6.
Triazole-carboxylic addition provided water-solubility while preserving potency.
All analogues possessed concentration-dependent vasorelaxant activity on pulmo-
nary arteries (40% of maximal effective concentration (EC40) = 29–90 ꢀm,
maximum response = 60–90% at 300 ꢀm), while compounds (1–8) were weakly
acting in aorta (maximum response <40%). Only demethoxycurcumin (2) and
analogues 5, 8, 9 had endothelium-dependent actions. Sildenafil was highly potent
(EC40 = 0.04 ꢀm) and highly endothelium dependent in pulmonary artery but
weak on intact aorta (EC40 = 1.8 ꢀm). Activity profiles suggest actions through
additional cell pathways for promoting vasorelaxation.
E-mail: krongkarnc@gmail.com
Abbreviations
Ach, acetylcholine; BSA, bovine serum
albumin; cGMP, guanosine cyclic
monophosphate; DMSO, dimethylsulphoxide;
EC40, 40% of maximal effective concentra-
tion; ESMS, electrospray mass spectrometry;
IC50, half maximal inhibitory concentration;
NMR, nuclear magnetic resonance; PAH, pul-
monary artery hypertension; PDE,
phosphodiesterase; PE, phenylephrine; QAE,
quaternary aminoethyl anion exchanger; sGC,
soluble guanylyl cyclase.
Conclusions Curcumin analogues are potential leads for developing efficacious
and selective PDE5 inhibitors and other pathologies of pulmonary hypertension.
Received April 21, 2014
Accepted July 4, 2014
doi: 10.1111/jphp.12302
Introduction
Idiopathic pulmonary arterial hypertension (PAH) is a rela-
tively rare lung disorder (prevalence ∼6 cases/million
people[1]) with a poor prognosis. Pulmonary arteries
become constricted, thus reducing blood flow to the lungs
and increasing pulmonary arterial pressure. This PAH
increases the load on the right ventricle, leading to right
heart failure and death.[2] Several mechanisms for the
disease have been proposed and shown to produce PAH in
animal models, but none appears to reflect the human con-
dition. Because of this unclear aetiology, current drug
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–••
1

Curcumins inhibit PDE5
Oraya Kruangtip et al.
treatments have focused on symptomatic treatment in the
form of vasodilators. Of these, sildenafil, tadalafil, vardenafil
and similar drugs have received the most widespread
application. These compounds are inhibitors of
phosphodiesterase-5 (PDE5), which is found in many cells
including the smooth muscle cells of pulmonary vascular
tree expressed as PDE5A.[3] This enzyme is upregulated in
pulmonary hypoxia and PAH[4] making it an attractive drug
target for treatment of these pathologies. The target
substrate for PDE5 is the inactivation of guanosine cyclic
monophosphate (cGMP) via hydrolysis. Cytosolic cGMP
constituatively activates in particular, adenosine
triphosphate and big potassium current channels[5] and the
resulting increased K-permeability maintains vasorelaxant
tone. Thus inhibiting PDE5 favours cGMP accumulation
hence promoting vasodilatation.[6,7]
Sildenafil was initially registered as an oral drug for erec-
tile dysfunction, then later approved for PAH treatment[6]
and is now licensed for Raynaud’s disease.[7] There is a
growing list of successful clinical trials with PDE5 inhibi-
tors[8] that may lead to extended licensing by regulatory
bodies.[7] Clearly, these inhibitors also cause vasodilatation
in other tissues[8] and more especially have retinal-related
complications most commonly associated with additional
PDE6 blockade.[9] Therefore, the hunt continues for drugs
that are specific for diseased targets including the pulmo-
nary arterial circulation and especially erectile dysfunction.
Several Curcuma species are known to be vasorelaxant
including Curcuma longa,[10] while curcumin reduces pul-
monary arterial pressure.[11] We have shown that extracts of
these plants exhibit PDE5 inhibitory activity.[12] This sug-
gests that curcuminoids might provide leads for the devel-
opment of a new generation of selective PDE5 inhibitors.
Therefore, here we aimed to explore the activities of both
natural and synthetic curcumin analogues on the inhibition
of PDE5 in cell-free assay and on the vasorelaxation of
freshly isolated rat pulmonary arteries in vitro.
Sante Animale (Libourne, France). Quaternary aminoethyl
anion exchanger (QAE) resin (QAE Sephadex A-25) was
purchased from GE Healthcare (Upsala, Sweden). The
PDE5 and PDE6 were prepared from rat lung tissues and
chicken retinas, respectively.
Preparations of compounds
The natural compounds, (1) curcumin, (2) demethoxy-
curcumin and (3) bisdemethoxycurcumin, were obtained
from the rhizomes of C. longa.[13] Compound 4 was synthe-
sised by the demethylation reaction of compound 1 as
described previously.[13] Compounds 5–7 were prepared by
aldol condensation of substituted cinnamones (10a and
10b) and substituted cinnamaldehydes (11a, 11b and 11c)
under a base-catalysed condition, which has recently been
reported[14] (Figure 1). Compounds 8 and 9, the acid group-
containing analogues of 7, were synthesised by coupling the
alkyne analogues 12 and 13 with the azide 14 as shown in
Figure 2. Briefly, the cinnamones 10c and 10b were coupled
with cinnamaldehydes 11c and 11d by the same method
used to synthesise compounds 5–7[14] to yield compounds
12 (32%) and 13 (34%). Each of compounds 12 and 13 was
later reacted with 2-azidoacetic acid (14) under click condi-
tions[15] to give compounds 8 and 9 in 69% and 38% yields,
1
respectively. Compound 8: H nuclear magnetic resonance
(NMR) (400 MHz, acetone-d6) δ (ppm): 5.36 (s, 2H), 5.38
(s, 2H), 6.62 (d, J = 15.3 Hz, 1H), 6.82 (d, J = 7.8 Hz, 1H),
7.01–7.08 (m, 5H), 7.21 (t, J = 7.8 Hz, 1H), 7.32 (d,
J = 7.7 Hz, 1H), 7.36 (d, J = 16.1 Hz, 1H), 7.42 (t, J = 7.7 Hz,
1H), 7.51 (ddd, J = 15.3, 6.6, 3.4 Hz, 1H), 7.73 (d,
J = 7.6 Hz, 1H), 7.93 (d, J = 16.1 Hz, 1H), 8.22 (s, 1H). 13C
NMR (100 MHz, acetone-d6) δ (ppm): 51.3, 63.0, 114.0,
114.5, 117.1, 119.7, 122.1, 125.0, 126.3, 126.6, 128.3, 130.1,
130.7, 130.8, 132.4, 138.2, 138.8, 142.0, 143.6, 144.0,
158.6, 158.5, 168.5, 189.2. electrospray mass spectrometry
1
(ESMS) (−ve): m/z 430 [M − H]−. Compound 9: H NMR
(400 MHz, acetone-d6) δ (ppm): 5.26 (s, 2H), 5.34 (s, 2H),
6.74 (d, J = 15.2 Hz, 1H), 6.90 (t, J = 7.4 Hz, 1H), 6.98 (d,
J = 8.1 Hz, 1H), 7.03 (d, J = 8.1 Hz, 1H), 7.09 (d,
J = 15.6 Hz, 1H), 7.17–7.33 (m, 6H), 7.54 (dd, J = 15.1,
10.8 Hz, 1H), 7.69 (d, J = 7.7 Hz, 1H), 8.04 (d, J = 16.1 Hz,
1H), 8.15 (s, 1H). 13C NMR (100 MHz, acetone-d6) δ
(ppm): 51.2, 62.4, 113.8, 116.6, 117.1, 120.8, 121.2, 122.9,
126.0, 126.1, 128.7, 129.4, 130.6, 130.7, 132.4, 138.5, 138.9,
141.5, 143.3, 144.3, 157.7, 159.9, 168.7, 189.2. ESMS (−ve):
m/z 430 [M − H]−. The purity of all the test compounds was
more than 95% as determined by NMR and thin layer chro-
matography. The samples were stored at −20°C until use.
Materials and Methods
Materials
Krebs’ solution (mm): (NaCl 122; KCl 5; (N-(2-
hydroxyethyl)piperazine-N′-(2-ethane-sulfonic acid)) 10;
KH2PO4 0.5; NaH2PO4 0.5; MgCl2 1; glucose 11 and CaCl2
1.8; adjusted to pH 7.3 with 1-N NaOH), acetylcholine
(ACh), phenylephrine (PE), ingredients in buffer
1
(100-mm Tris-HCl, pH 7.5; 100-mm imidazole; 15-mm
MgCl2; 1.0-mg/ml bovine serum albumin (BSA) and 2.5-
mg/ml snake venom from Crotalus atrox), ingredients in
Buffer 2 (100-mm Tris-HCl, pH 7.5 100-mm imidazole,
15-mm MgCl2, 1.0-mg/ml BSA and 0.5-mg/ml histone)
were obtained from Sigma (St Louis, MO, USA). Pentobar-
bital sodium solution (Nembutal) was obtained from Ceva
Determination of phosphodiesterase-5
inhibitory activity
All samples were dissolved in dimethylsulfoxide (DMSO)
and diluted with water giving a final DMSO concentration
2
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–••

Oraya Kruangtip et al.
Curcumins inhibit PDE5
R4
R1
O
O
R5
R6
R2
R3
H
10a: R1 = H, R2 = OMe,R3 = OH
10b: R1 = OH, R2 = R3 = H
10c: R1 = OCH2C CH,R2 = R3 = H
11a: R4 = H, R5 = OMe,R6 = OH
11b: R4 = R5 = H, R6 = OH
11c: R4 = R6 = H, R5 = OH
(a)
(a)
11d: R4 = R6 = H, R5 = OCH2C CH
(b)
R1
2'
R4
2"
O
5
1
7
R2
3'
R5
3"
1'
1"
3
6'
6"
R3
R6
4'
4"
5'
5"
+
→
→
→
→
→
10a 11a
5: R1 = R4 = H, R2 = R5 = OMe, R3 = R6 = OH
6: R1 = R4 = R5 = H, R2 = OMe,R3 = R6 = OH
+
10a 11b
1
5
2
3
4
6
+
10b 11c
7: R = R = OH, R = R = R = R = H
1
+
10c 11c
12: R = OCH22C CH,R2 = R3 = R4 = R6 = H, R5 = OH
1
3
4
6
5
+
10b 11d
13: R = OH, R = R = R = R = H, R = OCH2C CH
Figure 1 Preparation of compounds 5–7, 12 and 13. Reagents and conditions: (a) propargyl bromide, K2CO3, acetone, room temperature.
(b) 20% aq. NaOH, EtOH, 0oC− room temperature.
O
Br
OH
(a)
1”’
3”’
O
2”’
O
O
N
N
5
7
1
2”
O
O
HO
5”’
N
3” OH
1’
6’
1”
6”
4”’
3’
4’
3
O
6
2
4
OH
N3
4”
OH
5’
8
5”
12
14
(b)
O
OH
O
N
N
OH
O
N
5”’OH
4”’
5
1
7
2”
O
1’
3” O
1”
6”
3’
4’
3
2”’
3”’
4
6
2
1”’
13
6’
4”
9
5’
5”
Figure 2 Preparation of compounds 8 and 9. Reagents and conditions: (a) NaN3, H2O, room temperature. (b) CuSO4, sodium ascorbate, THF : H2O
(9 : 1).
of 5%. PDE5 was extracted from rat lung tissue as described
previously.[16] The PDE5 assay was conducted using the two-
step radioactive procedure, which has been modified from
Sonnenburg et al., 1998.[17] Twenty microlitres of the follow-
ing reagents was added to 96-well plates: buffer 1, EGTA,
PDE5 solution and test samples or control (5% DMSO in
buffer). The reaction was started by adding 20 ꢀl of 5-ꢀm
[3H]cGMP (∼50 000 cpm) to the reaction mixture and
incubated at 30°C for 40 min. Then, 100 ꢀl of 50% QAE
resin in water was added to the wells to purify the
hydrolysate. The plate was shaken for 10 min and left for
20 min to allow the resin to sediment. The supernatants
(100 ꢀl) were transferred to new microplate wells contain-
ing 100 ꢀl of fresh 50% QAE resin. The plate was again
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–••
3

Curcumins inhibit PDE5
Oraya Kruangtip et al.
shaken and left to permit sedimentation. Then, 100-ꢀl
supernatant was mixed and shaken with 200 ꢀl of
Microscint (Perkin Elmer, Waltham, MA, USA) 20 for 2 h.
The radioactivity was counted by a TopCount NXT
(PerkinElmer, Boston, MA, USA), each well for 1 min. The
PDE5 activity in the study was standardised to have a
hydrolysis activity of 20–25% of the total substrate counts.
The calculation of hydrolysis is shown in equation (1). The
PDE5 inhibitory activity was calculated using equation (2).
housed under standard environmental conditions at
25 2°C, 12-h light and dark cycle, fed with standard
rodent diet and tap water in the Center for Animal
Research, Naresuan University, Thailand.
Tissue preparation and vascular protocols
Rats were anaesthetised by pentobarbital (65-mg/kg, i.p.
injection) and the lungs and aorta were isolated. Intra-
pulmonary artery was removed from lung and soaked in
Krebs’ solution to wash off the surrounding loose connec-
tive tissue. The vessel was cut into rings 2–3 mm in length
and mounted in tissue chambers via a pair of intraluminal
wires. The chambers contained Krebs’ solution at 37°C and
bubbled with air. The rings were incubated for 45–60 min at
an optimum tension of 1 g during which the solution was
replaced every 15 min. The wires were connected to force
transducers to measure isometric tension via a MacLab A/D
converter (Chart V5, A.D. Instruments, Castle Hill, NSW,
Australia), stored and displayed on personal computer. The
arterial contraction and relaxation was tested by sequential
application of 10-ꢀm PE and 10-ꢀm ACh. Only vessels
showing 80–100% relaxation to ACh were considered as
endothelium intact, while in some experiments, the
endothelium was predenuded mechanically, and relaxations
of <20% were considered as successfully denuded. After
washing for 45–60 min, vessels were precontracted by
adding 10-ꢀm PE again. When stable contractions were
obtained, the samples (containing compounds 1–9) at con-
centrations of 0.1–100 ꢀm were cumulatively added
(Figure 3). The samples were dissolved in DMSO and then
diluted with water to obtain the final concentrations of 300,
100, 30, 10, 3, 1, 0.1 ꢀm in 2-ml tissue baths (final solutions
contained <0.1% DMSO). Sildenafil was similarly diluted to
working concentrations of 0.0001–100 ꢀm.
⎛
⎜
CPMsample − CPMbackground
⎝ CPMtotal count − CPMbackground
⎞
(
)
%hydrolysissample
=
×100
⎟
⎠
(
)
(1)
where CPMsample is the radioactive count rate of the assay
with enzyme and CPMbackground is the count rate without
enzyme. CPMtotal
is the count rate of 20 ꢀl of substrate
count
plus 100 ml of buffer 1.
%hydrolysis
sample ⎞ ⎤
⎡
⎣
⎛
%PDE5 inhibition = 1−
×100
⎟
⎥
(2)
⎜
⎝
⎢
⎠
%hydrolysiscontrol
⎦
where %hydrolysissample and %hydrolysiscontrol were the
enzyme activities of the sample and solvent (1% DMSO)
used in the assay, respectively. The IC50 values were deter-
mined using the test samples at >80% PDE5 inhibition.
Determination of phosphodiesterase-6
inhibitory activity
PDE6 activity was conducted using the procedure previ-
ously reported,[14] which has been modified from Huang et
al., 1998.[18] Twenty-five millilitres of the following reagents
was added to tube: buffer 2, EGTA, PDE6 solution and test
samples or control (5% DMSO in buffer). The reaction was
started by adding 25 ꢀl of 5-ꢀm [3H]cGMP and incubated
at 30°C for 10 min. Then, the reaction was stopped by
placing the tube in boiling water for 1 min and cooled for
5 min. The second step of reaction used 25 ꢀl of 2.5-mg/ml
snake venom added to the reaction, incubated at 30°C for
5 min. After that, 250 ꢀl of 20-mm Tris-HCl, pH 6.8 (buffer
I) was added. The reaction was transferred to a QAE resin
column and eluted four times with 500 ꢀl of buffer I.
The eluent was mixed with a scintillation cocktail, and the
Statistical analysis
Data were expressed as the mean standard error of the
mean. Statistical analysis was conducted using Student’s
t-test: unpaired and one-way analysis of variance, followed
by Tukey’s post-hoc test. P-values of < 0.05 were considered
significant.
Results
radioactivity was measured using
a β-counter. The
%hydrolysis of PDE6 was similarly calculated as for PDE5.
Inhibition of phosphodiesterase-5 and
phosphodiesterase-6 by curcumin analogues
Animals
Three naturally occurring curcuminoids (1–3) and six syn-
thetic analogues (4–9) were tested using the PDE5 and
PDE6 inhibition assays. The highest potency on PDE5 was
for compounds 7 and 9 with IC50s of ∼4 ꢀm. Nevertheless,
compounds 7 and 9 were ∼100-fold less potent than
sildenafil, but more effective than the naturally occurring
Male Wistar rats (200–250 g) were obtained from the
National Laboratory Animal Center, Mahidol University,
Nakhornpathom, Thailand. The study was approved by the
Animal Ethics Committee, Naresuan University,
Phitsanulok, Thailand (NU-AE540416). Animals were
4
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–••

Oraya Kruangtip et al.
Curcumins inhibit PDE5
10-μM ACh
10
0.1
1
3
Compound 3, μM
30
100
300 μM
Wash
Tension
0.1 g
Wash
10 min
10-μM PE
10-μM PE
10-μM PE
Figure 3 An example time-course showing vasorelaxation of an endothelium-intact pulmonary artery to ACh followed by the relaxant effect of
compound 3 at 0.1 to 300 ꢀM. PE, phenylephine, ACh, acetylcholine, K, high potassium solution.
1–3 (Table 1). The other compounds showed IC50 values
that varied between 10 and 100 ꢀm. For PDE6, the IC50s for
compounds 2, 5, 6, 7, 8 and 9 were in the range of 3–20 ꢀm.
Only 1, 3 and 4 showed IC50s >100 ꢀm (Table 1).
Discussion
Structure–activity relationship of curcumin
analogues on phosphodiesterases
Cell-free experiments showed that all curcumin analogues
were moderately active PDE5 inhibitors (Table 1). When
the metamethoxyl group was missing from 1, the activities
of 2 and 3 were reduced suggesting that this group is
Vasorelaxation of pulmonary artery and
aorta by curcumin analogues
In isolated pulmonary arteries with intact endothelium,
the natural curcuminoids (1–3) and synthetic analogues
(4–9) demonstrated concentration-dependent vasodilata-
tion (Figures 3 and 4). Sildenafil achieved a supramaximal
relaxation, but limited solubility and lower potency pre-
vented us from determining the supramaximal effect for
any of the test analogues while some of maximal responses
did not even reach 50% of the maximal sildenafil
response. Therefore, all potencies were expressed as an
EC40 (for sildenafil both EC50 and EC40 were determined).
Every test analogue showed similar potencies (mean
EC40s were 58–111 ꢀm) on intact pulmonary arteries
(Table 2).
Endothelial denudation of pulmonary arteries showed
smaller relaxations with compounds 2, 5, 8 and 9 (Table 2)
(Figure 4b, 4e, 4h and 4i). Similarly, sildenafil potency was
substantially lower in endothelium-denuded pulmonary
arteries (Figure 4j). For the remaining compounds 1, 3, 4, 6
and 7 (Figure 4a, 4c, 4d, 4f and 4g), there was no change in
potency with denudation.
important for PDE5 inhibitory activity. Demethylation of
1 gave the more polar analogue 4 resulting in a twofold
lower inhibitory activity while removing both methoxyl
groups yielded compound 3, which produced a fourfold
reduction in activity. It might be possible that these
metapositions need these bulky substituents for binding to
the active site.[19]
Replacement of the ketomethylene group in 1 giving 5
slightly decreased the inhibitory activity whereas a similar
structural modification of 2–6 increased the activity. Taking
compound 6 and transposing the two hydroxyl groups from
positions 4′ to 2′ and 4″ to 3″ and removing the aromatic
methoxyl group yielded the highly potent compound 7.
Water solubility was a major challenge for curcumin ana-
logues (1–7) and solubility was improved by adding a
triazole carboxylic group to the 2′ hydroxyl group in 7 to
give 8 but this compromised PDE5 potency. In contrast, a
similar substitution on the opposite 3″-hydroxyl group of 7
yielding 9 preserved the inhibitory activity as well as offer-
ing superior water solubility.
In contrast, most of the test compounds produced
weaker actions on aorta compared with intact pulmonary
artery except compound 9, which was equally potent on
both aorta and pulmonary artery (Figure 4i). Sildenafil was
∼40-fold less effective on the aorta than intact pulmonary
artery; that is, only ∼5-fold more potent than denuded pul-
monary artery (Figure 4j). This could be a consequence of
more PDE5 in pulmonary artery but we are unaware of any
direct comparison of PDE5 in vascular smooth muscle on
different arteries.
The inhibitory activity on PDE6 is another concern
because it disrupts the cGMP signalling pathway used in
retinal transduction and this is avoided in the highly selec-
tive PDE5 inhibitor, tadalafil.[19] The inhibition by these
compounds on PDE6 suggests that 1, 3 and 4 had weak
actions compared with the corresponding actions on PDE5
and accords with the 10-fold selectivity of sildenafil.[20]
These results suggest that further modification of
curcuminoid analogues could achieve the required specific-
ity and high activity needed to realise clinical usefulness.
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–••
5

Curcumins inhibit PDE5
Oraya Kruangtip et al.
Table 1 The inhibitory effects of curcumin and its analogues on PDE5 and PDE6. Values are means standard error of the mean (n = 3)
IC50against
Compounds
1
Chemical structures
PDE5 (ꢀM)
PDE6 (ꢀM)
18.8 2.1a
113.9 20.9b
2
3
4
50.6 3.3c
94.4 5.2d
44.5 1.5c
12.6 2.8a
>500c
>700d
5
6
30.5 5.1b
27.6 5.7b
18.1 9.7a
7.6 1.8a
7
8
9
4.4 1.6a
17.1 2.0a
3.9 0.6a
4.0 2.1a
5.1 2.0a
2.8 2.3a
Sildenafil
0.03 0.01e
ND
PDE, phosphodiesterase. a–eDifference within columns (samples not connected by the same letter are statistically diferent at P < 0.05). PDE,
phosphodiesterase.
previous work in rat pulmonary artery[21] and aorta.[22]
Vasorelaxant effects of curcumin analogues
Furthermore, the potency of sildenafil here was similar to
In these experiments, sildenafil potency on intact
pulmonary arteries expressed as EC40 was 0.04 ꢀm
(0.074 0.016 ꢀm calculated as EC50), which accords with
the cell-free action on PDE5 protein. Endothelial removal
caused a dramatic decrease in sildenafil potency (200-fold
less) both here and in previous work on aorta.[20] This
6
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–••

Oraya Kruangtip et al.
Curcumins inhibit PDE5
100
80
60
40
20
0
100
80
60
40
20
0
(a)
(b)
#
Compound 1
Compound 2
#*
#*
#
Pulmonary artery
#*
#*
#
#
intact
-endothelium
aorta
0.1
1
10
100 μM
0.1
1
10
100 μM
100
80
60
40
20
0
100
80
60
40
20
0
(c)
(d)
#
Compound 4
Compound 3
#*
#
#
#
#*
#*
10
0.1
1
10
100 μM
0.1
1
100 μM
100
80
60
40
20
0
100
80
60
40
20
0
(e)
(f)
#
Compound 6
#*
Compound 5
#*
#
#
#
#*
0.1
1
10
100 μM
0.1
1
10
100 μM
100
80
60
40
20
0
100
80
60
40
20
0
#*
(g)
(h)
#
#
Compound 7
#
Compound 8
#*
#*
#*
0.1
1
10
100 μM
0.1
1
10
100 μM
*
#*
100
80
60
40
20
0
100
80
60
40
20
0
#*
(i)
(j)
*
*
#*
Compound 9
Sildenafil
#*
#*
*
#*
#*
*
0.1
1
10
100 μM
Concn. of compound
0.0001 0.001 0.01 0.1
1
10
100 μM
Concn. of compound
Figure 4 Concentration–vasorelaxation plots for the nine curcumin analogues and sildenafil. Each value is mean standard error of the mean
(n = 5–6). *P < 0.05 compared relaxation of pulmonary arteries with and without endothelium, #P < 0.05 comparing endothelium-intact pulmonary
arteries with aortas.
confirms that the vascular smooth muscle relaxation was
largely mediated through the endothelium-releasing vasodi-
lator factors even though sildenafil is acting on vascular
smooth muscle.[22]
All the curcuminoids were vasorelaxant using
endothelium-intact pulmonary artery and all had similar
potencies (Table 2). Compounds 2, 5, 8, 9 and possibly 4
had actions that indicate that the endothelium was neces-
sary. But, these four compounds had similar potencies
that did not reflect those variations seen in the cell-free
assays on either PDE5 of PDE6. Thus for compounds
2, 5, 8 and 9, there was clearly endothelium dependency,
but the poor potency correlation with the cell-free studies
does not clearly indicate that they are acting on PDE5.
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–••
7

Curcumins inhibit PDE5
Oraya Kruangtip et al.
Table 2 Vasorelaxant actions of curcumin and its analogues on rat
endothelium-intact and denuded pulmonary arteries and aorta. (n = 5–
6), P-values listed are for differences between endothelium-intact
versus endothelium-denuded pulmonary arteries.
some selectivity for the pulmonary artery. This alone indi-
cates that the compounds may form the basis for the devel-
opment of drugs that selectively target the pulmonary
circulation. Finally, the multiple cellular actions of
curcuminoids[26–28] may be an asset in the treatment of PAH
where there are multiple pathologies including inflamma-
tion, PDE5 upregulation, ionchannelopathies, vasoconstric-
tion, endothelial dysfunction and vascular hyperplasia.
EC40 (ꢀM)
Pulmonary
Pulmonary artery
denuded
Compounds
artery intact
Aorta-intact
1
109 23
58 10
111 29
58 20
76 24
98 28
93 30 (P = 0.82)
>300 0 (P < 0.0001)
184 31 (P = 0.14)
121 20 (P = 0.058)
>300 0 (P < 0.0001)
147 21 (P = 0.17)
52 6 (P = 0.74)
208 17 (P < 0.0001)
>300 0 (P < 0.0001)
8.4 0.8 (P < 0.0001)
>300 0#
>300 0#
>300 0#
>300 0#
>300 0#
>300 0#
>300 0#
>300 0#
46 11
2
3
Conclusions
4
Curcumin analogues showed PDE5 inhibitory activity with
varying potencies and some showed selectivity for PDE5
over PDE6. There were clear endothelium-dependent
vasorelaxant effects to which the pulmonary artery was
more sensitive compared with the aorta. These results
suggest that these curcuminoids could underpin the further
development of highly selective and potent compounds,
which could discriminate the pulmonary arterial circulation
by targeting several coincident pathologies of PAH includ-
ing PDE5 upregulation.
5
6
7
59
59
8
8
8
9
71 15
Sildenafil
0.042 0.009
1.8 1.0#
#P < 0.0001 for endothelium-intact pulmonary arteries versus intact
aorta.
Compounds 1 and 7 were unaffected by endothelial denu-
dation suggesting that they acted directly on vascular
smooth muscle through
a mechanism probably not
Declarations
directly on PDE5. These might include actions on soluble
guanylyl cyclase[23] on β-receptors, or cytosolic Ca2+
handling.[24]
There are two important differences between the cell-free
and vascular relaxation studies, which might affect potency
of our compounds: (1) the compounds have to gain access
to the cell interior, and numerous bioavailability studies
have shown that the membrane permeability of at least
curcumin itself is very poor;[25] (2) a vast number of cellular
effects for curcuminoids have been described.[26]
Conflict of interest
The Author(s) declare(s) that they have no conflicts of
interest to disclose.
Funding
The financial supports from Naresuan University, the
Center of Excellence for Innovation in Chemistry (PERCH-
CIC), Office of the Higher Education Commission, Minis-
try of Education, Thailand and The Thailand Research
Fund are acknowledged.
However, the very high concentrations needed to have
any effect on the aorta suggest that these compounds have
3. Sebkhi
A
et al. Phosphodiesterase
and smooth muscle tone in rat iso-
lated aorta. A study with selective
inhibitors. Biochem Pharmacol 1987;
36: 3965–3972.
References
type 5 as a target for the treatment
of hypoxia-induced pulmonary hyper-
tension. Circulation 2003; 107:
1. Ling Y et al. Changing demographics,
epidemiology, and survival of incident
pulmonary arterial hypertension:
results from the pulmonary hyperten-
sion registry of the United Kingdom
and Ireland. Am J Respir Crit Care Med
2012; 186: 790–796. doi: 10.1164/
rccm.201203-0383OC.
3230–3235.
.0000074226.20466.B1.
4. Scipioni et al. Immunohisto-
doi:
10.1161/01.CIR
6. Galiè N et al. Guidelines on diagnosis
and treatment of pulmonary arterial
hypertension The Task Force on Diag-
nosis and Treatment of Pulmonary
Arterial Hypertension of the European
Society of Cardiology. Eur Heart J
2004; 25: 2243–2278. doi: 10.1016/
j.ehj.2004.09.014.
A
chemical localisation of PDE5 in rat
lung during pre- and postnatal
development.
J Biomed Biotechnol
2. Barst RJ et al. Diagnosis and differen-
tial assessment of pulmonary arterial
hypertension. J Am Coll Cardiol 2004;
43(Suppl. 12 S): 40S–47S. doi:
10.1016/j.jacc.2004.02.032.
2009; 2009: 1–7. doi: 10.1155/2009/
932961.
5. Schoeffter P et al. Role of cyclic AMP-
and cyclic GMP-phosphodiesterases in
the control of cyclic nucleotide levels
7. Chrysant SG. Effectiveness and safety
of phosphodiesterase 5 inhibitors in
patients with cardiovascular disease
8
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–••

Oraya Kruangtip et al.
Curcumins inhibit PDE5
and hypertension. Curr Hypertens Rep
2013; 15: 475–483. doi: 10.1007/
s11906-013-0377-9.
chloro-3-((4-phenyl-1H-1,2,3-triazol-
Pharmacol 2003; 139: 513–522. doi:
1-yl)methyl)quinoline derivatives via
10.1038/sj.bjp.0705277.
click chemistry approach. Eur J Med 22. Mochida
Chem 2010; 45: 3142–3146. T-1032, phosphodiesterase type
et al. Kaempferia
H
et al. Sildenafil and
8. Kukreja RC et al. Emerging new uses
5
of phosphodiesterase-5 inhibitors in 16. Temkitthawon
P
inhibitors, showed a different vasore-
laxant property in the isolated rat
cardiovascular diseases. Exp Clin
Cardiol 2011; 16: e30–e35.
parviflora, a plant used in traditional
medicine to enhance sexual perfor-
mance contains large amounts of low
aorta. Eur
J
Pharmacol 2002;
9. Kerr NM, Danesh-Meyer HV.
Phosphodiesterase inhibitors and the
eye. Clin Experiment Ophthalmol
2009; 37: 514–523. doi: 10.1111/
j.1442–9071.2009.02070.x.
440: 45–52. doi: 10.1016/S0014-
2999(02)01339-0.
affinity
PDE5
inhibitors.
J
Ethnopharmacol 2011; 137: 1437– 23. Dash JR, Parija SC. Spasmolytic effect
1441. doi: 10.1016/j.jep.2011.08.025.
17. Sonnenburg WK et al. Identification,
quantitation, and cellular localization
of curcumin on goat ruminal artery is
endothelium independent and by
activation of sGC. Res Vet Sci 2013; 95:
588–593. doi: 10.1016/j.rvsc.2013
.04.029.
10. Adaramoye OA et al. Hypotensive
and endothelium-independent vaso-
relaxant effects of methanolic extract
from Curcuma longa L. in rats. J
Ethnopharmacol 2009; 124: 457–462.
doi: 10.1016/j.jep.2009.05.021.
of
PDE1
calmodulin-stimulated
cyclic nucleotide phosphodiesterases.
Methods 1998; 14: 3–19. doi: 10.1006/ 24. Xu P-H et al. The relaxant effect of
meth.1997.0561.
curcumin on porcine coronary arterial
ring segments. Vascul Pharmacol 2007;
47: 25–30. doi: 10.1016/j.vph.2007
.03.003.
18. Huang D et al. Molecular determi-
nants of cGMP binding to chicken
cone photoreceptor phosphodie-
11. Lin Q et al. Effect of curcumin on pul-
monary hypertension and wall colla-
gen of pulmonary arterioles of
chronic hypoxic hypercapnic rats.
sterase. J Biol Chem 2004; 279: 48143– 25. Wahlang
48151. doi: 10.1074/jbc.M404338200.
B et al. Identification of
permeability-related hurdles in oral
delivery of curcumin using the Caco-2
cell model. Eur J Pharm Biopharm
2011; 77: 275–282. doi: 10.1016/
j.ejpb.2010.12.006.
Chin J Appl Physiol 2006; 22: 257–261. 19. Cahill KB et al. Identification of
12. Temkitthawon P. Investigation of
phosphodiesterase-5 inhibitory activity
from Thai medicinal plants. Naresuan
University, Thailand: PhD thesis, 2012.
13. Changtam C et al. Curcuminoid ana-
logues with potent activity against
amino acid residues responsible for
the selectivity of tadalafil binding to
two closely related phosphodie-
sterases, PDE5 and PDE6. J Biol Chem 26. Shishodia S. Molecular mechanisms of
2012; 287: 41406–41416. doi: 10.1074/
jbc.M112.389189.
curcumin action: gene expression.
Biofactors 2013; 39: 37–55. doi:
10.1002/biof.1041.
Trypanosoma and Leishmania species. 20. Teixeira CE et al. Differential effects of
Eur J Med Chem 2010; 45: 941–956.
14. Chuprajob et al. Synthesis,
the phosphodiesterase type 5 inhibi- 27. Metzler M et al. Curcumin uptake and
T
tors sildenafil, vardenafil, and tadalafil
metabolism. Bio Factors 2013; 39:
cytotoxicity against human oral cancer
KB cells and structure–activity rela-
tionship studies of trienone analogues
in rat aorta. J Pharmacol Exp Ther
14–20. doi: 10.1002/biof.1042.
2006; 316: 654–661. doi: 10.1124/ 28. Bronte E et al. Role of curcumin in
jpet.105.092544.
idiopathic pulmonary arterial hyper-
tension treatment: A new therapeutic
possibility. Med Hypotheses 2013; 81:
923–926. doi: 10.1016/j.mehy.2013
.08.016.
of curcuminoids. Bioorg Med Chem 21. Pauvert O et al. Effect of sildenafil on
Lett 2014; 24: 2839–2844.
cyclic nucleotide phosphodiesterase
activity, vascular tone and calcium sig-
naling in rat pulmonary artery. Br J
15. Kategaonkar AH et al. Synthesis
and biological evaluation of new 2-
© 2014 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, ••, pp. ••–••
9