Modulation of the transient receptor potential vanilloid channel TRPV4 by 4α-phorbol esters: A structure-activity study

Article abstract of DOI:10.1021/jm9001007

The mechanism of activation of the transient receptor potential vanilloid 4 (TRPV4) channel by 4α-phorbol esters was investigated by combining information from chemical modification of 4α-phorbol-didecanoate (4α-PDD, 2a), site-directed mutagenesis, Ca

Full text of DOI:10.1021/jm9001007

J. Med. Chem. 2009, 52, 2933–2939
2933
Modulation of the Transient Receptor Potential Vanilloid Channel TRPV4 by 4r-Phorbol
Esters: A Structure-Activity Study
Thomas Kjær Klausen,†,§ Alberto Pagani,‡ Alberto Minassi,‡ Abdellah Ech-Chahad,‡ Jean Prenen,† Grzegorz Owsianik,†
Else Kay Hoffmann,§ Stine Falsig Pedersen,§ Giovanni Appendino,*,‡ and Bernd Nilius*,†
KU LeuVen, Department of Molecular Cell Biology, Laboratory Ion Channel Research, Campus Gasthuisberg, Herestraat 49, bus 802, LeuVen,
Belgium, Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e Farmacologiche, Via BoVio 9, 28100 NoVara, Italy, Department of
Biology, Section of Cell and DeVelopmental Biology, UniVersity of Copenhagen, 13 UniVersitetsparken, 2100 Copenhagen Ø, Denmark
ReceiVed October 24, 2008
The mechanism of activation of the transient receptor potential vanilloid 4 (TRPV4) channel by 4R-phorbol
esters was investigated by combining information from chemical modification of 4R-phorbol-didecanoate
(4R-PDD, 2a), site-directed mutagenesis, Ca2+ imaging, and electrophysiology. Binding of 4R-phorbol esters
occurs in a loop in the TM3-TM4 domain of TRPV4 that is analogous to the capsaicin binding site of
TRPV1, and the ester decoration of ring C and the A,B ring junction are critical for activity. The lipophilic
ester groups on ring C serve mainly as a steering element, affecting the orientation of the diterpenoid core
into the ligand binding pocket, while the nature of the A,B ring junction plays an essential role in the
Ca2+-dependence of the TRPV4 response. Taken together, our results show that 4R-phorbol is a useful
template to investigate the molecular details of TRPV4 activation by small molecules and obtain information
for the rational design of structurally simpler ligands for this ion channel.
Introduction
Currently, the best described small molecule TRPV4 ligands
are bisandrographolide A (BAA, 1, Figure 1), a plant dimeric
diterpenoid,11 4R-phorbol-12,13-didecanoate (4R-PDD, 2a), a
semisynthetic nontumor promoter phorbol ester,12,13 and
GSK1016790A (3), a synthetic peptide.14,15 No recognized
structural relationship exists between these compounds, and
limited information is available on their structure-activity
relationships. 4R-Phorbol esters have been extensively employed
in biomedical research as negative controls for the activity of
their corresponding phorbol esters, a class of ultrapotent
biological analogues of the secondary messenger 2-arachidoyl-
glycerol (2-AG) for the activation of protein kinase C (PKC).16
The discovery that 4R-PDD (2a) strongly interacts with TRPV4
has further emphasized that the phorbol core, a per se biologi-
cally inert structure, is a pleiotropic framework for the induction
of bioactivity by acylative modification.17
TRPV1 and TRPV4 show some degree of similarity (60%
homology, 40% identity) in the TM3-4 region, an element
involved in vanilloid binding and activation of TRPV1,18 and
site-directed mutagenesis experiments have indeed confirmed
the location of a ligand-binding element in this region. Thus,
mutation of L584A and of W586A in TM4 of TRPV4
specifically inhibited activation by heat and 4R-phorbol-esters,19
while the mutations Y591A and R594A inhibited activation by
all stimuli, suggesting a role in gating rather than in ligand
binding.13 The YS motif at the N-terminal end of TM3 in
TRPV4 (Y556 and S557) is strongly reminiscent of the capsaicin
domain of TRPV1,20 and mutating Y556 to alanine strongly
inhibited the activation of TRPV4 by 4R-phorbol-esters.21 This
effect correlated with the length of the aliphatic side chains of
the 4R-phorbol esters, pointing to the involvement of Y556 in
binding to the aliphatic side chain of these compounds.13
Binding of 4R-phorbol-12-myristate-13-acetate (2b), an ana-
logue of 4R-PDD (2a) with potent TRPV4 agonistic activity
(EC50 ∼ 3 µM), was strongly affected by the Y556A mutation,
while binding of 4R-phorbol diacetate (2c) was unaffected by
this mutation,19 suggesting that the long acyl moiety at C-12 is
critically involved in binding. Conversely, mutation of S557 to
The transient receptor potential (TRPa) channel proteins form
cation-permeable structures of low evolutionary relation, grouped
into seven subfamilies structurally characterized by six trans-
membrane (TM) domains and a pore region between TM5 and
TM6.1-5 Most TRP channels are Ca2+ permeable and sense
multiple physical stimuli (cold, heat, osmolarity, voltage) as well
as a plethora of noxious compounds and intracellular signaling
molecules like endocannabinoids and phospholipase C (PLC)-
and phospholipase A2 (PLA2)-derived products.1-5 Recently,
evidence has been mounting that TRP channels, and especially
those of the TRP vanilloid (TRPV)-type, are involved in a host
of human inflammatory pathologies.6 Thus, compounds capable
of preventing activation of TRPV1 by classic receptor antago-
nism or by desensitization have been extensively investigated,
and several capsaicinoid agonists and a host of synthetic TRPV1
antagonists have reached clinical studies.7 The large number
of TRPV1 ligands from the natural products pool has undoubt-
edly contributed to the wealth of studies on this channel, while
much less is known regarding ligand interactions of related TRP
channels (TRPV2-4).8
TRPV4 is widely expressed in epithelial cells and has also
been found in the brain, endothelium, liver, and trachea.9 Several
lines of evidence have pointed out the involvement of TRPV4
in important pathological conditions such as hypotonic hyper-
algesia, thermal hyperalgesia, asthma, and neuropathic pain.10
* To whom correspondence should be addressed. For B.N.: phone, +32
16 34 5937; fax, 32 16 34 5991; E-mail, bernd.nilius@med.kuleuven.be.
For G.A.: phone, +39 0321375744; fax, +39 0321375621; E-mail,
appendino@pharm.unipmn.it.
†
KU Leuven, Department of Molecular Cell Biology, Laboratory Ion
Channel Research.
§
Department of Biology, Section of Cell and Developmental Biology,
University of Copenhagen.
‡
Dipartimento di Scienze Chimiche, Alimentari, Farmaceutiche e
Farmacologiche.
a Abbrevations: TRP, transient receptor potential; TRPV, TRP vanilloid;
TM, transmembrane; PLC, phospholipase C; PLA2, phospholipase A2; PKC,
protein kinase C; 2-AG, 2-arachidoylglycerol, NT, nontransfected.
10.1021/jm9001007 CCC: $40.75
2009 American Chemical Society
Published on Web 04/10/2009

2934 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 9
Klausen et al.
13,20 diesters could be easily prepared from 4R-phorbol (2d,
Scheme 1), while methanolysis of the primary allylic 20-ester
groups eventually afforded the 12,13-diesters 2e-2j and the
13-monoester 2k from their corresponding triestes and the
diesters.24 In a similar way, the 4-deoxy analogue of 4R-PDD
(4c) was obtained from 4R-4-deoxphorbol (4b), a byproduct
from the isolation of phorbol from croton oil.22 Finally, lumi-
4R-PDD (5) was prepared from 4R-PDD by photochemical [2π
+ 2π] intramolecular photocycloaddition.22
Biological Evaluation
The activation of TRPV4 by 4R-phorbol deriviatives was
evaluated by measuring their capacity to elicit an increase in
the free intracellular Ca2+ concentration ([Ca2+]i) in mTRPV4
transfected HEK 293 cells, while electrophysiological studies
were performed in the whole cell patch clamp mode. Because
TRPV4 is activated by cell swelling, the osmolarity of this
solution was changed by omitting mannitol from an isotonic
solution (see Experimental Section).19
Results and Discussion
Ring C Acylation Pattern. Activity was strongly affected
by the acylation pattern of the ring C hydroxyls. Within the
series of esters investigated, the most potent compound was the
12,13-dihexanoate (2f) (4R-PDH, EC50 ) 70 nM, Table 1,
Figure 2). The closely related diacetate (2c), dioctanoate (2g),
and dinonanoate (2h) showed only marginal activity at the
maximal concentration assayed (50 µM), insufficient to estimate
EC50 values (Table 1, Figure 2). On the other hand, while the
12,13-dimyristate (2i) showed decreased activity compared to
4R-PDD (2a), the 12,13-distearate (2j) turned out to be too
lipophilic and insoluble to allow a reliable measurement of
activity. These observations show that the relationship between
the length of the side chain and the activity is not linear, with
two maxima of activity at C-6 and C-10 carbon length,
respectively. This differs markedly from the bell-shape curve
of potency vs acyl moiety length observed for the activation of
PKC by phorbol diesters.16 A bell-shaped relationship also holds
for the pungency of vanillamides, an indirect measurement of
TRPV1 activation,25 but this relationship is presumably more
complex because the introduction of unsaturations has different
effects in vanillamides with long (C-18) and medium (C9) length
aliphatic chains, with a substantial dissociation of TRPV1
affinity and pungency.25,26
Figure 1. Structural formulas of the TRPV4 ligands BAA (1), 4R-
PDD (2a), GSK1016790A (3), of the new ligands investigated (2e-k,
4a-c, and 5), and of the reference phorboids 2b-d.
alanine, a detrimental maneuver for capsaicin binding to
TRPV1,20 had little effect on 4R-PDD (2a) binding to TRPV4.21
Taken together, these data show that 4R-PDD (2a) is as an
interesting template to investigate the molecular details of the
interaction of TRPV4 with small molecules and obtain informa-
tion for the rational design of structurally simpler synthetic
analogues. For this purpose, we have focused on what have
emerged as the two critical elements for the biological profile
of 4R-PDD, namely the esterification pattern of ring C and the
nature of the A,B rings and their junction.
Previous site-selective mutagenesis data have suggested that
the binding of 4R-phorbol-diesters to TRPV4-Y556 involves
only the C-12 ester groups.19 However, the diester 4R-PDD (2a)
and its 13-monoester (2k) showed similar EC50 values (370 nM
vs 450 nM, Table 1), although the maximal increase in [Ca2+
]
i
elicited, i.e., the efficacy, of the monoester (2k) was lower than
that of the diester 4R-PDD (2a) (Figure 3). The relationship
between location and length of the acyl moieties and TRPV4
activation by 4R-phorbol-esters is complex and better accom-
modated by a model of interaction where the ring C ester groups
have an orienting- rather than a direct binding role, in contrast
with a previous proposal.19
We next focused on the characterization of the TRPV4
response by the most potent compound of the series, 4R-PDH
(2f) (EC50 value ) 70 nM, Figure 4A). Given the poor activity
of the dioctanoate 2g and the dinonanoate 2h, the almost 5-fold-
higher potency of the dihexanoate 2f compared to 4R-PDD 2a
raised the question whether the observed [Ca2+]i increase indeed
reflected TRPV4 activation. This was, however, clearly the case
because: (i) 2f did not affect [Ca2+]i in nontransfected HEK293
Chemistry
Phorbol (4a) and 4-deoxy-4R-phorbol (4b) were obtained
from commercial croton oil,22 while 4R-phorbol (2d) was
prepared from phorbol (4a) by base-catalyzed retro-aldol
epimerization.23 The reactivity pattern of the hydroxyls of
phorbol is rather peculiar and is dominated by the surprisingly
facile esterification of the tertiary C-13 hydroxyl and the easy
removal of the allylic primary ester group at C-20.22,24 The same
pattern of reactivity was found in 4R-phorbol (2d) and in its
4-deoxyderivative (4b), whose acylative decoration, a critical
element for TRPV4 binding,19 could be varied by applying the
protocols developed for phorbol esters.22,24 Thus, by simply
varying the amounts of acylating agent, 12,13,20-triesters and

Modulation of TRPV4 by 4R-Phorbol Esters
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 9 2935
Scheme 1. Synthesis of the 4R-Phorbol Diesters 2e-j, the 4R-4-Deoxyphorbol Diester 4c, and the 4R-Phorbol Ester 2k
Table 1. Affinity Responses for Compounds 2a,c,e-k, 4c, and 5a
EC50 (µM)
compd
carbons in acyl moiety
nHill
WT
Y556A
6.3*b ( 0.3
S557A
W586A
>6*b
2a
2c
2e
2f
2g
2h
2i
2j
2k
4c
5
10
2
4
6
8
2.2
0.37 ( 0.08
>50
22.47 ( 4.04
0.07 ( 0.01
>50
0.32 ( 0.15
2.6
1.5
2.79 ( 1.15
0.36 ( 0.01
2.84 ( 1.12
9
>50
14
18
10
10
10
2.0
2.0
1.0
2.83 ( 0.62
0.45 ( 0.08
>50
0.175 ( 0.07
9.36 ( 19.98
0.24 ( 0.20
3.64 ( 3.47
a EC50 values and Hill coefficients where calculated via eq 1 (See Experimental Section). b *Values taken from ref 19.
Figure 3. Comparison of the effects of 4R-PDD (2a) and the monoester
4R-PD (2k) on [Ca2+]i in HEK293 cells. The relative increase in [Ca2+
]
i
(shown as the increase in 340/380 nm fura-2 fluorescence ratio after
agonist addition, ∆Fura-2 ratio) was estimated as a function of the bath
concentration of 4R-PDD (2a, black circles) and 4R-PD (2k, gray
triangles). Solid lines represent a fit to the Hill equation. Points are
∆Fura-2 ratio for 18-32 individual cells from three independent
transfections. * indicates significant differences at the level p < 0.05.
Figure 2. Relationship between EC50values and side chain length of
4R-phorbol-diesters. EC50 values were calculated using eq 1. Results
for each compound are based on >15 individual cells from at least
three independent transfections.
exhibited substantial run-down, suggesting washout of an
internal factor necessary for TRPV4 activation (Figure 5A,B).
Site directed mutagenesis has highlighted the relevance of
residues L586, W586, and Y556 in 4R-phobol ester binding to
TRPV4.19 To assess whether 2f activates TRPV4 by interaction
with the same binding pocket as 4R-PDD (2a), dose-response
experiments were performed on wild type, W586A, S557A, and
Y556A mTRPV4 constructs expressed in HEK293 cells. Each
mutation shifted the dose-response curves to higher concentra-
tions, indicating that, indeed, 2f activates TRPV4 by binding
cells (data not shown), (ii) the response was dependent on
extracellular Ca2+ (Figure 3B), and (iii) the increase in [Ca2+
]
i
was inhibited by the TRP antagonist ruthenium red (RR) (Figure
4C). In whole cell patch clamp studies of mTRPV4 transfected
cells, 2f activated currents with outward-inward rectification
(Figure 5A) with less pronounced rectification in Ca2+ free
solution (Figure 5B). The current could be reactivated, but

2936 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 9
Klausen et al.
Figure 4. Effects of compound 2f on [Ca2+]iin mTRPV4 transfected
HEK 293 cells. (A) Dose-response relationship for 2f. ∆Fura-2 ratio
was estimated by single cell Ca2+ imaging as described in the
Experimental Section. Points are ∆Fura-2 ratio estimated for 25-61
individual cells from three independent transfections. The solid line
represents a fit to the Hill equation. (B) The response to 2f is dependent
on extracellular Ca2+. Compound 2f (1 µM) and extracellular Ca2+ were
present as indicated by top bars. The figure is representative of three
independent experiments for each condition and cell type. (C) The
[Ca2+]i response to 2f (1 µM) is inhibited by ruthenium red (RR, 1
µM). The presence of RR and 2f in the extracellular solution is indicated
by the top bars. The figure is representative of three independent
experiments for each condition and cell type.
Figure 5. Characteristics currents activated by 2f. Whole cell currents
were analyzed in mTRPV4-transfected HEK293 cells as described in
the Experimental Section. (A) Left: Whole cell mTRPV4 current
development as a function of time. In Ca2+ containing solutions, currents
in mTRPV4 transfected cells were monitored via ramp protocols from
-100 mV to +100 mV. Compound 2f was included in the extracellular
solution as indicated by the top bar. The traces are representative of
five individual cells in >2 transfections. Right: I/V relationship of the
2f activated currents. Same experiments as in Figure 4A. The numbers
refer to ramps obtained at the times indicated by numbers in left panel.
(B) Left: Experiment as in A, except for the absence of Ca2+ in the
intra- and extracellular solutions. The traces are representative of five
individual cells in >2 transfections. Right: I/V relationship of the 2f
activated currents in the absence of Ca2+. Same experiments as in Figure
4C. The numbers refer to ramps indicated by numbers in Figure 4C.
in the same pocket as 2a (Figure 5). Y556 is part of a YS motif
in TM3, similar to that shown to be important for capsaicin
activation of TRPV1.20 On the other hand, while the residue
TRPV1 equivalent (S512) to S557 in TRPV4 is important for
binding of capsaicin to TRPV1,20 its mutation to alanine had
little effect on 4R-PDD binding to TRPV4 (Table 1).21 Notably,
the S557A mutation shifted the dose-response curve for
activation of TRPV4 by 2f toward higher concentrations.
Although this shift was less pronounced than in the Y556A
mutant (Figure 6), it suggests that the increased potency of 2f
compared to 2a may be related to the presence of a further
binding element in TRPV4 for 2f. Because serine is a polar
element, it is unlikely that this residue directly interacts with
the acyl moiety of 2f, supporting the view that, compared to
other side chains, the hexanoate moiety positions the polar
diterpenoid core in a more optimal orientation for interaction
with TRPV4 rather than directly interacting with the binding
site. The involvement of both Y556 and S557 in TRPV4 gating
is reminiscent of similar observations reported for TRPV120 and
suggests that these two channels share similar gating properties.
Changes to the A,B Ring Junction. The A,B ring junction
is a critical structural feature of the phorboid TRPV4
pharmacophore. To further explore this dependence, 4R-PDD
(2a) was subjected to photocyclization, generating its cor-
responding cage-like lumiphorbol derivative (5), while the
4-deoxygenated 4R-PDD analogue (4c) was prepared from
4R-4-deoxphorbol (4b). Compound 4c was unable to activate
TRPV4 (Table 1), suggesting that the 4-hydroxyl is crucial
for agonist activation of TRPV4. Conversely, 4R-LPDD (5)
outperformed 4R-PDD (2a) in terms of TRPV4 activation,
with an almost 2-fold lower EC50 value (175 ( 0.07 vs 370
nM, Table 1 and Figure 7A for 5). At 0.1 µM concentration,
Figure 6. (D) Dose-response relationship for 2f in TM3-4 mTRPV4
mutants. Dose-response relationship for 2f in mTRPV4 mutants.
HEK293 cells were transiently transfected with either wild type
mTRPV4 (WT, square symbols), or the mutants Y556A (circles),
S557A (downward triangles), or W586A (upward triangles). Agonist
activity of 2f was tested by Ca2+ imaging as described in the
Experimental Section, and shown as ∆Fura-2 ratio. Data for each point
are calculated from >16 individual cells from three independent
transfections. * indicates significant differences from wild type (p <
0.05).
shown, p < 0.01). On the other hand, at concentrations above
10 µM, 5 elicited [Ca2+]i responses also in nontransfected
cells (not shown). These [Ca2+]i responses were not increased
by the expression of TRPV1-3, indicating that 4R-LPDD
(5) did not affect related TRP channels (data not shown).
However, at less than 10 µM, there was no effect of 5 in
5 could still elicit a significantly greater increase in [Ca2+
]
i
in mTRPV4 transfected HEK293 cells than did 2a (data not

Modulation of TRPV4 by 4R-Phorbol Esters
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 9 2937
Figure 7. Effects of 4R-LPDD (5) on [Ca2+]i in mTRPV4 transfected
HEK 293 cells. (A) Dose-response relationship for 5. ∆Fura-2 ratio
was estimated by single cell Ca2+ imaging as described in the
Experimental Section. Points are ∆Fura-2 ratio estimated for 24-71
individual cells from three independent transfections. The solid line
represents a fit to the Hill equation. (B) The response to 5 is dependent
on extracellular Ca2. Compound 5 (500 nM) and extracellular Ca2+
were present as indicated by the top bars. The figure is representative
of three independent experiments. (C) The [Ca2+]i response to 5 (500
nM) is inhibited by ruthenium red (RR, 1 µM). The presence of RR
and 5 in the extracellular solution is indicated by the bars. The figure
is representative of three independent experiments.
Figure 8. Characteristics of currents activated by 4R-LPDD (5). Whole
cell currents were analyzed in mTRPV4-transfected HEK293 cells as
described in the Experimental Section. (A) Left: Whole cell mTRPV4
current development as a function of time. In Ca2+ containing solutions,
currents in mTRPV4-transfected cells were monitored via ramp
protocols from -100 mV to +100 mV. Compound 5 was included in
the extracellular solution as indicated by the bar. The figure is
representative of five individual cells from >2 transfections. Right:
I/V relationship for currents activated by 5. Same experiment as in
Figure 7A. (B) Left: Conditions as in A, except that the experiment
was carried out in the absence of Ca2+ in both the intra- and extracellular
solutions. Note the dramatic difference in current duration. The figure
is representative of five individual cells from >2 transfections. Right:
nontransfected cells, neither in Ca2+ imaging experiments
(Figure 7B,C) nor in patch clamp studies (data not shown).
Furthermore, the [Ca2+]i response to 5 was dependent on
extracellular Ca2+ and was inhibited by RR (Figure 7B,C),
strongly suggesting that it reflects activation of TRPV4.
In patch clamp studies, 5 activated typical TRPV4
outward-inward rectifying currents in both the presence
(Figure 7A) and in the absence (Figure 8B) of extra- and
intracellular Ca2+. A number of significant differences
between the TRPV4 currents triggered by the 4R-lumiphorbol
ester (5) and by 4R-phorbol esters (2a and 2f) were observed.
Most notably, the same current densities were observed
irrespective of the presence of Ca2+ in recording solutions
(Ca2+ vs no Ca2+; 505 ( 102 pA/pF vs 635 ( 107 pA/pF; p
> 0.4 at 100 mV, n ) 5-6). This is in contrast to 4R-PDD
(2a), where Ca2+-dependent potentiation of TRPV4 is well
described27,28 and to currents activated by 4R-PDH (2f) (Ca2+
vs no Ca2+; 354 ( 112 pA/pF vs 806 ( 132 pA/pF; p <
0.05 at 100 mV, n ) 6-7) Second, in contrast to the transient
current activation by 4R-phorbol esters (Figure 5), currents
activated by the lumiphorbol (5) were maintained in Ca2+
free solutions (Figure 8B). Hence, currents activated by 4R-
PDH (2f) inactivated with an initial slope of -16.5 ( 4.9%
min-1, while the 4R-LPDD (5) had an almost constant slope
of 1.0 ( 1.0% min-1 after activation that was significantly
different (P ) 0.018, n ) 4-5). These findings highlight
the critical role for the ring A,B-junction for functional
agonist activity.
I/V relationship for currents activated by 5 in the absence of Ca2+
.
Same experiment as in Figure 7C.
Figure 9. Dose-response relationship for 4R-LPDD (5) in TM3-4
mTRPV4 mutants: Dose-response relationship for the [Ca2+]i response
elicited by 5 in HEK 293 cells transfected with wild type mTRPV4
(WT, square symbols), or the mutants Y556A (circles), S557A
(downward triangles), or W586A (upward triangles). The agonist
activity of 2f was tested by Ca2+ imaging as described in the
Experimental Section and shown as ∆Fura-2 ratio. Data for each point
are calculated from n > 15 individual cells from three independent
transfections. * indicates significant differences from WT (p < 0.05).
previously published results for 4R-PDD (2a)21 (Table 1).
Because 2a differs from 5 in the headgroup and from 2f only
in the length of the acyl groups, the fact that the S557A mutation
attenuates binding of 2f but not of 2a and 5 must be related to
the acyl decoration of the terpenoid core, seemingly due to a
suboptimal orientation of the terpenoid core with S557. Because
4R-PDD (2a) and its lumiphorbol derivative (5) apparently bind
the same pocket of TRPV4, the substantially different features
of the currents elicited by these ligands suggest, in fact, that
To gain information on the requirements for binding of 5 to
TRPV4, dose-response experiments for the 4R-LPDD-induced
increase in [Ca2+]i were carried out on wild type TRPV4 and
on W586A, S557A, and Y556A mutant TRPV4. The W586A
and Y556A mutations resulted in rightward shifts of the
dose-response curve (Figure 9), whereas the S557A mutation
had no effect. These observations are in agreement with

2938 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 9
Klausen et al.
the filtrate was evaporated to afford 870 mg of crude 4R-phorbol
12,20-didecanoate. The latter was dissolved in methanol (7 mL),
and the solution was brought to pH 2 by the addition of few drops
of 65% HClO4. After stirring overnight at room temperature, the
reaction was worked up by the addition of solid NaHCO3, filtration,
and evaporation. The residue was purified by gravity column
chromatography on silica gel (6 g, petroleum ether-EtOAc 7:3 as
eluant) to afford 48 mg (13% from 4R-phorbol) 2k as a colorless
oil.
Synthesis of LUMI 4r-PDD (5). A solution of 4R-PDD (2a)
in ethanol was degassed in a flow of nitrogen for 40 min and then
irradiated in an immersion well photoreactor with a low-pressure
mercury lamp. After 24 h of irradiation, the reaction was evaporated
and the residue was purified by gravity column chromatography
on silica gel (5 g, petroleum ether-EtOAc 8:2 as eluant) to afford
190 mg (95%) 5 as a colorless oil.
Cell Culture and Transfection. Human embryonic kidney cells,
HEK293, were grown in DMEM containing 10% human serum, 2
mM L-glutamine, 2 units/mL penicillin and 2 mg/mL streptomycin,
in a 37 °C/10% CO2 incubator. Transient transfection of HEK293
cells with mTRPV4 (accession number Q9EPK8) in the pCAGGS/
Ires-GFP vector29 was performed using Mirus TransIT-293 trans-
fection agent (Mirus Corporation; Madison, WI) and cells where
incubated overnight before experiments. For both [Ca2+]i measure-
ments and electrophysiological experiements, the cells were
trypsinized and seeded on poly-L-lysine coated coverslips on
the day of the experiment. The mTRPV4 expressing cells were
identified by GFP expression. GFP-negative cells from the same
coverslips were used as untransfected controls.
their terpenoid cores interact in a different way with the ligand
binding site of the receptor.
Conclusions
We have provided evidence that 4R-phorbol esters are useful
tools to investigate the mechanism of activation of TRPV4 by
small molecules. Our findings strongly indicate that the lipo-
philic side chains on the C-ring, although critical for TRPV4
activation, serve mainly as a steering element for a correct
positioning of the terpenoid core into the phorboid binding
pocket of TRPV4, where a critical role has been identified for
the TM3 YS motif. Moreover, we have demonstrated that the
nature of the A,B ring junction plays a major role in binding of
4R-phorbol esters to TRPV4. Finally, the marked difference
between 4R-phorbol- and 4R-lumiphorbol derivatives in their
effects on TRPV4 activity imply the existence of multiple
mechanisms to translate affinity into functional activity for this
class of compounds. In this context, the possibility of producing
continuous rather than transient TRPV4 activation could has
major implications for the therapeutic exploitation of TRPV4
agonists (e.g., bladder dysfunction14,28). Taken together, these
results illustrate the potential of combining structure-activity
studies and site-directed mutagenesis to map binding pockets
of “difficult” proteins, providing information useful to design
synthetic small molecule agonists for these proteins.
Experimental Section
Solutions. For [Ca2+]i measurements, cells were constantly
perfused with Krebs solution containing (in mM): 150 NaCl, 6 KCl,
1 MgCl2, 1.5 CaCl2, 10 HEPES, 10 glucose, pH 7.4. Where
indicated, CaCl2 was emitted from the solution. For electrophysi-
ological studies, the extracellular solution contained (in mM): 150
NaCl, 6 CsCl, 1 MgCl2, 5 CaCl2, 10 glucose, and 10 HEPES, pH
7.4. For Ca2+ free experiments, CaCl2 was omitted from the
solution. The pipet solution contained (in mM): 20 CsCl, 100
aspartate, 1 MgCl2, 10 HEPES, 4 Na-ATP, 10 K-1,2-bis(2-
aminophenoxy)ethane N,N,N′,N′-tetraacetate (BAPTA), pH 7.4.
Ca2+ was buffered to a free concentration of 200 nM with CaCl2
as calculated using the software CaBuf (G. Droogmans) or omitted
for Ca2+ free pipet solutions.
Electrophysiological Recording. All electrophysiological data
were obtained using 400 ms linear ramp protocols between -100
mV and 100 mV after a 50 ms prepulse at -100 mV as previously
described.19 Ramps were performed every 2 s, and the cells were
kept at a holding potential of -30 mV between ramps. Data were
recorded with 2 kHz sample rate using an EPC10 amplifier
controlled by PatchMaster software (Heka, Lambrecht, Germany)
and filtered at 2.9 kHz using the built-in Bessel filter. Pipets had
resistances of 2-4 MΩ in asymmetrical recording solutions and
an Ag-AgCl wire was used as reference electrode. Capacitance and
70-80% of the series resistance were electronically compen-
sated.
Materials. Gravity column chromatography: Merck silica gel
(70-230 mesh). IR: Shimadzu DR 8001 spectrophotometer. NMR:
1
Jeol Eclipse (300 and 75 MHz for H and 13C, respectively). For
1H NMR, CDCl3 as solvent, CHCl3 at δ ) 7.26 as reference. For
13C NMR, CDCl3 as solvent, CDCl3 at δ ) 77.0 as reference.
Commercial dry CH2Cl2 and THF were filtered from neutral alumina
before use. Reactions were monitored by TLC on Merck 60 F254
(0.25 mm) plates that were visualized by UV inspection and/or
staining with 5% H2SO4 in ethanol and heating. Organic phases
were dried with Na2SO4 before evaporation. A Beckmann apparatus
equipped with a UV detector set at 254 nm (refraction index for 5)
was used to assess purity (>95%) of all final products. A LUNA
SI60 column was used, with isocratic elution with MeOH-H2O
4:1, and 2.0 mL/min as flow rate. Rt ranged from 17 min (2c) to
34 min (2j).
Synthesis of 4r-Phorbol Diesters (2e-j) and the
4r-4-Deoxyphorbol Diester (4c): Synthesis of 4r-PDH (2f) as
Representative for the Process. To a solution of 4R-phorbol (2d,
500 mg, 1.37 mmol) in CH2Cl2-THF (1:1, 10 mL), hexanoic acid
(858 µL, 6.71 mmol, 5 mol equiv), EDC (1.58 g, 8.12 mmol, 6
mol equiv), and DMAP (251 mg, 2.1 mmol, 1.5 mol equiv) were
added. After stirring at room temperature for 4 h, the reaction was
worked up by the addition of brine and extraction with CH2Cl2.
The organic phase was filtered over neutral alumina (5 g) to remove
unreacted hexanoic acid, and the filtrate was evaporated to afford
870 mg of crude 4R-phorbol 12,13,20-trihexanoate. The latter was
dissolved in methanol (7 mL), and the solution was brought to pH
2 by the addition of few drops of 65% HClO4. After stirring
overnight at room temperature, the reaction was worked up by the
addition of solid NaHCO3, filtration, and evaporation. The residue
was purified by gravity column chromatography on silica gel (13
g, petroleum ether-EtOAc 7:3 as eluant) to afford 245 mg (30%
from 4R-phorbol) (2f) as a colorless oil.
Synthesis of the 4r-Phorbol Monoester (2k). To a solution of
4R-phorbol (2d, 200 mg, 0.55 mmol) in CH2Cl2-THF (1:1, 10
mL), decanoic (capric) acid (284 mg, 1.65 mmol, 3 mol equiv),
EDC (370 mg, 1.93 mmol, 3.5 mol equiv), and DMAP (67 mg,
0.55 mmol, 1 mol equiv) were added. After stirring at room
temperature for 2 h, the reaction was worked up by the addition of
brine and extraction with CH2Cl2. The organic phase was filtered
over neutral alumina (2 g) to remove unreacted decanoic acid, and
[Ca2+]i Measurements. Cells were preincubated for 30 min at
37 °C with 4 µM Fura-2 AM (TefLab, Austin, TX) in normal
growth medium and mounted on the stage of an IX81 inverted
microscope (Olympus, Tokyo, Japan). The imaging system con-
sisted of a MT10 filter-based illumination system (Olympus) and
a CCD Camera (Olympus). Illumination and imaging was controlled
by CellM software (Olympus). [Ca2+]i was estimated as the ratio
of the emitted fluorescence at 510 nm after excitation at 340 and
380 nm, respectively (fura-2 ratio). The background fluorescence
was digitally subtracted from individual experiments before the ratio
was calculated.
Data Analysis and Statistics. Electrophysiological data where
analyzed using Patchmaster software (HEKA, Lambrecht, Ger-
many), WinACD (Guy Drogmans, ftp://ftp.cc.kuleuven.ac.be/pub/
droogmans/winascd.zip) and Origin (OrigineLab, Northampton,
MA). Fura-2 AM fluorescence ratios were calculated in CellM
Software (Olympus) and Origin (OrigineLab, Northampton, MA).

Modulation of TRPV4 by 4R-Phorbol Esters
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 9 2939
(14) Thorneloe, K. S.; Sulpizio, A. C.; Lin, Z.; Figueroa, D. J.; Clouse,
A. K.; McCafferty, G. P.; Chendrimada, T. P.; Lashinger, E. S.;
Gordon, E. A.; Evans, L.; Misajet, B. A.; Demarini, D. J.; Nation,
J. H.; Casillas, L. N.; Marquis, R. W.; Votta, B. J.; Sheardown, S. A.;
Xu, X.; Brooks, D. P.; Laping, N.; Westfall, T. D. GSK1016790A, a
Novel and Potent TRPV4 Channel Agonist Induces Urinary Bladder
Contraction and Hyperactivity: Part I. J. Pharmacol. Exp. Ther. 2008,
326, 432–442.
∆Fura-2 ratio was calculated as (Rt - Rt)0)/Rt)0, and maximal
response plus time of maximal response were analyzed for
individual cells. EC50 values were calculated using the Hill equation:
Cn
Rmax ) Vmax
(1)
ECn50 + Cn
(15) Willette, R.; Bao, W.; Nerurkar, S.; Yue, T. L.; Doe, C. P.; Stankus,
G.; Turner, G. H.; Ju, H.; Thomas, H.; Fishman, C.; Sulpizio, A.;
Behm, D.; Hoffman, S.; Lin, Z.; Lozinskaya, I.; Casillas, L.; Lin, M.;
Trout, R. E. T.; Thorneloe, K.; Lashinger, E. S.; Figueroa, D. J.; Votta,
B.; Marquis, R.; Xu, X. Systemic Activation of the Transient Receptor
Potential V4 Channel Causes Endothelial Failure and Circulatory
Collapse: Part 2. J. Pharmacol. Exp. Ther. 2008, 326, 443–452.
(16) Goel, G.; Makkar, H. P.; Francis, G.; Becker, K. Phorbol Esters:
Structure, Biological Activity, and Toxicity in Animals. Int. J. Toxicol.
2007, 26, 279–288.
(17) Rios, M. Y.; Aguilar-Guadarrama, A. B. Nitrogen-Containing Phorbol
Esters from Croton ciliatoglandulifer and their Effects on Cyclooxy-
genases-1 and -2. J. Nat. Prod. 2006, 69, 887–890.
(18) Gavva, N. R.; Klionsky, L.; Qu, Y.; Shi, L.; Tamir, R.; Edenson, S.;
Zhang, T. J.; Viswanadhan, V. N.; Toth, A.; Pearce, L. V.; Vanderah,
T. W.; Porreca, F.; Blumberg, P. M.; Lile, J.; Sun, Y.; Wild, K.; Louis,
J. C.; Treanor, J. J. Molecular Determinants of Vanilloid Sensitivity
in TRPV1. J. Biol. Chem. 2004, 279, 20283–20295.
where Rmax is the maximal ratio at the given concentration (C),
Vmax is the maximal response, n is the Hill coefficient, and EC50
denotes the concentration eliciting half-maximal response.
Data are summarized as mean ( standard error of the mean
(SEM). Individual experiments shown are representative of at least
three independent experiments.
Acknowledgment. We thank all the members of the Ion
Channel Research Laboratory at the KU Leuven for helpful
discussions. This work was supported by grants from Interuni-
versity Attraction Poles Program-Belgian State-Belgian Sci-
ence Policy (P6/28), the Research Council of the KU Leuven
(GOA 2004/07), and the Flemish Government (Excellentiefin-
anciering, EF/95/010, to B.N.), by Danish Natural Sciences
Research Foundation (E.K.H. and S.F.P.), the Gangsted Founda-
tion (S.F.P.), and the Lundbeck Foundation (E.K.H.).
(19) Vriens, J.; Owsianik, G.; Janssens, A.; Voets, T.; Nilius, B. Deter-
minants of 4R-Phorbol Sensitivity in Transmembrane Domains 3 and
4 of the Cation Channel TRPV4. J. Biol. Chem. 2007, 282, 12796–
12803.
Supporting Information Available: Full spectroscopic data for
all final products. This material is available free of charge via the
Internet at http://pubs.acs.org.
(20) Jordt, S. E.; Julius, D. Molecular Basis for Species-Specific Sensitivity
to “Hot” Chili Peppers. Cell 2002, 108, 421–430.
(21) Vriens, J.; Watanabe, H.; Janssens, A.; Droogmans, G.; Voets, T.;
Nilius, B. Cell Swelling, Heat, and Chemical Agonists Use Distinct
Pathways for the Activation of the Cation Channel TRPV4. Proc. Natl.
Acad. Sci. U.S.A 2004, 101, 396–401.
(22) Appendino, G.; Bertolino, A.; Minassi, A.; Annunziata, R.; Szallasi,
A.; De Petrocellis, L.; Di Marzo, V. Synthesis and Biological
Evaluation of Phorbol-Resiniferatoxin RTX) Hybrids. Eur. J. Org.
Chem. 2004, 3413-3421.
(23) Appendino, G.; Carello, G. P.; Enriu`, R.; Jakupovic, J. A New
Rearrangement of a Phorbol Derivative. J. Nat. Prod. 1995, 58, 284–
287.
(24) Marquez, N.; Calzado, M. A.; Sanchez-Duffhues, G.; Perez, M.;
Minassi, A.; Appendino, G.; Diaz, L.; Mun˜oz-Ferna´ndez, A. A.;
Mun˜oz, E. Differential Effects of Phorbol-13-Monoesters on Human
Immunodeficiency Virus Reactivation. Biochem. Pharmacol. 2008, 75,
1370–1380.
References
(1) Pedersen, S. F.; Owsianik, G.; Nilius, B. TRP Channels: an Overview.
Cell Calcium 2005, 38, 233–252.
(2) Owsianik, G.; D’Hoedt, D.; Voets, T.; Nilius, B. Structure-Function
Relationship of the TRP Channel Superfamily. ReV. Physiol. Biochem.
Pharmacol. 2006, 156, 61–90.
(3) Voets, T.; Talavera, K.; Owsianik, G.; Nilius, B. Sensing with TRP
Channels. Nat. Biol. Chem. 2005, 1, 85–92.
(4) Montell, C. The TRP Superfamily of Cation Channels. Sci. STKE 2005,
2005, re3.
(5) Venkatachalam, K.; Montell, C. TRP Channels. Annu. ReV. Biochem.
2007, 76, 387–417.
(6) Jordt, S. E.; Ehrlich, B. E. TRP Channels in Disease. Subcell. Biochem.
2007, 45, 253–271.
(7) Appendino, G.; Szallasi, A. Clinically Useful Vanilloid Receptor
TRPV1 Antagonists: Just Around the Corner (or Too Early to Tell)?
Prog. Med. Chem. 2006, 44, 145–180.
(8) Appendino, G.; Minassi, A.; Pagani, A.; Ech-Chahad, A. The Role of
Natural Products in the Ligand Deorphanization of TRP Channels.
Curr. Pharm. Des. 2008, 14, 2–17.
(9) Vennekens, R.; Owsianik, G.; Nilius, B. Vanilloid Transient Receptor
Potential Cation Channels: An Overview. Curr. Pharm. Des. 2008,
14, 18–31.
(10) Nilius, B.; Owsianik, G.; Voets, T.; Peters, J. A. Transient Receptor
Potential Channels in Disease. Physiol. ReV. 2007, 87, 165–217.
(11) Smith, P. L.; Maloney, K. N.; Pothen, R. G.; Clardy, J.; Clapham,
D. E. Bisandrographolide from Andrographis Paniculata Activates
TRPV4 Channels. J. Biol. Chem. 2006, 281, 29897–29904.
(12) Watanabe, H.; Davis, J. B.; Smart, D.; Jerman, J. C.; Smith, G. D.;
Hayes, P.; Vriens, J.; Cairns, W.; Wissenbach, U.; Prenen, J.; Flockerzi,
V.; Droogmans, G.; Benham, C. D.; Nilius, B. Activation of TRPV4
Channels (hVRL-2/mTRP12) by Phorbol Derivatives. J. Biol. Chem.
2002, 277, 13569–13577.
(13) Vriens, J.; Owsianik, G.; Janssens, A.; Voets, T.; Nilius, B. Deter-
minants of 4R-Phorbol Sensitivity in Transmembrane Domains 3 and
4 of the Cation Channel TRPV4. J. Biol. Chem. 2007, 282, 12796–
12803.
(25) Appendino, G. Capsaicin and Capsainoids. In Modern Alkaloids:
Structure Isolation, Synthesis, and Biology; Fattorusso, E., Taglialatela-
Scafati, O., Eds.; Wiley: Weinheim, 2007; pp 75-112.
(26) Appendino, G.; Minassi, A.; Morello, A. S.; De Petrocellis, L.; Di
Marzo, V. N-Acylvanillamides: Development of an Expeditious
Synthesis and Discovery of New Acyl Templates for Powerful
Activation of the Vanilloid Receptor. J. Med. Chem. 2002, 45, 3739–
3745.
(27) Strotmann, R.; Schultz, G.; Plant, T. D. Ca2+-Dependent Potentiation
of the Nonselective Cation Channel TRPV4 Is Mediated by a
C-Terminal Calmodulin Binding Site. J. Biol. Chem. 2003, 278,
26541–26549.
(28) Gevaert, T.; Vriens, J.; Segal, A.; Everaerts, W.; Roskams, T.; Talavera,
K.; Owsianik, G.; Liedtke, W.; Daelemans, D.; Dewachter, I.; Van
Leuven, F.; Voets, T.; De Ridder, D.; Nilius, B. Deletion of the
Transient Receptor Potential Cation Channel TRPV4 Impairs Murine
Bladder Voiding. J. Clin. InVest. 2007, 117, 3453–3462.
(29) Nilius, B.; Prenen, J.; Wissenbach, U.; Bo¨dding, M.; Droogmans, G.
Differential Activation of the Volume-Sensitive Cation Channel TRP12
(OTRPC4) and Volume-Regulated Anion Currents in HEK-293 Cells.
Pflu¨gers Arch. 2001, 443, 227–233.
JM9001007