Importance of non-selective cation channel TRPV4 interaction with cytoskeleton and their reciprocal regulations in cultured cells.

Importance of Non-Selective Cation Channel TRPV4
Interaction with Cytoskeleton and Their Reciprocal
Regulations in Cultured Cells
Chandan Goswami1,2*, Julia Kuhn1, Paul A. Heppenstall3¤, Tim Hucho1
1 Signal Transduction in Pain and Mental Retardation, Department for Molecular Human Genetics Max Planck Institute for Molecular Genetics, Berlin, Germany, 2National
Institute of Science Education and Research, Bhubaneswar, India, 3 Klinik fu¨r Anaesthesiologie und Operative Intensivmedizin, Charite´ Universita¨tsmedizin Berlin, Campus
Benjamin Franklin, Berlin, Germany
Abstract
Background: TRPV4 and the cellular cytoskeleton have each been reported to influence cellular mechanosensitive
processes as well as the development of mechanical hyperalgesia. If and how TRPV4 interacts with the microtubule and
actin cytoskeleton at a molecular and functional level is not known.
Methodology and Principal Findings: We investigated the interaction of TRPV4 with cytoskeletal components
biochemically, cell biologically by observing morphological changes of DRG-neurons and DRG-neuron-derived F-11 cells,
as well as functionally with calcium imaging. We find that TRPV4 physically interacts with tubulin, actin and neurofilament
proteins as well as the nociceptive molecules PKCe and CamKII. The C-terminus of TRPV4 is sufficient for the direct
interaction with tubulin and actin, both with their soluble and their polymeric forms. Actin and tubulin compete for binding.
The interaction with TRPV4 stabilizes microtubules even under depolymerizing conditions in vitro. Accordingly, in cellular
systems TRPV4 colocalizes with actin and microtubules enriched structures at submembranous regions. Both expression
and activation of TRPV4 induces striking morphological changes affecting lamellipodial, filopodial, growth cone, and neurite
structures in non-neuronal cells, in DRG-neuron derived F11 cells, and also in IB4-positive DRG neurons. The functional
interaction of TRPV4 and the cytoskeleton is mutual as Taxol, a microtubule stabilizer, reduces the Ca2+-influx via TRPV4.
Conclusions and Significance: TRPV4 acts as a regulator for both, the microtubule and the actin. In turn, we describe that
microtubule dynamics are an important regulator of TRPV4 activity. TRPV4 forms a supra-molecular complex containing
cytoskeletal proteins and regulatory kinases. Thereby it can integrate signaling of various intracellular second messengers
and signaling cascades, as well as cytoskeletal dynamics. This study points out the existence of cross-talks between non-
selective cation channels and cytoskeleton at multiple levels. These cross talks may help us to understand the molecular
basis of the Taxol-induced neuropathic pain development commonly observed in cancer patients.
Citation: Goswami C, Kuhn J, Heppenstall PA, Hucho T (2010) Importance of Non-Selective Cation Channel TRPV4 Interaction with Cytoskeleton and Their
Reciprocal Regulations in Cultured Cells. PLoS ONE 5(7): e11654. doi:10.1371/journal.pone.0011654
Editor: Hiroaki Matsunami, Duke University, United States of America
Received December 3, 2009; Accepted June 15, 2010; Published July 19, 2010
Copyright: � 2010 Goswami et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Funding from Max Planck Institute for Molecular Genetics (Berlin, Germany) is gratefully acknowledged. CG is now supported by National Institute of
Science Education and Research, India. The authors declare that the funders had no role in study design, data collection and analysis, decision to publish, or
preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: goswami@molgen.mpg.de
¤ Current address: EMBL Monterotondo, Adriano Buzzati-Traverso Campus, Monterotondo, Italy
Introduction
Transient receptor potential vanilloid sub type 4 (TRPV4) is a
member of TRP super family of Ca2+-permeable non-selective
cation channels. This polymodal receptor is involved in cellular
processes such as mechanosensation, osmosensation and thermo-
sensation [1–5]. Some of these sensory functions are well conserved
in different species. For example, mammalian TRPV4 can rescue
mechanosensitive defects observed in C.elegans OSM-9 mutants [5].
In higher organisms TRPV4 is endogenously expressed in
nociceptive dorsal root ganglion (DRG) neurons but also in many
non-neuronal tissues and cells such as skin, kidney corneal epithelial
cells [6], cerebral microvascular endothelial cells [7], cortical
astrocytes [8], tracheal epithelial cells [9], keratinocyte cell lines [10]
and in other cells. The widespread distribution of TRPV4 is
indicative of its involvement in various physiological functions.
Indeed, TRPV4 is of importance in shear stress-induced vasodila-
tion [11] as well as in auditory functions [12–13]. Recently TRPV4
gained importance as it has been linked with the development of
different pathophysiological conditions such as neuropathic pain,
cystic fibrosis, brachyolmia and cancer [14–18].
From several reports, the involvement of cytoskeleton can be
correlated with the localization and function of TRPV4. For
example, TRPV4 is found in structures like cilia in various tissues
and cells [9,19–21] and in lamellipodia, where it regulates the
dynamics of cytoskeleton [22–23]. Many cellular functions
involving TRPV4 are known to require active participation of
the cytoskeleton. For example, TRPV4 activity is central to
PLoS ONE | www.plosone.org 1 July 2010 | Volume 5 | Issue 7 | e11654

cytoskeleton-dependent/mediated regulatory volume decrease of
cells [6,10,24], a process where actin-binding proteins contribute
to cell volume regulatory ion channel activation [24–26]. In
addition, TRPV4 has a conserved role in mechanotransduction, a
complex process that involves both actin and microtubule
cytoskeletal components [27–29]. The interplay of TRPV4 with
microtubule cytoskeleton also appears on a behavioural level,
where alteration of microtubule dynamics by Taxol induces a
TRPV4-dependent painful peripheral neuropathy [30]. While all
these cellular and behavioural studies strongly suggest that TRPV4
shares a functional relation with the cytoskeleton, so far a direct
link of TRPV4 with the cytoskeleton has not been demonstrated.
Thus, a molecular mechanism for the role of TRPV4 and the
cytoskeleton in pain, mechanosensation as well as other cellular
functions remains elusive.
Recently, we have established a functional interplay between
TRPV1, a close homologue of TRPV4, and the microtubule
cytoskeleton [31–35]. We demonstrated the physical interaction of
microtubule cytoskeleton with TRPV1 via two novel tubulin-
binding motifs [36–37]. Based on our previous experiments done
on TRPV1 and the sequence homology between TRPV1 and
TRPV4, we predicted that TRPV4 might interact with tubulin via
its C-terminal domain. Therefore, in this work we set out to
explore if TRPV4 physically and functionally interacts with actin
and microtubule cytoskeletal components.
Results
TRPV4 interacts with endogenous actin and tubulin
In order to test if TRPV4 interacts with cytoskeletal proteins like
tubulin and actin, we performed co-immunoprecipitation experi-
ments with affinity purified TRPV4 antibodies. CHO-KI-TRPV4
stable cell lines were used, which express low levels of TRPV4. In
immunoblot analysis, we observed that TRPV4 antibodies
precipitated TRPV4 together with actin and tubulin proteins
(Fig. 1a). Presence of tubulin and actin was not observed when a
similar co-immunoprecipitation was performed from the same cell
extract using an antibody, which was not raised against TRPV4. To
confirm further that the tubulin interaction is occurring even in
endogenous tissues, we isolate DRG neurons from rat and
performed similar co-immunoprecipitation experiments with affin-
ity purified TRPV4 antibodies. We observed that tubulin co-
immunoprecipitated with TRPV4 even from DRG neurons (1b).
The C-terminus of TRPV4 is sufficient for interaction with
actin and tubulin
To identify, which part of the TRPV4 interacts with actin and/
or tubulin proteins, we performed a pull down experiment using
maltose-binding-protein (MBP)-fused to the N- and C-termini of
TRPV4. According to our prediction [37], at least one tubulin-
binding site is located within the C-terminus of TRPV4 (Fig. S1)
Figure 1. Interaction of soluble tubulin and actin with TRPV4. a. Co-immunoprecipitation of actin and tubulin with TRPV4. Cell extracts from
CHO-KI cells stably expressing TRPV4 (lane 1) was immunoprecipitated by TRPV4 antibody (lane 2) or by a non-specific antibody (lane 3). Blots were
probed for TRPV4 (left side), tubulin (middle) and actin (right side). b. Co-immunoprecipitation of tubulin with TRPV4. Extracts from DRG (lane 1) was
immunoprecipitated by TRPV4 antibody (lane 2) or by a non-specific antibody (lane 3). Blots were probed for TRPV4 (upper panel) and tubulin (lower
side). c. MBP-TRPV4-Ct (lane 2-3) but not MBP-LacZ (lane 4–5) forms specific complexes when incubated with mammalian brain extract (lane 1), both
in presence (lane 2 and 4) or absence (lane 3 and 5) of Ca2+ (1 mM). Presence of PKCe, actin and tubulin are observed only in lane 2 and 3.
Neurofilament in the pull down samples is visible only after exposing for a prolonged time. Presence of CamKII is noted only in the presence of Ca2+
(lane 2). Note that the amount of MBP-LacZ used, as a negative control for the pull down experiment is much more than MBP-TRPV4-Ct. d. Tubulin
interacts with TRPV4-Ct directly. MBP-LacZ (lane 1–2) or MBP-TRPV4-Ct (lane 3–4) was incubated with buffer only (lane 1 and 3) or with purified
tubulin (lane 2 and 4). Pulled down samples were probed for different isotype-specific and different post-translationally modified tubulins. e. Actin
interacts directly with TRPV4-Ct. MBP-TRPV4-Ct (lane 1–2) or MBP-LacZ (lane 3–4) was incubated with purified actin (lane 1–4) either in the presence
(lane 1-and 3) or absence (lane 2 and 4) of Ca2+ (1 mM) and subsequently probed for bound actin. f. Soluble tubulin and actin competes for the C-
terminal cytoplasmic fragment of TRPV4. MBP-TRPV4-Ct was incubated with only tubulin (lane 1), with only actin (lane 2), or both tubulin and actin in
a sequential manner (lane 3–4). Prior incubation of tubulin inhibits further binding of actin (lane 3). Similarly, prior incubation of actin significantly
reduces the further binding of tubulin (lane 4).
doi:10.1371/journal.pone.0011654.g001
TRPV4 Regulates Cytoskeleton
PLoS ONE | www.plosone.org 2 July 2010 | Volume 5 | Issue 7 | e11654

and as the expression of the N-terminal cytoplasmic fragment
remained poor, we restricted our study to the C-terminus of
TRPV4 only. Pull down was performed from adult porcine brain
homogenate as well as from F11 cell lysate, a fusion-cell of rat
DRG neurons and mouse neuroblastoma cells [38]. In immuno-
blot analysis, we observed MBP-TRPV4-C-terminus fusion
protein (MBP-TRPV4-Ct) but not the control fusion protein
MBP-LacZ to pull down soluble tubulin and actin (Fig. 1c, S2).
When probed for the presence of another cytoskeletal element,
namely soluble neurofilament proteins, only minimal amounts of
NF116 kDa and no NF200 kDa were detected.
As activation of TRPV4 results in high Ca2+-influx, we tested if
higher concentration of Ca2+ could modulate these interactions.
However, we observed that tubulin and actin interaction with
TRPV4 do not depend on Ca2+ (Fig. 1c, S2).
TRPV4-Ct forms a Ca2+-sensitive supra-molecular
signaling complex made of actin, tubulin as well as
nociceptive signaling components PKCe and CamKII
Previously we have demonstrated that mechanical hyperalgesia
induced agonists of novel estrogen receptor GPR30 strongly
correlates with the translocation of PKCe in IB4 (+) neurons [39].
Moreover, mechanical hyperalgesia can be induced by activation of
the epsilon isoform of PKC (PKCe), a well-described pro-
nociceptive signaling molecule [39–43]. In addition, calmodulin,
and CamKII function has been linked with the chronic inflamma-
tory pain [44–46]. Thus, we tested if the TRPV4-actin/tubulin
complex also contains PKCe and/or CamKII. We found CamKII
to be present in the eluates of the pull-down material both from
soluble brain and F11 cell extract, but only in the presence of Ca2+
(Fig. 1c, Fig. S1). In addition, we also detected PKCe in the MBP-
TRPV4-Ct complex, both in presence and absence of Ca2+.
Presence of PKCe or CamKII was not observed with the control
protein MBP-LacZ. Thus, using two different biological sources,
our results indicate that TRPV4-Ct can form supra-molecular
complexes consisting of structural and signaling proteins.
MBP-TRPV4-Ct interacts directly with soluble tubulin and
actin
To test if TRPV4-Ct interacts directly with tubulin and actin, we
performed pull down experiments with the purified MBP-TRPV4-
Ct, tubulin, and actin. As expected, MBP-TRPV4-Ct but not MBP-
LacZ pulls down tubulin (Fig. 1d). Tubulin is subjected to different
types of post-translational modification, which regulate the process
of microtubule stabilization, destabilization and maturation. Thus,
we tested for the presence of various post-translationally modified
tubulins in the complex formed with MBP-TRPV4-Ct. Indeed we
found a large number of these post-translationally modified tubulins
and neuron-specific b-III tubulin (Fig. 1d).
Next we tested if also soluble actin interacts directly with MBP-
TRPV4-Ct. In immunoblot analysis, we observed that the MBP-
TRPV4-Ct pulls down purified actin, while under the same
conditions MBP-LacZ was unable to pull down any actin (Fig. 1e).
Again, we observed no influence of Ca2+ on this interaction.
Tubulin and actin compete for binding to MBP-TRPV4-Ct
To understand if tubulin binding affects actin binding and vice
versa, we performed a competition experiment between soluble
tubulin and soluble actin for the binding to MBP-TRPV4-Ct. Soluble
tubulin and actin were added to MBP-TRPV4-Ct sequentially before
being analysed for their binding to TRPV4 (Fig. 1f). As control either
soluble tubulin or soluble actin were used alone and washed in the
same manner. We observed that the amount of bound tubulin was
strongly reduced if MBP-TRPV4-Ct was initially incubated with actin
(Fig. 1f). Inversely, the amount of bound actin is very low if MBP-
TRPV4-Ct is initially incubated with tubulin. These results indicate
that both actin and tubulin compete for binding to MBP-TRPV4-Ct.
MBP-TRPV4-Ct interacts with polymerized actin and
tubulin filaments and favours formation of stable
microtubules
Next we addressed whether TRPV4 can also bind to polymerized
filaments. We observed that MBP-TRPV4-Ct co-sedimented with
polymerized actin and thus appeared in the pellet fraction (Fig. 2a).
Under the same conditions, MBP only showed no interaction with
polymerized actin filaments.
Here we probed if the MBP-TRPV4-Ct can also interact with
polymerized microtubules. We observed that MBP-TRPV4-Ct but
not MBP alone co-sedimented with Taxol-stabilized polymerized
microtubules (Fig. 2b, left side). In addition, we observed that if
MBP-TRPV4-Ct was added to saturated tubulin dimer solution
during GTP-induced microtubule formation, co-sedimentation of
MBP-TRPV4-Ct with polymerized microtubules was also observed
(Fig. 2b, right side). Again, only MBP, the control protein failed to
bind polymerized microtubules. These results confirm that MBP-
TRPV4-Ct directly interacts with filamentous microtubules.
We tested if MBP-TRPV4-Ct can change the physico-chemical
properties of the microtubules. For that purpose, we analysed the
stability of the microtubules by comparing the amount of
microtubules formed under depolymerization conditions, both in
the presence or absence of MBP-TRPV4-Ct. We observed that the
presence of MBP-TRPV4-Ct favours microtubule formation even
in presence of Nocodazole, Ca2+, or both (Fig. 2c). In contrast, in
absence of MBP-TRPV4-Ct much lower amount of microtubules
was formed. MBP as a control protein does not interact with the
polymerised microtubules and thus fails to provide stability to the
microtubules. This result suggests a stabilization effect of MBP-
TRPV4-Ct on microtubules.
TRPV4 localizes to actin- and microtubule-enriched
regions in F11 cell
Based on a RT-PCR analysis demonstrating the amplification of
small mRNA fragment, endogenous expression of TRPV4 in F11
cell has been proposed [47]. However, by immunofluorescence
analysis two different polyclonal antibodies which recognise the C-
terminal region of TRPV4, we could not observe any specific
endogenous expression of TRPV4 in F11 cells. To investigate if
the biochemical interaction under in vitro conditions also occurs in
vivo, we expressed TRPV4 in F11 cells as well as in other non-
neuronal cells and performed co-localization experiments. The
purpose of TRPV4 overexpression in F-11 cells is to mimic the
physiological situation in DRG neurons.
Using fluorescent labelled Phalloidin, we observed co-localiza-
tion of TRPV4 with actin at various actin cytoskeleton enriched
regions, such as the actin ribs of the cortical membranous regions,
and at filopodial and lamellipodial structures (Fig. S3).
As some of these structures are very dynamic and sensitive to
environmental changes such as temperature drops or media
composition, we attempted to confirm that the observed co-localization
in these structures was not a fixation artefact. Thus, we performed live
cell imaging by expressing TRPV4-GFP and RFP-actin in F11 cells
and we observed co-localization at filopodia, at lamellipodia, and also
at cortical actin-rich structures (Fig. 3a-c).We also noted co-localization
at actin-rich structures, which resemble focal adhesion points (Fig. 3b).
This is in agreement with the fact that focal adhesion points are
important for cellular mechanosensory functions [48].
TRPV4 Regulates Cytoskeleton
PLoS ONE | www.plosone.org 3 July 2010 | Volume 5 | Issue 7 | e11654

Next, we addressed if TRPV4 similarly co-localizes with
microtubules. We observed co-localization between TRPV4 and
microtubules in fixed F11 cells all along neurite-like structures. Co-
localization was observed at the plasma membrane and at
membrane ruffles that are enriched in TRPV4 (Fig. 3d-e). In
addition, we noted the presence of numerous microtubule ends at
the submembranous regions. Often these microtubule ends
extended to the plasma membrane and seemed to be stabilized at
the membranous regions containing TRPV4. Co-localization of
tubulin and TRPV4 was also observed in filopodial structures that
were developed from growth cones, from neurite-like structures as
well as from cell bodies. This kind of submembranous tubulin
accumulation was not observed in non-transfected cells (Fig. 3f),
implying that this effect is primarily due to the presence of TRPV4.
Activation of TRPV4 results in fast retraction of growth
cones in F11 cell
TRPV4 localizes to growth cones when expressed in F11 cells.
So we tested if TRPV4 activation can alter the morphology and
movement of growth cones. For that purpose we expressed
TRPV4-GFP along with RFP-actin and performed live cell
imaging. We observed that in response to 4a-phorbol-didecanoate
(4aPDD, 1 mM), a TRPV4-specific agonist [22], neurites from
TRPV4-GFP expressing F11 cells show rapid change in its
morphology and induce multiple varicosities (Fig. S4). In response
to 4aPDD, TRPV4-GFP containing growth cones retract quickly
(Fig. 4a). Under the same conditions growth cones developing
from non-transfected cells do not show any retraction, thus
assuring specificity of the pharmacological treatment. These results
strongly suggest that TRPV4 can regulate the growth cone
motility.
Long-term exposure to a TRPV4 agonist restricts neurite
outgrowth in a subset of primary DRG neurons
To validate the effect of TRPV4 activation on growth cones in a
system with endogenous TRPV4 expression, we used primary
neurons and tested the effect of TRPV4 activation. DRG neurons
from adult male rat were cultured for 5 days and treated with
4aPDD (1 mM) for short-term (20 minutes). Live cell-imaging
revealed indications of slow retraction of neurites in response to
Figure 2. TRPV4 interacts directly with polymerized actin and microtubule filaments. MBP and MBP-TRPV4-Ct were centrifuged at
70000 g/30 min/4uC and only soluble proteins present in the supernatant were used for all co-sedimentation experiments. a. MBP-TRPV4-Ct co-
sediments with polymerized actin filaments. Actin was polymerized either in presence of MBP-TRPV4-Ct (lane 1 and 4), in presence of MBP only (lane
2 and 5) or in buffer only (lane 3 and 6). Polymerized actin filaments and associated proteins were isolated from remaining soluble actin and unbound
proteins by centrifugal separation of pellets (P, lane 1–3) from corresponding supernatants (S, lane 4–6). The entire amount of MBP remains in the
supernatant (lane 5) while a significant amount of MBP-TRPV4-Ct appears in the pellet (lane 1). Arrows indicate the position of respective proteins. b.
MBP-TRPV4-Ct co-sediments with microtubules. Taxol-stabilized microtubules (left panel) were incubated with MBP (lane 1–2), MBP-TRPV4-Ct (lane 3–
4) or with buffer only (lane 5–6) followed by the centrifugal separation of pellet (P) consisting MT and bound proteins from supernatant (S) consisting
of soluble tubulin and other unbound proteins (left side panel). In right side panel, soluble tubulin and GTP was incubated with MBP (lane 1–2), MBP-
TRPV4-Ct (lane 3–4) or buffer only (lane 5–6) followed by separation of pellet (P) and supernatant (S). Note the specific presence of MBP-TRPV4-Ct in
the pellet in both cases (in lane 4). c. MBP-TRPV4-Ct stabilizes microtubules against depolymerizing factors. Microtubules was formed form soluble
tubulin in buffer (lane 1), along with MBP (lane 2) or along with MBP-TRPV4-Ct (lane 3) in control condition (left most panel), in presence of
Nocodazole (middle left panel), in presence of Ca2+ (middle right side) or in presence of both Nocodazole and Ca2+ (right most). Microtubules and
bound proteins present in the pellet fraction (P) were isolated from unpolymerized tubulin and unbound proteins remaining in the supernatant (S)
by centrifugal separation. Note the enhancement of polymerized microtubules (represented by tubulin present in lane 3, P fraction in every
conditions) due to the presence of MBP-TRPV4-Ct.
doi:10.1371/journal.pone.0011654.g002
TRPV4 Regulates Cytoskeleton
PLoS ONE | www.plosone.org 4 July 2010 | Volume 5 | Issue 7 | e11654

Figure 3. TRPV4 co-localizes with actin and microtubule cytoskeleton. a–c. Shown are the live-cell confocal images of F11 cells expressing
TRV4-GFP (green) and RFP-actin (red). Presences of TRPV4-GFP specifically in actin-enriched structures are shown. a. Enlarged view of lamellipodia
and at the tip of the actin filaments are shown. b–c. Enlarged view of focal adhesion point-like structures (b) and cell cortex with actin ribs (c) are
shown. Arrows indicate the localization of TRPV4-GFP at filopodial tips. d–f. TRPV4 co-localizes with microtubule cytoskeleton. Shown are the
confocal images of F11 cells immunostained for TRPV4 (green) and tyrosinated tubulin (red). Arrows indicate presence and acumulation of
microtubules in thin filopodial structures (d, upper panel) and thin lamellipodial structures (e, middle panel). The status of the microtubules in the
non-transfected cells are shown in below (f, lower panel).
doi:10.1371/journal.pone.0011654.g003
TRPV4 Regulates Cytoskeleton
PLoS ONE | www.plosone.org 5 July 2010 | Volume 5 | Issue 7 | e11654