TRPV1–tubulin complex: involvement of membrane tubulin in the regulation of chemotherapy-induced peripheral neuropathy


Address correspondence and reprint requests to Chandan Goswami, National Institute of Science Education and Research, Sachivalaya Marg, Bhubaneswar, Orissa, India.



Existence of microtubule cytoskeleton at the membrane and submembranous regions, referred as ‘membrane tubulin’ has remained controversial for a long time. Since we reported physical and functional interaction of Transient Receptor Potential Vanilloid Sub Type 1 (TRPV1) with microtubules and linked the importance of TRPV1-tubulin complex in the context of chemotherapy-induced peripheral neuropathy, a few more reports have characterized this interaction in in vitro and in in vivo condition. However, the cross-talk between TRPs with microtubule cytoskeleton, and the complex feedback regulations are not well understood. Sequence analysis suggests that other than TRPV1, few TRPs can potentially interact with microtubules. The microtubule interaction with TRPs has evolutionary origin and has a functional significance. Biochemical evidence, Fluorescence Resonance Energy Transfer analysis along with correlation spectroscopy and fluorescence anisotropy measurements have confirmed that TRPV1 interacts with microtubules in live cell and this interaction has regulatory roles. Apart from the transport of TRPs and maintaining the cellular structure, microtubules regulate signaling and functionality of TRPs at the single channel level. Thus, TRPV1-tubulin interaction sets a stage where concept and parameters of ‘membrane tubulin’ can be tested in more details. In this review, I critically analyze the advancements made in biochemical, pharmacological, behavioral as well as cell-biological observations and summarize the limitations that need to be overcome in the future.

Abbreviations used

Electron Microscopy


Fluorescence Resonance Energy Transfer


Sub Membranous Region


Transient Receptor Potential


Transient Receptor Potential Vanilloid Sub Type 1


Transient Receptor Potential Vanilloid Sub Type 4

Since discovery, αβ-tubulin dimer has been mainly represented as the major constituent of microtubule cytoskeleton present in the cytoplasm. However, more than 40 years have passed since biochemical experiments demonstrated that tubulin is also a bona fida constituent of biological membrane. Indeed, biochemical fractions isolated from several biological systems such as mitochondrial membrane, synaptic membrane, plasma membrane reveal presence of tubulin there. During the 1970s, Feit and Barondes used membranes from mouse brain and demonstrated for the first time that tubulin is present in the membrane fraction (Feit and Barondes 1970). In 1974, Blitz and Fine prepared whole synaptosomal membrane and analyzed the proteins present in this membrane fraction (Blitz and Fine 1974). By performing two-dimensional gel electrophoresis, they have identified a protein of 55 kDa to be present in this membrane fraction and the peptides of this protein matched well with the peptides derived from purified tubulin. As this 55-kDa protein is present in the membrane fraction, it suggested that tubulin can be a part of synaptosomal membranes (Blitz and Fine 1974). Interestingly, it was demonstrated that tubulin is generally enriched in the membrane fraction as the amount of tubulin present in the soluble synaptosomal fraction was much less (Blitz and Fine 1974). During 1975, Bhattacharyya and Wolff also provided evidence for the presence of tubulin in membrane by using two different tissue systems, namely membrane from brain and thyroid (Bhattacharyya and Wolff 1975, 1976). They demonstrated that the binding constants for colchicine and vinblastin to membrane-bound tubulin or cytoplasmic tubulin are same. They showed that tubulin interaction with membrane is not an artifact, as less than 0.1% of the labeled tubulin remains associated with membrane when added during the preparation. Subsequently, several laboratories reported the occurrence of tubulin in biochemically enriched fractions derived from plasma membrane, synaptic membranes as well as from post-synaptic density (Kornguth and Sunderland 1975; Walters and Matus 1975; Therien and Mushynski 1976; Zenner and Pfeuffer 1976; Matus and Taff-Jones 1978; Kelly and Cotman 1981; Carlin et al. 1982). However, these studies do not indicate if tubulin interaction with the membrane fraction is direct or because of some other membrane-bound proteins such as ion channels, receptors, etc.

Subsequently, there were few in vitro studies which demonstrated that tubulin can interact with membrane directly. In 1979, Caron and Berlin demonstrated that tubulin can be adsorbed directly on the artificially prepared liposomes (Caron and Berlin 1979). This liposome-adsorbed tubulin forms extensive intermolecular disulfide bridges and remains inert to reducing agents in the aqueous medium, suggesting a direct tubulin-lipid interaction (Caron and Berlin 1987). In the similar context, few biochemical studies have demonstrated the interaction of tubulin with membrane fraction can be indirect too, that is via few transmembrane and membrane-associated proteins (discussed later). In the same context, proteomic studies have detected tubulin in the synaptic structures including the post-synaptic density fractions (Yoshimura et al. 2000, 2004). This not only leads to the emergence of novel functionalities of submembranous microtubule cytoskeleton but also strengthens the concept of ‘membrane tubulin’ (Bernier-Valentin et al. 1983; Stephens 1986; Ravindra 1997).

In spite of these handfuls of biochemical experiments, the presence of tubulin in the membranous fraction remained uncertain for long. Especially, membrane association of tubulin faced high criticism mainly because of possible contamination of isolated membranes by soluble and denatured tubulin during preparation. For example, the in vitro association of tubulin to liposomes was criticized as ‘preparation artifacts’ (Melchior et al. 1980). However, subsequent reports suggested that the association of tubulin with membrane is specific. For example, fractionation studies with different detergents from synaptic vesicular membranes confirmed that α-tubulin acts as an integral membrane protein whereas β-tubulin acts as peripheral protein (Zisapel et al. 1980). In the same context, in vitro interaction of tubulin as well as tubulin–colchicine complex with brain microsomal membrane was also shown (Bernier-Valentin et al. 1983; Rodríguez and Barra 1983). Reconstruction experiments have confirmed interaction of purified tubulin with purified ciliary membrane (Stephens 1983). All together these reports were only suggestive for a potential importance of membrane tubulin.

Interestingly, much later studies have confirmed the presence of tubulin and components of the microtubule cytoskeleton-like CLIPR-59 (Lallemand-Breitenbach et al. 2004), tubulin-specific chaperone A, and KIF13 (Li et al. 2004) in submembraneous compartments such as detergent-resistant lipid rafts. A large number of reports suggest direct interaction of tubulin with several transmembrane proteins as diverse as ion channels, G-protein coupled receptors, ion pumps, T cell receptors, and gap junction proteins (Table 1). The number of membrane-associated peripheral proteins interacting with different microtubule cytoskeletal components is increasing rapidly. In many cases, these transmembrane and peripheral membrane proteins themselves contain cytoskeletal-binding motifs which are conserved throughout evolution.

Table 1. Example of membrane proteins physically interacting with microtubule cytoskeleton
Polycystin-2 (PC2)SignalingMontalbetti et al. (2007) and Li et al. (2006)
TRPV1MT dynamics, signalingGoswami et al. (2004, 2007b)
TRPV4MT dynamics, signalingGoswami et al. (2010) and Huai et al. (2012)
TRPC1TranslocationBollimuntha et al. (2005)
TRPC5SignalingGoel et al. (2005)
TRPC6SignalingGoel et al. (2005)
NMDAMT dynamics, synaptic MT reorganizationvan Rossum et al. (1999)
VDACApoptosis, mitochondrial signaling,Carre et al. (2002), Rostovtseva et al. (2008), Rostovtseva and Bezrukov (2008) and Rostovtseva et al. (2012)
Na+, K+-ATPase pumpSignaling, development of brain?Vladimirova et al. (2002) and Casale et al. (2001)
MGluR1SignalingCiruela et al. (1999) and Ciruela and McIlhinney (2001)
MGluR7SignalingSaugstad et al. (2002)
G-Protein coupled receptorsSignalingPopova et al. (1997), Roychowdhury and Rasenick (1997), Roychowdhury et al. (1999) and Sarma et al. (2003)
P2X(2)Transport, signaling?Gendreau et al. (2003) and Guimaraes (2008)
Slo1TransportMa et al. (2007)
RasSignalingThissen et al. (1997) and Chen et al. (2000)
Estrogen Receptor aSignalingAzuma et al. (2004)
H+-ATPaseGlucose uptake in yeastCampetelli et al. (2005)
GRK2MT dynamics, signaling.Haga et al. (1998)
GABAASignalingItem and Sieghart (1994)
CD2T-cell activationOffringa and Bierer (1993)
STIM1Calcium signalingGrigoriev et al. (2008)
CLIC (p64-related)Signaling, transportSuginta et al. (2001)
heag2 channelSignaling, ion conductivityBracey et al. (2008) and Wray (2009)
AsxMT dynamicsKramer and Hawley (2003)
Connexin 43Cell–cell communicationGiepmans et al. (2001)
CARCell migrationFok et al. (2007)
PMCASignaling, brain function?Monesterolo et al. (2008)

Important functional evidence for membrane tubulin was provided from genetic studies conducted on C. elegans. Chalfie and coworkers demonstrated that mec12, and mec7, required for touch sensitivity encode α- and β-tubulin, respectively (Savage et al. 1989; Gu et al. 1996; Fukushige et al. 1999; Ernstrom and Chalfie 2002). They further demonstrated that these two genes are not needed for the normal outgrowth of the sensory cells, as loss of mec-7 or mec-12 activity and the concomitant loss of these tubulins still allowed for normal neuronal development. Further genetic studies confirmed an interaction between mec12 and mec2 (encodes an integral membrane protein), which regulate the mechanotransduction (Huang et al. 1995). Notably, the distal ends of the microtubules in all these mutants are distributed normally and remain in a position where they can contact the plasma membrane. Indeed, the electron microscopy (EM) studies confirmed that the microtubule ends in these mutants are associated with electron-diffuse material that appears to contact the membrane. These genetic studies combined with EM images therefore strengthen the importance of membrane tubulin for highly specific functions that are complex enough to understand at present. Taken together, all these recent evidence helped establish that membrane tubulin has specific functions and is of scientific interest.

What is ‘membrane tubulin’?

In general terms, the fraction of tubulin, which is associated with plasma membrane, other subcell organelle membrane (like mitochondrial or golgi membrane) and many other membranes like synaptic membrane, ciliary membrane are collectively known as ‘membrane tubulin’ (detailed historic background is described in Stephens 1986; Wolff 2009). Several techniques, namely one-dimensional and two-dimensional sodium dodecyl sulfate–polyacrylamide gel electrophoresis, immunolabeling, peptide sequencing of proteins present in the membrane fraction, binding of radiolabeled tubulin-specific drugs, and competition with artificially added tubulin, immunofluorescence analysis, have confirmed the existence of membrane tubulin. Interaction of tubulin to the biological membranes is mostly indirect but can also be direct (Zisapel et al. 1980). ‘Membrane tubulin’ has a similar molecular weight and mobility on sodium dodecyl sulfate–polyacrylamide gel electrophoresis with respect to cytosolic tubulin. But, membrane tubulin differs from cytoplasmic forms by isoelectric point, non-polar amino acid substitutions, lacks C-terminal tyrosine residues, carbohydrate content (membrane tubulin has more glycosylation), selective ability to associate and reassociate with lipids, and their ability to incorporate into the existing microtubules (Stephens 1986; Wolff 2009). Membrane tubulin has a poor capacity to form microtubules in presence of lipids, but individual dimers purified from membranes can copolymerize with cytosolic tubulin (Bhattacharyya and Wolff 1976). Membrane tubulin has increased heat stability because of the membrane environment as solubilization of membrane by NP-40, a non-ionic detergent, abolishes the heat stability (Bhattacharyya and Wolff 1975).

Membrane tubulin has different solubility against different detergents. Solubility of membrane tubulin is maximum in the presence of Nonidet P40, but a fraction of membrane tubulin is resistant against Triton X100 (another non-ionic detergent) (Feit and Barondes 1970; Bhattacharyya and Wolff 1976). According to some estimates, the membrane tubulin represents 0.2–0.4% of total detergent-soluble membrane proteins and in saturated conditions 5–10% of the total membrane protein (by weight) (Bernier-Valentin et al. 1983). It is highly heterogeneous in nature as it contains several isotypes of α- and β-tubulin as well as their different post-translationally modified forms (Stephens 1986). Among different post-translational forms, acetylated- and palmitylated-tubulin, which have better affinity for a lipophilic environment, are predominant. The membrane tubulin is mostly susceptible to proteases, but resistant against some proteases like cathepsin-D (Bracco et al. 1982; Stephens 1986). It is also accessible to antibodies and tubulin-binding pharmacological agents (Lagnado et al. 1971; Stephens 1986). The interaction of tubulin with membrane fraction is fast, reversible, saturable, and non-competitive with other acidic proteins like actin and serum albumin (Bernier-Valentin et al. 1983; Stephens 1986). In different systems, the interaction of tubulin with membrane is a time- and temperature-dependent events (Lagnado et al. 1971; Stephens 1986).

The apparent affinity constant measured for tubulin – membrane interaction is 1.5–3 x 107 M−1 (Bernier-Valentin et al. 1983). Therefore, the tubulin interaction with membrane can be observed at physiological concentration within cells. Notably, this affinity constant is much lower than the critical concentration of microtubule polymerization observed in most cases. The association of tubulin with the membrane is increased with the temperature and dissociation is increased with lowering the temperature (at 0°C) (Bhattacharyya and Wolff 1976; Bernier-Valentin et al. 1983). However, some fraction of tubulin remains associated with membrane even after prolonged incubation at low temperature (Lagnado et al. 1971). The tubulin – membrane interaction cannot be inhibited by low ionic strength or by presence of microtubule-depolymerizing drugs like nocodazole or colchicine (Stephens 1986). But, phosphate buffer is known to inhibit tubulin interaction with membrane (Bernier-Valentin et al. 1983). Taken together, all these aspects suggest that membrane tubulin have distinct characteristics than rest of the cytoplasmic tubulin. Therefore, membrane tubulin can be instrumental for regulation of membrane proteins in a complex manner.

TRPV1–tubulin complex: functional example of membrane tubulin

During 2004, for the first time we reported that TRPV1 interacts physically with tubulin dimer as well as with polymerized microtubules (Goswami et al. 2004). Subsequently, in the last few years, few groups have demonstrated the cross-talk of TRP channels with microtubule cytoskeleton at various levels (Bollimuntha et al. 2005; Goel et al. 2005; Goswami et al. 2006; Li et al. 2006; Montalbetti et al. 2007; Clark et al. 2008; Goswami and Hucho 2008a,b; Laínez et al. 2010; Huai et al. 2012; Storti et al. 2012). Using biochemical experiments, we established the interaction of TRPV1 with tubulin dimer (Goswami et al. 2004, 2007a,b). Interestingly, this interaction has been mapped to the two short sequences that are able to interact with tubulin dimer independently. However, a recent report also demonstrated that the N-terminus of TRPV1 is also able to bind tubulin independently (Laínez et al. 2010). These results strongly suggest that the TRPV1 may have multiple tubulin-binding sites. In the same notion, Storti et al. have demonstrated that microtubule cytoskeleton helps to form the TRPV1 tetramer at the membrane (Storti et al. 2012). At present, it may seem speculative, but it is possible that tubulin interaction with TRPV1 is dynamic in nature and the availability of the contact sites change depending on the conformational change of the channel. This in turn can regulate the opening – closing cycle of TRPV1. Such a possibility has recently been demonstrated for TRPP channels and VDAC where cytoskeletal components such as actin and tubulin regulate the ion channel functions (Montalbetti et al. 2005, 2007; Li et al. 2006; Sheldon et al. 2011). However, it is unknown if tubulin regulates TRPV1 channel function that remains to be explored in future. Nevertheless, TRPV1–tubulin complex provides an excellent example where membrane tubulin plays an important function such as tetramer formation.

Evolutionary significance of TRPV1–Tubulin interaction

Previously, we have demonstrated that the C-terminus of TRPV1 can interact with tubulin dimers, particularly in the absence of GTP (Goswami et al. 2004, 2007b). In the same context, interaction of tubulin dimer with the N-terminal cytoplasmic region of TRPV1 has also been demonstrated (Laínez et al. 2010). Interestingly, the C-terminus of TRPV1 interacts with different tubulin dimmers as well as almost all types of post-translationally modified tubulins that are tested so far, such as tyrosinated tubulin, detyrosinated tubulin, polyglutamylated tubulin, acetylated tubulin, and phospho (Serine) tubulin (Goswami et al. 2007b). In addition, two short peptide sequences have been identified that independently interact with tubulin dimer as well as with polymerized microtubules (Goswami et al. 2007b). Interestingly, these two stretch sequences contain positively charged residues that are conserved throughout evolution (Sardar et al. 2012). These two stretch sequences from different species suggest that these regions of TRPV1 can form putative alpha helical structures projecting the positive-charged residues at one side of the helix. Such distribution of positive charges form ‘+XXX+XX++XX+XX+XXXXXX’ motif sequence where + can be any positively charged residue (mostly lysine or arginine) and X can be any amino acid except negatively charged residues and helix-breaking residues (Sardar et al. 2012). This motif can facilitate the interaction with tubulin dimer, especially the E-hook region of tubulin that contains stretches of negatively charged residues (Fig. 1). In this regard, it is important to mention that an in-depth biophysical characterization indicates that the C-terminal cytoplasmic domain of TRPV1 is mostly unstructured (Aguado-Llera et al. 2012). Therefore, tubulin interaction can provide stability to certain region, especially the interacting region.

Figure 1.

Distribution of positively charged residues that are possibly involved with tubulin interaction. (a) The distribution of positively charged residues located within TBS-1 and TBS-2 of human Transient Receptor Potential Vanilloid Sub Type 1 (TRPV1) are represented in a 360° circular wheel. (b) The potential helix formed by TBS-1 and TBS-2 from different species are superimposed and the distribution of positively charged residues are represented in a 360° circular wheel. The distributions of positively charged residues in both Tubulin-binding stretch sequences are evolutionary conserved. Residues lysine (K), arginine (R), and histidine (H) are represented by violet, green, and the indigo dots, respectively. Image adapted from Sardar et al. (2012).

Evolutionary conservation of tubulin-binding motif in TRPV1 therefore suggests that this region may play an important role in the channel function. The presence of this motif may also explain the basis of tubulin interaction observed for some other TRP channels. Indeed, we noted the presence of the critical features of this motif sequence in TRPV2, TRPV3, TRPV4, TRPV5, TRPV6, TRPC1, TRPC2, TRPC3, TRPC4, TRPC5, TRPC6, and TRPM8 (Goswami et al. 2010; Sardar et al. 2012). Among which, tubulin interaction with TRPV4, TRPC1, TRPC5, TRPC6, and TRPP channels have already been demonstrated experimentally (Bollimuntha et al. 2005; Goel et al. 2005; Montalbetti et al. 2005, 2007; Goswami et al. 2010; Huai et al. 2012; Shin et al. 2012).

Microtubule - protofilament - tubulin oligomer – tubulin dimer: which fits well?

Recently, a few reports have confirmed the cross-talk of TRPV1 with microtubule skeleton. However, there are few aspects of the studies which are apparently contradictory and need further attention. In this context, it is worth to mention that ‘microtubule filament’, ‘microtubule protofilament’, ‘tubulin oligomer’, ‘soluble tubulin dimer’, and ‘membrane tubulin’, all represent entities that differ at the level of architecture with different geometry, size, structural, biochemical, and biophysical properties. So far, the exact nature of physical interaction between TRPV1 and microtubule skeleton is not very clear. It is especially unclear if TRPV1 interacts with microtubule filament' or ‘microtubule protofilament’ or ‘soluble tubulin dimer’ or ‘membrane tubulin or all of these in a more context-dependent manner. In this aspect, relative sizes give a logical understanding of the possible mode of interaction (Fig. 2). Recently cryo EM structure of TRPV1 complex with lipid membrane has been solved (Moiseenkova-Bell et al. 2008). Even a computer-predicted three-dimensional structural model of TRPV1 has been developed that fits well with many of the parameters reported so far (Fernández-Ballester and Ferrer-Montiel 2008). On the basis of the available information, we tested if tubulin dimers, tubulin oligomers, microtubule protofilaments, or whole microtubules filaments can fit well to the TRPV1 tetramer with an assumption that the surface areas of the binding partners will be at the matching scale. The calculated surface area of one C-terminal globular region (residue 716–839) is around 81.7 nm2. Therefore, the area of the blunt ends of the microtubules (with 13 protofilament, 25 nm in diameter) are 490 nm2, which seems to be much larger for binding to the C-terminal region, especially at the two distinct regions which show microtubule binding (the narrow hook/groove-like structure). Important to note here is that the diameter of microtubule is larger even than the whole functional TRPV1 made of four homo-polypeptides (490 nm2 and 412.5 nm2, respectively) and are therefore unlikely to bind because of mismatch in shape (planner vs spherical). In contrast, αβ-tubulin dimers measure to the average size of 4.6 nm × 8.0 nm × 6.5 nm (Nogales et al. 1998). Thus, an individual tubulin dimer or tubulin unit present at the tip of single microtubule protofilaments (plus end of open, straight or frayed conditions) can fit into the groove formed by the TRPV1-Ct surface. Accordingly, in a random trial of 200 possible tubulin–TRPV1 interactions, 155 possibilities showed an interaction of tubulin dimers with TRPV1-Ct indicating a high binding probability at the level of tubulin dimer or at least at the protofilament level (Goswami et al. 2011).

Figure 2.

Comparative size of Transient Receptor Potential Vanilloid Sub Type 1 (TRPV1) and microtubule. The Cryo-EM of TRPV1-tetramer suggests a hanging bucket-like structure with a portion at the cytoplasmic side which has a size of nearly 100 Å (in length) and 100 Å (in height). The microtubule filaments have a cross-section of 25nm. Therefore, it is unlikely that the total microtubule as a blunt end can interact with TRPV1. The interaction is likely to be mediated by protofilaments and/or by tubulin dimer.

Using red fluorescent protein (RFP)-tagged α-tubulin and Fluorescence Resonance Energy Transfer (FRET) as an experimental system, Storti and colleagues have proposed that at membrane, TRPV1 preferably binds to the intact microtubules but not to the tubulin dimer (Storti et al. 2012). However, this result contradicts in part to our biochemical results described above which mainly demonstrated that TRPV1 interacts with soluble tubulin dimer as well as with polymerized microtubules. This difference seems to be mainly because of the choice of tubulin that has been used for FRET analysis (Storti et al. 2012). Ironically, Storti et al. have used alpha tubulin tagged with RFP for FRET analysis. Using cross-linking experiments previously, we have demonstrated that the C-terminus of TRPV1 interacts with αβ-tubulin dimer very fast (within a minute). However, the C-terminus of TRPV1 preferably interacts with β-tubulin rather than α-tubulin (Goswami et al. 2007b). This preferential interaction with β-tubulin is also indicative of the nature of the contact site as the plus ends of microtubules are always decorated with β-tubulin and not with α-tubulin. In addition, the interaction of Maltose-Binding Protein-TRPV1-Ct with tubulin dimer was sensitive to Ca2+ and apparently the amount of pulled-down β-tubulin is more when compared with α-tubulin. Therefore, it could be possible that Storti et al. have missed certain responses that are only detectable if β-tubulin is used as the FRET partner. In addition, contribution of a complex system such as microtubule cytoskeleton, which is highly heterogeneous (because of presence of different post-translationally modified tubulin and isotypes) cannot be interpreted fully by FRET analysis. Also, tagging a fluorescent protein such as RFP at the N-terminus of tubulin can interfere with the interaction of tubulin with TRPV1 in native condition. More studies are needed to validate these differences. The interpretation by Storti et al. regarding the contribution of microtubules in the organization of functional TRPV1 complex was primarily based on the application of Nocodazole, which causes shift in the microtubule dynamics. In this context, it is important to mention that Nocodazole not only disrupt microtubule dynamics but also results in massive microtubule disassembly that may interfere drastically with the vesicular trafficking of TRPV1. Destabilization of microtubule could also account for the self-aggregation/oligomerization of TRPV1 with partial loss of functionality as observed experimentally (Storti et al. 2012). Nevertheless, interaction of TRPV1 with tubulin sets a stage where function of ‘membrane tubulin’ can be studied in great details.

How TRPV1–tubulin interaction is regulated?

The exact mechanism by which microtubule cytoskeleton is involved in regulation of TRPV1-related processes remains unclear. Increasing number of reports also suggest that apart from transport, microtubules, especially the dynamics of microtubule cytoskeleton plays a role in the regulation of TRP channels. The microtubule dynamics seem to be important as this primarily regulates the signaling complexes that are formed at the membrane scaffolds which need both TRP channels and microtubule (Goswami et al. 2011). In addition, recent reports suggest that microtubule cytoskeleton regulates multiple important aspects relevant for TRPV1 function (Fig. 3). First, microtubules regulate the tetramer organization from dimeric units of TRPV1 (Storti et al. 2012). In this regard, it is important to mention that C-terminal dimerization is prerequisite for further activation of TRPV1 (Wang and Chuang 2011). In addition, microtubule dynamics regulate the sensitization by key kinases such as by PKCε at S800 position of TRPV1 (Mandadi et al. 2006; Goswami et al. 2011). Recently, it has been demonstrated that the interaction of tubulin dimer to the Maltose-Binding Protein-TRPV1-Ct is dependent on the status of phosphorylation at S800 position. We reported that tubulin interaction with TRPV1-Ct is considerably less if the S800 position is phosphorylated previously (Goswami et al. 2011). Conversely, tubulin interaction considerably reduces the phosphorylation by PKCε as tubulin dimer seems to cover the S800 site and accessibility of this site to PKCε is less in the presence of tubulin. Phosphorylation of S800 position of TRPV1 by PKCε is directly involved with the sensitization of TRPV1 (Mandadi et al. 2006). Therefore, sensitization of TRPV1 is directly linked and dependent on the integrity as well as dynamics of microtubule cytoskeleton. This in turn also regulates the animal behavior such as mechanical hyperalgesia and pain perception (Goswami et al. 2011). A similar aspect has also been reported for TRPV4, which interacts with both soluble actin as well as with soluble tubulin and both compete with each other for the C-terminal cytoplasmic region of TRPV4 (Goswami et al. 2010). Interestingly, it has been demonstrated that Ser 824 residue of TRPV4 is phosphorylated by glucocorticoid-induced protein kinase1 and this phosphorylation regulates TRPV4 interaction with F-actin and microtubules in a complementary manner (Shin et al. 2012). Such phosphorylation-mediated regulation of interaction between different ion channels with tubulin has been reported by others and such regulation might be relevant for many physiological processes. For example, in the case of VDAC, phosphorylation by either glycogen synthase kinase-3β (GSK3β) or cAMP-dependent protein kinase A (PKA) regulates the tubulin interaction (Sheldon et al. 2011). These reports largely suggest that cytoskeletal dynamics and kinase activities regulate TRP channels in a reciprocal manner. Such regulations may also alter the voltage sensitivity, oligomer formation, and sensitization–desensitization cycle to a large extent. At present, it is not clear if microtubule skeleton can regulate the function at the level of single ion channels too. However, much more work on such type of regulations is needed.

Figure 3.

A plausible model depicting the involvement of microtubule cytoskeleton in the multimerization of Transient Receptor Potential Vanilloid Sub Type 1 (TRPV1). In step 1a, the monomeric TRPV1 forms a stable dimer. In native condition, tubulin dimer and microtubule filaments remain in a dynamic equilibrium (Step 1b). The TRPV1 dimer interacts with tubulin dimer and forms complexes where at least one dimer is associated with one TRPV1 dimer (Step 2). Subsequently, the complexes become attached with at least one microtubule filament (MT) via its protofilament (Step 3). The interaction with microtubule filament helps to form functional tetramer, a condition where the TRPV1-tetramer is associated with at least one microtubule filament (Step 4). Dissociation of microtubule and/or tubulin dimer from this complex is required to sensitize the TRPV1 complex by phosphorylation (Step 5). Further activation of TRPV1 (Step 6 and Step 7) results in several signaling events as well as Ca2+-influx which causes rapid disassembly of microtubules and shifts the equilibrium towards the tubulin dimer as a feedback mechanism (Step 8). A thin arrow indicates the direction of a reaction which is less likely to happen.

X-Y and Z factors of pain module and conclusion

In-depth understanding of the physical interaction between tubulin with TRP channels and the biochemical–biophysical properties of these complexes are extremely important. These are important mainly for their implications in basic research as well as in clinical applications, especially in the context of peripheral neuropathy. The prime importance of detailed understanding of TRPV1–tubulin complex exists in their potential therapeutic value in cancer, cancer pain, and chemotherapeutics-induced pain (Dina et al. 2003; Goswami and Goswami 2010). So far, the exact nature of molecular interaction and subsequent functional repercussions are not clear yet. It is well known that administration of Paclitaxel, a microtubule stabilizer to cancer patients, as chemotherapy develops severe neuropathic pain primarily because of alteration of the both microtubule organization and dynamics as well as by other factors in a more complex manner (Röyttä and Raine 1985; Blagosklonny and Fojo 1999; Bhalla 2003; Lee and Swain 2006; White and Rao 2008; Canta et al. 2009; Almeida-Souza et al. 2011; Fidanboylu et al. 2011; Okubo et al. 2011; Xiao et al. 2011; Zheng et al. 2011; Jaggi and Singh 2012; Xiao and Bennett 2012). Similarly, application of Vincristine, another regulator of microtubule dynamics also causes neuropathic pain (Lee and Swain 2006; White and Rao 2008). It has been shown that either stabilization of microtubules by short-term application of Paclitaxel (30 min) or destabilization of microtubules by Nocodazole/Colchicine/Vinca drugs alters pain sensitization (Table 2) (Dina et al. 2003; Kuhn et al. 2008; Goswami et al. 2011). These studies suggest that the integrity of microtubules can indeed modulate pain sensitivity (Bhave and Gereau 2003). Changes in the microtubule structures as well as in the mitochondrial shape and function in response to Paclitaxel has also been shown (Tanner et al. 1998; Topp et al. 2000). These side effects of microtubule regulators confirmed the involvement of microtubule cytoskeleton in the signaling of pain and thus correlate with TRP channels in such functions. However, the exact molecular mechanism by which microtubule stabilizers, especially Paclitaxel induces this long-lasting neuropathic pain is not clear.

Table 2. Involvement of TRPV1–tubulin complex in pain signaling
ConditionPriming byStimuliPost-stimuli treatment% change in nociceptive thresholdPossible reasonReferences
  1. 0, No change in nociceptive threshold (no alteration in hyperalgesia); +, % change in nociceptive threshold goes slightly up (pre-conditioning for hyperalgesia); ++, % change in nociceptive threshold goes up (development of hyperalgesia); +++, % change in nociceptive threshold goes up very high (extreme sensitivity to mechanical pain); −, % change in nociceptive threshold goes down (less hyperalgesia); −−, % change in nociceptive threshold goes up very down (insensitivity to mechanical pain); TRPV1 −/−, Anti-sense oligo-deoxynucleotides-mediated knockdown of TRPV1; Tax, Paclitaxel; Noc, Nocodazole; Col, Colchicine; Vin, Vinca drugs; Epi, Epinephrine.

TRPV1 +/+xxx0In absence of any nociceptive stimuli there is no hyperalgesia.Dina et al. (2003)
TRPV1 +/+Taxxx+Paclitaxel alone changes baseline nociceptive threshold slightly. This is possibly because of less tubulin at the SMR which cannot prevent spontaneous sensitization of TRPV1 by PKCε.Dina et al. (2003)
TRPV1 +/+Noc/Col/Vinxx0Noc/Col/Vin alone does not cause in the% change nociceptive threshold baseline. In absence of stimuli, there is no hyperalgesia.Dina et al. (2003)
TRPV1 +/+xEpix+++Development of strong hyperalgesia because of epinephrine application.Dina et al. (2003)
TRPV1 +/+TaxEpix+++Less tubulin at SMR cannot prevent sensitization (phosphorylation) more sensitization and strong hyperalgesia develops in response to epinephrine.Dina et al. (2003)
TRPV1 +/+Noc/Col/VinEpix0/+Disruption of microtubule by Noc/Col/Vin reduces the effect of Epi significantly. Noc/Col/Vin-mediated disruption of MT increases the tubulin dimer at the cell and also at SMR. Excess tubulin at SMR prevents sensitization of pain receptors.Dina et al. (2003)
TRPV1 +/+xG1x+++Less tubulin at SMR cannot prevent sensitization (phosphorylation) more sensitization and more hyperalgesia in response to G1.Kuhn et al. (2007)
TRPV1 +/+PKCε inhibitorG1x0In absence of PKCε activity, no pain signaling is possible, indicating the essential role of PKCε in this signaling event.Kuhn et al. (2007)
TRPV1 +/+NocG1x0Disruption of microtubule by Noc blocks effect of G1. Noc application increases the tubulin dimer at the cell and also at SMR. Excess tubulin at SMR blocks all potential phosphorylation sites and thus prevents phosphorylation-mediated sensitization of both TRPV1 and ‘Z’.Goswami et al. (2011)
TRPV1 +/+TaxG1x+++Less tubulin at SMR cannot prevent sensitization (phosphorylation) more sensitization and more hyperalgesia in response to G1.Goswami et al. (2011)
TRPV1 +/+xG1Tax+++Application of Paclitaxel after G1 has no effect. Less tubulin at SMR cannot prevent sensitization (phosphorylation), more sensitization, and more hyperalgesia in response to G1.Goswami et al. (2011)
TRPV1 +/+xG1Noc++Application of Noc after G1 has no effect. Less tubulin at SMR during the application of stimuli cannot prevent sensitization (phosphorylation), more sensitization, and more hyperalgesia in response to stimuli.Goswami et al. (2011)
TRPV1 −/−0Only knockdown of TRPV1 does not cause change in pain sensitivity in absence of stimuli.Goswami et al. (2011)
TRPV1 −/−G1++G-1-induced mechanical hyperalgesia does not depend on TRPV1. Suggest for another yet similar factor (termed as Z).Goswami et al. (2011)
TRPV1 −/−Taxxx0Only Paclitaxel application in TRPV1 −/− background does not cause change in pain sensitivity in absence of stimuli.Goswami et al. (2011)
TRPV1 −/−Nocxx0Only Noc application in TRPV1 −/− background does not cause change in pain sensitivity in absence of stimuli.Goswami et al. (2011)
TRPV1 −/−TaxG1xStabilization of microtubule reduces tubulin concentration at the SMR. This results in non-availability of tubulin for complex/scaffold formation on ‘Z’. Thus, no signaling events are possible in response to G1.Goswami et al. (2011)
TRPV1 −/−NocG1x++Destabilization of microtubule enhances tubulin concentration at the SMR. This results in availability of tubulin for complex/scaffold formation on ‘Z’. Thus, signaling events are possible in response to G1.Goswami et al. (2011)

In this context, recent work indicates that TRPV1–tubulin and TRPV4–tubulin complexes are important keys that can mediate several pain-signaling events (Table 2) (Goswami et al. 2004, 2006, 2007a,b, 2010, 2011; Suzuki et al. 2003). These works also indicate a possible mechanism and involvement of other factors that are so far unknown. For example, G1 (an estrogen analog) -induced mechanical hyperalgesia is highly contextual and dependent on both presence of TRPV1 and dynamics of microtubule cytoskeleton (Goswami et al. 2011). Surprisingly, presence of TRPV1 alone is not sufficient for this regulation as knocking down of the TRPV1 does not abolish the G1-mediated signaling leading to development of hyperalgesia (Goswami et al. 2011). This suggests the presence of an unknown yet similar factor which can be termed as ‘Z’ factor. Molecular properties of ‘Z’ are expected to be similar to TRPV1, but not same. For example, ‘Z’ should be able to interact with membrane tubulin. Therefore, ideally the ‘Z’ factor can be a transmembrane protein. It can be another TRP channel or some other ion channel per se that forms molecular complexes with membrane tubulin at the submembranous region and forms scaffolds. However, available literature suggests that TRPV4 maybe the best candidate as ‘Z’. This is because of the fact that TRPV4 interacts with tubulin and microtubule at membrane (Goswami et al. 2010). Moreover, in case of Paclitaxel-induced neuropathic pain, TRPV1, TRPV4, and TRPA1 are sensitized in a similar manner by protease-activated receptor 2 suggesting that these three channels can share a certain degree of redundancy (Chen et al. 2011; Materazzi et al. 2012). In addition, TRPV4 has recently been considered as an essential factor for chemotherapy-induced neuropathic pain as mechanical hyperalgesia induced by Paclitaxel and Vincristine was strongly reduced in trpv4−/− mice (Alessandri-Haber et al. 2004, 2008). However, more studies are needed to understand the molecular identity of ‘Z’.

It has been demonstrated that stabilization as well as destabilization of microtubules by prior application of Paclitaxel or Nocodazole, respectively, do not cause any hyperalgesia, especially in absence of any other nociceptive stimuli (Dina et al. 2003). Often, the effect of microtubule stabilization and destabilization prior to application of certain nociceptive stimuli are opposite. Similarly, effect of certain nociceptive stimuli in wild-type and TRPV1-knocked-down system is opposite. For example, prior application of Paclitaxel or nocodazole before application of G1 (a functional analog of estrogen) in wild-type and TRPV1-knocked-down system is just opposite (Goswami et al. 2011). It suggests that factors other than microtubule cytoskeleton are also involved in such complex signaling events (Dina et al. 2003; Goswami et al. 2011). Taken together, this may suggest that the formation of at least one binary complex at the submembranous region is favored by TRPV1–tubulin complex. The unknown binary complex, which sequesters on TRPV1–tubulin complex can be termed as ‘X-Y’ factor of pain module. It seems that TRPV1–tubulin complex can provide a scaffold at the submembranous region where ‘X-Y’ complex as well as other regulatory molecules will form a complex network of signaling events (Fig. 4). Therefore, the TRPV1–tubulin interaction probably plays a catalytic role only. For such a purpose, ionic conductivity via the TRPV1 channel may not be needed as formation of TRPV1–tubulin complex at the plasma membrane is independent of the channel function and even can be achieved with a partially truncated but non-functional TRPV1 polypeptide (Goswami et al. 2011). Based on the experimental results, certain characteristics and the molecular identity of this ‘X-Y’ factor can be derived (Dina et al. 2003; Goswami et al. 2011). For example, it can be postulated that factor ‘X’ and ‘Y’ cannot form an active ‘X-Y’ complex at the cytoplasm by their own and needs a highly specialized scaffold made of TRPV1–tubulin complex (alternatively ‘Z’–tubulin complex in absence of TRPV1) at the submembranous region and a nociceptive stimuli such as G1 or epinephrine to become active. It can also be possible that ‘X-Y’ complex is a substrate for PKCε phosphorylation. However, further studies are needed to test this hypothesis.

Figure 4.

‘X-Y-Z’ module of mechanical hyperalgesia. A plausible model demonstrating the involvement of transient receptor potential vanilloid (TRPV)–tubulin complex(es) in development of mechanical hyperalgesia. Presence of TRPV1 but not the channel function (Ca2+-influx) is essential for development of mechanical hyperalgesia induced by certain stimuli such as G1 (estrogen analog). This can be explained if TRPV1 acts as a scaffold which facilitates the formation of a complex made of two unknown factors (here termed as ‘X’ and ‘Y’). As this complex is inactive in absence of nociceptive stimuli, it seems that this complex become activated only at the submembranous region because of TRPV1-mediated signaling events. As knocking down of TRPV1 also results in mechanical hyperalgesia in response to certain conditions, suggesting a presence of another unknown yet similar channel (termed as ‘Z’) which can also scaffold ‘X’ and ‘Y’ and allow activation of ‘X-Y’ complex. The TRPV1 (or ‘Z’ in absence of TRPV1) interacts with soluble tubulin dimer and forms a scaffold at the submembranous region (SMR) (Step 1). Subsequently, these scaffolds allow forming nociceptive complex ‘X-Y’ at the submembranous region (Step 2). Stimuli such as kinase activation may result in the sensitization of TRPV1 (or ‘Z’ in absence of TRPV1) and subsequent channel opening (Step 3). This also allows release of ‘XY’ the activated nociceptive complex and induces Ca2+ influx leading to development of hyperalgesia (Step 4). Stabilization of cellular microtubules by Paclitaxel reduces the availability of soluble tubulin at the SMR and thus allows more sensitization of the TRPV1 because of more accessibility of the kinases for the key residues of TRPV1. In contrast, Nocodazole, which disrupts the microtubules and increase the availability of soluble tubulin at the SMR. This condition thus prevents sensitization of the TRPV1 because of non-accessibilities of the kinases for the key residues.


Funding from Department of Biotechnology (Govt. India, grant number BT-BRB-TF-2-2011) and National Institute of Science Education and Research are acknowledged. I thank Vera Moiseenkova-Bell (Cleveland, Ohio, USA) for providing the cryo-EM images of TRPV1 for this review. Scientific input from Prof. Raymond E. Stephens (Boston University) is appreciated. Input from all the laboratory members is appreciated. CG declares no conflict of interest associated with this review article.