2.5 Purinergic mechanisms involved in pain
2.5.2 Specific properties of pro-nociceptive P2X3 receptor
Homomeric P2X3 (and heteromeric P2X2/3 in some species) receptors are almost exclusively expressed in peripheral sensory neurons (Chen et al., 1995; Lewis et al., 1995). The current evidence suggests that they are involved in chronic, inflammatory and neuropathic pain (Cockayne et al., 2000; Souslova et al., 2000; McGaraughty et al., 2003; Chen et al., 2005;
Burnstock and Di Virgilio, 2013). They are expressed in C- and Aδ– sensory nerve fibers, participating in bone innervation as shown in Fig 2 (Mach et al., 2002; Kaan et al., 2010;
Castañeda-Corral et al., 2011; Falk et al., 2012; Zhao and Levy, 2014). Indeed, there is a growing body of evidence for the involvement of P2X3 receptors in cancer bone pain (Kaan et al., 2010; Rumney et al., 2012; Wu et al., 2012; Franceschini and Adinolfi, 2014). It has also been demonstrated that bone cancer is associated with an elevated expression of P2X3 receptors in CGRP-positive nociceptive sensory nerves, which are widely present in the bone and involved in pathological peripheral sensitization (Gilchrist et al., 2005; Wu et al., 2016).
Consistent with this, a blockade of peripheral P2X3 receptors has effectively reduced the severity of tumor-induced cancer pain (González-Rodríguez et al., 2009; Kaan et al., 2010;
Hansen et al., 2012). One important factor is that, unlike the situation in rodents, human nociceptive neurons do not co-express P2X3 with P2X2 subunits (Serrano et al., 2012) . This means that homomeric P2X3 receptors are the predominant nociceptive receptor type involved in purinergic pain signaling in human bone cancer.
Figure 9. Architecture and pore structure human P2X3 (hP2X3) receptor (G-I) in an apo/resting state, (A-C) an agonist-bound (unbound agonist state)/open-pore state, (D-F) an agonist-bound/closed-pore/ desensitized state. Adopted from Mansoor et al., 2016.
The P2X3 receptors, when activated by ATP, are characterized by rapidly activating, and quickly desensitizing inward currents. The rate of this receptor desensitization depends on the type of agonist (ATP or its analogues) and the concentration. Subsequent studies revealed the presence of two types of receptor desensitization: the fast process, evoked by micromolar agonist concentrations, developing in the range of tens of milliseconds and a slow process induced by nanomolar agonist concentrations applied for tens of seconds, that inhibited receptors without macroscopic signs of activation (Sokolova et al., 2006). The latter phenomenon has been called ‘high-affinity desensitization’ (HAD) (Sokolova et al., 2006).
HAD likely develops via a direct transition of receptors from their closed states to inactive states (Sokolova et al., 2006).
The recent study conducted by Mansoor (Mansoor et al., 2016), reporting X-ray crystal structures of human P2X3 (hP2X3) receptor in apo/resting, bound/open-pore, agonist-bound/desensitized and antagonist-bound
closed states, extended our understanding of this receptor functioningP2X3 receptor. Like the other P2X receptor subtypes, it shares a common topology containing a large hydrophilic extracellular domain, six α-helices forming the transmembrane (TM1 and TM2) domain, and intracellular termini (Fig 8, 9) (North, 2002; Mansoor et al., 2016)
The TM2 lines the pore lumen, which contains residues I323, V326, T330, and V334 fronting a chanel pore (Fig 9 g, h, i) (Egan et al., 1998; Kawate et al., 2009; Hattori and Gouaux, 2012b;
Mansoor et al., 2016). In the apo state this residues define a chanel as closed due to defining narrow P2X3 pore, which is not possible for dehydrated Na+ ions to pass (Fig 9 h, I and 10).
Residue I323 defines the extracellular border of the receptor gate in the apo state (pore radius 0.3 Å). T330 located at the cytoplasmic border of the gate and defines a pore radius 0.7 Å (Fig 9 I and 10). Another important residue, V326, contributes to the pore occlusion (Mansoor et al., 2016). Notably, the residues lining of the receptor gate are similar for both antagonist-bound structures and the apo state structure, demonstrating that these are stabilized by the competitive antagonists (Mansoor et al., 2016).
The open state structure of hP2X3 contains ATP in the ligand-binding pocket and an open pore (Fig 9 a–c). In this state hP2X3 has a continuous open pore through the TM with a radius of 3.2 Å (Fig 9 d, e, f), which penetrable to partially hydrated Na+ (Degrève et al., 1996) (Fig 9 b, c). Residues I323 and V326 to open the pore are translated upward to the extracellular surface and rotated outward, away from the pore’s center (Mansoor et al., 2016). Whereas, T330 defines the narrowest region of the chanel pore in the open state as represented on Fig 10 (Mansoor et al., 2016). There is only two hydrophilic residues in the hP2X3 pore lining, T330 and S331. Notably, a rP2X2 threonine in the equivalent residue to T330 of hP2X3 receptor, has been implicated in ion selectivity (Migita et al., 2001), suggesting that T330 interact with cations permeating the pore.
ATP binding pushing the left flipper outward and leads to the closure of the cleft between the head and dorsal fin (Hattori and Gouaux, 2012b; Jiang et al., 2012). These structural rearrangements are transmitted to the lower body, and leads to the outwards movement of the β1, β9, β11 and β14 strands (Fig 8, 9)(Mansoor et al., 2016). This outward flexing of β1 and β14 strands (which are directly connected to the TM1 and TM2, see Fig 8, 9) pulls the extracellular parts causing the expansion of this helixes outward and thereby opening the pore (Li et al., 2010; Hattori and Gouaux, 2012b). Further, flexing of the lower body pulls on TM2, thus the helix rotates outward which promotes the rotation move away from the pore center of I323 and T330 (Mansoor et al., 2016). Notably, important function have been described by Mansoor et al., that unlike in zfP2X4 (Hattori and Gouaux, 2012b), the movement of TM2 in hP2X3 receptor not only rigid-body transformation but additional movement from an α -helix to a 310-helix occur to open the chanel. This additional change in helical position allows the movements of TM2 associated with channel opening and desensitization.
Figure 10. A cartoon representation of the gating cycle of P2X receptors. Adopted from Mansoor et al.,2016.
In the desensitized state, P2X3 receptor structure also has ATP in the pocket, but a closed ion
pore (Fig 9 d–f). Interestingly, a single residue, V334, determines the construction site of the desensitized state with a pore radius of 1.5 Å, which is not penetrable to the hydrated Na+ (Fig 9 e, f). V334 translates towards the extracellular surface and rotates inward to block the pore. Transition from open to the desensitized state is mediated by the cytoplasmic portion of TM2 which slightly rotates and then the short 310-helix reverts to an α–helix (Mansoor et al., 2016). This movement closes the pore with reversed move of V334. Thus, meaning that in the desensitized and open states, hP2X3 adopts similar conformations in the extracellular domain and binding pocket, but possessing differences in the cytoplasmic TM2 domain and at the gate (Fig 10). Thus, these points to the existence of a P2X3 desensitized state not previously observed for any other P2X structure. Furthermore, it has been shown that the apo-structure of zfP2X4 reveals several unique features of the hP2X3 apo-structure, including a more complete TM domain, different residues defining the pore constriction, and an Mg2+
ion bound in the head domain near the ATP binding pocket (Fig9 g). Notably, the 'left flipper'
(Kawate et al., 2009) is a specific and important region in the ectodomain of P2X receptors since it faces the ATP binding pocket (Mansoor et al., 2016). We previously reported that the conserved serine 275 in the left flipper region plays an essential role for the function of ATP-gated P2X3 receptors (Petrenko et al., 2011). In particular, we proposed that this residue contributes to both agonist binding and receptor desensitization. This role of left flipper in agonist binding was recently confirmed with a new structure of P2X3 receptor in the ATP-bound state (Mansoor et al., 2016).
In summary, the P2X3 receptor is one of the best-characterized members of P2X family. Its selective expression in nociceptive fibers and involvement in chronic and cancer pain has designated it as an important therapeutic target. In fact, the particular combination of agonist-specific mechanisms such as desensitization, including a specific process of HAD, could help to modulate the pro-nociceptive responsiveness of sensory neurons to endogenous P2X3 receptor agonists. For instance, it might be possible to generate sustained inhibition of P2X3 receptors for controlling chronic pain with a low, subthreshold concentration of a desensitizing agonist acting via the HAD mechanism (Viatchenko-Karpinski et al., 2016). Therefore, it seemed an attractive aim to elucidate if there were powerful selective ’desensitizers’ of P2X3 receptors in order to discover their impact on physiological and pathological pain conditions.