ATP-gated P2X7 receptors, expressed in immune and glial cells, have been implicated in various chronic pain conditions and recently they have attracted considerable attention as potential therapeutic targets. However, all of their functional properties are still poorly understood. Therefore, in the third part of the study (study III), we explored the functional role of the residue F288 located in the functionally important left flipper region in the P2X7 receptor. Therefore, we studied the basic properties of the non-sensitized, desensitized, and sensitized states of the P2X7 receptor.
5.3.1 Kinetic characteristics of the WT and F288S rat P2X7 receptors
First, we compared the kinetic characteristics of currents mediated by the WT and F288S mutant of rP2X7 receptor. In this experiment, we used two protocols with short (2 second) and long (20 second) applications of the natural P2X7 agonist ATP (1 mM). The first protocol involved short ATP applications; we found that the most notable difference between the WT and F288S receptors was the slow decay (deactivation) time in the latter. The decay time (τ) for both receptors was well fitted with the mono-exponential curve but this parameter for F288S mutant had approximately 3 times longer than in the WT receptor. Thus, in the WT, it was 0.8 ± 0.1 s (n = 29) whereas in the F288S mutant it was 2.6 ± 0.2 s (n = 24; p < 0.05). These data are shown in the study III, Fig.1A, B.
We also showed that the desensitization of the mutant receptor measured as the decay of the current in the presence of agonist (τDS) was significantly slower than in the WT. Thus, in the WT, it was 0.45 ± 0.5 s (n = 7) whereas in the F288S mutant, the τDS was 3.09 ± 1.35 s (n =7; p <
0.05; study III, Fig. 1C). Interestingly, when constructing the dose-responses curves, we found no significant differences in the potency between the WT and F288S receptors. The only difference was present in the maximal responses to 3 and 5 mM ATP. Thus, in the F288S mutant, there was a lower amplitude of maximal response than in the WT P2X7 receptor (study III, Fig. 1D).
Next, using longer application (20 s application protocol) of 1 mM ATP, we were able to observe more complex current kinetics through the P2X7 receptor. In particular, in these conditions, we observed the secondary peak generation in the WT P2X7 receptor (study III, Fig 2A left). In contrast, there was a lack of a secondary peak in F288S (study III, Fig 2A right).
It is known that the secondary peak is due to the process called ‘sensitization’ (Yan et al., 2010; Khadra et al., 2013). Our finding suggested that, unlike the WT receptor, the F288S P2X7
mutant has a lower probability to enter into the sensitized state (study III, Fig. 2A). Consistent with short ATP applications, the current decay after the end of 20 s long ATP application was 3 times slower in the F288S mutant in comparison to the WT. Thus, in the WT, the decay was 0.6 ± 0.1 s (n = 5), whereas in the F288S mutant, this parameter was 1.9 ± 0.2 s (n =6; p < 0.01;
study III, Fig. 2A,B). Further, to investigate the ability of the F288S to be sensitized, we used a higher ATP concentration (5 mM applied for 20 s). As expected, the WT receptor showed an initial current peak then the fast current decline followed by a gradually growing secondary peak (study III, Fig. 2C, left). This shape was previously observed by others, and was explained by initial desensitization after ‘normal’ channel opening followed by the sensitization (Yan et al., 2008; Khadra et al., 2013). Notably, with this high concentration of ATP, both phenomena (desensitization and sensitization) were observed not only in the WT but also in the F288S receptor (study III, Fig. 2C, right). Interestingly, the development of sensitization in the F288S mutant was associated with accelerated deactivation (study III, Fig.
2C, D). Thus, the significant difference in the decay time of the WT and F288S which was clearly visible with 1 mM ATP, disappeared at high ATP concentrations (WT: τ = 0.5 ± 0.1 s, n = 8; F288S: τ = 0.4 ± 0.1 s, n = 10; p > 0.05).
These data suggested different mechanisms of deactivation of the P2X7 receptors, with and without sensitization. The results obtained identified the different kinetics of deactivation in the WT and F288S receptor linked to the key amino acid in position 288 in the left flipper region of the rP2X7 receptor.
5.3.2 Kinetic modelling of the WT and F288S rP2X7 receptors activated by ATP
In the kinetic modelling, in collaboration with Dr. A. Skorinkin (Kazan), we used the previously published model for BzATP (Khadra et al., 2013) to describe the action of ATP on the P2X7 receptor. The principal structure of the kinetic model is presented in publication III, Fig. 3. With our modified model, we were able to replicate the currents mediated both by the WT and the F288S receptor during either 2 or 20 s application of 1 mM ATP. The results are presented in study III, Fig. 4A, C and B, D. where the rate constants used in the model for the WT and F288S receptor currents are also presented (Fig. 3 B and C). In order to simulate the slow current decay in the F288S mutant, we reduced the speed of agonist unbinding (transition from A3R to R; study III, Fig 3C). To simulate the absence of the transient initial current decline and the lack of the secondary peak in the F288S mutant receptor (study III, Fig. 4B,D), we decelerated transitions from the non-sensitized open receptors (A3R) to sensitized (A3S) or desensitized (A3D) receptor states.
5.3.3 Model validation and predictions
We performed a validation experiment to test if our model could replicate the dependence of the current decay on the agonist concentration for both the WT and F288S. We found that in the model of the WT receptor as well as in our experimental conditions with the WT, the current decay was independent of the concentration or application duration (study III, Fig.
5A, where the experimental decay is shown as white columns vs the black columns of the model). Furthermore, our model was able to reproduce an unusual feature of F288S mutant that gives very different deactivation rates, with and without sensitization (study III, Fig. 5B).
5.3.4 Characterization of rP2X7 non-selective pore opening
Next, as an independent approach to characterize the function of the P2X7 receptor, we quantified the uptake of the fluorescent dye YO-PRO1 via the WT and F288S mutant receptor expressed in HEK cells using flow cytometry cell sorting (FACS). To stimulate sensitization in the WT and F288S receptors, we used the prolonged ATP application (1 mM, 1−2 min) in the presence of a YO-PRO1 dye in the suspension solution. Consistent with our electrophysiological data, we observed that the dye uptake in cells expressing WT rP2X7 receptors was significantly increased after ATP application. However, in the F288S mutant, the dye uptake was not significant different from control conditions prior to ATP application (study III, Fig. 6A–C). It is also noteworthy that pre-incubation of cells with the P2X7 antagonist (5µM) prevented previously observed dye uptake in WT rP2X7 and the rP2X7 F288S mutant (study III, Fig. 6C).
Next, as noted in the electrophysiological experiments, in order to test the ability of the F288S rP2X7 mutant to take up the dye with the stronger activation, we used 5 mM ATP. In agreement with the electrophysiological tests, we were able to detect a significant dye uptake, analogues to that of the WT (WT: 26.0 ± 1.5% vs control 9.2 ± 1.5%, n = 6, p < 0.05; F288S: 22.4
± 3.5% vs control 7.8 ± 0.8%, n = 7, and p < 0.05). Taken together, this data confirm that the F288S mutant requires stronger stimulation than the WT rP2X7 to show significant YO-PRO1 uptake.
5.3.5 ATP action on human WT, Y288F and Y288S P2X7 receptors
Human P2X7 receptor, unlike the rat P2X7 receptor, has a tyrosine at position 288. For this reason, we studied the functional properties of two hP2X7 mutants: Y288F (rP2X7-like) and Y288S (288S mutant-like). To be consistent with experiments on rat P2X7 receptors, we used the same protocol with short 2 s applications of 1 mM ATP. We did not find any difference in the decay time for hP2X7 WT and Y288S (study III, Fig. 8). However, the hP2X7 mutant, Y288F (rP2X7-like), showed the fastest decay time, which was 2 times faster than that of WT hP2X7 or the Y288S mutant (Y288F: τ = 0.45 ± 0.04 s, n = 25; WT: τ = 0.74 ± 0.05 s, n = 38; Y288S:
τ = 0.73 ± 0.05 s, n = 25; study III, Fig. 8B). It is noteworthy that tyrosine and serine are both polar amino acids, probably similarly contributing to the prolonged deactivation time of the WT hP2X7 and the Y288S mutant. Our observation that phenylalanine at position 288 determines the fast deactivation of the rat WT receptor was replicated in the Y288F (rat-like) mutant.
5.3.6 Molecular modelling of the WT P2X7 receptor and F288S mutant
As an additional approach, in collaboration with Dr. K. Khafizov (Moscow), we also constructed the homology models of the ATP-bound form for both the WT receptor and the F288S mutant (study III, Fig. 9). Our molecular model shows that the side chain of S288 potentially can form a hydrogen bond with an ATP molecule. Thus, we suggest a direct role for the S288 residue in the binding of agonist molecules. A similar role was also recently suggested based on the P2X3 X-ray structure (Mansoor et al., 2016). Our molecular modelling is consistent with kinetic modelling and experimental data by proposing that residue 288 plays a key role in stabilization of the non-sensitized P2X7 receptor states and also determines the slow rate of ATP unbinding.
6 Discussion
In the present work, we elucidated the anti-nociceptive effects of two groups of polyphosphates via the state-dependent inhibition of the pro-nociceptive P2X3 receptors in sensory neurons. The specificity of these effects was confirmed in experiments with recombinant P2X2, P2X3 and P2X7 receptors (studies I, II). Additionally, we demonstrated the functional role of residue 288 in the ectodomain of the pro-inflammatory P2X7 receptor (study III).
In the first study, we found that two stable synthetic Ap4A-analogues (AppCH2ppA and AppNHppA) exhibited considerably high inhibitory activity on rat and human P2X3 receptors. These effects correlated with anti-nociceptive activity in the pain models in vivo (study I).
In the second study of present work, we demonstrated that the NBP-induced compound ApppI (ATP-analogue) provided even stronger inhibition of pain transducing P2X3 receptors. This inhibition was achieved with even low nanomolar concentrations of ApppI with similar specificity towards the P2X3 receptor subtype (study II). These results suggest a novel therapeutic perspective of tested compounds for analgesic treatments in cancer and bone pain.
In the third study of the thesis, we assessed the functional role of the residue at position 288 in the left flipper of the rat and human P2X7 receptors (study III). The left flipper itself is an important region for the normal function of all ATP-gated P2X receptors. One of our key findings is that the mutation of F (phenylalanine) in position 288 to S (serin) significantly slowed receptor deactivation and lowered the probability of the transition of the receptor into the sensitized state. These data provided previously unknown properties of this residue in the P2X7 receptor, which can be used for construction of novel drugs modulating these widely distributed pro-inflammatory receptors.
6.1 SELECTIVE INHIBITION OF P2X3 RECEPTORS BY NOVEL