• Ei tuloksia

Anti-nocicEption induced by the bisphosphonate-induCed ATP-analogue

> 100 µM 0.11 ± 0.02 µM 0.28 nmol/kg (0.23mg/kg)

AppCH2pp A

34.66 ± 13.21 µM 0.29 ± 0.05 M 0.72 nmol/kg (0.61 mg/kg)

It is noteworthy that the calculated in vivo IC50 values for the tonic phase are in good agreement with the IC50 values determined on rP2X3 receptors (study I, Fig 2 and 3; table 2).

Subsequently, we tested analgesic properties in a model of chronic thermal hyperalgesia induced by the CFA injection into rat hind paw. Therefore, we used Hargreaves plantar test to monitor the possible anti-nociceptive effects after CFA-induced thermal hyperalgesia and the subsequent injection of AppNHppA or AppCH2ppA. We found that anti-nociceptive effects could be observed during the first 200 min after intra-plantar injection of AppNHppA and AppCH2ppA into the inflamed hind paw (study I, Fig 10 A-B). Thus, intrathecal injection of AppCH2ppA (20mM) induced a significantly weaker and delayed the anti-nociceptive effect on CFA-induced thermal hyperalgesia in comparison to 180 min with intra-planar injection (study I, Fig 10 C). Moreover, the observed anti-nociceptive effect by intrathecal administration was only 22 %, while the anti-nociceptive effect following intra-plantar administration was 58% inhibition (study I, Fig 10 C-D).

5.2 ANTI-NOCICEPTION INDUCED BY THE BISPHOSPHONATE-INDUCED ATP-ANALOGUE APPPI (STUDY II)

The second part of the thesis (study II) was carried out to address a possible anti-nociceptive effect of the bisphosphonate-induced ATP-analogue ApppI via inhibition of P2X3 receptors in peripheral sensory neurons and expressed in host HEK cells.

5.2.1 Activation and inhibition of rat recombinant P2X3 receptors by ApppI

First, we tested whether ApppI can activate pro-nociceptive P2X3 receptors. In an attempt to clarify its possible agonist effect on the P2X3 receptors, we used rat P2X3 receptors (rP2X3) expressed in HEK cells. To evaluate this function of ApppI, we compared ApppI-induced responses with those induced by the full agonist of P2X3 receptors, -meATP. Our data demonstrated that ApppI is a partial agonist of the rP2X3 receptor. ApppI was able to activate P2X3 currents only at relatively high concentrations (study II. Fig 1). Thus, the EC50 (concentration which induces the half of the maximal effect) was relatively high (16.4 ± 0.9 µM, n = 7; study II Fig. 1C). Significantly, P2X3 receptor (like P2X1 receptor) is distinct from the other P2X receptors by fast rate of desensitization and by ability to develop HAD (Chen et al., 1995; Lewis et al., 1995; Cook and McCleskey, 1997; Jiang et al., 2005; Sokolova et al., 2006). We found that P2X3 receptor mediated currents induced by both -meATP and ApppI are prone to desensitization (Fig. 1B). However, using paired-pulse protocol with

variable intervals (5–120 seconds), we demonstrated that the recovery time of rP2X3 receptors from desensitization was longer in the case of ApppI (study II, Fig. 1D, B).

Next, we tested the ability of ApppI to induce HAD of P2X3 receptors. HAD, providing long-lasting inhibition of P2X3 receptors, could contribute to anti-nociception in pain states mediated by the pro-nociceptive effects of endogenous ATP (Sokolova et al., 2006; Giniatullin and Nistri, 2013). Thus, we used different concentrations of ApppI to evaluate its inhibitory potency on rP2X3 receptors. Surprisingly, we found that the inhibitory effect on P2X3 receptors already with 10 nM ApppI, which was administered for 120 s between two -meATP test (10 µM for 2 s) applications. ApppI in this concentration decreased the -meATP-induced membrane currents to 36 ± 8 % of control values (P=0.0013, n=10; study II, Fig.2B). With increased concentrations, the inhibitory effect of ApppI further increased. Thus, with 100 nM of ApppI, the currents were reduced to only 15.4 ± 3.9 % (p=0.02, n=4), whereas in 1 mM, ApppI reduced currents to 7.1 ± 3.9 % (p=0.03, n=4; study II, Fig.2C).

Then, we tested if the anti-nociceptive effect of ApppI could be reproduced when P2X3 receptors were activated with the natural agonist, ATP instead of -meATP. Notably, we found a similar strong inhibitory potency of 10 nM ApppI on P2X3 receptors activated by ATP (study II, Fig.2C, D). Further, to evaluate the potency of the ApppI-induced inhibition on rP2X3, we constructed the dose-response curve for -meATP alone and in the presence of 10 nM ApppI. We found that 10 nM ApppI strongly decreased currents induced by different concentrations of -meATP. This result was indicative of reduced efficacy without any significant changes in affinity (study II, Fig.2F).

In summary, in this part of the study, we found that ApppI was only a weak partial agonist and a non-competitive antagonist of P2X3 receptors operating via a HAD mechanism. These data suggested that ApppI was a powerful anti-nociceptive agent.

5.2.2 The role of extracellular calcium in ApppI inhibition

Bone cancer is often associated with profound disturbances of calcium homeostasis resulting in certain clinical manifestations such as hypercalcemia (Mantyh, 2014). As many membrane receptors are calcium dependent, including P2X3 subtype, hypercalcemia can be implicated in bone pain. In particular, the function of the pro-nociceptive P2X3 receptor could be enhanced in the presence of elevated extracellular calcium (Giniatullin and Nistri, 2013).

Notably, BPs treatment commonly used for bone disorders, including bone cancer, can effectively reduce elevated calcium levels (Lopez-Olivo et al., 2012; Lluch et al., 2014).

Therefore, BPs can provide pain relief via this indirect mechanism. To investigate potential role of calcium in the inhibitory effect of ApppI on rP2X3 we tested ApppI anti-nociception in different concentrations of extracellular calcium. We found that in a low concentration (0.2 mM) of extracellular calcium, 10 nM ApppI had the strongest depressant action on rP2X3 receptors. Thus, membrane currents decreased to 12 ± 3% of control values (p=0.0002, n=14;

study II, Fig.3A and C). However, further increase of the extracellular calcium to 4 mM or to 10 mM largely reduced the ApppI-induced inhibition (68 ± 8%, p=0.035, n=4; and 86 ± 10%, p>0.1, n=7). Thus, in conditions mimicking hypercalcemia, in the presence of 10 mM calcium, the inhibition was almost absent, (study II, Fig.3B, C).

5.2.3 Selectivity of ApppI effects on P2X3 receptors

To determine the specificity of ApppI action on the pro-nociceptive P2X3 receptors, we tested the action of this compound on other P2X receptors. First, we tested the action of ApppI on recombinant rP2X2 receptors. In contrast to rP2X3, we did not detect any agonist activity of ApppI even at relatively high (10 M) concentrations on rP2X2 receptors, although we did observe a large response of these receptors to 10 M ATP (n=4; study II, Fig. 4A). In contrast to P2X3 receptors, we did not detect any inhibitory effect of ApppI on rP2X2 receptors (n 5 4;

study II, Fig. 4B).

Similarly, we identified neither agonist nor antagonist (inhibitory) properties of ApppI on rP2X7 receptors (n=4; study II, Fig. 4C and D). Thus, these data pointed to the specificity of ApppI action on P2X3 receptors.

We further elucidated the possibility that ApppI can interfere with the hP2X7 and hP2X2 receptor subtypes. We used similar protocols as used for rP2X2 and rP2X7 receptors to evaluate both the agonist and inhibitory action of ApppI. We used 1 mM ATP for sufficient stimulation of the low affinity hP2X7 receptors (study II, Fig. 4, G and H). Similarly to the situation with rat receptors, we noted that ApppI exerted neither any agonist nor inhibitory effects on hP2X7 and hP2X2 receptors (study II, Fig. 4, C, D, G, and H), again evidence that ApppI is selective for P2X3 subunit–containing receptors in both rats and humans.

Next, we compared the action of ApppI on currents activated by -meATP via native P2X receptors expressed in trigeminal and nodose sensory ganglia. The nodose sensory neurons, responsible for the detection of visceral pain, express preferentially heteromeric P2X2/3 subunits that produce sustained currents that are much less prone to desensitization as shown in study I. In contrast, trigeminal ganglion nociceptive neurons, supplying the head and face tissues and taking part in innervation of the cranial periosteum (Zhao and Levy, 2014) express mainly P2X3 subunits with fast desensitization and HAD (I).

In trigeminal ganglion neurons, we found that fast desensitizing currents activated with

meATP were potently inhibited by 10 nM ApppI (to 44 ± 6.4%, P=0.0007, n=17; study II, Fig. 5A, C). Further, we found that ApppI-induced inhibition on native rP2X3 is calcium dependent as observed in recombinant rP2X3 receptors. In conditions with low calcium levels (0.2 mM), 10nM ApppI reduced currents to 9 ± 4% (p=0.05, n=7; study II, Fig 5C), while in a condition mimicking hypercalcemia (10 mM), we detected no inhibitory action of ApppI on trigeminal neurons (87.7 ± 4.9%, p>0.5, n=10; study II, Fig 5C). Notably, slow currents recorded in ND neurons were mostly insensitive to ApppI (90.4 ± 3.4%, p>0.1, n=8; study II, Fig 5B).

A summary of this data from recombinant receptors, we suggest that ApppI inhibition is selective for homomeric P2X3 receptors and this proposal was reinforced by the intense calcium-dependent ApppI inhibitory action on homomeric rP2X3 receptors.

5.2.4 ApppI-induced effects on human P2X3 receptors

To explore the translational aspects of our findings, we further tested the effects of ApppI in human receptors. First, we tested the inhibitory action of nanomolar ApppI on human P2X3

(hP2X3) receptors expressed in HEK cells. Importantly, we detected a significant depressant effect of 10 nM ApppI on hP2X3 receptors, even stronger than in the rat homolog. These data were obtained with physiological concentrations of extracellular calcium (2 mM). In these conditions, concentrations as low as 10 nM ApppI, produced a depressant effect of hP2X3 receptor, inhibiting their activity down to 2.7 ± 1.0 % of control values (p=0.02, n=8; study II, Fig 6 A, C). In an attempt to explore the role of hypercalcemia, we also characterized ApppI inhibition in the presence of a high (10 mM) calcium concentration. As expected, in this surrogate model of hypercalcemia, the inhibitory action of 10 nM ApppI was virtually absent (84 ± 4% of control, n=14, p>0.05; study II, Fig 6B, C). In summary, our data confirmed the strong calcium dependent inhibitory action of ApppI both in rat and in human homomeric P2X3 receptors. The observed strong effect of low ApppI in human pro-nociceptive P2X3 receptors hinted at a potential analgesic action of this polyphosphate compound in hyperalgesia conditions in humans.

5.2.5 Biodegradation of the ApppI in living tissues

If it is to be used in vivo, the prototype analgesic compound should be relatively stable.

However, it is commonly accepted that polyphosphates undergo fast degradation in living tissues due to enzymatic activity of multiple extracellular nucleotidases (Yegutkin et al., 2008, 2016). To test the speed of ApppI degradation in living tissues, we compared the breakdown of ApppI with one of the endogenous polyphosphates, ATP. In these experiments, we used the lead nitrate method, which is based on the precipitation of lead orthophosphate (Yegutkin, 2008). This method was optimized recently to estimate the nucleotidase activity in rat meninges (Yegutkin et al., 2016). The amount of orthophosphate precipitated is proportional to the extent of phosphate hydrolysis; these precipitates could be clearly visualized as brown deposits around the main arteries in the meninges. We found that the intensity of the tissue labelling was low in control (28.4 ± 2.4 a.u., n =4), slightly more pronounced with 300 µM ApppI (59.6 ± 6.3 a.u., p = 0.0033, n =4), and was very strong with 300 µM ATP (221.1 ± 8.7 a.u., p=0.0001, n = 4). These results indicated a relatively slow rate of ApppI degradation in living tissues compared to the natural polyphosphate ATP (study II, Fig. 8, A–D). Thus, our data suggest that ApppI exerts a possibly longer effect on P2X3 receptors in vivo.

5.2.6 Testing the direct action of the NBP zoledronate on P2X3 receptors

The NBP, which are able to produce ApppI (Mönkkönen et al., 2006), can evoke analgesia not only via the production of this novel endogenous polyphosphate but, most likely also directly affecting the activity of the pro-nociceptive receptors. Therefore, next we tested a possible direct effect of the widely used NPB, zoledronate, on rat P2X3 receptors. We found that 10 µM zoledronate exerted no effect on the P2X3 receptor function (study II, Fig.7E, F). This result suggested that this NBP does not exert any direct action on the P2X3 receptor.

However, we used the ATP-luminescent assay and observed that zoledronate triggered the release of ATP from cells to the extracellular milieu. Thus, the 24 h treatment of the TG neuronal culture with 100 µM zoledronate significantly increased the concentration of extracellular ATP from 0.70 ± 0.26 nM (n=12) to 3.2 ± 0.4 nM (p=0.018, n=6; study II, Fig.7G).

Such strong effect of zoledronate may indicate that other purines can be released in a similar manner during bone cancer treatment with BPs. For instance, we can hypothesize that not

only generation of ApppI but also the release of endogenous ApppI could be initiated during prolonged zoledronate treatments.

Further, we clarified the lack of involvement in the inhibitory action of ApppI of its two metabolites, IPP and AMP since potentially these compounds could also contribute to the inhibition of P2X3 receptors during ApppI exposure. Thus, we tested the action of IPP and AMP on the P2X3 receptors. However, application of 1 µM IPP or 1 µM AMP on P2X3 receptors did not exert any inhibitory effect (study II, Fig.7, C and D), suggesting that ApppI has to exist as the whole molecule in order to achieve the inhibition of P2X3 receptors.

5.3 THE ROLE OF THE LEFT FLIPPER FOR P2X7 RECEPTOR FUNCTION