Pharmacological Profile of the Purinergic P2Y Receptors That Modulate, in Response to ADPβS, the Vasodepressor Sensory CGRPergic Outflow in Pithed Rats

Calcitonin gene-related peptide (CGRP), an endogenous neuropeptide released from perivascular sensory nerves, exerts a powerful vasodilatation. Interestingly, adenosine triphosphate (ATP) stimulates the release of CGRP by activation of prejunctional P2X2/3 receptors, and adenosine 5′-O-2-thiodiphosphate (ADPβS), a stable adenosine diphosphate (ADP) analogue, produces vasodilator/vasodepressor responses by endothelial P2Y1 receptors. Since the role of ADP in the prejunctional modulation of the vasodepressor sensory CGRPergic drive and the receptors involved remain unknown, this study investigated whether ADPβS inhibits this CGRPergic drive. Accordingly, 132 male Wistar rats were pithed and subsequently divided into two sets. In set 1, ADPβS (5.6 and 10 µg/kg·min) inhibited the vasodepressor CGRPergic responses by electrical stimulation of the spinal T9–T12 segment. This inhibition by ADPβS (5.6 µg/kg·min) was reverted after i.v. administration of the purinergic antagonists MRS2500 (300 µg/kg; P2Y1) or MRS2211 (3000 µg/kg; P2Y13), but not by PSB0739 (300 µg/kg; P2Y12), MRS2211 (1000 µg/kg; P2Y13) or the KATP blocker glibenclamide (20 mg/kg). In set 2, ADPβS (5.6 µg/kg·min) failed to modify the vasodepressor responses to exogenous α-CGRP. These results suggest that ADPβS inhibits CGRP release in perivascular sensory nerves. This inhibition, apparently unrelated to activation of ATP-sensitive K+ channels, involves P2Y1 and probably P2Y13, but not P2Y12 receptors.


Introduction
Calcitonin gene-related peptide (CGRP) is a member of the endogenous peptides family, formed by 37 amino acids [1], which (i) was identified in plasma in the 1980s and subsequently in the spinal cord [2]; (ii) produces a potent vasodepressor effect, erythema and an increase in local blood flow [3,4]; and (iii) is mostly released by C sensory nerve fibres mainly emerging from dorsal root ganglia, trigeminal ganglia and heterogeneous small and medium-sized neurons [5][6][7]. Once released by sensory neurons, CGRP binds to the CGRP receptor, which is coupled to G αs proteins and consists of two proteins, namely: (i) calcitonin-like receptor (CLR) and (ii) receptor activity modifying protein (RAMP 1 ) [8,9].
Our results suggest that ADPβS-induced inhibition of the vasodepressor sensory CGRPergic outflow, which seems to be unrelated to activation of ATP-sensitive K + channels, could be mediated by activation of prejunctional P2Y 1 and probably P2Y 13 , but not P2Y 12 , receptors.

Effect of Vehicle or ADPβS Infusions on the Vasodepressor Responses by Electrical Sensory Stimulation or Exogenous α-CGRP
Electrical sensory stimulation and the i.v. bolus administration of exogenous α-CGRP resulted in frequency-dependent (upper panel) and dose-dependent (lower panel) decreases in DBP represented as a percentage change in DBP. When comparing the vasodepressor responses of the control subgroup with those of the subgroup receiving vehicle (bidistilled water; 0.02 mL/min), no significant differences were found in the responses produced by sensory electrical stimulation or exogenous α-CGRP (p > 0.05) ( Figure 1A,E). Thus, these vasodepressor responses were reproducible during our experimental protocols.
In contrast, the infusions of 5.6 and 10 µg/kg·min ADPβS produced a significant inhibition (compared to control) of the vasodepressor responses generated by sensory electrical stimulation at 1.8, 3.1 and 5.6 Hz ( Figure 1C,D; p < 0.05), whereas 3 µg/kg·min ADPβS was inactive ( Figure 1B; p > 0.05). However, as shown in Table 2, 5.6 and 10 µg/kg·min ADPβS produced practically the same degree of inhibition; in other words, this inhibition was not dose dependent. On this basis, the infusion dose of 5.6 µg/kg·min ADPβS was selected to investigate (i) its effect on the vasodepressor responses produced by exogenous α-CGRP; and (ii) the pharmacological profile of the P2Y receptors mediating ADPβS-induced inhibition of the vasodepressor sensory CGRPergic drive. Indeed, Figure 1F indicates that 5.6 µg/kg·min ADPβS failed to inhibit the vasodepressor responses to exogenous α-CGRP (contrasting with Figure 1C). With these results, we suggest that the inhibition by ADPβS of the vasodepressor sensory CGRPergic drive is prejunctional in nature, but this does not shed further light on the specific pharmacological profile of the P2Y receptors involved.
ADPβS produced practically the same degree of inhibition; in other words, this inhibit was not dose dependent. On this basis, the infusion dose of 5.6 μg/kg•min ADPβS w selected to investigate (i) its effect on the vasodepressor responses produced by exogen α-CGRP; and (ii) the pharmacological profile of the P2Y receptors mediating ADPβS duced inhibition of the vasodepressor sensory CGRPergic drive.
Indeed, Figure 1F indicates that 5.6 μg/kg•min ADPβS failed to inhibit the vaso pressor responses to exogenous α-CGRP (contrasting with Figure 1C). With these resu we suggest that the inhibition by ADPβS of the vasodepressor sensory CGRPergic dr is prejunctional in nature, but this does not shed further light on the specific pharma logical profile of the P2Y receptors involved.   One fundamental experimental condition that would facilitate the pharmacological analysis of ADPβS-induced inhibition of the neurogenic vasodepressor sensory CGRPergic responses is that these responses remain unaffected after the administration of a vehicle or P2Y receptor antagonists alone. For this purpose, Figure 2 compares the control vasodepressor CGRPergic responses by electrical stimulation (without treatment) with those produced after i.v. treatment with (A) bidistilled water (1 mL/kg; Figure 2A); (B) MRS2500 (300 µg/kg; P2Y 1 antagonist, Figure 2B); (C) PSB0739 (300 µg/kg; P2Y 12 antagonist, Figure 2C); (D) MRS2211 (1000 µg/kg; P2Y 13 antagonist, Figure 2D); and (E) MRS2211 (3000 µg/kg; P2Y 13 antagonist, Figure 2E). of exogenous α-calcitonin gene-related peptide (α-CGRP). Data are shown as percentage (%) change in diastolic blood pressure. Solid symbols indicate a significant difference (p < 0.05) against the corresponding control response (○). Values are indicated as means ± SEM (n = 6 for each subgroup). Table 2. Percentage of the inhibition produced by the infusions (after 10 min) of vehicle (bidistilled water; 0.02 mL/min) or ADPβS (3, 5.6 and 10 μg/kg•min) on the vasodepressor CGRPergic responses (peak changes) by electrical sensory stimulation in pithed rats.

General
The pithed rat is a well-established experimental model for studying cardiovascular function [4,51] and has been further optimised for investigating the pharmacological profile of the receptors that modulate, at the peripheral level, the activity of the sympathetic and sensory CGRPergic nerve terminals that innervate the cardiovascular system [23,[26][27][28][29][30]36,[52][53][54][55][56][57]. Since central functions are not operative in this model, any change in blood pressure (and/or heart rate) produced by i.v. administration of any compound can be

General
The pithed rat is a well-established experimental model for studying cardiovascular function [4,51] and has been further optimised for investigating the pharmacological profile of the receptors that modulate, at the peripheral level, the activity of the sympathetic and sensory CGRPergic nerve terminals that innervate the cardiovascular system [23,[26][27][28][29][30]36,[52][53][54][55][56][57]. Since central functions are not operative in this model, any change in blood pressure (and/or heart rate) produced by i.v. administration of any compound can be attributed exclusively to peripheral (rather than central)  [23,[26][27][28][29][30]36,[52][53][54][55][56][57]. On this basis, within the context of the present study, it is reasonable to assume that the inhibition of the vasodepressor sensory CGRPergic drive produced by ADPβS is peripheral in nature (i.e., at the level of perivascular sensory CGRPergic nerves, and unrelated to baroreceptor compensatory reflex mechanisms or central actions).
Hence, the present study has analysed the pharmacological profile of the purinergic P2Y receptors modulating the functionality of the CGRPergic neurovascular junction at the specific level of systemic resistance blood vessels, which are determinant for peripheral vascular tone and, consequently, for DBP. For this purpose, ADPβS (which is a preferential agonist at purinergic P2Y 1 , P2Y 12 and P2Y 13 receptors [33,36,39,47,58]) was used as it has recently been shown to produce (when given i.v.) (i) acute vasodepressor responses in anaesthetized rats [33]; and (ii) cardiac sympatho-inhibition in pithed rats by activation of purinergic P2Y 12 receptors, and less prominently by P2Y 13 receptors [36].
Our results show that ADPβS (5.6 µg/kg·min) is capable of producing a prejunctional inhibition of the vasodepressor sensory CGRPergic drive (implying an inhibition of CGRP release from perivascular sensory nerves) as it induced (i) inhibition of the vasodepressor responses produced by electrical stimulation of perivascular sensory CGRPergic nerves ( Figure 1C); and (ii) no effect on the vasodepressor responses produced by exogenous α-CGRP ( Figure 1F).
Moreover, it is to be noted that the electrically induced CGRP release from perivascular sensory nerves was not directly measured in our experiments but, alternatively, was determined by the evoked vasodepressor responses, as previously reported [26][27][28][29][30]54,55], which are specifically blocked by CGRP receptor antagonists [4,57].
On the other hand, since glibenclamide blocked the response to ADPβS (Table 1), it would seem tempting to suggest the blockade of K ATP channels a priori, given that these channels play a role in adenosine-induced vasodilatation [63]. Despite the fact that the glibenclamide vehicle also blocked this response (Table 1), we still considered it important to analyse the effect of glibenclamide on the ADPβS-induced inhibition of the vasodepressor sensory CGRPergic drive.

Effect of ADPβS on the Vasodepressor Sensory CGRPergic Drive
As shown in Figure 1 (upper panel), only 5.6 and 10 µg/kg·min ADPβS induced a significant inhibition of the vasodepressor CGRPergic responses by electrical stimulation at 1.8, 3.1, 5.6 Hz, but the degree of inhibition produced by the two infusions was practically identical (Table 2), probably producing a maximal (5.6 µg/kg·min) and a supramaximal (10 µg/kg·min) inhibition. Consequently, 5.6 µg/kg·min ADPβS was selected for the subsequent pharmacological analysis with exogenous α-CGRP and P2Y receptor antagonists. Hence, the fact that the vasodepressor responses to exogenous α-CGRP were not significantly modified (p > 0.05) by 5.6 µg/kg·min ADPβS (Figure 1, lower panel) suggests that ADPβS (i) inhibits the electrically-induced vasodepressor CGRPergic responses by activating prejunctional receptors; and (ii) does not interact with post-junctional (vascular musculotropic) receptors that might oppose CGRP-induced vasodilatation, for example, by activation of P2Y 1 receptors producing vasoconstriction [42,50,59].

Inhibition of the Vasodepressor Sensory CGRPergic Drive by ADPβS: Possible
Pharmacological Correlation with the Purinergic P2Y 1 , P2Y 12

and P2Y 13 Receptor Subtypes
Once the prejunctional inhibition by ADPβS of the vasodepressor sensory CGRPergic drive was established (Figure 1), our next step was to analyse the pharmacological profile of this response. For this purpose (as shown in Table 3), it is important to consider that (i) ADPβS can activate (and displays affinity for) P2Y 1 , P2Y 12 and P2Y 13 receptors [37,39,46,49]; and (ii) some antagonists for these receptors, which include MRS2500 (P2Y 1 ), PSB0739 (P2Y 12 ) and MRS2211 (P2Y 13 ) [49], display specific binding affinities for these receptors. Values are presented as pK i /pEC 50 /pIC 50 in human receptors. No effect was observed until a 100 µM; b 1 µM; c 10 µM. Data taken and modified from [36,37].
As a first step of our pharmacological investigation, we decided to explore the possible effects of each of these antagonists on the neurogenic vasodepressor CGRPergic responses produced by electrical stimulation. For this reason, they were administered during i.v. continuous infusions of the ADPβS vehicle (bidistilled water). Since no significant differences versus the control subgroup were found, this finding reinforces the view that these antagonists have no effects on baseline DBP (Table 1) and on the neurogenic vasodepressor CGRPergic responses (Figure 2). On the basis of their affinities shown in Table 3, their profile of blockade of cardiovascular responses in pithed rats [36], and the dosage considerations described above, the fact that the ADPβS-induced sensory inhibition was only reversed by i.v. MRS2500 (300 µg/kg) or MRS2211 (3000 µg/kg), but not by PSB0739 (300 µg/kg) or MRS2211 (1000 µg/kg) (Figure 3) suggests the main involvement of P2Y 1 , and probably P2Y 13 , but not P2Y 12 , receptors.
Our findings, suggesting the possible involvement of prejunctional P2Y 13 receptors inhibiting CGRP release from perivascular sensory nerves ( Figure 3E), are consistent with other studies reporting that ADPβS inhibits (i) CGRP release from rat sensory neurons in dural arteries and trigeminal ganglion by MRS2211-sensitive P2Y 13 receptors [33]; and (ii) noradrenaline release from cardioaccelerator sympathetic nerves in pithed rats by activation of purinergic P2Y 12 receptors and less prominently by P2Y 13 receptors [36].
Regarding the possible transduction mechanisms of P2Y 13 receptors associated with inhibition of neuronal CGRP release, some in vitro studies indicate that P2Y 13 receptors have several transduction pathways [74][75][76] including, amongst others: (i) G i/o protein activation with ADP, leading to inhibition of adenylate cyclase with a resulting decrease in cAMP production; (ii) phosphorylation of the PI3K/Akt/GSK3 axis that produces release of β-catenin and Nrf2 (transcription factors) promoting cell survival; and (iii) G αq coupling, with a resulting increase in [Ca ++ ] and activation of phospholipase C/PKC/ERK/CREB7DUSP2. Moreover, the βγ subunits can activate RhoA with a resulting decrease in Ca ++ channel activity that modulates neurotransmitter release [74].
Our findings supporting the role of prejunctional P2Y 1 , and probably P2Y 13 , receptors in the inhibition of CGRP release from perivascular sensory nerves may complement the general concept of purinergic modulation of CGRP release in sensory neurons. With this concept in mind, activation of sympathetic postganglionic neurons results in the release of noradrenaline and ATP as a cotransmitter; in turn, ATP would activate P2X 2/3 receptors (ATP-gated Na + , K + and Ca ++ channels [59,77]) on sensory nerves with an increase in CGRP release [24]. Subsequently, ATP at the neuroeffector junction would be hydrolysed to ADP by ecto-nucleoside triphosphate diphosphohydrolase (ecto-NTPDase 2,3,8) [78]; then ADP could stimulate P2Y 1 and P2Y 13 receptors on sensory neurons (as suggested in the present study) with a decrease in CGRP release.

Are K ATP Channels Involved in the Inhibition of the Vasodepressor CGRPergic Drive by ADPβS?
K ATP channels are expressed in vascular smooth muscle and modulate vascular tone, blood flow and blood pressure; when opened, they produce membrane hyperpolarization of vascular smooth muscle, relaxation and vasodilation [79][80][81][82]. Hence, glibenclamide (a K ATP channel blocker) was used to pharmacologically discern the possible role of K ATP channels in the CGRPergic sensory inhibition produced by ADPβS ( Figure 5). However, under our experimental conditions, glibenclamide, which had no effect on DBP (Table 1), attenuated the vasodepressor sensory CGRPergic drive ( Figure 5B), as previously reported [53]. Certainly, this effect could have overshadowed the ADPβS-induced sensory inhibition and would help explain why glibenclamide failed to revert ADPβS-induced inhibition of the vasodepressor sensory CGRPergic drive (compare Figure 5C with Figure 5A).
This inactivity of glibenclamide, notwithstanding, does not seem to be a compelling finding to rule out the role of K ATP channels in ADPβS-induced prejunctional sensoryinhibition because we hypothesise that two fundamental mechanisms are operative in our experimental model, namely: (i) ADPβS-induced hyperpolarization of CGRPergic sensory nerves; and (ii) CGRP-induced systemic vasodilatation.
On the other hand, the CGRP-induced systemic vasodilatation (via the activation of G αs protein-coupled CGRP receptors) is mediated by two pathways: (i) direct smooth muscle vasorelaxation involving activation of adenylate cyclase and, consecutively, an increase in cAMP levels, PKA activity and phosphorylation of K ATP channels; and (ii) endothelial vasorelaxation resulting from a sequential increase in PKA activity, NO production diffusing to vascular smooth muscle, guanylate cyclase activity, cGMP levels, and phosphorylation of K ATP channels leading to vasodilation [10,13,14,18]. We would, finally, like to put forward (with no direct experimental evidence) that these transduction mechanisms activated by ADPβS at vascular level might also occur in perivascular sensory CGRPergic nerves.

Limitations of the Study
Based on the above, and considering the neurovascular junction, it is clear that glibenclamide may have blocked K ATP channels at both prejunctional (perivascular sensory nerves) and postjunctional (vascular) levels and that, as a result, may have inhibited the actions of ADPβS and CGRP, respectively. For this reason, it was not possible to discern, under our experimental conditions, the actions of glibenclamide (blocking K ATP channels) at prejunctional and postjunctional levels. These experimental limitations may be approached in other studies with additional technologies involving, among others, molecular biology and immunohistochemistry.
On the other hand, we have to recognize that (i) the comparison of affinities of agonists and antagonists at P2Y 1 , P2Y 12 and P2Y 13 receptors, shown in Table 3, consists of data obtained from human P2Y receptors; and (ii) as far as we know, this binding data comparison does not exist for rodents. Nevertheless, these binding data may be transferrable from humans to rodents for several reasons: (i) for ADPβS, the affinity is the same for human and rat P2Y 12 receptors [65,83], but there are only limited differences between rat and human P2Y 13 receptors [64]; (ii) for ADP, only comparable affinities exist for P2Y 1 and P2Y 13 receptors, which is equipotent on human and rat P2Y 1 receptors [66,84], but it seems slightly more potent on human than on rat P2Y 13 receptors [85]. The main finding of the present study is the blockade produced by MRS2500 on P2Y 1 receptors ( Figures 3B and 4D), which displays a comparable affinity for human and rat P2Y 1 receptors [86]. To our knowledge, rat binding data do not exist for both PSB0739 and MRS2211; however, for Ticagrelor (an FDA approved P2Y 12 receptor antagonist), there was no difference in affinity for rodent and human P2Y 12 receptors [87], suggesting a similar pharmacology.

Perspectives and Potential Clinical Significance
Purinergic P2Y receptors play an important role in numerous cardiovascular diseases including endothelial dysfunction, which is characterized by vasoconstriction, increased vascular permeability and a prothrombotic and proinflammatory state [44,46]. On the other hand, it has been suggested that CGRP is involved in cardiovascular pathologies such as hypertension [12,15,17,20,88] or neurovascular disorders such as migraines [6,33,89].
Based on the inactivity of PSB0739 (300 µg/kg; Figure 3C) to revert ADPβS-induced sensory inhibition, our results imply that, in healthy animals, there is no physiological relevance of purinergic P2Y 12 receptors modulating CGRP release from perivascular sensory nerves. In keeping with this view, P2Y 12 receptors are highly expressed on platelets and megakaryocytes, exerting a prothrombotic function. Nevertheless, in pathological conditions such as hypoxia, heart failure, hypertension, sepsis, atherosclerosis, tissue damage and inflammation (among others) P2Y 12/13 receptors become relevant, generating (at an endothelial level) increased permeability, thrombosis and angiogenesis [46]. Significantly, the effectivity of the P2Y 1 receptor antagonist MRS2500 (300 µg/kg; Figures 3B and 4D) to revert ADPβS-induced inhibition strongly suggests that purinergic P2Y 1 receptors may play a role in modulating the release of CGRP at a prejunctional level, in addition to their vasodilator effects. This would strengthen the role of P2Y 1 receptors in vascular diseases such as hypertension and migraines.

General Methods
A total of 132 male normotensive Wistar rats (380-420 g, 18-22 weeks of age) were used in the present investigation. The animals were maintained at 22 ± 2 • C room temperature, 50% humidity and a 12/12-h light/dark cycle (light beginning at 07:00 h) with food and water freely available in their home cages.
After bilateral cervical vagotomy, the rats were cannulated with polyethylene catheters which were placed in: (i) the left and right femoral veins for the continuous infusions of methoxamine and ADPβS (or vehicle), respectively; (ii) the left jugular vein for the continuous infusion of hexamethonium; and (iii) the right jugular vein, for the bolus injections of gallamine or the P2Y receptor antagonists (of vehicles). Subsequently, the left carotid artery was connected to a Grass pressure transducer (P23 XL), for the recording of blood pressure. Both, heart rate (measured with a 7P4F tachograph) and blood pressure were recorded simultaneously by a model 7D Grass polygraph (Grass Instrument Co., Quincy, MA, USA). The body temperature of each pithed rat (monitored with a rectal thermometer) was maintained at 37 • C by a lamp".

Protocol I: Selective Electrical Stimulation of the Vasodepressor Sensory CGRPergic Drive
In the first set (n = 114), "the stainless-steel rod was replaced by an enamelled electrode whose uncovered segment was located at T 9 -T 12 of the spinal cord to allow selective stimulation of the vasodepressor sensory CGRPergic drive", as previously reported [26][27][28][29][30]54,55,57]. Before electrical stimulation, the animals were pre-treated with gallamine (25 mg/kg, i.v.), a nondepolarizing neuromuscular blocking agent, to avoid the electrically induced muscular twitching [25,36,56]. In order to obtain vasodepressor responses, "DBP was initially increased and maintained at around 100-120 mm Hg by an i.v. continuous infusion of methoxamine (15-20 µg/kg·min) during and until the end of the experiments", as previously established by our group [26][27][28][29][30]54,55,57]. Then, the animals received i.v. continuous infusions of hexamethonium (2 mg/kg·min), a nicotinic ganglion blocker, to block the sympathetic vasopressor responses generated by electrical stimulation of the spinal T 9 segment [4,25,56,57]. When haemodynamic conditions were stable, baseline values of heart rate and DBP (a more accurate indicator of peripheral vascular resistance) were determined [36], and the 114 animals were then divided into six groups (n = 18, 24, 12, 12, 24, 24, respectively) for spinal T 9 -T 12 electrical stimulation (see Figure 6). It must be emphasised that, prior to electrical spinal stimulation (and also prior to i.v. bolus injections of exogenous α-CGRP) to produce vasodepressor responses (depending on the specific protocol for each subgroup; see below), 10 min were allowed to elapse after each i.v. bolus injection of compound and after each i.v. continuous infusion of compound.
Spinal T 9 -T 12 electrical stimulation consisted of applying trains of 10 s to selectively stimulate the vasodepressor sensory CGRPergic drive (monophasic rectangular pulses of 2 ms and 50 V) at increasing frequencies (0.56, 1.0, 1.8, 3.1 and 5.6 Hz). When DBP had returned to baseline levels, the next frequency was applied (at intervals of about 5-10 min) until the S-R curve was completed (around 50 min).
The doses of the above antagonists/blockers have been shown to abolish the responses mediated by their corresponding receptors/mechanisms in pithed rats [36,53].

Supplementary Procedures
It is to be noted that the doses of (i) vehicle (bidistilled water) or ADPβS were continuously infused (i.v.) at a rate of 0.02 mL/min by a KDS100 model infusion pump (KD Scientific Inc., Holliston, MA, USA); and (ii) vehicles or antagonists were given as i.v. bolus injections in volumes of 1 mL/kg.
The intervals between the different stimulation frequencies or α-CGRP doses depended on the duration of the vasodepressor responses (5 min), as we waited until DBP had returned to baseline values.

Data Presentation and Statistical Evaluation
"All data in the text and figures are presented as the mean ± SEM. The peak changes in DBP by electrical stimulation or exogenous α-CGRP were expressed as the percent change from baseline", as previously described in pithed rats [26][27][28][29][30]54,55,57]. "A one-way ANOVA was used to compare the absolute values of DBP obtained during the continuous infusions of methoxamine, before and 10 min after the administration of all compounds before starting the electrical stimulation.
Moreover, the decreases in DBP induced electrically or by exogenous α-CGRP in the different subgroups of animals were evaluated with the Student-Newman-Keuls post hoc test, once a two-way repeated measures ANOVA (randomized block design) showed that the samples represented different populations [93]", as reported in previous studies [26][27][28][29][30]54,55,57]. Statistical significance was accepted at p < 0.05. Statistical analysis was performed using SigmaPlot 12.0 (Systat Software, Inc. SigmaPlot for Windows).
The graphics were performed using Prism 6 software (GraphPad Software, Inc., San Diego, CA, USA).

Conclusions
Our results, taken together, allow us to suggest that the inhibition of vasodepressor sensory CGRPergic outflow produced by 5.6 mg/kg·min of ADPβS in healthy pithed rats (i) is apparently unrelated to activation of ATP-sensitive K + channels; and (ii) could be mediated by activation of prejunctional P2Y 1 and probably P2Y 13 , but not P2Y 12 , receptors.