Capsaicin-sensitive extrinsic sensory nerves and submucosal vasodilator neurons provide important vasodilator input to submucosal arterioles, but relatively little is known about the signaling between these populations and the sympathetic vasoconstrictor innervation. This study examined whether release of sympathetic purines can modulate dilator nerves. In vitro submucosal preparations from guinea pig ileum were modified to leave the parent mesenteric artery intact so that perivascular sympathetic and extrinsic afferent nerves could be activated by a bipolar stimulating electrode placed on the parent artery, and submucosal vasodilator neurons were activated using focal electrodes placed on submucosal ganglia. The outside diameter of submucosal arterioles was monitored using videomicroscopy, and dilator responses were examined after preconstricting vessels 80–95% with prostaglandin F2α (400 nM). Mesenteric nerve stimulation evoked a frequency-dependent dilation, with suramin (100 μM) present throughout to inhibit P2X receptor-mediated vasoconstrictions. In the presence of guanethidine (10 μM) to inhibit sympathetic purine release, superfusion of ATP (200 nM-6 μM) caused a concentration-dependent inhibition of nerve-evoked dilations. Vasodilations to substance P (10 nM) were not inhibited by ATP in the presence of guanethidine, implicating a presynaptic effect of ATP on neurotransmitter release. The inhibitory effect of ATP was blocked by the adenosine receptor antagonist 8-phenyltheophylline (8-PT; 10 μM). In addition, 8-PT increased the amplitude of nerve-evoked dilations, suggesting a tonic inhibitory effect of adenosine receptors on vasodilator release. Dilations evoked by electrical stimulation of submucosal ganglia were also inhibited almost 50% by ATP (2 μM) and its nonhydrolyzable analog, α,β-methylene-ATP (10 μM). These data suggest that sympathetic varicosities release ATP or a related purine that can act at presynaptic adenosine receptors on extrinsic sensory and submucosal vasodilator neurons to inhibit neurotransmitter release.
- purinergic neurotransmission
- extrinsic afferent neuron
- substance P
effective mucosal function in the intestine is highly dependent on adequate mucosal blood flow. Regulation of this blood flow occurs predominantly at the level of the submucosal arterioles, which are the final resistance vessels in the gastrointestinal tract (10, 18, 33). Neural reflexes play a key role in this regulation, in both physiological and pathophysiological settings. These reflexes involve the combined actions of vasoconstrictor and vasodilator innervation of submucosal arterioles. Enteric motor neurons with cell bodies in submucosal ganglia innervate and vasodilate arterioles, and the axons of sympathetic postganglionic neurons act as vasoconstrictors of submucosal arterioles (33). In addition, the axon collaterals of extrinsic primary afferent neurons with cell bodies in dorsal root ganglia release neuropeptides, including substance P, which also cause vasodilation (30). Classically, it is believed that the neural regulation of blood flow is achieved by the balance of these opposing postjunctional vasodilator and vasoconstrictor effects on arteriolar lumen. However, it is likely that a more complex regulation exists, including presynaptic inhibitory effects of neurotransmitters on the release of other vasoactive transmitters. There are many examples of this within the enteric nervous system, including norepinephrine activation of presynaptic α2-adrenoceptors, which inhibits neurotransmitter release from several enteric neuron subtypes, and presynaptic inhibitory effects of muscarinic receptor activation (8, 34).
ATP has been identified as the major sympathetic vasoconstrictor in submucosal arterioles of guinea pig ileum (7). Ectonucleotidases, including CD39, are thought to play important roles in regulating purinergic receptor activation by catalyzing the hydrolysis of ATP (35). One of the end products of ATP catabolism is adenosine, which can activate metabotropic adenosine receptors (6, 35). Several studies of vas deferens have established a prejunctional role for adenosine receptors in the modulation of norepinephrine release (5, 11). What remains to be determined is whether purine release from sympathetic vasoconstrictor nerves in the gut can modulate vasodilator pathways by activating adenosine receptors. This question is of particular interest given that recent studies suggest that purinergic neurotransmission is altered during intestinal inflammation, due to the upregulation of purinergic catabolism which would be predicted to increase adenosine levels (2, 17, 23). In the present study, we exploited the ability to monitor submucosal arteriole diameter in vitro using a videomicroscopy system while simultaneously stimulating vasoconstrictor and intrinsic or extrinsic vasodilator neurons to examine the purinergic presynaptic interactions between these neural pathways.
MATERIALS AND METHODS
All methods were approved by the Animal Care Committee at Queen's University and conform with the principles and guidelines of the Canadian Council on Animal Care. Guinea pigs (150–225 g; Charles River, Montreal, Canada) were anesthetized with isoflurane inhalation and killed by cervical transection and exsanguination. Distal ileum with attached mesentery was removed, opened along the mesenteric border, and pinned flat with the mucosa facing upward in a Sylgard-lined (Dow-Corning, Midland, MI) petri dish containing Krebs solution (in mM: 126 NaCl, 2.5 NaH2PO4, 1.2 MgCl2, 2.5 CaCl2, 5 KCl, 25 NaHCO3, and 11 glucose) that had been bubbled with 95% O2 and 5% CO2. Submucosal preparations obtained following removal of the mucosa and muscularis externae were pinned in small Sylgard-lined organ baths, mounted on the stage of an inverted microscope, and continuously superfused with Krebs solution warmed to yield a bath temperature of 35–36°C. Extrinsic afferent and sympathetic nerves course along mesenteric arteries and follow these vessels as they branch into the submucosal arterioles (Fig. 1A). These nerves were stimulated electrically (20 Hz, 2 s) by placing a bipolar electrode on the parent artery just proximal to where it entered the submucosa, 4–5 mm from where arteriolar responses were measured (Fig. 1A). The electrode location allowed simultaneous stimulation of sympathetic and extrinsic sensory axons. Frequency-response relationships were studied using constant-duration (2-s) trains of stimuli at varying frequencies. Therefore, the number of stimulus pulses applied increased with higher frequencies. Submucosal ganglia were stimulated with a focal monopolar electrode (20 Hz, 2 s) to examine interaction between sympathetic and enteric vasomotor pathways.
Vasoconstrictions were monitored by continuously measuring the outside diameter of individual submucosal arterioles using a computer-assisted videomicroscopy system (Diamtrak; Flinders University of South Australia), as previously described (21). Briefly, an Imaging Technology PCVision frame-grabber board in an IBM PC-AT computer was used to digitize television images of the arteriole. This was converted to an analog signal and stored on a chart recorder. The resolution of the system was <1 μm. The vasodilator responses were monitored by first preconstricting the arterioles to 80–95% of the maximum they could constrict from their resting diameter using 400 nM 9,11-dideoxy-11α9α-epoxy-methanoprostaglandin F2α. The magnitude of the dilation was expressed in two ways, one represented the area under the curve and the other represented the peak amplitude of the dilation. The area under the curve was used in measurement of all nerve-evoked dilations and peak amplitude when measuring responses to exogenous vasodilators. Peak dilation amplitude measurements were expressed as a percentage of the maximal dilation of the vessel, as we have previously reported (29). Previous studies have shown that the maximal dilation evoked by exogenous application of muscarine is ∼85% of the preconstricted diameter, i.e., a maximal stimulus does not fully dilate the vessel back to the resting diameter (22). Therefore, the percentage dilation was calculated by dividing the peak amplitude of the dilation by 85% of amplitude of the prostaglandin-evoked constriction before the onset of the dilation. The area under the curve was estimated as the product of the amplitude times the duration at half peak amplitude, as we have previously described (24).
The drugs used were purchased from Sigma Aldrich (St. Louis, MO) unless noted otherwise: α,β-methylene-ATP, ATP, substance P, 4–4-diphenylacetoxy-N-methylpiperidine methiodide (4-DAMP), prazosin, 8-phenyltheophylline (8-PT), idazoxan, guanethidine, tetrodotoxin, capsaicin, suramin (Tocris, Ellisville, MO), and prostaglandin F2α (U-46619; Caymen Chemical, Ann Arbor, MI). Stock solutions of drugs were diluted in oxygenated Krebs and added to preparations by superfusion. Capsaicin was dissolved in Tween 80, alcohol, and saline (10:10:80%). Data are expressed as means ± SE, and the effects of drug treatments were compared with paired two-way Student's t-tests. A P < 0.05 was taken to indicate statistical significance.
Electrical field stimulation of mesenteric nerves resulted in large transient vasoconstrictions in arterioles that had not been preconstricted by superfusion of prostaglandin F2α (Fig. 1B). Consistent with previous reports (7), application of the nonselective P2 purinoceptor antagonist suramin (100 μM) abolished the vasoconstriction, implicating ATP or a related purine as the major sympathetic vasoconstrictor in submucosal arterioles of guinea pig ileum. When submucosal arterioles were preconstricted by prostaglandin F2α (400 nM), mesenteric nerve stimulation evoked a small transient vasoconstriction, followed by a long-lasting (>10 s) vasodilation (Fig. 1B). The remaining experiments in this study were performed in prostaglandin F2α preconstricted arterioles. Superfusion of tetrodotoxin (1 μM) abolished the constriction (n = 8 arterioles) and markedly reduced but did not abolish the dilation response to nerve stimulation (58.3 ± 11.8% inhibition, n = 5). The M3 receptor antagonist 4-DAMP (1 μM) did not affect the vasodilator response to nerve stimulation (data not shown), suggesting that cholinergic vasomotor dilator pathways in the submucosal plexus were not activated following mesenteric nerve stimulation. These findings suggest vasodilation of submucosal arterioles during mesenteric nerve stimulation is the result of activation of capsaicin-sensitive extrinsic afferent nerves, consistent with previous reports (30–32).
Guanethidine is taken up into sympathetic varicosities and inhibits the action potential-induced release of neurotransmitter. We examined whether guanethidine (10 μM) affected vasodilation amplitude by measuring the area of submucosal arteriolar dilation in preconstricted arterioles in the presence of suramin. Superfusion of guanethidine for 5 min markedly increased the area of vasodilations during mesenteric nerve stimulation (Fig. 2; n = 4 arterioles), which is consistent with sympathetic varicosities releasing a transmitter that inhibits extrinsic afferent nerve-induced vasodilation.
ATP is the major vasoconstrictor neurotransmitter released from sympathetic varicosities innervating submucosal arterioles in the guinea pig ileum (7). Therefore, we examined whether ATP could mimic the sympathetic inhibition of electrical field stimulation-induced vasodilation in the presence of suramin to block postjunctional P2 receptors. ATP at concentrations between 0.2 and 6 μM caused a concentration-dependent reduction in the area of nerve-evoked vasodilations (Fig. 3). Our next step was to examine whether this inhibitory effect was mediated via prejunctional inhibition of neuropeptide release or reduction of postjunctional responsiveness to neuropeptides. To distinguish between these two possibilities, we examined the effect of ATP, in the presence of suramin (which was present for 3 min before ATP superfusion), on the postjunctional response to superfusion of substance P (Fig. 4). Substance P (10 nM) caused large vasodilations in preconstricted arterioles that were not affected by prior superfusion of ATP (n = 6; paired t-test). Taken together, these data suggest that ATP or a metabolite acts presynaptically on non-P2 receptors to inhibit neuropeptide release from extrinsic afferent nerve terminals.
Several studies have previously documented presynaptic inhibitory effects of adenosine receptors (25–28) on the release of neuropeptides from extrinsic afferent nerve terminals. Given the role of ectonucleotidases in regulating the extracellular concentration of ATP by hydrolyzing it to adenosine (35), an agonist of A1 receptors, we hypothesized that adenosine receptor activation may underlie the presynaptic inhibitory effect of sympathetic neurotransmitter release on extrinsic afferent terminals. Superfusion of the nonselective adenosine receptor antagonist 8-PT (10 μM) enhanced the vasodilator effect of stimulating mesenteric nerves (n = 5; Fig. 5). These data indicate that sympathetic release of ATP or a related purine leads to activation of presynaptic A1 receptors on extrinsic afferent nerve terminals. Consistent with this, 8-PT reversed the inhibition of nerve-evoked vasodilation by exogenous ATP in the presence of suramin (n = 3 arterioles; Fig. 6). We next examined whether the nonhydrolyzable analog to ATP, α,β-methylene-ATP (10 μM), could presynaptically inhibit neuropeptide release. Dilations to nerve stimulation in the presence of guanethidine and suramin (1.74 ± 0.3 mm2) were significantly reduced by superfusion of α,β-methylene-ATP for 3 min before nerve stimulation (0.98 ± 0.23 mm2) (n = 5; P < 0.05; paired t-test).
We examined whether the inhibition of neurotransmitter release by ATP also occurred in the other vasodilator neuron population: submucosal cholinergic vasodilator neurons. To do this, we monitored vasodilations, in response to focal electrical stimulation of adjacent submucosal ganglia, of submucosal arterioles that were preconstricted with prostaglandin F2α (400 nM). As previously reported (22), these vasodilations were abolished by an M3 muscarinic receptor antagonist (data not shown). Comparison of nerve-evoked vasodilation amplitude in the presence of guanethidine (10 μM) and suramin (100 μM) before and during superfusion of 2 μM ATP revealed a significant inhibition of vasodilation by the purine (control: 35 ± 11% dilation vs. ATP: 18.8 ± 7.3% dilation; n = 6, P < 0.05 paired t-test). This suggests that purines can also presynaptically inhibit neurotransmitter release from enteric vasodilator neurons.
The neural regulation of the intestinal microcirculation is mediated by sympathetic vasoconstrictor neurons and the opposing vasodilator actions of the extrinsic sensory nerves and intrinsic submucosal vasodilator neurons (33). The aim of the present study was to further our understanding of the complexity of this neural regulation by examining the possibility that purines released from sympathetic varicosities constrict submucosal arterioles not only through postjunctional mechanisms but also by presynaptically inhibiting the release of neurotransmitters from extrinsic afferent nerve terminals and submucosal vasodilator nerves. We found that activating presynaptic adenosine receptors decreased release of vasodilator transmitters from extrinsic sensory nerves, whereas inhibition of these receptors with 8-PT enhanced vasodilation. We also found that activation of adenosine receptors on intrinsic submucosal vasodilator neurons had similar effects. Together, these findings demonstrate multiple neural vasoconstrictor mechanisms regulating the gastrointestinal microcirculation, which provides an additional level of complexity underlying neural regulation of mucosal blood flow.
Extrinsic sensory neurons have cell bodies in the dorsal root ganglia and innervate the submucosal arterioles via their projections in the perivascular plexus. The local activation of these extrinsic sensory nerves by electrical stimulation or agonists acting at transient receptor potential vanilloid 1 channels releases neurokinins and calcitonin gene-related peptide, which are potent vasodilators of guinea pig submucosal arterioles (13, 30). In the current study, electrical stimulation of these nerves evoked frequency-dependent vasodilations of submucosal arterioles that were inhibited >50% by exogenous ATP in the presence of suramin to block P2 purinoceptors. This effect was concluded to be presynaptic because ATP had no effect on the vasodilator action evoked by the exogenous application of substance P. We hypothesized that sympathetic varicosities were a source of the inhibitory purine, since ATP is the principal neurotransmitter released by sympathetic vasoconstrictor nerves in guinea pig ileum. Consistent with this, we found that vasodilations were markedly increased in preconstricted blood vessels in the presence of guanethidine, which blocks the sympathetic release of ATP. Given the virtual absence of nerve-evoked vasoconstrictions when arterioles were preconstricted, and the lack of overlap in the time course of nerve-evoked constrictions and dilations, the increase in dilation is consistent with the loss of presynaptic inhibition of the vasodilator response. We also found that the nerve-evoked dilations were enhanced by a nonselective adenosine receptor antagonist, again suggesting that nerve stimulation releases ATP from sympathetic nerves, which acts presynaptically on the extrinsic afferent nerves to inhibit vasodilation.
Exogenous ATP dose-dependently inhibited nerve-evoked vasodilations. This inhibitory action of ATP and that of the nerve-evoked inhibition (i.e., via sympathetic nerves) was unaffected by suramin but blocked by the P1 receptor antagonist, 8-PT, suggesting a role for presynaptic adenosine receptors in modulating neurotransmission from extrinsic sensory afferent nerves. Previous studies have shown that adenosine can also evoke presynaptic inhibition of extrinsic sensory nerves in other neural networks outside the gut (27). It is well known that ATP is rapidly hydrolyzed to form AMP and adenosine in many tissues and thus it is possible that neurally released ATP is converted to adenosine before having its presynaptic effect. However, previous studies examining the actions of ATP in the presence of the ectonucleotidase inhibitor, α,β-methylene-ADP, to prevent the degradation to adenosine, suggest that ATP itself is able to cause presynaptic inhibition via activation of adenosine receptors (25). Furthermore, the hydrolysis-resistant β,γ-methylene-ATP was active at the same concentration as ATP, and its inhibitory action blocked by 8-PT. Similarly, we found that α,β-methylene-ATP was able to inhibit nerve-evoked vasodilations, consistent with an ability of ATP to directly activate 8-PT-sensitive presynaptic adenosine receptors. In contrast to our current findings of a presynaptic inhibitory role for adenosine receptors, activation of adenosine A2 receptors has previously been reported to vasodilate rat gastric submucosal arterioles (20). Taken together, these data suggest that adenosine receptors can enhance or inhibit submucosal vasodilation, depending on whether the receptors are located pre- or postjunctionally.
Submucosal vasodilator neurons are the final common pathway of intrinsic vasodilator reflexes mediated by mucosal stimulation in both the submucosal and myenteric plexus (33). Previous studies in the guinea pig myenteric plexus demonstrated that contractile responses evoked by capsaicin or mesenteric nerve stimulation of extrinsic sensory neurons were also inhibited by presynaptic adenosine receptors (28), suggesting that these receptors may be found on other enteric neurons, including the submucosal vasodilator neurons (1, 9). We found that ATP also inhibited enteric submucosal vasodilator nerves. Although it was not directly tested in this study, the close apposition of the cholinergic enteric nerve terminals with the sympathetic nerve terminals on submucosal arterioles makes it highly feasible that purine release from sympathetic nerves could have a similar action on enteric neurons as observed with the extrinsic afferent neurons. Alternatively, or in addition, purines derived from nonneural sources could play important roles in modulating vasodilator transmitter release (3).
Neural regulation of the intestinal microcirculation can have a profound influence on mucosal blood flow, which in turn is an important determinant of epithelial barrier function. Several recent studies have highlighted defects in blood flow regulation during gastrointestinal diseases and suggest the importance of gaining a more thorough understanding of the reflex circuits underlying neural regulation of mucosal blood flow (12, 13, 19). We have employed precise in vitro techniques to directly demonstrate presynaptic inhibitory effects of the sympathetic vasoconstrictor pathway on the release of neurotransmitters from the two vasodilator neural pathways. Our studies suggest that the balance between vasoconstriction and dilation involves not only the actions of neurotransmitters on the postsynaptic membrane but also the release of purines, presumably from sympathetic nerves terminals, that result in a significant presynaptic inhibition of vasodilation. When combined with presynaptic actions of other sympathetic cotransmitters, norepinephrine (14) and neuropeptide Y (15), there is great scope for the reduction of submucosal arteriolar diameter by sympathetic innervation. Recent studies of inflammatory bowel disease (IBD) patients and mouse models of IBD have shown that purinergic neural signaling is dramatically altered during inflammation (2, 16, 17, 23). Moreover, there may be greater availability of adenosine from nonneural sources, including endothelial cells and platelets (4), during vascular inflammation (6), which could act at presynaptic adenosine receptors, thereby blunting the vasodilator reflexes, increasing mucosal injury, and inhibiting tissue healing.
This work was supported by an operating grant to S. J. Vanner from the Canadian Institutes of Health Research and a Research Scientist Award from the Crohn's and Colitis Foundation of Canada.
No conflicts of interest are declared by the authors.
We gratefully acknowledge the technical support of Lenisa Atwood, Dan Patton, and Margaret O'Reilly and the helpful comments of Ian Spreadbury on the manuscript.
- Copyright © 2010 the American Physiological Society