| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
NEUROREGULATION AND MOTILITY
1Department of Biomedical Science, The University of Sheffield, Western Bank, Sheffield, United Kingdom; and 2Department of Human Physiology and Centre for Neuroscience, Flinders University, Bedford Park, Adelaide, South Australia, Australia
Submitted 14 February 2007 ; accepted in final form 20 June 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
-conotoxin MVIIC (1 µM intra-arterially) or the P2X receptor/ion channel blocker PPADS (50 µM). However, ruthenium red (50 µM), a blocker of nonselective cation channels, greatly reduced the flow-off response and also abolished the vasodilator response to capsaicin. Our data support the concept that mesenteric afferents sense changes in vascular flow during flow off through direct mechanisms, possibly involving nonselective cation channels. Passive distortion in the fan, caused by changes in blood flow, may represent a natural stimulus for these afferents in vivo.
mesenteric blood vessels; blood flow; mechanotransduction
In our previous study (6), we suggested that mesenteric afferents may also sense changes in gastrointestinal blood flow. This sensitivity persisted after removal of the gut itself and, therefore, was a property of the mesenteric membrane and its vascular supply, prior to where they penetrate the gut wall. We speculate that this sensitivity is part of a novel mechanism for detecting and responding to hemodynamic disturbances. However, the mechanism by which mesenteric sensory nerves detect changes in flow was not identified. Therefore, in the present study, we used a novel in vitro mesenteric vascular preparation in combination with anterograde nerve labeling experiments to investigate 1) the location and 2) the mechanism of afferent sensitivity to vascular perfusion. We offer evidence that flow sensitivity arises as a consequence of direct distortion of the afferent endings on blood vessels rather than being secondary to mediator release from the vascular endothelium or mechanical responses of vascular smooth muscle.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Male Wistar rats (350–400 g) were anesthetized with pentobarbitone sodium (60 mg/kg), and a midline laparotomy was performed. The terminal ileum (
3 cm rostral to the ileocaecal junction) and the arterial supply were identified, and the vessels were exposed by carefully removing connective tissue from around the superior mesenteric artery and vein. A branching section of an artery with a clear projection to a segment of the ileum was chosen, and unwanted side branches were ligated, as previously described (6).
The ileal segment, complete with the mesenteric arcade and an accompanying piece of the mesenteric artery, was then removed and quickly immersed in ice-cold saline. After the gut lumen was flushed, the tissue was placed in a purpose-built chamber consisting of a recording chamber and a common bathing and perfusion chamber (Fig. 1). The gut "tube" was removed by cutting along the mesenteric border, and the mesenteric fan was pinned flat in the bathing chamber and perfused with Krebs-bicarbonate-saline solution [composed of (in mM) 143.5 Na+, 5.9 K+, 126 Cl–, 2.5 Ca2+, 1.2 Mg2+, 1.2 H2PO4, 1.2 SO42–, and 25 HCO3–; pH was maintained at 7.4 with 95% O2-5% CO2; flow rate was 5 ml/min]. The mesenteric artery was cannulated to permit intravascular perfusion, and the system was connected to a pressure transducer to allow monitoring of vascular perfusion pressure (VPP). VPP was used as an index of vascular tone/blood flow. The artery was perfused with modified Krebs-bicarbonate-saline solution [+3% dextran (40,000 mol wt), 10 mM glucose, and 0.6 mM glutamine: 305 mosM/kg water], vigorously gassed with 95% O2-5% CO2, and perfused via a bubble trap using a peristaltic pump, which resulted in flow that was not pulsatile, at 1.4 ml/min. The mesenteric arcade was also pinned in the bathing chamber. Under a stereomicroscope, a paravascular nerve bundle was teased out from between the small artery and vein supplying the segment, pulled through an aperture into a separate recording compartment, and wrapped around one arm of a bipolar platinum recording electrode. A similar length of connective tissue was also drawn into the chamber and attached to a second electrode. The aperture linking the two compartments was then sealed with grease and a glass coverslip before the recording compartment was filled with heavy liquid paraffin. The preparation was allowed to warm slowly and reached a working temperature of 32–34°C before a paravascular nerve recording was obtained.
|
Electrodes were connected to a Neurolog headstage (NL 100), and the signal was amplified (NL 104, x20,000), filtered (NL 125, bandwidth: 100–1,000 Hz), and then relayed to a spike processor (Digitimer D130) to allow discrimination of action potentials from noise using a manually set amplitude and polarity window. The whole nerve recording was displayed on a storage oscilloscope (Tektronix 5111A) and digitized (PCM-2 A/D VCR adapter, Medical Systems) to allow recording on a VHS videotape for future off-line analysis. Whole nerve activity was continually monitored as spike discharge (in impulses/s) and stored on a PC using Spike 2 software (CED).
Anterograde Labeling in Whole Mounts
After functional recordings had been completed, some preparations were used to visualize the distribution of axons originating from the paravascular nerve bundle under study. The nerve bundle was left attached to the recording electrode and anterogradely filled with biotinamide (5%) applied at a recording site (sites A or B; Fig. 2) as previously described (30). The nerve bundle was washed using drops of artificial "intracellular medium" [composed of (in mM) 150 monopotassium L-glutamic acid, 7 MgCl2, 5 glucose, 1 EGTA, 20 HEPES, 5 disodium ATP, 0.02% saponin, and 1% DMSO] (26) and placed under the paraffin oil. A drop of biotinamide (5% in intracellular medium) was then applied to the nerve bundle under paraffin oil, and the preparation was covered with supplemented culture medium (DMEM-F-12, 10% FBS, 1.8 mM CaCl2, 100 IU/ml penicillin, 100 µg/ml streptomycin, 2.5 µg/ml amphotericin B, 20 µg/ml gentamycin, Cytosystems; pH 7.4) in the perfusion chamber. Preparations were covered, humidified, and incubated with 5% CO2-95% O2. After 12–15 h, preparations were fixed with modified Zamboni's fixative (15% saturated picric acid and 2% formaldehyde in 0.1 M phosphate buffer; pH 7), cleared with DMSO (30 min), and rinsed in PBS. Labeled axons were visualized with Texas red-conjugated streptavidin (1:50, 4 h). Some preparations were labeled overnight at room temperature with antiserum to 
-CGRP (rabbit anti-human, 1:1,600, Peninsula Labs). After being rinsed in PBS, these preparations were incubated with goat anti-rabbit FITC for 2–4 h (1:1,200, Jackson Immunochemicals). After being rinsed again in PBS, preparations were mounted in Vectamount (Vector Labs) and viewed using a Zeiss Axioscope 2 fluorescent microscope.
|
Location of afferent sensitivity to stop of vascular perfusion (stopped flow). To assess the site of sensitivity, basal afferent firing and afferent responses when vascular perfusion was stopped for 5 min ("flow off") were examined. The first response was recorded in preparations in which the blood vessels had been cut close to the point where they entered the gut body. The second response was recorded after removal of the terminal vessels (cut 1), as shown in Fig. 2D. After each experiment, afferent responses and effects on vascular perfusion pressure induced by capsaicin (100 nM intra-arterially) were assessed in preconstricted preparations (L-phenylephrine, 10 µM intra-arterially).
Mechanism of afferent activation by stopped flow. Preparations were largely immobilized by being pinning flat, which prevented excessive distortions as the entire fan collapses once arterial perfusion is stopped. Although not preventing all movement in the fan, the pinning allowed us to examine the effects of more subtle perivascular changes such as vasoconstriction and localized distortion, which may act as stimuli for perivascular afferents. We investigated roles of the following: 1) extracellular Ca2+, 2) changes in vascular smooth muscle contractility, 3) neuronal stimulation by ATP, 4) neuronal stimulation by ruthenium red-sensitive channels, and 5) involvement of N-type Ca2+ channels.
After a 30-min equilibration period, two successive responses to 5-min flow off were recorded 20 min apart. Only preparations with two reproducible responses were used. After the second response, perfusion solutions were modified or pharmacological agents were added (see below). After a further 20 min, a third response to flow off was established. In experiments investigating extracellular Ca2+, all perfusion solutions were replaced with Ca2+-free equivalent solutions (CaCl2 replaced with MgCl2) after the second response to flow off. A similar protocol was used to test the effects of water-soluble antagonists (PPADS, ruthenium red, and -conotoxin). Effects of wortmannin were examined by comparing two sets of experiments: one group pretreated with wortmannin (1 µM intra-arterially) and another set of vehicle controls (0.1% DMSO). In these experiments, the inhibitory action of this light chain kinase was confirmed by comparing vasoconstriction (maximum increase in VPP) caused by L-phenylephrine (10 µM intra-arterially) after each experiment.
Chemicals
The following compounds were purchased from Sigma: dextran (mol wt 40,000), L-phenylephrine HCl, capsaicin, and ruthenium red. Capsaicin was dissolved in DMSO and diluted with saline; 0.01% DMSO was the final concentration for intra-arterial administration. PPADS and wortmannin were purchased from Tocris Bioscience (Bristol, UK).
Data Analysis
The afferent response to flow off was calculated as the total number of impulses (area under response profile) using Spike 2 software. Baseline afferent discharge was calculated as the average firing rate over 1 min (unless stated otherwise) during steady-state activity after an equilibration for 30 min and expressed as impulses per second. Afferent responses to capsaicin were determined as the total number of impulses in the first minute of the response [area under the curve (AUC)]. Data are expressed as means ± SE. Two means were compared using a paired or unpaired t-test on the raw data. Values of P < 0.05 were considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Mesenteric afferents respond to the stop of vascular perfusion even when all connections to the body of the gastrointestinal tract have been severed (6). We explored the location of this sensitivity within the mesenteric fan by 1) visualizing the morphological arrangement of nerves within it and 2) examining the functional effects of removing the terminal blood vessels (Fig. 2D).
Morphological experiments. Biotinamide applied to one of the paravascular nerve bundles was used to visualize the pattern of mesenteric innervation. The nerve bundle loosely followed the dividing vascular tree (Fig. 2A). No intimate associations between nerves and blood vessels were observed when paravascular nerve bundles were filled from recording site A. Since the intensity of fluorescence reduces as the distance from the recording site increases, the arrangement of nerves further down the vascular tree was examined by applying biotinamide closer to the mesenteric border with the gut body (site B). Under these conditions, a perivascular nerve plexus was observed, largely confined to the terminal arteries, close to the point where they penetrated the serosa (Fig. 2B). Some of these neurobiotin-filled axons contained CGRP immunofluorescence (Fig. 2C) suggesting that some of the perivascular nerves visualized were afferent in nature. No other specialized axonal endings were seen in either the mesenteries or serosa; the only other filled structures were bundles of smooth axons of passage.
Functional experiments. In the absence of the gut body, the basal afferent discharge was 6.4 ± 2.3 impulses/s, and there was a robust response to stop of vascular perfusion (513 ± 98 impulses). Cutting off all terminal vessels had a nonsignificant effect on perfusion pressure (from 64.3 ± 5.9 mmHg before to 56.6 ± 2.7 mmHg after) but significantly reduced both baseline discharge and the response to flow off (Fig. 3). However, after these terminal arteries were removed, preparations showed maintained vasoconstrictor responses to L-phenylephrine and vasodilator responses to capsaicin (data not shown).
|
To determine the mechanisms underlying the increase in afferent firing in response to flow off, the response was examined following pharmacological treatments designed to modulate mediator release, including that from the vascular endothelium and vascular contractility.
Effects of Ca2+-free media. Under Ca2+-free conditions, there was an augmentation of baseline afferent firing (Fig. 4B) and a dramatic increase in the response to stopped flow (Fig. 4C). A representative example is shown in Fig. 4A. The effect of removal of Ca2+ was reversible (data not shown). The vasoconstrictor effect of L-phenylephrine persisted under Ca2+-free conditions (data not shown). The increased excitability of mesenteric afferents following removal of Ca2+ and the maintained vasoconstrictor effect of L-phenylephrine made it difficult to distinguish contributions from mediator release and vascular contractility. Further experiments sought to distinguish between the two.
|
|
Effects of -conotoxin MVIIC.
The potential for mediators to increase afferent firing via an action dependent on N-, P-, and Q-type Ca2+ channels was investigated using the Ca2+ channel blocker -conotoxin MVIIC (100 nM). -Conotoxin had no effect on baseline discharge (data not shown) or responses to stopped flow (flow off) (Fig. 5B).
Effects of ruthenium red. Ruthenium red is an inorganic dye that blocks nonselective cation channels (NSCCs), including capsaicin-activated voltage-regulated (VR)1 receptors [transient receptor potential (TRP)V1 channels]. Ruthenium red (50 µM) reduced baseline firing (from 10.4 ± 3.5 to 5.5 ± 2.4 impulses/s, n = 4), although not statistically significantly, and dramatically attenuated the response to flow off (Fig. 5A). Ruthenium red also significantly reduced the afferent response to 100 nM capsaicin (from 425 ± 46 to 33.5 ± 14 impulses, AUC first minute, P < 0.05 by unpaired t-test) and blocked the capsaicin-induced vasodilation (data not shown).
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Visualizing Paravascular Nerves in the Mesentery
Paravascular nerves (including CGRP-containing extrinsic afferents) were shown to follow the mesenteric vascular tree and form a perivascular plexus around terminal arteries prior to penetrating the gut wall. Previous studies (12, 28) have demonstrated that CGRP-containing fibers extend to the whole mesenteric arterial tree. However, immunohistochemical studies have not identified that axons take to their terminals. From our study, it is clear that afferent fibers travel some distance in the paravascular bundles before coming into close proximity with the vessel wall. This is important because it shows where the axons, which contributed to our recordings, make their specialized endings. Fibers in the main mesenteric paravascular bundles, in which robust flow-off responses were recorded, rarely make close contact with the underlying artery but appear to preferentially supply the smaller branches just prior to their entry into the gut wall. These terminal arteries thus are important structures for sensory innervation and may play an important role in the detection of, or its responses to, disturbed blood flow. This is fully compatible with the location of mechanosensitive sites identified by strong focal compression, both in vivo and in vitro, which are typically densest on the terminal vascular branches and nearby sites on the serosa (3, 5, 10, 21–22).
Site of Afferent Sensitivity Within the Mesentery
Sensory nerves, originating from cell bodies in dorsal root ganglia, are known to project along the mesenteric vasculature and send collateral projections to small arteries and arterioles (2, 14). Vagal afferents also supply the small bowel, but these terminate in endings within the muscularis and mucosa (2) and so are unlikely to contribute in this mesenteric preparation. Little is known about functional afferent sensitivity within the mesenteric fan itself, apart from the fact that some afferents can be activated by nonphysiological levels of focal compression with blunt probes or von Frey hairs (3, 5, 10, 21–22). We have proposed that some of these mesenteric afferents are sensitive to vascular perfusion of the mesentery, responding, in particular, to stop of arterial perfusion. We (6) have previously shown that this is not simply a response to hypoxia but is caused by the change in perfusion. Afferents are excited by flow off, and activation still occurs when the gut body is removed. However, afferent sensitivity in the mesentery itself appears to reside mainly in the finer branches of the vascular tree originating from the second arterial bifurcation, although some sensitivity remains more proximal to this. Hence, mesenteric vessels are not merely a "passive" neurovascular highway but rather appear to be an important site of sensory transduction.
Afferent sensitivity to stopped flow was associated with the same areas of the mesentery where blood vessels had an intimate association with neurobiotin-labeled fibers. Previous studies (3, 5, 10, 21–22) that have used probing to identify regions of mechanosensitivity have identified "hot spots" at the branch points of arteries close to the serosal surface. It is likely therefore that the endings responding to stopped flow also respond to other stimuli that cause distortion of the mesenteric fan, such as blunt probing.
Mechanism of the Afferent Response to Stopped Flow
Gastrointestinal afferents can also be activated by ischemia in vivo (11, 15, 18). However, these responses are probably more complex than the afferent responses observed when vascular perfusion is stopped in vitro. In particular, hypoxia cannot account for afferent activation in our model, since previous work has shown that removal of oxygen from the perfusing solution does not activate afferents (6). Moreover, since the responses are preserved after removal of the gut body, it is likely that the "machinery" is present within the mesenteric fan itself.
Indirect Afferent Activation Following the Release of Endothelial Mediators?
Preliminary experiments aimed at selectively damaging the endothelium proved inconclusive, since treatments also altered afferent viability (data not included). In addition, attempts to disable transduction sites in the glycocalyx with neuraminidase were also difficult to interpret (6). To investigate a role for endothelial signals, Ca2+-free solutions were used to prevent mediator release through influx of extracellular Ca2+ triggering fast exocytosis (31). The response to flow off persisted in Ca2+-free medium. Indeed, the response was considerably exaggerated after removal of Ca2+, as described in other populations of sensory neurons (29), probably reflecting a generalized increase in the excitability of afferent axons. Because extracellular Ca2+ is not necessary for the flow-off response, it is unlikely that endothelial factors such as EDRF and EDHF are necessary for this response, although mediator release caused by Ca2+ release from intracellular stores cannot be excluded (27).
Ca2+ is also necessary for the contractile machinery in vascular smooth muscle that may be necessary for the action of EDRF and EDHF. However, it is unlikely that changes in tone are necessary to mediate the effect of flow off, as indicated in Fig. 6B, since afferent responses persisted when vascular smooth muscle was paralyzed by wortmannin to block myosine light chain kinase (25), a treatment that abolished the vasconstrictor response to L-phenylephrine.
|
The abrupt and prolonged nature of the afferent responses to flow off suggests a rapid and constant stimulus such as a mechanical stress. Direct activation of neuronal ion channels could play a role, and while neuronal Ca2+ channels are unlikely to be involved, mechanosensitive channels could be suitable candidates. Since the NSCC blocker ruthenium red effectively and consistently reduced afferent responses to flow off, one or more types of NSCCs may play a role. While the pharmacological characterization of afferent responses to flow off is by no means complete, two fundamental observations support the involvement of NSCCs. Afferent responses are 1) augmented by the removal of extracellular Ca2+ (with raised [Mg2+]) and 2) are inhibited by ruthenium red.
Removal of extracellular Ca2+ adds further support to the involvement of a NSCC, since Ca2+ can both carry current through these channels and also modulating the passage of monovalent cations (29).
Ruthenium red is an inorganic dye that blocks NSCCs, including capsaicin-activated TRPV1 channels and other TRP channels (1, 17, 23). Some of these channels are mechanically gated [such as TRPA1 (23)], raising the possibility that there is a direct link between the mechanical forces generated by flow off and activation of afferent firing. However, ruthenium red is not highly specific and may act on other channels not directly linked to mechanosensitivity. For example, the TRPV1 channel is sensitive to treatment with ruthenium red but is not considered to be mechanically gated, even though the intestinal afferent response to distension is attenuated in the TRPV1 knockout mouse (24). Clearly, the clarification of channels mediating direct mechanosensitivity of mesenteric afferents awaits better pharmacological tools.
Physiological/Pathological Relevance?
Sensitivity to capsaicin is often taken as synonymous with nociception, and it is conceivable that signals generated from the vasculature as a consequence of stopped flow may be relevant for pain signaling. Certainly, high-threshold nociceptive afferents can be sensitized by mediators released by ischemia and injury (4). The extent to which ischemia may modulate responses to stopped flow has not been investigated, and therefore the impact of the current findings for nociception is unknown. However, it is clear that the sensitivity to vascular perfusion may impact on the mechanisms that control mesenteric blood flow. Our study clearly points to an afferent role for nerve fibers in the perivascular plexus. Previous emphasis has been on the release of vasoactive mediators from these endings and their contribution to axon reflexes. The sensitivity of these endings to vascular perfusion might suggest that this system offers a rapid "protective" mechanism against gut ischemia. Afferent terminals alone could provide the machinery necessary to "detect and protect," where any potentially damaging fall in blood flow would signal and induce rapid changes in afferent discharge and the consequent release of vasoactive mediators such as CGRP. Vessel diameter, and hence blood flow, could thus be controlled locally, without reflex mechanisms.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
