|
|
||||||||
-adrenoreceptor-induced C-fiber activation in healthy human skin
1Department of Neurology, Johannes Gutenberg Universität Mainz, D-55101 Mainz; and 2Department of Neurology and 3Institute for Physiology and Experimental Pathophysiology, Friedrich Alexander Universität Erlangen, 91054 Erlangen, Germany
Submitted 11 September 2003 ; accepted in final form 3 December 2003
| ABSTRACT |
|---|
|
|
|---|
-adrenoreceptor agonists alone do not activate nociceptors sufficiently to induce axon reflex flare. sympathetic nervous system; axon reflex; laser-Doppler imaging
Sympathetic activation, and even injection of NE, usually is pain free. Therefore, it was astonishing that axon reflex vasodilation was demonstrated during iontophoresis when NE was the ion used to deliver current instead of saline (11). According to these results, one might speculate that NE binds to and stimulates receptors on primary afferent neurons and contributes to C-fiber excitation and axon reflex vasodilation in human skin. However, iontophoresis of catecholamines has disadvantages because the current itself could activate C fibers (10). To overcome this shortcoming, we administered NE and other
-adrenergic agonists via microdialysis in the present study. Then, if axon reflexes were observed, it would support the hypothesis that catecholamines may contribute to the axon reflex response previously observed during iontophoresis. In combination with laser-Doppler imaging (LDI), microdialysis has been previously shown to be sensitive to assess local skin blood flow changes (5, 27).
| METHODS |
|---|
|
|
|---|
Microdialysis. Before insertion of the microdialysis fibers, nerve trunks of superficial branches of the radial nerve proximal to the wrist were blocked with 2% lidocaine. We watched clinically that this block persisted throughout the experiment. Thereby, efferent sympathetic nerve fibers or dorsal root reflexes were blocked, and the microdialysis fiber could be inserted pain free. Three (or four, see below) single plasmapheresis hollow fibers (0.4 mm in diameter, cut-off 3,000 kDa; Dermal Dialysis, Erlangen, Germany) were then inserted intracutaneously into the anesthetized skin on the back of the hands. Insertion depth was measured by ultrasound in previous studies and found to be 0.65 mm, on average (25). The fibers were inserted at a length of 1.5 cm via a 25-gauge cannula. A distance of at least 1.5 cm was kept between each fiber. The hollow fibers were perfused with physiological saline via a microdialysis pump (Pump 22, Harvard Apparatus, Holliston, MA). The flow rate was 4.0 µl/min. The dialysate was collected in calibrated glass capillaries to control perfusion and to minimize outflow resistance.
After a baseline period of 60 min of saline perfusion, perfusion medium was switched to NE, clonidine (CL), or phenylephrine (PE) in saline for a further 30 min. NE concentrations ranged from 10-10 to 10-3 M; CL and PE (all substances from Sigma, St. Louis, MO) were applied at concentrations of 10-8-10-3 M. All solutions were freshly prepared from stock solutions (10-3 M, stored at -50°C, renewed every week) before each experiment.
In all subjects, three fibers were inserted in parallel (n = 90 fibers in total) and perfused with different concentrations of either NE, CL, or PE. In nine of our subjects, an additional fourth fiber was inserted and perfused with saline for control. All numbers in RESULTS refer to the number of fibers.
LDI. Superficial blood flow was quantified by using a laser-Doppler scanner (LDI, Moor Instruments, Devon, UK). LDI image series were recorded at 15-min intervals during baseline perfusion (60 min) and at 4-min intervals during stimulation periods. The size of LDI scanned skin area was determined individually; one picture took
2.30 min (scan resolution 4 ms/pixel) to be completed. In any case, scanned skin area exceeded the borders of microdialysis or iontophoreses by at least 3-4 cm. Thereby, even very remote changes of microcirculation would have been detected. Based on visual screening of LDI pictures, we made the decision to quantify skin blood flow in a rectangular area (2 x 10 mm) directly above the microdialysis fiber and in a putative axon reflex flare area of 0.75 cm beneath (Fig. 1). The mean blood flow in perfusion units was calculated by using dedicated software offline (Moor LDI image processing, Moor Instruments, London, UK). For statistical analysis, flux values were normalized and expressed as percent change of baseline (last picture before stimulation).
|
Statistics. For statistical evaluation, a multivariate ANOVA for repeated measures (eight pictures: baseline and during drug application) was calculated. Main factors were the course of VC (subsequent pictures, within-subject factor), the different drugs (saline, NE, CL, PE), and the different concentrations of each drug (between-subjects factors). To allocate significant differences, the mean blood flow of seven pictures after stimulation was compared by post hoc tests with Bonferroni's correction. In this study, all values are presented as means ± SE. Statistical significance is considered at P < 0.05.
| RESULTS |
|---|
|
|
|---|
CL (total n = 20) also induced VC (concentration effect: F = 2.915, P < 0.001). Post hoc analysis revealed significant differences between saline and 10-5 M CL (n = 3; 48.7 ± 7.7%) and 10-3 M CL (n = 8; 67.7 ± 4.7%). The other values [10-8 (n = 2, 87.9 ± 9.4%), 10-7 (n = 3, 91.7 ± 7.7%), 10-6 (n = 2, 68.3 ± 9.4%), and 10-4 M (n = 2, 61.1 ± 9.5%)] were not significant.
In a similar way, PE (total n = 18) induced VC (concentration effect: F = 2.9, P < 0.001). Post hoc analysis revealed significant differences between saline and 10-3 M PE (n = 6; 61.0 ± 6.3%). PE concentrations of 10-8 (n = 2, 80.0 ± 11.0%), 10-7 (n = 3, 101.5 ± 9%), 10-6 (n = 2, 65.4 ± 11.0%), 10-5 (n = 2, 58.5 ± 8.9%), and 10-4 M (n = 3, 60.2 ± 11%) failed to reach significant difference in the post hoc analysis.
There was no significant difference in VC induced by NE, CL, or PE (F = 0.421, not significant; multivariate ANOVA adjusted for different concentrations). Results are summarized in Fig. 2, A-C, respectively.
|
Adrenoreceptor-mediated axon reflexes. At a concentration of 10-3 M, NE (n = 9), CL (n = 8), and PE (n = 6) induced significant VC, which was not different among the three drugs (Fig. 3A). This clearly supraphysiological concentration was chosen to assess possible axon reflexes quantitatively. In the presumed axon reflex area, there was no significant difference in skin blood flow among saline (89.8 ± 3.7%), NE (81.6 ± 3.5%), CL (90.4 ± 3.9%), or PE (90.3 ± 4.5%) (F = 1.408, not significant) (Fig. 3B).
|
Moreover, we watched every single picture during our experiments. There were, in total, 210 LDI scans with drug administration. In none of these LDI scans could any vasodilation, suspicious for axon reflex vasodilation, be detected, neither in concentrations below nor in concentrations exceeding VC threshold.
| DISCUSSION |
|---|
|
|
|---|
-adrenegic drugs did not activate nociceptive C fibers sufficiently to produce axon reflex flares in healthy human skin. Thus we could not provide evidence for interaction between sympathetic efferent fibers and afferent nociceptors in healthy human skin.
In the present study, we used different
-adrenegic agents: NE, CL, and PE. When applied via dermal microdialysis, these drugs dose-dependently induced VC, and, with concentrations of 10-3 M, maximum VC and skin blanching could be achieved. VC was not different, despite the different affinity to
1- and
2-receptors of each drug. That means the microdialysis system in our study obviously was able to deliver drugs quantitatively to blood vessels.
In neuropathic pains, NE may intensify pain and allodynia, in particular if these symptoms were alleviated before by sympathetic blocks (1, 26). In these cases, a pathological cross talk between sympathetic efferent and primary afferent fibers has been shown, either in dorsal root ganglia (18) or in the periphery (24). If sympathetic activation of afferent fibers is a major mechanism of neuropathic pain, it is labeled "sympathetically maintained pain" (3). Injection of NE into intact human skin, however, is not painful. Activation of the sympathetic nervous system usually helps to suppress pain perception. This difference between neuropathic and physiological reaction would be explained if
1- and
2-receptors on primary afferents were either newly formed, increased in number, or sensitized in response to axonal damage. To further study these different mechanisms of sympathetic-nociceptive interaction, efforts were undertaken to find human models of sympathetically maintained pain. The results were ambiguous. Physiological activation of the sympathetic nervous system had no impact on capsaicin-induced pain, neither in the skin (4) nor in muscles (13). On the other hand, thermal hyperalgesia, which frequently occurs after capsaicin injection, was enhanced if NE was coapplied by iontophoresis (8). Recently, even sensitization of heat-sensitive C fibers after injection of catecholamines has been demonstrated in normal skin without capsaicin pretreatment (14). Correspondingly, there was one study showing a current-dependent vasodilation after iontophoresis of NE in normal human skin (11). Inside a ring-shaped iontophoresis electrode, a current-dependent vasodilatation was observed, which was significantly greater than with saline. It must be attributed to an axon reflex mechanism, because it was restricted to the iontophoresis site, and vasodilation could be prevented by pretreatment with local anesthetics. If indeed there were constitutively expressed adrenoreceptors on primary afferent neurons, a rapid adrenergic responsiveness of C fibers after nerve lesion (2) and "sympathetically maintained" neuropathic pain would be better understood. However, the findings in the present study were converse. Even at supramaximum concentrations for local VC, neither NE, CL, nor PE induced axon reflex vasodilation, neither in the vicinity of the microdialysis fibers nor somewhere on the LDI pictures. The most important difference between the previously reported findings and our results is the route of application of adrenergic agents. For iontophoreses, it has been shown that the current itself induces C-fiber activation and axon reflexes (15). Saline iontophoresis, for example, could increase permeability of superficial skin layers (10), and vascular and even sensory nerve function might become compromised. Furthermore, the iontophoretic current could disrupt the perineurium of nociceptors in the skin (12). Activation of C fibers is even more obvious when NE is applied by injection (14). These drawbacks of iontophoretic or injection techniques do not apply for dermal microdialysis. Although insertion of the fibers induces skin trauma as well, these inflammatory skin changes usually subsided after 1 h (22). Thereafter, all substances could be delivered atraumatically. As demonstrated by dose-dependent VC in this study and by previous results (6), there is a linear correlation between concentration of substances in the perfusion medium and delivery through the microdialysis fibers into the tissue. It could be argued that afferent fibers are located more superficially in the epidermis and, therefore, might be reached easier by iontophoresis compared with microdialysis. However, it is obvious that the nociceptors, which are responsible for the axon reflex, must have a close relation to the subepidermal vessels, because epidermal diffusion of CGRP is very limited (27). This microdialysis mode of dermal application was used to provoke axon reflex vasodilation by various inflammatory mediators, such as bradykinin, histamine, and prostaglandin E2 (20, 22). As the physiological source of catecholamines in the skin is the endings of vasoconstrictor neurons close to skin vessels, application via microdialysis might resemble the physiological conditions even better than the epidermal application via iontophoresis. The fact that skin blood flow was even slightly decreased in the presumed axon reflex area indicates that we indeed quantified skin blood flow at the border of VC. However, our results do not entirely rule out the possibility that the adrenergic system may still be involved in excitation of afferent C fibers. The adrenergic system may lower thresholds for axon reflexes to another stimuli (e.g., iontophoretic current). This could explain algesic and sensitizing effects when NE was iontophoresed (11), injected (14), or coapplied with capsaicin (8, 17). Furthermore, adrenoreceptor-induced VC could augment axon reflex vasodilation, in particular if VC surrounds the flare area. In both studies claiming NE-induced axon reflex flares, a ring-shaped iontophoresis electrode was used (11, 28). In this cases, vasodilatory neuropeptides from current-excited (iontophoreses) primary afferents may accumulate because they could not be sufficiently cleared from the skin (9).
In conclusion, our results demonstrate that there is no good evidence for direct interaction of sympathetic and nociceptive skin fibers under physiological conditions in human skin. Therefore, human models of sympathetically maintained pain have to be interpreted with care. The mechanisms of how sympathoafferent coupling leads to sympathetically maintained pain has to be explored in studies basically involving patients rather than normal subjects.
| GRANTS |
|---|
|
|
|---|
| 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 |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. Dogrul, I. Coskun, and T. Uzbay The contribution of alpha-1 and alpha-2 adrenoceptors in peripheral imidazoline and adrenoceptor agonist-induced nociception. Anesth. Analg., August 1, 2006; 103(2): 471 - 7, table of contents. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. L. Houghton, J. R. Meendering, B. J. Wong, and C. T. Minson Nitric oxide and noradrenaline contribute to the temperature threshold of the axon reflex response to gradual local heating in human skin J. Physiol., May 1, 2006; 572(3): 811 - 820. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. Hermann, T. Schlereth, T. Vogt, and F. Birklein Clonidine induces nitric oxide- and prostaglandin-mediated vasodilation in healthy human skin J Appl Physiol, December 1, 2005; 99(6): 2266 - 2270. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |