|Home | About | Journals | Submit | Contact Us | Français|
Subcutaneous injection of endothelin-1 (ET-1) into the glabrous skin of the rat’s hind paw is known to produce impulses in nociceptors and acute nocifensive behavioral responses, such as hind paw flinching, and to sensitize the skin to mechanical and thermal stimulation. Here we show that, in contrast to the responses in glabrous skin, ET-1 injected subcutaneously into rat hairy skin causes transient antinociception. Concentrations of 1-50 uM ET-1 (in 0.05 mL) depress the local nocifensive response to noxious tactile probing at the injection site with von Frey filaments, for 30 - 180 mins.; distant injections have no effect at this site, showing that the response is local. Selective inhibition of ETA, but not of ETB receptors inhibits this antinociception, as does co-injection with nimodipine (40μM), a blocker of L-type Ca2+ channels. Local subcutaneous injection of epinephrine (45uM) also causes antinociception, through alpha-1 adrenoreceptors, but such receptors are not involved in the ET-1-induced effect. Both epinephrine and ET-1, at antinociceptive concentrations, reduce blood flow in the skin; the effect from ET-1 is largely prevented by subcutaneous nimodipine. These data suggest that ET-1-induced antinociception in the hairy skin of the rat involves cutaneous vasoconstriction, presumably through neural ischemia, resulting in conduction block.
The pain-inducing effects of endothelin-1 have been well-documented in glabrous skin of the rat, a frequently used test site. The opposite behavioral effect, antinociception, occurs from endothelin-1 in hairy skin, and is correlated with a reduction in blood flow. Vasoactive effects are important in assessing mechanisms of peripherally acting agents.
Endothelin-1 (ET-1) is an endogenous peptide of 21 amino acids 31, 67, whose effects are mediated by two G-protein-coupled receptors, ETA and ETB49,40,41. ET-1, and ETA and ETB receptors are localized in the epidermis and subcutaneous vasculature1, as well as in dorsal root ganglia and peripheral nerve, where, respectively, ETA receptors have been identified on neuronal cell bodies and axons and ETB receptors on satellite cells and Schwann cells55.
The most prominent physiological action of ET-1 is transient vasoconstriction, triggered by activation of ETA receptors on vascular smooth muscle and a corresponding increase in intracellular Ca2+, 57. This vasoconstriction is reversed by an activation of ETB receptors on the vascular endothelium that leads to the release of nitric oxide 65. More recent work suggests that ET-1 is an important mediator of hyper-nociceptive responses to injury and disease41. The use of selective antagonists of ETA and ETB receptors has revealed that ET-1 contributes importantly to elevated pain from nerve injury46, from inflammation 14, 19, 30, from cancer 22 and from incision of the skin 50.
Because ET-1 was known to be involved in a variety of inflammatory events, the pain associated with its elevation at sites of injury was initially believed to result from its vascular actions. More recently it has been shown that ET-1 applied directly to nerve12, 18 or injected into the plantar hindpaw23, at >100μM, leads rapidly to pain in the affected limb. Lower ET-1 concentrations, 2-20μM, injected into the paw result in sensitization of the skin to mechanical and thermal stimulation2,19. These local, acute actions of ET-1 occur within a few minutes, can persist for several hours, and often are followed by desensitization which lasts for 24h or longer18.
High concentrations of ET-1 injected into glabrous paw skin activate not only ETA receptors, causing pain behavior, but also ETB receptors, resulting in anti-hyperalgesia42. Such anti-hyperalgesia is mediated in part through the release of the opioidergic peptide β-endorphin from keratinocyes and its subsequent binding to μ-opiate receptors (MORs) on nociceptor endings in the epidermis42.
Despite the well-documented vasoconstrictive effects of ET-1, a role for such constriction, and the likely occurrence of local ischemia, has not been established in the pain-related effects of peripherally applied ET-1. It now appears that such events may occur, and even dominate, the overall sensory response to ET-1, although they are not dominant in the glabrous paw skin. In the present study, we report the results of experiments to determine the role of ET-1-induced vasoconstriction and related behavior in the hairy skin of the rat.
All procedures were approved by the Standing Committee on Animals of Harvard Medical School and the Animal Care and Use Committee at Indiana University School of Medicine, Indianapolis, IN. Male, Sprague-Dawley rats (Charles River Laboratories, Cambridge, MA) weighing 225-320g were used for all experiments and were housed in pairs on a 12-hour light:dark cycle with food and water ad libitum. For studies measuring nociceptive behaviors, rats were handled for 5-7 days until they were accustomed to sitting still on the bench top for up to 5 minutes at a time, and their local skin responses to innocuous touching had become extinguished (see next passage)63. Rats had the hair clipped on the dorsum of the trunk prior to the day of the experiment.
Responses were evaluated by cutaneous trunci muscle reflexes (CTMR) of subcutaneous muscles responding to stimulation with von Frey filaments, near the dorsal medio-thoracic site of drug injection, and on the contralateral dorsum16,50. The CTMR is a response to noxious tactile stimulation involving the local contraction of skeletal muscle beneath the skin with concurrent movement of the nearby skin over the rat dorsum64. The sensory input for this reflex, after the rats have been handled for several days to accommodate the reactions to light touch, is conducted by fine myelinated and non-myelinated nociceptor afferents64. An absence of the response by locally applied agents is considered an indication of local anti-nociception, unless the same agent has systemic actions, in which case analgesia, mediated by the CNS, cannot be ruled out. The rats were probed with von Frey hairs (VFHs) at points located 0.5, 1 and 2 cm ipsilateral and 2 cm contralateral from the center of the injection site, along a line that is perpendicular to the spinal midline. A “graded response” was assessed by probing the back skin with VFHs of varying stiffness/force, each one applied four times at any given distance with an interstimulus interval of ~2s16. Each filament was applied perpendicularly to the skin until the tip bent, at which point the force exerted is maximum. Each response is graded 1, 0.5 or 0, with 1 defined by a brisk, large amplitude and short latency CTMR response, 0.5 by a slower, smaller response and a grade of 0 indicating no detectable response. The response for 4 applications of any one force/VFH, at a given distance, was then summed and divided by 4 to get the normalized “graded response” (range = 0 - 1).
Measurements using VFHs were made every 15-minute for the first hour after injection and every 30-minute thereafter until the rat’s responses recovered to baseline. Baseline responses were taken for 3 days before any s.c. injections to get an averaged baseline response with which to compare the results of injection. Clipping the back on the day before the experiment, to facilitate visualization of CTMR, did not change this baseline responsiveness.
All solutions were adjusted to pH 7.4 and the experimenter was blinded to the subcutaneous treatments received by rats. ET-1 (Axxora, San Diego, CA) was freshly prepared by dissolving it in phosphate buffered saline (PBS). ETA and ETB receptor antagonists, BQ-123 and BQ-788, respectively, (both purchased from American Peptide, Sunnyvale, CA), were dissolved, respectively, in PBS only and in PBS with 40% DMSO (v:v) at stock concentrations of 6mM. These two antagonists have The final concentration of DMSO, used in the VEHICLE control solutions, had no effect on the behavioral response.
Rats were anesthesized briefly (~1min) with sevoflurane (Abbott Laboratories, North Chicago, IL) delivered by nose inhalation and the pre-shaved back skin was swiped with alcohol. Subcutaneous injections were made on the medial region of the dorsum of the rat’s back, 1 cm from the midline. Circles of radii 1 and 2 cm were drawn around the injection site, for locating test sites (see above) and animals were returned to their cage for recovery.
ET-1 was administered at concentrations of 1, 5, 10, 50 and 100uM (n=4), in s.c. injections of 50 μL. Two forces were applied, a weak force giving ~0.3 baseline score on the graded response scale (1.0 is the maximum value,) and a strong force giving ~0.7 on this scale; sensitivity was followed for up to 5 h, or until the behavioral changes had returned to baseline. To further quantify the effects of drugs, the area-under-the-curve (AUC) was calculated. Area-under-the-curve was determined as the area between the graded response after ET-1, and the baseline graded response value extended over the full time that behavior was tested, usually 3-4 h. Since the graded response is a dimensionless parameter, the AUC has the value of time (hours).
For the experiments testing ET receptor antagonists, rats were co-injected with ET-1 (10uM, 50uL) with or without BQ123 (0.1-3 mM), or BQ788 (3 mM). Control experiments followed the graded response after injection of 50 uL of PBS, with or without DMSO (control for BQ788). These antagonists have very similar affinities for their respective receptors 35, 36 so that comparable blockade of cutaneous receptors probably occurs from the equal, highest dosing used here.
Naloxone was purchased from Sigma Aldrich, as the hydrochloride salt, and was dissolved in PBS to give a stock concentration of 27.5mM, and then equal volumes of 5.5 mM naloxone and 20 μM ET-1 were mixed to give a final injectate concentration of 2.75mM and 10 μM, respectively.
Prazosin (Sigma Aldrich) was dissolved in slightly acidified buffer (pH = 6.8) and heated to 60°C to accelerate dissolution, to a stock concentration of 120 μM then cooled to room temperature for final dilution and injection. Yohimbine (Sigma Aldrich) was dissolved directly in PBS to a stock concentration of 130 μM and was co injected with ET-1 at 20 μM.
Local analgesia from ET-1, applied epineurially in vivo, diminishes endoneurial blood flow on rat sciatic nerve68, an effect that is inhibited by systemically administered (i.p.) nimodipine, a blocker of L-type Ca2+ channels. A stock solution of nimodipine (Sigma Aldrich), for co-injection with ET-1, was made to 24mM by dissolving in 100% ethanol (Fisher Scientific), and then sequentially diluted in PBS to give 40uM, containing <0.2% ethanol.
For experiments on cutaneous blood flow measurement by laser Doppler flowmetry, male Sprague Dawley rats weighing 225 to 250 gms were utilized. They were housed in a light controlled room (light from 6:00 to 19:00) at a constant temperature of 22°C. Food and water were available ad libitum. Endothelin-1 (ET1) was purchased from Alexis Biochemicals, and nimodipine was purchased from Sigma Chemical Company (St. Louis, MO). The ET-1 was reconstituted in PBS to a concentration of 60 μM and was stored at -20°C and was diluted to 30 μM before the experiment. The nimodipine was initially dissolved in 1-methyl, 2-pyrrolidinone (MPL; Aldrich Chemical Co., Milwaukee, WI) to a concentration of 400 mM, and then diluted to 40 μM in PBS. Blood flow was measured with a BLF21D laser Doppler flowmeter (Transonic systems Inc.). The rats were anaesthetized with sodium thiopental (100mg/kg) throughout the experiment. The hair from the paw was shaved with a trimmer and the animals were placed on a heated platform to maintain body temperature at 37° C. The laser flowprobe was placed on the hairy skin of the back of the paw of the rat. After an initial baseline recording for 15 minutes to establish baseline, 10 μL of 40 μM nimodipine or vehicle (0.01% MPL in PBS) was injected intradermally, followed in 5 min by a second 10 μL injection of 30μM ET-1. The blood flow response was recorded for a subsequent 30 min.
Results were expressed as graded response ± standard error of the mean (S.E.M.) We used t-test (independent) for comparing groups of two, and 1-way ANOVA for multiple group comparisons. All statistical analyses were performed using Origin 7 (OriginLab Corp., Northampton, MA). A P value <.05 was considered to be significant.
All data are presented as mean response (measured over 3 min) ± S.E.M. Blood flow was recorded as tissue perfusion units (TPU). To compare the effects of the different drugs on blood flow, the average response over 15 minutes to injection of each drug/combination of drugs was subtracted from the average baseline with needle over 15 minutes to give the average evoked response over 15 minutes. One way ANOVA and Tukey’s multiple comparison post hoc test was used to compare the average evoked responses between different treatment groups.
Subcutaneous injection of ET-1 resulted in tactile antinociception, measured as reduced or absent responses to von Frey hair stimulation. Antinociception was detected at the first measured time, 15 minutes after injection of 5, 10 and 50uM (Fig. 1A, B). Partial antinociception, which lasted for less than 1h, followed injection of 1uM or 5uM ET-1 and was significantly greater than the response to vehicle only at the 15 min time point (Figure 1A, B). Ten micromolar ET-1 produced significant antinociception which lasted for 3 hours, whereas the maximum, reversible duration of antinociception was seen with 50uM ET-1, with antinociception lasting for 4 hours when tested at the weaker force (Figure 1A) and 3 hours at the stronger one (Figure 1B). Although in one study we attempted using 100 uM ET-1, 3 of the 4 rats injected became hyperactive, constantly licked at the site of injection, defecated more frequently, and died within 2 hours.
Considering the responses to the strong force only, and correcting for the brief antinociception from vehicle, the ET-1 concentration that reduced the CTMR response by half at its peak effect, i.e., that caused 50% antinociception, is ~1 μM. Analysis of area-under-the-curve for these anti-nociceptive actions shows that ~ 5μM ET-1 gave about 50% of the maximum area (Figure 1C).
ET-1’s antinociceptive effect was also seen at distances of 1 and 2 cm from the injection site, although the durations were shorter at these locations. There was less than 0.25 reduction in the graded response seen at 2 cm from the site of injection at 15-min after 5uM ET-1 (p>0.05 compared to vehicle), compared to the drop of 0.6, constituting complete antinociception, seen within the injection-raised wheal (0.5 cm).
A decline of the antinociceptive effect after one injection of ET-1 could be due to cellular uptake of the peptide, 15, 53 or to proteolytic degradation13, or to desensitization of the active receptors 8 or their downstream pathways. We tested this second possibility by injecting a second, higher dose of ET-1 (5uM, 50uL) at 4h after the initial dose of 1 μM ET-1, when the antinociceptive effect had just returned to baseline. Injecting the first dose gives a robust response similar to the responses observed in figure 1. In contrast, with the second dose there was a much more rapid return to baseline compared to the first dose. The AUC from the second injection of ET-1 was reduced to 75±14%, for the strong force, and to 53±35%, for the weak force, when compared to responses to a single injection of 5μM ET-1 (Figure 2).
The receptor specificity for antinociception was determined by co-injecting animals with ET-1 plus selective ET receptor antagonists. Treating rats with an ETA receptor antagonist, BQ-123, effectively abolished ET-1-induced analgesia in a concentration-dependent manner (Figures 3A, 3B, and 3C) with peak antinociception blocked by more than 70% for the weak force, and more than 90% for the strong force (Figure 3E). Analysis of maximum possible effect (%MPE) for this inhibition gave an IC50 of ~ 0.3mM. In contrast, blockade of ETB receptors with the selective antagonist BQ-788, at 3mM, had no effect on ET-1-induced analgesia (Figure 3D), showing the exclusive involvement of ETA receptors in this phenomenon.
Previous studies have shown that an antinociceptive action of ET-1, mediated through ETB receptors, in the rat’s glabrous skin involves the direct participation of β-endorphin and is blocked by antagonists of μ-opiate receptors40. We tested for such a mechanism in the antinociceptive actions of ET-1 in the hairy skin. Co-injection of the non-selective opiate receptor antagonist naloxone (2.75 mM) with 10μM ET-1 did not alter antinociception (data not shown), showing that opiate receptors are not involved in the hairy skin.
Earlier studies had shown that ET-1 applied to the rat sciatic nerve in vivo caused analgesia and conduction block, accompanied by a transient reduction in endoneurial blood flow68. These effects were prevented by systemic (i.p.) administration of nimodipine, a selective inhibitor of L-type voltage-gated Ca2+ channels. To determine if local antinociception caused by s.c. ET-1 also involved such Ca2+ channels, we co-injected nimodipine (40μM) with ET-1 (10 μM). This concentration of the channel inhibitor strongly attenuated the effects of ET-1 (Figure 4), although lower concentrations, of 5uM or 10uM, were ineffective (data not shown).
In order to determine if vasoconstriction and the associated reduction in blood flow could, in general, produce antinociception, we examined the effects of the classic vasoconstrictor epinephrine3. Subcutaneous injection of epinephrine (45uM) in the back resulted in transient antinociception, detectable at 15 min and lasting for about 2h (Figure 5 A,B), as previously reported43. When epinephrine was co-co-injected with the α1- adrenoreceptor antagonist prazosin (10μM) the antinociceptive effect was reduced by more than 90% (Figures 5 A-C). Antinociception from epinephrine selectively involved α1- adrenoreceptors since yohimbine, an α2-adrenoreceptor antagonist, did not significantly affect analgesia induced by epinephrine (Figure 5C).
Endothelin receptors are present on many types of cells in the skin, including those in vascular tissue and in keratinocytes themselves (see review41,) and it is possible that a series of linked stimulus-release reactions could couple exogenous ET-1 to the release of endogenous epinephrine and subsequent activation of α1-adrenoreceptors, and thus cause antinociception. This possibility was eliminated by results that showed that co-injection of ET-1 with prazosin, at the same inhibitor concentration that almost abolished epinephrine-induced analgesia (10μM), had no effect on ET-1-induced antinociception (Figure 5C).
The question remains whether a common vasoconstrictive effect, caused by either ET-1 acting through activation of ETA receptors or by epinephrine acting through α1- adrenoreceptors, might account for the antinociceptive effect of these agents. To initially test this notion, we confirmed that ET-1 at 30 μM, a concentration within those that produced antinociception, caused local vasoconstriction. Laser-Doppler measurements of blood flow in the hairy skin showed a rapidly developing, constant reduction of flow for at least 50 mins after injection of 30μM ET-1, from a baseline value of 18.3 ± 1.6 TPU to 6.9 ± 0.6 TPU (Figures (Figures6A6A and and7).7). Injection of nimodipine (40 μM) 5 min prior to the injection of ET-1 almost abolished the reduction in blood flow caused by ET-1; the blood flow decreased from 19.6 ± 1.2 TPU to 15.7 ± 2.4 TPU (Figures (Figures6B,6B, ,7).7). Nimodipine by itself had no effect on blood flow (Figure 6C), nor did the injection of vehicle (Figures (Figures6D,6D, ,77).
This paper reports that: 1) subcutaneous injection of ET-1 causes a local, transient antinociception to punctate mechanical stimulation of the hairy back skin of the rat that is concentration-dependent, with complete, reversible absence of response lasting for ~3 h with 50uM ET-1; 2) antinociception correlates pharmacologically with reduced blood flow, being mediated exclusively through ETA receptors and involving L-type Ca2+ channels. ET-1 is known to have many diverse biological actions, including its traditional vasoactive effects, positive inotropic actions on the heart, and stimulation of gene expression49,32. However, this is the first report of antinociception resulting from ET-1. In contrast, it was previously reported that ET-1’s injection into the glabrous skin of the rat plantar hindpaw, or application directly to sciatic nerve, causes ETAR-dependent flinching behavior 12, 18, 23, which is inhibited by morphine, testifying to the pain-like nature of this response. Injection of ET-1 into the joints causes motor incapacitation, apparently due to elevated pain19 and tissue levels of ET-1 are elevated in arthritic conditions, suggesting a role for endogenous ET-1 in arthritic pain14.
Although both ET-1 and epinephrine appear to cause antinociception in hairy skin primarily through vasoconstriction, they operate via different pathways. Epinephrine is a less potent vasoconstrictor44, and, consistent with this, antinociception resulting from epinephrine (45uM) develops more slowly over time, with an initial 10-20min latency before detectable functional loss after injection, and has a smaller effect, a maximum loss of response to pinprick, of ~80%, occurring 30-40 min after injection with full recovery by 90 min43. In contrast, ET-1 (50uM) produces a more rapid onset, causing complete antinociception at 15 min after injection that lasts for several hours. Such a rapidly developing block is consistent with the very rapid drop in blood flow that occurs after ET-1 injection (Figure 6). Interestingly, epinephrine caused observable blanching of the injected hairy skin but ET-1 caused no local pallor or flare. Perhaps the locations of the ETA receptors that cause vasoconstriction in the skin are different from those for the α1-adrenoreceptors, even though they result in the same gross vascular response.
Nimodipine attenuated ET-1 induced antinociception through its selective blockade of L-type channels. 25 ET-1 is known to activate a phosphoinositide cascade, which in turn produces IP3 that triggers Ca2+ release by mobilizing intracellular stores and extracellular Ca+2 59. Our finding lends further support to the role of ET-1 in a biphasic elevation of cytosolic Ca+2 58 and acute vasoconstriction of microvessels induced by epineurial ET-1, resulting in endoneurial ischemia and axonal conduction block68.
A second injection of ET-1, delivered after recovery to baseline of the antinociception from the first injection, produced an attenuated effect, demonstrating tachyphylaxis, as seen with nociception induced by direct application of ET-1 to sciatic nerve18. Such tachyphylaxis is consistent with the known desensitization of the ETA receptor8, 21.
The question remains whether there is an anatomic or physiologic basis for the opposite effects on pain from ET-1 injected into hairy versus glabrous skin. The answer is complex, because the behavioral effects of ET-1 result from the local effects on cutaneous neurons, on skin cells, e.g., keratinocytes, that release neuroactive substances42, 61, and on vascular tissue essential for supporting local metabolism as well as for delivering bioactive agents. Some differentiating features between glabrous and hairy skin are listed in Table 1. In the glabrous skin, pain from s.c. exogenous ET-1 is suppressed by local co-injection of the ETB receptor agonist, IRL-1620, in a naloxone-reversible manner, a reflection of a local opioidergic mechanism for ET-1-induced anti-hyperalgesia40, 42. This mechanism was not involved in the antinociception observed here, since naloxone had no effect on the reduced response to tactile stimulation caused by ET-1. There are several possible explanations for this difference between skin types, including an absence of ETB receptors on keratinocytes in hairy skin, a lack of coupling between such ETB receptors and release of β-endorphin, or an absence of μ-opiate receptors on nociceptor endings in hairy skin. Although data are not available for rat skin, human hairy skin is known to express both β-endorphin in keratinocytes and μ-opiate receptors, on both keratinocytes and non-myelinated nerve endings4. However, the presence of ETB receptors on keratinocytes and their coupling to release β-endorphin has not been shown in hairy skin.
In glabrous skin epinephrine, like ET-1, exerts the opposite behavioral effect that it does in hairy skin.39 Acting through β-adrenoreceptors, epinephrine induces tactile hyperalgesia in the glabrous skin, contrasting with the α-adrenoreceptor-induced anti-nociception in hairy skin. Although the authors of this study proposed from in vitro experiments that epinephrine is acting directly on cutaneous afferents to induce hyperalgesia 39, others have presented in vivo results that suggest that epinephrine (released from sympathetic nerve endings) only modulates cutaneous afferents indirectly, through vascular control. 17 It seems apparent that interactions between nociceptive afferents and blood vessels in the skin are reciprocal and involve several potential feedback loops 17, 27-29 and that ET-1 may influence these two systems in a number of different, dynamically responding ways.
We hypothesize that the pro-algesic actions of ET-1, initiating nociceptor firing via ETA receptors, occur in both glabrous (paw) and hairy (back) skin, but that the anti-hyperalgesic, vasoconstrictive actions dominate on the back and are less prominent on the paw. Both vasoconstriction and vasodilatation have been observed after ET-1 administration, the former occurring at higher concentrations of ET-1 than the latter10,44. Whereas vasoconstriction was inhibited by Ca2+ channel blockers44, 45, as observed in the present study, vasodilatation from ET-1 was blocked by local H1 antagonists and also by s.c. local anesthetics, suggesting a mechanism involving the local release of histamine10. We did not see any evidence of increased blood flow after ET-1 in the Laser-Doppler flowmetry records from the hairy skin, but we have observed a delayed rubor in the plantar paw following an initial blanching after ET-1 injection there (initial blanching of the paw develops at ~1 min and lasts for 15-24 min after subcutaneous injection of 200 uM ET-1 -1; 10 uM ET-1 also induces blanching, that appears no sooner than 5-6 min after injection. A. Khodorova, personal observations). Activation of ETA receptors known to be present on smooth muscle cells of the neural microvessels9,11 causes a temporary, local reduction in blood flow6,7,10. Studies using radionuclide (133Xe) clearance techniques to measure skin blood flow60 have shown a > 60% reduction in flow caused by 10pmol ET-1 injection7,10, consistent with our own finding (albeit after 250 pmol of ET-1), but using Laser Doppler flowmetry62. It is noteworthy, however, that 133Xe clearance measures the blood flow through all layers of the skin, whereas Laser Doppler flowmetry selectively captures the flow from the more superficial layers34. The overall effect on blood flow by ET-1’s sub-dermal injection is likely to result from the difference between vasoconstrictive and vasodilatory actions, perhaps with the former dominating in the back and causing anti-nociception, and the latter occurring in the paw. Vasoconstriction may also result from a stronger effect of ET-1 on sub-dermal vessels, located in the deeper levels that are closer to the locus of ET-1’s delivery. A differential control of superficial compared to the deeper cutaneous vasculature in glabrous skin is known to result from the differential innervation by primary afferent and sympathetic fibers 27, with the former accounting for local vasodilation and the latter for vasoconstriction. 29
With respect to differences in vasculature, back skin is known to contain predominantly nutritive, capillary vessels with high resistance to flow while paw skin has a high density of arterioles and venules, with low resistance, high flow vessels, at a 3-fold higher density than in the back56. ET-1 is reported to contract isolated human resistance vessels in vitro33, consistent with a stronger vasoconstriction on the back than on the paw skin. Skin vascular anatomy and physiology alone, therefore, might explain the difference in pain reactions to ET-1.
The innervation of hairy and glabrous skin also differs, with a substantially higher density of epidermal neurites in the glabrous skin 32, 38, 51. Physiological responses to thermal stimuli differ between these skin types in humans, with a lower heat detection threshold but a higher heat pain threshold in the palm than on hairy skin24, correlated with the abundance of heat-sensitive C-fibers in the glabrous skin. Glabrous skin also contains cold-activated fibers which are virtually absent in hairy skin 54, as well as generally smaller receptive fields for mechano-receptive afferents that are sensitive to punctate stimuli. 54 However, there are no general rules for differences in sensory innervation between these skin types, exemplified by similar densities of Merkel cells 20 and of CGRP-expressing neurites 47 in both skin types but differences in the acid-sensing ion channel expression between them 37. It is possible that the generally lower density innervation of the hairy skin accounts in part for its greater behavioral susceptibility to ischemia, since equal degrees of nerve block in glabrous and hairy skin might leave the latter with a subliminal afferent output when the former was still able to evoke sensation.
Although we have no direct evidence of conduction block from ischemia in these studies, others have shown that ischemia of peripheral nerve leads to increased neuronal accommodation5 and slowing of conduction and lengthening of the refractory period26, all three being indicators of reduced axonal excitability. These changes in conduction parameters were made in large myelinated axons, and it was determined that prolonged membrane depolarization could not alone account for the conduction defects, suggesting that some neuroactive substance or metabolic shift had a direct effect on Na+ channel inactivation48. If such a substance also is elevated by ischemia of small, cutaneous afferents then, given the much larger surface:volume ratio in these fibers, the putative blocking substance could reach even higher concentrations. But regardless of the mechanism, there is ample published evidence of ischemia-induced peripheral nerve block that would couple reduced blood flow to sensory deficits.
In conclusion, sensory losses, exemplified by the antinociceptive effects of ET-1, can result from local vascular actions of many types of neuroactive agents. Such deficits will affect physiological and behavioral assays of drug actions, and can thereby confound analyses when only targets on neuronal cells are considered. A broad accounting of the several factors that can alter neuronal activity in vivo is necessary in understanding drug actions.
This research was supported by grants from the NIH (NIH/NCI CA08053, to GS, and NS048565 to MRV). Some of the work was conducted in a facility constructed with support from Research Facilities Improvement Program Grant Number C06 RR015481- 01 from the National Center for Research Resources, National Institutes of Health.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.