- Email: [email protected]
Functional reinnervation of the vasculature of the adult cat paw pad by axons originally innervating vessels in hairy skin
Neuroscience Vol. 67, No. 1, pp. 245-252, 1995
~
Pergamon
Ltd
0306-4522(95)00033-X
Elsevier Science Copyright © 1995 IBRO Printed in Great Britain. All rights reserved 0306-4522/95 $9.50 + 0.00
F U N C T I O N A L REINNERVATION OF THE VASCULATURE OF THE A D U L T CAT PAW PAD BY AXONS ORIGINALLY I N N E R V A T I N G VESSELS IN HAIRY SKIN M. K O L T Z E N B U R G , * H.-J. H , A B L E R t and W. J,ANIG]':~ *Neurologische Universitfits-Klinik, Josef-Schneider-Stral3e 11, D-97080 Wiirzburg, Germany tPhysiologisches Institut, Christian-Albrechts-Universit/it zu Kiel, Olshausenstrage 40, D-24098 Kiel, Germany Abstract--Sympathetic vasoconstrictor neurons that had previously innervated blood vessels in hairy skin were made to reinnervate the vasculature of the hairless skin of the paw pad by suturing the central stump of the cut sural nerve to the distal stump of the cut tibial nerve. After allowing sufficient time for the reinnervation, electrical stimulation of the vasoconstrictor pathway in the lumbar sympathetic trunk produced a reduction of the blood flow that was significantly greater than in control animals. There was also a clear sign of a "denervation supersensitivity" of the blood vessels as evidenced by a significantly increased vasoconstriction that followed the systemic application of the alphal-adrenoceptor specific agonist phenylephrine. Neurogenic vasodilatation evoked by antidromic excitation of small diameter primary afferent neurones was significantly impaired although myefinated and unmyelinated primary afferents had re-grown into the target tissue. Electrical stimulation of the intact tibial nerve (containing sympathetic vasoconstrictor axons and nociceptive primary afferent fibres) in control animals, always produced vasodilatation indicating that the neurogenic vasodilatation can override the sympathetic vasoconstrictor response. By contrast, electrical stimulation of cross-unioned nerves consistently produced a robust vasoconstriction. We conclude that sympathetic vasoconstrictor neurons have a high capacity to functionally reinnervate autonomic effector organs in the adult cat. Despite this functional recovery, the blood vessels exhibited stronger than normal responses to an alpha~-adrenoceptor agonist. The impaired neurogenic vasodilatation mediated by small diameter afferents may be due to their poor ability to re-establish their efferent vasodilatory function. Alternatively it may be masked by the strong vasoconstriction.
The blood vessels of the hairless skin of the cat paw pads are supplied by three types of unmyelinated fibres. These are sympathetic vasoconstrictor fibres, sympathetic vasodilator fibres ~2'1s and primary afferent fibres. Cutaneous vasoconstrictor neurons, particularly to the distal extremities, are important for thermoregulation. They are normally activated during adaptation to cold and under conditions of stress. Heat load decreases the activity in these neurons. The functions of vasodilator neurons are unknown although it is believed that they are also involved in thermoregulation. 3 Small afferent fibres with nociceptive function generate local vasodilation when activated, 2 probably by release of neuropeptides (substance P and/or calcitonin gene-related peptide, C G R P ) which act on the precapillary resistance vessels, ll'2j In addition there are sweat glands which receive a sympathetic innervation which is cholinergic. Hairless skin differs from hairy skin (in non-primate mammals) in that the latter contains no sweat glands. The calibre of blood vessels in hairy skin is affected by vasoconstrictor activity and by release of vasodilator neuropeptides from primary afferent :~To whom correspondence should be addressed. Abbreviation: CGRP, calcitonin gene-related peptide.
endings; the sympathetic vasodilator component is weak or absent and pilomotor muscles of the distal parts of the extremities are probably not innervated.12 The functions of unmyelinated fibres in cutaneous nerves of the cat hindlimb have been extensively characterized (for reviews see Refs 12, 13, 31). The paw pads are primarily supplied by axons of the tibial nerve (by way of the medial and lateral plantar nerves). The sural nerve supplies only hairy skin over the lateral part of the paw. These anatomical separations provide a useful preparation to examine the ability of different types of nerve fibres to reinnervate the vessels of the pad. Peripheral nerve injuries are followed by changes of cutaneous blood flow which are produced by denervation and reinnervation, and the question as to the extent of the functional reinnervation may be complicated by changes in the reactivity of the vasculature. 19'27 It is known that postganglionic fibres can regenerate to supply arterial blood vessels but the functional result of this reinnervation is p o o r if the vessel is distant. 17 In traumatic nerve injuries, the regeneration process is also dependent on the conditions at the site of the lesion, so that even for large m o t o r and sensory fibres the outcome is limited if there is more that a crush injury. 245
246
M. Koltzenburg et al.
Here we have studied a model of peripheral nerve injury in which we have examined the extent o f functional r e i n n e r v a t i o n o f the b l o o d vessels of the cat paw p a d by p o p u l a t i o n s of axons some o f which differed functionally from the original innervation. This would be the c o m m o n consequence of a n injury in which a m a j o r nerve t r u n k was transected a n d fibres r a n d o m l y innervate w h a t e v e r target o r g a n they h a p p e n to contact. W e d e n e r v a t e d the cat paw p a d a n d directed into it (by m e a n s of cross-suturing the nerve trunks) p r i m a r y afferent a n d sympathetic vasoconstrictor axons which h a d previously supplied hairy skin. After m a n y m o n t h s we tested the b l o o d flow responses to electrical stimulation of nerves above the suture site a n d c o m p a r e d t h e m with those elicited in u n o p e r a t e d control animals. Moreover, the b l o o d flow responses to activation of purely s y m p a t h etic axons (via the l u m b a r sympathetic chain) were c o m p a r e d with responses to activation o f the entire reinnervating nerve t r u n k in order to assess antidromic vasodilatation.
EXPERIMENTAL PROCEDURES
Cross-union o f inappropriate nerves
Seven adult cats of either sex weighing 2.44.9 kg at the time of the operation were used for the cross-union experiments between inappropriate nerves. In these animals the proximal stump of the cut sural nerve was sutured to the distal stump of the tibial nerve, All experimental procedures had been approved by the local animal care committee of the state administration and were conducted in accordance with German Federal Law. After induction with ketamine hydrochloride (KetanestR; 15 mg/kg, i.m.) and diazepam (ValiumR; 0.2 mg/kg, i.m.) adequate surgical anaesthesia was maintained with repeated injections of methohexital (BrevimytalR; 10 20 mg per bolus, i.m.). With antiseptic precautions the tibial nerve was exposed at the ankle. After transection the central stump was ligated and the distal stump was repositioned over the lateral gastrocnemius muscle. Following transection of the sural nerve, the proximal stump was sutured with two or three epineurial stitches (Ethicon R 10-0, atraumatic) to the distal stump of the tibial nerve. The distal stump of the sural nerve was resected over a length of 1-2 cm. The exposure was then closed in layers. The distance from the nerve suture
~
spinal cord .
_r-L stim LST
site to the central pad of the paw ranged from 110-155 mm in different animals. After the initial postoperative recovery, the animals did not display any abnormal behaviour indicative of spontaneous pain and healing of the incision was uneventful. As expected, there was severe wasting of small foot muscles. However, the operation interfered with stance and gait only slightly and gross motor power of the limbs was well preserved allowing all animals to perform vertical jumps exceeding 1 m. Progress of nerve regeneration could be assessed by the flexion reflex elicited by pin-prick applied to the innervation territory of the tibial nerve as it was progressively invaded by regenerating inappropriate nerve fibres. In all animals, superficial wounds of the plantar surface which developed after denervation and which were insensitive to pin-prick disappeared once reinnervation had occurred. Terminal experiments
Acute and final experiments were conducted on 14 adult cats. Seven un-operated cats weighing 2.9-4.8 kg served as controls. The seven operated animals weighed 2.7-5.1 kg on the day of the terminal experiment 9-21.5 months after the operation. Following induction with ketamine hydrochloride (15-20 mg/kg, i.m.) anaesthesia was maintained with alphaD-gluco-chloralose (40-50 mg/kg, i.p.). Supplementary doses (5 10 mg/kg, i.v.) were given as required to maintain deep anaesthesia as judged by the persistence of miotic pupils, assessed by frequent inspection, and by the absence of blood pressure and heart rate fluctuations. Drugs were injected via a cannula into the left superficial jugular vein. Heart rate and blood pressure were continuously monitored after cannulation of the common carotid artery. Animals were immobilized by repeated injections of pancuronium bromide (Pancuronuimr~; 0.2 mg/kg, i.v.) and artificially ventilated through a tracheal cannula keeping the end-expiratory CO 2 concentration at 3-4% (v/v). Body core temperature was continuously monitored via an intraoesophageal thermistor and kept close to 38%C by a heating pad. After the final experiments the animals were killed by intravenous injection of a saturated aqueous solution of potassium chloride. Other data from some of the animals have been reported elsewhere. ~4 Nerve stimulation
The left lumbar sympathetic trunk and the peripheral nerves were prepared for electrical stimulation (Fig. 1). Using a lateral approach, the sympathetic trunk was exposed retroperitoneally and isolated from surrounding tissue with plastic sheaths between the ganglia L3 and L4 or L4 and Lh. The interconnecting trunk was then placed on
hairless skin [paw)
,¢.~
_r-L stim peripheral nerve
cross-union
recording of blood flow
Fig. 1. Schematic drawing of the preparation. Either lumbar sympathetic trunk (LST) or peripheral nerve were stimulated electrically (stim) and blood flow was recorded on the hairless skin of the central pad of the paw. Cross-unions were performed by suturing the proximal end of the cut sural nerve (SU) to the distal stump of the cut tibial nerve (TIB).
Reinnervation of skin blood vessels
247
CONTROL
A
relative blood flow
,v.] i
i71'6"
''1/32"
1/%
174
!
i Hz 2min
172
CROSS-UNION
/V 1132
g
1/16
1/8
1/2
1/4
1Hz
0v*.s
O control •
cross-union
O3 ¢-,
61
O O.
O
"------ -50O
..go ,,O
- 100
•
'
'
I
0.1
'
stimulation frequency
'
'
Hz
'
"
I
1
Fig. 2. Blood flow responses produced by electrical stimulation of the lumbar sympathetic trunk. (A) Stimulation with trains of five pulses delivered at 1/32-1 Hz at intertrain intervals of 1-2 rain in a control cat and in a cat after cross-union of the sural and the tibial nerve. Decrease of voltage indicates reduction of blood flow. (B) Stimulus-response functions of the integrated blood flow responses produced by electrical stimulation of the lumbar sympathetic trunk (trains of five pulses at 1/32 1 Hz) in control animals (n = 3) and following cross-union of the sural nerve and the tibial nerve (n = 4). There is a significantly stronger blood flow reduction following nerve cross-union (F~,33 = 15.1; P < 0.001; ANOVA). a pair o f platinum wire electrodes and the exposure covered with warm mineral oil in a pool made from surrounding tissue. Electrical stimulation of the preganglionic axons in the lumbar sympathetic trunk at this position results in the excitation of postganglionic sympathetic fibres innervating the hindlimb. ~624 In operated animals the sural nerve was exposed and prepared for electrical stimulation proximal to the nerve suture. All remaining intact nerves innervating the hind paw were cut. This always included the saphenous nerve and the common peroneal nerve. In control animals the intact tibial nerve was prepared for electrical stimulation. The lumbar sympathetic trunk and the peripheral nerves were stimulated with supramaximal pulses of 0.2~0.5 ms duration at frequencies of 1/32 to 4 Hz. Usually, the stimulus intensities were 5 15 V for the lumbar sympathetic
trunk, 15-25 V for the sural nerve and 30-60 V for the tibial nerve. As determined in earlier experiments these stimulus strengths are sufficient to excite all axons in the nerves. Peripheral nerve stimulation was avoided prior to the assessment of neurogenic vasodilatation. At the end of the terminal experiments it was tested by neurophysiological means whether afferent nerve fibres had reinnervated the plantar skin of the hindpaw. Nerve filaments were isolated from the peripheral end of the cut sural nerve proximal to the lesion site using teased fibre techniques and put on a recording electrode. Activity in the filament was recorded and amplified by a low noise amplifier. The plantar skin of the paw pads was then probed with both innocuous and noxious mechanical stimuli in order to assess whether activation of afferent fibres occurred in the recorded nerve filaments. It was determined by
248
M. Koltzenburg et al. CONTROL
A I V ~ relative bloodflow 1401 MAP H91
mm
ii0 J
t
t
0.05
t
O.l
t
0.2
t
0.4
0.8 pg/kg
CROSS - UNION
Data collection and statistical tests Recordings of mean blood pressure and blood flow were continuously displayed on an oscilloscope and digitized (Minc; PDP 11) for further analysis. All values are given as mean_+ S.E.M. For statistical comparison the data was evaluated with analysis of variance (ANOVA), analysis of covariance (ANCOVA) and the Newman-Keuls post hoc test as appropriate using the CSS software package (StarSoft, Tulsa, OK, U.S.A.).
140]
I IOJ
t 0.05
t 0.1
t 0.2
t , , t 0.4 Imin 0.8 pg/k9
B O-
!
~ o
o
o
tissue. However, the relative change of flow is measure as voltage readings [V*] which are linearly related to the number and velocity of the red blood cells passing under the probe such that the complete absence of flow corresponds to a value of 0 V. ~s Blood flow changes were integrated over time until the blood flow reaction had returned to baseline after each stimulus. This results in an apparent saturation at higher stimulation frequencies when the responses fuse (see Figs 2B and 4B). In control animals and in some animals with cross-union of sural and tibial nerves the vasoconstrictor action of the alpha~-adrenoceptor agonist phenylephrine was tested. Bolus injections of 0.05 3.2/~g in 1.0 ml normal saline were injected i.v. (followed by 2.0 ml of normal saline). Injections were given in ascending order of dose and administered when recovery from the blood flow reduction of a previous injection had reached baseline for at least 1 min.
RESULTS
o
V'.e
Reinnervation of the foot pad
-50-
! P -I00-_= • CrOSS--UOiOn -150
.
.
.
.
.
.
i
0.I Phenylephrine dose
1 .
.
.
.
.
.
pg/k 9
.
i
Fig. 3. Blood flow responses produced by an alpha~-adrenoceptor agonist. (A) Reactions of blood flow and mean arterial pressure (MAP) following intravenous bolus injection of the alpharselective adrenoceptor agonist phenylephrine in a control animal and following cross-union of the sural and the tibial nerve. Decrease of voltage indicates reduction of blood flow. (B) Stimulus-response functions of the integrated blood flow response following intravenous bolus injection of phenylephrine. There is a significantly stronger response of the blood flow in animals with a cross-union of the sural and the tibial nerve (n = 5) than in control animals (n = 6) (FL5~= 20.1; P < 0.001; ANOVA).
electrical stimulation of the tibial nerve distal to lesion site whether the afferents activated by physiological stimulation were unmyelinated or myelinated.
Measurement of cutaneous bloodflow Cutaneous blood flow through the dermis of the hairless skin of the paw was measured using a lase~Doppler flowmeter (Periflux PF2 and standard probe PF 108; Perimed) with a time constant set at 1.5 s and amplification factors of 1 and 3. The laser-Doppler device does not provide an absolute value of the blood flow per volume of
The foot pad of all operated animals had been reinnervated by axons of the sural nerve. This was demonstrated by the presence of a flexion reflex with pin-prick stimulation of the foot pad in the unanaesthetized animals. Regeneration of mechanoreceptive primary afferent neurons was confirmed by recording of whole-nerve activity proximal to the suture site evoked by light and strong mechanical stimuli applied to the hairy and hairless skin of the foot. Teased fibre techniques were used to record activity of both myelinated and unmyelinated fibres that had crossed the suture site and which were activated by electrical stimulation of the tibial nerve and/or by mechanical stimulation applied to the foot pads. These studies therefore were consistent with previous anatomical and electrophysiological observations which have determined that unmyelinated fibres can successfully regenerate across a lesion site. 14'2°
Vasoconstriction evoked by electrical stimulation of the lumbar sympathetic trunk following cross-union Following cross-union of the sural and tibial nerves there was evidence of functional reinnervation of the foot pad by sympathetic vasoconstrictor fibres. Electrical stimulation of preganglionic vasoconstrictor axons in the lumbar sympathetic trunk produced strong reductions in paw blood flow in all animals (shown as a reduction in voltage from the laser-Doppler device, Fig. 2A). While stimulation with single impulses produced short periods of blood flow reduction, stimulation with trains of impulses at submaximal frequencies for vasomotor responses but higher than 1/8 Hz typically resulted in a fused and
Reinnervation of skin blood vessels
normal un-operated animals, intravenous injection of phenylephrine (0.05-0.8 # g/kg) produced virtually no change in the blood flow through the foot pad (Fig. 3A upper panel; Fig. 3B control). Doses higher than 0.8 #g/kg were required to produce clear paw vasoconstriction in the normal animal. After cross-union of sural and tibial nerves, the vasoconstriction produced by phenylephrine was always markedly enhanced and the dose-response curve was shifted to the left. Often doses as low as 0.05 #g/kg were sufficient to induce clear vasocon-
maintained vasoconstriction. Quantitative comparison of the integrated blood flow reaction over time showed that vasoconstriction was significantly stronger in the operated animals than in the normal controls (Fig. 2B).
Hyper-reactivity of blood vessels Despite the functional recovery of neurally-mediated vasoconstriction after nine months or longer there were clear signs of persisting hypersensitivity to the alpha]-adrenoceptor agonist phenylephrine. In
A
249
CONTROL relative blood flow
•
•
•
•
•
mlml
1/32
•
1/16
i
m
I/8
D
I/4
m
I/2
I
Hz
CROSS-UNION
,vt i •
•
•
•
I
l i l l e
I/32
B
m
I / 16
I/B
2rnin
i
i
I/4
I/2
Hz
ioo control
V * - s
g
•
cross-union
.o O
0 0
"o
0
0 0
"0
E ~ o o
- I00
'
'
'
'
I
'
'
'
'
'
'
0,1
'
I
I
Hz
stimulation frequency
Fig. 4. Blood flow responses produced by electrical stimulation of the peripheral nerve. (A) Stimulation of the nerve with series of five single pulses delivered at 1/32-1 Hz in a control animal and after cross-union of the sural and the tibial nerve. Increase of voltage indicates increase of blood flow whilst a decrease indicates blood flow reduction. (B) Stimulus-response functions of the integrated blood flow response following electrical stimulation of the peripheral nerve. While there is a blood flow increase in control animals (n = 4) there is a decrease of the blood flow in animals with a cross-union of the sural and the tibial nerve (n = 5) (FI,59 = 138.3; P < 0.001; ANOVA).
M. Koltzenburg et al.
250
striction (Fig. 3A). At higher doses there was a concentration-dependent increase in blood flow reduction in terms of both the absolute levels and the duration of the response. There was a highly significant difference (Fi,sl = 20.1; P < 0.001; ANOVA) between the integrated blood flow reaction over time in normal and operated animals (Fig. 3B). At all doses of phenylephrine tested there was a significant difference between normal controls and animals with nerve cross-union (P <0.05, Newman-Keuls post hoc test).
Neurogenic vasodilatation following stimulation of peripheral nerves Supramaximal electrical stimulation of peripheral nerves produces a vasodilatation that is brought about by the excitation of unmyelinated and small myelinated afferent fibres. This response persists for several minutes after short trains of stimulation and is so powerful that, at the stimulation frequencies used in the present experiments, it overrides the short-lasting sympathetically-mediated vasoconstrictionJ s'28 In the present investigation, this response pattern was always seen in both hairless and hairy skin following electrical stimulation of the tibial or sural nerve, respectively, (Fig 4A, upper record) in normal animals. However, following cross-union of sural and tibial nerves, a completely different response was observed. Rather than producing vasodilatation, electrical stimulation of the sural nerve at all frequencies evoked a strong vasoconstriction (Fig. 4A, lower record). This vasoconstriction was sometimes followed by a rebound increase of blood flow indicating that antidromic vasodilatation mediated by small diameter afferents was not totally absent. There was a highly significant difference (FI,59 = 138.3; P < 0.001; ANOVA) between the integrated blood flow reaction over time in normal animals and cats with cross-union of sural and tibial nerves (Fig. 4B). Further, the integrated blood flow reduction evoked by electrical stimulation of the cross-sutured peripheral nerve was not statistically different (F1,42 = 2.3; P > 0.1; ANCOVA, using stimulation frequency as covariate) from the blood flow responses induced by electrical stimulation of the lumbar sympathetic trunk in the same animals (compare curves with filled symbols in Figs. 2B and 4B).
nist produced a much stronger vasoconstriction in the reinnervated skin than in the control paw; and fourth, no clear signs of a neurogenic vasodilatation evoked by excitation of unmyelinated and thin myelinated afferent fibres were seen in the reinnervated paw, although sometimes a rebound increase of blood flow followed stimulation-induced vasoconstriction indicating that neurogenic vasodilatation might not have been totally absent. These changes were present in all animals studied and were unlikely to have arisen because of an inadequate postoperative period of regeneration. Assuming a rate of growth of 1 2 ram/day for regenerating unmyelinated fibres, 17'2°'27the survival period of more than 270 days was more than adequate for anatomical regeneration over the distance of 110-155 mm to the paw from the site of nerve suture. By the same token, the changes observed in the present investigation were unlikely to be transient phenomena occurring during the establishment of the reinnervation, because they persisted for over twelve months after the regenerated fibres would have established a functional relationship with the blood vessels in the hairless skin.
Functional recovery of vasoconstrictor function Reinnervated blood vessels in the paw pad constricted strongly in response to stimulation of either preganglionic or postganglionic axons. Vasoconstriction induced by electrical stimulation of the lumbar sympathetic trunk was more profound after the nerve cross-union than in the control condition. This is surprising as the sural nerve contains only 1/5 to 1/3 of the postganglionic fibres found in the tibial nerve. 24 The good functional recovery /n vivo contrasts with previous in vitro investigations which have demonstrated poor restoration of neuroeffector transmission after freeze lesions of vasoconstrictor fibres to branches of the mesenteric artery and to the tail artery in the rat. 1°'~7These neurophysiological results are also supported by poor histochemical evidence of regeneration. Even after long survival periods a significant portion of the target tissues was neither functionally nor anatomically reinnervated. The present data, however, indicate that nerve-evoked vasoconstriction may be greater than the control in the presence of reduced innervation and further studies will be necessary to explain this discrepancy.
Persistent hyper-reactivity of blood vessels DISCUSSION
The principal results of the present investigation are: first, paw blood vessels were reinnervated by vasoconstrictor neurons originally supplying hairy skin; second, the vasoconstriction induced by sympathetic stimulation is stronger than normal despite the fact that the available number of postganglionic axons in the sural nerve was at best about 30% of those present in the tibial nerve (see Ref. 24); third, systemic application of an alpha~-adrenoceptor ago-
It has been known for some time that denervated blood vessels become more sensitive to circulating agonists. 6'8 It is thought that this hyper-reactivity arises because of the loss of sympathetic fibres and so should disappear when the fibres regenerate. In the present investigation we have demonstrated that a marked hyper-reactivity of the blood vessels remains after functional restoration of reinnervation by postganglionic sympathetic fibres. The dose-response curve for the alphas-selective adrenoceptor agonist
Reinnervation of skin blood vessels phenylephrine was shifted markedly to the left both for the magnitude and the duration of the vasoconstriction. Recent in vitro investigations of reinnervated smooth muscle cells of the rat tail artery ~7 revealed that the excitatory junction potentials evoked by electrical stimulation of postganglionic fibres recovered their control amplitude in segments of the reinnervated artery within a few centimetres of the site of axon lesion. However, there was a significant increase of the depolarization mediated by neurally-released noradrenaline acting through alpha2-adrenoceptors suggesting a persistent hypersensitivity. It is presently not clear whether this effect is: first, due to noradrenaline reaching more extrajunctional receptors; second, caused by a modification of postreceptor mechanisms of the smooth muscle cell; or, third, whether it is the consequence of reduced re-uptake of noradrenaline into sympathetic varicosities. On the basis of the greater responsiveness to phenylephrine described here, the second possibility seems the most likely (see Ref. 8). This would be consistent with the observation that the vasoconstrictor responses to excitation of sympathetic axons were markedly enhanced. Absence o f neurogenic vasodilatation After reinnervation strong vasoconstriction was invariably elicited by the simultaneous electrical stimulation of all sympathetic and afferent fibres in the peripheral nerve as well as by selective electrical stimulation of the sympathetic supply. There was no statistically significant difference between the magnitude of vasoconstriction evoked by either stimulation indicating that vasodilatation evoked by antidromic stimulation of afferent fibres was virtually absent. Using teased fibre techniques, we have determined that unmyelinated afferent fibres can cross the lesion site and do respond to natural and electrical stimulation, 9 although we cannot exclude the possibility that the appropriate axons had failed to regenerate so as to affect blood vessels. Regenerating unmyelinated afferent fibres of rodents can occupy a larger than normal skin territory, 3° but the magnitude of neurogenie inflammation is known to be severely compromised following regeneration. L4'~8'2°'26 It has been recognized that the impairment of neurogenic inflammation after nerve regeneration correlates with a reduction of the substance P content of the regenerated axons 5 and the absence of these fibres in distal skin. 25 Thus, the impairment of vasodilatation in the present investigation could be explained by a limited ability of Substance P/CGRP fibres to regenerate. One could also argue that the vasodilatation induced by electrical stimulation of the peripheral nerve was masked by the more powerful vasoconstriction
251
evoked by excitation of the regenerated noradrenergic axons when the peripheral nerve was stimulated; i.e. that the hyper-reactivity of the reinnervated blood vessels to alpha-adrenoceptor agonists simply allowed the constriction which is induced by activation of the vasoconstrictor axons now to override the dilatation. This could account for the absence of short-lasting vasodilatation, and the rebound increase of blood flow sometimes following the vasoconstriction is in agreement with this interpretation, since the vasodilatation induced by antidromic excitation of unmyelinated afferent fibres usually outlasts the relatively short-lasting vasoconstriction and probably is related to long actions of peptides released. 3,15,28 Clinical implications The results of the present study bear on our understanding of autonomic dysfunction in neuropathies. Patients with severe degeneration of postganglionic sympathetic fibres as seen in pure autonomic failure generally develop hyper-reactivity of their blood vessels to circulating catecholamines such that low concentrations of exogenous catecholamines produce exaggerated blood pressure increasesfl3 However, the present data suggest that such hyper-reactivity may occur even when there is only partial denervation of blood vessels and a mild deterioration of neurally-mediated sympathetic control. It is possible that hyper-reactivity develops after partial denervation and regeneration of sympathetic fibres or after partial denervation alone. Moreover, some patients suffering from post-traumatic neuralgia can develop increased vasoconstrictor reactivity of blood vessels presenting with excessive cooling of the affected extremity. In contrast, direct microneurographic recordings of sympathetic multiunit activity have shown normal activity of regenerated units 29 or evidence of reduced noradrenaline release from sympathetic fibres in these patients. 7 The findings of the present investigation may reconcile these seemingly paradoxical observations in that normal or only slightly compromised neural sympathetic activity could interact with an abnormally sensitive vascular bed. Since it seems capable of overriding the vasodilator effects of afferent peptidergic sensory fibres, the consequences of this vasoconstriction may become even more exaggerated. Acknowledgements--We would like to thank Elspeth McLachlan for her valuable comments. We thank Nanke Bluhm, Barbara Howaldt and Eike Tallone for their skilful help. We are grateful to Dr Helmut Blumberg for generously letting us borrow his laser-Doppler flowmeter. This work was supported by Deutsche Forschungsgemeinschaft.
REFERENCES
I. Allnatt J. P., Dickson K. E. and Lisney S. J. W. (1990) Saphenous nerve injury and regeneration on one side of a rat suppresses the ability of the contralateral nerve to evoke plasma extravasation. Neurosci. Lett. 118, 219-222.
252
M. Koltzenburg et al.
2. Bayliss W. M. (1901) On the origin from the spinal cord of the vaso-dilator fibres of the hind-limb, and on the nature of these fibres. J. Physiol. 26, 173-209. 3. Bell C., J~inig W., Kiimmel H. and Xu H. (1985) Differentiation of vasodilator and sudomotor responses in the cat paw pad to preganglionic sympathetic stimulation. J. Physiol. 364, 93-104. 4. Bharali L. A. and Lisney S. J. W. (1992) The relationship between unmyelinated afferent type and neurogenic plasma extravasation in normal and reinnervated rat skin. Neuroscience 47, 703-712. 5. Bisby M. A. and Keen P. (1986) Regeneration of primary afferent neurons containing substance P-like immunoreactivity. Brain Res. 365, 8545. 6. Cannon W. B. and Rosenbleuth A. (1949) The Supersensitivity of Denervated Structures: A Law of Denervation. Macmillan, New York. 7. Drummond P. D., Finch P. M. and Smythe G. A. (1991) Reflex sympathetic dystrophy: The significance of differing plasma catecholamine concentrations in affected and unaffected limbs. Brain 114, 2025-2036. 8. Fleming W. W. and Westfall D. P. (1988) Adaptive supersensitivity. In Handbook o f Experimental Pharmacology, Vol. 90/1. Catecholamines I (eds Trendelenburg U. and Weiner N.), pp. 509-559. Springer, Heidelberg. 9. H/ibler H.-J., Jfinig W. and Koltzenburg M. (1987) Activation of unmyelinated afferents in chronically lesioned nerves by adrenaline and excitation of sympathetic efferents in the cat. Neurosci. Lett. 82, 35-40. 10. Hill C. E., Hirst G. D. S., Ngu M. C. and Van Helden D. F. (1985) Sympathetic postganglionic reinnervation of mesenteric arteries and enteric neurones of the ileum of the rat. J. auton, nerv. Syst. 14, 317 334. 11. Holzer P, (1988) Local effector functions of capsaicin-sensitive sensory nerve endings: involvement of tachykinins, calcitonin gene-related peptide and other neuropeptides. Neuroscience 24, 739-768. 12. J/inig W. (1985) Organization of the lumbar sympathetic outflow to skeletal muscle and skin of the cat hindlimb and tail. Rev. Physiol. Biochem. Pharmac. 102, 119 213. 13. J/inig W. (1988) Pre- and postganglionic vasoconstrictor neurons: differentiation, types and discharge properties. A. Rev. Physiol. 511, 525-539. 14. JS.nig W. and Koltzenburg M. (1991) Plasticity of sympathetic reflex organization following cross-union of inappropriate nerves in the adult cat. J. Physiol. 436, 309 323. 15. J~inig W. and Lisney S. J. W. (1989) Small diameter myelinated afferents produce vasodilatation but not plasma extravasation in rat skin. J. Physiol. 415, 477-486. 16. J/inig W. and McLachlan E. M.(1986) The sympathetic and sensory components of the caudal lumbar sympathetic trunk in the cat. J. comp. Neurol. 245, 62-73. 17. Jobling P., McLachlan E. M., J/inig W. and Anderson C. R. (1992) Electrophysiological responses in the rat tail artery during reinnervation following lesions of the sympathetic supply. J. Physiol. 454, 107-128. 18. Kolston J. and Lisney S. J. W. (1993) A study of vasodilator responses evoked by antidromic stimulation of Aft afferent nerve fibers supplying normal and reinnervated rat skin. Microvasc. Res. 46, 143-157. 19. Langley J. N. (1897) On the regeneration of preganglionic and of postganglionic visceral nerve fibres. J. Physiol. 22, 215-230. 20. Lisney S, J. W. (1989) Regeneration of unmyelinated axons after injury of mammalian peripheral nerve. Q. JI exp. Physiol. 74, 757-784. 21. Lisney S, J. W. and Bharali L. A. (1989) The axon reflex: an outdated idea or a valid hypothesis. News Physiol. Sci. 4, 45-48. 22. Lynn B. (1988) Neurogenic inflammation. Skin Pharma. 1, 217-224. 23. Mathias C. J. and Bannister R. (1992) Investigation of autonomic disorders. In Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System, 3rd edn (eds Bannister R. and Mathias C. J.), pp. 255-290. Oxford University Press, Oxford. 24. McLachlan E. M. and J~inig W. (1983) The cell bodies of origin of sympathetic and sensory axons in some skin and muscle nerves of the cat hindlimb. J. comp. Neurol. 214, 115-130. 25. McLachlan E. M., Keast J. R. and Bauer M. (1994) SP- and CGRP-immunoreactive axons differ in their ability to reinnervate the skin of the rat tail. Neurosci. Lett. 176, 147-151. 26. Ringkamp M., Reeh P. W. and Magerl W. (1993) The functional regeneration of fine afferent nerve fibers in rat skin following peripheral nerve lesion. I. Antidromic vasodilatation. Pfliigers Arch. 422, Suppl. 1, R51. 27. Sunderland S. (1991) Nerve Injuries and their Repair. Churchill Livingstone, London. 28. Szolcsany J. (1988) Antidromic vasodilatation and neurogenic inflammation. Agents Actions 23, 4-11. 29. Wallin B. G., Torebj6rk H. E. and Hallin R. G. (l 976) Preliminary observations on the pathophysiology of hyperalgesia in the causaligic pain syndrome. In Sensory Functions of the Skin in Primates (ed. Zotterman Y.), pp. 489-499. Pergamon Press, Oxford. 30. Wiesenfeld-Hallin Z., Kinnman E. and Aldskogius H. (1989) Expansion of innervation territory by afferents involved in plasma extravasation after nerve regeneration in adult and neonatal rats. Expl Brain Res. 76, 88-96. 31. Willis W. D. and Coggeshall R. E. (1991) Sensory Mechanisms of the Spinal Cord, 2nd edn. Plenum Press, New York, London. (Accepted 5 January 1995)
~
Pergamon
Ltd
0306-4522(95)00033-X
Elsevier Science Copyright © 1995 IBRO Printed in Great Britain. All rights reserved 0306-4522/95 $9.50 + 0.00
F U N C T I O N A L REINNERVATION OF THE VASCULATURE OF THE A D U L T CAT PAW PAD BY AXONS ORIGINALLY I N N E R V A T I N G VESSELS IN HAIRY SKIN M. K O L T Z E N B U R G , * H.-J. H , A B L E R t and W. J,ANIG]':~ *Neurologische Universitfits-Klinik, Josef-Schneider-Stral3e 11, D-97080 Wiirzburg, Germany tPhysiologisches Institut, Christian-Albrechts-Universit/it zu Kiel, Olshausenstrage 40, D-24098 Kiel, Germany Abstract--Sympathetic vasoconstrictor neurons that had previously innervated blood vessels in hairy skin were made to reinnervate the vasculature of the hairless skin of the paw pad by suturing the central stump of the cut sural nerve to the distal stump of the cut tibial nerve. After allowing sufficient time for the reinnervation, electrical stimulation of the vasoconstrictor pathway in the lumbar sympathetic trunk produced a reduction of the blood flow that was significantly greater than in control animals. There was also a clear sign of a "denervation supersensitivity" of the blood vessels as evidenced by a significantly increased vasoconstriction that followed the systemic application of the alphal-adrenoceptor specific agonist phenylephrine. Neurogenic vasodilatation evoked by antidromic excitation of small diameter primary afferent neurones was significantly impaired although myefinated and unmyelinated primary afferents had re-grown into the target tissue. Electrical stimulation of the intact tibial nerve (containing sympathetic vasoconstrictor axons and nociceptive primary afferent fibres) in control animals, always produced vasodilatation indicating that the neurogenic vasodilatation can override the sympathetic vasoconstrictor response. By contrast, electrical stimulation of cross-unioned nerves consistently produced a robust vasoconstriction. We conclude that sympathetic vasoconstrictor neurons have a high capacity to functionally reinnervate autonomic effector organs in the adult cat. Despite this functional recovery, the blood vessels exhibited stronger than normal responses to an alpha~-adrenoceptor agonist. The impaired neurogenic vasodilatation mediated by small diameter afferents may be due to their poor ability to re-establish their efferent vasodilatory function. Alternatively it may be masked by the strong vasoconstriction.
The blood vessels of the hairless skin of the cat paw pads are supplied by three types of unmyelinated fibres. These are sympathetic vasoconstrictor fibres, sympathetic vasodilator fibres ~2'1s and primary afferent fibres. Cutaneous vasoconstrictor neurons, particularly to the distal extremities, are important for thermoregulation. They are normally activated during adaptation to cold and under conditions of stress. Heat load decreases the activity in these neurons. The functions of vasodilator neurons are unknown although it is believed that they are also involved in thermoregulation. 3 Small afferent fibres with nociceptive function generate local vasodilation when activated, 2 probably by release of neuropeptides (substance P and/or calcitonin gene-related peptide, C G R P ) which act on the precapillary resistance vessels, ll'2j In addition there are sweat glands which receive a sympathetic innervation which is cholinergic. Hairless skin differs from hairy skin (in non-primate mammals) in that the latter contains no sweat glands. The calibre of blood vessels in hairy skin is affected by vasoconstrictor activity and by release of vasodilator neuropeptides from primary afferent :~To whom correspondence should be addressed. Abbreviation: CGRP, calcitonin gene-related peptide.
endings; the sympathetic vasodilator component is weak or absent and pilomotor muscles of the distal parts of the extremities are probably not innervated.12 The functions of unmyelinated fibres in cutaneous nerves of the cat hindlimb have been extensively characterized (for reviews see Refs 12, 13, 31). The paw pads are primarily supplied by axons of the tibial nerve (by way of the medial and lateral plantar nerves). The sural nerve supplies only hairy skin over the lateral part of the paw. These anatomical separations provide a useful preparation to examine the ability of different types of nerve fibres to reinnervate the vessels of the pad. Peripheral nerve injuries are followed by changes of cutaneous blood flow which are produced by denervation and reinnervation, and the question as to the extent of the functional reinnervation may be complicated by changes in the reactivity of the vasculature. 19'27 It is known that postganglionic fibres can regenerate to supply arterial blood vessels but the functional result of this reinnervation is p o o r if the vessel is distant. 17 In traumatic nerve injuries, the regeneration process is also dependent on the conditions at the site of the lesion, so that even for large m o t o r and sensory fibres the outcome is limited if there is more that a crush injury. 245
246
M. Koltzenburg et al.
Here we have studied a model of peripheral nerve injury in which we have examined the extent o f functional r e i n n e r v a t i o n o f the b l o o d vessels of the cat paw p a d by p o p u l a t i o n s of axons some o f which differed functionally from the original innervation. This would be the c o m m o n consequence of a n injury in which a m a j o r nerve t r u n k was transected a n d fibres r a n d o m l y innervate w h a t e v e r target o r g a n they h a p p e n to contact. W e d e n e r v a t e d the cat paw p a d a n d directed into it (by m e a n s of cross-suturing the nerve trunks) p r i m a r y afferent a n d sympathetic vasoconstrictor axons which h a d previously supplied hairy skin. After m a n y m o n t h s we tested the b l o o d flow responses to electrical stimulation of nerves above the suture site a n d c o m p a r e d t h e m with those elicited in u n o p e r a t e d control animals. Moreover, the b l o o d flow responses to activation of purely s y m p a t h etic axons (via the l u m b a r sympathetic chain) were c o m p a r e d with responses to activation o f the entire reinnervating nerve t r u n k in order to assess antidromic vasodilatation.
EXPERIMENTAL PROCEDURES
Cross-union o f inappropriate nerves
Seven adult cats of either sex weighing 2.44.9 kg at the time of the operation were used for the cross-union experiments between inappropriate nerves. In these animals the proximal stump of the cut sural nerve was sutured to the distal stump of the tibial nerve, All experimental procedures had been approved by the local animal care committee of the state administration and were conducted in accordance with German Federal Law. After induction with ketamine hydrochloride (KetanestR; 15 mg/kg, i.m.) and diazepam (ValiumR; 0.2 mg/kg, i.m.) adequate surgical anaesthesia was maintained with repeated injections of methohexital (BrevimytalR; 10 20 mg per bolus, i.m.). With antiseptic precautions the tibial nerve was exposed at the ankle. After transection the central stump was ligated and the distal stump was repositioned over the lateral gastrocnemius muscle. Following transection of the sural nerve, the proximal stump was sutured with two or three epineurial stitches (Ethicon R 10-0, atraumatic) to the distal stump of the tibial nerve. The distal stump of the sural nerve was resected over a length of 1-2 cm. The exposure was then closed in layers. The distance from the nerve suture
~
spinal cord .
_r-L stim LST
site to the central pad of the paw ranged from 110-155 mm in different animals. After the initial postoperative recovery, the animals did not display any abnormal behaviour indicative of spontaneous pain and healing of the incision was uneventful. As expected, there was severe wasting of small foot muscles. However, the operation interfered with stance and gait only slightly and gross motor power of the limbs was well preserved allowing all animals to perform vertical jumps exceeding 1 m. Progress of nerve regeneration could be assessed by the flexion reflex elicited by pin-prick applied to the innervation territory of the tibial nerve as it was progressively invaded by regenerating inappropriate nerve fibres. In all animals, superficial wounds of the plantar surface which developed after denervation and which were insensitive to pin-prick disappeared once reinnervation had occurred. Terminal experiments
Acute and final experiments were conducted on 14 adult cats. Seven un-operated cats weighing 2.9-4.8 kg served as controls. The seven operated animals weighed 2.7-5.1 kg on the day of the terminal experiment 9-21.5 months after the operation. Following induction with ketamine hydrochloride (15-20 mg/kg, i.m.) anaesthesia was maintained with alphaD-gluco-chloralose (40-50 mg/kg, i.p.). Supplementary doses (5 10 mg/kg, i.v.) were given as required to maintain deep anaesthesia as judged by the persistence of miotic pupils, assessed by frequent inspection, and by the absence of blood pressure and heart rate fluctuations. Drugs were injected via a cannula into the left superficial jugular vein. Heart rate and blood pressure were continuously monitored after cannulation of the common carotid artery. Animals were immobilized by repeated injections of pancuronium bromide (Pancuronuimr~; 0.2 mg/kg, i.v.) and artificially ventilated through a tracheal cannula keeping the end-expiratory CO 2 concentration at 3-4% (v/v). Body core temperature was continuously monitored via an intraoesophageal thermistor and kept close to 38%C by a heating pad. After the final experiments the animals were killed by intravenous injection of a saturated aqueous solution of potassium chloride. Other data from some of the animals have been reported elsewhere. ~4 Nerve stimulation
The left lumbar sympathetic trunk and the peripheral nerves were prepared for electrical stimulation (Fig. 1). Using a lateral approach, the sympathetic trunk was exposed retroperitoneally and isolated from surrounding tissue with plastic sheaths between the ganglia L3 and L4 or L4 and Lh. The interconnecting trunk was then placed on
hairless skin [paw)
,¢.~
_r-L stim peripheral nerve
cross-union
recording of blood flow
Fig. 1. Schematic drawing of the preparation. Either lumbar sympathetic trunk (LST) or peripheral nerve were stimulated electrically (stim) and blood flow was recorded on the hairless skin of the central pad of the paw. Cross-unions were performed by suturing the proximal end of the cut sural nerve (SU) to the distal stump of the cut tibial nerve (TIB).
Reinnervation of skin blood vessels
247
CONTROL
A
relative blood flow
,v.] i
i71'6"
''1/32"
1/%
174
!
i Hz 2min
172
CROSS-UNION
/V 1132
g
1/16
1/8
1/2
1/4
1Hz
0v*.s
O control •
cross-union
O3 ¢-,
61
O O.
O
"------ -50O
..go ,,O
- 100
•
'
'
I
0.1
'
stimulation frequency
'
'
Hz
'
"
I
1
Fig. 2. Blood flow responses produced by electrical stimulation of the lumbar sympathetic trunk. (A) Stimulation with trains of five pulses delivered at 1/32-1 Hz at intertrain intervals of 1-2 rain in a control cat and in a cat after cross-union of the sural and the tibial nerve. Decrease of voltage indicates reduction of blood flow. (B) Stimulus-response functions of the integrated blood flow responses produced by electrical stimulation of the lumbar sympathetic trunk (trains of five pulses at 1/32 1 Hz) in control animals (n = 3) and following cross-union of the sural nerve and the tibial nerve (n = 4). There is a significantly stronger blood flow reduction following nerve cross-union (F~,33 = 15.1; P < 0.001; ANOVA). a pair o f platinum wire electrodes and the exposure covered with warm mineral oil in a pool made from surrounding tissue. Electrical stimulation of the preganglionic axons in the lumbar sympathetic trunk at this position results in the excitation of postganglionic sympathetic fibres innervating the hindlimb. ~624 In operated animals the sural nerve was exposed and prepared for electrical stimulation proximal to the nerve suture. All remaining intact nerves innervating the hind paw were cut. This always included the saphenous nerve and the common peroneal nerve. In control animals the intact tibial nerve was prepared for electrical stimulation. The lumbar sympathetic trunk and the peripheral nerves were stimulated with supramaximal pulses of 0.2~0.5 ms duration at frequencies of 1/32 to 4 Hz. Usually, the stimulus intensities were 5 15 V for the lumbar sympathetic
trunk, 15-25 V for the sural nerve and 30-60 V for the tibial nerve. As determined in earlier experiments these stimulus strengths are sufficient to excite all axons in the nerves. Peripheral nerve stimulation was avoided prior to the assessment of neurogenic vasodilatation. At the end of the terminal experiments it was tested by neurophysiological means whether afferent nerve fibres had reinnervated the plantar skin of the hindpaw. Nerve filaments were isolated from the peripheral end of the cut sural nerve proximal to the lesion site using teased fibre techniques and put on a recording electrode. Activity in the filament was recorded and amplified by a low noise amplifier. The plantar skin of the paw pads was then probed with both innocuous and noxious mechanical stimuli in order to assess whether activation of afferent fibres occurred in the recorded nerve filaments. It was determined by
248
M. Koltzenburg et al. CONTROL
A I V ~ relative bloodflow 1401 MAP H91
mm
ii0 J
t
t
0.05
t
O.l
t
0.2
t
0.4
0.8 pg/kg
CROSS - UNION
Data collection and statistical tests Recordings of mean blood pressure and blood flow were continuously displayed on an oscilloscope and digitized (Minc; PDP 11) for further analysis. All values are given as mean_+ S.E.M. For statistical comparison the data was evaluated with analysis of variance (ANOVA), analysis of covariance (ANCOVA) and the Newman-Keuls post hoc test as appropriate using the CSS software package (StarSoft, Tulsa, OK, U.S.A.).
140]
I IOJ
t 0.05
t 0.1
t 0.2
t , , t 0.4 Imin 0.8 pg/k9
B O-
!
~ o
o
o
tissue. However, the relative change of flow is measure as voltage readings [V*] which are linearly related to the number and velocity of the red blood cells passing under the probe such that the complete absence of flow corresponds to a value of 0 V. ~s Blood flow changes were integrated over time until the blood flow reaction had returned to baseline after each stimulus. This results in an apparent saturation at higher stimulation frequencies when the responses fuse (see Figs 2B and 4B). In control animals and in some animals with cross-union of sural and tibial nerves the vasoconstrictor action of the alpha~-adrenoceptor agonist phenylephrine was tested. Bolus injections of 0.05 3.2/~g in 1.0 ml normal saline were injected i.v. (followed by 2.0 ml of normal saline). Injections were given in ascending order of dose and administered when recovery from the blood flow reduction of a previous injection had reached baseline for at least 1 min.
RESULTS
o
V'.e
Reinnervation of the foot pad
-50-
! P -I00-_= • CrOSS--UOiOn -150
.
.
.
.
.
.
i
0.I Phenylephrine dose
1 .
.
.
.
.
.
pg/k 9
.
i
Fig. 3. Blood flow responses produced by an alpha~-adrenoceptor agonist. (A) Reactions of blood flow and mean arterial pressure (MAP) following intravenous bolus injection of the alpharselective adrenoceptor agonist phenylephrine in a control animal and following cross-union of the sural and the tibial nerve. Decrease of voltage indicates reduction of blood flow. (B) Stimulus-response functions of the integrated blood flow response following intravenous bolus injection of phenylephrine. There is a significantly stronger response of the blood flow in animals with a cross-union of the sural and the tibial nerve (n = 5) than in control animals (n = 6) (FL5~= 20.1; P < 0.001; ANOVA).
electrical stimulation of the tibial nerve distal to lesion site whether the afferents activated by physiological stimulation were unmyelinated or myelinated.
Measurement of cutaneous bloodflow Cutaneous blood flow through the dermis of the hairless skin of the paw was measured using a lase~Doppler flowmeter (Periflux PF2 and standard probe PF 108; Perimed) with a time constant set at 1.5 s and amplification factors of 1 and 3. The laser-Doppler device does not provide an absolute value of the blood flow per volume of
The foot pad of all operated animals had been reinnervated by axons of the sural nerve. This was demonstrated by the presence of a flexion reflex with pin-prick stimulation of the foot pad in the unanaesthetized animals. Regeneration of mechanoreceptive primary afferent neurons was confirmed by recording of whole-nerve activity proximal to the suture site evoked by light and strong mechanical stimuli applied to the hairy and hairless skin of the foot. Teased fibre techniques were used to record activity of both myelinated and unmyelinated fibres that had crossed the suture site and which were activated by electrical stimulation of the tibial nerve and/or by mechanical stimulation applied to the foot pads. These studies therefore were consistent with previous anatomical and electrophysiological observations which have determined that unmyelinated fibres can successfully regenerate across a lesion site. 14'2°
Vasoconstriction evoked by electrical stimulation of the lumbar sympathetic trunk following cross-union Following cross-union of the sural and tibial nerves there was evidence of functional reinnervation of the foot pad by sympathetic vasoconstrictor fibres. Electrical stimulation of preganglionic vasoconstrictor axons in the lumbar sympathetic trunk produced strong reductions in paw blood flow in all animals (shown as a reduction in voltage from the laser-Doppler device, Fig. 2A). While stimulation with single impulses produced short periods of blood flow reduction, stimulation with trains of impulses at submaximal frequencies for vasomotor responses but higher than 1/8 Hz typically resulted in a fused and
Reinnervation of skin blood vessels
normal un-operated animals, intravenous injection of phenylephrine (0.05-0.8 # g/kg) produced virtually no change in the blood flow through the foot pad (Fig. 3A upper panel; Fig. 3B control). Doses higher than 0.8 #g/kg were required to produce clear paw vasoconstriction in the normal animal. After cross-union of sural and tibial nerves, the vasoconstriction produced by phenylephrine was always markedly enhanced and the dose-response curve was shifted to the left. Often doses as low as 0.05 #g/kg were sufficient to induce clear vasocon-
maintained vasoconstriction. Quantitative comparison of the integrated blood flow reaction over time showed that vasoconstriction was significantly stronger in the operated animals than in the normal controls (Fig. 2B).
Hyper-reactivity of blood vessels Despite the functional recovery of neurally-mediated vasoconstriction after nine months or longer there were clear signs of persisting hypersensitivity to the alpha]-adrenoceptor agonist phenylephrine. In
A
249
CONTROL relative blood flow
•
•
•
•
•
mlml
1/32
•
1/16
i
m
I/8
D
I/4
m
I/2
I
Hz
CROSS-UNION
,vt i •
•
•
•
I
l i l l e
I/32
B
m
I / 16
I/B
2rnin
i
i
I/4
I/2
Hz
ioo control
V * - s
g
•
cross-union
.o O
0 0
"o
0
0 0
"0
E ~ o o
- I00
'
'
'
'
I
'
'
'
'
'
'
0,1
'
I
I
Hz
stimulation frequency
Fig. 4. Blood flow responses produced by electrical stimulation of the peripheral nerve. (A) Stimulation of the nerve with series of five single pulses delivered at 1/32-1 Hz in a control animal and after cross-union of the sural and the tibial nerve. Increase of voltage indicates increase of blood flow whilst a decrease indicates blood flow reduction. (B) Stimulus-response functions of the integrated blood flow response following electrical stimulation of the peripheral nerve. While there is a blood flow increase in control animals (n = 4) there is a decrease of the blood flow in animals with a cross-union of the sural and the tibial nerve (n = 5) (FI,59 = 138.3; P < 0.001; ANOVA).
M. Koltzenburg et al.
250
striction (Fig. 3A). At higher doses there was a concentration-dependent increase in blood flow reduction in terms of both the absolute levels and the duration of the response. There was a highly significant difference (Fi,sl = 20.1; P < 0.001; ANOVA) between the integrated blood flow reaction over time in normal and operated animals (Fig. 3B). At all doses of phenylephrine tested there was a significant difference between normal controls and animals with nerve cross-union (P <0.05, Newman-Keuls post hoc test).
Neurogenic vasodilatation following stimulation of peripheral nerves Supramaximal electrical stimulation of peripheral nerves produces a vasodilatation that is brought about by the excitation of unmyelinated and small myelinated afferent fibres. This response persists for several minutes after short trains of stimulation and is so powerful that, at the stimulation frequencies used in the present experiments, it overrides the short-lasting sympathetically-mediated vasoconstrictionJ s'28 In the present investigation, this response pattern was always seen in both hairless and hairy skin following electrical stimulation of the tibial or sural nerve, respectively, (Fig 4A, upper record) in normal animals. However, following cross-union of sural and tibial nerves, a completely different response was observed. Rather than producing vasodilatation, electrical stimulation of the sural nerve at all frequencies evoked a strong vasoconstriction (Fig. 4A, lower record). This vasoconstriction was sometimes followed by a rebound increase of blood flow indicating that antidromic vasodilatation mediated by small diameter afferents was not totally absent. There was a highly significant difference (FI,59 = 138.3; P < 0.001; ANOVA) between the integrated blood flow reaction over time in normal animals and cats with cross-union of sural and tibial nerves (Fig. 4B). Further, the integrated blood flow reduction evoked by electrical stimulation of the cross-sutured peripheral nerve was not statistically different (F1,42 = 2.3; P > 0.1; ANCOVA, using stimulation frequency as covariate) from the blood flow responses induced by electrical stimulation of the lumbar sympathetic trunk in the same animals (compare curves with filled symbols in Figs. 2B and 4B).
nist produced a much stronger vasoconstriction in the reinnervated skin than in the control paw; and fourth, no clear signs of a neurogenic vasodilatation evoked by excitation of unmyelinated and thin myelinated afferent fibres were seen in the reinnervated paw, although sometimes a rebound increase of blood flow followed stimulation-induced vasoconstriction indicating that neurogenic vasodilatation might not have been totally absent. These changes were present in all animals studied and were unlikely to have arisen because of an inadequate postoperative period of regeneration. Assuming a rate of growth of 1 2 ram/day for regenerating unmyelinated fibres, 17'2°'27the survival period of more than 270 days was more than adequate for anatomical regeneration over the distance of 110-155 mm to the paw from the site of nerve suture. By the same token, the changes observed in the present investigation were unlikely to be transient phenomena occurring during the establishment of the reinnervation, because they persisted for over twelve months after the regenerated fibres would have established a functional relationship with the blood vessels in the hairless skin.
Functional recovery of vasoconstrictor function Reinnervated blood vessels in the paw pad constricted strongly in response to stimulation of either preganglionic or postganglionic axons. Vasoconstriction induced by electrical stimulation of the lumbar sympathetic trunk was more profound after the nerve cross-union than in the control condition. This is surprising as the sural nerve contains only 1/5 to 1/3 of the postganglionic fibres found in the tibial nerve. 24 The good functional recovery /n vivo contrasts with previous in vitro investigations which have demonstrated poor restoration of neuroeffector transmission after freeze lesions of vasoconstrictor fibres to branches of the mesenteric artery and to the tail artery in the rat. 1°'~7These neurophysiological results are also supported by poor histochemical evidence of regeneration. Even after long survival periods a significant portion of the target tissues was neither functionally nor anatomically reinnervated. The present data, however, indicate that nerve-evoked vasoconstriction may be greater than the control in the presence of reduced innervation and further studies will be necessary to explain this discrepancy.
Persistent hyper-reactivity of blood vessels DISCUSSION
The principal results of the present investigation are: first, paw blood vessels were reinnervated by vasoconstrictor neurons originally supplying hairy skin; second, the vasoconstriction induced by sympathetic stimulation is stronger than normal despite the fact that the available number of postganglionic axons in the sural nerve was at best about 30% of those present in the tibial nerve (see Ref. 24); third, systemic application of an alpha~-adrenoceptor ago-
It has been known for some time that denervated blood vessels become more sensitive to circulating agonists. 6'8 It is thought that this hyper-reactivity arises because of the loss of sympathetic fibres and so should disappear when the fibres regenerate. In the present investigation we have demonstrated that a marked hyper-reactivity of the blood vessels remains after functional restoration of reinnervation by postganglionic sympathetic fibres. The dose-response curve for the alphas-selective adrenoceptor agonist
Reinnervation of skin blood vessels phenylephrine was shifted markedly to the left both for the magnitude and the duration of the vasoconstriction. Recent in vitro investigations of reinnervated smooth muscle cells of the rat tail artery ~7 revealed that the excitatory junction potentials evoked by electrical stimulation of postganglionic fibres recovered their control amplitude in segments of the reinnervated artery within a few centimetres of the site of axon lesion. However, there was a significant increase of the depolarization mediated by neurally-released noradrenaline acting through alpha2-adrenoceptors suggesting a persistent hypersensitivity. It is presently not clear whether this effect is: first, due to noradrenaline reaching more extrajunctional receptors; second, caused by a modification of postreceptor mechanisms of the smooth muscle cell; or, third, whether it is the consequence of reduced re-uptake of noradrenaline into sympathetic varicosities. On the basis of the greater responsiveness to phenylephrine described here, the second possibility seems the most likely (see Ref. 8). This would be consistent with the observation that the vasoconstrictor responses to excitation of sympathetic axons were markedly enhanced. Absence o f neurogenic vasodilatation After reinnervation strong vasoconstriction was invariably elicited by the simultaneous electrical stimulation of all sympathetic and afferent fibres in the peripheral nerve as well as by selective electrical stimulation of the sympathetic supply. There was no statistically significant difference between the magnitude of vasoconstriction evoked by either stimulation indicating that vasodilatation evoked by antidromic stimulation of afferent fibres was virtually absent. Using teased fibre techniques, we have determined that unmyelinated afferent fibres can cross the lesion site and do respond to natural and electrical stimulation, 9 although we cannot exclude the possibility that the appropriate axons had failed to regenerate so as to affect blood vessels. Regenerating unmyelinated afferent fibres of rodents can occupy a larger than normal skin territory, 3° but the magnitude of neurogenie inflammation is known to be severely compromised following regeneration. L4'~8'2°'26 It has been recognized that the impairment of neurogenic inflammation after nerve regeneration correlates with a reduction of the substance P content of the regenerated axons 5 and the absence of these fibres in distal skin. 25 Thus, the impairment of vasodilatation in the present investigation could be explained by a limited ability of Substance P/CGRP fibres to regenerate. One could also argue that the vasodilatation induced by electrical stimulation of the peripheral nerve was masked by the more powerful vasoconstriction
251
evoked by excitation of the regenerated noradrenergic axons when the peripheral nerve was stimulated; i.e. that the hyper-reactivity of the reinnervated blood vessels to alpha-adrenoceptor agonists simply allowed the constriction which is induced by activation of the vasoconstrictor axons now to override the dilatation. This could account for the absence of short-lasting vasodilatation, and the rebound increase of blood flow sometimes following the vasoconstriction is in agreement with this interpretation, since the vasodilatation induced by antidromic excitation of unmyelinated afferent fibres usually outlasts the relatively short-lasting vasoconstriction and probably is related to long actions of peptides released. 3,15,28 Clinical implications The results of the present study bear on our understanding of autonomic dysfunction in neuropathies. Patients with severe degeneration of postganglionic sympathetic fibres as seen in pure autonomic failure generally develop hyper-reactivity of their blood vessels to circulating catecholamines such that low concentrations of exogenous catecholamines produce exaggerated blood pressure increasesfl3 However, the present data suggest that such hyper-reactivity may occur even when there is only partial denervation of blood vessels and a mild deterioration of neurally-mediated sympathetic control. It is possible that hyper-reactivity develops after partial denervation and regeneration of sympathetic fibres or after partial denervation alone. Moreover, some patients suffering from post-traumatic neuralgia can develop increased vasoconstrictor reactivity of blood vessels presenting with excessive cooling of the affected extremity. In contrast, direct microneurographic recordings of sympathetic multiunit activity have shown normal activity of regenerated units 29 or evidence of reduced noradrenaline release from sympathetic fibres in these patients. 7 The findings of the present investigation may reconcile these seemingly paradoxical observations in that normal or only slightly compromised neural sympathetic activity could interact with an abnormally sensitive vascular bed. Since it seems capable of overriding the vasodilator effects of afferent peptidergic sensory fibres, the consequences of this vasoconstriction may become even more exaggerated. Acknowledgements--We would like to thank Elspeth McLachlan for her valuable comments. We thank Nanke Bluhm, Barbara Howaldt and Eike Tallone for their skilful help. We are grateful to Dr Helmut Blumberg for generously letting us borrow his laser-Doppler flowmeter. This work was supported by Deutsche Forschungsgemeinschaft.
REFERENCES
I. Allnatt J. P., Dickson K. E. and Lisney S. J. W. (1990) Saphenous nerve injury and regeneration on one side of a rat suppresses the ability of the contralateral nerve to evoke plasma extravasation. Neurosci. Lett. 118, 219-222.
252
M. Koltzenburg et al.
2. Bayliss W. M. (1901) On the origin from the spinal cord of the vaso-dilator fibres of the hind-limb, and on the nature of these fibres. J. Physiol. 26, 173-209. 3. Bell C., J~inig W., Kiimmel H. and Xu H. (1985) Differentiation of vasodilator and sudomotor responses in the cat paw pad to preganglionic sympathetic stimulation. J. Physiol. 364, 93-104. 4. Bharali L. A. and Lisney S. J. W. (1992) The relationship between unmyelinated afferent type and neurogenic plasma extravasation in normal and reinnervated rat skin. Neuroscience 47, 703-712. 5. Bisby M. A. and Keen P. (1986) Regeneration of primary afferent neurons containing substance P-like immunoreactivity. Brain Res. 365, 8545. 6. Cannon W. B. and Rosenbleuth A. (1949) The Supersensitivity of Denervated Structures: A Law of Denervation. Macmillan, New York. 7. Drummond P. D., Finch P. M. and Smythe G. A. (1991) Reflex sympathetic dystrophy: The significance of differing plasma catecholamine concentrations in affected and unaffected limbs. Brain 114, 2025-2036. 8. Fleming W. W. and Westfall D. P. (1988) Adaptive supersensitivity. In Handbook o f Experimental Pharmacology, Vol. 90/1. Catecholamines I (eds Trendelenburg U. and Weiner N.), pp. 509-559. Springer, Heidelberg. 9. H/ibler H.-J., Jfinig W. and Koltzenburg M. (1987) Activation of unmyelinated afferents in chronically lesioned nerves by adrenaline and excitation of sympathetic efferents in the cat. Neurosci. Lett. 82, 35-40. 10. Hill C. E., Hirst G. D. S., Ngu M. C. and Van Helden D. F. (1985) Sympathetic postganglionic reinnervation of mesenteric arteries and enteric neurones of the ileum of the rat. J. auton, nerv. Syst. 14, 317 334. 11. Holzer P, (1988) Local effector functions of capsaicin-sensitive sensory nerve endings: involvement of tachykinins, calcitonin gene-related peptide and other neuropeptides. Neuroscience 24, 739-768. 12. J/inig W. (1985) Organization of the lumbar sympathetic outflow to skeletal muscle and skin of the cat hindlimb and tail. Rev. Physiol. Biochem. Pharmac. 102, 119 213. 13. J/inig W. (1988) Pre- and postganglionic vasoconstrictor neurons: differentiation, types and discharge properties. A. Rev. Physiol. 511, 525-539. 14. JS.nig W. and Koltzenburg M. (1991) Plasticity of sympathetic reflex organization following cross-union of inappropriate nerves in the adult cat. J. Physiol. 436, 309 323. 15. J~inig W. and Lisney S. J. W. (1989) Small diameter myelinated afferents produce vasodilatation but not plasma extravasation in rat skin. J. Physiol. 415, 477-486. 16. J/inig W. and McLachlan E. M.(1986) The sympathetic and sensory components of the caudal lumbar sympathetic trunk in the cat. J. comp. Neurol. 245, 62-73. 17. Jobling P., McLachlan E. M., J/inig W. and Anderson C. R. (1992) Electrophysiological responses in the rat tail artery during reinnervation following lesions of the sympathetic supply. J. Physiol. 454, 107-128. 18. Kolston J. and Lisney S. J. W. (1993) A study of vasodilator responses evoked by antidromic stimulation of Aft afferent nerve fibers supplying normal and reinnervated rat skin. Microvasc. Res. 46, 143-157. 19. Langley J. N. (1897) On the regeneration of preganglionic and of postganglionic visceral nerve fibres. J. Physiol. 22, 215-230. 20. Lisney S, J. W. (1989) Regeneration of unmyelinated axons after injury of mammalian peripheral nerve. Q. JI exp. Physiol. 74, 757-784. 21. Lisney S, J. W. and Bharali L. A. (1989) The axon reflex: an outdated idea or a valid hypothesis. News Physiol. Sci. 4, 45-48. 22. Lynn B. (1988) Neurogenic inflammation. Skin Pharma. 1, 217-224. 23. Mathias C. J. and Bannister R. (1992) Investigation of autonomic disorders. In Autonomic Failure: A Textbook of Clinical Disorders of the Autonomic Nervous System, 3rd edn (eds Bannister R. and Mathias C. J.), pp. 255-290. Oxford University Press, Oxford. 24. McLachlan E. M. and J~inig W. (1983) The cell bodies of origin of sympathetic and sensory axons in some skin and muscle nerves of the cat hindlimb. J. comp. Neurol. 214, 115-130. 25. McLachlan E. M., Keast J. R. and Bauer M. (1994) SP- and CGRP-immunoreactive axons differ in their ability to reinnervate the skin of the rat tail. Neurosci. Lett. 176, 147-151. 26. Ringkamp M., Reeh P. W. and Magerl W. (1993) The functional regeneration of fine afferent nerve fibers in rat skin following peripheral nerve lesion. I. Antidromic vasodilatation. Pfliigers Arch. 422, Suppl. 1, R51. 27. Sunderland S. (1991) Nerve Injuries and their Repair. Churchill Livingstone, London. 28. Szolcsany J. (1988) Antidromic vasodilatation and neurogenic inflammation. Agents Actions 23, 4-11. 29. Wallin B. G., Torebj6rk H. E. and Hallin R. G. (l 976) Preliminary observations on the pathophysiology of hyperalgesia in the causaligic pain syndrome. In Sensory Functions of the Skin in Primates (ed. Zotterman Y.), pp. 489-499. Pergamon Press, Oxford. 30. Wiesenfeld-Hallin Z., Kinnman E. and Aldskogius H. (1989) Expansion of innervation territory by afferents involved in plasma extravasation after nerve regeneration in adult and neonatal rats. Expl Brain Res. 76, 88-96. 31. Willis W. D. and Coggeshall R. E. (1991) Sensory Mechanisms of the Spinal Cord, 2nd edn. Plenum Press, New York, London. (Accepted 5 January 1995)