Injury-induced plasticity of the flexor reflex in chronic decerebrate rats

Neuroscience Vol. 16,No. 2, pp. 395404, 1985 Printed in Great Britain

0306-4522/85 $3.00 + 0.00 Pergamon Press Ltd 0 1985 IBRO

INJURY-INDUCED PLASTICITY OF THE FLEXOR REFLEX IN CHRONIC DECEREBRATE RATS C. J. WOOLF* and S. B. MCMAHON Cerebral Functions Research Group, Department of Anatomy, University College London, Gower Street, London WClE 6BT, U.K. Abstract-The hindlimb-flexor-withdrawal reflex elicited by stimulation of the skin of the hindpaw has been examined in chronic decerebrate rats. This flexor reflex manifests as a typical phasic avoidance response when measured either behaviourally in the decerebrate rat or electrophysiologically in the decerebrate-spinal preparation. Once the threshold of the cutaneous flexor-reflex afferents in the skin have been exceeded a brief burst of activity with only a short afterdischarge occurs in the flexor motoneurones. The response to sustained stimuli adapts rapidly. In the absence of any treatment to the hindlimb the threshold, duration and responsiveness of the reflex remains stable when tested repeatedly. Thermal or chemical stimuli of sufficient intensity to produce tissue injury and prolonged local inflammation in a hindpaw of the chronic decerebrate rat result in marked and long-lasting (several weeks) alterations in the ipsilateral withdrawal reflex. The mechanical threshold necessary to elicit the reflex by stimulation of the hindpaw falls so that light touch or brush can now elicit a response instead of the firm pressure or pinch required pre-injury. Suprathreshold stimuli to the inflamed skin generate a sustained oscillating pattern of flexion in contrast to the brief flicking movement found in control animals. Electromyographic recordings from the hamstring flexor muscles ipsilateral to the inflamed hindpaw show decreased mechanothresholds, increased spontaneous activity, prolonged afterdischarges to brief stimuli and a slowly adapting tonic response to sustained stimulation. Populations of single cutaneous mechanoreceptive C-primary afferents recorded both from untreated decerebrate rats and from rats with an inflamed hindpaw are indistinguishable in terms of their response properties. There is no difference in threshold, spontaneous activity or afterdischarge between the two populations. The possible mechanisms responsible for the conversion of the high threshold phasic flexor reflex into a low threshold tonic reflex are discussed as are the possible implications for sensory disorders that accompany chronic injury in man.

The hindlimb-flexor-withdrawal reflex elicited in the rat by noxious stimulation of the hindpaw is a typical phasic avoidance reflex consisting of a rapid flicking of the limb away from the source of the stimulus. The brief flick is produced by a limited, explosive burst of activity in flexor motoneurones once an adequate afferent input arrives in the spinal cord. The withdrawal reflex is qualitatively identical in intact, decerebrate and decerebrate-spinal animals although the threshold may vary between the different preparations.30.32In all preparations a feature of the phasic flexor reflex is its stability. There is little change in threshold, latency or duration if the reflex is tested repeatedly over several days with stimuli that do not produce tissue injury.-” Acute thermal injury to the hindpaw of a decerebrate rate results, however, in marked alterations in the flexor withdrawal reflex.3o There is a fall in the cutaneous threshold of the reflex and an increase in its responsiveness and duration which, in many respects, resembles the post-injury pain hypersensitivity that occurs in man.‘,” Sensory blockade of the site of the injury with a local anaesthetic does not reverse the post-injury-increased excitability of the flexor reflex3’ *To whom all correspondence Abbreviation:

should be addressed. EMG, electromyogram. 395

which implies that the injury-induced afferent input triggers, but is not necessary to sustain the excitability increase. Electrical stimulation of unmyelinated primary afferents can mimic the injury-induced reflex changes29 with brief conditioning stimuli (1 Hz for 20 s at C-fibre strength) producing up to 90 min of increased excitability. Further evidence that the excitability increase reflects an adaptive dynamic change within the spinal cord is the finding that acute thermal injuries, identical to those that increase the reflex excitability, fail to alter the response properties of single cutaneous C-fibre afferents in a way that can fully explain the changes found in the flexor motoneurones.‘5 Recently, using behavioural measurements, we have examined some of the longer lasting-effects of peripheral tissue injury on the flexor reflex.3’ These experiments were performed in chronic decerebrate rats. Decerebrate rats continue to display intact spinal and brainstem reflexes although all their higher brain functions are absent.29 It is therefore possible to study the consequences of peripheral inflammatory lesions on spinal reflexes”~” without the ethical problems that would prevent such experiments being performed in intact animals. With intense peripheral lesions a hypersensitivity of the flexor reflex can be measured for several

C. .l. WOOLF and S. B. MCMAHON

396

weeks.” In the present experiments we have examined the neurophysiological mechanisms which might underlie these persistent changes in threshold and responsiveness of the reflex. The response properties of primary afferent C-fibres and flexor motoneurons have been recorded S-10 days after the creation of a thermal or chemical injury of a type sufficient to induce behavioural changes. We have thus been able to compare the long-term consequences of peripheral tissue injury on both the afferent and the efferent limbs of the flexor reflex arc. EXPERIMENTAL PROCEDURES Experiments were performed on 3.5 chronic decerebrate rats (20&300 g).

The animals were decerebrated under pentobarbitone anaesthesia by aspiration of all the cranial contents rostra1 to the mesencephalon. The techniques used for the decerebration, post-surgical management of the rats and iongterm care of the chronic decerebrate rat have all been previously described in detail.” When the condition of the decerebrate rats had become stable (24-48 h post-surgery) the mechanical threshold of the hindlimb-flexor-withdrawal reflex was monitored daily using mono~lament nylon Von Frey hairs. Nine of the decerebrate animals had thermal lesions made to a hindpaw on one side by heating the skin on the lateral edge of the paw at the glabrous/hairy skin line to 200°C for 3 s. This resulted in a zone of tissue necrosis at the site of heating surrounded by a zone of inflammation which appeared several hours later. By one week the thermal lesion consisted of a scar at the point of heating and a variable extent of red swollen skin surrounding it. In 16 other decerebrate rats the chemical irritant turpentine (0. I ml) was injected S.C.into the plantar surface of the hindpaw on one side. This resulted in an immediate inflammatory swelling of the foot which remained present for at least one week. The other IO decerebrate rats were not treated further and acted as controls for the decerebrate rats with peripheral inflammatory lesions. Ei~trophysioIo~cal experiments were performed on all 35 chronic decerebrate rats: at S-10 days following production of the inflammatory lesion for the experimental groups and at It&14 days post-decerebration for the control group. The animals were initially anaesthetized in ether, a tracheal cannula inserted and one carotid artery cannulated. Using intra-arterial doses of Althesin (alphaxalone/ alphadalone) a short-acting anaesthetic, the spinal cord was transected through a small lamin~tomy at TX-TIO. The anaesthetic was then discontinued. Two categories of electrophysiological experiments were then performed. In the first single primary afferent C-fibres were isolated from small filaments of the tibia1 or sural nerves by tine dissection in the popliteal fossa and their cutaneous mechanical response properties studied. In the second el~tromyograms (EMGs) from the posterior head of the biceps femoris muscle on each side were recorded by needle electrodes. Following the EMG recordings the animals were paralysed with gallamine and ventilated. The nerve to the posterior head of biceps and semitendinosus was then exposed in the popliteal fossa and small filaments dissected free so that activity in single flexor alphamotoneurones could be recorded.32 RESULTS

Behaviourai analysis

The chronic decerebrate rat displays an elaborate if stereotyped behavioural repertoire including pro-

nounced righting reflexes, spontaneous locomotion, primitive grooming and pseudoexploratory behaviour.29,3’ Stimulation of the hindpaw of these animals with moderately intense (non-tissuedamaging) mechanical, chemical or thermal stimuli results in a flexion-withdrawal reflex, vocalization and orientation of the body to the site of the stimulus. Prior to the electrophysiological experiments the behaviour of the rats was closely monitored. In the control rats without inflammatory lesions the hindlimb-flexor-withdrawal reflex was stable, consisting of a flicking movement in response to firm pressure or pinch. Following production of the thermal or chemical inflammatory lesions the withdrawal reflex changed in a way identical to that previously reported.3’ The mechanical threshold of reflexes on the ipsilateral hindpaw (Fig. 1) and on the dorsal surface of the hindpaw contralateral to the injury fell significantly and the responsiveness of the reflex changed so that instead of eliciting a flicking movement, suprathreshold stimuli now produced sustained flexion lasting for up to 60s. These changes manifested behaviourally as spontaneous flexion and flexion in response to light touch. Ekctromyogram

analysis

Electromyographic recordings were made from the posterior head of the biceps femoris on both sides in 23 chronic decerebrate rats once they had been spinahzed at the level of the lower thoracic spinal cord. The Von Frey thresholds necessary to elicit a reflex response in the posterior biceps femoris by mechanical stimulation of the plantar and dorsal surfaces of the hindpaw are shown in Fig. 1. In the spinalized preparation using the EMG the Von Frey thresholds in the control non-injured rats are lower than those found from the behavioural analysis of the same animals before they were spinalized. This reflects either the greater sensitivity of the EMG in detecting the reflex before overt movement can be seen or the removal of a tonic descending inhibition. The Von Frey mechanical thresholds in animals with chronic thermal injuries or tur~ntine-indu~d inflammations obtained from the EMG analysis were, like the behavioural measurements, considerably lower than untreated control animals (Fig. 1). The Von Frey thresholds of the hindlimb contralateral to the lesion were also significantly lower than the control levels but not to the same extent as the ipsilateral hindlimb [e.g. for the thermal injury, the threshold for the plantar surface of untreated rats was 47 + 9 g (n = 10) The same threshold ipsilateral to the injury was 6 &-I g (n = 9) and that contralateral 22 f 7g (n = 9)]. The EMG response of one rat with a thermai injury to graded mechanical stimulation

is illustrated

ipsilateral and contralateral

to the lesion

in Fig. 2. In the control decerebrate-spinal rats without a peripheral inflammatory lesion a brief (3 s) standard (15Og) pinch applied by a calibrated sprung clip to

398

C. J. WOOLFand S. B. MCMAHON

the middle three toes elicited a short lasting explosive burst of activity in the posterior biceps femoris muscle (Fig. 3A). In 5 of the 10 animals the burst of activity was followed by an afterdischarge lasting for a mean duration of 17 _t 9 s. If the standard pinch was sustained for 20 s the discharge from the muscle continued to decline rapidly so that for the final 60% of the stimulus no response was elicited (Fig. 3B). This is a typical example of a phasic flexion reflex. Heating the plantar surface of the foot of these rats to 50°C for 10s also elicited a flexion reflex which, while longer lasting than that evoked by sustained pinch, still adapted greatly during the maintained stimulus and only produced a period of afterdischarge lasting a few second (Fig. 3C). When identical stimuli were applied to the hindpaws of the rats with chronic (7-10 days) thermal or turpentine lesions the responses recorded from the posterior biceps femoris muscle ipsilateral to the lesion differed from that obtained in control rats. Brief (3 s) standard pinch stimuli to the ipsilateral hindpaw evoked a prolonged afterdischarge in 88% of the animals (Fig. 4A). which lasted for 61 + 14 s (n = 16). Such stimuli applied to the contralateral hindpaw resulted in afterdischarges in only 30% of the animals (5.8 + 3 s). Sustained pinch (20 s) of the middle 3 toes ipsilateral to the lesions produced a sustained discharge in the flexor muscle (Fig. 4B) which is quite unlike the phasic responses obtained in control animals (Fig. 3B). Similar tonic responses with long afterdischarges were evoked by heating the ipsilateral but not the contralateral hindpaw to 50°C for 10 s. No qualitative or quantitative differences in the pattern of EMG recordings obtained ipsilateral to either the thermal or the turpentine lesions were found. Electromyogram recordings were also made from biceps femoris muscles contralateral to the site of injury. The right hand panels of Figs 4A-C show that the responses in these muscles were essentially similar to those obtained from control uninjured tissue (cf. Fig. 3), i.e. the increased excitability following injury was not a general change but was restricted to flexor motoneurones appropriate to the tissue injury. Primary afferent analysis

Thirty-seven single cutaneous afferents with conduction velocities less than 1.2 m s-’ (Fig. 5) were studied in 4 untreated rats. Four of these units had relatively low mechanical thresholds (< 5 g). one had no cutaneous mechanoreceptive field but responded to a noxious thermal stimulus (>SO”C), while the remainder had thresholds typical of high-threshold C-mechanoreceptors ranging from 5-68 g. Table 1 shows the mean Von Frey threshold for this population. The vast majority of units responded to a pinch of their receptive fields with a fairly slowly adapting response which was followed in 5 units by a few seconds of afterdischarge and in 4 units with a prolonged afterdischarge (~30 s) (Fig. 5). Only 5

A

40’

O-

40r

Ob

! c

40r

O-

020

Time

(8)

Fig. 3. Ratemeter records of typical EMG responses obtained from the posterior biceps femoris muscle in two control decerebrate- animals to (A) a brief (3 s) standard (150 a) Dinch of the middle 3 toes. CB)a sustained (20 s) &nchid these toes and (C) heating the‘piantar surface df thk foot to 50°C for 10 s. The vertical scale represents the number of action potentials per SOOms bin.

units displayed some background activity which varied from 4 to 14 action potentials per minute. Thirty-three single C-primary afferent fibres were recorded, in 7 animals, that innervated skin which was obviously inflamed as a result of prior tur~ntine injection (1 to 7 days). The mean Von Frey threshold of these afferents did not differ significantly from those of the C-fibres which innervated normal skin, nor did the proportion of units showing spontaneous discharge or afterdischarge to pinch (see Table 1). We were unable to find any obvious feature of the mechanoreceptive field properties of these C-afferents that differed from that of the control C-afferents although there was a suggestion that some of the sural C-afferents had receptive fields extending more onto the glabrous skin than we had found in control

Functional plasticity of the flexor reflex

o---P

0

-_--

2

.., .! 1

._;

__.I

0 P

,-m’ _

,J 1’

_-

0

L--:

0

399

C. J. WOOLF and S.

400

‘59

MCMAHON

units had high-threshold mechanosensitive receptive fields and brief afterdischarges were seen in only 2 units. In making single-unit recordings from these fine-diameter afferents, filaments containing several hundred large A/3 fibres were also examined. The presence of chronic tissue injury had no obvious effect on the receptive field properties of the vast majority of these fibres. However, a handful of afferents found only in the experimental animals showed a very regular, relatively high level (lo-20 Hz) of background firing occurring intermittently but for prolonged periods. This type of discharge pattern is reminiscent of that arising from damaged, regrowing axons in neuromas. These results indicated that this form of chronic injury does not appear to induce any long-lasting, gross alteration in the properies of cutaneous high-threshold primary afferents.

I

109

B.

68g

Single

5s

cc-motoneuron

analysis

The mechanoreceptive field properties of 40 singlefiexor a-motoneurones innervating that posterior biceps femoris/semitendinosus muscles were recorded in control decerebrate-spinal rats, 32 were recorded ipsilateral to either a thermal or turpentine lesion and 22 contralateral to the lesions. The results obtained closely resembled those found from the EMGanalysis with very similar Von Frey thresholds. An

PINCH

ipsilateral

contralateral

25

1

5s

Fig. 5. A single high mechanothreshold tibia1 C-afferent fibre. The top trace shows the activation of the unit following electrical stimulation of its receptive field in the skin with pin electrodes (conduction distance 49 mm). The middle traces show the response of this unit to Von Frey hairs of increasing force. The bottom trace shows the after discharge developing after a 3-s pinch to the receptive field.

O-

JL

25-

effort was devoted to recording from single C-afferents but in the course of this, we recorded from 20 A6 afferent fibres. Again there were no apparent differences between units recorded in animals with or without tissue injuries: 16 of the 20 animals.

Our

major

Table 1. The properties of single cutaneous C-afferent fibres recorded from control untreated rats and those which innervated skin with a turpentine-induced inflammation Control Mechanothresholds (g) y0 Units with spontaneous activity % Units showing an afterdischarge to mechanical stimulation Number of units

28 + 5 17

Inflamed skin 31 +8 23

24

28

37

33

o-

1

-IL 010 Time

(5)

Fig. 6. Ratemeter recordings of single responses made from 4 single hamstring flexor a-motoneurones showing the response to a brief (3 s) standard (I 50 g) pinch to the middle 3 toes. The 2 units on the left were ipsilateral to a chronic thermal lesion while the 2 units on the right were recorded contralateral to a chronic thermal injury. The vertical scale represents the number of action potentials per 200 ms bin.

401

Functional plasticity of the flexor reflex

afterdischarge to brief (3 s) standard pinches was only observed in those neurones ipsilateral to the injured hindpaw (44%) and not at all in the motoneurones in control decerebrate-spinal rats or in those recorded contralateral to the lesion. Responses to pinch in units ipsilateral and contralateral to a thermal injury are shown in Fig. 6. Following a sustained (20s) pinch 30% of the units ipsilateral to either type of lesion displayed a slowly adapting response, while only 8% of the units in the control animals and only 17% of the units recorded contralateral to the lesion had such tonic responses. The intention of studying the single motor units in this series of experiments was to try and elucidate to what extent the changed properties of the motoneurones ipsilateral to the thermal injury were the result of changes in the primary afferents innervating the inflamed area in a way different from the

recording of single afferents as described above. The approach used for this was the production of a complete sensory block at the site of the injury and surrounding tissue with a local anaesthetic. Changes in the excitability of the flexor reflex could then be measured by looking at the responses evoked by stimulation of the sural nerve at a strength that will activate C-fibres.26,30 A post-stimulus histogram of the responses evoked in a flexor alpha-motoneurone in a control decerebrate-spinal animal by stimulation of the sural nerve is illustrated in Fig. 7(A). A local sensory block with 0.1 ml of 2% lignocaine injected into the ipsilateral hindpaw failed to modify the sural-evoked response indicating that the stimulus-response relations in this preparation are independent of any ongoing afferent input. A similar post-stimulus time histogram is shown in Fig. 7(B), except that the 5

-00 Time

(msec)

Fig. 7. Post-stimulus time histograms (8 sweeps, 1 s duration) of the activity evoked in 3 single hamstring flexor motomeurones by stimulation of the sural nerve at a strength that activates C-fibres (5 mA, 500 pts, 0.5 Hz). The extreme left series of post-stimulus time histograms were made IO min before, the middle 5 min after and the right 30 min after the injection of 5 mg/kg lignocaine. In (A) the lignocaine was injected into a non-inflamed hindpaw. In (B) the same dose of lignocaine injected into an inflamed hindpaw suppressed the C-evoked activity, while in (C) the same dose of lignocaine injected i.p. also considerably reduced the sural-evoked activity. (Bin width 4ms.)

402

C. J.

WOOLF and

response is recorded in a motoneurone ipsilateral to a chronic thermal lesion. Injection of 0.1 ml of lignoCaine into the vicinity of the lesion and surrounding tissue resulted here in a marked diminution of the sural-evoked responses. Similar results were obtained in 6 of 8 animals with thermal lesions and in the 2 rats tested which had chemical inflammatory lesions. These results seem to imply that removal of afferent input generated by the inflamed tissue decreases the excitability of the spinal cord. To test against the possibility that the lignocaine was exerting an effect other than a local conduction block the following experiments were performed. Sural-evoked responses were recorded in a motoneurone contralateral to a chemical inflamed hindpaw. Lignocaine (0.1 ml) was injected into the uninflamed hindpaw, this had no effect on the suralevoked response. However, when the same quantity of lignocaine was injected into the inflamed hindpaw contralateral to the recording the sural-evoked responses were suppressed. This may mean either that a contralateral afferent barrage from inflamed tissue can modify the sural stimulus-response relations or that the lignocaine is taken up into the systemic circulation when injected into inflamed tissue. To test the latter possibility the same dose of lignocaine was injected intraperintoneally (Fig. 7C). This too reduced the sural-evoked response in flexor motoneurones. It was then found that i.v. injections at doses as low as 1 mg/kg suppressed sural-evoked polysynaptic reflexes. The dose of lignocaine injected locally into the inflamed tissue varied from 5 to lOmg/kg. Clearly it is not possible to produce a sensory block of inflamed tissue without generating a systemic effect as well.

DISCUSSION

Functional plasticity in the nervous stystem That the nervous system possesses the capacity to modify its functional performance in the face of changes in its external or internal environment has long been recognized. This plasticity is most obvious during development and following lesions to the peripheral or central nervous system where structural changes usually occur. A more subtle and different form of plasticity occurs in the mature central nervous system. This is a functional plasticity representing a highly dynamic, usually reversible adaptive response to an alteration in input. Examples of this are the homosynaptic facilitation or depression of synapses following tetanic input”.” and the changes in functional organization in the visual,2’ auditory” and somatosensory’” cortices. The spinal cord has been extremely useful for studying lesion-induced plasticity. Dorsal root secperipheral nerve section’ and spinal tion,” transection” all produce marked alterations in the organization of the spinal cord. The extent to which

S. B. MCMAHON

such changes are due to the sprouting of axons into vacant synaptic sitesI or to the unmasking of ineffective synapses24 in the different circumstances is still not fully resolved. What is clear though is that the spinal cord in the adult does not consist of rigidly wired networks and that there is much compensatory plasticity in response to deafferentation.22.25.27 The spinal cord, however, also shows prolonged changes in function in response to interventions other than lesions. Disuse of the Ia afferent fibre pathway following tetrodotoxin treatment for example, results in a substantial elevation in the size of the excitatory postsynaptic potentials evoked by those fibres while changes in the amplitude of the monosynaptic stretch reflex can be seen in the monkey if such changes are rewarded.2x In our laboratory we have found that dorsal horn cells can show expanded receptive fields and lower thresholds after peripheral injuryI and that the flexor withdrawal reflex increases its excitability both following acute tissue injury’(‘and brief conditioning stimuli to primary afferent C-fibres.‘” Because these changes occur in spinalized preparations, are heterosynaptic and outlast the period of stimulation they must reflect a prolonged alteration within the spinal cord. At the time that flexor motoneurones show a post-conditioning increase in excitability there is no change in the amplitude of their monosynaptic reflex.” The site of the acute C-fibreinduced excitability increase must lie somewhere other than in the motoneurone. Injury -induced plasticity in the spinal cord The present series of experiments show that following severe peripheral inflammatory lesions in the hindpaw of the chronic decerebrate rat there are major modifications in the functional performance of hamstring flexor reflex. In many respects these changes, a fall in threshold, an increase in spontaneous activity and the development of slowly adapting responses and prolonged afterdischarges represent the mechanisms underlying the behavioural hypersensitivity previously observed.” These changes resemble the alterations found after acute injury3’ although the slowly adapting tonic responses to sustained stimuli and the duration of the afterdischarges are much more exaggerated following chronic than acute inflammatory lesions. Essentially, the high-threshold phasic flexor withdrawal response has become a low-threshold tonic one. This represents a remarkable degree of functional plasticity of the system. Because injections of lignocaine into the inflamed hindpaw resulted in a systemic uptake of the drug which then presumably acted centrally to suppress C-evoked activity, 4 it was not possible from that approach to establish the extent, if any, of the contribution a constant afferent barrage from the site of the injury might make to the changes observed in the function of the flexor I-motoneurones. However, the failure to observe any differences in the response

Functional plasticity of the flexor reflex properties of the mechanoreceptive C-fibres innervating normal or chronically inflamed skin is highly suggestive that the altered motor output of the spinal cord is not just a reflection of an abnormal input. Repeated moderate thermal stimuli have been found to produce both sensitization and desensitization of nociceptors’,*.” and acutely, a severe thermal stimulus also produces a mixture of such increases and decreases in the thresholds of cutaneous C-fibres.15 When looked at in terms of overall population changes, however, both moderate and severe thermal stimuli fail to produce any significant change in rat C-fibre response properties.6,‘5 Acute thermal injury does, however, increase the excitability of the flexor reflex.30 Therefore for both acute and chronic injury-induced alterations in the flexor reflex it is unlikely that the increased excitability of the reflex requires an ongoing afferent input. Somehow the peripheral injury triggers a prolonged alteration in the spinal cord. In the case of acute effects this could be mediated by a neuromodulator released by afferents activated by the tissue-damaging stimulus which then sensitizes the spinal cord. On the basis of our experiments using conditioning stimuli to peripheral nerves we believe the C-afferents are most likely to trigger the altered excitability. 26It is possible that neuropeptides such as substance P, which produces prolonged excitability changes in dorsal horn cells23 may be responsible. The neuromodulator is then likely to act on a second messenger system in some neurones such as calcium, adenosine 3’: S-phosphate or guanosine

403

3’:5’-phosphate, which then alters the activity of intracellular enzymes like the protein kinases which by phosphorylating membrane-bound proteins can modify membrane excitability.*’ The mechanisms responsible for the alterations in the stimulus-response relationships of the spinal cord following chronic peripheral injuries are not known. These could be either repeated triggerings of the acute excitability effects or another mechanism altogether involving for example active transport of signals from the injured area to the spinal cord. What we do not know yet is whether a prolonged abnormal input can produce prolonged or even permanent changes in the excitability of the spinal cord. If this were so it might mean that chronic pain results more from changes within the central nervous system than to changes in the input. Concluding remarks

Teleologically the increased excitability of the spinal cord following acute peripheral injury can be considered to represent an adaptive functional plasticity of the system minimizing further injury and assisting the healing process. The persistence or exaggeration of this plasticity following chronic injury may be a maladaptive or pathological response leading to chronic pain, an example perhaps of dysfunctional plasticity. Acknowledgements-We wish to thank Professor for his useful advice and comments and the Wellcome Trust for financial support.

P. D. Wall

MRC and

REFERENCES 1. Beck P. W., Handwerker

2.

3. 4. 5. 6. 7. 8. 9. 10. 1 I. 12. 13. 14. 15. 16.

H. 0. and Zimmermann M. (1974) Nervous outflow from the cats foot during noxious radiant heat stimulation. Bruin Res. 67, 373-386. Bessou P. and Per1 E. R. (1969) Response of cutaneous sensory units with unmyelinated axons to noxious stimuli. J. Neurophysiol. 32, 1025- 1043. Cook A. J. and Woolf C. J. (1985) C-afferent fibre induced excitability increase of the flexor reflex is not due to change in motoneurone excitability. Neurosci. Left. Suppl. 21, 586. De Jong R. H., Robles R. and Corbin R. W. (1969) Central action of lidocaine-synaptic transmission. Anesthesiology 30, 19-23. Devor M. and Wall P. D. (1981) Plasticity in the spinal cord sensory map following peripheral nerve injury in rats. J. Neurosci. 1, 679-684. Fleischer E., Handwerker H. 0. and Joukhader (1983) Unmyelinated nociceptive afferents in two skin areas of the rat. Brain Res. 267, 81-92. Gallego R., Kuno M., Nunez R. and Snider W. D. (1979) Disuse enhances synaptic efficacy in spinal motoneurone. J. Physiol., Lond. 291, 191-205. Goldberger M. E. and Murray M. (1974) Restitution of function and collateral sprouting in the cat spinal cord: the deafferented animal. J. camp. Neural. 158, 37-54. Hardy J. D., Wolff H. G. and Goode11 H. (1952) Pain Sensations and Reacfions. Williams and Wilkins, Baltimore. Kraus W. and Disterhoff J. F. (1982) Response plasticity of single neurons in rabbit auditory association cortex during tone-signalled learning. Bruin Res. 246, 205-215. Lewis T. (1942) Pain. Macmillan, London. Liu C. N. and Chambers W. W. (1958) Intraspinal sprouting of dorsal root axons. Arch. Neural. Psych&r. 79,4&61. Lynn B. and Carpenter S. E. (1982) Primary afferent units from the hairy skin of the rat hindlimb. Brain Res. 238, 29-43. McMahon S. B. and Wall P. D. (1984) Receptive fields of rat lamina I projection cells move to incorporate a region of nearby injury. Pain 19, 235-247. McMahon S. B. and Woolf C. J. (1984) The effects of peripheral injury on the mechanosensitivity of rat unmyelinated cutaneous afferents. J. Physiol., Land. 353, 30P. McNaughton B. L. (1982) Long-term synaptic enhancement and short-term potentiation in rat fascia dentata act through different mechanisms. J. Physiol., Lond. 324, 249-262.

C. J.

404

WOOLF

and S. B. MCMAHON

17. Mendell L. M. (1984) Modifiability of spinal synapses. f’hysiol. Reo. 64, 260-324. 18. Merzenich M. M., Nelson R. J., Stryker M. P., Cyanader M. S., Schoppman A. and Zook J. A. (1984) Somatosensory cortical map changes following digit amputation in adult monkeys. J. camp. Neural. 224, 591-605. 19. Nelson S. G. and Mendell L. M. (1979) Enhancement in Ia-motoneurone synaptic transmission caudal to chronic spinal cord transection. J. Neurophysiol. 42, 642-654. 20. Nestler E. J. and Greengard P. (1983) Protein phosphorylation in brain. Nature 305, 583-588. 21. Singer W., Tretter F. and Yinon U. (1982) Evidence for long-term functional plasticity in the visual cortex of adult cats. J. Physiol., Lond. 324, 239-248. 22. Sugimoto T. and Gobel S. (1984) Dendritic changes in the spinal dorsal horn following transection of a peripheral nerve. Bruin Res. 321, 199-208. 23. Urban L. and Randic M. (1984) Slow excitatory transmission in rat dorsal horn: possible mediation by peptides. Bruin Res. 290, 336342. 24. Wall P. D. (1977) The presence of ineffective synapses and the circumstances which unmask them. Phil. Trans. R. Sot. Ser. B. 278, 361-372.

25. Wall P. D. and Devor M. (1981) The effect of peripheral nerve injury on dorsal root potentials and on transmission of afferent signals into the spinal cord. Bruin Res. 209, 95-l 11. 26. Wall P. D. and Woolf C. J. (1984) Muscle but not cutaneous C-afferent input produces prolonged increases in the excitability of the flexion reflex in the rat. J. Physiol., Lond. 356, 443458. 27. Wall P. D., Fitzgerald M. and Woolf C. J. (1982) Effects of capsaicin on receptive fields and on inhibitions in rat spinal cord. Expl Neural. 78, 425-436. 28. Wolpaw J. R., Noonan P. A. and O’Keefe J. A. (1984) Adaptive plasticity and diurnal rhythm in the primate spinal stretch reflex are independent phenomena. Brain Res. 300, 385-391. 29. Woods J. W. (1964) Behaviour of chronic decerebrate rats. J. Neurophysiol. 27, 635-644. 30. Woolf C. J. (1983) Evidence for a central component of post-injury pain hypersensitivity. Nature 306, 686688. 31. Woolf C. J. (1984) Long term alterations in the excitability of the flexion reflex produced by peripheral tissue injury in the chronic decerebrate rat. Pain 18, 325-343. 32. Woolf C. J. and Swett J. E. (1984) The cutaneous contribution to the hamstring flexor reflex in the rat: an electrophysiological and anatomical study. Bruin Res. 303, 299-3 12. (Accepfed

10 May 1985)