Pain, 51 (1992) 335-342 0 1992 Elsevier Science
335 Publishers
B.V. All rights reserved
0304-3959/92/$05.00
PAIN 02163
Selective changes of receptive field properties of spinal nociceptive neurones induced by noxious visceral stimulation in the cat Fernando Cervero, Jennifer M.A. Laird ’ and Miguel A. Pozo 2 Department of Physiology. University of Bristol Medical School, UniLlersityWalk, Bristol BS8 I TD (UK) (Received
21 April 1992, revision received
16 June 1992; accepted
22 June 1992)
This study was designed to examine the central changes in the receptive field properties of dorsal Summary horn neurones induced by a period of visceral noxious stimulation. The aim of this investigation was to establish whether noxious stimulation of the visceral input to the spinal cord could influence transmission of cutaneous information through dorsal horn neurones. Single-unit electrical activity was recorded in the lower thoracic spinal cord of anaesthetized cats from dorsal horn neurones with a somatic receptive field in the ipsilateral flank. Changes in the properties of these receptive fields induced by reversible spinalization (by means of a cold block 4 or 5 segments rostra1 to the recording electrodes) and by a conditioning noxious stimulation of the biliary system (3 successive distensions of the gall bladder for 30 set at 65-80 mm Hg at 1-min intervals) were analysed. Nineteen neurones have been studied, 10 of which could be driven by stimulation of the gall bladder. All of these 10 cells showed increases in the size of their cutaneous receptive fields following conditioning noxious stimulation of the biliary system. The increases were large and lasted for at least 20 min. None of the 9 spinal cord neurones without an input from the gall bladder were affected by the conditioning visceral stimulus even though 7 showed changes in receptive field size when the animals were spinalised. These results show that noxious stimulation of viscera can evoke increases in the somatic receptive fields of spinal cord neurones but only of those neurones which are also driven by the visceral stimulus. Key words: Dorsal horn; Spinal cord; Visceral pain; Receptive fields; Plasticity; Biliary system
Introduction
There is currently considerable interest in the peripheral and central mechanisms that underlie hyperalgesic states. It has been known for some time that an injury to the skin and somatic tissues results in sensitization of peripheral nociceptors and that this mecha-
’ Present address: Neuroscience Research Centre, Laboratories, Terlings Park, Harlow, UK. 2 Present address: Instituto cante, Alicante, Spain.
de Neurociencias,
Merck
Universidad
Research
de Ali-
Correspondence to: Dr. F. Cervero, Dept. of Physiology, University of Bristol, Medical School, University Walk, Bristol BS8 1TD. UK.
nism contributes to the hyperalgesia evoked by low-intensity stimuli applied to the injured site (Raja et al. 1984; Meyer et al. 1985). However, attention has also been focussed on the central changes that result from the arrival of a nociceptive barrage to the spinal cord (see refs. in Dubner 1992). Much of the information about these central alterations has been obtained by examining changes in the receptive field properties of dorsal horn neurones. These changes are interpreted as an expression of the CNS reactivity to a peripheral injury and as the underlying mechanism for the increased central excitability that characterises hyperalgesic states (Woolf 1989). One of the first reports of changes in the properties of dorsal horn neurones following a brief period of noxious stimulation originated from this laboratory (Cervero et al. 1984). In this study it was shown that a train of 20 stimuli at 1 Hz applied to the sural nerve at
336
an intensity supramaximal for C fibres, evoked an increase in the size of the receptive field of dorsal horn neurones driven by sural afferents. Other reports have described similar changes in the receptive field size of dorsal horn neurones and in the magnitude of flexor reflexes following similar periods of electrical stimulation of a peripheral nerve or after the application of a variety of noxious stimuli to, or close to, the receptive field of the cells (McMahon and Wall 1984; Cook et al. 1987; Schaible et al. 1987; Hoheisel and Mense 1989; Hylden et al. 1989; Woolf 1989; Laird and Cervero 1989, 1990). Virtually all the reports quoted above examined only the responsiveness of dorsal horn cells to somatic noxious stimuli. Comparatively little is known about how dorsal horn cells react to prolonged noxious stimulation of internal organs. Yet, this information is of great clinical relevance for two reasons. Firstly, hyperalgesia referred to somatic and visceral structures is a frequent sequel of visceral pain states. Conditions such as irritable bowel syndrome are characterised by an increased sensitivity to physiological gastrointestinal stimuli, so that normal gut activity becomes painful (Thompson 1991). Secondly, it is a common clinical experience that the area of referral of visceral pain can change in size depending on the intensity of the pain (MacKenzie 1909; Morley 1931). For instance colic pain referred to the surface of the abdomen produces large areas of superficial tenderness at the peaks of the colic attacks and smaller areas during periods of quiescence. The present study was designed to examine the central changes in the receptive field properties of dorsal horn neurones induced by a period of visceral noxious stimulation. We have recorded from neurones in the lower thoracic spinal cord of anaesthetized cats and have looked at changes in the sizes of their cutaneous receptive fields following prolonged noxious stimulation of the gall bladder and biliary system. One of our aims was to establish whether noxious stimulation of the visceral input to the spinal cord could influence the transmission of cutaneous information through dorsal horn neurones. Since the conditioning stimulus used in this study was restricted to internal organs, our approach also eliminated ambiguities as to the peripheral or central origin of any observed effect on the somatic receptive fields of the neurones. Some of the results presented here have been published in abstract form (Cervero et al. 1992).
Cold
Fig.
1. Schematic
diagram
spinal cord stimulation
block
of the
experimental
arrangement.
A:
of the ItJth thoracic dorsal root, recording
the 9th thoracic segment and cold block. B: mechanical of the biliary system; Tr = pressure
transducer.
in
stimulation
For further
details
see text.
curonium
bromide
tive-pressure
(0.06 mg/kg/h,
iv.1 and ventilated
with a posi-
pump. The general methods for monitoring
and main-
tenance of the physiological state of the animals have been described previously in detail (Cervero The
lower
thoracic
from T9 to Tl2. TS-T6
1983a, bJ.
spinal cord was exposed
A second smaller
by a laminectomy
laminectomy
to allow reversible spinalization
was performed
means of a cold block (see below). The animals were mounted rigid frame
at
of the animals at this level by in a
and pools were made with skin flaps over the exposed
areas of the spinal cord. Recording ing the vertebral
stability was improved
column, by infiltration
cord and by a bilateral
pneumothorax.
filled with warm paraffin
by clamp-
of 3% agar around the spinal All spinal cord pools were
oil at 38°C.
Recording techniques Extracellular croelectrodes
was X-20 Mf2J. the
right
searching
in T10
measured
at I KHz
for
through
neurones.
of the
segments
T9-TlO.
or Tl I was electrically ball-tipped
Recordings
oscilloscope
and analysed ‘on-line’
puter.
IA
Fig.
(impedance
grey matter
dorsal rootlet
V, 0.1 msec, I Hz)
when
recordings were made through glass m-
Recordings were made from dorsal horn neurones in
side of the
ipsilateral (I-3
single-unit
filled with 4 M NaCl
silver wire electrodes were
and ‘off-line‘
shows a schematic
An
stimulated
diagram
of
displayed
on an
using a microcomthe
experimental
arrangements.
Stimulation of somatic receptille fields Somatic
afferent
their receptive
fibres were activated
by natural
stimulation
of
fields. Stimuli included both innocuous (i.e., brushing
and stroking) as well as noxious stimuli (i.e., pinching and squeezing). The
boundaries
neurones
were
of the somatic mapped
receptive
fields of the dorsal
for both low-threshold
responses and marked on the flank of the animals with felt-tip of different receptive paper
colours. At the end of the study of each neurone,
field boundaries
were transferred
and the areas measured
horn
and high-threshold
to transparent
using a digitising tablet
pens the
drawing
and micro
computer.
Methods
Stimulation of Lisceral afferent fibres The conditioning
Experiments
body
of distensions applied
weights between 2.8 and 3.9 kg. All animals were anaesthetized
with
tion of the gall bladder
halothane;
conducted
(60-70
30 mg/kg.
mg/kg,
on Y adult i.v.;
i.v., maintenance
initial
male
noxious stimulus used in this study was a series
cats with
alpha-chloralose
were
dose
under
doses), paralysed
2.5”i
with pan-
previously in Fig.
described
to the biliary system of the animals. Stimulaand biliary ducts was achieved by a method
in detail (Cervero
1B. It involved
the cannulation
1982. 1983aJ and illustrated of the fundus
of the gall
337 bladder with a double-barrelled catheter. One barrel was connected to a pressure transducer and the other to a reservoir containing warmed saline solution. The common bile duct was ligated so that pressure changes could be imposed on the biliary system by raising the reservoir to various heights above the preparation.
25 20 15 Imp/set 10
Reversible spinalization
5
This was achieved by circulating cold fluid through a silver-plated thermode placed over the cord at T5-T6 level. Details of this technique have been published previously as well as tests of the efficacy of this method in providing a reversible spinalization (Cervero and Plenderleith 1985; Laird and Cervero 1990).
0
Experimental protocol The experimental protocol applied to each identified neurone consisted of the following sequential steps. (8 A brief, noxious distension of the gall bladder (5 XC, 50-70 mm Hg) was performed to ascertain if the cell responded to this stimulus. (ii) The somatic receptive field was mapped on the flank of the animal and marked as described above. This was recorded as the initial receptive field of the cell. (iii) A reversible spinalization of the animal was then performed and the receptive field was re-mapped in the spinal state. This was recorded as the spinal receptive field of the neurone. (iv) A brief, noxious distension of the gall bladder (parameters as above) was then applied to check for responsiveness of the cell to this stimulus in the spinal state. (v) The cord was rewarmed and the receptive field was again mapped to ensure that, if any changes had been observed in the spinal state, it had returned to the initial receptive field size. (vi) A conditioning noxious stimulation of the biliary system was then applied to the animal. This consisted of a series of 3 stimuli delivered at I-min intervals. Each stimulus was a pressure pulse of 65-80 mm Hg lasting for 30 sec. This conditioning stimulus was designed to mimic the pressure changes occurring during a series of colic contractures of the gall bladder. (vii) After the administration of the visceral conditioning stimulus the somatic receptive field was again mapped at 5-min intervals. This was recorded as the receptive field after gall bladder stimulation. Only a few neurones were studied in each animal to avoid cumulative effects of the conditioning stimulus. At least 1 h was allowed to elapse between the study of 2 successive neurones.
Histological methods The position of the recording microelectrode was marked by ionophoretic deposition of pontamine sky blue in the last track of each experiment. Marks were made in this track at l-mm intervals. The section of the cord was removed and frozen. The recording sites of the neurones were calculated from these spots recovered in 50-pm sections counterstained with haematoxylin-eosin. Details of this technique for the location of recording sites have been previously published (Cetvero 1983a).
Results Sample of new-ones
The results presented in this paper are based on a sample of 19 neurones recorded in the lower thoracic spinal cord. The search stimulus was electrical stimulation of a dorsal rootlet entering the same or an immediately caudal segment of the cord at an intensity supramaximal for all A fibres in the rootlet. Neurones included in the present sample were those responding to this stimulation and having a cutaneous receptive
0
50
100 Time
GE
Press.
,G
i
150
200
(sacs)
~
(mmHg) ,3 L M 1 Fig. 2. Response of a GB+ neurone in the thoracic spinal cord to repeated distensions of the biliary system. Top: ratemeter output of the spike discharge of the neurone. Bottom: biliary pressure.
field on the ipsilateral flank from which they could be driven by noxious mechanical stimulation. Some neurones showed, in addition, responses to innocuous mechanical stimulation of the receptive field (see below). Recordings judged to be made from axons rather than from cell bodies were discarded. Responses to gall bladder distension
All 19 neurones were challenged with distensions of the gall bladder at noxious intensities (see Methods). The sample of cells was divided into 2 groups depending on whether the neurones responded to this stimulus. Ten of the neurones responded to these distensions and were thus classified as having an input from the gall bladder (GB + >. Fig. 2 shows the responses of one such neurone to the visceral-conditioning stimulus. Characteristically the cells of this group responded to the increase in biliary pressure with an initial phasic discharge which adapted to a lower level of tonic response throughout the duration of the stimulus. The levels of background activity of this group of cells increased progressively in between successive stimuli (see Fig. 21. The other 9 cells did not respond to any of the distensions of the biliary system and were thus classified as lacking an input from the gall bladder (GB- ). All the cells in this group had been tested for gall bladder inputs while the animals were reversibly spinalised to ensure that descending inhibition was not masking a gall bladder input to the cells. None responded to biliary distension in the spinalised state. Somatic receptive fields
The neurones were also classified according to their responses to natural stimulation of their cutaneous receptive fields. Fifteen neurones responded to both innocuous and noxious stimulation and were thus classified as class 2 (or wide-dynamic-range). Four cells responded exclusively to noxious stimulation of their
338
jGB+i
GB-
Class
2
lomino
m
Fig. 3. Locations of the recording superimposed
and GB-
have also been divided
tive field
properties
nil?
RF
neurones
acdording to their somatic recep-
(class 2 and class 3). The
in the laminar
maps
is shown
grey matter
in the
central
area
diagram
(dotted square).
Fig. 5. Increase neurone
in the size of the somatic receptive
following
change after repeated
receptive fields and were classified as class 3 (nociceptive-specific). As shown in Fig. 3, the sample was equally divided when taking into account the presence or absence of an input from the gall bladder and the properties of the cutaneous receptive fields. The GB+ group comprised 8 class-2 and 2 class-3 cells and the GB- group included 7 class-2 and 2 class-3 neurones. Location of neuroiws
All neurones were recorded within the ipsilateral grey matter, mainly in the dorsal horn (Fig. 3). There was a similar distribution of superficial and deep neurones between the GB+ and GB- groups. Most cells were recorded in the deep dorsal horn, in or around lamina V.
Class 2
lamina
V
GE+
Fig. 4. Increases in the size of the somatic receptive neurone following reversible spinalization bladder
GB-
on standard sections of the lower thoracic spinal cord.
Neurones magnified
sites of GB+
initial
V
(bottom)
field of a GB+ and repeated
gall
distension (top). The initial receptive field is shown in black.
The location of the recording site of the neurone left corner of the figure.
is shown in the top
reversible
spinalization
field of a GB-
(bottom)
and
lack
of
gall bladder distension (top). The initial recep-
tive field is shown in black. The location of the recording
site of the
neurone is shown in the top left corner of the figure.
Receptille field changes after spinalization and after noxious stimulation of the biliary system
The receptive fields of all cells were examined before and after reversible spinalization of the animals and before and after the application of the conditioning visceral stimulus. Quantitative measurements of the receptive field areas were obtained from 8 GB+ and 8 GB- neurones, Two types of response were observed depending on whether the cells responded to gall bladder stimulation. (i) GB+ neurunes. All but one of the neurones having an input from the gall bladder showed increases in the size of their cutaneous receptive fields while in the spinalized state (Figs. 4 and 6). The receptive field of the remaining neurone decreased in size in the spinal state. After rewarming the spinal cord, the receptive fields of all neurones returned to their initial size. All GB+ cells showed increases in the sizes of their receptive fields after the conditioning visceral stimulation (Figs. 4 and 6). The increases lasted for at least 20 min and, in some cases, the receptive fields did not return to their initiai sizes until 40-45 min after stimulation. The increases were quite large (range: 20-136% larger than the original size) but were generally smaller than the increases induced by spinalization (range: 50-339s larger than the original size) (see Figs. 4 and 6). Increases in receptive field size affected both the low- and high-threshold areas of the receptive fields of class-2 cells. However, the quantitative data presented in Fig. 6 apply only to the high-threshold areas of class-2 cells so that a comparison can be established with the receptive fields of class-3 cehs. The only cell whose receptive field decreased in size in the spinalized state was a class-3 cell.
339 Cutaneous
Receptive
Field
Changes
Fig. 6. Changes in the size of the somatic receptive fields of GB+ and GB- neurones following reversible spinalization and repeated gall bladder distensions. Neurones have also been divided according to their somatic receptive field properties (class 2 and class 3). Note that none of the GB- neurones changed their somatic receptive fields after gall bladder stimulation even though most of them had larger receptive fields in the spinal state.
(ii) GB- neurones. The receptive fields of all but 2 of the neurones that did not have an input from the gall bladder increased in size after reversible spinalization (range: 122-373% larger than the original size) (Figs. 5 and 6). After rewarming the spinal cord, the receptive fields of these neurones returned to their initial size. The other 2 cells did not change the size of their receptive fields in the spinalised state. In contrast, none of the cells in the GB- group showed changes in the size of their receptive fields following conditioning visceral stimulation (Figs. 5 and 6). Thus, even though the receptive fields of many neurones in this group could increase in size after spinalization, they remained unchanged after the conditioning noxious visceral stimulation.
Discussion The results presented in this paper show that the somatic receptive fields of spinal cord neurones can change in size following a period of noxious visceral stimulation. This observation indicates that noxious stimuli applied to internal organs induce alterations in the spinal processing of somatosensory information. Since the conditioning noxious stimulus was applied to visceral afferent fibres and the changes observed were expressed on the cutaneous receptive fields of the neurones it is fairly certain that these alterations in excitability took place centrally rather than peripherally. In addition, our results also indicate a high degree of selectivity in the expression of these central changes depending on whether or not the spinal cord cells are
driven by the visceral fibres activated by the conditioning stimulus. This study extends current observations on the functional plasticity of spinal neurones to those driven by visceral afferent fibres. We have studied cells in the lower thoracic spinal cord driven by noxious stimulation of the gall bladder, a group of neurones which are putative candidates for the signalling of pain from the biliary system (Cervero and Tattersall 1985). The conditioning stimulus used was designed to reproduce the pattern of a biliary colic attack and to maximise the afferent barrage triggered on high threshold biliary afferents (Cervero 1982). However, this kind of stimulus is insufficient to produce long-lasting inflammatory or traumatic changes in the biliary system and therefore the alterations observed in the present study should be interpreted as those likely to underlie acute, rather than chronic, visceral pain states. Peripheral versus central effects
There is some debate as to whether peripheral sensitization of nociceptors or central sensitization of spinal neurones play a major role in the triggering of hyperalgesic states. The present study clearly shows that the afferent barrage triggered by the visceral stimulus induced changes that were predominantly central. For the changes in the cutaneous receptive fields of the neurones to have had a peripheral origin it would be necessary to postulate that the noxious stimulation of the biliary system could have altered the responsiveness of cutaneous receptors. While we cannot discard this possibility with our data, this seems unlikely. There are also some descriptions in the literature of afferent
340
fibres with branches in somatic and in visceral nerves (i.e., see Dawson et al. 1992). These fibres are extremely rare and, in the absence of data as to whether or not they are connected to active receptor sites in the skin and viscera, their functional significance remains obscure. Our interpretation of the observed changes in receptive field size is that the barrage evoked in visceral afferent fibres by the noxious stimulus removed central inhibition of spinal neurones. This increase in excitability meant that the cells could now be activated by input from what was previously within their subliminal fringes (Zieglgansberger and Herz 1971; Woolf and King 1989). Therefore, regardless of whether the visceral nociceptors were sensitized, the change in the somatic receptive field of the neurones was clearly a central phenomenon. The long duration of these effects also suggests that the central changes were not maintained by a continuous barrage from the stimulated visceral afferents. ~gh-threshold nociceptors in the gall bladder and biliary ducts have very low levels of background activity and tend to desensitize on repetitive stimulation (Cervero 1982). Therefore, it seems that although the initial barrage triggered by the conditioning stimulus was the cause of the central changes, a persistent drive from the periphery was not necessary for the maintenance of these effects. It is more likely that the positive feed-back loops between brain-stem nuclei and viscerosomatic neurones in the spinal cord that we have described (Cervero 1988) played a more important role in sustaining central excitability. Effects of visceral stimulation
One of the key observations of the present study is the high degree of selectivity observed between different spinal neurones in the expression of excitability changes. Our results show that all neurones driven by gall bladder afferents increased their somatic receptive fields after visceral stimulation. In contrast, none of the cells without a gall bladder input did so. These data suggest that an excitatory input from the viscus that undergoes noxious stimulation is necessary for the cell to express changes in its somatic receptive field. We cannot, however, extend this conclusion to suggest that a noxious visceral stimulus would influence all cells driven by visceral afferents, including those not directly activated by the conditioning stimulus. The thoracic spinal cord contains 2 broad classes of neurone according to whether they have a visceral input: (i) somatic neurones, driven only by somatic afferents and (ii) viscera-somatic neurones, driven by both somatic and visceral afferent fibres (Cervero and Tattersall 1985). While all the neurones of our study with an input from the gall bladder were, by definition, viscera-somatic we cannot say that all of the cells in the
other group were not. In order to limit surgical trauma to the abdominal skin and muscle layers we did not prepare the splanchnic nerves for electrical stimulation, our usual procedure to activate the entire visceral input to the lower thoracic cord (Cervero 1983a, b). So it is possibie that some (or all) of the neurones not driven from the gall bladder had inputs from other viscera and were therefore viscera-somatic. We know that more than 75% of all cells in the lower thoracic cord are viscera-somatic (Cervero and Tattersall 19851 and that only a one-third of them are driven by biliary afferents (Cervero 1983a). We also know that somatic ceils tend to be driven only by low-threshold mechanoreceptors from the skin (Cervero 1983a, bl and that only a minority of them are of the class-2 or -3 type. Therefore, it is very likely that our GB- group contained some viscera-somatic celis. It has been reported, by this and by other laboratories, that some spinal cord neurones do not alter their receptive field properties after the application of conditioning stimuli (Laird and Cervero 1989). In order to eliminate the possibility that all of our GB- neurones belonged to this category, we tested all cells for receptive field changes following reversible spinalization. Removal of tonic descending influences is a very useful tool to examine the capacity of spinal cord cells to change their receptive field properties (Schaible et al. 1991). This procedure also eliminates the need to apply peripheral noxious stimuli that can sensitize nociceptors and induce long-lasting central changes. Using reversible spinalization we were able to show that many cells of both the GB+ and GB- groups had a capacity to alter their somatic receptive fields. Therefore the lack of change observed on cells of the GBgroup after noxious stimulation of the gall bladder cannot be attributed to a general inability of these cells to express changes in the size of their somatic receptive fields. Selectirlity of the central changes
The central effects of a conditioning visceral stimulus appear to be selective to those cells driven by the stimulated viscera1 afferents. This observation adds to the body of experimental evidence that shows that central sensitization of spinal cord neurones is a highly selective process (Cervero et al. 1988; Laird and Cervero 1989, 19901. Work from this laboratory has shown that the mechanisms of central sensitization of dorsal horn neurones are fairly selective depending on the type of afferent input received by the cells. For instance, relatively minor noxious stimuli do not evoke changes in class-3 cells but can readily induce increases in the size of the receptive fields of class-2 (Laird and Cervero 1989). Further differences have been shown between class-3 neurones located in the superficial dorsal horn and those located in the deeper layers
341
Gird and Cervero 1990). Moreover, sub-populations of class-2 and class-3 cells also show differences in their ability to encode small intensity changes of noxious mechanical stimuli (Laird and Cervero 1991). In the present study we have confirmed many of these observations. The only class-3 cell recorded in the superficial dorsal horn was not driven by gall bladder stimulation and did not change its receptive field properties in the spinal state. Class-3 cells located in the deep dorsal horn expressed changes after spinalization and after noxious visceral stimulation similar to those of class-2 neurones. Clinical implications
Our results help to explain some of the clinical features of visceral pain. It is well known that visceral pain is often referred to the surface of the body and that the area of referral can vary in size depending on the intensity of the pain (MacKenzie 1909; Morley 1931). This phenomenon has been analysed recently by Ness et al. (1990) in a psychophysical study in man in which pain was induced by controlled colorectal distension. These authors documented the observation that repeated painful distensions of the colorectal region resulted in increased pain sensitivity and in an enlargement of the superficial area of the body to which the pain was referred. Our results provide a neuronal correlate for these observations by showing that the somatic receptive fields of spinal cord neurones can increase in size after noxious visceral stimulation. It has been proposed that the size of the somatic receptive fields of viscerosomatic neurones determines the surface area to which a visceral stimulus is referred (Ruth 1946). Therefore increases in the size of such receptive fields will result in larger areas of referral of the visceral noxious stimulus. The long duration of the central effects observed in our experiments is also in line with the clinical experience that somatic referrals of visceral pain tend to outlast the duration of the visceral stimulus.
Acknowledgements
The excellent technical assistance of Steve Allen and Simon Lishman and the financial support of the British MRC are gratefully acknowledged.
References Cervero, F., Afferent activity evoked by natural stimulation of the biliary system in the ferret, Pain, 13 (1982) 137-151. Cervero, F., Somatic and visceral inputs to the thoracic spinal cord of the cat: effects of noxious stimulation of the biliary system, J. Physiol., 337 (1983a) 51-67. Cervero, F., Supraspinal connections of neurones in the thoracic
spinal cord of the cat: ascending projections and effects of descending impulses, Brain Res., 275 (1983b) 251-261. Cervero, F., Visceral pain. In: R. Dubner, G.F. Gebhart and M.R. Bond (Eds.), Proc. Vth World Congresson on Pain, Elsevier, Amsterdam, 1988, pp. 216-226. Cervero, F. and Plenderleith, M.B., C-fibre excitation and tonic descending inhibition of dorsal horn neurones in adult rats treated at birth with capsaicin, J. Physiol., 365 (1985) 223-237. Cervero, F. and Tattersall, J.E.H., Somatic and visceral sensory integration in the thoracic spinal cord. In: F. Cervero and J.F.B. Morrison (Eds.), Visceral Sensation, Prog. Brain Res., Vol. 67, Elsevier, Amsterdam, 1986, pp. 189-205. Cervero, F., Schouenborg, J., Sjolund, B.H. and Waddell, P.J., Cutaneous inputs to dorsal horn neurones in adult rats treated at birth with capsaicin, Brain Res., 301 (1984) 47-57. Cervero, F., Handwerker, H.O. and Laird, J.M.A, Prolonged noxious mechanical stimulation of the rat’s tail: responses and encoding properties of dorsal horn neurones, J. Physiol., 404 (1988) 419436. Cetvero, F., Laird, J.M.A. and Pozo, M.A., Plastic changes of spinal nociceptive neurones induced by noxious visceral stimulation in the anaesthetized cat, J. Physiol., 452 (1992) 240P. Cook, A.J., Woolf, C.J., Wall, P.D. and McMahon, S.B. Dynamic receptive field plasticity in rat spinal cord dorsal horn following C-primary afferent input, Nature, 325 (1987) 151-153. Dawson, N.J., Schmid, H. and Pierau, F.-K., Pre-spinal convergence between thoracic and visceral nerves of the rat, Neurosci. Lett., 138 (1992) 149-152. Dubner, R., Hyperalgesia and expanded receptive fields, Pain, 48 (1992) 3-4. Hoheisel, U. and Mense, S., Long-term changes in discharge behaviour of cat dorsal horn neurones following noxious stimulation of deep tissues, Pain, 36 (1989) 239-247. Hylden, J.L.K., Nahin, R.L., Traub, R.J. and Dubner, R., Expansion of receptive fields of spinal lamina I projection neurones in rats with unilateral adjuvant-induced inflammation: the contribution of central dorsal horn mechanisms, Pain, 37 (1989) 229-243. Laird, J.M.A. and Cervero, F., A comparative study of the changes in receptive-field properties of multireceptive and nocireceptive rat dorsal horn neurons following noxious mechanical stimulation, J. Neurophysiol., 62 (1989) 854-863. Laird, J.M.A. and Cervero, F., Tonic descending influences on receptive-field properties of nociceptive dorsal horn neurons in sacral spinal cord of rat, J. Neurophysiol., 63 (1990) 1022-1032. Laird, J.M.A. and Cervero, F., Signalling of a step-like intensity change of noxious mechanical stimuli by dorsal horn neurones in the rat spinal cord. J. Physiol., 434 (1991) 561-575. MacKenzie, J., Symptoms and their interpretation, Shaw and Sons, London, 1909, pp. l-297. McMahon, S.B. and Wall, P.D., The receptive fields of rat lamina I projection cells move to incorporate a nearby region of injury, Pain, 19 (1984) 235-247. Meyer, R.A., Campbell, J.N. and Raja, S.N., Peripheral neural mechanisms of cutaneous hyperalgesia. Adv. Pain Res. Ther., 9 (1985) 53-71. Morley, J.A.. Abdominal Pain, Livingstone, Edinburgh, 1931, pp. 1-134. Ness, T.J., Metcalf, A.M. and Gebhart, G.F., A psychophysiological study in humans using phasic colonic distension as a noxious visceral stimulus, Pain, 43 (1990) 377-386. Raja, S.N., Campbell, J.N. and Meyer, R.A., Evidence for different sensory mechanisms of primary and secondary hyperalgesia following heat injury to glabrous skin, Brain, 107 (1984) 1179-1188. Ruth, T.C., Visceral sensation and referred pain. In: J.F. Fulton (Ed.), Howell’s Textbook of Physiology, Saunders, Philadelphia, 1946, pp. 385-401. Schaible, H.-G., Schmidt, R.F. and Willis, W.D., Jr., Enhancement
342 of the responses of ascending tract cells in the cat spinal cord by acute inflammation of the knee joint, Exp. Brain Res., 66 (19871 489-499. Schaible, H.-G., Neugebauer, V., Cervero. F. and Schmidt, R.F.. Changes in tonic descending inhibition of spinal neurons with articular input during the development of acute arthritis in the cat, J. Neurophysiol., 66 (1991) 1021-1031. Thompson, W.G., Irritable bowel syndrome. In: SF. Phillips, J.H. Pemberton and R.G. Shorter (Eds.1, The Large Intestine: Physiology, Pathophysiology and Disease, Raven Press, New York. 1991. pp. 593-610.
Woolf. C.J.. Afferent induced alterations of receptive field properties. In: F. Cervero, G.J. Bennett and P.M. Headley (Eds.), Processing of Sensory Information in the Superficial Dorsal Horn of the Spinal Cord, Plenum Press, New York, 1989, pp. 443-462. Woolf, C.J. and King, A.E., Subthreshold components of the cutaneous mechanoreceptive fields of dorsal horn neurons in the rat lumbar spinal cord, J. Neurophysiol.. 62 (1989) 907-916. Zieglglnsberger. W. and Herz. A., Changes of cutaneous receptive fields of spinocervical tract neurones and other dorsal horn neurones by microelectrophoretically administered amino-acids, Exp. Brain Res., 13 (1971) 111-126.
335 Publishers
B.V. All rights reserved
0304-3959/92/$05.00
PAIN 02163
Selective changes of receptive field properties of spinal nociceptive neurones induced by noxious visceral stimulation in the cat Fernando Cervero, Jennifer M.A. Laird ’ and Miguel A. Pozo 2 Department of Physiology. University of Bristol Medical School, UniLlersityWalk, Bristol BS8 I TD (UK) (Received
21 April 1992, revision received
16 June 1992; accepted
22 June 1992)
This study was designed to examine the central changes in the receptive field properties of dorsal Summary horn neurones induced by a period of visceral noxious stimulation. The aim of this investigation was to establish whether noxious stimulation of the visceral input to the spinal cord could influence transmission of cutaneous information through dorsal horn neurones. Single-unit electrical activity was recorded in the lower thoracic spinal cord of anaesthetized cats from dorsal horn neurones with a somatic receptive field in the ipsilateral flank. Changes in the properties of these receptive fields induced by reversible spinalization (by means of a cold block 4 or 5 segments rostra1 to the recording electrodes) and by a conditioning noxious stimulation of the biliary system (3 successive distensions of the gall bladder for 30 set at 65-80 mm Hg at 1-min intervals) were analysed. Nineteen neurones have been studied, 10 of which could be driven by stimulation of the gall bladder. All of these 10 cells showed increases in the size of their cutaneous receptive fields following conditioning noxious stimulation of the biliary system. The increases were large and lasted for at least 20 min. None of the 9 spinal cord neurones without an input from the gall bladder were affected by the conditioning visceral stimulus even though 7 showed changes in receptive field size when the animals were spinalised. These results show that noxious stimulation of viscera can evoke increases in the somatic receptive fields of spinal cord neurones but only of those neurones which are also driven by the visceral stimulus. Key words: Dorsal horn; Spinal cord; Visceral pain; Receptive fields; Plasticity; Biliary system
Introduction
There is currently considerable interest in the peripheral and central mechanisms that underlie hyperalgesic states. It has been known for some time that an injury to the skin and somatic tissues results in sensitization of peripheral nociceptors and that this mecha-
’ Present address: Neuroscience Research Centre, Laboratories, Terlings Park, Harlow, UK. 2 Present address: Instituto cante, Alicante, Spain.
de Neurociencias,
Merck
Universidad
Research
de Ali-
Correspondence to: Dr. F. Cervero, Dept. of Physiology, University of Bristol, Medical School, University Walk, Bristol BS8 1TD. UK.
nism contributes to the hyperalgesia evoked by low-intensity stimuli applied to the injured site (Raja et al. 1984; Meyer et al. 1985). However, attention has also been focussed on the central changes that result from the arrival of a nociceptive barrage to the spinal cord (see refs. in Dubner 1992). Much of the information about these central alterations has been obtained by examining changes in the receptive field properties of dorsal horn neurones. These changes are interpreted as an expression of the CNS reactivity to a peripheral injury and as the underlying mechanism for the increased central excitability that characterises hyperalgesic states (Woolf 1989). One of the first reports of changes in the properties of dorsal horn neurones following a brief period of noxious stimulation originated from this laboratory (Cervero et al. 1984). In this study it was shown that a train of 20 stimuli at 1 Hz applied to the sural nerve at
336
an intensity supramaximal for C fibres, evoked an increase in the size of the receptive field of dorsal horn neurones driven by sural afferents. Other reports have described similar changes in the receptive field size of dorsal horn neurones and in the magnitude of flexor reflexes following similar periods of electrical stimulation of a peripheral nerve or after the application of a variety of noxious stimuli to, or close to, the receptive field of the cells (McMahon and Wall 1984; Cook et al. 1987; Schaible et al. 1987; Hoheisel and Mense 1989; Hylden et al. 1989; Woolf 1989; Laird and Cervero 1989, 1990). Virtually all the reports quoted above examined only the responsiveness of dorsal horn cells to somatic noxious stimuli. Comparatively little is known about how dorsal horn cells react to prolonged noxious stimulation of internal organs. Yet, this information is of great clinical relevance for two reasons. Firstly, hyperalgesia referred to somatic and visceral structures is a frequent sequel of visceral pain states. Conditions such as irritable bowel syndrome are characterised by an increased sensitivity to physiological gastrointestinal stimuli, so that normal gut activity becomes painful (Thompson 1991). Secondly, it is a common clinical experience that the area of referral of visceral pain can change in size depending on the intensity of the pain (MacKenzie 1909; Morley 1931). For instance colic pain referred to the surface of the abdomen produces large areas of superficial tenderness at the peaks of the colic attacks and smaller areas during periods of quiescence. The present study was designed to examine the central changes in the receptive field properties of dorsal horn neurones induced by a period of visceral noxious stimulation. We have recorded from neurones in the lower thoracic spinal cord of anaesthetized cats and have looked at changes in the sizes of their cutaneous receptive fields following prolonged noxious stimulation of the gall bladder and biliary system. One of our aims was to establish whether noxious stimulation of the visceral input to the spinal cord could influence the transmission of cutaneous information through dorsal horn neurones. Since the conditioning stimulus used in this study was restricted to internal organs, our approach also eliminated ambiguities as to the peripheral or central origin of any observed effect on the somatic receptive fields of the neurones. Some of the results presented here have been published in abstract form (Cervero et al. 1992).
Cold
Fig.
1. Schematic
diagram
spinal cord stimulation
block
of the
experimental
arrangement.
A:
of the ItJth thoracic dorsal root, recording
the 9th thoracic segment and cold block. B: mechanical of the biliary system; Tr = pressure
transducer.
in
stimulation
For further
details
see text.
curonium
bromide
tive-pressure
(0.06 mg/kg/h,
iv.1 and ventilated
with a posi-
pump. The general methods for monitoring
and main-
tenance of the physiological state of the animals have been described previously in detail (Cervero The
lower
thoracic
from T9 to Tl2. TS-T6
1983a, bJ.
spinal cord was exposed
A second smaller
by a laminectomy
laminectomy
to allow reversible spinalization
was performed
means of a cold block (see below). The animals were mounted rigid frame
at
of the animals at this level by in a
and pools were made with skin flaps over the exposed
areas of the spinal cord. Recording ing the vertebral
stability was improved
column, by infiltration
cord and by a bilateral
pneumothorax.
filled with warm paraffin
by clamp-
of 3% agar around the spinal All spinal cord pools were
oil at 38°C.
Recording techniques Extracellular croelectrodes
was X-20 Mf2J. the
right
searching
in T10
measured
at I KHz
for
through
neurones.
of the
segments
T9-TlO.
or Tl I was electrically ball-tipped
Recordings
oscilloscope
and analysed ‘on-line’
puter.
IA
Fig.
(impedance
grey matter
dorsal rootlet
V, 0.1 msec, I Hz)
when
recordings were made through glass m-
Recordings were made from dorsal horn neurones in
side of the
ipsilateral (I-3
single-unit
filled with 4 M NaCl
silver wire electrodes were
and ‘off-line‘
shows a schematic
An
stimulated
diagram
of
displayed
on an
using a microcomthe
experimental
arrangements.
Stimulation of somatic receptille fields Somatic
afferent
their receptive
fibres were activated
by natural
stimulation
of
fields. Stimuli included both innocuous (i.e., brushing
and stroking) as well as noxious stimuli (i.e., pinching and squeezing). The
boundaries
neurones
were
of the somatic mapped
receptive
fields of the dorsal
for both low-threshold
responses and marked on the flank of the animals with felt-tip of different receptive paper
colours. At the end of the study of each neurone,
field boundaries
were transferred
and the areas measured
horn
and high-threshold
to transparent
using a digitising tablet
pens the
drawing
and micro
computer.
Methods
Stimulation of Lisceral afferent fibres The conditioning
Experiments
body
of distensions applied
weights between 2.8 and 3.9 kg. All animals were anaesthetized
with
tion of the gall bladder
halothane;
conducted
(60-70
30 mg/kg.
mg/kg,
on Y adult i.v.;
i.v., maintenance
initial
male
noxious stimulus used in this study was a series
cats with
alpha-chloralose
were
dose
under
doses), paralysed
2.5”i
with pan-
previously in Fig.
described
to the biliary system of the animals. Stimulaand biliary ducts was achieved by a method
in detail (Cervero
1B. It involved
the cannulation
1982. 1983aJ and illustrated of the fundus
of the gall
337 bladder with a double-barrelled catheter. One barrel was connected to a pressure transducer and the other to a reservoir containing warmed saline solution. The common bile duct was ligated so that pressure changes could be imposed on the biliary system by raising the reservoir to various heights above the preparation.
25 20 15 Imp/set 10
Reversible spinalization
5
This was achieved by circulating cold fluid through a silver-plated thermode placed over the cord at T5-T6 level. Details of this technique have been published previously as well as tests of the efficacy of this method in providing a reversible spinalization (Cervero and Plenderleith 1985; Laird and Cervero 1990).
0
Experimental protocol The experimental protocol applied to each identified neurone consisted of the following sequential steps. (8 A brief, noxious distension of the gall bladder (5 XC, 50-70 mm Hg) was performed to ascertain if the cell responded to this stimulus. (ii) The somatic receptive field was mapped on the flank of the animal and marked as described above. This was recorded as the initial receptive field of the cell. (iii) A reversible spinalization of the animal was then performed and the receptive field was re-mapped in the spinal state. This was recorded as the spinal receptive field of the neurone. (iv) A brief, noxious distension of the gall bladder (parameters as above) was then applied to check for responsiveness of the cell to this stimulus in the spinal state. (v) The cord was rewarmed and the receptive field was again mapped to ensure that, if any changes had been observed in the spinal state, it had returned to the initial receptive field size. (vi) A conditioning noxious stimulation of the biliary system was then applied to the animal. This consisted of a series of 3 stimuli delivered at I-min intervals. Each stimulus was a pressure pulse of 65-80 mm Hg lasting for 30 sec. This conditioning stimulus was designed to mimic the pressure changes occurring during a series of colic contractures of the gall bladder. (vii) After the administration of the visceral conditioning stimulus the somatic receptive field was again mapped at 5-min intervals. This was recorded as the receptive field after gall bladder stimulation. Only a few neurones were studied in each animal to avoid cumulative effects of the conditioning stimulus. At least 1 h was allowed to elapse between the study of 2 successive neurones.
Histological methods The position of the recording microelectrode was marked by ionophoretic deposition of pontamine sky blue in the last track of each experiment. Marks were made in this track at l-mm intervals. The section of the cord was removed and frozen. The recording sites of the neurones were calculated from these spots recovered in 50-pm sections counterstained with haematoxylin-eosin. Details of this technique for the location of recording sites have been previously published (Cetvero 1983a).
Results Sample of new-ones
The results presented in this paper are based on a sample of 19 neurones recorded in the lower thoracic spinal cord. The search stimulus was electrical stimulation of a dorsal rootlet entering the same or an immediately caudal segment of the cord at an intensity supramaximal for all A fibres in the rootlet. Neurones included in the present sample were those responding to this stimulation and having a cutaneous receptive
0
50
100 Time
GE
Press.
,G
i
150
200
(sacs)
~
(mmHg) ,3 L M 1 Fig. 2. Response of a GB+ neurone in the thoracic spinal cord to repeated distensions of the biliary system. Top: ratemeter output of the spike discharge of the neurone. Bottom: biliary pressure.
field on the ipsilateral flank from which they could be driven by noxious mechanical stimulation. Some neurones showed, in addition, responses to innocuous mechanical stimulation of the receptive field (see below). Recordings judged to be made from axons rather than from cell bodies were discarded. Responses to gall bladder distension
All 19 neurones were challenged with distensions of the gall bladder at noxious intensities (see Methods). The sample of cells was divided into 2 groups depending on whether the neurones responded to this stimulus. Ten of the neurones responded to these distensions and were thus classified as having an input from the gall bladder (GB + >. Fig. 2 shows the responses of one such neurone to the visceral-conditioning stimulus. Characteristically the cells of this group responded to the increase in biliary pressure with an initial phasic discharge which adapted to a lower level of tonic response throughout the duration of the stimulus. The levels of background activity of this group of cells increased progressively in between successive stimuli (see Fig. 21. The other 9 cells did not respond to any of the distensions of the biliary system and were thus classified as lacking an input from the gall bladder (GB- ). All the cells in this group had been tested for gall bladder inputs while the animals were reversibly spinalised to ensure that descending inhibition was not masking a gall bladder input to the cells. None responded to biliary distension in the spinalised state. Somatic receptive fields
The neurones were also classified according to their responses to natural stimulation of their cutaneous receptive fields. Fifteen neurones responded to both innocuous and noxious stimulation and were thus classified as class 2 (or wide-dynamic-range). Four cells responded exclusively to noxious stimulation of their
338
jGB+i
GB-
Class
2
lomino
m
Fig. 3. Locations of the recording superimposed
and GB-
have also been divided
tive field
properties
nil?
RF
neurones
acdording to their somatic recep-
(class 2 and class 3). The
in the laminar
maps
is shown
grey matter
in the
central
area
diagram
(dotted square).
Fig. 5. Increase neurone
in the size of the somatic receptive
following
change after repeated
receptive fields and were classified as class 3 (nociceptive-specific). As shown in Fig. 3, the sample was equally divided when taking into account the presence or absence of an input from the gall bladder and the properties of the cutaneous receptive fields. The GB+ group comprised 8 class-2 and 2 class-3 cells and the GB- group included 7 class-2 and 2 class-3 neurones. Location of neuroiws
All neurones were recorded within the ipsilateral grey matter, mainly in the dorsal horn (Fig. 3). There was a similar distribution of superficial and deep neurones between the GB+ and GB- groups. Most cells were recorded in the deep dorsal horn, in or around lamina V.
Class 2
lamina
V
GE+
Fig. 4. Increases in the size of the somatic receptive neurone following reversible spinalization bladder
GB-
on standard sections of the lower thoracic spinal cord.
Neurones magnified
sites of GB+
initial
V
(bottom)
field of a GB+ and repeated
gall
distension (top). The initial receptive field is shown in black.
The location of the recording site of the neurone left corner of the figure.
is shown in the top
reversible
spinalization
field of a GB-
(bottom)
and
lack
of
gall bladder distension (top). The initial recep-
tive field is shown in black. The location of the recording
site of the
neurone is shown in the top left corner of the figure.
Receptille field changes after spinalization and after noxious stimulation of the biliary system
The receptive fields of all cells were examined before and after reversible spinalization of the animals and before and after the application of the conditioning visceral stimulus. Quantitative measurements of the receptive field areas were obtained from 8 GB+ and 8 GB- neurones, Two types of response were observed depending on whether the cells responded to gall bladder stimulation. (i) GB+ neurunes. All but one of the neurones having an input from the gall bladder showed increases in the size of their cutaneous receptive fields while in the spinalized state (Figs. 4 and 6). The receptive field of the remaining neurone decreased in size in the spinal state. After rewarming the spinal cord, the receptive fields of all neurones returned to their initial size. All GB+ cells showed increases in the sizes of their receptive fields after the conditioning visceral stimulation (Figs. 4 and 6). The increases lasted for at least 20 min and, in some cases, the receptive fields did not return to their initiai sizes until 40-45 min after stimulation. The increases were quite large (range: 20-136% larger than the original size) but were generally smaller than the increases induced by spinalization (range: 50-339s larger than the original size) (see Figs. 4 and 6). Increases in receptive field size affected both the low- and high-threshold areas of the receptive fields of class-2 cells. However, the quantitative data presented in Fig. 6 apply only to the high-threshold areas of class-2 cells so that a comparison can be established with the receptive fields of class-3 cehs. The only cell whose receptive field decreased in size in the spinalized state was a class-3 cell.
339 Cutaneous
Receptive
Field
Changes
Fig. 6. Changes in the size of the somatic receptive fields of GB+ and GB- neurones following reversible spinalization and repeated gall bladder distensions. Neurones have also been divided according to their somatic receptive field properties (class 2 and class 3). Note that none of the GB- neurones changed their somatic receptive fields after gall bladder stimulation even though most of them had larger receptive fields in the spinal state.
(ii) GB- neurones. The receptive fields of all but 2 of the neurones that did not have an input from the gall bladder increased in size after reversible spinalization (range: 122-373% larger than the original size) (Figs. 5 and 6). After rewarming the spinal cord, the receptive fields of these neurones returned to their initial size. The other 2 cells did not change the size of their receptive fields in the spinalised state. In contrast, none of the cells in the GB- group showed changes in the size of their receptive fields following conditioning visceral stimulation (Figs. 5 and 6). Thus, even though the receptive fields of many neurones in this group could increase in size after spinalization, they remained unchanged after the conditioning noxious visceral stimulation.
Discussion The results presented in this paper show that the somatic receptive fields of spinal cord neurones can change in size following a period of noxious visceral stimulation. This observation indicates that noxious stimuli applied to internal organs induce alterations in the spinal processing of somatosensory information. Since the conditioning noxious stimulus was applied to visceral afferent fibres and the changes observed were expressed on the cutaneous receptive fields of the neurones it is fairly certain that these alterations in excitability took place centrally rather than peripherally. In addition, our results also indicate a high degree of selectivity in the expression of these central changes depending on whether or not the spinal cord cells are
driven by the visceral fibres activated by the conditioning stimulus. This study extends current observations on the functional plasticity of spinal neurones to those driven by visceral afferent fibres. We have studied cells in the lower thoracic spinal cord driven by noxious stimulation of the gall bladder, a group of neurones which are putative candidates for the signalling of pain from the biliary system (Cervero and Tattersall 1985). The conditioning stimulus used was designed to reproduce the pattern of a biliary colic attack and to maximise the afferent barrage triggered on high threshold biliary afferents (Cervero 1982). However, this kind of stimulus is insufficient to produce long-lasting inflammatory or traumatic changes in the biliary system and therefore the alterations observed in the present study should be interpreted as those likely to underlie acute, rather than chronic, visceral pain states. Peripheral versus central effects
There is some debate as to whether peripheral sensitization of nociceptors or central sensitization of spinal neurones play a major role in the triggering of hyperalgesic states. The present study clearly shows that the afferent barrage triggered by the visceral stimulus induced changes that were predominantly central. For the changes in the cutaneous receptive fields of the neurones to have had a peripheral origin it would be necessary to postulate that the noxious stimulation of the biliary system could have altered the responsiveness of cutaneous receptors. While we cannot discard this possibility with our data, this seems unlikely. There are also some descriptions in the literature of afferent
340
fibres with branches in somatic and in visceral nerves (i.e., see Dawson et al. 1992). These fibres are extremely rare and, in the absence of data as to whether or not they are connected to active receptor sites in the skin and viscera, their functional significance remains obscure. Our interpretation of the observed changes in receptive field size is that the barrage evoked in visceral afferent fibres by the noxious stimulus removed central inhibition of spinal neurones. This increase in excitability meant that the cells could now be activated by input from what was previously within their subliminal fringes (Zieglgansberger and Herz 1971; Woolf and King 1989). Therefore, regardless of whether the visceral nociceptors were sensitized, the change in the somatic receptive field of the neurones was clearly a central phenomenon. The long duration of these effects also suggests that the central changes were not maintained by a continuous barrage from the stimulated visceral afferents. ~gh-threshold nociceptors in the gall bladder and biliary ducts have very low levels of background activity and tend to desensitize on repetitive stimulation (Cervero 1982). Therefore, it seems that although the initial barrage triggered by the conditioning stimulus was the cause of the central changes, a persistent drive from the periphery was not necessary for the maintenance of these effects. It is more likely that the positive feed-back loops between brain-stem nuclei and viscerosomatic neurones in the spinal cord that we have described (Cervero 1988) played a more important role in sustaining central excitability. Effects of visceral stimulation
One of the key observations of the present study is the high degree of selectivity observed between different spinal neurones in the expression of excitability changes. Our results show that all neurones driven by gall bladder afferents increased their somatic receptive fields after visceral stimulation. In contrast, none of the cells without a gall bladder input did so. These data suggest that an excitatory input from the viscus that undergoes noxious stimulation is necessary for the cell to express changes in its somatic receptive field. We cannot, however, extend this conclusion to suggest that a noxious visceral stimulus would influence all cells driven by visceral afferents, including those not directly activated by the conditioning stimulus. The thoracic spinal cord contains 2 broad classes of neurone according to whether they have a visceral input: (i) somatic neurones, driven only by somatic afferents and (ii) viscera-somatic neurones, driven by both somatic and visceral afferent fibres (Cervero and Tattersall 1985). While all the neurones of our study with an input from the gall bladder were, by definition, viscera-somatic we cannot say that all of the cells in the
other group were not. In order to limit surgical trauma to the abdominal skin and muscle layers we did not prepare the splanchnic nerves for electrical stimulation, our usual procedure to activate the entire visceral input to the lower thoracic cord (Cervero 1983a, b). So it is possibie that some (or all) of the neurones not driven from the gall bladder had inputs from other viscera and were therefore viscera-somatic. We know that more than 75% of all cells in the lower thoracic cord are viscera-somatic (Cervero and Tattersall 19851 and that only a one-third of them are driven by biliary afferents (Cervero 1983a). We also know that somatic ceils tend to be driven only by low-threshold mechanoreceptors from the skin (Cervero 1983a, bl and that only a minority of them are of the class-2 or -3 type. Therefore, it is very likely that our GB- group contained some viscera-somatic celis. It has been reported, by this and by other laboratories, that some spinal cord neurones do not alter their receptive field properties after the application of conditioning stimuli (Laird and Cervero 1989). In order to eliminate the possibility that all of our GB- neurones belonged to this category, we tested all cells for receptive field changes following reversible spinalization. Removal of tonic descending influences is a very useful tool to examine the capacity of spinal cord cells to change their receptive field properties (Schaible et al. 1991). This procedure also eliminates the need to apply peripheral noxious stimuli that can sensitize nociceptors and induce long-lasting central changes. Using reversible spinalization we were able to show that many cells of both the GB+ and GB- groups had a capacity to alter their somatic receptive fields. Therefore the lack of change observed on cells of the GBgroup after noxious stimulation of the gall bladder cannot be attributed to a general inability of these cells to express changes in the size of their somatic receptive fields. Selectirlity of the central changes
The central effects of a conditioning visceral stimulus appear to be selective to those cells driven by the stimulated viscera1 afferents. This observation adds to the body of experimental evidence that shows that central sensitization of spinal cord neurones is a highly selective process (Cervero et al. 1988; Laird and Cervero 1989, 19901. Work from this laboratory has shown that the mechanisms of central sensitization of dorsal horn neurones are fairly selective depending on the type of afferent input received by the cells. For instance, relatively minor noxious stimuli do not evoke changes in class-3 cells but can readily induce increases in the size of the receptive fields of class-2 (Laird and Cervero 1989). Further differences have been shown between class-3 neurones located in the superficial dorsal horn and those located in the deeper layers
341
Gird and Cervero 1990). Moreover, sub-populations of class-2 and class-3 cells also show differences in their ability to encode small intensity changes of noxious mechanical stimuli (Laird and Cervero 1991). In the present study we have confirmed many of these observations. The only class-3 cell recorded in the superficial dorsal horn was not driven by gall bladder stimulation and did not change its receptive field properties in the spinal state. Class-3 cells located in the deep dorsal horn expressed changes after spinalization and after noxious visceral stimulation similar to those of class-2 neurones. Clinical implications
Our results help to explain some of the clinical features of visceral pain. It is well known that visceral pain is often referred to the surface of the body and that the area of referral can vary in size depending on the intensity of the pain (MacKenzie 1909; Morley 1931). This phenomenon has been analysed recently by Ness et al. (1990) in a psychophysical study in man in which pain was induced by controlled colorectal distension. These authors documented the observation that repeated painful distensions of the colorectal region resulted in increased pain sensitivity and in an enlargement of the superficial area of the body to which the pain was referred. Our results provide a neuronal correlate for these observations by showing that the somatic receptive fields of spinal cord neurones can increase in size after noxious visceral stimulation. It has been proposed that the size of the somatic receptive fields of viscerosomatic neurones determines the surface area to which a visceral stimulus is referred (Ruth 1946). Therefore increases in the size of such receptive fields will result in larger areas of referral of the visceral noxious stimulus. The long duration of the central effects observed in our experiments is also in line with the clinical experience that somatic referrals of visceral pain tend to outlast the duration of the visceral stimulus.
Acknowledgements
The excellent technical assistance of Steve Allen and Simon Lishman and the financial support of the British MRC are gratefully acknowledged.
References Cervero, F., Afferent activity evoked by natural stimulation of the biliary system in the ferret, Pain, 13 (1982) 137-151. Cervero, F., Somatic and visceral inputs to the thoracic spinal cord of the cat: effects of noxious stimulation of the biliary system, J. Physiol., 337 (1983a) 51-67. Cervero, F., Supraspinal connections of neurones in the thoracic
spinal cord of the cat: ascending projections and effects of descending impulses, Brain Res., 275 (1983b) 251-261. Cervero, F., Visceral pain. In: R. Dubner, G.F. Gebhart and M.R. Bond (Eds.), Proc. Vth World Congresson on Pain, Elsevier, Amsterdam, 1988, pp. 216-226. Cervero, F. and Plenderleith, M.B., C-fibre excitation and tonic descending inhibition of dorsal horn neurones in adult rats treated at birth with capsaicin, J. Physiol., 365 (1985) 223-237. Cervero, F. and Tattersall, J.E.H., Somatic and visceral sensory integration in the thoracic spinal cord. In: F. Cervero and J.F.B. Morrison (Eds.), Visceral Sensation, Prog. Brain Res., Vol. 67, Elsevier, Amsterdam, 1986, pp. 189-205. Cervero, F., Schouenborg, J., Sjolund, B.H. and Waddell, P.J., Cutaneous inputs to dorsal horn neurones in adult rats treated at birth with capsaicin, Brain Res., 301 (1984) 47-57. Cervero, F., Handwerker, H.O. and Laird, J.M.A, Prolonged noxious mechanical stimulation of the rat’s tail: responses and encoding properties of dorsal horn neurones, J. Physiol., 404 (1988) 419436. Cetvero, F., Laird, J.M.A. and Pozo, M.A., Plastic changes of spinal nociceptive neurones induced by noxious visceral stimulation in the anaesthetized cat, J. Physiol., 452 (1992) 240P. Cook, A.J., Woolf, C.J., Wall, P.D. and McMahon, S.B. Dynamic receptive field plasticity in rat spinal cord dorsal horn following C-primary afferent input, Nature, 325 (1987) 151-153. Dawson, N.J., Schmid, H. and Pierau, F.-K., Pre-spinal convergence between thoracic and visceral nerves of the rat, Neurosci. Lett., 138 (1992) 149-152. Dubner, R., Hyperalgesia and expanded receptive fields, Pain, 48 (1992) 3-4. Hoheisel, U. and Mense, S., Long-term changes in discharge behaviour of cat dorsal horn neurones following noxious stimulation of deep tissues, Pain, 36 (1989) 239-247. Hylden, J.L.K., Nahin, R.L., Traub, R.J. and Dubner, R., Expansion of receptive fields of spinal lamina I projection neurones in rats with unilateral adjuvant-induced inflammation: the contribution of central dorsal horn mechanisms, Pain, 37 (1989) 229-243. Laird, J.M.A. and Cervero, F., A comparative study of the changes in receptive-field properties of multireceptive and nocireceptive rat dorsal horn neurons following noxious mechanical stimulation, J. Neurophysiol., 62 (1989) 854-863. Laird, J.M.A. and Cervero, F., Tonic descending influences on receptive-field properties of nociceptive dorsal horn neurons in sacral spinal cord of rat, J. Neurophysiol., 63 (1990) 1022-1032. Laird, J.M.A. and Cervero, F., Signalling of a step-like intensity change of noxious mechanical stimuli by dorsal horn neurones in the rat spinal cord. J. Physiol., 434 (1991) 561-575. MacKenzie, J., Symptoms and their interpretation, Shaw and Sons, London, 1909, pp. l-297. McMahon, S.B. and Wall, P.D., The receptive fields of rat lamina I projection cells move to incorporate a nearby region of injury, Pain, 19 (1984) 235-247. Meyer, R.A., Campbell, J.N. and Raja, S.N., Peripheral neural mechanisms of cutaneous hyperalgesia. Adv. Pain Res. Ther., 9 (1985) 53-71. Morley, J.A.. Abdominal Pain, Livingstone, Edinburgh, 1931, pp. 1-134. Ness, T.J., Metcalf, A.M. and Gebhart, G.F., A psychophysiological study in humans using phasic colonic distension as a noxious visceral stimulus, Pain, 43 (1990) 377-386. Raja, S.N., Campbell, J.N. and Meyer, R.A., Evidence for different sensory mechanisms of primary and secondary hyperalgesia following heat injury to glabrous skin, Brain, 107 (1984) 1179-1188. Ruth, T.C., Visceral sensation and referred pain. In: J.F. Fulton (Ed.), Howell’s Textbook of Physiology, Saunders, Philadelphia, 1946, pp. 385-401. Schaible, H.-G., Schmidt, R.F. and Willis, W.D., Jr., Enhancement
342 of the responses of ascending tract cells in the cat spinal cord by acute inflammation of the knee joint, Exp. Brain Res., 66 (19871 489-499. Schaible, H.-G., Neugebauer, V., Cervero. F. and Schmidt, R.F.. Changes in tonic descending inhibition of spinal neurons with articular input during the development of acute arthritis in the cat, J. Neurophysiol., 66 (1991) 1021-1031. Thompson, W.G., Irritable bowel syndrome. In: SF. Phillips, J.H. Pemberton and R.G. Shorter (Eds.1, The Large Intestine: Physiology, Pathophysiology and Disease, Raven Press, New York. 1991. pp. 593-610.
Woolf. C.J.. Afferent induced alterations of receptive field properties. In: F. Cervero, G.J. Bennett and P.M. Headley (Eds.), Processing of Sensory Information in the Superficial Dorsal Horn of the Spinal Cord, Plenum Press, New York, 1989, pp. 443-462. Woolf, C.J. and King, A.E., Subthreshold components of the cutaneous mechanoreceptive fields of dorsal horn neurons in the rat lumbar spinal cord, J. Neurophysiol.. 62 (1989) 907-916. Zieglglnsberger. W. and Herz. A., Changes of cutaneous receptive fields of spinocervical tract neurones and other dorsal horn neurones by microelectrophoretically administered amino-acids, Exp. Brain Res., 13 (1971) 111-126.
