Brain Research Reviews 40 (2002) 29–44www.elsevier.com/ locate/brainresrev

Review

T he whole body receptive field of dorsal horn multireceptive neurones*Daniel Le Bars

´INSERM, 2, Rue d’ Alesia, 75014Paris, France

Abstract

Multireceptive neurones are found in the spinal dorsal horn and may be projection neurones and/or interneurones for polysynapticreflexes. The cutaneous receptive field of a multireceptive neurone exhibits a gradient of sensitivity with the centre responding to anymechanical stimulus, including hair movements and light touch, while the periphery responds only to noxious stimuli. These neuronesalso receive signals from viscera, muscles and joints. This convergence of inputs means that multireceptive neurones are continuouslycapturing all the information from both the interface with the external environment (the skin) and the internal milieu (the viscera, muscles,etc.). This information constitutes a ‘basic somaesthetic activity’ that could help the somatosensory system build a ‘global representationof the body’. In addition to be seen as a global entity, the output of multireceptive neurones should be understood in dynamic terms sincethe size of the peripheral fields of the individual neurones may change, as a result of the plasticity of both excitatory and inhibitorysegmental processes. Furthermore, the activity of these neurones can be inhibited from most of the remaining parts of the body viasupraspinal mechanisms. These diffuse noxious inhibitory controls (DNIC) are triggered by peripheral Ad- and C-fibres, involve brainstructures confined to the caudal-most part of the medulla including the subnucleus reticularis dorsalis (SRD) and are mediated bydescending pathways in the dorsolateral funiculi. A painful focus that both activates a segmental subset of neurones and inhibits theremaining population can seriously disrupt this basic activity, resulting in the distortion of the body representation in favour of the painfulfocus, which becomes pre-eminent and (relatively) oversized. 2002 Elsevier Science B.V. All rights reserved.

Keywords: Pain; Dorsal horn; Multireceptive neurones; Wide-dynamic-range neurones; DNIC; Body schema; Body image

Contents

1 . Introduction ............................................................................................................................................................................................ 292 . Segmental excitatory processes................................................................................................................................................................. 303 . Potential plasticity of segmental excitatory processes ................................................................................................................................. 324 . Segmental inhibitory processes ................................................................................................................................................................. 335 . Potential plasticity of segmental inhibitory processes ................................................................................................................................. 336 . Propriospinal inhibitory processes............................................................................................................................................................. 347 . Diffuse noxious inhibitory controls ........................................................................................................................................................... 348 . Spatial summation ................................................................................................................................................................................... 359 . Summary and conclusions ........................................................................................................................................................................ 39Acknowledgements ...................................................................................................................................................................................... 40References................................................................................................................................................................................................... 41

1 . Introduction more so, in deep areas centred on lamina V. They areactivated both by a variety of nociceptive stimuli and by

‘Multireceptive’ neurones are found in the dorsal horn weak mechanical stimulation. Various other names haveof the spinal cord—both in the superficial layers and, even been used by different authors for these neurones, e.g.,

‘trigger’, ‘lamina V type’, ‘wide-dynamic-range’, ‘class 2’‘convergent’ neurones [142]. Multireceptive neurones in-*Tel.: 133-1-4078-9358; fax:133-1-4588-1304.

E-mail address: lebars@broca.inserm.fr(D. Le Bars). clude interneurones involved in polysynaptic reflexes

0165-0173/02/$ – see front matter 2002 Elsevier Science B.V. All rights reserved.PI I : S0165-0173( 02 )00186-8

30 D. Le Bars / Brain Research Reviews 40 (2002) 29–44

[34,94,102,119,121,123,125] and neurones projecting in unusual property has some unexpected functional conse-ascending pathways such as the spino-thalamic and spino- quences.reticular tracts [142]. Interestingly, the most frequentlyencountered somatosensory neurones in the dorsal horn ofnon-anaesthetised fully conscious animals had wide dy- 2 . Segmental excitatory processesnamic range properties in the sheep [52,53] but not in thecat [103,129]. Most of the common procedures used for The nature of the excitatory receptive fields of mul-the alleviation of pain in humans (e.g., systemic or tireceptive neurones is well known and has been describedintrathecal morphine, transcutaneous electrical nerve by many laboratories (Fig. 1). They are fairly restricted instimulation, dorsal column stimulation) also result in a size with well-circumscribed and obvious boundaries.reduction in the responses of multireceptive cells to Although they are larger than those of nociceptive-specificnoxious stimulus. For this reason amongst others, these neurones, they are still compatible with providing usefulneurones are believed to play a key role in pain processes. information regarding stimulus location [142]. Usually, theIn addition however, they constitute a strategic site where receptive field exhibits a gradient of sensitivity: in thevarious types of excitatory and inhibitory influences con- centre, any mechanical stimulus including small hairverge. movements or light touch, can activate the neurone; by

Determining the receptive fields of such neurones at first contrast, at the periphery, only more intense stimuli elicitmay seem like a simple task. However it is one that must neuronal responses [56,69,137]. When stimuli of increasingbe approached with care. Some of the related problems are intensities are applied to the centre of the receptive field,obviously of a technical nature and can be solved by both the rate and the duration of neuronal firing increase.adequate experimental protocols. For example, the physio- This is best demonstrated by the use of a single modalitylogical condition of the animal is monitored by recording of stimulation under standardised temporal conditions andheart rate, blood pressure and core temperature in most suggests that such neurones are able to encode the intensityexperiments, but peripheral temperature and vascular tone of a peripheral stimulus by varying both the rate andcould modify the responsiveness of a neurone to a given duration of their spike discharges. Examples are thestimulus. This is particularly relevant in pharmacological increasing discharges to mechanical stimuliexperiments in which the administration of a drug may [56,88,90,136,138] and the graded responses to tempera-unintentionally produce such peripheral changes, or in ture [48,66,91,106] (see insert in Fig. 1). Studies in thephysiological experiments in which electrical stimulation awake monkey illustrate that these neurones are better ableof the brain or other structures may produce such changes to encode noxious stimuli than are nociceptive-specific[33,78]. The use of noxious stimuli can induce sensitisa- neurones [59].tion and/or desensitisation of peripheral receptors as well When a cutaneous nerve is stimulated electrically, it isas discrete inflammatory processes [132]. Thus the prior possible to define which categories of afferent fibresapplication of search stimuli and/or the testing of neurones activate multireceptive neurones. In fact they receivewhose receptive fields have already been stimulated could information from all three groups of cutaneous afferent,be problematic. One has therefore to bear in mind that any i.e., Ab-, Ad- and C-fibres [46,91,108,107,136]. Record-part of the body, and in particular the area that is ings from multireceptive neurones have also been made indesignated as the receptive field for a given neurone, has the medulla in both nucleus caudalis and nucleus oralis ofits own history before recordings are made from that the trigeminal system [4]. Interestingly, the receptive fieldsparticular neurone. of such neurones recorded in awake, behaving monkeys

In this review, these problems will not be emphasised, can expand or contract, depending of the attentional statenor will those related to temporal summation or central of the animal [32,50,59].sensitisation, although multireceptive neurones are highly It is often believed that under non-pathological con-susceptible to these phenomena. For instance, the repetitive ditions, these neurones respond with a higher frequency ofapplication of electrical stimuli at strength sufficient to firing to any form of noxious stimulation than to any formexcite C-fibres produces the well-known phenomenon of of innocuous stimulation. Indeed this seems to be the case‘wind up’ in multireceptive neurones, which consists of a during recordings of spino-thalamic neurones in the mon-progressive increase in their discharge to each successive key anaesthetised with barbiturates [142]. However, this isstimulus [88,90,107,108,136]. These matters have been the not the case in the rat, either under halothane anaesthesiasubjects of several recent magnificent reviews [55,64,144] or non-anaesthetised following spinalisation. As shown for

¨(see also the contributions by Hunsgaard, Sandkuhler and an individual neurone in the insert of Fig. 1, the dischargeWillis in this volume). The present paper will focus on the rate increased in relation to the temperature applied to thephysiological properties of multireceptive neurones in neurone’s receptive field with the highest discharge beingterms of their receptive fields under normal conditions, and produced by noxious intensities, although a very high levelmore particularly, on the concept that they exhibit whole was also elicited by adequate repetitive innocuous me-body receptive fields. This fundamental and disconcerting chanical stimuli [72] (and see Ref. [102]). Price et al. [109]

D.

Le

Bars

/B

rainR

esearchR

eviews

40 (2002) 29–4431

Fig. 1. Schematic organisation of the segmental receptive field of a multireceptive neurone. It is located on the skin and is made up of an excitatory (EF) and an inhibitory field (IF). The centre of theEF is activated by both innocuous (blue areas) and nociceptive (red areas) stimuli. The periphery of the EF is activated only by intense stimuli (red areas). The inhibitory field (green area) is activatedmainly by non-nociceptive stimuli, above all if they are applied in a repetitive and rapid fashion (hair movement, rubbing, vibration, etc.). The receptive field very often also includes a visceral (andsometimes a muscular or articular) component. One can then see a peculiar convergence of information onto a single neurone. In summary, peripheral influences are both excitatory (1) and inhibitory(2). One can relieve pain by reinforcing segmental inhibitory controls through electrical stimulation of nerves at high frequency and low intensity orby other empirical physical means (friction,rubbing, scraping, scratching, etc.). A ratemeter record of a lumbar multireceptive neurone studied in the rat is shown in the insert; the record shows the responses to stimuli applied on the centre of theexcitatory field. The first four sequences show the responses of the neurone to applications of radiant heat (as indicated by bars with the temperature achieved): the neurone responded with anincreasing firing rate as the temperature increased within the noxious range. A response of the neurone to repetitive innocuous stimuli (light pressure applied by means of a blunt probe) is shown on theextreme right; note the high level of firing achieved with these repetitive stimuli. Adapted from Ref. [72].

32 D. Le Bars / Brain Research Reviews 40 (2002) 29–44

reported that in trigeminal nucleus caudalis of the monkey, reception of information relevant to the integrity of thethe firing frequencies evoked in multireceptive neurones by body. One can imagine for example that such an activity44–488C radiant heat applications could often also be assists the proprioceptive system and other sensory sys-attained by innocuous mechanical stimuli. These authors tems, notably the vestibular and visual, in the building ofattributed this observation to the effects of anaesthetics. our ‘body schema’ or ‘body image’ [51,118]. MoreHowever our observations in the rat were identical in the specifically, Head and Holmes [51] proposed both aanaesthetised ‘intact’ animal and in the non-anaesthetised ‘postural model of ourselves’ and ‘another schema’ allow-‘spinal’ animal, suggesting that the level of anaesthesia ing the ‘recognition of the locality of the stimulated spot’.was not the explanation for such observations [72] (and see The ‘basic somaesthetic activity’ of multireceptive neuro-Ref. [52]). nes might well constitute a continuous source of infor-

With regard to the overlapping of excitatory fields, the mation for the building and rebuilding of such a secondspatial organisation of the convergence probably plays an schema.essential role in the elaboration of the information fromthis class of neurone. A tactile stimulus applied to a giventerritory, will activate a limited number of neurones, 3 . Potential plasticity of segmental excitatorynotably all those with receptive field centres situated processeswithin the territory. On the other hand, a nociceptivestimulus applied to the very same place will activate not It seems essential that multireceptive neurones should beonly these same neurones, but also others, which have the thought of in dynamic terms, given the fact that the sizesfringes of their receptive fields within the territory. It of their excitatory fields may change. The anatomicalfollows that, any noxious stimulus applied to a given convergence of inputs onto a single neurone is actuallysurface area, will activate a larger number of multirecep- larger than is observed functionally under normal con-tive neurones than any innocuous stimulus. Thus in order ditions. Such an imbalance signifies that, on fringes of theto understand the functional role of these neurones, one has excitatory field of a neurone, there is a zone the stimulationto think not only in simple terms of the activity of single of which triggers a cellular activation in the form of aneurones, but also in terms of the global volume of depolarisation. However, under physiological conditions,information that a neuronal population sends to the target this depolarisation is insufficient to elicit the actionstructures, notably in the brain [23,72,110]. potentials that are the only messages sent to higher centres.

Another important property of multireceptive neurones For instance, the cutaneous receptive fields of spinallies in their capability to be activated from different types neurones are smaller when natural stimuli are used forof territory [40,105]. For instance, they can also receive mapping than when electrical stimulation is used [89];signals from the viscera: one then refers to viscero-somatic synchronisation of inputs by electrical stimulation is likelyconvergence [16,39,65]. This property could be the func- to underlie such observations. The concept of ‘relativelytional and anatomical basis for referred pains, such as pain ineffective’ or ‘quiescent’ synapses developed progressive-in the left arm, testicle or right shoulder in cases of angina ly [112,139,143] on the basis of the idea that this neuronalpectoris, renal colic, or biliary colic, respectively. Finally, fringe, normally activated at a sub-liminal level, could, atit should be noted that nociceptive signals of muscular and least in some conditions, constitute a significant mecha-articular origin could also activate multireceptive neurones nism for information transfer. Thus, expanded receptive[2,93]. fields can be seen after transient repetitive electrical

Although knowledge of these convergence phenomena stimulation of C-fibres [18,24], injury [67,84], noxiousis sometimes very useful indeed for the diagnosis of some pinch [19,54,70,120], intradermal injection of capsaicindiseases, it is possible that their importance applies beyond [128] or noxious stimulation of viscera or muscle [20,62].such clinical curiosities. In this respect one must underline Interestingly, expanded receptive fields of multireceptivethe capacity of multireceptive neurones to capture all the neurones were also reported following non-noxious stimuliinformation coming from both the interfaces with the in non-anaesthetised sheep [54].external environment (the skin) and the internal milieu Such synapses can be made efficacious by partially(viscera, muscles, etc.). This set of information constitutes depolarising the recorded neurone. For example this can bea ‘basic somaesthetic activity’ for which a function has to done experimentally by applying an excitatory amino acidbe found. It is very possible that these neurones are not by micro-electrophoresis to the neurone’s membrane: oneexclusively involved in pain processes and, by transmitting then sees an increase, sometimes a considerable increase,such a ‘basic somaesthetic activity’, inform the brain that in the size of the neurone’s excitatory field [148]. Such anthe organism is suffering no special perturbation from increase is also observed when one blocks inhibitoryeither the external or the internal milieu. Such ‘basic mechanisms by pharmacological means, thus effectivelysomaesthetic activity’ could assist the specific lemniscal depolarising the neuronal membrane [28,114,115,126,somatosensory system in the elaboration of a ‘body 145,146].representation’, in the sense of a continual and global This demonstrates the potential plasticity of these sys-

D. Le Bars / Brain Research Reviews 40 (2002) 29–44 33

tems—a plasticity that could be revealed during the course reminiscent of the lateral inhibition described in otherof pathological processes, be they from peripheral or systems.central origins. In general terms, a sensitisation of excitat- That a contrasted picture is built by such neurones canory mechanisms or a deficiency of inhibitory mechanisms also be inferred from experiments on withdrawal reflexeswill produce both an increased activity in the neuronal (see the contribution by Schouenborg in this volume).population concerned with the painful focus and a growth Indeed, withdrawal reflex pathways receive powerful seg-of this population. The resulting information, which de- mental excitatory and inhibitory nociceptive inputs. Thesevelops in the spinal cord, is then transmitted to the brain, inputs are organised with a ‘modular’ pattern. Mostwhere it is decoded in the form of hyperalgesia. This can muscles in the limb—be they flexor, extensor or otherwisebe illustrated by the increased size of excitatory receptive (supinator, pronator)—can contract during nociceptivefields during acute or chronic inflammation [26,47,60,63, stimulation of a well-defined region of skin and thus each101,113,117]. muscle has its own ‘nociceptive cutaneous receptor field’

[121,124]. There is a nearly perfect match between thefields of cutaneous receptors and the cutaneous territorieswhich are removed from the nociceptive stimulus [122]. In

4 . Segmental inhibitory processes addition, nociceptive stimulation of other well-definedregions of skin can elicit powerful inhibition of the reflexes

In addition to having an excitatory receptive field, a in the same muscles [140]. The noxious stimulation of thismultireceptive neurone will frequently exhibit a cutaneous region prevents any inappropriate movement of the limbinhibitory receptive field which is often situated proximally towards the stimulus. It is very likely that multireceptive[5,56,136,138]. Most mechanical stimuli, notably those of neurones are involved in these processes.low intensity, when applied to this area, can inhibit theactivity of multireceptive neurones. An example is shownin Fig. 2, where non-noxious stimuli applied repetitively to 5 . Potential plasticity of segmental inhibitorythe inhibitory fields of these neurones were able to elicit processesstrong inhibitions of glutamate-induced firing. The activa-tion of large diameter cutaneous afferents can elicit strong Indeed, if one thinks in terms of neuronal populations,inhibitory effects on the responses of spinal neurones to one must take account of the mutual overlap of excitatorynociceptive stimuli [48]. It is generally thought that these and inhibitory fields of neighbouring neurones to under-phenomena are triggered only by the activation of Ab- stand the impact of the spatial organisation of the conver-fibres. This is most probably the case when vibratory gence in the elaboration of signals emanating from thisstimuli are applied to the skin [116]. However, the neuronal class [72,110]. The organisation of these inhib-activation of Ad-fibres results in the most powerful inhib- itory fields offers an explanation for how the application ofitions [21,22,79]. From such observations, one can infer multiple non-nociceptive stimuli to a large area of the bodythat the activation of a pool of Ad-fibre terminals will both surface (incorporating consequently numerous ‘centres’ ofactivate those neurones with excitatory receptive fields in excitatory fields), does not generate an ‘inaccurate’this particular area and inhibit those neurones with over- nociceptive signal, at least in non-pathological conditions.lapping inhibitory receptive fields. As described below, the Such an eventuality is prevented in healthy organisms byfinal output of the system constitutes a contrasted picture the concomitant stimulation of numerous inhibitory fields

Fig. 2. Rate meter record of a lumbar multireceptive neurone studied in a ‘spinal’ cat where firing was elicited both by continuous pinch of the excitatoryfield and by periodic electrophoretic injections of glutamate (using 10-s, 25-nA current pulses as indicated by the lower bars). Both synaptically mediated(about 40 spikes/s) and chemically induced firing were strongly depressed during the period of stimulation of the inhibitory field (IF, rapid repetitivestroking of the fur). Adapted from Ref. [5].

34 D. Le Bars / Brain Research Reviews 40 (2002) 29–44

that combine to attenuate the overall response of the 119,127,131] as well as to the trigeminal nuclei caudalisneuronal population. On the other hand, a deficit of such and oralis [14,25,30,61,92,96]. The neurones affectedinhibitory mechanisms will result in the triggering of a include some which project to the thalamus [29] and otherslarge amount of neuronal activity by harmless stimuli. which are involved in spinal reflexes [34,119]. It should beSuch information, elaborated in the spinal cord and then noted that the inhibitions triggered by heterotopic noxioustransmitted to the brain, could then be decoded in the form stimuli are highly sensitive to the anaesthetic regimen—anof ‘allodynia’ (pain triggered by a non-noxious stimulus). observation which could explain some reports of lesserA notable example is the ‘dynamic’ form of allodynia inhibitory effects [1,17,44,99,100,131].triggered by brushing the skin with cotton wool, which is a DNIC are triggered by conditioning stimuli applied tocharacteristic sign of some neuropathies. any part of the body distant from the excitatory receptive

Thus, the spatial organisation of the convergence of field of the neurone under study, provided that the stimuliexcitatory and inhibitory influences probably plays an are clearly noxious. There is a clear relationship betweenessential role in the elaboration of significant signals from the intensity of the conditioning stimulus and the strengththis class of neurone. One can then conceive that of the resultant DNIC [77,133]. With strong stimuli, thepathological processes may completely disrupt the equilib- inhibitory effects are powerful indeed and are followed byrium between excitatory and inhibitory influences, notably long-lasting post-stimulus effects, which can persist forwhen inhibitory controls are lacking. In addition, these several minutes. In some cases, the inhibitory effects canproperties may explain the hypoalgesic effects elicited by involve a complete abolition of activity for a long periodphysical stimulation (e.g., rubbing, electrical stimulation, of time following removal of the conditioning stimulietc.) of parts of the body close to the painful focus. The (‘switch-off’) although the activity can be restored tofact that activities in large and fine cutaneous afferent pre-conditioning levels by further manipulations of thefibres interact, has been known for a long time [85], but it excitatory receptive field (‘switch-on’) [11].is important to stress that this does not happen with the With regard to the viscera, some differences should beviscera, which are innervated exclusively by unmyelinated noted: visceral stimuli, e.g., distension of the colon orfibres. urinary bladder, generally produce inhibitions with slower

rates of onset and recovery but starting with intensitiesbelow what would constitute a painful level in man [12]. Itwas proposed that these differences may have reflected

6 . Propriospinal inhibitory processesdifferent amounts and patterns of activity in the relevantprimary afferent fibres rather than being due to different

In addition to the excitatory and inhibitory influences ofcentral neural mechanisms.

segmental origin discussed above, there is evidence thatIn any case, these data suggest that DNIC are triggered

multireceptive neurones are inhibited by nociceptivespecifically by the activation of peripheral receptors whose

stimuli applied to areas of the body remote from thesignals are carried by Ad- and C-fibres. In order to further

excitatory and inhibitory segmental fields. These effectsinvestigate the types of peripheral fibres that trigger DNIC,

involve propriospinal and supraspinal mechanisms. Theadvantage was taken of the facts that: (1) multireceptive

former can be observed in the dorsal horn of spinalneurones respond with relatively steady discharges to the

animals [13,37,44,83,130,136], but the underlying mecha-electrophoretic application of excitatory amino acids, and

nisms have not been subject to detailed investigation. One(2) DNIC act by a final postsynaptic inhibitory mechanism

of these, triggered by visceral afferents, involves the upperinvolving hyperpolarisation of the neuronal membrane, as

cervical cord [58,147]. The latter are observed in the intactshown in Fig. 3 [77,134,133]. When multireceptive neuro-

but not in the spinal animal, are termed diffuse noxiousnes were directly excited by the electrophoretic application

inhibitory controls (DNIC) and have been studied far moreof glutamate, the percutaneous application of single

extensively.square-wave, electrical stimuli to the tail always induced abiphasic depression of the resultant activity [7]. Both theearly and late components of this inhibition occurred with

7 . Diffuse noxious inhibitory controls shorter latencies when the base rather than the tip of thetail was stimulated. Such differences in latency were used

In the rat (see Refs. below), the mouse [38], the cat to estimate the mean conduction velocities of the peripher-[97,98] and probably the monkey [10,44], most mul- al fibres triggering the inhibitions, which fall into the Ad-tireceptive and some nociceptive-specific neurones can be and C-fibre ranges, respectively.strongly inhibited by noxious inputs applied outside their Such biphasic inhibitions could be elicited from any partreceptive fields. Such effects do not appear to be organised of the body and recorded from any multireceptive neuro-somatotopically but apply to the whole body. Conversely, nes, illustrating the notion of a whole body inhibitorythey apply to neurones in the dorsal horn of various receptive field for these neurones (Fig. 4). Indeed they cansegments of the spinal cord [11–14,57,74,96,99,100 be excited and inhibited from their segmental receptive

D. Le Bars / Brain Research Reviews 40 (2002) 29–44 35

Fig. 3. Example of the effect of various nociceptive conditioning stimuli on glutamate-induced activities of a multireceptive lumbar neurone recorded inthe rat. The 20-nA electrophoretic current was applied regularly (horizontal bars below recordings) every minute. Conditioning stimuli (arrows),as depictedon the insert drawing to the left, were pinch of the muzzle, the contralateral hind paw or the tail, or immersion of the tail in a 528C water bath (from top tobottom, respectively). Note that all these conditioning stimuli virtually blocked the glutamate-induced firing, with these inhibitions remainingfor severalminutes in most cases. Adapted from Ref. [134].

fields and inhibited by stimulation of Ad- and C-fibres in characteristics which suggest that they have a key rolethe remaining parts of the body. From now on, it is specifically in the processing of nociceptive informationnecessary to remember that, probably due to an increase in [135]. Indeed, they are preferentially or exclusively acti-the peripheral drive (see above), these controls are ex- vated by nociceptive stimuli from ‘whole-body’ receptiveacerbated during clinical pains, for example when an fields, they encode the intensity of cutaneous and visceralanimal suffers from monoarthritis, polyarthritis or a stimulation within noxious ranges and they are excitedperipheral mononeuropathy [15,26,27]. exclusively by activity in cutaneous Ad- or Ad- and C-

Since DNIC disappear in animals in which the spinal fibres. In addition, they send descending projectionscord has been sectioned [13,97], it is obvious that the through the dorso-lateral funiculus that terminate in theunderlying mechanisms involve the brain. The ascending dorsal horn at all rostro-caudal levels of the spinal cord.and descending limbs of this loop travel through the The fact that the supraspinal loop sustaining DNIC isventro-lateral and dorso-lateral funiculi, respectively confined within the most caudal part of the medulla was[77,133]. It has been proposed that DNIC result from a confirmed in a series of experiments where the potency ofphysiological activation of some of the brainstem struc- DNIC was tested in animals with complete sections attures that produce descending inhibition. Surprisingly, different levels of the brain stem [8] (see Fig. 6C).DNIC were not modified by lesions of the periaqueductalgrey (PAG), cuneiform nucleus, parabrachial area, locuscoeruleus/subcoeruleus or rostroventral medulla (RVM). 8 . Spatial summationBy contrast, lesions of subnucleus reticularis dorsalis(SRD) in the caudal medulla strongly reduced DNIC Very few studies have approached from a neuro-[77,133]. The SRD is located ventral to the cuneate physiological point of view, the role of spatial summationnucleus, between trigeminal nucleus caudalis and the in the integration of nociceptive information. Indeed,nucleus of the solitary tract and contains neurones with although one can conceive that the activity of a spinal

36 D. Le Bars / Brain Research Reviews 40 (2002) 29–44

Fig. 4. Example of heterotopic activation of Ad- and C-fibres triggering inhibitions in a spinal multireceptive neurone. (A) Schematic representation of theexperimental design. Recordings were made in the lumbar dorsal horn from a multireceptive neurone with a receptive field located on the extremity of theipsilateral hind paw. The continuous electrophoretic application of the excitatory amino acid,DL-homocysteic acid (DLH) induced a steady discharge fromthe neurone under study. The repetitive application of individual percutaneous electrical stimuli of adequate intensities to the contralateral muzzle (a), thebase (b) or the tip (c) of the tail induced biphasic depressions of the neuronal activity. (B) Peri-stimulus histograms (bin width: 5 ms) prepared during thecontinuous electrophoretic application (15 nA) of DLH onto the membrane of the neurone. The broken vertical white lines show the timing of percutaneouselectrical stimulation (10 mA; 2 ms duration; 0.66 Hz; 200 ms delay; 100 sweeps) of the contralateral muzzle (a), the base (b) or the tip (c) of the tail. Thebroken horizontal black line represents the mean firing calculated during the pre-stimulation control period (2200 to 0 ms). Two waves of inhibition can beseen. They occurred earlier when the base (b) as opposed to the tip (c) of the tail was stimulated. The time gaps are shown as grey areas between thehistograms, for both inhibitory components. The gaps were 7 and 25 ms for the beginning and end of the first component; they were 210 and 330 ms forthe beginning and end of the second component. Knowing that the distance between b and c was 100 mm, one can easily calculate the conductionvelocities of the fibres that elicited the first and second components: 4–14 and 0.3–0.5 m/s, respectively. These fibres therefore belong to the Ad- andC-fibre groups, respectively. Adapted from Ref. [71].

2neurone will increase as a nociceptive stimulus covers an stimulation of 1.9 and 4.8 cm were identical in the two2increasingly large part of its excitatory receptive field, the groups. On the other hand, beyond 4.8 cm , the frequency

responses evoked by stimulation of areas larger than that of discharge decreased significantly in the ‘intact’ animalsof the field have not been studied. Given that any but increased, albeit not significantly, in the ‘spinal’multireceptive neurone exhibits a whole-body receptive animals.field with an excitatory part surrounded by a large inhib- These experiments show that spatial summation triggersitory part (see above), this is an important question which inhibitory feed-back mechanisms onto multireceptive neu-needs to be addressed. rones when the surface stimulated reaches a critical area,

Multireceptive neurones, with excitatory fields of be- namely that of between two and three times the surface oftween one and three digits on the ipsilateral posterior paw, the excitatory field of the neurones under study. In order towere studied in a series of experiments during which clarify the relationship between inhibitions triggered byvarious lengths of the paw were immersed in a water bath spatial summation and the descending inhibitory systemsset at 488C [9,41,42]. Individual examples and the relation which had previously been identified as capable of modu-between the surface of stimulation and the amplitude of the lating the spinal transmission of nociceptive information,cellular responses recorded in ‘spinal’ and ‘intact’ rats are further experiments were performed. In these, the re-

2shown in Fig. 5. The levels of activity reached with sponses elicited by stimulation of 4.8 and 18 cm were

D. Le Bars / Brain Research Reviews 40 (2002) 29–44 37

Fig. 5. (A) Individual examples of the responses of a multireceptive neurone recorded in a ‘spinal’ rat. Each histogram represents the response elicited bythe immersion of one stimulus area (indicated at the top of the figure) for 15 s (horizontal bar) in a thermostatically controlled water bath set at 488C.

2Increasing the stimulus area from 1.9 to 7.5 cm induced a slight increase in the responses of the neurone; thereafter the discharge rate plateaued. (B)Individual examples of the responses of a multireceptive neurone recorded in an ‘intact’ rat. In this case, the profile of activity was quite different: the

2 2responses elicited by stimulating 1.9 and 4.8 cm were equivalent; the frequency of discharge then decreased moderately for stimulation of 7.5 cm and2more markedly for the stimulation of 18 cm . (C) Relationship between the mean firing of multireceptive neurones during the 15-s stimulation period

(ordinate) and the area stimulated (abscissa). In the spinal animals (hatched squares), there were no significant differences related to the size of the2stimulated area. In the intact animals (black squares): the responses elicited by immersion of 1.9 and 4.8 cm were not significantly different, whereas the

2responses evoked by immersion of 7.5 and 18 cm , were significantly smaller (*P,0.05; **P,0.001). (D) Identical results expressed as bar charts2 2representing the responses to stimulation of the whole paw (18 cm ) as percentages of the responses to stimulation of the digits alone (4.8 cm ) In this

representation, a value greater than 100% means that the response from stimulation of the large surface was higher, whereas a value lower than 100%means that it was the weaker. Adapted from Ref. [9].

studied in rats, which had previously undergone various the order of 50–60%; it included sham-operated controlssections of the brainstem. The animals were distributed in and the two groups of animals submitted to the mostsix groups on the basis of: four rostro-caudal levels of rostral sections of the brainstem. In the second class, thebrainstem sections (as confirmed histologically), a ‘sham’ ratio was in the order of 110–120%; it included the twoand a ‘spinal’ control group (Fig. 6A). The six groups of groups of animals submitted to the most caudal sections ofanimals could be separated into two classes on the basis of the brainstem and the ‘spinal’ animals. The anatomicalthe ratio of the responses elicited by the stimulation of 18 limit between these two classes of animals was a coronal

2and 4.8 cm (Fig. 6B). In the first class, the ratio was in plan about 500mm caudal to the caudal limit of the

38 D. Le Bars / Brain Research Reviews 40 (2002) 29–44

D. Le Bars / Brain Research Reviews 40 (2002) 29–44 39

´nucleus raphe magnus. During the course of these experi- studies were much greater than those used in the morements, the animals were also tested to investigate the recent ones.effects of a nociceptive heterotopic stimulus on the C-fibre The observation in human volunteers, that painfulresponses of the multireceptive neurone under study (Fig. thermal stimuli evoke pleasant warmth when applied to a6C). There was a remarkable overlap of the anatomical larger area of skin [86] and, more generally, the classicallimits defined by the functional criteria of either spatial lack of correlation between pain and the extents of lesionssummation or DNIC. in clinical situations, may also involve such mechanisms.

These results confirm that the inhibitory phenomena Thus, during World War II, Beecher made the observationtriggered by spatial summation result from the activation that many soldiers, who were severely wounded duringof control mechanisms of supraspinal origin. In addition, combat, exhibited a surprising delay in the onset of painthey demonstrate that the brain structures which are from their injuries [3]. Such observations, which have beeninvolved, are confined to the most caudal part of the confirmed more recently in civilian accident victims [87],medulla. These phenomena seem rather closely related to emphasised the lack of a one-to-one correspondenceDNIC both from the point of view of the implicated between the extent of an injury and the intensity of pain oranatomical structures and from the physiological con- suffering.ditions for their activation. From a more general point ofview, these results underline the pivotal role of the mostcaudal regions of the medulla in the control of the spinal 9 . Summary and conclusionstransmission of nociceptive signals, a role which hadpreviously been suggested on both electrophysiological When all is said, a nociceptive focus activates aand anatomical grounds [135]. segmental subset of multireceptive neurones and inhibits

On the basis of the present results, one can propose that the remaining population of such cells. We have seen thatthe feedback, which involves the activation of descending these neurones respond to non-nociceptive stimuli (pres-inhibitory controls and modulates the spinal transmission sure, rubbing, hair movements, etc.). Such neurones areof nociceptive signals, is dependent on the recruitment of a therefore activated in unpredictable but permanent ways bycritical number of spinal noci-responsive neurones. These all non-noxious stimuli provided by the environment. Thecontrols are different from the multiple modulatory con- resulting ‘basic somaesthetic activity’ when transmitted totrols originating from more rostral brainstem structures higher centres could constitute a ‘noise’, from which these[35,36,81]. Notably, the PAG–RVM system is not directly centres would have difficulty extracting a clear signal ofinvolved in such a negative feedback loop, as shown in the pain. DNIC could provide the filter which would allowanimals with sections which completely disconnected the such an extraction to be achieved. When a painful focusspinal cord from the RVM. In these animals, neither the occurs in a region of the body, the corresponding neuronesgeneral electrophysiological characteristic of the mul- in the dorsal horn actually send their excitatory signalstireceptive neurones nor DNIC were different from the toward higher centres. The latter activate DNIC to inhibitcontrol sham preparations. all those spinal and trigeminal multireceptive neurones that

It turns out that the spatial summation of nociceptive were not directly affected by the initial stimulus. Thisperipheral inputs results in both an increase in the number mechanism improves the signal /noise ratio by increasingof activated neurones and, beyond a critical level, a the contrast between the activities of the activated segmen-decrease in the responses of these neurones. These consid- tal pool of neurones and the silence of the remainingerations may explain some of the apparent discrepancies in population [72,75] (see also Ref. [6]). After all, this isreports of psychophysical investigations concerning the what happens within any large gathering: silence is neces-effects of spatial summation of nociceptive inputs on the sary if a contributor is to be heard. DNIC establish such aresulting pain sensation. Indeed, early studies concluded silence amongst spinal neurones. Conversely, a hubbub orthat there was no or little spatial summation for heat pain racket does not reduce the contributor to silence, but cover[45,49], whereas in more recent studies, a relationship was up his speech to the extent that one cannot hear him. Thisfound between the surface area stimulated and pain is what morphine does: indeed DNIC are extremely[31,68,80,111]. It is likely that such differences were due sensitive to low doses of morphine [76,141].mainly to the fact that the areas stimulated in the early If the ‘basic somaesthetic activity’ plays any role in

Fig. 6. (A) Drawing of a sagittal section of the brainstem [104], showing the rostral and caudal limits (shaded areas) of brainstem sections in four groups ofanimals. Abbreviations: 12, hypoglossal nucleus; AP, area postrema; CGM, central gray medial; Gi, gigantocellular nucleus; GiA, gigantocellularnucleuspars alpha; GiV, gigantocellular nucleus ventral; IO, inferior olive; PnV, pontine reticular nucleus; py, pyramidal tract; pyx, pyramidal decussation; RMg,nucleus raphe magnus; RPa, nucleus raphe pallidus; ROb, nucleus raphe obscursus; tz, trapezoid body. (B) Bar charts representing the responses evoked by

2 2noxious heat (488C, 15 s) stimulation of the whole paw (18 cm ) as percentages of the responses to stimulation of the digits alone (4.8 cm ) in sixdifferent groups of animals. (C) Bar charts illustrating the percentage inhibitions of C-fibre-evoked responses during conditioning stimulation (immersionof the tail in a 528C water bath). (**P,0.01, ***P,0.001 versus sham operated). Adapted from Ref. [41].

40 D. Le Bars / Brain Research Reviews 40 (2002) 29–44

elaborating or ‘dressing’ the body schema, then one could From an experimental point of view, the triggering ofconceive that the schema might be profoundly perturbed DNIC could change the properties of the segmentalby a painful focus, which would thus come to occupy an excitatory receptive field. Indeed to reach dorsal hornoversized and pre-eminent domain. It is appropriate here to neurones with micropipettes, a laminectomy is required.come back to visceral sensations, notably those triggered Any trauma to the bone, especially to the periosteum,by distension of hollow organs. Indeed one should re- constitutes a noxious focus. One can therefore confidentlymember that heterotopic inhibitory processes are triggered predict that such a surgical wound will trigger DNIC andin hollow organs by intensity levels lower than those be a source of descending inhibitory processes that couldwhich trigger pain. Interestingly, the urgent need to urinate change receptive field properties [82]. The possibility ofor defecate, that could be described as ‘imperious’ when it tuning withdrawal reflexes by an effect of DNIC on thecannot be accomplished, distorts one body schema to the balance between flexion and extension influences must alsoextent that one’s whole being is devoted to the physiologi- be considered [95,96].cal need. Thus answering ‘the call of nature’—an act In summary, multireceptive neurones do not constitute awhich is usually so discreet—is experienced not only as linear, single-channelled transmission network towards therelief, but often as a true pleasure when it allows re- brain. Nociceptive information is processed by complexconciliation with one’s body schema. mechanisms, which are able to modify the gain of the

If one accepts that multireceptive neurones have a role nociceptive system. The modulation of this gain is de-in nociception, then a second direct implication of the termined by the tri-dimensional characteristic of the pain-model relates to interactive phenomena between nocicep- ful focus—not only its intensity and duration but also itstive signals elicited from distant parts of the body and, surface area. Beyond a critical level, an increase in theconsequently, interactions between pains from distinct number of activated noci-responsive spinal neurones istopographic origins. Indeed, a noxious stimulus can dimin- associated with a decrease in their individual responses.ish or even mask a pain resulting from a focus situated on The functional consequences of such apparently opposinga remote part of the body. Such an observation goes back effects remains to be investigated during recordings fromto ancient times as illustrated by Hippocrate’s aphorism:if supraspinal target neurones [43]. In any case, it may bea patient be subject to two pains arising in different parts suggested that a complete analysis of nociceptive processesof the body simultaneously, the stronger blunts the other. must consider not only a large population of neurones butNumerous popular therapeutic methods for the alleviation also their mutual interactions. A volume transmissionof pain, some used spontaneously by patients, take advan- involving the whole cord must then be considered astage of this common clinical observation. They are often constituting the significant signal which is sent towards thegrouped together under the terms of ‘counter-irritation’ or brain centres which are responsible for the elaboration of‘counter-stimulation’ (see Ref. [77]). Combined psycho- pain. Indeed, the data described briefly above suggests thatphysical measurements and recordings of nociceptive the origin of painful sensations cannot be found in thereflexes (R ) in both normal volunteers and patients with simple increase of activity in neuronal populations of aIII

lesions in the spinal cord or brain strongly suggest the wired unidirectional system, which is specific in itsexistence of DNIC in man (see Refs. [77,73]). In short, the modality of activation. Rather, it might result from gra-following features are shared by dorsal horn multireceptive dients in the activities of different neuronal populations.neurones in the rat spinal cord and by the R reflex in Painful stimuli encountered in clinical practice are notIII

humans: (1) The R reflex and the responses of mul- punctuate and presumably involve a large number ofIII

tireceptive neurones to electrical stimulation of their excitatory receptive fields of peripheral fibres and centralcutaneous receptive fields are similarly inhibited by vari- neurones. Thus spatial summation is presumably involvedous heterotopic nociceptive stimuli applied on the skin. (2) in most clinical situations. When several foci coexist, oneThey are also both inhibited by visceral stimulation; in this can conceive that the global signal from multireceptivecase the inhibitions could be elicited with stimuli below neurones is polymorph with the resulting pain showingthe level which produces pain. (3) The extent of the fluctuations, instability or even versatility. The complexityinhibitions is directly related to the intensity of the of such a signal then depends on the multiplicity of mutualconditioning stimulus. (4) The inhibitions are followed by influences between the painful foci.after-effects, which can last for several minutes. (5) Theinhibitions are mediated by a spino–bulbo–spinal loop, theascending part of which is composed of the spino-reticulartract, and which has synaptic relays in the brainstem. (6) A cknowledgementsThe ascending pathways of the loop are mainly crossedwhile the descending pathways run ipsilaterally to the The author is very grateful to Dr. Cadden for advice inrecording site. (7) The inhibitions are blocked by low the preparation of the manuscript. This work was sup-doses of morphine. Such similarities allow one to conclude ported by L’Association pour la Recherche sur le Cancerthat the existence of DNIC in humans is more than likely. (ARC) and l’Institut UPSA de la Douleur (IUD).

D. Le Bars / Brain Research Reviews 40 (2002) 29–44 41

[22] J.M. Chung, K.H. Lee, Y. Hori, K. Endo, W.D. Willis, FactorsR eferencesinfluencing peripheral nerve stimulation produced inhibition ofprimate spinothalamic tract cells, Pain 19 (1984) 277–293.

´[1] G. Alarcon, F. Cervero, Effects of two anesthetic regimes on the [23] R.C. Coghill, D.J. Mayer, D.D. Price, The roles of spatial recruit-heterotopic inhibition of rat dorsal horn neurones, J. Physiol. ment and discharge frequency in spinal cord coding of pain: a(London) 416 (1989) 19P. combined electrophysiological and imaging investigation, Pain 53

[2] L. Arendt-Nielsen, P. Svensson, Referred muscle pain: basic and (1993) 295–309.clinical findings, Clin. J. Pain 17 (2001) 11–19. [24] A.J. Cook, C.J. Woolf, P.D. Wall, S.B. McMahon, Dynamic receptive

[3] H.K. Beecher, Pain in man wounded in battle, Ann. Surg. 123 field plasticity in rat spinal cord dorsal horn following C-primary(1946) 96–105. afferent input, Nature 325 (1987) 151–153.

[4] D.A. Bereiter, H. Hirata, J.W. Hu, Trigeminal subnucleus caudalis: [25] R. Dallel, P. Raboisson, A. Woda, B. Sessle, Properties of nocicep-tive and non-nociceptive neurons in trigeminal subnucleus oralis ofbeyond homologies with the spinal dorsal horn, Pain 88 (2000)the rat, Brain Res. 521 (1990) 95–106.221–224.

[26] N. Danziger, J. Weil-Fugazza, D. Le Bars, D. Bouhassira, Alteration[5] J.M. Besson, R.F.H. Catchlove, P. Feltz, D. Le Bars, Furtherof descending modulation of nociception during the course ofevidence for postsynaptic inhibitions on lamina V dorsal hornmonoarthritis in the rat, J. Neurosci. 15 (1999) 2394–2400.interneurons, Brain Res. 66 (1974) 531–536.

[27] N. Danziger, M. Gautron, D. Le Bars, D. Bouhassira, Activation of[6] D.C. Bolser, S.F. Hobbs, M.J. Chandler, R.D. Foreman, Inhibitorydiffuse noxious inhibitory controls (DNIC) in rats with an ex-effects of phrenic afferent fibers on primate lumbosacralperimental peripheral mononeuropathy, Pain 91 (2001) 287–296.spinothalamic tract neurons, Brain Res. 557 (1991) 162–166.

[28] D. Denny-Brown, N. Yanagisawa, The function of the descending[7] D. Bouhassira, D. Le Bars, L. Villanueva, Heterotopic activation ofroot of the fifth nerve, Brain 96 (1973) 783–814.Ad- and C- fibres triggers inhibition of trigeminal and spinal

[29] A.H. Dickenson, D. Le Bars, Diffuse noxious inhibitory controlsconvergent neurones in the rat, J. Physiol. (London) 389 (1987)(DNIC) involve trigeminothalamic and spinothalamic neurones in301–317.the rat, Exp. Brain Res. 49 (1983) 174–180.

[8] D. Bouhassira, D. Chitour, L. Villanueva, D. Le Bars, The spinal[30] A.H. Dickenson, D. Le Bars, J.M. Besson, Diffuse noxious inhib-

transmission of nociceptive information: modulation by the caudalitory controls (DNIC) Effects on trigeminal nucleus caudalis

medulla, Neuroscience 69 (1995) 931–938.neurones in the rat brain, Brain Res. 200 (1980) 293–305.

[9] D. Bouhassira, O. Gall, D. Chitour, D. Le Bars, Dorsal horn [31] D.K. Douglass, E. Carstens, L.R. Watkins, Spatial summation ofconvergent neurones: negative feed-back triggered by spatial sum- human thermal pain perception: comparison within and betweenmation of nociceptive afferents, Pain 62 (1995) 195–200. dermatomes, Pain 50 (1992) 197–202.

[10] T.J. Brennan, U.T. Oh, S.F. Hobbs, D.W. Garrison, R.D. Foreman, [32] R. Dubner, D.S. Hoffman, R.L. Hayes, Neuronal activity in medul-Urinary bladder and hindlimb afferent input inhibits activity of lary dorsal horn of awake monkeys trained in a thermal discrimina-primate T2-T5 spinothalamic tract neurons, J. Neurophysiol. 61 tion task. III. Task-related responses and their functional role, J.(1989) 573–588. Neurophysiol. 46 (1981) 444–464.

[11] S.W. Cadden, The ability of inhibitory controls to ‘switch-off’ [33] A.W. Duggan, B.T. Griersmith, P.M. Headley, J.B. Maher, The needactivity in dorsal horn convergent neurones in the rat, Brain Res. to control skin temperature when using radiant heat in test of628 (1993) 65–71. analgesia, Exp. Neurol. 61 (1978) 471–478.

[12] S.W. Cadden, J.F.B. Morrison, Effects of visceral distension on the [34] S. Falinower, C. Willer, J.L. Junien, D. Le Bars, A C-fibre reflexactivities of neurones receiving cutaneous inputs in the rat lumbar modulated by heterotopic somatic stimuli in the rat, J. Neurophysiol.dorsal horn: comparison with effects of remote noxious somatic 72 (1994) 194–213.stimuli, Brain Res. 558 (1991) 63–74. [35] H.L. Fields, Pain modulation: expectation, opioid analgesia and

[13] S.W. Cadden, L. Villanueva, D. Chitour, D. Le Bars, Depression of virtual pain, Prog. Brain Res. 122 (2000) 245–253.activities of dorsal horn convergent neurones by propriospinal [36] H.L. Fields, J.M. Besson (Eds.), Pain Modulation, Progress in Brainmechanisms triggered by noxious inputs: comparison with diffuse Research, Vol. 77, Elsevier, Amsterdam, 1988.noxious inhibitory controls (DNIC), Brain Res. 275 (1983) 1–11. [37] M. Fitzgerald, The contralateral input to the dorsal horn of the spinal

[14] B. Calvino, L. Villanueva, D. Le Bars, The heterotopic effects of cord in the decerebrate spinal rat, Brain Res. 236 (1982) 275–287.visceral pain: behavioural and electrophysiological approaches in the [38] A. Fleischmann, G. Urca, Clip-induced analgesia: noxious neckrat, Pain 20 (1984) 261–271. pinch suppresses spinal and mesencephalic neural responses to

[15] B. Calvino, L. Villanueva, D. Le Bars, Dorsal horn (convergent) noxious peripheral stimulation, Physiol. Behav. 46 (1989) 151–157.neurons in the intact anaesthetized arthritic rat. II Heterotopic [39] R.D. Foreman, Integration of viscerosomatic sensory input at theinhibitory influences, Pain 31 (1987) 359–379. spinal level, Prog. Brain Res. 122 (2000) 209–221.

[16] F. Cervero, J.M. Laird, Visceral pain, Lancet 353 (1999) 2145– [40] R.D. Foreman, R.F. Schmidt, W.D. Willis, Convergence of muscle2148. and cutaneous input onto primate spinothalamic tract neurons, Brain

[17] F. Cervero, A. Morales, Heterotopic noxious heating of the skin Res. 124 (1977) 555–560.inhibits dorsal-horn neurones in normal rats and in rats treated at [41] O. Gall, D. Bouhassira, D. Chitour, D. Le Bars, Involvement of thebirth with capsaicin, J. Physiol. (London) 398 (1988) 29P. caudal medulla in negative feed-back mechanisms triggered by

¨[18] F. Cervero, J. Shouenborg, B.H. Sjolund, P.J. Waddell, Cutaneous spatial summation of nociceptive inputs, J. Neurophysiol. 79 (1998)inputs to dorsal horn neurones in adult rats treated at birth with 304–311.capsaicin, Brain Res. 301 (1984) 47–57. [42] O. Gall, D. Bouhassira, D. Chitour, D. Le Bars, Effect of systemic

[19] F. Cervero, H.O. Handwerker, J. Laird, Prolonged noxious me- morphine on the responses of convergent neurons to noxious heatchanical stimulation of the rat’s tail: responses and encoding stimuli applied over graded surface areas, Anesthesiology 90 (1999)properties of dorsal horn neurones, J. Physiol. 404 (1988) 419–436. 1129–1136.

[20] F. Cervero, J.M. Laird, M.A. Pozo, Selective changes of receptive [43] O. Gall, L. Villanueva, D. Bouhassira, D. Le Bars, Spatial encodingfield properties of spinal nociceptive neurones induced by noxious properties of subnucleus reticularis dorsalis neurons in the ratvisceral stimulation in the cat, Pain 51 (1992) 335–342. medulla, Brain Res. 873 (2000) 131–134.

[21] J.M. Chung, Z.R. Fang, Y. Hori, K.H. Lee, W.D. Willis, Prolonged [44] K.D. Gerhart, R.P. Yezierski, G.J. Giesler Jr., W.D. Willis, Inhibitoryinhibition of primate spinothalamic tract cells by peripheral nerve receptive fields of primate spinothalamic tract cells, J. Neurophysiol.stimulation, Pain 19 (1984) 259–275. 46 (1981) 1309–1325.

42 D. Le Bars / Brain Research Reviews 40 (2002) 29–44

[45] L.C. Greene, J.D. Hardy, Spatial summation of pain, J. Appl. [66] D.R. Kenshalo, R.B. Leonard, J.M. Chung, W.D. Willis, Responsesof primate spinothalamic tract neurones to graded and to repeatedPhysiol. 13 (1958) 457–464.noxious heat stimuli, J. Neurophysiol. 42 (1979) 1370–1389.[46] M. Gregor, M. Zimmermann, Characteristics of spinal neurones

[67] D.R. Kenshalo, R.B. Leonard, J.M. Chung, W.D. Willis, Facilitationresponding to cutaneous myelinated and unmyelinated fibres, J.to the responses of primate spinothalamic cells to cold and tactilePhysiol. (London) 221 (1972) 555–576.stimuli by noxious heating of the skin, Pain 12 (1982) 141–152.[47] B.D. Grubb, R.U. Stiller, H.G. Schaible, Dynamic changes in the

[68] I. Kojo, A. Pertovaara, The effects of stimulus area and adaptationreceptive field properties of spinal cord neurons with ankle input intemperature on warm and heat pain thresholds in man, Int. J.rats with chronic unilateral inflammation in the ankle region, Exp.Neurosci. 32 (1987) 875–880.Brain Res. 92 (1993) 441–452.

[69] G.M. Kolmodin, C.R. Skoglund, Analysis of spinal interneurons[48] H.D. Handwerker, A. Iggo, M. Zimmerman, Segmental and sup-activated by tactile and niociceptive stimulation, Acta Physiol.raspinal actions on dorsal horn neurons responding to noxious andScand. 50 (1960) 337–355.non-noxious skin stimuli, Pain 1 (1975) 145–165.

[70] J.M. Laird, F. Cervero, A comparative study of the changes in[49] J.D. Hardy, H.G. Woolf, H. Goodel, in: Pain Sensations andreceptive-field properties of multireceptive and nocireceptive ratReaction, William and Wilkins, Baltimore, MD, 1952.dorsal horn neurons following noxious mechanical stimulation, J.

[50] R.L. Hayes, R. Dubner, D.S. Hoffman, Neuronal activity in medul-Neurophysiol. 62 (1989) 854–863.

lary dorsal horn of awake monkeys trained in a thermal discrimina-[71] D. Le Bars, The receptive fields of dorsal horn neurones: a simple

tion task. II. Behavioral modulation of responses to thermal andquestion, in: F. Cervero, G.J. Bennett, P.M. Headley (Eds.), Process-

mechanical stimuli, J. Neurophysiol. 46 (1981) 428–443.ing of Sensory Information in the Superficial Dorsal Horn of the

[51] H. Head, G. Holmes, Sensory disturbances from cerebral lesions, Spinal Cord. Nato ASI Series, Series A, Vol. 176, Plenum, NewBrain 34 (1911) 102–245. York, 1989, pp. 167–176.

[52] J.F. Herrero, P.M. Headley, Cutaneous responsiveness of lumbar [72] D. Le Bars, D. Chitour, Do convergent neurons in the spinal cordspinal neurons in awake and halothane-anesthetized sheep, J. discriminate nociceptive from non-nociceptive information?, Pain 17Neurophysiol. 74 (1995) 1549–1562. (1983) 1–19.

[53] J.F. Herrero, P.M. Headley, The dominant class of somatosensory [73] D. Le Bars, J.C. Willer, Pain modulation triggered by high-intensityneurone recorded in the spinal dorsal horn of awake sheep has wide stimulation: implication for acupuncture analgesia, in: A. Satohdynamic range properties, Pain 61 (1995) 133–138. (Ed.), International Congress Series 1238, Elsevier, Amsterdam,

[54] J.F. Herrero, P.M. Headley, Sensitization of spinal neurons by 2002.non-noxious stimuli in the awake but not anesthetized state, [74] D. Le Bars, A.H. Dickenson, J.M. Besson, Diffuse noxious inhib-Anesthesiology 82 (1995) 267–275. itory controls (DNIC): I Effects on dorsal horn convergent neurones

[55] J.F. Herrero, J.M. Laird, J.A. Lopez-Garcia, Wind-up of spinal cord in the rat, Pain 6 (1979) 283–304.neurones and pain sensation: much ado about something?, Prog.[75] D. Le Bars, A.H. Dickenson, J.M. Besson, Diffuse noxious inhib-Neurobiol. 61 (2000) 169–203. itory controls (DNIC): II. Lack of effect on non convergent

[56] P. Hillman, P.D. Wall, Inhibitory and excitatory factors influencing neurones, supraspinal involvement and theoretical implications, Painthe receptive fields of lamina 5 spinal cord cells, Exp. Brain Res. 9 6 (1979) 305–327.(1969) 284–306. [76] D. Le Bars, D. Bouhassira, L. Villanueva, Opioids and diffuse

[57] S.F. Hobbs, U.T. Oh, T.J. Brennan, M.J. Chandler, K.S. Kim, R.D. noxious inhibitory controls in the rat, in: B. Bromm, J.E. DesmedtForeman, Urinary bladder and hindlimb stimuli inhibit T1-T6 spinal (Eds.), ‘Pain and the Brain: From Nociceptor To Cortical Activity’,and spinoreticular cells, Am. J. Physiol. 258 (1990) R10–20. Advances in Pain Research and Therapy, Vol. 22, Raven Press, New

[58] S.F. Hobbs, U.T. Oh, M.J. Chandler, Q.G. Fu, D.C. Bolser, R.D. York, 1995, pp. 517–539.Foreman, Evidence that C1 and C2 propriospinal neurons mediate [77] D. Le Bars, L. Villanueva, D. Bouhassira, J.C. Willer, Diffusethe inhibitory effects of viscerosomatic spinal afferent input on noxious inhibitory controls (DNIC) in animals and in man, Patol.primate spinothalamic tract neurons, J. Neurophysiol. 67 (1992) Fiziol. Eksp. Ter. 4 (1992) 55–65.852–860. [78] D. Le Bars, M. Gozariu, S.W. Cadden, Animal models of nocicep-

[59] D.S. Hoffman, R. Dubner, R.L. Hayes, T.P. Medlin, Neuronal tion, Pharmacol. Rev. 53 (2001) 597–652.activity in medullary dorsal horn of awake monkeys trained in a [79] H.K. Lee, J.M. Chung, W.D. Willis, Inhibition of primatethermal discrimination task. I. Responses to innocuous and noxious spinothalamic tract cells by TENS, J. Neurosurg. 2 (1985) 276–287.thermal stimuli, J. Neurophysiol. 46 (1981) 409–427. [80] H. Machet-Pietropaoli, S. Chery-Croze, Spatial summation of

[60] U. Hoheisel, S. Mense, Long-term changes in discharge behaviour thermal pain in human beings, Sensory Processes 3 (1979) 183–187.of cat dorsal horn neurones following noxious stimulation of deep [81] P. Mason, Contributions of the medullary raphe and ventromedialtissues, Pain 36 (1989) 239–247. reticular region to pain modulation and other homeostatic functions,

[61] J.W. Hu, Response properties of nociceptive and non-nociceptive Annu. Rev. Neurosci. 24 (2001) 737–777.neurons in the rat’s trigeminal subnucleus caudalis (medullary dorsal [82] B. Matthews, Peripheral and central aspects of trigeminal nocicep-horn) related to cutaneous and deep craniofacial afferent stimulation tive systems, Phil. Trans. R. Soc. London B Biol. Sci. 308 (1985)and modulation by diffuse noxious inhibitory controls, Pain 41 313–324.(1990) 331–345. [83] S. McGaraughty, J.L. Henry, Effects of noxious hindpaw immersion

[62] J.W. Hu, B.J. Sessle, P. Raboisson, R. Dallel, A. Woda, Stimulation on evoked and spontaneous firing of contralateral convergent dorsalof craniofacial muscle afferents induces prolonged facilitatory horn neurons in both intact and spinalized rats, Brain Res. Bull. 43effects in trigeminal nociceptive brain-stem neurones, Pain 48 (1997) 263–267.(1992) 53–60. [84] S.B. McMahon, P.D. Wall, Receptive fields of rat lamina 1 projec-

[63] J.L. Hylden, R.L. Nahin, R.J. Traub, R. Dubner, Expansion of tion cells move to incorporate a nearby region of injury, Pain 19receptive fields of spinal lamina I projection neurons in rats with (1984) 235–247.unilateral adjuvant-induced inflammation: the contribution of dorsal [85] R. Melzack, P.D. Wall, Pain mechanisms: a new theory, Science 150horn mechanisms, Pain 37 (1989) 229–243. (1965) 971–979.

[64] R.R. Ji, C.J. Woolf, Neuronal plasticity and signal transduction in [86] R. Melzack, G. Rose, D. McGinty, Skin sensitivity to thermalnociceptive neurons: implications for the initiation and maintenance stimuli, Exp. Neurol. 6 (1962) 300–314.of pathological pain, Neurobiol. Dis. 8 (2001) 1–10. [87] R. Melzack, P.D. Wall, T.C. Ty, Acute pain in an emergency clinic:

[65] S.K. Joshi, G.F. Gebhart, Visceral pain, Curr. Rev. Pain 4 (2000) latency of onset and descriptor patterns related to different injuries,499–506. Pain 14 (1982) 33–44.

D. Le Bars / Brain Research Reviews 40 (2002) 29–44 43

[88] L.M. Mendell, Physiological properties of unmyelinated fiber pro- stimulation of monkey’s face, Neurophysiology 39 (1976) 936–jection to the spinal cord, Exp. Neurol. 16 (1966) 316–332. 953.

[89] L.M. Mendell, Modifiability of spinal synapses, Physiol. Rev. 64 [110] D.D. Price, R.L. Hayes, M.A. Ruda, R. Dubner, Spatial and(1984) 260–324. temporal transformations of input to spinothalamic tract neurons

and their relation to somatic sensations, J. Neurophysiol. 41 (1978)[90] L.M. Mendell, P.D. Wall, Response of single dorsal cord cells to933–947.peripheral cutaneous unmyelinated fibres, Nature 206 (1965) 97–99.

´ [111] D.D. Price, J.G. McHaffie, M.A. Larson, Spatial summation of[91] D. Menetrey, G.J. Giesler Jr., J.M. Besson, An analysis of responseheat-induced pain: influenced of stimulus area and spatial sepa-properties of spinal cord dorsal horn neurones to nonnoxious andration of stimuli on perceived pain sensation intensity and unpleas-noxious stimuli in the spinal rat, Exp. Brain Res. 27 (1977) 15–33.antness, J. Neurophysiol. 62 (1989) 1270–1279.[92] I.D. Meng, J.W. Hu, A.P. Benetti, D.A. Bereiter, Encoding of corneal

[112] L.M. Pubols, Electrical stimulation reveals relatively ineffectiveinput in two distinct regions of the spinal trigeminal nucleus in thesural nerve projections to dorsal horn neurons in the cat, Brain Res.rat: cutaneous receptive field properties, responses to thermal and371 (1986) 109–122.chemical stimulation, modulation by diffuse noxious inhibitory

controls, and projections to the parabrachial area, J. Neurophysiol. [113] K. Ren, J.L. Hylden, G.M. Williams, M.A. Ruda, R. Dubner, The77 (1997) 43–56. effects of a non-competitive NMDA receptor antagonist, MK-801,

on behavioral hyperalgesia and dorsal horn neuronal activity in rats[93] S. Mense, Nociception from skeletal muscle in relation to clinicalwith unilateral inflammation, Pain 50 (1992) 331–344.muscle pain, Pain 54 (1993) 241–289.

[114] N.E. Saade, N.R. Banna, A. Khoury, S.J. Jabbur, P.D. Wall,[94] M.M. Morgan, Direct comparison of heat-evoked activity ofCutaneous receptive field alterations induced by 4-aminopyridine,nociceptive neurons in the dorsal horn with the hindpaw withdrawalBrain Res. 232 (1982) 177–180.reflex in the rat, J. Neurophysiol. 79 (1998) 174–180.

[115] N.E. Saade, S.J. Jabbur, P.D. Wall, Effects of 4-aminopyridine,[95] M.M. Morgan, P.K. Whitney, Behavioral analysis of diffuse noxiousGABA and bicuculline on cutaneous receptive fields of cat dorsalinhibitory controls (DNIC): antinociception and escape reactions,horn neurons, Brain Res. 344 (1985) 356–359.Pain 66 (1996) 307–312.

[116] M.W. Salter, J.L. Henry, Differential responses of nociceptive vs[96] M.M. Morgan, M.M. Heinricher, H.L. Fields, Inhibition and facilita-non-nociceptive spinal dorsal horn neurones to cutaneously appliedtion of different nocifensor reflexes by spatially remote noxiousvibration in the cat, Pain 40 (1990) 311–322.stimuli, J. Neurophysiol. 72 (1994) 1152–1160.

[117] H.G. Schaible, R.F. Schmidt, W.D. Willis, Enhancement of the[97] C.R. Morton, B. Maisch, M. Zimmermann, Diffuse noxious inhib-responses of ascending tract cells in the cat spinal cord by acuteitory controls of lumbar spinal neurons involve a supraspinal loop ininflammation of the knee joint, Exp. Brain Res. 66 (1987) 489–the cat, Brain Res. 410 (1987) 347–352.499.[98] C.R. Morton, H.J. Du, H.M. Xiao, B. Maisch, M. Zimmermann,

[118] P. Schilder, The Image and Appearance of the Human Body, KeganInhibition of nociceptive responses of lumbar dorsal horn neuronesPaul, Trensh & Trubner, London, 1935, International Universitiesby remote noxious afferent stimulation in the cat, Pain 34 (1988)Press, New York, 1950.75–83.

[119] J. Schouenborg, A.H. Dickenson, The effects of a distant noxious[99] T.J. Ness, G.F. Gebhart, Interactions between visceral and cutaneousstimulation on A and C fibre-evoked flexion reflexes and neuronalnociception in the rat. I. Noxious cutaneous stimuli inhibit visceralactivity in the dorsal horn of the rat, Brain Res. 328 (1985) 23–32.nociceptive neurones and reflexes, J. Neurophysiol. 66 (1991) 20–

28. [120] J. Schouenborg, A. Dickenson, Long-lasting neuronal activity in ratdorsal horn evoked by impulses in cutaneous C fibres during[100] T.J. Ness, G.F. Gebhart, Interactions between visceral and cuta-noxious mechanical stimulation, Brain Res. 439 (1988) 56–63.neous nociception in the rat. II. Noxious visceral stimuli inhibit

¨cutaneous nociceptive neurones and reflexes, J. Neurophysiol. 66 [121] J. Schouenborg, J. Kalliomaki, Functional organization of the(1991) 29–39. nociceptive withdrawal reflexes I. Activation of hindlimb muscles

in the rat, Exp. Brain Res. 83 (1990) 67–78.[101] V. Neugebauer, H.G. Schaible, Evidence for a central component inthe sensitization of spinal neurons with joint input during develop- [122] J. Schouenborg, H.R. Weng, Sensorimotor transformation in ament of acute arthritis in cat’s knee, J. Neurophysiol. 64 (1990) spinal motor system, Exp. Brain Res. 100 (1994) 170–174.

¨299–311. [123] J. Schouenborg, B.H. Sjolund, Activity evoked by A- and C-[102] K. Nishioka, Y. Harada, L.M. Kitahata, S. Tsukahara, J.G. Collins, afferent fibers in rat dorsal horn neurons and its relation to a

Role of WDR neurons in a hind limb noxious heat evoked flexion flexion reflex, J. Neurophysiol. 50 (1983) 1108–1121.withdrawal reflex, Life Sci. 56 (1995) 485–489. [124] J. Schouenborg, H. Holmberg, H.R. Weng, Functional organization

[103] K. Ota, T. Yanagidani, K. Kishikawa, Y. Yamamori, J.G. Collins, of the nociceptive withdrawal reflexes II. Changes of excitabilityCutaneous responsiveness of lumbar spinal dorsal horn neurons is and receptive fields after spinalization in the rat, Exp. Brain Res.reduced by general anesthesia, an effect dependent in part on 90 (1992) 469–478.GABAA mechanisms, J. Neurophysiol. 80 (1998) 1383–1390. [125] J. Schouenborg, H.R. Weng, J. Kalliomaki, H. Holmberg, A survey

[104] G. Paxinos, C. Watson, in: The Rat Brain in Stereotaxic Coordi- of spinal dorsal horn neurones encoding the spatial organization ofnates, Academic Press, San Diego, CA, 1986. withdrawal reflexes in the rat, Exp. Brain Res. 106 (1995) 19–27.

[105] B. Pomeranz, P.D. Wall, W.V. Weber, Cord cells responding to fine [126] K. Semba, H.M. Geller, M.D. Egger, 4-Aminopyridine inducesmyelinated afferents from viscera, muscle and skin, J. Physiol. expansion of cutaneous receptive fields of dorsal horn cells, Brain(London) 199 (1968) 511–532. Res. 343 (1985) 398–402.

[106] D.D. Price, A.C. Browe, Spinal cord coding of graded nonnoxious [127] G.D. Sher, D. Mitchell, IntrathecalN-methyl-D-aspartate inducesand noxious temperature increases, Exp. Neurol. 48 (1975) 201– hyperexcitability in rat dorsal horn convergent neurones, Neurosci.221. Lett. 119 (1990) 199–202.

[107] D.D. Price, I.H. Wagman, Physiological roles of A and C fiber [128] D.A. Simone, T.K. Baumann, J.G. Collins, R.H. LaMotte, Sensiti-inputs to the spinal dorsal horn ofMacaca mulatta, Exp. Neurol. zation of cat dorsal horn neurons to innocuous mechanical stimula-29 (1970) 383–399. tion after intradermal injection of capsaicin, Brain Res. 486 (1989)

185–189.[108] D.D. Price, C.D. Hull, N.A. Buchwald, Intracellular responses ofdorsal horn cells to cutaneous and sural nerve A and C fiber [129] L.S. Sorkin, T.J. Morrow, K.L. Casey, Physiological identificationstimuli, Exp. Neurol. 33 (1971) 291–309. of afferent fibers and postsynaptic sensory neurons in the spinal

cord of the intact, awake cat, Exp. Neurol. 99 (1988) 412–427.[109] D.D. Price, R. Dubner, J.W. Hu, Trigeminothalamic neurons innucleus caudalis responsive to tactile, thermal, and nociceptive [130] J.E. Tattersall, F. Cervero, B.M. Lumb, Viscerosomatic neurons in

44 D. Le Bars / Brain Research Reviews 40 (2002) 29–44

the lower thoracic spinal cord of the cat: excitations and inhibitions [140] H.R. Weng, J. Schouenborg, Cutaneous inhibitory receptive fieldsevoked by splanchnic and somatic nerve volleys and by stimulation of withdrawal reflexes in the decerebrate spinal rat, J. Physiol.of brain stem nuclei, J. Neurophysiol. 56 (1986) 1411–1423. (London) 493 (1996) 253–265.

[131] R.W.W. Tomlinson, B.G. Gray, J.O. Dostrovsky, Inhibition of rat [141] J.C. Willer, D. Le Bars, Electrophysiological studies on morphinespinal cord dorsal horn neurons by non-segmental, noxious cuta- analgesia in man, in: B. Bromm, J.E. Desmedt (Eds.), Pain and theneous stimuli, Brain Res. 279 (1983) 291–294. Brain: From Nociception To Cognition. Advances in Pain Research

[132] R.D. Treede, R.A. Meyer, S.N. Raja, J.N. Campbell, Peripheral and and Therapy, Vol. 22, Raven Press, New York, 1995, pp. 541–557.central mechanisms of cutaneous hyperalgesia, Prog. Neurobiol. 38 [142] W.D. Willis, R.E. Coggeshall, in: Sensory Mechanisms of the(1992) 397–421. Spinal Cord, Plenum, New York, 1991.

[133] L. Villanueva, D. Le Bars, The activation of bulbo-spinal controls [143] C.J. Woolf, A.E. King, Subthreshold components of the cutaneousby peripheral nociceptive inputs: diffuse noxious inhibitory con- mechanoreceptive fields of dorsal horn neurons in the rat lumbartrols, Biol. Res. 28 (1995) 113–125. spinal cord, J. Neurophysiol. 62 (1989) 907–916.

[134] L. Villanueva, S.W. Cadden, D. Le Bars, Evidence that diffuse[144] C.J. Woolf, M.W. Salter, Neuronal plasticity: increasing the gain in

noxious inhibitory controls (DNIC) are mediated by a final post-pain, Science 288 (2000) 1765–1769.

synaptic inhibitory mechanism, Brain Res. 298 (1984) 67–74.[145] T. Yokota, Y. Nishikawa, Actions of picrotoxin upon trigeminal[135] L. Villanueva, D. Bouhassira, D. Le Bars, The medullary subnu-

subnucleus caudalis neurons in the monkey, Brain Res. 171 (1979)cleus reticularis dorsalis (SRD) as a key link in both the transmis-369–373.sion and modulation of pain signals, Pain 67 (1996) 231–240.

[146] T. Yokota, N. Nishikawa, Y. Nishikawa, Effects of strychnine upon[136] I.H. Wagman, D.D. Price, Responses of dorsal horn cells ofM.different classes of trigeminal subnucleus caudalis neurons, BrainMulatta to cutaneous and sural A and C fiber stimuli, J. Neuro-Res. 168 (1979) 430–434.physiol. 32 (1969) 803–817.

[147] J. Zhang, M.J. Chandler, R.D. Foreman, Thoracic visceral inputs[137] P.D. Wall, Cord cells responding to touch, damage and temperatureuse upper cervical segments to inhibit lumbar spinal neurons inofthe skin, J. Neurophysiol. 23 (1960) 197–210.rats, Brain Res. 709 (1996) 337–342.[138] P.D. Wall, The laminar organisation of dorsal horn and effects of

¨descending impulses, J. Physiol. 188 (1967) 403–423. [148] W. Zieglgansberger, A. Herz, Changes of cutaneous receptive fields[139] P.D. Wall, The presence of ineffective synapses and the circum- of spino-cervical tract neurones and other dorsal horn neurones by

stances which unmask them, Phil. Trans. R. Soc. London Ser. B microelectrophoretically administered aminoacids, Exp. Brain Res.278 (1977) 361–372. 13 (1971) 111–126.