Download Neurophysiology of Pain Perception: Convergence of Noxious Stimuli and Role of C Fibers and more Study notes Physiology in PDF only on Docsity!
ACTA NEUROBIOL. EXP. 1975, 35: 715- Memorial Paper in Honor of Jerzy Kono,rski
RECENT ADVANCES ON THE NEUROPHYSIOLOGICAL BASES
OF PAIN SENSATION 1
Denise ALBE-FESSARD and Alfred FESSARD Laboratoire de Physiologie des Centres nerveux Universitk Pierre et Marie Curie et Institut Marey, Paris, France
Our paper aims at presenting a rapid survey of the most recent advances on the neuroanatomical and neurophysiological bases of pain sensation, with a relative emphasis given to the work done i n our labo- ratory.
- Specificity of pain receptors as evidenced by that of their peripheral fibers
Is pain a specific sensation? This question introduces a discussion which started a century ago between Von Frey (1895) and Goldscheider (1898) and has remained a question to the present: do pain messages have, like the other sensory messages, their specific receptors, pathways, cent- ers; or, on the contrary, is pain a quantitative component of specific sensations? We shall see that, even if some of the peripheral properties of afferent fibers plead in favor of the specific hypothesis, it is certain that a great deal of psychophysiologica1 observations and of experiments per- formed on the relay centers of pain messages cannot be interpreted in the simple terms of a strict specificity. On the other hand, receptors inducing pain have not themselves been the object of recent attempts for their de-
1 The present article is. dedicated to the memory of Jerzy Konorski with whom we had kept up 25 years of constant friendship, associated with a high appreciation of his pioneer work and original ideas. These were devoted, his life long, to a better understanding of the various conditioned reflexes and of The in-
tegrative activity of the brain - his latest important book (1967). This is why we
like to think that our friend would have had more than one reason to pay attention to the following pages.
716 D. ALBE-FESSARD AND A. FESSARD
finite identification, so that their precise anatomical substrates are still un- known. On the contrary, a better knowledge has been acquired about the type of fibers connected with these unidentified receptors: fibers of the myelinated group (A) and those of the non-myelinated one (C) have been found which are activated by stimuli called "noxious". Sherrington was the first to use the term "nociceptive" (1906). He suggested that all painful peripheral stimuli have in common the possi- bility of being a threat to the integrity of the skin tissues. From this time on, the term "noxious" has been used extensively to designate any stimulus which, applied to an animal, is capable of giving rise to a mwsage generally felt as painful by human subjects. We must how- ever keep in mind that certain of these "noxious" stimuli, like pricking, pinching, even if unpleasant or even very painful, are not necessarily harmful. Thus, when we use here the term noxious, it is not exactly with the same meaning as that proposed by Sherrington. The term "noxious" will be used here, as elsewhere to qualify nerve impulses and messages engendered in animals by pain-inducing stimuli i n man. Different groups of workers, (see references in Burgess and Iggo 1974, Per1 1973) using microphysiological techniques, have recently searched for the fibers activated by noxious stimuli in cats and in mon- keys. In both animals, nerve fibers specific for certain types of stimula- tion (tactile, articular, etc.) had since long been identified. In general: when the stimulus has been strongly intensified so that when applied to the experimenter it becomes painful to him these specific fiber res- ponses do not increase and thus do not convey noxious messages. On the contrary, two groups of fibers are certainly connected with noci- ceptors: a former group which first responds to a light non pain-produ- cing stimulus, but in which stronger stimulations, then called "noxious" and experienced as painful by man, enhance its activity; a second group, which responds only when a noxious stimulus is applied. The noxious stimulus can be of one type only, or on the contrary may correspond to different noxious agents (burning, pinching, etc.); in this latter case these fibers are said to be connected to polymodal nociceptors. Fibers transmitting noxious impulses are present in both the mye- linated and non-myelinated groups. In the A group, they are found in all the diameter ranges, but they are more numerous in the gamma-delta range. In primates, afferent C fibers seem essentially involved in the conduction of nociceptive afferents. In man, pain involves activity of fibers of the A and C groups as was demonstrated with gross nerve recordings by Landau and Bishop (1958). Recently, these facts have been confirmed in human subjects by two groups of workers who ob- served the psychophysical correlates of the A and C fiber activities. The nerve responses were recorded in normal subjects by means of a mi-
718 D.^ ALBE-FESSARD^ AND^ A.^ FESSARD
Fig. 1. At right: Schema of the Rexed laminae a t lumbar level (cat). Stippled zones: areas where cells responding to noxious messages have been found. At left, upper part: Neuronal organization of the dorsal horn from a drawing by Cajal (1909). Note the dense arborizations of the collateral endings of myelinated fibers (F) and the large cells in deeper layers (a); their axons are going to form the ascending pathways. At left, lower part: Same region schematized by Rethelyi and Szentagothai (1973). Levels of termination of A and C fibers are illustrated together with two cells, one being an intra-dorsal-horn neuron, the other a cell with an ascending axon. Descending terminations (DES) are also represented.
has shown in the cat that it is possible to distinguish different layers within the cord, Fig. 1 at right gives an example from the lumbar Re- xed maps, the equivalent of the dorsal horn illustrations in the Cajal anatomical description. In order to go farther than the patterns of connection, and having in mind to locate the cells the neurons of which receive nociceptive af-
NEUROPHYSIOLOGICAL BASES OF PAIN SENSATION i 19
ferents, an electrophysiological exploration with microelectrodes had to be performed. This was done by several workers using different tech- niques of stimulation:
- Stimulation of nerve fibers of the A and C groups;
- Natural stimulation by noxious procedures (pricking, pinching, burning, deep pressure);
- Injection of bradykinin into a proximal artery of the peripheral field, as it was first proposed by Lim (1968). With these different techniques, it was demonstrated that in the cat and the monkey noxious stimuli activate cells of layer I (Christensen and Per1 1970), essentially those in layer V (Wall 1960) and also those in part of layers VII and VIII (Trevino et al. 1971, 1973, Levante and Albe-Fessard 1972). However, cells in layer V at least have a peculiarity the layers I and VIII cells do not have: light as well as strong tactile stimulations activate these cells. Consequently a convergence of noxious and non-noxious messages is realized in layer V at unitary level. More-
Natural stimulations Brady kinin
A h c i ~ smovements
- - n A
U
pressure
B
5 0
Fig. 2. Comparison of properties of tvpical cells in layer IV ( A ) and layer V ( B J in the cat (reorganized from Btsson et al. 1972, 1973). At left: A, receptive peripheral field and type of response during a light tactile contact on a toe; note t h e on-off burst produced. B, I n the receptive field, note that t h e center is sensitive to a light touch, and that t h e peripheral field is activated only by stronger stimuli; below, tonic response of the cell to pinching. At right: Comparison of the actions of a n intra-arterial injection of Bradykinin in t h e proximity of the receptive fields of the two preceding types of cells. Note clear response of lamina V cell as opposed to doubtfui response in lamina IV.
NEUROPHYSIOLOGICAL BASES OF PATN SENSATION
C A T MONKEY
Horizontal plane
Fig. 3. Schematic outlines of horizontal sections in the thalamus of cat and monkey. They summarize the results of Boivie (1972-1974) obtained with degeneration tech- niques after lesions of anterolateral pathway sparing the Morin bundle. Regions displaying degenerated terminals are represented by dots. Black areas are regions of dense terminations; note their absence in cat's VPL.
work of J. Boivie (personal communication) has brought out good evi- dence of this being so. It is illustrated in the schema of Fig. 3 which shows the thalamic terminal sites of the anterolateral pathways in cats and monkeys.
4. Cells of origin of t h e spino-thalamic bundles
Several investigations have been made recently by anatomists and physiologists to locate cells of origin corresponding to the different spi- no-thalamic bundles. Those connected to layer I or V cells have received particular attention. In the cat, cells of origin have been searched for by means of the antidromic technique. Only very few cells of the dorsal horn layers V and VI of the cervical enlargement have been found which send their impulses to the lemniscus (Dilly et al. 1968). Trevino et al. (1971) located in layers VII and VIII some cells connected to the thalamic posterior area. We have ourselves practically found no dorsal horn cells connected without relay to the thalamus. The only direct connection we found was between layers VII-VIII cells, reticular formation and medial tha- lamus (Levante and Albe-Fessard 1972). The Morin bundle is the only important thalamic ascending pathway having its origin in layers IV and V (see ref. in Brown 1973).
722 D. ALBE-FESSARD^ AND^ A. FESSARD
In the monkey, the organization is different. Using antidromic acti- vation from the thalamic VP regions, Willis et al. (1974). Levante et al. (1973), Albe-Fessard et al. (1974ab), discovered a direct connection be- tween dorsal horn cells and contralateral VP. We have until now stu- died, in 14 macacas a total of 56 cells antidromically activated from dif- ferent places in VP. The majority of these cells were found in the me- dial dorsal horn (Fig. 4) and they had the characteristics of layer V cells in the cat. A pathway coming from surrounding layers, in parti- cular from layer I has also been found by Trevino et al. (1973). The layer V cell type which we had found connected to the thala- mus had a peripheral field whose center responded to light touch where- as its surroundings responded to pricking and, more lastingly, t o pin- ching. On the other hand, 21 of these same cells, antidromically activa- ted, were tested for their responses to injections of Bradykinin: 18 res- ponded (Fig. 5) (Levante et al. 1975). When a short and well-localized electrical stimulation was applied
Fig. 5. At left, spinal cord hemi-section showing (surrounded with a circle) the locus of the injected dye substance applied to a cell of the dorsal horn. At right: Anti- dromic spikes provoked by stimulation of VP. From top to bottom, note its stable latency, the collision of the antidromic spike with an orthodromic one, the rapid rate of stimulation which the antidromic response can follow. Lower part, frequency of the fluctuating spontaneous activity and its modification by a Bradykinin intra- arterial injection (from Levante et al. 1975b).
Fig. 6. Responses of a lamina V cell type of the dorsal horn in a monkey, to an increased electrical stimulation of the receptive field. From top to bottom, intensity of the same type of stimulus is increased, which results in eliciting bursts of spikes of progressively longer latencies. Note under each burst the peripheral speed of t h e fibers involved (inferior tracing).
bodykinin activated o un-modifled a inhibited
Bradylcinin @activated g u n modified A inhibited
Fig. 7. Thalamic sites of cells which had their activities modified by Bradykinin intra-arterial injection a t the level of their receptive fields. Lightly anesthetized cats. a: Data^ presented^ in^ different^ anterior^ stereotaxic^ planes^ (nomenclature^ from^ Rinvik 1968). Note that the lna~orityof Bradykinin activated cells are a t the limit of VP, in PO or suprageniculate (from Guilbaud et al. 1974). + ts b: Data presented in the CM-Pf region (after the Ajmone-Marsan atlas). Note activation or inhibition by Brady- kinin injection (from Conseiller et al. 1972).
~L‘EUROPHYSIOLOGIC 11, BASES OF PAIN SENSATION 725
Only projections to V P have been studied until now. We found cel- lular activities resembling those of layer V cells of the spinal cord. They have a cutaneous peripheral field sensitive at its center to light pres-
7 2 6 D. ALBE-FESSARD^ AND A.^ FESSARD
sure and at its periphery to pinching or pricking (Fig. 8); until now, histological control has revealed in two monkeys only a total section of both dorsal column and Morin bundle. In these animals, 50 cells could be activated by means of various stimulations applied to the par- tially deafferented posterior limb. The result was: 7 cells activated by weak tactile pressures, 37 by pink-prick orland pinching, 6 by passive movements (Levante et al. 1975~). These data suggest thzt in primates, the^ neospinothalamic^ bundle conducts messages engendering a relatively well localized pain, prece- ded by a non-painful component, qualified as tactile. Our latency values correspond to impulse velocities (see Fig. 8) in the range of those of myelinated fibers which constitute this bundle (1-100 mlsec modal value around 60 mlsec as measured by the antidromic technique, Albe-Fes- sard et al. 1 9 7 4 ~ ). With regard to the paleo-spinothalamic tract and the spinoreticu- lo-thalamic component, origin and role are not so clear. The wider dis- persion of the antidromically activated cells as found by Trevino et al. (1973), Willis et al. (1974) is probably due to the fact that these authors stimulated the part of thalamus posterior to VP, that is PO, and possibly the reticular formation. In fact, inasmuch as pain perception is concer- ned, the role of the projection to medial thalamus through the antero- lateral bundle remains to be established in the monkey.
Fig. 9. Schematic representation summarizing the various spino-thalamic relayed and non-relayed pathways of nociceptive afferents (interrupted lines), with their spinal cells of origin represented in a section of the lumbar spinal cord (the non- relayed dorsal column afferent system is not represented). continuous lines sche- matize the known controls which can act upon these pathways and show their origins and the levels where they exert their action. The conjectural projections to the cortex of the thalamic cells driven by nociceptive messages through relayed spinal pathways are shown with a question mark. Abbreviations: A (0-@, messages through myelinated fibers; C, through nonmyelinated fibers; Cerv. lat., n. cervical lateralis of the spinal cord; GB, Go11 and Burdach nuclei; RF, reticular formation; GC, grisea centralis; v p , ventralis posterior; P o , posterior oral group of^ nuclei; Pf, CM, CL, MD, Parafascicularis nucleus, part cf centre median, Centralis lateralis nucleus, part of medial dorsal nucleus; Mot., motor cortex; C Orb., orbital cortex; Pyr. tract, pyramidal tract.
728 D.^ ALBE-FESSARD^ AND^ A.^ FESSARD
circulatory changes under the control of sympathetic efferents. It is a n old assumption, frequently forgotten but which now seems to regain in- terest. However, it is generally at synaptic levels, in the relays involved, that these controls have been analyzed, with the help of various tech- niques. A recent neurological study on macacas (Denny-Brown et al. 1973) has brought out new evidence that a n afferent message initially travel- ling through a dorsal root can be modulated by actions originating in the neighboring root areas. These modulations are either inhibitory or facilitatory. At spinal level, they are conveyed by the Lissauer bundle, a propriospinal tract which connects neighboring segments (only a small part of it contains C fiber primary afferents). Its lateral part conducts inhibitory impulses, its medial part facilitatory ones. The inhibitory ef- fect can be reversed by strychnine, which makes probable a post-syna- ptic site of action. The same authors have described similar interactions in relation to the afferents from the face, in particular between trige- minal and cervical afferents (Denny-Brown and Yanagisawa 1973). An inhibitory modulation originating in proximal segments has also been demonstrated by the electrophysiological signs associated with eith- e r pre- or post-synaptic processes. The negative dorsal root potential (DRP) is generally considered as the sign of an axono-axonic depolariza- tion of the afferent terminals, this being the cause of a "pre-synaptic inhibition". However, a recent suggestion has been made that it would be produced by and accumulation of Kf ions (KrnjeviE and Morris 1974, Vyklicky et al. 1974). On the other hand, recordings in layer V cells activated by glutamate iontophoresis have shown that the inhibitory ef- fect elicited by stimulation of an inhibitory peripheral field is at least in part of post-synaptic origin (Besson et al. 1974). Facilitation has also been produced in layer V cells following activa- tion of fine myelinated fibers and of C fibers (Hillman and Wall 1969). The correlation between these facilitations and a positive DRP elicited by a n activation of C fibers (Mendell and Wall 1964) is still a matter of discussion. It is often assumed that this positive potential and the resul- ting facilitation might be due to the supression of a tonic inhibitory in- fluence (see references in Schmidt 1973) and not to a presynaptic hyper- polarization. Of major importance to our present subject is the considerable amount of work that has been devoted to the systematic study of the descending controls exerted on layers IV and V cells; a study still pur- sued with the purpose of locating the origin of this control in the brain and of identifying their descending pathways. Stimulation of cortical areas gives rise to dorsal root potentials
NEUROPHYSIOLOGICAL BASES OF PAIN SENSATION 729
(DRP), but these potentials may be the sign of various actions on cells subserving different roles. For the purpose of the present report, inhibi- tion of responses to noxious stimulations is the one that has retained our interest. According to our experimental procedures, this inhibition displayed itself through either layer V cell activities or behavioral reac- tions. Layers IV and V cells can be inhibited from the cortex by way of the pyramidal tract (Fetz 1968). But these descending effects (tested on Layer V cells) are more marked on the early components of evoked activities due to noxious stimuli than on the late ones (Coulter et al. 1974). The same observation was reported by Maillard et al. (1971) after stimulation of the orbital cortex (Wyon-Maillard 1972). These results lead us to think that the cortical inhibitory controls would not be very efficient on noxious afferents, but on the contrary would act upon the messages arising from lighter stimuli. However, other descending inhibitory controls have been disclosed. They come from regions situated in the brainstem (Carpenter et al. 1962, 1966; Hugo and Jankowska 1967). Their existence was clearly demon- strated by the considerable increase of spontaneous and provoked layer V cell activities during a reversible spinalization (cooling method) (see Fig. 10, and Besson et al. 1975). One of these descending inhibitory effects has its origin in the gri- sea centralis region surrounding the Raphe nuclei. Behavioral obser- vations on rats first provided evidence that such a control is exerted from this level (Reynolds 1969, and Mayer et al. 1971). Liebeskind with some of our group (Liebeskind et al. 1973) applied the same technique to cats, using strong pinching or pulpar dental stimulation, and imme- diately searching for the maximal pain-amending effect. This was reco- gnized when the electrode was implanted a t the level of either the an- terior or the posterior Raphe nuclei (Liebeskind et al. 1973; Oliveras et al. 1975) (Fig. 11). Following these observations, the same animals were tested in acute experiments, during which layer V cells were activated by pinching. It was then shown that stimulation of the same Raphe nuclei regions had a strong inhibitory effect on activation by noxious messages while this effect was not so clear when light stroking was used to activate the cell. From these convergent results coming from different laboratories it can be concluded that the inhibitory controls exerted upon the inte- grating activities of layer V cells, either by cortical or by mesencephalic centers, cannot be confounded: the cortical control bears upon activities elicited by light specific stimuli, the mesencephalic one attenuates or cancels the late effects of messages produced by pinching, pricking or Bradykinin injection (Besson et al. 1974) (Fig. 12).
NEUROPHYISIOLOGICAL BASES OF PAIN SENSATION
Fig. 11. Different regions of the grisea centralis of the cat from which stimulation provokes an inhibiticn of reactions to noxious stimuli in the cat. Note the proxi- mity of the stimulated efficacious area to the different Raphe nuclei. (From Oliveras et al. 1975.)
- Effects of Raphe stimulation compared to morphine action
The effects of Raphe stimulation seem to be related to the action of morphine. Takagi, in a pioneer work (Takagi et al. 1955) suggested that this drug was acting through a local action on the cells of Raphe nuclei. Herz et al. (1970) relate experiments in which behavioral reactions to dental stimulations were inhibited by morphine injection into the IVth ventricle. Samanin and al. (1970), Samanin and Valzelli (1971) reduce or enhance the analgesic action of morphine by destruction or by stimula- tion, respectively, of the Raphe nuclei. Akil et al. (1972) showed in rats that the analgesia elicited by stimulation of the Raphe nuclei is anta- gonized by Naloxone, an antagonist of morphine. Similarly, analgesia produced in cats by stimulation of posterior Raphe is suppressed by Nalorphine.
732 D.^ ALBE-FESSARD^ AND^ A.^ FESSARD
An intermediary serotoninergic pathway seems to intervene in the action of morphine as well as in that of Raphe stimulation. The effect of this stimulation was reduced by parachlorophenylalamine. (Akil et al. 1972) and, in acute preparations, the inhibition of layer V cells is suppressed by LSD, while other types of inhibition were not influenced (Guilbaud et al. 1973). In a recent work, Marthe Vogt (1974) has shown that experimental depletion in serotonin within the spinal cord and the brain stem reduces the morphine analgesia. Thus, the hypothesis can be advanced that morphine would act by the mediation of a descending pathway having serotonin as a mediator. The exact point of impact of the drug is however far from being established. Another aspect of its action, puzzling enough at first sight, has been made clearer recently. It is the fact that morphine can have an action on a spinal animal while its action Is concealed in decerebrate ones. In fact, in these latter, it is as if a strong descending inhibition was con- stantly blocking the noxious messages, so strongly that it is useless to add to it. Undoubtedly this explains certain discrepancies existing in the literature. This phenomenon was clearly exemplified in our group by Le Bars (1974) when he used Bradykinin to activate layer V cells (Fig. 12).^ In decerebrate cats, their spontaneous activity was reduced and Bradyki- nin injection had no effect. If a reversible spinalization had been pro- duced by cooling the cervical cord, then spontaneous activity was incre- ased and the stimulating effect of Bradykinin reappeared. One can understand how under these conditions the action of mor- phine is affected, given the fact that morphine inhibition is exerted through the same pathways as those which mediate a normal descending control. In decerebrate preparations where this control seems to be at its highest point, morphine cannot increase this inhibition and seems to be without effect. Consequently, painful reactions cannot be but difficult to study in decerebrate preparations. The inhibitory controls exerted at spinal levels on noxious messages are not the only ones. Others have been described which act at reticular and thalamic levels, originating from cortical primary areas and corpus striatum. They have not been analyzed in detail as the preceding ones with regard to their action on noxious messages. A study of the corres- ponding relays has now to be undertaken with this purpose in mind. Finally, a question of major importance for our problems remains open to further research; it is that of the cortical projections from the cell type which receives noxious messages at thalamic level. Little is known on this point (see Fig. 9). However, now that the thalamic cells
