Exp Brain Res (1992) 90:384-392
Experimental BrainRe search 9 Springer-Veflag 1992
Corticofugal actions on lemniscal neurons of the cuneate, gracile and lateral cervical nuclei of the cat J.D. Cole* and G. Gordon University Laboratory of Physiology, Oxford OX1 3PT, United Kingdom Received April 15, 1991 / Accepted March 19, 1992
Summary. Extracellular records were made from single identified lemniscal neurons of the cell-cluster regions of the cuneate and gracile nuclei, and of the lateral cervical nucleus, in pentobarbitone-anaesthetized cats. Forepaw, hind paw or face regions of the contralateral Sm I cortex were identified by recording through an inserted microelectrode which was then used for stimulation. The effect of a double cortical shock or train of shocks was usually inhibition: occasionally facilitation was observed, or mixed effects with facilitation preceding inhibition. Effects were seen in about half the cells studied in all three nuclei. Some cells of the lateral cervical nucleus were strongly excited, an effect not seen in the other nuclei. No component of these responses depended on suprathreshold stimulus intensities. Some lateral cervical cells were studied after deafferentiation by section of the dorsolateral spinal white matter; the same pattern of effects was seen. With an upper stimulus limit of 200 ~tA, cuneate but not gracile cells were affected from the cortical forepaw region, and gracile but not cuneate cells from the hind paw region. With threshold stimuli in an identified part of the forepaw cortical representation it was clear that cuneate cells with cutaneous receptive fields in corresponding parts of the forepaw had the lowest thresholds (minimum 6 ~tA). Threshold rose steeply with distance across the paw, suggesting quite sharp focusing of corticofugal effects in this system. When using similar procedures with the lateral cervical nucleus, with an upper limit of 200 ~tA, stimulation of forelimb cortex, or of facial cortex, affected both neurons with forelimb and those with hind limb fields. With near-threshold stimuli (minimum 11 ~tA) a broad but inconsistent somatotopic relationship emerged between cortical site and cutaneous receptive field. We conclude that under these circumstances corticofugal actions are Abbreviations: DCN, dorsal column nuclei; LCN, lateral cervical nucleus; SCT, spinocervical tract; Sm L sensorimotor cortex; fore-
limb digits numbered I (medial) to V (lateral) * Present address: Southampton General Hospital, Shirley, Southampton SO9 4XY, U.K. Correspondence to: J.D. Cole
much more sharply focused spatially on lemniscal neurons of the dorsal column nuclei than on those of the lateral cervical nucleus. Key words: Medullary sensory nuclei - Corticofugal inhibition - Somatosensory system - Intracortical stimulation
Introduction Although corticofugal actions on subcortical somaesthetic nuclei are well known, their operational significance has proved hard to assess from existing anatomical and physiological knowledge, the latter having been derived almost entirely from recordings made in the nuclei while stimulating the cortex. In a previous paper (Cole and Gordon 1983) we described a new feature of corticofugal actions on the dorsal column nuclei by showing that the latency for action on the gracile nucleus is about twice that for the cuneate nucleus. This asymmetry allowed speculation about involvement of the corticofugal system in sensorimotor co-ordination, since the cortical action on each nucleus is matched in time with the arrival there of sensory feedback from hind or forelimb respectively when they are moved in response to a cortical command. In the present study we have investigated such systems further by comparing actions on the dorsal column nuclei with those on another prominent tactile pathway in the carnivores, the spinocervicothalamic system, studied at the level of the lateral cervical nucleus. These two pathways are believed to play different but complementary roles in tactile sensory function (see e.g. Norrsell 1966; Kitai and Weinberg 1968; Brown and Gordon 1977). In terms of numbers of cells and projecting axons the dorsal column system is much the larger (see the discussion in Brown 1981, p 85); the input from the hind half of the body on each side, studied in the upper cervi-
385 cal gracile fascicle, c o n t a i n s s o m e 25,000 a x o n s ( H w a n g et al. 1975) a n d the gracile nucleus c o n t a i n s s o m e 15,000 p o s t s y n a p t i c cells t h a t p r o j e c t in the m e d i a l l e m n i s c u s ( B l o m q v i s t 1980). T h e cell-cluster r e g i o n o f this n u c l e u s f r o m w h i c h we r e c o r d e d in the p r e s e n t s t u d y i s the densest source o f these l e m n i s c a l n e u r o n s (see e.g. K u y p e r s a n d T u e r k 1964). I n r e c o r d i n g s f r o m single l e m n i s c a l a x o n s in the c a t excited b y d o r s a l c o l u m n s t i m u l a t i o n ( B r o w n et al. 1974), t h o s e w i t h c u t a n e o u s receptive fields in the fore p a r t o f the b o d y a n d f o r e l i m b o u t n u m b e r e d t h o s e in the h i n d p a r t b y a b o u t ten times, so it is p o s s i b l e t h a t in t o t a l the d o r s a l c o l u m n nuclei give rise to s o m e 165,000 such l e m n i s c a l a x o n s o n e a c h side. I n c o n t r a s t , o n l y a b o u t 3000 a x o n s e n t e r e a c h l a t e r a l cervical nucleus f r o m the s p i n o c e r v i c a l t r a c t (van B e u s e k o m 1955), w h i c h c o n t a i n s n e a r l y e q u a l n u m b e r s f r o m f o r e a n d h i n d halves o f the b o d y ( H e a t h 1978; E n e v o l d s o n a n d G o r d o n 1989); a n d e a c h nucleus c o n t r i b u t e s o n l y a b o u t 8000 p o s t s y n a p t i c a x o n s to the m e d i a l l e m n i s c u s ( F l i n k a n d W e s t m a n 1986). M o s t o f the p e r i p h e r a l i n p u t to the d o r sal c o l u m n nuclei, especially to the cell-cluster regions, is f r o m p r i m a r y afferent fibres, w h e r e a s a m a j o r i t y o f s p i n o c e r v i c a l a x o n s a n d o f l a t e r a l cervical n u c l e a r cells c a r r y a c o n v e r g e n t i n p u t f r o m b o t h tactile a n d n o c i c e p tive c u t a n e o u s sense o r g a n s (see e.g. B r o w n 1973; B r o w n et al. 1989). B e h a v i o u r a l defects f o l l o w i n g lesions to either o r b o t h o f these p a t h s are difficult to i n t e r p r e t , b u t s h o w t h a t either a l o n e is sufficient to s u s t a i n reactions to tactile stimuli, a n d t h a t a c o m b i n e d lesion p r o duces severe a n d even p e r m a n e n t loss o f such r e a c t i o n s . T o t a l lesions o f the d o r s a l c o l u m n s alone, h o w e v e r , l e a d to defects in p e r f o r m i n g m o r e c o m p l e x b e h a v i o u r a l t a s k s i n v o l v i n g s p a t i a l a n d t e m p o r a l s e n s o r y sequences (see e.g. A l s t e r m a r k et al. 1986; a n d for a review N o r r s e l l 1980). Since b o t h the d o r s a l c o l u m n nuclei (see e.g. Towe 1973) a n d the n u c l e a r r e g i o n s o f the s p i n o c e r v i c o t h a l a m ic p a t h ( B r o w n a n d S h o r t 1974; P e t o 1980) a r e u n d e r c o r t i c o f u g a l c o n t r o l , it is o f i n t e r e s t to see h o w the spatial a s p e c t s o f the c o r t i c o f u g a l a c t i o n s c o r r e s p o n d w i t h w h a t is k n o w n o f t h e s t r u c t u r e a n d f u n c t i o n o f the t w o systems. Single units were r e c o r d e d f r o m e a c h nucleus u n d e r closely c o m p a r a b l e c o n d i t i o n s . C o r t i c o f u g a l p a t h s were a c t i v a t e d b y i n t r a c o r t i c a l m i c r o s t i m u l a t i o n ; we h a v e att a c h e d p a r t i c u l a r i m p o r t a n c e to e s t a b l i s h i n g r e l a t i o n ships b e t w e e n t h e r e c e p t i v e fields o f the n u c l e a r cells a f f e c t e d a n d the p o s i t i o n s o f the a r e a s s t i m u l a t e d w i t h i n the c o r t i c a l b o d y m a p .
Methods
General Cats weighing 1.9-3.0 kg were anaesthetized by intravenous infusion of pentobarbitone sodium (initially 38 mg/kg) after incubation with intramuscular ketamine hydrochloride (22 mg/kg). Deep anaesthesia, sufficient to abolish all spontaneous movement and withdrawal reflexes, was maintained throughout the experiment with supplements of 5-15 mg pentobarbitone. Arterial cannulation was avoided so as not to compromise circulation in any of the limbs or the head : a satisfactory indication of the state of the circulation at stimulation and recording sites was given by microscopic inspec-
tion of cortical and particularly medullary surfaces, where it was important that arteries were conspicuous and well-coloured, and that there was no cellular ' sludging' in veins or tissue swelling. During recording, flaccid paralysis was induced with gallamine triethiodide in an initial intravenous dose of 15 mg, and maintained with subsequent doses of 7.5 mg as required. The effect of each dose lasted 40-60 min, in contrast with that of the anaesthetic which varied little over several hours. By deliberately allowing each period of paralysis to wear off it was possible to make regular assessments of anaesthetic level. While paralysed the animal was artificially ventilated and end-tidal CO2, monitored with an infrared analyser (Godart Capnograph Model BE), maintained at 3.54.5%. After starting ventilation a bilateral pneumothorax was formed and the pneumothorax tubes closed externally with rubber balloons to prevent dehydration. Rectal temperature was kept at 38~ C by a thermostatic heating blanket.
Recording from identified lemniscal neurons in DCN and L C N With the head ventriflexed by 25-30 ~ with respect to the body, the dorsal medulla and first cervical segment were exposed by removal of most of the arch of the first vertebra and the occipitoatlantoid ligament. This gave access to the cell-cluster regions of the cuneate and gracile nuclei and to a substantial part of the lateral cervical nucleus for subsequent single-cell recording. A transverse incision was made across the scalp and a small trephine hole made 5.5 mm from the midline in the Horsley-Clarke frontal plane 4.5 overlying the medial lemniscus. This was enlarged to allow insertion of a transverse array of six steel needles, separated by 1 mm and insulated except for 0.5 mm at their tips, to an appropriate depth in the midbrain. Stimulating current could be passed through any adjacent pair in this array. A high-resistance recording microelectrode was inserted in the cell-cluster region of the cuneate nucleus of the opposite side and stimuli delivered to the lemniscal region: the optimal depth of the stimulating array, and the optimal pair of needles, for exciting cuneate cells antidromically were then determined by lowering the array through the tissue and repeatedly switching between electrodes until the lowest threshold for a substantial group of cells was found; the array was then fixed accordingly. Minimal thresholds for lemniscal axons with this electrode configuration, and using single rectangular pulses of 0.4 ms at 1 Hz, were 3040 gA; if antidromic excitation thresholds rose slightly during the lengthy experiment a further similar adjustment was made. While antidromic responses always occurred at constant latency and threshold, the only acceptable criterion of antidromic response was cancelling of the antidromic spike by orthodromic impulses, discharged either spontaneously or in response to skin stimuli, with critical timing. The optimal electrode position from which to activate cuneothalamic axons was also found to be optimal for gracilothalamic and cervicothalamic axons; at this level of the brainstem these projections are known to overlap (Busch 1961). It was easy to place the recording electrode in the cell-cluster region of either cuneate or gracile nucleus by using conspicuous surface markings and employing existing knowledge of the rostrocaudal relations of these regions with respect to the obex (see e.g. Gordon and Jukes 1964a; Blomqvist 1980; Gordon and Grant 1982). In the gracile nucleus they lie mainly between 2 and 5 mm caudal to the obex, and in the cuneate nucleus they extend caudally from the obex level for about 2 mm. The lateral cervical nucleus was found by inserting the electrode about 0.5 mm lateral to the dorsal root line at C1, angled 15~ lateral to medial. The more rostral insertions usually traversed the spinal nucleus of the trigeminal nerve containing cells with facial receptive fields, before entering the LCN, whose cells had fields on forelimb, trunk or hind limb. Cells with fields in cervical dermatomes 1-3 usually had no lemniscal projection : these were assumed to be dorsal horn cells. In each experiment about 15 electrode penetrations were made into the LCN, working from rostral to caudal and ending around the C2 root zone. A photograph taken and printed at the beginning
386 of each experiment was used to record the position of each microelectrode insertion in medulla or cervical cord.
Cortical stimulation Bone was removed over the Sm I sensorimotor cortex of the side opposite to that of medullary recording, the exposed frontal sinuses were plugged with wax, and a pool of mineral oil (paraffin) formed over the exposed cortex. A photograph of the exposed area was used to record the site of the recording electrode, which was inserted in a chosen position in the location of the somatotopic representation of the forelimb, hind limb or face, and then adjusted so that vigorous multiunit responses were recorded from tactile stimulation of a clear-cut area of skin (e.g. one or two toe-pads or an equivalent area of hairy skin). The electrode was then switched electronically to a stimulating mode, and cathodal stimuli delivered consisting of either paired 0.4 ms rectangular pulses separated by 2 ms, or trains of six 0.4 ms pulses at the same pulse separation, at 5 s intervals. This electrode was moved at least once in the course of an experiment, and the above procedure repeated. Stimulus current, where quoted, was monitored from the potential drop across a 100 O resistor in series with the electrode.
Deafferentation of the LCN In some experiments an additional laminectomy was made at C3C4 and the dorsolateral fascicle, containing the spinocervical tract, was divided by longitudinal tearing between two pairs of watchmakers' forceps. This was achieved by opening up the natural tissue plane separating the two fascicles; the lateral surface of the cuneate fascicle was thus exposed together with its pial vascular covering and was not touched by the forceps. The whole dorsolateral fascicle, in which the spinocervical tract lies quite superficially, was destroyed in this way under direct observation and this was judged not to need checking histologically; the lesion abolished the peripheral receptive fields of LCN cells on the side of recording. The area was packed with gelatin sponge and the incision closed.
Recording procedure Tungsten-in-glass electrodes (Merrill and Ainsworth 1972) for extracellular single unit recording had impedances of 0.5-2 Mr2, those for cortical recording (and stimulating) 1.5-40 kO, at I kHz. Each was coupled to a conventional amplifying system by a high impedance unity gain solid state stage (Analog Devices AD52 3KH; input impedance 10l~ f2). All recorded activity was monitored visually on a storage oscilloscope. Selected portions were fed in parallel to a magnetic tape recorder, and via a pulse-height analyser to a Neurolog NL750 signal averager (Digitimer Ltd) used in the pulse counting mode to generate peristimulus time (PST) histograms which were plotted by pen recorder, either on-line, or off-line from stored material (see below).
Studies of corticofugal inhibition and facilitation As a recording electrode was advanced through one or other nucleus in steps of 10 Ixm or less, an appropriate area of the body surface (for the LCN, the whole surface) was rapidly and continuously scanned by hand with light tactile stimulation. In accord with previous studies, the majority of recorded cells projecting into the medial lemniscus had a resting or 'spontaneous' discharge (see e.g. Gordon and Jukes 1964 for the DCN; Brown et al. 1989, Peto 1980 for the intact and deafferented LCN respectively); and we studied the effect of cortical conditioning stimuli on this discharge in single cells. Such a discharge is absent in most primary afferent cutaneous axons, and must therefore be generated in nuclear regions. The physical presence of a recording electrode near a cell
can increase the impulse frequency (Cremers 1971); however its presence in single lemniscal axons shows the discharge to occur independently of this. There is no evidence to suggest that it is a feature of any particular functional category of cell. Since the instability of recording conditions usually precluded long study of single cells, we looked for cells with visually significant lowered or increased spike density in the display following a train of cortical shocks (see above). We then observed this effect during cumulative superimposition of spikes in the PST display through 16, 32 or 64 sweeps according to the resting discharge frequency. We describe the effects respectively as inhibitory or facilitatory consequences of cortical stimulation. Any series of observations in which steady resting discharge was interrupted by one or more 'spontaneous' high-frequency bursts was discarded. Given a sufficient period of time before the cell was lost, in our later experiments this process was repeated many times on line with different stimulus strengths in order to determine an approximate threshold for the effect by successive approximation. Thus only threshold, duration, and, if long enough in relation to sweep time, latency for these effects were quantifiable. In determining threshold, it was nearly always apparent that duration was directly, and latency inversely, related to stimulus strength. Thus the use of threshold stimuli does not allow meaningful comparison of latency of corticofugal effect on different groups of cell, as it did in Cole and Gordon's (1983) experiments where stimuli were maximized with respect to latency of response. Use of the terms inhibition and facilitation is based purely on firing frequency and does not assume any particular underlying mechanism.
Histology At the bottom of successful recording electrode tracks electrolytic lesions were made by passing current (e.g. 10 gA for 10 s) with the electrode negative. Each experiment was terminated by giving a lethal overdose of pentobarbitone and then immediately perfusing the upper half of the animal in situ with isotonic saline followed by 5% formaldehyde-saline from a cannula inserted through the dorsal chest wall into the thoracic aorta. The exposed tissues were covered by swabs soaked in fixative. The head and neck, protected from evaporation, were left in the stereotaxic frame for about 12 h, at the end of which blocks of cortex and of the spinomedullary region were cut out with blades attached to the manipulators used to insert the cortical and medullary electrodes respectively, thereby ensuring that all insertions were parallel to the faces of the blocks. After further fixation these were embedded in low-viscosity nitrocellulose, sectioned transversely at 50 gm, and serial sections stained with cresyl violet (cortex) or thionin (medulla). Electrode tracks and marking lesions were readily identified in sections of this thickness, and were always verified in DCN, LCN and cortex.
Results U n i t s were isolated in gracile, c u n e a t e or lateral cervical nucleus. M u c h o f o u r s u b s e q u e n t p r o c e d u r e was constrained b y the difficulty o f r e c o r d i n g for a n y length o f time f r o m a n y one cell, especially f r o m c u n e a t e or gracile n e u r o n s , in this very pulsatile region. Tactile comp o n e n t s o f receptive fields were q u i c k l y assessed. A l m o s t all n e u r o n s studied r e s p o n d e d in a r a p i d l y a d a p t i n g fashi o n to d i s p l a c e m e n t of hairs or (in the D C N o n l y ) to lightly s t r o k i n g forefoot or h i n d foot toe-pads. Fields were characteristically small in distal limb areas a n d progressively larger i n p r o x i m a l l i m b or t r u n k ; there was n o significant difference in field size b e t w e e n the cells f r o m the different nuclei. O n e D C N cell gave a slowly a d a p t i n g response to m o v i n g a claw. H i g h threshold
387 Imp 4-
A
1
Fig 1A, B. Corticofugal actions on cuneate and lateral cervical nuclear cells elicited by trains of six pulses well above threshold, showing inhibition followed by 'rebound' facilitation. 32 superimposed sweeps. In all plots, sweep time is 500 ms, and bin width
2 ms. A Cuneate cell : receptive field medial forepaw. Cortical stimulation site: representation of forepaw digit I. B LCN cell: receptive field lateral hind paw. Cortical stimulation site: representation of angle of mouth
components of receptive fields were not investigated. All cells referred to were identified as projecting into the contralateral medial lemniscus. When stimulating cortex, particular attention was paid to the occurrence of inhibition, facilitation, or both, of resting discharge (see Methods), and to the relationship between the body representation at the site of cortical stimulation and the position of the receptive field of the cell studied. The aim was to sample cells influenced by such stimulation but not to determine overall the proportion of cells affected. The fact that only just over half were affected may merely reflect the experimental choice of particular sites for stimulating and recording in each instance, as well as the effect of anaesthesia, and the relative insensitivity o f our initial visual determination of the presence of an effect. More cells might have been affected if we had been able to test each from a number of cortical stimulation sites. Most L C N cells were isolated in the fully innervated nucleus, and some after deafferentation (see Methods). The classification of the latter as L C N cells - rather than cells in the adjacent spinal nucleus of the trigeminal nerve - depended on histological confirmation of electrode position, on knowledge of the position and depth of the trigeminal nucleus from previous exploration in the same experiment, and on exclusion by repeated tactile stimulation of the trigeminal skin area of any input from that source. It was sometimes difficult to distinguish the boundary between the two nuclei on histological evidence alone. Some effects of a cortical stimulus on L C N cells may be direct, and others secondary to actions of spinocervical cells; and while studies of the discharge o f L C N cells will not in themselves separate these effects, deafferentation at least isolates those that are direct.
Cuneate and gracile nuclei. Fifty-four cuneate and 34 gracile cells influenced by cortical stimuli were isolated and held long enough for the above investigations to be made. The effect of a cortical stimulus was usually inhibition, occasionally facilitation and sometimes apparently mixed effects with a period of facilitation preceding the inhibition. Inhibition alone, often lasting as long as 100 ms, occurred in 78% of this sample, facilitation alone in 9%, and mixed effects in 13%. Firing frequency following a period of inhibition was sometimes unstable, with a facilitatory ' r e b o u n d ' apparently followed by one or more further cycles of depression and facilitation (Fig. 1 A). Stimulation at a 'forelimb' cortical site had no effect on gracile cells, nor had stimulation at a hind limb site any affect on cuneate cells. Within a population of cells with receptive fields on either forelimb or hind limb, however, the exact position of the cortical electrode within the corresponding limb representation was not apparently critical: stimuli well above threshold delivered in a 'proximal hind limb' area of cortex could affect cells with fields distributed from hind paw to thigh, for instance, and stimuli in the cortical representation of the medial digit of the hind paw could affect cells with fields distributed between proximal tail and lateral hind paw digits. Similarly stimulation at a cortical forepaw focus affected not only cells with fields on the forepaw but also those with fields on elbow or even axilla. When, in two experiments, the cortical receptive focus and the cells' receptive fields on the forepaw exactly corresponded (four cells in all), inhibition was more intense or prolonged, or both, than in other forelimb cells studied. This trend was investigated further (see below) with threshold stimuli as it seemed likely that in the experiments just described the size of stimulus could have affected the results significantly.
Patterns of inhibition and facilitation produced by unquantified cortical shocks
Lateral cervical nucleus. Responses to cortical stimula-
In a preliminary survey the effects of intracortical stimuli not exceeding 200 gA were studied on the resting discharge of cells of the cuneate, gracile and lateral cervical nuclei. In this phase o f the study, thresholds for the cort~cofugal effects were not determined.
tion were seen in 48 cells isolated in the fully innervated lateral cervical nucleus, and in another 26 after deafferentation (see Methods). The usual effect of cortical conditioning on a spontaneously firing cell was similar to that seen on D C N cells (Fig. 1 B); facilitation and mixed effects were rather less common, but the same tendency for instability of firing was sometimes observed follow-
388 5-
A
5-
B
,,,,,,,,i,,,
i,,,,,,,,, ,,Li,, ,,I
i 5-
i=,, ,,,,
C
,=, ,
I,I=,,
,J,,
,,,,,,,,,,I=
=.,,
.,,
,,, ,,,,, ll ,=h, ,,
,.I
!
ing the inhibitory period. O f the 48 fully innervated cells, 89% showed pure inhibition, 5% pure facilitation, and 6% mixed effects. The cortical sites chosen for stimulation were in forelimb cortex lying medial, and in cortex with facial receptive fields lying just lateral to the coronal sulcus. A particular reason for investigating the latter was the conclusion of Peto (1980) that the origin of the corticofugal p a t h acting on the L C N lay in part o f the facial cortex. Stimulation o f a forelimb area of cortex led to effects not only on L C N cells with forelimb receptive fields, but also on hind limb cells. Stimulation o f facial cortex also had effects on b o t h forelimb and hind limb L C N cells. There was no evidence of inhibition produced by stimulation in corresponding and non-corresponding areas of cortex differing significantly either in depth or duration. It is clear therefore that under these conditions the corticofugal effects observed in the L C N were somatotopically diffuse c o m p a r e d with those in the cuneate and gracile nuclei. Corticofugal effects on deafferented cells, whether elicited f r o m forelimb or facial cortex, were similar, 25 showing pure inhibition and one a mixed effect. Figure 2 C shows an inhibitory effect on such a cell, elicited by a stimulus in facial cortex.
Corticofugal actions with quantified stimuli In seven experiments comparisons were m a d e between corticofugal actions on cuneate and lateral cervical nuclear cells in the same animal, using stimuli at or near threshold in an attempt to achieve better definition of
I
Fig. 2A-D. Corticofugal actions on cuneate and lateral cervical nuclear cells elicited by trains of six stimuli at or near threshold. In A-C, sweep time is 500 ms, and bin width 2 ms. In D, sweep time is 250 ms, and bin width 1 ms. A Shows some facilitation followed by inhibition of a cuneate cell with receptive field on forepaw digit I. Cortical site: representation of lateral forepaw. Stimulus 10 IxA. 32 sweeps. B Inhibition of an LCN cell with receptive field on abdomen and flank. Cortical site: representation of lateral forepaw. Stimulus 60 gA. 32 sweeps. C Inhibition of a deafferented LCN cell. Cortical site: representation of side of face. Stimulus 75 gA. 64 sweeps. D Strong facilitation of an LCN cell with receptive field on medial forepaw. Cortical site: representation of lateral forepaw. Stimulus 40 IxA. 16 sweeps
the areas o f cortex controlling these cells. Sites in forelimb cortex were used in all experiments, while in some cases a facial region located lateral to the coronal sulcus was also stimulated. Where possible, threshold was determined by approximation in successive on-line PST histogram plots, a time-consuming procedure often terminated prematurely by loss of the cell. Observations were started with stimuli of an estimated twice-threshold intensity, which was reduced in a successful series to values giving a just-observable effect (' threshold' stimuli). The resting discharge of a cell could be affected by a train of six intracortical pulses at 500 Hz; with some cells a pair of pulses with similar current was almost equally effective. The threshold value for an effect on nine cuneate and 25 L C N cells in four experiments ranged between 61 and 103 ~tA, with no single value below 30 gA: in the remaining three experiments, for which all values are shown in Fig. 3, 19 out of 46 cells (41%) had thresholds of 20 gA or less, the lowest observed being 6 gA. The reasons for these individual variations a m o n g experiments are not known, but m a y include level of anaesthesia. We regard the three with the lowest thresholds as the m o s t informative. The effect at threshold was usually inhibition. The cell illustrated in Fig. 2 A showed in addition an early facilitation, giving a mixed effect. Cells showing these effects m a d e up about half the cells isolated in each nucleus. Instability in firing frequency following the initial inhibitory period, mentioned earlier in relation to stimuli well above threshold (Fig. 1), was sometimes seen at threshold also (Fig. 2A). In two experiments there was a general rise in thresholds with time (see e.g. Fig. 3 A), possibly reflecting deterioration of the tissues.
389 150 - -
IIA
100
--
A
C
B
..J
§ 50--
oo
~ o
9
9
9
I
I CN
f
1
LCN
go
1
9
9 I
9
9 n
t
CN
Fig. 3A-C. Plot of approximate thresholds for corticofugal action on cuneate (CAr) and lateral cervical (LCN) nuclear cells by a train of six shocks in the contralateral cortical representation of the forepaw (solid circles) or of the face (circles with vertical line). The plots show data from three experiments, A-C, in the sequence,
Cuneate nucleus. Twenty-six cuneate cells were investigated with the intracortical electrode in forelimb cortex. In 18 cells, only inhibition was seen, in seven there were mixed effects, and in one, facilitation alone. It is apparent that mixed effects seen in such circumstances cannot represent suprathreshold p h e n o m e n a as b o t h c o m p o nents were present at threshold (Fig. 2A). With the cortical electrode in a k n o w n part of the limb representation particular interest was taken in thresholds for action on cuneate cells with fields in corresponding or neighbouring regions of the limb; and the results are best described by examples of such situations. Giving stimuli at a cortical focus corresponding to digits IV and V in the lateral forepaw, the thresholds for cortical action on two cuneate cells with receptive fields on digit I I I were 9 and 10 gA, and for another, with its field on digits II and III, 10 gA. A m u c h larger (70 gA) stimulus was needed for a cell with a field on digit I. These are the first four cells plotted in Fig. 3 A: the fifth cell had a forepaw receptive field that was not fully investigated. In another experiment with the cortical electrode in a similar position, seven o f the nine cells studied had fields which were satisfactorily located. One had a field on digit IV and two on digit V, corresponding with the cortical position, and these were inhibited at thresholds o f 7, 6 and 20 l.tA respectively, two with fields on digit I at 10 gA, and one with a field on the p a l m a r pad at 60 gA (Fig. 3B and Table 1). In b o t h examples, there is a qualitative m a t c h between spatial order and change o f threshold which indicates a basically somatotopic pattern in the corticofugal effects; quantitative aspects are considered in the Discussion. Lateral cervical nucleus. Fifty-four cells were investigated, including four isolated after deafferentation o f the nucleus. As with the cuneate sample, m o s t (approximately 76%) responded with inhibition alone to near-threshold cortical stimuli; this sample included the deafferented cells. Three (5%) showed a mixed effect with early
LCN
I--'7 CN
r---1 LCN
from left to right, in which the cells were recorded in each. The vertical arrow in B indicates deafferentation of the LCN (see Methods). Information about receptive fields is provided in the text and in Table I
Table 1. Relation between pogition of receptive field and threshold for corticofugal action on individual cells in cuneate and lateral cervical nuclei in one experiment Stimulating electrode in cortical representation of forepaw digits IV-V Cuneate Forepaw
Palmar pad
LCN digit V digit IV digit V digit I digit I
6 ~A 7 ~A 10 ~xA 10 ~A 10 ixA 60 ~xA
Forepaw digit III Forearm Forepaw digit III Shoulder/chest Flank Forelimb (field not fully specified) Flank Hind paw digit IV Upper arm/shoulder
12 gA 15 gA 30 gA 35 p,A 35 pA 40 p~A 60 I-tA 60 IxA 80 I-tA
These cells are listed in order of increasing threshold: in Fig. 3 B they are listed in the serial order in which they were isolated. Cuneate cell 6 and LCN cell 3 of Fig. 3 B, whose receptive fields were not adequately recorded, have been omitted from this table facilitation followed by inhibition, and eleven (19%) showed facilitation alone. Four of the latter, with receptive fields on trunk or forelimb, showed a clear burst of firing (Fig. 2D), their frequency of firing exceeding the prestimulus frequency by over three hundred times - a strong effect not seen in cuneate cells. N o facilitatory effects of any sort were seen on cells with hind limb fields (nine out of fifty-seven in this sample); the sample however is too small for this to be considered significant. These effects all occurred at the threshold for any response to be observed; in a sample of seven cells, each held long enough for the effects o f different stimulus strengths to be systematically investigated, increased current led to an increase in duration, intensity, or both, and decrease in latency, of the effects, but in no case was the type of response (i.e. facilitatory or inhibitory) altered. As already stated, stimulation of forelimb cortex
390 produced effects on LCN cells with fields on forelimb and on those with fields on hind limb; in addition in this sample, four cells with forelimb fields were affected by stimuli given in facial cortex (Fig. 3 C). It was of interest whether thresholds for corticofugal action differed according to the stimuli being given in somatotopically corresponding or non-corresponding areas of cortex. Such an assessment was difficult because a valid series of observations could only be made within single experiments, where acquiring a suitably large sample was rare. In one experiment with nine LCN cells having receptive fields reasonably distributed over the body, and with the stimulus in the cortical representation of the lateral forepaw, there was a clear tendency for the thresholds for cells with fields on distal forelimb to be lower than for others with fields on shoulder, trunk or hind limb (Fig. 3 B and Table 1). The relationship was inconsistent and change in threshold with distance across the body surface much less steep than that described for DCN cells under similar conditions (see above). A similar pattern is suggested within the small sample of cells whose thresholds for corticofugal action are plotted in Fig. 3 C; the three cuneate cells had receptive fields on the forepaw, whereas of the lateral cervical cells, all of which had low thresholds, the first two had fields on the forepaw, and the third on the hind paw. With stimuli in the cortical face representation, thresholds for action on LCN cells were either equal to or higher than the highest found in the same experiment with stimuli in forelimb cortex (see e.g. Fig. 3A). Deafferentation of the nucleus was followed by an increase in threshold values for inhibition (Fig. 3 B), but the data are too few for this to be established as significant.
Cytoarchitectonic areas stimulated
The main criterion in selecting the site of electrode insertion into the cortex was that vigorous activity was recorded on tactile skin stimulation. Histological study showed that in the preliminary series the electrode tips had been in sites fairly evenly distributed between areas 4, 3 a and b. In the studies with threshold stimuli, where the electrode was inserted either just medial or just lateral to the coronal sulcus slightly rostral to the postcruciate dimple on the posterior sigmoid gyrus, or just lateral to the sulcus on the coronal gyrus, most insertions (seven out of nine) were about 1 mm deep in area 3 b, one in area 2 and one at the 3 a/4 border. In these insertions, the electrode tip lay in a range between laminae III and VI. There were insufficient data for comparison of corticofugal effects produced by stimulating in different cytoarchitectonic areas or laminar positions.
Discussion
The present study shows that the effects of intracortical stimulation on the cells of the cluster regions of the dorsal column nuclei and on the spinocervicothalamic path-
way as seen in the lateral cervical nucleus are similar, consisting most often of pure inhibition, less commonly pure facilitation or a mixed effect. Lowest thresholds for actions on individual cells were also similar in the two paths. This evidence is consistent with other studies (see e.g. Cole and Gordon 1983, dorsal column nuclei; Peto 1980, lateral cervical nucleus). Earlier evidence had suggested that excitatory and inhibitory actions on t h e dorsal column nuclei were exerted through different pathways and upon different cell-groups (see the review by Towe 1973 for early references): our study has been restricted to the cluster regions where Gordon and Jukes (1964b) found only inhibition in single oscilloscope sweeps. Facilitatory effects are probably more readily observed with extensive sweep superposition and averaging as in the present study. The much higher incidence of mixed effects in the experiments of Cole and Gordon (1983) can probably be attributed to their inclusion of cells in other parts of the nuclei and their use of large stimuli which maximized these effects. It is a new observation that facilitation occurred at the same threshold as inhibition in both DCN and LCN, suggesting that they depend on excitation of the same or contiguous cells : contiguity might make them equally accessible to an electrical stimulus while operating separately under normal circumstances. One effect seen in some LCN cells but not in the DCN was excitation such as that shown in Fig. 2D; excitation was seen also by Peto (1980) in part of his cell sample. Actions on the fully innervated lateral cervical nucleus must in part be secondary to corticofugal actions on spinocervical cells. Surface stimulation of Sm I cortex revealed a broadly somatotopic arrangement in the effective areas for corticofugal inhibition on lumbosacral (Brown and Short 1974) and brachial (Heath 1978) SCT cells. With intracortical stimulation, inhibitory actions on lumbosacral SCT cells were elicited from a circumscribed focus in the medial (hind limb) area of Sm I cortex (Brown et al. 1977). Direct actions on the LCN were first reported by Peto (/980) in a sample of cells most of which had been deafferented by section of the dorsolateral spinal fascicle. A proportion of our sample of LCN cells (30 out of the total of 128) was also studied after irreversible deafferentation, and showed responses similar to those of the fully innervated cells, and to those of the deafferented cells studied by Peto. Absence of knowledge of the receptive field distribution of deafferented LCN cells is unfortunate: as a result our findings throw no light on somatotopic patterns within the system acting directly on the LCN. The preferable method of blocking the spinocervical tract reversibly by cooling was not feasible because of the difficulty in 'holding' cells in the spinomedullary region, the time-consuming nature of the determination of threshold for corticofugal action on each cell, and the added delay of some ten minutes (see e.g. Noble and Riddell 1989) needed to establish a block of the dorsal spinal white matter. The main contrast between these two systems, which were studied in our experiments under the same conditions and with identical criteria, was in the somatotopic patterns of corticofugal action. In the DCN, cuneate
391
cells were only affected by stimuli in forelimb cortex and gracile cells only from hind limb cortex. Cells in the fully innervated LCN, however, were affected whether stimulating in the somatotopic representation of forelimb, hind limb or face; and it was only with near-threshold stimuli that a broadly somatotopic relationship became apparent, the corresponding area of cortex being then the most effective. A similar influence from facial cortex was noted by Brown and Short (1974) and by Heath (1978) on SCT axons. Peto (1980), moving his stimulating electrode through multiple tracks in the cortex to establish 'best points' for inhibitory actions on 13 deafferented LCN cells, found that these lay in a cylindrical area of cortex and white matter in the lateral wall of the anterior sigmoid gyrus and upper bank of the coronal sulcus. Cytoarchitecturally this corresponded to area 3 a or the 3 a/4 border. It was judged by reference to published maps to lie in part of the face representation, although no recordings were made from the cortex to establish this. Our results may seem in conflict with this conclusion: although we also found stimuli in facial cortex to be effective, those in forelimb cortex were more so. Our experiments were conducted in a different way with fixed stimulating positions and it could be argued that if we had explored the cortex in individual cases, areas of lower threshold might have been found. It is also possible that as almost all our evidence on effects generated at threshold came from fully innervated LCN cells, we failed to detect direct heterotopic actions on the nucleus in the face of the corticofugal effects which originate earlier in the spinocervical pathway. It must also be asked how much our findings could have resulted from stimulus spread. Estimates of effective spread of current for direct and indirect (synaptic) excitation of pyramidal tract neurons in cat's cortex vary, but not enough to affect the present argument: Stoney et al. (1968) give a value of around 125 gm for direct excitation with currents of 20 gA, and Peto (1980) gives about 1 mm and 1.6 mm for indirect excitation with currents of 50 and 150 gA respectively. These distances are much too small to account for spread from the representation of the forelimb to that of the face or vice versa, areas separated by the coronal sulcus. Perhaps the most likely cause of any discrepancy lies in identification of the facial area of cortex. I n the present experiments this area was located by recording and was always lateral to the coronal sulcus: in previous experiments (Cole 1984) no representation of facial afferents could be found medial to the sulcus. Our most detailed study was made with threshold stimuli in a sample of cuneate cluster cells with receptive fields on the forepaw, where corticofugal actions were found to be much more sharply focused than in the sample of LCN cells. Lowest thresholds were found when these fields corresponded to the position of the cortical electrode within the forepaw representation, and increased proressively for cells with fields on more distant digits. Stimulus spread, as estimated above, is not however sufficient to account for the increase, or the rate of increase, of threshold with increasing distance.
A series of cortical mapping studies conducted under similar conditions (Cole 1984) showed for example that the representation of digit I at the 3 a/3 b border is distanced 2.5-4 mm from that of the lateral forepaw whereas in one example the threshold for a stimulus in the latter cortical area for inhibiting a cuneate digit I cell was 70 gA, from which the total effective stimulus spread would only be about 1 ram. Consideration of the other data presented in these examples (see Results) shows the same tendency consistently: it therefore seems likely that the corticofugal system is not organized in a point-to-point fashion, but in overlapping pools of neurons, the centre of each pool spatially in register with the sensory map but the extent of the pool not limited by the 'grain' seen in mapping the anaesthetized cortex. Such mapping does not in any case provide a true representation of the structural or functional potential of the somaesthetic area of cortex; in particular, both the size (McKenna et al. 1981 ; Duncan et al. 1982) and the functional complexity (see e.g. Hyv/irinen and Poranen 1978) of the receptive fields of single cells are greatly affected by anaesthesia. The main thrust of this investigation has been to give further emphasis to the high degree of spatial order in the cell-cluster regions of the dorsal column nuclei which, with their massive lemniscothalamic projections, form the main core of the so-called dorsal column system. The fine focusing of the corticofugal path would be expected to have operational significance in spatial discrimination; whether this is expressed in sharpening sensory impulse patterns, in precise deletions of unwanted portions of the sensory input, or in both, cannot be revealed by this type of study. Many studies on the conscious human have however shown a local reduction in tactile sensitivity during voluntary movement (see e.g. Coquery 1978). Schmidt et al. (1990) have shown that in finger movement such 'gating' is very significantly greater for elementary sensations in the finger moved than in its neighbours. Part of this effect persisted after local anaesthesia, suggesting the involvement of descending inhibition. The observed spatial pattern of thresholds for corticofugal inhibition in the cat's cuneate nucleus provides an interesting parallel with such observations.
Acknowledgements. We thank Patricia Cordery and Ankaret Wimshurst for invaluable help during these experiments. The work was done during tenure of a Schorstein Research Fellowship by J.D. Cole and of a project grant to G. Gordon from the British Medical Research Council. G. Gordon also acknowledges a personal grant from the Wellcome Trust.
References Alstermark B, G6rska T, Johannisson T, Lundberg A (1986) Effects of dorsal column transection in the upper cervical segments on visually guided forelimb movements. Neurosci Res 3 : 462-466 Beusekom GT van (1955) Fibre analysis of the anterior and lateral funiculi of the cord in the cat. Doctoral thesis, Eduard Ijdo, Leiden Blomqvist A (1980) Gracilo-diencephalic relay cells: a quantitative study in the cat using retrograde transport of horseradish peroxidase. J Comp Neurol 193:1097-1125
392 Brown AG (1973) Ascending and long spinal pathways: dorsal columns, spinocervical tract and spinothalamic tract. In: Iggo A (ed) Somatosensory system. Handbook of sensory physiology, Vol II. Springer, Berlin Heidelberg New York, pp 315-338 Brown AG (1981) Organization in the spinal cord. The anatomy and physiology of identified neurones. Springer, Berlin Heidelberg New York Brown AG, Gordon G, Kay RH (1974) A study of single axons in the cat's medial lemniscus. J Physiol (Lond) 236: 225-246 Brown AG, Short AD (1974) Effects from the somatic sensory cortex on transmission through the spinocervical tract. Brain Res 74: 338-341 Brown AG, Coulter JD, Rose PK, Short AD, Snow PJ (1977) Inhibition of spinocervical tract discharges from localized areas of the sensorimotor cortex in the cat. J Physiol (Lond) 264:1-16 Brown AG, Gordon G (1977) Subcortical mechanisms concerned in somatic sensation. Br Med Bull 33:121-128 Brown AG, Maxwell DJ, Short AD (1989) Receptive fields and in-field inhibition of neurones in the cat's lateral cervical nucleus. J Physiol (Lond) 413:119-137 Busch HFM (1961) An anatomical analysis of the white matter in the brain stem of the cat. Doctoral thesis, van Gorcum, Leiden Cole JD, Gordon G (1983) Timing of corticofugal actions on the gracile and cuneate nuclei of the cat. J Physiol (Lond) 341:139152 Cole JD (1984) A physiological study of the organisation of descending effects within the somaesthetic system of the anaesthetised cat. DM thesis, University of Oxford Coquery J-M (1978) Role of active movement in control of afferent input from skin in cat and man. In: Gordon G (ed) Active touch. Pergamon Press, Oxford New York, pp 161-169 Cremers PFLJM (1971) Responses of single units in the nervus suralis and the nucleus gracilis to stimulation of a hair receptor of cat and rat. Doctoral thesis, Nijmegen, The Netherlands Duncan GH, Dreyer DA, McKenna TM, Whitsel BL (1982) Doseand time-dependent effects of ketamine on SI neurones with cutaneous reseptive fields. J Neurophysiol 47: 677 699 Enevoldson TP, Gordon G (1989) Spinocervical neurons and dorsal horn neurons projecting to the dorsal column nuclei through the dorsolateral fascicle: a retrograde HRP study in the cat. Exp Brain Res 75:621-630 Flink R, Westman J (1986) Different neuron populations in the feline lateral cervical nucleus: a light and electron microscopic study with the retrograde axonal transport technique. J Comp Neurol 250: 265-281 Gordon G, Jukes MGM (1964a) Dual organization of the exteroceptive components of the cat's gracile nucleus. J Physiol (Lond) 173:263-290
Gordon G, Jukes MGM (1964b) Descending influences on the exteroceptive organizations of the cat's gracile nucleus. J Physiol (Lond) 173:291-319 Gordon G, Grant G (1982) Dorsolateral spinal afferents to some medullary sensory nuclei. An anatomical study. Exp Brain Res 46 : 1~23 Heath JP (1978) The cutaneous sensory input to the spinocervical tract of the cat and the corticofugal modulation of transmission from the forelimb component. PhD thesis, University of Edinburgh Hwang Y-C, Hinsman EJ, Roesel OF (1975) Caliber spectra of fibers in the fasciculus gracilis of the cat cervical spinal cord: a quantitative electron microscopic study. J Comp Neuroi 162:195-203 Hyv/irinen J, Poranen A (1978) Receptive field integration and submodality convergence in the hand area of the post-central gyrus of the alert monkey. J Physiol (Lond) 283 : 539-556 Kitai ST, Weinberg J (1968) Tactile discrimination study of the dorsal column-medial lemniscal system and spino-cervico-thalamic tract in cats. Exp Brain Res 6 : 234-246 Kuypers HGJM, Tuerk JD (1964) The distribution of the cortical fibres within the nuclei cuneatus and graeilis in the cat. J Anat (Lond) 98:143-162 McKenna TM, Whitsel BL, Dreyer DA, Metz CB (1981) Organization of cat anterior parietal cortex: relations among eytoarchitecture, single neuron functional properties and interhemispherie connectivity. J Neurophysiol 45 : 667-697 Merrill EG, Ainsworth A (1972) Glass-coated platinum-plated tungsten micro-electrodes. Med Biol Eng 10:662-672 Noble R, Riddell JS (1989) Descending influences on the cutaneous receptive fields of postsynaptic dorsal column neurones in the cat. J Physiol (Lond) 408:167-183 Norrsell U (1966) The spinal afferent pathways of conditioned reflexes to cutaneous stimuli in the dog. Exp Brain Res 2:269282 Norrsell U (1980) Behavioral studies of the somatosensory system. Physiol Rev 60:327-354 Peto TEA (1980) Corticofugal actions on the lateral cervical nucleus of the cat. Exp Neurol 68:531-547 Schmidt RF, Torebj6rk HE, Schady WJL (1990) Gating of tactile input from the hand II. Effects of remote movements and anaesthesia. Exp Brain Res 79:103-108 Stoney SD, Jr, Thompson WD, Asanuma H (1968) Excitation of pyramidal tract cells by intracortical microstimulation: effective extent of stimulating current. J Neurophysiol 31:659-669 Towe AL (1973) Somatosensory cortex: descending influences on ascending systems. In: Iggo A (ed) Somatosensory system. Handbook of sensory physiology, Vol II. Springer, Berlin Heidelberg New York, pp 701-718