- Email: [email protected]
Rhythmic antidromic discharges of single primary afferents recorded in cut dorsal root filaments during locomotion in the cat
Brain Research, 359 (1985) 375-378 Elsevier
375
BRE 21240
Rhythmic antidromic discharges of single primary afferents recorded in cut dorsal root filaments during locomotion in the cat R. DUBUC, J.-M. CABELGUEN* and S. ROSSIGNOL Centre de Recherche en Sciences Neurologiques, Department of Physiology, Faculty of Medicine, Universit~ de Montrdal, P.O. Box 6128, Station A, Montrdal, QuL H3C 3J7 (Canada)
(Accepted August 5th, 1985) Key words: locomotion - - dorsal root - - primary afferent - - antidromic discharge - - decorticate cat - - spinal cat
Single units recorded in the proximal stump of cut dorsal root filaments were found to antidromically discharge rhythmically during fictive locomotion in decorticate and paralyzed cats. Some units fired throughout the period of flexor or extensor nerve activity, whereas other units discharged near the transitional phases. Similar findings were made in acutely spinalized and paralyzed cats injected with L-DOPA, as well as in non-paralyzed decorticate cats walking on a treadmill. These results suggest that different types of primary afferents may be depolarized cyclically at different specific time in the step cycle by the central pattern generator for locomotion, and that this central control of the primary afferents may be involved in the modulation of the reflex transmission observed during locomotion.
During movement, there is a continuous interaction between the motor c o m m a n d and sensory feedback. Thus, although the sensory feedback may serve to adjust the execution of the movement to accommodate to prevailing conditions, the motor command itself may also exert control over the sensory information which it receives. For instance, during locomotion, stimulation of cutaneous reflex pathways originating from the dorsum of the paw may evoke large excitatory responses in flexor muscles, during the swing phase of the locomotor cycle, but not during the stance phase. Conversely, the same stimulation may evoke excitatory responses in extensor muscles during the stance phase, but not during the swing phase6, 7. This suggests that the neuronal elements responsible for generating the locomotor pattern may also phasically modulate sensory transmission of at least some cutaneous pathways. Such central control could be exerted at different sites in the reflex pathways, the first of which being the primary afferents themselves. Indeed, during locomotor activity in paralyzed cats (fictive locomotion), it has
been reported that primary afferents are cyclically depolarized2,3, 5. Recordings of dorsal root potentials (DRPs) have revealed, in most cases, a maximum of depolarization close to the time of transition between the end of flexor muscle nerve activity and the onset of activity in the extensor muscle nerve. Minimal depolarization was reached at the extension-flexion transition. By testing the excitability of different populations of primary afferents with Wall's technique is, it was shown that the different groups of afferents (muscular and cutaneous) were depolarized together in the step cycle. The above observations then suggest that the sensory transmission should be facilitated in one phase and reduced in the opposite phase. This is difficult to reconcile with the specific modulation of different cutaneous reflexes observed in different phases 6.7 as described above. For this reason, presynaptic mechanisms were reinvestigated during fictive locomotion, in decorticate paralyzed cats. In the course of these experiments, it was striking to find that single primary afferents, recorded in the proximal stump of sec-
* Present address: Laboratoire de Neurophysiologie compar6e, Universit6 Pierre et Marie Curie, 9 Quai Saint-Bernard, 75005 Paris, France. Correspondence: S. Rossignol, Centre de Recherche en Sciences Neurologiques, Department of Physiology, Faculty of Medicine, Universit6 de Montr6al, P.O. Box 6128, Station A, Montr6al, Qu6. H3C 3J7, Canada. 0006-8993/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
376 tioned dorsal roots, often discharged rhythmically at the periodicity of the step cycle. It was deemed worthwhile to study these discharges in more details, since they presumably resulted from a suprathreshold depolarization of primary afferent terminals induced by the central locomotor activity in absence of phasic peripheral inputs. Ten cats were decorticated 13 under methohexital sodium (7 mg/kg). As soon as they started to exhibit spontaneous walking movements, they were paralyzed with gallamine triethiodide (10 mg/kg; repeated when needed) and artificially ventilated. Peripheral nerves were dissected and mounted for recording with conventional bipolar Ag/AgC1 hook electrodes in a pool of warm paraffin oil. An extensive laminectomy gave access to dorsal roots from the level of L4-$2. The spinal cord was covered with paraffin oil and maintained at around 38 °C using a feedback controlled heating element. In experiments in which fictive locomotion did not appear spontaneously, it could sometimes be induced by stimulating a dorsal root or part of it with a short train (100 ms, 300 Hz, 0.1 ms pulses). Dorsal rootlets were cut one at a time, dissected with fine forceps from the rest of the root and mounted on bipolar hook electrodes. The preamplifier bandwidth was set between 300 Hz and 10 kHz to filter out DRPs from the records. In 10 cats with fictive locomotion, 194 single units with spontaneous activity were recorded in cut dorsal root filaments. It is noteworthy that these antidromic discharges could be recorded in filaments cut several hours previously; it is thus unlikely that they should represent injury discharges. 116 units (60%) fired in short bursts, or as single spikes, but without any relation to the locomotor cycle. Some of these units could be modulated by passive manipulations of the hindlimbs or the forelimbs, as well as by manual skin stimulation of the hindlimb. Most importantly, 78 units (40%) recorded in 9 of the cats clearly discharged rhythmically with bursts occurring at a fixed time relationship with the step cycle. Fig. 1 illustrates 3 examples taken from two different experiments. In Fig. 1A, the unit is recorded from a cut L7 dorsal root filament. Before the onset of this spontaneous bout of locomotion, the large unit is silent. It is recruited in the first cycle, at the onset of the activity in the flexor nerve Sartorius Lateralis
A
AF351 6
L7
. . . . . . . . . . . . . . . . . . • trr" ~ ' ~ - , " ~
"'-'~r .......... . . . ilillll
B
AF3514
L7
.....J -W
......
|
lit
..... all
" !1
1
A,,[
_1
~71
~
II
1
iSartL iSmA C
AF4407
L7
1s Fig. 1. Rhythmic antidromic discharges recorded from single primary afferents in the proximal stump of the cut 7th lumbar (L7) dorsal root in decorticate and paralyzed cats. A: a unit discharging during the activity of the nerve to the ipsilateral (i) Sartorius Lateralis (iSartL). B: a unit discharging atthe transition between iSartL and the ipsilateral Semimembranosus Anterior (iSmA). C: a unit discharging two bursts of activity in certain step cycles.
(iSartL) and lasts throughout the greater part of its activity. Other units of smaller amplitude are active during that time, but do not appear to discharge with any clear relationship with the ongoing rhythm. When the rhythmic activity stopped, the large unit also ceased firing. Fig. lB. taken from the same experiment, shows a unit in a different L7 rootlet. which, in this case, discharges near the transition between the end of iSartL and the onset of the extensor nerve Semimembranosus Anterior (iSmA). It ]s noteworthy that the unit fired consistently at the same time in the step cycle, even though there could be missing bursts in iSartL (left hand side) or iSmA (right hand side). It is thus quite clear, from the two previous examples. that the dorsal root units do not all discharge in the same periods of the step cycle. Fig. 1C shows an example of an even more complex discharge pattern.
377 This unit, taken from another experiment, fired a burst of activity every time iSartL discharged. Moreover, a second burst of discharge appeared during iSmA activity in cycles where it was particularly prolonged. In cycles with short iSmA activity, only the burst associated with iSartL could clearly be distinguished. It was important to determine whether these antidromic discharges could result from the influence of descending pathways which are known to be phasically active in such decorticate and paralyzed preparations 12. Therefore, in 5 of the paralyzed decorticate animals, the spinal cord was sectioned at Thl3 and nialamide (50 mg/kg) and L-DOPA (50 mg/kg) were injected i.v. to induce locomotor activity in the hindlimbs 9. The results obtained in such spinal preparation are examplified in Fig. 2A and B which illustrate A
AF3203 I ItllIL
i
I
S
lilltii
a
iVL
illt~l
r
,
'
t
, L, --,~,,,L . . . .
L
'
,jill[ [
.I Jtllill
~
\
B
AF3209
iVL
,,
C L7
EX37 ~.i, LJLaIIItlII IiIIJL//ild/,,,tlll..Mi ,LI,, '11~1
....
ist iVL
i,iii "~'!
IL
i
i
,i i
i
iJ
[
~ili~i~i~ii'Wlili.,¶li'~li~ll '~W W V
coSt I l~ " ~ coVL~ i i l / I
~- 1 I ' ' ~ rl l i i ' ,,il I l l
1
~
W
l
~- I ~i- .,.i. ii~' ls
Fig. 2. Rhythmic antidromic discharges of single primary afferents in the proximal stump of cut dorsal roots during fictive locomotion in a spinal cat and real locomotion in a decorticate cat. A: a S1 unit discharging during the activity of the nerve to the ipsilateral Vastus Lateralis (iVL) in an acute low spinal cat injected with nialamide and L-DOPA. B: in the same preparation, a S1 unit discharges before and throughout the activity of iSartL nerve, the iVL nerve being silent. C: decorticate cat walking on a treadmill belt; the electromyograms are recorded from the Semitendinosus (St) and the VL muscles ipsilaterally (i) and contralaterally (co) with implanted copper wires. The unit was recorded in a long filament of L7 which provided stable recordings despite the movements of the animal.
two dorsal root units in S1 discharging antidromically at the periodicity of the locomotor rhythm. In Fig. 2A, the unit followed quite closely the output of the extensor nerve Vastus Lateralis (iVL). For instance, in the 4th burst, the unit discharges for a longer time and at a higher frequency and corresponds to the larger and longer burst of activity in the iVL neurogram. The 5th burst of the unit is of shorter duration corresponding also to an iVL activity which is shorter and smaller than in the previous cycle. In the same experiment, but in a different S1 dorsal rootlet (Fig. 2B), one unit discharges throughout the period of activity of the flexor iSartL and even starts before the onset of activity in that nerve. Note that there was no activity in the iVL nerve in this case. Finally, it was of interest to know if such units could similarly be recorded in animals actually walking on a treadmill. Two cats were decorticated and the electromyographic (EMG) a&ivity of selected hindlimb muscles was recorded. Fig. 2C shows a unit in L7 which is gradually recruited as the walking pattern becomes well organized. This unit fires mainly throughout the period of activity of the ipsilateral Semitendinosus (iSt) and starts to discharge just at the transition between the end of iVL and the start of iSt. This pattern of firing recorded in a walking cat, closely resembles the unit recorded in the paralyzed cat shown in Fig. 1A. The present results indicate clearly that, during real and fictive locomotion, the primary afferents are subjected to changes in polarization that may lead to antidromic firing which can be recorded as unitary discharges in cut dorsal root filaments. Spontaneous antidromic unitary discharges in cut dorsal root filaments have been recorded before in anesthetized cats ]4. Unit discharges related to rhythmic DRPs have been briefly mentioned by Bayev and Kostyuk 5 for fictive locomotion and were illustrated for fictive scratching 4. The interest of the present results is to emphasize that different primary afferents may be subjected to different depolarizing drives. Indeed, some afferents fired during the periods of flexor and/or extensor activity or else near transitional phases between the flexors and the extensors in the step cycle. This complex pattern of firing is more in keeping with the already well-documented complexity of the centrally generated locomotor patternSA3 and suggests that the transmission in primary affer-
378 ents themselves is under close control of the central pattern generator. O n e could postulate from these results that the transmission in some afferents will be m o d u l a t e d as a function of the different phases of the step cycle. It is possible that different groups of primary afferents, or afferents originating from different parts of the limb, may be influenced at different times during the step cycle. H o w e v e r , the present experimental approach did not allow the identification of the origin and type of the different afferents. The extracellular technique allows to record presumably only a fraction of units whose m e m b r a n e polarization is cyclically changed during locomotion. Indeed, experiments in progress, using intracellutar recordings of primary afferents, suggest that some units are only subjected to subliminal changes in their m e m b r a n e polarization during fictive locomotion. The present results may have two o t h e r secondary consequences. Firstly, it is generally assumed that, during fictive locomotion, only tonic afferent activity remains since the limbs are not moving. A l t h o u g h there are indeed no phasic afferent inputs originating from the stationary limbs, some primary afferents can apparently also be brought to fire cyclically by the central p a t t e r n g e n e r a t o r for locomotion. There-
1 Arshavsky, Yu. I., Berkinblit, M.B., Fukson, O.I., Gelfand, I.M. and Orlovsky, G.N., Origin of modulation in neurones of the ventral spinocerebeilar tract during locomotion, Brain Research, 43 (1972) 276-279. 2 Bayev, K.V., Periodic changes in primary afferent depolarization during fictitious locomotion by thalamic cats, Neurophysiology, 10 (1978) 316-317. 3 Bayev, K.V., Polarization of primary afferent terminals in the lumbar spinal cord during fictitious locomotion, Neurophysiology, 12 (1980) 305-311. 4 Bayev, K.V. and Kostyuk, P.G., Primary afferent depolarization evoked by the activity of spinal scratching generator, Neuroscience, 6 (1981) 205-215. 5 Bayev, K.V. and Kostyuk, P.G., Polarization of primary afferent terminals of lumbosacrat cord elicited by the activity of spinal locomotor generator, Neuroscience, 7 (1982) 1401-1409. 6 Forssberg, H., Grillner, S. and Rossignol, S., Phase-dependent reflex reversal during walking in chronic spinal cats, Brain Research, 85 (1975) 103-107. 7 Forssberg, H., Grillner, S. and Rossignol, S., Phasic gain control of reflexes from the dorsum of the paw during spinal locomotion, Brain Research, 132 (1977) 121-139. 8 GriUner, S., Control of locomotion in bipeds, tetrapods and fish. In J.M. Brookhart and V.B. Mounteastle (Eds.), Handbook of Physiology, Section 1: The Nervous System,
fore, if the afferents discharged by presynaptic depolarization can still exert some postsynaptic effects, it is possible that second order neurons could be rhythmically m o d u l a t e d in this m a n n e r during fictive locomotion (see ref. 1 as an example). Secondly. when recording from dorsal root ganglion cells whose axons are in continuity with the spinal cord in chronic preparations H, techniques such as backward spike triggered averaging should be used as done by Loeb l0 to insure that the recorded activity results from peripheral stimulation and not from a central activation as r e p o r t e d here. H o w e v e r . there is up to now. no experimental evidence that such antidromic discharges are present in intact animals. This work was s u p p o r t e d by a G r o u p grant from the Medical Research Council of Canada. J.-M.C. was supported by a fellowship from the F r a n c e - C a nada exchange p r o g r a m ( C N R S - C N R C ) as well as by the Centre de Recherche en Sciences Neurologiques of the Universit6 de Montreal. R . D . was awarded an F . C . A . C . studentship. J. Provencher, S. Bergeron, R. Bouchoux and E. Rupnik are gratefully acknowledged for their technical contributions.
Vol. II, Part 2, American Physiological Society, 1981, pp. 1179-1236. 9 Grillner, S. and Zangger, P., On the central generation of locomotion in the low spinal cat, Exp. Brain Res., 34 (1979) 241-261. 10 Loeb, G.E., Somatosensory unit input to the spinal cord during normal walking, Can. J. Physiol. Pharmacol., 59 (1980) 627-635. 11 Loeb, G.E., Bak, M.J. and Duysens, J., Long-term unit recording from somatosensory neurons in the spinal ganglia of the freely walking cat. Science. 197 (1977) 1192-1194 12 Perret, C.. Neural control of locomotion in the decorticate cat. In R.M. Herman. S. Grillner. P.S.G. Stein and D.G. Stuart IEds.), Neural Control of Locomotion. Plenum Press. New York. 1976. pp. 587-615. 13 Perret, C and Cabelguen. J.-M.. Main characteristics of the hindlimb locomotor cycle in the decorticate cat with special reference to bifunctional muscles. Brain Research. 187 (1980) 333-352 14 Repkin, A.H.. Wolf. P. and Anderson. E.G.. Non-GABA mediated primary afferent depolarization. Brain Research, 117 (1976) 147-152. 15 Wall, P.D.. Excitability changes in afferent fibre terminations and their relation to slow potentials. J. Physiol. (London). 142 (19581 1-21.
375
BRE 21240
Rhythmic antidromic discharges of single primary afferents recorded in cut dorsal root filaments during locomotion in the cat R. DUBUC, J.-M. CABELGUEN* and S. ROSSIGNOL Centre de Recherche en Sciences Neurologiques, Department of Physiology, Faculty of Medicine, Universit~ de Montrdal, P.O. Box 6128, Station A, Montrdal, QuL H3C 3J7 (Canada)
(Accepted August 5th, 1985) Key words: locomotion - - dorsal root - - primary afferent - - antidromic discharge - - decorticate cat - - spinal cat
Single units recorded in the proximal stump of cut dorsal root filaments were found to antidromically discharge rhythmically during fictive locomotion in decorticate and paralyzed cats. Some units fired throughout the period of flexor or extensor nerve activity, whereas other units discharged near the transitional phases. Similar findings were made in acutely spinalized and paralyzed cats injected with L-DOPA, as well as in non-paralyzed decorticate cats walking on a treadmill. These results suggest that different types of primary afferents may be depolarized cyclically at different specific time in the step cycle by the central pattern generator for locomotion, and that this central control of the primary afferents may be involved in the modulation of the reflex transmission observed during locomotion.
During movement, there is a continuous interaction between the motor c o m m a n d and sensory feedback. Thus, although the sensory feedback may serve to adjust the execution of the movement to accommodate to prevailing conditions, the motor command itself may also exert control over the sensory information which it receives. For instance, during locomotion, stimulation of cutaneous reflex pathways originating from the dorsum of the paw may evoke large excitatory responses in flexor muscles, during the swing phase of the locomotor cycle, but not during the stance phase. Conversely, the same stimulation may evoke excitatory responses in extensor muscles during the stance phase, but not during the swing phase6, 7. This suggests that the neuronal elements responsible for generating the locomotor pattern may also phasically modulate sensory transmission of at least some cutaneous pathways. Such central control could be exerted at different sites in the reflex pathways, the first of which being the primary afferents themselves. Indeed, during locomotor activity in paralyzed cats (fictive locomotion), it has
been reported that primary afferents are cyclically depolarized2,3, 5. Recordings of dorsal root potentials (DRPs) have revealed, in most cases, a maximum of depolarization close to the time of transition between the end of flexor muscle nerve activity and the onset of activity in the extensor muscle nerve. Minimal depolarization was reached at the extension-flexion transition. By testing the excitability of different populations of primary afferents with Wall's technique is, it was shown that the different groups of afferents (muscular and cutaneous) were depolarized together in the step cycle. The above observations then suggest that the sensory transmission should be facilitated in one phase and reduced in the opposite phase. This is difficult to reconcile with the specific modulation of different cutaneous reflexes observed in different phases 6.7 as described above. For this reason, presynaptic mechanisms were reinvestigated during fictive locomotion, in decorticate paralyzed cats. In the course of these experiments, it was striking to find that single primary afferents, recorded in the proximal stump of sec-
* Present address: Laboratoire de Neurophysiologie compar6e, Universit6 Pierre et Marie Curie, 9 Quai Saint-Bernard, 75005 Paris, France. Correspondence: S. Rossignol, Centre de Recherche en Sciences Neurologiques, Department of Physiology, Faculty of Medicine, Universit6 de Montr6al, P.O. Box 6128, Station A, Montr6al, Qu6. H3C 3J7, Canada. 0006-8993/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
376 tioned dorsal roots, often discharged rhythmically at the periodicity of the step cycle. It was deemed worthwhile to study these discharges in more details, since they presumably resulted from a suprathreshold depolarization of primary afferent terminals induced by the central locomotor activity in absence of phasic peripheral inputs. Ten cats were decorticated 13 under methohexital sodium (7 mg/kg). As soon as they started to exhibit spontaneous walking movements, they were paralyzed with gallamine triethiodide (10 mg/kg; repeated when needed) and artificially ventilated. Peripheral nerves were dissected and mounted for recording with conventional bipolar Ag/AgC1 hook electrodes in a pool of warm paraffin oil. An extensive laminectomy gave access to dorsal roots from the level of L4-$2. The spinal cord was covered with paraffin oil and maintained at around 38 °C using a feedback controlled heating element. In experiments in which fictive locomotion did not appear spontaneously, it could sometimes be induced by stimulating a dorsal root or part of it with a short train (100 ms, 300 Hz, 0.1 ms pulses). Dorsal rootlets were cut one at a time, dissected with fine forceps from the rest of the root and mounted on bipolar hook electrodes. The preamplifier bandwidth was set between 300 Hz and 10 kHz to filter out DRPs from the records. In 10 cats with fictive locomotion, 194 single units with spontaneous activity were recorded in cut dorsal root filaments. It is noteworthy that these antidromic discharges could be recorded in filaments cut several hours previously; it is thus unlikely that they should represent injury discharges. 116 units (60%) fired in short bursts, or as single spikes, but without any relation to the locomotor cycle. Some of these units could be modulated by passive manipulations of the hindlimbs or the forelimbs, as well as by manual skin stimulation of the hindlimb. Most importantly, 78 units (40%) recorded in 9 of the cats clearly discharged rhythmically with bursts occurring at a fixed time relationship with the step cycle. Fig. 1 illustrates 3 examples taken from two different experiments. In Fig. 1A, the unit is recorded from a cut L7 dorsal root filament. Before the onset of this spontaneous bout of locomotion, the large unit is silent. It is recruited in the first cycle, at the onset of the activity in the flexor nerve Sartorius Lateralis
A
AF351 6
L7
. . . . . . . . . . . . . . . . . . • trr" ~ ' ~ - , " ~
"'-'~r .......... . . . ilillll
B
AF3514
L7
.....J -W
......
|
lit
..... all
" !1
1
A,,[
_1
~71
~
II
1
iSartL iSmA C
AF4407
L7
1s Fig. 1. Rhythmic antidromic discharges recorded from single primary afferents in the proximal stump of the cut 7th lumbar (L7) dorsal root in decorticate and paralyzed cats. A: a unit discharging during the activity of the nerve to the ipsilateral (i) Sartorius Lateralis (iSartL). B: a unit discharging atthe transition between iSartL and the ipsilateral Semimembranosus Anterior (iSmA). C: a unit discharging two bursts of activity in certain step cycles.
(iSartL) and lasts throughout the greater part of its activity. Other units of smaller amplitude are active during that time, but do not appear to discharge with any clear relationship with the ongoing rhythm. When the rhythmic activity stopped, the large unit also ceased firing. Fig. lB. taken from the same experiment, shows a unit in a different L7 rootlet. which, in this case, discharges near the transition between the end of iSartL and the onset of the extensor nerve Semimembranosus Anterior (iSmA). It ]s noteworthy that the unit fired consistently at the same time in the step cycle, even though there could be missing bursts in iSartL (left hand side) or iSmA (right hand side). It is thus quite clear, from the two previous examples. that the dorsal root units do not all discharge in the same periods of the step cycle. Fig. 1C shows an example of an even more complex discharge pattern.
377 This unit, taken from another experiment, fired a burst of activity every time iSartL discharged. Moreover, a second burst of discharge appeared during iSmA activity in cycles where it was particularly prolonged. In cycles with short iSmA activity, only the burst associated with iSartL could clearly be distinguished. It was important to determine whether these antidromic discharges could result from the influence of descending pathways which are known to be phasically active in such decorticate and paralyzed preparations 12. Therefore, in 5 of the paralyzed decorticate animals, the spinal cord was sectioned at Thl3 and nialamide (50 mg/kg) and L-DOPA (50 mg/kg) were injected i.v. to induce locomotor activity in the hindlimbs 9. The results obtained in such spinal preparation are examplified in Fig. 2A and B which illustrate A
AF3203 I ItllIL
i
I
S
lilltii
a
iVL
illt~l
r
,
'
t
, L, --,~,,,L . . . .
L
'
,jill[ [
.I Jtllill
~
\
B
AF3209
iVL
,,
C L7
EX37 ~.i, LJLaIIItlII IiIIJL//ild/,,,tlll..Mi ,LI,, '11~1
....
ist iVL
i,iii "~'!
IL
i
i
,i i
i
iJ
[
~ili~i~i~ii'Wlili.,¶li'~li~ll '~W W V
coSt I l~ " ~ coVL~ i i l / I
~- 1 I ' ' ~ rl l i i ' ,,il I l l
1
~
W
l
~- I ~i- .,.i. ii~' ls
Fig. 2. Rhythmic antidromic discharges of single primary afferents in the proximal stump of cut dorsal roots during fictive locomotion in a spinal cat and real locomotion in a decorticate cat. A: a S1 unit discharging during the activity of the nerve to the ipsilateral Vastus Lateralis (iVL) in an acute low spinal cat injected with nialamide and L-DOPA. B: in the same preparation, a S1 unit discharges before and throughout the activity of iSartL nerve, the iVL nerve being silent. C: decorticate cat walking on a treadmill belt; the electromyograms are recorded from the Semitendinosus (St) and the VL muscles ipsilaterally (i) and contralaterally (co) with implanted copper wires. The unit was recorded in a long filament of L7 which provided stable recordings despite the movements of the animal.
two dorsal root units in S1 discharging antidromically at the periodicity of the locomotor rhythm. In Fig. 2A, the unit followed quite closely the output of the extensor nerve Vastus Lateralis (iVL). For instance, in the 4th burst, the unit discharges for a longer time and at a higher frequency and corresponds to the larger and longer burst of activity in the iVL neurogram. The 5th burst of the unit is of shorter duration corresponding also to an iVL activity which is shorter and smaller than in the previous cycle. In the same experiment, but in a different S1 dorsal rootlet (Fig. 2B), one unit discharges throughout the period of activity of the flexor iSartL and even starts before the onset of activity in that nerve. Note that there was no activity in the iVL nerve in this case. Finally, it was of interest to know if such units could similarly be recorded in animals actually walking on a treadmill. Two cats were decorticated and the electromyographic (EMG) a&ivity of selected hindlimb muscles was recorded. Fig. 2C shows a unit in L7 which is gradually recruited as the walking pattern becomes well organized. This unit fires mainly throughout the period of activity of the ipsilateral Semitendinosus (iSt) and starts to discharge just at the transition between the end of iVL and the start of iSt. This pattern of firing recorded in a walking cat, closely resembles the unit recorded in the paralyzed cat shown in Fig. 1A. The present results indicate clearly that, during real and fictive locomotion, the primary afferents are subjected to changes in polarization that may lead to antidromic firing which can be recorded as unitary discharges in cut dorsal root filaments. Spontaneous antidromic unitary discharges in cut dorsal root filaments have been recorded before in anesthetized cats ]4. Unit discharges related to rhythmic DRPs have been briefly mentioned by Bayev and Kostyuk 5 for fictive locomotion and were illustrated for fictive scratching 4. The interest of the present results is to emphasize that different primary afferents may be subjected to different depolarizing drives. Indeed, some afferents fired during the periods of flexor and/or extensor activity or else near transitional phases between the flexors and the extensors in the step cycle. This complex pattern of firing is more in keeping with the already well-documented complexity of the centrally generated locomotor patternSA3 and suggests that the transmission in primary affer-
378 ents themselves is under close control of the central pattern generator. O n e could postulate from these results that the transmission in some afferents will be m o d u l a t e d as a function of the different phases of the step cycle. It is possible that different groups of primary afferents, or afferents originating from different parts of the limb, may be influenced at different times during the step cycle. H o w e v e r , the present experimental approach did not allow the identification of the origin and type of the different afferents. The extracellular technique allows to record presumably only a fraction of units whose m e m b r a n e polarization is cyclically changed during locomotion. Indeed, experiments in progress, using intracellutar recordings of primary afferents, suggest that some units are only subjected to subliminal changes in their m e m b r a n e polarization during fictive locomotion. The present results may have two o t h e r secondary consequences. Firstly, it is generally assumed that, during fictive locomotion, only tonic afferent activity remains since the limbs are not moving. A l t h o u g h there are indeed no phasic afferent inputs originating from the stationary limbs, some primary afferents can apparently also be brought to fire cyclically by the central p a t t e r n g e n e r a t o r for locomotion. There-
1 Arshavsky, Yu. I., Berkinblit, M.B., Fukson, O.I., Gelfand, I.M. and Orlovsky, G.N., Origin of modulation in neurones of the ventral spinocerebeilar tract during locomotion, Brain Research, 43 (1972) 276-279. 2 Bayev, K.V., Periodic changes in primary afferent depolarization during fictitious locomotion by thalamic cats, Neurophysiology, 10 (1978) 316-317. 3 Bayev, K.V., Polarization of primary afferent terminals in the lumbar spinal cord during fictitious locomotion, Neurophysiology, 12 (1980) 305-311. 4 Bayev, K.V. and Kostyuk, P.G., Primary afferent depolarization evoked by the activity of spinal scratching generator, Neuroscience, 6 (1981) 205-215. 5 Bayev, K.V. and Kostyuk, P.G., Polarization of primary afferent terminals of lumbosacrat cord elicited by the activity of spinal locomotor generator, Neuroscience, 7 (1982) 1401-1409. 6 Forssberg, H., Grillner, S. and Rossignol, S., Phase-dependent reflex reversal during walking in chronic spinal cats, Brain Research, 85 (1975) 103-107. 7 Forssberg, H., Grillner, S. and Rossignol, S., Phasic gain control of reflexes from the dorsum of the paw during spinal locomotion, Brain Research, 132 (1977) 121-139. 8 GriUner, S., Control of locomotion in bipeds, tetrapods and fish. In J.M. Brookhart and V.B. Mounteastle (Eds.), Handbook of Physiology, Section 1: The Nervous System,
fore, if the afferents discharged by presynaptic depolarization can still exert some postsynaptic effects, it is possible that second order neurons could be rhythmically m o d u l a t e d in this m a n n e r during fictive locomotion (see ref. 1 as an example). Secondly. when recording from dorsal root ganglion cells whose axons are in continuity with the spinal cord in chronic preparations H, techniques such as backward spike triggered averaging should be used as done by Loeb l0 to insure that the recorded activity results from peripheral stimulation and not from a central activation as r e p o r t e d here. H o w e v e r . there is up to now. no experimental evidence that such antidromic discharges are present in intact animals. This work was s u p p o r t e d by a G r o u p grant from the Medical Research Council of Canada. J.-M.C. was supported by a fellowship from the F r a n c e - C a nada exchange p r o g r a m ( C N R S - C N R C ) as well as by the Centre de Recherche en Sciences Neurologiques of the Universit6 de Montreal. R . D . was awarded an F . C . A . C . studentship. J. Provencher, S. Bergeron, R. Bouchoux and E. Rupnik are gratefully acknowledged for their technical contributions.
Vol. II, Part 2, American Physiological Society, 1981, pp. 1179-1236. 9 Grillner, S. and Zangger, P., On the central generation of locomotion in the low spinal cat, Exp. Brain Res., 34 (1979) 241-261. 10 Loeb, G.E., Somatosensory unit input to the spinal cord during normal walking, Can. J. Physiol. Pharmacol., 59 (1980) 627-635. 11 Loeb, G.E., Bak, M.J. and Duysens, J., Long-term unit recording from somatosensory neurons in the spinal ganglia of the freely walking cat. Science. 197 (1977) 1192-1194 12 Perret, C.. Neural control of locomotion in the decorticate cat. In R.M. Herman. S. Grillner. P.S.G. Stein and D.G. Stuart IEds.), Neural Control of Locomotion. Plenum Press. New York. 1976. pp. 587-615. 13 Perret, C and Cabelguen. J.-M.. Main characteristics of the hindlimb locomotor cycle in the decorticate cat with special reference to bifunctional muscles. Brain Research. 187 (1980) 333-352 14 Repkin, A.H.. Wolf. P. and Anderson. E.G.. Non-GABA mediated primary afferent depolarization. Brain Research, 117 (1976) 147-152. 15 Wall, P.D.. Excitability changes in afferent fibre terminations and their relation to slow potentials. J. Physiol. (London). 142 (19581 1-21.