- Email: [email protected]
Central locomotor programming in the rabbit
Brain Research, 168 (1979) 57-73
57
© Elsevier/North-Holland Biomedical Press
C E N T R A L L O C O M O T O R P R O G R A M M I N G IN T H E RABBIT
CATHERINE VIDAL*, DENISE VIALA** and PIERRE BUSER Laboratoire de Neurophysiologie Compar~e, Universitd Pierre et Marie Curie, 75230 Paris - Cedex 05, (France)
(Accepted September 14th, 1978)
SUMMARY In decorticate, unanaesthetized and curarized rabbit preparations, with both hindlimbs deafferented, locomotor-like discharges were recorded from nerves to flexors and extensors and their time patterns were compared. The bursts of rhythmic activities recorded from nerves to several flexor muscles acting at either joint were shown to be synchronous, with no differences in their time of onset. The same was true for the extensors. Nerves to bifunctional muscles (biceps posterior, semi-tendinosus and tenuissimus) acting on two consecutive joints (knee flexors, hip extensors) could display two consecutive bursts in each locomotor cycle, one being pure flexor and the other an extensor discharge. It is shown here that this functional bivalence is programmed centrally and can be modulated, i.e. the relative importance of the flexor and the extensor bursts can be changed in a predictable way through afferent (proprioceptive or cutaneous) influences, or through activation of the descending monoaminergic pathways. In extreme cases, complete functional reversal was observed in these bifunctional muscle nerves.
INTRODUCTION Locomotion is currently studied in freely moving animals through kinematic methods, which indicate how different segments move and how joint angles change with time 3A1,17, and through E M G recordings. The latter technique provides information on the functional categories of the involved muscles, and on their chronology of activation within a step cycle 1°,11,16. As a result of these converging methods of investigation, hip movements, in mammalian quadrupeds, were shown to differ from knee and ankle movements; * Present address: Centre d'Etude du Syst6me Nerveux, C.N.R.S. 91190, Gif-sur-Yvette, France. * ~ To whom correspondence should be addressed.
58 correspondingly, hip flexors and part of the extensors displayed patterns of activation distinct from those of the knee and ankle. These differences were attributed either to the pattern of organization of the central generating system itself (now currently called 'locomotor programme') or to peripheral modulating influences due to the movements themselves (phasic or kinetic) or to the instantaneous posture (tonic or static) or finally to mechanical constraints. Engberg and Lundberg tl and Lundbergz2 emphasized the importance of proprioceptive reflexes in modulating a possibly 'common' central programme, i.e. providing a simultaneous activation of hindlimb flexors alternating with a simultaneous activation of the antagonistic extensors. However, it was later shown in the cat 20 that the pattern of activation of the different muscles remained unchanged after deafferentation of both hindlimbs, and that the central programme for hindlimb locomotion already included differences in activation of distinct groups of muscles. Our aim here was to investigate the pattern of organization of the central locomotor system of the domestic rabbit, and to identify some of the modulating factors which influence this organization. In decorticate, unanaesthetized and paralyzed rabbit preparation, hindlimb muscle nerves quite frequently display spontaneous sequences of discharges whose pattern closely mimicks a 'hopping' type of locomotion ('fictive locomotion') since they alternate in nerves to antagonistic muscles on one side, and are symmetrical in homonymous nerves on both sides. In these preparations, paralysis suppresses all kinetic feedback: if, in addition to this, both hindlimbs are deafferented, thus also eliminating tonic sensory control, the pattern of central locomotor generation can be studied in isolation. This was undertaken here, with systematic comparisons of discharges from a large number of hip, knee and ankle muscle nerves. A preliminary account of this work has already been published 42. METHODS Experiments were carried out on 80 rabbits (males, 2.5-3 kg body weight). Under general methohexital anaesthesia (Brietal, 10 mg/kg, i.v.), they were tracheotomized and their head was placed in a rigid stereotaxic contention frame. After craniotomy, extensive decortication was performed. Animals were then paralyzed through gallamine triethyiodide (Flaxedil, 5 mg/kg, i.v.) and placed under artificial ventilation. In most experiments, and unless specified, both hindlimbs were extensively deafferented through transsecting all cutaneous, joint and muscle nerves as indicated by Bessou and Laporte 4. The nerves to the following flexor and extensor muscles were selected for recording from their central cut ends: sartorius (Sar) and rectus femoris (RF), hip flexors; semitendinosus (ST), tenuissimus (Ten) and biceps femoralis posterior (BP), knee flexors; tibialis anterior (TA), ankle flexor; extensor digitorum longus (EDL) and brevis (EDB), physiological digit flexors, biceps femoralis anterior (BA) and semi-membranosus (SM), hip extensors; vastus lateralis (VL), knee extensor; gastrocnemius medialis (GM) and gastrocnemius lateralis-soleus (GL-Sol),
59 ankle extensors; flexor digitorum longus (FDL) and musculi lumbricales (L), physiological extensors. The dissected nerves were kept in mineral oil, poured into a pool formed by the skin flaps and maintained at body temperature (38 °C). For recording, the nerves were placed on silver electrodes connected to short time-constant conventional amplifiers. In about 20 ~ of the experiments, the left and right cutaneous sural nerves were mg/kg, i.v.) or D,L-5-HTP (150 mg/kg, i.v.). These drugs were dissolved in saline controlled at any time through electrodes placed on the proximal part of the same nerve. Single shocks of short duration (0.1 msec) were applied to the nerve using the oscilloscopic control; their intensity was adjusted to recruit the small myelinated afferents, thus exciting Aa, Bfl, A~ fibre groups. In other experiments (16 rabbits), animals were injected with L-DOPA (100 mg/kg, i.v.) or D,L-5-HTP (150 mg/kg, i.v.). These drugs were dissolved in saline through stirring and gentle heating and adjusted at pH 7 with phosphate buffer. The drugs were then immediately injected at low rate (50 mg/min) through an automatic injector. Heart rate, blood pressure and end-tidal CO2 were constantly monitored. Nerve discharges were displayed on a 4 trace oscilloscope, either directly or after storing on an 8 channel tape recorder. Running films were taken, using one or two reference nerves (in particular TA and/or G M muscle nerves) for comparison with other tested nerves. Since this paper basically deals with time relationships between discharges, the latter will be represented in most pictures by block-diagrams of real duration but standard arbitrary amplitude. RESULTS Central locomotor pattern in various muscle nerves of the deafferented hindlimb Recording locomotor-like discharges from various nerves to hindlimb flexor and extensor muscles has shown two distinct features which will be considered separately. Simple cases In previous studies32,35, ag, we had only recorded from the nerves to TA and to G M muscles. We attempted to test here whether nerves to other flexor or extensor muscles of the same limb were discharging in simultaneity* with either the TA or the G M muscle nerve, depending on their function. Our results are rather clearcut: with the used speed of film recording (100 mm/sec), no time-lag in the onset or difference in duration of the discharges could be observed between all tested flexor muscle nerves (TA and Sar, RF, Ten, ST, EDL and EDB, see examples in Fig. 1) and all extensor muscle nerves (GM and BA, SM, VL, * It is true that conduction distances between spinal cord and recording sites were not identical on different nerves. However, with respect to the recording speed, lags which could thus be introduced are negligible.
60 Sar ST TA EDB
SM VL GM TA
Sar BA TA GM
500 ms
Tell
VL TA GM Fig. 1. Locomotor discharge timing in nerves to hindlimb flexor and extensor muscles. Examples of spontaneous locomotor activities recorded in different muscle nerves during the same experiment. Upper record: simultaneous record from different physiological flexor muscle nerves; Sar (hip joint), ST (knee joint), TA (ankle joint) and EDB (digit joints). Second record: simultaneous record from 3 physiological extensor muscle nerves; SM (hip joint), VL (knee joint) and GM (ankle joint) and from an ankle flexor nerve (TA). Third and fourth records: compared burst timing in different flexor and extensor muscle nerve3; Sar and BA (hip flexor and extensor), Ten and VL (knee flexor and extensor) TA and GM (ankle flexor and extensor).
G L - S o l , F D L a n d L, e x a m p l e s in Fig. l). T h e previously o b s e r v e d a l t e r n a t i o n p a t t e r n between T A a n d G M nerve discharges thus also exists between o t h e r flexor and e x t e n s o r muscle nerves (Fig. 1). Therefore, the discharges in a n t a g o n i s t i c T A a n d G M nerves will serve as 'reference activities' in this study.
Particular activation characteristics o f Ten, S T and BP nerves Three muscle nerves d i s p l a y e d p a r t i c u l a r p a t t e r n o f activation.
61 A Ter]
....
ST TA GM
.V. Ju ~ J
~J:.t
C
B •
W
D
v
E
,.imm~
~r~r
fit ii
nil
~
-
1 sec
F
F'
G
-G '
Ten ST TA GM
= w
Fig. 2. Discharge pattern in nerves to flexor, to extensor and to bifunctional muscles. Simultaneous recording from nerves to tenuissimus (Ten), semi-tendinosus (ST), tibialis anterior (TA) and gastrocnemius medialis (GM). A, B, C, D and E are examples from different experiments (for details, see text). F and F' were recorded from the same experiment, but the general state of excitability of the animal was higher in F' than in F. G and G': another example of such increase in the general excitability of the animal in another rabbit (see text).
62 Ten and STnerves. The behaviour of Ten and ST nerves as typical flexor nerves was not the most frequently encountered in our experiments. In some preparations, Ten and ST nerves discharged continuously throughout the locomotor sequence, with a maximum density during the TA flexor burst (Fig. 2A). In most preparations, these nerves could display two separate bursts in each cycle, one during the TA flexor activity, and one during the GM extensor activity (Fig. 2B, C, D). The 'flexor burst' could be shorter than that in TA, although it started most often simultaneously. The 'extensor burst' could also be shorter than that in G M and in addition its onset could also be delayed. Finally, in still other preparations, ST nerve could display clearcut extensor type discharges with a silence during flexor activity (Fig. 2E). These various aspects o f T e n and ST discharge patterns could be observed in the same animal, depending on its overall locomotor activity or its 'excitation level' (which could undergo considerable changes with time in this type of preparation). This is illustrated on Fig. 2 (F-G'): with low rates of spontaneous TA bursts (F), Ten and ST nerves were discharging together with TA while GM was silent. In F', a slight increase in TA rhythms took place, with G M nerve now also discharging; ST nerve then displayed double bursts while the Ten pattern of discharge remained unchanged. In G', extensor-type discharges were added to the flexor ones on Ten and ST nerves, which last ones were the only visible in a lower state of excitation of the animal (in G). BP nerve. The nerve trunk to BP divides into 3 branches before entering into 3 corresponding parts of the muscle. The anterior branch (BPa) contains a few muscle fibres with a majority of cutaneous fibres, while the medial (BPm) and posterior (BPp) branches are only muscular. When recording from the entire BP nerve trunk, the obtained discharges were of long duration, simultaneous with both flexor and extensor activities (Fig. 3A). Recording separately from the 3 nerve branches gave the following results: BPa discharges (when present) were always of the flexor type (Fig. 3B, C, D); those in BPm were, in most preparations, of the extensor type (Fig. 3B, D) and in some cases, of a mixed type (Fig. 3C); BPp discharges displayed the largest variations. resembling those observed in Ten and ST nerves. The usually accepted functional classification of BP as a flexor muscle is based on its participation to the flexion reflex31. Therefore, we tested separately the BPa, BPm and BPp branches to a single shock stimulation of all group A fibres of the ipsilateral sural nerve (see methods). A reflex discharge developed in BPa, BPm and BPp branches during periods of locomotor rest when these nerves displayed a 'flexor ~ participation to the locomotor pattern just before testing for reflex discharges (Fig. 3C and D). Thus, the functional differences between the 3 BP nerve branches as observed during fictive locomotion are also observed when these branches are tested for their participation to the ipsilateral flexion reflex in static conditions. Table I summarizes the frequencies of occurrence of the various patterns of activation of Ten, ST and BP in 60 out of 80 preparations which demonstrated stability throughout the experiment (5 h). As can be seen, BPa activities were constantly of the flexor type, while double activation could be widely observed in Ten, ST and BPp nerves. Exclusive extensor activation most often occurred in BPm nerve, seldom in BPp and ST and never in Ten nerve.
63
A BP TA III
GM
|
II l i b
II
I
I
II
I
I
I I
B BPa
I
BPrn
BPp TA GM
C •lll•llllllllllml,
. . . . . . . . . . . . . . . . . . . ._
]L:.T,L-_:
,-1..,,,
'
II
i
;L ;/;.LI
_..
Ji .......... I
2 5 c,s I
500 ms
D . L,.m~,
L. LA.L l
i
m
II
.ua~.
•~
, , ~
-~r ~
I
,
A
~
"'~'J'
.r~.~,
~ . - . ,
', ...... I.Jl
I ~
~f~'
~
a
m
i
' ~,,.
•
I
•
Fig. 3. Comparison of locomotor discharge pattern in nerves to TA, G M and nerve branches to BP, a bifunctional muscle (anterior, a; medial, m; posterior, p). A, B, C and D are examples from different experiments. In C and D, the ipsilateral flexor reflex response was tested in the corresponding nervous branch through stimulation of the ipsilateral sural nerve (single shock, recruiting all A fibres, see Methods) during locomotor pauses, i.e. between spontaneous locomotor sequences.
J
64 TABLE I Frequencies of occurrence of the variouspatterns of activation displayed in nerves or nerve branches to bifunctional muscles (knee flexor and hip extensor muscles)
Numbers indicate per cent of experiments (60 animals) in which one of the 3 central patterns of activation was encountered in each muscle nerve. BP: biceps posterior (a, anterior; m, medial; p, posterior) nerve. ST: semi-tendinosus muscle nerve. Ten: tenuissimus muscle nerve. Pattern of activation/ Muscular nerves
Flexor
Extensor
Mixed
BPa BPm BPp ST Ten
100 0 25 42 56
0 91 9 8 0
0 9 66 50 44
Modulation o f the central locomotor pattern o f central or peripheral origin
In the above section, fictive locomotion was described as it occurs when neither phasic afferences (immobilized animal) nor even tonic ones (deafferented limbs) could interfere. H o w this 'central locomotor p r o g r a m m e ' can be altered ('modulated') by various factors will be considered now. Posture o f the intact limb
This modulating factor had already been briefly reported earlier as: in curarized preparation, but with intact (uncut) afferences, the limb posture influences the relative importance of flexor and extensor nerve activities. We considered this point in more details here in 20 preparations where no deafferentation was performed. In fact, for surgical reasons, ST was the only bifunctional muscle which was left inserted at its both ends and on which passive limb position could be studied then. (i) When both hindlimbs were placed in an intermediate position (half flexed) the locomotor discharge pattern closely resembled those of the deafferented limbs. (ii) With the afferented limbs being placed in a complete extended position, the balance of activities was modified in all tested animals: activities in pure flexor nerves prevailed, with longer duration than those in extensor nerves, and the ST nerve now only displayed a long flexor discharge (Fig. 4, 'EXT', A, B, C, D). Correspondingly, the locomotor sequence most usually started with a flexor discharge (in 75 % of the experiments; see examples in Fig. 4, B and D). (iii) The opposite feature was observed with both hindlimbs placed in forced flexion: activity in G M nerve was more developed than that in TA, while ST nerve displayed various possible patterns (Fig. 4, ' F L E X ' ) : short flexor burst (A), silence (B), extensor burst (C) or double short burst, one flexor and one extensor (D). The sequence this time always began with an extensor burst. It is thus interesting to note that in the same preparation, ST discharges could change from the pure flexor to the pure extensor type (Fig. 4C), depending on whether the hindlimbs were completely extended or flexed.
65 EXT
FLEX
A
TA ~
,
•
•
1~
~]
•
.mmnm sr gYAWAgTA B
i
I •
i
•
C
I
•
•
•
•
D
l
n
•
i
•
iiim is
Fig. 4. Influence of the afferented limb posture on central locomotor activation of ST as compared to that of TA and GM. Locomotor discharges are schematized and their only duration and moment of occurrence on the time axis is considered (see text). Black: TA and GM discharges. Hatched: ST discharges. A, B, C and D: examples from 4 different animals whose both hindlimbs were first placed in a complete extended position (left column, EXT), then in a complete flexed position (right column, FLEX).
Sural nerve stimulation in deafferented limb discharges Going back to the deafferented hindlimb preparation, we could observe that the repetitive 10/sec stimulation of both large and small myelinated afferents delivered to the ipsi- or the contralateral sural nerve could alter the discharge patterns ~2, particularly those in the ST and Ten muscle nerves. Compared to the locomotor discharges occurring spontaneously (Fig. 5T), those elicited through stimulation of the contralateral sural nerve were modified as follows: (i) the first discharge was often of the extensor type whereas it was merely of the flexor type in a locomotor sequence occurring spontaneously; (ii) T A discharges were relatively abredged and those in G M were lengthened and (iii) Ten and ST activities shifted towards the extensor pattern. Stimulation of the ipsilateral sural nerve had the opposite effect, that of favouring the flexor bursts which always initiated the elicited locomotor sequence; in Fig. 5 (A and B), Ten and ST nerves now displayed pure flexor activities during stimulation whatever their prior discharge pattern. Similar effects (not illustrated here) were obtained with the BPp nerve.
66 T
A
n m •
TA
GM
ARFc.
ARFi.
,
n
__m_mm_
m
n
Ten ST
B
llll
iI_lt_
mm m m
1s I
#
Fig. 5. Influence of sural nerve stimulation in deafferented hindlimb discharges of TA, G M , Ten and ST muscle nerves. T: control spontaneous activities. A R F c : activities during repetitive stimulation (10/sec) of all cutaneous A fibres of the contralateral sural nerve (see Methods). ARFi : same type of stimulation applied to the ipsilateral sural nerve. A and B records obtained in two different experiments.
Pharmacologicalactivation of monoaminergic descendingpathways Our previous studies had indicated that systemic injection of DOPA, a catecholamine precursor, in intact or decorticate preparations, leads to an increase in duration of the GM discharges with respect to those in TA. An opposite effect was produced after injection of 5-HTP, the precursor of serotonin 3~,35. These pharmacological actions on fictive locomotion were particularly tested here on bifunctional muscle nerves in 20 preparations. Fig. 6 illustrates 4 different cases of the effect of DOPA. Control discharges occurred in BPp (case A) and BPm (case B) during both 'flexion' and 'extension'; in both cases, DOPA clearly abolished the flexor burst. An extreme case is also shown (Fig. 6, D) where DOPA completely reversed the pattern of discharge of Ten and ST from flexor to extensor one. Fig. 7 provides examples of the inverse effect of 5-HTP; total reversals (from extensor to flexor burst now) could still be obtained. DISCUSSION
In curarized rabbit, locomotor-like discharges which develop spontaneously (i.e. in the absence of an identified cause) in the muscle nerves of a deafferented hindlimb (with no sensory control from the tonic receptors) are the expression of the central locomotor programme. Summing up our data, it appears that all flexor muscles, whatever their anatomical level of action, proximal or distal, are simultaneously
67 T
A
DOPA
JIJLII
GM ~ BPa [~ ~
T
C ,,
~
I I ~J
~J
"
DOPA
lIB
GM J t - B L m t ~ Ten [~1[~J~1 [~/]
g'AlleTAlg'/ l g'AIIF/Atlg'/~l
D
B
_ii__it,Ji_
iRe
•
•
I
I
I__
I
I
I
m e
I
ls
L
Fig. 6. Effect of DOPA on the firing pattern of TA, GM and bifunctional muscle nerves. A, B, C and D: groups of records from 4 different experiments. In each case, locomotor activities are shown before (T) and 30 min. after a L-DOPA injection (100 mg/kg, I.V.). In A and B, BPa, BPm and BPp discharges are considered, and those ofTen, ST in C and D, TA and GM discharges being in each case taken as reference. In control records, BPp (A) and BPm (B) displayed a flexor and an extensor discharge; the flexor component was abolished by DOPA. In C, Ten and ST also displayed flexor and extensor discharges in controls. After DOPA administration, extensor discharges became predominant. D is an extreme case where Ten and ST displayed flexor discharges before and extensor discharges after DOPA injection.
activated, in clear alternation with the activation of the antagonistic extensor muscles. In the cat, various bifunctional muscles whose main function is exerted at hip level, but which have also an anatomically opposite function at knee joint level, can display double EMG bursts 11. When studied in the rabbit, the central activation pattern of these muscles was always unique, consisting either in flexor (in Sar and RF muscle nerves) or in extensor activations (in SM muscle nerve). None of the different means used in our experiments to increase or modulate the central pattern activation (as in Results, second section) allowed us to disclose a double central programming for these muscles. Another interesting difference concerns EDB activation which occurred in the rabbit together with that of the general set of flexors, while in the cat EDB behaves like an extensor during locomotion1°,11,2° although it was long considered as a flexor when tested for its reflex activations from the cutaneous afferents3L This difference may be due to the digitigrad progression of the cat, as opposed to the plantigrad pacing of the rabbit. It can be concluded from the above mentioned differences that in the rabbit (on the contrary of the cat), hip (Sar, RF) and digit (EDB) agonistic muscles receive the same simple central programming as the ankle muscle considered as reference (TA). The same occurs for SM compared to GM reference muscle. Besides these simple cases in the rabbit, 3 muscles (ST, BP and Ten) mainly regarded as knee flexors by physiology but anatomically related to bifunctional
68 T
5-HTP
A
o.
m m
BPm
~
n
n
m
~::~
B
n
m
C
I
n
iS
I
TA GM
m |
•
Ten
ST Fig. 7. Effect of 5-HTP on the firing pattern of TA, GM and BP (A and B) and Ten, ST (C) muscle nerves. A, B and C: examples provided by 3 different experiments. Locomotor activities are shown before (T, left column) and 30 min after OL-5-HTP administration, 150 mg/kg i.v. (right column). In A and B, BPm discharged as an extensor before drug action. These discharges were suppressed (A) or reduced (B) after 5-HTP injection. In B, BPp and in C, Ten and ST, displayed both flexor and extensor control discharges. The extensor discharge was suppressed in B or reduced in relative importance in C.
muscles (knee flexors and hip extensors) clearly displayed in our experiments a complex pattern of central activation: Ten and ST nerves behaved either as flexors or as extensors, and ST could even exhibit a pure extensor activity. As for ST, a complex distribution of activations was observed in the three nervous branches innervating BP muscle: BPa branch always conveyed flexor type activities, BPm almost exclusively pure extensor type ones, and BPp branch exhibited a pattern of discharge closely resembling that of ST. It is interesting that, when eliciting the ipsilateral flexion reflex during locomotor pauses, this reflex occurred in the muscle nerves which displayed flexor or flexor and extensor bursts, but was never obtained when extensor bursts were only present. The conclusion we thus reach, from inspection of the discharge pattern in these bifunctional muscle nerves in deafferented limb, is that the double function is also part of the central programme. These data fit with Grillner and Zangger's data 20 in cat but not with those of Perret and Calbelguen 27 who consider the flexor burst as being of reflex origin.
69 The next correlated question is whether a given a motoneurone can receive both types of activation, or whether functionally distinct and specialized subpools of motoneurones exist. The frequent opposite pattern of bursty firing in two nervous branches to the same BP muscle (flexor type in BPa and extensor type in BPm branches) seems to favour the second hypothesis. Recent data obtained in the same type of rabbit preparation with recordings from a ST, BP or Ten isolated nerve fibresa4 support this view: all investigated fibres displayed only one type of activity, either flexor or extensor, at the same time when the entire nerve disclosed double discharges. All these observations also seem to be at variance with those obtained in the cat where double bursts were described in the same ST motoneuronesza. Another point of discussion concerns the durations of the double discharges in bivalent muscle nerves. These were generally shorter than the reference TA or GM discharges. The flexor bursts started most often together with TA bursts, but they were briefer. The same was observed for the EMG bursts recorded in chronic spinal cats on a treadmill: bursts in ST were shorter than those in TA which occupied the half cycle9. On the other hand, a short latency flexion reflex (evoked in ST through weak ipsilateral electrical stimulation of the foot dorsum) was observed by Forssberg et al. 15 in ST throughout the half-cycle, even after termination of the ST flexor burst. This phasic reflex modulation is central2, 40 and would be achieved at least by a phasic (rhythmic) influence from the spinal generator of locomotion on short latency reflex pathways to flexors or extensors, depending on the phase of the step cycle2. These data thus indicate that interneurones mediating short latency flexor reflex to ST motoneurones get, from the generator, the same temporal facilitation than the TA pool of motoneurones. This would suggest that: (a) the spinal generator would have a common action towards both groups of motoneurones, TA and ST flexor type ones; this is consistent with the possible development, in our experiments (see Fig. 1, and 2), of ST and Ten flexor bursts with simultaneous onset and same duration as TA reference bursts in cases when the only flexor activation is conveyed to these bifunctional muscles and (b) the shortening of discharges in bifunctional muscles would then be achieved through private inhibitory interconnections of last order interneurones on the subpools of motoneurones; these interneurones have to be independent from those transmitting the short latency reflex effects. The closer temporal link between onset of TA and flexor bursts in bivalent muscles, unlike extensor ones compared to GM bursts, might be linked to asymetry in the genesis of locomotor activities to the flexor and extensor motoneurones: in the rabbit, as in the cockroach~6, the spinal generator of locomotion appears as a flexor oscillator driving the flexor motoneurones while inhibiting the extensor motoneurones37,41. It was shown in the second part of this work that external influences can drastically modify the pattern of locomotor discharges. Changes in the balance between flexor and extensor activations were especially striking when taking the bivalent knee flexors-hip extensors (Ten, ST and BP muscle nerves) as indices. As a matter of fact, in our experiments, two distinct classes of actions were isolated: changes in the limb position, modifying the tonic inflow from muscles and
70 joints, and the effect of repetitive stimulation of cutaneous afferents (with shocks recruiting all myelinated fibres, thus including presumably 'noxious' afferents). 'Tonic reversals' (provoked by passive changes of limb position) had already b~en described for various 'reflex' activities evoked from cutaneous afferents, such as short latency reflexes23, ~4,3°, spino-bulbo-spinal reflexes (G. Viala, unpublished data) and finally late long lasting dischargeslS,19, ~9 described and studied by Lundberg's group 1,21. The latter switching displays direct links with the modulations observed in our experiments when reversing passively hindlimb position, since late discharges can in fact be considered as the first term of a sequence of locomotor discharges in spinal 36 and in intact 4° or decorticate preparations 32 as well. Passive limb position has a functional significance at onset of the locomotor sequence since it will determine the first locomotor movement to be performed: a large extensor burst will initiate the sequence when hindlimbs are flexed, and vice-versaaS; this influence was also clearly observed in our experiments when limbs could be placed in extreme positions (Fig. 4B, D). In our particular experimental conditions (unchanged proprioceptive inputs throughout a spontaneous locomotor sequence in curarized rabbit), it is noteworthy that, besides the initial 'reversal', passive limb position also modulated the relative duration of each antagonistic burst of the sequence. The latter influence might even totally modify ST burst activation (Fig. 4C): flexor type discharges were disclosed during passive extension, while extensor type discharges were the only present during passive flexion. In the present work, the relative importance of joint and/or muscle afferents in above modulations was not stated. Muscle inflow might be of importance since reversals of late discharges were achieved from such afferents in the cat 29. Other modulations of flexor and extensor activations and, here again, some total reversals in bifunctional muscle discharges were displayed through cutaneous nerve stimulation by changing the stimulated side. These reversals should be clearly distinguished from 'phase dependent reflex reversals' during spontaneous locomotion 12,14,15, already mentioned above: they correspond to a control of short latency reflexes from cutaneous afferents by the generator of locomotion 2. We were interested here, as others 5-s,2~, in the effect of cutaneous afferents on the pattern of locomotor bursting. The point which may be raised concerns the site of action of these influences. Actually, two distinct modifications have to be distinguished which have distinct meanings: changes in the overall amount of discharges ('activation'), and changes in the balance of flexor and extensor phase ('modulation'). Activatory actions may be exerted at a 'pregenerator' level, where triggering of the generator would occur by a desinhibition processa2,3a, 41, modulations on the contrary are likely exerted at the generator level itself: desinhibition of the spinal generator, because of its all-or-none character 37, can indeed account neither for shortening of the step cycle (in elicited activities compared to spontaneous ones) nor for modulation of the balance between flexor and extensor activities; both variables can only be ruled by direct inputs on the generator itself. With this difference in mind, one can stress the point that changes in the tonic postural information related to limb position (flexed or extended) are only modula-
71 tory, but not excitatory. On the other hand, the effect of cutaneous nerve stimulation on locomotor activities revealed more complex, since this stimulation elicited both activation (triggering and holding locomotor discharges) and modulation (controlling the flexor/extensor balance). It must therefore be considered that the activatory effect of cutaneous afferents is exerted on the 'trigger' pregenerator interneurones with tonic influence on the generator itself 33, while modulation of the flexor/extensor balance may underlie an action on the generator itself. In the same line, it was assumed that, in the cat, the resetting of spontaneous locomotor activities by short trains of cutaneous stimuli (applied on the plantar surface of the foot) would be achieved by action of cutaneous afferents on the rhythm generator 5. It is still interesting to notice that DOPA and 5-HTP which had been previously32 shown to act on bulbo-spinal monoaminergic pathways could also modify the relative development of flexor and extensor bursts in the rabbit, again with both activatory and modulatory effects which would occur at two distinct spinal levels. In this respect, systemic injection of DOPA was shown efficient on the subset of 'trigger' interneurones, even in the spinal preparation 3a,41. Finally, 3 sites of action are assumed to account for the results of various modulating factors on the pattern of locomotor discharges: (1) a pregenerator level on which inhibitory effects from cutaneous afferents and monoaminergic bulbo-spinal endings are converging 3a, releasing so the spinal generator of locomotion by a desinhibition process 33,41, (2) the generator level where peripheral (cutaneous and/or proprioceptive) inputs and monoaminergic terminals may modulate the basic pattern of antagonistic flexor/extensor activation and (3) a premotoneuronal level, considered for bifunctional muscles (BP, ST and Ten), with postulated private inhibitions to the flexor and extensor subpools of motoneurones; such inhibitions would increase the contrast between flexor and extensor activations, possibly leading to reversals. It is concluded that the central locomotor programming in the rabbit is less diversified than in the cat, what might be correlated with the stereotyped locomotion in that species. From the different accounts outlined above, it is postulated that the spinal flexor generator would drive simultaneously all hindlimb flexor motoneurones and inhibit tonic activity in all extensor motoneurones. The more subtle patterns observed in the central programming (for BP, ST and Ten in the rabbit) would be achieved at a preterminal level (premotoneuronal), by more or less important inhibition of the generator drive from last order interneurones.
REFERENCES 1 Anden, N. E., Jukes, M. G., Lundberg, A. and Vyklicky, L., The effect of DOPA on the spinal cord. I. Influence on transmission from primary afferents, Acta physiol, scand., 67 (1966) 373-386. 2 Andersson, O., Forssberg, H., Grillner, S. and Lindquist, M., Phasic gain control of the transmission in cutaneous reflex pathways to motoneurones during 'fictive' locomotion, Brain Research, 149 (1978) 503-507. 3 Arshavsky, Y. I., Kots, Y. M., Orlovsky, G. N., Rodionov, I. M. and Shik, M. L., Investigation of the biomechanics of running by the dog, Biophysics, 10 (1965) 737-746 (Translation from the Russian Journal Biofizika).
72 4 Bessou, P. et Laporte, Y., Technique de pr6paration d'une fibre aff6rente Iet d'une fibre aff6rente I 1 innervant le m6me fuseau musculaire, chez le chat, J. Physiol. (Paris), 57 (1965) 511-520. 5 Duysens, J., Reflex control of locomotion as revealed by stimulation of cutaneous afferents in spontaneously walking premammillary cats, J. NeurophysioL, 40 (1977a) 737-751. 6 Duysens, J., Fluctuations in sensitivity to rhythm resetting effects during the cat's step cycle, Brain Research, 133 (1977b) 190-195. 7 Duysens, J. and Pearson, K. G., lpsilateral extensor reflexes and cat locomotion. In R. M. Herman, S. Grillner, P. Stein and D. Stuart (Eds.), Neural Control of Locomotion, Plenum Press, New York, 1976, pp. 763-765. 8 Duysens, J. and Pearson, K. G., The role of cutaneous afferents from the distal hindlimb in the regulation of the step cycle of thalamic cats, Exp. Brain Res., 24 (1976) 245-256. 9 Edgerton, V. R., Grillner, S., Sj6str6m, A. and Zangger, P., Central generation of locomotion in Vertebrates. In R. M. Herman, S. Grillner, P. Stein and D. Stuart (Eds.), Neural ControlofLocomotion, Plenum Press, New York, 1976, pp. 439-464. 10 Engberg, 1., Reflexes to foot muscles in cat, Acta physiol, scand., 62 suppl. 235 (1964) 1-64. 11 Enberg, I. and Lundberg, A., An electromyographic analysis of muscular activity in the hindlimb of the cat during unrestrained locomotion, Aeta physiol, sccnd., 75 (1969) 614-630. 12 Forssberg, H., Grillner, S. and Rossignol, S., Phase dependent reflex reversal during walking in chronic spinal cats, Brain Research, 85 (1975) 103-107. 13 Forssberg, H., Grillner, S. and Rossignol, S., Phase dependent reflex reversal during locomotion in cats. In a Communication to Third International Symposium on Motor Control, Varna, 1976. 14 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. 15 Forssberg, H., GriUner, S., Rossignol, S. and Wallen, P., Phasic control of reflexes during locomotion in Vertebrates. In R. M. Herman, S. Grillner, P. Stein and D. Stuart (Eds.), Neural Controlof Locomotion, Plenum Press, New York, 1976, p. 647-674. 16 Gambarian, P. P., Orlovsky, G. N., Protopopova, T. G., Severin, F. B. and Shik M. L., The activity of muscles during different gaits and adaptative changes of moving organs in family Felidae. Morphology and ecology of vertebrates, Proc. Inst. Acad. Sei. U.S.S.R. Zool. Sci., 48 (1971) 220-239. 17 Goslow, G. E., Reinking, R. M. and Stuart, D. G., The cat step cycle: hind limb joint angles and muscle lengths during unrestrained locomotion, J. Morph., 141 (1973) 1-41. 18 Grillner, S., Locomotion in the spinal cat. In R. B. Stein, K. G. Pearson, R. S. Smith and J. B. Redford (Eds.), Control of Posture and Locomotion, Plenum Press, New York, 1973, pp. 515-535. 19 Grillner, S. and Rossignol, S., Contralateral reflex reversal controlled by limb position in the acute spinal cat injected with clonidine, Brain Research, 144 (1978) 411-414. 20 Grillner, S. and Zangger, P., How detailed is the central pattern generation for locomotion? Brah~ Research, 88 (1975) 367 371. 21 Jankowska, E., Jukes, M. G., Lund, S. and Lundberg, A., The effect of DOPA on the spinal cord. 5. Reciprocal organization of pathways transmitting excitatory action to alpha motoneurones of flexors and extensors, Acta ph.vsiol, stand., 70, 369-388. 22 Lundberg, A., Refex control of stepping. In The Nansen MemorialLecture V, Universitetsforlaget, Oslo, 1969, p. 1-42. 23 Magnus, R., Zur Regelung des Bewegungen durch das Zentralnervensystem. Mitteilung II, Pfliigers Arch. ges. Physiol., 130 (1909) 253-269. 24 Magnus, R., In Korperstellung, Springer, Berlin, pp. 24-49. 25 Pearson, K. G. and Duysens, J., Function of segmental reflexes in the control of stepping in cockroaches and cats. In R. M. Herman, S. Grillner, P. Stein and D. Stuart (Eds.), Neural Control o[ Locomotion, Plenum Press, New York, 1976, pp. 519-537. 26 Pearson, K. G., Fourtner, C. R. and Wong, R. K., Nervous control of walking in the cockroach. In R. B. Stein, K. G. Pearson, R. S. Smith and J. B. Redford (Eds.), Control of Posture andLocomotion, Plenum Press, New York, 1973, pp. 495-514. 27 Perret,C. and Calbelguen,J. M., Central and reflex participation in the timing oflocomotor activations o f a bifunctional muscle, the semi-tendinosus, in the cat, Brain Research, 106 (1976) 390-395. 28 Perret, C. and Calbelguen, J. M., Efferent organization of the hindlimb locomotor cycle in the cat. In Satellite Symposium on Neurophysiological Mechanisms of Locomotion, Paris, 1977. 29 Rossignol, S., Ipsi- and contralateral reflex adaptation to locomotor activity. In Satellite Symposium on Neurophysiological Mechanisms of Locomotion, Paris, 1977. 30 Sherrington, C. S., On the innervation of antagonistic muscles. Sixth note, Proc. Roy. Soc. B, 66 (1900) 66-67.
73 31 Sherrington, C. S., Flexion reflex of the limb, crossed extension reflex and reflex stepping and standing, J. Physiol. (Lond.), 40 (1910) 28-121. 32 Viala, D., Etude Neurophysiologique et Pharmaeologique de FAction de Deux Syst~mes Monoaminergiques Centraux Descendants sur les Activit~s Locomotrices chez le Lapin, Th~se de Doctorat d'l~tat (Paris), 1975, pp. 1-201. 33 Viala, D., Donn6es sur l'activation du g6n6rateur spinal de la locomotion par les votes noradr6nergiques et par les aff6rences du r6flexe de flexion, J, Physiol. (Paris), 72 (1975) 31-32A. 34 Viala, D., Activation centrale de fibres alpha innervant des muscles bifonctionnels au cours de la locomotion fictive chez le lapin, J. Physiol. (Paris), 73 (1977) 60A. 35 Viala, D. and Buser, P., The effects of DOPA and 5-HTP on rhythmic efferent discharges in hindlimb nerves in the rabbit, Brain Research, 12 (1969) 437-443. 36 Viala, D., Valin, A. and Buser, P., Relationship between the 'late reflex discharge' and locomotor movements in acute spinal cats and rabbi ts treated with DOPA, Arch. ital. Biol., 112 (1974) 299-306. 37 Viala, D. and Viala, G.,Caract~reset concomitants de la programmationlocomotricechezlelapin, J. Physiol. (Paris), (1979) in press. 38 Viala, G., Variabilit6 de la configuration des d6charges eff6rentes d'origine supraspinate chez le lapin, J. Physiol. (Paris), 59 (1967) 307. 39 Viala, G. and Buser, P., D6charges eff6rentes rythmiques dans les pattes post6rieures chez le lapin et leur m~canisme, J. Physiol. (Paris), 57 (1965) 287-288. 40 Viala, G. andBuser~P.~Activit6s~c~m~tricesrythmiquesst6r6~typ6eschez~e~apins~usanesth6sie 16g~re, Exp. Brain Res., 8 (1969) 346-363. 41 Viala, G. and Viala, D., Elements of locomotor programming in the rabbit. In Satellite Symposium on Neurophysiological Mechanisms of Locomotion, Paris, 1977. 42 Vidal, C. and Viala, D., Programmation des activit6s locomotrices pour les diff6rents segments du membre post6rieur chez le lapin, J. Physiol. (Paris), 71 (1975) 348-349A.
57
© Elsevier/North-Holland Biomedical Press
C E N T R A L L O C O M O T O R P R O G R A M M I N G IN T H E RABBIT
CATHERINE VIDAL*, DENISE VIALA** and PIERRE BUSER Laboratoire de Neurophysiologie Compar~e, Universitd Pierre et Marie Curie, 75230 Paris - Cedex 05, (France)
(Accepted September 14th, 1978)
SUMMARY In decorticate, unanaesthetized and curarized rabbit preparations, with both hindlimbs deafferented, locomotor-like discharges were recorded from nerves to flexors and extensors and their time patterns were compared. The bursts of rhythmic activities recorded from nerves to several flexor muscles acting at either joint were shown to be synchronous, with no differences in their time of onset. The same was true for the extensors. Nerves to bifunctional muscles (biceps posterior, semi-tendinosus and tenuissimus) acting on two consecutive joints (knee flexors, hip extensors) could display two consecutive bursts in each locomotor cycle, one being pure flexor and the other an extensor discharge. It is shown here that this functional bivalence is programmed centrally and can be modulated, i.e. the relative importance of the flexor and the extensor bursts can be changed in a predictable way through afferent (proprioceptive or cutaneous) influences, or through activation of the descending monoaminergic pathways. In extreme cases, complete functional reversal was observed in these bifunctional muscle nerves.
INTRODUCTION Locomotion is currently studied in freely moving animals through kinematic methods, which indicate how different segments move and how joint angles change with time 3A1,17, and through E M G recordings. The latter technique provides information on the functional categories of the involved muscles, and on their chronology of activation within a step cycle 1°,11,16. As a result of these converging methods of investigation, hip movements, in mammalian quadrupeds, were shown to differ from knee and ankle movements; * Present address: Centre d'Etude du Syst6me Nerveux, C.N.R.S. 91190, Gif-sur-Yvette, France. * ~ To whom correspondence should be addressed.
58 correspondingly, hip flexors and part of the extensors displayed patterns of activation distinct from those of the knee and ankle. These differences were attributed either to the pattern of organization of the central generating system itself (now currently called 'locomotor programme') or to peripheral modulating influences due to the movements themselves (phasic or kinetic) or to the instantaneous posture (tonic or static) or finally to mechanical constraints. Engberg and Lundberg tl and Lundbergz2 emphasized the importance of proprioceptive reflexes in modulating a possibly 'common' central programme, i.e. providing a simultaneous activation of hindlimb flexors alternating with a simultaneous activation of the antagonistic extensors. However, it was later shown in the cat 20 that the pattern of activation of the different muscles remained unchanged after deafferentation of both hindlimbs, and that the central programme for hindlimb locomotion already included differences in activation of distinct groups of muscles. Our aim here was to investigate the pattern of organization of the central locomotor system of the domestic rabbit, and to identify some of the modulating factors which influence this organization. In decorticate, unanaesthetized and paralyzed rabbit preparation, hindlimb muscle nerves quite frequently display spontaneous sequences of discharges whose pattern closely mimicks a 'hopping' type of locomotion ('fictive locomotion') since they alternate in nerves to antagonistic muscles on one side, and are symmetrical in homonymous nerves on both sides. In these preparations, paralysis suppresses all kinetic feedback: if, in addition to this, both hindlimbs are deafferented, thus also eliminating tonic sensory control, the pattern of central locomotor generation can be studied in isolation. This was undertaken here, with systematic comparisons of discharges from a large number of hip, knee and ankle muscle nerves. A preliminary account of this work has already been published 42. METHODS Experiments were carried out on 80 rabbits (males, 2.5-3 kg body weight). Under general methohexital anaesthesia (Brietal, 10 mg/kg, i.v.), they were tracheotomized and their head was placed in a rigid stereotaxic contention frame. After craniotomy, extensive decortication was performed. Animals were then paralyzed through gallamine triethyiodide (Flaxedil, 5 mg/kg, i.v.) and placed under artificial ventilation. In most experiments, and unless specified, both hindlimbs were extensively deafferented through transsecting all cutaneous, joint and muscle nerves as indicated by Bessou and Laporte 4. The nerves to the following flexor and extensor muscles were selected for recording from their central cut ends: sartorius (Sar) and rectus femoris (RF), hip flexors; semitendinosus (ST), tenuissimus (Ten) and biceps femoralis posterior (BP), knee flexors; tibialis anterior (TA), ankle flexor; extensor digitorum longus (EDL) and brevis (EDB), physiological digit flexors, biceps femoralis anterior (BA) and semi-membranosus (SM), hip extensors; vastus lateralis (VL), knee extensor; gastrocnemius medialis (GM) and gastrocnemius lateralis-soleus (GL-Sol),
59 ankle extensors; flexor digitorum longus (FDL) and musculi lumbricales (L), physiological extensors. The dissected nerves were kept in mineral oil, poured into a pool formed by the skin flaps and maintained at body temperature (38 °C). For recording, the nerves were placed on silver electrodes connected to short time-constant conventional amplifiers. In about 20 ~ of the experiments, the left and right cutaneous sural nerves were mg/kg, i.v.) or D,L-5-HTP (150 mg/kg, i.v.). These drugs were dissolved in saline controlled at any time through electrodes placed on the proximal part of the same nerve. Single shocks of short duration (0.1 msec) were applied to the nerve using the oscilloscopic control; their intensity was adjusted to recruit the small myelinated afferents, thus exciting Aa, Bfl, A~ fibre groups. In other experiments (16 rabbits), animals were injected with L-DOPA (100 mg/kg, i.v.) or D,L-5-HTP (150 mg/kg, i.v.). These drugs were dissolved in saline through stirring and gentle heating and adjusted at pH 7 with phosphate buffer. The drugs were then immediately injected at low rate (50 mg/min) through an automatic injector. Heart rate, blood pressure and end-tidal CO2 were constantly monitored. Nerve discharges were displayed on a 4 trace oscilloscope, either directly or after storing on an 8 channel tape recorder. Running films were taken, using one or two reference nerves (in particular TA and/or G M muscle nerves) for comparison with other tested nerves. Since this paper basically deals with time relationships between discharges, the latter will be represented in most pictures by block-diagrams of real duration but standard arbitrary amplitude. RESULTS Central locomotor pattern in various muscle nerves of the deafferented hindlimb Recording locomotor-like discharges from various nerves to hindlimb flexor and extensor muscles has shown two distinct features which will be considered separately. Simple cases In previous studies32,35, ag, we had only recorded from the nerves to TA and to G M muscles. We attempted to test here whether nerves to other flexor or extensor muscles of the same limb were discharging in simultaneity* with either the TA or the G M muscle nerve, depending on their function. Our results are rather clearcut: with the used speed of film recording (100 mm/sec), no time-lag in the onset or difference in duration of the discharges could be observed between all tested flexor muscle nerves (TA and Sar, RF, Ten, ST, EDL and EDB, see examples in Fig. 1) and all extensor muscle nerves (GM and BA, SM, VL, * It is true that conduction distances between spinal cord and recording sites were not identical on different nerves. However, with respect to the recording speed, lags which could thus be introduced are negligible.
60 Sar ST TA EDB
SM VL GM TA
Sar BA TA GM
500 ms
Tell
VL TA GM Fig. 1. Locomotor discharge timing in nerves to hindlimb flexor and extensor muscles. Examples of spontaneous locomotor activities recorded in different muscle nerves during the same experiment. Upper record: simultaneous record from different physiological flexor muscle nerves; Sar (hip joint), ST (knee joint), TA (ankle joint) and EDB (digit joints). Second record: simultaneous record from 3 physiological extensor muscle nerves; SM (hip joint), VL (knee joint) and GM (ankle joint) and from an ankle flexor nerve (TA). Third and fourth records: compared burst timing in different flexor and extensor muscle nerve3; Sar and BA (hip flexor and extensor), Ten and VL (knee flexor and extensor) TA and GM (ankle flexor and extensor).
G L - S o l , F D L a n d L, e x a m p l e s in Fig. l). T h e previously o b s e r v e d a l t e r n a t i o n p a t t e r n between T A a n d G M nerve discharges thus also exists between o t h e r flexor and e x t e n s o r muscle nerves (Fig. 1). Therefore, the discharges in a n t a g o n i s t i c T A a n d G M nerves will serve as 'reference activities' in this study.
Particular activation characteristics o f Ten, S T and BP nerves Three muscle nerves d i s p l a y e d p a r t i c u l a r p a t t e r n o f activation.
61 A Ter]
....
ST TA GM
.V. Ju ~ J
~J:.t
C
B •
W
D
v
E
,.imm~
~r~r
fit ii
nil
~
-
1 sec
F
F'
G
-G '
Ten ST TA GM
= w
Fig. 2. Discharge pattern in nerves to flexor, to extensor and to bifunctional muscles. Simultaneous recording from nerves to tenuissimus (Ten), semi-tendinosus (ST), tibialis anterior (TA) and gastrocnemius medialis (GM). A, B, C, D and E are examples from different experiments (for details, see text). F and F' were recorded from the same experiment, but the general state of excitability of the animal was higher in F' than in F. G and G': another example of such increase in the general excitability of the animal in another rabbit (see text).
62 Ten and STnerves. The behaviour of Ten and ST nerves as typical flexor nerves was not the most frequently encountered in our experiments. In some preparations, Ten and ST nerves discharged continuously throughout the locomotor sequence, with a maximum density during the TA flexor burst (Fig. 2A). In most preparations, these nerves could display two separate bursts in each cycle, one during the TA flexor activity, and one during the GM extensor activity (Fig. 2B, C, D). The 'flexor burst' could be shorter than that in TA, although it started most often simultaneously. The 'extensor burst' could also be shorter than that in G M and in addition its onset could also be delayed. Finally, in still other preparations, ST nerve could display clearcut extensor type discharges with a silence during flexor activity (Fig. 2E). These various aspects o f T e n and ST discharge patterns could be observed in the same animal, depending on its overall locomotor activity or its 'excitation level' (which could undergo considerable changes with time in this type of preparation). This is illustrated on Fig. 2 (F-G'): with low rates of spontaneous TA bursts (F), Ten and ST nerves were discharging together with TA while GM was silent. In F', a slight increase in TA rhythms took place, with G M nerve now also discharging; ST nerve then displayed double bursts while the Ten pattern of discharge remained unchanged. In G', extensor-type discharges were added to the flexor ones on Ten and ST nerves, which last ones were the only visible in a lower state of excitation of the animal (in G). BP nerve. The nerve trunk to BP divides into 3 branches before entering into 3 corresponding parts of the muscle. The anterior branch (BPa) contains a few muscle fibres with a majority of cutaneous fibres, while the medial (BPm) and posterior (BPp) branches are only muscular. When recording from the entire BP nerve trunk, the obtained discharges were of long duration, simultaneous with both flexor and extensor activities (Fig. 3A). Recording separately from the 3 nerve branches gave the following results: BPa discharges (when present) were always of the flexor type (Fig. 3B, C, D); those in BPm were, in most preparations, of the extensor type (Fig. 3B, D) and in some cases, of a mixed type (Fig. 3C); BPp discharges displayed the largest variations. resembling those observed in Ten and ST nerves. The usually accepted functional classification of BP as a flexor muscle is based on its participation to the flexion reflex31. Therefore, we tested separately the BPa, BPm and BPp branches to a single shock stimulation of all group A fibres of the ipsilateral sural nerve (see methods). A reflex discharge developed in BPa, BPm and BPp branches during periods of locomotor rest when these nerves displayed a 'flexor ~ participation to the locomotor pattern just before testing for reflex discharges (Fig. 3C and D). Thus, the functional differences between the 3 BP nerve branches as observed during fictive locomotion are also observed when these branches are tested for their participation to the ipsilateral flexion reflex in static conditions. Table I summarizes the frequencies of occurrence of the various patterns of activation of Ten, ST and BP in 60 out of 80 preparations which demonstrated stability throughout the experiment (5 h). As can be seen, BPa activities were constantly of the flexor type, while double activation could be widely observed in Ten, ST and BPp nerves. Exclusive extensor activation most often occurred in BPm nerve, seldom in BPp and ST and never in Ten nerve.
63
A BP TA III
GM
|
II l i b
II
I
I
II
I
I
I I
B BPa
I
BPrn
BPp TA GM
C •lll•llllllllllml,
. . . . . . . . . . . . . . . . . . . ._
]L:.T,L-_:
,-1..,,,
'
II
i
;L ;/;.LI
_..
Ji .......... I
2 5 c,s I
500 ms
D . L,.m~,
L. LA.L l
i
m
II
.ua~.
•~
, , ~
-~r ~
I
,
A
~
"'~'J'
.r~.~,
~ . - . ,
', ...... I.Jl
I ~
~f~'
~
a
m
i
' ~,,.
•
I
•
Fig. 3. Comparison of locomotor discharge pattern in nerves to TA, G M and nerve branches to BP, a bifunctional muscle (anterior, a; medial, m; posterior, p). A, B, C and D are examples from different experiments. In C and D, the ipsilateral flexor reflex response was tested in the corresponding nervous branch through stimulation of the ipsilateral sural nerve (single shock, recruiting all A fibres, see Methods) during locomotor pauses, i.e. between spontaneous locomotor sequences.
J
64 TABLE I Frequencies of occurrence of the variouspatterns of activation displayed in nerves or nerve branches to bifunctional muscles (knee flexor and hip extensor muscles)
Numbers indicate per cent of experiments (60 animals) in which one of the 3 central patterns of activation was encountered in each muscle nerve. BP: biceps posterior (a, anterior; m, medial; p, posterior) nerve. ST: semi-tendinosus muscle nerve. Ten: tenuissimus muscle nerve. Pattern of activation/ Muscular nerves
Flexor
Extensor
Mixed
BPa BPm BPp ST Ten
100 0 25 42 56
0 91 9 8 0
0 9 66 50 44
Modulation o f the central locomotor pattern o f central or peripheral origin
In the above section, fictive locomotion was described as it occurs when neither phasic afferences (immobilized animal) nor even tonic ones (deafferented limbs) could interfere. H o w this 'central locomotor p r o g r a m m e ' can be altered ('modulated') by various factors will be considered now. Posture o f the intact limb
This modulating factor had already been briefly reported earlier as: in curarized preparation, but with intact (uncut) afferences, the limb posture influences the relative importance of flexor and extensor nerve activities. We considered this point in more details here in 20 preparations where no deafferentation was performed. In fact, for surgical reasons, ST was the only bifunctional muscle which was left inserted at its both ends and on which passive limb position could be studied then. (i) When both hindlimbs were placed in an intermediate position (half flexed) the locomotor discharge pattern closely resembled those of the deafferented limbs. (ii) With the afferented limbs being placed in a complete extended position, the balance of activities was modified in all tested animals: activities in pure flexor nerves prevailed, with longer duration than those in extensor nerves, and the ST nerve now only displayed a long flexor discharge (Fig. 4, 'EXT', A, B, C, D). Correspondingly, the locomotor sequence most usually started with a flexor discharge (in 75 % of the experiments; see examples in Fig. 4, B and D). (iii) The opposite feature was observed with both hindlimbs placed in forced flexion: activity in G M nerve was more developed than that in TA, while ST nerve displayed various possible patterns (Fig. 4, ' F L E X ' ) : short flexor burst (A), silence (B), extensor burst (C) or double short burst, one flexor and one extensor (D). The sequence this time always began with an extensor burst. It is thus interesting to note that in the same preparation, ST discharges could change from the pure flexor to the pure extensor type (Fig. 4C), depending on whether the hindlimbs were completely extended or flexed.
65 EXT
FLEX
A
TA ~
,
•
•
1~
~]
•
.mmnm sr gYAWAgTA B
i
I •
i
•
C
I
•
•
•
•
D
l
n
•
i
•
iiim is
Fig. 4. Influence of the afferented limb posture on central locomotor activation of ST as compared to that of TA and GM. Locomotor discharges are schematized and their only duration and moment of occurrence on the time axis is considered (see text). Black: TA and GM discharges. Hatched: ST discharges. A, B, C and D: examples from 4 different animals whose both hindlimbs were first placed in a complete extended position (left column, EXT), then in a complete flexed position (right column, FLEX).
Sural nerve stimulation in deafferented limb discharges Going back to the deafferented hindlimb preparation, we could observe that the repetitive 10/sec stimulation of both large and small myelinated afferents delivered to the ipsi- or the contralateral sural nerve could alter the discharge patterns ~2, particularly those in the ST and Ten muscle nerves. Compared to the locomotor discharges occurring spontaneously (Fig. 5T), those elicited through stimulation of the contralateral sural nerve were modified as follows: (i) the first discharge was often of the extensor type whereas it was merely of the flexor type in a locomotor sequence occurring spontaneously; (ii) T A discharges were relatively abredged and those in G M were lengthened and (iii) Ten and ST activities shifted towards the extensor pattern. Stimulation of the ipsilateral sural nerve had the opposite effect, that of favouring the flexor bursts which always initiated the elicited locomotor sequence; in Fig. 5 (A and B), Ten and ST nerves now displayed pure flexor activities during stimulation whatever their prior discharge pattern. Similar effects (not illustrated here) were obtained with the BPp nerve.
66 T
A
n m •
TA
GM
ARFc.
ARFi.
,
n
__m_mm_
m
n
Ten ST
B
llll
iI_lt_
mm m m
1s I
#
Fig. 5. Influence of sural nerve stimulation in deafferented hindlimb discharges of TA, G M , Ten and ST muscle nerves. T: control spontaneous activities. A R F c : activities during repetitive stimulation (10/sec) of all cutaneous A fibres of the contralateral sural nerve (see Methods). ARFi : same type of stimulation applied to the ipsilateral sural nerve. A and B records obtained in two different experiments.
Pharmacologicalactivation of monoaminergic descendingpathways Our previous studies had indicated that systemic injection of DOPA, a catecholamine precursor, in intact or decorticate preparations, leads to an increase in duration of the GM discharges with respect to those in TA. An opposite effect was produced after injection of 5-HTP, the precursor of serotonin 3~,35. These pharmacological actions on fictive locomotion were particularly tested here on bifunctional muscle nerves in 20 preparations. Fig. 6 illustrates 4 different cases of the effect of DOPA. Control discharges occurred in BPp (case A) and BPm (case B) during both 'flexion' and 'extension'; in both cases, DOPA clearly abolished the flexor burst. An extreme case is also shown (Fig. 6, D) where DOPA completely reversed the pattern of discharge of Ten and ST from flexor to extensor one. Fig. 7 provides examples of the inverse effect of 5-HTP; total reversals (from extensor to flexor burst now) could still be obtained. DISCUSSION
In curarized rabbit, locomotor-like discharges which develop spontaneously (i.e. in the absence of an identified cause) in the muscle nerves of a deafferented hindlimb (with no sensory control from the tonic receptors) are the expression of the central locomotor programme. Summing up our data, it appears that all flexor muscles, whatever their anatomical level of action, proximal or distal, are simultaneously
67 T
A
DOPA
JIJLII
GM ~ BPa [~ ~
T
C ,,
~
I I ~J
~J
"
DOPA
lIB
GM J t - B L m t ~ Ten [~1[~J~1 [~/]
g'AlleTAlg'/ l g'AIIF/Atlg'/~l
D
B
_ii__it,Ji_
iRe
•
•
I
I
I__
I
I
I
m e
I
ls
L
Fig. 6. Effect of DOPA on the firing pattern of TA, GM and bifunctional muscle nerves. A, B, C and D: groups of records from 4 different experiments. In each case, locomotor activities are shown before (T) and 30 min. after a L-DOPA injection (100 mg/kg, I.V.). In A and B, BPa, BPm and BPp discharges are considered, and those ofTen, ST in C and D, TA and GM discharges being in each case taken as reference. In control records, BPp (A) and BPm (B) displayed a flexor and an extensor discharge; the flexor component was abolished by DOPA. In C, Ten and ST also displayed flexor and extensor discharges in controls. After DOPA administration, extensor discharges became predominant. D is an extreme case where Ten and ST displayed flexor discharges before and extensor discharges after DOPA injection.
activated, in clear alternation with the activation of the antagonistic extensor muscles. In the cat, various bifunctional muscles whose main function is exerted at hip level, but which have also an anatomically opposite function at knee joint level, can display double EMG bursts 11. When studied in the rabbit, the central activation pattern of these muscles was always unique, consisting either in flexor (in Sar and RF muscle nerves) or in extensor activations (in SM muscle nerve). None of the different means used in our experiments to increase or modulate the central pattern activation (as in Results, second section) allowed us to disclose a double central programming for these muscles. Another interesting difference concerns EDB activation which occurred in the rabbit together with that of the general set of flexors, while in the cat EDB behaves like an extensor during locomotion1°,11,2° although it was long considered as a flexor when tested for its reflex activations from the cutaneous afferents3L This difference may be due to the digitigrad progression of the cat, as opposed to the plantigrad pacing of the rabbit. It can be concluded from the above mentioned differences that in the rabbit (on the contrary of the cat), hip (Sar, RF) and digit (EDB) agonistic muscles receive the same simple central programming as the ankle muscle considered as reference (TA). The same occurs for SM compared to GM reference muscle. Besides these simple cases in the rabbit, 3 muscles (ST, BP and Ten) mainly regarded as knee flexors by physiology but anatomically related to bifunctional
68 T
5-HTP
A
o.
m m
BPm
~
n
n
m
~::~
B
n
m
C
I
n
iS
I
TA GM
m |
•
Ten
ST Fig. 7. Effect of 5-HTP on the firing pattern of TA, GM and BP (A and B) and Ten, ST (C) muscle nerves. A, B and C: examples provided by 3 different experiments. Locomotor activities are shown before (T, left column) and 30 min after OL-5-HTP administration, 150 mg/kg i.v. (right column). In A and B, BPm discharged as an extensor before drug action. These discharges were suppressed (A) or reduced (B) after 5-HTP injection. In B, BPp and in C, Ten and ST, displayed both flexor and extensor control discharges. The extensor discharge was suppressed in B or reduced in relative importance in C.
muscles (knee flexors and hip extensors) clearly displayed in our experiments a complex pattern of central activation: Ten and ST nerves behaved either as flexors or as extensors, and ST could even exhibit a pure extensor activity. As for ST, a complex distribution of activations was observed in the three nervous branches innervating BP muscle: BPa branch always conveyed flexor type activities, BPm almost exclusively pure extensor type ones, and BPp branch exhibited a pattern of discharge closely resembling that of ST. It is interesting that, when eliciting the ipsilateral flexion reflex during locomotor pauses, this reflex occurred in the muscle nerves which displayed flexor or flexor and extensor bursts, but was never obtained when extensor bursts were only present. The conclusion we thus reach, from inspection of the discharge pattern in these bifunctional muscle nerves in deafferented limb, is that the double function is also part of the central programme. These data fit with Grillner and Zangger's data 20 in cat but not with those of Perret and Calbelguen 27 who consider the flexor burst as being of reflex origin.
69 The next correlated question is whether a given a motoneurone can receive both types of activation, or whether functionally distinct and specialized subpools of motoneurones exist. The frequent opposite pattern of bursty firing in two nervous branches to the same BP muscle (flexor type in BPa and extensor type in BPm branches) seems to favour the second hypothesis. Recent data obtained in the same type of rabbit preparation with recordings from a ST, BP or Ten isolated nerve fibresa4 support this view: all investigated fibres displayed only one type of activity, either flexor or extensor, at the same time when the entire nerve disclosed double discharges. All these observations also seem to be at variance with those obtained in the cat where double bursts were described in the same ST motoneuronesza. Another point of discussion concerns the durations of the double discharges in bivalent muscle nerves. These were generally shorter than the reference TA or GM discharges. The flexor bursts started most often together with TA bursts, but they were briefer. The same was observed for the EMG bursts recorded in chronic spinal cats on a treadmill: bursts in ST were shorter than those in TA which occupied the half cycle9. On the other hand, a short latency flexion reflex (evoked in ST through weak ipsilateral electrical stimulation of the foot dorsum) was observed by Forssberg et al. 15 in ST throughout the half-cycle, even after termination of the ST flexor burst. This phasic reflex modulation is central2, 40 and would be achieved at least by a phasic (rhythmic) influence from the spinal generator of locomotion on short latency reflex pathways to flexors or extensors, depending on the phase of the step cycle2. These data thus indicate that interneurones mediating short latency flexor reflex to ST motoneurones get, from the generator, the same temporal facilitation than the TA pool of motoneurones. This would suggest that: (a) the spinal generator would have a common action towards both groups of motoneurones, TA and ST flexor type ones; this is consistent with the possible development, in our experiments (see Fig. 1, and 2), of ST and Ten flexor bursts with simultaneous onset and same duration as TA reference bursts in cases when the only flexor activation is conveyed to these bifunctional muscles and (b) the shortening of discharges in bifunctional muscles would then be achieved through private inhibitory interconnections of last order interneurones on the subpools of motoneurones; these interneurones have to be independent from those transmitting the short latency reflex effects. The closer temporal link between onset of TA and flexor bursts in bivalent muscles, unlike extensor ones compared to GM bursts, might be linked to asymetry in the genesis of locomotor activities to the flexor and extensor motoneurones: in the rabbit, as in the cockroach~6, the spinal generator of locomotion appears as a flexor oscillator driving the flexor motoneurones while inhibiting the extensor motoneurones37,41. It was shown in the second part of this work that external influences can drastically modify the pattern of locomotor discharges. Changes in the balance between flexor and extensor activations were especially striking when taking the bivalent knee flexors-hip extensors (Ten, ST and BP muscle nerves) as indices. As a matter of fact, in our experiments, two distinct classes of actions were isolated: changes in the limb position, modifying the tonic inflow from muscles and
70 joints, and the effect of repetitive stimulation of cutaneous afferents (with shocks recruiting all myelinated fibres, thus including presumably 'noxious' afferents). 'Tonic reversals' (provoked by passive changes of limb position) had already b~en described for various 'reflex' activities evoked from cutaneous afferents, such as short latency reflexes23, ~4,3°, spino-bulbo-spinal reflexes (G. Viala, unpublished data) and finally late long lasting dischargeslS,19, ~9 described and studied by Lundberg's group 1,21. The latter switching displays direct links with the modulations observed in our experiments when reversing passively hindlimb position, since late discharges can in fact be considered as the first term of a sequence of locomotor discharges in spinal 36 and in intact 4° or decorticate preparations 32 as well. Passive limb position has a functional significance at onset of the locomotor sequence since it will determine the first locomotor movement to be performed: a large extensor burst will initiate the sequence when hindlimbs are flexed, and vice-versaaS; this influence was also clearly observed in our experiments when limbs could be placed in extreme positions (Fig. 4B, D). In our particular experimental conditions (unchanged proprioceptive inputs throughout a spontaneous locomotor sequence in curarized rabbit), it is noteworthy that, besides the initial 'reversal', passive limb position also modulated the relative duration of each antagonistic burst of the sequence. The latter influence might even totally modify ST burst activation (Fig. 4C): flexor type discharges were disclosed during passive extension, while extensor type discharges were the only present during passive flexion. In the present work, the relative importance of joint and/or muscle afferents in above modulations was not stated. Muscle inflow might be of importance since reversals of late discharges were achieved from such afferents in the cat 29. Other modulations of flexor and extensor activations and, here again, some total reversals in bifunctional muscle discharges were displayed through cutaneous nerve stimulation by changing the stimulated side. These reversals should be clearly distinguished from 'phase dependent reflex reversals' during spontaneous locomotion 12,14,15, already mentioned above: they correspond to a control of short latency reflexes from cutaneous afferents by the generator of locomotion 2. We were interested here, as others 5-s,2~, in the effect of cutaneous afferents on the pattern of locomotor bursting. The point which may be raised concerns the site of action of these influences. Actually, two distinct modifications have to be distinguished which have distinct meanings: changes in the overall amount of discharges ('activation'), and changes in the balance of flexor and extensor phase ('modulation'). Activatory actions may be exerted at a 'pregenerator' level, where triggering of the generator would occur by a desinhibition processa2,3a, 41, modulations on the contrary are likely exerted at the generator level itself: desinhibition of the spinal generator, because of its all-or-none character 37, can indeed account neither for shortening of the step cycle (in elicited activities compared to spontaneous ones) nor for modulation of the balance between flexor and extensor activities; both variables can only be ruled by direct inputs on the generator itself. With this difference in mind, one can stress the point that changes in the tonic postural information related to limb position (flexed or extended) are only modula-
71 tory, but not excitatory. On the other hand, the effect of cutaneous nerve stimulation on locomotor activities revealed more complex, since this stimulation elicited both activation (triggering and holding locomotor discharges) and modulation (controlling the flexor/extensor balance). It must therefore be considered that the activatory effect of cutaneous afferents is exerted on the 'trigger' pregenerator interneurones with tonic influence on the generator itself 33, while modulation of the flexor/extensor balance may underlie an action on the generator itself. In the same line, it was assumed that, in the cat, the resetting of spontaneous locomotor activities by short trains of cutaneous stimuli (applied on the plantar surface of the foot) would be achieved by action of cutaneous afferents on the rhythm generator 5. It is still interesting to notice that DOPA and 5-HTP which had been previously32 shown to act on bulbo-spinal monoaminergic pathways could also modify the relative development of flexor and extensor bursts in the rabbit, again with both activatory and modulatory effects which would occur at two distinct spinal levels. In this respect, systemic injection of DOPA was shown efficient on the subset of 'trigger' interneurones, even in the spinal preparation 3a,41. Finally, 3 sites of action are assumed to account for the results of various modulating factors on the pattern of locomotor discharges: (1) a pregenerator level on which inhibitory effects from cutaneous afferents and monoaminergic bulbo-spinal endings are converging 3a, releasing so the spinal generator of locomotion by a desinhibition process 33,41, (2) the generator level where peripheral (cutaneous and/or proprioceptive) inputs and monoaminergic terminals may modulate the basic pattern of antagonistic flexor/extensor activation and (3) a premotoneuronal level, considered for bifunctional muscles (BP, ST and Ten), with postulated private inhibitions to the flexor and extensor subpools of motoneurones; such inhibitions would increase the contrast between flexor and extensor activations, possibly leading to reversals. It is concluded that the central locomotor programming in the rabbit is less diversified than in the cat, what might be correlated with the stereotyped locomotion in that species. From the different accounts outlined above, it is postulated that the spinal flexor generator would drive simultaneously all hindlimb flexor motoneurones and inhibit tonic activity in all extensor motoneurones. The more subtle patterns observed in the central programming (for BP, ST and Ten in the rabbit) would be achieved at a preterminal level (premotoneuronal), by more or less important inhibition of the generator drive from last order interneurones.
REFERENCES 1 Anden, N. E., Jukes, M. G., Lundberg, A. and Vyklicky, L., The effect of DOPA on the spinal cord. I. Influence on transmission from primary afferents, Acta physiol, scand., 67 (1966) 373-386. 2 Andersson, O., Forssberg, H., Grillner, S. and Lindquist, M., Phasic gain control of the transmission in cutaneous reflex pathways to motoneurones during 'fictive' locomotion, Brain Research, 149 (1978) 503-507. 3 Arshavsky, Y. I., Kots, Y. M., Orlovsky, G. N., Rodionov, I. M. and Shik, M. L., Investigation of the biomechanics of running by the dog, Biophysics, 10 (1965) 737-746 (Translation from the Russian Journal Biofizika).
72 4 Bessou, P. et Laporte, Y., Technique de pr6paration d'une fibre aff6rente Iet d'une fibre aff6rente I 1 innervant le m6me fuseau musculaire, chez le chat, J. Physiol. (Paris), 57 (1965) 511-520. 5 Duysens, J., Reflex control of locomotion as revealed by stimulation of cutaneous afferents in spontaneously walking premammillary cats, J. NeurophysioL, 40 (1977a) 737-751. 6 Duysens, J., Fluctuations in sensitivity to rhythm resetting effects during the cat's step cycle, Brain Research, 133 (1977b) 190-195. 7 Duysens, J. and Pearson, K. G., lpsilateral extensor reflexes and cat locomotion. In R. M. Herman, S. Grillner, P. Stein and D. Stuart (Eds.), Neural Control of Locomotion, Plenum Press, New York, 1976, pp. 763-765. 8 Duysens, J. and Pearson, K. G., The role of cutaneous afferents from the distal hindlimb in the regulation of the step cycle of thalamic cats, Exp. Brain Res., 24 (1976) 245-256. 9 Edgerton, V. R., Grillner, S., Sj6str6m, A. and Zangger, P., Central generation of locomotion in Vertebrates. In R. M. Herman, S. Grillner, P. Stein and D. Stuart (Eds.), Neural ControlofLocomotion, Plenum Press, New York, 1976, pp. 439-464. 10 Engberg, 1., Reflexes to foot muscles in cat, Acta physiol, scand., 62 suppl. 235 (1964) 1-64. 11 Enberg, I. and Lundberg, A., An electromyographic analysis of muscular activity in the hindlimb of the cat during unrestrained locomotion, Aeta physiol, sccnd., 75 (1969) 614-630. 12 Forssberg, H., Grillner, S. and Rossignol, S., Phase dependent reflex reversal during walking in chronic spinal cats, Brain Research, 85 (1975) 103-107. 13 Forssberg, H., Grillner, S. and Rossignol, S., Phase dependent reflex reversal during locomotion in cats. In a Communication to Third International Symposium on Motor Control, Varna, 1976. 14 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. 15 Forssberg, H., GriUner, S., Rossignol, S. and Wallen, P., Phasic control of reflexes during locomotion in Vertebrates. In R. M. Herman, S. Grillner, P. Stein and D. Stuart (Eds.), Neural Controlof Locomotion, Plenum Press, New York, 1976, p. 647-674. 16 Gambarian, P. P., Orlovsky, G. N., Protopopova, T. G., Severin, F. B. and Shik M. L., The activity of muscles during different gaits and adaptative changes of moving organs in family Felidae. Morphology and ecology of vertebrates, Proc. Inst. Acad. Sei. U.S.S.R. Zool. Sci., 48 (1971) 220-239. 17 Goslow, G. E., Reinking, R. M. and Stuart, D. G., The cat step cycle: hind limb joint angles and muscle lengths during unrestrained locomotion, J. Morph., 141 (1973) 1-41. 18 Grillner, S., Locomotion in the spinal cat. In R. B. Stein, K. G. Pearson, R. S. Smith and J. B. Redford (Eds.), Control of Posture and Locomotion, Plenum Press, New York, 1973, pp. 515-535. 19 Grillner, S. and Rossignol, S., Contralateral reflex reversal controlled by limb position in the acute spinal cat injected with clonidine, Brain Research, 144 (1978) 411-414. 20 Grillner, S. and Zangger, P., How detailed is the central pattern generation for locomotion? Brah~ Research, 88 (1975) 367 371. 21 Jankowska, E., Jukes, M. G., Lund, S. and Lundberg, A., The effect of DOPA on the spinal cord. 5. Reciprocal organization of pathways transmitting excitatory action to alpha motoneurones of flexors and extensors, Acta ph.vsiol, stand., 70, 369-388. 22 Lundberg, A., Refex control of stepping. In The Nansen MemorialLecture V, Universitetsforlaget, Oslo, 1969, p. 1-42. 23 Magnus, R., Zur Regelung des Bewegungen durch das Zentralnervensystem. Mitteilung II, Pfliigers Arch. ges. Physiol., 130 (1909) 253-269. 24 Magnus, R., In Korperstellung, Springer, Berlin, pp. 24-49. 25 Pearson, K. G. and Duysens, J., Function of segmental reflexes in the control of stepping in cockroaches and cats. In R. M. Herman, S. Grillner, P. Stein and D. Stuart (Eds.), Neural Control o[ Locomotion, Plenum Press, New York, 1976, pp. 519-537. 26 Pearson, K. G., Fourtner, C. R. and Wong, R. K., Nervous control of walking in the cockroach. In R. B. Stein, K. G. Pearson, R. S. Smith and J. B. Redford (Eds.), Control of Posture andLocomotion, Plenum Press, New York, 1973, pp. 495-514. 27 Perret,C. and Calbelguen,J. M., Central and reflex participation in the timing oflocomotor activations o f a bifunctional muscle, the semi-tendinosus, in the cat, Brain Research, 106 (1976) 390-395. 28 Perret, C. and Calbelguen, J. M., Efferent organization of the hindlimb locomotor cycle in the cat. In Satellite Symposium on Neurophysiological Mechanisms of Locomotion, Paris, 1977. 29 Rossignol, S., Ipsi- and contralateral reflex adaptation to locomotor activity. In Satellite Symposium on Neurophysiological Mechanisms of Locomotion, Paris, 1977. 30 Sherrington, C. S., On the innervation of antagonistic muscles. Sixth note, Proc. Roy. Soc. B, 66 (1900) 66-67.
73 31 Sherrington, C. S., Flexion reflex of the limb, crossed extension reflex and reflex stepping and standing, J. Physiol. (Lond.), 40 (1910) 28-121. 32 Viala, D., Etude Neurophysiologique et Pharmaeologique de FAction de Deux Syst~mes Monoaminergiques Centraux Descendants sur les Activit~s Locomotrices chez le Lapin, Th~se de Doctorat d'l~tat (Paris), 1975, pp. 1-201. 33 Viala, D., Donn6es sur l'activation du g6n6rateur spinal de la locomotion par les votes noradr6nergiques et par les aff6rences du r6flexe de flexion, J, Physiol. (Paris), 72 (1975) 31-32A. 34 Viala, D., Activation centrale de fibres alpha innervant des muscles bifonctionnels au cours de la locomotion fictive chez le lapin, J. Physiol. (Paris), 73 (1977) 60A. 35 Viala, D. and Buser, P., The effects of DOPA and 5-HTP on rhythmic efferent discharges in hindlimb nerves in the rabbit, Brain Research, 12 (1969) 437-443. 36 Viala, D., Valin, A. and Buser, P., Relationship between the 'late reflex discharge' and locomotor movements in acute spinal cats and rabbi ts treated with DOPA, Arch. ital. Biol., 112 (1974) 299-306. 37 Viala, D. and Viala, G.,Caract~reset concomitants de la programmationlocomotricechezlelapin, J. Physiol. (Paris), (1979) in press. 38 Viala, G., Variabilit6 de la configuration des d6charges eff6rentes d'origine supraspinate chez le lapin, J. Physiol. (Paris), 59 (1967) 307. 39 Viala, G. and Buser, P., D6charges eff6rentes rythmiques dans les pattes post6rieures chez le lapin et leur m~canisme, J. Physiol. (Paris), 57 (1965) 287-288. 40 Viala, G. andBuser~P.~Activit6s~c~m~tricesrythmiquesst6r6~typ6eschez~e~apins~usanesth6sie 16g~re, Exp. Brain Res., 8 (1969) 346-363. 41 Viala, G. and Viala, D., Elements of locomotor programming in the rabbit. In Satellite Symposium on Neurophysiological Mechanisms of Locomotion, Paris, 1977. 42 Vidal, C. and Viala, D., Programmation des activit6s locomotrices pour les diff6rents segments du membre post6rieur chez le lapin, J. Physiol. (Paris), 71 (1975) 348-349A.