- Email: [email protected]
Absence of impulse activity in cortical neurons with transient projections to the cerebellum
Developmental Brain Research, 50 (1989) 241-249 Elsevier
241
BRESD 50987
Absence of impulse activity in cortical neurons with transient projections to the cerebellum D.L. Tolbert Francis and Doris Murphy, Neuroanatomy Research Laboratory, Department of Anatomy and Neurobiology and Surgery (Neurosurgery), St. Louis University, St. Louis, MO 63104 (U.S.A.) (Accepted 6 June 1989)
Key words: Central nervous system development; Transient projection; Impulse activity; Cerebellum; Cerebral cortex; Collateral
During the second postnatal week of development in cats, neurons in layer V of the primary sensorimotor cortex project transiently, by way of collaterals of pyramidal tract axons, to the cerebellum. All cerebrocerebellar collaterals are subsequently eliminated, while the collaterals in the pyramidal tract persist into the adult. To determine if the transience of the projection to the cerebellum could be due to the lack of functional activity in cerebrocerebellar projection neurons, single-unit extracellular recordings were made from neurons in the primary somatosensory cortex (S-I) in 8-14-day-old kittens. Projection neurons were identified by their antidromic activation from pyramidal tract or cerebellum. Collision experiments confirmed that some neurons had collateral projections to both structures. Recordings from both generally anesthetized as well as locally anesthetized, but awake preparations, indicated that pyramidal tract and cerebrocerebellar projection neurons never fired action potentials spontaneously or were orthodromically activated following stimulation of the medial lemniscus. Stimulation of the medial lemniscus did synaptically activate neurons in the cortex, but these were always located superficial to the antidromically activated projection neurons. These findings indicate that pyramidal tract and/or cerebrocerebellar S-I projection neurons are physiologically silent during the period of development that cortical axons are transiently present in the cerebellum, suggesting that cerebrocerebellar projections may be eliminated because of the lack of impulse activity.
pulse activity on incipient synapses is indicated primarily by findings that n o r m a l l y transient connections will persist if afferents with which they overlap in the target are e x p e r i m e n t a l l y r e m o v e d 14'29, pharmacologically silenced 2'x4'25'26'34, or if they have
not being spontaneously active, responsive to orthodromic activation, or capable of initiating action potentials; action potentials initiated at the cell b o d y m a y not be p r o p a g a t e d by the axon or some collateral branches; transient axons m a y not form functional synapses 22 o r contact phenotypically app r o p r i a t e target cells prior to their elimination 23. A c o m m o n event during the n o r m a l ontogenesis of neocortical efferent pathways a p p e a r s to be the d e v e l o p m e n t of transient c o r t i c a l - c o r t i c a l 1°,18-21 and c o r t i c a l - s u b c o r t i c a l p r o j e c t i o n s 1,11,27'32,37,41,42. In
altered functional activity5,6. While p a t t e r n e d impulse activity m a y m e d i a t e the stabilization or elimination of some connections, the transience of o t h e r p r o j e c t i o n s m a y be due to the lack of functional activity as the result of: the p r o j e c t i o n neurons
neonatal cats, collaterals of p y r a m i d a l tract axons project transiently to the cerebellar cortex and d e e p nuclei 3°'43. C e r e b r o c e r e b e l l a r p r o j e c t i o n s arise from layer V p y r a m i d a l n e u r o n s located p r e d o m i n a t e l y in p r i m a r y s e n s o r i m o t o r areas of the f r o n t o p a r i e t a l
INTRODUCTION T h e stabilization or regression of developing i n t e r n e u r o n a l connections m a y be d e p e n d e n t upon the p a t t e r n of impulse activity in axons converging on t a r g e t neurons 9'12'13A6'35'44. The effects of im-
Correspondence: D.L. Tolbert, Department of Anatomy & Neurobiology, St. Louis University School of Medicine, 1402 S. Grand Blvd., St. Louis, MO 63104, U.S.A. 0165-3806/89/$03.50 (~ 1989 Elsevier Science Publishers B.V. (Biomedical Division)
242 cortex 42. With continued maturation of the brain, all cerebrocerebellar collaterals are eliminated while the collaterals in the pyramidal tract persist into the adult 43. This study is the first of a series of experiments to determine if the elimination of the cerebrocerebellar pathway could be due to the lack of functional activity for any of the reasons noted above. Single unit extracellular recordings were obtained from pyramidal tract and cerebrocerebellar projection neurons in the primary somatosensory cortex (S-I). In both general and locally anesthetized but awake preparations, S-I projection neurons were never spontaneously active nor were they orthodromically activated following electrical stimulation of the medial lemniscus or mechanical stimulation of the body surface. These findings indicate that during the developmental period that cortical axon collaterals are transiently present in the cerebellum, their parent cells are functionally silent. The transience o f cerebrocerebellar axons may therefore be due to the absence of impulse activity in this pathway. MATERIAL AND METHODS Recordings were made from cats ranging in age from 8 to 14 postnatal days (PND). All kittens were obtained from a breeding colony maintained in our vivarium. Kittens found in the morning, which were not present the previous day, were considered 1 PND. Kittens whose weights were within normal ranges, were active in their cages, and were considered healthy when used in the experiments. Most experiments were carried out on kittens anesthetized with a-chloralose (50 mg/kg, i.p.). Catheters placed in the left femoral artery and vein were used for monitoring arterial blood pressure, heart rate, and blood gases, as well as for injecting drugs. All animals were intubated, mounted in a Kopf sterotaxic apparatus, artificially ventilated with 100% oxygen, and paralyzed with pancuronium chloride (Pavulon 0.08 mg/kg). Body temperature was monitored with a rectal thermoprobe and maintained at 36.5-38.5 °C. Craniotomies were performed on the left side. A pair of monopolar stainless steel stimulating electrodes, oriented anterior-posteriorly and separated by 1-2 mm, were inserted posteriorly into the medulla on the left side at the level of the obex.
The orientation of this electrode rake and its positioning in the brainstem was such that one monopolar electrode was located in the predecussating fibers of the pyramidal tract (PT) whereas the other electrode was located in the area of the medial lemniscus (ML). The polarity of stimulation was reversed in an attempt to switch from anti- to orthodromic activation of S-l projection neurons, respectively. A pair of horizontally oriented stimulating electrodes were positioned into the left cerebellar nuclei and were used to activate cerebrocerebellar axons. The left frontal and rostral parietal cortex was exposed for recording. All exposed areas of the brain were covered with a 4% agar-saline mixture. Stimulus pulses 0.1-0.2 ms in duration were used to activate cortical axons. Extracellular recordings from S-I were made with glass microelectrodes filled with 2 M sodium chloride. Responses evoked following stimulation to the cerebellum (CB) and the brainstem were displayed and also recorded onto VCR tapes for offiine analysis, At the conclusion of the experiments the animals were perfused with 10% buffered formalin and the location of the stimulating electrodes was histologically verified. To determine whether the lack of impulse activity in cerebrocerebellar projection neurons (see Results) was due to the anesthetic or a genuine developmental phenomena, some experiments were carried out on locally anesthetized, but awake preparations (n = 4). The kittens were anesthetized with 1-2% isoflurane or halothane for all surgical procedures. All other aspects of the preparation of these animals for recording were identical to these described above. Just prior to recording the general anesthesia was discontinued and the edges of all skin incisions and pressure points were infiltrated with 2% lidocaine hydrochloride. To ensure that the animals were not perceiving any pain, a noxious stimulus (gentle pinching with a hemostat) was applied periodically to an area of the skin (midback) away from the locally anesthetized sites. This stimulus caused an immediate increase in heart rate and blood pressure which, after removing the stimulus, returned to physiological levels. These changes in heart rate and blood pressure in response to a noxious stimulus indicated that the 2% lidocaine prevented any perception of pain from the locally anesthetized skin openings or pressure points.
243 RESULTS Antidromic and orthodromic evoked unitary responses were recorded following stimulation of the brainstem and the CB. Unitary responses recorded after stimulation of the pyramidal tract (PT) or CB were considered to be antidromically evoked if they met at least 3 of 4 criteria: (1) the latency to the onset of the action potential did not vary more than 0.3 ms when the stimulus intensity was increased from threshold to suprathreshold levels; (2) when the frequency of suprathreshold stimulation was increased from 1 to 10 Hz; (3) the unit was capable of following suprathreshold stimulation greater than 30 H z (Fig. 1)3'17'28'38-40; and (4) the critical interstimulus interval was identical (within 0.3 ms) in collision experiments. Collision of an orthodromically propagated spike with a stimulus-evoked action potential being propagated antidromically, another criterion normally used to confirm antidromic activation could not be applied (see Discussion). Responses that were recorded when the microelectrode was near a cell body were distinguished from those made from axons by the presence, in the former, of an initial segment s o m a - d e n d r i t e (IS-SD) break on the rising phase of the action potential (Fig. 1B) 5'6. Thirty-one units were activated antidromically from the pyramidal tract. The latencies of activation varied between 11.6 and 56 ms with most responses occurring at 36-56 ms (Fig. 2). A distinction be-
tween fast- and slow-conducting PT neurons was inferred by a few neurons which were activated at 12-16 ms and the relatively greater n u m b e r of units activated between 36 and 44 ms 28. Seventeen units were activated antidromically from the cerebellum. The latency of cerebellar antidromic activation was between 8 and 36 ms with a peak latency at 16 ms. Previous fluorescent dye double-labeling experiments indicated that many of the cortical projections to the cerebellum were collaterals of corticobulbar and corticospinal axons 3°'43. Collision experiments to demonstrate collateral branching electrophysiologically 39, were performed on units which appeared to be antidromically activated from both the PT and CB (Fig. 3A,B). Suprathreshold conditioning and test stimuli were applied separately to the PT or CB and the interval between the stimuli was shortened until the response evoked by the test stimulus was consistently suppressed (Fig. 3D,G). This was identified as the critical interstimulus interval (CIS) for that order of stimulation. The order of stimulation was then reversed and again the
LATENCY OF ANTIDROMIC ACTIVATIONFROM TIE PYRAMI)AL TRACT B--
7--
V
6-5-4-3-2 I
Antidromic Responses
4
B
12
16 20
24 2B 52
:56 40 44 48
52 56 60 (MSI
LATENCY OF ANTIDROMIC ACTIVATION FROM THE CEREBELLUM
kL. :.i llmli
PT 1H=
•
P T 6OH=
Fig. 1. Antidromic responses of a cortical neuron to stimulation of the ipsilateral pyramidal tract (PT). A: the antidromic response to threshold stimulation of the PT (large arrow). The small arrow indicates the monophasic response recorded from a thalamocortical axon. B: the responses to suprathreshold stimulation of the PT at a frequency of 50 Hz. Notice the initial segment-soma-dendritic inflection (arrow) on the rising phase of the action potential identifying the recording as being obtained from the cell body. In B it was not possible to show the complete trace because of superimposition of the stimulus artifact. The latency to antidromic activation was not different from that in A.
7 6 5 4-5-2-I -I B
I 12
I i L i~ i . i I ~ 1 i I 16 20 24 2B 3:P SB 40 44 48
I I I I 52 56 60 (MS)
* LATENCY OF UNITS ACTIVATED FROM BOTH SITES
Fig. 2. Latency histograms of antidromic responses from the pyramidal tract (top) and cerebellum (bottom). The "Jr' identifies the latency of antidromic activation for neurons with collateral projection to both structures.
244 interstimulus interval was decreased to the critical interval. If the CIS intervals were identical (within 0.3 ms) and greater than the refractory p e r i o d of the neuron, as d e t e r m i n e d by paired stimulation of either the PT or CB (Fig. 3I,J), the neuron was considered to be antidromically activated via collateral p r o j e c t i o n s to both structures. Occasionally a n t i d r o m i c a l l y - e v o k e d spikes with similar amplitudes and waveform were r e c o r d e d following stimulation of both the PT and CB, but collision of the responses never occurred. This suggests that the recordings were from two different neurons, one being activated antidromically from the PT and the o t h e r from the CB. This would be consistent with anatomical findings that CB and PT projection
Antidromic Responses
:. PT
Cll
Collision
..._.
PT
CII
CiB
Refractory I
.
PT
Period J
..
AA
C8
C8
Fig. 3. The antidromic responses evoked following stimulation of the PT (A) and cerebellum (CB) (B). The collision experiments between the antidromically evoked response is shown in C-H. In C a suprathreshold test stimulus to the CB was preceded by a suprathreshold conditioning stimulus to the PT. As the interval between the stimuli was decreased, the critical interstimulus interval (CIS) was reached (D) at which the antidromic response from the CB was consistently suppressed. At the CIS when the conditioning stimulus to the PT was reduced to infrathreshold levels, the antidromic response evoked by the CB stimulus could again be observed (E). In F - H the order of stimulation was reversed and similar observations were made. Note that the CIS for both sequences of stimulation (D,G) were identical. In I and J the refractory period of the neuron was determined by applying two successive suprathreshold stimuli to the CB. Note that the interstimuli interval in J is equal to the refractory period of the neuron and is appreciably shorter than the CIS in D and G.
Orthodromic
B
Responses
I..
_5.:1 5 rnS+
ML 5 H z
Fig. 4. Synaptic activation of a cortical neuron following stimulation of the medial lemnicus (ML) at 1 (A) and 5 Hz (B). Note the variable latency to the onset of the evoked action potentials indicating the synaptic nature of the activation.
neurons are closely a d j a c e n t to each o t h e r in lamina V of the s e n s o r i m o t o r cortex 43. A n o t h e r possibility is that action potentials e v o k e d by the conditioning stimulus and p r o p a g a t e d in one collateral failed to invade the o t h e r collateral activated by the t e s t stimulus. Six neurons were identified in collision experiments as having collateral p r o j e c t i o n s to both the cerebellum and the spinal cord, B a s e d upon the latency of antidromic activation for these 6 units, it a p p e a r e d that most cells with collateral projections to both structures have relative i n t e r m e d i a t e conduction velocities (Fig. 2). Unitary responses, different from those e v o k e d antidromically, were also r e c o r d e d following stimulation of the caudal medulla or cerebellum. These spikes were considered to be o r t h o d r o m i c a l l y e v o k e d because: (1) there was considerable variability in the latency to the onset of the spike when the stimulus intensity was increased from threshold to suprathreshold levels and when the frequency of stimulation was varied from 1 to 10 H z (Figs. 4, 5); and (2) these units did not follow a suprathreshold stimulus at a frequency of stimulation g r e a t e r than 10 Hz. Two types of synaptically activated units were r e c o r d e d following stimulation of the caudal medulla. First, unitary m o n o p h a s i c responses were e v o k e d at a latency of 5 - 2 0 ms (Fig. 1A). G e n e r a l l y these spikes were followed at 1-3 ms by a negative field potential that was also e v o k e d by t h e b r a i n s t e m stimulus (Figs. 1A, 4). These units were identified as
245
5 Hz
I Hz ~m o
o
° 0
1~3
15
2O M-ec
° 0
.,
I
,,
,
15
20
,
,
i
25
30
36
Msec
7.5 Hz
2 . 5 Hz ,-
°--
~m
ar4~
o
o
it)
c0
•
i
~/,sec
i
i
Msec
Fig. 5. Poststimulus time histograms for a cortical neuron activated synaptically following stimulation to the medial lemnicus. Note the differences in the number of responses recorded when the frequency of stimulation was increased from 1 to 7.5 Hz and the
increase in the latency of activation. This unit could not follow greater than 7.5 Hz rate of stimulation.
thalamocortical axons. The second type of synaptically activated unit was usually a biphasic response. These responses were superimposed on the negative field potential (Fig. 4) with an I S - S D break apparent on the rising phase of the action potential. Increasing or decreasing the stimulus intensity or frequency of stimulation resulted not only in changes in the latency of the evoked unitary responses but also changes in the amplitude of the field potential as well (Figs. 4B, 5). The field potential was considered to be a result of synaptic potentials of thalamocortical afferents whereas the unitary responses were recorded from cortical neurons activated by the thalamic afferents. Stimulating the CB occasionally evoked similar field potentials and unitary responses in S-I, but the amplitude of the field was always less than that evoked by the brainstem stimulus as was the number of unitary responses. Orthodromically activated neurons were always located in the cortex superficial to units
antidromically activated from the PT or CB. Once PT and/or CB projection neurons were antidromically identified, the recordings were routinely continued for variable periods of time (up to 15 min) to determine if these units fired spontaneously or were synaptically activated following electrical stimulation of the medial lemniscus or mechanical stimulation of the contralateral body surface. The PT and/or CB units were periodically activated antidromically to verify that recordings could still be obtained from the projection neurons. During these recording periods PT and/or CB S-I projection neurons were functionally silent, i.e. they were not spontaneously active nor were they activated orthodromically following stimulation of the medial lemniscus. Only one unit fired very intermittently in response to brushing the skin on the contralateral forelimb. Since the objective of these studies was to determine if S-I neurons with transient cerebroce-
246
rebellar projections fire action potentials, the absence of spontaneous or evoked activity in the projection neurons in the above experiments could be a genuine developmental phenomenon, or alternatively, due to the effects of the anesthetic on cortical neuronal excitability. To resolve this, recordings were carried out in a limited number of kittens (n = 4) which were awake but locally anesthetized at all pressure points and incision sites. In these experiments, recordings were made from 2 PT and 2 CB projection neurons; these units were neither spontaneously active nor orthodromically activated following brainstem stimulation. DISCUSSION Orthograde and retrograde intra-axonal labeling studies have previously shown that as part of the normal ontogenesis of corticofugal pathways, direct projections to the cerebellum develop but then are subsequently eliminated 1"n'42. This report presents electrophysiological evidence for direct cerebrocerebellar projections in 2-week-old cats. Unitary responses were recorded from S-I neurons following electrical stimulation of the brainstem and cerebellum. The presence of an I S - S D break on the rising phase of the evoked action potential indicated that the recordings were obtained from cell bodies and not from axons 7"8. The criteria used to distinguish between antidromically versus orthodromically evoked responses were similar to those used in recording from adult brain 4. Constant latency in the onset of the evoked response when the stimulus intensity was increased from threshold to suprathreshold level and the unit's ability to follow relatively high frequency stimulation (>30 Hz) were criteria which were primarily used to distinguish antidromically-activated units from orthodromicallyevoked responses. However, recordings from a few neurons showed latency shifts of 0.4-2.0 ms when the stimulus intensity was increased from threshold to suprathreshold levels (Fig. 3A,C and B,F). Nevertheless, these responses followed high frequency stimulation without any latency dispersion. One unit was activated from both the CB and p T and the critical interstimulus interval was identical in the collision experiments (Fig. 3D,G). Because these unitary responses meet the latter criteria, they were
considered to be antidromically ew)ked. Previous studies have reported latency dispersions of 2 ~I--:L3 ms in recordings from immature cortical projection neurons following stimulation of the pyramidal tract 3. Intraceilular recordings confirmed, however. that these responses were evoked antidromically. These variations in antidromic latencies could be the result of immature membrane properties of the unmyelinated or poorly myelinated cortical axons ~. In the present study the latencies of antidromic activation of most responses did not vary more than 0.3 ms when stimulation intensity was increased from threshold to suprathreshold levels and therefore the classification of these units as PT and/or CB projection neurons is valid. Since impulse activity has been shown to be associated with the stabilization of incipient connections (see below), the principal objective for these experiments was to determine if the elimination of cerebrocerebellar projections could be due to the absence of impulse activity in the S-I projection neurons. Once neurons were identified electrophysiologically as projecting to the CB and/or PT. the recordings were continued to determine if they fired action potentials spontaneously or as a result of activation of medial lemniscal-thalamocortical projections (either directly by electrically stimulating the medial lemniscus or by cutaneous mechanical stimulation). In generally anesthetized animals, CB and/or PT projection neurons were silent during the period of the recordings. This absence of activity was shown to be a genuine developmental phenomenon since recordings from S-I projection neurons in awake animals revealed a similar absence of spontaneous or orthodromically evoked unitary activity in CB and/or PT projection neurons. A number of factors could account for the lack of spontaneous and evoked activity in S-I PT and/or CB projection neurons. First, the absence of impulse activity may be due to intrinsic morphophysiologicai properties of the projection neurons. Structurally, projection neurons and particularly their dendrites may not have developed to the point to support afferent innervation. Intracellular H R P staining of PT projection neurons in neonatal cats have shown that the apical dendrites of PT neurons extend to the pial surface in newborns and that the basal dendrites are nearly completely developed at the end of the
247 first postnatal week 28. With expansion of the thickness of the cortex, the dendrites continue to elongate and more spines develop on the dendrites. The morphology of PT projection neurons nevertheless appears to be relatively adult-like during the developmental period that cortical axons are present in the cerebellum. Alternatively, the lack of impulse activity in projection neurons may be due to intrinsic biophysical properties of the neurons. Intracellular injections of depolarizing current evokes action potentials in immature cortical projection neurons, indicating that the spike-generating mechanism of the cells is functional 3'24. Similarly, the application of exogenous glutamate evokes trains of action potentials in projection neurons suggesting that receptors have developed sufficiently for synaptic activation of the cell to occur24. Finally, the resting membrane potential of projection cells in neonates is not significantly different from that in mature animals, and therefore, it is unlikely that the lack of excitability is due to hyperpolarization of the cells. There are, however, changes in the amplitude and waveform of the action potentials evoked in projection neurons during maturation suggesting that there are ongoing changes in the membrane properties of the cells. For example, membrane resistance of the immature projection neuron is greater than in mature cells, but this does not necessarily mean that the cells are less responsive to synaptic input. Quite the contrary, increased input resistance could provide a compensatory mechanism to insure cell activation from poorly developed synaptic inputs TM. Collectively, the above findings suggest that the lack of impulse activity in projection neurons must be due to factors extrinsic to the projection neurons; specifically, the lack of afferent innervation. Thalamocortical axons synapse directly on and are capable of evoking spike activity in adult S-I projection neurons 16'45. In newborn cats peripheral cutaneous stimulation evokes multiunit activity in S-I neurons confirming that the medial lemniscalthalamocortical pathway is functional at the time of birth 31. However, findings reported in this paper indicate that this evoked activity is restricted to neurons located in layers of the cortex superficial to the layer V projection neurons. This suggests that thalamocortical input in 2-week-old cats has devel-
oped only to the point that impulse activity can be evoked in cortical interneurons but not in projection neurons. Taken together, these observations would infer that S-I projection neurons may not be responsive to the peripheral stimulation because either thalamocortical axons or cortical interneurons have not developed functional synaptic contacts with the projection neurons. The influence of impulse activity on developing connections has been most convincingly shown in the visual system 5'6'12'34'35. When tetrodotoxin, which selectively blocks impulse activity, is injected into one eye, normally transient retinal projections from the non-injected eye to areas of the ipsilateral superior colliculus persist in the adult 14. Similarly, in the visual cortex, geniculocortical projections relaying input from both eyes initially converge onto the same cortical neurons, but later this input becomes segregated leading to the formation of ocular dominance columns. Blocking activity from the eyes prevents the development of the ocular dominance columns 34. It is important to point out, though, that the above findings are primarily examples of the effects of impulse activity on the stabilization or elimination of the terminal arborization and synapses of projections which initially overlap in the target. In this situation impulse activity serves to refine diffuse connections into topographic maps. The relationship between impulse activity and the elimination of long collateral pathways, such as the transient cerebrocerebellar projection, has not been investigated until this present study. The findings that S-I cerebrocerebellar projection neurons do not fire action potentials, either spontaneously or as the result of orthodromic activation, suggests that the collateral projection to the cerebellum may be eliminated because of the lack of impulse activity in this pathway. From this one could conclude that during development there is a critical period which, after a collateral forms its terminal arborization, it must be functionally activated or be eliminated. If projections to the cerebellum are eliminated because of the lack of impulse activity, and since many cerebrocerebellar projections are collaterals of pyramidal tract axons, why then do the projections in the PT persist into the adult? This may be due to differences in the timing of development of the arborizations of dif-
248 ferent p y r a m i d a l tract collaterals. For example, at the time c e r e b r o c e r e b e l l a r collaterals have formed mossy fiber-type terminals in the cerebellar cortex, but the S-I p r o j e c t i o n neurons are silent; collateral corticospinal projections are just beginning to form their terminal arbors in the spinal gray m a t t e r 36. The critical p e r i o d for activation of the corticospinal collaterals accordingly would then occur later in d e v e l o p m e n t than the critical p e r i o d for activation of the c e r e b r o c e r e b e l l a r collaterals. In cats, corticospinal p r o j e c t i o n s from the p r i m a r y m o t o r cortex do not a p p e a r adult-like in their distribution until about the seventh p o s t n a t a l week of d e v e l o p m e n t 37. Resting unitary activity in PT neurons in the primary m o t o r cortex gradually increases from a low rate (2/s) in the second postnatal w e e k to adult levels by the ninth postnatal week 3. A caveat to the hypothesis that the elimination of c e r e b r o c e r e b e l l a r projections is due to the absence of impulse activity in S-I p r o j e c t i o n neurons is that c e r e b r o c e r e b e l l a r projections will persist in adult cats if o t h e r afferent projections to cerebellar targets are eliminated 29. O n e possibility for the persistence
of c e r e b r o c e r e b e l l a r projections u n d e r these conditions is that cerebellar deafferentation causes a more rapid d e v e l o p m e n t of impulse activity in cortical projection neurons. A m o r e plausible explanation. however, is that the critical p e r i o d for the development of impulse activity in c e r e b r o c e r e b e l l a r axons has been e x t e n d e d as the result of the lesion-induced changes in the e n v i r o n m e n t of the partially deafferented cerebellum. A x o n - t a r g e t neuron interactions could occur in a less competitive milieu and could then sustain the transient collaterals until impulse activity develops. This would be consistent with the hypothesis that impulse activity and target interactions are necessary for the stabilization of incipient connections ~3"44.
REFERENCES
10 Dehay, C., Kennedy, H. and Bullier, J., Characterization of transient cortical projections from auditory somatosensory, and motor cortices to visual areas 17, 18, and 19 in the kitten, J. Comp. Neurol., 272 (1988) 68-89. 11 Distel, H. and Hollander, H., Autoradiographic tracing of developing subcortical projections of the occipital region in fetal rabbits, J. Comp. Neurol., 192 (1980) 505-518. 12 Dubin, M.W., Stark, L.A. and Archer, S.M., A role for action-potential activity in the development of neuronal connections in the kitten retinogeniculate pathway, J. Neurosci., 6 (1986) 1021-1036. 13 Fawcett, J.W. and O'Leary, D.D.M., The role of electrical activity in the formation of topographic maps in the nervous system, Trends Neurosci., 8 (1985) 201-206. 14 Fawcett, J.W., O'Leary, D.D.M. and Cowan, W.M., Activity and the control of ganglion cell death in the rat retina, Proc. Natl. Acad. Sci. U.S.A., 81 (1984) 5589-5593. 15 Hebb, D.O., Organization and Behavior, Wiley, New York, 1949. 16 Heilweg, E-C., Schuitz, W. and Creutzfeldt, O.D., Extracellular and intracellular recordings from cat's cortical whisker projection area: thalamus-cortical response transformation, J. Neurophysiol., 40 (1977) 463-479. 17 Huttenlocher, ER., Development of cortical neuronal activity in the neonatal cat, Exp. Neurol., 17 (1967) 247-262. 18 Innocenti, G.M., Growth and reshaping of axons in the establishment of visual callosal connections, Science, 212 (1981) 824-827. 19 Innocenti, G.M. and Caminiti, R., Postnatal shaping of callosal connections from sensory areas, Exp. Brain Res.,
1 Adams, C.E., Mihailoff, G.A. and Woodward, D.J., A transient component of the developing corticospinal tract arises in visual cortex, Neurosci. Lea., 36 (1983) 243-248. 2 Archer, S.M., Dubin, M.W. and Stark, L.A., Abnormal development of kitten retino-geniculate connectivity in the absence of action potentials, Science, 217 (1982) 743-745. 3 Amassian, V.E. and Ross, R.J., Electrophysiological correlates of the developing higher sensorimotor control system, J. Physiol. (Paris), 74 (1978) 185-201. 4 Bantli, H., Bloedel, J.R. and Tolbert, D., Activation of neurons in the cerebellar nuclei and ascending reticular formation by stimulation of the cerebellar surface, J. Neurosurg., 45 (1976) 539-554. 5 Berman, N.E. and Payne, B.R., Alterations in connections of the corpus callosum following convergent and divergent strabismus, Brain Res., 274 (1983) 201-212. 6 Berman, N.E.J. and Payne, B.R., An exuberant retinocollicular pathway in Siamese kittens: effects of competition and abnormal activity on its maturation, Dev. Brain Res., 22 (1985) 197-209. 7 Bishop, P.O., Burke, W. and Davis, R., Single unit recording from antidromically activated optic radiation neurones, J. Physiol. (Lond.), 162 (1962) 432-450. 8 Bishop, P.O., Burke, W. and Davis, R., The identification of single units in central visual pathways, J. Physiol. (Lond.), 162 (1962) 409-431. 9 Changeux, J.P. and Danchin, A., Selective stabilisation of developing synapses as a mechanism for the specification of neuronal networks, Nature (Lond.), 264 (1976) 705-711.
ACKNOWLEDGEMENTS The authors gratefully a c k n o w l e d g e the d e d i c a t e d technical assistance of Ms. Terri Harris and the assistance of Mrs. W e n d y Anzilotti in the p r e p a r a tion of this manuscript. S u p p o r t e d by N I H G r a n t NS20227.
249 38 (1980) 381-394. 20 Innocenti, G.M. and Clarke, S., Bilateral transitory projection to visual areas from auditory cortex in kittens, Dev. Brain Res., 14 91984) 143-148. 21 Ivy, G.O. and Killackey, H.P., Ontogenetic changes in the projections of neocortical neurons, J. Neurosci., 2 (1982) 735-743. 22 Jeffery, G., Arzymonow, B.J. and Lieberman, A.R., Does the early exuberant retinal projection to the superior colliculus in the neonatal rat develop synaptic connections? Dev. Brain Res., 14 (1984) 135-138. 23 Mason, C.A. and Gregory, E., Postnatal maturation of cerebeilar mossy and climbing fibers: transient expression of dual features on single axons, J. Neurosci., 4 (1984) 1715-1735. 24 McCormick, D.A. and Prince, D.A., Post-natal development of electrophysiological properties of rat cerebral cortical pyramidal neurones, J. PhysioL (Lond.), 393 (1987) 743-762. 25 Meyer, R.L., Tetrodotoxin inhibits the formation of refined retinotopography in goldfish, Dev. Brain Res., 6 (1983) 293-298. 26 O'Leary, D.D.M., Fawcett, J.W. and Cowan, W.M., Topographic targeting errors in the retinocollicular projection and their elimination by selective ganglion cell death, J. Neurosci., 6 (1986) 3692-3705. 27 O'Leary, D.D.M. and Stanfield, B.B., A transient pyramidal tract projection from the visual cortex in the hamster and its removal by selective collateral elimination, Dev. Brain Res., 27 (1986) 87-99. 28 Oka, H., Samejima, A. and Yamamoto, T., Post-natal development of pyramidal tract neurones in kittens, J. Physiol. (Lond.), 363 (1985) 481-499. 29 Panneton, W.M., The persistence of a normally transient cerebrocerebellar pathway in the cat, Dev. Brain Res., 30 (1986) 133-139. 30 Panneton, W.M. and Tolbert, D.L., The collateral origin of a transient cerebrocerebellar pathway in kittens. A study using fluorescent double-labeling techniques, Dev. Brain Res., 14 (1984) 247-254. 31 Rubel, E.W., A comparison of somatotopic organization in sensory neocortex of newborn kittens and adult cats, J. Comp. Neurol., 143 (1971) 447-480. 32 Stanfield, B,B. and O'Leary, D.D.M., The transient
33 34 35 36 37 38 39 40
41 42 43
44 45
corticospinal projection from the occipital cortex during the postnatal development of the rat, J. Comp. Neurol., 238 (1985) 236-248. Stent, G.S., A physiological mechanism for Hebb's postulate of learning, Proc. Natl. Acad. Sci. U.S.A., 70 (1973) 997-1001. Stryker, M.P. and Harris, W.A., Binocular impulse blockade prevents the formation of ocular dominance columns in cat visual cortex, J. Neurosci., 6 (1986) 2117-2133. Swindale, N.V., Absence of ocular dominance patches in dark-reared cats, Nature (Lond.), 290 (1981) 332-333. Swink, T.D. and Tolbert, D,L., The development of corticospinal projections in neonatal cats, Soc. Neurosci. Abstr., 15 (1989) 69. Theriault, E. and Tatton, W.G., Postnatal redistribution of pericruciate motor cortical projections within the kitten spinal cord, Dev. Brain Res., 45 (1989) 219-237. Tolbert, D.L., Bantli, H. and Bloedel, J.R., Anatomical and physiological evidence for a cerebellar nucleo-cortical projection in the cat, Neuroscience, 1 (1976) 205-217. Tolbert, D.L., Bantli, H. and Bloedel, J.R., Multiple branching of cerebellar efferent projections in cats, Exp. Brain Res., 31 (1978) 305-316. Tolbert, D.L., Bantli, H., Hames, E.G., Ebner, T.J., McMullen, T.A. and Bloedel, J.R., A demonstration of the dentato-reticulospinal projection in the cat, Neuroscience, 5 (1980) 1479-1488. Tolbert, D.L., Dunn, R.C. and Vogler, G.A., The postnatal development of corticotrigeminal projections in the cat, J. Comp. Neurol., 228 (1984) 478-490. Tolbert, D.L. and Panneton, W.M., Transient cerebrocerebellar projections in kittens: postnatal development and topography, J. Cornp. Neurol., 221 (1983) 216-228. Tolbert, D.L. and Panneton, W.M., The transience of cerebrocerebellar projections is due to selective elimination of axon collaterals and not neuronal death, Dev. Brain Res., 16 (1984) 301-306. Udin, S.B. and Fawcett, J.W., Formation of topographic maps, Annu. Rev. Neurosci., 1l (1988) 289-327. White, E.L., Termination of thalamic afferents in the cerebral cortex. In E.G. Jones and A. Peters (Eds.), Sensory-Motor Areas and Aspects of Cortical Connectivity, Vol. 5, Plenum, New York, 1986, pp. 271-289.
241
BRESD 50987
Absence of impulse activity in cortical neurons with transient projections to the cerebellum D.L. Tolbert Francis and Doris Murphy, Neuroanatomy Research Laboratory, Department of Anatomy and Neurobiology and Surgery (Neurosurgery), St. Louis University, St. Louis, MO 63104 (U.S.A.) (Accepted 6 June 1989)
Key words: Central nervous system development; Transient projection; Impulse activity; Cerebellum; Cerebral cortex; Collateral
During the second postnatal week of development in cats, neurons in layer V of the primary sensorimotor cortex project transiently, by way of collaterals of pyramidal tract axons, to the cerebellum. All cerebrocerebellar collaterals are subsequently eliminated, while the collaterals in the pyramidal tract persist into the adult. To determine if the transience of the projection to the cerebellum could be due to the lack of functional activity in cerebrocerebellar projection neurons, single-unit extracellular recordings were made from neurons in the primary somatosensory cortex (S-I) in 8-14-day-old kittens. Projection neurons were identified by their antidromic activation from pyramidal tract or cerebellum. Collision experiments confirmed that some neurons had collateral projections to both structures. Recordings from both generally anesthetized as well as locally anesthetized, but awake preparations, indicated that pyramidal tract and cerebrocerebellar projection neurons never fired action potentials spontaneously or were orthodromically activated following stimulation of the medial lemniscus. Stimulation of the medial lemniscus did synaptically activate neurons in the cortex, but these were always located superficial to the antidromically activated projection neurons. These findings indicate that pyramidal tract and/or cerebrocerebellar S-I projection neurons are physiologically silent during the period of development that cortical axons are transiently present in the cerebellum, suggesting that cerebrocerebellar projections may be eliminated because of the lack of impulse activity.
pulse activity on incipient synapses is indicated primarily by findings that n o r m a l l y transient connections will persist if afferents with which they overlap in the target are e x p e r i m e n t a l l y r e m o v e d 14'29, pharmacologically silenced 2'x4'25'26'34, or if they have
not being spontaneously active, responsive to orthodromic activation, or capable of initiating action potentials; action potentials initiated at the cell b o d y m a y not be p r o p a g a t e d by the axon or some collateral branches; transient axons m a y not form functional synapses 22 o r contact phenotypically app r o p r i a t e target cells prior to their elimination 23. A c o m m o n event during the n o r m a l ontogenesis of neocortical efferent pathways a p p e a r s to be the d e v e l o p m e n t of transient c o r t i c a l - c o r t i c a l 1°,18-21 and c o r t i c a l - s u b c o r t i c a l p r o j e c t i o n s 1,11,27'32,37,41,42. In
altered functional activity5,6. While p a t t e r n e d impulse activity m a y m e d i a t e the stabilization or elimination of some connections, the transience of o t h e r p r o j e c t i o n s m a y be due to the lack of functional activity as the result of: the p r o j e c t i o n neurons
neonatal cats, collaterals of p y r a m i d a l tract axons project transiently to the cerebellar cortex and d e e p nuclei 3°'43. C e r e b r o c e r e b e l l a r p r o j e c t i o n s arise from layer V p y r a m i d a l n e u r o n s located p r e d o m i n a t e l y in p r i m a r y s e n s o r i m o t o r areas of the f r o n t o p a r i e t a l
INTRODUCTION T h e stabilization or regression of developing i n t e r n e u r o n a l connections m a y be d e p e n d e n t upon the p a t t e r n of impulse activity in axons converging on t a r g e t neurons 9'12'13A6'35'44. The effects of im-
Correspondence: D.L. Tolbert, Department of Anatomy & Neurobiology, St. Louis University School of Medicine, 1402 S. Grand Blvd., St. Louis, MO 63104, U.S.A. 0165-3806/89/$03.50 (~ 1989 Elsevier Science Publishers B.V. (Biomedical Division)
242 cortex 42. With continued maturation of the brain, all cerebrocerebellar collaterals are eliminated while the collaterals in the pyramidal tract persist into the adult 43. This study is the first of a series of experiments to determine if the elimination of the cerebrocerebellar pathway could be due to the lack of functional activity for any of the reasons noted above. Single unit extracellular recordings were obtained from pyramidal tract and cerebrocerebellar projection neurons in the primary somatosensory cortex (S-I). In both general and locally anesthetized but awake preparations, S-I projection neurons were never spontaneously active nor were they orthodromically activated following electrical stimulation of the medial lemniscus or mechanical stimulation of the body surface. These findings indicate that during the developmental period that cortical axon collaterals are transiently present in the cerebellum, their parent cells are functionally silent. The transience o f cerebrocerebellar axons may therefore be due to the absence of impulse activity in this pathway. MATERIAL AND METHODS Recordings were made from cats ranging in age from 8 to 14 postnatal days (PND). All kittens were obtained from a breeding colony maintained in our vivarium. Kittens found in the morning, which were not present the previous day, were considered 1 PND. Kittens whose weights were within normal ranges, were active in their cages, and were considered healthy when used in the experiments. Most experiments were carried out on kittens anesthetized with a-chloralose (50 mg/kg, i.p.). Catheters placed in the left femoral artery and vein were used for monitoring arterial blood pressure, heart rate, and blood gases, as well as for injecting drugs. All animals were intubated, mounted in a Kopf sterotaxic apparatus, artificially ventilated with 100% oxygen, and paralyzed with pancuronium chloride (Pavulon 0.08 mg/kg). Body temperature was monitored with a rectal thermoprobe and maintained at 36.5-38.5 °C. Craniotomies were performed on the left side. A pair of monopolar stainless steel stimulating electrodes, oriented anterior-posteriorly and separated by 1-2 mm, were inserted posteriorly into the medulla on the left side at the level of the obex.
The orientation of this electrode rake and its positioning in the brainstem was such that one monopolar electrode was located in the predecussating fibers of the pyramidal tract (PT) whereas the other electrode was located in the area of the medial lemniscus (ML). The polarity of stimulation was reversed in an attempt to switch from anti- to orthodromic activation of S-l projection neurons, respectively. A pair of horizontally oriented stimulating electrodes were positioned into the left cerebellar nuclei and were used to activate cerebrocerebellar axons. The left frontal and rostral parietal cortex was exposed for recording. All exposed areas of the brain were covered with a 4% agar-saline mixture. Stimulus pulses 0.1-0.2 ms in duration were used to activate cortical axons. Extracellular recordings from S-I were made with glass microelectrodes filled with 2 M sodium chloride. Responses evoked following stimulation to the cerebellum (CB) and the brainstem were displayed and also recorded onto VCR tapes for offiine analysis, At the conclusion of the experiments the animals were perfused with 10% buffered formalin and the location of the stimulating electrodes was histologically verified. To determine whether the lack of impulse activity in cerebrocerebellar projection neurons (see Results) was due to the anesthetic or a genuine developmental phenomena, some experiments were carried out on locally anesthetized, but awake preparations (n = 4). The kittens were anesthetized with 1-2% isoflurane or halothane for all surgical procedures. All other aspects of the preparation of these animals for recording were identical to these described above. Just prior to recording the general anesthesia was discontinued and the edges of all skin incisions and pressure points were infiltrated with 2% lidocaine hydrochloride. To ensure that the animals were not perceiving any pain, a noxious stimulus (gentle pinching with a hemostat) was applied periodically to an area of the skin (midback) away from the locally anesthetized sites. This stimulus caused an immediate increase in heart rate and blood pressure which, after removing the stimulus, returned to physiological levels. These changes in heart rate and blood pressure in response to a noxious stimulus indicated that the 2% lidocaine prevented any perception of pain from the locally anesthetized skin openings or pressure points.
243 RESULTS Antidromic and orthodromic evoked unitary responses were recorded following stimulation of the brainstem and the CB. Unitary responses recorded after stimulation of the pyramidal tract (PT) or CB were considered to be antidromically evoked if they met at least 3 of 4 criteria: (1) the latency to the onset of the action potential did not vary more than 0.3 ms when the stimulus intensity was increased from threshold to suprathreshold levels; (2) when the frequency of suprathreshold stimulation was increased from 1 to 10 Hz; (3) the unit was capable of following suprathreshold stimulation greater than 30 H z (Fig. 1)3'17'28'38-40; and (4) the critical interstimulus interval was identical (within 0.3 ms) in collision experiments. Collision of an orthodromically propagated spike with a stimulus-evoked action potential being propagated antidromically, another criterion normally used to confirm antidromic activation could not be applied (see Discussion). Responses that were recorded when the microelectrode was near a cell body were distinguished from those made from axons by the presence, in the former, of an initial segment s o m a - d e n d r i t e (IS-SD) break on the rising phase of the action potential (Fig. 1B) 5'6. Thirty-one units were activated antidromically from the pyramidal tract. The latencies of activation varied between 11.6 and 56 ms with most responses occurring at 36-56 ms (Fig. 2). A distinction be-
tween fast- and slow-conducting PT neurons was inferred by a few neurons which were activated at 12-16 ms and the relatively greater n u m b e r of units activated between 36 and 44 ms 28. Seventeen units were activated antidromically from the cerebellum. The latency of cerebellar antidromic activation was between 8 and 36 ms with a peak latency at 16 ms. Previous fluorescent dye double-labeling experiments indicated that many of the cortical projections to the cerebellum were collaterals of corticobulbar and corticospinal axons 3°'43. Collision experiments to demonstrate collateral branching electrophysiologically 39, were performed on units which appeared to be antidromically activated from both the PT and CB (Fig. 3A,B). Suprathreshold conditioning and test stimuli were applied separately to the PT or CB and the interval between the stimuli was shortened until the response evoked by the test stimulus was consistently suppressed (Fig. 3D,G). This was identified as the critical interstimulus interval (CIS) for that order of stimulation. The order of stimulation was then reversed and again the
LATENCY OF ANTIDROMIC ACTIVATIONFROM TIE PYRAMI)AL TRACT B--
7--
V
6-5-4-3-2 I
Antidromic Responses
4
B
12
16 20
24 2B 52
:56 40 44 48
52 56 60 (MSI
LATENCY OF ANTIDROMIC ACTIVATION FROM THE CEREBELLUM
kL. :.i llmli
PT 1H=
•
P T 6OH=
Fig. 1. Antidromic responses of a cortical neuron to stimulation of the ipsilateral pyramidal tract (PT). A: the antidromic response to threshold stimulation of the PT (large arrow). The small arrow indicates the monophasic response recorded from a thalamocortical axon. B: the responses to suprathreshold stimulation of the PT at a frequency of 50 Hz. Notice the initial segment-soma-dendritic inflection (arrow) on the rising phase of the action potential identifying the recording as being obtained from the cell body. In B it was not possible to show the complete trace because of superimposition of the stimulus artifact. The latency to antidromic activation was not different from that in A.
7 6 5 4-5-2-I -I B
I 12
I i L i~ i . i I ~ 1 i I 16 20 24 2B 3:P SB 40 44 48
I I I I 52 56 60 (MS)
* LATENCY OF UNITS ACTIVATED FROM BOTH SITES
Fig. 2. Latency histograms of antidromic responses from the pyramidal tract (top) and cerebellum (bottom). The "Jr' identifies the latency of antidromic activation for neurons with collateral projection to both structures.
244 interstimulus interval was decreased to the critical interval. If the CIS intervals were identical (within 0.3 ms) and greater than the refractory p e r i o d of the neuron, as d e t e r m i n e d by paired stimulation of either the PT or CB (Fig. 3I,J), the neuron was considered to be antidromically activated via collateral p r o j e c t i o n s to both structures. Occasionally a n t i d r o m i c a l l y - e v o k e d spikes with similar amplitudes and waveform were r e c o r d e d following stimulation of both the PT and CB, but collision of the responses never occurred. This suggests that the recordings were from two different neurons, one being activated antidromically from the PT and the o t h e r from the CB. This would be consistent with anatomical findings that CB and PT projection
Antidromic Responses
:. PT
Cll
Collision
..._.
PT
CII
CiB
Refractory I
.
PT
Period J
..
AA
C8
C8
Fig. 3. The antidromic responses evoked following stimulation of the PT (A) and cerebellum (CB) (B). The collision experiments between the antidromically evoked response is shown in C-H. In C a suprathreshold test stimulus to the CB was preceded by a suprathreshold conditioning stimulus to the PT. As the interval between the stimuli was decreased, the critical interstimulus interval (CIS) was reached (D) at which the antidromic response from the CB was consistently suppressed. At the CIS when the conditioning stimulus to the PT was reduced to infrathreshold levels, the antidromic response evoked by the CB stimulus could again be observed (E). In F - H the order of stimulation was reversed and similar observations were made. Note that the CIS for both sequences of stimulation (D,G) were identical. In I and J the refractory period of the neuron was determined by applying two successive suprathreshold stimuli to the CB. Note that the interstimuli interval in J is equal to the refractory period of the neuron and is appreciably shorter than the CIS in D and G.
Orthodromic
B
Responses
I..
_5.:1 5 rnS+
ML 5 H z
Fig. 4. Synaptic activation of a cortical neuron following stimulation of the medial lemnicus (ML) at 1 (A) and 5 Hz (B). Note the variable latency to the onset of the evoked action potentials indicating the synaptic nature of the activation.
neurons are closely a d j a c e n t to each o t h e r in lamina V of the s e n s o r i m o t o r cortex 43. A n o t h e r possibility is that action potentials e v o k e d by the conditioning stimulus and p r o p a g a t e d in one collateral failed to invade the o t h e r collateral activated by the t e s t stimulus. Six neurons were identified in collision experiments as having collateral p r o j e c t i o n s to both the cerebellum and the spinal cord, B a s e d upon the latency of antidromic activation for these 6 units, it a p p e a r e d that most cells with collateral projections to both structures have relative i n t e r m e d i a t e conduction velocities (Fig. 2). Unitary responses, different from those e v o k e d antidromically, were also r e c o r d e d following stimulation of the caudal medulla or cerebellum. These spikes were considered to be o r t h o d r o m i c a l l y e v o k e d because: (1) there was considerable variability in the latency to the onset of the spike when the stimulus intensity was increased from threshold to suprathreshold levels and when the frequency of stimulation was varied from 1 to 10 H z (Figs. 4, 5); and (2) these units did not follow a suprathreshold stimulus at a frequency of stimulation g r e a t e r than 10 Hz. Two types of synaptically activated units were r e c o r d e d following stimulation of the caudal medulla. First, unitary m o n o p h a s i c responses were e v o k e d at a latency of 5 - 2 0 ms (Fig. 1A). G e n e r a l l y these spikes were followed at 1-3 ms by a negative field potential that was also e v o k e d by t h e b r a i n s t e m stimulus (Figs. 1A, 4). These units were identified as
245
5 Hz
I Hz ~m o
o
° 0
1~3
15
2O M-ec
° 0
.,
I
,,
,
15
20
,
,
i
25
30
36
Msec
7.5 Hz
2 . 5 Hz ,-
°--
~m
ar4~
o
o
it)
c0
•
i
~/,sec
i
i
Msec
Fig. 5. Poststimulus time histograms for a cortical neuron activated synaptically following stimulation to the medial lemnicus. Note the differences in the number of responses recorded when the frequency of stimulation was increased from 1 to 7.5 Hz and the
increase in the latency of activation. This unit could not follow greater than 7.5 Hz rate of stimulation.
thalamocortical axons. The second type of synaptically activated unit was usually a biphasic response. These responses were superimposed on the negative field potential (Fig. 4) with an I S - S D break apparent on the rising phase of the action potential. Increasing or decreasing the stimulus intensity or frequency of stimulation resulted not only in changes in the latency of the evoked unitary responses but also changes in the amplitude of the field potential as well (Figs. 4B, 5). The field potential was considered to be a result of synaptic potentials of thalamocortical afferents whereas the unitary responses were recorded from cortical neurons activated by the thalamic afferents. Stimulating the CB occasionally evoked similar field potentials and unitary responses in S-I, but the amplitude of the field was always less than that evoked by the brainstem stimulus as was the number of unitary responses. Orthodromically activated neurons were always located in the cortex superficial to units
antidromically activated from the PT or CB. Once PT and/or CB projection neurons were antidromically identified, the recordings were routinely continued for variable periods of time (up to 15 min) to determine if these units fired spontaneously or were synaptically activated following electrical stimulation of the medial lemniscus or mechanical stimulation of the contralateral body surface. The PT and/or CB units were periodically activated antidromically to verify that recordings could still be obtained from the projection neurons. During these recording periods PT and/or CB S-I projection neurons were functionally silent, i.e. they were not spontaneously active nor were they activated orthodromically following stimulation of the medial lemniscus. Only one unit fired very intermittently in response to brushing the skin on the contralateral forelimb. Since the objective of these studies was to determine if S-I neurons with transient cerebroce-
246
rebellar projections fire action potentials, the absence of spontaneous or evoked activity in the projection neurons in the above experiments could be a genuine developmental phenomenon, or alternatively, due to the effects of the anesthetic on cortical neuronal excitability. To resolve this, recordings were carried out in a limited number of kittens (n = 4) which were awake but locally anesthetized at all pressure points and incision sites. In these experiments, recordings were made from 2 PT and 2 CB projection neurons; these units were neither spontaneously active nor orthodromically activated following brainstem stimulation. DISCUSSION Orthograde and retrograde intra-axonal labeling studies have previously shown that as part of the normal ontogenesis of corticofugal pathways, direct projections to the cerebellum develop but then are subsequently eliminated 1"n'42. This report presents electrophysiological evidence for direct cerebrocerebellar projections in 2-week-old cats. Unitary responses were recorded from S-I neurons following electrical stimulation of the brainstem and cerebellum. The presence of an I S - S D break on the rising phase of the evoked action potential indicated that the recordings were obtained from cell bodies and not from axons 7"8. The criteria used to distinguish between antidromically versus orthodromically evoked responses were similar to those used in recording from adult brain 4. Constant latency in the onset of the evoked response when the stimulus intensity was increased from threshold to suprathreshold level and the unit's ability to follow relatively high frequency stimulation (>30 Hz) were criteria which were primarily used to distinguish antidromically-activated units from orthodromicallyevoked responses. However, recordings from a few neurons showed latency shifts of 0.4-2.0 ms when the stimulus intensity was increased from threshold to suprathreshold levels (Fig. 3A,C and B,F). Nevertheless, these responses followed high frequency stimulation without any latency dispersion. One unit was activated from both the CB and p T and the critical interstimulus interval was identical in the collision experiments (Fig. 3D,G). Because these unitary responses meet the latter criteria, they were
considered to be antidromically ew)ked. Previous studies have reported latency dispersions of 2 ~I--:L3 ms in recordings from immature cortical projection neurons following stimulation of the pyramidal tract 3. Intraceilular recordings confirmed, however. that these responses were evoked antidromically. These variations in antidromic latencies could be the result of immature membrane properties of the unmyelinated or poorly myelinated cortical axons ~. In the present study the latencies of antidromic activation of most responses did not vary more than 0.3 ms when stimulation intensity was increased from threshold to suprathreshold levels and therefore the classification of these units as PT and/or CB projection neurons is valid. Since impulse activity has been shown to be associated with the stabilization of incipient connections (see below), the principal objective for these experiments was to determine if the elimination of cerebrocerebellar projections could be due to the absence of impulse activity in the S-I projection neurons. Once neurons were identified electrophysiologically as projecting to the CB and/or PT. the recordings were continued to determine if they fired action potentials spontaneously or as a result of activation of medial lemniscal-thalamocortical projections (either directly by electrically stimulating the medial lemniscus or by cutaneous mechanical stimulation). In generally anesthetized animals, CB and/or PT projection neurons were silent during the period of the recordings. This absence of activity was shown to be a genuine developmental phenomenon since recordings from S-I projection neurons in awake animals revealed a similar absence of spontaneous or orthodromically evoked unitary activity in CB and/or PT projection neurons. A number of factors could account for the lack of spontaneous and evoked activity in S-I PT and/or CB projection neurons. First, the absence of impulse activity may be due to intrinsic morphophysiologicai properties of the projection neurons. Structurally, projection neurons and particularly their dendrites may not have developed to the point to support afferent innervation. Intracellular H R P staining of PT projection neurons in neonatal cats have shown that the apical dendrites of PT neurons extend to the pial surface in newborns and that the basal dendrites are nearly completely developed at the end of the
247 first postnatal week 28. With expansion of the thickness of the cortex, the dendrites continue to elongate and more spines develop on the dendrites. The morphology of PT projection neurons nevertheless appears to be relatively adult-like during the developmental period that cortical axons are present in the cerebellum. Alternatively, the lack of impulse activity in projection neurons may be due to intrinsic biophysical properties of the neurons. Intracellular injections of depolarizing current evokes action potentials in immature cortical projection neurons, indicating that the spike-generating mechanism of the cells is functional 3'24. Similarly, the application of exogenous glutamate evokes trains of action potentials in projection neurons suggesting that receptors have developed sufficiently for synaptic activation of the cell to occur24. Finally, the resting membrane potential of projection cells in neonates is not significantly different from that in mature animals, and therefore, it is unlikely that the lack of excitability is due to hyperpolarization of the cells. There are, however, changes in the amplitude and waveform of the action potentials evoked in projection neurons during maturation suggesting that there are ongoing changes in the membrane properties of the cells. For example, membrane resistance of the immature projection neuron is greater than in mature cells, but this does not necessarily mean that the cells are less responsive to synaptic input. Quite the contrary, increased input resistance could provide a compensatory mechanism to insure cell activation from poorly developed synaptic inputs TM. Collectively, the above findings suggest that the lack of impulse activity in projection neurons must be due to factors extrinsic to the projection neurons; specifically, the lack of afferent innervation. Thalamocortical axons synapse directly on and are capable of evoking spike activity in adult S-I projection neurons 16'45. In newborn cats peripheral cutaneous stimulation evokes multiunit activity in S-I neurons confirming that the medial lemniscalthalamocortical pathway is functional at the time of birth 31. However, findings reported in this paper indicate that this evoked activity is restricted to neurons located in layers of the cortex superficial to the layer V projection neurons. This suggests that thalamocortical input in 2-week-old cats has devel-
oped only to the point that impulse activity can be evoked in cortical interneurons but not in projection neurons. Taken together, these observations would infer that S-I projection neurons may not be responsive to the peripheral stimulation because either thalamocortical axons or cortical interneurons have not developed functional synaptic contacts with the projection neurons. The influence of impulse activity on developing connections has been most convincingly shown in the visual system 5'6'12'34'35. When tetrodotoxin, which selectively blocks impulse activity, is injected into one eye, normally transient retinal projections from the non-injected eye to areas of the ipsilateral superior colliculus persist in the adult 14. Similarly, in the visual cortex, geniculocortical projections relaying input from both eyes initially converge onto the same cortical neurons, but later this input becomes segregated leading to the formation of ocular dominance columns. Blocking activity from the eyes prevents the development of the ocular dominance columns 34. It is important to point out, though, that the above findings are primarily examples of the effects of impulse activity on the stabilization or elimination of the terminal arborization and synapses of projections which initially overlap in the target. In this situation impulse activity serves to refine diffuse connections into topographic maps. The relationship between impulse activity and the elimination of long collateral pathways, such as the transient cerebrocerebellar projection, has not been investigated until this present study. The findings that S-I cerebrocerebellar projection neurons do not fire action potentials, either spontaneously or as the result of orthodromic activation, suggests that the collateral projection to the cerebellum may be eliminated because of the lack of impulse activity in this pathway. From this one could conclude that during development there is a critical period which, after a collateral forms its terminal arborization, it must be functionally activated or be eliminated. If projections to the cerebellum are eliminated because of the lack of impulse activity, and since many cerebrocerebellar projections are collaterals of pyramidal tract axons, why then do the projections in the PT persist into the adult? This may be due to differences in the timing of development of the arborizations of dif-
248 ferent p y r a m i d a l tract collaterals. For example, at the time c e r e b r o c e r e b e l l a r collaterals have formed mossy fiber-type terminals in the cerebellar cortex, but the S-I p r o j e c t i o n neurons are silent; collateral corticospinal projections are just beginning to form their terminal arbors in the spinal gray m a t t e r 36. The critical p e r i o d for activation of the corticospinal collaterals accordingly would then occur later in d e v e l o p m e n t than the critical p e r i o d for activation of the c e r e b r o c e r e b e l l a r collaterals. In cats, corticospinal p r o j e c t i o n s from the p r i m a r y m o t o r cortex do not a p p e a r adult-like in their distribution until about the seventh p o s t n a t a l week of d e v e l o p m e n t 37. Resting unitary activity in PT neurons in the primary m o t o r cortex gradually increases from a low rate (2/s) in the second postnatal w e e k to adult levels by the ninth postnatal week 3. A caveat to the hypothesis that the elimination of c e r e b r o c e r e b e l l a r projections is due to the absence of impulse activity in S-I p r o j e c t i o n neurons is that c e r e b r o c e r e b e l l a r projections will persist in adult cats if o t h e r afferent projections to cerebellar targets are eliminated 29. O n e possibility for the persistence
of c e r e b r o c e r e b e l l a r projections u n d e r these conditions is that cerebellar deafferentation causes a more rapid d e v e l o p m e n t of impulse activity in cortical projection neurons. A m o r e plausible explanation. however, is that the critical p e r i o d for the development of impulse activity in c e r e b r o c e r e b e l l a r axons has been e x t e n d e d as the result of the lesion-induced changes in the e n v i r o n m e n t of the partially deafferented cerebellum. A x o n - t a r g e t neuron interactions could occur in a less competitive milieu and could then sustain the transient collaterals until impulse activity develops. This would be consistent with the hypothesis that impulse activity and target interactions are necessary for the stabilization of incipient connections ~3"44.
REFERENCES
10 Dehay, C., Kennedy, H. and Bullier, J., Characterization of transient cortical projections from auditory somatosensory, and motor cortices to visual areas 17, 18, and 19 in the kitten, J. Comp. Neurol., 272 (1988) 68-89. 11 Distel, H. and Hollander, H., Autoradiographic tracing of developing subcortical projections of the occipital region in fetal rabbits, J. Comp. Neurol., 192 (1980) 505-518. 12 Dubin, M.W., Stark, L.A. and Archer, S.M., A role for action-potential activity in the development of neuronal connections in the kitten retinogeniculate pathway, J. Neurosci., 6 (1986) 1021-1036. 13 Fawcett, J.W. and O'Leary, D.D.M., The role of electrical activity in the formation of topographic maps in the nervous system, Trends Neurosci., 8 (1985) 201-206. 14 Fawcett, J.W., O'Leary, D.D.M. and Cowan, W.M., Activity and the control of ganglion cell death in the rat retina, Proc. Natl. Acad. Sci. U.S.A., 81 (1984) 5589-5593. 15 Hebb, D.O., Organization and Behavior, Wiley, New York, 1949. 16 Heilweg, E-C., Schuitz, W. and Creutzfeldt, O.D., Extracellular and intracellular recordings from cat's cortical whisker projection area: thalamus-cortical response transformation, J. Neurophysiol., 40 (1977) 463-479. 17 Huttenlocher, ER., Development of cortical neuronal activity in the neonatal cat, Exp. Neurol., 17 (1967) 247-262. 18 Innocenti, G.M., Growth and reshaping of axons in the establishment of visual callosal connections, Science, 212 (1981) 824-827. 19 Innocenti, G.M. and Caminiti, R., Postnatal shaping of callosal connections from sensory areas, Exp. Brain Res.,
1 Adams, C.E., Mihailoff, G.A. and Woodward, D.J., A transient component of the developing corticospinal tract arises in visual cortex, Neurosci. Lea., 36 (1983) 243-248. 2 Archer, S.M., Dubin, M.W. and Stark, L.A., Abnormal development of kitten retino-geniculate connectivity in the absence of action potentials, Science, 217 (1982) 743-745. 3 Amassian, V.E. and Ross, R.J., Electrophysiological correlates of the developing higher sensorimotor control system, J. Physiol. (Paris), 74 (1978) 185-201. 4 Bantli, H., Bloedel, J.R. and Tolbert, D., Activation of neurons in the cerebellar nuclei and ascending reticular formation by stimulation of the cerebellar surface, J. Neurosurg., 45 (1976) 539-554. 5 Berman, N.E. and Payne, B.R., Alterations in connections of the corpus callosum following convergent and divergent strabismus, Brain Res., 274 (1983) 201-212. 6 Berman, N.E.J. and Payne, B.R., An exuberant retinocollicular pathway in Siamese kittens: effects of competition and abnormal activity on its maturation, Dev. Brain Res., 22 (1985) 197-209. 7 Bishop, P.O., Burke, W. and Davis, R., Single unit recording from antidromically activated optic radiation neurones, J. Physiol. (Lond.), 162 (1962) 432-450. 8 Bishop, P.O., Burke, W. and Davis, R., The identification of single units in central visual pathways, J. Physiol. (Lond.), 162 (1962) 409-431. 9 Changeux, J.P. and Danchin, A., Selective stabilisation of developing synapses as a mechanism for the specification of neuronal networks, Nature (Lond.), 264 (1976) 705-711.
ACKNOWLEDGEMENTS The authors gratefully a c k n o w l e d g e the d e d i c a t e d technical assistance of Ms. Terri Harris and the assistance of Mrs. W e n d y Anzilotti in the p r e p a r a tion of this manuscript. S u p p o r t e d by N I H G r a n t NS20227.
249 38 (1980) 381-394. 20 Innocenti, G.M. and Clarke, S., Bilateral transitory projection to visual areas from auditory cortex in kittens, Dev. Brain Res., 14 91984) 143-148. 21 Ivy, G.O. and Killackey, H.P., Ontogenetic changes in the projections of neocortical neurons, J. Neurosci., 2 (1982) 735-743. 22 Jeffery, G., Arzymonow, B.J. and Lieberman, A.R., Does the early exuberant retinal projection to the superior colliculus in the neonatal rat develop synaptic connections? Dev. Brain Res., 14 (1984) 135-138. 23 Mason, C.A. and Gregory, E., Postnatal maturation of cerebeilar mossy and climbing fibers: transient expression of dual features on single axons, J. Neurosci., 4 (1984) 1715-1735. 24 McCormick, D.A. and Prince, D.A., Post-natal development of electrophysiological properties of rat cerebral cortical pyramidal neurones, J. PhysioL (Lond.), 393 (1987) 743-762. 25 Meyer, R.L., Tetrodotoxin inhibits the formation of refined retinotopography in goldfish, Dev. Brain Res., 6 (1983) 293-298. 26 O'Leary, D.D.M., Fawcett, J.W. and Cowan, W.M., Topographic targeting errors in the retinocollicular projection and their elimination by selective ganglion cell death, J. Neurosci., 6 (1986) 3692-3705. 27 O'Leary, D.D.M. and Stanfield, B.B., A transient pyramidal tract projection from the visual cortex in the hamster and its removal by selective collateral elimination, Dev. Brain Res., 27 (1986) 87-99. 28 Oka, H., Samejima, A. and Yamamoto, T., Post-natal development of pyramidal tract neurones in kittens, J. Physiol. (Lond.), 363 (1985) 481-499. 29 Panneton, W.M., The persistence of a normally transient cerebrocerebellar pathway in the cat, Dev. Brain Res., 30 (1986) 133-139. 30 Panneton, W.M. and Tolbert, D.L., The collateral origin of a transient cerebrocerebellar pathway in kittens. A study using fluorescent double-labeling techniques, Dev. Brain Res., 14 (1984) 247-254. 31 Rubel, E.W., A comparison of somatotopic organization in sensory neocortex of newborn kittens and adult cats, J. Comp. Neurol., 143 (1971) 447-480. 32 Stanfield, B,B. and O'Leary, D.D.M., The transient
33 34 35 36 37 38 39 40
41 42 43
44 45
corticospinal projection from the occipital cortex during the postnatal development of the rat, J. Comp. Neurol., 238 (1985) 236-248. Stent, G.S., A physiological mechanism for Hebb's postulate of learning, Proc. Natl. Acad. Sci. U.S.A., 70 (1973) 997-1001. Stryker, M.P. and Harris, W.A., Binocular impulse blockade prevents the formation of ocular dominance columns in cat visual cortex, J. Neurosci., 6 (1986) 2117-2133. Swindale, N.V., Absence of ocular dominance patches in dark-reared cats, Nature (Lond.), 290 (1981) 332-333. Swink, T.D. and Tolbert, D,L., The development of corticospinal projections in neonatal cats, Soc. Neurosci. Abstr., 15 (1989) 69. Theriault, E. and Tatton, W.G., Postnatal redistribution of pericruciate motor cortical projections within the kitten spinal cord, Dev. Brain Res., 45 (1989) 219-237. Tolbert, D.L., Bantli, H. and Bloedel, J.R., Anatomical and physiological evidence for a cerebellar nucleo-cortical projection in the cat, Neuroscience, 1 (1976) 205-217. Tolbert, D.L., Bantli, H. and Bloedel, J.R., Multiple branching of cerebellar efferent projections in cats, Exp. Brain Res., 31 (1978) 305-316. Tolbert, D.L., Bantli, H., Hames, E.G., Ebner, T.J., McMullen, T.A. and Bloedel, J.R., A demonstration of the dentato-reticulospinal projection in the cat, Neuroscience, 5 (1980) 1479-1488. Tolbert, D.L., Dunn, R.C. and Vogler, G.A., The postnatal development of corticotrigeminal projections in the cat, J. Comp. Neurol., 228 (1984) 478-490. Tolbert, D.L. and Panneton, W.M., Transient cerebrocerebellar projections in kittens: postnatal development and topography, J. Cornp. Neurol., 221 (1983) 216-228. Tolbert, D.L. and Panneton, W.M., The transience of cerebrocerebellar projections is due to selective elimination of axon collaterals and not neuronal death, Dev. Brain Res., 16 (1984) 301-306. Udin, S.B. and Fawcett, J.W., Formation of topographic maps, Annu. Rev. Neurosci., 1l (1988) 289-327. White, E.L., Termination of thalamic afferents in the cerebral cortex. In E.G. Jones and A. Peters (Eds.), Sensory-Motor Areas and Aspects of Cortical Connectivity, Vol. 5, Plenum, New York, 1986, pp. 271-289.