Brain Research, 576 (1992) 277-286 (~) 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/92/$05.00
277
BRES 17576
Effect of superior colliculus lesions on sensory unit responses in the intralaminar thalamus of the rat George M. Krauthamer, Jennifer G. Krol and Barry S. Grunwerg Department of Neuroscience and Cell Biology, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, Piseataway, NJ 08854 (U.S.A.)
(Accepted 19 November 1991) Key words: Intralaminar thalamus; Sensory input; Superior colliculus lesion; Kainic acid; Extracellular unit; Rat
The effects of kainic acid lesions of the intermediate and deep layers of the superior colliculus on the sensory input to the intralaminar thalamus of the rat were determined. Ipsiversive circling and contralateral sensory neglect were consistently seen after lesion placement. Two to 7 days later, the intralaminar thalamus was systematically explored for extracellular mechanoreceptive unit responses to high threshold and low threshold stimuli. On the side ipsilateral to the lesion the number of responsive units was reduced by 51%. The loss was particularly marked for nociceptive units (80%), and low threshold and complex units with orofacial receptive fields (73%). This effect may involve a partial deafferentation of the intralaminar thalamus as well as altered excitatory thresholds of thalamic neurons. It is suggested that the functionally distinct direct tectothalamic projection as well as the indirect tecto-reticulo-thalamic pathway are implicated.
INTRODUCTION The intralaminar thalamus (ILT) plays an important role in the central processing of pain 2 and in the regulation of thalamocortical arousal 3°'7°. It also provides the major subcortical extrinsic input to the neostriatum 6'3s' 7t, an input presumably linked to the complex sensorimotor functions of the basal ganglia 4'15'37. Numerous studies have shown that ILT neurons operate as state dependent integrators of sensory and central information 3'5°. While all sensory modalities contribute, the relative predominance of somatosensory, visual and auditory inputs varies among species. In the rat, much of this information originates from somatosensory receptors and reaches ILT by several parallel pathways. Best known are the spinothalamic and spinoreticulothalamic tracts and their trigeminal equivalents 8'12'41'57. Much less consideration has been given to the superior colliculus (SC) as a source of sensory inputs to ILT. Recent studies from our laboratory have provided evidence in support of such a role 27'78. Identified intermediate and deep tectal neurons antidromically activated from ILT display the same sensory properties as other intermediate and deep SC neurons with unidentified projections 45'51'64. These neurons display a general somatotopic organization with a very prominent orofacial representation. They consist essentially of two functionally distinct populations: low threshold neurons with con-
tralateral receptive fields optimally responsive to abruptly applied sensory stimuli and high threshold nociceptive neurons. In turn, stimulation of SC activates ILT neurons which also respond to sensory inputs 26. These results point to the potential importance of ILT in the elaboration of orienting behavior commonly attributed to SC. The superior coUiculus is central to attention related orienting movements in response to significant external events 13'2°'35. Somatosensory information which reaches the intermediate and deep layers of SC is conveyed rostrally to ILT and caudally to reticular and premotor regions of the brainstem and spinal cord, largely by tectal projection neurons with ascending and descending axon collaterals 25'39. This anatomical arrangement insures the concurrent activation of both pathways and indicates that ILT participates in integrating the sensory, postural and arousal or attention related components of orienting responses and in relaying them to cortex and neostriatum 9,44. The ascending projection directly links the intermediate and deep layers of SC with ILT. This substantial ipsilateral projection terminates diffusely throughout the central lateral (CL), paracentral (Pc) and parafascicular nuclei (pf)34,79. The targets of the descending deep collicular projections include wide areas of the mesencephalic and pontine reticular formation 34'63. These reticular areas give
Correspondence: G.M. Krauthamer, UMDNJ Robert Wood Johnson Medical School, Department of Neuroscience and Cell Biology, 675 Hoes Lane, Piscataway, NJ 08854, U.S.A.
278 rise to extensive ascending reticulothalamic projections to ILT and regulate thalamocortical arousal functions. Hence, activation of the reticulothalamic projection by tectoreticular neurons constitutes a second, indirect input from SC to ILT73. Unilateral lesions of SC result in ipsilateral curvature of the trunk or circling and contralateral sensory neglect 2°'42. Similar symptoms follow unilateral destruction of S S 40'66 and they have been attributed to the nigrotectal projection 32'49. These postural and sensory abnormalities may reflect, not only a disturbed output from the basal ganglia but also alterations of its input from ILT TM by way of the thalamostriatal and thalamocortical projections. In the present study we wished to determine whether SC contributed significantly to the flow of sensory information reaching ILT by systematically comparing the effect of a unilateral SC lesion on the somatosensory responses of ILT neurons ipsilateral and contralateral to the tectal lesion.
with pentobarbital Na, transcardially perfused with 0.9% saline followed by neutral buffered 10% formalin. Brains were removed and cryoprotected. Frozen sections were cut on a cryostat at 40/~m and stained with Cresyl violet for histological reconstruction of the recording tracts and analysis of lesion site and extent of gliosis. Units were classified as TP if they responded to light taps, as NX if they responded only to noxious stimuli, and as CX or complex if they responded to some combination of somatosensory stimuli such as light tapping, needle jabs, tail pinch and, in some cases, auditory stimuli. The number of units on each side, their sensory modality and their anatomical location in CL or Pf were analyzed statistically by analysis of variance and Z2 tests.
RESULTS
Extent of lesion In all cases the lesions produced an extensive subtotal destruction of the intermediate and deep layers in the anterior half of the superior colliculus, destroying approximately 40-60% of these layers (Fig. 1). The lesions were largely confined to portions of the intermediate and ROSTRAL
MATERIALS AND M E T H O D S
Preparation and lesioning procedure Fifteen adult male Long-Evans rats (225-300 g) were used for these experiments. Following induction of anesthesia with chloral hydrate, 350 mg/kg i.p., the skull was trephined for insertion of a 5/tl Hamilton syringe. In 12 animals, the needle tip was positioned midlaterally in the anterior half of the intermediate to deep layers of SC. Three additional animals received a control lesion limited to the superficial SC layers. A lesion was produced by slowly injecting 0.5 pl kainic acid (1.5 /~g/pl). The needle was withdrawn 5 min after the end of the injection, the scalp incision was closed and the animal was returned to its recovery cage. The lesions were made on the left side in ten rats and on the right side in the two others. During the survival period daily observations were made of their behavior in their home cages.
Recording and stimulation procedures Following a survival period of 2-7 days, the animals were reanesthetized, intubated, immobilized with gallamine triethiodide (Flaxedil) and artificially ventilated. Body temperature was maintained near 37°C and supplemental doses of anesthetic and Flaxedil were given when needed. The intralaminar thalamus on the lesioned and on the intact side was systematically explored with tungsten microelectrodes, 8-12 MQ (Haer & Co.). An average of 15 descents were made on each side between stereotaxic coordinates Ant. 6.5-4.5 and Lat. 0.8-1.6 using the atlas of Paxinos and Watson 55 as guide. The descents were spaced at least 200 p m apart. The number of descents and their stereotaxic coordinates were always identical on both sides of the thalamus. Extracellular recordings were made of neurons responsive to mechanical somatosensory stimulation. This consisted of non-noxious light tapping with a small cotton tipped applicator, passive rotation of joints and deep pressure. Pin pricks and tail pinch with serrated forceps were used for noxious stimulation. For each isolated neuron the entire body was explored and receptive fields were noted. The activity was stored on magnetic tape or a digital data recorder (RC Electronics) for off-line analysis.
Histology and data analysis At the end of the experiment the rats were deeply anesthetized
CAUDAL I mm Fig. 1. Extent of kainic acid lesion of superior colliculus. The injection site straddled the intermediate and deep layers of the right superior colliculus. The area of tissue destruction due to the penetration of the needle is shown in black (three upper sections of right column). The area of marked gliosis is indicated in dark gray and included the superificial layers. The area of cell loss, shown in light gray, was much more extensive than the area of intense gliosis and included most of the intermediate and deep layers of the anterior half of the superior colliculus.
279 deep layers although in a few cases some gliosis encroached on the superficial layers adjacent to the needle track or the tegmental region and central grey just below and medial to the deep white layer. Although the same volume and concentration of kainic acid were always used, the extent of neuron loss and intensity of gliosis varied. Cell loss and gliosis were maximal within an ovoid zone of about 600-800 /~m in diameter beyond which cell loss and gliosis gradually diminished. Loss of neurons was much more extensive than the area of intense gliosis; particularly striking was the total absence of the large neurons normally scattered throughout the intermediate layer (Fig. 2).
Post-lesion behavior In all animals the kainic acid lesion resulted in contralateral sensory inattention and ipsilateral postural deviations. The postural abnormalities were manifested immediately upon recovery from anesthesia; they consisted
of a marked and persistent ipsilateral curvature of the trunk, head, and tail. All animals also engaged in prolonged periods of tight ipsiversive circling. These postural abnormalities gradually waned over a period of 1-2 days after which no gross postural deviations were apparent in their spontaneous activity. Sensory inattention or neglect was profound on the side contralateral to the lesion. The animals made no overt responses to light tapping or fur stroking of any part of the contralateral body surface, whisker deflection, or moving a small light spot through the contralateral visual field. Unlike the postural abnormalities, the sensory inattention did not lessen during the length of the observation period (maximum of 7 days). Identical stimuli applied to the ipsilateral side caused an immediate and intense behavioral reaction. The animals appeared to be hypersensitive on the lesioned side even to innocuous tactile stimuli applied with a cotton tipped applicator which they frequently attempted to attack.
A
Fig. 2. Example of a typical lesion of the superior colliculus produced by the injection of 0.5/~1 kainic acid. The injection site is located laterally in the left intermediate layers. The areas enclosed by the squares are shown at higher magnification in the two lower panels. The extensive gliosis and cell loss on the lesioned site (left) contrast with the corresponding area on the intact side (right). The complete loss of the large neurons throughout the lesioned side was a consistent finding. Cresyl violet stain, 40 ktm cryostat sections. Bar = 200 ~m.
280
Somatosensory response properties The somatosensory response properties of 165 intralaminar thalamic units were analyzed in 12 rats. In each rat the same n u m b e r of descents was made on the intact and lesioned side. O n both sides the largest number of neurons were NX and CX units. Together, these two submodalities constitued 58% (n = 96/165) of all responsive units (Fig. 3).
The high threshold NX units (n = 12) responded only to tail pinch or needle pricks applied anywhere on the body. CX units (n = 84) responded optimally to noxious stimuli but they could also be activated by non-noxious light tapping and, in some instances, by auditory stimuli such as loud hand claps. Most CX units (n = 60) were activated by stimulation of any part of the body. However, 24 other CX units had contralateral receptive
A,
B,
2
2
3
3
L
...I,L ...... ,_j . . . . . . . . . . . . . . . . . . . .
~ ~LJ
|
Fig. 3. Examples of intralaminar thalamic single unit responses to somatosensory stimuli and location of recording electrode tracks. A 1 illustrates the responses of a TP unit in the anterior central lateral nucleus to light tapping of the contralateral shoulder. Another TP unit (A2), located posteriorly in the central lateral nucleus was responsive to deflection of the contralateral vibrissae, a 3 shows a parafascicular NX unit responding with brief bursts to needle jabs to the contralateral shoulder. All stimuli were applied repetitively at approximately 0.5-0.8 s intervals. The Nissl-stained coronal sections in column B show representative electrode tracks at levels corresponding to the unit responses illustrated in column A: anterior CL (B1), posterior CL (B2), and Pf (B3). Calibration for A is 320 ms.
281 fields. Some of the receptive fields (n = 13) were orofacial and included the snout, others consisted of the shoulder and forelimb (n = 11). In some cases responses could also be elicited by stimulation of the corresponding ipsilateral zone but in these instances the ipsilateral response was always w e a k e r than the contralateral one. The r e m a i n d e r (n = 69) consisted of T P units which res p o n d e d to light taps. Some TP units had large bilateral receptive fields encompassing most of the b o d y (n = 7) but the r e m a i n d e r had contralateral orofacial (n = 49) or forelimb and shoulder (n = 13) receptive fields.
LESION
INTACT
Effects of lesions The lesions resulted in a dramatic reduction in the n u m b e r of somatosensory units r e c o r d e d on the lesioned side (Fig. 4). The overall reduction a m o u n t e d to 51% (n = 111 vs 54) and a Z2 analysis showed this difference to be highly significant (Zz = 19.691, df = 1, P < 0.001). To obtain an overall view of the changes resulting from the tectal lesion, the results were t r e a t e d by a twoway analysis of variance. The responsive units were g r o u p e d by anatomical location in anterior CL, posterior CL and Pf, by sensory submodality and by laterality (Table I). The results indicated significant differences be-
LESION
INTACT
"-,
[ ,
)
1
/
.,,,m
. . . . . .
/
,111;-',
': . . . . . . . . . . . . . . . "
'
VL',
/>.
/ / // / / t \
i
5.2 ~
........ .
\
I
/
k Ii / I\
o/-/
5.7
"X~Q~.
l
z/
J
MD
// \
\\
)/
"
\
I
\
~' I
/
~_-'~
•
,/ /"
• responses
~ ~ - ~
• CX
• NX
~",
//
. no responses
• TP
Fig. 4. Distribution of units responsive to sensory stimulation in the intralaminar thalamus. The four representative cross sections on the left (based on Paxinos and Watson55) show the location and submodality of somatosensory units on the side ipsilateral and contralateral to the collicular lesion. The horizontal section on the right represents the pattern of exploratory microelectrode descents in one rat. It indicates the number of tracks which yielded and failed to yield sensory responses on each side of the intralaminar thalamus.
282 TABLE I Classification, anatomical distribution and percent reduction of neurons in the intralaminar thalamus responding to somatosensory stimulation following a unilateral lesion of the superior colliculus
The number and percent reduction (%) of responsive cells by submodality (NX, nociceptive; CX, complex; TP, tap) and total (n) are given for intact (int) and lesioned (les) side of anterior (ant) and posterior (post) central lateral nucleus and parafascicular (Pf) nucleus. Cl-ant
NX CX TP n
CL-post
Pf
int
lsn
%
int
lsn
%
int
lsn
%
4 36 17 57
0 16 12 28
0 44 70 49
0 5 12 17
0 0 8 8
-
6 18 13 37
2 9 7 18
33 50 54 49
0 67 47
tween the intact and lesioned sides (P = 0.017, df = 1,8), between anatomical regions (P = 0.017, df = 2,8), between submodalities (P = 0.005, df = 2,8), and a significant interaction between regions and submodalities ( P = 0.028, df = 2,8). Effect on neuron type. While the overall effect of the lesion was the same in CL and Pf the three submodalities were differentially affected. NX units showed the largest reduction (80%, n = 2 vs 10) followed by CX units which showed a reduction of 58% (n = 25 vs 59). TP units appeared least affected since their loss was only
36% (n = 27 vs 42). There was, however, a differential breakdown within this submodality and within CX units with respect to receptive fields (Table II). CX and TP units with contralateral orofacial receptive fields were reduced by 73% (n = 13 vs 49) whereas units with receptive fields on body and forelimb-shoulder showed a smaller reduction (25%, n = 39 vs 52). The loss of orofacial units was significant (2'2 = 7.875, df = 1, P < 0.001) but the loss of units with other receptive fields or whole body responses was not significant. Effects o f lesion on C L and Pf. Although the loss in the n u m b e r of responsive units was the same in CL (36/ 70) and Pf (16/31), the analysis of variance indicated that the different submodalities were not equally affected when
TABLE II Distribution on intact and lesioned sides and P value (Ze) of nociceptive-specific units and other units with orofacial or other receptive fields Cell type
NX CX + TP oro CX + TP body
Fig. 5. Schematic diagram of the major input-output channels of the intralaminar thalamus. Neurons of the intermediate and deep layers of the superior colliculus (SC) convey information to the thalamus directly by ascending axon collaterals and indirectly by descending axon collaterals terminating in the brainstem reticular formation (RF) which, in turn, provides the reticulothalamic input. The other major inputs from the spinothalamic tract (STT) and trigeminothalamic tract (TIT) are indicated as are the two major thalamic output targets (neostriatum and cortex). It is not known to what extent tectal and other inputs converge synaptically on the same thalamic neurons and to what extent direct contacts are established with thalamostriatal and thalamocortical neurons. Some of the reticulothalamic input is cholinergic and all of these pathways are presumed to be excitatory.
the units were grouped by location in anterior or posterior CL and Pf. The difference was attributed to the large number of CX and TP units concentrated in anterior CL where they constituted 57% (n = 81/143) of all non-NX units. DISCUSSION It is evident that the effectiveness of sensory stimuli to excite ILT neurons is markedly reduced by SC lesions, indicating that, in addition to the traditional inputs to ILT, the ascending influence of SC must be taken into account. The consequences of such a partial deafferentation of ILT are likely to be widespread in view of the extensive projections of this region to cortex and neostriatum. This raises the question whether the altered responsivity of the thalamic neurons plays a significant role in the abnormal posture and the sensory neglect typically associated with unilateral lesions of the intermediate and deep collicular layers. Behavioral effects The ipsiversive circling and the contralateral sensory
ILT Intact
Lesion
P
10 49 52
2 13 39
0.043
277
BRES 17576
Effect of superior colliculus lesions on sensory unit responses in the intralaminar thalamus of the rat George M. Krauthamer, Jennifer G. Krol and Barry S. Grunwerg Department of Neuroscience and Cell Biology, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, Piseataway, NJ 08854 (U.S.A.)
(Accepted 19 November 1991) Key words: Intralaminar thalamus; Sensory input; Superior colliculus lesion; Kainic acid; Extracellular unit; Rat
The effects of kainic acid lesions of the intermediate and deep layers of the superior colliculus on the sensory input to the intralaminar thalamus of the rat were determined. Ipsiversive circling and contralateral sensory neglect were consistently seen after lesion placement. Two to 7 days later, the intralaminar thalamus was systematically explored for extracellular mechanoreceptive unit responses to high threshold and low threshold stimuli. On the side ipsilateral to the lesion the number of responsive units was reduced by 51%. The loss was particularly marked for nociceptive units (80%), and low threshold and complex units with orofacial receptive fields (73%). This effect may involve a partial deafferentation of the intralaminar thalamus as well as altered excitatory thresholds of thalamic neurons. It is suggested that the functionally distinct direct tectothalamic projection as well as the indirect tecto-reticulo-thalamic pathway are implicated.
INTRODUCTION The intralaminar thalamus (ILT) plays an important role in the central processing of pain 2 and in the regulation of thalamocortical arousal 3°'7°. It also provides the major subcortical extrinsic input to the neostriatum 6'3s' 7t, an input presumably linked to the complex sensorimotor functions of the basal ganglia 4'15'37. Numerous studies have shown that ILT neurons operate as state dependent integrators of sensory and central information 3'5°. While all sensory modalities contribute, the relative predominance of somatosensory, visual and auditory inputs varies among species. In the rat, much of this information originates from somatosensory receptors and reaches ILT by several parallel pathways. Best known are the spinothalamic and spinoreticulothalamic tracts and their trigeminal equivalents 8'12'41'57. Much less consideration has been given to the superior colliculus (SC) as a source of sensory inputs to ILT. Recent studies from our laboratory have provided evidence in support of such a role 27'78. Identified intermediate and deep tectal neurons antidromically activated from ILT display the same sensory properties as other intermediate and deep SC neurons with unidentified projections 45'51'64. These neurons display a general somatotopic organization with a very prominent orofacial representation. They consist essentially of two functionally distinct populations: low threshold neurons with con-
tralateral receptive fields optimally responsive to abruptly applied sensory stimuli and high threshold nociceptive neurons. In turn, stimulation of SC activates ILT neurons which also respond to sensory inputs 26. These results point to the potential importance of ILT in the elaboration of orienting behavior commonly attributed to SC. The superior coUiculus is central to attention related orienting movements in response to significant external events 13'2°'35. Somatosensory information which reaches the intermediate and deep layers of SC is conveyed rostrally to ILT and caudally to reticular and premotor regions of the brainstem and spinal cord, largely by tectal projection neurons with ascending and descending axon collaterals 25'39. This anatomical arrangement insures the concurrent activation of both pathways and indicates that ILT participates in integrating the sensory, postural and arousal or attention related components of orienting responses and in relaying them to cortex and neostriatum 9,44. The ascending projection directly links the intermediate and deep layers of SC with ILT. This substantial ipsilateral projection terminates diffusely throughout the central lateral (CL), paracentral (Pc) and parafascicular nuclei (pf)34,79. The targets of the descending deep collicular projections include wide areas of the mesencephalic and pontine reticular formation 34'63. These reticular areas give
Correspondence: G.M. Krauthamer, UMDNJ Robert Wood Johnson Medical School, Department of Neuroscience and Cell Biology, 675 Hoes Lane, Piscataway, NJ 08854, U.S.A.
278 rise to extensive ascending reticulothalamic projections to ILT and regulate thalamocortical arousal functions. Hence, activation of the reticulothalamic projection by tectoreticular neurons constitutes a second, indirect input from SC to ILT73. Unilateral lesions of SC result in ipsilateral curvature of the trunk or circling and contralateral sensory neglect 2°'42. Similar symptoms follow unilateral destruction of S S 40'66 and they have been attributed to the nigrotectal projection 32'49. These postural and sensory abnormalities may reflect, not only a disturbed output from the basal ganglia but also alterations of its input from ILT TM by way of the thalamostriatal and thalamocortical projections. In the present study we wished to determine whether SC contributed significantly to the flow of sensory information reaching ILT by systematically comparing the effect of a unilateral SC lesion on the somatosensory responses of ILT neurons ipsilateral and contralateral to the tectal lesion.
with pentobarbital Na, transcardially perfused with 0.9% saline followed by neutral buffered 10% formalin. Brains were removed and cryoprotected. Frozen sections were cut on a cryostat at 40/~m and stained with Cresyl violet for histological reconstruction of the recording tracts and analysis of lesion site and extent of gliosis. Units were classified as TP if they responded to light taps, as NX if they responded only to noxious stimuli, and as CX or complex if they responded to some combination of somatosensory stimuli such as light tapping, needle jabs, tail pinch and, in some cases, auditory stimuli. The number of units on each side, their sensory modality and their anatomical location in CL or Pf were analyzed statistically by analysis of variance and Z2 tests.
RESULTS
Extent of lesion In all cases the lesions produced an extensive subtotal destruction of the intermediate and deep layers in the anterior half of the superior colliculus, destroying approximately 40-60% of these layers (Fig. 1). The lesions were largely confined to portions of the intermediate and ROSTRAL
MATERIALS AND M E T H O D S
Preparation and lesioning procedure Fifteen adult male Long-Evans rats (225-300 g) were used for these experiments. Following induction of anesthesia with chloral hydrate, 350 mg/kg i.p., the skull was trephined for insertion of a 5/tl Hamilton syringe. In 12 animals, the needle tip was positioned midlaterally in the anterior half of the intermediate to deep layers of SC. Three additional animals received a control lesion limited to the superficial SC layers. A lesion was produced by slowly injecting 0.5 pl kainic acid (1.5 /~g/pl). The needle was withdrawn 5 min after the end of the injection, the scalp incision was closed and the animal was returned to its recovery cage. The lesions were made on the left side in ten rats and on the right side in the two others. During the survival period daily observations were made of their behavior in their home cages.
Recording and stimulation procedures Following a survival period of 2-7 days, the animals were reanesthetized, intubated, immobilized with gallamine triethiodide (Flaxedil) and artificially ventilated. Body temperature was maintained near 37°C and supplemental doses of anesthetic and Flaxedil were given when needed. The intralaminar thalamus on the lesioned and on the intact side was systematically explored with tungsten microelectrodes, 8-12 MQ (Haer & Co.). An average of 15 descents were made on each side between stereotaxic coordinates Ant. 6.5-4.5 and Lat. 0.8-1.6 using the atlas of Paxinos and Watson 55 as guide. The descents were spaced at least 200 p m apart. The number of descents and their stereotaxic coordinates were always identical on both sides of the thalamus. Extracellular recordings were made of neurons responsive to mechanical somatosensory stimulation. This consisted of non-noxious light tapping with a small cotton tipped applicator, passive rotation of joints and deep pressure. Pin pricks and tail pinch with serrated forceps were used for noxious stimulation. For each isolated neuron the entire body was explored and receptive fields were noted. The activity was stored on magnetic tape or a digital data recorder (RC Electronics) for off-line analysis.
Histology and data analysis At the end of the experiment the rats were deeply anesthetized
CAUDAL I mm Fig. 1. Extent of kainic acid lesion of superior colliculus. The injection site straddled the intermediate and deep layers of the right superior colliculus. The area of tissue destruction due to the penetration of the needle is shown in black (three upper sections of right column). The area of marked gliosis is indicated in dark gray and included the superificial layers. The area of cell loss, shown in light gray, was much more extensive than the area of intense gliosis and included most of the intermediate and deep layers of the anterior half of the superior colliculus.
279 deep layers although in a few cases some gliosis encroached on the superficial layers adjacent to the needle track or the tegmental region and central grey just below and medial to the deep white layer. Although the same volume and concentration of kainic acid were always used, the extent of neuron loss and intensity of gliosis varied. Cell loss and gliosis were maximal within an ovoid zone of about 600-800 /~m in diameter beyond which cell loss and gliosis gradually diminished. Loss of neurons was much more extensive than the area of intense gliosis; particularly striking was the total absence of the large neurons normally scattered throughout the intermediate layer (Fig. 2).
Post-lesion behavior In all animals the kainic acid lesion resulted in contralateral sensory inattention and ipsilateral postural deviations. The postural abnormalities were manifested immediately upon recovery from anesthesia; they consisted
of a marked and persistent ipsilateral curvature of the trunk, head, and tail. All animals also engaged in prolonged periods of tight ipsiversive circling. These postural abnormalities gradually waned over a period of 1-2 days after which no gross postural deviations were apparent in their spontaneous activity. Sensory inattention or neglect was profound on the side contralateral to the lesion. The animals made no overt responses to light tapping or fur stroking of any part of the contralateral body surface, whisker deflection, or moving a small light spot through the contralateral visual field. Unlike the postural abnormalities, the sensory inattention did not lessen during the length of the observation period (maximum of 7 days). Identical stimuli applied to the ipsilateral side caused an immediate and intense behavioral reaction. The animals appeared to be hypersensitive on the lesioned side even to innocuous tactile stimuli applied with a cotton tipped applicator which they frequently attempted to attack.
A
Fig. 2. Example of a typical lesion of the superior colliculus produced by the injection of 0.5/~1 kainic acid. The injection site is located laterally in the left intermediate layers. The areas enclosed by the squares are shown at higher magnification in the two lower panels. The extensive gliosis and cell loss on the lesioned site (left) contrast with the corresponding area on the intact side (right). The complete loss of the large neurons throughout the lesioned side was a consistent finding. Cresyl violet stain, 40 ktm cryostat sections. Bar = 200 ~m.
280
Somatosensory response properties The somatosensory response properties of 165 intralaminar thalamic units were analyzed in 12 rats. In each rat the same n u m b e r of descents was made on the intact and lesioned side. O n both sides the largest number of neurons were NX and CX units. Together, these two submodalities constitued 58% (n = 96/165) of all responsive units (Fig. 3).
The high threshold NX units (n = 12) responded only to tail pinch or needle pricks applied anywhere on the body. CX units (n = 84) responded optimally to noxious stimuli but they could also be activated by non-noxious light tapping and, in some instances, by auditory stimuli such as loud hand claps. Most CX units (n = 60) were activated by stimulation of any part of the body. However, 24 other CX units had contralateral receptive
A,
B,
2
2
3
3
L
...I,L ...... ,_j . . . . . . . . . . . . . . . . . . . .
~ ~LJ
|
Fig. 3. Examples of intralaminar thalamic single unit responses to somatosensory stimuli and location of recording electrode tracks. A 1 illustrates the responses of a TP unit in the anterior central lateral nucleus to light tapping of the contralateral shoulder. Another TP unit (A2), located posteriorly in the central lateral nucleus was responsive to deflection of the contralateral vibrissae, a 3 shows a parafascicular NX unit responding with brief bursts to needle jabs to the contralateral shoulder. All stimuli were applied repetitively at approximately 0.5-0.8 s intervals. The Nissl-stained coronal sections in column B show representative electrode tracks at levels corresponding to the unit responses illustrated in column A: anterior CL (B1), posterior CL (B2), and Pf (B3). Calibration for A is 320 ms.
281 fields. Some of the receptive fields (n = 13) were orofacial and included the snout, others consisted of the shoulder and forelimb (n = 11). In some cases responses could also be elicited by stimulation of the corresponding ipsilateral zone but in these instances the ipsilateral response was always w e a k e r than the contralateral one. The r e m a i n d e r (n = 69) consisted of T P units which res p o n d e d to light taps. Some TP units had large bilateral receptive fields encompassing most of the b o d y (n = 7) but the r e m a i n d e r had contralateral orofacial (n = 49) or forelimb and shoulder (n = 13) receptive fields.
LESION
INTACT
Effects of lesions The lesions resulted in a dramatic reduction in the n u m b e r of somatosensory units r e c o r d e d on the lesioned side (Fig. 4). The overall reduction a m o u n t e d to 51% (n = 111 vs 54) and a Z2 analysis showed this difference to be highly significant (Zz = 19.691, df = 1, P < 0.001). To obtain an overall view of the changes resulting from the tectal lesion, the results were t r e a t e d by a twoway analysis of variance. The responsive units were g r o u p e d by anatomical location in anterior CL, posterior CL and Pf, by sensory submodality and by laterality (Table I). The results indicated significant differences be-
LESION
INTACT
"-,
[ ,
)
1
/
.,,,m
. . . . . .
/
,111;-',
': . . . . . . . . . . . . . . . "
'
VL',
/>.
/ / // / / t \
i
5.2 ~
........ .
\
I
/
k Ii / I\
o/-/
5.7
"X~Q~.
l
z/
J
MD
// \
\\
)/
"
\
I
\
~' I
/
~_-'~
•
,/ /"
• responses
~ ~ - ~
• CX
• NX
~",
//
. no responses
• TP
Fig. 4. Distribution of units responsive to sensory stimulation in the intralaminar thalamus. The four representative cross sections on the left (based on Paxinos and Watson55) show the location and submodality of somatosensory units on the side ipsilateral and contralateral to the collicular lesion. The horizontal section on the right represents the pattern of exploratory microelectrode descents in one rat. It indicates the number of tracks which yielded and failed to yield sensory responses on each side of the intralaminar thalamus.
282 TABLE I Classification, anatomical distribution and percent reduction of neurons in the intralaminar thalamus responding to somatosensory stimulation following a unilateral lesion of the superior colliculus
The number and percent reduction (%) of responsive cells by submodality (NX, nociceptive; CX, complex; TP, tap) and total (n) are given for intact (int) and lesioned (les) side of anterior (ant) and posterior (post) central lateral nucleus and parafascicular (Pf) nucleus. Cl-ant
NX CX TP n
CL-post
Pf
int
lsn
%
int
lsn
%
int
lsn
%
4 36 17 57
0 16 12 28
0 44 70 49
0 5 12 17
0 0 8 8
-
6 18 13 37
2 9 7 18
33 50 54 49
0 67 47
tween the intact and lesioned sides (P = 0.017, df = 1,8), between anatomical regions (P = 0.017, df = 2,8), between submodalities (P = 0.005, df = 2,8), and a significant interaction between regions and submodalities ( P = 0.028, df = 2,8). Effect on neuron type. While the overall effect of the lesion was the same in CL and Pf the three submodalities were differentially affected. NX units showed the largest reduction (80%, n = 2 vs 10) followed by CX units which showed a reduction of 58% (n = 25 vs 59). TP units appeared least affected since their loss was only
36% (n = 27 vs 42). There was, however, a differential breakdown within this submodality and within CX units with respect to receptive fields (Table II). CX and TP units with contralateral orofacial receptive fields were reduced by 73% (n = 13 vs 49) whereas units with receptive fields on body and forelimb-shoulder showed a smaller reduction (25%, n = 39 vs 52). The loss of orofacial units was significant (2'2 = 7.875, df = 1, P < 0.001) but the loss of units with other receptive fields or whole body responses was not significant. Effects o f lesion on C L and Pf. Although the loss in the n u m b e r of responsive units was the same in CL (36/ 70) and Pf (16/31), the analysis of variance indicated that the different submodalities were not equally affected when
TABLE II Distribution on intact and lesioned sides and P value (Ze) of nociceptive-specific units and other units with orofacial or other receptive fields Cell type
NX CX + TP oro CX + TP body
Fig. 5. Schematic diagram of the major input-output channels of the intralaminar thalamus. Neurons of the intermediate and deep layers of the superior colliculus (SC) convey information to the thalamus directly by ascending axon collaterals and indirectly by descending axon collaterals terminating in the brainstem reticular formation (RF) which, in turn, provides the reticulothalamic input. The other major inputs from the spinothalamic tract (STT) and trigeminothalamic tract (TIT) are indicated as are the two major thalamic output targets (neostriatum and cortex). It is not known to what extent tectal and other inputs converge synaptically on the same thalamic neurons and to what extent direct contacts are established with thalamostriatal and thalamocortical neurons. Some of the reticulothalamic input is cholinergic and all of these pathways are presumed to be excitatory.
the units were grouped by location in anterior or posterior CL and Pf. The difference was attributed to the large number of CX and TP units concentrated in anterior CL where they constituted 57% (n = 81/143) of all non-NX units. DISCUSSION It is evident that the effectiveness of sensory stimuli to excite ILT neurons is markedly reduced by SC lesions, indicating that, in addition to the traditional inputs to ILT, the ascending influence of SC must be taken into account. The consequences of such a partial deafferentation of ILT are likely to be widespread in view of the extensive projections of this region to cortex and neostriatum. This raises the question whether the altered responsivity of the thalamic neurons plays a significant role in the abnormal posture and the sensory neglect typically associated with unilateral lesions of the intermediate and deep collicular layers. Behavioral effects The ipsiversive circling and the contralateral sensory
ILT Intact
Lesion
P
10 49 52
2 13 39
0.043
