Brain Research Bulkfin,
Vol. 3, pp. 533-539. printed in the U.S.A.
Neuronal Activity in the Midbrain Reticular Formation Related to Behavioral Habituation and Sensitization JOHN P. KOOTZ’ AND HARRY M. SINNAMON2 Laboratory
of Neuropsychology,
Wesleyan
(Received
University,
9 January
Middletown,
CT 06457
1978)
KOOTZ, J. P. AND H. M. SINNAMON. Neuronul ucfivity in the midhroin reticular formution related to hehuviorul BRAIN RES. BULL. 3(5) 533-539, 1978.-Single units of the midbrain reticular formation (MRF) were recorded in the active rat (N=lO) during the presentation of renetitive trains of clicks or flashes. The movement and orienting behavior elicited by the stimuli, measured by direct observation and in terms of electrical disturbances in the recording cable, showed habituation in 14 cases, sensitization in six cases, and no decrement in nine cases. Of the 18 MRF units recorded during at least one of these behavioral conditions, seven were classed as congruent, i.e., they showed activity that paralleled the behavioral response. The majority of MRF units were behaviorally incongruent; in the most frequently observed case, their activity failed to change reliably when behavior habituated. Although the activity of congruent units generally showed moderate correlations with behavior that followed stimulation, disassociations between the two responses were commonly found. While the data are consistent with a role for the MRF in behavioral habituation and sensitization, they provide little evidence that the region is closely linked to the systems in direct control of behavior.
hubituution and sensitization.
Reticular formation
Habituation
Chronic unit recording
THE BRAINSTEM reticular formation has been a particular focus of interest in the search for the neural basis of habituation. In the most recent theoretical treatment, Groves and Lynch 151 suggested that the smaller cell regions of the reticular formation, e.g., the nucleus cuneiformis in the midbrain, lie at the early stages of the processing chain responsible for habituation and sensitization and transmit their information to the larger cells of the reticular formation, e.g., nucleus gigantocellularis, which are more tightly coupled to behavior. Electrophysiology of single units in the reticular formation has consistently shown that the area has the characteristics required for the control of habituation. Activity of single units in the midbrain reticular formation (MRF) shows patterns of habituation, dishabituation, and sensitization [2, 6, 9, 10, 171 that resemble the patterns typical of behavioral habituation. The pontomedullary reticular formation contains cells showing similar patterns of response [8, 14, 16, 17, 191. With few exceptions [9,18], studies of response plasticity in neurons of the reticular formation have used preparations that were anesthetized or otherwise immobilized. A critical question raised by many of these studies is how the changes in neuronal response relate to actual behavior. There is a need for studies which examine reticular formation unit activity during behavior elicited by novel stimuli. In the present study, single units in the MRF of the active rat were
Sensitization
Orienting response
recorded during the presentation of repetitive auditory and visual stimuli. Cases of behavioral habituation and sensitization were observed and unit activity was found to correlate positively in some cases. However, many instances of disassociation between MRF unit activity and behavior were found, relationships which suggest a complex role of the MRF in the control of behavior during habituation. METHOD
Animals Male (N=6) and female (N=4) DA Agouti rats from the Wesleyan colony were used. They were between 4 and 8 months at time of surgery and housed in individual cages with a reversed light cycle. Testing occurred during the dark phase of the cycle. Surgev
The rats were prepared for chronic recording of single units by the placement of a 6.1 mm, circular, threaded, nylon well on the skull dorsal to the MRF. For this procedure the animals were anesthetized with intraperitoneal injections of Nembutal (30-35 mg/kg), supplemented by local application of 2% Xylocaine. The rat was mounted in a stereotaxic apparatus and a hole of 5 mm dia. was drilled in the skull.
‘Present address: Child Study Center, Yale University, New Haven, CT 06510. ‘Requests for reprints should be sent to: H. M. Sinnamon, Laboratory of Neuropsychology,
Copyright o 1978 ANKHO International
Attention
Wesleyan University, Middletown, CT 06457.
Inc.-0361-9230/78/050533-07$01.20/O
534
Bacymycin, an antibiotic ointment, was applied to the dura which was left intact. The nylon well with its base sealed with a layer of Partilm was placed over the opening and secured with acrylic cement. Five stainless steel screws distributed over the skull served to anchor the cement. An uninsulated copper wire, wrapped around the screws, served as the ground connection. Recording electrodes were also placed in the hippocampus and over the frontal cortex although the results of these recordings are not reported here. All electrodes were connected to Amphenol pins which were inserted into an Amphenol connector strip. The strip was also secured to the skull with acrylic cement and provided the means of attachment for the recording cable. A recovery period of 5-7 days was allowed.
The test chamber was a cylinder made of clear Plexiglas, 27.9 cm dia. and 27.9 cm high from the base, which was also made from Plexiglas. The cylinder was surrounded by a copper screen which served as shielding against 60 Hz interference. Opaque paper covered the lower 20 cm of the chamber and blocked the view of the subject. A trio of speakers (76 mm dia.) were placed 90” apart on the outside of the cylinder just above the base. The wall of the chamber had openings at the points where the speakers were attached. A flashhead of a Grass photo stimulator was mounted 0.6 m above the chamber. Also mounted above the chamber were a counterbalanced arm assembly with a nine-channel commutator, and a small convex mirror that allowed indirect observation of the animal’s behavior. The testing occurred in a dimly lit room with minimal ambient noise. Masking noise (approximately 60 db) produced by a white noise generator was amplified and delivered to the center speaker of the testing chamber. Two types of stimuli were presented to the rat. The first type was used in the standard tests for sensory responsiveness that preceded the tests for habituation. The standard visual stimulus was a grid made of white hardware cloth covering a black rectangle (10x 15 cm). The auditory stimulus was a click produced by a microswitch held about 10 cm from the rat’s head. Tactile stimuli were produced by probing the rat’s head and body with a pencil Stimuli used in tests for habituation were clicks and flashes controlled by solid-state logic circuitry. The frequency of the stimuli varied from 10 to 100 Hz and the train length was held constantly at approximately 1 sec. The clicks were delivered to one of the two opposing speakers of the chamber. The speaker selection was controlled manually and changed periodically to maintain the novelty of the stimulus. The loudness of the clicks was adjusted between approximately 75 and 88 db so that a locomotor movement or head turn, but not a startle or escape response, was produced. The flashes used in the tests for habituation were produced by a Grass PS 22 photostimulator. Intensity was set at a nominal level of 16.
Single neuronaf activity was recorded through tungsten microelectrodes that were insulated with glass up to 10 PM from their tips which were approximately of I J.LMdia. The electrode was lowered into the brain by means of a small microdrive fitted to the nylon well on the skuI1 of the rat. A short wire from the electrode led directly to a miniature pre-
KOOTZ AND SINNAMON amplifier mounted nearby on the end of a cable composed of Microdot wires. Additional details of the system may be found in previous reports [15,20]. The output of the amplifier was subjected to high frequency filtering and led to one channel to a Tektronix 502 oscilloscope for additional amplification. The output of the vertical amplifier of the oscilloscope was led to a spike discriminator which produced a 5 V pulse when spikes of a selected size and duration occurred. The disc~minator output was sent to a solid-state counter which printed accumulated counts at intervals controlled by the stimulus control circuitry. The discriminator output was also integrated over 1 set periods and displayed on one channel of a Beckman R411 polygraph. The accuracy of spike disc~mination was continuously monitored by one experimenter, a precaution necessitated by the changes in unit amplitude that sometimes occurred with tests lasting several minutes. The unit activity was also led to one channel of a Sony tape recorder. Movement of the animal in response to the stimuli presented during the habituation tests was measured in two ways. The simpler and generalIy more effective method was direct observation of the rat by one experimenter during the stimulus and for 3 set thereafter. This measure was dichotomous, indicating only whether a behavioral response occurred or not, and did not reflect the intensity, latency or duration of the response. In order to obtain quantitative measures of the behavioral response, two unte~inated wires were attached to the recording cable and led to the differential inputs of a Grass P15 preamplifier. Electrical disturbances in these wires caused by the rats’ movements were displayed on one channel of the polygraph along with stimulus markers and the integrated unit activity. The amplitude of pen deflection on the movement detector failed to relate systematically to the intensity of behavior as judged by direct observation but the duration of the deflection often did correlate with the time course of the response observed. Therefore, the movement record was measured in terms of the time over which the tracing was above a threshold level. However, a dficulty was encountered due to individual differences between rats in terms of the topography of behavior elicited. For some rats, the movement detector was insensitive to the behavior directly observed to be elicited by the stimuli. Adjustment of the movement detector sensitivity before the habituation tests was precluded by the need to preserve the general responsiveness of rat to the test stimuli. Thus the principal behavioral measure was the qualitative index, which was supplemented in some cases by the quantitative movement detector.
Prior to testing, the rats were lightly anesthetized with halothane (0.50/c, 1 liter per min), and the microdrive and recording cable were attached. The microelectrode was lowered into the cortex dorsal to the MRF in order to assess the recording capabilities of the electrode. If well isolated, units with peak-to-peak amplitudes greater than 300 PV were not found within 2 mm, the electrode was replaced and another penetration was made in a new location. Once adequate recording was established, the anesthesia was discontinued. A period of 45 min was allowed for recovery from the anesthesia and for adaptation to the chamber. Testing for habituation and sensitization began only after a single unit was isolated for at least 5 min without appreciable changes in amplitude or signs of injury (e.g., notched
MRF ACTIVITY
535
AND HABITUATION
waveforms). Prior to the main tests, the sensory responsiveness of the unit was assessed by presenting the visual, auditory and tactile stimuli several times. Generally, the microswitch clicks were presented first at various points in a IO cm radius around the rat’s head. Next, the moving visual stimulus was presented to nasal and temporal visual fields on both sides. Finally, tactile sensitivity was tested by probing the head, vibrissa, neck, back and lateral aspects of the limbs. Habituation, of course, was a potential problem in these tests since several units were tested on each rat, but care was taken to assure that the rat showed at least a behavioral response to these stimuli. A unit was considered responsive if the two experimenters agreed that a clear increase or decrease attended each stimulus presentation. After sensory testing the more formal habituation tests began. First a series of control trials were given in which all of the logic and analysis apparatus operated without the presentation of test stimuli. Pilot work had shown that occasionally the rats reacted to noises made by coincidental equipment, e.g., the printer, and the control trials habituated these responses. Next, the series of click or flash stimuli were presented. In each series the stimulus was kept constant in frequency. To apply the criteria for habituation or sensitization of behavior and unit activity, a minimum of six trials was required. Generally about four series of stimuli were presented with each unit. Not all series resulted in usable data. Most often the series were terminated before the required six trials when it was apparent that neither a behavioral or unit activity response was evoked. The loss of isolation of the single unit was another principle reason for the discounting of a series. The stimuli were changed as needed in frequency, speaker duration and type (flash or click) in order to revive the behavioral responses. No one order of stimulus presentation was given to all animals. Occasionally the clicks and flashes were paired on one trial after a habituation occurred as a test for dishabituation. Trials were begun only when the rat was still. Consequently, the interval between trials was not constant; generally it was between 5 and 10 sec. Occasionally a movement occurred between the beginning of the 1 set prestimulus period and the stimulus. When this movement was large enough to obscure any response to the stimulus the trial was not used.
For series of stimulus presentations with up to 16 trials, the first three trials were compared to each successive sets of three subsequent trials. For series with 17 to 24 trials, sets of four trials were compared. The same analysis was applied to the counts of single unit activity in the four 1-set bins of a trial. RESULTS
Generai Characteristics
of MRF Units
The units were characterized in terms of location, responsiveness during the standard sensory tests, baseline firing rate, and general responsiveness to the stimuli presented during habituation tests. Locations of all units recorded, including those in the MRF, are shown in Fig. 1. The 24 units classed as reticular are shown lateral to the central gray and ventral and medial to the dashed lines. This area generally conforms to the nucleus cuneiformis as described by Valverde 1211. Another 12 units outside of the MRF were recorded and tested. These units, located in the area of mergence between the MRF and tectum (N=4), the medial geniculate nucleus (N=2), the hippocampal formation (N=2) and the neocortex (N=4), were a diverse group which served for comparison to the MRF units.
Histology A small electrolytic lesion (100 PA, 20 set) was made at the deepest point of the electrode penetration. The rat was anesthetized with Nembutal and perfused through the heart with physiological saline followed by 10% Formalin. The brains were cut in transverse sections (16 pm) with a microtome cryostat and stained with cresyl violet. Locations of units were dete~ined relative to the marker lesion. Data Analysis Classifications of behavioral habituation and sensitization were based primarily on the dichotomous measure of movement as directly observed. Habituation was delcared if: (a) a movement occurred on two of the initial three trials of a series, and (b) no movement occurred on two of any three consecutive and subsequent trials. The criterion for sensitization was the converse. The quantitative movement detector scores were averaged over four I-set periods beginning with the onset of stimulation and were tested for significant (~~0.05) increases or decreases by Mann Whitney U tests.
FIG. 1. Sensory responsiveness and locations of units tested projected onto plates adapted from the atlas of Kiinig and Klippel 1121. The stimuli were clicks, a moving visual stimulus and tactile stimulation of the head and body. The area enclosed with the dashed lines is the MRF.
The response of the single units to auditory, tactile and visual stimuli presented during the standard sensory tests that preceded tests for habituation are shown in Fig. I. All of the 23 MRF units on which data was available responded to at least one of the stimuli. By contrast, the units in the non-
KOOTZ AND SINNA~ON
536
MRF areas were significantly less responsive (Fisher Exact Test, p=O.O09) with one third failing to show a response during the preliminary sensory testing. The MRF units favored tactile stimulation, with over 80% (18/23) of the units responding. SignificantIy fewer b-0.09) of the units in the other areas responded to tactile stimulation. MRF units were afso more responsive to the auditory stimulus 0, =0.04), 16 of 23 responsive MRF units compared to four of 12 non-MRF units. The visual stimulus was the least effective for both the MRF units (eight of 23 responsive) and the non-MRF units (two of 12 responsive). The MRF units showed significantly more multimodal responses than units in other areas @ =0.008). These multimodal responses were distributed throughout the MRF regions studied and are shown as filledin symbols of Fig. 1. The mean baseline firing rates of the MRF units were determined from the I-see prestimulation periods in all stimulation trials in the tests for habituation. It should be noted that trials were begun only when the rat was immobile. The rates ranged from 0.04 to spikeslsec to 29.7 spikeslsec with a median of 2.4 spikes/set. One third of the units fired at rates of llsec or less. Tactile and auditory responses were found in units with a wide range of baseline rates but visual responses were confined, with one exception, to units with slower than average baseline rates (~0.5 spikes/see). No tendency was discerned for units with slow rates to congregate in any particular region of the MRF. The general responsiveness of each MRF unit to the stimuli presented during the tests for habituation and sensitization was assessed by paired 1 tests comparing the rates in all of the prestimulation periods to the rates in the periods with stimulation. All stimulus conditions were combined so that a general index of responsivity was obtained rather than a specific measure of response to a particular stimulus. Nearly all MRF units (20 of 24) showed signi~cant responses @
Vol. 3, pp. 533-539. printed in the U.S.A.
Neuronal Activity in the Midbrain Reticular Formation Related to Behavioral Habituation and Sensitization JOHN P. KOOTZ’ AND HARRY M. SINNAMON2 Laboratory
of Neuropsychology,
Wesleyan
(Received
University,
9 January
Middletown,
CT 06457
1978)
KOOTZ, J. P. AND H. M. SINNAMON. Neuronul ucfivity in the midhroin reticular formution related to hehuviorul BRAIN RES. BULL. 3(5) 533-539, 1978.-Single units of the midbrain reticular formation (MRF) were recorded in the active rat (N=lO) during the presentation of renetitive trains of clicks or flashes. The movement and orienting behavior elicited by the stimuli, measured by direct observation and in terms of electrical disturbances in the recording cable, showed habituation in 14 cases, sensitization in six cases, and no decrement in nine cases. Of the 18 MRF units recorded during at least one of these behavioral conditions, seven were classed as congruent, i.e., they showed activity that paralleled the behavioral response. The majority of MRF units were behaviorally incongruent; in the most frequently observed case, their activity failed to change reliably when behavior habituated. Although the activity of congruent units generally showed moderate correlations with behavior that followed stimulation, disassociations between the two responses were commonly found. While the data are consistent with a role for the MRF in behavioral habituation and sensitization, they provide little evidence that the region is closely linked to the systems in direct control of behavior.
hubituution and sensitization.
Reticular formation
Habituation
Chronic unit recording
THE BRAINSTEM reticular formation has been a particular focus of interest in the search for the neural basis of habituation. In the most recent theoretical treatment, Groves and Lynch 151 suggested that the smaller cell regions of the reticular formation, e.g., the nucleus cuneiformis in the midbrain, lie at the early stages of the processing chain responsible for habituation and sensitization and transmit their information to the larger cells of the reticular formation, e.g., nucleus gigantocellularis, which are more tightly coupled to behavior. Electrophysiology of single units in the reticular formation has consistently shown that the area has the characteristics required for the control of habituation. Activity of single units in the midbrain reticular formation (MRF) shows patterns of habituation, dishabituation, and sensitization [2, 6, 9, 10, 171 that resemble the patterns typical of behavioral habituation. The pontomedullary reticular formation contains cells showing similar patterns of response [8, 14, 16, 17, 191. With few exceptions [9,18], studies of response plasticity in neurons of the reticular formation have used preparations that were anesthetized or otherwise immobilized. A critical question raised by many of these studies is how the changes in neuronal response relate to actual behavior. There is a need for studies which examine reticular formation unit activity during behavior elicited by novel stimuli. In the present study, single units in the MRF of the active rat were
Sensitization
Orienting response
recorded during the presentation of repetitive auditory and visual stimuli. Cases of behavioral habituation and sensitization were observed and unit activity was found to correlate positively in some cases. However, many instances of disassociation between MRF unit activity and behavior were found, relationships which suggest a complex role of the MRF in the control of behavior during habituation. METHOD
Animals Male (N=6) and female (N=4) DA Agouti rats from the Wesleyan colony were used. They were between 4 and 8 months at time of surgery and housed in individual cages with a reversed light cycle. Testing occurred during the dark phase of the cycle. Surgev
The rats were prepared for chronic recording of single units by the placement of a 6.1 mm, circular, threaded, nylon well on the skull dorsal to the MRF. For this procedure the animals were anesthetized with intraperitoneal injections of Nembutal (30-35 mg/kg), supplemented by local application of 2% Xylocaine. The rat was mounted in a stereotaxic apparatus and a hole of 5 mm dia. was drilled in the skull.
‘Present address: Child Study Center, Yale University, New Haven, CT 06510. ‘Requests for reprints should be sent to: H. M. Sinnamon, Laboratory of Neuropsychology,
Copyright o 1978 ANKHO International
Attention
Wesleyan University, Middletown, CT 06457.
Inc.-0361-9230/78/050533-07$01.20/O
534
Bacymycin, an antibiotic ointment, was applied to the dura which was left intact. The nylon well with its base sealed with a layer of Partilm was placed over the opening and secured with acrylic cement. Five stainless steel screws distributed over the skull served to anchor the cement. An uninsulated copper wire, wrapped around the screws, served as the ground connection. Recording electrodes were also placed in the hippocampus and over the frontal cortex although the results of these recordings are not reported here. All electrodes were connected to Amphenol pins which were inserted into an Amphenol connector strip. The strip was also secured to the skull with acrylic cement and provided the means of attachment for the recording cable. A recovery period of 5-7 days was allowed.
The test chamber was a cylinder made of clear Plexiglas, 27.9 cm dia. and 27.9 cm high from the base, which was also made from Plexiglas. The cylinder was surrounded by a copper screen which served as shielding against 60 Hz interference. Opaque paper covered the lower 20 cm of the chamber and blocked the view of the subject. A trio of speakers (76 mm dia.) were placed 90” apart on the outside of the cylinder just above the base. The wall of the chamber had openings at the points where the speakers were attached. A flashhead of a Grass photo stimulator was mounted 0.6 m above the chamber. Also mounted above the chamber were a counterbalanced arm assembly with a nine-channel commutator, and a small convex mirror that allowed indirect observation of the animal’s behavior. The testing occurred in a dimly lit room with minimal ambient noise. Masking noise (approximately 60 db) produced by a white noise generator was amplified and delivered to the center speaker of the testing chamber. Two types of stimuli were presented to the rat. The first type was used in the standard tests for sensory responsiveness that preceded the tests for habituation. The standard visual stimulus was a grid made of white hardware cloth covering a black rectangle (10x 15 cm). The auditory stimulus was a click produced by a microswitch held about 10 cm from the rat’s head. Tactile stimuli were produced by probing the rat’s head and body with a pencil Stimuli used in tests for habituation were clicks and flashes controlled by solid-state logic circuitry. The frequency of the stimuli varied from 10 to 100 Hz and the train length was held constantly at approximately 1 sec. The clicks were delivered to one of the two opposing speakers of the chamber. The speaker selection was controlled manually and changed periodically to maintain the novelty of the stimulus. The loudness of the clicks was adjusted between approximately 75 and 88 db so that a locomotor movement or head turn, but not a startle or escape response, was produced. The flashes used in the tests for habituation were produced by a Grass PS 22 photostimulator. Intensity was set at a nominal level of 16.
Single neuronaf activity was recorded through tungsten microelectrodes that were insulated with glass up to 10 PM from their tips which were approximately of I J.LMdia. The electrode was lowered into the brain by means of a small microdrive fitted to the nylon well on the skuI1 of the rat. A short wire from the electrode led directly to a miniature pre-
KOOTZ AND SINNAMON amplifier mounted nearby on the end of a cable composed of Microdot wires. Additional details of the system may be found in previous reports [15,20]. The output of the amplifier was subjected to high frequency filtering and led to one channel to a Tektronix 502 oscilloscope for additional amplification. The output of the vertical amplifier of the oscilloscope was led to a spike discriminator which produced a 5 V pulse when spikes of a selected size and duration occurred. The disc~minator output was sent to a solid-state counter which printed accumulated counts at intervals controlled by the stimulus control circuitry. The discriminator output was also integrated over 1 set periods and displayed on one channel of a Beckman R411 polygraph. The accuracy of spike disc~mination was continuously monitored by one experimenter, a precaution necessitated by the changes in unit amplitude that sometimes occurred with tests lasting several minutes. The unit activity was also led to one channel of a Sony tape recorder. Movement of the animal in response to the stimuli presented during the habituation tests was measured in two ways. The simpler and generalIy more effective method was direct observation of the rat by one experimenter during the stimulus and for 3 set thereafter. This measure was dichotomous, indicating only whether a behavioral response occurred or not, and did not reflect the intensity, latency or duration of the response. In order to obtain quantitative measures of the behavioral response, two unte~inated wires were attached to the recording cable and led to the differential inputs of a Grass P15 preamplifier. Electrical disturbances in these wires caused by the rats’ movements were displayed on one channel of the polygraph along with stimulus markers and the integrated unit activity. The amplitude of pen deflection on the movement detector failed to relate systematically to the intensity of behavior as judged by direct observation but the duration of the deflection often did correlate with the time course of the response observed. Therefore, the movement record was measured in terms of the time over which the tracing was above a threshold level. However, a dficulty was encountered due to individual differences between rats in terms of the topography of behavior elicited. For some rats, the movement detector was insensitive to the behavior directly observed to be elicited by the stimuli. Adjustment of the movement detector sensitivity before the habituation tests was precluded by the need to preserve the general responsiveness of rat to the test stimuli. Thus the principal behavioral measure was the qualitative index, which was supplemented in some cases by the quantitative movement detector.
Prior to testing, the rats were lightly anesthetized with halothane (0.50/c, 1 liter per min), and the microdrive and recording cable were attached. The microelectrode was lowered into the cortex dorsal to the MRF in order to assess the recording capabilities of the electrode. If well isolated, units with peak-to-peak amplitudes greater than 300 PV were not found within 2 mm, the electrode was replaced and another penetration was made in a new location. Once adequate recording was established, the anesthesia was discontinued. A period of 45 min was allowed for recovery from the anesthesia and for adaptation to the chamber. Testing for habituation and sensitization began only after a single unit was isolated for at least 5 min without appreciable changes in amplitude or signs of injury (e.g., notched
MRF ACTIVITY
535
AND HABITUATION
waveforms). Prior to the main tests, the sensory responsiveness of the unit was assessed by presenting the visual, auditory and tactile stimuli several times. Generally, the microswitch clicks were presented first at various points in a IO cm radius around the rat’s head. Next, the moving visual stimulus was presented to nasal and temporal visual fields on both sides. Finally, tactile sensitivity was tested by probing the head, vibrissa, neck, back and lateral aspects of the limbs. Habituation, of course, was a potential problem in these tests since several units were tested on each rat, but care was taken to assure that the rat showed at least a behavioral response to these stimuli. A unit was considered responsive if the two experimenters agreed that a clear increase or decrease attended each stimulus presentation. After sensory testing the more formal habituation tests began. First a series of control trials were given in which all of the logic and analysis apparatus operated without the presentation of test stimuli. Pilot work had shown that occasionally the rats reacted to noises made by coincidental equipment, e.g., the printer, and the control trials habituated these responses. Next, the series of click or flash stimuli were presented. In each series the stimulus was kept constant in frequency. To apply the criteria for habituation or sensitization of behavior and unit activity, a minimum of six trials was required. Generally about four series of stimuli were presented with each unit. Not all series resulted in usable data. Most often the series were terminated before the required six trials when it was apparent that neither a behavioral or unit activity response was evoked. The loss of isolation of the single unit was another principle reason for the discounting of a series. The stimuli were changed as needed in frequency, speaker duration and type (flash or click) in order to revive the behavioral responses. No one order of stimulus presentation was given to all animals. Occasionally the clicks and flashes were paired on one trial after a habituation occurred as a test for dishabituation. Trials were begun only when the rat was still. Consequently, the interval between trials was not constant; generally it was between 5 and 10 sec. Occasionally a movement occurred between the beginning of the 1 set prestimulus period and the stimulus. When this movement was large enough to obscure any response to the stimulus the trial was not used.
For series of stimulus presentations with up to 16 trials, the first three trials were compared to each successive sets of three subsequent trials. For series with 17 to 24 trials, sets of four trials were compared. The same analysis was applied to the counts of single unit activity in the four 1-set bins of a trial. RESULTS
Generai Characteristics
of MRF Units
The units were characterized in terms of location, responsiveness during the standard sensory tests, baseline firing rate, and general responsiveness to the stimuli presented during habituation tests. Locations of all units recorded, including those in the MRF, are shown in Fig. 1. The 24 units classed as reticular are shown lateral to the central gray and ventral and medial to the dashed lines. This area generally conforms to the nucleus cuneiformis as described by Valverde 1211. Another 12 units outside of the MRF were recorded and tested. These units, located in the area of mergence between the MRF and tectum (N=4), the medial geniculate nucleus (N=2), the hippocampal formation (N=2) and the neocortex (N=4), were a diverse group which served for comparison to the MRF units.
Histology A small electrolytic lesion (100 PA, 20 set) was made at the deepest point of the electrode penetration. The rat was anesthetized with Nembutal and perfused through the heart with physiological saline followed by 10% Formalin. The brains were cut in transverse sections (16 pm) with a microtome cryostat and stained with cresyl violet. Locations of units were dete~ined relative to the marker lesion. Data Analysis Classifications of behavioral habituation and sensitization were based primarily on the dichotomous measure of movement as directly observed. Habituation was delcared if: (a) a movement occurred on two of the initial three trials of a series, and (b) no movement occurred on two of any three consecutive and subsequent trials. The criterion for sensitization was the converse. The quantitative movement detector scores were averaged over four I-set periods beginning with the onset of stimulation and were tested for significant (~~0.05) increases or decreases by Mann Whitney U tests.
FIG. 1. Sensory responsiveness and locations of units tested projected onto plates adapted from the atlas of Kiinig and Klippel 1121. The stimuli were clicks, a moving visual stimulus and tactile stimulation of the head and body. The area enclosed with the dashed lines is the MRF.
The response of the single units to auditory, tactile and visual stimuli presented during the standard sensory tests that preceded tests for habituation are shown in Fig. I. All of the 23 MRF units on which data was available responded to at least one of the stimuli. By contrast, the units in the non-
KOOTZ AND SINNA~ON
536
MRF areas were significantly less responsive (Fisher Exact Test, p=O.O09) with one third failing to show a response during the preliminary sensory testing. The MRF units favored tactile stimulation, with over 80% (18/23) of the units responding. SignificantIy fewer b-0.09) of the units in the other areas responded to tactile stimulation. MRF units were afso more responsive to the auditory stimulus 0, =0.04), 16 of 23 responsive MRF units compared to four of 12 non-MRF units. The visual stimulus was the least effective for both the MRF units (eight of 23 responsive) and the non-MRF units (two of 12 responsive). The MRF units showed significantly more multimodal responses than units in other areas @ =0.008). These multimodal responses were distributed throughout the MRF regions studied and are shown as filledin symbols of Fig. 1. The mean baseline firing rates of the MRF units were determined from the I-see prestimulation periods in all stimulation trials in the tests for habituation. It should be noted that trials were begun only when the rat was immobile. The rates ranged from 0.04 to spikeslsec to 29.7 spikeslsec with a median of 2.4 spikes/set. One third of the units fired at rates of llsec or less. Tactile and auditory responses were found in units with a wide range of baseline rates but visual responses were confined, with one exception, to units with slower than average baseline rates (~0.5 spikes/see). No tendency was discerned for units with slow rates to congregate in any particular region of the MRF. The general responsiveness of each MRF unit to the stimuli presented during the tests for habituation and sensitization was assessed by paired 1 tests comparing the rates in all of the prestimulation periods to the rates in the periods with stimulation. All stimulus conditions were combined so that a general index of responsivity was obtained rather than a specific measure of response to a particular stimulus. Nearly all MRF units (20 of 24) showed signi~cant responses @
