Brain Research Bulletin, Vol. 25, pp. 5543.
0361-9230/m $3.00 + .OO
0 Pergamon Press plc, 1990. Printed in the U.S.A.
Conditioning-Related Changes of Unit Activity in the Dorsal Lateral Geniculate Nucleus of Urethane-Anaesthetized Rats DORIS ALBRECHT, HELGA DAVIDOWA AND HANS-J. GABRIEL Institute of Physiology, Medical School (Charit&), Humboldt University Berlin Hessische Str. 3-4, Berlin-1040, GDR Received 3 August 1989
ALBRECHT, D., H. DAVIDOWA AND H.-J. GABRIEL. Conditioning-related changes of unif activiry in the dorsal lateral geniculate nucleus of urethane-anaesthetized rats. BRAIN RES BULL 25(l) 5.5-63, 1990. -Changes in geniculate unit activity of urethane-anaesthetized and freely moving rats were investigated during conditioning. The conditioned stimulus (CS) was a flash which was paired with an electrical stimulation of the tail as unconditioned stimulus (US). The discharge rates evoked by the CS during forward conditioning were significantly higher in responding units than those evoked by reversal of CS and US (backward conditioning) or by pseudoconditioning. Tail stimulation alone did not cause significant changes in the firing rate of most of the neurons. In 25% of the investigated neurons the facilitation of activity evoked by forward conditioning persisted during an extinction period of more than 15 min. The effect of conditioning on neuronal activity appeared to be comparable in urethane-anaesthetized rats and in freely moving ones which responded to the US with a slight freezing behavior. Conditioning
dLGN
Unit activity
Freely moving and urethane-anaesthetized
IT has been shown that neurons of the dorsal lateral geniculate nucleus (dLGN) change their responses to light during condition. . mg, i.e., by pairing light with a reward or with aversive stimuli, a modification of CS-evoked discharge can be obtained as a function of associative training (3, 6, 7, 11, 18, 24). In this way the gain of signal transmission in the visual system is adjusted to the information needs of the organism (28,29). The probability of occurrence of changes at the retinal level is rather small. Wild and Cohen (30) did not observe changes of the retinal output during visual learning. Furthermore, avoidance conditioning is not only possible in awake animals, but it has been reported to occur under urethane anaesthesia (12). Unit activity changes in relation to conditioning have also been observed under urethane, in the frontal cortex for example (23,27). Recently we have shown that urethane alters the responses to light of single units in the dLGN (1). Latencies of the primary responses are prolonged and the fiing frequency is reduced. Now the question arises whether, in spite of the changed transmission of signals in the dLGN, conditioning-related activity changes can be elaborated in geniculate neurons of urethane-anaesthetized rats. Further questions that arise are: How far are the neuronal conditioned responses to light under urethane comparable to those observed in awake and freely moving rats? Do response changes persist during an extinction period? Can conditioning-related changes in geniculate unit responses be regarded as resulting from
associative
rats
mechanisms,
or
can
similar
changes
be
observed
during reversal of CS and US or pseudoconditioning? METHOD
The experiments were performed on adult male albino rats in two series: 1) acute, and 2) combined chronic and acute ones. All animals were housed under similar conditions (in cages of six; for chronic experiments individually 3 days between surgery and recording, respectively) and were provided with food and water ad lib. Prior to preparation and recording, the reactivity of each rat to electrical tail stimulation (US) was tested. The stimulus intensity of the electrical tail stimulation was adjusted to provoke a slight freezing behavior. Then in the acute series, each rat was anaesthetized with a single dose of urethane (1.2 g/kg, IP). Additional doses (10% of the initial dose) were given when necessary. In order to ensure a fairly constant level of anaesthesia across animals, the following criteria were used: (a) no vibrissae movement; (b) no reflexive withdrawal to a fii tail pinch; (c) a firm tail pinch elicited a short desynchronization of the EEG. The head of the animal was fixed in a stereotaxic instrument. Two small holes (approximately 1 mm in diameter) were drilled into the skull. The dura mater was carefully removed. The electrode was positioned 5.0 mm anterior to lambda and 3.5-4.0 mm lateral to the midline (15), and lowered approximately 4.0 mm
55
ALBRECHT, DAVIDOWA AND GABRIEL
BIG. 1. Example of iontophoretic marking of recording sites with Trypan blue solution
below the cortical surface, until the dLGN was recognized by cellular responses to a light beam moved across the visual field. Extracellular recordings were performed with glass microelectrodes (tip diameter 1 p,m, 15-35 Ma resistance) filled with saturated Trypan blue solution. Rectal temperature was maintained at 37-38°C. In the second series, the rats were prepared under ether
anaesthesia. Recording electrodes (insulated NiCr-wires, 50 pm, assembled in bundles of 5) were implanted under electrophysiological control into the left or right dLGN. On the same side stainless steel electrodes were implanted over the visual cortex (VC) for EEG recording. In addition, a holder for a light emitting diode (LED, Fairchild Semiconductor EVL 352, 7 mcd, 560 MI) was attached to the head. The recording sessions were started after
TABLE 1 RESULTS OF COMPARISON BETWEEN PAIRS OF RESPONSE PARAMETERS ON- AND OFF-LIKE UNITS (WILCOXON TEST)
On-Like (N=78)
Cell Type Experimental Conditions Compared parameters Primary latency Number of impulses during first excitation first inhibition Number of impulses during postexcitatory inhibition postexcitatory excitation Number of impulses during second excitation second inhibition Integrated activity Maintained activity
Off-Like (N = 53)
A
B
C
D
A
-
-
J
J
----
tt
-
t T
t
-
-
tt
t t
t
OF
B
C
t
-
tl-
-
-
T
tt
-
D
tt
t t -
t tt
-
Denotations concerning the experimental conditions: (A) flash before/flash during BC, (B) flash before/flash a&r BC, (C) flash before/flash during FC, (D) flash before/flash after FC (extinction); -no sign&ant changes, one arrow: significant changes at the 0.05 level, two arrows: ~~0.01, t incmase of activity from the fast to the second condition, 4 decrease of latency, N: number of neurons.
CONDITIONING-RELATED CHANGES IN GENICULATE UNITS
maintained
integrotcd
activity
2. 90 light stimuli of 15 msec duration (flash); 3. /(5) Backward conditioning (BC): in 90 trials tail stimulation was followed by flash with an interstimulus interval (ISI) of 700 or 900 msec (US-flash), in some recordings CS and US were presented randomly (pseudoconditioning); 4. 90 flashes; 5. /(3) Forward conditioning (FC): at least in 90 trials flash was followed by tail stimulation (flash-US), IS1 as above; Additionally, in some experiments a block with shortened IS1 (400 or 500 msec) was interposed. 6. Extinction period: at least 90 trials in which flashes were given alone; 7. 30 trials tail stimulation.
1
i
activity
1 flash 2 US - ftush
3 flash-l.6 4 Nash [ext.) ms
I
4
peak ducharge rate
I
I3 on-tikc
(N- 491
El off-like
(N =171
2
primary
51
3
c
lotcncy
FIG.
2. Averaged impulse rates (imp./sec.; mean-+S.D.) of maintained and integrated activity, peak discharge rates of the first excitation or postinhibitory excitation, respectively, and primary response latencies for on- and off-like cells of dLGN to flash, backward and fonvard conditioning and extinction.
a recovery of about 3 days.
Conventional recording techniques were used as described previously (3). Unit activity was displayed on a storage oscilloscope and recorded on tape. Recordings from neuronal somata were differentiated from fiber recordings by conventional criteria (5). Experiments were carried out in a dimly illuminated room (back~und luminance about 0.01 cd/m’). The LED was placed within a distance of 5 mm to the eye contralaterally to the recording side. After an adaptation period of at least 30 min to the dark and after isolation of one unit, the following schedule with an intertrial interval of about 10 set was applied: 1. 30 light stimuli of 500 msec duration to characterize on- and off-responses of the neuron;
Forward and backward pairings were alternated in their sequence in about half of the recordings. Due to the mean intertrial interval of about 10 set, one block of 90 trials lasted 15 min, which means that the whole recording procedure of one unit lasted as much as 90 min. Some units could be studied for periods of up to 5 hr, but some were lost before finishing the schedule. Therefore, the cell numbers vary in the results. In general, in each rat only one cell was tested in order to prevent influences of preceding conditioning procedure on the investigated unit activity. In seven experiments the activity of two cells was simultaneously recorded and separated by means of a window discriminator. In the second series the same procedure was carried out in freely moving animals, and was repeated one or two days later under anaesthesia. Only those recordings were included in the results in which the stability of the potentials before and after anaesthesia could be secured by supe~sition (see spike examples in Figs. 3 and 4). In most of the recordings tail stimulation caused a short desynchronization of the EEG activity of VC in freely moving as well as in urethane-anaesthetized rats. After the completion of the experiments iontophoretic marking (series 1) (Fig. 1) or electrolytical coagulation (series 2) were performed to confirm the electrode position histologically. Pe~stimulus activity was analyzed off--line by compiling histograms (PSTH) of 5-msec bin width from 30 trials. In such a way three PSTH resulted from one block of 90 stimuli. The fist 300 msec of the PSTH preceding the light stimulation served to assess the maintained activity. The statistical significance of responses was assessed by Chi2-test with Yates-correction. By means of this test the relative impulse frequency of every bin (number of impulses/number of single trials) was compared with the mean of m~ntained activity. Examples for the resulting signi~cance levels @
0361-9230/m $3.00 + .OO
0 Pergamon Press plc, 1990. Printed in the U.S.A.
Conditioning-Related Changes of Unit Activity in the Dorsal Lateral Geniculate Nucleus of Urethane-Anaesthetized Rats DORIS ALBRECHT, HELGA DAVIDOWA AND HANS-J. GABRIEL Institute of Physiology, Medical School (Charit&), Humboldt University Berlin Hessische Str. 3-4, Berlin-1040, GDR Received 3 August 1989
ALBRECHT, D., H. DAVIDOWA AND H.-J. GABRIEL. Conditioning-related changes of unif activiry in the dorsal lateral geniculate nucleus of urethane-anaesthetized rats. BRAIN RES BULL 25(l) 5.5-63, 1990. -Changes in geniculate unit activity of urethane-anaesthetized and freely moving rats were investigated during conditioning. The conditioned stimulus (CS) was a flash which was paired with an electrical stimulation of the tail as unconditioned stimulus (US). The discharge rates evoked by the CS during forward conditioning were significantly higher in responding units than those evoked by reversal of CS and US (backward conditioning) or by pseudoconditioning. Tail stimulation alone did not cause significant changes in the firing rate of most of the neurons. In 25% of the investigated neurons the facilitation of activity evoked by forward conditioning persisted during an extinction period of more than 15 min. The effect of conditioning on neuronal activity appeared to be comparable in urethane-anaesthetized rats and in freely moving ones which responded to the US with a slight freezing behavior. Conditioning
dLGN
Unit activity
Freely moving and urethane-anaesthetized
IT has been shown that neurons of the dorsal lateral geniculate nucleus (dLGN) change their responses to light during condition. . mg, i.e., by pairing light with a reward or with aversive stimuli, a modification of CS-evoked discharge can be obtained as a function of associative training (3, 6, 7, 11, 18, 24). In this way the gain of signal transmission in the visual system is adjusted to the information needs of the organism (28,29). The probability of occurrence of changes at the retinal level is rather small. Wild and Cohen (30) did not observe changes of the retinal output during visual learning. Furthermore, avoidance conditioning is not only possible in awake animals, but it has been reported to occur under urethane anaesthesia (12). Unit activity changes in relation to conditioning have also been observed under urethane, in the frontal cortex for example (23,27). Recently we have shown that urethane alters the responses to light of single units in the dLGN (1). Latencies of the primary responses are prolonged and the fiing frequency is reduced. Now the question arises whether, in spite of the changed transmission of signals in the dLGN, conditioning-related activity changes can be elaborated in geniculate neurons of urethane-anaesthetized rats. Further questions that arise are: How far are the neuronal conditioned responses to light under urethane comparable to those observed in awake and freely moving rats? Do response changes persist during an extinction period? Can conditioning-related changes in geniculate unit responses be regarded as resulting from
associative
rats
mechanisms,
or
can
similar
changes
be
observed
during reversal of CS and US or pseudoconditioning? METHOD
The experiments were performed on adult male albino rats in two series: 1) acute, and 2) combined chronic and acute ones. All animals were housed under similar conditions (in cages of six; for chronic experiments individually 3 days between surgery and recording, respectively) and were provided with food and water ad lib. Prior to preparation and recording, the reactivity of each rat to electrical tail stimulation (US) was tested. The stimulus intensity of the electrical tail stimulation was adjusted to provoke a slight freezing behavior. Then in the acute series, each rat was anaesthetized with a single dose of urethane (1.2 g/kg, IP). Additional doses (10% of the initial dose) were given when necessary. In order to ensure a fairly constant level of anaesthesia across animals, the following criteria were used: (a) no vibrissae movement; (b) no reflexive withdrawal to a fii tail pinch; (c) a firm tail pinch elicited a short desynchronization of the EEG. The head of the animal was fixed in a stereotaxic instrument. Two small holes (approximately 1 mm in diameter) were drilled into the skull. The dura mater was carefully removed. The electrode was positioned 5.0 mm anterior to lambda and 3.5-4.0 mm lateral to the midline (15), and lowered approximately 4.0 mm
55
ALBRECHT, DAVIDOWA AND GABRIEL
BIG. 1. Example of iontophoretic marking of recording sites with Trypan blue solution
below the cortical surface, until the dLGN was recognized by cellular responses to a light beam moved across the visual field. Extracellular recordings were performed with glass microelectrodes (tip diameter 1 p,m, 15-35 Ma resistance) filled with saturated Trypan blue solution. Rectal temperature was maintained at 37-38°C. In the second series, the rats were prepared under ether
anaesthesia. Recording electrodes (insulated NiCr-wires, 50 pm, assembled in bundles of 5) were implanted under electrophysiological control into the left or right dLGN. On the same side stainless steel electrodes were implanted over the visual cortex (VC) for EEG recording. In addition, a holder for a light emitting diode (LED, Fairchild Semiconductor EVL 352, 7 mcd, 560 MI) was attached to the head. The recording sessions were started after
TABLE 1 RESULTS OF COMPARISON BETWEEN PAIRS OF RESPONSE PARAMETERS ON- AND OFF-LIKE UNITS (WILCOXON TEST)
On-Like (N=78)
Cell Type Experimental Conditions Compared parameters Primary latency Number of impulses during first excitation first inhibition Number of impulses during postexcitatory inhibition postexcitatory excitation Number of impulses during second excitation second inhibition Integrated activity Maintained activity
Off-Like (N = 53)
A
B
C
D
A
-
-
J
J
----
tt
-
t T
t
-
-
tt
t t
t
OF
B
C
t
-
tl-
-
-
T
tt
-
D
tt
t t -
t tt
-
Denotations concerning the experimental conditions: (A) flash before/flash during BC, (B) flash before/flash a&r BC, (C) flash before/flash during FC, (D) flash before/flash after FC (extinction); -no sign&ant changes, one arrow: significant changes at the 0.05 level, two arrows: ~~0.01, t incmase of activity from the fast to the second condition, 4 decrease of latency, N: number of neurons.
CONDITIONING-RELATED CHANGES IN GENICULATE UNITS
maintained
integrotcd
activity
2. 90 light stimuli of 15 msec duration (flash); 3. /(5) Backward conditioning (BC): in 90 trials tail stimulation was followed by flash with an interstimulus interval (ISI) of 700 or 900 msec (US-flash), in some recordings CS and US were presented randomly (pseudoconditioning); 4. 90 flashes; 5. /(3) Forward conditioning (FC): at least in 90 trials flash was followed by tail stimulation (flash-US), IS1 as above; Additionally, in some experiments a block with shortened IS1 (400 or 500 msec) was interposed. 6. Extinction period: at least 90 trials in which flashes were given alone; 7. 30 trials tail stimulation.
1
i
activity
1 flash 2 US - ftush
3 flash-l.6 4 Nash [ext.) ms
I
4
peak ducharge rate
I
I3 on-tikc
(N- 491
El off-like
(N =171
2
primary
51
3
c
lotcncy
FIG.
2. Averaged impulse rates (imp./sec.; mean-+S.D.) of maintained and integrated activity, peak discharge rates of the first excitation or postinhibitory excitation, respectively, and primary response latencies for on- and off-like cells of dLGN to flash, backward and fonvard conditioning and extinction.
a recovery of about 3 days.
Conventional recording techniques were used as described previously (3). Unit activity was displayed on a storage oscilloscope and recorded on tape. Recordings from neuronal somata were differentiated from fiber recordings by conventional criteria (5). Experiments were carried out in a dimly illuminated room (back~und luminance about 0.01 cd/m’). The LED was placed within a distance of 5 mm to the eye contralaterally to the recording side. After an adaptation period of at least 30 min to the dark and after isolation of one unit, the following schedule with an intertrial interval of about 10 set was applied: 1. 30 light stimuli of 500 msec duration to characterize on- and off-responses of the neuron;
Forward and backward pairings were alternated in their sequence in about half of the recordings. Due to the mean intertrial interval of about 10 set, one block of 90 trials lasted 15 min, which means that the whole recording procedure of one unit lasted as much as 90 min. Some units could be studied for periods of up to 5 hr, but some were lost before finishing the schedule. Therefore, the cell numbers vary in the results. In general, in each rat only one cell was tested in order to prevent influences of preceding conditioning procedure on the investigated unit activity. In seven experiments the activity of two cells was simultaneously recorded and separated by means of a window discriminator. In the second series the same procedure was carried out in freely moving animals, and was repeated one or two days later under anaesthesia. Only those recordings were included in the results in which the stability of the potentials before and after anaesthesia could be secured by supe~sition (see spike examples in Figs. 3 and 4). In most of the recordings tail stimulation caused a short desynchronization of the EEG activity of VC in freely moving as well as in urethane-anaesthetized rats. After the completion of the experiments iontophoretic marking (series 1) (Fig. 1) or electrolytical coagulation (series 2) were performed to confirm the electrode position histologically. Pe~stimulus activity was analyzed off--line by compiling histograms (PSTH) of 5-msec bin width from 30 trials. In such a way three PSTH resulted from one block of 90 stimuli. The fist 300 msec of the PSTH preceding the light stimulation served to assess the maintained activity. The statistical significance of responses was assessed by Chi2-test with Yates-correction. By means of this test the relative impulse frequency of every bin (number of impulses/number of single trials) was compared with the mean of m~ntained activity. Examples for the resulting signi~cance levels @
