Exp Brain Res (1992) 88:229-232
Experimental BrainResearch © Springer-Verlag 1992
Research Note
Simultaneous recording of lagged and nonlagged cells in the cat dorsal lateral geniculate nucleus E. Hartveit Institute of Basic Medical Sciences, Department of Neurophysiology, University of Oslo, POB 1104 Blindern, N-0317 Oslo 3, Norway Received June 7, 1991 / Accepted September 23, 1991
Summary. It has been suggested that lagged and nonlagged cells in the cat dorsal lateral geniculate nucleus (LGN) represent state-dependent response modes of the same class of L G N cells. In two separate experiments with single-unit recording in the L G N of anaesthetized and paralysed adult cats, a lagged and a nonlagged X-cell were recorded simultaneously with the same microelectrode. For each pair of cells, the amplitude of the action potentials was sufficiently different to allow separate compilation of peri-stimulus-time-histograms. For all 4 cells, the visual response pattern to a stationary flashing light spot was typical of their respective cell class. These findings support the hypothesis that lagged and nonlagged cells are separate cell classes and indicate that the population of L G N cells do not appear as lagged during one state of modulatory input and as nontagged during another. Key words: Dorsal lateral geniculate nucleus - Lagged cells - Nonlagged cells - Sensory processing - Cat
Introduction In the A-laminae of the cat lateral geniculate nucleus (LGN), the X relay cells consist of lagged and nonlagged cells (Mastronarde 1987a, b). When stimulated with a stationary flashing light spot, the lagged cells differ from the classical nonlagged geniculate cells primarily by their longer latency visual responses and by an initial suppression before the response rises to a maintained level. Simultaneous recordings from a single lagged cell and the retinal ganglion cell mediating its excitatory input, indicated that the differentiation of lagged and nonlagged cells occurs in the L G N (Mastronarde 1987b). There is evidence that the lagged X-cells are morphologically different from the nonlagged X-celts (Humphrey and Weller 1988b). There is also evidence for the existence of lagged Y-cells (Heggelund and Hartveit 1990; Saul and Humphrey 1990; Mastronarde et al. 1991).
Recently it was found that the lagged and nonlagged cells also differ pharmacologically. The excitatory input to the lagged cells is strongly dependent on glutamate receptors of the N-methyl-D-aspartate (NMDA) type (Heggelund and Hartveit 1990) whereas that to the nonlagged cells is more dependent on glutamate receptors of the quisqualate/kainate (non-NMDA) type (Hartveit and Heggelund 1990). There is evidence that the initial suppression and delay of the response of the lagged cells is caused by inhibitory input mediated by 7-aminobutyric acid (GABA). Ionophoresis of GABA antagonists reduces the initial suppression, and the latency of the visual response and the visual response pattern of the lagged cells become more similar to that of nonlagged cells and retinal ganglion cells (Heggelund and Hartveit 1990). Thus, an eventual physiologically occurring strong reduction in the action of the geniculate inhibitory circuitry might change the visual response properties of the lagged cells and make them more similar to their retinal input and the nonlagged cells. It is also possible that a strong facilitation of the geniculate relay cells could have a similar effect. An important substrate for modulatory input to the L G N is the peribrachial region (PBR) of the brain stem (reviewed in Steriade and McCarley 1990). The PBR contains cholinergic, noradrenergic and serotonergic neurones with ascending inputs to the LGN, but the cholinergic input is the most extensive (Fitzpatrick et al. 1989). Electrical stimulation of the PBR has a facilitatory effect on the visual response of L G N cells (Francesconi et al. 1988). There is evidence from in vitro experiments that acetylcholine (ACh) has a direct effect both on the relay cells (McCormick and Prince 1987) and on the inhibitory interneurones (McCormick and Pape 1988) in the LGN. The existence of the modulatory brain stem input to the L G N raises the question whether the differences in visual response properties between lagged and nonlagged cells are dependent on the state of the modulatory input (Humphrey and Weller 1988a; Uhlrich et al. 1990). In the experiments reported here two pairs of X-cells, each
230 consisting o f one lagged and one nonlagged cell, were recorded simultaneously with the same microelectrode. This demonstrates that b o t h cell classes can occur in the same state o f the animal. Thus, the cells in the L G N do n o t simply a p p e a r as lagged in one state and as n o n lagged in another.
~-
0
,
,~""
'
' ~"
"h,,
1.0
Methods Adult cats (2.4 and 3.0 kg) were prepared acutely for single unit recording according to methods described previously (Heggelund and Hartveit 1990). During the recording period, the cats were anaesthetized with halothane (0.3-0.8%) in a mixture of N20 and 02 (70: 30). In one experiment, the cat also received small intravenous injections of pentobarbitone sodium. Single cells in the A-laminae of the LGN were recorded with glass pipettes filled with 0.2 M NaC1 and 0.05 M 2-amino-2(hydroxymethyl)-l,3-propanediol (tris; pH 7.4, tip diameter 1.25-1.50 gin, impedances of 30--60 Mf~ in vivo). The LGN was approached through a craniotomy at Horsley-Clarke stereotaxic coordinates anterior 6 mm, lateral 15.6 mm, and with the electrode angled 32 deg from the vertical. The entry of the electrode into lamina A was detected by the shift in response properties from those typical of perigeniculate cells to those typical of LGN cells. As the penetrations were approximately along visual projection lines in the LGN, the location of the receptive fields was relatively stationary within each of the A4aminae. The borders between lamina A and A1, and between A1 and C, were detected by the concomitant shift in ocular dominance. After isolation of a single cell, its receptivefield centre was plotted on a tangent screen 86 cm in front of the cat's eyes with the use of stationary and moving light and dark spots, as well as grating stimuli. The quantitative tests were made with a small, stationary flashing light spot. The size of the spot was adjusted so that it approximately filled the receptive-field centre of the cell. The spot presentation was made by a computer-controlled light projector system. The computer also generated peri-stimulustime-histograms (PSTHs) on-line. The position of the receptive-field centres on the tangent screen was checked regularly to control for eye movements. The neuronal activity could also be recorded on a tape recorder. Selected epochs of the signals were digitized off-line at 10 kHz. The lagged cells were identified by their characteristic temporal response properties, and primarily by their long latency to the onset and to half-rise and half-fall of the visual response (Mastronarde 1987a). The luminance of the spot was ~0.9 log units above the background luminance (~ 7 cd/m2). The cells were classified into X (brisk sustained) or Y (brisk transient) as described in Heggelund and Hartveit (1990). The experiment was terminated when the mean blood pressure fell below 90 mm Hg. The animal was then deeply anaesthetized with pentobarbitone sodium (50 mg/kg i.v.) and perfused transcardially with saline and 4% formaldehyde in saline. A block containing the LGN was cut at 25 gm on a freezing microtome and Nissl stained with cresyl violet. Tracks could be retrieved because the electrodes had been advanced for a few mm after finishing the penetration through the A-laminae.
Results The two pairs o f lagged/nonlagged cells were recorded in two different experiments. T h e first pair was recorded in lamina A. The lagged cell o f this pair was an O N - c e n t r e X-cell. Its receptive field was located a b o u t 25 deg f r o m the area centralis a n d the m e a n diameter o f its centre was 0.6 deg. The visual response pattern was typical o f a lagged X-cell (Fig. 1A).
o~
c
250
B
C
0
m
,~
1.0 Time (s)
Fig. IA, B. Peri-stimulus-time histograms that show the response of 2 ON-centre cells to a stationary flashing spot of light. Visual latencies refer to the histograms shown here. A Lagged X-cell. Latency to onset 78 ms, latency to half-rise 118 ms, latency to half-fall 53 ms. B Nonlagged X-cell. Latency to onset 43 ms, latency to half-rise 58 ms, latency to half-fall 33 ms. Each histogram shows the response of the cell during a 500 ms time window when the light was on, marked by the horizontal line under the histogram, and during an equally long time window after the light was turned off. Histograms were compiled from 16 (A) and 20 (B) stimulus presentations, respectively. Binwidth 5 ms
The cell's average latency to onset o f the visual response was 83 ms. Its average half-rise and half-fall latencies were 135 and 55 ms, respectively. W i t h o u t c h a n g i n g the position o f the electrode, the activity o f this cell was recorded in isolation for a b o u t 50 min. D u r i n g this period, there was minimal change in the shape o f the action potentials a n d at no time did the cell change its response pattern or visual latencies to those o f nonlagged cells. T h e n the action potentials o f a second unit a p p e a r e d a n d gradually b e c a m e larger until their amplitude was a b o u t 2 times that o f the first unit. F o r the second unit, slow potentials (S-potentials; Bishop et al. 1962) could be recorded in addition to the action potentials. The m e a n diameter o f its receptive-field centre was 0.6 deg. It was an O N - c e n t r e X-cell and had a visual response pattern typical o f a nonlagged cell (Fig. 1B). The average latency to onset o f the visual response was 45 ms and the average half-rise and half-fall latencies were 55 and 33 ms, respectively. The separation between the two receptive-field centres was 0.5 deg. The amplitudes o f the respective action potentials were sufficiently different to allow separate recording o f a P S T H for one unit in isolation f r o m the other. The two units could be recorded together for a b o u t 10 min. T h e n the action potentials o f the lagged cell suddenly disappeared, a n d at the same time the S-potentials and action potentials o f the nonlagged cell b e c a m e larger. The second pair o f cells was recorded in lamina A1. The n o n l a g g e d cell h a d large positive-negative action potentials a n d S-potentials could be recorded in addition to the action potentials (Figs. 2E, F). It was an O N centre X-cell with receptive field located a b o u t 37 deg f r o m area centralis. T h e m e a n diameter o f its centre was 1.1 deg. The visual response pattern was typical o f a
231
A
200
200
C
~9
c~ 0
....
1.0
1.0
C~
100
100
C
'
t
°
1.0 Time (s)
1.0 Time (s)
50 ms
F
50 ms
Fig. 2. A, B Peri-stimulus-time histograms that show the response of 2 ON-centre cells to a stationary flashing spot of light. Visual latencies refer to the histograms shown here. A Nonlagged X-cell. Latency to onset 48 ms, latency to half-rise 63 ms, latency to half-fall 43 ms. B Lagged X-cell. Latency to onset 78 ms, latency to half-rise 93 ms, latency to half-fall 73 ms. Each histogram shows the response of the cell during a 500 ms time window when the light was on, marked by the horizontal line under the histogram, and during an equally long time window after the light was turned off. Histograms were compiled from 15 (A) and 50 (B) stimulus presentations, respectively. Binwidth 5 ms. C, D Peristimulus-timehistograms that show the responses of the same cells as in A and B to a single light spot that covered part of the receptive-field centre
of each cell. C Nonlagged X-cell. Latency to onset 48 ms, latency to half-rise 63 ms, latency to half-fall 33 ms. D Lagged X-cell. Latency to onset 93 Ins, latency to half-rise 103 ms, latency to half-fall 78 ms. Time window for stimulus presentation 500 ms. Histograms were compiled from 20 (C) and 70 (D) stimulus presentations, respectively. Binwidth 5 ms. E, F Plots of digitized traces. The large action potentials and the S-potentials belong to the nonlagged X-cell and the smaller action potentials belong to the lagged X-cell. Horizontal line below the trace marks the duration of the light-ON period. E The trace corresponds to the histogram in B. Notice that the nonlagged cell responded weakly to the stimulus. F The trace corresponds to the histograms in C, D
n o n l a g g e d cell (Fig. 2A). T h e cell's average latency to onset o f the visual response was 48 ms. Its average latency to half-rise was 63 ms and to half-fall it was 43 ms. T h e activity o f this cell was recorded for a b o u t 25 min u n d e r stable conditions. A t this time, the action potentials o f a second unit b e c a m e gradually larger until their amplitude was a b o u t 1/2 o f the first unit. The second unit had positive-negative action potentials (Figs. 2E, F). It t o o was an O N - c e n t r e X-cell, and it h a d a visual response pattern typical o f a lagged X-ceU (Fig. 2B). T h e m e a n diameter o f the receptive-field centre was 1.t deg. The average latency to onset o f the visual response was 83 ms and to half-rise and half-fall it was 100 a n d 75 ms, respectively. The separation between the two receptive-
field centres was 0.9 deg. Because the receptive-field centres were elongated a p p r o x i m a t e l y perpendicularly to a line connecting them, the degree o f overlap was relatively small. P S T H s for one unit could be compiled in isolation f r o m the other. F r o m plots o f digitized traces o f the n e u r o n a l activity, the S-potentials and action potentials o f the weakly responding nonlagged cell could be seen together with the action potentials o f the strongly res p o n d i n g lagged cell when the latter was appropriately stimulated in its receptive-field centre (Fig. 2E). T o ascertain that the cells could r e s p o n d with patterns typical o f lagged and n o n l a g g e d cells simultaneously, the size a n d location o f the light spot was adjusted so that it covered a p p r o x i m a t e l y h a l f o f b o t h receptive-field cen-
232 tres. Figures 2C, D show that even when the two cells received this suboptimal visual stimulus the visual response patterns and latencies were very similar to those in the separate control recordings. Figure 2F shows a digitized trace with both cells responding to the suboptimal stimulus. The amplitudes of the respective action potentials remained sufficiently different to allow compilation of PSTHs for about 1 1/2 h. During this time neither cell changed its response pattern to that of the other class. However, the mean firing rate could vary spontaneously for both cells. The amplitude o f the action potentials o f the lagged cell gradually increased, and after a while, the two cells could not be separated based on the amplitude of the action potentials. By moving the microelectrode 20 ~tm further down into lamina A1, the potentials of the nonlagged cell disappeared and the action potentials of the lagged cell remained in isolation. PSTHs compiled at this time were virtually identical to those compiled when the nonlagged cell was still present.
Discussion There is strong physiological (Mastronarde 1987a, b), morphological (Humphrey and Weller 1988b) and pharmacological (Hartveit and Heggelund 1990; Heggelund and Hartveit 1990) evidence for treating lagged and nonlagged cells as separate cell classes. The present finding that lagged and nonlagged cells can be recorded simultaneously supports this hypothesis. Uhlrich et al. (1990) recently examined the influence of electrical PBR stimulation on the visual response of lagged and nonlagged cells. They reported that PBR stimulation could reduce the latency to half-rise below 70 ms for 6 out of 8 lagged cells recorded. It was claimed that this had converted the visual responses o f lagged cells to those of nonlagged cells, and the authors put forward the hypothesis that the response properties characteristic of lagged and nonlagged cells may represent state-dependent response modes of the same L G N cells. One version o f this hypothesis might be that the cells in the L G N appear as lagged in one arousal state and as nonlagged in another state. Accordingly, it should not be possible to record lagged and nonlagged cells simultaneously in the L G N . The present result shows that this is indeed possible. This result is consistent with the experience that lagged and nonlagged cells can be recorded in succession in a single electrode penetration (Mastronarde 1987a; Heggelund and Hartveit 1990). An alternative version o f the hypothesis might be that different L G N cells are subject to different degrees o f influence of the modulatory input, such that they would switch between lagged and nonlagged response patterns at different levels of modulatory input. This interpretation cannot be rejected by the present results. However, there are no reports that a cell can change from a nonlagged to a lagged response pattern. In fact, the available evidence indicates that during reduced modulatory input the response pattern o f nonlagged cells becomes more transient (e.g. Sawai et al. 1988) and accordingly less
similar to that of lagged cells. Furthermore, some of the previously reported differences between lagged and nonlagged cells cannot be state-dependent. In particular, lagged and nonlagged cells have almost non-overlapping antidromic latencies to electrical stimulation o f the visual cortex (Mastronarde 1987a; Humphrey and Welter 1988a) and almost non-overlapping morphologies (Humphrey and Weller 1988b).
Acknowledgements. I thank Paul Heggelund for continuous support and encouragement and for valuable comments on the manuscript, Allen L. Humphrey for constructive criticism of an earlier version of the manuscript, Ingeborg Spinnangr for histological and technical assistance, Torsten Eken for help with digitizing and plotting of neuronal signals and Arve Stavo for writing the on-line computer programs. Gallamine triethiodide (Flaxedil) was kindly provided by Cyanamid Nordiska and halothane (Fluothane) by ICI-Pharma. The research was financially supported by The Norwegian Research Council for Science and The Humanities (NAVF). References Bishop PO, Burke W, Davis R (1962) The interpretation of the extracellular response of single lateral geniculate cells. J Physiol (Lond) 162:451-472 Fitzpatrick D, Diamond IT, Raczkowski D (1989) Cholinergic and monoaminergic innervation of the cat's thalamus: comparison of the lateral geniculate nucleus with other principal sensory nuclei. J Comp Neurol 288:647-675 Francesconi W, Mfiller CM, Singer W (1988) Cholinergic mechanisms in the reticular control of transmission in the cat lateral geniculate nucleus. J Neurophysiol 59:1690-1718 Hartveit E, Heggelund P (1990) Neurotransmitter receptors mediating excitatory input to cells in the cat lateral geniculate nucleus. II. Non-lagged cells. J Neurophysiol 63:1361-1372 Heggelund P, Hartveit E (1990) Neurotransmitter receptors mediating excitatory input to cells in the cat lateral geniculate nucleus. I. Lagged cells. J Neurophysiol 63:1347-1360 Humphrey AL, Weller RE (1988a) Functionally distinct groups of X-cells in the lateral geniculate nucleus of the cat. J Comp Neurol 268, 429447 Humphrey AL, Weller RE (1988b) Structural correlates of functionally distinct groups of X-cells in the lateral geniculate nucleus of the cat. J Comp Neurol 268: 448M68 Mastronarde DN (1987a) Two classes of single-input X-cells in cat lateral geniculate nucleus. I. Receptive field properties and classification of cells. J Neurophysiol 57:357-380 Mastronarde DN (t987b) Two classes of single-input X-cells in cat lateral geniculate nucleus. IL Retinal inputs and the generation of receptive field properties. J Neurophysiol 57: 381-4 13 Mastronarde DN, Humphrey AL, Saul AB (199t) Lagged Y cells in the cat lateral geniculate nucleus. Vis Neurosci 7:191-200 McCormick DA, Pape H-C (1988) Acetylcholine inhibits identified interneurons in the cat lateral geniculate nucleus. Nature 334:246--248 McCormick DA, Prince DA (1987) Actions of acetylcholine in the guinea-pig and cat medial and lateral geniculate nuclei, in vitro. J Physiol (Lond) 392:147-165 Saul AB, Humphrey AL (1990) Spatial and temporal properties of lagged and nonlagged cells in cat lateral geniculate nucleus. J Neurophysiol 64:206-224 Sawai H, Morigiwa K, Fukuda Y (1988) Effects of EEG synchronization on visual responses of the cat's geniculate relay cells: a comparison among Y, X and W cells. Brain Res 455: 394-400 Steriade M, McCarley RW (1990) Brainstem control of wakefulness and sleep. Plenum, New York London Uhlrich D J, Tamamaki N, Sherman SM (1990) Brainstem control of response modes in neurons of the cat's lateral geniculate nucleus. Proc Natl Acad Sci USA 87:2560-2563
Experimental BrainResearch © Springer-Verlag 1992
Research Note
Simultaneous recording of lagged and nonlagged cells in the cat dorsal lateral geniculate nucleus E. Hartveit Institute of Basic Medical Sciences, Department of Neurophysiology, University of Oslo, POB 1104 Blindern, N-0317 Oslo 3, Norway Received June 7, 1991 / Accepted September 23, 1991
Summary. It has been suggested that lagged and nonlagged cells in the cat dorsal lateral geniculate nucleus (LGN) represent state-dependent response modes of the same class of L G N cells. In two separate experiments with single-unit recording in the L G N of anaesthetized and paralysed adult cats, a lagged and a nonlagged X-cell were recorded simultaneously with the same microelectrode. For each pair of cells, the amplitude of the action potentials was sufficiently different to allow separate compilation of peri-stimulus-time-histograms. For all 4 cells, the visual response pattern to a stationary flashing light spot was typical of their respective cell class. These findings support the hypothesis that lagged and nonlagged cells are separate cell classes and indicate that the population of L G N cells do not appear as lagged during one state of modulatory input and as nontagged during another. Key words: Dorsal lateral geniculate nucleus - Lagged cells - Nonlagged cells - Sensory processing - Cat
Introduction In the A-laminae of the cat lateral geniculate nucleus (LGN), the X relay cells consist of lagged and nonlagged cells (Mastronarde 1987a, b). When stimulated with a stationary flashing light spot, the lagged cells differ from the classical nonlagged geniculate cells primarily by their longer latency visual responses and by an initial suppression before the response rises to a maintained level. Simultaneous recordings from a single lagged cell and the retinal ganglion cell mediating its excitatory input, indicated that the differentiation of lagged and nonlagged cells occurs in the L G N (Mastronarde 1987b). There is evidence that the lagged X-cells are morphologically different from the nonlagged X-celts (Humphrey and Weller 1988b). There is also evidence for the existence of lagged Y-cells (Heggelund and Hartveit 1990; Saul and Humphrey 1990; Mastronarde et al. 1991).
Recently it was found that the lagged and nonlagged cells also differ pharmacologically. The excitatory input to the lagged cells is strongly dependent on glutamate receptors of the N-methyl-D-aspartate (NMDA) type (Heggelund and Hartveit 1990) whereas that to the nonlagged cells is more dependent on glutamate receptors of the quisqualate/kainate (non-NMDA) type (Hartveit and Heggelund 1990). There is evidence that the initial suppression and delay of the response of the lagged cells is caused by inhibitory input mediated by 7-aminobutyric acid (GABA). Ionophoresis of GABA antagonists reduces the initial suppression, and the latency of the visual response and the visual response pattern of the lagged cells become more similar to that of nonlagged cells and retinal ganglion cells (Heggelund and Hartveit 1990). Thus, an eventual physiologically occurring strong reduction in the action of the geniculate inhibitory circuitry might change the visual response properties of the lagged cells and make them more similar to their retinal input and the nonlagged cells. It is also possible that a strong facilitation of the geniculate relay cells could have a similar effect. An important substrate for modulatory input to the L G N is the peribrachial region (PBR) of the brain stem (reviewed in Steriade and McCarley 1990). The PBR contains cholinergic, noradrenergic and serotonergic neurones with ascending inputs to the LGN, but the cholinergic input is the most extensive (Fitzpatrick et al. 1989). Electrical stimulation of the PBR has a facilitatory effect on the visual response of L G N cells (Francesconi et al. 1988). There is evidence from in vitro experiments that acetylcholine (ACh) has a direct effect both on the relay cells (McCormick and Prince 1987) and on the inhibitory interneurones (McCormick and Pape 1988) in the LGN. The existence of the modulatory brain stem input to the L G N raises the question whether the differences in visual response properties between lagged and nonlagged cells are dependent on the state of the modulatory input (Humphrey and Weller 1988a; Uhlrich et al. 1990). In the experiments reported here two pairs of X-cells, each
230 consisting o f one lagged and one nonlagged cell, were recorded simultaneously with the same microelectrode. This demonstrates that b o t h cell classes can occur in the same state o f the animal. Thus, the cells in the L G N do n o t simply a p p e a r as lagged in one state and as n o n lagged in another.
~-
0
,
,~""
'
' ~"
"h,,
1.0
Methods Adult cats (2.4 and 3.0 kg) were prepared acutely for single unit recording according to methods described previously (Heggelund and Hartveit 1990). During the recording period, the cats were anaesthetized with halothane (0.3-0.8%) in a mixture of N20 and 02 (70: 30). In one experiment, the cat also received small intravenous injections of pentobarbitone sodium. Single cells in the A-laminae of the LGN were recorded with glass pipettes filled with 0.2 M NaC1 and 0.05 M 2-amino-2(hydroxymethyl)-l,3-propanediol (tris; pH 7.4, tip diameter 1.25-1.50 gin, impedances of 30--60 Mf~ in vivo). The LGN was approached through a craniotomy at Horsley-Clarke stereotaxic coordinates anterior 6 mm, lateral 15.6 mm, and with the electrode angled 32 deg from the vertical. The entry of the electrode into lamina A was detected by the shift in response properties from those typical of perigeniculate cells to those typical of LGN cells. As the penetrations were approximately along visual projection lines in the LGN, the location of the receptive fields was relatively stationary within each of the A4aminae. The borders between lamina A and A1, and between A1 and C, were detected by the concomitant shift in ocular dominance. After isolation of a single cell, its receptivefield centre was plotted on a tangent screen 86 cm in front of the cat's eyes with the use of stationary and moving light and dark spots, as well as grating stimuli. The quantitative tests were made with a small, stationary flashing light spot. The size of the spot was adjusted so that it approximately filled the receptive-field centre of the cell. The spot presentation was made by a computer-controlled light projector system. The computer also generated peri-stimulustime-histograms (PSTHs) on-line. The position of the receptive-field centres on the tangent screen was checked regularly to control for eye movements. The neuronal activity could also be recorded on a tape recorder. Selected epochs of the signals were digitized off-line at 10 kHz. The lagged cells were identified by their characteristic temporal response properties, and primarily by their long latency to the onset and to half-rise and half-fall of the visual response (Mastronarde 1987a). The luminance of the spot was ~0.9 log units above the background luminance (~ 7 cd/m2). The cells were classified into X (brisk sustained) or Y (brisk transient) as described in Heggelund and Hartveit (1990). The experiment was terminated when the mean blood pressure fell below 90 mm Hg. The animal was then deeply anaesthetized with pentobarbitone sodium (50 mg/kg i.v.) and perfused transcardially with saline and 4% formaldehyde in saline. A block containing the LGN was cut at 25 gm on a freezing microtome and Nissl stained with cresyl violet. Tracks could be retrieved because the electrodes had been advanced for a few mm after finishing the penetration through the A-laminae.
Results The two pairs o f lagged/nonlagged cells were recorded in two different experiments. T h e first pair was recorded in lamina A. The lagged cell o f this pair was an O N - c e n t r e X-cell. Its receptive field was located a b o u t 25 deg f r o m the area centralis a n d the m e a n diameter o f its centre was 0.6 deg. The visual response pattern was typical o f a lagged X-cell (Fig. 1A).
o~
c
250
B
C
0
m
,~
1.0 Time (s)
Fig. IA, B. Peri-stimulus-time histograms that show the response of 2 ON-centre cells to a stationary flashing spot of light. Visual latencies refer to the histograms shown here. A Lagged X-cell. Latency to onset 78 ms, latency to half-rise 118 ms, latency to half-fall 53 ms. B Nonlagged X-cell. Latency to onset 43 ms, latency to half-rise 58 ms, latency to half-fall 33 ms. Each histogram shows the response of the cell during a 500 ms time window when the light was on, marked by the horizontal line under the histogram, and during an equally long time window after the light was turned off. Histograms were compiled from 16 (A) and 20 (B) stimulus presentations, respectively. Binwidth 5 ms
The cell's average latency to onset o f the visual response was 83 ms. Its average half-rise and half-fall latencies were 135 and 55 ms, respectively. W i t h o u t c h a n g i n g the position o f the electrode, the activity o f this cell was recorded in isolation for a b o u t 50 min. D u r i n g this period, there was minimal change in the shape o f the action potentials a n d at no time did the cell change its response pattern or visual latencies to those o f nonlagged cells. T h e n the action potentials o f a second unit a p p e a r e d a n d gradually b e c a m e larger until their amplitude was a b o u t 2 times that o f the first unit. F o r the second unit, slow potentials (S-potentials; Bishop et al. 1962) could be recorded in addition to the action potentials. The m e a n diameter o f its receptive-field centre was 0.6 deg. It was an O N - c e n t r e X-cell and had a visual response pattern typical o f a nonlagged cell (Fig. 1B). The average latency to onset o f the visual response was 45 ms and the average half-rise and half-fall latencies were 55 and 33 ms, respectively. The separation between the two receptive-field centres was 0.5 deg. The amplitudes o f the respective action potentials were sufficiently different to allow separate recording o f a P S T H for one unit in isolation f r o m the other. The two units could be recorded together for a b o u t 10 min. T h e n the action potentials o f the lagged cell suddenly disappeared, a n d at the same time the S-potentials and action potentials o f the nonlagged cell b e c a m e larger. The second pair o f cells was recorded in lamina A1. The n o n l a g g e d cell h a d large positive-negative action potentials a n d S-potentials could be recorded in addition to the action potentials (Figs. 2E, F). It was an O N centre X-cell with receptive field located a b o u t 37 deg f r o m area centralis. T h e m e a n diameter o f its centre was 1.1 deg. The visual response pattern was typical o f a
231
A
200
200
C
~9
c~ 0
....
1.0
1.0
C~
100
100
C
'
t
°
1.0 Time (s)
1.0 Time (s)
50 ms
F
50 ms
Fig. 2. A, B Peri-stimulus-time histograms that show the response of 2 ON-centre cells to a stationary flashing spot of light. Visual latencies refer to the histograms shown here. A Nonlagged X-cell. Latency to onset 48 ms, latency to half-rise 63 ms, latency to half-fall 43 ms. B Lagged X-cell. Latency to onset 78 ms, latency to half-rise 93 ms, latency to half-fall 73 ms. Each histogram shows the response of the cell during a 500 ms time window when the light was on, marked by the horizontal line under the histogram, and during an equally long time window after the light was turned off. Histograms were compiled from 15 (A) and 50 (B) stimulus presentations, respectively. Binwidth 5 ms. C, D Peristimulus-timehistograms that show the responses of the same cells as in A and B to a single light spot that covered part of the receptive-field centre
of each cell. C Nonlagged X-cell. Latency to onset 48 ms, latency to half-rise 63 ms, latency to half-fall 33 ms. D Lagged X-cell. Latency to onset 93 Ins, latency to half-rise 103 ms, latency to half-fall 78 ms. Time window for stimulus presentation 500 ms. Histograms were compiled from 20 (C) and 70 (D) stimulus presentations, respectively. Binwidth 5 ms. E, F Plots of digitized traces. The large action potentials and the S-potentials belong to the nonlagged X-cell and the smaller action potentials belong to the lagged X-cell. Horizontal line below the trace marks the duration of the light-ON period. E The trace corresponds to the histogram in B. Notice that the nonlagged cell responded weakly to the stimulus. F The trace corresponds to the histograms in C, D
n o n l a g g e d cell (Fig. 2A). T h e cell's average latency to onset o f the visual response was 48 ms. Its average latency to half-rise was 63 ms and to half-fall it was 43 ms. T h e activity o f this cell was recorded for a b o u t 25 min u n d e r stable conditions. A t this time, the action potentials o f a second unit b e c a m e gradually larger until their amplitude was a b o u t 1/2 o f the first unit. The second unit had positive-negative action potentials (Figs. 2E, F). It t o o was an O N - c e n t r e X-cell, and it h a d a visual response pattern typical o f a lagged X-ceU (Fig. 2B). T h e m e a n diameter o f the receptive-field centre was 1.t deg. The average latency to onset o f the visual response was 83 ms and to half-rise and half-fall it was 100 a n d 75 ms, respectively. The separation between the two receptive-
field centres was 0.9 deg. Because the receptive-field centres were elongated a p p r o x i m a t e l y perpendicularly to a line connecting them, the degree o f overlap was relatively small. P S T H s for one unit could be compiled in isolation f r o m the other. F r o m plots o f digitized traces o f the n e u r o n a l activity, the S-potentials and action potentials o f the weakly responding nonlagged cell could be seen together with the action potentials o f the strongly res p o n d i n g lagged cell when the latter was appropriately stimulated in its receptive-field centre (Fig. 2E). T o ascertain that the cells could r e s p o n d with patterns typical o f lagged and n o n l a g g e d cells simultaneously, the size a n d location o f the light spot was adjusted so that it covered a p p r o x i m a t e l y h a l f o f b o t h receptive-field cen-
232 tres. Figures 2C, D show that even when the two cells received this suboptimal visual stimulus the visual response patterns and latencies were very similar to those in the separate control recordings. Figure 2F shows a digitized trace with both cells responding to the suboptimal stimulus. The amplitudes of the respective action potentials remained sufficiently different to allow compilation of PSTHs for about 1 1/2 h. During this time neither cell changed its response pattern to that of the other class. However, the mean firing rate could vary spontaneously for both cells. The amplitude o f the action potentials o f the lagged cell gradually increased, and after a while, the two cells could not be separated based on the amplitude of the action potentials. By moving the microelectrode 20 ~tm further down into lamina A1, the potentials of the nonlagged cell disappeared and the action potentials of the lagged cell remained in isolation. PSTHs compiled at this time were virtually identical to those compiled when the nonlagged cell was still present.
Discussion There is strong physiological (Mastronarde 1987a, b), morphological (Humphrey and Weller 1988b) and pharmacological (Hartveit and Heggelund 1990; Heggelund and Hartveit 1990) evidence for treating lagged and nonlagged cells as separate cell classes. The present finding that lagged and nonlagged cells can be recorded simultaneously supports this hypothesis. Uhlrich et al. (1990) recently examined the influence of electrical PBR stimulation on the visual response of lagged and nonlagged cells. They reported that PBR stimulation could reduce the latency to half-rise below 70 ms for 6 out of 8 lagged cells recorded. It was claimed that this had converted the visual responses o f lagged cells to those of nonlagged cells, and the authors put forward the hypothesis that the response properties characteristic of lagged and nonlagged cells may represent state-dependent response modes of the same L G N cells. One version o f this hypothesis might be that the cells in the L G N appear as lagged in one arousal state and as nonlagged in another state. Accordingly, it should not be possible to record lagged and nonlagged cells simultaneously in the L G N . The present result shows that this is indeed possible. This result is consistent with the experience that lagged and nonlagged cells can be recorded in succession in a single electrode penetration (Mastronarde 1987a; Heggelund and Hartveit 1990). An alternative version o f the hypothesis might be that different L G N cells are subject to different degrees o f influence of the modulatory input, such that they would switch between lagged and nonlagged response patterns at different levels of modulatory input. This interpretation cannot be rejected by the present results. However, there are no reports that a cell can change from a nonlagged to a lagged response pattern. In fact, the available evidence indicates that during reduced modulatory input the response pattern o f nonlagged cells becomes more transient (e.g. Sawai et al. 1988) and accordingly less
similar to that of lagged cells. Furthermore, some of the previously reported differences between lagged and nonlagged cells cannot be state-dependent. In particular, lagged and nonlagged cells have almost non-overlapping antidromic latencies to electrical stimulation o f the visual cortex (Mastronarde 1987a; Humphrey and Welter 1988a) and almost non-overlapping morphologies (Humphrey and Weller 1988b).
Acknowledgements. I thank Paul Heggelund for continuous support and encouragement and for valuable comments on the manuscript, Allen L. Humphrey for constructive criticism of an earlier version of the manuscript, Ingeborg Spinnangr for histological and technical assistance, Torsten Eken for help with digitizing and plotting of neuronal signals and Arve Stavo for writing the on-line computer programs. Gallamine triethiodide (Flaxedil) was kindly provided by Cyanamid Nordiska and halothane (Fluothane) by ICI-Pharma. The research was financially supported by The Norwegian Research Council for Science and The Humanities (NAVF). References Bishop PO, Burke W, Davis R (1962) The interpretation of the extracellular response of single lateral geniculate cells. J Physiol (Lond) 162:451-472 Fitzpatrick D, Diamond IT, Raczkowski D (1989) Cholinergic and monoaminergic innervation of the cat's thalamus: comparison of the lateral geniculate nucleus with other principal sensory nuclei. J Comp Neurol 288:647-675 Francesconi W, Mfiller CM, Singer W (1988) Cholinergic mechanisms in the reticular control of transmission in the cat lateral geniculate nucleus. J Neurophysiol 59:1690-1718 Hartveit E, Heggelund P (1990) Neurotransmitter receptors mediating excitatory input to cells in the cat lateral geniculate nucleus. II. Non-lagged cells. J Neurophysiol 63:1361-1372 Heggelund P, Hartveit E (1990) Neurotransmitter receptors mediating excitatory input to cells in the cat lateral geniculate nucleus. I. Lagged cells. J Neurophysiol 63:1347-1360 Humphrey AL, Weller RE (1988a) Functionally distinct groups of X-cells in the lateral geniculate nucleus of the cat. J Comp Neurol 268, 429447 Humphrey AL, Weller RE (1988b) Structural correlates of functionally distinct groups of X-cells in the lateral geniculate nucleus of the cat. J Comp Neurol 268: 448M68 Mastronarde DN (1987a) Two classes of single-input X-cells in cat lateral geniculate nucleus. I. Receptive field properties and classification of cells. J Neurophysiol 57:357-380 Mastronarde DN (t987b) Two classes of single-input X-cells in cat lateral geniculate nucleus. IL Retinal inputs and the generation of receptive field properties. J Neurophysiol 57: 381-4 13 Mastronarde DN, Humphrey AL, Saul AB (199t) Lagged Y cells in the cat lateral geniculate nucleus. Vis Neurosci 7:191-200 McCormick DA, Pape H-C (1988) Acetylcholine inhibits identified interneurons in the cat lateral geniculate nucleus. Nature 334:246--248 McCormick DA, Prince DA (1987) Actions of acetylcholine in the guinea-pig and cat medial and lateral geniculate nuclei, in vitro. J Physiol (Lond) 392:147-165 Saul AB, Humphrey AL (1990) Spatial and temporal properties of lagged and nonlagged cells in cat lateral geniculate nucleus. J Neurophysiol 64:206-224 Sawai H, Morigiwa K, Fukuda Y (1988) Effects of EEG synchronization on visual responses of the cat's geniculate relay cells: a comparison among Y, X and W cells. Brain Res 455: 394-400 Steriade M, McCarley RW (1990) Brainstem control of wakefulness and sleep. Plenum, New York London Uhlrich D J, Tamamaki N, Sherman SM (1990) Brainstem control of response modes in neurons of the cat's lateral geniculate nucleus. Proc Natl Acad Sci USA 87:2560-2563
